Changes in connective tissue in patients with pelvic organ prolapse—a review of the current literature

The vaginal wall is composed of four layers: a superficial layer of stratified squamous epithelium, a subepithelial dense connective tissue layer, composed primarily of collagen and elastin, a layer of smooth muscle referred to as the muscularis and an adventitia, which is composed of loose connective tissue. The vaginal subepithelium and muscularis together form a fibromuscular layer beneath the vaginal epithelium, providing longitudinal and central support.

The connective tissue underlying the vagina contains relatively few cells: Beside fat cells and mast cells, mainly fibroblasts are found, producing components of the extracellular matrix (ECM). The ECM contains fibrillar components (collagen and elastin) embedded in a non-fibrillar ground substance. This ground substance consists of non-collagenous glycoproteins, proteoglycans and hyaluronan. In addition, with the exception of the ATFP, these tissues contain a significant amount of smooth muscle cells [16]. The fibrillar component is thought to contribute the most to the biomechanical behaviour of these tissues. The quantity and quality of collagen and elastin are regulated through a precise equilibrium between synthesis, maturation and degradation. This process results in a dynamic process of constant remodelling.

In 1954, Ramachandran and Kartha [17] discovered the triple helical structure of collagen. In the endoplasmatic reticulum, α chains are formed, followed by posttranslational modifications of proline and lysine residues. Each collagen molecule is made of a precise combination of three α-polypeptide chains. Depending on the collagen type, the three α-polypeptide chains vary. The three helices are twisted together into a triple helix, stabilised by numerous hydrogen bonds. There is some covalent cross-linking within the triple helices and a variable amount of covalent cross-linking between collagen helices, resulting in tissue-residing collagen of different maturity.

Once the triple helix called procollagen is formed intracellularly, it is secreted into the extracellular space. Tropocollagen molecules are formed by the action of carboxy-terminal and amino-terminal peptidases. The tropocollagen molecules undergo self-assembly into collagen fibrils, which in turn associate to form fibres and fibre bundles (Fig. 1). The shape and behaviour of a tissue are determined in part by the correct positioning of collagen fibrils within a fibre and fibres within a matrix.

There are 28 types of collagen. The fibrillar collagens I, III and V, present in the vagina, and its supportive tissues are thought to be the principal determinants of soft tissue strength. Collagen I fibres are universally present and are flexible and offer great resistance to tension. Collagen III is predominant in tissues that require increased flexibility and distension and that are subject to periodic stress [18]. It is the primary collagen subtype in vagina and supportive tissues. Collagen III is the major collagen in skin at birth before it is replaced by collagen I later in life. Both type I and type III are found in granulation tissue during wound repair [19, 20]. Collagen V is of minor quantitative importance. It forms small fibres with very low tensile strength and has been found to be also important in wound healing and fibrillogenesis [21]. The role of collagen V in the vagina and supportive tissue has not been elucidated yet.

Collagen I copolymerises with collagen III and V to form fibrils with controlled diameters. These fibrils influence the biomechanical characteristics of a given tissue [21]. An increase in collagen III and V decreases the mechanical strength of connective tissue by decreasing fibre size [22]. It is generally agreed that a higher I/III ratio in the ligament is indicative of greater strength, whereas a lower ratio may result in tissue laxity.

Age is a risk factor for POP. In the POSST study, there was a 10% increased risk of POP for each decade of life [23]. Olsen et al. [3] found that the cumulative incidence of primary operation for POP and incontinence increased from 0.1% in the age group 20–29 up to 11.1% in the age group 70–79.

Reay Jones et al. [15] assessed the Uterosacral ligament Resilience (UsR) by tensiometry in patients with and without POP to determine whether it influenced uterine or pelvic floor mobility or whether it varied with age, vaginal delivery, menopause or histological variations in the ligament. He found that the UsR was significantly reduced (P = 0.02) in symptomatic POP. There was a significant decrease in UsR with menopause (P = 0.009) and older age (P = 0.005), suggesting that where pelvic floor muscles are weakened, a decrease in pelvic connective tissue resilience—related to the age and menopause—may facilitate progression to symptomatic POP.

Two mechanisms of maturation of collagen have been identified. The first involves the enzymatically controlled lysine aldehyde cross-links. The initial enzymatic controlled divalent cross-links dehydro-hydroxy lysinonorleucine and hydroxylysinoketo-norleucine are converted to stable trivalent cross-links, histidino-hydroxylysinonorleucine and hydroxylysyl-pyridinoline, as the tissue matures [24]. The relative proportion of the initial divalent cross-links to mature cross-links enables an assessment of the maturation of the tissue. The mechanism of creating strength of the collagen by interfibrillar cross-links is currently under investigation.

Secondly, the mature, slowly metabolising collagen is susceptible to the so-called non-enzymatic cross-linking, also known as glycation or Maillard reaction. It involves the fairly random addition of glucose to the collagen, as the turnover of collagen is generally rather low. The products of this glycation ultimately react further to form intermolecular cross-links. It has been well established that these advanced glycated endproducts (AGEs) of collagen accumulate with age (Fig. 2). This mechanism is the major cause of substantial dysfunction of the collagenous tissues and is responsible for the complications of connective tissues seen at older age. The “overmature” collagen is stiffer and therefore more fragile than the collagen that exhibit only enzymatic cross-links [24, 25]. The glycation of other proteins involves the same mechanisms. But the long biological half-life of collagen ensures that it plays an important role in ageing. The nature and relative importance of the cross-links formed in vivo, in comparison to those reported formed in vitro with model compounds, still needs to be established as the structure of only a few AGEs has been established.

With the knowledge of changes of collagen with age, more research of POP, especially in young women, will elucidate the underlying pathophysiology of the disease.

The balance between synthesis and degradation of collagen is important for maintaining tissue integrity and tensile strength during continuous tissue remodelling. Degradation depends upon the combined activity of matrix metalloproteinases (MMP) and their regulation of release, activation or sequestration of growth factors, growth factor binding proteins, cell surface receptors and cell–cell adhesion molecules [26]. MMPs are being synthesised intracellularly and secreted as pro-enzymes into the extracellular space, which requires conversion to the active form for enzymatic activity. Twenty-three different members of the MMP family have been identified in humans. They all can degrade one or more extracellular matrix components, however, with different specificities. The interstitial and neutrophil collagenases (MMP-1, MMP-8, MMP-13) are capable of cleaving fibrillar collagen, while the gelatinases (MMP-2 and MMP-9) degrade the resulting denatured peptides. Acid cathepsins depolymerise collagen fibres by cleaving near cross-link sites. The combined action of these enzymes is capable of degrading all components of the extracellular matrix.

To limit connective tissue degradation, the activity of MMPs is regulated by modulation of pro-enzyme production [27]. Overdegradation is also regulated by endogenous inhibitors: serum-borne inhibitors and the tissue-derived inhibitors of metalloproteinases (TIMPs). They bind stoichiometrically to MMPs to inhibit their activity. TIMP-1, as well as TIMP-3, binds to MMP-1 and MMP-9; TIMP-2 binds to MMP-2 [28, 29]. Recently, those that have been formed in vitro have shown that active MMPs can also be inactivated spontaneously by degradation into smaller fragments. This process is referred to as autocatalysis [30, 31].

The mechanical properties of tissues are also dependent on the proportion of elastin, an insoluble polymer that is formed by the assembly of tropo-elastin monomers followed by catalysis of cross-link formation by lysyl oxidases (LOX). Elastin allows the tissue to stretch and return to its original shape without energy input [32]. This property of resilience is presumed important for reproductive tissues. It accommodates the enormous expansion in pregnancy and involution after parturition. Production of elastin is unique among connective tissue proteins. In most organs, elastin biosynthesis is limited to a brief period of development. The assembly of elastic fibres is complete by maturity when tropo-elastin synthesis ceases. In undisturbed tissues, elastic fibres produced in the third trimester of foetal life last the rest of life [33]. In the female reproductive tract, however, elastic fibre turnover appears to be continuous. Recently, it is found that LOX is essential for elastic fibre homeostasis in many tissues, including the female pelvic organs. Mice lacking LOXL1 are unable to synthesise elastin polymers in adult tissue, whereas collagen synthesis appears to proceed normally. These mice also fail to replenish mature elastin fibres in the reproductive tract after parturition. They develop spontaneous prolapse [33, 34]. Fibulin-5, which is an elastic binding protein crucial for elastic fibre assembly [35], is believed to act as a bridge between cells and tropo-elastin for effective cross-linking and assembly of tropo-elastin into mature elastic fibres. Increased synthesis of tropo-elastin and fibulin-5 may be necessary to counteract for disruption of elastic fibres and to regenerate elastic fibres in the vaginal wall. In fibulin-5 knock-out mice, POP was similar to that in primates, suggesting that synthesis and assembly of elastic fibres are crucial for recovery of pelvic organ support after damage. Disordered elastic fibre homeostasis seems to be a primary event in the pathogenesis of POP in mice [36].

Studies on changes in the quantity and ratios of subtypes of collagen produced inconclusive data. Both increase and reduction of total collagen content of vaginal and pelvic floor supportive tissues in patients with POP have been reported (Table 1). Different methods of quantification of collagen content and also the lack of information on tissue histology and biopsy site of the vagina or supportive tissue that was analysed make direct comparison between studies difficult. Table 1 shows that tissue obtained from uterine ligaments of patients with POP seems to show a decreased total collagen content [11, 37‐39] with a higher concentration of collagen III [40‐43] irrespective of age or parity. An increased collagen III content may suggest tissue repair after overstretching of the supportive connective tissue of the pelvic floor. Moalli et al. [41] also found a significantly increased expression of active MMP-9 in women with POP relative to controls. An increase in collagen III in combination with an increase in active MMP-9 is typical of tissue that is remodelling after injury or a tissue that is remodeling to accommodate a progressively increasing mechanical load [44]. A higher expression of tenascin, an extracellular matrix glycoprotein that reappears around healing wounds [42], supports this theory.

There is more to collagen than just the total amount and the subtypes. Barbiero et al. [45] performed a qualitative analysis of type I collagen in the parametrium of patients with and without POP. It was demonstrated that the parametrium consists of shorter, thinner and more disorderly arranged collagen fibres in patients with uterine prolapse compared to healthy controls.

The maturity of the tissue depends on the relative proportion of the divalent cross-links to the mature (trivalent) cross-links. Furthermore, mature, slowly metabolising collagen is susceptible to non-enzymatic glycation and some of these AGEs, such as pentosidine, are additional cross-linking compounds that may further inhibit the turnover of collagen.

Jackson [13] showed that POP was associated with a significant rise in the immature cross-link. The mature pyridinoline cross-linking was not altered. Pentosidine was increased significantly in the prolapse tissue, and pentosidine concentrations increased with increasing age in both groups, demonstrating that non-enzymatic glycation of collagen occurs slowly over a long time. This greater pentosidine concentration in prolapse tissue makes the tissue less soluble than in controls, reflecting the maturity of the tissue [46].

Only two studies describe the cross-linking of collagen fibres in POP [11, 47]. However, both studies only analysed a specific subset of cross-links, limiting their conclusions to these subsets. Söderberg et al. [11] found a decrease in extractability by pepsin digestion at young age in both women with and without POP compared to older women. This is an indicator of cross-links in the collagen molecule. It is considered a physiological effect of normal ageing. Chen et al. [47] analysed pyridinoline, the major mature collagen cross-link in fascia, in anterior vaginal wall. There was no difference between the incontinent women with POP and the continent women without POP. This is in accordance with the findings of Jackson [13].

The majority of studies on MMPs focus on synthesis (pro-MMP or MMP messenger RNA (mRNA)). Although the active form is the most relevant form with respect to tissue degradation, the analysis of the entire expression profile of these enzymes, including pro-enzyme, active and autocatalytic forms, should provide a more conclusive insight [48, 49].

Jackson [13] suggested an increased metabolic turnover of collagen, since MMP-2 and MMP-9 were significantly higher in prolapse tissue than in normal tissue. He did not assess the expression of other MMPs or the TIMPs. As shown in Table 2, an increase in pro-MMP-2 and active MMP-2 was confirmed by two other studies [43, 50]. Summarising the data of Table 2 concerning the activity of MMP point to a condition in which the degradation of connective tissue is accelerated in the vagina and the supportive tissue of patients with POP by an increased expression of pro- and/or active metalloproteinases in combination with a decrease in TIMP-1 mRNA and active TIMP-1 expression [47].

The importance of elastic fibres in maintaining vaginal support is demonstrated by genetic connective tissue diseases like Marfan’s syndrome (mutations in fibrillin-gene) and cutis laxa (mutation in elastin and fibulin-5 genes). An increased incidence of POP is seen in women affected by these connective tissue diseases [51, 52].

Some changes in elastin expression in relation to POP are reported (Table 3). But the ways by which elastin is measured also vary: via mRNA level, precursor protein levels or mature elastin levels. Elastin mRNA is a few steps away from the actual elastin. Also, the protein precursor of elastin, tropo-elastin, can be measured without actually measuring the amount of mature elastin [48, 53]. A common way to quantify mature elastin used by Jackson is by indirectly measuring its cross-links with desmosine [54]. This measurement could be inaccurate, however, when studying a disease such as POP in which the cross-linking process may be disrupted and desmosine concentration is more reflective of diminished cross-linking than the total amount of elastin [48]. The most direct and thus appropriate way to measure elastin is by immunohistochemistry [53].

Even though the techniques used to obtain the data vary and thus provide inconclusive data, there is circumstantial evidence that a deficient synthesis and degradation of elastic fibres may be associated with POP [42, 53, 55‐57].

If the collagen content in POP is changed, this may be caused by differences in the number of fibroblasts (cellularity) in the connective tissue [58, 59]. But, the quality, i.e. functionality of the fibroblast, may also be important in the pathogenesis of POP. There is some indication that the contractibility of the vaginal (myo)fibroblasts is decreased in POP patients, which may result in deficient collagen [60]. In models of wound healing, skin myofibroblasts control the contractile and strength of the tissue, which is regulated by the so-called endothelin-1 (ET-1) system. Poncet et al. [60] compared cultures of α-smooth muscle actin-positive myofibroblasts that were established from POP patients to myofibroblasts from primiparous women with respect to their expression of the ET-1 system and contractile properties. They found that spontaneous contraction of myofibroblasts from estrogen-treated women with POP was significantly lower than from young primiparous women. Addition of exogenous endothelin-1 decreased the spontaneous contraction of myofibroblasts, which is opposite to observations in skin myofibroblasts.

Makinen et al. [61] studied the rates of collagen synthesis and procollagen mRNA levels in cultured fascia fibroblasts of patients with POP. They found that these fibroblasts exhibited rates of collagen synthesis similar to or slightly higher than those from age-matched controls. This finding suggests that POP is not related to defects in the capacity of vaginal fibroblasts to synthesise or process procollagen. The existence of a possible qualitative change in type I and or type III collagen could not be ruled out in this study. There are no indications of mutations in the various polypeptides in collagen type I and III. Mutations could produce minor changes in collagen fibres that would make connective tissue less able to withstand the stresses of an individual lifespan and could explain predisposition to common diseases affecting connective tissue.

With respect to the production of elastin, the elastin gene expression and protein synthesis in fibroblasts derived from cardinal ligaments of patients with POP is markedly lower than in non-POP patients [57]. This may be a result of a decreased expression of the wild-type p53 mRNA and wild-type p53 protein in fibroblasts of women with POP. Cells fail to enter quiescence (G0 phase) that may lead to a decrease in synthesis and deposition of elastin.

The synthesis of components of the ECM by fibroblasts is influenced by stretch. Whether it is cause or effect, prolapsed tissue is stretched tissue. Ewies demonstrated recently that mechanical stretch disturbs the fibroblasts’ ability to maintain the cytoskeleton architecture. The use of estrogens did not reverse the process or protect the cells from the effect of stretch, but significantly increased the rate of fibroblast proliferation, suggesting their role in the healing process [62].

These results suggests, in contrast to the findings of collagen production, that functional changes in the fibroblasts of the cardinal ligaments are involved in the mechanism of prolapse development [57, 63, 64]. In relation to POP in which tissue is stretched, it is demonstrated that mechanical stretch disturbs the fibroblasts’ ability to maintain their cytoskeleton architecture and we speculate that they may disrupt ligamentous integrity and result in clinical prolapse.

As estrogen deficiency is a known risk factor for POP [65], estrogen replacement therapy traditionally has been used to improve structural integrity of the pelvic tissue with favourable effects on urinary incontinence[66‐69]. Previously, estrogen receptors (ER) were identified in the nuclei of connective tissue and of the smooth muscle cells of the bladder trigone, urethra, vaginal mucosa, levator ani stromal and uterosacral ligament. Two different subtypes have been found in human cells: ER-α is the dominant receptor in the adult uterus, whereas ER-β is expressed at high levels in other estrogen-target tissues such as prostate, testis, ovary, smooth muscle, vascular endothelium and immune system. These receptors participate in maintaining the supportive system of the pelvic by a.o. increasing synthesis or by decreasing breakdown of collagen and other extracellular matrix proteins [70]. Few studies have been done to assess the tissue expression levels of sex steroid hormone receptors in patients with POP. Lower estrogen receptor expression in patients with POP have been found in combination with lower serum concentrations of estrogen [65, 71]. In contrast, higher sex steroid hormone receptor expression in POP patients has also been described [72].

Several studies report an increase in the mRNA expression for collagen I and III in estrogen replacement therapy [30, 73, 74]. These findings suggest that estrogen increases the turnover of connective tissues of the pelvic floor. It is also suggested that estrogen restores the collagen metabolism to a premenopausal state [41]. In a double-blind, placebo-controlled trial with postmenopausal women with urinary stress incontinence treated with estradiol therapy, Jackson [13] found strong evidence for both new synthesis and degradation. The immature cross-links were increased, indicating newly synthesised collagen. However, the ratio of type I/III collagen was unchanged in the estradiol treated group, and the total collagen content was significantly decreased. Also, an elevation of both the pro-active and active forms of MMP-2 and MMP-9 in women treated with estradiol compared to controls was found. This resulted in a decrease in total collagen content [75]. Also, a combination of upregulation of MMPs and suppression of TIMP by estrogen resulted in an increase of ECM breakdown [76].

Inhibition of MMP by estrogen therapy is also reported [77, 78]. Zong et al. [49] found that only E2 combined with progesterone decreased the active form of MMP-1, which suggests that both hormones are necessary to maintain the integrity of the female pelvic floor.

With respect to elastin, Moalli et al. [40] found no differences between pre-, post- and postmenopausal women on hormone therapy.

17β-Estradiol may have a suppressive effect on proliferation of fibroblasts, derived from cardinal ligaments in patients with POP. Therefore, 17β-estradiol may have a role in inducing POP by negatively affecting the concentration of fibroblasts in pelvic organ connective tissue [79].

Estrogen therapy, thus, induces turnover of collagen, but the precise role of estrogen in collagen metabolism related to POP is still unclear. Data suggest an increased mRNA expression of collagen type I and III as a reaction to hormone replacement therapy, with a concomitant increased synthesis of these collagen types. However, an increased activity of MMP, resulting in an increase in collagen degradation, is also reported. Whether sex steroid receptors are a primary cause or a downstream effect of POP remains unknown.