Increasing mTORC1 Pathway Activity or Methionine Supplementation during Pregnancy Reverses the Negative Effect of Maternal Malnutrition on the Developing Kidney

Significance Statement

Kidney development ceases at the end of the third trimester of pregnancy, with no new nephrons forming after birth even with numerous injuries. Therefore, the intrauterine environment, as the maternal nutritional state, has a high effect on the risk of kidney disease when the fetus reaches adulthood. Impairment of mTOR pathway activity and methionine metabolism in nephron progenitor cells play a pivotal role in mediating the effect of caloric restriction during pregnancy on nephron endowment in a mouse model. Increasing the activity of the mTOR pathway or methionine supplementation during pregnancy reverses the negative effect of maternal malnutrition on the developing kidney. These results highlight new interventions to improve nephrogenesis in high-risk pregnancies.

Visual Abstract

Abstract

Background

Low nephron number at birth is associated with a high risk of CKD in adulthood because nephrogenesis is completed in utero. Poor intrauterine environment impairs nephron endowment via an undefined molecular mechanism. A calorie-restricted diet (CRD) mouse model examined the effect of malnutrition during pregnancy on nephron progenitor cells (NPCs).

Methods

Daily caloric intake was reduced by 30% during pregnancy. mRNA expression, the cell cycle, and metabolic activity were evaluated in sorted Six2 NPCs. The results were validated using transgenic mice, oral nutrient supplementation, and organ cultures.

Results

Maternal CRD is associated with low nephron number in offspring, compromising kidney function at an older age. RNA-seq identified cell cycle regulators and the mTORC1 pathway, among other pathways, that maternal malnutrition in NPCs modifies. Metabolomics analysis of NPCs singled out the methionine pathway as crucial for NPC proliferation and maintenance. Methionine deprivation reduced NPC proliferation and lowered NPC number per tip in embryonic kidney cultures, with rescue from methionine metabolite supplementation. Importantly, in vivo, the negative effect of caloric restriction on nephrogenesis was prevented by adding methionine to the otherwise restricted diet during pregnancy or by removing one Tsc1 allele in NPCs.

Conclusions

These findings show that mTORC1 signaling and methionine metabolism are central to the cellular and metabolic effects of malnutrition during pregnancy on NPCs, contributing to nephrogenesis and later, to kidney health in adulthood.

The prevalence of CKD is increasing worldwide, with an estimated global prevalence of 10%.^1,2 Prevention efforts are directed at the modifiable factors exacerbating renal injuries, such as obesity, hypertension, and diabetes. However, suboptimal kidney development in utero has long-term effects on the risk of CKD at later stages of life. In humans, nephrogenesis starts approximately at the fifth week of pregnancy and proceeds through reciprocal interactions between the ureteric bud (UB) and the metanephric mesenchyme comprising the nephron progenitor cells (NPCs) that lead to the formation of nephron epithelia via mesenchymal to epithelial transition. In humans, the majority of nephrons are formed in the last trimester of pregnancy.³ Nephrogenesis stops at the 35th week of gestation when the NPC population differentiates with a coincident disappearance of the embryonic environment. From that point on, no new nephrons will be generated to replace those lost due to injury or disease.^4,5 Unlike humans, where nephrogenesis is completed in utero, nephrogenesis in mice continues for another 3 days after birth.^5–11 A combination of genetic and environmental factors affects the total number of nephrons formed, which ranges from 200,000 to >2 million nephrons per kidney. One of the key drivers for this variation is insufficient nutrition.^12,13

The intrauterine environment and mainly, the nutritional status of the pregnant mother have significant effects on the development of various organs. Thus, poor nutrition during pregnancy increases the susceptibility of the offspring to different diseases in adulthood, including heart disease, diabetes, and kidney disease.¹⁴ It was hypothesized that a poor intrauterine environment impairs the developmental programming of the embryonic kidney, resulting in a low nephron number. Low nephron number amplifies the harmful effect of nephron loss due to hypertension or other factors increasing hyperfiltration by the residual nephrons; this, in turn, increases the risk of progressive sclerosis of the nephrons, accelerating onset of CKD.^13,15–21 Offspring of mothers with a history of malnutrition during pregnancy have a high risk of kidney disease secondary to impaired intrauterine development.^{12,17,19,21–24} A correlation between impaired nephron endowment and high risk for hypertension and proteinuria has been demonstrated in human postmortem studies and in rodent models of malnutrition, mainly induced by a low-protein diet during pregnancy.^25–30

The cellular and molecular effects of poor maternal nutrition on NPCs in utero, which will ultimately reduce nephron number, are still obscure.^31–33 NPCs may require various macromolecules, such as essential amino acids, cofactors for various enzymatic activities, rare earth minerals, and other nutrients, all of which may be limited or missing when nutrition is poor. Metabolic influx and derivatives are necessary for the vigorous activities of stem cells, including the function of chromatin-modifying enzymes.^34,35 Increased cellular glycolysis supports the self-renewal of NPCs, whereas low glycolytic activity supports their differentiation.³⁶

To achieve a better understanding of the underlying mechanisms connecting maternal nutrition during pregnancy with nephron endowment, we established a calorie-restricted diet (CRD) mouse model during pregnancy. Undernutrition reduced NPC proliferation and mTORC1 pathway activity. Moreover, we show that the amino acid methionine or its metabolites are the key limiting factors in nephrogenesis during maternal CRD. The effect of caloric restriction can be compensated for by either removing one allele of the mTORC1 inhibitor, Tsc1, or by methionine supplementation.

Methods

Mice Husbandry

All mice were maintained at the Hebrew University Specific Pathogen–Free Animal Facility Unit. All experiments were performed in CD1 outbred mice, including transgenic mice. When stated, we used the transgenic mice lines Tg(Six2-EGFP/cre)1Amc (herein Six2 Cre^+/tg) and Tsc1^f/f.³⁷ Heterozygous Six2 Cre^+/tg embryos were used in order to sort GFP⁺ NPCs, generated by mating 6- to 8-week-old Six2 Cre^+/tg male mice with wild-type (WT) CD1 females. Six2 Cre^tg/+ males were mated with Tsc1^f/f females to generate Six2 Cre^+/tg Tsc1^+/f pups, which were identified using genotyping by the following primers—Six2 Cre forward: GCATTACCGGTCGATGCAACGAG; Six2 Cre reverse: GAGTGAACGAACCTGGTCGAAATCAGTGCG; Tsc1 forward: CAGCTCCGACCATGAAGTG; and Tsc1 Reverse: AGGAGGCCTCTTCTGCTACC. For all mice, the pregnancy date was determined by vaginal plug expulsion. The morning of plug detection was designated as day 0.5 of pregnancy. In our preliminary experiment, pregnant females were separated into individual cages, and their daily food consumption was monitored throughout pregnancy (data not shown). On the basis of the average daily food consumption, 70% of the average daily intake was provided once daily to CRD-treated pregnant females for all of the experiments as described. Unrestricted diet (URD)–treated mice were not subjected to any food restriction during pregnancy. At different embryonic dates, pregnant females were euthanized using ketamine/xylazine and cervical dislocation. The embryos and newborn pups were dissected, and kidneys were excised. In some experiments, bromodeoxyuridine (BrdU; B5002; Sigma-Aldrich) was diluted in PBS and injected intraperitoneally to pregnant females (100 mg/kg) 24 hours before being euthanized, and the kidneys were excised for immunostaining. In some experiments, the drinking water was supplemented with 0.5% l-methionine (64319; Sigma-Aldrich) or 0.5% l-Asparagine (A0884; Sigma-Aldrich).

Nephron Count

Nephron count was performed as previously described.^37,38 Briefly, a single kidney of every pup was dissected. The kidneys were chopped into small pieces and digested in 5 ml of 6 N HCl at 37°C for 90 minutes. Tissue was further dissociated by repeated pipetting. Twenty-five milliliters of ddH₂O was added to each sample and kept at 4°C. To count the number of nephrons, 200 μl of well-mixed suspension was placed in a 1-×1-cm area of a P100 culture dish marked with gridlines. The number of glomeruli was counted under an inverted microscope three times. The total number of glomeruli was estimated as follows: total nephron number per kidney equaled the average measured number of glomeruli in 30 ml of the suspension.

FACS of Six2 Cre^+/Tg Nephron Progenitor Cells

Kidneys were excised in ice-cold HBSS buffer. The kidneys were sliced and chopped into approximately 0.5- to 1-mm pieces on ice using a surgical scalpel. The chopped kidneys were transferred into HBSS solution containing 1 μg/μl collagenase/dispase (10269638001; Sigma-Aldrich) and incubated for 25 minutes at 37°C. The cells were filtered through a 40-μm nylon cell strainer (Corning) and washed twice with cold HBSS. Six2 Cre^+/tg GFP⁺ NPCs were isolated by cell flow cytometry-based cell sorting (Hebrew University Faculty of Medicine Core Facility). For RNA extraction, 80,000–130,000 GFP⁺ NPCs in each biological sample were FACS-based sorted, washed cells were washed in cold HBSS, and total RNA was extracted using peqGOLD TriFast (PeqLab Biotechnologie). As cell number in each group varied, each sample was normalized to housekeeping genes as mentioned.

Metabolomics Analyses

For the measurement of intracellular metabolites, 250,000 GFP⁺ NPCs in each biological sample were used, and the metabolites were washed with 2 ml ice-cold PBS twice on ice. The cells were extracted quickly on dry ice by scraping in 0.2 ml methanol/acetonitrile/water (50:30:20, vol/vol/vol) solution at −20°C. All of the metabolite samples were stored at −80°C for at least 2 hours. Protein-free metabolite extractions were prepared by spinning the samples at 20,000×g for 20 minutes at 4°C twice. Samples were subsequently analyzed using the liquid chromatography-mass spectrometry method. Samples were chromatographically separated on a SeQuant ZIC-pHILIC column (2.1×150 mm, 5 μm; EMD Millipore) using an HPLC system (Ultimate 3000 Dionex LC system; Thermo Fisher Scientific, Inc.). The flow rate was set to 0.2 ml min⁻¹, column compartment was set to 30°C, and the autosampler tray was maintained at 4°C. Mobile phase A consisted of 20 mM ammonium carbonate and 0.01% (vol/vol) ammonium hydroxide. Mobile Phase B was 100% acetonitrile. The mobile phase linear gradient (percentage B) was as follows: 0 minutes, 80%; 15.0 minutes, 20%; 15.1 minutes, 80%; and 23.0 minutes, 80%. Mass spectrometry detection was performed using a Q Exactive Hybrid Quadrupole Orbitrap high-resolution mass spectrometer with an electrospray ionization source (Thermo Fisher Scientific, Inc.). Ionization source parameters were the following: sheath gas, 25 U; auxiliary gas, 3 U; spray voltage, 3.3 and 3.8 kV in negative and positive ionization mode, respectively; capillary temperature, 325°C; S-lens RF level, 65; and auxiliary gas temperature, 200°C. Metabolites were analyzed in the range of 72–1080 m/z. The retention time of metabolites in the chromatogram was identified by corresponding pure chemical standards. Data were analyzed with MAVEN and commercially available pathway analysis software, MetaboAnalyst (www.metaboanalyst.ca/).³⁹

NPC Cell Cycle Analysis

For NPC cell cycle analysis, pooled embryonic kidneys from each litter were chopped and dissociated as described above. Five thousand sorted NPCs were incubated with HBSS solution containing 10 μg/ml Hoechst 33342 (Sigma-Aldrich) for 20 minutes at room temperature before analysis by LSRII flow cytometry and FlowJo software.

RNA Sequencing and Quantitative RT-PCR Validation

The integrity of the RNA was evaluated by the TapeStation system using an RNA ScreenTape kit (Agilent Technologies). RNA was quantified using a Qubit apparatus (Qubit DNA HS Assay kit; Invitrogen). Libraries were prepared from RNA samples using a KAPA Stranded mRNA-Seq Kit (https://rochesequencingstore.com/wp-content/uploads/2017/10/KAPA-StrandedmRNASeq-Kit_KR0960-%E2%80%93-v6.17.pdf). The libraries were bar coded and pooled for multiplex sequencing (1.5 pM total, including 1.5% PhiX control library). The pooled cDNA was loaded on the NextSeq 500 High Output v2 kit (75-cycles) cartridge (Illumina) and sequenced on the Illumina NextSeq 500 System using sequencing conditions of 75 cycles, single read. Library preparation and sequencing were performed at the Core Facility of the Hebrew University Faculty of Medicine. For further validation, RNA was extracted and used for quantitative RT-PCR with the following primers: mKI-67 forward: 5′-TCA​CCT​GGT​CAC​CAT​CAA​GC-3′; mKI-67 reverse: 5′-TCA​ATA​CTC​CTT​CCA​AAC​AGG​CA-3′; Top2a forward: 5′-TGG​TTT​TAC​GGA​GCC​AGT​TTT-3′; Top2a reverse: 5′-TCA​CGT​CAG​AGG​TTG​AGC​AC-3′; GAPDH forward: 5′- ACC​CTT​AAG​AGG​GAT​GCT​GC-3′; and GAPDH reverse: 5′- CCC​AAT​ACG​GCC​AAA​TCC​GT-3′.

Gene Set Enrichment Analyses

Cutoff-dependent enrichment pathway analysis of the significantly differentiated genes (P_adj<0.1) was done using Ingenuity Pathway Analysis (QIAGEN Inc.; https://digitalinsights.qiagen.com/products-overview/discovery-insights-portfolio/content-exploration-and-databases/qiagen-ipa/). Differential expression data of NPCs obtained from unrestricted diet versus calorie-restricted mice were subjected to gene set enrichment analysis (GSEA; Broad Institute). GSEA uses ranked differential expression data (cutoff independent) to determine whether prior-defined sets of genes show statistically significant and concordant differences between two biologic states. GSEA was run against the hallmark gene set collections from the molecular signatures database.

Section Preparation and Immunostaining

Kidneys at different embryonic ages, as indicated, were dissected in ice-cold PBS and fixed in fresh 4% formaldehyde in PBS. Kidneys were embedded in paraffin and sectioned. For overall histology, tissue sections were stained with hematoxylin/eosin. IF staining was performed as previously described.³⁷ Briefly, paraffin-embedded tissue sections (4–6 μm) were deparaffinized, hydrated, and incubated overnight at 4°C with the following reagents, according to the manufacturer’s instructions: rabbit anti-Six2 antibody (11562–1-AP, 1:200; Proteintech), mouse anti-BrdU (MS-1058-P, 1:500; NeoMarkers), mouse antiproliferating cell nuclear antigen (anti-PCNA; 180110, 1:500; Invitrogen), mouse anti-S6 ribosomal protein (2317, 1:100; Cell Signaling ), and rabbit antiphospho-S6 ribosomal protein (pS6; 2211, 1:100; Cell Signaling). The sections were incubated with either Cy3-conjugated goat anti-rabbit or Cy5-conjugated goat anti-mouse antibody according to the manufacturer’s instruction (Jackson Immuno Research Laboratories). In some experiments, Six2 Cre^+/tg NPCs were sorted as previously described, spun onto glass slides using Cytospin 4 Cytocentrifuge, air dried, fixed in 4% PFA, and stained with the indicated antibodies. Sections and slides were visualized with a confocal A1R microscope. For the detection of apoptotic cells, kidney sections were prestained with anti-Six2 antibody before the detection of apoptotic cells using the In Situ Cell Death Detection Kit, TMR red, according to the manufacturer’s instructions (12156792910; Roche).

Western Blotting

Kidneys were extracted and homogenized in cold RIPA buffer containing 150 mM NaCl, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, and 25 mM Tris (pH 7.4) supplemented with protease/phosphatase inhibitors (4906837001; MERCK). An equal amount of protein extract was analyzed by SDS-PAGE as previously described using the antibodies as above, rabbit anti-pS6 ribosomal protein, and mouse anti-S6 ribosomal protein (2317; Cell Signaling), according to the manufacturer’s instructions.^40,41

Biochemical Analyses

For kidney function assessments, serum from the indicated mice in each group was collected and analyzed for serum urea using the QuantiChrom Urea assay kit (DIUR-100; BioAssay Systems). Serum levels of creatinine were determined by using the Cobas C-111 chemistry analyzer (Roche). Glucose and β-hydroxybutyrate (β-HB) levels were measured at embryonic days 12.5 and 18.5 twice daily before and 8 hours after feeding. Both serum glucose and β-HB levels were measured using FreeStyle Optium test strips.

Kidney Organ Culture and Whole-Mount Kidney Staining

For kidney organ cultures, kidneys were dissected, immediately placed on SPL Insert Hanging 0.4-µm polycarbonate membranes (35024; Greenpia Technology), and incubated in a 24-well plate containing DMEM high-glucose media either with (01–056–1A, 30 mg/L) or without l-methionine (06–1052–13–1A; Biologic Industries), 100 µM S-(5′adenosyl)-l-methionine chloride (A7007; Sigma-Aldrich), and 100 µM S-(5′adenosyl)-l-homocysteine (A9384; Sigma-Aldrich) supplemented with 10% FCS for 48 hours. In different experiments, as indicated, 24 hours after the kidneys were incubated, the media were supplemented with 10 μg/ml BrdU. At the end of the incubation time, the kidneys were fixed and either sectioned or whole-mount immunostained.

Whole-mount staining of the cultured kidneys was done as previously described.^37,42 Briefly, kidneys at different embryonic ages as indicated were excised in ice-cold HBSS buffer and fixed in cold 4% PFA for 13–14 minutes, depending on embryonic age. The kidneys were washed using PBS and preincubated in PBS/0.5% Triton for 2 hours. The kidneys were then incubated in CAS-Block (008120; Life Technologies) containing the antibodies rabbit anti-Six2 antibody (11562–1-AP, 1:200; Proteintech), mouse anti-BrdU (MS-1058-P, 1:500; NeoMarkers), and mouse anti-Cytokeratin (C-2562, 1:100; Sigma-Aldrich) for 48 hours according to the manufacturer’s instructions. After extensive washing, the kidneys were incubated with either Cy3-conjugated goat anti-rabbit or Cy5-conjugated goat anti-mouse antibody (Jackson Immuno Research Laboratories) for an additional 48 hours before washing. The kidneys were mounted using Fluoromount Aqueous Mounting solution (F4680; Sigma-Aldrich), visualized using the Nikon Multiphoton A1MP microscope, and analyzed using Imaris version 9.0.2 software.³⁷

Statistical Analyses

Values are presented as means ± SEM. Statistical significance was determined using a two-tailed t test or where indicated, one-way ANOVA followed by post hoc Tukey multiple range test (P=0.05).

Results

Nephron Endowment and Kidney Function Are Impaired in Offspring of Calorie-Restricted Female Mice

To study the effect of nutrition during pregnancy on nephron endowment, we established a CRD mouse model during pregnancy. Plugged WT CD1 females were separated into discrete cages and randomly assigned to URD (ad libitum) or a diet containing 70% of the average daily food consumed by the ad libitum group (herein CRD). Although the weight of the pregnant females in the CRD group was indistinguishable from URD females (Figure 1A), the weight of CRD offspring was significantly lower at P0 (Figure 1, B and E). There was no effect on litter size (Supplemental Figure 1). Kidney weight was reduced in the CRD offspring. However, the kidney-body weight ratio was unaffected (Figure 1, C and D). The CRD pups had complete catchup growth for weight already at age P28 in both males and females (Supplemental Figure 2). To see whether the reduced kidney weight is associated with low nephron number, kidneys of CRD and URD offspring were excised at P20, and nephron number was measured by acid maceration. The nephron number was at least 50% lower in CRD relative to control URD pups (Figure 1F). Kidney-body ratio was similar in CRD and URD pups at P30 (Supplemental Figure 3). We then asked if this reduction in nephron number affects kidney function in adulthood. Serum urea levels of both young (P30) and older (P200) mice were higher in the offspring of the CRD group compared with the URD group (Figure 1, G and H). Serum creatinine levels at P30 were stable (Supplemental Figure 4). Therefore, a reduction of 30% in the daily caloric intake during pregnancy leads to a decrease in nephron number, as well as an impairment in kidney function in offspring at young and older ages.

Figure 1.

Caloric restriction during pregnancy impairs embryonic nephron endowment and kidney function. (A) The weight gain of pregnant CD1 female mice with CRD or URD relative to the first day of pregnancy, presented as percentage of weight gain (n=4 CRD, n=6 URD). (B) The weight of pups from CRD- and URD-fed mothers at P0 (n=19–21, three litters). (C) Weight of two kidneys of each pup as in (B) at P0. (D) Weight of the two kidneys as in (C) compared with total body weight at P0 (n=19–21, three litters). (E) Representative pups from CRD and URD groups at P2 and P7. (F) Nephron number at P20 of mice as above, counted using acid maceration (n=6 URD, n=6 CRD). (G and H) Serum urea of mice from CRD or URD mothers at ages P30 and P200 (n=7 URD, n=5 CRD). *P=0.05. P, postnatal day.

Caloric Restriction during Pregnancy Modifies Gene Expression in Nephron Progenitor Cells and Reveals Decreased Cell Proliferation and mTORC1 Pathway Signaling

To study the effect of caloric restriction during pregnancy on gene expression and signaling pathways in NPCs, Six2 Cre^+/tg males were mated with WT CD1 females. The plugged WT CD1 females were randomly assigned to either CRD or URD. Half of the embryos are expected to be Six2 Cre^+/tg genotype with GFP⁺ NPCs for sorting purposes. The transgene expresses eGFP only in the NPCs to enable GFP⁺ NPCs sorting by FACS. Kidney GFP⁺ NPCs of E18.5 embryos were isolated from Six2 Cre^+/tg offspring of females assigned to CRD or URD (Figure 2A). CRD Six2 Cre^+/tg offspring had a reduced nephron number compared with URD Six2 Cre^+/tg offspring (Supplemental Figure 5), similar to WT CD1 CRD offspring (Figure 1F). Gene expression was analyzed using RNA sequencing to identify genes and pathways that were affected by caloric restriction during pregnancy (Figure 2, B and C, Supplemental Figure 6). Among the different pathways identified, RNA sequencing and pathway analysis identified major perturbations in mTORC1 and G2M cell cycle regulation in kidneys of embryos subjected to caloric restriction. Additionally, there was a decrease in genes related to oxidative phosphorylation (Figure 2, D and E). We proceeded to validate the effects of CRD on the cell cycle and the mTORC1 pathway.

Figure 2.

Caloric restriction during pregnancy modifies expression profiles of NPCs, including mTORC1 pathway and cell cycle control. (A) E18.5 kidneys were resected from Six2 Cre^+/tg embryos of CRD and URD mothers. Six2⁺ GFP⁺ NPCs were sorted by FACS. (B and C) Heat map visualization and principal component analysis of RNA sequencing performed on sorted NPCs from Six2 Cre^+/tg embryos of CRD- and URD-fed mothers (n=3 litters, 12 mice on average in each litter). (D) Selected Ingenuity Pathway Analysis–enriched canonical pathways (BH-P value >1.3). Bar plot showing changes in the expression of various pathways on the basis of gene expression analysis of the data as in (B). (E) GSEA analysis of selected pathways in NPCs from embryos of CRD compared with URD mothers. GSEA of the whole-transcriptomic data of NPCs obtained from CRD versus URD. Molecular signatures database hallmark of mTORC1 signaling (left panel), G2M checkpoint (center panel), and oxidative phosphorylation (right) gene sets were significantly enriched (FDR<0.0001) in the downregulated genes of the NPCs obtained from CRD versus URD. EIF2, eukaryotic initiation factor 2; eIF4, eukaryotic initiation factor 4; FDR, false discovery rate; FITC-A, fluorescein isothiocyanate A; FSC-A, forward scatter area; G2M, gap 2 mitosis; GP6, glycoprotein 6; NES, Normalized Enrichment Score; NRF2, nuclear factor erythroid 2-related factor 2; PTEN, phosphatase and tensin homolog; TCA cycle, tricarboxylic acid cycle.

NPC Proliferation, but Not Apoptosis, Is Affected by Caloric Restriction during Pregnancy in the Embryonic Kidney

NPCs proliferate continually to compensate for the splitting of the niche by the emerging branch tips and replace NPCs that exit the niche and differentiate into new nephrons.^43,44 The NPC expression profiles suggest that maternal CRD reduced their proliferation rate (Figure 2). To validate these findings, we measured the expression of key proliferation markers, KI67 and Top2a by quantitative RT-PCR in RNA isolated from Six2 Cre^+/tg CRD-exposed NPC at E15.5, relative to URD NPC (Figure 3A). For an orthogonal validation, we coimmunostained kidney sections from WT CD1 URD and CRD-exposed offspring for NPCs (Six2) and the proliferation marker PCNA at ages E15.5 and P2. The fraction of PCNA, Six2 double-positive NPCs was significantly lower in the offspring of CRD at E15.5 (Figure 3, B and C). However, after birth, CRD and URD kidneys were indistinguishable (Figure 3, B and D). Finally, to directly measure the effect of maternal CRD on NPC proliferation, we injected BrdU, a thymidine analog that incorporates into the DNA during S phase. BrdU was injected into pregnant WT CD1 females subjected to either CRD or URD at E14.5, which is characterized by high proliferation rate.⁴⁴ The embryos were removed 24 hours after injections, and the kidneys were processed for histology (Figure 3E). Punctuated, nuclear BrdU labeling was evenly distributed in the NPCs, indicating S phase in progress. Importantly, BrdU labeling in Six2⁺ NPCs from CRD kidneys was significantly reduced relative to URD (Figure 3, F and G). There was no difference in NPC apoptosis, as demonstrated by terminal deoxynucleotidyl transferase–mediated digoxigenin-deoxyuridine nick-end labeling (TUNEL) assay of Six2⁺ cells at E15.5 (Figure 3H, Supplemental Figure 7). For reinforcement, kidneys of CRD- and URD-exposed Six2 Cre^+/tg embryos were excised at E15.5, dissociated, stained with Hoechst, and analyzed by flow cytometry. CRD led to a significant decrease in the G2M cell population in GFP⁺ NPCs, concomitant with an increase in the G1-phase cell population (Figure 3, I and J). Collectively, these results indicate that CRD during pregnancy reduces NPCs proliferation in the offspring with no significant effect on apoptosis, which could account for the reduction in the nephron number in these embryos (Figure 1).

Figure 3.

Caloric restriction during pregnancy alters cell cycle and proliferation in offspring NPCs. (A) Six2⁺ GFP⁺ NPCs were extracted and FACS sorted from E15.5 kidneys of Six2 Cre^+/tg embryos from CRD and URD mothers, and the expression of proliferation markers Ki67 and Top2a was quantified by quantitative RT-PCR (n=4, three litters). (B) Renal sections of E15.5 and P2 of CRD and URD WT mothers stained for Six2 (NPCs), PCNA, and DAPI. Scale bars: 20 μm. Magnification: ×40. (C and D) Quantification of proliferating NPCs as in B (n=4 from the same litter). (E) Schematic representation of the experimental time course for BrdU. BrdU was injected to pregnant females subjected to either URD or CRD at E14. Embryonic kidneys were resected after 24 hours. (F) Kidneys were stained for BrdU, Six2, and DAPI. Scale bars: 20 μm. Magnification: ×40. (G) Quantification of proliferating NPCs as in (F) (n=9 from 2 litters). (H) Quantification of Six2 and terminal deoxynucleotidyl transferase–mediated digoxigenin-deoxyuridine nick-end labeling (TUNEL) immunostaining of E15.5 kidney sections of WT embryos of URD- and CRD-fed mothers, presented as the percentage of apoptotic NPCs in each group (n=7 sections from different kidneys, two litters) (Supplemental Figure 6). (I) The cell cycle plot of E15.5 Six2 Cre^+/tg NPCs by FACS analysis. (J) Quantification as for (I), demonstrating a decrease in the percentage of cells at the G2/M phase in the CRD NPCs (n=3, three litters). *P=0.05. E, embryonic day.

Maternal Caloric Restriction Decreases mTORC1 Activity in Embryonic Kidney Progenitor Cells

We previously showed that a removal of one mammalian target of rapamycin (mTOR) allele in NPCs was detrimental to nephrogenesis.³⁷ mTORC1 functions as a sensor of the metabolic environment, but its role in mediating the changes in the intrauterine environment on kidney development was not studied.³⁷ Ribosomal protein (S6) is a key target of mTORC1, and its phosphorylation by S6K enhances the translation of a set of proteins, among them those that are involved in proliferation.^41,45,46 To test if mTORC1 activity is altered in response to caloric restriction during pregnancy, whole kidneys of CRD- and URD-treated offspring were excised, and mTORC1 activity was monitored. Western blotting showed that CRD reduced pS6 of S6 ratio in kidney extracts from CRD E15.5 embryos and P0 pups (Figure 4, A and B). Measuring pS6 levels in NPCs by western blot is challenging due to the low cell number and the difficulty of separating them from the rest of the kidney parenchyma. Therefore, GFP⁺ NPCs from E15.5 kidneys isolated from Six2 Cre^+/tg offspring of CRD and URD mothers were FACS sorted, cytospinned, and immunostained for either S6 or pS6. Although NPCs from both URD- and CRD-treated embryos contained similar amounts of total S6 ribosomal protein, pS6 was only detected in URD NPCs with only negligible staining in CRD embryo kidneys (Figure 4C). Our results indicate that caloric restriction impairs mTORC1 activity in NPCs.

Figure 4.

mTORC1 activity mediates the effect of caloric restriction during pregnancy on NPCs. (A) Western blot for pS6 and total ribosomal S6 in WT kidney extracts from E15.5 and P0 offspring of CRD and URD mothers. (B) Quantification of the western blots as in (A) at the indicated ages (n=4–5, from the same litter). (C) Representative images of cytospinned NPCs from URD and CRD kidneys of Six2 Cre^+/tg mice at E15.5 stained for pS6 or S6 and DAPI, Scale bars: 20 μm. Magnification: ×40. (D) Nephron number of kidneys from Six2 Cre^+/tg Tsc1^+/f (TSC1^+/−) mice offspring (P28–P32) subjected to URD or CRD (n=5 URD, n=10 CRD, two different litters). (E) Serum urea of Six2 Cre^+/tg Tsc1^+/f (TSC1^+/−) mice subjected to CRD or URD at P200 (n=20 URD, n=11 CRD, three different litters). *P=0.05. DAPI, 4′,6-diamidino-2-phenylindole; E, embryonic day; P, postnatal day.

mTORC1 Is a Central Mediator of Caloric Restriction on Nephron Progenitor Cells, and Removal of Tsc1 Rescues the Impaired Renal Development

The decrease in mTORC1 activity in NPCs, the reduction in nephron number in Six2 Cre^+/tg mTOR^+/f pups,³⁷ and the effect of low-calorie diet on kidney development all suggest that mTORC1 mediates the effect of maternal environment on offspring NPCs. This hypothesis motivated us to study if a genetic reduction of a negative mTORC1 regulator will ameliorate the negative effect of CRD on nephron number. For this purpose, Six2 Cre^+/tg males were mated with Tsc1^f/f females to generate litters with the offspring genotypes Six2 Cre^+/tg Tsc1^+/f and Tsc1^+/f. Pregnant females were randomly assigned to URD or CRD, and nephron number was measured. No change in nephron number was detected when comparing URD with CRD-treated Six2 Cre^+/tg Tsc1^+/f offspring (Figure 4D). Moreover, we could not detect any significant difference in the serum urea levels between URD and CRD Six2 Cre^+/tg Tsc1^+/f offspring at P200 (Figure 4E). These data collectively indicate that the mTORC1 pathway is a crucial sensor of caloric restriction in NPCs and that mTORC1 pathway activation by a reduction in hamartin levels rescues the effect of CRD on nephrogenesis.

Caloric Restriction in Pregnancy Reduces the Levels of Methionine and Metabolites in Nephron Progenitor Cells

The data thus far indicate that CRD impairs nephrogenesis by affecting the mTOR pathway and cell cycle, perhaps downstream of mTORC1. Caloric restriction during pregnancy can also independently restrict the supply of essential metabolites to the NPCs. Identifying these metabolites can improve our ability to mitigate the effect of poor nutrition on nephrogenesis. To analyze the metabolites in NPCs, we FACS-sorted GFP⁺ NPCs from E15.5 Six2 Cre^+/tg kidneys isolated from embryos of CRD and URD mothers, as described above. The metabolites were extracted and analyzed by hybrid triple-quadrupole mass spectrometry (Figure 5A). Among the different pathways identified, we focused on methionine metabolism, which was affected by CRD (Figure 5B). Methionine is an essential amino acid supplied by nutrition. Methionine and its metabolites play crucial roles in regulating translation, DNA methylation, and antioxidant balance.^47,48 The dietary oscillation in methionine supply alters the levels of metabolic substrates in one-carbon metabolism (e.g., S-adenosylmethionine [SAM] and S-adenosylhomocysteine [SAH]) and changes the expression of genes involved in DNA methylation (Figure 5C).⁴⁸ Indeed, we found lower levels of methionine, as well as its derivatives SAM and SAH, in NPCs from CRD E15.5 embryos compared with URD embryos, as was measured by mass spectrometry (Figure 5D).

Figure 5.

Maternal malnutrition impairs metionine metabolism in NPCs from offspring of CRD pregnant mice. (A) NPCs were sorted from kidneys of CRD- and URD-treated Six2 Cre^+/tg mice at E15.5. Metabolites were extracted and analyzed from cell homogenates. The main modified metabolites are presented by heat map (n=3, repeated at least twice). (B) The bar plot shows key metabolic pathways affected by CRD. The methionine pathway is highlighted. (C) Scheme of methionine metabolism. (D) The levels of methionine and its metabolites in NPCs extracted from embryo kidneys of URD- and CRD-treated Six2 Cre^+/tg mice as measured by mass spectrometry (n=3 samples). (E) Representative images and (F) quantification of NPCs in the metanephric niche in cultured kidneys. E13.5 WT embryonic kidneys were incubated in organ culture for 48 hours in standard media, low-methionine media (−Met), low-methionine supplemented with 100 μM SAM media (−Met +SAM), or low methionine supplemented with 100 μM SAH media (−Met +SAH). The cultured kidneys were stained with markers for NPCs and UBs: Six2 and cytokeratin, respectively (n=4–7 kidneys, three different litters). Statistical significance was determined using one-way ANOVA followed by post hoc Tukey multiple range test. (G and H) Representative images and quantification of proliferating NPCs stained for Six2⁺ BrdU⁺ in cultured kidneys. E13.5 WT embryonic kidneys were incubated in organ culture for 24 hours in standard or low-methionine media, followed by an additional 24 hours of incubation with BrdU (control n=3, −Met=4, two different litters). *P=0.05. E, embryonic age.

Nephrogenesis Is Impaired after Methionine Deprivation in Ex Vivo Organ Culture

Although methionine and its metabolites are essential in different types of embryonic cells,⁴⁷ their role in nephrogenesis has not been studied. We first used organ cultures to ask if methionine and its metabolites are important to kidney development. Embryonic kidneys were excised at E13.5, cultured in methionine-free or methionine-containing medium for 48 hours, and then, whole-mount stained for Six2 (NPC) and cytokeratin (a UB marker) to determine the number of NPCs per UB tip. Methionine deprivation reduced the number of NPCs per tip. Supplementing the methionine-free medium with the major metabolites of methionine, SAM or SAH, restored the number of NPCs per tip to the levels seen with methionine (Figure 5, E and F). Therefore, we concluded that caloric restriction decreases the levels of methionine and its derivatives in NPCs. In kidney organ cultures, methionine deprivation impairs embryonic kidney development, and supplementation with methionine and its metabolic products reverses this effect.

Low Methionine Reduces Proliferation of Nephron Progenitor Cells

Caloric restriction in pregnant mice decreased embryonic NPC proliferation (Figure 3). To study the effect of methionine deprivation on NPC proliferation, kidneys from E13.5 were excised and cultured in media with or without methionine for 48 hours. BrdU was added during the last 24 hours of incubation (Figure 5, G and H). Kidney sections were stained for Six2 and BrdU as markers for NPCs and proliferation, respectively. The percentage of double-stained Six2⁺/BrdU⁺ cells representing proliferating NPCs was decreased in the low-methionine kidneys compared with the control kidneys incubated in methionine-containing medium. These results indicate that methionine is a crucial factor for NPC proliferation and suggest that the methionine metabolic pathway contributes to nephrogenesis.

Dietary Methionine Alleviates the Effect of Caloric Restriction on Nephron Endowment

To determine if methionine is sufficient to reverse the effect of CRD on nephron endowment and kidney function, we added methionine to the drinking water of calorie-restricted pregnant mice to determine if this improves nephrogenesis in the offspring. WT CD1 pregnant mice were subjected to CRD without or with 0.5% l-methionine in the drinking water.⁴⁹ At P22–P30, the kidneys of the offspring in the indicated groups were excised, and nephron numbers were measured. Methionine supplementation to CRD pregnant mice significantly increased nephron numbers in offspring, compared with CRD-treated offspring subjected to standard drinking water (Figure 6A), with no effect on body weight or kidney-body ratio (Figure 6B, Supplemental Figure 8). Similar effects on nephron number were observed in both males and females (Supplemental Figure 9). Methionine is a gluconeogenic amino acid and may prevent the effect of CRD on nephrogenesis by modifying carbohydrate metabolism. However, methionine supplementation did not affect serum glucose or ketone bodies level in CRD pregnant mice in E12.5 or E18.5 (Figure 6, C and F, Supplemental Figure 10). In addition, supplementation of drinking water with asparagine, which also functions as a gluconeogenic amino acid, did not improve nephrogenesis (Figure 6A). Serum β-HB was high in both E12.5 and E18.5 in either URD or CRD mice as ketogenesis level is increased during pregnancy (Figure 6, E and F, Supplemental Figure 9).⁵⁰ Ketones level was even higher at E18.5 before feeding (Figure 6E). A ketogenic diet can slow the progression of cystic kidney disease independently of caloric restriction.⁵¹ However, methionine prevented impairment in nephrogenesis in CRD with no effect on the level of β-HB at the end of pregnancy. The additional independent effect of ketone bodies on nephrogenesis cannot be ruled out.

Figure 6.

Dietary methionine (Met) alleviates the effect of caloric restriction on nephron endowment. (A) Nephron number of P22–P30 offspring kidneys from WT mice subjected to CRD, CRD supplemented with 0.5% Met, or CRD supplemented with 0.5% asparagine (Asn) in the drinking water throughout pregnancy (n=10–22, at least two litters in each group). The dashed line represents the average nephron number of URD pups in the same age measured previously. (B) Two kidney to body ratio of P0 pups from mothers subjected to CRD, CRD supplemented with Met, or CRD supplemented with Asn. The data of the CRD group also appear in Figure 1D and were collected from the same experiment. (C and D) Maternal serum glucose level of mothers subjected to URD, CRD, CRD supplemented with Met, or CRD supplemented with Asn at E18.5 (C) before or (D) 8 hours after feeding (n=4–6). (E and F) Maternal β-HB levels of mothers subjected to URD, CRD, CRD supplemented with Met, or CRD supplemented with Asn at E18.5 (E) before or (F) 8 hours after feeding (n=4–6). Statistical significance was determined using one-way ANOVA followed by the Fisher LSD test. *P=0.05. E, embryonic day; P, postnatal day; LSD, least significant difference.

Discussion

Here, we describe the cellular and metabolic pathways that mediate the effect of caloric restriction during pregnancy on NPCs and suggest a supplementation strategy that can improve kidney development and prevent kidney disease in adulthood (Figure 7). We show that a modest 30% reduction in caloric intake during pregnancy reduces nephron number in offspring, resulting in an increased risk of kidney disease at an older age. Maternal caloric restriction modified the cell cycle of NPCs, decreasing proliferation. By analyzing the NPCs expression profiles, we show significant changes in pathways downstream of mTORC1 activity in this mouse model. We validated the findings by showing that the levels of pS6, the major downstream target of mTORC1, are reduced in the embryonic CRD kidneys. Importantly, the reduction in Tsc1, a component of the primary inhibitor of the mTORC1 pathway, prevented the decrease in nephron endowment by caloric restriction. Using metabolomic analysis, we discovered that caloric restriction during pregnancy impaired the methionine metabolism pathway in the NPC. Methionine deprivation impaired the number of NPCs per tip and their proliferation rate in organ culture. Replenishing methionine-deprived medium with metabolic intermediates of methionine improved nephrogenesis, as demonstrated by an increase in the number of NPCs in the metanephric mesenchyme. Importantly, methionine supplementation in the drinking water of calorie-restricted pregnant mice ameliorated the harmful effect of caloric restriction on nephron endowment in the offspring.

Figure 7.

Caloric restriction during pregnancy modifies cellular and metabolic activities in NPCs, including cell proliferation, mTORC1 activity, and methionine metabolism. The resulting low nephron number increases the risk of CKD in adulthood.

Caloric restriction during pregnancy led to major effects on the expression profiles of NPCs, including pathways that play a role in the cell cycle, epigenetics, and metabolic activity of the cell.

As could have been predicted, we found that mTORC1 activity is significantly reduced by the change in maternal nutritional status. mTORC1 is extremely sensitive to a favorable cellular growth environment, such as amino acid levels, growth factors like IGF1, and energy. As the main intrinsic regulator of cellular metabolism, mTORC1 integrates these inputs to control the rates of cellular activities, such as proliferation, metabolism, and differentiation.^37,52,53 Indeed, deletion of one mTOR allele in NPCs impaired kidney development to a similar extent as caloric restriction.³⁷ TSC is a heterodimer acting as a key negative regulator of mTORC1. Hemizygous deletion of the Tsc1 gene in NPCs accelerates nephrogenesis by a modest expansion of NPCs lifespan that increases nephron numbers in a mouse model.³⁷ Our data indicate that caloric restriction during pregnancy can be reversed by TSC1 deletion, most probably by altering mTORC1 signaling within the NPCs. However, Tsc1 may also act by mTORC1-independent cellular activities.³⁷ mTORC1 is a major metabolic sensor. The mechanism of mTORC1 control in NPCs is still obscure. Although mTORC1 senses the environment by different signaling, such as growth factors, it can also interacts with other metabolic sensors as AMP-activated protein kinase, which may affect both TSC and mTORC1 activities independently.^54,55

NPCs can either proliferate and increase the pool of the progenitor cells or differentiate into 1 of over 20 cell types comprising the mature nephron. This balance is controlled in part by FGF20,^38,56 GDNF,⁵⁷ and stromal signals.^58–61 The fate of NPCs is also determined by interactions with adjacent NPCs as well as other cell populations in the embryonic kidney.⁶² In addition, the branching pattern of the UB may have an important effect on the nephron number as well.⁶³ There is a constant reciprocal interaction between NPCs and UB tip cells. Wnt9b secretion by the UB affects the self-renewal and differentiation of NPCs.^7,64 Therefore, maternal malnutrition may interrupt kidney development by affecting various cell populations in the embryonic kidney.

The effect of a low-protein diet on kidney development emphasizes the role of amino acid supply in kidney development.^25–30 Interestingly, using metabolomics, methionine metabolism has a pivotal role in mediating the effect of caloric restriction on kidney development. Methionine is an essential amino acid that must be supplied by nutrition. Methionine has crucial roles in translation regulation, DNA methylation, and antioxidant balance, as well as in different types of embryonic cells.⁴⁷ Changes in methylation of gene promoters and enhancers in the embryonic kidney are essential for kidney development.^65,66 Different components of the methionine pathway serve as methyl donor groups for DNA methylation.^48,67 Therefore, epigenetic changes may contribute to the changes in gene expression in NPCs as a result of caloric restriction through methionine deprivation. The role of epigenetics in mediating the effect of nutrition together with other metabolic pathways that were revealed in this metabolomic assay opens new portals to understanding renal embryogenesis and the effects on kidney function later in life.

In summary, we report that the untoward effects of caloric restriction during pregnancy, which lead to low nephron number at birth with long-term effects on kidney health in adulthood, are mediated in large part by methionine deficiency and mTORC1. Although caloric restriction may affect all cell types in the kidney and act through multiple cellular, signaling, and metabolic changes, these effects can be ameliorated by elevating mTORC1 activity in the NPCs or by supplementing the restricted diet with methionine. These observations may provide new targets to improve kidney development in high-risk, marginalized populations to reduce the global burden of CKD.

Disclosures

R. Kopan reports research funding from the National Institutes of Health; patents and inventions with Washington University; scientific advisor or membership via the Developmental Cell Editorial Board; and other interests/relationships as a member in American Association for the Advancement of Science and the kidney community. O. Volovelsky reports scientific advisor or membership with the Tuberous Sclerosis Alliance Professional Advisory Board and speakers bureau as a speaker in intensive care units for Baxter on AKI. All remaining authors have nothing to disclose.

Funding

This work was supported by United States-Israel Binational Science Foundation grant 2017153 (to O. Volovelsky), Israel Science Foundation grant 2358/18 (to O. Volovelsky), Ministry of Science and Techology of Israel Eshkol award (to Y. Makayes), Hadassah Research Foundation Startup, bridging and physician scientist grants (to E. Resnick and O. Volovelsky) and National Institutes of Health grant DK106225 (to R Kopan).

Data Sharing Statement

The sequencing data have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus database (accession no. GSE159901; reviewer’s token: epwfiseshxejxub).

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2020091321/-/DCSupplemental.

Supplemental Figure 1. Litter size at birth is unaffected by CRD during pregnancy.

Supplemental Figure 2. CRD pups catch up weight by P28.

Supplemental Figure 3. CRD and URD pups have a similar kidney-body ratio at P30.

Supplemental Figure 4. Serum creatinine levels are unaffected by calorie restricted diet in P30 pups.

Supplemental Figure 5. Nephron number is decreased in calorie-restricted Six2 Cre^+/tg pups.

Supplemental Figure 6. Heat map of the top 70 genes in RNA sequencing of NPCs from CRD-treated embryos compared with URD.

Supplemental Figure 7. Apoptosis rate of NPCs is unchanged in CRD-treated embryos.

Supplemental Figure 8. Kidney and body weight at P0 are unaffected by supplementation of methionine or asparagine to CRD during pregnancy.

Supplemental Figure 9. The effect of CRD without or with methionine supplementation on nephron endowment.

Supplemental Figure 10. Maternal serum glucose and β-hydroxybutyric acid during pregnancy in the CRD model.