Autosomal dominant polycystic kidney disease (ADPKD): Genetics of the disease and mechanisms of cyst growth

INTRODUCTION — Autosomal dominant polycystic kidney disease (ADPKD) is a common disorder, occurring in approximately 1 in every 400 to 1000 live births [1,2]. It is estimated that less than one-half of affected individuals will be diagnosed during their lifetime since the disease is often clinically silent [1,3].

The genetics of ADPKD and the mechanisms of cyst growth will be reviewed here. Issues related to the diagnosis of and screening for ADPKD and for autosomal recessive polycystic kidney disease, which is a disease of children, are discussed separately. (See "Autosomal dominant polycystic kidney disease (ADPKD) in adults: Epidemiology, clinical presentation, and diagnosis" and "Autosomal recessive polycystic kidney disease in children".)

GENETICS — Mutations in one of two genes, PKD1 or PKD2, account for most cases of ADPKD. The two disease loci segregate independently since they reside on separate chromosomes. The PKD1 gene is located on chromosome 16p13.3, and the PKD2 gene is located on chromosome 4q21. (See 'PKD1 and PKD2 genes' below.)

The estimated minimal prevalence of ADPKD is approximately 1 per 1000 based upon pathogenic protein-truncating PKD1 and PKD2 mutations identified by population-based whole genome and exome sequencing. The disease prevalence is estimated to be higher if nontruncating but likely pathogenic variants are included [3].  

There is evidence that the PKD1 and PKD2 disease loci can have synergistic effects. A family has been reported with bilineal inheritance, in which one parent had a mutation in PKD1 and the other parent had a mutation in PKD2 [4]. The two family members with both mutations progressed to end-stage kidney disease (ESKD) approximately 20 years earlier than family members with only one PKD1 or PKD2 mutation.

A few families with ADPKD are not linked to either locus [5-7].

A third ADPKD gene has not been identified. A study that re-evaluated five such families has shown that the absence of linkage was due to misdiagnosis, DNA contamination, and genotype errors and did not support the presence of a third gene [8]. On the other hand, at least seven genes (ie, ALG5, ALG8, ALG9, SEC 61B, SEC63, GANAB, and PRKCSH) have been identified for autosomal dominant polycystic liver disease (ADPLD), which can be associated with very mild to moderate kidney phenotype (ie, a few to multiple but not innumerable kidney cysts) [9-13]. Additionally, DNAJB11 and IFT140 mutations can also be a rare cause of atypical ADPKD [14,15]. Some patients with DNAJB11 mutations have CKD and cystic kidney atrophy rather than enlargement [14]. Patients harboring mutations in one of these genes may be misdiagnosed with ADPKD but are generally not at risk for progressive chronic kidney disease. (See 'PLD gene products modify polycystin-1 dose' below.)

Prevalence of PKD1 and PKD2 — Initial studies found that mutations in PKD1 were responsible for 96 percent of cases of ADPKD [16,17]. Subsequent reports found a lower incidence of PKD1 of 86 percent in a study from Europe [18] and 74 percent in a study from Canada [19].

A possible contributing factor to these findings is that patients with PKD2 have a later onset of clinically apparent disease than patients with PKD1, which can lead to delayed diagnosis. Macroscopic cysts occur later in PKD2, as do clinical findings such as hypertension, kidney pain, and development of ESKD. (See "Autosomal dominant polycystic kidney disease (ADPKD) in adults: Epidemiology, clinical presentation, and diagnosis", section on 'Epidemiology'.)

The following findings are illustrative:

●In a report from the European PKD1-PKD2 study group that included 333 patients with PKD1 and 291 with PKD2, the patients with PKD2 were less likely to have hypertension (odds ratio [OR] 0.25) or gross (macroscopic) hematuria and progressed to ESKD at an older age (median 74 versus 54 years) [20].

●In a study from Canada that included 484 patients, the presence of at least one affected family member with ESKD at age ≤55 years was highly predictive of a PKD1 mutation [19]. In contrast, maintenance of sufficient kidney function to not require dialysis or the development of ESKD until after 70 years of age in at least one affected family member was highly predictive of a PKD2 mutation. The positive predictive value was 100 percent for both findings with a sensitivity of 72 to 74 percent.

●A study performed in the Genkyst cohort conducted mutation analysis on 741 patients from 519 pedigrees in the Brittany region of France. PKD1 and PKD2 mutations were identified in 75.5 percent and PKD2 in 18.3 percent of patients, respectively. The median age of ESKD was 58.1 years in PKD1 mutation carriers versus 79.7 years in those individuals with PKD2 mutations. The mean age of hypertension diagnosis was approximately 10 years earlier in PKD1 compared with PKD2 (37 years versus 49 years, respectively) [21].

PKD1 and PKD2 genes — The genes for both PKD1 and PKD2 have been identified. The PKD1 gene is located on chromosome 16p13.3 [16,17,22-25], and the PKD2 gene is located on chromosome 4q21 [18,25-27]. The gene products, polycystin-1 and polycystin-2, and their function are discussed below. (See 'PKD gene products' below.)

The PKD1 gene is both a large and complex gene (46 exons). Exons 1 to 34, which comprise two-thirds of the coding sequence, are highly similar to several nearby genes, making it challenging to perform mutation screening in this "duplicated" region. In contrast, the PKD2 gene is much smaller, and mutation screening of this gene is straightforward [28].

The PKD1 gene is adjacent to a disease gene for tuberous sclerosis (TSC2) [22], a disorder that is characterized primarily by renal angiomyolipomas and kidney cysts. Contiguous deletions of both the PKD1 and TSC2 genes cause a rare but severe form of PKD associated with ESKD that typically occurs in childhood or teenage years [29,30]. (See "Renal manifestations of tuberous sclerosis complex", section on 'TSC2/PKD1 contiguous gene syndrome' and "Tuberous sclerosis complex: Genetics and pathogenesis", section on 'Genotype-phenotype correlations'.)

The availability of protocols for locus-specific amplification of the PKD1 gene has enabled comprehensive mutation screening in large cohorts of patients with ADPKD [31,32]. A large number of mutations have been described in both the PKD1 and PKD2 genes, with most mutations being unique to single families [21,31]. Most of these mutations are due to nonsense, frameshift, or splice-site alterations that are predicted to be protein truncating and inactivating. The clinical significance of up to one-third of the other PKD1 mutations is less clear. Advances in next-generation sequencing technologies have been used for high-throughput mutation screening of both PKD1 and PKD2, with a "proof of principle" study showing promising results [33]. The adaptation of this new technology for molecular diagnostics in ADPKD has the potential to facilitate mutation screening at reduced costs.

There is variability in the kidney manifestations of ADPKD. Some PKD1 missense variants reduce, but do not abolish, PKD1 activity and are called "hypomorphic alleles." These alleles are associated with a milder form of ADPKD, similar to PKD2 [21,34-36]. On the other hand, the presence of a hypomorphic allele may modify disease severity in rare families with bilineal inheritance of two PKD1 mutations (ie, one from each parent), resulting in prenatal PKD in the offspring who carry both PKD1 mutations [37]. In addition, within-family variability in ADPKD kidney manifestations suggests a significant contribution from modifier genes, independent of the germline (inherited) mutation [38-41]. Identification of these modifiers is expected to be difficult since multiple genes are likely to be involved, each with only a modest effect size [41].

The role of genetic factors in the rate of progression of ADPKD is discussed elsewhere. (See "Autosomal dominant polycystic kidney disease (ADPKD) in adults: Epidemiology, clinical presentation, and diagnosis", section on 'Epidemiology'.)

Some patients with PKD do not have detectable mutations in either PKD1 or PKD2 genes; some of these patients may have somatic mosaicism. Patients with mosaicism have two genetically different populations of cells due to a de novo mutation occurring during early embryonic development [42,43]. This additional de novo mutation, affecting a subset of cells, results in the dilution of the mutant gene dose. Cells collected from these patients produce a low signal-to-noise ratio on Sanger sequencing, potentially leading to false negative results for PKD1 or PKD2 gene mutations [43]. In one study, approximately 15 percent of patients suspected to have ADPKD did not have a mutation detected by Sanger sequencing in either the PKD1 or PKD2 genes [36]. Most of these patients did not have a family history of ADPKD and they exhibited atypical findings on imaging [42]. In another study, patients from 20 such families (without detectable PKD1 or PKD2 mutations) were diagnosed as having somatic mosaicism using next-generation sequencing with high read-depth [44].      

PKD gene products — The PKD1 and PKD2 genes encode proteins called polycystin-1 and polycystin-2, respectively [22-25,27,28,45].

PKD1 is a large gene (46 exons), encoding polycystin-1, which is more than 4000 amino acids in length [22-24,28]. Polycystin-1 is predicted to contain 11 transmembrane domains, a short intracellular C-terminus, and a large extracellular region with a novel combination of motifs [22,23], suggesting that it participates in adhesive cell-cell or cell-matrix interactions. The protein undergoes a variety of cleavage events, including at a juxtamembrane G protein-coupled receptor (GPCR) proteolytic site (GPS) motif. This cleavage does not require an exogenous protease and is therefore said to occur via "cis-autoproteolysis." Cleavage at the GPS site results in two fragments that remain noncovalently associated. This cleavage event is required to maintain tubular morphology since a knock-in mouse model with a mutation at the GPS site results in ADPKD with cysts primarily affecting the collecting duct [46,47]. In addition, mechanical stimuli result in the intracellular release of the cytoplasmic tail [48,49]. This protein fragment is then translocated into the nuclei to activate transcriptional pathways such as activating protein-1 (AP-1) [48,50,51]. One report suggests that the polycystin-1 cytoplasmic tail also translocates to the mitochondrial matrix where it could regulate mitochondrial structure and function [52].

The PKD2 gene is much smaller than PKD1 (15 exons) and encodes polycystin-2, which is less than 1000 amino acids in length [28]. Polycystin-2 is a calcium-permeable channel with six transmembrane (TM) spanning elements that is a member of the transient receptor potential (TRP) family of nonselective cation channels [45]. Polycystin-2 also has homology to the last six transmembrane domains of polycystin-1 [27]. Four polycystin-2 channels form a tetrameric structure with a voltage-sensing domain, a pore loop between TM5 and TM6, and a novel "TOP" domain comprised of the large extracellular loop between TM1 and TM2. [53,54]. The TOP domain is postulated to be critical for channel assembly and/or function since it is a hotspot for missense disease-causing variants [54]. Polycystin-2 interacts with Polycystin 1 via their C-termini and possibly TOP domains [53-55].  

PLD gene products modify polycystin-1 dose — At least five genes have been identified to cause ADPLD with a mild kidney phenotype, including ALG8, SEC 61B, SEC63, GANAB, and PRKCSH, which respectively encode alpha-1,3-glucosyltransferase, SEC61-beta, SEC63p, and alpha- and beta-catalytic subunits of glucosidase II [11,12]. The proteins encoded by these genes all function in the endoplasmic reticulum (ER) posttranslational protein biogenesis pathway, and mutations in any of these genes have been shown to reduce the maturation and cell membrane expression of polycystin-1. (See 'Polycystin function' below.)

Polycystin localization — The sites of localization of polycystin-1 and polycystin-2 have been controversial due to issues of antibody specificity [24,56,57]. The following features are characteristic of both proteins:

●Polycystins are localized in renal tubular epithelia, hepatic bile ductules, and pancreatic ducts, all of which are sites of cyst formation in ADPKD. Polycystins also are expressed in endothelial and vascular smooth muscle cells and are found in major blood vessels such as the aorta, aortic branches, and the Circle of Willis [58-60].

●Polycystins are overexpressed in most, but not all, cyst epithelial cells in kidneys from patients with ADPKD.

●Polycystins are integral membrane proteins that are found primarily in plasma membranes but also in the primary cilium. The primary cilium is an organelle that arises from the surface of most cells and acts as cell sensors that facilitate the interactions between the cell and its environment [61].

Several studies have shown that there are various cellular pools of polycystin-1 and polycystin-2. Polycystin-1 and polycystin-2 interact in the endoplasmic reticulum, and this interaction is required for the mature polycystin complex to reach the cell surface, including the cilium. Similarly, polycystin-1 cleavage is necessary for ciliary localization of both polycystin-1 and polycystin-2 [62-64].

Polycystin function — Studies analyzing the proposed protein structure of the PKD1 gene product, polycystin-1, and its membrane localization in cells in culture suggest that it is involved in adhesive protein-protein, cell-cell, and/or cell-matrix interactions [23,56]. As an example, polycystin-1 localizes to the lateral membranes of cells, making cell-cell contact in vitro [56]. In addition, polycystin-1 may exist in a complex with E-cadherin and the catenins, two proteins that are involved in cell adhesion [57].

Thus, an abnormality in polycystin-1 may impair cell-cell and cell-matrix interactions, leading to abnormal epithelial cell differentiation and the various phenotypic expressions of ADPKD. The demonstration of impaired in vitro tubulogenesis in cyst-derived cells from ADPKD kidneys is compatible with this hypothesis [65].

A ligand for the polycystin complex has not been identified [25]. However, it has been proposed that the cleaved N-terminus of polycystin-1 can act as a soluble ligand that activates the heteromeric polycystin-1/polycystin-2 complex [66].

The PKD2 gene product, polycystin-2, is thought to be involved in calcium signaling [25,45]. There is strong evidence that polycystin-2 functions as a calcium-regulated nonselective cation channel:

●Expression studies using polycystin-L (PKDL), a protein with high homology to polycystin-2 and TRP cation channels, show that PKDL functions as a cation channel that is permeable to calcium, sodium, and potassium [67]. In the kidney, it is predominantly localized to principal cells in the inner medullary collecting ducts [68].

●Interaction between polycystin-1 and polycystin-2 creates a new, calcium-permeable, nonselective cation current, although the exact role for polycystin-1 is unclear [45]. Such activity is not observed with either protein alone or with disease-associated mutant proteins incapable of heterodimerization.

●Direct ciliary patch clamp recordings from cultured cells lacking polycystin-2 demonstrated a ciliary current that is dependent on PKD2 [69]. This channel appears to have high conductance with permeability to K⁺>Ca²⁺>Na⁺. This is in keeping with the idea that polycystin-2 is a nonselective cation channel.

●Cryogenic electron microscopy studies of the six polycystin-2 transmembrane domains (S1-S6) in lipid nanodiscs show that in the absence of polycystin-1, the polycystin-2 channel has a tetrameric structure that forms the channel pore [53]. This shortened polycystin-2 channel conducts Na⁺ and K⁺ but has a lower permeability and smaller single-channel conductance to Ca²⁺. Polycystin-2 channels have a unique structural feature, called the Polycystin or tetragonal opening for polycystins (TOP) domain, which is a large extracellular loop between S1 and S2 [53,54]. These domains work with other domains forming a lid-like structure that may open the channel pore in response to voltage-dependent conformational changes. Missense mutations in the TOP domain de-stabilize the channel structure and impair channel opening without altering the localization of cilia [70].

Since the clinical features of ADPKD1 and ADPKD2 are identical, it has been suggested that polycystin-1 and polycystin-2 interact in a common signaling pathway(s) [25,28,45]. Several lines of evidence provide support for this hypothesis:

●Polycystin-1 and polycystin-2 are expressed in similar cellular and subcellular locations, although the overlap is not complete [24,56,57].

●Polycystin-1 appears to activate the JAK-STAT pathway, thereby inducing cell cycle arrest [71]. Polcystin-2 functions as an essential cofactor for this process. Activation of the pathway is prevented by mutations that disrupt the interaction between polycystin-1 and polycystin-2.

●Polycystin-1 and polycystin-2 interact via their carboxy terminal coiled domains and TOP domains [53-55]. There are naturally occurring pathogenic mutations in the coiled domains that disrupt this interaction.

●Both polycystins appear to regulate G protein signaling [72]. The carboxy terminal end of polycystin-1 can bind heterotrimeric G proteins, while polycystin-2 interacts with polycystin-1 to antagonize signaling.

The precise function of polycystin-1/polycystin-2 heteromers are unclear. Cryogenic electron microscopy studies of shortened polycystin protein fragments suggest heteromer assembly with a 1:3 polycystin-1/polycystin-2 ratio [73]. In this structure, the last six transmembrane domains of polycystin-1 (those with homology to polycystin-2) interdigitate with polycystin-2 to form a structure similar to other TRP-like channels, including the polycystin-2 homomer. Whether this heteromeric channel is functional, however, remains a question because the inclusion of polycystin-1 appears to disrupt the pore structure [73].

Genome editing of PKD1 and PKD2 — Genome editing technologies have provided the possibility of future therapeutic interventions to correct the primary defect of ADPKD. Genetic reactivation of PKD1 or PKD2 in adult mouse models of ADPKD resulted in reversal of ADPKD even with moderately severe disease. These experiments demonstrated extensive plasticity of the kidneys in tissue repair and remodeling [74].

MECHANISM OF CYST FORMATION AND GROWTH — The underlying mechanisms by which cysts form in ADPKD are not fully understood [50,75]. A number of theories have been proposed, including weakening of tubular basement membranes and intratubular obstruction from hyperplastic cells. However, there has been little evidence to support these hypotheses. Furthermore, the observation that human PKD cells, but not normal renal epithelial cells, can form cysts when grown in culture suggests a genetic defect that directly promotes cyst formation [76,77].

Kidney cysts in ADPKD are derived from different nephron segments, and the epithelial cells lining the cysts appear to retain their transport functions [75,78]. Sodium concentrations vary within different cysts, generally being <60 mEq/L (low-sodium cysts) or >75 mEq/L (high-sodium cysts) [79]. It has been proposed that cysts with high fluid sodium derive from the proximal tubule (particularly in cysts with a sodium concentration similar to that of plasma), while cysts with low fluid sodium derive from the distal nephron (the site at which a low urine sodium concentration is normally achieved). However, careful histological examination has been unable to confirm these relationships since there is a continuum of cyst sodium concentrations, not just high or low values [79]. Fluid secretion in the cysts may be mediated in part by the cystic fibrosis transmembrane conductance regulator (CFTR). (See 'Abnormal fluid secretion' below.)

Early cysts begin as dilatations of intact tubules that are in contact with the nephron and fill by glomerular filtration [27,75]. In contrast, enlarging cysts lose their connection to functioning nephrons as they reach a size of more than 2 to 3 mm. Cyst growth in this setting results from secretion of fluid into the cysts (not glomerular filtration) and is associated with hyperplasia of the cyst epithelium that may reflect underlying maturational arrest [75].

Human PKD cyst fluid, via an undefined mechanism, promotes additional fluid secretion by a lipid secretagogue [80] and cyst formation when added to renal tubular cells in culture [76]. The rate of increase in total kidney volume was evaluated in a clinical study in which 232 patients with ADPKD and no azotemia were followed for three years with serial magnetic resonance imaging (MRI) studies used to measure the rates of change in total kidney and cyst volume [81]. The following findings were noted:

●Total kidney volume was 1060 mL at baseline and increased at a mean rate of 204 mL (5.3 percent) at three years. The mean increase was greater with PKD1 than PKD2 mutations (245 versus 136 mL).

●Total cyst volume increased by a mean of 218 mL at three years and therefore accounted for all of the increase in total kidney volume.

Higher baseline total kidney volume was associated with a more rapid decrease in kidney function. A baseline total kidney volume >1500 mL was present in 135 patients and was associated with a mean decline in glomerular filtration rate (GFR) of 4.3 mL/min per year.

Focal cyst formation and the 'second hit' hypothesis — A notable feature of ADPKD is variable phenotypic disease expression. Even though the germline (inherited) genetic defect is present in all cells, cysts form in <10 percent of tubules, and, within tubules, cystic dilatation is focal [25,28,75]. This observation led to a "second-hit" hypothesis of cystogenesis for both PKD1 [82-84] and PKD2 [85]. Under the "second hit" hypothesis, cysts form in the presence of an inherited PKD1 or PKD2 mutation only if the remaining normal copy of PKD1 or PKD2 develops a somatic (acquired not inherited) mutation. Accordingly, studies of individual PKD1 kidney cysts, which are monoclonal and derived from a single cell, show both loss of heterozygosity and somatic PKD1 mutations [82,83,86].

This model of cystogenesis has been questioned by some because the original studies detected a low percentage (<30 percent) of PKD1 cysts with loss of heterozygosity or somatic PKD1 gene mutations [28]. The low detection rate of somatic mutation, however, was likely due to the challenges of performing mutation screening in the duplicated region of PKD1 (see 'PKD1 and PKD2 genes' above). By comparison, when the entire PKD2 gene is screened, approximately 80 percent of kidney and liver cysts from individuals with germline (inherited) PKD2 mutations have inactivating PKD2 somatic mutations [85,87]. Now, with state of the art sequencing methods, such as whole genome sequencing or next generation sequencing of long range PCR products, somatic mutations of PKD1 or PKD2 can be detected in the vast majority (>90 percent) of kidney cysts [88].

Additional evidence in support of the "second-hit" hypothesis comes from studies in murine models. The following observations are illustrative:

●Kidney and pancreatic cysts develop in mice homozygous for PKD1 or PKD2 null (lacks molecular function) mutations, but do not occur in heterozygotes [84].

●An unstable PKD2 allele was designed to undergo random homologous-recombination, which results in a null allele that can occur in cells after birth [89]. This mimics the ongoing accumulation of somatic (acquired not inherited) mutations that is postulated to occur in human ADPKD. When this unstable allele was combined with a germline PKD2 knock-out allele, mice developed kidney cystic disease, resulting in kidney failure and early death.

●Mutant mice with aberrant mRNA splicing that significantly reduces the level of functional polycystin-1 develop kidney cysts, which is consistent with the hypothesis that cyst formation can be triggered by polycystin levels below a critical threshold [90,91].

The above findings, along with reports of cystic disease in patients with homozygous hypomorphic (reduced level of activity) PKD1 mutations or with a heterozygous hypomorphic PKD1 mutation associated with a second inactivating PKD1 mutation, suggest a "threshold mechanism of cystogenesis" model of disease [25,92]. As an example, a dose of functional polycystin that falls below a critical threshold (approximately 10 to 30 percent of normal) within a tubular epithelial cell may be sufficient to initiate cyst formation. It has been proposed that falling below this threshold leads to abnormal fluid secretion, dysregulated cell proliferation, and apoptosis, which promote cyst growth via multiple signaling pathways [25,75,77,91,93,94].

The time at which the PKD1 gene is inactivated is another factor that may modify kidney disease severity, which suggests the existence of a developmental switch. In a murine model, inactivation of the PKD1 gene prior to 13 days after birth, while the kidney is still developing, resulted in severe cystic disease within three weeks. In contrast, inactivation of the gene after 14 days of age resulted in a more indolent course, with cyst development after five months [95]. Kidney cystic disease in the late induction model can be accelerated by acute kidney injury or tubular calcium oxalate crystal deposition (also called the "third-hit model") [96-98]. These observations suggest that additional factors such as the "on and off" states of certain cell-signaling pathways are needed for kidney cyst growth.

Abnormal fluid secretion — Increased fluid secretion into cysts, possibly mediated by the CFTR, appears to play a role in cyst growth. Fluid accumulation within ADPKD cysts is thought to be driven primarily by 3′,5′-cyclic adenosine monophosphate (cAMP)-dependent chloride secretion [99]. One hypothesis is that chloride transport across the apical membrane occurs via the CFTR. CFTR has the appropriate characteristics since it functions as a cAMP-dependent chloride channel and has been localized to the apical membranes of ADPKD cyst lining cells. Therefore, CFTR has been postulated to contribute to cyst growth [80,100,101]. (See 'Role of cyclic AMP and intracellular calcium' below.)

Suggestive evidence supporting an important pathogenic role for CFTR is provided by the following observations:

●Case reports of a few families with ADPKD and cystic fibrosis reported that individuals with both ADPKD and cystic fibrosis gene mutations may have less severe cystic disease in the kidney and liver compared with family members with ADPKD but without cystic fibrosis [102]. (See "Autosomal dominant polycystic kidney disease (ADPKD): Treatment".)

●Unlike normal renal epithelial cells, in vitro growth of ADPKD cells, as well as enlargement of associated cysts, is markedly stimulated by the addition of cAMP to the culture medium [103]. As mentioned above, CFTR functions as a cAMP-dependent chloride channel.

●Small molecule inhibitors of CFTR slow cyst expansion in both cell culture and mouse models of PKD [104].

These findings support a pathogenic role of epithelial chloride secretion in the generation and maintenance of fluid-filled cysts in ADPKD and provide novel targets for potential therapeutic intervention.

Role of cyclic AMP and intracellular calcium — Derangements in cAMP related to decreased intracellular calcium signaling may underlie the development of cysts via increased fluid secretion and cell proliferation [105-108]. As noted in the preceding section, the CFTR functions as a cAMP-dependent chloride channel.

Strategies that target cAMP signaling and/or calcium homeostasis have been studied in rodent models of PKD with the following observations:

●Dual deletion of PKD1 and adenylate cyclase 6 in principal cells of the collecting duct markedly decreased cystogenesis, improved kidney function, and increased survival. Inhibitors of specific adenylate cyclase isoforms are not available, but this proof-of-concept study is consistent with other data showing that decreasing levels of cAMP are protective in ADPKD [109].

●The administration of a vasopressin V2 receptor antagonist prevents kidney enlargement and inhibits cystogenesis by decreasing kidney cAMP levels [110]. This approach has been validated in two large, randomized, control trials to provide the first mechanism-based therapy for patients with ADPKD [111,112]. (See "Autosomal dominant polycystic kidney disease (ADPKD): Treatment", section on 'Tolvaptan'.)

●Calcium channel blockers, such as verapamil, increase PKD progression in a rat model of PKD that is not caused by mutations in either PKD1 or PKD2 [113].

Role of mammalian (mechanistic) target of rapamycin (mTOR) — Activation of the mTOR protein may contribute to cyst growth in ADPKD [114]. In contrast, inhibition of mTOR with rapamycin (sirolimus) preserved kidney function and inhibited epithelial cell proliferation and fibrosis in a mouse model of ADPKD in which the PKD1 gene was conditionally deleted [115].

Clinical trials have suggested that mTOR inhibitors may slow the increase in liver and kidney volumes in patients with ADPKD, but do not appear to preserve kidney function, at least over the short term. The limited efficacy may be due in part to low tissue concentrations. In addition, toxicity is a major issue that may limit the therapeutic efficacy of these drugs.

More selective rapamycin delivery that preferentially targets cystic tissues might increase therapeutic efficacy and minimize rapamycin toxicity. One approach that has been evaluated is the administration of folate-conjugated rapamycin [116]. The folate receptor is highly expressed in cyst-lining cells in both ADPKD and mouse models. The folate conjugate is cleaved intracellularly to reconstitute the active drug. In a mouse model of PKD, folate-conjugated rapamycin slowed the growth of kidney cysts and preserved kidney function [116]. mTOR activity was reduced in kidneys, but not other organs, which may minimize rapamycin toxicity.

Impaired glucose metabolism and possible role for metformin — Altered glucose metabolism may play an important role in kidney cyst formation. In a knock-out mouse model of PKD1 and in kidneys from humans with ADPKD, there is a preferential shift of energy metabolism to aerobic glycolysis in renal cells [117]. Targeting this pathway in the knock-out mouse by the administration of a nonmetabolized glucose analog, 2-deoxyglucose, led to a reduction in the severity of the kidney cystic disease by modulating both mTOR and adenosine monophosphate-activated protein kinase (AMPK) (see 'Role of mammalian (mechanistic) target of rapamycin (mTOR)' above).

Metformin is a modulator of glucose metabolism that has been proposed as a treatment for ADPKD. Metformin is an indirect pharmacologic activator of AMPK, which downregulates mTOR and CFTR, proteins that drive cyst expansion through proliferative and secretory effects. Metformin has been shown to slow kidney cyst growth in mouse and miniature pig models of ADPKD, at least in part by inhibiting both the mTOR and CFTR pathways [118-120]. Similarly, salsalate, a prodrug dimer of salicylate, which is a direct AMPK activator, has also been shown to slow cystic disease progression in an adult-onset Pkd1 conditional mouse model [121]. These findings provide the rationale for repurposing AMPK activators for clinical testing in patients with ADPKD. Two phase 2 trials of metformin in patients with ADPKD and eGFR >50 have shown that the agent is safe and tolerable [122,123]. Larger trials will be necessary to determine efficacy.

Dietary interventions that affect glucose metabolism may also hold promise as a means of slowing cyst growth. Dietary caloric restriction (10 to 40 percent) in Pkd1^RC/RC mice (mice homozygous for a missense PKD1 mutation that affects protein trafficking) beginning at six weeks of age for six months or at 5.5 months of age for three months both demonstrated a dose-dependent reduction in the kidney cystic disease severity by suppressing the mTOR pathway by activating AMPK signaling [124,125]. In addition, experimental intermittent fasting without caloric restriction and ketogenic diet may also be beneficial [126].

Role of JAK-STAT signaling — STAT6, which is activated by the cytokines interleukin (IL)-4 and IL-13, may be a therapeutic target in ADPKD. Studies in mutant animal models have shown that not only is STAT6 activated in cyst lining epithelia, but cyst fluid contains elevated levels of IL-13, suggesting that there may be a positive feedback loop involving autocrine signaling by IL-13. Interestingly, microdissected cysts from ADPKD kidneys also had elevated expression of IL-13 receptor isoforms. Treatment of a nonorthologous mouse model of PKD with leflunomide, which has several molecular targets including STAT6, ameliorated cystic kidney disease [127].

Angiogenesis — Angiogenesis may play a role in cyst growth and disease progression in ADPKD [128,129]. This is supported by the observation in kidneys of patients with ADPKD that a rich network of capillary neovascularization is embedded within the walls of macrocysts and that kidney cysts display a profile of gene and protein signatures indicative of active angiogenesis [94,129,130]. It is unclear if antiangiogenic interventions have a therapeutic role in ADPKD.

Abnormal cilia function — Virtually all of the protein products that have been implicated in a wide array of human and animal cystic kidney diseases localize to the primary cilia of renal tubular cells [131,132]. These protein products include polycystin-1 and polycystin-2 in ADPKD, fibrocystin in autosomal recessive polycystic kidney disease, and the nephronophthisis gene products. (See 'PKD gene products' above and "Autosomal recessive polycystic kidney disease in children", section on 'Pathogenesis' and "Genetics and pathogenesis of nephronophthisis", section on 'Genetics'.)

There are many potential ways impaired function of primary cilia may impact cyst formation in ADPKD. One hypothesis is that loss of primary cilia can induce kidney cyst formation by disrupting the normal orientation of tubular epithelial cells, which is called planar cell polarity [133,134]. Instead of maintaining a normal elongated tubular geometry, defective planar cell polarity signaling results in saccular dilatation of tubules, which eventually evolves into a cyst.

Although both polycystin-1 and polycystin-2 are located in the primary cilia, the mechanism(s) by which defective polycystins may affect cilia function and lead to cyst formation in ADPKD remains unclear. One possibility is that the inability to sense fluid flow by the defective ciliary polycystin complex may result in reduced intracellular calcium flux, which then stimulates cAMP signaling, causing dysregulated cyst growth [25,132,135,136]. One study calls this hypothesis into question since depletion of polycystin-2 in the cilia of retinal pigment epithelial cells does not alter ciliary calcium currents measured by direct patch clamp methods. However, depletion of polycystin homologs, PKD2L1 and PKD1 L1, which interact with one another, appear to mediate ciliary calcium currents [137] (see 'PKD gene products' above). One line of evidence suggesting that polycystins may act as ciliary sensors comes from the finding that PKD2 mutant mice have situs inversus, a phenotype that is classically associated with ciliary function in the embryonic node [138].

The connection between cilia, growth, and PKD is complex. Mutations in genes involved in ciliary assembly such as TG737 and Kif3a result in PKD in mice. Studies using a combination of mutations in mice have shown that deleting cilia unexpectedly suppresses cyst growth. One explanation is that cyst growth is dependent on intact cilia. When cilia are deleted in the absence of polycystins, growth is reduced [139]. Insight into the connection between abnormal cilia function and dysregulated cell growth may also be provided by observations that cilia are connected to centrosomal structures [140]. The centrosome, in turn, may have a significant role in the regulation of cell-cycle progression and control, with abnormalities in the cilia/centrosome apparatus possibly resulting in abnormal cellular proliferation. Additional support for this is provided by the observation that most of the proteins associated with the Bardet-Biedl syndrome (which has cystic/dysplasia) also localize to centrosomal structures [141,142].

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Polycystic kidney disease (The Basics)")

●Beyond the Basics topic (see "Patient education: Polycystic kidney disease (Beyond the Basics)")

SUMMARY

●Genetics of autosomal dominant polycystic kidney disease – Autosomal dominant polycystic kidney disease (ADPKD) occurs in approximately 1 in every 400 to 1000 live births. Fewer than one-half of cases are diagnosed during the patient's lifetime as the disease is frequently clinically silent. (See 'Introduction' above.)

•PKD1 and PKD2 mutations – Multiple causative genetic defects have been described, most of which are within the PKD1 locus on chromosome 16, resulting in ADPKD1 disease. A smaller number of patients have a defect in the PKD2 locus on chromosome 4 (resulting in ADPKD2 disease), and a few families have neither PKD1 nor PKD2 mutations. (See 'Prevalence of PKD1 and PKD2' above.)

•PKD gene products – PKD1 and PKD2 encode proteins called polycystin-1 and polycystin-2, respectively. (See 'PKD1 and PKD2 genes' above.)

•Polycystin localization, function, and interaction

-Polycystin-1 – Polycystin-1 is a membrane protein localized at sites of cyst formation, including renal tubular epithelia, hepatic bile ductules, and pancreatic ducts. Possible roles of polycystin-1 include the regulation of protein-protein, cell-cell, and cell-matrix interactions and of intracellular signaling pathways involved in the regulation of cell proliferation and survival. (See 'Polycystin localization' above and 'Polycystin function' above.)

-Polycystin-2 – Polycystin-2 is expressed in the distal tubules, collecting duct, and thick ascending limb and is involved in cell calcium signaling. (See 'Polycystin localization' above and 'Polycystin function' above.)

-Interactions between polycystin-1 and polycystin-2 – Polycystin-1 and polycystin-2 also interact in common pathways and function together as an important regulator of calcium influx as a nonselective calcium ion channel. (See 'Polycystin function' above.)

●Mechanism of cyst formation and growth – The mechanism by which cysts form is unclear. Cyst formation may result from multiple aberrant processes including abnormal fluid secretion and cell proliferation and apoptosis. Early cysts begin as dilatations in intact tubules that are in contact with the nephron. However, enlarging cysts lose their connection to functioning nephrons. Cyst growth in this setting results from secretion of fluid into the cysts and is associated with hyperplasia of the cyst epithelium. (See 'Mechanism of cyst formation and growth' above.)

•The 'second hit' hypothesis – In addition to the inheritance of an abnormal PKD1 gene, cysts appear to form only when there is a second, acquired somatic loss of the normal haplotype; this is referred to as the second-hit hypothesis. More recent data suggest a "threshold mechanism" whereby functional polycystin levels below a critical threshold (ie, 10 to 30 percent) within a tubular epithelial cell may trigger the cystogenic process. As a result, although the genetic defect is present in all cells, cysts form in <10 percent of tubules, and cystic dilatation is focal within tubules. (See 'Focal cyst formation and the 'second hit' hypothesis' above.)

•Cyclic AMP and mammalian (mechanistic) target of rapamycin – Elevations in cyclic adenosine monophosphate (cAMP) related to decreased intracellular calcium signaling may enhance cyst formation by increasing fluid secretion and cell proliferation. Vasopressin V2 receptor antagonists, increases in the intracellular calcium concentration, and calcimimetics all may decrease cyst growth. Aberrant activation of mammalian (mechanistic) target of rapamycin (mTOR) signaling pathway may also enhance kidney cyst growth. (See 'Role of cyclic AMP and intracellular calcium' above and 'Role of mammalian (mechanistic) target of rapamycin (mTOR)' above.)

•Angiogenesis – Angiogenesis may contribute to cyst growth. (See 'Angiogenesis' above.)

•Abnormal cilia function – Abnormalities of kidney cilia function may contribute to kidney cyst formation. The potential causes and consequences of cilia dysfunction in ADPKD remain unclear. (See 'Abnormal cilia function' above.)