Methods and reagents: Purification of GST fusion proteins

Methods and reagents is a unique monthly column that highlights current
discussions in the newsgroup bionet.molbio.methds-reagnts, available on the
Internet.  This month's column discusses some common problems associated with
the construction of protein fusions with glutathione S-transferase and with
hybridization membranes. For details on how to partake in the newsgroup, see
the accompanying box.

In past discussions, several netters have reported seeing bands that correspond
to larger proteins than expected when using glutathione S-transferase (GST)
fusions to purify their favorite protein. For example, when affinity purifying a
GST fusion protein synthesized in Escherichia coli using a glutathione-sepharose
4B matrix from Pharmacia, one netter observed an extra band indicating a protein
of approximately 69 kDa in addition to the expected fusion protein of 52 kDa.
This person thought that the extra protein could have been a contaminant that
bound GSH, or that it might have come from the binding matrix. However,
pre-washing the GSH matrix with 1 M NaCl did not help eliminate the extra band.
As the 52 kDa band was later confirmed to be the fusion protein, because it
reacted with an anti-GST antibody and the larger one did not, this netter could
only explain the presence of the larger protein by concluding that it came from
E. coli and somehow could co-purify with GST.

Another netter also found that after purifying a GST fusion protein using the
GST-pGEX system, extra bands appeared that were larger then expected.  Although
the fusion product was also present, it was at a very low concentration compared
with the others.  This person thought that the extra proteins could have
resulted from recognition of an exogenous promoter on the pGEX vector, but was
unsuccessful in eliminating the extra bands by varying the IPTG concentration
from 0.1 mM to 1.0 mM during the induction.

Another person had a similar problem. In addition to an 18 kDa fusion protein, a
band of approximately 32 kDa was seen on an SDS-PAGE gel. Although not fully
explaining how proteins less than twice the molecular mass of the expected
fusion protein could appear on the gel, netters agreed that this could be caused
by dimerization of the fusion product. Some wrote that this was likely the
reason since GST is known to form a homodimer in solution. [2] Others pointed
out a study in which domain 1 (D1) of rat CD2, a transmembrane glycoprotein
present on the surface of T cells, was synthesized in E. coli as a GST fusion.
Stable dimers of the CD2 D1 protein were presumably formed through interactions
between the fused GST segments. [3]

Similarly, when fusions were made between GST and the human Bcr-Abl locus,
multimeric forms of the fusion protein were observed in SDS-PAGE gels after
crosslinking with glutaraldehyde, showing that some proteins allow the formation
of aggregates after being fused to GST. [4]

Avoiding proteases
******************
Some netters also described having the opposite problem - GST fusion proteins
that are smaller than expected.  One person who tried to produce a fusion
protein of approximately 64 kDa found that only rarely did the protein appear to
be the correct size. After many attempts to synthesis the protein from the same
clone, a smaller protein of 50 kDa was routinely isolated. This can be explained
more easily than the appearance of larger proteins because smaller ones could be
the result of protease activity.  Some think that the engineered thrombin
cleavage site or some other cleavage site within the recombinant protein might
be recognized by an unidentified protease present in E. coli lysates, and that
this is especially evident in the E. coli strain BL21.  To avoid this kind of
protease activity, netters suggest that the following precautions can help
reduce the cleavage:

(1) Use a relatively fresh transformant made using plasmid DNA stored in the
freezer.  The bacteria should be no more than one month old and should not be
streaked from glycerol stocks because this has been found to be unreliable for
proper synthesis of GST fusions. Also, several different strains of E. coli
should be used.

(2) Grow the E. coli in the presence of 0.25-0.5% glucose to keep expression
from the tac promoter to a minimum before induction with IPTG.

(3) Grow the cells at the lower temperature of 22-30 degrees C.  This will most
likely lower the yield of recombinant protein, but should lessen the activity of
the unknown protease and give a higher proportion of undegraded protein.

(4) Keep the induction time relatively short (2-4 hours).

(5) Include a protease inhibitor when lysing the cells.

One netter wrote that mixing in 10 mM EDTA and 0.1 mM benzamidine and keeping
everything cold, or using 1 mM phenylmethylsulfonyl fluoride (PMSF), a serine
protease inhibitor, did not help remove smaller, apparently degraded proteins.

Care should be taken when using PMSF as its ability to inhibit proteases is pH
dependent, and it has a very short half life. However, this is good news,
because on one hand PMSF could inhibit either thrombin and Factor Xa in the
cleavage step and, on the other hand, it could be inactive by the cleavage
stage.  Even so, most netters suggest using a cocktail of protease inhibitors in
all buffers, and that the mix should include those active against different
classes of proteases, including cysteine proteases.

Churchgoers
************
Netters are continuing to adjust to the change over of membrane formulations
that occurred over the past year (see TiBS 20,522-523 for a discussion of
membrane SWAPs). Recently, there was a flurry of postings about problems netters
are having with high background radiation when preforming blots with the new
formula Hybond N membrane from Amersham.

Although Amersham has clearly labeled their boxes of membranes with a warning
that states the chemical composition of the membrane is now different, some
people are still unaware of the switch and are perplexed by changes in their
results when using the new membrane under their old conditions. It is estimated
that about one-third of Hybond N users are likely to face problems with the new
membrane and might have to find a new set of conditions for best results.
Unfortunately, the end-users seem to be stuck with the unenviable job of working
out the best conditions for the newer membranes on the market.

Mike Gruidl (mgruidl@com1.med.usf.edu) recently wrote that he was getting
excellent blots with the older Hybond N membrane when pre-hybridizing in 2x
Denhardt's/5x SSC, and then hybridizing in 2x Denhardt's/5x SSC/0.3% SDS plus
50-100 ug ml-1 salmon sperm DNA as carrier, but that under these same conditions
the newer membrane produced a tremendous blotchy background. The high background
could not be removed by washing in 0.1% SSC/0.1% SDS at 65 degrees C, or even by
boiling the membrane in 0.5% SDS.  After switching to a different hybridization
solution which includes pyrophosphate, he found that the background problem went
away.

Another netter wrote that he has been using the new Hybond N+ membranes for
Southerns without any background problems after switching from a hybridization
solution that contained formamide to one that is composed of 25 mM Tris, 1M
NaCl, 10% dextran sulfate, 1% SDS, pH 7.4. It is unclear if formamide is the
reagent that is causing the trouble, and a complete study of all factors and
components that affect the hybridization and signal-to-noise ratio has not been
done.  It could simply be coincidence that these two people were having trouble
with solutions containing formamide.  In any case, it might be worth avoiding
formamide if background problems persist.

Netters now recommend that Church's buffer composed of 0.5 M NaHPO4, 1 mM EDTA,
7% SDS, 1% BSA, pH 7.2, [Ref. 5] be used for hybridizations with the new Hybond
N membrane.

References
**********

[1] Smith, D. B. and Johnson, K. S. (1988)
    Gene 67,31-40

[2] Ji, X. et al. (1992)
    Biochemistry 31,10169-10184

[3] Murray, A. J. et al. (1995)
    Proc. Natl. Acad. Sci. U.S.A. 92,7337-7341

[4] Maru, Y. et al. (1996)
    J. Biol. Chem. 271,15353-15357

[5] Church, G. M. and Gilbert, W. (1984)
    Proc. Natl. Acad. Sci. U.S.A. 81,1991-1995

*******************************************************************************
Any statements made by the author are not meant to advocate the use of a
particular commercial product or endorse any company. All opinions are those of
the author and do not reflect the opinion of the National Cancer Institute or
the National Institutes of Health.

Copyright: This manuscript is not copyrighted by Elsevier Publishing Company.
However, you may not reproduce any portion for resale or edit the text for
redistribution, sale, or otherwise without written permission from the author.

You found this at the World Wide Web (WWW) Uniform Resource Locator (URL)
ftp://ftp.ncifcrf.gov/pub/methods/TIBS/oct96.txt

Any reference to this column must be cited as the following published article:
Hengen, P. N. 1996. Methods and reagents: Purification of GST fusion proteins
Trends in Biochemical Sciences 21(10):400-401.

*******************************************************************************
* Paul N. Hengen, Ph.D.                           /--------------------------/*
* National Cancer Institute                       |Internet: pnh@ncifcrf.gov |*
* Laboratory of Mathematical Biology              |   Phone: (301) 846-5581  |*
* Frederick Cancer Research and Development Center|     FAX: (301) 846-5598  |*
* Frederick, Maryland 21702-1201 USA              /--------------------------/*
*******************************************************************************