Developing a targeted approach to halt CKD progression remains challenging due to the complexity of molecular and cellular mechanisms underlying kidney fibrosis45,46. Transcriptional and functional heterogeneity of the key effector population contributing to aberrant ECM deposition and pathologic tissue remodeling – kidney fibroblasts – remains a crucial challenge in the field. While recent research has made significant progress in unraveling the molecular nature of fibrotic pathologies2, a comprehensive strategy to exclusively trace and target kidney fibroblasts is still elusive, due to the lack of specificity of currently used markers47–56. The current study presents Gucy1α1 as a newly validated kidney marker capable of labelling fibroblasts both comprehensively and selectively.
We and others recently used single-cell or single-nucleus RNA-seq to dissect the heterogeneity of kidney fibroblasts15,31,57. Consistent with Li. et al.57, we observed transcriptional and functional heterogeneity among kidney fibroblasts, which were divided based on their gene expression enrichments into “secretory” (“Fibro 1”), “contractile” (“Fibro 2”) and “migratory” (“Fibro 3”) clusters31. Gucy1α1 comprehensively marked secretory (high in ECM related genes such as Col1a1/2, Col3a1, Fn1), contractile (Acta2, Myh11, Myl9 high) and migratory (Sfrp2, Pdgfrβ, Amotl1 high) fibroblasts with no off-target expression in any other epithelial, endothelial, immune or podocyte populations. We therefore examined Gucy1α1 as a novel marker that might allow for tracing and targeting kidney fibroblasts in an exclusive manner at multiple AKI-to-CKD transition stages. With that goal, we established two independent clinically relevant murine models of kidney fibrosis caused by UIR or UUO58 which both elicited progressive kidney fibrotic remodeling. We found that Gucy1α1 mRNA and protein levels progressively increased as fibrosis in both models advanced. Moreover, we discovered a significant direct correlation between Gucy1α1 and classical fibrotic remodeling markers, such as Pdgfrβ, αSma and Vim. Pdgfrβ is traditionally used as a kidney fibroblast marker in many studies12,15. Mildly expressed in the normal kidney stroma, including glomerular mesangium and interstitial fibroblasts59, Pdgfrβ becomes significantly elevated in both murine and human kidney fibrosis60. Our previous study31 showed that Pdgfrβ is present in all three fibroblast fractions at advanced stages of UIR and UUO. Acta2, which marks contractile “myofibroblast” phenotype, predominantly labelled “Fibro 2” populations of control and fibrotic kidneys with mild to absent expression in other fibroblast clusters. We also showed that Vim, while exhibiting medium expression levels in all three fibroblast populations, was also present in many off-target clusters, including immune cells and podocytes. In contrast, Gucy1α1 marked Vim- and αSma-expressing fibroblast fractions and exhibited near-total overlap with Pdgfrβ-positive interstitial fibroblasts in the control kidney and throughout all stages of fibrosis progression in both models. Since regional heterogeneity might represent a challenge to comprehensive kidney fibroblast labelling39, we separately assessed Gucy1α1 potential in cortex and medulla and found that it equally comprehensively labels kidney fibroblasts in both regions. The trajectory of double Gucy1α1/αSma and Gucy1α1/Vim positivity might reflect the process of resident fibroblast activation and “contractile” phenotype acquisition by a portion of them. Our analysis identified a regional molecular heterogeneity inside kidney stroma, while only a fraction of cortical Gucy1α1-positive fibroblasts retained αSma expression throughout the whole course of both injuries, most medullary fibroblasts remained double Gucy1α1/αSma positive at all stages of UIR and UUO. This distinction might reflect regional heterogeneity in “contractile” phenotype acquisition by activated kidney fibroblasts. Also, a higher percentage of cortical and medullary fibroblasts from UUO treated kidneys exhibited double Gucy1α1/αSma positivity, probably reflecting the more severe fibrotic response caused by UUO compared to UIR31,57.
While our scRNA-seq did not capture Pdgfrβ gene expression in podocytes, multiple rounds of IF reproducibly demonstrated significant Pdgfrβ expression which colocalized with podocyte marker Nphs1 in both control and fibrotic kidneys, compared to no Gucy1α1 intraglomerular presence. Intraglomerular Pdgfrβ expression was also corroborated by another independent study,61 which showed robust Pdgfrβ upregulation in the cells occupying glomerular Bowman’s space triggered by focal segmental glomerulosclerosis. Furthermore, recent scRNA-seq62 identified PDGFRΒ expression in many glomerular cell types, including almost all human and some murine podocytes, murine parietal epithelial cells along with human and murine mesangial-like cells. Notably, Kidney Precision Medicine Project data also demonstrated lack of GUCY1Α1 and presence of PDGFRΒ expression in the human podocytes.63 The discrepancy between our scRNA-seq and IF data might reflect the differences between Pdgfrβ RNA and protein expression patterns, or the ability of scRNA-seq to capture murine Pdgfrβ-expressing podocytes. However, we and others independently demonstrated that while Pdgfrβ labels interstitial fibroblasts, it also exhibits off-target expression among many glomerular populations on the protein level. Collectively, these findings indicate that Gucy1α1 might serve as a better marker for exclusive labeling of all interstitial quiescent and activated fibroblasts with no off-target intraglomerular effects.
Due to the increased importance of accounting for sex differences in the kidney molecular
landscapes and injury response64–66, we conducted a set of experiments to evaluate Gucy1α1 in the female murine CKD model. We found that Gucy1α1 mRNA and protein levels reflected the degree of ECM deposition in the female UIR induced fibrosis. Female kidneys exhibited intratubular pattern of Gucy1α1 expression, which underwent robust upregulation following UIR and overlapped with Pdgfrβ-, αSma- and Vim-positive interstitial areas. These findings suggest that Gucy1α1 holds a potential as a specific fibroblast marker in both sexes.
Another challenge in the field is to identify reliable fibroblast labelling strategies that are applicable to both experimental models and human specimens with kidney fibrosis16. Since Gucy1α1 levels are reflective of fibrosis degree in the murine UIR and UUO models, it might serve as a predictor of kidney injury progression trajectory either towards recovery or exacerbated fibrotic remodeling and CKD/ESKD. GUCY1α1 expression in αSMA- and VIM-expressing cells of the normal human kidney, and its upregulation in the fibrotic human kidney, suggests that it can be used as a marker of kidney fibroblasts and predictor of kidney injury outcome in humans. Our finding corroborated scRNA-seq and single nuclear RNA-seq (snRNA-seq) predicted GUCY1Α1 expression in the human kidney stromal cells, such as fibroblasts, myofibroblasts and pericytes, reported by Kidney Precision Medicine Project63. We also revealed mild GUCY1α1 expression in the normal human lung which was elevated in interstitial pulmonary fibrosis (IPF), colocalizing with αSMA- and VIM. This human finding is consistent with the previous murine scRNA-seq data, showing Gucy1a1 expression in the pulmonary ECM-producing Acta2-positive populations67.
Recent scRNA-seq analysis68 identified Gucy1a1 as a pericyte marker in lung and kidney.
The authors defined pericytes based upon strong expression of Cspg4 and Pdgfrβ. However, none of these genes is exclusive to pericytes. According to our scRNA-seq data31, Pdgfrβ labels all three fibroblast populations including “contractile” Acta2-rich cluster. Likewise, Cspg4 was present in the “contractile” fibroblasts along with Fibro3 “migratory” population. Thus, Cspg4-/Pdgfrβ-positive clusters identified as “pericyte-enriched” or even “stringent pericytes” might represent a subset of fibroblasts or fibroblast/pericyte mixture. To further address the multiorgan potential of Gucy1α1, we conducted a series of studies in other organs prone to fibrosis, such as heart and liver. In both organs, Gucy1α1 levels parallelled the trajectory of fibrotic injury progression and co-labelled baseline Pdgfrβ-/Vim-positive and activated Pdgfrβ-/Vim-/αSma-expressing fibroblasts. Our finding is consistent with the recent scRNA-seq which showed Gucy1a1 enrichment in the portal fibroblasts and hepatic stellate cells in two other models of liver fibrosis69. Overall, our findings highlight Gucy1α1 as a novel marker which selectively labelled kidney fibroblasts in both sexes and exhibited translational multiorgan potential.