U19/Eaf2 binds to and stabilizes von hippel-lindau protein

Abstract

Studies have firmly established a key regulatory role for the tumor suppressor pVHL in the regulation of the vascular system and normal spermatogenesis. Here, we report that knockout of the newly identified tumor suppressor U19/Eaf2 also caused vascular system abnormalities and aspermatogenesis, suggesting a potential link between U19/Eaf2 and pVHL. Coimmunoprecipitation and in vitro binding assays showed an association between U19/Eaf2 and pVHL, whereas deletion mutagenesis revealed the requirement of the NH(2) terminus of U19/Eaf2 and both the alpha and beta domains of pVHL for this binding. U19/Eaf2 stabilizes pVHL, as shown by protein stability and pulse-chase studies. Testes and mouse embryonic fibroblasts (MEF) derived from U19/Eaf2 knockout mice expressed reduced levels of pVHL, indicating that full in vivo expression of pVHL indeed requires U19/Eaf2. As expected, U19/Eaf2 knockout MEF cells exhibited an increased level and activity of hypoxia-inducible factor 1alpha (HIF1alpha), a protein typically regulated via a pVHL-mediated degradation pathway. Furthermore, angiogenesis in a Matrigel plug assay was significantly increased in U19/Eaf2 knockout mice. The above observations argue that U19/Eaf2 can modulate HIF1alpha and angiogenesis, possibly via direct binding and stabilization of pVHL.

Figure 1

Nonmalignant phenotypes in U19/Eaf2-null mice ages ≥22 mo. A, a representative wild-type testis and a regressed U19^−/− testis in old mice. The pictures show gross, H&E (10× and 100× objectives), Ki-67 (100× objective), and TUNEL (100× objective) stainings of the wild-type testis and the regressed U19/Eaf2-null testis. B, incidence rate of aspermatogenesis in wild-type and U19^−/− mice. C, wet weight of testis in wild-type (U19^+/+) and U19/Eaf2-null (U19^−/−) mice. D, extramedullary hemopoiesis in U19^−/− mice. Top, an animal with extramedullary hemopoiesis (arrow). Bottom, H&E, Ki-67, and TUNEL stainings (100× objective) of the wild-type and U19/Eaf2-null spleen.

Figure 2

Interaction of U19/Eaf2 with pVHL. A, Flag-tagged pVHL was cotransfected with Myc-tagged U19/Eaf2 or empty Myc vector into Cos-7 cells. Cell lysates were immunoprecipitated (IP) with anti-Myc antibody–conjugated agarose beads, and blots were probed with anti-Flag antibody (top image) or anti-Myc antibody (bottom image). B, Myc-U19 was cotransfected with HA-pVHL or empty HA vector into 293 cells. Cell lysates were immunoprecipitated with anti-HA antibody–conjugated agarose beads, and blots were probed with anti-Myc antibody (top image) or anti-HA antibody (bottom image). C, whole-cell lysates from RCC4/VHL cells exposed to hypoxia (1% O₂) for 16 h were subjected to immunoprecipitation with control rabbit IgG or rabbit anti-U19/Eaf2 polyclonal antibody, separated by SDS-PAGE, and immunoblotted with anti-HA and anti-U19/Eaf2 antibodies. D, in vitro translated ³⁵S-labeled pVHL was incubated with bacterially purified GST-U19/Eaf2 fusion protein attached to glutathione-agarose beads. GST served as negative control. After incubation, the beads were extensively washed and subjected to SDS-PAGE. Association of ³⁵S-VHL and U19/Eaf2 in vitro was detected by autoradiography. Bottom image, inputs of GST and GST fusion proteins are shown as Coomassie blue–stained gel.

Figure 3

Domain mapping of the interaction between U19/Eaf2 and pVHL. A, pVHL interacted with the NH₂-terminal region of U19/Eaf2 in coimmunoprecipitation assays. HA-pVHL was cotransfected with Myc-tagged U19/Eaf2 COOH-terminal deletion mutants (1–67, 1–113, 1–161, and 1–245) and NH₂-terminal deletion mutants (36–260, 68–260, 114–260, and 162–260) into 293 cells. Wild-type Myc-U19/Eaf2 was cotransfected with HA empty vector as a negative control. Cell lysates were immunoprecipitated with anti-HA antibody–conjugated agarose beads, and blots were probed with anti-Myc antibody (top image) or anti-HA antibody (bottom image). B, both the α and β domains of pVHL were required for interaction with U19/Eaf2 in coimmunoprecipitation assays. HA-U19/Eaf2 was cotransfected with the Myc-tagged major β domain (1–154), α and major β domains (1–192), α and entire β domains, α domain deletion mutant (Δ155–192), or major β domain deletion mutant (Δ64–154) into 293 cells. Wild-type Myc-pVHL cotransfected with HA empty vector served as a negative control. Cell lysates were immunoprecipitated with anti-HA antibody–conjugated agarose beads, and blots were probed with anti-Myc antibody (top image) or anti-HA antibody (bottom image).

Figure 4

Stabilization of pVHL by U19/Eaf2. A, the effect of wild-type U19/Eaf2 (left) or U19/Eaf2 deletion mutant (114–260; right) expression on transfected pVHL protein levels. Left, Cos-7 cells were cotransfected with 2.0 μg of Flag-VHL in the absence (−) or presence of increasing amounts of Myc-U19. Myc empty vector was used to complement the Myc-U19. Right, 293 cells were cotransfected with 2.0 μg of Flag-VHL in the absence (−) or presence of increasing amounts of Myc-U19 (114–260). Myc empty vector was used to complement the Myc-U19 (114–260). B, effect of U19/Eaf2 on pVHL protein stability. Two micrograms of HA-VHL were cotransfected with 2.0 μg of Myc-U19/Eaf2, Myc empty vector, or Myc-U19/Eaf2 (114–260) into 293 cells. Twenty hours after transfection, cells were treated with cycloheximide (CHX) at 50 μg/mL for the indicated number of hours. C, pulse-chase assay to determine HA-VHL protein half-life in the absence or presence of Myc-U19/Eaf2. Two micrograms of HA-VHL expression vector were cotransfected with Myc-U19/Eaf2 or Myc vector into 293 cells. Forty-eight hours after transfection, cells were metabolically labeled with [³⁵S]Met/Cys for 2 h followed by pulse-chase analysis for the indicated time periods. HA-VHL was immunoprecipitated from the cell lysate with anti-HA antibody–conjugated agarose beads and signals were detected by autoradiography. D, endogenous pVHL expression in the testes and MEF cells derived from U19^+/+ and U19^−/− mice. Testes from three wild-type and three U19/Eaf2 knockout mice were used. Two of the three U19/Eaf2 knockout testes were normally sized, whereas the third (center) was reduced. The lack of U19/Eaf2 expression in the U19/Eaf2-null MEFs was verified (30). Blots of the testicular tissue were probed with anti-pVHL, anti–β-actin, or anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibodies. The relative expression level of pVHL was quantitated using the Photoshop CS 7.0 software and normalized to β-actin or GAPDH expression.

Figure 5

U19/Eaf2 modulation of the protein level and transcriptional activity of HIF1α. A, the degradation kinetics of HIF1α in wild-type and U19^−/− MEFs following release from hypoxia. U19^+/+ and U19^−/− MEFs were maintained in a hypoxia chamber (~1% O₂) for 16 h to achieve HIF1α stabilization and then released into normoxia. The lysates were collected at the indicated time points for Western blot analysis of HIF1α. B, HIF1α transcriptional activity in wild-type and U19/Eaf2-null MEF cells. MEF cells were transiently transfected with wild-type (wHRE) or mutant (mHRE) reporter constructs, and transcriptional activity was measured after 16 h in normoxic (21% O₂) or hypoxic (1% O₂) conditions.

Figure 6

The effect of U19/Eaf2 deletion on angiogenesis in Matrigel plugs s.c. implanted in mice. Left, representative digital images of microvessels, stained with anti-CD31 antibody, in Matrigel plugs in the presence of 2 ng/mL VEGF in the wild-type and U19/Eaf2 knockout mice. Right, quantitative analysis of the microvessel densities (MVD) in the Matrigel plugs.