The Future of Polycystic Kidney Disease Research—As Seen By the 12 Kaplan Awardees

Peter C. Harris

Autosomal dominant PKD (ADPKD) shows considerable phenotypic variability in terms of severity of renal disease and the occurrence of extrarenal manifestations (Figure 1). The gene involved, PKD1 (approximately 85% of cases) or PKD2 (approximately 15%), is strongly associated with renal disease severity, with ESRD occurring on average approximately 20 years earlier in PKD1 than PKD2 (approximately 58 years versus approximately 79 years).¹ Genetic background, along with environmental effects, also influence the phenotype, manifesting as significant intrafamilial phenotypic variability. Analysis of common single nucleotide polymorphisms (genome-wide association studies) and rare variants (whole exome analysis) is likely to shed light on the importance of specific modifier variants/genes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Inherited and acquired determinants of cystogenesis. The formation of cysts in ADPKD is viewed as being dependent on the balance between ADPKD gene dosage (and hence protein expression, particularly of PC1 at the plasma membrane) and the susceptibility of the renal tubule epithelium to its effects, the “cystogenic threshold.” The gene dosage is dependent on the nature of germline mutations in PKD1 and PKD2, other variants at the disease locus, modifier genes including PKHD1, HNF1B, and other ciliopathy genes, and somatic mutations (the classic “second hit”), and other factors affecting the expression levels of these genes.² The cystogenic threshold seems to be strongly dependent on cell autonomous contextual factors such as the developmental stage and superimposition of environmental agents such as renal injury and infection, and can also potentially be affected by modifier genes. Cell nonautonomous local effects may alter the cystogenic threshold and account for a “snowball effect” in cyst appearance. In the graph, as gene dosage (solid blue line) dips below the cystogenic threshold (dashed red line), cysts will form (green area). As such, the interplay between decreasing gene dosage and increasing cystogenic threshold determines the onset, progression, and severity of cystic disease.

There has been a recent focus on allelic effects with several lines of evidence indicating that the PKD1 genotype is significantly associated with renal disease severity, and can explain some extreme phenotypes. Viable patients homozygous or compound heterozygous for pathogenic PKD1 missense variants indicate that some alleles are incompletely penetrant (hypomorphic); hypomorphic heterozygotes develop very mild cystic disease without ESRD.^3,4 Furthermore, the in trans combination of an inactivating and hypomorphic PKD1 allele can explain rare, very early onset ADPKD (phenotypically similar to ARPKD).^3,5,6 An in trans combination of PKD1 hypomorphic alleles can also result in ARPKD-like disease with a negative family history, but without congenital hepatic fibrosis.⁴

A mouse model mimicking the best characterized PKD1 hypomorphic allele, p.R3277C, recapitulates the human phenotypes with slowly progressive disease in homozygotes, generating a good model for preclinical testing, and rapidly progressive disease in compound heterozygotes with a null allele.^7,8 Detailed analysis of the model indicates that the level of the mature glycoform of the PKD1 protein (polycystin 1; PC1) is associated with disease severity, strongly supporting a dosage model of pathogenesis. Consistent with related mechanisms in cystic diseases, a combination of a PKD1 allele and mutations in a second cystogene can also result in severe PKD.⁵

Analysis of large ADPKD populations indicates that a significant proportion of nontruncating mutations are hypomorphic with an average age at ESRD of approximately 67 years in nontruncating mutation patients compared with 55 years in those with truncating changes.¹ Molecular screening in ADPKD can, hence, be of prognostic as well as diagnostic value, and can be employed to identify patients with rapidly progressive disease, suitable for clinical trials and future treatments⁹; although, truncating PKD1 mutation mosaics can have mild disease.^10–12 However, a significant proportion of nontruncating mutations are fully inactivating and so presently the prognostic value of knowing the mutation for the individual patient is limited.

Gregory G. Germino

What makes cysts form? We know that genetics plays a key role. Although the disease is inherited as an autosomal dominant trait, numerous lines of evidence indicate that it is recessive on a cellular level, with cysts arising when the total functional activity of the two alleles of either PKD1 or PKD2 falls below an ill-defined threshold (Figure 1). This most commonly arises as a result of acquired mutations in renal epithelial cells that have a germline mutation of either locus.^17–21 The nature of the germline mutation likely plays a dominant role in determining the severity of disease.¹ Recent studies suggest that some individuals instead may have two germline, hypomorphic alleles whose combined activity falls below the threshold.^3,7 There is also evidence to suggest that the threshold may be different in different cell types and life stages.^7,22 Genetic modifiers are likely important in three steps of this process: determining the rate of somatic mutation, setting the cystogenic threshold, and in downstream effector pathways.

There are several other lines of evidence, however, that suggest other, nongenetic factors may also be important. For example, we have shown that adult mouse kidneys take months to develop cysts after inactivation of Pkd1.²³ Metanephric Pkd null kidneys develop cysts in vivo and yet do not when cultured in vitro.²⁴ We have generated multiple mouse cell lines lacking either Pkd1 or Pkd2 and find that they make tubules in three-dimensional cultures regardless of genetic status (L. Menezes, unpublished observation). We have applied network strategies and gene expression analyses to better understand the processes responsible for cyst formation.²⁵ We compared gene expression patterns of a test set of mutant (Pkd1^cond/cond Cre⁺) and control (Pkd1^cond/cond Cre^–) mice induced at around P7 and harvested at P12–P24 and identified a mutant gene signature that properly clustered an independent validation set of mouse samples by genotype. Comparing the network structures of P12 and P14 control and mutant samples, we found that they were highly similar, arguing against a developmental block in cysts. There was, however, a change in the network architecture of P12 versus P14 samples for both controls and mutants. The genes in this cluster were highly enriched for metabolic pathways, suggesting that metabolic context could be an important disease modifier. Perturbation of the metabolic state seems to alter disease severity.

Dorien J.M. Peters

Most patients with ADPKD carry a heterozygous mutation in either the PKD1 or the PKD2 gene. However, the phenotype is also influenced by additional genetic and environmental factors (Figure 1).

In ADPKD, cysts develop in a minority of nephrons, suggesting that a heterozygous mutation in PKD1 or PKD2 is not sufficient to induce cyst formation. In fact, the mutation primes the kidneys for cyst formation, but the likelihood of cyst formation strongly increases when the level of functional PKD1 or PKD2 gene product drops below a critical threshold.^3,17,26 A variety of data suggest that the rate of cyst formation depends on the context of renal tissue. For example, both in mice and in humans, cyst growth is much faster during renal development than it is during adulthood.^23,27,28 It is likely that growth promoting conditions during renal development accelerate the process.^23,27 Furthermore, the relatively slow process of cyst formation and progression in adult mice can be accelerated by acute renal injury treatments.^29–31 Renal injury is accompanied by a combination of processes including repair associated proliferation, secretion of growth factors, and inflammation. Accelerated cyst formation seems to be the result of increased susceptibility to cyst-promoting stimuli, e.g., cytokines or growth factors, due to altered integrity of the cells, rather than the result of uncontrolled cell proliferation.^30,32 In addition, the cysts themselves can have a profound impact on the surrounding tissue by inducing mechanical stress on neighboring nephrons, which locally increases the likelihood of cyst formation.³¹

In patients, recurrent urinary tract infections and toxic (waste) products may cause local injury. The effects of these may accumulate over years and contribute to impaired renal function and renal aging. Individual variation in exposure to environmental factors that may cause renal injury, as well as renal aging, may contribute to the variations in severity of renal cystic disease among patients.

During renal development, injury/repair, or normal homeostasis, the PKD proteins modulate a complex network of signaling pathways (Figure 2) needed to establish and maintain the tubular architecture. Especially during phases in which the full potential of these pathways is required, the likelihood that this signaling network gets out of balance may increase, making renal tissue more prone to cyst formation, particularly when the PKD proteins are absent.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Diagram depicting signaling pathways that have been found to be increased or decreased in PKD, and potential drugs that act on targets in these pathways and that might ameliorate PKD.

Lisa M. Guay-Woodford

ARPKD is a single gene disorder in which the renal and biliary lesions can be quite variable. This phenotypic variability is typical even among family members who share identical PKHD1 mutations, indicating modulating effects of other genetic (i.e., co-inherited modifier genes or quantitative trait loci [QTL]), epigenetic, or environmental factors (summarized previously³³).

Efforts to identify QTL in human ARPKD cohorts are confounded by several factors, most notably: (1) ARPKD is a relatively rare disorder; (2) the PKHD1 gene is transcriptionally complex³⁴; and (3) most patients are compound heterozygotes for different PKHD1 mutations.³⁵ Genetic studies in experimental models would be a reasonable alternative, but of the eight currently available mouse Pkhd1 models, none expresses a renal lesion that phenocopies the human disease. In comparison, the renal and biliary disease in the congenital polycystic kidney (cpk) model closely phenocopies human ARPKD. As an alternative experimental model, the cpk mouse has several advantages: (1) it is the most extensively characterized PKD model; (2) the Cys1^cpk mutation disrupts the protein cystin, which is normally expressed in the primary cilium, as is the PKHD1-encoded protein; and (3) in contrast to Pkhd1, Cys1 is a small, transcriptionally uncomplicated gene.³⁶

In previous studies, we generated a cohort of F₂ affected mice, characterized the renal cystic disease as a series of quantitative traits, and used statistical analyses to identify a highly significant QTL complex on chromosome 4 that exerts a dominant effect of the mus musculus castaneus (CAST) haplotype on renal disease severity.³³ This interval comprises a complex of at least three QTL (Mpkd1–3), and contains approximately 1300 genes. We have identified Kif12, a novel kinesin-encoding gene, as the major Mpkd2 candidate using a combination of gene expression profiling, computational analyses, single nucleotide polymorphism-based haplotype mapping, and congenic strain generation. In addition, we have determined that Kif12 colocalizes with cystin in the primary cilium and that its expression pattern is highly correlated with Cys1 and Pkhd1 in the GUDMAP kidney gene expression atlas database (http://www.gudmap.org/Menu_Index/Gene_Expression.html). In related studies, we have collaborated with Greg Germino’s group and identified a genetic interaction between Pkhd1 and Pkd1 that exacerbates the renal cystic phenotype in the Pkhd1^del3–4 model.³⁷ Therefore, these mouse studies have identified Kif12 and Pkd1 as putative genetic modifiers of the ARPKD phenotype.

In parallel studies we have identified a PKHD1 founder mutation (c.T1880A; p.M627K) in a South African Afrikaner cohort.³⁸ Varied disease expression was documented among the 27 affected individuals who were homozygous for this founder mutation. Therefore, this Afrikaner cohort represents a unique, foundational resource for directed testing of candidate recessive PKD QTL identified in our mouse model studies.

James P. Calvet

One of the more perplexing questions about polycystic kidneys is the nature of the abnormal cell proliferation required to form the growing cyst. There is no doubt that cyst growth involves increased cell proliferation. However, this process proceeds over months and years, and although perhaps unrelenting, it cannot be characterized as runaway cell growth. However, neither are most cancerous tumors, which probably take many years of growth before they reach a size that can be detected by screening.

Cyst growth in PKD has many parallels with cancer, including elevated proto-oncogene activation, Ras/B-Raf/mitogen-activated protein kinase, mTOR, and EGF receptor activation, elevated cell cycle activity, macrophage infiltration, elevated cytokines including TNF-α, altered basement membrane deposition, increased integrin-linked kinase activity and cell migration, elevated periostin, increased angiogenesis, and tumor-like defective glucose metabolism (Figure 2).^40,41 Uremia associated with ESRD can often cause acquired PKD and renal cell carcinoma (RCC). Furthermore, there is recent evidence indicating that even less severe CKD can increase the incidence of RCC and urothelial cancer.⁴² Yet, in PKD, numerous studies have shown that very little RCC is seen⁴³ unless the kidneys are removed and examined histologically.^44,45

This very low level of diagnosed RCC in ADPKD suggested that there may be a degree of protection from cancer associated with PKD mutations. As such, we carried out an epidemiologic study to answer this question.⁴⁶ Data from the Scientific Registry of Transplant Recipients, which contains information on all solid organ transplant recipients in the United States, were linked to 15 population-based United States cancer registries that included data on 50 different cancers. Cancer incidence was compared in PKD versus non-PKD renal transplant recipients, and incidence rate ratios adjusted for age, sex, race/ethnicity, dialysis duration, and time since transplantation were determined. The study included >10,000 kidney recipients with PKD and >100,000 without PKD. After multivariable adjustment, overall cancer incidence was found to be significantly lower in patients with ADPKD by about 16%–17%, suggesting that there is cancer protection associated with PKD.⁴⁶ This was also found to be true for ARPKD, in a study that examined the incidence of colon cancer in unaffected carriers of a recessive PKHD1 mutation.⁴⁷

The reason for the lower cancer risk in patients with ADPKD is not known, but may relate to biologic characteristics of ADPKD itself or to better cancer risk behaviors with these patients. However, the fact that there was a decreased incidence in ARPKD carriers, who would not know their gene status, makes it more likely that an underlying cellular mechanism is protective. Such a mechanism might lie in the unique biology of PKD cells, in which there is loss of intracellular calcium homeostasis and decreased basal calcium.^48,49 This lower calcium allows cAMP activation of the Ras/B-Raf/mitogen-activated protein kinase pathway and would contribute to high levels of c-myc expression, a characteristic PKD hallmark. We have also shown that lower calcium in PKD cells leads to lower levels of Cox-2 expression.

How might these observations explain the cancer protective mechanism? The answer may lie in recent studies demonstrating that high Myc levels⁵⁰ and Cox-2 inhibition⁵¹ prevent cancer metastasis. As such, the unusual PKD biology may paradoxically promote neoplastic cyst growth while preventing progression to metastasis, leading to lower cancer incidence.

Stefan Somlo

The past decade has seen a remarkable evolution in the understanding of the importance of primary cilia in the pathogenesis of structural kidney diseases that manifest with cysts and fibrosis. Primary cilia are single nonmotile hair-like projections on many mammalian cells, including most renal tubular epithelial cells where they reside on the apical surface. They are highly privileged subcellular compartments composed of a microtubular scaffold with overlying plasma membrane that is devoid of other subcellular organelles. All component proteins must be synthesized in the cell body and the cilia are able to select which of these enter and leave the cilia compartment and the respective rates for each of these processes. The importance of cilia to the pathogenesis of PKD was brought to the fore by studies that show that mutations affecting cilia structure and composition produce cysts in mammalian kidney tubules and liver bile ducts and that many of the protein products associated with human and mouse fibrocystic diseases are expressed in and around the cilia compartment.^52,53 Most notably, both PC1 and PC2 are expressed in cilia.^54,55 This led to the hypothesis that polycystin function is critical to cilia function and that defects in cilia function are critical to the pathogenesis of PKD. Although such a unifying hypothesis is both appealing and reasonable, certain features of the respective cilia mutant and polycystin mutant phenotypes are inconveniently inconsistent. Most notably, loss of cilia results in markedly less pronounced polycystic disease than is seen in ADPKD.^56,57

This apparent paradox led us to investigate the genetic interrelationship between inactivation of polycystin proteins alone, of cilia alone, and of polycystin proteins and cilia together.⁵⁸ We found that, as previously noted, inactivation of polycystins alone resulted in severe PKD whereas inactivation of cilia alone resulted in mild disease. Surprisingly, simultaneous inactivation of polycystins and cilia together resulted in a marked decrease in severity of PKD when compared with polycystin-only inactivation. These findings proved universal—they were applicable to Pkd1 and Pkd2, to all segments of the nephron, and to the bile duct and were independent of the timing of gene inactivation (developmental or adult). We were able to show that the severity of PKD was directly related to the period of time for which intact cilia persisted following loss of polycystins. In aggregate, these data suggest that mechanistically polycystins are inhibitory signals that normally modulate a pathway which is yet to be identified, but that requires intact cilia to function. When this pathway is de-repressed due to loss of polycystins, ADPKD ensues.

Jing Zhou

Polycystin-1 and PC2 are integral membrane proteins forming a receptor channel complex.⁵⁹ Over a decade ago, PC1 and PC2 were found on the primary cilia.^54,55,60 Kidney tubular epithelial cells respond to flow shear stress with a calcium signal. Cells without functional PC1 or PC2 are unable to respond to flow-induced shear stress with a calcium signal.^60,61 This calcium signal, visualized by ratiometric calcium imaging, is detectable several seconds after flow stimulation and is amplified by a calcium-induced calcium release mechanism involving a ryanodine receptor.⁶¹ Using a recently developed calcium indicator, the calcium signal induced by flow shear stress can now be detected in the primary cilia.^61,62 This PC2-dependent calcium signal precedes an increase in cytosolic calcium level via a ryanodine receptor, supporting the previous findings.^61,62 This PC1- and PC2-dependent calcium signaling may modulate a number of cellular activities (Figure 2). The ciliary compartment appears to be a separate compartment for calcium signaling.^62,63 Although the ciliary functions of PC1 and PC2 are likely not restricted to calcium signaling, studies defining the local as well as global impact of the calcium signal mediated by ciliary PC1 and PC2 will be instructive in understanding polycystin function.

ADPKD is now known as a ciliopathy. Ciliopathies refer to a group of diseases caused by structural or functional defects of the primary cilia. Ciliopathies have a wide range of phenotypes with cystic kidneys as a common feature. What are the proximal events of ciliary polycystins? We searched for polycystin interaction partners. We found that PC2 interacts with the ARPKD protein, fibrocystin/polyductin,⁶⁴ and nek8 (namely nephronophthisis 9).⁶⁵ Recently we found that PC1 interacts with several Bardet–Biedl syndrome (BBS) proteins that are part of the BBSome protein complex known to function in the transport of a set of proteins to the cilia.⁶⁶ Deficiency of BBS1 and BBS3 affects the ciliary trafficking of PC1. Hence, the polycystins are cargos of the BBSome and physical interactions between the polycystins and BBS proteins may underlie the overlapping renal phenotypes in these two diseases. Other components of the trafficking pathways may well be important for polycystin targeting to the cilia.

Polycystins are present at other subcellular compartments besides the primary cilia.⁵⁹ We recently found that PC1 is present in the lamellipodia in migrating kidney tubular epithelial cells.⁶⁷ binds to an F-bar protein Pacsin 2 and regulates Pacsin 2 interaction with N-Wasp, an activator of the actin nucleator Arp2/3 protein complex. We found that disorganization of the actin cytoskeleton is a feature of PKD in vivo and that a novel PC1-Pacsin 2-N-Wasp complex is required for the actin remodeling and directional cell migration.⁶⁷ Directional cell migration is essential for the regeneration and maintenance of the epithelium and kidney development.⁶⁸ We propose that a defective actin cytoskeleton and directional cell migration contributes cyst formation.

Friedhelm Hildebrandt

Nephronophthisis-related ciliopathies (NPHP-RC) are recessive multisystem disorders that affect kidney, retina, liver, and cerebellum either by prenatal-onset dysplasia or by childhood-onset degeneration and fibrosis. Identification of >19 disease-causing genes (NPHP1 to NPHP19) revealed that their products are located at primary cilia and centrosomes. However, the proximal disease mechanisms remain poorly understood.

We identified, by whole exome resequencing, mutations that affect the centrosomal proteins, namely, FAN1, MRE11, ZNF423, and CEP164, as novel causes of NPHP-RC. Surprisingly, these ciliopathy genes serve functions within the DNA damage response (DDR) pathway.⁶⁹

• (1) ZNF423 interacts with the DNA ds-break sensor PARP1, which recruits MRE11 (MRN) and ATM to sites of DNA damage. ATM, in turn, is activated by MRE11 (MRN) and the ‘TIP60 complex’. We demonstrate colocalization to TIP60-positive (and SC35-positive) nuclear foci or protein–protein interactions for the following products of genes mutated in NPHP-RC: SDCCAG8/NPHP10, ZNF423, CEP164, OFD1, RUVBL1, RUVBL2, NPHP5, NPHP1, and ATXN10. OFD1, RUVBL1, RUVBL2, are known to play a role in DDR.

• (2) In addition, in four different families with NPHP-RC, we identify recessive mutations of CEP164 as a novel cause of NPHP-RC. CEP164 acts in the ATR-Chk1-related arm of DDR, where it is necessary for ATR-dependent Chk1 activation upon induced replication stress.⁷⁰

Furthermore, to identify single-gene causes of renal fibrosis/CKD we performed homozygosity mapping and whole exome resequencing in a model disorder for renal fibrosis known as karyomegalic interstitial nephritis (KIN).⁷¹

We identified recessive mutations of the Fanconi anemia-associated nuclease 1 gene (FAN1) as causing the NPHP-like renal phenotype of karyomegalic interstitial nephritis (KIN) in nine of 10 families ascertained. KIN causes CKD with renal histology indistinguishable from NPHP, except for the presence of karyomegaly. FAN1 has nuclease activity and acts in DNA interstrand crosslinking (ICL) repair within the Fanconi anemia (FA) pathway of DNA damage response (DDR). Interestingly, ICL-causing genotoxins generate a KIN-like phenotype. We demonstrate that cells from individuals with FAN1 mutations exhibit sensitivity to the ICL agent mitomycin C. We complement ICL sensitivity with wild-type FAN1 but not mutant cDNA from individuals with KIN. The FAN1 defect was not epistatic with the Fanconi anemia pathway. By depletion of fan1 in zebrafish we recapitulated increased DDR, apoptosis, and kidney cysts akin to NPHP.⁷²

We suggest a working hypothesis for the pathogenesis of certain forms of NPHP-RC proposing the following cascade of events: defects of DDR lead to a lack of Chk1 (Chk2) activation, thereby causing inadequate G₂/M cell cycle arrest. This would lead in high proliferation states (high replication stress) during morphogenesis to dysplastic phenotypes (Meckel syndrome) and in low proliferation states (low replication stress) during tissue maintenance and repair to tissue degeneration and fibrosis (nephronophthisis).

Corinne Antignac

Virtually all epithelial cells display primary cilia. However, whereas in immature rat glomeruli, podocytes express cilia, these tend to disappear during development.⁷⁴ Hence, our identification of a missense mutation (p.P209L) in the TTC21B gene encoding the intraflagellar transport (IFT) 139 protein in seven families with hereditary focal-segmental glomerulosclerosis, discovered during late adolescence or early adulthood, raised the question of a role of IFT139 (and potentially other IFT proteins) beyond the primary cilium.⁷⁵ Mutations in TTC21B had previously been reported in patients with nephronophthisis,⁷⁶ but careful re-analysis of the clinical and histologic features of all patients bearing the p.P209L mutation clearly showed that they present with both glomerular and tubulo-interstitial involvement. In agreement, we found that IFT139 was predominantly expressed in distal tubules, as expected for a nephronophthisis-causing gene, but was also strongly expressed in glomerular podocytes. IFT139 was mainly localized at the base of the primary cilium in developing podocytes from human fetal tissue and in undifferentiated cultured podocytes. In contrast, in nonciliated adult podocytes and differentiated cultured cells, IFT139 relocalized along the extended microtubule network.

Our functional studies in undifferentiated podocytes showed that the p.P209L mutation has a hypomorphic effect on podocyte ciliogenesis, but was not sufficient to inhibit podocyte differentiation. These data suggested that the glomerular defects observed in patients with this mutation were, rather, due to a nonciliary alteration of IFT139 in mature podocytes, uncovered by the hypomorphic p.P209L mutation. As actin and microtubule cytoskeletons are key regulators of the delicate architecture, plasticity, and contractility of podocytes,⁷⁷ and as IFT139 is redistributed along the microtubule network in mature podocytes, we studied the effect of the p.P209L mutant on cytoskeleton organization in differentiated podocytes. Interestingly, the depletion of IFT139 led to increased alterations to the cell surface associated with actin cytoskeleton, such as short and misorganized stress fibers and microtubule rearrangement into bike-wheel-like shape. All these defects were fully rescued by the wild-type protein. In contrast, p.P209L re-expression led to the rescue of the cell-size defect, but actin and microtubule networks remained severely altered. In addition, microtubule repolymerization after nocodazole treatment resulted in multiple nucleation sites dispersed within the cytoplasm in IFT139-depleted differentiated podocytes and these abnormalities were partially rescued upon p.P209 mutant expression.

Vicente E. Torres

Understanding the pathogenesis of increased fluid secretion and epithelial cell proliferation downstream from PKD1 or PKD2 mutations is important to identify therapies (Figure 2).^40,41 Those targeting proximal mechanisms are more likely to be effective than those targeting distal mechanisms.

Aberrant cross-talk between intracellular calcium and cAMP signaling likely is one of the first effects of PKD mutations. Disrupted calcium may enhance cAMP and protein kinase A signaling through activation of calcium-inhibitable adenylyl cyclases and inhibition of calcium-dependent phosphodiesterases (PDE1 and indirectly cGMP-inhibited PDE3).⁷⁸ Enhanced protein kinase A activity may in turn disrupt intracellular calcium homeostasis through hyperphosphorylation of calcium cycling proteins in the endoplasmic reticulum. Preclinical studies have targeted calcium signaling with some success. Only triptolide (an activator of the polycystin-2 channel) is being clinically investigated. Gαs protein-coupled receptor antagonists (i.e., vasopressin V2 receptor antagonists) and Gαi protein-coupled receptor agonists (i.e., somatostatin analogs) have demonstrated encouraging results in preclinical and clinical trials. Protein kinase A-induced phosphorylation of cystic fibrosis transmembrane conductance regulator (CFTR) allows chloride and fluid secretion into the cysts. Anoctamin-1, a calcium-activated chloride channel, may synergistically interact with CFTR.⁷⁹ CFTR inhibitors inhibit cystogenesis in vitro and in kidney-specific Pkd1 knockout mice, whereas anoctamin-1 inhibitors have been effective in vitro.

Protein kinase A activation inhibits cell proliferation in wild-type cells, but has a stimulatory effect in PKD cells. Calcium deprivation in wild-type cells and delivery of calcium in PKD cells reverse these effects. A proposed mechanism for the proliferative response in PKD and calcium-deprived wild-type cells is inhibition of phosphoinositide 3-kinase and AKT releasing B-Raf from AKT inhibition.⁴⁹ This leads to dysregulation of signaling pathways (B-Raf/MEK/ extracellular signal-regulated kinase [ERK]; AMP-activated protein kinase [AMPK]/mTOR; arguably Wnt/β-catenin) and transcription factors (hypoxia-inducible factor 1, Myc, P53, signal transducer and activator of transcription 3) that control cell cycle progression, energy metabolism, and the microenvironment. Overexpression of growth factors, cytokines, chemokines, and their receptors further contributes to disease progression. Src is a nonreceptor tyrosine kinase activated downstream from G-protein–coupled and growth factor receptors. SKI-606 (bosutinib) is an Src/Abl inhibitor. KD019 inhibits Src and receptor tyrosine kinases (EGF receptor, ERBB2, vascular endothelial growth factor receptor). Both have shown benefit in preclinical studies. Clinical trials are ongoing. On the other hand, targeting of B-Raf/MEK/ERK has given inconsistent results.

Overwhelming evidence indicates that mTOR signaling is enhanced in cystic tissues. Preclinical studies of rapalogs and mTOR antisense oligonucleotides⁸⁰ have been encouraging. Clinical trials have been disappointing^81–83 likely because clinically achievable blood levels do not inhibit mTOR in the kidney. Strategies to overcome systemic toxicity and limited renal bioavailability are being investigated. Pan-sirtuin (nicotinamide) and sirtuin-1–specific inhibitors, AMPK activators (metformin), thiazolidinediones, and signal transducer and activator of transcription 3 inhibitors have been effective in animal models.⁸⁴ Nicotinamide is being tested in an uncontrolled, open label clinical trial. Recent studies suggest that PKD cells, like cancer cells, reprogram their energy metabolism from oxidative phosphorylation to aerobic glycolysis.⁸⁵ Whether drugs interfering with glycolysis will be safe and effective clinically is uncertain.

Gerd Walz

The mTORC1 kinase cascade is almost universally activated in cystic kidney disease independent of the underlying disease-causing gene mutation^87,88 (Figure 2). Although the molecular basis for this activation remains incompletely understood, primary cilia appear to curtail mTORC1 activity through a flow-dependent activation of the LKB1/AMPK signaling pathway, involving stabilization of the tuberous sclerosis complex 1 (TSC1)/TSC2 complex and inhibition of Rheb.⁸⁹ In addition, PC1 might influence the activity of this small GTPase through interaction with the TSC complex,^90,91 and inhibit MEK/ERK-mediated phosphorylation of TSC2 to control mTORC1 activity,⁹² whereas c-Cbl is sequestered in the Golgi apparatus in the absence of polycystin-1, allowing the HGF receptor c-Met to escape degradation after stimulation.⁹³

First shown in the Cy/+ rat model,⁹⁴ the efficacy of mTORC1 inhibitors to reduce cyst growth and to ameliorate disease progression was subsequently demonstrated in several animal models of cystic kidney disease, including Pkd1-deficient mice.^90,95–100 Despite these encouraging preclinical data, mTOR inhibitors did not yield the expected benefits in patients with ADPKD. While everolimus slowed cyst growth in ADPKD patients, this inhibition did not translate into an improvement of renal function. However, other pilot studies yielded more promising results,¹⁰¹ leaving open the role of mTOR inhibition as an ADPKD target.^102,103

To determine the involvement of mTORC1 in cyst growth and ADPKD disease progression, we abrogated ciliogenesis in the thick ascending limb of Henle and distal tubular segments, generating Kif3a fl/fl*KspCre mice. These mice rapidly developed cystic kidney disease, and died before 12 weeks of age (median survival 7.3 weeks). In contrast, the additional elimination of Raptor (Raptor fl/fl*Kif3a fl/fl*KspCre) delayed cyst growth and dramatically increased the survival by more than three times (median survival 24.0 weeks), revealing the importance of mTORC1 in promoting cyst growth. However, despite the remarkable delay, the Raptor-deficient mice nevertheless developed cystic kidney disease. Gene expression profiling revealed that cyst formation and disease progression in the absence of functional mTORC1 were driven by other growth-promoting pathways, including AKT and ERK.

Jared J. Grantham

Cysts originate in tubules and are distinguished from simple tubule dilations by having increased numbers of cells expanding the wall beyond normal boundaries. Cysts are the primary suspects causing the decline of GFR (Figure 3). This is supported by the following findings: (1) cyst enlargement compresses the interstitial matrix and surrounding tubules entraining, distorting, and destroying arterioles, venules, capillaries, and lymphatics; (2) most cysts separate from the tubules from which they formed; (3) an isolated collecting duct cyst can block urine flow in the branching arcade upstream of it that would ordinarily drain hundreds of filtering nephrons^105,106; (4) the glomerulo–tubule junctions of blocked nephrons become thinned and eventually the tubules separate creating atubular glomeruli and apoptotic proximal tubules^107–109; (5) cysts lead to hypertension and impair urine concentrating capacity before GFR declines¹¹⁰; (6) morphologic studies of human and animal kidneys reveal that cysts can form in the absence of interstitial pathology¹⁰⁵; (7) in ADPKD, cortical fibrosis in the absence of cysts nearby is likely secondary to the downstream blockade of collecting ducts¹⁰⁹; (8) a seminal study in PKHD1-null rats unable to synthesize vasopressin proved that fibrosis and renal insufficiency do not develop in the absence of cyst formation, but do when the hormone is directly administered causing cysts to form.¹¹¹ What causes renal failure in PKD? The cyst did it!

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Natural history and pathogenesis of PKD. Total kidney volume (red line) exhibits exponential growth at an average rate of 5% per year, presumably due to cyst epithelial cell proliferation and fluid secretion, although this rate can range widely from patient-to-patient. This causes progressive compression of neighboring structures leading to nephron destruction. Early in the disease, remaining intact nephrons exhibit compensatory hyperfiltration sufficient to preserve GFR (blue line). Further renal injury is mediated by multiple additional mechanisms (purple block arrows) including inflammation and fibrosis. Late in the disease, nephron loss exceeds renal compensatory capacity and GFR starts to decline, reaching a rapid phase generally when the total kidney volume reaches 1500 ml (black arrow), with an inexorable descent toward ESRD (58 years for PKD1, 79 years for PKD2). Advancing uremia is likely to be accompanied by superimposed acquired cystic disease, but apparently without significant increase in the risk of metastatic RCC.