Adjuvant-independent immunization by immunotargeting antigens to MHc and non-MHc determinants in vivo

Molecular Immunology, Vol. 28, No. 3, pp. 261-267, Printed in Great Britain.

1991

0161-5890/91 $3.00 + 0.00 Pergamon Press plc

ADJUVANT-INDEPENDENT IMMUNIZATION BY IMMUNOTARGETING ANTIGENS TO MHC AND NON-MHC DETERMINANTS IN VW0 GEORGE CARAYANNIOTIS,*DANNA L. SKEA, MARK A. LUSCHER and BRIAN H. BARBER? Department of Immunology, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada MS.5 lA8 (Accepted 1 May 1990) Abstract-Using avidin as a model protein antigen, and biotinylated monoclonal antibodies as a convenient means of forming stable complexes with avidin, we have investigated the adjuvant-independent immunization of three mouse strains, C57BL/6, C3H and (C57BL/6 x C3H)Fl, with immunoconjugates targeted to different class II MHC and non-MHC sites. The results confirm the effectiveness of anti-I-AL and anti-I-Ab immunoconjugates with respect to priming for secondary IgG responses in (H-2b x H-2’)Fl mice, while indicating a lack of response in strains which are homozygous for the targeted allele. In terms of non-MHC targets in the monocyte-macrophage lineage, neither anti-MAC-l nor anti-MAC-2 immunoconjugates were effective in any of the three strains. However, the 33Dl anti-dendritic cell antibody gave significant responses in all three strains, with the Fl response being more than IO-fold greater than the anti-class II immunoconjugates in either strain. These findings indicate that immunotargeting a protein antigen to a non-MHC determinant on dendritic cells in oiuo can be an effective means of inducing an adjuvant-independent serological response, and that this approach can have significant advantages over anti-class II MHC immunotargeting.

INTRODUCTION The recently acquired ability to clone and express, in quantity, individual pathogen genes, or to chemically synthesize peptides which mimic B and T cell epitopes on known protective antigens, affords the potential for a new era of subunit vaccine agents (Zanetti et al., 1987). However, as these immunogens become more defined in composition and construction, they often also become less immunogenic, thus increasing the need for safe and effective adjuvants to augment their immunogenicity. At the moment, alum is the only adjuvant licensed for use in man. This leaves a significant gap between the much stronger adjuvants (such as Freund’s complete adjuvant), available to test subunit vaccine constructs in experimental

animals, and the only option available for human vaccines. In an effort to explore subunit immunization strategies which avoid the need for adjuvants altogether, we have investigated the use of monoclonal antibodies as vehicles for the delivery of antigens to specific cell surface structures in uiuo (Carayanniotis and Barber, 1987, 1990; Carayanniotis et al., 1988; Barber and Carayanniotis, 1988). For this purpose, immunoconjugates consisting of protein or synthetic peptide epitopes coupled to monoclonal antibodies specific for the class II gene products of the major *Present address: Department

of Medicine, Health Sciences Center, Memorial University of Newfoundland, St John’s, Newfoundland, Canada AlB 3V6. tAuthor to whom correspondence should be addressed.

histocompatibility complex (MHC) were constructed. We have demonstrated that immunization with these immunoconjugates, in the absence of an adjuvant, can result in good secondary antibody responses to the delivered antigen. In this paper, we compare immunotargeting of the model protein antigen, avidin, to class II MHC and non-MHC determinants in MHC heterozygous and homozygous mice. The results clearly indicate the potential for improving the response to immunotargeted antigens by a judicious choice of target structure on an appropriate antigen presenting cell. MATERIALS AND

METHODS

Mice Female C57BL/6 (H-2b) (abbreviated B6), C3H (H-2k), and (C57BL/6 x C3H)Fl (H-2b x H-2k) [abbreviated (B6 x C3H)Fl] mice were obtained from Charles River Laboratories (St Constant, PQ). All mice were 9-11 weeks old at the time of priming. Monoclonal antibodies The hybridoma cell lines 10-3.6.2 (anti I-Ak, Oi et al., 1978), 25-9-173 II (anti I-Ab, Ozato and Sachs, 1981), 33Dl (anti-dendritic cell; Nussenzweig et al., 1982), Ml/70.15.11.5HL (anti-MAC-l; Springer et al., 1978), M3/38.1.2.8HL.2 (anti-MAC-2; Ho and Springer, 1982 and H16-LlO-4R5 (anti-influenza A nucleoprotein; Yewdell et al., 1981), were obtained from the American Type Culture Collection. Each was cultured in RPM1 1641 supplemented with 10% 261

262

G.

CARAYANNIOTIS

fetal bovine serum (Gibco, Grand Island, NY), L-glutamine and antibiotics. Monoclonal antibodies were purified from culture supernatants by affinity chromatography on protein G-Sepharose 4 Fast Flow (Pharmacia Canada, Baie D’UrfC, PQ). Biotinylation Monoclonal antibodies were dialyzed against 0.1 M NaHCO, (pH 8.2) and adjusted to a final concentration of 1 mg/ml. Sulfosuccinimidyl 6-(biotinamido) hexanoate (Pierce, Rockford, IL) was dissolved (1 mg/ml) in 0.1 M NaHCO, and this solution was added to each of the antibodies (60 pl/mg antibody). The reaction mixtures were incubated for 1 hr at room temperature and then dialyzed against PBS (pH 7.4). Monovalent avidin Egg avidin (Sigma Chemical Co., St Louis, MO) was dissolved (1 mg/ml) in PBS. The biotin-binding capacity of this solution, on a molar basis, was estimated using 4-hydroxyazobenzene-2’-carboxylic acid (HABA), according to the method of Green (1970). Biotin (Sigma Chemical Co.) was then added slowly, with stirring, to reduce the binding capacity to 1.16 moles of biotin per mole of avidin. By calculations, made assuming the non-cooperative irreversible binding of biotin to avidin (Green, 1975) it was estimated that this solution contained 25.4% fully occupied (inactivated) avidin, 41.5% monovalent avidin, 25.4% bivalent avidin, and 7.6% other species. Immunoconjugates Conjugates were prepared by the addition of monovalent avidin to the biotinylated monoclonal antibodies in the molar ratio of 1.34: 1, or about 1: 1 with respect to the species of avidin in the mixture still able to bind biotin. Conjugates were analyzed by HPLC on a 7.5 x 600mm TSK-250 gel filtration column (Bio-Rad, Richmond, CA), and were used without further purification. Immunization of mice Mice were primed by S.C. injection of 30 pg of antibody-avidin conjugate (containing 10 pg of avidin) in 0.15 ml of PBS. Control mice received 10 pg of free avidin emulsified in Freund’s complete adjuvant (FCA-Difco Laboratories, Detroit, MI), or 10 pg of free avidin in PBS. Five mice of each strain were immunized with each of the antigen preparations. Three weeks later, all of the mice were boosted by i.p. injection of 1Opg of free avidin in 0.2ml of PBS. One week later the mice were bled from the retro-orbital sinus. The blood samples were allowed to clot at room temperature for 1 hr, and were then centrifuged at 1200g for 5 min. Serum samples were stored at 4°C after the addition of 5 ~1 of 10% sodium azide to each sample.

et al.

Measurement of anti-avidin antibody An enzyme-linked immunosorbent assay was used to measure anti-avidin antibody in the serum samples. Microtitre plate (Dynatech-Fisher Scientific, Toronto, ON) wells were coated with 0.1 ml of 10 pg/ml avidin, overnight, at 4°C. The wells were blocked with 1% bovine serum albumin in PBS (BSA-PBS) at room temperature for 1 hr, after which they were washed three times with PBS containing 0.05% Tween-20 (PBS-Tw). Sera were diluted 1: 100 or 1: 5000 with BSA-PBS. Duplicate 0.1 ml aliquots of the diluted sera were incubated in the coated wells at room temperature for 3 hr. The wells were then washed three times with PBS-Tw. Goat anti-mouse IgG conjugated to alkaline phosphatase (ICN ImmunoBiologicals, Montreal, PQ) was diluted 1: 1000 with BSA-PBS, and 0.1 ml of the diluted conjugate were incubated in each well at room temperature for 2.5 hr. The wells were then washed three times with PBS-Tw. Colour development was performed by adding, to each well, 0.1 ml of a substrate solution consisting of 1 mg/ml p-nitrophenyl phosphate (Sigma) in 10% diethanolamine-HCl (pH 9.8). The reaction was allowed to proceed at room temperature for 30 min, at which time it was terminated by the addition, to each well, of 0.025 ml of 5 N NaOH. The optical density at 405 nm was determined using a Titertek Multiskan Plus microELISA plate reader. A standard curve of antibody binding was constructed using a monoclonal anti-avidin antibody (6E8-9H-6B, mouse IgG2a, kindly provided by Dr M. McDermott, McMaster University, Hamilton, ON). A sigmoidal curve was fitted to the standard curve by a computer program (Davis et al., 1980) adapted to BASIC programming by Dr M. Schiff of the University of Toronto. The binding activity in a serum sample was computed from its optical density result in ELISA by interpolation of the fitted curve. The data are expressed as the geometric mean of five mice. Variance is given as the range of the standard error of the geometric mean; this was calculated by taking the anti-logarithm (base 10) of the values resulting from the addition, and subtraction, of the logarithm of the standard error of the geometric mean, to, and from, the logarithm of the geometric mean. Statistical significance of antibody responses was determined within each strain, by comparing the antibody responses of experimental groups to the antibody response of the control group immunized with the H16-LlO-4R5 (anti-influenza A nucleoprotein)-avidin conjugate, using Dunnett’s t-test. RESULTS

Assessment of immunoconjugate formation between different biotinylated MAbs and monovalent avidin Because avidin is tetravalent with respect to its biotin binding capacity (Green, 1975), and therefore

Adjuvant-inde~ndent

~rnuni~tion

has the ability to crosslink biotinylated MAbs into high molecular weight aggregates, we chose to reduce the valency of the avidin by the addition of free biotin. The intention was to create a “functionally monovalent” preparation of avidin, restricted in its ability to create high molecular weight aggregates by crosslinking. A comparison of the complexes formed between a biotinylated MAb and either tetravalent avidin or monovalent avidin to which a 3 : 1 molar ratio of biotin has been added (i.e. monovalent avidin) is depicted in Fig. 1. Even by monitoring the absorbance of these interacting protein mixtures at 28Onm, it is possible to observe the more extensive aggregation resulting from the interaction of tetravalent avidin with the biotinylated MAb, as reflected by an increasing apparent optical density, caused by light scattering from the larger aggregates which form over time. In contrast, the monovalent avidin preparation reflects a stable absorbance at 280 nm, consistent with its more limited potential for crosslinking. Based on these charactersitics of complex formation, we chose to use the monovalent avidin preparation as our model antigen for an assessment of the immunotargeting properties of a series of different immunoconjugates. To define further the charactersitics of the different immunoconjugates being assessed, each combination of biotinylated MAb and monovalent avidin was examined by gel filtration HPLC. The results, depicted in Fig. 2, indicate that for each combination of biotinylated MAb and avidin, conjugate formation was apparent. The extent of immunoconjugate formation was variable, however, with the anti-I-Ak MAb being the least effective participant in this pairing. These gel filtration profiles reflect the species present in the different mixtures assessed as immunogens in vivo. Serological response to anti-class II immunoconjugates injected into class 11 ~omoz~gous or heterozygow recipients

The anti-I-A’ monovalent avidin and anti-I-Abmonovalent avidin immunoconjugates were injected subcutaneously into C3H(H-2k), B6(H-2b) and

“.”

0

2b

eb

4b

100

120

Time6frnin.)

Fig. 1. A comparison in the change in optical density at 28Onm observed with time when the biotinylated 10-3.6.2 (anti-I-Ak) MAb (SOOpg) was mixed in a total volume of 1 ml PBS with 114 pg of unmodified avidin (O), or avidin made monovalent by the addition of biotin (0).

by immunotargeting

263

(B6 x C3H)Fl (H-2b x H-2’) mice. Three weeks after the immunoconjugate priming, all mice were boosted intraperitoneally with free avidin in PBS. The pooled serum IgG response to avidin is depicted in Fig. 3 for the two anti-class II-monovalent avidin immunoconjugates and the control anti-influenza nucleoprotein (FLU-NP j-monovalent avidin immun~onjugate. As described previously (Carayanniotis and Barber, 198’7), immunization with the anti-I-AL monovalent avidin immunoconjugate results in a significant adjuvant-independent anti-avidin response in the (86 x C3H)Fl mice, and a near-background response in B6 mice, bearing a class II gene product (I-Ab) which is unreactive with the anti-I-A’. Background responsiveness is defined by the control antiinfluenza NP-avidin immun~onjugate, which fails to elicit a significant anti-avidin IgG response in the Fl, B6 or C3H mice. However, somewhat unexpected was the low level of response seen with the anti-I-Ak immunoconjugate in the C3H mice, which are homozygous for the expression of 1-A’. This lack of responsiveness in the homozygote versus the Fl mice was mirrored by the results with the anti-I-Ab-monovalent avidin immun~onjugate. Again, a good antiavidin IgG response was seen in the Fl mice, but essentially a background response was observed in the B6 mice homozygous for the I-Ab allele. A comparison of serological responses to avidin delivered by d@erent MHC and non-MHC specific imm~noconjugates

In an effort to determine whether or not immunotargeting to class II MHC antigens represents the optimal “entry point” for adjuvant-independent delivery in vivo, a comparison was made between immunoconjugates of differing specificity containing the same antigen (avidin). The choice of target structures in this experiment was limited to certain MAbdefined differentiation antigens on cells of the monocyte-macrophage cell lineage. The 33D1 antigen, defined by Steinman and co-workers (Nusse~weig et al., 1982) as a marker of murine dendritic cells (DC), was chosen to assess the effect of antigen delivery to this important class of antigen-presenting cells. For comparison, MAbs specific for MAC-l and MAC-2, which recognize differentiation antigens on macrophage subpopulations, were also examined (Ho and Springer, 1982a, 6). The results obtained in terms of the quantity of avidin-specific IgG antibody, elicited by each of these immunoconjugates in the different strains of mice are depicted in Table 1. Quantitation of the anti-I-A’, anti-I-Ab and anti-influenza NP responses depicted in Fig. 3 are reported in Table 1. Statistical significance of the responses obtained for the different immunoconjugates is judged relative to the response seen for the anti-NP control conjugate. For reference, the responses observed when avidin is injected in PBS only, or emulsified in FCA, are also included. The data clearly indicate that immunotargeting avidin with the anti-DC MAb (33Dl)

G. CARAYANNIOTIS et al.

264

Mac-1

Avidin only

40.-60 1-Ab

FLU NP

Fig. 2. A comparison of the HPLC gel filtration profiles obtained on a 7.5 x 600 mm TSK-250 column for biotinylated MAb alone, avidin alone, or the complexes formed by the mixture of the different indicated biotinylated MAbs with an approximately equimolar amount of monovalent avidin.

results in a serological response to avidin which is more than an order of magnitude greater than the response seen with either of the anti-class II MHC immunoconjugates, in the Fl animals. For the antiDC MAb, significant responses were also seen in the homozygous mouse strains. The anti-MAC-l and anti-MAC-2 monovalent avidin immunoconjugates failed to generate statistically significant serological responses to avidin in any of the mouse strains, indicating that the coupling of antigen to MAb which binds to any cell surface determinant is not sufficient to enhance its immunogenicity. It is also noteworthy that anti-MAC-l and anti-MAC-2 antibodies are rat monoclonal antibodies, as is the anti-DC MAb, indicating that the high response seen with the antiDC MAb cannot simply be attributed to its xenogeneic nature with respect to the mouse recipient. All

three rat MAbs were seen to form good immunoconjugates with avidin, as judged by the gel filtration HPLC (Fig. 2), suggesting that the lower responses seen with anti-MAC-1 and anti-MAC-2 were unrelated to the level of complex formation with avidin. DISCUSSION

The experiments described in this paper serve to confirm and extend the utility of MAb-mediated delivery of antigen as a means of generating adjuvantindependent serological responses in viva. In particular, they indicate that at least one non-MHC target structure, the 33Dl antigen on mouse dendritic cells, can provide a more effective entry point than class II MHC for the promotion of serological responses to a coupled protein antigen.

265

Adjuvant-independent immunization by immunotargeting

10

Reciprocal

l&l Serum

1000 Dilufion

.

10 Reciprocal

10

Reciprocal

100 Serum

1OD

Serum

1000

.

..‘.1

10000

Dilution

1000

Dilution

Fig. 3. A comparison of the titration curves observed for the anti-avidin IgG responses to different immunoconjugates (upper panel, anti-I-Ak; middle panel anti-I-Ab; lower panel the control anti-NP) in three different mouse strains C3H (H-2’) (0); B6 (H-2b) (a), and the (B6 x C3H) Fl (+). Sera used represent equal volume pools of four or five animals in each group.

Deridritic cells have been known for some time to be very effective antigen presenting celfs, particularly with respect to the stim~fation of primary responses [reviewed by Steinman and Inaba (1989)]. In a series of in vitro experiments, Inaba and Steinman (1985) demonstrated the crucial role played by MAb 33D1positive dendritic cells in promoting both primary and memory T-helper cell responses to protein antigens. More recently, dendritic cells have also been described as potent stimulators of both alloreactive and virus-specific cytotoxic T-Iymph~yt~ (Boog d al., 1988; Macatonia et af., 1989). The efficiency of dendritic cells as antigen presenting cells is prabably related in their relatively high level of expression of class I and class II MWC gene products, but may also involve antigen-independent features of their ability to interact with T-celfs (Xnaba et at., 1989). Although dendritic cetfs express only modest amounts of 33Dt antigen fapprox. 14,000 copies per cell; Nussenzweig

et d. =(1982)j, it would appear that the immune conjugate spe&c for this determinant permits sufficient avidin to be presented to promote a good secondary serological response. It is interesting to note that adjuvant-independent serological responses can be obtained to an immunotargeted antigen without “directing” the antigen to B-cells. The class II MHC gene product was initially chosen as the target molecule because it was expressed on both antigen-seine B-c&s, and other cells (e.g. macrophages and dendritic cells) capable of presenting processed antigen fragments to T-helper cells. The observation that even better serological responses were obtained when avidin was targeted to dendritic cells with the 33Dl MAb, indicates that effective B-cell priming can occur without a requirement to target antigen to B-cells. Although class 11 MIX-positive macrophages are known to be effective antigen presenting cells in z&w

G. CARAYANNIOTISet al.

266

Table 1. IgG antibody responses to avidin immunotargeted

to MHC and non-MHC determinants

Anti-avidin antibody response Targeting antibody

Specificity

(B6 x C3H)F?g’m1)’

C57BL/6

C3H

10-3.6.2 (mouse IgGZa)

I-Ak

2.07t (1.0&4.30)$

0.087 (0.028-X1.27)$

0.14 (0.062-0.34)$

25-9-179 II (mouse IgGZa)

I-Ab

1.42t (0.78-2.60)

0.064 (0.03&0.11)

0.28 (0.1 l-0.73)

33Dl (rat IgG2b)

Dendtitic cells

27.43t (18.05-41.68)

1.077 (0.35-3.32)

5.25t (3.12-8.82)

M1/70.15.11.5.HL (rat IgGZb)

MAC-l

0.27 (0.1 l-0.66)

0.67 (0.20-2.21)

0.69 (0.35-1.39)

M3/38.1.2.8.HL.2 (rat IgG2a)

MAC-2

0.12 (0.04Wo.32)


0.14 (0.041-0.41)

H16-LlO-4r5 (mouse IgGZa)

0.046 (0.02(M.10)

0.036 (0.018-0.074)

0.025 (0.010-0.061)

PBS

Influenza A Nucleoprotein -

0.032 (0.0140.075)

0.42 (0.30-0.59)

ND

FCA

-

1637.597 (1550.02-1730.11)

192.99t (111.7M33.17)

401.14t (286.12-562.41)

*Geometric mean of five mice. tsignificantly different from the response made by syngeneic mice immunized using the control antibody (H16-L10-4R5, anti-influenza nucleoprotein); P < 0.05, by Dunn&t’s I-test. fRange of the SEM. ND = not determined.

(Unanue, 1981) directing the immunoconjugates to macrophages bearing the MAC-l or MAC-2 antigens in uivo was not particularly successful in eliciting an adjuvant-independent antibody response. MAC-l is expressed constitutively on class II MHC positive and negative macrophages in the mouse (Ho and Springer, 19826), whereas MAC-2 is predominantly expressed on inflammatory macrophages, particularly those induced by thioglycolate administration in the peritoneal cavity (Ho and Springer, 1982~). This latter point potentially explains the absence of any response to the anti-MAC-2 conjugates, whereas the low responses seen for the anti-MAC-l conjugates in B6 and C3H mice might reflect the non-productive delivery of antigen to MAC-l bearing macrophages which fail to express class II MHC gene products. At this point, we have no information on the location or number of cells which encounter the subcutaneously injected immunoconjugates. In the case of the anti-class II MHC immunoconjugates, all indications are that they exert their effect on the serological response to the conjugated antigen by immunospecific recognition of their target class II determinants (Carayanniotis and Barber, 1987, 1990). Until this paper, these data had been collected by assessing responses for anti-I-A’ immunoconjugates in (H-2b x H-2’)Fl mice, using the B6 (H-2b) mice as controls for non-specific responses. The lack of response seen in mice homozygous for the targeted allele (Fig. 3, Table 1) is an observation which could provide further insight with respect to the recognition occurring in uiuo. It is well established from in vitro studies that anti-class II MAbs can block class II-restricted T-cell recognition of processed antigen fragments on antigen presenting cells (Schwartz, 1986). Likewise, large quantities of anticlass II MAbs administered repeatedly in oiuo can be

immunosuppressive, presumably by blocking essential class-II-restricted recognition events (Perry and Williams, 1985). If anti-class II immunoconjugates were to “fix” on class II molecules expressed on cells at or near the site of injection, then in those situations where the target class II molecule was the same as the gene product required for interaction with T-helper cells, blockage of this class II allele could serve to abrogate the T-cell recognition event required for promotion of antibody responses to the delivered antigen. This would be the situation in mice which were homozygous for the targeted class II allele (i.e. anti-I-Ak in C3H or anti-I-Ab in B6 mice), but would not necessarily hold for the (H-2b x H-2k)F1 mice, where the opportunity to present antigen fragments on the alternative (i.e. unblocked) class II allele could circumvent the disruption of interactions via the targeted allele. This would serve to explain why both anti-I-Ak and anti-I-Ab immunoconjugates are effective in the Fl mice, but not in the corresponding parental strains. Casten et al. (1988) have shown, using B-cells as antigen presenting cells in vitro, that it is possible to elicit 1-E’ recognition of processed antigen fragments through immunotargeting the pep tide fragment to I-Ak, H-2Kk or surface Ig expressed on the same B-cell. More recently, Snider and Segal (1989) also demonstrated class-II-restricted presentation of protein antigen fragments by splenic B-cells, when immunoconjugates were directed to surface IgG, IgM or class I MHC gene products. These in vitro studies demonstrate the potential for immunoconjugates recognizing certain cell surface determinants, other than the class II restriction element in question, to deliver antigen to the intracellular compartments required for processing and class-IIrestricted presentation of appropriate antigen fragments. The positive serological response that we

Adjuvant-independent

immun lization by immunotargeting

observed with the immunotargeting of avidin via the 33D1 MAb presumably represents an in viuo example of this phenomenon. In conclusion, the immunotargeting approach to eliciting adjuvant-independent serological responses has now been extended to non-MHC targets within the murine immune system. Using defined, soluble conjugates between the model protein antigen, avidin, and 33D1, an MAb directed against mouse dendritic cells, it has been possible to prime for secondary antibody responses approximately one order of magnitude stronger than those induced with anti-class II MHC immunoconjugates. Further work now will be required to determine the extent to which immunotargeting to dendritic cells might be an effective means of promoting T-cell functions other than those required for secondary IgG responses. Ac~nowledgemenrs-Support for this research was provided by Connaught Laboratories Ltd and the Medical Research Council of Canada. D.L.S. is the recipient of an MRC Postdoctoral Fellowship, and M.A.L. is the recipient of an MRC studentship. REFERENCES Barber B. H. and Carayarmiotis G. (1988) Antigen delivery by immuno~rgeting: a possibility for adjuvant-free vacc&es? In Tech~olog~coIk&awes-in Vaccine ~~~e~o~ment (Edited bv Laskv L.). D. 471. Alan R. Liss. New York. Boo8 C. J. P., Bo& J. ‘and Melief C. J. M. (1988) Role of dendritic cells in the regulation of class I restricted cytotoxic T-lymphocyte responses. J. Immun. 140, 3331-3337. Carayanmotis G. and Barber B. H. (1987) Adjuvant-free IgG responses induced with antigen coupled to antibodies against class II MHC. Nature 327, 59-61. Carayanniotis G. and Barber B. H. (1990) Characteri~tion of the adjuvant-free serological response to protein antigens coupled to antibodies specific for class II MHC determinants. Vuccine 8, 137-i44. Caravanniotis G.. Vizi E.. Parker J. R. M.. Hodees R. S. anh Barber B. H. (1988)Delivery of synthetic peitides by anti-class II MHC monoclonal antibodies induces specific adjuvant-free IgG responses in vivo. Molec. Immun. 25, 907-911. Casten L. A., Kaumaya P. and Pierce S. K. (1988) Enhanced T-class responses to antigenic peptides targeted to B cell surface Ig, Ia or class I molecules. J. Zmmun. 168. 171-180. Davis S. E., Munson P. J., Jaffe M. L. and Rodbard D. (1980) Radioimmunoassav data nrocessina with a small programmable calculator.. J. Imnknoassayl, 15-25. Green N. M., (1970) Spectrophotometric determination of avidin and biotin. In Methods in Enzymology (Edited by McCormick D. B. and Wright L. D.), Vol. 18, part A, p. 418. Academic Press, New York.

267

Green N. M. (1975) Avidin. Ads. Protein Cirem. 29,85-133. Ho M.-K. and Springer T. A. (1982a) Mac-2, a novel 32,000 M, mouse macrophage subpopulation-specific antigen defined by monoclonal antibodies. J. Intmun. 128. 1221-1228. Ho M.-K. and Springer T. A. (1982b) Mac-l antigen: quantitative expression in macrophage subpopulations and tissues, and immunofluorescent localization in spleen. J. Zmmun. 128, 2281-2286. Inaba K., Romani N. and Steinman R. M. (1989) An antigen-ind~endent contact mechanism as an early step in T-cell proliferative responses to dendritic cells. j. exi. Med. 170. 527-542. Inaba K. and Steinman R. M. (1985) Protein-specific helper T-lymphocyte formation initiated by dendritic cells. Science 229, 475-479. Macatonia S. E., Taylor P. M., Knight S. C. and Askonas B. A. (1989) Primary stimulation by dendritic cells induces anti-viral proliferative and cytotoxic T cell responses in vitro. J. exp. Med. 169, 1255-1264. Nussenzweig N. C., Ste~n~an R. M., Witmer M. D. and Gutchirov B. (1982) A mon~lonal antibodv mecific for mouse dendri& cells. Proc. nain. Acad. SC;.. &S.A. 79, 161-165. Oi V. T., Jones P. P., Goding J. W., Herzenberg L. A. and Herzenberg L. A. (1978) Properties of monoclonal antibodies to mouse Ig allotypes, H-2 and Ia antigens. Curr. Topics Microbial. Zmmun. 81, 115-129. Ozato K. and Sachs D. H. (1981) Monoclonal antibodies to mouse MHC antigens: hybridoma antibodies reacting to antigens of the H-2b haplotype reveal genetic control of isotype expression. J. Zmmun. 126. 317-321, Perry-L. L: and Williams I. R. (1985) Regulation of transplantation immunity in vivo by monoclonal antibodies recognizing host class II restriction elements. I. Genetics and specificity of anti-I-A immunotherapy in murine skin allograft recipients. J. Immun. 134, 2935-294 1. Schwartz R. H. (1986) Immune response genes of the murine major histocompatibility complex. Adv. Zmmun. 38, 31-201. Snider D. P. and Segal D. M. (1989) Efficiency of antigen presentation after antigen targeting to surfa& IgD, IgM, MHC, FCy RI1 and B220 molecules on mu&e snlenic B cells. J. Iimun. 143, 59-65. Springer T. A., Galfre G., Secher D. S. and Milstein C. (1978) Monoclonal xenogeneic antibodies to murine cell surface antigens: identification of novel leukocyte differentiation antigens. Eur. J. Zmmun. 8, 539-551. Steinman R. and Inaba K. (1989) Immunogenicity: role of dendritic cells. SioEssays 10, 145-152. Unanue E. R. (1981) The regulatory role of macrophages in antigenic stimulation. Part Two: Symbiotic relations~p between lymphocytes and macrophages. Adv. Immun. 31, l-135. Yewdell J. W., Frank E. and Gerhard W. (1981) Expression of influenza A virus internal antigens on the surface of infected P815 cells. J. Immun. 126, 181&1819. Zanetti M., Sercarz E. and Salk J. (1987) The immunology. of new generation vaccines. Zmmun. Today 8, 18-25.