View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Kidney International, Vol. 39 (1991), pp. 450—463

Tubulocystic epithelium PATRICIA D. WILSON and ANN C. SHERWOOD

Department of Physiology and Biophysics, UMDNJ—Robert Wood Johnson Medical School,Piscataway, NewJersey, USA

Pathology of human renal cystic disease when parallels or contrasts have bearing on studies in human There are several pathologic conditions in humans whichtissues. result in tubulocystic enlargement in the kidney [1, 2]. In some RPKD is a rare but lethal form of polycystic kidney disease. cases, renal cystic development is genetically determined andIt occurs with an estimated frequency of 1 in 10,000 and maybe the primary disease event (Table 1). The most commontypically children are born with enlarged, malfunctioning, mul- example is autosomal dominant polycystic kidney diseaseticystic kidneys. For these reasons, this form of disease has (ADPKD) which is estimated to afflict 1 in 1000 live births [3].been studied rarely, but recently a mutant mouse strain C57/6J Autosomal recessive polycystic kidney disease (RPKD) occurs(cpk) has been characterized and shown to represent an excel- less frequently (1 in 100,000 live births), but is more aggressivelent animal model for this disease [7, 8]. This fortunate circum- in its course since it is usually lethal to infants within their firststance does not apply to the more common ADPKD, which year. Another significant inherited condition leading to renalaffects approximately 500,000 people in United States. No tubule cyst formation is the nephronophthisis-medullary cysticknown animal model has been characterized. This disease disease complex, which has been described as an autosomaltypically shows a slow progression in which renal cysts increase recessive trait in the juvenile form and autosomal dominant inin number and size, gradually replacing normal renal tubule the adult form of disease. Tubulocystic enlargement may alsomass and leading to renal insufficiency in the fourth decade of occur sporadically in association with hereditary syndromes,life, although symptoms of reduced ability to concentrate the such as von-Hippel Lindau disease and tuberous sclerosis. Inurine and hypertension often precede loss of renal function [9, yet other cases, renal tubule cysts may develop in small or large10]. Classical, pioneering microdissection studies of ADPKD numbers in response to non-genetic factors. For instance,kidneys in the early stages of development of the disease simple cysts occur with increasing frequency with advancingshowed that cysts can arise as focal dilations originating from age, but rarely compromise renal function. By contrast, ac-any proximal or distal segment of the nephron [11—13]. This was quired renal cystic disease, which affects approximately 40% oflater confirmed by discriminatory lectin staining patterns [14]. non-ADPKD patients receiving hemodialysis for end-stage re-In addition, ion and pH analysis of cyst fluid aspirated from nal failure, may result in the development of multiple cysts [4].ADPKD kidneys had shown wide variations and led to the Additional reports have described renal cysts after long-termbroad classification of two types of cysts, designated "proxi- administration of lithium chloride [5] and exposure to diphenyl-mal" and "distal" [15—171. Macroscopic examination of end- amine [6]. stage ADPKD kidneys shows that cyst enlargement is asyn- A renal tubule cyst is defined as a segment of the nephronchronous and can reach enormous proportions. Cyst size enlarged to a diameter of 200 microns or more, and consists ofcommonly ranges < 1 mm to >10cm. This contrasts with a fluid-filled, distended sac lined by a single layer of epithelium.RPKD where cysts derive exclusively from distal collecting It would seem from the pattern of occurrence in humans andtubules, and cysts are smaller and more homogeneous in size. from experimental studies in animals and tissue culture, that The genes responsible for ADPKD and RPKD have not yet renal tubule cysts can form in response to genetic, non-geneticbeen isolated, although a major ADPKD gene (ADPKD-1) is and toxic influences. Although there is a common end prod-known to be linked to a highly polymorphic marker on the short uct—a tubulocystic epithelium lining a fluid-filled expandedarm of chromosome 16 [18]. However, it appears that in lumen—it maybe spurious to consider that the mechanism(s) ofapproximately 10% of cases of ADPKD, the gene on chromo- cyst formation are necessarily identical in all cases. Conceptu-some 16 is not responsible, since families with inheritance of ally, it seems more likely that several pathogenetic pathwaysADPKD unlinked to the 3'HVR region of the a-globin locus converge on a single means of phenotypic expression. Thehave now been found in United States and in Europe [19—21]. following discussion will center on what little is known aboutThe apparent existence of a minimum of two alleles for ADPKD the cell biology and underlying mechanisms governing cysthas not only complicated certainty of diagnosis, but unfortu- formation in human genetic polycystic kidney diseasesnately increases the complexity of the task of isolating and (ADPKD, RPKD), but will draw occasionally on results ob-characterizing this genetic defect at the molecular level. How- tained from some appropriate experimental models (Table 2),ever, significant progress has been made by several laboratories towards this goal. Progress has also been made at the cellular level. Light and © 1991 by the International Society of Nephrology electron microscopy of human ADPKD kidneys shows that

450 Wilson and Sherwood: Tubulocyslic epithelium 451

Table 1. Renal tubular cystic disease into certain ADPKD-associated symptoms. For instance, the Disease entity Causative factors Reference lack of ability of arginine vasopressin (AVP) to stimulate adenylate cyclase in cyst lining epithelia [23], would help to Autosomal dominant Genetic: gene 1, on [3, 18] explain the concentrating defect which occurs early in ADPKD polycystic chromosome 16 kidney disease gene 2, unknown [191 patients. The mechanism of this modification is not caused by a (ADPKD) complete inability to increase cAMP levels in ADPKD epithe- Autosomal recessive Genetic: gene and [1] ha, since forskolin and PGE2 generate a cAMP induction, polycystic kidney chromosome although this was < 50%of unknown the normal response [23, 25, 26]. disease (RPKD) These observations suggest an additional pre-cAMP lesion at Polycystic kidney Genetic: autosomal [1] disease: medullary dominant the level of AVP-receptor interaction, and/or faulty transduc- type tion of the signal via modified interactions with membrane G Juvenile Genetic: autosomal [11 proteins which act as the link between hormone-stimulated nephronophthisis recessive receptors and the catalytic subunit of adenylate cyclase. A Acquired renal Dialysis, drugs [4, 5] cystic disease possible explanation of hypertension, another early symptom of Simple cysts Advancing age [3] ADPKD patients, has been suggested by studies at the cellular von-Hippel Lindau Associated with [21 level. Using specific antibodies and tissue sections from 19 disease hereditary cases of ADPKD using specific antibodies, both an increase in (cerebroretinal syndrome number and an abnormal localization of renin-containing cells angiomatosis) Tuberous sclerosis Associated with [21 immunolocalization showed 50% of these cells were seen in the (Bouneville' s hereditary thin attenuated blood vessels that supply the walls of cysts, and disease) syndrome others were seen in the intervening in fibrous tissue [27]. Models of renal cystic disease Table 2. Experimental models of renal cystic disease The morphological characterization of cyst lining cells as Tubule simple epithelia has been confirmed and is retained in tissue Causative agent affected Species Reference culture, both in flat, two-dimensional monolayers grown on In Vivo dried collagen [23], and in three-dimensional explants embed- Genetic CT Cpk mouse [7, 81 ded in collagen gels [28, 29]. Although the ADPKD epithelial Drugs: diphenyl CT Rat [321 cells contain relatively few organelles and are morphologically thiazole less specialized than renal proximal tubules, thick ascending phenol II CT Rat [591 limbs of Henle or collecting tubule cells in vivo or in vitro, they Corticoids CT Rat, rabbit, [168] hamster remain polarized with apical tight junctions and microvilli In Vitro facing upward in the dish. ADPKD epithelial are similar to Metanephric PT Mouse [124] normal microdissected human renal tubules in vitro [29], in organ culture, their requirement for type 1 collagen or some other matrix hydrocortisone protein to establish this polarity. To confirm the validity of Monolayer PT,TAL,CT Human [231 ADPKD primary culture monolayers of microdissected human ADPKD epithelia cysts as a model system, a series of comparative studies was Explant ADPKD PT,TAL,CT Human [281 carried out in freshly isolated human cysts and confluent epithelia cultures. Good correlations were seen between the levels of Monolayer RPKD CT Human [169] epithelia marker enzyme activities and adenylate responsiveness to Monolayer RPKD CT Mouse [1701 hormones, which suggested the suitability of this primary epithelia culture system as an in vitro model system for the analysis of MDCK cell line in "Distal" Dog [1711 renal cyst formation. Since these ADPKD epithelia grow easily collagen gel and rapidly in culture and can be maintained in primary culture Abbreviations are: PT, proximal tubule; CT, collecting tubule; and for up to six weeks or passaged for up to 15 transfers, this TAL, thick ascending limb of Henle's loop. provides a practical, reliable and reproducible system for the analysis of cellular alterations in ADPKD epithelia [23]. The advantages of viable epithelial monolayers are numerous and cysts are lined by a single layer of polarized epithelium, inobvious, allowing for direct comparisons with similar epithelia which rudimentary microvilli and intercellular tight junctionsderived from normal renal proximal convoluted and straight line the apical cell surface at the lumenal aspect of the epithelialtubules (PCT, PST), cortical and medullary thick ascending sheet lining the cyst [22—24]. So few differentiated characteris-limbs of Henle's loop (cTAL, mTAL) and cortical and medul- tics of the original tubule segment remain that it is usuallylary collecting tubules (CCT, MCT) [25]. This is by no means impossible, in fixed tissue slices, to determine the derivation ofthe only model available for analysis of cyst formation, but is cysts based solely on morphological characteristics. Even at aone of the few employing human-derived epithelia. Some other functional level, isolated cystic epithelia have such dramaticallycommonly used model systems are listed in Table 2. Evaluation altered levels of marker enzyme activities and diminishedof every cellular model of cyst formation and its relevance to responses to hormone induction that characterization is dif-the mechanisms underlying human cyst formation in ADPKD ficult. However, these types of studies have provided insightand RPKD is beyond the scope of this review. 452 Wilson and Sherwood: Tubulocystic epithelium S

Fig. 1.Electronmicrograph of a cyst lining epithelial cell from an ADPKD kidney, undergoing mitosis. Note the absence of a nuclear membrane and chromosome condensation. Organdies are distributed at the periphery. Fixation was 2.5%glutaraldehydein 100 m sodium cacodylate buffer containing 5% sucrose for 4 hours. x 3,360.

Theories of polycystic kidney disease hyperplasia. The presence of hyperplastic nodules in ADPKD Over the years there have been several theories to explain thecysts was originally reported by Evan, Gardner and Bernstein derivation of renal cysts. Originally the obstruction theory was[34] and confirmed by others in ADPKD and cystic diseases favored [30]. This suggested that fluid accumulation was thethat predispose to malignant transformation, such as von- result of occlusion of the renal tubule at a point distal to theHippel Lindau disease [24]. Although the occurrence of hyper- glomerulus, and thus a build-up of pressure resulted in cystplastic nodules is not sufficiently frequent or extensive to formation. This theory has now been discounted since measure-invoke obstruction as a universal course of cyst formation, ments of intracystic pressures are not sufficiently high tocalculations of cell height and cyst size suggest that cell division account for tubular distension [31]. Also, it has been demon-must occur during cyst enlargement [35]. This has been con- strated that the majority of cysts become isolated from theirfirmed by detection of mitotic profiles in cyst lining epithelial of nephron of origin as an early event and not late as thoughtADPKD kidneys in vivo (Fig. 1). Mitotic figures have not been previously. seen in electron micrographs of normal human renal tubules. This theory gave way to one of increased compliance of the To explain formation of cysts instead of solid tumors, there cystic basement membrane. This still holds some sway as amust be an accompanying defect which results in fluid accumu- unifying concept which could explain the systemic nature oflation, presumably by secretion [36], since cysts have been cyst development in several organs of patients with ADPKDshown to close off from their nephron segment of origin early [101. Major basement membrane alterations are seen in drug-after the onset of expansion. The ability of cysts to accumulate induced cyst formation in rats [321, in cultured human ADPKDfluid very rapidly has been confirmed in vivo by injection of cells and in the later stages of dominant and recessive cystic3H-water, the majority of which was detected in cyst fluids after disease. However, measurements of compliance in acquiredfive hours [371. and mouse ADPKD cysts were not consistent with increased It can be concluded that whichever defective gene product is elasticity as the exclusive cause of cyst expansion [331. responsible for ADPKD or RPKD, the two basic criteria This theory has now given way slightly to one favoringnecessary for cyst formation to occur are increased epithelial Wilson and Sherwood: Tubulocystic epithelium 453 cell proliferation and altered fluid and transport. Alterations in the proliferative potential of epithelia can be the result of: modified epithelial/stromal interactions; altered susceptibilities a or synthesis of peptide growth factors; or abnormal expression 12 34 5678 of growth related genes including protooncogenes. Renal fluidFig. 2. Northern blot of total RNA probed with 32P-cDNA for type IV transport is directed by electrochemical ion gradients estab-collagen, (HT2I, Dr. D. Prockop, Jefferson, PA). Lanes 1 and 2, lished by the sodium pump and normally results in reabsorp- normal adult kidney; lanes 3 and 4 early stage ADPKD; lanes 5, 6, 7 and tion. Fluid accumulation in renal cysts could therefore be the8 end-stage ADPKD kidney. Note increase in steady state levels of type IV collagen mRNA, particularly in end-stage kidneys. Courtesy of Dr. result of alterations in ion pumping or reflect alterations ofiT. Norman, UCLA Division of Nephrology. polarity of cell structure and/or function. Evidence for each of these alterations will be considered.

The proliferative defect in tubulocystic epithelia UCLA). Thus, alterations in the structural component of cystic basement membranes may occur, but that as a late event is Matrix alterations unlikely to be a primary change initiating cyst enlargement. No Stromal-epithelial interactions are of central importance inchanges were found in type I collagen, laminin B 1 mRNA, or the patterning of growth and differentiation in the developingprotein expression. kidney [381. Although the precise mechanisms involved are not More dramatic alterations were seen in the heparan sulfate fully understood, it is clear that the dynamic composition of theproteoglycan (HSPG) component of cystic basement mem- extracellular matrix is responsible, at least in part, for control ofbranes. This integral component of the basement membrane all normal epithelial cell migration, spatial organization, prolif-consists of a core protein with glycosaminoglycan (GAG) side eration and differentiation. It is therefore not surprising thatchains. These molecules can vary considerably in size in a basement membrane defects are commonly found in diseasestissue specific fashion [5 1—53] but are thought to belong to a associated with loss of normal cell division, including humansingle gene family. Proteoglycans have been assigned roles in tumors of the breast [391, ovary [401, and in chemical carcino-cell adhes4on [54], interstitial collagen and fibronectin binding gen-induced rat mammary tumors [411. The structural defects[551, and cell proliferation [56—58]. In several cell systems described include thinning, attenuation, layering and thickeningproteoglycans have been shown to inhibit cell proliferation. In of the basement membrane. Similar descriptions have beenthis context it maybe important to note that both in drug- reported for the basement alterations seen surrounding theinduced cystic development and in human ADPKD cultured renal medullary cysts of nephronophthisis [42], in pathologicalepithelia, reduced structural integrity of the heparan sulfate specimens of human ADPKD [23, 43, 441, and in diphenyithia-proteoglycan component of basement membranes has been zole-induced renal cystic disease in rats [45]. Thickening hasreported [23, 45], and in both conditions this was associated also been described in renal aging and in diabetes [461. with increased epithelial cell proliferation [23, 43, 44, 59]. In the kidney, as in all organs, the basement membraneDespite a rapid incorporation of H2[35S04] into human ADPKD separates epithelia from the underlying interstitium and servesmonolayers, no complete basement membrane structure was several functions, including cell adhesion, tensile strength andsynthesized. Instead, ruthenium red positive spheroids were permselectivity. Normal renal tubule basement membranes aresecreted on the basal surfaces of confluent monolayers. Immu- made up of type IV collagen, and the non-collagenous glyco-noprecipitation studies in normal and ADPKD cultures in which proteins laminin, entactin and heparan sulfate proteoglycansshort-term labelling with 35S-methionine was used as an index [reviewed in 47]. The interstitial components include types I, IIIof core protein synthesis, no differences were found (Fig. 3, and V collagen [48] and their relative abundance is develop-courtesy Dr. J. Hassell, Pittsburgh). This suggests that alter- mentally regulated in the kidney [48, 49]. Morphological alter-ations in HSPG synthesis does not occur at the level of protein ations in basement membranes in renal cysts could derive fromprecursor synthesis, but rather at a later step when sulfated side alterations in any of these extracellular matrix components andchains (GAG) are inserted into the core protein. It remains to be could reflect defects in synthesis, post-translational modifica-determined whether the structural abnormality of ADPKD tions, secretion, extracellular assembly or degradation. Studiesbasement membrane HSPG is due to inappropriate post-trans- of the structural component of type IV collagen, and laminin B Ilational processing or to differential degradation, presumably and B2 chains which influence adhesion, growth and differen-controlled by the action of extracellular proteases [60]. Thus, it tiation in cpk mice, showed that steady state levels of mRNAsis possible to speculate that a reduction in the integrity of the were altered with the progression of RPKD. There was, how-normal spatial deposition of FISPG in the basement membranes ever, no correlation between gene expression as determined byof cystic epithelia could result in some loss of inhibition of cell Northern blot analysis and protein expression as measured byproliferation. immunofluorescence, since basement membrane staining be- Cell growth is under the control of several interactive intra- came weaker and mRNA levels increased in endstage kidneyscellular and extracellular factors. An abnormality in cell growth [50]. Our own findings have also shown that human ADPKDcontrol, such as the limited hyperplasia seen in cyst formation, epithelia continue to synthesize immunoreactive laminin andcould be the result of abnormalities at any of these levels of type IV collagen in vitro and in vivo [43, 44]. It is also of interestcontrol. Extracellular controlling factors fall into two catego- that steady state levels of type IV collagen mRNAs increasedries: insoluble and soluble. The former includes extracellular significantly in ADPKD, but not markedly until the final stagesmatrix proteins as discussed above and the latter includes of disease progression (Fig. 2, courtesy of Dr. J.T. Norman,peptide growth factors. 454 Wilson and Sherwood: Tubulocystic epithelium

— S S HSPG core — • LamininA 200 — a Laminin Bs

Fig. 3. Autoradiogram of immunoprecipitation of laminin A and B chains (lane b) and HSPG core protein (lane c) in confluent cultures of normal proximal tubule (N) and ADPKD (P) epithelia. Cells were labelled with 35S-methionine (1 mCi/mi) for 1 hour at 37°C. Lane a shows control samples incubated with pre-immune serum. Courtesy of Dr. J. Hassell, University of N abc p abc Pittsburgh, School of Medicine.

Growth factor abnormalities insulin act early in G1. Progression factors act later in G1 and Orderly cell division depends on an exquisite coordination ofinclude insulin-like growth factors (IGF-1, IGF-2) and glücocor- events that regulate the cell cycle during which cells progressticoids [67]. Switches in and out of G1 are thought to be the from one mitosis to the next. During this cycle, cells initiated bymajor determinants of post-embryonic cell proliferation, and a mitogenic stimulus pass through: a quiescent phase (G1),therefore, defective control would lead to conditions such as during which RNA and protein synthesis take place; followedhyperplasia. by S phase, during which DNA synthesis takes place; and after Abnormal cell proliferation maybe the result of aberrant DNA replication has stopped, into G2 phase which culminatesexpression of normal mitogenic pathways, the molecular medi- in mitosis [61]. In the normal kidney, rapid cycling of prolifer-ators of which are a family of peptide growth factors that ating cells takes place during certain specific stages of fetalcirculate in the blood or are released locally and interact with organogenesis, but after epithelial cell differentiation into highlytarget cells through specific receptors. These types of growth segmented nephron units is complete, division stops and cellsfactors have three potential modes of action: endocrine, in remain in the prereplicative (G0) state. Thus, in normal adultwhich their action is analogous to that of hormones, in that kidneys, the mitotic index in renal tubules is very low (0.1%),factors released from a distant site exert effects on different cell reflecting the very low level of cell renewal. Renal tubule cellstypes in a remote target tissue, after circulation; paracrine, in are not, however, irrevocably "terminally" differentiated,which growth factors released from one cell type stimulate since they can be stimulated to re-enter the cell cycle andproliferation of different cell types nearby, without involvement complete several rounds of cell division in response to injury inof systemic circulation; and autocrine, in which growth factors vivo or by the addition of serum or specific growth factors inproduced by one cell type have direct mitogenic effects on the vitro [62—64]. same cell [68]. Since a prerequisite for growth factor response is Several factors act in concert to control epithelial cell prolif-occupancy of specific cell surface receptors, an increased eration including: interactions with specific extracellular matrixresponse to normal mitogenic regulators in cystic hyperplasia proteins [65]; the mitogenic influence of cell type specificcould simply be the result of increased receptor number or soluble growth factors; and the expression of cell cycle regu-affinity. Alternatively, since many growth factor receptors lated genes [reviewed in 661. Certain growth factors are knowncontain tyrosine kinase activity, alterations in phosphorylation to act at specific stages of the cell cycle. For instance, platelet-and/or alterations in cellular substrate may be crucial. After derived growth factor (PDGF) and fibroblast growth factorbinding to receptors, a cascade of events is initiated leading to (FGF) exert mitogenic effects on cells early in the cell cycle, atdiverse early cellular responses including increased phosphati- the G0/G1 interface, and are thus termed competence factors.dylinositol turnover, expression of competence genes, activa- "Entry" factors, such as epidermal growth factor (EGF) andtion of phospholipase A2, changes in cell shape and ion changes Wilson and Sherwood: Tubulocystic epithelium 455

[69]. The relationship between these alterations and the laterreceptors. It is also conceivable that these increased levels of event of stimulation of DNA synthesis is unclear. EGF are partially responsible for the inhibition of the hydrooso- Normal renal tubule epithelia possess receptors for a widemotic effects of AVP, as has been demonstrated experimentally variety of growth factors and the kidney is itself, a rich sourcein rabbit CCT [871. It is also of note that EGF has been reported for several growth factors [70]. Therefore, it is possible that into decrease sodium reabsorption [881, which may have some the hyperplasia underlying cyst formation, there might beimpact on fluid accumulation. It would also be of interest to inappropriate growth factor and/or growth factor receptorexamine the occurrence and effects of TGFa in cystic disease, expression in the affected epithelia. Under normal conditions insince this structural homolog of EGF is transiently synthesized the kidney, there is a balance between mitogenic and inhibitoryduring fetal renal development and expression of the gene for growth factors, which presumably normally favors inhibitionboth TGFa and its receptor are frequently associated with since so little cell renewal takes place. An overexpression ofcarcinomas [89]. mitogenic growth factors or loss of sensitivity to inhibitory The overall proliferative response of any cell type in vivo factors are therefore potential mechanisms underlying the inap-reflects the effects of interactions between several growth propriately increased epithelial proliferation in renal cyst en-factors [90]. In the kidney, it is likely that TGF/3 interactions largement. would be important, since under normal conditions this is a Epidermal growth factor (EGF) is a 6.1 kD peptide, whichpotent inhibitor of renal tubule proliferation. In fact, TGF2 acts via interaction with a specific, high-affinity 170 kD receptorwas first isolated from a renal epithelial-derived cell line, BSC-l which possesses intrinsic tyrosine kinase activity [71, 721.[91]. TGF1 and 2 belong to the same gene family [92] and are Overexpression of EGF receptors has been demonstrated inmultifunctional growth factors which exert differential effects several human cancers and tumorigenic cell lines [73—76] and inaccording to the target cell type. They have been shown to be liver regeneration [77]. The normal adult kidney has beenmitogenic for cells of mesodermal origin, including fibroblasts, shown to be a major source of EGF mRNA, although relativelybut inhibit cell division of epithelia. In the normal kidney, as in little peptide is stored there [78]. Immunoreactivity for EGF hasother organs, it has been shown that TGF/3 antagonizes the been detected in distal tubule cells including thick ascendingmitogenic action of EGF, dexamethasone and insulin [83, 85, limbs, distal convoluted tubules, and medullary collecting tu-93]. TGFf3 is a ubiquitous, 25 kD peptide [941 which interacts bules and EGF receptors are paradoxically present on thewith the equally ubiquitous membrane receptors [95]. Little is basolateral membranes of proximal tubule cells. The normalknown about the mechanism of TGFI3 receptors 1961, except role of EGF in the kidney is unclear at present, but it has beenthat they do not possess tyrosine kinase activity [971. A clearly demonstrated as a potent mitogen for primary culturesdisturbance in the normal balance between positive and nega- of rabbit proximal tubules and cortical collecting tubules [63,tive growth regulators, such as TGF/3 could be responsible for 79] and for the proximal and distal tubule-derived cell lines,uncontrolled cell growth as has been demonstrated in human LLC-PK1 and MDCK [80, 81]. Since EGF had been detected inbreast tumor cells [98], retinoblastoma cells [991, and tumor- human breast cyst fluid, although virtually undetectable inpromoter-stimulated mouse epidermis cells [1001. Our studies plasma [82], we were prompted to examine the localization ofon the effects of TGF on ADPKD epithelial proliferation in EGF and its receptor in ADPKD tissues and cultured cells. Thevitro showed a marked loss of inhibitory activity of TGF/3 with mitogenic action of EGF was significantly (5 1%) increased inregard to EGF, dexamethasone and insulin mitogenesis [83, 84], cultured ADPKD epithelial monolayers by comparison to age-and, therefore, suggests a further potential mechanism for the matched control cultures derived from human proximal tubules,enhanced proliferation of ADPKD cyst lining epithelia. It thick ascending limbs or collecting tubules. Examination ofremains to be determined whether the alterations in TGF/3 immunoreactivity showed that not only cyst-lining epithelialsensitivity are due to a decreased number of receptors or cells but also cyst fluids contained large amounts of immuno-decreased proteolytic activation [101, 102]. fluorescence. This was particularly marked in the small cysts in Other growth factors relevant to renal development, includ- the early stages of development of the disease [83, 84]. Thising IGF-l and 2 [103], have been tested for possible discrimi- correlated with radioimmunoassay findings in which detectablenatory proliferative effects in cultures of human normal renal levels of EGF were found in ADPKD and von-Hippel Lindautubule and ADPKD epithelia, but little difference has been cyst fluid contents [851. In addition, EGF has been shown toseen. It was of interest, however, that a significant proliferative exert hyperplastic effects in cystic proximal tubules in mouseresponse to PDGF was found in ADPKD fibroblasts and that metanephric organ culture [86]. This, therefore, suggests thatimmunoreactivity for this peptide was demonstrated in cyst hypersensitivity to EGF can lead to the hyperplasia necessarylining epithelia. This appears to be a case of abnormal paracrine for cystic enlargement in the kidney. In ADPKD, EGF receptorregulation of fibroblast proliferation exerted by tubulocystic immunoreactivity has been demonstrated on apical as well asepithelia, and may correlate with findings that BSC-l cells basal cell surfaces of cyst lining epithelia, and specific bindingsecrete a PDGF-like protein [1041. This may therefore have of '251-EGF was demonstrable both on the apical and basalrelevance with regard to fibrosis, which is another characteris- surfaces of ADPKD epithelia grown to confluence on perme-tic of ADPKD and medullary cystic kidneys. able membrane supports. These results support the notion that part of the hyperplastic response underlying increased cell Cell-cycle regulated gene expression proliferation in cyst formation is an autocrine effect, since in Little is known about the expression of cell-cycle regulated ADPKD, EGF is synthesized by the epithelia cells lining thegenes in the kidney. During embryonic renal development the cysts, secreted apically into the cyst lumen, and is capable ofN-myc proto-oncogene has been demonstrated by in situ hy- interacting with the same cells by virtue of apically locatedbridization to be confined to differentiating epithelial cells [1051. 456 Wilsonand Sherwood: Tubulocystic epithelium

In adult rabbits, during proliferative renal repair after acute Table 3. NaK-ATPase isoforms nephrotoxic injury, transient increases in steady-state mRNA Protein Transcript Chromosome levels were seen: for the proto-oncogenes c-fos, c-myc and a-subunit c-Ha-ras; for the growth-regulated structural protein genes a aA 3 vimentin and f3-actin; and for the transport protein genes a aB 7 NaK-ATPase, ADP-ATP translocase and calcyclin [621. A aC ? single study in the cpk mouse model of RPKD using similar aD 11091 ? techniques has demonstrated an increase in c-myc proto-onco- allI all! [1111 1 [172] gene, but paradoxically this was not seen until late in the/3-subunit /3 1 [172] /32 ? [173] progression of the disease [1061. This was a surprising finding, since c-fos and c-myc are usually associated with proliferation in the early stages. Further studies will be required to charac- terize early response genes at stages of disease developmentinto the endoplasmic reticulum and processed in the Golgi, that correlate with the onset of proliferation. The involvementwhere maturation of the glycosylated /3 subunit involves con- of c-myc in renal cystic development is further implicated byversion of high mannose oligosaccharide side chains. It is recent reports that transgenic mice with the c-myc gene developgenerally suggested that equimolar amounts of a and /3 subunit renal cysts and die prematurely of renal failure [107]. form a2 132complexesinserted into plasma membranes with most of the a subunit on the intracellular side and most of the /3 Fluid transport defects in renal cyst formation subunit on the extracellular side. Tamkun and Fambrough [1201 have also shown that a significant proportion of NaK- Alterations in NaK-ATPase ATPase subunits may remain unassembled and intracellular, NaK-ATPase is a ubiquitous, heterodimeric, integral mem-and may present an additional means for rapid post-transla- brane protein responsible for establishing and maintaining thetional modification. electrochemical gradients for Na and K. In polarized epithe- NaK-ATPase activity in the kidney can be increased ha, such as renal tubules, NaK-ATPase enzyme complexessignificantly by several means, including the steroid hormones are segregated in basolateral membranes only, consistent withaldosterone [121, 122], thyroid hormone [123, 124] and triiodo- its function of vectorial active pumping of Na out of cells intothyronine [124], in remnant kidneys [125] and by increased the blood. In light of its major reabsorptive function, it is notcellular Na concentration [1261. A significant finding by Avner surprising that the kidney is a rich source of NaK-ATPase.et al [124] showed that the induction of NaK-ATPase by Distribution is non-uniform along the nephron and the numbertriiodothyronine in metanephric organ cultures was associated of pump units correlates with the hydrolytic activity of individ-with proximal tubule cyst formation. Subsequent studies have ual nephron segments, being maximal in thick ascending limbsshown that increased NaK-ATPase activity is also a consis- of Henle and distal convoluted tubules (5 x l0 enzyme U/cell),tent property of RPKD in the cpk mouse [86] and in human high in proximal convoluted tubules, lower in proximal straightADPKD tissues [23, 127—129]. Since NaK-ATPase is the tubules and collecting tubules, and barely detectable in thinactive sodium pump enzyme and sodium ions are the major limbs of Henle's loop [108]. osmotic determinant in the kidney, it is not surprising that The NaK-ATPase enzyme complex is made up of twoincreased NaK-ATPase would accompany conditions of dissimilar types of subunit (a and /3). The larger (100 kD) aincreased ion and fluid transport. subunit is a polypeptide which spans the cell membrane at least six times and is the catalytic moiety, bearing the phosphoryla- Membrane polarity defects tion-ATP and ouabain-binding sites. The smaller (55 kD) /3 Normal renal tubule epithehia are highly polarized cells with a subunit is a glycoprotein with a single membrane spanningdifferent complement of proteins in their apical membranes region and is essential for correct insertion of the enzymefrom that in their basolateral membranes [reviewed in 130, 131]. complex into the appropriate cell membrane [109, 110]. EachApical tight junctions separate these two sets of proteins and subunit is synthesized separately as a product of distinct genesprevent them from intermixing within the lipid bilayer of the on separate chromosomes. The a subunit is encoded by a highlyplasma membrane [1321. Establishment of epithehial polarity in conserved multigene family and several isoforms have beenthe kidney occurs early in fetal development [131] and is under detected in various species, including three in humans, thethe control of the laminin A chain [65]. Typically, renal tubule genes for which are located on different chromosomes (Tableepithelia apical brush border membranes are characterized by 3). The expression of these isoforms is developmentally regu-the presence of sodium-coupled transporters, brush border and lated [111, 112] and tissue dependent [109, 113, 114], the acoated pit-specific antigens, and ectoenzymes anchored by isoform predominating in the kidney [1151. Although biochem-covalent links to glcosyl-phosphatylinositol (GPI) including ical data concerning the existence of multiple isoforms of the /3alkaline phosphatase, y-glutamyl transpeptidase and renal subunit are sparse, multiple mRNAs have been isolated indipeptidase [133, 134]. Renal tubule basolateral membranes sheep and rat [116], and Fambrough has demonstrated twocontain receptors for basement membrane proteins, hormone isoforms in chicken [1171. Thus, there is a complex pattern ofand growth factor receptors, the Cl/HC03 exchanger, and NaK-ATPase subunit isoform expression, and this can bethe sodium pump enzyme NaK-ATPase. specifically and differentially modulated by sodium flux and A loss of membrane protein polarity would be catastrophic pressure changes [118, 1191. Like all proteins, NaK-ATPasefor renal tubule function and can be induced experimentally in subunit precursors are synthesized on ribosomes, translocatedmonolayer cultures of renal epithelia by deprivation of extra- Wilson and Sherwood: Tubulocystic epithelium 457 cellular calcium, or in vivo in rat kidney proximal tubules bypermeable to water [37], since 80% of radioactive label had limited ischemic insult [131, 135]. In both cases a mixing ofaccumulated in cyst fluids after five hours. This suggests that membrane proteins was seen, including NaK•ATPase andADPKD cell membranes are different from normal ascending alkaline phosphatase, and this was consistent with a loss in thelimbs of Henle which are impermeable to water, and to collect- integrity of tight junctions. After repair by restoration of normaling tubules which are impermeable in the absence of vasopres- (mM) concentrations of calcium or restoration of blood flow,sin. Since vasopressin has no effect on ADPKD epithelia with newly synthesized proteins were correctly inserted into polar-regard to increasing cAMP second messenger [23], it is unlikely ized membranes. to have any effect on water permeability in cysts. Thus, The precise intracellular mechanisms by which Na K-permeability to water could reflect a proximal tubule origin, or ATPase subunits are sorted and targeted to basolateral mem-conceptually more likely, an increase in membrane fluidity. branes remains obscure, although a role for the expression of /3 It seems that some general properties are characteristic of the subunit in directing the a chain to the membrane is clear. It isfluid expansion observed in cyst development, including in- not known, however, whether fatty acid acylations play a rolecreased NaK-ATPase activity and reversal of vectorial fluid in targeting and docking proteins. Once in the membrane,transport from normal to apical cell surfaces. This capacity for polarity of NaK-ATPase is maintained by interactions withfluid secretion has also been demonstrated in experimentally- membrane-associated proteins which are linked to the actinderived MDCK cell cysts in collagen gels [140]. Another cytoskeleton [136]. It has also been proposed that a complexgeneral feature in MDCK and in metanephric organ culture was interaction with soluble pools of these proteins (fodrin andthat cyst formation could be inhibited by ouabain, further ankyrin) maybe involved in insoluble complex formation [1371.confirming dependence on NaK-ATPase [141]. In ADPKD A dramatic and consistent finding in 20 specimens of ADPKDepithelia in vitro, apically applied ouabain also inhibited kidneys has been that, in cystic epithelia, all immunoreactivityNaK-ATPase and 22NaC1 transport into the apical medium, for the a subunit of NaK-ATPase was localized exclusivelybut basally applied ouabain had no effect, further confirming the on the apical cell membranes, lining the fluid-filled cyst lumensmislocation of NaK-ATPase to the apical membranes of [127—1291. This was shown to be a functionally competentADPKD epithelia [142]. enzyme by cytochemical staining techniques, ouabain sensitiv- ity and biochemical assays on isolated apical and basolateral Cytoskeletal alterations in renal cystic disease membrane vesicles [1291. This defect was seen very early in the The epithelial cell cytoskeleton is comprised of microfila- onset of tubule dilation in ADPKD kidneys which containedments (actin), microtubules (tubulins), and intermediate fila- some cystic and some normal regions. In those kidneys,ments (cytokeratins) [reviewed in 143]. These structural pro- NaK-ATPase location in normal, non-dilated tubules wasteins confer cell-type specific size, shape, and motility. exclusively basolateral [127]. ADPKD monolayers grown onIntermediate filament proteins are highly insoluble and are permeable membrane supports in which apical and basal mediathought to be present largely in filamentous form [144]. Micro- compartments were separated showed apical membrane immu-filaments and microtubules, however, may exist in both soluble no- and enzyme cytochemical staining and vectorial of trans-and insoluble pools and thus contribute to the morphology of port of 22NaCI from basal to apical media which was inhibitablethe cell and to the viscoelastic properties of the cytoplasm by ouabain [129]. Thus, an early defect in cyst formation in[144—146]. Since cellular morphology may change profoundly ADPKD was the reversal of polarized insertion of NaK-with the growth and differentiation state of the cell, it is not ATPase into apical membranes with consequent redirection ofsurprising that the nature of the cytoskeletal proteins expressed net Nat-transport from the serosa into the tubule lumen.in a cell together with their localization and associations with Similar apical NaKtATPase is found in the choroid plexusbinding proteins may change. where it is associated with the vectorial transport of sodium In polarized cells, microfilaments and intermediate filaments from the plasma into the cerebrospinal fluid [138]. However, inare major components inserting into intercellular junctions, and this case there is also a reversal of polarity of several otherhence are essential for the maintenance of membrane polarity membrane proteins, such as Na/H antiporter to the basal[143, 147]. Actin is also a major component of the cortical surface and Ca2-ATPase in the apical membrane surface. Thiscytoskeleton, the filamentous web observed just beneath the is not the case in ADPKD epithelia where the Nat/H fantiportplasma membrane, and is particularly prominent in the terminal remained apical, and the anion exchanger and basement mem-web, at the apical aspect of epithelial cells with a brush border, brane proteins remained basal in location. A survey of severalsuch as renal proximal tubules [147]. Actin appears to play a polarized markers led to the conclusion that there was a specificrole in anchoring integral membrane proteins, including en- mislocation of NaK-ATPase to the apical membranes ofzymes and transporters, via interactions with ankyrin, fodrin cystic epithelia, rather than a complete reversal of cell polarity.and related proteins [136, 137, 148, 149]. Cytokeratins share the This was confirmed morphologically by the presence of apicalrole of establishing intercellular junctions with microfilaments, tight junctions and rudimentary microvilli and basal laminin,and recent experiments indicate that they may also bind ankyrin type IV collagen and HSPG immunostaining [127, 129]. Anor fodrin [144]. Convincing evidence implicates microtubules in additional example of apical NaK-ATPase has been seen inthe transport of organelles and in the delivery of apical mem- association with four cases of human arachnoid cysts [139], andbrane proteins to the correct domain [150—154]. may suggest this as a general mechanism associated with fluid The importance of the cytoskeleton in renal cyst formation is secretion. just beginning to be explored. As discussed above, renal cyst It has been demonstrated directly, by infusion of 3H-waterformation requires expansion of tubules by limited hyperplasia. into ADPKD patients, that ADPKD cyst walls are highlyThe observed defects in basement membranes in cystic epithe- 458 Wilson and Sherwood: Tubulocystic epithelium lia might have impact on, or even be directed by the actinlining epithelia and cyst fluid of expanding cysts, as discussed cytoskeleton via intracellular-extracellular interactions [148,above. 149, 155]. Cells with a proliferative defect share some proper- The fluid accumulation defect in cystic epithelia, in which ties with the transformed phenotype of tumor cells including aaltered targeting and/or anchoring of membrane transport pro- loss of specialized ultrastructure and aberrant basement mem-teins appear to play a role, also implicate modifications in the brane attachments. A direct association between growth abnor-cytoskeleton. The importance of intact cytoskeletal elements in malities and cytoskeletal properties in cystic epithelia remainsnormal cellular processes, including secretion, is clear. Thus, to be explored. there could be a fundamental role for cytoskeletal abnormalities The role of microtubules in cell division is well characterizedin the initiation or maintenance of tubulocystic disease in which while the function of microfilaments and intermediate filamentsseveral cellular processes are perturbed. is not so clearly understood. The expression of intermediate As described above, Nelson et a! have shown that the filament protein genes is highly cell-type specific [144, 145].NaK-ATPase is anchored in the basolateral membrane of Cytokeratins are the major intermediate filament proteins foundcultured kidney cells (MDCK cells) to the actin cytoskeleton in epithelial cells while vimentin predominates in cells ofvia ankyrin and fodrin [136, 137]. Microtubules do not appear to mesenchymal origin [144]. Although some renal epithelia are inhave a direct effect on the targeting or anchoring of basolateral fact derived from embryonic mesoderm, fully differentiatedmembrane proteins, but microtubular disruption by coichicine tubule cells express cytokeratins typical of their epithelialand vinblastine has suggested their integrity is essential for nature [156—158]. It is of interest, however, that coexpressiontransport of apically destined proteins to the correct plasma rats treated with daunomycin showed coexpression of cytoker-membrane domain [150—154], although the possibility that the atins and vimentin in renal epithelia, which was correlated witheffects of these agents were due to perturbation of the Golgi could not be ruled out. An analogous process appears to act in the degree of cystic expansion and proliferation [158]. How- ever, coexpression of these proteins was also observed inthe delivery of secretory vesicles to the apical surface of polarized cells, and calpactin I has recently been implicated in several other diseases and may represent a non-specific re-the fusion process required for regulated secretion events [167]. sponse to tubule injury. The possibility that apically-destined membrane proteins are The actin cytoskeleton also responds to the growth state ofdelivered via the same processes as secretory granules remains the cell. Expression of different isoforms of actin can beto be established, Thus, it can be speculated that the specific induced in cells treated with tumor promoters, agents whichmistargeting of NaK-ATPase to the apical plasma membrane facilitate but cannot initiate neoplastic transformation [159].in ADPKD cyst enlargement, might be the result of alterations Although actin isoform expression has not yet been character-of interaction with those cytoskeletal proteins responsible for ized in renal development, a novel isoform is transiently ex-NaK-ATPase intracellular targeting or anchoring. Alterna- pressed in intestinal epithelia from neonatal mice [160]. Thetive possibilities remain to be explored including: possible significance of multiple actin isoforms, which differ only by amodifications in fatty acid acylation targeting and anchoring; few amino acids [161], is not fully understood, but it could bealtered Golgi processing of NaK-ATPase subunits; and al- that different isoforms have different affinities for specificterations at the level of the gene NaK-ATPase a and /3 binding proteins. Our studies have shown that actin is exces-subunits. sively abundant in epithelia from ADPKD renal tissue [128, 162]. The altered actin content appears to result from a post- Conclusions and perspectives translational change or a change in isoform expression, since The evidence presented suggests that numerous cellular differences in mRNA levels for f3-actin in ADPKD cells cannotchanges can occur during the initiation and progression of renal account for the excessive amounts of actin observed [129].cystic diseases of genetic and non-genetic origin. At least two mRNA levels for /3-actin have also been measured in recessivebasic criteria are necessary for cystic expansion of renal tu- PKD in cpk mice and no significant differences were detectedbules: aberrant epithelial cell proliferation and lumenal fluid [50]. In addition to the increased abundance of actin protein inaccumulation. Epithelial cell growth can be regulated by insol- ADPKD cells, immunoperoxidase and fluorescence studiesuble factors in the extracellular matrix and by soluble peptide have shown colocalization with apical NaK-ATPase in cystfactors. Both mechanisms are implicated in renal cystic devel- lining epithelia. Visualization of ADPKD cyst epithelial actinopment since abnormalities of heparan sulfate proteog!ycan filaments with rhodamine-conjugated phalloidin also showed amatrix deposition and decreased sensitivity to TGFI3 have been consistent codistribution with the calpactin II (lipocortin I), thedemonstrated, thus potentially reducing inhibitory regulation of calcium-phospholipid actin binding protein [163]. In addition,epithelial proliferation. In addition, hypersensitivity to the calpactin II is excessively abundant in ADPKD [162]. Thisepithelial mitogen, EGF has been found in genetic polycystic observation is of particular interest because calpactin II is alsokidney disease, where it is thought to be an autocrine factor and a major substrate for the tyrosine kinase activity of the epider-is associated with the presence of aberrant apical membrane ma! growth factor receptor [1641. The phosphorylation of thisreceptors and increased protein amounts of the intracellular protein correlates with the activation of phospholipase A2 andEGF receptor tyrosine kinase substrate, lipocortin. Increased subsequent release of plasma membrane arachidonate [165,NaK-ATPase activity is also a common feature of renal 166]. Our hypothesis is that the over-abundance of calpactin IIepithelial cysts of genetic and non-genetic origin and is associ- may be an amplification response to prolonged stimulation ofated with the observed fluid accumulation. In human autosomal the EGF pathway in ADPKD epithelia, since EGF is a potentdominant polycystic kidney disease (ADPKD), a specific rever- mitogen for these cells and is present in large quantities in cystsal of polarity of NaK-ATPase location has been demon- Wilson and Sherwood: Tubulocystic epithelium 459 strated in the apical membranes of cyst lining epithelia, and 4. DUNNILL MS, MILLARD PR, OLIvER D: Acquired cystic disease shown to be responsible for basal to apical transport of Na and of the kidneys: A hazard of long-term intermittent maintenance can be inhibited by apically applied ouabain. hemodialysis. J C/in Pathol 30:868-.877, 1977 5. HESTBECH J, HANSEN HE, AMDISEN A, OLSEN S: Chronic renal Little information is available concerning the sequence of lesions following long-term treatment with lithium. Kidney mt events necessary to result in the phenotypic expression of renal 12:205—213,1977 tubule cysts. However, it is clear from studies on early stages of 6.GARDNER KD, SOLOMON 5,FITZGERRELWW, EvAN AP: Func- development of RPKD in cpk mice and in ADPKD kidneys with tion and structure in the diphenylamine-exposed kidney. J Cl/n > 50% morphologically-normal renal tubules, that the enhanced Invest 57:796—806, 1970 7.PREMINGERGM, KOCH WE, FRIED FA, MCFARLAND E, MURPHY activity and reversed membrane polarity of NaK-ATPase is E, MANDELL J: Murine congenital polycystic kidney disease: A an early event. The presence of immunoreactive EGF in the model fur studying development of cystic disease. J Urol 127:556— fluid of small cysts in early stage ADPKD kidneys also suggest 560, 1982 autocrine EGF production might be an early modification. By 8.AVNERED, STUDNICKI FE, YOUNG MC, SWEENEY WE, PIESCO NP, ELLIS D, FETTERMAN GH: Congenital murine polycystic contrast, alterations in the type IV collagen components of kidney disease. 1. The ontogeny of tubular cyst formation. Pediatr cystic basement membranes appear to be later events. The Nephrol1:587—596,1987 proliferative and fluid transport defects appear to be coordi- 9. MARTINEZ-MALDONADO M, YIUM JJ, EKNOYANG,SuKI WN: nated in cyst formation, and must be a different process from Adult polycystic kidney disease: Studies of the defect in urine the hyperplasia of tumor formation in which rapid uncontrolled concentration. Kidneymt2:107—113, 1972 10. GABOW PA, IKLE DW, HOLMES JH: Polycystic kidney disease. growth leads to solid cell masses. Nevertheless, complete Progressive analysis of nonazotemic patients and family members. analysis of expression of growth related genes, including pro- Ann Intern Med 101:238—247,1984 tooncogenes during cystic development, would be instructive. 11.BAERT L: Hereditary polycystic kidney disease (adult form): A It maybe speculated that one of the observed defects in cystic microdissection study of two cases at an early stage of the disease. epithelia could be a primary change to which all the other Kidneymt 13:519—525, 1978 12.LAMBERTPP: Polycystic diseaseof the kidney. ArchPathol manifestations are linked in a secondary fashion. For instance, 44:34—58, 1947 there are well documented relationships between the actin- 13. POTTER EL: Normal and Abnormal Development of the Kidney. based cytoskeleton and NaK4-ATPase polarity [136, 137], Chicago, Year Book, 1972 secretion [148—1521, matrix [153—155, 1681 and growth factor 14. FARAGGLANA T, BERNSTEIN J, STRAUSS L, CHIRG J: Use of lectins in the study of histogenesis of renal cysts. Lab Invest 53:575—579, receptor substrate interactions [1631. In addition, there is evi- 1985 dence of tight interactions between matrix proteins and growth 15. GARDNER KD: Composition of fluid in twelve cysts of a polycystic factors [169—171], matrix and polarity [172—1731 and intracellu- kidney. N Engl J Med 28 1:985—988, 1969 lar and extracellular Na ion contents and proliferation [174, 16. HUSEMAN R, GRADY A, WELLING D, GRANTHAM JJ: Macropunc- 175]. Until more evidence has been collected in normal renal ture study of polycystic kidney disease in adult human kidneys. Kidneymt 18:375—385, 1980 tubule epithelia, however, the primary defect will remain elu- 17.GRANTHAM JJ: Polycystic kidney disease: a predominance of sive. The sobering possibility must also be considered that none giant nephrons. Am J Physiol 244:F3—Fl0, 1983 of the changes documented to date is the underlying cause, but 18. REEDERS ST, BREUNING MH, DAVIES KE, NICHOLLS RD, JAR- merely reflect the effects of cystic enlargement. A search for MANAP,HIGGS DR, REARSONPL,WEATHERALLDi:A highly polymorphic DNA marker linked to adult polycystic kidney dis- alterations at the fundamental levels of signal transduction and ease on chromosome 16. Nature 317:542—544, 1985 transcription factors remains to be explored. 19. KIMBERLINGWi,FAINPR,KENYON JB, GOLDGAR D, SUJANSEY E,GABOWPA: Linkage heterogeneity of autosomal dominant polycystic kidney disease. N Engl J Med 319:913—918, 1988 Acknowledgments 20.ROMEOG, DavoTo M, COSTA G, RoNCuzzi L, CATIZONE L, This work was supported in part by NIH R01 DK 40698 and P01 DK ZUCCHELLI P, GERMINO GG, KEITH T, WEATHERALLDi,REED- 34039. Tissue was procured by National Disease Research Interchange ERS ST: A second genetic locus for autosomal dominant polycystic and International Institute for the Advancement of Medicine, Philadel- kidneydisease. LancetJuly 2,8—10, 1988 phia, Pennsylvania, USA. The authors are indebted to Dr. J.T. Norman 21.NORBY S, SORENSEN AWS, BOESEN P: Non-allelic genetic heter- and Dr. J. Hassell for their generous contributions of experimental ogeneity of autosomal dominant polycystic kidney disease? in effort and discussion, to Eugene Guerra for technical assistance and to Genetics of Kidney Disorders. Prog C/in Biol Res 305:83—88, 1989 Mrs. Gloria Trechock in typing the manuscript. Additional thanks are 22. CUPPAGE FE, HUSEMANRA,CHAPMAN A, GRANTHAM JJ: Ultra- also due to Dr. C.R. Burrow, whose detailed sleuthing and critical structure and function of cysts from human adult polycystic evaluations were not only stimulating but gave an added dimension to kidneys. Kidney Irn 17:372—381,1980 the discussion. 23. WILSONPD, SCHRIERRW,BRECKONRD,GABOw PA: A new method for studying human polycystic kidney disease epithelia in culture. Kidneymt 30:371—378, 1986 Reprint requests to Patricia D. Wilson, M.D., Department of Phys- 24.BERNSTEIN J, EVAN AP, GARDNER KD: Epithelial hyperplasia in iology and Biophysics, UMDNJ-Robert Wood Johnson Medical human polycystic kidney diseases. Am JPathol 129:92—101, 1987 School, 674 Hoes Lane, Piscataway, New Jersey 08854-5635, USA. 25. WILSONPD,DILLINGHAM MA, BRECKON R, ANDERSONRi: Defined human renal tubular epithelia in culture: Growth, charac- terization and hormonal response. Am J Physiol 248:F436—F443, References 1985 26.VANPUTTEN V,PRZEKWA5 i, GABOW P. SCHRIER RW, GRANOT 1. MCKUSICK V: Mendelian inheritance in man. Y: Parathyroid hormone and vasopressin resistance in autosomal 2.BERNSTEINJ: The classification of renal cysts. Nephron11:91— dominant polycystic kidney disease epithelial cells. (abstract) Am 100, 1973 Soc Nephrol, 1989, p. l92A 3.DALGAARDOZ: Polycystic disease of the kidneys, in Disease of 27. GRAHAM PC, LINDOP GBM: The anatomy of the renin-secreting the Kidney, edited by STRAUSS MB, WELT LG, Boston, Little cell in adult polycystic kidney disease. Kidney mt33:1084—1090, Brown & Co., 1963, pp. 1223—1258 1988 460 Wilsonand Sherwood: Tubulocystic epithelium

28. MCATEER JA, CARONE FA, GRANTHAM JJ, KEMPSON ST, GARD- basement membrane proteoglycans are derived from a large NER KD, EVAN AP: Explant culture of human polycystic kidney. precursor. J Cell Biol 106:963—970, 1988 Lab Invest 59:126—136, 1988 53. STOW JL, FARQUHAR MG: Distinctive populations of basement 29. WILSON PD: Characteristics of microdissected normal human membrane and cell membrane heparan sulfate proteoglycans are renal tubules and polycystic kidney epithelia grown in culture. produced by cultured cell lines. J Cell Biol 105:529—539, 1987 Proc MDRJ 2nd mt Conference, The Use of Human Tissue and 54. COUCHMAN JR. AUSTRIA R, WOODS A, HUGHES RC: Adhesion Organs for Research, 1985, pp. 151—157 defective BHK mutant has cell surface heparan sulfate proteogly- 30. EVAN AP, GARDNER KD: Nephron obstruction in nordihydroguai- can of altered properties. J Cell Physiol 136:226—236, 1988 aretic acid-induced renal cystic disease. Kidney mt 15:7—19, 1979 55. SAUNDERS S, BERNFELD M: Cell surface proteoglycan binds 31. BJERLE P, LINDQUIST B, MICHAELSON G: Pressure measurements mouse mammary epithelial cells to fibronectin and behaves as a in renal cysts. ScandJ Lab C/in Invest 27:135—138, 1971 receptor for interstitial matrix. J Cell Biol 106:423—430, 1988 32. CARONE FA, ROWLAND RG, PERLMAN SG, GAN0TE CE: The 56. YAMAGUCHI Y, RUOSLAHTI E: Expression of human proteoglycan pathogenesis of drug-induced renal cystic disease. Kidney mt in Chinese hamster ovary cells inhibits cell proliferation. Nature 5:411—421, 1974 336:244—246,1988 33. GRANTHAM JJ, DONOSO VS, EVAN AP, CARONE FA, GARDNER 57. HAMATI HF, BRITTON EL, CAREY DJ: Inhibition of proteoglycan KD: Viscoelastic properties of tubule, basement membranes in synthesis alters extracellular matrix deposition, proliferation and experimental renal cystic disease. Kidney mt 32:187—197, 1987 cytoskeletal organization of rat aortic smooth muscle cells in 34. EVAN AP, GARDNER KD, BERNSTEIN J: Polypoid and papillary culture. J Cell Biol 108:2495—2505, 1989 epithelial hyperplasia: A potential cause of ductal obstruction in 58. CASTELLOT JJ, ADDONIZIO ML, ROSENBERG R, KARNOVSKY MJ: adult polycystic disease. Kidney mt 16:743—750, 1979 Cultured endothelial cells produce a heparin-like inhibitor of 35. WELLING LW, WELLING D: Theoretical models of cyst formation smooth muscle cell growth. J Cell Biol 90:372—379, 1981 CARONE and growth. Scanning Microsc 2:1097—1102, 1988 59. FA, HOLLENBERG PF, NAKAMURA S, PUNYARIT P, 36. GRANTHAM JJ, GEISER JL, EVAN AP: Cyst formation and growth GLOGOWSKI W, FLOURET G: Tubular basement membrane change in autosomal dominant polycystic kidney disease. Kidneymt occurs pan passu with the development of cyst formation. Kidney lnt35:1034—1040,1989 31:1145—1152, 1987 60. WERB Z, DINGLE JT, REYNOLDS JJ, BARRETT AJ: Proteoglycan- 37. JACOBSSON K, LINDQVIST B, MICHAELSON G, BJERLE P: Fluid degrading enzymes of rabbit fibroblasts and granulocytes. Bio- turnover in renal cysts. Acta Med Scand 202:327—329, 1977 chemJ173:949—958, 1978 38. EKBLOM P, VESTWEBER D, KEMLER R: Cell-matrix interactions 61. BASERGA R: The cell cycle. N Engl J Med 304:453—459, 1981 and cell adhesion during development. Ann Rev Cell Biol 2:27—47, 62. NORMAN iT, BOHMAN R, FISCHMANN G, BOWEN JW, MCDON- 1986 OUGH A, SLAMON D, FINE L: Patterns of mRNA expression 39. HALTER SA, GLICK AD: Basement membrane changes in human during early cell growth differ in kidney epithelial cells destined to breast tissue in explant culture: A morphometric and ultrastruc- undergo compensatory hypertrophy versus regenerative hyperpla- tural study. Ultrastruct Pathol 4:23—34, 1983 sia. Proc NatI Acad Sd USA 85:6768—6772, 1988 40. STENBACH F, WA5ENIUS V: Basement membrane structures in 63. WILSON PD, HORSTER M: Differential response to hormones of tumors of the ovary. EurJ Obstet Gynec ReprodBiol20:357—371, defined distal nephron epithelia in culture. Am J Physiol 24:C 166— 1985 C174, 1983 41. ORMEROD EJ, WARBURTON MJ, GUSTERSON B, HUGHES CM, 64. WILSON PD, ANDERSON RJ, BRECKON RD, NATHRATH WJ, RUDLAND PS: Abnormal deposition of basement membrane and SCHRIER RW: Retention of differentiated characteristics by cul- connective tissue components in dimethylbenzanthracene-in- tures of defined rabbit kidney epithelia. J Cell Physiol 130:245— duced rat mammary tumours: An immunocytochemical and ultra- 254, 1987 structural study. Histochem J 17:1155—1166, 1985 65.EKBLOMP: Developmentally regulated conversion of mesen- 42. COHEN AH, H0YER JR: Nephronophthisis a primary tubular chyme to epithelium. FASEB J 3:2141—2150, 1989 basement membrane defect. Lab Invest 55:564—572, 1986 66. RITTLING SR, BASERGA R: Expression of growth-related genes in 43. Wnso PD, HRENIUK D, GABOW PA: Relationship between senescent human diploid fibroblasts, in Growth Control During abnormal extracellular matrix and escessive growth of human CellAging,edited by WANG E, WARNER HR, New York, CRC adult polycystic kidney disease epithelia. J Cell Physiol (in press) Press, 1989, pp. 19—29 44. WILSON PD, HRENIUK D, GABOW PA: Influence of extracellular 67. PARDEE AB: G1 events and regulation of cell proliferation. Sci- matrix components on the enhanced growth response of polycys- ence 246:603—608,1989 tic kidney disease epithelia. (abstract) J Cell Biol 107:796, 1988 68. SPORN M, ROBERTS AB, WAKEFIELD LM, ASSOIAN R: Trans- 45. KANWAR YS, CARONE FA: Reversible changes of tubular cell and forming growth factor /3: Biological function and chemical struc- basement membrane in drug-induced renal cystic disease. Kidney ture. Science233:532—534,1986 mt 26:35—43, 1984 69. ROZENGURT E: Early signals in the mitogenic response. Science 46. HALEY DP, BULGER RE: The aging male rat: Structure and 234:162—166, 1986 function of the kidney. Ann JAnat 167:1—13, 1983 70. MENDLEY SR, TOBACK FG: Autocrine and paracrine regulation of 47. MARTINEZ-HERNANDEZ A, AMENTA A: The basement membrane kidney epithelial cell growth. Ann Rev Physiol 51:33—50, 1988 in pathology. Lab Invest 48:656-677, 1983 71. CARPENTER G, ZENDEGUI JG: Epidermal growth factor, its recep- 48. EKBLOM P, LEHTONEN E, SAXEN J, TLMPL R: Shift in collagen tor and related proteins. ExpCell Res164:1—10, 1986 type as an early response to induction in the metanephric mesen- 72. DEFIZE LHK, BOONSTRA J, MEISENHELDER J, KRUIJER W, chyme. J Cell Biol 89:267—283, 1981 TERTOOLEN LGJ, TILLY BC, HUNTER T, VAN BERGEN PM, 49. MARTINEZ-HERNANDEZ A, GAY S, MILLER EJ: Ultrastructural MOOLENAAR WH, DE LAAT SW: Signal transduction by epidermal localization of type V collagen in rat kidney. J Cell Biol 92:343— growth factor occurs through the subclass of high affinity recep- 349, 1982 tors. J Cell Biol 109:2495—2507, 1989 50. EBIHARA I, KILLEN PD, LAURIE GW, HUANG T, YAMADA Y, 73. FABRICANT RN, DELARCO JE, TODARO GJ: Nerve growth recep- MARTIN GR, BROWN KS: Altered mRNA expression of basrnent tors on human melanoma cells in culture. Proc NatI Acad Sci USA membrane components in a murine model of polycystic kidney 74:565—569, 1977 disease. Lab Invest 58:262—269, 1988 74. HOLLSTEIN MC, SMITS AM, GALIANA C, YAMASAKI H, BoS JL, 51. HAsSELLJR, LEYSHON WC, LEDBETTERSR, TYREEB, SUZUKI S, MANDARD A, PARTENSKY C, MONTESANO R: Amplification of KATO M, KIMATA K, KLEINMAN H: Isolation of two forms of epidermal growth factor receptor gene but no evidence of ras basement membrane proteoglycans. J Biol Chem 260:8098—8105, mutations in primary human oesophageal cancers. Cancer Res 1985 48:5119—5123, 1988 52. KLEIN DJ, BROWN DM, OEGEMA TR, BRENCHLEY PE, ANDER- 75. Ro J, NORTH SM, GALLICK GE, HORTOBAGYI GN, GUTTERMAN SON JC, DICKINSON MAJ, HORIGAN EA, HASSELL JR: Glomerular JU, BUCK M: Amplified and overexpressed epidermal growth Wilson and Sherwood: Tubulocystic epithelium 461

factor receptor gene in uncultured primary human breast carci- Some recent advances in the chemistry and biology of transform- noma. CancerRes48:161—164, 1988 ing growth factor-beta. J Cell Biol 105:1039—1045, 1987 76. DI FloRa PP, PIERCE JH, FLEMING TP, HAZAN R, ULLRICH A, 98. KNABBE C, LIPPMAN ME, WAKEFIELD LM, FLANDERS KC, KING AR, SCHLESSINGER J, AARONSON SA: Overexpression of KASID A, DERYNCK R, DICKSON RB: Evidence that transforming the human EGF receptor confers an EGF-dependent transformed growth factor-/3 is a hormonally regulated negative growth factor phenotype to NIH 3T3 cells. Cell 51:1063—1070, 1987 in human breast cancer cells. Cell 48:417—428, 1987 77. JOHNSON AC, GARFIELD SH, MERLINO GT, PASTAN I: Expres- 99. KIMCHI A, WANG X, WEINBERG RA, CHEIFETZ 5, MASSAGUE J: sion of epidermal growth factor receptor proto-oncogene mRNA Absence of TGF-/3 receptors and growth inhibitory responses in in regenerating rat liver. Biochem Biophys Res Commun 150:412— retinoblastoma cells. Science 240:196—198, 1988 418, 1988 100. AKI-IURST RJ, FEE F, BALMAIN A: Localized production of TGF-13 78. FISHER DA, SALIDO EC, BARAJAS L: Epidermal growth factor and mRNA in tumour promoter-stimulated mouse epidermis. Nature the kidney. Annu Rev Physiol 5 1:67—80. 1989 331:363—365, 1988 79. NORMAN J, BADIE-DEZFOOLY B, NORD E, KURTZ I, SCHLOSSER 101. O'CoNNoR-McCOURT MD, WAKEFIELD LM: Atent transforming J, CHAUDHARI A, FINE L: EGF-induced mitogenesis in proximal growth factor-$ in serum. J Biol Chem 262:14090-14099, 1987 tubular cells: Potentiation by angiotensin II. Am J Physiol 253: 102. LYONS RM, KESKI-OJA J, MOSES HL: Proteolytic activation of F299—F309, 1987 latent growth factor-$ from fibroblast-conditioned media. J Cell 80. MULLIN JM, MCGINN MT: Epidermal growth factor-induced Biol 106:1659—1665, 1988 mitogenesis in kidney epithelial cells (LLC-PKI). Cancer Res103. HAMMERMAN MR: The growth factor hormone-insulin-like 48:4886—4891, 1988 growth factor axis in the kidney. Am J Physiol 257:F503—F514, 81. GOODYER PR, KACHRA Z, BELL C, ROZEN R: Renal tubular cells 1989 are potenetial targets or epidermal growth factor. Am J Physiol 104. KARTHA S, BRADHAM DM, GROTENDORST GR, TOBACK FG: 255:Fll9l—F1l96, 1988 Kidney epithelial cells express c-sis protooncogene and secrete 82. BOCCARDO F, VALENTI G, ZANARDI 5, CERRUTI G, FASSIO T, PDGF-like protein. Am J Physiol 255:F800—F806, 1988 BRUZZI P, DE FRANCHIS V, BARRECA A, DEL MONTE P, MINUT0 105. HIRVONEN H, SANDBERG M, KALIMO H, HUKKANEN V, Vuoiio F: Epidermal growth factor in breast cyst fluid: Relationship with E, SALMI TT, ALITALO K: The N-myc-proto-oncogene and IGF-II intracystic cation and androgen conjugate content. Cancer Res growth factor mRNAs are expressed by distinct cells in human 48:5860—5863, 1988 fetal kidney and brain. JCell Biol108:1093—1104, 1989 83. WILSON PD, Du J, GUERRA E, NORMAN iT: Autocrine, endocrine 106. COWLEY BD, SMARDO FL, GRANTI-IAM JJ, CALVET JP: Elevated and paracrine regulation of growth abnormalities in autosomal c-myc protooncogene expression in autosomal recessive polycys- dominant polycystic kidney disease. FASEB J (submitted for tic kidney disease. ProcNail Acad Sci USA 84:8394—9398,1987 publication) 107. TRUDEL M, D'AGATI V, CONSTANTINI F: The c-myc oncogene 84. DAS K, GABOW PA, WILSON PD: Abnormal growth factor re- induces kidney cysts in transgenic mice. (abstract) Kidney mt sponse by human adult polycystic kidney disease epithelia in vitro. 35:364, 1989 (abstract) Kidney mt 35:225, 1988 108. KATZ AL: Overview: Role of Na-K-ATPase in kidney function, in 85. MOSKOWITZ D, BONAR SL, MARCUS MD, CLAYMAN RV, AVNER The Na,K-Pump, Part B, Cellular Aspects, New York, Alan ED: Epidermal growth factor content of human and mouse (cpk) Liss Inc., 1988, pp. 207—232 renal cyst. (abstract) Clin Res 37:497A, 1989 109. SHULL MM, LINGREL JB: Multiple genes encode the human 86. AVNER ED, SWEENEY WE: Epidermal growth factor induces Na,K-ATPase catalytic subunit. Proc Nail Acad Sd USA hyperplastic tubular cysts and increased Na-K-ATPase activity in 84:4039—4043, 1987 vitro. (abstract) Am Soc Nephrol 1989, p. 24 110. GEERING K: Overview: Biosynthesis, membrane insertion and 87. BREYER MD, JACOBSON HR, BREYER JA: Epidermal growth maturation of Na,K-ATPase, in The Na,K-Pump, Part B, factor inhibits the hydroosmotic effect of vasopressin in the Cellular Aspects, New York, Alan Liss Inc., 1988, pp. 19—33 isolated perfused rabbit cortical collecting tubule. J Clin Invest Ill. SVERDLOV ED, AKOPYANZ NS, PETRUKHIN KE, BROUDE NE, 82:1313—1320, 1988 MONASTYRSKAYA OS, MODYANOV NN: Na,K-ATPase: Tissue 88. VEHASKARI VM, HERING-SMITH KS, MosKowITz DW, WEINER specific expression of genes coding for a-subunit in diverse human ID, HAMM LL: Effect of epidermal growth factor on sodium tissues. FEBSLett239:65—68, 1988 transport in the cortical collecting tubule. Am J Physiol 256:F803— 112. HERRERA VL, EMANUEL JR, RUIZ-OPAZO N, LEVENSON R, F809, 1989 NADAL-GINARD B: Three differentially expressed Na,K-ATPase 89. DERYNCK R: Transforming growth factor a. Cell 54:593—595, 1988 a subunit isoforms: Structural and functional implications. JCell 90. SPORN MB, ROBERTS AB: Peptide growth factors are multifunc- Biol105:1855—1865,1987 tional. Nature 332:217—219, 1988 113. EMMANUEL JR, GARETZ 5, STONE L, LEVENSEN R: Differential 91. TUCKER RF, SHIPLEY GD, MOSES HL, HOLLEY RW: Growth expression of Na,K-ATPase a and /3 subunit mRNAs in rat inhibitor from BSC-l cells closely related to platelet type b tissues and cell lines. ProcNat/Acad Sci USA 84:9030—9034,1987 transforming growth factor. Science 226:705—707, 1984 114. SHYJAN AW, LEVENSEN R: Antisera specific for the al, a2, a3 92. DUKE P, HANSEN P, IWATA KK, PIELER C, FOULKES JG: Iden- and /3 subunits of the Na,K-ATPase: Differential expression of a tification of another member of the transforming growth factor and /3 subunits in rat tissue membranes. Biochem 28:4531—4535, type f3 gene family. Proc Nod Acad Sci USA 85:4715—4719, 1988 1989 93. MULDOON LL, RODLAND KD, MAGUN BE: Transforming growth 115. MERCER RW, SCHNEIDER JW, BENZ EJ: Molecular cloning and factor f3 modulates epidermal growth factor-induced phosphoi- characterization of a-subunit isoforms of the Na,K-ATPase, in nositide metabolism and intracellular calcium levels. J Biol Chem The Na,K-Pump, Part B, Cellular Aspects, New York, Alan 263:5030—5033, 1988 Liss Inc, 1988, pp. 119—126 94. FLANDERS KC, THOMPSON NL, CISSEL DS, VAN OBBERGHEN- 116. YOUNG RM, SHULL GE, LINGREL JB: Multiple mRNAs from rat SCHILLING E, BAKRE CC, KASS ME, ELLINGSWORTH LR, RoB- kidney and brain encode a single Na,K-ATPase /3 subunit ERTS AB, SPORN MB: Transforming growth factor-/31: Histochem- protein. J Biol Chem 262:4905—4910, 1987 ical localization with antibodies to different epitopes. J Cell Biol 117. FAMBROUGH DM, BAYNE EK: Multiple forms of (Na K)- 108:653—660, 1989 ATPase in the chicken. JBiol Chem 258:3926—3935,1983 95. CHEIFETZ S, LIKE B, MASSAGUE J: Cellular distribution of type I 118. BOWEN JW, MCDONOUGH A: Pretranslational regulation of Na- and type II receptors for transforming growth factor /3. JBiol K-ATPase in cultured canine kidney cells by low K. Am J Chem 261:9972—9978, 1986 Physiol 252:C179—C189, 1987 96. BOYD FT, MASSAGUE J: Transforming growth factor-$ inhibition 119. HERRERA VL, CHOBANIAN AV, RUIZ-OPAZO N: Isoform-specific of epithelial cell proliferation linked to the expression of a 53-kDa modulation of NaK-ATPase a-subunit gene expression in hy- membrane receptor. JBiol Chem 264:2272—2278,1989 pertension. Science241:221—230,1988 97. SPORN MB, ROBERTS AB, WAKEFIELD LM, DE CROMBROGGHE B: 120. TAMKUN MM, FAMBROUGH DM: The (Na=K)-ATPase of chick 462 Wilsonand Sherwood: Tubulocystic epithelium

sensory neurons. Studies on biosynthesis and intracellular trans- 143. COUDRIER E, KERJASCHKI D, LOUVARD D: Cytoskeleton organi- port.JBiol Chem 261:1009—1019, 1986 zation and submembranous interactions in intestinal and renal 121. DOUCET A: Function and control of Na-K-ATPase in single brush borders. Kidney mt 34:309—320, 1988 nephron segments of the mammalian kidney. Kidney ml 34:749— 144. STEINERT PM, RooP DR: Molecular and cellular biology of 760, 1988 intermediate filaments. Ann Rev Biochem 57:593—626, 1988 122. VERREY F, SCFIAERER E, ZOERKLER P, PACCOLAT MP, GEERING 145. POLLARD TD, COOPER JA: Actin and actin-binding proteins. Ann K, KRAEHENBUHL JP, ROSSIER BC: Regulation of aldosterone of Rev Biochem 55:987—1036, 1986 Na,K-ATPasemRNAs,protein synthesis and sodium 146.transport KIRSCHNER MW: Microtubule assembly and nucleation. mt Rev in cultured kidney cells. J Cell Biol 104:1231—1237, 1987 Cytol 54: 1—71, 1978 123. GUERNSEY DL, EDELMAN IS: Loss of thyroidal inducibility of 147. MOOSEKER MS, TILNEY LG: Organization of actin-filament-mern- Na,K-ATPase with neoplastic transformation in tissue culture. J brane complex: Filament polarity and membrane attachment in BiolChem 261:11956—11961,1986 the microvilli of intestinal epithelial cells. JCell Biol 67:725—743, 124. AVNERED, SWEENEY WE, PIEsco NP, ELLIS D:Triiodothyro- 1975 nine-induced cyst formationin metanephric organ culture: The 148. GEIGER B: A 130K protein from chicken gizzard: Its localization at roleof increasedNa-K-adenosine triphosphatase activity. J Lab the termini of microfilament bundles in cultured chicken cells. Cell Clin Med 109:441—453, 1987 18:193—205, 1978 125. SALEHMOGHADDAM S,BRADLEYT,MIKHAILN, BADIE-DEZ- 149.BURRIDGEK, FERAMISCO R: Microinjection and localization of a FOOLY B, NORD E,TRIZNAW,KHEYFETSR,FINEL:Hypertro- 130K protein in living fibroblasts: A relationship to actin. Cell phyof basolateral Na-K pump activity in the proximal tubule of 19:587—595, 1980 the remnant kidney. Lab Invest 53:443—450, 1985 150.BENNETTG, CARLET E, WILD G, PARSONS S: Influence of 126. WOLITSKY BA, FAMBROUGH DM: Regulation of the (Na-K)- colchicine and vinbiastine on the intracellular migration of secre- ATPase in cultured chick skeletal muscle. Modulation of expres- tory and membrane glycoproteins. Am J Anat 170:545—566, 1984 sion by the demand for ion transport. J Biol Chem 261:9990—9999, 151. BLOK J, GINSEL A, MOLDER-STAPEL A, ONDERWATER ii, DAEMS 1986 WT: The effect of colchicine on the intracellular transport of 127.WILSON PD,HRENIUKD:Altered polarity of NaK-ATPase in 3H-fucose-labelled glycoproteins in the absorptive cells of cul- epitheliawith a genetic defect and abnormal basement membrane. tured small intestine tissue. Cell Tissue Res 215:1—12, 1981 (abstract) J Cell Biol 105:1760, 1987 152. DANIELSEN EM, COWELL GM, POULSON SS: Biosynthesis of 128.WILsoNPD: Cystic epithelia in vitro, in Proceedings of the microvillar proteins: Evidence for an intracellular sorting taking Polycytic Kidney Foundation, TP&TInc., 1988 place in or shortly after exit from the Golgi complex. Eur J 129. WILsoNPD,SHERWOODAC,PALLAK,Du J,WATSON R, Biochem 152:493—499, 1983 NORMANJT: Reversed polarity of NaK-ATPase: Mistocation153. RINDLER Mi, IvANov IE, SABATINI DD:Microtubule acting to apical plasma membranes in human polycystic kidney disease drugslead to non-polarized delivery of the influenza hernagglutinin epithelia. J Clin Invest (submitted for publication) to the cell surface of polarized Madin-Darby canine kidney cells. 130. HANDLER iS: Overview of epithelial polarity. Ann Rev Physiol JCellBio! 104:231—241, 1987 51:729—740,1989 154. ACHLER C, FILMER D, MERTE C, DRENCKHAHN D: Role of 131.RODRIGUEZ-BOULAN E, NELSON WJ: Morphogenesis of the po- microtubules in polarized delivery of apical membrane proteins to larized epithelial cell phenotype. Science 245:718—725, 1989 thebrush border ofintestinal epithelium. J Cell Biol 109:179—189, 132. SIM0N5 K, FULLER SD: Cell surface polarity in epithelia. Ann Rev 1989 Cell Biol 1:243—288, 1985 155. HYNES RO, DESTREE AT: Relationships between fibronectin 133. Low MG: Glycoryl-phosphatidylinositol: A versatile anchor for (LETS protein) and actin. Cell 15:875—866, 1978 cell surface proteins. FASEB J 3:1600—1608, 1989 156.HOLTHOFERH, MIETTINEN A,PASSIVO R,LEI-ITO VP, LINDRI E, 134. LIsANrI MP, SARGIACOMO M, GRAEVE A, SALTIEL AR, ROD- ALTHAN 0, VIRTANEN i:Cellular origin and differentiation of RIGUEZ-BOULAN E: Polarized apical distribution of glyconyl- renalcarcinomas: A fluorescent microscopic study with kidney phosphatidylinositol anchored proteins in a renal epithelial cell specific antibodies, intermediate filament antibodies and lectins. line. Proc NatlAcad Sci USA 85:9557—9561, 1988 Lab Invest 49:317—326, 1983 135. MOLIT0RIs BA, WILSON PD, SCHRIER RW, SIMON FR: Ischemia 157. BAcHMANN 5, KRIZ W, KUHN C, FRANKE WW: Differentiation of induces partial loss of surface membrane polarity and accumula- cell types in the human kidney by immunofluorescence micros- tion of putative calcium ionophores. J Clin Invest 76:2097—2105, copy using antibodies to intermediate filament proteins and des- 1985 moplakins. Histochem 77:365—394, 1983 136. NELSON WJ, HAMMERTON RW: A membrane-cytoskeletalcorn- 158. GRONE H, WEBER K, GRONE E, HELMCHEN U, OSBORN M: plexcontaining NaK-ATPase, ankyrin and fodrin in Madin- Coexpression of keratin and vimentin in damaged and regenerat- Darby canine kidney (MDCK) cells. Implications for the biogen- ing tubular epithelia of the kidney. Am J Pathol 129:1—8, 1987 esis of epithelial cell polarity. J Cell Biol 108:893—902, 1989 159. GERSTENFELD LC, FINER MH, BOEDTKER H: Altered /3-actin 137. NELSON Wi, VESHNOCK Pi: Ankyrin binding to (Na,K')-AT- gene expression in phorbol myristate acetate treated chondrocytes Pase and implications for the organization of membrane domains and fibroblasts. Mol Cell Biol 5:1425—1433, 1985 in polarized cells. Nature 328:533—536, 1987 160. HARTMAN AL, SAWTELL NW, LESSARD JL: Expression of actin 138. SPECTOR R, JOHANSON CE: The mammalian choroid plexus. Sci isoforms in developing rat intestinal epithelium. JHistochem Am:68—74,1989 Cyrochem37:1225—1233, 1989 139. Go KG, HOUTHOFF H, HARTSUIKER J, BLAAUW EH, HAVINGA 161. VANDEKERCKHORE i, WEBER K: At least six different actins are P: Fluid secretion in arachnoid cysts as a clue to cerebrospinal expressed in a higher mammal: An analysis based on the amino fluid absorption at the arachnoid granulation. J Neurosurg 65:642— acid sequence of the amino-terminal tryptic peptide. J Mo! Biol 648, 1986 126:783—792, 1978 140. GRANTHAM JJ, UCHIC M, CRAGOE EJ, KORNI-IAUS J, GRANTHAM 162. SHERWOOD AC, WILSON PD: Actin cytoskeleton in autosomal JA, DoNoso V, MANGOO-KARIM R, EVAN A, MCATEER J: dominant polycystic kidney disease epithelial cell. (abstract) Kid- Chemical modification of cell proliferation and fluid secertion in ney mt37:229, 1989 renal cysts. Kidney mt 35:1379—1389, 1989 163. KLEE CB: Ca2-dependent phospholipid-(and membrane-) binding 141. AVNER ED, SWEENEY WE, ELLIS D: In vitro modulation of proteins. Biochem 27:6645—6653, 1988 tubular cyst regression in murine polycystic kidney disease. 164. SAWYER ST, COHEN S: Epidermal growth factor stimulates the Kidneymt36:960—968, 1989 phosphorylation of the calcium-dependent 35,000 dalton substrate 142.WILSONPD, PALLA K, GUERRA E, HRENIUK D: Effects of in intact A-431 cells. J Biol Chem 260:8233—8236, 1985 reversed polarity of NaK-ATPase on vectorial Na transport in165. DAVIDSON FF, DENNIS EA: Biological relevance of lipocortins cultured human polycystic kidney disease epithelia. AmSoc and related protein inhibitors of phospholipase A2. Biochem Nephol1989, p. 414 Pharm38:3645—3651,1989 Wilson and Sherwood: Tubulocys tic epithelium 463

166. FLOWER RJ: Lipocortin and the mechanism of action of the171. PANAYOTOU G, END P, AUMAILLEY M, TIMPI R, ENGEL J: glucocorticoids. BrJ Pharmacol 94:987—1015,1988 Domains of laminin with growth-factor activity. Cell56:93—101, 167. NAKATA T, SOBUEK,HIROKAWA N: Conformational change and 1989 localization of calpactin I complex involved in exocytosis as 172. PARRY G, CULLEN B, KAETZEL CS,KRAMERR, Moss L: Regu- revealed by quick freeze deep-etch microscopy and immunocyto- lation of differentiation and polarized secretion in mammary chemistry. JCell Biol 110:13—25, 1990 epithelial cells maintained in culture: Extracellular matrix and 168. WooDs A, COUCHMAN JR, HooK M: Heparan sulfate proteogly- membrane polarity influence. J Cell Biol 105:2043—2051, 1987 cans of rat embryo fibroblasts. A hydrophobic form may link 173. RAPRAEGER A, JALKANEN M, BERNFJELD M: Cell surface proteo- cytoskeleton and matrix components. JBiol Chem 260:10872— glycan associates with the cytoskeleton at the basolateral cell 10879, 1985 surface of mouse mammary epithelial cells. JCell Biol 103:2683— 169. BAssoLs A, MASSAGUE J: Transforming growth factor /3 regulates 2696, 1986 the expression and structure of extracellular matrix chondroitin! 174. WALSH-REITZ MM, GLUCK SL, WAACK 5, TOBACK FG: Lowering dermatan sulfate proteoglycans. JBiol Chem 263:3039—3045,1988 extracellular Na concentration releases autocrine growth factors 170. THOMPSON KL, AssolAN R, ROSNERMR:Transforming growth from renal epithelial cells. ProcNatl Acad Sci USA 83:4761—4768, factor /3-increased the transcription of the genes encoding epider- 1986 mal growth factor and fibronectin in normal rat kidney fibroblasts. 175. STANTON B, KAI55LING 5: Regulation of renal ion transport and JBiolChem 263:19519—19524,1988 cell growth by sodium. Am J Physiol 257:F1—F10, 1989