I. INTRODUCTION

Top Previous Next References

The last Physiological Reviews article on the mechanism of milk secretion was written in 1971 (118) at a time when morphological information from electron microscopy was being combined with physiological and biochemical knowledge to provide the basis of our current understanding of milk secretion at the cellular level. Many of the hypotheses on the secretory mechanisms for the various milk constituents have survived more extensive work and amplification at the cellular and molecular levels, and it is not our intention to cover this older ground other than in outline. Rather, our intention is to extend the explanatory reductionist approach to examine the transport and secretory mechanisms for the major milk components or their precursors and their integration within the economy of the secretory cell.

The mammary gland is, in a number of respects, an unusual exocrine gland. Secretion, once initiated at about the time of parturition, is continuous but relatively slow in terms of fluid volume output per unit time. The secretion, milk in full lactation, is a complex mixture arising from different secretory pathways across, within, and sometimes between the mammary epithelial cells, comprising membrane-bound fat droplets, casein micelles, and an aqueous phase that usually contains lactose and sometimes complex carbohydrates, minerals in various ionic or bound forms, other proteins, and a bewildering array of soluble components, some of which may be biologically active in the young consuming the milk (78, 165). Another but related unusual feature is that secretion is stored within the lumen of the gland, either in the secretory alveoli or duct system, until it is removed at intervals (varying from several hours in many mammals, to once daily in rabbits, to once every 2 days in tree shrews, and to once a week in some pinnipeds) by the sucking young (46).

Studies in a number of species have shown that while there may be species-specific changes in milk composition with stage of lactation corresponding to growth phases of the young (a phenomenon particularly marked in marsupials), milk composition is markedly resistant to environmental, including nutritional, changes, whereas the quantity of milk secreted shows considerable plasticity (162).

The control of the rate of milk secretion will not be covered in detail, except for the effects of prolactin and cell stretch, since it has been reviewed extensively in recent years, particularly the autocrine control of milk secretion that acts to match supply by the mother to demand by the young within each individual mammary gland within the hormonal milieu of lactation that maintains the cells in a secretory state (163, 168, 169, 247-249).

The mammary gland is controlled by a number of hormones and local factors. When these are added to various mammary gland preparations, effects on transport may be observed. However, it is difficult to determine whether a hormone is having a primary, direct effect on a particular transport system or whether the effect is due to the overall response of the cell, to increase the degree of differentiation, for example, or simply to prolong its survival. Therefore, caution must be exercised in interpreting the individual effects of hormones on transport systems in such highly integrated differentiation, secretion, and transport systems.

	II. EXPERIMENTAL METHODS USED TO STUDY MAMMARY GLAND SOLUTE TRANSPORT

Top Previous Next References

The first physiological studies on the mechanism of milk secretion were done using conscious undisturbed animals. Although the range of measurements that can be made is limited, the major advantage is that artifacts are not introduced as part of the experimental procedure of removing tissue and trying to maintain it alive and undamaged. It is relatively easy to measure transepithelial solute gradients across the lactating mammary gland, and moreover, because of the slow rates of milk secretion, it is possible to make some meaningful electrophysiological measurements; the effect of changing the luminal contents on the transepithelial potential difference can be readily tested, for example (27).

Many studies on the mechanism of solute transport have been done using mammary tissue slices, otherwise termed fragments and explants. Explants are still a popular choice because they are easily and rapidly prepared; they do not require biochemical treatment, such as collagenase digestion, which could possibly alter the properties of the alveolar cells. Although explants are a mixed cell population, the vast majority, at least in those prepared from lactating tissue, are secretory cells. The alveolar nature of mammary tissue means that explants will give a good measure of transport across the basolateral membranes of the secretory cells particularly in experiments employing short time courses. However, the relatively large extracellular space of mammary tissue explants together with the presence of large unstirred fluid layers places limitations on the design of experiments. Isolated mammary epithelial cells and acini are an alternative to explants to study membrane transport. Like explants, they are relatively easy to prepare; however, the isolation process requires that the cells be exposed to digestive agents. The extracellular space is minimal compared with that associated with explants, but it is difficult to distinguish between transport occurring at the apical and basolateral membranes in both these preparations. There is also the possibility that dissociation of mammary tissue leads not only to a loss of polarity but also decreases cell viability.

The perfused lactating mammary gland has been successfully employed to characterize solute transport systems in both guinea pig and rat. This technique allows the transport properties of the blood-facing aspect of the mammary epithelium to be characterized under near-physiological conditions. One major assumption is that the endothelium does not constitute a major barrier to solute movement between blood and the interstitium. Recent work suggests that this is a reasonable assumption, since sucrose (an extracellular marker) has a larger dilution space than albumin (an intravascular marker) in the lactating bovine mammary gland (176). The major drawback is that the skills required to achieve a successful preparation take time to acquire. The perfused mammary gland, used in conjunction with the rapid, paired-tracer dilution technique (255), has been successfully employed to study amino acid transport.

Cultured mammary cells have also been used to study mammary transport processes. For example, cells cultured on collagen gels to form monolayers and subsequently mounted in Ussing chambers have allowed the polarity of the mammary epithelium to be examined. However, there is justifiable concern that cultured mammary epithelial cells are not fully active in terms of secretion, although an exception to this may be mammary epithelial cells cultured on a reconstituted basement membrane. Cells grown in this manner form polarized acinar structures that enclose a hollow, sealed lumen and secrete in a vectorial fashion (13, 42, 89). These acini, termed mammospheres, may prove particularly useful for characterizing transport across the basolateral membranes of secretory cells. Unfortunately, a technique that allows purified basolateral membrane vesicles to be prepared from mammary secretory cells has not yet been developed.

Membrane fractions isolated from milk have been used in attempts to study the transport properties of the apical (luminal) membrane. Wooding et al. (254) identified membrane-bound cytoplasmic droplets in goat's milk containing fat and protein particles, mitochondria, and Golgi apparatus (but no nuclei); further studies showed that the droplets were capable of triacylglycerol biosynthesis (44). Because the cytoplasmic fragments originate from the apical aspect of mammary secretory cells, it is reasonable to assume that the surrounding membrane is indeed representative of the apical membrane. It has been shown that cytoplasmic droplets transport ions into an osmotically active space (213).

Because the milk fat globule becomes enveloped by membrane during its extrusion from the cell, the milk-fat-globule membrane has long been assumed to be representative of the apical membrane of the secretory cell. However, the finding that bovine milk-fat-globule membrane vesicles are impermeable to K⁺ casts doubt on the accepted origin or maintenance of original properties of these membranes (195, 196) (Fig. 1). There is the possibility that the milk-fat-globule membrane is derived from intracellular membranes rather than the apical membrane (228, 253). On the other hand, membrane vesicles prepared from bovine skim milk, which are distinct from milk-fat-globule membranes, are able to transport solutes into an intravesicular compartment (196). Skim milk membrane vesicles do appear to originate from the apical membrane (172).

View larger version (19K): [in this window] [in a new window]   Fig. 1. K+ (Rb+) uptake by milk-fat-globule membrane vesicles (squares) and skim milk membrane vesicles (circles) prepared from bovine milk in the absence (open symbols) and presence (solid symbols) of valinomycin. Previously, it has been argued that both of these membrane preparations originate from the apical aspect of mammary secretory cells. It is apparent that skim milk membrane vesicles are permeable to K+ in the absence and presence of valinomycin. In contrast, milk-fat-globule membranes only display a significant permeability to K+ in the presence of valinomycin. Data are means ± SE (n = 3). The results suggest that skim milk membranes may be a better marker for the apical membrane of mammary secretory cells than milk-fat-globule membranes. [Adapted from Shennan (195).]

In ending this section, it should be noted that the relatively complex morphology of the mammary gland precludes the use of several techniques to study solute transport. For example, the mammary epithelium is not flat and therefore, unlike frog skin and toad bladder, cannot be mounted in Ussing chambers.

	III. ROUTES OF SECRETION

Top Previous Next References

There are five major, known routes of secretion across the mammary secretory epithelium from the blood side to milk, four transcellular and one paracellular (116, 118) (Fig. 2): 1) membrane route, 2) Golgi route, 3) milk fat route, 4) transcytosis, and 5) paracellular route.

View larger version (73K): [in this window] [in a new window]   Fig. 2. Diagram showing the known major routes of secretion across the mammary epithelium. Routes 1, 2, 3, and 5 correspond to the descriptions given in text. For the sake of clarity, route 4 (transcytosis) is not shown. [Adapted from Linzell and Peaker (118).]

1) In the membrane route, substances may traverse the apical cell membrane (and for those directly derived from blood, the basolateral membrane). Examples are water, urea, glucose, Na⁺, K⁺, and Cl (116).

2) In the Golgi route, secretory products are transported to or sequestered by the Golgi apparatus and secreted into the milk space by exocytosis. Examples are casein, whey proteins, lactose, citrate, and calcium (116, 118, 267).

3) In the milk fat route, milk fat globules are extruded from the apex of the secretory cell surrounded by membrane (milk-fat-globule membrane); some cytoplasm is sometimes included (126). Examples are milk fat, lipid-soluble hormones and drugs, some unknown growth factors (in milk-fat-globule membrane) (244), and the product of the ob gene, leptin (214).

4) In transcytosis, vesicular transport involves various organelles, possibly, in some cases also involving extrusion by route 2 (88, 154). Examples are immunoglobulins during colostrum formation (88), transferrin (154), and prolactin (154).

5) In the paracellular route, there is direct passage from interstitial fluid to milk.

In full lactation, routes 1, 2, 3, and, possibly, 4 predominate.

Transcellular transport requires solutes to cross the constituent plasma membranes of the epithelium, whereas paracellular transfer involves movement between the cells via leaky tight junctions. In most species, solute transport across the lactating mammary epithelium appears to be mainly transcellular, i.e., the zonulae occludentes are tight junctions. However, there are situations where a significant flux occurs via a paracellular pathway. Movement of solutes across the paracellular shunt pathway occurs during pregnancy and, in some species, during late lactation. The early (160) and later work (150) on this aspect of mammary transport including physiological and morphological findings have been reviewed, and it is not our intention to provide a further detailed review here.

The transepithelial electrical resistance is a good measure of the tightness (or leakiness) of an epithelium. However, although it is possible to measure the transepithelial resistance of cultured mammary cell monolayers, it is more difficult to obtain reliable values of the electrical resistance across the lactating mammary gland in vivo. A common way of assessing paracellular permeability in the lactating mammary gland is to measure the solute content of milk; the concentrations of lactose and K⁺ in milk decrease (moving down their concentration gradients), whereas those of Na⁺ and Cl increase when the junctions become leaky. However, care must be exercised when relying solely on Na⁺ and K⁺ gradients because a change may simply reflect an alteration in the transport of these ions across transcellular pathways. A more preferable way of assessing tight junction permeability in vivo is to measure the rate of appearance of lactose in plasma or of an inert disaccharide introduced into plasma in milk (119, 216).

In general, the paracellular pathway is absent during lactation and present during pregnancy. However, in the rabbit during late lactation, when the rate of secretion is still relatively high, milk lactose and K⁺ concentrations decrease while those of Na⁺ and Cl increase. At this stage, the epithelium is permeable to lactose and sucrose and more permeable to Na⁺ and Cl (167). These changes were reversed by exogenous prolactin (121). Recent evidence suggests that prolactin may also be involved in some way in maintaining the integrity of the tight junctions (judged by the ability of the gland to maintain solute concentration gradients) in the rat mammary gland (71). There is also evidence that a paracellular pathway appears during other physiological states. Characteristic changes in milk composition were observed in many goats for 1-2 days up to 4 days before estrus (164), but the controlling mechanism remains unknown.

There is a major change in the permeability of the paracellular pathway at the onset of copious milk secretion around the time of parturition (119). There is also good evidence which suggests that glucocorticoids are involved in controlling the formation of tight junctions during mammary development and differentiation in pregnancy. The glucocorticoid dexamethasone increased the transepithelial electrical resistance and decreased the transepithelial movements of labeled mannitol and inulin across monolayers of a mouse mammary cell line (261). Glucocorticoid-induced formation of mouse mammary tight junctions involves an increase in the amount of the tight junction-associated protein ZO-1 and phosphorylation/dephosphorylation events since okadaic acid, a protein serine/threonine phosphatase inhibitor, attenuates the dexamethasone-induced increase in transepithelial resistance (212). However, mammary development is a complex series of events, and transforming growth factor (TGF)- was found to antagonize the effect of dexamethasone (252). There is evidence from in vivo studies that glucocorticoids may control the permeability of tight junctions. In late-pregnant goats, an intraluminal injection of cortisol had marked effects on the composition of fluid within the lumen: K⁺ concentration increased and Na⁺ decreased, which is consistent with the closing of tight junctions (224). However, an effect of cortisol on transcellular ion transport cannot be ruled out at this stage. Hormonal events, other than or as well as a rise in circulating glucocorticoid concentration, appear to be involved in the final closure near parturition, with the hormonal events leading to parturition and the onset of copious milk secretion (lactogenesis stage II, Ref. 68) being closely linked. For example, in late pregnancy, the goat mammary gland synthesizes PGF₂; near term, this production stops. Unilateral treatment with a PGF₂ analog prevented the normal changes in milk composition and increase in the rate of secretion (127). The current view on the control of tight junctional permeability during lactogenesis has been well summarized by Nguyen and Neville (150)

We propose that progesterone inhibits tight junction closure during pregnancy, possibly through or along with local factors such as TGF- and PGF₂. As the level of progesterone drops during parturition or with ovariectomy, its inhibitory effect is removed, allowing the mammary epithelial tight junctions to close. However, glucocorticoid may be required for final differentiation of the mammary epithelium from the pregnant animal into the epithelium of the lactating animal with its highly impermeable tight junctions. Prolactin appears to play a role in the maintenance of the mammary epithelium in the lactating state and inhibition of programmed cell death.

The paracellular pathway reappears at the cessation of lactation. When milk removal was stopped in goats during late lactation, the key process was determined to be the local event of stretching the mammary epithelium by milk accumulating within the lumen of the gland, leading to a decrease in secretory and metabolic activity and an increase in the permeability of the tight junctions. Systemic control by hormones was excluded in this species as was pressure of the accumulated milk simply bursting the taut epithelium (69, 161). Similarly, in the cow, milk stasis leads to an increase in the paracellular pathway that can be reversed by milking (217). These as well as other studies indicate an inverse link between secretory activity and permeability of the tight junctions. Whether maintenance of impermeable junctions is part of the secretory activity of the cell, like the synthesis of milk components, or whether disruption of the junctions is followed by partial depolarization of the cell which affects secretory activity, or whether there is signaling from the junctions to the synthetic and secretory pathways is not known (150).

The presence of a paracellular pathway may confound studies on milk composition and mammary epithelial permeability. Disruption of the epithelium is evident in mastitis but also when supraphysiological doses of oxytocin are given to contract the myoepithelium and obtain a milk sample. Care must be exercised when interpreting data on milk composition, particularly of the aqueous phase (160).

	IV. SODIUM, POTASSIUM, AND CHLORIDE TRANSPORT

Top Previous Next References

The mammary gland is able to generate and maintain large Na⁺, K⁺, and Cl gradients between milk and plasma. These ions make a substantial contribution to the osmolality of milk, especially in those species where the lactose content is relatively low. On the basis of measurements of the concentration gradients between interstitial or extracellular fluid, intracellular fluid, and milk, and the electrical potential gradients between these three compartments, Linzell and Peaker in 1971 (117, 118) proposed a scheme to account for the movement of Na⁺, K⁺, and Cl across the lactating mammary epithelium and the maintenance of milk composition (Fig. 3). In their hypothesis, the main features are that the Na⁺ concentration inside the cell is kept low, and the K⁺ concentration high, by the action of a Na⁺-K⁺ pump on the basolateral membrane, whereas these ions are freely distributed between the interior of the cell and milk according to the electrical potential difference across the apical membrane. Milk is electrically positive with respect to the cell, so the concentrations of Na⁺ and K⁺ are lower in milk, but the ratio of these ions is similar in both compartments, i.e., ~3 K⁺:1 Na⁺. The situation for Cl is, however, different from that of Na⁺ and K⁺. The concentration of Cl is higher inside mammary cells than that expected for an equilibrium distribution, suggesting that a mechanism exists in both the basolateral and apical membranes which "actively" pumps Cl into the cells. Many experimental observations, at the level of individual transport systems, supporting the original hypothesis of Linzell and Peaker have since been made. The relations between lactose concentration, apical membrane potential difference, and ionic concentrations have also been considered (160, 159).

View larger version (19K): [in this window] [in a new window]   Fig. 3. Profile of ion distribution and membrane potentials (both measured and calculated) in guinea pig mammary tissue. [Modified from Linzell and Peaker (118).]

Three carrier mechanism were postulated in the scheme of Linzell and Peaker: 1) a Na⁺-K⁺ pump on the basolateral membrane, 2) an inwardly directed Cl pump on the apical membrane, and 3) a similar Cl transporter in the basolateral membrane of the secretory cell (117, 118). Evidence for the presence of a Na⁺-K⁺ pump in the blood-facing aspect of the mammary epithelium is strong, and several candidates for Cl pumps have been identified.

Many studies have shown that lactating mammary tissue from a variety of species displays ouabain-sensitive Na⁺-K⁺-ATPase activity (242) and ouabain-sensitive Na⁺ and K⁺ transport (3, 61, 62, 117, 194). In addition, there is good histochemical evidence that Na⁺-K⁺-ATPase is located on the basal and lateral membranes, but not the apical membranes, of mammary secretory cells (96, 105). These findings taken together support the scheme of Linzell and Peaker that lactating mammary tissue expresses a functional Na⁺-K⁺ pump on the blood-facing pole of the epithelium. Experiments using cultured mammary epithelial cells have provided evidence that the Na⁺-K⁺ pump is located on the basolateral, but not the apical, membranes (24-26). However, there is evidence for Na⁺-K⁺-ATPase activity on the milk-fat-globule membrane, albeit at a relatively low specific activity (55).

Na⁺-K⁺-Cl cotransport could be a candidate for the Cl pump in the model of Linzell and Peaker. Na⁺-K⁺-Cl cotransport has been identified in lactating rat mammary tissue (194, 203). Thus K⁺ (Rb⁺) transport (both influx and efflux) by mammary tissue explants is Na⁺ dependent, Cl dependent, and inhibited by the loop diuretic furosemide. In addition, Cl efflux from rat mammary tissue explants is sensitive to furosemide (203). On the basis that cotransport activity was detected in explants, it is predicted that the Na⁺-K⁺-Cl cotransport system is situated in the blood-facing aspect of the epithelium. Supporting evidence of Na⁺-K⁺-Cl expression in mammary tissue is the finding that bovine milk-fat-globule membrane binds bumetanide (an analog of furosemide) in a fashion that is dependent on Na⁺, K⁺, and Cl (166).

The triple cotransporter is able to drive the uphill transport of Cl (2). The Na⁺-K⁺-Cl cotransport system is electroneutral and is generally assumed to operate with a stoichiometry of 1 Na⁺:1 K⁺:2 Cl (43, 79) The free-energy (G) available to a cotransporter operating with such a stoichiometry is given by

(1)

where subscripts i and o refer to intracellular and extracellular, respectively. Applying the Na⁺, K⁺, and Cl concentrations of guinea pig mammary tissue and plasma (117) to Equation 1 gives a value +0.79RT for G, suggesting that the triple cotransport system operating with a coupling ratio of 1 Na⁺:1 K⁺:2 Cl would require an input of free energy to drive inward transport. Alternatively, a Na⁺-K⁺-Cl cotransporter with a stoichiometry of 2 Na⁺:1 K⁺:3 Cl, as found in the squid axon (190), would have sufficient freeenergy (G = 1.03RT) associated with the ion gradients to accomplish inward transport. However, it is clear that further experiments are required to determine both the stoichiometry and locus of the mammary Na⁺-K⁺-Cl cotransport mechanism.

Cl/HCO₃ exchange could also be a candidate for a Cl pump in lactating mammary tissue (120). For example, a Cl/HCO₃ antiport drives the uphill accumulation of Cl into smooth muscle cells (2, 50). A DIDS-sensitive anion exchanger that accepts Cl, SO₄², I, and NO₃ as substrates has been reported in lactating rat mammary tissue explants (194, 203). However, it remains to be shown whether the rat mammary tissue anion-exchanger transports HCO₃ and, if so, whether it is able to act as a Cl pump.

The 1971 model of Linzell and Peaker predicts that the apical membrane of mammary secretory cells possesses Na⁺, K⁺, and Cl channels. Several lines of experimental evidence support this prediction. First, Blatchford and Peaker (27) have shown that the transepithelial potential difference across the lactating goat mammary gland can be altered by varying the Na⁺, K⁺, and/or Cl concentration in the lumen of the gland in a fashion that is consistent with the presence of channels for each of these ions in the apical aspect. However, it was found that the apical membrane is unable to discriminate between Na⁺ and K⁺, suggesting that the apical membrane may express a nonspecific cation channel rather than separate channels for Na⁺ and K⁺. In this connection, nonspecific cation channels have been identified in cultured mouse mammary epithelial cells (72). Second, cytoplasmic fragments isolated from goat's milk that are surrounded by apical membrane, transport K⁺(Rb⁺) via barium- and quinine-sensitive channels (213). Third, skim milk membranes, isolated from bovine milk, which appear to be derived from the apical aspect of secretory cells, express K⁺ and Cl channels (196). The channels identified in the membrane preparations need to be characterized using electrophysiological techniques.

Prolactin was found to decrease the Na⁺ and increase the K⁺ content of mammary tissue taken from rabbits treated with bromocriptine, a drug that prevents prolactin release; ouabain inhibited the action (62, 63). In addition, the unidirectional uptake of K⁺(Rb⁺) by rabbit mammary explants was stimulated by prolactin via a pathway sensitive to ouabain (61). These observations are consistent with an effect of prolactin-induced differentiation of the cells on the number of Na⁺-K⁺ pumps present either induced directly or through an increase in Na⁺ permeability. In this connection it is interesting to note that prolactin upregulates Na⁺-K⁺-Cl cotransport in rat mammary tissue (202). Work using cultured cells supports the claim that prolactin-induced differentiation involves changes in mammary Na⁺ and K⁺ transport; prolactin increases the mucosal-to-serosal transfer of Na⁺ across monolayers of cultured mouse mammary epithelial cells via a ouabain-sensitive pathway (24-26).

	V. RELATIONS BETWEEN SODIUM, POTASSIUM, AND CHLORIDE TRANSPORT AND WATER MOVEMENTS

Top Previous Next References

It is recognized that the synthesis and secretion of lactose is responsible for the majority of water in milk. However, this cannot be the whole story, since water still appears in the milk of those species (i.e., pinnipeds) which contain little or even no lactose or any other sugar (151). In a number of secretory epithelia, it is accepted that the intracellular accumulation of Cl, via Na⁺-K⁺-Cl cotransport across the basolateral membrane, is the driving force for the secretion of ions and water across the apical membrane (43, 153). The same process could occur in the mammary secretory cell, since the necessary gradients exist. Therefore, a situation could arise whereby the secretion of water by the mammary gland could be driven partly by the secretion of lactose and partly by the secretion of ions. Variations in the concentrations of lactose and ions within and between species could be related to the relative flux between these two mechanisms. This hypothesis, together with the involvement of paracellular ion flow in some species, may explain how secretion of the aqueous phase is achieved in all mammals.

	VI. PHOSPHATE TRANSPORT

Top Previous Next References

The suckling requires a considerable amount of phosphate for normal growth and development. There are at least three chemical forms of phosphate in milk: free inorganic orthophosphate (P_i) in solution as HPO₄² and H₂PO₄; colloidal phosphate, the majority of which is calcium phosphate associated with casein micelles; and casein phosphate (86). For example, of the total phosphate (20.5 mM) present in goat's milk, 6.7 mM is P_i, 8.9 mM is colloidal, and 4.9 mM is casein bound (146). Despite the complex nature of phosphate in milk, it can be predicted that milk phosphate is derived from plasma P_i that has to be transported across the basolateral membranes of the secretory cells. In accordance with this prediction, the mammary gland has the capacity to extract large quantities of P_i from blood. For example, it is estimated that the entire pool of plasma P_i is replaced every 10 min in lactating rats (30).

The mechanism of P_i transport by the lactating mammary gland has only recently received attention. P_i uptake by rat mammary tissue explants and the perfused rat mammary gland occurs predominantly via a Na⁺-dependent process (210).The Na⁺-dependent moiety of P_i uptake in the latter preparation operates with a Michaelis constant (K_m) of 40 µM with respect to the extracellular P_i concentration. Furthermore, P_i efflux from mammary tissue can be stimulated by reversing the trans-membrane Na⁺ gradient. These findings are consistent with the presence of a Na⁺-P_i cotransport system in the basolateral aspect of the mammary epithelium. Na⁺-P_i cotransport mechanisms have been identified in a number of epithelia including the intestine, kidney, and placenta (21, 23, 33), several of which have been cloned and sequenced (125). There are, however, several differences between the properties of the Na⁺-dependent P_i transporter in mammary tissue and those described in other epithelia. First, P_i uptake by mammary tissue is only weakly inhibited by structural analogs; arsenate and phosphonoformic acid (PFA) were found to be weak or ineffective inhibitors of P_i transport. In contrast, both of these compounds have been shown to be high-affinity blockers of Na⁺-dependent P_i transport in other tissues (103). Second, the kinetic parameters, with respect to Na⁺ activation, are markedly different; the stimulation of P_i uptake by external Na⁺ did not exhibit sigmoidal kinetics as found, for example, in the kidney (125).

Under certain experimental conditions P_i transport in lactating rat mammary tissue occurs via an anion-exchange system (210). External P_i is able to trans-accelerate P_i efflux via a mechanism sensitive to DIDS. It appears that the P_i self-exchange system is distinct from the anion exchanger that accepts Cl and SO₄² as substrates (194), since P_i efflux is not trans-stimulated by Cl and SO₄² and P_i does not trans-accelerate SO₄² efflux (210). Moreover, the finding that the P_i self-exchanger is not Na⁺ dependent suggests that the exchange mechanism is distinct from the Na⁺-P_i cotransport system. However, the physiological significance of the P_i exchanger remains to be established.

The finding that P_i in milk is higher than that of plasma could, at first sight, be taken as evidence that the apical membrane expresses a P_i transport mechanism. However, the major pathway for P_i secretion into milk is believed to be via the Golgi vesicle route. This hypothesis is based on two important findings. First, Neville and Peaker (146) found that the apical aspect of the lactating goat mammary epithelium is almost impermeable to P_i. Second, the time course of P_i secretion into milk, following an intravascular injection of radiolabeled P_i, was similar to that of casein phosphate and colloidal phosphate. This finding raises questions about the nature of P_i transport across the Golgi membrane. There is, of course, the possibility that Golgi secretory vesicles do not need to transport P_i per se from the cytosol because sufficient P_i is generated in the Golgi lumen by the hydrolysis of UDP during lactose synthesis (Fig. 4). However, this mode of P_i generation will not occur in the Golgi lumen of pinnipeds.

View larger version (30K): [in this window] [in a new window]   Fig. 4. A schematic diagram of the mechanisms that may be involved in the transport and metabolism of phosphate and Ca2+ in mammary secretory cells. Phosphate is transported across the basolateral membrane via a (Na+-Pi) cotransport mechanism and possibly by an anion-exchange system. Pi is formed within the lumen of the Golgi during the synthesis of lactose as a consequence of the hydrolysis of UDP-galactose. In addition, Pi is generated during the process of casein phosphorylation. The mechanism of Ca2+ uptake across the basolateral membrane has not been precisely identified but may be a Ca2+ channel. Stretch-activated Ca2+ channels may also provide a pathway for Ca2+ uptake. Ca2+ is pumped across Golgi membranes by a Ca2+-ATPase. Pi and Ca2+ are transported into milk along with lactose, casein, and citrate after exocytosis of Golgi-derived vesicles at the apical aspect of the secretory cells. LS, lactose synthase; CK, casein kinase; A, ADPase.

	VII. CALCIUM TRANSPORT

Top Previous Next References

Milk is a calcium-rich fluid (94). For example, the total calcium concentration in rabbit's milk can approach a concentration of 100 mM. Accordingly, during lactation, the mammary gland extracts large quantities of calcium from plasma to meet the requirements of the growing neonate. Milk calcium, like phosphate, exists in several chemical forms: free ionized calcium, casein-bound calcium, and calcium complexed to inorganic anions such as phosphate and citrate. In most milks, calcium associated with casein micelles comprises the majority of the total calcium present in milk (144). Neville and Peaker (146) found that the mammary gland does not transport calcium in the milk-to-blood direction, suggesting that calcium cannot cross the apical membrane or tight junctions. Thus it can be inferred that calcium transport into milk is transcellular. Furthermore, because the time course of appearance of labeled milk calcium is similar to that of casein and lactose, it has been suggested that calcium is secreted via the Golgi vesicle route (146). It follows that casein-bound calcium and colloidal calcium, along with free ionized calcium, are all derived from plasma Ca²⁺, indicating that Ca²⁺ must be able to cross both the basolateral and Golgi membranes before secretion.

The first question that arises is, What mechanism transports calcium across the basolateral membrane? Surprisingly, there is a paucity of information about the system(s) available for calcium transport across the blood-facing membrane except for the finding that it is a temperature-sensitive process that increases during the transition from pregnancy to lactation (147). Whatever the nature of the system(s), it must operate with a high capacity given the large amount of calcium secreted into milk. Calcium channels would of course fulfill such a role. However, until the mechanism of calcium uptake by mammary secretory cells is thoroughly studied, the presence of calcium channels in the basolateral membrane can only remain a suggestion.

Mammary epithelial cells maintain intracellular Ca²⁺ at a low concentration (56, 75). This observation prompts the next question, How do mammary epithelial cells maintain a low cytosolic Ca²⁺ concentration while facilitating a large transcellular calcium flux? A low calcium concentration could be maintained if Ca²⁺ were pumped into intracellular organelles at a rate that matched the entry of calcium across the basolateral membranes of the secretory cells. In this connection, there is evidence suggesting that Golgi membranes express an ATP-driven Ca²⁺ pump (149, 241, 245). Golgi-derived vesicles are able to accumulate Ca²⁺ by a saturable process (K_m = 0.8 µM) that requires the hydrolysis of ATP (Fig. 4). A Ca²⁺ pump situated in the basolateral membranes of mammary secretory cells could also act to keep intracellular Ca²⁺ at a low concentration. However, no evidence for such a transport system has been presented. In addition, no evidence for a basolateral Na⁺/Ca²⁺ exchanger, another transport system which could regulate intracellular Ca²⁺, has been found (143).

Ca²⁺ uptake by mammary tissue may be a hormonally regulated process. Ca²⁺ accumulation by cultured mouse mammary explants, presumably via the basolateral membranes, is stimulated by 1,25-(OH)₂D₃, the hormonal form of vitamin D, in a dose-dependent fashion (137), and Ca²⁺ transfer into goat's milk is increased by parathyroid hormone-related protein (15). However, in both of these examples, the target mechanism is unknown.

	VIII. CITRATE TRANSPORT

Top Previous Next References

Citrate exists in the aqueous phase of milk in a variety of chemical forms such as calcium citrate, citrate², and citrate³. The relative proportion of each chemical form depends on factors such as H⁺, Ca²⁺, and Mg²⁺ and thus citrate can be considered as an important milk buffer. The lactating goat mammary epithelium is impermeable to citrate in both directions, suggesting that citrate is synthesized within the secretory cells and released into milk after exocytosis of Golgi vesicles (116). In this connection, a Golgi-enriched fraction isolated from the bovine mammary gland accumulated citrate in a fashion sensitive to temperature, suggesting that a citrate transport system may be present in the Golgi apparatus membrane (267). The putative citrate transport mechanism in the Golgi membranes remains to be fully characterized.

	IX. IODIDE TRANSPORT

Top Previous Next References

The concentration of iodide in milk can be 20- to 30-fold higher than that found in maternal plasma (32), matching the requirements of the suckling for the synthesis of thyroid hormones. Iodide transport was investigated relatively early because of the contamination of land and milk supplies by radioactive iodine isotopes from nuclear explosions or accidents. In addition, it was thought that an understanding of the mammary iodide transport mechanism could help limit the transfer of radiolabeled iodide into the food chain via milk after environmental contamination.

The iodide transport system in the thyroid gland has recently been cloned (47). It is a Na⁺-dependent mechanism sensitive to anions such as perchlorate and thiocyanate (155). A body of evidence suggests that mammary epithelial cells express a similar iodide transport mechanism. The lactating mammary gland has the ability to accumulate iodide into milk against a high concentration gradient by a system that is inhibited by perchlorate and thiocyanate (31, 32, 53, 114). Iodide uptake by mammary tissue explants prepared from midpregnant mice is a Na⁺-dependent process stimulated by prolactin (184, 185). Moreover, mRNA encoding for the Na⁺-dependent iodide transporter has been identified in the lactating human mammary gland (215).

Although the Na⁺-dependent mechanism is probably the predominant pathway for iodide uptake by mammary cells, there is evidence that iodide also enters via a DIDS-sensitive anion-exchange pathway; sulfate efflux from rat mammary tissue can be stimulated by extracellular iodide (194).

	X. STRETCH-SENSITIVE ION TRANSPORT MECHANISMS

Top Previous Next References

Milk accumulation in the alveolar lumen leads to distension of the secretory alveoli, and cellular deformation may play an important role in determining secretory activity during the cessation of lactation. An increase in intramammary pressure at the end of lactation has been shown to reduce the rate of milk secretion in the goat (161), suggesting that, in addition to local, autocrine chemical control of milk secretion by the milk protein FIL, mammary distension could also play a part in controlling the rate of milk secretion in response to demand by the young (48, 163, 168, 169, 247-249).

Cell distension has marked effects on mammary gland ion transport. Enomoto et al. (58) found that mechanical stimulation of cultured mouse mammary cells, induced by touching with a blunt microelectrode, resulted in a short depolarization (1-8 s) followed by a prolonged (10-40 s) hyperpolarization. At present, the locus of the stretch-sensitive channels (apical and/or basolateral) on the mammary epithelial cells is not known. The mechanism underlying the depolarization remains to be determined, but it has been established that the hyperpolarization can be attributed to the activation of quinine-sensitive, Ca²⁺-dependent K⁺ channels (58, 59, 72). Accordingly, mechanical stimulation of cultured mouse mammary cells leads to an increase in the intracellular Ca²⁺ concentration. However, the increase in the Ca²⁺ concentration is not confined to the touched cell but also spreads to surrounding cells, even to those that have no direct physical contact with the stimulated cell, suggesting that a paracrine factor may be involved (60, 73). There is evidence that ATP is released from cultured mammary cells upon mechanical stimulation (73). In this connection, extracellular ATP, acting via P₂-purinergic receptors, has been shown to increase intracellular Ca²⁺ in human breast tumor cells (70). Cell swelling, induced by hyposmotic shock, increases the intracellular Ca²⁺ concentration in acini isolated from lactating mouse mammary gland (221). The effect consists of a rapid and large transient increase in intracellular Ca²⁺ followed by a sustained plateau phase. The increase in Ca²⁺ is due to increased Ca²⁺ influx from the incubation medium rather than release from intracellular stores. The nature and precise locus of the Ca²⁺ transport pathway is unknown, but it is insensitive to nifedipine (221). Moreover, it is not yet known if the primary stimulus is membrane stretch or a change in the cellular hydration state.

Based on these still sparse observations, we hypothesize that stretch induces ATP release into the milk or interstitial space, where it interacts with neighboring cells leading to an increase in calcium entry into these cells.

	XI. CHOLINE AND CARNITINE TRANSPORT

Top Previous Next References

The organic cation choline is required by the suckling for normal growth and development including the synthesis of ACh, phosphatidylcholine, and sphingomyelin (258). The concentrations of choline (and its metabolites) in rat's milk and plasma are 1.5 mM and 12 µM, respectively, suggesting that lactation may put a considerable strain on the maternal choline stores (259). The concentration of free choline in rat's milk, 182 µM, is consistent with active choline transport. Lactating rat mammary epithelial cells possess a high-affinity (K_m = 35 µM), Na⁺-dependent choline carrier that can be inhibited by hemicholinium-3 (41). Rat mammary epithelial cells can maintain high concentrations of choline (41), suggesting that the Na⁺-dependent carrier is situated in the basolateral membrane and that the transepithelial choline gradient is generated at the basal membrane of the mammary cell.

L-Carnitine (4-N-trimethylammonia-3-hydroxybutanoate) is essential for the transport of long-chain fatty acids between cell compartments. For example, the formation of long-chain acylcarnitines from their respective coenzyme A esters enables the acyl moieties to cross the mitochondrial membrane where they are subjected to -oxidation (256). L-Carnitine-dependent oxidation of long-chain fatty acids is particularly important for brain development in the neonate. The transport of L-carnitine into milk is physiologically significant because the newborn has a limited capacity for L-carnitine biosynthesis. In rats, the lactating female depletes its reserves of L-carnitine through secretion into milk; the rate of L-carnitine secretion is high during the first 3-4 days of lactation but thereafter falls as the young develop the capacity to synthesize L-carnitine (186). In the early stages of lactation, the L-carnitine concentration in rat milk is almost an order of magnitude greater than that of plasma (320 vs. 35 µM), suggesting that mammary epithelial cells possess a carrier for L-carnitine. In accordance with this prediction, a concentrative, Na⁺-dependent L-carnitine carrier has been identified in mammary explants isolated from rats during the early phase of lactation (201, 257). The carrier operates with a K_m of 132 µM and maximum velocity (V_max) of 100 pmol · h¹ · mg intracellular water¹. If, as expected, the carrier is situated on the basolateral membrane of the alveolar cells it will not be saturated with plasma L-carnitine under physiological conditions. However, the K_m value obtained with explants is probably an overestimate because of the presence of unstirred fluid layers. The rat mammary L-carnitine carrier is not stereospecific, since D-carnitine is an effective cis-inhibitor. In addition, acetylcarnitine, but not choline or taurine, inhibits L-carnitine uptake by rat mammary explants. These characteristics suggest that the mammary tissue L-carnitine carrier is similar to that of renal brush-border membrane vesicles, neuroblastoma cells, and JAR cells (140, 174, 182). The activity of the Na⁺-dependent L-carnitine transport system was found to decrease as lactation progressed (201).

L-Carnitine efflux from lactating rat mammary tissue is a slow process that is Na⁺ independent and cannot be stimulated by external L-carnitine. However, cell-swelling, induced by a hyposmotic shock, increases the fractional release of radiolabeled L-carnitine from rat mammary tissue (201). The possibility that volume-activated L-carnitine release is mediated via the volume-sensitive amino acid transport pathway should be investigated.

	XII. GLUCOSE TRANSPORT

Top Previous Next References

The concentration of lactose in the milk of most species is relatively high, and lactose is one of the major milk osmolytes; for example, the concentrations in human and bovine milk are, respectively, 204 and 140 mM (158). Lactose is synthesized within the lumen of the Golgi apparatus by an enzyme complex collectively known as lactose synthase. Exocytosis of the Golgi secretory vesicles then occurs at the luminal side of the cell. Golgi and apical membranes of mammary secretory cells are impermeable to lactose; therefore, because these membranes are freely permeable to water, the volume of milk secreted is closely related to the rate of lactose synthesis.

The lactating mammary gland has a high demand for D-glucose because it is the obligate precursor for lactose synthesis and, depending on species, is also used to synthesize lipids and amino acids (110, 115). Glucose transport across both the basolateral and Golgi membranes of the secretory cells is therefore an important process in milk secretion.

A.  Glucose Transporters

1.  GLUT1 expression

Lactating mammary tissue transports D-glucose via a facilitative transport system similar, if not identical, to GLUT1. Amato and Loizzi (4) found that guinea pig mammary tissue slices are capable of transporting 2-deoxyglucose, a substrate of GLUT1 type transporters (74), via a system inhibited by cytochalasin B. Rat mammary acini also transport 2-deoxyglucose and 3-O-methyl-D-glucose via a temperature-sensitive and saturable mechanism (K_m for 2-deoxyglucose = 16 mM) that is inhibited by cytochalasin B and phloretin (225). Carrier-mediated D-glucose uptake with characteristics of GLUT1 has also been described in mouse mammary acini (177); thus 3-O-methyl-D-glucose uptake by mouse mammary cells is via a saturable (K_m = 14 mM), cytochalasin B-sensitive pathway.

In accordance with the flux studies quoted in the previous paragraph, it has been shown that rat mammary tissue expresses mRNA that hybridizes with a cDNA probe for GLUT1 as well as a protein that reacts with an antibody raised against the COOH terminus region of the human erythrocyte glucose GLUT1 transporter (35, 124). Immunofluorescence and immunoelectron studies have shown that GLUT1 expression is on the basolateral aspect of the lactating rat mammary epithelium (40, 222), consistent with the finding that 2-deoxyglucose transport occurs at the blood-facing aspect of the intact mammary gland (226). There are also several reports of GLUT1 expression, both at the mRNA and protein level, in lactating bovine mammary tissue (262-264).

Although the available evidence suggests that GLUT1 mediates the majority of D-glucose uptake by lactating mammary tissue, this conclusion would be strengthened if sensitivity to phloretin and cytochalasin could be demonstrated in the perfused lactating mammary gland and if the sequence of the lactating mammary tissue GLUT1 mRNA were known. It is notable, however, that guinea pig mammary tissue apparently has more than one type of saturable transport system for D-glucose (4). Thus cytochalasin B and phloretin do not completely inhibit the saturable component of 2-deoxyglucose uptake by guinea pig mammary tissue explants. In addition, a significant proportion of 2-deoxyglucose uptake by guinea pig mammary tissue slices is via a nonsaturable pathway. Therefore, it appears that GLUT1 may not be the sole mechanism for glucose transport across the basolateral membrane of mammary secretory cells.

No evidence has been found for the insulin-sensitive D-glucose transporter GLUT4 in lactating rat mammary epithelial cells (35, 40, 124). However, GLUT4 mRNA is expressed in mammary tissue taken from virgin and pregnant rats, but expression decreases as gestation progresses, suggesting that GLUT4 is expressed in mammary adipocytes rather than mammary epithelial cells (35, 40). No detectable expression of GLUT4 mRNA is evident in lactating bovine mammary tissue (263).

Lactating rat mammary epithelial cells do not appear to express GLUT2 or GLUT5 (35, 40). Similarly, lactating bovine mammary tissue does not express GLUT2 mRNA. Although low levels of GLUT3 and GLUT5 mRNA are found in the bovine mammary gland, the cellular location (i.e., epithelial or stromal) of these transcripts in the gland is not known (263).

There is evidence for the presence of a Na⁺-dependent glucose transporter in lactating rat mammary tissue (198). 1) The efflux of 3-O-methyl-D-glucose from lactating rat mammary tissue explants is stimulated by reversing the transmembrane Na⁺ gradient. This finding is consistent with a Na⁺-dependent D-glucose transporter operating in the reverse mode or the inhibition of D-glucose (re)uptake via a Na⁺-dependent mechanism. 2) mRNA extracted from the rat mammary gland contains a 4-kb transcript that hybridizes with the cDNA for the rabbit intestinal Na⁺-dependent D-glucose transporter (SGLT1). 3) A protein reacting specifically with an antibody raised against a synthetic hydrophilic nonadecapeptide corresponding to a proposed extramembranous loop of the rabbit intestinal SGLT1 has been found in Western blots of rat mammary cell membranes. 4) RT-PCR, using primers derived from the sequence of lamb intestinal SGLT1, generated a product (670 bp) from total rat mammary RNA identical in size to that obtained from intestinal RNA (211). Moreover, the sequence was identical to the corresponding region of intestinal SGLT1. In addition, the ovine mammary gland also expresses a SGLT1 type D-glucose transporter; the sequence is 1,995 bp in length and has the potential to code for a protein containing 664 amino acids with a sequence identical to that of the lamb intestinal SGLT1 (211). SGLT1 mRNA has also been found in lactating bovine mammary tissue (265).

The precise locus and functional significance of the mammary SGLT1-like transporter remains to be determined. If it is situated in the basolateral aspect of the mammary epithelium, then it may operate in parallel with GLUT1 to provide mammary acinar cells with D-glucose. However, an intracellular or apical location cannot be ruled out at this stage.

D-Glucose, along with a number of other monosaccharides, is present in milk of most species at a much lower concentration than that found in plasma (65). The available experimental evidence is consistent with the presence of a D-glucose transport system on the apical membrane of mammary secretory cells. Indeed, it has been suggested that milk D-glucose concentrations can be taken as a measure of the intracellular concentration (65, 142). D-Glucose and galactose introduced into the lactating goat mammary gland via the teat canal are rapidly lost from the milk space. The finding that fructose does not cross the mammary epithelium, even after 16 h, suggests that D-glucose and galactose cross the mammary epithelium via a transcellular route and not via a paracellular pathway (64). This conclusion is supported by the finding that radiolabeled 3-O-methyl-D-glucose introduced into the milk space of the goat mammary gland via the teat enters venous blood at a faster rate than radiolabeled sorbitol (65). The nature and significance of D-glucose transport across the apical membrane remains to be determined. If D-glucose is able to cross the apical membrane of rat mammary epithelial cells, then the available evidence suggests that the pathway is one other than GLUT1 because GLUT1 protein could not be detected on the apical membrane (40).

In contrast to most species (i.e., eutherian mammals), the Tammar wallaby can, in late lactation, maintain higher concentrations of glucose and galactose in milk than those in plasma (133, 134). This finding raises interesting questions about the nature of D-glucose and galactose transport across the apical membrane of the mammary epithelium in this species and other marsupials. For example, why aren't these monosaccharides rapidly lost from the milk space as would be predicted if the apical membrane possessed a facilitative transport mechanism?

D-Glucose must cross the Golgi membrane to reach the site of lactose synthesis. A Golgi vesicle fraction prepared from lactating rat mammary tissue was found to transport D-glucose; it was postulated that a pore was involved (246). However, the precise nature of D-glucose transport across the Golgi membrane remains to be established, although one report indicates that Golgi membranes possess a protein similar to GLUT1. Therefore, it is possible that GLUT1 transports D-glucose to the site of lactose synthesis; alternatively, Golgi-derived GLUT1 proteins may be destined for the plasma membrane (124).

Lactose synthesis and glucose uptake by the mammary gland increase abruptly at the time of parturition and decline rapidly as the gland stops secretion before involution (49, 69). Accordingly, the transport of 3-O-methly-D-glucose by isolated mouse mammary cells increased during progression from the pregnant to the midlactating state (177). A sharp decline in uptake was observed in the immediate postlactational period. A change in the V_max but not the K_m of 3-O-methyl-D-glucose uptake suggests that the number of transporters is affected, rather than the affinity of the carrier, by the stage of lactation. The expression of GLUT1 mRNA and protein fall within 24 h of litter removal in the rat (40). These observations suggest that the differences in D-glucose transport by mammary tissue during development, lactation, and involution reflect the expression and function of GLUT1 under the prevailing hormonal milieu of these physiological states.

Milk secretion in the rat is controlled by both prolactin and growth hormone. Fawcett et al. (66) investigated the effect of these hormones on the expression of GLUT1 protein in mammary secretory cell membranes isolated from lactating rats receiving various treatments for 48 h. Administration of bromocriptine reduced GLUT1 in mammary membranes, as would be expected. In contrast, GLUT1 levels in membranes from animals that had been treated with an anti-rat growth hormone serum were not significantly altered. However, when the two treatments were combined, GLUT1 expression was reduced to a level below that found with bromocriptine treatment alone. The results suggest that prolactin and growth hormone act synergistically to maintain GLUT1 transporter expression in rat mammary gland plasma membranes. A role for prolactin in the regulation of GLUT1 expression in cultured mouse mammary epithelial cells has also been reported (87). Prolactin stimulates 2-deoxyglucose uptake by mammary gland explants isolated from midpregnant mice (170). Although the target mechanism was not identified, it can be predicted, on the basis of experiments using rats, that the prolactin effect is on GLUT1.

Zhoa et al. (264) found that administration of bovine growth hormone releasing factor, but not bovine growth hormone itself, increased by 21% the amount of GLUT1 mRNA in the lactating bovine mammary gland. However, the physiological significance is unclear, since the increase in message was not accompanied by a rise in GLUT1 protein in mammary cell membranes.

The uptake of 2-deoxyglucose by the lactating rat mammary gland in vivo was found in one study to be enhanced by insulin (226). This observation is difficult to reconcile with the failure of insulin to produce acute stimulation of 3-O-methyl-D-glucose transport by isolated lactating mouse mammary epithelial cells (177). If insulin does have a direct effect on D-glucose transport by lactating mammary epithelial cells, then the target must be a transporter other than GLUT4 (see sect. XIIA). It is possible that insulin stimulates D-glucose uptake by the rat mammary gland as a result of altering the magnitude of the transmembrane D-glucose gradient rather than by affecting the transport proteins directly (34).

Overnight starvation of lactating rats reduces 2-deoxyglucose and 3-O-methyl-D-glucose uptake by the mammary gland by >90%. (226); the arteriovenous difference in D-glucose concentration across the mammary gland is decreased to a similar extent (156). However, refeeding starved animals rapidly restores the uptake of glucose analogs and increases the arteriovenous difference. It is apparent that the effect of overnight fasting is a consequence of a reduction of D-glucose transport via GLUT1. Thus starvation decreases cytochalasin B-sensitive 3-O-methyl-D-glucose uptake by isolated lactating mouse mammary epithelial cells. This reduction in transport occurs as a result of a decrease in the V_max of uptake and is accompanied by a decrease in the number of cytochalasin B binding sites on mammary cell plasma membranes (175). However, it has been reported that starvation does not alter the total GLUT1 protein content of lactating rat mammary tissue, suggesting that the reduction in D-glucose transport induced by starvation is brought about by the translocation of GLUT1 carriers from the plasma membrane to an intracellular site (40).

	XIII. AMINO ACID TRANSPORT

Top Previous Next References

The lactating mammary gland has a large demand for amino acids to meet the requirements of milk protein synthesis. For example, a dairy cow secreting 35 l milk/day (by no means exceptional) with a protein content of 3.3% needs over 1 kg/day of amino acids to sustain milk protein synthesis. The first studies on amino acid transport by the lactating mammary gland focused on arteriovenous differences in amino acid concentrations and combining these with measurement of mammary blood flow to establish whether amino acids were the blood source of milk proteins and then whether their quantitative uptake was sufficient to account for milk output of protein (129). Many studies in goats, cows, sheep, pigs, guinea pigs, and rats have shown that the arteriovenous differences across the gland are substantial, especially for some of the essential amino acids (82, 119, 129, 136, 240). This extraction of amino acids from blood is generated by amino acid transport across the basolateral membranes of the secretory cells. Despite the importance of amino acids to milk secretion and mammary metabolism, it was some time before the cellular mechanisms responsible for amino acid uptake by the mammary gland received the attention they deserved. Although our current understanding of mammary amino acid transport systems and their regulation lags behind what is known in other tissues, it is now becoming clear that the lactating mammary gland possesses amino acid transport systems analogous to those described in other organs such as the