WIP1 is a novel specific target for growth hormone action

Abstract

DNA damage repair (DDR) is mediated by phosphorylating effectors ATM kinase, CHK2, p53, and γH2AX. We showed earlier that GH suppresses DDR by suppressing pATM, resulting in DNA damage accumulation. Here, we show GH acting through GH receptor (GHR) inducing wild-type p53-inducible phosphatase 1 (WIP1), which dephosphorylated ATM and its effectors in normal human colon cells and three-dimensional human intestinal organoids. Mice bearing GH-secreting xenografts exhibited induced colon WIP1 with suppressed pATM and γH2AX. WIP1 was also induced in buffy coats derived from patients with elevated GH from somatotroph adenomas. In contrast, decreased colon WIP1 was observed in GHR^-/- mice. WIP1 inhibition restored ATM phosphorylation and reversed GH-induced DNA damage. We elucidated a novel GH signaling pathway activating Src/AMPK to trigger HIPK2 nuclear-cytoplasmic relocation and suppressing WIP1 ubiquitination. Concordantly, blocking either AMPK or Src abolished GH-induced WIP1. We identify WIP1 as a specific target for GH-mediated epithelial DNA damage accumulation.

Keywords: biochemistry; molecular biology; physiology.

Graphical abstract

Figure 1

GH induces WIP1 (A–C) Western blots of (A) hNCC line #1, (B) hNCC line #2, and (C) MCF12A cells treated with GH (500 ng/mL) for indicated times; C, untreated control. (D) Colon tissue derived from male nude mice implanted with HCT116 transduced with either lentivirus expressing murine GH (Lenti-mGH) or empty vector (Lenti-V) and sacrificed 5 weeks later. Each lane represents sample analysis derived from an individual animal. (E) Peripheral blood buffy coats derived from patients with non-functioning adenoma (NFA; n = 4) or acromegaly (n = 3). Females and males shown separately. Each blot lane represents an individual patient. Representative blots from at least 3 independent experiments are shown. ImageJ quantifications of Western blots are depicted in Figure S1.

Figure 2

GH induces WIP1 through GHR signaling (A) Western blots of hNCC line #1 pretreated with 20 mg/mL pegvisomant (Peg) for 1 h then treated with GH (500 ng/mL) for an additional 24 h. C, untreated control. (B) Colon tissue derived from 3- and 24-month-old WT and GHR^−/− male mice. Each lane represents sample analysis derived from an individual animal. (C) hNCC line #1 transduced with lentivirus expressing GH (Lenti-GH) or empty vector (Lenti-V) and analyzed 10 days later. (D) iPSC-derived human intestinal organoid lines from 3 different patients were transduced with Lenti-V or Lenti-GH and analyzed 5 weeks later. Representative blots from one human intestinal organoid line are shown. (E) hNCC line #1 transduced with lentivirus expressing control shRNA (shControl) or GH shRNA (shGH). (F) Three different human intestinal organoid lines were transduced with GFP-expressing Lenti-V or Lenti-GH and cultured for 5 weeks, then sorted for GFP-negative cells. GFP-negative cells co-existing in organoids with either Lenti-GH or Lenti-V expressing cells were analyzed. All lines exhibited similar results. Representative blots from one human intestinal organoid line are shown. Representative blots from at least 3 independent experiments are depicted. ImageJ quantifications of western blots are depicted in Figure S2.

Figure 3

Blocking WIP1 increases DDR protein phosphorylation and decreases unrepaired DNA damage (A and B) Western blots of (A) hNCC line #1 treated with WIP1 inhibitor (GSK2830371) for 1 h followed by GH (500 ng/mL) for an additional 24 h and (B) organoids treated with WIP1 inhibitor and GH (500 ng/mL) for 48 h. Untreated cells served as control. Representative blots from at least 3 independent experiments are shown. ImageJ quantification of western blots is depicted in Figure S3. (C) Comet assay of hNCC line #1 treated with WIP1 inhibitor for 1 h followed by GH (500 ng/mL) for an additional 24 h. C, control. Single-cell gel electrophoresis was conducted and Olive tail moment assessed on at least 400 cells/per slide for each experiment. Results are shown as mean of 3 experiments ±SEM. Differences were assessed with non-parametric Kruskal-Wallis test followed by Dunn’s multiple comparison test. ∗∗p < 0.01 versus control; ##p < 0.01 GH versus GH + WIP1 inhibitor.

Figure 4

GH triggers HIPK2 cytoplasmic relocation and induces nuclear WIP1 stability by reducing ubiquitination (A) hNCC line #1 was treated with GH (500 ng/mL) for 24 h, and cell lysates were immunoprecipitated (IP) with IgG or anti-ubiquitin (Ub) antibodies and immunoblotted for WIP1. C, control. (B) Western blot of whole-cell lysates. hNCC line #1 was treated with 20 μM etoposide for 24 h, then with GH (500 ng/mL) after washout for an additional 24 h. (C) WIP1 and ubiquitin (Ub) interactions were analyzed in hNCC line #1 treated with 20 μM etoposide for 24 h, then with GH (500 ng/mL) after washout for an additional 24 h. Cell lysates were immunoprecipitated with IgG or anti-ubiquitin antibodies and immunoblotted for WIP1 and Ub. (D) WIP1 and HIPK2 interactions analyzed in hNCC line #1 treated with 20 μM etoposide for 24 h, then with GH (500 ng/mL) after washout for an additional 24 h. Cell lysates were immunoprecipitated with IgG or anti-HIPK2 antibodies and immunoblotted for WIP1 and HIPK2. (E) WIP1, HIPK2, and Ub interactions analyzed in hNCC line #1 treated with 20 μM etoposide for 24 h, then with GH (500 ng/mL) after washout for additional 24 h. Cell lysates were immunoprecipitated with IgG or anti-WIP1 antibodies and immunoblotted for anti-HIPK2, Ub, and WIP1. (F) Western blots of cytoplasmic and nuclear fractions of expressed proteins in hNCC line #1 left untreated (C) or treated with GH (500 ng/mL) for 6 h and 24 h. ImageJ quantifications of western blots are depicted in Figure S4. (G) Confocal images of hNCC line #1 untreated (Control) or treated with GH (500 ng/mL) for 24 h and stained for HIPK2 (red), WIP1 (green), DAPI (blue), and phalloidin (white). Scale bar, 20 μm.

Figure 5

GH induces WIP1 by activating Src/AMPK (A and B) Western blots of (A) hNCC line #1 treated with 1 μM AMPK inhibitor (Compound C) for 1 h followed by GH (500 ng/mL) for an additional 24 h and (B) human intestinal organoids treated with 2 μM AMPK inhibitor and GH (500 ng/mL) for 48 h. (C–F) Western blots of hNCC line #1 was treated with (C) GH (500 ng/mL) for up to 60 min; (D) 25 μM Src inhibitor for 24 h followed by GH (500 ng/mL) for 30 min–60 min; (E) 25 μM Src inhibitor for 1 h followed by GH (500 ng/mL) for 24 h; and (F) 25 μM Src inhibitor for 24 h followed by GH (500 ng/mL) for 3 h. Untreated cells/organoids served as control (C). (G–I) Western blots of hNCC line #1 transduced with lentivirus expressing Control shRNA (shControl) or Src shRNA (shSrc) (G) analyzed 72 h after transduction; (H) treated with GH (500 ng/mL) for 3 h; and (I) treated with GH (500 ng/mL) for 24 h. ImageJ quantifications of western blots are depicted in Figure S5.

Figure 6

Blocking GH signaling inhibited GH-related WIP1 induction (A–D) Western blots of hNCC line #1 treated with (A) increasing doses of BM001 (GHR synthesis inhibitor) for 24 h; (B) 25 nM BM001 for 24 h followed by GH (500 ng/mL) for 30 min; (C) 25 nM BM001 for 24 h followed by GH (500 ng/mL) for 3 h; and (D) 25 nM BM001 for 1 h followed by GH (500 ng/mL) for an additional 24 h. (E) Colon tissue derived from 3-month-old female mice treated i.p. with DMSO (C, control) or BM001 (3 mg/kg) 3 times/week for 3 weeks. Mice were sacrificed 24 h after last injection. Each lane represents an individual animal. C, control. ImageJ quantifications of western blots are depicted in Figure S6.

Figure 7

Proposed mechanisms for GH-mediated WIP1 induction GH activates Src/AMPK phosphorylation. Phospho-AMPK, in turn, phosphorylates HIPK2, which is relocated to the cytoplasm, weakening binding to intranuclear WIP1, decreasing WIP1 ubiquitination, and increasing WIP1 stability. Induced WIP1 dephosphorylates ATM, thus suppressing DDR activity and leading to accumulated DNA damage. Created with BioRender.com.