Full text PDF - Biotechnology and Applied Biochemistry

Biotechnol. Appl. Biochem. (2010) 57, 127–138 (Printed in Great Britain) doi:10.1042/BA20100279 127 

ETRAP (efficient trapping and purification) of target protein 

polyclonal antibodies from GST–protein immune sera 1 

Dan L. Crimmins 2 , Nancy A. Brada, Christina M. Lockwood, Terry A. Griest, 

Rachel J. Waldemer, Mark A. Cervinski, Matthew F. Ohlendorf 3 , Jay J. McQuillan 

and Jack H. Ladenson 

Washington University School of Medicine, Department of Pathology and Immunology, Division of Laboratory and Genomic 

Medicine, 660 South Euclid Avenue, Box 8118, Saint Louis, MO 63110, U.S.A. 

Recombinant GST (glutathione transferase) proteins 

are widely used as immunogens to generate polyclonal 

antibodies. Advantages of using GST proteins include: 

commercially available cloning vectors, vast literature 

for protein expression in Escherichia coli, the ease 

of protein purification, immunogen can be used as 

an ELISA standard and GST can be removed in 

some systems. However, there are disadvantages: GST 

oligomerization, inclusion body formation and target 

protein insolubility after GST removal. Perhaps the 

most detrimental is the significant generation of anti- 

GST antibodies by the host animal. A two-column 

procedure using a glutathione-GST column and a 

glutathione-(GST–protein) column can yield affinitypurified 

anti-(GST–protein) polyclonal antibody. Several 

passes over the first column are often required, 

though, to completely extract the anti-GST antibodies 

from the immune sera. We reasoned that knowledge 

of the target protein linear epitope(s) would allow 

construction of a peptide affinity resin for a singlepass 

‘one and done’ purification termed ETRAP 

(efficient trapping and purification). In the present 

paper, we describe our efforts and present data on 

rabbits and sheep immunized with GST proteins 

having target protein molecular masses of ∼8, 21 and 

33 kDa. The titre and purity of the target antibodies 

using the ETRAP protocol were comparable to the 

more laborious multi-column purifications but with a 

considerable saving in time. 

Introduction 

Polyclonal antibodies are typically produced in medium to 

large animals such as rabbits, sheep, goats and donkeys 

[1,2]. There are essentially two types of amino acid 

immunogens, peptides attached to carrier proteins and 

large fragments of or full-length protein targets. Affinity 

purification of the corresponding immune sera will yield 

anti-peptide and anti-target protein antibodies respectively. 

www.babonline.org 

It is relatively simple to prepare the immunogen for the 

former as a synthetic peptide [3] that is then conjugated 

to a carrier protein, e.g. keyhole limpet haemocyanin, BSA 

or ovalbumin [3,4]. We and others have observed that 

these anti-peptide antibodies are highly successful reagents 

for Western blot analyses, low to moderately useful in 

immunohistochemistry and have a poor to low favourable 

rating in ELISAs [5,6] and immunoprecipitation [1]. In 

contrast, the anti-protein antibodies have a moderate to 

good rate of success in immunohistochemistry, ELISAs [5,6] 

and immunoprecipitation [1] in addition to a high proficiency 

in Western blotting. However, the immunogen must be 

prepared from a purified protein or, as frequently is the 

case, a recombinant protein such as a GST (glutathione 

transferase)-fusion product [7]. We employ a dual strategy 

of making anti-peptide antibodies that are primarily used in 

Western blot analyses of target tissue homogenates to verify 

the presence of the target protein and recombinant GST– 

protein antibodies predominantly used for sandwich ELISAs. 

Such an approach has the advantage of rapidly generating 

anti-peptide antibodies in approx. 2 months while the cloning 

and expression conditions are being developed. Ultimately 

we wish to develop sandwich immunoassays for diagnostic 

purposes, e.g. Creatine Kinase MB [8] and cardiac Troponin 

I [9]. Such assays require highly characterized and purified 

antibodies. 

Identification of antibody epitopes is a major asset 

in their overall characterization [10]. A minimum of two 

non-competing antibodies are required for sandwich ELISAs 

[5,6] and knowledge of the antibodies’ epitopes greatly 

Key words: epitope mapping, GST–recombinant protein immunogens, 

one-step affinity chromatography. 

Abbreviations used: ETRAP, efficient trapping and purification; GST, 

glutathione transferase; the standard three letter code for the amino acids 

is used. 

1 Portions of this work were presented at PepTalk 2010 in San Diego, CA in 

January 2010. 

2 To whom correspondence should be addressed (email 

crimmins@pathology.wustl.edu). 

3 We dedicate this article to the memory of our dear friend and colleague 

Matt Ohlendorf (1967–2009). 

C○ 2010 Portland Press Limited 

Biotechnology and Applied Biochemistry

128 D. L. Crimmins and others 

facilitates selection of the appropriate antibody pairs. For 

anti-peptide antibodies, the epitope is contained in the 

sequence of the synthetic peptide [3]. Affinity purification 

of the desired antibodies with concomitant removal of 

anti-carrier protein antibodies is then rather trivial by 

simply using the cognate peptide attached to resin [2– 

4]. Animals that have been immunized with GST protein 

will almost assuredly possess multiple epitopes spanning 

different regions of the entire target sequence in addition 

to anti-GST antibodies. Purification then requires a more 

involved route by removing the anti-GST antibodies with 

a GST–glutathione column prior to affinity purification of 

the desired target antibodies using the immunogen attached 

to a glutathione resin [11]. The non-bound flow-through 

fraction from the GST–glutathione column must be assayed 

for anti-GST activity before the affinity purification step is 

performed. If anti-GST activity is present, then multiple 

cycles of anti-GST removal are obligatory, thus increasing 

the overall purification time. By leveraging our knowledge 

of the epitope assets, we constructed an epitope peptide 

resin essentially similar in principle to those prepared 

for anti-peptide antibody affinity purification described 

above. The epitopes were chosen from only intensely 

immunostained spots on spot-peptide membrane arrays 

[12], typically two to four linear sequences in number. 

This strategy termed ETRAP (efficient trapping and 

purification), permits a single-column ‘one and done’ 

procedure for both removing anti-GST antibodies and 

isolating those antibodies that have the highest reactivity to 

specific epitopes within the target protein. Three examples 

of the ETRAP protocol are described for the immune 

sera of rabbits and sheep immunized with recombinant 

GST proteins having target protein molecular masses of 

∼8, 21 and 33 kDa (referred to hereafter as 8K, 21K and 

33K). It will be shown that the titre and purity for the 

target protein antibodies obtained by ETRAP are congruent 

with those acquired from the time-consuming two-column 

purification scheme. Furthermore, since the ETRAP process 

delineates the epitopes of these antibodies, the development 

of sandwich ELISAs are facilitated. 

Materials and methods 

General 

Peptides were synthesized at Biomolecules Midwest 

(Waterloo, IL, U.S.A.), the secondary antibodies goat 

anti-rabbit alkaline phosphatase (111–005-047) and rabbit 

anti-sheep alkaline phosphatase (313–055-047) were 

purchased from Jackson ImmunoResearch Laboratories 

(West Grove, PA, U.S.A.), and 5-bromo,4-chloro,3indolylphosphate/nitrobluetetrazolium 

was obtained from 

Kirkegaard and Perry Laboratories (Gaithersburg, MD, 

U.S.A.). Pierce Chemical Company (Rockford, IL, U.S.A.) 


supplied the glutathione resin (No. 78250), glutathione precoated 

96-well plate (No. 15140), the glutathione-GST resin 

(No. 20205) and the SulfoLink resin (No. 44999). 

Cloning 

Nucleotide sequences encoding the 8, 21 and 33K proteins 

were inserted into pGEX-4T-1 (GE Healthcare Biosciences, 

Pittsburg, PA, U.S.A.) for the production of protein in 

Escherichia coli. 

The recombinant proteins were purified with immobilized 

glutathione following the manufacturer’s protocol. 

Quality control of the expressed proteins and calculation 

of the molar absorption coefficient (ε) at 280 nm have been 

described previously [12]. 

Antibody production 

Two rabbits and two sheep were each immunized with the 

same recombinant GST protein at Harlan Bioproducts for 

Science (Madison, WI, U.S.A.). Pre-immune and immune sera 

were collected as per the manufacturer’s protocol. Following 

purification (see below), the purified antibodies were titred 

on a glutathione-GST plate and on a covalently attached 

cognate (GST protein) to a glutathione plate. Antibodies 

were stored in PBS, pH 7.2, containing 0.05 % (w/v) NaN 3 

at 4 ◦ C. Antibody solutions were scanned from 240 to 

340 nm with protein concentration estimated using ε 280 = 

1.35 ml/(mg·cm). 

Antibody purification with the standard protocol 

Three to five ml of the immune sera was applied to a 

glutathione-GST column according to the product inserted 

to remove anti-GST antibodies. The bound material was acid 

eluted and the column re-equilibrated for another cycle of 

purification if necessary. The flow-through was assayed for 

anti-GST activity using a direct binding ELISA to glutathione- 

GST (see below). This process was repeated until no 

detectable anti-GST activity was present in the flow-through 

after which, this material was applied to a glutathione-(GST– 

protein) column as specified by the manufacturer to affinity 

purify the target antibodies. The affinity-purified antibodies 

were subjected to direct binding ELISAs on glutathione-GST 

and glutathione-(GST–protein) plates as described below. 

Antibody purification with the ETRAP protocol 

Following epitope mapping (see below), the sequences 

corresponding to the most intensely immunostained 

spots were tabulated. Typically two to four linear sequence 

ranges were readily identified. Synthetic peptides were 

prepared to span these single-epitope 10–20 amino acids 

and contained a glycine tripeptide as an N-terminal addition 

followed by an N-terminal cysteine residue. For multiple 

epitope columns, Gly-Gly-Ser-Gly-Gly spacers were

Figure 1 Scheme for the standard protocol for the affinity-purification of GST–antigen polyclonal antibodies 

See the Materials and methods section for column preparation. Abs, antibodies; flow-thru, flow-through. 

inserted between epitopes, glycine tripeptides were placed 

at each terminus, and an N-terminal cysteine residue was 

added for coupling to the SulfoLink resin. A 5 mg portion of 

the epitope peptide was added to 2 ml of the SulfoLink resin 

for covalent attachment using reaction conditions according 

to the manufacturer. We determined that the reaction was 

quantitative by C 18 reversed-phase HPLC of peptide aliquots 

before and after the coupling reaction. A 3–5 ml volume 

of the appropriate anti-serum was applied to the SulfoLink 

column, and the non-bound flow-through fraction and the 

acid-eluted bound fraction were titred for anti-GST antibodies 

and anti-(GST–protein) antibodies on glutathione- 

GST and glutathione-(GST–protein) plates respectively. 

Direct binding ELISA to glutathione-GST and 

glutathione-(GST–protein) 

These immunoassays were performed according to the 

manufacturer’s instructions using microtitre plates precoated 

with glutathione. Briefly, wells were washed with 

Tween/saline, coated with 1 μg/ml recombinant E. coli 

GST or GST–target protein diluted in TBS (Tris-buffered 

saline), and incubated with shaking for 1 h at room 

temperature (20 ◦ C). Wells were washed and serial dilutions 

of column flow-through or purified antibody were applied 

in duplicate and incubated for 1 h. After washing, an 

appropriate anti-species detection antibody labelled with 

alkaline phosphatase was applied and incubated for 1 h. 

Wells were washed and the plate was developed with the 

AttoPhos reagent (Promega, Madison, WI, U.S.A.) according 

to the manufacturer’s instructions. Results are plotted in 

the Figures as signal/noise, where signal is the fluorescent 

‘One-and-done’ polyclonal antibody purification 129 

response for the 1 μg/ml sample divided by noise defined as 

the response for a buffer-only sample. 

Epitope mapping 

Antibody linear epitopes were assigned from spotpeptide 

membrane arrays synthesized at MIT Biopolymers 

Laboratory (Cambridge, MA, U.S.A.) [12]. Each spot 

comprised a 10-mer peptide anchored to the membrane 

at the C-terminus. For the 8K and 21K target proteins, spot 

one corresponded to residues 1–10, spot 2 to residues 2– 

11, spot 3 to residues 3–12, etc., until the entire sequence 

of the target protein was covered. A two-residue offset 

was used for the 33K target protein such that spot 2 

corresponded to residues 3–12, spot 3 to residues 5–14, 

etc. Primary antibodies were used at 4 μg/ml for 4 h and the 

secondary antibodies, goat anti-rabbit alkaline phosphatase 

or rabbit anti-sheep alkaline phosphatase, at 1/1000 for 1 h. 

The membrane was developed with 5-bromo,4-chloro,3indolylphosphate/nitrobluetetrazolium 

substrate. 

Results and discussion 

Antibodies to 8K target protein 

We adapted a standard protocol for the affinity-purification 

of GST–antigen polyclonal antibodies [11]. Figure 1 depicts 

such a scheme for this two column process. A 3–5 ml 

of polyclonal anti-sera was applied to a glutathione-GST 

column to remove anti-GST antibodies. The non-bound 

flow-through was assayed for anti-GST activity using a direct 

binding ELISA format as above. If this fraction was positive 

for these antibodies then the column was regenerated and 

the flow-through was reloaded for another cycle of anti-GST 

C○ 2010 Portland Press Limited


Figure 2 The standard protocol for the removal of GST antibodies and purification of 8K target protein antibodies 

Elution profiles of (A) removal of rabbit anti-GST antibodies with a glutathione column; (B) purification of rabbit anti-(GST–8K) antibodies with a 

glutathione-(GST–8K) column; (C) removal of sheep anti-GST antibodies with a glutathione column; and (D) purification of sheep anti-(GST–8K) antibodies 

with a glutathione-(GST–8K) column. Fractions of approx. 1 ml were collected for all. Five (A) and four (C) cycles of anti-GST antibody removal were necessary 

for each antibody preparation. Acid-eluted proteins from (B) and(D) were buffer exchanged and concentrated. 

antibody removal. This procedure was repeated n times until 

the level of these non-target antibodies was minimal. Each 

complete regeneration cycle (column plus flow-through 

assay) required up to 8 h when using the columns and 

reagents supplied with the kits. 

An example of this process is illustrated in Figures 2(A) 

and 2(C) for rabbit and sheep anti-sera removal of anti- 

GST antibodies from GST–8K immunization respectively. 

For the rabbit, five rounds of glutathione-GST column 

and corresponding direct binding ELISA were necessary to 

deplete the anti-GST antibodies to an acceptable level. A 

similar acceptable level was achieved for the sheep antisera 

in this case after fours cycles. It is likely that a 

larger glutathione-GST column could reduce the number of 

elution/regeneration cycles following our standard loading 

conditions. Note, however, that this would not obviate the 

second chromatographic step in the standard protocol used 

to affinity-purify the target protein antibodies. 

The flow path for the purification scheme (Figure 1) 

followed the ‘No’ arrow to a glutathione-(GST–antigen) 

column. Here the desired target antibodies bound to the 

column with other serum constituents including non-target 

antibodies in the flow-through fraction. Following acid 

elution, buffer exchange and concentration, the desired 

antibody preparation is completed. The elution profiles for 

anti-(GST–8K) antibodies are displayed in Figures 2(B) and 

2(D) for rabbit and sheep respectively. These antibodies 

were then titred for anti-GST and anti-(GST–8K) activity 

using the appropriate direct binding plate assays. Thus using 


the kit-supplied columns and reagents, this process required 

several days to obtain purified target 8K antibodies that are 

devoid of anti-GST antibodies and which can be applied 

in sandwich ELISA development that uses GST–8K as a 

standard. By using a ‘cleavable’ GST–protein and attaching 

this cleaved protein to a resin it should be possible to obtain 

purified antibodies in one step. In our hands though, both 

the recovery of the cleaved protein after proteolysis and its 

solubility were in general poor. It is difficult to accurately 

assess overall recovery with the standard protocol because 

of slight differences in column volumes, fraction volumes and 

variable losses with each round of purification and antibody 

concentration. One can calculate, however, the amount of 

material obtained in the final preparation compared with the 

initial anti-sera load. These data are listed in Table 1 for all 

the examples presented herein. 

Epitope determination is a major part of antibody 

characterization [10]. We have previously used spot-peptide 

membrane array epitope mapping to successfully determine 

the linear epitopes of standard protocol affinity-purified 

antibodies [12]. Typically, one observes two to four intensely 

immunostained regions for these polyclonal affinity-purified 

antibodies. Figure 3 shows the results of such an experiment 

for rabbit (Figure 3A) and sheep (Figure 3B) antibodies to 

the target 8K protein. Each spot contains a continuous 

10-mer peptide such that spot one corresponds to residues 

1–10, spot 2 to 2–11, spot 3 to 3–12, etc., until the entire 

sequence of the 8K protein is covered. The arrow for each 

shows the last spot at the end of the sequence. There

Table 1 Amount of purified antibody per ml of applied anti-sera obtained 

from the standard or ETRAP protocol a 

Rabbit Sheep 

Standard ETRAP Standard ETRAP 

Target protein 

8K 0.16 0.23 0.04 0.09 

21K 0.4 0.08 b 0.16 1.4 

33K 0.36 0.37 0.33 0.18 c 

a Values represent the total mg of antibody obtained after buffer exchange and 

concentration normalized per ml of applied anti-sera. 

b The peak fractions from runs 1 through 3 in Figure 9(A) were combined 

before buffer exchange and concentration. 

c The peak fractions from runs 1 and 2 in Figure 13(B) were combined before 

buffer exchange and concentration. 

are four major intensely stained immunogenic regions for 

the rabbit map (Figure 3A) spanning spots 5–21, 24 –28, 

41–45 and 55–58. The wide spread of the reactive regions 

is probably a consequence of the nature of the polyclonal 

response. Many of the same immunoreactive regions were 

found for the sheep epitope map (Figure 3B), but there 

were differences. The 55–58 region was nested in a larger 

epitope of 52–60 and the extreme C-terminal 10-mer spot 

(arrow) was intensely stained. It was not possible to directly 

compare the intensities for the epitope map of the two 

animals since different secondary antibodies were used. We 

note that spot-peptide array epitope mapping within species, 

Figure 3 Spot-peptide membrane array epitope-mapping for standard protocol-purified anti-(GST–8K) antibodies 


for example, two rabbits or two sheep immunized with the 

same immunogen, will generate non-identical maps in both 

intensity and coverage for identical animals probably due to 

biological variation (results not shown). 

Our previous experience with the affinity purification 

of anti-peptide antibodies indicated that simple epitope 

peptide affinity resins can provide efficient trapping and 

purification of the target antibodies in a one-step process. 

A variant of this strategy termed ETRAP is illustrated in 

Figure 4. Thus knowledge of the epitope(s) is a prerequisite 

for ETRAP of anti-(GST–antigen) antibodies. The twofaceted 

strategy of preparing anti-peptide and anti-(GST– 

antigen) antibodies for a given target protein proved highly 

useful in the 8K antigen case. The two-rabbit anti-peptide 

antibodies corresponded to synthetic peptide immunogens 

represented by spot regions 6–16 and 49–58 (results not 

shown) which encompass the intensely immunostained 

regions 5–21 and 41–45/55–58 for anti-(GST–8K) antibodies 

(Figure 3A). We serendipitously had prepared these epitope 

affinity columns for anti-peptide antibody purification and 

therefore decided to use these for ETRAP of the 8K 

target protein antibodies. [We will not be so fortunate 

for the other two examples and will indeed need to design 

the epitope(s) peptide and conjugate them to the affinity 

columns.] Figure 5 displays the acid elution profile following 

ETRAP (peptide from spots 49 to 58) for both the rabbit and 

sheep anti-sera. ETRAP protein yields for these antibodies 

(Table 1) were approx. 0.23 and 0.09 mg respectively. 

This should be compared with the standard protocol on 

Each spot comprises a 10-mer peptide, spots were offset by 1 residue, and there were 20 spots per row. Arrow indicates end of the 8K sequence. (A) Rabbit 

antibodies and (B) sheep antibodies. Several of the immunostained regions overlap for both animals. 



Figure 4 ETRAP scheme to affinity-purify GST–antigen polyclonal antibodies 

in one step 

See the Materials and methods section for column preparation. Abs, antibodies. 

purification giving 0.16 mg for rabbit and 0.04 mg for sheep. 

Overall, the ETRAP process exacted approx. 8 h once the 

column was constructed. 

Figure 5 ETRAP elution profile of anti-(GST–8K) with an epitope peptide affinity column 

A critical issue when comparing the standard protocol 

and the ETRAP process, apart from the yield, was the 

purity of the final preparation. This was assessed by anti- 

GST antibody activity and percentage of ETRAP target 

antibody titre compared with the standard protocol. We 

can answer each in the affirmative for both rabbits 

(Figure 6A) and sheep (Figure 6B). The removal of anti-GST 

antibodies was either identical (Figure 6A) or slightly better 

(Figure 6B), and the 8K target protein antibody activity 

from the ETRAP process was 83 % and 87 % respectively. 

This is quite reasonable considering that we selected a 

specific subset of linear epitope sequences from the entire 

mapped repertoire. Importantly, we obtained final antibody 

preparations having similar purification characteristics for 

the 8K target antibodies in ∼8 hforETRAPcomparedwith 

several days for the standard procedure. A sandwich ELISA 

format using ETRAP (P-4974) rabbit GST–8K as capture 

antibody and biotinylated rabbit anti-peptide 8K (P-4973) 

as capping antibody was superior to all combinations of 

rabbit GST–8K obtained by the standard purification and 

rabbit anti-peptide P-4973 and P-4974 antibodies (results 

not shown). 


Figure 7 is a composite of the elution profiles from a 

glutathione-GST and a glutathione-(GST–21K) column for 

(A) Rabbit antibodies and (B) sheep antibodies. Fractions of approx. 1 ml were collected for all. Acid-eluted proteins from (A) and(B) were buffer exchanged and 

concentrated. 


Figure 6 Comparison of the purified preparation of anti-(GST–8K) antibodies from the standard and ETRAP processes 

(A) Rabbitand(B) sheep. The concentration of each antibody (Ab) was 1 μg/ml. S/N, signal/noise. 

rabbit (Figures 7A and 7B) and sheep (Figures 7C and 

7D) anti-sera respectively. The anti-sera from each animal 

required four regeneration (Figures 7A and 7C) cycles 

prior to affinity purification with an anti-(GST–21K) resin 



(Figures 7B and 7D). The yield was approx. 0.4 mg for the 

rabbit and 0.16 for the sheep (Table 1). The spot-peptide 

membrane array immunostaining is shown in Figure 8 with 

each row containing 24 spots. Note that all the antibodies 



with a glutathione-(GST–21K) column. Fractions of approx. 1 ml were collected for all. Four cycles of anti-GST antibody removal were necessary for both animals 

(A and C). Acid-eluted proteins from (B) and(D) were buffer exchanged and concentrated. 



Figure 8 Spot-peptide membrane array epitope-mapping for standard protocol purified anti-(GST–21K) antibodies 

Each spot comprises a 10-mer peptide, spots were offset by 1 residue, and there were 24 spots per row. The arrow indicates the end of the 21K sequence. 

(A) Rabbit antibodies and (B) sheep antibodies. Several of the immunostained regions overlap for both animals. 

are directed linear epitopes in the N-terminal region 

of the 21K target protein. Like the 8K target antibodies, 

there were more regions with similar intensity than those 

that were different. In this case, we chose a synthetic peptide 

that comprised residues 1–26 (spots 1–17) of the 21K 

protein as we expected the majority of the anti-(GST– 

21K) antibodies from each animal to bind to that peptide– 

SulfoLink resin (compare row 1 of Figures 8A and 8B). 

ETRAP for the rabbit anti-(GST–21K) sample is shown 

in Figure 9(A). Three consecutive runs were made and the 

column was regenerated between runs. These data indicate 

that the ETRAP procedure is qualitatively reproducible, 

not unlike multiple runs in the standard purification mode 

(results not shown). The sheep anti-sera results are found 

in Figure 9(B). The additional two initial fractions were 

due to collection of the last two re-equilibration washing 

steps prior to sample application. Both the standard and 

ETRAP methods were equally effective in removing anti-GST 

antibodies for rabbit (Figure 10A) and sheep (Figure 10B). 

There was an apparent improvement of approx. 29 % in the 

titre for the rabbit ETRAP sample compared to the standard 

protocol. Interestingly, the second rabbit of the immunized 

pair also showed an increased titre of 50 % (results not 

shown). For the sheep antibodies, 86 % of activity was 


observed in relation to the more laborious two-column 

process. ETRAP yields (Table 1) for this antigen from the 

rabbit were 0.08 mg and 1.4 mg for the sheep. 


The last comparative example describes the antibodies to 

a 33K target protein. Figure 11 displays the results of 

the elution profiles from the standard purification process 

of both rabbit (Figure 11A) and sheep (Figure 11B) antisera. 

The overall yield is approx. 0.36 mg and 0.33 mg 

respectively (Table 1). The epitope map for this 33K protein 

is given in Figure 12. The spots are offset by two residues 

instead of one due to the longer sequence of this target 

protein. Clearly, there are four non-overlapping heavily 

stained regions for the standard protocol-purified rabbit 

anti-sera (Figure 12A) as region 1 = spot 1; region 2 = 

spots 5–9; region 3 = spots 40–43; and region 4 = spots 

79–81. One could make a case for up to ten potential 

immunogenic sequences for the sheep anti-sera (Figure 12B). 

Some of these sequences overlapped with those epitopes 

of the rabbit antibodies. A 60-mer 3×epitope peptide 

was constructed, with inter-epitope Gly-Gly-Ser-Gly-Gly 

spacers, additional Gly-Gly-Gly termini, and an N-terminal 

cysteine residue as described in the Materials and methods



(A) Rabbit antibodies showing three consecutive runs following a column regeneration step and (B) sheep antibodies. Two extra wash fractions were collected 

before the sample was applied. Fractions of approx. 1 ml were collected for all. Acid-eluted proteins from (A) (peak fractions from three consecutive runs were 

combined) and (B) were buffer exchanged and concentrated. 

section, to include regions 2, 3 and 4 of Figure 12(A). 

Table 1 lists these ETRAP antibody yields at 0.37 mg for the 

rabbit (Figure 13A) and 0.18 mg for the sheep (Figure 13B). 

Unexpected results were found in the rabbit anti-(GST– 

33K) antibody activity as depicted in Figure 14(A). Only 

approx. 33 % of the antibody activity was retained using the 

3×epitope affinity column. The significantly stained spot 1 

(Figure 12A) was purposely omitted from the constructed 


(A) Rabbitand(B) sheep. Note the apparent increase of antibody (Ab) activity for the rabbit anti-(GST–21K) antibodies for ETRAP versus the standard protocol. 

See the Results and discussion section. The concentration of each antibody was 1 μg/ml. S/N, signal/noise. 






with a glutathione-(GST–33K) column. Fractions of approx. 1 ml were collected for all. Two cycles of anti-GST antibody removal were necessary for both animals 

(A and C). Acid-eluted proteins from (B) and(D) were buffer exchanged and concentrated. 

Figure 12 Spot-peptide membrane array epitope-mapping for standard protocol purified anti-(GST–33K) antibodies 

Each spot comprises a 10-mer peptide, spots were offset by 2 residues, and there were 24 spots per row. The arrow indicates the end of the 33K sequence. 

(A) Rabbit antibodies and (B) sheep antibodies. Several of the immunostained regions overlap for both animals but note also the unique regions in the sheep 

epitope map. 




(A) Rabbit antibodies and (B) sheep antibodies. Fractions of approx. 1 ml were collected for all. Acid-eluted proteins from (A) and(B) (peak fractions from two 

consecutive runs were combined) were buffer exchanged and concentrated. 

epitope peptide. Our past experiences make us suspicious 

of a single spot at the N-terminus (unpublished work). In 

this case, we posit that the epitope was directed towards 

the linker or joining peptide that connects the fusion 

protein GST and the target protein. Such occurrences 

have been reported previously [13]. Possible verification 

of this comes from: (i) a lone N-terminal spot or lack of 

staining of the membrane; (ii) a positive Western blot for 


(A) Rabbitand(B) sheep. Note the apparent loss of antibody (Ab) activity for the rabbit anti-(GST–33K) antibodies for ETRAP versus the standard protocol. See 

the Results and discussion section. The concentration of each antibody was 1 μg/ml. S/N, signal/noise. 



GST–antigen, but a negative Western blot for the cleaved 

(e.g. following thrombin proteolysis) antigen; or (iii) a 

positive direct binding ELISA for GST–antigen with inhibition 

by joining region peptide and negative for direct binding 

using the cleaved antigen. For the corresponding sheep antisera, 

the epitope map (Figure 12B) shows approx. 10 linear 

regions of various immunoreactivities. The first two Nterminal 

spots are moderately reactive suggesting that this 

sequence is not as highly immunogenic compared with that 

spot for the rabbit anti-sera (Figure 12A). Moreover, there 

was little loss of activity for the sheep anti-sera following 

the ETRAP procedure as 87 % of activity was observed 

(Figure 12B). 

Conclusion 

Polyclonal antibodies to amino acid immunogens are 

versatile reagents in many areas of study. Anti-protein 

antibodies are particularly useful in ELISAs as these 

antibodies tend to be of high affinity compared to 

anti-peptide antibodies. GST–recombinant proteins are 

frequently the immunogens in this case. Furthermore, the 

GST–protein is an ideal standard in ELISAs particularly 

when the native target protein is limited. This necessitates 

the removal of anti-GST antibodies which could otherwise 

seriously compromise the immunoassay. We have found 

that for the three examples presented here, several cycles 

of anti-GST removal must occur before the final affinity 

column purification step. Depending on the severity of the 

anti-GST response this can take up to 5 days. A one-step 

epitope affinity column process termed ETRAP successfully 

generated pure target protein antibodies from anti-(GST– 

protein) sera in all three cases. The remaining activity of 

anti-GST antibodies was negligible for either protocol, a 

mandatory prerequisite for using the GST–protein as an 

ELISA standard. For all six antibodies, the combined median 

antibody titre using ETRAP was 84.5 % of the antibody 

activity obtained by the standard procedure. We were 

unable to discern any salient trends regarding yields from 

the standard or ETRAP procedures for the results given 

in Table 1. The total amount of antibody obtained as the 

final product in these six examples was increased in three 

cases (rabbit 8K, sheep 8K and 21K), decreased in two 

cases (rabbit 21K and sheep 33K) and equal (rabbit 33K) 

for one comparison. Importantly, high-quality antibodies 

were produced using the ETRAP strategy compared to the 

lengthier standard two-column procedure. It is anticipated 


that polyclonal antibodies generated against other fusion tag 

recombinant proteins will be amenable to a similar ETRAP 

process. 

Funding 

This work was supported by funds from the Department of 

Pathology and Immunology, Washington University School 

of Medicine and Siemens Healthcare Diagnostics. 

References 

1 Harlow, E. and Lane, D. (1988) Antibodies: A Laboratory 

Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, 

NY 

2 Thermo Scientific Pierce (2009) Antibody Handbook, 

Rockford, IL 

3 Van Regenmortel, M. H. V., Briand, J. P., Muller, S. and 

Plaué, S. (1998) Synthetic Polypeptides as Antigen, Elsevier, 

Amsterdam 

4 Catty, D. and Raykundalia, C. (1989) In Antibodies (Catty, D., 

ed.), Volume 1, pp. 19–80, IRL Press, Oxford 

5 Crowther, J. R. (2001) In The ELISA Guidebook (Crowther, 

J. R., ed), pp. 9–44, Humana Press, Totowa, NJ 

6 Ekins, R. P. (1997) In Principles and Practice of Immunoassay 

(Price, C. P. and Neuman, D. L., eds.), pp. 173–208, Macmillian 

Reference Ltd, London 

7 Smith, D. B. and Johnson, K. S. (1988) Gene 67, 31–40 

8 Ladenson, J. H., Vaidya, H. C., Dietzler, D. N. and Maynard, A. 

Y. (1990) Creatine kinase MB determination method. U. S. Pat. 

4,912,033 

9 Bodor, G. S., Porter, S., Landt, Y. and Ladenson, J. H. (1992) 

Clin. Chem. 38, 2203–2214 

10 Atassi, M. (1984) Eur. J. Biochem. 145, 1–20 

11 Bar-Peled, M. and Raikhel, N. V. (1996) Anal. Biochem. 241, 

140–142 

12 Laterza, O. F., Modur, V. R., Crimmins, D. L., Olander, J. V., 

Landt, Y., Lee, J.-M. and Ladenson, J. H. (2006) Clin. Chem. 52, 

1713–1721 

13 Agaton, C., Falk, R., Guthenberg, I. H., Göstring, L., Uhlén, M. 

and Hober, S. (2004) J. Chromatogr. A 1043, 33–40 

Received 3 September 2010/1 November 2010; accepted 5 November 2010 

Published as Immediate Publication 5 November 2010, doi:10.1042/BA20100279

Full text PDF - Biotechnology and Applied Biochemistry

Create successful ePaper yourself

Delete template?

Save as template?