The role of the gubernaculum in the descent and undescent of the testis

Abstract

Testicular descent to the scrotum involves complex anatomical rearrangements and hormonal regulation. The gubernaculum remains the key structure, undergoing the ‘swelling reaction’ in the transabdominal phase, and actively migrating out of the abdominal wall to the scrotum in the inguinoscrotal phase. Insulin-like hormone 3 (Insl3) is the primary regulator of the first phase, possibly augmented by Müllerian inhibiting substance/anitmüllerian hormone (MIS/AMH), and regression of the cranial suspensory ligament by testosterone. The inguinoscrotal phase is controlled by androgens acting both directly on the gubernaculum and indirectly via the genitofemoral nerve, and release of calcitonin gene-related peptide from its sensory fibres. Outgrowth of the gubernaculum and elongation to the scrotum has many similarities to an embryonic limb bud.

Cryptorchidism occurs because of both failure of migration congenitally, and failure of elongation of the spermatic cord postnatally. Germ cell development postnatally is disturbed in congenital cryptorchidism, but our current understanding of germ cell biology suggests that early orchidopexy, around 6 months of age, should provide a significant improvement in prognosis compared with a previous generation. Hormone treatment is not currently recommended. Acquired cryptorchid testes may need orchidopexy once they no longer reach the scrotum, although this remains controversial.

Introduction

Testicular descent is a complex process that only occurs in mammals. The location of the scrotum varies between species, as does the scrotal development. In the human, however, baby boys are born with fully descended testes in a pendulous scrotum, which reaches half way down the thigh. The laxity of the tissues is related to prenatal hormones, as a few weeks later the scrotum appears smaller. Postnatal testosterone levels are high between 2–4 months, but after this the scrotum shrinks and the testes are quite retractile until the onset of puberty in the second decade.

Given the complexity of the developmental process, it is no surprise that undescended or cryptorchid testes are common anomalies. Orchidopexy is one of the most common operations in childhood, and this is a significant financial as well as emotional burden for parents. Long-term outcomes after orchidopexy also remain problematic, with decreased fertility and increased risk of testicular cancer still serious public health issues. To address these clinical problems we need a deeper understanding of the normal process, and the reasons for abnormality, with the long-term aim of developing new therapeutic approaches.

Normal testicular descent

It is now accepted that normal testicular descent occurs in two separate stages that have different anatomy and hormonal control [Hutson and Hasthorpe, 2005]. After the onset of sexual differentiation in humans at 7–8 weeks’ gestation, the developing testis in the male secretes testosterone and Müllerian inhibiting substance/anti-müllerian hormone (MIS/AMH). MIS/AMH causes the Müllerian ducts to regress, so that the boy will not form fallopian tubes, uterus and upper vagina. The androgen is secreted in a probably exocrine manner down the Wolffian duct and triggers its development into epididymis and vas deferens, as well as formation of a distal bud to become the seminal vesicle [Tong et al. 1996]. By contrast, in the female the lack of MIS/AMH allows Müllerian duct development to proceed, to complete development into female internal genitalia under the influence of maternal and fetal oestrogens. The Wolffian duct regresses in the absence of androgens.

Between 10-15 weeks of development the testis remains near the future inguinal canal, while in females the ovary ascends away from the groin as the abdominal cavity enlarges [Shono et al. 1994]. Around 25-28 weeks the testis descends rapidly through the inguinal canal, which has just formed, and then migrates across the pubic region and down into the scrotum, arriving there about 35–40 weeks [Heyns and Hutson, 1995].

The key to understanding the anatomy of testicular descent is the ligamentous attachments of the urogenital ridge. Cranially the upper poles of the regressing mesonephros and developing gonad are anchored by the cranial suspensory ligament [CSL] [Van Der Schoot and Elger, 1993]. Caudally the genito-inguinal ligament, or ‘gubernaculum’, connects the lower pole of the gonad and epididymis to the future inguinal canal. The gubernaculum was so-named by John Hunter in the eighteenth century because he thought that it steered the testis to the scrotum. Prior to descent the gubernaculum ends in the inguinal abdominal wall, with the oblique and rectus muscles forming around it [Backhouse, 1982]. The gubernaculum has its own nerve supply, the genitofemoral nerve (GFN), descending on the anteromedial surface of the psoas muscle from L1-L2 segments and exiting through the back of the future inguinal canal, to reach the scrotum and groin skin.

The processus vaginalis develops as a diverticulum of the peritoneal membrane inside the gubernaculum, dividing it into three parts. The outer rim of mesenchyme is where the cremaster muscle forms, while the central column or cord attaches to the caudal epididymis and testis. Caudal to the processus vaginalis is the solid tip of the gubernaculum which contains abundant undifferentiated mesenchyme and glucopolysaccharides. It is quite bulky, and is about the same size as the testis [Barteczko and Jacob, 2000].

Enlargement of the distal gubernaculum or ‘bulb’ during the transabdominal phase produces the ‘swelling reaction’, making the male gubernaculum short and fat. This swelling reaction anchors the testis to the future inguinal canal. By contrast, in the female there is no swelling reaction and the gubernaculum merely elongates in proportion with the enlarging abdominal cavity. This produces relative descent of the testis compared with the ovary, which is actually ascending away from the internal inguinal ring [Shono et al. 1994]. It is only the growth of the uterus and broad ligament which keeps the ovary in the pelvis in humans, but its connection to the inguinal canal is indirect via the gubernaculum forming the ligament of the ovary and the round ligament [Miller et al. 2004].

The primary hormone regulating transabdominal descent is insulin-like hormone 3 (Insl3), also known as relaxin-like factor (RLF). Insl3 was first recognised in cloning studies using testis in the 1990s [Agoulnik, 2007], and is part of the insulin superfamily. A pre-protein of 131 amino acids is processed into an active molecule comprising an A-chain (26 AAs) and a B-chain (32 AAs). Insl3 is secreted by the Leydig cells and stimulates the swelling reaction by the LGR9 receptor on the gubernaculum [Fu et al. 2004]. Synthesis of Insl3 continues through life and putative functions include a role in spermatogenesis, but some functions probably are still unknown.

The function of Insl3 was discovered when a transgenic mouse with a disrupted Insl3 gene unexpectedly was found to have intra-abdominal undescended testes [Nef and Parada, 1999; Zimmermann et al. 1999]. In this mutant the gubernaculum is feminine like a round ligament. In vitro studies confirm that Insl3 stimulates mitosis in the gubernaculum, and also suggests a supplementary role for MIS/AMH [Scott et al. 2005]. The role of Insl3 in cryptorchidism remains uncertain, as most studies show a low rate of Insl3/LGR8 mutations in boys with undescended testes [Foresta et al. 2008].

The second or inguinoscrotal phase of descent requires the gubernaculum to change from a relatively inert, static structure ending in the inguinal muscles into an elongating, migrating organ that extends across the pubis and into the scrotum in the perineum. The anatomy of the gubernaculum varies between species during this phase, with the bulky, gelatinous bulb of the rodent shrinking prior to migration. By contrast, in the pig and the human the gubernacular bulb remains bulky and gelatinous until after migration is complete. Indeed, in premature babies, the bulky gubernaculum is still palpable below the testis for a few weeks after it has arrived in the scrotum. The distance required for the gubernaculum to transverse is considerable, being more than 4 cm in many fetuses, when the gubernaculum itself is only 1 cm in diameter. In rodents, the intra-abdominal gubernaculum partly everts at the onset of inguinoscrotal descent, which occurs postnatally, but eversion alone cannot account for the distance to the scrotum [Lam et al. 1998].

As the gubernaculum contains a peritoneal diverticulum, processus vaginalis, intra-abdominal pressure plays a supplementary role in inguinoscrotal descent [Attah and Hutson, 1993]. Conditions with lower abdominal pressure, such as gastroschisis, have a significant incidence of cryptorchidism. In this case, not only is abdominal pressure in the processus reduced, but also the gubernacular cord is often torn, enabling the testis to prolapse out of the defect along with the bowel [Pham et al. 2005].

To reach the scrotum, the gubernaculum needs several key properties:

1. a mechanism for growing and elongating

2. a directional signal to ensure migration in the correct direction

3. a mechanism for closing the processus vaginalis after testicular descent is complete.

It was once thought that the testis was pushed out of the abdominal cavity by passive pressure, which implied that the cremaster muscle was an elongated segment of the internal oblique muscle [Shenker and Hutson, in press]. This notion envisaged the gubernaculum as an inert, gelatinous ligament that was stretched out in a pulsion diverticulum.

The most recent evidence suggests that the gubernaculum, far from being an inert ligament, actually acquires specific growth properties similar to an embryonic limb bud [Nightingale et al. 2008; Tickle, 2003]. We have found that the site of maximum cell division is in the caudal tip of the bulb [Hrabovszky et al. 2002], and that both the processus vaginalis and the cremaster muscle grow maximally from their distal end [Harnaen et al. 2007]. Using fluorescent vital dyes taken up into cell membranes, the distal gubernaculum bulb contains an active zone of proliferation analogous to a “progress zone” in an embryonic limb bud [Na et al. 2007]. In addition, the tip of the gubernaculum can be amputated and grafted, suggesting that this is the site of maximal growth [Ng et al. 2009].

As the genetic programmes in limb buds, branchial arches and genital tubercle are all closely related, it suggests that body-wall outgrowths are all regulated by a common set of genes [Tickle, 2003]. We examined gene expression in wholemounts of mouse fetuses and found that Fgf10 and Hoxa10 were expressed in the gubernaculum and the inguinal mesenchyme immediately superficial to the bulb [Nightingale et al. 2008]. Not only are these genes known to be active in limb development, but also Hoxa10 mutant mice have undescended testes [Satokata et al. 1995]. Indeed, we have found that the unguinoscrotal migration of the gubernaculum is absent in these mice (unpublished). In the Long-Evans cryptorchid rat (ORL), there are also genetic defects in muscle and cytoskeletal development [Barthold et al. 2008].

For many years it was known that androgens control inguinoscrotal descent, as this phase is completely absent in girls with complete androgen insensitivity, where the androgen receptor is non-functional [Hutson, 1985]. However, the way androgens worked remained enigmatic, particularly as unilateral cryptorchidism did not seem easy to explain with a generalised hormonal deficiency. The clue to how androgen regulates the gubernaculum came in the 1980s, when we repeated an experiment first described in 1948 by Lewis [Lewis, 1948]. He transected the GFN in neonatal rats to test whether the cremaster might pull the testis to the scrotum. He reasoned that denervation should block this, and, indeed, the rats were cryptorchid. We postulated that cutting the nerve could block the effect of a circulating hormone if the nerve was acting as a ‘second messenger’ for the hormone [Beasley and Hutson, 1987; Hutson and Beasley, 1987].

On a search for a neuropeptide that might act as a second messenger for androgen we found calcitonin gene-related peptide (CGRP) in the GFN [Anderson and Seybold, 2004]. After 20 years of study, we now have good evidence that CGRP is a key regulator of gubernacular development in the inguinoscrotal phase [Hutson and Hasthorpe, 2005]

CGRP is a 37-amino acid peptide generated by alternative splicing of the calcitonin gene messenger RNA [Rosenfeld et al. 1983]. In the nervous system, CGRP has numerous sensory, motor and autonomic functions. In the rodent gubernaculum, CGRP causes rhythmic contractility (similar to cardiac muscle) in organ culture, and also stimulates mitosis and suppresses apoptosis [Goodman and Iversen, 1986]. Intriguingly, CGRP also produces a concentration gradient that can provide a chemotactic signal to the gubernaculum, suggesting that this may provide key directional signal in migration to the scrotum [Yong et al. 2008].

Obliteration of the processus vaginalis after testicular descent is once of the functions of CGRP that has been demonstrated in humans [Hutson et al. 2000]. Indeed, one day inguinal hernia might be treated by a depot injection of CGRP rather than by herniotomy.

Androgens appear to act both directly and indirectly on the gubernaculum, via the GFN releasing CGRP from its sensory nerve endings [Schwindt et al. 1999]. However, it is not clear how androgens trigger the ligamentous gubernaculum to become an outgrowth like a limb bud. Limb buds require mesenchyme with the right segmental programming and nerves to innervate the developing structures, but also provide some trophic signals. Most importantly, the initiating signals come from the ectoderm: the apical ectodermal ridge (AER), which is highly specialised ridge of ectoderm on the dorsoventral boundary of the embryo. By contrast, in the gubernaculum the only nearby skin is the inguinal skin just outside the external inguinal ring. Interestingly, this is where the scrotum forms in marsupials, such as the wallaby and kangaroo [Renfree and Short, 1988]. We are currently exploring the possibility that the inguinal skin may provide some initiating signals to stimulate the gubernaculum to grow out of the abdominal wall.

Cryptorchidism

All our work on testicular descent would suggest that cryptorchidism will be caused by any breakdown of the complex anatomical and mechanical process [Hutson and Hasthorpe, 2005]. Intra-abdominal testes, secondary to failure of the swelling reaction, is uncommon: this is likely because Insl3 mutations may interfere with spermatogenesis, and therefore lead to infertility [Adham and Agoulnik, 2004]. Moreover, the trans-abdominal phase is a simple mechanism that would not go wrong often. This is contrasted with the high frequency of palpable undescended testes in the groin, where inguinoscrotal migration has failed, likely because the mechanism is so complex [Virtanen and Toppari, 2008]. There is speculation that inadequate androgens, or environmental synthetic oestrogen analogues may cause cryptorchidism, but the issue remains controversial [Fisher et al. 2003]. Certainly, fetal pituitary anomalies lead to cryptorchidism, and placental failure, with insufficient human chorionic gonadotrophin (hCG), may also be a cause [Ferlin et al. 2007]. Nevertheless, the high frequency of unilateral cryptorchidism suggests that aberrations of the ipsilateral GFN regulation of gubernacular migration may be important. We have proposed that perineal testicular ectopia is caused by abnormal chemotaxis from the GFN [Hutson and Hasthorpe, 2005]

Recent studies suggest that acquired cryptorchidism may be common, although this is still not accepted universally [Hack et al. 2007; Rabinowitz and Hulbert, 1997]. We proposed some years ago that acquired maldescent was secondary to failure of the spermatic cord to elongate as the boy grows, as the scrotum moves away from the groin with age [Smith et al. 1989]. The likely explanation is suboptimal CGRP from the GFN in the perinatal period, leaving the processus vaginalis behind, although often obliterated [Clarnette et al. 1997]. Seen in this light, the acquired maldescent is the least severe anomaly in the spectrum of inguinal hernia and hydrocele [Clarnette et al. 1997].

Postnatally the primitive germ cells in the testis undergo a series of developmental transformations that convert the neonatal gonocyte into a ‘A’ then ‘B’ type spermatogonium and then into a primary spermatocyte [Huff et al. 2001]. Evidence now supports the notion that type-A spermatogonia, which appear between 3 and 12 months of age, are the putative stem cells for postpubertal spermatogenesis [Franke et al. 2004; Maclaughlin and Donahoe, 2004]. Importantly, transformation of gonocytes at 3–6 months of age is impaired in cryptorchidism, which is the likely cause of subsequent infertility, because of a deficient stem cell pool [Kolon et al. 2004]. Persistence of neonatal gonocytes is a likely cause of an increased cancer risk, as testicular tumours carry markers of gonocytes [Hasthorpe et al. 2007].

It has been suggested in the literature that gonocyte transformation is caused by transiently high serum androgens during a ‘mini-puberty’ at 3–6 months of age [Hadziselimovic et al. 2005]. This is the rationale for recommending hormone treatments that stimulate androgen production, such as hCG and LHRH [Hadziselimovic et al. 1984]. Our own work suggests that androgens are not involved, as gonocytes transform normally in the testicular feminisation (TFM) mouse, with a nonfunctional androgen receptor [Zhou et al. in press]. In addition, there is some evidence that MIS/AMH may be involved [Cortes et al. 2003], but there may be other regulatory factors that remain unknown.

In many parts of Europe, hormone treatment for cryptorchidism has become less fashionable, as a recent consensus suggested that it was not only of low efficacy, but also may be potentially detrimental to germ cell development [Thorsson et al. 2007; Cortes et al. 2000]. In particular, hormone treatments that stimulate androgens may induce pubertal changes in the germ cells, rather than the intended transformation of gonocytes to type-A spermatogonia. [Ritzen et al. 2007].

There was a time, not long ago, when orchidopexy was delayed until puberty, as it was known that many cryptorchid testes descended then. However, in retrospect we think the testes that descended at puberty were probably acquired cryptorchids, and that these testes would be predicted to have suppressed fertility because of secondary heat-induced loss of stem cells [Nistal et al. 1984]. Since we now have a much better understanding of postnatal germ cell development, the optimal time for orchidopexy appears to be during the first year [Thorup et al. 2007]. This would allow normal germ cell maturation, as long as the abnormality was secondary to high temperature and was reversible. We are just beginning to obtain direct evidence supporting early surgery at 9 months [Kollin et al. 2006], and we propose elective orchidopexy for congenital cryptorchidism around 6 months of age [Thorup et al. 2007]. For boys who develop acquired cryptorchidism, we recommend orchidopexy when the testis can no longer remain in the scrotum without manual traction by the examiner [Hutson and Hasthorpe, 2005].

As we learn more about testicular biology and the process of descent, it is quite possible that new ways of treating cryptorchidism will appear. It may be that our increasing knowledge of gubernacular migration may permit some trophic treatment. Certainly it may already be possible to contemplate topical or injectable treatment for inguinal hernia, and we await the development of such putative therapies.

Conflicts of interest statement

None declared.