Physiol Rev Journal of Applied Physiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Rev. 78: 1165-1191, 1998;
0031-9333/98 $15.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by SULLIVAN, L. P.
Right arrow Articles by GRANTHAM, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by SULLIVAN, L. P.
Right arrow Articles by GRANTHAM, J. J.

PHYSIOLOGICAL REVIEWS   Vol. 78 No. 4 October 1998, pp. 1165-1191
Copyright ©1998 by the American Physiological Society

Epithelial Transport in Polycystic Kidney Disease

LAWRENCE P. SULLIVAN, DARREN P. WALLACE, AND JARED J. GRANTHAM

Departments of Molecular and Integrative Physiology, Medicine, and Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City, Kansas

I. INTRODUCTION
II. GENETIC FACTORS
III. MORPHOLOGY
IV. ANIMAL AND CELL MODELS
V. ABNORMALITIES IN FUNCTION IN THE POLYCYSTIC KIDNEY
VI. CYST FLUID
    A. Composition
    B. Exchange With the Circulation
VII. CYST GROWTH
    A. In Vivo
    B. Cyst Formation by Cultured Cells
VIII. TRANSPORT MECHANISMS INVOLVED IN FLUID SECRETION
    A. Madin-Darby Canine Kidney Cell Model
    B. Autosomal Dominant Polycystic Kidney Disease Tissue
IX. ABSORPTION BY THE CYSTIC EPITHELIUM
X. DILEMMA POSED BY THE GRADIENT CYST
XI. HEPATIC CYSTS IN AUTOSOMAL DOMINANT POLYCYSTIC KIDNEY DISEASE
XII. A CYST ACTIVATING FACTOR PRESENT IN CYST FLUID
XIII. QUESTIONS REMAINING
XIV. COMPARISON OF THE CYSTIC CELL TO OTHER EPITHELIAL CELLS
    A. Renal Epithelial Cells
    B. Intestinal Crypt and Lung Cells
XV. CONCLUSIONS
REFERENCES

    ABSTRACT
Top
References

Sullivan, Lawrence P., Darren P. Wallace, and Jared J. Grantham. Epithelial Transport in Polycystic Kidney Disease. Physiol. Rev. 78: 1165-1191, 1998. --- In autosomal dominant polycystic kidney disease (ADPKD), the genetic defect results in the slow growth of a multitude of epithelial cysts within the renal parenchyma. Cysts originate within the glomeruli and all tubular structures, and their growth is the result of proliferation of incompletely differentiated epithelial cells and the accumulation of fluid within the cysts. The majority of cysts disconnect from tubular structures as they grow but still accumulate fluid within the lumen. The fluid accumulation is the result of secretion of fluid driven by active transepithelial Cl- secretion. Proliferation of the cells and fluid secretion are activated by agonists of the cAMP signaling pathway. The transport mechanisms involved include the cystic fibrosis transmembrane conductance regulator (CFTR) present in the apical membrane of the cystic cells and a bumetanide-sensitive transporter located in the basolateral membrane. A lipid factor, called cyst activating factor, has been found in the cystic fluid. Cyst activating factor stimulates cAMP production, proliferation, and fluid secretion by cultured renal epithelial cells and also is a chemotactic agent. Cysts also appear in the intrahepatic biliary tree in ADPKD. Normal ductal cells secrete Cl- and HCO-3. The cystic ductal cell also secretes Cl-, but HCO-3 secretion is diminished, probably as the result of a lower population of Cl-/HCO-3 exchangers in the apical membrane as compared with the normal cells. Some segments of the normal renal tubule are also capable of utilizing CFTR to secrete Cl-, particularly the inner medullary collecting duct. The ability of Madin-Darby canine kidney cells and normal human kidney cortex cells to form cysts in culture and to secrete fluid and the functional similarities between these incompletely differentiated, proliferative cells and developing cells in the intestinal crypt and in the fetal lung have led us to suggest that Cl- and fluid secretion may be a common property of at least some renal epithelial cells in an intermediate stage of development. The genetic defect in ADPKD may not directly affect membrane transport mechanisms but rather may arrest the development of certain renal epithelial cells in an incompletely differentiated, proliferative stage.

    I. INTRODUCTION
Top
Next
References

Autosomal dominant polycystic kidney disease (ADPKD) is characterized by the appearance and slow growth of fluid-filled cysts within the parenchyma of the kidney. The cysts begin as focal microscopic enlargements or diverticula at almost any point along the nephron (7, 103, 120). As the cysts enlarge, they begin to distort the renal parenchyma. Blood vessels and functioning nephrons are pushed aside, and remodeling of the extracellular matrix occurs. Over a period of decades, the growth of the cysts enlarges the kidney to four to six times its normal size, but the number of functioning nephrons declines and renal failure progresses. In the end-stage kidney, most of the renal parenchyma has been replaced by cysts and fibrotic tissue.

The course of the disease varies widely. It usually becomes clinically apparent in the third to seventh decade of life, although it has been found in the fetus and in children on occasion. Approximately one-half of the patient population will progress to chronic renal failure and require replacement therapy. Other abnormalities appear in some patients. These include vascular aneurysms and the appearance of cysts in the biliary tracts within the liver. Autosomal dominant polycystic kidney disease is one of the most common, potentially lethal, genetic diseases known, occurring with a frequency of 1 in 1,000 births worldwide (52, 56). There is a recessive form of the disease (ARPKD) that is usually lethal within the first year of life and an acquired form that is often seen in patients undergoing chronic dialysis treatment. There are other diseases that produce cysts in the kidney, and this has made it difficult to interpret the literature written before the inheritable forms were fully recognized and classified (12). Autosomal dominant polycystic kidney disease is emphasized in this review.

In the autosomal dominant form of the disease, the genetic mutations do not appear to interfere with the normal development of the kidney. Glomerular function usually appears to be normal well into adult life, and changes in tubular function are relatively slight. However, in a minority of developed nephrons, cysts begin to form as a single tubular cell, or a small, localized group of cells begins to proliferate. The initiating factor is unknown. It has been postulated that in the process of repair of a localized injury, the replacement of damaged cells goes awry (67). The regulation of proliferation and differentiation of new cells is apparently abnormal, and proliferation of relatively undifferentiated cells proceeds indefinitely. This can occur at any point from Bowman's capsule to the terminal collecting duct. The proliferating cells initially may cause a bulge in the tubule or may form an out pocket or diverticulum in the wall of the tubule. As the cysts develop and fill with fluid, the adjoining extracellular matrix undergoes remodeling. The basement membrane may thicken or become laminated, and inflammation and fibrogenesis occur in the later stages of the disease.

Studies of the function of the polycystic kidney have uncovered abnormalities in the ability to osmotically concentrate urine, to respond to an acid challenge, and to correct extracellular fluid volume expansion. The cause of these abnormalities has not been pursued extensively. However, the transport functions of cystic cells and the mechanisms of fluid accumulation within the cysts have been investigated in detail because it is considered that cyst growth is a major factor in the progress to renal failure. Describing the results of these investigations is the object of this review. Research into this complex disease has raised questions about the function and structure of genes and the proteins for which they are responsible; about tubular morphology; cellular proliferation and differentiation; formation of basement membrane and the interstitial matrix; inflammation, hormonal, paracrine, and autocrine factors; and signal transduction mechanisms. All of these questions may have bearing on the abnormalities of epithelial transport, but a review of all of these subjects is impractical. We have attempted to limit the review to a short description of the new genetic findings, which do not yet indicate a direct link to transport mechanisms, and to those investigations that directly bear on the function of the cystic epithelium.

    II. GENETIC FACTORS
Top
Previous
Next
References

Progress in understanding the genetics of ADPKD has been remarkable in the last decade. Mutations of three different genes can produce forms of the disease. The first gene PKD1 is located on chromosome 16 (16p13.3), and mutations of it account for ~85% of the cases of ADPKD (143). It contains 52 kb of DNA that encode a 14-kb mRNA transcript and a protein, polycystin I, a molecule weighing 460 kDa and consisting of ~4,300 amino acids. The protein contains several putative membrane-spanning regions and extracellular domains that are analogous to those found in cell-to-cell or cell-to-matrix proteins. These include four fibronectin-related domains, leucine-rich repeats, and C-type lectins (1, 23, 47, 87, 91). Polycystin I is highly expressed in the developing kidney; it is expressed in low amounts or not at all in the adult kidney, but some reports indicate that it is overexpressed in cyst epithelial cells. Attempts at localizing it within renal cells has produced variable results (65, 78, 90, 130, 174, 175).

The PKD2 gene is located on chromosome 4 (4q23-p23). It also encodes a large protein, polycystin II, of ~110 kDa with a section that is homologous to PKD1 (99, 121). Mutations of this gene account for 10-15% of patients with ADPKD. It is hypothesized that the COOH-terminal tails of polycystin I and II interact, but colocalization of the two proteins within cells has not yet been demonstrated (141, 172). The onset of end-stage renal disease usually occurs later in the life of patients with the PKD2 genotype (56, 131). Mutations of a third gene, whose structure has not been determined, account for a small percentage of the patient population (39). Most of the mutations that have been found to occur in the PKD1 and PKD2 genes appear to inactivate the genes (24).

In ADPKD, every kidney cell contains a mutated allele, yet cysts appear in only a small fraction of nephrons. It is postulated that random somatic mutations of the wild-type allele occur in some cells, and this results in clonal growth of cells. Evidence for this loss of heterozygosity in cells forming the cysts has been obtained in patients with the PKD1 genotype. This "second hit" hypothesis may also account for the highly variable course of the disease among family members with the same germ-line mutation (17, 24, 140). The little information that is available on the genesis of the cysts strongly suggests that the polycystin proteins play a role in proliferation and differentiation of renal epithelial cells. The ongoing studies of the function of these proteins are likely to provide major insights into these processes and the development of the nephrons. There is as yet no evidence of a direct link of these proteins to transport mechanisms.

    III. MORPHOLOGY
Top
Previous
Next
References

In examining the literature on the morphology of the ADPKD kidney, the major questions considered were as follows: 1) At what sites along the tubule do cysts appear? 2) Does the appearance of the cells lining the cysts offer clues on the site of origin of the cyst or on the process of cyst development? 3) Are the cysts connected to functioning tubules or glomeruli that provide the fluid contained within the cyst?

Early anatomic studies of the cystic kidney did not produce consistent results. Some did not fully distinguish among the various types of diseases that cause cysts to appear in the kidney (138). Later studies of adult polycystic kidneys, from patients whose family history indicated an hereditary connection, dealt mostly with the end-stage kidney in which the parenchyma is grossly altered by the cysts and is extensively fibrotic (86, 103). Using Peter's reconstruction method (137), Lambert (103) reported that cysts were found in Bowman's capsule, in all sections of the nephron, and in the collecting tubules. None of the examined glomerular cysts was connected to patent nephrons, and most contained glomerular capillary tufts that were slightly reduced in size but appeared to be functional. A minority of tubular cysts existed as termination of nephrons with normal glomeruli; the rest were connected to patent tubules. Most of the cysts found in collecting tubules were also connected to patent tubules. Lambert's use of Peter's graphical reconstruction method may have led him to concentrate primarily on quite small cysts. Using a microdissection technique, Heggo (86) also found that cysts appeared in the glomeruli and all sections of the tubules. They were most numerous in the glomeruli, in the angle of the loop of Henle, and especially in the collecting tubules. Heggo noted that cysts frequently occurred at the bifurcations of the collecting tubule and that abnormalities of collecting tubule branching occurred. No mention was made of the number of cysts with connections to patent tubules. Baert (7) was able to obtain kidneys from two patients that were in the early stage of disease. Glomerular function was normal in both cases. Approximately 100 nephrons and 20 collecting tubules were microdissected from each kidney. Cysts were randomly distributed throughout the glomeruli and the tubules; all the cysts in these early stage kidneys were connected to tubules. No abnormal branching of the collecting tubules was noted. It should be pointed out that the site of location of cysts that have lost tubular connections cannot be identified by microdissection, and these cysts would probably be ignored in the process of dissection.

Transmission electron microscopy studies of cyst walls indicated a reduction or disappearance of microvilli and reduced basolateral infoldings and intracellular organelles. Polyps were frequently present in the examined cysts, and the tubular basement membrane was thickened and, in many cases, laminated. Interstitial fibrosis was often noted (13, 35, 49, 55). Scanning electron microscopy investigations provided much more detail (34, 48, 50, 70). The appearance of the cystic epithelium varied widely among cysts from the same kidney; the cells of some cysts bore a close resemblance to normal tubular or glomerular epithelial cells, whereas most did not (Fig. 1). A few exhibited distinct signs of hyperplasia in the form of micropolyps, adenomas, or linear cordlike projections (13, 50, 70). In a study of 387 cysts in ADPKD kidneys from 10 patients (70), the epithelium of the large majority of cysts was not typical of that of a normal tubular segment (Table 1). The appearance of the apical surface suggested that dedifferentiation had occurred (Fig. 1B). The epithelium of a small minority of cysts resembled that of various tubular segments. About 5% of the cysts were distinctly hyperplastic. All of these various types of cystic epithelium were found in at least four of the kidneys examined, with the exception of the cysts in which the epithelium resembled that of the proximal tubule. A detailed analysis of cell size and number and cyst size clearly indicated that proliferation of cells occurred as the cysts expanded; the cysts were not caused by a simple ballooning of the tubule. Because a few cysts contained cells of more than one phenotype, it cannot be unequivocally concluded that proliferation occurred by clonal growth from a single cell. Examination of a single half of 259 bisected cysts found tubular openings in only 25. An additional examination of both halves of 11 bisected cysts uncovered tubular openings in 3. Thus fluid can accumulate within the majority of the cysts only by secretion across the cystic epithelium (70).


View larger version (132K):
[in this window]
[in a new window]
  		FIG. 1.   A: scanning electron microscopy of a hemisection of a cyst found in a kidney from a patient with autosomal dominant polycystic kidney disease. Surrounding tubular structures appear to be compressed by cyst. B: a higher magnification of inner surface of cyst. Cyst is lined with a flattened, simple, poorly differentiated epithelial cell layer.

 
View this table:
[in this window] [in a new window]
  		TABLE 1.   Morphological characteristics of autosomal dominant polycystic kidney disease cysts

All of these morphological studies have focused on the cysts. There do not appear to be any detailed studies of cellular architecture in noncystic segments of the tubules in ADPKD kidneys, although all the tubular cells presumably carry the mutated gene. This information will be difficult to obtain, since the studies must be performed on renal tissue removed from patients that have not reached the end stage of renal failure.

In conclusion, cysts may develop at any point from Bowman's capsule to the terminal collecting duct. Only a small minority of the tubules are involved. Cyst growth is accompanied by proliferation of cells that are usually poorly differentiated. The large majority of the cysts are lined by cells not resembling any normal tubular phenotype. At some point in their development, the majority of cysts lose their tubular connection. Accumulation of fluid within these cysts after that point must occur by secretion across the cystic epithelium.

    IV. ANIMAL AND CELL MODELS
Top
Previous
Next
References

Several animal models of polycystic disease have been discovered or developed, and these have been extensively reviewed (33, 64, 79, 119). The murine models especially have been used extensively to study the genetic and environmental factors that modify disease progression. However, very little information on function is available in these models. The Han:SPRD rat has been a particularly useful model. Discovered in Hanover, Germany in 1989, the spontaneous mutation in the Sprague-Dawley strain resulted in a slowly progressive, autosomal dominant disease with many of the characteristics of the human disease (34, 64, 95, 96, 146). However, the results of recent studies indicate that the mutation is not in the PKD1 or PKD2 genes (14, 122). The homozygous (cy/cy) rat develops massively enlarged cystic kidneys, profound azotemia, and dies at about 3 wk of age. There have been no reports as yet of patients that were known to be homozygous for mutations of PKD1 or PKD2; however, ADPKD can be severe in the fetus and in young children, and the homozygous fetus may not survive in utero. Cysts in the kidney of the heterozygote rat (cy/+) enlarge over many months. The disease is more severe, and life expectancy is less in the male rat than in the female, a characteristic also of the human form of the disease (52).

Transgenic models of polycystic kidney disease have also been developed. Renal cysts appear in transgenic mice developed with the use of simian virus 40 (SV40)-early region with large-T antigen and in mice made transgenic by using a c-myc oncogene driven by an SV40 enhancer and a human beta -globin promoter (64). Cysts can also be chemically induced in vivo by administration of the antioxidants nordihydroguaiaretic acid, diphenylamine, and diphenylthiazole. Glucocorticoids will induce cysts in newborn animals (33, 64, 119). An in vitro organ culture model has been developed (5, 119). Mouse kidneys removed from the fetus at 2 wk survive for extended periods and progress through certain stages of nephrogenesis. The addition of corticosteroids or thyroxine to the organ culture medium reversibly promotes the development of cysts.

Cultured cell models of renal cysts have been developed and studied intensively with the objective of understanding the mechanisms involved in cyst growth and in the accumulation of fluid within the cysts. These include a subtype of Madin-Darby canine kidney (MDCK) cells (64, 110, 111, 118), primary cultures of human kidney cortex (HKC) cells (123), and primary cultures of ADPKD cyst cells (75, 113, 159). This began after McAteer et al. (118) discovered that fluid-filled cysts will develop by clonal growth when wild-type MDCK cells are seeded within a collagen matrix. Isolation of a single cyst and culture of the cells forming that cyst resulted in the development of a subtype of MDCK cells that reproducibly and efficiently formed cysts (73). Techniques were developed for measuring fluid transport into the cultured cysts and changes in cell volume (165). Microelectrode techniques were applied (108), and epifluorimetric methods were adapted for measuring cellular fluid composition (158). A method of measuring fluid transport by monolayers of these cells grown on a permeable support was developed (73, 124), and Ussing chamber methods have been used to study the transport properties of the cells (75, 113).

    V. ABNORMALITIES IN FUNCTION IN THE POLYCYSTIC KIDNEY
Top
Previous
Next
References

Glomerular filtration rate (GFR) in the polycystic kidney disease patient may remain in the normal range long after cysts are discernable by radiologic or sonigraphic techniques. Clearance of p-aminohippurate ( PAH ) and GFR may also remain in the normal range after development of hypertension, but the filtration fraction may be elevated (170). However, the polycystic kidney disease patient with a normal GFR commonly exhibits a reduced ability to maximally concentrate the urine (30, 37, 58, 116, 139). In one comprehensive study, 177 nonazotemic subjects with ADPKD and 123 unaffected family members were subjected to 12 h of water deprivation and an injection of vasopressin (58). Maximum urine osmotic concentration averaged 680 ± 14 mosmol/kgH2O in the first group and 812 ± 13 mosmol/kgH2O in the second group. Examination of the cause of defective concentrating ability uncovered a variable capacity to increase free water reabsorption under conditions of increasing osmolar clearance by ADPKD patients with normal filtration rates (37, 116). However, the ability to increase free water clearance with increasing delivery of filtrate to the distal nephron was not affected in one study (116). These results suggest that salt reabsorption by the thick ascending limb is not impaired but that water reabsorption from the collecting tubule may be defective. These results could be attributed to an effect of medullary cyst growth on the architecture of the medulla, resulting in altered function of the countercurrent mechanism, or to an effect of cortical cysts on medullary blood flow rate or the delivery of tubular fluid to long loops of Henle. It is also possible that the cellular response to vasopressin is defective.

Examination of the expression of water channels in the ADPKD kidney has not disclosed an absence or mislocalization of aquaporin (AQP)-2, the water channel regulated by vasopressin (6, 42, 85) or of AQP-3, the water channel found in the basolateral membrane of principal cells in the collecting tubule (85). However, Western immunoblot analysis showed a decline in AQP-2 content in early stage ADPKD kidneys and abnormalities in glycosylation patterns (42). More research is necessary to determine if this is the cause of the abnormality in concentrating ability. In particular, the collecting tubules rather than the cysts should be the focus for examination.

A concentrating defect has not yet been documented in the Han:SPRD rat. However, proton NMR studies of renal tissue from normal and heterozygous rats have indicated reduced concentrations of the osmolytes betaine, taurine, and glycerophosphocholine but not inositol in the heterozygous animal. These osmolytes normally accumulate in medullary cells when the environment is hypertonic. The osmotic concentration of the urine in the two groups did not differ; however, these rats did not undergo water deprivation and vasopressin injection (128).

A concentrating defect was also found to be present in the diphenylthiazole-treated rat model in the presence of normal GFR (27, 28). The ability to dilute the urine was not impaired. Micropuncture experiments indicated no change in proximal or distal tubular fluid-to-plasma ratios of inulin concentration or in the distal tubular fluid-to-plasma osmotic concentration ratio in rats treated for 4 days with diphenylthiazole. However, the Na+ concentrations in medullary and papillary tissue slices were reduced, and microscopic changes in the morphology of collecting tubular cells occurred. It was concluded that the defect in concentrating ability was because of impaired water transport by the collecting tubular epithelium (28). A later study indicated that the rise in medullary adenylate cyclase activity, caused by administration of vasopressin, was reduced in diphenylthiazole-treated rats (43). The relevance of these studies to the ADPKD concentrating defect is not known.

The natriuretic response to volume expansion in ADPKD patients has been reported to be blunted (37) or exaggerated (38). In a study comparing ADPKD patients with normal GFR to normal control subjects, Torres et al. (170) found that baseline fractional excretion of Na+ was higher in the patients, although absolute rates of excretion were similar between the two groups. The filtration fraction was also higher in the patients in the baseline period. During saline infusion, fractional Na+ excretion also increased to a greater extent in the patients. The curve relating natriuresis to arterial pressure was shifted to the right in the hypertensive ADPKD patients but not in the normotensives. Plasma renin levels tended to be higher and aldosterone levels lower in the hypertensive patients before expansion, and both levels fell in response to volume expansion. Atrial natriuretic factor (ANF) levels did not differ from control subjects (170). A similar shift in the pressure-natriuresis curve was reported in ADPKD patients with chronic expansion (147). The question arises, Is the alteration in the reabsorption of Na+ due to an intrinsic alteration in tubular transport mechanisms or to an alteration in control mechanisms?

Hypertension is common among patients with ADPKD, and its development usually precedes reduction in renal function. In a comparison of hypertensive to normotensive ADPKD patients with normal GFR, Bell et al. (10) found no differences in plasma volume, a greater increase in cardiac index with exercise, increased renal vascular resistance, and higher ANF levels in the hypertensive patients. Renin and aldosterone levels were not different. However, use of an angiotensin-converting enzyme (ACE) inhibitor during a high-sodium diet increased renin levels only in the hypertensive patients. Torres et al. (169) performed an additional comparison of the response of hypertensive ADPKD patients and normotensive control subjects to volume expansion. The results confirmed that hypertensive patients with ADPKD have lower renal plasma flow, higher renal vascular resistance and filtration fraction, and an altered pressure-natriuresis relationship during volume expansion. Administration of an ACE inhibitor increased renin secretion and renal plasma flow, reduced renal vascular resistance and filtration fraction, and restored the pressure-natriuresis relationship to normal. In a study that compared the effects of ACE inhibition on hypertensive ADPKD patients and on patients with essential hypertension (29), the former exhibited higher plasma renin and aldosterone levels and responded to ACE inhibition with a greater increase in renin secretion. Long-term ACE inhibition also increased renal plasma flow and decreased filtration fraction in the ADPKD patients.

Renal angiography has indicated that peripheral renal vessels are attenuated in ADPKD (32, 46). An immunocytochemistry study of ADPKD kidneys demonstrated an increase in renin-containing cells in the juxtaglomerular apparatus, persistence of renin-containing cells in atrophied glomeruli, and an increase of these cells in proximal afferent arterioles and intralobular arteries (66). Thin, attenuated arterioles in the wall of cysts often contained these cells, and they were also found in connective tissue separate from the arterial tree. Thus it is reasonable to suggest that intrarenal levels of renin are increased despite the fact that circulating levels of renin may not differ between normotensive and hypertensive ADPKD patients (28). Therefore, it seems likely that alterations in Na+ excretion in ADPKD are the result of changes in the control mechanisms and that a defect in tubular Na+ transport mechanisms may not be present.

Preuss et al. (139) reported that two of four ADPKD patients given an acid challenge could not reduce urine pH and increase ammonium excretion to the same extent as control subjects. A similar finding in three of six patients was reported by Milutinovic et al. (120). Martinez-Maldonado (115), however, reported that the ability of four ADPKD patients to reduce urine pH was normal. In a more extensive study of ADPKD patients and a control group of subjects, Torres et al. (168) found that only one of eight patients could not reduce urine pH below 5.3. No significant differences existed between the ADPKD patients and control subjects. However, the ability of the patients to increase ammonium excretion in response to an acid challenge was reduced. To determine if a defect in proton secretion existed in the collecting tubules of these patients, they measured the urine to blood PCO2 difference during diuresis caused by infusion of NaHCO3 and found no significant impairment. They also measured the concentration of ammonium in cyst fluid of nine patients in an attempt to determine if a gross defect existed in ammonium production. The concentration of ammonium in nongradient cysts (cysts containing fluid with a Na+ concentration and pH similar to plasma) was higher than that reported in renal venous blood of normal subjects. Ammonium concentration was higher and urine pH lower in gradient cysts (cysts with a low Na+ concentration and a low pH). The calculated concentration of the base, NH3, in cyst fluid of the two types of cyst did not differ but that level was higher than that usually measured in renal venous blood. The authors considered this to be circumstantial evidence indicating that ammonium production is not reduced in ADPKD. They pointed out the role of the countercurrent system in the excretion of ammonium and suggested that disruption of the medullary architecture may decrease the efficiency of this system and account for the reduced ability to excrete ammonium.

Organic anion secretion has been examined in the Han:SPRD rat (163). In 7- to 16-wk-old heterozygous rats, GFR and the maximal rate of PAH secretion did not vary from that of healthy control littermates. Examination of proximal cysts in vivo using fluorescence microscopy indicated that 27 of 29 cysts secreted sulfonefluorescein, which is transported by the organic anion system. Secretion of PAH has been shown to generate fluid secretion in isolated segments of rabbit proximal straight tubules (71). However, micropuncture examination of segments of superficial proximal tubules or cysts, isolated by upstream and downstream wax blocks, failed to show fluid secretion when PAH was infused intravenously (163).

    VI. CYST FLUID
Top
Previous
Next
References

A. Composition

A number of studies have focused on the composition of cyst fluid as a means of understanding the properties of the lining epithelium and the origin of the fluid (15, 59-61, 88, 182). As noted initially by Gardner (59), cysts can be divided into two categories: nongradient or proximal cysts and gradient or distal cysts. This categorization is based primarily on the Na+ concentration of their fluids as related to normal plasma values. However, the terms also apply to other common tubular fluid constituents. Cyst fluids with a low Na+ concentration tend to have high K+ and H+ concentrations and low Cl- concentrations, whereas the concentrations of these substances in nongradient cysts fall approximately into the range usually seen in plasma. Generally, approximately two-thirds of the total population of the cysts sampled fall into the nongradient category and 26-33% fall into the gradient category, depending on the authors' definition of the two categories (60, 88). Huseman et al. (88) noted that the proportion of the two types of cysts approximated that of the population of the proximal and distal tubules found on the surface of the normal kidney. Nongradient cysts have been observed in all kidneys studied, whereas gradient cysts have not been identified in every case. Gardner et al. (60) have contended that adult polycystic kidneys differ from each other because of this. A statistical analysis applied to the available data indicated that it is unlikely that a sampling error contributed to this discrepancy. However, in those kidneys in which gradient cysts were not found, only 5-12 cysts were sampled per kidney out of a population that probably exceeds 1,000.

All the reports of cyst fluid osmolality indicate that the values fall into a range that is close to plasma osmolality. However, the gradient cyst values were somewhat higher than those of nongradient cysts in one study: 312 vs. 285 mosmol/kgH2O (60). The low Na+ and Cl- concentrations of fluid from gradient cysts imply that other solutes contribute significantly to the total solute concentration, and Gardner (59) found that amino acids accounted for 50-100 mosmol/kgH2O. More remarkable is the finding that the glucose concentration of proximal or nongradient cysts is similar to that of serum; the concentration in distal or gradient cysts is higher (59, 61, 88). This finding implies that the proximal cysts do not reabsorb glucose or that the cyst walls are quite permeant to it. The latter may be quite probable.

B. Exchange With the Circulation

The fluid within each cyst is not a stagnant pool. Jacobsson et al. (92) found that 2-5 h after intravenous injection of tritiated water, the concentration in cyst fluid averaged 88% of the concentration in plasma. The average turnover was more rapid in smaller cysts than in the larger. However, even in cysts with volumes greater than 10 ml, the mean value for tritiated water concentration was 82% of the plasma level. The cyst walls also exhibit some permeability to much larger molecules. An early study by Lambert (103) and a second by Bricker and Patton (19) indicated that inulin entered the cyst fluid of the majority of cysts studied. They attributed this to filtration into cysts connected to glomeruli, but a later study by Wickre and Bennett (182) of a single, nonuremic patient found that the amount of inulin recovered in cyst fluid would require a single nephron GFR of 68-250 times the normal value. Technetium-labeled diethylenetriaminepentaacetic acid (DPTA), 131I-hippurate, and PAH also entered cysts in amounts too large to be due to filtration. The DPTA is not secreted by normal proximal tubules, but it is possible that hippurate and PAH were actively secreted into the cyst cavity. Infections commonly occur in cystic tissue, and it became apparent that treatment may require the use of antibiotics that will reach an effective concentration within cyst fluid. Many will penetrate the cyst wall, some of these are weak electrolytes, and accumulation within cysts appears to depend on pH gradients. Clindamycin and metronidazole, for example, are cationic, lipid-soluble drugs that will accumulate in gradient cysts that contain fluid with a low pH (150). Penicillin derivatives evidently enter cysts very slowly, suggesting that the organic anion secretion mechanism functions poorly in many proximal cysts (11).

An electron microscopic study of cyst walls provided a morphological basis for the evident large permeability of nongradient cysts and for the difference between nongradient and gradient cysts (35). Eleven of 13 nongradient cysts from 5 patients were lined by epithelia with open or short, closed zonulae occludens that were permeable to lanthanum. The fluid from the two remaining nongradient cysts, each from different patients, contained Na+ concentrations equal to or greater than serum concentrations. However, unlike the other nongradient cysts, the fluid-to-serum glucose concentration ratios were substantially less than one (0.07 and 0.17). The epithelia of these cysts possessed "tight" junctions with fusion of the outer lamellae of the plasma membranes of adjacent cells for a length of at least 50 nm. Thus, in those nongradient cysts with glucose concentrations close to that of serum, glucose may have penetrated through the leaky zonulae occludens. All seven of the gradient cysts (fluid-to-serum Na+ concentration ratios <0.4) exhibited long, closed junctions between adjacent cells that were impermeable to lanthanum (35). Gardner et al. (60) confirmed that the average fusion depth of the zonula occludens differed between nongradient and gradient cysts, 37 vs. 204 nm.

    VII. CYST GROWTH
Top
Previous
Next
References

A. In Vivo

Microdissection of human polycystic kidneys in an early stage of the disease revealed that small cysts look like a ballooning in a bicycle inner tube or a knob in the sidewall of a tire (7). Indeed, an early hypothesis of cyst formation proposed that the tubular basement membrane is abnormally compliant (28, 43). A study of the relationship between tubular hydrostatic pressure and outer diameters of segments of rabbit tubules had shown that the compliance of the normal basement membrane would maintain tubular diameter well below 100 µm at pressures up to 60 cmH2O or 44 mmHg (179). These pressures are higher than those recorded in vivo within cysts of ADPKD patients and in the cystic nephrons of rats treated with nordihydroguaiaretic acid (16, 48). However, it was considered possible that defective formation of the basement membrane could result in tubular expansion at normal or even less than normal intratubular pressures. To investigate this hypothesis, Welling and Welling (180) developed mass balance equations for fluid movement into and out of the cysts by filtration, secretion, absorption, and distal outflow. An initial postulate of a constant and normally transporting cellular and basement membrane mass in the stretching cyst was obviously untenable on geometric grounds, since the tubular wall would become impossibly thin at cyst diameters beyond 1 mm. Cellular growth or hyperplasia must be invoked to explain the size of cysts routinely seen in cystic kidneys. However, they also showed that cyst growth would stop when the absorptive capacity of the hyperplastic tubular wall rose to equal the inflow (filtration) rate. In a cyst in which distal outflow is totally obstructed and each added cell absorbs at a normal rate, the spherical diameter would not increase beyond 2 mm. Additional cyst growth beyond that point would require a reduction in the absorptive capacity of the lining cells (180). Thus a combination of proliferation of cells with a low reabsorptive capacity and an abnormally compliant basement membrane could conceivably lead to cyst formation, particularly if a downstream obstruction increased tubular pressure.

Tubular obstruction, caused by polyps appearing on the inner tubular wall, has been considered to be an initiating factor in the dilatation of a tubule and the subsequent formation of a cyst. Polyps have been observed in ADPKD cysts and on occasion are found at a position suggesting that they may obstruct an outflow pathway (13, 49, 50). Polyps that impinged on outflow lumens of collecting tubular cysts were found in rats treated with nordihydroguaiaretic acid (48). Intratubular pressures did not differ among nondilated and cystic nephrons in the treated rats and nephrons in control animals, but microperfusion of the cystic nephrons caused pressure to rise significantly higher than it did in the others. The excretion of [3H]inulin after microinjection into cystic nephrons was reduced and delayed compared with that in nondilated tubules. Similar findings were obtained in rats in which cystic changes were produced by diphenylamine and diphenylthiazole (49). Tanner et al. (162) repeated these experiments in the Han:SPRD rat. The kidneys in the examined cystic rats were twice the size of those in normal control animals, but mean arterial pressure, GFR, and urine flow rate did not differ in the two groups. Pressures in the cystic nephrons were modestly elevated above that registered in the nephrons of the control rats, 18.5 versus 14.3 mmHg. Pressures in the noncystic tubules of the cystic rats did not differ statistically from those in the cysts or in the normal rats. The response to microinfusion into the cysts at 15 and 50 nl/min was highly variable. Thirty-nine percent of the cysts displayed a pressure increase above 30 mmHg and were considered to be obstructed. In the remaining 61%, the pressure increase was much less and similar to that obtained in control animals. When lissamine green was injected intravenously, the dye appeared in all the observed surface cysts, and washout of the dye occurred, albeit at a delayed rate, indicating that these cysts were connected to functioning tubules. Scanning electron microscopy studies of these cystic kidneys did find evidence of obstruction in some cysts in the form of tubular casts or narrowing of the tubular lumen. Polyps were not observed. These results suggest that in this genetic model, as compared with the chemical models, tubular obstruction may contribute to cyst formation but is not a necessary component (162). Similar conclusions were reached in studies of two genetic mouse models (62, 160). These models differ from the human form of ADPKD in which the majority of cysts do not retain a tubular connection (70).

A more direct assessment of basement membrane deformability was made by applying a negative pressure to a tubular or cystic wall via a micropipette and measuring the distance the wall was pulled into the pipette (69). Viscoelastic creep was determined by measuring the time-dependent effect of pipette aspiration on membrane deformation. Four models of cystic disease were studied: a mouse genetic model of recessive polycystic kidney disease [C57 BL/6J (cpp/cpk)], two chemically induced models (rats fed either diphenylthiazole or nordihydroguaiaretic acid), and an environmental model (the CFWw mouse develops cystic disease when raised in a conventional environment rather than in a germ-free environment). No differences appeared in the measurements of deformability and viscoelastic creep between normal and cystic tubules in any of the animal models. The calculated transtubular pressure that would be required to dilate the tubules to the extent seen in these rodent models was much higher than the intratubular pressures that have been measured.

Thus tubular obstruction and a reduction in the tensile strength of the basement membrane are not constant features of cystic disease. However, cellular proliferation occurs in all the animal models and in the human. Every cyst that has been examined has shown signs of proliferation, including those cysts that did not exhibit polyps or cordlike hyperplasia. It seems apparent that this proliferation, with each cell making its contribution to the basement membrane, is the factor that initiates cyst formation. Studies of this proliferation have concentrated on the cultured cell models.

B. Cyst Formation by Cultured Cells

The microcyst model pioneered by McAteer et al. (118) has proven to be very useful in studies of growth of cysts by renal epithelial cells. Madin-Darby canine kidney cells seeded within a collagen matrix form cysts by clonal growth of individual cells. These microcysts are filled with fluid, and the cells in the lining monolayer are polarized so that the apical surface faces the lumen and the basolateral surface is in contact with the collagen matrix. (These cysts formed by cultured cells are referred to as microcysts to distinguish them from the cysts formed in vivo in the polycystic kidney.) The only source for the fluid is secretion across the cyst wall. Hydrostatic pressure within the microcyst was found to exceed that in the bath by an average of 6.7 mmH2O; thus the secretion must involve the addition of solute to the cyst cavity (73). Cell proliferation in MDCK monolayers can be stimulated by stretching the underlying membrane support (166). However, secretion of fluid into the microcysts (presumably stretching the cyst wall) is not necessary for cell proliferation. Ouabain stopped fluid secretion by the MDCK microcysts, but the cells continued to proliferate and a solid tumorlike structure was formed (73). Prostaglandin E1, which stimulates cAMP production in MDCK cells (167), increased the growth of fluid-filled microcysts in defined medium, suggesting that cAMP may play a role both in cell proliferation and in fluid secretion (73). In a more detailed study (112), it was found that arginine vasopressin (AVP), cholera toxin, and forskolin increased the number and volume of MDCK microcysts and the proliferation of cells in monolayer cultures. Prostaglandin E1, cholera toxin, and forskolin also increased the level of cAMP in the microcysts, and 8-bromo-cAMP increased the number and volume of microcysts and the proliferation of cells in monolayer cultures. The rate of fluid secretion by the monolayers was measured and found to be stimulated by forskolin, PGE1, and 8-bromo-cAMP. The effects of these agonists on microcyst formation and fluid secretion were potentiated by 3-isobutyl-1-methylxanthine (IBMX), an inhibitor of phosphodiesterase. By itself, IBMX also stimulated cell proliferation but did not induce cyst formation. Levels of cAMP generated in response to IBMX were lower than those obtained in response to forskolin, PGE1, and cholera toxin, but IBMX was as potent as the other agents in generating cell proliferation. This suggests that relatively low levels of cAMP are sufficient to maximally stimulate proliferation but that higher levels are required to induce fluid secretion. Both proliferation and fluid secretion are required to induce microcyst formation by MDCK cells (112).

The ability of primary cultures of cortical cells removed from normal human kidneys (HKC) to form microcysts was examined (111, 123, 125). Histochemical analysis indicated that microcysts and monolayers were formed primarily by cells derived from the distal nephron (123). The polarity of these cells was the same as that of cells in the MDCK microcysts; the apical membrane faced the lumen of the cyst. In contrast to MDCK cells, agents that increase cAMP levels within cells did not stimulate microcyst formation and growth in the absence of epidermal growth factor (EGF) or transforming growth factor ( TGF ) -alpha (111, 123). Either of these growth factors was sufficient to promote microcyst initiation and growth. Their effect was strongly augmented by inclusion of insulin in the growth medium. Platelet-derived growth factor, TGF-beta , and basic fibroblast growth factor did not initiate microcyst formation; in fact, TGF-beta inhibited the effect of EGF. Receptor-mediated agonists of the cAMP pathway, PGE1, AVP, parathyroid hormone, vasoactive intestinal polypeptide, and isoproterenol enhanced the initiation and enlargement of the microcysts when EGF and insulin were present in the growth medium. Forskolin, cholera toxin, IBMX, and 8-bromo-cAMP were also effective, as was a relatively high concentration (10-7 M) of ANF, which increases intracellular levels of cGMP. In contrast to their effect on microcyst formation and growth, forskolin, PGE1, AVP, and 8-bromo-cAMP induced fluid secretion by monolayers of cells in the absence of EGF. Epidermal growth factor itself did not induce fluid secretion; however, it potentiated fluid secretion in the presence of forskolin (111, 123, 125).

End-stage ADPKD kidneys removed from patients for medical reasons have been the major source of tissue for developing primary cultures. These cultures have been derived from explants of cyst walls (185) or from cells that were freed from minced cyst walls by use of collagenase (111). McAteer et al. (117) found that when a section of cyst wall was embedded in and covered with collagen, the proliferating cells commonly formed an epithelial sac in which fluid accumulated. Cells seeded within a collagen matrix also formed cysts by clonal growth (111). As with the HKC cultures, EGF was required for the initiation of microcyst formation, but, unlike the HKC cultures, EGF alone was not sufficient. Agonists that induce the production of cAMP were also required. These agonists also induced fluid secretion by monolayers of the ADPKD cells as they did in monolayers of the MDCK and HKC cells (67, 111).

The studies described above make it clear that cAMP is involved in the proliferation of the cultured cells and in generating fluid secretion. The roles of additional factors potentially involved in cystic cell proliferation, including endocrine, paracrine, and autocrine substances, and oncogenes have been studied (for reviews, see Refs. 33, 183). They are not reviewed here. The studies performed with the MDCK and HKC cells also indicate that abnormalities in the PKD1 or PKD2 gene are not required to induce cyst formation or fluid secretion by renal epithelial cells.

    VIII. TRANSPORT MECHANISMS INVOLVED IN FLUID SECRETION
Top
Previous
Next
References

A. Madin-Darby Canine Kidney Cell Model

Mangoo-Karim et al. (112) analyzed the composition of the fluid secreted by MDCK microcysts and monolayers using electron-probe analysis. The ratio of Cl- concentrations of the secreted fluid to that of the bathing medium was 1.12 for the microcysts and 1.18 for the monolayers. The corresponding Na+ concentration ratios were 0.98 and 1.02. These values suggested that Cl- is secreted by MDCK cells. Secretion of Cl- is known to drive fluid secretion by a variety of epithelia, and Simmons and co-workers (22, 152, 153) had reported that monolayers of MDCK cells grown on a permeable membrane secreted Cl- in response to forskolin, vasopressin, PGE1, and vasoactive intestinal polypeptide.

In the standard model for Cl- secretion, Na+-K+-ATPase, located in the basolateral membrane and acting in concert with K+ channels, establishes the requisite chemical and electrical gradients that are utilized by the Cl- transporters (106). The Na+-K+-2Cl- cotransporter, also located in the basolateral membrane, brings Cl- into the cell. A Cl- channel in the apical membrane provides the efflux pathway. Secretagogues, often acting through the cAMP signal transduction system, activate the Cl- channel. The cotransporter and K+ conductance pathways in the basolateral membrane may also be activated by initiation of secretion. The activation of Cl- channels in the apical membrane and increased efflux of K+ across the basolateral membrane establishes a transepithelial electrical gradient, lumen negative, that drives net Na+ diffusion through the paracellular pathway. The addition of NaCl to the luminal fluid provides the osmotic force to drive fluid secretion.

A variety of studies have focused on the Cl- transport mechanisms in MDCK cells (see Ref. 153 for review). The presence of a large-conductance Cl- channel in MDCK cells, activated by epinephrine, was reported by Kolb et al. (102). Agents that raise cellular cAMP levels increased a Cl- conductance that is blocked by 9-anthracene carboxylate and diphenylamine-2-carboxylate (DPC) (18, 104) and the Cl- channel blocker 5-nitro-2-(3-phenylpropyl-amino)benzoic acid inhibited Cl- secretion (152). Bumetanide, an inhibitor of Na+-K+-2Cl- cotransport, reduced cell Cl- accumulation (151). Measurements of [3H]bumetanide binding to monolayers located the cotransporter on the basolateral surface and indicated that the density of the cotransporters was similar to that of the Na+-K+-ATPase mechanism (21).

Tanner and co-workers (164, 165) developed a novel technique to investigate the role of these Cl- transport mechanisms in fluid secretion by MDCK microcysts. A small section of collagen gel containing the cysts was placed in a perfusion chamber on the stage of an inverted microscope and superfused with culture medium or with Ringer solution. A single microcyst was selected for examination, and net fluid transport by the epithelium was measured by following changes in cyst cavity volume. This volume was determined by morphometric measurement of the diameter of the cyst cavity in videotaped images of the equatorial plane of the cyst. The volume of the cells lining the microcyst was calculated by subtraction of the cyst cavity volume from the total cyst volume. This technique was modified by Sullivan et al. (158) to permit epifluorescent measurements in single microcysts of cell pH with the use of 2',7'-bis(carboxyethyl)-5,6-carboxyfluorescein and cell potential changes with the use of bis-oxonol (158). Macias et al. (108) used double-barreled microelectrodes to measure electrical gradients and Cl- activity in the cells and in the cyst cavity. The subtype of MDCK cells used in the microcyst experiments was also cultured to form monolayers of cells on permeable membranes. Fluid secretion by these monolayers was measured in 24-h periods, and monolayers were also mounted in Ussing chambers to permit measurements of transepithelial voltage (Vt), resistance (Rt), and short-circuit current (Isc) (75, 113).

Madin-Darby canine kidney microcysts superfused with control media absorbed fluid from the cyst cavity at a slow rate (158, 164). Microcysts superfused with dibutyryl cAMP, forskolin, or AVP, often in conjunction with IBMX, secreted fluid at a constant rate for 60-90 min. Cell volume fell upon initiation of secretion and did not recover within 60 min (158, 164). Ouabain (0.1 mM), applied to the basolateral surface of the microcysts, inhibited fluid secretion induced by the secretagogues. The direction of the Isc and the magnitude of the increase induced by the application of forskolin to the monolayers indicated that fluid secretion is probably driven by active anion transport from the basolateral to the apical surface (75). The cAMP content of the monolayers was also increased by forskolin.

Microelectrode measurements conducted on the unstimulated microcyst by Macias et al. (108) indicated that the transepithelial electrochemical gradient across the cyst wall opposed passive Cl- secretion. The cellular Cl- activity, 60 ± 1 mM, exceeded the value that would exist if electrochemical equilibrium for Cl- prevailed across the basolateral and apical membranes. Thus Cl- entry into the cell across the basolateral membrane must occur against an electrochemical gradient, whereas exit across the apical membrane could occur passively in response to the favorable electrochemical gradient. The results of the microelectrode study suggested that a Cl- conductance pathway may be present in the apical membrane. This hypothesis was supported by the results of experiments in which the anion channel blocker DPC was applied to the apical surface of monolayers (113). At 3 mM, DPC inhibited fluid secretion initiated by application of forskolin. It also depolarized the monolayers and inhibited Isc, confirming the results of earlier studies (18, 104, 152).

Tanner et al. (164) investigated the nature of the mechanism transporting Cl- into the cell across the basolateral membrane with the use of Cl- transport inhibitors applied to microcysts stimulated to secrete fluid by AVP plus IBMX. Secretion rates were measured in separate control and experimental groups. Furosemide (0.1 mM) inhibited secretion, but bumetanide (0.01 mM) caused an insignificant decrease of 23%, and chlorothiazide (1 mM) had no effect. The nominal absence of bicarbonate and CO2 inhibited secretion, and pretreatment with DIDS (0.1 mM) inhibited secretion as did amiloride (1 mM). However, acetazolamide (0.1 mM) did not. Also, DIDS inhibited the loss of cell volume when the cyst was exposed to a solution containing no Cl-. Microelectrode measurements on cysts not exposed to secretagogues indicated that the cell Cl- level fell when the microcyst was exposed to a high HCO-3 medium and when 1 mM amiloride was added to the medium (108). The results of these studies suggested that a Cl-/HCO-3 exchanger, paired with a Na+/H+ exchanger, operated in the basolateral membrane. The authors proposed that the Cl-/HCO-3 exchanger is responsible for Cl- entry across the basolateral membrane when secretion is stimulated (108, 164). However, the results of similar experiments did not support these conclusions (158). In this study, control and experimental measurements were made sequentially in 30-min periods on each microcyst. Secretion initiated by AVP plus IBMX was not affected by subsequent addition of DIDS (0.1 mM) to the bathing medium. Cell pH might be expected to change if initiation of secretion accelerated Cl-/HCO-3 exchange. However, epifluorescent measurements of cell pH indicated that no transient or long-lasting change in cell pH occurred. The application of bumetanide (0.1 mM) did not alter the slow rate of fluid reabsorption or cell volume when it was added to the bathing medium in the absence of a secretagogue. However, 0.01 mM bumetanide reduced the rate of secretion to values not different from zero after the administration of AVP plus IBMX, an effect that was partly reversed by the removal of the inhibitor. It was concluded that Cl- enters the cell during secretion via a Na+-Cl- or a Na+-K+-2Cl- cotransporter. The Cl-/HCO-3 exchanger may maintain the high Cl- concentration within the resting cell, but activation of the cotransporter maintains the supply of Cl- to the cell during secretion (158). Studies conducted in other secretory epithelia are consistent with this interpretation. The binding of benzmetanide, a potent inhibitor of the cotransporter, to cell membranes of the shark rectal gland is greatly increased by the activation of secretion (53, 107). The binding of bumetanide to cultured canine airway epithelia is also enhanced by activation of secretion (83).

The microcyst cell volume fell ~7% whenever secretion was initiated by AVP or forskolin, indicating a loss of cell solute (158, 164). Apparently, activation of the cotransporter limited that loss, since a further loss of cell volume occurred after the application of 0.1 mM bumetanide (158). Epifluorescent measurements of changes in cell electrical potential indicated that application of 0.1 mM bumetanide, after initiation of secretion, hyperpolarized the cell. This may have been the result of a fall in the depolarizing efflux of Cl- across the apical membrane resulting from inhibition of Cl- entry into the cell via the electrically neutral cotransporter. Bumetanide also reduced forskolin-stimulated fluid secretion and Isc in monolayers of the ADPKD cells (113).

The role of electrical gradients in Cl- and fluid secretion was investigated in experiments on MDCK microcysts in which Ba2+ (1 mM) was added subsequent to the initiation of secretion by AVP plus IBMX (158). Barium depolarized the cell, blocked fluid secretion, and caused an increase in cell volume (Fig. 2). These changes were completely reversed by removal of Ba2+. It was considered that the depolarization of the cell resulted from blockade of K+ channels in the basolateral membrane. The depolarization evidently reduced the electrochemical force driving Cl- efflux across the apical membrane, resulting in the inhibition of fluid secretion. The retention of K+ and Cl- within the cell may have caused the expansion of cell volume that was observed. It had been demonstrated earlier that Ba2+-induced depolarization of the basolateral membrane also inhibited Cl- secretion by the canine tracheal epithelium (181). Greger and Schlatter (77) found that depolarization of the basolateral membrane by Ba2+ reduced the electrochemical gradient for Cl- efflux from the cells of the shark rectal gland.


View larger version (20K):
[in this window]
[in a new window]
  		FIG. 2.   Effect of barium on a cultured Madin-Darby canine kidney cell microcyst induced to secrete fluid by superfusion with arginine vasopressin (AVP) and 3-isobutyl-1-methylxanthine (IBMX). Cavity volume and volume of cell layer were determined by morphometric measurement of a videotaped image of cyst with microscope focused on equatorial plane of cyst (see Fig. 3A). Changes in membrane electrical potential were determined by epifluorimetric measurement of cellular bis-oxonol levels. Top: cavity volume. Solid lines are regression lines for each period and were used to calculate fluid transport rates. Before application of secretagogues, volume was stable. Superfusion of cyst with a medium containing 10 mU/ml AVP and 0.1 mM IBMX initiated fluid secretion at a rate of 7.2 nl·min-1·cm2 cavity surface area-1. Addition of 1 mM barium to superfusion medium reversed fluid secretion to absorption at a rate of -3.37 nl·min-1·cm-2. After removal of barium, secretion resumed at a rate of 5.45 nl·min-1·cm-2. Middle: experimental-to-control ratio (E/C) of cell volume. Horizontal dashed line indicates an E/C value of 1.00. Control cell volume was 1.10 µl/cm2 cyst outer surface area. Upon initiation of secretion, cell volume fell rapidly to ~91% of control level. After addition of barium, volume transiently rose to 95% of control level. Bottom: E/C of epifluorescence of bis-oxonol. A fall in ratio indicates membrane hyperpolarization; a rise indicates depolarization. Addition of secretagogues caused membrane hyperpolarization, and subsequent exposure to barium caused membranes to depolarize. [From Sullivan et al. (158). Used with permission from Kidney International.]

These studies strongly support the conclusion that Cl- is secreted by MDCK cells by mechanisms that are very similar to those employed by a variety of secretory epithelia. In addition, the data indicate that fluid secretion by monolayers and microcysts formed by these cells is driven by the active transport of Cl-. Monolayers and microcysts derived from primary cultures of human kidney cortex cells also secreted fluid in response to adenylyl cyclase agonists (Table 2).

 
View this table:
[in this window] [in a new window]
  		TABLE 2.   Fluid transport rates in preparations of cyst-forming epithelia

B. Autosomal Dominant Polycystic Kidney Disease Tissue

Perrone and co-workers (133, 136) were the first to attempt to investigate the transport mechanisms present in ADPKD tissue by mounting segments of cyst walls in Ussing chambers. The cysts from which the segments were dissected were classified as gradient or nongradient cysts on the basis of Na+ and K+ concentrations in the cyst fluid. The cyst epithelium removed from nongradient cysts exhibited a high conductance (20-80 mS/cm2) and quite low Vt and Isc values. Amiloride, ouabain, acetazolamide, and bumetanide did not affect these values or the unidirectional fluxes of Na+ and Cl-. The conductances of two gradient cysts were lower, 3.1 and 6.5 mS/cm2, Vt and Isc were higher, and the unidirectional fluxes of Na+ and Cl- were lower than in the nongradient cysts. Amiloride inhibited the Isc in the gradient cysts. The effects of agents that increase cellular cAMP levels were not investigated in these early studies. Another attempt to mount segments of cyst walls in Ussing chambers was unsuccessful (unpublished data). The cyst walls were fibrous, and the thickness of the tissue was highly irregular. An inability to obtain a tight electrical seal of the chamber wall to the tissue may have caused the low and highly variable Rt and Vt values measured. However, the cyst walls are active as another method has documented (190).

Ye and Grantham (190) applied a gravimetric method to directly demonstrate the ability of the cystic epithelium to secrete fluid in vitro. Intact cysts were carefully dissected from ADPKD kidneys. Cyst fluid was aspirated to determine its volume so that cyst surface area could be calculated. One-third of the volume of cyst fluid was reinjected into the cavity of some cysts; culture medium was injected into the cavity of the remaining cysts. All the cysts were then incubated in the culture medium. The net rate of fluid transport was determined by weighing the cysts at 24-h intervals. Forskolin was added to the bathing medium after the control period. Cysts containing natural cyst fluid secreted fluid at a rate of 0.87 µl·cm-2·h-1. The addition of forskolin did not significantly alter that rate. The cysts containing culture fluid did not secrete fluid at a significant rate until forskolin was added to the bathing medium (Table 2). The rate of fluid secretion by cysts containing cyst fluid and bathed in medium containing forskolin was inhibited by ouabain (0.01 or 0.1 mM) added to the bath. This study indicated that the cystic epithelium secretes fluid by an active process and that secretion can be increased by an unidentified secretagogue present in cyst fluid. Monolayers developed from primary cultures of ADPKD cells also secreted fluid when agonists that raise the levels of cellular cAMP were applied (75, 111, 113, 124), as did the microcysts developed in culture (178) (Table 2). In fact, the cultured cells in both the monolayer and cyst configuration exhibited the ability to drive net fluid transport in either direction. Monolayers and microcysts reabsorbed fluid in the control state in the absence of cAMP agonists and rapidly reversed the direction of fluid transport to secretion when exposed to forskolin (75, 113, 178) (Table 2).

In a more detailed study of intact cysts, 30 were dissected from the kidneys of 2 different donors (75). The cysts were filled with culture medium and incubated in the same medium. Forskolin (10 µM) was added to the bath of 10 cysts. All the cysts secreted fluid during the 24-h period at a mean rate of 0.19 ± 0.04 µl·cm-2·h-1. Ouabain (10 µM) was added to the cavity medium and forskolin to the bath of another group of 10 cysts. This application of ouabain to the apical surface did not affect the rate of secretion (0.18 ± 0.07 µl·cm-2·h-1). Both forskolin and ouabain were added to the bath (basolateral surface) of a third group of 10 cysts. The rate of secretion by this group was substantially lower (-0.14 ± 0.15 µl·cm-2·h-1). Similar results were obtained from monolayers of cultured ADPKD cells. Ouabain (0.1 µM) added to the apical surface did not affect forskolin-induced secretion; ouabain added to the basolateral surface greatly inhibited secretion. Thus Na+-K+-ATPase, accessible from the basolateral surface, does participate in the solute transport that drives fluid secretion.

Monolayers of cultured ADPKD cells were also successfully studied in Ussing chambers (75, 113). In 36 such preparations derived from cysts removed from 7 kidneys, the Rt averaged 156 Omega ·cm2; the apical surface was negative with respect to the basolateral surface in all of the monolayers, averaging -0.9 mV, and the Isc registered as a positive current flowing from the apical to the basolateral surface (Table 3). Forskolin increased the conductance, hyperpolarized Vt, and increased Isc. Forskolin (10 µM) also caused a threefold elevation of cAMP in the monolayers (75). The direction of Vt and the Isc before and after administration of forskolin are compatible with cation reabsorption or anion secretion. Because forskolin also stimulates fluid secretion by these monolayers (Table 2), it is apparent that the tissues secrete anion (113). The forskolin-stimulated current was inhibited by basolateral application of ouabain (10 µM) but not by apical application (75). The effect of forskolin on Isc and fluid secretion was also inhibited by basolateral application of bumetanide and by apical application of DPC (113). The forskolin-stimulated Isc was not affected by administration of DIDS to the basolateral surface. Glibenclamide (200 µM) applied to the apical surface reduced Isc by 31% (unpublished data). The effect of these inhibitors suggests that Cl- enters the cell via a Na+-Cl- or a Na+-K+-2Cl- cotransporter and exits via a Cl- conductance pathway.

 
View this table:
[in this window] [in a new window]
  		TABLE 3.   Transepithelial electrical properties of cultured autosomal dominant polycystic kidney disease monolayers

Several approaches were taken to determine if the anion secreted by ADPKD tissue is Cl-. Monolayers of ADPKD cells were exposed bilaterally to solutions in which Cl- was replaced by cyclamate. The absence of Cl- reduced Isc, and forskolin did not significantly increase it (178). The Cl- efflux rate constant was measured in monolayers of cells grown on a plastic surface and loaded with 36Cl. Forskolin increased the rate constant for that efflux. Diphenylamine-2-carboxylate reduced the control efflux rate constant and prevented the forskolin-induced increase (40).

Experiments performed on microcysts formed by cultured ADPKD cells also verified that the anion secreted is Cl- (178). Single ADPKD microcysts (Fig. 3A) were isolated and studied as described in section VIIIA for the MDCK microcyst. Fluid transport was measured by determining changes in cyst cavity volume with time, and cell volume was also monitored. In the control state, cavity volume declined with time at a steady rate for at least 60 min without a change in cell volume, indicating that the cells were reabsorbing fluid (Fig. 3B). The application of forskolin plus IBMX or 8-bromo-cAMP plus IBMX reversed the direction of fluid transport within 5 min, and the secretion was maintained at a steady rate for 60 min. Cell volume rapidly fell when secretion was initiated and did not recover. The loss of cell volume averaged 7.5% after 20 min of exposure to 8-bromo-cAMP and IBMX. Bumetanide rapidly blocked fluid secretion and caused an additional 4.2% loss of cell volume. Fluid secretion resumed after the removal of bumetanide, and cell volume increased 2.6%. Changes in intracellular Cl- concentration ([Cl-]i) were monitored with the use of the fluorescent indicator 6-methoxy-N-ethylquinolinium chloride (178). No noticeable change in [Cl-]i occurred when secretion was initiated with 8-bromo-cAMP. The rapid loss of cell volume without a change in [Cl-]i indicated that Cl-, as well as water, was lost when secretion was initiated. The effect of