M. Reagin, A. Merla, D. Meyers; Amersham Biosciences, 800 Centennial Ave, Piscataway, NJ 08855

Preparing plasmid templates for DNA sequencing is the most time consuming step in the sequencing process. Current template preparation methods rely on a labor intensive, multi-step procedure that takes up to 24 hours to produce DNA of varying quality and quantity. The TempliPhi DNA Sequencing Template Amplification Kit has been developed that eliminates the requirement for extended bacterial growth prior to sequencing and saves you hands-on time by eliminating the centrifugation and transfer steps currently required by older preparatory methods. In addition, costly purification filters or columns are not necessary as amplified product can be added directly to a sequencing reaction. Starting material can be any circular template from a colony, culture, glycerol stock, or plaque. Based on rolling circle amplification (RCA) and employing bacteriophage Phi29 DNA polymerase, the process can produce 2-3 ug of template directly from a single bacterial colony in 4 hours. We will demonstrate how the implementation of these procedures in a laboratory or core sequencing facility can decrease costs, save on tips, plates, and other plasticware, while at the same time increasing throughput.

C. Jecmen, J. T. Medalle, M. Randesi, B. S. Imai; The Rockefeller University, 1230 York Ave., New York, NY 10021

Core facilities constantly search for products that enhance and improve their services to the research community. The bottom-line for acquiring these products relies on the ability of the facility to work within their budget. Some facilities are fortunate to receive funding from a university; but those without such funding or budget conscious facilities must find ways to stretch their dollar. As a cost cutting measure many labs choose between store-bought and homemade reagents. Sequencing facilities, from low to high throughput, prepare reagents such as 10X TBE or 5X BigDye buffer for diluting the enzyme. Fluorescent dye terminator kits for automated DNA sequencing, such as BigDye Version 3.0 offered by Applied Biosystems, are vital reagents used by sequencing facilities. Diluting BigDye with 5X BigDye buffer one-to-one can defray the costs that arise from the use of BigDye chemistries; a 5000 reaction kit can be optimized to a 10,000 reaction kit. This cost-effective strategy will be discussed through various dilution examples of ABI Big Dye V3.0 on different types of samples handled by core facilities.

H. G. Hills, M. Blomberg-Wirschell, G. B. Witman, B. Rajeshkumar; University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655

Samples containing high "G-C content" often present problems when using ABI BigDye chemistry. The ABI dGTP generally sequences these areas with little or no difficulty but often results in less than optimum data due to the number of compressions. Using the BigDye and dGTP chemistries together often produce better results even though ABI does not support mixing the two chemistries. Using 3 or 4 parts of BigDye with one part of dGTP mix will often produce very good results. The combination chemistry was used to sequence a 240 base micro satellite repeat region from Chlamydomonas reinhardtii. The Amersham Dynamic ET Terminator Kit sometimes sequences high "G-C" templates better than the BigDyes but not better than the BigDye and dGTP combination. Areas of high "AT" content are often sequenced better with the Amersham chemistry. The Amersham chemistry also generally reads through a polyT area with fewer errors after the polyT. The Amersham chemistry can be used with the "E" filter set, although some loss of resolution may occur. This has the advantage that both chemistries can be run on the same gel.

R. Pershad, A. Scarpa; University of Texas. M.D.Anderson Cancer Center, 1515 Holcombe Blvd, Houston, Texas 77030

One of the goals of the investigators at the Cancer Center is to be able to rapidly screen patient tumor samples and families with a history of cancer for mutations.

In an effort to eliminate the need for individual laboratories to perform labor intensive SSCP (single stranded conformation polymorphism) screening and MDE (Mutation Detection Enhancement) gel analysis, the use of the WAVE instrument has been implemented in the DNA Core facility to provide a centralized service for mutation screening.

In this Core the WAVE has been used successfully to identify deletions, insertions and heterozygote mutations associated with familial cancers. The facility has screened over 22 different exons for 12 principal investigators. All the variants identified by WAVE analysis have been confirmed by sequencing. Here we will present one scenario of how to operate the WAVE as a core instrument.

L. Wen; San Diego State University, 5500 Campanile Drive, San Diego, CA 92182

I recently demonstrated that combining the annealing step with the extension step at one temperature (60^oC) reduces premature terminations in DNA sequences that regularly contain premature terminations when the three temperature steps are used (Molecular Biotechnology 17, pp. 135-142 (2001)). I have studied a novel class of 100 base pairs in length DNA sequences upstream from the point that cause premature terminations. Thermodynamics of 49 DNA sequences with premature terminations at three temperature steps were determined from DNA mfold profiles. Sequencing results of the 28 samples were improved by the two step cycle sequencing and the 21 sequences were not improved or severed by the two step cycle sequencing reactions. Nearest neighbor thermodynamic parameters for all 49 sequences were compared at temperature 50^oC and 60^oC. The parameters predict thermodynamic free base (external loop) energies DD G^o) are significantly different between these two groups. DNA templates that have improved the premature terminations by the two-step cycle sequencing method, the values for DDG^o60 for the external loop are higher than DDG^o50. In contrast, DNA sequences that have not improved the premature terminations by the two-step cycle protocol, the values for DDG^o60 for the external loop are no longer higher than DDG^o50. The results of the investigation are reported here.

G. J. Wiebe, E. Tanaka, D. N. Drechsel; Max Planck Institute of Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, 01307 Dresden, Germany

We have been evaluating the production of high-quality sequencing template using the TempliPhi^TM kit from Amersham Pharmacia Biotech. This kit utilizes rolling circle amplification to replicate plasmid DNA directly from bacterial colonies or cultures with minimal handling. The protocol consists of a few simple steps performed within a single tube: a small amount of bacterial culture is denatured, incubated with enzyme overnight, and finally heat-inactivated. A small volume may then be added directly to a sequencing reaction without further purification.

Successful results have been obtained from high and low copy number Amp^R plasmids cultured in different types of growth media; Kan^R plasmids still require further investigation. A 96-well format has been maintained throughout, in anticipation of future high-throughput applications. This method may offer an alternative to costly robotics based template preparation.

S. P. Yadav¹, T. Haqqi², X. Zhao², A. Panciu²; ¹The Lerner Research Institute, 9500 Euclid Avenue, Cleveland, OH 44195, ²The Lerner Res. Institute, Cleveland, OH

The presence of several copies of the same class of repetitive element in DNA templates increases the probability of ambiguous base calling caused by band compression artifacts in dye terminators sequencing method. The presence of di-, tri- and tetranucleotide repeats and short tandem repeat sequences all over in the genome pose a daunting routine sequencing roadblocks in the Core laboratories where samples originate from a variety of DNA templates. These compressions are due to the formation of secondary structures, such as hairpins and intramolecular base pairing between guanine and cytosine bases, on the nascent strand resulting in to anomalies in the migration of certain DNA fragments.

To overcome these undesired sequencing artifacts and formation of secondary structure several modifications of the sequencing reactions parameters have been applied in the past, including replacement of dGTP nucleotide by base analogs dITP, or 7-deaza-DGTP in the reaction mix, and adding denaturants in the reaction mix. We have earlier reported that the use of 1M betaine in the reaction mix is advantageous to sequence through AACCCC-type repeats found in telomeres. In this study we report that adding 1M betaine in the reaction mix also results in reduction of band compressions in other repeat elements and thus provides a higher accuracy of sequencing.

D. Adam, F. Dobie, H. McDonald; University of British Columbia, Rm.237 - 6174 University Blvd., Vancouver, B.C. V6T 1Z3 Canada

Commercial suppliers frequently develop new and improved DNA Sequencing premix, attempting to provide a more robust chemistry suitable for core facilities which encounter a wide range of samples. Templates which display characteristics such as high GC or AT content, homopolymer regions, G-stops and profound secondary structure have proven very difficult to sequence with manufacturers' recommended protocols. We present the data obtained when evaluating the performance of Applied Biosystems' Big Dye Chemistries (BDv1.0, BDv2.0, BDv2.0 dGTP and BDv3.0) and Amersham's DYEnamic ET Terminators, with and without facilitators such as DMSO and Betaine, on a wide variety of templates. The findings of our study are presented here.

F. H. Wolven, B. Etzel, A. M. Segall, L. Wen; San Diego State University, 5500 Campanile Drive CSL326, San Diego, CA 92182-1030

A large increase of samples from plasmid isolation and sequencing projects, cultured in 96 well plates, has created significant challenges in achieving high throughput DNA sequencing for a small core facility with staff of two. The experiences we learned were: (1) replacing the ABI377 with an ABI 3100 CE sequencer increased throughput while eliminating staggered workshifts, (2) it is difficult to isolate enough plasmid template grown in 2 mL deep 96 well blocks using standard protocols; to obtain sufficient yield requires chloramphenicol amplification and 2X LB, (3) when transitioned from 377 to the 3100, it was discovered that a second wash of 70% ethanol, with subsequent removal of residual ethanol, was needed to produce good data on the 3100, and (4) clones containing bacteriophage inserts were difficult to grow and would not sequence with our standard protocols due to low copy number. The thermocycle program was modified, and data collection injection time increased from 30 sec to 90 sec. to achieve optimal signal strength on the 3100. Details of our protocols, descriptions of the problems we encountered along the way, and their solutions will be presented.

S. Chen, G. Amparo, E. Kuo, K. Hacker, B. Johnson, S. Spurgeon; Applied Biosystems, 850 Lincoln Centre Dr., Foster City, CA 94404

The new ABI PRISM® dGTP BigDye™ Terminator v3.0 Ready Reaction Cycle Sequencing kit with AmpliTaq® DNA Polymerase, FS was introduced earlier this year with the BigDye Terminator v3.0 Cycle Sequencing kits. The dGTP BigDye Terminator v3.0 sequencing kit, like the dGTP BigDye Terminator Cycle Sequencing kit, was introduced as a special chemistry solution to overcome certain problems with difficult templates. These are templates which when sequenced with the BigDye Terminator v3.0 sequencing kit or BigDye Terminator sequencing kit fail to give good read length due to certain sequence motifs that result in stops in data.. Since the introduction of the dGTP BigDye Terminator v3.0 Cycle Sequencing kit we have further studied this chemistry and made improvements to protocols for use with this chemistry. In this presentation we will discuss our study and findings.

I. Y. Hlede, S. Wegener, R. T. Pon; University of Calgary, 3330 Hospital Drive NW, Calgary, Alberta T2N 4N1 Canada

Our primary goals as a university core service laboratory are to provide accurate (99.9%) results, instruction in DNA sequencing analysis and sample preparation, and troubleshooting of results. We have optimized throughput, data quality, and communications through a number of changes. These affect the way samples are processed and how results are communicated to clients. These changes include: software and sequencer upgrades, use of Sephadex spin columns, switching from TBE to TTE (TRIS, TAPS, EDTA) buffer, sample (sequencing and genescan) loading on porous combs, template quantitation on a UV plate reader, synthesis of home-made lane guide, offering results in 3 data formats on a FTP/Fetch server, Apple Script short cuts, and email responses with web-references to sequencing customers.

M. Vazquez; City College of New York, 140th and Convent. Ave.; T-250, New York, NY 10032

The flexibility of the microfabricated format creates unique opportunities for study of the electrophoretic process. This understanding is essential to the design of all electrophoresis apparati including electrophoretic microdevices. The present work utilizes digital images to capture the motion of DNA samples during loading and injection. A systematic study of strong sample stacking (sample concentration effects) was performed in order to analyze realistic DNA analysis conditions within microdevices. These effects are shown to be highly useful and can be separately enhanced using high-voltage injection. The sample profiles within the injector have been analyzed by deconvolving the geometrical sample profile into different velocity groups. This analysis illustrates the evolution of molecular separation into distinct migrating populations within the injector itself. The present study performed DNA injections within microfabricated devices imposing run voltages between 85 V/cm - 850 V/cm. Data from 3 different offset lengths of a double-T cross-injector, 10 different applied voltages, and 2 different sample preparation protocols are presented.

C. Potter¹, M. K. Johansson², D. Dick², R. Cook²; ¹Biosearch Technologies, 81 Digital Dr., Novato, CA 94949-5750, ²Biosearch Technologies, Inc., Novato, CA

Passive reference (PR) dyes are frequently used in real-time PCR to normalize fluorescence of reporter dyes & correct for fluctuations in fluorescence that are non-PCR based. An ideal PR dye maintains steady & constant fluorescence, does not affect the PCR, & has a fluorescent signal that can be distinguished from that of any reporter & quencher dyes used in the reaction. PR dyes that emit around 600nm, such as the commonly used ROX, can be distinguished from reporter dyes in real-time PCR, because they are further red than the reporters most often used. Because these dyes are not efficiently excited by a 488nm excitation source, this can be problematic due to the fact that several of the most popular real-time PCR instruments are limited to this single excitation source. To generate sufficient signal, high concentrations of PR dye must be added, which may lead to signal instability, such as the 'ROX drop.' This study evaluates a variety of PR candidates, including a novel dye with a tremendous Stokes shift, which can be efficiently excited at 488nm & has an emission maximum of 620nm. All data is gathered from the 7700 to assess all the PR candidates, based on stability of signal & reproducibility.

L. M. Overbergh, A. Giulietti, D. Valckx, R. Bouillon, C. Mathieu; KULeuven/UZ-Gasthuisberg, Herestraat 49, O&N8, Leuven, 3000 Belgium

The analysis of cytokine profiles helps to clarify functional properties of immune cells. Real-time RT-PCR is becoming a widely used method to quantify cytokines from cells, body fluids or tissues. The method allows for the direct detection of PCR product during the exponential phase of the reaction, combining amplification and detection in one single step.

Using the TaqMan chemistry and the 7700 SDS, we validated a very large panel of murine and human cytokines. Although the method allows a fast, sensitive and accurate quantification, different control assays are necessary for the method to be reliable. By construction of cDNA plasmid clones, standard curves are generated which allow direct quantification of every unknown sample. Furthermore, the choice of a reliable houskeeping gene is very important. Finally, co-amplification of contaminating genomic DNA is avoided by designing primers located in different exons, or on intron/exon junctions.

In summary, the real-time RT-PCR technique is very accurate, sensitive, allows a high throughput and can be performed on very small samples; therefore it is the method of choice for quantification of cytokine profiles in immune cells or inflamed tissues.

I. V. Foulds¹, A. Kapoor¹, A. Castle², C. Gubala¹, U. J. Krull¹, P. A. Horgen¹; ¹University of Toronto, 3359 Mississauga Rd. N., Mississauga, Ontario L5L 1C6 Canada, ²Brock University, St. Catherines, Ontario Canada

The presence of waterborne pathogens in drinking (lakes/reservoirs as well as bottled water) and recreational water is traditionally quantified by time consuming culturing techniques or by chemical characterization. We have developed a more rapid, real-time PCR-based detection method for Escherichia coli, and toxic microcystin-producing Microcystis aeruginosa. Dual-labeled TaqMan probes have been designed for regions of the lacZ and microcystin synthetase genes, respectively. We have transferred our technology to the analysis of field collected water samples and commercially sold bottled water, and we are able to detect down to 3 gene copies per sample. Detection of pathogens can be done within about 2 hr after DNA extraction and sample setup, rather than up to several days needed for traditional detection methods.

W. C. Kusser, B. Lowe, M. Gleeson; Invitrogen, 1600 Faraday Ave., Carlsbad, CA 92008

A real time PCR method based on a novel design of fluorescent primers is presented. The reporting system is based on a single fluorescent dye molecule attached close to the 3’end. A primer design software was developed that identifies a primer with a suitable labeling site in a target of interest, based on a set of rules for optimum signal development. Subsequently it attaches a short tail (3-5 nucleotides) to the 5’end of the primer that is complementary to the 3’end. The resulting hairpin oligonucleotide and the sequence context around the fluorophore result in an initially low background fluorescence. Typically, JOE and FAM have been used as fluorescent dyes coupled to the primer. Upon incorporation of the primer into an amplicon during real time PCR, the fluorescence of the dye molecule increases several fold and provides an efficient reporting system compatible with most real time PCR instruments. We present results on the successful use of this system (termed Lux for "Light Upon extension") on a series of genes including house keeping genes, cytokines, signal transduction, cancer and apoptosis genes.

M. Brisson, L. Huiet, R. Park, L. Tan, K. Hamby; Bio-Rad Laboratories, Hercules, CA

Proper experimental design for real-time PCR is essential for studies involving relative or absolute comparisons between two or more genes amplified in the same tube. Several parameters must be evaluated and optimized independently to achieve accurate quantitation. Low amplification efficiency can reduce the dynamic range of quantitation for a particular target and compromise the data. For gene expression studies that require precise and accurate quantitation of multiple genes in the same samples, the iCycler iQ displays powerful multiplexing capabilities. Here we demonstrate 4-gene multiplexing capabilities. Calculated threshold values for a single gene target were identical when amplifying alone or with 3 other genes under optimized conditions. This was true for plasmid or genomic DNA targets. In multiplex, individual targets can be reliably quantitated over at least 6 orders of magnitude for a plasmid or 4 orders for a genomic DNA target. In addition, wide target concentration differences between 2 amplified gene products could be shown under optimized conditions. We quantitated as few as 10²copies of one gene in the presence of up to 10⁹copies of the second gene target.

R. T. Pon, S. Yu; University of Calgary, 3350 Hospital Drive NW, Calgary, AB T2N 4N1 Canada

New reagents with cleavable 3'-ester linkages use phosphoramidite coupling chemistry to attach the 3'-terminal nucleoside to a solid-phase support. Inexpensive amino supports may then be used as "Universal" supports in high-throughput DNA synthesizers. The linker phosphoramidites also allow multiple oligonucleotides, linked end-to-end in tandem, to be made. Tandem synthesis allows two or more oligos to be produced in a single run. Thus, a pair of PCR primers or both strands of a duplex DNA can be made in one synthesis. This doubles the number of oligonucleotides which can be produced per run and halves the number of individual samples processed. The tandem products can be used as a single mixture without separation, so end users also have fewer samples to handle. Both individual and tandem syntheses can be performed simultaneously and no changes to conventional synthesis or deprotection conditions are required.

A. T. Yeung¹, B. P. Holloway², P. S. Adams³, K. Mills⁴, S. Scaringe⁵, G. A. Buck⁶, K. Lilley⁷, M. Gunthorpe⁸; ¹Fox Chase Cancer Center, ²Centers for Disease Control, ³Trudeau Institute, ⁴Millennium Pharmaceuticals Inc., ⁵Dharmacon Research, Inc., ⁶Virginia Commonwealth University, ⁷University of Cambridge, ⁸University of San Francisco

The high cost of DNA probes for Fluorescence Resonance Energy Transfer (FRET) experiments has long impeded the full utilization of real-time PCR in genomic research. The high cost is partly due to limitations in the traditional synthesis chemistry that demanded extensive purification of these probes. Limited access to proprietary quencher reagents has also discouraged the development of better synthesis protocols. Fortunately, an opportunity has arisen with the recent availability of non-fluorescent quencher reagents to core DNA synthesis facilities. We are testing the synthesis of real-time PCR probes labeled with a fluorescence reporter on the 5' termini and with either a non-fluorescent quencher or tetramethylrhodamine on the 3' end. Both crude products of DNA synthesis and the purified probes will be tested in real-time PCR to judge the effect of probe quality on assay outcome. If improved synthesis protocols are found, we propose to verify their utility by expanding the testing to all interested laboratories.

M. Lemaitre¹, J. Truffert¹, N. Thelwell², L. Brown²; ¹EUROGENTEC SA, Herstal, Belgium Belgium, ²OSWEL Ltd, Southampton, United Kingdom

We have developed a number of novel analogues for use in real-time PCR systems. Development of these analogues was undertaken to improve fluorescent probing systems and to resolve problems that have arisen with the standard probing systems, such as Scorpion primers and TaqMan probes. The oligonucleotide modifications discussed will include dark quenchers that can be used in either of the 2 systems to quench a wide range of dyes across the spectrum. Data comparing quenchers will be shown for TaqMan probes and Scorpion primers. In addition, a system that incorporates FRET into a Scorpion primer for detection in channel 2 of the LightCycler will be described. This system works in both the standard stem-loop Scorpion format and the new Duplex Scorpion, which will be described in detail.

1) Whitcombe,D., Theaker,J., Guy,S.P., Brown,T. and Little,S., Nature Biotechnol., 1999, 17, 804.

2) Thelwell,N., Millington,S., Solinas,A., Booth,J. and Brown,T., Nucleic Acids Research, 2000, 28, 3752.

F. Schäfer, K. Steinert, C. Feckler, J. Drees, J. Ribbe; QIAGEN GmbH, Max-Volmer-Str. 4, Hilden, 40724 Germany

The TAGZyme System allows efficient and precise exoproteolytic removal of N-terminal affinity tags from proteins. In combination with Ni-NTA technology, the TAGZyme System provides high-purity proteins free of vector-encoded amino acids for use in applications such as preparation of pharmceuticals that demand recombinant reagents, an absence of non-specific cleavage, and a complete removal of all impurities from the target protein preparation.

New vectors, based on the pQE series for highly efficient expression of 6xHis-tagged proteins, have been constructed. The vectors’ multiple cloning site sequences have been designed to allow complete exoproteolytic removal of the encoded N-terminal affinity tag regardless of the restriction site used for cloning. Numerous examples of obtained expression rates and the successful downstream processing of the Ni-NTA–purified proteins are shown.

The extremely efficient cleavage reaction and removal of the His-tagged TAGZyme enzymes by subtractive Ni-NTA chromatography is demonstrated for 6xHis-Il-1β and 6xHis-TNFα. In both cases, more than 99.8% of TAGZyme proteolytic activity was separated from recovered, detagged proteins. N-terminal analyses by Edman degradation revealed the predicted sequences of the native proteins and indicated a purity >99%. Furthermore, data on the successful and fast removal of tags of between 8 and 25 amino acids in length are shown. Finally, the processing of high molecular weight protein complexes and their recovery by subtractive Ni-NTA chromatography is presented.

U. Schweiger-Hufnagel, M. Lubeck, C. Baessmann, A. Ingendoh; Bruker Daltonik GmbH, Fahrenheitstrasse 4, Bremen, 28359 Germany

Expression of recombinant proteins in biochemical laboratories can be complicated with certain degrees of primary sequence errors. Such deviations could result in modified enzymatic activities or more profoundly in modified 3-D structure.

LC-MS/MS provides ultimate sensitivity and specificity. Capabilities of full automation and exact analysis of the primary sequence can be utilized to provide a fast and reliable approach to confirm successful expression. We have developed a system that yields high sequence coverage and a sophisticated software package that would allow matching of signals from an LC-MS/MS run to a known sequence. Here, an esquire3000plus ion trap was used in combination with nano-LC for confirmation of the expected amino acid sequence of recombinant proteins.

Isolated and purified proteins were digested and separated on a nanobore column. Eluted peptides were analyzed by automated nano-ESI-MS/MS. Pepmap Creator, was used to generate a combined MS and MS/MS peak list, which was then automatically compared with the expected protein sequence in Biotools. We show here that this intuitive software is a powerful tool for detecting synthesis errors during protein expression.

A. P. Wulandari; THe University of Japan, Yayoi 1-1-1, Bunkyo-ku, Tokyo, Tokyo 153-8657 Japan

The lysS gene encoding for homocitrate synthase , the key enzyme involved in the lysine biosynthesis of T. thermophilus was over expressed in E.coli. SDS-PAGE shows a single protein specific of Mr 43kD. The multimeric structure of the native enzyme were confirmed by cross-linking with glutaraldehyde followed by electrophoresis. The enzyme follows Michelis-menten kinetic to 2-oxoglutarate (Km= 0.031 mM) and exhibit allosteric condition with acetyl Co-A (nH= 2.6 and S[0.5]=0.032mM.

Thermodymic of denaturation enzyme exhibited that the reaction catalyzed by HCS from T. thermophilus is temperature dependence reaction. The enzyme strongly inhibited by lysine with Ki =0.0094mM, and showed the additive effect in the present of arginine

Withdrawn

X. Zhang, X. Xu; Zhengzhou University Medical College, 40 Daxue Road, ZhengZhou, Henan 450052 China, People's Republic of

Xiaoli Zhang Xia Xu Lexun Xue *

E-mail: lxxue@public2.zz.ha.cn

Tel/Fax: 86-371-699-9548

P. S. Adams¹, T. B. Miller¹, D. Vignali², K. Vignali², L. S. Cauley¹, L. Haynes¹, P. Arnold², S. L. Swain¹, D. Woodland¹; ¹Trudeau Institute, ²St. Jude Children's Research Hospital

MHC class II molecules play a critical role in the presentation of peptide antigens on the cell surface for recognition by CD4+ T cells. However, it has proven difficult to reliably express soluble forms of MHC class II molecules to study their role in infectious disease. To resolve this problem, we have generated fusion proteins in which the extracellular MHC class II a and b chains are linked via a fos/jun leucine zipper, and attached to the hinge and Fc domain of murine IgG2a. These were expressed as either ‘empty’ dimers or loaded with specific peptides recombinantly attached to the MHC class II b chain. We have generated 22 different MHC class II:Ig multimers using four alleles (H-2A^k, H-2A^b, H-2A^g7, H-2E^k). All are expressed as functional dimers by Drosophila S2 cells and can be purified by conventional protein A chromatography. Multimerization of peptide loaded MHC class II:Ig multimers with fluorescently labeled protein A offers a useful tool for tracking CD4+ T cells by flow cytometry. A preparation scheme and examples of experimental use of multimers representing several alleles and peptide antigens will be presented.

J. VanEe, A. Mannochia, T. Thannhauser, T. Stelick, R. Halpin; Cornell University

Large scale, multi-user DNA sequencing facilities face special challenges. Acceptable levels of customer service and reasonable prices must be maintained: Only by increasing operational efficiency can a facility meet these challenges. Cornell University's BioResource Center DNA Sequencing Facility has faced sustained exponential growth, leading to the development of several generations of data management systems. The latest, the "BRC Online" web portal, provides the ability to manage every step of the sequencing process. MySQL provides the database backbone of the system. PHP (PHP Hypertext Processor) and an Apache Web Server provide the interface for the portal. Standard networking technologies coupled with server based scripting integrate various instrumentation with the central database. Special functions include automatic robotics work list generation, creation of .plt files for ABI 3700 sequencers and generation of several quality control metrics for efficient review and interpretation of outgoing data. This system provides managers with an overview of the operation's status and an audit trail for troubleshooting.

R. Ehrnström, A. Palm, S. Wallenborg, M. Gustafsson, P. Andersson, A. Hedström, E. Togan-Tekan, G. Ekstrand; Gyros AB, Uppsala Science Park, Uppsala, SE-75183 Sweden

The sensitivity of MALDI mass spectrometry has increased immensely since its introduction and the technique is widely used for rapid protein identification. The need for improvements in sample preparation methods of protein digests prior to identification by mass spectrometry is well recognized. Until now an automated process for reliable, routine preparation of peptide digests at concentrations approaching the detection limit for mass spectrometry, has been missing. We have developed a system in which key steps for sample preparation prior to MALDI analysis are integrated and run in a CD microlaboratory. Protein digests are concentrated, desalted, mixed with matrix and crystallized on MALDI target areas on the CD. Up to 96 samples are processed in parallel. The CD is transferred to a MALDI instrument and no separate target plate is required. By optimizing the entire process we have significantly enhanced sample recovery and concentration so that analysis is performed close to the detection limit of commercial MALDI mass spectrometers. Results with sensitivities in the attomole to low femtomole can be achieved, facilitating the identification of low abundance proteins.

I. P. Smirnov, P. L. Ross, L. R. Hall, L. A. Haff; Applied Biosystems, 500 Old Connecticut Path., Framingham, MA 01701

In the past decade MALDI-TOF MS has emerged as one of the main analytical tools for proteins, peptides and short DNA fragments. The main obstacle of this method is its susceptibility to salts and surfactants, usually present in the sample as buffer components. Their elimination requires laborious desalting procedures and substantially decelerates overall analysis time. Here we report a novel solution to this problem by employing MALDI sample plates coated with various ion-exchange polymers, such as polyacrylic acid and polyethyleneimine. By using their ability to bind biomolecules sample isolation and purification can be done directly on the MALDI plates, thus significantly simplifying the sample preparation step. Various polymer films have been characterized for peptides, proteins and oligonucleotides of various size. The importance and role of different factors have been studied, such as composition of ion exchange film and compatibility of the MALDI matrices together with mass accuracy optimization in order to obtain the best performance. This method makes possible a high throughput sample preparation for MALDI analysis in proteome research.

M. E. McComb, A. Lim, T. B. Prokaeva, L. H. Connors, M. Skinner, C. E. Costello; Boston University School of Medicine, 715 Albany Street, Boston, MA 02118

Amino acid substitution in TTR, a transport protein in plasma, may result in the formation of amyloid fibrils. Familial transthyretin amyloidosis (ATTR) is associated with deposition of TTR amyloid fibrils in tissues and organs. More than 80 variants are known. Effective treatment requires liver transplantation, thus correct diagnosis is critical. We have characterized TTR variants using mass spectrometry (MS). Intact TTR and peptide mapping suggests the variant. For conclusive identification it is often necessary to sequence the individual peptides. To this end, we have explored the use of tandem MS (MSMS) on an ABI QSTAR quadrupole-orthogonal-time-of-flight (QoTOF) MS equipped with MALDI and ESI sources. Using MALDI-MS, ESI-MS and on-line capillary LC-ESI-MS with information dependent acquisition (IDA) we have successfully characterized several TTR variants by MSMS. Whereas manual MALDI-MSMS and ESI-MSMS may be tedious, LC-ESI-IDA-MS/MSMS affords the ability to purify, separate, obtain peptide maps and sequence information on an entire proteolytic digest in one automated experiment. We demonstrate this while characterizing amyloidogenic TTR variants.

R. Tyldesley¹, D. O. Gostick¹, J. I. Langridge², J. B. Hoyes³, K. Schneider⁴, A. O. Wattenberg⁴; ¹Micromass UK Ltd, Floats Road, Wythenshawe, Manchester M23 9LZ United Kingdom, ²Micromass UK Ltd., Floats Road, Wythenshawe, Manchester M23 9LZ United Kingdom, ³MICROMASS UK LTD, Almere, ⁴GSK, Harlow, Essex United Kingdom

Protein identification by mass spectrometer is now a routine in many biochemistry laboratories. There are several alternative methods for producing an identification of an unknown protein, but these nearly always require the mass measurement of peptides. The analysis of the peptides is typically performed by either ESI or MALDI. These two ionisation techniques have until now always been performed on different types of mass spectrometers. In this presentation we compare the MS and MS/MS spectra from the analysis of protein digests using ESI and MALDI on the same instrument. This is now possible using the dual source available on the Q-Tof mass spectrometer.

The two ionisation techniques are known to produce different MS and MS/MS spectra from the analysis of the same sample. In this poster we investigate these differences and illustrate their complimentary nature for protein identification using database searching and bioinformatics.

B. Wu¹, P. Bondarenko¹, A. Land¹, V. Kovtoun¹, G. Stafford¹, S. Swedberg², K. Miller¹, P. Shieh¹, B. Hancock¹; ¹ThermoFinnigan Corp., 355 River Oaks Parkway, San Jose, CA, ²ThermoFinnigan, San Jose, CA

Protein identification using peptide mass maps and database searching has become an important technique for proteomic studies. MALDI-TOF MS has been used extensively to identify samples extracted from gel spots by this approach. Also proteins purified from multi-dimensional separation techniques have been characterized by off-line coupling with MALDI-TOF MS. However, MALDI-TOF MS has the disadvantage that MS/MS is better performed on tandem mass spectrometers. The power to fragment selected ions further can be key to assigning the position of modifications as well as determining the structure of the modification.

As an alternative, this report described an atmosphere-pressure MALDI source coupled to an ion-trap MS (LCQ-Deca-XP), which was used to assign peptide sequences by both MS and MS/MS spectra. The structure of modifications (e.g. disulfide linkage or phosphorylation) was determined by further fragmentation in this device (e.g. MS to 3 in the ion trap).

This new MALDI-Ion Trap is commercially available and compatible with the current Xcalibur control software and Bioworks data-handling software, which allows samples to be run and data to be analyzed automatically.

M. D. Person, K. Tikoo, T. J. Monks, S. Lau; University of Texas at Austin, 2501 University Avenue, Austin, TX 78712

Identification of post-translational modifications of proteins remains a major challenge in mass spectrometry. Our approach uses a comparative analysis of the proteolytic digest and MALDI-MS spectra of control and treated proteins to target differences due to modifications. Peptide mass mapping is used to identify peptides present. Accurate mass measurements indicate potential modification types for the peptides. HPLC-ESI-MS/MS is then used to sequence peptides and determine the type of modification present. This approach presents a broad strategy for identifying modifications without initial assumption as to type or residue localization. The MALDI screening step is used to reduce the analysis time. The MS/MS allows determination of sequence and modification. Successful characterization of a multiply modified histone protein is given as an example of the use of this strategy. Hyperacetylated histone 4 is separated on a gel into 5 bands representing different levels of acetylation. Bands are digested with chymotrypsin in-gel and then analyzed by MALDI-MS and HPLC-ESI-MS/MS to identify the sites of methylation and acetylation.

S. H. Bhattacharya, T. J. Raiford , K. K. Murray; Lousiana State University, Choppin Hall, Baton Rouge, LA 70803

Desorption/ionization on porous silicon (DIOS) is a form of laser desorption mass spectrometry in which a silicon surface acts as the energy-absorbing medium. Conventional MALDI methods employ an organic matrix for analyte desorption and ionization which can lead to matrix interference in the region below m/z 1000. DIOS allows the analysis of low molecular weight analytes since there is no low mass matrix interference. UV DIOS requires the preparation of a porous silicon surface by electrochemical etching in a hydrofluoric acid solution. We have recently employed a tunable mid-IR infrared laser to perform DIOS on unmodified single crystal doped silicon surfaces. Using the IR laser at 2.9 micrometers, it was possible to ionize peptides up to 2000 Da on an untreated single crystal silicon (100) surface with no additional solid or liquid matrix. The analyte was deposited on the surface in an ethanol solution, or in some cases, simply sprinkled onto the surface in powder form. In ongoing work, the effects of laser wavelength and surface treatment are being investigated.

S. J. Berger¹, D. Wall¹, R. Karol¹, J. W. Finch¹, H. Y. Xiong², S. A. Cohen¹; ¹Waters Corporation, Milford, MA, ²Micromass Inc., Beverly, MA

MALDI and ESI are the predominant ionization methods for introducing biomolecules into the gas phase prior to mass analysis. Until recently, ESI has monopolized the online analysis of complicated mixtures of proteins due to the ease in coupling liquid phase separations to ESI sources. However, the propensity of MALDI to generate ions with fewer charge states is appealing for the analysis of protein mixtures, as multiple overlapping charge states often arising from ESI complicate spectral interpretation. Here, we demonstrate a device for continuous deposition of eluent from analytical/capillary scale RPHPLC protein separations onto MALDI plates precoated with matrix. MALDI analyzed separations of intact proteins, protein mixtures, and complexes are presented, and comparisons to LC/ESI-TOFMS analyses are shown. Our results show that LC-MALDI plate capture of separations can be automated, and that captured separations can be archived for re-examination. Furthermore, we demonstrate that automated LC-MALDI plate collection of proteins is accomplished without sacrificing highly resolved LC separations.

N. Leymarie, E. Berg, P. O. OConnor, J. Grogan, F. G. Oppenheim, C. E. Costello; Boston University School of Medicine, 715 Albany Street, R806, Boston, MA 02118-2526

The salivary proline-rich proteins (PRP) are involved in the maintenance of tooth integrity. Recognition of the difficulties in obtaining PRP structural information has led to the development of methodologies based on mass spectrometry. These methodologies utilize enzymatic digestion followed by the characterization of PRP by classical MS/MS techniques (CID, post-source decay). The abundance of y-type fragment ions associated with proline (proline-effect) gives rise to difficulties in obtaining complete sequence data.

Recently, a new method of fragmentation, electron capture dissociation (ECD), has emerged. It cleaves peptide backbones at the Ca-N bonds rather than at the amide linkages.

The goal of this poster is to demonstrate the power of ECD for the determination of the structure of an acidic PRP and to obtain more complete sequence information of the proteolytic peptides derived from it. Due to the cyclic structure of proline, ECD spectra do not contain c and z· product ions on the N-terminal side of proline. Additionally the localization of post-translational modifications using ECD and the complementarity of ion fragmentation methods such as CID and ECD will be discussed.

W. Sandoval¹, K. O’Connell¹, J. T. Stults², W. J. Henzel¹; ¹Genentech, South San Francisco, CA, ²Genentech, Inc., South San Francisco, CA

We have explored the use of both Lys-C and Lys-N endoproteases to obtain more complete sequence coverage for the de novo sequencing of peptides from MS/MS spectra. Lysine-C from Achromobacter lyticus is a well-characterized endoprotease that is stable in a variety of denaturants including urea and SDS. Endoproteases that cleave at the N-terminal side of lysine have also been purified from several fungal sources. We have compared MS/MS data obtained from digestions of model proteins with Lys-N from Grifoloa frondosa and with Lys-C from > Lysobacter enzymogenes. Each enzyme produces peptides with the same sequences except the lysine residues are at opposite end of the peptide. This different placement of a basic residue results in significantly different MS/MS fragmentation patterns. The complementarities of these spectra result in higher sequence coverage and/or higher confidence in sequence assignment.

H. A. Sawyer¹, B. T. Ruotolo², J. T. Marini², G. F. Verbeck, IV¹, K. J. Gillig², D. H. Russell²; ¹Texas A&M University, College Station, Texas 77840, ²Laboratory for Biological Mass Spectrometry, Texas A&M University, mail stop 3255, College Station, Texas 77843

Metal ions play important roles in many biological processes and there is considerable interest in developing mass spectrometry methods for determining structure and binding sites of peptide/protein metal ion complexes. Ion mobility-mass spectrometry couples volume-to-charge and mass-to-charge measurements and yields information about gas phase structure. For a homologous ion series, a plot of ion drift time versus mass-to-charge over a limited mass range yields a near-linear relationship. However, alkali ion adducts of bradykinin and subunits of bradykinin (des-Arg⁹, des-Arg¹, fragments 2-7, 5-9, 1-5, 1-6, and 1-7) deviate from the expected linear relationship. This result suggests that [M+Cs] ⁺ have similar mobilities to that of [M+H] ⁺, suggesting similar cross-sections. These deviations are rationalized in terms of conformational change to the peptide upon alkali metal adduction. H/D exchange experiments as well as molecular modeling will be coupled with ion mobility data to address the conformational effects of cesium ion adduction as well as the effects of primary structure and chain length on the metal ion binding site.

B. D. Miller, C. A. Miller, F. E. Kuhlmann; Agilent Technologies, Inc., 1601 California Avenue, Palo Alto, California 94304

Protein identification by protein database search of peptide-map data produced by ion trap LC/MS/MS instrumentation is a well-established technique. In this technique it is common to employ automated triggering of MS/MS experiments based on real-time evaluation of MS spectral information content. Driven by increasingly complex proteomics research questions and expanded analytical demands, efforts continue to improve both the sample throughput and the data quality available from unattended, automated LC/MS/MS analyses of protein digest mixtures.

In this work, we demonstrate the use of two novel precursor-ion selection criteria to improve data quality and identification results in the automated ion trap-LC/MS/MS analysis of protein digest mixtures: preferred charge-state and preferred mass-list. Preferred charge-state mode provides for real-time, data-dependent exclusion of singly charged peptide precursor ions and/or preferential selection of doubly charged peptide precursor ions. Preferred mass-list mode biases on-the-fly precursor selection in favor of targeted peptide precursor ions, e.g., chemically modified peptides that may be present at low relative abundance in a complex digest mixture.

J. L. Whitney, M. E. Hail, D. J. Detlefsen; Novatia, 301A College Road East, Princeton, NJ 08540

In recent years pharmaceutical and biotechnology companies have initiated wide-reaching programs in genomics and proteomics. These ongoing initiatives are resulting in a flood of new proteins and oligonucleotides used as both targets and potential therapeutics. Characterizing these biomolecules throughout the discovery and development process is of utmost importance. Rapid and automated analytical methods are needed to address important issues of molecular weight confirmation, purity assessment, sequence determination, purification heterogeneity, elucidation of chemical and biological modifications, bio-production monitoring, etc. To answer many of these questions it is often necessary to analyse the intact biomolecule. LC/MS, utilizing electrospray ionization, has emerged as a powerful technique in these application areas. However, to date, there are no fully automated approaches for analyzing and processing data obtained on intact biomolecules using LC/MS. Here we will illustrate several applications of biomolecule characterization using our totally automated LC/MS system.

S. Leicester; Micromass UK Ltd., Wythenshawe, Manchester United Kingdom

Current databank search techniques facilitate the identification of peptides from MS/MS data. However, it is becoming clear that peptides are missed in the identification process due to non-specific cleavage, amino acid substitution, databank errors and the absence of the peptide from the databank being searched. We present a methodology that allows chaining of different search algorithms together to form an overall search strategy. By combining databank searching with PTM analysis, non-specific cleavage sites, amino acid substitution and de novo peptide sequence determination, we show that peptides can be identified that would otherwise be missed by databank searching alone.

J. Leszyk; University of Massachusetts Medical School, 222 Maple Avenue, Shrewsbury, MA 01545

Matrix-Assisted-Laser-Desorption/Ionization Time of Flight (MALDI-TOF) mass spectrometry has emerged as an important tool in the identification of gel purified proteins for various proteomic studies. MALDI -TOF has been somewhat limited to the determination of intact mass information from several peptides followed by enzyme specific mass correlation analysis to make an identification. This limitation is due in part to the difficulty in obtaining good MS/MS information via Post Source Decay (PSD) fragmentation for further MS/MS database correlation analysis. I describe here the use of the Kratos Axima CFR MALDI-TOF mass spectrometer as a powerful tool for the identification of proteins from both MS and MS/MS information. The Curved Field Reflectron (CFR) is capable of acquiring complete PSD spectra with isotopic resolution in a single acquisition phase at very high sensitivity (1-2 fmols). Other types of Reflectrons, which require voltage stepping and multiple acquisitions to assemble complete spectra often, take 20 to 40 minutes to acquire a single spectrum. PSD fragmentation spectra are not easily interpreted de novo but provide numerous fragments which allow definitive identifications via the Protein Prospector MS-Tag program. A powerful attribute of this program is its allowance of internal fragments for the correlation analysis of PSD spectra. The Axima CFR does not require any tuning or extensive instrument maintenance to obtain ultra high sensitivities. MALDI TOF-MS and PSD data from silver stained protein bands obtained from both 1D and 2D gel formats will be presented.

T. A. Shaler¹, P. V. Bondarenko¹, S. Swedberg²; ¹Thermo Finnigan, San Jose, CA, ²ThermoFinnigan, San Jose, CA

The entire human serum from a commercial pooled source was denatured, reduced and digested with trypsin prior to LC/MS/MS analysis using an ion trap mass spectrometer. The digest products were eluted from a 75 μm ID C18 column with a linear gradient 0 – 60% B over 500 min at 0.1 μL/min. The mass spectrometer was operated in a data-dependent MS/MS mode, when one full-scan MS was followed by three tandem scans on precursor ions selected "on the fly" from the MS scan. Using TurboSequest search algorithm, the MS/MS data were correlated with peptide sequences from a homo-sapiens database. A half of proteins were identified by only one matching peptide. Although correlation scores for these peptides were above conventional threshold, probability of true identification using one matching peptide was still not high. To improve this probability, the entire LC/MS/MS analysis was repeated three times. Those low abundant proteins, which were identified with only one matching peptide, but detected in two or three runs and with similar elution time, were included in the final list of identified proteins. Total 200 human serum proteins have been identified using described algorithm.

Guido F. Verbeck, Brandon T. Ruotolo, Kent J. Gillig, David H. Russell

Analytical chemists are always striving for techniques that add dimensions of orthogonality with increased throughput and sample complexity. Ion mobility (IM) time-of-flight(TOF) is a system that adds new dimensions to mass spectrometry(MS). IM separates gas phase ions based on their collision cross-section. IM can be coupled with time-of-flight mass spectrometry to yield a powerful tool used in the identification and characterization of proteins and peptides. One particular area of interest is in the separation/identification of different conformers of peptide and protein ions. IM-TOF-MS can simultaneously collect a spectrum of mobility through a drift gas (drift time) and mass-to-charge (m/z). Figure 1 illustrates the orthogonality of achieved by coupling these two separation techniques. The figure contains plots of drift time vs. m/z for a sum of peptide and carbon cluster ions. The ball-like structure of fullerenes and carbon clusters means their ions have high mobility (short drift times) relative to that for the peptide ions. These trend lines are seen because ions that possess similar structural features will have a constant of proportionality between mass and collision cross-section. We can probe structural differences between proteins and peptides based on this assumption. Figure 2 shows two peptides with similar mass but different mobilities. With the aid of molecular dynamics calculations and IM predictions, proposed structures of these ions can probe these different conformations. In this presentation the fundamentals of ion mobility will be covered with emphasis on applications to structural analysis of proteins and peptides.

J. M. Hettick, D. H. Russell; Laboratory for Biological Mass Spectrometry, Texas A&M University, TAMU 3255, College Station, TX 77843-3255

Tandem mass spectrometry is a powerful tool for the identification and characterization of proteins, and may be used to sequence peptides de novo or to increase the specificity of a peptide mass mapping/database searching experiment. Tandem MS requires that the analyte of interest be activated prior to fragmentation. Conventional approaches rely upon the collision of the analyte with a neutral target molecule(s). Research in our laboratory has focused on the development of a MALDI tandem TOF MS that utilizes radiation from a 193-nm ArF excimer laser for ion activation. Both photodissociation and conventional collision-induced dissociation (CID) produce sequence-informative fragment ions from protonated peptide molecules ([M+H] ⁺), however the types of fragment ions produced can be quite different. In particular, photodissociation produces side-chain cleavage ions important for differentiating isomass residues. In this study, the fragmentation produced by photodissociation is compared to that produced by CID of electrosprayed ions in a commercial quadrupole TOF instrument. The utility of the two methods for proteomic analysis will be discussed.

C. A. Miller, S. M. Fischer; Agilent Technologies, Inc., 1601 California Avenue, Palo Alto, CA 94304

Proteins are commonly separated by RPLC using wide-pore silica-based columns. Typically separations are done using acetonitrile/water gradients with TFA modifier. The limiting factor in reverse-phase protein separations is mass transport, i.e., diffusion of the proteins in and out of the pores. This was improved by the development of non-porous 2-micron silica-based columns. However, the limitation for this packing material was capacity. The development of superficially porous silica-microsphere packing with a porous outer shell yielded silica-based columns with the capacity advantage of traditional protein columns and the rapid mass transport of 2-micron non-porous columns. Using this column technology with electrospray mass spectrometry, rapid separation of protein mixtures with molecular weight determination can be done. This work investigates separation conditions that optimize mass spectrometric detection with minimal chromatographic impact.

K. Bernardo, N. Pakulat, O. Krut, M. Macht, H. Frank, H. Seifert, S. Fleer; University of Cologne, Joseph Stelzmann 9, Cologne, Cologne D-50931 Germany

Katussevani Bernardo^§, Norbert Pakulat^§, Marcus Macht^$, Silke Fleer^§, Oleg Krut^§ and Martin Kroenke^§.

M. E. Hail, J. L. Whitney, D. J. Detlefsen; Novatia, 301A College Road East, Princeton, NJ 08540

The recent increase in genomic and proteomic discovery has increased the need for high-throughput automated tools for biomolecule characterization. The flood of new protein targets will soon demand more efficient tools for evaluation of expressed proteins for drug discovery and structural biology studies. Intact protein molecular weight measurement using electrospray ionization mass spectrometry (ESI/MS) remains one of the most important methods to assess identity, purity, and the possibility of post-translational modifications. However, to date, there are no fully automated systems to perform these types of analyses. Here we describe a system, called ProMassXcali^TM, which utilizes on-line LC/MS, automated biomolecule deconvolution and web-based reporting for totally automated biomolecule molecular weight profiling. This system can be used to process data from a wide variety of biomolecules including large proteins, oligonucleotides, peptides, etc. An overview of ProMassXcali^TM will be presented along with selected applications that highlight its utility in a high-throughput environment.

M. Lubeck¹, C. Baessmann¹, U. Schweiger-Hufnagel¹, M. Shen², A. Ingendoh¹; ¹Bruker Daltonik GmbH, Fahrenheitstrasse 4, Bremen, 28359 Germany, ²Bruker Daltonics, Inc., Billerica, MA

Automated data dependent LC-MS/MS strategies have been utilized for analyses and subsequent identification of digested proteins. One of the challenges in working with physiologically derived proteins is that sample amount could be a limiting factor. We have demonstrated previously that nano-LC systems coupled on-line with esquire3000plus ion trap can provide an effective solution to protein identification in low femtomole and even attomole ranges.

The unparalleled high scan power of esquire3000plus facilitates fast switching between MS and MS(n) modes. Even so, the time needed for automated MS(n) is limited by typical peak widths of up to 30s.

We show here that "peak parking" was integrated by implementing a switching valve that could be controlled seamlessly through HyStar. The eluted peaks were recognized and preferentially selected by their charge states. The valve was used to block the flow from the pump when a peak was detected, and the peak was sent to the mass spectrometer by an additional pump with an 8-fold lower flow rate. Several proteins were digested and analyzed, and sequence information was obtained by de novo sequencing in Biotools using MS/MS and MS(3) spectra.

A. H. Henschen-Edman; University of California, Irvine, Biological Sciences 2, Irvine, CA 92697-3900

Cyanogen bromide cleavage was used to analyze posttranslational modifications in human fibrinogen, a 340 kDa dimeric protein with 33 Met per monomer. CNBr cleavage was performed in 75% formic acid at a protein concentration of 10 mg/ml. CNBr concentrations compared were 100, 10, 1 and 0.1 mg/ml. Reaction times were 2 or 24 h. Extents of cleavage and possible side reactions were evaluated by gel filtration chromatography, HPLC, N-terminal sequence analysis and MALDI-TOF mass spectrometry. The most complete cleavage was only obtained with 100 mg/ml for 2 h and 10 mg/ml for 24 h, i.e. concentrations considerably higher than described in the literature. The shorter cleavage time caused less oxidative damage to Trp residues. About half of the Met residues occurred to a small part in an uncleaved, modified form that was resistant to the acid in gel filtration, HPLC and mass spectrometry as well as to SDS-PAGE. However, these modified Met residues always cleaved during N-terminal sequencing, the sequence following the specific Met appearing both in the first and the second cycle indicating cleavage on both sides of these residues.

S. Wong, W. J. Henzel; Genentech, South San Francisco, CA

We have designed a hydrolysis sample preparation station (HSPS) which is used with PICOTAG hydrolysis vessels (Waters Corp.) to control the rapid succession of vacuum and nitrogen purging in the hydrolysis vessel. The HSPS is attached between the vacuum pump and the sample hydrolysis vessel and consist of several relays and an electronic counter. An electronic vacuum gauge containing outpoint relays allow nitrogen gas to be delivered to the hydrolysis vessel before bumping can occur. The counter is set to the number of purging/vacuum cycles desired. Sufficient cycles are required to prevent bumping when the last cycle is reached and high vacuum is applied. The hydrolysis vessel is then removed from the device and placed in an oven.

An improved sample clean-up procedure to process large number of samples containing high salts and buffer components for amino acid analysis has been developed. This procedure utilized the new Millipore Multiscreen 96-well sample plate with high capacity PVDF-psq membrane. Multiple samples may be applied to the plate and can be processed at the same time for further washing or concentrating.

These protocols have greatly reduced the time spent on sample preparation and improved the quality of analysis.

P. Jandik, J. Cheng, M. Weitzhandler, D. Farnan, C. Loran, N. Avdalovic; Dionex Corp., 445 Lakeside Drive, Sunnyvale, CA 94085

Anion exchange chromatography with IPAD detection (AEC/IPAD) is a recently introduced new tool for amino acid analysis. It avoids derivatization as a detection-enabling step. Amino sugars and amino acids can be separated and quantified in a single chromatographic run.

We have evaluated the use of this new tool for amino acid analysis after gel electrophoresis and electroblotting and also in fractions collected from HPLC.

In the study of electroblotted PVDF spots, we followed the format of the 1999 ABRF Amino Acid Survey Report. Several different levels of a test compound were analyzed and resulting data were used to search the PROPSEARCH and ExPASy sites for protein identification.

To assess the applicability of AEC/IPAD for amino analysis of HPLC fractions, we carried out two-dimensional separations (cation exchange/ reversed phase) of intact proteins and enzyme digested proteins. Using the same approach as with PVDF spots, we submitted the compositional data obtained from HPLC fractions to the Internet sites to test the reliability of identification.

K. D. Linse, T. E. Taub-Montemayor, J. W. Kent, M. Rankin; University of Texas at Austin, 2500 Speedway Blvd, Austin, Texas 78712

Reversed phase HPLC has traditionally been used to purify neuropeptide hormones from the corpora cardiaca glands (CC) of insects. Using RP-HPLC, we have previously identified an Adipokinetic Hormone (AKH) from the CC of the migratory grasshopper Melanoplus sanguinipes. This hydrophobic decapeptide has the same primary structure as locust (Lom) AKH I. Initially, only one HPLC peak with the same retention time as Lom AKH I was detected; however, we found and characterized a second putative AKH within this fraction using Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS) and HPLC tandem Electrospray Ionization Mass Spectrometry (ESI-MS). Armed with this information, we were able to improve the resolution of the native AKHs by HPLC. Our results show the power of coupling HPLC and MS methods to resolve and identify very similar peptides in minute amounts.

S. Kwak, N. Laliberte, M. Hincapie, R. Chicz; Zycos, Inc, 44 Hartwell Ave, Lexington, MA 02421

We present a rapid and efficient automated immunoaffinity chromatography method for the purification of Major Histocompatibility Complex (MHC) molecules. The speed of purification allows for greater sample throughput and better preserves the integrity of the non-covalently bound peptide repertoire of these receptors. Up to 4 different antibody columns can be serially connected, allowing simultaneous purification of multiple human leukocyte antigen (HLA) allotypes from a single cell source. High purity protein is isolated from these HLA specific antibody columns and the HLA molecules are free of contaminant proteins.

The recovery of purified HLA can range from one microgram to several milligrams depending on the protein expression level from as little as one gram of harvested cultured cells. This is illustrated by the purification of both HLA class I and class II protein from multiple preparations of human lymphoblastoid preparations. This method has been successfully applied for the isolation and subsequent sequence analysis of both HLA class I and class II T cell epitopes.

J. Masuda¹, E. J. Unsworth¹, D. M. Maynard¹, M. Nishimura², T. Ueda³, J. A. Kowalak¹, S. P. Markey¹; ¹NIMH/NIH, 10 Center Drive Room 3N321, Bethesda, Maryland 20892-1262, ²Shimadzu Scientific Instrumets, Inc., Columbia, Maryland, ³Shimadzu Corporation, Kyoto, Kyoto Japan

Over the past few years, several studies have been published on two-dimensional HPLC separation methods for proteome analyses. In general, a combination of strong cation exchange (SCX) chromatography and reversed phase (RP) chromatography is used for a two-dimensional HPLC with the sample fraction eluting from a cation exchange column being directly loaded onto a pre-equilibrated RP column. Analytical methods combining two orthogonal separation techniques have been directly coupled to electrospray ionization tandem mass spectrometers for the analysis of proteolytic hydrolysates of complex protein mixtures. Constituent proteins are subsequently identified using algorithms that utilize MS/MS data to search genomic databases.

In this paper, a new automated two-dimensional HPLC system is described that is based on micro-scale HPLC techniques and utilizes six capillary trapping pre-columns, CapTrap (Michrom BioResources, Inc.), mounted on a 14-port rotary valve to capture sample fractions after sequential elution from a micro-bore SCX column. SCX fractions trapped on CapTraps are effectively desalted, and sequentially transferred, by RP gradient elution, onto a pre-equilibrated capillary RP column. The chromatographic resolution of peaks that could not be separated using only RP chromatography was achieved when analyzing a tryptically digested protein mixture with ESI-ion trap as the detector. Data will be presented that illustrate the utility of a micro-scale two-dimensional HPLC system for complex protein digest mixtures in proteomic analysis.

J. S. Smith, E. R. Surriga, A. Kurosky; University of Texas Medical Branch, Rm. 2.226 BSB, Galveston, Texas 77555-0645

The Applied Biosystems 173A Microblotter has proven to be a valuable resource for isolating peptides at the pmol level after in-gel or on-PVDF membrane enzyme hyrdrolyses. On occasion and, for a variety of reasons, these samples may contain nonvolatile buffers, salts, and detergents that can interfere with the capillary HPLC performed by the microblotter. An on-line sample cleanup method for the microblotter was developed using commercially available Michrom BioResources peptide traps and nonionic detergent removal cartridges. In this procedure, we replaced the Microblotter’s HPLC sample loop with a rheodyne cartridge holder containing the peptide or nonionic detergent trap. Samples were loaded on to the trap and manually washed with 0.1% TFA prior to injection. The resulting chromatograms showed that the salt peaks and detergent artifacts were virtually eliminated and the baseline greatly improved when compared with control non-trap results. The resolution of the peptide fractions was also considerably enhanced. Finally, we have found that the traps also served as an effective guard column preventing back-pressure problems and extending the life of the column.

S. A. Cohen, R. S. Plumb, J. W. Finch, C. L. Stumpf; Waters Corporation, 34 Maple Street, Milford, MA 01757

The identification of proteins via the peptides produced from their tryptic digests is now the approach of choice in proteomics. When dealing with a complex mixture such as a protein digest it is necessary to have as much chromatographic performance and mass spectrometric sensitivity as possible. The use of nanoflow MS has significantly increased instrument sensitivity when dealing with the trace levels of material encountered in protein analysis. Nanoflow chromatography does, however, present serious problems, namely gradient fidelity at very low flow rates, band broadening due to sample loading, analyte losses due to adsorption and poor column performance. Here we will discuss the use of small trapping and focusing columns for sample introduction and discuss their benefits and drawbacks. The selection of tubing and components to reduce sample loss will also be discussed along with potential solutions. Finally we will consider the effects of the chromatographic column, and the merits of pulled tip and/or fritted columns. Procedures that produce highly efficient, reproducible 75 micron columns will be demonstrated while highlighting the beneficial effects on MS sensitivity and spectral quality.

M. Weitzhandler, H. Schwartz, C. Saini, M. Rey, N. Avdalovic; Dionex Corp., 445 Lakeside Drive, Sunnyvale, CA 94085

We have developed a new monosaccharide column that gives faster, more efficient separations of glycoprotein monosaccharides with better spacing of monosaccharides. The new column (3mm x 150mm) is based on a new chemistry that reduced the resin bead particle size from 10 microns to 7 microns. Faster, more efficient separations of glycoprotein monosaccharides with better spacing was achieved across a range of isocratic NaOH concentrations (8 – 20 mm NaOH) at lower flow rates. Sensitivity of monosaccharide detection was improved as a consequence of the higher efficiency resin and narrower column diameter, enabling routine low pmol monosaccharide determinations. Fucose, galactosamine, glucosamine, galactose, glucose, and mannose were baseline separated and eluted in less than eight minutes (cycle time 30 min) at a flow rate of 0.5 mL/min (back pressure < 800psi). Further, when used with a 3mm AminoTrap guard column, amino acids did not interfere with monosaccharide determinations, enabling the direct monosaccharide profiling of glycoproteins with low levels of glycosylation such as monoclonal antibody hydrolysates.

F. Maroto¹, M. Scigelova², J. Vazquez³; ¹Thermo Finnigan, San Jose, California, ²Thermo Finnigan UK, Boundary way, Hemel Hempstead, HP2 7GE United Kingdom, ³Centro de Biología Molecular Severo Ochoa, Madrid, Spain

Post-translational modifications (PTM) are prominent features of proteins in higher organisms. Automated LC/MS analysis of peptides is a high throughput technique for protein identification. Several strategies have been developed to detect PTMs in an automated fashion suitable for high throughput peptide characterization.

We have compared the merits of several different strategies using a correlation of either peptide masses obtained from full scan spectra or sequence tags derived de novo from MS/MS spectra. Both approaches use a systematic search of peptide mass differences that can identify a modified peptide. The key benefit of using the de novo MS/MS strategy is that it is able to detect an unexpected post-translational modification and its exact location.

J. I. LANGRIDGE¹, A. Millar², C. HUGHES³, H. VISSERS⁴, T. Dourdeville⁵, J. B. Hoyes⁶; ¹Micromass UK Ltd, FLOATS ROAD, MANCHESTER, M23 9LZ United Kingdom, ²Micromass UK Ltd., Floats Rd, Wythenshawe, Manchester M23 9LZ United Kingdom, ³Micromass UK, MANCHESTER, ⁴Micromass Europe, Almere, Netherlands, ⁵Waters Corp., Milford, MA, ⁶MICROMASS UK LTD, Almere

Mass spectrometry has firmly established itself as the primary technique for identifying proteins. Currently, the mass spectrometry technique providing the highest degree of specificity and sensitivity is electrospray LC-MS/MS. However in the case of post translationally modified peptides only a very limited sub set of the peptides present are required to be fragmented and often these low intensity peaks can be missed. A solution to this problem is a method that allows specific post translationally modified peptides to be identified during the course of an LC-MS experiment. In the case of phosphotyrosine, a low mass immonium ion at m/z 216 can be detected. This characteristic ion is used to direct the mass spectrometer to fragment potential phosphopeptide precursor ions which are present at that time point in the low energy data. In this case several precursor ions may require MS/MS interrogation at one decision making time-point and implementation of a chromatographic technique known as variable flow chromatography allows greater time to interrogate these peaks. This approach will be discussed with examples of where this methodology has been used for the targeted analysis of phosphorylated peptides.

A. S. Woods¹, K. Fuhrer², M. Gonin², T. Egan², J. A. Schultz²; ¹NIDA IRP, NIH, 5500 Nathan Shock Drive, Baltimore, MD 21208, ²Ionwerks, Houston, TX

Spectra acquired using MALDI and Ion-Mobility-Orthogonal MALDI instruments show that phosphorylated residues form salt bridge with quaternary ammonium compounds. Each phosphate group forms a non-covalent complex with one or two quaternary ammonium compounds. When both the mono- and di-ammonium compounds are present in the mixture we see MH+ of the phosphorylated compound and two mono- or di-quaternary ammonium molecules, or MH+ to which one mono- and one di-ammonium added to the same phosphate.

IM coupled with MALDI is metamorphosing from laboratory curiosity to useful analytical technique. Laser Desorbed peptide ions are drifted through 2 Torr of He by a 40 V/cm field. The ions drift with a speed related to the ratio of their volume/charge (shape), which is slightly different in many cases than their mass/charge ratio. IM allows separation based on the conformation or overall shape of the MH+. Sensitivity for peptides is not far from conventional MALDI instruments. We can separate phosphorylated peptides from non-phosporylated ones, and achieve even larger separations between peptides as a class when compared to inorganic salts and small aromatic organic molecules. This ability to do "plasma chromatography" within the time scale of the combined MALDI experiment makes this instrument attractive for analysis of complex biological samples. We will present data showing the separation of phosphorylated peptides or lipids, and their non-covalent complexes with quaternary ammonium compounds by Ion-Mobility.

R. Shen¹, P. Shieh¹, W. S. Hancock²; ¹ThermoFinnigan Corp., 355 River Oaks Pkwy, San Jose, CA 95134, ²Thermofinnigan corp., San Jose, CA

Phosphorylation is an important post-translational modification critical in various cell functions including receptor-mediated signal transduction. The often transient phosphorylation of minute amount of a protein has made the determination of site of phosphorylation rather challenging. We explored the potential of using magnet-immobilized anti-phosphotyrosine or metal chelating matrix for capturing phosphorylated proteins/peptides derived from A431 cells. The tryptic peptides were analyzed by LC/MS and the identities and sites of phosphorylation revealed by database search. Several proteins bound by the anti-pY-coated beads are proteins with known regulatory function, e.g., proteasome activator, SH3-containing adaptor, and hu MCM2, topoisomerase I, etc. The identification of two or more peptides for some of these proteins and the absence of these proteins in the control beads suggest that the binding is antibody-dependent. A phosphopeptide derived from epidermal growth factor receptors (EGFR)phosphorylated at a tyrosine reside was identified. We've also identified and mapped the sites of phosphorylation of several peptides captured by Ga+3-NTA agarose. These approaches involve minimal sample manipulation and are desirable when dealing with limited amounts of samples.

S. Wu, P. Shieh, W. S. Hancock; ThermoFinnigan Corp., 355 River Oaks Parkway, San Jose, CA 95134-1991

Purpose: To obtain detailed carbohydrate and peptide structures for a glycosylated protein using a high throughput, automated procedure.

Method: LC-MSn</ on an ion trap using data dependent ion selection and dynamic exclusion.

Results: The majority of the protein sequence and glycosylation sites were identified in one MS/MS run using a small aliquot of a tryptic digest mixture. A second experiment was performed confirming the glycoform and linked peptide structures via data dependent MS2 through MS5.

Conclusions. A. The complicated peptide mixture was easily identified through the MS/MS fragment ion match in Turbosequest search software. B. The post-translational glycosylation sites were identified and the glyco-structures were clearly elucidated by MS from 2 to 5th on the fly. C. The overlooked and low-intensity ions (e.g. glycopeptides) can still be automatically picked up for MS/MS fragmentation by dynamic exclusion software on the fly. D. The assembly of Surveyor capillary LC coupling with ion-trap MSn (LCQ) for sample analysis and equipped with Turbosequest software for data analysis is automatic, robust, and easy to use.

H. Tran¹, A. Huhmer², S. Swedberg¹; ¹ThermoFinnigan, 355 River Oaks Parkway, San Jose, CA 95134, ²Thermo Finnigan, San Jose, CA

Calmodulin (CaM) is the major high affinity Ca2+ -binding protein in most eukaryotic cells that regulates a variety of cellular Ca2+ -dependent signaling pathways. CaM has an unusual amino acid composition with a high proportion of acidic residues (> 34%) and is particularly rich in methionine (Met) residues, accounting for more than 5% of all amino acid residues. It has been demonstrated that Met residues are the primary targets of oxidative changes in CaM in vivo and in vitro. We discuss experimental approaches for data dependent MSn analysis exploiting the unique capabilities of ion-trap MS for the identification and characterization of oxidative modifications in CaM. Different data-dependent acquisition methods are evaluated to maximize sequence coverage and sequence information for those peptide fragments with two possible oxidation sites. We also discuss search strategies for the automatic identification of those post-translational modifications by TURBOSequest.

L. K. Barbacci, J. M. Moeller, M. C. Harvey, J. G. Dapron, W. K. Kappel; Sigma-Aldrich Biotechnology, 2909 Laclede Ave, St. Louis, MO 63103

The analysis of protein phosphorylation sites is a major endeavor in the field of proteomics. Current techniques commonly employ protein digests with subsequent analysis by mass spectrometry (MS). Generally there is a need to enrich the phosphopeptides prior to MS analysis, often utilizing immobilized metal affinity capture (IMAC) matrices. A novel chelate was attached to both resin and 96-well plates followed by loading with iron (Fe⁺³). Tryptic digests of β-casein were used to evaluate the specificity of phosphopeptide capture. Enrichment of the phosphopeptides (75–90% purity) was observed using HPLC-MS. Using purified monophosphopeptide (MPP) from β-casein, capacity was determined as approximately 0.5 nmole per well for the plate and 1 μmole per mL for the resin. Capacity for the purified tetraphosphopeptide (TPP) was greater than the MPP in both formats. In the peptide mixtures, preferential binding of the TPP was observed. Various eluents compatible with MS analysis were evaluated. Significant differences in recovery efficiencies were noted. Optimal utilization of elution conditions is paramount in ensuring representative recovery of complex phosphopeptide mixtures.

L. Nuwaysir, T. Settineri, C. Hunter, B. Halpern, D. Schaeffer; Applied Biosystems, Foster City, CA

Mass spectrometry has become one of the primary techniques for identifying proteins and studying protein expression levels and protein function. Many software programs have been developed to identify proteins from mass spectrometric data. The software programs Pro ID and Pro ICAT use a patent-pending algorithm called InterrogatorTM for identifying proteins from uninterpreted MS/MS data. In addition to finding specific modifications, error and modification-tolerant searching is possible by specifying a generic "zone modification" mass to add or subtract to the calculated peptide molecular weight. The zone modification can be due to specific chemical modifications, amino acid substitutions (homologous peptides), truncations, non-specific or missed cleavages, or any other modification that results in the experimental mass differing from the expected database mass. The Interrogator search algorithm returns the modification mass and the region of the peptide where the modification is localized. This powerful feature will be demonstrated with examples of proteins that have been identified from post-translationally modified peptides, non-specifically cleaved peptides, and disulfide-bound peptides.

M. Scigelova¹, G. Woffendin², M. Ward³, H. Byers³, D. Hanger⁴; ¹Thermo Finnigan UK, Boundary way, Hemel Hempstead, HP2 7GE United Kingdom, ²Thermo Finnigan, Hemel Hempstead, United Kingdom, ³Proteome Sciences plc, De Crespigny Park, London, SE5 8AF United Kingdom, ⁴Kings College London, London, United Kingdom

Phosphorylation of proteins plays very important roles in regulation of metabolic processes. Neurofilament proteins are important structural features of the neuronal cytoskeleton. Abnormal phosphorylation of these proteins is associated with some neurodegenerative diseases.

The MS/MS spectra of peptides provide a wealth of information on the assignment of peptide sequence and identification of modified residues. Phosphorylated peptides usually exhibit a prominent loss of a phosphate group in their MS/MS spectra. A neutral loss of 49 amu (for doubly charged peptides) indicates a possible phosphopeptide candidate. An identification of the peptide sequence is then obtained with SEQUEST database search engine, and the exact location of the phosphorylated residue is confirmed.

We present a simple strategy enabling an unambiguous identification of the phosphorylated residue of a peptide analysed by LC/MS/MS in a complex peptide mixture.

G. Woffendin¹, M. Scigelova², E. Tarelli³, S. Pouria⁴, P. Corran⁵; ¹Thermo Finnigan, Hemel Hempstead, United Kingdom, ²Thermo Finnigan UK, Boundary way, Hemel Hempstead, HP2 7GE United Kingdom, ³St. George's Hospital Medical School, London, United Kingdom, ⁴Kings College Hospital, London, United Kingdom, ⁵London School of Hygiene and Tropical Diseases, London, United Kingdom

Glycosylation is a prominent post-translational modification (PTM) of proteins in higher organisms. The heterogeneity of glycan population, a multitude of combinations of sugar building blocks, and a relatively large size makes the study of glycopeptides a challenging task.

The use of MS/MS is essential in analyses of PTMs such as glycosylation. We have used a straightforward routine to identify glycosylated peptides in complex mixtures by automated capillary LC/MS on an ion trap by profiling specific losses of sugar moieties. Multiple sugar losses observed under optimised fragmentation conditions afforded important structural information about the glycan portion of the molecule.

J. D. Haley¹, A. Thelemann¹, H. Pan¹, D. Fenyo², T. Settineri³, K. Iwata¹, A. Bruskin¹; ¹OSI Pharmaceuticals Inc, 106 Charles Lindbergh Blvd, Uniondale, NY 11553-3649, ²Proteometrics LLC, New York, NY, ³Applied Biosystems, Foster City, CA

Enhanced activation of the epidermal growth factor receptor (EGFr) is frequently observed in human cancers and correlates with poor prognosis. Affinity chromatography/MS (MALDI and Qq-TOF) was used to define phosphotyrosine-containing protein complexes. 638 MALDI spectra from multiple experiments and separations were evaluated (RADARs; Proteometrics LLC) allowing tentative identification of 129 proteins based on matches common to both trypsin and GluC (20-30ppm). The predominant protein identified by all methods was EGFr. Peptides phosphorylated on P1 and P2 C-terminal tyrosines were observed. Proteins from a specific internalization/ degradation pathway were confirmed by Western transfer. 34,931 MS-MS spectra identified 1813 peptides using stringent selection criteria (Pro-ID; ABI-Sciex). 34 individual EGFr peptides were sequenced from multiple MS-MS scans. 81 of 129 MALDI-assigned proteins were identified by uLC-MS-MS using established cutoff criteria. Proteins of unknown function and novel complexes were detected. ICAT labeling is being used to measure protein changes associated with the blockade of EGFr kinase activity by the specific quinazoline inhibitor OSI-774 (Tarceva).

H. A. Remmer¹, N. Ambulos², L. F. Bonewald³, J. J. Dougherty², E. Eisenstein⁴, E. Fowler², J. Johnson², A. Khatri², N. Ritter², S. T. Weintraub²; ¹University of Michigan, ², ³University of Missouri, ⁴National Institute of Standards and Technology, Rockville, MD

The Peptide Standards Project, conducted by the Peptide Standards Project Committee (PSPC) of the ABRF, entails production and establishment of 3 synthetic peptides as registered and certified peptide reference standards by collaboration with the National Institute of Standards and Technology (NIST).

The committee’s mission is to provide active support and advice to NIST in conducting this project. Therefore the PSPC focuses on the following:

(1) Large scale synthesis the standard peptides in highest purity.

(2) Packaging of the peptides in small quantities.

(3) Analysis of peptide quality by the ABRF member laboratories in a collaborative study.

(4) Stability tests for a variety of physical and chemical conditions and for long-term storage and handling.

Results of synthesis optimization and the set-up for large scale synthesis will be presented and discussed as well as peptide design, quality control and a project overwiew.

The Peptide Standards Project is the first ABRF project funded by NIST. Collaboration with NIST provides additional recognition and visibility for the ABRF. Vice versa, the ABRF membership is a valuable resource of information, data and data validation for NIST. For the Peptide Standards Project, the collaborative analysis of the peptides by the ABRF membership will yield the required certificate of analysis.

J. P. Wiley¹, K. A. Hughes², R. J. Kaiser¹, G. Li¹, D. D. Lucas¹, K. P. Lund², C. L. Pershing¹; ¹Prolinx, Inc., 22322 20th Ave SE, Bothell, WA 98021, ²Prolinx Inc., Bothell, WA

Surface plasmon resonance is a powerful, sensitive and label-free methodology for the analysis of biomolecular interactions. However, the utility of this technology is limited by the cost, throughput, and complexity of available instrumentation and surface chemistries. Prolinx®, Inc. has developed a new SPR instrument, the Octave™ Molecular Interaction Analysis System, which will address these limitations and will enable SPR biosensors to become ubiquitous in life science research and drug discovery laboratories. This technology is the result of combining the Texas Instruments Spreeta® 2000 chip with Prolinx Versalinx™ Chemical Affinity Tools. The Octave incorporates eight independent sensors operating in parallel. The gold sensor surface has been modified to minimize nonspecific binding and utilizes Versalinx Chemical Affinity Tools to facilitate the immobilization of macromolecular ligands for binding studies. Samples are introduced to the sensors using an integrated liquid handling robot, enabling walk-away automated analysis. The versatility and ease of use of the Octave provide affordability and accessibility of molecular interaction analysis for both academic and industrial laboratories.

O. Beg; University of Maryland, 701 East Pratt St., Baltimore, MD 21202

Lipoprotein Lipase (LPL) is a key enzyme in Lipid Metabolism. Recent evidence suggests that LPL could play a role in the process of atherosclerosis by retaining lipoproteins (LP) at the surface of cells, which when taken up the cells contribute in the process of atherosclerosis. In view of these functions of LPL, a study was conducted by making use of synthetic peptides corresponding to potential heparin and LP binding sites to study the role of critical residues in LPL which bind its ligands. In contrary to the site directed mutagenesis results these studies map a binding site for both LP's and heparin to a specific region in LPL. These results also suggest that peptides corresponding to specific sequences in LPL are also responsible in inducing structural change(s) to accommodate an increase in binding its ligands. These preliminary findings suggest a possible mechanism of allosteric regulation in LPL. Data from these studies will be presented to delineate a different approach to study heparin and LP binding sites in LPL.

(This work was supported by a grant from American Heart Association-TN affiliate to OUB. Author acknowledges this support).

M. Raida, M. Schirle, T. Werner, B. Kuster; CellZome GmbH, Meyerhofstrasse 1, Heidelberg, n/a D-69117 Germany

Most cellular processes are maintained by functional multi-protein complexes rather than by single proteins. In a complex, proteins of different biological functions are assembled to exercise highly specific overall function. As proteins are used as drug targets, it is advantageous to gather information about their interaction partners as to provide more insight into the molecular environment in which drug targets operate. Intact protein complexes can be purified from cells by using the tandem affinity purification. By applying specific protocols, complexes from different subcellular compartments can be purified. From yeast, complexes containing up to 100 proteins have been purified. In general, proteins are first separated by 1D SDS-PAGE and subsequently identified by mass spectrometric methods.

Depending on the complexity, different mass spectrometric approaches have to be selected. While well-separated proteins from known small genomes can easily identified by MALDI-TOF MS instruments, more complex mixtures require LC-MS/MS approaches. The application of one- and multi-dimensional LC separations coupled to either quadrupole TOF or ion trap mass spectrometers will be discussed in detail.

D. Bintzler, S. Sherwani; University of Cincinnati, 231 Bethesda Avenue, Cincinnati, Ohio 45267-0524

Human DNA profiling, also referred to as DNA fingerprinting, is a method of fragment analysis used for assigning a genetic identity to a human being. In the US, human DNA testing has been used as evidence in police forensics, other criminal cases and parentage issues since 1988. It has also been used to settle immigration cases, disputes involving purebred dogs and in animal conservation. Since these issues can be very sensitive, absolute accuracy is the foremost requirement for DNA profiling. The University of Cincinnati DNA Core Facility provides DNA profiling as a new service. DNA profiling uses 10 certified markers to amplify 10 specified areas of the human genome, including a marker to identify X and Y chromosomes. The product sizes range from 106 base pairs to 343 bases pairs. Various protocols in method design were tested to improve upon the accuracy. The Applied Biosystems DNA profiling allelic ladder with known base pair values provided a good standard to calculate error. It was determined that the gel percentage was a key factor. Since accuracy is an extremely important factor for any laboratory providing fragment analysis as a service, results from this study are reported here.

C. R. Kemper, B. Schulenberg, W. Leung, W. F. Patton; Molecular Probes, Inc., 4849 Pitchford Avenue, Eugene, Oregon 97402

Recombinant proteins are often linked to oligohistidine tags and purified by immobilized metal-ion affinity chromatography. Typically, oligohistidine-tagged proteins must be electroblotted and probed using anti-oligohistidine antibody/ secondary antibody-reporter enzyme conjugates or biotin-nitrilotriacetic acid/streptavidin-reporter enzyme conjugates, to verify fusion tag expression and monitor fusion protein purification. Two fluorophore- nitrilotriacetate conjugates, Pro-Q Sapphire 365 and Pro-Q Sapphire 488 oligohistidine gel stains, permit fluorescence detection of oligohistidine-tagged proteins directly in SDS-polyacrylamide gels. Pro-Q Sapphire 365 dye exhibits bright blue fluorescence when illuminated with a UV transilluminator while Pro-Q Sapphire 488 dye exhibits bright green fluorescence when illuminated with a 470-488 nm laser-based scanner. 25-65 ng of fusion protein is detectable using either stain. Gels may then be post-stained with the red-fluorescent SYPRO Ruby protein gel stain to reveal the total protein pattern.

B. S. Miller, B. Fryksdale, G. Ganshaw, P. T. Jedrzejewski, D. Wong, G. England, A. Gaertner; Genencor International Inc., 925 Page Mill Road, Palo Alto, CA 94304

Semi-automated 2D gel systems have been developed and are entering service in a number of industrial and academic laboratories. We describe the advantages and applications of such a system applied to an industrial enzyme project in a product based biotechnology company. We discuss the generation of high quality data that allowed protein identification using a fungal exoproteome as an example. We also discuss the limitations of the technology and illustrate our findings with examples from product composition analysis, whole cell and cell fractionation proteomics experiments.

Our conclusions are based on the development of reliable analytical methods using deglycosylation of the proteins that facilitates peptide assignments via MALDI-TOF mass spectrometry. Also described are analytical investigations on the "charge trains" seen in 2D gel electrophoresis.

G. L. Hansen¹, L. Stevens¹, F. Gharahdaghi², S. Kim¹, S. M. Mische¹, V. Papov¹; ¹Boehringer Ingelheim Pharmaceuticals, 175 Briar Ridge Rd., Ridgefield, CT 06877, ²AstraZeneca, Willington, DE

Two-dimensional gel electrophoresis (2-DE) is currently the favored method for the separation of proteins not only for the highly achievable resolution but also because 2-DE allows direct visual comparison of the changes in protein expression or modification between the proteomes under study. For many years proteins separated by 2-DE have generally been stained with either the well-known Coomassie blue (CB) or the more sensitive silver staining.

The increase in popularity of proteome analysis has fueled numerous efforts to develop "new" stains and protocols in an effort to maximize the limit of protein detection while maintaining compatibility with mass spectrometric analysis. Here, we compared some of the newer commercially available stains with the stains and dyes already in widespread use. Further, we examined a double stain of colloidal CB, followed by silver stain to determine whether an enhancement in stain sensitivity could be achieved above either stain alone.

Stains/dyes and methods were compared in terms of limit of detection, compatibility with mass spectrometry and ease of use. Band intensities in 1D gels were visually compared to determine the sensitivity of each stain. The number of proteins detected in 2DE gels and mass profiling were summarized. The results of these experiments allowed us to develop a coherent strategy for the visualization of proteins for proteome comparison without sacrificing analyte sensitivity in the mass spectrometric analysis of peptides which resulted from in-gel digests of selected protein spots.

K. J. Martin, B. Schulenberg, C. R. Kemper, W. F. Patton; Molecular Probes, Inc., 4849 Pitchford Ave., Eugene, OR 97402

For specific protein identification, researchers typically wrestle with the time-consuming task of running replicate gels; one gel being stained for total protein while the other is electroblotted and probed using standard immunoblotting methods. Ambiguity arises due to difficulty in aligning the gel and the electroblot. Simultaneous dichromatic detection, on one blot, eliminates the inherent uncertainty arising from running replicate gels. A two-color fluorescence detection method is described based upon covalently coupling BODIPY TR-X succinimidyl ester to electroblotted proteins, followed by detection of target proteins using ELF 39 phosphate in combination with alkaline-phosphatase-conjugated reporter molecules. This results in all proteins in the profile being visualized as fluorescent red signal while those detected specifically with the alkaline phosphatase conjugate appear as fluorescent green signal. The dichromatic method permits detection of low nanogram amounts of protein and allows unambiguous identification of target proteins relative to the entire protein profile on a single electroblot.

N. E. Caceres¹, T. R. Brockett¹, I. Jordon-Thaden², R. L. Cerny², G. Sarath¹; ¹University of Nebraska-Lincoln. Center for Biotechnology-Proteomics Core Facility, 19 th & Vine, Lincoln, NE 68588-0665, ²Mass Spectrometry Core Facility, Lincoln, NE

Our facility handles a broad variety of samples from diverse sources. This poses several challenges for providing effective separation and imaging services. Successful 2DE greatly relies on efficient extraction and solubilization of proteins. Our goal was to develop standard protocols to achieve the best separation and reproducibility for all of the samples. However, the complexity of the material specified the necessity for customized sample procedures. Proteins from bacterial crude extracts and purified protein fractions were recovered with 10 % TCA or precipitated with cold acetone if the samples were more diluted. Spot patterns were improved when extracts were further cleaned with phenol extraction. For protein preparations from plant tissues and fungal mycelia we used 10 % TCA in acetone containing 0.07 % β-mercaptoethanol. Proteins prepared by this protocol could be focused in absence of DTT. Our equilibration buffer contained 5 mM TBP. Under these conditions, streaking in the 2nd dimension was greatly minimized. Data obtained from our experiments will be presented. Our experience revealed that most samples required substantially different conditions for extraction as well as focusing.

A. cherkasskiy¹, L. lacy¹, J. J. Broadbent², J. loveless¹, K. Appasani²; ¹perkinelmer life sciences, 100 e. dedham, boston, ma 02118, ²PerkinElmer Life Sciences, Boston, Massachusetts

Total RNA was isolated from Jurkat and HL-60 cell lines. Separate labeling reactions were run using 100, 75, 25, or 10 ug of input total RNA using the MICROMAXTM Direct Labeling Kit. Same vs. same hybridizations were performed at all mass levels using Jurkat in both channels and scatter plots were examined. The above all masses showed differential expression within two-fold limits for greater than 99% of the genes. For the lowest levels of mass (10 ug), an overall weaker signals and an increased scatter was observed, indicating that the possibility of false positive data from the weakest signals. Differential expression patterns were studied using hybridizations comparing Jurkat to HL-60 expression. Again, lowest mass levels resulted in increased scatter on the lowest signal and also somewhat reduced expression ratios on highly expressed genes. We conclude that lowering mass levels of input RNA, especially to the 10 ug range may result in a loss of sensitivity, specificity and differential expression data.

F. J. Hilerio¹, D. T. Campogan¹, K. A. Hughes², D. Spicer¹, A. L. Springer², L. R. Booth¹; ¹Prolinx, Inc., 22322 20th Avenue SE, Bothell, Washington 98021, ²Prolinx Inc., Bothell, WA

Nucleic acid microarrays allow for the profiling of thousands of genes in a single experiment and have become an increasingly valuable tool in gene expression analysis. A typical microarray is composed of a 3 x 1 glass slide containing an array of nucleic acids to which fluorescently-labeled cDNAs are hybridized. Considerable time and expense can be spent in preparing RNA from which the labeled cDNA is derived. The removal of excess dyes from the reverse transcription reactions, while maintaining recovery of message, is essential to the accuracy of experimental results. Prolinx^®, Inc. has developed a new method for the removal of excess Cy3- and Cy5-dye labeled dNTPs from reactions using direct Cy-labeled nucleotide incorporation. The superparamagnetic particle-based technology is quick to use and affords minimal loss of product, making it ideal when sample size is limited. High cDNA recoveries also reduce the time and expense of multiple RNA preps when replicates are required and allow for the possibility of improved detection of rare message. The procedure can be rapidly optimized to eliminate excess dyes while maximizing product recovery, and can be automated.

K. Appasani¹, M. Tyler¹, A. cherkasskiy², J. J. Broadbent¹, L. Lacy¹; ¹PerkinElmer Life Sciences, 549 Albany Street, Boston, Massachusetts 02118, ²perkinelmer life sciences, 100 e. dedham, boston, ma 02118

RNA was isolated from HL-60 and Jurkat cell lines, and two masses (75 ug and 20 ug) of total RNA was used. The mRNA was isolated and converted to label cDNA. After labeling and hydrolysis using the Direct Labeling kit; both Cyanine-3 and Cyanine-5 reactions were purified and hybridized to MICROMAX 4800 cDNA arrays. Our results indicated in duplicate analyses where the input mRNA was 1.4 ug and 0.4 ug, we observed ratios of ratios (ROR) for differential genes to be 99% and 98.5% respectively. When we compared the mRNA (1.4 ug) and total RNA (75 ug) labeling in identical conditions, we found that ROR agreement for differential genes was 99%, where as, it was 96%, when we used mRNA (0.4 ug) and total RNA (20 ug). The top 200 differentially expressed genes were selected and grouped into different clusters. Such analysis, pathway identification and differential expression analysis will be described in detail. In conclusion, our results indicated that input of 20 ug total RNA or 0.4 ug mRNA is sufficient enough to perform microarray experiments, and the differential analysis is equivalent in both scenarios, except that image qualities were much better in the mRNA labeling.

J. Li, A. Viale; Memorial Sloan Kettering, New York, NY

Microarray technology has changed the way scientists study biological phenomena. Whether their goal is gene discovery or pathway analysis, scientists have access to different array platforms. The recent trend is to get more information from less tissue and to monitor gene expression profiling from biopsies. However, the amount of starting material is still an obstacle in many cases. The processing of samples for the Affymetrix genechip technology involves an in vitro transcription step during which biotinylated nucleotides are incorporated into cRNA. Several biotechnology companies have taken up the challenge of labeling small amounts of total RNA samples (between 100ng and 10μg ). This study is designed to evaluate the reliability of one labeling kit in particular, the Ambion MesAmp aRNA kit. This kit permits a small amount of total RNA sample (1μg with one amplification or 100 ng with two rounds of amplification) to be labeled in comparison to the 10 micrograms used in the Invitrogen cDNA synthesis kit. After hybridization on the U74Av2.0 array, fold changes between the three conditions are generated using the M.A.S 4.0 algorithm (Affymetrix) and statistical analyses are performed.

K. L. Knudtson¹, C. Griffin², A. Brooks³, A. Iacobas⁴, K. Johnson⁵, G. Khitrov⁶, K. Lilley⁷, A. Massimi⁴, A. Viale⁸, W. Zhang⁴, Y. Bao⁹, G. Grills¹⁰, H. Thaler⁸, C. Cox³; ¹University of Iowa, 323 EMRB, Iowa City, Iowa 52242, ²UCSF GCRC/SFGH, ³U. of Rochester, ⁴AECOM, ⁵Jackson Labs, ⁶The Rockefeller U., ⁷University of Cambridge, Department of Biochemistry, Cambridge, CB21QW United Kingdom, ⁸MSKCC, ⁹U. of Virginia, ¹⁰Harvard U.

Users of DNA microarray technology must overcome many challenges to obtain results in which the variability of the data is due solely to biological activity. It is important to be aware of and account for potential sources of variability in the experimental design and results. The goal of this study is to identify non-biological factors that contribute to variation in microarray results. A retrospective study, using data collected by ABRF MARG members in their core labs, was conducted using both Affymetrix GeneChip and spotted microarray technologies. The GeneChip study analyzed the metrics and meta data from a total of over 900 murine U74A, human U95A, and Test 2/3 arrays. The spotted microarray study analyzed the effects of slide fabrication, hybridization and scanner settings from over 100 slides in which the same reference RNA was used. The effect of tissue type, array lot, hybridization procedure and scanner settings on the variability of results was investigated for both types of microarray technologies. The results of these studies provide insight on potential sources of experimental error with microarray technologies and suggest experimental strategies to correct them.

D. Spicer¹, L. R. Booth¹, D. T. Campogan¹, S. T. Clary², A. S. Gall¹, K. A. Hughes², R. J. Kaiser¹, K. P. Lund²; ¹Prolinx, Inc., Bothell, WA, ²Prolinx Inc., Bothell, WA

Proteomics, the systematic study of protein structure and function, is a natural extension of genomics. As with genomics, proteomics requires a high throughput platform that enables screening of multiple proteins in a single assay. Protein microarrays holds promise as such a platform. However, due to the diversity and fragile nature of proteins, tools and technologies for DNA arrays have found little success when applied to protein arrays. Prolinx has developed a protein array platform based on its Versalinx Chemical Affinity Tools. The platform is comprised of 3X1" glass slides with a 3D structure of salicylhydroxamic acid (SHA), and protein conjugation reagents based on phenyl(di)boronic acid (P(D)BA). P(D)BA is conjugated to proteins with little to no affect on protein structure or activity. Any P(D)BA-protein can then be spotted directly onto the universal SHA substrate without purifying the conjugates or blocking the slide. Furthermore, because protein conjugation is done in solution, separate from immobilization, automation of the conjugation reaction is possible. Results will be presented demonstrating automated optimization of protein conjugates for array preparation.

L. R. Booth¹, D. T. Campogan¹, S. T. Clary², A. S. Gall¹, K. A. Hughes², R. J. Kaiser¹, K. P. Lund², D. Spicer¹; ¹Prolinx, Inc., 22322 20th Ave NE, Bothell, WA 98021, ²Prolinx Inc., Bothell, WA

The success of nucleic acid microarrays has created a demand for similar tools for proteomics. The transfer of tools and formats utilized in nucleic acid microarrays, however, has proven difficult. Issues inherent in the fundamental nature of proteins-maintenance of functional 3D structures and the optimal presentation of active regions-must be addressed by the selected protein immobilization strategy. Prolinx has developed a protein microarray platform utilizing Versalinx Chemical Affinity Tools that help to overcome these challenges. The tools are based on a small molecule chemical affinity pair-phenyl(di)boronic acid and salicylhydroxamic acid-which allow for immobilization of macromolecules on solid phases. The technology, when applied to protein arrays, provides high capacity surfaces that require no blocking but afford low background. Protein conjugation uses low MW reagents minimizing the effect on protein structure and biological activity. Purification of conjugates is not required, saving time and allowing for automated array preparation. Data will demonstrate the effectiveness of this platform for protein arrays, as assessed by the quality of arrays produced using this technology.

A. A. Karavanov; Ciphergen Biosystems, Inc., 6611 Dumbarton Circle, Fremont, CA 94555

SELDI (Surface Enhanced Laser Desorption Ionization) technology is a combination of a variety of active chromatographic and affinity surfaces with LDI-TOF MS detection.It,also,allows protein-protein interactions analysis in complex biological mixes.The objective of this study was to test a possibility of a specific receptor capture from mouse and human brain and human fibroblasts membrane preparations on two phospholipase A2 ligands covalently attached to preactivated surfaces (PS 20) of ProteinChip® arrays.The specificity of interaction was verified by on-chip competition experiments and by SELDI analysis of semi-purified preparations of known receptors.Clear specific capture of two known receptors from mouse and human brain, and human fibroblasts membrane extracts was demonstrated.High MW and high degree of glycosylation make these type of proteins difficult targets for MS detection. We demonstrate that 1 pm of 180 kDa heavily glycosylated recombinant receptor in semi-purified form, including its 360 kDa dimer, could be easily detected using H4(C16 resin equivalent) ProteinChip array.Protein and membrane samples for this study were provided by Dr. Gerard Lambeau, IPMC,CNRS(France).

K. A. Hughes; Prolinx Inc., Bothell, WA

Proteomics requires the development of new technologies that will enable the complete characterization and understanding of the myriad of protein structures and functions present in the human proteome. One emerging technology that holds significant promise for the high throughput analysis of protein-ligand interactions is protein microarrays. Generically similar to the nucleic acid microarrays now commonly employed for functional genomics, protein microarray technology awaits the successful development of a platform chemistry enabling the efficient, robust and flexible attachment of proteins to the array substrate. Issues inherent in the fundamental nature of proteins - maintenance of functional three-dimensional structure and optimal presentation of active regions - must be addressed by the selected protein immobilization strategy. Chemical affinity surface modification based upon rationally designed, synthetic small molecules is novel approach to the preparation of protein microarrays. Strategies for producing microarray substrates based on this approach, along with data assessing the quality of protein microarrays produced utilizing these substrates, will be presented.

M. G. Pluskal¹, M. Lopez¹, M. Robinson¹, S. Gutierrez², A. M. Pitt²; ¹Proteome Systems, 14 Gill Street, Woburn, MA 01801, ²Millipore Corporation, Danvers, MA

Proteomic analysis of complex samples, such as plasma or serum is frequently influenced by the presence of high abundance proteins such as, albumin or immunoglobulin. These proteins can lead to loss of resolution in two-dimensional electrophoresis (2DE) and chromatographic separations or excessive background noise in LC-MS/MS protein analysis methods. As a result, sample complexity reduction to lower the level of these abundant proteins is rapidly becoming an essential first step of any high throughput analysis approach. Data will be presented to show applications of affinity ligands for removal of abundant proteins from serum or plasma. The impact of this approach for removal of albumin and immunoglobulin is assessed using an integrated Proteome analysis technology platform, employing 2DE analysis. Removal of these two major components was shown to considerably improve the detection of low abundance proteins in the resulting 2DE gels. Alternative separation and purification techniques were also investigated. The sample complexity reduction technology described in this presentation has resulted in a kit employing small volume centrifugal devices. This kit format provides a convenient and efficient process for removal of abundant proteins from a complex sample, such as plasma or serum.

K. G. Mawuenyega¹, Y. Yamauchi², T. Shinkawa², M. Taoka¹, H. Kaji¹, T. Isobe¹; ¹Tokyo Metropolitan University, Department of Chemistry, Graduate of Science, 1-1 Minami Osawa, Hachioji, Tokyo, 192-0397 Japan, ²Integrated Proteomics System Project, MEXT, Tokyo, Japan

The complete genome of C. elegans provides a unique opportunity for a comprehensive assessment of the naturally occurring proteins, in order to substantiate the existence of predicted proteins. Both humans and C. elegans share features of most proteins regulating the mechanisms of their biological activities. Focused proteomes of native-acidic, -basic and membrane functional complexes were analyzed using μ2DLC-MS/MS techniques. Current achievements in both chromatography separations and MS detection offer better accessibility to all classes of proteins, and makes μ2DLC-MS/MS process amenable to automation at enhanced speeds. This analysis is a fully integrated approach to non-gel based protein profiling and identification. The identification performance using the system was 600-800 proteins per analysis. In the membrane proteome, 800 proteins were identified, out of which 205 were integral and 240 were peripheral. Data will be presented on proteins identified as well as details of analytical methodologies.

B. H. Glatz, R. Schuster, P. Goodley; Agilent Technologies, Hewlett Packard Str. 8, Waldbronn, BW 76337 Germany

For the structure elucidation of proteins in complex mixtures it has been shown that 2D-LC in combination with Ion Trap mass spectrometry analysis significantly increases protein identification capacity. Nano-LC in combination with Nanoelectrospray MS/MS allows reliable detection of fmol to amol amounts of peptides from enzymatic protein digests.

The bottlenecks of these techniques are high operational skills and the general robustness of the system, e.g.reliable gradient delivery of flow rates in the 100nl/min range, combined with small delay volumes for true gradient profiles, low dispersion connections and the overall automation capability of the complete system are essential for a high performance yet robust system. MS with data-dependent automatic scan functions including Auto-MS⁵, ActiveExclusion and MaxResScan, as well as a fully automated protein database search (MASCOT) complete the identification of enzymatic peptide fragments and the overall protein identification with high sequence coverage. Altogether we describe a concept/system, which combines the robustness of classical HPLC with the sensitivity impovements of Nano-LC and Nano-Electrospray Ion Trap mass spectrometry.

W. Kopaciewicz¹, E. Chernokalskaya², A. Dedeo², D. Brewster²; ¹Millipore, 17 Cherry Hill Drive, Danvers, MA 01923, ²Millipore Corp., 17 Cherry Hill Dr., Danvers, MA 01923

Reversed phase (RP) chromatography has long been the most versatile method for concentrating, fractionating and desalting peptides during protein structural characterization. Of the many types of hydrophobic phases available, C18 is used most frequently in both HPLC and sample preparation devices. C18 has the features of rapid equilibration, strong adsorption of hydrophobic solutes and desorption with volatile mobile phases. With such favorable attributes, RP is used in the vast majority of peptide separations. However, it is not without limitations. Namely, hydrophilic peptides do not adsorb and the stationary phase must be pre-wetted with organic for efficient adsorption.

In addition to RP, strong cation exchange (SCX) chromatography has been successfully used for the separation of peptides and amino acids. Since every peptide has at least one positive change (+2 for tryptic peptides, the negatively charged surface is relatively universal for polypeptide capture at acidic pH. We discuss, in this poster, the use of SCX resin immobilized in a pipette tip (ZipTipSCX) for the sample preparation of tryptic peptides prior to MALDI-TOF MS using dilute alcoholic NH4OH as the eluant. The tryptic peptide coverage of several proteins was compared to ZipTipC18 at the 100 fmol level. ZipTipSCX compared favorably to C18 and was more effective at capturing lower molecular weight (hydrophilic) peptides. The combined peak lists of both methods provides additional peptides for improved data base searches.

I. E. Jordon-Thaden¹, R. T. Hilger¹, R. L. Cerny²; ¹UNL, Hamilton Hall, Lincoln, NE 68588-0304, ²Mass Spectrometry Core Facility, Lincoln, NE

Accurate and reproducible results form the foundation of a modern high throughput proteomics laboratory with robotic systems. With a 96 well format, robotic systems in our lab are an in-gel digestion system, a robotic sample spotter, and automated MALDI-TOF data acquisition. The performance of the in-gel digester was evaluated by using 96 standard gel pieces containing 10pmols of BSA. The standard digestion protocol (washing, reduction, alklylation, digestion, and extraction) was followed by MALDI analysis. Uniform spotting of the 96 well MALDI plate was explored. Data were acquired in an automated mode using a close external calibration. The masses of the peptides were compared to the theoretical values and basic statistical tests were done to determine the consistency of the results. Our results showed a 25-60ppm margin of error for the major BSA peptide masses from the tryptic digest on all 96 wells of the entire system.

W. S. Hancock; ThermoFinnigan Corp., 355 River Oaks Parkway, San Jose, CA 95134

This proteomics studies presentation is divided into two approaches: global and focused proteomics. This division is helpful in understanding the applied technology to such complex challenges as the characterization of human plasma. Human plasma offers important opportunities both in the development of suitable proteomic tools as well as yielding greater information about human physiology.

De novo characterization of protein microheterogeneity is required in disease studies. To characterize modifications it is necessary to digest the protein mixture to produce the corresponding peptide mixture. Shot-gun sequencing of the proteome is based on the characterization of the key fragments produced by such a digest. Thus, a glycopeptide and hence a specific glycosylation motif will be identified by a unique mass and then a diagnostic MS/MS spectrum.

This presentation describes new instrumentation based on capillary-based separation systems. This approach can be used for stand-alone shot-gun sequencing or integrated with 2D-gel proteomic studies e.g., identification of low level proteins present in plasma and the characterization of complex glycosylation by the use of MS to the 5th on the fly.

D. Wall¹, J. W. Finch¹, S. A. Cohen¹, R. Karol¹, S. J. Berger¹, J. Brown², D. Gostic³; ¹Waters Corporation, 34 Maple Street, Milord, MA 01757, ²Micromass, Floats Road, Wythenshawe, Manchester, M23 9LZ United Kingdom, ³MicromassUK Ltd, Manchester, UK United Kingdom

This paper investigates the advantages for MALDI that result from the continuous deposition of RP HPLC eluent onto MALDI targets using an LC-MALDI interface. Peptides are derived from standard and ribosomal protein mixture tryptic digests. Samples vary in complexity from 4 standard proteins to purified yeast large ribosomal subunits. The 4 standard proteins are mixed in different amounts to determine how well low abundance proteins can be detected in the presence of high abundance proteins. Purified yeast large ribosomal subunits are analyzed as total digests and as band digests from 1D-SDS PAGE. All digests are then separated by RP HPLC and collected onto MALDI targets as continuous tracks using the LC-MALDI interface. Tracks are analyzed using MALDI-TOF MS and MALDI-QTOF MS/MS. Peptide mass mapping and sequencing results are submitted to a database search to determine protein identity. Peptide mass mapping resulting from the LC-MALDI work is compared to that obtained by traditional sample loading methods.

M. Hincapie, A. J. Tomlinson, S. Kwak, N. Laliberte, J. A. Corbo, J. R. Lill, G. E. Morris, R. Chicz; Zycos, Inc, 44 Hartwell Ave, Lexington, MA 02421

TransAtlas is the catalog of translated proteins from a cell or tissue source. This database is created using novel methods that characterize the protein content of cells or tissues samples.

One approach relies on a highly conserved relationship between a cell’s proteolytic processes and the different multi-ligand receptor (MLR) systems; cells display a profile of peptides representing expressed proteins on their surface. Such Expressed Protein Tags (EPTs) are of sufficient length to differentiate particular proteins and/or genes. In a second approach in vitro global protein digests are produced from cell fractionation preparations. The peptide repertoires are termed Global Protein Tags (GPTs).

We have developed a fast micro-2-D HPLC fractionation method for rapid analysis of complex proteome mixtures. Peptides are separated using ion-exchange chromatography followed by microbore reverse phase HPLC prior to analysis by LC/MS/MS. This protocol has been employed to identify over 1,000 proteins from an example gastric carcinoma. Bioinformatic approaches have been used to prioritize a therapeutic target list for accelerated target validation.

E. Chernokalskaya, A. Dedeo, D. Brewster, B. Kopaciewicz; Millipore Corp., 17 Cherry Hill Dr., Danvers, MA 01923

The affinity interaction between Ni++ ions and histidine residues is well–known. Nickel ion chelated chromatography resins have proven to be effective in capturing recombinant proteins that have a ca. 4 – 8X histidines grafted onto the N- or C-terminus. This system is very convenient as the result of the wide variety of plasmid constructs and chelated nickel chromatography products commercially available. With these products, DNA inserts can easily be cloned, expressed and analyzed.

The Ni++/His tag system is becoming increasingly important in proteomics for expressing/isolating low abundant proteins and generating sufficient quantities of proteins for interaction studies and structure analysis.

We report in this poster on the use of a 10 ul pipette tip containing ca. a 0.5 ul of Ni++ chelated resin immobilized at the distal end (ZipTipNi). This device was effective in isolating 10 – 100 ng of native or denatured His-tagged protein from ca. 10 to 50 ul of cleared cell lysate in about 5 minutes. The amount of protein recovered was sufficient for detection in a mini-gel using colloidal Coomassie stain. Alternatively, eluted protein can be directly spotted onto a target for MALDI TOF MS analysis. For more precise analysis, the adsorbed protein can be digested with trypsin in situ generating peptides for fingerprint analysis by MALDI TOF MS and data base search. The pipette tip format offers a convenient sized tool for the rapid characterization of expressed His-tagged clones from only a small sample of cell lysate.

I. Mohtashemi, A. Kishiyama, E. A. Luis, L. Gonzalez, J. T. Stults, D. Arnott; Genentech, Inc., South San Francisco, CA

Technologies such as 2D electrophoresis or LC-MS/MS of digested protein mixtures have been used to analyze complex protein expression patterns, but only a fraction of a cell's proteins are detectable in a single experiment. It is therefore desirable to reduce the complexity of protein mixtures to identify more, and lower abundance proteins. Fractionation of sub-cellular organelles is one such approach. Plasma membrane proteins are of particular interest because of their roles in signal transduction, cell-cell interactions, and as targets of therapy. Although many protocols are described in the literature, isolation of plasma membranes from the other cellular membranes is considered difficult. Three methods for plasma membrane isolation were evaluated: a crude membrane preparation, sucrose layer sedimentation, and affinity two-phase partitioning. Each was characterized for enrichment and recovery of specific plasma membrane markers. A more comprehensive view was obtained by solubilization and digestion of the membrane fractions, and identification of as many proteins as possible by LC-MS/MS. The merits of each protocol will be discussed as well as their suitability for specific experiments.

B. L. Williamson¹, S. G. Honholt¹, D. K. Han², L. M. Nuwaysir¹; ¹Applied Biosystems, 500 Old Connecticut Path, Framingham, MA 01701, ²University of Connecticut Health Center, Farmington, CT

The use of isotope coded affinity tags for the identification and quantification of differentially expressed proteins was first reported back in 1999. In this approach proteins are first covalently tagged with the ICAT reagents, followed by proteolysis of the combined-labeled protein samples. Peptides can be separated into 30-50 fractions by IX. The resulting fractions are then purified on an avidin column to isolate the ICAT reagent labeled peptides prior to analysis by cLC MS/MS for quantitation and identification information. As the number of peptide fractions increase there is a clear need to automate these chromatographic steps on a single HPLC system and develop software to interpret the data generated.

We have developed an improved method for the ICAT reagent technology which uses complete automation of the chromatography steps, coupled with fully automated software processing of the data generated. In this method, crude protein digests are injected onto an ion-exchange column followed by gradient elution of the peptides into as many as 50 fractions. Each of these fractions is then automatically passed over an avidin purification column to isolate the cysteine labeled peptides. In turn each of these fractions are analyzed by nano-LC MSMS and a new software program is used to automatically quantify and identify the differentially expressed proteins. To illustrate the potential of this method, proteins from microsomal fractions of human myeloid leukemia cells were identified and quantified using this procedure.

Z. Smilansky; Compugen, 72 Pinchas Rosen, Tel Aviv, 69512 Israel

Two dimensional polyacrylamide gel electrophoresis (2D-PAGE) is currently the method of choice for separating complex protein mixtures. Large-scale studies, however, typically consist of hundreds of gels, and it becomes impractical to analyze the vast number of gel images using traditional computer programs. In addition, large-scale experiments call for statistical analysis, aiming to detect proteins whose expressions show interesting dependence on the controlled experimental parameters. Compugen’s Z4000 system will implement a computational model that will allow the user to organize and analyze a large multi dimensional gel collection. The system will provide raw data for statistical analysis, and support visual verification of the analysis results. We propose a new scalable computational model, in which only a few gels are analyzed at a time. The analyzed sets of gels form processing units, which in turn can be combined into higher-level units, and so on in a hierarchical manner. A natural hierarchical scheme will group together gels that share common experimental attributes. For example, in an experiment involving normal and mutated species, the basic hierarchical level naturally corresponds to the different species, while the next level naturally corresponds to the dimension of normality versus mutation.

H. Antelmann¹, B. S. Miller², W. Weyler², B. Fryksdale², G. Chotani², E. Ferrari², A. Gaertner², M. Hecker¹; ¹Institut fur Mikrobiologie und Molekularbiologie, Ernst-Moritz-Arndt-Universitat Greifswald, Greifswald, Germany, ²Genencor International Inc., 925 Page Mill Road, Palo Alto, CA 94304

The intracellular and extracellular proteome of a gram-positive, spore-forming bacterium Bacillus subtilis was profiled using two-dimensional electrophoresis followed by protein identification using mass spectrometry. More than 300 proteins in both the intra- and extracellular compartments have been identified and their relative induction and repression levels were determined across batch culture time courses. Comparisons were also made between strains bearing pleiotropic mutations that affect sporulation and other late growth phase phenotypes. It was found that these mutations, including scoC4, affected the levels of a number of proteins involved in diverse cellular functions.

K. L. Foster, J. M. Zobrist, L. K. Barbacci, J. G. Dapron, T. C. Hassell, W. K. Kappel; Sigma-Aldrich Biotechnology, P.O. Box 14508, Saint Louis, MO 63178

With the recent growth in the field of proteomics, there is a subsequent need for high throughput characterization of proteins and peptides. Some techniques involve biotinylation of proteins at specific residues such as attachment of isotope coded affinity tags (ICAT). Following digestion, these procedures rely on capture and recovery of biotinylated peptides for analysis by mass spectrometry (MS). To accelerate this process, unique high capacity (HC) streptavidin (SA) coated 96-well plates were utilized in a manner compatible with analysis by MS. In a model system, lysozyme was denatured, reduced, and biotinylated. The protein was tryptically digested and fractionated using ion exchange HPLC. The fractions were directly collected into the wells of a HC SA plate, quickly enabling the purification of the biotinylated peptides. The results demonstrate that the biotinylated peptides were selectively captured. More importantly, the captured peptides were eluted in a sufficient quantity for detection using MS compatible reagents. The HC SA plate facilitates the high throughput purification of biotinylated peptides in significantly less time than current methods, such as monomeric avidin agarose.

E. claude¹, D. O. Gostick¹, A. A. Alaiya², G. Auer², J. I. LANGRIDGE¹; ¹Micromass UK Ltd, floats road, Wythenshaw, Manchester m23 9lz United Kingdom, ²Karolinska Hospital, Stockholm, Sweden

Cancer development and progression is a multi-step process in which control of cell growth is progressively perturbed due to either loss of tumor suppressor genes or activation of oncogenes. Colorectal cancer represents an ideal model system to study development and progression of human tumors. This is because the epithelial cells of the colon follow a systematic process of cellular proliferation, differentiation and adenoma – carcinoma transformation. The p53 gene is actively involved during the different stages of colo-rectal carcinogenesis. In this study, we have examined the role of p53 in a human colorectal cell line (HCT 116) a diploid cell line deficient in DNA mismatch repair mechanism.

To understand the role of p53 in the development of colon cancer, we have studied the protein expression profiles in HCT 116 cells with two wild type p53 alleles (p53+/+) and their p53 knockout counterparts (p53-/-). Representative 2D gel images were derived from the two samples and proteins were identified by automate MALDI-TOF-MS.

C. Hughes¹, J. I. LANGRIDGE², A. Millar³, M. Bailey⁴; ¹MICROMASS UK LTD, Floats Road, MANCHESTER, M23 9LZ United Kingdom, ²Micromass UK Ltd, FLOATS ROAD, MANCHESTER, M23 9LZ United Kingdom, ³Micromass UK Ltd., Floats Rd, Wythenshawe, Manchester M23 9LZ United Kingdom, ⁴Upchurch Scientific, Oak Harbor, WA

Any reversed phase LC-MS/MS run will naturally involve a column re-equilibration step after the gradient has been completed, where the initial solvent starting conditions are flushed through the column to prepare it for injection of the next sample. This consequently leads to a 'dead time' in the experiment during which no data can be acquired. The removal of this dead time would lead to a reduction in the total experiment time, especially when large numbers of samples are to be analysed. Here we describe an experimental arrangement where two sets of analytical and trapping columns are operated in parallel. This allows the time required for re-equilibration for one set of columns to be absorbed into the elution step for the opposite set of columns. The system uses a ternary solvent delivery system, a single ESI spray tip and two ten-port switching valves. One of these valves has an extremely low internal volume, which is essential for use post column to reduce the effect of post-column dead volume on chromatographic peak broadening. This arrangement allows two samples to be analysed in less than one hour while running a 25-minute gradient for each sample.

G. R. Asbury, S. Loyland; Amersham Biosciences, 800 Centennial Ave., Piscataway, NJ 08855-1327

Identification of proteins by peptide mass fingerprinting using a MALDI-TOF instrument has become routine in the past several years. This identification is done by comparing the measured masses of peptides from an enzymatic digestion (normally trypsin) to those predicted for all proteins contained within a database. This strategy is particularly useful for organisms that have complete protein databases such as saccharomyces cerevisiae, but has limited use for organism without a complete databases. In these cases, however it may be possible to gain information by searching peptides against expressed sequence tag (EST) databases. One requirement for searching EST databases is that sequence information must be obtained on a given peptide. This requirement has made the use of MALDI-TOF instruments for these searches to be of little use because of the difficulty to obtain fragmentation data using post source decay (PSD). This difficulty has, however been overcome by using a combination of chemical modifications and the use of a specially designed reflectron. Using this combination EST databases were interrogated using CAF-MALDI for tomato proteins.

P. Juhasz, B. Williamson, J. Marchese, A. Falick, A. Graber, N. Khainovski, A. Romeo, D. Patterson, S. Martin; Applied Biosystems, 500 Old Connecticut Path, Framingham, MA 01701

On-line LC ESI/MS and MS-MS methods for proteomics are likely to have their limitations with the increasing number of components to be quantified and identified in a single experiment. Newly developed MALDI tandem TOF instrumentation offers the off-line coupling of LC to MS as an attractive alternative to run complex experiments where data acquisition, processing, and interpretation can be freely combined to maximize the information obtained. Mixtures of standard proteins and cell lysates from two strains of H. ducreyi were analyzed. After reduction/alkylation (iodoacetamide or ICAT reagent), the digested mixtures were subjected to SCX-avidin-HPLC fractionation. HPLC fractions were collected directly on MALDI plates. For quantitative analysis the MALDI acquisition parameters were optimized to compensate for the spot-to-spot variations of the ion signal. MS-MS measurements utilized novel dynamic exclusion algorithms to maximize data quality and information content. To process very large sets of data, Oracle queries and report generating features have been utilized. In the highest level of data summary, information from every ICAT pairs from the same protein was folded in and evaluated statistically.

B. L. Allen¹, R. C. Jones¹, M. Smith², M. R. Pisano¹, L. H. Malkas², P. Wills², R. J. Hickey², R. D. Edmondson¹; ¹Proteomic Research Services, 4401 Varsity Drive, Suite E, Ann Arbor, MI 48103, ²University of Maryland, Baltimore, MD

Previous studies have isolated and partially characterized a fully functional multiprotein complex that mediates DNA synthesis designated the DNA synthesome. We recently performed initial experiments employing SDS-PAGE with mass spectrometry to enable further characterization of the complex and associated proteins. Here we are optimizing the analytical approach with respect to three major issues: sample complexity, dynamic range of protein quantity and sensitivity.

The utility of SDS-PAGE is tested when dealing with mixtures of high dynamic range in which the proteins of interest are at low levels, here the co-migration of multiple proteins to the same point on a gel can hinder their identification. Recent approaches utilize solution digestion with multi-dimensional LC (MDLC). We have exploited SDS-PAGE and MDLC both individually and in tandem. Peptides were analyzed by LC/MS/MS (Q-Tof2; with gel digests and/or SCX fractions in MDLC) using PicoFrit columns with fast gradients at 50 nL/min flow-rates; the use of a vented-column system allows fast load times with no sample losses and cycle times around 20 min (with 15-20 μL loaded).

W. H. Vensel, C. K. Tanaka, W. J. Hurkman; U. S. Department of Agriculture, 800 Buchanan Street, Albany, CA 94710

Environmental factors, during grain-fill, influence wheat flour quality and composition. Our experiments suggest that the most significant effect on grain development is high temperature. We are using proteomics to identify the metabolic pathways central to wheat grain development. Wheat (Triticum aestivum L. cv. Butte 86) plants were grown in a climate-controlled greenhouse. Grain was harvested at pre-selected time points and endosperm collected. Salt soluble proteins were extracted and separated by 2-D gel electrophoresis. Gel patterns for the different developmental stages were characterized using computer-based image analysis. Identification of the proteins from the 2-D gel spots required a number of different procedures. Proteins were identified using Western blots, Edman degradation, ESI and MALDI mass spectrometry. Peptide mass mapping allowed us to identify about only 30% of the gel spots. Proteins identified could be grouped into those involved in transport and metabolism of amino acids and carbohydrates, defense against predators, posttranslational modification, and transcription

D. Goodlett¹, J. Eng¹, S. Leicester², R. O'Malley², A. Millar², J. I. Langridge²; ¹The Institute for Sytems Biology, Seattle, Washington, ²Micromass UK Ltd., Wythenshawe, Manchester United Kingdom

Whilst a non-gel based strategy using direct analysis of tryptic digests by LC-MS/MS has been utilised for qualitative proteomics, until recently relative quantitation of protein expression levels from this data has proved problematical. Recently, however, a great deal of interest has been generated on the use of isotope coded affinity tags (ICAT) for the quantitative study of gene expression at the proteome level.

The technique is based upon chemically modifying the cysteine residues of proteins isolated from cells in two different states with light and heavy isotopically labeled reagents. The two cell states are then combined, digested with trypsin and the cys containing peptides purified by binding to an avidin column. Interrogation of the resulting peptides by capillary LC-ESI MS and MS/MS coupled with bioinformatics processing of the resulting data sets, provides both relative quantitative information and identification of the associated proteins contained within the sample.

In this paper we describe the analysis of ICAT labelled protein samples obtained from the microbe Halobacterium and describe new mass spectrometric methods for both data acquisition and processing of the resulting data.

H. Byers¹, J. Westbrook¹, A. Millar², J. Langridge³, M. Ward¹; ¹Proteome Sciences plc, De Crespigny Park, London, United Kingdom, ²Micromass UK Ltd., Floats Rd, Wythenshawe, Manchester M23 9LZ United Kingdom, ³MICROMASS UK LTD, Wythenshawe

Following heart transplantation, patients often suffer episodes of acute rejection, which can occur in the absence of clinical signs and symptoms. Currently, there are no non-invasive methods of monitoring these episodes and endomyocardial biopsies are often performed. Two-dimensional gel electrophoresis was used to visualise differential protein expression and biopsy proteins responding to episodes of acute rejection were identified by mass spectrometry.

Here we describe a set of complementary experiments involving the use of LC/MS/MS to provide supporting evidence to further validate the findings of the 2DE approach. Anti-heart antibody reactive proteins were detected in human heart left ventrical tissue using transplant recipient serum. Bands from a Coomassie Blue stained 1D gel of human heart left ventricle proteins, corresponding to immunoreactive bands of interest on Western blots, were excised and analysed by LC/MS/MS to identify potential antigens.

Taken together, these results suggest that, as a consequence of rejection, heart proteins circulate in the blood and this has lead to the development of a panel of non-invasive markers, which can be used to monitor this condition.

R. G. Biringer¹, M. Harrington², W. Stochaj², P. Bondarenko³, A. Huhmer¹, H. Amato¹, G. Chu¹, S. Swedberg⁴; ¹Thermo Finnigan, 355 River Oaks Parkway, San Jose, CA 95134, ²Huntington Medical Research Institutes, Pasadena, CA, ³ThermoFinnigan Corp., San Jose, CA, ⁴ThermoFinnigan, San Jose, CA

The combination of 2D-gels and ion-trap nanospray mass spectrometry is clearly a versatile methodology with significant impact in the evolving field of proteomics. It is an ideal combination with which to follow changes in protein expression and posttranslational modification that characterize disease states, a requirement for the development of therapeutics and diagnostics. A key step in this method involves the extraction of peptides from individual proteolyzed protein gel plugs obtained from 2D-gels. Rigorous characterization of the proteins of interest requires high sequence coverage which in turn requires efficient extraction, especially for low abundance proteins. For this purpose, we have developed an aqueous solvent-based methodology for production and subsequent extraction of trypsin-derived peptides from gel plugs. We have used it in conjunction with an LCQ ion trap spectrometer equipped with a nanoflow solvent delivery system and polymeric reversed-phase peptide trap to identify proteins and their isoforms from human cerebrospinal fluid (CSF). Total recovery of peptides from gels was considerably better than that obtained using standard methods.

Y. Moriwaki, T. Sasagawa, M. Kojima, A. Motoki, Y. Mizuno; Toray Res. Ctr., 1111,Tebiro, Kamakura, Kanagawa 248-8555 Japan

Excellent resolution and sensitivity make 2D gel electrophoresis (2DE), especially differential 2DE, indispensable technique in proteomics. Protein, however, can not be recovered from the gel and bio-active neither. Alternatively, 2D chromatography based on ion exchange and reversed phase has been used. Unfortunately, the reversed phase chromatography is not necessary good choice, because of poor recovery and of the fact that protein is recovered in denatured state. In order to get around these problems, we have developed 2D chromatographic system. The first dimensional chromato-focusing is the emulation of isoelectric focusing of 2DE. The second dimensional MALDI-TOF MS, which is also the emulation of SDS-PAGE, is used for high sensitive monitoring of the separation. Unresolved protein is further subjected to chromatography of different elution mode, e.g. size exclusion, metal chelate affinity chromatography and so on. As expected, by the combination of these two techniques, 2D chromatogram was obtained which bears close resemblance to 2D electrophoregram. With the assistance of initial differential 2DE, target protein is easily purified by this system. We show several successful examples.

S. Guertin; Applied Biosystems, 500 Old Connecticut Path, Framingham, MA 01701

One of the major difficulties in analyzing the proteome of human serum is the dynamic range of protein concentrations present in the sample. Often the presence of the highly abundant proteins such as albumin (HSA) and immunoglobulin (IgG) interfere with the analysis of the less abundant proteins. In this poster, we will describe the performance of immunoaffinity chromatography cartridges which we have designed to remove these two high abundant proteins from human serum. These cartridges have been characterized with respect to their capacity, lifetime, carryover, non-specific binding and cross-reactivity. The results of our development work will be presented.

C. Shieh, R. Shen, M. Chen, T. Zhang, Y. Lu, W. S. Hancock; Thermofinnigan corp., 355 River Oaks Pky., San Jose, CA 95134

Two-dimensional LC-MS/MS method using packed tips with nanospray mass spectrometry was reported in the literature (Nature (1999) 17:676-682. We report here an automated capillary 2D-LC-MS/MS system using a strong cation exchange column and two reverse phase C18 column for the analysis of complex protein mixtures.

Tryptic peptide mixtures from cell lysates were loaded on a strong cation exchange column, then separated by a C18 column. The peptides were analyzed by a LCQTM Deca XP mass spectrometer. With dynamic exclusion, thousands of unique MS/MS spectra were obtained and hundreds of proteins were rapidly identified using TurboSEQUESTTM software. As compared to the current 2DE-MS method, this on-line 2D LC-LC-MS/MS system has the advantages of higher sensitivity, higher throughput and a higher degree of automation. We have analyzed protein mixtures from human plasma and cell lysates to demonstrate the sensitivity, robustness and throughput of the system.

R. Plumb¹, S. A. Cohen², H. Liu²; ¹Waters, 34 Maple street, Milford, MA 01757, ²Waters Corporation, 34 Maple Street, Milford, MA 01757

Protein identification in the proteomics arena is conventionally performed by 2D gel electrophoresis followed by LC/MS/MS. The use of 2D gels, whilst imparting significant resolving power to the experiment, has the draw back of irreversibly absorbing high and low pKa proteins. Those with low copy numbers are also poorly represented in 2D gels. In this paper we present the use of 2D chromatography coupled to time of flight mass spectrometry for the separation and identification of intact proteins from complex mixtures. We will show how the gradients have been optimized for resolution and peak fidelity. In this paper we will also discuss how the mobile phase of the final dimension has been balanced to give optimal performance for both the chromatographic separation and mass spectrometric response. Finally we will show how this approach has been employed for the direct analysis of proteins in real samples using complex mixtures such as whole ribosomal preparations. This approach has several advantages over the 2D gel method, namely it does not suffer from irreversible absorption, the detection process is on-line and the process can easily be scaled up to the preparative scale for drug discovery purposes.

W. R. Stochaj, M. Meys, P. T. Snell; Beyond Genomics, Inc., 40 Bear Hill Road, Waltham, MA 02451

With the higher demands on separation and sample preparation techniques required to enable proteomic analysis of complex organisms, a need for a high throughput method of protein digestion has developed.

The need to analyze increasingly complex sample matrices with greater selectivity and sensitivity has begun to be filled by the application of multidimensional liquid chromatography (MDLC) coupled with peptide analysis by LC/MS or peptide mass fingerprinting by MALDI-TOF MS. These analytical techniques benefit from a method of in-solution or on-column reduction, denaturation and digestion of proteins that is amenable to high throughput, fully automated processing.

Although a number of digestion methods have been detailed in the literature, a direct comparison of these methods for simple and complex protein mixtures has not been described. This work compares standard urea denaturation followed by in-solution tryptic digestion with two alternative methods: 1) Thermal denaturation and in-solution tryptic digestion and 2) On-column immobilized enzyme digestion. Advantages and limitations of each method are discussed.

M. Chen, R. Shen, P. Shieh, W. S. Hancock; ThermoFinnigan, 355 River Oaks PKY, san Jose, CA 95134

Large scale protein identification (shotgun sequencing) using peptide mass maps has become an important technique for the proteomic study. Peptide fragments from protein digested are separated and identified by a capillary LC/MS/MS system. Comparing the measured fragment ions of the peptides to those predicated by the sequence can elucidate the corresponding protein identification through database search. In this paper, we reported an optimized LC/MS/MS condition for the analysis of complex peptides mixture. Proteins from cell A431 lysate are analyzed under different gradient conditions and different mass dependent search. Results showing longer gradient and more MS/MS spectra increase the # of proteins identified by capillary LC/MS/MS.

V. S. Asirvatham, B. Watson, L. W. Sumner; The Noble Foundation, 2510 Sam Noble Parkway, Ardmore, OK 73401

Proteomics is being incorporated as part of an integrated functional genomics study of Medicago truncatula. This report will provide an overview of our program and the technical approaches being employed. For example, analytical (gel-to-gel) and biological (plant-to-plant) variance associated with the 2-DE profiling of leaf proteins has been quantified. The implications of these variance studies on interpretation of protein expression level changes in large proteomics projects will be discussed. Multiple searching strategies employing protein and nucleotide databases including ESTs will be compared and associated success rates summarized. Finally, an overview of the systematic identification of a large number of tissue specific proteins from Medicago truncatula will be presented. This data is being deposited into a new internet-based 2D PAGE database to assimilate, store and share protein data of plants, particularly Medicago truncatula.

O. Chertov¹, D. Yang², Q. Chen², I. Mirkina², J. Wooters³, J. Oppenheim²; ¹SAIC Frederick contract DHHS N01-CO-12400, Bldg 560, rm 31-19, NCI-FCRF, Frederick, Maryland 21702, ²National Cancer Institute at Frederick, Frederick, Maryland, ³Genetics Institute, Cambridge, Massachusetts

The acute stage of inflammation as typified by neutrophil infiltration is followed by mononuclear cell infiltration. This infiltration may be neutrophil dependent. Earlier we identified some of chemotactic proteins from neutrophils using Edman sequencing. Because we were unable to purify other chemotactic proteins in quantity sufficient for sequence analysis, we used mass-spectrometry for their identification. We fractionated the neutrophil granule extract by RP-HPLC. The fractions were separated by SDS-PAGE and stained. The bands were excised and digested with trypsin. The digest was analyzed using Finnigan LCQ mass spectrometer. The chemotactically active fraction contained peptide LL-37, which was chemotactic. The perusal of other proteins yielded interesting information. A protein from respiratory syncytia virus was detected implying that the donor was infected with the virus. The combination of RP-HPLC and PAGE presents a number of advantages for analysis of protein mixtures: a) RP-HPLC has very high resolution and yields fractions that do not require any special manipulation before biologic testing and SDS-PAGE, b) proteins with very different dynamic range of concentrations can be detected.

G. Choudhary, S. Wu, P. Shieh, W. S. Hancock; ThermoFinnigan Corp., 355 River Oaks Parkway, San Jose, CA 95134

Protein identification using peptide mass maps has become an important technique for the proteomic study. The marriage of electrospray ionization and tandem mass spectrometry has led to widespread use of these techniques. By digesting the protein with a site-specific enzyme (e.g. trypsin) and comparing the measured fragment ions of the peptides (e.g. in tandem MS or ion trap) to those predicated by the sequence one can identify the corresponding protein through database search. The strategy of multiple enzymatic digestion is frequently used to achieve complete sequence coverage for recombinant proteins. In this poster we present data on advantages of employing multiple enzymatic digests for complex proteomics mixtures. Human plasma was digested with three separate enzymes (Trypsin, Lys-C and Asp-N) and analyzed using capillary LC-MS/MS. Approximately 200-300 proteins were analyzed for every run. A combination of the data obtained from the above three enzymatic digests results in an increased protein sequence coverage that enhances the confidence in identification of proteins comprising this complex mixture.

C. J. Rudd¹, J. De², D. Houck², H. Amato¹, R. G. Biringer¹, D. Recktenwald², S. Swedberg³; ¹Thermo Finnigan, 355 River Oaks Parkway, San Jose, CA 95134-1991, ²BD Biosciences, San Jose, CA, ³ThermoFinnigan, San Jose, CA

A goal of proteomics is the ability to detect all proteins in a particular biological subsystem, such as a cell type or cellular subfraction. Many cell surface proteins are integral membrane proteins of low abundance, therefore are not easily detected using standard proteomics methods. However, liquid chromatography (LC) with tandem mass spectrometry (MS/MS) is proving to be a sensitive method for identification of proteins. Fluorescence-activated cell sorting (FACS) was used to rapidly purify and characterize up to 15 million normal CD4+ and CD8+ human lymphocytes from peripheral blood. Proteins in purified plasma membranes from both normal cells and cell lines were identified from their tryptic peptides using nanoflow LC/MS/MS. The multiple peptide fragments were separated, analyzed and linked to their parent proteins using TurboSequest® software to cross-correlate the MS/MS spectra of individual peptides with computer-generated MS/MS spectra calculated for tryptic fragments of proteins in the human database. The database searches resulted in the identification of many cell surface and membrane-associated proteins, including lymphocyte-specific CD proteins.

B. S. Phinney¹, K. Wilkerson², J. E. Froehlich², R. McAndrew³, D. A. Gage¹; ¹Michigan State University, Room 11 Biochemistry, E. Lansing, MI 48824, ²DOE-Plant Research Laboratory, Michigan State University, E. Lansing, MI, ³Department of Plant Biology Michigan State University, E. Lansing, MI

With the completion of the Arabidopsis genome and with the rapid increase in the amount of other plant genome and EST data, plant proteomics is rapidly becoming a very active field. We have pursued a high-throughput mass spectrometry based proteomics approach to identify and characterize membrane proteins localized to either the pea or Arabidopsis thaliana chloroplast envelope membranes. Because membrane proteins typically resolve and behave poorly using traditional two dimensional gel electrophoresis, we instead relied on one-dimensional electrophoresis followed by in gel digestion and automated nano-flow LC/MS/MS to identify the membrane proteins in the chloroplast. We will compare this technique in terms of total dynamic range and numbers of membrane proteins identified to protease digestion of crude chloroplast membranes followed by automated multi-dimensional chromatography-nano-flow mass spectrometry. The preliminary results from this high throughput proteomic study of the chloroplast envelope membranes will be presented. To date no successful systematic analysis of the envelope proteome has been carried out.

K. Kubota¹, C. Sakikawa¹, M. Katsumata¹, T. Nakamura², K. Wakabayashi¹; ¹Sankyo Co., Ltd., 2-58, Hiromachi 1-Chome, Shinagawa-ku, Tokyo 140-8710 Japan, ²RIKEN, Wako, Tokyo Japan

Osteoclasts and osteoblasts are responsible for strict bone maintenance with a balance between bone formation and resorption by interacting with each other. Recently, it has been revealed that osteoblasts regulate differentiation of osteoclasts by two factors, RANKL on the plasma membrane and secreted OPG. However, no factors have yet been reported by which osteoclasts regulate osteoblasts. To elucidate the possibility of signal transduction from osteoclasts to osteoblasts, we studied the conditioned medium of mouse osteoclast-like myeloma cell line RAW264.7 treated with RANKL. We found that this medium contains a factor that inhibits differentiation of mouse osteoblast precursor-like cell line MC3T3-E1 to osteoblasts induced by BMP-4, and named this factor osteoblastogenesis inhibitory factor (OBIF). OBIF was purified by successive three-step chromatography by heparin affinity, anion exchange, and reverse-phase columns. OBIF activity made one peak in each chromatography step, showing the factor is a single entity. Active fractions were loaded on SDS-PAGE and bands of proteins were excised, digested by trypsin, and analyzed by LC/MS/MS. As a result, we have identified this factor to be platelet-derived growth factor BB homodimer (PDGF BB). Furthermore, this identification of PDGF BB as OBIF was confirmed by neutralization of the inhibitory activity of the medium with anti-PDGF antibody. These results demonstrate, for the first time, that osteoclasts regulate osteoblasts directly and suggest that PDGF BB is a key factor in bone remodeling.

J. Brown¹, D. O. Gostick²; ¹Micromass, Floats Road, Wythenshawe, Manchester, M23 9LZ United Kingdom, ²Micromass UK Ltd, Wythenshawe, Manchester United Kingdom

Axial MALDI-TOF mass spectrometer data may require a small correction to the mass scale to compensate for ion desorption velocities, gap changes between the sample and the source lenses, crystal structure of the sample/matrix, and other thermal effects. Addition of an internal reference at a similar concentration to the analyte can provide mass accuracy better than 10ppm RMS. As analyte concentrations are not always known this becomes impractical for large sample sets. Correction may also be achieved using a reference nearby on the sample plate, this provides mass accuracy better than 50ppm RMS. An algorithm has been developed that adjusts a calibration factor to minimise the overall difference between the expected and observed mass sufficiency for "typical" peptides [1]. It also filters the data for noise peaks or non-typical peptides (such as glycosylated species). Data from standards and from 96 mouse protein digests are tested and the effect of noise peaks and the presence of modified peptides is discussed. Differences between the theoretical and matched peptide masses are better than 20ppm RMS.

[1] Mann, Possible peptide masses. 43rd ASMS. 1995

J. Brown¹, D. O. Gostick², E. claude², J. I. Langridge³, K. Howes², W. Chen⁴, E. Bouvier⁴, P. J. Lee⁴, J. C. Gebler⁴; ¹Micromass, Floats Road, Wythenshawe, Manchester, M23 9LZ United Kingdom, ²Micromass UK Ltd, Wythenshawe, Manchester United Kingdom, ³Micromass UK Ltd., Floats Road, Wythenshawe, Manchester M23 9LZ United Kingdom, ⁴Waters Corporation, Milford, MA

This work presents a new MALDI sample plate designed for direct preparation of biological samples. Significant concentration of very dilute samples and subsequent removal of inorganic salt contamination is performed in-situ on the MALDI plate. The sample preparation of protein digests resulting from in gel digestion of 2D electrophoresis gel spots are evaluated. Data is presented showing that the sensitivity is significantly improved for faint Coomassie blue or silver stained gel pieces. The method is compared with alternative methods of sample preparation and its amenability for higher throughput automation is demonstrated.

K. Nugent, L. A. Upton, C. Stine, H. Wang; Michrom BioResources, Inc., 1945 Industrial Drive, Auburn, CA 95603

Now that the human genome has been sequenced, attention is shifting to proteomics and the role that proteins play in the structure, function and control of biological systems. Although 2D PAGE is the most widely accepted technique for separation of complex protein mixtures, it has several drawbacks including being biased towards highly abundant proteins, selective in protein recovery, labor intensive and difficult to automate. With recent developments in nanoLC-MS/MS, proteins in simple mixtures can be easily identified at the femtomole level. More complex protein mixtures (hundreds to thousands of proteins) still require some degree of fractionation prior to nanoLC-MS/MS. Although 2D HPLC may not be able to achieve the same degree of resolution as 2D PAGE, it is sufficient to fractionate complex proteomic samples into simple mixtures for analysis by nanoLC-MS/MS. 2D HPLC can be easily automated to give reproducible separations with quantitative recoveries at femtomole levels. Using a new multidimensional HPLC system, several schemes for orthogonal 2D HPLC of proteins and peptides were explored to find optimum conditions for low level protein identification in complex proteomic samples.

F. J. Yang, E. Tom, D. Wu, A. Moreno; Micro-Tech Scientific, 140 South Wolfe Road, Sunnyvale, CA 94086

This presentation will discuss the concept of an multi-dimensional capillary column methodology and the design of the instrument that can perform fully automated two dimensional capillary IC- RPLC with an integrated software and hardware for proteomics applications. Examples for biological sample analysis including desalting in a sequence of multiple automated software steps based on NT-4.- or Finnigan X’Calibur MS control will also be presented. For example,

Step 1: Load the entire protein/tryptic digest sample from a zero loss microtiter plate A/S into a SCX capillary column

Step 3. Load sample eluted from the SCX column using 50mM NH4OAC into a C18 precolumn. Desalt the precolumn with 100% water from pump 3 Step 4. Engage the pre-column with the nano-LC column. Pump1 and pump 2 apply solvent gradient for high resolution reverse phase nano-LC elution into MS Step 7. Return to step 1 for loading new sample . Repeat step 2 to 6 until all samples in the autosampler microtiter plate have been analyzed.

The above 2-dimensional steps can be performed using a 10 port switching valve for high through-put application.

W. Xu, J. W. Finch, R. S. Plumb, C. L. Stumpf, S. A. Cohen; Waters Corporation, 34 Maple St, Milford, MA 01757

While 2-D gel electrophoresis is currently the method of choice for proteome analysis, this approach does have some disadvantages namely in the analysis of membrane proteins, low copy number proteins, and proteins with high or low isoelectric points. To overcome some of these disadvantages, multi-dimensional HPLC methods have been explored as a possible alternative.

In this study we present results utilizing a fully automated multidimensional capillary HPLC system to separate complex peptide mixtures from protein digests where the eluent is directly coupled to a flow-thru nanospray source of a quadrupole-TOF hybrid (QTOF) instrument. We will show the applicability of 2D LC/MS/MS for peptide separation and protein identification, illustrate its advantage over 2D gel electrophoresis, and demonstrate the sensitivity of the QTOF data-directed analysis (DDA) approach. Additional discussion will include the chromatographic fluidic path considerations, optimization of column performance and modifications of the nanospray source for robustness and ease of use.

N. Tang, P. Tornatore, S. R. Weinberger; Ciphergen Biosystems, Inc., 6611 Dumbarton Circle, Fremont, CA 94555

ProteinChip® Array technology has emerged as a promising tool in proteomics. Studying protein-protein interaction is one of its important applications. We describe an approach for rapid epitope mapping based on protein-protein recognition on ProteinChip® Arrays followed by on-chip enzymatic digestion. Antibodies or receptors were first covalently linked to ProteinChip® arrays followed by specific extraction of antigens or ligands through molecular recognition motifs. Protease was then applied to each array to allow digestion of bound species. At times, sequential protease digestions were carried out to obtain proximal epitope sequence. Aggressive, non-specific proteases were used to achieve smaller peptide fragments. After digestion, each array was stringently washed to remove non-specifically bound peptides. The antigenic peptides remained bound to the array and were analyzed using ProteinChip® single and tandem mass spectrometry. The utility of this technique will be demonstrated by presenting applications of epitope mapping for Tumor Necrosis Factor, Carcinoembryonic Antigen, and other biomarkers of interest.

J. L. Duff¹, F. G. Hopwood¹, P. E. Smith¹, C. J. Hill¹, P. S. Gandhi¹, N. H. Packer¹, P. W. Cooley², D. B. Wallace², K. L. Williams¹, A. A. Gooley¹; ¹Proteome Systems Ltd., Sydney, Australia, ²MicroFab Technologies, Inc, Plano, TX

Two-dimensional polyacrylamide electrophoresis generates high-resolution arrays of proteins expressed by an organism. Chemical Printing is a concept where the chemistry is brought to the protein using piezoelectric pulsing for rapid and accurate microdispensing of picolitre volumes of reagents. This technology has been applied to proteomics to enable high throughput peptide mass fingerprinting (PMF) from protein arrays electrophoretically transferred onto membranes. The on-membrane digestion protocol has been streamlined to a simple automated process that has eliminated the multi-step procedures that are usually required for traditional protein in-gel digestion. A major advantage of the Chemical Printing technology is the ability to store the membrane for future analyses. This is particularly important for rare and valuable samples. Multiple endoproteinase reactions are also possible on the same protein spot because of the small area of protein that is manipulated. Peptide mass fingerprinting data resulting from the analysis of samples such as human plasma and platelets and the microorganism E.coli will be presented demonstrating the sensitivity of the Chemical Printer technology.

P. T. Jedrzejewski, G. Ganshaw, B. Fryksdale, G. England, B. S. Miller, A. Gaertner; Genencor International Inc., 925 Page Mill Road, Palo Alto, ca 94304

The study of the proteome is a critical step towards the functional analysis of the genome. However, robust, sensitive, and comprehensive methodologies for proteome separation and analysis similar to genomics methods (e.g., gene arrays) are not yet available. The current standard proteomic methodology (2D-PAGE) suffers from limitations. In order to overcome these limitations, we have focused on the development of alternative methods, specifically, multi-dimensional chromatographic methods, for proteome analyses.

In this presentation, we will show our efforts in this area. As a model biological system, we have selected the exoproteome of a cellulose production strain of filamentous fungi. Using this exoproteome we have evaluated the effectiveness of two multi-dimensional chromatographic methods: multi-dimensional chromatography on the protein level versus multi-dimensional chromatography on the peptide level following a total exoproteome digest. For either technique proteins were identified using ESI-MS/MS in conjunction with database search software. In this presentation we will detail the advantages and limitations of these techniques for the characterization of proteomes.

R. Zhang, F. E. Regnier; Purdue University, Department of Chemistry, West Lafayette, IN 47907-1393

Labeling reference analytes with stable isotopes has been widely used to create internal standards for quantitative analysis. Subsequent to labeling the chromatographic behavior of analytes is frequently modified. Similar strategies are now being applied in proteomics for quantifying protein expression in proteomics where internal standards are created by derivatizing reactive functional groups of analytes in a control sample with one form of the labeling agent while analytes in the experimental sample are labeled with a second, isotopically distinct form of the reagent. Subsequent to proteolysis and mixing of the samples, derivatized peptides were fractionated by chromatography and the isotope ratio in peptides determined by mass spectrometry. It was observed that during reversed phase chromatography the resolution of peptides differentially labeled with the ²H₈- and ¹H₈-ICAT reagent was substantially larger than that of isoforms labeled with GIST reagents. Ease and accuracy of quantification was seen to increase when the resolution of isotopically coded isoforms was minimized.

C. Hunter¹, L. Nuwaysir¹, T. Settineri¹, E. Witkowska¹, S. U¹, X. Zhang¹, P. Walter², M. Shigenaga²; ¹Applied Biosystems, Foster City, CA, ²Children's Hospital Oakland Research, Oakland, CA

Mitochondrial permeability transition (MPT), a key component of most forms of apoptosis, is a sequence of events culminating in the release of proteins from the mitochondria. The escape of some mitochondrial components may exacerbate MPT and continue the apoptosis cascade. Our goal is to characterize in vitro a time course of the mitochondrial proteins released under the conditions of MPT using the isotope-coded affinity tag method of relative protein quantitation using ICAT(TM) reagents and mass spectrometry. Changes in relative abundance of proteins released at early vs. late stages of MPT were investigated utilizing ICAT reagent labeling of their representative cysteine-containing peptides and identification/quantitiation by mass spectrometry. On the basis of the ICAT reagent-based quantitation of the released proteins, subtle differences in the temporal release patterns were noted for many of the proteins identified. Data analysis was performed with the new quantitation and identification algorithms of the ProICAT software package. Analysis of the quantitation results for all time points was performed with SpotFire DecisionSite software.

S. Krishnan, S. Zobbi; Applied Biosystems, 500 Old Connecticut Path, Framingham, MA 01701

A multidimensional chromatographic approach substituting the SDS PAGE gel separation of protein extract is used in the analysis of Schizosaccharomyces pombe proteome. The MDLC approach offers several advantages, such as automation, throughput, analysis of all classes of proteins, peptide recovery aided by in-solution digestion and the identification of low abundance proteins. The proteins are separated using ion exchange chromatography as a first dimension followed by reverse phase chromatography of the ion exchange fractions in the second dimension. The chromatographic fractions are proteolytically digested followed by MALDI-TOF mass spectrometric analysis. The lower resolution of the chromatographic approach resulted in fractions containing mixtures of proteins. The confidence in the protein identifications of such mixtures is overcome largely by current database search algorithms and the quality of the mass spectral data obtained. The speed and the automation of the MDLC approach make it attractive for large scale proteome studies. This approach coupled with quatitation technologies such as ICAT could further help in understanding the protein regulation pattern in the biological system.

M. X. Lin¹, J. M. Campbell¹, A. K. Walker², P. Andrews²; ¹Applied Biosystems, Framingham, MA, ²University of Michigan Medical School, Ann Arbor, MI

We have demonstrated collision-induced fragmentation (CID) of proteins up to 18,000 daltons using a matrix assisted laserdesorption ionization time-of-flight/time-of-flight MALDI TOF/TOF mass spectrometer. The high energy CID capability of the MALDI TOF/TOF mass spectrometer allows fragmentation of large peptides and small proteins while maintaining the sensitivity associated with MALDI. Discrete masses were observed for singly-charged ions resulting from fragmentation primarily at prolyl and aspartyl residues. Partial ion series were observed for some proteins, providing partial sequence tags. Fragment coverage ranged from 50% to 100% with redundant fragmentation information for most proteins. Effective fragmentation could be obtained using alpha-cyano 4-hydroxy cinnamic acid, dihyroxybenzoic acid, trihydroxyacetophenone, or sinapinic acid as matrices. The amount of protein required to provide good quality spectra ranged from 200 fmol to 10 pmol in these initial studies. CID of intact small proteins provides the opportunity to identify them via database searches or to directly map the sites of post-translational modifications without having to resort to enzymatic digests.

R. van Soest¹, R. Swart², J. Chervet²; ¹LC Packings(USA)/Dionex, 80 Carolina Street, San Francisco, CA 94103, ²LC Packings/Dionex, Amsterdam, Netherlands

2-D gel electrophoresis followed by MS analysis of tryptic fragments of protein spots is being used extensively in proteome analysis. Although such studies have revealed significant results, there are several problems related to sensitivity. The most important limitation is the maximum protein load on the gel. This dictates the amount and number of proteins that can be visualized for further characterization. Secondly, when crude extracts are used for 2-D gels, only the most abundant proteins are represented on the gel, with the minor components being below detection limits. Thirdly, the overlapping of protein spots hinders identification.

Multidimensional chromatographic separation followed by mass spectrometric analysis has been used to overcome some of these shortcomings. The use of fully automated on-line 2-D Capillary LC/MS will be presented for the identification of low abundance proteins in complex samples. The 2-D strategy is based on strong cation exchange (SCX) chromatography as the first separation dimension, followed by reversed chromatography as the second separation dimension. The study of low-abundance protein expression will be used to illustrate the potential of this technique.

D. Barnidge¹, G. Choudhary², L. E. Bonilla²; ¹Neuromics, Minneapolis, Minnesota, ²ThermoFinnigan Corp., 355 River Oaks Parkway, San Jose, CA 95134

Isotopically coded affinity tag (ICAT) reagents are an integral part of mass spectrometry based schemes for the relative quantification of proteins. In this poster we describe the use of ICAT in conjunction with bidimensional chromatography and ion trap mass spectrometry to quantify the levels of the membrane ion channel P2X3. The receptor P2X3 belongs to the P2X class of purinergic receptors which are gated by extracellular ATP. P2X3, in particular, has been implicated in nociception and its role in the modulation of chronic pain is currently an active topic of study. This work focuses on evaluating the performance of the ICAT method in providing relative quantitation of P2X3 levels in vitro with emphasis on using the technique for analyzing the changes in expression levels of cultures exposed to antisense oligonucleotides designed to specifically ‘knock down’ P2X3.

G. Browse, M. Deery, T. Dunkley, S. Hester, J. Howard, C. Jackson, A. Razzaq, K. Lilley; University of Cambridge, Department of Biochemistry, Cambridge, United Kingdom

The quantitative study of the protein content of cells is emerging as one of the main methods employed to make sense of the vast amount of DNA sequence and gene expression data currently being generated. Many novel techniques are emerging to aid high-throughput comparative protein analysis. One such method is Difference Gel Electrophoresis (DiGE), a technique that allows two different protein samples to be run on the same 2D gel by virtue of differential labelling using spectrally distinct fluorescent cyanine dyes. This method is very powerful in the study of changes in expression of soluble proteins between a test and control sample. It avoids the use of separate gels to study the variation between two samples, thereby removing the problems associated with gel to gel variability.

Here we discuss our evaluation of this technique, in terms of its reproducibility, sensitivity, linearity and dynamic range. We also discuss the use of this technique in showing quantitative changes in protein expression of membrane-associated proteins from two ecotypes of Arabidopsis.

L. Mouledous¹, S. Hunt², R. Harcourt², J. Harry², K. L. Williams², H. B. Gutstsein¹; ¹MD Anderson Cancer Center, 1515 Holcombe box 110, Houston, TX 77030, ²Proteome Systems Ltd., Sydney, NSW Australia

The anatomical complexity of tissues such as the brain makes characterization of proteomic changes in specific groups of cells very difficult. Laser Capture Microdissection (LCM) is a technique that is precise enough to dissect single cells within a tissue section. However, it is not clear whether staining techniques used to guide LCM interfere with protein recovery and resolution on 2D gels. The goal of this study was to determine the effects of staining procedures on protein extraction and separation. LCM samples of rat brain striatum obtained after several histochemical and immunostaining methods and unstained manually dissected samples underwent protein extraction and 2D gel electrophoresis. Gels were analyzed using Proteome Systems’ image analysis program, ImagepIQ to determine the number of proteins detected from stained samples compared with unstained controls. Selected proteins were identified using peptide mass fingerprinting on Proteome Systems’ proteome analysis platform, ProteomIQ. Our results indicate that substantial modifications of conventional staining techniques and careful selection of stains are necessary to ensure the fidelity of LCM samples.

P. Metalnikov¹, P. O'Donnel¹, G. Vassilovski¹, K. Ashman²; ¹Samuel Lunenfeld Research Institute, 600 University Ave., Toronto, Ontario M5G 1X5 Canada, ²Samuel Lunenfeld Institute, 600 University Ave, Toronto, Ontario M5G1X5 Canada

The development of robotic sample preparation systems has allowed the throughput of protein analysis to be accelerated. However, one of the rate limiting steps in preparing protein samples for analysis by mass spectrometry is their proteolytic cleavage. Robots have been successfully used for the in gel tryptic digestion of proteins as well as the de-salting, concentration and loading of protein digests onto MALDI targets. The advantages of these instruments are fast and stable sample processing and precise spotting. It seems reasonable to use these advantages for the entire protein sample preparation in a single instrument. In this paper, we describe a method to achieve this goal by performing protein sample concentration, all the digestion chemistry steps and MALDI target loading on ZipTips using a rapid and fully automated method.

A. M. Pitt¹, E. Kellard¹, J. Tilghman², J. VanDinther², S. Gutierrez¹, C. Hapenney³, V. Tisdale³, M. Engelstein¹; ¹Millipore Corporation, 17 Cherry Hill Drive, Danvers, MA 01923, ²Packard BioScience, Downers Grove, IL, ³Applied Biosystems, Framingham, MA

The classical method for protein identification entails techniques including gel electrophoresis, tryptic digestion, and mass spectrometry. One of the major bottlenecks with this approach is the extensive sample preparation required for the in-gel digestion process. This poster describes the automation of of the in-gel digestion steps prior to MS analysis. A 96-well filter plate containing the excised gel pieces is placed on the vacuum manifold where gel destaining, dehydration and trypsin addition takes place. Once the digestion is complete, the peptide extraction solution is added and peptides collected. The digested peptides are spotted on the MALDI target and sample processing software manages MS analysis and subsequent database searching of the indexed samples. The combination of vacuum filtration, quality controlled reagents, robotic liquid handling and analysis software enables a convenient, reproducible and high throughput method for in-gel digestion, sample clean up, and protein identification. The successful automation of such complicated and lengthy procedures provides the throughput, reproducibility, and simplicity anticipated to meet the challenges of Proteomics.

K. Steinert, F. Schäfer, H. Lubenow, C. Feckler, J. Ribbe; QIAGEN GmbH, Hilden, 40724 Germany

Methods based on well-established Ni-NTA technology for the isolation of 6xHis-tagged proteins offer a purification method highly suited for proteomics and functional genomics projects. The robust one-step purification process meets the challenge of purifying thousands of proteins with differing structures and characteristics. To allow protein purification in a high-throughput, automated format, protocols have been developed which can be run on QIAGEN^® BioRobot^® workstations.

Ready-to-run protocols covering a wide range of purification and assay applications are available. Ni-NTA Magnetic Agarose Beads allow micro-scale purification of up to 15 μg of 6xHis-tagged protein per well in addition to various types of assays using structurally active proteins. If larger amounts of protein are needed, the Ni-NTA Superflow 96 BioRobot protocol provides a convenient medium-scale method for purification of up to 300 μg of 6xHis-tagged protein in a 96-well format.

In our efforts to further increase the yield of 6xHis-tagged proteins, we have developed a novel protocol that allows purification of up to several milligrams of highly pure protein per well. The protocol is based on vacuum-driven filtration and Ni-NTA Superflow affinity chromatography and can process 96 samples, each derived from up to 25 ml culture volume (E. coli), in less than 3 hours. Examples of purification and assay applications using the protocols mentioned above, and data on reproducibility and cross-contamination–free processing are presented.

M. Kuschel¹, M. Kratzmeier¹, M. Dittmann¹, P. Barthmaier¹, S. Mouradian²; ¹Agilent Technologies, Hewlett-Packard-Str.8, Waldbronn, 76337 Germany, ²Caliper Technologies, Mountain View, CA

The Agilent 2100 bioanalyzer, the first commercial lab-on-a-chip (LOAC) system, was developed for fast and automated analysis of nucleic acids and proteins. Recently the system was enhanced to support flow cytometric analysis of cell fluorescent parameters. The LOAC system consists of disposable chips, application-specific reagent kits, computer-controlled instrumentation, and user-friendly software. Utilizing LOAC technology for protein analysis allows integrating electrophoretic separation, staining, destaining or fluorescence detection into a single process, and combining it with data analysis.

A variety of protein samples, such as cell lysates, column fractions or purified proteins were analyzed with the LOAC system. It allowed, to detect overexpression of recombinant proteins in cell lysates, check IMAC or reversed phase column fractions for protein purity, and analyze antibodies under reducing and non-reducing conditions. The protein assay allows sizing and performing relative quantitation based on internal standards or absolute quantitation based on user-defined standards. Furthermore, the performance of this system was compared to conventional SDS-PAGE. The LOAC system has several advantages compared to conventional SDS-PAGE including fast analysis times, reduced manual labor, automated data analysis, and good reproducibility. Using such a system allows tracking the protein of interest during the whole purification procedure, e.g. from cell lysates through column fractions to purified proteins.

G. Thorsen, H. Derand, S. Wallenborg, J. Engström, M. Inganäs, M. Holmquist, U. Selditz, M. Ljungström, P. Andersson; Gyros AB, Uppsala Science Park, Uppsala, SE-751 83 Sweden

In this work we present a generic microfluidic platform utilizing a CD microlaboratory. As the CD spins, liquids move through microstructures as a result of centrifugal force, eliminating the need for pumps, connectors or high voltage power supplies. The CD contains a large number of microstructures so that nanoliter sample volumes can be handled in parallel. Fluidic functions, such as mixing and volume definition, combined with the use use of functionalized beads in the channels, enable the creation of application-specific microstructures.

For fast, efficient capture of analyte molecules on a packed bed, as in a sandwich assay, it is necessary to precisely control flow rate. This is done by a combination of spinning speed, dimensional channel design.

Controlled surface chemistry is essential in microscale systems due to the large surface-to-volume ratio. We show an enzymatic assay that requires a surface with minimal protein adsorption. Precise volume definition is also used in this quantitative assay.

Detection options include laser induced fluorescence detection, performed while spinning the CD, or off-line detection and protein identification with MALDI mass spectrometry.

S. Park, S. Cho, M. Kim, Y. Kim, B. Kim; Seoul National University, Seoul, Korea, Republic of

Microfabricated capillary electrophoresis devices have been successfully applied to DNA assay and enzymatic assay. While pre- and post-column reactors were fabricated and showed versatile applications, a new design was necessary for each application. We describe a simple method of importing a reactor into micro-devices without pre-determined structures. A micro-device was sequentially constructed by fabricating the capillary channels on a PDMS substrate, spotting and drying a drop of dissolved nitrocellulose on a glass substrate, adsorbing protein on the formed nitrocellulose film, and assembling these two substrates. By means of this method, it is possible to vary the size of NC film reactor and to place the reactor at any position within the microchip device. We tested the performance of this reactor through β-galactosidase reaction. The β-Gal was assayed using fluorescein di-β-D-galactopyranoside. Enzymatic reaction and the inhibition for PETG compared well between a microchip and a conventional assay. The overall enzyme assays required only 1μl enzyme solution. This NC film method can be applicable to multiple and sequential enzyme reactions as well as affinity capture assay.

K. Ashman¹, J. Zhang¹, A. Lau¹, G. Vassilovski¹, P. O'Donnel²; ¹Samuel Lunenfeld Institute, 600 University Ave, Toronto, Ontario M5G1X5 Canada, ²Samuel Lunenfeld Research Institute, Toronto, Ontario Canada

The genome sequencing projects have given us a number of genome sequences. The next challenge is to translate the linear genomic sequence data into functional relationships between the many individual and sets of proteins that make up a living organism.

One of the major problems to be faced when analysing the large numbers of protein samples that are analysed in a proteomics project is to keep track of the samples: their storage; processing; analysis, and the reporting of results. This becomes especially onerous when working with diverse research groups internally, nationally and internationally. To address these problems a Web Browser Based Proteomics Sample Tracking Database, the STARS (sample tracking and results storage) database, has been developed. Filemaker Pro 5.5 was chosen as the software platform because of its: cross-platform compatibility (PC, MAC and Linux); ease of use; good scripting; easy translation to a web browser, security and low cost. The structure of the database and examples of its use will be described.

T. M. Smith, S. G. Porter; Geospiza, 2442 NW Market St. #344, Seattle, WA 98107

Recent years have seen an increased number of laboratories undertake projects in genomic sequencing. Data management tools such as the Finch Server, allow laboratories to analyze several genomic sequencing steps, saving time and money by quickly identifying problems.

Random (shotgun) sequencing is a common strategy that depends on high quality clone libraries. Often, these libraries are constructed by using physical shearing to randomly fragment the target DNA. Alternative methods can include transposon insertion and digestion with high frequency cutting restriction enzymes. Using the Finch Server, one can analyze DNA library quality by counting the number of vector and short clones, and the number of clones with poor quality sequence data. In addition, the Assembly Manager, in conjunction with SQL (Structured Query Language) queries can be used to assess library randomness and thus compare different methods for library construction. In this project, we show how the Finch-Server is used to compare physical shearing to restriction digests, and show how clone bias and chimeric clones, created by ligating non-contiguous restriction fragments during the cloning step, can be identified.

T. Le, H. Dang; Incellico, Inc., 2327 Englert Drive, Suite 205, Durham, NC 27713

One of the many challenges of ‘post-genomic-sequencing’ biology is to correlate and evaluate data from many different sources. Incellico’s cross-referencing technology called the Coded Electronic Life Library (CELL) and an associated microarray data management/collaboration application facilitate the expression data management and the assimilation of divergent data types. In CELL, biological entities (i.e. genes and proteins) and their aliases are represented as nodes in a graph-based ontology. Such a database allows the correlation and/or clustering of genes based on very different information sources, such as expression data and Medline search results. To demonstrate the functionality of CELL Suite, we analyzed the gene expression and drug interaction data from the NCI60 cancer cell lines and clustered the genes against PROSITE and INTERPRO domain entries, associated tissues, and gene ontology component classification. These analyses indicate that the presence of a cluster of highly expressed genes in melanoma cell lines are consistent with the existence of melanin biosynthesis and melanosome biogenesis in these cells, reflecting their tissue of origin.

J. I. LANGRIDGE¹, A. Millar², P. Young³, R. O'Malley², N. Swainston³, J. Skilling ⁴, J. B. Hoyes³, K. Richardson³; ¹Micromass UK Ltd, FLOATS ROAD, MANCHESTER, M23 9LZ United Kingdom, ²Micromass UK Ltd., Floats Rd, Wythenshawe, Manchester M23 9LZ United Kingdom, ³MICROMASS UK LTD, MANCHESTER, ⁴Maximum Entropy Data Consultants, Cambridge, United Kingdom

Large quantities of MSMS spectra can be acquired and database-searched in an automated manner, allowing rapid identification of large numbers of protein samples. A challenge remains, however, in that many of the spectra acquired do not provide matches when searched against known protein sequences. The nature of database searching is such that only peptides that match exactly those within the databank will be identified. Consequently, many good quality spectra of novel peptides remain unmatched. Currently the solution to this would be to extract these spectra manually from the data set, derive some degree of peptide sequence and performed further database searching. This, however, can be time consuming. With the introduction of an automated computer sequencing algorithm, MS/MS spectra can be identified by direct derivation of the novel peptide sequence. Here the intelligent batch-submission of a number of MSMS spectra is introduced, such that, upon completion of a database search, all unmatched spectra are automatically de novo sequenced. The results are integrated in a single Java interface, from which novel peptide sequences can undergo homology searching through submission to BLAST.

S. M. McDonald, M. J. Dudek, C. L. Fisher, K. Ramnarayan; Structural Bioinformatics, Inc., 10929 Technology Place, San Diego, CA 92127

The botulinum and tetanus neurotoxins are multidomain proteins that are released by the pathogens Clostridium botulinum and Clostridium tetani. The catalytic domains of these toxins are metalloendopeptidases, which act on specific proteins involved in vesicular transport processes. These neurotoxins thus inhibit the release of acetylcholine at neuromuscular junctions, leading to symptoms of muscle paralysis common in the diseases botulism and tetanus. In order to direct experimental studies of the catalytic mechanisms of these toxins, we have developed homology models for 44 members of this toxin family. Structural quality features, putative active site residues, and electrostatic properties of these structural models are discussed. A structural comparison of the catalytic domain of a recent crystal structure of a botulinum neurotoxin type B with one of the models is made, including a detailed analysis of zinc coordination and the conformation of the belt region. Docking calculations performed with the toxin models in complex with peptides from known presynaptic proteins are discussed with implications for substrate specificity and catalytic efficiency.

J. rousseau; University of Cape Town Faculty of Health Sciences, falmouth road, Observatory, Cape Town, Western Cape 7800 South Africa

Fluorescent STR-PCR DNA genotyping (ProFiler-Plus, Applied Biosystems) is performed by the Western Province Blood Transfusion Service for their Disputed paternity testing service. The amplified DNA samples are sent to our facility for gel electrophoresis (ABI 373), analysis (Genescan), interpretation and reporting.

The alleles of 9 genetic loci generated from more than 500 unrelated individuals were transferred to a database so that their distribution, frequency and percentage heterozygosity could be determined. The results will be presented.

G. E. Tarr; Boston University, 36 Cummington St, Rm 312, Boston, MA 02215

What could be more mundane than radiolabelling a protein? One buys a standard reagent, follows a standard protocol, and obtains labelled product. What could possibly go wrong? The protein starring in this short (and embarrassing) story is organophosphate hydrolase (OPH). Attached to RBCs, it protects rats from high doses of OP nerve agent; our goal now is to engineer an effective and safe defence for humans. Radiolabel allows circulating protein to be monitored independent of enzyme activity. I-129 was effective, but a less hazardous label was desired. Thus OPH was tritiated with commercial propionyl hydroxysuccinide ester. No H-3 was incorporated. Perhaps an unadvertised nucleophile in the buffer? Extensive dialysis, same result. Bad reagent? Another protein was labelled with high efficiency. The desalting column consistently disgorged a peak of radioactivity just after the protein and before the hydrolysed reagent was expected. What was that? Nothing by PAGE. Figured it out? Hint: I almost had, until the confusion of the desalting column led me off on a WGC. If you want the answer, suspicion confirmation, or just a laugh, visit the poster.

D. Harris, J. Smith, J. Leonard; Millipore Corporation, 17 Cherry Hill Drive, Danvers, MA 01923

An important step in current genome sequencing strategies is the cloning of large fragments of genomic DNA into F factor based vectors called BACs (bacterial artificial chromosomes). The cloned fragments are generally in the range of 100-300Kb while the most commonly used BAC vectors are 7-10Kb. Once cloned into the BAC vector, the fragments can be sequenced either directly or following PCR amplification of a specific region. One obstacle to adapting this strategy to a production sequencing environment is the low copy number of these vectors. Since the F factor is generally maintained at only 1-2 copies per cell, the BAC DNA yield obtained from a given volume of culture is significantly lower than that for plasmid DNA. Therefore, it has been necessary to alter the growth and DNA purification protocols, as well as sequencing reaction conditions, in order to maximize yields and obtain high quality DNA sequence data from BAC samples generated by mini preparation procedures.

We have developed a size exclusion based system for purification of BAC clones in a 96 well kit format which provides DNA of sufficient quantity and purity for direct BAC end sequencing. Cultivating the BAC clones in a 96 well culture block, we are able to reproducibly isolate 0.5-1.0 g of BAC DNA per well by this procedure. Using BigDyeTM v.3.0 terminator chemistry (Applied Biosystems) and the ABI 3700 DNA sequencer in conjunction with our new BAC purification system, 85-95% pass rates (>100 phred 20 bases) are consistently achieved for 96-clone BAC library plates.

A. Amit; Compugen, 1250 Oakmead Parkway, Suite 210, Sunnyvale, California 94085

The enormous high throughput potential of microarrays and DNA chips as a research tool has led many scientists to invest significant resources and time in conducting experiments, while overlooking the importance of accurate probe design. With the availability of the human genome sequence, it is now possible to create an accurate and comprehensive transcriptome database, which encompasses biological phenomena such as alternative splicing. By applying its LEADS technology, Compugen has developed Gencarta, an annotated collection of the genome, transcriptome and proteome database. Using Gencarta, it is now possible to accurately design probes for manufacturing microarrays. Compugen’s DNA Chip Design platform provides custom chip and oligo designs for Affymetrix or oligonucleotide based chips, while taking into account alternative splicing, probe specificity, distance from the 3’ end, secondary structure, common melting temperatures and more. Using this platform, Compugen, together with Sigma-Genosys, has designed and manufactured comprehensive libraries of oligos corresponding to 19,000 human genes. These OligoLibraries now provide the life sciences research community with access to oligonucleotide collections for high throughput gene expression analysis, drug discovery and functional assays. Examples and findings resulting from the use of these libraries will be discussed.

D. L. McMinimy, T. Carver, J. P. Giard, T. B. Radcliff, J. K. Naggert; the Jackson Laboratory, 600 Main Street, Bar Harbor, ME 04609

The WICGR collection of simple sequence length polymorphisms (SSLPs) markers have greatly facilitated gene mapping. However, when analyzing strains for which the allele sizes have not been published, the small amount of publicly available size information remains problematic. In addition, one can often observe size discrepancies of between 2 to12 bases from those on the Lander list (which contains allele sizes for all Mit markers, but only for 12 mouse strains) and those generated in the lab. This limited strain information and lack of precise size information often limits the ability to pool markers together.

To address these limitations, our core lab surveyed the Jackson Laboratory Staff and asked them to list strains of mice they anticipated using in conjunction with the Allele Typing service. From the staff response we compiled a list of 48 different strains of mice. To date we have run some 250 mouse Mit markers on these DNAs. Size information will be made available via the web.

F. S. Siegman¹, H. Lubenow², C. Schade²; ¹QIAGEN Inc., 28159 Avenue Stanford, Valencia, CA 91355, ²QIAGEN GmbH, Max-Volmer-Str. 4, Hilden, 40724 40724 Germany

High-throughput projects such as genome sequencing analysis or genotyping projects require standardized and reliable front-end sample processing, including liquid handling and sample purification. Workflow can be streamlined by integrating a variety of different instruments for full automation of sequential tasks. We present a modular approach to laboratory automation that provides streamlined workflow while ensuring adaptability to changing application repertoires and growing throughput requirements.

BioRobot® 3000 extended arm systems are designed to accommodate complementary instruments on both, left and right sides. These include microplate readers and the BioRobot® RapidPlate®, a 96-channel pipetting system with capabilities for 96-as well as 384-well pipetting. These integrated systems provide rapid and fully automated processing for applications such as sample dilution and transfer, DNA template normalization, and sequencing and amplification reaction setup. Other integration allows for fully automated, streamlined PCR applications. Already existing non-extended arm BioRobot 3000 systems can be easily modified to accommodate the BioRobot RapidPlate, adding 96/384-well pipetting to their application range.

BioRobot Twister systems can be integrated with BioRobot robotic workstations, as well as the BioRobot RapidPlate, to increase storage capacity for microplates, blocks and disposable tips allowing longer unattended processing.

S. Iacobas, D. Spray, D. Iacobas; Albert Einstein College of Medicine, 1410 Pelham Prkwy South, Bronx, NY 10461

The Theory of Genomic Patholog provides a method to express mathematically the properties of genomic profiles obtained in DNA array experiments. By hybridizing extracts from the same type of tissue with several arrays, one estimates the variability of transcription levels, incorporating them it in the expression scores to quantify the real alteration in the test tissue. The Cartesian product of possible value intervals of expression scores is organized as pre-Hilbert space where every point represents a possible genomic profile and every curve a possible genomic evolution. The genomic patholog of an altered profile is the Euclidian distance separating the representative point by the 0-centered hypersphere with the radius 2.58 that is the (p < 0.01) subspace of normal and stable profiles in the reference. The theory classifies stages of genetic disorders according the range of patholog values. We found that stably transfection with Cx36 altered the genomic profile of N2A cells with 78.64E-6/expressed gene, while depletion of Cx43 modified the brain expression profile with 19.19E-6/expressed gene. Concepts potentially applicable for gene therapy evaluation and optimization are mathematically defined.

F. A. Sherwani, I. Parwez; Aligarh Muslim University, 4/656 Gule Rana, Nagla Road, Dodhpur, Aligarh, UP 202002 India

Membrane bound enzyme sodium/potassium dependent adenosine triphosphatase (Na⁺/K⁺-ATPase) and mitochondrial enzyme succinic dehydrogenase (SDH) have often been implicated in osmoregulatory adjustment of teleosts at higher salinities. The present study reports changes in activity profiles of these two enzymes in gills and kidney following transfer of catfish, Clarias batrachus to 30% and 35% seawater (SW). From each group, 4-5 fish were sampled at 3h, 24h and 3 and 6 days post-transfer to 30% and 35% SW. The gills were excised and kidney was dissected out to estimate the enzyme activity. Results showed no significant change in the activity profile of Na⁺/K⁺-ATPase in the gills. While in kidney, it showed significant increase from 3h (p<0.001) and 24h (p<0.001) in 35% and 30% SW respectively, for up to 6 days. SDH activity was significantly increased from 24h in gills both in 30% and 35% SW (p<0.01) and remained significantly elevated up to 6 days of transfer, an activity profile pattern quite opposite to that of Na⁺/K⁺-ATPase. However, no significant change was observed in SDH activity profile of kidney at higher salinities.

D. G. Myszka¹, F. Arisaka², O. Byron³, M. L. Doyle⁴, E. Eisenstein⁵, P. Hensley⁶, C. Lombardo⁷, F. Schwarz⁵, W. Stafford⁸, J. Thomson⁹; ¹University of Utah, 50 N. Medical Dr. School of Medicine Rm 4A417, Salt Lake City, Utah 84132, ²Tokyo Institute of Technology, Yokohama, Japan, ³University of Glasgow, Glasgow, United Kingdom, ⁴Bristol-Myers Squibb PRI, Rt 206 and Provinceline Rd, Princeton, NJ 08543, ⁵National Institute of Standards and Technology, Rockville, MD, ⁶Pfizer Global Research and Development, Groton, CT, ⁷The Burnham Institute, La Jolla, CA, ⁸Boston Biomedical Research Institute, Watertown, MA, ⁹Pfizer/Agouron Global Research and Development, La Jolla, CA

Characterizing in detail the interactions between biomolecules requires information on the assembly state, affinity, kinetics, and thermodynamics associated with complex formation. The analytical technologies often utilized to measure biomolecular interactions include analytical ultracentrifugation (AUC), isothermal titration calorimetry (ITC), and surface plasmon resonance (SPR). In order to evaluate the capabilities of Core Facilities to implement these technologies, the Molecular Interactions Research Group (MIRG) developed a standardized model system and distributed it to a panel of AUC, ITC, and SPR operators. The model system was composed of a well-characterized protein-ligand pair: bovine carbonic anhydrase II (CA-II) and 4-carboxbenzensulfonamide. Study participants were asked to measure one or more of the following: 1) the affinity and thermodynamics for complex formation by ITC, 2) the affinity and kinetics of complex formation by SPR, and 3) the molecular mass, homogeneity, and assembly state of CA-II by AUC. The results from this study provide a benchmark for comparing the capabilities of individual laboratories and to define the utility of the different instrumentation.

D. A. Bintzler¹, P. S. Adams², Y. Bao³, D. Bartley⁴, L. Kasch⁴, R. Keefe⁵, L. Petukhova⁶, C. Rosato⁷; ¹University of Cincinnati, ²Trudeau Institute, ³University of Virginia School of Medicine, ⁴Johns Hopkins University, ⁵Wadsworth Center, ⁶The Rockefeller University, ⁷Oregon State University

The Fragment Analysis Research Group has organized a study to investigate laboratory protocols for multiplexing fluorescent-labeled markers that amplify microsatellite regions from human DNA templates using PCR. By amplifying more than one marker in a single PCR, laboratories can reduce their reagent expenses as well as the time spent setting up reactions for DNA fragment analysis projects. Participants who volunteered for this study received two DNA template samples and five fluorescently-labeled primer pairs, and were asked to amplify each marker as a multiplex reaction containing all 5 markers or individually and pooled (post-PCR multiplexing). Participants then reported the conditions they used for carrying out their PCR and the sizes of the 5 microsatellite alleles (in base pairs) obtained using software analysis after electrophoresis of PCR products. The aim of this study was to evaluate the methods used by other facilities to optimize multiplex PCR, and to present a summary of their results so that other laboratories might learn helpful hints and possible pitfalls associated with trying to simultaneously carry out PCR involving multiple primers. Results from the study are reported here.

S. Buckel¹, R. Cook², M. Crawford³, D. Dupont⁴, B. Madden⁵, J. Neveu⁶, L. Steinke⁷, J. Fernandez⁸; ¹Amgen, Boulder, CO, ²Baylor College of Medicine, Houston, TX, ³Yale University, New Haven, CT, ⁴Applied Biosystems, Foster Cirt, CA, ⁵Mayo Clinic, Rochester, MN, ⁶Harvard, Cambridge, MA, ⁷Nebraska Medical Center, Omaha, NE, ⁸Rockefeller University, 1230 York Ave., New York, NY 10021

The ABRF-2002ESRG sample is the 14^th study designed as an education and self-evaluation tool for laboratories that perform Edman Sequence Analysis. This year’s study, one of the more challenging protein samples distributed by an ABRF research group, is a known protein with a staggered amino terminus which was a "real-life" sample submitted to a ESRG member lab. The protein was purified using commercially available, precast SDS-PAGE gels and transferred to PVDF. Protein bands were distributed to 72 members who requested ABRF-2002ESRG, along with a data instruction sheet and a brief survey. Participating members were requested to report observed raw data, interpret the data as they normally would for an investigator, and identify the protein using a BLAST search. Results from the study will be presented from 29 responses to show how labs fare with a difficult, but sequencable protein.

K. L. Knudtson¹, C. Griffin², A. Brooks³, D. A. Iacobas⁴, K. Johnson⁵, G. Khitrov⁶, K. Lilley⁷, A. Massimi⁸, A. Viale⁹, W. Zhang⁸, Y. Bao¹⁰, G. Grills¹¹, H. Thaler⁹, C. Cox³; ¹University of Iowa, 323 EMRB, Iowa City, Iowa 52242, ²UCSF GCRC/SFGH, ³U. of Rochester, ⁴Albert Einstein College of Medicine, 1410 Pelham Prkwy South, Bronx, NY 10461, ⁵Jackson Labs, ⁶The Rockefeller U., ⁷University of Cambridge, Department of Biochemistry, Cambridge, CB21QW United Kingdom, ⁸AECOM, ⁹MSKCC, ¹⁰U. of Virginia, ¹¹Harvard U.

Users of DNA microarray technology must overcome many challenges to obtain results in which the variability of the data is due solely to biological activity. It is important to be aware of and account for potential sources of variability in the experimental design and results. The goal of this study is to identify non-biological factors that contribute to variation in microarray results. A retrospective study, using data collected by ABRF MARG members in their core labs, was conducted using both Affymetrix GeneChip and spotted microarray technologies. The GeneChip study analyzed the metrics and meta data from a total of over 900 murine U74A, human U95A, and Test 2/3 arrays. The spotted microarray study analyzed the effects of slide fabrication, hybridization and scanner settings from over 100 slides in which the same reference RNA was used. The effect of tissue type, array lot, hybridization procedure and scanner settings on the variability of results was investigated for both types of microarray technologies. The results of these studies provide insight on potential sources of experimental error with microarray technologies and suggest experimental strategies to correct them.

D. P. Arnott¹, M. Gawinowicz², R. A. Grant³, W. S. Lane⁴, L. C. Packman⁵, K. Speicher⁶, K. Stone⁷; ¹Genentech, Inc., 1 DNA Way MS63, South San Francisco, CA 94080, ²Columbia University, New York, NY, ³Proctor & Gamble Co., Cincinnati, OH, ⁴Harvard University, Cambridge, MA, ⁵Cambridge University, Cambridge, United Kingdom, ⁶The Wistar Institute, Philadelphia, PA, ⁷Yale University, New Haven, CT

Identification of multiple components in a protein mixture represents a continuing technological challenge. Whereas the main component of a simple mixture (e.g. a 1 or 2D PAGE band or spot) is often identifiable, a greater problem is posed when the proteins span a range of abundances and identification of the minor components is important. This is often the case with pull-down assays where bands may contain several components due to the complexity of the recovered protein profile, or from proteins that associate non-covalently. The Proteomics Research group therefore prepared and distributed a multicomponent sample for analysis by ABRF members. This sample of up to 6 proteins was supplied as a tryptic digest. Some components were present at up to 2 pmol each; others at a several-fold lower level. Participants identified as many components as they could by any techniques available to them. The results should indicate which mass spectrometric or chemical approaches different laboratories applied, the levels of sensitivity that they achieved, and their ability to analyze results with different search programs. Problem areas identified in the analyses may provide a focus for future studies.

A. T. Yeung¹, M. Gunthorpe², K. Lilley³, S. Scaringe⁴, P. S. Adams⁵, B. P. Holloway⁶, T. Thannhauser⁷, K. Mills⁸, G. A. Buck⁹; ¹Fox Chase Cancer Center, ²University of San Francisco, ³University of Cambridge, Department of Biochemistry, Cambridge, CB21QW United Kingdom, ⁴Dharmacon Research, Inc., ⁵Trudeau Institute, ⁶Centers for Disease Control, ⁷Cornell U., ⁸Millennium Pharmaceuticals Inc., ⁹Virginia Commonwealth University

In 2001, the Nucleic Acids Research Group conducted a brief email survey to all the ABRF laboratories that listed DNA/RNA synthesis as a service in the current ABRF directory. In doing so, this networking will facilitate information exchange on nucleic acid synthesis and quality control, and to enlist interested members in experimentation with new synthesis chemistries. Further, with more direct contact, the NARG will determine the reasons for low participation numbers in past studies and also determine the future of DNA/RNA synthesis services in core facilities. Afterwards, a discussion with the audience will also provide the NARG with information about participants in regards to a possible study regarding fluorescence resonance energy transfer (FRET) probes in real-time PCR and to acquire suggestions for useful studies in the future.

J. Hawes¹, D. Bartley², T. C. Hunter³, E. Jackson-Machelski⁴, D. Leviten⁵, R. Pershad⁶, M. Robertson⁷, D. Spicer⁸; ¹Indiana University School of Medicine, Indianapolis, IN, ²John Hopkins University, Baltimore, MD, ³University of Vermont, Burlington, VT 05405, ⁴Washington University School of Medicine, 660 S. Euclid Avenue, P.O. Box 8103, St. Louis, MO 63110, ⁵ICOS, Corp., 22021 20th Ave SE, 98021, ⁶University of Texas. M.D.Anderson Cancer Center, 1515 Holcombe Blvd, Houston, Texas 77030, ⁷University of Utah, Salt Lake City, UT, ⁸Prolinx, Inc., Bothell, WA

DSRG has solicited the ABRF community to participate in its biannual general survey to obtain comprehensive information regarding facility composition and configuration. DNA Sequencing Core Facilities provide integral DNA services that expedite research at their respective institutes. New methods and instrumentation are continuously developed and brought online that impact the make-up of the typical DNA Sequencing Core Facility. A web-based survey was utilized to collect information from participating laboratories regarding: staffing, funding, instrumentation, chemistries employed, throughput, and services provided. The information gathered from this survey will help depict a more accurate profile of the current DNA Sequencing Core Facility.

D. Leviten¹, J. Hawes², T. Hunter³, E. Jackson-Machelski⁴, R. Pershad⁵, D. Spicer⁶, D. Bartley⁷, K. L. Knudtson⁸, G. Grills⁹, M. Robertson¹⁰, J. VanEe¹¹, T. Thannhauser¹¹; ¹ICOS, Corp., 22021 20th Ave SE, 98021, ²Indiana University School of Medicine, Indianapolis, IN, ³Vermont Cancer Center, Burlington, VT, ⁴Washington University School of Medicine, 660 S. Euclid Avenue, P.O. Box 8103, St. Louis, MO 63110, ⁵University of Texas. M.D.Anderson Cancer Center, 1515 Holcombe Blvd, Houston, Texas 77030, ⁶Prolinx, Inc., Bothell, WA, ⁷John Hopkins University, Baltimore, MD, ⁸University of Iowa, 323 EMRB, Iowa City, Iowa 52242, ⁹Harvard-Partners Genome Center, Boston, MA, ¹⁰University of Utah, Salt Lake City, UT, ¹¹Cornell University

SNPs are new to the sequencing core environment and are fast becoming an important tool for researchers in both academic and commercial settings. Therefore, the DSRG was particularly interested and concerned about how DNA sequencing core facilities could/would be able to handle SNP sequencing projects and hence, launched a study.

The goal of this study is to determine whether Single Nucleotide Polymorphisms (SNPs) and different ratios of SNP mixtures can be sequenced accurately using the equipment and chemistries currently being used in participating member laboratories. Some major questions to be answered are how SNPs are ‘called’ and what ratios can be correctly determined.

The DSRG launched an internal study to determine whether our current equipment and chemistries could accomplish the above goal. It was determined that the goal could be accomplished, and these results were given at the ABRF 2001 conference. Since then, an external study has been launched, and data from this study is compared and contrasted to the results from the internal study.

H. Remmer ¹, N. Ambulos¹, L. F. Bonewald¹, J. J. Dougherty¹, E. Eisenstein², E. Fowler¹, J. Johnson¹, A. Khatri¹, N. Ritter¹, S. T. Weintraub¹; ¹, ²National Institute of Standards and Technology, Rockville, MD

The Peptide Standards Project Committee (PSPC)

The Peptide Standards Project is the production and establishment of three synthetic peptides as registered and certified peptide reference standards. This project is conducted by the Peptide Standards Project Committee (PSPC) of the ABRF in collaboration with the National Institute of Standards and Technology (NIST). The project entails the following:

(1) Large scale synthesis of the standard peptides in high (>98%) purity; (2) Packaging of the peptides in small quantities for distribution and analysis; (3) Analysis of the physical and chemical properties of the peptide standards by the ABRF member laboratories; (4) Tests for stability and shelf life of the peptides.

The Peptide Standards Project was originally initiated by joint interactions between the Quality Compliance Committee (QCC) and the Peptide Synthesis Research Group (PSRG). It is the first ABRF project funded by NIST.

Detailed information about the project and its current status will be presented as well as data on peptide design and analysis and results of optimization for large-scale synthesis.

E. Jackson-Machelski¹, T. C. Hunter², D. Leviten³, R. Pershad⁴, D. Spicer⁵; ¹Washington University School of Medicine, 660 S. Euclid Avenue, P.O. Box 8103, St. Louis, MO 63110, ²University of Vermont, Burlington, VT 05405, ³ICOS, Corp., 22021 20th Ave SE, 98021, ⁴University of Texas. M.D.Anderson Cancer Center, 1515 Holcombe Blvd, Houston, Texas 77030, ⁵Prolinx, Inc., Bothell, WA

Members of the DNA Sequencing Research Group will host a roundtable discussion immediately following the research group presentation. Topics to be covered include: Setting Up a Sequencing Core, Capillary Sequencers in a Core Laboratory, Pharmacogenomics and Troubleshooting. The discussion will include an opportunity for questions from the audience and audience participation is strongly encouraged.

G. A. Grant, M. W. Crankshaw, C. L. Crankshaw; Washington University, St. Louis, MO 63110

This roundtable discussion will provide an opportunity to discuss the interpretation of results from database searching using experimentally derived mass values for the identification of unknown proteins. A brief presentation of the factors that influence the scores will be presented, followed by an open session where the audience can comment. In addition to the speakers listed above, a panel of experts will be on hand to answer questions.

P. S. Adams¹, J. Hawes², D. S. Grove³, B. P. Holloway⁴, T. C. Hunter⁵, K. L. Knudtson⁶, G. L. Shipley⁷; ¹Trudeau Institute, ²Indiana University School of Medicine, Indianapolis, IN, ³Pennsylvania State University, ⁴Centers for Disease Control, ⁵University of Vermont, Burlington, VT 05405, ⁶University of Iowa, 323 EMRB, Iowa City, Iowa 52242, ⁷University of Texas-Houston Medical School

Real time quantitative PCR using Fluorescence Resonance Energy Transfer (FRET) technology has become an invaluable technique for quantitating gene expression, viral and bacterial detection and quantitation, allelic differentiation and Single Nucleotide Polymorphism (SNP) detection. Many core facilities are being asked to provide these services to their researchers. Factors to consider in setting up and running this service such as equipment selection, level of service, SYBR green vs. specific probes, one-step vs two step, type of probes, automation, charges, etc. will be presented. A panel of core facility directors who already provide this service will discuss their experiences. Audience participation in a roundtable discussion format will be encouraged. Technical as well as service related discussions will be welcome.

M. E. McComb¹, L. Licklider², S. T. Weintraub³, M. Raida⁴, B. Williamson⁵, R. Martin⁶, G. Valaskovic⁷; ¹Boston University School of Medicine, 715 Albany Street, Boston, MA 02118, ²Harvard Medical School, 240 Longwood Ave, Boston, MA 02115, ³University of Texas Health Sciences Center at San Antonio, San Antonio, TX, ⁴CellZome GmbH, Meyerhofstrasse 1, Heidelberg, n/a D-69117 Germany, ⁵Applied Biosystems, Framingham, MA, ⁶Micromass, Inc., Beverly, MA, ⁷New Objective, Inc., Woburn, MA

On-line LCMS and LCMSMS affords the opportunity to purify, pre-concentrate and separate complex peptide mixtures derived from proteolytic digests of proteins in order to introduce them as single components to the mass spectrometer. Removal of interfering matrix compounds/analyte ions which interfere with the ESI process results in improved S/N and dynamic range. Capillary/nano LC columns (300/75 μm id.) are compatible with microspray thus enhancing concentration dependent ESI-MS allowing analysis of sub pmol amounts of material from complex mixtures. Software driven data dependent tandem MS capability of new instruments combined with improved data manipulation increases throughput and reliability of the analysis. The entire process of separation-MS-MSMS-data analysis may be completely automated allowing for the analysis of large numbers of samples in a high-throughput proteomics approach. This workshop will focus on several practical aspects of capillary and nanoscale LCMS/MSMS. A number of examples will be presented and emphasize instrumentation, separation methodology, column design, 2D separations, the column/ESI interface, QIT and QoTOF mass spectrometers and automated sequencing.

M. L. Doyle¹, O. Byron², A. Cooper², C. Lombardo³, D. G. Myszka⁴; ¹Bristol-Myers Squibb PRI, Rt 206 and Provinceline Rd, Princeton, NJ 08543, ²University of Glasgow, Glasgow, United Kingdom, ³The Burnham Institute, La Jolla, CA, ⁴University of Utah, 50 N. Medical Dr. School of Medicine Rm 4A417, Salt Lake City, Utah 84132

A major need in the molecular interactions field is to measure accurate molecular-level binding parameters (affinity, stoichiometry, kinetics, and thermodynamics). This need grows with the increasing list of protein-protein and other biomolecular interactions. Ultimately, our ability to predict biology will require a quantitative understanding of these interactions, including how are they regulated by effector molecules, expression levels, post-translational modifications, and formation of various combinations of protein-protein and protein-nucleic acid multi-molecular complexes. The purpose of this roundtable is to provide an informal gathering for discussion and question and answers about quantitative measurement of molecular interactions. Several biophysical methods (e.g., surface plasmon resonance, calorimetry, light scattering, analytical ultracentrifugation, capillary electrophoresis, etc.) are immediately germane to the discussion. Input about other methods will also be encouraged. It is anticipated that the discussion will be driven by audience participation.

L. Licklider, C. Thoreen, S. P. Gygi; Harvard Medical School, 240 Longwood Ave, Boston, MA 02115

Methods for polyacrylamide gel electrophoresis remain an integral part of mass spectrometry-based proteomics. Widespread application of cellular extracts to gels for final purification before MS analyses has called for development of high-sensitivity nanocolumn chromatography / tandem MS methods to identify and characterize complex protein mixtures in gels following in-situ trypsin digestion. An fully-automated on-line strong-cation exchange LC / nanoflow reverse-phase LC / ion trap MS method was developed to add selectivity for and to efficiently enrich for peptides in coomassie- and silver-stained gel samples. A vented column strategy was employed to transfer peptides at microliter flowrates between each dimension and to permit optimal sensitivity and resolution at 100 nanoliter / min. The method was augmented by a versatile web browser interface for interpreting results for database searches performed with uninterpreted MS / MS spectra. Enhancements in the efficiency for identifying and characterizing proteins after PAGE will be the subject of the presentation.

N. S. Weich¹, L. Poisson², R. He², M. Kim², J. Carroll², R. Hoffman³, J. Gutierrez-Ramos²; ¹Millennium Pharmaceuticals Inc., 75 Sidney Street, Cambridge, MA 02139, ²Millennium Pharmaceuticlas Inc., Cambridge, MA, ³Medicine/Hematology/Oncology University of Illinois, Chicago, IL Canada

A pairwise comparison and self organizing map (SOM) clustering analysis of data obtained through the utilization of microarray technology to monitor differential gene expression in non-human primate models of response to and recovery from irradiation and carboplatin treatment has been performed. In general, the two animals receiving irradiation exhibited very similar patterns of degree of differential gene expression at all four timepoints examined as compared to pretreatment. The carboplatin treated animal displayed a distinctive pattern from the irradiated animals with significant levels of differential gene expression detected by day 7. Direct comparison of the behavior of specific genes at each timepoint in the irradiated animals determined that response to treatment was most similar on a gene-by-gene basis at days 14 and 21 with greater than 80% of the genes upregulated in one animal also being upregulated in the other. Cellular pathway analysis of the samples generated from the irradiated animals identified activation and suppression of many pathways.

D. A. Brenner¹, B. Schnabl¹, R. Bataller¹, J. Lora², N. S. Weich²; ¹University of North Carolina School of Medicine, 156 Glaxo Bldg, Campus Box 7038, Chapel Hill, NC 27599-7038, ²Millennium Pharmaceuticals Inc., Cambridge, MA

Hepatic fibrosis is a common result of chronic liver diseases, including hepatitis C and alcoholic liver disease. Hepatic fibrosis of any etiology results from the activation of the hepatic stellate cell from a quiescent retinoid storing cell to a proliferating cell secreting extracellular matrix protein and a variety of cytokines and chemokines. When quiescent stellate cells are purified from a normal liver, they undergo a similar activation process in culture similar to activation by fibrogenic agents in vivo. The aim of the study was to analyze gene expression in activated human hepatic stellate cells (HSCs) as a fingerprint of the fibrogenic process in patients with progressive fibrosis due to hepatitic C virus. In addition, the gene expression in an immortal HSC line was compared to early activated HSCs. Altered patterns of gene expression were assessed using DNA microarray analysis and confirmed by realtime PCR or by measuring protein levels. Initially, the early activated HSCs and early fibrotic liver disease predominantly express extracellular matrix protein genes. Subsequently, a more inflammatory gene expression pattern predominates in the senescent HSCs and in end stage cirrhosis.

D. N. Farlow¹, R. M. Walker¹, L. Zhang², M. Muesing², D. Ho², J. Gutierrez-Ramos³; ¹Millennium Pharmaceuticals, 75 Sidney St, Cambridge, MA 02139, ²Aaron Diamond AIDS Research Center, New York, NY, ³Millennium Pharmaceuticlas Inc., Cambridge, MA

Current anti-HIV therapies target specific viral genes to disrupt HIV replication. HAART is a combination of protease and reverse transcription inhibitors reduces viral levels which increases patient survival, but does not eradicate the virus. We present transcriptional profiling of more than 10,000 different genes probed with RNA isolated from PBMCs of 4 HIV-infected patients before and after treatment. More than 300 genes, for example, interferon-induced proteins 1-8D, 6-16, P27 and P78 and proteasome regulatory subunits are detected at higher levels before treatment in all patients. More than 600 genes, including interleukin receptors 1, 6, 7, and 9 are expressed at higher levels after treatment demonstrating that some immune function has been restored in all 4 patients. Analysis of differential expression of other genes not previously known to be regulated during drug therapy, for example, more than 20 kinases downregulated in this experiment, reveals many potential new targets. Study of all genes changing significantly during treatment of HIV-infected patients will lead to development of new therapy which targets host cell genes influenced by the virus during its lifecycle.

J. Henion, G. A. Schultz, X. Huang, G. Rule, S. Prosser, T. Corso, C. van Pelt, S. Zhang; Advion BioSciences, Inc., 15 Catherwood Rd., Ithaca, NY 14850

Mass spectrometry (MS) is a valuable analytical tool for protein characterization. Matrix assisted laser desorption ionization (MALDI) and electrospray ionization (ESI) provide complementary molecular information on proteins and protein digests. To-date, MALDI has the advantage of automation while ESI is more easily coupled with nanoscale separations. This lecture will describe the ESI ChipTM, an automated nanoelectrospray platform, that is an important new analytical tool for studies in proteomics.

The ESI ChipTM automatically analyzes samples from a conventional 96-well microtiter plate using disposable pipette tips with no sample-to-sample carryover. We are exploring novel strategies to provide chip-based electrospray MS applications amenable to studies in proteomics as well as small molecule applications in the pharmaceutical industry.

The ESI Chip is a microfabricated nanoelectrospray nozzle array etched from the surface of a silicon substrate. This nozzle can be coupled with a variety of analytical techniques including off-line sample preparation, CE, HPLC, and CEC. This ESI ChipTM is robotically controlled to provided unattended, nanoelectrospray analysis and is designed to be compatible with commercially available ESI-MS systems. Nanoelectrospray examples will be shown coupled with ion trap, time-of-flight, and triple quadrupole systems. Examples will include protein identification of in-gel tryptic digests, as well as the determination of peptides from non-gel digests.

A. L. Burlingame, L. Huang, C. Sihlbom, R. Chalkley, K. F. Medzihradszky, D. Maltby, M. A. Baldwin; UCSF, 513 Parnassus Av., San Francisco, CA 94143-0446

To gain an understanding of cell homeostasis and its perturbation by genetic and/or extracellular cues, development of the capacity to identify and define global protein expression states rapidly and reliably is essential. This is challenging per se, and the need is as widespread as the pursuit of a molecular understanding of living systems. Over the last decade, the technologies of mass spectrometry have made impressive inroads in the identification of proteins and their covalent modifications based on analysis of one or a few proteins at a time. This presentation will focus on the development of the capacity to identify thousands of proteins per day based on the new ABI MALDI TOF TOF MS technology using a 200Hz laser. Applications to identification of components involved in nucleocytoplasmic protein transport [N. Allen, L. Huang, A. L. Burlingame and M. Rexach, J. Biol. Chem. 2001, 276, 29268-29274] and other projects will be discussed. This work was supported by NCRR Grant RR 01614.

P. R. Griffin, E. Rohde, J. J. Cummings, J. Mehl, N. A. Yates, M. Southern, W. A. Hanlon, L. Kochanski; Merck Research Laboratories, 123 East Lincoln Ave, Rahway, NJ 07065

2DE is a powerful separation technique for complex protein mixtures. The introduction of a variation of this technique, 2-DIGE, has contributed to a reduction in gel-to-gel variations. Up to three protein samples were labeled separately with different fluorophores prior to 2DE then mixed and separated on the same gel. Proteins were detected in a sequential fashion at different fluorophore specific wavelengths. Resulting superimposable protein expression maps were analyzed and compared using dedicated software. This presentation describes the utility of 2-DIGE for the analysis of sera and tissue lysates. We show that proteins in extensively pre-processed sample fractions were successfully labeled with dye and separated. Gels were stained with Sypro Ruby and spots were excised by a robot. Proteins were digested in a semi-automated fashion using a Tecan robot. The resulting tryptic digests were analyzed by automated MALDI-TOF-MS or automated microspray LC-MS. Acquired spectra were automatically submitted to Sequest. An assessment of the feasibility of protein identification from analytical and preparative gels using 2-DIGE will be presented.

D. Landsman; NCBI NLM NIH, 45 Center Drive MSC 6510, Building 45 Room 6AN12J, Bethesda, MD 20892

The identification of open reading frames in a genome is a challenge that is being met both computationally and experimentally and there are considerable efforts underway to expedite the determination of many of the protein folds and structures. However, the regulatory networks which underpin the normal functioning of cells and which represent the interactions between the protein and RNA products are less well understood. In the yeast, Saccharomyces cerevisiae, there are predicted to be about 300 DNA-binding proteins with a wide variety of specific or non-specific DNA binding. Many of the sites that these proteins bind to are undiscovered and several methods for prediction have been developed. We sought to decrease the rate of false positive detection of these sites by incorporating expression profile data into a consensus pattern-based search methodology. Based on our analysis, we have developed a web-based tool called PROSPECT, which allows consensus pattern-based searching of gene clusters obtained from microarray data. In two other unrelated projects, we will discuss our analyses of the human genome sequences in GenBank with particular reference to L1 retrotransposons and to processed pseudogenes.

N. V. Grishin; HHMI/UTSouthwestern, 5323 Harry Hines Blvd, Dallas, TX 75390-9050

Approaches integrating sequence, structure and functional information with evolutionary considerations have been proven to be most efficient for understanding weak similarities between proteins. Several examples of remote homology detection using combination of computational methods will be discussed. In particular, power of transitive sequence similarity searches in reliable detection of homologs at close to and below random sequence identity will be illustrated. Several pairs of proteins with statistically supported sequence similarity that adopt different structural folds will be shown.

G. B. Fields; Florida Atlantic University, 777 Glades Road, Boca Raton, Florida 33431

Tumor cell invasion involves a complex series of correlated macromolecular interactions, resulting in tumor cell receptor recognition of ECM components, induction of "outside-in" signaling, dissolution of the ECM, and promotion of tumor cell motility. To better understand tumor cell behavior, we have developed triple-helical "mini-collagen" model systems. These mini-collagens have specifically been utilized to examine (a) tumor cell interactions with a glycosylated region of type IV collagen and (b) the ability of tumor cell enzymes to degrade triple-helices. For (a), a novel procedure was developed for synthesis of glycosylated Hyl derivatives. Glycosylation was subsequently found to have a profound influence on tumor cell adhesion and spreading. For (b), FRET substrates were designed using Mca and Dnp derivatives. Kinetic parameters and activation energies were determined for "collagenolytic" proteases. Overall, mini-collagens mimic the 3D structure of native collagens, yet allow for analysis of sequence-specific events that occur within these collagens. The incorporation of unusual amino acid derivatives within mini-collagens facilitates the mechanistic dissection of ECM invasion by tumor cells.

P. E. Dawson; Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037

Chemical ligation methods have faciliatated the synthesis of large polypeptides and have enabled the synthesis of a variety of proteins and protein domains. Native chemical ligation utilizes a highly chemoselective reaction between two fully unprotected peptides, one containing an N-terminal Cys and the other a C-terminal thioester group, to yield a native amide bond at the site of ligation. Some of the practical considerations of applying this technique to protein targets will be discussed. In addition, progress will be reported on the development of related techniques including solid phase chemical ligation and the synthesis of proteins without Cys residues.

J. B. Fenn; Virginia Commonwealth University, 1001 West Main St., Richmond, Virginia 23284-2006

Some 33 years ago Malcolm Dole proposed that evaporation of solvent from charged droplets of a dilute solution of molecules could produce intact gas phase ions of those molecules no matter how large or involatile they might be. Dole's attempts to prove the feasibility of such Electrospray Ionization" (ESI) failed to attract any followers for 15 years. Then in 1983 a paper from our lab showed that ESI could indeed produce intact ions from small molecules of species that could not be vaporized without catastrophic decomposition. Dole's vision was further verified in 1988 by the discovery that ESI could produce intact ions from proteins with molecular weights of at least 50,000. That report triggered an avalanche of activity which produced 1400 publications in 2001 alone and still seems to be growing. Moreover, that number is dwarfed by the unpublished results languishing in the files of pharmaceutical companies.

From his personal experience, perspective and prejudice, this balding and grizzled veteran will review the origins and evolution of the ESI process, why it remains so enigmatic, and what the future may hold for a couple of prospective applications.

L. Huang¹, C. Sihlbom¹, R. Chalkley¹, N. Allen², M. Rexach², A. L. Burlingame¹; ¹UCSF, 513 Parnassus Av., San Francisco, CA 94143-0446, ²Stanford University, Stanford, CA

The S. cerevisiae nuclear pore complex (NPC) is a supramolecular assembly of 30 nucleoporins that cooperatively facilitate nucleocytoplasmic transport. Thirteen nucleoporins containing FG peptide repeats (FG Nups) are proposed to function as stepping stones in karyopherin-mediated transport pathways. To understand its transport mechanism at molecular level, protein interactions occurring at distant locations within NPC were sampled using immobilized nucleoporins and yeast extracts [Allen NP et al. J Biol Chem. 2001, 276, 29268]. To obtain a global view of the nucleocytoplasmic transport network, hundreds of proteins were obtained using various nucleoporins and karyopherins under different growth conditions. To facilitate rapid identification of all of the interacting proteins on such large scale, we have developed an integrated high throughput protein identification system using robotic sample preparation techniques, a novel MALDI-TOF/TOF tandem mass spectrometer, and automated data collection and analysis. The nucleoporin interacting proteins were identified using peptide mass mapping (2 sec/digest) and peptide sequencing (5 sec/peptide) on this single instrument system. Supported by NIH NCRR 01614.

S. R. Weinberger, T. Yip, L. Lomas, N. Tang, P. Tornatore, T. Pham, E. Fung, A. A. Karavanov; Ciphergen Biosystems, Inc., 6611 Dumbarton Circle, Fremont, CA 94555

Surface Enhanced Laser Desorption / Ionization Mass Spectrometry is an advanced analytical technique providing facile protein analysis of complex biological mixtures. Commercially, SELDI-MS has been embodied within Ciphergen’s ProteinChip® Array technology (Fremont, CA, USA).

Proteinchip Array Technology is typically employed to perform differential display between control and experimental groups. Following pre-fractionation, experimental and control group samples are directly analyzed using an array of different surfaces and appropriate wash conditions. After detection, protein profiles of each group are analyzed for obvious instances of up or down regulation. The recent application of multivariant, pattern-matching algorithms have proven quite successful in finding clusters of proteins positively associated with the presence of neoplasm, monogenic and polygenic disease, as well as with toxicological insult. Once potential biomarkers have been identified, focus shifts towards the identification and characterization of these markers.

This talk presents recent advancements in the effort to identify and characterize potential biomarkers discovered using ProteinChip Array Technology.

K. W. Turteltaub, K. Dingley, J. S. Vogel; Lawrence Livermore National Laboratory, 7000 East Ave., Livermore, California 94550

Accelerator mass spectrometry (AMS) is a mass spectrometric technique for measuring isotope ratios. Attomole (10 ^-18 moles) sensitivity with radiolabeled compounds is routine. The technique uses a Van de Graaff electrostatic accelerator at megavolt (mV) energies coupled to momentum, velocity, and energy-loss selection to separate and count nuclei of isotopes such as ³H and ¹⁴C. The technique has been applied successfully in ADME, dose-response, biochemical and most recently in human subject research. AMS also offers the potential to reduce or eliminate radioactive waste and to reduce the need for high specific activity agents or libraries in drug screening studies. This work performed under the auspices of the US DOE by LLNL (W-7405-ENG-48) and partially supported by CA55861, ESO4705, and RR13461.

P. Wagner; Zyomyx, Inc., 26101 Research Road, Hayward, CA 94545

Novel high-throughput biomolecular analysis in genomics, proteomics, drug discovery, disease diagnosis and the development and application of patient-specific medicines require highly parallel, ultrasensitive, miniaturized device technologies. While technological innovation in form of DNA microarrays (gene chips) and other formats have adapted the analysis of genetic material to a miniaturized format, the more delicate nature and diversity of proteins in function, structure, stability and abundance have precluded the development of analogous tools for proteomic anaysis. In addition, a comprehensive proteomic analysis would require measurement and characterization of protein abundance and chemical modifications as well as discovery of unknown proteins, new pathways and functional linkages.

Protein biochips have started to emerge recently based on new developments and integration efforts in advanced materials, protein engineering and detection physics. Recent developments and selected examples will be presented with an emphasis on the technical challenges in surface and assay methodologies.

M. H. Cardone; Merriamck Pharmaceuticals, 50 Church Street, Cambridge, MA 02139

Microarray technology allows the simultaneous analysis of thousands of parameters with a single experiment. This technology has been successfully used to profile mRNA expression and has recently been adapted to explore protein abundance using antibodies arrayed at high density. We have used antibodies arrayed on glass slides and on MALDI mass spectrometry surface to analyze proteins from cell extracts. Using these formats we are able to explore phosphorylation states, protein abundance, and protein translocation to or from different compartments of the cell in response to perturbations of the cell steady state. Additionally we have used this format to explore protein-protein interactions among members of functional families of proteins. This is done by performing affinity capture of these proteins by the printed cognate partner.

C. Williams; Millennium Pharmaceuticals, 640 Memorial Drive, Cambridge, MA 02139

Protein/Protein interaction analysis is an integral part of many proteomics platforms. There are many applications, both in vivo and in vitro, that contribute technologies to this area. Currently there is an increased development effort in the area of protein arrays. With the continued success of DNA arrays for research it is a natural progression to develop and use arrayed proteins in the discovery process. Antibody arrays are some of the earliest and most convenient forms of protein arrays. These types of arrays can provide useful information about protein expression profiles as well as possible pathway interaction partners. One weakness of these arrays is that they usually require two antibodies to the same protein for detection, one for capture and one for detection. This could significantly increase the number of reagents required to make a meaningful array. This talk will focus on our current efforts to develop and validate arrays using surface plasmon resonance as a label free detection strategy.

A. J. Jones, J. B. Briggs, S. Ma, R. G. Keck; Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080

Our focus is the characterization of the glycosylation status of recombinant proteins expressed in CHO cells to support their development as pharmaceuticals. We routinely analyze mixtures of released oligosaccharides by MALDI-TOF MS in both positive ion mode (for neutral structures) and negative ion mode (for acidic or sialylated) structures. Data from the two methods can be "linked" to yield overall glycan distributions. These data allow us to derive quantitative information of the numbers of terminal saccharides which are key to understanding the impact of the glycosylation status on activity and/or metabolic clearance. Derivatization is a common way of introducing fluorophores which enable high sensitivity detection for chromatographic or electrophoretic separations. We have recently developed a label which also allows all oligosaccharides to be detected in a single MALDI-TOF spectrum with high sensitivity. This is proving to be both powerful and versatile and the use of CE, ion-exchange and normal-phase chromatography, coupled with MALDI-TOF MS and glycosidase treatment will be illustrated.

J. Hirabayashi, K. Kasai; Teikyo University, Sagamiko, Kanagawa 199-0195 Japan

Among post-translational modifications, protein glycosylation is a central issue in the post-genome sciences, because glycans play dominant roles in various cell-cell recognition events. Hence, a core strategy for analyzing glycosylated proteins under the concept of the "glycome" should be established. Though the term glycome is defined as "a whole set of glycans produced in a single organism", we specifically focus on glycans attached to glycoproteins produced in genome organisms (Proteomics 1 (2001) 295). Principal objectives in the project are to identify: i) which genes encode glycoproteins (i.e., genome information); ii) which sites among potential glycosylation sites are actually glycosylated (i.e., site information); iii) what are the structures of glycans (i.e., structural information); and iv) what are the functions of glycosylation (functional information). For these purposes, we introduced 2 affinity technologies recently. One is named the "glyco-catch method" to identify genes encoding glycoproteins, and the other is reinforced "frontal affinity chromatography". The concept and strategy of glycomic approaches referring to the on-going project of C. elegans will be described.

A. Cooper; University of Glasgow, Joseph Black Building, Glasgow, Scotland G12 8QQ United Kingdom

Most physico-chemical processes have an associated heat effect, and biomolecular interactions are no exception. Calorimetric instrumentation is now available to measure this quite routinely in biomolecular systems with reasonable amounts of material. Until recently, microcalorimetry has been seen as rather arcane technique, suitable only for the most dedicated pursuit of absolute thermodynamic parameters. But it is now being used more and more as a relatively straightforward, non-destructive, non-invasive analytical technique for studying biomolecular stability and interactions, even by those who care little for thermodynamics. This session (together with the associated tutorial) will concentrate specifically on the analytical uses of microcalorimetry (as opposed to the fundamental thermodynamics) with examples of current applications in biomolecular research and biotechnology. Techniques will cover differential scanning (DSC), isothermal titration (ITC) and pressure perturbation (PPC) calorimetry.

G. A. Holdgate, W. H. Ward; AstraZeneca Pharmaceuticals, Alderley Park, Macclesfield, Cheshire SK10 4TG United Kingdom

Isothermal titration calorimetry (ITC) monitors the heat change when a ligand binds to a protein. This unique detection method gives ITC its broad application. It may permit accurate and precise measurement of affinity. ITC is direct, facilitating interpretation, since there is often no necessity for competing molecules. Measurements do not require immobilisation or chemical modification. ITC can allow characterisation of protein fragments and mutant enzymes lacking catalytic competency. It can measure stoichiometry, allowing evaluation of the proportion of the sample that is functional. ITC is the only technique that directly measures the enthalpy of binding (ΔH°). Changes in structure at protein-ligand interfaces often have an imperceptible effect on binding affinity, because changes in enthalpy of binding tend to be compensated by similar changes in the entropic penalty. ITC reveals this enthalpy-entropy compensation and so gives insight into biomolecular structure activity relationships (SAR). Titration in the presence of other ligands can reveal information relating to the mechanism of action of test compounds, identifying intermolecular complexes that are relevant for biological activity and structure-based drug design. It is due to its wide application and recently improved sensitivity, that ITC is being used increasingly as a valuable tool throughout drug discovery.

J. Ramsey; Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6142

After several years as a background research activity in the chemical separations community, microfabricated fluidics devices (or microchips) are receiving much attention from the federal and commercial sectors. Microfluidics microchips promise to deliver many similar advantages that have been demonstrated by microelectronics devices over the past two decades. Microchips already allow biochemical experiments to be performed in an automated fashion providing quantitative results two orders of magnitude faster and using four to six orders of magnitude smaller quantities of materials than conventional technology. Further increases in throughput are likely to be achieved through refinement of devices and use of parallel biochemical processing. In addition materials consumption will also likely decrease as technology is developed to store and manipulate the small quantities of materials microchip devices require. We will present some of our recent developments in microchip technology as related to biochemical measurement and discuss probably new directions.

B. H. Warrington, S. Y. Wong; GlaxoSmithKline, NFSP (N), Third Avenue, Harlow, Essex CM19 5AW United Kingdom

The competitive world of drug market forces pharmaceutical companies to an expensive search for suitable leads. Although automation and new technologies have increased the productivity of the individual chemist, this has hardly scratched the enormous and diverse universe of the possible compounds and has led to logistical problems of reagent consumption, sample storage and waste production. Our objective is to exploit the advantages of fast reaction and product separation offered by microchannel systems to rapidly prepare minute amounts of products that will remain compatible with concurrently advancing miniaturized screening techniques. It is anticipated that as future screening could require only a few hundred atoms, the nanomole of test compound targeted will be more than sufficient. The aim is to create a device in which reagents are introduced in a constant flow stream, to obtain products that can then be chromatographycally separated and directed towards an integrated screening system.

A. Kopf-Sill; Caliper Technologies Corp., Austin, TX

We are using microfluidics technology to develop a flexible laboratory testing platform. The devices we make are essentially serial processors. Reaction plugs move through a fluid channel and each one experiences the same conditions – much like an assembly line. The paradigm offers extremely high data reproducibility and quality. It also allows for informed experimentation. The results of one experiment can be used to design the subsequent one. Other features are nanoliter reaction volumes, automated reagent assembly, and integrated functionality. Products are now on the market that make use of this technology to offer productivity advantages for individual researchers, as well as for high throughput centralized laboratories.

M. P. Washburn¹, G. Oshiro², R. Ulaszek¹, C. Deciu², E. Winzeler², J. Yates¹; ¹Torrey Mesa Research Institute, 3115 Merryfield Row, San Diego, CA 92121-1125, ²Genomics Instititue of the Novartis Research Foundation, San Diego, CA

Proteomic technologies are being developed to determine large-scale changes in protein expression levels. Several quantitative proteomic methods have recently been published where the protein expression levels are compared between two different growth conditions by metabolic isotopic labeling of proteins or post-growth isotopic labeling of proteins in a sample. To fully understand how growth conditions effect biological systems, both protein and mRNA expression levels must be analyzed. We have applied metabolic labeling strategies to S. cerevisiae in order to determine the changes in protein and mRNA expression levels. S. cerevisiae was grown in 15N enriched minimal media and compared to S. cerevisiae grown in rich media. Each sample was analyzed via mRNA expression array and multidimensional protein identification technology. By combining these methods we were able to correlate the changes in protein and mRNA expression levels of several hundred gene products.

N. J. Dovichi, S. Hu, D. Michels, A. Dambrowitz, L. Zhang, R. Schoenherr; University of Washington, Seattle, WA 98195-1700

We are developing technology to monitor the proteome of a single cancer cell by use of comprehensive two-dimensional capillary electrophoresis to separate proteins into components and laser-induced fluorescence to detect the proteins. Our goals are to provide protein fingerprints as a prognostic indicator of cancer, as well as providing a tool to monitor changes in protein expression during development and to study the expression of selected regulatory proteins at the lowest possible levels.

S. A. Cohen, H. Liu, R. S. Plumb, J. W. Finch, D. Wall, S. J. Berger, W. Xu; Waters Corporation, 34 Maple Street, Milford, MA 01757

Difficulties in automating 2D gel electrophoresis as well as problems with the analysis of several protein types including membrane proteins have provided incentive to explore alternate methods for proteomics analysis. We have studied 2D liquid chromatography for separating extremely complex peptide mixtures derived from protein mixtures. Here we present data showing how optimizing the chromatographic system peak capacity can dramatically improve the quality of the MS/MS data acquisition. Less well studied has been multi-dimensional separations of intact proteins. We have employed a comprehensive 2D system linking an ion-exchange column and two reversed-phase columns to permit the resolution and analysis such mixtures. Offline analysis is accomplished through the collection of fractions followed by MALDI-ToF MS of either the intact proteins or peptides produced via their enzymatic digestion. The offline data are also compared to on-line analysis of the intact molecules as provided by an ESI-ToF MS. Data from both standard protein mixtures and yeast ribosomal proteins were used to evaluate the effectiveness of these approaches.

T. Tuschl, S. M. Elbashir, J. Harborth, K. Bechert, M. Lagos-Quintana, J. Martinez, A. Yalcin, A. Patkaniowska, W. Lendeckel, K. Weber; Max Planck Institute for Biophysical Chemistry, Am Fassberg 11, Goettingen, 37077 Germany

21-Nucleotide RNAs are important specificity determinants for post-transcriptional regulation of gene expression in the processes of RNA interference (RNAi) and microRNA-guided translation control. In RNAi, duplexes of 21-nt RNAs are produced by Dicer RNase III cleavage from long dsRNAs. The 21-nt small interfering RNAs (siRNAs) guide sequence-specific mRNA degradation of homologous single-stranded target RNAs. When siRNA duplexes are transfected into mammalian somatic cells, gene-specific silencing is observed without activation of the unspecific interferon response. Therefore, siRNA-mediated gene silencing provides a new tool for the analysis of mammalian gene function in cultured cells.

Cloning of endogenous, size-fractionated total RNA of fly embryos and human HeLa cells lead to the identification of more than 30 novel genes coding for 21-nt RNAs. These RNAs are related to the previously described small temporal RNAs lin-4 and let-7 in C. elegans, and are referred to as microRNAs. Dicer RNase III and probably other cofactors such as the Argonaute proteins excise them from 30 base-pair stem-loop precursor molecules. MicroRNAs are believed to control translation of target mRNAs by binding to complementary sequences within the 3'-UTR, and are not believed to cause mRNA degradation.

L. Marino-Ramirez, J. Minor, J. C. Hu; Texas A&M Univ., College Station, TX

With the availability of genomic sequences it is now possible to rapidly identify, map and characterize oligomerization domains on a genome-wide scale. Genetic approaches to the study of protein interactions will be reviewed, emphasizing n-hybrid approaches. We are using bacteriophage λ repressor fusions to find homotypic interaction domains encoded by the genomes of S. cerevisiae and E. coli. Random fragments of genomic DNA were used to construct libraries of λ repressor fusions expressed at low levels in E. coli. Repressor fusions encoding homotypic interactions were recovered by selection for immunity to phage infection, and inserts encoding functional oligomerization domains were identified by sequencing from both ends. In this way, we identified oligomerization units in 35 yeast proteins 28 E. coli proteins so far. The oligomerization domains found by this method include a mixture of homotypic interactions that have been previously characterized at varying levels of detail, novel interactions with predicted coiled-coils, and novel domains of unknown structure. Interaction domains found in this way could be useful targets for structure determination, and could also provide tools for genetics and drug discovery.

A. J. Link, J. L. Jennings, A. Canutescu, S. Sanders, T. Weil; Vanderbilt University Medical Center, 1161 21st Ave S, Nashville, TN 37232

Most proteins in the cell work within multi-subunit complexes. We are employing various approaches to isolate native protein complexes followed by multidimensional mass spectrometry analysis to identify the purified proteins and post-translationally modified amino acids. From our proteomics analysis is emerging a complex network of protein interactions.

K. S. Misono¹, E. Folta-Stogniew², K. Williams², J. Philo³; ¹Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195, ²Yale University, New Haven, CT, ³Alliance Protein Laboratories, Thousand Oaks, CA

Natriuretic and vasorelaxant activities of a cardiac hormone, atrial natriuretic peptide (ANP), are mediated by cell-surface receptors coupled to guanylate cyclase (GCase). The receptor is a single-chain transmembrane protein that contains an extracellular ANP-binding domain (ECD), a transmembrane region, and an intracellular kinase-like regulatory domain and a GCase catalytic domain. Binding of ANP causes GCase activation by an as yet unknown mechanism. Several lines of evidence suggest that activation by ANP may involve receptor dimerization. To elucidate the signaling mechanism, we expressed and purified the ECD, and characterized its structure-function relationships. By SEC-LS, we analyzed ANP binding to the ECD and its effect on the association behavior of the ECD. Our results showed that the ECD is in a concentration-dependent dynamic equilibrium between the monomer and dimer states (Kdissoc ~0.5 micromolar) and that this equilibrium is strongly shifted toward the dimer by ANP binding. The results of SEC-LS analysis will be discussed in correlation with data from analytical ultra-centrifugation and the 3- dimensional structure of the ECD determined by X-ray crystallography.

E. J. Folta-Stogniew; Yale University, 333 Cedar Street, New Haven, CT 06520-8024

Size exclusion chromatography (SEC) coupled with "on-line" laser light scattering (LS), refractive index (RI) and ultraviolet (UV) detection provides an elegant approach to determining the molecular weights of proteins and their complexes in solution. SEC serves solely as a fractionation step in order to minimize the ambiguity that can result from the fact that light scattering provides the weight average molecular weight, MW, of all species present in solution. I will discuss realistic expectations for MW determination using LS coupled with SEC, define sample requirements and identify possible limitations of the SEC/LS analysis. Analyses of 15 protein standards that range from 6.5 kDa to 475 kDa suggest that molecular weights of native proteins may be determined in a single SEC/LS experiment with an accuracy of ±5%. The method is suitable for glycoproteins as well as membrane proteins solubilized in non-ionic detergents. SEC/LS analysis represents a fast and robust approach to determining MW and to monitoring protein oligomerization in solution.

J. M. Neveu¹, W. J. Henzel²; ¹Harvard Microchemistry Facility, Cambridge, MA, ²Genentech, 1 DNA way, South San Francisco, CA 94080

Protein samples requiring analysis often contain high levels of salts and or detergents. In this tutorial, we will discuss methods for protein and peptide sample cleanup and concentration for the purpose of Edman degradation. We will discuss the methods and the limitations of ultrafiltration, reversed-phase desalting including Zip tips, electrophoresis and electroblotting. We also discuss innovated approaches to obtaining sequence on blocked samples. A handout will be provided summarizing these methods with references.

M. Randesi; HHMI / The Rockefeller University, 1230 York Ave. box 105, New York, New York 10021

With the seemingly ever-increasing speed of computers and therefore the instruments in which they control, many of the services that core labs offer are being transformed from low volume to high volume. This transformation brings upon many new challenges, one being the management of samples and their data. Fortunately, this increase in computing speed also can be used to assist in the tracking of samples, and the distribution of the data related to those samples. There are commercially available software packages, which one can purchase, but in general, these are quite costly for the average core facility. An overview of one lab’s alternative approach, using Apple scripts and Excel macros to handle all of these tasks will be presented.

J. T. Medalle; The Rockefeller University, 1230 York Ave. Box 105, New York, New York 10021

Well thought-out computer languages accomplish the automation of manipulating data and their files on PCs, Macintoshes, and servers. All facilities funnel data into specified formats through manual or electronic protocols. Transiting streams of data from client to facility can be a bottleneck for any size lab. In a small throughput facility, lab personnel manually enter data into notebooks or electronic files. For high throughput labs, database software programs perform data stream transition, tracking, and storage. For facilities who are at this interface between low and high throughput, the issues of purchasing complex database software are upon them. The processing of data streams into reaction logs, client receipts, email notification, and data storage are overwhelming on notebooks and confusing electronically. Examples of computer languages, such as Apple scripts and Excel macros will be presented to demonstrate a cost-effective solution to handle sporadic sample flux and their data streams.

B. S. Imai; The Rockefeller University, 1230 York Ave, New York, NY 10021

Web based data submittal and dissemination offers an attractive potential for streamlining data handling chores in the small to medium sized facility. Since all facilities will have access to some sort of personal computer systems, whether PC or Macintosh based, a system that is compatible with these systems is preferable in terms of cost and ease of file maintenance. Automated data collection systems can allow a limited number of personnel to accumulate ever-increasing amounts of data. This makes data archiving a significant issue even for small core facilities. Plunging computer hardware costs and academic software discounts coupled with relatively simple to use web development applications have placed web server based data systems within the reach of even modest budgets. A model system used for dissemination of protein analysis data will be presented to demonstrate how virtually any data that can be printed can be easily distributed via the web. Security issues will also be considered.

E. Lader; Ambion, Austin, TX

Reliable expression analysis by real-time RT-PCR necessitates careful attention to sample handling prior to RNA isolation, RNA isolation itself, and RNA storage. As real-time RT-PCR is a very accurate and reproducible method of quantitation, it is particularly sensitive to differences in the quality of the starting material. Important factors include variability in the intactness of RNA samples, the presence of residual genomic DNA or other contaminants carried over from either the sample or the isolation method. The use of RNAlater for RNA stabilization in tissue and cell samples, and the use of the Agilent Bioanalyzer for the analysis of RNA will be discussed. In addition, a simple real-time assay to quantify residual DNA contamination and methods for removing residual DNA will be covered. Finally, several interesting observations about the relationship between the amount of RNA in a sample and the amount of reverse transcriptase in the RT reaction will be presented.

G. L. Shipley; UT-Houston HSC, 6431 Fannin St., Houston, TX 77030

The methodologies for transcript quantitation have evolved quite rapidly in the last decade from Northern transfer/hybridization and RNAse protection to competitive RT-PCR and finally real-time RT-PCR. Utilization of the latest methodology successfully however, requires acquired skills in 3 main areas. The first is the development of the assay itself. The 3 steps in this process are selection of the sequence to be used to generate the assay, selection of a probe and primer set and testing the reagents under varying assay conditions resulting in a optimal assay. Second, there are many options when running the assay. Some examples are what quencher dye to use, what normalizer dye to use, one-step vs two-step reactions, homemade chemistry vs kits, cycle times and hot start vs normal taq. The third topic is data analysis which includes issues such as: setting the baseline and threshold and how those settings affect the slope, y-intercept and the data, examples of inter- and intra-assay variability and selecting a transcript for normalization of the final data.

B. P. Holloway, J. Limor, K. McCaustland; Centers for Disease Control, 1600 Clifton road, Atlanta, Ga 30333

The establishment of Real-Time quantitative PCR (RT-QPCR) analysis as a new activity within many core facilities has resulted in an ever increasing need for dual-labeled, Fluorescence Resonance Energy Transfer (FRET) DNA probes. Many facilities performing RT-QPCR also maintain an active DNA synthesis activity, however the synthesis of FRET probes is often delegated to commercial vendors. The absence of practical protocols as well as perceived complexities in synthesis chemistries have discouraged many DNA synthesis facilities from synthesizing their own FRET probes. This presentation will provide practical protocols which any core facility can utilize to synthesize, purify and characterize these probes. Traditional chemistries using a fluorescent quencher as well as non-fluorescent quenchers will be presented.

D. Dick¹, J. I. Knecht², R. Cook¹; ¹Biosearch Technologies, Inc., Novato, CA, ²Biosearch, 81 Digital, Novato, Ca 94949

Black Hole Quencher (BHQ) dyes are a family of non-fluorescent quenchers designed to enhance performance of fluorogenic probe assays. In this presentation, preparation and analysis of BHQ probes will be discussed in an attempt to pass on some of the practical experience gained in working with these dyes. A primary focus will be the workup conditions for FAM/BHQ-1 probes, including stability of these dyes to ammonia and ammonia methylamine (AMA) workup. Recommendations will be made for deprotection conditions of FAM/BHQ-1 probes prepared using both standard and fast deprotect amidites in ammonia and AMA deprotection systems. Preparation of higher wavelength probes incorporating BHQ-2 and BHQ-3 quenchers can be complicated by the instability of most higher wavelength reporter dyes to traditional workup methodologies. Alternative, mild deprotection strategies to facilitate synthesis of these high wavelength probes will be discussed. Finally, a review of any known byproducts or artifacts commonly seen in analysis of BHQ probes by HPLC and mass spectroscopy will be given.

C. Potter; Biosearch Technologies, 81 Digital Dr., Novato, CA 94949-5750

This talk is for molecular biologists, who would like to understand how fluorescence is regulated in fluorophore-quencher assays. I will explain (1) how fluorescence is controlled by quenchers and (2) how dark quenchers enable multiplexing in multi-reporter assays. FRET or Fluorescence Resonance Energy Transfer (a.k.a. Förster RET) is the transfer of the excited state energy from a donor (fluorophore) to an acceptor (quencher). A number of quenching mechanisms, including FRET, collisional, and static quenching all contribute in varying degrees for different dye-quencher pairs. Several probe designs will be addressed with respect to their specifications and subsequent quenching efficiencies. Quencher dyes must control fluorescence for any PCR experiment to yield usable data; in order to multiplex effectively, the quencher must have no native fluorescence that contributes to background and must allow a large range of fluorophores to be used.

S. Levy, B. Boone; Vanderbilt University, 702 Light Hall, Nashville, TN 37323-0615

DNA microarrays allow the parallel measurement of relative gene expression levels and have begun to revolutionize biomedical research. The evolution of these technologies has been driven by the need for more powerful analytical approaches to utilize the enormous amount of genomic data and resources being acquired through the various genome projects. Due to the large investments necessary to develop a successful microarray program, many institutions are developing centralized core support facilities. This tutorial will concentrate on two key areas in the development of a successful microarray core facility. The first is the development of a complete technology platform that produces high quality, reproducible and meaningful RNA expression profiles. The second is the development and implementation of a bioinformatics program that provides efficient access to the large amount of data produced during a microarray experiment, supports data annotation standards, and provides statistical and higher-order analysis support. Example technologies and methodologies from current microarray facilities will be shown and important questions to ask when evaluating a microarray support facility will be discussed.

N. R. Barnes; Millennium Pharmaceuticals, Inc., 640 Memorial Drive, Cambridge, MA 02139

Fluorometric microvolume assay technology (FMAT) offers a rapid, homogeneous assay alternative to conventional heterogeneous ELISA methodologies. Although ELISA has long been routinely used to quantitate proteins and other cellular metabolites in microtiter plate formats, it requires a number of time-consuming plate wash and incubation steps. In contrast, fluorescent bead-based FMAT assays, i.e., fluorescence-linked immunosorbent assays (FLISA), require no plate wash steps and typically only a single incubation step. These homogeneous "mix-incubate-read" FMAT FLISAs are therefore amenable to both low and high-throughput screening applications, as well as to applications necessitating rapid turnaround of analytical data. FMAT relies on the combined power of a confocal imaging system and a He-Ne laser to measure fluorescently labeled cells or beads that have settled to the bottom of microtiter plate wells. The objectives of this tutorial are to briefly summarize the underlying photophysical principles of FMAT, describe a typical FMAT FLISA experimental protocol, and present a sample of development data from a recent high-throughput protein assay project.

G. S. Long; Packard BioSience Company, 800 Research Parkway, Meriden, CT 06405

AlphaScreen is a versatile and adaptable screening chemistry that allows for reliable measurement and analysis of a wide range of different types of biomolecular interactions, activities and affinities. AlphaScreen relies on the use of "Donor" and "Acceptor" beads coated with a layer of hydrogel providing functional groups for bioconjugation. When a biological interaction between molecules brings the beads into proximity, a cascade of chemical reactions is initiated to produce a greatly amplified signal. Upon laser excitation, a photosensitizer in the "Donor" bead converts ambient oxygen to a more excited singlet state. This reacts with chemiluminescent and fluorescent dyes in the "Acceptor" bead to produce light signal that is measured on an AlphaScreen plate reader.

The technology has markedly short assay development times, and is amenable for both automation and miniaturization. The principles and underlying chemistry of AlphaScreen will be explained and examples for various assay systems will be presented with particular emphasis on the broad range of targets and assay types that can be performed.

B. O. Fanger; IGEN International, Inc., 16020 Industrial Drive, Gaithersburg, MD 20877

There are many choices for detection technologies in basic research, each of which has attributes that provide performance for a limited number of applications. However, these platforms may not be applied broadly without sacrificing performance and quality of results. IGEN's ORIGEN Technology is capable of detecting a wide range of biological compounds, including proteins, small molecules, nucleic acids and microorganisms because of its revolutionary process that uses labels designed to emit light when electrochemically stimulated. This seminar will focus on applications currently used in biological research including immunoassays, enzymatic assays (kinase and HIV1 integrase) and transcription factor binding assays.

D. Bartley¹, D. Bintzler², R. Ashworth¹; ¹Johns Hopkins University, Baltimore, MD, ²University of Cincinnati, 231 Bethesda Avenue, Cincinnati, Ohio 45267-0524

The Fragment Analysis Research Group (FARG) will present a tutorial on database querying to identify genetic markers for genotyping projects. Genotyping facility personnel may be asked to identify markers that would be useful to carry out a genotyping project. Facility users may have a genetic disorder mapped out to a causative chromosome or section on a chromosome but are not sure which markers are available in that region. Facilities should be able to assist in identifying markers using available database web tools.

This tutorial will identify different web tools, show their strengths and weaknesses, and explain how to use these resources to build a successful and cost-effective genotyping study. The tutorial will also illustrate how to use this information and apply it in the laboratory in planning out reaction conditions, multiplexing and overall throughput requirements. Using this skill is one that can ultimately save an investigator time and it's ever-important counterpart money.

W. F. Patton; Molecular Probes, Inc., 4849 Pitchford Avenue, Eugene, Oregon 97402

Large sets of biological samples are commonly encountered in biomedical research, so rigorous and reliable matching of proteins from numerous 2-D gels is required. The Multiplexed Proteomics (MP) platform employs the same dye to measure proteins across all gels in the database, and reserves additional dyes to highlight specific functional attributes. The MP platform allows parallel determination of both protein expression levels and functional attributes of proteins (glycosylation, drug-binding capabilities, drug-metabolizing capabilities) within a single 2-D gel. First, a set of 2-D gels is fluorescently stained and imaged to reveal some functional attribute of the proteins, such as glycosylation. Next, SYPRO Ruby protein gel stain is used to reveal protein expression levels in the same gels. Differential display analysis is accomplished by computer, using image analysis software. Any two images can then be re-displayed as a single pseudo-colored map, allowing visual inspection of differences. Quantitative information is readily retrieved in tabular form, with differential expression data calculated.

N. A. Morrice; University of Dundee, Dow Street, Dundee, Scotland DD1 5EH United Kingdom

Mass spectrometry is now used routinely for protein identification and the identification of posttranslational modifications. It does however fall short of being the method of choice for phosphorylation site analysis. This is because phosphopeptides are often not observed in the mass spectrum of complex mixtures (e.g tryptic digests) as their signal appears to be suppressed. .We routinely use a combination of mass spectrometry and solid phase Edman sequencing to determine sites of phosphorylation from ³²P labelled proteins. By using ³²P labelling in vivo and in vitro the recovery of phosphopeptides from in-gel digestions and chromatographic separations can be determined and phosphoamino acid analysis performed . Mass spectral analysis of isolated phosphopeptides indicates their primary sequence and the release of ³²P during solid phase Edman sequencing indicates the site(s) of phosphorylation. This method , although slow compared to mass spectral methods, is virtually independent of the size of the phosphopeptide, the number of phosphorylation sites and the residue that is phosphorylated.

R. S. ANNAN; GlaxoSmithKline, Box 1539, King of Prussia, PA 19406

Much of the activity in the proteome is regulated by reversible protein phosphorylation. On the other hand, not all protein phosphorylation appears to functionally relevant. Thus the challenge for understanding how phosphorylation modulates a given biological pathway is to map all given phosphorylation sites on a protein and then determine which are the relevant ones. This presentation describes how mass spectrometry can be used to address these important issues.

D. S. Rinehart; Tripler Army Medical Center, Hawaii

In this tutorial we will discuss customer complaints as opportunities to build client-service provider relationships and improve business processes. Techniques will be presented to help managers and staff effectively resolve customer complaints and identify problems that cause service failures. Elements of a systematic approach to realize full benefits from complaints will be discussed.

Complaints are potentially valuable customer feedback unavailable from other sources. Not every dissatisfied customer gives you a chance to respond to a problem. Studies show that, even though customers are bothered by poor service and mistakes, a relative few of them take time to voice their displeasure. Many go elsewhere for service without saying anything.

A complaint means that a customer feels so strongly about an error or poor service that he/she has taken his/her valuable time to voice their displeasure. For the business, this is an opportunity to keep a customer and build loyalty, if it is handled well. If it is mishandled, the customer is apt to be forever lost. A complaint also means that something in the business process has gone wrong and unless it is fixed, the error is likely to recur.

D. A. Iacobas, A. B. Massimi, M. Urban, S. Iacobas, D. C. Spray; Albert Einstein College of Medicine, 1410 Pelham Prkwy South, Bronx, NY 10461

We introduce a complete set of procedures to maximize information obtainable in cDNA array studies. They proved effective when comparing Cx36 stably transfected and parental N2A cells, Cx43 KO and wild type mouse brains, and normal and STZ diabetic rat smooth muscle. The experimental design in three steps (calibration, variability, and score) provides global and local correction factors, estimates the biological variability of individual gene transcription levels, and calculates the expression ratios. A strong mathematical algorithm reduces drastically the effect of technical errors. The newly introduced expression score incorporates the biological variability and considerably increases the resolution of the method. In the case of Cx43 KO mouse brain for example, the classical procedure detected 35 up-regulated genes compared to the wild type brain, out of which 12 false results have been identified by the independence test. Our procedure clearly detected 187 up-regulated genes (p < 0.05), including all 23 good results based on expression ratio. The original cluster analysis allows the user to group the genes in sets of co-coordinated expression scores and facilitate identification of key genes.