The advantage of this study was measuring both total adiponectin levels and the proportions of adiponectin isoforms, which had not been previously investigated in pediatric patients with NS. In this study, total adiponectin level increased significantly in pediatric patients with NS at the onset and initial remission of the disease compared with that in control subjects. We also observed increases in the levels of HMW, MMW, and %MMW as well as a decrease in %LMW at the onset of NS. These findings suggest that the total adiponectin level and the proportions of the adiponectin isoforms change in NS.
The finding of significantly increased adiponectin levels in patients with NS is consistent with that of previous reports [12, 13]. Zoccali et al. reported that the levels of serum adiponectin levels significantly increased in adult patients with NS secondary to different etiologic causes and were strongly related to the degrees of proteinuria [12]. Similarly, Bakkaloglu et al. reported that serum adiponectin levels remarkably increased in steroid-sensitive NS relapse compared with those in steroid-sensitive NS remission in pediatric patients [13]. The increase in adiponectin level parallels the progression of CKD, and high adiponectin levels are found in patients with ESRDs [21, 22]. In patients with CKD, uremia may reduce adiponectin clearance due to renal dysfunction and lead to impaired biodegradation and abolition [9]. However, adiponectin levels increase in patients with NS despite normal renal function. Therefore, uremic conditions and reduced clearance are unlikely to be associated with the increased adiponectin in NS.
The mechanism for elevated serum adiponectin levels and the direct effects of adiponectin on podocytes in the nephrotic state in NS remain unknown. However, increased serum adiponectin levels were unlikely to be associated with renal dysfunction because most patients with NS maintain renal functions. In urine, adiponectin exists largely as a trimer (LMW), which has a similar or lesser molecular weight than that of albumin [23–25]. Increased glomerular permeability due to podocyte dysfunction can cause hyper-filtration of adiponectin into urine. Hence, the changes in the distribution of adiponectin isoforms observed in our study may be associated with LMW leakage from the kidney. Recent evidence in mice suggests that the adiponectin regulates albuminuria and exerts a protective effect on the podocytes [8]. The adiponectin receptor (AdipoR1) is expressed at a high level in the podocytes and adiponectin signaling via adenosine monophosphate-activated protein kinase (AMPK) regulates NADPH oxidase 4 (Nox4) expression, which is linked to oxidative stress, the fusion of podocyte foot processes, and albuminuria [8]. These findings indicate a possibility that the adiponectin is upregulated in NS to attenuate podocyte damages. Zoccali et al. speculated an alternative possibility that the depletion of albumin and protein in NS triggers metabolic changes that may lead to hyperlipidemia and hyperfibrinogenemia in the liver as well as a parallel increase in the synthesis of adiponectin in fat cells [12]. This adaptation would be appropriate to limit the acute inflammatory phase reactions associated with NS [12]. In our study, three control patients had hypoalbuminemia (< 2.5 g/dL) without proteinuria or adiponectin elevation, which indicates that the prominent proteinuria, rather than the hypoalbuminemia was the trigger for the elevating adiponectin levels. Otherwise, elevated adiponectin levels in NS may be caused by an abnormal response to adiponectin in the kidney due to reduced AdipoR1 density, or signal coupling, or the production of an aberrant form of the molecule, as Sethna et al. implied [11]. It is difficult to know whether the increase in adiponectin levels occurs before the emergence of proteinuria, and thus, further investigations including animal studies are required to clarify the mechanisms of adiponectin upregulation and the its effects on the podocyte in NS.
There are few reports available on adiponectin isoform levels. Nishimura et al. reported that total serum adiponectin, HMW, MMW, and LMW levels are 5.5–6.4 µg/mL, 2.2–3.1 µg/mL, 1.6–1.8 µg/mL, and 1.3–1.7 µg/mL, respectively, in school-age children [26]. Compared with those values, some of our patients with NS had higher levels for all adiponectin isoforms. Regarding the association between the adiponectin isoforms and diseases, the HMW levels may serve as a predictor of future cardiovascular events in patients with coronary artery disease, and the HMW-to-total adiponectin ratio may be positively associated with aortic stiffness in patients undergoing hemodialysis [1]. In addition, physiological activities, such as vascular protective activities and the insulin sensitivity, are also affected by HMW or %HMW [19]. Pajvani et al. reported that the HMW-to-total serum adiponectin ratio is useful for monitoring the improvement of insulin sensitivity in response to thiazolidiones in cases of type 2 diabetes [27]. Goto et al. reported that decreases in LMW, HMW, and total adiponectin levels are associated with diabetes [28]. To the best of our knowledge, the changes in the proportion of adiponectin isoforms in NS have not been investigated previously. In our study, both HMW and the %HMW increased in patients with NS, and thus, both HMW and %HMW may promote the protection of podocytes.
This study has some limitations. First, all patients with NS were treated with steroids, and steroid treatment could affect adiponectin levels. Fallo et al. reported that the glucocorticoids decreased plasma adiponectin levels in males, as shown in both healthy subjects receiving acute exogenous administration and patients with chronic endogenous hypercortisolism [29]. By contrast, steroids have an inhibitory effect on TNF-α, which decreases adiponectin level [30]. Uchida et al. reported that adiponectin levels were significantly augmented after glucocorticoid pulse therapy in patients with IgA nephropathy [31]. Furthermore, some reports have shown that steroids have no effect on the adiponectin level [32, 33]. As mentioned above, the effects of steroid treatment on adiponectin are somewhat controversial, but many of studies suggest that steroid treatment increases adiponectin expression [34]. We emphasize that high serum adiponectin levels at the onset were under untreated condition and that all patients with NS were treated with steroids. Second, urine adiponectin levels were not measured in our study. The distribution of adiponectin isoforms in blood may differ from that in urine. In an experimental model, it is reported that the glucocorticoid-mediated tightening can reduce the flux of adiponectin across endothelial monolayers, possibly owing to alterations in the expression profiles of the tight junction proteins [35]. Therefore, the amount of each adiponectin isoform leaked from the kidney may be altered by the steroid treatment. Finally, only a small number of participants were enrolled in this study. Larger-scale studies are needed to further delineate the mechanisms between proteinuria and adiponectin complexes.
In conclusion, elevated serum adiponectin levels in patients with NS are believed to be secondary to proteinuria, but the mechanism remains unknown. This study suggests that total adiponectin levels and proportions of adiponectin isoforms change in the course of NS. A regulatory mechanism may exist between adiponectin and proteinuria.