Regulatory T Cells and Oxidative Stress in Minimal Change Nephropathy

The suggestion about a pathogenetic role of both innate immunity and reactive oxygen species (ROS) in MCN derives from studies carried out in experimental models that reproduce, in the acute phase, minimal glomerular lesions and then evolve to more structured glomerulosclerosis. PAN and ADR are two compounds that can be classified as oxidants and have been for years utilized to induce proteinuria in rats [27, 28]. Metabolic studies and protection by antioxidants support an entirely oxidative stress in these models. Infusion of lipopolysaccharide (LPS) induces transient proteinuria in mice [23]. This model of proteinuria is of interest for studying the immunomodulatory link since LPS upregulates B7-1, a co-stimulatory molecule expressed by B cells and by other antigen-presenting cells that serves as signal for T cells. Mice lacking B7-1 are protected from LPS proteinuria [23] that suggests this is the mechanism for LPS nephropathy. In spite of transient proteinuria, LPS nephropathy is characterized by progressive signs of renal involvement resulting in glomerulosclerosis that mimics in some way the natural history of patients with MCN. Findings in humans with posttransplant recurrence of glomerulosclerosis treated with abatacept, the inhibitor of B7-1 molecule [34, 35], support the same conclusion. Connected with the B7-1 story is the possibility that the direct activation of B7-1 in podocyte by an inflammatory trigger producing LPS is the true mechanism for proteinuria and is independent of T or B cells. Experiments in severe combined immunodeficiency (SCID) mice, missing both cell lineages but still developing proteinuria after LPS, are central to this demonstration [23]. Overall, this conflicts with the general idea of a circulating factor and contributes to maintain unsolved clues that require further explanations.

Polymorphonuclear leukocytes (PMN) in children affected by INS produce high quantities of oxidants. Results indicate that soluble factors deriving from circulating cells regulate oxidant production and that this regulatory circuit is altered in MCN [36]. It has been proposed that Treg and probably B cells play a key role in the regulation of oxidant production by PMNs. The evidence for Tregs comes from a unique observation in a child with INS associated with immune dysregulation, polyendocrinopathy, enteropathy x-linked (IPEX) syndrome that is due to loss of function mutation of forkhead box P3 (Foxp3), a transcriptional factor specific to Tregs that makes these cells functionally negative [36]. Oxidant production in this child was 100 times higher during exacerbation of clinical symptoms and restored a near normal level in remission. On the side of B cells, there is the finding that rituximab decreased ROS production by 60 % [36].

Studies by Bertelli and colleagues [37] showed that the percentage of Tregs (CD39⁺CD4⁺CD25⁺) was markedly reduced in patients with nephrotic syndrome. It was also noted that generation of oxidants by PMN was indirectly correlated with the amount of CD39 on Tregs. Apyrase (CD39) is the key enzyme that transforms ATP into adenosine, the latter metabolite playing an inhibitory role on oxidant production. Overall, this indicated a key function of Tregs in modulating the oxidative stress in MCN (see below).

An indirect confirmation of these findings comes from the observation that, concomitant with proteinuria, a significant part of serum albumin is oxidated in MCN patients. Candiano and colleagues studied serum albumin in children with MCN by mean of mass spectrometry and showed that albumin is highly oxidized during the proteinuric phase. The site of oxidation was shown to be a free cystein of the albumin sequence 31Cys that was chemically transformed in a sulfonic group upon oxidation [38, 39].

Tregs could be involved in MCN as a second step in a cascade where the first hit remains unidentified. The evidence of a Treg involvement in MCN is entirely based on results coming from experimental nephrosis (i.e., in Buffalo/Mna rats and in ADR nephrosis that are two recognized models of chronic proteinuria leading to glomerulosclerosis and renal failure) [31, 40]. More recently, experimental data have been expanded to LPS nephropathy (see above) and some unexpected results to new pathogenic horizons [33]. Le Berre and colleagues [31] utilized Buffalo/Mna rats that spontaneously develop glomerulosclerosis showing that pre- and posttransplant proteinuria was reduced by infusion of Tregs; regression of the nephropathy was obtained as well. Several authors reported the same protective role of Tregs in ADR nephrosis that was mediated by adenosine or by direct infusion of FOXP3-transduced T cells [41].

IL2 has been utilized to enhance Treg proliferation and lifespan. To escape from bystander effects of this cytokine due to its wide-ranging action, the association of IL2 and anti-IL2 antibodies has demonstrated as a valid stratagem to improve its in vivo selectivity toward Treg population [42, 43]. In this case the binding of the cytokine to the α and β receptor chains, which are present on the CD8 and NK cell surface, is inhibited by preferential binding of anti-IL2, whereas the γ receptor chain, uniquely expressed by activated CD4+ (or effector T cell (Teff)) and by Treg cells, remains available for IL2 recognition and activity. The administration of low doses of IL2 coupled to this specific antibody was indeed proven to induce high levels of Tregs, being ineffective on CD8, on NK, and also on Teff counterpart. Polhill and colleagues [32] induced Treg expansion by IL2/anti-IL2 antibodies in rats with ADR nephrosis and documented improvement of renal function, reduced inflammation, and less pathologic injury. More articulated are the results of the study by Bertelli and colleagues [33] who utilized the same approach with IL2/ anti-IL2 antibodies in mice with LPS nephropathy. In fact, IL2/anti-IL2 antibody administration in mice exposed to LPS had no effect on the progression of the resulting renal damage, enhancing significantly peripheral tissue Treg levels, whereas IL2 infused alone elicited some protection despite less Tregs. Therefore, these results partially contradict a direct role of Tregs (higher after IL2/anti-IL2 administration) and supported hitherto undefined mechanisms.

Several of the molecules described in the context of MCN participate in Treg regulation. A description of regulatory compounds and functions may simplify the comprehension of pathology.

Tregs are a dynamic cell population whose levels can be rapidly modified and made active, with alternative functions, in any inflammatory context [44]. At equilibrium, Treg concentration in periphery is low, but when necessary they are produced from nonactivated T conventional (Tconv) cells through induction of FOXP3 in CD4⁺. Tconv can also differentiate into activated CD4+ CD25+ cytotoxic cells (Teff) by IL2. Mature Tregs play a bifunctional role: the first is to switch off the inflammatory burst by means of anti-inflammatory cytokines (IL10) and/or by adenosine deriving from ATP (see below) and the second is that pro-inflammatory cytokine is activated by IL17 that induces a Th17 phenotype [45]. Therefore, there are at least three cell subsets (e.g., Treg/Teff/Th17), all deriving from the same progenitor, that play antagonistic effects and determine the thin limit between inflammation and normal status: when Teff/Tregs or Th17/Tregs is high, the signal is to maintain inflammation, and, vice versa, when Tregs are higher than Teff and Th17, inflammation is switched off [46]. Co-stimulatory molecules, CD80 and CD86 (also known as B7-1 and B7-2, respectively), play a key role in this balance by interacting with CD28 that is expressed on both Teff and Tregs and with CTLA-4, a second ligand only present on activated T cell and on Tregs [47