DNA damage-mediated induction of a chemoresistant niche

Abstract

While numerous cell-intrinsic processes are known to play decisive roles in chemotherapeutic response, relatively little is known about the impact of the tumor microenvironment on therapeutic outcome. Here, we use a well-established mouse model of Burkitt's lymphoma to show that paracrine factors in the tumor microenvironment modulate lymphoma cell survival following the administration of genotoxic chemotherapy. Specifically, IL-6 and Timp-1 are released in the thymus in response to DNA damage, creating a "chemo-resistant niche" that promotes the survival of a minimal residual tumor burden and serves as a reservoir for eventual tumor relapse. Notably, IL-6 is released acutely from thymic endothelial cells in a p38-dependent manner following genotoxic stress, and this acute secretory response precedes the gradual induction of senescence in tumor-associated stromal cells. Thus, conventional chemotherapies can induce tumor regression while simultaneously eliciting stress responses that protect subsets of tumor cells in select anatomical locations from drug action.

Figure 1. The thymus represents a chemoprotective niche that harbors surviving lymphoma cells following doxorubicin treatment

(A) Lymphoma-bearing mice were imaged for whole body fluorescence prior to treatment and 4 days following a single dose of 10mg/kg doxorubicin. Representative mice are shown. (B) Ratios of live GFP-tagged Eμ-myc p19^Arf-/- B lymphoma cells in the thymus versus peripheral lymph nodes were quantified by flow cytometry, before (n=4 mice) and 48 hours after (n=5 mice) doxorubicin treatment. Average ratios are indicated with a line. (C) Hematoxylin and eosin (H&E) sections of lymph node and thymus from a tumor-bearing bearing mouse 48 hours after doxorubicin treatment. The dotted line in the thymus demarcates a small region of infiltrating lymphocytes neighboring a larger region of surviving lymphoma cells. Representative fields are shown at 40× magnification. (D) A western blot showing γ-H2AX levels in FACS sorted GFP-positive lymphoma cells from the thymus and peripheral lymph nodes following doxorubicin treatment. β-Tubulin serves as a loading control. The untreated sample is a lysate from cultured lymphoma cells. (E) A Kaplan-Meier curve showing the overall survival of tumor-bearing C57BL/6 (n=8) or C57BL/6 Rag1^-/- (n=5) mice following doxorubicin treatment. The p value was calculated using a logrank test. See also Figure S1.

Figure 2. Thymic conditioned media contains soluble chemoprotective factors

(A) A graph showing lymphoma cell survival in the presence of doxorubicin alone or in the presence of conditioned media. The data are represented as mean +/- standard error of the mean (SEM) (n=3). (B) A graph showing the growth of lymphoma cells cultured in the absence or presence of conditioned media. The data are represented as mean +/- SEM (n=3). (C) Cytokine array analysis of conditioned media from untreated and doxorubicin treated lymph nodes and thymus. The data is represented graphically as normalized signal intensity. Conditioned media was pooled from 3 or 4 mice for each array.

Figure 3. IL-6 and Timp-1 are chemoprotective in vitro and in vivo

(A) A graph showing the fold change in lymphoma cell number 72 hours after treatment with doxorubicin as a single agent or doxorubicin plus recombinant IL-6. The data are represented as mean +/- SEM (n=4 independent experiments). (B) A graph showing the relative survival of cultured lymphoma cells at 24 hour intervals following treatment with doxorubicin alone, doxorubicin plus recombinant IL-6 or Timp-1, or doxorubicin plus both IL-6 and Timp-1. The data are represented as mean +/- SEM (n=3 independent experiments). (C) A schematic diagram of the lymphoma transplant experiment, showing injection of IL-6^+/+ lymphoma cells into both IL-6^+/+ and IL-6^-/- recipients. (D) A Kaplan-Meier curve showing post-treatment survival of IL-6^+/+ (n=17) or IL-6^-/- (n=5) mice bearing IL-6^+/+ lymphomas. All mice were treated with a single dose of 10mg/kg doxorubicin. The p value was calculated using a logrank test. (E) H&E stained sections of lymphomas 72 hours following doxorubicin treatment. The black dotted line shown in the thymus from the IL-6^+/+ recipient mouse demarcates a zone of surviving lymphoma cells that is absent in the other sections. Representative fields are shown at 20× magnification. See also Figure S2.

Figure 4. Doxorubicin induces the release of IL-6, and inhibition of this cytokine signaling sensitizes tumor cells to chemotherapy

(A) Quantification of IL-6 levels in conditioned media from the thymus or lymph nodes of untreated mice (n≥10) or mice treated for 18 hours with 10mg/kg doxorubicin (n≥7). Values were normalized by tissue weight. The data are represented as mean +/- SEM. (B) Quantification of IL-6 levels in conditioned media derived from tumor-bearing thymuses or lymph nodes of doxorubicin treated (n=3) or untreated (n=3) mice. The data are represented as mean +/- SEM. (C) A bar graph showing the fold change in number of live cells following treatment with doxorubicin alone or in combination with conditioned media plus or minus a Jak2 inhibitor. The data are represented as mean +/- SEM (n=3). (D) A Kaplan-Meier curve showing tumor free survival of lymphoma-bearing mice treated with doxorubicin (n=9) or doxorubicin plus two doses of 50mg/kg AG-490 m-CF₃ (n=4). The p value was calculated using a logrank test. (E) H&E sections of lymphomas 72 hours after treatment with doxorubicin or doxorubicin plus AG-490 m-CF₃. Black dotted lines distinguish surviving lymphoma cells, which are largely absent in the presence of AG-490 m-CF₃, from infiltrating immune cells. Representative fields are shown at 20× magnification. See also Figure S3.

Figure 5. Endothelial cells secrete IL-6 and Timp-1 in response to DNA damage in a p38 MAP kinase dependent manner

(A) IL-6 and (B) Timp-1 levels were quantified by ELISA in conditioned media derived from sorted thymic cell populations The data are represented as mean +/- SEM (n≥3 independent experiments). Values were normalized to the number of cells sorted. (C) A graph showing lymphoma cell survival in response to 20nM doxorubicin, with or without endothelial cell co-culture. Fold change in cell number was assessed at 48 and 72 hours post treatment. The data are represented as mean +/- SEM (n=6 independent experiments) (D) A western blot for Bcl-XL levels in lymphoma cells in the presence or absence of TCM for 24 hours. The blot is representative of three independent experiments. (E) A graph showing the results of a GFP competition assay in cells partially transduced with a Bcl-XL shRNA or a control vector. Fold change in GFP percentage was assessed 48 hours following treatment with 20nM doxorubicin. The data are represented as mean +/- SEM (n=3). (F) A bar graph showing the amount of IL-6 in conditioned media from endothelial cells sorted from the thymus of untreated mice (n=5), mice treated with doxorubicin (n=8) or mice treated with doxorubicin plus SB203580 (n=4). Values were normalized to cell number. The data are represented as mean +/- SEM. (G) A graph showing the amount of IL-6 present in conditioned media from untreated and treated human vascular endothelial cells (HUVECs). The data are represented as mean +/- SEM (n=3). (H) A graph showing the amount of IL-6 present in conditioned media from HUVECs 48 hours after treatment with doxorubicin alone or doxorubicin plus either SB203580 or KU55933. The data are represented as mean +/- SEM (n=3). See also Figure S4.

Figure 6. Genotoxic damage promotes cellular senescence in thymic stromal cells and subsequent IL-6 mediated thymic rebound

(A) β-galactosidase staining of normal and tumor-bearing thymuses and lymph nodes in the presence or absence of doxorubicin-induced DNA damage. Representative fields are shown at 20× magnification. (B) A graph showing relative thymic and splenic weight following genotoxic damage in the presence (n=14) and absence (n=6) of IL-6. Organ weights are shown as the ratio of individual irradiated thymus or spleen weights relative to the average un-irradiated thymic or spleen weight for each genotype. The data are represented as mean +/- SEM. See also Figure S5.

Figure 7. DNA damage acutely induces IL-6 in human hepatocellular carcinoma, promoting both cellular survival and senescence

(A) IL-6 levels were quantified in conditioned media derived from Focus cells treated with 200nM doxorubicin for 24 hours. The data are represented as mean +/- SEM (n=3). (B) A graph showing the amount of IL-6 present in Focus cells following treatment with either SB203580 or KU55933, in the presence or absence of doxorubicin. The data are represented as mean +/- SEM (n=3). (C) A graph showing the results of an acute cell survival assay in which Focus cells were treated with doxorubicin and increasing doses of Ag490, as indicated, for 4 days. The data are represented as mean +/- SEM (n=3 independent experiments) (D) A colony formation assay showing Focus cells that were treated with doxorubicin, Ag490 or both for 24 hours before replating. Results are representative of 3 independent experiments.