Selective upregulation of interleukin-8 by human rhabdomyosarcomas in response to hypoxia: therapeutic implications

Abstract

Rhabdomyosarcoma (RMS) is the most common soft-tissue sarcoma of adolescence and childhood. Because RMS tumors are highly vascularized, we sought to determine which factors secreted by RMS cells are crucial in stimulating angiogenesis in response to hypoxia. To address this issue, we evaluated expression of several proangiogenic factors [interleukin (IL)-8, vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF)-2, stromal-derived factor (SDF)-1, hepatocyte growth factor (HGF) and leukemia inhibitory factor (LIF)] in 8 human RMS cell lines in both normal steady-state and hypoxic conditions. We found by real-time quantitative polymerase chain reaction (RQ-PCR) and confirmed by enzyme-linked immunosorbent assay (ELISA) that from all the factors evaluated, IL-8, whose expression is very low in normoxia, had been very highly expressed and secreted by RMS cells lines during hypoxic conditions ( approximately 40-170 times). Interestingly, this upregulation was not affected by knocking down hypoxia-inducible factor (HIF)-1alpha, but was inhibited by mitogen-activated protein kinase (MAPK)p42/44 and phosphatidylinositaol 3-kinase (PI3K)/AKT pathway inhibitors. This suggests that IL-8 expression is regulated in an activating protein (AP)-1- and nuclear factor (NF)-kappaB-dependent manner. Furthermore, we found that conditioned media (CM) harvested from RMS cells exposed to hypoxia activated and stimulated chemotactic responses in human umbilical vein endothelial cells (HUVECs) and that IL-8 was responsible for hypoxia-related effects. Finally, by employing shRNA, the expression of IL-8 in human RH-30 cells was downregulated. We noticed that such RMS cells, if injected into skeletal muscles of immunodeficient mice, have a reduced ability for tumor formation. We conclude that IL-8 is a pivotal proangiogenic factor released by human RMS cells in hypoxic conditions and that the targeting of IL-8 may prove to be a novel and efficient strategy for inhibiting RMS growth.

Figure 1. Hypoxia up-regulates expression of IL-8 in human RMS both at the mRNA and protein levels

Panel A: RMS cell lines were incubated under normal (21%) and reduced oxygen conditions (1%). Sixteen hrs later, RNA was extracted to evaluate the level of mRNA for IL-8 by real time PCR. Panel B: Similarly, at the same time point, CM were harvested to measure the level of IL-8 by ELISA. Data from four separate experiments are pooled together. * p<0.0001. Panel C: Expression of IL-8 was also measured by immunofluorescence directly in tumor xenografts (RH30 tumors growing in SCID mice). Intratumoral hypoxia was detected by immunohistochemistry by employing Hypoxyprobe-1 HP-2 FITC). IL-8 was detected by PE-anti-IL-8 Abs. The experiment was repeated three times with similar results. A representative study is shown. Left panel: RH 30 tumor. Middle panel: RH-30 tumor. Enlarged inset from left panel: yellow staining shows co-localization of hypoxia and IL-8 expression. Right panel: murine skeletal muscles with no tumor.

Figure 2. CM from RMS cell cultured in hypoxia possesses stronger chemotactic activity for HUVECs and shows angiogenic potential

Panel A: CM collected from RH30 cells cultured in hypoxia (CM RH30 H) have higher chemotactic activity as compared to CM collected from RH30 cells growing in normoxia (CM RH30 N). When CM from RH30 cells cultured in hypoxia were precleared with anti-IL-8 blocking Abs (CM RH30 H + Ab), chemotactic activity decreased. Control: chemotaxis to culture medium alone. Data from four separate experiments are pooled together. * p<0.0001. Panel B: Similar experiment with CM harvested from RH30 cells stable transfected with shRNA against IL-8, growing in normoxia (CM RH30 IL-8 KD N) and hypoxia (CM RH30 IL-8 KD H). CM RH30 N and CM RH30 H: CM harvested from RH30 cells cultured in normoxic and hypoxic conditions respectively. Control: chemotaxis to culture medium alone. Data from four separate experiments are pooled together. * p<0.0001. Inset: ELISA data for IL-8 secretion by wild type RH30 and shRNA manipulated RH30 cells. Panels C and D: CM collected from RH30 cells cultured in hypoxia possesses a higher ability to form capillaries by HUVECs on Matrigel-coated plates. This effect was reduced when CM were collected from RH30 cells with shRNA-directed, reduced expression of IL-8 cultured in hypoxic conditions. Data from three separate experiments are pooled together. * p<0.0001. Representative pictures are shown.

Figure 3. PI3K-AKT and MAPK are responsible for hypoxia-induced production of IL-8 in RMS

Panel A: RH30 cells were incubated in normoxia (N) or hypoxia (H). Sixteen hrs later, total protein was extracted and analyzed for activation/phosphorylation of PI3K-AKT and MAPK pathways. The experiment was repeated three times with similar results. A representative study is shown. Panels B and C: RH30 cells were incubated in hypoxia and normoxia in presence of wortmanin (2μM), MEK inhibitor (2μM), or a combination of both. Sixteen hrs later, the level of expression of IL-8 was analyzed by real-time PCR (panel B) and the protein level in culture media was measured by ELISA (panel C). Data from four separate experiments are pooled together. * p<0.0001.

Figure 4. HIF-1α does not influence hypoxia-induced expression of IL-8 in RMS

Panel A: Expression of HIF-1α in RH30 cells non-exposed or exposed for 16 hrs to HIF-1α shRNA or scrambled shRNA. The experiment was repeated three times with similar results. A representative study is shown. Panels B and C: RH30 cells were not exposed (RH30) or exposed to HIF-1α scrambled- (RH30 scrambled) or HIF-1α-shRNA (RH30 HIF-1 KD) in hypoxic and normoxic conditions. Sixteen hrs later, the level of IL-8 mRNA in RH30 extracts was evaluated by real time RT-PCR (Panel A) and IL-8 secretion to CM was measured by ELISA (Panel B). Data from four separate experiments are pooled together.

Figure 5. Effect of IL-8 promotes angiogenesis and tumor growth in vivo

Panel A: Level of hemoglobin (Drabkin assay) in control (empty) Matrigel implants and implants containing IL-8 (IL-8) implanted for 4 weeks into SCID-Beige inbred mice. Data from three separate experiments (5 animals/group) are pooled together. * p<0.0001. Inset: Representative pictures of vascularization of Matrigel plugs (arrows point to blood vessels). Panel B: RH30, RH30 scrambled, and RH30 IL-8 KD cells were inoculated into the hind limb muscles of SCID-Beige inbred mice. Tumors were measured every 7 days for 28 days. Data from three separate experiments are pooled together. * p<0.0001. Panel C: Human IL-8 level in murine peripheral blood in the same mice as measured at end of the experiments by ELISA. Data from three separate experiments are pooled together. * p<0.0001. Panel D: Detection of CD31 positive endothelial cells in tumor sections. Representative images of CD31 expression in tumor sections after RH30 (left), RH30 scrambled (middle) and RH30 IL-8 KD cells (right) inoculation. Panel E: Level of hemoglobin (Drabkin assay) in supernatants from homogenized tumors formed after inoculation of RH30, RH30 scrambled, and RH30 IL-8 KD cells into the hind limb muscles of SCID-Beige inbred mice.

Figure 6. MAPKp42/44-AP-1 and PIK-AKT-NFκB pathways regulate hypoxia-induced expression of IL-8 in RMS

RMS cells exposed to hypoxia activate PI3K-AKT and MAPK42/44 pathways, which lead to increased production of IL-8 through binding of NF-κB and AP-1 transcription factors to the IL-8 promoter. IL-8 secreted by the growing RMS tumor chemoattracts endothelial progenitors and promotes angiogenesis. Furthermore, IL-8 is also probably playing an important role in chemoattraction of stroma precursors for the expanding RMS tumor.