TYPE Brief Research Report PUBLISHED 09 February 2023 DOI 10.3389/fimmu.2023.1112761 SAMHD1 expression modulates OPEN ACCESS innate immune activation and EDITED BY Nadine Laguette, Centre National de la Recherche correlates with ovarian Scientifique (CNRS), France REVIEWED BY cancer prognosis Renate König, Paul-Ehrlich-Institut (PEI), Germany Yea-Lih Lin, Lucı́a Gutiérrez-Chamorro 1†, Eudald Felip 1,2†, Université de Montpellier, France Adrià Bernat-Peguera 2, Ifeanyi Jude Ezeonwumelu 1, Iris Teruel 2, *CORRESPONDENCE Margarita Romeo Anna Martı́nez-Cardús 2, Bonaventura Clotet 1,3, mromeo@iconcologia.net Eva Riveira-Muñoz 1, Margarita Romeo 2*, Mireia Margelı́ 2* Mireia Margelı´ mmargeli@iconcologia.net and Ester Ballana 1* Ester Ballana 1 IrsiCaixa AIDS Research Institute – and Health Research Institute Germans Trias i Pujol (IGTP), Hospital eballana@irsicaixa.es Germans Trias i Pujol, Universitat Autònoma de Barcelona, Badalona, Spain, 2 Medical Oncology † These authors have contributed Department, Catalan Institute of Oncology (ICO), B-ARGO (Badalona Applied Research Group in equally to this work Oncology), Health Research Institute Germans Trias i Pujol (IGTP), Universitat Autònoma de Barcelona, Badalona, Spain, 3 Consorcio Centro de Investigación Biomédica en Red de Enfermedades Infecciosas SPECIALTY SECTION (CIBERINFEC), Instituto de Salud Carlos III, Madrid, Spain This article was submitted to Molecular Innate Immunity, a section of the journal Frontiers in Immunology Purpose: SAMHD1 is a deoxynucleotide triphosphate (dNTP) triphosphohydrolase RECEIVED 30 November 2022 ACCEPTED 25 January 2023 which has been proposed as a putative prognostic factor in haematological PUBLISHED 09 February 2023 cancers and certain solid tumours, although with controversial data. Here, we CITATION evaluate SAMHD1 function in ovarian cancer, both in vitro and in ovarian cancer Gutiérrez-Chamorro L, Felip E, patients. Bernat-Peguera A, Ezeonwumelu IJ, Teruel I, Martı´nez-Cardús A, Clotet B, Riveira-Muñoz E, Romeo M, Margelı´ M and Methods: SAMHD1 expression was downregulated in ovarian cancer cell lines Ballana E (2023) SAMHD1 expression OVCAR3 and SKOV3 by RNA interference. Gene and protein expression changes in modulates innate immune activation and immune signalling pathways were assessed. SAMHD1 expression in ovarian cancer correlates with ovarian cancer prognosis. Front. Immunol. 14:1112761. patients was evaluated by immunohistochemistry and survival analysis was doi: 10.3389/fimmu.2023.1112761 performed according to SAMHD1 expression. COPYRIGHT © 2023 Gutiérrez-Chamorro, Felip, Results: SAMHD1 knockdown induced a significant upregulation of Bernat-Peguera, Ezeonwumelu, Teruel, Martı´nez-Cardús, Clotet, Riveira-Muñoz, proinflammatory cytokines concomitant to increased expression of the main Romeo, Margelı´ and Ballana. This is an open- RNA-sensors, MDA5 and RIG-I, and interferon-stimulated genes, supporting the access article distributed under the terms of idea that the absence of SAMHD1 promotes innate immune activation in vitro. To the Creative Commons Attribution License (CC BY). The use, distribution or assess the contribution of SAMHD1 in ovarian cancer patients, tumours were reproduction in other forums is permitted, stratified in SAMHD1-low and SAMHD1-high expressing tumours, resulting in provided the original author(s) and the copyright owner(s) are credited and that significantly shorter progression free survival (PFS) and overall survival (OS) in the original publication in this journal is SAMHD1-high expression subgroup (p=0.01 and 0.04, respectively). cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not Conclusions: SAMHD1 depletion correlates with increased innate immune cell comply with these terms. signalling in ovarian cancer cells. In clinical samples, SAMHD1-low expressing tumors showed increased progression free survival and overall survival irrespective of BRCA mutation status. These results point towards SAMHD1 modulation as a new therapeutic strategy, able to enhance innate immune activation directly in tumour cells, leading to improved prognosis in ovarian cancer. KEYWORDS SAMHD1, ovarian cancer, RLR (RIG-I like receptors), interferon, inflammation Frontiers in Immunology 01 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 Introduction Materials and methods Ovarian cancer is the most lethal gynaecologic cancer and long Cell lines and generation of SAMHD1 term outcomes are still unsatisfactory, irrespectively of the advent of knock-down cells new treatment strategies (1). At present, it is clear that genetic alterations such as mutations in BRCA1 and BRCA2 tumour Human OVCAR3 cells were obtained from American Type suppressor genes and other alterations in DNA repair machinery Culture Collection (ATCC) and grown in RPMI 1640 medium influence ovarian cancer development in a significant proportion of supplemented with 10% heat-inactivated fetal bovine serum (FBS; cases (2). In this context, the introduction of poly (ADP-ribose) ThermoFisher), 100 U/mL penicillin, 100 mg/mL streptomycin (Life polymerase inhibitors (PARPi) represented an important therapeutic Technologies) and insulin solution human (0.01 mg/mL) (Sigma- step forward particularly, but not only, for BRCA mutated ovarian Aldrich) and maintained at 37°C in a 5% CO2 incubator. Human cancer patients (3). Furthermore, the function of the immune system SKOV3 cells were obtained from ATCC and grown in Gibco™ has become a matter of extensive research in ovarian cancer, McCoy’s 5A (Modified) Medium (ThermoFisher) supplemented prompting the development of different immunotherapeutic with 10% of heat-inactivated fetal bovine serum (FBS, approaches as putative effective treatments for ovarian cancer. ThermoFisher) and antibiotics (100 U/mL penicillin, 100 mg/mL Immunotherapy in ovarian cancers aims at the stimulation of streptomycin (Life Technologies)) and maintained at 37°C in a 5% antigen-presenting cells, the induction of antitumor immunity and CO2 incubator. the attenuation of immunosuppressive microenvironment, albeit with Ovarian cancer cells were transfected following standard inconsistent results in clinical trials (4). At the same time, new procedures. In brief, siRNAs targeting SAMHD1 gene (siSAMHD1, insights on the complex interaction between the immune system L-013950-01, ON-TARGETplus Human SAMHD1 siRNA and tumour cells in the tumour microenvironment (TME) have Smartpool, Dharmacon, Cultek, Spain) were mixed with emerged, suggesting that ovarian cancer initiation and development Lipofectamine 3000 reagent (ThermoFisher) at a final concentration as well as immune infiltration into the tumour may be influenced by a of 100 nM and let stand for 20 min. Then, lipofectamine-siRNA complex chemokine-signalling network (5, 6). However, the factors complexes were mixed with 1.6x105 cells and seeded in 24-well plates leading to this anti-tumor immune activation are still under in the absence of serum, using OPTIMEM medium (Invitrogen). investigation. Some recent works show that damaged cancer cells After 24h, medium with serum was added and left untreated 24h can induce anti-tumoral innate immunity by releasing nucleic acids more. 48h after transfection, downregulation of gene expression was that are detected by nucleic acid-sensing receptors (7). In addition, it assessed at RNA (1x105 cells were used) and protein (8x105 cells were has also been described that cytosolic DNA sensing and subsequent used) level. A non-targeting siRNA (siNT, D-001810-10, ON- stimulation of innate immunity might represent a relevant pathway in TARGETplus Non-targeting Pool, Dharmacon, Cultek) was used as anti-tumoral immune response through type I interferon (IFN) a control for putative off-targets effects in all experiments. production (8). Thus, a better understanding of nucleic acid sensing pathways, including upstream and downstream regulators, is key for a better understanding of ovarian cancer immunity, the identification Generation of SAMHD1 knock-out cell lines of novel putative therapeutic targets and the development of new potential therapeutic approaches. T47D cells were obtained from Sigma-Aldrich-ECACC SAMHD1 (sterile alpha motif and histidine/aspartic acid domain- (European Collection of Authenticated cell cultures, 85102201-1VL) containing protein 1) is a cellular deoxynucleotide (dNTP) and grown in complete Dulbecco’s modified Eagle’s medium triphosphohydrolase that has been recently linked to tumour (DMEM, ThermoFischer) supplemented with 10% of heat- initiation and development, although with controversial findings, inactivated fetal bovine serum (FBS, ThermoFisher) and antibiotics either being recognized as a tumour suppressor in haematological 100 U/ml penicillin, 100 mg/ml streptomycin (Life Technologies) and cancers (9, 10) or tumour promoting in several solid tumours, maintained at 37°C in a 5% CO2 incubator. For generation of knock- including ovarian cancer (11). On the other hand, mutations in out (KO) cells, T47D cells were transfected with a plasmid expressing SAMHD1 are linked to a severe congenital autoinflammatory a CRISPR-Cas9 construct designed to disrupt the sequence disease known as Aicardi–Goutières syndrome (AGS) characterized corresponding to exon 5 of SAMHD1 gene that encodes for HD by a dysregulated interferon (IFN) signalling due to defects in self and domain (CRISPR-SAMHD1), as described previously (11). nonself nucleic acids recognition (12). Indeed, SAMHD1 has been recently proposed as a key regulator of cellular RNA homeostasis, demonstrating a relevant role in the recognition and buffering of Drugs and in vitro treatment immunogenic self RNAs, a process that regulates innate immune responses (13–15). Moreover, SAMHD1 has also been linked to DNA OVCAR3 cells were treated at indicated doses with the damage response, suggesting that it can influence anticancer therapy corresponding drugs for 24 hours or left untreated as a control. following DNA damage induction (16). Thus, here we evaluated the Lipopolisaccaride (LPS) and Polyinosinic–polycytidylic acid (Poly(I:C) role of SAMHD1 expression and function in ovarian cancer in vitro were purchased from Sigma-Aldrich (L2630-10MG and P1530-25MG, and in patient cohorts. respectively). Carboplatin was obtained from ThermoFisher (J60433.06). Frontiers in Immunology 02 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 RNA-Sequencing and library preparation quantitative RT-PCR and normalized to GAPDH mRNA expression using the DDCt method. Primers and DNA probes were TaqMan Cellular RNA was extracted from cells using the NucleoSpin RNA Gene expression assays from Life Technologies (DDX58, TaqMan II kit (Magerey-Nagel), as recommended by the manufacturer, Hs01061436_m1; IFIH1, TaqMan Hs00223420_m1; MB21D1 including the DNase I treatment step. RNA-sequencing samples were TaqMan Hs00403553_m1; TMEM173 TaqMan Hs00736955_g1; prepared in biological duplicates. After quality control check, RNA IL6, TaqMan Hs00174131_m1; SAMHD1, TaqMan library was constructed using Illumina TruSeq Stranded mRNA LT Hs00174103_m1; IL8, TaqMan Hs00174103_m1; TNF, TaqMan Sample Prep Kit and sequencing was performed using NovaSeq 6000 Hs00174128_m1; IL18, TaqMan Hs01038788_m1; CXCL10, System with 150 bp paired ends reads (Macrogen). Sequencing files can TaqMan Hs00171042; ISG15 TaqMan Hs00192713_m1). be accessed on gene expression omnibus repository (GSE215309). Transcriptomic analysis Western blot analysis Transcriptomic analysis was performed as implemented in the Cells were rinsed in ice-cold phosphate-buffered saline (PBS) and computational workflow for the detection of differentially expressed extracts were prepared in lysis buffer (50 mM Tris HCl pH 7.5, 1 mM genes and pathways from RNA-seq data (17). Reads were aligned to EDTA, 1 mM EGTA, 1 mM NaV 3 O 4 , 10 mM sodium b- the human GRCh38 (annotation NCBI_109.20200522) using glycerophosphate, 50 mM NaF, 5 mM sodium pyrophosphate, 270 HISAT2. Low-expression genes with at least one zero counts were mM sucrose and 1% Triton X-100) supplemented with protease filtered out and the remaining reads normalized with Relative Log inhibitor cocktail (Roche) and 1 mM phenylmethylsulfonyl fluoride. Expression (RLE) method as implemented in DESeq2 R library. Samples were electrophoresed in SDS-polyacrylamide gels and blotted Differential gene expression between the control and treatment onto nitrocellulose membranes. Blocked membranes were incubated groups was estimated with the DESeq2 Wald test. Sequencing files overnight at 4°C with the following antibodies: anti-human Hsp90 can be accessed on gene expression omnibus repository (GSE215309). (1:10000; 610418, BD Biosciences); anti-human GAPDH (1:10000; Gene Set Enrichment Analysis (GSEA) was performed on a pre- ab9485, Abcam); anti-SAMHD1 (1:2000; ab67820, Abcam); anti- ranked GSEA list based on Log 2FC values of differentially expressed Cleaved PARP1 (E51) (1:1000; ab32064, Abcam); anti-Cleaved genes (DEGs: Log2FC > 1, p-value < 0.05), against Molecular caspase 3 (D175) (5A1E) (1:1000; 9664, Cell Signaling); anti-IRF7 Signatures Database (MsigDB v7.4) “Reactome” gene-set. Weighted (1:1000; 4920; Cell Signaling); anti-MDA5 (1:1000; 5321; Cell enrichment statistics were based on 1000 permutations. Significantly Signaling); anti-RIG-I (1:1000; 3743; Cell Signaling); anti- enriched gene-sets with FDR adjusted q-value<0.1 were selected for phosphoSTAT1 (Y701) (1:1000; 9167, Cell Signaling); anti-IFITM2 Enrichment map visualization as previously described. Briefly, (1:1000; 13530, Cell Signaling); anti-b-Actin (1:1000; A5441, Sigma). enrichment files were inputted into the Enrichment Map app After washing, the membranes were incubated with a secondary within the Cytoscape program for visualization. Parameters were conjugated horseradish peroxidase-conjugated secondary antibodies set at default values (node cutoff FDR Q value 0.1; Jaccard Overlap for 1h at room temperature and then revealed with SuperSignal West combined coefficient cutoff 0.375, k-constant 0.5). Nodes were Pico Chemiluminescent Substrate (Pierce Chemical). manually laid out and combined into a common biological process for clarity using the AutoAnnotate app. Patient cohort Human cytokine network array Total RNA was extracted using the NucleoSpin RNA II kit A cohort of 22 ovarian cancer patients was collected from the (Macherey-Nagel), as recommended by the manufacturer, including Medical Oncology department of our hospital. Ovarian cancer the DNase I treatment step. Reverse transcription was performed diagnosis occurred between 1991 and 2018. All patients had been using the PrimeScript ™ RT-PCR Kit (Takara) following primarily treated with debulking surgery plus platinum-based manufacturer instructions. Cytokine expression was evaluated by chemotherapy. The study was conducted under the ethics principles using the commercial TaqMan Human Cytokine Network array of the Declaration of Helsinki and approved by the Research and (4414124, ThermoFisher) which included primers and probes for Ethics Committee of Hospital Germans Trias i Pujol. Samples were 28 cytokine network associated genes and 4 endogenous control obtained from the Biobank of the Institut d’Investigació Germans genes. Relative expression of the distinct cytokine genes was Trias i Pujol. All patients provided written informed consent. measured by two-step quantitative RT-PCR and normalized to Variables studied: age at diagnosis, histologic subtype (high grade GAPDH expression by using the DDCt method. serous versus others), BRCA genes status (pathologically mutated versus wild type, variants of unknown significance or unknown), SAMHD1 immunochemistry (positivity defined by cellular positivity Quantitative RT-polymerase chain reaction 25%, as previously performed in AML cancers (11), progression free survival (PFS, defined as the time between first treatment and Total RNA was extracted and reverse transcribed as above progression or death, whatever occurs first), overall survival (OS, described. mRNA levels of all genes were measured by two-step defined as time from diagnosis to death from any cause). Frontiers in Immunology 03 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 Construction of tissue microarrays lines, previously developed in our group (11), to perform a whole and immunohistochemistry transcriptome profiling (Figure 1C; Supplementary Figure 1). Hierarchical clustering of wild-type and SAMHD1-KO cell clones TMA were constructed using a TMA workstation MTA-1 using the union of all differentially expressed genes (DEGs) revealed (Beecher instruments) and 3 different areas/tumor were selected distinct genetic signatures among them, while distinct knock-out and included (cylinders of 0.6 mm in diameter of each block of clones presented more similar signatures, compared to wild-type paraffin-embedded tissue). Then, TMA was cut in 5 micrometers (Figure 1C). To identify pathways specifically affected by the sections for subsequent analysis. SAMHD1 expression was evaluated downregulation of SAMHD1 in breast cancer cells, we performed by immunohistochemistry (1:200, polyclonal rabbit anti-SAMHD1 gene-set enrichment analysis (GSEA) using the Reactome gene-sets antibody, cat. no. 12586-1-AP, Proteintech) in an automated (Figure 1D). Overall, SAMHD1 knockout induced a global detection system (Ultraview, Ventana 9 after antigen retrieval), as downregulation of several signalling pathways, especially at previously reported (11). Evaluation of SAMHD1 expression in the transcription and RNA processing level. More interestingly, TMA sections was performed blinded by experienced pathologists downregulation of the cytosolic sensors of pathogen associated and the percentage of SAMHD1 positive tumor cells was recorded. molecular patterns (PAMPs) was also observed in SAMHD1-KO Independent triplicate evaluations were performed for each tumor. cells, suggesting that downregulation of SAMHD1 may influence Then, tumors were classified as SAMHD1 positive or negative, immune signalling and response in vitro (Figures 1D, E). Although depending on the percentage of SAMHD1 stained cells, which was transcriptomic data may vary between cell types, our results in breast arbitrarily defined, being SAMHD1 positive cases those with cellular cancer cells are also in accordance with previous transcriptomic data positivity ≥25%, as previously performed (11). Histopathological from monocytic cells where SAMHD1-KO clones showed unit of Hospital Germans Trias i Pujol performed all the dysregulation of several immune signalling pathways, such as RIG-I immunohistochemical analyses. like receptors, IFN and cytokine signalling pathways (18) and thus, prompted the evaluation of similar pathways in ovarian cancer models. Statistical analysis Considering the key role of SAMHD1 in the induction of IFN- mediated immune activation derived from its role in DNA damage Experimental data were analysed with the PRISM statistical repair (16, 19), together with reported deficiencies in nucleic acid package and expressed as mean ± SD of at least three independent sensing and subsequent loss of innate immune activation in ovarian experiments performed in duplicate. p-values were calculated using cancer (20), we focused our attention on innate immune response and an unpaired, two-tailed, t-student test. cytosolic pattern recognition receptors (PRRs). First, we evaluated Clinical variables were analysed using the SPSS statistical package. changes in the most common RNA and DNA PRR (Figure 2A). Descriptive statistics were medians and percentages, as appropriate. Interestingly, significant upregulation of both RNA sensors RIG-I, Correlation of SAMHD1 immunostaining positivity with clinical encoded by the DDX58 gene and MDA5, encoded by the IFIH1 gene characteristics was studied with the Pearson Chi-square test (2- was observed in siSAMHD1 cells, whereas DNA sensors cGAS tailed). Median times for PFS and OS were estimated using the (MB21D1) and STING (TMEM173) expression did not change Kaplan–Meier method and reported with their confidence intervals upon SAMHD1 downregulation (Figure 2A). In addition, when (CI) at the 95% level. Log rank was used to compare Kaplan–Meier ovarian cancer cells were exposed to LPS, known to recognize and Curves. Cox analyses were used to estimate the effect of different activate TLR, or Poly(I:C), known to activate the cytosolic RNA variables on survival outcomes (hazard ratios). helicases as RIG-I and MDA-5 (21), induction of IFN-stimulated genes (ISG) was only observed with poly(I:C) treatment, further supporting the more prominent role of RNA sensors in IFN- Results mediated response in ovarian cancer cells (Supplementary Figure 2). Then, we evaluated a comprehensive set of 28 cytokine To determine the contribution of SAMHD1 in ovarian cancer, we associated genes included in the commercial TaqMan Human effectively downregulated SAMHD1 expression in the ovarian cancer Cytokine Network array, finding significant transcriptional changes cell lines OVCAR3 and SKOV3, leading to a 70-80% reduction in associated to SAMHD1 depletion in IFNA7, IFNB1, IL16, IL18, IL4, SAMHD1 RNA and protein expression levels (Figures 1A, B). We IL6, IL8, LTA and TNF (Figure 2B). These findings were confirmed in have previously shown that SAMHD1 knockout breast cancer cells independent experiments, indicating an increased IFN-induced presented increased DNA damage and apoptosis, an effect that was signalling upon SAMHD1 downregulation (IL6 p=0.0259, IL8 enhanced upon platinum-based treatment (11). In SKOV3 and p=0.0173 and TNF p=0.023, respectively) (Figure 2C, upper panel). OVCAR3 ovarian cancer cell lines, SAMHD1-depleted cells also Interestingly, further evaluation of additional innate immune showed increased expression of the apoptotic markers, cleaved activation pathways showed similar results, i.e., expression of IFN- PARP and cleaved caspase-3 expression, although differences were stimulated genes (ISG) as CXCL10 (p=0.0327) and ISG15 were also not statistically significant (Figure 1B), indicating the existence of upregulated in SAMHD1-depleted cells (Figure 2C, bottom panel) additional mechanisms in contrast to previous data (11). and SAMHD1 knockdown also induced increased protein expression To investigate the signalling pathways affected by SAMHD1 of the PRR, MDA5 and RIG-I, IRF7 transcription factor and IFN- depletion, we took advantage of SAMHD1-KO breast cancer cell induced transmembrane protein IFITM2 as well as enhanced Frontiers in Immunology 04 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 A B C D E FIGURE 1 Depletion of SAMHD1 enhances apoptosis and regulates innate immune response. (A) Effective SAMHD1 downregulation by RNA interference in ovarian cancer cell lines. siRNAs targeting SAMHD1 gene (siSAMHD1) were transfected into SKOV3 cells (up) and OVCAR-3 cells (bottom). Gene expression was evaluated by RT-qPCR. (B) SAMHD1 knockdown induces apoptosis in ovarian cancer cells. Expression of apoptotic markers, cleaved PARP and cleaved Caspase 3 proteins, was measured by western blot in siRNA treated OVCAR3 (upper panel) and SKOV3 (bottom panel) cell lines. Representative western blot (left) and quantification (right) showing specific siRNA-mediated inhibition of SAMHD1 and increased cleaved PARP and cleaved caspase 3 expression. (C) Heatmap representation of gene expression changes in SAMHD1-KO and SAMHD1-WT cells. Heatmap was generated by unsupervised hierarchical clustering of significantly expressed genes (normalized enrichment score NES; p < 0.05). (D) Reactome Gene set enrichment map of significantly enriched pathways for SAMHD1-KO cells. Reactome Gene set clusters are annotated, and nodes manually laid out for clarity. Node size represents number of genes, node color represents significance (NES), and edge thickness represents number of shared genes. (E) Enrichment plot of the reactome cytosolic sensors of pathogen associated DNA geneset. Profile of the running ES Score of this geneset supports downregulation signature of cytosolic sensors and signalling pathways. Significantly down- or up-regulated genes (C) and gene sets (D) are highlighted in blue or red, respectively. *p<0.05; **p<0.005; ***p<0.0005. phosphorylation of STAT1 (Figure 2D), all suggestive of enhanced Median age at diagnosis was 63.00 years (min-max 51-82 years), the IFN-mediated inflammation. most frequent histology was high-grade serous subtype (n=17, In view of these findings, the role of SAMHD1 was also evaluated 77.3%), and 4 patients were known to harbour germline in a cohort of 22 ovarian cancer patients, previously described (11). pathologic BRCA1/2 mutations (18.2%, all of them with high-grade Frontiers in Immunology 05 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 A B C D FIGURE 2 SAMHD1 knockdown modulates RLR (RIG-I like receptor) expression and innate immune signaling. (A) Gene expression of DNA sensors (MB21D1 encoding cGAS protein, TMEM173 encoding STING protein) and RNA sensors (DDX58 encoding RIG-I protein, IFIH1 encoding MDA5 protein), upon SAMHD1 knock-down. (B) SAMHD1 knockdown induces proinflammatory cytokine expression. Heatmap showing fold change increase expression in siSAMHD1 cells compared to non-targeting control, evaluated using the TaqMan Human Cytokine Network array. (C) Gene expression of distinct IFN- stimulated genes (ISG) in SAMHD1 knockdown cells. Increased IL6, IL8, TNF, IL18, CXCL10 and ISG15 expression upon SAMHD1 depletion was confirmed in additional experiments. (D) Representative western blot (left) and quantification (right) showing increased protein expression of distinct IFN-stimulated proteins in SAMHD1 knockdown cells. Protein expression of RNA sensors MDA5 and RIG-I, transcription factor IRF7, phosphorylation of STAT1 and IFITM2 was determined by western blot. Mean +/- SD of at least three independent experiments is shown. *p<0.05; ***p<0.0001. siNT, non-targeting siRNA used as control; siSAMHD1, siRNA specifically targeting SAMHD1. ns, non-significant. serous histology). Median progression free survival (PFS) and evaluated by immunohistochemistry in ovarian cancer archival median overall survival (OS) of the whole sample were 16.00 biopsies (Figure 3A), that were classified as SAMHD1 positive or months (95% CI 5.66-26.34), and 66.00 months (95% CI 33.03- negative depending on the percentage of SAMHD1 stained tumoral 98.96), respectively (Table 1). SAMHD1 expression was re- cells [positivity was arbitrarily defined as those with cellular positivity Frontiers in Immunology 06 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 TABLE 1 Baseline characteristics of the clinical sample and survival outcomes. Descriptive statistics (%) Median PFS p-value Median OS p-value (months, 95%CI) (months, 95%CI) Age at diagnosis (years) 63 NA NA – (51-82 years) Histologic subtype 0.040 0.144 High grade serous 17 (77.30%) 16.00 (13.02-18.98) 66.00 (37.59-94.40) Others* 5 (22.70%) 52.00 (0.00-112.12) 157.00 (0-359.77) BRCA status 0.372 0.857 Mut** 4 (18.20%) 12.00 (4.16-19.84) 113.00 (0-238.75) Wild type/UK*** 18 (81.80%) 23.00 (6.37-39.63) 62.00 (34.97-89.03) SAMHD1 0.010 0.040 <25 7 (31.80%) 52.00 (0.00-123.86) 157.00 (0.00-343.66) >25 15 (68.20%) 15.00 (9.95-20.05) 62.00 (26.83-97.17) Total 22 16.00 (5.66-26.34) 66.00 (33.03-98.96) – *4 clear cells and 1 low grade serous tumour; **Pathologically mutated; *** UK, unknown. NA, not-applicable. ≥25%, based on previous reported thresholds (11)]. As previously prognosis in ovarian cancer, a mechanism that is regulated by reported, SAMHD1 positivity correlated with high-grade serous SAMHD1, as demonstrated in vitro (Supplementary Figure 4). histology (p=0.007), but not with BRCA1/2 status (p=0.144). More importantly, SAMHD1 expression showed a statistically significant effect on survival outcomes. Median progression free survival (PFS) of Discussion the SAMHD1-high expression subgroup was statistically significantly shorter than those of the SAMHD1-low expression subgroup (15.00 Ovarian cancer (OC) is the third most diagnosed gynaecologic [95% CI 9.95-20.05] vs. 52.00 [95% CI 0.00-123.86], p=0.010) cancer in Europe (23). However, despite the recent advances in (Figure 3B, upper panel). Median overall survival (OS) of the surgery and chemotherapy, it remains the most lethal gynaecologic SAMHD1-high expression subgroup was also shorter than those of cancer. In this context, a better understanding of the induction and the SAMHD1-low expression subgroup (62.00 [95% CI 26.83-97.17] modulation of the anti-tumor innate immunity is key to develop new vs. 157.00 [95% CI 0.00-343.66], p 0.040) (Figure 3B, bottom panels). therapies. Here, we provide evidence of the involvement of SAMHD1 Hazard ratio for PFS was 4.54 (95% CI 1.27-16.23, p=0.020), and in the induction and modulation of anti-tumoral immunity in hazard ratio for OS was 3.564 (95% CI 0.99-12.56, p=0.052), favouring ovarian cancer. the low-expression subgroup. These differences remained statistically Our evaluation of innate immune activation pathways revealed an significant when individually analysing the BRCA wild type or increased IFN-induced signalling upon SAMHD1 downregulation, unknown subgroup (Figure 3B, lower panel). concomitant with an upregulation of MDA5 and RIG-I RNA sensors. The contribution of the observed differences in PFS and OS Indeed, in agreement with our data, SAMHD1-deficiency and depending on SAMHD1 expression might be partially affected by subsequent accumulation of endogenous RNA substrates is a cause previous platinum-based chemotherapy which was common to all of type I interferonopathies, characterized by an upregulation of patient cohort. However, in vitro exposure to carboplatin did not distinct IFN-stimulated genes (ISGs) (20, 24). Moreover, it has also induce a differential response in SAMHD1 knock-down cells, been shown that chronic interferon response in SAMHD1-KO mice compared to non-targeting control (Supplementary Figure 3), was driven by the MDA5 pathway in close concordance with our data suggesting that SAMHD1 expression might be the main contributor in ovarian cancer cells (25) and further supporting the idea that to the observed correlation with low-SAMHD1 and better SAMHD1 depletion is able to enhance innate immune activation and survival outcomes. inflammation in cancer cells, a process that might have an important Overall, clinical data allowed us to propose SAMHD1 as a impact on ovarian cancer clinical outcome. In view of the promising prognostic marker in ovarian cancer, whose function might pre-clinical data, we evaluated the role of SAMHD1 in 22 ovarian putatively induce antitumoral innate immune response, as cancer patients. Interestingly, SAMHD1 expression was significantly demonstrated in vitro in cell lines (Figure 3C). Interestingly, associated with tumor histology, being high-grade serous histology exploring the ovarian cancer proteome using TCGA ovarian tumors those presenting the highest proportion of SAMHD1, transcriptomics data obtained from Human Protein Atlas database as previously reported in other cancer types (11). Although no (www.proteinatlas.org) (22), we observed that high expression of significant association between SAMHD1 and BRCA1/2 status was several innate immune activation hallmark genes was associated to found, all BRCA1/2 mutated patients showed high expression of improved ovarian cancer survival, supporting the idea that SAMHD1, suggesting a correlation between these two variables. In upregulation of innate immune response is linked to better this sense, it has been described that unprotected stalled replication Frontiers in Immunology 07 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 A B C FIGURE 3 SAMHD1 expression is a prognostic factor in ovarian cancer patients. (A) Representative microscopy images of SAMHD1 expression in paraffin-embedded ovarian tumour biopsies. Images on the left represent SAMHD1-low expressing tumours and on the right positive expressing tumours. SAMHD1 expression was evaluated by immunohistochemistry in tumour samples. High expression of SAMHD1 observed in lymphocytic cells infiltrating in the tumours was used as a positive control of immunohistochemistry for negative or low expressing biopsies. (B) Kaplan-Meier curves of progression-free survival (PFS, top) and overall survival (OS, bottom) stratified according to SAMHD1 status, i.e, SAMHD1-low (SAMHD1 expression below 25% in cancer cells, red lines) or SAMHD1-high (SAMHD1 expression equal or above 25%, black lines). Median survival times with 95% CI of both groups are shown. Log rank test was used to test the significance and censored patients are indicated by a vertical line. (C) Working model of the innate immune signalling pathways triggered by the absence or the low expression of SAMHD1 in ovarian cancer. Low SAMHD1 expression induces increased levels of distinct IFN-stimulated genes in ovarian cancer cells, subsequently leading to increased antitumoral immunity and better prognosis in patients. Abbreviations are as follows: MDA5, melanoma differentiation-associated protein 5; RIG-I, retinoic acid-inducible gene-I; IFN, interferon; IFNAR, interferon a/b receptor 1; JAK, Janus kinase 1; TYK2, tyrosine kinase 2; STAT, signal transducer and activator of transcription; P, phosphoryl group; ISG, interferon-stimulated genes; IL6, Interleukin 6; IL8, Interleukin 8; TNF, Tumour Necrosis Factor; CXCL10, C-X-C motif chemokine ligand 10. Created with BioRender. Frontiers in Immunology 08 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 forks in SAMHD1-deficient cells can produce cytoplasmic DNA (26), BRCA1/2 abrogation (34), we suggest that TILs recruitment processes which could mimic DNA fragments generated from stalled could be exploited to develop potential novel immunotherapy treatment. replication forks in BRCA1/2-mutant cells during the S-phase or DNA damage repair. Thus, high expression of SAMHD1 could represent a compensatory mechanism in response to the Data availability statement inflammatory response associated with the release of cytosolic DNA as a consequence of defective DNA repair secondary to BRCA1/2 The data presented in the study are deposited in the omnibus mutations. More interestingly, SAMHD1 positivity was significantly repository repository, accession number GSE215309. Other original associated with poorer prognostic clinical outcomes, including contributions presented in the study are included in the article/ decreased median progression free survival (PFS) and median Supplementary Material. Further inquiries can be directed to the overall survival (OS). Overall, clinical data allow us to propose corresponding authors. SAMHD1 as a prognostic marker in ovarian cancer, whose function might putatively induce antitumoral proinflammatory innate immune response, as demonstrated in vitro in cell lines. Ethics statement Interestingly, exploring the ovarian cancer proteome using TCGA transcriptomics data obtained from Human Protein Atlas database The studies involving human participants were reviewed and (www.proteinatlas.org) (22), we observed that high expression of approved by Hospital Germans Trias i Pujol. The patients/ several innate immune activation hallmark genes was associated to participants provided their written informed consent to participate improved ovarian cancer survival, supporting the idea that in this study. upregulation of innate immune response is linked to better prognosis in ovarian cancer, a mechanism that is regulated by SAMHD1, as demonstrated in vitro (Supplementary Figure 4). Author contributions Although inflammation and cancer onset and progression are closely interrelated, in ovarian tissue, inflammation is a double-edge sword that LG-C and EF were responsible of all experimental tasks, including has been associated with either tumour progression or suppression (27, biological analysis in cell lines and immunohistochemistry of patient 28), highlighting the importance of characterizing specific inflammatory samples and manuscript drafting. IJE performed RNAseq data pathways. Along the same line, increasing amounts of data are pointing analysis. IT, AM-C and MR were responsible of patient selection, towards the importance of nucleic acid-sensing pathways in cancer clinical data collection and clinical data analysis. AB-P and ER-M patients progression (29). In concordance with our data, several helped in obtaining experimental data, contributed to data analysis evidences indicate the capability of tumour cells to generate and were responsible of methodological set-up. BC was responsible of inflammatory factors, representing key signals that determine the study conception and writing review. MR, MM and EB were cross-talk between tumour and immune cells and ultimately affect the responsible of formal analysis, writing and review. EB and MM had mechanisms of immunosuppression by which tumour cells circumvent substantial contribution to the conception, conceptualization, formal innate and adaptive immune responses (30). In our study, we show that analysis, manuscript writing and final review. All authors contributed depletion of SAMHD1 in ovarian cancer cells leads to upregulation of to the article and approved the submitted version. RNA helicases and also several IFN-stimulated genes, as cytokines and chemokines, suggesting an activation of innate immune signalling pathways that could trigger an inflammatory response in the tumour Funding site (Figure 2C) and may ultimately affect patient prognosis. In fact, it has been largely described that some cytokines can act as potent This research was supported by grants from Instituto de Salud chemoattractant for different cell subsets; for example, IL6 and IL8 are Carlos III, Fondo de Investigació n Sanitaria (FIS) PI17/00624 and direct mediators of T cell migration (31, 32). Moreover, the presence of T PI21/00642 cofinanced by FEDER. EF and EB are research fellows cells in ovarian tumours has been associated with a survival advantage in from ISCIII-FIS (CM20/00027 and MSII19/00012, respectively). LG- distinct studies across diverse patient cohorts (33). However, whether C is a research fellow from Generalitat de Catalunya AGAUR. IJE is a tumoral cells are capable of effectively initiating an antitumoral IFN- research fellow from la Caixa (LCF/BQ/IN18/11660017) cofunded by mediated response in vivo remains to be characterized. the European Union’s Horizon 2020 research and innovation In conclusion, our data provides evidence of the involvement of program under the Marie Sklodowska-Curie grant agreement SAMHD1 in ovarian cancer, as previously reported in other cancer types No. 713673. (11). Next steps should aim to validate these results in larger series, as well as prospectively explore the correlation between SAMHD1 expression, innate immune response, and inflammatory chemokines directly in Acknowledgments ovarian cancer patients. Moreover, given that the activation of innate immunity in response to the inactivation of SAMHD1 described in this We thank the Biobank of the Institut d’Investigació Germans work, together with the described innate immune response triggered by Trias i Pujol (IGTP) for sample management. Frontiers in Immunology 09 frontiersin.org Gutiérrez-Chamorro et al. 10.3389/fimmu.2023.1112761 Conflict of interest organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or The authors declare that the research was conducted in the claim that may be made by its manufacturer, is not guaranteed or absence of any commercial or financial relationships that could be endorsed by the publisher. construed as a potential conflict of interest. Publisher’s note Supplementary material The Supplementary Material for this article can be found online at: All claims expressed in this article are solely those of the https://www.frontiersin.org/articles/10.3389/fimmu.2023.1112761/ authors and do not necessarily represent those of their affiliated full#supplementary-material References 1. World Cancer Research Fund. World cancer research fund international. Ovarian 18. Oh C, Ryoo J, Park K, Kim B, Daly MB, Cho D, et al. A central role for PI3K-AKT Cancer statistics (2022). Available at: https://www.wcrf.org. signaling pathway in linking SAMHD1-deficiency to the type I interferon signature. Sci 2. The cancer genome atlas research network. integrated genomic analyses of ovarian Rep (2018) 8(1):84. doi: 10.1038/s41598-017-18308-8 carcinoma. Nature (2011) 474:609–15. doi: 10.1038/nature10166 19. Kretschmer S, Wolf C, König N, Staroske W, Guck J, Häusler M, et al. SAMHD1 3. Xie H, Wang W, Xia B, Jin W, Lou G. Therapeutic applications of PARP inhibitors prevents autoimmunity by maintaining genome stability. Ann Rheum Dis (2015) 74(3): in ovarian cancer. BioMed Pharmacother (2020) 127:110204. doi: 10.1016/j.biopha. e17–7. doi: 10.1136/annrheumdis-2013-204845 2020.110204 20. de Queiroz NMGP, Xia T, Konno H, Barber GN. Ovarian cancer cells commonly 4. Palaia I, Tomao F, Sassu CM, Musacchio L, Benedetti Panici P. Immunotherapy for exhibit defective STING signaling which affects sensitivity to viral oncolysis. Mol Cancer ovarian cancer: Recent advances and combination therapeutic approaches. Onco Targets Res (2019) 17(4):974–86. doi: 10.1158/1541-7786.MCR-18-0504 Ther (2020) 13:6109–29. doi: 10.2147/OTT.S205950 21. Kato H, Takeuchi O, Sato S, Yoneyama M, Yamamoto M, Matsui K, et al. 5. Turner TB, Buchsbaum DJ, Straughn JM, Randall TD, Arend RC. Ovarian cancer Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses. Nat and the immune system — the role of targeted therapies. Gynecol Oncol (2016) 142 (2006) 441(7089):101–5. doi: 10.1038/nature04734 (2):349–56. doi: 10.1016/j.ygyno.2016.05.007 22. Uhlen M, Zhang C, Lee S, Sjöstedt E, Fagerberg L, Bidkhori G, et al. A pathology 6. Mlynska A, Salciuniene G, Zilionyte K, Garberyte S, Strioga M, Intaite B, et al. atlas of the human cancer transcriptome. Science (2017) 357(6352):eaan2507. doi: Chemokine profiling in serum from patients with ovarian cancer reveals candidate 10.1126/science.aan2507 biomarkers for recurrence and immune infiltration. Oncol Rep (2019) 41(2):1238–52. 23. Ferlay J, Ervik M, Lam F, Colombet M, Mery L, Piñeros M, et al. Global cancer doi: 10.3892/or.2018.6886 observatory. In: Cancer today. Available at: https://gco.iarc.fr/. 7. McWhirter SM, Jefferies CA. Nucleic acid sensors as therapeutic targets for human 24. Crow YJ, Manel N. Aicardi–goutières syndrome and the type I interferonopathies. disease. Immunity. (2020) 53(1):78–97. doi: 10.1016/j.immuni.2020.04.004 Nat Rev Immunol (2015) 15(7):429–40. doi: 10.1038/nri3850 8. Corrales L, Matson V, Flood B, Spranger S, Gajewski TF. Innate immune signaling 25. Schumann T, Ramon SC, Schubert N, Mayo MA, Hega M, Maser KI, et al. and regulation in cancer immunotherapy. Nat Publ Gr (2016) 27:96–108. doi: 10.1038/ Deficiency for SAMHD1 activates MDA5 in a cGAS/STING-dependent manner. J Exp cr.2016.149 Med (2023) 220(1):e20220829. doi: 10.1084/jem.20220829 9. Herold N, Rudd SG, Sanjiv K, Kutzner J, Myrberg IH, Paulin CBJ, et al. With me or 26. Coquel F, Silva M-J, Té cher H, Zadorozhny K, Sharma S, Nieminuszczy J, et al. against me: Tumor suppressor and drug resistance activities of SAMHD1. Exp Hematol SAMHD1 acts at stalled replication forks to prevent interferon induction. Nat (2018) 557 (2017) 52:32–9. doi: 10.1016/j.exphem.2017.05.001 (7703):57–61. doi: 10.1038/s41586-018-0050-1 10. Clifford R, Louis T, Robbe P, Ackroyd S, Burns A, Timbs AT, et al. SAMHD1 is 27. Zhao H, Wu L, Yan G, Chen Y, Zhou M, Wu Y, et al. Inflammation and tumor mutated recurrently in chronic lymphocytic leukemia and is involved in response to DNA progression: signaling pathways and targeted intervention. Signal Transduct Target Ther damage. Blood (2014) 123(7):1021–31. doi: 10.1182/blood-2013-04-490847 (2021) 6(1):263. doi: 10.1038/s41392-021-00658-5 11. Felip E, Gutié rrez-Chamorro L, Gó mez M, Garcia-Vidal E, Romeo M, Morá n T, 28. Savant S, Sriramkumar S, O’Hagan H. The role of inflammation and inflammatory et al. Modulation of DNA damage response by SAM and HD domain containing mediators in the development, progression, metastasis, and chemoresistance of epithelial deoxynucleoside triphosphate triphosphohydrolase (SAMHD1) determines prognosis ovarian cancer. Cancers (Basel) (2018) 10(8):251. doi: 10.3390/cancers10080251 and treatment efficacy in different solid tumor types. Cancers (Basel) (2022) 14(3):641. 29. Yan-Fei H, Han Y, Yan-Ting Z, Hui Y, Yu-Qing Y, Ipsita P, et al. Dysregulation in doi: 10.3390/cancers14030641 nucleic acid-sensing pathway genes is associated with cancer patients’ prognosis. Cancer 12. Rice GI, Bond J, Asipu A, Brunette RL, Manfield IW, Carr IM, et al. Mutations Sci (2020) 111(7):2212–22. doi: 10.1111/cas.14450 involved in aicardi-goutières syndrome implicate SAMHD1 as regulator of the innate 30. Burkholder B, Huang R-Y, Burgess R, Luo S, Jones VS, Zhang W, et al. Tumor- immune response. Nat Genet (2009) 41(7):829–32. doi: 10.1038/ng.373 induced perturbations of cytokines and immune cell networks. Biochim Biophys Acta - 13. Schott K, Majer C, Bulashevska A, Childs L, Schmidt MHH, Rajalingam K, et al. Rev Cancer (2014) 1845(2):182–201. doi: 10.1016/j.bbcan.2014.01.004 SAMHD1 in cancer: curse or cure? J Mol Med (2022) 100(3):351–72. doi: 10.1007/s00109- 31. Weissenbach M, Clahsen T, Weber C, Spitzer D, Wirth D, Vestweber D, et al. 021-02131-w Interleukin-6 is a direct mediator of T cell migration. Eur J Immunol (2004) 34(10):2895– 14. Chen S, Bonifati S, Qin Z, St. Gelais C, Wu L. SAMHD1 suppression of antiviral 906. doi: 10.1002/eji.200425237 immune responses. Trends Microbiol (2019) 27(3):254–67. doi: 10.1016/j.tim.2018.09.009 32. Taub DD, Anver M, Oppenheim JJ, Longo DL, Murphy WJ. T Lymphocyte 15. Maharana S, Kretschmer S, Hunger S, Yan X, Kuster D, Traikov S, et al. SAMHD1 recruitment by interleukin-8 (IL-8). IL-8-induced degranulation of neutrophils releases controls innate immunity by regulating condensation of immunogenic self RNA. Mol Cell potent chemoattractants for human T lymphocytes both in vitro and in vivo. J Clin Invest (2022) 82(19):3712–3728.e10. doi: 10.1016/j.molcel.2022.08.031 (1996) 97(8):1931–41. doi: 10.1172/JCI118625 16. Cabello-Lobato MJ, Wang S, Schmidt CK. SAMHD1 sheds moonlight on DNA 33. Hwang W-T, Adams SF, Tahirovic E, Hagemann IS, Coukos G. Prognostic double-strand break repair. Trends Genet (2017) 33(12):895–7. doi: 10.1016/ significance of tumor-infiltrating T cells in ovarian cancer: A meta-analysis. Gynecol j.tig.2017.09.007 Oncol (2012) 124(2):192–8. doi: 10.1016/j.ygyno.2011.09.039 17. Chen Y, Lun ATL, Smyth GK. From reads to genes to pathways: Differential 34. Reisländer T, Lombardi EP, Groelly FJ, Miar A, Porru M, Di Vito S, et al. BRCA2 expression analysis of RNA-seq experiments using rsubread and the edgeR quasi- abrogation triggers innate immune responses potentiated by treatment with PARP likelihood pipeline. F1000Res (2016) 5:1438. doi: 10.12688/f1000research.8987.1 inhibitors. Nat Commun (2019) 10(1):3143. doi: 10.1038/s41467-019-11048-5 Frontiers in Immunology 10 frontiersin.org