Eur. J. Immunol. 1990. 20: 2763-2768

Els J. M. Hogervorsto, Maja AgterbergO, JosCe P. A. WagenaarO, Henriette AdriaanseO, Claire J. P. Boogn, Ruurd Van der Zeen, Jan D. A.Van EmbdenA, Willem Van Edenn and Jan TommassenO Institute of Infectious Diseases and Immunology, Department Immunologyo, Faculty of Veterinary Medicine, Institute of Molecular Biology and Medical Biotechnology, Department of Molecular Cell Biologyo, University of Utrecht, Utrecht and Laboratory of Bacteriology, National Institute of Public Health and Enviromental ProtectionA, Bilthoven

Tcell recognition of an cpitope expressed in E. coli PhoE

Efficient recognition by rat T cell clones of an epitope of mycobacterial hsp 65 inserted in Escherichia coli outer membrane protein PhoE* PhoE is a pore-forming protein, abundantly expressed in the Escherichia coli outer membrane. Previous investigations have shown the possibility of inserting antigenic determinants in cell surface-exposed regions of PhoE by recombinant DNA techniques without disturbing the biogenesis and the functioning of the protein. This method proved to be successful for foot-and-mouth disease virus B cell determinants. We have now shown for the first time that PhoE can also be used as a carrier molecule for Tcell epitopes. A well-characterized Tcell epitope (180-188) of the 65-kDa heat-shock protein (hsp 65) of Mycobacterium tuberculosis was expressed in PhoE and tested for recognition by specificTcel1clones. Specific and efficient T cell proliferation was found after stimulation with this protein construct in vitro. Interestingly, paraformaldehyde fixation of antigen-presenting cells did not abrogateTce11recogition.Thus, in contrast to hsp 65 itself, recognition of epitope 180-188 in the context of PhoE appeared to be independent of antigenprocessing events. At the level of polyclonal Tcell responses the epitope in the context of PhoE is recognized more efficiently than 180-188 as synthetic peptide or in the context of the hsp 65 molecule itself.These findings indicate that PhoE may serve as attractive vaccine carrier not only for B, but also for Tcell epitopes. Furthermore, the possibility for expression of PhoE constructs in attenuated Salmonella typhimurium strains offers the exciting prospect of new types of live oral vaccines expressing selected combinations of B and Tcell epitopes.

1 Introduction Recombinant DNA techniques can be used for the highlevel expression of potentially protective antigens or antigenic determinants of pathogenic organisms in a safe and fast growing organism such as Escherichia coli K 12. The expression of these antigenic determinants at the bacterial cell surface is expected to be advantageous for the development of new types of (live oral) vaccines or serodiagnostic reagents. Therefore, we were interested in developing a carrier system based on the well-characterized outer membrane protein PhoE of E. coli K 12. PhoE protein is an abundant outer membrane protein of 36 kDa of E. coli K 12. Its synthesis is induced when cells are grown under phosphate limitation [1]. In its functional trimeric form the protein forms transmembrane channels through which small hydrophilic solutes can pass [2]. According to a model for the folding of PhoE protein [3,4], the polypeptide traverses the outer membrane 16 times in

[I 87071

*

2763

This work was supported by the Foundation for Medical and Health Research (MEDIGON), grant nr. 900-515-003 (H.A.).

Correspondence: Willem Van Eden, Institute of Infectious Diseases and Immunology, Department of Immunology, Faculty of Veterinary Medicine, University of Utrecht, P. 0. Box 80.165, NL-3508 TD Utrecht, The Netherlands Abbreviations: A A: Adjuvant arthritis M T Mycobacterium tuberculosis hsp 65:65-kDa heat-shock protein of Myobacterium bovis BCG MBP: Myelin basic protein 0 VCH Verlagsgcscllschaft mbH, D-6940 Weinheim, 1990

antiparallel fi sheet structure, thereby exposing eight regions at the cell surface (Fig. 1). Theoretically, all these exposed regions can be used for the insertion of antigenic determinants, provided that the assembly of the hybrid proteins is not disturbed by the insertions. Previously, we have described the insertion of different combinations of two B cell determinants of VP1 protein of foot-and-mouth disease virus (FMDV), into the fourth and the fifth exposed region of PhoE protein [5, 6].The hybrid proteins were correctly incorporated into the outer membrane and the inserted epitopes were exposed at the cell surface. Immunization experiments in mice, using partially purified proteins, showed that the FMDV determinants in all constructs were immunogenic [6]. One construct, containing both epitopes, also elicited significant levels of neutralizing antibodies in some animals. This hybrid protein was used for protection experiments in guinea pigs and all animals appeared to be protected against challenge with the virus [7]. However, for a good and long-lasting protection against pathogens, a cell-mediated immune response is also required.To explore the possibilities of using PhoE protein as a carrier to present Tcell determinants, we have chosen a T cell determinant of Mycobacterium tuberculosis (MT), that is of crucial importance in the experimental model of adjuvant arthritis (AA). In the A A model, rats are immunized with whole heat-killed mycobacteria in oil and arthritis develops [8]. From such rats arthritogenic and protective Tcell clones, named A2b and A2c, were raised [9, lo], which recognized in addition to a mycobacterial epitope an epitope associated with cartilage proteoglycans [ll, 121. The MT epitope was characterized as the amino acids at position 180-188 of the 65-kDa heat-shock protein 0014-2980/90/1212-2763$3.50+ .25/O

2764

E. J. M. Hogervorst, M. Agterberg, J. P. A. Wagenaar et al. LA LV LA LV

--

Eur. J. Irnmunol. 1990. 20: 2763-2768 PhoE control

G1y'x I

l

l

T FG LQ LELT

PhoE Tepitope

Asp S e r

Asp A mA l a S eLr y r Ser Asp Uel Gly Tyr Asp His Gln Me( Ser Ala Tyr 110 Lys Arg Val Lys Phe Gly. GIy Tyr Phe Val Lyr Asp Cly Leu Glu Lyr Thr Asn Gln Gly 110 Ala Asp Asn Glu Lyt Asp I l e Asn Tyr

AspT h r A l G a ln Ser Glu Gln Ala Lys Lys Thr Arg Asn Leu Gly Ala Al. Phe Phe Ala Glu Ala Gly Glu Leu Lyr Trp Arg Tyr Lys Gly Tyr Asp Leu Gly Thr Gly Ser Leu Phe Gln

Gly Phe Glu P ro Phe

Met Asp Thr Trp Ala Glu Val Asp Tyr Leu Ala Gly Leu Asn Arg Gly Tyr Asp

Atn

Phe Met Thr Lya Arg Ala Ser Gly Leu Ala Thr Tyr Arg Arn Thr Asp Phe Phe

Am Glu Asn Lya G II Gln Tyr Gln Leu Thr Leu Arn Leu Gly ASP 11s Val Gly

Atn Gln Glu Asn Thr Arg Asp Ser Asn Thr Tyr Ala Gly

Lyr Gln Atn Gly Asp G ~ Y Phe Gly Thr Ser Leu Ser Thr Tyr Ile ASP Ala Phe Phe GIY ASP G l y Ser

Leu Gln Ser Arg Gly Thr Gly Lys Arg Ala Glu A18 Trp Ala Thr Gly Leu Lyr Tyr Asp

Thr Ile Pro Thr Met Lys Arg Thr Glu Ser Tyr Phe Thr Aia Leu Tyr Ile Asn Am Ala

Gly 110 Gly Gly Gly Lyr Phe Glu Asp Asn Ala 110 Glu Asp Asn Asp Asp Ser Lys Lyr Leu Asp Thr Gly Val Leu Gln Lys Asn Gln Asn Ser T y r Asn Phs Leu Ile 110 Glu Lys V a l Asp Ala Tyr Val Tyr Val Gly Gly Asp Leu Ala Val Ala Gln Ser Thr Phe Tyr Pro Tyr Ala Gln Ser Arg Tyr Phe Leu Phe Mot Asp Gly Am Asn Phe LYS

Leu Asn Ile Asn Asn Asp Asp 11s Val Ala Va I Gly

Met Thr

Tir Gln Phe

Figure 1. Model for the topology of PhoE in the outer membrane of E. coli K 12 [4].Theboxed residue Arg 158is cell surface exposed and the position where the Tcell epitope is inserted is indicated.

(hsp 65) of M. bovis BCG by screening the clones for proliferation in the presence of genetically engineered truncated polypeptides of hsp 65 and of synthetic peptides [13,14]. In rats, responsiveness to this epitope was shown to determine susceptibility to A A (Hogervorst et al., submitted). Immunization with hsp 65 appeared to induce resistance to A A and this was also observed in other related animal models of arthritis [13, 15, 161. In this report we show for the first time that PhoE can be used as a carrier molecule for a T cell epitope. A PhoET-epitope hybrid protein is efficiently presented toTcells by APC, even in the absence of processing when fixed APC were used. Also at the polyclonal level, after immunization with hsp 65 or the 180-188 peptide and boostering with MT, good responses to the PhoE-T hybrid protein were obtained.

2 Materials and methods 2.1 Bacterial strains and growth conditions E. coli K 12 strain CE1224 [ 171 produces no pore proteins due to mutations in the phoE and ompR genes. E. coli strain CE1248 [18] also containsphoE and ompR mutations and, in addition, a phoR mutation resulting in constitutive expression of the pho regulon. Cells were grown overnight at 37°C under aeration in L broth [17] or in a phosphatelimited medium described by Levinthal et al. [19]. Where necessary, the medium was supplemented with chloramphenicol (25 kg/ml). 2.2 Construction of the plasmids

Plasmid DNA was prepared as described by Birnboim and Doly [20] followed by CsC1-ethidium bromide isopycnic

centrifugation. Oligonucleotides were automatically synthesized on a Biosearch 8600 DNA synthesizer (Milligen/Biosearch, Burlington, MA) and were purified by preparative gel electrophoresis on a 20% polyacrylamide gel in the presence of 7 M urea. Expression vector pMR05 [21] contains a unique restriction site for Nru I in the DNA, coding for the fourth exposed region of PhoE protein.This plasmid was linearized with Nru I and ligated to a linker coding for theTcell epitope (Fig. 2). This linker, consisting of two complementary synthetic oligonucleotides, was designed in such a way that the insertion in thephoE gene is in the correct reading frame. Furthermore, the linker contains a restriction enzyme recognition site and a stop codon, as indicated in Fig. 2, to facilitate the screening for correct recombinants. The DNA preparation was used to transform strain CE1224, selecting for chloramphenicolresistant colonies. One of the recombinant plasmids was designated pMR20. Nucleotide sequencing according to the dideoxy chain termination method [22] confirmed that the linker was correctly inserted into the Nru I site. As a control, a hybrid protein containing an insertion of the hydrophobic peptide LALVLALY in the fourth exposed region of PhoE was used. This hybrid protein is encoded by plasmid pMR5H2 [23]. 2.3 Isolation and characterization of cell fractions

Trimers of wild-type PhoE (PhoE-WT) and of the hybrid proteins were isolated from 3L overnight cultures as described previously [6, 71. Briefly, cell envelope fractions were isolated by differential centrifugation after ultrasonic disintegration of the cells [24]. These fractions were incubated in a buffer containing 2% SDS for 30 min at 40 "C and the protein-peptidoglycan complexes were isolated by ultracentrifugation [25]. To dissociate PhoE from the peptidoglycan, the complexes were incubated at 40 "C for 30 min in the 2% SDS buffer supplemented with 0.6 M NaCl

Tccll rccognition of an cpitope expressed in E. coli PhoE

Eur. J. Immunol. 1900. 20: 2763-2768

180 T F

G

L

Q

L

E

L

188 T

A ACG TTC GGA CTA CAG CTG GAG CTT ACG TC T TGC AAG CTT GAT GTC GAC CTC GAA TGC AG PVUII STOP Figure 2. The amino acid sequence of theTcell epitope of hps 65 of MT. The numbers correspond t o the positions in hsp 65. Also the nucleotide sequences of two complementary synthetic oligonucleotides, which encode the T cell determinant, are shown. The indicated Pvu I1 restriction site and stop codon were included in the oligonucleotides t o facilitate screening of recombinant plasmids.

[26]. After ultracentrifugation the trimers were isolated from the SN by ethanol precipitation. The protein patterns of whole cells and of cell fractions were analyzed by SDS-PAGE [24]. To determine whether the E . coli PhoE protein and the hybrid proteins are correctly assembled in the outer membrane an ELISA, using whole cells as immobilized antigen, was performed as earlier described [27]. PhoE-specific mAb PP1-3 and PP2-1 were used which recognize conformational epitopes at the cell surfaceexposed parts of the PhoE protein.

2765

0.5% paraformaldehyde (Merck, Darmstadt, FRG) in PBS for 20 min at room temperature. The cells were washed three times with PBS plus 10% FCS at 4°C. After the third washing, the cells were resuspended in PBS plus 10% FCS and incubated for 20 min at 37°C to remove all paraformaldehyde. The cells were washed twice, resuspended in medium, incubated with antigen for 2 h at 37 "C and washed twice before adding to the Tcell clones. 2.7 Polyclonal response to PhoE Lewis rats were immunized i.p. with hsp 65 (2 mg) or the 180-188 peptide (1 mg) in IFA and boostered 8 weeks later with MT (1 mg MT in IFA intracuteanously in the base of the tail). At day 57 after MT immunization popliteal LN cells were isolated and 2 x lo5 cells were tested for proliferation in t h e persence of various concentrations of the hsp 65, the 180-188 peptide and the PhoE hybrid proteins.

3 Results 3.1 Characterization of the hybrid proteins

2.4 T cell clones

The isolation, maintenance and properties of A2b and A2c have been described previously [28, 291. Briefly,Tcell line A2 was isolated from draining LN of Lewis rats immunized with MT (Difco, Detroit, MI) in IFA. Subcloning of A2 resulted in arthritogenic clone A2b and protective T cell clone A2c. The clones were maintained in vitro in Iscove's modified Dulbecco's medium (Gibco, Paisley, Scotland) supplemented with 10% FCS, 10% EL4 SN as IL 2 source, 2 mM L-glutamine and antibiotics.The clones were restimulated every 10 days by incubating 4 x lo5 cells/ml for 3 days with MT in the presence of 1 x 107/mlirradiated syngeneic thymocytes as APC. Control Tcell line Z l a , a myelin basic protein (MBP)-specific T cell line isolated from an MBPimmunized Lewis rat [28], was maintained under similar conditions as A2b and A2c, but restimulated with MBI?

Plasmid pMR20, which contains a phoE gene with an insertion of a linker encoding the 180-188 Tcell epitope of hsp 65 (Fig. 2), was constructed as described in Sect. 2.2. Plasmid pMR5H2 had been previously constructed according to the same procedure [23].This plasmid, which contains a linker encoding eight hydrophobic amino acids, inserted in the same position of the PhoE gene, was used as a control. The expression of the hybrid proteins was investigated in strain CE1224, in which expression of the pho regulon is induced by growth under phosphate limitation and in the constitutive strain CE1248. Correct assembly of

2.5 T cell proliferation assay

Proliferation was determined by incubating 2 x lo4 cell/well for 4 days with various amounts of antigen (0.01 to 50 pg/ml) in the presence of 2 x lo6 irradiated (2500 rad) syngeneic thymocytes. For the last 16 h [3H]dThd was added, the cells were harvested on fiberglass filters and incorporation was measured in a liquid scintillation counter. Besides the PhoE constructs also MT, hsp 65 and the 180-188 peptide of M . bovis BCG hsp 65 were tested. hsp 65 from M . bovis BCG was purified from an overexpressing recombinant E. coli strain [13]. Synthetic peptide 180-188 was prepared by solid-phase techniques according to Steward and Young [30].

2.6 Paraformaldehyde fiiation of APC Thymocytes were fixed according the method of Kovac and Schwartz [31]. Briefly, 6 x lo6 cells were fixed with 50 ml

PhoE--

MS

I

a

b CE1224

I

a

b CE1248

Figure 3. SDS-PAGE protein patterns of whole cells of plasmid containing derivatives of strains CE1224 and CE1248.The plasmids present were pMR05, encoding wild-type PhoE protein (lane a); pMR5H2, encoding PhoE protein with an insertion of eight hydrophobic amino acids (lane b) and pMR20, encoding PhoE with the inserted Tcell epitope (lane c). MS indicates molecular mass standard proteins.

2766

Eur. J. Immunol. 1990. 20: 2763-2768

E. J. M. Hogervorst, M. Agterberg, J. €? A . Wagenaar et al.

the hybrid PhoE proteins into the outer membrane, was investigated by binding of antibodies using an ELISA and whole cells as antigen. mAb PP1-3 and PP2-1 were applied, recognizing conformational epitopes at the cell surfaceexposed parts of PhoE [27]. PP1-3 and PP2-1 could bind to the cells producing the wild-type gene products. Binding of antibody PP1-3 was hardly or not affected by the insertions, indicating that the mutant proteins are correctly incorporated into the outer membrane (data not shown). The binding of antibody PP2-1 was completely abolished by the insertions and this is consistent with the earlier finding that the antigenic determinant recognized by this antibody is located in the region of the insertion [5]. Proteins of whole cell preparations were analyzed by SDS-PAGE (Fig. 3). The amount of hybrid protein expressed in strain CE1224 was comparable to wild-type PhoE protein level (Fig. 3). However, the level of expression of the hybrid protein in strain CE1248 was drastically decreased, as described previously for the pMR5H2encoded protein, but this level was comparable low for both hybrid proteins (Fig. 3). Sufficient amounts of the hybrid proteins could be isolated from the CE1224 derivatives, following the standard procedures, to perform in vitro proliferation experiments.

was found in the presence of APC from BN rats (MHC haplotype RTIN;data not shown) or in the absence of APC (see Table 3, Sect. 3.4). From these data it appeared that epitopes expressed in PhoE are recognized at lower concentrations than epitopes in the native antigen itself. Optimal proliferation was found at a concentration of 1 pg/ml, but even very low concentrations of PhoE-Tprotein (0.01 yglml) were recognized by Tcell clone A2b (stimulation index 13.8; data not shown). The 180-188 peptide, on the other hand, showed optimal proliferation at a concentration of 10 pg/ml. The same concentration of hsp 65 is necessary for optimal proliferation of the clones. In comparison, 1 yg/ml PhoE hybrid protein contains 0.03 pM epitope, 10 yg/ml hsp 65 contains 0.17 p~ epitope and 10 pglml180-188 peptide corresponds with 10 PM epitope, which indicates that epitopes are presentedlrecognized more efficiently in the context of a protein, than as a synthetic peptide. Z la , an MBP-specific T cell clone, showed no response to PhoE protein containing the epitope or one of the control proteins. These results showed that the response to the PhoE-T protein is specific for the inserted 180-188 epitope.

3.2 T cell proliferation in the presence of the PhoE-T hybrid protein

3.3 Polyclonal T cell recognition of the 180-188 epitope in the PhoE-T protein

We now show that the PhoE protein can be used as expression system for T cell determinants. Both T cell clones, A2b and A2c, were able to proliferate in the presence of the 180-188 epitope in the context of the PhoE protein (PhoE-T),while the wild-type PhoE protein (PhoEWT) or the hybrid protein carrying an insertion of eight hydrophobic amino acids (PhoE-H2) did not stimulate the Tcell clones. A representative experiment is shown inTable 1.ThisTcell recognition is MHC restricted because theTcell clones only proliferated in the presence of syngeneic irradiated APC (MHC haplotype RTIL). N o proliferation

To investigate the polyclonal T cell recognition of the 180-188 epitope in the context of PhoE, animals were immunized with hsp 65 or with the 180-188 peptide and boostered with MT. Fifty-seven days later a response was detected against the PhoE-T protein, while the wild-type PhoE protein showed no proliferation '(Table 2). This demonstrated that, even at the polyclonal level, the 180-188 epitope was recognized in the context of PhoE. Recognition of the epitope inserted in PhoE appeared to be more efficient than the recognition of the epitope in the native hsp 65 or assynthetic peptide (Table 2). In the case of immunization with the 180-188 peptide, recognition was only observed in the context of PhoE.

Table 1. Tcell clones A2b and A2c proliferate in the presence of epitope 180-188 inserted in the outer membrane protein PhoE of E. colia)

Antigens usedh) None PhoE-T (1)') PhoE-WT (1) PhoE-H2 (1) MT hsp 65 180- 188 MBP

(10) (10) ( 10)

(5)

Proliferative responses of T cell cloncs Zla A2b A2c

Table2. Recognition of the epitope inserted in PhoE. In vitro prolifcration of rat LN cells after immunization with hsp 65 or peptide 180-188 and boostering with MT

In vitro proliferative response 26 17 216 41 50

43 848 13 106 1344.3 32

67 78

20 24 37 3')

27 801 25 044 19614 23

20 29 I4 5719

26 9 45s

a) The results are given in cpm. b) Concentrations of the antigens (in pglml) are indicated in parentheses. c) 1 pg/rnl PhoE-T, 10 pglml hps 65 and 10 pglml peptide 180-188 correspond with 0.03 pM, 0.17 yM and 10 pM of the epitope. rcspectively.

Imniunization hsp 65

180-1 88

Booster

Antigen

MT

None PhoE-T PhoE-WT lisp 65 180-188

( (10) (25) (25)

5 729 373 1086 390

None PhoE-T PhoE-WT hsp 65 180-188

(10) (10) (25) (25)

I532 13800 1 837 2 095 1996

MT

CPm

403

a) Concentrations of the antigens (in pglml) are indicated in parentheses.

Tccll recognition of an epitope expressed in E. coli PhoE

Eur. J. Immunol. 1990. 20: 2763-2768

3.4 Recognition of the phoE-T protein in the absence of antigen processing Recognition of a protein by Tcells requires processing of the protein by APC. Peptides, however, can also be presented without processing.This is found for the 180-188 peptide in the presence of fixed APC (Table 3). Unexpectedly, we found that the PhoE-Tprotein also induced proliferation in the absence of preceeding processing events, although in this situation higher concentrations of the antigens were necessary for optimal proliferation. In the presence of fixed APC, no response was obtained after stimulation with the 180-188 epitope as part of the native hsp 65 or the whole MT. Preincubation of APC with hsp65 and subsequent fixation, however, was found to induce proliferative responses, showing that the MHC class I1 molecules were still intact (data not shown). I n the latter situation higher concentrations of the antigens were necessary and may possibly be explained by a slight alteration of the MHC molecules by the fixation procedure. No proliferation was found in the absence of APC (Table 3).

Table 3. Fixed APC efficiently present epitope 180-188 in the context of PhoE to T cell clone A2b”)

Antigen

with APC

without APC 40 15

(SO)

(1) (1)

60 37-15 (SO) -1s (SO) 44 (50

21

(50)

13848 (10) 13 106 (10) 13413 (10)

41 (SO) 47 (SO) 19481 (SO)

33 23 61

(10)

None PhoE-T PhoE-WT PhoE-H:!

26 17216

MT hsp 65 180-188

APC with fixed APC

41

50

43

(SO)

(SO) (SO)

a) The results are given in cpm. b) Concentrations of the antigens (in pg/ml) are indicated in parentheses.

4 Discussion In this study, we showed that a genetically engineered bacterial outer membrane protein, PhoE, can be used as an efficient carrier for Tcell epitopes. The epitope used is the arthritis-associated epitope characterized as amino acid sequence 180-188 of the hsp 65 [13]. The assembly of the PhoE protein into the outer membrane was not disturbed by the insertion of this epitope, although the expression in strain CE1248 was drastically reduced.

MHC-restricted Tcell recognition was demonstrated by rat T cell clones A2b and A2c in response to the PhoE-T protein. Low concentrations of t h e PhoE-T protein stimulated the clones, especially when it is considered that only a small part of the PhoE-T-hybrid protein consists of the epitope. Obviously, the epitope is more efficiently recognized in the context of a protein, than as a single peptide. One possible reason for this could be that the epitope,when inserted in PhoE, is less susceptible to degradation than the 180-188 peptide.

2767

The efficient recognition of the 180-188 epitope was not only found when Tcell clones were used, but also when polyclonal Tcells were tested. In animals primed with hsp 65 or the 180-188 epitope and boostered with MT, the proliferative responses in vitro to the 180-188 epitope in the context of PhoE exceeded the responses to the native hsp 65 or the 180-188 peptide. This can perhaps also be explained by the above-mentioned reason. Such efficient recognition of T cell epitopes in the context of PhoE, obviously raises expectations for the use of these constructs for vaccination purposes. Previous investigations have already shown that PhoE can be very useful for vaccination aims. For example, PhoE hybrid proteins containing B cell determinants of theVP1 protein of foot-and-mouth disease virus, were able to induce complete protection in guinea pigs by the induction of neutralizing antibodies [7]. Unexpectedly, we have noted that the epitope inserted in the PhoE protein can be recognized in the absence of proper antigen processing, using fixed APC. However, this presentation appeared not to be as efficient as in the presence of normal APC and higher concentrations of peptide antigen were necessary.This can be explained by a slight alteration of the MHC class I1 molecules of the APC, as result of the fixation. However, the fixation procedure used appeared to be very succesful and no proliferation was found when hsp 65 or MT were used. Moreover, preincubation of APC with hsp 65 or MT before fixation, showed normal proliferative responses. The process of antigen processing and presentation has been the subject of many investigations and thus far it is common knowledge that native proteins are not presented to T helper cells unless they have been processed by APC. Although, the exact pathways of antigen processing and presentation are still unknown, proteins need to unfold or to be proteolytically cleaved, depending on the antigen and theTcells that are used [32]. Nevertheless, an exception has been reported by Lee and Allen [33]. Human fibrinogen, a 340-kDa protein, did not appear to require processing for recognition by two specific murine hybridomas. This was shown by using pre-fixed and chloroquine-treated M@ as APC. It was demonstrated that no involvement of proteases, either on the M a surface or in the culture medium in the presence of fibrinogen occurred. The epitope recognized by the murine hybridomas, located at the carboxyterminal portion of the a chain, turned out to be very hydrophilic and had no defined secundary structure, allowing conformational flexibility resulting in a direct association with the murine I-Ekclass I1 molecule and theTcR. For the 180-188 epitope situated in the unproccessed PhoE, it is unlikely that the same occurs. The membrane-spanning segments of PhoE are antiparallel fi sheets with a very tight structure, resulting in less flexibility.The epitope is inserted in a hydrophilicpart of the PhoE protein but, except for the knowledge that the PhoE protein has a trimeric structure, nothing is known about the conformation after purification. Alternatively, it is possible that the PhoE protein is degraded before fitting in the binding site of the MHC molecule. Such degradation could be caused by the presence of cleavage sites close to the epitope, although it is known that PhoE, in its native structure, is relatively inaccessible for proteases. The insertion of the epitope however, could have changed this property. This is currently being investigated b y the use of protease inhibitors and/or

2768

E. J. M. Hogervorst, M. Agterberg, J. P. A. Wagenaar et al.

by inserting the epitope in other surface-exposed regions of PhoE . Thus, we have shown that an engineered bacterial cellsurface molecule, such as PhoE, may serve as an attractive candidate for the use as a carrier for not only B cell determinants, but also for Tcell determinants. The use of PhoE is also very interesting because of the finding that E. coli PhoE can be normally expressed and incorporated into the outer membrane of two attenuated Salmonella typhimurium strains [34]. Therefore, insertion of combinations of B and Tcell determinants of proteins of bacteria, viruses or other pathogenic organisms in the PhoE protein may be useful for vaccine purposes and could open the way for development of new types of live oral vaccines.

5 References 1 Overbeeke, N. and Lugtenberg, B., FEBS Lett 1980. 112: 229. 2 Benz, R. and Bauer, K., Eur. J. Biochem. 1988. 176: 1. 3 Van der Ley, I?, StruyvC, M. and Tommassen, J., J. Biol. Chern. 1986. 261: 12222. 4 Tommassen, J., in Op den Kamp, J. A. F. (Ed.),Mernbrane Biogenesis, Springer-Verlag. Berlin, Heidelberg 1988, p. 352. 5 Agterberg. M., Adriaanse, H. and Tommassen, J.. Gene 1987. 59: 145. 6 Agterberg, M., Adriaanse, H.. Lankhof, H., Meloen. R. and Tommassen, J., Vaccine 1990. 8: 85. 7 Agterberg, M., Adriaanse, H., Bartelings, S.,Van Maancn, K. and Tommassen, J., Vaccine 1990, in press. 8 Pearson, C. M., Proc. SOC. Exp. Biol. Med. 1956. 92: 95. 9 Holoshitz, J., Matitiau, A. and Cohen, I. R., J. Clin. Owest. 1984. 73: 211. 10 Cohen, I. R., Holoshitz, J. and Van Eden,W., Arthritis Rheum. 1985. 29: 841. 11 Van Eden, W., Holoshitz, J., Ncvo, Z., Frcnkel, A., Kiajman. A. and Cohen. I. R., Proc. Natl. Acad. Sci. USA 1985. 82: 5117.

Eur. J. Immunol. 1990. 20: 2763-2768

12 Van Eden,W., Hogervorst, E. J. M., Hensen, E. J.,Van der Zee, R., Van Embden, J. D. A. and Cohen, I. R., Curr. Top. Microhiol. Irnmunol. 1989. 145: 27. 13 Van Eden, W. ,Thole, J. E. R.,Van der Zee, R., Noordzij, A., Van Embden, J. D. A., Hensen, E . J. and Cohen, I. R., Nature 1988. 331: 171. 14 Van der Zee, R.,Van Eden,W., Meloen, R., Noordzij, A. and Van Embden. J. D. A., Eur. J. Irnrnunol. 1989. 19: 43. 15 Van den Broek, M. F., Hogervorst, E. J. M.,Van Bruggen, M. C. J. ,Van Eden, W. ,Van der Zee, R. and Van den Berg, W., J. Exp. Med. 1989. 170: 449. 16 Billingham, M. E. J., Catney, S., Butler, R., Colston, M. J., J. Exp. Med. 1990. 171: 339. 17 Tommasscn, J..VanTol, H. and Lugtenberg. B., EMBOJ. 1983. 2: 1275. 18 Kortcland, J., Overbeeke, N., De Graaff, I?, Overduin. P. and Lugtenberg, B., Eur. J. Biochern. 1985. 152: 691. 19 Levinthal, C., Signer, E . R. and Fetherolf. K., Proc. Natl. Acad. Sci. USA 1962. 48: 1230. 20 Birnboim. M. C. and Doly, J., Nucleic Acids Res. 1979. 7: 1513. 21 Agterberg, M., Benz, R. and Tommassen, J., Eur. J. Biochem. 1987. 169: 65. 22 Sanger, F.. Nicklen, S. and Coulson, A. R., Proc. Natl. Acad. Sci. USA 1977. 74: 5463. 23 Agterberg, M., Adriaanse, H.,Van Bruggen, A., Karperien, M. and Tommassen, J.. Gene 1990. 88: 37. 24 Lugtenberg, B., Meijers, J., Peters, P. ,Van der Hoek, I-! and Van Alphen, L., FEBS Lett. 1975. 58: 254. 25 Rosenbusch, J. €!, J. Biol. Chern. 1974. 249: 8019. 26 Nakamura, K. and Mizushima, S., J. Biochem. 1976. 80: 1411. 27 Van der Ley, P., Amesz, H . ,Tommassen, J. and Lugtenberg, B., Eur. J. Biochem. 1985. 147: 401. 28 Ben-Nun, A., Werkerle, H., Cohen, I. R., Eur. J. Irnrnunol. 1981. 11: 195. 29 Holoshitz, J., Naparstek, Y , Ben-Nun, A., Cohen, I. R., Science 1983. 219: 56. 30 Steward, J. M. and Young, J. D., Solidphase peptide synthesis. Pearce Chemical Co., Illinois 1984. 31 Kovac, Z . and Schwartz, R., J. Irnmunol. 1985. 134: 3233. 32 Allen, P. M., Irnrnul. Today 1987. 8: 270. 33 Lee, l?, Matsueda, G. R. and Allen, F? M., J. Irnrnunol. 1988. 140: 1063. 34 Agterberg, M., Fransen, R. and Tommassen, J., FEMSMicrohiol. Lett. 1988. 50: 295.