Introduction

In 2017, chronic kidney disease (CKD) was included in the top ten global burden of disease with a prevalence of 843.6 million worldwide1. CKD is also a significant public health concern in Thailand, affecting 4.6–17.5% of the Thai population2. Persistence of kidney damage or loss of kidney function over 3 months is employed to discriminate CKD from acute conditions. Moreover, clinical criteria including an estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2, albuminuria, presence of abnormal urinary sediment, and abnormal renal ultrasound are used for staging categorization3. Early screening with robust methods is beneficial for the development of effective treatment regimens. Thus, markers with proper sensibility and specificity are necessary to recognize CKD early. Although different markers have been reported and applied in routine practice, such can mostly diagnose CKD in advanced disease stages4.

Cyclophilin A (CypA) is an indicator of inflammation and oxidative stress5, and the association between CypA and diabetes mellitus (DM) has been described previously. Long-term hyperglycemia and dysregulation of the metabolic environment trigger the production of advanced glycation end products, activation of cellular signaling pathways, and the generation of free radicals6,7. In particular, monocytes secrete CypA, resulting in increased plasma CypA levels8. The secretion of CypA into the peripheral blood can also be observed in inflammatory conditions, infectious diseases, and solid organ damage, such as rheumatoid arthritis, asthma, and periodontitis9. Moreover, various reports have emphasized the strong relationship between macrophages and diabetic nephropathy (DN)10,11. As CypA is freely filtered from the glomeruli, elevation of urinary CypA (uCypA) levels may reflect an increase in plasma levels or directly reflect renal injury itself. Rather than causing phenotypic damage to resident kidney cells, activated macrophages release profibrotic cytokines, which increase cell proliferation and matrix volume expansion, and induce fibrosis12,13. More than 20% of patients with DM are estimated to develop CKD, with a 3% incidence 10–20 years after diabetes onset7.

A previous report has associated plasma CypA levels with DN and the severity of kidney damage. Urine as the less-invasive sample is the choice for CypA investigation. Thus, uCypA may reflect the occurrence at the kidney regional site. This study aimed to develop an enzyme-linked immunoassay (ELISA)-based method to identify uCypA cutoff levels for early-stage CKD. To simplify the detection procedure, an in-house CypA detection method was established using a lateral flow system. Furthermore, the correlation between uCypA and common noncommunicable diseases, including DM, DN, and CKD, was investigated.

Results

Characteristics of the study groups

Laboratory and clinical parameters of the participants of the study are shown as means and standard deviation (SD), while demographics, behaviors, and medicinal implementation are presented as numbers and percentages in Table 1. Among the 357 patients, patients with T2DM, hypertension, and hyperlipidemia condition accounted for 40.1%, 44%, and 71.2%, respectively. The highest levels of fasting blood sugar (FBS) and hemoglobin A1c (HbA1c) were found in patients with type 2 diabetes mellitus (T2DM) and early-stage CKD. In patients with late-stage CKD with and without T2DM, the higher creatinine levels corresponded to the lower eGFR, which was less than 60 mL/min/1.73 m2. The highest albumin creatinine ratio (ACR) was found in patients with late-stage CKD with T2DM, followed by those with late-stage CKD and early-stage CKD with T2DM.

Table 1 Baseline characteristics of the study participants.
Full size table

The biochemical parameters stratified by groups (T2DM < 5 y, T2DM ≥ 5 y, T2DM + early CKD, T2DM + late CKD, early-stage CKD and late-stage CKD) were compared to the control group using the Kruskal–Wallis test, as shown in Supplementary Table S1. Compared to parameters indicating kidney function, significant differences in urine microalbumin (umAlb) and ACR values were observed comparing the control groups with the remaining groups, except for T2DM < 5 y. Interestingly, among the seven study groups, the levels of uCypA ranged from 0.02 to 39.60 µg/mL. Significant differences in uCypA levels were observed between the control and remaining groups (p-value < 0.001).

The association of uCypA levels and other parameters

The association between uCypA and other characteristic factors was explored using linear regression analysis. uCypA was associated with age, FBS, low density lipoprotein (LDL), eGFR, umAlb, ACR, and metformin levels (Table 2). Consequently, multiple linear regression analysis was performed, identifying an association between uCypA and FBS, eGFR, ACR, and use of metformin (p < 0.003, R2 = 0.530). uCypA was increased by 0.021 and 0.156 µg/mL for each increase of 1 mg/dL in FBS and decrease of 1 mL/min in eGFR, respectively.

Table 2 Linear regression analysis of variables association with uCypA levels.
Full size table

The association of uCypA with T2DM and CKD

uCypA was significantly increased under T2DM with and without CKD conditions, and CKDs. The association of increasing uCypA levels was most prominent in patients with T2DM > 5 years and in the early-stage CKD group compared with the control group (Table 3). The aORs describe that for each 1-unit increase in uCypA, the higher the probability of T2DM ≥ 5 years and early-stage CKD, accounting for 1.817 and 1.808 times compared to the control group, respectively.

Table 3 Multinomial logistic regression analysis of uCypA levels with associated factors.
Full size table

The accuracy of uCypA for CKD diagnosis

The capacity of uCypA levels (µg/mL) for CKD diagnosis was evaluated using receiver operating characteristic (ROC) curve analysis. The area under curve (AUC) showed a high predictive ability to differentiate the early-stage CKD group (n = 85) from the control group (n = 90; Fig. 1; Table 4). A cutoff value of 0.48 µg/mL provided 98.82% sensitivity and 92.21% specificity with an AUC at 0.982 (95% confidence interval [CI] 0.989–1.000). Moreover, a cutoff of 2.89 µg/mL of uCypA discriminated early-stage (n = 85) and late-stage CKD (n = 96) with 80.21% sensitivity and 50.59% specificity with an AUC of 0.777 (95% CI 0.709–0.845) (Table 5).

Fig. 1
figure 1

ROC curve of urinary cyclophilin A for the (a) control and early-CKD and (b) early and late-CKD groups. ROC receiver operating characteristic, CKD chronic kidney disease.

Full size image
Table 4 Sensitivity and specificity of uCypA in the control and early-stage CKD groups.
Full size table
Table 5 Sensitivity and specificity of uCypA in the early and late-stage CKD groups.
Full size table

Establishment of a uCypA rapid screening method

To establish a rapid screening test for CKD, lateral flow strip (LFS) platform was developed to monitor CypA levels in urine samples. uCypA levels in the control, T2DM with early-stage CKD, and early-stage CKD groups were evaluated using LFS. A reddish band was observed on the control line in the control group (Fig. 2a). Meanwhile, the appearance of a red band in both the control and test lines was detected in T2DM with early-stage CKD and early-stage CKD groups. Additionally, quantitative detection of uCypA by LFS was performed using a LFS reader (Fig. 2b). The results showed that the intensities of the test line in the control group were 3.2 times lower than those in the T2DM with early-stage CKD group and 2.4 times lower than those in the early-stage CKD group.

Fig. 2
figure 2

LFS detection results of uCypA levels. (a) Images of LFS detection of uCypA levels in urine samples, including the running buffer, control group, T2DM with early-stage CKD group, and early-stage CKD group. (b) Corresponding test line intensity for uCypA levels in urine samples, including the running buffer, control group, T2DM with early-stage CKD group, and early-stage CKD group. LFS lateral flow system, uCypA urinary cyclophilin A, T2DM type 2 diabetes mellitus, CKD chronic kidney disease.

Full size image

Discussion

CKD is defined as abnormalities in kidney structure or function that are present for more than three months. Other clinical manifestations, including eGFR < 60 mL/min/1.73 m2, presence of pathological findings, markers of kidney damage, or abnormal imaging tests may occur14. To evaluate renal function, clinical biochemical markers, such as eGFR, are assessed. eGFR is based on creatinine or cystatin C levels, which are freely filtered at the glomeruli15. Moreover, a combination of surrogate factors, including age, sex, ethnicity, and body size, are included in the eGFR equation. Elevation of serum creatinine (sCr) in patients with late-stage kidney disease reflects kidney malfunction, in which by-products cannot be excreted from the blood circulation. The decrease in eGFR values in patients with late-stage CKD (36.3 ± 17.3) and late-stage CKD with T2DM (36.3 ± 15.6) was associated to the staging criteria of CKD3. Additionally, the increase in umAlb and ACR levels in late-stage CKD and late-stage CKD with T2DM can be explained by a defect in glomerular permeability. To predict the severity and progression of the disease, > 30 mg/g ACR was designated as the CKD criterion3. Moreover, > 30 mg/g of ACR in patients with T2DM ≥ 5 y may reflect an abnormality of glomerular filtration in patients with poorly controlled DM and a probability of CKD progression. Besides, elevated uCypA and normal eGFR value can be observed in DM < 5 y. It can be described by the production of oxidative stress and inflammatory cytokines resulting from metabolic alteration and persistent low-grade inflammation under chronic hyperglycemic condition16,17,18. These evidences triggered vascular complication and impacted the local organs. The pathological changes of kidney such as glomerular basement membrane thickening, glomerular ECM production, fibrosis, nephrosclerosis and podocyte apoptosis were occurred prior to the functional changes18,19. It was evinced by normal/increase eGFR, and normal/intermittent albuminuria in the early DM, while the long-term DM condition led to decline of eGFR and albuminuria20,21,22 (Fig. 3).

Fig. 3
figure 3

The schematic of pathology related to diabetes and renal complications. Under hyperglycemic conditions, imbalanced glucose homeostasis and persistent low-grade inflammation generated oxidative stress and inflammatory cytokines, which caused vascular complications and impacted the local organs16,17,18. At the kidney site, oxidative stress by-products and inflammatory cytokines triggered pathological changes such as glomerular basement membrane thickening, glomerular ECM production, fibrosis, nephrosclerosis, and podocyte apoptosis18,19. Normal/increased eGFR and normal/intermittent albuminuria can be observed during the structural changes, while secretion of inflammatory biomarkers such as CypA can also be observed5,20,21,22. The long-term effects of these matters have led to functional changes in the kidney, resulting in microalbuminuria, proteinuria, and a decline in eGFR. Not only the regional site but also peripheral blood monocytes can secrete CypA as a result of an inflammatory response. Secreted CypA was then freely filtered through glomeruli and observed in urine. Chronic kidney damage or injury also triggered the CypA secretion through the ROS-Rho pathway. It can perform both autocrine and paracrine functions, stimulating cell proliferation, migration, and inflammatory cascade and promoting cytokines and chemokines release5. ECM extra cellular matrix,eGFR estimated glomerular filtration rate,CypA cyclophilin A,ROS reactive oxygen species.

Full size image

The elevation of uCypA levels in CKD-associated conditions can be explained by the presence of both local and systemic inflammation23. The inflammatory response exerts paracrine and autocrine effects by secreting extracellular Cyclophilin A (eCypA). According to kidney injury, increased expression of CD147 bound with circulating eCypA activates the ERK/NF-kB pathway, leading to an inflammatory cascade24,25. Moreover, eCypA exerts a chemotactic effect, resulting in the infiltration of various inflammatory cells at regional sites5. In summary, increased uCypA levels were observed in patients since early-stage CKD. Moreover, uCypA was predominantly found in late-stage CKD, as well as in patients with T2DM with and without late-stage CKD; such is evidenced by responses to inflammatory stimuli such as hyperglycemia and oxidative stress. eCypA secreted from peripheral blood monocytes increases plasma eCypA level8. High glucose levels can also activate the STAT3-CypA interaction, which induces inflammation, oxidative stress, and apoptosis in podocytes26. Since circulating eCypA is freely filtered through the glomeruli, high levels of uCypA reflect elevated eCypA secretion from responsive cells and/or increased glomerular filtration or leakage27,28,29. It can be emphasized that the elevation of uCypA reflects the increase of circulation CypA as well as localized inflammation.

The correlation between uCypA and CKD can be explained by the exposure to generated inflammatory mediators in the vascular system and kidneys5. An association between eCypA and nephrotic conditions, including CKD, DN, and acute kidney injury (AKI) has been reported25,28,30. Long-term inflammation activation impairs renal function, limits tissue regeneration after cellular injury, and leads to progressive decline in renal function23. The use of uCypA as a biomarker for early-stage CKD may play an important role in providing effective treatments for patients with early-stage CKD. A cutoff value of 0.48 for patients in the early-stage CKD and control groups resulted in over 90% sensitivity and specificity, with an AUC of 0.982. A previous report found a significant difference in serum CypA levels between patients with non, early stage, and late-stage CKD31. Another cross-sectional study on T2DM also reported high accuracy of CypA in the early diagnosis of DN, with an AUC of 0.91432. Although routinely employed markers such as blood urea nitrogen, cystatin C, creatinine, and albuminuria have been established for CKD diagnosis, such cannot be used to distinguish the early stages of CKD33,34,35. The eGFR, which is the gold standard for determining kidney function and disease progression, might not be specific under certain conditions, such as extrarenal cystatin C or creatinine metabolism in patients with human immunodeficiency virus (HIV) infection, chronic liver disease, and cardiovascular diseases36,37. Moreover, a lack of association between eGFR and albuminuria has been observed in patients with end-stage renal disease (ESRD)38,39,40. Implementation of routine markers combined with proteomic or metabolomic markers may provide effective screening outcomes41,42. However, the high cost of biomarker detection and inconsistent detection methods limit their clinical application43. The development of a rapid screening method for uCypA using LFS could facilitate access to screening methods for the general population. Moreover, application of the lateral flow system provided impressive results for screening early-stage CKD with and without T2DM in our study. To improve detection sensitivity, capture systems should be developed.

The promising high sensitivity and specificity at a 0.48 µg/mL uCypA cutoff could be crucial for early CKD screening. uCypA levels also elucidate the renal complications of DM, indicating a strong association with other kidney and glycemic parameters, including eGFR, ACR, and FBS. Rather than palliative treatment and health expenditures, the addition of uCypA in routine practices may enable the initiation of effective treatment regimens for patients with early-stage CKD. For applying uCypA in predictive/prognostic approach, an extension of deteriorated renal subjects should be implied. In addition, the T2DM durational effect of the patient with and without CKD should be a concern. Exclusion of poor glycemic control patient group should be considered for further study.

Materials and methods

Sample population

A total of 375 patients with metabolic syndrome were recruited from the Chronic Kidney Disease Northeast Thailand (CKDNET) project. The patient with waist circumference ≥ 90 cm for male or ≥ 80 cm for female, and the presence of any two criteria as follows: (1) hypertriglyceridemia (≥ 150 mg/dL), (2) low HDL (< 40 mg/dL in male or < 50 mg/dL in female), (3) elevated blood pressure (SBP ≥ 130 mmHg and DBP ≥ 85 mmHg or drug treatment for hypertension), and (4) fasting blood sugar ≥ 100 mg/dL or drug treatment for diabetes mellitus, was diagnosed as metabolic syndrome44. The study protocol conformed to the ethical guidelines based on the Declaration of Helsinki was approved by the Human Ethics Committee of Khon Kaen University (HE 601035), and all participants provided written informed consent before the study. Patients with inflammation, infection, liver disease, malignancy, or autoimmune diseases were excluded from the study. Additionally, the patient with serious conditions regarding the diabetes complications were also excluded. The remaining 357 participants were divided into the following seven groups:

Group I

control group (subjects who were free from underlying diseases, inflammation, infection, and medicinal treatment; n = 90);

Group II

patients with T2DM diagnosed less than five years prior (T2DM < 5 y; n = 45);

Group III

patients with T2DM diagnosed within and more than five years prior (T2DM ≥ 5 y; n = 41);

Group IV

patients with T2DM and early-stage CKD, i.e., stages I and II (T2DM + early CKD; n = 38);

Group V

patients with T2DM and late-stage CKD, i.e., stages III, IV and V (T2DM + late CKD; n = 52);

Group VI

patients with early-stage CKD, i.e., stages I and II (early CKD; n = 47);

Group VII

patients with late-stage CKD, i.e., stages III, IV, and V (late CKD; n = 44).

CKD staging classification was indicated using eGFR level from KDIGO guideline3. According to various CKD treatment guidelines, the indication for a nephropathic referral depends on eGFR, which is less than 60 mL/min/1.73 m245,46. Moreover, intensive monitoring was undertaken in the patient with a particular eGFR value. Therefore, 2 stages were categorized in this study, including early stage (stage I and II; eGFR ≥ 60 mL/min/1.73 m2), and late stage (stage III, IV and V; eGFR < 60 mL/min/1.73 m2). As per the clinical stages describing diabetic nephropathy progression20,47, 2 subgroups of patients without CKD were designated. Group II represented stage II DN describing the silent progression of kidney defect with less than 5 years after DM onset. Meanwhile group III illustrated stage III DN with more than 5 years after DM onset and prior for CKD diagnoses, coherent with various cohort studies reporting the estimating median time of 5–13 years for developing microalbumin and ESRD after DM onset21,48.

Baseline measurements

Fasting blood samples were collected from all the participants. Laboratory parameters including FBS, HbA1c, LDL cholesterol, alanine transaminase (ALT), uric acid, serum creatinine (sCr), and eGFR were determined. Morning urine was collected to examine urinary creatinine (uCr) and microalbumin (umAlb) levels and then calculate the ACR. Clinical examinations assessed the blood pressure (BP), including systolic blood pressure (SBP) and diastolic blood pressure (DBP). All data were collected twice within three months. A questionnaire inquired about demographics (age and sex), behaviors (exercise, current water consumption, smoking, and consumption of high-purine compounds), and use of medicines (metformin, glipizide, gliben, mixtard, Nph, amlodipine, enalapril, losartan, doxazosin, simvas, and atorvas).

ELISA for determining uCypA

uCypA levels were analyzed using an indirect ELISA method. Urine samples were diluted in a carbonate buffer (pH 9.5). A total 50 µL of the diluted urine was coated on the microplate (Thermo Scientific, MA, USA) and incubated overnight at 4 °C. The unbound antigen was washed three times with phosphate-buffered saline with Tween 20 (PBST). The wells were blocked with 100 µL of 3% skim milk-PBST for 2 h at 37 °C and washed three times with PBST. One hundred microliters of primary antibody (1:250; mouse monoclonal anti-human CypA antibody, Thermo Scientific, MA, USA) was added and then incubated for 2 h at 37 °C. After washing three times with PBST, 100 µL of conjugated anti-mouse immunoglobulin G (IgG) linked with alkaline phosphatase (Thermo Scientific, MA, USA) was added and incubated at 37 °C for 2 h. The plates were washed three times with PBST, followed by three washes with phosphate-buffered saline (PBS). Color development was executed by adding 100 µL of p-nitro-phenyl phosphate substrate solution and the reaction was then measured with an ELISA reader (PerkinElmer, MA, USA) at 405 nm. Calibration curves of CypA concentration levels at 0.001, 0.01, 0.1 and 1 ug/mL were utilized for calculation.

Statistical analyses

All statistical analyses were performed using the STATA statistical software package, version 13.0 (StataCorp LLC, TX, USA). Characteristic data for each variable were reported as means and standard deviations (SD) for continuous variables, and counted numbers and percentages for categorical variables. Normality testing using Kolmogorov–Smirnov test was performed for each biochemical parameter prior to comparison among the groups. Analysis of variance (ANOVA) was used to compare parametric data, whereas the Kruskal-Wallis test was applied for non-parametric data. The associations between uCypA levels and all characteristic factors were determined using multiple logistic regression analysis, and potential confounders (demographics, laboratory and clinical parameters, behaviors, and use of various drugs) were controlled. The association of uCypA among the groups was analyzed using multinomial logistic regression. Statistical significance was set at a p-value < 0.05.

Preparation of the LFS for uCypA detection

The LFS consisted of four parts: a sample pad (C048, Millipore Corporation, Billerica, MA, USA), a glass microfiber filter conjugate pad (GF33, Whatman Schleicher & Schuell, Dassel, Germany), a nitrocellulose membrane (UniSart® CN140, Sartorius Stedim Biotech SA, Goettingen, Germany), and an absorbent pad (Whatman ABS PAD #470, Sigma-Aldrich, MA, USA). The anti-CypA polyclonal antibody (Thermo Scientific, MA, USA) conjugated gold nanoparticle in a suspension buffer (10 mM phosphate buffer pH 7 + 1% bovine serum albumin (BSA) + 20% sucrose + 0.05% sodium azide) at an optical density of ∼5 was dropped onto the conjugated pad until reaching a volume of 20 µL before being dried at 37 °C for 2 h. Then, 1 mg/mL of mouse monoclonal anti-human CypA antibody (Thermo Scientific, MA, USA) and 1 mg/mL of goat anti-mouse IgG polyclonal antibody (Lampire Biological Laboratories, Pipersville, PA, USA) were dispensed (1 µL/cm) onto the nitrocellulose membrane as the test and the control line, respectively, by using a lateral flow dispenser (XYZ3000 Dispensing Platform; BioDot Inc., Irvine, CA, USA). Next, the membranes were dried at 37 °C for 1 h. The samples and absorbent pads were used without any treatment. Subsequently, all four parts were assembled on a backing card with a 2 mm overlap between successive components. Finally, the LFS was cut into narrower 4 × 60 mm strips using a guillotine cutter (BioDot Inc., Irvine, CA, USA).

LFS analytical procedure

To detect uCypA levels using the LFS platform, urine samples from the control (n = 5), T2DM + early CKD (n = 5), and early-stage CKD (n = 5) groups were randomly selected. Urine samples (10 µL) were diluted with 100 µL of running buffer (100 mM phosphate buffer + 1% Tween20 + 2% BSA), and the 110 uL mixture was subsequently released onto the sample pad and the sample was allowed to flow laterally for 15 min. After the assay was completed, test line intensities were visualized and quantified using a RapidScan ST5-W Lateral Flow Test Strip Reader (Pacific Image Electronic Co., Ltd., Taipei, Taiwan).