Created: 1st December 2000, last updated: 30th December 2000, © 2000 ABRF
Jennifer L. Ware, Laurie Moran, Ching-Lun Lin, and Barton Slatko
New England Biolabs, Inc., Beverly, MA
New England Biolabs (NEB) sequencing core facility provides automated sequencing services to support various company-wide projects in house, but on a very small scale of about 1000 to1500 reactions per month. A procedure has been implemented at the NEB core sequencing facility to integrate simplified methods and robotics to provide a more efficient small-scale process. This has been done using a Beckman Biomek 2000 robot combined with an MJ DNA Engine, 96-well plate cycler (PTC-200), AB 373 and 377 sequencers, BMA Singel gels, and several other materials that help reduce the time required for otherwise lengthy procedures in a cost-efficient manner. Protocols have also been developed for efficient sequencing of a variety of templates submitted to the NEB core facility. (J Biomol Tech 2000;11:151-154)
Key Words: automation, robotics, sequencing template cleanup, small-scale DNA sequencing.
Address correspondence and reprint requests to: Jennifer L. Ware, New England Biolabs, Inc., 32 Tozer Road, Beverly, MA 01915 (email: ware@neb.com).
Genome sequencing and analysis technology has advanced in a manner that allows companies to generate enormous amounts of data per day, via automated processes, in an effort to complete genome projects and use that information as a source of possible disease treatment and control. Unfortunately, smaller core facilities, which have less of a need to produce such high-throughput data, are not the primary targets of application of these advanced technologies and are in need of simplified methods. The core sequencing laboratory at New England Biolabs, Inc. (NEB, Beverly, MA) has combined recent advances in sequencing technology to efficiently simplify the process from beginning to end. The sequencing core facility provides additional measures of quality control of appropriate DNA products for sale at NEB, such as sequencing primers and other oligonucleotides, cloning vectors by sequencing reaction assay, and vector and phage display products to ensure correct sequence. For our in-house research program, sequences are required from vector and overexpression constructs of clones, cDNA, polymerase chain reaction (PCR), phage display, and other sequencing templates, including our involvement in the Filarial Genome Project (http://www.neb.com/fgn/filgen1.html). Among those templates submitted for analysis are plasmids, M13 derivatives, lambda phage, PCR products, cosmids, and bacterial artificial chromosomes (BACs), each of which requires a different amount of template and/or primer per reaction and sometimes different sequencing reaction conditions as well.
To make this process efficient from beginning to end, researchers start by submitting an electronic form to the core facility in which key information necessary to perform the sequencing is provided. The form is accessible on the DNA Sequencing Core Facility intranet site (on a Unix server) through the user's desktop Macintosh or PC. The intranet site provides the home base for submission and receipt of sequence information to and from the core facility. By selecting the "submit your template to DNASEQ lab" icon, the researcher can fill out an html form in which they describe the project, identify the type of template being submitted, and list any special instructions concerning the reaction (eg, "please include DMSO; high G+C content"). It is necessary that the instructions to the sequencing core be precise. We require both the templates and primers be submitted at the proper concentrations for the reaction conditions. The core facility maintains a stock of common "standard" primers such as T3, T7, SP6 promoter primers, and pUC-based forward and reverse primers, and the researcher need not provide them. The form also asks the submitter about the method of purification of the template to enable the core facility to track statistics of success and failure with different methods and to identify any failure correlations. In addition, the form requires some substantiation of DNA concentration to ensure that the submitter has indeed verified template concentration. One major class of reaction failures is the result of too much or too little DNA. A second class is suboptimal (or no) sequence data, usually attributable to template impurity or primer design issues. We consider a failure any sequence that does not give a reasonable amount of sequence (eg, 500-600 "good" nucleotide base calls for double-stranded plasmid DNA templates).
After the form is submitted to and received by the core sequencing facility, a message to the sender thanks them for their submission. If the form is not completely filled out, it will not be submitted, and an error message appears to the submitter identifying the problem and asking them to "try again."
We chose a Unix-based server as the site for submission and return of sequence files because of its faster speed and higher data-handling capabilities, especially when downloading sequences to the end user. We also run CAP (Columbia Appletalk Package for Unix) on our Web server so the sequence data from the sequencers can be uploaded easily. The submission-data porting script was written at NEB and provides a rapid and simple way to handle data electronically.
After the submission form is sent, the researcher physically submits the template and primer in a small plastic bag with their name on it by placing it in a box marked "TEMPLATES IN" in a small freezer in the core facility. Template and oligonucleotide are submitted in pairs--for every template submitted there must be a companion primer tube, and vice versa. The sequencing reactions are then performed according to the instructions on the submission sheet using the corresponding primers and templates submitted to the core facility. After the sequence data are obtained, and retracked and recalled if necessary, a copy of the raw data is directly ported through the server back to a folder automatically created at the DNA Sequencing Core Facility intranet site. Users simultaneously receive an email message, delivered to the email address they submitted on the form, stating, "Your DNA sequence data is ready." By returning to the intranet site and choosing the "pick up your sequence" icon, the researcher can go directly to the server to download the sequences to his or her own computer for analysis. Sequenced templates and primers can then be picked up by the user from the core facility freezer "TEMPLATES OUT" box.
The NEB core sequencing facility uses the Applied Biosystems (Foster City, CA) AB 373 and 377 automated sequencers to process approximately 84 to 96 dye terminator samples per day. Dye terminator sequencing chemistry is used to accommodate the variety of template types and vector constructs being sequenced and to maximize the number of primers that can be used.
The Biomek 2000 Robot (Beckman Instruments, Palo Alto, CA) is used to assemble the components of sequencing reactions. Using adaptors for 1.5-mL centrifuge tubes, the robot can build any number of desired reactions so that in-house customers can submit in standard-sized tubes. The reaction components are transferred directly into a 96-well tray on the MJ cycler with a Power Bonnet heated lid (MJ Research, Watertown, MA) attached to the robot and thermally cycled immediately on addition of all reagents, as specified by a preset program.
The robot has limitations that affect volume per tube, namely, a lack of ability to withdraw small volumes of reagent. It is necessary to provide at least 10 µL of reagent if 5 µL is to be withdrawn by the robot. Similarly, for 1 µL to be withdrawn, 5 µL must be present. For our sequencing protocol, double-stranded and single-stranded DNA templates must be at 100 ng/µL (5 µL is used, thus 10 µL must be submitted), and PCR products must be at 8 ng/µL (5 µL is used, thus 10 µL must be submitted). We require oligonucleotide concentrations to be 3.2 pmol/µL for double-stranded DNA, and 5 µL must be provided per primer tube although only 1 µL is added per reaction. For single-stranded DNA, primers are submitted at 0.8 pmol/µL with 5 µL per primer tube; again, only 1 µL is added per reaction.
To use the robot to assemble the components of these reactions, a simple program was composed to transfer samples between the robot benchtop platform and the 96-well plate on the connected MJ cycler. A script is available from Beckman, as well as from MJ Research, that allows the cycler to close the heated lid and start cycling on reaction setup completion.
We have found that reactions containing one half of the manufacturer's prescribed quantity of Amplitaq DNA Polymerase FS premix (AB PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit using 4 µL premix per reaction with the AB recommended amount of DNA template and primer), yielding a total volume of 10 µL, works better and is more cost-efficient than the standard AB protocol that includes 8 µL premix in each reaction, yielding a total volume of 20 µL.1
Protocols have been designed and optimized for sequencing reactions other than standard double-stranded or single-stranded DNA templates. These sequences are done manually and cycled on an MWG cycler (Biotech, High Point, NC), but can be done on the robot with the appropriate programming.
With Amplitaq FS terminator chemistry, BAC sequencing requires a large-scale reaction for maximum output of at least 500 bases per reaction and minimum background. Sixteen microliters of premix is combined with 1 µg of DNA template and 50 pmol primer and dH2O to give a total volume of 40 µL.2
Cosmid and lambda phage reactions are performed using 500 ng DNA (the standard amount of DNA used in double-stranded and single-stranded reactions), but with 12.8 pmol of primer (approximately 4 times the standard 3.2 pmol amount) and 8 µL premix to provide a total reaction volume of 20 µL.
For reaction product purification, Centrisep spin columns (Princeton Separations, Adelphia, NJ) are used on an individual basis for the reactions performed manually, as described in the product manual. A 96-well system (Edge Biosystems, Gaithersburg, MD) is used for product purification of the robot-prepared automated sequencing reactions performed in the 96-well cycler. Product purification is also done with the assistance of the robot. A simple program was written to transfer the cycled sequencing reactions directly onto the center of each well of a 96-well separation plate with gel bed of hydrated sephadex mixture. After all reaction mixtures are transferred to this separation plate, the plate is centrifuged at 800 rpm in a Beckman GS-15 microtiter tray centrifuge with a collection tray underneath it to collect the purified sequencing sample. This tray is then placed in a Jouan R6-1010 vacuum centrifuge (Jouan, Winchester, VA) to dry the samples. The dried samples, as well as those dried samples prepared via Centrisep columns, are then resuspended in 2-µL Accutrac dye (3 µL for the AB 373 sequencer; Commonwealth BioTechnologies, Richmond, VA), which provides a unique dye marker for each lane on the gel image regardless of whether or not the reaction has failed. This reduces software errors in lane tracking. After the samples are resuspended in dye, they are vortexed and centrifuged at maximum speed for 30 seconds and then heat denatured at 90°C for 2 minutes. The samples are then loaded onto the sequencers in a staggered manner in which the odd samples are loaded first and subjected to electrophoreses for 10 minutes, and then the even samples are loaded.
Biowhitaker prepackaged Singel gels (Long Ranger Singel packs for AB 373 and 377; BMA, Rockland, ME) are used as a simple, efficient alternative to making our own gel solutions. It takes 12 minutes to prepare the gels for pouring. Using mild alkanox detergent and isopropanol to clean the glass plates, the gel solution is placed in a side-arm flask for pouring onto the plates using an Owl Scientific Otter sequencing gel caster (Owl Separation Systems, Portsmouth, NH). A period of 2 to 2.5 hours is necessary to allow the gels to completely polymerize; other steps in the process are performed during this period.
The core facility currently processes 1000 to 1500 sequence reactions per month, with about 60 facility users submitting templates for sequencing. Data from each sequence run is immediately backed up on a gigabyte tape drive and copied onto a CD for permanent storage. In addition, a summary of the sequencing results is transferred to tape for short-term storage. Gel image files are kept for 3 days to enable retracking or for use of alternative base-calling programs, if necessary.
When novice users first submit templates for sequencing, they are given a written set of instructions concerning template preparation and quantitation. They are also instructed in how to configure their Netscape Navigator (Netscape Communications Corp., Mountain View, CA) to download sequences from the intranet site. Netscape Navigator must be configured using a sequence trace editor to view the sequence chromatogram files generated by the AB sequencers. They are also shown how to use the analysis programs. This information is also present on the intranet site, along with information on sequencing chemistry and helpful hints on reading sequences. Numerous copies of computer programs for analysis are mounted on individual Macintosh or Windows-based PC computers with access by limited user-number licenses. In addition, each user is given a copy of EditView for Macintosh users (available by FTP from the AB Web site) or Chromas for PC users (Techlysium, PTY LTD, Helensvale, Australia) to do simple editing and text file saving. The core facility provides support and training for certain computer programs for assembling sequencing projects, such as Autoassembler (AB), Sequencher (Gene Codes, Ann Arbor, MI), and GCG Programs (Pharmacopeia, Princeton, NJ). Other Mac- and PC-based programs are also in common use.
Despite the focus on high-throughput automated sequencing, many laboratories, such as NEB, have a medium-throughput level. From the submission/data porting script to the template cleanup, we have designed simplified methods to make each step more efficient and less time consuming for a small-scale core facility. Using this system has enabled us to efficiently perform DNA sequencing service, generally achieving a 2-day turnaround between submission and return of sequence information.
This work was supported by New England Biolabs, Inc. (NEB). We sincerely thank the computer group at NEB for their essential aid in the core facility. All programs mentioned in this article can be obtained from Beckman, MJ Research, or NEB.
1. ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit with Amplitaq DNA Polymerase FS Protocol, revision A. Foster City, CA: Applied Biosystems, August 1995.
2. Boysen C, Simon M, Hood L. Fluorescence-based sequencing directly from bacterial and P1-derived artificial chromosomes. BioTechniques 1997;23:978-982.