Inhibition of Cdc25A Suppresses Hepato-Renal Cystogenesis in Rodent Models of Polycystic Kidney and Liver Disease

Abstract

Background & Aims

In polycystic kidney (PKD) and liver (PLD) diseases, the normally non-proliferative hepato-renal epithelia acquire a proliferative, cystic phenotype, which is linked to overexpression of Cdc25A and cell cycle deregulation. We investigated the effects of Cdc25A inhibition in mice and rats, via genetic and pharmacological approaches.

Methods

Cdc25A^+/− mice (which have reduced levels of Cdc25A) were cross-bred with Pkhd1^del2/del2 mice (which have increased levels of Cdc25A and develop hepatic cysts). Cdc25A expression was analyzed in livers of control and PCK rats, control and Pkd2^ws25/− mice, healthy individuals, and patients with PLD. We examined effects of pharmacologic inhibition of Cdc25A with Vitamin K3 (VK3) on the cell cycle, proliferation, and cyst expansion in vitro; hepato-renal cystogenesis in PCK rats and Pkd2^ws25/− mice; and expression of Cdc25A and the cell cycle proteins regulated by Cdc25A. We also examined effects of the Cdc25A inhibitor PM-20 on hepato-renal cystogenesis in Pkd2^ws25/− mice.

Results

Liver weights and hepatic and fibrotic areas were decreased by 32%–52% in Cdc25A^+/−:Pkhd1^del2/del2 mice, compared to Pkhd1^del2/del2 mice.VK3 altered the cell cycle and reduced proliferation of cultured cholangiocytes by 32%–83% and decreased growth of cultured cysts by 23%–67%. In PCK rats and Pkd2^ws25/− mice, VK3 reduced liver and kidney weights and hepato-renal cystic and fibrotic areas by 18%–34%. PM-20 decreased hepato-renal cystogenesis in Pkd2^ws25/− mice by 15%.

Conclusions

Cdc25A inhibitors block cell cycle progression and proliferation, reduce liver and kidney weights and cyst growth in animal models of PKD and PLD, and might be developed as therapeutics for these diseases.

Introduction

Different genetic mutations in polycystic kidney and liver diseases (PKD/PLD) initiate the formation of hepato-renal cysts that continue to grow progressively.^1-3 Substantial evidence suggests that a switch from a non-proliferative normal epithelia to a proliferative cystic epithelia is associated, in particular, with accelerated cell proliferation (due to elevated cAMP and decreased [Ca²⁺]_i) and cell cycle deregulation (due to overexpression of the Cdc25A phosphatase).^4-10 Indeed, suppression of cAMP and restoration of [Ca²⁺]_i decreases abnormal cell proliferation and inhibits cyst growth.^{4, 5, 8, 9} Moreover, in PKD/PLD patients, treatment with somatostatin analogs halted progression of hepatic and renal cysts and improved quality of life.^11-14

The cell cycle regulator, Cdc25A is of particular interest to us for the following reasons: (i) Cdc25A regulates all phases of the cell cycle;^{15, 16} (ii) Cdc25A is overexpressed in cancers and its suppression reduces cancer cell growth;^{17, 18} and (iii) Cdc25A elevation in PCK rats, an animal model of PKD/PLD, is linked to decreased levels of its regulator, microRNA-15a (miR-15A); experimental increase of miR-15A in PCK cholangiocytes reduced Cdc25A expression inhibiting cyst growth.⁴ Taken together, these observations suggest that Cdc25A might represent a potential therapeutic target in the treatment of PKD/PLD.

Multiple agents including analogues of natural Vitamin K (VK) have been developed as pharmacological inhibitors of Cdc25 phosphatases. VK3 (2-methyl-1, 4-naphtoquinone or menadione), in particular, has anti-proliferative capacity irreversibly inhibiting Cdc25A through covalent modification of its active site.^19-21 Another Cdc25A inhibitor, PM-20 [N-(4-biphenyl)-3, 4-bis-(2-hydroxy-ethylsulphanyl)-maleimide], acts as a phosphatase antagonist binding to Cdc25A active site and subsequently inhibiting cancer cell growth in vitro and in vivo.^{20, 22} While VK3 and PM-20 as anti-cancer agents had been tested in vitro and in vivo and recently in clinical trials,^{19, 23, 24} their role in benign hyperproliferative conditions in general and in PKD/PLD in particular has not been explored.

In the present study, we used genetic and pharmacological approaches to assess further the pathogenic role of Cdc25A in hepato-cystogenesis and to provide evidence supporting its therapeutic potential in PKD/PLD. First, we crossbred Cdc25A^+/−mice (characterized by reduced levels of Cdc25A)²⁵ with our own model of PKD/PLD, Pkhd1^del2/del2 mice (characterized by progressive hepatic but not renal cyst growth and elevated Cdc25A levels).²⁶ We hypothesized that if Cdc25A is involved in hepatic cystogenesis, the crossbred mice would have less severe cystic disease due to decreased Cdc25A levels. Second, we examined the effects of pharmacological targeting of Cdc25A by VK3 on: (i) cell cycle and proliferation in vitro; (ii) cyst expansion in 3-D cultures; (iii) hepato-renal cystic areas and fibrosis in vivo in the PCK rat and Pkd2^ws25/− mice; and (iv) the expression of Cdc25A and the cell cycle proteins known to be regulated by Cdc25A (i.e., Cdk1, 2, 4/6, cyclin A, B, D and E, Rb and PCNA). We also tested the effect of PM-20 on progression of hepato-renal cystic disease in vivo in Pkd2^ws25/− mice. Both the PCK rat and Pkd2^ws25/− mice resemble human pathology in PKD/PLD and have been used in preclinical trials.^{2, 6} The PCK rat (a model of Autosomal Recessive PKD/PLD) develops liver and kidney cysts at birth which gradually expand over time. Pkd2^ws25/− mice (a model of Autosomal Dominant PKD/PLD) have progressive hepato-renal cystic disease with onset at 5-6 month of age. We recently described the phenotype of PCK rats and Pkd2^ws25/− mice in detail.^{27, 28}

We showed that hepatic cystogenesis is inhibited in Cdc25A^+/−:Pkhd1^del2/del2 crossbreds compared to their Pkhd1^del2/del2 counterparts, providing support for an essential role of Cdc25A in cyst growth. VK3 reduced the proliferation of cystic cholangiocytes, caused cell cycle arrest, suppressed cyst growth in 3-D cultures, inhibited progression of hepato-renal cystic disease in vivo in animal models of PKD/PLD, and affected the expression of cell cycle proteins. Targeting of Cdc25A with PM-20 also attenuated hepato-renal cystogenesis in Pkd2^ws25/− mice. These data suggest that Cdc25A represents a potential molecular target in PKD/PLD and its inhibitors might be beneficial in the treatment of these pathological conditions.

Materials and Methods

Animals and cell culture

Animals (number of rats and mice used is indicated for each experimental condition) were maintained on a standard diet after Mayo Institutional Animal Care and Use Committee approval. They were anesthetized with pentobarbital (50 mg/kg). Blood was collected from PCK rats by cardiac puncture. Liver and kidneys were fixed and embedded in paraffin for histology. For in vitro study, cholangiocytes were isolated from normal and PCK rats and cultured as described.⁷ Normal and diseased human liver tissue were obtained from Mayo Clinical Core and National Disease Research Interchange.

Development of Cdc25A^+/−:Pkhd1^del2/del2 mice

The Cdc25A^+/− mice were bred to Pkhd1^del2/del2 mice (both of C57BL6 background) and offspring’s (i.e., Cdc25A^+/−:Pkhd1^+/del2) were backcrossed with Pkhd1^del2/del2 mice to produce Cdc25A^+/−:Pkhd1^del2/del2 animals. Mice were genotyped by PCR. The primers and PCR conditions are described in Supplementary Methods. We analyzed body, liver and kidney weighs and cystic and fibrotic areas in age matched 7-9 month old littermates of Cdc25A^+/− (n=3), Pkhd1^del2/del2 (n=5) and Cdc25A^+/−:Pkhd1^del2/del2 (n=3) mice.

Flow cytometry

Normal (n=5) and PCK cholangiocytes (n=6) were treated with 50, 100 and 200 μM of VK3 for 24h, fixed in ethanol and suspended in 50 μg/ml propidium iodide containing 0,1mg/mL RNase. Cell cycle analysis was performed at Mayo Advanced Genomics Technology Center.

Immunofluorescence confocal microscopy

Microscopy was performed with Zeiss LSM-510 microscope using tissue of normal humans (n=5) and patients with ADPKD, ARPKD and CHF (n=5 for each condition); normal (n=5) and PCK (n=6) rats; and normal (n=4) and Pkd2^ws25/− (n=5) mice. Liver sections were incubated with antibodies against Cdc25A, Cdk4 and 6, cyclin D1 and D3, p-Rb, (Santa Cruz Biotechnology, Santa Cruz, CA; 1:50); Cdk1 and cyclin B (Cell Signaling, Danvers, MA; 1:50); cyclin A and D2 (Abcam, Cambridge, MA; 1:50); Cdk2 (BD Biosciences, Franklin Lakes, NJ; 1:50); cyclin E (Upstate; Lake Placid, NY; 1:50) and PCNA (Santa Cruz, 1:100). Respective secondary antibodies (Invitrogen, Carlsbad, CA; 1:200) were applied. Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay with the ApopTag Peroxidase In Situ Apoptosis Detection Kit (Chemicon, Billerica, MA) was used to detect apoptosis. For mitotic and apoptotic indices (calculated as a percent of PCNA- or TUNEL-positive cells, respectively), 1000 nuclei in randomly selected fields of liver and kidney sections were counted.

Western Blot

Cholangiocytes isolated from livers of normal humans (n=3) and ADPKD patients (n=3); normal (n=5) and PCK rats [non-treated (n=5) or VK3-treated (n=4)]; normal (n=6), Pkd2^ws25/− (n=6), Cdc25A^+/− (n=3), Pkhd1^del2/de/2 (n=6) mice and Cdc25A^+/−:Pkhd1^del2/de/2 crossbreeds (n=3); and cultured normal (n=3) and PCK (n=3) cholangiocytes were used. The details are described in Supplementary Methods.

Cell proliferation

Normal (n=5) and PCK (n=5) cholangiocytes (5000 cells/well) were grown in regular DMEM/F12 medium at 37°C (5% CO2 and 100% humidity) for 48 hours before treatment. VK3 (50, 100 and 200 μM) was added daily for a total of 5 days. Cell proliferation were determined every 24 h by CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega, Madison, WI) as described.⁴

3D culture and cyst measurement

Freshly isolated bile ducts of normal (n=5) and PCK (n=7) rats were kept between 2 layers of collagen (1.5 mg/mL, BD Biosciences) plus 10% Matrigel (BD Biosciences).⁵ VK3 (50, 100 and 200 μM) was added daily. Growth of cystic structures was assessed by light microscopy at days 0 (24 hours after seeding), 1, 3 and 5. The circumferential area of each cyst (n=20 for each condition) was measured using Image J software (National Institutes of Health) as described.²⁹ Data were expressed as a percent change in cystic area compared to day 0.

VK3 and PM-20 treatment protocol

PCK rats (n=22) and Pkd2^ws25/− mice (n=11) received 0.15 g/1 L of VK3 (Sigma, St. Louis, MO) dissolved in 1L of drinking water every other day for 4 and 8 weeks; control rats (n=22) and mice (n=10) received water. Pkd2^ws25/− mice (n=12) received daily intraperitoneal injections of PM-20 (Sigma; 1 mg/kg for 4 weeks); the control group (n=7) was injected with DMSO (Supplementary Figure 4). Doses of VK3 and PM-20 was chosen based on published studies.^{20, 22, 30, 31} We assessed: body weight, serum biochemistry, liver and kidney weights, hepatic and renal cystic areas and fibrosis. Since no differences was observed in liver and kidney weights between male and female PCK rats and Pkd2^ws25/− mice, for statistical analysis male and female data were combined. Cystic and fibrotic areas were analyzed as described.⁵ Briefly, livers and kidney sections were stained with H&E or picrosirius red collagen to assess, respectively, cystic or fibrotic areas. Measurements were done by Meta-Morph software (Universal Imaging, West Chester, PA) installed on a Pentium IBM-compatible computer (Dell OptiPlex, Dell Ins, Round Rock, TX) following image acquisition using a light microscope and color digital camera (Nikon DXM 1200). Hepatic and renal cystic and fibrotic areas were expressed as a percent of total hepatic or renal parenchyma, respectively.

3-dimentional reconstruction of the intrahepatic biliary tree

The intrahepatic biliary tree of non-treated (n=4) and VK3-treated for 8 weeks (n=4) PCK rats were reconstructed in three-dimensions. Briefly, anesthetized rats were injected via the common bile duct with a silicone polymer MV-122 (Flow Tech, Carver, MA). Liver lobes were fixed, suspended within a plastic cylinder and scanned using a micro-computed tomography (micro-CT) scanner with 20 μm resolution. Quantitation was performed by Analyze (Biomedical Imaging Resource; Mayo Clinic) as described.²⁷

Results

Cell cycle phase distribution is altered in cystic cholangiocytes and rate of cell proliferation is increased

We assessed the cell cycle distribution in normal and PCK rats. In cystic cholangiocytes, the percent of cells in G0/G1 phase was reduced, in S phase was not altered, and in G2/M phase was increased compared to normal cholangiocytes (Supplementary Figure 1). We also examined the rate of proliferation in humans and rodents by PCNA expression. Most of cystic cholangiocytes in patients with PKD/PLD, in PCK rats and Pkd2^ws25/− mice showed PCNA immunoreactivity while in normal cholangiocytes PCNA-positive nuclei were practically undetectable (Supplementary Figure 2A). Western blots further confirmed that PCNA levels are increased in vivo in cholangiocytes isolated from ADPKD patients, PCK rats and Pkd2^ws25/− mice and in vitro in PCK-derived cholangiocytes compared to respective controls (Supplementary Figure 2B).

Cdc25A is overexpressed in cystic cholangiocytes

Here we confirm and extended our previous finding⁴ showing the increase in Cdc25A immunoreactivity in cystic cholangiocytes of patients with PKD/PLD, in PCK rats and Pkd2^ws25/− mice compared to normal respective conditions (Figure 1A). In agreement with immunofluorescent data, western blots demonstrate that Cdc25A is increased in vivo in cholangiocytes isolated from ADPKD patients, PCK rats and Pkd2^ws25/− mice and in vitro in cultured cholangiocytes derived from PCK rats compared to respective controls (Figure 1B). Interestingly, Cdc25A expression in normal humans and mice appears to be stronger compared to rats which might be a reflection of species differences.

Figure 1.

Cdc25A is overexpressed in cystic cholangiocytes. (A) Relatively low expression of Cdc25A (red) is found in normal human (n=5), rat (n=5) and mouse (n=4) cholangiocytes by confocal microscopy. Cdc25A appears to be increased in cystic cholangiocytes of patients with ARPKD, ADPKD, CHF, in PCK rats and Pkd2^ws25/− mice (n=5 for each condition). Magnification, x40. White boxes show higher power of bile duct depicted by an arrow. (B) Representative western blots (n=3 for each set of data) demonstrate that Cdc25A levels are elevated in vivo in human patients with ADPKD, in PCK rats and Pkd2^ws25/− mice and in vitro in cultured PCK cholangiocytes compared to normal controls, respectively.

Cdc25A plays an important role in hepatic cystogenesis

To assess further the potential pathogenic role of Cdc25A in hepatic cyst growth and provide evidence supporting its role as a therapeutic target, we hypothesized that crossbreeding of Cdc25A^+/− mice with our own Pkhd1^del2/del2 mice might decrease hepatic cystogenesis. Cdc25A^+/− mice have reduced Cdc25A protein by ~50% compared to wild type resulting in impaired proliferative capacity.²⁵ Pkhd1^del2/del2 animals develop hepatic cysts by the age of 7-8 months while no renal cysts are present²⁶ and have increased levels of Cdc25A (Supplementary Figure 3). The scheme describing development of Cdc25A^+/− :Pkhd1^del2/del2 crossbred mice is depicted in Supplementary Figure 3.

Body and kidney weights were not different in age matched 7-9 month old Cdc25A^+/−, Pkhd1^del2/del2 and Cdc25A^+/−:Pkhd1^del2/del2 crossbreds (Supplementary Table 1). In contrast, liver weights were higher by 62% in Pkhd1^del2/del2 mice compared to Cdc25A^+/− mice. A 33% reduction in liver weight was observed in Cdc25A^+/−:Pkhd1^del2/del2 mice compared to Pkhd1^del2/del2 animals (Figure 2A and D, Supplementary Table 1). Multiple cysts occupied a substantial portion of liver in Pkhd1^del2/del2 mice; no cysts were present in Cdc25A^+/− mice; and hepatic cystogenesis were decreased in Cdc25A^+/−:Pkhd1^del2/del2 crossbreds (Figure 2A-C). Quantitatively, hepatic cysts represented 40% of total liver parenchyma in Pkhd1^del2/del2, 0% in Cdc25A^+/− and 18% in Cdc25A^+/−:Pkhd1^del2/del2 mice (Figure 2E). Hepatic fibrosis was also reduced in Cdc25A^+/−:Pkhd1^del2/del2 crossbreds compared to Pkhd1^del2/del2 mice (Figure 2F).

Figure 2.

Cdc25A is involved in hepatic cystogenesis. (A) Representative photographs of livers from Cdc25A^+/− (n=3), Pkhd1^del2/del2 (n=5) and Cdc25A^+/− :Pkhd1^del2/del2 (n=3) mice. Bar, 10 mm. (B) Representative images (H&E staining) demonstrate that no cysts are present in Cdc25A^+/− mice, many hepatic cysts observed in Pkhd1^del2/del2 mice, and Cdc25A^+/−:Pkhd1^del2/del2 crossbreeds have reduced hepatic cystogenesis. Bar, 500 mm. Arrow indicates the liver area shown with the higher magnification in panel C. (C) Representative images (Picrosirius Red staining) of liver areas depicted by arrows in panel B. Magnification, x4. (D) Liver weight, hepatic cystic (E) and fibrotic (F) areas are decreased in Cdc25A^+/−:Pkhd1^del2/del2 mice compared to Pkhd1^del2/del2 counterparts. C - Cdc25A^+/−, P - Pkhd1^del2/del2 and C:P - Cdc25A^+/−:Pkhd1^del2/del2. * -P <.05.

Targeting of Cdc25A by VK3 in vitro suppresses cholangiocyte proliferation and alters cell cycle phase distribution

Non-treated PCK cholangiocytes proliferate at a much higher rate (Figure 3A) then normal ones (Figure 3B). In response to all VK3 doses (50, 100 and 200 μM), proliferation of both PCK (Figure 3A) and normal cholangiocytes (Figure 3B) was inhibited; however, VK3 decreased the proliferation of cystic cholangiocytes to a higher extent than in normal cholangiocytes (Figure 3C).

Figure 3.

VK3 decreases cholangiocyte proliferation and alters the cell cycle distribution. VK3 inhibits proliferation in both (A) PCK (n=5) and (B) normal (n=5) cholangiocytes. (C) Changes in proliferation of PCK cholangiocytes at day 5 after treatment (compared to non-treated condition) were greater than in normal cholangiocytes. (D) VK3 decreases a percent of PCK cholangiocytes (n=3) during S phase and leads to cell accumulation in G2/M phase. (E) VK3 did not affect the cell cycle in normal cholangiocytes (n=3). * - P<.05.

Non-treated PCK cholangiocytes were distributed during the cell cycle as followed: 53.04% (G0/G1 phase), 26.88% (S phase), and 20.08% (G2/M phase). VK3 in all doses caused cell cycle arrest of PCK cholangiocytes in G2/M phase, did not affect G0/G1 phase and decreased the cell proportion in S phase (Figure 3D). In contrast, VK3 did not change the distribution of normal cholangiocytes (Figure 3E).

VK3 inhibits growth of cystic structures in 3-D culture

PCK cystic structures grown in 3-dimentions expanded more rapidly then normal ones (Figure 4). VK3 suppressed PCK cyst growth as shown by light microscopy and quantitative assessment of their circumferential areas compared to non-treated counterparts (Figure 4A). VK3 also inhibited growth of normal cysts (Figure 4B). However, VK3 suppressed expansion of normal cystic structures to a lesser degree. By day 5, in response to 50, 100 and 200 μM of VK3, the growth of PCK cysts was decreased 1.9, 2.8 and 3.7-fold, respectively; under the same conditions development of normal cystic structures was inhibited 1.2, 1.5 and 1.7-fold.

Figure 4.

VK3 suppresses growth of hepatic cysts in 3-D cultures. Representative images of cystic structures derived from (A) PCK (n=7) and (B) normal rats (n=5). They were treated with VK3 (50 μM) daily for total of 5 days. Magnification, x4 (A) and x10 (B). Graphs show the effects of different doses of VK3 on cyst enlargement assessed by changes in their circumferential areas. * - P<.05.

VK3 decreases total volume of the intrahepatic biliary tree and hepato-renal cystic and fibrotic areas in vivo

We next tested the effects of Cdc25A targeting by two inhibitors, VK3 and PM-20, in vivo. The detailed treatment protocol is described in Supplementary Figure 4.

No visible defects (such as weight loss) were observed in VK3 and PM-20 treated animals. Serum markers of the PCK rats (which are within the range of normal rats) also showed no differences in treated and non-treated animals (Supplementary Table 2).

We first analyzed the VK3 effects on hepatic cystogenesis in PCK rats by assessing the total volume of the intrahepatic biliary tree. Representative images demonstrate that VK3 improved biliary tree architecture reducing its total volume by 26% (Figure 5A).

Figure 5.

VK3 attenuates hepato-renal cystic disease in PCK rats. (A) Representative images and bar graph demonstrate that VK3 significantly reduced the biliary tree volume in VK3 treated (n=4) for 8 weeks PCK rats compared to non-treated rats (n=4). Bar, 5 mm. (B) Representative images of picrosirius red-stained liver and (C) kidney sections and bar graphs demonstrate that VK3 decreases hepatic and renal cystic and fibrotic areas of PCK rats compared to non-treated counterparts. Magnification, x4. Bar, 500 μm. NT – non-treated. * - P<.05.

VK3 treatment reduced liver and kidney weights up to 18%, hepatic and renal cyst areas up to 36%, hepatic and renal fibrosis up to 31%, and the mitotic and apoptotic indices up to 34% in both PCK rats (Supplementary Table 3 and Figure 5B and C) and Pkd2^ws25/− mice (Supplementary Table 4 and Figure 6A-B). Moreover, longer treatment (i.e., for 8 weeks) led to more significant inhibition of all parameters studied compared to respective controls. PM-20 had similar effects on hepato-renal cystogenesis in Pkd2^ws25/− mice: liver and kidney weights, hepatic and renal cyst areas, hepatic and renal fibrosis, and mitotic and apoptotic indices were decreased by 10-15% compared to non-treated mice (Supplementary Table 5 and Figure 6C-E).

Figure 6.

VK3 and PM-20 decreases hepato-renal cystic and fibrotic areas in Pkd2^ws25/− mice. Representative images and bar graphs demonstrate that hepatic and renal cystic and fibrotic areas were decreased in VK3-treated (A and B) and PM-20 (C-E) mice compared to non-treated animals. Liver and kidney were stained with picrosirius red. Bar, 2500 μm. * - P<.05.

VK3 affects the expression of cell cycle machinery

Cdc25 phosphatases are recognized as only dephosphorylation molecules of Cdks which in turn are important regulators of the cell cycle transitions.³² Cdk/Cyclin complexes play an important role during cell cycle: G1 phase is regulated by Cdk4/Cyclin D and Cdk6/Cyclin D; G1/S phase - by Cdk2/Cyclin E; S phase - by Cdk2/Cyclin A; G2 phase – by Cdk1/Cyclin A, and M phase - by Cdk1/CyclinB (Supplementary Figure 5).³³ Thus, we examined the expression of Cdc25A and its downstream targets in response to VK3 (Figure 7). VK3 inhibited the expression of Cdc25A and the cell cycle proteins known to be regulated by Cdc25A (i.e., Cdk1, 2, 4/6, cyclin A, B, D and E, Rb and PCNA), increased the phosphorylated status of Cdk1 and Cdk2 and decreased Erk1/2 phosphorylation.

Figure 7.

VK3 affects the expression of the cell cycle proteins. (A) Confocal microscopy and western blots (B) demonstrates that VK3 treatment changes the expression of the cell cycle proteins stained in green. Magnifications: x100.

Discussion

We described here that: (a) cystic cholangiocytes are hyper-proliferative and have an abnormal cell cycle; (b) Cdc25A is overexpressed in cystic cholangiocytes of humans and rodents with PKD/PLD; (c) in Cdc25A^+/− :Pkhd1^del2/del2 crossbred mice, liver weights and cystic and fibrotic areas are decreased compared to Pkhd1^del2/del2 counterparts supporting the importance of Cdc25A in hepatic cystogenesis; (d) VK3 decreases proliferation and causes cell cycle arrest of PCK cholangiocytes; (e) VK3 reduces cyst growth in 3-D cultures; (f) the intrahepatic biliary tree volume is significantly less after VK3; (g) in vivo in PCK rats and Pkd2^ws25/− mice, liver and kidney weights, hepato-renal cystic and fibrotic areas, and mitotic and apoptotic indices are all decreased after VK3 treatment; (h) PM-20 attenuates hepato-renal cystogenesis in Pkd2^ws25/− mice; and (i) VK3 affects the expression of multiple cell cycle proteins. These findings provide new insights into the pathogenesis of PKD/PLD and present a rationale for the use of Cdc25A inhibitors to halt cyst progression.

Cdc25A over-expression in cystic cholangiocytes is likely linked to disturbances in several signaling pathways. First, elevated Cdc25A is associated with reduced expression of its regulator, miR-15a.⁴ Second, Cdk2 (i.e., a down-stream effector of Cdc25A) phosphorylates Cdc25A in a positive feedback loop increasing its activity.³⁴ We observed (unpublished data) that Ckd2 is up-regulated in cystic cholangiocytes; therefore its overexpression may add to Cdc25A elevation. Importantly, Cdk2 is also a miR-15a target. Third, the levels of many RTK’ (such as EGFR, VGFR, IGFR, HGFR) and several components of the RTK pathway (i.e., Raf, MEK and ERK) are increased in hepato-renal cysts.^{2, 7} Raf kinase forms a functional complex with Cdc25A while ERK is known to directly phosphorylate Cdk2 further suggesting a connection between mitogenic signaling and the cell cycle machinery.³⁵ Additional studies are underway in our laboratory to directly assess all these possibilities. Be that as it may, accumulating data suggest a major role for Cdc25A in hepatic cystogenesis.

Based on our published evidence that hepatic cystogenesis is linked to elevated Cdc25A in cholangiocytes,⁴ we hypothesized that its reduction might decrease cyst growth. To this end, we cross-breed the Pkhd1^del2/del2 mice (they develop hepatic cysts and have increased Cdc25A expression) with Cdc25A^+/− mice that have reduced Cdc25A levels. We anticipated that in crossbred (i.e., Pkhd1^del2/del2:Cdc25A^+/− mice), cystogenesis should be less due to Cdc25A diminution. As expected, in Cdc25A^+/−:Pkhd1^del2/del2 mice, decreased Cdc25A expression was associated with reduced liver weights and less evident cystogenesis. The mechanism/s by which Cdc25A depletion affects hepatic cystogenesis is not yet understood. However, our indirect in vitro and in vivo data suggest that cell cycle progression and cell proliferation might be involved since suppression of Cdc25A by VK3 affected the distribution of the cell cycle phases, inhibited rate of cell proliferation and cyst growth.

Various pharmacological inhibitors of Cdc25 phosphatases have been developed.^{20, 21, 36, 37} Studies testing the effects of VK family members on cancer cell growth have shown that VK3 has the strongest antiproliferative capacity directly affecting Cdc25A.³⁸ Mechanisms of VK3 action might relate to: (i) an oxidative effect produced by redox cycling of VK3; (ii) modulation of transcriptional factors; (iii) perturbation of cell proliferation; and (iv) cell cycle arrest with consequent inhibition of proliferation due to direct binding of inhibitor to Cdc25A.^{21, 23, 38-42} Consistent with these possibilities, we observed inhibited proliferation of PCK cholangiocytes which was coupled with cell cycle arrest after treatment. In contrast, VK3 has no effect on the cell cycle in normal cholangiocytes suggesting that suppression of their proliferation might involve other mechanisms including those described above. VK3 also slows down the growth of cystic structures in 3-D cultures. Finally, in vivo in two animal models of PKD/PLD, VK3 halted hepato-renal cystogenesis by decreasing mitotic indices and reducing PCNA expression. Another inhibitor used in this study, PM-20, also attenuated hepato-renal cystogenesis in vivo in Pkd2^ws25/− mice. PM-20 is known to bind and inhibit the Cdc25A activity subsequently affecting cell proliferation and tumor growth.^{20, 22} Indeed, mitotic indices were decreased in PM-20 treated mice.

To begin to elucidate the mechanisms by which VK3 affects hepatic cyst growth, we analyzed the expression of the cell cycle proteins. We found that treatment of PCK rats with VK3 decreased levels of Cdc25A and proteins known to be regulated by Cdc25A; reduced the phosphorylated status of Rb, and increased the phosphorylated status of Cdk1 and Cdk2. These observations are in agreement with previous reports.^{36, 41, 42} Increased phosphorylation of Cdks suggest that VK3 might affect cell proliferation and cell cycle via direct inactivation of Cdc25A which is usually correlated with protein levels.^{35, 43, 44} Consistent with this view, treatment of PCK rats with VK3 for 8 weeks decreased Cdc25A expression. Finally, we evaluated the phosphorylated status of Erk1/2. We have recently shown the significance of Erk1/2 signaling in hepatic cystogenesis.⁷ In particular, expansion of PCK-derived hepatic cysts in 3-D cultures as well as enhanced proliferation of cystic cholangiocytes was Mek/Erk1/2-dependent.⁷ In the present study, VK3 affected Mek/Erk1/2 pathway by decreasing the level of Erk1/2 activation. The mechanism of this inhibition is currently unclear; however, suppressed Erk1/2 phosphorylation was observed in vivo in jck mice (the animal model of PKD) after 5 weeks of treatment with Cdk inhibitor, roscovitine.⁴⁵

Taken together, our study provides the first experimental evidence that Cdc25A targeting with VK3 and PM-20 decreases proliferation of cystic cholangiocytes halting the expansion of hepatic cysts. Moreover, VK3 and PM-20 also has positive effects on renal cytogenesis.

In conclusion, we believe that Cdc25A represents a potential therapeutic target and this study provides a rationale for examining the role of VK3, PM-20 or other Cdc25 inhibitors that are now at developmental/testing stage³⁷ in the treatment of PKD/PLD.

Abbreviations

PKD

polycystic kidney disease

PLD

polycystic liver disease

ADPKD

autosomal dominant PKD

ARPKD

autosomal recessive PKD

CHF

congenital hepatic fibrosis

Pkhd1

polycystic kidney and hepatic disease 1

Pkd2

polycystic kidney 2

PCK

polycystic kidney

Cdc25

cell division cycle 25

VK3

Vitamin K3

cAMP

cyclic adenosine monophosphate

miR-15A

microRNA 15A

C57BL6

c57 black 6

Cdk

cyclin dependent kinase

p-Rb

retinoblastoma protein

PCNA

proliferating cell nuclear antigen

TUNEL

terminal deoxynucleotidyl transferase dUTP nick-end labeling

DMEM/F12

Dulbecco Modified Eagle Medium: Nutrient Mixture F-12

H&E

hematoxylin and eosin

MV-122

low-viscosity radiopaque polymer Microfil MV-122

micro-CT

micro-computed tomography

G0/G1 phase

gap 0/1

S phase

synthetic phase

G2/M phase

gap2/mitosis

3-D

three dimensional

RTK

receptor tyrosine kinases

EGFR

epidermal growth factor receptor

VGFR

vascular endothelial growth factor receptor

IGFR

insulin-like growth factor receptor

HGFR

hepatocyte growth factor receptor

ERK

extracellular signal-regulated protein kinase