Medicine (RMH) - Theses

Permanent URI for this collection

http://hdl.handle.net/11343/194

Filters

2022 1
1
Reset filters

Settings
Statistics
Citations
Subject: genetics × Author: Harraka, Philip Adam ×

Search Results

Now showing 1 - 1 of 1
  • Item
    No Preview Available
    Retinal drusen and genetics in glomerulonephritis
    Harraka, Philip Adam ( 2022)
    Glomerulonephritis comprises a heterogenous group of diseases that represent a common cause of impaired kidney function. Many are immune-mediated with antibodies directed against various self-antigens that form immune complexes and activate complement through the classical, alternative or lectin-binding pathways. Complement deposits are found in IgA, membranous and lupus nephropathy, anti-GBM disease, Focal and Segmental Glomerulosclerosis (FSGS) and fibrillary glomerulonephritis, and often correlate with more severe disease and a worse prognosis. The evidence for complement activation in glomerulonephritis includes glomerular complement deposits, low plasma complement levels in active disease, milder disease where complement is absent from the deposits, mild disease in animal models where serum complement levels are depleted, and the association with genetic complement variants in different forms of glomerular disease. There are also case series and reports of retinal drusen in C3 glomerulopathy, systemic lupus erythematosus, IgA and some other forms of glomerulonephritis. Retinal drusen are white-yellow deposits of oxidised lipids, immunoglobulins, complement and amyloid in Bruch’s membrane that are characteristic of age-related macular degeneration. The pathogenesis of drusen is well described with 37 genetic loci that affect multiple pathways including the complement system. Drusen have a similar composition to the glomerular immune deposits and their occurrence in glomerulonephritis is further evidence for complement activation in these conditions but also suggests that drusen represent a model for glomerular immune deposition and a biomarker for disease activity. Hence, studying drusen in glomerulonephritis may help identify further shared pathways in the pathogenesis of glomerular disease. The pathogenesis of immune-mediated glomerulonephritis is incompletely understood and treatment is still not optimal. Complement activation may represent a common mechanism of disease that can be targeted by novel therapies. Chapter 2 of this thesis examined 122 individuals with specialist-diagnosed IgA glomerulonephritis for retinal drusen from the renal clinics at two teaching hospitals. Drusen were counted using a grid overlay by two trained graders and more than ten central drusen in either eye was considered abnormal. Central drusen number was higher and abnormal drusen counts (>10) were more common in individuals with IgA nephropathy (9 +/- 27; 23, 19%) than the hospital controls (2 +/- 7; 8, 7%) (p<0.001, p=0.006). Drusen in IgA disease were more common with longer disease duration (p=0.03) and larger with stronger mesangial IgA staining (p=0.004). These findings suggest a shared pathogenesis between drusen and IgA glomerulonephritis. Chapter 3 of this thesis also examined drusen in other forms of glomerulonephritis some of which were associated with glomerular immune deposits and others which were not. These included membranous (n=8) and anti-glomerular basement membrane (GBM) glomerulonephritis (n=6) which are associated with antibodies against glomerular antigens that are also expressed in the retina; FSGS (n=49), which is associated with glomerular IgM and complement deposits; and ANCA-mediated vasculitis (n=15) and minimal change glomerulonephritis (n=6) where immunoglobulin and complement deposits are sparse. Individuals with glomerulonephritis and immune deposits (membranous nephropathy and anti-GBM disease) were compared with participants who had glomerulonephritis without immune deposits (pauci-immune ANCA-associated vasculitis and minimal change disease) and with healthy matched hospital controls. Mean central drusen number was 21 +/- 45 for individuals with immune deposits and four (29%) had abnormal drusen counts compared with a mean of 2 +/- 6 for those without immune deposits and one (5%) with abnormal counts (p=0.28, p=0.13). Membranous and anti-GBM glomerulonephritis are likely associated with drusen because the target glomerular antigens are also expressed in the retina. However, drusen were much less common in glomerulonephritis without immune deposits despite complement activation occurring in ANCA-associated vasculitis. Mean central drusen number was increased in FSGS (9 +/- 25) compared with hospital controls (3 +/- 8) (p=0.02). Abnormal drusen counts affected nine individuals with FSGS (19%) and four controls (8%) (p=0.23). Drusen were larger in FSGS (20, 41%; 10, 20%) (p=0.048). Abnormal drusen counts were seen in secondary (2/6, 33%), viral-associated (2/6, 33%) and genetic FSGS (3/10, 30%) and therefore were not specific to any aetiology. Previous studies have suggested that an unknown neo-epitope is exposed in FSGS following glomerular injury which, if present in the retina, might have contributed to drusen formation through IgM deposition and classical complement pathway activation. Thus, treatments targeting complement may still be beneficial for FSGS and other forms of glomerulonephritis regardless of the nature of the initial podocyte injury. Chapter 4 of this thesis investigated 8 families with IgA nephropathy for a causative pathogenic variant in a gene from one of five candidate lists: for IgA disease (n=102), the complement system (n=56), drusen pathogenesis (n=46), Alport syndrome (COL4A3–COL4A5) and FSGS (n=47). In addition, all variants were investigated for any pathogenic change that segregated with disease within the family and explained the phenotype. Chapter 4 investigated eight unrelated individuals with kidney-biopsy-proven IgA nephropathy and at least one other affected family member. Three of the index cases in these families had drusen (38%). Variants were prioritised based on the autosomal dominant (AD) inheritance model in all but one family who had X-linked disease. The index cases and other family members (median n=4, range n=1 to n=12) underwent Whole Exome Sequencing (Illumina NovaSeq platform at the Australian Genome Research Facility or Otogenetics) followed by variant calling and annotation using the Melbourne Bioinformatics pipeline using the Spartan supercomputer to identify predicted pathogenic variants. Variants were considered Pathogenic or Likely Pathogenic based on the ACMG criteria. Variants were examined in genes from the candidate lists and assessed for a CADD score >10 and a gnomAD minor allele frequency (MAF) <0.05, using the in-silico pathogenicity tools Polyphen2 (damaging), SIFT (deleterious) and MutationTaster (disease causing); for conservation in mice and birds; and for previous reports of pathogenicity in ClinVar. Of the 8 families, two had Gly substitutions in a collagen IV gene that segregated with disease (25%). These were p.Gly624Asp in COL4A5 (Family 10) and p.Gly395Glu in COL4A3 (Family 2). The p.Gly624Asp variant is a known hypomorphic variant (PP5, PM1, PM5, PP2, PP3) that results in X-linked Alport syndrome in later life. The p.Gly395Glu variant (PP5, PM1, PM2, PP2, PP3) results in a thinned GBM in AD Alport syndrome but not necessarily kidney failure. The association of mesangial IgA deposits and Alport syndrome is well-recognised and probably occurs because the thinned GBM predisposes to mesangial IgA deposition. Family 7 with three affected males had a Likely Pathogenic missense variant p.Pro437Leu (PM2, PP2, PP3, PP5) in HNF1B which segregated with the disease. This gene is associated with Renal Cysts And Diabetes syndrome suggesting that genes causing other kidney diseases may explain other familial forms of IgA glomerulonephritis. In this family two of the three affected individuals underwent retinal imaging and both had multiple retinal drusen. The index case in Family 1 had IgA glomerulonephritis and drusen. He was demonstrated to have a pathogenic nonsense variant in C9 (p.Cys54Ter) (PVS1, PP5, PM2, PP3). C9 is a late component of the complement pathway. This variant was found in four affected individuals and one unaffected family member consistent with incomplete penetrance. This variant causes partial C9 deficiency and predisposes to severe infections. C3, C4 and C9 deficiency have been described in IgA glomerulonephritis. This finding contrasts with the observation that complement activation is increased in IgA disease and suggests that the complement deficiency predisposes to more severe infection and increased IgA production. Hence, as for lupus nephritis, IgA disease in some may involve a genetic complement deficiency. These studies suggest that familial IgA nephropathy may be caused by IgA deposits which occur in 16% of the otherwise normal population being found together with another genetic kidney disease, as for the COL4A3–COL4A5 genes, or with genes that predispose to autoimmune disease, such as the Complement deficiencies. No pathogenic variants were found in the candidate list for FSGS or in the IgA list in this cohort. Chapter 5 investigated the genetic aetiology of fibrillary glomerulonephritis in a young woman from a consanguineous family with retinal drusen. The fibrils were confirmed to comprise DNAJB9 and were intermediate in diameter between those seen in renal amyloid and immunotactoid glomerulopathy. Fibrillary glomerulonephritis is typically a disease of middle-aged-elderly women who go on to develop kidney failure within two years. It is usually not familial but there are occasional reports of families where the disease occurs in successive generations and inheritance is presumed to be AD. The only gene implicated to date is MEFV in an individual who also had Familial Mediterranean Fever. The pathogenesis is presumed to be similar to that of amyloid because of the physical resemblance of their fibrils. The index case with fibrillary glomerulonephritis was investigated for a homozygous pathogenic variant that was absent from her three normal brothers. DNA was examined from all siblings and both parents. Two strategies were used to identify the pathogenic variant. One was to identify homozygous variants present only in the affected individual and not in her unaffected brothers; the second was to identify homozygous variants present in the affected individual and one other sibling, assuming the variant was of reduced penetrance. The same strategy was used as described above for IgA glomerulonephritis. The first strategy identified 31 homozygous variants but only two were Pathogenic or Likely Pathogenic. The second strategy identified seven homozygous variants but none was Pathogenic. A candidate variant was identified and may have influenced the deposition of DNAJB9 in the kidney. Thus, this thesis has described retinal drusen in different forms of glomerulonephritis. It found that drusen occurred commonly in IgA glomerulonephritis and were found more often with longer disease duration and were larger with stronger mesangial IgA staining. It also found that drusen occurred commonly in other forms of glomerulonephritis even when it appeared to be secondary to structural kidney damage (FSGS). Since drusen in macular degeneration depend on complement activation these observations suggest that treatments targeting complement may be useful in different forms of glomerulonephritis. These studies suggest that familial IgA glomerulonephritis occurs commonly because IgA is deposited in other genetic kidney diseases. The results also demonstrate overlap between the different kidney diseases and that it is not sufficient to examine a family with genetic kidney disease for variants in only one gene panel. Interestingly pathogenic variants in the complement pathway and drusen susceptibility genes did not appear to be common in these families. Finally, a new gene candidate was identified in autosomal recessive fibrillary glomerulonephritis with drusen. Future studies of drusen in glomerulonephritis should use a more sensitive method of drusen detection and correlate drusen with disease activity on kidney biopsy. Drusen may represent a biomarker for glomerulonephritis and in predicting the likelihood of kidney failure or response to treatment. Complement-targeted therapies should be investigated for efficacy for these forms of glomerulonephritis. More families with IgA glomerulonephritis should be investigated for pathogenic variants in genes previously described in genetic kidney disease. The role of COL4A3–COL4A5 variants in predisposing to IgA nephropathy needs to be determined. The association with complement deficiency, particularly C9, needs to be investigated. While complement pathway defects were not common, they may contribute to disease in a subset of individuals with IgA nephropathy. Lastly, more individuals with hereditary fibrillary glomerulonephritis need to be investigated for a genetic variant, particularly children or others with familial disease. Laboratory studies are needed to confirm the role of our gene candidate in the deposition of DNAJB9.