INTRODUCTION

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

PHYSIOLOGICAL ANALYSIS OF embryonic epithelial membrane transporters is providing increasing data on gene messages and their specific proteins in defined embryonic cell populations. Although recent studies on the expression of membrane transporters in differentiated organs have been summarized in outstanding reviews (e.g., 26, 30, 37, 79), work of the past decade on transporters in embryonic epithelial organs and their cell types has not yet been reviewed critically. In particular, a comparative description of membrane transporter expression, in terms of temporal and spatial patterns, of the same transporter or transporter family in different embryonic epithelial organs has not been made. Organogenesis of epithelial systems, in general, is governed by basic processes that are similar in different epithelial organs, such as lung (38), kidney (40), pancreas (34, 69), and salivary gland. Among these processes are early induction of mesenchymal blastema and its transition to epithelium, branching morphogenesis of induced epithelial primordia, and actions of signal molecules, as illustrated for some of these events in renal organogenesis (Fig. 1). Because stages of morphogenesis are well defined, and morphogenic factors and mechanisms are being evaluated increasingly for different epithelial organs (e.g., 38, 40), the present article attempts to relate changes in membrane transporter expression to developmental stages. Although progress in the area of epithelial transport differentiation would appear to have been substantial since the publication of an early review¹ on renal transport, most of the key processes such as the role of established transcription factors (e.g., Pax-2, WT-1, HNF-1a, SP1, and SP3) in the temporospatial expression of embryonic membrane transporter patterns remain to be understood.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Scheme of general mechanisms and regulatory factors involved in morphogenesis of the kidney. Some of these, e.g., branching morphogenesis and mesenchyme-to-epithelium transition (MET), are common to various epithelial organs, whereas most regulatory factors are organ specific. Expression of plasma membrane transporters is associated with defined periods in morphogenesis, such as MET (47). Schemes of general mechanisms in organ morphogenesis, e.g., of lung (38), have been presented elsewhere. HGF, hepatoctye growth factor; GDNF, glial cell line-derived neurotrophic factor; ECM, extracellular matrix.

	AMILORIDE-SENSITIVE EPITHELIAL SODIUM CHANNEL

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

The highly selective amiloride-sensitive epithelial sodium channel (ENaC) mediates electrogenic sodium transport and is the rate-limiting step of sodium absorption in epithelia of the distal nephron and colon, airways, and salivary ducts. Although renal, colonic, and salivary ENaC-mediated sodium absorption maintains extracellular sodium homeostasis, pulmonary sodium transport is the primary driving force for lung liquid clearance at birth. Inherited ENaC mutations are inactivating (loss of function), as in pseudohypoaldosteronism type I (PHA-1), or activating (gain of function), as in Liddle's syndrome. Targeted disruption of subunit loci in the embryo has uncovered much of their functional roles in ENaC-mediated sodium transport.

Expression of ENaC subunits (12) has been measured specifically in embryonic and perinatal epithelia of lung (1, 24, 48, 61, 67, 71, 72, 74, 75, 87, 88, 98, 99, 101) and kidney (44, 62, 49, 97, 100) in humans (61, 98, 99), guinea pigs (24), rats (44 67, 71, 74, 75, 87, 97, 99), and mice (1, 48, 49, 62, 72, 87).

In embryonic rat lung (Fig. 2), -ENaC expression begins (67, 88) in the late canalicular and saccular stage [embryonic days (E)17/18-E19], concomitant with Na-ATPase (66), but not in the earlier pseudoglandular stage; it rises to the highest mRNA subunit abundance around term, with a simultaneous increase (75) of aquaporin-4 (AQP4; see Fig. 4), and it decreases from postnatal days (P)1-P7 (88). The - (87) and -subunit expressions, by contrast, are induced postnatally (P1-P7). Similar to this divergent temporal pattern is the effect of exogenous corticosteroids (88). Dexamethasone, given to mothers between E17 and E19, increased embryonic -subunit expression but not that of the - and -subunits within 8 h after injection, much like the effect of the glucocorticoid and thyroid-stimulating hormone together (67), whereas thyroid hormone triiodothyronine (T3) alone, or combined with dexamethasone, did not increase any of the subunits (88). Because the -ENaC subunit alone suffices to express part of the sodium channel activity (12), the sole rise of this subunit either in early postnatal or in embryonic lung, depending on the species (24, 61, 87), should be capable of driving initial alveolar fluid clearance. Indeed, the first report (48) on the -null phenotype [-ENaC (/)] indicated that lungs of these newborn mice, although histologically normal, showed patchy filling of the alveolar space, and the animals died within 40 h after the transition to air-breathing. By contrast, the -ENaC deletion (62), although associated with a lethal phenotype, did not result in the development of respiratory functional or morphological failure, suggesting that - and -subunits alone might generate sodium channel activity just appropriate for perinatal respiratory survival.

View larger version (77K): [in this window] [in a new window]   Fig. 2.   Expression patterns of mRNA abundance of amiloride-sensitive epithelial sodium channel (ENaC) subunits in kidney and lung. Temporal patterns of some transporters (Figs. 2-5) and their subunits are presented for rat embryonic and postnatal expression only, because 1) data in other species are less abundant and 2) comparison between species is limited by the fact that organ morphogenesis (e.g., nephrogenesis) in some species (mouse, rat) extends beyond embryonic life, whereas in others (sheep and human) it is completed before birth. Relative abundance of mRNA expression at different stages of morphogenesis, as delineated by embryonic (E) and postnatal (P) age, is represented by width of horizontal lines. Thick vertical line, transition from embryonic to postnatal epithelial cell differentiation; circles, ref. nos.

The sensitivity of pulmonary ENaC expression to oxygen content of the medium had been established earlier in cultured fetal rat lung epithelia (71), where, remarkably, the switch from 3 to 21% O₂ increased the abundance of -, -, and -ENaC subunits, implying that a similar rise on P1 (75) might be induced by the change in blood O₂ content after birth. Although the fetal expression change in response to O₂ (from 3 to 21%) was homogeneous in all subunits (71), the unstimulated pattern (87) showed early embryonic temporal patterns that were differential for the subunits. The expression of the -subunit follows the -ENaC pattern, but that of -ENaC rises gradually from P1 to adult (87). The embryonic spatial patterns are similar for - and -subunits, with a prevalence of the -subunit in early embryonic acinar and in small airway cells, whereas the -subunit localized mostly to small airway cells. Remarkably, - and -subunits in the adult appear to localize to the AII cell type (87). The O₂-induced ENaC subunit mRNA levels (71, 74) imply the induction of O₂-responsive elements (71) and, indeed, a change in intracellular superoxide expression was associated with the activation of a redox-sensitive transcription factor, nuclear factor-B (74), which was blocked, as was the O₂-induced increase in ENaC, by a superoxide scavenger.

Subunit temporal expression patterns were further examined (88, 101) in lung and colon and found to be differential both within the lung and particularly between these organs. Lung -abundance is consistently higher than that of - (88) or - (101), most impressively so from E19 to P2 and from P9 to P28 (Fig. 2). Remarkably, the mRNA amount of each subunit on P1 is almost as high as on P28. Thus the early expression specifically of the -ENaC subunit (Fig. 2), even before the need for ENaC protein function, is to ensure neonatal fluid removal (clearance), in concert with the early expression of AQP (see Fig. 5) from epithelial surfaces. Nevertheless, pulmonary differences in temporal expression of mRNA abundance (Fig. 2) suggest divergent and independent functions of the subunits that might also imply differential subunit O₂ sensitivities.

In human lung, ENaC mRNA appeared to rise continuously from fetal to adult expression level (99), but -ENaC message was found only at the end of the second trimester (61), whereas all subunits were first expressed in week 21 (98). The low and differential expressions of the subunits in week 24 cultured lung (98) were increased by a factor of two to three by dexamethasone, not by thyroid (T3) hormone (88, 98), mainly at the transcriptional level. These effects on ENaC mRNA abundances suggest that promoter regions of ENaC subunits might have glucocorticoid response element sequences, but they do not imply any functional (posttranslational) consequences.

Renal ENaC expression patterns (Fig. 2), by contrast, although established early in nephrogenesis mostly by principal cells (44), are increasingly responsible for postnatal sodium homeostasis; it is relevant in this context that the temporospatial pattern of nephron differentiation from deep to superficial cortex (40) puts the renal workload at birth, in several species, on the almost mature nephron generations in the deep cortex. In humans, however, nephrogenesis is completed at about weeks 34-35 of gestation. Total renal embryonic rat ENaC subunit mRNA abundance is very low but increasing from E16 to E19 (97), and it rises abruptly from E19 to E22 to almost adult levels of expression at P1-3 in all subunits (Fig. 2), indicating that the distal nephron does not have the divergent embryonic subunit mRNA expression pattern that characterizes pulmonary ENaC development.

A detailed study of -ENaC in the ureteric bud (UB) revealed that, surprisingly, this subunit and the sodium channel protein are expressed and functional (44) in rat UB (and embryonic collecting tubule) well before filtrate delivery, suggesting that the other subunits (100) also might be functional in this progenitor cell of the cortical collecting duct (CCD) cell populations. Moreover, the acquisition of apicobasal cell polarity, a hallmark of nephrogenesis (39) after mesenchyme-to-epithelium transition (45), also pertains to the sided expression of the ENaC functional protein (41, 44). The early expression (E17) was confirmed because each of the subunit antisense riboprobes hybridized to collecting ducts, whereas only - and -subunits were detected in the most distal uroepithelia (100). Although whole kidney -ENac mRNA abundance increased, and - and -subunit abundance decreased after birth, there was a striking shift of the collecting duct expression pattern from inner medullary to cortical localization (100).

The functional roles of the ENaC subunits in epithelial sodium transport were studied in several mouse organ systems and subunit-knockout models (1, 48, 49, 72). As mentioned for the -ENaC (/) phenotype, failure of pulmonary fluid clearance was the primary cause of early postnatal death although -ENaC transcripts could not be detected in both lung and kidney (48). In subsequent work, the -ENaC (/) with transgenic expression of -ENaC resulted in a partial restoration of -mediated sodium reabsorption specifically in lung, and liquid clearance was sufficient for the newborn to survive the early postnatal period; however, renal manifestations of reduced -ENaC expression were salt wasting and metabolic acidosis with elevated aldosterone, thus resembling a compensated PHA-1 syndrome. When the -subunit gene was disrupted (72), the mutant newborn mice (-ENaC m/m) still were able to clear their lung liquid despite a sixfold reduction in ENaC-mediated tracheal transport, but compensated metabolic acidosis and elevated aldosterone levels indicated a compensated PHA-1 phenotype on a normal-salt diet. On low-salt intake, however, the reduced -ENaC mutants revealed the PHA-1 phenotype with renal and colonic expression, such as elevated serum potassium (6 mM), lower blood pressure, and elevated aldosterone (72). Of interest, -ENaC (/) mice, which are hyponatremic and hyperkalemic with corresponding urinary excretion patterns and elevated aldosterone, display the collecting duct dysfunction phenotype of human PHA-1 without defects in pulmonary or renal morphology (62). Last, the -ENaC (/) resulted in a lethal PHA-1 phenotype (1), which, however, did not lead to respiratory failure during the first 24 h postnatally but to severe hyperkalemia of probably renal origin that appeared to be the leading cause of death in the -mutants (1). Colonic ENaC expression patterns were qualitatively similar for all subunits, the increase in their mRNA abundance being gradual, with -subunit mRNA having the highest value throughout (101).

In conclusion, expression patterns of ENaC subunits in kidney, lung (Fig. 2), and colon differ fundamentally, and regulatory mechanisms, e.g., O₂ and steroid hormones, appear to act in organ-specific ways in embryonic and postnatal differentiation.

	INWARDLY RECTIFYING K+ CHANNELS

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

Inwardly rectifying K⁺ (Kir) channels constitute diverse subfamilies, such as ATP-sensitive and ATP-dependent potassium channels, and they participate, among many diverse functions, in maintaining potassium homeostasis and resting conductance. For the Kir family, as for ENaC, distinct temporal expression patterns have been described, specifically for some Kir members in embryonic (Braun GS, Huber SM, Veh RW, Segerer S, and Horster MF, unpublished observations) and in postnatal (Braun GS, Huber SM, Veh RW, Segerer S, and Horster MF, unpublished observations; 77, 111) renal epithelial development. Thus ROMK, Kir6.1, and its associated sulfonylurea receptor SUR2, were found to be expressed in a development-dependent mode (Braun GS, Huber SM, Veh RW, Segerer S, and Horster MF, unpublished observations) in defined stages of collecting duct morphogenesis, in mesenchyme-to-epithelium transition, and in nephron formation, where they might serve specific developmental functions. It is of particular interest that Kir6.1 (protein) is faintly expressed in mesenchymal blastema, upregulated at the mRNA and protein level in condensed mesenchyme, and downregulated in the S-shaped body (47). In collecting duct ontogeny, Kir6.1 and SUR2 mRNAs are downregulated with time, whereas ROMK2 mRNA is upregulated by cAMP, suggesting that vasopressin might function as transcriptional regulator in development (Braun GS, Huber SM, Veh RW, Segerer S, and Horster MF, unpublished observations). The ROMK isoforms, when analyzed by immunofluorescence, are selectively expressed in the apical plasma membrane of mature thick ascending limb of Henle and of CCD principal cells (30, 37). By contrast, ROMK is not expressed in the nephrogenic zone or in midcortical CCD of postnatal week 1 rat kidney, whereas midcortical CCDs (week 3) were clearly stained (111). This postnatal change in ROMK protein abundance may be responsible for the increase in the number of open potassium channels, i.e., the number of functional channel proteins reported earlier in postnatal rabbit principal cells (77).

It may be concluded that embryonic ROMK, Kir6.1, and SUR2 expressions are involved in developmental processes such as cell cycle, mitosis, and mesenchyme conversion rather than in potassium homeostasis and that these latter functions are gradually acquired for the mature nephron (30).

	CHLORIDE TRANSPORTERS

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

Chloride (anion) transporters (channels) not only have a leading role in cell volume regulation, but they also participate in stabilizing potassium-mediated membrane potential and in acidifying intracellular compartments. Chloride transporter genes from distinct families that are developmentally expressed include the cystic fibrosis transmembrane conductance regulator (CFTR) (22, 34, 41, 59, 60, 63, 86, 90, 91, 110), ClC-2 (42, 43, 65), ICln (42, 102), the bumetanide-sensitive Na-K-Cl symporter (NKCC2) (51, 78, 89, 96), the Cl/HCO₃ antiporter (58), and the Cl/OH antiporter (80).

CFTR

In embryonic kidney (Fig. 3), by contrast, CFTR mRNA abundance in UB cells declined between gestational E15 and postnatal P1, rising transiently by a factor of 10 between P1 and P7 (41), a period corresponding to the late stage of rat kidney branching morphogenesis (40), and was downregulated thereafter.

View larger version (99K): [in this window] [in a new window]   Fig. 3.   Expression patterns of cystic fibrosis transmembrane conductance regulator (CFTR), ClC2, and ICln mRNA in kidney. Epithelial cell functional differentiation was followed from ureteric bud (E15) to cortical collecting duct (P14) during branching morphogenesis. Relative abundance of mRNA expression is represented by width of lines; circles, ref. nos.

Of interest in this context is the finding (41) that wild-type and truncated forms of CFTR mRNA have different patterns of developmental regulation in UB. The human renal CFTR expression pattern (22) is characterized by early appearance (weeks 12-19) of the multifunctional (79) CFTR protein in the apical plasma membrane of UB, but not in comma- or S-shaped body, a pattern similar to that in pancreas (34) and lung during branching morphogenesis.

To summarize, the early expression of CFTR, its transient decrease in the last gestational (rat) trimester (Fig. 3), its bronchial centrifugal expression gradient, and the shift from apolar to apical localization might suggest CFTR regulatory functional roles not only for other chloride channels (e.g., ClC-2, ICln) but also for ENaC, which is upregulated concomitantly (44).

ClC-2

In lung, ClC-2 mRNA is expressed in an almost identical temporal pattern, and downregulation of ClC-2 apparently is associated with the switch of embryonic respiratory epithelia from fluid secretion to reabsorption at birth, when -ENaC (88) and AQP1 (94, 107) are upregulated.

To conclude, the changing functional role (79) of the CFTR mRNA pattern in nephrogenesis, as mentioned, might be that of a molecular switch downregulating ClC-2-mediated chloride (and fluid) secretion and upregulating ENaC-mediated sodium (and fluid) absorption on initiation of glomerular filtration. The inversely related mRNA abundances of these transporters in different embryonic and perinatal stages appear to support this hypothesis.

Expression of ICln (Fig. 3), a further swelling-activated, volume-regulatory channel or channel regulator (102), was compared with that of ClC-2 during renal branching morphogenesis (42). Although ClC-2 mRNA abundance increased from E15 to E17, peaked around P3, and decreased from P3 to P7 (Fig. 3), similar to the temporal pattern in lung (65), mRNA expression of ICln increased during the entire embryonic and postnatal period of development (42), extending data in mouse kidney where a single transcript was found to be expressed at similar levels in embryonic stages between E7 and E17 (102). The pattern of ICln mRNA abundance, to summarize, follows that of increasing vectorial solute transport after P7 with its, by now, enhanced requirement for volume regulation (47), whereas the ClC-2 temporal expression pattern suggests specific embryonic functions in epithelia undergoing branching morphogenesis.

Embryonic epithelial cell volume regulation and its developmental course have been evaluated in UB cells and CCD cells in culture, and it was shown that the mature type of hypotonic swelling-activated chloride conductance may evolve already between E17 and P1 (47). Embryonic UB cells had a high and constitutively active chloride conductance, whereas perinatal and postnatal cells, by contrast, expressed a chloride conductance that was activated by hypotonic cell swelling (47), and the embryonic conductance type was downregulated simultaneously.

The chloride transporter NKCC2 is highly and early (E16.5, mouse) expressed in the metanephric kidney (51), and three isoforms have been identified in embryonic human kidney (96), where loss-of-function mutations can cause the antenatal (inherited) hypokalemic alkalosis of Bartter's syndrome (85). The first embryonic expression (E14.5, mouse) is localized to the distal limb of the immature loop of Henle, in cells close to future glomeruli, before the morphological differentiation of the loop into thin and thick ascending segments (51). These data were confirmed and extended to the postnatal nephrogenic (P3/P4, rat) zone and to deeper cortical areas (78). NKCC2, as in embryonic mouse, is first expressed in macula densa cells of the thick ascending limb, and it appears earlier than the sodium-phosphate cotransporter (NaPi) in the pars recta (78). These and other transporters are expressed embryologically even before completion of organ structure (e.g., of the post-S-shaped stages in the nephrogenic zone) and the onset of filtration or of ventilation, such as in embryonic (sheep) lung epithelia in culture, where NKCC2 may drive the chloride secretory pathway (89), suggesting that these molecules might have a morphoregulatory role. This function is by now also ascribed to polycystin, an integral membrane protein and polycystic kidney disease 1 (PKD-1) gene product, in human (29, 50) and in mouse (28, 31, 95) embryonic kidney, lung, salivary gland, and pancreas, where polycystin is highly expressed throughout major stages of organogenesis, such as mesenchymal condensation and tissue patterning. In the kidney, polycystin expression is very high in differentiating and very low in differentiated UB and CCD mouse epithelia, respectively (29). Transcript levels were detected in UB branches on E12 (31), peaked from E16 to E19, and were abruptly downregulated to very low adult levels in week 1/2. Indeed, the PKD-1 gene product polycystin was shown to regulate branching morphogenesis of mouse UB (95), and its mode of function may resemble that of a channel/receptor subunit. In human kidney, the PKD-1 protein is expressed in collecting duct and distal convoluted tubule on week 20 (50), and at a high level even earlier (week 14) in UB/CCD apical membrane but not in mesenchyme (29), with a general decrease after week 18 in kidney and pancreatic duct to undetectable levels in adult (29).

The chloride/anion antiporters, Cl/OH in proximal straight tubule (80) and Cl/HCO₃ in mesonephric collecting tubule (58), have distinct developmental expression patterns. In mesonephric kidney, which in rabbit is functional from E13 to E19 (in sheep from E15 to E55, and in humans from week 5 to week 16), the collecting duct expresses (in ~1/3 of the CD cells) the intercalated A-type cell of the later metanephric kidney, as indicated by apical H-ATPase and basolateral Cl/HCO₃ immunoexpression (58), suggesting that lineages of the metanephric collecting duct cell populations are already expressed in the mesonephros. In proximal straight tubule (rabbit), the functional activity of the Cl/OH antiporter was about sevenfold less in neonatal compared with adult S2 segments (80) and notably similar to the rate change in the Na/H antiporter (NHE3), indicating that developmental changes in proximal tubule NaCl transport are mediated by similar changes of these cooperating transporters (80). The heterogeneous families of chloride transporters, to summarize, appear to serve various undefined embryonic and postnatal functions, as suggested by the distinct temporal patterns of up- and downregulation in different morphogenic periods.

	SODIUM-GLUCOSE SYMPORTER-GLUCOSE UNIPORTER

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

Transcellular vectorial movement of glucose in epithelial cells, i.e., uptake and release, results from concerted actions of members of two families of regulated glucose transporters, the isoforms of sodium-glucose symporter (SGLT) in the apical and of glucose uniporter (GLUT) in the lateral plasma membrane. Their embryonic expression is tissue specific and regulated developmentally.

SGLT and GLUT have been evaluated in embryonic and postnatal stages of intestine (9, 14, 20, 56, 57, 73, 82, 92), kidney (4, 108), lung (35, 84), and pancreas (36, 69) in humans (20, 56), pigs (73), cats (9), rabbits (4), and rats (14, 35, 36, 57, 69, 82, 84, 92, 108).

SGLT-1, the high-affinity symporter in human intestine, is expressed first in week 17 of gestation (20), and it increases from a low embryonic expression (57) with development. Apical glucose (SGLT-1) and D-fructose uptake (rat), when studied from P1 to P42 (92), had a steep rate increase after P21 to adult, and it declined thereafter. Interestingly, the ratio of intestinal glucose uptake to that for galactose and for fructose changes with postembryonic ontogeny (73). Although the galactose-to-glucose and fructose-to-glucose transport ratios have a similar course between birth and P10, galactose-to-glucose ratio decreases sharply after P10, and that of fructose to glucose, as might be anticipated, increases further (73). Monosaccharide transporters, in general, are already acquired embryonically before substrate delivery to these transporters (73), similar to ENaC expression in kidney and lung before the onset of filtration and ventilation, respectively.

SGLT-2-mediated glucose transport in late embryonic renal brush-border membranes has the same affinity but a reduced maximal velocity (V_max) compared with that in adults, suggesting that the increase in SGLT-2 expression (4) may also represent an adaptive intestinal change of transporter availability in response to the gradual postnatal shift in diet (9). This latter factor appears to be relevant for the biphasic expression pattern of the ileal sodium-dependent bile acid (taurocholate) transporter ASBT mRNA (81). Both mRNA and protein abundance are high on E22, downregulated in P1-P7 relative to E22, and markedly increased at weaning (P21, rat). This pattern is similar to rBAT mRNA expression, a cystine-related absorption system in embryonic renal pars recta (25).

Expression patterns of organic ion transporters related to the organic cation transport family (OCT) were examined recently in murine development (69a). Although NKT/mOAT (55), Roct, OCT-1, and NLT transcript levels increase during embryonic development almost pari passu in proximal tubule, their mRNA expression in extrarenal sites decreases impressively, thus raising the possibility that OCTs might have a role in morphogenesis.

Renal SGLT-1 mRNA is expressed on E18 (rat), and it increases up to P1 (108). By contrast, SGLT-2 expression began on E17, rose up to E19, and declined between E19 and P1 when, notably, the size of the message changed from an embryonic (2.6-kb) to the postnatal (2.2-kb) form, suggesting developmental splice variants of the low-affinity symporter similar to the CFTR embryonic variant (41).

GLUT-5 in human small intestine is immunoexpressed as early as gestational week 15 in the apical membrane (56), whereas GLUT-2 mRNA has been detected already in week 11 (20). This differential temporal pattern of GLUT expression pertained to all isoforms; i.e., GLUT-5 in embryonic intestine was expressed at lower abundance than GLUT-2 and both increase from low embryonic expression with development (57), whereas GLUT-1 has an early and high embryonic expression (20) and decreases postnatally, as is true for pulmonary (35) GLUT-1. Immunoreactive localization in embryonic period E10-E20 confirmed that the GLUT-1 (high-affinity) transport protein is expressed earlier than -2 and -5, and it disappears in the newborn enterocyte, whereas the others are increasingly expressed. The progressive increase in GLUT-5 transcript level was correlated with the fructose uptake pattern (14, 73); i.e., GLUT-5 mRNA expression in P22 is upregulated by luminal fructose after completion of weaning, and dietary fructose levels induce GLUT-5 mRNA abundance specifically and instantaneously (82).

Pulmonary GLUT-1 mRNA, protein, and glucose uptake in embryonic AII cells (E20, rat) are upregulated during culture in glucose-free medium and decreased in high-glucose medium (84). Mechanisms regulating transcriptional rates in response to the transported substrate for GLUT-1 and -5 (82) are unknown. In embryonic lung, in addition, the estradiol-induced magnitude of change in GLUT-1 mRNA was similar to that of protein and of glucose uptake rate, suggesting a steroid-induced increase in membrane transporter density (35).

Pancreatic immunohistochemical GLUT-2 is absent in E20 -cells (36), and it is faintly expressed in P1, suggesting that glucose-induced insulin secretion is developmentally reduced at the transporter level. However, in E11 pancreas, GLUT-2 protein is expressed in endoderm before bud morphogenesis, but the protein- and insulin-positive cells in acini and in duct epithelia are lost by E19, indicating embryonic and postnatal lineages of insulin expression (69).

SGLT and GLUT, to conclude, apparently have distinct isoform expression patterns, which might be similar, however, in different organs for a given isoform, as in GLUT-1, and dependent on substrate load, as in GLUT-5.

	NHE

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

The renal proximal convoluted tubule (PCT) has a central role in acid-base homeostasis, and its major part of proton secretion is carried by NHE via the low intracellular sodium concentration established by the Na-K-ATPase. The NHE isoforms (1-4) are expressed developmentally.

Embryonic and early postnatal NHE expression has been evaluated by transport activity, by mRNA and protein abundance, or by immunolocalization, in kidney of sheep (33), rabbit (2, 5), rat (3, 6, 17, 109), and in rat jejunum (16, 17).

PCT embryonic apical membrane abundance of Na/H (NHE3) protein and of amiloride-insensitive sodium uptake into brush-border vesicles (E29-30, rabbit) is much lower compared with adult (5). Although V_max of the embryonic transporter was about threefold lower than in the adult, suggesting an increase in the number of functional transporters or in the turnover number of individual transporters, the Michaelis-Menten coefficient value (K_m) was similar throughout ontogeny. Previous work had reported transport rates, V_max, K_m, and amiloride sensitivity of NHE3 to be higher at P7 than at P21 or the adult stage (109), indicating an increased capacity for antiport at the early stage. As might be anticipated, maternal glucocorticoid application (5) induced V_max of the embryonic NHE3 antiport activity to rise precociously to an adult level, which might have been due to an increase in transcriptional efficiency. NHE3, when immunolocalized on P1, was not detected in vesicle or in early S-shaped body of the nephrogenic zone, and it was first expressed in the apical plasma membrane of the fully formed S-shaped body (6), whereas medullary thin and thick loops of Henle expressed the adult staining pattern already at this P1 stage. NHE3 mRNA abundance (rabbit) increased from very low on E30 and P1, only slightly in week 1, and sharply in week 3 to an adult level (2). In subsequent stages (rat: weeks 2, 3, 6, and adult), the pH-dependent renal sodium uptake (17) was higher in weeks 2-6 than in adult, and the abundance of NHE1 and NHE4 mRNA decreased similarly. By contrast, intestinal uptake increased slightly from week 2 to week 6 but steeply from there to adult, whereas intestinal NHE1 message remained constant (17). NHE3 expression is probably regulated at several levels, because the relative contribution of NHE3 and NHE2 to total transport activity changed with ontogeny. In early stages (weeks 2 and 3), NHE2 and NHE3 carried 40 and 60% of transport activity, respectively, whereas later (week 6) the contribution of NHE3 had increased to some 90% (16). Intestinal NHE3 mRNA and protein abundance were increased by corticosteroids on P18 about twofold (16).

Besides corticosteroids (5), thyroid hormones regulate ontogenic transcriptional activities, and these were studied in P17-20 (rat) with respect to NHE3 antiporter activity, mRNA, and protein abundance (3). Hypothyroid renal cortical NHE3 mRNA abundance was similar to that of euthyroids, but transporter activity and apical membrane protein abundance were lower. In hyperthyroids, by contrast, mRNA, protein, and transporter activity were all higher than in euthyroids (3), suggesting a regulatory role for thyroid hormones in PCT functional differentiation.

The glucocorticoid-induced embryonic expression (sheep) of NHE3 activity in PCT is similar to that during natural embryonic-postnatal transition (33); i.e., V_max increases and K_m is constant, suggesting an increase in transporter numbers under the regulatory control of corticosteroids. Remarkably, total renal sodium reabsorption was unchanged, which emphasizes the differential effects of the glucocorticoid on transport expression. NHE1 and NHE4 mRNA abundance (rat) in cortical basolateral plasma membranes decreased with postnatal ontogeny from week 2 to adult, concomitant with pH-dependent sodium uptake into basolateral vesicles (17), whereas NHE1 mRNA abundance (rabbit) was expressed at adult level already on P1 and did not change further (2) and protein profiles of both NHE3 and NHE1 were similar to those of their mRNAs. Glucocorticoid treatment in week 1 (rabbit) increased NHE3 mRNA (~2-fold) and protein (~3-fold) abundance but not that of the high-NHE1 expression level (2). To summarize, it appears that not only the contributions of NHE isoforms to pH and sodium homeostasis change with embryonic and, specifically, postnatal ontogeny but also their response to steroid-induced expression.

	NAPI

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

A systematic evaluation of the intracortical developmental (rat: P1, P13, P22, adult) profile of NaPi expression, by immunolocalization and in situ hybridization, revealed that NaPi protein and mRNA are expressed, as expected, only in those PCT of inner cortex, specifically in juxtamedullary PCT, that had acquired a brush border (93). Surprisingly, however, NaPi-2 was not expressed in any of the nephrogenic structures, i.e., vesicle, comma- and S-shaped body, and it appeared rather homogeneously throughout the cortex only on P13, i.e., when nephrogenesis (rat) is almost completed (40). The adult pattern (week 6) was reached after abundance in superficial PCT appeared downregulated compared with juxtamedullary PCT (93). NaPi transport on P21 in renal brush-border membrane vesicles of hypothyroid rat (23) was specifically and acutely restored by T3, in both K_m and V_max, and T3 also increased the abundance of NaPi-2 mRNA to control values, independently of corticosteroid effects (23).

	ATPASES

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

Na-K-ATPase is the key transporter in cellular solute and water homeostasis. With the increasing demand for transcellular solute and fluid movement in epithelia after the embryonic period of development, expression of transporters such as ENaC (88, 101), SGLT (57), and NHE (5) is upregulated and may contribute to the ubiquitous rise in the abundance of Na-K-ATPase.

Na-K-ATPase expression in embryonic lung (15, 19, 52, 66, 88) and kidney (10, 32, 70, 103) has been measured in mice (19), rats (52, 66, 103), guinea pigs (32), rabbits (15), sheep (70), and humans (10), whereas expressions of renal and gastric H-K-ATPase were studied postnatally in rabbit (18) and mouse (64), respectively.

Embryonic distal lung cells are to become air space epithelia, and their high Na-K-ATPase activity serves sodium absorption and thus fluid clearance after birth, whereas much of the ATPase abundance in embryonic airway (bronchiolar) cells is part of the chloride-secretory and thus fluid-secretory process that is required in early lung morphogenesis. In lung, epithelial cell Na-K-ATPase undergoes two major changes, i.e., in turnover rate and in the amount of enzyme in plasma membranes (15). At term (E30-P1), a fourfold increase in the ouabain-sensitive Rb uptake indicates a major change in the antiporter turnover rate, whereas postnatal development is characterized by a threefold increase of binding sites per cell, i.e., in the number of antiporters without an apparent change in the turnover rate (15). The embryonic Na-K-ATPase 1- and 1- subunit mRNAs (66) were found in cultured distal epithelia to be faintly expressed on E17, increased on E20, and declined on E20-21. Remarkably, however, the two subunits appeared to be regulated differentially because the E17 mRNA steady-state level was higher for the 1-subunit, but the increase in late canalicular/early saccular stage (E20) was higher for 1 than for 1 (66). Also, the major rise in Na-K-ATPase mRNA was at the transition from canalicular to saccular stage, whereas the enzyme activity increased at the saccular stage.

Pulmonary embryonic mRNA abundance profiles of these two isoforms of the enzyme increased coordinately from mouse E14.5 (where they are similar to adult values) to the highest levels (6- to 8-fold higher than adult) at P1, representing the terminal saccular stage (19), and mRNA abundance of both isoforms decreased from P3 to reach final adult values at P13. The spatial pattern, by in situ hybridization, is characterized by intense 1-isoform signals in proximal bronchial epithelia on P1, whereas 1 is expressed increasingly from E14.5 to P1 in both proximal and distal epithelia (19). It is of interest that maternal glucocorticoid treatment (52) changes the steady-state levels of the isoforms differentially. When dexamethasone was applied specifically from E14 to E16, both 1-subunit mRNA and protein increased on P17, but mRNA and protein of the 1-subunit remained unchanged (52).

Renal embryonic Na-K-ATPase activity rose in sheep (E132) after cortisol (48 h) to levels of P1 activity; this change, however, was much higher than that of mRNA and protein (70), suggesting that most of this short-term effect was posttranscriptional, i.e., not consequent to de novo synthesis of the 1-subunit, and thus similar to the effect in embryonic guinea pig kidney (32) and rat lung (88). By contrast, cortisol effects on the renal Na/H antiporter in embryonic sheep (E132) appear to be preferentially transcriptional (33). The unstimulated temporal pattern of Na-K-ATPase activity in rat juxtamedullary PCT shows an almost twofold increase from E21 to P1 (103) but much less enzyme activity in the induced mesenchyme. Detailed analysis of PCT enzyme activity (32) during embryonic-neonatal transition (guinea pig) revealed an intricate pattern, i.e., a constant K_m and a large increase in V_max and - (by a factor of 7) and -subunit (by a factor of 4) protein abundance but only a small increase (by factor of 1.5-2) in subunit mRNA abundance, indicating either posttranscriptional upregulation of Na-K-ATPase protein synthesis or lower degradation (32).

In the metanephros of humans (10), as well as in those of other species (39), the acquisition of basolateral membrane sidedness of the enzyme is developmentally regulated because both apical and basolateral localization of the -subunit is seen throughout a major part of gestation, and final polarization may be organized at the level of the 2-isoform expression involving the sorting pathway of the Na-K-ATPase. This important conclusion is based on the fact that 2-subunit mRNA and protein, but not 1-subunit protein, are expressed apically in distal nephron and collecting duct segments during early and midgestational metanephrogenesis. Postnatally, 2-subunit mRNA and protein are downregulated whereas the 1-subunit protein, by contrast, was upregulated (10).

H-K-ATPase protein in CCD intercalated cells is expressed early (weeks 1-2, rabbit) in apical plasma membrane (18), and thus ensures, together with the V-ATPase (18), acid excretion in the transitional period. Closer examination of the (mouse) gastric H-K-ATPase expression by membrane immunoassay revealed apparently similar control mechanisms for subunit expression, i.e., very low levels for both the - and -subunit protein from P1 to P15, with a sharp rise to adult level on P30, and the increase was mainly due to increased subunit mRNA levels (64). It appears justified to conclude that recent work has emphasized differential developmental roles of the Na-K-ATPase - and -subunits at the mRNA and protein level, in both the process of sided expression and the enzyme's upregulation under the control of steroids.

	AQP

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

AQPs are a family of highly conserved transmembrane channel proteins generally responsible for rapid osmotic water movements, and some nine members of this transporter family have been cloned in mammals to date. Their distribution patterns are organ and cell type specific, and their pulmonary and renal temporal developmental patterns differ fundamentally.

Expression of the aquaporins AQP1-4 in renal epithelia (7, 11, 21, 27 , 68, 76, 83, 104, 105, 106) and of AQP1, 4, and 5 in lung epithelia (13, 53, 54, 75, 94, 107) is acquired with embryonic and/or perinatal development in humans (21), sheep (11, 104), rabbits (13), rats (7, 53, 54, 75, 76, 83, 94, 105-107), and mice (27).

Temporospatial patterns are expressed differentially for each of the aquaporins in both kidney and lung (Fig. 4). In the kidney (rat), AQP1-4 are hardly detectable on E16 by immunohistochemistry (105). AQP1 mRNA rises sharply to a near-adult level in week 1 whereas AQP2 is expressed already on E18 and increases steadily up to week 4. This early AQP2 transcript appearance is accompanied by vasopressin receptor V₂R mRNA expression (in situ hybridization) already at E16 in embryonic collecting duct (68). At the protein level, the earliest expression (E18) of AQP2 is in UB apical membrane, whereas basolateral AQP3 is barely detected at this stage and highly expressed on P1 (105). Except for this embryonic divergence, basolateral AQP3 and apical AQP2 patterns appear to be similar. It is relevant in the context of CCD cell lineage expression that on P1 all of the cells were AQP2 positive, whereas on P11 principal cells (by exclusive AQP2 expression) and intercalated cells (by V-ATPase expression) were clearly distinct (76). Incidentally, some CCD cells on P11 expressed both AQP2 and V-ATPase in the apical plasma membrane, suggesting developmental transitional states between these cell types. AQP4 mRNA has a faint embryonic expression (Fig. 4), and it is continuously expressed after birth in kidney and lung up to P40 (75, 105).

View larger version (78K): [in this window] [in a new window]   Fig. 4.   Expression patterns of mRNA abundance of aquaporin (AQP) subunits in kidney and lung. Relative abundance of mRNA expression at different stages of morphogenesis, E and P, is represented by width of horizontal lines; thick vertical line, transition from E to P epithelial cell differentiation; circles, ref. nos.

Because epithelial nephrogenesis (39) in many species, including rat, extends from embryonic into early postnatal life (P1-14), expression patterns of several epithelial membrane transporters also change during this later developmental phase (40). The increasing expression of AQP2 at the mRNA and protein level between P10 and P40 is paralleled by a similar rise in urine osmolality (27, 106). AQP2 mRNA abundance rose almost twofold from P10 to P20 (Fig. 4), and its protein in the principal cell apical plasma membrane stained faintly from P2 to P6, but strongly from P6 to P40 (106).

Remarkably, an adequate challenge to the signal cascade (e.g., dehydration or arginine vasopressin) when applied in the early (P1) postnatal phase (7) increases both AQP2 expression and, to a higher extent, protein insertion from the intracellular pool into the apical membrane of CCD, specifically of the glycosylated form, whereas urine osmolality remained at the nonactivated level. These latter data imply that factors in addition to AQP2 availability, V₂R mRNA expression (68), and cAMP-mediated signal system capacity, as measured by developmental responses of osmotic water permeability in inner medullary collecting duct (83), contribute to the ontogeny of countercurrent function. AQP2, at any rate, is not a major limiting factor of early postnatal urinary osmotic concentration. Of particular interest, the pattern of glycosylation of the AQP2 protein switched from a preferentially nonglycosylated early form to the expression of both glycosylated and nonglycosylated forms of AQP2 in week 2 (7). When its expression was induced by glucocorticoid on P10 (106), AQP2 mRNA abundance rose by a factor of two and protein stain by a factor of eight, but there was no effect on AQP2 in adult kidney.

In human kidney, AQP1 and AQP2 are expressed as early as gestational week 12, when they appear, as shown by antibody localization, in the apical membrane of PCT and of UB, respectively (21). The pattern of glycosylation from embryonic to postnatal changed in AQP1, similar to that in rat AQP2 (7), such that the nonglycosylated protein was present throughout ontogeny, but the glycosylated subunits of AQP1 were detected only after birth, and AQP-1 in mature PCT and descending thin limb has been localized to the apical and basolateral membrane (21). AQP2, by contrast, is expressed only in UB-derived structures, beginning at gestational week 12 (21). Remarkably, V-ATPase was not detected at any embryonic stage in UB-derived cells (21), which further characterizes the collecting duct progenitor cell as a principal cell-like type (43, 46).

AQP1 mRNA is not expressed in mesonephros (E15-60, sheep); it appears in the metanephric PCT at a low level on E40 (11), and transcript abundance rises some sevenfold in the last trimester of gestation to reach adult levels in week 6 (104). AQP1 mRNA expression, similar to AQP2 in rat (106), increased by 43% on midgestational E74 (but not earlier) after glucocorticoid administration (104).

Lung water channel transcript expression, when analyzed systematically from E19 to P21 in rat (94), reveals distinct patterns for AQP1, 4, and 5 (Fig. 4) whereas AQP3 is not expressed in distal lung (54). AQP1 transcript expression, more than that of AQP4, increased simultaneously with osmotic water permeability immediately after birth in rabbit (13). In embryonic rat lung, by contrast, AQP1 protein is the first to appear (54), probably similar to in sheep lung bud (11), and its level rises severalfold from E20 to P1 (53) or to P5 (94), concomitant with the transition from secretion to absorption of lung liquid (53). AQP1 transcript expression had a sharp rise to its maximal transcript value at P2 (94) or P1 (107), and it rose severalfold after maternal treatment with glucocorticoid not only embryonically but also in P1 and even in adult lung (53). The AQP1 ontogenic expression pattern in embryonic lung, and the AQP2 pattern in kidney, appear to parallel those of ENaC (Fig. 5); however, AQP1 is not expressed in epithelia of large airways and of alveoli. AQP4 mRNA has an expression pattern similar to AQP1 (107), beginning on E20 with a sharp rise by a factor of eight to P1/3 (75), and it declines from there by a factor of two during week 1 to almost the adult level (75). This unstimulated pattern of AQP4 is lung specific, and the protein localizes to postnatal basolateral membranes of all airway cells but not to alveolar membranes. It is very relevant functionally that AQP4 responds to a change in O₂ from 3 to 21% (75) in a manner similar to the change in ENaC expression, and maximal coexpression of ENaC subunits and AQP4 occurs on P1 (75). AQP4 mRNA increases by some 50% after -adrenergic agonists and glucocorticoids (107). The AQP5 transcript, by contrast, has a slow and steady increase from E20 to P5 (94, 107), its maximum expression is after week 1, and there is no change after glucocorticoid treatment (54).

View larger version (93K): [in this window] [in a new window]   Fig. 5.   Synopsis of expression patterns of principal AQPs and of -ENaC in kidney and lung. Relative abundance of mRNA expression at different stages of morphogenesis, E and P, is represented by width of horizontal lines; thick vertical line, transition from E to P epithelial cell differentiation; circles, ref. nos.

In sum, developmental physiology of AQPs has been proven to be far more intricate than anticipated from work in other transporter families, because fluid transport functions most likely involve multiple AQPs acting in concert to maintain the delicate fluid balance not only on respiratory and secretory surfaces but also within the intracellular space during developmental transitional states.

	PERSPECTIVES

TOP ABSTRACT INTRODUCTION AMILORIDE-SENSITIVE EPITHELIAL... INWARDLY RECTIFYING K+ CHANNELS CHLORIDE TRANSPORTERS SODIUM-GLUCOSE SYMPORTER-... NHE NAPI ATPASES AQP PERSPECTIVES REFERENCES

Embryonic temporospatial expression patterns of some plasma membrane transporters have been described comparatively between epithelial organs, and they have been related to morphogenetic stages and evolving organ function. The mere description of embryonic expression patterns at the mRNA and immunoprotein level, however, can only provide a basis on which to evaluate causal relations in morphogenesis, cell polarization, and differentiation. Among the presently most interesting approaches to uncover causal links between regulatory entities are those that measure transporter and transcription factor patterns in defined transitional states (47) and interfere with those gene expressions in lung and kidney (e.g., 48, 62), the major embryonic transporter organs. However, we do not wish to leave the impression that this sole approach will enable us to draw a coherent picture of how embryonic epithelial membrane transporter expression is regulated.