Collapsing glomerulopathy associated with inherited mitochondrial injury

The patient is an 18-month-old boy from Romania, whose parents are healthy and non-consanguineous. In October 2005, periorbital and lower extremities edema were first noted while the child was in general good conditions and without any sign of acute infectious problem. For generalized edema, he was admitted at a pediatric nephrology unit where heavy proteinuria (>10 g day-1), with serum high cholesterol levels (>500 mg per 100 ml) and hypoalbuminemia (<1.5 g per 100 ml) were rapidly discovered. The urine microscopic analysis showed moderate microscopic hematuria, epithelial cells, and granular casts. Serum creatinine was normal. Serology was negative for hepatitis B, hepatitis C, EBV, HIV, and parvovirus B19. Complement profile was normal and tests for autoimmunity (ANA, nDNA, and ANCA) were negative. No focal neurological defects were observed. He was immediately treated with diuretics and infusions of albumin and soon started prednisone 2.3 mg kg-1 day-1. After the first 4 weeks of therapy, his edema increased and he developed anasarca. ACE inhibitors (5 mg per 1.73 m2) were added to the therapy, and a genetic approach considering major inherited diseases was adopted that excluded mutations relative to NPHS1 and NPHS2. Then, a renal biopsy was performed. As the clinical features did not improve, the patient underwent a unilateral nephrectomy and started a peritoneal dialysis program. Part of the clinical history of this patient has been described in detail elsewhere,1 while the findings on renal biopsy and nephrectomy that are crucial for a correct differential diagnosis are described together in the next paragraph. Sections from the needle biopsy (Figure 1) showed a fragment of skeletal muscle as well as renal cortex with 19 glomeruli, some of which revealed collapsing features of glomerular damage. Sections from the nephrectomy (Figure 1) showed a more advanced disease with 25% globally sclerotic glomeruli, 25% of the glomeruli having global collapse, and 50% with segmental collapse. No mesangial sclerosis was identified. Marked podocyte hypertrophy and hyperplasia with either a foamy appearance or large vacuoles, and protein reabsorption droplets were present. In addition, PAS and silver-positive small inclusion in perinuclear location, probably representing accumulation of mitochondria, were seen in podocytes and other cell types. Proximal tubules contained numerous PAS-positive vacuoles as well as optically clear vacuoles. Large microcysts, and patchy interstitial fibrosis and tubular atrophy were noted, and were also more pronounced in the nephrectomy specimen compared with the biopsy, indicating progression of the disease. Mild inflammation of lymphocytes, monocytes, and plasma cells accompanied the areas of fibrosis. Arterioles and arteries were unremarkable. Immunohistochemical analysis was performed with selected antibodies and main results are reported in Figures 2 and 3 and Table 1. Glomeruli revealed segmentally reduced synaptopodin expression and positive staining for Ki-67 in the areas of pseudocrescents formation. Contrariwise, synaptopodin expression was well preserved in the areas where podocytes were markedly hypertrophic but not hyperplastic. Podocin and -dystroglycan expression were also preserved. Podocytes were not dysregulated in their phenotype as the expression of WT-1 was maintained and PAX2 expression, normally regulated by WT-1, was absent in podocyte but positive in parietal cells nuclei. CK 17 staining was very focally and segmentally weakly positive in parietal cells, and podocytes with a perinuclear pattern. CK 7, 8, and 19 were strongly positive, in few hyperplastic and hypertrophic podocytes and in parietal cells (1% of glomeruli). CK 20 was negative (not shown). On ultrastructural analysis, the glomerular capillary walls were wrinkled and folded. Hypertrophic podocytes revealed focal loss of primary processes, extensive foot process effacement, and contained numerous vacuoles. Many of the intracytoplasmic vacuoles appeared to be dismorphic mitochondria, from 0.78 to 1.2 m in diameter, with absent or truncated cristae. In addition, a central double membrane ring was occasionally noted, but no classic 'parking lot inclusions' (paracrystalline intramitochondrial inclusions) were identified. Abnormal mitochondria were also in parietal, endothelial, and mesangial cell cytoplasm in glomeruli and fibroblast, tubular cells, smooth muscle and endothelial cells of the tubulointerstitial compartment. No electron-dense deposits or tubuloreticular inclusions were present. The fragment of skeletal muscle revealed abnormal myocytes with loss of the typical striate appearance and formation of a reddish band at the border of the cells resembling ragged red fibers. On occasion, myocytes appeared filled with vacuolated PAS-positive granules, which also were fuschinophilic on trichrome staining, in the perinuclear areas and in the region underneath the sarcolemma. After the results of the renal biopsy, additional studies on mitochondria functionality and genetic analysis were performed on the patient's DNA. The results of the molecular workup have been described in details elsewhere.1 Briefly, genes involved in Mendelian forms of mitochondrial pathologies were analyzed. First, prenyltransferase-like mitochondrial protein, the human ortholog of the murine gene responsible for a spontaneous mouse model of CG was negative. Instead, the child was found to carry a compound heterozygous mutation in the gene encoding for CoQ2 (c.590G>A (p.Arg197His) and a c.683A>G (p.Asn228Ser), which was associated with markedly decreased CoQ10 levels and complex II+III activities in the renal cortex. The c.590G>A was also heterozygous in the mother, whereas the p.Asn228Ser mutation was also heterozygous in the father and in the healthy brother. Other family members showed no relevant renal symptoms. The final diagnosis was CoQ2-associated CG. At the last follow-up, nearly 2 years after the initial presentation, the patient is on chronic peritoneal dialysis and is growing well. He is currently awaiting a renal transplantation and is treated with oral ubiquinone supplementations. His neurological examination has been checked periodically and has always been found to remain normal. Mitochondrial injury may be due to primary insults or may be the result of secondary events; the former are known as mitochondriopathies or mitochondrial cytopathies. Mitochondriopathies are due to sporadic or spontaneous mutations in the mitochondrial DNA (mtDNA) or nuclear DNA (nDNA), in genes encoding structural or functional mitochondrial proteins. They display extreme heterogeneity, which is both genetic and clinical, and make unpredictable the extent and manifestations of disease presentation.2 These disorders have been considered as neuromuscular diseases, but recently, involvement of other organs has been described.3 The screening for mitochondrial disorders includes the determination of lactate, pyruvate, ketone bodies and their molar ratios in fasted and fed patients, polarographic, and spectrometric studies to evaluate the different enzymatic complexes of the respiratory chain, histologic studies (generally muscle biopsies), and genetic analysis.4 Renal disease may be the first sign of a mitochondrial disorder,5 or it may appear simultaneously with neurological and neuromuscular signs (Table 2). Fanconi's syndrome is particularly frequent in newborns, whereas tubule–interstitial damage is more frequent in children and young adults, and can be associated with focal segmental glomerulosclerosis (FSGS).6 Hereditary FSGS has also been described in rare cases associated with mitochondrial tRNALeu and tRNATyr gene mutation.7, 8, 9, 10, 11, 12, 13 Two genetically transmitted forms of nephrotic syndrome associated with mutations of a gene of the CoQ cascade have lately been reported. The first form of the disease is associated with mutations in the gene COQ2 encoding the para-hydroxybenzoate-polyprenyl-transferase enzyme of the CoQ10 synthesis pathway. Patients with mutations in this gene present with morphologic features of CG or classic FSGS, and it has been suggested that, due to morphologic heterogeneity, the disease should be termed COQ2 nephropathy.1 The other form was described in a child who carried mutations of the PDSS2, gene encoding for decaprenyl diphosphate synthase, which is the first enzyme of the CoQ10 biosynthetic pathway.14 This patient presented with Leigh's syndrome, including severe postnatal hypotonia, early-onset epileptic encephalopathy, and respiratory failure, and developed nephrotic syndrome at 7 months of age.14 Brain magnetic resonance imaging (MRI) showed bilateral symmetric areas of increased T2- and decreased T1-signal intensity in the basal ganglia. mtDNA-associated Leigh syndrome (subacute necrotizing encephalomyelopathy) and NARP (neurogenic muscle weakness, ataxia, and retinitis pigmentosa) are part of a continuum of progressive neurodegenerative disorders observed in members of the same family caused by abnormalities of mitochondrial energy generation. In particular, Leigh syndrome is characterized by an early onset of neurologic symptoms (3–12 months of age) including hypotonia, spasticity, movement disorders, cerebellar ataxia, and peripheral neuropathy. Extraneurologic manifestations may include hypertrophic cardiomyopathy. Eleven mutations in several mitochondrial genes have been associated with the disease (MTATP6, MTTL1, MTTK, MTND1, MTND3, MTND4, MTND5, MTND6, MTCO3, MTTW, and MTTV), and approximately 10–20% of individuals with Leigh syndrome have either the T8993G or T8993C MTATP6 mutation. Interestingly, PDSS2 is the human homolog of prenyltransferase-like mitochondrial protein encoding a gene mapped on mouse chromosome 10, and resulting in a spontaneous mouse model of CG.15, 16 The genetic defect is phenotypically reflected by the presence of numerous dismorphic mitochondria in several cell types, including podocytes. Mice develop exclusively renal symptoms by 8 weeks of life, without any other organ being functionally affected.16 Primary mitochondrial dysfunction and respiratory chain deficiency were also described in two boys and a girl with severe congenital nephrotic syndrome and normal nephrin expression, but none of the patients had collapsing pattern of glomerular injury.17 Secondary mitochondrial abnormalities may also result from a few primary injuries at the level of the slit diaphragm, suggesting a crucial role of these organelles for limiting severe podocyte injury and fatal renal failure.18 Children with congenital nephrotic syndrome of Finnish type, for example, have mitochondria abnormalities, reflected by downregulation of COX I mRNA to 25% of normal kidney values. In addition, transcripts of other proteins encoded by mtDNA are downregulated in kidneys with congenital nephrotic syndrome of Finnish type, whereas mitochondria proteins encoded in the nDNA are generally expressed at comparable levels to normal kidney.18 Secondary mytochondrial dysfunction with reduction of the respiratory chain enzymatic activity, oxygen consumption, and mtDNA copy number have also been described in experimental FSGS induced by puromycin.19 Whether functional defects associated with mitochondrial abnormalities (such as neurological, and muscle problems or functional defects of the proximal renal tubule) are widespread in several organs or restricted to one cell type, it is unpredictable in which cell type and to which degree a functional damage will manifest. In cases where the mutation is affecting mtDNA, heteroplasmy may explain this phenomenon, but this is not applicable to cases where the mutation is affecting the nDNA. Cells most affected by mitochondriopathies are those with little post-birth mitotic activity, such as podocytes or neuronal cells. Threshold expression may also affect the cell phenotype.23 Much has to be learned in the field of mitochondrial defects and podocytes damage. It is reasonable to think that, in the presence of mitochondrial abnormalities in podocytes, cell death will occur and, according to the cell depletion hypothesis, segmental sclerosis may then develop.24 Different from expectation, in CoQ2-associated CG, proliferation and glomerular collapse, rather then podocyte death and glomerulosclerosis, were the pathologic findings. The mechanism by which podocytes react to mitochondria malfunctioning with proliferation rather than cell death is not fully clear. It has been suggested that ion channels in the inner membranes of mitochondria have a potential role in redox regulation and have a central role in the decision between life and death in ischemic myocardial injury. Reduction of the cellular redox state may, on one hand, induce programmed cell death, and, on the other hand, activate transcription factor hypoxia-inducible factor 1 with consequent increase expression of adaptive enzymes, resulting in protection of the cells from death.25, 26 The recent demonstration of hypoxia-inducible factor 1 potency to modulate podocyte phenotype and induction of proliferation is in support of these observations.27, 28, 29 It is possible that environmental factors may also be involved in the manifestation and progression of the disease phenotype through mechanisms involving the redox state of these organelles. To support this hypothesis is the evidence that germfree conditions markedly reduce the appearance of mitochondrial phenotype in the kd/kd mouse.30 In conclusion, although CG associated with mutation in a nuclear gene encoding mitochondrial proteins seems to be a rare condition, evaluation of mitochondrial morphology is an important screening in patients with unexplained proteinuria. The described findings also stress the importance of accurate clinical–pathologic correlation and call for complete genetic analyses to avoid unnecessary interventions in patients. The authors state no conflict of interest.