Reproductive functions of Kisspeptin/KISS1R Systems in the Periphery

The expression of Kiss1 and Kiss1r was first demonstrated in the rodent ovary [4]. To date, the expression of Kiss1/Kiss1r has been found in the ovaries of different animals, such as hamsters [20], mice [21], rats [22], chickens [23], cats [24], dogs [25], pigs [26], humans and marmoset primates [27]. Because ovarian Kiss1 mRNA is mainly expressed in rat granulosa cells during proestrus, granulosa cells are likely the main site of kisspeptin synthesis [28]. The LH surge may directly stimulate kisspeptin synthesis through LH receptors on granulosa cells [29], and prevention of the preovulatory gonadotropin surge can block the upregulation of ovarian Kiss1 expression [22]. The expression of ovarian Kiss1 mRNA shows a distinctive cell- and stage-specific pattern under regulation of LH [22, 29, 30], whereas Kiss1r mRNA expression remains low and does not significantly fluctuate with the oestrous cycle or gonadotropin treatment in rats [28‐30]. Interestingly, in both rodent and human growth follicles, kisspeptin is present in theca cells of the growing follicle; in preovulatory follicles, kisspeptin begins to appear in the basal cells of the granular layer; after ovulation, positive immunostaining can be observed in non-luteinized granulosa cells of newly ruptured ovulation follicles; and in the corpus luteum (CL), intense kisspeptin immunoreactivity can be detected in steroidogenic granulosa lutein cells, with a gradual increase with gradual maturation of the CL [22, 27]. These results demonstrate that kisspeptin and its receptor have a highly conserved expression pattern in rodent, monkey and human ovaries. The distribution of kisspeptin in the ovary has significant temporal and spatial specificity, suggesting that the kisspeptin/KISS1R system performs multiple functions at different physiological stages in the ovary.

The expression of ovarian Kiss1 mRNA gradually increases from infancy to adolescence [28]. The immature ovary shows negligible Kiss1 expression [22], and there is no significant difference in ovarian weight between Kiss1/Kiss1r-deficient mice and normal mice before puberty [31]. However, after puberty, the ovaries in Kiss1r^−/− and Kiss1^−/− mice shrink compared with those in control mice, likely due to the loss of kisspeptin-mediated regulation of follicular development, not defects in gonadotropin secretion because follicular development cannot be rescued by gonadotropin replacement [32]. In fact, although the role of the HPG axis cannot be completely ruled out, follicles at all stages and the CL are present in mice with targeted removal of kisspeptin and Kiss1r neurons (> 90%), suggesting that local kisspeptin in the ovary plays a very important role in follicle development [1].

Under conditions of a healthy nutrient supply, the administration of kisspeptin in the ovary reduces the number of antral follicles and increases the number of preovulatory follicles, and these structural changes can be reversed by the administration of the kisspeptin antagonist peptide 234 (P234). Furthermore, kisspeptin administration increases plasma anti-Mullerian hormone (AMH) in 6- and 10-month-old rats. AMH, a marker of ovarian reserve, is mainly secreted by secondary and small sinus follicles and can inhibit the activation of primordial follicles by negative feedback; moreover, P234 administration reduces plasma AMH levels in rats [33]. The FSH/follicle stimulating hormone receptor (FSHR) axis is responsible for follicular growth [34], but kisspeptin can block the increase in FSHR expression by isoproterenol (ISO, a β-adrenergic agonist). Collectively, kisspeptin negatively regulates the development of preantral follicles by inducing the production of AMH and reduces the sensitivity to FSH by inhibiting the induction of FSHR expression by sympathetic activators, thereby reducing the recruitment of primary follicles (Fig. 1a). In the future, an ovarian-specific Kiss1/Kiss1r knockout model will be established to further elucidate the role of kisspeptin in follicle development.

The addition of kisspeptin to FSH-rich medium for porcine cumulus-oocyte complexes (COCs) promotes oocyte maturation, indicating a direct effect of kisspeptin on oocytes [35], and the mechanism may involve upregulating the expression of C-MOS, growth differentiation factor 9 (GDF 9) and bone morphogenetic protein 15 (BMP 15) [36]. Even in the absence of cumulus cells, kisspeptin can increase the maturity of oocytes because Kiss1r is expressed in oocytes during in vitro maturation (IVM). Thus, kisspeptin may act continuously and directly on oocytes in an autocrine-paracrine manner. Interestingly, the absence of FSH results in failed oocyte maturation, even in IVM medium supplemented with kisspeptin, confirming a critical role of gonadotropins in the maturation of oocytes in vitro. Moreover, the addition of FSH to COCs induces a significant increase in Kiss1r expression, reflecting the permissive action of FSH on kisspeptin.

When a mouse oocyte acquires meiotic capacity, Kiss1 mRNA expression increases 82.2-fold [36]. However, when the oocyte progresses through the first and second divisions of meiosis (MII), Kiss1 mRNA expression decreases by 5.4- and 12-fold, respectively [36]. During the progression from germ-vesicle I to MII, the expression of Kiss1r remains stable. However, kisspeptin treatment fails to affect the percentage of MII eggs [36]. Therefore, the upregulation of Kiss1 expression may be related to the ability to undergo meiosis and may affect the recovery of meiosis but not the progression of MII. Taken together, these data suggest that the effect of kisspeptin on oocyte maturation may be accomplished through the regulation of meiosis (Fig. 1b).

The LH peak plays a crucial role in ovulation by inducing the upregulation of COX-2 and prostaglandin production [37]. The COX-2 inhibitor NS398 and the COX non-selective inhibitor indomethacin significantly inhibited Kiss1 mRNA expression in the rat ovary and decreased the efficiency of rat ovulation, suggesting that Kiss1 may be a downstream target of COX-2 (Fig. 1c). Furthermore, administration of prostaglandin E₂ can reverse the antagonism of indomethacin on kiss1 expression. The anti-progestin RU486 ameliorates ovulation defects caused by indomethacin but cannot reverse the regulation of ovarian Kiss1 expression [27], implying the existence of other pathways that regulate ovulation. In fact, the indispensable role of ovarian kisspeptin in ovulation is suspect because gonadotropins can induce ovulation in Kiss1-deficient mice with mild hypogonadism and in women with homozygous KISS1R mutations [38].

Kisspeptin stimulates progesterone secretion by rat luteal cells and by chicken and porcine granulosa cells. Our previous study showed that recombinant KP-10 significantly enhances basal and human chorionic gonadotropin (hCG)-induced progesterone levels in cultured rat luteal cells and upregulates the transcription of key steroidogenic enzymes (StAR, CYP11A, and 3β-HSD) [30]. Moreover, KP-10 promotes the secretion of progesterone by cultured chicken follicular granulosa cells in vitro, accompanied by the upregulation of StAR, CYP11A, and 3β-HSD expression [23]. In addition, KP-10 significantly enhances progesterone production and prevents the efflux of oestradiol from granulosa cells of porcine large follicles [23]. Furthermore, KP-10 increases the phosphorylation of the mitogen-activated protein kinase Erk1/2 but not of P38 MAPK and Akt in cultured rat luteal cells, suggesting that kisspeptin may stimulate progesterone secretion via the Erk1/2 signalling pathway in these cells [30]. However, treatment with KP-54 alone did not alter steroidogenesis or the expression of gonadotropin receptors [39], indicating that KP-54 may require gonadotropins to promote steroidogenesis [30] or that different kisspeptin isoforms (such as KP-10) may have different affinities for ovarian KISS1R [23].

Unlike progesterone, KP-10 does not promote the basal or hCG-induced secretion of oestrogen by rat luteal cells [30]. Currently, the best data on the effects of kisspeptin on luteal cell function are from luteinized granulosa cell cultures. KP-54 significantly augments the expression of oestrogen receptors alpha and beta (ESR1 and ESR2) in human granulosa lutein cells, suggesting that kisspeptin may increase sensitivity to oestrogen [39].

Additional studies have indicated that serum kisspeptin levels are significantly higher in women with polycystic ovary syndrome (PCOS), which is characterized by hyperandrogenism and ovulatory dysfunction [40]. Notably, serum levels of kisspeptin are negatively correlated with FSH but positively correlated with LH, testosterone and dehydroepiandrosterone (DHEA) [41]. Mouse KP-10 and KP-52 can significantly increase serum testosterone levels in mice [42]. Furthermore, ovary-derived kisspeptin has been shown to regulate the secretion of LH [43].

There are not only significant differences in the distribution of testicular kisspeptin and KISS1R between mammals and non-mammalian species but also diverse distribution patterns in the same or similar species [5, 44‐48] (summarized in Table 1). For example, a previous study reported kisspeptin and Kiss1r immunoreactivity in round spermatids in immature mice [5]. However, another study showed kisspeptin immunoreactivity mainly in Leydig cells and sperm cells at different stages, not in only round sperm cells [47]. Therefore, the different results in the same species may be related to the age of the experimental mice and largely influenced by experimental methods. For example, when the LacZ gene was inserted into the Kiss1 and Kiss1r alleles to allow β-galactosidase staining to detect gene expression, unique structural changes in sperm (deformation) resulted in inactivation of β-galactosidase after the round spermatid stage, making it impossible to determine whether kisspeptin is expressed in prolonged spermatid and spermatozoa [5].

In non-mammalian species, subcutaneous injection of synthetic Kiss1 pentadecapeptide can speed up spermatogenesis in prepubertal male chub mackerel [47]. In mammals, gene expression profiling revealed that the initiation of Kiss1/Kiss1r expression in mouse testis coincides with the formation of spermatozoa [5], suggesting a link between spermatogenesis and the testicular kisspeptin/Kiss1r system in mammals. In addition, kisspeptin exerts anti-metastatic effects by inhibiting cell chemotaxis and migration, which play important roles in the early stage of spermatogenesis [49]. Furthermore, in the late stage of spermatogenesis, KP-13 can induce human sperm motility changes and hyperactivation, possibly caused by the increase in sperm intracellular Ca²⁺ concentration ([Ca²⁺]i) [45]. The positive association between kisspeptin concentration in seminal plasma and semen quality supports the importance of the kisspeptin system in spermatogenesis [50]. However, peripheral kisspeptin may not be essential for spermatogenesis in mammals. First, Kiss1 and Kiss1r mutant mice still show low levels of spermatogenesis on a phytoestrogen diet [51]. Second, male patients with KISS1R mutations respond to exogenous hormonal therapy and successfully achieve fertility [52]. Collectively, testicular kisspeptin may not be necessary for mammalian spermatogenesis but is an important regulator of this process.

Androgens (mainly testosterone) are steroid hormones secreted by Leydig cells in the testes of males. Thus far, there is no verdict as to whether peripheral kisspeptin has an effect on androgen production in Leydig cells. First, the interruption of Kiss1 expression is associated with decreased testosterone levels in rats [53], and the kisspeptin antagonist P234 reduces the production of hCG-activated testosterone in vitro [54], but local injection of P234 does not alter plasma testosterone levels in adult rhesus monkeys [55]. Second, although the immortalized Leydig cell line MA-10 expresses Kiss1r, it does not respond to KP-10 stimulation [5]. In addition, Sertoli cells respond to kisspeptin and stimulate the production of androgen-binding protein (ABP), indicating a potential role of kisspeptin in ABP production [46].

In the human female genital tract, KISS1/KISS1R is mainly expressed in epithelial and stromal cells of the uterus but not of the myometrium [6]. In mice, Kiss1 and Kiss1r mRNA expression levels are generally low from day 1 to 4 of pregnancy, which is the stage of zygote to blastocyst transformation (Fig. 2a). KISS1 and KISS1R proteins are mainly localized at low levels in the luminal and glandular epithelium. However, Kiss1 and Kiss1r mRNA expression level dramatically increase with the progression of uterine decidualization, and attenuated expression of Kiss1 can significantly inhibit the expression of stromal cell decidualization markers, indicating that the kisspeptin/kiss1r system plays an important role in the decidualization process [56]. However, the underlying mechanism is unknown.

Calder et al. found that in kiss1 mutant mice, gonadotropin and oestradiol replacement could restore ovulation, mating, and fertilization but not lead to pregnancy; moreover, leukaemia inhibitory factor (Lif), a crucial cytokine required for implantation, is weakly expressed in these mice [57]. Lif secreted by the uterine glands promotes embryo-uterine communication and contributes to embryo attachment and decidualization [58, 59]. Oestrogen upregulates Lif expression in the uterus, and supplementation with Lif restores implantation and decidualization in ovariectomized mice and mice lacking uterine oestrogen receptor expression [60, 61]. Furthermore, in Kiss1 knockout mice, exogenous administration of Lif, but not E₂, partially rescues implantation failure [57], and our data demonstrated that E₂ significantly increases the expression of uterine kiss1 mRNA in ovariectomized mice [56]. These data suggest that kisspeptin signalling may act downstream of E₂ to stimulate uterine Lif expression and is beneficial for promoting embryo implantation and decidualization in mice (Fig. 2b).

There is evidence that the primary source of circulating kisspeptin is trophoblast cells of the placenta [12, 62]. In rat placental cells, Kiss1 expression is upregulated by GnRH and neurokinin B, and all of these neuropeptides can increase hCG expression [63]. Serum KP-54 levels increase several thousand fold during pregnancy and return to normal within 15 days after delivery, suggesting that the placenta produces large quantities of kisspeptin during pregnancy [4, 62, 64]. Moreover, low circulating kisspeptin levels during pregnancy are associated with an increased risk of miscarriage. Therefore, plasma kisspeptin levels are a potential biomarker for miscarriage in the first and third trimesters [65, 66]. As one of the biomarkers of pregnancy, peripheral kisspeptin has multiple functions, including the regulation of placental invasion and migration (discussed in detail below) [62], the apoptosis of embryonic and placental cells, and foetal development [67, 68].

Kisspeptin administration increases the apoptosis of embryonic cells cultured in vitro by upregulating pro-apoptotic genes [69]. The expression of the pro-apoptotic gene BAK1 in blastocysts increased 3.5-fold at 24 h after kisspeptin treatment, but no significant change was observed in the expression of the anti-apoptotic gene Bcl-2 [35]. In addition, the apoptosis index (AI), the ratio of the pro-apoptotic protein BAX to the anti-apoptotic protein Bcl-2, determines whether the cell will initiate apoptosis [70]. Interestingly, the AI and KISS1/KISS1R expression in the placenta are much higher in late pregnancy than at term delivery in humans [68]. Furthermore, external administration of kisspeptin increases AI and induces apoptosis in placental explants in a dose-dependent manner [68]. Taken together, these data indicate that kisspeptin may be a pro-apoptotic placental factor during pregnancy.

In addition, studies have indicated that the kisspeptin/KISS1R system in the embryo may affect human foetal adrenal function synergistically with adrenocorticotropic hormone and corticotropin-releasing hormone secretion by increasing the production of DHEA in mid to late gestation (Fig. 2c) [71, 72].

Kisspeptin was originally called metastin because it can inhibit tumour metastasis. Interestingly, the invasion processes of placental and tumour cells are markedly similar [73, 74]. The highest expression of Kiss1 and Kiss1r in gestational trophoblast cells is consistent with peak trophoblast invasion [62, 75]. Thus, kisspeptin is thought to inhibit trophoblast migration and invasion in the placenta. A series of studies demonstrated that kisspeptin can regulate trophoblast migration and invasion by a variety of mechanisms. First, kisspeptin stimulates Erk1/2 phosphorylation in trophoblast cells and inhibits the expression of matrix metalloproteinases (MMPs), such as MMP-2, thereby regulating placental invasion [74, 76]. Second, KP-10 inhibits the migration of HTR8SVneo cells by stimulating complex Erk1/2-GSK3β-FAK feedback interactions in vitro [77]. Third, kisspeptin suppresses angiogenesis by downregulating vascular endothelial growth factor A (Fig. 2b) [78]. In addition, the active kisspeptin/KISS1R system not only suppresses the migration of trophoblast cells but also inhibits their growth in placental explants [35].