Introduction

MicroRNAs (miRNA) are small non-coding RNAs that post-transcriptionally regulate gene expression by affecting both translation and stability of complementary mRNAs [1]. Bioinformatics predictions indicate that mammalian miRNAs regulate 30–50% of all protein-coding genes; each miRNA can bind several mRNAs and each mRNA can be targeted by different miRNAs, thus giving rise to complex regulatory networks that take part in the regulation of almost all physiological pathways [2, 3]. As a consequence, miRNA mutations, dysregulation of their expression or dysfunction of miRNA biogenesis have a key role in different pathological processes, with oncogenesis as the most investigated field [4,5,6].

miRNA genes can be present singularly in the genome or clustered together, with their own promoter or hosted by coding or non-coding genes, thus generally sharing their transcription promoter [7]. They are transcribed by RNA polymerase II as long primary transcripts (pri-miRNAs), wherein miRNA sequences fold into hairpin structures, recognized and excised by Drosha and DGCR8, the microprocessor complex, generating 60–80 nt precursors (pre-miRNAs). Pre-miRNAs are exported to the cytoplasm where they are processed by Dicer in miRNA duplexes. Finally, the mature miRNA strand that could derive from the 5′ arm or 3′ arm of the pre-miRNA (miRNA-5p or miRNA-3p, respectively) is loaded onto an Argonaute protein within the RISC complex to bind and silence complementary mRNA targets [8].

Surprisingly, genomic distribution analysis revealed that the human X chromosome has a higher density of miRNAs when compared to autosomes [9]; in contrast, the Y chromosome has only 4 miRNA sequences, but not experimentally validated, with 2 produced by the PAR1 (pseudoautosomal region 1) shared by both sex chromosomes. The miRNA enrichment on the X chromosome was observed in several mammalian species, but does not extend to species other than mammals [9]. The evolutionary conserved higher density of miRNAs on the X chromosome may suggest that X-linked miRNAs could contribute to some X-related conditions, properties or functions in mammals.

It has long been known that the presence of two X chromosomes in females and only one X chromosome in males requires mechanisms to equalize gene dosage between sexes and relative to autosomes to avoid a potentially lethal double-dose of genes residing on the X chromosome [10]. This mechanism involves two processes: X chromosome inactivation (XCI), i.e., the silencing of almost all genes on the one X chromosome leading to partial functional X monosomy, and X chromosome upregulation leading to increased gene expression on the single active X chromosome in males or females. It has been shown that up to 15% of the X-linked genes escape permanent silencing (“escapees”), with large variability in their number and tissue distribution within a given individual and between individuals: the escape from silencing or skewed XCI allow the expression of some genes by both X chromosomes in females [11, 12]. Furthermore, during early female embryonic development, the process of XCI is random across alleles in all cells, i.e., occurs irrespective of the parental origin of X chromosome and is clonally maintained once established, thus resulting in females being a functional mosaic for the active X chromosome across cell types. The XCI is a multistep process, involving a mechanism of counting and choice of the chromosome that will start XCI. The entire course of action is directed by the X inactivation center (** gene KDM5C contributes to ovarian development via downregulating miR-320a. Hum Genet 136(2):227–239. https://doi.org/10.1007/s00439-016-1752-9 " href="#ref-CR19" id="ref-link-section-d60481855e611">19,20,21], even though the higher density of miRNAs mapped on the X chromosome and their recognized regulatory role in biological processes.

In this review, we analyze the literature and databases about X-linked miRNAs, aiming at understanding how miRNAs could contribute to emerging gender-biased functions and to highlight some gaps in the knowledge, particularly in terms of possible implications and pathological perspectives for X chromosome aneuploidy syndromes.

Map** of miRNA sequences on X chromosome

A general overview

The human X chromosome contains 10% of all miRNAs detected so far in the human genome. According to the miRbase database, 118 miRNAs are located on the X chromosome, with 62 classified as “with confidence”, i.e., experimentally validated. Exploring Ensembl and miRbase databases (https://www.ensembl.org/Homo_sapiens/; https://www.miRBase.org), positions of the miRNA sequences were mapped on the X chromosome as well as their genomic context, i.e., their possible position in host gene and transcribed strand. The full miRNA list is available as Supplementary Material, whereas some statistical parameters are reported in Table 1: approximately 70% of miRNA sequences are localized on Xq, the phylogenetically oldest part of the X chromosome [22]; more than the half (62.7%) are intragenic, suggesting a co-regulation of miRNAs and host genes, in particular when they are transcribed by the same strand of host gene, and indeed this is the most often the case. We focused our attention on the 62 miRNAs experimentally validated and they are reported in Fig. 1; most of them are clustered (77.3%), and hosted in either coding or non-coding genes (66%).

Table 1 Information about miRNA sequence map** on X chromosome, based on Ensembl and miRBase databases
Full size table
Fig. 1
figure 1

The figure was inspired by Pinheiro et al. (2011) [35]

Map of microRNA sequences on human X chromosome. Information about validated miRNA sequence position is based on Ensembl and miRBase databases. On the left, chromosome bands are indicated and on the right, the miRNA names, simplified by eliminating « hsa-miR- » and indicated only by the number. Some miRNAs are intragenic and the host gene is indicated (name of host gene). Several miRNAs are clustered (within 10 kbp) and are shown on the same line separated by ‘,’, with the exception of miR-514b (*) that is located within 10.3 kbp of adjacent cluster; miRNA name is repeated when it can be considered belonging to different clusters. Some miRNAs are mainly involved in immunity ( ), cancer ( ) and cardiovascular homeostasis ( ), as detailed in the text. LOC10798567, uncharacterized non-coding RNA; CTPS2, CTP synthase 2; CLCN5, chloride voltage-gated channel 5; HUWE1, HECT, UBA and WWE domain containing E3 ubiquitin protein ligase 1; EDA, ectodysplasin A; FTX, FTX transcript, XIST regulator; CHM, CHM Rab escort protein; HTR2C, 5-hydroxytryptamine receptor 2C; WDR44, WD repeat domain 44; SEPTIN6, septin 6; MIR503HG, miR-503 host gene; FGF13, fibroblast growth factor 13; LOC105373347, periphilin-1-like; LOC101928863, uncharacterized non-coding RNA; GABRE, gamma-aminobutyric acid type A receptor epsilon subunit; GABRA3, gamma-aminobutyric acid type A receptor alpha3 subunit; SNORA36A, smal nucleolar RNA, H/ACA box 36A; DKC, dyskerin pseudouridine synthase 1.

Full size image

Genes related to X-linked syndromes: any relationships with hosted miRNAs?

Some of the hosting coding genes are directly involved in X-linked syndromes. This is the case of HUWE1, an E3 ubiquitin ligase required for the ubiquitination and subsequent degradation of different targets, such as the anti-apoptotic protein Mcl1 [myeloid cell leukemia sequence 1 (BCL2-related)], the tumor suppressor p53, core histones and DNA polymerase beta; mutations in the HUWE1 gene are associated with X-linked syndromic cognitive disability, where researchers found copy number variations [23], that perhaps could have similarities with some of the neurocognitive challenges seen in some individuals with Turner syndrome. It has also been proposed that an increased propensity to harbor copy number variations should be present in Turner syndrome (45,X) [24], while no studies so far have investigated this in the other sex chromosome aneuploidy syndromes. Defects in the EDA gene, hosting miR-676 and encoding a type II membrane protein belonging to the tumor necrosis factor family, are a cause of ectodermal dysplasia, anhidrotic, which is also known as X-linked hypohidrotic ectodermal dysplasia [32,33,34]. A number of X-linked miRNAs impacting the immune system integrity and function potentiate the concept of the immunological advantage of females [reviewed in 35]. Among the X-linked miRNAs involved in immune regulation, miR-223 is probably the most studied so far. miR-223, which is also involved in cancer pathology, is expressed in the myeloid lineage in the bone marrow, and is a regulator of neutrophil differentiation from myeloid precursors; it is also a negative modulator of the inflammatory response [36, 37]. Other miRNAs map** on X chromosome are involved in hematopoietic lineage differentiation. miR-106 was shown to negatively control monocytopoiesis by targeting AML-1; intriguingly, all miRNAs of the miR-106–363 cluster have oncogenic potential and been implicated in human T cell leukemias, where they are overexpressed [38]. In addition, two clustered miRNAs, miR-424 and miR-503, together with miR-222, all X-linked, promote monocytic differentiation [39]. The involvement of X-linked miRNAs in gender-biased immunity is also supported by a study reporting differential expression of six X-linked miRNAs (miR-221, miR-222, miR-98, miR-532, miR-106 and miR-92a) in PBMC (peripheral blood mononuclear cells) between males and females affected by rheumatoid arthritis, an autoimmune disease that affects females three times more often than men [40].

An example of a miRNA-dependent, sex-specific regulation of immune responses and cancer immunosurveillance is represented by the PD-1/PD-L1 pathway. The PD1 receptor, by binding to PD-L1 (or PD-L2), induces T cells to undergo immunosurveillance against tumors; PD-1 is also considered an immune checkpoint key molecule. Intriguingly, a number of X-linked miRNAs directly targets the stability and/or the translation of PD-L1 transcript (i.e., miR-20a/b, miR-106 a/b, miR-513, miR-424), or interfere with the transcription factors modulating its expression, i.e., miR-18a/b targeting HIF and miR-221/222 targeting STAT3 [31].

Several forms of cancer show a gender difference in terms of incidence, prevalence, and response to therapy [41]. miR-221 and miR-222 are the most extensively studied in tumors of different origins, wherein they act as oncomirs (oncogenic miRNAs) controlling the development and progression of the tumor through the down-regulation of several key targets [42, 43]. Both miRNAs are located in a cluster on the X chromosome; their deregulation is a hallmark of several types of cancer, probably because one of their targets is the cell cycle regulatory protein p27Kip1/CDKN1B [44]. In contrast to miR-221 and miR-222, different studies point to a tumor suppressor role for miR-503: in different cancer cell lines, it has been shown to target modulators of the cell cycle, such as cyclin-dependent kinases (CCND1, D2, D3) and cyclins (D1 and D3); consistently, miR-503 was found down-regulated in cancer tissues [reviewed in 45]. However, a possible differential contribution of miR-503 to tumors between males and females can presently not be evaluated since the gender context was ignored in current studies. This consideration is applicable also to other X-linked miRNAs implicated in different types of cancer [35].

With regard to breast cancer, the most common cancer in women, two profiling studies, performed on very large patient cohorts, indicate that circulating miRNAs from the two X-linked miR-106a–363 and miR-532–502 clusters are promising non-invasive biomarkers for diagnosis since they were upregulated in patients in comparison to healthy volunteers [18], which is also differentially expressed in Turner syndrome [14]. Furthermore, a number of circulating miRNAs were differentially detected in 45,X Turner syndrome females as compared with 46,XX normal women, whose role should be further clarified [19].

Presence of an extra X-chromosome characterizes Klinefelter syndrome males, 1 in every 660 male births. Klinefelter syndrome is characterized by a 47,XXY karyotype in about 80–90% of men, whereas the remaining cases are represented by mosaicism, additional sex chromosome (e.g., 48,XXXY; 48,XXYY; 49,XXXXY) or X chromosome structural abnormalities. The low testosterone levels, hypergonadotropic hypogonadism observed in a high proportion of these men only explain a few of the classical characteristics of Klinefelter syndrome, such as tall stature, gynecomastia, while the frequent infertility encountered is a consequence of the dramatic demise of the seminiferous tubules which takes place early in life [76]. Also in the case of Klinefelter syndrome, the wide variable phenotypic spectrum and the different severity of symptoms suggest a role for epigenetic factors. Two different studies have profiled miRNAs in peripheral blood cells from a small group of non-mosaic Klinefelter syndrome patients and healthy men and found 89 [20] or 2 [21] miRNAs differentially detectable. Also in the case of Klinefelter syndrome, the functional contribution of miRNAs to the phenotype has not been explored.

Potential key X-linked genes responsible for the comorbidities seen in Turner syndrome, Klinefelter syndrome and other X chromosome aneuploidy syndromes have been identified. One example is represented by KDM6A, involved in germ cell development, differentially expressed, and methylated in both syndromes [14]; another example is KDM5C, which could play a role in the neurocognitive development of the syndromes [77]; TIMP1/TIMP3 genes are involved in the bicuspid aortic valve, probably accounting for higher incidence of aortic dissection in Turner syndrome [78]. The SHOX gene (short stature homeobox), located on Xp22.23, is the only X-chromosome gene that is truly linked with a phenotypic trait in Turner syndrome, i.e., short stature and skeletal growth [79]. The function is dosage dependent, thus causing short stature in Turner syndrome and increased stature in other X aneuploidy condition characterized by extra X chromosomes (47,XXX; 47,XXY; 48,XXYY) [80].

With regard to miRNAs, currently only speculations are possible. For example, taking into account the number of X-linked miRNAs with a role in immunity, it is paradigmatic that a plethora of autoimmunity stigmata are observed in Turner syndrome patients [81]; consistently, Klinefelter syndrome males (47,XXY) have a risk similar to female to develop not only systemic lupus erythematosus (SLE), an autoimmune disease with a striking female preponderance [82], but also other autoimmune conditions [83] and as such both missing an X chromosome (like in Turner syndrome) and having too many (like in Klinefelter syndrome) can lead to an increased risk of autoimmune disease. The number of X chromosomes, and the consequent altered X-linked genes and miRNAs dosage, is probably critical for the maintenance or the loss of the immune homeostasis. Again, a role for X-linked miRNA in infertility of Klinefelter syndrome men could be hypothesized, based on some data in the literature discussed above: miRNAs are essential for male fertility, as shown by spermatogonial knock-out of their processing enzyme Dicer [61, 84]; the highest proportion of X-linked miRNAs is observed in mouse testis; 86% of the X-linked miRNAs escape from X inactivation during meiosis and thus might be crucial for gene expression regulation during spermatogenesis [85]. Finally, we inquired also if the targetome of X-linked miRNAs was already known to be involved in diseases using OMIM as reference database (DAVID search): five out eight statistically significant items were “diabetes”; consistently, “Insulin signaling pathway” is one of significantly enriched pathways according to KEGG. These bioinformatics results should be experimentally investigated, taking into account that diabetes is seen with four to sixfold increased frequency in Turner syndrome and Klinefelter syndrome adult patients [75, 86].

Overall, focusing the research on miRNA contribution to Turner syndrome and Klinefelter syndrome phenotype would take the understanding of syndrome development and associated morbidities to a new level, likely also in terms of therapeutic approach. For example, systematic studies on a correlation between the absence or extra-copies of specific X regions containing miRNAs, with consequent dysregulation of their potential targets, and specific clinical features should be performed. Furthermore, miRNA profiling on different tissues other than blood cells could prompt functional analysis on those miRNAs differentially expressed between X-aneuploidy and normal individuals. These studies could also pave the way for the development of new therapeutic approaches (e.g., co-administration of GH or sex hormone with an “essential” synthetic miRNA in TS).

Conclusion and perspectives

Females appear to be equipped with a larger miRNA machinery than males: X chromosome contains an unexpected high number of miRNAs (118), in comparison to Y chromosome (4 miRNA sequences predicted) and an average of 40–50 localized on autosomes. Taking into consideration the regulatory power of these small non-coding RNAs, the scenario depicted above suggests a strong miRNA involvement in sex-specific phenotypes and some differences in pathogenetic mechanisms and/or pathological responses. This is also supported by the fact that also at cellular level, differences in response to stressful stimuli have been observed between cells carrying XX or XY chromosomes, with a possible direct role of miRNAs; as an example, miR-548am has been shown to act as a key modulator of sex difference in the susceptibility to mitochondria-based apoptosis in primary dermal fibroblasts [29]. However, most of the miRNA-based studies ignore the gender context. Some sex-biased expression of specific miRNAs could account for sex different pathophysiological conditions. This research field should be explored in the next future, to fill the gap between clinical data and our understanding of molecular mechanisms underlying some gender-biased diseases, also in terms of possible new therapeutic approaches. Noteworthy, to date most of the X-linked miRNAs have no described functions. In addition, in this review, only X-linked validated miRNAs have been considered, but they represent approximately the half of miRNA content of X chromosome; all the other miRNAs should be validated to have a comprehensive picture of their contribution to X chromosome properties and implication in the pathogenesis of X aneuploidy syndromes.