Construction and characterization of M13 bacteriophages displaying functional IgG-binding domains of Staphylococcal protein A

Gene, 151(1994)45-51 0 1994 Elsevier Science B.V. All rights reserved.

45

0378-1119/94/$07.00

GENE 08308

Construction and characterization of M 13 bacteriophages displaying functional IgG-binding domains of Staphylococcal protein A (Display vectors; fusion proteins; panning; phagemid; surface display)

Ashima Kushwaha, Partha Sarathi Chowdhury, Kajal Arora, Smita Abrol and Vijay K. Chaudhary Department of Biochemistry, University of Delhi South Campus, New Delhi 110021, India Received by G.N. Godson:

4 April 1994; Revised/Accepted:

8 June/24

June 1994; Received at publishers:

1 August

1994

SUMMARY

Staphylococcal protein A (SPA) is ranked as a versatile probe in immunoassays because of its immunoglobulin G (IgG)-binding capability. However, poor binding of SPA to the IgG of some laboratory animals and its inability to bind human IgG3 restricts its universal utility. In the present study, DNA encoding the four IgG-binding domains of SPA (E, D, A and B) or the B domain alone has been fused, in separate phagemid vectors, to the 5’ end of gene ZZZof the phage M13. Upon infection by helper phage M13K07, phagemid particles encapsulating single-stranded DNA were produced. Dot immunoblot and Western blot analyses showed the presence of fusion proteins on the Ml3 surface. Binding of rabbit IgG-horseradish peroxidase (IgG-HRP) complex to the phage particles confirmed that the fusion proteins possessed functional IgG-binding domains. The interaction of these phages with immobilised human IgG and its various subclasses was studied by the phage capture immunoassay where the captured phages were detected by a monoclonal antibody to the major coat protein encoded by gene I/III (gVZZZ).The phages showed maximal binding to IgGlk, followed by IgG& and showed negligible binding to the IgG3K and IgG3h subclasses. The specificity of IgG-binding phages was confirmed in a phage capture and elution assay where the binding of these phages to immobilised human IgGlK was abolished in the presence of excess of soluble protein A. Moreover, IgG-binding phages could be enriched approx. lOOO-fold over non-specific phages in a single round of panning. These IgG-binding phages will be useful in selecting novel protein A molecules with wider specificity and higher affinity.

INTRODUCTION

Staphylococcal protein A (SPA) is a cell-wallassociated protein of Staphylococcus aureus having an Correspondence to: Dr. V.K. Chaudhary, Department of Biochemistry, University of Delhi South Campus, Benito Juarez Road, New Delhi 110021, India. Tel. (91-11) 678-876 or 601-955; Fax (91-11) 688-5270 or 688-6427. Abbreviations: A, absorbance (1 cm); aa, amino acid(s); Ab, antibody (ies); ABTS, 2,2’-azino-bis(3’-ethylbenzthiazoline-6-sulfonic acid); AP, alkaline phosphatase; bla, gene encoding B-lactamase (Bla); BSA, bovine serum albumin; cfu, colony-forming unit(s); CH, constant domain of the heavy chain of an Ab; ELISA, enzyme-linked immunosorbent assay; Fc, crystallisable fragment of an Ab; gIIIp, gene-Ill-encoded minor coat protein of bacteriophage M13; gVIIIp, gene-Fill-encoded major coat protein of bacteriophage M13; HRP, horseradish peroxidase; IgG, SSDI 0378-1119(94)00519-2

extracellular portion that binds to the Fc portion of IgG. One of the major applications of SPA is in the field of immunodiagnostics where its conjugates with reporter enzymes such as alkaline phosphatase (AP) or horseimmunoglobulin

G;

IPTG,

laczpo, lac promoter-operator; deionised water containing

isopropyl-S-D-thiogalactopyranoside; LB, Luria-Bertani (medium); M9, sterile 5 x M9 salts/l mM MgS0,/0.0005%

thiamineHCl/O.l mM CaC1,/0.2% tion; ori, origin of DNA replication; phoresis; PBS, phosphate-buffered buffer/l50 mM NaCl pH 7.4); PEG, signal sequence; pfu, plaque-forming

glucose; moi, multiplicity of infecPAGE, polyacrylamide-gel electrosaline (20 mM sodium phosphate poly(ethylene glycol); pelBss, pelB unit(s); RBS, ribosome-binding

site(s); SDS, sodium dodecyl sulfate; SPA, staphylococcal protein A; spa, gene encoding SPA; ss, single strand(ed); TBS, Tris-buffered saline (50 mM Tris/lSO mM NaCl pH 7.4); XGal, 5-bromo-4-chloro-3indolyl-B-D-galactopyranoside; 2 x YT (per litre), 16 g Bactotryptone/lO

g yeast extract/5

g NaCl pH 7.2.

46

radish peroxidase (HRP) are used as substitutes for enzyme-linked secondary Ab. However, the inability of SPA to bind human IgG3 and its poor binding to IgG from some animal species may limit the use of these reagents in some immunoassays. The IgG-binding portion of SPA consists of five highly homologous domains, namely, E, D, A, B and C, of about 58 aa each (Uhlen et al., 1984) and all the domains can individually bind IgG (Moks et al., 1986). X-ray crystallography (Deisenhofer, 1981) and NMR studies (Gouda et al., 1992) of domain B of SPA complexed with the Fc fragment of human IgG have shown the presence of two a-helices in the B domain that make multiple contacts with CH2 and CH3 domains of Fc. Inability of SPA to bind human IgG3 has been attributed to the His435+Arg substitution in this subclass of immunoglobulins (Deisenhofer, 1981; Haake et al., 1982). Synthetic IgGbinding domains of SPA have been constructed and expressed in E. coli in various forms and their mutants have been analysed for interactions with IgG (Gore et al., 1992; Cedergren et al., 1993). Recently, the phage display system (Parmley and Smith, 1988) has emerged as a powerful and versatile technique for rapid selection of high-affinity variants of proteins which are expressed as fusions with phage coat proteins (Hoess, 1993; Lowman and Wells, 1993; Smith, 1993; Larrick et a1.,1993). We have described the high level expression and properties of a chimeric protein consisting of IgG-binding domains of SPA and a high specific activity variant of E. coli alkaline phosphatase (APK328A) as a substitute for secondary Ab (Chowdhury et al., 1994). In order to produce a more versatile Staphyloccocal protein-A-alkaline phosphatase fusion protein (SPA-AP), mutants of SPA with wider specificity and higher affinity can be created by employing phage display system. In this paper, we describe a phagemid system to display either the first four IgG-binding domains (E, D, A and B) or only the B domain of SPA fused to the N terminus of gene III protein (gIIIp) of filamentous phage M13. The displayed proteins have been shown to be fully functional and specific as studied by their interactions with the human IgG and its various subclasses.

RESULTS

AND DISCUSSION

(a) Construction of phagemid vectors The phagemid vectors constructed to display SPA or its B domain fused to the N terminus of gene III protein (gIIIp) on the Ml3 surface are shown in Fig. 1. In these vectors, the gene fusion is under the control of lacZpo, which allows tight regulation of the fusion protein synthesis due to overexpressed lac repressor from the host

XLl-blue. The gene fusion consists of DNA sequence encoding pelB signal sequence (pelBss) for transport of fusion proteins across the inner bacterial membrane (Lei et al., 1987), the first four IgG-binding domains of SPA (pVCSPA3426) or only the B domain (pVCB3426) followed by codons for a decapeptide tag (c-myc) and a flexible spacer (SGGGG) fused to the gene III (gill). Mutagenesis of spa gene is difficult owing to its repetitive sequence. Therefore, B domain display vector (pVCB3426) was constructed to enable mutagenesis in the a-helices that are involved in interaction with Fc portion of IgG. It also incorporates the codons for the first 6 aa of the E domain of SPA in between the pelBss and the B domain (Fig. 1C). The aa of the E domain have been shown to facilitate correct processing of the protein A signal sequence and efficient secretion of mature protein (Abrahmsen et al., 1985). (b) Fusion proteins are displayed on the phage particles E. coli XLl-blue cells carrying pVCSPA3426 or pVCB3426 were grown in 2 x YT medium with appropriate antibiotics and 1% glucose. The phage particles encapsulating phagemid ssDNA were produced by superinfecting these cultures with helper phage M13K07 at a multiplicity of infection (moi) of 10 and the expression of fusion proteins was induced by diluting the cultures in glucose-free medium. The phagemid yield in terms of colony forming units (cfu) was (3-4) x 10” per ml of culture while the plaque forming units (pfu) were (l-2) x lo9 per ml. The pfu arises due to the encapsulation of ssDNA from M13K07. The low titre of M13K07 is due to its poor replication in the presence of phagemid DNA. Inside the cell, both gIIIp from M13K07 and the fusion protein from the phagemid are produced which then compete for incorporation in the phage coat during the phage particle assembly. Therefore, the extruded phage particles should have both gIIIp and the fusion protein on their surface. The phage particles were purified to remove bacterial cells and soluble proteins. In dot blot analysis, SPA and B phages showed dose-dependent reactions with rabbit IgG coupled to HRP (RIgG-HRP) (Fig. 2, panel I, lanes B and C) similar to SPA-AP fusion protein (lane A). KS phages (from pBluescript KS +, Stratagene, La Jolla, CA, USA) and CD4 phages (Abrol et al., 1994) did not bind to RIgG-HRP, indicating the absence of IgG-binding fusion protein on their surface (lane D, 1 and 2). SPA, B and CD4 phages contain a decapeptide tag from c-myc (Evan et al., 1985) in between the foreign peptide and gIIIp. All of these phages reacted with 9E10, a mouse monoclonal Ab against this decapeptide tag (Fig. 2, panel II). But KS phages which did not carry any fusion protein were not reactive to anti-c-Myc Ab (lane Dl). The SPA-AP fusion protein showed weak

47

pVCSPA3426 b/a

L

gill

MAEASQBDPATSADN 1 ATGGCTGAGGCTAGCCAACACGATGAAGCTACTAGTGCGGATAAC

Fig. 1. Schematic representation of protein A display vectors. (A) pVCSPA3426 consists of a gene fusion cassette comprising the ribosome-binding site (RBS), DNA encoding the p&3 signal sequence (pe&s), domains E, D, A and B of protein A, a decapeptide tag (c-myc) and codons 3-406 of gene 111,inserted between Hind111 and EcoRI sites of pUC119. The gene fusion is under the control of 1acZpo. The vector also contains fl ori (F+), ColEl ori (ori) and b/a. (B) Gene fusion cassette in pVCSPA3426. (C) Gene fusion cassette in pVCB3426. Shown in box are the aa of domain E of SPA. Vertical arrow indicates the cleavage site of signal sequence (Chowdhury et al., 1994). Methods: pVCSPA3426 and pVCB3426 were constructed by inserting the DNA sequence encoding either the first four IgG-binding domains of SPA or that of B domain, respectively, as NheI-MluI fragment into pVCCD43426 in place of CD4. pVCCD43426 is a derivative of pUC119 that carries DNA encoding first two extracellular domains of human CD4 fused to gIlI of Ml3 phage under the control of 1acZpo (Abrol et al., 1994). The first four IgG-binding domains of spa were excised by NheI and MluI from pVNLSPA-AP1306 which is a derivative of pVNLSPA-APO216 (Chowdhury et al., 1994). B domain was amplified by PCR, using a HindIII-EcoRI fragment of pVNLSPA-AP1306 as a template along with a 5’ primer, 5’.AAGCTAAAGCTAGCCAACACGATGAAGCTACTAGTGCGGAAATTCAACAAGGAACAACAA, which incorporated an NheI site (underlined), first six codons of E domain and an SpeI site (underlined) at the 5’ end of B domain and a 3’ primer T-CATTTCGGGGGTACCGCCCCCGCCCTACGCGTTAC CCTTAGGTGCTTGAGC, which incorporated the MluI site (underlined) at the 3’ end of the B domain.

reaction with 9E10 and this was probably due to the ability of its protein A component to bind to mouse IgG and HRP-linked goat anti-mouse IgG used in the assay. This indicates that IgG-binding portions of SPA and B phages may also bind to these Ab. But this artifact is not significant, since mouse monoclonal Ab and goat Ab react very weakly to protein A. This is reflected by a large difference between the intensity of dots in lane A in comparison to that in lanes B and C. All the phage prepara-

tions showed a very strong reaction with polyclonal rabbit anti phage Ab (Fig. 2, panel III, lanes B and C). These results clearly demonstrated that purified preparations of SPA and B phages contained IgG-binding moieties. These results were further confirmed by Western blot analysis (Fig. 3). SPA and I3 phages showed bands corresponding to 85 and 65 kDa, respectively, which reacted with RIgG-HRP, anti-c-Myc Ab and rabbit anti-phage Ab. The calculated sizes of SPAgIIIp and B-gIIIp are 68

1

2

3

4

B C D

Fig. 2. Dot blot of phages displaying IgG-binding domains. Lanes: Al-4, SPA-AP fusion protein (3.75 ng, 7.5 ng, 15 ng and 30 ng, respectively); Bl-4, SPA phages, (3 x lOlo, 6 x lOi’, 1.2 x 10I1, 2 4 x 10” particles respectively); Cl-4, B phages, (3 x lOi’, 6 x lOi’, 1.2 x lo”, 2.4 x 10” particles, respectively); Dl, KS phages and D2, CD4 phages (2.4 x 10” particles each); D3, NET buffer (100 mM NaCl/l mM EDTA/lO mM Tris pH 8.0). Blots were probed with rabbit IgG-HRP conjugate (panel I), 9ElO (ascitis from hybridoma 9El0, ATCC CRL 1729) followed by goat anti-mouse IgG-HRP conjugate (Jackson ImmunoResearch Laboratories, West Baltimore Pike, PA, USA) (panel II) and rabbit anti phage Ab followed by goat anti-rabbit IgG HRP conjugate (panel III). Methods: To produce various phage particles, individual cultures of XLl-blue cells carrying pVCSPA3426, pVCB3426, pKS or pVCCD43426 in 2 x YT medium containing ampicillin ( 100 ug/ml), tetracycline (10 ug/ml) and glucose ( 1%) were grown to an A 600nnl of 0.3 at 37°C and infected with M13K07 (Vieira and Messing, 1987) at moi of 10. The cultures were grown for 1 h with slow shaking and then diluted lo-fold in glucose-free 2 x YT medium containing ampicillin (100 pg/ml), tetracycline (10 ug/ml) and kanamycin (70 ug/ml). The incubation at 37°C was continued for 16 h. The culture supernatant was collected by pelleting the cells at 12000 x g. Phage particles were collected by two rounds of PEG precipitation by adding PEG 8000 (Sigma, St. Louis, MO, USA) and NaCl to a final concentration of 2.2% and 0.4 M, respectively (Smith and Scott, 1993). Further purification was done by pelleting the phage particles by ultracentrifugation at 200000 x g for 2.5 h and phages were suspended in NET buffer (100 mM NaCl/l mM EDTA/lO mM Tris pH 8.0) to a final concentration of lOi cfu/ml. In case of SPA phages, the phages were further purified by CsCl density gradient by the method of Lin et al. (1980). 150 ul of each sample in NET buffer was applied onto nitrocellulose membranes using dot blot apparatus (SchleicherkSchuell, Keene, NH, USA). The nitrocellulose membranes were blocked with PBST containing 3% gelatin (PBS/O.OS% Tween-20). All the Ab dilutions were made in PBST with 1% gelatin and the incubations with Ab carried out at room temperature for 1 h with shaking. The washing between incubations was done with PBST containing 1% gelatin. Finally the blots were washed three times with PBS and developed with substrate solution (1 mg diaminobenzidine/l ul 30% H,OZ per ml PBS).

1 2 3 4 *mL :; = .“)=yty

kDa

43 -

-*”

*9-

&!

18.4 14 -

I

II

III

Fig. 3. Western blot analysis of phages displaying IgG-binding domains. Samples were boiled in 1 x sample buffer (Laemmli, 1970), analysed on 0.1% SDS-12.5% PAGE, transferred onto nitrocellulose membranes and probed with rabbit IgG-HRP conjugate (panel I) or 9E10, mouse anti-c-Myc Ab followed by goat anti-mouse IgG-HRP (panel II) or rabbit anti phage Ab followed by goat anti-rabbit IgG-HRP (panel III). Lanes: 1, 2 x 10” SPA phages; 2, 2 x 10” B phages; 3, 2 x 10” KS phages; 4, 5 x 10’ XLl-blue cells. a and b indicate the position of SPA-gIIIp and B-gIIIp fusion proteins, respectively. The position of high-molecular-mass markers (BRL, Gaithersburg, MD, USA) in kDa is also shown. Methods After electrophoresis and transfer onto nitrocellulose, the membranes (I, II and III) were processed as described in legend to Fig. 2.

and 50 kDa, respectively. The large apparent sizes of the fusion proteins in SDS-PAGE are due to the abberant mobility of gIIIp component which migrates as a 55-70-kDa protein instead of 44 kDa (Goldsmith and Konigsberg, 1977). KS phages which did not display any fusion protein showed reaction only with anti phage Ab with bands corresponding to gIIIp and gVIIIp (Fig. 3, panel III, lane 3). SPA and B phages, besides showing a band of gVIIIp, also showed a strong band corresponding to gIIIp (Fig. 3, panel III, lanes 1 and 2, respectively).

This is due to the fact that in this controlled expression and display system, only a fraction of gIIIp is present as fusion protein, the rest being native gIIIp contributed by M13K07. In Fig. 3, panel III, SPA phages showed very clear bands of the SPA-gIIIp fusion protein of 85 kDa, gIIIp of 60 kDa and gVIIIp of 6 kDa, when developed with rabbit anti phage Ab, whereas B and KS phages showed several other bands which were characteristic of host cells (lane 4). These extra bands can also be seen in blots developed with anti-c-Myc Ab. The presence of

49 extra bands is due to low purity of these phage preparations as compared to SPA phages. Nevertheless, the results clearly demonstrated that the fusion proteins of appropriate sizes and properties were displayed on phage particles. (c) Phages displaying IgG-binding domains can be selectively captured by human IgG

To further investigate IgG-binding property of SPA and B phages, these were captured on immobilized polyclonal human IgG or its various subclasses and bound phages were detected either by ELISA using culture supernatant of anti-gVIIIp hybridoma or by eluting the bound phages followed by a biological assay, i.e., their colony forming potential. In phage ELISA, both SPA and B phages bound to polyclonal human IgG in a dose-

dependent manner (data not shown). As shown in Fig. 4, both SPA and B phages showed strong binding to human polyclonal IgG, IgGlK and IgG2K subclasses and negligible binding to IgG3K and IgG3h. KS and CD4 phages showed negligible binding to all subclasses thereby indicating the absence of IgG-binding domains on their surface. These results were further substantiated in biological assays where SPA, B and KS phages were captured on human 1gGlK in the absence and presence of soluble SPA and the number of eluted phages was estimated by infecting XLl-blue cells. About 3-4% of SPA and B phages were captured under experimental conditions while only 0.01% of non-specific KS phages could bind (Fig. 5). This could be due to the fact that in this M13K07-assisted phagemid display system, only 10% of the phage population would display one molecule of the fusion protein per phage particle (Bass et al., 1990). Therefore, only a frac-

1SPA-Phage

4

B-Phage

2 B E g

2

n 1

”

B

SPA

KS

CD4

2

3 0

/

Phages Fig. 4. Phage capture phages

on microtitre

immunoassay. plate coated

ELISA showing with different

Methods: The wells of a 96-well microtitre NY, USA) were coated

1 capture

human

plate (Corning,

with 100 ~1 of 10 pg/ml

of various

IgG subclasses. New York,

of polyclonal

human

2

Fig. 5. Phage capture to human to human

3

and elution

IgGlK. Bar diagram IgGlK

in the presence

and 3) of soluble

protein

assay. Binding

showing

A (Sigma)

as competitor.

(bars

shaded

phages,

respectively.

h. The wells were washed six times with TBST (50 mM Tris pH 7.4/150 mM NaCl/O.l% Tween-20) followed by blocking with TBST

as compared to that of SPA and B phages. Methods: I pg human IgGlK in 100 pl of 50 mM NaHCO, buffer (pH 9.4) was coated in maxisorp ELISA strips (Nunc, Naperville, IL, USA) at 37°C for 2 h. After washing

added to the wells in duplicates

and incubated

at 37°C for 2 h followed

1

The open (1 and

2) and

Binding

represent

of phage binding

2 and 4) and absence

IgG (Jackson ImmunoResearch Laboratories) or various subclasses of human IgG (Sigma) in 50 mM bicarbonate buffer pH 9.4 at 37°C for 2

containing 3% gelatin at 37°C for 2 h. lo9 phage particles of each type of phage preparation in 100 pl of TBST containing 1% gelatin were

(3 and 4) bars

of SPA and B phages

the percentage

(bars

5

4

the binding

of SPA and B

of KS phages (bar 5) was less than 0.01%

the wells with TBST, blocking at 37°C for 2 h. To the blocked

was done in TBST containing wells, IO” phages

3% gelatin

or 10”’ phages+

100

by washing of the wells ten times with TBST. This was followed by addition of 100 pl of 100-fold diluted culture supernatant of anti-gVIIIp

ng protein A in 100 ~1 of TBST containing 1% gelatin were added in duplicates and incubated at 37°C for 2 h with constant shaking. The

hybridoma (D. Bhardwaj, S.S. Singh, S.A. and V.K.C., data not shown) in TBST containing 1% gelatin and the plate incubated at 37°C for 2

unbound phages were removed and the wells were washed ten times with TBST and five times with TBS. The bound phages were eluted by

h. Wells were washed with TBST and goat anti-mouse added and incubated at 37°C for 2 h. After washing TBST and finally by TBS, the binding was quantitated 100 ~1 of substrate solution (1 mg ABTS and 1 pl H,02

IgG-HRP was the wells with by addition of per ml citrate-

the addition of 100 ~1 of elution buffer (0.1 M HCl adjusted to pH 2.2 by addition of solid glycine and containing 1 mg/ml BSA) to each well followed by incubation at room temperature for 10 min with constant shaking. The phage eluates were rapidly neutralised by addition of 2

phosphate buffer (76 mM citric acid/21.8 mM Na2HP0, pH 4.5)). Reaction was allowed to proceed at 37°C for 10 min and then stopped by adding 100 (rl 10% SDS. Absorbance was recorded at 405 nm using an ELISA reader (Bio-Tek Instruments, Winooski, VT, USA).

M Tris base. The colony forming units in eluates were determined by infecting 100 ~1 log phase XLl-blue cells with serial dilutions of the eluates followed by incubation at 37°C for 1 h and plated on LB plates containing ampicillin (100 pg/ml) and glucose (1%).

50 tion of these fusion protein bearing phages would bind to the immobilised IgG. Moreover, the binding of SPA and B phages was highly specific as this could be inhibited by the presence of soluble SPA. The specificity of SPA phages was further confirmed in an enrichment experiment where SPA and KS phages in the ratio 1:2000 were incubated with immobilised human IgG. After elution, the phages were plated on M9 medium containing ampicillin, XGal and IPTG. The ratio of SPA and KS phages could be determined directly from the ratio of white and blue colonies. The results showed that due to specific interaction with IgG, SPA phages were enriched approx. lOOO-fold over KS phages in a single round of panning (data not shown). Recently, Djojonegoro et al. (1994) have also developed a phagemid system to display fragment B of SPA fused to gene III of phage Ml3 and have shown that these phages could be selectively enriched from a mixed population using IgG-Sepharose. (d) Conclusions and perspectives (1) The present study demonstrates that by a controlled expression of SPA- or B-gIIIp fusion proteins along with wild-type gIIIp, a large number of virions [(3-4) x lOlo cfu/ml] can be produced that display functional IgGbinding domains on their surface. Binding studies indicate that 3-4% phages bind to immobilised IgG. (2) The biopanning experiments peformed on immobilised human IgG and its different subclasses show that SPA and B phage particles can bind with several hundred fold higher specificity as compared to non-specific KS phage particles, hence confirming that binding is mediated by a highly specific interaction between human IgG and protein A borne on phages. (3) The combination of random mutagenesis in the first and second helices of B domain and their phage display is anticipated to yield a vast collection of mutants that can be rapidly screened by biopanning (Parmley and Smith, 1988) to select higher affinity variants of SPA.

ACKNOWLEDGEMENTS

The authors are grateful to Drs. J.K. Batra, Dinkar Sahal and P.K. Ghosh for careful reading of the manuscript, and Dr. Raj Raghupati for extending his help in preparing anti phage antisera. Authors are also thankful to Mr. Devesh Bhardwaj for providing anti-gVIIIp monoclonal antibody and Mr. Rajiv Chawla for help in preparing the manuscript. A.K and S.A. are recipients of Senior Research Fellowships from the Council of Scientific and Industrial Research, India. This work was supported by the Department of Biotechnology and the

Department India.

of Science and Technology, Government

of

REFERENCES AbrahmsCn, L., Moks, T., Nilsson, B., Hellman, U. and UhlBn, M.: Analysis of signals for secretion in the staphylococcal protein A gene. EMBO J. 4 (1985) 3901-3906 Abrol, S., Sampath, A., Arora, K. and Chaudhary, V.K.: Construction and characterisation of Ml3 bacteriophages displaying gp120 binding domains of human CD4. Ind. J. Biochem. Biophys. 31 ( 1994) 302-309. Bass, S., Greene, R. and Wells, J.A.: Hormone phage: an enrichment method for variant proteins with altered binding properties. Proteins 8 (1990) 309-314. Cedergren, L., Andersson, R., Jansson, B., Uhltn, M. and Nilsson, B.: Mutational analysis of the interaction between staphylococcal protein A and human IgG,. Protein Eng. 6 (1993) 441-448. Chowdhury, P.S., Kushwaha, A., Abrol, S. and Chaudhary, V.K.: An expression system for secretion and purification of a genetically engineered thermostable chimera of protein A and alkaline phosphatase. Protein Expr. Purif. 5 (1994) 89-95. Deisenhofer, J.: Crystallographic refinement and atomic models of a human Fc fragment and its complex with fragment B of protein A from Staphy/ococcus aweus at 2.9- and 2.8-A resolution. Biochemistry 20 (1981) 2361-2370. Djojonegoro, B.M., Benedik, M.J. and Willson, R.C.: Bacteriophage surface display of an immunoglobulin-binding domain of Staphylococcus aureus protein A. Bio/Technology 12 (1994) 169-172. Evan, G.I., Lewis, G.K., Ramsay, G. and Bishop, J.M.: Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product. Mol. Cell. Biol. 5 (1985) 3610-3616. Goldsmith, M.E. and Konigsberg, W.H.: Adsorption protein of the bacteriophage fd: isolation, molecular properties and location in the virus. Biochemistry 16 (1977) 2686-2694. Gore, M.G., Ferris, W.F., Popplewell, A.G., Scawen, M. and Atkinson, T.: pH sensitive interactions between IgG and a mutated IgG-binding protein based upon two B domains of protein A from Staphylococcus aureus. Protein Eng. 5 (1992) 577-582. Gouda, H., Torigoe, H., Saito, A., Sato, M., Arata, Y. and Shimada, I.: Three dimensional solution structure of the B domain of staphylococcal protein A: comparisons of the solution and crystal structures. Biochemistry 31 (1992) 9665-9672. Haake, D.A., Franklin, E.C. and Frangione, B.: The modification of human immunoglobulin binding to staphylococcal protein A using diethyl pyrocarbonate. J. Immunol. 129 (1982) 190-192. Hoess, R.H.: Phage display of peptides and protein domains. Curr. Opin. Struct. Biol. 3 (1993) 572-579. Laemmli, U.K.: Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227 (1970) 680-685. Larrick, J.W., Balint, R.F. and Houghten, R.A. (Eds.): Pharmaceutical Design Using Epitope Selection Technologies, Palo Alto. Gene 137 (1993) 1-149. Lei, S.-P., Lin, H.-C., Wang, S.-S., Callaway, J. and Wilcox, G.: Characterisation of the Erwinia carotovora pel B gene and its product pectate lyase. J. Bacterial. 169 (1987) 4379-4383. Lin, T.-C., Webster, R.E. and Konigsberg, W.: Isolation and characterisation of the C and D proteins coded by gene IX and gene VI in the filamentous bacteriophage fI and fd. J. Biol. Chem. 255 (1980) 10331-10337. Lowman, H.B. and Wells, J.A.: Affinity maturation of human growth hormone by monovalent phage display. J. Mol. Biol. 234 (1993) 564-578.

51 Moks, T., Abrahmstn, L., Nilsson, B., Hellman, U., Sjiiquist, J. and Uhlen, M.: Staphylococcal protein A consists of five IgG-binding domains. Eur. J. Biochem. 156 (1986) 637-643. Parmley, S.F. and Smith, G.P.: Antibody-selectable filamentous Ed phage vectors: affnity purification of target genes. Gene 73 (1988) 305-318. Smith, G.P. (Ed.): Surface display and peptide libraries. Banbury Conference. Gene 128 (1993) 1-144.

Smith, G.P. and Scott, J.K.: Libraries of peptides and proteins displayed on filamentous phage. Methods Enzymol. 217 (1993) 228-257. Uhltn, M., Guss, B., Nilsson, B., Gatenbeck, S., Philipson, L. and Lindberg, M.: Complete sequence of the staphylococcal gene encoding protein A. J. Biol. Chem. 259 (1984) 169551702. Vieira, J. and Messing, J.: Production of single-stranded plasmid DNA. Methods Enzymol. 153 (1987) 3-11.