Fig S1: Ibrutinib induces apoptosis and inhibits proliferation of Mino and Rec-1 cells but the effect is impaired by stromal cells. Fig S2: Stromal cells confer ibrutinib resistance by a tumor cell non-autonomous mechanism. Fig S3: Disrupting MCL cell interaction with stromal cells by blocking the adhesion molecule VLA4 overcomes ibrutinib resistance in vitro. Fig S4: Attachment to stromal cells supports MCL cell regrowth after ibrutinib removal. Fig S5: Stromal cells do not influence ibrutinib-mediated down-regulation of NF-κB target genes but impair ibrutinib-mediated down-regulation of pAKT signaling. Fig S6: p110α inhibition overcomes stromal cell-mediated ibrutinib resistance by restoring the ability of ibrutinib to down-regulate pAKT levels. Fig S7: Effects of ibrutinib or BYL719 on mice weight and expression of NF-kB target genes in the MCL tumors. Fig S8: p110α inhibitor BYL719 or ibrutinib do not induce MS-5 stromal cell apoptosis.

For most lymphomas, including diffuse large B-cell lymphoma (DLBCL), the male incidence is higher, and the prognosis is worse compared to females. The reasons are unclear; however, epidemiological and experimental data suggest that estrogens are involved. With this in mind, we analyzed gene expression data from a publicly available cohort (EGAD00001003600) of 746 DLBCL samples based on RNA sequencing. We found 1293 genes to be differentially expressed between males and females (adj. p-value < 0.05). Few autosomal genes and pathways showed common sex-regulated expression between germinal center B-cell (GCB) and activated B-cell lymphoma (ABC) DLBCL. Analysis of differentially expressed genes between pre- vs. postmenopausal females identified 208 GCB and 345 ABC genes, with only 5 being shared. When combining the differentially expressed genes between females vs. males and pre- vs. postmenopausal females, nine putative estrogen-regulated genes were identified in ABC DLBCL. Two of them, NR4A2 and MUC5B, showed induced and repressed expression, respectively. Interestingly, NR4A2 has been reported as a tumor suppressor in lymphoma. We show that ABC DLBCL females with a high NR4A2 expression showed better survival. Inversely, MUC5B expression causes a more malignant phenotype in several cancers. NR4A2 and MUC5B were confirmed to be estrogen-regulated when the ABC cell line U2932 was grafted to mice. The results demonstrate sex- and female reproductive age-dependent differences in gene expression between DLBCL subtypes, likely due to estrogens. This may contribute to the sex differences in incidence and prognosis.

Supplementary Table S1 showing QPCR primers used in the study

Supplementary Materials and Methods

Refractory and relapsed B cell lymphomas are often driven by the difficult-to-target oncogene MYC. Here, we report that high MYC expression stimulates proliferation and protects B lymphoma cells from apoptosis un-der normal oxidative stress levels and that compounds including N-acetylcysteine (NAC) and vitamin C (VitC) induce apoptosis by reducing oxidative stress. NAC and VitC injections effectively reduce tumor growth in lymphoma cells with high MYC expression but not in those with low MYC expression. MYC knockdown confers tumor resistance to NAC and VitC, while MYC activation renders B cells sensitive to these compounds. Mechanistically, NAC and VitC stimulate MYC binding to EGR1 through Cys117 of MYC, shifting its transcriptional output from cell cycle to apoptosis gene expression. These results identify a redox-controlled mechanism for MYC's role in maintaining proliferation and preventing apoptosis, offering a potential therapeutic rationale for evaluating NAC or VitC in patients with MYC-driven B cell lymphoma.

Mantle cell lymphoma (MCL) is a non-Hodgkin lymphoma with one of the highest male-to-female incidence ratios. The reason for this is not clear, but epidemiological as well as experimental data have suggested a role for estrogens, particularly acting through estrogen receptor β (ESR2). To study the ESR2 effects on MCL progression, MCL cells sensitive and resistant to the Bruton tyrosine kinase inhibitor ibrutinib were grafted to mice and treated with the ESR2-selective agonist diarylpropionitrile (DPN). The results showed that the DPN treatment of mice grafted with both ibrutinib-sensitive and -resistant MCL tumors resulted in impaired tumor progression. To identify the signaling pathways involved in the impaired tumor progression following ESR2 agonist treatment, the transcriptome and ESR2 binding to target genes were investigated by genome-wide chromatin immunoprecipitation in Granta-519 MCL tumors. DPN-regulated genes were enriched in several biological processes that included cell–cell adhesion, endothelial–mesenchymal transition, nuclear factor-kappaB signaling, vasculogenesis, lymphocyte proliferation, and apoptosis. In addition, downregulation of individual genes, such as SOX11 and MALAT1, that play a role in MCL progression was also observed. Furthermore, the data suggested an interplay between the lymphoma cells and the tumor microenvironment in response to the ESR2 agonist. In conclusion, the results clarify the mechanisms by which estrogens, via ESR2, impair MCL tumor progression and provide a possible explanation for the sex-dependent difference in incidence. Furthermore, targeting ESR2 with a selective agonist may be an additional option when considering the treatment of both ibrutinib-sensitive and -resistant MCL tumors.

Abstract Acquired resistance to cancer drugs is common, also for modern targeted drugs like the Bruton tyrosine kinase (BTK) inhibitor ibrutinib, a new drug approved for the treatment of the highly aggressive and relapsing mantle cell lymphoma (MCL). The tumor microenvironment often impacts negatively on drug response. Here, we demonstrate that stromal cells protect MCL cells from ibrutinib-induced apoptosis and support MCL cell regrowth after drug removal by impairing ibrutinib-mediated downregulation of PI3K/AKT signaling. Importantly, the stromal cell–mediated ibrutinib resistance was overcome in vitro by inhibiting AKT activity using the PI3K catalytic p110α subunit–specific inhibitor BYL719. This was seen both for MCL cell lines and primary MCL cells. Furthermore, inhibition of p110α activity by BYL719 potentiated the ability of ibrutinib to inhibit MCL tumor growth in vivo in a mouse xenograft model. The stromal cell–mediated ibrutinib resistance was found to be due to a direct interaction with MCL cells and involves the integrin VLA-4, as disrupting stromal cell–MCL cell interaction using a VLA-4 blocking antibody abrogated the ibrutinib resistance. This suggests that combined treatment with ibrutinib and a p110α inhibitor, alternatively by disrupting stromal cell–MCL cell interaction, may be a promising therapeutic strategy to overcome stromal cell–mediated ibrutinib resistance in MCL. Mol Cancer Ther; 17(5); 1090–100. ©2018 AACR.

Well-defined physiological functions of estrogens are mediated via nuclear estrogen receptors α (ESR1) and β (ESR2). With regard to hematological malignancies, expression of ESR2 has been found in both B and T cell lymphomas. In addition to endogenous estrogens or selective ESR2 agonists, ESR2 signaling may be affected by both environmental synthetic estrogen-mimicking compounds and dietary phytoestrogens. In the present study, we demonstrate that oral exposure with either the synthetic compound bisphenol A (BPA) or the dietary phytoestrogen genistein reduced the growth of grafted murine T cell (EG7) and human B cell (Granta-519 mantle cell) lymphomas which both express ESR2. Suppression of lymphoma growth was due to reduced proliferation (BPA and genistein) and induction of apoptosis (genistein). Inhibition of lymphoma growth was seen at a BPA dose of 50 µg/kg body weight (BW)/day considered to be safe human exposure dose or a genistein dose of 1 mg/kg BW/day orally, which is reached in soy-rich diets. Thus, our study indicates that the environmental xenoestrogens BPA and genistein have anti-proliferative effects on ESR2-expressing lymphomas. Our data suggest that phytoestrogens may be considered as a dietary supplement for lymphoma patients and possibly for prevention of lymphoid malignancies.

Diffuse large B-cell lymphoma (DLBCL) shows a higher incidence in males versus females. Epidemiological studies have shown that female gender is a favorable prognostic factor, which may be explained by estrogens. Here we show that when grafting human DLBCL cells to immunocompromised mice, tumor growth in males is faster. When treating mice grafted with either germinal center or activated B-cell like DLBCL cells with the selective estrogen receptor beta (ER beta) agonist diarylpropionitrile, tumor growth was significantly inhibited. Furthermore, nuclear ER beta 1 expression analysis in primary DLBCL's by immunohistochemistry revealed expression in 89% of the cases. Nuclear ERb1 expression was in a univariate and multivariate analysis, an independent prognostic factor for adverse progression-free survival in Rituximab-chemotherapy treated DLBCL (p = 0.02 and p = 0.04, respectively). These results suggest that estrogen signaling through ERb1 is an interesting future therapeutic target for treatment of DLBCL, and that ERb1 expression can be used as a prognostic marker.

Splenic cells from one BALB/c mouse and one C57/BL mouse, immunized with purified rat liver gluco- corticoid receptor (GR), were fused with the mouse myeloma cell line Sp 2/0-Ag 14. Screening for production of anti-GR- antibodies by the hybridomas was carried out with an enzyme- linked immunosorbent assay, using partially purified rat liver GR as antigen. Further screening was by a second-antibody immunoprecipitation assay using (3H)triamcinolone aceto- nide-GR complex from rat liver cytosol as tracer. Hybrid- omas from 10 different microplate wells, positive in both as- says, were successfully cloned by the limiting dilution method to monoclonality. The different origins of the monoclonal anti- bodies were confirmed by their various isoelectric points when analyzed by isoelectric focusing. Four of the monoclonal hy- bridoma cell lines secreted IgM antibodies; two, IgGl; three, IgG2a; and one, IgG2b. The GR-antibody complex was iden- tified in glycerol density gradients by a shift of the 4S GR to an 8.5S or 19S GR-antibody complex when incubated with mono- clonal IgG or IgM antibody, respectively. The 10 monoclonal antibodies recognized different determinants on the GR, all situated on that domain of the receptor that is separate from the ligand and DNA-binding domains. Also, the cross-reactiv- ity to the mouse liver GR varied among the monoclonal anti- bodies. No cross-reactivity was observed to the human lym- phocytic GR. NaDodSO4 electrophoresis of a 0.5% pure GR preparation followed by immunoblotting using one of the

Most lymphomas show higher incidence and poorer prognosis in males compared to females. However, the endocrine contribution to this gender difference is not entirely known. Here we show that castration accelerates lymphoma growth in C57BL6 male mice grafted with murine EG7 T cell lymphoma cells. However, the androgen receptor antagonist Bicalutamide did not affect lymphoma growth, suggesting no impact of androgen receptor signaling on lymphoma progression. In contrast, inhibition of androgen-to-estrogen conversion by the aromatase inhibitor (AI) Letrozole induced faster lymphoma growth in mice, suggesting that androgens impact lymphoma growth through its conversion to estrogens. This was supported by the inability of dihydrotestosterone, which is not converted to estrogens by aromatase, to influence lymphoma growth in castrated male mice. Lymphoma growth was also stimulated in immunocompromised mice grafted with human B cell lymphoma (Granta-519) and treated with either reversible or irreversible AIs, showing that the blockage of estrogen synthesis caused enhanced growth of both murine T and human B cell lymphomas and with different AIs. Additionally, AI-treated EG7 lymphomas showed accelerated growth not only in male but also in intact female mice. Altogether, our results demonstrate that aromatase inhibition accelerates lymphoma growth but not androgens per se, highlighting a protective role of estrogens in lymphoma pathogenesis. These results also raise concern that the use of AIs in women with breast cancer might enhance lymphoma progression.

It has been known for decades that lowering the circulating glucocorticoid (GC) concentration as in Addison's disease or after removing the adrenals results in thymus enlargement, largely due to thymocyte expansion. This has been attributed to the loss of the proapoptotic effects on thymocytes by adrenal GCs. Here, we test this concept and report that ACTH directly controls thymic growth post-adrenalectomy (ADX) independent of the proapoptotic effect of GCs. This was supported by the finding that ADX caused thymus enlargement and a 1.7-fold (P < 0.001) increase in thymocyte number in GR(LckCre) mice resistant to GC-induced thymocyte apoptosis, similar to the increase seen in wild-type mice (2.2-fold; P < 0.01). We show by immunostaining that melanocortin receptor subtype 2, which selectively binds ACTH, is partly expressed on the thymic epithelium. Furthermore, ACTH in comparison to vehicle induced a 2.0-fold (P < 0.01) increase in fetal thymic organ culture thymocyte numbers in vitro and enhanced 2.2-fold (P < 0.05) the expression of delta-like ligand 4, a factor that supports T-cell development. Additionally, adrenalectomized GR(LckCre) mice treated with ACTH under conditions that repressed endogenous ACTH secretion showed increased thymocyte cellularity (1.9-fold; P < 0.01) and splenic naive T-cell numbers (2.5-fold; P < 0.001) compared to when treated with PBS. Altogether, our results show that ACTH directly controls thymocyte homeostasis independent of GCs. These results revise the old paradigm behind compensatory thymus growth following ADX, now demonstrating that ACTH has a central role in regulating thymocyte expansion when systemic GC concentration is low.

Besides generating immunocompetent T lymphocytes, the thymus is an established site of de novo extra-adrenal glucocorticoid (GC) production. Among the compartments of the thymus, both stromal thymic epithelial cells (TECs) and thymocytes secrete biologically active GCs. Locally produced GCs secreted by the various thymic cellular compartments have been suggested to have different impact on thymic homeostasis. TEC-derived GCs may regulate thymocyte differentiation whereas thymocyte-derived GCs might regulate age-dependent involution. However the full biological significance of thymic-derived GCs is still not fully understood. In this review, we summarize and describe recent advances in the understanding of local GC production in the thymus and immunoregulatory steroid production by peripheral T cells and highlight the possible role of local GCs for thymus function.

Glucocorticoids (GCs) strongly impact on different T cell subsets inducing generally immunosuppressive effects, whereas much less is known about the effect of GC on natural killer (NK) cells. The aims of this study were to investigate the effects of GC on T cell functions, including T cell-mediated anti-tumor immune response, and on NK cells. We have used lck-GR mice, which overexpress a transgenic rat GR in both T and NK cells. These mice were found to have decreased both CD4(+) and CD8(+) T cell populations in the periphery. In contrast, both NK and NKT cells were found in normal numbers in lck-GR mice. To identify genes and pathways affected by GR overexpression in our system in T cells, we have compared gene expression profiles in wild-type and lck-GR T cells. Among the genes upregulated in T cells from lck-GR mice, the microarray analysis has identified genes regulating expansion of regulatory T cells. The analysis of genes downregulated in lck-GR mice has identified genes and gene associated with the regulation of immune response. With regard to the effects on T cell functions in lck-GR mice, transgenic expression of GR had a suppressive effect on killer cell activity in vitro. In addition, lck-GR mice showed an increased tumor growth in murine tumor model in vivo, which may be a possible consequence of reduced T cell numbers and activity. We conclude that an increased expression of the GR strongly affects numbers and possibly functions of T cell subsets, but has little effect on NK cells.

Most lymphomas show an increased incidence and poorer prognosis in males vs females, suggesting endocrine regulation. We have previously shown that tumor growth in vivo of a murine T-cell-derived lymphoma is repressed following activation of estrogen receptor β (ERβ, ESR2). By using ERβ-deficient mice, we now demonstrate that this inhibition is mediated via a direct effect on the tumor cells and not on the microenvironment. Furthermore, we show that the growth-suppressing effects of ERβ agonist are also valid for human B-cell lymphomas as demonstrated in tumors derived from Granta-519 mantle cell lymphoma (MCL) and Raji Burkitt lymphoma (BL) cells. In Granta-519 MCL tumors, activation of ERβ reduced expression of BAFF and GRB7, 2 important molecules involved in B-cell proliferation and survival. Importantly, activation of ERβ inhibited angiogenesis and lymphangiogenesis, possibly mediated by impaired vascular endothelial growth factor C expression. Furthermore, using disseminating Raji BL cells, we show that ERβ activation reduces dissemination of grafted Raji BL tumors. We also show by immunohistochemistry that ERβ is expressed in primary MCL tissue. These results suggest that targeting ERβ with agonists may be valuable in the treatment of some lymphomas, affecting several aspects of the malignant process, including proliferation, vascularization, and dissemination.

The glucocorticoid receptor is a nuclear receptor, and can be activated by glucocorticoid ligands. Mitogen- and stress-activated protein kinase (MSK1), when activated by p38 and ERK mitogen-activated protein kinases (MAPKs), plays a major role in chromatin relaxation via phosphorylation of histone H3 S10. The glucocorticoid receptor can target MSK1 as part of its anti-inflammatory mechanism. Here, we studied the converse mechanism, i.e. the impact of MSK1 on glucocorticoid receptor-mediated transactivation. Upstream MSK1-activating kinases concentration-dependently enhanced glucocorticoid response element (GRE)-regulated promoter activity. Correspondingly, MSK1 inhibition, via H89, or combined p38 and ERK MAPK inhibition, via SB203580 and U0126, diminished maximally stimulated GRE-regulated promoter activity using high concentrations of glucocorticoids. Concomitantly, the combination of these agents does not seem to alter site-specific phosphorylations of murine glucocorticoid receptor S212 or S220. Paradoxically, we reveal that a sub-maximally activated GRE-mediated promoter activity, by using lower concentrations of glucocorticoids, is consistently enhanced by H89 or a combination of SB203580 and U0126, irrespective of the GRE promoter context. Furthermore, we show that the glucocorticoid-induced nucleocytoplasmic translocation of MSK1 occurs in a glucocorticoid concentration-dependent manner. The observed glucocorticoid concentration-dependent effect of MSK1 or MAPK inhibition on glucocorticoid receptor transactivation warrants further research into the applicability of combined glucocorticoid and kinase inhibitor strategies for anti-inflammatory purposes.

Estrogens regulate various normal and pathophysiological processes including cancers. Cellular signaling by estrogens is mediated by estrogen receptor α (ERα) and β (ERβ), respectively. Binding of agonists to the ERs affects gene transcription. The main endogenous estrogen, 17β-estradiol (E2), binds to both ERα and ERβ with similar affinity. However, the ligand-binding pocket of ERα and ERβ are slightly different which has allowed the development of selective ER ligands. Importantly, while estrogens via ERα stimulate proliferation, signaling via ERβ inhibits proliferation and promotes apoptosis. In both normal and cancer cells the ERs are co-expressed with ER splice variants which may modify the transcriptional activity of the wild-type receptors. Estrogens have prominent effects on immune functions and both ERα and ERβ are expressed in immune cells and lymphoid malignancies. With regard to lymphoid malignancies, most show estrogen influence as several epidemiological studies of lymphoid cancers demonstrate gender differences in incidence and prognosis with males being more affected. In line with these findings, recent results generated by us have shown that ERβ selective agonists inhibit growth and induce apoptosis in human and murine lymphomas in vivo in xenograft experiments. This suggests that ERβ selective agonists in the future may be useful in the treatment of lymphomas.

•Extraadrenal de novo glucocorticoid synthesis has been identified in various organs.•The locally produced glucocorticoids usually act in a paracrine or autocrine manner.•Locally produced glucocorticoids influence cell differentiation and inflammation.•In the thymus both thymic epithelial cells and thymocytes produce glucocorticoids.•Local glucocorticoid production is suggested to affect thymic homeostasis.

Background: Keratinocyte growth factor (KGF) is a member of the fibroblast growth factor family. KGF protects the oral and intestinal mucosa against damage induced by irradiation or chemotherapy. Previous studies have found the expression of KGF in chondrocytes and suggested that KGF promotes the wound healing process in injured cartilage. KGF also has important effects on the immune system such as the regeneration of thymus tissue and the formation of regulatory T cells (Treg) in the periphery. Aim: Here we investigated the effect of KGF on collagen type II induced arthritis (CIA) and anti-collagen antibody induced arthritis (CAIA) in order to discriminate between immunoregulatory effect and direct protective effect on chondrocytes. Methods: CIA was induced by immunization with CII and CAIA by treatment of mice with a cocktail of four different anti-CII antibodies. The effect of KGF on the thymus and spleen was analyzed by FACS and by immunohistochemistry. Results: We have found that KGF treatment delayed the onset of CIA but had no effect on CAIA. Our results show that KGF treatment leads both to an outflow of naïve T cells from the thymus and to a statistically significant increase in the percentage of CD4+Foxp3+ Tregs in the periphery. Conclusions: We suggest that the effect of KGF on CIA depends on immunoregulatory mechanisms. KGF may delay the aging of the cellular immune system and thus improve the resilience of the immune system against autoimmune reactions.