Myeloma Idiotype Protein-Pulsed Dendritic Cells Produce Strong Anti-Myeloma Effects After Syngeneic Stem Cell Transplantation in Mice

Myeloma Idiotype protein-pulsed dendritic cells produce strong anti-myeloma effects after syngeneic stem cell transplantation in mice

M Zeis1, H Frenzke3, N Schmitz1, L Uharek3 and J Steinmann2

1Department of Internal Medicine II, University of Kiel, Germany; 2Institute of Immunology, University of Kiel, Germany; and 3Department of Internal Medicine II, University of Leipzig, Germany

Summary: Multiple myeloma (MM) is a malignant plasma cell dis- order derived from a B cell clone.1 Median survival time Dendritic cell (DC) vaccination represents an interesting of untreated MM patients is less than 1 year. Treatment immunotherapeutic option in the treatment of several with alkylating agents induces a clinical response in about malignancies. In multiple myeloma (MM) patients, vac- 50% of the patients and prolongs survival to a median of cination with autologous idiotype (Id) protein-pulsed approximately 3 years. Although autologous2,3 or allo- DC is feasible, but their antitumoral effectiveness was geneic stem cell transplantation4 can improve the outcome rather limited. To improve the therapeutic potential of in subgroups of patients, the course of MM is still almost DC therapy, we studied the immunological effects of invariably fatal. New therapeutic approaches to control or syngeneic peripheral blood stem cell transplantation eradicate the malignant clone are definitely needed. (PBSCT) given in conjunction with Id-loaded DC. The idiotype (Id) that is expressed on the immuno- Balb/c mice were inoculated i.p. with 5 × 105 of HOPC globulin (Ig) produced by B cell lymphomas and myelomas myeloma cells (Balb/c origin). Animals were immunized is clonally restricted to tumour cells and thus can serve as with three injections of 5 × 105 DC pulsed with the a tumour-specific antigen. Early studies by Sirisinha and IgG2aHOPC or with a control immunoglobulin (Igcontrol). Eisen5 and Lynch et al6 showed that immunisation with Some experimental groups of myeloma-bearing animals purified Ig derived from a mineral oil-induced plasma- received total body irradiation (7.5 Gy) and a sub- cytoma (MOPC) induced anti-idiotypic immunity in syn- sequent transplant of 2 × 107 syngeneic peripheral blood geneic mice. These findings could be confirmed in several progenitor cells (PBPC) followed by DC therapy begin- other B cell tumour models.7–9 In most of these studies, ning at day 10 post transplant. Animals receiving DC anti-Id immunity was dependent on T cells rather than on therapy or syngeneic PBPCT alone neither induce long- anti-Id antibodies. Bogen et al have shown that Id-specific term survival nor tumor-specific CTL reactivity in vitro. CD4+ T cells are essential for tumour protection in the In marked contrast, combination of syngeneic PBPC MOPC myeloma model.9–12 Several experimental and clini- transplantation and subsequent DC therapy resulted in cal data suggest that dendritic cells (DC) specialised to opti- 78% survival after a follow-up of 180 days. In addition, mally present antigens can be used to enhance the anti- this treatment modality conferred a generation of Id tumour T cell response.13–18 In a murine B cell lymphoma peptide-specific CD8-mediated T cell reactivity. These model, Flamand et al16 have shown that a specific anti-Id data provide a rationale for DC-based vaccination in immunity can be generated in the host by immunisation multiple myeloma patients administered post syng- with DC loaded in vitro with the Id protein. This anti-Id eneic transplantation. immunity confers protection against a subsequent challenge Bone Marrow Transplantation (2002) 29, 213–221. DOI: with a lethal dose of tumour cells. 10.1038/sj/bmt/1703327 Despite the induction of minimal tumour load with high- Keywords: dendritic cells; idiotype; immunoglobulin, dose chemotherapy protocols followed by autologous complementary determining region PBPCT, the therapeutic efficacy of DC vaccination remained limited.19 These observations may be at least partly due to the defective T cell arm of patients’ immune systems. Hence, the objective of the current study was to assess whether transplantation of syngeneic PBPC can enhance the antitumoral efficacy of DC therapy. Our data Correspondence: Dr M Zeis, II. Medizinische Klinik, Knochenmarktrans- plantationseinheit, Universita¨t Kiel, Schwanenweg 20, D-24105 Kiel, show that DC-based vaccination is particularly effective in Germany animals receiving DC therapy early after syngeneic stem Received 30 July 2001; accepted 18 October 2001 cell transplantation. Dendritic cell therapy and PBPCT M Zeis et al 214 Materials and methods OPTI-MEM medium (Life Technologies) for 16 h at 37°C and 5% CO2. Before in vivo use, matured DC were washed Animals, tumour cells and tumour diagnosis twice in Dulbecco’s solution. By flow cytometric analysis,

d these matured DC expressed high levels of CD45 Balb/c (H-2 ) mice were bred and kept in the animal facili- (95% ± 2.3), CD11c (82% ± 1.9), CD18 (90% ± 3.6), CD80 ties of our institute. All animals were housed in conven- (86% ± 4.2), CD86 (82% ± 3.9), class I (92% ± 2.1) and tional cages, seven to 10 animals per cage, and were given class II MHC (91% ± 1.9) antigens. Moreover, morphologi- non-sterilised food and water ad libitum. Cotrimoxazole cal analysis and MLR assay exhibited typical properties of was added to the water for 40 days after bone marrow mature DC (data not shown). transplantation. HOPC-1F was established in culture from a mineral oil- induced Balb/c plasmacytoma, HOPC 1. It produces an Immunisation IgG2a, lambda 1 light chain. HOPC myeloma cells Two days post inoculation of 5 × 105 HOPC-1F myeloma expressed high levels (Ͼ80%) of MHC class I (H2-Dd and d cells (i.p.) Balb/c mice were immunised with three subcut- H2-K ), CD54, CD58 and Fas antigen, but were negative aneous injections (given in the flank) of 5 × 105 pulsed or for MHC class II, NK, T cell markers and lambda idiotype. ␮ Ͻ unpulsed DC (in a volume of 200 l), at weekly intervals. Low levels ( 5%) of the antigens B7–1, B7–2, B220 and Other control groups had been treated with 50 ␮g HOPC- CD19 were detectable (data not shown). MOPC-315 cell 1F-IgG2a␭ (s.c.) or the same dose of the control protein line, an IgA-secreting myeloma, was used as a control tar- given with incomplete Freund’s adjuvant (IFA, Sigma; get. Both cell lines were obtained from American Type 50 ␮l IFA admixed with 50 ␮l protein). Myeloma-bearing Culture Collection (Manassas, VA, USA). animals receiving syngeneic PBPC were treated with the Animals were examined daily for palpable abdominal identical immunisation protocol as described above begin- tumours and necropsied after death. Animals dying of ning on day 10 post transplant. myeloma exhibit bulky abdominal tumours with malignant ascites. Generation of CTL Proteins and peptides Seven days after the last immunisation, tumour-bearing animals were killed under ether anaesthesia. The spleens were HOPC-1F immunoglobulin (Ig) and the control Ig (IgA- removed and pressed through wire mesh screens to obtain MOPC 315) were obtained from Sigma (Deisenhofen, single cell suspensions. Mononuclear cells were isolated by Germany). For ELISPOT assays the following synthetic density gradient centrifugation (Lympholyte M; Cedarlane peptides were used: pp65495–503 (NLVPMVATV) as a Laboratories, Hornby, Canada). Splenocytes were restimu- control peptide, HOPC-1F CDR III light chain214–222 lated at a concentration of 5 × 106 cells/ml with 1 × 106 Id- (ALWYSNHWV) and the HOPC CDR III heavy pulsed DC. In some experiments, splenocytes obtained + chain93–102 (GDYYRRYFDL). All peptides were obtained from DC-treated animals had been enriched for CD4 and from Biosyntan, Berlin, Germany. CD8+ T cells by immunomagnetic isolation (see Enrich- ment of CD4+ and CD8+ T cells). Afterwards, 5 × 106/ml × 6 Bone marrow-derived dendritic cells T cells were coincubated with 1 10 /ml protein-loaded DC in 24-well plates (Nunc). After 5 days, viable cells were As previously reported by Mayordomo et al,15 lymphocyte- harvested and tested in a 51Cr release assay for their ability depleted murine bone marrow cells were cultured overnight to kill HOPC myeloma cells. in complete medium (CM, RPMI 1640 + 10% FCS sup- plemented with 5 × 10−5 m 2-mercaptoethanol) in 24-well plates (106 cells in 1 ml culture medium per well), and then Cytotoxicity assay replated in the presence of cytokines (2.5 × 105 cells per Target cells were obtained from cell culture and were ␮ well) in 1 ml CM containing recombinant murine GM-CSF labelled with 200 Ci of NaCrO4 (Amersham-Buchler, (103 U/ml; CellConcept, Umkirch, Germany) and recombi- Braunschweig, Germany) in 0.5 ml complete medium for nant murine IL-4 (103 U/ml; CellConcept). On day 8, 1 h. They were washed three times with complete medium immature dendritic cells were harvested by gentle pipetting. and added at a concentration of 1 × 104 cells/well in round- bottomed microtitre plates (Nunc). Effector cells were Pulsing and maturation of dendritic cells added at various effector/target ratios in a final volume of 200 ␮l/well. The plates were incubated for 4 h at 37°Cin × 6 Bone marrow-derived dendritic cells (1 10 /ml) were a humid atmosphere with 5% CO2. Maximum chromium incubated in 1 ml OPTI-MEM (Gibco Life Technologies, release was ensured by the addition of 10% Triton-X, and Eggenstein, Germany) in a 6-well plate (Nunc, Oslo, spontaneous release was assessed by adding complete Denmark) for 16 h at 37°C in the presence of 50 ␮g protein medium (RPMI 1640 + 10% FCS) to the target cells. The (1 ␮g/␮l) derived from the myeloma cell lines HOPC-1F culture supernatant was harvested semi-automatically with ␭ (IgG2a␭) or MOPC-315 (IgA ). Cells were then washed a Scatron Titertek System (Scatron, Suffolk, UK) and coun- twice in Dulbecco’s solution (PBS; Gibco, Eggenstein, ted in a gamma counter (Beckmann, Heidelberg, Germany). Germany). Afterwards, immature DC (1 × 106/ml) were The percentage of specific lysis was calculated as incubated with murine TNF-␣ (1000 U/ml; CellConcept) in (experimental cpm − spontaneous cpm)/(maximum

Bone Marrow Transplantation Dendritic cell therapy and PBPCT M Zeis et al 215 cpm − spontaneous cpm) × 100. All determinations were (1G10), anti-CD86 (GL1), anti-MHC class I (H2-Dd, made in hexaplicate. 34–2-12), anti-H-2Kd (clone SF1–1.1) and anti-MHC class II (AMS-32.1). With two exceptions, all monoclonal anti- Blocking inhibition assays with monoclonal antibodies bodies were obtained from Pharmingen; anti-CD3 (KT3) (MoAbs) and anti-CD8 (KT15) antibodies were obtained from Serotec (Wiesbaden, Germany). For MoAb-blocking experiments, effector cells or target cells were incubated for 30 min with MoAbs against CD4 + + (clone RM4–4, Pharmingen, San Diego, CA, USA), CD8 Enrichment of CD4 and CD8 T cells (clone KT15; Serotec, Wiesbaden, Germany) and MHC In some experiments, enriched CD3+/CD4+ and d class I (H-2K clone SF1–1.1; Pharmingen) at a final CD3+/CD8+ T cells derived from the spleens of DC-IgHOPC- b dilution of 1:10. As a control, anti-mouse H-2D (clone treated mice were isolated by immunomagnetic cell separ- KH95; Pharmingen) was used. The cells were washed twice ation with an anti-CD4 or anti-CD8 antibody coupled to with complete medium and tested in a chromium release magnetic microbeads (Miltenyi Biotec, Bergisch Gladbach, assay as described above. Germany). This procedure resulted in a cell population con- sisting of more than 92% of CD4+ (±2.1%) or CD8+ ELISPOT assay (±1.5%) lymphocytes and showed fewer than 2% of B cells, NK cells and monocytes. Interferon-␥ ELISPOT assays were done as described by Okamoto et al.20 One week after the last vaccination, primed splenocytes were harvested and incubated with pep- TNF-␣ ELISA assay tide-pulsed DC for 4 h in OPTI-MEM medium (MNC:DC ratio 10:1). ELISPOT assays were performed in nitrocellu- As described above, splenocytes obtained from animals lose-backed 96-well plates (MAHA-S45; Milipore, which had been immunised with pulsed DC or the complete Bedford, UK), with the anti-mouse interferon-␥ antibodies myeloma protein, were incubated with syngeneic loaded DC in a ratio of 5:1 (splenocytes:DC) in a 24-well plate at R4–6A2 for capture and biotinylated XMG1.2 for detection ° (both Pharmingen). Spots were detected using chromogenic 37 C. After 48 h of incubation, supernatants were harvested and tested for TNF-␣ in a standardised ELISA Assay peroxidase substrate (Sigma), and quantified using the ␣ Bioreader 2000 (Biosys, Freiberg, Germany). (TiterZyme murine TNF- ; PerSeptive Diagnostics, Cam- bridge, MA, USA). Cytokine levels were calculated with reference to standard curves which were constructed from Hematopoietic stem cell transplantation (SCT) the analysis of supernatants containing known amounts of ␣ Transplantation was performed as described elsewhere.21 In recombinant TNF- . brief, a Cs137 source was used for total body irradiation (TBI) of the recipients. A lethal dose of 7.5 Gy was chosen Statistical analysis for conditioning. Syngeneic Balb/c mice intended to serve as donors for PBPC were splenectomized under general Survival of animals was calculated according to the method anesthesia at least 14 days prior to the donation of PBPC. of Kaplan and Meier. Group comparisons were made by Starting 5 days prior to collection of PBPC, 5 ␮g of rhu- Student’s t-test. The calculations were done on a PC with met-G-CSF (Filgrastim; Amgen, Thousand Oaks, CA, ‘Statistica’ data analysis software (StatSoft, Tulsa, OK, USA) were injected s.c. twice daily; the last injection of USA). Filgrastim was given 2 h prior to harvesting the PBPC. The mice were then anticoagulated by heparin, anesthetised and killed by cervical dislocation. The peripheral blood was col- Results lected under sterile conditions by dissection of both carotid arteries. Erythrocytes and debris were removed by density Id-loaded dendritic cells did not induce long-term gradient centrifugation for 20 min at 1500 g at room tem- survival in myeloma-bearing animals perature. Two days post i.p. inoculation of 1 × 105 HOPC- myeloma cells animals received 7.5 Gy of TBI followed by Balb/c mice i.p. inoculated with 5 × 105 of HOPC-myeloma an i.v. injection of 2 × 107 PBPC. cells were treated with three s.c. injections (spaced one week apart) of 5 × 105 specifically pulsed syngeneic DC Immunophenotyping (obtained from healthy Balb/c animals) given 2 days after tumour injection. Whereas untreated animals died within 41 For marker analysis, spleen cells were incubated with days, treatment with specifically pulsed DC (DC-Ig-HOPC) appropriate primary and secondary antibodies and analysed resulted in a significantly prolonged median survival time using a flow cytometer (FACScan; Becton Dickinson, (MST: 62 days) when compared to the tumour control Mountain View, CA, USA). The following antibodies were group (MST: 35 days, P Ͻ 0.01). However, this type of used: anti-CD3 (KT3), anti-CD8 (KT15), anti-CD4 (RM4– vaccination therapy did not elicit long-term survival. Sur- 4), anti-Thy1.2 (30-H12), anti-NK 5E6 (5E6), anti-CD19 vival was not improved when unpulsed DC (MST: 40 (1D3), anti-CD11b (MAC-1), anti-CD11c (HL3), anti- days), DC loaded with the control Ig (DC-Ig-control), or CD45 (30-F11), anti-CD18 (GAME-46), anti-CD80 the complete HOPC-IgG2a combined with IFA (MST: 41

Bone Marrow Transplantation Dendritic cell therapy and PBPCT M Zeis et al

216 100 100 90 tumor control (n = 15) 90 tumor control (n = 12) Ig-HOPC (n = 10) 80 80 TBI only (n = 10) Ig-control (n = 10) PBPCT only (n = 10) 70 70 DC-Ig-HOPC (n = 10) + Ig control (n = 10) 60 DC-Ig-control (n = 10) 60 + Ig-HOPC (n = 10) 50 DC (n = 10) 50 40 40 Survival (%)

Survival (%) 30 30 20 20 10 10 0 0 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 160 180 Days post tumor injection Days post tumor injection

Figure 1 Antimyeloma effects of DC therapy in tumour-bearing animals. Figure 3 Survival of animals after syngeneic PBPCT. Two days after Balb/c mice were given (i.p.) a dose of 5 × 105 HOPC myeloma cells. inoculation of 5 × 105 HOPC cells Balb/c mice were treated with a lethal Two days later, mice received three subcutaneous injections of unmanipul- dose of TBI (7.5 Gy) followed by a transfer of 2 × 107 syngeneic G-CSF- ated DC (DC only), DC pulsed with IgGHOPC (DC-Ig-HOPC), DC pulsed mobilised PBPC. Ten days after transplantation animals received three s.c. with the control Ig (DC-Ig-control), IgHOPC + IFA (Ig-HOPC) or vaccinations spaced 1 week apart containing the myeloma-specific protein Igcontrol + IFA (Ig-control). or the control antibody admixed with incomplete Freund’s adjuvant. Stat- istical analysis: IgHOPC vs PBPCT NS, Igcontrol vs PBPCT NS, TBI vs tumour control P Ͻ 0.01, PBPCT vs tumour control P Ͻ 0.001. days) were administered (see Figure 1). To determine whether this therapeutic concept resulted in a myeloma- longation in MST when compared to the tumour control specific CTL response, tumour-bearing animals were killed Ͻ 1 week after the last vaccination. As depicted in Figure 2, group (see Figure 3; 67 days vs 40 days; P 0.01). On day no specific cytotoxic activity against the HOPC-1F mye- 10 post transplant, animals received three s.c. vaccinations × 5 loma and no significant numbers of IFN-␥-secreting cells spaced 1 week apart containing 5 10 unpulsed DC or reacting to idiotype-specific peptides could be measured those loaded with either the myeloma-specific Ig or the con- (data not shown). trol Ig. As shown in Figure 3, vaccination of transplanted animals with IgHOPC (MST: 67 days) or Ig-control (MST: 57 days) administered in combination with IFA did not sig- High antitumoural efficacy of DC therapy after syngeneic nificantly improve survival compared to the experimental PBPCT group receiving PBPCT only. Similar findings were To evaluate the antitumoural efficacy of Id-pulsed DC syn- obtained by injecting unpulsed DC (MST 56 days) or DC geneic transplantation, animals received 2 days after inocu- loaded with the control protein (MST: 65 days), respect- lation of 5 × 105 HOPC cells, a lethal dose of TBI (7.5 Gy) ively (Figure 4). followed by a transfer of 2 × 107 syngeneic G-CSF-mobil- In marked contrast, animals receiving vaccination with ised PBPC. This procedure resulted in a significant pro- DC pulsed with the HOPC-myeloma protein experienced a powerful antimyeloma effect. Eleven of 14 animals (78%) were alive and free from myeloma after an observation per- 50 DC 100 40 DC-Ig-control 90

DC-Ig-HOPC 80 30 70 60 tumor control (n = 12) 20 50 +DC only (n = 10)

Specific lysis (%) 40 +DC-Ig-control (n = 10) Survival (%) 10 30 +DC-Ig-HOPC (n = 10) 20 0 10 6.25:1 12.5:1 25:1 50:1 0 E:T ratio 0 20 40 60 80 100 120 140 160 180 Days post tumor injection Figure 2 No significant induction of CTL against HOPC myeloma cells in animals bearing a high tumour load. Animals receiving a dose 5 × 105 Figure 4 High antitumoural efficacy of DC therapy after syngeneic HOPC-myeloma cells i.p. were treated with three subcutaneous injections PBPCT. Two days after inoculation of 5 × 105 HOPC cells Balb/c mice of 5 × 105 idiotype protein-pulsed DC spaced 1 week apart (see Immuni- were treated with a lethal dose of TBI (7.5 Gy) followed by a transfer of sation protocol in Methods). Control groups received either DC pulsed 2 × 107 syngeneic G-CSF-mobilised PBPC. Ten days after transplantation, with the control-Ig or untreated DC. One week after the last vaccination, animals received three s.c. vaccinations spaced 1 week apart containing splenocytes were harvested and restimulated for 5 days in vitro with 5 × 105 DCs pulsed with either the myeloma-specific protein or the control specifically pulsed syngeneic DC. Afterwards, CTL were tested for their antibody. Statistical analysis: DC-Ig-HOPC vs DC-Ig-control P Ͻ 0.0001, cytotoxic activity against HOPC-1F. Data were pooled from three differ- DC-Ig-HOPC vs DC only P Ͻ 0.0001, DC-Ig-control vs DC only NS, DC ent experiments. only vs PBPCT NS.

Bone Marrow Transplantation Dendritic cell therapy and PBPCT M Zeis et al 217 iod of 180 days. Moreover, in separate tumour challenge cantly higher levels of TNF-␣ when compared to the experiments a second injection of an otherwise lethal dose control groups (Figure 5b; DC-Ig-HOPC vs DC-Ig-control of 5 × 104 HOPC-myeloma cells given i.p. 3 weeks after the P Ͻ 0.005, DC-Ig-HOPC vs DC only P Ͻ 0.005). last vaccination could be completely eradicated in animals To look for specific reactivity against epitopes derived receiving DC-Ig-HOPC post syngeneic PBPCT without any from the CDRIII region of the HOPC immunoglobulin, histologic evidence of myeloma (data not shown). ELISPOT assays were conducted. In transplanted recipients treated with IgHOPC-pulsed DC or with the complete IgHOPC Induction of idiotype-specific T cell reactivity after DC (admixed with IFA) a significant increase in the frequency therapy post transplant of tumour-specific (heavy chain-derived CDRIII-peptide) T cells could be detected (see Table 1). DC-therapy induced To test whether bone marrow-derived DC pulsed with the higher numbers of IFN-␥-secreting spot forming cells if myeloma-specific protein are capable of inducing a tumour- compared to the experimental groups receiving IgHOPC (DC- specific T cell response post syngeneic PBPCT, splenocytes IgHOPC vs IgHOPC, P Ͻ 0.001). of treated animals were harvested 1 week after the last vac- in vitro + cination and restimulated for 5 days with syngneic Idiotype-primed CD8 T cells specifically lyse HOPC Id-pulsed DC. As shown in Figure 5a, T cells primed with myeloma cells IgHOPC-loaded DC exhibited relevant cytotoxic activity against the HOPC-myeloma target whereas the control tar- To investigate the MHC restriction of the antigen recog- get MOPC 315 was not attacked (data not shown). In paral- nition of these CTL, the ability of CD4, CD8 and MHC I lel, supernatants of these cell cultures contained signifi- antibodies to inhibit cytotoxicity against the MHC class II

50 DC

40 DC-Ig-control

30 DC-Ig-HOPC

20 Specific lysis (%) 10

0 6.25:1 12.5:1 25:1 50:1 E:T ratio b

400

350

300

250 pg/ml a 200 TNF 150

100

Ig-HOPC nonimm. Ig-control DC-Ig-HOPC DC-Ig-control

Figure 5 Induction of idiotype-speciﬁc CTL reactivity after treatment of myeloma-bearing animals with syngeneic PBPCT followed by DC therapy. Ten days post transplant, mice received three subcutaneous injections of unmanipulated DC (DC only), DC pulsed with IgGHOPC (DC-Ig-HOPC), DC pulsed with the control Ig (DC-Ig-control), IgHOPC + IFA (Ig-HOPC) or Igcontrol + IFA (Ig-control). Primed splenocytes were restimulated for 5 days in vitro and tested both for their cytotoxic activity against HOPC-1F (a) and for their potential to secrete TNF-␣ (b). Data were pooled from three different experiments.

Bone Marrow Transplantation Dendritic cell therapy and PBPCT M Zeis et al 218 Table 1 Signiﬁcant increase of VHHOPC-CDRIII-induced ␥-IFN-secreting cells/106 MNC in animals treated with Id-pulsed DC post syngeneic PBPCT

Experimental groups No vaccination Post vaccination

DC-IgHOPC DC-Igcontrol DC only IgHOPC

DC only 9.0 ± 2.4 10.7 ± 2.1 12.0 ± 2.9 9.3 ± 2.0 13.3 ± 2.6 DC + irrelevant peptidea 5.7 ± 2.1 12.0 ± 2.4 14.0 ± 1.6 15.0 ± 1.7 12.7 ± 0.9 DC + HOPC-CDRIII heavy chain peptideb 8.0 ± 3.7 67.7 ± 9.4 12.0 ± 2.9 8.0 ± 2.6 29.7 ± 4.8 DC + HOPC-CDRIII light chain peptidec 9.3 ± 4.2 13.7 ± 5.4 13.3 ± 1.7 10.0 ± 2.0 19.7 ± 3.3

a pp65495–503 (NLVPMVATV). b HOPC Ig-derived CDR III heavy chain93–102 (GDYYRRYFDL). c HOPC Ig-derived CDR III light chain214–222 (ALWYSNHWV). On day 10 post syngeneic PBPCT, myeloma-bearing animals were treated with DC-IgHOPC, DC-Igcontrol,IgHOPC (+ IFA) or unpulsed DC, as described in Methods. One week after the last vaccination splenocytes were harvested and restimulated with unpulsed DC or DC pulsed with different peptides in an ELISPOT assay. Data were pooled from three different experiments.

negative HOPC myeloma was tested. The results, illustrated mechanism in advanced disease of multiple myeloma. in Figure 6, indicate a MHC I-mediated, CD8-dependent Besides the induction of tumour antigen-specific T cell tol- antigen recognition. In addition, IgHOPC-primed enriched erance, several other strategies of malignant cells to avoid CD8+ T cells specifically killed the HOPC target without recognition by cytotoxic T lymphocytes have been affecting the control target MOPC whereas Id-primed CD4+ described.24 For example in our model, HOPC myeloma T cells elicited no relevant effects. cells produce high amounts of interleukin-10 (data not To determine the MHC-restriction of Id-specific CTL shown), an immunosuppressive cytokine capable of generated from mice receiving specifically pulsed DCs inhibiting antitumoural T cell immunity.25 Several clinical (Figure 7) the ability of CD4, CD8 and MHC I antibodies trials were initiated using pulsed dendritic cells derived to inhibit cytotoxicity against the MHC class II negative from monocytes26–28 or CD34+ progenitor cells.29 Wen and HOPC myeloma was tested. The results indicate that Id- coworkers26 treated advanced-stage myeloma patients with primed CTL mediated an MHC class I and CD8-dependent idiotype-pulsed DC. Id-specific immune responses were idiotype recognition. generated, characterised by MHC-dependent T cell prolifer- ative responses with cytokine release and the production of anti-Id antibodies. The immune responses were associated Discussion with a transient minor fall in the serum Id level and were not ablated by high-dose myeloablative chemotherapy. In the present study, we show that treatment with syngeneic Reichardt et al19 reported on 12 MM patients who had been Id-loaded DC elicits most powerful antitumoural effects immunised with Id-pulsed DC given 3 to 7 months post when administered post syngeneic PBPCT. Our data pro- high-dose therapy and autologous PBPCT. Two of 12 vide evidence that Id-specific DC therapy generates a patients developed an Id-specific, cellular immune response tumour-specific immune response mediated by CD8+ T and remained in complete remission whereas patients cells. Furthermore, this vaccination concept enhances the achieving only partial responses post transplant did not pro- frequency of myeloma CDR III peptide-specific T cells. duce any detectable tumour-specific activity post DC ther- Dendritic cells are large, bone marrow-derived, pro- apy. These data show that vaccination with Id-pulsed DC fessional antigen-presenting cells which efficiently stimu- can elicit humoral and in some cases specific T cell late naive T cells to recognise and respond to specific anti- responses. However, despite the induction of a status of gens. Several experimental studies have documented that minimal tumour load by autologous stem cell transplan- DC pulsed with peptide epitopes induce cytotoxic immune tation, the therapeutic effectiveness of idiotype vaccination responses against various tumour antigens.13–18 In our concepts remained limited. One possible explanation of present study, DC therapy failed to induce a lasting thera- these findings may rely on the induction of T cell tolerance peutic effect in animals bearing high numbers of HOPC to the myeloma idiotype proteins, as described above. myeloma cells. These observations could be supported by To overcome this problem, we conducted DC vacci- studies from Bogen.22 Utilising T cell receptor (TCR) trans- nation studies after transfer of naive unprimed T cells con- genic mice specific for a model tumour antigen expressed tained in the graft of syngeneic healthy stem cell donors. on the MOPC-315 myeloma, they showed that in animals This therapeutic concept resulted in very powerful antimy- with high tumour burden the vast majority of autologous eloma effects with the majority of animals being tumour idiotype-specific T cells in the peripheral lymphoid organs free after an observation period of more than 180 days post and in the tumour were functionally deleted. Yi et al23 PBPCT. In addition, tumour challenge experiments in sur- reported that in myeloma patients at an early stage of dis- viving animals revealed that DC therapy produced long- ease, T cell responses to autologous idiotype protein were term antitumoural immunity. In a murine B cell lymphoma significantly higher than in patients with more advanced model, Kwak et al8 showed that vaccination with the stage. These data show that the induction of T cell tolerance tumour-derived immunoglobulin given 14 days post trans- to idiotype proteins represents an important tumour escape plant induced both humoral and cellular antitumour

Bone Marrow Transplantation Dendritic cell therapy and PBPCT M Zeis et al

40 40 219 CD8+ Id-primed T cells 35

30 30

25 20 20 Specific lysis (%) 15 10 Specific lysis (%) 10

5 0 12.5:1 25:1 50:1 0 control antibody anti-CD8 E:T ratio anti-MHC class 1 anti-CD4

40 Figure 7 Effect of blocking antibodies on the cytotoxicity of HOPC CD4+ Id-primed T cells myeloma reactive idiotype-speciﬁc T cells. In vivo primed splenocytes were restimulated in vitro in the presence of Id-pulsed DC. Five days later, blocking experiments using monoclonal antibody against determinants of 30 CD4, CD8 and MHC class I molecules were performed in a 4h 51chromium release assay against HOPC target cells (effector:target ratio 50:1). Data were pooled from three different experiments.

20 ment. Consequently, the precursor frequency of tumour

Specific lysis (%) antigen-specific T cells may be significantly increased if vaccination strategies were carried out during this interval. 10 These observations could be supported by our findings that treatment with Id loaded DC post syngeneic PBPCT resulted in a significant increase of Id peptide-specificT 0 cells. In addition, the current study shows that DC therapy 12.5:1 25:1 50:1 provides the most effective vaccination concept in generat- ing Id-specific T cells if compared to those treatment E:T ratio modalities using unpulsed DC or the complete idiotype protein admixed with IFA (see Table 1) documenting that opti- vs HOPC-1F mal antigen presentation could be accomplished by utilising specifically loaded matured dendritic cells only. vs MOPC-315 Several experimental and clinical studies provide evi- + Figure 6 Cytotoxic reactivity of CD8+ IgHOPC-primed T cells against dence for the existence of CD4 T cells capable of recognis- 9–12,17 HOPC-1F and MOPC 315 myeloma targets. In vivo presensitised spleno- ing myeloma idiotype-specific determinants. In spite cytes of myeloma-bearing mice treated with syngeneic PBPCT and DC of the fact that the vast majority of myelomas failed to were enriched for CD4+ and CD8+ T cells and restimulated in vitro in the express MHC class II molecules, CD4+ T cells may play an presence of Id-pulsed DC. Five days later, CTL activity was measured in important role in tumour immunity. Qin and Blankenstein31 a4h51chromium release assay. Data were pooled from three different experiments. recently showed that cellular immunity against MHC class II negative tumours can be mediated by CD4+ T cells through an IFN-␥-dependent mechanism. In the MOPC-315 responses and protected some mice from an otherwise plasmocytoma model, Bogen and coworkers showed that lethal tumour dose. In our study, striking therapeutic effects the adoptive transfer of a CD4+ T cell clone specific for a were detected when the first dose of idiotype-pulsed DC defined peptide of the ␭-light chain variable region of the was administered on day 10 post syngeneic PBPCT during IgMOPC-315 elicited specific antitumoral effects.12 As hematopoietic engraftment. The rationale to attempt tumour expected, Id-primed CD4+ T cells did not exhibit relevant vaccination in the early post-transplant period during cytotoxicity against the myeloma target in a chromium immune reconstitution is the potential to skew the release assay in our model. It cannot be ruled out that T developing T cell repertoire towards the recognition of helper cells support CD8 effector cells by releasing type 1 tumour antigens. Consistent with this hypothesis are data cytokines (IFN-␥, GM-CSF and TNF-␣) and may act by generated by Mackall et al30 who found that the spectrum inflammatory cytokine effects on nonhematopoietic cells, of antigen specificities of the polyclonal T cell population as shown by Qin and Blankenstein.31 Consistent with this that repopulates after SCT is influenced by encountering hypothesis, Cull et al27 reported two myeloma patients antigens in the early phase post transplant during engraft- responding with an idiotype-specific Th1-directed immune

Bone Marrow Transplantation Dendritic cell therapy and PBPCT M Zeis et al 220 response after receiving a combination of Id-pulsed DC and responses elicited by tumor-derived immunoglobulin M and GM-CSF. its molecular subunits. J Immunol 1987; 139: 2825–2833. The data presented here indicate that vaccination with 8 Kwak LW, Campbell MJ, Zelenetz AD, Levy R. Transfer of idiotype protein-pulsed DCs generate a tumour-specific specific immunity to B-cell lymphoma with syngeneic bone class I-restricted T cell response. Similar findings could be marrow in mice: a strategy for using autologous marrow as an anti-tumor therapy. Blood 1991; 78: 2768–2772. obtained in an allogeneic bone marrow transplant setting. 9 Bogen B, Munthe L, Sollien A et al. Naive CD4+ T cells Utilising the same myeloma model we recently showed that confer idiotype-specific tumor resistance in the absence of immunisation of healthy MHC-identical bone marrow antibodies. Eur J Immunol 1995; 25: 3079–3086. donors with the HOPC-derived idiotype protein induced a 10 Lauritzsen GF, Weiss S, Dembic Z, Bogen B. Naive idiotype- powerful antitumoural effect post transplant in which Id- specific CD4+ T cells and immunosurveillance of B-cell specific CD8+ T cells were involved.32 These data are in tumors. Proc Natl Acad Sci USA 1994; 91: 5700–5704. line with recent studies in humans by Li et al33 demonstrat- 11 Lauritzsen GF, Bogen B. The role of idiotype-specific, CD4+ ing that idiotype-specific CD8+ T cells derived from immu- T cells in tumor resistance against major histocompatibility nised healthy stem cell donors can recognise and kill fresh complex class II molecule negative plasmacytoma cells. Cell autologous myeloma cells. In the autologous setting, O¨ ster- Immunol 1993; 148: 177–188. borg and coworkers34 treated myeloma patients with the 12 Lauritzsen GF, Weiss S, Bogen B. Anti-tumor activity of idiotype-specific, MHC-restricted Th1 and Th2 clones in vitro purified serum M-component by repeated intradermal injec- and in vivo. Scand J Immunol 1993; 7:77–85. tions together with soluble granulocyte–macrophage col- 13 Porgador A, Gilboa E. Bone marrow-generated dendritic cells ony-stimulating factor (GM-CSF). In the majority of pulsed with a class I-restricted peptide are potent inducers of patients an Id-specific predominantly MHC class I-restric- cytotoxic T lymphocytes. J Exp Med 1995; 182: 255–260. ted type T cell response could be detected. To further eluci- 14 Celluzzi CM, Mayordomo JI, Storkus WJ et al. Peptide-pulsed + date the role of CD4 and CD8 T cell subsets in eradicating dendritic cells induce antigen-specific, CTL-mediated protec- residual tumour cells in vivo depletion studies using cyto- tive tumor immunity. J Exp Med 1996; 183: 283–287. toxic antibodies directed against CD4 and CD8 antigens 15 Mayordomo JI, Zorina T, Storkus WJ et al. Bone marrow- are required. derived dendritic cells pulsed with synthetic tumor peptides Taken together, our data show that vaccination concepts elicit protective and therapeutic antitumor immunity. Nat Med with Id protein-pulsed DCs elicit strong anti-myeloma 1995; 1: 1297–1302. effects when administered in the context of syngeneic 16 Flamand V, Sornasse T, Thielemans K et al. Murine dendritic PBPCT. These findings provide a rationale for the use of cells pulsed in vitro with tumor antigen induce tumor resistance in vivo. Eur J Immunol 1994; 24: 605–610. idiotype protein-pulsed DCs in myeloma patients after allo- 17 Timmerman JM, Levy R. Dendritic cell vaccines for cancer geneic stem cell transplantation in which no clinically overt immunotherapy. Ann Rev Med 1999; 50: 507–529. graft-versus-host disease occurred. 18 Greten TF, Jaffee EM. Cancer vaccines. J Clin Oncol 1999; 17: 1047–1060. 19 Reichardt VL, Okada CY, Liso A et al. Idiotype vaccination Acknowledgements using dendritic cells after autologous peripheral blood stem cell transplantation for multiple myeloma – a feasibility study. Blood 1999; 93: 2411–2419. This work was supported by the Deutsche Krebshilfe (Grant No. 20 Okamoto Y, Gotoh Y, Kobayashi Y et al. Detection of 10–1167 and 70–2428-Schm2). interleukin-12 (IL-12)-secreting cells in normal mice with enzyme-linked immunospot (ELISPOT) assay. J Immuno- References assay 2000; 21:25–37. 21 Glass B, Uharek L, Zeis M et al. Allogeneic peripheral blood 1 Bataille R, Harousseau JL. Multiple myeloma. New Engl J progenitor cell transplantation in a murine model: evidence Med 1997; 336: 1657–1664. for an improved graft-versus-leukemia effect. Blood 1997; 90: 2 Barlogie B, Jaganath S, Vesole D, Tricot G. Autologous and 1694–1700. allogeneic transplants for multiple myeloma. Semin Hematol 22 Bogen B. Peripheral T-cell tolerance as a tumor escape mech- + 1995; 32:31–44. anism: deletion of CD4 T-cells specific for a monoclonal 3 Attal M, Harousseau JL, Stoppa AM et al. A prospective, ran- immunoglobulin idiotype secreted by a plasmocytoma. Eur J domised trial of autologous bone marrow transplantation and Immunol 1996; 26: 2671–2679. chemotherapy in multiple myeloma. New Engl J Med 1996; 23 Yi Q, Osterborg A, Bergenbrant S et al. Idiotype-reactive T- 335:91–97. cell subsets and tumor load in monoclonal gammopathies. 4 Cavo M, Bandini G, Benni M et al. High-dose busulfan and Blood 1995; 86: 3043–3049. cyclophosphamide are an effective conditioning regimen for 24 Pawelec G, Heinzel S, Kiessling R et al. Escape mechanisms allogeneic bone marrow transplantation in chemosensitive in tumor immunity: a year 2000 update. Crit Rev Oncol 2000; multiple myeloma. Bone Marrow Transplant 1998; 22:27–32. 2:97–133. 5 Sirisinha S, Eisen HN. Autoimmune-like antibodies to the 25 Salazar-Onfray F. Interleukin-10: a cytokine used by tumors ligand-binding sites of myeloma proteins. Proc Natl Acad Sci to escape immunosurveilance. Med Oncol 1999; 16:86–94. USA 1971; 68: 3130–3135. 26 Wen YJ, Ling M, Bailey-Wood R, Lim SH. Idiotypic protein- 6 Lynch RG, Graff RJ, Sirisinha S et al. Myeloma proteins as pulsed adherent peripheral blood mononuclear cell-derived tumor-specific transplantation antigens. Proc Natl Acad Sci dendritic cells prime immune system in multiple myeloma. USA 1972; 69: 1540–1544. Clin Cancer Res 1998; 4: 957–962. 7 Campbell MJ, Carroll W, Kon S et al. Idiotype vaccination 27 Cull G, Durrant L, Stainer C et al. Generation of anti-idiotype against murine B cell lymphoma: humoral and cellular immune responses following vaccination with idiotype-protein

Bone Marrow Transplantation Dendritic cell therapy and PBPCT M Zeis et al 221 pulsed dendritic cells in myeloma. Br J Haematol 1999; 107: involves inhibition of angiogenesis that is dependent on IFN 648–655. gamma receptor expression by nonhematopoietic cells. 28 Lim SH, Bailey-Wood R. Idiotypic protein-pulsed dendritic Immunity 2000; 12: 677–686. cell vaccination in multiple myeloma. Int J Cancer 1999; 83: 32 Zeis M, Steinmann J, Petrela E et al. Transfer of idiotypic 215–222. protein primed allogeneic marrow grafts elicits potent graft- 29 Titzer S, Christensen O, Manzke O et al. Vaccination of mul- versus-myeloma effects in mice. Bone Marrow Transplant tiple myeloma patients with idiotype-pulsed dendritic cells: 2001; 27: 279–285. immunological and clinical aspects. Br J Haematol 2000; 108: 33 Li Y, Bendandi M, Deng Y et al. Tumor-speciﬁc recognition 805–816. of human myeloma cells by idiotype-induced CD8+ T cells. 30 Mackall CL, Bare CV, Granger LA et al. Thymic-independent Blood 2000; 96: 2828–2833. T-cell regeneration occurs via antigen-driven expansion of 34 O¨ sterborg A, Yi Q, Henriksson L et al. Idiotype immunization peripheral T-cells resulting in a repertoire that is limited in combined with granulocyte–macrophage colony-stimulating diversity and prone to skewing. J Immunol 1996; 156: factor in myeloma patients induced type I, major histocompat- 4609–4616. ibility complex-restricted, CD8- and CD4-speciﬁc T-cell 31 Qin Z, Blankenstein T. CD4+ T cell-mediated tumor rejection responses. Blood 1998; 91: 2459–2466.

Bone Marrow Transplantation