Normal and cancer fibroblasts differentially regulate TWIST1, TOX and cytokine gene expression in cutaneous T-cell lymphoma

Cutaneous T-cell lymphoma (CTCL) is a heterogeneous group of T cell malignancies that develop from the proliferation and transformation of mature skin-homing T cells, the most common types include mycosis fungoides (MF) and Sézary syndrome (SS) [1‐3]. MF is an indolent variant that progressively advances primarily in the skin. Skin histology of early patch MF lesions show a low tumor burden with T cell infiltration characterized by Th1 cytokine bias, with increased expression of IL-2 and IFN-γ [3‐6]. In addition, Th1 chemokines such as CXC chemokine ligand 9 (CXCL 9) and CXCL10 are also expressed in lesional skin of early CTCL, when epidermotropism of tumor cells is evident [7]. SS is an aggressive variant of CTCL characterized by erythroderma, lymphadenopathy and malignant T cells circulating in the blood. SS can have eosinophilia, a high level of IgE and chemokine ligand 17 (CCL17) in patients [8, 9]. MF and SS share similarities in gene expression and a subset of MF progresses to SS. Immune analysis of the skin in SS shows a Th2 cytokine profile [10] and the malignant T cells exhibit a Th2 cytokine pattern with increased IL-4 [11]. From gene profiling studies, a unique gene expression phenotype of SS has been uncovered [12]. Gene expression changes in SS, such as decreased expression of IFN-γ, and increased expression of unique biomarker genes identified in SS such as TWIST1, and TOX are frequent and represent important features of CTCL [13‐15].

Recent studies have established that both the tumor microenvironment (TME) and the activity of tumor-infiltrating stromal cells affect cancer phenotypes [16]. The contribution of the TME to cancer prognosis was highlighted by a recent analysis of 39 malignancies that revealed that TME gene signatures are better predictors of survival than genes expressed in malignant tumor cells [17]. The TME is comprised of abundant fibroblasts and immune cells, as well as endothelial cells and extracellular matrix (ECM) components, which closely interact with tumor cells. Crosstalk between the TME and tumor cells can regulate cancer progression either positively or negatively. Fibroblasts have been shown to play an important role in maintaining the ECM and regulating epithelial differentiation by stromal–epithelial crosstalk to establish an invasion-permissive TME [18]. In B-cell lymphomas, fibroblasts have an inverse correlation with survival outcomes compared to carcinomas [19]. In CTCL, fibroblasts are an important component of the TME and have been shown to promote tumorigenesis by augmenting Th2 and attenuating Th1 immune responses [20]. In MF lesional skin, fibroblast-derived periostin promotes the production of thymic stromal lymphopoietin (TSLP) [21]. TSLP subsequently activates immature myeloid dendritic cells (DCs) to produce the Th2-attracting cytokine CCL17 [22], suggesting that fibroblasts from CTCL may nurture a Th2-dominant TME in MF lesions through TSLP secretion. Moreover, fibroblasts contribute to the high level of exotoxin-3 in lesional skin of CTCL, which interacts with CCR3 and controls the Th2-dominant TME in CTCL [23].

A Th1 bias has been described in the skin in early MF, when malignant cells are sparse, but how this immune bias develops is unclear. A Th1 cytokine pattern in the TME may suggest the presence of anti-tumor immunity that inhibits disease progression, which is consistent with an indolent clinical course of MF seen in the majority of patients. This observation was complemented by the finding that T cell clones isolated from early MF skin lesions lack a Th2-polarized cytokine pattern [24]. The interaction of tumor T cells with fibroblasts in MF is not well studied, but normal fibroblasts have variable activities in cancer and can exert suppressive functions against tumor cells [25]. With the indolent nature of MF, one hypothesis is that an interaction between skin fibroblasts and malignant T cells influences malignant T cell growth. The underlying propensity for immune bias is illustrated when culturing benign host T cells from SS patients in vitro away from the malignant Th2 cells, which leads to an enhanced Th1 cytokine pattern [26]. These findings suggest an important role for the TME in immune bias.

To better understand the interactions between fibroblasts and neoplastic T cells in CTCL, we studied immune changes and biomarker regulation using in vitro culture of skin fibroblasts and MF cells. Here we describe one of the first studies using a novel 2-dimensional co-culture method to demonstrate the immune regulation by skin fibroblasts of CTCL cells, and investigate how fibroblasts promote changes in CTCL.

Primary fibroblasts were co-cultured with MyLa cells as previously described [29] (Fig. 1). Briefly, normal and tumor fibroblasts (5 × 10⁵ cells/ml) were seeded into separate 6-well plates, and cultured in RPMI 1640 medium containing 10% FBS and antibiotics until ~ 70% confluence was reached. Upon ~ 70% confluence, MyLa cells (3 × 10⁵/ml), derived from a patient with advanced MF, were added with fresh medium, and cultured with the fibroblasts for 5 days. MyLa cells were also cultured in the absence of fibroblasts as a control. After 5 days, co-cultured MyLa cells and normal/lesional fibroblasts were trypsinized and re-plated into new 6-well plates for 40 min to allow fibroblasts to adhere to the plastic, leaving MyLa cells in suspension. Separated MyLa cells along with their no-fibroblast controls were lysed for RNA extraction.

Studies have shown that tumor growth is preceded or accompanied by activation of local host stroma [53], which plays a major role in disease evolution and response to therapy [54]. CAFs are stromal cells that are abundant in a variety of cancers and have diverse tumor-restraining/promoting roles [55‐58]. The cutaneous TME in CTCL includes abundant stromal fibroblast, but their influence on malignant T cell is poorly characterized.

The majority of MF patients experience indolent disease limited to the skin and have an excellent prognosis unless lesions thicken and progress to tumor stage [59]. In contrast, SS is a CTCL with more aggressive and rapidly progressive disease. Early MF skin lesions are characterized by a low burden of malignant T cells accompanied by a reactive immune cell infiltrate. In addition to the immune cells, the skin ME background consists of mesenchymal stromal cells, many of which are fibroblasts [60]. Malignant T cell burden increases with progression from patch to plaque to tumor lesions in MF. Concomitant changes in the TME can be detected, such as increased angiogenesis and stromal fibroblasts expressing matrix metalloproteinase-2 (MMP2) [61]. As disease stage progresses, the skin architecture is disrupted by TME-related changes including epidermal fibrosis, increased Th2 cells and cytokines, and declining expression of Th1 factors [3, 62‐64]. Th2-like malignant T cells proliferate in the TME in the presence of MF fibroblasts. Upregulation of CTLA-4 on the surface of malignant T cells further suppresses host immunity [65]. Highlighting the importance of this immune shift in disease progression is that restoring cytokines seen in early MF by treating advanced CTCL with IFN-α and IFN-γ is an effective strategy for treatment [66]. These observations suggest that fibroblasts in early skin lesions may contribute to the indolent nature of early MF by ameliorating disease-promoting gene expression in malignant T cells.

The current study is a first step to elucidate the regulatory role of fibroblasts in CTCL compared to normal fibroblasts. The present study focuses on the dermal fibroblasts between CTCL and normal, and specifically directed at major differences in their interaction with malignant T cells. In the development of CTCL lesions in the skin, there are no privilege sites where lesions have not been observed and in this initial report, we are focused on analyzing properties in fibroblasts that is independent of location. However future studies comparing properties of fibroblasts from different sites in the skin will be important in yielding understanding into the mechanism of progression and site predilection.

In the present study, we isolated fibroblasts from patient tissue, using established methods to elute and analyze fibroblast properties [67‐70]. Here we described a novel 2-dimensional co-culture system by incubating primary fibroblasts with malignant T cells. These fibroblasts were rested in culture and were not subjected to any mechanical stress at the time of analysis. In our experiments, both normal and CTCL fibroblast were co-cultured with MyLa cells simultaneously to study disparate properties of these different fibroblast populations.

It has been observed that CAFs greatly influence the TME via the secretion of cytokines and chemokines [71, 72], regulate the plasticity of cancer stem cells [73], and play a significant role in the development of drug resistance [74]. Our findings demonstrated that MF fibroblasts have upregulated expression of IL32 and FAPα compared to normal fibroblasts (Fig. 2a). IL-32 expression has been described in keratinocytes (KCs) in lesional skin of MF patches and plaques, and in atypical T cells in the dermis in MF tumors [36]. IL-32 expression was also detected in KCs in atopic dermatitis [75]. However this study is the first to show differential IL32 expression in MF fibroblasts compared to normal fibroblasts, which resembles a recent study where IL-32 was found to be abundantly expressed in CAFs in the breast cancer TME [76]. We also showed that fibroblasts in MF differ from normal fibroblasts in the expression of FAP-α, a known CAF marker. We observe expression in MF fibroblasts that is consistent with the previous report of FAP-α in MF lesions [34]. Furthermore, the current study shows increasing expression of FAP-α in more advanced MF stages, and demonstrates for the first time that FAP-α expression can be correlated to stage (Fig. 2f-i) compared to normal donors (Fig. 2b-c).

CTCL presents with a variable clinical presentation, nevertheless there are recurrent morphologic features that can serve as clues in diagnosis and classification. We believe that the distinctive phenotype may reflect signature genes important in Sezary syndrome (SS). Indeed, from transcriptome studies of SS, a number of genes can be frequently detected to be abnormally expressed that are biomarkers [15, 77]. Therefore, we selected a number of biomarker genes to identify CTCL cell lines that are representative of SS for in vitro studies. Of the four cell lines tested (Fig. 3d), only MyLa cells expressed TWIST1, TOX and cytokine genes that are seen in primary SS cells. Based on these results, MyLa cells were chosen as a model SS cells for our experiments.

Using the novel co-culture system in this report, we describe for the first time that normal fibroblasts in co-culture can induce gene expression changes in MyLa T cells, suggesting that fibroblasts from normal skin may suppress disease-promoting gene expression in malignant T cells. Specifically, we showed that fibroblasts can modulate MyLa cells to suppress Th2 genes (Fig. 5a) and enhance Th1-related gene expression (Fig. 4d), suppress expression of the MF biomarkers TWIST1 (Fig. 4a) and TOX genes (Fig. 4b), and inhibit proliferation of MyLa cells (Fig. 5d). The results suggest that fibroblasts in early MF skin lesions may be a microenvironment that is less hospitable for proliferation. We also demonstrated that MF tumor-derived fibroblasts differ from normal fibroblasts in their ability to alter disease-associated gene expression in MyLa cells (Figs. 4a-b and 5a). In contrast to normal fiborblasts, we showed that MF fibroblasts promote expression of Th2 cytokine genes (Fig. 5b-c) and lack the ability to suppress expression of TWIST1 and TOX genes (Fig. 4a-b). Whether this is similar to CAF from solid tumors is unclear. ACTA2, which is highly expressed in carcinomas [37‐39], is not upregulated in MF fibroblasts compared to their normal counterparts (Fig. 3b). Collectively, these results suggest that signals from normal fibroblasts may mimic the TME of early MF skin lesions, creating an environment inhospitable for proliferation.

Interestingly, these findings are reminiscent of studies on diffuse large B-cell lymphoma (DLBCL), where the finding of stromal gene signature representing fibroblasts and extracellular matrix components associated with good survival, and creating a TME not conducive for lymphoma progression [19]. Lenz et al. identified two sets of stromal gene signatures in DLBCL patients [78]. The stromal-1 gene signature was found to be associated with better survival in DLBCL patients, which includes the genes that are associated with poor survival in other carcinomas [66]. However, the mechanism behind the suppressive effect of normal fibroblasts in CTCL is unclear, and will need further study. Our findings have implications for the understanding of tumor progression in MF in the early stages when malignant cells are sparse, when the skin architecture is preserved with the presence of normal fibroblasts, and when the immune infiltrate consists primarily of nonmalignant (reactive) Th1 cells and cytotoxic CD8+ T cells [3, 79]. The results we found from co-culturing normal fibroblasts with MyLa cells suggest that the Th1 immunophenotype of early stage MF skin lesions may be affected by fibroblasts that increase IFNG and TBX21 expression in T cells. Our observations indicate that MF progression is accompanied by changes in fibroblast phenotype in the TME. We show that MF fibroblasts have a modulatory effect on malignant T cells, supporting their expression of disease-associated cytokines (Fig. 5b-c) and CTCL biomarkers (Fig. 4a-b). When fibroblasts from tumor stage lesions were co-culture with MyLa cells, IL4 and IL16 expression (Fig. 5b-c) was increased, which is consistent with a malignant CTCL phenotype [3, 80]. In that setting, high levels of TWIST1 and TOX expression were maintained, in marked contrast to the significant decreases observed when MyLa cells were co-culture with normal fibroblast (Fig. 4a-b). The results from these experiments demonstrate for the first time that fibroblasts from normal skin and MF tumors differ functionally in their ability to modulate gene expression of malignant T cells.

In summary, our results describe novel activities of normal and MF fibroblasts to modulate gene expression in malignant T cells (Fig. 6). In this report, we focus on the transcription regulation induced by the interaction between fibroblasts and MyLa cells, and not on the function of these gene products such as GATA3 and TBX21 that are biomarkers. These novel findings suggest that the TME changes during the evolution of MF tumors, and suggest that fibroblasts in the TME play a role in disease pathogenesis and progression. Whether MF tumor fibroblasts may protect malignant T cells from cytotoxic and genotoxic therapies remains unclear. Further detailed global transcriptomic studies and functional studies such as gene silencing and blocking receptors will be needed to explore the complex interacting pathways between CTCL fibroblasts and primary CTCL cells that alters the regulation of gene expression in CTCL cells. Importantly, understanding of the mechanism important in dysregulated gene expression by skin fibroblasts may lead to the development of new strategies to discover novel compounds for the treatment of CTCL.