Sexually dimorphic renal expression of Klotho is directed by a kidney-specific distal enhancer responsive to HNF1b

Abstract

Transcription enhancers are genomic sequences regulating common and tissue-specific genes and their disruption can contribute to human disease development and progression. Klotho, a sexually dimorphic gene specifically expressed in kidney, is well-linked to kidney dysfunction and its deletion from the mouse genome leads to premature aging and death. However, the sexually dimorphic regulation of Klotho is not understood. Here, we characterize two candidate Klotho enhancers using H3K27ac epigenetic marks and transcription factor binding and investigate their functions, individually and combined, through CRISPR-Cas9 genome engineering. We discovered that only the distal (E1), but not the proximal (E2) candidate region constitutes a functional enhancer, with the double deletion not causing Klotho expression to further decrease. E1 activity is dependent on HNF1b transcription factor binding site within the enhancer. Further, E1 controls the sexual dimorphism of Klotho as evidenced by qPCR and RNA-seq. Despite the sharp reduction of Klotho mRNA, unlike germline Klotho knockouts, mutant mice presented normal phenotype, including weight, lifespan, and serum biochemistry. Lastly, only males lacking E1 display more prominent acute, but not chronic kidney injury responses, indicating a remarkable range of potential adaptation to isolated Klotho loss, especially in female E1 knockouts, retaining renoprotection despite over 80% Klotho reduction.

INTRODUCTION

Transcriptional enhancers allow for increased expression of genes under their control and help build tissue-specific expression profiles.¹ Enhancers bind transcription factors and facilitate the subsequent construction of the mediator complex forming a bridge between enhancer and the gene promoter. Thanks to next generation sequencing methods, such as ChIP-seq, visualization of enhancer histone modifications and DNA binding proteins, including transcription factors, and RNA Pol II became possible. There are multiple databases gathering enhancer mutations relevant for human disease and making them an appealing target of investigation.^2,3 However, due to the diverse structure, range, and orientation of enhancers relative to their target genes, experimental validation of potential enhancers is a labor-intensive yet essential aspect of unraveling the gene regulatory landscape.

Epigenetic regulation of renal gene expression is an active target of investigation. Cornerstone literature on acute kidney injury, polycystic kidney disease and diabetes were published only recently and illustrated hundreds of candidate transcription elements changing activity and shaping the response to renal pathology.^4–7 Among the transcription factors responsible for those changes, HNF1b is inextricably linked to kidney health.⁸ It plays a role in fetal development, ion transport and mitochondrial health, while human mutations of HNF1b result in various nephropathies.^9–12 Taken together, those insights into renal enhancer activity open new avenues of investigation, as in some cases, the usual gene deletion approach is not viable to dissect the gene’s role, and reducing gene expression through enhancer deletion can provide new understanding of clinically relevant mechanisms.

The role of the Klotho gene, first described in 1997.¹³ was initially explored in gene knockout mice, which displayed acute phenotype resembling ageing. Accelerated osteoporosis, hyperphosphatemia, skin atrophy and infertility of both sexes indicated that Klotho was crucial for survival.^14,15 The combination of later studies dissecting Klotho transcripts and their impact on phosphate-calcium homeostasis,^16,17 soon elevated it to a status of a potential kidney injury marker, though it is still debated whether KLOTHO’s depletion is a cause or the consequence of disease development.¹⁸

Despite repeated observation that Klotho expression decreases in kidney injury setting,¹⁹ exact mechanisms governing this change are unknown. Additionally, because 50% of Klotho knockout mice die at six weeks of age, investigating Klotho in adult research animals remained a challenging task. While renal proximal and distal tubule, as well as parathyroid-specific Klotho knockout mice exist,^20–23 their phenotype is much milder, and it’s unknown whether they properly recapitulate the intricacies of human disease. Mouse model where gene expression is evenly reduced but not null throughout the body is likely to be a more robust tool to investigate gene function and the degree of decreased expression’s impact on phenotype, than tissue-restricted or inducible ones.²⁴

An additional challenge in the assessment of the function of Klotho is its sexually dimorphic expression. Female mice are underrepresented in the kidney injury literature, as their sex is linked with renoprotection against acute injury. Similar difference is observed between men and women, with the latter often indicated as less likely to suffer acute injury, but more prone to developing chronic disease.^25,26 It is recognized that male mice express more Klotho mRNA, and recent reports link high testosterone and Klotho levels, suggesting the relevance of experimental data for clinical setting and the need to better understand the regulatory mechanisms of Klotho.^27,28

In this manuscript, we follow up on our previous study, where we dissected the renal transcriptional landscape before and after injury using RNA-seq and ChIP-seq, identifying hundreds of potential gene regulatory elements.⁴ We investigate the hypothesis that two DNA elements located upstream of the Klotho gene, displaying decrease in activity caused by injury accompanied by lower gene expression, are functional enhancers of Klotho. We constructed several lines of mice carrying deletions of different sizes in the two candidate enhancers and investigated the biological consequences of the deletions on gene expression, sexual dimorphism, and the effects of enhancer depletion on acute kidney injury and fibrosis models. Mice lacking one of the enhancer candidates displayed a significant decrease in Klotho expression but had no premature aging phenotype and were fertile unlike Klotho knockout mice. The effect of the deletion was dependent on impacting the HNF1b transcription factor binding site, illustrating its direct link to Klotho regulation. Female renal Klotho mRNA levels were impacted much stronger than male, further intensifying the inherent dimorphism in gene expression. Surprisingly, though Klotho deficiency resulted in higher susceptibility to male ischemia-reperfusion injury, the development of fibrosis was not impacted by Klotho levels, and female mice retained their relative resilience to kidney injury despite markedly lower Klotho levels.

METHODS

Mice

General care.

All animals were housed in the same environmentally controlled room (22–24 °C, with 50 ± 5% humidity and 12 h/12 h light–dark cycle) and handled according to the Guide for the Care and Use of Laboratory Animals (8th edition) and all animal experiments were approved by the Animal Care and Use Committee (ACUC) of National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK, MD) and performed under the NIDDK animal protocol K089-LGP-20. All mice were 12–16 weeks old at the time of respective experiments unless stated otherwise.

Generation of mutant mice.

CRISPR-Cas9 targeted mice were generated using B6D2F1/J (deletion of enhancer 1, E1) or C57BL/6N (E2) (Charles River) or by the transgenic core of the National Heart, Lung, and Blood Institute (NHLBI). Single E1 KO mouse strains were developed using one sgRNA, while E2 KO were generated using two sgRNAs. Due to the proximity of the enhancers, E1/E2 double knockouts were obtained by using E2 sgRNAs on homozygous E1 KO mice rather than generating them independently. All mice were genotyped by PCR amplification and Sanger sequencing (Quintara Biosciences) with genomic DNA from mouse tails. Sequences of genotyping primers and sgRNAs are in Supplementary Table S1. Sequences of all deletions are available in Supplementary Table S2.

Ischemia-reperfusion surgery.

Randomized litters of homozygous-bred mice were used for all the procedures. We chose approximately 3–4 months old mice to balance avoiding mortality in our severe model (2.3% for procedures presented) while aiming for more pronounced injury than in young animals. To perform warm renal ischemia-reperfusion, in random order and with the surgeon blinded to genotype, but not sex, mice were anesthetized with ketamine/xylazine mix (100mg/kg and 10mg/kg respectively). Hair was removed from the mouse retroperitoneal area using sterilized electrical clippers and skin was cleared and prepared using betadine and ethanol swabs. Next the mice were placed over temperature-controlled heating pad maintained at 38°C. Core temperature of the mice was sustained at approximately 35.5 – 36°C as measured by a rectal probe. Renal Ischemia was induced by clamping the renal artery for 30 minutes, either uni- or bilaterally. Then the clamp was removed and skin was closed using sterile wound clips, which were removed at the time of euthanasia or 14 days after surgery depending on follow-up time. Finally, the animals were injected with 1ml saline to replenish fluids, provided analgesia (sustained release Buprenorphine, 1mg/kg), and allowed to recover in a cage heated to approximately 37°C until anesthesia wore off.

Ovariectomy.

3-week-old female WT mice were used for the procedure. Mice were prepared for surgery as above. A dorsal skin incision was made parallel and lateral to the spine from the mid-thoracic curvature to end of curvature, through subcutaneous tissue and muscle to the abdominal cavity. The fat pads were lifted outwards using forceps to exteriorize the ovary. The ovary was then removed by cauterizing between the ovary and oviduct and removing all ovarian tissue. Incision was closed using 6–0 absorbable interrupted suture and staples, then mice were given analgesic and put in recovery as above. Staples were removed 14 days after surgery.

Chromatin Immunoprecipitation and Sequencing (ChIP-seq) and Data Analysis

Renal tissues were frozen with dry ice and stored in −80°C prior to use. Each ChIP library was prepared by combining tissue from four animals of the same genotype to reduce variability. After grinding the tissue with mortar and pestle under liquid nitrogen, chromatin was fixed with formaldehyde (1% final concentration) for 15 min at room temperature and then quenched with glycine (0.125 M final concentration). Nuclei were isolated with Farnham Lysis Buffer (5 mM PIPES pH 8.0, 85 mM KCl, 0.5% NP-40, PMSF and proteinase inhibitor cocktails). The chromatin was fragmented to 200–500 bp using sonicator 3000 (20 s pulse/30 s rest, 24 minutes active time, Misonix Sonicators) and further lysed in RIPA buffer. Approximately one milligram of chromatin was immunoprecipitated with Dynabeads Protein A (Novex) coated with antibodies. The following antibodies were used for ChIP-seq: H3K27ac (Abcam, ab4729), H3K4me3 (Millipore, 07–473), HNF1b (Invitrogen, 720259). Then, 5–10 ug of antibodies were added to 1 mg of total proteins. After serial bead washes, ChIP DNA was reverse crosslinked at 65 °C overnight in the presence of 1% SDS and 1 mg/ml of Proteinase K (Invitrogen), and DNA was purified with QIAquick PCR Purification Kit (Qiagen). The DNA fragments were blunt-ended and ligated to the Illumina index using the NEBNext Ultra II DNA Library Prep kit for Illumina (New England BioLabs). Libraries for next-generation sequencing were prepared and sequenced with a NovaSeq 6000 instrument (Illumina). Quality of received raw data files was assessed with FastQC and data under quality threshold was removed using Trimmomatic.²⁹ Bowtie2 was used for alignment using the reference genome mm10 and MACS2 for peak calling.^30,31 Integrative Genomics Viewer was used for data visualization.³² In addition to original data, GEO series GSE114292 and GSE104907 were used to obtain H3K27ac, H3K4me3, PolII, GR and Esrrγ data visible in Fig.1 a and b.

Fig. 1. Klotho locus includes two putative enhancers.

Histone modification and PolII ChIP-seq data of Klotho (red) locus on mouse chromosome 5 (a). Candidate enhancers are marked by H3K427ac peaks and differentiated from gene promoters through lack of H3K4me3 peaks. Closeup of Klotho enhancers E1 and E2, as well as transcription factors binding within them (red), relative to Klotho promoter (grey) (b). Representative histone modification ChIP-seq showing extent of the deletions (red) in male single enhancer and double knockout mice and a closeup (c, d).

Total RNA Sequencing (Total RNA-seq) and Data Analysis

Total RNA was extracted from whole frozen renal tissue from wild-type and mutant mice and purified with RNeasy Plus Mini Kit (Qiagen, 74134). Ribosomal RNA was removed from 1 μg of total RNAs, and cDNA was synthesized using SuperScript III (Invitrogen). Libraries for sequencing were prepared according to the manufacturer’s instructions with TruSeq Stranded Total RNA Library Prep Kit with Ribo-Zero Gold (Illumina, RS-122–2301), and paired-end sequencing was done with a NovaSeq 6000 instrument (Illumina). Read quality control was done using FastQC and Trimmomatic.²⁹ RNA STAR was used to align the reads to mm10 genome.³³ HTSeq and DeSeq2 were used to obtain gene counts and compare genotypes.^34,35 The data were pre-filtered keeping only those genes, which have at least ten reads in total. Genes were categorized as significantly differentially expressed with an adjusted p-value below 0.05. Differentially expressed genes were visualized with ComplexHeatmap R package.³⁶

RNA Isolation and Quantitative Real-Time PCR (qRT-PCR)

Total RNA was extracted from frozen whole or cortical renal tissue of wild-type and mutant mice using a homogenizer and the PureLink RNA Mini kit according to the manufacturer’s instructions (Thermo Fisher Scientific). Total RNA (1 μg) was reverse transcribed for 50 min at 50 °C using 50 μM oligo dT and 2 μl of SuperScript III (Thermo Fisher Scientific) in a 20 μl reaction. Quantitative real-time PCR (qRT-PCR) was performed using TaqMan probes: mouse Klotho (Mm00502002_m1), Havcr1 (Mm00506686_m1), Acta2 (Mm0156133_m1), Tgfb1 (Mm01178820) Gapdh (Mm99999915_g1), Thermo Fisher Scientific) on the CFX384 Real-Time PCR Detection System (Bio-Rad) according to the manufacturer’s instructions. PCR conditions were 95 °C for 30 s, 95 °C for 15 s, and 60 °C for 30 s for 40 cycles. All reactions were done in duplicate. Relative differences in PCR results were calculated through the CFX Manager software (Bio-Rad) using the comparative cycle threshold (CT) method and normalized to Gapdh levels.

Serum Component Measurements

Serum FGF-23 was measured using ELISA kit according to manufacturer’s instructions (CY-4000 Kainos Laboratories) and serum creatinine with a colorimetric kit (Diazyme). Full renal panel was performed by VRL Diagnostics.

Fibrotic Area Assessment

Kidneys were fixed in 10% neutral buffered formalin for 24 hours, washed and stored in 70% ethanol. Masson Trichrome staining was performed by Histoserv. Keyence BZ-9000 microscope was used to take serial, randomized photographs of cortical kidneys At least 5 photographs were taken per animal. Photographs with at least 90% tissue coverage were then processed with countcolors (https://CRAN.R-project.org/package=countcolors) R package to obtain percentage of photograph occupied by pixels in blue spectrum reflective of Masson Trichrome stain.

Statistics and Reproducibility

GraphPad PRISM 10 was used to analyze experimental data. Normal distribution test (Shapiro-Wilk) was performed before assessing statistical significance of the findings by using appropriate measure detailed in each figure description. All tests used two-tailed p-value and statistical significance was set at p<0.05. Levels of statistical significance were described on graphs as follows: * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. Group sizes are reported in detail in figure descriptions. Error bars on graphs represent standard error of the mean (SEM). Any statistically analyzed data is derived from biological replicates.

RESULTS

Candidate Klotho Enhancers and Construction of the Enhancer-Deficient Mice

The two candidate Klotho enhancers are located approximately 40 kbp upstream of the gene body and promoter region (Fig. 1a). Those sequences are characterized by H3K27ac marks⁴ and the occupation by transcription factors, including hepatocyte nuclear factor 1b (HNF1b), glucocorticoid receptor (GR) and the estrogen related receptor gamma (Errγ), further marking them as potential enhancers (Fig. 1b). Only a known HNF1b motif³⁷ can be found within the enhancer sequences. Pol II coverage of the enhancer loci is observed as well, suggesting transcription and potential presence of enhancer RNAs (eRNAs), however little to no total RNA-seq reads map to the enhancer loci (Supplementary Fig. S1). The Pds5b gene, located further upstream, is expressed at a relatively lower level than Klotho (mean 1686 vs. 21134 WT male read counts as measured by RNA-seq, Supplementary Spreadsheet 1) in renal tissues and has no established impact on kidney disease,³⁸ thus is unlikely to be regulated by the putative enhancers. In addition to the relative proximity of the enhancers to the gene and transcription factor occupancy, height of the H3K27ac peaks were observed to lower after renal injury in tandem with the gene promoter marks, suggesting their role in gene expression modulation.⁴

To distinguish between the two candidate enhancer regions, we called the distal one E1 and the proximal one E2. Initially, we attempted to excise the entire E1 fragment using two sgRNAs in the CRISPR-Cas9 system. However, two attempts at generating a mouse yielded no viable pups for undetermined reasons. We only obtained viable mice after targeting the HNF1b motif located within E1. Due to the imprecise nature of CRISPR-Cas9,³⁹ we observed a range of deletions in the desired locus and developed four separate mouse lines, further called 1744 del (or E1 KO), 1001 del, 145 del, and 31 del. A wild type (WT) mouse line was generated in parallel, by crossbreeding heterozygous E1 KO mice. Deletion of E2, targeted with two sgRNAs, resulted in a 900 bp deletion in the desired region (E2 KO). We chose the E1 KO mice as the ones best representing entire E1 deletion and used them as a source to obtain mice lacking both E1 and E2 (E1/E2 KO) aiming to investigate any compensatory mechanisms a singular deletion might have. To confirm the effects of the deletion on the enhancer structure and activity, we performed ChIP-seq to assess acetylation and methylation of the enhancer and promoter regions (Fig. 1c). We observed that while E2 deletions retains some of the E1 marks, E1 deletion diminishes E2 peak as well (Fig. 1d).

Klotho Enhancer Function is Dependent on HNF1b Binding

First, we investigated whether the enhancer deletions impacted Klotho expression and mRNA levels and whether the size of the deletion impacted their effects. The range of E1 deletions relative to the HNF1b binding site is shown in Fig. 2a. The E1 (1744 bp) and 145 bp deletions cover the entire HNF1b motif, while the 31 bp deletion only removes half of the binding motif. The 1001 bp deletion begins 8 nucleotides downstream of the binding site, leaving it intact. Only the mouse lines where HNF1b binding site was disrupted displayed significantly lower (approximately 50%) renal Klotho expression levels (Fig. 2b). The 1001 bp deletion, despite removing a significant portion of the putative enhancer, did not cause a decrease of Klotho. Surprisingly, complete deletion of the proximal E2 enhancer had no effect on gene expression (Fig. 2c) suggesting that this sequence is not an enhancer in its own right. To investigate whether E2 could function synergistically with E1, we generated mice lacking both E1 and E2. Notably, Klotho expression in E1/E2 KO mice was similar to E1 KO, suggesting that E2 has no significant impact on Klotho expression and elicits no compensatory effect in absence of E1 (Fig. 2d).

Fig. 2. Enhancer deletions impacting HNF1b transcription factor binding site have a marked effect on Klotho expression.

Coverage of E1 deletions relative to HNF1b transcription factor binding site originally targeted by sgRNA (a). qPCR of renal Klotho expression levels in wild type and mutant E1 mouse lines with strains where HNF1b binding site is impacted in red (b). qPCR of renal Klotho expression levels in wild type and mutant E2 mouse line (c) and wild type, E1 and double knockout mouse lines (d). Expression shown as fold change compared to control. Bar = SEM; b, c – n=4, d – n=6, One-way ANOVAs with group mean comparisons, *** P < 0.001, **** P < 0.0001.

Expression of Klotho in Females Lacking the E1 Enhancer Decreases by 90% Without Visible Physiological Consequences

Since Klotho gene expression is sexually dimorphic⁴⁰ it was imperative to investigate the physiological consequences of the enhancer deletions in both sexes. All the obtained mutant lines displayed no significant deviation in phenotype compared to wildtype. As opposed to Klotho knockout mice, they had no shortened lifespan, reaching over a year with no unusual health issues (only rare cases of dermatitis characteristic for the background at >6 months old were observed) and gained weight normally (Supplementary Fig. S2). Both WT and E1 KO female mice displayed decreased renal Klotho expression compared to males, approximately 50% and 90%, respectively (Fig. 3a). In combination with the baseline 50% decrease in renal Klotho, this means that in similar injury models, female mice might be more severely affected. Next, we found that serum FGF23, the main cofactor of Klotho,⁴¹ while remaining at the baseline level in male E1 KO mice, is significantly elevated in females (Fig. 3b). Assuming that this difference might be bound to androgen and estrogen activity, we measured Klotho expression in kidneys of pre-pubescent mice and found a disparity mirroring adults, though less pronounced in E1 KO mice, most likely due to lower overall expression levels (Supplementary Fig. S3a). Additionally, we performed ovariectomy in WT mice at 3 weeks old and euthanized them after five weeks to investigate potential impact of female sex hormones on Klotho expression but observed normal expression levels at 8 weeks old (Supplementary Fig. S3b).

Fig. 3. Female E1 knockout mice display significantly lower Klotho expression at the baseline and after E1 deletion.

qPCR of renal Klotho (a) and serum FGF23 levels (b) in wild type and mutant E1 mouse lines. Representative histone modification ChIP-seq of Klotho locus histone modifications in female (blue) and male (red) wild type and E1 mice (c). Serum phosphate and calcium concentrations in male and female WT and E1 mice (d, e). Expression shown as fold change compared to control. Bar = SEM; a, d, e – n=4, b – n=4 for female and n=6 for male mice, Two-way ANOVAs with group mean comparison, ** P < 0.01, *** P < 0.001.

Next, we assessed the activity at the enhancer and promoter elements of Klotho in the E1 deletion mice, as those changes can help narrow down the causes of sexually dimorphic gene expression. To visualize the Klotho enhancer and promoter regions, we performed ChIP-seq and saw that active promoter marks are being preserved in male E1 KO mice compared to females (Fig. 3c). We observed loss of H3K27ac marks in both E1 and E2 regions upon E1 deletion, indicating a link between E1 and E2, despite apparent irrelevance of E2 for Klotho expression regulation, and E2 deletion didn’t lower the gene expression below baseline in females as well (Supplementary Fig. S3c). Despite the increase in FGF23 and apparent sexual dimorphism, serum phosphate and calcium as well as other serum components (albumin, bicarbonate, blood urea nitrogen, chloride, creatinine, glucose, potassium, sodium, and total protein) in all experimental groups remained at the same, physiological level (Fig. 3 d, e, Supplementary Fig. S4).

Transcriptional Consequences of E1 Deletion.

To investigate genetic programs altered by E1 deletion, we performed RNA-seq on male and female WT and E1 KO kidneys with the goal to observe organ-wide changes. Even at the baseline, WT male and female kidneys are significantly sexually dimorphic, with 939 deregulated genes (DEGs) (Figure 4a, Supplementary Spreadsheet 1). Significant difference in Klotho expression is clearly visible in E1 male vs. female comparison (mean read count 12656 vs. 4685 respectively), among a similar number of 1021 DEGs. The appearing difference strongly suggests that E1 deletion impacts females more. Within sexes, there are only 42 DEGs in male WT vs. E1 and 57 in females, with Klotho holding the highest significance score in the latter group, but not clearing the minimum 2-fold decrease threshold in males required to be included, thus again indicating strong sexual dimorphism. Small amount of DEGs and the fact that this last two, sex-dependent DEG lists only have three common elements (The Gulo gene and two non-gene elements), fits with the fact that drastic decrease in Klotho expression by itself has little impact on steady state homeostasis, as no common pathways are disturbed. We visualized the 20 most significantly deregulated genes in female WT vs. E1 comparison side by side with the other groups as a heatmap (Fig. 4b). In addition to those genes being different between lines, several of them display sexual dimorphism as well (Jchain, Gulo, Hdc). RNA-seq tracks comparing peak height and exon read distribution (Fig. 4 c, Supplementary Fig. S5) reveal overall trend similar to the qPCR results, but curiously the difference lessens when focused on exon 1 of Klotho, except for the low amount of reads in E1 KO females, again indicating presence of a sexually dimorphic promoter element remaining active in E1 KO males (Fig. 4d, e).

Fig.4. E1 deletion causes shift in kidney gene expression.

Number of genes detected by renal RNA-seq analysis as significantly deregulated between WT and E1 KO males and females (a). Heatmap of the 20 most deregulated genes in female genotype comparison, normalized to male WT expression (b). RNA-seq alignment showing read density at the Klotho locus (c) Read counts mapping to exons 1 and 2 of Klotho in male and female WT and E1 mice (d, e). Bar = SEM. d, e – n=4.

Impact of Acute Kidney Injury is Increased by Klotho Depletion Only in Male Mice

Next, recognizing the sexual dimorphism of our model, we performed renal ischemia-reperfusion injury in 3-month-old WT, E1 KO and E1/E2 KO males and females. We chose a relatively severe, 30-minute bilateral model to ensure induction of injury in female mice, which are known to be more resilient to AKI.⁴² We were able to elicit similar levels of serum creatinine increase and body weight loss, indicating successful induction of injury (Fig. 5a, b). Mice were euthanized after 24 hours and renal Klotho and Havcr1 (Kidney injury molecule 1⁴³) expression was measured as direct indicators of renal health and acute injury response. Compared to the baseline of their respective sexes, the decrease of Klotho after injury is more pronounced in males than in females, though the final levels after AKI are similar (Fig. 5c, d). In contrast, while we observed an increase in renal Havcr1 in both sexes, it was much relatively more increased in males, with only mutant males indicating more severe injury than WT controls (Fig. 5e, f).

Fig. 5. Female kidneys remain protected against AKI despite Klotho depletion.

Serum creatinine (a) and % weight loss (b) at 24 hours after AKI in male and female wild type, E1 KO and E1/E2 KO mice. Renal Klotho (c, d) and Havcr1 (e, f) expression 24 hours after AKI in male and female wild type, E1 KO and E1/E2 KO mice. Expression shown as fold change compared to control. Bar = SEM; a, b, c, d – n = 4, Two-way ANOVA with group mean comparisons, separate for WT vs. E1 and WT vs. E1/E2, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001; e, f – n = 4, non-paired T-test between IRI WT and IRI E1 groups, * P < 0.05.

Klotho Depletion Does Not Exacerbate Fibrosis in Male or Female E1/E2 KO mice

Finally, we investigated the effects of diminished Klotho expression on development of renal fibrosis. We used 4-months old male and female WT and E1/E2 KO mice in a 30-minute unilateral ischemia-reperfusion model and euthanized them after 28 days. Unexpectedly, we saw no further decrease in Klotho expression in the injured kidney (Fig. 6a, b). While we saw a marked increase in Tgfβ and Acta2 in male mice, indicating successful induction of fibrosis, no such increase was present in female mice (Fig 6c–f). The levels of fibrosis markers did not differ between WT and E1/E2 KO mice, and the similar fibrosis development within same sexes was confirmed by Masson Trichrome staining, with female mice displaying less renal collagen deposition (Fig. 6g, h, Supplementary Fig. 6).

Fig. 6. Klotho depletion alone does not exacerbate fibrosis in double knockout mice.

Renal Klotho (a, b) Acta2 (c, d), and Tgfb (e, f) expression 28 days after AKI in injured and contralateral male and female wild type and E1/E2 KO mice. Quantification of Masson Trichrome staining in male and female wild type and E1/E2 KO kidneys (g, h). Expression shown as fold change compared to control. Bar = SEM; a-h – n = 6 except male E1/2 group, where n = 5 (mouse excluded due to death during surgery recovery, cause undetermined), Two-way ANOVAs with group mean comparisons, ** P < 0.01, **** P < 0.0001.

DISCUSSION

This is the first report to identify and validate the function of a transcription enhancer in kidney that controls dimorphic Klotho expression. We show that the principle of excising tissue-specific enhancers allows for a detailed investigation and regulation of target genes. The concept of enhancer disruption and modification contributing to disease development has been well established in literature. Possibly the most versatile example of that is enhancer hijacking in cancer, where enhancers are able to activate oncogenes after transposition.⁴⁴ Deletions, methylation or SNP accumulation in enhancers is known to be able to induce a number of congenital and chronic diseases.^45–47 Majority of disease enhancers are first reported as prediction based on ChIP-seq and similar techniques. Although validating the activity of potential enhancers through animal models is essential, it remains infrequent due to its time-consuming nature, expenses and uncertain outcomes. Literature both reports potential for significant impact of the deletions on phenotype in vitro or in vivo.^48,49 and, no less importantly, no differences in knockout strains.⁵⁰ To our knowledge, this is the first report on the functionality of an enhancer of a known renal disease marker. The only comparable work we are aware of is the recent detailed dissection of Cyp27b1 locus enhancers,⁵¹ a gene which by itself is not currently widely considered a kidney disease risk factor, but is heavily involved in the PTH-FGF23 axis, just like Klotho.

It was well within expectations for deletion of the Klotho enhancers to elicit strong dysregulation of transcription, as the activity of those elements decreases after kidney injury in tandem with gene expression.⁴ Kidney pathology displays large shifts in its epigenetic landscape, as recently discussed in the literature. For example, as was illustrated for polycystic kidney disease, enhancer activity drives cystogenesis and majority of PKD-upregulated genes have associated enhancer elements.^5,6 Enhancers also help maintain renal homeostasis, for example by establishing cell identity of podocytes.⁵² This process is regulated by the WNT pathway, crucial for kidney health.⁵³ Because of the sparsity of data, it’s hard to pinpoint transcription factors driving renal transcription, and a wide range of contributors like NRF2, FOSL1 or STAT proteins were proposed.^54–56 HNF1B, the key regulator of kidney development, is also intertwined in regulating aforementioned WNT signaling.⁵⁷ HNF1b knockout model is embryonically lethal, while kidney-specific deletion results in premature death due to organ underdevelopment and heterozygous point mutation mimicking human can result in significant renal deformity.^58–60 We are unaware of any reports indicating direct Klotho regulation by HNF1b as shown here. In fact, while the literature indicates potential HNF1b target genes and discusses its importance, there are only a few examples of validated HNF1b-enhancer interactions. It was reported to directly bind to regulatory elements in Lef1 and Axin2 gene loci in vitro,⁶¹ and several more targets were indicated in a Xenopus model.⁶² HNF1b itself is regulated by a Pax-8-responsive enhancer, and thus might be a promising target in a subsequent study.⁶³

Investigation of Klotho depletion is a relatively rare approach, as since the discovery of its anti-aging properties, a variety of research fields have used Klotho supplementation to reverse disease progression and attenuate injury. Klotho proved to be an effective therapeutic agent in several disease models such as bone regeneration,⁶⁴ pulmonary fibrosis⁶⁵ and myocardial ischemia.⁶⁶ The need for this supplementation could be avoided if the mechanisms regulating native Klotho expression were better understood. So far, attempts to dissect epigenetics of the Klotho locus and preserve its levels were limited to non-specific treatments, such as HDAC inhibition.⁶⁷ Klotho expression is usually considered to be kidney specific. The organ not only boasts the highest expression of the membrane protein, but also is responsible for the vast majority of circulating protein. Low levels of expression are observed in other tissues such as parathyroid gland, and liver Klotho has been suggested to contribute to rodent longevity.⁶⁸ Since Klotho expression in other tissues is negligible and the ChIP-seq datasets to date do not indicate co-presence of enhancer and promoter marks in extrarenal tissues, it is challenging to assess to what degree our intervention affected those secondary expression locations.

What we know, is that the deletion of the validated E1 enhancer impacted the kidneys in a sexually dimorphic manner. Descriptions and validation of such dimorphic enhancers in the literature are sparse, as most studies focus on single sex of experimental animals or conduct global analyses only indicating potential loci of interest.^69,70 The effects of E1 enhancer deletion only partially resembled those shown in tissue-specific Klotho knockouts. While tissue-specific knockouts lack an exon but are still able to initiate making of a dysfunctional protein, our model allows the cells for maintaining low Klotho expression levels. Since the gene promoter can function independently of enhancer, mice lacking the E1 enhancer potentially allow for additional compensatory effects at the promoter, which we saw evidenced more strongly in male mice. Despite marked decrease in Klotho mRNA, normal phenotype remained mostly intact. Both male and female tamoxifen-inducible proximal tubule-specific knockouts reported by Takeshita had significantly lower body mass than controls and had impacted phosphate homeostasis despite the renal Klotho still being present through distal tubule expression.²¹ Without distinction between male and female mice, the decrease in expression was approximately 70%. Targeted deletion of the gene in distal tubule, thought to be the main source of KLOTHO, resulted in no differences in body weight but still impacted phosphate handling caused by similar to E1 KO 50% decrease in expression level.²³ In the same distal tubule-specific Ksp/Kl^−/− mouse strain, soluble Klotho protein level was reported to decrease to nearly zero, which predisposes mice to aortic aneurysms.²⁰ Regrettably, the lack of comprehensive information on the sexual dimorphism of the other models complicates the identification of factors responsible for the disparities between the mutant strains.

In addition, to our knowledge none of those models had been subjected to renal injury models. Higher susceptibility to injury can be assumed based on deregulated phosphate-calcium homeostasis involving hypertension and hyperphosphatemia, but as our results show, Klotho deficiency alone might not be enough to disturb phosphate/calcium homeostasis and thus be the defining factor in injury response, especially in females. Our data indicates only slight predisposition to injury in E1 mutant mice, however considering similar Havcr1 expression trend in E1 and E1/E2 double knockout mice and similar level of Klotho decrease in those strains, subsequent studies focused on injury modulation should elucidate extent of this susceptibility. In a chronic disease setting, isolated Klotho depletion does not worsen the outcome in mice, suggesting that in a clinical setting it’s an effect, not a cause of declining health, or that an unknown compensatory effect exists allowing for normal repair process.

Due to the female mice being more resilient than males in AKI setting, they are far less often used in injury models. This protective phenomenon is mirrored in humans, and though historically the female sex was considered to be a risk factor, current analyses indicate high testosterone, rather than absence of estrogen, to be an important contributor to AKI development.^26,71,72 The human differences in Klotho expression remain to be investigated in detail. Recent studies show no significant difference between the levels of soluble Klotho in the serum,⁷³ though high testosterone has also been associated with high soluble Klotho levels.⁷⁴ The disparities might be caused by different assays used to measure serum Klotho. A study comparing two different immunoassays found that either there is no difference between men and women, or that serum Klotho levels are higher in women.⁷⁵ To further confound those findings, different primate species, such as bonobos and chimpanzees, can display the opposite direction of sexually dimorphic Klotho expression.⁷⁶ While the hypothesis was posed that those differences might stem from androgen and estrogen regulation of Klotho, the direct interaction between androgen receptor (AR) and Klotho is yet to be confirmed, as the AR motif indicated in Klotho promoter is not well conserved and the only evidence for this interaction comes from rat in vitro experiments.⁷⁷ Recent report by Xiong et al. dissects in detail the range of renal transcriptome’s sexual dimorphism and its dependence on androgen receptor signaling but observes no significant change in Klotho expression or dimorphism caused by orchiectomy/ovariectomy or nephron-specific androgen and estrogen receptor knockouts in bulk RNA-seq analysis.⁷⁸

The information our model adds to the renal injury literature is significant. Marked decrease in Klotho increased the AKI severity in male mice but is not the solely responsible for the degree of injury, as female mice were less affected despite the 90% decrease in renal Klotho expression. One of the few human studies available indicates that the renal Klotho levels correspond to the injury severity.³⁶ There are three possible explanations for our finding: 1) either the soluble and membrane Klotho proteins have distinct predictive powers, 2) it’s not the pre- but only post-injury Klotho that allows for differentiating injury levels, 3) the sexual dimorphism in the injury response itself stems from Klotho-independent factors. More surprisingly, we were unable to confirm any effect of Klotho depletion on the development of fibrosis. While we observed increased fibrotic marker expression in our model in male mice, females remained undisturbed by the injury. In both sexes, there was no further decrease in Klotho expression after injury, though there is a possibility of the expression levels recovering during the 4-week fibrosis development period. This, however, would still contradict the large body of literature clearly indicating that decreased Klotho has a strong impact on the progression of chronic kidney disease and renal fibrosis.^79–81 The evidence encompasses Klotho-derived peptides for treatment of fibrosis and microRNA silencing of Klotho worsening the injury.^82,83 One possible explanation is that our model requires a follow-up period longer than a month to be able to differentiate between experimental groups, though the lack of any decrease in Klotho expression at the day 28 timepoint makes it less probable.

In conclusion, our study highlights the importance of taking sexual dimorphism into consideration while exploring the role of epigenetic modifications. Our findings clearly demonstrate the involvement of Klotho in the FGF23 signaling pathway and likely kidney injury response, though those observations are dependent on the sex of research subjects. We demonstrate the value of validating enhancer function by their selective deletion and show that genetically manipulating regulatory cis-elements can change gene expression level and result in unpredictable effects on phenotype.

Funding Statement

This work was supported by the Intramural Research Programs of National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, US (J.J., H.K.L., and L.H.), Intramural Research Programs of National Heart, Lung, and Blood Institute, National Institutes of Health, US. (C.L.) and the Austrian Science Fund (J.W. - P30373).

DATA AVAILABILITY STATEMENT

All ChIP-seq and RNA-seq datasets generated for this study were deposited in Gene Expression Omnibus (GEO) with the accession number GSE243946 (direct link: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE243946). Data is currently set to private and will be released upon acceptance for publication. Please access the data using reviewer token: qpkbkcoohnidbsn.

In addition to original data, GEO series GSE114292 and GSE104907 were used to obtain H3K27ac, H3K4me3, PolII, GR and Esrrγ ChIP-seq data used in Figure 1a and b. Further, data necessary to replicate figures, including original photographs, was deposited in Zenodo data sharing repository with the DOI: 10.5281/zenodo.10672248. Any additional data or materials are available on request.