Introduction
Milk (including colostrum) is a complex biological fluid produced by mammals for the nutritional and immunological benefit of their offspring, whose chance of survival without it would be slim to none. In addition to the major milk components protein, fat, lactose, minerals and vitamins milk contains hundreds of minor components which are present in very low concentrations and often have still poorly understood bioactive properties (Stelwagen et al., Reference Stelwagen, Carpenter, Haigh, Hodgkinson and Wheeler2009; Foroutan et al., Reference Foroutan, Guo, Vazquez-Fresno, Lipfert, Zhang, Zheng, Badran, Budinski, Mandal, Ametaj and Wishart2019). Although the relative concentrations of the major milk components may vary between species (Jenness, Reference Jenness1986) they and/or their building blocks, such as ammino acids and fatty acids, have to come from the blood supplied to the mammary gland. Once they leave the blood into the interstitial fluid, they can follow an intracellular route involving active transport mechanisms in the mammary secretory epithelial cell or follow a non-selective passive diffusion route between adjacent mammary secretory cells. This paracellular route is controlled by tight junctions (TJ) and constitutes the mammary blood-milk barrier (BMB; Wellnitz and Bruckmaier, Reference Wellnitz and Bruckmaier2021).
The TJ, or zonula occludens, is part of the junctional complex in epithelial cells, including the mammary secretory cell. The various components of the junctional complex are shown in Fig. 1. Collectively, they play a crucial role in the cell's functioning and maintenance of its three-dimensional structure. A detailed discussion of the non-TJ components of the junctional complex is beyond the scope of the current review and can be found elsewhere (Fu et al., Reference Fu, Jiang, Li, Zhu and Zhang2022; Canse et al., Reference Canse, Yildirim and Yaba2023). Briefly, adherens junctions, desmosomes and hemidesmosomes are anchoring junctions, helping to anchor the cell to adjacent cells (adherens junctions and desmosomes) and to their extra cellular matrix (hemidesmosomes) and are linked to the cell's cytoskeletal micro- or actin filaments and intermediate filaments, respectively. Gap junctions, consisting of bundles of connexin proteins, are cell-to-cell communication junctions, allowing for chemical signals to diffuse directly between the cytoplasm of adjacent cells.
A schematic overview of the junctional complex of the mammary secretory cell (detailed description in the main text; modified from: Brennan et al., Reference Brennan, Offiah, McSherry and Hopkins2010).
The remaining component of the junctional complex, the TJ, is the only occluding junction, meaning it can restrict the two-directional free flow of small molecules and solutes between the blood and milk, providing a physical barrier between the “blood side” (i.e., extracellular fluid side) and the “milk side” of the mammary cell. In addition to this so-called barrier function, TJ also help to separate the plasma membrane of the cell into distinct apical and basolateral domains (i.e., fence function; Schneeberger and Lynch, Reference Schneeberger and Lynch1992). The TJ is located on the apical side of the cell and consist of proteins with multiple transmembrane loops (Fig. 1). Intracellularly these transmembrane proteins are linked to the micro- or actin filaments of the cytoskeleton. In the mammary cell the main transmembrane proteins, occludin and claudins, are the true TJ proteins. Despite occludin, a single 65 kDa protein, being a transmembrane TJ protein, its role in the barrier function of the TJ has recently been questioned (Zhou et al., Reference Zhou, Lu, Xu, Wang, Zhang and Lu2020). Instead, it appears to play a role in facilitating the cell's milk protein secretion and may help to prevent apoptosis (Beeman et al., Reference Beeman, Baumgartner, Webb, Schack and Neville2009; Zhou et al., Reference Zhou, Lu, Xu, Wang, Zhang and Lu2020). Claudins, on the other hand, constitute a large family of at least 25 proteins of approximately 25 kDa and are instrumental in the barrier function of mammary TJ (Baumgartner et al., Reference Baumgartner, Rudolph, Ramanathan, Burns, Webb, Bitler, Stein, Kobayashi and Neville2017). However, not all claudins are expressed in the mammary epithelium and for those that are, expression depends on the physiological state of the gland (Kobayashi and Kumura, Reference Kobayashi and Kumura2011; Baumgartner et al., Reference Baumgartner, Rudolph, Ramanathan, Burns, Webb, Bitler, Stein, Kobayashi and Neville2017).
Finally, there are three zonula occludens proteins (ZO-1, ZO-2 and ZO-3), also called scaffolding proteins, that link the true TJ proteins to the cytoskeleton and all three have been shown to be expressed in normal and malignant mammary tissue (Martin et al., Reference Martin, Watkins, Mansel and Jiang2004).
In the normal non-malignant mammary gland TJ are open or “leaky” when the tissue is non-lactating but close, becoming “tight”, when the gland switches to a lactating state. This is also the period of colostrogenesis and coincides with major changes in the level of several hormones important for mammary functioning (Bigler et al., Reference Bigler, Gross, Baumrucker and Bruckmaier2023). Therefore, the objective of this review is to discuss the role of the TJ, constituting the BMB, in milk secretion and in particular endocrine factors affecting the permeability of mammary TJ, primarily in cows and goats. However, where relevant or where no research on cows and goats is available, information from other species is referred to.
Measuring blood-milk barrier permeability
In order to examine the importance of the BMB it is important to be able to measure the permeability of TJ (i.e., open or closed). The most common method in vitro is to grow mammary cells (primary or cell line) to confluence in a small insert chamber with a permeable membrane, placed in the well of a multi-well plate, allowing for culture media to be added separately to the basal side (i.e., inside the well) and to the apical side (i.e., inside the insert). Once the cells have grown to confluence and TJ are established the effects of treatments added to the apical and/or basolateral side on TJ state can be examined by placing an electrode on each the basolateral and apical side and measure the transepithelial electrical resistance (TEER) across the cell monolayer or, alternatively, the paracellular movement of small, labelled marker molecules (Kobayashi, Reference Kobayashi2023). A reduction in TEER and/or a movement of marker molecules across the cell monolayer are indicative of increased TJ permeability.
Measuring TJ state in live animals, especially in large animals such as goats and cows, is slightly more complicated due to their size and natural behaviour and movement. However, it is feasible to measure a small (15–30 mV) potential difference across the mammary epithelium. Similar to the TEER measurement in vitro, a small electrode is inserted via the teat orifice inside the intramammary milk pool (i.e., representing the apical side of the mammary epithelium) and the other electrode is inserted in a blood vessel (e.g., the mammary vein; representing the basolateral side) (Peaker, Reference Peaker1977; Stelwagen et al., Reference Stelwagen, Davis, Farr, Prosser and Sherlock1994). It is also possible to study the movement of markers across the BMB by administering small radio-labelled-molecules via the teat into the milk and measuring their presence in the blood, or vice versa (Linzell and Peaker, Reference Linzell and Peaker1974) or injecting a dye (e.g., Evans blue) via the teat orifice into the milk and measuring its appearance in a mammary efferent lymph vessel (Stelwagen et al., Reference Stelwagen, Farr, Davis and Prosser1995). Although the latter requires surgical catheterization of a lymph vessel.
Milk synthesis and secretion is unique to mammary tissue and therefore milk components should not be present in blood in animals with healthy, non-involuting mammary glands. Hence, the measurement of small milk components, such as lactose and α-lactalbumin, in blood have been validated to be reliable indicators of mammary TJ permeability during established lactation (Stelwagen et al., Reference Stelwagen, Farr, McFadden, Prosser and Davis1997). Single or repeated blood samples (via an indwelling jugular vein catheter) can be easily obtained in large animals and allow for the assessment of changes in TJ permeability.
Finally, changes in the milk concentrations of sodium (Na), potassium (K) and chloride (Cl) can also be used as indirect indicators of TJ permeability (Neville and Peaker, Reference Neville and Peaker1981; Stelwagen et al., Reference Stelwagen, Farr, Davis and Prosser1995). An increase in the milk concentration of Na and Cl and a concomitant decrease in that of K are indicative of leaky TJ and vice versa. Although, milk samples can be easily and non-invasively obtained, it is difficult to conduct repeated sampling without each time initiating a milk let-down response, which in itself can cause transient opening of TJ (Herve et al., Reference Herve, Quesnel, Lollivier, Portanguen, Bruckmaier and Boutinaud2017).
The role of the blood-milk barrier in milk secretion
When the mammary gland switches from a non-lactating state to one of copious milk production around parturition and again from a lactating to an involuting state at cessation of milk removal, the gland undergoes structural changes that are consistent with reducing (Linzell and Peaker, Reference Linzell and Peaker1974) and increasing (Peaker, Reference Peaker1980), respectively, the permeability of the BMB. This suggests a role for mammary TJ in establishing and maintaining milk synthesis. Moreover, in goats and cows during established lactation, it was shown that after approximately 18 to 21 hours of milk accumulation following the last milk removal (i.e., milk stasis), the BMB becomes compromised and TJ start to become leaky, corresponding with a decrease in the rate of milk secretion (Stelwagen et al., Reference Stelwagen, Davis, Farr, Prosser and Sherlock1994). More direct evidence for a role of TJ in maintaining milk production comes from experiments where TJ were deliberately disrupted in lactating glands. Maintaining the concentration of extracellular calcium (Ca) is a requirement to maintain the integrity of mammary TJ (Pitelka et al., Reference Pitelka, Taggart and Hamamoto1983), hence administering a Ca chelator via the teat allows for experimentally disrupting TJ. A single iso-osmotic dose of the Ca chelators citrate (Neville and Peaker, Reference Neville and Peaker1981) or EGTA (Stelwagen et al., Reference Stelwagen, Farr, Davis and Prosser1995) administered unilaterally via the teat orifice into a lactating gland of goats, not only caused rapid transient disruption of the TJ, it also resulted in an immediate concomitant reduction in milk secretion in the treated gland only. Once the effect of the Ca chelator wore off and TJ started to close again, milk secretion began to return to normal.
The mechanism by which TJ affect milk production is not fully understood. However, it is likely that disrupting TJ results in a perturbation of the cytoskeleton of the mammary epithelial cell, as they are closely interlinked (Fig. 1) and that this in turn affects cell functioning. Disruption of the mammary cell's cytoskeleton has been shown to decrease milk synthesis and secretory activity, including milk protein synthesis and secretion, in the lactating mammary gland (Patton, Reference Patton1978; Nickerson et al., Reference Nickerson, Smith and Keenen1980; Akers and Nickerson, Reference Akers and Nickerson1983). Presumably, due to disruption of the cytoskeleton-facilitated movement of secretory vesicles toward the apical membrane of the cell. Consistent with this notion is the fact that the TJ protein occludin has been shown to play a role in intracellular milk protein synthesis and secretion (Zhou et al., Reference Zhou, Lu, Xu, Wang, Zhang and Lu2020) and also that in transgenic mice, constitutively expressing the Rho effector protein PKN1, TJ closure was prevented (Fischer et al., Reference Fischer, Stuckas, Gluth, Russell, Rudolph, Beeman, Bachmann, Umemura, Ohashi, Neville and Theuring2007). Rho proteins belong to a family of GTPases that can act as molecular switches and, as such, are known to control many cellular functions, including cytoskeletal reorganization and cell polarity (Ellenbroek and Collard, Reference Ellenbroek and Collard2007).
Further, being polarised is a requirement for mammary epithelial cells to lactate (Berga, Reference Berga1984) and disruption of TJ results in a loss of potential difference across the mammary epithelium (Stelwagen et al., Reference Stelwagen, Davis, Farr, Prosser and Sherlock1994). Finally, increased TJ permeability not only increases the Na to K ratio in the milk, it also does so in the cytoplasm of the cell and this, in turn, can reduce milk production (Stelwagen et al., Reference Stelwagen, Farr and McFadden1999a).
Collectively these studies indicate an important role for the BMB, and consequently TJ, in establishing and maintaining lactation in the healthy, non-malignant, mammary gland. It must be noted that the BMB, and TJ, may be affected by mastitis, an inflammation of the mammary gland, often caused by a pathogenic infection. However, in the diseased gland paracellular leakage of milk and/or blood components across the BMB is not only due to TJ being compromised but is also likely due to alveolar cell damage caused by the infection. Moreover, mastitis induces a massive influx of neutrophils in the gland. The discussion of the role of the BMB and TJ in the diseased mammary gland is beyond the scope of the current paper and has been discussed previously (Stelwagen and Singh, Reference Stelwagen and Singh2014).
Hormonal control of the blood-milk barrier
As discussed above, the BMB is established when the gland switches from a non-lactating state to a lactating one around parturition. This coincides with major changes in systemic hormone profiles occurring during this time (Fig. 2; Convey, Reference Convey1974; Bigler et al., Reference Bigler, Gross, Baumrucker and Bruckmaier2023). In particular the sudden drop in progesterone appears to trigger the closure of mammary TJ at this time (Nguyen et al., Reference Nguyen, Parlow and Neville2001). However, progesterone is not the only hormone affecting mammary TJ status and in the following sections the various known endocrine factors involved in establishing and maintaining mammary TJ, and therefore the BMB, will be discussed. It must be emphasized that all of the hormones discussed are involved in multiple biological processes throughout the body but only those activities relating to mammary TJ will be discussed here.
Stylised changes in blood hormone profiles around parturition in the cow (based on data from Convey, Reference Convey1974).
Progesterone and estrogen
The steroid hormones progesterone and estrogen play an important role in maintaining pregnancy, requiring an elevated systemic level of progesterone and a relatively low level of estrogen. However, towards the end of pregnancy the level of estrogen starts to increase approximately 30 days before parturition, whereas that of progesterone shows a marked and sudden drop approximately two days prior to parturition (Fig. 2, Convey, Reference Convey1974). The timing of this coincides with the closure of mammary TJ and, indeed, it is speculated that in particular the sudden drop in the concentration of progesterone may trigger the closure of TJ and trigger subsequent milk production. Using an ovariectomised mouse model in combination with administering combinations of progesterone or a progesterone antagonist, Nguyen et al. (Reference Nguyen, Parlow and Neville2001) showed clearly that the presence of progesterone prevents closure of mammary TJ.
Agenäs et al. (Reference Agenäs, Lundström and Holtenius2019) looked at the effect of estrogen (17β-estradiol) on mammary TJ permeability in non-pregnant late-lactation dairy cows. Daily administration of 17β-estradiol during six days decreased milk yield immediately, but did not affect TJ permeability, based on concentrations of lactose in the urine and, until after the third day of 17β-estradiol injections. This suggests that 17β-estradiol has no direct effect on mammary TJ or may initially have had a “protecting” effect on TJ and the delayed increase in plasma lactose concentrations is the results of mammary involution rather than leaky TJ.
Recently Zhu et al. (Reference Zhu, Jia, Zhang, Liu, Yang, Han, Chen and Ding2022) investigated the effect of different concentrations of progesterone and 17β-estradiol administered together on TJ protein expression in cultured primary epithelial cells isolated from goat mammary tissue. Expression of the TJ proteins occludin, claudin-1 and claudin-3 were not really affected by 17β-estradiol in the presence of progesterone, whereas its effect on the scaffolding protein ZO-1 was ambiguous. In contrast, when the level of progesterone, in the presence of 17β-estradiol, was reduced by approximately 85% it reduced the protein expression of claudin-1 and claudin-3. It did not affect that of occludin and the effect on ZO-1 was again ambiguous. Given the importance of these steroids in mammary function (Tucker, Reference Tucker2000), it is surprising there has not been more research on their effect on TJ in the nonmalignant mammary gland.
Phytoestrogens
These are nonsteroidal compounds derived from plants and, as such are not synthesized in the body and are solely derived from the diet. However, due to their chemical structures, once absorbed into the body, they can bind to estrogen receptors and elicit either an agonistic or antagonistic estrogen response in the body (Glazier and Bowman, Reference Glazier and Bowman2001). Tsugami et al. (Reference Tsugami, Matsunaga, Suzuki, Nishimura and Kobayashi2017) demonstrated in cultured mouse mammary epithelial cells that the phytoestrogens coumestrol, genistein and daidzein all increased TJ permeability and reduced the expression of TJ proteins claudin-3 and 4 and occludin. However, more recently they showed that a phytoestrogen metabolite (equol) decreased TJ permeability and increased the expression of claudin-3 in cultured bovine mammary epithelial cells (Tsugami et al., Reference Tsugami, Wakasa, Kawahara, Nishimura and Kobayashi2022).
Glucocorticoids
Glucocorticoids, including cortisol, are steroid hormones produced by the adrenal glands and promote differentiation of mammary epithelial tissue (Tucker, Reference Tucker2000). Their systemic levels increase and peak shortly before parturition (Fig. 2; Convey, Reference Convey1974), when mammary TJ are closing, suggesting a role in the formation and/or regulation of the BMB (i.e., TJ).
Zettl et al. (Reference Zettl, Sjaastad, Riskin, Parry, Machen and Firestone1992), using a mouse mammary cell line, demonstrated for the first time that glucocorticoids could directly induce the formation and closure of mammary TJ. To explore a possible effect of glucocorticoids in vivo, Thompson (Reference Thompson1996) injected hydrocortisone (i.e., cortisol) via the teat orifice into one mammary gland of late-pregnant goats, using the contralateral gland as a control. A transient decrease in the concentration of Na in milk and a corresponding increase of that of K, suggested a tightening effect of hydrocortisone on the BMB. Similarly, Stark et al. (Reference Stark, Wellnitz, Dechow, Bruckmaier and Baumrucker2015) showed that, whilst a combination of progesterone and estrogen can be used to induce lactation in non-lactating, non-pregnant animals, additional systemically administered dexamethasone, a synthetic glucocorticoid, was necessary for the formation of TJ and subsequent colostrogenesis. During established lactation in cows mammary TJ become leaky after 18 hours if milk is allowed to accumulate in the gland (Stelwagen et al., Reference Stelwagen, Davis, Farr, Prosser and Sherlock1994, Reference Stelwagen, Farr, McFadden, Prosser and Davis1997), as discussed earlier. Using this model, Stelwagen et al. (Reference Stelwagen, Van Espen, Verkerk, McFadden and Farr1998), showed that when cows were injected with adrenocorticotrophin, to significantly increase their natural cortisol levels in the blood, the opening of TJ around 18 hours of milk accumulation could be prevented. In contrast, Nguyen et al. (Reference Nguyen, Parlow and Neville2001), injecting dexamethasone into late-pregnant mice, found that there was no effect on mammary TJ closure. Taken together these studies indicate that glucocorticoids can stimulate the closure of mammary TJ in vivo in goats and cows, but not necessarily in mice, suggesting potential species differences.
Despite the lack of an in vivo effect of glucocorticoids on mammary TJ in mice, TJ in mouse-derived mammary cell lines do respond to glucocorticoids. Addition of dexamethasone to non-transformed 31EG4 mouse mammary cells increased the TEER and the expression of the TJ scaffolding protein ZO-1 (Zettl et al., Reference Zettl, Sjaastad, Riskin, Parry, Machen and Firestone1992; Singer et al., Reference Singer, Stevenson, Woo and Firestone1994; Woo et al., Reference Woo, Vha, Singer and Firestone1996). Similarly, dexamethasone significantly increased the TEER in HC11 and COMMA-D1 mouse mammary cell lines (Stelwagen et al., Reference Stelwagen, McFadden and Demmer1999b). Further, it increased the expression of the scaffolding protein ZO-1, as in the previous studies, and also that of the TJ transmembrane protein occludin. Since then, Kobayashi et al. (Reference Kobayashi, Tsugami, Matsunaga, Oyama, Kuki and Kumura2016), using cultured primary epithelial cells isolated from mouse mammary glands, confirmed the tightening effect of glucocorticoids on TJ. Moreover, dexamethasone increased the expression of the TJ proteins claudin-3 and claudin-4 in this study, indicating that it has a direct effect on the barrier function of TJ.
Taken together, both in vitro and in vivo data indicate that glucocorticoids play an important role in establishing the BMB and are doing so by upregulating members of the claudin family of TJ proteins, as well as that of the scaffolding proteins that link TJ to the cytoskeleton.
Prolactin
The protein hormone prolactin is well-known for its lactogenic properties (Tucker, Reference Tucker2000) and its systemic level increases and peaks immediately prior to parturition, when the gland switches from a non-lactating state to one of copious milk secretion (Fig. 2; Convey, Reference Convey1974). However, surprisingly little research to study its effect on mammary TJ has been conducted in vivo.
Prolactin administered to lactating rabbits decreased the Na and Cl concentration in milk and increased that of K (Linzell et al., Reference Linzell, Peaker and Tayler1975). It also decreased the paracellular movement of labelled 14C-sucrose molecules from the blood into milk. Flint and Gardner (Reference Flint and Gardner1994) injected bromocriptine, a known inhibitor of prolactin release, into lactating rats and showed a significant decrease in the concentration of milk K. These studies suggest that prolactin may reduce the permeability of the BMB by promoting the closure or tightening of mammary TJ. However, Nguyen et al. (Reference Nguyen, Parlow and Neville2001), inhibiting prolactin release in mice, found that it had no effect of on the permeability of TJ. This agrees with results from in vitro studies by Zettl et al. (Reference Zettl, Sjaastad, Riskin, Parry, Machen and Firestone1992). Using 31EG4 mouse mammary cells, they showed that addition of prolactin to the culture medium had no effect on the permeability of TJ. However, using the mouse mammary cell lines HC11 and COMMA-1D, the addition of prolactin increased TEER and decreased the paracellular movement of labelled inulin (Stelwagen et al., Reference Stelwagen, McFadden and Demmer1999b). In addition, in this study prolactin was shown to increase the expression of the TJ protein occludin, indicating a direct effect of prolactin on TJ closure. Although, the effect of prolactin alone on TJ permeability and on the expression of occludin was less than that of dexamethasone by itself and the largest effect was observed when the cells were exposed to a combination of prolactin and glucocorticoids. Moreover, differences in responsiveness were observed between the two cell lines used. In contrast, investigating the effect of prolactin on TJ in cultured primary mouse-derived mammary cells, Kobayashi et al. (Reference Kobayashi, Tsugami, Matsunaga, Oyama, Kuki and Kumura2016) found that prolactin increased TJ permeability, corresponding with a downregulation of the TJ barrier proteins claudin-3 and claudin-4. However, when the cells were treated with a combination of prolactin and glucocorticoids, TJ leakiness was decreased and the expression of claudin-3 was increased, thus indicating tighter TJ.
The limited research on a direct effect of prolactin on mammary TJ indicates that its effect varies between studies and may perhaps be cell line-specific. Alternatively, it may depend on the phosphorylation state of prolactin, as phosphorylated and unphosphorylated prolactin impact mammary cell functionality in distinctly different ways (Huang et al., Reference Huang, Ueda, Chen, Chen and Walker2008). Only phosphorylated prolactin stimulated TJ closure and the expression of the scaffolding protein ZO-1 in cultured HC11 cells (Ueda et al., Reference Ueda, Huang, Nguyen, Ferreira, Andre and Walker2011). Moreover, prolactin in combination with glucocorticoids, as would be the case in vivo, appears to have a consistent tightening effect on mammary TJ, mediated via increased expression of at least the claudin-3 TJ protein.
Parathyroid hormone-related peptide and serotonin
Maintaining extracellular Ca levels is a critical requirement to sustain the integrity of mammary TJ, both in vitro (Pitelka et al., Reference Pitelka, Taggart and Hamamoto1983) and in vivo (Neville and Peaker, Reference Neville and Peaker1981; Stelwagen et al., Reference Stelwagen, Farr, Davis and Prosser1995). Parathyroid hormone-related peptide (PTHrP) is an important hormone involved in regulating extracellular Ca homeostasis and transport in tissues throughout the body, including the mammary gland (Philbrick et al., Reference Philbrick, Wysolmerski, Galbraith, Holt, Orloff, Yang, Vasavada, Weir, Broadus and Stewart1996). Moreover, it is expressed in the mammary gland in response to prolactin and levels are highest in milk during established lactation (Thiede, Reference Thiede1994). Further, Thompson et al. (Reference Thompson, Ratcliffe, Hughes, Abbas and Care1994) showed that in goats the concentration of PTHrP was highest in the milk from glands that were unilaterally milked thrice daily and lowest in the once-daily milked contralateral glands. Given this information and the fact that mammary TJ become leaky when the frequency of milk removal is reduced to once-daily, Stelwagen (Reference Stelwagen2001) postulated a possible role for PTHrP in establishing and/or maintaining the BMB. However, little research has been conducted on PTHrP and its potential effect on TJ. Adding PTHrP to a mouse mammary cell line (HC11) increased the TEER, making TJ tighter, and also increased the expression of the TJ protein occludin, but these effects were only observed when extracellular Ca concentrations were low (Stelwagen and Callaghan, Reference Stelwagen and Callaghan2003). It is, therefore, tempting to speculate that PTHrP plays an important role during the onset of lactation when the BMB forms and extracellular Ca levels may be low due to the massive sudden drain of Ca from the gland via the milk. It also suggests a role for PTHrP in maintaining established TJ through maintaining extracellular Ca concentrations. Such an effect of PTHrP may be indirect via the stimulation of Ca-channel activity, as it was shown to stimulate Ca-channel activator activity but had no effect on Ca-channel blocker activity (Stelwagen and Callaghan, Reference Stelwagen and Callaghan2003).
Closely associated with PTHrP is serotonin. Known as a neurotransmitter, serotonin or 5-hydroxytryptamine (5-HT) can also act as a hormone, regulating the synthesis of PTHrP in the mammary gland (Hernandez, Reference Hernandez2017). As such, it plays also key role in regulating Ca homeostasis in the mammary gland and it has been shown to directly regulate mammary TJ in vitro. Addition of 5-HT or its receptor antagonists to human mammary epithelial cells (MCF10A) decreased and increased TEER, respectively, but only when administered to the basal side of the cells (Stull et al., Reference Stull, Pai, Vomachka, Marshall, Jacob and Horseman2007). This indicates that 5-HT has an “opening” effect on mammary TJ and that it acts via receptors on the basolateral side of the cell. Moreover, the complete loss of expression of the TJ scaffolding proteins ZO-1 and ZO-2 indicates that it can mediate cellular function through interfering with the link between TJ and the cell's cytoskeleton. Consistent with these in vitro results, Kessler et al. (Reference Kessler, Wall, Hernandez, Gross and Bruckmaier2019) showed that in periparturient cows selected for high circulating concentrations of 5-HT the BMB was more permeable than that in cows selected for low circulating levels of 5-HT.
As discussed earlier, at the end of lactation, when milk removal ceases, the permeability of the BMB and that of TJ increases, prior to the onset of mammary involution (Peaker, Reference Peaker1980). Given the “opening” effect of 5-HT on TJ as demonstrated by Stull et al. (Reference Stull, Pai, Vomachka, Marshall, Jacob and Horseman2007), Field et al. (Reference Field, Davidson, Hoerl, Dado-Senn, Hernandez and Laporta2023) recently investigated if increasing 5-HT levels in late-lactation cows, by administering a 5-HT precursor, could accelerate the drying-off process of the mammary gland in cows. Indeed, the various indicators of BMB state showed increased permeability in the treated cows only, indicating a loss of TJ integrity. This also allowed for the entry of innate immune factors such as lactoferrin into the intra-alveolar space. Presumably to help protect the gland against the (re)occurrence of intra-mammary infections, commonly associated with drying-off (Vilar and Rajala-Schultz, Reference Vilar and Rajala-Schultz2020). Furthermore, analyses of biopsied mammary tissue revealed down-regulated expression of all the key mammary TJ genes (occludin, claudin-1, claudin-2, and claudin-3) and all the genes encoding the scaffolding proteins (ZO-1, ZO-2 and ZO-3) (Field et al., Reference Field, Davidson, Hoerl, Dado-Senn, Hernandez and Laporta2023).
Collectively, these data show that both PTHrP and 5-HT play a role in regulating TJ permeability, albeit apparently opposing roles. Whereas limited research shows that PTHrP may have a protecting effect on TJ, helping to maintain tightness during the onset of lactation and/or established lactation, 5-HT clearly functions to open TJ and may play a role in speeding up the mammary involution process and in helping to protect the gland from mastitis at drying-off, by allowing innate immune factors to enter the mammary tissue at his critical time.
Other endocrine factors
There are a number of other hormones and growth factors that may affect TJ state. However, research, either in vitro or in vivo, is still very limited at this stage.
Oxytocin
This is a small peptide hormone that facilitates the contraction of smooth muscle cells and in the mammary gland it induces the milk let-down response by inducing transient contraction of the myoepithelial cells (Sagi et al., Reference Sagi, Gorewit and Wilson1980). Injection or infusion (i.v.) of oxytocin into lactating goats increased the movement of labelled sucrose (14C) from blood into milk and the concentrations of Na and Cl in milk also increased, whereas that of K decreased in milk (Linzell and Peaker, Reference Linzell and Peaker1971). Similarly, Allen (Reference Allen1990) showed that administration of exogenous oxytocin in lactating cows increased the concentration of Na in milk and decreased that of K and more recently Herve et al. (Reference Herve, Lollivier, Quesnel and Boutinaud2018) showed that exogenous oxytocin increased plasma lactose concentrations, but not when it was administered in combination with an oxytocin receptor antagonist. Although these results suggest that oxytocin can increase directly the permeability of the BMB, it must be emphasised that very high, supraphysiological doses of oxytocin were administered in these studies. Rather than having a regulatory activity on TJ, it is more likely that the high doses of oxytocin, resulting in a rapid contraction of myoepithelial cells and, as a result, that of the mammary alveoli. This likely disrupts TJ between the mammary alveolar cells through mechanical shear forces (Herve et al., Reference Herve, Lollivier, Quesnel and Boutinaud2018).
Somatotropin
Somatropin or growth hormone is another lactogenic hormone important for mammary function (Tucker, Reference Tucker2000). Its concentration in blood increases around parturition when the gland switches to a state of copious milk production (Convey, Reference Convey1974) suggesting a possible role in establishing the BMB. However, Flint and Gardner (Reference Flint and Gardner1994) suppressed circulating somatotropin level in lactating rats and showed no effect on the concentrations of Na and K in milk. Consistent with these findings Stelwagen et al. (Reference Stelwagen, Davis, Farr, Prosser and Sherlock1994) found no tightening effect on TJ in cows treated with exogenous somatotropin and milked once-daily to challenge mammary TJ. Therefore, somatotropin does not appear to affect TJ permeability.
Insulin and insulin-like growth factor 1
Both of these endocrine peptides play a role in mammary function, albeit predominantly in the developing gland (Tucker, Reference Tucker2000). They have been shown to increase the permeability of TJ when added to the basolateral side of colonic epithelial cells in vitro (McRoberts et al., Reference McRoberts, Aranda, Riley and Kang1990; McRoberts and Riley, Reference McRoberts and Riley1992). However, Singer et al. (Reference Singer, Stevenson, Woo and Firestone1994) observed no effect of added insulin on mammary TJ in cultured 31EG4 epithelial cells. Similarly, Flint and Gardner (Reference Flint and Gardner1994), found no effect on TJ permeability of suppressing circulating insulin-like growth factor 1 (as a result of actively suppressing circulating somatotropin) in lactating rats. Therefore, the effect of these two endocrine peptides on TJ permeability appears to be cell type-specific, with no effect on mammary cells.
Transforming growth factor-β
This is another endocrine peptide that plays a role during mammary gland development (Tucker, Reference Tucker2000), when TJ are known to be leaky. Woo et al. (Reference Woo, Vha, Singer and Firestone1996) showed that in cultured 31EG4 mammary cells transforming growth factor-β did not have a direct effect on TEER, but that in the presence of glucocorticoids it reversibly inhibits the closing effect of glucocorticoids on TJ. Moreover, unlike glucocorticoids, it had no effect on the expression scaffolding protein ZO-1.
Conclusions
The TJ between adjacent mammary epithelial secretory cells make up the BMB. As such, mammary TJ play an important role in the initiation and maintenance of milk production and they are under hormonal control as elucidated through both in vitro and in vivo studies. Progesterone appears to prevent the closure of TJ and the sudden drop in its concentration approximately two days before parturition is a prerequisite for TJ closure at the start of lactation. Elevated glucocorticoid levels promote TJ closure and, although prolactin may have a small effect itself on TJ, depending on its phosphorylation status, its main role appears to be to work synergistically with glucocorticoids to close TJ, through upregulating the expression of claudin TJ proteins (in particular that of claudin-3) and the scaffolding protein ZO-1, that links TJ to the cell's cytoskeleton. Maintaining extra cellular Ca concentration is important to maintain TJ closure. Two hormones involved in regulating Ca homeostasis, PTHrP and 5-HT are also important in regulating TJ. Whereas PTHrP reduces TJ permeability, necessary for establishing and maintaining milk production, 5-HT has an opening effect on TJ. The latter effect may help to speed up mammary involution and help facilitate the movement of immune factors into the mammary gland, to help prevent intramammary infections. Estrogen appears to have little effect on mammary TJ. Similarly, somatoropin and oxytocin, important lactogenic and milk let-down hormones, respectively, appear to have no direct effect on mammary TJ and, therefore the BMB.
Competing interest
The author declares no conflict of interest.
Open access
