Conditional Mesenchymal Disruption of Pkd1 Results in Osteopenia and Polycystic Kidney Disease

Conditional deletion of Pkd1 in osteoblasts using either Osteocalcin(Oc)-Cre or Dmp1-Cre results in defective osteoblast-mediated postnatal bone formation and osteopenia. Pkd1 is also expressed in undifferentiated mesenchyme that gives rise to the osteoblast lineage. To examine the effects of Pkd1 on prenatal osteoblast development, we crossed Pkd1 flox/flox and Col1a1(3.6)-Cre mice, which has been used to achieve selective inactivation of Pkd1 earlier in the osteoblast lineage. Control Pkd1 flox/flox and Pkd1 flox/+, heterozygous Col1a1(3.6)-Cre;Pkd1 flox/+ and Pkd1 flox/null, and homozygous Col1a1(3.6)-Cre;Pkd1 flox/flox and Col1a1(3.6)-Cre;Pkd1 flox/null mice were analyzed at ages ranging from E14.5 to 8-weeks-old. Newborn Col1a1(3.6)-Cre;Pkd1 flox/null mice exhibited defective skeletogenesis in association with a greater reduction in Pkd1 expression in bone. Conditional Col1a1(3.6)-Cre;Pkd1 flox/+ and Col1a1(3.6)-Cre;Pkd1 flox/flox mice displayed a gene dose-dependent decrease in bone formation and increase in marrow fat at 6 weeks of age. Bone marrow stromal cell and primary osteoblast cultures from homozygous Col1a1(3.6)-Cre;Pkd1 flox/flox mice showed increased proliferation, impaired osteoblast development and enhanced adipogenesis ex vivo. Unexpectedly, we found evidence for Col1a1(3.6)-Cre mediated deletion of Pkd1 in extraskeletal tissues in Col1a1(3.6)-Cre;Pkd1 flox/flox mice. Deletion of Pkd1 in mesenchymal precursors resulted in pancreatic and renal, but not hepatic, cyst formation. The non-lethality of Col1a1(3.6)-Cre;Pkd1 flox/flox mice establishes a new model to study abnormalities in bone development and cyst formation in pancreas and kidney caused by Pkd1 gene inactivation.


Introduction
Polycystin-1 (PC1), the PKD1 gene product, is a highly conserved, multi-domain membrane protein widely expressed in various cell types and tissues [1,2]. The specific biological functions of PC1 (PKD1) in different tissues are currently being elucidated; best understood are the kidney functions. In renal epithelium, PKD1 forms a complex with the calcium channel, PKD2 that co-localizes to primary cilia and functions as a flow sensor regulating cell proliferation and polarity. Loss of PKD1 or PKD2 function in renal tubular epithelial cells causes Autosomal Dominant Polycystic Kidney Disease (ADPKD) [3,4] that is the result of abnormal cell proliferation and cell polarity and leads to cystic kidney disease. PKD1 mutations also lead to cystic disease of liver and pancreas in some patients with ADPKD, also because of abnormal proliferation of ductal epithelial cells in these tissues [5,6,7]. PKD1 may also have a role in the development and function in endothelial-and mesenchymal-derived cells. PKD1 mutations lead to vascular (intracranial and aortic aneurysms) [8,9,10,11] and lung (bronchiectasis) abnormalities [12]. PKD1 and PKD2 are required for placental development [13]. Pkd1 mutations in mouse models also cause abnormalities of the skeleton [14,15,16,17,18] and human subjects with polycystic kidney disease appear to have earlier elevation of the bone-derived hormone FGF23 [19].
It has been difficult to determine the specific extra-renal functions of Pkd1/PKD1using mutations of Pkd1/PKD1 in mice and humans. Global ablation of Pkd1 in mice leads to a complex, embryonically lethal, phenotype [7]. Multiple abnormalities, including renal and pancreatic cysts and pulmonary hyperplasia, are observed in Pkd1 2/2 mice when they survive to embryonic day 15.5 post coitum (E15.5). In global knockout mice it is difficult to differentiate between indirect extra-renal abnormalities due to the effects of the complex metabolic alterations caused by renal cystic disease from direct effects caused by loss of Pkd1 functions in affected tissues. In addition, in humans, ADPKD is a heterozygous state, whereby mutations leading to loss of one PKD1 or PKD2 allele is combined with somatic mutations in the kidney (i.e., a second hit) to cause renal cystic disease [20,21,22,23,24]. The resulting residual function of the non-mutated PKD1 or PKD2 allele in extra-renal tissues may also mask discovery of PKD1 or PKD2 functions in non-renal tissues.
The Pkd1 flox/flox mouse model has been used to define the tissue selective function of Pkd1 in vivo [25]. A low frequency of renal Pkd1 gene inactivation and only a few renal cysts and more frequent hepatic cysts is reported from the conditional deletion of Pkd1 in MMTV-Cre mice [25], whereas the broadly expressed tamoxifen-Cre inducible inactivation of the Pkd1 gene in mice resulted in massive cystic transformation of renal tissue [26]. The selective deletion of Pkd1 in kidney by using Ksp-Cre, or more broadly Nestin-Cre, also leads to the formation of polycystic kidneys resembling human ADPKD [27,28].
In addition, use of Pkd1 flox/flox mice and bone-specific Cre mice has defined previously unrecognized functions of polycystin-1 in bone. In this regard, the selective deletion of Pkd1 in osteoblasts by using Osteocalcin(Oc)-Cre and in osteocytes by using Dmp1-Cre results in osteopenia in adult mice because of defects in osteoblastmediated bone formation [17,18]. At present, however, it is unclear whether the functions of Pkd1 are limited to mature osteoblasts and osteocytes or involve earlier stages in osteoblast development.
To explore the effects of Pkd1 on early pre-osteoblast stage and prenatal bone development, we used the Col1a1(3.6) promoter to drive Cre-recombinase expression (Col1a1(3.6)-Cre) in mesenchymal precursors. Col1a1(3.6) promoter driven Cre expression begins at E10 and peaks between E12.5 and E14.5 in developing skeletal elements [29,30]. Col1a1(3.6)-Cre is thought to be specific for the osteoblast lineage and it has been extensively used to conditionally delete genes early in osteoblastic development and to study bonespecific function of many genes [29,30,31,32].
We found that conditional deletion of Pkd1 from osteoblasts precursors within the mesenchymal lineage resulted in defective bone formation that was associated with abnormal osteoblastic development and enhanced adipogenesis. Unexpectedly, we found that Col1a1(3.6)-Cre was not bone specific and resulted in deletion of Pkd1 in multiple tissues, leading to cyst formation in the kidney and pancreas, but not the liver of adult mice.
A gene dose-dependent effect of Col1a1 (3.6)-Cremediated conditional deletion of Pkd1 in newborn and postnatal bone formation Indeed, we failed to observe abnormalities of skeletal development in homozygous Pkd1 Col1a1(3.6)-cKO newborn mice ( Fig. 2A, 2C, 2E and 2G). In contrast, bone structural abnormalities were observed in homozygous Col1a1(3.6)-Cre;Pkd1flox/null newborn mice, including delayed bone mineralization in calvarial and vertebral bone tissues (Fig. 2B), a short and less mineralized femur (Fig. 2D), and a significant reduction in both trabecular bone volume (Fig. 2F) and cortical bone thickness (Fig. 2H). Because the Cre;Pkd1 flox/null strategy leads to greater Cre-mediated reduction in Pkd1 conditional deletion compared to Cre;Pkd1 flox/flox approach [33], the more severe phenotype that we observed is likely the result of greater reductions in Pkd1 gene dose during embryogenesis.
In accordance with decreased osteogenesis in bone, we also observed an increased adipogenesis in bone marrow and in bone marrow stromal cultures from homozygous Pkd1 Col1a1(3.6)-cKO mice. In this regard, homozygous Pkd1 Col1a1(3.6)-cKO mice showed a higher number of adipocytes and volume of fat droplets in decalcified tibias stained with Oil Red O and OsO4 (Fig. 4A). BMSC cultures derived from Pkd1 Col1a1(3.6)-cKO mice exhibited a marked increase of Oil Red O stained adipocytes (Fig. 4B). In addition, PPARc, an adipocyte transcription factor, and adipocyte markers such as aP2 (adipocyte fatty acid-binding protein 2) were also significantly increased in BMSC cultures of Pkd1 Col1a1(3.6)-cKO mice compared to Pkd1 flox/flox control mice ( Fig. 4C and 4D), consistent with increased adipogenic markers including adiponectin, aP2, and Lpl (lipoprotein lipase) in long bone samples of Pkd1 Col1a1(3.6)-cKO mice ( Table 1).
Timed pregnancies were generated to analyze heterozygous Col1a1(3.6)-Cre;Pkd1 flox/+ and homozygous Pkd1 Col1a1(3.6)-cKO fetuses at various developmental stages. We did not observe renal or pancreatic cysts in heterozygous Col1a1(3.6)-Cre;Pkd1 flox/+ mice during embryogenesis or at 6-weeks of age (data not shown). Pancreatic cysts first became evident in homozygous Pkd1 Co-l1a1(3.6)-cKO embryos was at E15.5 ( Fig. 7A and 7B). In addition, the dilatation of pancreatic cysts progressed as a function of age in homozygous Pkd1 Col1a1(3.6)-cKO ( Fig. 7A and 7B). The size of renal cystic lesions also increased as a function of age. Masson-trichrome stained sections of the kidney showed that the cystic fibrosis of polycystic kidney started at P7 and became progressively more severe by 6-weeks of age ( Fig. 8A and 8B). Real-time RT-PCR showed that mesenchymal-to-epithelial transition and fibrosis formation genes were significantly up-regulated in cystic kidney from homozygous Pkd1 Col1a1(3.6)-cKO mice compare with control Pkd1 flox/flox mice ( Fig. 9A-9F), but no such alterations were observed in heterozygous Col1a1(3.6)-Cre;Pkd1 flox/+ mice ( Fig. 9A-9F). In addition, homozygous Pkd1 Col1a1(3.6)-cKO mice had significantly higher levels of serum BUN, PTH, and FGF23, but lower levels of phosphorus, and no changes in calcium levels at 6 weeks of age (Table 2), consistent with the development of renal impairment caused by polycystic kidney disease and secondary induction of compensatory hormonal changes ( Table 2).

Discussion
Pkd1, Pkd2, and primary cilia are present in mature osteoblasts and osteocytes [16], where the primary cilia-polycystin complex plays an important role in postnatal osteoblast and osteocyte regulation of bone formation and mechanosensing [17,18]. Primary cilia-polycystin complexes are known to have a role in regulating developmental pathways in other tissues, such as leftright patterning in embryos and kidney development [35], but the role of Pkd1 in osteoblast development and bone embryogenesis, while suggested by skeletal alterations in global Pkd1 knockout mice [16,17,18,36], has not been confirmed by selective ablation of Pkd1 early in the osteoblast lineage. In the current study, Pkd1 was conditionally deleted in mesenchymal precursors that are destined to become multiple cell types including osteoblast lineage [37], renal tubule cells [38], pancreas duct epithelial cells [39], and bile ductal plate cells [40] by creating Pkd1 Col1a1(3.6)-cKO or Col1a1(3.6)-Cre;Pkd1 flox/null mice, which differ in the degree of Pkd1 deletion [33].
We found that reduction of Pkd1 in mesenchymal precursors in Pkd1 Col1a1(3.6)-cKO mice resulted in impaired osteoblast-mediated bone formation and low bone mass in the adult mouse, consistent with the observations in Oc-Creand Dmp1-Cre-mediated conditional knockout of Pkd1 [17,18]. Primary osteoblast cultures showed higher cell replication and lower osteoblastic differentiation markers in E17.5 homozygous Pkd1 Col1a1(3.6)-cKO mice, similar to the defects of osteoblastic maturation in Oc-Cre-mediated conditional knockout of Pkd1 [17]. The increased proliferation observed in Pkd1-deficient osteoblasts is similar to increased renal effect was observed during skeletogenesis between Col1a1(3.6)-Cre;Pkd1 flox/flox and Col1a(3.6)-Cre; Pkd1 flox/null newborn mice. Data represent the mean 6 SD from three to four individual samples. *Significant difference from control mice (Pkd1 flox/flox or Pkd1 flox/+ ) at P,0.05. doi:10.1371/journal.pone.0046038.g002 cell proliferation caused by loss of Pkd1 in renal epithelial cells, suggesting that an important function of Pkd1 is to regulate cell proliferation, which is typically inversely correlated with differentiation [41,42]. In addition, the marrow fat content of bone was increased and enhanced adipogenesis was found in bone marrow stromal cell cultures, which expressed decreased levels of the osteoblast differentiation factor Runx2 and increased levels of PPARc (peroxisome-proliferator-activated receptor-c), which regulates adipocyte development. These findings are consistent with previous studies showing that Pkd1 stimulates osteogenesis and inhibits adipogenesis through a Pkd2-calcium dependent Runx2 expression and that compound heterozygous Pkd1 and Runx2 deficient mice have additive effects to induce osteopenia [17,18,36]. Further studies are needed to determine the signaling pathways linking Pkd1 inactivation in pre-osteoblasts to increments in PPARc expression.
We were surprised by the difference in skeletal abnormalities between newborn Col1a1(3.6)-Cre;Pkd1 flox/null and Pkd1 Co-l1a1(3.6)-cKO mice. Col1a1(3.6)-Cre;Pkd1 flox/null had evidence of abnormal skeletogenesis, albeit less severe than the previously described global Pkd1 2/2 mouse models [14,15,16]. Col1a1(3.6)-Cre, which is highly active between E12.5 and E14.5, and the resulting bone abnormalities are consistent with known effects of Pkd1to regulate Runx2, an essential transcriptional factor controlling osteoblast development [43,44]. In contrast, newborn Pkd1 Col1a1(3.6)-cKO mice lacked a demonstrable bone phenotype, consistent with studies in Osx-Cre;Pkd1 flox/flox mice, which also failed to find evidence for a role of Pkd1 in skeletogenesis [45]. A gene-dose dependent effect on skeletogenesis and differences in Table 1. Gene-expression profiles in 6-week-old mice.  methods of conditional gene targeting might explain these discrepancies. In this regard, the less severe skeletal phenotype in newborn Pkd1 Col1a1(3.6)-cKO and Osx-Cre;Pkd1 flox/flox mice compared to Col1a(3.6)-Cre;Pkd1 flox/null mice may be due to insufficient reductions in Pkd1 expression to cause abnormal osteoblast development [33]. Additional studies are needed using the heterozygous floxed allele paired with a ''null'' mutant allele along with earlier osteoblast lineage specific and less leaky promoters to define the function of Pkd1 in osteoblast development during different stages of embryogenesis [17,33,46]. The role of Pkd1 in post-natal bone remodeling is further documented by the current studies. Bone remodeling that occurs postnatally is characterized by the recruitment of bone marrow mesenchymal stem cells to differentiate into osteoblasts that refill resorptive cavities. Our studies suggest that a reduction of Pkd1 expression of more than 50% results in abnormalities in osteoblastmediated bone formation in adult mice through Pkd1 regulation of critical transcription factors involved in osteoblastogenesis and adipogenesis [47,48,49]. We have previously shown that the bonespecific deletion of Pkd1 using Oc-Cre or Dmp1-Cre had a direct role in adult bone formation [17,18]. However, we observed a more severe osteopenia in adult Pkd1 Col1a1(3.6)-cKO compared to Oc-Cre or Dmp1-Cre-mediated Pkd1 deletion, which could result from the effects of Col1a1(3.6)-Cre to delete Pkd1 during embryogenesis or to the leakiness of this promoter leading to alterations in systemic factors that lead to secondary effects on bone. Although we observed significant increases of serum PTH and FGF23 levels in association with elevated BUN in adult Pkd1 Col1a1(3.6)-cKO mice, the expected PTH-induced increases in bone transcripts were not observed, rather we found that Akp2, FGF23, Trap, Osterix and Runx2 transcripts were decreased in Pkd1 Col1a1(3.6)-cKO , suggesting a direct effect of Pkd1 on bone. In addition, our finding of abnormal skeletogenesis in newborn Col1a1(3.6)-Cre;Pkd1 flox/null mice and impairment in osteoblast differentiation maturation in E17.5 immortalized osteoblast cultures from Pkd1 Col1a1(3.6)-cKO mice, are consistent with direct effects of Pkd1 in osteoblasts, as previously reported [16].
Interestingly, in this model, serum phosphate was low and FGF23 was high, consistent with other reports that FGF23 regulation may be abnormal in ADPKD [19]. However, we failed to find evidence for increased FGF23 mRNA expression in bone or in osteoblasts cultures derived from Pkd1 Col1a1(3.6)-cKO mice; rather FGF23 mRNA levels were decreased in bone and osteoblasts Osteoblasts from control, heterozygous, and homozygous mice displayed timedependent increments in ALP activity during 21 days of culture, but ALP activity was gene dose-and time-dependently decreased in heterozygous and homozygous osteoblasts compared to control osteoblasts. (D) Quantification of mineralization. Alizarin Red-S was extracted with 10% cetylpyridinium chloride and quantified as described in Materials and Methods. A time-dependent increment of Alizarin Red-S accumulation was observed in control, heterozygous, and homozygous osteoblasts during 21 days of culture, but the accumulation was gene dose-dependently decreased in heterozygous and homozygous osteoblasts cultures compared to control osteoblasts at day 21 of culture. (E-I) Gene expression profiles by real-time RT-PCR. Osteoblastic markers such as Runx2, Akp2, and FGF23 were gene dose-dependently reduced during 18 days of osteogenic culture from heterozygous and homozygous osteoblasts. In contrast, a marked increase of adipocyte markers such as PPARc2 and aP2 was observed from heterozygous and homozygous osteoblasts under the same osteogenic media when compared with control osteoblasts. Data are expressed as the mean 6 SD from three independent experiments. *Significant difference from control (Pkd1 flox/flox ); # significant difference from heterozygous Col1a1(3.6)-Cre; Pkd1 flox/+ mice at P,0.05, respectively. doi:10.1371/journal.pone.0046038.g005 There was no cyst formation in the liver, kidney, and pancreas in heterozygous Col1a1(3.6)-Cre;Pkd1 flox/+ mice, whereas agematched homozygous Col1a1(3.6)-Cre;Pkd1 flox/flox mice developed severe renal and pancreatic cysts at 6 weeks of age. (B-D) Hematoxylin-eosinstained sections (5X) of liver, pancreas, and kidney from 6-week-old mice. Cysts were not observed in the livers from heterozygous and homozygous mice, and renal and pancreatic cysts were also not found in kidney and pancreas tissues from heterozygous Col1a1(3.6)-Cre;Pkd1 flox/+ mice. However, homozygous Col1a1(3.6)-Cre;Pkd1 flox/flox mice exhibited massive cyst formation in both the pancreas and kidney. Interestingly, glomeruli formation in cultures from these mice, suggesting that reduction of Pkd1 in osteoblast lineage diminishes FGF23 production. The disparity between serum FGF23 and bone expression of FGF23 is consistent with recent findings that FGF23 is regulated by both transcriptional and post-transcriptional mechanisms [50].
In addition to skeletal abnormalities, we found that Col1a1(3.6)-Cre-mediated deletion of Pkd1 resulted in cyst formation in the kidney and pancreas. Although Pkd1 is expressed in undifferentiated mesenchyme during embryogenesis [1,2,51] and Pkd1 mutations lead to cyst formation in the kidney, pancreas and liver in hereditary polycystic diseases [52,53,54,55,56], extraskeletal abnormalities due to deletion of Pkd1 in mesenchymal precursors using Col1a1(3.6)-Cre was unexpected. Indeed, the Col1a1(3.6) promoter has been purported to specifically target cells of the osteoblastic lineage, and in previous reports was found to have minimal expression of Cre-recombinase in liver, pancreas, and kidney [30,32]. However, unlike our studies, the prior reports did not examine Cre-mediated recombination in extraskeletal tissues. Indeed, we found evidence for Col1a1(3.6)-Cre deletion of Pkd1 in all tissues tested, including kidney, pancreas and liver. Pancreatic and renal cysts developed by E15.5, just after highly expression of Col1a1(3.6)-Cre activity in skeletal tissues at E12.5 [29,30,31]. The size of polycystic pancreas and kidney varied between individual mice, and only 30% of individual homozygous Pkd1 Col1a1(3.6)-cKO mice had both polycystic pancreas and kidney, indicating the the kidney and endocrine islet formation in pancreas appeared to be unaffected. In addition, expansion of pancreatic ducts formed large pancreatic cysts that led to massive acinar cell loss, formation of abnormal tubular structures, and appearance of endocrine cells in ducts. doi:10.1371/journal.pone.0046038.g006  possible effects of a mixed genetic background or other factors affecting cyst formation in these mice. This observation, along with the fact that mesenchymal cells give rise to both osteoblasts during bone development and renal tubular epithelia through mesenchymal-to-epithelial transition during kidney development [37,57,58,59], and pancreatic mesenchyme gives rise to pancreatic ducts by epithelial-mesenchymal interaction during pancreas development [60,61,62], suggests a broad role of Pkd1 in mesenchymal development pathways. Also, consistent with the known gene-dose dependent effect of Pkd1 in cystogenesis that requires a second hit in humans to cause ADPKD, there was no evidence of cysts formation in the pancreas or kidney of heterozygous Col1a1(3.6)-Cre;Pkd1 flox/+ mice. Approximately 30% of homozygous Pkd1 Col1a1(3.6)-cKO adult mice exhibited a polycystic pancreas containing solitary unilocular cysts, whereas pancreatic cysts have been reported in only 5% of patients with ADPKD [63,64].
Interestingly, we did not observe cyst formation in the liver of 6week-old homozygous Pkd1 Col1a1(3.6)-cKO mice, in spite of the known propensity for liver cyst formation in hereditary cystic disorders. The reason for the absence of cyst formation in the liver is not clear from our studies, because Col1a1(3.6)-Cre is expressed in the liver and results in excision of the Pkd1 flox allele in this tissue, similar to the pancreas and kidney. In contrast, MTTV-Cremediated conditional deletion of Pkd1 results in liver cysts by 10 weeks-of-age [25], suggesting that the lack of liver cysts in our study was a consequence of limitations of the Col1a1(3.6)-Cre or the need for additional time for cyst development in the liver. Regardless, the broad expression of Col1a1(3.6)-Cre in multiple tissues limits the conclusion that can be drawn from this targeting strategy [32].
Regardless, unlike the embryonic lethality of global Pkd1 null mice [7,15], the conditional Pkd1 null mice are born alive and exhibit a 50% 6-week survival rate, thereby creating a new model to study polycystic kidney and pancreatic cystic disease postnatally. We found that the expression of growth factor such as TGF-b was significantly increased in homozygous Pkd1 Col1a1(3.6)-cKO kidney at 6 weeks of age, which would stimulate the epithelial cell phenotype's transformation, producing myofibroblasts and secreting extracellular matrix (ECM). In fact, snail1 and vimentin, the EMT markers, and a-SMA and Col1a1, the fibrosis markers, were markedly upregulated in homozygous Pkd1 Col1a1(3.6)-cKO kidney compared with controls at 6 weeks of age. These findings agree with previous reports that polycystic kidney disease triggers the onset of epithelial-mesenchymal transition (EMT) and renal fibrosis [65,66,67,68,69,70].
In conclusion, the conditional deletion of Pkd1 from mesenchymal lineage results in both defective bone formation and polycystic kidney and pancreatic but not liver disease, indicating that val function in mesenchymal precursors to regulate skeletal, renal, and pancreatic development. The long-term survival of Pkd1 Co-l1a1(3.6)-cKO mice establishes a potential model to study postnatal interventions to retard cyst formation.
Bone densitometry, histomorphometric, marrow adipocyte staining, and microcomputed tomography analyses BMD of femurs was assessed at 6 weeks of age with a LUNAR PIXIMUS bone densitometer (Lunar Corp., Madison, WI, USA). Calcein (Sigma-Aldrich, St. Louis, MO, USA) double labeling of bone and histomorphometric analyses of periosteal MAR in tibias were performed using the osteomeasure analysis system (OsteoMetrics, Decatur, GA, USA) [71,72]. The distal femoral metaphyses were also scanned with a Scanco mCT 40 (Scanco Medical AG, Brüttisellen, Switzerland). 3D-images were analyzed to determine bone volume/trabecular volume and cortical thickness as previously described [71]. For detection of bone marrow adipocytes, whole intact femurs or tibiae with encapsulated marrow were dissected from 6-week-old mice, fixed for 48 h in phosphate-buffered paraformaldehyde, decalcified in 14% EDTA, and then stained with aqueous osmium tetroxide (OsO4) for quantification of fat volume, density, and distribution by mCT analysis [17]. In addition, the cryosectioning was performed for Oil Red O lipid staining as previously reported in our laboratory [17].

Serum biochemistry
Serum osteocalcin levels were measured using a mouse Osteocalcin EIA kit (Biomedical Technologies, Inc., Stoughton, MA, USA). Serum BUN was determined using a BUN diagnostic kit from Pointe Scientific, Inc (Canton, MI, USA). Serum calcium (Ca) was measured by the colorimetric cresolphthalein binding method, and phosphorus (P) was measured by the phosphomolybdate -ascorbic acid method (Stanbio Laboratory, Boerne, TX, USA). Serum osteoprotegerin (OPG) and Rank ligand (RankL) were measured using mouse ELISA kits (Quantikine H , R&D Systems, Minneapolis, MN, USA), and serum tartrate-resistant acid phosphatase (TRAP) was assayed with the ELISA-based SBA Sciences mouse TRAP TM assay (Immunodiagnostic Systems, Fountain Hills, AZ, USA). Serum parathyroid hormone (PTH) levels were measured using the Mouse Intact PTH ELISA kit (Immutopics, Carlsbad, CA, USA). Serum FGF23 levels were measured using the FGF23 ELISA kit (Kainos Laboratories, Tokyo, Japan).
Bone RNA isolation and real-time reverse transcriptase (RT)-PCR For quantitative real-time RT-PCR, 1.0 mg total RNA isolated from calvaria, kidney, liver, and whole tibias of 6-week-old control and Col1a1(3.6)-Cre-mediated Pkd1-deficient mice was reverse transcribed as previously described [34]. PCR reactions contained 20 gg template (cDNA or RNA), 375 gM each forward and reverse primers, and 1X SsoFast TM EvaGreenH supermix (Bio-Rad, Hercules, CA, USA) in a total of 10 ml reaction volume. The threshold cycle (Ct) of tested gene product from the indicated genotype was normalized to the Ct for cyclophilin A. Expression of total Pkd1 transcripts was performed using the following Pkd1allele-specific primers in exons 2-4: forward primer of normal Pkd1 + transcript (Pkd1 + plus Pkd1 flox ): 59-ATA GGG CTC CTG GTG AAC CT-39, and reverse primer: 59-CCA CAG TTG CAC TCA AAT GG-39. The normal Pkd1 + vs. cyclophilin A was normalized to the mean ratio of five control mice, which was set to 1. The percentage of conditional deleted Pkd1(Pkd1 Dflox ) transcripts was calculated from the relative levels of the normal Pkd1 + transcripts in different Pkd1-deficient mice [73]. All primer information of other genes used in real-time RT-PCR can be found in our previous report [18].
Cell proliferation, osteoblastic differentiation, and gene expression profiles in immortalized osteoblast cultures Calvaria from E17.5 control and Pkd1-deficient embryos were used to isolate primary osteoblasts by sequential collagenase digestion at 37uC. To engineer immortal osteoblast cell lines, isolated primary osteoblasts were infected using a retroviral vector carrying SV40 large and small T antigen as previously described [34,74]. Briefly, cells were grown in 100-mm plates at 50-60% confluence the day before infection. On the day of infection, the medium was removed and replaced with medium containing SV40 large and small T antigen-helper-free viral supernatant in the presence of 4 mg/ml of polybrene (Sigma, St. Louis, MO, USA) for 48 h. The cells were allowed to recover for 72 h followed by selection with 1 mg/ml puromycin (Sigma) for up to 15 days. The immortalized osteoblasts were cultured in a-MEM containing 10% FBS and 1% penicillin and streptomycin (P/S) and characterized following the protocols below. Cell proliferation was detected by BrdU incorporation assays following the manufacturer's directions (QIA58, Calbiochem, Gibbstown, NJ, USA). To induce differentiation, the immortalized osteoblasts were plated at a density of 2610 4 cells per well in a 12-well plate and 4610 4 cells per well in a 6-well plate and grown up to 21 days in a-MEM containing 10% FBS supplemented with 5 mM bglycerophosphate and 25 mg/ml ascorbic acid. ALP activity and Alizarin red-S histochemical staining for mineralization were performed as previously described [16,34]. Total DNA content was measured with a PicoGreen H dsDNA quantitation reagent and kit (Molecular Probes, Eugene, OR, USA). Protein concentrations of the supernatant were determined with a Bradford protein assay kit (Bio-Rad, Heumrcules, CA, USA). For gene expression profiles, 1.0 mg of total RNA was isolated from primary osteoblasts cultured 4, 12, and 21 days in differentiation media. The cDNAs were generated using an iScript reverse transcription kit (Bio-Rad). PCR reactions contained 20 gg template (cRNA or cDNA), 375 gmol each forward and reverse primers, 1X SsoFast TM EvaGreenH supermix (Bio-Rad) in a total of 10 ml reaction volume. The Ct of tested gene product from the indicated genotype was normalized to the Ct for cyclophilin A as previously described [16,34,36].

Statistical analysis
We evaluated differences between two groups by unpaired t-test and multiple groups by one-way analysis of variance. All values are expressed as means 6 SD. All computations were performed using GraphPad Prism5 (GraphPad Software Inc. La Jolla, CA, USA).