T-Helper Cell-Response to MHC Class II-Binding Peptides of the Renal Cell Carcinoma-Associated Antigen RAGE-1

Immunobiol. (2001) 203, pp. 743-755 © 200 1 Urban & Fischer Verlag http://www.urbanfischer.de/journals/immunobiol

IDepartment of Tumor Progression and Immune Defense, German Cancer Research Center (DKFZ), Heidelberg, 2Department of Applied Genetics, University of Karlsruhe, Germany, and 3Roche, Nutley, N.]., USA

T-Helper Cell-Response to MHC Class II-Binding Peptides of the Renal Cell Carcinoma-Associated Antigen RAGE-l MARIKE

J. J.

G.

1

3

3

STASSAR , LAURA RADDRIZZANI , JORGEN HAMMER ,

and MARGOT ZOLLER

1

,2

Received November 1,2000 . Accepted in revised form March 13,2001

Abstract Recently, epitope prediction software for HLA-D R binding sequences has become available. In view of the importance ofT helper (Th) cell activation in immunotherapy of cancer and evidences supporting immunogenicity of renal cell carcinoma (RCC), we have tested 4 peptides of RAGE-I binding promiscuously to HLA-DR molecules for induction of an immune response. The peptides predicted by the TEPITOPE program using a stringent threshold were derived from the open reading frame 2 and 5 of RAGE-I. Induction of response was evaluated by culturing peripheral blood mononuclear cells (PBMC) in the presence of peptide-loaded dendritic cells (DC) to determine proliferative activity and cytokine expression. Two out of 5 donors did not respond to any of the 4 peptides, 2 donors responded to one peptide and one donor responded to two other peptides. Notably, as revealed by blocking studies and T cell subtype definition, peptides bound to MHC class II molecules and peptide pulsed DC exclusively activated CD4+ T cells, which were of the Thl subtype. With respect to clinical application it is important that (un)responsiveness of individual donors' PBMC was a very consistent feature. Though we have not tested explicitly whether these peptides correspond to naturally processed peptides, the possibility to define those patients whose Th might respond to in silico predicted peptides of RAGE-I, by an in vitro assay, could well be a helpful step towards setting up a RAGE-I based immunotherapeutic protocol.

Introduction At present, the prognosis of patients with RCC remains poor. One third of patients already have a metastatic disease at the initial presentation; 30-40% develop distant metastases after resection of the primary tumor (1). The median survival time of metastatic RCC is only 6-8 months and the 5-year survival rate is less than 5%. Abbreviations: DC = dendritic cell; MHC = major histocompatibility complex; ORF = open reading frame; PI = proliferation index; RAGE = renal cell carcinoma associated gene; RCC = renal cell carcinoma; Th = T-helper cell 0171-2985/011203/05-743 $ 15.00/0

744 . M. J. J. G.

STASSAR

et al.

Chemotherapy rarely appears successful since more than 800/0 of the tumors express the phenotype of multidrug resistance that is not reversible by tamoxifen, cyclosporin, quinine or verapamil (2-4). Although RCC is known to be immunogenic (e.g. 5-10), immunotherapy using cytokines had in most cases a limited impact (e.g. 11-14). Additionally, the adoptive transfer of autologous tumor infiltrating lymphocytes (15, 16) remained poorly effective, and the use oflymphokine-activated killer cells combined with high dose interleukin-2 revealed contradictory results (17, 18). These data underline the urge for improved strategies in the immunotherapy of RCC. Although the central role of Th in the initiation of an immune response is well known, concepts for immunotherapy of cancer have only recently started to focus on Th (19-24). Particularly in RCC patients vaccination either with DC loaded with tumor extracts or vaccination with tumor cell dendritic cell hybrids has been proven to be highly efficient and to induce regression of metastatic RCC (25-27). One approach of eliciting a tumor-specific Th-response relies on loading DC with MHC class II-binding peptides derived from a tumor antigen. Using the melanoma-associated antigen gp 100, we demonstrated recently that DC loaded with peptides that have been predicted by the TEPITOPE program (28, 29) to bind to MHC class II molecules were able to induce a specific Th response in vitro as well as in vivo (30). Encouraged by these results, we decided to explore the efficacy of Th activation by the RCC-related antigen RAGE-l (31). Available data on the frequency of RAGE-l expression in human RCC vary from very rare (1/57) (31) to fairly frequent (7/14) (32), RAGE-l represents one of the very few immunogenic RCC-related antigens (33) defined until today. Thus, it is an interesting candidate target molecule for T-cell-based immunotherapy. In the study presented herein, we describe MHC class II-binding peptides of RAGE-l that elicit, after being loaded onto autologous DC, a specific Th-driven immune response in vitro.

Materials and Methods Collection of peripheral blood and HLA-DR typing

Heparinized blood was collected from 5 healthy volunteers (3 male, 2 female, 27-55 years). PBMC were isolated from heparinized blood by Ficoll gradient centrifugation. HLA-DR typing was done by the PCR-SSP method (34): donor MZ (HLA-DRB1 *0101/1101), OC (HLA-DRB1 *1501/1501), BC (HLA-DRB1 *0408/0701), MS (HLA-DRB1 *8001/1501), and WS (HLA-DRB1 *0101/0301). Monoclonal antibodies

The following hybridomas were used: OKT4 (anti-human CD4, ATCC), OKT8 (anti-human CD8, ATCC) , W6/32 (anti-human MHC class I, ATCC) , 9-3F10 (anti-human MHC class II, ATCC), HNK1 (anti-human NK, ATCC), 63D3 (anti-human monocytes, ATCC). Culture supernatants were purified by passage over ProteinG-Sepharose 4B. The eluted fractions were dialyzed against PBS, concentrated to 1 mg/ml and filter sterilized. Anti-human CD25, CD40L, IL-2, IL-4 IFN-y and FITC or phycoerythrin (PE)-labeled, secondary antibodies were obtained commercially (Pharmingen, Hamburg, Germany). Selection of RAGE-l peptides

The TEPITOPE software was used to predict potential HLA-DR binding peptides with promiscuous binding characteristics as described elsewhere (28, 29). The prediction threshold was set at 20/0 and pep-

T-helper cell-response to RAGE-1 .

745

tides were picked that were predicted to bind to at least 4 of the following 7 HLA-DR molecules DRB1*0101, DRB1*0301, DRB1*0401, DRB1*0701, DRB1*0801, DRB1*1101 and DRB1*1501, which are among the most frequent alleles. Four peptides, two using the ORF 2 and two using the ORF 5 of RAGE-I, consisting of 13AA-16AA have been synthesized by the Central Unit of Peptide Synthesis, German Cancer Research Center, Heidelberg, Germany: peptide 2374 (ORF 2: NRIRNTSTNNQFV), peptide 2375 (ORF 2: NQFVPTMPLPPAR), peptide 2288 (ORF 5: PKLKSGVVRLSSYPT) and peptide 2289 (ORF 5: PVLRPLKAIPASK). Generation of dendritic cells and lymphocyte separation

DC were generated according to a slight modification (35) of the protocol described by Xu et al. (36). Briefly, DC were generated in 24-well plates, seeding 1 X 106 PBMC/well. Mter 24 h of culture, nonadherent cells were removed and plastic-adherent cells were cultured in IMEM/ 10% autologous serum supplemented with 150U GM-CSF, 50U IL-4 and 50U IFN-y. Mature DC were verified by FACS analysis (MHC class II+, CD40+ CD80+ CD86+, CD 14-) and microscopy (veiled cells). Dendritic cells were loaded for 2 h at day 10 of culture with 10 f.1g peptide, if not indicated otherwise. Cultures were washed to remove unbound peptides and autologous PBMC were added. Where indicated, T cells were enriched in PBMC samples by passage over a nylon wool column. In some experiments CD4+ cells were depleted by panning on OKT4 coated plates according to the method ofWYSOCKI and SATO (37). The panning procedure was done twice. Proliferation assay

Autologous PBMC (1 x 106 /ml) were cultured for 3 d on peptide-loaded DC adding 10 f.1Ci/ml [3H]_ thymidine during the last 16 h. Cells were harvested and the incorporation of [3H] -thymidine was determined in a {3-counter. Flow cytometry

FACS analysis followed standard procedures. Briefly, 3-5 X 10 5 cells were stained with the first antibody for 1 h at 4°C, washed, incubated with the second dye-labeled antibody for 30 min at 4 °C in the dark and washed again. For the intracellular staining of cytokines, cells were fixed and permeabilized before staining.

Results Proliferative response after stimulation with DC loaded with HLA-DR binding peptides of RAGE-l

The TEPITOPE program was used for the selection of 4 peptides from the RAGE-1 molecule, which likely bind to the most frequent HLA-DR haplotypes of the Caucasian population. Two peptides originated from the open reading frame 2 and 2 from the open reading frame 5 of the RAGE-1 gene. Autologous PBMC were coincubated with DC which had been loaded with these 4 peptides and proliferative activity was determined after 3 days of culture. PBMC cultured with unloaded DC served as control. Responses were graded according to the proliferation index (PI) as distinct (PI 1.5-2), strong (PI> 2-5) and very strong (PI> 5). To evaluate the consistency of response, 2-10 independent bleedings were collected from each of the 5 donors. As can be seen from Figure 1, from a total of 112 assays, only 2 inconsistencies were observed, i.e. PBMC of two donors mostly not responding to a defined peptide showed one time a borderline and one time a strong response.

746 . M. J. J. G.

STASSAR

et al.

Furthermore, although there was some day to day variability in the strength of the response of individual donors towards a defined peptide, the pattern remained consistent inasmuch as responses were weak (e.g. donor MZ towards 2288) or strong (e.g. donor OC towards 2289). Peptide 2274 did not induce a proliferative response in any of the donors, whereas peptide 2275 and 2288 elicited a consistent response with PBMC of one donor (MZ). PBMC of 2 donors strongly responded to peptide 2289. Thus, 3 out of 5 donor reacted with at least 1 peptide, while 2 donors did not respond. Notably, donor OC is homozygous at the HLA-DRB1 *1501 locus. Since he reacted strongly with the peptide 2289, it is likely that this peptide binds to HLA-DRB1*1501. Yet, donor MS, which also carries one allele ofHLA-DRB 1*1501, does not react with the peptide. This suggests, that either binding of the peptide might be rather weak or that additional factors, e.g. other MHC class II molecules, are involved. Further, donor MZ is the only one that responded to peptides 2288 and 2275. Since only donor MZ carries one allele of the HLA-DRB1 *1101, it might well be that this haplotype efficiently binds peptides 2288 and 2275. Yet, for firm conclusions on this aspect a larger collection of identical as well as of additional HLA-DR haplotypes remains to be tested.

B

A 6

6 peptide 2275

peptide 2274

~ 4

~ c

J::

~ 3

D

~

-g, 2 -------------------------------------------------------1-----

ws

OC

BC

MS

6

ws

MZ

Donor

c

'-'-'-'-'-'-'-'-'-'-'-'i-'-'-'-'-'-'-';'-'-'-'-'-'-'-'i'-'-'-'-'-'-'-'-'-'-'-'

--~---tI,I-------l;--------i -

+--

o

o

o

8

peptide 2288

OC

BC

MS

MZ

Donor

9

peptide 2289

X Ii.

~ 6

~ 5 +---;::J::r--

~ 4

~

~ 3 • ~ -g, 2 -------------------------------------------------------a----

._._._Q._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._.

ws

OC

•

II

BC

MS

Donor

MZ

.2

e4 ~

o

-g,3

2

_ J::

-----.-----------t-----------A.-----------------------------ws

OC

BC

MS

MZ

Donor

Figure 1. Induction of a proliferative response towards MHC class II binding peptides of RAGE-I: PBMC of 5 healthy donors (WS: DR*OI01/0301, OC: DR*1501/1501, BC: DR*0408/0701, MS: DR*8001/1501, MZ: DR*OI01/1101) were tested repeatedly for proliferative activity after coculture with autologous peptide loaded (10 J-Lgi m!) DC. Proliferation indices were calculated from triplicate cultures as the ratio of 3H-thymidine uptake in response to peptide loaded versus unloaded DC. Proliferation indices were committed as -: < 1.5, +: > 1.5-2 (above dotted line), ++: >2-5 (above dashed line), +++: >5 (above full line). (A): peptide 2274 (ORF 2); (B): peptide 2275 (ORF 2); (C): peptide 2288 (ORF 5); (D) peptide 2289 (ORF 5).

T-helper cell-response to RAGE-I .

747

To estimate whether peptide binding may be a limiting factor for induction of response, DC of donor OC were loaded with increasing concentrations (0.1-50 J,.Lg/ml) of peptide (Figure 2). First to mention, proliferative activity clearly increased in dependence in the peptide concentration. Second, peptide loading is saturable, i.e. only a minor increase in proliferative activity was observed when pulsing DC with 50 J..Lg/ml as compared to 20 J,.Lg/ml peptide and no increase in proliferative activity was seen with higher doses of peptide (data not shown). Third and importantly, unresponsiveness was not abrogated by increasing peptide concentrations, which strengthens our interpretation that peptide binding as well as the T cell response are specific. The importance of peptide concentration for induction ofT cell proliferation was strengthened by an additional experiment. Since peptide loading was saturable, we had assumed that the low proliferative response induced by DC loaded with less than 10 J,.Lg/ml peptide may have been due to incomplete saturation of MHC II molecules. Such a deficit should be correctable by the continuous presence of the peptide during the culture period. This has, indeed, been the case. When PBMC were cultured with the preloaded DC in the presence of peptide (1 J,.Lg/ml), maximal proliferative responses were observed already at 10 J,.Lg/ml and at low peptide concentrations proliferation indices significantly exceeded those observed when adding T cell to loaded DC without providing an additional source for "loading". The plateau level as such was unaltered. We next controlled whether peptides do bind to MHC class II molecules. Dendritic cells were incubated with either a non-binding antibody, an anti-MHC I or an antiMHC II antibody. After 30 minutes cells were washed and peptide was added for 2 hours. Cells were washed again and T cells were added. 3H-thymidine incorporation was A5

B 5

4

4 ~

~ .~

"0

~ 3

B

.~

c:

;

~

It::::

~

~ 2

3

2

10

20

30

~mlpeptide

40

50

10

20

30

40

50

~mlpeptide

Figure 2. Proliferative responses in dependence on peptide concentration: (A) DC of donor OC were loaded with increasing concentration of peptides 2288 and 2289. After loading and before addition ofPBMC, non-bound peptides were washed oft (B) DC of donor OC were loaded as described above adding in addition 1 J.,Lg/ml peptide during the period of coculture with PBMC. 3H-thymidine incorporation was determined after 3 days of culture. Proliferation indices of one representative experiment (out of 3) are shown.

748 . M.].]. G. STASSAR et al. Table 1. T lymphocyte subset distribution after stimulation by RAGE-l peptide loaded DC Stimulus a unloaded unloaded unloaded unloaded peptide peptide peptide peptide

DC DC DC DC

2288-loaded 2288-loaded 2288-loaded 2288-loaded

DC DC DC DC

Blocking antibodyb

Proliferation indexc

none irrelevant specificity anti-MHC I anti-MHC II

0.91 1.19 1.00 1.06

none irrelevant specificity anti-MHC I anti-MHC II

3.13 3.22 2.43 1.22*

DC and PBMC were from donor MZ, who responds to peptide 2288; DC were loaded with 10 ~/ml peptide; b Before loading with peptide, DC were incubated for 30 minutes with the indicated antibodies at a concentration of 20 ~g/ml. Antibodies were washed out and peptides were added immediately; cOne of 3 independently performed experiments which revealed similar results is shown. Significance of differences (p < 0.05) is indicated by an asterisk. a

evaluated after 3 days (Table 1). Whereas incubation with the antibody of irrelevant specificity had no bearing on the proliferation index, the proliferation index was slightly decreased after incubation with the anti-MHC I antibod~ However, the proliferation index was significantly reduced only when DC had been incubated with an MHC II binding antibody. Thus, peptides likely bound to MHC II molecules. Dendritic cells loaded with MHC class II binding peptides induce a Th response

In order to characterize the immune response elicited by peptide-loaded DC, a flow cytometric analysis of the cell population after stimulation was perfomed (Table 2). PBMC isolated from donor OC were stimulated with autologous DC that were loaded with peptide 2289 or unloaded. A higher number of cells was recovered from cultures containing peptide 2289 loaded DC (data not shown). Furthermore, there was a relative increase in the percentage ofCD4+ cells (63.7 versus 37.9) and a slight decrease in the percentage of CD8+ cells (15.3 versus 25.9). Importantly, there was a strong increase in CD25+ and CD40L+ cells. An increased percentage of CD4+ CD25+ and CD40L+ cells was also recovered when PBMC of donor MZ were cultured with autologous DC loaded with peptide 2288. In the following experiment, T cells of donor OC were enriched by passage over a nylon wool column and were incubated with DC which were unloaded or loaded with either peptide 2288 (no response) or peptide 2289 (response). Control cultures (unloaded DC and DC loaded with peptide 2288) contained 46.6% and 42.1 % CD4+ cells, while after incubation with peptide 2289 loaded DC 62.30/0 of the recovered cells were CD4+. A relative increase in the percentage of CD25+ and CD40L+ cells was also selectively observed when DC had been pulsed with peptide 2289. To further support the idea of a selective activation/expansion of CD4+ T cells, blocking antibodies were added to the coculture of loaded DC with PBMC. The proliferative

T-helper cell-response to RAGE-1 ·749 Table 2. T lymphocyte subset distribution after stimulation by RAGE-1 peptide loaded DC Stimulusa

Responder population

%

stained cells b

CD4

CD8

CD25

CD40L

unloaded DC (OC) peptide 2289-loaded DC (OC)

PBMC PBMC

37.9 63.7*

25.9 15.3*

12.5 59.9*

13.0 37.4*

unloaded DC (MZ) peptide 2288-loaded DC (MZ)

PBMC PBMC

36.0 69.3*

38.9 34.8

23.5 47.5*

11.6 28.9*

unloaded DC (OC) peptide 2288-loaded DC (OC) peptide 2289-loaded DC (OC)

T cells T cells T cells

46.6 42.1 62.3*

nt nt nt

16.9 18.7 38.6*

10.4 12.0 26.2*

a

b

DC and PBMC/T cells were from donor OC, who responds to peptide 2289, but does not respond to peptide 2288 or from donor MZ, who responded to peptide 2288; Mean values of 3 experiments are shown, significance of differences (p < 0.05) is indicated by an asterisk; nt: not tested. 12

DonorOC

unloaded

peptide 2289 loaded

DonorMZ

unloaded

peptide 2275 loaded

Figure 3. Induction of proliferation by peptide loaded DC can be blocked by anti-CD4: DC of donor OC were loaded with peptides 2289 (10 J..Lg/ml); DC of donor MZ were loaded with peptides 2275 (10 J..Lg/ml). During the period of coculture with PBMC, 10 J..Lg/ml of either OKT4 or OKT8 were added. 3H-thymidine incorporation was determined after 3 days of culture. Mean cpm (per 1 X 10 5 PBMC) ± SO of triplicate cultures are shown. Significance of differences is indicated by an asterisk. The experiment was repeated 3 times revealing consistent results.

response ofPBMC in the absence of peptide and DC was neither influenced byanti-CD8 nor anti-CD4 (data not shown). The same accounted for PBMC cultured in the presence of unloaded DC (Figure 3). When PBMC were stimulated by peptide-loaded DC (donor OC with peptide 2289, donor MZ with peptide 2288), the proliferative activity was reduced by 48% and 560/0 in the presence of 0 KT4, but only by less than 150/0 in the presence ofOKT8. The finding clearly demonstrates that DC loaded with MHC class II binding peptides derived from RAGE-1 initiate preferentially a CD4+ T cell response.

750 . M.].]. G.

STASSAR

et al.

Finally, it was of interest to see whether loading of DC with MHC class II binding peptides of RAGE-1 would activate Th1 orTh2 cells. This was evaluated by the analysis if eytokine expression. PBMC of donor OC and MZ were cultured in the presence of unloaded or peptide-loaded DC (Table 3). When PBMC were cultured with loaded DC a higher percentage of cells expressed IL-2 (28.6%) and IFN-y (38.3%) than in cultures with unloaded DC (IL-2: 14.70/0, IFN-y: 15.2%). A relatively high percentage of cells expressed IL-4, but this was independent of whether DC were unloaded or peptide loaded. Furthermore, a lower percentage of eytokine expressing cells was recovered when PBMC, which had been depleted ofCD4+ cells by panning on OKT4 coated plates, were cultured with DC. This was independent of whether DC were unloaded or peptide loaded. Thus, DC loaded with RAGE-l derived peptides predicted to bind to MHC class II support activation and expansion of CD4+ T cells and these CD4+ T cells express a Thl cytokine pattern.

Table 3. Cytokine expression after stimulation with MHC class II binding RAGE-l peptides Stimulusa

Responder population

0/0 stained cells b IL-2

IL-4

IFN-y

unloaded DC (DC) peptide 2289-loaded DC (DC)

PBMC PBMC

14.7 28.6*

33.7 32.1

15.2 38.3*

unloaded DC (MZ) peptide 2288-loaded DC (MZ)

PBMC PBMC

18.8 33.4*

23.3 27.5

18.8 35.2*

unloaded DC (DC) peptide 2289-loaded DC (DC)

CD4 depleted CD4 depleted

11.7 12.3

18.8 22.1

9.0 15.4

a

b

DC and PBMC/PBMC depleted of CD4+ cells were from donor DC, who responds to peptide 2289, but does not respond to peptide 2288 or from donor MZ, who responded to peptide 2288; Mean values of 3 experiments are shown, significance of differences (p < 0.05) is indicated by an asterisk.

Discussion Vaccination against tumor antigens is hoped to become a powerful weapon against cancer (38, 39). Since DC are known to be the most efficient antigen presenting cell type (40), several studies have been performed using tumor antigen-loaded DC for induction of response (41). Naturally, the induction of an immune response proceeds via activation ofTh cells, that support activation of CTL as well as of effector cells of the non-adaptive immune system (42). However, until recently the knowledge ofMHC class II-restricted peptides derived from tumor antigens has been limited. The new computer program TEPITOPE for the prediction of HLA-DR binding peptides has already proven its reliability (30, 43--45). Encouraged by these results, we have chosen the RCC-associated antigen RAGE-I, which has been the first RCC-associated tumor antigen described to be recognized by autologous CTL (31), for the prediction of MHC class II-restricted pep-

T-helper cell-response to RAGE-l . 751

tides and evaluated the efficacy of these peptides in the activation ofPBMC derived from normal donors. It was found that the RAGE-1 gene contains several ORFs, and until now it remains unclear which of the ORFs codes for an antigenically relevant gene product or whether several of the ORFs are utilized relating e.g. to certain stages of the cell-cycle or to stages of differentiation. However, there are evidences that at least the ORF 2 and 5 are functional. GAUGLER et al. (31) described a peptide recognized by CTL that originated from ORF 2, whereas by elution from MHC class I molecules of a RCC cell line, a peptide from ORF 5 could be identified (46). In addition, it is supposed that ORF 3 also might be transcribed (46). Using the TEPITOPE program 4 different peptides predicted to bind to the most frequent HLA-DR molecules of the Caucasian population were selected, two originating from ORF 2 (2274, 2275) and two from ORF 5 (2288, 2289). To examine the immunogenic properties of the different peptides, DC from 5 healthy donors with different MHC class II haplotypes were loaded and the induction of a proliferative response of autologous PBMC was analyzed. We did not observe any activation using the peptide 2274, and only 1 of 5 donors responded to peptides 2275 and 2288. Peptide 2289 elicited a proliferative response of PBMC of 2 donors. Thus, 3 out of 5 donors reacted with at least 1 peptide, while 2 donors did not respond. As already mentioned in Results, there is evidence that peptide 2289 binds to the HLA-DRB 1*1501 haplotype, albeit weakly, as well as to the HLA-DR*0301 haplotype. Peptides 2275 and 2288 apparently bind to the HLA-DR*1101 molecule. Although larger panels of HLA-DR haplotypes need to be tested, it should be mentioned that the overall frequency is well in line with published evidences. A recent report on TEPITOPE to identify candidate T cell epitopes in the tumor antigen MAGE-3 shows that PBMC of a healthy donor responded strongly to one out of five peptides (44). In search for DR4-restricted MART-1 epitopes, 3 out of 6 predicted peptides were actually binding (45). We observed that PBMC of several healthy donors and melanoma patients mostly responded to 1 or 2 out of 7 peptides of gp 100 selected for promiscuous binding to different HLA-DR haplotypes. Thus, although the peptides were selected for promiscuous binding to a wide range of HLA- 0 R haplotypes, the observed response profiles were much more restricted. Therefore, in addition to a prediction software, a functional selection step, e.g. induction of a proliferative response in vitro, will be essential for recruiting possibly responding patients. Taking into account the demonstrated consistency of response, a reliable prediction in clinical settings using the combination ofTEPITOPE selection and an in vitro evaluation ofT cell activation can be surmised. In the clinical setting, where the patients' tumor and PBMC are available, it will be possible to control, in addition, whether the predicted peptides likely correspond to naturally processed ones by assaying for a secondary response profile after priming with peptide loaded DC and a challenge with DC which have been loaded with a tumor extract. Alternatively and as far as the protein is available, the likelihood of peptides selected by TEPITOPE to correspond to naturally processed ones can also be evaluated by restimulation with protein-loaded DC, as we have demonstrated for gp100/gp100derived peptides (30). One other aspect should be discussed. Proliferation indices increased with the amount of peptide added to the DC, but reached a plateau between 10-20 f..Lg/ml peptide, i.e. occupancy of MHC II molecules was saturable, but saturation was reached at a rather high amount of peptide. In line with this finding was the observation that, particularly at low peptide concentrations, a higher response rate was observed when the peptide was

752 . M.].]. G. STASSAR et al. present during the period of coculture of DC with T cells. We interpret this finding in the sense that in many instances peptides may not bind strongly to a given HLA-DR haplotype. Thus, although the off rate is low, the continuous presence of peptide may help to sustain loading of a sufficient number of MHC class II molecules. This assumption needs to be controlled by defining the affinity of binding of the individual peptides with any particular HLA-DR haplotype. It is, however, interesting to note that several recent reports demonstrate that transfection of DC is superior to loading for induction of response (47, 48). Taking these reports and our findings together, the quantity of loaded DR molecules and the persistence of loading appear to be very important factors for induction of response against tumor-associated antigens, which in most instances are known to be weakly immunogenic (49). Finally, it should be noted that, as far as PBMC did not respond to a given peptide, unresponsiveness was neither influenced by higher peptide concentrations nor by addition of peptides during the coculture of PBMC with DC. Therefore, it is recommendable to evaluate individually the optimal peptide concentration after a selective screening for response, which in our setting reliably worked with 10 f..Lg/ml peptide. Without question, TEPITOPE predicted peptides bind to HLA-DR rather than to HLA-A/B/C molecules and MHC class II loading of DC initiates preferentially activation ofTh cells. Hardly any response was observed when DC were incubated with antiMHC class II antibody peptide loading, while incubation with an anti-MHC class I antibody had only a minor effect. In addition, we noted expansion selectively of CD4+ T cells, Furthermore, responses could be blocked by anti-CD4, but only insignificantly by anti-CD8. Finally, the analysis of cytokine expression provided evidence for activation of Thl cells. Which observations are well in line with a previous study evaluating MHC class II peptide binding of the melanome antigen gpIOO (30). There is ample evidences that activation ofTh has a major impact on the efficacy of immunotherapeutic protocols in cancer (reviewed in 50). Although it remains to be tested whether the selected peptides correspond to naturally processed ones, we consider the fact that a reasonable percentage, i.e. 3 out of 5 donors mounted a Th response against at least one of 4 RAGE-I-derived peptides predicted to bind MHC class II encouraging to follow this line in search for a vaccine against RCC. Acknowledgements This investigation was supported by the Mildred Scheel-Stiftung fur Krebshilfe (MZ). We thank Dr. loannis Mytelineos, Dep. Transplantational Immunology, University of Heidelberg, Heidelberg, Germany, for the identification of the HLA-DR haplotypes.

References 1. RAVAUD, A., and M. DEBLED. 1999.Present achievements in the medical treatment of metastatic renal cell carcinoma. Crit. Rev. Oncol. Hematol. 31: 77. 2. SAMUELS, B. L., D. L. TRUMP, and G. ROSNER. 1994. Multidrug resistance (MDR) modulation in renal cell carcinoma (RCC) using cyclosporin A (CSA) or tamoxifen (CALGB 9163). Proc. Am. Soc. Clin. Oncol. 13: 793. 3. MOTZER, R.]., LYN, FISHER, LlANES, R. L. NGO, C. CORDON-CARDa, and]. O'BRIAN. 1995. Phase IIII trial of dexverapamil plus vinblastine for patients with advanced renal cell carcinoma. J. Clin. Oncol. 13: 1958.

r

r

r

r

T-helper cell-response to RAGE-1 .

753

4. RAVAUD, A., C. CHEVREAU, and L. GEOFFROIS. 1998. Phase II trials with multidrug resistance (MDR) modulation using high dose tamoxifen alone or with quinine in metastatic renal cell carcinoma (MRCC). Proc. Am. Soc. Clin. Oncol. 17: 1313. 5. MILLER, G. A., J. E. PONTES, R. r HUBEN, and M. H. GOLDROSEN. 1985. Humoral immune response of patients receiving specific active immunotherapy for renal cell carcinoma. Cancer Res. 45: 4478. 6. ABUBAKR, Y. A., T. H. CHOU, and B. G. REDMAN. 1994. Spontaneous remission of renal cell carcinoma: a case report and immunological correlates. J. Urol. 152: 156. 7. SAHIN, U., O. TORECI, H. SCHMITT, B. COCHLOVIUS, T. JOHANNES, R. SCHMITS, F. STENNER, G. Luo, I. SCHOBERT, and M. PFREUNDSCHUH. 1995. Human neoplasms elicit multiple specific immune response in the autologous host. Proc. Natl. Acad. Sci. USA 92: 11810. 8. BROUWESTIjN, N., B. GAUGLER, K. M. KRUSE, C. W VAN DER SPEK, A. Mulder, S. Osanto, B. J. van den Eynde, and r I. SCHRIER. 1996. Renal-cell carcinoma-specific lysis by cytotoxic T-lymphocyte clones isolated from peripheral blood lymphocytes and tumor-infiltrating lymphocytes. Int. J. Cancer 68: 177. 9. M. KALLFELZ, D. JUNG, C. HILMES, A. KNUTH, E. JAEGER, C. HUBER, and B. SELIGER. 1999. Induction of immunogenicity of a human renal-cell carcinoma cell line by TAP I-gene transfer. Int. J. Cancer 81: 125. 10. JANTZER, r, and D. J. SCHENDEL. 1998. Human renal cell carcinoma antigen-specific CTLs: antigen-driven selction and long-term persistence in vivo. Cancer. Res. 58: 3078. 11. SAVAGE, r D., and H. B. Muss. 1995. Renal cell cancer. In: DE VITA, V. T., S. HELLMAN, and S. A. ROSENBERG (eds.) Biological Therapy of Cancer. J. B. Lippincott Press, Philadelphia: 373. 12. PALMER, r A., J. ATZPODIEN, T. PHILIP, S. NEGRIER, H. KIRCHNER, H. VON DER MAASE, r GEERTSEN, rEVERS. E. LORIAux, and R. OSKAM. 1993. A comparison of two modes of administration of recombinant interleukin-2-continuous intravenous infusion alone versus subcutaneous administration plus interferon alpha in patients with advanced renal cell carcinoma. Cancer Biother. 8: 123. 13. ROSENBERG, S. A., M. T. LOTZE, J. C. YANG, M. W LINEHAN, C. SEIPP, S. CALBRO, S. E. KARP, R. M. SHERRRY, S. STEINBERG, and D. E. WHITE. 1989. Combination therapy with interleukin-2 and alpha-interferon from the treatment of patients with advanced cancer. J. Clin. Oncol. 7: 1863. 14. SMALL, E.]., G. R. WEISS, U. K. MALIK, l?]. WALTHER, D. JOHNSON, G. WILDLING, T. KUNZEL, A. BEJAMONDE, and V. PATON. 1998. The treatment of human metastatic renal cell carcinoma patients with recombinant gamma interferon. Cancer J. Scient. Am. 4: 162. 15. BUKOWSKI, R. M., and A. C. NOVICK. 1997. Clinical practice guidelines: renal cell carcinoma. Cleve. Clin. J. Med. 64: (Suppl. 1): 11. 16. BELDEGRUN, A., W PIERCE, R. KABOO, C. L. Tso. H. SHAU. r TURCILLO, N. MOLDAWER, S. GOLUB, J. DE KERNION, and R. FIGLIN. 1993. Interferon-alpha primed tumor-infiltrating lymphocytes combined with interleukin-2 and interferon-alpha as therapy for metastatic renal cell carcinoma. J. Urol. 150: 1384. 17. LAw, T. M., R. J. MOTZER, M. MAZUMDAR, K. W SELL, r J. WALTHER, M. O'CONNEL, A. KHAN, V. VLAMIS, N. J. VOGELZANG, and D. F. BAJORIN. 1995. Phase III randomized trial of interleukin-2 with or without lymphokine-activated killer cells in the treatment of patients with advanved renal cell carcinoma. Cancer 76: 824. 18. ROSENBERG, S. A., M. T. LOTZE, L. M. MUUL, A. E. CHANG, F. r AVIS, S. LEITMAN, M. W LINEHAN, C. N. ROBERTSON, R. E. LEE, and J. T. RUBIN. 1987. A progress report on the treatment of 157 patients with advanced cancer using lymphokiner-activated killer cells and interleukin-2, or high-dose interleukine-2 alone. N. Engl. J. Med. 316: 889. 19. ZITVOGEL, L., J. I. MAYORDOMO, T. TJANDRAWAN, A. B. DELEO, M. R. CLARKE, M. T. LOTZE, and W J. STORKUS. 1996. Therapy of murine tumors with tumor peptide-pulsed dendritic cells: dependence on T-cells, B7 costimulation and T helper cell-1 associated cytokines. J. Exp. Med. 183: 87.

754 .

M.

J. J.

G. STASSAR et al.

20. OSTANKOVITCH, M., F. A. LE GAL, F. CONNAN, D. CHASSIN, J. CHOPPIN, and J. G. GUILLET. 1997. Generation ofMelan-NMART-l-specific CD8+ cytotoxicT lymphocytes from human naive precursors: helper effect requirement for efficient primary cytotoxic T lymphocyte induction in vitro. Int. J. Cancer 72: 987. 21. OSSENDORP, E, E. MENGEDE, M. CAMPS, R. FILIUS, and C. J. MELIEF. 1998. Specific T helper cell requirement for optimal induction of cytotoxic T lymphocytes against major histocompatibility complex class II negative tumors. J. Exp. Med. 187: 693. 22. COCHLOVIUS, B., M. LINNEBACHER, M. ZEWE-WELSCHOF, and M. ZOLLER. 1999. Recombinant gp 100 protein presented by dendritic cells elicity a T-helper cell-response in vitro and in vivo. Int. J. Cancer 83: 547. 23. BARRET-BoYES, S. M., A. VLAD, and o. J. FINN. 1999. Immunization of chimpanzees with tumor antigen MUCI mucin tandem repeat peptide elicits bith helper and cytotoxic T-cell responses. Clin. Cancer Res. 5: 1918. 24. PULASKI, B. A., V. K. CLEMENTS, M. R. PIPELING, and S. OSTRAND-RoSENBERG. 2000. Immunotherapy with vaccines combining MHC class II/CD80+ tumor cells with interleukin-12 reduces established metastatic renal disease and stimulates immune effectors and monokine induced by interferon gamma. Cancer Immunol. Immunother. 49: 34. 25. RIESER, C., R. RAMONER, L. HOLTL, H. ROGATSCH, C. PAPESH, A. STENZL, G. BARTSCH, and M. TURNHER. 2000. Mature dendritic cells induce T-helper type-I-dominant immune responses in patients with metastatic cell carcinoma. Urol. Int. 63: 151. 26. KUGLER, A., G. STUHLER, ~ WALDEN, G. ZOLLER, A. ZOBYWALSKI, ~ BROSSART, U. TREFZER, S. ULLRICH, C. A. MULLER, V. BECKER, A. J. GROSS, B. HEMMERLEIN, L. KANZ, G. A. MULLER, and R. H. RINGERT. 2000. Regression of human metastatic renal cell carcinoma after vaccination with tumor cell-dendritic cell hybrids. Nat. Med. 6: 332. 27. MULDERS, ~, C. L. Tso, B. GITLITZ, R. KABOO, A. HINKEL, S. FRAND, S. KrERTSCHER, M. D. ROTH, J. DEKERNION, R. FIGLIN, and A. BELLDEGRUN. 1999. Presentation of renal rumor antigens by human dendritic cells activates tumor-infiltrating lymphocytes against autologous tumor: implications for live kidney cancer vaccines. Clin. Cancer Res. 5: 445. 28. HAMMER, J., T. STURNIOLO, and F. SINIGAGLIA. 1998. HLA class II binding specificity and autoimmunity. Adv. Immunol. 66: 647. 29. STURNIOLO, T., E. BONO, J. DING, L. RADDRIZZANI, O. TUERECI, U. SAHIN, M. BRAXENTHALER, E GALLAZZI, M. ~ PROTTI, E SINIGAGLIA, andJ. HAMMER. 1999. In silico generation of tissue-specific and promiscuous HLA ligand databases using chips and virtual HLA class II matrices. Nature Biotechnol. 17: 555. 30. COCHLOVIUS, B., M. J. J. G. STASSAR, O. CHRIST, L. RADDRIZZANI, J. HAMMER, I. MYTELINEOS, and M. ZOLLER. 2000. In vitro and in vivo induction of a helper T cell response towards peptides of the melanoma-associated gpl00 protein selected by the TEPITOPE program. J. Immunol. 165: 4731. 31. GAUGLER, B., N. BROUWENSTIJN, V. VANTOMME, J. ~ SZIKORA, C. W VAN DER SPEK, J. J. PATARD, T. BOON, ~ SCHRIER, and B. J. VAN DEN EYNDE. 1996. A new gene coding for an antigen recognized by autologous cytolytic T lymphocytes on a human renal carcinoma. Immunogenetics 44: 323. 32. NEUMANN, E., A. ENGELSBERG, J. DECKER, S. STORKEL, E. JAGER, C. HUBER, and B. SELIGER. 1998. Heterogeneous expression of the tumor-associated antigens RAGE-I, PRAME and glycoprotein 75 in human renal cell carcinoma: candidates for T-cell-based immunotherapy. Cancer Res. 58: 4090. 33. GAUDIN, C., ~ Y. DIETRICH, S. ROBACHE, M. GUILLARD, B. ESCUDIER, M. J. LACOMBE, A. KUMAR, E TRIEBEL, and A. CAIGNARD. 1995. In vivo local expansion of clonal T cell subpopulations in renal cell carcinoma. Cancer Res. 55: 685. 34. OLERUP, 0., and H. ZETTERQUIST. 1992. HLA-DR typing by PCR amplification with sequence-specifjc primers (PCR-SSP) in 3 hours: an alternative to serological DR typing in clinical practice including donro-recipient matching in cadvaric transplantation. Tissue Antigens 39: 225.

T-helper cell-response to RAGE-l .

755

35. COCHLOVIUS, B., A. PERSCHL, G. J. ADEMA, and M. ZOLLER. 1999. Human melanoma therapy in the SCID mouse: Targeting and activation of melanoma-specific cytotoxic T cells by bispecific antibody fragments is curative. Int. J. Cancer 81: 486. 36. Xu, H., M. KRAMER, H. ~ SPENGLER, andJ. H. PETERS. 1995. Dendritic cells differentiated from human monocytes through a combination of IL-4, GM-CSF and IFN-gamma exhibit phenotype and function of blood dendritic cells. Advanc. Exp. Med. BioI. 378: 75. 37. WYSOCKI, L. ]., and V. L. SATO. 1978. Panning for lymphocytes: a method for cell selection. Proc. Natl. Acad. Sci. USA. 75: 2844. 38. FORNI, G., ~ L. LOLLINI, ~ MUSIANI, and M. ~ COLOMBO. 2000. Immunoprevention of cancer: is the time ripe? Cancer Res. 60: 2571. 39. BREMERS, A. ]., and G. PARMIANI. 2000. Immunology and immunotherapy of human cancer: present concept and clinical developments. Crit. Rev. Oncol. Hematol. 34: 1. 40. CELLA, M., E SALLUSTO, and A. LANZAVECCHIA. 1997. Origin, Maturation and antigen presenting function of dendritic cells. Curro Opine Immunol. 9: 10. 41. FONG. L., and E. G. ENGLEMAN. 2000. Dendritic cells in cancer immunotherapy. Annu. Rev. Immunol. 18: 245. 42. POST-WHITE, J. 1996. The immune system. Sem. Oncol. Nurs. 12: 89. 43. DE LALLA, C., T. STURNIOLO, L. ABBRUZZESE, J. HAMMER, A. SIDOLI, F. SINIGAGLIA, and ~ PANINA-BORDIGNON. 1999. Cutting edge: identification of novel T cell epitopes in Lol p5a by computational prediction. J. Immunol. 163: 1725. 44. MANICI, S., T. STURNIOLO, M. A. IMRO, J. HAMMER, F. SINIGAGLIA, C. NOPPEN, G. SPAGNOLI, B. MAZZI, M. BELLONE, ~ DELLABONA, and M. ~ PROTTI. 1999. Melanoma cells present a MAGE-3 epitope to CD4(+) cytotoxic T cells in association with histocompatibility leukocyte antigen DRl1. J. Exp. Med. 189: 871. 45. ZAROUR, H. M.,]. M. KIRKWOOD, L. S. KIERSTEAD, W HERR, V. BRUSIC, C. L. SLINGLUFF JR., J. SIDNEY, A. SETTE, and W J. STORKUS. 1999. Melan-NMART-l s1 _73 represents an immunogenic HLA-DR4-restricted epitope recognized by melanoma-reactive CD4+ T cells. Proc. Natl. Acad. Sci. USA 97: 400. 46. FLAD, T., B. SPENGLER, H. KALBACHER, ~ BROSSART, D. BAIER, R. KAUFMANN, ~ BOLD, S. METZGER, M. BLUGGEL, H. E. MEYER, B. KURZ, and C. A. MULLER. 1998. Direct identification of major histocompatibility complex class I-bound tumor-associated peptide antigens of a renal carcinoma cell line by a novel mass spectrometric method. Cancer Res. 58: 5803. 47. REEVES, M. E., R. E. ROYAL, J. S. LAM, S. A. ROSENBERG, and ~ Hwu. 1996. Retroviral transduction of human dendritic cells with a tumor-associated antigen gene. Cancer Res. 56: 5672. 48. TUTING, T., C. C. WILSON, D. M. MARTIN, Y. L. KASAMON, J. ROWLES, D. I. MA, C. L. SLINGLUFF, S. N. WAGNER, ~ VAN DER BRUGGEN,]. BAAR, M. T. LOTZE, and W J. STORKUS. 1998. Autologous human monocyte derived dendritic cells genetically modified to express melanoma antigens elicit primary cytotoxic T cell responses in vitro: enhancement by cotransfection of genes encoding the Th I-biasing cytokines IL-12 and IFN-alpha. J. Immunol. 160: 1139. 49. KIRKIN, A. E, E DZHANDZHUGAZYAN, and J. ZEUTHEN. 1998. The immunogenic properties of melanoma-associated antigens recognized by cytotoxic T lymphocytes. Exp. Clin. Immunogenet. 15: 19. 50. PARDOLL, D. M., and S. L. TOPALIAN. 1998. The role of CD4+ T cells in antitumor immunity. Curro Opine Immunol. 10: 588. Prot Dr. MARGOT ZOLLER, Dept. Tumor Progression and Immune Defense (D0600), German Cancer Research Center (DKFZ) , 1m Neuenheimer Feld 280, D-69120 Heidelberg, Germany. Tel. (+49) 6221-4224 54; Fax (+49) 6221-424769; [email protected]