Identification of druggable targets from the interactome of the Androgen Receptor and Serum Response Factor pathways in prostate cancer

Background: The Androgen Receptor (AR) pathway is crucial in driving the progression of prostate cancer (PCa) to an advanced state. Despite the introduction of second-generation AR antagonists, such as enzalutamide, majority of patients develop resistance. Several mechanisms of resistance have been identified, including the constitutive activation of the AR pathway, the emergence of AR spliced variants, and the influence of other signalling pathways. The Serum Response Factor (SRF) was previously identified as a possible player of resistance involved in a crosstalk with the AR signalling pathway. Elevated SRF levels in PCa patients were associated with disease progression and resistance to enzalutamide. However, the molecular mediators of the crosstalk between SRF and AR still need to be elucidated. The objective of this study was to identify common interactors of the AR/SRF crosstalk as therapeutic targets.

Methods: Here we used affinity purification mass spectrometry (MS) following immunoprecipitation of SRF and AR, to identify proteins that interact with both SRF and AR. The list of common interactors was expanded using STRING. Four common interactors were functionally validated using MTT assays.

Results: Seven common interactors were identified, including HSP70, HSP0AA1, HSP90AB1, HSAP5, PRDX1, and GAPDH. Pathway analysis revealed that the PI3k/AKT pathway was the most enriched in the AR/SRF network. Moreover, pharmacological inhibition of several proteins in this network, including HSP70, HSP90, PI3k, and AKT, significantly decreased cellular viability of PCa cells.

Conclusions: This study identified a list of AR/SRF common interactors that represent a pipeline of druggable targets for the treatment of PCa.

Methods
Co-Immunoprecipitation Assay

Prior to cellular lysis, the antibody solution was prepared adding 2μg of either SRF (Novus, Biotechne), AR (Santa Cruz, California, US), Ms IgG or Rb IgG to 20μL A/G protein beads (Pierce) and 300μL PBS (Gibco). The beads/antibody mixture was incubated for an hour on a rotator at 4°C. Following incubation, the mixture was washed with ice-cold lysis buffer 3 times.  Cells were scraped with 500μL of lysis buffer (500μL of 1% Triton x100, 1mL of 20mM Tris-HCl pH7.5, 1.5mL of 150mM NaCl and 50μL of 1mM MgCl2) and incubated for 10 mins on ice. 1mg of protein was added to the beads/antibody mix and incubated for 1 hour on a rotator at 4°C. Samples were washed 3 times in ice-cold lysis buffer.

Peptide elution and digestion

Following Co-IP, the peptides in each sample were eluted with 60μL of ice-cold Elution Buffer I (0.012g Urea, 50μL of 1M Tris-HCl pH7.5) and 5μg/mL Trypsin (Promega, Seq Grade Modified) for 30 mins at RT. Samples were then centrifuged at 3000rpm for 30s. The supernatant was collected into a new Eppendorf tube, and 20μL of Elution Buffer II was added to each sample. This step was repeated twice. The supernatant was collected into a new centrifuge tube with a total volume of 110μL. Samples were left to digest overnight at 37°C at 300rpm.

Liquid Mass spectrometry (Bruker timsTOF Pro) and data analysis with MaxQuant

A Bruker timsTOF Pro MS (Bruker Daltonics, Bremen, Germany) connected to an EvoSep One chromatography system was used. The timsTOF Pro MS was run using positive ion polarity with TIMS (Trapped Ion Mobility Spectrometry) and PASEF (Parallel Accumulation Serial Fragmentation) modes. Accumulating ramp times for the TIMS were set at 100ms, with the ion mobility ranging from 0.6 to 1.6 Vs/cm. A mass range from 100 to 1,700 m/z was the set range to record the spectra of ions. The precursor MS Intensity Threshold was set to 2,500 and the precursor Target Intensity was set to 20,000. Each PASEF cycle included one MS ramp for precursor detection, accompanied by 5 PASEF MS/MS ramps and a total cycle time of 1.03s. Peptides were separated using reverse-phase C18 Endurance column (15cm x 150µm ID, C18, 1.9µm) using the Evosep pre-set 30 SPD method. Mobile phase A consisted of 0.1% (v/v) formic acid in water and phase B included 0.1% (v/v) formic acid in acetonitrile. Peptides were separated by increasing gradient of solvent B for 44 minutes with a flow rate of 0.5µL/min. MaxQuant v1.6.17.0 was used by applying the Homo sapiens subset of the Uniprot Swissprot database against the raw data. A contaminants database was included in the search and the ‘Match Between Runs’ and ‘Label free quantification’ were selected. The minimum peptide length allowed was 7 amino acids.  False discovery rate (FDR) for peptides was set at 0.01. Protein intensity of each identified protein was normalised to obtain the label free quantification intensity (LFQi) value. A ProteinGroups.txt output file generated by MaxQuant was used for subsequent data analysis.