Peroxiredoxins as Markers of Oxidative Stress in IgA Nephropathy, Membranous Nephropathy and Lupus Nephritis

Chronic kidney disease (CKD) is a growing public health problem, affecting approximately 8–13% of the population (Brück et al. 2016; Hill et al. 2016). Furthermore, it is projected that in 2040, CKD will be the fifth leading cause of death in the world (Foreman et al. 2018). Glomerulonephropathies (GNs) such as IgA nephropathy (IgAN), membranous nephropathy (MN), and lupus nephritis (LN) are immune-mediated, although they have different etiologies, and are among the most frequent causes of CKD (Pippias et al. 2017). GNs account for about 20% of CKD cases in most countries, usually affecting young people and carry a lifelong CKD burden (Floege and Amann 2016). IgAN and MN belong to the primary GNs, and their annual incidence is estimated at 2–5 and 1–2 cases per 100,000 adults, respectively (McGrogan et al. 2011). LN develops secondary to systemic disease, and its incidence is estimated at 0.4–0.7 cases per 100,000 population per year (Patel et al. 2006). Unfortunately, GNs frequently progress asymptomatically or present with proteinuria, erythrocyturia or hematuria, edema, and hypertension (Vassalotti et al. 2010). These symptoms are neither specific nor sensitive enough for any GN. The diagnosis is difficult and frequently too late, and still requires histopathological evaluation by kidney biopsy, an invasive procedure with known risks (Mucha et al. 2016). Therefore, alternative, specific, reproducible, and safer methods are needed to facilitate noninvasive diagnosis.

There are specific markers that may help to diagnose glomerular diseases, including galactose-deficient IgA1 and IgG autoantibodies that correlate with IgAN (Placzek et al. 2018), anti-phospholipid 2 receptor antibodies that correlate with the histological picture of MN, and anti-double-stranded DNA antibodies associated with LN activity (Na et al. 2017). Multiple urine and serum proteins (Gao et al. 2018; Krata et al. 2018; Moszczuk et al. 2021; Mucha et al. 2014) or gene polymorphisms (Xie et al. 2020; Pac et al 2021) have been proposed as markers of different kidney diseases in the last decade. However, their diagnostic and/or prognostic utility remains to be validated (Krata et al. 2018; Selvaskandan et al. 2020; Sethi et al. 2019, 2020; Yanagawa et al. 2014). In this study, we focused on oxidative stress-related markers, 2-cysteine peroxiredoxins (2-Cys PRDXs), as potentially discriminatory in renal diseases.

Oxidative stress (OS) is one of the mechanisms involved in the progression of every type of CKD, including GN (Krata et al. 2018). Indeed, specific CKD-related conditions may lead to the overproduction of reactive oxygen species (ROS). It was reported that CKD patients have increased levels of plasma thiol oxidation and carbonylation, but the role of PRDXs in the pathophysiology of kidney diseases remains unknown (Cachofeiro et al. 2008; Krata et al. 2018). PRDXs, which are similar in function to well-known antioxidant enzymes such as catalase and glutathione peroxidase, possess the ability to reduce excessive levels of hydrogen peroxide, one of the major OS mediators (Jeong et al. 2012; Yang and Lee 2015). Importantly, kinetics measurements imply that PRDXs reduce more than 90% of cellular peroxides (Adimora et al. 2010; Perkins et al. 2015; Winterbourn 2008), which predisposes them to being a crucial factor in cellular OS regulation. Oxidative stress has been reported in kidney disease, due to both antioxidant depletions as well as increased production of ROS (Daenen et al. 2019; Irazabal and Torres 2020). The kidney is a highly metabolic organ, rich in oxidation reactions in mitochondria, which makes it vulnerable to damage caused by ROS (Aranda-Rivera et al. 2021). Therefore, OS can accelerate kidney disease progression. Different PRDX isoforms were reported to be involved in diabetic nephropathy (Lee and Lee 2018), ischemia/reperfusion damage (Sharapov et al. 2020), obstructive kidney disease (Hwang et al. 2019), ciliopathies (Zacchia et al. 2020) and acute tubular necrosis (Wu et al. 2017). However, the role of PRDXs in the pathophysiology of glomerular diseases is not well known.

In the current study, we hypothesized that 2-Cys PRDXs could be differentially involved in IgAN, MN, and LN. If so, the PRDX family could serve as additional markers of specific GNs. Therefore, the aim of this study was to evaluate PRDX 1–5 serum concentrations in IgAN, MN, and LN patients and healthy controls.

In this study, we show that serum levels of PRDX 1–4 were significantly elevated in IgAN, MN, and LN patients compared to healthy individuals. In addition to the differences between disease and control groups, we also found that each GN type revealed a distinct PRDX pattern. There is no straight explanation why there would be differences in oxidative stress markers between different types of GN. The data on biological reasons for these differences are largely missing.

Oxidative stress is defined as a disturbance in the natural ability of cells to maintain a balance between pro- and antioxidant systems (Krata et al. 2018; Selvaskandan et al. 2020; Sethi et al. 2019, 2020; Yanagawa et al. 2014). Although moderate levels of hydrogen peroxide are necessary for numerous cellular processes, in excessive concentrations it can cause cell and tissue damage (mediated by oxidation of lipids, proteins, or DNA). Several mechanisms are responsible for the removal of reactive oxygen species, including superoxide dismutase, catalase, peroxidase, the peroxide–redox–thioredoxin–thioredoxin reductase enzymatic chain, and a number of non-enzymatic antioxidants (Descamps-Latscha et al. 2001). PRDXs are known as biomarkers for cancer (Basu et al. 2011; O'Leary et al. 2014), bacterial infections (Yang et al. 2018), and neurodegenerative (Goemaere and Knoops 2012) and inflammatory-related diseases (Park et al. 2016). Their role in the pathophysiology and/or diagnostics of IgAN, MN, and LN has not been characterized yet.

The fundamental method of following up CKD, including GN patients, is eGFR estimation. We observed an inverse correlation between eGFR and PRDX 2 in IgAN and LN but not MN patients and PRDX3 in IgAN and MN individuals. For the obvious reasons, a simple link between eGFR and serum PRDX probably does not exist. However, declining eGFR is the evident effect of kidney disease progression involving, e.g., oxidative stress. Coexisting secondary anemia or other comorbidities could additionally influence PRDX levels in serum (discussed below). Another gold standard for GN follow-up is 24-h urine protein loss analysis. Higher proteinuria usually indicates more advanced or active GN. However, we observed only one correlation between PRDX 2 and 24 h proteinuria in MN patients. The implication of this is unclear. The highly variable degree of proteinuria in MN patients, related to the MN biology could explain this finding. On the other hand, a potential relation to protein loss may be suspected also in other GNs based on the observed PRDX association with serum proteinogram changes. Therefore, the causes of proteinogram changes are probably multifactorial and OS is differentially involved. Moreover, to date, there has not been much published data about PRDX and GN, particularly in the context of proteinuria. Of note, proteinuria may additionally serve as an indirect indicator of the GN activity. One could speculate that due to the relative unspecificity of OS, PRDX serum levels might be expected to reflect disease activity rather than disease itself. This speculation could not be elucidated with our study design, since each studied GN has distinct activity scale. Thus, the comparison of disease activity between these groups is impossible. Moreover, diagnostic renal biopsies were performed in history, not at the time of sampling.

OS is implicated in various or almost all disease states, including atherosclerosis, cardiovascular disease, obesity, diabetes, cancer, neurodegeneration, aging, drugs (e.g., allopurinol) and many others. Such conditions could confound the results and anemia belongs to the most potent OS confounders. Along with CKD progression, the uremic toxins increase red blood cell (RBC) damage and erythropoietin deficiency reduces HGB production, resulting in tissue hypoxia. The latter may stimulate OS responses, including, e.g., increased PRDXs (Gwozdzinski et al. 2021). Indeed, we found a correlation between HGB and serum PRDX 2 concentrations in IgAN and LN patients. It was reported that PRDX 2 plays an important role in the protection of RBCs from OS through HGB autoxidation (Johnson et al. 2005), e.g., in iron-deficiency anemia (Nagababu et al. 2008). The protective function of PRDX 2 in HGB stability has been investigated in PRDX 2 knockout mice and RBCs from patients with hereditary hemolytic anemia. The authors of this study suggested that PRDX 2 could bind to HGB and protect it from oxidative denaturation and aggregation in RBCs (Han et al. 2012). What is very important in our study is the fact that our patients had stage 2 CKD and an average eGFR of 71, 74, and 84 mL/min (for IgAN, MN, and LN, respectively), thus had normal RBC and HGB levels. One of the limitations of our study was relatively low patient numbers in each GN group, which precluded analysis of the effect other comorbidities on PRDX concentrations. Therefore, the question is, what does the increased serum PRDX really mean? Do PRDX levels behave as a damage marker or damage type marker? Is it possible that it is just a very early marker of preexisting hypoxia and/or OS in CKD patients long before the development of even early stages of anemia? The fact that different GNs enhance different PRDX classes also suggests different disease-specific mechanisms of hypoxia, which needs additional research. Another question is whether the serum concentrations of PRDXs should be only used as markers for a specific disease type or can also reflects the activity of this particular disease at the time of sample collection? In addition, oxidative stress activity in serum may not always reflect that of changes cellular microdomains in the kidney, which should be taken under consideration while interpreting the results of the current study. Finally, it is also worth adding that evaluation of all GN groups vs. healthy individuals showed significant differences in PRDX 1 and 2 (as illustrated on a heatmap). Therefore, it is an open question whether serum PRDX 1 and 2 might help in the future to distinguish patients suspected of having IgAN, MN-, or LN-related oxidative stress?

Complement activation and its regulation are complex phenomena. It is not definitive whether these phenomena are the direct primary cause or the effector mechanism of kidney injury, but they are involved in GN development and progression; thus, inhibiting them has become an emerging treatment option (Kaartinen et al. 2019). In patients with IgAN, the immunoglobulin A deposited in the mesangial area of the kidney can activate the complement system through either the lectin pathway or an alternative pathway. This may amplify the local inflammatory response and contribute to renal injury (Daha and van Kooten 2016). The immunopathology of human MN shows evidence of complement activation within immune deposits (Beck and Salant 2010). Therefore, it is thought likely that the complement system also plays a substantial role in MN pathogenesis. Although serum levels of complement proteins are usually normal in MN patients, it was recently reported that measuring the circulating complement activation products may be a way to detect ongoing complement activation (Zhang et al. 2019). Complement also has an important role in the pathogenesis of LN. On the one hand, a deficiency of complement components predisposes to lupus, while on the other hand, excess complement activation increases renal damage, and measuring it is done to assess disease activity (Sharma et al. 2020). Although OS involvement in renal injury being driven by complement activation seems obvious, to the best of our knowledge, to date there are no available data linking complement activation in GNs to PRDXs. We assume that the disease-specific PRDX associations with circulating C3 and C4 complement components observed in our study suggest differential OS responses depending on the GN etiology and the mode of complement activation. If so, PRDX subclass assessment might be useful as an adjunct in the diagnosis of GNs.

Another finding of our study is the positive correlation between PRDX 5 and BMI in the IgAN group. It was previously demonstrated in obese mouse models that PRDX 5 inhibits adipogenesis by modulating ROS generation and adipogenic gene expression, implying that it may serve as a potential target to prevent and treat obesity (Kim et al. 2018). Moreover, in vitro and in vivo experiments suggested that PRDX 5 functions as a protective regulator in fatty liver disease and may be a valuable therapeutic target for the management of obesity-related metabolic diseases (Kim et al. 2020). Taking these findings and our results together, it is possible that PRDX 5 is upregulated in response to chronic nephritis depending on BMI value, but this hypothesis needs further elucidation.

It is important to mention that in the current study the serum levels of PRDXs were evaluated, which might not reflect the intracellular changes in PRDXs expression within the tissue. As PRDXs can be released as damage-associated molecular pattern markers from injured tissues (He et al. 2019), this subject warrants further investigations.

Generally, our study indicates potential applicability of antioxidant supplementation in renal disease. The potential compounds to be used are, for instance: coenzyme Q10, Vitamins B, C, D, and E, L-carnitine, statins, or N-acetylcysteine (Liakopoulos et al. 2019). Indeed, application of coenzyme Q10 has been reported beneficial in the treatment of chronic diseases, including CKD (Gutierrez-Mariscal et al. 2020).

In conclusion, our results highlight the link between PRDXs and GNs (IgAN, MN, and LN). Our study indicates that individual PRDXs can play roles in pathophysiology of selected GNs and that their concentrations in serum may become useful as new supplementary diagnostic markers in IgAN, MN, and LN. Validation studies including other kidney diseases are required to better understand GNs and enable their less invasive diagnosis.