Abstract
Nicotinamide adenine dinucleotide (NAD+) is a ubiquitous coenzyme involved in electron transport and a co-substrate for sirtuin function. NAD+ deficiency has been demonstrated in the context of acute kidney injury (AKI).
We studied the expression of key NAD+ biosynthesis enzymes in kidney biopsies from human allograft patients and patients with chronic kidney disease (CKD) at different stages. We used ischaemia–reperfusion injury (IRI) and cisplatin injection to model AKI, urinary tract obstruction [unilateral ureteral obstruction (UUO)] and tubulointerstitial fibrosis induced by proteinuria to investigate CKD in mice. We assessed the effect of nicotinamide riboside (NR) supplementation on AKI and CKD in animal models.
RNA sequencing analysis of human kidney allograft biopsies during the reperfusion phase showed that the NAD+de novo synthesis is impaired in the immediate post-transplantation period, whereas the salvage pathway is stimulated. This decrease in de novo NAD+ synthesis was confirmed in two mouse models of IRI where NR supplementation prevented plasma urea and creatinine elevation and tubular injury. In human biopsies from CKD patients, the NAD+de novo synthesis pathway was impaired according to CKD stage, with better preservation of the salvage pathway. Similar alterations in gene expression were observed in mice with UUO or chronic proteinuric glomerular disease. NR supplementation did not prevent CKD progression, in contrast to its efficacy in AKI.
Impairment of NAD+ synthesis is a hallmark of AKI and CKD. NR supplementation is beneficial in ischaemic AKI but not in CKD models.
What is already known about this subject?
Alterations of mitochondrial metabolism are key players in kidney disease.
Nicotinamide adenine dinucleotide (NAD+) is crucial for mitochondrial function. NAD+ deficiency has been described in acute kidney injury (AKI) mouse models, but nothing is known about CKD.
Nicotinamide riboside (NR) is an orally available efficient precursor of NAD+.
What this study adds?
The NAD+ biosynthesis pathway is downregulated in experimental models of AKI and CKD as well as in biopsies from the acute phase of ischaemia in kidney allograft patients and in >200 biopsies from CKD patients of diverse causes.
NAD+ replenishment therapy with a highly potent orally available NAD+ precursor, NR, alleviates kidney injury in two murine models of AKI.
NAD+ replenishment therapy with NR does not efficiently prevent CKD progression in two classic models of CKD.
What impact this may have on practice or policy?
NR could be used to prevent AKI in high-risk patients and before kidney donation/transplantation.
In CKD, more studies are needed to identify which NAD+ replenishment therapy should be used and which subpopulations are likely to benefit most, given the lower response observed experimentally with NR.
INTRODUCTION
Acute kidney injury (AKI) is a major risk factor for CKD progression [1]. Chronic kidney disease (CKD) is a worldwide health problem leading to decreased quality of life, increased mortality and a large financial burden [2]. Mitochondrial dysfunction is part of the pathophysiology of AKI [3], whereas alterations of tubular cell metabolism and mitochondrial dysfunctions may also lead to CKD progression [4]. Nicotinamide adenine dinucleotide (NAD+) is a ubiquitous coenzyme that serves as an electron shuttle in mitochondria and as a substrate of sirtuins and poly-adenosine triphosphate ribose polymerases, thus controlling mitochondrial biogenesis and metabolism. NAD+ synthesis can be achieved from tryptophan through de novo synthesis or from precursors such as nicotinamide, nicotinic acid or nicotinamide riboside (NR) [5].
Deficient NAD+ synthesis caused by alteration of the de novo pathway plays an important role in tubular lesions of AKI, with a key role of decreased activity of the quinolinate phosphoribosyltransferase (QPRT) enzyme in humans and mice [6]. In ischaemia–reperfusion injury (IRI), supplementation with NAD+ precursor nicotinamide attenuated renal injury in mice and prevented human AKI in a pilot study of patients undergoing cardiac surgery [6]. Stimulation of the de novo synthesis of NAD+ can be achieved through inhibition of α-amino-β-carboxymuconate-ε-semialdehyde decarboxylase (ACMSD), a critical enzyme strongly expressed in the kidneys and liver [7]. ACMSD inhibition improved mitochondrial function, increased NAD+ and lifespan in worms and prevented AKI in mice [7]. The NAD+ pathway is thus an important therapeutic target that is being explored currently for AKI prevention. Mitochondrial dysfunction is also present in CKD and recently it has been shown that preventing mitochondrial damage could reduce fibrogenesis [8]. Whether alterations of NAD+ synthesis also participate in non-ischaemic CKD progression and fibrogenesis through metabolic alterations is unknown.
Several precursor molecules for NAD+ are available. NR has a very good oral bioavailability, enhances sirtuin activity and is described as the most potent NAD+-replenishment therapy [5, 9]. NR has no recorded side effects in humans [10]. In addition, oral administration appears as a more relevant and realistic option in chronic diseases. Deficiency in NAD+ could thus be rescued by the administration of NR, which could be an interesting therapy in both AKI and CKD.
In this study, we first demonstrate that NAD+ synthesis is altered not only in acute kidney disease but also CKD in both humans and mice, leading to lower NAD+ content in kidney tissue. We show in mouse models that NR is able to prevent AKI but does not attenuate CKD progression.
MATERIALS AND METHODS
Microarray data analyses of human kidney biopsies
We used published Affymetrix microarray expression data from the European Renal cDNA Bank–Kröner-Fresenius Biopsy Bank (ERCB-KFB; CKD: GSE99340; living donors: GSE32591, GSE35489 and GSE37463). Sample collection, RNA isolation and preparation and microarray analysis were performed as described previously [11]. For the differential gene expression analysis in the glomerular and tubulointerstitial compartments, patients were grouped into different CKD Stages 1–5 according to their estimated glomerular filtration rate (GFR) calculated by the Chronic Kidney Disease Epidemiology Collaboration equation [12] (CKD1, n = 56; CKD2, n = 46; CKD3, n = 37; CKD4, n = 26; CKD5, n = 10). All available biopsies irrespective of the primary disease were included in the analysis. Pre-transplantation kidney biopsies from healthy living donors (living donors, n = 42) served as the control group. To identify differentially expressed genes, the significance analysis of microarrays (SAM) method was applied using TiGR software (MeV, version 4.8.1) [13]. A q-value <5% was considered to be statistically significant.
Renal allograft RNA sequencing
Genome-wide gene expression profiling using RNA sequencing was performed in kidney allograft recipients as previously described [14]. Briefly, 43 patients were enrolled at the University Hospitals of Leuven. For each of them, we compared a protocol biopsy performed after reperfusion (at the end of the surgical procedure) and 12 months after transplantation. Living kidney donors and donors after cardiac death were excluded, leaving 25 patients for analysis. Raw gene counts were normalized by the trimmed mean of M-values method (TMM) using the edgeR package and genes with null expression in one sample were excluded. Counts were expressed in counts per million (CPM). Recovery biopsies were identified as previously described using a transcription analysis [14]. Biopsies performed at the time of reperfusion and from donors after cardiac death (n = 13) were excluded from the analysis. Comparisons of gene expression were performed using a gene-wise negative binomial linear model with the quasi-likelihood test. Values are expressed as TMM-normalized CPM. Raw data are available at Gene Expression Omnibus (accession GSE126805).
Animal experiments
All the animals had free access to food and water and were acclimatized for at least 1 week before any procedure. For all experiments, littermate males were used at 8–10 weeks of age.
For unilateral ureteral obstruction (UUO), 8-week-old C57BL/6J male mice underwent isoflurane anaesthesia followed by unilateral ligature of the ureter as described previously [15]. Mice were euthanized 7 or 10 days after the procedure. The contralateral non-ligated kidney was used as a control. For NR experiments, mice were sacrificed 7 days after UUO. Since NR experiments were repeated twice, data were corrected to the mean of the controls of each repetition and presented as the fold change.
For the proteinuric model of progressive tubulointerstitial fibrosis, POD-ATTAC mice were generated as described [16]. Eight- to 10-week-old male FVB mice were used for the experiment and randomized into four experimental conditions. One week after the start of the diet regimen, dimerizer (AP20187; Clontech Laboratories, Mountain View, CA, USA) was prepared following the product data sheet. POD-ATTAC mice on FVB background were injected once with 0.5 µg/g body weight of dimerizer for groups sacrificed after 7 days and injected five times with 0.2 µg/g of dimerizer in the groups sacrificed after 14 or 28 days. The dimerizer induces dose-dependent podocyte apoptosis, leading to glomerular proteinuria and secondary tubulointerstitial lesions. Control mice were either FVB mice injected with dimerizer or POD-ATTAC mice injected with vehicle. For NR experiments, mice were injected five times with 0.2 µg/g of dimerizer and sacrificed after 14 days. Since NR experiments were repeated three times, data were corrected to the mean of the controls of each repetition and presented as the fold change.
For IRI, 9-week-old C57BL/6J males were randomized and separated into three groups. On Day 10 after the start of the diet, regimen mice were subjected to isoflurane anaesthesia, both their kidneys were exposed via flank incisions and renal pedicles were occluded with vascular clamps. After 25 min, the clamps were released and the kidneys were allowed to reperfuse. The surgical site was then sutured. To prevent dehydration, 1 mL of physiological saline was administered intraperitoneally (i.p.) after closing the wound. The sham controls were injected with the same volume of 0.9% saline. After the surgery and for the following 48 h, until the experiment termination, mice were single-housed with appropriate environmental enrichment. Animals were sacrificed 48 h after the surgery. One experiment was performed.
For cisplatin-induced AKI, 8-week-old C57BL/6J males were randomized and separated into three groups. On Day 10 after the start of the diet, regimen cisplatin was injected i.p. at a dose of 20 mg/kg of body weight. The sham controls were subjected to similar surgical procedures, except that the occluding clamp was administered with the same volume of 0.9% saline. Animals were sacrificed 72 h after cisplatin injection. One experiment was performed.
One week (for POD-ATTAC and UUO) or 10 days (IRI) before the start of renal disease induction, mice were randomly assigned to either normal chow containing vehicle (water) or NR (NovAliX). NR was added to freely accessible normal chow at either 400 mg/kg/day (IRI) or 800 mg/kg/day (POD-ATTAC and UUO). In a 24-h repeated observation period, the food intake of the groups was similar. The treatment was maintained until sacrifice with a 4-h fasting period before anaesthesia.
For sacrifices, mice were anaesthetized and blood was sampled via cardiac puncture followed by organ collection. Plasma was collected by centrifugation at 4000 rpm for 10 min at 4°C, separated into a fresh tube and stored at −80°C. Collected tissues were snap-frozen in liquid nitrogen and stored at −80°C.
All animal experiments were approved by the Institutional Ethical Committee of Animal Care in Geneva and Lausanne and cantonal authorities. All animal experiments were conducted in accordance with Swiss legislation on animal welfare under the animal authorization numbers GE-18218 and VD-3313.
Transcutaneous GFR measurement in CKD
GFR was measured in a subset of animals and assessed transcutaneously by fluorescein isothiocyanate sinistrin excretion rate measurement with a minicamera as described previously [11]. Results calculated as millilitre per minute per kilogram of body weight were normalized to controls and expressed as a fold change. As the contralateral kidney compensates for the ligated kidney, we did not perform kidney function assessment in the UUO model.
Kidney NAD+ content measurement
A 3-mm3 piece of kidney cortex was weighted and flash-frozen. NAD+ was extracted using the acidic extraction method and analysed by high-performance liquid chromatography (HPLC)–mass spectrometry as previously described [7]. In brief, ∼10 mg of frozen tissue samples were used for NAD+ extraction in 10% perchloric acid and neutralized in 3 M potassium carbonateon ice. After final centrifugation, the supernatant was filtered and the internal standard (NAD+-C13) was added and loaded onto a column (150 Å, ∼2.1 mm; Kinetex EVO C18, 100 Å). HPLC was run for 1 min at a flow rate of 300 mL/min with 100% buffer A [methanol/water (H2O), 80/20% v/v]. Then a linear gradient to 100% buffer B (H2O + 5 mM ammonium acetate) was performed (at 1–6 min). Buffer B (100%) was maintained for 3 min (at 6–9 min) and then a linear gradient back to 100% buffer A (at 9–13 min) started. Buffer A was then maintained at 100% until the end (at 13–18 min). NAD+ eluted as a sharp peak at 3.3 min and was quantified on the basis of the peak area ratio between NAD+ and the internal standard and normalized to tissue weight.
Plasma urea and creatinine measurements
Plasma urea and creatinine were measured at the Zurich Integrative Rodent Physiology facility at the University of Zurich using the UniCel DxC 800 analyser (Beckman Coulter, Brea, CA, USA).
Histological analyses
For AKI experiments, half of a kidney was fixed with 10% neutral buffered formaldehyde, processed and embedded in paraffin. Tissue sections were stained with haematoxylin and eosin. Two different pathologists (4 and 7 years of experience) performed histopathological scoring in a blinded and independent way. The scoring system was adapted with slight modifications [17]. The following parameters were scored: tubular cell necrosis, infiltration of inflammatory cells, tubular dilation and cast formation.
For the POD-ATTAC and UUO mouse models, harvested kidneys were embedded in paraffin and cut into longitudinal sections. One representative median slide of the kidney stained with Sirius red was scanned with an Axioscan image scanner (Zeiss, Oberkochen, Germany) at 20× magnification. Using Tissue Phenomics software (Definiens, Munich, Germany), we manually defined the whole cortical area to be analysed. The software detected the red marker based on the intensity of the signal, excluding the zones displaying a width of >5 µm3 for proteinuric mice to avoid intratubular signal. The percentage of red-stained area over the total area was quantified by the software.
Real-time quantitative polymerase chain reaction (PCR)
RNA extraction and reverse transcription were done as described previously [11]. The real-time PCR was performed using a StepOne Plus Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) or a QuantStudio 5 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA). ΔΔCt values between the gene of interest and a reporter gene (RPLP0) were normalized to controls and expressed as a fold change or log2 fold change. The sequences of the primers are provided in Supplementary data, Table S1.
Western blotting
Protein extraction and dosage were performed as previously described [11]. Proteins of 20–30 μg were loaded. We performed a sodium dodecyl sulphate–polyacrylamide gel electrophoresis and transferred the proteins to nitrocellulose (Amersham Protran Premium 0.45 NC) membranes. After blocking with 5% milk–tris-buffered saline and polysorbate 20, membranes were incubated at 4°C overnight with the primary antibodies. The following primary antibodies were used: rabbit polyclonal 3-hydroxyanthranilate 3,4-dioxygenase (HAAO; AB 106436, 1:1000) and rabbit polyclonal glyceraldehyde 3-phosphate dehydrogenase (GAPDH; AB 9485, 1:3000). Goat anti-rabbit HRP (1:5000, BD 554021) secondary antibodies were used. The HAAO band density was quantified with ImageJ software (National Institutes of Health, Bethesda, MD, USA) and normalized to GAPDH band density. Results were presented as the fold change in protein expression compared with the control or sham samples.
Statistical analyses
For microarray data analyses, we used the SAM method with TiGR (MeV, version 4.8.1) software. Statistical significance was reached with a q-value <5%.
For other analyses, we performed one-way analysis of variance with Sidak's multiple comparisons test for more than two groups and t-test for two groups using GraphPad Prism 7 software (GraphPad Software, San Diego, CA, USA). Data are shown as mean ± standard deviation (SD). P-values <0.05 were considered significant. All analyses were two-tailed.
RESULTS
NAD+ synthesis is dysregulated in human and experimental IRI
Expression of enzymes involved in the de novo NAD+ synthesis pathway was downregulated during acute ischaemia in kidney allograft patients when comparing the reperfusion time to a stable phase after transplantation. The expression of enzymes of the salvage pathway was preserved and some enzymes were even enhanced (Figure 1A). A similar pattern of NAD+ biosynthetic enzymes expression was observed in mice subjected to IRI, with fewer conserved NAD+ salvage pathway enzymes (Figure 1B). At the protein level, the expression of HAAO, a key enzyme of the de novo pathway, was also decreased (Supplementary data, Figure S1 A).
Regulation of NAD+ synthesis in acute ischaemic kidney injury. (A) Expression of genes implicated in the de novo (IDO1, KMO, KYNU, HAAO and QPRT) and salvage NAD+ biosynthesis pathways (NAMPT, NMNAT2, NADSYN1 and NRK1) in human kidney graft biopsies during reperfusion (n = 25) compared with the post-recuperation time point (n = 23). Results represent individual values of patients during the reperfusion phase compared with the mean of controls (log2 fold change). (B) Expression of genes implicated in the de novo (Ido1, Afmid, Kmo, Kynu, Haao, Acmsd and Qprt) and salvage NAD+ biosynthesis pathways (Nampt, Nmnat1, Nmnat3 and Nmrk1) in mice subjected to 24 h (n = 10) of acute reperfusion injury (IRI). Controls are sham-operated mice (n = 5). *P < 0.05. Results are presented as the log2 fold change.
NR efficiently prevents experimental IRI
To further test the significance of altered NAD+ biosynthesis, mice were administered NR for 10 days before the induction of IRI. Mice treated with NR displayed attenuated plasma urea and creatinine elevation 48 h after the reperfusion (Figure 2A), with alleviated tubular injury (Figure 2C), as reflected by lower histopathological scores (Supplementary data, Figure S2A). Similar results were obtained in a toxic AKI model, the cisplatin mouse model (Figure 2B and D and Supplementary data, Figure S2B).
Effect of NAD+ precursor supplementation in AKI. (A) Blood urea nitrogen (BUN) and creatinine levels (mg/dL) and histological damage assessed by the cumulative Jablonski score in sham-operated (sham, n = 5) versus ischaemia–reperfusion-subjected mice (IR-AKI, n = 5) and ischaemia–reperfusion-subjected mice treated with 400 mg/kg/day of NR (IR-AKI + NR, n = 5). (B) BUN and creatinine levels (mg/dL) and histological damage assessed by the cumulative Jablonski score in controls (CTL, n = 6) versus cisplatin-induced AKI (cis-AKI, n = 6) and cisplatin-induced AKI treated with 400 mg/kg/day of NR (cis-AKI + NR, n = 6). Representative images of kidney sections in (C) sham, IR-AKI and IR-AKI + NR mice and in (D) CTL, cis-AKI and cis-AKI + NR mice. Black arrows show tubular cell necrosis, asterisks indicate tubular dilation and casts and white arrowheads show interstitial inflammatory cells. *P < 0.05, **P < 0.01, ***P < 0.0005, ****P < 0.0001. Results are presented as mean ± SD.
NAD+ synthesis is dysregulated in human and experimental CKD
To determine whether NAD+ synthesis was altered in human CKD, we used expression data generated by the ERCB-KFB. The tubulointerstitial expression of several enzymes involved in de novo NAD+ synthesis, including the rate-limiting QPRT enzyme, was downregulated in a CKD stage-dependent manner compared with biopsies obtained from healthy donors. The salvage pathway appeared preserved, implying that only the de novo NAD+ synthesis is altered in CKD according to disease severity (Figure 3A). In the glomerular compartment, the same pattern of downregulation was observed (Supplementary data, Figure S3A). We then assessed whether similar alterations of the NAD+ biosynthetic pathway were observed in CKD in two different mouse models, the proteinuric POD-ATTAC model (model of progressive secondary tubulointerstitial lesions induced by high-range glomerular proteinuria) and the classic UUO model. In accordance with human kidney disease, we observed a well-defined downregulation of enzymes involved in the salvage and de novo NAD+ pathways in mouse CKD models (Figure 3B and C). The protein expression of HAAO followed the same pattern in the POD-ATTAC and UUO mouse models (Supplementary data, Figure S1B and C). These changes in enzyme expression resulted in lower cortical NAD+ levels (Figure 3D and E) thatpersisted after correction for the level of fibrosis (Supplementary data, Figure S3B and C). As positive controls for gene expression, we show that the expression of C-myc and other metabolic regulators was increased or unchanged and inflammatory markers were upregulated (Supplementary data, Figure S3D and E).
Regulation of NAD+ synthesis in CKD. (A) Analysis of genes implicated in the de novo (IDO1, TDO2, KMO, KYNU, HAAO and QPRT) and salvage NAD+ biosynthesis pathways (NAMPT, NMNAT2 and NADSYN1) in the Affymetrix microarray expression data set obtained in the ERCB-KFB, sorted by CKD stage (CKD1, n = 56; CKD2, n = 46; CKD3, n = 37; CKD4, n = 26; CKD5, n = 10). Biopsies from kidney donors (n = 42) are used as controls. *P <0.05–<0.0001. (B) Transcript levels of genes implicated in the de novo (Ido2, Tdo2, Afmid, Kmo, Kynu, Haao, Acsmd and Qprt) and salvage NAD+ biosynthesis pathways (Nampt, Nmnat1, Nmnat3 and Nmrk1) in a proteinuric model of progressive CKD (POD-ATTAC). Controls, n = 6; mild CKD (7-day post-CKD induction), n = 6; moderate CKD (14-day post-CKD induction), n = 7; severe CKD (28-day post-CKD induction), n = 6. *P <0.05–<0.0001. (C) Transcript levels of genes implicated in the de novo (Ido2, Tdo2, Afmid, Kmo, Kynu, Haao, Acsmd and Qprt) and salvage NAD+ biosynthesis pathways (Nampt, Nmnat1, Nmnat3 and Nmrk1) in kidney cortex after 7 or 10 days of UUO (7 days: controls, n = 7; obstructed, n = 7; UUO 10 days: controls, n = 6; obstructed, n = 6). *P <0.05–<0.0001. (D) NAD+ content in the control (CTL, n = 7) and experimental kidney cortex (DIM, n = 8) of the POD-ATTAC mouse model. (E) NAD+ content in the control (CTL, n = 9) and obstructed kidney (UUO, n = 9) of the UUO mouse model of kidney fibrosis. Results are presented as mean ± SD. *P < 0.05, **P < 0.01, *** P < 0.0005. Results are presented as mean ± SD.
NR does not prevent CKD progression experimentally
Mice were pretreated with NR for 7 days and then subjected to UUO or chronic proteinuria (POD-ATTAC). NR supplementation did not improve measured GFR or plasma urea or creatinine levels (Figure 4A), although it was efficient in increasing kidney NAD+ content in POD-ATTAC mice (Figure 4B). NR also failed to improve interstitial fibrosis as assessed by Sirius red staining in both proteinuric and UUO mice (Figure 4C and D). Finally, neither metabolic nor pro-inflammatory or pro-fibrotic pathway gene expression was improved by NR administration (Supplementary data, Figure S4).
Effect of NAD+ replenishment therapy on kidney function in two CKD mouse models. (A) Renal function as assessed by measured GFR (mL/min/kg), blood urea nitrogen (BUN) and creatinine (mg/dL) in POD-ATTAC mice pretreated 7 days before CKD induction followed by 14 days with 800 mg/kg/day of NR in the diet. The represented groups are control mice (CTL, n = 7), control mice treated with NR (CTL + NR, n = 8), POD-ATTAC mice injected with dimerizer (DIM, n = 8) and POD-ATTAC injected with DIM and treated with NR (DIM + NR, n = 12). (B) NAD+ content (μmol/mg of kidney cortex) in POD-ATTAC mice (CTL, n = 9; CTL + NR, n = 8), supplemented as described with NR. Representative images of Sirius red stainings and corresponding quantification of (C) POD-ATTAC and (D) UUO kidney sections in mice with and without NR supplementation. For POD-ATTAC: CTL, n = 7; CTL + NR, n = 8; DIM, n = 8; DIM + NR, n = 12. For UUO: CTL, n = 9; CTL + NR, n = 8; UUO, n = 9; UUO + NR, n = 8. *P < 0.05, **P < 0.01. Results are presented as mean ± SD.
DISCUSSION
Recent studies have convincingly shown that alterations of the expression and function of enzymes involved in the de novo NAD+ synthesis pathway result in disturbed renal NAD+ homoeostasis in toxic and ischaemic animal models of AKI [6, 7]. We first extended this observation to human IRI in allograft kidneys sampled during the reperfusion phase of the transplantation compared with control biopsies obtained in a stable phase at some time from transplantation. We further showed that NR, an orally bioavailable and well-tolerated NAD+ precursor, was nephroprotective in both toxic and ischaemic models of AKI. Given our observations in kidney allograft recipients, administration of NR before donation may improve renal recovery in this specific population.
The expression of NAD+ synthesis enzymes was clearly altered in human and mouse models of CKD, with lower tissue NAD+ levels in mice that persisted after correction for the degree of fibrosis. In humans, deficient NAD+ synthesis was observed in the tubular and glomerular compartments of the kidney. However, systemic NR supplementation was unable to improve disease progression in the two studied CKD mouse models, in contrast to the beneficial effects of NR observed in AKI. Thus the observed alterations of the NAD+ pathway could be markers of advanced tubular lesions, irreversible mitochondrial lesions or cellular dedifferentiation in chronic conditions, but factors in the pathogenesis of the acute but reversible mitochondrial injury of AKI. In contrast, a recent study using high doses of i.p. nictotinamide at later time points of mouse UUO observed a dose-dependent protective phenotype [18]. The NAD+ content and the renal function were not assessed in this study and only histological lesions were shown. This difference is yet unexplained, but the precursor used, the administration route and the dosage were different. In our study, we ensured that the food intake was sufficient in our animals and we used a higher dose than in the acute model, where it was efficient, and as a result, the therapy increased NAD+ levels in the kidney. Another possibility could be that restoring the NAD+ pathway, for example, through de novo pathway stimulation, may also be more potent and successful in the context of CKD. Altogether, these results indicate that CKD is, at a minimum, less responsive than AKI to NAD+ restoration, at least using NR as a NAD+ precursor.
Our observation is limited to two experimental models of CKD, but these models are complementary and representative since they arise from different physiopathologies and display similar dysregulation of the NAD+ biosynthetic pathways [16, 19]. POD-ATTAC is a progressive model of secondary tubulointerstitial lesions induced by glomerular proteinuria, mimicking chronic human glomerular diseases, whereas UUO is a model of primary tubular lesions leading to tubular atrophy and interstitial fibrosis. Our observation does not preclude a beneficial effect of NAD+ replenishment therapies on CKD progression through the avoidance of acute episodes of renal injury that often participate in the progression.
Overall, we confirm the alterations of the de novo NAD+ synthesis pathway in human and experimental acute kidney injuries. These alterations can be rescued by stimulation of the salvage NAD+ synthesis pathway via NR administration in the context of two different models of experimental AKI. Moreover, to our knowledge, we are the first to report alterations of NAD+ synthesis in all kidney compartments in human CKD patients, with a severity proportional to the CKD stage. This observation was confirmed in two mouse models of CKD accompanied by decreased renal NAD+ content. However, NR administration failed to prevent CKD progression in these models while successfully attenuating AKI.
SUPPLEMENTARY DATA
Supplementary data are available at ndt online.
ACKNOWLEDGEMENTS
We thank all participating centres of the ERCB-KFB and their patients for their cooperation. Active members at the time of the study, see [N. Shved et al., Scientific reports 7, 8576 (2017)].
We are deeply indebted Phillip E. Sherer (Texas A&M Health Science Center School of Medicine Bryan/College Station, TX) who kindly provided us with the POD-ATTAC mouse model.
We thank Nicolas Liaudet, from the Bioimaging Core Facility of the University of Geneva, Nadine Stokar-Regenscheit and Ruthra Perumal from the Institute of Animal Pathology in EPFL.
FUNDING
The ERCB-KFB was supported by the Else Kröner-Fresenius Foundation. S.d.S. is supported by grants from the Swiss National Science Foundation (PP00P3-187186/1), Jules Thorn Foundation and Swiss National Centre of Competence in Research Kidney Control of Homeostasis. A.F. is the recipient of a grant from the Swiss National Science Foundation (323530_191224). J.A. was funded by grants from the Ecole Polytechnique Fédérale de Lausanne and the Swiss National Science Foundation (310030 A-179435).
AUTHORS’ CONTRIBUTIONS
A.F. performed experiments, analysed the data, corrected the manuscript and drew the figures. S.d.S. designed and oversaw the project, analysed the data and drafted the manuscript. T.V. performed experiments. R.D.R. provided intellectual input and analysed the data. C.H. performed experiments. M.L. and C.C. performed the microarray data analyses of human kidney biopsies. M.N. provided the human allograft RNA sequence data. E.K. and A.M. performed and analysed the IR-AKI and cisplatin-AKI experiments. E.F. provided intellectual input and interpreted the data. A.L. and F.A. performed experiments and provided data. P.C. contributed with mouse RNA sequencing data in IRI. J.M.R. provided the POD-ATTAC mouse model. D.L. performed biostatistical analyses. J.A. designed and oversaw the IR-AKI project. All authors revised and approved the manuscript.
CONFLICT OF INTEREST STATEMENT
J.A. has a patent on NAD-related technologies pending and is a consultant to Mitobridge-Astellas, MetroBiotech and TES pharma, companies that develop NAD boosting therapies. No other authors reported any conflicts of interest.
Comments