Immunobiology of Cutaneous T Cell Lymphoma

Braverman, Berger*, Cresswell, Edelson, Girardi*, Hanlon*, Heald*, Imaeda*, Kacinski, Pamer, Rothstein*, Schechner*

For the past twelve years Richard Edelson, Chairman of the Department of Dermatology, has coordinated a broad based research program directed at the elucidation of the basic and clinically relevant properties of cutaneous T cell lymphoma (CTCL) cells. This program now has several interrelated components, integrating the contributions of investigators both inside and outside of Dermatology. The uniquely large referred CTCL patient base at Yale has facilitated a substantial number of clinical trials aimed at finding better treatments both for patients with "early" as well as with advanced disease. These studies have involved very productive, long-standing collaborative associations among Drs. Edelson, Irwin Braverman, and Peter Heald (Dermatology) with Dr. Barry Kacinski (Therapeutic Radiology).

Dr. Irwin Braverman has had a long and productive relationship with Dr. Barry Kacinscki (Therapeutic Radiology) which has contributed substantially to our understanding of the efficacy of total skin electron beam therapy (ESBT) in CTCL. They have previously demonstrated that ESBT in conjunction with modest intensity chemotherapy prolongs the survival of patients with stage 2 CTCL. In collaboration with Drs. Edelson and Heald (Dermatology), Drs. Braverman and Kacinski compared survival of CTCL patients who received photopheresis in conjunction with remission inducing total skin electron beam radiotherapy with those who received electron beam alone. Patients with tumors or erythroderma who received adjuvant photopheresis had better overall survival (100% at five years) than those who received electron beam alone (50%), suggesting that an adjuvant nontoxic regimen of extracorporeal photochemotherapy may prolong survival in advanced stage CTCL patients. These results are the basis for a recently initiated prospective, randomized trial studying the impact of photopheresis in conjunction with electron beam radiotherapy, compared to electron beam followed by skin-directed PUVA treatment.

The large CTCL patient base at Yale continues to provide a highly advantageous source of clonal T cell populations for a substantial number of distinct, but interrelated basic studies aimed ultimately at finding alternative, more efficacious treatments for these patients. One major collaborative focus of several YSDRCC investigators (Drs Berger, Cresswell, Edelson, Hanlon, Heald, Imaeda, and Pamer) involves characterizing tumor specific class I major histocompatibility complex (MHC) peptides distinctive to individual clones of malignant T cells, with the goal of determining whether these molecules can serve as clinically relevant immunogens.

Drs. Richard Edelson and Carole Berger have collaborated closely on research focused on the investigation of the biologic features of CTCL cells for over twenty years. They have recently made substantial progress in their efforts to identify tumor specific antigens on the malignant cells of CTCL patients. Using a technique which they had developed and perfected, populations of malignant cells were obtained by cell sorting and incubated with family-specific Vb T cell receptor antibodies. CD8 T cell lines, reacting selectively with the tumor cells, were propagated by incubation with the CTCL cells of the same patient in the presence of interleukin-2. CD8 lines were subcloned, to obtain homogeneous populations of anti-CTCL cytotoxic lymphocytes. Their findings revealed that CTCL patients have CD8 populations which recognize tumor specific peptides of 8-11 amino acid length, which reside in the surface clefts of class I molecules. This is the first such demonstration of tumor-specific antigens on CTCL cells. They then used the CD8 lines to further characterize the composition of the tumor-specific peptides. In two distinct ways, they were able to demonstrate that these peptides originate, at least in large measure, from distinctive components of the T cell receptor chains of the malignant cells. In collaboration with Suguru Imaeda, Peter Heald, Jack Longley (formerly here at Yale and now at Columbia University), and Peter Cresswell (Immunobiology and Dermatology), they prepared short amino acid chain segments from the VDJ hypervariable portion of the Vb chain of the patient's T cell receptor. Using appropriate controls, they found that these peptides specifically sensitized autologous B lymphoblasts to attack by the cultured CD8 T cells. Next, they isolated all of the surface class I peptides into individual fractions. They determined which of these fractions sensitized the autologous B lymphocyte blasts to attack by the same CD8 cells and continued to further fractionate until they came up with a single peptide. That peptide was then sequenced by an entirely separate laboratory, revealing that it also came from a peptide of the Vb chain of the same patient's CTCL cells T-cell receptor. Together, these results demonstrated for the first time that CTCL cells express tumor specific peptide antigens which are truly immunogenic. These peptides are derived from the clone-specific T cell receptor, and the results may also explain how normal T cells are regulated through an idiotypic network.

In follow-up collaborative studies involving Dr. Suguru Imaeda (Dermatology), Dr. Eric Pamer (Infectious Diseases) and Dr. Peter Cresswell (Immunobiology and Dermatology) with Drs. Edelson and Berger, all of the surface class I peptides from this same CTCL patient were harvested and separated into various fractions using HPLC. Further experiments then determined which of these fractions sensitized autologous B lymphocyte blasts to attack by the same CD8 cells described immediately above. Ultimately, a single peptide was isolated; subsequent sequencing by Yale's Keck Resource Laboratory revealed that it also came from a peptide of the Vb chain of the same patient's CTCL cells T-cell receptor. Together, these results demonstrated for the first time that CTCL cells express tumor specific peptide antigens which are truly immunogenic. These peptides are derived from the clone-specific T cell receptor, and the results may also explain how normal T cells are regulated through an idiotypic network.

In other studies, Drs. Berger and Edelson have recently shown that CTCL cells express a cell surface 78 kDa heat shock protein, recognized by the BE2-78 monoclonal antibody. The relevance of this molecule for peptide transport, presentation and transfer is being pursued. This molecule differs in its peptide map from known members of the heat shock protein 70 family and cloning and sequencing of the molecule is in progress. These studies are being performed in collaboration with Pramod Srivastava at the University of Connecticut.

Peter Cresswell, Ph.D. (Immunobiology and Dermatology) is an internationally recognized expert in the arena of the mechanisms of MHC-restricted antigen processing. In recent years they have made substantial inroads into understanding the molecular components required for loading peptides into both MHC class I and class II molecules. His laboratory has determined that two related ER chaperones, calnexin and calreticulin, play a role in MHC class I assembly and have defined the "loading complex" in the ER as a multi-subunit assembly containing MHC class 1, b2 microglobulin, calreticulin, the thiolreductase ERp57, the two subunits of the Transporter associated with Antigen Processing (TAP), and a newly-defined product of an MHC-linked gene, tapasin. They have also identified roles for the invariant chain, HLA-DM, and HLA-DO molecules in the loading of MHC class 11 molecules with endocytically generated peptides.

Douglas Hanlon, Ph.D*., completed three years of postdoctoral training under the co-supervision of Dr. Cresswell and Dr. Edelson and very recently accepted (effective 7/1/98) a full time appointment in the Department of Dermatology as an Associate Research Scientist. During the initial portion of his fellowship with Dr. Cresswell, Dr. Hanlon efforts focused on strategies to develop human cell lines capable of displaying MHC class I alleles of choice in an "empty" (peptide-free) configuration at the cell surface of antigen presenting cells. The goal of these studies was to derive cells expressing defined MHC I alleles which could potentially serve as flexible peptide "acceptors" capable of displaying large densities of tumor-specific peptides to CD8+ cytotoxic T-cells (CTL). This project, with Dr. Cresswell as the Principal Investigator, formed the basis for a YSDRCC P/F study funded during the 05 year of this Center; these studies continue to be pursued in Dr. Cresswell's group (involving the continuing collaboration with Dr Hanlon and Dr. Edelson not only because of their utility in further dissecting the complex MHC I antigen processing/presentation pathway, but also because of the potential for such engineered cells to be used extremely sensitive targets to monitor antigen-specific T cell responses in CTCL, melanoma, etc.

Dr. Hanlon's training in Dr. Cresswell's laboratory allowed him to become familiar with the broad range of cell biological and immunochemical methods that can, as Dr. Cresswell's international reputation so clearly shows, be used to elucidate the pathways involved in antigen processing and presentation for both MHCI and MHC II antigens. During the past year, during which he has been working closely with Dr. Edelson, he has utilized these newly acquired skills to try to elucidate the mechanisms involved in the apparently enhanced immunogenicity observed in cells which have been exposed to 8-MOP/UVA. His preliminary studies have not only verified a striking TAP-dependent increase in cell surface MHC I following such pretreatment, but quite unexpectedly, that after UVA exposure, a major proportion of 8-MOP is bound to a strikingly limited number of cellular proteins (rather than exclusively to DNA). These preliminary studies formed the basis for a new P/F project (#25), in which Dr. Hanlon and Dr. Edelson (with Dr. Cresswell as a very able consultant) propose to identify these cellular targets of 8-MOP/UVA and determine the relationship(s) to antigen processing/presentation pathways.

Dr. Peter Heald* has contributed significantly to elucidating the natural history of CTCL and to devising a convenient and reliable method to purify malignant CTCL cells for subsequent studies (e.g., analysis of putative anti-CTCL cell specific immune responses mediated by the nonmalignant T cells in patients who respond to photopheresis). In collaboration with Drs. Berger and Longley, Dr. Heald has been examining the utility of a panel of monoclonal antibodies to a variety of antigens expressed preferentially on the malignant cell population. These include epitopes on particular TCR chains, the CD45RO isoform preferentially expressed by memory T cells, and the cutaneous lymphoid associated antigen (CLA) shown by Picker and his colleagues to be the homing receptor for skin-seeking T cells. In a productive collaborative study involving the YSDRCC with the SDRCC at U.T. Southwestern, a collaboration which was the direct outgrowth of the YSDRCC's enrichment program (seminar series), Drs. Heald, Edelson, Tigelaar, and Picker (UT Southwestern) demonstrated that the malignant cell in CTCL is a CD45RO+ ("memory"-type) CLA+ T cell. Dr. Heald also has demonstrated that many erythrodermic CTCL patients with normal absolute blood lymphocyte counts have an otherwise unsuspected, and frequently very striking, replacement of the normal lymphocyte population by CTCL cells ("nonmalignant lymphopenia"). This observation provides a plausible explanation for why these patients frequently develop HIV-like opportunistic complications. Furthermore, they reveal a remarkable heterogeneity of immunocompetence in CTCL patients, probably explaining the variable responses to the immunotherapeutic regimen photopheresis. In this regard, Dr. Heald, in collaboration with Dr. Edelson, has found that whereas <10% of erythrodermic CTCL patients with very low CD8 T cell counts respond significantly to that therapy, nearly 70% with near normal CD8 counts have a very substantial and often complete response.

Dr. Michael Girardi is the principal investigator in a multidisciplinary collaboration with Drs. Berger and Tigelaar in Dermatology, Dr. Hayday (Biology) and Dr. Janet Brandsma (Comparative Medicine) which is exploring the potential for using DNA encoding (portions of) the T cell receptor expressed on clonally expanded pathogenic T cells (such as in CTCL, and possibly also a variety of autoimmune diseases) in a variety of vaccination schemes aimed at generating cytotoxic and/or regulatory T cell responses against the (oligo)clonal population of pathogenic T cells. This project, supported initially by YSDRCC P/F funds and very recently by a K08 award from NIAMS, utilizes an established murine model of T cell lymphoma where the subcutaneous injection of the 2B4.11 T cell hybridoma cells into syngeneic mice results in tumor growth and decreased survival. Inoculations with TCR-DNA plasmids encoding portions of the a and b chain of the 2B4 TCR are being assayed for their ability to elicit specific anti-tumor (anti-2B4.11) immune responses, as determined by in vitro proliferation and cytotoxicity assays. Inoculated mice are also be tested in vivo for their ability to resist 2B4.11 tumor challenge. Direct comparisons will assay the relevance of inoculation technique (intradermal vs. intramuscular vs. gene gun). Furthermore, several strategies designed to enhance the immunotherapeutic potential of TCR-DNA plasmid immunization will be examined, including: (1) co-inoculation with immunostimulatory oligonucleotides, (2) co- or pre-injection with cytokine expression plasmids (3) inoculation with chimeric TCR/cell-trafficking protein plasmids, and (4) inoculation with dendritic cells transformed in vitro with TCR-DNA plasmids.

Dr. Girardi also has ongoing studies, carried out in collaboration with Drs. Hayday and Tigelaar to investigate the immunobiology of squamous cell carcinoma (SCC). That the immune system plays a role in tumor surveillance against this very common skin cancer squamous cell carcinoma (SCC) is supported by several data obtained from clinical observation of immunosuppressed transplantation-recipient patients, the cellular infiltration observed in sections of SCC, and experimental murine models of SCC. Dr. Girardi' studies are analyzing the relative roles of T cell subsets in the oncogenesis, tumor surveillance, and tumor rejection of SCC by utilizing the crossbreeding of genetically engineered mice. Specifically, there are three experimental murine models being studied: (1) injection of the PDV squamous cell carcinoma cell line into TCR knockout mice, (2) the crossbreeding of TCR knockouts onto the chemically-induced skin tumor-susceptible FVB background, and (3) the crossbreeding of skin-specific ras transgenic mice with TCR knockout mice.

Dr. Suguru Imaeda* has three major areas of research focus: 1) T cell lymphoma immunotherapy; 2) T cell receptor-derived major histocompatibility complex class I (MHC class I) epitopes in CTCL; and 3) Immunomodulation of lymphoma using photoactivated psoralen. Several models are under active investigation, including use of D10.G4.1, a "normal" cloned CD4 Th2 T cell responsive to conalbumin presented in the context of I-Ak. This project was begun as a postdoctoral fellow in the laboratory of Dr. Charles A. Janeway, (Immunobiology). and has been continued after the completion of his fellowship. Using the known major histocompatibility complex class I (MHC I) sequence motif for H-2Kk, "candidate" peptides derived from the TCR protein sequence have been identified, synthesized and used to stimulate cytotoxic T cells (CTLs) generated against D10.G4.1. MHC I peptides have been eluted from the D10.G4.1 cells, separated by reversed phase high performance liquid chromatography (RP-HPLC) and are being used to stimulate the anti-D10.G4.1 CTLs to determine whether the synthetic peptides correspond to the natural epitopes. A second model, investigated in collaboration with Dr. Carole Berger, uses a similar approach with the 2B4.11, B10.AxAKR murine hybridoma. Since 2B4.11 is a rapidly growing lethal tumor, these studies will hopefully determine whether TCR-derived MHC I epitopes can immunoprotect and/or abort tumor progression. The third system, the investigation of CTCL in human subjects, involves collaboration with Drs. Berger, Edelson, and Longley, involves the identification and immunization with TCR-derived peptides in CTCL patients. Peptides acid eluted from the malignant T cell clone are separated by RP-HPLC for use in CTL stimulation assays. The determination of potential "candidate" peptides using the known HLA sequence motifs has been performed. All of these studies have involved the collaborative efforts of Dr. Eric Pamer (Infectious Disease), whose studies of CTL responses to Listeria-infected macrophages were among the first to utilize peptide elution/RP-HPLC separation for purification of peptides recognized by CD8 cells.

Dr. Imaeda's studies of the effects of photoactivated psoralen on MHC I expression and antigenicity have involved collaboration with Dr. Frank Gasparro, a former YSDRCC member , who is now in the Department of Dermatology at Thomas Jefferson University. These studies have shown that the murine T cell tumor RMA when exposed to 8-MOP/UVA increases its surface expression of MHC I in a dose dependent fashion. In addition, a large number of "empty" fillable MHC I molecules can be generated under the proper thermodynamic conditions. These findings suggest that photoactivated psoralen can render a tumor cell more immunogenic by increasing normally underexpressed MHC I epitopes and by forming "empty" MHC I molecules capable of exogenous peptide loading. The current goal is to determine whether photoactivated psoralen can indeed make tumor cells: 1) more immunogenic, 2) more susceptible to CTL lysis, and 3) capable of presenting exogenously loaded antigen. In addition, photoactivated psoralen may make dendritic cells and other professional antigen presenting cells more capable of enhancing anti-tumor responses when present in conjunction with tumor cells during phototherapy.

Dr. David Rothstein* (Nephrology) has been collaborating with Drs. Halaban, Edelson and Berger, in addressing questions of potential importance to understanding CTCL cell proliferation. Observations originally made by Berger, Edelson and Heald at Yale and by Nickoloff at the University of Michigan indicated that CTCL cells may be driven to proliferate by stimuli endogenous to skin. Whereas a high percentage of those CTCL cells localized to or near the epidermis express activation markers, their removal from the body and brief incubation in vitro causes them to lose these markers and apparently enter a more quiescent state. These findings suggested that CTCL cells, at least early in the malignant process, receive and can process activation signals from the microenvironment of the skin. Since they express memory T cell markers, Dr. Rothstein, in collaboration with the mentioned Dermatology colleagues, has focused some of his skin-related research efforts on the roles of CD45 and its isoforms in CTCL cell activation signaling and is defining the role of individual CD45 isoforms in the regulation of early tyrosine phosphorylation events in both normal and malignant human T cells. In studies that are being extended, he and Drs. Halaban and Edelson have determined that CTCL cells, which primarily express the CD45RO isoform, manifest significant but poorly understood signaling defects. In particular, they express high levels of Csk but decreased overall Csk activity. Interestingly, Csk has been reported to be a negative regulator of Src-kinases and a positive regulator of CD45. How Csk activity itself is regulated is unclear. A better understanding of the signals involving these molecules in human T cells will contribute to their understanding of the defects present in these CTCL cells. Following a recent report suggesting that the CD45RO isoform may be involved in apoptosis and in adhesion, very recent pilot studies in collaboration with Dr. Carole Berger have suggested that antibodies against this isoform selectively increases Bc1-2 levels. Further exploration of the role of CD45 in a regulation of Fas-ligand are being performed on a collaboration with Dr. Berger and with Dr. Laurie Owen-Schaub ( MD Anderson).

Dr. Jeffrey Schechner* completed his dermatology residency at Yale in 1994. He then undertook three years of postdoctoral fellowship training in the laboratory of Dr. Jordan Pober (Pathology and Dermatology), with Dr. Edelson as his secondary advisor. Following completion of his fellowship he was recruited to the Dermatology Department, where he is currently an assistant Professor. His research has focused on the expression of adhesion molecules on lymphocytes, endothelial cells and epithelial cells, and what role they may play in the infiltration of the skin by lymphocytes in benign inflammatory dermatoses and CTCL. For the past four years he has been extensively utilizing chimeric human-immunodeficient mouse models in which to study the interactions between human skin or skin derived cells and human leukocytes (This expertise has resulted in his being identified as Co-director of the YSDRCC SCID Mouse: Skin Xenograft Core, a new P/F study). A collaborative study, involving Drs. Pober, Heald, and Madison directed toward identifying endothelial cell and lymphocyte adhesion molecules that participate in infiltration of the skin by CTCL cells and making extensive use of the SCID mouse model, was submitted as a P/F project. Preliminary studies have revealed several immunohistologic characteristics of CTCL in allogeneic skin grafts implanted on SCID-beige mice who received i.p. injections of peripheral blood mononuclear cells (PBMCs) obtained from affected patients, and the malignant clone of T cells has been detected both in the mouse circulation and in the human skin grafts. Further development of a validated animal model of CTCL will allow evaluation of the mechanism of skin specific homing of malignant cells which may lead to new therapeutic strategies.