HACE1 blocks HIF1α accumulation under hypoxia in a RAC1 dependent manner

Uncovering the mechanisms that underpin how tumor cells adapt to microenvironmental stress is essential to better understand cancer progression. The HACE1 (HECT domain and ankyrin repeat-containing E3 ubiquitin-protein ligase) gene is a tumor suppressor that inhibits the growth, invasive capacity, and metastasis of cancer cells. However, the direct regulatory pathways whereby HACE1 confers this tumor-suppressive effect remain to be fully elucidated. In this report, we establish a link between HACE1 and the major stress factor, hypoxia-inducible factor 1 alpha (HIF1α). We find that HACE1 blocks the accumulation of HIF1α during cellular hypoxia through decreased protein stability. This property is dependent on HACE1 E3 ligase activity and loss of Ras-related C3 botulinum toxin substrate 1 (RAC1), an established target of HACE1 mediated ubiquitinylation and degradation. In vivo, genetic deletion of Rac1 reversed the increased HIF1α expression observed in Hace1–/– mice in murine KRasG12D-driven lung tumors. An inverse relationship was observed between HACE1 and HIF1α levels in tumors compared to patient-matched normal kidney tissues, highlighting the potential pathophysiological significance of our findings. Together, our data uncover a previously unrecognized function for the HACE1 tumor suppressor in blocking HIF1α accumulation under hypoxia in a RAC1-dependent manner.

. Genes differently expressed in HEK293 cells stably expressing HACE1 or MSCV under hypoxia vs normoxia HEK293 cells expressing HA-HACE1 or vector alone were exposed to hypoxia (1% O2) for 3hours. For each cell line and culture condition, two independently isolated RNA samples were hybridized to Affymetrix GeneChip Human Exon 1.0 ST (HuEx 1.0) arrays. The criteria defined for selection of differentially expressed genes between hypoxia and normoxia in HEK293 HA-HACE1 and MSCV cells was based on fold-difference of at least 2 or greater and FDR <0.05. All data management and analysis were conducted using R software.

Immunoprecipitation and immunoblotting
For immunoprecipitation, cells were lysed in Nonidet P-40 lysis buffer containing 100 mM NaCl, 5 mM MgCl2, 2 mM EDTA, 1 mM DTT, 10 mM Tris-HCl (pH 7.6), 0.5 % Nonidet P-40, 0,05% sodium deoxycholate, and 1 mM phenylmethylsulfonyl fluoride (PMSF). Cell debris was removed by centrifugation at 13,000 rpm for 20 min, and cell extracts (800 µg protein) were incubated with indicated antibodies overnight. Protein G-or A-Sepharose beads (Thermofisher, Waltham, MA, USA) (50 µl) were used to pull down antibody-protein complexes and were washed with lysis buffer for 3 times to remove non-specific binders. Bound proteins were eluted by boiling in Laemmli buffer for 5-minutes. Immunoblotting was performed using standard methods. Lysates (10-30 μg per lane) were separated with SDS-PAGE gels and transferred to nitrocellulose membranes. The proteins of interest were examined by blotting with respective antibodies overnight at +4C and anti-mouse or rabbit IgG-HRP for 1 hr at RT. All antibodies were used at a dilution of 1:1000 unless otherwise stated. Signals were detected using Enhanced Chemiluminescence (ECL).

Immunofluorescence microscopy
HEK293 cells expressing HA-tagged HACE1 or the vector alone were grown on coverslips. Cells were exposed to various stresses as described above and fixed with 3.5% paraformaldehyde. Cells were permeabilized with 0.5% Nonidet P-40 for 5-minutes and nonspecific signaling was blocked with 3% skim-milk for 10-minutes. HEK293 cells were transfected with plasmids encoding HRE-GFP and Cherry and then cultured under 1% O2 for 16 hrs. GFP and Cherry signals were examined by fluorescence microscopy using a Zeiss LSM 780 confocal microscope (Carl Zeiss, Thornwood, NY) and quantified using ImageJ software.

RAC1 G-LISA activation assay
25 μg total of protein was added wells pre-coated with RAC-GTP-binding protein. Then, the plate was incubated at 4°C for 30-minutes followed by incubation with 50 μl of anti-RAC1 (1/50 in Antibody Dilution Buffer) for 45-minute at RT. After rinsing 3 times with the wash buffer, the plate was incubated with a secondary antibody conjugated with HRP (1/100 in Antibody Dilution Buffer) for 45 min. Finally, 50 μl of HRP detection reagent was added to wells and incubated for 20-minutes. Lastly, 50 μl HRP stop solution was added to each well to stop the reaction and the absorbance was recorded at 490 nm.

HIF1α target gene hybridization assays
Hybridization assays were carried out using the Human HIF-regulated cDNA plate array system (Cat# AP-0111, Signosis BioSignal Capture, Santa Clara, CA, USA), according to according to the manufacturer's protocols. Briefly, RNA samples prepared from 1% O2 or normoxia treated HEK293 cells were reverse transcribed into cDNA in the presence of biotin-dUTP. The cDNA was then incubated with gene-specific oligonucleotides pre-coated in the individual wells of a 96-6 well plate. The captured cDNA was detected with streptavidin-HRP with a dilution of (1:500) on a microplate luminometer. The data were analyzed following normalization to actin levels.

Soft agar colony assays
Soft agar assays were carried out as described (Zhang et al., 2007). SKNEP1 cells expressing wildtype or mutant HIF1α-P402A/P564A were suspended in soft agar and seeded in 6-well culture dish at a density of 45,000 cells per well. The cells were grown in a hypoxia chamber for 2 weeks.
After colonies are stained, wells were imaged, and colony numbers were counted using ImageJ.
7 Quantitative analysis of IHC samples was conducted using ImageJ software. Data represented as average value ± SEM for n = 15 HPFs in 3 tumors/group.
Tissue microarrays (TMAs) from Children's Hospital of Philadelphia were prepared according to standard protocols and consisted of formalin-fixed, paraffin-embedded human tissue cores from primary tumors of Wilms' tumors (9 cases) and 18 cases sarcomas, including Ewing sarcoma (5 cases), alveolar rhabdomyosarcoma (3 cases), embryonal rhabdomyosarcoma (4 cases), and synovial sarcoma (6 cases). These were stained for HACE1 and HIF1α expression as above. Cores were scored for the percentages of cells positively staining for HACE1 and HIF1α as well as for staining intensities. For the latter, a 4-point scale (0-3+) was used according to standard methods.

Gene expression profiling
HEK293 cells expressing HA-HACE1 or vector alone were exposed to hypoxia (1% O2) for 3hours. RNA samples were prepared using TRIZOL (Invitrogen, Waltham, MA, USA), and 5 µg of each RNA sample was used for microarray target synthesis and hybridization as described in the Affimetrix GeneChip manual (Affymetrix, Santa Clara, CA). For each cell line and culture condition, two independently isolated RNA samples (from different cell culture plates, i.e., biological replicates) were hybridized to Affymetrix GeneChip Human Exon 1.0 ST (HuEx 1.0) arrays. Raw data were read and processed by frozen RMA normalization (1). Only "core" probesets were retained for the downstream analysis. Differential expression between hypoxia and normoxia condition was computed by probe-level expression change averaging (PECA) procedure (2). The criteria defined for selection of differentially expressed genes between hypoxia and normoxia in HEK293 HA-HACE1 and MSCV cells was based on fold-difference of at least 2 or greater and FDR <0.05. All data management and analysis were conducted using R software.