A molecular circadian clock operates in the parathyroid gland and is disturbed in chronic kidney disease associated bone and mineral disorder

Egstrand, Søren; Nordholm, Anders; Morevati, Marya; Mace, Maria L.; Hassan, Alia; Naveh- many, Tally; Rukov, Jakob L.; Gravesen, Eva; Olgaard, Klaus; Lewin, Ewa

Published in: Kidney International

DOI: 10.1016/j.kint.2020.06.034

Publication date: 2020

Document version Publisher's PDF, also known as Version of record

Document license: CC BY-NC-ND

Citation for published version (APA): Egstrand, S., Nordholm, A., Morevati, M., Mace, M. L., Hassan, A., Naveh-many, T., Rukov, J. L., Gravesen, E., Olgaard, K., & Lewin, E. (2020). A molecular circadian clock operates in the parathyroid gland and is disturbed in chronic kidney disease associated bone and mineral disorder. Kidney International, 98(6), 1461-1475. https://doi.org/10.1016/j.kint.2020.06.034

Download date: 26. sep.. 2021 www.kidney-international.org basic research

A molecular circadian clock operates in the parathyroid gland and is disturbed in chronic OPEN kidney disease associated bone and mineral disorder Søren Egstrand1,2, Anders Nordholm1,2, Marya Morevati2, Maria L. Mace2, Alia Hassan3, Tally Naveh-Many3, Jakob L. Rukov2, Eva Gravesen2, Klaus Olgaard2 and Ewa Lewin1,2

1Nephrological Department, Herlev Hospital, University of Copenhagen, Copenhagen, Denmark; 2Nephrological Department, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark; and 3Minerva Center for Calcium and Bone Metabolism, Nephrology Services, Hadassah Hebrew University Medical Center, Jerusalem, Israel

Circadian rhythms in metabolism, hormone secretion, cell Kidney International (2020) 98, 1461–1475; https://doi.org/10.1016/ cycle and locomotor activity are regulated by a molecular j.kint.2020.06.034 circadian clock with the master clock in the suprachiasmatic KEYWORDS: aorta; CKD; secondary hyperparathyroidism; PTH; vascular nucleus of the central nervous system. However, an internal calcification Copyright ª 2020, International Society of Nephrology. Published by clock is also expressed in several peripheral tissues. Elsevier Inc. This is an open access article under the CC BY-NC-ND license Although about 10% of all genes are regulated by clock (http://creativecommons.org/licenses/by-nc-nd/4.0/). machinery an internal molecular circadian clock in the parathyroid glands has not previously been investigated. Parathyroid hormone secretion exhibits a diurnal variation Translational Statement and parathyroid hormone gene promoter contains an E- box like element, a known target of circadian clock Circadian rhythms in metabolism, hormone secretion, proteins. Therefore, we examined whether an internal cell cycle, and locomotor activity are regulated by a molecular circadian clock is operating in parathyroid molecular circadian clock. Disruption of the circadian glands, whether it is entrained by feeding and how it rhythm may cause serious health problems and responds to chronic kidney disease. As uremia is associated abnormal cell growth. Secondary hyperparathyroidism is with extreme parathyroid growth and since disturbed a severe complication of chronic kidney disease associ- circadian rhythm is related to abnormal growth, we ated with increased morbidity and mortality, and the control of parathyroid hyperplasia is a clinical challenge. examined the expression of parathyroid clock and clock- This article describes the existence of an internal mo- regulated cell cycle genes in parathyroid glands of normal lecular circadian clock that is operating in the para- and uremic rats. Circadian clock genes were found to be thyroid gland and which is disturbed in chronic kidney rhythmically expressed in normal parathyroid glands and disease. Deregulation of the circadian clock in the this clock was minimally entrained by feeding. Diurnal parathyroids might therefore potentially call for the use regulation of parathyroid glands was next examined. of chronotherapy in future clinical studies. Significant rhythmicity of fibroblast-growth-factor- receptor-1, MafB and Gata3 was found. In uremic rats, deregulation of circadian clock genes and the cell cycle ircadian rhythms are controlled by the molecular clock regulators, Cyclin D1, c-Myc, Wee1 and p27, which are mechanisms and have a systemic impact on all aspects influenced by the circadian clock, was found in parathyroid C of physiology, including hormonal secretion, meta- glands as well as the aorta. Thus, a circadian clock operates bolism, and cell cycle. Core clock genes and their products are in parathyroid glands and this clock and downstream cell involved in transcriptional-translational feedback loops and cycle regulators are disturbed in uremia and may are revealed as important regulators of organ function. The contribute to dysregulated parathyroid proliferation in possible existence of an internal circadian clock machinery secondary hyperparathyroidism. in the parathyroid gland has not previously been examined. Calcium homeostasis is maintained primarily by the parathy- roid glands.1 The parathyroid gland senses minor changes in þ extracellular ionized calcium (Ca2 ) via the calcium-sensing Correspondence: Ewa Lewin, Nephrological Department B 109, Herlev Hos- receptor (CaSR) and secretes parathyroid hormone (PTH) 2,3 pital, DK 2730 Herlev, Copenhagen, Denmark. E-mail: [email protected] or in response to hypocalcemia. PTH acts primarily on bone [email protected] to release calcium from bone storage and on the kidney to Received 5 January 2020; revised 17 June 2020; accepted 18 June 2020; enhance renal calcium reabsorption and increase published online 25 July 2020 1,25(OH)2-vitamin D synthesis, which increases intestinal

Kidney International (2020) 98, 1461–1475 1461 basic research S Egstrand et al.: Circadian clock in the parathyroid gland

þ calcium absorption. In addition to extracellular Ca2 , the of targets: induction of Wee1 for inhibition of Cyclin B1 at the 30 parathyroid gland also responds to changes in plasma phos- G2/M (Gap2/Mitosis) transition, induction of p20 and p21 31,32 phate (p-P), recently shown to be mediated via an effect of at the G1/S (Gap1/Synthesis) transition, and regulation of 4 33–36 P on the CaSR, activation of the vitamin D receptor, fibro- p27, Cyclin D1, and c-Myc for progression through G1. blast growth factor 23 (FGF23) via fibroblast growth factor Cell turnover rate is very low in the parathyroid gland receptor 1 (FGFR1) in combination with FGFR1 cofactor under normal conditions; with an estimated mean life – Klotho and recently proposed to be affected by leptin.4 8 span of normal parathyroid cells being 19 years in PTH and FGF23 are phosphaturic hormones reducing P reab- humans and 2 years in rats.37 This rate increases sorption in the renal proximal tubules.9 PTH secretion ex- dramatically in secondary hyperparathyroidism (sHPT) hibits a diurnal variation with acrophase during the inactive caused by chronic kidney disease (CKD). sHPT is a period in both humans and rodents.10,11 The parathyroid frequent cause of clinically significant bone disease and glands are not controlled by a superior “hypothalamic-pitui- soft-tissue and vascular calcification. sHPT in CKD oc- tary-axis” as seen in some endocrine glands and are likely to curs with a combination of functional and structural use other, hitherto unknown, regulatory mechanisms for sus- changes in the parathyroid glands that are resulting in – taining a circadian variation in hormone secretion.12 14 PTH increased PTH biosynthesis, secretion, and in parathyroid gene promotor contains an E-box–like motif in both humans hyperplasia.38 and rodents.7 E-box motifs are known targets of the core MafB, Gata3, and glial cells missing homolog 2 (Gcm2) are circadian clock proteins, BMAL1 (Brain and Muscle indispensable transcription factors in the development of the ARNTL-like 1) and CLOCK.15 parathyroid gland,39,40 but continue to influence parathyroid The circadian clock is an internal, self-sustaining pace- functions into adulthood: Gcm2 was found to maintain maker that operates with a periodicity of 24 hours to coor- parathyroid cell mass and proliferation in adult mice41 and dinate proper rhythms in behavior, hormone secretion, both Gcm2 and MafB may play a role in sHPT.42,43 metabolism, and cell cycle.16 The suprachiasmatic nucleus in The present study is the first to investigate the internal mo- hypothalamus generates the central circadian rhythm essential lecular circadian clock in the parathyroid gland and whether for directing cycles of locomotor activity and coordinating the such an internal molecular circadian clock is related to the peripheral and central clocks via neuronal, hormonal, and circadian rhythm of p-PTH and parathyroid function. metabolic signaling pathways. However, in addition to this Furthermore, we examine the potential role of deregulation of central pacemaker, a molecular clock has been found in a such a circadian clock in the development of sHPT in CKD– number of peripheral tissues.17 At the core of the circadian mineral and bone disorder (MBD) and expand the impact of clock is the activator BMAL1, which combines with either deregulation of the clock in CKD-MBD to uremic vasculopathy. CLOCK or NPAS2 for binding to E-box motifs to induce transcription of the repressor genes Period 1-3 (Per1-3) and RESULTS 18 Cryptochrome 1-2 (Cry1-2). The PERIOD and CRYPTO- An internal molecular circadian clock operates in the CHROME proteins then translocate to the nucleus and parathyroid gland interact with BMAL1/CLOCK to repress their own tran- Rats kept in 12:12 hour light-dark cycle and fed ad libitum scription and thereby completing an autoregulatory were sacrificed at 4-hour intervals for 24 hours and para- transcriptional-translational feedback loop.19 An extra feed- thyroid glands were investigated for gene and protein back loop involves BMAL1 induction of REV-ERBs20 and expression of core clock components (protocol 1). Rhyth- RORs, which in turn regulate Bmal1 transcription. In addi- micity of the gene expression was assessed by cosinor analysis, tion, posttranslational modification regulates the clock ma- fitted to a periodicity of 24 hours and was significant for chinery. For instance, a phosphorylation-dephosphorylation Bmal1 (P < 0.0001), Npas2 (P < 0.0001), Per1 (P < 0.02), cycle of CK1ε regulates the activity of the Period genes by Per2 (P < 0.0001), Per3 (P < 0.0001), Cry1-2 (P < 0.0001), hindering nuclear translocation and promoting degradation and Rev-Erba (P < 0.001). No significant circadian rhythm of PER/CRY complexes.21 Similarly, CLOCK undergoes was found for CK1ε and Clock (Figure 1). phosphorylation in a circadian manner, which is coupled to JTK_CYKLE (Hughes Lab, Pulmonary and Critical Care nuclear translocation and the subsequent degradation of the Medicine, Washington University School of Medicine, St. protein.22 CK1ε does not cycle but is constitutively expressed, Louis, MO) analysis confirmed the significant circadian whereas Clock only cycles in some tissues.23 The circadian rhythmicity of core clock genes (Table 1). clock regulates an estimated 8% to 10% of all genes expressed Western blot showed significant circadian rhythmicity of in peripheral tissues.24,25 Regulation by different constituents core clock protein BMAL1 in the parathyroid glands (P < of the molecular clock occurs at various levels from tran- 0.05) with a phase delay of approximately 4 hours compared – scription to secretion and is cell-specific.22,26 28 The circadian with that of the gene expression profile of Bmal1 clockwork is entrained by environmental cues. Entrainment (Supplementary Figure S1). Further support for an oper- signals, which differ between tissues, are primarily blue light ating circadian clock on the protein level was found in our input via retina or feeding signals.29 The circadian clock has phosphoproteomics study with CK1E and CLOCK identified been implicated in gating control of the cell cycle at a number in the rat parathyroid glands (data not shown).

1462 Kidney International (2020) 98, 1461–1475 S Egstrand et al.: Circadian clock in the parathyroid gland basic research

Figure 1 | The circadian clock operates in the parathyroid gland. Expression profiles of core circadian clock genes in the parathyroid gland harvested every fourth hour for 24 hours (n ¼ 38). (a) Bmal1,(b) Clock,(c) Npas2,(d–f) Per1-3,(g,h) Cry1-2,(i) Rev-Erba, and (j) CK1ε. Dots represent each animal. Circadian rhythmicity was assessed by cosinor analysis (solid line). Resulting P values are shown; P < 0.05 considered significant. Gray areas indicate dark periods and white areas indicate light periods. Dashed line is the mesor. Zeitgeber time (ZT; “time-giver”)is time since light onset. Conc, concentration.

Plasma PTH has circadian rhythmicity, but Pth gene expression showed significant rhythmicity (P < 0.05), expression is not rhythmic whereas the expression of FGFR1 cofactor Klotho was not P-PTH exhibited diurnal variation as previously shown by us rhythmic. Likewise, Casr and Vdr had no significant rhyth- 10,11 and others, and its statistical significance was confirmed micity (Figure 2). by cosinor analysis (Figure 2a), revealing acrophase in the inactive period and nadir during the active period. Para- Regulators of cell cycle and proliferation of the parathyroid thyroid expression of the Pth gene showed no diurnal varia- gland show diurnal variation tion (Figure 2d). The gene expression of cell-cycle regulators under circadian clock influence was examined: Wee1, Cyclin D1, p27, and c- Expression profiles of parathyroid genes of importance for Myc (Figure 3). Wee1 expression showed significant rhyth- PTH secretion micity (P < 0.0005), as did Cyclin D1 (P < 0.03), whereas no The gene expression of important regulators of PTH secretion significant rhythmicity was found for p27 (Supplementary was examined: Casr, Vdr, Klotho (Kl), and Fgfr1. Fgfr1 Figure S2)orc-Myc.

Kidney International (2020) 98, 1461–1475 1463 basic research S Egstrand et al.: Circadian clock in the parathyroid gland

Table 1 | Circadian parameters of core clock genes estimated by JTK_CYCLE analysis in the parathyroids from controls, CKD rats, and rats exposed to restricted feeding Mesor (CI) Amplitude (CI) Acrophase (ZT) (CI) Period (h) P value (JTK) Control Bmal1 0.73 (0.58, 0.89) 0.66 (0.44, 0.88) 0.48 (23.27, 1.69) 24 <0.0001 Clock 0.79 (0.69, 0.89) 0.05 (0.09, 0.20) 8.65 (22.62, 18.68) 16 1 Npas2 0.88 (0.69, 1.08) 0.64 (0.36, 0.91) 0.20 (22.61, 1.79) 24 <0.0001 Per1 0.71 (0.58, 0.83) 0.24 (0.06, 0.42) 10.72 (8.01, 13.43) 24 0.57 Per2 0.89 (0.73, 1.05) 0.72 (0.49, 0.94) 14.82 (13.61, 16.03) 24 <0.0001 Per3 0.82 (0.66, 0.98) 0.64 (0.40, 0.87) 11.64 (10.28, 12.99) 24 <0.0001 Cry1 0.99 (0.86, 1.12) 0.61 (0.43, 0.79) 18.17 (16.98, 19.36) 20 <0.0001 Cry2 0.90 (0.79, 1.00) 0.42 (0.26, 0.57) 14.07 (12.67, 15.47) 24 0.0001 Rev-Erba 0.47 (0.29, 0.65) 0.62 (0.37, 0.88) 8.61 (7.06, 10.16) 24 <0.0001 CK1ε 0.78 (0.64, 0.92) 0.11 (0.09, 0.31) 15.52 (8.665, 22.38) 12 1 CKD Bmal1 0.79 (0.70, 0.87) 0.64 (0.52, 0.76) 1.22 (0.55, 1.89) 24 <0.0001 Clock 0.90 (0.81, 1.00) 0.14 (0.004, 0.28) 3.18 (23.46, 6.90) 12 0.27 Npas2 0.57 (0.49, 0.66)a 0.52 (0.40, 0.64) 2.89 (2.02, 3.76)a 24 <0.0001 Per1 1.08 (0.94, 1.22)a 0.19 (0.02, 0.40) 10.82 (6.84, 14.80) 12 0.16 Per2 1.41 (1.10, 1.72)a 0.96 (0.52, 1.40) 15.72 (13.92, 17.52) 20 <0.0001 Per3 1.11 (0.87, 1.34) 1.20 (0.86, 1.54) 14.13 (13.10, 15.16)a 20 <0.0001 Cry1 1.11 (0.92, 1.31) 0.71 (0.43, 0.99) 20.77 (19.44, 22.11)a 20 <0.0001 Cry2 1.10 (0.95, 1.25) 0.40 (0.17, 0.62) 13.38 (11.36, 15.40) 20 0.0005 Rev-Erba 1.27 (1.04, 1.50)a 1.76 (1.42, 2.09)a 9.5 (8.79, 10.21) 24 <0.0001 CK1ε 0.96 (0.83, 1.09) 0.12 (0.06, 0.29) 6.93 (0.59, 13.27) 16 0.017 RF Bmal1 0.68 (0.54, 0.81) 0.35 (0.16, 0.54) 23.82 (21.78, 25.86) 16 <0.0001 Clock 0.78 (0.67, 0.89) 0.03 (0.13, 0.18) 23.89 (0.00, 23.99) 16 1 Npas2 0.63 (0.51, 0.76) 0.15 (0.03, 0.34)a 20.94 (16.55, 1.33) 20 0.010 Per1 0.84 (0.73, 0.94) 0.21 (0.06, 0.37) 14.94 (12.17, 17.71) 24 0.46 Per2 1.04 (0.87, 1.21) 0.56 (0.32, 0.80) 13.67 (12.04, 15.30) 24 0.00093 Per3 1.21 (0.88, 1.55) 0.83 (0.35, 1.31) 9.74 (7.61, 11.87) 24 0.00039 Cry1 0.99 (0.80, 1.17) 0.36 (0.11, 0.61) 16.91 (14.10, 19.72) 24 0.055 Cry2 1.13 (0.85, 1.40) 0.67 (0.27, 1.06) 12.78 (10.58, 14.98) 24 0.019 Rev-Erba 0.68 (0.50, 0.87) 0.47 (0.21, 0.73) 6.13 (3.92, 8.34) 24 0.00051 CK1ε 0.84 (0.69, 0.98) 0.04 (0.17, 0.24) 4.54 (0.00, 23.99) 12 1 CI, confidence interval; CKD, chronic kidney disease; RF, restricted feeding; ZT, Zeitgeber time (measured in hours since lights on). Data shown as mean (95% CI). aThere is a significant difference when compared with control (P < 0.05). P value (JTK) is the adjusted P value reported from JTK_CYKLE analysis.

For the regulators of parathyroid proliferation and PTH blood samples were drawn, then feeding was restricted to the secretion, MafB, Gata3, and Gcm2, expression was examined light phase (ZT2–ZT12) for 4 weeks, blood samples were (Figure 3), and significant rhythmicity was demonstrated for repeated and parathyroid glands harvested every 4 hours for MafB (P < 0.003) and Gata3 (P < 0.03), but not for Gcm2. 24 hours and gene expression examined (protocol 2). Sig- nificant rhythmicity was confirmed for p-PTH with compa- The parathyroid secretory response to induction of acute rable acrophase timing to that found in protocol 1 when rats hypocalcemia does not vary according to the time of the day were fed ad libitum (P < 0.0002) (Figure 5a). Significant Parathyroid function was assessed by acute induction of hypo- rhythmicity was found also for p-P (P < 0.0001) (Figure 5b) calcemia in vivo in the rat (protocol 3).3,12,13 Ethylene- and p-urea (P < 0.0001) (data not shown). P-Ca (Figure 5d) bis(oxyethylene-nitrilo)tetraacetic acid infused into the femoral and creatinine (data not shown) remained constant during vein resulted in progressive hypocalcemia starting at either ZT4 the day. Feeding restricted to the inactive phase markedly (light phase) or ZT16 (dark phase). ZT refers to time since light shifted the phase of p-PTH and p-P (Figure 5a and b) and onset (zeitgeber time; “time-giver”). Hypocalcemia was induced urea (not shown), now exhibiting significant circadian pattern at similar rates in the 2 groups (Figure 4a). Samples were drawn that mirrors that of rats with continuous access to feeding from a catheter in femoral artery to determine the secretory (P < 0.0001). In restricted feeding, p-Ca gained significant response to hypocalcemia. PTH secretion did not differ signif- rhythmicity with a small amplitude (P < 0.03) (Figure 5d). icantly between the 2 time points (Figure 4b). We also measured p-FGF23 during restricted feeding and Changing feeding pattern shifts the phase of PTH secretion, found a highly significant rhythmicity (P < 0.0001) and but only slightly impacts the internal circadian clock in the reversal of the phase similarly to that of P and PTH parathyroid gland (Figure 5c). Rats were acclimatized to the standard conditions of 12:12 Despite the dramatic shifted pattern of the circadian hour light-dark cycle and ad libitum feeding for 2 weeks and rhythminp-PTH,thegeneexpressionoftheparathyroid

1464 Kidney International (2020) 98, 1461–1475 S Egstrand et al.: Circadian clock in the parathyroid gland basic research

Figure 2 | Parathyroid hormone (Pth) gene expression is constitutively expressed and similarly the gene expressions of important regulators of PTH secretion, calcium-sensing receptor (CaSR), vitamin D receptor (VDR), and Klotho show no overall diurnal variation. (a) Plasma PTH shows circadian rhythmicity, but (d) Pth gene expression shows no diurnal variation. Diurnal variation was found for (b) Fgfr1 gene expression but not for the Fgfr1 cofactor, (c) Klotho, or for (e) Casr or (f) Vdr (n ¼ 38). Dots represent each animal. Data are fitted by cosinor regression (solid line) and resulting P values are shown; P < 0.05 is considered significant. Gray areas indicate dark periods and white areas indicate light periods. Dashed line is the mesor. Zeitgeber time (ZT; “time-giver”) is time since light onset. Conc, concentration. circadian clock genes in restricted feeding showed only Npas2 only, with significantly reduced amplitude, but did slightly changed profiles without a shift in phase not significantly affect the acrophase timing, mesor, or (Figure 6). Compared with ad libitum–fed rats in proto- amplitude of Bmal1, Per1-3, Cry1-2, and Rev-Erba col I, restricted feeding disrupted the circadian rhythm of (Table 1).

Figure 3 | Regulators of cell cycle and proliferation are rhythmically expressed in the parathyroid glands. Cell-cycle regulators (a) Wee1 and (b) Cyclin D1 show significant rhythmicity, whereas (c) c-Myc does not (P ¼ 0.25). Regulators of parathyroid cell proliferation (d) MafB and (e) Gata3 exhibit significant circadian rhythm. (f) Gcm2 shows no diurnal variation (n ¼ 38). Dots represent each animal. Data are fitted by cosinor regression (solid line) and resulting P values are shown, with P < 0.05 considered significant. Gray areas indicate dark periods and white areas indicate light periods. Dashed line is the mesor. Zeitgeber time (ZT; “time-giver”) is time since light onset. Conc, concentration.

Kidney International (2020) 98, 1461–1475 1465 basic research S Egstrand et al.: Circadian clock in the parathyroid gland

Figure 4 | Parathyroid function, expressed as the secretory response of parathyroid hormone (PTH) to hypocalcemia, is stable between day and night. (a) Hypocalcemia was induced by a continuous i.v. ethylene-bis(oxyethylene-nitrilo)tetraacetic acid infusion and the resulting decrease in plasma Ca2þ was similar between the 2 groups: examined at day (ZT4) or nighttime (ZT16). (b) Plasma PTH secretion in response to hypocalcemia showed no significant difference between day and nighttime. Area under the curve (pg/ml min) was at ZT4: 21,042 5702 and ZT16: 32,475 4757; P ¼ 0.156; n ¼ 6 in each group. Shown are mean SEM. ZT (zeitgeber time; “time-giver”) is time since light onset.

The parathyroid circadian clock is disturbed in uremic nuclear antigen (PCNA) (P < 0.04) (Figure 8b), a proliferation hyperplastic glands marker expressed during DNA synthesis. CKD and sHPT were induced by 5/6 nephrectomy followed Circadian clock genes were severely deregulated in uremic by high-P diet for 8 weeks (protocol 4) or for 24 weeks rats compared with control rats with significant upregulation (protocol 5), respectively. Plasma parameters are presented in of Per1 (P < 0.0002), Per2 (P < 0.03), Rev-Erba (p < 0.004) Supplementary Table S1 and Supplementary Figure S3, while significant downregulation of Npas2 (P < 0.05) was comparing the control and the CKD groups. In CKD, we found (Figure 7). Cosinor analysis demonstrated abolished found significantly increased p-PTH, p-urea, and p-creati- rhythmicity of Per1 in uremic rats, in contrast to Per1 in nine, as well as significantly reduced animal weight. p-P was control rats (p < 0.02) (Figure 7). unchanged, whereas p-Ca was significantly decreased. JTK_CYKLE analysis of the gene expressions showed that Results from protocol 4, with a duration of CKD and high-P in CKD the best fitting period of rhythmicity was reduced to diet of 8 weeks and parathyroid glands harvested every fourth 20 hours as opposed to the normal 24 hours for Per1, Per2, hour for 24 hours (Figures 7 and 8)demonstratedthatgene Per3, and Cry2. Significant shifts in the acrophase were found expression of Pth was significantly elevated in sHPT (P < for Npas2, Per3, and Cry1. Furthermore, a significant increase 0.0001) (Figure 8a), as was gene expression of proliferating cell in amplitude was found for Rev-Erba (Table 1, Figure 7i).

Figure 5 | Restricted feeding shifts the phase of the circadian rhythm of plasma (p) phosphate (P), parathyroid hormone (PTH), and fibroblast growth factor 23 (FGF23). Restricted feeding to the inactive phase reverses the circadian rhythm of (a) pPTH, (b) P, and (c) FGF23. (d) Plasma calcium had little diurnal variation, which was only significant when feeding was restricted (n ¼ 14). Dots represent each animal. Data are fitted by cosinor regression (solid lines) and resulting P values are shown with P < 0.05 considered significant. Gray areas indicate dark periods and white areas indicate light periods. Striped area indicates interval of restricted feeding (ZT2–ZT12). Dashed horizontal line is the mesor. Zeitgeber time (ZT; “time-giver”) is time since light onset.

1466 Kidney International (2020) 98, 1461–1475 S Egstrand et al.: Circadian clock in the parathyroid gland basic research

Figure 6 | (a–j) Effect of restricted feeding on the circadian clock in the parathyroid glands. Expression profiles of core circadian clock genes in the parathyroid gland after 4 weeks of restricted feeding (RF) from ZT2 to ZT12 are fitted by cosinor regression (blue lines) (n ¼ 39) and compared with expression profiles of ad libitum–fed animals from protocol 1 (black lines) (n ¼ 38). Circadian rhythmicity was assessed by cosinor analysis. The resulting P values are shown within figures with P < 0.05 considered significant. Gray areas indicate dark periods and white areas indicate light periods. Zeitgeber time (ZT; “time-giver”) is time since light onset. Asterisk * indicates significant difference (P < 0.05) between the groups at specific time points. Conc, concentration; Ctrl, control.

Regulators of cell cycle under influence of the circadian these 3 transcription factors and the circadian clock- clock were deregulated in uremia; the circadian rhythm of controlled regulator of cell cycle Wee1 (Figure 9). Wee1 and Cyclin D1 was abrogated and downregulation of Results from protocol 5, in which the duration of CKD and Wee1 (P < 0.0001) and p27 (P < 0.0001) was found high-P diet was 24 weeks and parathyroid glands were har- (Figure 8), whereas no change was found for c-Myc vested between ZT3 and ZT8 are presented in Supplementary (Supplementary Figure S4). Figures S5 and S6. Significant deregulation was found for Likewise, significant deregulation was found for regulators Clock (P < 0.05), Npas2 (P < 0.005), Per1 (P # 0.02), Per2 of parathyroid proliferation and PTH secretion; the circadian (P < 0.02),Per3(P < 0.003), Cry2 (P < 0.0001), Rev-Erba rhythm of MafB and Gata3 was abrogated in uremia; and (P < 0.004), and CK1ε (P < 0.0001). Bmal1 expression was MafB, Gata3, and Gcm2 were significantly upregulated not affected by uremia similar to the findings after 8 weeks of (Figure 8). In addition, a strong correlation was found among CKD in protocol 4.

Kidney International (2020) 98, 1461–1475 1467 basic research S Egstrand et al.: Circadian clock in the parathyroid gland

Figure 7 | (a–j) The expression of circadian clock genes in the parathyroid glands is disturbed in chronic kidney disease (CKD). Expression profiles of core circadian clock genes in the parathyroid gland after 8 weeks of uremia (red lines) (n ¼ 44) and compared to normal controls (Ctrl) (black lines) (n ¼ 38). Dots represent each animal. Data are fitted by cosinor regression, and the resulting P values are shown within figures. Gray areas indicate dark periods and white areas indicate light periods. Zeitgeber time (ZT; “time-giver”) is time since light onset. The significant difference in mesor is indicated on top of the single figures. Asterisk * indicates significant difference (P < 0.05) between the groups at specific time points. Conc, concentration.

In protocol 6, phosphoproteomic analysis was performed significant difference was found for phosphorylation of on parathyroid glands from normal rats and rats with CKD CK1E in parathyroid glands of uremic rats compared with and sHPT, induced by adenine diet. Several phosphoryla- that of control rats. tion sites in Clock protein were found. A significant increase in phosphorylation of S852 residues of the Clock protein The circadian clock is disturbed in the uremic aorta (HRT(0.21)DS(0.729)LT(0.061)DPSKVQPQ) was found in Vascular calcification is part of CKD-MBD. The vascular wall CKD rats (P < 0.0005). Phosphorylation at this site, has been shown to have a functioning circadian clock. The conserved in mammals, alters clock nuclear localization and impact of CKD on the vascular circadian clock has not pre- hence function as indicated by a search of PhosphoSitePlus viously been investigated. Induction of vascular calcification (version 6.5.9.3; Cell Signaling Technology, Danvers, MA; in CKD models requires long-term uremia, administration of https://www.phosphosite.org/siteGroupAction.action?id¼ high-P diet, and active vitamin D (protocol 7), while vascular 23094535&protOrg¼23174&showAllSites¼true).22,27 No pathology in CKD rats does not include robust calcification.44

1468 Kidney International (2020) 98, 1461–1475 S Egstrand et al.: Circadian clock in the parathyroid gland basic research

Figure 8 | The circadian rhythm of genes related to cell-cycle regulation and proliferation is disturbed in the uremic parathyroid gland. (a) The Pth gene and (b) proliferation marker Pcna do not express circadian rhythmicity. The levels are, however, significantly increased in uremia. Cell-cycle regulators (c) Wee1 and (d) Cyclin D1 show significant rhythmicity in normal parathyroid glands, which is abolished in chronic kidney disease (CKD), while the expression of (e) p27 is significantly decreased. Regulators of parathyroid cell proliferation (d) MafB and (e) Gata3 exhibited significant circadian rhythm in normal parathyroids, which is abolished in uremia. (f) Gcm2 showed no diurnal variation. Dots represent each animal. Circadian rhythmicity was assessed by cosinor analysis (solid lines), and the resulting P values are shown within figures with P < 0.05 considered significant. CKD rats (red lines) (n ¼ 44) are compared with normal controls (Ctrl) (black lines) (n ¼ 38). Gray areas indicate dark periods and white areas indicate light periods. Zeitgeber time (ZT; “time-giver”) is time since light onset. (Continued)

Kidney International (2020) 98, 1461–1475 1469 basic research S Egstrand et al.: Circadian clock in the parathyroid gland

Figure 9 | The cell-cycle regulator gene, Wee1, is strongly correlated to MafB, Gata3, and Gcm2 in the parathyroid glands (PTGs). Correlations between the gene expression of the regulator of cell cycle, Wee1, and regulators of parathyroid proliferation and parathyroid hormone (PTH) secretion are presented. Strong correlations were found in the normal rat parathyroid glands: (a) MafB (P ¼ 106), (b) Gata3 (P ¼ 109), and (c) Gcm2 (P ¼ 107)(n ¼ 22). In the long-term (24 weeks) uremic rat parathyroid glands a significant correlation was found for (d) MafB (P ¼ 0.001) and (f) Gcm2 (P ¼ 0.0001), but not for (e) Gata3 (P ¼ 0.08) (n ¼ 10). Dots represent each animal. Data are fitted by linear regression (solid lines), and resulting P values and R2 values are shown.

Results from RNA sequencing (RNAseq) data from the downregulated in the uremic calcified aorta (P < 0.001). The calcified rat aorta45 in CKD compared with that from the circadian clock–controlled cell-cycle regulators Wee1 and p21 aorta of control rats have been submitted to The Gene were significantly downregulated (P < 0.001), while Cyclin D1 Expression Omnibus (GSE146638). RNAseq data on core was upregulated (P < 0.001) in the uremic aorta. No change circadian genes and main output genes of circadian clock are was found for p27. presented in Table 2. A significant upregulation of Clock (P < In protocol 4, aorta was removed at 4-hour intervals for 0.001) and a significant downregulation of Per1 (P < 0.001), analysis of the rhythmicity of the expression of the circadian Per2 (P < 0.003), Per3 (P < 0.001), and Rev-Erba (P < 0.05) clock genes in rats at an early stage of CKD with no active were clearly found in CKD. Rev-Erbb (P < 0.05) was also vitamin D administration, compared with control rats. In downregulated in CKD, whereas RORa was unchanged. Al- control rats, rhythmicity of Bmal1 (P < 0.0001), Clock (P < bumin D box–binding protein (DBP) and thyrotroph em- 0.001), Npas2 (P < 0.0001), Per2 (P < 0.0001), Per3 (P < bryonic factor (TEF) are proline and acidic amino acid–rich 0.0001), Cry1 (P < 0.0001), and Rev-Erba (P < 0.001) was basic leucine zipper transcription factors with rhythmic found, whereas Per1, Cry2, and CK1ε did not exhibit signif- expression, known to be controlled by the circadian clock, icant rhythmicity. JTK_CYCLE analysis confirmed the similar to their repressive counterpart, nuclear factor inter- rhythmicity, which was demonstrated by cosinor analysis leukin 3 regulated (NFIL3)46. Dbp, Tef, and Nfil3 were all (Supplementary Table S2).

= Figure 8 (Continued) The significant difference in mesor is indicated on top of the single figures. Asterisk * indicates significant difference (P < 0.05) between the groups at specific time points. Conc, concentration.

1470 Kidney International (2020) 98, 1461–1475 S Egstrand et al.: Circadian clock in the parathyroid gland basic research

Table 2 | Circadian clock and clock-controlled genes in the uremic calcified aorta Gene Name Control aorta (Rpkm) Uremic calcified aorta (Rpkm) P value

Bmal1/Arntl Aryl hydrocarbon receptor nuclear 7 8 0.52 translocator-like 1 Clock Circadian locomotor output cycles 25<0.001 kaput Npas2 Neuronal PAS domain protein 2 18 13 0.052 Per1 Period circadian clock 1 87 51 <0.001 Per2 Period circadian clock 2 51 32 0.003 Per3 Period circadian clock 3 16 7 <0.001 Cry1 Cryptochrome 1 3 3 1 Cry2 Cryptochrome 2 10 11 0.91 Rev-Erba/Nr1d1 Nuclear receptor subfamily 1, group D, 145 42 <0.05 member 1 Rev-Erbb/Nr1d2 Nuclear receptor subfamily 1, group D, 10 7 <0.05 member 2 CK1ε/CSNK1ε Casein kinase 1, epsilon 15 12 0.24 Dbp D site of albumin promoter (albumin 206 28 <0.001 D-box) binding protein Tef Thyrotrophic embryonic factor 32 17 <0.001 Nfil3 Nuclear factor, interleukin 3 regulated 73 32 <0.001 Wee1 WEE1 G2 checkpoint kinase 26 10 <0.001 Ccdn1 Cyclin D1 22 86 <0.001 Cdkn1b/p27 Cyclin-dependent kinase inhibitor 1B 33 28 0.43 Ckdn1a/p21 Cyclin-dependent kinase inhibitor 1A 40 78 <0.001 Icam1 Intercellular adhesion molecule 1 12 28 <0.001 Vcam1 Vascular cell adhesion molecule 1 80 293 <0.001 Ccl2 Chemokine ligand 2 2 26 <0.001 Thbd Thrombomodulin 40 23 <0.001 Rpkm, reads per kilobase per million mapped reads. n ¼ 6. Vascular calcification was induced by long-term uremia and treatment with calcitriol. The aorta gene expressions were analyzed by RNA sequencing.

In CKD, significant upregulation of Rev-Erba (P < 0.02), syndrome.”48 This mechanism could explain the timing of the Clock (P < 0.0001), Cry2 (P < 0.01), and CK1ε (P < 0.05) downward slope of p-P in our results (in active period in ad was found, whereas Per1 was downregulated (P < 0.02) libitum feeding and just after feeding time at ZT2 in the (Figure 10). Finally, a significant change in the amplitude for restricted feeding regime). Circadian rhythmicity of urea might Rev-Erba was found (Supplementary Table S2). be controlled by the circadian clock in the liver,26 known to be entrained by food intake.29 This might explain the observed DISCUSSION shift in the phase of urea. Interestingly, feeding restricted to the This is the first investigation showing an internal circadian inactive phase shifted the phase of p-PTH, but did not clock operating in the parathyroid gland. It is demonstrated significantly shift the phase of core circadian clock genes in the that the expression of the clock genes: Bmal1, Npas2, Per1-3, parathyroids. This demonstrates that feeding only minimally Cry1-2, and Rev-Erba show significant 24-hour rhythmicity affects the circadian clock in the parathyroid gland, in contrast in the rat parathyroid gland. The circadian expression profiles to some other peripheral tissues.29,47,49,50 Furthermore, the were similar to that found in the aorta in the present study. observed dissociation between p-PTH cycle and the intrinsic CK1ε, as expected, did not exhibit rhythmicity.23 parathyroid circadian clock cycle makes it unlikely that PTH Circadian clocks are in some peripheral tissues entrained secretion is primarily regulated by a circadian clock. A constant primarily by food intake rather than light input to the retina. expression of the Pth gene in the parathyroid glands during the To examine the relative importance of these prime zeitgebers in day and no difference in the parathyroid secretory response to the parathyroid gland, we dissociated feeding inputs from hypocalcemia between day and night strongly support this lighting inputs by restricting food availability to only the notion. As such it is still not clear how p-PTH rhythmicity is habitual inactive period of the rats (ZT2–ZT12) for 4 weeks, regulated. Possibly, the rhythmicity might be due to trans- providing ample time for the phase of the circadian clock to lational or posttranslational modifications or secretory mech- accommodate. In previous investigations on peripheral tissues, anisms, as have been shown for insulin.51 the sufficient time for a shift in the phase of the circadian clock PTH has a phosphaturic function in addition to its cal- to take place after a change in feeding time has been <12 ciotropic effect and elevated P has been shown to stimulate days.47 Restricted feeding completely inversed the acrophase of PTH secretion, both directly via CaSR inhibition and – p-P, p-PTH, p-FGF23, and p-urea. p-P is known to decrease increasing Pth mRNA stability4,5,52 54 and indirectly by þ following food intake, primarily due to an insulin-driven shift decreasing p-Ca2 , which further inhibits the CaSR.4,53,54 of phosphate into cells, most evidently in the “refeeding One might speculate that the diurnal variation in P,

Kidney International (2020) 98, 1461–1475 1471 basic research S Egstrand et al.: Circadian clock in the parathyroid gland

Figure 10 | (a–j) Disturbed expression of circadian clock genes in uremic aorta after 8 weeks of chronic kidney disease (CKD). Expression profiles of core circadian clock genes in the aorta after 8 weeks of uremia (red lines) (n ¼ 44) and compared with normal controls (Ctrl) (black lines) (n ¼ 39). Dots represent each animal. Data are fitted by cosinor regression, and the resulting P values are shown within figures. Gray areas indicate dark periods and white areas indicate light periods. Zeitgeber time (ZT; “time-giver”) is time since light onset. In CKD, significant upregulation of Rev-Erba (P < 0.02), Clock (P < 0.0001), Cry2 (P < 0.01), and CK1ε (P < 0.05) was found, whereas Per1 was downregulated (P < 0.02). Asterisk * indicates significant difference (P < 0.05) between the groups at specific time points. resulting from timing of food intake, could drive the diurnal uremia and extended at 24 weeks of uremia to also include variation in PTH. As an indication of that, we found that the Cyclin D1 and c-Myc. phase of the P cycle closely paralleled and preceded that of BMAL1 requires heterodimerization with either CLOCK or PTH in both feeding regimes. NPAS2 for translocation to the nucleus. This might explain sHPT induced by CKD and high-P diet had a significant the downregulation of the circadian clock–regulated cell cycle impact on the expression of the circadian clock genes in the components, despite unchanged Bmal1 gene expression, parathyroid glands. Wee1, Cyclin D1, p27, and c-Myc are cell- because the gene expression of Npas2 and CLOCK phos- cycle regulators, which are controlled by the circadian clock. phorylation was affected by CKD. Significant rhythmicity of the gene expression of Wee1 and The change in phosphorylation of CLOCK in CKD in- Cyclin D1 was found in the parathyroid glands and this was dicates an additional level of regulation of the circadian clock abolished in CKD. Furthermore, a marked downregulation of machinery in the parathyroids, which is disturbed in uremia. cell cycle components Wee1 and p27 was found at 8 weeks of This change might result from upregulation in CLOCK

1472 Kidney International (2020) 98, 1461–1475 S Egstrand et al.: Circadian clock in the parathyroid gland basic research

phosphorylation or from a shift in the phase of CLOCK mice have excessive cell proliferation in the salivary glands phosphorylation. The CLOCK phosphorylation is known to and propensity to cancer.35,36 be circadian, and the glands for our phosphoproteomic study In conclusion, this is the first study to demonstrate an were harvested within a short time interval and not over 24 internal circadian clock machinery operating in the para- hours. thyroid gland. The phase of this clock is not entrained by Our results are in line with results from an elegant study by feeding. CKD and secondary parathyroid growth are associ- Sadowski et al.,55 who examined the gene expression profile in ated with severe deregulation of the circadian clock and clock- human parathyroids and found downregulation of Cry2 in dependent factors, regulating cell cycle and tissue growth. sHPT and downregulation of Per1-2, Cry1-2, and c-Myc in CKD also led to disturbance of the circadian clock and primary hyperparathyroidism. However, the circadian downstream genes in the uremic aorta, indicating that rhythmicity in the parathyroid glands and its disturbance in deregulating of the circadian clock machinery is an important uremia as found in the present study have not been shown feature of CKD-MBD. before. Downregulation of p27 in sHPT has been linked to development of parathyroid hyperplasia and adenoma for- METHODS 56 57 mation. Experimental design Gcm2 is a zinc finger-type transcription factor, which is All rats were acclimatized for 2 weeks at our facility. indispensable for organogenesis of the parathyroid glands, but Protocol 1. A total of 38 male Wistar rats (Charles River, also with retained expression in adult glands. Gcm2 has been Erkrath, Germany) were kept in a 12-hour light-dark cycle and fed proposed to maintain parathyroid cell mass and proliferation a standard diet (0.9% calcium, 0.5% phosphorus, and vitamin D, in adult mice41 and to uphold high expression of CaSR.58,59 1050 IU/kg chow) with ad libitum access to water and food. At 4- Gcm2 requires the transcription factors Gata3 and MafB for hour intervals, 6 to 7 rats were anesthetized, blood was drawn – parathyroid organogenesis and PTH stimulation.38 40,60 In from tails, and parathyroid glands were harvested by microdis- sectionandimmediatelyfrozeninliquidnitrogen.Ratsinvesti- normal rat parathyroid glands, we found significant rhyth- gated in the dark period in all protocols were carefully not exposed micity in the expression of Gata3 and MafB. Furthermore, to blue light by using red light during anesthesia and covering the strong correlations among the gene expressions of MafB, eyes once anesthetized. Gata3, and Gcm2 to that of Wee1 were found in both normal Protocol 2. Another 39 rats had blood samples obtained at 4- and uremic glands, which may suggest a regulation by the hour intervals. Then, feeding was restricted to the inactive light circadian clock, which directly or indirectly is linked to phase from ZT2 to ZT12 for 4 weeks, and blood samples were ob- WEE1. tained again as described above. Weights were recorded and rats Results of studies on the expression of GCMB, the human followed standard growth rate charts. Animals were then killed, and analog of Gcm2, in sHPT have been conflicting, both with parathyroid glands were harvested every 4 hours. increased expression in primary HPT and sHPT43 and also Protocol 3. Twelve rats were studied at either ZT4 or ZT16. with unchanged expression in sHPT and decreased expression Hypocalcemia was induced by infusion of 40 mmol/l ethylene- in primary HPT.61 Another study has looked at MafB in the bis(oxyethylene-nitrilo)tetraacetic acid (Sigma, St. Louis, MO) at context of uremic sHPT in mice, where MafB hap- 4.5 ml/h, by a Braun infusion pump via a catheter inserted in the femoral vein. Blood samples were obtained for determination of p- loinsufficiency suppressed parathyroid proliferation and PTH 2þ 42 fi PTH and p-Ca at times 0, 5, 10, 20, 30, 40, and 50 minutes from a secretion. In the present study, an initial signi cant increase catheter in the femoral artery.3,12,13 The sample volume of 0.6 ml was in the gene expression of Gcm2 and its required transcription replaced by an equal amount of saline. factors, Gata3 and MafB, was found in sHPT at the time of Protocol 4. CKD was induced by one-step 5/6 nephrec- increased parathyroid proliferation, confirmed by an tomy62,63 in 44 rats. Thirty-eight normal rats were used as controls. increased PCNA gene expression, and accompanied by CKD rats received high-P diet (0.9% Ca, 1.4% P, 600 IU vitamin D3 increased Pth gene expression after 8 weeks of uremia. This per kilogram chow). After 8 weeks, parathyroid glands and aorta was followed by a significant downregulation of Gcm2, Gata3, were harvested every 4 hours for 24 hours. and MafB at 24 weeks of uremia. Because GCM2 upholds Protocol 5. Rats were randomization to CKD (n ¼ 10) or parathyroid expression of CaSR,58,59 which is downregulated control (n ¼ 22). CKD was introduced by 5/6 nephrectomy62 and in uremia, and leads to reduced responsiveness of the para- high-P diet. After 24 weeks, parathyroid glands were harvested be- thyroid glands to extracellular calcium and calcimimetics, it tween ZT3 and ZT8. might be important to study the potential regulation of this Protocol 6. For phosphoproteomic analysis, CKD and sHPT transcription factor activity by the circadian clock in future was induced in 8 rats by 0.75% adenine and high-P diet for 14 days.64 Parathyroid glands were harvested from ZT3 to ZT4 in CKD research. rats and from 9 control rats. It is still an open question, what the primary event is: the Protocol 7. Severe vascular calcification was induced in 6 rats deregulation of circadian clock that favors alteration of the by 5/6 nephrectomy, high-P diet, and vitamin D.45 Aorta between the cell cycle or whether the unrestricted growth of the tissue renal arteries and the aortic root was removed from uremic rats as affects the circadian genes. Some evidence suggests that well as from 6 control rats between ZT3 and ZT6. We have submitted disruption of the circadian rhythm may cause abnormal cell raw and processed RNAseq data files to the Gene Expression growth and even an increased risk of cancer. Per2 mutant Omnibus (GSE146638). Some of the vascular calcification–related

Kidney International (2020) 98, 1461–1475 1473 basic research S Egstrand et al.: Circadian clock in the parathyroid gland

RNAseq data have previously been reported,45 while the present SUPPLEMENTARY MATERIAL investigation extends the pool of data to focus on circadian molec- Supplementary File (PDF) ular clock–related genes. Supplementary Methods Figure S1. The core circadian clock protein BMAL1 exhibits Biochemistry rhythmicity in the parathyroid gland. Plasma urea, creatinine, phosphate, total calcium, PTH, and FGF23 Figure S2. The p27 gene expression is not rhythmic in the were analyzed. For details, see the Supplementary Methods. parathyroid gland. Figure S3. Diurnal variation of plasma urea and plasma calcium in Quantitative reverse transcription polymerase chain reaction CKD, as compared to control. Parathyroid and aorta gene expressions were analyzed by quantitative Figure S4. The expression of c-Myc is unchanged after 8 weeks of reverse-transcriptase polymerase chain reaction. For detailed method CKD. Figure S5. Circadian clock genes in the parathyroid gland are and primers, see Supplementary Methods and Supplementary deregulated in long-term uremia. Table S3. Figure S6. Regulators of cell cycle and parathyroid gland proliferation are deregulated in long-term uremia. Western blot Table S1. Plasma biochemistry. Protein expression was analyzed by Western blot. For detailed Table S2. Circadian parameters of core clock genes estimated by method and antibodies, see the Supplementary Methods. JTK_CYCLE analysis in the aortas from control and CKD rats. Table S3. Primer sequences. RNAseq The whole method and dataset can be found in the Supplementary Methods and at the Gene Expression Omnibus (https://www.ncbi. REFERENCES nlm.nih.gov/geo/query/acc.cgi?acc¼GSE146638). 1. Lewin E, Wang W, Olgaard K. Rapid recovery of plasma ionized calcium after acute induction of hypocalcaemia in parathyroidectomized and nephrectomized rats. Nephrol Dial Phosphoproteomics Transplant. 1999;14:604–609. Phosphoproteomics of parathyroid glands from normal and uremic 2. Brown EM, Gamba G, Riccardi D, et al. Cloning and characterization of an rats were analyzed by liquid chromatography–tandem mass spec- extracellular Ca(2þ)-sensing receptor from bovine parathyroid. Nature. trometry. For details, see the Supplementary Methods. 1993;366:575–580. 3. Lewin E, Wang W, Olgaard K. Reversibility of experimental secondary – Statistics hyperparathyroidism. Kidney Int. 1997;52:1232 1241. 4. Centeno PP, Herberger A, Mun H-C, et al. Phosphate acts directly on the All statistical analyses were performed using R (version 3.5.3; R Core calcium-sensing receptor to stimulate parathyroid hormone secretion. Team 2018, R Foundation, Vienna, Austria). For cosinor analysis, Nat Commun. 2019;10:4693. data were fitted to a linear model using the least-squares method 5. Nielsen PK, Feldt-Rasmussen U, Olgaard K. A direct effect in vitro of minimizing the residual sum of the squares: phosphate on PTH release from bovine parathyroid tissue slices but not from dispersed parathyroid cells. Nephrol Dial Transplant. 1996;11:1762– 2p$zt 2p$zt 1768. YðztÞ¼mesor þ b$ cos þ g$ sin ; 6. Mace ML, Gravesen E, Nordholm A, et al. Fibroblast growth factor (FGF) 24 24 23 regulates the plasma levels of parathyroid hormone in vivo through where zt ¼ zeitgeber time, b ¼ A$ cos4, g ¼ A$ sin 4,A ¼ the FGF receptor in normocalcemia, but not in hypocalcemia. Calcif Tissue Int. 2018;102:85–92. 4 ¼ fi amplitude, acrophase. The period is xed at 24 hours. Sig- 7. Kim MS, Fujiki R, Murayama A, et al. 1Alpha,25(OH)2D3-induced nificant rhythm is found when the 95% confidence intervals of b or transrepression by vitamin D receptor through E-box-type elements in g do not include 0. the human parathyroid hormone gene promoter. Mol Endocrinol. For JTK_CYCLE analysis, the R package, MetaCycle (version 2007;21:334–342. 65 8. Lopez I, Pineda C, Raya AI, et al. Leptin directly stimulates parathyroid 1.2.0) was used. – ’ – hormone secretion. Endocrine. 2017;56:675 678. Welch s 2 sample t-test or Mann Whitney test was used for 9. Yuan Q, Sato T, Densmore M, et al. FGF-23/Klotho signaling is not comparing uremic and control groups. Results were classified as essential for the phosphaturic and anabolic functions of PTH. J Bone significant at P < 0.05. Miner Res. 2011;26:2026–2035. 10. el-Hajj Fuleihan G, Klerman EB, Brown EN, et al. The parathyroid — Ethics hormone circadian rhythm is truly endogenous a general clinical research center study. J Clin Endocrinol Metab. 1997;82:281–286. The study was approved by the Danish Animal Inspectorate (refer- 11. Nordholm A, Egstrand S, Gravesen E, et al. Circadian rhythm of activin A ence no. 2017-15-0201-01214) and executed in accordance with and related parameters of mineral metabolism in normal and uremic national guidelines for laboratory animals. rats. Pflugers Arch. 2019;471:1079–1094. 12. Lewin E, Almaden Y, Rodriguez M, et al. PTHrP enhances the secretory response of PTH to a hypocalcemic stimulus in rat parathyroid glands. DISCLOSURE Kidney Int. 2000;58:71–81. All the authors declared no competing interests. 13. Lewin E, Garfia B, Almaden Y, et al. Autoregulation in the parathyroid glands by PTH/PTHrP receptor ligands in normal and uremic rats. Kidney – ACKNOWLEDGMENTS Int. 2003;64:63 70. Special thanks to Nina Vasiljevna Sejthen for her skilled technical 14. Huan J, Olgaard K, Nielsen LB, et al. Parathyroid hormone 7-84 induces hypocalcemia and inhibits the parathyroid hormone 1-84 secretory support. response to hypocalcemia in rats with intact parathyroid glands. JAm Financial support was obtained from The Eva and Henry Fraenkel Soc Nephrol. 2006;17:1923–1930. Foundation, The Kirsten and Freddy Johansen Foundation, and The 15. Gekakis N, Staknis D, Nguyen HB, et al. Role of the CLOCK protein in the Helen and Ejnar Bjoernow Foundation. mammalian circadian mechanism. Science. 1998;280:1564–1569. Proteomic analysis was performed in the Smoler Proteomics 16. Chaix A, Zarrinpar A, Panda S. The circadian coordination of cell biology. Center at the Israel Institute of Technology, Haifa, Israel. J Cell Biol. 2016;215:15–25.

1474 Kidney International (2020) 98, 1461–1475 S Egstrand et al.: Circadian clock in the parathyroid gland basic research

17. Patke A, Young MW, Axelrod S. Molecular mechanisms and 43. Kebebew E, Peng M, Wong MG, et al. GCMB gene, a master regulator of physiological importance of circadian rhythms. Nat Rev Mol Cell Biol. parathyroid gland development, expression, and regulation in 2020;21:67–84. hyperparathyroidism. Surgery. 2004;136:1261–1266. 18. Landgraf D, Wang LL, Diemer T, et al. NPAS2 compensates for loss of 44. Gravesen E, Nordholm A, Mace M, et al. Effect of inhibition of CBP- CLOCK in peripheral circadian oscillators. PLoS Genet. 2016;12:e1005882. coactivated beta-catenin-mediated Wnt signalling in uremic rats with 19. Takahashi JS. Transcriptional architecture of the mammalian circadian vascular calcifications. PLoS One. 2018;13:e0201936. clock. Nat Rev Genet. 2017;18:164–179. 45. Rukov JL, Gravesen E, Mace ML, et al. Effect of chronic uremia on the 20. Preitner N, Damiola F, Lopez-Molina L, et al. The orphan nuclear receptor transcriptional profile of the calcified aorta analyzed by RNA sequencing. REV-ERBalpha controls circadian transcription within the positive limb of Am J Physiol Renal Physiol. 2016;310:F477–F491. the mammalian circadian oscillator. Cell. 2002;110:251–260. 46. Gachon F. Physiological function of PARbZip circadian clock-controlled 21. Meng QJ, Logunova L, Maywood ES, et al. Setting clock speed in transcription factors. Ann Med. 2007;39:562–571. mammals: the CK1 epsilon tau mutation in mice accelerates circadian 47. Damiola F, Le Minh N, Preitner N, et al. Restricted feeding uncouples pacemakers by selectively destabilizing PERIOD proteins. Neuron. circadian oscillators in peripheral tissues from the central pacemaker in 2008;58:78–88. the suprachiasmatic nucleus. Genes Dev. 2000;14:2950–2961. 22. Luciano AK, Zhou W, Santana JM, et al. CLOCK phosphorylation by AKT 48. Kawamura H, Tanaka S, Uenami Y, et al. Hypophosphatemia occurs with regulates its nuclear accumulation and circadian gene expression in insulin administration during refeeding by total parenteral nutrition in peripheral tissues. J Biol Chem. 2018;293:9126–9136. rats. J Med Invest. 2018;65:50–55. 23. Ko CH, Takahashi JS. Molecular components of the mammalian circadian 49. Balsalobre A, Brown SA, Marcacci L, et al. Resetting of circadian time in clock. Hum Mol Genet. 2006;15:R271–R277. peripheral tissues by glucocorticoid signaling. Science. 2000;289:2344– 24. Storch KF, Lipan O, Leykin I, et al. Extensive and divergent circadian gene 2347. expression in liver and heart. Nature. 2002;417:78–83. 50. Crosby P, Hamnett R, Putker M, et al. Insulin/IGF-1 drives PERIOD 25. Rey G, Cesbron F, Rougemont J, et al. Genome-wide and phase-specific synthesis to entrain circadian rhythms with feeding time. Cell. 2019;177: DNA-binding rhythms of BMAL1 control circadian output functions in 896–909. mouse liver. PLoS Biol. 2011;9:e1000595. 51. Perelis M, Marcheva B, Ramsey KM, et al. Pancreatic beta cell enhancers 26. Lin R, Mo Y, Zha H, et al. CLOCK acetylates ASS1 to drive circadian regulate rhythmic transcription of genes controlling insulin secretion. rhythm of ureagenesis. Mol Cell. 2017;68:198–209. Science. 2015;350:aac4250. 27. Robles MS, Humphrey SJ, Mann M. Phosphorylation is a central 52. Naveh-Many T, Rahamimov R, Livni N, et al. Parathyroid cell proliferation mechanism for circadian control of metabolism and physiology. Cell in normal and chronic renal failure rats. The effects of calcium, Metab. 2017;25:118–127. phosphate, and vitamin D. J Clin Invest. 1995;96:1786–1793. 28. Koike N, Yoo SH, Huang HC, et al. Transcriptional architecture and 53. Kilav R, Silver J, Naveh-Many T. Parathyroid hormone gene expression in chromatin landscape of the core circadian clock in mammals. Science. hypophosphatemic rats. J Clin Invest. 1995;96:327–333. 2012;338:349–354. 54. Martin DR, Ritter CS, Slatopolsky E, et al. Acute regulation of parathyroid 29. Stokkan KA, Yamazaki S, Tei H, et al. Entrainment of the circadian clock in hormone by dietary phosphate. Am J Physiol Endocrinol Metab. 2005;289: the liver by feeding. Science. 2001;291:490–493. E729–E734. 30. Matsuo T, Yamaguchi S, Mitsui S, et al. Control mechanism of the 55. Sadowski SM, Pusztaszeri M, Brulhart-Meynet MC, et al. Identification of circadian clock for timing of cell division in vivo. Science. 2003;302:255– differential transcriptional patterns in primary and secondary 259. hyperparathyroidism. J Clin Endocrinol Metab. 2018;103:2189–2198. 31. Laranjeiro R, Tamai TK, Peyric E, et al. Cyclin-dependent kinase inhibitor 56. Buchwald PC, Åkerström G, Westin G. Reduced p18INK4c, p21CIP1/WAF1 p20 controls circadian cell-cycle timing. Proc Natl Acad Sci U S A. and p27KIP1 mRNA levels in tumours of primary and secondary 2013;110:6835–6840. hyperparathyroidism. Clin Endocrinol (Oxf). 2004;60:389–393. 32. Grechez-Cassiau A, Rayet B, Guillaumond F, et al. The circadian clock 57. Tokumoto M, Tsuruya K, Fukuda K, et al. Reduced p21, p27 and vitamin D component BMAL1 is a critical regulator of p21WAF1/CIP1 expression receptor in the nodular hyperplasia in patients with advanced secondary and hepatocyte proliferation. J Biol Chem. 2008;283:4535–4542. hyperparathyroidism. Kidney Int. 2002;62:1196–1207. 33. Hunt T, Sassone-Corsi P. Riding tandem: circadian clocks and the cell 58. Mizobuchi M, Ritter CS, Krits I, et al. Calcium-sensing receptor expression cycle. Cell. 2007;129:461–464. is regulated by glial cells missing-2 in human parathyroid cells. J Bone 34. Casey T, Crodian J, Suarez-Trujillo A, et al. CLOCK regulates mammary Miner Res. 2009;24:1173–1179. epithelial cell growth and differentiation. Am J Physiol Regul Integr Comp 59. Canaff L, Zhou X, Mosesova I, et al. Glial cells missing-2 (GCM2) Physiol. 2016;311:R1125–R1134. transactivates the calcium-sensing receptor gene: effect of a dominant- 35. Gery S, Komatsu N, Baldjyan L, et al. The circadian gene Per1 plays an negative GCM2 mutant associated with autosomal dominant important role in cell growth and DNA damage control in human cancer hypoparathyroidism. Hum Mutat. 2009;30:85–92. cells. Mol Cell. 2006;22:375–382. 60. Han S-I, Tsunekage Y, Kataoka K. Gata3 cooperates with Gcm2 and MafB 36. Fu L, Pelicano H, Liu J, et al. The circadian gene Period2 plays an to activate parathyroid hormone gene expression by interacting with important role in tumor suppression and DNA damage response in vivo. SP1. Mol Cell Endocrinol. 2015;411:113–120. Cell. 2002;111:41–50. 61. Correa P, Akerstrom G, Westin G. Underexpression of Gcm2, a 37. Wang Q, Palnitkar S, Parfitt AM. Parathyroid cell proliferation in the rat: master regulatory gene of parathyroid gland development, in effect of age and of phosphate administration and recovery. adenomas of primary hyperparathyroidism. Clin Endocrinol (Oxf). Endocrinology. 1996;137:4558–4562. 2002;57:501–505. 38. Naveh-Many T, Silver J. Transcription factors that determine parathyroid 62. Lewin E, Garfia B, Recio FL, et al. Persistent downregulation development power PTH expression. Kidney Int. 2018;93:7–9. of calcium-sensing receptor mRNA in rat parathyroids when 39. Kamitani-Kawamoto A, Hamada M, Moriguchi T, et al. MafB interacts with severe secondary hyperparathyroidism is reversed by an Gcm2 and regulates parathyroid hormone expression and parathyroid isogenic kidney transplantation. J Am Soc Nephrol. 2002;13: development. J Bone Miner Res. 2011;26:2463–2472. 2110–2116. 40. Grigorieva IV, Mirczuk S, Gaynor KU, et al. Gata3-deficient mice 63. Hofman-Bang J, Martuseviciene G, Santini MA, et al. Increased develop parathyroid abnormalities due to dysregulation of the parathyroid expression of klotho in uremic rats. Kidney Int. 2010;78:1119– parathyroid-specific transcription factor Gcm2. J Clin Invest. 2010;120: 1127. 2144–2155. 64. Hassan A, Durlacher K, Silver J, et al. The fibroblast growth factor 41. Yamada T, Tatsumi N, Anraku A, et al. Gcm2 regulates the maintenance receptor mediates the increased FGF23 expression in acute and chronic of parathyroid cells in adult mice. PLoS One. 2019;14:e0210662. uremia. Am J Physiol Renal Physiol. 2016;310:F217–F221. 42. Morito N, Yoh K, Usui T, et al. Transcription factor MafB may play an 65. Hughes ME, Hogenesch JB, Kornacker K. JTK_CYCLE: an efficient important role in secondary hyperparathyroidism. Kidney Int. 2018;93: nonparametric algorithm for detecting rhythmic components in 54–68. genome-scale data sets. J Biol Rhythms. 2010;25:372–380.

Kidney International (2020) 98, 1461–1475 1475