Functional screening reveals HORMAD1-driven gene dependencies associated with translesion synthesis and replication stress tolerance

HORMAD1 expression is usually restricted to germline cells, but it becomes mis-expressed in epithelial cells in ~60% of triple-negative breast cancers (TNBCs), where it is associated with elevated genomic instability (1). HORMAD1 expression in TNBC is bimodal with HORMAD1-positive TNBC representing a biologically distinct disease group. Identification of HORMAD1-driven genetic dependencies may uncover novel therapies for this disease group. To study HORMAD1-driven genetic dependencies, we generated a SUM159 cell line model with doxycycline-inducible HORMAD1 that replicated genomic instability phenotypes seen in HORMAD1-positive TNBC (1). Using small interfering RNA screens, we identified candidate genes whose depletion selectively inhibited the cellular growth of HORMAD1-expressing cells. We validated five genes (ATR, BRIP1, POLH, TDP1 and XRCC1), depletion of which led to reduced cellular growth or clonogenic survival in cells expressing HORMAD1. In addition to the translesion synthesis (TLS) polymerase POLH, we identified a HORMAD1-driven dependency upon additional TLS polymerases, namely POLK, REV1, REV3L and REV7. Our data confirms that out-of-context somatic expression of HORMAD1 can lead to genomic instability and reveals that HORMAD1 expression induces dependencies upon replication stress tolerance pathways, such as translesion synthesis. Our data also suggest that HORMAD1 expression could be a patient selection biomarker for agents targeting replication stress.

: Doxycycline-inducible HORMAD1 expression drives genomic instability. Fig. S2: RNAi screen quality control data. Fig. S3: Fourteen DDR-associated gene candidates identified in the primary screen.     respectively. STR valiadated cell line stocks were used and all cell lines used were within 30 passages of this STR typing. All cells were routinely tested for mycoplasma Lentiviral p20 HORMAD1 and GFP construct generation HORMAD1 or GFP cDNAs were cloned into the pInducer20 lentivirus expression vector (a gift from Stephen Elledge, Addgene plasmid # 44012) according to a protocol described previously (1). Lentiviral particles were produced in HEK293T cells by cotransfection of the relevant pINDUCER20 plasmids with the pMD2.G (Addgene plasmid # 12259) and psPAX2 (Addgene plasmid # 12260) packaging vectors, which were a gift from Didier Trono. Virus-containing cell culture media was collected for downstream transduction of cell lines and the virus titre was estimated in SUM159 using serial dilution of viral supernatants and selection with 0.5mg/ml Geneticin (Gibco, Thermo Fisher Scientific, Waltham, MA, USA).

Generation of isogenic and clonal cell systems
Parental (heterogenous) SUM159 cells were isolated in 96 well plates using serial dilutions to obtain a final density of 1 cell/well. Colony formation was monitored by light microscopy and single-cell colonies were expanded to generate "parental clones".
Parental clones were plated into 6-well plates, infected with lentiviral vectors at a multiplicity of infection (MOI) of 1. After two days, selection was initiated by adding 0.5 centred luminescence value for siRNA in the doxycycline arm, = log2 platecentred luminescence value for siRNA in the DMSO arm.
Where: = Drug Effect for siRNA , = Median Drug Effect for entire library, =Median Absolute Deviation of Drug Effects across entire library. This methodology was described previously (5). For subsequent interrogation, we selected genes with a DE-Z score of less than -3, indicating a HORMAD1/doxycycline induced gene sensitivity. We removed genes which also had a DE-Z of less than -2 in the parental cell line. We also removed inherently toxic genes (Z score of less than -3).

Single oligo siRNA deconvolution RNA interference validation
Secondary validation experiments were performed with the same methodology as the primary screen but using sets of four individual targeting siRNAs (Dharmacon, Lafayette, CO, USA), as well as siRNA SMARTpools (Dharmacon, Lafayette, CO, USA) utilised in the primary screen. Data were analysed and presented using the normalised percentage inhibition, NPI= 3/4 # 4 $ /4 # *100. Where: = raw luminescence value for siRNA , / = average raw luminescence value for non-targeting siRNA control (siALLSTAR), 5 = average raw luminescence value for targeting siRNA control (siPLK1), as described previously (6). Data from triplicate wells of a single replicate are shown.

Validation of gene knockdown by RT-qPCR
Gene knockdown following treatment with each siRNA was determined by RT-qPCR.

CRISPR-Cas9-based viability assays
pLentiCRISPR-mCherry was a gift from Beat Bornhauser (Addgene plasmid # 75161, ). Edit-R crRNA sequences targeting POLH were custom-designed using CRISPR v4.2 software (TEFOR) (8). Target sequences of each crRNA are listed in Table S7. Following primary antibody incubation, membranes were washed 4x with TBS-T for 5 minutes at room temperature, and then incubated with a secondary antibody (raised against the species of the primary antibody) conjugated to HRP. Membranes were washed a further 4 times for 5 minutes at room temperature before imaging using electrochemiluminescence (ECL) (GE Healthcare, Chicago, IL, USA).

Immunofluorescence analysis of γH2AX
Clonally-derived HORMAD1-inducible SUM159 cells were treated with doxycycline

STRING Protein Network analysis
STRING Protein Network Analysis (9) was carried out by importing the list of 63 candidate genes into the multiple proteins search tab and selecting Homo sapiens as the organism. MCL clustering with inflation parameter of 3 was used as the clustering method. For network settings, the following settings were used: 1) meaning of network edges: evidence; 2) active interaction sources: textmining, experiments, databases, co-expression, neighborhood, gene fusion and co-occurrence; 3) minimum required interaction score: medium confidence (0.400); 4) network display mode: interactive svg.

Statistical analysis
All data are presented as the mean with standard deviation, from 3 replicate experiments. All pair-wise statistical analysis was performed using parametric t-tests, unless otherwise stated.