Cell culture conditions
HEK293T cells were cultured in DMEM (Invitrogen #41965) supplemented with 10% FCS (Invitrogen #10500) and 1% penicillin/streptomycin 10,000 U (Invitrogen #15140). UOK-262 and UOK-262-FH cells were cultured in DMEM (Invitrogen #21969) supplemented with 2 mM L-glutamine and 10% FCS. Cells were cultured at 37°C in a 5% CO2 atmosphere. All cell lines were tested via “Multiplex cell Contamination Test” as previously described .
Packaging of lentiviral pool
DECIPHER library human Module 1 (#DHPAC-M1-P) and human Module 2 (DHDAC-M2-P) were obtained from Cellecta Inc. (Mountain View, U.S.). For packaging into lentiviral particles, plasmid DNA from Modules 1 and 2 was co-transfected into HEK293T cells together with the second generation helper plasmids psPAX2 and pMD2.G. One day before transfection, HEK293T cells were seeded into nine 175 cm2 flasks (Greiner Bio One #660175) at 7.0E + 04 cells per cm2 in standard DMEM medium with freshly added L-Glutamine (2 mM final) and without penicillin/streptomycin. At the day of transfection, 60 μg from each Module was mixed with 100 μg pMD2.G, 200 μg psPAX2 and 600 μl PLUS™ Reagent (Invitrogen #11514-015) into a total volume of 13 ml Opti-MEM® (Invitrogen #11058-021). The plasmid mix was incubated for 15 min at RT before mixing with 14 ml Opti-MEM® containing 900 μl Lipofectamine™ (Invitrogen #18324020). The complete transfection mix was incubated for another 15 min at RT before adding 2.9 ml to each cell culture flask containing HEK293T cells. At 24 h post transfection the medium in each culture flask was replaced by 20 ml high serum DMEM (30% FCS) containing DNase (1 U/ml, QIAGEN #19101) and MgCl2 (4 mM). Addition of DNase helps reducing carryover of plasmid DNA into the viral pool. At 24 h after addition of high serum medium, the lentivirus containing supernatant was harvested and stored at 4°C. To each culture flask 20 ml fresh high serum medium were added and flasks were incubated for an additional 24 h. Finally, the complete lentiviral harvest was pooled to a total volume of 360 ml and passed through a 0.45 μm PES filter (Nalgene #166-0045). Lentiviral pools were stored in aliquots at −80°C for further use.
Pooled RNAi screen
HEK293T cells were seeded 24 h prior to transduction. A total of 5.0E + 07 cells were seeded for transduction with each Module into 25 cell culture flasks the size of 175 cm2. Cells were transduced with the lentivirally packaged Modules at low multiplicity of infection (MOI = 0.3) in standard DMEM containing 3.5 μg/ml Polybrene® (Sigma-Aldrich #107689). As a result, we anticipate that each of the 27,500 shRNA expression plasmids present in each Module integrated into the genome of 540 individual cells. The viral supernatant was replaced 24 h later with standard DMEM culture medium containing 2 μg/ml cell culture tested puromycin dihydrochloride (Sigma-Aldrich #5858-2). Selection was continued for an additional 5 days after which cells were transferred into fresh standard DMEM medium without antibiotics and grown for an additional 24 h. Following this, 2.8E + 07 cells transduced with either Module were transferred into four new 175 cm2 flasks in standard DMEM without antibiotics and freshly added L-Glutamine (2 mM). Cells in two flasks were transfected with a mix of two siRNAs targeting the expression of Fumarate Hydratase (siFH) while cell in the two additional flasks were transfected with of an AllStars negative control siRNA (siCTRL). For siFH transfection, 18.5 ml Opti-MEM® containing 1 ml HiPerfect (QIAGEN #301707) were mixed with 220 μl siFH_2 (20 μM QIAGEN #SI00002989) and 220 μl siFH_3 (20 μM QIAGEN #SI00002996). For the siCTRL transfection, 440 μl AllStars negative control siRNA (20 μM QIAGEN #1027281) was mixed with identical amounts of Opti-MEM® and HiPerfect. Both master mixes were mixed by vortexing and incubated at RT for 10 min. Finally transfection complexes were added dropwise to two culture flasks from each Module. At 40 h post-transfection, cell from each of the four samples (Module1/siFH, Module1/siCTRL, Module2/siFH, and Module2/siCTRL) were transferred into 20 cell culture flasks (175 cm2) in standard DMEM medium at 7,0E + 05 cells per cm2 and grown for an additional 80 h. By the end of the screen, cells were harvested and cell pellets from each sample were stored in three aliquots (~4.0E + 07 cells per aliquot) at −20°C for further analysis.
Cell pellets from each sample were thawed and resuspended in 5 ml buffer P1 (QIAGEN #19051) supplemented with 100 μg/ml RNase A (QIAGEN #19101) and 0.5% SDS. After 5 min incubation at RT the DNA was sheared sonicating the cells for 5 sec. Following sonication genomic DNA was extracted from each sample using the DNeasy Blood and Tissue Kit (QIAGEN #69504). Total DNA yield was around 300 μg from each sample. PCR amplification of barcodes was carried out in two separate rounds using the Titanium® Taq PCR Kit (Clonetech #639211). The reaction composition for the first round PCR was 10 μl 10× Titanium Taq PCR Buffer, 8 μl dNTPs (2.5 mM each), 1 μl Titanium Taq, 3 μl FwdHTS Primer (10 μM), 3 μl RevHTS Primer (10 μM) and 50 μg template genomic DNA adjusted to a total of 100 μl with PCR-grade water. Four reactions were prepared from each sample resulting in a total of 200 μg genomic DNA being used for barcode amplification. The PCR program for the first round PCR was (1) 94°C for 3 min, (2) 94°C for 30 sec, (3) 65°C for 10 sec, (4) 72°C for 20 sec, (5) goto (2) 15 times, (6) 68°C for 2 min, (7) 10°C forever. For each sample, the four first round reactions were pooled and 2 μl of this pool were used as template for second round PCR. In addition to the 2 μl template from the first round PCR, the reaction composition of the second round PCR was 10 μl 10× Titanium Taq PCR Buffer, 8 μl dNTPs (2.5 mM each), 1 μl Titanium Taq, 5 μl FwdGEX Primer (10 μM), 5 μl RevGEX Primer (10 μM) adjusted to a total of 100 μl with PCR-grade water. The PCR program for the first round PCR was (1) 94°C for 3 min, (2) 94°C for 30 sec, (3) 65°C for 10 sec, (4) 72°C for 10 sec, (5) goto (2) 9 times, (6) 68°C for 2 min, (7) 10°C forever. For each sample three second round PCRs were performed and every reaction was purified individually using QIAquick PCR purification Kit (QIAGEN #28106) and eluted in 50 μl elution buffer (EB). Eluted PCR products were analyzed individually via gel electrophoresis and Nanodrop ND1000. PCR products were 106 base pairs long for every reaction and amounts were almost identical. Triplicate reactions from each sample were pooled and purified via gel electrophoresis (3.5% agarose) using QIAquick Gel Extraction Kit (QIAGEN #28704). Gel excision was performed without directly staining the 106 bp band with ethidium bromide but smaller bands which were run alongside with the band for excision. After gel purification the PCR products were eluted in 2× 10 μl elution buffer, heat denatured at 95°C for 5 min and placed on ice. The gel purified fragments were adjusted to equal concentrations (50 ng/μl). Finally 2 μl of each fragment were analyzed on a 3.5% agarose gel and found to be of correct size. Following successful recovery, the barcodes were quantified using a Genome Analyzer IIx (Illumina) and the GexSeqN sequencing primer at a final concentration of 500 nM. Barcodes from Module 1 and Module 2 were sequenced combined in one lane for siFH or siCTRL respectively. In average 3.0E + 07 reads were sequenced per lane. The sequences for all primers used are shown in Additional file 5.
Analysis of sequencing data
For quantification of read counts per barcode, the Barcode Deconvoluter software was used which is available for download from the Cellecta website. Read counts of individual barcodes were normalized to the average read count in each lane to adjust for varying total read counts in different sequencing lanes. Barcodes with less than 100 reads in the control lane were excluded from analysis (Module 1 = 2.3%; Module 2 = 2.7%). From the remaining barcodes the ratios between read counts after FH knockdown and negative control were calculated. All log2 ratios used for further analysis are shown in Additional file 2. Each module included twenty-one negative control shRNA expression constructs targeting the gene Luciferase (Luc) for knockdown. The mean standard deviation from those constructs was calculated as an estimate of variance in the screen. Standard deviations were 0.1595 and 0.1863 for Modules 1 and 2, respectively. Expression constructs with log2 ratios lower than the negative standard deviation of Luciferase constructs were considered to be specifically depleted in the FH knockdown sample. The number of constructs per gene that passed those filter criteria was counted. Genes represented by more than half of the constructs they were targeted by, were considered as potentially synthetic lethal interaction partners of FH. In total 340 genes (Additional file 3) out of the 10,455 genes targeted by both Modules (Additional file 1) were considered as candidate genes and used for signaling pathway analysis.
Pathway enrichment analysis of candidate synthetic lethal genes
A hypergeometric test was used to compute an enrichment p-value for 861 KEGG and Reactome pathways. P-values were FDR corrected for multiple testing. Significantly overlapping pathways were filtered by iteratively going over the list of pathways (sorted by FDR corrected p-value), removing pathways with more than 50% overlap with previous pathways in the list.
Cloning of shRNA expression constructs
Candidate oligonucleotide sequences were synthesized and desalted (Sigma). Sequences are shown in Additional file 4. At 5′ the guide strand was synthesized with an additional ACCG overhang and the passanger strand with CGAA to allow subsequent cloning into the BpiI (Fermentas #ER1011) digested pRSI9 vector. Following phosphorylation (T4 PNK, Fermentas #EK0031) and annealing of guide and passanger strand (99°C to 4°C at −0.1°C per second) the double stranded oligonucleotides were ligated into BpiI digested pRSI9 vector with T4 ligase (Fermentas #EL0011) for 1 h at RT. Ligated plasmids were transformed into chemically competent E. coli (strain STBL3) using standard heat shock protocol. Transformed E. coli were transferred onto agar plates supplemented with 100 μg/ml carbenicillin and incubated at 37°C overnight. Single clones were used to inoculate 5 ml LB (100 μg/ml carbenicillin). After 18 h at 37°C with shaking at 200 rpm, plasmids were isolated using QIAprep Spin Miniprep Kit (QIAGEN #27106). Purified plasmids were checked for correctly inserted sequences via sequencing (GATC) with FwdU6 primer. From E. coli cultures harboring the correct plasmid, glycerol stocks were prepared and stored at −80°C for further use.
Quantitative real-time PCR
Cells were pelleted and total RNA was extracted using Qiagen RNeasy Mini Kit (QIAGEN #74104). RNA concentration was adjusted to 50 ng/μl. The OneStep RT-PCR Kit (QIAGEN #210210) was used according to instructions in the manual and reactions were prepared in a 384 well format with 50 ng template per well. QIAGEN QuantiTect primer shown in Additional file 5 were used. Reactions were run on run on a LightCycler 480 (Roche).
Samples were run on a 10% SDS-PAGE gel and blotted on nitrocellulose membrane. Primary antibodies were incubated overnight at 4°C. FH antibody (Sigma #2500433) was diluted 1:10,000 and beta-actin antibody (Sigma #A3853) was diluted 1:5,000. Secondary horseradish peroxidase conjugated antibodies were incubated for 1 h at RT. Anti-goat was diluted 1:5,000 and anti mouse was diluted 1:2,000.
Cells were seeded into 96 well plates at 600 cells per well and transduced at MOI > 1 in appropriate cell culture medium supplemented with 5 μg/ml polybrene. At 6 days post transduction, cell culture medium was replaced by 50 μl medium containing 20 μg/ml resazurine (Acros Organics #418900010). Following incubation fluorescence intensity was detected (Ex = 544 nm/Em =590 nm).
Colony formation assay
Colony formation assays were performed using a modified version of the sulforhodamine B colorimetric protocol developed by Vichai & Kirtikara, . 600 cells were plated in triplicate into 6-well plates infected with lentiviral particles against the indicated genes or treated with the indicated doses of adenylate cyclase inhibitor MDL-12,330A. At the end of the treatment cells were allowed to recover for 4 days and then fixed using trichloroacetic acid (TCA) at the final concentration of 3% followed by 1 h incubation at 4°C. The plates were washed four times with water before being allowed to air-dry at room temperature overnight. Colonies were stained by incubation with 1 ml 0.05% (wt/vol) sulforhodamine B solution in 1% (vol/vol) acetic acid for 30 min. Unbound dye was removed and the plate washed four times with 1% (vol/vol) acetic acid. Colonies were counted manually and by a ColCount automated colony counter (Oxford Optronix).
2 × 105 cells were plated onto 6-well plates and cultured in standard medium for 12 h. The medium was replaced with fresh medium supplemented with drugs at the indicated concentrations or DMSO at the maximal concentration used in the drug treatments. At the end of the incubation period, cells were lysed with 250 μl 0.1% DMSO in 0.1 M HCl for 20 min at room temperature. The lysate was centrifuged at 1,000 × g for 5 min and the supernatant split for protein quantification by BCA assay (20 μl) and cAMP quantification (200 μl). The cAMP levels were quantified using the Direct cAMP ELISA kit (Enzo Life Sciences) following manufacturer’s instructions. Absorbance at 405 nm (A405) was detected using an Infinite 200 PRO plate reader (Tecan). The average A405 from blank wells was subtracted from all other wells and the cAMP concentration of samples was interpolated from a four-parameter logistic curve fitted to cAMP standards before normalising by the total protein concentration in each sample. In each individual experiment, drug treatments were performed in triplicate and the cAMP and protein quantifications for each sample were performed in duplicate.