Background: Renal glycogen synthase kinase-3 beta (GSK3β) overactivity has been associated with a diverse range of kidney diseases. GSK3β activity in urinary exfoliated cells was reported to predict the progression of diabetic kidney disease (DKD). We compared the prognostic value of urinary and intrarenal GSK3β levels in DKD and nondiabetic chronic kidney disease (CKD). Methods: We recruited 118 consecutive biopsy-proved DKD patients and 115 nondiabetic CKD patients. Their urinary and intrarenal GSK3β levels were measured. They were then followed for dialysis-free survival and rate of renal function decline. Results: DKD group had higher intrarenal and urinary GSK3β levels than nondiabetic CKD (p < 0.0001 for both), but their urinary GSK3β mRNA levels were similar. Urinary p-GSK3β level is statistically significantly correlated with the baseline estimated glomerular filtration rate (eGFR), but urinary GSK3β level by ELISA, its mRNA level, the p-GSK3β level, or the p-GSK3β/GSK3β ratio had no association with dialysis-free survival or the slope of eGFR decline. In contrast, the intrarenal pY216-GSK3β/total GSK3β ratio significantly correlated with the slope of eGFR decline (r = −0.335, p = 0.006) and remained an independent predictor after adjusting for other clinical factors. Conclusion: Intrarenal and urinary GSK3β levels were increased in DKD. The intrarenal pY216-GSK3β/total GSK3β ratio was associated with the rate of progression of DKD. The pathophysiological roles of GSK3β in kidney diseases deserve further studies.

Chronic kidney disease (CKD) is an important and costly noncommunicable disease [1]. Although the kidney consists of multiple cell types that are arranged in a delicate 3-dimensional architecture, podocyte has long been implicated as the primary focus of many kidney diseases [2, 3], and markers of podocyte injury have been extensively studied as biomarkers of kidney diseases [4‒6]. However, renal tubular cell is the most abundant cell type in the kidney; proximal tubular cells are particularly active metabolically and play a pivotal role in the progression of many kidney diseases [7, 8]. Recently, glycogen synthase kinase-3 beta (GSK3β) has emerged as a valuable marker of both glomerular and tubular injury.

Glycogen synthase kinase-3 is a highly conserved, ubiquitously expressed serine/threonine protein kinase, which was originally identified as a key transducer of the insulin signaling cascade that regulates glycogenesis [9, 10]. The enzyme has two highly homologous isoforms, alpha and beta, of which GSK3β is predominantly expressed both in the glomeruli and proximal tubular cells of the renal cortex [11, 12]. Renal GSK3β overactivity has been associated with a diverse range of kidney diseases. For example, analysis of publicly available kidney transcriptome database demonstrated that patients with progressive CKD exhibited GSK3β overexpression in the renal tubulointerstitial space [13]. Treatment of renal tubular epithelial cells in vitro by transforming growth factor beta-1 (TGF-β1) augmented GSK3β expression and led to cellular dedifferentiation, accumulation of extracellular matrix, and overproduction of other profibrotic cytokines [13]. In a mouse model of progressive CKD, GSK3β ablation in renal tubules mitigated the profibrogenic plasticity of renal tubular epithelial cells, with attenuation of interstitial fibrosis and tubular atrophy [13]. Patients with diabetic kidney disease (DKD) had an increased expression of GSK3β in glomeruli and renal tubules, and the magnitude of increase correlated with the severity of DKD [14]. More importantly, the activity of GSK3β in urinary exfoliated cells was an independent risk factor for the progression of renal impairment in diabetic patients [14], suggesting that GSK3β level in urinary sediment is a promising biomarker for the progression of DKD. In the present study, we compared the prognostic value of urinary and intrarenal GSK3β levels in DKD and nondiabetic CKD.

The study was approved by the Joint Chinese University of Hong Kong – New Territories East Cluster Clinical Research Ethics Committee (approval number CRE-2019.283). All study procedures were in compliance with the Declaration of Helsinki, and all subjects provided written informed consent. We recruited 118 consecutive patients who had kidney biopsy-proved DKD (the DKD group) and 115 nondiabetic patients with CKDs (73 with typical features of hypertensive nephrosclerosis, 42 with non-specific glomerulosclerosis) as the nondiabetic CKD group. All kidney biopsy specimens were assessed by a single experienced pathologist (F.M.L.). A whole-stream early morning urine was collected on the day of kidney biopsy for gene expression and biochemical studies. We also reviewed the demographic and clinical data including serum creatinine and proteinuria. The estimated glomerular filtration rate (eGFR) was calculated by a standard equation [15].

RNA Extraction

The method of mRNA extraction and quantification in urinary sediment has been described previously [16]. Briefly, urine samples were centrifuged immediately after collection at 3,200 g, for 15 min at 4°C. The supernatant was removed, the pellet suspended in 1.5 mL diethyl pyrocarbonate treated phosphate buffered saline, and then centrifuged at 120,00 g for 5 min at 4°C. The washed pellet was resuspended in lysis buffer (RNeasy®; Qiagen, Germantown, MD, USA) and then purified using an RNeasy mini kit (Qiagen). The cDNA was prepared by the SuperScript™ IV First-Strand Synthesis System (ThermoFisher, Germany).

GSK3β mRNA Expression Level

Quantitation of GSK3β mRNA was performed by the StepOnePlus real-time quantitative polymerase chain reaction system (Applied Biosystems, Foster City, CA, USA), using TaqMan™ Fast Advanced Master Mix (ThermoFisher), with a final volume of 10 μL per reaction. Commercially available Taqman primers and probes, including two unlabeled PCR primers and one FAM dye-labeled TaqMan MGB probe, were used for GSK3β and β-actin genes (all from ThermoFisher). Each sample was run in triplicate. Results were analyzed with the use of Sequence Detection software, version 1.9 (Applied Biosystems). Gene expression for each signal was calculated by using the difference-in-threshold-cycle procedure. For the quantification of the target mRNA abundance, differences of threshold cycles between target genes were calculated by 2−ΔΔCT method.

Urine GSK3β and p-GSK3β Level

Urine GSK3β levels were determined by a commercially available ELISA kit (Bioassay Technology Laboratory, China) following the manufacturer’s protocol. The urinary levels were adjusted to urinary creatinine level, which was measured by a creatinine colorimetric assay kit (Millipore Sigma, USA). Each sample was measured in duplicates. An ELISA kit for phosphorylated GSK3β (p-GSK3β) (Cell Signaling Technology, Danvers, MA, USA) was used to quantify the phosphoprotein present. In this part of the experiment, lysates were processed and the assays performed according to manufacturer’s instructions. Bicinchoninic acid assay (BCA) was performed on each lysate, and the lysates were diluted such that 20 μg of protein lysate was used in each ELISA assay. The amount of phosphoprotein in ng per 20 μg of total protein lysate was then determined by comparing the measured 450 nm absorbance of the sample to the standard curve.

Intrarenal GSK3β and pY216-GSK3β Level

Intrarenal GSK3β levels were determined by Western blot study with β-actin level as the reference. The protein electrophoresis, transfer apparatus, and acrylamide gel were obtained from Bio-Rad (Hercules, CA, USA). Total protein from the kidney biopsy specimen was washed with radioimmunoprecipitation assay buffer containing protease inhibitor cocktail. The protein concentrations were quantified by a BCA Protein Assay Kit (Beyotime Institute of Biotechnology, Shanghai, China). Individual proteins were then separated from 20 μg of total protein extract by acrylamide gel electrophoresis in Mini-PROTEAN® Cell (Hercules Bio-Rad, CA, USA) transferred to Hybond-P PVDF membrane and then probed with primary antibodies against GSK3β (1:1,000, Abcam, Cambridge, UK), pY216-GSK3β (1:1,000, Abcam, Cambridge, UK), and β-actin (1:1,000, Abcam, Cambridge, UK). The corresponding secondary antibody was obtained from Abcam. The membrane was exposed to Amersham Hyperfilm Blue. The areas of the bands were estimated by the software Image J. The protein expression level of a sample was calculated by dividing the area of the protein interested by the area of β-actin of the same sample.

Morphometric Study of Kidney Biopsy

The method of morphometry study of renal scarring has been described in previous studies [17, 18]. Briefly, Jones’ silver staining was performed on 5 μm thick sections of renal biopsy specimen. Semi-quantitative computerized image analysis was performed with the Leica Twin Pro image analysis system (Leica Microsystems, Wetzlar, Germany), which was connected to a Leica DC500 digital camera on a Leica DMRXA2 microscope with a ×40 objective (final calibration: 0.258 mm/pixel). Image analysis was performed by MetaMorph 4.0 image-analyzing software (Universal Imaging Corporation™, Downingtown, PA, USA). Ten glomeruli and 10 randomly selected areas were assessed in each patient, and the average percentage of scarred glomerular and tubulointerstitial areas, as represented by the area with positive silver staining, were computed.

Outcome Measures

All patients were followed for at least 12 months. All patients were followed by nephrologists, and their treatment was not affected by the study. Kidney function was monitored every 3 months. The primary outcome measures are dialysis-free survival and renal event-free survival. Renal event was defined as death from any cause, need for dialysis, or ≥40% decline in eGFR as compared to baseline. Secondary outcome measure includes the rate of eGFR decline, which was calculated by the least-squares regression method.

Statistical Analysis

Statistical analysis was performed by SPSS for Windows software version 17.0 (SPSS, Chicago, IL, USA). All the results were presented in mean ± SD for data normally distributed and median (lower and upper quartiles) for the others. Since data on gene expression levels were highly skewed, we used the Mann-Whitney U test to compare gene expression levels between groups and Spearman’s rank-order correlations to test associations between gene expression levels and other parameters. Data were further analyzed with log rank test with dialysis-free survival and renal event-free survival. In addition to GSK3β levels, the multivariate Cox regression model was constructed by age, sex, diagnosis group (DKD vs. CTL), baseline eGFR, and proteinuria. In view of the univariate analysis result, a multivariate linear regression model was constructed to adjust for sex, age, baseline eGFR, proteinuria, the severity of glomerulosclerosis and tubulointerstitial fibrosis, and the intrarenal pY216-GSK3β/GSK3β ratio to determine the independent predictor of the rate of renal function decline. A p value of below 0.05 was considered statistically significant. All probabilities were two tailed.

A total of 233 patients were enrolled, including 118 patients with biopsy-proved DKD and 115 nondiabetic CKD. Their baseline clinical, biochemical, and pathological information is summarized in Table 1. In essence, the DKD group was older and had more proteinuria than the control group, but there was no significant difference in any histological parameter between the groups.

Table 1.

Baseline demographic and clinical data

DKDNondiabetic CKDp value
Patients, n 118 115  
Sex (M:F) 80:38 56:59 0.041a 
Age, years 61.3±12.3 55.4±15.1 0.031b 
Serum creatinine, μmol/L 217.9±162.2 273.3±228.5 0.036b 
eGFR, mL/min/1.73 m2 41.4±31.3 38.0±32.0 0.528b 
Proteinuria, g/day 3.2±2.4 2.6±2.1 0.007b 
Histological damage (%) 
 Glomerulosclerosis 32.7±21.7 25.8±22.5 0.144b 
 Tubulointerstitial fibrosis 30.2±17.4 23.8±22.8 0.584b 
DKDNondiabetic CKDp value
Patients, n 118 115  
Sex (M:F) 80:38 56:59 0.041a 
Age, years 61.3±12.3 55.4±15.1 0.031b 
Serum creatinine, μmol/L 217.9±162.2 273.3±228.5 0.036b 
eGFR, mL/min/1.73 m2 41.4±31.3 38.0±32.0 0.528b 
Proteinuria, g/day 3.2±2.4 2.6±2.1 0.007b 
Histological damage (%) 
 Glomerulosclerosis 32.7±21.7 25.8±22.5 0.144b 
 Tubulointerstitial fibrosis 30.2±17.4 23.8±22.8 0.584b 

DKD, diabetic kidney disease; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate.

Data were compared by aχ2 test and bStudent’s t test.

Urinary and Intrarenal GSK3β Levels

The urinary GSK3β level by ELISA, its corresponding urinary mRNA level, and intrarenal level by Western blotting are summarized and compared between DKD and nondiabetic CKD groups in Figure 1a–c, respectively. The urinary p-GSK3β level by ELISA, intrarenal p-GSK3β level by Western blotting, and the intrarenal pY216-GSK3β/total GSK3β ratio by Western blotting are summarized and compared between DKD and nondiabetic CKD groups in Figure 1d–f, respectively. The urinary GSK3β level and intrarenal GSK3β level in the DKD group were significantly higher than those of the nondiabetic CKD group. In contrast, the intrarenal pY216-GSK3β/GSK3β ratio in the nondiabetic CKD group was significantly higher than those of the DKD group. However, there was no significant difference in urinary GSK3β mRNA expression between the two groups. The internal correlation of urinary and intrarenal GSK3β levels is summarized in Table 2. In essence, only urinary GSK3β level by ELISA was significantly negative correlated with intrarenal level by Western blotting.

Fig. 1.

Comparison of GSK3β levels: urinary level by ELISA (a); urinary mRNA level (b); intrarenal level by Western blotting (c); urinary p-GSK3β level by ELISA (d); intrarenal p-GSK3β level by Western blotting (e); and intrarenal pY216-GSK3β/total GSK3β ratio by Western blotting (f), between diabetic kidney disease (DKD) and nondiabetic chronic kidney disease (CKD) groups. Error bars denote standard deviations. Data were compared by Mann-Whitney U test.

Fig. 1.

Comparison of GSK3β levels: urinary level by ELISA (a); urinary mRNA level (b); intrarenal level by Western blotting (c); urinary p-GSK3β level by ELISA (d); intrarenal p-GSK3β level by Western blotting (e); and intrarenal pY216-GSK3β/total GSK3β ratio by Western blotting (f), between diabetic kidney disease (DKD) and nondiabetic chronic kidney disease (CKD) groups. Error bars denote standard deviations. Data were compared by Mann-Whitney U test.

Close modal
Table 2.

Internal correlation of urinary and intrarenal GSK3β levels

Urinary level by ELISAIntrarenal level
Urinary level by ELISA 
 Entire cohort r = −0.075, p = 0.475 
 DKD group r = −0.247, p = 0.027 
 Nondiabetic CKD group r = 0.115, p = 0.721 
Urinary mRNA level 
 Entire cohort r = 0.110, p = 0.106 r = 0.034, p = 0.746 
 DKD group r = 0.060, p = 0.521 r = −0.061, p = 0.588 
 Nondiabetic CKD group r = −0.105, p = 879 r = −0.025, p = 0.938 
Urinary level by ELISAIntrarenal level
Urinary level by ELISA 
 Entire cohort r = −0.075, p = 0.475 
 DKD group r = −0.247, p = 0.027 
 Nondiabetic CKD group r = 0.115, p = 0.721 
Urinary mRNA level 
 Entire cohort r = 0.110, p = 0.106 r = 0.034, p = 0.746 
 DKD group r = 0.060, p = 0.521 r = −0.061, p = 0.588 
 Nondiabetic CKD group r = −0.105, p = 879 r = −0.025, p = 0.938 

DKD, diabetic kidney disease; CKD, chronic kidney disease.

Person’s correlation coefficients are depicted.

Relation with Clinical and Histological Parameters

The correlation between urinary GSK3β level by ELISA, its corresponding urinary mRNA level, its p-GSK3β level by ELISA, or its ratio and clinical and histological parameters is summarized in Table 3. The urinary p-GSK3β level is statistically significantly correlated with the baseline eGFR. There was a significant but only modest correlation between baseline eGFR and GSK3β mRNA expression in DKD patients. The urinary p-GSK3β/GSK3β ratio was not correlated with any clinical or histological parameters in any group.

Table 3.

Relation between GSK3β levels and clinicopathological parameters

eGFRProteinuriaGlomerulosclerosisTubulointerstitial fibrosis
(A) Urinary levels 
Urinary level by ELISA 
 Entire cohort r = −0.057, p = 0.406 r = 0.020, p = 0.775 r = 0.069, p = 0.434 r = 0.075, p = 0.390 
 DKD group r = 0.013, p = 0.893 r = 0.058, p = 0.546 r = −0.160, p = 0.085 r = −0.169, p = 0.069 
 Nondiabetic CKD group r = −0.010, p = 0.922 r = 0.015, p = 0.885 r = −0.293, p = 0.122 r = −0.234, p = 0.213 
Urinary mRNA level 
 Entire cohort r = 0.096, p = 0.149 r = 0.063, p = 0.357 r = 0.067, p = 0.435 r = 0.018, p = 0.836 
 DKD group r = 0.230, p = 0.012 r = 0.057, p = 0.552 r = −0.080, p = 0.393 r = −0.024, p = 0.798 
 Nondiabetic CKD group r = 0.140, p = 0.155 r = 0.129, p = 0.201 r = −0.238, p = 0.213 r = −0.223, p = 0.236 
Urinary p-GSK3β level 
 Entire cohort r = −0.054, p = 0.447 r = −0.053, p = 0.464 r = −0.029, p = 0.745 r = −0.034, p = 0.699 
 DKD group r = 0.325, p = 0.001 r = −0.100, p = 0.323 r = −0.016, p = 0.869 r = −0.021, p = 0.259 
 Nondiabetic CKD group r = −0.109, p = 0.292 r = −0.071, p = 0.500 r = 0.159, p = 0.437 r = −0.047, p = 0.818 
p-GSK3β/total GSK3β ratio 
 Entire cohort r = −0.006, p = 0.932 r = −0.060, p = 0.415 r = −0.035, p = 0.692 r = −0.051, p = 0.562 
 DKD group r = 0.049, p = 0.620 r = −0.040, p = 0.693 r = −0.010, p = 0.917 r = 0.015, p = 0.881 
 Nondiabetic CKD group r = −0.066, p = 0.532 r = −0.002, p = 0.983 r = 0.035, p = 0.865 r = −0.096, p = 0.633 
(B) intrarenal levels 
Total intrarenal GSK3β 
 Entire cohort r = 0.146, p = 0.164 r = 0.052, p = 0.631 r = 0.187, p = 0.086 r = 0.152, p = 0.165 
 DKD group r = 0.100, p = 0.375 r = 0.016, p = 0.893 r = 0.012, p = 0.913 r = −0.060, p = 0.596 
 Nondiabetic CKD group r = −0.231, p = 0.470 r = 0.323, p = 0.332 r = −0.352, p = 0.562 r = −0.448, p = 0.450 
Intrarenal pY216-GSK3β level 
 Entire cohort r = 0.266, p = 0.013 r = −0.219, p = 0.053 r = −0.046, p = 0.681 r = −0.137, p = 0.220 
 DKD group r = 0.259, p = 0.020 r = −0.201, p = 0.085 r = −0.064, p = 0.575 r = −0.166, p = 0.144 
 Nondiabetic CKD group r = 0.384, p = 0.453 r = −0.789, p = 0.113 r = 0.918, p = 0.260 r = 0.858, p = 0.344 
pY216-GSK3β/total GSK3β ratio 
 Entire cohort r = −0.124, p = 0.255 r = −0.188, p = 0.097 r = −0.140, p = 0.210 r = −0.144, p = 0.197 
 DKD group r = 0.303, p = 0.006 r = −0.218, p = 0.062 r = −0.053, p = 0.642 r = −0.149, p = 0.190 
 Nondiabetic CKD group r = 0.575, p = 0.232 r = −0.693 p = 0.195 r = −0.352, p = 0.562 r = 0.866, p = 0.333 
eGFRProteinuriaGlomerulosclerosisTubulointerstitial fibrosis
(A) Urinary levels 
Urinary level by ELISA 
 Entire cohort r = −0.057, p = 0.406 r = 0.020, p = 0.775 r = 0.069, p = 0.434 r = 0.075, p = 0.390 
 DKD group r = 0.013, p = 0.893 r = 0.058, p = 0.546 r = −0.160, p = 0.085 r = −0.169, p = 0.069 
 Nondiabetic CKD group r = −0.010, p = 0.922 r = 0.015, p = 0.885 r = −0.293, p = 0.122 r = −0.234, p = 0.213 
Urinary mRNA level 
 Entire cohort r = 0.096, p = 0.149 r = 0.063, p = 0.357 r = 0.067, p = 0.435 r = 0.018, p = 0.836 
 DKD group r = 0.230, p = 0.012 r = 0.057, p = 0.552 r = −0.080, p = 0.393 r = −0.024, p = 0.798 
 Nondiabetic CKD group r = 0.140, p = 0.155 r = 0.129, p = 0.201 r = −0.238, p = 0.213 r = −0.223, p = 0.236 
Urinary p-GSK3β level 
 Entire cohort r = −0.054, p = 0.447 r = −0.053, p = 0.464 r = −0.029, p = 0.745 r = −0.034, p = 0.699 
 DKD group r = 0.325, p = 0.001 r = −0.100, p = 0.323 r = −0.016, p = 0.869 r = −0.021, p = 0.259 
 Nondiabetic CKD group r = −0.109, p = 0.292 r = −0.071, p = 0.500 r = 0.159, p = 0.437 r = −0.047, p = 0.818 
p-GSK3β/total GSK3β ratio 
 Entire cohort r = −0.006, p = 0.932 r = −0.060, p = 0.415 r = −0.035, p = 0.692 r = −0.051, p = 0.562 
 DKD group r = 0.049, p = 0.620 r = −0.040, p = 0.693 r = −0.010, p = 0.917 r = 0.015, p = 0.881 
 Nondiabetic CKD group r = −0.066, p = 0.532 r = −0.002, p = 0.983 r = 0.035, p = 0.865 r = −0.096, p = 0.633 
(B) intrarenal levels 
Total intrarenal GSK3β 
 Entire cohort r = 0.146, p = 0.164 r = 0.052, p = 0.631 r = 0.187, p = 0.086 r = 0.152, p = 0.165 
 DKD group r = 0.100, p = 0.375 r = 0.016, p = 0.893 r = 0.012, p = 0.913 r = −0.060, p = 0.596 
 Nondiabetic CKD group r = −0.231, p = 0.470 r = 0.323, p = 0.332 r = −0.352, p = 0.562 r = −0.448, p = 0.450 
Intrarenal pY216-GSK3β level 
 Entire cohort r = 0.266, p = 0.013 r = −0.219, p = 0.053 r = −0.046, p = 0.681 r = −0.137, p = 0.220 
 DKD group r = 0.259, p = 0.020 r = −0.201, p = 0.085 r = −0.064, p = 0.575 r = −0.166, p = 0.144 
 Nondiabetic CKD group r = 0.384, p = 0.453 r = −0.789, p = 0.113 r = 0.918, p = 0.260 r = 0.858, p = 0.344 
pY216-GSK3β/total GSK3β ratio 
 Entire cohort r = −0.124, p = 0.255 r = −0.188, p = 0.097 r = −0.140, p = 0.210 r = −0.144, p = 0.197 
 DKD group r = 0.303, p = 0.006 r = −0.218, p = 0.062 r = −0.053, p = 0.642 r = −0.149, p = 0.190 
 Nondiabetic CKD group r = 0.575, p = 0.232 r = −0.693 p = 0.195 r = −0.352, p = 0.562 r = 0.866, p = 0.333 

DKD, diabetic kidney disease; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate.

Person’s correlation coefficients are depicted.

The correlation between intrarenal total GSK3β, pY216-GSK3β level, and their ratio by Western blotting with clinical parameters is summarized in Table 3. In fact, there was a significant but only modest correlation between baseline eGFR and intrarenal pY216-GSK3β level in DKD patients. The pY216-GSK3β/total GSK3β ratio significantly correlated with the baseline eGFR in the DKD group. In contrast, there was no correlation between the intrarenal total GSK3β or pY216-GSK3β and clinical parameters in the nondiabetic CKD group.

Relation with Clinical Outcome

The entire cohort was followed for an average 30.2 ± 31.2 months. During this period, none of the patients died; 106 patients progressed to dialysis-dependent kidney failure, and another 44 patients had 40% decline in GFR. The average rate of eGFR decline was −9.7 ± 24.2 mL/min/1.73 m2 per year. The relation between urinary and intrarenal GSK3β levels and clinical outcome by univariate analysis is summarized in Table 4. In essence, urinary GSK3β level by ELISA, its mRNA level, the p-GSK3β level, or the p-GSK3β/GSK3β ratio had no association with dialysis-free survival or the slope of eGFR decline in DKD or nondiabetic CKD groups.

Table 4.

Relation between GSK3β Levels and clinic outcome

Dialysis-free survivalaRenal event-free survivalaSlope of eGFR declineb
(A) Urinary levels 
Urinary level by ELISA 
 Entire cohort 1.00 (0.99–1.02), p = 0.693 1.00 (0.98–1.02), p = 0.849 r = −0.029, p = 0.698 
 DKD group 0.99 (0.98–1.02), p = 0.858 1.00 (0.98–1.02), p = 0.957 r = −0.004, p = 0.970 
 Nondiabetic CKD group 0.94 (0.87–1.01), p = 0.102 0.89 (0.81–0.99), p = 0.031 r = 0.081, p = 0.474 
Urinary mRNA level 
 Entire cohort 1.02 (0.86–1.22), p = 0.808 0.96 (0.80–1.14), p = 0.623 r = −0.011, p = 0.876 
 DKD group 1.06 (0.84–1.33), p = 0.645 0.95 (0.76–1.18), p = 0.625 r = 0.020, p = 0.851 
 Nondiabetic CKD group 0.96 (0.74–1.24), p = 0.754 0.97 (0.73–1.30), p = 0.849 r = −0.074, p = 0.507 
Urinary p-GSK3β level 
 Entire cohort 0.80 (0.52–1.23), p = 0.307 0.92 (0.65–1.32), p = 0.655 r = 0.018, p = 0.816 
 DKD group 0.72 (0.42–1.21), p = 0.212 0.83 (0.53–1.30), p = 0.412 r = 0.020, p = 0.856 
 Nondiabetic CKD group 1.07 (0.58–1.99), p = 0.829 1.02 (0.60–1.73), p = 0.941 r = −0.096, p = 0.372 
p-GSK3β/total GSK3β ratio 
 Entire cohort 0.00 (0.00–33.92), p = 0.278 0.00 (0.00–30.51), p = 0.553 r = 0.062, p = 0.382 
 DKD group 0.00 (0.00–24.18), p = 0.242 0.00 (0.00–26.23), p = 0.399 r = 0.041, p = 0.712 
 Nondiabetic CKD group 369.50 (0.00–62.00), p = 0.735 1,058.93 (0.00–57.61), p = 0.829 r = −0.093, p = 0.401 
(B) Intrarenal levels 
Total GSK3β 
 Entire cohort 1.14 (0.76–1.70), p = 0.534 1.02 (0.695–1.51), p = 0.905 r = −0.220, p = 0.054 
 DKD group 0.86 (0.48–1.54), p = 0.619 0.94 (0.51–1.72), p = 0.850 r = −0.122, p = 0.328 
 Nondiabetic CKD group 0.14 (0.00–6.83), p = 0.139 1.59 (0.21–12.10), p = 0.655 r = 0.135, p = 0.730 
pY216-GSK3β 
 Entire cohort 0.93 (0.67–1.29), p = 0.652 1.24 (0.91–1.68), p = 0.168 r = −0.228, p = 0.058 
 DKD group 0.93 (0.67–1.30), p = 0.670 1.26 (0.92–1.72), p = 0.152 r = −0.221, p = 0.075 
 Nondiabetic CKD group 0.62 (0.04–9.73), p = 0.731 1.14 (0.11–11.70), p = 0.915 r = −0.632, p = 0.368 
pY216-GSK3β/total GSK3β ratio 
 Entire cohort 0.95 (0.51–1.76), p = 0.870 1.21 (0.72–2.02), p = 0.477 r = 0.069, p = 0.568 
 DKD group 0.82 (0.26–1.2.53), p = 0.724 2.47 (0.89–6.82), p = 0.081 r = −0.335, p = 0.006 
 Nondiabetic CKD group 0.64 (0.01–24.17), p = 0.811 1.61 (0.08–33.8), p = 0.759 r = −0.600, p = 0.400 
Dialysis-free survivalaRenal event-free survivalaSlope of eGFR declineb
(A) Urinary levels 
Urinary level by ELISA 
 Entire cohort 1.00 (0.99–1.02), p = 0.693 1.00 (0.98–1.02), p = 0.849 r = −0.029, p = 0.698 
 DKD group 0.99 (0.98–1.02), p = 0.858 1.00 (0.98–1.02), p = 0.957 r = −0.004, p = 0.970 
 Nondiabetic CKD group 0.94 (0.87–1.01), p = 0.102 0.89 (0.81–0.99), p = 0.031 r = 0.081, p = 0.474 
Urinary mRNA level 
 Entire cohort 1.02 (0.86–1.22), p = 0.808 0.96 (0.80–1.14), p = 0.623 r = −0.011, p = 0.876 
 DKD group 1.06 (0.84–1.33), p = 0.645 0.95 (0.76–1.18), p = 0.625 r = 0.020, p = 0.851 
 Nondiabetic CKD group 0.96 (0.74–1.24), p = 0.754 0.97 (0.73–1.30), p = 0.849 r = −0.074, p = 0.507 
Urinary p-GSK3β level 
 Entire cohort 0.80 (0.52–1.23), p = 0.307 0.92 (0.65–1.32), p = 0.655 r = 0.018, p = 0.816 
 DKD group 0.72 (0.42–1.21), p = 0.212 0.83 (0.53–1.30), p = 0.412 r = 0.020, p = 0.856 
 Nondiabetic CKD group 1.07 (0.58–1.99), p = 0.829 1.02 (0.60–1.73), p = 0.941 r = −0.096, p = 0.372 
p-GSK3β/total GSK3β ratio 
 Entire cohort 0.00 (0.00–33.92), p = 0.278 0.00 (0.00–30.51), p = 0.553 r = 0.062, p = 0.382 
 DKD group 0.00 (0.00–24.18), p = 0.242 0.00 (0.00–26.23), p = 0.399 r = 0.041, p = 0.712 
 Nondiabetic CKD group 369.50 (0.00–62.00), p = 0.735 1,058.93 (0.00–57.61), p = 0.829 r = −0.093, p = 0.401 
(B) Intrarenal levels 
Total GSK3β 
 Entire cohort 1.14 (0.76–1.70), p = 0.534 1.02 (0.695–1.51), p = 0.905 r = −0.220, p = 0.054 
 DKD group 0.86 (0.48–1.54), p = 0.619 0.94 (0.51–1.72), p = 0.850 r = −0.122, p = 0.328 
 Nondiabetic CKD group 0.14 (0.00–6.83), p = 0.139 1.59 (0.21–12.10), p = 0.655 r = 0.135, p = 0.730 
pY216-GSK3β 
 Entire cohort 0.93 (0.67–1.29), p = 0.652 1.24 (0.91–1.68), p = 0.168 r = −0.228, p = 0.058 
 DKD group 0.93 (0.67–1.30), p = 0.670 1.26 (0.92–1.72), p = 0.152 r = −0.221, p = 0.075 
 Nondiabetic CKD group 0.62 (0.04–9.73), p = 0.731 1.14 (0.11–11.70), p = 0.915 r = −0.632, p = 0.368 
pY216-GSK3β/total GSK3β ratio 
 Entire cohort 0.95 (0.51–1.76), p = 0.870 1.21 (0.72–2.02), p = 0.477 r = 0.069, p = 0.568 
 DKD group 0.82 (0.26–1.2.53), p = 0.724 2.47 (0.89–6.82), p = 0.081 r = −0.335, p = 0.006 
 Nondiabetic CKD group 0.64 (0.01–24.17), p = 0.811 1.61 (0.08–33.8), p = 0.759 r = −0.600, p = 0.400 

DKD, diabetic kidney disease; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate.

Results are depicted as aunadjusted hazard ratio (95% confidence interval) and bPearson’s correlation coefficient.

For intrarenal levels, the total GSK3β, pY216-GSK3β, or their ratio do not predict dialysis-free survival or renal event-free dialysis survival in the DKD group. Intrarenal Y216-GSK3β had a modest but insignificant correlation with the rate of eGFR decline. The pY216-GSK3β/total GSK3β ratio significantly correlated with the slope of eGFR decline (r = −0.335, p = 0.006). After adjusting for other clinical parameters, the intrarenal pY216-GSK3β/total GSK3β ratio remained an independent predictor of the slope of eGFR decline (unstandardized B −24.1, 95% confidence interval −9.4 to −38.8, p = 0.002). In contrast, intrarenal total GSK3β, pY216-GSK3β level, or their ratio were not associated with dialysis-free survival or the slope of eGFR decline in nondiabetic CKD patients.

GSK3β, a highly conserved, redox-sensitive serine/threonine protein kinase, is centrally involved in insulin signaling as well as a number of other cellular signal pathways that are crucial for pathophysiological processes related to kidney injury, repair, and regeneration [19]. Notably, GSK3β is a critical regulator of cellular function in podocytes [20]. The role of GSK3β in DKD is an active area of research. Existing evidence suggests that GSK3β dysregulation is related to type 2 diabetes mellitus. GSK3β overexpression and hyperactivity were reported in the muscles of diabetic patients and in the adipose tissues of obese diabetic mice [21‒23]. In the kidney, GSK3β mediates podocyte autonomous injury by integrating multiple podocytopathic signaling pathways [24, 25]. Activation of GSK3β by sodium nitroprusside ameliorates diabetes-induced kidney injury in diabetic mice [26]. More recently, GSK3β activation in the kidney and in urinary exfoliated cells was reported to be associated with the development or progression of DKD [14].

The results of our present study are in line with but slightly different from the report of Liang et al. [14]. Our study showed that urinary and intrarenal GSK3β levels are specifically elevated in DKD as compared to nondiabetic CKD, and urinary sediment GSK3β mRNA level had a modest but significant correlation with eGFR in DKD but not nondiabetic CKD, both of which were consistent with the findings of the previous study [14]. In our present study, neither urinary GSK3β level or its mRNA expression was associated with the rate of progression of DKD, while Liang et al. [14] showed that urinary phosphorylated GSK3β (p-GSK3β) predicted DKD progression [14]. In this study, we also found that urinary sediment p-GSK3β level correlated with the baseline eGFR. More importantly, we found that the intrarenal pY216-GSK3β/total GSK3β ratio level was associated with the slope of eGFR decline in DKD patients. The pathophysiological role of pY216-GSK3β in the progression of DKD deserves further study.

An important edge of our study was the inclusion of a large group of nondiabetic CKD patients for comparison because a major objective of our study was to determine whether total or activated GSK3β level is a specific marker of DKD or a generic marker of kidney damage irrespective of the underlying cause. We found that the urinary GSK3β level and intrarenal GSK3β level in the DKD group were significantly higher than those of the nondiabetic CKD group, suggesting that the abnormal GSK3β level was at least partly specific for DKD. However, there was no significant difference in urinary GSK3β mRNA expression between the two groups.

Our study carries several limitations. First, we did not have a control group with normal kidney biopsy. Nonetheless, we found no correlation between GSK3β level (either intrarenal or urinary) and any clinical parameter in the nondiabetic group, suggesting that GSK3β level is not affected by kidney damage per se. Second, unlike the previous study of Liang et al. [14], we only measured the total GSK3β level in urinary supernatant due to the limitation of experimental design and the availability of urine sample. Finally, the kidney biopsy of nondiabetic CKD, and probably DKD, may represent a heterogeneous group, although we did exclude cases with obvious immune deposit or underlying causes. Further studies are necessary to explore the expression of GSK3β in specific kidney diseases and its relationship with their disease progression.

In summary, our results showed that intrarenal and urinary GSK3β levels were increased in DKD as compared to nondiabetic CKD. Neither urinary GSK3β nor its p-GSK3β levels predicted the progression of DKD, and the prognostic value of the urinary GSK3β level in nondiabetic CKD was also limited. However, the intrarenal pY216-GSK3β/total GSK3β ratio was associated with the rate of progression of DKD. Our findings indicated that urinary GSK3β may not be a suitable biomarker for DKD, but the pathophysiological roles of GSK3β in DKD progression deserve further studies.

The study was approved by the Joint Chinese University of Hong Kong – New Territories East Cluster Clinical Research Ethics Committee (approval number CRE-2019.283). All study procedures were in compliance with the Declaration of Helsinki, and all subjects provided written informed consent.

The authors declare no other conflict of interest.

This study was supported by the Research Grant Council Research Impact Fund (project reference R4012-18) and the Chinese University of Hong Kong research accounts 6905134, 2410026, and 3133200. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Research idea and study design, data analysis/interpretation, and statistical analysis: L.Z. and C.C.S.; data acquisition: L.Z., W.W.S.F., G.C.S.C., and J.K.C.N.; supervision or mentorship: K.M.C. and C.C.S. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.

1.
Levey
AS
,
Becker
C
,
Inker
LA
.
Glomerular filtration rate and albuminuria for detection and staging of acute and chronic kidney disease in adults: a systematic review
.
JAMA
.
2015 Feb 24
313
8
837
46
.
2.
Kerjaschki
D
.
2015 homer W. Smith award: the podocyte from periphery to center stage
.
J Am Soc Nephrol
.
2016 Nov
27
11
3266
70
.
3.
Torban
E
,
Braun
F
,
Wanner
N
,
Takano
T
,
Goodyer
PR
,
Lennon
R
.
From podocyte biology to novel cures for glomerular disease
.
Kidney Int
.
2019 Oct
96
4
850
61
.
4.
Zeng
L
,
Szeto
CC
.
Urinary podocyte markers in kidney diseases
.
Clin Chim Acta
.
2021 Dec
523
315
24
. Epub 2021 Oct 16.
5.
Szeto
CC
,
Lai
KB
,
Chow
KM
,
Szeto
CYK
,
Yip
TWC
,
Woo
KS
.
Messenger RNA expression of glomerular podocyte markers in the urinary sediment of acquired proteinuric diseases
.
Clin Chim Acta
.
2005 Nov
361
1–2
182
90
.
6.
Wang
G
,
Lai
FM
,
Kwan
BC
,
Lai
KB
,
Chow
KM
,
Li
PK
.
Podocyte loss in human hypertensive nephrosclerosis
.
Am J Hypertens
.
2009 Mar
22
3
300
6
.
7.
Chevalier
RL
.
The proximal tubule is the primary target of injury and progression of kidney disease: role of the glomerulotubular junction
.
Am J Physiol Renal Physiol
.
2016 Jul 1
311
1
F145
61
.
8.
Liu
BC
,
Tang
TT
,
Lv
LL
,
Lan
HY
.
Renal tubule injury: a driving force toward chronic kidney disease
.
Kidney Int
.
2018 Mar
93
3
568
79
.
9.
Xu
W
,
Ge
Y
,
Liu
Z
,
Gong
R
.
Glycogen synthase kinase 3β dictates podocyte motility and focal adhesion turnover by modulating paxillin activity: implications for the protective effect of low-dose lithium in podocytopathy
.
Am J Pathol
.
2014
;
184
(
10
):
2742
56
.
10.
Ali
A
,
Hoeflich
KP
,
Woodgett
JR
.
Glycogen synthase kinase-3: properties, functions, and regulation
.
Chem Rev
.
2001
;
101
(
8
):
2527
40
.
11.
Li
C
,
Ge
Y
,
Dworkin
L
,
Peng
A
,
Gong
R
.
The β isoform of GSK3 mediates podocyte autonomous injury in proteinuric glomerulopathy
.
J Pathol
.
2016
;
239
(
1
):
23
35
.
12.
Zhou
S
,
Wang
P
,
Qiao
Y
,
Ge
Y
,
Wang
Y
,
Quan
S
.
Genetic and pharmacologic targeting of glycogen synthase kinase 3β reinforces the Nrf2 antioxidant defense against podocytopathy
.
J Am Soc Nephrol
.
2016
;
27
(
8
):
2289
308
.
13.
Chen
B
,
Wang
P
,
Liang
X
,
Jiang
C
,
Ge
Y
,
Dworkin
LD
.
Permissive effect of GSK3 on profibrogenic plasticity of renal tubular cells in progressive chronic kidney disease
.
Cell Death Dis
.
2021 Apr 30
12
5
432
.
14.
Liang
X
,
Wang
P
,
Chen
B
,
Ge
Y
,
Gong
AY
,
Flickinger
B
.
Glycogen synthase kinase 3β hyperactivity in urinary exfoliated cells predicts progression of diabetic kidney disease
.
Kidney Int
.
2020 Jan
97
1
175
92
.
15.
Levey
AS
,
Stevens
LA
,
Schmid
CH
,
Zhang
YL
,
Castro
AF
3rd
,
Feldman
HI
.
A new equation to estimate glomerular filtration rate
.
Ann Intern Med
.
2009
;
150
(
9
):
604
12
.
16.
Wang
G
,
Szeto
CC
.
Methods of microRNA quantification in urinary sediment
.
Methods Mol Biol
.
2013
;
1024
:
211
20
.
17.
Wang
G
,
Lai
FM
,
Chow
KM
,
Kwan
BC
,
Pang
WF
,
Luk
CC
.
Urinary mRNA levels of ELR-negative CXC chemokine ligand and extracellular matrix in diabetic nephropathy
.
Diabetes Metab Res Rev
.
2015 Oct
31
7
699
706
.
18.
Wei
PZ
,
Kwan
BC
,
Chow
KM
,
Cheng
PM
,
Luk
CC
,
Lai
KB
.
Urinary mitochondrial DNA level in non-diabetic chronic kidney diseases
.
Clin Chim Acta
.
2018 Sep
484
36
9
.
19.
Doble
BW
,
Woodgett
JR
.
GSK-3: tricks of the trade for a multi-tasking kinase
.
J Cell Sci
.
2003
116
Pt 7
1175
86
.
20.
Hurcombe
JA
,
Hartley
P
,
Lay
AC
,
Ni
L
,
Bedford
JJ
,
Leader
JP
.
Podocyte GSK3 is an evolutionarily conserved critical regulator of kidney function
.
Nat Commun
.
2019 Jan 24
10
1
403
.
21.
Nikoulina
SE
,
Ciaraldi
TP
,
Mudaliar
S
,
Mohideen
P
,
Carter
L
,
Henry
RR
.
Potential role of glycogen synthase kinase-3 in skeletal muscle insulin resistance of type 2 diabetes
.
Diabetes
.
2000
;
49
:
263
71
.
22.
Nikoulina
SE
,
Ciaraldi
TP
,
Mudaliar
S
,
Carter
L
,
Johnson
K
,
Henry
RR
.
Inhibition of glycogen synthase kinase 3 improves insulin action and glucose metabolism in human skeletal muscle
.
Diabetes
.
2002
;
51
(
7
):
2190
8
.
23.
Henriksen
EJ
,
Kinnick
TR
,
Teachey
MK
,
O’Keefe
MP
,
Ring
D
,
Johnson
KW
.
Modulation of muscle insulin resistance by selective inhibition of GSK-3 in Zucker diabetic fatty rats
.
Am J Physiol Endocrinol Metab
.
2003
284
5
E892
900
.
24.
Gong
R
,
Rifai
A
,
Ge
Y
,
Chen
S
,
Dworkin
LD
.
Hepatocyte growth factor suppresses proinflammatory NFkappaB activation through GSK3beta inactivation in renal tubular epithelial cells
.
J Biol Chem
.
2008
;
283
(
12
):
7401
10
.
25.
Li
C
,
Ge
Y
,
Peng
A
,
Gong
R
.
The redox sensitive glycogen synthase kinase 3β suppresses the self-protective antioxidant response in podocytes upon oxidative glomerular injury
.
Oncotarget
.
2015
;
6
(
37
):
39493
506
.
26.
Mariappan
MM
,
Prasad
S
,
D’Silva
K
,
Cedillo
E
,
Sataranatarajan
K
,
Barnes
JL
.
Activation of glycogen synthase kinase 3β ameliorates diabetes-induced kidney injury
.
J Biol Chem
.
2014
;
289
(
51
):
35363
75
.