The discontinuity of eukaryotic genome requires that introns are spliced out and exons joined to form a continuous RNA transcript. However, certain exons deviate from this regulation resulting alternatively included in the mature mRNA. This process, known as alternative splicing (AS), allows yielding multiple transcripts from the same gene expanding the proteome diversity and complexity of the genome. This process is catalysed by the spliceosome, a ribonucleoprotein complex, assisted by auxiliary splicing factors
[89]. Genome-wide analyses have revealed how this process is highly coordinated during cell differentiation and tissue development
[90], proving to be a powerful means to finely tune gene expression during fundamental biological processes. Among organs, AS is particularly widespread in testis relying on activation of specific RNA processing programs that contributes to temporally regulate expression of genes essential for proper development of male gametes
[91][92][93][94][95]. This suggests that splicing abnormalities might contribute to the development of pathological conditions. Accordingly, aberrant expression of splicing factors, as SF1, QK1 and RBFOX RNA-binding proteins, influences TGCTs’ incidence
[96][97][98]. Furthermore, genome-wide profiling of TGCT datasets retrieved from The Cancer Genome Atlas revealed that numerous AS events are significantly associated with risk of disease progression
[99]. Collectively, these observations point out AS dysregulation as a potential key driver that could promotes acquisition of hallmark traits in TGCTs, underpinning specific oncogenic processes and/or contributing to adaptive resistance toward chemotherapeutic agents.
In TGCTs, a large number of apoptosis- and cell cycle-related factors have been found to be regulated via AS
[100][101]. The Ser/Thr-kinase NEK2 is a new key player regulating the interplay between splicing and cellular signalling
[100][102]. In TGCTs, overexpression of NEK2 has been documented
[103][104] and previous observations reported its activation during G2/M progression of male germ cells and its involvement in chromatin condensation during the meiotic divisions of mouse spermatocytes
[105][106]. Recently, identification of NEK2 splicing isoforms allowed the identification of a novel function for this kinase, adding another layer of complexity with regard to its oncogenic activity
[102]. NEK2 AS results in the differential expression of three isoforms. The canonical splice variant NEK2A results localized in the centrosome, and partially also in nucleus and cytoplasm, whereas NEK2B and NEK2C variants showed cytoplasmic and nuclear localization, respectively. In TGCTs, NEK2 was found enriched in the nucleus of several cancer cells, including testicular seminomas, where it interacts with and phosphorylates numerous splicing factors, including the oncogenic SR protein SRSF1. Notably, NEK2-mediated phosphorylation increases SRSF1 splicing activity toward its apoptotic target genes, promoting antiapoptotic variants of BCL-X, MKNK2 and BIN1. Accordingly, knockdown of NEK2 favoured pro-apoptotic splice variants, promoting apoptosis and sensitizing cells to chemotherapeutic treatment with cisplatin
[102]. Besides NEK2, immunohistochemistry studies on seminomas and normal testis tissues also reported the cellular mis-localizzation of MAD2, a mitotic factor that act as a component of spindle assembly checkpoint (SAC)
[107]. AS of MAD2 transcripts yields three isoforms, full length MAD2α, and MAD2β and MAD2γ, lacking exon 3 and exon 2 and 3, respectively. MAD2α was present in both the nucleus and the cytoplasm, while MAD2γ mainly localized to the nucleus and reduced mitotic arrest. Interestingly, in TGCTs patients, the overexpression of endogenous MAD2γ, but not MAD2α, was associated with resistance to cisplatin-based chemotherapy
[108][109]. Since Nek2 interacts with MAD2
[110] and SAC has emerged as a promising target for cancer therapy
[111], modulation of NEK2 activity and/or MAD2 AS could be exploited therapeutically in TGCTs patients displaying resistance to cisplatin-based chemotherapy. Accordingly, it has been shown that depletion of MAD2 induces apoptosis and restores sensitivity to cisplatin therapy in a cisplatin resistant lung cancer model
[112]. Another gene that undergoes AS modulation, representing a valuable therapeutic target in cisplatin-resistant TGCTs, is the member of the inhibitor of apoptosis protein (IAP) family, Livin
[100]. Although its expression in normal testicular tissue is still elusive
[100][113][114], in a large cohort of TGCT patients the expression of both Livin α and Livin β splice variants was found to be strongly related to the histological subtype, resulting in frequently expressed seminoma
[100]. Since Livin has been identified as a target for immune-mediated tumor destruction
[115], clarifying the involvement of these isoforms in drug resistance might provide a therapeutic option for cisplatin-resistant patients.
In addition to evasion of apoptosis, metabolic adaptation, also known as aerobic glycolysis (or the Warburg effect), is an emerging hallmark of cancer
[116]. Glycolytic key regulators have been found differentially expressed in TGCTs and associated with tumor aggressiveness
[117][118], highlighting the potential role of metabolic adaptation among plausible causative processes of tumor progression. The contribution of AS to metabolism reprogramming has been reported in several tumors
[119]. Notably, screening performed to find novel cancer-associated immunogenic gene products allowed the identification of four cancer-restricted splice variants of testis Lactate Dehydrogenase C (LDHC) in several cancers
[117]. Three splice variants skip exons encoding NAD binding domains and/or L-lactate dehydrogenase active sites, thus yielding aberrant LDHC proteins devoid of specific enzymatic activity, whereas one isoform encodes for a fully active enzyme
[117]. Lactate dehydrogenase catalyses the interconversion of lactate and pyruvate in the glycolytic pathway, and lactate has a pivotal role in spermatogenesis
[120] and exerts an antiapoptotic effect
[121]. Interestingly, a gene expression profile identified LDHC among genes differentially expressed in seminoma samples
[122]. Although the impact of LDHC variants has not been investigated in TGCTs, these protein isoforms might exhibit different functional properties in terms of substrate specificity that may be beneficial for the metabolic adaptation, survival and proliferation of tumor germ cells.
R-Loop
R-loops are three-stranded nucleic acid structures characterised by a DNA:RNA hybrid and a displaced single-stranded DNA that frequently form in connection with transcription. Their programmed formation occurs physiologically, and it contributes to important cellular processes including transcription initiation and termination, immunoglobulin class switching, replication of mitochondrial DNA and epigenetic modifications
[123][124][125]. Besides their role in normally replicating cells, a large body of evidence suggests that mis-regulated formation of R-loops occurs in cancer cells. Unscheduled formation of R-loops is associated with transcription elongation defects, hyper-recombination and DNA damage, all of which might contribute significantly to cancer-related genome instability
[125]. New technology, including the application of the S9.6 antibody, has led to the identification of R-loops interactome and knowledge of genome-wide distribution of R-loops
[126][127][128]. One such RNA:DNA hybrid binding protein is Senataxin
[128], a putative RNA:DNA helicase whose mutation is responsible for rare neurological disorders
[129]. As deduced from GEO expression data and experimental results in mouse
[130], Senataxin is highly expressed in testis, and its mutation in humans
[131] or deletion in mice
[132] causes germ cell arrest at pachynema, and unscheduled formation of R-loops in spermatocytes, likely as consequence of a conflict between transcription and meiotic recombination intermediates
[132]. R-loops also accumulate in proliferating cells of Setx-/- mice testes
[130][132], indicating a role for the helicase in resolving R-loops that occur in the germ cell lineage, due to the collision of the replication fork and the transcriptional machinery, as previously described in mitotic somatic cells
[132][133][134]. Notably, double strand breaks (DSBs) induced by topoisomerase I treatment augment R-loops accumulate in proliferating germ cells of the testis, in both wild type and Setx-/- mice
[130]. Since R-loops normally occur at pause sites during transcription
[135], this indicates that DNA lesions can stall the transcription machinery, which in turn causes R-loop accumulation. This is further confirmed by the accumulation of R-loops in mitotic germ cells of the testis in mice deleted for Atm or Tdp1, two DNA-damage response genes required for repair of DNA breaks
[136][137][138]. In this regard, it has been also demonstrated that defects in the homologous recombination (HR) proteins BRCA1 and BRCA2
[139][140], in the nucleotide excision repair (NER) proteins XPG and XPF
[141] and in the Fanconi anemia (FA) pathway
[142][143][144], lead to R-loop accumulation, thus indicating that several DNA repair pathways contribute to R-loop regulation
[145].
To date, there are no reports in literature investigating R-loop levels in testicular germ cell tumors. Thus, whether an alteration of formation, stabilization or resolution of R-loops associates with TGCTs development is unknown. According to the exquisite sensitivity of TGCTs to cisplatin-induced damage, we demonstrated that embryonal carcinoma TGCT cell lines that are sensitive to drug treatment have a reduced proficiency of DSBs repair by HR
[146]. In addition, analysis of NER protein expression in TGCT-derived cell lines revealed that levels of XPA, ERCC1 and XPF DNA repair proteins are reduced with respect to somatic tumor cells
[146][147], indicating that repair of DNA damage by NER might be compromised. Moreover, in a recent study from our laboratory, in which EC cell lines sensitive and resistant to cisplatin were compared, we demonstrated that cisplatin-sensitive cell lines have a reduced expression of FANCD2 with respect to cisplatin-resistant cell lines
[148]. Collectively these observations suggest that often, TGCTs have a reduced recombinative efficacy, which is one of the mechanisms that have been proposed to account for their unique sensitivity to DNA damage
[149]. Given the importance of HR, NER and FANC pathways’ deficiency in R-loop accumulation in mitotic cells, it is plausible that R-loops might arise in replicating germ-cell tumor cells, contributing to increase genome instability and tumor progression. In accordance with this hypothesis, studies on TGCT tissues aimed at evaluating the activation of the DNA-damage response (DDR) in TGCTs have shown that DDR is not activated in pre-invasive carcinoma in situ (CIS) lesions, while it was found to be activated in a subset of seminomas and in embryonal carcinomas at the invasive stage
[150]. This suggests that, at least in a subset of tumors, R-loops might arise along with persistent replication-associated DNA damage. Their accumulation might contribute to increased genome instability
[151] and perhaps tumor progression. Interestingly, by studying a series of nonseminomatous GCTs it has also been found that tumors resistant to cisplatin express low levels of the mammalian serine/arginine-rich protein-specific kinase 1 (SRPK1)
[152]. The latter, by phosphorylating the splicing factor SRSF1 promotes its subcellular nuclear localization
[153]. Given that inactivation of SRSF1 has been demonstrated to promote genome instability via formation of R-loops
[154], one can speculate that low expression of SRPK1 in GCTs may favour genome instability, promoting genome rearrangements leading to the acquisition of resistance to cisplatin.
As the role of R-loops in tumorigenesis will advance in the years to come, and the interactome of DNA:RNA hybrid will expand, it will be interesting to investigate further their potential role in the pathogenesis of TGCTs.