ER stress and basement membrane defects combine to cause glomerular and tubular renal disease resulting from Col4a1 mutations in mice

Abstract

Collagen IV is a major component of basement membranes, and mutations in COL4A1, which encodes collagen IV alpha chain 1, cause a multisystemic disease encompassing cerebrovascular, eye and kidney defects. However, COL4A1 renal disease remains poorly characterized and its pathomolecular mechanisms are unknown. We show that Col4a1 mutations in mice cause hypotension and renal disease, including proteinuria and defects in Bowman's capsule and the glomerular basement membrane, indicating a role for Col4a1 in glomerular filtration. Impaired sodium reabsorption in the loop of Henle and distal nephron despite elevated aldosterone levels indicates that tubular defects contribute to the hypotension, highlighting a novel role for the basement membrane in vascular homeostasis by modulation of the tubular response to aldosterone. Col4a1 mutations also cause diabetes insipidus, whereby the tubular defects lead to polyuria associated with medullary atrophy and a subsequent reduction in the ability to upregulate aquaporin 2 and concentrate urine. Moreover, haematuria, haemorrhage and vascular basement membrane defects confirm an important vascular component. Interestingly, although structural and compositional basement membrane defects occurred in the glomerulus and Bowman's capsule, no tubular basement membrane defects were detected. By contrast, medullary atrophy was associated with chronic ER stress, providing evidence for cell-type-dependent molecular mechanisms of Col4a1 mutations. These data show that both basement membrane defects and ER stress contribute to Col4a1 renal disease, which has important implications for the development of treatment strategies for collagenopathies.

Keywords: Basement membrane; Collagen IV; Endoplasmic reticulum stress; Extracellular matrix; Kidney disease.

Fig. 1.

Analysis of renal function. (A-D) In vivo renal function analysis of 1-, 3- to 4- and 6- to 8-month-old Col4a1^+/Raw (black bars) and WT (white bars) mice. (A) Reduced mean arterial blood pressure in Col4a1^+/Raw animals at all ages. (B) Reduced sodium excretion in Col4a1^+/Raw animals at all ages. (C) Inulin clearance assays uncovered reduced glomerular filtration rate per gram body weight at 3 months (∼40% reduction), but no further decline with age. (D) Measurement of sodium excretion in the presence of the diuretics furosemide and thiazide remains reduced in mutant animals. Blockade of ENaC by amiloride abolished the difference between WT and mutant mice. (E) Western blotting showed ∼16-fold increased levels of total NHE3 (NHE3) in Col4a1^+/Svc (SVC) mice but unaltered phosphorylated NHE3 (p-NHE3). Representative band of total protein stain is given as loading control (entire gel is provided in Fig. S6). (F) ImageJ densitometry analysis of total and p-NHE3. *P<0.05, **P<0.01, ***P<0.001. n=5-7 in A-E; n=3-5 in F.

Fig. 2.

Diabetes insipidus in Col4a1 mutant mice. (A) Metabolic cage study revealed ∼1.8-fold increased 24 h (24H) urine production in Col4a1^+/Raw (RAW) mice per gram body weight (g BW; n=6). (B) In vivo urine flow was ∼1.7-fold increased in 3- to 4-month-old mice. Older mutant mice displayed similar trends. (n=4-6). (C) Metabolic cage study revealed increased daily (24 h) urine production in 4- to 5-month-old Col4a1^+/Svc (SVC) mice per gram body weight (n=4-6). (D) Metabolic cage study indicates ∼4.2-fold increased daily water consumption in Col4a1^+/Svc mice (WT, n=6; Col4a1^+/Svc, n=10). (E) Col4a1^+/Raw mice develop a reduced urine osmolality at 3 months, which becomes more severe with age. Urine osmolality of 3- to 4-month-old Col4a1^+/Svc mice is provided in Fig. S1D (n=5-7). (F) Increased urine production in 2.5-month-old Col4a1^+/Svc mice compared with controls (n=3). (G) Effect of water deprivation (WD) on urine production in wild-type and Col4a1^+/Svc mice (n=4-5). (H) Percentage weight loss following 24 h water deprivation in 2.5- (2.5M) and 4- to 5-month-old animals (2.5M, n=3; 4-5 months old, n=4-5). (I) Urine production after vasopressin injection. (J) Analysis of Aqp2 protein levels in 4- to 6-month-old Col4a1^+/Svc mice. Representative band of total protein stain is given as loading control (entire gel is provided in Fig. S6) (n=3) with densitometry analysis using ImageJ. *P<0.05, **P<0.01, ***P<0.001.

Fig. 3.

Erythropoietin treatment of Col4a1 mutant mice. (A) Four-month-old Col4a1^+/Svc (black bars) and WT (white bars) mice were treated with Epo, and haematological analysis was performed before (−Epo) and afer Epo (+Epo) treatment, which showed that treatment rescued haematocrit values in Col4a1^+/Svc mice (WT 44% pretreatment, 44% post-treatment; Col4a1^+/Svc 36% pre-treatment, 49% post-treatment; n=4-5). (B) Tail-cuff plethysmography revealed a rescue of blood pressure in Epo-treated mice (WT 122 mmHg pre-treatment, 114 mmHg post-treatment; Col4a1^+/Svc 97 mmHg pre-treatment, 110 mmHg post-treatment; n=4-5). *P<0.05, **P<0.01, ***P<0.001, Student's unpaired t-test.

Fig. 4.

Analysis of kidney histopathology. (A) Bowman's capsule defects in 3- to 4-month-old Col4a1^+/Raw (RAW) and Col4a1^+/Svc (SVC) mice include thickening of the capsule (arrowhead) and formation of multiple cell layers of parietal epithelial cells, which can develop a more cuboid appearance. (B) Haematoxylin and Eosin staining revealed atrophy of the medulla (indicated by black box) in 8-month-old Col4a1^+/Raw and Col4a1^+/Svc mice that develops in adulthood (see Table S3). (C) Protein cast (arrowhead) in kidney sections of Col4a1^+/Raw and Col4a1^+/Svc mice (n=3, 6 months old). (D) Perls' Prussian Blue staining of renal sections identified haem deposits (arrowhead) in 4-month-old Col4a1^+/Svc mice as a sign of haemorrhage. Perls' Prussian Blue staining does not detect intact red blood cells (n=3; scale bars: 20 µm). (E) Semi-quantitative analysis of urine samples using a dipstick revealed haematuria in mutant mice. Absence of haematuria was set as a value of 1 (see Fig. S4C for haematuria scale; n=6). ***P<0.001.

Fig. 5.

Investigation of mechanisms underlying Col4a1 kidney disease. (A) Transmission electron microscopic analysis revealed irregular thickening of the BM of 4-month-old Col4a1^+/Svc Bowman's capsule (indicated by black arrowhead; white arrowhead indicates thinner region). The presence of swollen ER vesicles can also be observed (red arrow). (B) The vascular BM in 4-month-old Col4a1^+/Svc mice (SVC) displayed focal interruptions (right black arrow) and an irregular less dense appearance with BM fragments (left arrow). (C) The GBM in 4-month-old Col4a1^+/Svc mice showed irregular thickening (black arrow), but no interruptions were detected. Scale bars: 2 µm in A-C (n=3). (D) Immunostaining against Col4a1 and perlecan in 4-month-old Col4a1^+/Svc kidney sections followed by confocal microscopy revealed reduced deposition of Col4a1 and perlecan in BMs of Col4a1^+/Svc mice (n=3). Scale bar: 20 μm. (E) ImageJ analysis of fluorescence staining in D. (F) Immunostaining against Col4a3 (green) and podocin (red) revealed deposition in 3- to 4-month-old mutant BM of Bowmans capsule (white arrows). Scale bar: 20 µm. (G) Fraction of Bowman's capsules that stain positive (black) and negative (white) for Col4a3 in WT (0% positive) and Col4a1^+/Svc mice (32% positive; n=85 glomeruli across three animals, Fisher's exact test). (H) ImageJ analysis of podocin fluorescence staining in H revealed ∼9.8-fold increase in SVC. *P<0.05, ***P<0.001, Student's unpaired t-test in E,H; Fisher's exact test in G.

Fig. 6.

ER stress and renal disease. (A) Significantly increased protein levels of ER stress markers Bip (∼2.9-fold increase) and Atf4 (∼4.6-fold increase) in 4-month-old Col4a1^+/Svc mice (n=3). Representative band of total protein stain is given as loading control (entire gel is provided in Fig. S6). (B) Densitometric analysis of Bip and Atf4 using ImageJ. (C) Examination of protein levels of Bip and CHOP analysis across Col4a1^+/Raw animals of different ages (40 days, 3-4, 8 and 12 months). Increased protein levels of Bip in Col4a1^+/Raw animals precedes elevated levels of CHOP, which is a marker for chronic ER-stress-induced apoptosis. Activation of CHOP coincides with development of atrophy of the medulla (see Table S3; n=3). (D) Col4a1^+/Svc mice display thinning of the BM of Bowman's capsule at 40 days of age. Scale bar: 2 μm. (E) Average BM thickness as measured by ImageJ (WT ∼257 nm; Col4a1^+/Svc ∼121 nm; n=3). (F) Distended ER vesicle (black star) in epithelial cell of Bowman's capsule in 40-day-old Col4a1^+/Svc. Scale bar: 2 μm. (G) Analysis of Bip protein levels in 40-day-old Col4a1^+/Svc mice (n=3). Representative band of total protein stain is given as loading control (entire gel is provided in Fig. S6). (H) Densitometric analysis of G. (I) Immunostaining against Bip (red) and Col4a1 (green) on kidneys of 3-month-old mice revealed increased expression in epithelial cell of Bowman's capsule (white arrow), and apparent more intense staining in Col4a1^+/Svc. Scale bar: 20 µm (n=3).