Dendritic cells in Melanoma
Polak, M.E.
Citation
Polak, M. E. (2008, September 16). Dendritic cells in Melanoma. Retrieved from https://hdl.handle.net/1887/13100
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/13100
Note: To cite this publication please use the final published version (if applicable).
Chapter VIII
Summary and conclusion
Metastatic melanoma of either cutaneous or uveal origin is poorly responsive to
conventional treatment, and is associated with a median survival of 7–9 months. There is therefore an urgent need for new treatments. Melanomas are immunogenic tumours and the possibility of successful melanoma treatment with anti-tumour vaccines fires the imagination, but although numerous strategies have been tested, no truly successful vaccine has as yet been developed. Despite the presence of potent anti-tumour immune cells in patient’s blood, in more than 95% of patients, tumour growth remains unaffected [1-5]. The co-existence of systemic anti-tumour immunity and tumour progression in the same individual remains one of the major paradoxes of melanoma immunology. As the last paper in the thesis summarizes (Polak et al., submitted), little progress has been observed in the field of melanoma vaccines, in spite of ample studies, many of which were performed between 2003 and 2008. As the numerous approaches explore every possible vaccine design, mode of administration and use of potent immune adjuvants, and in majority successfully generate populations of functional tumour-specific T
lymphocytes in vitro and in the peripheral blood, clearly a major physiological obstacle exists, which prevents the vaccines from working in patients. To investigate the
mechanisms that help melanomas to escape from the cytotoxicity of T lymphocytes after activation by vaccination, we examined the local microenvironment created by this tumour, focusing on the presence of immunomodulatory mechanisms.
It is of interest to compare the immune escape mechanisms of uveal and cutaneous melanoma, as these two tumours share a common origin, but develop in two different immunological microenvironments. As uveal melanoma develops in an
immunoprivileged site, it is already in a much advantaged position. Nevertheless, it contributes to the ocular immunosuppression by secreting a range of immunomodulatory factors. On the contrary, cutaneous melanoma grows in skin, which in itself constitutes an immunological barrier for pathogens, and comprises a network of active immune system effectors; therefore, the melanoma cells need to avoid intense immune surveillance for successful growth of the tumour. Since immune escape is critical for successful cutaneous melanoma growth, it is intriguing that this cancer metastasizes via the lymphatic pathway, where the probability of recognition is highest.
127
Dendritic cells orchestrate the immune response in the peripheral tissues. They sense the microenvironment and recognize potentially dangerous pathogens, initiating the immune reaction against them while tolerizing T lymphocytes towards harmless antigens. We therefore attempted to investigate whether the development of a melanoma-associated local immunosuppressive microenvironment has an impact on the biology and function of peripheral dendritic cells.
To determine the phenotype of DCs residing in tumours, we studied primary uveal melanoma and lymph nodes with advanced cutaneous melanoma metastasis. A population of peripheral DCs is very heterogeneous, and a common conundrum in studying their biology is lack of a population-specific marker. To visualize tissue DC irrespective of their origin, functional subset or maturation status, we chose FXIIIa, which has previously been used by others in studies of dermal DCs [6]. The morphology of positively-stained cells was further assessed by microscopic examination by a
pathologist and their identity as DC confirmed by morphometry. The presence of DCs within the tumour tissue was shown in 70% of primary uveal melanomas and 70% of cutaneous melanomas. DCs residing in peripheral tissues were immature, able of antigen incorporation and motile. Upon antigen uptake, appropriate secondary stimulation, and migration to the regional lymph node, they differentiate into mature DC stimulating immune responses. We found however, that although both uveal and cutaneous melanoma residing-DC lacked expression of maturation markers (CD40, CD83), DCs residing in uveal melanoma had a more mature morphology. In one case of uveal melanoma, DCs at a final stage of maturation were observed, at the site of tumour
necrosis. This indicates a profound immunosuppression of a lymph node with the deposit of melanoma. In both types of tumour, the distribution of DC in tissue suggested that they actively concentrated in hotspots, possibly associated with the release of chemokines by tumour cells. The observed DCs expressed HLA-DR, and were therefore able to present antigen, but did not express the co-stimulatory molecule, CD40, necessary for the correct dialogue with T lymphocytes. It is conceivable that antigen presentation by such
immature DC, lacking immune-activating abilities, will suppress antigen-specific immune responses, anergize T lymphocytes and promote tumour progression.
128
Nevertheless, the presence of CD83-positive DC in the vicinity of necrotic uveal
melanoma suggests, that the reversal of the immunosuppressed phenotype is possible, if the local microenvironment delivers appropriate stimulatory signals.
To further characterize the exact character of the melanoma-associated local immunosuppressive microenvironment, we examined the presence of suppressor T lymphocytes and tolerizing DCs, the expression of immunosuppressive cytokines (IL-10, TGFβ1 and TGFβ2) and the enzyme indoleamine 2,3-dioxygenase (IDO) using qRT–
PCR and immunohistochemistry in paired samples of primary skin melanomas, negative and positive sentinel lymph nodes (SLN), and a separate group of lymph nodes with advanced melanoma metastases. Our investigation showed, that cutaneous melanoma indeed creates a highly immunosuppressive microenvironment, with regard to both the production of immunosuppressive cytokines and enzymes and to the presence of
immunomodulatory cells. We confirmed an earlier finding by Reed (Reed 1994) that the expression of TGFβ2, but not TGFβ1 was correlated with the presence of melanoma cells, and we showed that the expression of all other immunosuppressive factors
increased with the progression of melanoma, with a peak concentration in positive SLN.
Both tolerogenic DC and immunosuppressive T lymphocytes were present at all stages of melanoma progression, and ubiquitously expressed TGFβR1. Most importantly, however, we observed, that the negative SLN were profoundly immunosuppressed, in comparison with both the primary tumour site, as well as the positive SLN, prior to any signs of melanoma dissemination. We hypothesized that DC tolerized by TGFβ2 produced by primary melanoma migrate to SLN, secret IDO and TGFβ1, and thereby effectively create an immuno-privileged site and thus pre-condition SLN for subsequent melanoma spread. Interestingly, despite the reported abundance of TGFβ in the aqueous humour and in uveal melanoma [7, 8] and the ubiquitous expression of TGFβR1 by DC we did not see any production of IDO by DC in uveal melanoma. We therefore conclude that the
microenvironment in uveal melanoma prevents maturation of DC, but does not render them tolerogenic. It suggests that either molecular mechanisms of the DC response to melanoma-derived immunosuppression differ for DCs residing in uveal or cutaneous
129
melanoma, or mechanisms of immunosuppression differ for these two tumours despite the similarities in their cytokine spectrum.
Despite being very potent immune system regulators, DCs are infrequent in the blood and their numbers are even lower in peripheral tissues. However much desirable, functional studies on DC isolated from tumour tissue are practically unachievable, due to
insufficient numbers of target cells after isolation and difficulties with purification of a population of DCs completely devoid of tumour cells. To investigate the possible reversal of DC-created immunosuppression, it was essential to create a model, using peripheral tissue DCs from an immunosuppressed site. Palatine tonsils obtained from patients with recurrent tonsillitis have long been a source of human lymphoid tissue. They are
particularly interesting because they are constitutively immunosuppressed as part of the mucosal-associated lymphoid tissue (MALT), but are still able to evoke a potent immune response to pathogens [9, 10]. A number of DC subsets have been identified in tonsil.
Like all DCs, they express HLA-DR in the absence of the lineage-associated markers CD3, CD19, CD14 & CD16 but are distinguishable by the intensity of HLA-DR and the expression of CD11C & CD123 [11, 12]. Our model cell culture comprised of
plasmacytoid DC, of myeloid DC and infrequently of tissue macrophages, which we believe creates an adequate representation of peripheral tissue antigen-presenting cells.
We found, that tonsil DC were considerably immunosuppressed when compared with Monocyte-Derived DC (MoDC), commonly used as a model population for DC studies.
On tonsil DCs, expression of dendritic cell activation markers (HLA Class I, HLA-DR, CD40, CD80, CD86) was weak in comparison with laboratory-generated DC, even tonsil DCs: only very few tonsil CD11cpositive/HLA DRhigh cells expressed CD83. In our study, we examined the effect of pathogenic material of bacterial origin (LPS) and recombinant nucleotide sequence (poly: IC) simulating viral infection as well as the influence of IFNy and CD40L binding on the activation and maturation process of highly enriched mixed population of DC isolated from tonsils. Despite the fact that the adjuvants that we used were reported to be very potent activators of MoDC we found that none of them, or the most powerful combination, resulted in the high levels of antigen presenting and co- stimulatory molecules seen using monocyte derived DC. Tonsil DC however underwent
130
activation in pure culture media, in comparison to the original, immunosuppressed phenotype. Activation of tonsil DC by immunoadjuvants, especially IFNγ, resulted in production of measurable quantities of IL12p70 which is the main cytokine secreted by activated myeloid DC. Surprisingly, Poly: IC inhibited the maturation and activation of tonsil DC, both in terms of IL12p70 secretion and surface antigen expression. This effect was observed when Poly: IC was used alone, or in combination with other adjuvants.
The greater part of today’s understanding of the biology of dendritic cells was acquired in studies of monocyte or blood precursors derived dendritic cells, generated and matured in vitro. We found that the discrepancies between these two populations were very
profound, and that their responses to stimulation differed significantly. This finding is of great importance, as it shows, that conclusions cannot be based only on the study of blood-derived laboratory-generated dendritic cells.
As a potentially immunogenic tumour, melanoma need to develop strategies to escape the immune response, and the research described in this thesis certainly suggests that the creation of a local immunosuppressive microenvironment is a critical step in melanoma progression.
As the key function of peripheral DCs is discrimination between harmful and safe antigens, we postulate a mechanism for melanoma immune evasion, which includes modulation of antigen presenting and co-stimulatory skills of peripheral and sentinel lymph node DC, inhibiting their maturation or turning them into a tolerogenic subtype through secretion of immunosuppressive cytokines.
Antigen presentation by immature or tolerogenic DC leads to the anergy, inactivation or suppression of tumour-sensitive lymphocytes. This may greatly support inhibition of anti- tumour immune responses in situ, but upon melanoma cell migration via the lymphatic pathway, it would also affect the very centre of immunity, and effectively provide a mechanism similar to clonal deletion of tumour-specific lymphocytes. Under these circumstances, the main mechanism of immune avoidance in case of melanoma would represent attack, not escape. Our hypothesis also explains why the presence of tumour-
131
infiltrating antigen-presenting cells and expression of HLA-DR are poor prognostic factors in cutaneous and uveal melanoma [13-15].
Although some mechanisms for the interaction between melanoma cells and the immune system have been discovered, including alteration of dendritic cell function, maturation by tumour-derived cytokines, and concentration of numerous suppressive and regulatory T lymphocytes, the exact molecular pathways need to be investigated. A deeper
understanding of the relationship between melanoma and the immune system is necessary to enable successful immunotherapy of melanoma. Additionally, further investigations of local and systemic immune suppression would allow the discrimination of immune system-related key events in melanoma progression, and may lead to the discovery of convenient therapeutic targets per se.
We conclude that the adjuvant-based activation of peripheral DCs is challenging, but possible, if the immunosuppressive microenvironment can be neutralised. Discovery of an adjuvant able to stimulate DCs in melanoma will probably require an extensive research program, and a profound understanding of molecular mechanisms underlying immunosuppression. Thus, however difficult, extended studies on the biology of intratumoral DC may considerably influence the management of cancer since when activated successfully, they may significantly support the treatment of this devastating malignancy.
References
1. F Arienti, J Sule-Suso, F Belli, L Mascheroni, L Rivoltini, C Melani, M Maio, N Cascinelli, MP Colombo, G Parmiani: Limited antitumor T cell response in melanoma patients vaccinated with interleukin-2 gene-transduced allogeneic melanoma cells. Hum Gene Ther 1996, 7:1955-63.
2. JF Baurain, D Colau, N van Baren, C Landry, V Martelange, M Vikkula, T Boon, PG Coulie: High frequency of autologous anti-melanoma CTL directed against an antigen generated by a point mutation in a new helicase gene. J Immunol 2000, 164:6057-66.
3. A Soruri, A Fayyazi, R Gieseler, T Schlott, TM Runger, C Neumann, JH Peters:
Specific autologous anti-melanoma T cell response in vitro using monocyte- derived dendritic cells. Immunobiology 1998, 198:527-38.
132
4. KH Lee, MC Panelli, CJ Kim, AI Riker, MP Bettinotti, MM Roden, P Fetsch, A Abati, SA Rosenberg, FM Marincola: Functional dissociation between local and systemic immune response during anti-melanoma peptide vaccination. J Immunol 1998, 161:4183-94.
5. V Appay, C Jandus, V Voelter, S Reynard, SE Coupland, D Rimoldi, D Lienard, P Guillaume, AM Krieg, JC Cerottini, P Romero, S Leyvraz, N Rufer, DE Speiser: New generation vaccine induces effective melanoma-specific CD8+ T cells in the circulation but not in the tumor site. J Immunol 2006, 177:1670-8.
6. R Cerio, CE Griffiths, KD Cooper, BJ Nickoloff, JT Headington: Characterization of factor XIIIa positive dermal dendritic cells in normal and inflamed skin. Br J Dermatol 1989, 121:421-31.
7. P Esser, S Grisanti, K Bartz-Schmidt: TGF-beta in uveal melanoma. Microsc Res Tech 2001, 52:396-400.
8. P Maier, A Broszinski, U Heizmann, D Boehringer, T Reinhard: Determination of active TGF-beta2 in aqueous humor prior to and following cryopreservation. Mol Vis 2006, 12:1477-82.
9. JM Bernstein, GA Rich, C Odziemiec, M Ballow: Are thymus-derived lymphocytes (T cells) defective in the nasopharyngeal and palatine tonsils of children? Otolaryngol Head Neck Surg 1993, 109:693-700.
10. C Macaubas, RH DeKruyff, DT Umetsu: Respiratory tolerance in the protection against asthma. Curr Drug Targets Inflamm Allergy 2003, 2:175-86.
11. KL Summers, BD Hock, JL McKenzie, DN Hart: Phenotypic characterization of five dendritic cell subsets in human tonsils. Am J Pathol 2001, 159:285-95.
12. G Stent, JC Reece, DC Baylis, K Ivinson, G Paukovics, M Thomson, PU Cameron: Heterogeneity of freshly isolated human tonsil dendritic cells
demonstrated by intracellular markers, phagocytosis, and membrane dye transfer.
Cytometry 2002, 48:167-76.
13. C Ericsson, S Seregard, A Bartolazzi, E Levitskaya, S Ferrone, R Kiessling, O Larsson: Association of HLA class I and class II antigen expression and mortality in uveal melanoma. Invest Ophthalmol Vis Sci 2001, 42:2153-6.
14. T Makitie, P Summanen, A Tarkkanen, T Kivela: Tumor-infiltrating macrophages (CD68(+) cells) and prognosis in malignant uveal melanoma. Invest Ophthalmol Vis Sci 2001, 42:1414-21.
15. S Moretti, C Pinzi, E Berti, A Spallanzani, A Chiarugi, V Boddi, UM Reali, B Giannotti: In situ expression of transforming growth factor beta is associated with melanoma progression and correlates with Ki67, HLA-DR and beta 3 integrin expression. Melanoma Res 1997, 7:313-21.
133
