TIPS ARTICLES


Use of Affi-Gel Blue or Blue-Sepharose for Cleanup of Samples From Cultured Cells Before Sequencing.

Victor Hatcher and Ruth Hogue Angeletti
Albert Einstein College of Medicine.


Sample preparation is often the most challenging part of protein sequencing. In the face of unpalatable sequence data presented by core facilities, many users give up the project in frustration, or try core facilities at other institutions, usually with similar lack of success. The work summarized here represents a successful resolution of a difficult sequencing problem by an unusually persistent investigator. Our core facility (RHA) had consistent problems sequencing PVDF samples from one of our most careful user laboratories (VH). This lab was trying to sequence several proteins isolated from subcellular fractions of cultured cells. The electrophoretic separations and stained PVDF membranes had an excellent appearance, and did yield sequence data. However, some of the sequences obtained were derived from fragments of bovine serum albumin (BSA). The original cell cultures required serum- containing medium for growth. In spite of careful washing of the cells with serum-freed medium before the homogenization and cell fractionation steps, albumin sequences were always obtained from the resulting PVDF membranes. Since further improvements in cell manipulations were not possible without altering the biological experiment, advantage was taken of the known properties of albumin. After the cell fractions had been solubilized by Triton X-100, they were passed over a small Affi-Gel Blue (BioRad) column, which binds BSA. The proteins of interest, which were eluted in the breakthrough of the column, were prepared for electrophoresis and transfer. The material which bound to the Affi-Gel Blue was eluted and analyzed by SDS-PAGE. The Affi-Gel Blue column not only removed BSA, but also removed 5-10 higher molecular weight proteins present in low concentration. The background stain of the resulting PVDF blots was much whiter. Bands excised from these blots yielded unique sequence data with good signal-to-noise ratio.

Subsequent to this successful outcome, another laboratory (RHA) was attempting to identify peptides secreted by cultured choroid plexus epithelium. This required analysis of the tissue culture medium. Although the serum levels in the medium could be lowered from 10% to 2% for short periods, this condition compromised the cells over a longer period of time. Further, these "low" serum levels still have massive amounts of albumin, overwhelming any HPLC separations attempted.

However, if the culture medium was passed over a column of Blue-Sepharose (Sigma), HPLC separations with significant detail were obtained. Specifically, transthyretin, a 12,000 Dalton protein which binds thyroxin and retinoic acid, known to be secreted only by choroid plexus and liver, was clearly identified. Although this protein was eluted from a C-18 column slightly before the peak of remaining bovine albumin, unequivocal sequence data were obtained, free from any contaminating albumin sequences.

The Blue-Sepharose gels can be purchased either as loose media or as small, prepacked columns from companies such as Sigma or Pharmacia-LKB, while Affi-Gel Blue can be purchased from BioRad. The companies provide protocols for use.1 The resins will specifically retain enzymes which have binding sites for adenylate cofactors, in addition to binding serum albumin. Some proteins may also bind by hydrophobic interactions. Generally, the samples are loaded at pH 7.5-8.0 and low salt concentrations. If elution of the bound proteins is desired, then 1.5 M NaCl in buffer, chaotropic agents, or more specific eluants can be used. In the above examples, where the primary need is to remove the albumin, there is no need to elute unless the resin will be reused. The breakthrough of the column is the desired sample. We have found we can either load tissue culture media directly, at physiological salt concentrations, or after diluting it 1:1 in distilled water. For the first application above, a 3 ml column is sufficient. Small fractions were collected, and the samples were run on SDS-PAGE without further concentration. In the second application, 10 ml columns packed in the laboratory were used, one for the control and one for the experimental medium. When one ml of the breakthrough was injected directly on a 2.1 mm x 23 cm C-18 column using a HP1090 HPLC, a transthyretin peak of more than 0.125 absorbance at 214 nm was obtained.

1 Additional information can be found in Hermanson, G.T., Mallia, A.K., & Smith, P.K. (1992) Immobilized Affinity Ligand Techniques, Academic Press, pp 454.


Return to the The ABRF Home Page


Created: 29th July 1995
Last modified: 29th July 1995