Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of...

10
Interrelated role of Klotho and calcium-sensing receptor in parathyroid hormone synthesis and parathyroid hyperplasia Yi Fan a,b , Weiqing Liu a,b , Ruiye Bi b,c , Michael J. Densmore a , Tadatoshi Sato a , Michael Mannstadt c , Quan Yuan b , Xuedong Zhou b , Hannes Olauson d , Tobias E. Larsson d , Hakan R. Toka e , Martin R. Pollak e , Edward M. Brown f , and Beate Lanske a,c,1 a Division of Bone and Mineral Research, Harvard School of Dental Medicine, Boston, MA 02115; b State Key Laboratory of Oral Diseases, National Clinical Research Center for Oral Diseases, West China Hospital of Stomatology, Sichuan University, Chengdu, 610041 Sichuan, China; c Endocrine Unit, Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114; d Division of Renal Medicine, Clinical Sciences, Intervention and Technology, Karolinska Institutet, SE-171 77 Stockholm, Sweden; e Division of Nephrology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02115; and f Division of Endocrinology, Diabetes and Hypertension, Department of Medicine, Brigham and Womens Hospital, Boston, MA 02115 Edited by David W. Russell, University of Texas Southwestern Medical Center, Dallas, TX, and approved March 9, 2018 (received for review October 19, 2017) The pathogenesis of parathyroid gland hyperplasia is poorly understood, and a better understanding is essential if there is to be improvement over the current strategies for prevention and treatment of secondary hyperparathyroidism. Here we investigate the specific role of Klotho expressed in the parathyroid glands (PTGs) in mediating parathyroid hormone (PTH) and serum calcium homeostasis, as well as the potential interaction between calcium- sensing receptor (CaSR) and Klotho. We generated mouse strains with PTG-specific deletion of Klotho and CaSR and dual deletion of both genes. We show that ablating CaSR in the PTGs increases PTH synthesis, that Klotho has a pivotal role in suppressing PTH in the absence of CaSR, and that CaSR together with Klotho regulates PTH biosynthesis and PTG growth. We utilized the tdTomato gene in our mice to visualize and collect PTGs to reveal an inhibitory function of Klotho on PTG cell proliferation. Chronic hypocalcemia and ex vivo PTG culture demonstrated an independent role for Klotho in mediating PTH secretion. Moreover, we identify an in- teraction between PTG-expressed CaSR and Klotho. These findings reveal essential and interrelated functions for CaSR and Klotho during parathyroid hyperplasia. secondary hyperparathyroidism | FGF23 | chronic kidney disease | tdTomato | conditional knockout C hronic kidney disease (CKD) is a major, worldwide public health issue and has a high impact on mortality and mor- bidity. The overall CKD mortality has increased by 32% over the last decades (1). In the United States, Medicare spending for patients with CKD ages 65 and older exceeded $55 billion in 2015, representing 20% of all Medicare spending in this age group, consuming substantial financial and social resources (2). The most severe systemic problems are related to the mineral and hormonal imbalances that instigate secondary hyperpara- thyroidism (SHPT) accompanied by parathyroid gland (PTG) hyperplasia (3). The metabolic imbalances and hyperplasia then reinforce each other as the disease progresses. SHPT in CKD is associated with renal osteodystrophy, extraskeletal calcification, impaired bone mineralization, and cardiovascular disease (4). SHPT and the exacerbated mineral metabolism disturbances are the chief contributors to mortality in CKD patients (5). The current treatment is to artificially balance one of the metabolic factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under- standing of how the PTGs initiate and maintain the hyperplastic state is of high clinical significance. In healthy individuals, the PTG detects serum calcium levels via the calcium-sensing receptor (CaSR), a G protein-coupled receptor that modulates the secretion of parathyroid hormone (PTH) in response to low circulating levels of calcium to main- tain homeostasis (8, 9). PTH targets the kidneys to (i ) stimulate calcium reabsorption and (ii ) increase the production of active vitamin D, which in turn induces intestinal calcium absorption (10). Fibroblast growth factor 23 (FGF23) is secreted by bone in response to higher serum 1,25-dihydroxyvitamin D [1,25(OH) 2 D 3 ] and phosphate levels and as a consequence of the changes in calcium metabolism initiated by PTH (11). PTH plays an impor- tant role in promoting FGF23 transcription and serum FGF23 concentration. Previous studies have elucidated that PTH directly induces FGF23 production in a vitamin D-independent manner (1214). PTH expression and production are suppressed as serum levels of calcium, vitamin D, and FGF23 rise in response to PTHs actions, thus providing a negative control mechanism to maintain proper balance (1517). As CKD develops and progresses, how- ever, this regulatory network breaks down and serum PTH levels increase despite significant elevated serum FGF23 levels (15). The mechanism underlying this resistance to regulatory control is not Significance Secondary hyperparathyroidism (SHPT) is a severe consequence of chronic kidney disease. A better understanding of the mech- anisms controlling the progression of SHPT and the regulation of parathyroid hormone (PTH) production is clinically relevant. Uti- lizing parathyroid gland (PTG)-specific knockout mouse models, we demonstrated calcium-sensing receptor (CaSR) and Klotho together regulate PTH synthesis and PTG growth, and that Klotho contributes to PTH suppression in the absence of CaSR. Klotho exerts an independent function in mediating PTH secre- tion under chronic hypocalcemia and in suppressing PTG cell proliferation. Moreover, the results revealed a previously un- identified interaction between PTG-expressed CaSR and Klotho. These findings highlight the essential and interrelated roles for CaSR and Klotho to prevent parathyroid hyperplasia, suggesting potential treatment strategies to control PTH synthesis and hyperparathyroidism. Author contributions: Y.F., E.M.B., and B.L. designed research; Y.F., W.L., R.B., and T.S. performed research; T.E.L., H.R.T., M.R.P., and E.M.B. contributed new reagents/analytic tools; Y.F., W.L., R.B., M.J.D., T.S., M.M., Q.Y., X.Z., H.O., E.M.B., and B.L. analyzed data; and Y.F., M.J.D., and B.L. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Published under the PNAS license. 1 To whom correspondence should be addressed. Email: [email protected]. This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1717754115/-/DCSupplemental. Published online April 4, 2018. www.pnas.org/cgi/doi/10.1073/pnas.1717754115 PNAS | vol. 115 | no. 16 | E3749E3758 MEDICAL SCIENCES PNAS PLUS Downloaded by guest on July 20, 2020

Transcript of Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of...

Page 1: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

Interrelated role of Klotho and calcium-sensingreceptor in parathyroid hormone synthesis andparathyroid hyperplasiaYi Fana,b, Weiqing Liua,b, Ruiye Bib,c, Michael J. Densmorea, Tadatoshi Satoa, Michael Mannstadtc, Quan Yuanb,Xuedong Zhoub, Hannes Olausond, Tobias E. Larssond, Hakan R. Tokae, Martin R. Pollake, Edward M. Brownf,and Beate Lanskea,c,1

aDivision of Bone and Mineral Research, Harvard School of Dental Medicine, Boston, MA 02115; bState Key Laboratory of Oral Diseases, National ClinicalResearch Center for Oral Diseases, West China Hospital of Stomatology, Sichuan University, Chengdu, 610041 Sichuan, China; cEndocrine Unit, Departmentof Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114; dDivision of Renal Medicine, Clinical Sciences, Interventionand Technology, Karolinska Institutet, SE-171 77 Stockholm, Sweden; eDivision of Nephrology, Department of Medicine, Beth Israel Deaconess MedicalCenter, Harvard Medical School, Boston, MA 02115; and fDivision of Endocrinology, Diabetes and Hypertension, Department of Medicine, Brigham andWomen’s Hospital, Boston, MA 02115

Edited by David W. Russell, University of Texas Southwestern Medical Center, Dallas, TX, and approved March 9, 2018 (received for review October 19, 2017)

The pathogenesis of parathyroid gland hyperplasia is poorlyunderstood, and a better understanding is essential if there is tobe improvement over the current strategies for prevention andtreatment of secondary hyperparathyroidism. Here we investigatethe specific role of Klotho expressed in the parathyroid glands(PTGs) in mediating parathyroid hormone (PTH) and serum calciumhomeostasis, as well as the potential interaction between calcium-sensing receptor (CaSR) and Klotho. We generated mouse strainswith PTG-specific deletion of Klotho and CaSR and dual deletion ofboth genes. We show that ablating CaSR in the PTGs increases PTHsynthesis, that Klotho has a pivotal role in suppressing PTH in theabsence of CaSR, and that CaSR together with Klotho regulatesPTH biosynthesis and PTG growth. We utilized the tdTomato genein our mice to visualize and collect PTGs to reveal an inhibitoryfunction of Klotho on PTG cell proliferation. Chronic hypocalcemiaand ex vivo PTG culture demonstrated an independent role forKlotho in mediating PTH secretion. Moreover, we identify an in-teraction between PTG-expressed CaSR and Klotho. These findingsreveal essential and interrelated functions for CaSR and Klothoduring parathyroid hyperplasia.

secondary hyperparathyroidism | FGF23 | chronic kidney disease |tdTomato | conditional knockout

Chronic kidney disease (CKD) is a major, worldwide publichealth issue and has a high impact on mortality and mor-

bidity. The overall CKD mortality has increased by 32% over thelast decades (1). In the United States, Medicare spending forpatients with CKD ages 65 and older exceeded $55 billion in2015, representing 20% of all Medicare spending in this agegroup, consuming substantial financial and social resources (2).The most severe systemic problems are related to the mineraland hormonal imbalances that instigate secondary hyperpara-thyroidism (SHPT) accompanied by parathyroid gland (PTG)hyperplasia (3). The metabolic imbalances and hyperplasia thenreinforce each other as the disease progresses. SHPT in CKD isassociated with renal osteodystrophy, extraskeletal calcification,impaired bone mineralization, and cardiovascular disease (4).SHPT and the exacerbated mineral metabolism disturbances arethe chief contributors to mortality in CKD patients (5). Thecurrent treatment is to artificially balance one of the metabolicfactors, use a combination of medications to balance multiplefactors, or surgically remove the PTGs (6, 7). A deeper under-standing of how the PTGs initiate and maintain the hyperplasticstate is of high clinical significance.In healthy individuals, the PTG detects serum calcium levels

via the calcium-sensing receptor (CaSR), a G protein-coupledreceptor that modulates the secretion of parathyroid hormone

(PTH) in response to low circulating levels of calcium to main-tain homeostasis (8, 9). PTH targets the kidneys to (i) stimulatecalcium reabsorption and (ii) increase the production of activevitamin D, which in turn induces intestinal calcium absorption(10). Fibroblast growth factor 23 (FGF23) is secreted by bone inresponse to higher serum 1,25-dihydroxyvitamin D [1,25(OH)2D3]and phosphate levels and as a consequence of the changes incalcium metabolism initiated by PTH (11). PTH plays an impor-tant role in promoting FGF23 transcription and serum FGF23concentration. Previous studies have elucidated that PTH directlyinduces FGF23 production in a vitamin D-independent manner(12–14). PTH expression and production are suppressed as serumlevels of calcium, vitamin D, and FGF23 rise in response to PTH’sactions, thus providing a negative control mechanism to maintainproper balance (15–17). As CKD develops and progresses, how-ever, this regulatory network breaks down and serum PTH levelsincrease despite significant elevated serum FGF23 levels (15). Themechanism underlying this resistance to regulatory control is not

Significance

Secondary hyperparathyroidism (SHPT) is a severe consequenceof chronic kidney disease. A better understanding of the mech-anisms controlling the progression of SHPT and the regulation ofparathyroid hormone (PTH) production is clinically relevant. Uti-lizing parathyroid gland (PTG)-specific knockout mouse models,we demonstrated calcium-sensing receptor (CaSR) and Klothotogether regulate PTH synthesis and PTG growth, and thatKlotho contributes to PTH suppression in the absence of CaSR.Klotho exerts an independent function in mediating PTH secre-tion under chronic hypocalcemia and in suppressing PTG cellproliferation. Moreover, the results revealed a previously un-identified interaction between PTG-expressed CaSR and Klotho.These findings highlight the essential and interrelated roles forCaSR and Klotho to prevent parathyroid hyperplasia, suggestingpotential treatment strategies to control PTH synthesis andhyperparathyroidism.

Author contributions: Y.F., E.M.B., and B.L. designed research; Y.F., W.L., R.B., and T.S.performed research; T.E.L., H.R.T., M.R.P., and E.M.B. contributed new reagents/analytictools; Y.F., W.L., R.B., M.J.D., T.S., M.M., Q.Y., X.Z., H.O., E.M.B., and B.L. analyzed data;and Y.F., M.J.D., and B.L. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Published under the PNAS license.1To whom correspondence should be addressed. Email: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1717754115/-/DCSupplemental.

Published online April 4, 2018.

www.pnas.org/cgi/doi/10.1073/pnas.1717754115 PNAS | vol. 115 | no. 16 | E3749–E3758

MED

ICALSC

IENCE

SPN

ASPL

US

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 2: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

fully understood, although some key issues have been identified(18). These include the drop in vitamin D levels due to the highFGF23 and a reduction in the expression of FGF23’s requiredcofactor, Klotho, and its receptor, fibroblast growth factor re-ceptor 1 (FGFR1), in the PTGs, reducing the ability of FGF23 toefficiently control the PTG function (18, 19).Type I membrane-bound alpha-Klotho (Klotho, KL) is a

transmembrane protein that has been found in mice, rats, andhumans predominantly in the renal distal convoluted tubular cellsand to a lesser extent in proximal convoluted tubular epithelialcells and the PTGs (18, 20, 21). It is also expressed in variousorgan systems, including arterial, epithelial, reproductive, andneuronal tissues (22). The principal role of Klotho is to form thespecific receptor complex with FGFR1 that is required for FGF23signaling (23–25). Activation of the FGFR1–Klotho complex byFGF23 in the kidney regulates phosphate homeostasis by affectingthe expression of sodium-phosphate cotransporters, Napi2a andNapi2c, in the proximal tubules. It also inhibits 1,25(OH)2D3synthesis by altering the vitamin D-metabolizing enzymes CYP27b1and CYP24a1 (26–29). Noteworthy, Klotho is also highly expressedin the PTGs, but a link between Klotho and PTH in the regulationof serum calcium homeostasis has not yet been investigated.FGF23–Klotho signaling has been shown to inhibit PTH mRNAtranscription and hormone secretion in vitro (30) and to negativelyregulate PTH secretion in vivo (15). It has also been shown that thecalcineurin-mediated FGF23 signaling pathway in PTGs mediatessuppression of PTH (31). Contrary to these studies, Fgf23−/− miceoverexpressing human recombinant FGF23 under the control ofthe 2.3-kb collagen 1 promoter exhibited high serum PTH despiteunchanged serum calcium and 1,25(OH)2D3 levels, and despitehypophosphatemia (32). It has been suggested that FGF23 is along-term inducer of PTH during CKD (33). Klotho activity hasbeen implicated as fundamental for the stimulation of PTH se-cretion in hypocalcemia conditions by recruiting Na+, K+-ATPase(34), although the underlying mechanism has been challenged (35).In this study, we sought to determine the tissue-specific role of

Klotho in PTGs in controlling PTH and serum calcium homeo-stasis and to explore a potential interaction between Klotho andCaSR in proper PTH regulation. We therefore generated mousestrains with a parathyroid-specific deletion of Klotho or CaSRand a parathyroid-specific dual deletion of both genes. The re-sults of our study reveal an independent function of PTG Klothoto regulate PTH production and cell proliferation, as well as apossible interaction between Klotho and CaSR.

Materials and MethodsAnimals. Mice with a PTG-specific deletion of Klotho, CaSR, or both togetherwere generated using the Cre-LoxP recombination system. The derivation ofthe PTHcre;KLfl/fl mice was previously described (31). In short, LoxP sequenceswere introduced into the flanking regions of exon 2 of the Klotho gene.PTHcre;CaSRfl/fl mice were generated by mating mice in which the PTHpromoter drives the expression of Cre-recombinase (36) with mice in whichexon 3 of CaSR gene was flanked by LoxP sites (37), featuring a complete lossof CaSR RNA and protein, thereby avoiding the limitations of the previousmouse models (38). KLfl/fl and CaSRfl/fl mice were mated to obtain KLfl/fl;CaSRfl/fl

mice, which were then mated with PTHcre mice to generate PTHcre;KLfl/+;CaSRfl/+ mice. PTHcre;KLfl/+;CaSRfl/+ male mice were fertile and were interbredwith KLfl/fl;CaSRfl/fl females to attain KLfl/fl;CaSRfl/fl (control) and PTHcre;KLfl/fl;CaSRfl/fl. KLfl/fl;CaSRfl/fl, PTHcre;KLfl/fl, PTHcre;CaSRfl/fl, and PTHcre;KLfl/fl;CaSRfl/fl

mice at postnatal day 10 (P10) were used to analyze serum and urinary biochem-istries. Isolating PTGs in mice is extremely difficult. B6.Cg-Gt(ROSA)26Sortm14(CAG-tdTomato)Hze/J mice (here referred to as Tmfl/fl) were purchased fromJackson Laboratory and were interbred with our lines to generatePTHcre;Tmfl/+, PTHcre;Tmfl/+;KLfl/fl, PTHcre;Tmfl/+;CaSRfl/fl, and PTHcre;Tmfl/+;KLfl/fl; CaSRfl/fl mice. These mice were utilized to collect PTGs for RNA ex-pression at P10. We performed a calcium-deficient diet study with 3-wk-oldPTHcre;Tmfl/+, PTHcre;Tmfl/+;KLfl/fl, PTHcre;Tmfl/+;CaSRfl/+, and PTHcre;Tmfl/+;KLfl/fl;CaSRfl/+ mice. Mice were fed a low-Ca2+ diet (0% calcium and 0.4%phosphorus) based on American Institute of Nutrition Research Diets (AIN-93).The control calcium diet contained 0.6% calcium and 0.4% phosphorus (39). The

use of all animals in this study was approved by the Institutional Animal Careand Use Committee at the Harvard Medical School.

PTG Histology and Immunostaining. Thyro-PTG tissues were dissected andfixed in 10% formalin overnight at 4 °C, dehydrated, and embedded inparaffin. Five-micrometer sections were prepared for H&E staining andimmunostaining. Slides for immunostaining were deparaffinized and rehy-drated. Antigen retrieval was performed using citric acid buffer for 20 min.Sections used for immunohistochemistry staining were immersed in 3%H2O2 in methanol for 10 min. Sections were blocked with blocking buffer for30 min and then incubated with primary antibodies mouse anti-PTH (1:100;Serotec, Inc.), rat anti-KL (1:100; TransGenic Inc.), rabbit anti-CaSR (1:200, agenerous gift from E.M.B.), rabbit anti-Ki67 (1:100; Cell Signaling), andrabbit anti phospho-Erk1/2 (1:200; Cell Signaling) at 4 °C overnight. AlexaFluor-conjugated secondary antibodies were used for immunofluorescencestaining (Invitrogen). All sections were mounted with HardSet MountingMedium with DAPI (Vector Labs). Sections used for immunohistochemistrystaining were incubated with biotinylated secondary antibodies followed byVector ABC Reagent and developed with DAB substrate (Vector Labs). He-matoxylin was used for counterstaining. For TUNEL staining, apoptotic cellswere detected by In Situ Cell Death Detection Kit, Fluorescein (Roche) fol-lowed the manufacturer’s instructions. For counting Ki67+ cells and apo-ptotic cells, one in every 5-10 sections was stained and the ratios of thenumber of Ki67+/total cells and TUNEL+/total cells were calculated by ImageJ.

Coimmunoprecipitation Assay.HEK293 cells were cultured in DMEMwith 10%FBS, 4.5 g/L D-glucose, 584 mg/L L-glutamine, 110 mg/L sodium pyruvate,100 U/mL penicillin, and 100 μg/mL streptomycin (Invitrogen). The pRK5-GFP-CaSR plasmid was kindly provided by Jeremy M. Henley, University of Bristol,Bristol, UK (40). The pCS-mKLcFT plasmid was obtained fromMakoto Kuro-o,Jichi Medical University, Shimotsuke, Japan (41). Cells grown in 100-mmculture dishes were transfected with 4 μg of cDNA using PolyJet In Vitro DNATransfection Reagent (SignaGen Laboratories) according to the manufacturer’sprotocol. Cells were homogenized with radioimmunoprecipitation assay buffer(Alfa Aesar) containing phosphatase and protease inhibitor mixture tablets(Roche) and centrifuged at 14,000 × g. The supernatant was then transferred toa fresh tube. Protein concentration was measured by BCA protein assay (Pierce).Immunoprecipitation was performed using Protein A/G PLUS-Agarose Immu-noprecipitation Reagent (sc-2003; Santa Cruz) according to the manufacturer’sprotocol. Protein samples were heated with NuPAGE LDS Sample buffer andNuPAGE Reducing Agent (Life Technologies) at 70 °C for 10 min and thensubjected to NuPAGE 4–12% Bis-Tris SDS/PAGE using precast gels (Invitrogen)and NuPAGE Mes SDS running buffer (Life Technologies). The separated pro-teins in the gel were electrophoretically transferred to Hybond-P polyvinylidene-difluoride transfer membranes. After incubation in blocking solution for 1 h, themembranes were further treated with a rat anti-human Klotho monoclonalantibody (1:1,000, KM2076; Cosmo Bio) or rabbit anti-GFP antibody (1:1,000;Invitrogen). Horseradish peroxidase-conjugated anti-rat or rabbit IgG was usedas the secondary antibody (Jackson ImmunoResearch Laboratories), and signalwas detected by Amersham ECL Prime Western Blotting Detection Reagent(GE Healthcare).

Statistics. GraphPad Prism 6.0 (GraphPad Software Inc.) was used for statisticalanalysis. Comparisons between groups were evaluated by unpaired two-tailedStudent’s t test between two groups or by one-way ANOVA followed byTukey’s test for multiple comparisons. Two-way ANOVA and Bonferroniposttests were used to evaluate individual samples between control diet andlow-Ca2+ diet. All values are expressed as mean ± SEM. P values <0.05 wereconsidered significant for all analyses.

Extended methods and information about genotyping, serum and urinemeasurements, PTG isolation, ex vivo PTG culture, RNA isolation, and qRT-PCRanalyses, skeletal preparations and bone histology, and Western blot aredescribed in SI Materials and Methods.

ResultsGeneration of PTHCre;KLfl/fl, PTHCre;CaSRfl/fl, and PTHCre;KLfl/fl;CaSRfl/fl

Mice. Mice with PTG-specific deletion of Klotho or CaSR, andmice with specific deletion of both, were generated using Cre-LoxP recombination. Immunostaining and qRT-PCR using PTGRNA confirmed efficient deletion of Klotho and CaSR in thePTG tissue with the according genotype at protein and mRNAlevels, respectively (Fig. S1). These mice were born at the expec-ted Mendelian ratio. PTHCre;KLfl/fl mice were viable andappeared macroscopically normal in size and weight. However,

E3750 | www.pnas.org/cgi/doi/10.1073/pnas.1717754115 Fan et al.

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 3: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl mice were severelygrowth-retarded (Fig. 1A). These mice had reduced body weightcompared with control littermates (KLfl/fl;CaSRfl/fl) and died be-tween 1 and 2 wk of age. Moreover, PTHCre;KLfl/fl;CaSRfl/fl miceexhibited a significantly lower body weight and shorter life ex-pectancy than PTHCre;CaSRfl/fl mice (Fig. 1 B and C).

Serum Biochemistry. PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl

mice died within 2–3 wk of age; therefore, we compared serumparameters among all four genotypes at P10. We were interestedin a potential interaction between Klotho and CaSR in PTGtissues to modulate mineral ion homeostasis. We detected nodifferences in serum Ca2+, Pi, PTH, intact FGF23 (iFGF23), and1,25(OH)2D3 in PTHCre;KLfl/fl mice compared with controls(Fig. 1 D–H). However, both PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl mice displayed hypercalcemia, hypophosphatemia, andsignificantly increased serum PTH, iFGF23, and 1,25(OH)2D3levels (Fig. 1 D–H). Most importantly, serum PTH, iFGF23, and1,25(OH)2D3 levels in PTHCre;KLfl/fl;CaSRfl/fl mice were furtherelevated and significantly higher than those in PTHCre;CaSRfl/fl

mice, indicating a suppressive function of Klotho on PTH syn-thesis in the absence of CaSR (Fig. 1 F–H).

Renal and Bone Phenotype. We investigated potential secondaryeffects of PTG-specific deletion of the Klotho and/or CaSR geneson the main target organs of PTH signaling, namely kidney andbone. There were no significant alterations in urinary Ca2+ and Piexcretion of PTHCre;KLfl/fl mice at P10. Both PTHCre;CaSRfl/fl

and PTHCre;KLfl/fl;CaSRfl/fl mice exhibited hypercalciuria andhyperphosphaturia compared with control or PTHCre;KLfl/fl mice(Fig. S2 A and B). Moreover, these mice had significantly in-creased renal Cyp27b1 gene expression (Fig. S2C), which waspositively correlated with the higher serum 1,25(OH)2D3 levels.Renal Napi2a gene and protein expressions were markedly de-creased in PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl mice (Fig.S2 D–F). Radiographs of hind limbs showed no significant changesin bone due to Klotho deletion, while CaSR deletion in PTGs

led to markedly increased radiolucency in metaphysis and di-aphysis (Fig. S3A). Moreover, the formation of the secondaryossification center in the epiphyses of PTHCre;CaSRfl/fl andPTHCre;KLfl/fl;CaSRfl/fl mice was dramatically delayed (Fig. S3B).Alizarin Red and Alcian Blue staining showed that these twomouse mutants had severely undermineralized skeletons and thetypical rickets-like nodules in the ribs. PTHCre;KLfl/fl;CaSRfl/fl

mice exhibited a more severe skeletal phenotype compared withsingle CaSR-deletion mice (Fig. S3C).

Klotho and CaSR Regulate PTG Growth. Visualizing and collectingthe PTGs from mice for further analyses is difficult due to thesmall size of the glands. We thus introduced the tdTomato re-porter gene (Tmfl/fl) to generate new mouse lines, includingPTHcre;Tmfl/+, PTHcre;Tmfl/+;KLfl/fl, PTHcre;Tmfl/+;CaSRfl/fl,and PTHcre;Tmfl/+;KLfl/fl;CaSRfl/fl. The presence of the fluores-cent reporter facilitated detection and dissection of the PTGsusing a fluorescent stereomicroscope. tdTomato fluorescentimages from P10 mice suggested that the size of the PTG tissuewas significantly increased in PTHcre;Tmfl/+;CaSRfl/fl andPTHcre;Tmfl/+;KLfl/fl;CaSRfl/fl mice. This is especially significantin the presence of the dramatically decreased body weight ofthose mutant mice. Most notably, PTGs in PTHcre;Tmfl/+;KLfl/fl;CaSRfl/fl mice were even further enlarged compared with PTHcre;Tmfl/+;CaSRfl/fl. There was only a trend toward an increase in PTGsize in PTHcre;Tmfl/+;KLfl/fl mice at P10 (Fig. 2 A and B).We next examined the histology of the PTGs at P10. We gen-

erated paraffin sections and used H&E staining to confirm thesize difference (Fig. 2C). The results suggested that no significantstructural difference could be detected between PTHCre;KLfl/fl

and control mice. Both had the typical rope-like structure withdense parathyroid cells. Interestingly, cells in PTHCre;CaSRfl/fl

and PTHCre;KLfl/fl;CaSRfl/fl PTGs appeared to be larger. Also, theglands contained more eosinophilic areas with nodular forma-tions. The morphological structure is more severe and disruptedin double-mutant mice (Fig. 2C). A proliferation analysis revealeda significant increase in the percentage of Ki67-positive cells in

Fig. 1. Phenotype of parathyroid-specific deletion of Klotho, CaSR, and both. (A–C) Gross phenotype, body weight, and survival rate of control, PTHCre;KLfl/fl,PTHCre;CaSRfl/fl, and PTHCre;KLfl/fl;CaSRfl/fl mice at P10. (D–H) Measurements of serum parameters in control and mutant mice at P10. Significant changes areevident between PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl mice. *P < 0.05, **P < 0.01, ***P < 0.001 versus control; #P < 0.05, ##P < 0.01, ###P < 0.001 versusPTHCre;KLfl/fl; $P < 0.05, $$P < 0.01, $$$P < 0.001 versus PTHCre;CaSRfl/fl.

Fan et al. PNAS | vol. 115 | no. 16 | E3751

MED

ICALSC

IENCE

SPN

ASPL

US

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 4: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl PTGs comparedwith controls. This confirms that an increase in proliferationcontributed to the enlargement of the glands (Fig. 2 D and E).

Both Klotho and CaSR Control the Expression Pattern of PTH. Weperformed immunofluorescence double staining for PTH andCaSR to investigate the PTH expression pattern in the presenceand absence of CaSR. Confocal images of double-stained sec-tions showed a normal PTH distribution, coordinated with theCaSR expression pattern in PTHcre;KLfl/fl mice (Fig. 3 A and B).Interestingly, in PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl

PTGs, many nodules were visible at CaSR-negative areas thatwere also highly positive for PTH, indicating that CaSR-deficientcells are contributing to higher PTH synthesis (Fig. 3 A and B).Total RNA from PTGs was extracted at P10 for qRT-PCR anal-ysis; however, no changes were observed in PthmRNA expression,despite the markedly elevated serum PTH levels in PTHcre;CaSRfl/fl and PTHcre;KLfl/fl;CaSRfl/fl mice (Fig. 3C).Surprisingly, we observed areas of altered tissue structure in

some extremely hyperplastic PTGs of PTHcre;KLfl/fl;CaSRfl/fl micethat failed to express PTH protein (Fig. S4 A and B). To investigatewhether this could be due to cell death, we performed TUNELstaining. The results suggested that ablation of CaSR leads to in-creased parathyroid cell apoptosis, which is even more pronouncedin the PTHcre;KLfl/fl;CaSRfl/fl glands (Fig. S4 C and D).We next investigated a few consequences on gene transcription

upon deletion of Klotho/CaSR from PTGs. The results demon-strated that expression of Fgfr1, Egr1, and 1α(OH)ase were slightlyreduced in PTGs of Klotho-deficient mice. Furthermore, CaSRdeletion resulted in increased 24(OH)ase expression levels, whichexhibited comparatively lower expression in Klotho-ablatedglands. Moreover, mutant mice lacking Klotho, CaSR, or both

showed reduced expression of Vdr (Fig. S5 A–E). We next ex-plored the MAP kinase cascade using immunofluorescencestaining. As shown in Fig. S5 F and G, there was significantlyreduced signal for phosphorylated ERK1/2 in parathyroid tissuefrom Klotho- and/or CaSR-ablated mice, indicating that the MAPkinase pathway was markedly suppressed in Klotho- and CaSR-deficient PTGs.

Klotho and CaSR Regulate PTH Secretion Under Chronic Hypocalcemia.Our next goal was to determine the role of PTG-specific Klothoand CaSR deletion in development and progression of PTG hy-perplasia under conditions of chronic hypocalcemia. We thereforechallenged PTHCre;Tmfl/+;KLfl/fl, PTHCre;Tmfl/+;CaSRfl/+, andPTHCre;Tmfl/+;KLfl/fl;CaSRfl/+ mice with a low-Ca2+ diet for 3 wk,starting at weaning. PTHCre;Tmfl/+ mice were used as controls.Serum Ca2+ and Pi levels were significantly reduced in all geno-types after being on the low-Ca2+ diet (Fig. 4 A and B). Asexpected, the low-Ca2+ diet significantly increased serum PTHlevels by around sevenfold in PTHCre;Tmfl/+ mice compared withthose on a control diet (Fig. 4C). However, both in PTHCre;Tmfl/+;KLfl/fl and PTHCre;Tmfl/+;CaSRfl/+ mice, the increase in serumPTH was only around twofold and thus not as great as that incontrols. Notably, we observed the highest PTH levels under low-Ca2+ conditions in PTHCre;Tmfl/+;KLfl/fl;CaSRfl/+ mice (Fig. 4C).This is consistent with the previous observation that dual deletionof Klotho and CaSR in the PTGs at P10 resulted in excess secre-tion of PTH and confirms that ablation of both factors results inabnormal production of PTH.PTH immunostaining suggested that the pattern of PTH-

secreting cells appeared normal in PTHCre;Tmfl/+ and PTHCre;Tmfl/+;KLfl/fl mice under low-Ca2+ diet conditions compared withthose on a control diet. Partial deletion of CaSR caused diffused

Fig. 2. Klotho and CaSR regulate growth of PTGs. (A and B) Tomato fluorescence and size calculations for PTGs from P10 mice showed parathyroid-specificCaSR ablation leads to significantly larger glands relative to body weight. Ablation of both Klotho and CaSR in PTGs resulted in a further enlargement of PTGsize over single CaSR-deletion mice. There was a trend toward an increase in PTHCre;Tmfl/+;KLfl/fl PTG size compared with controls at P10. (Scale bar, 500 μm.)(C) H&E staining confirmed the changes of PTG size and showed that CaSR deletion affected parathyroid morphological structure and caused nodularformations. This was more severe for Klotho and CaSR dual ablation. (Scale bar, 200 μm in low magnification, 50 μm in high magnification.) Black arrowheadsindicate nodules. (D and E) Immunohistochemical staining of Ki67 and calculation of Ki67-positive cell ratio in PTGs, showing increased cell proliferation inPTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl mice at P10. n = 6. (Scale bar, 50 μm.) *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 versus control; #P < 0.05,##P < 0.01, ####P < 0.0001 versus PTHCre;KLfl/fl; $$$P < 0.001 versus PTHCre;CaSRfl/fl.

E3752 | www.pnas.org/cgi/doi/10.1073/pnas.1717754115 Fan et al.

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 5: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

PTH protein expression under control or low-Ca2+ diet condi-tions (Fig. 4D). Notably, we observed the PTG nodule forma-tions only in PTHCre;Tmfl/+;KLfl/fl;CaSRfl/+ mice fed a low-Ca2+

diet. This indicates that dual deletion of Klotho and CaSR leadsto nodule formation with excess PTH production, which is inaccord with observation of the highest serum PTH level inthis group.We next confirmed the effect of Klotho on PTH production

without the confounding influence of other circulating factors.We dissected the PTGs and performed PTG ex vivo culture. Theglands of PTHCre;Tmfl/+ and PTHCre;Tmfl/+;KLfl/fl mice weresubjected to low-Ca2+ (0.5 mM Ca2+) or high-Ca2+ (3 mM Ca2+)conditions over a 1.5-h period (Fig. 4E) and secreted PTH wasmeasured by ELISA. Regardless of calcium concentration,PTHCre;Tmfl/+;KLfl/fl PTGs secreted more PTH than controlglands. At low Ca2+ concentrations, PTH production was sig-

nificantly induced in both genotypes but showed a further ele-vation in Klotho-deficient glands (Fig. 4F).

Klotho and CaSR Suppress PTG Hyperplasia Under ChronicHypocalcemia. PTGs were dissected from mice using Tomatofluorescence as a guide and size was calculated by ImageJ. Theresults showed that under normal dietary conditions the size ofthe glands in PTHCre;Tmfl/+;KLfl/fl, PTHCre;Tmfl/+;CaSRfl/+, andPTHCre;Tmfl/+;KLfl/fl; CaSRfl/+ mice was significantly increasedcompared with PTHCre;Tmfl/+ mice. The effect was even morepronounced when Klotho was deleted (Fig. 5 A and B). A low-Ca2+ diet induced PTG hyperplasia in all genotypes, but a moresevere increase occurred in the absence of Klotho (Fig. 5 A andB). Histological analyses of H&E-stained paraffin sections ofPTGs showed that under a control diet PTG tissue fromPTHCre;Tmfl/+;KLfl/fl, PTHCre;Tmfl/+;CaSRfl/+, and PTHCre;

Fig. 3. Klotho and CaSR control the expression pattern of PTHs. (A) H&E staining of PTGs in control and mutant mice at P10. CaSR ablation caused noduleformations. Black arrowheads indicate nodules. n = 6. (Scale bar, 50 μm.) (B) Confocal images of immunofluorescent double staining of PTH and CaSR. NotePTG nodules were discovered coincident with CaSR-negative cells in PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl PTGs. White arrowheads indicate PTG nodules.n = 3. (Scale bar, 10 μm.) (C) qRT-PCR analysis of Pth transcript in PTGs showed comparable expression levels in control and mutant mice.

Fan et al. PNAS | vol. 115 | no. 16 | E3753

MED

ICALSC

IENCE

SPN

ASPL

US

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 6: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

Tmfl/+;KLfl/fl;CaSRfl/+ mice appeared to be more eosinophiliccompared with that of control glands (Fig. 5C). There was asignificant increase in the PTG volume in all groups in responseto hypocalcemia.Ki67 immunostaining showed that glands with Klotho or

Klotho/CaSR deletion exhibit increased cell proliferation undercontrol diet conditions compared with control glands (Fig. 6).Furthermore, chronic hypocalcemia induced by low-Ca2+ dietresulted in a significant up-regulation of Ki67-positive cells onlyin Klotho-ablated PTGs, indicating that an increase in cell pro-liferation contributed to the observed PTG hyperplasia in theabsence of Klotho (Fig. 6). TUNEL staining showed that cellapoptosis was not affected by Klotho or partial CaSR deletion orby diet conditions. These results suggest that Klotho and CaSRhave an important role in PTG growth by regulating cell pro-liferation. Deletion of either factor leads to PTG hyperplasia(Fig. S6).

Interaction of Klotho and CaSR in PTGs. We isolated PTGs usingtdTomato fluorescence to eliminate any contamination by sur-rounding thyroid tissue. qRT-PCR analyses and immunostaining

showed that deletion of Klotho leads to a tendency of lowerCaSR expression on either protein or mRNA levels at P10 (Fig.7 A and B). Interestingly, however, Klotho mRNA expressionlevels were significantly decreased in CaSR-deleted PTGs (Fig.7D), which was also confirmed at the protein level by immu-nostaining (Fig. 7C). This finding suggests a potential interactionbetween CaSR and Klotho. Dual deletion of Klotho and CaSRresulted in merely undetectable Klotho expression (Fig. 7C). Themoderately to severely elevated serum PTH in PTHCre;CaSRfl/fl

and PTHCre;KLfl/fl;CaSRfl/fl mice might relate to Klotho genedosage (Fig. 1F). We next intended to avoid the high serumcalcium levels in PTHCre;CaSRfl/fl and PTHCre;KLfl/fl;CaSRfl/fl

mice that could contribute to down-regulation of Klotho ex-pression in the PTGs. We therefore performed PTG ex vivocultures. The isolated glands of PTHCre;Tmfl/+ and PTHCre;Tmfl/+;KLfl/fl mice were subjected to low-Ca2+ (0.5 mM Ca2+) orhigh-Ca2+ (3 mM Ca2+) conditions. qRT-PCR results showedthat high Ca2+ had no effect on Klotho gene expression in controlPTGs (Fig. 7G), suggesting that the deletion of the CaSR itself isresponsible for the reduced Klotho expression in PTGs, andthat this down-regulation is independent of serum calcium

Fig. 4. Effects of calcium-deficient diet on PTH expression. (A and B) Low-Ca2+ diet significantly reduced serum calcium and phosphate levels. (C) Serum PTHlevels were significantly up-regulated in control mice with hypocalcemia, whereas the increase was not as high as the controls in PTHcre;Tmfl/+;KLfl/fl andPTHcre;Tmfl/+;CaSRfl/+ mice. PTHcre;Tmfl/+;KLfl/fl;CaSRfl/+ mice had the highest PTH levels on a low-Ca2+ diet. ****P < 0.0001 versus control diet in each ge-notype. ###P < 0.001 versus PTHcre;Tmfl/+;KLfl/fl mice in low-Ca2+ diet; $P < 0.05 versus PTHcre;Tmfl/+;CaSRfl/+ mice in low-Ca2+ diet. (D) PTH immunofluorescentstaining of control and mutant PTGs on control diet and low-Ca2+ diet. n = 3–6. (E) Illustration of ex vivo PTG culture experiments procedure. (F) Ex vivocultured PTG showed low-Ca2+ diet stimulated PTH excretion, with more pronounced increase in parathyroid Klotho deletion mice. n = 6. *P < 0.05, ***P <0.001 versus high-calcium media (3 mM) in each genotype. [Scale bars, 50 μm (D) and 1 mm (E).]

E3754 | www.pnas.org/cgi/doi/10.1073/pnas.1717754115 Fan et al.

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 7: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

levels. We investigated the effect of CaSR deletion on Klothoexpression at a later stage in 6-wk-old mice. Noteworthy, Klothoexpression was markedly decreased in PTHCre;Tmfl/+;CaSRfl/+

PTGs compared with controls (Fig. 7E). Concurrently, CaSRexpression was also found to be significantly reduced in Klotho-deficient PTGs at 6 wk (Fig. 7F).Furthermore, PTHCre;KLfl/fl;CaSRfl/+ mice had slightly higher

serum PTH values compared with PTHCre;KLfl/fl. A tendencytoward increased PTH levels was observed in PTHCre;KLfl/+;CaSRfl/fl compared with PTHCre;CaSRfl/fl. This was correlatedwith serum Ca2+ levels, suggesting Klotho and CaSR togethercontrol PTH synthesis and deletion of only one allele of eitherprotein resulted in more severe phenotype (Fig. S7 A and B).

Moreover, we have compared the serum PTH and Ca2+ values ofcontrol animals and mice with Klotho ablation alone, with orwithout heterozygous CaSR deletion, at 6-wk of age (Fig. S7 Cand D). The result was similar to what we observed at P10: Se-rum PTH levels exhibited a tendency to increase among theseanimals, confirming the synergistic effect of Klotho and CaSR incontrolling PTH synthesis.We further investigated the possibility of an interaction be-

tween Klotho and CaSR using coimmunoprecipitation assays.HEK293 cells were transfected with Klotho alone, CaSR alone,or Klotho+CaSR. Klotho and CaSR were then precipitated byKlotho (or Flag) or CaSR (or GFP) antibodies, respectively. Theresults demonstrated that Klotho coimmunoprecipitates with

Fig. 5. Klotho and CaSR regulate PTG hyperplasia under chronic hypocalcemia. (A and B) Tomato fluorescence and size calculations of PTGs from 6-wk-oldmice on a control diet or low-Ca2+ diet. Parathyroid Klotho deletion or heterozygous deletion of CaSR led to increased PTG volume on control diet. The PTGsize was further enlarged upon onset of chronic hypocalcemia. Two individual glands are shown as representative in the low Ca2+ diet groups of PTHcre;Tmfl/+;CaSRfl/+ and PTHcre;Tmfl/+;KLfl/fl;CaSRfl/+ mice. (Scale bar, 500 μm.) (C) H&E staining of PTGs from 6-wk-old control and mutant mice. (Scale bar, 50 μm.) *P < 0.05,**P < 0.01 versus control diet in each genotype.

Fig. 6. Effect of Klotho and CaSR deletion on PTG cell proliferation under chronic hypocalcemia. (A and B) Immunohistochemical staining of Ki67 andcalculation of Ki67-positive cell ratio in PTGs on control or low-Ca2+ diet at 6 wk of age. Klotho or CaSR deletion, as well as low calcium conditions, increasedparathyroid cell proliferation, with the highest elevation of Ki67-positive cells in Klotho-deficient PTGs. n = 3–6. (Scale bar, 25 μm.) ***P < 0.001 versus controldiet in each genotype.

Fan et al. PNAS | vol. 115 | no. 16 | E3755

MED

ICALSC

IENCE

SPN

ASPL

US

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 8: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

CaSR (Fig. 7H), suggesting a physical interaction between Klothoand CaSR.

DiscussionIn the current study we generated mice with a PTG-specificdeletion of Klotho alone or together with CaSR to better un-derstand their functional mechanisms and determine any po-tential interactions. Deleting Klotho in PTGs did not affect keyserum and urinary parameters. PTHCre;CaSRfl/fl mice displayedhypercalcemia, hypophosphatemia, and significantly elevatedserum PTH, iFGF23, and 1,25(OH)2D3 levels, resembling thephenotype of CaSR−/− mice (42). Combined deletion of Klothoand CaSR (PTHCre;KLfl/fl;CaSRfl/fl) resulted in significantlyhigher serum PTH, iFGF23, and 1,25(OH)2D3 levels comparedwith PTHCre;CaSRfl/fl mice. There are several ways to interpretthese data. First, Klotho deletion could impede the FGF23feedback loop that suppresses PTH in PTGs. The moderately or

severely elevated serum PTH observed in CaSR-deletion versusCaSR+Klotho-deletion mice might depend on Klotho gene-dosage effects. Second, PTG hyperplasia developed in bothmutants, but mice with dual deletion exhibited the largest effect.Pth mRNA levels were comparable between these mice, but thegreater hyperplasia in PTHCre;KLfl/fl;CaSRfl/fl mice might lead toan increased parathyroid cell number and thus enhanced PTHsynthesis and higher serum PTH levels. Third, PTHCre;KLfl/fl;CaSRfl/fl PTGs exhibited a more severe alteration of PTG mor-phology, accompanied by increased PTG nodule formations,which might also lead to higher PTH secretion. These resultscollectively indicate that Klotho is a negative regulator of PTHsynthesis in the absence of CaSR. Serum PTH levels remainedconstant in PTHcre;Klothofl/fl mice, suggesting that the CaSR hasa dominant function in modulating PTH production. Klotho mayserve as a supplementary factor in response to changes in serumcalcium, especially when CaSR function is diminished.

Fig. 7. Klotho and CaSR interaction in PTGs. (A) CaSR immunofluorescent staining showed a trend toward decreased expression levels of CaSR in Klotho-deficient PTGs at P10. n = 6. (Scale bar, 50 μm.) (B) A trend toward decreased CaSR gene expression was observed in Klotho-deficient PTGs at P10. n = 6–10.(C and D) Immunohistochemical staining and gene-expression analysis showed CaSR ablation caused a significant reduction in Klotho expression at proteinand transcript levels. n = 6–10. (Scale bar, 50 μm.) (E) Klotho immunostaining of 6-wk-old control and mutant mice showed a significant decrease in Klothoexpression in PTHcre;Tmfl/+;CaSRfl/+ mice. Efficient deletion of Klotho was detected in PTHcre;Tmfl/+;KLfl/fl and PTHcre;Tmfl/+;KLfl/fl;CaSRfl/+ PTGs. n = 6. (Scalebar, 25 μm.) (F) CaSR immunostaining also suggested reduced CaSR expression in parathyroid Klotho-deleted mice. Heterozygous parathyroid ablation ofCaSR confirmed reduced CaSR expression levels. n = 6. (Scale bar, 25 μm.) (G) Ex vivo cultured PTGs of control and PTHcre;Tmfl/+;KLfl/fl mice revealed differentcalcium levels have no effect on Klotho expression. n = 6. (H) Coimmunoprecipitation assay was performed on HEK293 cells overexpressing Flag-Klotho alone,GFP-CaSR alone, or both together. pRK5-GFP-CaSR and pCS-mKLcFT plasmids were used. Proteins were collected and immunoprecipitated with Klotho, Flag,CaSR, or GFP antibodies, respectively. Western blot was performed using Klotho or GFP antibodies to determine the interaction between Klotho and CaSR.The results revealed that Klotho and CaSR bind to each other. IP, immunoprecipitation. ***P < 0.001, ****P < 0.0001 versus control; ##P < 0.01 versus PTHCre;Tmfl/+;KLfl/fl.

E3756 | www.pnas.org/cgi/doi/10.1073/pnas.1717754115 Fan et al.

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 9: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

Renal Klotho is known to enhance renal Ca2+ absorption bystabilizing the transient receptor potential vanilloid 5 acting as aβ-glucuronidase (43) and is an important factor in the preventionof renal Ca2+ loss (44). We investigated the action of parathyroidKlotho in the proper response to serum calcium. A previous studyusing single injections of EGTA or calcium gluconate to changeserum calcium levels demonstrated that PTG Klotho is not es-sential for PTH secretion in response to acute alterations in serumcalcium (31). However, we challenged mice with a Ca2+-deficientdiet for 3 wk to investigate the physiological role of Klotho andCaSR during chronic hypocalcemia. The increase in serum PTHwas less in PTHcre;Tmfl/+;KLfl/fl and PTHcre;Tmfl/+;CaSRfl/+ micecompared with control mice under low-Ca2+ condition. This resultis in accord with a previous report that showed homozygousKlotho knockout mice did not secrete as much PTH as controlmice under hypocalcemia (34). Interestingly, ex vivo experimentsshowed that Klotho-deficient PTGs had more pronounced secre-tion of PTH under low-Ca2+ condition void of the influence byother circulating components, suggesting an independent role forKlotho in suppressing PTH secretion. Therefore, we speculatethat the inconsistency of in vivo and ex vivo PTH production mightbe due to some systemic inhibitory factors preventing PTH se-cretion in PTHcre;Tmfl/+;KLfl/fl mice under chronic hypocalcemia.Thus, additional investigations are required to determine the invivo regulatory network on PTH secretion.Accumulating evidence suggests that a chronic increase in

PTH production, characteristic of primary or secondary hyper-parathyroidism, is accompanied by an increase in PTG size (45–47). Determination of gland size is particularly difficult (48), sowe generated mice with a Tomato reporter gene in which redfluorescent protein is selectively expressed in parathyroid cells.Mice with ablation of PTG-CaSR at P10 or heterozygous PTG-specific deletion of CaSR at 6 wk of age had significantly en-larged glands compared with controls, confirming the pivotalfunction of CaSR in determining PTG growth. Klotho expressionis reduced as the PTG tissue becomes hyperplastic (19), as seenin patients with hyperphosphatemic familial tumoral calcinosis(49). It was unclear whether Klotho was a contributing factor tohyperplasia, a consequence of hyperplasia, or some combinationof the two. Our results revealed that Klotho has a more prom-inent role in the development of hyperplasia and/or SHPT. PTG-specific Klotho deletion only leads to a trend of increasing PTGsize at P10. However, the combined deletion of Klotho andCaSR caused significant PTG hyperplasia. The additional glandenlargement in PTHCre;KLfl/fl;CaSRfl/fl mice compared withPTHCre;CaSRfl/fl mice indicates a role of Klotho in preventingthe development of hyperplasia in the absence of a functionalCaSR. Interestingly, cell proliferation appears to be more im-portant than cell apoptosis to the increase in PTG volume. It isimportant to note that the critical role of Klotho in preventingPTG hyperplasia is demonstrated in our prolonged observationthat mice with PTG-specific Klotho ablation at 6 wk of ageexhibited a substantial increase in PTG size compared with controland PTHcre;Tmfl/+;CaSRfl/+ mice on a control diet. These resultsemphasize Klotho’s function in suppressing PTG hyperplasia.An increase in PTG volume, reportedly due to enhanced hy-

pertrophy, has been observed in hypocalcemic animals withnormal renal function on a calcium-deficient diet (50). Anotherfinding showed that parathyroid cell hyperplasia largely pre-vailed over hypertrophy in rats on a low-calcium diet (51).Analyses of human hyperparathyroidism samples found evidenceof apoptosis (52, 53). The percentage of PTG apoptotic cells incontrols was ∼0.15% in P10 mice and less than 0.1% in 6-wk-old

mice, but PTG-specific ablation of CaSR led to significantly in-creased cell apoptosis. This was more evident in dual deletion ofCaSR and Klotho at P10. Nevertheless, cell proliferation pre-vailed over apoptosis in these mice. Enhanced cell proliferationwas observed under chronic hypocalcemia, and cell apoptosiswas not altered in 6-wk-old mice. Thus, the higher rate of cellproliferation over apoptosis results in the enlargement of PTGvolume, similar to the observations in hyperplastic PTG tissue ofuremic patients (53, 54). We also observed that in some extremelyhyperplastic PTGs from PTHCre;KLfl/fl;CaSRfl/fl mice, PTH wasnot expressed in some areas of altered structure, probably due toincreased apoptosis. We found that an enlargement of PTG sizewas detected in all groups under hypocalcemia and revealed thathypocalcemia significantly enhanced cell proliferation. This wasmore pronounced in mice deficient in PTG Klotho, confirmingthat reduced Klotho expression could direct PTG hyperplasia bymediating cell proliferation.Previous studies showed Klotho and CaSR expression were

significantly decreased in PTGs of PHTP, SHPT patients, pa-tients after kidney transplantation, and those with end-stage re-nal disease (55–57). We could demonstrate a binding interactionbetween Klotho and CaSR that might play a synergistic effect incontrolling PTH synthesis and glandular hyperplasia. Severallines of experimental evidence support this tenet. PTG-specificdeletion of CaSR, for example, leads to significantly reducedKlotho transcript and protein levels at P10. Moreover, evenheterozygous PTG CaSR ablation resulted in markedly reducedKlotho expression. Ex vivo PTG culture revealed that Klothoexpression was not affected by changes in calcium levels, in-dicating that the observed reduction in Klotho expression in vivowas largely due to CaSR ablation. On the contrary, PTG Klothodeletion significantly reduced CaSR expression at 6 wk of age. Atrend toward decreased CaSR expression was already noted atP10 in PTHCre;KLfl/fl PTGs. Moreover, we were able to dem-onstrate that Klotho and CaSR bind to each other using coim-munoprecipitation. Technical issues related to the small size ofthe glands limited further investigations of in vivo coimmuno-precipitation experiments. However, we were able to show pro-tein colocalization of CaSR and Klotho by immunostaining (Fig.S7E), consistent with these two proteins acting together to me-diate PTH synthesis and PTG growth.In summary, we demonstrated that the specific deletion of

CaSR in the PTGs leads to elevated serum PTH and PTG hy-perplasia, and that additional deletion of Klotho in PTGs exac-erbated this condition. This suggests a pivotal function for Klothoin suppressing PTH biosynthesis and PTG growth in the absenceof CaSR. Moreover, we were able to demonstrate that Klothoexhibits an independent role to modulate PTH production underchronic hypocalcemia and acts as an inhibitory factor on para-thyroid cell proliferation, indicating a physiological function forKlotho in modulating the hyperplasia or SHPT. Most importantly,our findings propose a physical interaction between Klotho andCaSR. Nevertheless, the molecular mechanism by which Klothoand CaSR regulate each other requires further investigation. Theresults suggest that up-regulation or activation of Klotho couldprovide a potential treatment to control circulating PTH andhyperparathyroidism.

ACKNOWLEDGMENTS.We thank Dr.Wenhan Chang for providing themethodof ex vivo PTG culture and the Center for Skeletal Research Core for the bonehistology analysis. This work was supported by NIH Grant DK097105 (to B.L.),Harvard School of Dental Medicine (Y.F.), State Key Laboratory of OralDiseases Open Funding Grant SKLOD2016OF01, and Postdoctoral Foundationof Sichuan University Grant 2017SCU12052 (to Y.F.).

1. GBD 2015 Mortality and Causes of Death Collaborators (2016) Global, regional, and

national life expectancy, all-cause mortality, and cause-specific mortality for 249

causes of death, 1980-2015: A systematic analysis for the Global Burden of Disease

Study 2015. Lancet 388:1459–1544.

2. Saran R, et al. (2017) US Renal Data System 2016 annual data report: Epidemiology of kidney

disease in the United States. Am J Kidney Dis 69(Suppl 1):A7–A8, and erratum (2017) 69:712.3. Goodman WG, Quarles LD (2008) Development and progression of secondary hyperpara-

thyroidism in chronic kidney disease: Lessons frommolecular genetics. Kidney Int 74:276–288.

Fan et al. PNAS | vol. 115 | no. 16 | E3757

MED

ICALSC

IENCE

SPN

ASPL

US

Dow

nloa

ded

by g

uest

on

July

20,

202

0

Page 10: Interrelated role of Klotho and calcium-sensing receptor ...factors, use a combination of medications to balance multiple factors, or surgically remove the PTGs (6, 7). A deeper under-standing

4. Moe SM, et al. (2011) The pathophysiology of early-stage chronic kidney disease-mineral bone disorder (CKD-MBD) and response to phosphate binders in the rat.J Bone Miner Res 26:2672–2681.

5. Gutiérrez OM, et al. (2008) Fibroblast growth factor 23 and mortality among patientsundergoing hemodialysis. N Engl J Med 359:584–592.

6. Stracke S, Keller F, Steinbach G, Henne-Bruns D, Wuerl P (2009) Long-term outcomeafter total parathyroidectomy for the management of secondary hyperparathyroid-ism. Nephron Clin Pract 111:c102–c109.

7. Reichel H (2006) Current treatment options in secondary renal hyperparathyroidism.Nephrol Dial Transplant 21:23–28.

8. Brown EM, MacLeod RJ (2001) Extracellular calcium sensing and extracellular calciumsignaling. Physiol Rev 81:239–297.

9. Brown EM, et al. (1993) Cloning and characterization of an extracellular Ca(2+)-sensingreceptor from bovine parathyroid. Nature 366:575–580.

10. Lips P (2006) Vitamin D physiology. Prog Biophys Mol Biol 92:4–8.11. Lanske B, Razzaque MS (2014) Molecular interactions of FGF23 and PTH in phosphate

regulation. Kidney Int 86:1072–1074.12. Rhee Y, et al. (2011) Parathyroid hormone receptor signaling in osteocytes increases

the expression of fibroblast growth factor-23 in vitro and in vivo. Bone 49:636–643.13. Meir T, et al. (2014) Parathyroid hormone activates the orphan nuclear receptor

Nurr1 to induce FGF23 transcription. Kidney Int 86:1106–1115.14. Fan Y, et al. (2016) Parathyroid hormone 1 receptor is essential to induce

FGF23 production and maintain systemic mineral ion homeostasis. FASEB J 30:428–440.

15. Ben-Dov IZ, et al. (2007) The parathyroid is a target organ for FGF23 in rats. J ClinInvest 117:4003–4008.

16. Silver J, Russell J, Sherwood LM (1985) Regulation by vitamin D metabolites of mes-senger ribonucleic acid for preproparathyroid hormone in isolated bovine para-thyroid cells. Proc Natl Acad Sci USA 82:4270–4273.

17. Kemper B, Habener JF, Rich A, Potts JT, Jr (1974) Parathyroid secretion: Discovery of amajor calcium-dependent protein. Science 184:167–169.

18. Krajisnik T, et al. (2010) Parathyroid Klotho and FGF-receptor 1 expression declinewith renal function in hyperparathyroid patients with chronic kidney disease andkidney transplant recipients. Kidney Int 78:1024–1032.

19. Komaba H, et al. (2010) Depressed expression of Klotho and FGF receptor 1 in hy-perplastic parathyroid glands from uremic patients. Kidney Int 77:232–238, and er-ratum (2010) 77:834.

20. Kuro-o M, et al. (1997) Mutation of the mouse Klotho gene leads to a syndromeresembling ageing. Nature 390:45–51.

21. Hu MC, et al. (2010) Klotho: A novel phosphaturic substance acting as an autocrineenzyme in the renal proximal tubule. FASEB J 24:3438–3450.

22. Lim K, et al. (2015) α-Klotho expression in human tissues. J Clin Endocrinol Metab 100:E1308–E1318.

23. Urakawa I, et al. (2006) Klotho converts canonical FGF receptor into a specific receptorfor FGF23. Nature 444:770–774.

24. Kurosu H, et al. (2006) Regulation of fibroblast growth factor-23 signaling by Klotho.J Biol Chem 281:6120–6123.

25. Chen G, et al. (2018) α-Klotho is a non-enzymatic molecular scaffold forFGF23 hormone signalling. Nature 553:461–466.

26. Wolf M (2012) Update on fibroblast growth factor 23 in chronic kidney disease.Kidney Int 82:737–747.

27. Shimada T, et al. (2001) Cloning and characterization of FGF23 as a causative factor oftumor-induced osteomalacia. Proc Natl Acad Sci USA 98:6500–6505.

28. Shimada T, et al. (2004) Targeted ablation of Fgf23 demonstrates an essential phys-iological role of FGF23 in phosphate and vitamin D metabolism. J Clin Invest 113:561–568.

29. Larsson T, et al. (2004) Transgenic mice expressing fibroblast growth factor 23 underthe control of the alpha1(I) collagen promoter exhibit growth retardation, osteo-malacia, and disturbed phosphate homeostasis. Endocrinology 145:3087–3094.

30. Krajisnik T, et al. (2007) Fibroblast growth factor-23 regulates parathyroid hormoneand 1alpha-hydroxylase expression in cultured bovine parathyroid cells. J Endocrinol195:125–131.

31. Olauson H, et al. (2013) Parathyroid-specific deletion of Klotho unravels a novelcalcineurin-dependent FGF23 signaling pathway that regulates PTH secretion. PLoSGenet 9:e1003975.

32. DeLuca S, et al. (2008) Amelioration of the premature ageing-like features of Fgf-23 knockout mice by genetically restoring the systemic actions of FGF-23. J Pathol 216:345–355.

33. Kawakami K, et al. (2017) Persistent fibroblast growth factor 23 signalling in theparathyroid glands for secondary hyperparathyroidism in mice with chronic kidneydisease. Sci Rep 7:40534.

34. Imura A, et al. (2007) alpha-Klotho as a regulator of calcium homeostasis. Science 316:1615–1618.

35. Martuseviciene G, Hofman-Bang J, Clausen T, Olgaard K, Lewin E (2011) The secretoryresponse of parathyroid hormone to acute hypocalcemia in vivo is independent ofparathyroid glandular sodium/potassium-ATPase activity. Kidney Int 79:742–748.

36. Libutti SK, et al. (2003) Parathyroid gland-specific deletion of the mouse Men1 generesults in parathyroid neoplasia and hypercalcemic hyperparathyroidism. Cancer Res63:8022–8028.

37. Toka HR, et al. (2012) Deficiency of the calcium-sensing receptor in the kidney causesparathyroid hormone-independent hypocalciuria. J Am Soc Nephrol 23:1879–1890.

38. Toka HR (2014) New functional aspects of the extracellular calcium-sensing receptor.Curr Opin Nephrol Hypertens 23:352–360.

39. Sato T, et al. (2017) Parathyroid hormone controls paracellular Ca2+ transport in thethick ascending limb by regulating the tight-junction protein Claudin14. Proc NatlAcad Sci USA 114:E3344–E3353.

40. Bouschet T, Martin S, Henley JM (2005) Receptor-activity-modifying proteins are re-quired for forward trafficking of the calcium-sensing receptor to the plasma mem-brane. J Cell Sci 118:4709–4720.

41. Ravikumar P, et al. (2014) α-Klotho protects against oxidative damage in pulmonaryepithelia. Am J Physiol Lung Cell Mol Physiol 307:L566–L575.

42. Ho C, et al. (1995) A mouse model of human familial hypocalciuric hypercalcemia andneonatal severe hyperparathyroidism. Nat Genet 11:389–394.

43. Chang Q, et al. (2005) The beta-glucuronidase Klotho hydrolyzes and activates theTRPV5 channel. Science 310:490–493.

44. Alexander RT, et al. (2009) Klotho prevents renal calcium loss. J Am Soc Nephrol 20:2371–2379.

45. Drüeke TB (1995) The pathogenesis of parathyroid gland hyperplasia in chronic renalfailure. Kidney Int 48:259–272.

46. Parfitt AM (1997) The hyperparathyroidism of chronic renal failure: A disorder ofgrowth. Kidney Int 52:3–9.

47. Silver J, Sela SB, Naveh-Many T (1997) Regulation of parathyroid cell proliferation.Curr Opin Nephrol Hypertens 6:321–326.

48. Bi R, et al. (2016) Diphtheria toxin- and GFP-based mouse models of acquired hypo-parathyroidism and treatment with a long-acting parathyroid hormone analog.J Bone Miner Res 31:975–984.

49. Ichikawa S, et al. (2007) A homozygous missense mutation in human KLOTHO causessevere tumoral calcinosis. J Musculoskelet Neuronal Interact 7:318–319.

50. Szabo A, Merke J, Beier E, Mall G, Ritz E (1989) 1,25(OH)2 vitamin D3 inhibits para-thyroid cell proliferation in experimental uremia. Kidney Int 35:1049–1056.

51. Naveh-Many T, Rahamimov R, Livni N, Silver J (1995) Parathyroid cell proliferation innormal and chronic renal failure rats. The effects of calcium, phosphate, and vitaminD. J Clin Invest 96:1786–1793.

52. Wang W, Johansson H, Kvasnicka T, Farnebo LO, Grimelius L (1996) Detection ofapoptotic cells and expression of Ki-67 antigen, Bcl-2, p53 oncoproteins in humanparathyroid adenoma. APMIS 104:789–796.

53. Zhang P, et al. (2000) Apoptosis in parathyroid hyperplasia of patients with primary orsecondary uremic hyperparathyroidism. Kidney Int 57:437–445.

54. Drüeke TB (2000) Cell biology of parathyroid gland hyperplasia in chronic renalfailure. J Am Soc Nephrol 11:1141–1152.

55. Hong YA, Choi DE, Lim SW, Yang CW, Chang YK (2013) Decreased parathyroid Klothoexpression is associated with persistent hyperparathyroidism after kidney trans-plantation. Transplant Proc 45:2957–2962.

56. Latus J, et al. (2013) Analysis of α-Klotho, fibroblast growth factor-, vitamin-D andcalcium-sensing receptor in 70 patients with secondary hyperparathyroidism. KidneyBlood Press Res 37:84–94.

57. Latus J, et al. (2013) Involvement of α-Klotho, fibroblast growth factor-, vitamin-D-and calcium-sensing receptor in 53 patients with primary hyperparathyroidism.Endocrine 44:255–263.

E3758 | www.pnas.org/cgi/doi/10.1073/pnas.1717754115 Fan et al.

Dow

nloa

ded

by g

uest

on

July

20,

202

0