Inhibiting NLRP3 signaling in aging podocytes improves their life- and health-span

Middle-aged podocytes acquire an inflammatory phenotype

We have previously reported that aged 22 to 24-month-old (~70 human years) podocytes acquire an inflammatory phenotype including upregulation of NLRP3 signaling as measured by mRNA sequencing of isolated podocytes [20]. Based on these and published data on the importance of NLRP3 signaling, we further investigated the NLRP3 inflammasome in podocyte aging. First, the mRNA-seq data was confirmed by qRT-PCR in podocytes isolated from a different cohort of young and aged mice showing increased transcript levels for Nlrp3 and key pathway mediators, Casp1 and Pycard (Figure 1A). Next, to determine if the NLRP3 inflammasome was increased in middle-aged mice, we compared podocyte transcriptomes of young 4-month-old mice to middle-aged mice 19.5 months-old (~60 human years) (Supplementary Figure 1A). Gene set enrichment analysis (GSEA) again identified a signature consistent with an inflammatory phenotype (Supplementary Figure 1B). This includes pathways of the inflammatory response (Figure 1B), IL2/Stat5 signaling, IL6 signaling, interferon gamma response, allograft rejection, complement, and TNF signaling (Supplementary Figure 1B).

Figure 1. Increase in the NLRP3 inflammasome in middle-aged podocytes. (A) qRT-PCR of inflammasome components Nlrp3, Casp1 and Pycard comparing podocytes from 4 months (m)-old to 24 m-old mice. Graph shows increase in the NLRP3 signaling. Error bars correspond to standard deviation; data are compared using Student’s test and p-values are indicated above the bars. (B, C) GSEA plots comparing podocytes isolated from middle-aged to young mice show enrichment of inflammatory response (B) and the NLRP3 inflammasome (C) gene sets. (D–J) Nlrp3 immunoperoxidase staining (brown) comparing glomeruli of differently aged mice with no staining at 4 m (D) and 12 m (E), but increasing staining at 18 m (F), 24 m (G) and 27 m (H). Quantification of the staining intensity is shown in (J). NLRP3 staining was absent in NLRP3 null (−/−) mice aged 19.5 m (I). (K) NLRP3 immunofluorescent staining (red) was absent in podocytes (labeled in green) of young, 4 m-old Nphs1-EGFP reporter mice, but colocalized to podocytes of 24 m-old aged mice (labeled in yellow/orange). (L–O) NLRP3 immunostaining of human kidneys shows no staining in young (23 years old) glomeruli (L, M), but is detected in aged (82 years old) glomeruli (N, O). Panels (M) and (O) show enlarged views indicated by the red boxes in (L) and (N), respectively. (P–R) NLRP3 transcripts from micro-dissected human glomeruli. Higher Expression of NLRP3 is associated with increased percent of globally sclerosed glomeruli (Figure 1P), higher glomerular volume (Figure 1Q) and reduced podocyte density (Figure 1R).

Immunostaining for NLRP3 in glomeruli across the mouse lifespan showed no signal at 4 and 12 months of age, but a progressive increase from 18 months of age (referred to herein as middle-aged) through 24 and 27- months-old (referred to as aged) (Figure 1D–1H). As expected, staining was absent in middle-aged NLRP3 null mice (Figure 1I). Glomerular immunostaining from 18 months of age was predominantly in a podocyte distribution, with the increase confirmed by quantification using a computer-assisted machine learning (Figure 1J). Weak NLRP3 staining was also detected in parietal epithelial cells and tubular epithelial cells from middle-aged mice that was markedly increased in aged animals. Aged (24 months) Nphs1-EGFP podocyte reporter mice confirmed the increase in NLRP3 in a different mouse strain, which co-localized with the EGFP reporter in podocytes (Figure 1K).

Next, to address the translational potential, we examined NLRP3 immunostaining in podocytes of aged human kidneys. NLRP3 was not detected in glomeruli from young (23 years old) healthy human kidneys (Figure 1L, 1M). In contrast, NLRP3 staining was increased in glomeruli of aged (82 years old) human kidneys (Figure 1N, 1O). Higher transcript expression levels for NLRP3 in micro-dissected human glomeruli were associated with increased percent of globally sclerosed glomeruli (Figure 1P), higher glomerular volume (Figure 1Q) and reduced podocyte density (Figure 1R), assessed as previously described [19, 38].

Finally, we examined key NLRP3 pathway mediators. In line with the mRNA-seq and the qRT-PCR analyses (Figure 1A–1C), immunostaining for NLRP3 (Figure 2A, 2B, 2E), Caspase-1 (Figure 2F, 2G, 2J), IL-1β (Figure 2K, 2L, 2O) and IL-18 (Figure 2P, 2Q, 2T) were higher in glomeruli of middle-aged mice compared to young mice, predominantly in a podocyte distribution. Similarly, Caspase-1 staining was barely detected in young healthy human kidneys but increased in podocytes with age (Supplementary Figure 2A–2D).

Figure 2. Inhibiting/Deleting NLRP3 in middle-aged mouse podocytes. (A–Y) Comparison of glomeruli from young and middle-aged mice treated with vehicle (Saline) or MCC950, as well as middle-aged Nlrp3-null mice by immunohistochemistry for NLRP3 (A–E), inflammasome downstream signaling components Caspase-1 (F–J), Interleukin-1β (IL-1β) (K–O), IL-18 (P–T) as well as the senescence marker p16 (U–Y). All immunostainings (brown color) were quantified using automated approaches, depicted as integral optical density/area (IOD/area) and analyzed by Student’s t-test statistics. (Z–DD) Staining for SA-β-Gal (blue) and nuclear red (red). SA-β-Gal is not detected in young mice (Z, DD), but increased in the glomerulus in saline treated middle aged mice, including podocytes (AA, DD). SA-β-Gal was lowered by MCC950 (BB, DD) and in age-matched NLRP3 null mice (CC, DD). Note that in all the conditions staining increased in glomeruli of middle-aged saline-treated and was lower in MCC950-treated (C, E) and in aged-matched NLRP3 null mice. The scale bars in the images correspond to 25 μm.

Together, these data are consistent with an increase in the NLRP3 inflammasome and downstream mediators in mouse and human podocytes in middle-aged and aged kidneys.

NLRP3 expression is augmented in aged mice with experimental FSGS

Studies have shown an increase in NLRP3 in injured podocytes in disease. [23–37] This was confirmed in our experimental model of podocyte depletion induced FSGS, which showed an increase in NLRP3 staining (Supplementary Figure 3A, 3B). To determine if NLRP3 was further increased in aged mice with disease, experimental FSGS was induced in mice 24 months of age with a cytopathic anti-podocyte antibody. NLRP3 staining in the glomerulus was higher in aged mice with FSGS compared to age-matched mice without disease (Supplementary Figure 3C, 3D). These data support the notion that aging, and injury compound the increase in NLRP3 in podocytes.

Pharmacological and genetic approaches to reduce NLRP3 in podocytes from middle-aged mice

To address the biological role of an increased NLRP3 inflammasome in middle-aged podocytes we used both pharmacological and genetic approaches. 18-month-old C57B6 mice were randomized to receive 6 weeks of either the NLRP3 inhibitor MCC950 [39], or vehicle (saline) (Supplementary Figure 4A). MCC950 treatment had no impact on body weight (Supplementary Figure 4B). As previously reported [19] clinical measurements of kidney function, Albumin: Creatinine Ratio (ACR), Blood Urea Nitrogen (BUN) and Soluble Urokinase Plasminogen Activator Receptor (suPAR), were not increased in control middle-aged mice and were therefore not impacted by reducing NLRP3 pharmacologically (Supplementary Figure 4C–4E).

Immunostaining demonstrated that MCC950 treatment was effective when compared to vehicle-treated littermates by reducing podocyte staining for NLRP3 (Figure 2B, 2C, 2E) and its downstream targets Caspase-1 (Figure 2G, 2H, 2J), IL-1β (Figure 2L, 2M, 2O) and IL-18 (Figure 2Q, 2R, 2T). These changes were similar to those seen in middle-aged NLRP3 null mice, where staining for NLRP3 (Figure 2D, 2E), Caspase-1 (Figure 2I, 2J), IL-1β staining (Figure 2N, 2O) and IL-18 (Figure 2S, 2T) were not detected in podocytes. These results show that MCC950 treatment and the genetic deletion of NLRP3 reduce the NLRP3 inflammasome and downstream mediators in middle-aged mice.

Podocyte lifespan and glomerular ultrastructure were improved when NLRP3 was inhibited or absent

We next addressed whether NLRP3 inhibition in wildtype mice and/or genetic deletion of NLRP3 halted or even reversed podocyte aging in middle-age. As expected, compared to young mice, staining for the senescent activator gene p16^Ink4 was higher in both glomerular epithelial cells and tubular epithelial cells of control middle-aged mice (Figure 2U, 2V, 2Y). Glomerular staining for p16 was predominantly localized to podocytes. Both MCC950 treatment and gene deletion in NLRP3 null mice resulted in reduced p16 staining compared to aged-matched saline and wildtype controls, respectively (Figure 2W, 2X). Analysis of senescence-associated beta-galactosidase (SA-β-gal) as a marker of aging showed similar results (Figure 2Z–2DD).

A major hallmark of glomerular aging is the decrease in the podocyte’s lifespan, which is characterized by a decrease in podocyte density [17, 19, 40–42]. Podocyte number was measured by staining for p57 (Cdkn1c) and quantitated with computer-assisted machine learning. Podocyte density decreased in saline-treated middle-aged vs. young mice (Figure 3A, 3B, 3E). Podocyte density was restored in glomeruli of middle-aged mice treated with MCC950 and middle-aged NLRP3 null mice (Figure 3C–3E). The improvement in podocyte lifespan in the MCC950-treated and the NLRP3 null mice was accompanied by reduced glomerular scarring as measured by Collagen IV immunostaining (Figure 3A–3D, 3F). Measuring nuclear size as a surrogate for podocyte hypertrophy [43–46] demonstrated a reduction in MCC950-treated podocytes (Figure 3G). The reversal of the aged podocyte phenotype was not caused by podocytes re-entering the cell cycle as the proliferation marker Ki-67 was not detected in podocytes of any mouse group (not shown).

Figure 3. Podocyte lifespan and glomerular ultrastructure were improved by inhibiting and deleting NLRP3. (A–F) Immunostaining for p57 (black, nuclear) and Collagen IV (brown) comparing glomeruli of young, middle-aged saline-treated, MCC950-treated mice as well as NLRP3 null mice. Quantification of the number of p57-positive podocytes was used to determine podocyte density (E), Collagen IV staining to assess glomerular scaring (F) and their nuclear size was used to determine hypertrophy (G). Samples were compared using Student’s t-test and significance is indicated. Representative images are shown; the scale bars in the images correspond to 25 μm. (H–O) Analysis of glomerular ultrastructure using FLARE coupled to confocal microscopy; hydrogel-expanded mouse kidney tissue has been stained for primary amines (red, labels proteins), oxidized carbohydrates (green, labels basement membrane and mesangial matrix), and DNA (blue, labels nuclei). Representative images comparing glomeruli of young, middle-aged saline-treated, MCC950-treated mice as well as NLRP3 null mice are shown. All scale bars correspond to 5 μm pre-expansion. Abbreviations: P: podocyte; E: endothelial cells; M: mesangial cells. (P, Q) Quantification of glomerular basement membrane (GBM) thickness by measuring the average thickness of the oxidized carbohydrate stain in capillary loops (P) and of foot process width by determining the average thickness of the amine-stained foot processes at half-maximum (Q).

Finally, we assessed glomerular ultrastructure by FLARE and expansion microscopy (Figure 3H–3O). Quantification of these images demonstrated that the glomerular basement membrane (GBM) thickness increased (283 nm vs. 416 nm, P < 0.0001, Figure 3P) and foot process width increased (107 nm vs. 134.7 nm, P < 0.0001, Figure 3Q) in middle-aged saline-treated control mice compared to young mice. While GBM thickness did not change during the 6-week-long MCC950 treatment (405 nm vs. 416 nm, P = 0.37), it was reduced in the middle-aged NLRP3 null mice (373 nm vs. 416 nm, P = 0.002, Figure 3P). This is in line with the idea that GBM remodeling is not instantaneous [47–49] and likely takes longer than the 6 weeks of MCC950 treatment. However, foot process width was reduced in both MCC950-treated mice (123.6 nm vs. 134.7 nm, P = 0.012) and in NLRP3 null mice (107.1 nm vs. 134.7 nm, P < 0.001, Figure 3Q). In summary, the data demonstrate that decreasing or deleting NLRP3 improved podocyte lifespan in glomeruli of middle-aged mice.

Podocyte health was improved when NLRP3 was inhibited or absent

The podocyte’s health-span decreases with advancing age and is characterized by changes to their molecular, cellular and transcriptional landscape required for their normal physiology, structure and function [15]. Thus, to understand whether the NLRP3 inflammasome contributes to this aspect of podocyte aging and how the NLRP3 pathway impacts podocyte health-span, we initially focused on canonical podocyte genes essential for the highly specialized structure and function of these terminally differentiated cells. Compared to middle-aged/vehicle-treated mice, middle-aged/MCC950-treated mice had higher transcript levels for podocyte health-span genes such as Wt1 (P = 0.02), Nphs1 (P = 0.01), Nphs2 (P = 0.01), Synpo (P = 0.01) and Cdkn1c (P < 0.01) (Figure 4A). Higher podocyte protein expression was confirmed by immunostaining for Wilms tumor protein (Wt1), Nephrin (Nphs1), Podocin (Nphs2) and Synaptopodin (Synpo) in both middle-aged/MCC950-treated and middle-agedNLRP3 null mice compared to middle-aged/vehicle-treated mice (Figure 4B–4U).

Figure 4. Podocyte health was improved upon inhibition/absence of NLRP3. (A) Comparison of the transcripts for canonical podocyte genes (Wt1, Nphs1, Nphs2, Synpo, Cdkn1c, Vegfa) using the normalized mRNA-seq read counts from isolated podocytes of middle-aged saline-treated (red bars) and MCC950-treated mice (blue bars) compared to their young counterparts (grey bars). (B–Z) Immunostaining of kidneys from young, saline- and MCC95-treated middle-aged as well as middle-aged Nlrp3 null mice for WT-1/Wt1 (B–F), Nephrin/Nphs1 (G–K), Podocyin/Nphs2 (L–P), Synaptopodin/Synpo (Q–U) and VEGFA (V–Z). Representative images are shown; scale bars represent 25 μm. Staining intensities were quantified and depicted in the graphs as % tuft area.

Finally, we analyzed vascular endothelial growth factor (VEGF) which is critical for maintenance of podocyte/endothelial cell interactions [50], as a surrogate for podocyte synthetic function. Vegfa mRNA levels decreased in middle-aged/vehicle-treated mice but were partially restored in podocytes from middle-aged/MCC950-treated (P = 0.01) (Figure 4A). This was partially validated at the protein level by immunostaining for VEGFa, with a trend toward increased staining in the middle-aged/MCC950-treated mice and a significant increase in the middle-aged NLRP3 null mice (Figure 4V–4Z). Together, these results show that lowering/deleting NLRP3 improved podocyte health-span in middle-aged mice.

Transcriptomic changes in middle-aged podocytes regulated by the NLRP3 inflammasome

To obtain a better understanding of potential mechanisms responsible for the slowing of glomerular aging upon inhibition of NLRP3, we isolated podocytes from individual mice from each group (young, middle-aged/vehicle-treated and middle-aged/MCC950-treated) and performed transcriptome analyses. Principal component analysis showed excellent clustering of the individual groups (Figure 5A) with many differentially expressed transcripts between middle-aged/MCC950-treated vs. middle-aged/vehicle-treated podocytes (Figure 5B). As expected, GSEA analysis demonstrated that podocytes from mice treated with MCC-950 exhibited inhibition of the NLRP3 signaling pathway (Figure 5C). We next analyzed the significantly altered transcripts for their contribution to the aging process. Using a threshold of 2-fold and a P-value of 0.05 a total of 442 genes were down-regulated and 435 genes were up-regulated in middle-aged/MCC950-treated podocytes compared to middle-aged/vehicle-treated podocytes (Figure 5D, 5E). Of the down-regulated transcripts, 124 genes (9%) were also up-regulated in aged vs. young podocytes. Similarly, of the up-regulated transcripts, 75 genes (7%) were down-regulated in aged vs. young podocytes. GSEA analysis (Supplementary Figure 1B) confirmed the data shown in Figure 2 at the global level where both up- and down-regulated genes in response to senescence were restored by NLRP3i treatment (Figure 5F–5I). These findings were extended by examining a gene set specifically generated to assess kidney aging [51] which showed an even more pronounced impact (Figure 5J–5M). In summary, these data suggest activation of NLRP3 signaling is part of the natural podocyte aging process.

Figure 5. Transcriptomic analysis of NLRP3 inhibition. (A) Principal component analysis of the individual samples (young, middle-aged and middle-aged treated with MCC950 (NLRP3i)). (B) Volcano Plot comparing the saline- and MCC950-treated middle-aged podocyte transcriptomes; transcripts changed >2 and with a p-value > 0.05 are indicated in blue when down regulated upon NLRP3i and in red, when increased. (C) GSEA plot of the NLRP3 inflammasome gene set decreased upon treatment with MCC950. (D, E) Venn diagrams comparing the transcripts up-regulated in middle-aged podocyte and down-regulated upon NLRPi (D) and those down-regulated in the middle-aged and up-regulated upon NLRP3i (E) using the differentially expressed genes determined by the DSEQ2 analysis. (F–I) GSEA plots for the Reactome gene set cellular senescence comparing young to middle-aged (F) and middle-aged to middle-aged to NLRP3i-treated (G). Most differentially up- and down-regulated transcripts are compared along all three conditions in the histograms in panel H and I, respectively. (J–M) GSEA plots for the Rodwell aging kidney up (J) and down (I) gene sets comparing middle-aged to middle-aged to NLRP3i-treated. Most differentially expressed transcripts for either gene set are compared along all three conditions in the histograms (L, M).

Age-regulated inflammation and podocyte signaling were restored by inhibiting NLRP3

Interestingly, GSEA showed that inhibition of NLRP3 signaling was consistent with the reversal of the inflammatory phenotype in podocytes (Supplementary Figure 1C). In addition to the NLRP3 inflammasome, several additional inflammatory pathways that were increased in middle-aged podocytes compared to young were at least partially lowered in mice treated with MCC950. This included allograft rejection (Figure 6A), IL6 signaling (Figure 6B), Interferon alpha (Figure 6C), IL2/Stat5 signaling (Figure 6D), complement (Figure 6E) and NOD-like receptor signaling (Figure 6F). This effect was confirmed by immunostaining for phospho-STAT3, a marker of active interferon and IL6 pathways (Figure 6G–6J). Phospho-STAT3 immunostaining was not detected in podocytes of young mice, was increased in middle-aged saline-treated mice, but was reduced by MCC-950 treatment and was barely detected in the podocytes of NLRP3 null mice. A similar trend was observed for phospho-STAT5, a marker for active IL2 signaling (Figure 6K–6N) Importantly, not all inflammatory pathways were increased in middle-aged saline-treated mice nor reduced following NLRP3 inhibition. For example, immunostaining for phospho-IKKa/β, a marker of active TNF signaling, did not increase in middle-aged glomeruli (Figure 6O–6R).

Figure 6. Podocyte inflammaging is reduced by inhibiting NLRP3. (A–F) GSEA plots comparing the transcriptomic data from saline- and MCC950-treaded middle-aged mice for 5 hallmark gene sets (allograft rejection, IL-6/JAK/STAT3 signaling, interferon alpha response, IL-2/STAT5 signaling and complement) as well as one KEGG gene set (Nod-like receptor signaling pathway). (G–R) Immunoperoxidase staining comparing glomeruli from young mice, to middle-aged either treated with saline or MCC950 as well as middle-aged NLRP3 null mice using phospho-STAT3 (G–J), Phospho-STAT5 (K–N) or Phospho-IKKa/β (O–R). Representative images are shown; scale bars in the images correspond to 25 μm. Note that in line with the GSEA data phospho-STAT3 staining was strongly reduced by interfering with NLRP3 signaling in middle-aged mice, while STAT5 showed a more modest change.

NLRP3 inhibition does not majorly impact apoptosis, endoplasmic reticulum stress or autophagy

A reduction in canonical gene/protein expression in aged podocytes is normally accompanied by cell death, detachment and an inability of podocytes to self-renew to replace the lost cells [52]. Interestingly, besides the KRAS hallmark pathway, which has been shown to be functionally connected to NLRP3 signaling, [53, 54] none of those pathways were identified by the GSEA (Supplementary Figure 1B). To further substantiate this observation, we performed immunofluorescence studies comparing glomeruli of young, saline- and MCC950-treated middle-aged mice. Indeed, apoptosis as a contributor to podocyte death seems unlikely, as we were unable to detect increased glomerular Caspase-3 staining in control middle-aged mice (not shown). Other pathways shown to lead to podocyte loss [52] such as the endoplasmic reticulum stress (ERS) [55, 56] and autophagy stress pathways were also only slightly or not at all impacted by NLRP3i treatment. Immunostaining for the autophagy marker LC3 was barely detected (Figure 7A–7D). Conversely, the ERS associated proteins GRP94/Hsp90b1 were up-regulated in the middle-aged mice, but not significantly restored by MCC950 treatment (Figure 7E–7H). Finally, the active mTOR marker pS6 ribosomal protein (pS6RP) was higher in saline treated middle-aged mice but was unaffected by MCC950 treatment (Figure 7I–7L). Together, these data suggest that the pro-podocyte effect of NLRP3 inhibition at middle-age was primarily caused by inhibition of pyroptosis and only to a small extent by reduced ER stress and augmented autophagy.

Figure 7. MCC950 treatment improves autophagy signaling. Immunostaining for autophagy marker LC3 (A–D), for the ER stress response protein GRP94 (E–H) and the mTOR signaling readout pS6RP (I–L) comparing young to middle-aged saline- and MCC950-treated mice (I–L). Panels I–K were counterstained with Lotus Tetragonolobus Lectin (LTL, blue) to visualize proximal tubules. Abbreviation: BC: Bowman’s Capsule. Images were quantified and depicted as % tuft area Scale bars in the images correspond to 25 μm.

Impact of NLRP3 inhibition on non-podocyte glomerular cells

Compared to young mice, middle-aged/vehicle-treated mice had a decrease in the density of parietal epithelial cells (PAX8 staining) (Figure 8A–8D) accompanied by an increase in their activation (CD44 staining) (Figure 8E–8H). Glomerular endothelial cell density (ETS-related gene, ERG [57] staining) was also decreased in middle-aged/vehicle-treated mice (Figure 8I–8L), accompanied by an increase in the glomerular endothelial injury marker plasmalemmal vesicle associated protein-1 (PV1) [58] (Figure 8M–8P). The mesangial marker alpha 8 integrin was higher in middle-aged/vehicle-treated mice compared to young mice (Figure 8Q–8T). However, in contrast to the data shown earlier on middle-aged podocytes, none of the changes in non-podocyte glomerular cells were impacted by inhibiting NLRP3 with MCC950 (Figure 8). In contrast, the increase in the tubular epithelial cell injury marker KIM1 was reduced in middle-aged/MCC950-treated mice (Figure 8U–8X).

Figure 8. Impact of MCC950 treatment on other glomerular cells and proximal tubules. Immunostainings comparing young to middle-aged saline- and MCC950-treated mice. (A–D) PAX8 (brown, nuclear) and Collagen IV staining (blue) was used to visualize parietal epithelial cells (PECs) and Bowman’s capsule, respectively. Data were quantified and PEC density was calculated by the ratio of PAX8-positive cells and Bowman’s capsule (BC) length (D). (E–H) Immunostaining for CD44 (brown) and LRP2 (blue) was used to label activated PECs and proximal tubules, respectively (E–G). To determine the number of glomeruli with activated PECs, data were quantified based on the presence of CD44 staining in BC (H). (I–L) Endothelial cell number was determined staining with the endothelial cell marker ERG (nuclear brown staining) and the BC marker Collagen IV (blue) (I–K). Data were quantified as number of nuclei per mm² (L). (M–P) To assess endothelial injury, kidney sections were stained for PV-1 (brown) as a marker of injured endothelial cells (brown) and quantified as percentage of glomeruli with PV-1-positive cells (P). (Q–T) Alpha-8 integrin (ITGA8) staining (brown) was used as a measure of the mesangial area (brown) (Q–S) and quantified as the percent ITGA8-positive area within the glomerular tuft (T). (U–X) To determine the effect of MCC950 on proximal tubules, sections were stained with KIM1 (brown) (U–W) and quantified as IOD per area (X). In all panels representative images are shown and scale bars in the images correspond to 25 μm. In all the graphs, error bars are standard deviation, and the mean levels are stated by the number above the bars.

NLRP3 signaling is increased in injured human kidney organoids

To address whether inhibiting NLRP3 signaling improves podocyte injury and if NLRP3 increases senescent inducing genes in humans, we used a human kidney organoid model. Day 14 human kidney organoids were exposed to cytopathic anti-podocyte antibody with and without NRLP3 inflammasome inhibitor, MCC950. Induction of experimental FSGS caused upregulation of NRLP3 transcription along with other canonical NRLP3 inflammasome and senescence markers such as IL-1β, CASP1, PYCARD and CDKN2D (Supplementary Figure 5). Concurrent treatment with MCC950 resulted in a decreased expression of the NRLP3 inflammasome markers and CDKN2D (Supplementary Figure 5). These data support the notion that injury compounds the increase in NLRP3 and senescence in podocytes.

NLRP3 signaling in liver aging

We have previously shown that inhibition of PD1 signaling improves not only podocyte aging but also has a beneficial effect on liver aging [19]. Thus, we wondered whether lowering NLRP3 would have a similar anti-aging effect in the liver. As in the kidney, the age-associated increase in the NLRP3 target IL-1β was reduced by MCC950, consistent with NLRP3 inhibition (Supplementary Figure 6A–6C, 6E). This was accompanied by a decrease in the characteristic fat deposition in aging liver visualized using Oil Red O staining (Supplementary Figure 6F–6H, 6J). The effects from the MCC950 treatment were confirmed in middle-aged NLRP3 null mice (Supplementary Figure 6D, 6E, 6I, 6J). Together, these data suggest that the beneficial consequences of inhibition of NLRP3 signaling were not restricted to the kidney and that systemic NLRP3 inhibition has a more widespread anti-aging effect.

Animals

Experimental male C57BL/6J mice (WT, Strain #000664) aged 18 months (n = 20) were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). Animals received 10 mg/kg MCC950 sodium (Cat# HY-12815A, MedChemExpress) in saline (Sodium chloride 0.9% Injectable, BD), or saline vehicle alone, via twice weekly IP injections, for 6 weeks. Control male C57BL/6J (WT, Strain #000664) mice aged 4 (n = 5), 12 (n = 5), 18 (n = 5), 24 (n = 5), and 27 (n = 5) months were from our lab animal tissue archives and originally obtained from The Jackson Laboratory (Bar Harbor, ME, USA). Breeding pairs to generate male B6.129S6-Nlrp3tm1Bhk/J (Nlrp3KO, Strain #021302) were obtained from The Jackson Laboratory then aged to 19.5 months (n = 11).

Experimental focal segmental glomerulosclerosis (FSGS) was induced in 4-month-old (n = 6) and 24-month-old (n = 4) wildtype male C57BL/6J mice (Strain #000664, The Jackson Laboratory) by two intra-peritoneal (IP) injections, 24 hours apart, of a cytopathic sheep anti-glomerular antibody (9 mg/20 g body weight) [82–84]. Kidney tissue was collected at baseline 4 m (n = 5) and 24 m (n = 5) and 28 days following antibody injection.

Nphs1-FLPo|FRT-EGFP double transgenic mice were bred in house to express FLP recombinase as previously described by Goldberg et al. [85], resulting in the FLP mediated excision of an FRT flanked STOP in the RCE: FRT [86] (FRT-EGFP, Strain #010812) mouse available from The Jackson Laboratory (Bar Harbor, ME, USA). Resulting animals have permanent EGFP expression in all podocytes, for the lifetime of the animal. There were 2 groups of Nphs1-EGFP reporter mice: young (4 months, n = 15) and aged (24 months, n = 24).

All mice were housed in the animal care facility at the University of Washington under specific pathogen-free conditions. All animals were co-housed in social groups of 2–5 animals, provided with environmental enrichment (nestlets), maintained on a 14-hour light/10-hour dark cycle, at 70 ± 5 F and 50 ± 10% relative humidity, with ad libitum access to food and water. Aging animals (over 18 months) received additional monitoring according to our aging mice protocols to ensure healthy aging. Mice were sacrificed at the ages described above by cervical dislocation according to American Veterinary Medical Association (AVMA) guidelines for the euthanasia of animals by certified personnel. Animal protocols were approved (2968-04) by the University of Washington Institutional Animal Care and Use Committee. Weights were taken bi-weekly for the duration of the study for animals receiving interventions and all animals were checked daily by vivarium personnel.

Urine analysis

Urine was collected by pooling two spots on a single day at baseline and week 6 of MCC950 treatment. Albumin to creatinine ratio was calculated from urinary albumin measured by radial immunodiffusion assay (RID) as previously described [87] and creatinine measured by Creatinine (urinary) Colorimetric Assay Kit (Cayman Chemical, Ann Arbor, MI, USA).

Blood urea nitrogen and suPAR

Blood was collected at sacrifice. Blood Urea Nitrogen was measured with a colorimetric Urea Assay Kit (Abcam, Cambridge, United Kingdom). Circulating suPAR levels were measured in plasma samples collected at sacrifice, using Mouse uPAR DuoSet ELISA (Cat, #DY531, R&D Systems Minneapolis, MN, USA).

Immunostaining, quantification and visualization

Immunoperoxidase staining was performed on 4 μm thick formalin fixed paraffin-embedded (FFPE) mouse and human kidney sections as previously described [88]. Briefly, paraffin-embedded, formalin-fixed sections were incubated in Histoclear (National Diagnostics, Atlanta, GA, USA) and graded series of ethanol for rehydration. Sections were boiled in antigen retrieval buffer (Citric acid buffer pH 6.0 or EDTA buffer pH6.0, pH 8.0). To avoid nonspecific protein binding, Background Buster (Accurate Chemical and Scientific, Westbury, NY, USA) was used, as well as goat anti-rabbit Fab fragment and rabbit Fab fragment (Jackson ImmunoResearch Laboratories, West Grove, PA, USA). Suppression of endogenous biotin activity was performed using Avidin/Biotin Blocking Kit (Vector Laboratories, Burlingame, CA, USA). Initial primary antibodies were incubated overnight at 4°C. In the case of double immunostaining, subsequent primary antibodies were incubated either overnight at 4°C or for 3 hours at room temperature. Secondary antibodies and streptavidin conjugates were incubated for one hour at room temperature. The primary antibodies used in the study are summarized in Supplementary Table 1. Diaminobenzidine (Sigma-Aldrich, St. Louis, MO, USA) with or without 8% nickel chloride, was precipitated to visualize immunoperoxidase staining.

Immunofluorescence staining was performed on frozen sections that were thawed and washed in PBS. Gentle antigen retrieval was performed by warming citric acid buffer (pH 6.0) or EDTA buffer (pH 8.0) to 38°C. They underwent the same blocking and antibody incubation protocol as described for immunoperoxidase staining. All immunofluorescence samples were mounted using Vectashield mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA, USA).

To quantify immunohistochemistry, slides were scanned in brightfield with a 20× objective using a NanoZoomer Digital Pathology System (Hamamatsu City, Japan). The digital images were imported into Visiopharm software (Hoersholm, Denmark) and its Image Analysis Deep Learning module was trained to detect glomeruli and assess immunohistochemical staining-positivity for Collagen IV, p57, Nephrin, Synaptopodin and ERG. The glomeruli ROIs were processed in batch mode generating per area outputs, cell counts and analyzed from 100% of the tissue sections. Quantification of podocin, VEGFA, NLRP3, IL-1B, LC3, GRP94, p16 and SA-b-galactidase staining was performed by using ImageJ 1.46r software (National Institutes of Health).

Senescence-associated β-galactosidase (SA-β-Gal) staining

Frozen kidney sections (10 μm) were stained with SA-β-gal per kit instructions (Cell Signaling Kit #9860), then counter stained with nuclear fast red to determine senescence-associated-β-galactosidase [89, 90]. The percentage of glomeruli with SA-β-gal-positive staining in the glomerular tuft was quantified in outer cortex of each kidney section.

Assessment of liver aging

Immunostaining was performed for Oil Red O (to detect lipids and triglycerides) as previously described [91]. To determine morphological changes present in aged liver, frozen liver sections (6 μm) were stained for Collagen IV and the endothelial marker CD31/PECAM-1 as previously described [92].

FLARE staining and expansion microscopy

Kidney sections were cut to a thickness of 50 μm from cryo-preserved mouse kidney tissue blocks. Hydrogel-expansion and FLARE staining of kidney sections were performed according to the published FLARE protocol [93]. In brief, oxidized carbohydrates were labeled with ATTO 565 hydrazide, amines (proteins) were labeled with ATTO 647N NHS ester, and nuclei were labeled with the fluorescent DNA-binding dye SYBR Green I. Expanded samples were transferred onto a poly-L lysine coated coverslip (24 mm by 50 mm, no. 1.5; Fisher Scientific, #12544E) and imaged immediately. Images were acquired using a Nikon A1R inverted point-scanning confocal microscope at the University of Washington Biology Imaging Facility. A CFI Apo LWD Lambda S 40× objective lens with 1.15 numerical aperture was used on the microscope. Three-color 3D stacks were acquired with 130 nm lateral sampling and 570 nm axial sampling. The final FLARE images are red (amines), green (carbohydrates) and blue (DNA).

Magnetic activated cell sorting (MACS) of podocyte and non-podocyte cell fractions

Kidney tissue (w/o the kidney capsule and surrounding fat) was placed into ice cold RPMI 1640 medium (w/o L-glutamine and phenol red, GE Healthcare Bio-Sciences, Pittsburgh, PA, USA). After removal of the medulla, the remaining cortex was minced into fine pieces and digested in 0.2 mg/ml Liberase™ TL (Sigma-Aldrich, St. Louis, MO, USA), 100 U/ml DNAse I (Sigma-Aldrich, St. Louis, MO, USA) in RPMI 1640 medium (w/o L-glutamine and phenol red) by shaking at 37°C for 30 minutes. The digest was passed through an 18G needle (Becton Dickenson, Franklin Lakes, NJ, USA) 10 times and enzymes were inactivated by adding 5 ml of RPMI 1640 medium (w/o L-glutamine and phenol red) supplemented with 1 mM sodium pyruvate (ThermoFisher Scientific, Waltham, MA, USA), 9% Nu-Serum™ IV Growth Medium Supplement (Corning Incorporated-Life Sciences, Durham, NC, USA) and 100 U/ml Penicillin-Streptomycin (ThermoFisher Scientific, Waltham, MA, USA). The cell suspension was passed through a 100 μm and a 40 μm cell strainer (BD Biosciences, San Jose, CA, USA) and pelleted by centrifugation at 200G at 4°C for 5 minutes. Cells were were resuspended in media containing two rabbit anti-Nephrin antibodies [94] (1:100, Abcam, Cambridge, MA, USA). After 1 hour at 4°C, cells were pelleted, washed in media and incubated with anti-rabbit microbeads (Miltenyi Biotec, Auburn, CA, USA) along with Alexa Fluor^® 594-conjugated AffiniPure Donkey Anti-Rabbit IgG 1:200 (in order to visualize binding of the Nephrin antibodies to the podocytes) for 30 minutes at 4°C. Cells were pelleted and washed in PBS with 0.5% BSA and 2 mM EDTA and applied to MACS LS columns (Miltenyi Biotec, Auburn, CA, USA) to gently separate microbead-bound podocytes from the other kidney cells. Cells not retained by the magnetic field were collected, pelleted and designated non-podocyte (NP) fractions. LS columns were removed from the magnetic field then washed with PBS with 0.5% BSA and 2 mM EDTA and podocytes were collected. A small aliquot of each fraction was imaged using an EVOS FL Cell Imaging System to verify podocyte isolation, based on the presence of Nephrin antibody. Additionally, qRT-PCR for a panel of podocyte genes was performed in both podocyte and non-podocyte fractions to confirm cell type identity.

RNA isolation, qRT-PCR, library preparation and sequencing

mRNA was isolated using the RNeasy Mini Kit (Qiagen, Germantown, MD, USA) as per the manufacturer’s instructions and used for bulk mRNA sequencing or converted to cDNA by reverse transcription with the High-Capacity RNA-to-cDNA Kit (Thermo Fisher, Waltham, MA, USA) and utilized for quantitative real-time PCR analysis. qRT-PCR was performed using iTaq SYBR Green Supermix (Bio-Rad, Hercules, CA, USA) and a QuantStudio 6 Flex real-time PCR System (Applied Biosystems, Waltham, MA, USA) as we have previously described [20]. Relative mRNA expression levels were normalized to glyceraldehyde-3-phosphate dehydrogenase (Gapdh) levels. Library generation and bulk next-generation mRNA sequencing were performed by Psomagen, Inc. (Rockville, MD, USA) using TruSeq RNA Library Prep Kits (Illumina, San Diego, CA, USA) and the Illumina platform.

For the data analysis of the mRNA-seq data raw reads from fastq files were trimmed using TrimGalore, aligned to mm10 using Bowtie2 [95] and gene-level read counts were obtained using featureCounts [96]. Data were analyzed by Principal Component Analysis (PCA) using pcaExplorer [97]. DESeq2 [98] was used to identify differentially expressed genes (DEGs), which were defined as genes with a p-value of 0.05 and a >2-fold change comparing young to middle-aged or middle-aged to NLRP3 inhibition. Volcano Plots were generated using GGPlot2 (v3.4.0), Dplyr (v1.0.10) and Tidyverse (v1.3.2). Venn diagrams were generated using VennDiagram (v1.7.3), Dplyr (v1.0.10), Tidyverse v1.3.2.

Gene Set Enrichment Analysis (GSEA) was used to identify perturbed biological processes (GSE226796) (NCBI tracking system #23740500) [99]. Heatmaps were generated to demonstrate the differential expression of the genes part of individual gene sets (i.e., multiple pathways of the Hallmark data sets, the Cellular Senescence pathway of the C2.cp.reactome, and the Rodwell Aging Up/Down pathway of C2.cgp). The genes relevant to these pathways were extracted from the gene expression data used in the GSEA. The most upregulated genes in young, middle-aged, and MCC950-ionjected middle aged (NLRP3i) podocytes were mapped against one another. Heatmaps were generated using GGPlot2 (v3.4.0), Dplyr (v1.0.10), Tidyverse (v1.3.2), Patchwork (v1.1.2), ComplexHeatmap (v2.12.1). In all cases, the heatmaps were scaled between -1 and 1. All the R scripts are available upon request.

Gene expression analysis of human kidneys

Kidney tissue was obtained from the unaffected parts of kidneys removed from patients undergoing surgery at the University of Michigan and processed immediately via the tissue procurement service of the Department of Pathology as described [19, 38]. Clinical data were obtained through the honest broker office of the University of Michigan as we have reported.⁸⁶ Tissue was placed right away in RNAlater, micro-dissected into glomeruli and tubule-interstitial fractions, and isolated total RNA was used for gene expression analysis using polyA-RNA-sequencing as described [19]. Human kidney gene expression analysis studies were approved by the Institutional Review Board of the University of Michigan (IRB approval number: HUM00052918, HUM00165536). IRB 14-0019 from the University of Chicago was used for the immunostaining of human kidneys.

Human kidney biopsies

Paraffin-embedded formalin fixed human kidney biopsies used were from the archives of Department of Pathology, University of Chicago (Chicago, IL, USA). These include healthy young human kidney tissue (23 years old), healthy middle age human kidney tissue (43, 45) years old and aged human kidney tissue (greater than 65) years old. These tissues underwent similar immunostaining protocol as were for mice.

Human kidney organoids

Human kidney organoids derived from MANZ2 iPSCs were generated according to previously developed protocols [100–102]. Briefly, a cell culture treated 24 well plate (TPP 92012) was coated with anti-adherence solution (Fisher Scientific NC0488042) for 10 minutes, followed by 3 subsequent washes with sterile PBS. Day 14 organoids were placed into each well. Organoids were treated in three condition arms (n = 10–12 organoids/condition): untreated (control), 20 mg/mL cytotoxic antibody treated, or 20 mg/mL cytotoxic antibody + 10 mM MCC950 (MCE HY-12815A) treated every 24 hrs for a total of 48 hrs. Control treatment included equal volume of PBS to the antibody treatment arm. Organoids were harvested at day 16 by washing in PBS three times, removing supernatant, and resuspending in TRI reagent (Thermo Fischer AM9738). RNA was isolated using RNA isolation kit (Thermo Fisher 12-183-018A), and cDNA was synthesized using qScript cDNA SuperMix (Thermo Fisher NC0479275). qPCR was carried out and measured using QuantStudio 12K Flex and analysis was subsequently carried out in GraphPad Prism. Primer sequences for all qPCR markers were obtained from Massachusetts General Hospital PrimerBank: NRLP3-FOR, GATCTTCGCTGCGATCAACAG; NRLP3-REV, CGTGCATTATCTGAACCCCAC; IL1B-FOR, ATGATGGCTTATTACAGTGGCAA; IL1B-REV, GTCGGAGATTCGTAGCTGGA; CASP1-FOR, TTTCCGCAAGGTTCGATTTTCA; CASP1-REV, GGCATCTGCGCTCTACCATC; PYCARD-FOR, TGGATGCTCTGTACGGGAAG; PYCARD-REV, CCAGGCTGGTGTGAAACTGAA; CDKN2D-FOR, AGTCCAGTCCATGACGCAG; CDKN2D-REV, ATCAGGCACGTTGACATCAGC.

Statistical analysis

Data are shown as the mean ± S.E.M. Student’s t-test was applied for comparisons between two groups. P-values < 0.05 were considered statistically significant. Multiple groups were compared using one-way ANOVA with post hoc Tukey HSD test. P-values < 0.05 represented statistically significant differences.