The tyrosine phosphatase SHP-1 regulates hypoxia inducible factor-1α (HIF-1α) protein levels in endothelial cells under hypoxia

Abstract

Introduction: The tyrosine phosphatase SHP-1 negatively influences endothelial function, such as VEGF signaling and reactive oxygen species (ROS) formation, and has been shown to influence angiogenesis during tissue ischemia. In ischemic tissues, hypoxia induced angiogenesis is crucial for restoring oxygen supply. However, the exact mechanism how SHP-1 affects endothelial function during ischemia or hypoxia remains unclear. We performed in vitro endothelial cell culture experiments to characterize the role of SHP-1 during hypoxia.

Results: SHP-1 knock-down by specific antisense oligodesoxynucleotides (AS-Odn) increased cell growth as well as VEGF synthesis and secretion during 24 hours of hypoxia compared to control AS-Odn. This was prevented by HIF-1α inhibition (echinomycin and apigenin). SHP-1 knock-down as well as overexpression of a catalytically inactive SHP-1 (SHP-1 CS) further enhanced HIF-1α protein levels, whereas overexpression of a constitutively active SHP-1 (SHP-1 E74A) resulted in decreased HIF-1α levels during hypoxia, compared to wildtype SHP-1. Proteasome inhibition (MG132) returned HIF-1α levels to control or wildtype levels respectively in these cells. SHP-1 silencing did not alter HIF-1α mRNA levels. Finally, under hypoxic conditions SHP-1 knock-down enhanced intracellular endothelial reactive oxygen species (ROS) formation, as measured by oxidation of H2-DCF and DHE fluorescence.

Conclusions: SHP-1 decreases half-life of HIF-1α under hypoxic conditions resulting in decreased cell growth due to diminished VEGF synthesis and secretion. The regulatory effect of SHP-1 on HIF-1α stability may be mediated by inhibition of endothelial ROS formation stabilizing HIF-1α protein. These findings highlight the importance of SHP-1 in hypoxic signaling and its potential as therapeutic target in ischemic diseases.

Fig 1. SHP-1 expression and localization during hypoxia.

(A) The induction of HIF-1α protein levels by 1h and 4h hypoxia was confirmed by western blot (n = 4). (B) HIF-1α (green) and SHP-1 (red) were detected by immunofluorescence staining. DAPI (blue) was used to visualize nuclei. HIF-1α was induced by hypoxia (***p<0.001; n = 6). SHP-1 expression was also slightly induced by hypoxia (p<0.05; n = 6). Both SHP-1 and HIF-1α were shown to be predominantly located in the nucleus with moderate expression in the cytoplasm. Graphs next to photos show fluorescent intensities of SHP-1 and HIF-1α (n = 6 of 6 visual fields/sample) (C) Immunoprecipitation was performed for SHP-1. HIF-1α could be detected in precipitates of SHP-1 (n = 6). IgG: IgG isotype control antibody. (D) HIF-1α could successfully be immunoprecipitated by a specific antibody (lane 6 upper blot) and is tyrosine phosphorylated under normoxia and hypoxia as seen in MG132 (10μM) treated cells (lane 3 and 4 lower blot, n = 3). IgG: IgG isotype control antibody.

Fig 2. SHP-1 knock-down increased cell growth and VEGF synthesis and secretion during hypoxia.

(A) Cell growth was quantified by measurement of total protein content of single wells. Knock-down of SHP-1 resulted in increased cell growth during 24 hours of hypoxia compared to control Odn (*p<0.05, n = 18). (B) Intracellular VEGF was measured by FACS. SHP-1 knock-down led to increased levels of intracellular VEGF after 24 hours of hypoxia (*p<0.05, n = 10), which was blocked by echinomycin (Ecm, 10ng/ml) (*p<0.05 vs. control Odn; n = 10) as well as apigenin (Apg, 50μM) (*p<0.05 vs. control Odn; n = 10). (C) SHP-1 knock-down resulted in elevated VEGF concentration in supernatants of hypoxic HMECs, as measured by ELISA (*p<0.01, n = 18). Inhibition of HIF-1α activity by echinomycin (10ng/ml) or apigenin (50μM) reduced this in both control Odn and SHP-1 AS-Odn treated cells (*p<0.05 vs. control Odn; n = 8). (D) The increased cell growth during hypoxia by SHP-1 knock-down could be prevented by adding the HIF-1 inhibitor echinomycin (Ecm, 10 ng/ml) (*p<0.05, n = 11).

Fig 3. SHP-1 destabilizes HIF-1α protein during hypoxia.

(A) HIF-1α protein levels were further enhanced by SHP-1 knock-down during hypoxia (4h) (*p<0.05, n = 4, lane 4) compared to control conditions (lane 3). Graph shows the protein band density of HIF-1α in relation to loading control β-Actin. (B) HIF-1α levels in non transfected cells was compared to SHP-1 WT expressing cells showing no influence of transfection on this (upper blot; n = 4). HMECs expressing inactive SHP-1 (CS) showed increased (*p<0.05; n = 8, lane 2 in lower blot), whereas expression of constitutively active SHP-1 (E74A) showed decreased (*p<0.05; n = 9, lane 3 in lower blot) levels of HIF-1α compared to wildtype (WT) SHP-1 (lane 1 in lower blot). Graph shows HIF-1α protein band density in relation to the loading control β-Actin. (C) HIF-1α mRNA was quantified by qRT-PCR. Hypoxia (4h) significantly decreased HIF-1α mRNA compared to normoxic conditions (*p<0.05; n = 10). However, SHP-1 knock-down had no effect on HIF-1α mRNA levels (n = 10). (D) The regulatory effect of SHP-1 on HIF-1α levels shown in (A) could not be observed when protein degradation was prevented by using the proteasome inhibitor MG132 (10μM) (n = 4). Graph shows HIF-1α protein band density in relation to the loading control β-Actin. (E) Differences in HIF-1α translation between cells expressing catalytic inactive (CS) and wildtype (WT) SHP-1 were not detected (n = 8) and the decreased HIF-1α accumulation seen in cells expressing constitutively active (E74A) SHP-1 was rescued upon inhibition of proteasomal inhibition (MG132 10μM) (n = 8).

Fig 4. SHP-1 knock-down enhances ROS production during hypoxia.

(A) SHP-1 knock-down enhanced ROS formation during hypoxia (*p<0.001; n = 12) compared to control AS-Odn treated cells. (B) SHP-1 knock-down enhanced hypoxia induced intracellular superoxide production (*p<0.05, n = 8), as assessed by DHE fluorescence.