ANALYTICAL

BIOCHEMISTRY

lmmunoaffinity Kim

C. Williamson,*

206,359-362

(19%)

Chromatography Patrick

E. Duffy,?

Using Electroelution’

and David

C. Kaslow*

*Laboratory of Malaria Research, National Institute of Allergy and Infectious Maryland 20892; and TDepartment of Medicine, Walter Reed Army Institute

Received

May

Diseases, National Institutes of Health, of Research, Washington, D.C. 20307

Bethesda,

4,1992

Immunoaffinity chromatography and electroelution of protein from solid-phase matrices are two powerful tools often used to purify proteins. In this study, we combined these two techniques and found that antigen was effectively recovered from immunoaffinity resins by electroelution. Yields ranged from 90.5% to 62.8%, with a mean of 74.0 +- 7.4% (mean + standard deviation). A major portion of the eluate, 79.4 f 13.1%, was concentrated in a volume of 100 ~1 and 94.0 f 2.0% was recovered in 200 ~1, even when 1 ml of resin was used. Electroelution had no major effect on the electrophoretie mobility of the antigen. Two distinct antigens, a relatively hydrophilic 230-kDa protein and a hydrophobic 28-kDa protein were successfully electroeluted. In addition two types of immunoaffinity resins, monoclonal antibody covalently linked to CNBr-activated Sepharose or immobilized protein A, were found to be o 1992 Academic P~~TSS, I~C. compatible with this method.

Immunoaffinity chromatography is a powerful tool in protein purification, particularly when starting material is limited as in the case of Plasmodium falciparum and Plasmodium gallinaceum (1). One of the critical steps in immunoaffinity purification is the elution of the protein from the resin, especially when using high affinity monoclonal antibodies (mAbs)2 (2). Because each antigen an-

’ This investigation received financial support from the UNDP/ WORLD BANK/WHO Special Programme for Research and Training in Tropical Diseases (TDR). ’ Abbreviations: 2ME, 2-mercaptoethanol; SDS, sodium dodecyl sulfate; EDTA, ethylenediaminetetraacetic acid; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; Tris, Tris(hydroxymethyl)aminomethane; mAbs, monoclonal antibodies; Pfs230,230-kDa surface protein from Plasmodium fakiparum gametocytes; Pgs28,2&kDa surface protein from Plasmodium gallinaceum ookinetes; mAb lB3, monoclonal antibody P5E2-2F7-lB3; mAb B3B3, monoclonal antibody IID2-B3B3; PDVF, polyvinylidene difluoride; E350, upper 350 ~1 of eluate; E50, remaining 50 ~1 of eluate; RT, room temperature. 0003-2697192 $5.00 Copyright 0 1992 by Academic Press. All rights of reproduction in any form

tibody interaction is unique, an optimal strategy must be determined empirically for each purification. In designing an elution protocol the factors that need to be considered are yield, purity, and the residual effect elution has on the antigen and the mAb. Current elution techniques include extremes of pH (glycine (pH 3-2), diethylamine (pH ll)), urea, chaotropic agents, and denaturants (2-mercaptoethanol(2ME) and sodium dodecyl sulfate (SDS)) (2). None of these elution techniques were found to be satisfactory for the purification of a 230-kDa P. falciparum gamete surface protein (Pfs230) (3,4). An alternative is electroelution of antigen from affinity resins. This technique was originally described in 1977 (5). Early methods electroeluted ligand from the resin through a polyacrylamide gel therefore required long elution times and recovered very dilute antigen; or the elution apparatus had to be designed and constructed by the investigator (6-8). Consequently this technique has not been routinely used. Described here is an effective electroelution method that allows the recovery of concentrated antigen using a commercially available electroelution apparatus. Using this technique, Pfs230, a relatively hydrophilic protein and Pgs28, a hydrophobic surface protein on P. gallinaceum (9), were efficiently eluted and concentrated from mAb covalently bound to CNBr-activated Sepharose or crosslinked to immobilized protein A. MATERIALS

AND

METHODS

P. falciparum P. falciparum gametocytes (clone 3D7) were cultured in vitro (10). Mature gametocytes were harvested, trig-

gered to develop into gametes/zygotes, and then purified and surface labeled with 1251(3,4). The 1251-labeled gametes/zygotes were extracted on ice for 15 min in buffer A (0.15 M NaC1/5 mM EDTA/50 mM Tris (pH 7.4)/0.5% Triton X-100) with protease inhibitors (0.2 mM phenylmethylsulfonyl fluoride, 1 pg/ml pepstatin, 2 359

Inc. reserved.

360

WILLIAMSON,

DUFFY,

pg/ml aprotinin, and 2 pg/ml leupeptin (BoehringerMannheim)). The soluble extract was recovered after centrifugation at 14,000g for 20 min.

P. gallinaceum P. gallinaceum ookinetes were prepared as described Kaushal et al. (11). Ookinetes (1 X 10’) were extracted l-ml buffer A as described above.

Immunoafinity

by in

Resins

lB3-Sepharose. MAb P5E2-2F7-lB3 (lB3) (3) was purified from 10 ml of ascitic fluid on a protein A-Sepharose (Pharmacia) column (1.5 ml) using low salt elution (2). Antibody eluted in 100 mM glycine, pH 3.0, was neutralized with 1 M Tris to pH 8.0 then desalted on a PD10 column (Pharmacia) equilibrated with 0.5 M NaCl/ 0.1 M NaHCO, (pH 8.3). Seven milligrams of protein A-purified mAb lB3 (2.0 mg/ml) was added to acid washed CNBr-activated Sepharose 4B (1.5 ml) and mixed for 2 h at RT. The resin was pelleted by centrifugation, 1OOOg for 10 min, and the supernatant removed. To block unreacted groups, 0.1 M Tris (pH 8.0) was added to the resin and mixed for 2 h at RT. The mAb lB3-coupled resin was then washed with three cycles of alternating low and high pH: 0.1 M acetate (pH 4.0)/0.5 M NaCl followed by 0.1 M Tris (pH 8.0)/0.5 M NaCl. The washed mAb lB3-coupled resin was resuspended in buffer A with 0.5 M NaCl and 0.02% NaN, and stored at 4°C. MAb lB3-protein A and mAb IID2-B3B3 (B3B3)protein A. Monoclonal antibody lB3 or B3B3 (12) was coupled to immobilized protein A (Pierce) according to the manufacturer’s instructions. The column was stored in Binding Buffer (Pierce) at 4°C in 0.02% NaN,.

Immunoprecipitation Buffer A extract of P. falciparum gamete/zygotes was incubated with mAb lB3-protein A in a 2:l ratio of extract to resin overnight at 4OC. The same procedure was used with lB3-Sepharose, except that the NaCl concentration of the extract was increased to 0.5 M. The resins were pelleted by centrifugation at 1OOOg for 1 min, the supernatants removed, and the resins washed five times with five resin volumes of buffer A or when using lB3-Sepharose buffer A with 0.5 M NaCl. P. gallinaceum ookinete buffer A extract (2 ml) was incubated with 300 ~1 mAb B3B3-protein A for 1 h at 4°C. The resin was pelleted by centrifugation as above and washed three times, first with 10 resin volumes of buffer A, then 10 resin volumes of buffer A with 0.5 M NaCl, and finally 10 resin volumes of buffer A.

AND

KASLOW

Electroelution Bio-Rad electroeluter Model 422 was assembled according to the manufacturer’s instructions using 50 mM NH,HCO, (pH 7.8)/0.1% SDS as the elution buffer. Washed antigen-bound resin (50 ~1-1 ml) was resuspended in an equal volume of elution buffer and placed in the electroeluter tube. A current of 10 mA per tube was applied for 3-4 h. After elution, the upper reservoir buffer and resin were removed. The electroeluter tube was taken out of the apparatus and the membrane cap/ silicon gasket complex containing the eluted material was gently removed from the glass tube. The eluted material was collected in two aliquots. First the upper 350 ~1 (E350) of the eluate was drawn off and saved. The remaining 50 ~1 (E50) was pipetted up and down to resuspend the antigen then placed in a separate tube. The membrane cap was then washed five times with 50 ~1 (Wl-W5) of fresh elution buffer. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE) and Autoradiography Samples were concentrated to 50 ~1 in a Centricon 10 tube (Amicon) by centrifugation at 5000g for 1 h. The concentrated material was resuspended in 12.5 ~1 of 5~ Laemmli sample buffer (1 M Tris (pH 6.8)/16% SDS/ 20% glycero1/0.005% bromophenol blue) and size fractionated on a 4-20% polyacrylamide gel (Integrated Separation Systems) by the method of Laemmli (13). The gel was stained in 0.1% Coomassie blue, destained, dried, and autoradiographed. RESULTS Figure 1 shows the elution profile of radiolabeled Pfs230. In a series of 10 electroelutions using mAb lB3 linked to CNBr-activated Sepharose (100-500 ~1 of resin) an average of 74.0 + 7.4% (range 62.8 to 90.5%) of the bound radiolabel was eluted. The fractions E50 and Wl contained 79.4 f 13.1% of the total amount of radioactivity eluted and 94.0 f 2.0% was contained in E50W3 with a volume of 200 ~1. Similar results were obtained even when using 1 ml of resin (data not shown). (Larger volumes were not tested.) In contrast, the standard elution techniques usually required at least a column volume to remove most of the elutable antigen. When yield is critical all fractions, E350-W4 (600 pl), can be pooled and further concentrated to 50 ~1 or less in a Centricon 10 (or similar device). In addition to concentrating, the Centricon can be used to exchange buffers if necessary. Another advantage of electroelution is that the mAb lB3-Sepharose could be recovered from the elution tube and reused. Preelectroeluted resin bound 75.6 -I- 5.6% as much antigen as did fresh resin. This feature is important if antibody supply is limited; however, each mAb-

IMMUNOAFFINITY

CHROMATOGRAPHY

r

E350

E50

w1

w2

1 w3

w4

FIG. 1. Elution profile of Pfs230. Electroeluted Pfs230 was collected as described under Materials and Methods; (E350) upper 350 /rl of eluted material; (E50) remaining 50-~1 of eluted material; (Wl) first 50 ~1 wash; (WL) second 50-~1 wash; (W3) third 50-~1 wash; and (W4) fourth 50-p] wash. The radioactivity of each aliquot was measured directly in a gamma counter and expressed as the percentage of the total eluted material (the sum of all six aliquots). The results are presented as the mean of five experiments; the bar indicates the standard deviation.

resin should be tested. MAb B3B3-protein A did not bind antigen well after electroelution. MAb lB3-protein A was not tested. To directly compare electroelution to other techniques, radiolabeled antigen was bound to 500 ~1 of lB3 covalently crosslinked to immobilized protein A resin. The resin was washed and then aliquoted to separate tubes for elution. Equal distribution of radiolabel was monitored in a gamma counter. One aliquot (50 ~1 resin) was electroeluted for 3 h and collected as described above. The other aliquots were incubated for 30 min in 50 ~1 of buffer A, electroelution buffer (50 mM NH,HCO, (pH 7.8)/0.1% SDS), 100 mM glycine (pH 3), 50 mM diethylamine (pH ll), 8 M urea, 5% 2ME/3% SDS/buffer A, or 5% BME/buffer A. Following incubation the resin was pelleted by centrifugation and the buffer aspirated. The resin was then washed five times with 50 ~1 of fresh buffer. Elution of radiolabeled Pfs230 was monitored by measuring the radioactivity of each sample in a gamma counter. The elution yield was greater for electroelution (74.8%) than that for glycine (54.5%) and urea (68.6%); diethylamine (88.8%) was slightly more effective and 2ME/SDS (100.7%) was very effective (Fig. 2). Examination of eluted product by SDS-PAGE revealed that diethylamine, urea, and BME/SDS, although effective in eluting, had major disadvantages (Fig. 3). Urea caused the protein to aggre-

USING

361

ELECTROELUTION

gate and be retained at the origin of the gel; material eluted by diethylamine migrated as a diffuse band (Fig. 3). 2ME/SDS eluted a significant amount of antibody, thus contaminating the “purified” antigen and preventing the resin from being reused. Figure 2 also demonstrates that continued washing of antigen-bound mAb lB3-resin with electroelution buffer alone gradually elutes a significant portion of the antigen. Only 14.3 + 5.2% of the antigen was released with-each wash. The six column volumes of electroelution buffer required for a 85.6% recovery would be impractical when using 400 ~1 or more immunoaffinity resin. In addition to using the large, relatively hydrophilic Pfs230, electroelution was also tested using a small hydrophobic antigen, Pgs28 (9), and immunoaffinity resin made from mAb IID2-B3B3 (12). Using the same conditions employed for Pfs230/1B3-resin Pgs28 was efficiently eluted from a resin of monoclonal IID2-B3B3 covalently crosslinked to immobilized protein A (data not shown). This indicates that this method of electroelution is effective not only for the 230-kDa antigen/ lB3 monoclonal complex but may be generally applica-

E

w,

WT.

w3

w4

w5

FIG. 2. Comparison of elution techniques. Radiolabeled Pfs230 was eluted from 50 ~1 of immunoaffinity resin as described under Results using (X) 5% 2MEkmffer A, (W) electroelution buffer, (A) 50 mM diethylamine (pH ll), (Cl) electroelution, (0) 100 mM glycine (pH 3), (0) 8 M urea, and (A) 5% 2ME/3% SDS/buffer A. The x axis indicates the elution fraction, (E) the initial 50-~1 eluate (or in the case of electroelution pooled E350 and E50), and Wl, W2, W3, W4, and W5 represent the first through fifth washes. The cumulative percentage recovery at each step is plotted. To determine recovery the radioactivity of each sample was measured directly in a gamma counter and expressed as the percentage of the radioactivity bound to the immunoaffinity resin before elution 1% Recovery). The results presented are representative of two separate experiments.

362

WILLIAMSON.

DUFFY.

12345678910

-200

- 97.4 - 69 -46 -

30

-

21.5

FIG. 3.

Characterization of eluted antigen. Radiolabeled Pfs230 was eluted from 50 pl of immunoaffinity resin by the following techniques as described under Results; (lane 1) 5% BME/buffer A, (2) 5% 2ME/3% SDS/buffer A; (4) electroelution; (5) 100 mM glycine (pH 3); (7) 8 M urea; (8) 50 mM diethylamine (pH 11); (9) buffer A; (10) electroelution buffer. Lanes 3 and 6 contain molecular weight standards, myosin (200,000), phosphorylase b (97,400), bovine serum albumin (69,000), ovalbumin (46,000), carbonic anhydrase (30,000), and trypsin inhibitor (21,500) (Amersham); the M, x 10’ are indicated on the right. The eluants were pooled, concentrated, and run on a 4-20% polyacrylamide gel as described under Materials and Methods. The gel was analyzed by autoradiography.

AND

KASLOW

SDS). The use of 2ME/SDS also results in the contamination of eluted antigen with a large amount of antibody. This severely limits the use of the eluant, especially when the antigen comigrates on SDS-PAGE with the heavy or light chain of immunoglobulin. Additionally none of these nonelectrophoretic techniques concentrate the antigen, which is an important feature when using a large amount of immunoaffinity resin. A mechanism of electroelution was suggested by the gradual elution of antigen by electroelution buffer (50 mM NH,HCO, (pH 7.8)/0.1% SDS) alone. As the antigen is slowly dissociated from the resin it is carried by the electrical current to the impermeable membrane cap, where it concentrates. In the future it would be useful to develop membrane caps of microporous supports that immobilize protein such as polyvinylidene difluoride (PDVF). An immunoaffinity-purified protein could then be electroeluted directly on to PDVF, removed from the apparatus, and sequenced. ACKNOWLEDGMENT We thank

Roseanne

Hearn

for expert

technical

assistance.

REFERENCES

ble to the immunoaffinity purification of antigens with a range of molecular size and hydrophobicity.

1. Kaslow, Coligan,

D. C., Quakyi, I. A., Syin, C., Raum, J. E., McCutchan, T. F., and Miller,

333,74-76. 2. Harlow,

DISCUSSION

Electroelution is an effective technique for removing and concentrating antigen from immunoaffinity resin. This efficient method of elution was critical in purifying both a relatively hydrophilic 230-kDa protein from P. falciparum gametes and a hydrophobic 28kDa protein from P. gallinaceum. The yield of antigen ranges between 62.8 and 90.5%, with a mean of 74.0 & 7.4%. Electroelution did not permanently alter the antigens examined in this study or mAb lB3. The antigens migrated normally on SDS-PAGE and the lB3-Sepharose resin could be reused. The eluted material could be further concentrated to 50 ~1 using a Centricon 30 or 10. Standard eluants were either not effective (glycine (pH 3)) or significantly altered the antigen or the immunoaffinity resin (diethylamine (pH ll), urea, 2ME/

M. G., Keister, D. B., L. H. (1988) Nature

E., and Lane, D. (1988) Antibodies: ual, p. 511, Cold Spring Harbor Laboratory, NY.

3. Quakyi,

Miller, 4. Kumar,

I. A., Carter, R., Rener, L. H. (1987) J. Zmmunol. N. (1987)

Parasite

A Laboratory Cold Spring

J., Kumar, N., Good, 139,4213-4217.

Zmmunol.

ManHarbor, M. F., and

9, 321-335.

5. Dean,

P. D. G., Brown, P., Leyland, M. J., Watson, S., and Harvey, M. J. (1977) Biochem. Sot. Trans.

D. H., Angal, 6.1111-1113.

6. Morgan,

M. R. A., George, E., and Dean, P. D. G. (1980) Anal. Biochem. 105,1-5. Z., and Kasicka, V. (1985) J. Chromutogr. 320,81-88. 7. Prusik, a. Haff, 9. Kumar, 10. Ifediba,

L. A. (1981)

Electrophoresis

2, 287-290.

N. (1985) Mol. Biochem. Purusitol. 17,343-358. T., and Vanderberg, J. P. (1981) Nature 294,364-366.

11. Kausbal, D. C., Carter, R., Rener, J., Grotendorst, C. A., Miller, L. H., and Howard, R. J. (1983) J. Zmmunol. 131,2557-2562. 12. Grotendorst, C. A., Kumar, N., Carter, R., and Kaushal, D. C. (1984) Infect. Zmmun. 46, 775-777. 13. Laemmli, U. K. (1970) Nature 227,680-685.