mTOR and HIF-1alpha-mediated tumor metabolism in an LKB1 mouse model of Peutz-Jeghers syndrome

Abstract

Peutz-Jeghers syndrome (PJS) is a familial cancer disorder due to inherited loss of function mutations in the LKB1/ STK11 serine/threonine kinase. PJS patients develop gastrointestinal hamartomas with 100% penetrance often in the second decade of life, and demonstrate an increased predisposition toward the development of a number of additional malignancies. Among mitogenic signaling pathways, the mammalian-target of rapamycin complex 1 (mTORC1) pathway is hyperactivated in tissues and tumors derived from LKB1-deficient mice. Consistent with a central role for mTORC1 in these tumors, rapamycin as a single agent results in a dramatic suppression of preexisting GI polyps in LKB1+/- mice. However, the key targets of mTORC1 in LKB1-deficient tumors remain unknown. We demonstrate here that these polyps, and LKB1- and AMPK-deficient mouse embryonic fibroblasts, show dramatic up-regulation of the HIF-1alpha transcription factor and its downstream transcriptional targets in an rapamycin-suppressible manner. The HIF-1alpha targets hexokinase II and Glut1 are up-regulated in these polyps, and using FDG-PET, we demonstrate that LKB1+/- mice show increased glucose utilization in focal regions of their GI tract corresponding to these gastrointestinal hamartomas. Importantly, we demonstrate that polyps from human Peutz-Jeghers patients similarly exhibit up-regulated mTORC1 signaling, HIF-1alpha, and GLUT1 levels. Furthermore, like HIF-1alpha and its target genes, the FDG-PET signal in the GI tract of these mice is abolished by rapamycin treatment. These findings suggest a number of therapeutic modalities for the treatment and detection of hamartomas in PJS patients, and potential for the screening and treatment of the 30% of sporadic human lung cancers bearing LKB1 mutations.

Fig. 1.

Rapamycin reduces polyposis, mTORC1 signaling, and proliferation in Lkb1+/− polyps. (A) Top are images of whole stomach and duodenum and Bottom are images of the open stomachs (S) showing the exposed polyps (P) from Lkb1+/− mice treated with either vehicle (VEH) (i, iv) or rapamycin (RAPA) (ii, v) and Lkb1+/+ mice treated with vehicle (VEH) (iii, vi). (B) Immunohistochemical analysis of polyps from VEH- or RAPA-treated Lkb1+/− mice: H & E staining (i, ii), P-S6 staining (iii, iv), and Ki67 staining (v, vi). Results are representative of polyps from 5 mice of each treatment group. (C) A graph of the total polyp burden in Lkb1+/− mice treated with either VEH (○, n = 11) or RAPA (●, n = 10). The mean polyp burden for RAPA-treated mice (2.0 ± 1.2) was significantly reduced (∗, P = 0.00026; Student t test, 2 tail) compared with those mice treated with VEH (9.6 ± 5.5). (D) Average polyp number in VEH-treated mice (black bar, n = 11) or RAPA-treated mice (gray bar, n = 10). Only visible polyps between 1 and ≥5 mm were scored in both VEH- and RAPA-treated mice. The mean polyp number for RAPA-treated mice (2.8 ± 1.4) was significantly reduced (∗, P = 0.00022; Student t test, 2 tail) compared with VEH-treated mice (5.3 ± 1.8). (E) Average polyp size in VEH-treated mice (black bar, n = 11) or RAPA-treated mice (gray bar, n = 10). The mean polyp size in RAPA mice (1.2 ± 0.9) was significantly reduced (∗, P < 0.0001; Student t test, 2 tail) compared with VEH-treated mice (4.4 ± 0.8). (F) Average percentage of Ki67-positive epithelial cells in VEH-treated mice (black bar, n = 5) or RAP- treated mice (gray bar, n = 5). The mean percentage of Ki67-positive cells in RAPA mice (24.5 ± 6.1) was significantly reduced (∗, P < 0.0002; Student t test, 2 tail) compared with VEH-treated mice (59.7 ± 7.3).

Fig. 2.

Up-regulated HIF-1α and HIF-1α targets in LKB1-deficient polyps and fibroblasts are reduced by rapamycin. (A) Immunoblots of lysates made from GI tissue or polyps from Lkb1+/+ and Lkb1+/− mice treated VEH or RAPA. Immunoblots were probed against the indicated antibodies. (B) Immunohistochemical analysis of polyps from VEH- or RAPA-treated Lkb1+/− mice probed with antibodies against Glut1 or Hif-1α. Results are representative of polyps from 3 mice of each treatment group. (C) Immunoblots of lysates from Lkb1+/+ or Lkb1−/− MEFs (Left) or Ampk+/+ or Ampk−/− MEFs (Right) probed with antibodies against the indicated proteins. MEFs were either untreated (NT) or treated with RAPA or cobalt chloride (CoCl2).

Fig. 3.

Polyps from Lkb1+/− mice visualized by FDG PET analysis. (A) Left shows FDG PET images of axial, sagital and coronal views of untreated 12-month-old Lkb1+/+ mice. Right shows the same views of untreated Lkb1+/− mice. The FDG PET images of the mice are labeled accordingly: K, kidney; S, stomach; H, heart; B, bladder; and P, polyp. (B) Left shows FDG PET imaging of axial, sagittal, and coronal views of Lkb1+/+ and Lkb1+/− mice either treated with vehicle or rapamycin at 11 months of age. Right shows the same mice imaged after 1 month of receiving either vehicle or rapamycin. The images of the mice are labeled accordingly: K, kidney; S, stomach; H, heart; B, bladder; and P, polyp.

Fig. 4.

Polyps from human Peutz-Jeghers patients show increased P-S6, GLUT1, and HIF-1α expression. A and B represent immunohistochemistry performed on human small bowel samples from normal patients (Left) or Peutz Jeghers patients (Right) that were probed with antibodies against the mTORC1 marker P-S6. C–F represent immunohistochemistry performed on normal colonic mucosa (Left) and colonic Peutz-Jeghers polyps (Right) probed with antibodies against the GLUT1 protein (C and D) or the HIF-1α protein (E and F).