Helper-independent, L-selectinlow CD8+ T cells with broad anti-tumor efficacy are naturally sensitized during tumor progression

We recently reported that the CD4(+) T cell subset with low L-selectin expression (CD62L(low)) in tumor-draining lymph nodes (TDLN) can be culture activated and adoptively transferred to eradicate established pulmonary and intracranial tumors in syngeneic mice, even without coadministration of IL-2. We have extended these studies to characterize the small subset of L-selectin(low) CD8(+) T cells naturally present in TDLN of mice bearing weakly immunogenic tumors. Isolated L-selectin(low) CD8(+) T cells displayed the functional phenotype of helper-independent T cells, and when adoptively transferred could consistently eradicate, like L-selectin(low) CD4(+) T cells, both established pulmonary and intracranial tumors without coadministration of exogenous IL-2. Whereas adoptively transferred L-selectin(low) CD4(+) T cells were more potent on a cell number basis for eradicating 3-day intracranial and s.c. tumors, L-selectin(low) CD8(+) T cells were more potent against advanced (10-day) pulmonary metastases. Although the presence of CD4(+) T cells enhanced generation of L-selectin(low) CD8(+) effector T cells, the latter could also be obtained from CD4 knockout mice or normal mice in vivo depleted of CD4(+) T cells before tumor sensitization. Culture-activated L-selectin(low) CD8(+) T cells did not lyse relevant tumor targets in vitro, but secreted IFN-gamma and GM-CSF when specifically stimulated with relevant tumor preparations. These data indicate that even without specific vaccine maneuvers, progressive tumor growth leads to independent sensitization of both CD4(+) and CD8(+) anti-tumor T cells in TDLN, phenotypically L-selectin(low) at the time of harvest, each of which requires only culture activation to unmask highly potent stand-alone effector function.     

The shedding of CD62L (L-selectin) regulates the acquisition of lytic activity in human tumor reactive T lymphocytes

CD62L/L-selectin is a marker found on naïve T cells and further distinguishes central memory (Tcm, CD62L+) from effector memory (Tem, CD62L−) T cells. The regulation of CD62L plays a pivotal role in controlling the traffic of T lymphocytes to and from peripheral lymph nodes. CD62L is shed from the cell membrane following T cell activation, however, the physiological significance of this event remains to be elucidated. In this study, we utilized in vitro generated anti-tumor antigen T cells and melanoma lines as a model to evaluate the dynamics of CD62L shedding and expression of CD107a as a marker of lytic activity. Upon encounter, with matched tumor lines, antigen reactive T cells rapidly lose CD62L expression and this was associated with the acquisition of CD107a. By CD62L ELISA, we confirmed that this transition was mediated by the shedding of CD62L when T cells encountered specific tumor antigen. The introduction of a shedding resistant mutant of CD62L into the tumor antigen-reactive T cell line JKF6 impaired CD107a acquisition following antigen recognition and this was correlated with decreased lytic activity as measured by ⁵¹Cr release assays. The linkage of the shedding of CD62L from the surface of anti-tumor T cells and acquisition of lytic activity, suggests a new function for CD62L in T cell effector functions and anti-tumor activity.     

Adoptive immunotherapy of advanced tumors with CD62 L-selectin(low) tumor-sensitized T lymphocytes following ex vivo hyperexpansion

Tumor-draining lymph nodes (TDLN) contain sensitized T cells with the phenotype CD62 L-selectin(low) (CD62L(low)) that can be activated ex vivo with anti-CD3 mAb and IL-2 to acquire potent dose-dependent effector function manifested upon adoptive transfer to secondary tumor-bearing hosts. In this study advanced tumor models were used as a stringent comparison of efficacy for the CD62L(low) subset, comprising 5-7% of the TDLN cells, vs the total population of TDLN cells following culture in high dose IL-2 (100 U/ml). During the 9-day activation period the total number of CD8+ T cells increased 1500-fold, with equivalent proliferation in the CD62L(low) vs the total TDLN cell cultures. Adoptive transfer of activated CD62L(low) cells eliminated 14-day pulmonary metastases and cured 10-day s.c. tumors, whereas transfer of maximally tolerated numbers of total TDLN cells was not therapeutic. Despite their propagation in a high concentration of IL-2, the hyperexpanded CD62L(low) subset of TDLN cells functioned in vivo without exogenous IL-2, and CD8+ T cells demonstrated relative helper independence. Moreover, the anti-tumor response was specific for the sensitizing tumor, and long term memory was established. The facile enrichment of tumor-reactive TDLN T cells, based on the CD62L(low) phenotype, circumvents the need for prior knowledge of the relevant tumor Ags. Coupling the isolation of pre-effector T cells with rapid ex vivo expansion to >3 logs could overcome some of the shortcomings of active immunotherapy or in vivo cytokine treatment, where selective robust expansion of effector cells has been difficult to achieve.     

Role of C chemokine lymphotactin in mediating recruitment of antigen-specific CD62L(lo) cells in vitro and in vivo.

In this study we investigated whether T cells expressing high or low levels of CD62L were differentially susceptible to the T cell chemokine lymphotactin. We found that lymphotactin induced preferential migration of antigen-specific (CD62L(lo)) T cells over the nonspecific (CD62L(hi)) T cells in vitro and in vivo. The differing migratory abilities correlated with higher levels of mRNA encoding the lymphotactin receptor (XCR1) on the CD62L(lo) cells compared to the CD62L(hi) cells. Thus, we have identified a coupling mechanism between the activation of T cells and acquisition of new homing properties, in this case conferred by XCR1 expression. These data confirm that at least one function of lymphotactin includes mediating the recruitment of recently activated antigen-specific T cells.     

Reduced L-selectin (CD62LLow) expression identifies tumor-specific type 1 T cells from lymph nodes draining an autologous tumor cell vaccine

Reduced expression of CD62L can identify tumor-specific T cells in lymph nodes draining murine tumors. Here, we examined whether this strategy could isolate tumor-specific T cells from vaccinated patients. Tumor vaccine-draining lymph node (TVDLN) T cells of seven patients were separated into populations with reduced (CD62LLow) or high levels of CD62L (CD62LHigh). Effector T cells generated from CD62LLow cells maintained or enriched the autologous tumor-specific type 1 cytokine response compared to unseparated TVDLN T cells in four of four patients showing tumor-specific cytokine secretion. Interestingly, effector T cells generated from CD62LLow or CD62LHigh TVDLN were polarized towards a dominant type 1 or type 2 cytokine profile, respectively. For CD62LLow T cells the type 1 cytokine profile appeared determined prior to culture. Since a tumor-specific type 1 cytokine profile appears critical for mediating anti-tumor activity in vivo, this approach might be used to isolate T cells for adoptive immunotherapy.     

Prevention of lymph node metastases by adoptive transfer of CD4+ T lymphocytes admixed with irradiated tumor cells

CD4⁺8^– T lymphocytes with potent antitumor activity in vivo were obtained in peritoneal exudate cells by immunizing mice with irradiated MM48 tumor cells admixed with OK-432. These immune CD4⁺ T cells were used in adoptive immunotherapy for prevention of lymph node metastases after removal of the primary tumor. Complete cure of metastases was obtained by adoptive transfer of CD4⁺ T cells admixed with irradiated MM48 tumor cells, but not by CD4⁺ T cells alone. To analyze the curative effect of admixing tumor cells on the prevention of metastases, a model of 1-day tumor inoculated with macrophages was used. Administration of immune CD4⁺ T cells alone resulted in the regression of local tumor in more than half of the mice, although all of them eventually died of lymph node metastases. On the other hand, adoptive transfer of immune CD4⁺ T cells plus irradiated tumor cells resulted in the complete regression of local tumors in all the mice, which survived without any sign of metastasis. The curative effect of the immune CD4⁺ T cells obtained by admixing irradiated tumor cells was tumor-specific. Macrophages induced by OK-432 (tumoricidal), implanted together with tumor, assisted tumor regression more than did macrophages elicited by proteose peptone (nontumoricidal) in the same adoptive transfer system. Administration of recombinant interleukin-2 instead of stimulant tumor cells did not enhance, but rather eliminated the constitutive antitumor activity of CD4⁺ T cells. On the other hand, exogenous recombinant interleukin-1 was more effective in the enhancement of antitumor activity of the CD4⁺ T cells as compared with stimulant tumor cell administration. In this case, the activating states of macrophages at the implanted tumor site had no influence on the therapeutic efficacy. A possible role of macrophages for induction of tumor-specific cytotoxic T cells that were mediated by tumor-specific CD4⁺ T cells is discussed.     

Enhancement of anti-tumor immunity specific to murine glioma by vaccination with tumor cell lysate-pulsed dendritic cells engineered to produce interleukin-12

Aim: The aim of this study was to develop an immunotherapy specific to a malignant glioma by examining the efficacy of glioma tumor-specific cytotoxic T lymphocytes (CTL) as well as the anti-tumor immunity by vaccination with dendritic cells (DC) engineered to express murine IL-12 using adenovirus-mediated gene transfer and pulsed with a GL26 glioma cell lysate (AdVIL-12/DC+GL26) was investigated. Experimentl: For measuring CTL activity, splenocytes were harvested from the mice immunized with AdVIL-12/DC+GL26 and restimulated with syngeneic GL26 for 7 days. The frequencies of antigen-specific cytokine-secreting T cell were determined with mIFN-γ ELISPOT. The cytotoxicity of CTL was assessed in a standard 51Cr-release assay. For the protective study in the subcutaneous tumor model, the mice were vaccinated subcutaneously (s.c) with 1×10⁶ AdVIL-12/DC+GL26 in the right flanks on day −21, −14 and −7. On day 7, the mice were challenged with 1×10⁶ GL26 tumor cells in the shaved left flank. For a protective study in the intracranial tumor model, the mice were vaccinated with 1×10⁶ AdVIL-12/DC+GL26 s.c in the right flanks on days −21, −14 and −7. Fresh 1×10⁴ GL26 cells were inoculated into the brain on day 0. To prove a therapeutic benefit in established tumors, subcutaneous or intracranial GL26 tumor-bearing mice were vaccinated s.c with 1×10⁶ AdVIL-12/DC+GL26 on day 5, 12 and 19 after tumor cell inoculation. Results: Splenocytes from the mice vaccinated with the AdVIL-12/DC+GL26 showed enhanced induction of tumor-specific CTL and increased numbers of IFN-γ: secreting T cells by ELISPOT. Moreover, vaccination of AdVIL-12/DC+GL26 enhanced the induction of anti-tumor immunity in both the subcutaneous and intracranial tumor models. Conclusions: These preclinical model results suggest that DC engineered to express IL-12 and pulsed with a tumor lysate could be used in a possible immunotherapeutic strategy for malignant glioma.Korea Research Foundation Grant (KRF-2004-005-E00001).     

Antitumor activity exhibited by Fas ligand (CD95L) overexpressed on lymphoid cells against Fas+ tumor cells

Lymph node (LN) cells of Fas-mutant mice lpr/lpr (lpr) and lpr ^cg /lpr ^cg (lpr ^cg ) express an increased level of Fas ligand (FasL) (CD95L). We examined the antitumor potential of cell-bound FasL on these LN cells against Fas⁺ tumor cells. Fas⁺ F6b and Fas⁻ N1d cells were produced from murine hepatoma MH134 (Fas⁻) by gene transfection. lpr and lpr ^cg LN cells inhibited growth of F6b but not N1d cells in vitro. Neither gld/gld lpr/lpr (gld/lpr) LN cells, which lack both FasL and Fas, nor wild-type LN cells showed growth-inhibitory activities against F6b and N1d cells. The effector cells and molecule were CD4⁻CD8⁻ T cells and FasL, respectively. The tumor neutralization test and adoptive transfer demonstrated that lpr and lpr ^cg , but not gld/lpr, LN cells retarded the growth of F6b cells. Although anti-Fas antibody and FasL cause severe liver failure, wild-type mice injected with lpr LN cells appeared clinically normal. Adoptive transfer of lpr LN cells to F6b-bearing mice exerted the same antitumor activity in wild-type and gld/lpr recipient mice, indicating the applicability of cell-bound FasL for Fas-mediated target therapy of cancer. These results suggest that antitumor activity was dependent on the Fas-FasL system and that lymphoid cells overexpressing FasL can be powerful antitumor effector cells against Fas⁺ tumor cells. Received: 16 March 1998 / Accepted: 28 July 1998     

Interaction of rabbit lymph node cells with bacteria (Salmonella paratyphi B)

In five rabbits immunized withSalmonella paratyphi B antigen in all four paws, incubation of the regional lymph node cells with the same bacteria was followed by a significant shift of some of the bacteria from the supernatant of the culture medium to the sediment (as compared with the controls). In six rabbits, bactericidal activity was demonstrated in extracts prepared from the regional lymph nodes only 30 hours after immunization. No antibodies were present in extracts from control rabbits up to the fifth day after immunization.     

Genetic association between low expression phenotype of CD62L (L-selectin) in peripheral CD4+ T cells and the thid (T-helper immunodeficiency) phenotype in the LEC rat.

Genetic linkage analysis was performed between the low expression phenotype of peripheral CD4+ T cells and the thid (T-helper immunodeficiency) phenotype using (BN x LEC)F1 x LEC backcross progenies. In contrast to a previous result using a thid congenic strain that the low expression phenotype of CD62L was not correlated with the thid phenotype, our result in this study indicated that the low expression phenotype of CD62L was genetically linked with the thid phenotype. The discrepancy between the previous and present results may be due to the source of animals, congenic strain versus backcross progenies. It is suggested by this study that the thid locus controls the expression level of CD62L in peripheral CD4+ T cells.     

Spatial distribution of L-selectin (CD62L) on human lymphocytes and transfected murine L1-2 cells

Summary We have examined the topographical distribution of L-selectin on surface membrane domains of human lymphocytes and murine L1-2 cells transfected to express human L-selectin. L-selectin was immunolocalized using murine monoclonal DREG 200 Fab antibody and a 12 nm colloidal gold-conjugated secondary antibody. Cell surface morphology and surface distribution of gold-labelled L-selectin were visualized using backscatter electron images obtained by high-resolution, field emission scanning electron microscopy. The topographical morphologies of lymphocytes of both types were complex. The surface of human lymphocytes was composed of both microvilli and ruffles; that of the murine cells was composed of long microvilli and few, if any, ruffles. L-selectin on human lymphocytes was observed primarily as focal clusters on the apical surfaces of ruffles and microvilli. Similarly, on the transfected murine cells, L-selectin was detected predominantly on the apical surface of microvilli. We conclude that L-selectin has a common spatial distribution and clustered organization on all leukocytes examined to-date, and that these features of receptor expression likely facilitate rolling of circulating leukocytes on the endothelial surface.     

Selective killing of CD8+ cells with a 'memory' phenotype (CD62Llo) by the N-acetyl-D-galactosamine-specific lectin from Viscum album L

As reported previously by our group, among the toxic proteins from Viscum album L. only the mistletoe lectins (MLs) induce the apoptotic killing pathway in human lymphocytes. Although one may expect a homogenous distribution of carbohydrate domains on cell surface receptors for the carbohydrate binding B chains of the toxic protein, the sensitivity of cells to these B chains obviously differ. Here we report a selective killing of CD8+ CD62Llo cells from healthy individuals by the galNAc-specific ML III (and RCA60, which binds to gal and galNAc), while the gal-specific ML I was less effective. This selective killing is not sufficiently explained by protein synthesis inhibition alone, since this subset was not affected by other ribosome inhibiting proteins such as the lectin from Ricinus communis (RCA120), lectin from Abrus precatorus (APA), abrin A, and inhibitors of RNA, DNA and/or protein synthesis such as actinomycin D, mitomycin C, and cycloheximide. We conclude that CD8+ cells with 'memory' phenotype (CD62Llo) are more sensitive to the ML III-mediated killing than their CD8+ CD62Lhi counterparts, CD4+ T cells, and CD19+ B cells. These cells probably express a distinct receptor with galNAc domains that is missing or not active on CD8+ cells with a 'naive' phenotype.     

Induction of autoimmunity by expansion of autoreactive CD4+CD62Llow cells in vivo

The prerequisites of peripheral activation of self-specific CD4(+) T cells that determine the development of autoimmunity are incompletely understood. SJL mice immunized with myelin proteolipid protein (PLP) 139-151 developed experimental autoimmune encephalomyelitis (EAE) when pertussis toxin (PT) was injected at the time of immunization but not when injected 6 days later, indicating that PT-induced alterations of the peripheral immune response lead to the development of autoimmunity. Further analysis using IA(s)/PLP(139-151) tetramers revealed that PT did not change effector T cell activation or regulatory T cell numbers but enhanced IFN-gamma production by self-specific CD4(+) T cells. In addition, PT promoted the generation of CD4(+)CD62L(low) effector T cells in vivo. Upon adoptive transfer, these cells were more potent than CD4(+)CD62L(high) cells in inducing autoimmunity in recipient mice. The generation of this population was paralleled by higher expression of the costimulatory molecules CD80, CD86, and B7-DC, but not B7-RP, PD-1, and B7-H1 on CD11c(+)CD4(+) dendritic cells whereas CD11c(+)CD8alpha(+) dendritic cells were not altered. Collectively, these data demonstrate the induction of autoimmunity by specific in vivo expansion of CD4(+)CD62L(low) cells and indicate that CD4(+)CD62L(low) effector T cells and CD11c(+)CD4(+) dendritic cells may be attractive targets for immune interventions to treat autoimmune diseases.     

Ex vivo imaging of T cells in murine lymph node slices with widefield and confocal microscopes

Naïve T cells continuously traffic to secondary lymphoid organs, including peripheral lymph nodes, to detect rare expressed antigens. The migration of T cells into lymph nodes is a complex process which involves both cellular and chemical factors including chemokines. Recently, the use of two-photon microscopy has permitted to track T cells in intact lymph nodes and to derive some quantitative information on their behavior and their interactions with other cells. While there are obvious advantages to an in vivo system, this approach requires a complex and expensive instrumentation and provides limited access to the tissue. To analyze the behavior of T cells within murine lymph nodes, we have developed a slice assay ¹, originally set up by neurobiologists and transposed recently to murine thymus ². In this technique, fluorescently labeled T cells are plated on top of an acutely prepared lymph node slice. In this video-article, the localization and migration of T cells into the tissue are analyzed in real-time with a widefield and a confocal microscope. The technique which complements in vivo two-photon microscopy offers an effective approach to image T cells in their natural environment and to elucidate mechanisms underlying T cell migration.     

Involvement of CD166 in the activation of human gamma delta T cells by tumor cells sensitized with nonpeptide antigens

We previously reported that human Vgamma2Vdelta2-gammadelta T cells were activated by many human tumor cell lines treated with pamidronate (PAM) in a gammadelta TCR-dependent manner. In the present study, we indicated that a synthetic pyrophosphomonoester Ag, 2-methy-3-butenyl-1-pyrophosphate, could directly "sensitize" the tumor cells to activate gammadelta T cells independently of the host metabolism, while the sensitizing effect of PAM was reported to be dependent on the pharmacological activity. Some exceptional tumor cells that failed to be sensitized by PAM were incapable of activating gammadelta T cells by the treatment with 2-methy-3-butenyl-1-pyrophosphate either, suggesting a requirement of host factor(s) for the effective gammadelta T cell activation in addition to the nonpeptide Ags. By screening mAbs against a large panel of tumor cell lines, we found that the expression of CD166 closely paralleled the capacity of activating gammadelta T cells upon PAM treatment. The transfection of a CD166-negative tumor cell line with CD166 cDNA caused a marked enhancement of the capacity to activate gammadelta T cells following PAM treatment. On the contrary, down-regulation of the CD166 expression in a CD166-bearing tumor cell line by short hairpin RNA resulted in a significant reduction of PAM-induced gammadelta T cell-stimulatory activity. gammadelta T cells expressed CD6, a receptor of CD166, and CD6 and CD166 were recruited together to the center of synapse between gammadelta T cells and PAM-treated tumor cells, colocalizing with gammadelta TCR/CD3. The results suggested that the engagement of CD6 with CD166 on tumor cells played an important role in the gammadelta T cell activation by the tumor cells loaded with nonpeptide Ags either endogenously or exogenously.