The Circadian Clock Regulates Rhythmic Erythropoietin Expression in The Murine Kidney
Mar 11, 2022
For more information:ali.ma@wecistanche.com
Lina K. Sciesielsk et al
Cistanche tubulosa prevents kidney disease, click here to get the products and Cistanches
Generation of circadian rhythms is cell-autonomous and relies on a transcription/translation feedback loop controlled by a family of circadian clock transcription factor activators including CLOCK, BMAL1, and repressors such asCRY1 and CRY2. The aim of the present study was to examine both the molecular mechanism and the hemopoietic implication of circadian erythropoietin expression. Mutant mice with homozygous deletion of the core circadian clock genes cryptochromes 1 and 2 (Cry-null)were used to elucidate circadian erythropoietin regulation. Wild-type control mice exhibited a significant difference in kidney erythropoietin mRNA expression between circadian times 06 and 18. In parallel, a significantly higher number of erythropoietin-producing cells in the kidney (byRNAscope ) and significantly higher levels of circulating erythropoietin protein (by ELISA) were detected circadian time 18. Such changes were abolished in Cry-null mice and were independent of oxygen tension, oxygen saturation, or expression of hypoxia-inducible factor 2alpha, indicating that circadian erythropoietin expression is transcriptionally regulated by CRY1 and CRY2. Reporter gene assays showed that the CLOCK/BMAL1 heterodimer activates an E-box element in the 5’ erythropoietin promoter. RNAscope in situ hybridization confirmed the presence of Bmal1 in erythropoietin-producing cells of the kidney. In Cry-null mice, a significantly reduced number of reticulocytes was found while erythrocyte numbers and hematocrit were unchanged. Thus, circadian erythropoietin regulation in the normoxic adult murine kidney is transcriptionally controlled by master circadian activatorsCLOCK/BMAL1, and repressors CRY1/CRY2. This finding may have implications for kidney physiology and disease, laboratory diagnostics, and anemia therapy.
KEYWORDS: chronobiology; circadian rhythm; clock; cryptochrome; erythropoietin; hematopoiesis; hypoxia-inducible factor
Translational Statement Molecular clocks in nearly all cell types drive gene transcription in collaboration with tissue-specific factors. So far, circadian oscillatory mechanisms in the kidney have not been linked to the biology of erythropoietin (Epo). Herein, it has been elucidated that circadian Epo expression is regulated by master clock proteins (cryptochromes 1 and 2, Clock, and Bmal1). As EPO acts distinctively within the complex regulatory network of erythropoiesis, optimal use of recombinant human Epo in patients with kidney insufficiency may include its application just preceding its physiological circadian maximum around midnight.
Circadian rhythms in nephrology are largely unexplored but highly relevant because their disruption by shift work, lifestyle choices, and senescence is associated with an increased risk of various diseases, including cardiovascular disorders.1–4 Animal models of disturbed circadian clocks provide the first evidence of the negative impact of circadian dysregulation on the hematopoietic system (e.g., by increased numbers of aged erythrocytes).5,6
Concerning erythropoiesis, not only cellular effects but also rhythmicity of its primary regulator, erythropoietin (Epo), is of particular interest because athletes have been repetitively accused of blood doping with recombinant human Epo (who) because of time-of-day dependent differences in circulating Epo concentrations and hematological parameters.7,8 Diurnal variations in human serum Epo (S-Epo) levels were first described in 1981 in patients with chronic lung and coal workers respiratory disease,9 and subsequently in healthy subjects.10,11 To date, several reports indicate high S-Epo levels in the night (8 PM to 4 AM) and low S-Epo levels in the early morning (4 AM to 8 AM). The phase and amplitude of S-Epo oscillations, if present at all, are variable between human individuals.9,10,12,13 On average, S-Epo levels changed approximately 1.5-fold relative to their minimum.14–22 This diurnal rhythm was shown to be unaffected by aging,20 training,16, or altitude exposure.17 In contrast, the diurnal oscillation of S-Epo is abolished in patients with chronic obstructive pulmonary disease complicated by daytime hypoxemia, in myeloma with kidney failure, and in myelodysplasia.15,21,23
In 2009, we described circadian Epo mRNA expression in the murine kidney in a large-scale analysis of the promoters of clock-controlled genes but were unable to identify either a distinct genetic element in the EPO gene or a transactivating factor responsible for circadian oscillation.24 More recently, a rat model of hemorrhagic shock suggested, according to parallel expression patterns, that the clock genes (Bmal1 and Per2) were involved in the regulation of EPO secretion during hypoxia/ischemia.25
Generation of circadian rhythms is cell-autonomous and relies on a transcription-translation feedback loop controlled by a family of circadian clock transcription factors, including CLOCK, BMAL1, PER1, PER2, CRY1, and CRY2.26 The CLOCK/BMAL1 heterodimer activates the transcription of circadian clock genes PER1/2 and CRY1/2 via binding to Ebox elements in their promoters. PER and CRY proteins, however, provide negative feedback by inhibiting CLOCK/ BMAL1 activity, thereby reducing their own expression. The net results lead to the oscillation patterns of circadian gene expression and rhythmic changes in cellular and organ physiology.27
To understand the implications of the biological clock, various types of mutant mice with disrupted or ablated single-core clock transcription factors have been studied so far.28,29 Herein, we took advantage of Cry1–/–/Cry2–/– doublemutant mice (Cry-null), which lack the ability to express endogenous circadian rhythms.26,30 Combined in vivo and in vitro data demonstrate that Cry1/2 regulates circadian Epo expression via CLOCK/BMAL1-induced transcription in the normoxic kidney.
METHODS
Animal experiments
Homozygous Cry1–/– /Cry2–/– animals (Cry-null; male and female; C57BL/6J-based)31 and wild-type (WT) controls were bred (For-schungseinrichtungen für Experimentelle Medizin Charité) and raised for 5 to 7 months. The WT mice came from the breeding of the Cry-null colony.
The Cry-null genotype was confirmed by polymerase chain reaction (Supplementary Table S1). For entrainment, mice were group-housed and had food and water ad libitum at a 12-hour:12-hour light/dark cycle for 14 days. On day 2 after release into constant darkness, animals were sacrificed at circadian time (CT) 06 or 18 (n ¼ 13–15 for each group and time point). Tissues (liver and kidney) were quickly obtained and snap-frozen in liquid nitrogen. A subgroup of male and female WT and Cry-null animals were analyzed in detail for body weight and differential hemogram.
For blood gas analysis, animals were anesthetized (fentanyl, 0.075 mg/kg, midazolam, 1.5 mg/kg, and medetomidine, 0.75 mg/kg), tracheotomized, intubated, and ventilated (tidal volume, 9 ml/kg; respiratory rate, 160 min- 1; positive end-expiratory pressure, 2 cm H2O), as described.32 A polyethylene catheter was surgically introduced into the left carotid artery. After 5 minutes of stabilization, the experiment was terminated through rapid exsanguination via the carotid catheter, blood gases were analyzed (ABL-800; Radiometer; temperature controlled), and kidneys were excised for post hoc analyses.
All procedures were authorized by the Local Animal CareCommittee (T0307/08; G0100/17 with an addendum from January 2021) and performed in accordance with the guidelines and regulations of the German animal protection law.
Preparation of RNA and quantitative polymerase chain reaction analysis
Total RNA was extracted as described.33 A total of 1000 ng total RNA was reverse transcribed with SuperScript III reverse transcriptase (Thermo Fisher; No. 18080085) and random hexamers (Thermo Fisher; No. SO142), according to the manufacturer’s instructions. Quantitative polymerase chain reactions were run on a StepOnePlus cycler (Life Technologies) with intron-spanning primers or TaqMan assays (Supplementary Table S2). Absolute mRNA quantification was achieved by comparison with a standard curve from serial dilutions of polymerase chain reaction template.
Detection of Epo mRNA expression in the kidney byRNAscope technique
RNAscope assay was performed according to the manufacturer's protocols (ACD; technical note 320536). The 10-mm mid kidney transverse cryosections were stained with a C1 probe against DapB(negative control; ACD; No. 310043) or Epo (ACD; No. 315501). Hybridization steps using Amp 4-6 and detection of the red signals were omitted. Two independent, blinded researchers counted Epopositive cells at 200 original magnification, on 3 to 8 cryosectionsper animal at an Axioplan 2 imaging system (Zeiss).
Representative Epo quantification images and double fluorescent staining were performed using the RNAscope Multiplex FluorescentDetection Reagent V2 Kit (ACD; No. 323110), according to the manufacturer's protocols. The 1.5-mm mid kidney transverse para sections were stained with a C1 probe against Bmal1 (ACD; No.438741) and a C2 probe against Epo (ACD; No. 315501-C2). Opaleye 520 (Akoya BioSciences; No. FP1487001KT) was used with the theC1 probe, and opal dye 650 (Akoya BioSciences; No. FP1496001KT)was used with the C2 probe. At - 400 original magnification, Epopositive cells were imaged for Bmal1 colocalization at an Eclipse Ti2imaging system (Nikon). The 40,6-diamidino-2-phenylindole was used as counterstaining.
EPO serum concentrations
Blood samples were allowed to clot for 1 hour at room temperature before centrifuging for 20 minutes at 2000 - g. Serum was removed and immediately frozen at –80 C until performing the enzyme-linked immunosorbent assay for Epo (Quantikine; R&D Systems; No. MEP006) with undiluted samples. Absorbance was read with an iMARK Microplate Absorbance Reader (Bio-Rad) at 450 nm, with wavelength correction at 570 nm and a 4-parameter fit standard curve, as described previously.33
Blood cell counts
total and differentiated cells counts from EDTA-anticoagulated blood were measured by Synlab. vet Berlin with an ADVIA2120i(Siemens) automated cell counter for murine blood.
Reporter gene assays
Human embryonic kidney 293 cells (DSMZ; No. ACC305; passage numbers 3–10; mycoplasma negative) were grown in Dulbecco’s modified Eagle’s medium/Ham F12 (Biochrom; No. FG4815) supplemented with 10% fetal bovine serum (Merck; No. F7524). Cell transfection was performed in 12-well plates containing 1.67 - 105 cells/well. Each well was transfected with 333 ng plasmid DNA (1 of 10 of which was vanilla construct) and 1 ml Fugene 6 transfection reagent (Promega; No. E2691) as described.34 All constructs used are listed in Supplementary Table S3. Cells were lysed 48 hours after transfection with Passive Lysis buffer (Promega; No. E1941). Luciferase activity was determined with the Beetle-Juice and Renilla-Juice kits (both pjk GmbH; Nos. 102511/102531, respectively), at a Lumat LB9501 luminometer. Each experiment was performed in technical duplicates, and mean values were used for calculations.
Statistical analysis
In all animals, the circadian gene expression was analyzed; 2 animals were excluded as they showed outlier values (>1.5-fold the interquartile range) in 4 of 6 circadian expressed genes. Data were analyzed using IBM SPSS Statistics 27 and are presented as individual dots with the median or as bars with the mean and SD. Mann-Whitney U test or Kruskal-Wallis with Bonferroni as post hoc test was performed.
RESULTS
Ablation of circadian Epo expression in Cry-null mice
To elucidate the molecular mechanism of circadian Epo regulation, we analyzed Epo mRNA and protein expression in WT and Cry-null mutant mice. In WT kidney, canonical clock genes showed a time-of-day dependent expression, whereas this was abolished in Cry-null mice, as expected (Supplementary Figure S1). We previously reported circadian oscillation of Epo mRNA expression over 24 hours for adult WT murine kidneys.24 Focusing now on the minimal and maximal values, we observed a significant, w9-fold difference between CT06 (mouse sleeping period; expected minimum) and CT18 (activity phase; expected maximum; Figure 1a). In the kidneys of Cry-null mice, however, no significant difference in kidney Epo mRNA expression was detected. Notably, in Cry-null mice, the absolute amounts of kidney Epo transcript levels at both times were in between the median WT Epo mRNA levels at CT06 and CT18 (Figure 1a). EPO mRNA in corresponding livers was below the detection limit (data not shown).
Diurnal changes in Epo serum concentrations
To estimate the translation of circadian Epo mRNA expression into circulating Epo protein, we analyzed serum samples by enzyme-linked immunosorbent assay. Blood samples were taken before organ specimens. Serum Epo increased w2.3- fold between CT06 and CT18 in WT mice, but this difference was abolished in Cry-null mice (Figure 1b). Notably, the median S-Epo concentrations averaged over time (CT06 and CT18) were similar in WT and Cry-null mice (22 mU/ml [range, 3–59 mU/ml] vs. 21 mU/ml [range, 7–50 mU/ml]).
Circadian Epo expression in relation to arterial blood gas parameters and pulse oximetry
To investigate whether potential diurnal changes in blood and tissue oxygen levels could cause circadian oscillation of Epo expression, arterial blood specimens were obtained and blood gas analyses were performed in anesthetized, tracheotomized, and mechanically ventilated Cry-null and WT mice at CT06 or CT18, which correspond to the lowest and highest circadian Epo mRNA levels, respectively (Figure 1a). We detected neither significant differences in the arterial pH, partial pressure of CO2, or partial pressure of O2 nor in the standard base excess or lactate concentrations (Figure 2a–e) between CT06 and CT18 or WT and Cry-null mice. Oxygen saturation levels, measured by pulse oximetry, also did not show any differences (Figure 2f). Gene expression of the EPO master regulator hypoxia-inducible factor (HIF) 2a did not differ between CT06 and CT18, either in WT or Cry-null mice (Figure 3). Thus, under normoxic conditions, circadian Epo regulation is not caused by changes in oxygen tension.
Circadian on-off switch of Epo expression in renal Epo-producing cells
To study whether circadian Epo mRNA expression is mediated by switching on additional renal Epo-producing cells (REPCs) or solely by increasing Epo expression per cell, RNAscope in situ hybridization was used on mid kidney transverse sections (example in Supplementary Figure S2). In WT mice, the number of REPCs significantly increased between CT06 (median, 5; range, 0–25) and CT18 (median, 22; range, 5–82; 4.4-fold; P ¼ 0.010). In contrast, Cry-null kidneys displayed similar numbers of REPCs at CT06 (median, 18; range, 2–33) and CT18 (median, 15; range, 4–87; not significant; Figure 1c). We considered that Cry-null mice exhibit growth restriction due to impaired signaling of insulin-like growth factor 1 (IGF1), which results in a continuously increasing difference in body weight and organ size between Cry-null and WT mice.30 As we used relatively young animals, the absolute kidney weight of Cry-null mice was only slightly lower than in WT mice (–12% in Cry-null mice), but the relative kidney–to–body weight ratio did not differ (Supplementary Figure S3). Thus, the circadian increase in Epo mRNA expression seems to be regulated by an on the switch of additional REPCs.
Activation of the minimal EPO promoter by CLOCK/BMAL1
To identify the regulatory sequences responsible for circadian Epo expression (Figure 4a), luciferase reporter gene assays were performed in human embryonic kidney 293 cells. The human embryonic kidney 293 cell line was chosen not for its kidney origin but for its lack of an endogenous circadian clock. Thus, the stimulatory effect of CLOCK/BMAL1 could be tested in a low, nonoscillating CLOCK/BMAL1 background. Overexpression of CLOCK/BMAL1 significantly stimulated the activity of the minimal human EPO promoter (Figure 4b, I vs. II). If the E-box motif (–36 to –31 bp relative to the transcriptional start site) is mutated,35 this effect is blunted (Figure 4b, III).
To study whether the observed circadian regulation was cell type-dependent, we screened several EPO-expressing cell lines for endogenous clock activity by monitoring Bmal1 promoter-mediated oscillations of a luciferase reporter. Those with an endogenous rhythm included human hepatoma-derived HEP3B cells, human neuroblastoma-derived KELLY cells, and PDGFRbþ mouse kidney cells (formerly EPOexpressing mouse cell line FAIK1-10).36,37 Among the 3 cell lines, only KELLY cells exhibited EPO promoter-driven luciferase oscillations (Supplementary Figure S4A). Although PDGFRbþ cells exhibited a strong circadian rhythm of Bmal1 promoter activity, we did not detect EPO promotermediated oscillations, suggesting a lower amplitude of the EPO promoter-driven reporter construct (Supplementary Figure S4B).
Colocalization of Bmal1 and Epo in the kidneys of Cry-null versus WT mice
To further elucidate (i) the circadian regulation of kidney Epo production by recruiting renal Epo-producing cells and (ii) the colocalization of Bmal1 and Epo expression, we performed RNAscope in situ hybridization. Epo and Bmal1 colocalized, and microscopy on low magnification represents differences in the recruitment of REPCs (Figure 5), as quantified in Figure 1c.
Hematologic findings in Cry1/Cry2 deficiency To assess (i) the hematopoietic effects of the lack of circadian Epo expression and (ii) possible other hematologic abnormalities in mice without a functional clock, blood cell counts were analyzed. The time-of-day differences of Epo mRNA expression and S-Epo concentration were not mirrored by significant differences in peripheral reticulocyte counts at CT06 versus CT18 in WT mice. Reticulocyte counts, however, were significantly lower at both CT06 and CT18 in Cry-null animals compared with WT mice (Figure 6a). There was no significant difference in the erythrocyte numbers or hematocrit values in WT versus Cry-null mice. Both parameters, however, did not vary between CT06 and CT18 in both strains (Figure 6b and c) and did not correspond to the lower peripheral reticulocyte number in Cry-null mice (Figure 6a). To test whether nutritional deficiencies (e.g., iron and folate) due to impaired IGF1 signaling are involved in the discrepancy of reduced reticulocyte but normal erythrocyte numbers in Cry-null mice, red blood cell size and shape were analyzed but did not show any significant differences between Cry-null and WT mice (Supplementary Figure S5).
Notably, median platelet counts in Cry-null mice tended to be higher than in WT mice, but platelet numbers were not significantly different between CT06 and CT18 in both strains (Figure 6d). In contrast, white blood cell (WBC) counts differed significantly between CT06 and CT18 in WT but not in Cry-null mice. The median WBC number in Cry-null mice was similarly high, as in WT mice at CT06 (Figure 6e).
DISCUSSION
Herein, we demonstrate that circadian Epo expression is regulated at the transcriptional level. The analysis of Epo mRNA and S-Epo concentrations in arrhythmic Cry1 and Cry2 deficient mice redirected our search for “Epo’s clock” to their downstream target transcription factors CLOCK and BMAL1, which are repressed by CRY1 and CRY2.26 Ablation of Cry1/Cry2 leads to the loss of rhythmic repression of CLOCK/BMAL1, resulting in constant Epo transcript levels that were in between the median levels at CT06 and CT18 in WT mice (Figure 1a). In situ hybridization by RNAscope indicates that the circadian oscillation is achieved by switching on Epo mRNA expression in interstitial cells (Figure 1c), and in situ hybridization by RNAscope also revealed the expression of Bmal1 in REPCs (Figure 5).
More important, the significantly higher S-Epo levels at CT18 confirm that the circadian, transcriptional Epo regulation translates into circadian oscillations of circulating Epo protein under normoxic conditions (Figure 1b). The real maximum of S-Epo levels is expected slightly later than CT18 because de novo synthesis of EPO protein requires w80 to 120 minutes,38 and circadian times were chosen on the basis of maximal Epo mRNA levels. 24 We found that circadian Epo regulation is likely mediated by transcriptional activation of an E-box motif in the 5’ EPO promoter by CLOCK/BMAL1 (Figure 4 and Supplementary Figure S4), which is consistent with the described positive correlation between BMAL1 protein and S-Epo levels in a rat model of acute hemorrhage.25
There is evidence for a bidirectional regulation between the CLOCK/BMAL1 and the HIF pathways through direct protein-protein interaction.39,40 Upregulation of HIF2a, the major activator of the EPO promoter,41,42 resulted in altered expression levels of clock genes in human hepatoma cells.43 Furthermore, BMAL1 dimerizes with HIF1a and HIF2a proteins,44,45, and the circadian clock control of HIF activity are regulated in a tissue-specific manner.39 In mice, exposed to acute hypoxia (4 hours at 6% vs. 21% O2), EPO mRNA was excessively increased but did no longer show circadian differences.46 Thus, normoxic conditions are probably most appropriate for dissecting the molecular mechanism of circadian Epo regulation. In rodents, however, diurnal changes in blood and tissue oxygen levels have been reported.39,47 In rats, there is a low range of rhythmic daily changes in kidney oxygenation of zD3%, with a peak in the dark (¼ rodent activity).47 Such differences could not be detected in our experiments, in which only anesthetized, tracheotomized, and mechanically ventilated mice could be studied for animal regulatory reasons. However, analysis of arterial pH, partial pressure of CO2, and partial pressure of O2, standard base excess or lactate as well as oxygen saturation did not indicate major differences between WT and Cry-null mice at both CTs (Figure 2). Furthermore, Hif2a transcript levels in the kidney were also similar in all conditions (Figure 3). Thus, under normoxic conditions, circadian Epo regulation seems to be independent of diurnal changes in oxygen tension. However, the question to which extent high altitude or hypoxia (low pO2) influences the circadian oscillation of Epo production deserves further attention. Human S-Epo levels are generally higher at high altitude, whereas phase and amplitude are unchanged,17,22 and in healthy volunteers exposed to normobaric hypoxia, S-Epo concentrations show a pronounced oscillation.48
Circadian Epo regulation is probably most relevant for clinical nephrology and hematology, but also regarding laboratory diagnostics, such as blood testing for doping with erythropoiesis-stimulating agents. For evaluation of circulating Epo concentrations in doping analyses, the time of day (external time) at the collection of the blood and the chronotype of the person (internal time) need to be considered.49,50
In erythropoiesis, burst-forming units–erythroid is the first lineage-specific cells, followed by colony-forming units– erythroid, which show abundant expression of Epo receptor. After 2 days in culture with Epo, murine colony-forming units–erythroid produce erythroblast colonies. Once the stage of orthochromatic erythroblasts is reached after 7 days, the cells extrude their nuclei to become reticulocytes lacking Epo receptor expression.51 Considering the time it takes for the colony-forming units–erythroid to mature to reticulocytes or even fully mature erythrocytes, it is probably not surprising that no differences were observed between CT06 and CT18 in WT mice (Figure 6a–c). Notably, overall reticulocyte numbers were significantly higher in WT than in Cry-null mice (Figure 6a), suggesting that differentiation into reticulocytes is impaired in Cry-null mice. Previous data indicate that early erythroid progenitor cells also exhibit a circadian pattern of DNA synthesis. In vivo administration of rhEpo enhances the circadian rhythms of erythroid colony numbers.52 Lower reticulocyte numbers, despite relatively high overall S-Epo concentrations in the Cry-null mice, could result from abolishing circadian DNA synthesis on the level of erythroid progenitors. Such constellation of high S-Epo and reduced erythropoiesis has been observed in mice with genetic ablation of pineal melatonin production (C3H/HeN mice that lack a rate-limiting N-acetyl transferase),53 arguing for intrinsic circadian activities on the erythroid progenitor cell level.
Notably, Cry-null mice show an approximately 80% reduction in IGF1 levels, leading to reduced IGF1 signaling and a 30% reduction of body weight and organ size compared with WT, an effect that exacerbates over a lifetime.30 Although animals used in our experiments were relatively young (median age, 21–29 weeks), they displayed a moderate effect on total body and absolute kidney weights in Cry-null mice, but the kidney–to–body weight ratio was normal (Supplementary Figure S3). The latter fact may be relevant for the capacity of Epo protein synthesis (Figure 1b).
However, analysis of red blood cell size or shape (MCV, MCH, and MCHC; Supplementary Figure S5) did not suggest nutritional deficiencies (e.g., iron and folate) as an explanation for the discrepancy between reduced reticulocyte and normal erythrocyte counts in Cry-null mice (Figure 6a and b). Notably, Epo and IGF1 signaling synchronize cell proliferation and differentiation during erythropoiesis via interaction with the GATA-1/friend of the GATA-1 transcriptional complex.54 In this process, Epo activates the cellular AKT pathway and thereby increases the affinity of GATA-1, the major transcriptional regulator of erythropoiesis, to its cofactor friend of GATA-1. This mechanism is inhibited, however, if IGF1 signaling is abolished.54 Thus, a positive effect of relatively higher Epo in Cry-null mice on red cell differentiation may outweigh the reduced IGF1 activity in the relatively young Cry-null mice studied herein.
The general implication of complete circadian arrhythmicity on hematopoiesis is further elucidated by our analysis of platelet and WBC numbers. Previous results from mice expressing a dominant-negative form of Clock (ClockD19/D19) indicated that the disruption of the expression of thrombopoietin (the primary regulator of megakaryopoiesis) and its receptor Mpl results in increased numbers of mature marrow megakaryocytes and circulating platelet numbers.55 The described significant circadian oscillation of platelet numbers with at peak at zeitgeber time (ZT) 20 in WT mice as well as higher platelet numbers at ZT08 in ClockD19/D19 mice,55 resulting in a lack of circadian oscillation of platelet numbers, could not be confirmed in our mouse model (Figure 6d). Another important finding is the loss of time-of-day differences of circulating WBCs in Cry-null mice in contrast to WT mice (Figure 6e), which is consistent with the reported activity of CLOCK/BMAL1 on mature WBC production.56,57
For the interpretation of our data, differences in day-night activities in mice and humans need to be considered: The time CT06 in humans roughly corresponds to midnight (sleeping phase), whereas CT18 roughly corresponds to midday (activity phase). The opposite is true for nocturnal mice, but both organisms show the same circadian Epo expression pattern (low at CT/ZT06, high at CT/ZT18). This indicates that additional mechanisms (e.g., metabolic factors) could modify the rhythms of Epo expression in both organisms. The synchronization between Epo peak levels and Epo receptor expression in early erythroid progenitors (burstforming units–erythroid and even colony-forming units– erythroid cells)52 suggests that patients receiving rhEpo treatment (e.g., in end-stage kidney anemia or hematological disorders) would benefit from mimicking the normal circadian physiology by applying hypo during the night. Hematopoietic disorders have not been reported in humans with CRY1 (Online Mendelian Inheritance in Man *601933) or CRY2 (Online Mendelian Inheritance in Man *603732) loss-of-functions mutations yet. Clinical reports associate CRY1 variants primarily with attention-deficit/hyperactivity disorders, frequently accompanied by insomnia, anxiety, depression, or delayed sleep phase disorder.58,59 This deserves further investigation because such diseases can be worsened by anemia or hematopoietic disorders.
In conclusion, this study provides the first evidence that circadian Epo expression in the normoxic adult murine kidney is regulated at the transcriptional level by CLOCK/BMAL1-mediated activation of an E-box element in the 5’EPO promoter. As EPO acts distinctively within the complex regulatory network of erythropoiesis, optimal use of repo in patients with kidney insufficiency may include its application just preceding its physiological circadian maximum in the night.
