Gene-dosage issues: a recurrent theme in whole genome duplication events.

Two recent studies have addressed the long-term consequences of whole genome duplications (WGD). Specifically, they analyzed transcriptomes of the plant Arabidopsis thaliana and of four salmonids to assess the impact of WGD on gene expression. These studies point to commonalities in gene expression adjustments after polyploidization that we outline and discuss below.

Dominant negative variants and cotranslational assembly of macromolecular complexes.

Pathogenic variants occurring in protein-coding regions underlie human genetic disease through various mechanisms. They can lead to a loss of function (LOF) such as in recessive conditions or in dominant conditions due to haploinsufficiency. Dominant-negative (DN) effects, counteracting the activity of the normal gene-product, and gain of function (GOF) are also mechanisms driving dominance. Here, I discuss a few papers on these specific mechanisms. In short, there is accumulating evidence pointing to differences between LOF versus non-LOF variants (DN and GOF). The latter are thought to have milder effects on protein structure and, as expected, DN variants are enriched at protein interfaces. This tendency to cluster in 3D space can help improve the ability of computational tools to predict the pathogenicity of DN variants, which is currently a challenging issue. More recent results support the hypothesis whereby cotranslational assembly of macromolecular complexes can buffer deleterious consequences of variants that would otherwise lead to DN effects (DNEs). Indeed, subunits the variants of which are responsible for DNEs tend to elude cotranslational assembly, thus poisoning complexes involving wild-type subunits. The constraints explaining why the buffering of DNEs is not universal require further investigation.

The forkhead DNA-binding domain binds specific G2-rich RNA sequences.

Transcription factors contain a DNA-binding domain ensuring specific recognition of DNA target sequences. The family of forkhead (FOX) transcription factors is composed of dozens of paralogs in mammals. The forkhead domain (FHD) is a segment of about 100 amino acids that binds an A-rich DNA sequence. Using DNA and RNA PCR-SELEX, we show that recombinant FOXL2 proteins, either wild-type or carrying the oncogenic variant C134W, recognize similar DNA-binding sites. This suggests that the oncogenic variant does not alter the intrinsic sequence-specificity of FOXL2. Most importantly, we show that FOXL2 binds G2-rich RNA sequences whereas it virtually fails to bind similar sequences in DNA chemistry. Interestingly, a statistically significant subset of genes responding to the knock-down of FOXL2/Foxl2 harbor such G2-rich sequences and are involved in crucial signaling pathways and cellular processes. In addition, we show that FOXA1, FOXO3a and chimeric FOXL2 proteins containing the FHD of the former are also able to interact with some of the preferred FOXL2-binding sequences. Our results point to an unexpected and novel characteristic of the forkhead domain, the biological relevance of which remains to be explored.

Forkhead Transcription Factors in Health and Disease.

Forkhead box (FOX) proteins belong to an evolutionarily conserved family of transcription factors that has evolved by gene/genome duplication. FOX family members have undergone sequence and regulatory diversification. However, they have retained some degree of functional redundancy, in addition to playing specific roles, both during development and in the adult. Genetic alterations or misregulation of FOX genes underlie human genetic diseases, cancer, and/or aging. In this review, we provide an updated overview of the main characteristics of the members of this family, in terms of breadth of expression, protein domain composition, evolution, and function.

A cellular model provides insights into the pathogenicity of the oncogenic FOXL2 somatic variant p.Cys134Trp.

BACKGROUND: FOXL2 is a transcription factor expressed in ovarian granulosa cells. A somatic variant of FOXL2 (c.402 C > G, p.Cys134Trp) is the hallmark of adult-type granulosa cell tumours. METHODS: We generated KGN cell clones either heterozygous for this variant (MUT) or homozygous for the wild-type (WT) allele by CRISPR/Cas9 editing. They underwent RNA-Seq and bioinformatics analyses to uncover pathways impacted by deregulated genes. Cell morphology and migration were studied. RESULTS: The differentially expressed genes (DEGs) between WT/MUT and WT/WT KGN cells (DEGs-WT/MUT), pointed to several dysregulated pathways, like TGF-beta pathway, cell adhesion and migration. Consistently, WT/MUT cells were rounder than WT/WT cells and displayed a different distribution of stress fibres and paxillin staining. A comparison of the DEGs-WT/MUT with those found when FOXL2 was knocked down (KD) in WT/WT KGN cells showed that most DEGs-WT/MUT cells were not so in the KD experiment, supporting a gain-of-function (GOF) scenario. MUT-FOXL2 also displayed a stronger interaction with SMAD3. CONCLUSIONS: Our work, aiming at better understanding the GOF scenario, shows that the dysregulated genes and pathways are consistent with this idea. Besides, we propose that GOF might result from an enhanced interaction with SMAD3 that could underlie an ectopic capacity of mutated FOXL2 to bind SMAD4.

DNA Content, Cell Size, and Cell Senescence.

A recent study assessed the impact of cell size on various cell properties. Oversized yeast cells display slow cell division, cytoplasmic dilution, and transcriptomic and proteomic alterations. It highlights commonalities between aging yeast and mammalian cells, suggesting the existence of a range of DNA content: cell volume ratio ensuring optimal function.

The Muller's Ratchet and Aging.

Aging entails an irreversible deceleration of physiological processes, altered metabolic activities, and a decline of the integrity of tissues, organs, and organ systems. The accumulation of alterations in the genetic and epigenetic spaces has been proposed as an explanation for aging. They result, at least in part, from DNA replication and chromosome segregation errors due to cell division during development, growth, renewal, and repair. Such deleterious alterations, including epigenetic drift, irreversibly accumulate in a stepwise, ratchet-like manner and reduce cellular fitness, similar to the process known as Muller's ratchet. Here, we revisit the Muller's ratchet principle applied to the aging of somatic cell populations and discuss the implications for understanding the origins of senescence, frailty, and morbidity.

Recurrent missense variants in clonal hematopoiesis-related genes present in the general population.

Clonal hematopoiesis (CH) consists in an abnormal expansion of a hematopoietic stem cell bearing an advantageous somatic variant. A survey of known recurrent somatic missense variants in DNMT3A, SF3B1, SRSF2, and TP53, some of the most prominent genes underlying CH of indeterminate potential (CHIP), in gnomAD noncancer database shows the presence of 73 variants. Many of them reach frequencies higher than 0.01% in various populations and, in many cases, are enriched in specific populations. Consistent with a potential involvement in CHIP, we found that the age distribution of the carriers is shifted towards old ages. Moreover, the variant allele frequencies are on average lower than 50%, expected for germline heterozygous variants. The pervasive presence of some of such variants in blood DNA from elder individuals is compatible with CHIP of somatic origin. On practical grounds, CHIP can lead to misclassification of somatic variants in cancer-predisposition genes as inherited, which bear consequences for the affected individuals and their families.

Predictable increase in female reproductive window: A simple model connecting age of reproduction, menopause, and longevity.

With the ever-increasing lifespan along with societal changes, women can marry and procreate later than in previous centuries. However, pathogenic genetic variants segregating in the population can lead to female subfertility or infertility well before the average age of normal menopause, leading to counter-selection of such deleterious alleles. In reviewing this field, we speculate that a logical consequence would be the later occurrence of menopause and the extension of women's reproductive lifespan. We illustrate this point with a simple model that applies to other variants that contribute to female infertility, including epigenetic variation. We also consider the effect of medical interventions and lifestyle.

Pathogenic "germline" variants associated with myeloproliferative disorders in apparently normal individuals: Inherited or acquired genetic alterations?

Myeloproliferative syndromes (MPS) are hematologic malignancies due to the expansion of an abnormal hematopoietic stem cell. They include chronic myeloid leukemia (CML) and non-CML MPS such as polycythemia vera, essential thrombocythemia and primary myelofibrosis. The latter are distinguished by somatic pathogenic variants affecting JAK2, CALR, or MPL genes. Apparent germline pathogenic variants have been reported in the general population. Here, we found that two gnomAD data-sets report more homozygotes than expected for the JAK2 c.1849G > T(Val617Phe) variant. We propose that somatic gene conversion can explain the presence of those unexpected homozygotes in normal populations. Consistently, homozygous individuals are older than 65 years. We also found a lower-than-expected frequency of the JAK2 variant in younger individuals suggesting that somatic mutation can underlie its presence in (at least some) heterozygotes. Regarding pathogenic variants in MPL and CALR, they are also present in the gnomAD data-sets explored. However, we cannot conclude that such seemingly germline variants are in fact somatic alterations. These results suggest that apparently normal individuals bearing MPS-related variants can be subclinical/undiagnosed MPS cases of somatic origin. It would be interesting to assess the hematologic phenotype of such individuals and the presence of the relevant variants in other tissues.

Genomic exploration of the targets of FOXL2 and ESR2 unveils their implication in cell migration, invasion, and adhesion.

FOXL2 and ESR2 are key transcriptional regulators in ovarian granulosa cells. To explore their transcriptional roles and their interplay, we have depleted Foxl2 and Esr2 in mouse primary granulosa cells to assess their ability to bind their targets and/or to modulate gene expression and cellular functions. We show that FOXL2 is involved in a large number of regulatory actions essential for the maintenance of granulosa cell fate. A parallel ChIP-seq analysis showed that FOXL2 mainly binds to sites located in intergenic regions quite far from its targets. A bioinformatic analysis demonstrated that FOXL2-activated genes were enriched in peaks associated with the H3K27ac mark, whereas FOXL2-repressed genes were not, suggesting that FOXL2 can activate transcription through binding to enhancer sites. We also identified about 500 deregulated genes upon Esr2 silencing, of which one third are also targets of FOXL2. We provide evidence showing that both factors modulate, through a coherent feed-forward loop, a number of common targets. Many of the FOXL2/ESR2 targets are involved in cell motility and, consistently, granulosa cells depleted for either Foxl2 or Esr2 exhibit decreased migration, invasion and adhesion. This effect is paralleled by the depletion of their target Phactr1, involved in actin cytoskeleton dynamics. Our analysis expands the number of direct and indirect transcriptional targets of both FOXL2 and ESR2, which deserve investigation in the context of adult-type granulosa cell tumors whose molecular diagnostic hallmark is the presence of the C134W FOXL2 pathogenic variant.

FOXL2 in adult-type granulosa cell tumour of the ovary: oncogene or tumour suppressor gene?

A recurrent mutation in FOXL2 (c.402C>G; p.C134W) is present in over 95% of adult-type granulosa cell tumours (AGCTs). In contrast, various loss-of-function mutations in FOXL2 lead to the development of blepharophimosis, ptosis, and epicanthus inversus syndrome (BPES). BPES is characterised by an eyelid malformation often accompanied with primary ovarian insufficiency. Two recent studies suggest that FOXL2 C402G is a gain- or change-of-function mutation with altered DNA-binding specificity. Another study proposes that FOXL2 C402G is selectively targeted for degradation, inducing somatic haploinsufficiency, suggesting its role as a tumour suppressor. The latter study relies on data indicative of an FOXL2 allelic imbalance in AGCTs. Here we present RNA-seq data as genetic evidence that no real allelic imbalance is observed at the transcriptomic level in AGCTs. Additionally, there is no loss of protein expression in tumours harbouring the mutated allele. These data and other features of this mutation compared to other oncogenes and tumour suppressor genes argue strongly against FOXL2 being a tumour suppressor in this context. Given the likelihood that FOXL2 C402G is oncogenic, targeting the variant protein or its downstream consequences is the most viable path forward to identifying an effective treatment for this cancer. © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

A Fresh Look at 'Aging' Proteins.

Recent analyses of the degradation profiles of thousands of proteins by McShane et al. have shown that many proteins are less stable during the hours following their synthesis. Many of such nonexponentially degraded (NED) proteins are components of macromolecular complexes. This may explain why, in cases of trisomy, the effect of overexpression can be attenuated.

Insights into the pathogenicity of missense variants in the forkhead domain of FOX proteins underlying Mendelian disorders.

Forkhead box (FOX) proteins are members of a conserved family of transcription factors. Pathogenic variants in FOX genes have been shown to be responsible for several human genetic diseases. Here, we have studied the molecular and structural features of germline pathogenic variants in seven FOX proteins involved in Mendelian disorders and compared them with those of variants present in the general population (gnomAD). Our study shows that the DNA-binding domain of FOX proteins is particularly sensitive to damaging variation, although some family members show greater mutational tolerance than others. Next, we set to demonstrate that this tolerance depends on the inheritance mode of FOX-linked disorders. Accordingly, genes whose variants underlie recessive conditions are supposed to have a greater tolerance to variation. This is what we found. As expected, variants responsible for disorders with a dominant inheritance pattern show a higher degree of pathogenicity compared to those segregating in the general population. Moreover, we show that pathogenic and likely pathogenic variants tend to affect mutually exclusive sites with respect to those reported in gnomAD. The former also tend to affect sites with lower solvent exposure and a higher degree of conservation. Our results show the value of using publicly available databases and bioinformatics to gain insights into the molecular and structural bases of disease-causing genetic variation.

One Hundred Years of Gene Balance: How Stoichiometric Issues Affect Gene Expression, Genome Evolution, and Quantitative Traits.

A century ago experiments with the flowering plant Datura stramonium and the fruit fly Drosophila melanogaster revealed that adding an extra chromosome to a karyotype was much more detrimental than adding a whole set of chromosomes. This phenomenon was referred to as gene balance and has been recapitulated across eukaryotic species. Here, we retrace some developments in this field. Molecular studies suggest that the basis of balance involves stoichiometric relationships of multi-component interactions. This concept has implication for the mechanisms controlling gene expression, genome evolution, sex chromosome evolution/dosage compensation, speciation mechanisms, and the underlying genetics of quantitative traits.

Conventional and unconventional interactions of the transcription factor FOXL2 uncovered by a proteome-wide analysis.

Beyond the study of its transcriptional target genes, the identification of the various interactors of a transcription factor (TF) is crucial to understand its diverse cellular roles. We focused on FOXL2, a winged-helix forkhead TF important for ovarian development and maintenance. FOXL2 has been implicated in diverse cellular processes, including apoptosis, the control of cell cycle or the regulation of steroid hormone synthesis. To reliably identify partners of endogenous FOXL2, we performed a proteome-wide analysis using co-immunoprecipitation in the murine granulosa cell-derived AT29c and the pituitary-derived alpha-T3 cell lines, using three antibodies targeting different parts of the protein. Following a stringent selection of mass spectrometry data on the basis of identification reliability and protein enrichment, we identified a core set of 255 partners common to both cell lines. Their analysis showed that we could co-precipitate several complexes involved in mRNA processing, chromatin remodeling and DNA replication and repair. We further validated (direct and/or indirect) interactions with selected partners, suggesting an unexpected role for FOXL2 in those processes. Overall, this comprehensive analysis of the endogenous FOXL2 interactome sheds light on its numerous and diverse interactors and unconventional cellular roles.

High-throughput Exploration of the Network Dependent on AKT1 in Mouse Ovarian Granulosa Cells.

The PI3K/AKT signaling pathway is known to regulate a broad range of cellular processes, and it is often altered in several types of cancers. Recently, somatic AKT1 mutations leading to a strong activation of this kinase have been reported in juvenile granulosa cell tumors. However, the molecular role of AKT1 in the supporting cell lineage of the ovary is still poorly understood. To get insights into its function in such cells, we depleted Akt1 in murine primary granulosa cells and assessed the molecular consequences at both the transcript and protein levels. We were able to corroborate the involvement of AKT1 in the regulation of metabolism, apoptosis, cell cycle, or cytoskeleton dynamics in this ovarian cell type. Consistently, we showed in established granulosa cells that depletion of Akt1 provoked altered directional persistent migration and increased its velocity. This study also allowed us to put forward new direct and indirect targets of the kinase. Indeed, a series of proteins involved in intracellular transport and mitochondrial physiology were significantly affected by Akt1 depletion. Using in silico analyses, we also propose a set of kinases and transcription factors that can mediate the action of AKT1 on the deregulated transcripts and proteins. Taken altogether, our results provide a resource of direct and indirect AKT1 targets in granulosa cells and may help understand its roles in this ovarian cell type.

How the most common mitochondrial DNA mutation (m.3243A>G) vanishes from leukocytes: a mathematical model.

Mitochondrial diseases may be caused by alterations of the mitochondrial genome. The pathogenic variant m.3243A>G is one of the most frequent causes of mitochondrial disease and the most common mitochondrial DNA mutation. Patients with a variant in mitochondrial DNA can have a mixture of mutated and wild-type genomes (heteroplasmy). In the case of the pathogenic variant m.3243A>G, the degree of heteroplasmy (H) correlates to some extent with the severity of the disease. Several longitudinal studies, where H is measured at two different time-points, have shown an annual decline in leukocyte H values. Thus far, only an exponential decay of H with time has been noted but a mechanistic model is lacking. Here, I describe a deterministic mathematical model that accounts for the decline of H in leukocytes based on selective mechanisms acting at the stem cell level. The 'inverted-sigmoid' model provides estimates of at-birth H levels closer to those observed in post-mitotic tissues, such as skeletal muscle, than the estimates provided by an exponential decay. The new model never leads to predictions of H > 100% and provides a stronger correlation between at-birth H values in leukocytes and the scores of the Newcastle Mitochondrial Disease Scale for Adults, which can be of practical utility. This model could be extended to other mitochondrial DNA disease-causing variants.

Gene Duplicates: Agents of Robustness or Fragility?

A recent analysis of paralog deletion in yeast has shown that functional compensation and dependency occur equally. While theory predicts that compensation produces robustness, the authors hypothesized that stabilization of one paralogous protein through interaction with its copy could explain dependency. Here, we provide alternative explanations, such as selection for increased protein dosage and hypofunctionalization.