Pathological similarities between low birth weight-related nephropathy and nephropathy associated with mitochondrial cytopathy

From January 2006 to December 2011, we performed 472 kidney biopsies in our division. In these cases, there were four cases of FSGS among patients born with a birth weight under 2,500 g (according to the WHO definition of LBW). These four patients (LBW 1-4) were retrospectively evaluated pathologically and genetically. In addition, two patients (Mt 1 and 2) who were found to have mitochondrial DNA (mtDNA) mutations and underwent kidney biopsies to investigate the cause of their proteinuria were also evaluated. As normal controls for staining (N = 3), kidneys dissected because of kidney cancer were used.

Briefly, kidney specimens fixed in 10% neutral buffer formalin followed by embedding in paraffin were used for a light microscopy analysis with routine staining, as follows: hematoxylin and eosin (HE), periodic acid-Schiff (PAS), Masson's trichrome stain and periodic acid-methenamine-silver (PAM)-HE stain. The kidney specimens used for the electron microscopy analysis were fixed in 2% glutaraldehyde (pH 7.3-7.4) followed by 2% osmium tetroxide (pH 7.3-7.4). For cytochrome c oxidase subunit IV (COX IV) staining, paraffin-embedded specimens were used. Following deparaffinization and antigen activation using 10 mM of boiled citrate buffer (pH 6.0), rabbit anti-COX-IV antibodies (clone 3E11) (Cell Signaling Technology, Inc., Danvers, MA) were used as the primary antibody. SignalStain' Boost IHC Detection Reagent (HRP, Rabbit) (Cell Signaling Technology, Inc.) was used to detect the rabbit primary antibody, according to the manufacturer's instructions.

Blood samples (1 ml/each analysis) were collected in EDTA-2Na tubes. Following centrifugation, the pellets were used for the analysis. For the analysis of urine sediments, over 100 ml of urine was collected from each patient, and the pellets obtained after centrifugation were used for the analysis. Five slices (4 μm) of frozen kidney sections were used for the genetic analysis. A PCR-Luminex assay, which can be used to survey 61 different pathogenic mtDNA mutations, was performed according to our previously reported method [8].

This study was conducted in accordance with the 'Ethical Guidelines for Clinical Studies' (Revised on December 28, 2004, Ministry of Health, Labour and Welfare of Japan). All medical professionals involved in this study were required to comply with these ethical standards. The local Ethics Committee of Chiba-East Hospital approved the study protocol (No. 22), and all subjects provided their informed consent to participate in this study.

The background characteristics of the patients are summarized in Table 1. All four LBWN patients underwent kidney biopsies in order to determine the etiology of their persistent proteinuria. All four LBWN patients were male, with a mean birth weight of 1,692 g. The gestational ages and the reasons for the low birth weight could not be fully assessed in these cases (The gestational age of LBW1 was 28 weeks and that of LBW2 was 38 weeks). No subjects had symptoms of myopathy or encephalopathy, a history of stroke-like episodes or difficulty hearing. Although all patients had no subjective symptoms, they were found to have proteinuria on periodic medical examinations. Only one (LBW 3) of the LBWN patients suffered from diabetes mellitus (with a five-year history). The kidney size tended to be smaller than normal, except in LBW 3 (Table 1). The light microscopy analysis revealed that the glomeruli of all LBWN patients exhibited FSGS (Figure 1). Patient Mt 1 was initially found to have proteinuria at 39 weeks of gestation and was subsequently referred to our hospital because the proteinuria persisted for one year after delivery. Although she had no symptoms of MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis, stroke-like episodes), we found increased morphologically abnormal mitochondria in her podocytes (Figure 2A). Therefore, we evaluated her mtDNA at our hospital and detected an mtDNA mutation (3243 A > G). Patient Mt 2 was found to have proteinuria on a periodic medical checkup at 19 years of age. She also had no symptoms of MELAS. She refused an assessment of her serum lactate level. However, she had a brother with MELAS (3243 A > G), and a kidney biopsy revealed that her podocytes included increased mitochondria (data not shown). Her mtDNA mutation (3243 A > G) was diagnosed at another hospital. Although the characteristic glomerular change in patients with mitochondrial cytopathy is FSGS, as previously described in several reports [9]-[11], neither of our two patients with mtDNA mutations had apparent FSGS lesions in the evaluated glomeruli (data not shown). An electron microscopy analysis of the two patients with mtDNA mutations revealed that the glomerular epithelial cells contained an increased number of enlarged mitochondria and mild foot process effacement. On the other hand, we found no morphological abnormalities or increments in the number of mitochondria in any of the observed sections obtained from the LBWN patients (Figure 2B-D). However, the podocytes of the four LBWN patients showed mild foot process effacement (Figure 2B-D), similar to that observed in the patients with mtDNA mutations (Figure 2A).

The presence of granular swollen epithelial cells (GSECs) in the collecting ducts or distal tubules was recently reported to be a characteristic pathological change in patients with mitochondria cytopathy [14]. All four patients with LBWN as well as the two patients with mtDNA mutations had GSECs in their collecting ducts and/or distal tubules (Figure 3A-F). In addition, a portion of these GSECs had dropped out of the arrangement of the tubules. The electron microscopy analysis revealed that these GSECs contained an increased number of enlarged mitochondria (Figure 3G, H), and the nuclei occasionally seemed to be condensed.

Complex IV, which is composed of 13 subunits, is the terminal of the electron transfer chain in mitochondria. It was reported that the complex IV activity of podocytes changes in pathological conditions [15]. Therefore, we here stained for COX IV, which is one of the subunits. A COX IV expression analysis of normal controls showed that podocytes should be almost equally positive in the glomerulus (Figure 4A) and that tubular cells are also equally positive at the same luminal level (Figure 4B). However, in the LBWN patients, the expression of COX IV in the podocytes was unbalanced. Only a few podocytes appeared to intensely express COX IV, although the others expressed less COX IV (Figure 4C, E). In addition, the COX IV staining in a part of tubules of the LBWN patients exhibited a mosaic pattern, even at the same luminal level (Figure 4D,F). The cells strongly expressing COX IV appeared to correspond to GSECs. The expression patterns both of the glomerular podocytes and tubules in the patients with mtDNA mutations showed a mosaic pattern (Figure 4G, H), similar to that observed in the LBWN patients. The degree of the mosaic patterns in glomeruli and tubular cells should be more intense in patients with mtDNA mutations compared with LBWN patients.

According to the PCR-Luminex assay, none of the 61 mtDNA mutations were detected in the blood, urine or kidney sections of the four LBWN patients (Table 1). However, an mtDNA mutation (3243 A > G) was clearly detected in all samples of blood, urine sediment and kidney sections in patient Mit 1. Because another patient (Mit 2) had already been found to have 3243 A > G mtDNA at another hospital, we did not perform an mtDNA mutation analysis using RCR-Luminex due to ethical considerations.