Regeneration After Acute Tubular Necrosis

Kidney International, Vol. 41 (1992), pp. 226—246 NEPHROLOGY FORUM Regeneration after acute tubular necrosis

Principal discussant: F. GARY TOBACK

Department of Medicine and committee on cell Physiology, The University of chicago, chkago, Illinois

metabolic setting that accompanies death from uremia, a micro- Editors dissected nephron from this patient's kidney showed exuberant JORDAN J. COHEN focal proliferation of tubular epithelial cells (Fig. 1). The JOHN T. HARRINOTON surprising observation made 40 years ago that the regenerating NIcOLAO5 E. MAmAs nephron can contain collections of growing cells that project into the tubular lumen or interstitium has been confirmed by Managing Editor others [2—4]. What factors mediated the proliferation of renal CHERYL J. ZUSMAN epithelial cells in the disordered, catabolic internal milieu of this patient with acute renal failure? Is focal tubular cell multiplication evidence that growth-stimulatory factors were released State University of New York at Stony Brook locally by renal cells? If so, cellular mediators produced in the and injured, but healing, nephron might contribute to reepithelial- Tufts University School of Medicine ization after tubular necrosis. The relationships among renal cell dysfunction, morphologic damage, and impairment of glomerular filtration and renal blood flow in acute renal failure are poorly understood [5]. There are occasional examples in humans, and many more among exper- Case presentation imental animal models, in which extensive tubular cell necrosis A 44 year-old woman took six mercuric bichloride tablets (3.0 gm.) is associated with failure of renal function. Often, however, this and entered the hospital with nausea and abdominal cramps three hours is not the case. Renal tissue from many patients does not exhibit later. Her blood pressure was 86/60. Five hours after the ingestion of significant abnormalities at biopsy or autopsy despite oliguria; the poison, BAL treatment was begun and she was given 2 cc. every four hours for six days, and then 1 cc, every 12 hours for five days. this finding suggests that injury to the renal epithelium often During the first 42 hours the patient had diarrhea, vomited and suffered causes cellular dysfunction but not necrosis. When structural abdominal cramps. Fifty cc. of urine were passed in the first 24 hours, injury is present, it probably serves more as a guide to the then none for five days. The urinary output rose slowly for six days to severity of the lesion than as a marker of specific biochemical 935 cc. and then fell to an average of 120 cc. until death. The NPN on the second day was 53 mg. per cent rising to 450 on the 28th day. On the pathways that are disrupted and that are responsible for cellular 29th day she died, clearances of inulin, creatinine and PAH weredysfunction. Important biochemical abnormalities in acute re- infinitesimal on six occasions. nal failure are a reduction in the cytosolic concentration of ATP [Abbreviations: BAL, British antilewisite or dimercaprol; NPN, and an increase in cytosolic calcium, alterations in the cytoskel- nonprotein nitrogen (normal, 22—29mg per cent); PAH, para-aminohip- eton, generation of free radicals, and reduced protein synthesis purate.] [61. How a specific ischemic or toxic insult to the kidney Discussion induces biochemical perturbations that result in a marked DR. F. GARY TOBACK (Department of Medicine and Cellreduction in GFR, a lesser decrement in renal blood flow, and Physiology, and Professor of Medicine and Cell Physiology,oliguria is not clear. If the injury is mild, structural changes may University of Chicago, Chicago, Illinois): This case, taken fromnot be apparent at the light microscopic level. If the insult is the landmark paper on nephron structure in acute renal failuremore severe, alterations such as sloughing of proximal tubular- by Jean Oliver, Muriel MacDowell, and Ann Tracy, wasbrush-border microvilli can occur that are rapidly reversed after published in 1951, before the dialysis era began [1]. It is a briefadequate blood flow returns. If the injury is of sufficient chronicle of the relentless, malignant course of untreated mer-intensity and/or duration, cell death ensues. In most patients curic-chloride-induced tubular necrosis. Despite the abnormalwith acute failure of glomerular filtration, cellular dysfunction appears to be more important than frank structural damage because the two structural lesions found most consistently in humans, loss of microvilli and necrosis of individual tubular Presentation of the Forum is made possible by grants from Pfizer,epithelial cells, are found relatively infrequently [7]. Incorporated; Sandoz, Incorporated; Marion Merrell Dow Incorpo- When a renal insult induces both nonlethal injury of some rated; Merck Sharp & Dohme International; and Amgen Incorporated. cells and necrosis of others, recurrent insults probably continue © 1992 by the International Society of Nephrology after the onset of the acute renal failure syndrome and during

226 Nephrology Forum: Regeneration after acute tubular necrosis 227

apical surface of the cell with loss of barrier function of the plasma membrane [9]. At the cytoskeletal level, gaps in the terminal web of actin microfilaments develop, and actin redis- tributes in the cytoplasm [9]. If the intensity and/or duration of the insult is limited, the cell becomes dysfunctional but sur- vives. Injury that is more severe, however, probably leads to detachment of the cell from the tubular basement membrane, before or after necrosis. This analysis suggests two therapeutic strategies: one designed to facilitate recovery of cells after nonlethal injury, and a second designed to stimulate proliferation of noninjured and surviving injured cells that recovered to replace those dead cells that detached. Polypeptide factors that induce hyperplasia of renal epithelial cells have been identified [10], and these factors might play a role in the repair of nephrons after acute injury and necrosis (Fig. 2, pathway 1). Mechanisms that repair nonfatally injured cells (Fig. 2, pathway 2) are not well defined, although recovery of this population of cells is of major importance in the healing process. How repair of injured cells and proliferation to replace necrotic ones result in restoration of normal glomerular filtration and renal blood flow remains uncertain. In this Nephrology Forum, I will focus attention on mechanisms that mediate cellular regeneration after tubular necrosis. Prescott has shown that, under physiologic conditions, nearly 2 x 106 tubular epithelial cells, about one cell per human nephron, slough into the urine each day [11]. Some proliferation therefore is required under normal circumstances to replace the

r lost cells and to maintain structural integrity of the kidney. In acute renal failure, injury increases the rate of cell loss beyond Fig. 1. Portion of a microdissected nephron from the patient who died of mercury-induced acute rena/failure. The terminal medullary portion the capacity of basal physiologic proliferative mechanisms to of a proximal convoluted tubule is shown. The cortical portion of this keep pace and to maintain nephron integrity. To restore renal convolution showed the typical nephrotoxic necrosis due to mercury. structure and function, recovery must take place despite the This lesion is visible in the horizontal stretch of the tubule at point A, an retention of toxins and nitrogen, acidosis, electrolyte imbal- intact basement membrane outlining the dead mass of dark stained necrotic debris. At B there begins a long segment of tubule, greatly ance, anemia, hypervolemia, and a severely curtailed intake of swollen, in which the basement membrane has disintegrated completely nutrients. That restoration does occur demonstrates the ex- and the tubule wall is destroyed. The tubular destruction ends abruptly traordinary capacity of renal cells to regenerate. Renal repair (C) and the convolution continues, its basement membrane intact, to and regeneration are anabolic processes that must occur in the the passage into the thin portion of Henle's loop. At the point of greatest disruption (B'), irregular regeneration of the epithelium issetting of systemic catabolism. Although rarely stated, the evident in the clusters and masses of cells with large vesicular nuclei. It therapeutic principle that underlies the management strategy in is also evident that this regeneration has failed to repair the tubule wall acute renal failure is that the kidney heals itself when support- or reestablish a lumen, this failure apparently due to lack of an intact ive care such as dialysis is provided. Another strategy exists: basement membrane to direct the course of the proliferating cells (magnification x 82). Insert B' shows the disruptive lesion at B'acceleration of the pace of regeneration by provision of nutri- (magnification x 200). ents such as amino acids and by the administration of growth Insert A shows a typical nephrotoxic tubular lesion in the terminal factors [12]. The rationale for this strategy lies in the observa- segment of a proximal convolution of a nephron from the kidney of a tion that the rate of renal regeneration is ordinarily suboptimal dog that had received an intravenous injection of 75 mg of corrosive sublimate. The renal epithelium is completely necrotic and reduced to because systemic hypercatabolism increases the body's de- granular debris. Note the intact basement membrane, at places out ofmand for calories and nutrients, while at the same time, the focus and therefore invisible, which maintains the tubular contoursadverse effects of renal failure on renal excretory and gastroin- (magnification x 200). (Reproduced from Ref. I by copyright permis- testinal function decrease the supply of those nutrients. The sion of the American Society of Clinical Investigation.) role of dialysis is to correct acid-base, electrolyte, and volume imbalances and to remove retained toxins so that the extracellular environment for renal cellular repair is more favorable. the regenerative process [8]. A schematic response to renalProvision of nutrients, calories, and growth factors could re- injury is presented in Figure 2. At the subcellular level, mito-verse the catabolic state, permit accelerated repair of injured chondrial swelling and blebbing of the endoplasmic reticulumrenal tissue, and thereby speed recovery. I believe that we occur, and the proximal tubular brush border can undergoshould focus on developing and implementing this strategy in sloughing. In the plasma membrane, loss of protein orientationthe next decade [12]. and alterations in phospholipid content occur, such that Na/K- Rats with mercuric chloride-induced acute renal failure man- ATPase is displaced from its normal basolateral position to theifest stimulated synthesis of cellular protein, nucleic acids, and 228 NephrologyForum: Regeneration after acute tubular necrosis

Fig. 2. Renal epithelial cell repair after acute injury. A renal insult such as hypoxia can affect the cytosol, organelles, and plasma membrane. Cells can be uninjured, nonlethally injured, or killed. Dead cells could be replaced by proliferation of noninjured cells and/or nonlethally injured cells that have recovered (pathway 1). Nonlethally injured cells that are dysfunctional can die or can subsequently recover function and structure (pathway 2). Growth factors are likely to play a role in the proliferative response along pathway 1, and possibly in the recovery of nonlethally injured cells (pathway 2). phospholipids followed by rapid proliferation of tubular epithe- Table 1. Structural characteristics ofpolypeptidegrowth factorsa hal cells on the second day of the syndrome [13—15]. These Protein sizec observations indicate that anabolic mechanisms overcome the Growth unfavorable effects of systemic catabolism and the azotemic factor" Precursor Mature mRNA size4 environment that ordinarily inhibit cell growth. Previous stud-hTGF-/3l 390 112 2.5 ies in our laboratory showed that in rats with nephrotoxic acute mTGF-/32 414 112 2.4 renal failure, infused amino acids acted directly on regeneratinghEGF 1207 53 5.0 hTGF-a 160 50 4.8 renal cells to stimulate synthesis of phosphatidylcholine andhIGF-I 156 70 9.0, 5.3, 7.7 protein, essential for the biogenesis of organdies and surfacehIGF-Il 180 67 4.9, 6,Oe structures required for normal cell function [16]. We also hPDGF-A 211 125 1.9, 2.3, 2.8 observed improvement in renal function [17]. Extrapolation ofhPDGF-B 240 160 4.2 these results suggests that in patients with acute renal failure, haFGF 155 140 4.8 hbFGF 155 146 2.2, 4.6 amino acids delivered after the injury could act directly on renal cells to speed recovery of structure and function. Provision of a Adaptedfrom Ref. 10 bSourceof growth factor: h, human; m, monkey; TGF, transforming nutrients and effective dialysis also might promote regenerationgrowth factor; EGF, epidermal growth factor; IGF, insulin-like growth by facilitating the action of growth factors released by survivingfactor; PDGF, platelet-derived growth factor; aFGF, acidic fibroblast renal cells in the injured nephron. growth factor; bFGF, basic fibroblast growth factor Values are number of amino acids d Cell biology of renal growth factors Valuesare number of kilobases eValuesare for adult renal tissue; sizes differ in fetal kidney and Growth factors produced in the kidney could have initiated other tissues cell proliferation and the tufts of tubular cells observed in the patient presented, who died of mercuric-chloride-induced tubular necrosis (Fig. 1). Growth factors are polypeptide messengers that can stimulate dormant cells in the G0 phase of the cellof renal cells in vivo [20], suggest that growth inhibitory factors cycle to initiate DNA synthesis and subsequently to undergopredominate under physiologic conditions, although the set mitosis. Growth factors affect target cells by binding to specificpoint could be shifted when regeneration after injury is re- surface receptors. This ligand-receptor interaction results inquired. activation of an intracellular effector mechanism, such as the Figure 3 illustrates four mechanisms whereby cell-cell com- tyrosine kinase domain of the receptor or the phosphatidyhino-munication can be mediated by peptide messengers. Endocrine sitol signaling pathway [18, 191. Signal transduction is followedsignaling occurs when a protein is secreted into blood or lymph by a cascade of biochemical events that initiate DNA synthesis.for transport to distant cell target sites, as is the case with Recent work in many laboratories has identified the amino aciderythropoietin. A paracrine signaling cell secretes proteins that sequence of growth factors and the nucleotide sequence of theaffect only adjacent target cells that have functional growth genes that encode them [10]. Growth factors are derived fromfactor receptors on their surface. In autocrine signaling, cells larger precursor molecules by proteolytic processing (Table 1).respond to protein factors that they release; cells that utilize Epidermal growth factor (EGF) and insulin-like growth factorsthis mechanism thus secrete the growth factor and express (IGFs) act on renal cells to stimulate exit from the premitotic orreceptors for it on their plasma membrane [21]. A juxtacrine resting (G0!G1) phase of the cell cycle, whereas transformingmechanism recently has been described for TGF-a in which the growth factor-type /3 (TGF-f3) impedes the exit of renal cellsgrowth factor domain of the plasma-membrane-bound precur- from G1. These observations, and the low mitotic index (0.1%)sor molecule acts as a ligand for a receptor on an adjacent target Nephrology Forum: Regeneration after acute tubular necrosis 229

ENDOCRINE PARACRINE AUTOCRINE JUXTACRINE

Fig.3. Mechanisms of cell-cell communication by growth factors. Cells synthesize precursor molecules (depicted in the cytosol) that can be processed and released to act as growth factors on cells that produced them (autocrine), neighboring (paracrine), or distant cells (endocrine). Also, the precursor can be incorporated into the plasma membrane and act on adjacent cells (juxtacrine). The thickened curved and angular regions of the plasma membrane indicate receptor sites for growth-promoting and growth-inhibitory molecules, respectively. (•,stimulatorygrowth factors; A, inhibitory growth factors; w•and''A, precursor molecules.)

cell that is thereby stimulated to enter mitosis [22]. Consider-ability to induce production of autocrine or paracrine growth ation of the juxtacrine mechanism could provide fresh insightfactors by these renal cells (Fig. 4). into the as-yet-unknown role of the large amount of preproEGF Lowering extracellular potassium concentration stimulates protein present in the distal nephron of the murine kidney [23].growth of renal cells by an autocrine mechanism. More than 50 Receptor number, affinity, biosynthesis and recycling, and theyears ago, Schrader et al found that systemic potassium deple- integrity of signal transduction mechanisms at the level of thetion, achieved experimentally by feeding rats a diet deficient in plasma membrane, cytosol, cytoskeleton, and nucleus also arepotassium, induced renal growth [35]. Sustained mild hypokale- critical determinants of how cells respond to growth factors.mia in rats is associated with a doubling of renal mass due to These molecules could have physiologic effects that are unre-both hyperplasia and hypertrophy [36, 37]. An in vitro model of lated to their mitogenic properties. this phenomenon became available when we found that simply Unexpected observations from our laboratory led us tolowering the extracellular potassium concentration from 5.4 conclude that renal epithelial cells release growth factors in mM (control) to 3.2 mM in confluent cultures of BSC-l cells response to physiologic signals [24, 25]. Cultures of nontransformed renal epithelial cells derived from the African greeninduced accelerated growth [31]. Thus, these cells perceive a reduction in the extracellular potassium concentration as a monkey (Cercopithecus aethiops) BSC-l line [261 have been used, because this simple culture system circumvents many ofmitogenic signal. the problems inherent in studies of organ growth in a tissue as We then tested the hypothesis that proliferation in response complex as the kidney. High-density, quiescent cultures haveto low-potassium medium is mediated by the release of a been employed to simulate the low proliferative activity ofgrowth-promoting factor from the cells [38]. Low-potassium or highly differentiated kidney cells in vivo [27, 28]. Under thesecontrol medium was conditioned by placing it on confluent conditions, the epithelial monolayer forms small domes overly-cultures of BSC- 1 cells. The low-potassium-conditioned me- ing accumulated fluid; the cells thus might have the capacity fordium was aspirated from the dish, and its potassium concentra- transepithelial transport [29]. The nephron segment from whichtion was restored to the control value (5.4 mM) by the addition BSC-l cells originated is unknown, although their responsive-of potassium chloride. This normal-potassium-conditioned me- ness to vasopressin [30] suggests a relationship to the distaldium stimulated growth of fresh cultures of cells to the same nephron. extent as did unconditioned, low-potassium medium. Growth- Our studies show that small reductions in the extracellularstimulating activity in low-potassium conditioned medium first potassium or sodium concentrations induce DNA synthesis andappeared after one hour (Fig. 4) and was optimal when the cell multiplication [31, 32], that adenosine diphosphate (ADP) ispotassium concentration during conditioning was 3,2 mM. The the most powerful mitogen yet defined for BSC-1 cells [33], andgrowth factor(s) have an apparent Mr of 12,000—30,000, as that a cytokine of mesangial cell origin, interleukin-1f3, alsodetermined by ultrafiltration and dialysis. The factor(s) are promotes growth [34]. These four diverse mitogenic signals,highly potent because they retain their stimulatory activity at that is, changes in the extracellular concentration of twodilutions greater than 1:1000. Their activity is destroyed by cations, a nucleotide, and a cytokine, also have in common theheating to 56°C for 30 minutes, but is stable at 4°C for at least 3 230 Nephrology Forum: Regeneration after acute tubular necrosis

Low-K Low-Na PDGF 8r- growth factor growth factor B chain 0 x E 30 U0. a 7 (0 2.5 x 20 x -JCl) -J (0 0w 0q) z E 10 6 2.0 01 0 I I I I I I I I I 0 1 2 3 012345 0 1 2 012345 HOURS MINUTES HOURS HOURS Fig. 4. Production of autocrine and paracrine growth-promoting factors by renal epithelial cells exposed to low-potassium medium, low-sodium medium, adenosine diphosphate, or interleukin-1. Left panel: Appearance of growth-stimulating activity in low-potassium medium. BSC-l cells were exposed to medium containing 3.2 mM K and 0.01% calf serum for different amounts of time. The conditioned medium was collected, and its K concentration was adjusted to the control value (5.4 mM) by the addition of KCI. This low-potassium-conditioned medium was added to a fresh culture, and its effect on multiplication was assessed by counting the number of cells 3 days later. Release of growth-stimulating activity was maximal after one hour of exposure of cells to low-potassium medium. Center-left panel: Appearance of growth-promoting activity in low-sodium medium. At time zero, confluent cultures were exposed to low-sodium medium (130 mM) containining 0,01% serum, and the conditioned medium was collected at the times specified on the abscissa. After the Na concentration of the conditioned medium was adjusted to the control value (155 mM) by the addition of NaCI, the medium was added to a fresh culture, and its effect on multiplication was assessed by counting the numberof cells 4 days later. Maximal release of growth-promoting activity was detected by 3 minutes. Center-right panel: ADP stimulates release of PDGF-B chain. BSC-1 cells were exposed to ADP for I or 2 hours. The conditioned medium was collected, ADP in it was removed by dialysis, and mitogenic activity was assayed in cultures of normal rat kidney (NRK) fibroblasts. Human anti-PDGF IgG inhibited the capacity of ADP-conditioned medium to stimulate [3H]thymidine incorporation into DNA. After 1 hour of exposure to ADP, PDGF-like activity was about 10-fold greater than that in control conditioned medium. Right pane!: Interleukin- I/I induces expression of the gro1MGSA gene, whose product is a mitogen. BSC-l cells were exposed to interleukin-lj3 for different amounts of time. Total RNA was extracted, electrophoresed on a formaldehyde/agarose gel, transferred to a GeneScreen Plus membrane, and hybridized with a [32P]labeled probe prepared from the A2/gro clone. The gene for gro/MGSA was maximally expressed at 3 hours. (The data presented are adapted from references 32, 34 38, 50.) days. This activity differs from that of other, previously de-lated when the sodium concentration of the medium was scribed growth factors. The growth-promoting activity in low-reduced from 155 mM (control) to 130 mM [32]. The capacity of potassium-conditioned medium can be antagonized by thelow-sodium medium to increase cell proliferation depended on addition of purified TGF-j32, the autocrine inhibitor of BSC- 1the decreased sodium concentration, as growth stimulation was cell growth [39]. equivalent whether equimolar amounts of choline chloride or This study indicated that stimulation of renal epithelial cellisosmolar amounts of sucrose were used to replace sodium growth induced by lowering the extracellular potassium con-chloride deleted from the medium. centration is associated with the appearance of one or more After exposure of confluent cultures for only 3 to 5 minutes to mitogenic factors in the medium. This observation provided thelow-sodium medium (130 mM) (Fig. 4), the cells release two first evidence that cell growth induced by an alteration in the iongrowth-stimulatory factors (apparent Mr 6200 and 9000). Re- concentration of the extracellular fluid could be mediated by anlease does not depend on new protein synthesis, as it occurs in autocrine mechanism. It remains to be determined whetherthe presence of cycloheximide. The activity in conditioned these growth factors are produced by renal cells in animalsmedium is stable at 56°C for 30 minutes and is retained after rendered hypokalemic by a potassium-deficient diet. In vivo,freezing for many weeks. Treatment with trypsin abolishes the the factors could recruit unstimulated cells along the nephron togrowth-promoting effect, whereas dithiothreitol does not; these replicate, maintain the mitogenic signal, and/or serve an as-yet-observations suggest that the low-sodium growth factors prob- undefined purpose related to cell transport. ably are proteins that have no accessible disulfide bonds re- Reduced extra cellular sodium concentration induces renalquired for activity. The growth response to a reduction in cellgrowthand production of growth factors. The influx ofextracellular sodium concentration appears to be cell-type sodium ions is augmented during the onset of proliferation inspecific, because low-sodium medium does not stimulate pro- mammalian cells [40, 41J, so one might expect to find thatliferation of fibroblasts. Release of growth factor activity is also exposure of BSC-l cells to medium containing a reducedcell-type specific, as conditioned medium from fibroblasts ex- concentration of sodium would inhibit mitogenesis. We wereposed to a reduced extracellular sodium concentration does not surprised, therefore, to find that cell multiplication was stimu-stimulate renal epithelial cell growth. The low-sodium growth Nephrology Forum: Regeneration after acute tubular necrosis 231 factors appear chemically and functionally different from other Table 2.Effect of ADP on gene expression growth factors. Margaret Walsh-Reitz and Naga Aithal in my Gene Effect of ADP laboratory have purified the 6200 Mr low-sodium growth factor by reversed-phase high-performance liquid chromatography.Egr- 1, NAK1, c-fos, c-myc Expression induced Although its amino acid sequence has not yet been determined,c-Ha-ras, c-sis (PDGF-B chain), Constitutive expression increased TGF-p, transferrin receptor, amino acid compositional analysis indicates that it differs from fibronectin, tenascin, known growth factors of similar size. plasminogen activator inhibitor Low-potassium andlow-sodiumgrowth factors. The obser- EGF receptor, /3-actin, y-actin, Constitutive expression unaltered vations I have summarized indicate that exposure of renal vimentin, cathepsin L, thymosin f3 epithelial cells to low-potassium or low-sodium medium resultsIGF-I, IGF-H, preproEGF, TGF-Genes not expressed or induced in the rapid appearance of growth-promoting factors in the , PDGFreceptor, PDGF-A extracellular fluid. These autocrine products appear to differ in chain, v-src, v-fms, v-yes that (1) the apparent Mr of the growth factor(s) in low-potassium conditioned medium is 12,000—30,000, whereas the Mr in low- sodium conditioned medium is 6,200 and 9,000; (2) the low- potassium factors first appear in the medium after 60 minutes,sine, guanine, inosine, and the nucleotides of the latter two whereas the low-sodium factors are detected as early as 3substances. Furthermore, ADP stimulated DNA synthesis minutes; and (3) the growth-stimulating effect of low-potassiumthreefold more than any known growth-promoting agent for conditioned medium is destroyed by heating at 56°Cfor30 BSC- 1 cells. It is important that ADP and AMP were mitogenic minutes, whereas the low-sodium activity is stable under theseunder serum-free conditions, and their stimulatory effect appar- conditions. ently was not mediated by cAMP-dependent mechanisms [33]. Identification of the low-sodium growth factor and the obser-In addition, known metabolites such as adenosine, hypoxan- vation that lowering the potassium concentration of mediumthine, or inosine did not appear to mediate ADP or AMP from 5.4 to 3.2 mM for one hour releases a growth-promotingstimulation of DNA synthesis, because they were not as potent activity suggest that brief alterations in the ionic environmentwhen added to the culture medium. Autoradiography of cause renal epithelial cells and/or their extracellular matrix to[3H]thymidine-labeled cultures indicated that sixfold as many release autocrine products. The liberation of active factors bycells were synthesizing DNA in the presence of AMP than in its renal epithelial cell cultures suggests that renal tissue willabsence. No detachment of cells from the monolayer was respond similarly in vivo. As alterations in the serum concen-detectable under these conditions. The mitogenic effect of ADP tration of sodium and/or potassium are common in acute renalor AMP did not appear to be mediated through the adenosine failure, it is tempting to speculate that these derangements inreceptor, because the predicted stimulation of DNA synthesis the extracellular cation concentration could signal intact, non-was observed in the presence of a large excess of theophylline lethally injured, and regenerating cells to release autocrineor 5'-deoxy 5'-methylthioadenosine, which are potent adeno- growth factors that could speed the recovery process. sine receptor antagonists. Also, ATP stimulates [3H]thymidine These observations also suggest that modest reductions in theincorporation in primary cultures of rabbit proximal tubular extracellular concentration of potassium and sodium mightcells [47], whereas exogenous adenine nucleotides failed to result in the rapid appearance of physiologic mediators thatstimulate DNA synthesis in fibroblasts [331. could communicate information to other cells along the neph- To explore the molecular mechanisms by which ADP exerts ron. These factors might act as messengers between glomerularits mitogenic effect, we tested the hypothesis that stimulation of and tubular cells in response to changes in peritubular orDNA synthesis in renal epithelial cells of the BSC-l line is luminal ion concentrations. Although the factors released inmediated by gene activation. We studied the expression of 26 response to low-potassium and low-sodium media induce pro-genes and proto-oncogenes, most of which contribute to growth liferation in cell culture, their physiologic role might be differentcontrol, in BSC-l cells in the absence and presence of ADP in the intact kidney. Many mitogens stimulate fluxes of ions and[48—51]. The responses observed were divided into four pat- nutrients (42—461. One possible role, therefore, could be aterns (Table 2): genes not constitutively expressed, but induced modification of epithelial transport in response to changes in theby ADP; genes whose constitutive expression was increased or extracellular environment. Release of autocrine or paracrineunaltered by the nucleotide; and genes that were neither con- factors with this property suggests a previously undisclosedstitutively expressed nor induced. Exposure of renal epithelial mechanism for regulation of ion and nutrient transport incells to ADP is followed by rapid elevation of the level of Egr- 1 adjacent and distant epithelial cells. Additional studies will betranscripts at 10 minutes, NAK1 at 20 minutes, c-fos at 30 required to determine whether these factors contribute to theminutes, c-myc at one hour, maximal expression of c-Ha-ras control of renal epithelial cell growth and/or transport in vivo.mRNA at 12 hours, and stimulation of DNA synthesis between AdenOsine diphosphate, a potent mitogen for renal epithelial12 and 24 hours. The level of transcripts for the transferrin cells, stimulates production of paracrine and autocrine growthreceptor, another cell cycle-dependent gene, is also maximal at factors. During the course of a study to define the role of cAMP12 hours, whereas expression of the y-actin gene remains in renal epithelial cell growth, we observed that exogenousunchanged. Stimulation of Egr-1 by ADP and proto-oncogene adenosine 5'-monophosphate (AMP), used as a control, was aexpression occur during the G1 phase of the cell cycle, in that powerful mitogen [33]. Each of the adenine nucleotides stimu-maximal expression of these transcripts is detected prior to the lated DNA synthesis in confluent cultures of BSC-l cells. Bothincrement in DNA synthesis. These findings represent the first ADP and AMP were more potent mitogens than ATP, adeno-reported examples of increased expression of Egr- 1, NAK 1, 232 Nephrology Foru,n: Regeneration after acute tubular necrosis c-fos, c-myc, c-Ha-ras, and transferrin receptor genes in stim-immunoreactive PDGF-like protein. Most important, ADP rap- ulated, nontransformed epithelial cells. Most important, ADPidly induced release of PDGF-like activity, so that one hour activation of gene expression appears to be gene specific. after exposure of the cells to the nucleotide, the amount of Genes that are activated rapidly and transiently, and whosegrowth factor was increased by more than tenfold (Fig. 4). The induction is shared by diverse cell types when exposed toBSC-1 cells did not exhibit a mitogenic response to authentic different growth stimuli, are likely to be important in thePDGF, nor did they express the gene encoding the PDGF transduction of mitogenic signals and regulation of cell prolif-receptor. These results imply that if the PDGF-like protein is eration. The Egr-l cDNA sequence encodes a protein withreleased by renal epithelial cells in vivo, the protein could, via zinc-finger domains that could facilitate DNA binding [52]. Thisa paracrine mechanism, initiate proliferation of adjacent stro- protein is postulated to function as a transcriptional regulatorymal (fibroblastic) and/or vascular (smooth muscle) cells within factor that in turn could control induction of several genes. Thisthe organ. novel gene also is induced in serum-stimulated mouse fibro- Some transformed cells can release PDGF-like proteins and blasts, insulin-stimulated rat hepatoma cells, and phytohemag-express PDGF receptors [59, 60]. These findings suggest that an glutinin-stimulated human peripheral blood lymphocytes, thusautocrine mechanism contributes to their proliferation. Human sharing many features with c-fos and c-myc [49, 52]. In additionfetal kidney and Wilms' tumor cells have been shown to to increasing the expression of early growth response genes,produce PDGF [611; production of the growth factor by Wilms' ADP also stimulates expression of the gene for TGF-f3, whichtumor cells may be a marker for their embryonic ongin rather encodes a protein that inhibits growth of BSC-l cells, and whichthan the malignant phenotype. Like the BSC-l line, cells stimulates expression of the genes that encode fibronectin andderived from Wilms' tumor and fetal kidney bind little PDGF; tenascin, proteins found in the extracellular matrix (ECM). Inthis finding suggests that PDGF might have a paracrine function summary, ADP can mimic the action of a growth factor and canin vivo. stimulate expression of cell-cycle-specific genes (Egr- 1 and Figure 5 summarizes some of the mechanisms by which ADP proto-oncogenes), as well as the gene for TGF-f3, the expres-can initiate growth of renal epithelial cells. As Table 2 shows, sion of which could serve to limit cell proliferation by producingADP is presumed to bind to a specific plasma membrane a negative autocrine growth factor. receptor and to activate the early response genes Egr- 1, NAK 1, During the study of ADP-induced gene expression, we foundc-fos, and c-myc. There is rapid release of PDGF B-chain that quiescent BSC-l cells constitutively release platelet-derived growth factor (PDGF)-like activity [50]. The presence of ahomodimer, a paracrine factor; TGF-f32, an autocnne growth PDGF-like protein was unexpected because expression ofinhibitor; and an apparently novel autocrine growth stimulator PDGF was thought to be confined to cells of mesodermal originidentified by Sreedharan Kartha in our laboratory. Further, such as fibroblasts and smooth muscle cells [53]. This growthADP also activates the proto-oncogenes c-sis and c-Ha-ras as factor, released from platelets and macrophages during clottingwell as genes encoding growth factors. Thus ADP induces and wound healing, is a mitogen for cells of mesenchymalrelease of paracrine, and positive and negative autocrine, origin, a chemoattractant for smooth muscle cells and mono-growth factors. These observations provide evidence that a cytes, and a promoter of collagen, proteoglycan, and elasticsingle type of renal epithelial cell can generate diverse physio- fiber protein synthesis [54]. A dimeric glycoprotein of Mr 30,000logic messengers in response to one signal. The capacity of comprised of A and/or B chains linked by disulfide bonds [53],infused adenine nucleotides to enhance recovery from nephro- PDGF is most commonly found as a heterodimer (AB), al-toxic, ischemic, and obstructive renal injury might be mediated though biologically active homodimers (AA and BB) arein part by the capacity of the nucleotide to initiate mitogenesis known. The amino acid sequence of the B chain is nearlyand release a variety of growth factors [33, 48—51]. identical to the putative transforming protein of the simian Interleukin-1 /3 stimulates cell growth and induces expression sarcoma virus, which is encoded by the v-sis oncogene; theof gro/MGSA. Interleukin- 1, a protein released by mesangial cellular homologue is termed c-sis [55, 56]. Three classes ofcells, macrophages, and monocytes, has diverse biologic effects PDGF receptors have been defined, each of which comprises[62, 631. It stimulates proliferation of mesangial cells, lympho- two subunits located in the plasma membrane [57, 58]. The acytes, fibroblasts, and glioblastoma cells, inhibits the growth of and 13 receptor subunits are brought together by the presence ofcertain tumor cells in vitro, and contributes to the immune one of the three isoforms of PDGF to form the mature receptorresponse, inflammation, and hematopoiesis. This cytokine also dimer. The receptor a-subunit can bind PDGF-A chain orstimulates DNA synthesis in BSC-1 cells [34]. Interleukin-1j3 PDGF-B chain homodimers, whereas the /3-subunit can bindrapidly induces expression of the gro/MGSA gene (Fig. 4), only PDGF-B chains. whereas two other mitogens, ADP and serum, do not. The Conditioned medium from quiescent BSC-l cells was assayedgro/MGSA gene product has diverse growth-regulatory activi- for its mitogenic activity on normal rat kidney (NRK) fibro-ties and has been identified in three different settings: as the gro blasts, and its chemoattractant activity was assessed on bovinegene, which is expressed in tumorigenic but not nontumongenic aortic smooth muscle cells [50]. Both activities were found, andhamster fibroblasts [64]; as MGSA, the melanocyte growth- both could be blocked by human anti-PDGF IgG. The gene forstimulating activity secreted by melanoma cells in culture [65]; the B chain of PDGF (c-sis) was constitutively expressed inand as the KC gene product, which is expressed rapidly in these cells, whereas no PDGF A-chain mRNA was detected.PDGF-stimulated fibroblasts [66]. A recent study in our labo- Thus the cells appear to synthesize and secrete a PDGF-likeratory showed that purified MGSA initiates DNA synthesis in protein that is a BB homodimer. In this system, ADP increasedquiescent cultures of BSC-l cells (unpublished results). These by threefold the amounts of PDGF B-chain mRNA and secretedobservations suggest that the mesangial cell autocrine growth Nephrology Forum: Regeneration after acute tubular necrosis 233 9ADP

Paracrine growth factor Egr-1 c-myc c-fos c-Ha-ras

DNA:synthesis c-sis

Transcriptional reaulation N

Fig.5. Summary of growth-regulatory events in BSC-I cells following exposure to ADP. The ADP binds to the plasma membrane (PM) and subsequently acts in the nucleus (N) to stimulate the expression of genes that encode growth-stimulatory and -inhibitory factors, as well as transcriptional regulators. The ADP also stimulates rapid release (broken line) of growth factors such as PDGF-like protein, TGF-13, and a novel autocrine growth stimulator, an ADP growth factor (ADP GF). factor interleukin- 1 p[63]induces renal epithelial cells to ex-likely represents the cellular analogue of the protein product of press the gene for MGSA, which encodes a protein that is alsothe viral erb-B oncogene [69]. In BSC-l cells, the number of an autocrine growth factor. EGF receptors is approximately tenfold higher in sparse cultures than at high cell density, although receptor affinity appar- Growth factors as possible therapy in acute renal failure ently does not change [70]. In vivo, a renal insult could unmask Growth factors identified in renal tissue that could contributeEGF receptors on injured or intact renal cells when adjacent to regeneration after acute renal failure via autocrine, para-cells detach from the tubular basement membrane. Presumably crine, or juxtacrine mechanisms are described in Table 3.these surviving cells would be more responsive to the mitogenic Administration of purified growth factors as pharmacologiceffect of EGF produced at the site of injury. Following cell agents alone or in combination eventually might prove useful inproliferation and reepithelialization of the nephron, the recep- treating patients with this condition. tors again would be down-regulated. Epidermal growth factor. A potent renal epithelial cell mito- The role of EGF in regeneration after acute renal injury and gen, EGF is a single-chain peptide of 53 amino acids cross-as a pharmacologic agent has been investigated recently. The linked by 3 disulfide bonds [67]. The preproEGF mRNA inrenal synthesis of EGF falls rapidly after acute ischemic injury murine kidney is translated to preproEGF. The function ofin the rat; expression of preproEGF mRNA is reduced by 60% renal preproEGF is not known, but intracellular processing toat 2 hours and is undetectable by 24 hours, by which time the mature growth factor apparently does not occur [23].urinary excretion of EGF protein has declined by 97% [711. In Urinary EGF might be derived by proteolytic cleavage of thecontrast, expression of the early response genes (c-fox, Egr-l preproEGF protein localized in the plasma membrane. In theand c-myc) increases. It is important that binding of [125j] EGF mouse kidney, preproEGF mRNA is found in the distal neph-increased 3.6-fold in renal cortical tissue and 2.5-fold in the ron and loop of Henle [23], but we were unable to detect themedulla 24 hours after injury [71]. In a cis-platin model of acute mRNA by in situ hybridization in normal human renal tissuetubular injury, decreased preproEGF mRNA expression was (unpublished observations). The EGF receptor is found in thefound in cells of the distal convoluted tubule and thick ascend- basolateral membrane of the proximal tubule and collectinging limb of Henle's loop 12 to 72 hours after injury, whereas duct, and to a lesser extent in distal tubules and glomeruli ofsalivary gland preproEGF mRNA expression was unaffected rabbit nephrons [68]. The EGF membrane receptor protein[72]. Thus acute tubular injury initiates a specific decline in 234 NephrologyForum: Regenerationafter acute tubular necrosis

Table3. Growth factors that possibly participate in renal regeneration Factor Site of production Site of action Action Comment EGF Thick limb of Henle's loop Basolateral membrane Mitogenesis PreproEGF synthesis decreases Distal convoluted tubule Proximal tubule Increased type-I collagen in ARF Collecting tubule synthesis Speeds recovery of ARF in rats TGF-a Mesonephric tubules 9 Transformed phenotype Binds to EGF receptor Renal carcinoma cells TGF-p Collecting duct—cow 9 Mitogenesis-fibroblasts Hyperplastic response becomes Distal tubule—mouse Antimitogenic for renal hypertrophic epithelial cells Transforms with TGF-a or EGF Increases ECM synthesis IGF-I Medullary collecting duct—rat Basolateralsurface Hypertrophy of PCT Increases mRNA synthesis after throughout kidney cells in culture uninephrectomy Increases GFR and RBF Increases protein and mRNA synthesis in early DM in rats IGF-H Throughout kidney, 100 times more Activates PLC in fetal kidney PDGF Mesangial cells Mesangial cells Mitogenesis of mesangial Possible role in glomerular cell BSC-l cells Fibroblasts cells, SMC, fibroblasts growth Fibroblasts SMC Chemotactic for SMC Macrophages fibroblasts, SMC, Platelets macrophages FGF Isolated from renal tissue; cell of Fibroblasts Mitogenesis Embedded in ECM origin uncertain Myoblasts Chemotactic for vascular ? if in tubular basement Endothelial cells endothelial cells membrane in kidney Induces release of collagenases Low-K BSC-l cells BSC-l cells Mitogenesis Isolated from BSC-1 GF conditioned medium Low-Na BSC-l cells BSC-l cells Mitogenesis Isolated from BSC- I GF conditioned medium Renin/Al! JG cells Proximal tubule cells Low dose: hypertrophy Liver in culture and increased Na transport IL-i Mesangial cells Mesangial cells Mitogenesis IL-i mRNA in BSC-l cells Lymphocytes U Abbreviations:All, angiotension II; ARF, acute renal failure; DM, diabetes mellitus; ECM, extracellular matrix; GFR, glomerular filtration rate; GF, growth factor; IL, interleukin; JG, juxtaglomerular; PLC, phospholipase C; PCT, proximal convoluted tubule; RBF, renal blood flow; SMC, smooth muscle cells.

renal EGF gene expression, whereas surviving cells exhibit Transforming growth factor-n. A 50-amino acid polypeptide, increased affinity for the growth factor. TGF-a exerts its biologic effects by binding to the EGF receptor Three recent publications provide strong evidence that phar-[221. Recent evidence that the precursor form of TGF-a is macologic administration of EGF accelerates recovery frombiologically active suggests that both the processed soluble and experimental acute renal injury in the rat. An 8-day intrarenalmembrane-bound precursor forms of the growth factor might be infusion of EGF, or a single subcutaneous injection of themitogens, by autocrine and juxtacnne mechanisms respectively growth factor, each dampened the rise in serum creatinine(Fig. 3). Because TGF-a protein is secreted by developing following hypoxic renal injury and stimulated [3H] thymidinemouse mesonephric tubules, the major physiologic role of the incorporation into DNA in renal tubular cells 24 hours after thegrowth factor is thought to be in embryonic development [76]. insult [73, 74]. A similar augmentation of the repair processRenal carcinomas and other human tumors produce TGF-a, following subcutaneous administration of EGF was demon-which sometimes can be recovered from the urine of these strated in a model of mercuric-chloride-induced acute renalpatients [76]. The growth factor is not produced by renal failure [75]. Histologic examination showed that EGF treatmentepithelial cells of the BSC-l line (unpublished observations), did not alter the extent of tubular necrosis at 24 hours. Thus, inalthough a small amount of mRNA for this growth factor is three different models of acute tubular injury, EGF given soondetected in normal adult kidney [77]. The role of TGF-a in renal after a hypoxic or nephrotoxic insult enhanced regeneration ofrepair after injury has not been defined. Because this growth renal epithelial cells and recovery of glomerular filtration. factor utilizes the EGF receptor, we anticipate that infusion of Nephrology Forum: Regeneration after acute tubular necrosis 235

TGF-t would be as effective as EGF in hastening recovery fromII, and both have a high degree of homology with human acute tubular necrosis. proinsulin [91]. Transforming growth factor-$. The autocrine release of Two types of IGF receptors have been described [92]. Type-I growth-inhibitory activity was first inferred by Holley andreceptors bind IGF-I with an equal or greater affinity than coworkers in 1978 from observations made on crowded culturesIGF-II and bind insulin with low affinity, whereas type-Il of BSC-l cells [78]. They noted stimulation of [3H] thymidinereceptors bind IGF-II with greater affinity than IGF-I and do incorporation into DNA when conditioned medium containingnot bind insulin. The type-I receptor is structurally homologous serum was aspirated and replaced by fresh, serum-free medium.to the insulin receptor, with two disulfide-linked subunits that As no serum growth factors were added to the cells, it appearedbind the peptide, and two subunits that have intrinsic tyrosine that DNA synthesis was initiated by removal of a cell-derivedkinase activity. The type-Il receptor is a single polypeptide that growth inhibitor. Subsequent purification of the inhibitor pro-is identical to the mannose 6-phosphate receptor, a membrane tein indicated that it could arrest sparse cultures of growingprotein that directs proteins to lysosomes [93]. The type-I cells in the G1 phase of the cell cycle, but proliferation resumedreceptor likely mediates the mitogenic response to both IGF-I when the protein was removed [79]. This reversible effect onand II in most systems. The physiologic role for the type-Il cell growth, and the low concentration required for inhibition,receptor in mediating the tissue response to IGF-II is unknown suggested a physiologic role for the inhibitor in the control of[92]. The distribution of IGF receptors has been studied on proliferation. The growth-inhibitory effect could be overcomeproximal tubular cells isolated from canine kidney [94]. Insulin- by the addition of EGF to the culture medium; therefore,like growth factor-I-stimulated phosphorylation of the IGF-I proliferation of renal cells in culture can be regulated by factorsreceptor has been demonstrated only in preparations of baso- with opposing effects [80]. In 1984, the growth inhibitor waslateral membranes, and not in those from brush border. Insulin- reported to be biologically and chemically similar to TGF-13like growth factor-Il receptors appear to be equally distributed [81]. The complete amino acid sequence of the BSC-l cellon basolateral and brush-border membranes. growth inhibitor, determined from the nucleotide sequence of The site of production of IGF-I was originally considered to be the liver, from which it is released in response to growth the cDNA, has been shown to correspond to TGF-132 [39]. hormone [89]. However, evidence indicates that IGF-I is pro- Human TGF-/31 is a polypeptide with an apparent Mr of 25,000,duced by most, if not all tissues, and acts as an autocrine or composed of two chains linked by disulfide bonds. There is 71% paracrine effector [951.Productionof IGF-I and its growth- identity between the amino acid sequences of monkey TGF-/32hormone dependence have been demonstrated in rat kidney as and human TGF-/31 [82], closely related proteins with similarwell as in other rat tissues [96]. Insulin-like growth factor-I properties. mRNA is also present in adult human kidney, although its In most cells, TGF-f3 mRNA is constitutively expressed,distribution within the organ has not been determined [97]. In although production of the protein as detected by immunohis-rat kidney, immunoreactive IGF-I is localized primarily in the tochemistry appears limited in the adult mouse to cortical distalmedullary collecting duct and is not found in the proximal or tubular cells, and in bovine kidney to collecting duct [83, 84]. Indistal tubules [98]. Administration of IGF-I to fasted rats renal epithelial cells in culture, TGF-/3 inhibits [3H] thymidineincreases renal plasma flow and GFR while decreasing renal incorporation, whereas it stimulates growth of fibroblast colo-vascular resistance [99]. Insulin-like growth factor-Il is less nies in soft agar [81]. In primary cultures of rabbit proximaldirectly under growth hormone control and is believed to be an tubular cells, TGF-/32 can convert the mitogenic effect usuallyimportant fetal growth factor [100]. Insulin-like growth factor-Il exerted by insulin and hydrocortisone into a hypertrophic one;mRNA is present both in fetal and adult human kidney; this finding suggests a role for the factor in compensatory renalhowever it is 10- to 100-fold more abundant in the fetal organ. growth after uninephrectomy [85]. Also, TGF-/3 induces syn- Increased immunoreactive IGF-I has been demonstrated in thesis of fibronectin and its incorporation into ECM by lungkidneys after ischemic injury [101], and this growth factor can epithelial cells and fibroblasts in culture [86]. In a rat model ofinduce hypertrophy of proximal tubular cells in culture [102]. wound healing, TGF-p2 increased synthesis of type-I procolla-How IGF-I reaches proximal tubular cells in vivo remains gen by fibroblasts, and hastened the rate of healing [87]. Anuncertain, because its site of production is in the terminal antiserum to TGF-f3l appears to be efficacious in the treatmentcollecting duct. The contribution of IGFs to renal growth and of a murine model of mesangial proliferative glomerulonephritisrepair after injury remains to be clarified. [88]; this growth factor therefore might play a critical role in the Platelet-derived growth factor. Mesangial cells secrete a accumulation of basement membrane components in specificPDGF-like protein and express PDGF receptors [103]. Glomer- glomerulonephritides. ular endothelial and mesangial cells proliferate in response to Insulin-like growth factors. The IGFs, also termed so-PDGF; the factor thus could play a role in growth responses in matomedins, are peptide growth factors that were first isolatedthe glomerulus [103]. Although PDGF is not a mitogen for renal from human plasma fractions. The IGFs share the followingepithelial cells, it could contribute to regeneration after acute biologic activities: insulin-like activity on glucose metabolism,renal failure by attracting infiltrating inflammatory cells and stimulation of sulfate incorporation into cartilage, and a mito-stimulating adjacent fibroblasts to release factors that are ii- genic effect on different cell types [89]. Insulin-like growthgands for epithelial cell receptors. In this way PDGF might factor-I, also known as somatomedin C, is a basic, single-chaincoordinate tubular and glomerular proliferation after renal polypeptide of 70 amino acids with an apparent Mr of 7649 [90].injury. A slightly acidic peptide of 67 amino acids, IGF-II has an Fibroblast growth factors. The FGFs are two closely related apparent M of 7471; there is 62% identity between IGF-I andpolypeptides with different isoelectric points, designated as 236 NephrologyForum: Regeneration after acute tubular necrosis

. . BASEMENT MEMBRANE

Fig. 6. Growth regulation in regenerating renal epithelial cells. Cells at the edge of an injured segment of the nephron are shown during early renal regeneration following acute tubular necrosis. Pictured are non-necrotic cells, a dividing cell, and a migrating squamoid cell. Proliferation may be mediated in part by synthesis of growth-stimulatory factors, followed by cellular processing and release of the active molecules. Symbols for growth factors, precursors, and receptors are as in Figure 3. acidic FGF (aFGF) and basic FGF (bFGF). The cDNA clones Angiotensin II. Angiotensin converting enzyme (ACE) inhib- encoding both proteins have been isolated, and the predicteditors block compensatory renal hypertrophy alter partial renal amino acid sequences indicate 55% structural homology be-ablation in rats [110]. Although this observation has been tween these two growth factors [1041. The FGFs are mitogenicascribed to changes in intrarenal hemodynamics, direct cellular for most, if not all, nonterminally differentiated cells of meso-effects of angiotensin II have been defined. Angiotensin II can dermal or neuroectodermal origin. Also, FGFs are potentinduce hypertrophy of smooth muscle cells in culture [111], and inducers of blood vessel growth. They are chemotactic forit can stimulate sodium reabsorption both in intact nephrons vascular endothelial cells, and they induce expression of plas-and in cultures of proximal tubular cells, possibly via stimulated minogen activators and collagenases, proteolytic enzymes thatNa/H antiporter activity [112, 113]. In mouse proximal tubular presumably mediate tissue remodeling. cells of the MCT line, exposure to 10_8 M angiotensin II The FGFs bind to heparan sulfates, the principle molecules ofincreased cell size and protein synthesis; these effects were extracellular matrix in tissues and cell culture [1051. Theirblocked specifically by saralasin [114]. In MCT cells pretreated presence in extracellular matrix suggests that they can mediatewith EGF, the hypertrophic action of angiotensin II became a proliferation of cells adjacent to the basement membrane. Forproliferative one. Relatively high doses of amiloride blocked example, FGF is not secreted into the medium by endothelialangiotensin Il-induced protein synthesis and partially abolished cells, but FGF does appear to operate following its depositionthe increase in cell size; thus, increased cell sodium content into extracellular matrix. In this manner, FGF could be a localmight be required for hypertrophy to occur after stimulation growth regulator and could induce regeneration after denuda-with angiotensin II [114]. tion of the overlying cell layer. This property might explain Plasma renin activity is increased in humans and animals with some of its ability to promote wound healing. It has beenacute renal failure, and elevated intrarenal angiotensin II levels difficult to define the mechanism by which FGF is depositedcould contribute to abnormal autoregulation of blood flow in the into the extracellular matrix, as both aFGF and bFGF lack arecovering kidney [115, 116]. The direct renal cellular effects of classic leader sequence required for cellular secretion [104,angiotensin II on sodium transport and hypertrophy suggest 106]. It is not known whether FGF is released only fromthat this protein could have physiologic significance during damaged or dying cells, is delivered to sites of injury ortubular regeneration after injury. inflammation by stimulated macrophages, or is secreted by an as-yet-uncharacterized mechanism [106]. Conclusions Until recently, aFGF was thought to exist only in neural A major theme of this Forum is that renal cells release tissue, but both acidic and basic FGF have been isolated fromautocrine and paracrine growth factors in response to extracel- bovine kidney [107, 1081. Studies in embryonic kidney suggestlular signals and could thereby mediate repair of the nephron a role for FGF in renal angiogenesis [109]. During embryonicafter acute renal failure (Fig. 2). Examples of such signals are development, renal differentiation is driven by the interaction ofreductions in the extracellular potassium and sodium concen- different types of cells. Contact with the epithelial ureteric budtration, liberation of adenine nucleotides by injured or dying induces the nephrogenic mesenchyme to differentiate, whichcells, and cytokines released by neighboring cells (Fig. 4). As then leads to stimulation of blood vessel ingrowth. Little workgrowth factors alter ion and nutrient transport and other aspects has assessed the effects of FGF on kidney cells, although aFGFof cell metabolism, these effector molecules could mediate stimulates proliferation of BSC- 1 cells (unpublished observa-repair of biochemical and/or subcellular injuries and thereby tions). Whether FGF is present in glomerular or tubular base-allow an injured cell to recover, and subsequently to respond to ment membranes or participates in renal cell regeneration afterautocrine, paracrine, orjuxtacrine proliferative signals. Epithe- injury is not clear. hal cells at both ends of a necrotic segment of the nephron, and Nephrology Forum: Regeneration after acute tubular necrosis 237

other surviving isolated cells along its denuded surface, areinformation presented at this Forum provides a rationale for responsible for regenerating the tubular epithelial cell liningconsidering specific growth factors in combination with suffi- [14]. Proliferation of these cells with subsequent sliding alongcient nutrients, calories, and dialytic therapy to optimize recov- the basement membrane permits rapid restitution of the tubule.ery. An important goal of future research efforts is to identify Growth factors could stimulate the transition of these cells fromnew renal growth factors, and then to isolate and prepare them the G0 toG1 phase of the cell cycle (Fig. 6), and possibly speedfor use in combination with known factors to speed renal protein and phospholipid synthesis for the formation of brush-regeneration and improve the outcome in patients with acute border and basolateral membranes. The intact basement mem-renal failure. brane plays an important role in the healing process by providing a structural framework for the proliferating cells. Alterations in the Questions and answers extracellular matrix and its connections to the cytoskeleton occur with acute tubular injury and likely result in profound disruption of DR. JOHN C. LIESKE (Fellow, Section of Nephrology, The cellular function. Because several growth factors stimulate cells toUniversity of Chicago, Chicago, Illinois): Of the growth factors synthesize and release extracellular matrix components such asdiscussed, only EGF has been demonstrated to alter the course collagens, fibronectin, and tenascin, repair of the damaged base-of acute renal failure when administered systemically after an ment membrane also might be mediated by molecules of renal cellinsult in experimental animal models. However, intrarenal origin. At present the factors governing repair of the extracellularlevels of EGF mRNA and production of urinary EGF both matrix and cytoskeleton following acute tubular injury remaindecline precipitously during acute renal failure. What is the role unknown. of EGF in recovery after renal injury in vivo? Growth factors delivered to renal epithelial cells from local DR. TOBACK: The points you make argue against local release and systemic sources and acting through more than one mech-of EGF as a key mediator of recovery after acute renal failure. anism may orchestrate the proliferative repair of the nephronIf autocrine production of EGF plays a role in renal regenera- after cellular necrosis. Both IGF-I and EGF are produced bytion, one would have to propose it does so by binding to an renal epithelial cells; factors made by mesangial cells, such asincreased number of EGF receptors, as has been demonstrated PDGF and interleukins, also might contribute to tissue repair.in rat renal tissue after ischemic injury [711. The experimental The FGFs, which are found in the extracellular matrix, may besuccess of pharmacologic doses of EGF in attenuating the a stored form of growth factor for surviving cells that becomescourse of acute renal failure in rats appears to be mediated by operational after cell necrosis. Transforming growth factor-f3,the greater number of receptors as well. Another source of EGF which can convert a proliferative signal into a hypertrophic one,in vivo could be the systemic circulation; the plasma concen- may facilitate development of the nascent cells into a maturetration is in the range of 20 pg/mt [671. This relatively low epithelium, and act as a negative feedback signal to limitconcentration may achieve physiologic importance in the set- proliferation as repair of the nephron nears completion. ting of acute renal failure because the number of EGF receptors Renal growth factors also might act on a population of renalper cell is increased. cells arrested in the G2 phase of the cell cycle. Described by The physiologic function of urinary EGF remains undefined, Pederson and Gelfant in 1970 [117], these cells could be rapidlyas does the mechanism by which it enters the urine. Even induced to divide by paracrine or autocrine factors, therebythough the mRNA that encodes preproEGF is detected in the beginning replacement of cells that detached from the tubularthick loop of Henle and distal nephron in mice, and EGF can be basement membrane. Mesangial, endothelial, and/or infiltratingdetected immunohistochemically in the apical regions of these inflammatory cells at the site of renal injury represent additionalcells, little EGF can be extracted from renal tissue [67]. The sources of growth factors that could aid cells of the injuredamount of EGF excreted in human urine each day (approxi- nephron [118, 1191. Growth factors and possibly adenine nude-mately 50 zg) is too large to be accounted for by efficient renal otides released by regenerating and injured renal epithelial cellsclearance of the growth factor. At present it is thought that also could stimulate fibroblast cell growth. In addition, tubularpreproEGF is inserted in the luminal plasma membrane, where and fibroblastic cells may be stimulated by autocrine and/orit is subsequently cleaved by urinary proteases to release the paracrine mechanisms to synthesize and secrete extracellularmature growth factor [67]. The potential actions of EGF in matrix proteins, which could repair rents in the tubular base-distal tubular fluid are a matter of speculation. One possibility is ment membrane. As I pointed out earlier, it is interesting thatthat EGF stimulates proliferation of epithelial cells to replace the exuberantly proliferating surviving cells are located in thethose that have detached or died as a result of the acidic pH and region of a disrupted basement membrane (Fig. 1). Althoughrelatively high concentration of toxic metabolites in the lumen growth factors may have induced proliferation at this site, theyof the distal nephron, ureters, and bladder. apparently were unable to accomplish repair of the damaged There could be a practical problem with administering phar- basement membrane in this patient. macologic doses of EGF to patients in an effort to speed The National Kidney and Urologic Diseases Advisory Boardrecovery after acute renal failure. TGF-a, which binds to the 1990 Long-Range Plan concluded that acute renal failure is theEGF receptor, causes calcium resorption from mouse calvaria most costly kidney condition requiring hospitalization, that thein culture, and subcutaneous EGF and TGF-a administered number of cases has increased recently as a consequence ofover a 2-day period to mice induced mild hypercalcemia [122, complicated surgical procedures in older patients, and that the123]. The mechanisms of bone resorption and hypercalcemia mortality rate has not changed since the early 1950s [120, 1211.remain to be defined, but these might be consequences of the Thus, a considerable need exists for new therapeutic ap-interaction of EGF with the parathyroid hormone receptor proaches that can speed regeneration and reduce mortality. The[124]. Hyperphosphatemia is often found in acute renal failure; 238 NephrologyForum: Regeneration after acute tubular necrosis

thus the risk of dystrophic and metastatic calcification subse-low-potassium growth factors is that neither of them stimulate quent to further elevations in the calcium-phosphorus productthe growth of fibroblasts. Given that very limited information, after EGF administration ultimately might limit the clinicalthey appear relatively specific as mitogens for renal epithelial utility of the growth factor. cells, but that's the extent of our knowledge at the moment. DR. NICOLAOS E. MADIAS (Chief, Division of Nephrology, As to Jean Oliver's work, it is of interest that the most New England Medical Center, Boston, Massachusetts): Al-marked accumulation of cells noted in the microdissected though not well substantiated for ischemic acute renal failure,nephron from the patient described today (Fig. I) occurs in a for some models of nephrotoxic acute renal failure, specificallyregion of the tubule that exhibits tubulorrhexis; these prolifer- those induced by heavy metals or glycerol, it appears that theating cells overlie a disrupted portion of the basement mem- first insult confers on the kidney resistance to a second expo-brane. We know from our work using ADP as a mitogen, and sure to the same insult. Does modern biology offer any insightfrom studies by others, that as cells proliferate they express the into this phenomenon? genes that encode and secrete extracellular matrix proteins DR. TonAcK: The molecular basis for this observation is notsuch as collagen, fibronectin, and laminin [127]. Incorporation known. Animals recovering from certain types of acute renalof these proteins into the tubular basement membrane could failure often are resistant to a second exposure to the samefacilitate its repair after injury. If the patient with acute renal agent, for example, glycerol, uranium, or gentamicin [125]. Thisfailure that we are discussing here had been hospitalized today, is not the case in animals subjected to ischemic renal injury;she would have received nutritional and hemodialysis support. they are more susceptible to acute renal failure following anThis treatment might have provided sufficient time for these additional ischemic insult. Some agents confer resistance to aproliferating cells to have secreted enough extracellular matrix different nephrotoxin, whereas others do not. For example,material to repair the damaged tubular basement membrane glycerol-induced injury protects against exposure to mercuric(Fig. 1), and to have allowed eventual return of her renal chloride, but mercuric chloride does not protect against treat-function. In the rat, it appears that an intact tubular basement ment with gentamicin. membrane provides a scaffold on which regenerating epithelial Several factors may play a role in this phenomenon. Therecells migrate [2, 4,5].The cells move from both ends of the appears to be decreased sensitivity of mesangial cells to angio-damaged segment of the tubule until they meet and reepithe- tensin II in glycerol-induced acute renal failure. This couldlialize the nephron. Intact, embryonic rest or "stem" cells that blunt angiotensin-mediated mesangial contraction and possiblysurvived the renal injury also can proliferate, migrate, and minimize the expected decline in glomerular filtration thatcontribute to the repair process. The presence of acidic and would follow a second dose of glycerol [125]. In uranium-basic FGF in the extracellular matrix [128] may serve as an poisoned dogs, repopulation of degenerated tubules is accom-available reservoir of growth factors that could initiate repair plished by recovery of injured cells, mitosis, and ingrowth ofand proliferation of cells after injury, prior to the synthesis and tubular cells at the wound margin [126]. In this model, it is thesecretion of new molecules for this purpose. migrating new cells that appear resistant to injury by a second DR. LIESKE: Evidence is increasing that alterations in the exposure to uranium. Resistance of these regenerating cellscytoskeleton of renal cells might be responsible for functional might be a consequence of their decreased capacity to take upabnormalities in the nephron after acute injury. Is there any and concentrate the nephrotoxin. A solute diuresis in thisevidence that growth factors play a role in the formation or setting could have a salutary effect by preventing accumulationrepair of the cytoskeleton? of nephrotoxins in the kidney, and/or by relieving tubular DR. TOBACK: Connections between the extracellular matrix obstruction by casts of necrotic cells. and the cytoskeleton appear to be important in embryogenesis Along the lines discussed today, one could hypothesize thatand during morphologic differentiation of the renal epithelium. local production of growth factors occurs during the initial renalEnzymes, receptors, and carriers show an asymmetric distri- insult to facilitate recovery of injured cells and to stimulatebution between the apical and basolateral surfaces of polarized mitogenesis to replace necrotic ones. A second exposure to theepithelial cells, and the lipid content and physical properties of nephrotoxin might rapidly trigger synthesis and release of thethese two plasma membrane domains show distinct differences same factors and result in the brisk repair of the tubular lesion;[9].Theestablishment of asymmetry initially might involve this repair would be interpreted as relative resistance to thecell-cell contacts via surface glycoproteins called cell-adhesion second insult. Perhaps in this instance renal cells act like cellsmolecules, such as uvomorulin. Also important in the develop- of the immune system when confronted with an antigen thatment of cell polarity are contacts between the cell and the was encountered previously; they exhibit an anamnestic re-extracellular matrix. Cellular receptors for the extracellular sponse. matrix molecules laminin, collagen, and fibronectin belong to DR. JOHN T. HARRINGTON (Chief of Medicine, Newton-the integrin supergene family [129]. Laminin is secreted into the Wellesley Hospital, Newton, Massachusetts): I have two ques-basement membrane of cells destined to become renal epithe- tions: First, do the low-potassium and the low-sodium growthhum [130]. The gene encoding the A chain of laminin is factors stimulate renal cell growth only, or do they make cells inconstitutively expressed, whereas expression of B chain mRNA other organs of the body grow at a comparable rate? My secondprecedes commitment to an epithelial cell morphology in organ question refers to the work of Jean Oliver. As I recall, Oliverculture [1311. Furthermore, antibodies to the A chain can distinguished between simple cell death and what he calledprevent development of polarity in these cells. Production of tubulorrhexis, that is, destruction of the basement membrane.type-IV collagen also appears to be linked to development of Does that old distinction have any relevance today? renal epithelial cell polarity, as undifferentiated mesenchyme DR. TOBACK: The data we have about the low-sodium andproduces only type-I and -III collagen [132]. Syndecan, a Nephrology Forum: Regeneration after acute tubular necrosis 239

cell-surface proteoglycan, behaves as an ECM receptor. Itkidney after acute renal failure in the setting of systemic associates intracellularly with actin and is localized to theacidosis. basolateral surface of epithelial cells [133]. The extracellular Four growth factors—EGF, TGF-J3, PDGF, and FGF—each domain of the syndecan molecule contains heparan sulfatecan enhance synthesis of extracellular matrix proteins through chains, which can reversibly bind bFGF. Mammary epithelialdifferent mechanisms [67, 86, 141, 142]. As there are direct cells made syndecan-deficient with antisense cDNA lose epi-connections between the extracellular matrix and cytoskeleton thelial polarity and assume a fibroblastic morphology. via the cell proteins ankyrin and fodrin, growth factors could, Changes in the organization of cytoskeletal actin filamentsperhaps indirectly, facilitate repair of the cytoskeleton as well. represent early evidence of polarity in renal epithelial MDCKClearly the precise role of growth factors in repair of the cells in culture. Progressive organization of the actin network,cytoskeleton after acute injury remains to be defined. from the apical and basolateral surfaces, accompanies develop- DR. ANDREW KING (Division of Nephrology, New England ment of polarity in these cells during the first 12 hours in cultureMedical Center): I'm very interested in the idea of increased [9].Microtubulesalso appear to organize sequentially over fivegrowth factors with counterbalancing inhibitory factors. Has days. If formation of cell-cell contacts between MDCK cells inanyone measured growth factors in urine? culture is prevented by incubation in low-Ca medium, the DR. TOBACK: About 50 g of epidermal growth factor is polarized distribution of certain plasma membrane proteins isexcreted in the urine of normal humans each day [67]. This prevented [134]. Alternatively, other proteins still distributerepresents a large amount of a factor that is biologically active. normally if adhesion is prevented with low-Catmedium,butThe source of urinary EGF is not clear. It is small enough (Mr, not if formation of actin microfilaments is disrupted with6045) to be easily filtered, but given the low concentration in cytochalasin D [135]. Thus more than one mechanism ofplasma, the amount in urine is too large to be accounted for by localization of cellular proteins to specific regions of the plasmafiltration alone. The remainder likely originates from the kid- membrane exists, ney. It is not known exactly how EGF gets into the urine, but Alterations in cytoskeletal components can result from renalone possibility is that urinary proteases cleave precursor EGF injuries not severe enough to cause cell death. Permeability ofmolecules that protrude from the apical surface of cells, thereby tight junctions, as measured by ruthenium red penetration ofreleasing the growth factor into the tubular lumen. microperfused tubules, increases in a stepwise manner after 5, DR. KING: I would think that filtered factors would be 15, and 30 minutes of ischemia [136]. This phenomenonsusceptible to tubular endopeptidases. is accompanied by the abnormal redistribution of Na-K DR. TOBACK: In principle you are right. Epidermal growth ATPase from the basolateral into the apical membrane, indicat-factor does appear in the urine in two forms of 51 and 53 amino ing loss of so-called "fence" function mediated by tight junc-acids in length, so the extent of cleavage of this growth factor tions that separate apical and basolateral membrane domains.appears modest [67]. Perhaps the tertiary conformation of the Other changes of the cytoskeleton with ischemia include loss ofprotein confers relative resistance to protease action. brush-border terminal-web actin filaments after 15 minutes and It is interesting that the amount of mRNA encoding the EGF diffuse redistribution of actin throughout the cytoplasm after 50precursor declines progressively from 2 to 24 hours after the minutes [137]. Disrupting microfilaments of perfused tubulesonset of ischemic acute renal failure in the rat [71]. Concur- with cytochalasin D decreases their capacity for sodium reab-rently, the quantity of EGF protein in the urine falls to nearly sorption [137]; this finding suggests that morphologic alterationszero. Thus increased local production of EGF does not appear of the cytoskeleton after ischemia correlate with functionalto mediate nephron repair after injury. On the other hand, the derangements of the nephron. number of EGF receptors per renal cell rises approximately Changes in renal epithelial cell pH are associated withthreefold; thus, exogenous EGF delivered to the cells via the mitogenesis, cytoskeletal alterations, and modified receptorplasma could readily stimulate them to proliferate. The in- number. Renal mass increases in rats made chronically acidoticcreased number of receptors might be an important reason why by ammonium chloride administration [138], and acidification ofexogenously administered EGF attenuates the course of acute the culture medium is associated with mitogenesis of primaryrenal failure [73, 74]. Also, perhaps the pharmacologic dose of cultures of mouse renal tubular cells, an effect not observed inEGF administered in these experiments exceeds the capacity of hepatocytes or fibroblasts [139]. Stimulation of growth withserum and urinary proteases to degrade the growth factor so acidosis is preceded by an increase in intracellular sodium,that it reaches the tubular cells intact and induces proliferation. possibly mediated by the Na-H antiporter [139]. Acidifica- DR. RONALD PERR0NE (Division of Nephrology, New En- tion of porcine kidney cells in culture decreases the lateralgland Medical Center): You show nicely that several conditions mobility of vasopressin V2 receptors while increasing receptorstimulate release of growth factors from BSC-l cells in vitro. number [140]. The changes in receptor properties appear sec-Do maneuvers that injure cells in culture, for example, anoxic ondary to observed alterations in cytoskeletal actin, and couldinjury, cause release of growth factors in the same kind of be reproduced by cytochalasin B administration. Thus it ap-conditioned-media experiments that you described for low pears that intracellular acidosis can cause changes in thesodium and low potassium? cytoskeleton of renal epithelial cells, which in turn alter recep- DR. TOBACK: That's a good question. We have not done the tor function on the plasma membrane. It is interesting toexperiment you suggest. In collaboration with Eugene Gold- speculate that renal cell growth observed during extracellularwasser and Margaret Walsh-Reitz at the University of Chicago, acidosis likewise might be mediated through cytoskeletal alter-we did ask whether exposure of BSC-l cells to hypoxia would ations and subsequent changes in growth factor receptors. Thisinduce transcription of erythropoietin mRNA and secretion of phenomenon would be especially relevant to recovery of thethe protein. The answer was no. A recent report, however, 240 NephrologyForum: Regeneration after acute tubular necrosis

indicates that monocyte-derived macrophages exposed to hy-tions suggest that an insufficient amount of preproEGF or EGF poxia release several mitogens for endothelial cells, includingprotein during development might result in formation of collect- basic FGF [1431. ing duct cysts and renal failure. Da. JORDANJ.COHEN (Dean, School of Medicine, State DR. PERRONE: If you were to design a clinical study looking University of New York at Stony Brook, Stony Brook, Newat the efficacy of growth factors in tubular necrosis, what would York): Do growth factors play a role in the renal hypertrophyyou put in the concoction? evoked by uninephrectomy? If so, what do you think triggers DR. TOBACK: The strategy that I have in mind requires that them into action? the patient first be well dialyzed. Nutrition is another critical DR.TOBACK:This is an area of considerable interest, but theissue. It is important to provide calories, although the amount factors that initiate renal hypertrophy after uninephrectomywill depend on the presence or absence of a catabolic state such remain unknown. Insulin-like growth factor-I has been studiedas sepsis or trauma, and the nutritional status of the patient. as a possible mediator of this response because it is made inAdministration of amino acids is valuable to preserve muscle renal cells and induces hypertrophy of proximal tubular cells inmass by minimizing gluconeogenesis so that the infused amino culture [10]. Because the amount of the factor is elevated in theacids will serve primarily as precursors for protein synthesis compensating kidney 5 days after uninephrectomy, but not atrather than as fuel. After addressing these issues, it is worth one day, it appears unlikely that IGF-I initiates compensatoryconsidering several factors that could speed repair and regen- growth. Compensatory enlargement and an increase in renaleration of injured renal cells: (1) thyroxine promotes renal cell IGF-I content occur 7 days after uninephrectomy in dwarf ratsgrowth and clearly stimulates recovery after experimental that are deficient in growth hormone [144]. Thus increasednephrotoxic acute renal failure [151]; (2) infusion of ATP with IGF-I production in renal hypertrophy is independent of growthmagnesium chloride has been shown to hasten recovery after hormone. toxic, ischemic, and obstructive renal injury in animals [1521, Angiotensin II possibly plays a role in renal hypertrophybut its utility in humans remains to be determined, and (3) EGF because angiotensin converting enzyme inhibitors block com-is a growth factor that facilitates recovery after renal tubular pensatory growth after partial renal ablation in rats [110]. Thisinjury in the rat [73—75]. The critical factors that mediate repair has been ascribed to changes in intrarenal hemodynamics, butof tubule structure and function probably have not yet been more recently other direct cellular effects of angiotensin II havediscovered. For example, after we are able to purify a sufficient been defined, such as its capacity to stimulate hypertrophy ofamount of the low-sodium growth factor, we plan to evaluate its proximal tubular cells in culture [114, 145]. potential as a therapeutic agent to accelerate renal regeneration Many attempts have been made to isolate unique substancesin vivo. from blood that mediate compensatory renal growth after DR. HARRINOTON: When speaking about the factors that uninephrectomy. A recent report describes a factor isolatedcontrol growth of renal epithelial cells in culture, you referred from the plasma of female rats 24 hours after uninephrectomyto low-molecular-weight nutrients. To what were you referring? [146]. This factor was partially purified by gel filtration and DR. TOBACK: Glucose and amino acids. high-performance liquid chromatography, and has an Mr of DR. HARRINOTON: Would you speculate on how the signal 17,000 to 22,000. The factor stimulates DNA synthesis in malefrom the various growth factors is transmitted across the cell murine kidneys in vivo and porcine proximal tubular cells of themembrane and intracellularly? LLC-PK1 line in culture. The characteristics of this factor DR. TOBACK:Exactlyhow binding of a growth factor to the appear different from those of known renal growth factors. extracellular domain of its receptor sends a signal to the DR. PERRONE: Polycystic kidney disease is, among otherfunctional cytosolic domain of the molecule is not known. things, a model of increased cell proliferation. I was intriguedWhen the growth factor binds to its receptor, it is thought that about your comments regarding kidney cells being tonicallyconformational changes occur in the ligand-receptor complex inhibited. Are there any instances of cyst formation duringthat activate the receptor's functional domain, which is often a recovery from tubular necrosis? tyrosine kinase. The sequence of the 18 to 20 amino acids of the DR. TOBACK:Cysticchange during recovery after acute renalreceptor protein that span the plasma membrane is of particular failure must be quite rare, if it occurs at all. Investigators haveimportance because mutations in one or more of them can alter sought a role for growth factors in human cystic diseasesreceptor function [1531. The ligand-activated PDGF receptor, because cell proliferation is often detected in the cyst wall [147].for example, associates with key cytosolic mediators such as The concentration of EGF has been measured in fluid obtainedphospholipase C. The receptor then activates the enzyme by from patients with several types of cystic disease and found totyrosine phosphorylation, which subsequently triggers other be intermediate between that of urine and plasma [148]. signal transduction events such as production of inositol A contribution of EGF in cyst formation has been sought intrisphosphate and diacylglycerol [154]. cpk mice, a model of congenital polycystic kidney disease in DR. LIESKE: The plasma concentration of angiotensin II which large cysts develop in collecting tubules of neonatalincreases after acute renal failure. This abnormally high angio- animals. Unexpectedly, the amounts of preproEGF mRNA intensin II concentration appears to play a role in renal vasocon- renal tissue and urinary excretion of EGF protein are decreasedstriction and loss of autoregulation of renal blood flow. On the in affected mice compared with control littermates [149]. Thereother hand, angiotensin II can induce hypertrophy of proximal may be a generalized abnormality in the regulation of the genetubular cells in culture and promote hyperplasia of these cells if encoding preproEGF in cpk mice, because the normal burst ofEGF is present. Increased EGF binding is noted after acute growth factor synthesis in the salivary glands in response torenal failure in rats, so that an elevated angiotensin II concen- testosterone is not seen [150]. Taken together, these observa-tration and increased EGF binding could lead to mitogenesis Nephro/ogy Forum: Regeneration after acute tubular necrosis 241 and hastened epithelial regeneration. ACE inhibitors, by de-phy in the remaining kidney. Do you have any evidence of creasing angiotensin II levels, could have two opposing effectsage-related changes in the production of, or the response to, on renal function after acute renal failure: they could increasegrowth factors? renal blood flow and decrease renal epithelial cell proliferation. DR. TOBACK: Little information is available about this point. Has anyone investigated ACE inhibitors in patients with acuteIn humans the concentration of EGF in urine increases to a renal failure? Which cellular action of angiotensin II is morepeak at 1 to 2 years of age and declines slowly thereafter important in vivo after acute renal failure? throughout adolescence and adulthood [67]. The concentration DR. TOBACK: The role of the renin-angiotensin system in theis higher in females than in males and doubles during preg- pathogenesis of acute renal failure remains controversial de-nancy. EGF in urine and renal tissue of mice is lowest during spite much investigation [155]. Studies of ACE inhibition afterthe first 6 days of life; it then rises gradually to the adult level by acute renal failure have centered on the immediate post-injury3 to 4 months. An autocrine role for EGF in renal organogenesis period. In a dog model of mercuric-chloride-induced renaland development seems unlikely because preproEGF mRNA failure, intravenous captopril preserved renal blood flow andand EGF protein are not detectable in the organ until the GFR without attenuating an elevated FENa when assessed 3neonatal period in rodents. hours after the nephrotoxic insult [156]. Similarly in rats, DR. KURTIN: In ischemic tissues—for example the heart— captopril and saralasin each were able to prevent a decline inADP levels can rise. Is there evidence that ADP can induce renal blood flow 3 hours after glycerol-induced myoglobinuricgrowth factors in those tissues? acute renal failure, but the subsequent course of renal failure, as DR. TOBACK: Adenine nucleotides and magnesium chloride assessed by measurements of BUN and creatinine concentra-have been infused in shock to speed recovery and minimize the tion, was not changed by either agent [157].Thenewly de-extent of tissue injury after ischemia [160]. I am not aware that scribed direct actions of angiotensin H on renal tubular cellany experiments have been performed to look for production of sodium transport and growth may bear on the observation thatan adenine-nucleotide-induced growth factor by myocardial ACE inhibition can prevent deleterious immediate changes incells. renal function after acute injury, but not enhance long-term DR. KURTIN: In chronic renal failure, there is a spectrum of outcome. It could well be that the hemodynamic actions ofrenal cells from normal to damaged to necrotic, and in that ACE inhibition predominate in the early phases of acute renalsetting we find acquired cystic disease. Would you comment on failure, resulting in improved renal blood flow and GFR. Thethe role of growth factors in acquired cystic disease and why growth-promoting effect of angiotensin II on renal epithelialsome cysts undergo malignant degeneration? cells could be blunted by ACE inhibitors despite the salutary DR. TOBACK: The incidence of cell proliferation in the normal effect of the drug in increasing renal perfusion. The negativekidney is very low, so that detection of mitotic cells in the cellular effect might offset the immediate benefits on hemody-shrunken kidneys of patients undergoing chronic dialysis is namic parameters and thereby explain the overall ineffective-unexpected. One would expect only progressive fibrosis. Pon- ness of ACE inhibition in acute renal failure. dering the unexpected observation that cell proliferation occurs Da. MADIAS: In the experiments in which EGF was givenin the end-stage kidney might disclose new pathogenetic mech- exogenously to hasten recovery from acute renal injury, wasanisms that mediate progressive loss of renal function long after mitogenic activity increased in tissues other than the injuredthe initial insult has occurred. I have taken a similar approach in organ? pointing out the value of unexpected observations by citing DR. TOBACK: That's a good question. I'm not aware of anyexuberant cell proliferation in the case of acute renal failure published information on that point. EGF is a very potentdescribed by Oliver (Fig. 1) in this Nephrology Forum. growth factor, and it would not be surprising if it had a What is the stimulus for cell proliferation in the end-stage mitogenic effect on other organs. On the other hand, the kidneykidney? The answer is not known. We recently proposed the might be particularly sensitive to the growth factor because offollowing scenario [25]. Dialysis often is accompanied by wide the increased number of EGF receptors per cell demonstratedswings in the plasma concentrations of sodium and potassium. in the injured nephron [71]. EGF can facilitate wound healing inFurthermore, dialysis patients are often hypo- or hyperna- other tissues. When administered topically to the injured cor-tremic, and/or hypo- or hyperkalemic. Under these conditions, nea, it accelerates healing of the wounded epithelium [158]. the remaining renal epithelial cells might be stimulated to DR.MADIAS: Have you done the reverse experiment to thatproliferate as do BSC-l cells in culture exposed to altered in your low-potassium study, namely, does a high-potassiumconcentrations of electrolytes [28, 31, 32, 159]. Such perturba- environment lead to production of a growth-inhibiting activity?tions in the plasma potassium and/or sodium concentration DR. TOBACK: Raising the concentration of potassium in theduring progression of renal disease and maintenance dialytic culture medium of BSC-l cells stimulates DNA synthesis andtherapy could result in release of autocrine and paracrine cell multiplication [159].However,whether this is due togrowth-stimulatory factors, and restrict secretion of growth synthesis and/or release of a growth factor is not known. To findinhibitors by surviving renal cells. These growth factors thereby out, the excess potassium would have to be removed from thecould mediate proliferation of the remaining tubular cells, the conditioned medium, perhaps by dialysis, prior to testing forformation of renal cysts, and possibly malignant transforma- the presence of a growth-promoting factor. We have not yettion. performed that experiment. Similar alterations in the serum potassium or sodium concen- Da. PAUL KURTIN (Director, Dialysis Unit, New Englandtration exist in nonrenal diseases. Why isn't there similar Medical Center):The age at the time of renal injury or unine-evidence of renal hyperplasia? It could be that in these disease phrectomy in humans affects the degree of resultant hypertro-states, tonic inhibition of renal growth is mediated by TGF-/3 242 Nephrology Forum: Regeneration after acute tubular necrosis produced by renal cells. The loss of functional tubular cell masscited in the text were performed in collaboration with Drs. Sreedharan during renal failure could result in reduced production ofKartha, Margaret Walsh-Reitz, Naga Aithal, and Vivek Rangnekar. Research support for work in this laboratory was provided by USPHS- growth inhibitors so that the action of growth-stimulatoryNIH ROl grants DK39689, DKl84l3, DK37227, and T32 DK07510. molecules could go unopposed. DR. MADIAS: In your judgment, are the stimulation of the Reprint requests to Dr. F. G. Toback, The University of Chicago, sodium/hydrogen antiporter in response to mitogenic factorsDepartment of Medicine, Box 453,5841 S. Maryland Avenue, Chicago, and the resultant increase in cellular sodium content andIllinois 60637, USA cytosolic alkalinization critical steps for mitogenesis? DR. TOBACK: Activation of the sodium/hydrogen antiporter ISHIBASHI K, SASAKI S, SAKAMOTO H, NAKAMURA T, MARUMO F: usually occurs during the onset of cell growth, but neither itsExpression of hepatocyte growth factor and its receptor mRNA in kidney after renal ischemia or unanephrectomy. (abstract) J Am Soc activation nor cytosolic alkalinization are sufficient to initiateNephrol 2:648, 1991. and maintain proliferation [161]. Renal epithelial cell growth in culture has been shown to occur in the absence of antiporter References activity [1621. DR. HARRINGTON: Have we learned enough about these 1. OLIVER J, MACDOWELL M, TRAcY A: The pathogenesis of acute growth factors to be able to add them to the system used for the renal failure associated with traumatic and toxic injury. Renal ischemia, nephrotoxic damage and the ischemuric episode. J Clin preservation of kidneys prior to transplantation? Invest 30: 1305—1440, 1951 DR. TOBACK: I am not aware of any clinical studies address- 2, CUPPAGE FE, TATE A: Repair of the nephron following injury with ing this issue, but I hope that today's discussion will induce mercuric chloride. Am J Pathol 51:405—429, 1967 investigation along those lines. 3. SIEGEL FL, BULGER RE: Scanning and transmission electron DR. COHEN: You have underscored the remarkable ability of microscopy of mercuric chloride-induced acute tubular necrosis in rat kidney. Virchows Arch (Cell Pathol) 18:243—262, 1975 growth factors, both known and unknown, to mediate regener- 4. HAAGSMARH,POUNDAW:Mercuric chloride-induced tubulone- ation and full recovery of renal epithelial structure and function crosis in the rat kidney: The recovery phase. Br J Exp Pathol following both toxic and ischemic injury. Is there any reason to 61:229, 1980 take hope from what we know about growth factors that 5. SOLEZ K, WHELT0N A: Acute Renal Failure: Correlations be- glomeruli might be induced to regenerate in the aftermath of the tween Morphology and Function. Marcel Dekker, New York, 1984 various glomerulonephritides? Might one reasonably speculate 6. WEINBERG JM: The cell biology of ischemic renal injury. Kidney that any acute or chronic renal disease other than acute renal mt 39:476—500, 1991 failure ultimately might benefit from therapeutic manipulation 7. SOLEZ K, MOREL-MAROGER L, SRAER J-D: The morphology of at the level of growth factors? "acute tubular necrosis" in man: Analysis of 57 renal biopsies and a comparison with the glycerol model. Medicine (Baltimore) DR. TOBACK: Glomeruli are particularly complex because 58:362—376, 1979 each of the three cell types, mesangial, epithelial, and endothe- 8. DARMADY EM, STRANCK F: Microdissection of the nephron in lial, is likely to produce and respond to a different array of disease. Br Med Bull 13:21—26, 1957 growth factors. Thus far, the greatest progress has been made in 9. MOLITORIS BA, NELSON WJ: Alterations in the establishment and the study of mesangial cells [163], but I am not aware of reports maintenance of epithelial cell polarity as a basis for disease processes. J Clin Invest 85:3—9, 1990 of attempts to induce regeneration of glomerular cells following 10. MENDLEY SR. TOBACK FG: Autocrine and paracrine regulation of inflammatory cell injury. When more information is amassed kidney epithelial cell growth. Annu Rev Physiol 51:33—50, 1989 about each glomerular cell type, invading macrophages and 11. PRESCOTT LF: The normal urinary excretion rates of renal tubular leukocytes, and the interactions between them, therapeutic cells, leukocytes, and red blood cells. Clin Sci 31:425—435, 1966 12. TOBACK FG: Control of renal regeneration after acute tubular strategies may be developed to halt cell proliferation during the necrosis. Nephrology. Proc IXth mt Cong Nephrol 1:748—763, onset of glomerulonephritis, which is probably mediated in part 1985 by growth factors. Although the number of glomeruli in the 13. TOBACK FG, HAVENER FL, DODD RC, SPARGO BH: Phospholipid adult kidney apparently is fixed, it is tempting to suggest that metabolism during renal regeneration after acute tubular necrosis. administration of growth factor antagonists [88] or other agents Am J Physiol 232:E216—E222, 1977 14. CUPPAGE FE, CUNNINGHAM N, TATE AL: Nucleic acid synthesis could initiate repair and regeneration of those glomerular cells in the regenerating nephron following injury with mercuric chlo- that survive the acute inflammatory insult. ride. Lab Invest 21:449—457, 1969 Perhaps it is also worthwhile to consider the potential thera- 15. TOBACK FG, DODD RC, MAIER ER, HAvENER U: Amino acid peutic role of growth factors in the treatment of renal carci- administration enhances renal protein metabolism after acute tubular necrosis. Nephron 33:238—243, 1983 noma. If a growth factor that binds only to renal tubular 16. TOBACK FG: Amino acid treatment of acute renal failure, in epithelial cells is identified, it could be conjugated with a Contemporary Issues in Nephrology (vol 6), edited by BRENNER cellular toxin or radioactive isotope to produce powerful, new BM, STEIN JH, New York, Churchill Livingstone, 1980, pp chemotherapeutic agents to treat these neoplasms. 202—228 17. TOBACK FG; Amino acid enhancement of renal regeneration after Note Added in Proof acute tubular necrosis. Kidney mt 12:193—198, 1977 18. YARDEN Y, ULLRICH A: Molecular analysis of signal transduction The recently described hepatocyte growth factor, which also is by growth factors. Biochemistry 27:3113—3119, 1988 synthesized in the kidney, exhibits increased gene transcription after 19. BERRIDGE MI: Inositol trisphosphate and diacylglycerol: two ischemia and might play a role in renal regeneration. interacting second messengers. Annu Rev Biochem 56:159—193, Acknowledgments 1987 20. MCCREIGHT CE, SULKIN NM: Cellular proliferation in the kidney The author would like to thank Dr. John C. Lieske for valuable of young and senile rats following unilateral nephrectomy. J discussions and his efforts in the preparation of the manuscript. Studies Gerontol 14:440-443, 1959 Nephrology Forum: Regeneration after acute tubular necrosis 243

21. SPORN MB, TODARO GJ: Autocrine secretion and malignant regulation of fibroblast multiplication. Quantitative differences in transformation of cells. N Eng! J Med 303:878—880, 1980 nutrient and serum factor requirements for multiplication of nor- 22, MASSAGUE J: Transforming growth factor-a. J Biol Chem 265: mal and SV4O virus-transformed human lung cells. J Bio! Chem 21393—21396, 1990 256:2973—2981, 1981 23. RALL LB, SCOTT J, BELL GI, CRAWFORD RJ, PENSHOW JD, 46. WEBER MJ, EVANS PK, JOHNSON MA, MCNAIR TF, NAKAMURA NIALL HD, COGHLAN JP: Mouse prepro-epidermal growth factor KD, SALTER DW: Transport of potassium, amino acids, and synthesis by the kidney and other tissues. Nature 313:228—231, glucose in cells transformed by Rous sarcoma virus. Fed Proc 1985 43:107—112, 1984 24. TOBACK FG, WALSH-REITZ MM, MENDLEY SR, KARTHA 5: 47. Cwssmssici DA, MESSANA JM, HUMES HD: Adenosine triphos- Kidney epithelial cells release growth factors in response to phate stimulates thymidine incorporation but does not promote extracellular signals. Pediatr Nephrol 4:363—371, 1990 cell growth in primary cultures of rabbit proximal tubule cells 25. MENDLEYSR,TOBACK FG: Cell proliferation in the end-stage (abstract). Kidney mt 33:389A, 1987 kidney. Am J Kidney Dis 16:80—84, 1990 48. KARTHA 5, SUKHATME VP, TOBACK FG: ADP activates protoon- 26. Hops's HE, BERNHEIM BC, NISALAK A, Trio JH, SMADEL JE cogene expression in renal epithelial cells. Am J Physiol 252: Biologic characteristics of a continuous kidney cell line derived Fll75—Fl 179, 1987 from the African green monkey. J Immunol 91:416—424, 1963 49. SUKHATME VP, KARTHA 5, TOBACK FG, TAUB R, HOOVER RV, 27. ROLLASON HD: Compensatory hypertrophy of the kidney of the T5AI-M0RRI5 C-H: A novel early growth response gene rapidly young rat with special emphasis on the role of cellular hyperplasia. induced by fibroblast, epithelial cell and lymphocyte mitogens. AnatRec 104:263—285, 1949 Oncogene Res 1:343—355, 1987 28. TOBACK FG: Induction of growth kidney cells in culture by Nat 50. KARTHA 5, BRADHAM DM, GROTENDORST GR, TOBACK FG: Proc Nat! Acad Sci USA 77:6654—6656, 1980 Kidney epithelial cells express the c-sis proto-oncogene and 29. LEIGHTON J, BRADA Z, ESTES LW, JUSTH G: Secretory activity secrete PDGF-like protein: evidence for a paracrine mechanism. and oncogenicity of a cell line (MDCK) derived from canine Am J Physiol 255:F800—F806, 1988 kidney. Science 163:472—473, 1969 51. NAKAI A, KARTHA 5, SAKURAI A, TOBACK FG, DEGROOT L: A 30. WALSH-REITZ MM, TOBACK FG: Vasopressin stimulates growth human early response gene homologous to murine nur77 and rat of renal epithelial cells in culture. Am J Physiol 245:C365—C370, NGFI-B, and related to the nuclear receptor superfamily. Mo! 1983 Endocrino! 4: 1438-1443, 1990 31. WALSH-REITZ MM, TOBACK FG: Kidney epithelial cell growth is 52. SUKHATME VP, CAO X, CHANG LC, TSAI-MORRIS C-H, STAMEN- stimulated by lowering extracellular potassium concentration. Am KOVICH D, FERREIRA P, COHEN D, EDWARDS S, SHOWS T, J Physiol 247:Cl4-C19, 1984 CURRAN T, LEBEAU M, ADAMSON E: A zinc finger encoding gene 32. WALSH-REITZ MM, GLUCK SL, WAACK S, TOBACK FG: Lowering co-regulating with c-fos during growth, differentiation and after extracellular Na concentration releases autocrine growth factors cellular depolarization. Ce!! 53:37—43, 1988 from renal epithelial cells. Proc Nat! Acad Sci USA 83:4764—4768, 53. DEUEL TF, PIERCE GF, HSIU-JENG Y, SHAWVER LK, MILNER 1986 PG, KIMURA A: Platelet-derived growth factors/sis in normal and 33. KARTHA S, TOBACK FG: Purine nucleotides stimulate DNA syn- neoplastic cell growth. J Ce!! Physiol Supp! 5:95—99, 1987 thesis in kidney epithelial cells in culture. Am J Physiol 249:F967— 54. WESTERMARK B: The molecular and cellular biology of platelet- F972, 1985 derived growth factor. Acta Endocrino! (Copenh) 123:131—142, 34. RANGNEKAR VV, WAHEED S. DAVIES TJ, TOBACK FG, RANG- 1990 NEKAR VM: Antimitogenic and mitogenic actions of interleukin-l 55. Josir'ssors A, HELDIN C-H, WASTESON A, WESTERMARK B, in diverse cell types are associated with induction of gro gene DEUEL TF: The c-sis gene encodes a precursor of the B chain of expression: J Biol Chem 266:2415—2422, 1991 platelet-derived growth factor. EMBO J 3:921—928, 1984 35. SCHRADER GA, PRICKETT CO, SALMON WD: Symptomatology 56. JOSEPHS SF, GRI0 C, RATNER L, WONG-STAAL F: Human pro- and pathology of potassium and magnesium deficiencies in the rat. tooncogene nucleotide sequences corresponding to the transform- JNutr 14:85—110, 1937 ing region of simian sarcoma virus. Science 223:487—490, 1984 36. OLIVER J, MACDOWELL M, WELT LG, HOLLIDAY MA, H0L- 57. SEIFERT RA, HART CE, PHILLIPS PE, FORSTROM JW, Ross R, LANDER W J, WINTERS RW, WILLIAMS TF, SEGAR WE: The MURRAY MJ, BOWEN-POPE DF: Two different subunits associate renal lesions of electrolyte imbalance. I. The structural alterations to create isoform-specific PDGF receptors. J Bio! Chem 265:877 1— in potassium-depleted rats. J Exp Med 106:563—574, 1957 8778, 1989 37. SPARGO BH: Kidney changes in hypokalemic alkalosis in the rat. 58. MAT5uI T, HEIDARAN M, MIKI T, POPESCU N, LA ROCHELLE W, J Lab Clin Med 43:802—814, 1954 KRAUS M, PIERCE J, AARONSON 5: Isolation of a novel receptor 38. MORDAN Li, TOBACK FG: Growth of kidney epithelial cells in cDNA establishes the existence of two PDGF receptor genes. culture: evidence for autocrine control. Am J Physiol 246:C351— Science 243:800—804, 1989 C354, 1984 59. LE0F EB, PROPER JA, GOUSTIN AS, SHIPLEY GD, DICORLETO 39. HANKS SK, ARMOUR R, BALDWIN JH, MALDONADO F, SPIESS J, PE, MOSES HL: Induction of c-sis mRNA and activity similar to HOLLEY RW: Amino acid sequence of the BSC-1 cell growth platelet derived growth factor by transforming growth factor beta: inhibitor (polyergin) deduced from the nucleotide sequence of the a proposed model for indirect mitogenesis involving autocrine eDNA. Proc Nat! Acad Sci USA 85:79—82, 1988 activity. Proc Nat! Acad Sci USA 83:2453—2457, 1986 40. CONE CD JR, TONGIER M JR: Contact inhibition of division: 60. MAKELA TP, ALITALO R, PAULSSON Y, WESTERMARK B, HELDIN involvement of the electrical transmembrane potential. J Cell C-H, ALITALO K: Regulation of platelet-derived growth factor Physiol 82:373—386, 1973 gene expression by transforming growth factor beta and phorbol 41. SMITH JB, ROZENGURT E: Serum stimulates the Na/K pump in ester in human leukemia cell lines. Mo! Ce!!Bio! 7:3653—3662, 1987 quiescent fibroblasts by increasing Na entry. Proc Nat! Acad Sci 61. FRAIZER GE, BOWEN-POPE DF, VOGEL AM: Production of plate- USA 75:5560—5564, 1978 let-derived growth factor by cultured Wilms' tumor cells and fetal 42. HOLLEY RW, ARMOUR R, BALDWIN JH: Density-dependent kidney cells. J Ce!! Physio! 133:169—174, 1987 regulation of growth of BSC-1 cells in cell culture: control of 62. DINARELLO CA: Interleukin-1. Rev Infect Dis 6:51—95, 1988 growth by low molecular weight nutrients. Proc Nat! Acad Sci 63. LOVETT DH, SZAMEL M, RYAN JL, STERZEL RB, GEM5A D, USA 75:339—341, 1978 RESCH K: Interleukin 1 and the glomerular mesangium. I. Purifi- 43. KOCH KS, LEFFERT HL: Increased sodium ion influx is necessary cation and characterization of a mesangial cell-derived auto- to initiate rat hepatocyte proliferation. Ce!! 18:153—163, 1979 growth factor. J Immuno! 136:3700—3705, 1986 44. ROZENGURT E, HEPPEL LA: Serum rapidly stimulates ouabain- 64. ANISOWICZ A, BARD WELL L, SAGER R: Constitutive overexpres- sensitive 86Rb influx in quiescent 3T3 cells. Proc Nat! Acad Sci sion of a growth-regulated gene in transformed Chinese hamster USA 72:4492—4495, 1975 and human cells. Proc Nat! Acad Sci USA 84:7188—7192, 1987 45. MCKEEHAN WL, MCKEEHAN KA, CALKINS D: Extracellular 65. RICHMOND A, BALENTIEN E, THOMAS HG, FLAGOS G, BARTON 244 Nephrology Forum: Regeneration after acute tubular necrosis

DE, SPIESS J, BAIW0NI R, FRANCKE U, DERYNCK R: Molecular 85. FINE LG, HOLLEY RW, NA5RI H, BADIE-DEZFOOLY B: BSC-l characterizationandchromosomalmapping of melanoma growth growth inhibitor transforms a mitogenic stimulus into a hyper- stimulatoryactivity, a growth factor structurally related to beta- trophic stimulus for renal proximal tubular cells: Relationship to thromboglobulin. EMBOJ 7:2025—2033,1988 Na/H antiport activity. ProcNatlAcad Sci USA 82:6163—6166, 66. COCHRAN BH, REFFEL AC, STILES CD: Molecular cloning of gene 1985 sequences regulated by platelet-derived growth factor. Cell 33: 86. IGNOTZ R, MA55AGUE J: Transforming growth factor-beta stimu- 939—947,1983 lates the expression of fibronectin and collagen and their incorpo- 67.FISHER DA, SALIDO EC, BARAJAS L: Epidermal growth factor and ration into the extracellular matrix. J Rio! Chem 261:4337—4345, the kidney. Annu Rev Physiol 51:67—80, 1989 1986 68. BREYER MD, REDHA R, BRaYER JA: Segmental distribution of 87. PIERCE GE, MU5TOE TA, LINGELBACH I, MA5AKOWSKI yR. epidermal growth factor binding sites in rabbit nephron. AmJ GRIFFIN GL, SENIOR RM, DEUEL TF: Platelet-derived growth Physiol259:F553—F558, 1990 factor and transforming growth factor-p enhance tissue repalr 69. SPURR NK, SoLoMON B, JANSSON M,SHEERD,GOODFELLOW activities by unique mechanisms. J Cell Biol 109:429—440, 1989 PN, BODMER WE, VENNSTROM B: Chromosomal localization of 88. BORDER WA, OKUDA 5, LANGUINO LR,SPORNMB,RUO5LAHTI the human homologues to the oncogenes erbA and B. EMBOJ E: An antiserum agalnst transforming growth factor j31suppresses 3:159—163,1984 experimental glomerulonephritis. Nature 346:371—374,1990 70.HOLLEY RW, ARMOUR R, BALDWIN JH, BROWN KD, YEH Y-C: 89.BAXTER RC: The somatomedins: insulin-like growth factors. Adv Density-dependent regulation of growth of BSC-l cells in cell CliiiChem 25:49—1 15, 1986 culture: control of growth by serum factors. Proc Nat! Acad Sd 90. RINDERKNECHT E, HUMBEL RE: The amino acid sequence of USA74:5046—5050,1977 human insulin-like growth factor I and its structural homology 71. SAFERSTEIN R,PRICEPM, SAOOI SJ, HARRIS, RC: Changes in with proinsulin. JBiol Chem 253:2769—2776,1978 gene expression after temporary renal ischemia. Kidney mt 37: 91. RINDERKNECHT E, HUMBEL RE: Primary structure of human 1515—1521,1990 insulin-like growth factor II. FEBSLett 89:283—286,1978 72.SAFERSTEIN R, ZELENT AZ, PRICE PM: Reduced renal prepro- 92.RECHLERMM, NIS5LEY SP: Insulin-like growth factor (IGF)/ epidermal growth factor mRNAanddecreased EGE excretion in somatomedin receptor subtypes: structure, function, and relation- ARE.Kidney mt36:810—815, 1989 ships to insulin receptors and IGE carrier proteins. Horm Res 73.HUMES HD,CIELINSKIDA, COIMBRA TM,MESSANAJM,GAL- 24: 152—159, 1986 vAo C: Epidermal growth factor enhances renal tubular cell 93. MACDONALD RG, PFEFFER SR, COUSSENS L, TEPPER MA, regeneration and repair and accelerates the recovery of renal BROCKLEBANK CM, MOLE JE, ANDERSON JK, CHEN B, CZECH function in postischemic acute renal failure. J Clin Invest 84:1757— MP, ULLRICH A: A single receptor binds both insulin-like growth 1761, 1989 factor II and mannose-6-phosphate. Science 239:1134-1137, 1988 74. NORMAN J, TSAU Y-K, BACAY A, FINE LG: Epidermal growth 94. HAMMERMAN MR, ROGERS 5: Distribution of IGF receptors in the factor accelerates functional recovery from ischaemic acute tubu- plasma membrane of proximal tubular cells. Am J Physiol 253: lar necrosis in the rat: role of the epidermal growth factor F841—F847, 1987 receptor. Clin Sci78:445—450,1990 95.D'ERCOLEAJ, STILES AD, UNDERWOOD LE: Tissue concentra- 75. COIMBRA TM, CIESLINSICI DA, HUMES HD: Epidermal growth tionS of somatomedin C: Further evidence for multiple Sites of factor accelerates renal repalr in mercuric chloride nephrotoxicity. synthesis and paracrine or autocrine mechanisms of action. Proc Am J Physiol 259:F438—F443, 1990 Nat! Acad Sci USA 81:935—939, 1984 76. WILCOX JH, DERYNCK R: Developmental expression of trans- 96. MURPHY U, BELL GI, DUCKWORTH ML, FRIESEN HG: Identifi- forming growth factors alpha and beta in mouse fetus. Mol Cell cation, characterization, and regulation of a rat complementary Biol 8:3415—3422, 1988 deoxyribonucleic acid which encodes insulin-like growth factor-I. 77. GOMELLA LG, SARGENT ER, WADE T, ANGLARD P, LINEHAN Endocrinology 121:684—691, 1987 WM, KASID A: Expression of transforming growth factor in 97. BELL GI, GERHARD DS, FONG NM, SANCHEZ-PESCADOR R, RALL normal human adult kidney and enhanced expression of trans- LB: Isolation of the human insulin-like growth factor genes: forming growth factors a and /31 in renal cell carcinoma. Cancer Insulin-like growth factor II and insulin genes are contiguous. Res49:6972—6975,1989 ProcNatl Acad Sci USA82:6450—6454, 1985 78. HOLLEY RW, ARMOUR R, BALDWIN JH: Density-dependent 98. BORTZ ID, ROTWEIN P, DEVOL D, BECHTEL PJ, HANSEN BA, regulation of growth of BSC-l cells in cell culture: growth inhib- HAMMERMAN MR: Focal expression of insulin-like growth factor itors formed by the cells. Proc Nati Acad Sci USA 75:1864—1866, I in rat kidney collecting duct. JCell Rio! 107:811-819,1988 1978 99. KOPPLE JD, HIRSCHBERG R: Physiological effects of growth 79. HOLLEY RW, BOHLEN P,FAVAR, BALDWIN JH, KLEEMAN G, hormone and insulin-like growth factor-I on the kidney. Miner ARMOUR R: Purification of kidney epithelial cell growth inhibitors. Electrolyte Metab16:82—88,1990 ProcNat! Acad Sci USA 77:5989—5992, 1980 100. Scorr J, COWELL J, ROBERTSON ME, PRIESTLEY LM, WADEY R, 80.WALSH-REITZ MM, TOBACK P0, HOLLEY RW: Cell growth and HOPKINS B, PRITCHARD I, BELL 01, RALL LB, GRAHAM CE, net Na flux are inhibited by a protein produced by kidney KNOTT TJ: Insulin-like growth factor-Il gene expression in Wilms' epithelial cells in culture. ProcNatI Acad Sci USA 75:1864—1866, tumour and embryonic tissues. Nature 3 17:260-262, 1985 1984 101.ANDERSSONG,JENNISCHEB: IGE-I immunoreactivity is ex- 81. TUCKER RE, SHIPLEY GD, MOSES HL, HOLLEY RW: Growth pressed by regenerating renal tubular cells after ischemic injury in inhibitor from BSC-l cells closely related to platelet type /3 the rat. Acta Physiol Scand132:453—457,1988 transforming growth factor. Science 226:705—707, 1984 102. SEGAL R,FINELG:Polypeptidegrowth factors and the kidney. 82. DERYNCK R, JARRETT JA, CHEN EY, EATON DH,BELLJR, Kidney Int 36 (suppl 27):52—SlO,1989 AssolANRK, ROBERTS AB, Seoart MB, GOEDDEL DV: Human 103.ABBOUD HE, POPTIC E, DICORLETO P:Productionof platelet- transforming growth factor-p complementary DNA sequence and derived growth factorlike protein by rat mesangial cells in culture. expression in normal and transformed cells. Nature 316:701—705, JC!in Invest 80:675—683,1987 1985 104. ABRAHAM JA, MERGIA A, WHANO JL, TUMOLO A, FRIEDMAN J, 83. ELLINGSWORTII LR, BRENNAN JE, FOK K, ROSEN DM, BENTZ HJERRILDKA, GOSPODAROWICZ D, EIDDES JC: Nucleotide se- H, PIEz KA, SEYEDIN SM: Antibodies to the N-terminal portion quence of a bovine clone encoding the angiogenic protein, basic of cartilage-inducing factor-A and transforming growth factor /3. J libroblast growth factor. Science 233:545—548, 1986 BiolChem 261:12362—12367,1986 105. BAIRD A, LING N: Eibroblast growth factors are present in the 84. THOMPSON NL, FLANDERS KC, SMITH JM, ELLING5WORTH LR, extracellular matrix produced by endothelial cells in vitro: Impli- ROBERTS AB, SPORN MB: Expression of transforming growth cations for a role of heparinase-like enzymes in the neovascular factor-/31 in specific cells and tissues of adult and neonatal mice. J response. Biochem Biophys ResCommun 142:428—435,1987 CellBiol 108:661—669, 1989 106. JAYE M, HOWK R, BUROESS W, RICCA GA, CHIU I-M, RAVERA Nephrology Forum: Regeneration after acute tubular necrosis 245

MW, O'BRIENSJ,Morn WS, MAClAG T, DROHAN WN: Human 129. HYNES R: Integrins: a family of cell surface receptors. Cell endothelial cell growth factor: cloning, nucleotide sequence and 48:549—554, 1987 chromosomal localization. Science 233:541—545, 1986 130. EKBLOM P, ALITALO A, VAHERI R, TIMPL R, SAXEN L: Induction 107. BAIRD A, ESCH F, BOHLEN P, LING N, GOSPODAROWICZ D: of a basement membrane glycoprotein in embryonic kidney: Isolation and partial characterization of an endothelial growth possible role of laminin in morphogenesis. Proc NatI Acad Sd factor from bovine kidney: homology with basic fibroblast growth USA 77:485—489, 1980 131. KLEIN G, LAUGEGGER R, TIMPLE R, EKBLOM P: Role of laminin factor. Reg Pep: 12:201—213, 1985 A chain in the development of epithelial cell polarity. Cell 55:331— 108. GAUTSCHI-SOVA P, JIANG Z-P, FRATER-SCHRODER M, BOHLEN P: 341, 1988 Acidic fibroblast growth factor is present in nonneural tissue: 132. EKBLOM P, MIETTINEU A, VIRTANEN I, WAHLSTROM T, Isolation and chemical characterization from bovine kidney. Bio- DAWNAY A, SAXEN L: In vitro segregation of the metanephric chemistry 26:5844—5847, 1987 nephron. Dev Blot 84:88—95, 1981 109. RI5Au W, EKBLOM P: Production of a heparin-binding angiogen- 133. BERNFIELD M, SANDERSON RD: Syndecan, a developmentally esis factor by the embryonic kidney. J Cell Blot 103:1101—1107, regulated cell surface proteoglycan that binds extracellular matrix 1986 and growth factors. Phil Trans R Soc Lond[BJ 327:171—186, 1990 110. ANDERSON S, RENNKE HG, BRENNER BM: Therapeutic advan- 134. VEGAS-SALAS DE, SALAS PJI, GUNDERSON D, RODRIGUEZ- tage of converting enzyme inhibitors in arresting progressive renal BOULAN E: Formation of the apical role of epithelial (Madin- disease associated with systemic hypertension in the rat. J Clin Darby Canine Kidney) cells: Polarity of an apical protein is Invest 77:1993—2000, 1986 independent of tight junctions while segregation of a basolateral 111. POWELL IS, MUELLER RKM, ROUGE M, KUHN H, HEFTI F, marker requires cell-cell interactions. J Cell Blot 104:905—916, BAUMGARTNER HR: The proliferative response to vascular injury 1987 is suppressed by angiotensin-converting enzyme inhibition. J 135. OJAKIAN GK, SCHWIMMER R: The polarized distribution of an Cardiovasc Pharm 16 (suppl 4):S24—S48, 1990 apical cell surface glycoprotein is maintained by interactions with 112. XIE M-H, Liu F-Y, WONG PC, TIMMERMANS PBMWM, COGAN the cytoskeleton of Madin-Darby Canine kidney cells. J Cell Biol MG: Proximal nephron and renal effects of DuP 753, a nonpeptide 107:2377—2387, 1988 angiotensin II receptor antagonist. Kidney mt 38:473—479, 1990 136. M0LIT0RIS BA, FALK SA, DAHL RH: Ischemia-induced loss of 113. SACCOMANI G, MITCHELL KD, NAVAR LG: Angiotensin II stim- epithelial polarity. J Clin Invest 84:1334-1339, 1989 ulation of Na-H exchange in proximal tubule cells. Am J 137. KELLERMAN PS, CLARK RAF, H0ILIEN CA, LINAS SL, MOLITORIS Physiol 258:F1 188—Fl 195, 1990 114. WOLF G, NIELSON EG: Angiotensin II induces cellular hypertro- BA: Role of microfilaments in maintenance of proximal tubule phy in cultured murine proximal tubular cells. Am J Physiol structural and functional integrity. Am J Physiol 259:F279—F285, 259:F768—F777, 1990 1990 115. ROBINETTE JB, CONGER JD: Angiotensin and thromboxane in the 138. LOTSPIECH WD: Metabolic aspects of acid-base change. Science enhanced renal adrenergic nerve sensitivity of acute renal failure. 155:1066—1075, 1967 J Clin Invest 86:1532—1539, 1990 139. BLUMENTHAL SS, LEWAND DL, BUDAY MA, MANDEL NS, 116. MORAN SM, MYERS BD: Pathophysiology of protracted acute MANDEL GS, KLEINMAN JG: Effect of pH on growth of mouse renal failure in man. J Clin Invest 76:1440—1448, 1985 renal cortical tubule cells in primary culture. Am J Physiol 117. PEDERSON T, GELFANT S: G2-population cells in mouse kidney 27:C419—C426, 1989 and duodenum and their behavior during the cell division cycle. 140. JANS DA, PETERS R, JANS P. FAHRENHOLZ F: Ammonium chloride Exp Cell Res 59:32—36, 1970 affects receptor number and lateral mobility of the vasopressin 118. DANIEL TO, GIBBS VC, MILFAY DF, GAROVOY MR, WILLIAMS V2-type receptor in the plasma membrane of LLC-PK1 renal epithe- LT: Thrombin stimulates c-sis gene expression in microvascular hal cells: role of the cytoskeleton. Exp Cell Res 191:121—128, 1990 endothelial cells. J Biol Chem 261:9570—9582, 1986 141. GREENHALGH DG, SPRUGEL KH, MURRAY Mi, Ross R: PDGF 119. RAPPOLEE DA, MARK D, BANDA MJ, WEltn Z: Wound macro- and FGF stimulate wound healing in the genetically diabetic phages express TGF-alpha and other growth factors in vivo: mouse. Am J Pathol 136:1235—1246, 1990 Analysis by mRNA phenotyping. Science 241:708—712, 1988 142. PIERCE GF, MUSTOE TA, LINGELBACH J, MASAKOWSKI VR, 120. NATIONAL KIDNEY AND UROLOGIC DISEASES ADVISORY BOARD GRAMATES P, DEUEL TF: Transforming growth factor beta re- 1990 LONG-RANGE PLAN: Window on the 21st Century. NIH verses the glucocorticoid-induced wound-healing deficit in rats: publication number 90-583, 1990, pp 47—48 121. BUTKUS DE: Persistent high mortality in acute renal failure. Are possible regulation in macrophages by platelet-derived growth we asking the right questions? Arch Intern Med 143:209—212, 1983 factor. Proc Nail Acad Sd USA 86:2229—2233, 1989 122. STERN PH, KRIEGER NS, NI5sENS0N RA, WILLIAMS RD, WIN- 143. OGAWA 5, LEAVY J, CLAUSS M, DOGA S, SHREENIWAS R, KLER MS, DERYNCK R, STREWLER GJ: Human transforming JOSEPH-SILVERSTEIN J, FURIE M, STERN D: Modulation of endo- growth factor-alpha stimulates bone resorption in vitro. J Clin thelial cell function in hypoxia: alterations in cell growth and the Invest 76:2016—2019, 1985 response to monocyte-derived mitogenic factors. J Cell Biochem 123. TASHIJAN AH, VOELKEL EF, LLOYD W, DERYNCK R, WINKLER (suppl I5F):199A, 1991 ME, LEVINE L: Actions of growth factors on plasma calcium. J 144. EL NAHAS AM, LE CARPENTIER JE, BASSETT AH: Compensatory Clin Invest 78:1405—1409, 1986 renal growth: role of growth hormone and insulin-like growth 124. INSOGNA KL, STEWART AF, MoRRIs CA, HOUGH LM, MILSTONE factor-I. Nephrol Dial Transplant 5:123—129, 1990 LM, CENTRELLA M: Native and a synthetic analogue of the 145. POWELL IS, ROUGE M, MULLER RKM, BAUMGARTNER HR: The malignancy-associated parathyroid hormone-like protein have in proliferative response to vascular injury: the role of angiotensin II vitro transforming growth factor-like properties. J Gun Invest induction of PDGF and thrombospondin, in Growth Factors in 83:1057—1060, 1989 Health and Disease. Basic and Clinical Aspects, edited by WES- 125. HONDA N, HISHIDA A, IKUMA K, YONEMURA K: Acquired TERMARK B, BETSCHOLTZ C, HOKFELT B, Amsterdam, Elsevier, resistance to acute renal failure. Kidney mt 3 1:1233—1238, 1987 1990, pp 117—129 126. MACNIDER WDEB: The functional and pathological response of 146. MANZANO F, ESBRIT P, GARCIA-OCAfIA A, GARC1A-CAIERO R, the kidney in dogs subjected to a second subcutaneous injection of JIMENEz-CLAVER0 MA: Partial purification and characterization uranium nitrate. J Exp Med 49:411-431, 1929 127. TOBACK FG, KARTHA S, WALSH-REITZ MM: Determinants of of a renal growth factor from the plasma of uninephrectomised autocrine and paracrine growth factor release by kidney epithelial rats. Nephrol Dial Transplant 4:334—338, 1989 cells. Nephrology. Proc XIth mt Cong Nephrol 2:1322—1332, 1991 147. WILSON P: Aberrant epithelial cell growth in autosomal dominant 128. ZHANG G, ICHIMURA T, WALLIN A, KAN M, STEVENS JL: polycystic kidney disease. Am J Kidney Dis 17:634—637, 1991 Regulation of rat proximal tubule epithelial cell growth by fibro- 148. MosKowlTz DW, BONAR SL, MARCUS MD, CLAYMAN RV, blast growth factors, insulin-like growth factor-I and transforming AVNER ED: Epidermal growth factor content of human and mouse growth factor-p and analysis of fibroblast growth factors in rat (cpk) cysts. Am K Kidney Dis 14:435—436, 1989 kidney. J Cell Physiol, in press 149. HoRiKosi S, KUBOTA S, MARTIN GR, YAMADA Y, KLOTMAN PE: 246 NephrologyForum: Regeneration after acute tubular necrosis

Epidermal growth factor (EGF) expression in the congenital of vasoactive substances on early toxic acute renal failure in the polycystic mouse kidney. Kidneymt 39:57—62, 1991 dog. Nephrol Dial Transplant 2:332—340, 1987 150. GAi-FONE VH II, ANDREWS GK, Niu F-W, CHADWICK Li, 157. ISHIKAWA I,HOLLENBERO NK:Pharmacologic interruption of the KLEIN RM, CALVET JP: Defective epidermal growth factor gene renin-angiotensin system inmyohemoglobinuricacute renal fail- expression in mice with polycystic kidney disease. Dev Biol ure. Kidneymt 10:S183—S190, 1976 138:225—230, 1990 158.BRIGHTWELLJR,RIDDLE SL, EIFERMAN RA, VALENZUELA P, 151. SIEGEL N, GAUDIO KM, KATZ LA, REILLY HF, ARDITO TA, BARR PJ, MERRYWEATHER JP, SCHULTZ GS: Biosynthetic human HENDLER FG,KASHGARIANM: Beneficial effect of thyroxin on EGF accelerates healing of Neodecadron-treated primate corneas. recovery from toxic acute renal failure. Kidney mt 25:906—911,1984 Invest Ophthalmol Visual Sci 26:105—110, 1985 152. SIEGEL NJ, GAUDJO KM: Amino acids and adenine nucleotides in acute renal failure, in Acute Renal Failure (2nd ed), edited by 159.TOBACK FG, EKELMANKB,ORDOtEZ NG: Stimulation of DNA BRENNER BM, LAZARUS JM, New York, Churchill Livingstone, synthesisin kidney epithelialcellsin culture by potassium. Am J 1988, pp 857—873 Physiol 247:Cl4—C19, 1984 153. ESCOBEDO JA, BAR PJ, WILLIAMS LT: Role of tyrosine kinase and 160. CHAUDRY IH,SAYEED MM, BAUEAE: Effect of adenosine membrane-spanning domains in signal transduction by the platelet- triphosphate-magnesium chloride administration in shock. Surgery derived growth factor receptor. Mol Cell Biol 8:5126-5131, 1988 75:220—227, 1974 154.KUMJIANDA, WAHL MI, RHEE SG, DANIEL TO: Platelet-derived 161. GRINSTEIN S, ROTIN D, MASON MJ: Na/H exchange and growth factor (PDGF) binding promotes physical association of growth factor-induced cytosolic pH changes. Role in cellular PDGF receptor with phospholipase C. Proc Nail Acad Sd USA proliferation. Biochim Biophys Acta 988:73—97, 1989 86:8232—8236, 1989 162. MOHRMANN M,CANTIELLOHF,AUSIELLODA:Renalepithelial 155. HOLLENBERGNK,WILKESBM, SCHULMANG:The renin-angio- cell growth can occur in absence of Na -H exchanger activity. tensin system in acute renal failure, in Acute Renal Failure (2nd Am J Physiol 253:C633—C638, 1987 ed) edited by BRENNER B,LAZARUSJM, Church Livingstone, 163.VEISJH, YAMASHITA W,LIu YJ,OoiBS:The biology of New York, 1988, pp 119—141 mesangial cells in glomerulonephritis. Proc Soc Exp Biol Med 156. VANHOLDERR, LAEKMANG,HERMANA,LAMEIREN: Influence 195:160—167, 1990