On the potential origin of the zygote-like cancer stem cell with a focus on fusion for cell rescue.

The establishment of alternative concepts to explain the cell of origin of cancer is warranted, as survival rates in last stage cancer have seen only marginal improvement over the past century ( Liu, 2022 ). The emergence of a new tide of genomics data from cancer genome analysis in recent years points toward copy number and genomic structural changes that are inconsistent with the conventional somatic mutation theory ( Campbell et al., 2020 ). Next-generation sequencing data indicates that in tumorigenesis, the genomic rearrangements are highly improbable to have accumulated over time and instead holds that nearly all mutations occur during a one-off cellular crisis, promoting the evolution toward cancer ( Stephens et al., 2011 ). Following the avalanche of big genomic data, many parallels between development and the gametogenic program induction in cancer have been drawn in the last decades. The identification of cancer-testis - and trophoblast specific - antigens in tumors have generated provocative concepts, viewing cancer as a “somatic pregnancy” ( Old, 2007 ) or “chaotic embryogenesis”- based on McClintock’s heredity ( Liu, 2022 ). In an elegant tumor’s gametogenesis-related model, Liu et al. provides consistent arguments based on experimental data suggesting that the formation of tumor cells is in some way similar to fertilization, with PGC-like cells likely being an origin for somatic cancer initiation and metastasis via stage arrest resulting in tumors with immature tissues, and parthenogenetic activation of primary oocyte at meiotic arrest causing tumors with mature tissues ( Liu C. et al., 2022 ). Another mechanism of tumor formation holds that somatic cells can react to stress by increasing cell size, restructuring the genome for reproduction or neoplastic transformation and explaining the occurrence of polyploid giant cancer cells (PGCCs) in advanced tumors. This genomic restructuring not only creates a complex genotype but also activates an embryonic program forming the blastomere- or blastocyst-like embryo, since they resemble the pre-implantation embryo in morphology, gene expression, and the ability to generate germ cell tumors ( Liu, 2022 ). While these intriguing theories rely on the spontaneous activation of alternative developmental programs to generate pluripotent-like cellular states without zygotic fusion events, our concept distinguishes this model from alternative no-fusion mechanisms holding that activation of the developmental programme implies two cells involved in a failed fusion-driven cell rescue when the responders to cell stress signals are the primitive germline related stem cells. A consequence of such failed attempt of cell rescue by unselective gametic-like fusion could lead to recapitulation of the zygotic program in tumors and metastasis which seems to follow specific homing patterns to predefined niche - as in primitive cell migration during development and fetal-maternal microchimerism-rather than just random spread, in a tumor-host microchimerism-like manner. The migratory events of distant metastatic spread are reflected in the phenomenon of fetal-maternal microchimerism where primitive stem cells of the developing embryo are able to seed not only fetal structures but also various maternal tissues during gastrulation, persisting in the bone marrow and other primitive stem cell niche, in the steady state as tissue-resident pluripotent stem cells throughout the mother’s life with implications in both health and disease ( Cismaru et al., 2018 ; Cismaru et al., 2019 ). While pluripotent stem cells have been identified in various post-natal tissues and have been labeled as endothelial progenitor cells (EPCs), multipotent adult progenitor cells (MAPCs), marrow-isolated adult multilineage inducible cells (MIAMIs), multipotent adult stem cell (MACS), fetal microchimeric stem cells (FMSCs) or very small embryonic-like stem cells (VSELs) ( Cismaru et al., 2019 ; Ratajczak et al., 2019 ; Ratajczak et al., 2008 ) depending on the researcher groups, it is likely that in most cases, similar or overlapping populations of post-migratory and quiescent tissue resident pPSCs have been described since their molecular signature supports their epiblast/germ line origin ( Shin et al., 2010 ). Recapitulation of migratory events of gastrulation in tumorigenesis and their striking resemblance with fetal-maternal microchimerism lead us to hypothesize that following a putative fusion of a quiescent tissue resident pPSC with a mutation-harboring somatic cell for cell rescue, the nuclear transfer may potentially lead to genomic reprogramming and ZL-CSC emergence followed by passage through early Carnegie stages with the reenactment of the migratory events of early development both in the ectopically developing embryoid body-like tumor as well as in host tissues, mimicking the fetal-maternal microchimerism previously explored by our group ( Cismaru et al., 2019 ) ( Figure 3 ). The proposed cycle of tumorigenesis and metastasis centering the zygote-like cancer stem cell (ZL-CSC) and exemplified for the case of an anaplastic thyroid carcinoma (ATC) as a representative model for solid tumor transformation. We speculate that primitive (germline related) pluripotent stem cells (pPSCs) originating from migratory events of gastrulation (Carnegie stage 8) and residing in the steady state in primitive hematopoietic stem cell (HSC) niche of postnatal tissues, can be recruited via CXCL12/CXCR4 chemokine axis for the repair of mutation-harboring somatic cells signaling distress via SASP cytokines. Fusion of quiescent pPSCs with the damaged cell can lead to somatic cell nuclear reprogramming to a totipotent-like state similar to the zygote but with a high rate of developmental abnormalities caused by epigenetic remnants that alter gene expression regulation in the resulting ZL-CSC. Still, the ectopic gastrula-like/tumoroid-like developing tumor (subclinical stage) recapitulates faultily the early Carnegie stages of embryo development, with primitive cell migration and early metastatic spread resembling migratory events of gastrulation and fetal-maternal microchimerism. By recapitulation of fetal-maternal microchimerism, the cancer cell migration suggests that cancer cells don’t spread randomly but instead are homing to specific environments that support primitive stem cells. As the most lethal ATC often transforms from differentiated thyroid cancer the poorly understood but complex intratumor transformation process could result from a repeated cycle of tumorigenesis involving pPSCs fusion events. As a representative model for solid tumor transformation, anaplastic thyroid cancer (ATC), while being the deadliest type of undifferentiated thyroid malignancy, reflects many of the aspects related to the activation of the zygotic genome ( DeSouza et al., 2025 ) and retroelement reactivation in disease ( D’Alessandre et al., 2025 ). Since synchronous differentiated thyroid cancer regions have been reported on histopathology in about 70% of patients with ATC this suggests that it is dedifferentiation of DTC that leads to progression to ATC ( Santarpia et al., 2008 ; Molinaro et al., 2017 ). However, the passage from a diploid stage to an aneuploid genotype and transition to a mesenchymal phenotype over the course of progression from DTC to ATC as revealed by single-cell transcriptomics ( Lu et al., 2023 ) suggests that another mechanism for unlocking the zygotic genome in ATC could be fusion-driven ( Figure 3 ). Erasure of the gametic imprinting in PGC/germline-related pPSCs could bring more insights into how tumorigenesis may revoke some of the early stages of embryonic development. It becomes intriguing to assume that the ontogeny of stem cells originating in more primitive tissue-resident precursors such as PGC-related pPSCs could have crucial implications for the fate of the resulting hybrid cell that may be amenable to revoking specific stages of embryogenesis ectopically ( Cismaru et al., 2022 ; Ratajczak et al., 2019 ). This is since PGCs were shown to be able to undergo chromatin reorganization and imprint erasure with sequential epigenetic changes and genome-wide DNA demethylation to reset the epigenome for totipotency ( Hackett et al., 2012 ). Following a putative fusion, nuclear transfer and genomic reprogramming from a PGC-related pPSC to a mutation-harboring cell, this would potentially result in activation of early developmental genes and zygote-like cell formation but faultily, similar to that reported in the cloning technique with non-enucleated oocytes ( Rouillon et al., 2019 ). This implies that oncogenesis would presumably imply a deleterious and ectopic embryogenesis-like process without the classical fecundation trigger. As fecundation has been the de facto trigger of embryogenesis in the classical dogma of fertilization, research in the last decades has shown that embryogenesis can also occur in the absence of sperm-egg fusion using the nuclear transfer technique ( Lacham-Kaplan et al., 2001 ). However this requires an external manipulation of the cells such as the one used in cloning by nuclear transfer from a somatic cell in an enucleated oocyte leading to reactivation of genes previously inactivated by cell differentiation and triggering embryogenesis ( Tian et al., 2003 ). Furthermore, non-enucleated oocytes have also been successfully used in the nuclear transfer technique leading to the disappearance of the oocyte DNA ( Rouillon et al., 2019 ). Numerous concepts have been developed over the last years that tried to explain the origin of the cancer cell and the genetic and epigenetic mechanisms regulating the gene expression related to increased proliferation, blocked apoptosis, evasion of immune reactivity and preparation of metastasis and spread. Several works exploring the origin of CSCs suggest that they may arise by VSELs malignant transformation ( Bhartiya et al., 2024 ; Sharma and Bhartiya, 2021 ). While such pPSCs in postnatal tissues seem to be involved in tumorigenesis, we speculate that the mechanism of their malignant transformation is not spontaneous as consequence of environmental insults but rather fusion driven since further tumor development closely resembles early embryogenesis whereas metastasis resembles fetal-maternal microchimerism. Our concept implies that although the cancer originates from a single cell - the ZL-CSC, tumorigenesis involves two types of cells, one pluripotent with germ line traits–the tissue resident quiescent PGC/germline-related pPSC, and one somatic but damaged by environmental insults, which fuse together for somatic cell rescue but produce the ZL-CSC through nuclear reprogramming akin to an unselective fecundation ( Figure 4 ). Step-by-step diagram illustrating the core cellular and molecular mechanisms of a failed cell rescue mechanism by fusion, leading to epigenetic reprogramming and emergence of a ZL-CSC with chromosomal chaos. (A) Primitive pluripotent stem cell responding to damaged somatic cell distress; (B) Fusion and formation of a 4n hybrid cell; (C) Epigenetic reprogramming with DNA demethylation and activation of zygotic gene networks; (D) Chromosomal chaos producing an aneuploid cancer stem cell. It becomes intriguing to speculate that a spontaneous nuclear transfer by unselective cell-cell fusion could trigger a fecundation-like mechanisms leading to activation of embryonic developmental pathways such as Wnt, Notch, Hedgehog ( Collu et al., 2014 ) ( Figure 2 ), but ectopic and defectuous, with recapitulation of preimplantation and implantation stages, and even cell migration and tumor-host microchimerism at least up to a point, yet such a hypothesis has not been previously proposed explicitly ( Figure 5 ). Graphical representation of the evolution of oncogenesis models over time including the currently proposed concept of germ-somatic cell fusion hypothesis. Regarding the testable implications of our concept, potential discriminative prediction assays can clarify the emergence mechanisms of ZL-CSCs and differentiate germ-somatic fusion mechanisms from non-fusion reprogramming such as dedifferentiation or parthenogenesis. Assessing the genomic imprinting status using single-cell methylome sequencing, allele-specific RNA-seq and phasing single nucleotide polymorphisms (SNPs) can discriminate between parental genomes. Biparental imprinting marks and no genome-wide uniparental disomy pattern can indicate a fusion driven ZGA transcriptional reprogramming rather than parthenogenesis mechanisms which is associated with global skew toward maternal methylation patterns with regions of uniparental isodisomy. Moreover, single-cell whole-genome sequencing and mitochondrial haplotype tracing studies that show mixed SNP haplotypes from two clones and discordant mitochondrial genotypes can predict ZL-CSCs emerge via fusion. Otherwise, dedifferentiation would show a genotype that matches parental clone and no new haplotype combinations. CRISPR-Cas9 based barcode lineage tracing can further differentiate fusion from other non-fusion mechanisms of ZL-CSCs emergence ( Kester and van Oudenaarden, 2018 ) to strongly resolve the proposed model.

Preimplantation, implantation, growth, immunologic acceptance and fetal-maternal microchimerism are embryonic developmental processes that resemble the current hallmarks of cancer in many aspects like aerobic glycolysis/Warburg effect, phenotypic plasticity, cell invasiveness and migration, induction of neoangiogenesis, meiotic gene expression, epigenetic regulation, retrotransposon activity, protein profiling, immune escape, and proliferative signaling ( Ma et al., 2010 ; Hanahan, 2022 ; Krisher and Prather, 2012 ; Geck, 2013 ). However, in spite of their resemblance, some hallmarks of cancer such as genome instability, mutation and metastasis are not hallmarks of embryogenesis. Still, this suggests that the shared mechanisms of the two processes might involve recapitulation of embryonic developmental programmes at least up to a point that could potentially be evoked by a similar type of trigger such as cell-cell fusion in both occurrences. The last decades have brought about major advances in understanding cell-cell fusion. Hybrid cells resulting from fusion between gametes, myoblasts, epithelial cells, immune cells, trophoblasts, cancer, and other cells in physiological and pathological processes have been extensively explored ( Sapir et al., 2008 ; Posey et al., 2011 ) ( Table 1 ). Specific cell fusion processes were shown to proceed through related membrane rearrangements. Fusogens, the proteins required for cell fusion on either one of both fusing cells use varying mechanisms as some fusions are controlled by a single fusogen, while others depend on several proteins that either cooperate throughout the fusion pathway or are the drivers of distinct fusion stages ( Sapir et al., 2008 ). Nonetheless, some fusions require fusogens to be present on both fusing membranes, while in other contexts, fusogens are required only on one of the membranes. How and why cells fuse in both normal development and disease still remains an active area of research but important progresses have been made to explore this process by both in vitro and in vivo assays ( Brukman et al., 2019 ). Various types of cell fusion and their resulting entities in physiological, reparatory and cancer related conditions. The hybrid cells resulting from fusion retain genetic and functional characteristics of the fusing cells defining a “cell chimera” in which the genome of the less differentiated cell is dominant over the more differentiated cell, likely due to expression of key reprogrammers. The “cell chimera” can result from homotypic cell fusion when the fusing cells are identical, while heterotypic fusion occurs when cells of different origins of the organism fuse together forming a heterokaryon (stable di-nucleated hybrid cell) or synkarion (hybrid cell with a single nucleus but double in the genetic material) ( Lluis and Cosma, 2010 ). This type of “cellular chimerism” is different from the classical intercellular chimerism with primitive pluripotent fetal stem cells occurring early in pregnancy during gastrulation (fetal-maternal microchimerism) or with bone marrow derived stem cell transplantation (donor cell chimerism) in which fetal or donor cells home and coexist with host cells in various niche of the host tissues ( Cismaru et al., 2018 ; Cismaru et al., 2019 ; Cismaru et al., 2020 ). Starting from the evidence that primitive pluripotent stem cells (pPSCs) with germline traits reside in the quiescent state in postnatal tissues as epiblast derived vestiges from development (e.g., very small embryonic-like stem cells – VSELs) ( Ratajczak et al., 2019 ) and that fusion between stem and differentiated cells harboring mutations can occur unselectively for cell rescue ( Blau, 2002 ; Lluis and Cosma, 2010 ), the current analysis posits a hypothesis that, although all cancers originate from a single cell, the tumorigenesis process initially involves two types of cells, one pluripotent with germ line traits and one somatic but damaged by environmental insults which fuse together for cell rescue. It is due to the intrinsic ability of germline related pluripotent cells to fuse with somatic cells and reprogram the somatic genome using Tet1/Tet2 for imprint erasure ( Piccolo et al., 2013 ) that a potential result of such fusion may actually constitute a second hit in the damaged cell leading to the generation of a zygote-like cancer stem cell (Carnegie stage 1 equivalent) via somatic cell nuclear reprogramming which subsequently drives a tumor development that recapitulates the initial Carnegie stages of embryogenesis but with a high rate of developmental abnormalities caused by mutation and epigenetic remnants in the damaged somatic cell altering gene expression regulation. The functional interrelation between germ cells which already possess critical reprogramming factors for development and somatic cell reprogramming ( Hu et al., 2018 ) indicates that the (failed) rescue attempt by germline related cells using fusion-driven reprogramming is grounded in established molecular mechanisms. In the following paragraphs we bridged a wide array of complex, multidisciplinary concepts that explore physiological and pathological cell fusion, germline traits, homing mechanisms, miRNA regulation, and retrotransposon activation, supporting the current germ-somatic fusion hypothesis of ZL-CSCs emergence.

Cell fusion regulation is complex and tightly controlled, with growing evidence pointing to microRNAs (miRNAs) as essential modulators of the process. Quite expectedly, miRNAs, small (∼22 nucleotides), non-coding RNAs that regulate gene expression post-transcriptionally are not only ubiquitous regulators of gene expression, but also among the epigenetic mechanisms with the longest evolutionary history in cell fusion since early multicellular evolution ( Christodoulou et al., 2010 ). From tissue identity to muscle formation, placentation, immune function and cell repair, miRNAs play essential roles in the control of cell fusion across various biological systems ensuring cellular cooperation and individuality. In mioblast fusion, miR-1, miR-133, and miR-206 regulate the differentiation and fusion of myoblasts into multinucleated myotubes by targeting factors like Pax7, Hdac4, and Sox6 ( Chen et al., 2006 ). In syncytiotrophoblast formation, miR-193b and miR-17∼92 cluster have been implicated in trophoblast fusion by targeting key repressors of syncytin-1, a retroviral envelope protein essential for placental fusion ( Morales-Prieto et al., 2012 ). MiR-223 is an important regulator of cells of the hematopoietic lineage playing a pivotal role in macrophage and osteoclast fusion ( Sugatani and Hruska, 2007 ). SiRNA knockdown of miR-223 significantly increases osteoclast fusion while macrophages treated with mir-223 mimic ( Moore et al., 2016 ) or pre-miR-223 ( Sugatani and Hruska, 2007 ) lose their ability to undergo fusion in experimental models. There is growing evidence that the miR-200 family, miR-34, and other factors are linked to fusion-like activities and the epithelial–mesenchymal transition (EMT) in cancer stem cells, while their normal expression inhibits nuclear reprogramming of cells toward less differentiated phenotypes ( Shimono et al., 2009 ; Choi et al., 2011 ). MiR-17∼92 cluster (also known as oncomir-1), part of the let-7 family miRNAs is essential in early embryogenesis for hematopoietic and various fetal tissue development, and deficient miR-17∼92 cluster is associated with postnatal fatality ( Ventura et al., 2008 ). A comparison of miRNA expression in the mature oocyte, with zygote miRNAs, showed the same miRNA expression pattern in both cells, consistent with a maternal inheritance of these miRNAs from oocytes and not their transcription in the zygote itself ( Tang et al., 2007 ). Being essential for embryonic development, its germline transmission make it a pivotal regulator of gene expression in embryonic stem cells. It is becoming increasingly clear that in fusion capable cells such as primordial germ cells, embryonic stem cells and cancer stem cells, MiR-17∼92 cluster functioning is integral to the balance between “stemness” and differentiation, employed during development but hijacked in malignancy ( Cioffi et al., 2015 ). Murine studies have shown that upon egg fecundation, a maternal to zygote transition (MZT) occurs, which starts with the gradual clearance of maternal Dicer and miRNAs stored in the cytoplasm of oocytes, which are crucial to sustain the first cleavages in the first steps of early development ( Tang et al., 2007 ). MiRNAs act pleiotropically, blocking the translation of hundreds of maternal targets and miR-17∼92 cluster plays a pivotal role in this process. The transition to zygotic genome ensures that the maternal transcripts are actively degraded and replaced by zygotic products during zygotic genome activation (ZGA), a striking example of cellular reprogramming ( Rosa and Brivanlou, 2017 ). A new synthesis of miRNAs begins with the two-cell stage and includes the expression of miR-290 to miR-295 as the first detectable embryonic miRNAs ( Tang et al., 2007 ). It is noteworthy that miR-290 to miR-295 are also specifically expressed in embryonic stem cells (ESCs), being associated with maintenance of pluripotency ( Houbaviy et al., 2003 ). ESCs have a distinctive cell cycle with a very short G1 phase and deficient G1/S check point, ensuring them a quicker entrance in the S phase, similar with cancer cells ( Berthet and Kaldis, 2007 ). This mechanism of pluripotency maintenance through shortening G1 and extending S phase is assisted by miR-290–295 cluster through its effects on Cyclin D1 ( Yuan et al., 2017 ). The increased expression of Cyclin D1 that drives unchecked cellular proliferation is regulated postranscriptionally by miR-290–295 cluster in ZGA and is also recapitulated in cancer, promoting tumor growth and drug rezistance ( Montalto and De Amicis, 2020 ). The S phase regulating kinase CDK2 is not essential for mitotic cell cycle as it is compensated by CDK1 which promotes G1/S transition. However, in germ cells, CDK2 is essential for meiosis as it cannot be compensated by CDK1 ( Berthet and Kaldis, 2007 ). It was shown that failure to downregulate Ciclin D1 in primordial germ cells in order to go on cell arrest on schedule is associated with germ cell tumors susceptibility ( Lanza et al., 2016 ; Cook et al., 2011 ; Heaney et al., 2012 ). The expression of the miR-290–295 cluster in mice and its miR-371–373 homolog in humans is restricted to early embryos, stem cell lines derived from the early embryonic lineages, PGCs, and the germ line stem cell compartment of the adult testis ( Wu et al., 2014 ), providing a link between the stemness potential of PGCs and CSCs ( Figure 1 ). Despite having established roles in regulating post-transcriptional gene expression after fusion of diverse cells, there aren’t any miRNAs reported in the existent literature that regulate the expression of fusion genes such as IZUMO1 and ERVW-1 which are essential in gamete adhesion and fusion ( Soygur and Sati, 2016 ). As these genes evolved under strong transcriptional containment, with relatively minor selective pressure to develop complex miRNA-based repression layers, this suggests that regulation relies more on transcriptional silencing than on post-transcriptional modulation so their reactivation is thus mostly transcription-driven, not miRNA-limited. This supports the idea that, while serving as ancient epigenetic sentinels ensuring that post-fusion events occur only under tightly regulated conditions, miRNAs have fewer (if any - they are difficult to predict, rarely investigated, and biologically non-canonical) regulatory implications during the actual fusion. This is particularly important because dysregulated fusion may potentially lead to immunological dysfunction, defectuous histogenesis and even multinucleated or polyploid tumor cells ( Shabo et al., 2020 ). Epigenetics play a key role in development, determining cell fate by driving the transition from a single totipotent stem cell to an entire organism during embryogenesis. We are just beginning to understand the extensive epigenetic and functional resemblances between early developmental stages and cancer where transposable elements play a pivotal role ( Table 2 ). Genes and interacting miRNAs involved in the shared features of PGCs and CSCs. Transposable elements, also known as jumping genes, can move from one genomic locus to a new one and represent almost half of the human genome. Retrotransposons account for about 42% of the human genome and use an RNA intermediate to move to new genomic locations. They contribute to the genetic polymorphisms being regarded as important drivers of genome evolution through acquisition of new functional characteristics from homing to new coding regions of the genome. Transposable elements are almost always completely inactivated in healthy somatic cells ( Walsh et al., 1998 ). While epigenetic inactivation/silencing of transposable elements through histone methylation and deacetylation occurs in differentiated cells to circumvent their potential genotoxic effect, the primordial germ cells have active retrotransposons responsible for their specification and ensuring genetic variability in the future zygote ( Xiang et al., 2022 ). Human Endogenous Retroviruses (HERV) retrotransposons, vestiges of an ancestral retrovirus infection in the germ line, are still being actively transcribed in the placenta, hypothalamus, and testis ( Crowell and Kiessling, 2007 ; Kim et al., 2004 ). PGCs are precursors of both male and female germ cells. One representative epigenetic reprogramming event in PGCs is the global demethylation during their specification, which also puts transposable elements into a transcriptionally active state ( Lesch and Page, 2012 ). In a subsequent step, another DNA demethylation event takes place in a locus-specific manner to induce meiotic genes, germline genes required for gamete generation and imprint erasure ( Yamaguchi et al., 2012 ; Hackett et al., 2012 ). In embryogenesis, after sperm-egg fusion, the resulting zygote undergoes global demethylation ( Altun et al., 2010 ). During this reprogramming, some transposable elements keep their methylated state, similar to the maintenance of methylation of imprinted genes while others demethylate initially but soon remethylate through recruitment of the de novo methylation apparatus ( Theunissen et al., 2016 ). However, some transposable elements remain unmethylated and become transcriptionally active at this early stage regulating essential totipotency and pluripotency factors ( Rodriguez-Terrones and Torres-Padilla, 2018 ). Furthermore, active transposable elements of the placenta regulate key transcripts that are essential for invasion, immune modulation, neoangiogenesis, growth and proliferation, the placenta showing a DNA methylation landscape resembling that of tumors ( Robinson and Price, 2015 ). The repetitive nature of retrotransposons can result in incorrect recombination events and lead to translocations, deletions and insertions, resulting in genomic instability, a hallmark of cancer ( Hollister and Gaut, 2009 ). Long interspersed element 1 (LINE-1) is a highly active retrotransposon in the primordial germ cells being involved in key epigenetic reprogramming events of meiosis and gametogenesis ( Zhou et al., 2022 ). LINE-1 overexpression and retrotransposition are hallmarks of cancers, promoting genetic heterogeneity in tumors ( Mendez-Dorantes and Burns, 2023 ). Because their mobilization can result in deleterious effects, as highlighted by retrotransposon (RE)-associated diseases ( Belancio et al., 2009 ; Payer and Burns, 2019 ), cells have developed diverse mechanisms to repress their activity, primarily by epigenetic silencing, which includes DNA ( Haggerty et al., 2021 ) and histone ( Lindehell et al., 2023 ) methylation. De novo somatic insertions of the L1 REs have been observed in particular in colon epithelial cells ( Nam et al., 2023 ) and cortical neurons from both diseased and normal brain tissues, with potential implications in the genomic diversity of neurons for the later ( Zhao et al., 2019 ). Furthermore, retrotransposition events have been shown to contribute to hippocampal genomic variations and cognition processes ( Bachiller et al., 2017 ). Several lines of evidence indicate that neuronal progenitor cells seem to be a “hot spot” of retrotransposition activity in the brain ( Coufal et al., 2009 ; Muotri et al., 2005 ), linking the pluripotent phenotype of cells to an increase in retrotransposition events. Whole genome methylation analysis suggests that a decrease in methylation of L1 promoter occurs early in embryogenesis, reaching the lowest degree in gastrulation stage of the embryo. Post-gastrulation remethylation and subsequent L1 silencing is associated with the differentiation and maturation of the embryo and continues throughout life. In elderly an increase in retrotransposition events varies depending on the genetic imprinting inherited from parents, potentially leading to malignant transformation of cells ( Nam et al., 2023 ). Because genomic instability is a hallmark of cancer ( Negrini et al., 2010 ), and REs represent a major source of genomic instability ( Daskalos et al., 2009 ), it can be reasonable to say REs are important drivers of cancer pathogenesis, and high levels of retrotransposition is associated with cancer cell stemness ( Apostolou et al., 2015 ) a feature of highly aggressive tumors, relapse and resistance to treatment ( Chengizkhan et al., 2020 ). The undifferentiated stem cell phenotype, present in multiple tissues, and early in development, might be a consequence of REs activity. Therefore, genes that are regulated by REs in stem cells might overlap with the ones that are regulated by REs in tumor cells. It is known that regulation of the transcriptome in embryogenesis resembles the one found in malignant tumors, as reveled by a comprehensive analysis of expression data from 1,094 individual arrays derived from a broad spectrum of tumors. Clustering in relation to a developmental timeline expression pattern, resulted in three groups. Group 1 includes tumors enriched in genes upregulated in early embryogenesis, and related to stem cells phenotype, and downregulated genes linked to tissue specificity and development. Group 3 comprise tumors with gene expression patterns mostly related to the late development timeline, while group 2 falls in between. It is worth mentioning that group 1 comprised tumors with an aggressive phenotype and high proliferating rates, such as lung cancer. Nonetheless, a core gene set that map to early development was identified in approximately 50% of the tumor expression data analyzed ( Naxerova et al., 2008 ). As a link between the early embryogenesis stages and malignancy is highly probable to exist, it remains to question if REs are the ones that might drive these similar phenotypes of embryogenesis and tumorigenesis, yet different as we perceive them to be. A recent study points to such an assumption, showing that knockout of the L1 retrotransposons in mouse embryos lead to developmental arrest, while induced L1 expression is associated with upregulation of distant genes essential for zygotic activation and transition to blastocyst, in a transcriptional-dependent manner ( Naxerova et al., 2008 ), rather than chromatin remodeling as a previous study suggest ( Jachowicz et al., 2017 ). Therefore, if retrotransposon transcripts, such L1, might induce zygotic activation after fecundation, the question that might arise is related to the origin of these transcripts. The answer could possibly reside in the transcriptomic profile of the spermatozoa and oocytes. RNA sequencing analysis non-coding small non-coding RNAs (sncRNAs) from human mature sperm reveals that approximately 65% of transcript reads belong to repeat sequences. Among the known repeats, the LINE, LTR and SINE retrotransposons are the most prevailing ones. The remainder of the sncRNAs are represented by regulatory RNAs, such piwi interacting RNAs (piRNAs) and microRNAs (miRNAs) ( Krawetz et al., 2011 ). In addition, an abundance of coding mRNAs has also been identified, with functions in transcription regulation and cell cycles, among other ( Dadoune et al., 2005 ). The transcriptome of the human oocytes also shows expression of retrotransposons, though this study points to an abundance of LTR and SVA retrotransposons as compared to LINE ( Georgiou et al., 2009 ). It would be expected that differentiated cells, such as sperm cells and oocytes, would exhibit low levels of retrotransposons expression. Papers point to a genome-wide demethylation in the initial stages of sperm development, while the mature sperm cells exhibit an elevated methylation level and hence reduced transcriptional activity ( Deniz et al., 2019 ). This implies an accumulation of retrotransposition transcripts in the absence of transcriptional activity in mature germ cells, probably to limit the mutagenic potential of REs. After fecundation, the transcriptomic profiles of the germ cells bring the regulatory RNAs in one cell, the zygote, to initiate the embryonic development. It might be that the differences in the retrotransposon transcriptomic profile confer the germ cells different proprieties. It is known that syncytins derived from the envelope of HERV retrotransposons confer fusogenic properties to the cells encoding them and are involved in placentation of the mammalian embryos. A higher level of LTR retrotransposons-derived syncytins expression in the oocytes ( Georgiou et al., 2009 ) could make them susceptible to selective fusion with the sperm cells, as oocytes are not produced in their millions compared to sperm cells and are hence less likely to produce a wide-spread syncytium and infertility. The more abundant LINE retrotransposons in the sperm cells, alongside with the other regulatory transcripts ( Krawetz et al., 2011 ), would mediate activation of pluripotent phenotype in the new zygote cell, and proliferation to multiple cells that later would suffer the process of differentiation in specific tissues following an epigenetic pattern of regulation. Retrotransposons were initially regarded as “junk” DNA and later as drivers of evolution, but now they can be safely considered as important regulators of development, providing the cell with pluripotent characteristics. Such a phenotype is mandatory, both in early stages of embryogenesis, and later in adulthood. As transposable elements activate early developmental genes, they can also be key contributors to the malignant transformation of the cell. A mechanism similar to the cell fusion in fecundation where two terminally differentiated cells bring together different transcriptomic profiles, that initiate the expression of a pluripotent phenotype, and lead to cell proliferation and growth, might also be responsible for triggering early cancer pathogenesis. Furthermore, the same mechanisms that protect the developing embryo from the mother’s immune system might also help cancer cells to evade immune surveillance, as studies would suggest ( Zhu et al., 2021 ). Transcriptionally active retrotransposons are seen in CSCs and drive upregulation of numerous oncogenes by their endogenous promoter activity in a process termed onco-exaptation ( Lynch-Sutherland et al., 2020 ; Jang et al., 2019 ). While the mechanisms of retrotransposons activation in cancer is unknown, the epigenetic landscape in cancer resembles that of early development with transposable elements activation, hence dedifferentiation has been proposed as a potential mechanism of retrotransposon demethylation and reawakening in tumor cells ( Lynch-Sutherland et al., 2020 ). However, more recent investigations using single-cell transcriptomics of stem vs. differentiated and cancer vs. normal cells indicates a rather atavistic type of genomic signature in cancer stem cells rather than the result of dedifferentiation, putting the zygote-like cancer stem cell (unicellular-like state) as a central hallmark of cancer ( Vinogradov and Anatskaya, 2025 ). How this unicellular-like state is acquired in cancer stem cells is still unknown. Since sperm and eggs fuse together selectively and not randomly with any other cell type, the odd possibility of unselective fusion triggering subsequent reactions similar but ectopic to those occurring in normal fecundation could bring together all the hallmarks of cancer into one unified theory. This is supported by recent findings that spontaneous differentiation of germ cells from human embryonic stem cells has been shown to occur in vitro ( Geijsen and Daley, 2008 ) and while doubtful, we cannot exclude such an occurrence in-vivo with pPSCs (e.g., VSELs) as protagonists. In this context, unselective stem cell fusion leading to the emergence of a zygote-like cancer stem cell becomes even more intriguing since fertilizing oocytes using artificial fusion with somatic cells as male germ cells has already been demonstrated ( Lacham-Kaplan et al., 2001 ). Little is known about the potential acquisition of demethylated/active retrotransposons from fusion events. As germ cells carry active retrotransposons required for propagating themselves and ensuring genetic diversity in the future zygote ( Zhou et al., 2022 ), we suspect that a putative fusion of a PGC-related pPSC in postnatal tissues with a mutation-harboring somatic cell for cell rescue could potentially explain the acquisition of active retrotransposons in the resulting ZL-CSC. This could be responsible for tumor heterogeneity as seen in many cancers. Nonetheless, global demethylation following their fusion, as seen in the selective fusion that leads to zygote formation, could also explain the imprint erasure in a zygote-like cell resulting from unselective fusion. Regardless of their mechanism of activation, retrotransposons remain a distinctive feature in both embryogenesis and oncogenesis and their demethylation/activation mechanisms in cancer warrants further investigation.

Unlike CSCs which have been described as pluripotent and lineage restricted, zygote-like (8C-like) CSCs have been described as a rare metastable subpopulation of 1 in 1,000 cancer DUX4 + HLA-I - cells that are able to undergo ZGA and go through an early totipotent program with the expression of embryonic, trophectoderm, and mesenchymal markers ( Smith et al., 2023 ). Since DUX4 overexpression in humans is characteristic of the 2-4-cell stage zygote ( Hendrickson et al., 2017 ) and also of early germ cells ( Karlsson et al., 2021 ), they sit conceptually between cancer stem cells (CSCs) Induced pluripotent stem (iPS)-like states and pre-implantation embryonic cell states. We will further refer to these CSCs as zygote-like (ZL-CSCs) as they can be defined as a rare, highly plastic tumor cell population that expresses totipotency-associated genes and epigenetic features similar to those found in the zygote, enabling the cell to self-renew and differentiate into multiple divergent tumor cell lineages, thereby driving tumor initiation, progression, immune escape, and therapy resistance. In order to better understand the putative origin of ZL-CSCs, we delved deeper in ontogeny, to primitive progenitors emerging during gastrulation which have been shown to migrate and colonize fetal and maternal tissues, becoming a source of steady-state tissue committed pluripotent stem cells with a tight germ line kinship and overlapping molecular and functional traits with primordial germ cells (PGCs), primitive HSCs, primitive MSCs, and other pluripotent tissue resident precursors that have received different names depending on the teams that identified and described them ( Howell et al., 2003 ; Ratajczak et al., 2017 ) but most probably indicate similar or overlapping primitive populations of pluripotent cells ( Ratajczak et al., 2012 ; Cismaru et al., 2019 ; Cismaru et al., 2020 ). We will further address these cells as primitive pluripotent stem cells (pPSCs) since a constant feature indicating there primitiveness and linking them to other populations of PSCs is their induction in numbers in response to stimulating a vestige from their ontogeny – the orphan LHCG receptor by pituitary hormones, sex hormones, and chorionic gonadotropin ( Mierzejewska et al., 2015 ; Cismaru et al., 2020 ; Ratajczak, 2017 ; Abdelbaset-Ismail et al., 2015 ). Another characteristic of pPSCs in postnatal tissues is their susceptibility to tumorigenesis by still incompletely elucidated mechanisms which range from oncogenic transformation into CSCs and genomic instability following exposure to environmental insults ( Bhartiya, 2025 ), to fusion with somatic cells leading to heterokaryon formation, chromosomal aneuploidy and genomic instability leading to malignant transformation ( Ratajczak et al., 2009 ; Ratajczak et al., 2017 ). Having germ line markers, these pluripotent but lineage restricted stem cells are far more advanced in their differentiation pathway than a totipotent zygote, but are the closest in kinship with germ cells that acquire totipotency through fertilization ( Ratajczak, 2017 ). Our hypothesis for the potential origin of ZL-CSCs is that they could potentially emerge from the unselective but fecundation-like fusion of tissue resident pPCS harboring germline traits, with mutation harboring somatic cells, for cell rescue. The most versatile germ line cells are the PGCs which during development migrate by a well-defined route from the yolk sac into the genital ridge later giving rise to gametes throughout life ( Inst, 1948 ). They are the only cells in the human body capable of both mitosis and meiosis. During mitosis, they maintain their diploid state. During meiosis, the ploidy is reduced from diploid (2n) to haploid (1n) in the development of eggs and sperm cells. PGCs are programmed to go into apoptosis by BAX mediated programmed cell death if they don’t reach their destination ( Stallock et al., 2003 ), but some PGCs may mismigrate and hijack apoptosis later in life giving rise to germ cell tumors across their continued perineural migration along the midline of the developing embryo in distant sites of the sacral region, retroperitoneum, abdomen, anterior mediastinum, neck and midline of the brain as proposed by the embryonic rest hypotheses of Rudolf Virchow and Julius Cohnheim (19th century) ( Oosterhuis and Looijenga, 2019 ; Ratajczak et al., 2013 ). The striking resemblances between PGC-like cancer cells and PGCs in cell features like gene expression profiles such as CXCR4 and c-Kit, migration and tumor metastasis potential have been previously explored extensively ( Liu C. et al., 2022 ). The germline traits seen in tumors which are further discussed in the next section represent indirect evidence for the potential involvement of PGC-related pPSCs from postnatal tissues in tumorigenesis. Akin to the PGCs, many cancer cells express genes linked to different embryo developmental stages, such as germ cell-specific genes. This allows them to express meiotic proteins and undergo meiosis similar to PGCs ( McFarlane and Wakeman, 2017 ). Two genes regulating pluripotency in embryonic development that greatly influence meiosis in both PGCs and cancer cells are LIN28A and LIN28B also known as germ-cell/cancer (GC) genes ( Irie et al., 2015 ; Bruggeman et al., 2018 ). Germline-specific genes are often activated in cancer. Genes specific to PGCs, such as the germ cell determination gene DAZL, modulate tumorigenicity, stemness and drug resistance in cancer cells, being associated with the expression of stemness markers such as PROM1, SOX2, NANOG in glioblastoma cells ( Zhang et al., 2020b ), while in NSCLC it promotes migration, invasion and drug resistance by upregulating JAK2 and MCM8 ( Zhou et al., 2024 ). NANOS3, another germline specific gene that regulates inhibition of apoptosis during migration of PGCs, is also expressed in glioblastoma where it promotes migration, proliferation, invasion and drug resistance ( Zhang et al., 2020a ) while in lung cancer it promotes invasiveness by inducing EMT ( Grelet et al., 2015 ). DUX4 gene is an important regulator of ZGA for totipotency in humans and its silencing leads to disordered degradation of maternal transcripts and incomplete ZGA ( Liu Y. et al., 2022 ). While DUX4 is repressed in somatic cells, it is expressed in high levels in germ line cells ( Snider et al., 2010 ). A single–cell type transcriptomics map of human tissues and subcellular map of human proteome shows an increased expression in early germ cells but many cancer types also show a high expression of DUX4 RNA ( Karlsson et al., 2021 ; Thul et al., 2017 ). In cancer cell lines, DUX4 is only expressed in a subpopulation of 0.01% of cells under standard culturing conditions, but its transient (zygote-like) expression is sufficient to induce early embryonic and extraembryonic totipotent programs with expression of genes regulating trophoectoderm formation, and selective immune evasion by suppression of MHC class I ( Smith et al., 2023 ; Chew et al., 2019 ). Trophoblast specific genes such as TACSTD2 encoding Trop-2 expression are specific to trophoblast cells and developing embryo tissues warranting them invasiveness (e.g., invasion of the myometrium and induction of neoangiogenesis for access to host blood supply) and immune tolerance. However, Trop-2 overexpression is seen in many tumor types being associated with increased aggressiveness ( Shvartsur and Bonavida, 2015 ). Trophoblast cells also express HLA-G, a non-classical HLA Class I antigen involved in the induction of immunological acceptance of the embryo by the host tissues of the mother and recognized as a hallmark of the placenta ( Loke et al., 1994 ). HLA-G is strongly linked to embryogenesis being expressed in the zygote, blastocyst and early embryo, and its roles in spermatogenesis further links its expression to PGCs ( Yao et al., 2014 ). Our previous findings in vitro show that HLA-G expression is associated with the selection of more primitive and potent progenitors following human chorionic gonadotropin (HCG) exposure in primary cultures of BM derived stem cells ( Cismaru et al., 2020 ). In cancer, HLA-G expression is a common occurrence endowing cancer cells with immune evasion and metastatic capabilities ( Lin and Yan, 2015 ). Syncytins, the only known human fusogens, encoded by ERVW-1 gene, are proteins expressed by cytotrophoblast and syncytiotrophoblast cells of the feto-placental barrier that allows nutrient and waste exchange but blocks maternal immune cell infiltration ( Sha et al., 2000 ). Remarkably, syncytyn-1 is also expressed in microglia ( Antony et al., 2004 ), and testis ( Sha et al., 2000 ), potentially linking its expression to yolk sack progenitors such as primitive HSCs and PGCs since microglia is known to originate from the yolk sac progenitors and colonize the brain before the establishment of the blood-brain barrier during embryogenesis ( Dermitzakis et al., 2023 ). ERVW-1 gene and Syncytins expression is also upregulated in early germ cells and its expression is superposable to NANOS3 gene expression in germ cells and cancer cells ( Uhlén et al., 2015 ). It is noteworthy that various cancer cells express syncytyn-1 endowing them with fusion capabilities in breast cancer ( Bjerregaard et al., 2006 ), endometrial cancer, testicular cancer/seminoma, lung cancer/NSCLC, prostate cancer, colorectal cancer, bladder cancer, neuroblastoma and cutaneous T-cell lymphoma/mycosis fungoides ( Wang et al., 2023 ). Furthermore, IZUMO1 gene, encoding the Izumo sperm-oocyte fusion protein in sperm cells together with its receptor in egg cells (JUNO/folate-receptor delta isoform/SRPδ) is involved in gamete binding during fecundation and in meiosis ( Bianchi et al., 2014 ). IZUMO1 has been reported as a likely causal gene in various tumors such as melanoma, head and neck squamous cell carcinoma ( Iorio et al., 2018 ), colorectal cancer ( Law et al., 2019 ) and leukemia ( Halik et al., 2020 ). This link between germ line cells and cancer stem cells, underlined by the cancer/testis antigens encoded by genes with restricted expression to the germline, but also expressed in a range of cancers, has been extensively explored in the last years ( Bruggeman et al., 2018 ). The recapitulation of parts of the germline gene expression programme (gametic recapitulation) is reflected in oncogenesis through immortalization, invasiveness, immune escape, hypomethylation and migration/metastatic spread, underscoring the shared characteristics of germ cells and cancer cells ( Simpson et al., 2005 ). A comparison between germ cell transcriptome and combined data from public cancer genome databases (TCGA, GTEx) showed more than 700 genes that are highly germ cell-yet also cancer-specific, being essential for germ cell development, but also for tumor proliferation and metastasis ( Bruggeman et al., 2018 ; Wang C. et al., 2016 ). While expression of meiotic genes is a common trait in both germ and cancer cells, the mechanisms leading to such expression in CSCs are still elusive, with dedifferentiation potentially being evoked ( Lingg et al., 2022 ; Sou et al., 2023 ). However, given the striking similarities between PGCs and CSCs both genotypicaly and phenotypicaly ( Figure 1 ), fusion events could bring more insights into how a mutation-harboring cell could acquire meiotic genes and become a ZL-CSC. This is since erasure of the gametic imprinting of genes in the zygote or the primordial cells of the blastocyst and the totipotent stem cells of the very early embryo, control not only the fetal growth but also the feto-maternal interactions that are required to maintain a balance between contradictory fetal and maternal requirements including immunologic acceptance ( Ruvinsky, 1999 ). As erasure of the gametic imprinting is regarded as a hallmark of cancer ( Rainier and Feinberg, 1994 ), it becomes more clear how the dysregulated control mechanisms governing cell cycle progression in cancer resemble the unchecked cell division seen during embryonic development ( Vogelstein and Kinzler, 2004 ). Venn diagram displaying the overlapping traits shared by primordial germ cells (PGCs) and zygote-like cancer stem cells (ZL-CSCs) from the germline gene expression programme. Persistent and quiescent missmigrated/ectopic PGCs that fail to enter apoptosis and retain pluripotency are the most primitive PSCs from postnatal tissues and are presumably part of similar populations of developmentally related (epiblast derived) tissue resident quiescent pPSCs involved in tumorigenesis since their plasticity is well reported into all 3 germ layers and into germ cells. It is noteworthy that similarities between primitive germ line derived stem cells and CSCs exist not only in their genotype ( Table 1 ) but also in their phenotype and are also reflected in the migratory capabilities in both circumstances ( Figure 2 ). Exploring migratory events and homing of stem cells in predefined niche is an area of active research in both development and oncogenesis and is described in the following section. Graphical representation of the common traits shared by the developmental stage of a gastrula (Carnegie stage 8 embryo) and a subclinical stage of a tumor (tumoroid-like) in terms of the epithelial to mesenchymal transition (EMT) programme. In physiological conditions, anchorage dependence and contact inhibition of proliferation and migration are conventional mechanisms ensuring normal tissue homeostasis while a prolonged detachment usually triggers anoikis, a form of programmed cell death ( McClatchey et al., 2016 ). In the physiological cell migration of PGCs, NANOS3 gene ensures the suppression of both Bax-dependent and -independent apoptotic pathways, being essential in the early stage embryo to protect the migrating PGCs from apoptosis. However, in PGCs that fail to reach their destination in the prospective gonads, the pro-apoptotic gene Bax is required for the death of ectopic PGCs ( Stallock et al., 2003 ). Deregulated Bax expression is seen in ectopic germ cells that retain their primitive markers and are associated with germ cell tumors ( Cook et al., 2011 ). Expression of pro-apoptotic gene BAX is regulated by the tumor suppressor p53 but about 50% of all tumors have a disrupted P53 function ( Lomax et al., 1998 ). In embryogenesis, the collective migratory events of gastrulation and neural crest cell development have been extensively described and imply clusters of primordial cells which migrate and become specified at locations distant to where they will ultimately reside for their physiological function. Collective epithelial and mesenchymal cell migration during gastrulation is based on mechanisms of epithelial to mesenchymal transition ( Chuai et al., 2012 ) ( Figure 2 ). Seeding of primitive migratory cells in their prospected niche occurs physiologically during fetal development with the best examples being the migration of PGCs from the yolk sac to the genital ridge as future gametes, the migration of the neural crest cells to their destination in the nervous system, skin, and hair bulge as neurons/glia and melanocytes, or the migration of hemangioblasts from the yolk sac mesoderm to the primary vascular plexuses as angioblasts and of primitive HSCs to the aorta-gonad-mesonephros (AGM) then fetal liver and bone marrow ( Møllgård et al., 2010 ; Kim, 2010 ). It has been reported that most tissues in the developed organism accommodate niche containing clones of primitive HSCs derived from the yolk sac that are capable of reenacting extramedullary hematopoiesis in organs other than the bone marrow such as lymphnodes, liver, spleen, adrenal glands, cranial and spinal nerves, prostate, pleura, peritoneum, and other tissues ( Kim, 2010 ; Sohawon et al., 2012 ). Expression of CXCL12 (SDF1) is important for primitive HSC migration, homing, retention in the steady state and hematopoiesis in both the bone marrow and extramedullary sites ( Wang et al., 2014 ; Schyrr et al., 2024 ; Mendt and Cardier, 2015 ). Furthermore, CXCL12a protein, the most widespread splicing variant which is more potent in tissues than in the blood ( Janowski, 2009 ) is necessary to also guide PGCs toward the prospective gonad and ectopic expression of CXCL12a is capable of trapping PGCs in islands of high expression ( Aman and Piotrowski, 2010 ). It was shown that PGCs can be attracted to ectopic sources of CXCL12 expression in extragonadal tissues ( Doitsidou et al., 2002 ) therefore the expression of CXCL12 can be regarded as an important regulator of the suitability of the niche for both PGCs and primitive HSCs. Yolk sac derived primitive HSCs exist in many adult tissues where they undergo asymmetrical division to self renew and symmetrical division to differentiate and produce tissue-resident immune cells such as histiocytes of the brain (microglia), liver (Kupffer cells), bone (osteoclasts), skin (Langerhans cells), lung (interstitial macrophages/nerve- and airway associated macrophages), lymph node and spleen (interdigitating cells), pleura (pleural macrophages), peritoneum (peritoneal macrophages), synovial lining (aquaporin 1 + macrophages), etc. Their pools are maintained independently of BM monocytes having a more tolerogenic phenotype than their BM-derived counterparts ( van Furth, 1998 ; Bartnicki et al., 2024 ; Schonfeldova et al., 2025 ). Their niche with high CXCL12 expression seems to represent an appealing lodge for precursor cells including cancer cells ( Hashimoto et al., 2013 ; Cismaru et al., 2022 ; Casanova-Acebes et al., 2021 ). Reports showing increased CXCL12 expression in tissues that are amenable to accommodating both extramedullary hematopoiesis ( Kim, 2010 ) and metastases of various tumor types ( Zhao et al., 2011 ; Köhler and Milstein, 1975 ; Pan et al., 2006 ; Tang et al., 2019 ; Oonakahara et al., 2012 ) support the role of the hematopoietic niche in hosting migrating stem cells. Anticancer therapies focused on targeting CXCL12/CXCR4 axis hold considerable promise due to its role in promoting proliferation of tumour cells in protective niches ( Rueda et al., 2025 ). It has been postulated that tissue-resident HSCs of the adult have common ancestry with primitive precursors related to the germ line such as PGCs and other epiblast derived stem cells ( Mierzejewska et al., 2015 ; Ratajczak and Suszynska, 2016 ; Ratajczak et al., 2017 ). Evidence supporting this ontogeny is their expression of the orphan receptor - luteinizing hormone chorionic gonadotropin (LHCGR) that is stimulated by the pregnancy hormone - human chorionic gonadotropin (HCG) as well as pituitary and gonadal hormones (TSH, FSH, LH, estrogens, androgens) and their independence of erythropoietin which is required for definitive hematopoiesis ( Ratajczak, 2017 ; Abdelbaset-Ismail et al., 2015 ; Maggio et al., 2013 ). Both Notch 1 and erythropoietin (EPO) are regulators of hematopoiesis by HSCs while they are not required for hematopoiesis by primitive-HSCs (pHSCs) ( Huyhn et al., 1995 ; Pereda Tapiol and Niimi, 2008 ; Maggio et al., 2013 ). These observations emphasize the long documented positive effects of pituitary and placental hormones on stimulating hematopoiesis in bone marrow insufficiency ( Maggio et al., 2013 ), in normal pregnancy ( Dockree et al., 2021 ), in the hematopoietic effect of thyroid hormones ( Zhang et al., 2017 ) and even the rejuvenating effect of pregnancy on the mother and her post reproductive life expectancy ( Falick Michaeli et al., 2015 ; Grundy and Tomassini, 2005 ). Our group showed that HCG stimulates in vitro the selection of primitive precursors (CD133 + , NANOG + , SOX 2 + , OCT 4 + ) with a molecular profile linked to the gene expression programme of gastrulation ( Cismaru et al., 2020 ), while in vivo HCG stimulates the induction of primitive hematopoietic precursors (CD34 + , SCA1 + ) and leukocyte progenitor cells (CD29 + , CD11b + ) which lead to improved hematopoiesis and better survival in a murine model after peripheral blood stem cell transplantation ( Cismaru et al., 2024 ). Reports of high levels of fetal hemoglobin (HbF) in adults with increased levels of HCG from germ cell tumors ( Dainiak and Hoffman, 1980 ) also indicate that is rather the pHSCs that induce in numbers from the effect of HCG. A puzzling observation is the secretion of HCG not only by germ-cell tumors but also by various other tumor types. Paracrine secretion of β-HCG by malignant tumors such as osteosarcomas, lung, bladder, breast and colorectal cancers is associated with a worse prognosis ( Glass et al., 2015 ; Iles et al., 2010 ). From these observations we can speculate that it is the tissue resident primitive stem cells with a germline kinship that play a pivotal role in tumorigenesis through the effect of HCG on their proliferation and on the aggressiveness and therapy resistance of tumors as also proposed by Bhartiya (2025) . The group of Ratajczak et al. developed an elegant series of evidence showing that primitive stem cells with a germline kinship can have multiple implications in physiological and pathological conditions such as tissue regeneration, restoration of gametogenesis, hematopoiesis and even cancer ( Ratajczak et al., 2017 ; Ratajczak et al., 2013 ; Ratajczak et al., 2007 ; Ratajczak et al., 2019 ). While metastasis is a hallmark of oncogenesis, it does not represent a hallmark of embryogenesis. However, similar to embryonic migratory events of cells during gastrulation (Carnegie 8 stage embryogenesis), tumor cells in circulating clusters retain epithelial gene expression while also displaying a hybrid epithelial-mesenchymal phenotype to complete proliferation and metastasis in the early subclinical stage of tumor development ( Cheung and Ewald, 2016 ) ( Figure 2 ). Rather expectedly, distant sites of tumor cell clusters have been found in the early stages of cancer patients, which is in line with the observations that initiation of metastasis occurs in the subclinical stages of tumorigenesis ( Yang et al., 2019 ; Steinert et al., 2008 ). It is notable that subclinical tumor stages (1–2 mm 3 ) share common traits with Carnegie 8 stage embryogenesis in their development such as upregulation of mesenchymal gene programmes with reactivation of the same EMT factors — Snail/Slug, Twist, ZEB — to reduce cell-cell adhesion and enhance migratory potential ( Micalizzi et al., 2010 ), gastrulation signals such as Wnt, TGF-β, Notch, Hedgehog allowing invasion and migration ( Kalluri and Weinberg, 2009 ; Garg, 2017 ), E-cadherin downregulation to acquire invasive traits ( Scheibner et al., 2021 ; Kim et al., 2014 ) and dynamic EMT-MET ensuring plasticity which is essential for adaptation to changing microenvironments ( Huang et al., 2022 ; Pirlog et al., 2023 ) ( Figure 2 ). Invasive migration by individual and collective cell clusters involves proteolytic extracellular matrix (ECM) remodeling ( Wolf et al., 2007 ). The conventional mechanism of tumor metastasis holds that tumor cell migration begins with a single tumor cell and progresses through a number of intricate procedures before arriving and surviving at distant tissues and organs. Moreover, it has been demonstrated that the collective tumor cell migration, similar to physiological cluster migration during gastrulation, is more resistant to clinical treatments and has a larger capacity for invasion than the individual tumor cell migration. This is supported by findings that stem-cell-related and proliferation-related genes are abundant in circulating tumor cell clusters, and embryonic pluripotency transcription factors such as OCT4, NANOG, SOX2, and SIN3A have hypomethylated binding sites ( Gkountela et al., 2019 ). Apart from the classical hematological and lymphatic spread, the metastatic migration of cancer cells along nerve sheaths is known as the perineural spread ( Maroldi et al., 2008 ) while the abluminal perivascular migration of cancer cells is known as angiotropism and pericytic mimicry ( Lugassy et al., 2014 ), recreating mechanisms of primitive cell migration that are specific to early embryonic development.

The concept of CSCs lays at the basis of tumor heterogeneity but the exact mechanism of their occurrence is still incompletely elucidated. Stem cell fusion has been described more than 130 years ago but has been in the focus of research only recently since it represents a common occurrence in cancer and also in mutation-harboring nonmalignant tissues. As accumulating data suggests, this cell rescue mechanism could repair highly functional cells and even reprogram them through a cloning-like process endowing the resulting hybrids with new capabilities that result from nuclear transfer. While stem cell fusion can bring new insights into the mechanisms that lead to the emergence of cancer stem cells, there are still missing pieces in the puzzle of tumor development since this repair mechanism by fusion has been shown to promote cell differentiation and restoration of normal replication and apoptosis in cancer cells ( Blau, 2002 ). What could trigger the generation of totipotent ZL-CSCs that recapitulate embryogenesis is not completely understood but may depend on the suspected ontogeny of the fusing cells. The basic hypothesis of this speculative analysis and emerging notion is that stem cell fusion attempts the restoration of genetic abnormalities in functional cells which brings a new set of integral genes but can also induce a cloning-like genomic reprogramming and abnormal histogenesis, if germline-related pluripotent stem cells become involved in the rescue process. When quiescent tissue resident germ line related pPSCs attempt to rescue mutation-harboring cells by fusion, a fecundation-like cell reprogramming could potentially occur with the emergence of zygote-like/unicellular-like CSCs that demethylate their developmental meiotic genes. This speculative concept of oncogenesis becomes appealing for both solid and hematological malignancies since it corroborates past and current hallmarks of cancer into one unified theory, opening new avenues for cancer research and therapy. To our knowledge, this concept has not been proposed explicitly. Implantation of a ZL-CSC in the host tissue should induce invasion, immune tolerance, neoangiogenesis and distant spread which is not much different from the reactions triggered by a developing embryo outside the uterus in ectopic tissues such as the fallopian tubes and more rarely in the ovary, cervix, or outside the uterus and its annexes in the abdominal cavity in omentum, intestines, liver, spleen, lesser sac, and retroperitoneal space ( Agarwal et al., 2014 ; Ren et al., 2022 ). Specific fusogens such as syncytins and IZUMO1 involved in physiological fusions, are likely involved in the pPSC-somatic fusion in postnatal tissues since these fusogens are also expressed in multiple tumor types ( Wang et al., 2023 ). In normal trophoblast formation and placentation, cell fusion leads to genomic alterations and extensive mutation ( Coorens et al., 2021 ). However, while multiple mechanisms can stabilize the genome and ensure healthy and coordinated development in the post-implantation embryo, a failed fusion driven cell rescue could have chromoplexy and chromothripsis as drivers of unstable genomic changes in cancer development ( Stephens et al., 2011 ; Liu, 2022 ). How does the fusion evade immune detection, and how does it differ from physiological fusion events like trophoblast formation is still an area of active research. Following genomic reprogramming and activation of the zygotic genome in the resultant hybrid, the ZL-CSC appears to hijack the developmental programme for its own growth and immunologic acceptance. With the experimental evidence of fusion events in human tumors ( Wang R. et al., 2016 ; Rengstl et al., 2013 ; Dittmar et al., 2009 ), it is becoming increasingly difficult to escape the conclusion that the cellular crisis leading to activation of the embryonic programme might not be spontaneous in a single cell, but could potentially involve a two cell fusion-driven reprogramming, akin to development. Whether somatic mutations are the root cause of cancer or merely its consequence is still debatable. Quiescent stem cells are not expected to harbor somatic mutations but cancer cells with genomic instability can acquire somatic mutation following their clonal expansion. When coding and noncoding genomic elements are taken into account, cancer genomes typically feature four to five driver mutations; nevertheless, 5% of malignant tumors have none at all ( Campbell et al., 2020 ). Aging is correlated with accumulation of somatic mutant clones and the activation of oncogenic mutations ( Martincorena et al., 2018 ), indicating that a certain amount of genomic instability can be controlled during both development and aging. Since driver mutations are often seen in benign tissue, they seem necessary but not sufficient to trigger oncogenesis. Perhaps fusion is the second hit leading to tumor transformation. Our speculation is that fusion of a germ line related quiescent stem cell with a mutation-harboring somatic cell is what might actually induce reprogramming and genomic instability in the resulting CSC. Therefore, our hypothesis is that mutation in the somatic cell is not only the trigger but also a consequence of altered gene expression regulation perpetuated by fusion in tumorigenesis. This offers a necessary paradigm shift in cancer treatment, from destroying a mutant cell to stopping a rogue developmental program. The overlap between embryogenesis and oncogenesis is undeniable so understanding the very start of life is essential to cure the disease that so often ends it. Currently there is no direct observation tracking how ZL-CSCs emerge. However, due to the role of transposable elements in development and their interactions with meiotic genes and pluripotency factors such as OCT4 and NANOG, SOX2, PROM1, along with their transcriptionally active state in PGCs, and the activation of such factors in many cancers, we believe our hypothesis is worth pursuing, warranting further investigation. Experimental validation of this concept could challenge current cancer origin models, link developmental biology and oncology in a novel way and create premises for anti-reprogramming therapies. It remains a challenge to define the role of cell rescue by stem cell fusion, whether it is primarily reparative or oncogenic. Because such type of cell rescue correlates with both favorable and unfavorable biological outcomes, this suggests that the process might be a recent evolutionary event that could still be ongoing. As it takes millions of years for evolution to complete, we might have the rare opportunity to behold a new trend in maintaining genetic integrity in somatic cells by “intracellular chimerism”.

Syncytium formation can be induced by specific types of viral infections, the most extensively described being the respiratory syncytial virus while cell-cell fusion is also documented in human immunodeficiency virus and herpes simplex virus leading to syncytium formation and enabling the viral genome transfer into neighboring cells ( Leroy et al., 2020 ). Syncytium formation by fused pneumocytes was also observed in the severe stages of SARS-CoV-2 infection by pathology reports ( Lin et al., 2021 ). The hallmark cell of Hodgkin’s lymphoma is represented by Hodgkin and Reed/Sternberg (HRS) cell that does not resemble any normal cell in the body. These are large, often multinucleated cells with a dystrophic morphology and an uncharacteristic immunophenotype ( Küppers and Hansmann, 2005 ). The HRS cell, usually derived from B lymphocytes, was shown to be the result of fusion events. Until recently, the classical notion through which HRS cells developed from mononucleated Hodgkin cells was thought to be the mechanism of endomitosis. However, using continuous single-cell tracking of Hodgkin lymphoma cell lines by long-term time-lapse microscopy, it has been established that cell fusion represents the primary mechanism of HRS cell generation ( Rengstl et al., 2013 ). The implications of small mononuclear cell fusion of cells with different ancestors in the proliferative compartment of the Hodgkin lymphoma tumor clone goes beyond Hodgkin disease since HRS-like cells are frequently seen in various other diseases such as infectious mononucleosis, some Non-Hodgkin Lymphomas (e.g., T-cell lymphomas, B-cell chronic lymphocytic leukemia/small lymphocytic lymphoma - CLL/SLL), and even in EBV-associated lymphoproliferative disorders ( Küppers and Hansmann, 2005 ). As HRS cell proliferation alone is insufficient for the expansion or maintenance of the HRS clone, it is arguable whether HRS cells have a pivotal role in pathogenesis or are simple reliques. While HRS cell generation implies cell membrane fusion of Hodgkin daughter cells or cousin cells which results in long-lived multinucleated low-proliferative giant cells, only small Hodgkin cells can maintain the HRS cell clone in culture. This is in contrast with the other population of growth-induced HRS cells which become polyploid through incompletely elucidated mechanisms and remain quiescent ( Rengstl et al., 2013 ). Stem cell fusion can lead to tumor heterogeneity through diverse genetic alterations. This heterogeneity can lead to the emergence of subpopulations of cells with distinct phenotypes contributing to the clonal evolution and adaptation of tumors, making them more challenging to target effectively ( Marusyk and Polyak, 2010 ). Ploidy abnormalities in malignant tumors are common occurrences and are now considered a hallmark of cancer. Polyploidy and whole genome doubling has been documented in about 30%–37% of malignant tumors using whole genome sequencing ( Bielski et al., 2018 ; Zack et al., 2013 ). While the mechanisms of polyploid cell generation in premalignant and malignant cells is still elusive, it has been argued that the extra set of chromosomes is not redundant and their genome may play a tumor suppressor role limiting tumor transformation by providing an additional set of tumor suppressor genes as a repair mechanism ( Zhang et al., 2018 ; Lin et al., 2020 ). However, polyploidy reduction and missegregation following mitosis in replicating cells may potentially represent an early step in the initiation of oncogenesis ( Matsumoto et al., 2021 ). As aneuploidy is observed in more than 90% of all solid tumors ( Williams and Amon, 2009 ), this raises questions on how the aberrant caryotype may be part of tumor suppression or tumor progression mechanisms and what could be the actual implications of cell-cell fusion events in these processes. Reprogramming specialized cells to a pluripotent state outside the physiological conditions of normal fecundation can be achieved artificially through a) induced pluripotency using transfection of pluripotency transcription factors (Yamanaka factors), b) fusion of somatic cells with embryonic stem cells or embryonic germ cells, and c) somatic cell nuclear transfer (cloning) ( Yamanaka and Blau, 2010 ). Zygote formation by nuclear transfer technique from a somatic cell into an enucleated oocyte creates a clone with the donor cell’s genome ( Tian et al., 2003 ). Non-enucleated oocytes can also be used for cloning leading to the disappearance of the oocyte DNA, but developmental abnormalities occur, caused by epigenetic remnants that alter gene expression regulation ( Rouillon et al., 2019 ). Activation of the totipotent embryonic genome requires accessing the higher-order chromatin structure by transient epigenetic programs which are specific to the zygote development of 2cell/4cell transcription state in which DUX4, ZSCAN4, TCSTV, and HERVL retroelement network play key roles ( Hendrickson et al., 2017 ; Zhou and Dean, 2014 ; Defossez et al., 2022 ). This is only achieved during normal fecundation since germ cells are the only cells that undergo reprogramming to totipotency to the zygote state under physiological conditions. However, they also give rise to induced pluripotent stem cells (iPSCs) under the appropriate conditions in vitro since they already possess critical reprogramming factors. Despite that primordial germ cells (PGCs) are committed to produce unipotent cells, they express many of the master regulatory factors that facilitate pluripotency ( Durcova-Hills et al., 2001 ). The germline specific gene Prmt5, which is pivotal for PGC development together with the other genes expressed in PGCs and involved in reprogramming such as OCT 3/4 , SOX 2 , and LIN28, have the potential to reprogram somatic cells into iPSCs in vitro that exhibit germline transmission ( Nagamatsu et al., 2011 ). This indicates that there is a functional interrelation between germ cell development and somatic cell reprogramming ( Hu et al., 2018 ). Induced from PGCs, embryonic germ cells (EGCs) harbor an epigenome resembling to that of PGCs ( Nagamatsu et al., 2013 ). Following fusion with somatic cells, EGCs have a key advantage in reprogramming the somatic genome over embryonic stem cells (ESCs). Unlike ESCs which maintain DNA methylation in their imprinted control regions (ICRs), EGCs possess erased DNA methylation in ICRs and upon fusion with the somatic cell, erasure of DNA methylation in the ICRs of the somatic genome occurs which reprograms the somatic cell to a zygote-like state ( Ficz and Reik, 2013 ). This cell-fusion mediated imprint erasure and reprogramming in the somatic genome is modulated by Ten-Eleven Translocation proteins Tet1 and Tet2 leading to accumulation of 5-hydroxymethylcytosine (5hmC) and DNA demethylation at several ICRs following fusion with EGCs ( Piccolo et al., 2013 ). In vivo, the closest postnatal equivalent of an EGC is an early germ cell ( Zwaka and Thomson, 2005 ). Persistent missmigrated/ectopic PGCs that fail to enter apoptosis and retain pluripotency, have been sown to reside in postnatal tissues being amenable to tumorigenesis later in life ( Stallock et al., 2003 ; Heaney et al., 2012 ). Whether fusion events become involved in the missmigrated PGC related tumorigenesis still remains elusive but essential aspects such as the generation of multiple phenotypic tumor sub-clones that contribute to intratumoral heterogeneity, transition between epithelial, mesenchymal, and stem-like states, high capacity for tumor initiation in vivo , strong resilience to environmental stress and resistance to therapy linked to dormancy and reprogramming pathways could potentially be achieved in vivo by a germ-somatic fusion event connecting oncofetal and embryonic reprogramming pathways as further discussed in the next section. Stem cells have traits that entail them with high survival capabilities under unfavorable environments which include (I) quiescence, (II) active DNA-repair mechanism, (III) efflux transporters for toxic agents, (IV) high detoxification metabolism, and (V) apoptosis reluctance. Such mechanisms are believed to also be employed by CSC, to evade anti-cancer drugs ( Ng et al., 2013 ). CSCs have pivotal roles in tumorigenicity, drug resistance as well as recurrence ( Beck and Blanpain, 2013 ). Among cancer cells, CSCs were shown to possess several traits, such as a slower cycling or dormancy in some contexts, anti-apoptosis, downregulation of anti-proliferative pathways, drug resistance and more efficient DNA damage repair capabilities, which make them the protagonists of tumor drug resistance and recurrence ( Batlle and Clevers, 2017 ). The American Association of Cancer Research (AACR) defined in 2006 a CSC as any cancer cell that possessed stem cell-like properties of multi/pluripotency and unlimited self-renewal and specifically highlighted that the definition of a CSC does not imply that is has to be that initial cell in the body that caused cancer or its provenience from the tissue resident stem cells affected by tumor transformation ( Clarke et al., 2006 ). The widely accepted dogma of CSCs occurrence is that CSCs may be derived from normal stem cells, which may be affected by cancer causing somatic cell mutations ( Takebe et al., 2015 ; Trosko, 2021 ). However, as recurrent tumors often exhibit different characteristics and phenotypes compared with the original tumors, another hypothesis about the origin of CSCs is that CSCs are derived from the fusion of stem cells and differentiated cells ( Dittmar et al., 2009 ). With the presence of genomic hybrids and polyploid cells in both premalignant and malignant tissues, fusion events have gained a special interest as hallmarks of oncogenesis ( Spandidos Publications, 2025a ). The exact moment and mechanism by which a somatic cell or a stem cell becomes a CSC is not completely understood. As it may take years for mutations to develop and accumulate in a somatic cell, a germline mutation is acquired upon fecundation and represents the starting point of cancer susceptibility syndromes which can cause cancer in newborns and children ( Pruteanu et al., 2020 ; Sarkozy et al., 2009 ; Carvalho et al., 2021 ). However, how a mutated cell acquires or rejects tumorigenicity is still unknown but cannot exclude the involvement of stem cell fusion events in this process. This is since various oncogenic driver mutations have been observed in both nonmalignant and malignant lesions. Selected examples include BRAF mutation, an activating mutation in about 50% of melanoma cases, also present in more than 50% of benign nevi and atypical melanocytic nevi ( Uribe et al., 2003 ). ALK and EGFR activating mutations in non-small cell lung cancer (NSCLC) are common occurrences in benign lung lesions ( Hyde et al., 2023 ; Kamata et al., 2016 ) while PTEN activating mutation in endometrial cancer is also reported in endometriosis–a nonmalignant pathology involving ectopic endometrial tissue ( Govatati et al., 2014 ) ( Spandidos-Publications, 2025b ). The presence of IDH–mutant progenitor cells in non–cancerous regions of the brain in glioblastoma patients indicates that it is subsequent events that may or may not lead to the tumor transformation of IDH-mutation harboring glial and oligodendrocyte progenitor cells migrated during embryogenesis ( Park et al., 2026 ). As oncogenic driver mutations in cancer are also seen in benign lesions, it is arguable whether these mutations are sufficient by themselves for malignant transformation, or whether subsequent events such as maybe fusion for cell rescue could potentially trigger genomic reprogramming and oncogenesis at some point during the development of the lesion. Such potential sequence of unfavorable events seems to be contrary to the main objective of repair mechanisms by fusion so another pivotal trigger could drive the oncogenic process. It becomes intriguing to speculate that the ontogeny of unanticipated but versatile stem cells involved in cell rescue by fusion could drive a twist of fate in the resulting hybrid ( Cismaru et al., 2022 ).

While multiple physiological processes involve cell fusion, the prototype of selective cell fusion is best described in gamete fusion which occurs naturally in egg fertilization by sperm cells in humans. This further triggers the next stages of embryogenesis from cleavage, morulation, blastulation, implantation, gastrulation, neurulation, which involve host interactions that permit invasion of the miometrium, angiogenesis, cell proliferation, cell migration inside the embryo and at distant maternal seeding sites (e.g., fetal-maternal microchimerism), immunological modulation, reactivity restriction, and release of trophic hormones by the trophoblast (e.g., chorionic gonadotropin) leading to the development and growth of the fetus and its acceptance by the mother’s organism ( Handschuh et al., 2007 ). These stages, regulated by demethylation of imprinted genes in the primordial stem cells of the blastocyst produce changes in the host tissues which provide nutritional support by host blood supply and immunological acceptance for the developing organism ( Rossant and Tam, 2022 ; Ruvinsky, 1999 ). Syncytium formation by multiple cell fusions of uninuclear cells also occurs naturally in the decidualization process during embryo implantation ( Wang and Dey, 2006 ). A spontaneous cell fusion into syncytia also occurs naturally in muscle fibers during myogenesis. This interplay ensures the growth, maintenance and repair of muscle fibers throughout the life of an individual ( Lehka and Rędowicz, 2020 ). Cells use an intrinsic DNA damage response network of repair mechanisms and cell cycle checkpoints for dealing with genomic insults caused by replication errors, environmental agents, ionizing radiation and genotoxic chemicals ( Giglia-Mari et al., 2011 ). If DNA cannot be restored, the damaged cell transits to a senescent state and signals distress through senescence-associated secretory phenotype (SASP) inflammatory cytokines to initiate their clearance by immune cells and recruit progenitor cells to repopulate the tissue ( Muñoz-Espín and Serrano, 2014 ; Wang et al., 2024 ). CXCL12/CXCR4 chemokine axis signaling via neighboring pericytic mesenchymal progenitors is pivotal for the recruitment of stem cells for tissue repair at the site of cell damage ( Cismaru et al., 2022 ). Stem cells play important roles in tissue repair and regeneration, by mechanisms which range from replacement of senescent cells through replication and differentiation, to secretion of trophic biomolecules, and even transfer of mitochondria, an evolutionary conserved phenomenon of mesenchymal stem cells (MSCs) to reestablish mitochondrial function in cells harboring mitochondrial dysfunctions ( Gomzikova et al., 2021 ). Accumulating evidence supports the notion of a new type of cell rescue by stem cell fusion which supplies a mutation-harboring cell with new genes (e.g., tumour-suppressor genes in cancer cells), or correcting, up to a point, damaged cells harboring genetic defects continuously throughout life ( Blau, 2002 ; Brukman et al., 2019 ; Pesaresi et al., 2019 ). Bone marrow-derived MSCs have the ability to differentiate into a variety of cell types depending on their plasticity and are a potential source for epithelial tissue repair by fusion ( Pesaresi et al., 2019 ). MSCs expressing pluripotency markers have been isolated from the bone marrow and other tissues suggesting origins in more primitive precursors ( Cismaru et al., 2020 ; Ratajczak, 2015 ; Ratajczak et al., 2017 ). Several studies have demonstrated their ability to repopulate the gastrointestinal tract in bone marrow transplanted patients or in animal models of gastrointestinal carcinogenesis. However, mechanism of MSC epithelial differentiation still remains unclear and controversial with trans-differentiation or fusion events being evoked. In a study investigating the ability of MSCs to acquire epithelial characteristics, the authors showed that human bone marrow-derived stem cells (BMSCs) acquire epithelial characteristics through fusion with gastrointestinal epithelial cells ( Ferrand et al., 2011 ). Neural stem cells derived from embryonic stem cells were shown to fuse with microglia and mature neurons both in vitro and in vivo , the hybrid cells retaining genetic and functional characteristics of both fused cells being able to differentiate into neurons and astrocytes ( Cusulin et al., 2012 ). Dystrophic cardiomyocytes were shown to fuse with BMSCs for cell repair and not with normal myocytes in mice ( Bittner et al., 1999 ), while neurons were shown to fuse with BMSCs in humans with hematologic malignancies and not healthy controls, both types of fusions resulting in intracellular chimerism. In another study, BMSCs fusion with mutation-harboring hepatocytes after BM stem cell transplantation (BMSCT) was also shown to regenerate liver by generation of hybrid cells containing both donor and host genes, consistent with polyploid genome formation by fusion of host and donor cells in the liver ( Vassilopoulos et al., 2003 ). While polyploidy reduction by mitosis may lead to missegregation and cell abnormalities in daughter cells ( Matsumoto et al., 2021 ), it was proposed that mutation-harboring cell rescue by fusion with stem cells in this “cloning-like” rescue mechanism may function as a genomic restorative process for highly specialized cells that rarely (if ever) replicate and are not amenable for replacement by new non-functionalized cells in organs such as brain, muscle or liver ( Blau, 2002 ). Cell fusion for specific reprogramming purposes has been explored for more than a half of century in the technique of generating antibodies of predefined specificity and has been used for the production of monoclonal antibodies ( Köhler and Milstein, 1975 ). In recent years, the fusion with stem cells has been used to induce cell reprogramming. It was shown that pluripotent stem cells (PSCs), when fused with somatic cells, have the dominant capability to reprogram the somatic genome leading to somatic cell nuclear reprogramming ( Ficz and Reik, 2013 ). Fusion allows the somatic cell to acquire homologous characteristics by genomic reprogramming and to dedifferentiate into PSCs ( Piccolo et al., 2013 ). Artificial laser-induced fusion between single hepatocellular cancer cells (HepG2) and human embryonic stem cells (hESCs) allowed for the generation of cancer stem cells (CSCs). CSCs represent a distinct subpopulation of neoplastic cells that have the capacity to promote tumor growth, support self-renewal, resist conventional therapies and drive metastasis and relapse. The occurrence of CSCs in solid tumors of a wide range of organs is firmly supported by the accumulating evidence in recent years ( Clarke et al., 2006 ; Dittmar et al., 2009 ; Chen et al., 2013 ). With the newly acquired cancer- and stem cell - like characteristics, the resulting pluripotent hybrid cells resulted by fusion express increased tumorigenicity and drug resistance ( Wang R. et al., 2016 ).