Circulating Activin A, Kidney Fibrosis, and Adverse Events

Visual Abstract

Abstract

Background

Identification of reliable biomarkers to assess kidney fibrosis severity is necessary for patients with CKD. Activin A, a member of the TGF-β superfamily, has been suggested as a biomarker for kidney fibrosis. However, its precise utility in this regard remains to be established.

Methods

We investigated the correlation between plasma activin A levels, kidney fibrosis severity, and the incidence of major adverse kidney events in patients who underwent native kidney biopsies at a tertiary medical center. We performed RNA sequencing and histological analyses on kidney biopsy specimens to assess activin A expression. In vitro experiments were also conducted to explore the potential attenuation of TGF-β–induced fibroblast activation through activin A inhibition.

Results

A total of 339 patients with biopsy-confirmed kidney diseases were enrolled. Baseline eGFR was 36 ml/min per 1.73 m², and the urine protein/creatinine ratio was 2.9 mg/mg. Multivariable logistic regression analysis revealed a significant association between plasma activin A levels and the extent of tubulointerstitial fibrosis. Our RNA sequencing data demonstrated a positive correlation between kidney INHBA expression and plasma activin A levels. Furthermore, the histological analysis showed that myofibroblasts were the primary activin A–positive interstitial cells in diseased kidneys. During a median follow-up of 22 months, 113 participants experienced major adverse kidney events. Cox proportional hazards analysis initially found a positive association between plasma activin A levels and kidney event risk, but it became insignificant after adjusting for confounders. In cultured fibroblasts, knockdown of activin A significantly attenuated TGF-β–induced fibroblast–myofibroblast conversion.

Conclusions

Plasma activin A levels correlate with kidney fibrosis severity and adverse outcomes in various kidney disorders.

Introduction

Tubulointerstitial fibrosis plays a significant role in the progression of CKD and holds considerable importance in risk stratification.^1,2 However, the currently available methods for assessing kidney fibrosis severity have limitations.^3,4 While proteinuria and reduced eGFR serve as common indicators of kidney dysfunction, they do not provide histopathologic information for guiding treatment plans.^5,6 On the other hand, kidney biopsy is underutilized because of its invasive nature and associated risks, with only a small fraction of patients with kidney impairment undergoing the procedure.⁷ Recent imaging advancements provide a chance for noninvasive evaluation of kidney microstructure, but limited accuracy and accessibility hinder widespread clinical use.^8,9 In this regard, novel circulating biomarkers exhibit promising potential for identifying and monitoring kidney disease progression.^10–12 Although the performance of these biomarkers require evaluation in well-designed clinical trials, some of them such as activin A, a member of the TGF-β superfamily, have been implicated in the pathogenesis of a variety of kidney diseases.^13–15 Animal studies have demonstrated increased activin A expression in kidney injury models, promoting tubulointerstitial fibrosis and CKD progression.^16–19 Moreover, like certain molecules from injured kidneys, activin A may have systemic effects beyond its local impact.^20–23 Research indicates that elevated activin A levels correlate with disease severity, including complications like cachexia and chronic kidney disease–mineral and bone disorder.^22,23

Although evidence suggests activin A as a potential kidney biomarker, its association with kidney fibrosis extent and long-term outcomes in patients with advanced CKD is understudied. Some studies emphasize its detrimental role in kidney disease, but anti-inflammatory properties of activin A in human monocytes may potentially impede disease progression.^24,25 Furthermore, a continuous debate persists regarding the primary cell type responsible for producing activin A in the kidney. Some researchers report increased expression in tubular epithelium, while others identify myofibroblasts as the primary source in damaged kidneys.^16,19,26,27 In this study, we assessed plasma and kidney activin A levels and their correlation with kidney fibrosis severity and disease outcome in 339 participants with biopsy-proven kidney diseases. In vitro experiments explored the potential involvement of activin A in fibroblast activation and kidney fibrosis.

Methods

Patients in the Taipei Renal Transcriptomics and Outcomes Investigation

The Taipei Renal Transcriptomics and Outcomes Investigation (TRTOI) is an ongoing prospective observational study being conducted at the Taipei Veterans General Hospital in Taipei, Taiwan. This study focuses on a cohort of adult individuals (20 years and older) who underwent clinically indicated kidney biopsies or nephrectomies since October 2018. The primary aim of the TRTOI was to examine the kidney transcriptomic profiles and identify novel circulating biomarkers to assess the outcomes in these patients.²⁸ This study protocol received approval by the Institutional Review Board of Taipei Veterans General Hospital, and all procedures were conducted in accordance with the principles outlined in the Declaration of Helsinki. Before enrollment, written informed consent was obtained from all eligible patients. The study excluded patients who refused consent, kidney transplant recipients, those involved in other studies, and pregnant women, leaving 339 participants. On the same day as the biopsy, blood and urine samples were collected from each participant.

Gene Expression Profiling in the Tubulointerstitial Compartment from Patients in the TRTOI

A fraction of the obtained human kidney biopsy tissue was preserved in RNAlater (Invitrogen) and underwent manual microdissection to separate the glomerular and tubular compartments. The mRNA expression profile of the isolated tubulointerstitial portion was then sequenced using the Illumina NovaSeq 6000 Sequencing System, which generated paired-end reads with a length of 150 base pairs. Afterward, the processed reads were aligned to the human reference genome (GRCh37/hg19) using STAR (version 2.6.1). To quantify the expression levels, the RSEM software (version 1.3.1) was used to calculate the transcripts per million values for each sample.

Plasma Activin A Detection

Plasma samples were obtained from EDTA-anticoagulated whole blood and subsequently stored at −80°C until analysis. The quantification of plasma activin A was performed using a commercial ELISA kit (catalog number DAC00B; R&D Systems, Inc., Minneapolis, MN). The assay exhibited a dynamic range of detection from 15.6 to 1000 pg/ml, with intra-assay and interassay variabilities <5% and 10%, respectively. All samples were quantified in duplicates.

Histopathological Evaluation of Kidney Fibrosis

Kidney fibrosis parameters, including global glomerulosclerosis, interstitial fibrosis and tubular atrophy (IFTA), arterial sclerosis, and arteriolar sclerosis, were accessed as described in previous literatures.^1,28 The percentage of global glomerulosclerosis and the extent of IFTA were calculated and categorized into four severity levels: minimal (≤10%, score 0); mild (11%–25%, score 1); moderate (26%–50%, score 2); and severe (>50%, score 3). Vascular sclerosis was semiquantitatively assessed using a four-graded scale (none, mild, moderate, and severe), following the classification proposed by Srivastava et al.¹ In addition, the Mayo Clinic Chronicity Score (MCCS)²⁹ was also computed by summing the scores of global glomerulosclerosis (0–3), interstitial fibrosis (0–3), tubular atrophy (0–3), and arterial sclerosis (0–1). In this study, the IFTA scores were weighted by a factor of two to account for the separate scoring of IFTA in the original MCCS classification. The semiquantitative scale for arterial sclerosis was transformed into a binary value: 0 for none/mild and 1 for moderate/severe, aligning with the MCCS criteria.

Clinical Characteristics, Laboratory Data, and Major Adverse Kidney Events

Demographic information, underlying medical conditions, laboratory data, and clinical outcomes were extracted from the electronic medical record. Kidney function was accessed by the Modification of Diet in Renal Disease equation to estimate the eGFR, while the spot urine protein-to-creatinine ratio was used as a measurement of proteinuria severity. The primary outcome was a composite of major adverse kidney events (MAKE), defined as either a reduction in eGFR of 50% or more from baseline, kidney failure, or mortality attributed to cardiovascular or kidney diseases. All participants were followed until kidney failure, death, loss to follow-up, or the end of the study (January 31, 2023), whichever came first. Patients who were lost to follow-up were censored at the time of last eGFR measurement available.

Assess Kidney Activin A Expression through Immunostaining

We stained paraffin-embedded tissue samples from 22 patients with fibrotic kidney disorders (including diabetic kidney disease [DKD], nephrosclerosis, and focal segmental glomerulosclerosis) using activin A antibody (1:600; NBP1-30928, Novus Biologicals). We selected nine age-matched and sex-matched individuals with thin basement membrane disease and eGFR >90 ml/min per 1.73 m² as immunohistochemical staining controls. Double immunostaining for α-SMA was performed to confirm that myofibroblasts were the primary activin A–positive cells in the interstitium of fibrotic kidneys.

Cell Culture and Treatment

Normal rat kidney fibroblasts (NRK49F cells) were maintained in Dulbecco's Modified Eagle's Medium supplemented with 10% fetal bovine serum. On reaching 70%–80% confluence, the complete growth medium was replaced with serum-deprived medium, and cells were treated with exogenous activin A. To test the role of endogenous activin A in fibroblast activation, cells were transfected with short interfering RNA (siRNA) specific to Inhba (the gene encodes the inhibin β-A subunit, which forms activin A through homodimerization) or control siRNA using Lipofectamine 3000 (Invitrogen). After an 8-hour siRNA transfection, cells were treated with 10 ng/ml TGF-β and 200 ng/ml activin A (both from R&D Systems).

Real-Time Polymerase Chain Reaction and Western Blotting

Total RNA from NRK49F cells was extracted using NucleoZol reagent. Subsequently, 1 µg of RNA was reverse-transcribed into cDNA. The mRNA levels for specific genes were assessed through quantitative polymerase chain reaction and normalized to the Gapdh mRNA. A list of all primers used in this study can be found in Supplemental Table 1.

Proteins extracted from cultured cells were loaded onto an SDS-PAGE gel for electrophoresis and then transferred to a polyvinylidene fluoride membrane. The membrane was then incubated with primary antibodies specific to inhibin β-A, α-SMA, and collagen 3A1, followed by horseradish peroxidase-conjugated secondary antibodies. The abundance of the target protein was normalized to β-actin and analyzed using ImageJ software.

Statistical Analyses

The data were presented as mean±SD, median (interquartile range), or number (percentage) as appropriate. We assessed correlations between variables using Spearman's rank correlation test. Plasma activin A levels were either natural log-transformed or divided into tertiles for analysis. We conducted multivariable logistic regression, including baseline characteristics like age, sex, serum albumin, proteinuria, and eGFR, to examine the association between plasma activin A levels and chronic kidney injuries. The cumulative MAKE incidence was assessed on the basis of plasma activin A tertiles using Kaplan–Meier curves and the log-rank test. We also conducted Cox regression to determine the independent association of activin A with kidney outcomes, initially unadjusted, then adjusting for age, sex, serum albumin, urine protein/creatinine ratio, eGFR, and clinicopathological categories (DKD versus other diagnoses), with no significant collinearity in the multivariable analysis. Taking into account gonadal activin A secretion,³⁰ we studied its interaction with gender-concerning MAKE using statistical multiplicative terms. We also conducted sensitivity analysis with subdistribution hazards models considering the competing risk of death. All reported variables were complete with no missing data. To assess the effect of activin A knockdown under TGF-β treatment, a two-way ANOVA with Tukey's post hoc correction was used. Significance was defined as P < 0.05. Statistical analysis used SAS software (version 9.4; SAS Institute Inc., Cary, NC) and R software (version 3.5.2 for Windows).

Results

Baseline Patient Characteristics

Supplemental Figure 1 depicts the TRTOI participant enrollment flow chart, and Table 1 displays their baseline characteristics. The mean age was 55±17 years, with 58% being men. The median plasma activin A level was 382 pg/ml (257–526), eGFR was 36 ml/min per 1.73 m² (18–67), and urine protein/creatinine ratio was 2.9 (1.0–7.1). The most common primary clinicopathologic diagnoses were proliferative glomerulonephritis (29%), nonproliferative glomerulopathies (24%), and DKD (20%). Supplemental Table 2 lists clinicopathologic diagnoses for each category, and Supplemental Table 3 shows that patients with DKD had the highest plasma activin A levels.

Table 1.

Baseline characteristics of study participants in the Taipei Renal Transcriptomics and Outcomes Investigation

Parameters	Value (N=339)
Plasma activin A levels, pg/ml	382 (257–526)
Demographic and clinical characteristics
Age, yr	55±17
Male, n (%)	197 (58)
Body mass index, kg/m2	25.6±4.5
Diabetes mellitus, n (%)	98 (29)
Hypertension, n (%)	151 (45)
Prevalent CVDs, n (%)	45 (13)
Laboratory test results
Hemoglobin, g/dl	11.0 (9.3–13.1)
Leukocytes, 103/mm3	6.7 (5.1–8.1)
Albumin, g/dl	3.3 (2.6–3.9)
BUN, mg/dl	30 (18–49)
Creatinine, mg/dl	1.8 (1.0–3.5)
eGFR, ml/min per 1.73 m2	36 (18–67)
Proteinuria, mg/mg creatinine	2.9 (1.0–7.1)
Primary clinicopathologic diagnostic categories, n (%)
Proliferative glomerulonephritis	98 (29)
Nonproliferative glomerulopathies	81 (24)
Diabetic kidney disease	69 (20)
Vascular	34 (10)
Tubulointerstitial	22 (7)
Other	35 (10)
Grades of chronic changes
Global glomerulosclerosis, n (%)
≤10%	102 (30)
11%–25%	84 (25)
26%–50%	97 (29)
>50%	56 (17)
Interstitial fibrosis/tubular atrophy, n (%)
≤10%	93 (27)
11%–25%	70 (21)
26%–50%	119 (35)
>50%	57 (17)
Arterial sclerosis, n (%)
Minimal/mild	158 (47)
Moderate/severe	181 (53)
Arteriolar sclerosis, n (%)
Minimal/mild	241 (71)
Moderate/severe	98 (29)

Categorical variables are expressed as numbers and percentages. Continuous data are presented as mean±SD or median (interquartile range). CVDs, cardiovascular diseases.

We examined the relationships between plasma activin A, demographics, and laboratory parameters, uncovering negative correlations with hemoglobin, eGFR, and serum albumin levels (r_s=−0.38, −0.53, and −0.26, all P < 0.001) and positive correlations with age and proteinuria (r_s=0.37 and 0.42, both P < 0.001).

Associations of Plasma Activin A Level with Global Glomerulosclerosis, IFTA, and Vascular Sclerosis

We investigated the relationship between plasma activin A levels and chronic kidney injuries in biopsy samples. Figure 1 shows a significant positive association between plasma activin A and global glomerulosclerosis, IFTA, and vascular sclerosis. To further validate these findings, we performed multivariable logistic regression analysis, accounting for demographic and laboratory variables. As shown in Table 2, the results demonstrate a significant association between elevated plasma activin A and the presence of these chronic histopathological lesions, particularly in the cases of IFTA (odds ratio, 3.87; 95% confidence interval, 1.55 to 9.70) and arteriolosclerosis (odds ratio, 5.73; 95% confidence interval, 2.72 to 12.09).

Figure 1.

Association between circulating activin A levels and chronic histological lesions. The scatter plots depict the variations in plasma activin A concentrations among different severities of chronic histological lesions, including global glomerulosclerosis (A), interstitial fibrosis and tubular atrophy (IFTA) (B), arteriosclerosis (C), arteriolar sclerosis (D), and Mayo Clinic Chronicity Score (MCCS) (E). The data are presented as mean±SD, and statistical comparisons were performed using the Kruskal–Wallis test followed by post hoc Dunn's analysis.

Table 2.

Association between plasma levels of activin A with each category of chronic histopathologic lesion

Variables	Global GSa	IFTAb	Arteriosclerosisc	Arteriolosclerosisc	MCCS ≥5
	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Plasma activin A, pg/ml
Tertile 1: 220 (164–259)d	Reference	Reference	Reference	Reference	Reference
Tertile 2: 382 (333–414)d	1.57 (0.78 to 3.15)	1.88 (0.82 to 4.31)	2.14 (1.16 to 3.92)	3.19 (1.40 to 7.25)	2.65 (1.20 to 5.88)
Tertile 3: 621 (525–754)d	2.13 (0.90 to 5.07)	4.87 (1.60 to 14.82)	1.87 (0.92 to 3.79)	6.21 (2.54 to 15.14)	5.52 (2.17 to 14.02)
Per 1-Ln higher	2.14 (1.05 to 4.39)	3.87 (1.55 to 9.70)	1.65 (0.91 to 3.00)	5.73 (2.72 to 12.09)	3.61 (1.63 to 8.02)

Logistic regression models were constructed using each chronic histopathological lesion as the dependent variable and the tertiles and Ln-transformed concentrations of plasma activin A as the independent variables. Each individual model was adjusted for age, sex, serum albumin, proteinuria, and eGFR. CI, confidence interval; GS, glomerulosclerosis; IFTA, interstitial fibrosis and tubular atrophy; Ln, natural logarithm; MCCS, Mayo Clinic Chronicity Score; OR, odds ratio.

^aDependent variable is ≥11% of glomeruli affected.

^bDependent variable is ≥11% of cortical volume affected.

^cDependent variable is moderate-to-severe lesion severity.

^dMedian (interquartile range).

Unraveling Activin A Upregulation and Its Major Source in Diseased Kidney

To investigate gene expression in kidney disease, we conducted RNA sequencing on microdissected tubulointerstitial tissue from the first 71 recruited participants. The clinical characteristics of individuals whose tissue underwent RNA-seq are displayed in Supplemental Table 4, as well as those who did not. Using an IFTA threshold of >10% to define kidney fibrosis, we observed a significant increase in INHBA expression in fibrotic kidneys (n=45) compared with nonfibrotic samples (n=26) (Figure 2A). Furthermore, we found that kidney INHBA positively correlated with plasma activin A and IFTA, while negatively correlating with eGFR (Figure 2, B–D). Because activin A belongs to the TGF-β superfamily, its expression may resemble that of TGF-β, which increases in various cells within the fibrotic kidney.³¹ Thus, we performed immunostaining on kidney samples from patients with CKD with diverse etiologies to investigate the cellular source of activin A. We observed significantly stronger activin A staining in tissue sections from patients with fibrotic kidney diseases compared with normal kidney tissue (Figure 2, E–H). Notably, activin A was predominantly localized in the peritubular interstitium and some tubular cells. Our double-immunofluorescence staining revealed that most of the activin A–positive interstitial cells were positive for α-SMA, a marker for activated fibroblasts (Figure 2, I–L). To precisely pinpoint the cellular origin of activin A within the kidney, we used a single-nucleus RNA-Seq dataset for DKD,³² and we noted an upregulation of INHBA expression specifically in a subset of fibroblasts and parietal epithelial cells that were present in diseased kidneys (Supplemental Figure 2).

Figure 2.

Activin A increases in tubular epithelium and interstitial myofibroblasts in human CKD. (A) The volcano plot illustrates the differentially expressed genes, with INHBA showing upregulation in patients with fibrotic kidney disease compared with those without kidney fibrosis. The x axis represents the Log2 fold change, while the y axis represents −Log10 (P values). Scatter plots accompanied by Spearman's correlation analyses were used to examine the relationship between kidney INHBA expression and plasma concentrations of activin A (B), eGFR (C), and the extent of interstitial fibrosis and tubular atrophy (D). Immunohistochemical staining of activin A was performed on kidney biopsy sections from participants with CKD and healthy controls (E–H). While activin A was undetectable in normal kidneys (E and F), it showed upregulation in individuals with advanced CKD in both tubular cells and peritubular interstitial cells (G and H). The boxed areas were enlarged and are shown on the right (F and H). Arrows indicate activin A–positive interstitial cells. Representative images show Nomarski (I) and triple-labeled immunofluorescence images (J: activin A in red, K: α-SMA in green, DAPI in blue) of fibrotic kidney diseases. Panel L shows activin A and α-SMA colocalization in Nomarski and fluorescence image overlay (original magnifications ×200 [E and G] and ×400 [F, H, and I–L]). DAPI, 4',6-diamidino-2-phenylindole; Ln, natural logarithm; TPM, transcripts per million.

The Association between Plasma Activin A Levels and Kidney Outcomes

During a median follow-up of 22 (15–31) months, 33 individuals were lost, and 113 experienced a composite kidney outcome. Figure 3 presents Kaplan–Meier curves for MAKE on the basis of baseline activin A tertiles, highlighting the significant outcome differences among the tertile groups. Elevated plasma activin A levels were associated with a greater risk of kidney end points in Cox regression analysis (Table 3). This association remained significant after adjusting for clinical and laboratory parameters including eGFR. However, its significance in relation to MAKE diminished when considering the presence of DKD. There was no statistical evidence of interaction between plasma activin A and sex for MAKE (P for interaction: 0.84; Supplemental Table 5). Incorporating death as a competing risk in subdistribution hazards models revealed a significant association between plasma activin A levels and kidney outcomes (Supplemental Table 6).

Figure 3.

The rate of major adverse kidney events stratified by plasma activin A tertiles. The cumulative incidence of MAKE in kidney biopsy patients, categorized by plasma activin A tertiles, showed 2-year MAKE rates of 7% (tertile 1), 23% (tertile 2), and 51% (tertile 3). Significant differences in kidney outcomes were observed. MAKE was defined as a >50% decline in eGFR, kidney failure, or death, with activin A tertiles at 220 (164–259), 382 (333–414), and 621 (525–754) pg/ml. Survival analysis used the log-rank test. MAKE, major adverse kidney events.

Table 3.

Associations of plasma activin A with major adverse kidney events

Variables	Model 1	Model 2	Model 3
	HR (95% CI)	HR (95% CI)	HR (95% CI)
Plasma activin A, pg/ml
Tertile 1: 220 (164–259)a	Reference	Reference	Reference
Tertile 2: 382 (333–414)a	3.17 (1.61 to 6.26)	1.73 (0.86 to 3.52)	1.48 (0.72 to 3.02)
Tertile 3: 621 (525–754)a	8.38 (4.43 to 15.85)	3.10 (1.55 to 6.21)	2.05 (0.98 to 4.27)
Per 1-Ln increase	5.06 (3.46 to 7.40)	2.27 (1.40 to 3.68)	1.64 (0.99 to 2.73)

Model 1 is unadjusted. Model 2 is adjusted for age, sex, serum albumin, proteinuria, and baseline eGFR. Model 3 is Model 2 and further adjusted for primary clinicopathologic diagnosis. CI, confidence interval; HR, hazard ratio; Ln, natural logarithm.

^aMedian (interquartile range).

Activin A Regulates Fibroblasts Activation and Extracellular Matrix Production in CKD

To investigate the role of activin A in kidney fibroblast activation, we conducted in vitro experiments with NRK49F cells. As illustrated in Supplemental Figure 3A, Col3a1 expression was upregulated in a dose-dependent manner after exogenous activin A administration. On stimulation with the key fibroblast activator, TGF-β, we observed a significant increase in the expression of Inhba as well as key myofibroblast marker genes, including Acta2, Ccn2, and Col3a1, in NRK49F cells (Figure 4). Notably, targeted siRNA transfection designed to silence Inhba effectively suppressed inhibin β-A expression and mitigated fibrotic alterations. Treating Inhba siRNA-transfected NRK-49F cells with recombinant activin A partially increased Col3a1 levels but did not fully restore the antifibrotic effects of Inhba silencing. In addition, we observed a significant increase in mRNA expression of follistatin, a natural activin A antagonist, in fibroblast groups exposed to exogenous TGF-β (Supplemental Figure 3, B–C).

Figure 4.

Activin A–mediated TGF-β–induced activation of kidney fibroblasts and extracellular matrix formation. Treatment with TGF-β leads to a significant increase in Inhba levels in kidney fibroblasts. However, by using siRNA, we were able to successfully inhibit the expression of the Inhba gene, reducing it to nearly its original baseline level (A). The effect of suppressing the Inhba gene on the TGF-β–induced expression of myofibroblast marker genes, such as Acta2, Ccn2, and extracellular matrix marker Col3a1, was examined (B). Representative Western blotting images were used to illustrate the protein levels of Inhibin βA, collagen 3a1, and α-SMA in different experimental groups (C), and the densitometric analysis of the Western blot results was conducted (D).

Discussion

In this study, we found a strong correlation between plasma activin A levels and chronic kidney histopathological changes, particularly IFTA and arteriolosclerosis. In addition, individuals with increased plasma activin A were at a higher risk of MAKE. Our investigation revealed increased activin A expression in the kidneys of patients with CKD, mainly in interstitial myofibroblasts and tubular epithelial cells. Through in vitro studies, we also demonstrated that activin A serves as a critical mediator of fibroblast activation in kidney fibrosis.

Human activin A is a dimeric protein linked by disulfide bonds, ranging in molecular weight from 24 to 110 kDa because of post-translational processing.³³ Consequently, plasma activin A levels may accumulate as GFR declines. However, impaired activin A filtration may not be the only factor affecting its plasma concentrations in CKD. Our results demonstrated a strong correlation between plasma activin A concentrations and kidney INHBA expression. Furthermore, a previous study reported elevated plasma activin A levels in the unilateral ureteral obstruction model, but not in the unilateral nephrectomy animal model.³⁴ In preclinical kidney damage studies, elevated activin A was detected specifically in the kidney and serum, but not in other organs.^23,35 Collectively, these findings strongly suggest that the injured kidney serves as the primary source of circulating activin A in CKD. Our immunohistochemical staining and single-cell RNA analysis provided additional insights into the localization of activin A within the fibrotic kidney. We observed increased activin A protein in myofibroblasts and tubular cells in fibrotic kidneys, while mRNA increases were restricted to myofibroblasts, suggesting potential paracrine effects on neighboring tubular cells.

Several clinical studies evaluated the relationship between activin A and kidney function in CKD. Nevertheless, most of these investigations are cross-sectional in nature and suffer from a lack of kidney histological data. In this study, patients with biopsy-proven kidney diseases showed a correlation between plasma activin A and both kidney fibrosis and MAKE. Our findings revealed higher plasma activin A concentrations as being associated with higher risks of poor outcomes, even after adjusting for demographics and laboratory variables. However, this association lost significance after adjusting for DKD, suggesting that activin A might serve as an intermediary factor linking underlying kidney disease to adverse outcomes. This inference requires validation through randomized controlled trials. Moreover, our study involves patients with advanced CKD, where eGFR decline parallels kidney reserve loss, possibly limiting the utility of activin A in risk assessment. Future studies with individuals having preserved kidney function are needed to assess whether plasma activin A could surpass serum creatinine as an early kidney damage marker.

From our study, we propose that plasma activin A reflects kidney myofibroblast abundance and activity. Activin A promotes fibroblast differentiation and extracellular matrix production, suggesting it as a target to combat kidney fibrosis.²⁶ In our data, exogenous activin A promoted fibroblast activation, while Inhba knockdown attenuated TGF-β–induced fibroblast activation, implying a regulatory role for activin A. However, exogenous activin A was unable to reverse the effects of Inhba knockdown, highlighting differences between exogenous and endogenous activin A in influencing fibroblast activation. This may be due to increased follistatin levels in TGF-β–treated fibroblasts, blocking cell surface activin A access to its receptor and potentially serving as a negative feedback mechanism.^36–38

Many compounds aimed at mitigating the effects of activin A have been investigated for clinical use, and one such example is RAP-011, which has shown promise in reducing fibrosis in animal models of kidney injury.^27,39,40 In addition, we found a significant association between activin A and vascular damage because it was detected in arteriosclerotic injuries and may contribute to various vascular pathologies, including atherogenic plaque stabilization and vascular calcification in chronic kidney disease–mineral and bone disorder.^41–43 This suggests that activin A may initiate kidney fibrosis by affecting vascular health, although further research is needed to understand the mechanisms involved.

This study has several limitations worth noting. First, its observational design prevents establishing a direct causal relationship between baseline plasma activin A levels and kidney outcomes. Clinical trials are necessary to explore the potential benefits of inhibiting activin A in kidney disease progression. Second, despite adjusting for multiple variables in the Cox proportional hazard model, the possibility of residual confounding factors influencing the observed effects remains. Third, we did not assess urinary activin A concentrations in this study, which might have shown a stronger correlation with histology and disease progression.¹⁴ Fourth, while we used NRK49F cells in our in vitro experiment, using human primary kidney fibroblasts could provide more robust data, given their relevance to human biopsy information. Finally, plasma activin A levels from nonkidney organs may misinterpret CKD severity.⁴⁴ Thus, it is essential to take into account the clinical context when using activin A as a kidney biomarker.

In summary, our findings suggest that activin A could serve as a potential biomarker for assessing CKD severity and as a target to slow kidney fibrosis progression.

Supplementary Material

cjasn-19-169-s001.pdf ^{(566.9KB, pdf)}

Disclosures

K.-H. Lee reports employment with and research funding from Taipei Veterans General Hospital. S.-y. Li and S.-M. Ou report employment with Taipei Veterans General Hospital. C.-C. Lin reports employment with Taipei Veterans General Hospital, consultancy from WS Far Infrared Company, and honoraria from AstraZeneca, Baxter, and Boehringer Ingelheim. M.-T. Tsai reports employment with Taipei Veterans General Hospital, Taipei, Taiwan; was invited by the Genomics company to deliver a speech on Olink PEA technology in September 2023; and currently serves as the Deputy Chairperson of the Academy and Research Committee for the Taiwan Society of Nephrology.

Funding

This work is supported by Ministry of Science and Technology from 112-2628-B-075 -003 (M.-T. Tsai), 110-2314-B-075 -028 -MY3, 110-2314-038-125-MY3 (S.Y. Li), and Taipei Veterans General Hospital from V112C-063 (M.-T. Tsai), V108D42-003-MY3-3 (S.Y. Li).

Author Contributions

Conceptualization: Szu-yuan Li, Ming-Tsun Tsai.

Data curation: Kuo-Hua Lee, Szu-yuan Li, Ming-Tsun Tsai, Shuo-Ming Ou.

Formal analysis: Szu-yuan Li, Ming-Tsun Tsai, Shuo-Ming Ou.

Funding acquisition: Szu-yuan Li, Ming-Tsun Tsai.

Investigation: Kuo-Hua Lee, Szu-yuan Li, Chih-Ching Lin, Ming-Tsun Tsai.

Methodology: Kuo-Hua Lee, Szu-yuan Li, Chih-Ching Lin, Ming-Tsun Tsai, Shuo-Ming Ou.

Project administration: Chih-Ching Lin.

Resources: Szu-yuan Li.

Software: Szu-yuan Li, Shuo-Ming Ou.

Supervision: Szu-yuan Li.

Validation: Szu-yuan Li, Shuo-Ming Ou.

Visualization: Kuo-Hua Lee, Chih-Ching Lin.

Writing – original draft: Ming-Tsun Tsai.

Writing – review & editing: Szu-yuan Li.

Data Sharing Statement

The data for this study can be obtained by contacting the corresponding author.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/CJN/B837.

Supplemental Table 1. PCR primers used in this study.

Supplemental Table 2. Primary clinicopathologic diagnoses of the study population.

Supplemental Table 3. Kidney function, proteinuria, and plasma activin A levels by primary clinicopathologic diagnosis.

Supplemental Table 4. Comparing the patient characteristics of those with and without kidney tissue RNA seq.

Supplemental Table 5. Associations of plasma activin A with major adverse kidney events by sex.

Supplemental Table 6. Competing risk analysis of plasma activin A and kidney outcomes.

Supplemental Figure 1. Flow diagram of study participants.

Supplemental Figure 2. INHBA expression localization was studied in kidney cortex samples from control and diabetic kidney patients with disease.

Supplemental Figure 3. The effect of exogenous activin A on TGF-β–mediated fibroblast activation in Inhba siRNA transfected NRK49F cells.