4-Sodium phenyl butyric acid has both efficacy and counter-indicative effects in the treatment of Col4a1 disease

Abstract

Mutations in the collagen genes COL4A1 and COL4A2 cause Mendelian eye, kidney and cerebrovascular disease including intracerebral haemorrhage (ICH), and common collagen IV variants are a risk factor for sporadic ICH. COL4A1 and COL4A2 mutations cause endoplasmic reticulum (ER) stress and basement membrane (BM) defects, and recent data suggest an association of ER stress with ICH due to a COL4A2 mutation. However, the potential of ER stress as a therapeutic target for the multi-systemic COL4A1 pathologies remains unclear. We performed a preventative oral treatment of Col4a1 mutant mice with the chemical chaperone phenyl butyric acid (PBA), which reduced adult ICH. Importantly, treatment of adult mice with the established disease also reduced ICH. However, PBA treatment did not alter eye and kidney defects, establishing tissue-specific outcomes of targeting Col4a1-derived ER stress, and therefore this treatment may not be applicable for patients with eye and renal disease. While PBA treatment reduced ER stress and increased collagen IV incorporation into BMs, the persistence of defects in BM structure and reduced ability of the BM to withstand mechanical stress indicate that PBA may be counter-indicative for pathologies caused by matrix defects. These data establish that treatment for COL4A1 disease requires a multipronged treatment approach that restores both ER homeostasis and matrix defects. Alleviating ER stress is a valid therapeutic target for preventing and treating established adult ICH, but collagen IV patients will require stratification based on their clinical presentation and mechanism of their mutations.

Figure 1

PBA reduces ICH. (A) Overview of preventative PBA treatment from conception to point of analysis. (B) MRI image of untreated and treated small with vacuolar cataracts (SVC) mice showing ICH (red arrow). (C) Image analysis of MRI data reveals reduced ICH bleed volume (left graph) but not ICH number of bleeds (right graph) (WT n = 6, SVC n = 10, SVC PBA n = 6). (D) Perl’s staining of brains from 5-month-old untreated Col4a1^+/SVC and Col4a1^+/SVC treated from conception (blue staining). (E) Image analysis of Perl’s staining ICH (WT n = 7, SVC n = 6, SVC PBA n = 12). (F) Immunostaining against Iba1 (brown) on brain sections with detail of dashed square provided in a small square. (G) ImageJ analysis of staining is provided in graph (WT n = 6, SVC n = 5, SVC PBA n = 3). One-way analysis of variance (ANOVA) with post hoc test [Bonferroni (G), Tukey (E); ^*P-value < 0.05; ^***P-value < 0.001].

Figure 2

Effect of PBA treatment on adult eye and renal defects. (A) Daily urine output of five-month-old untreated WT littermate, untreated Col4a1^+/SVC mice and treated Col4a1^+/SVC (WT n = 20, SVC n = 18, SVC PBA n = 6). (B) Haematoxylin and eosin staining of kidneys revealed in untreated Col4a1^+/SVC (SVC) and PBA-treated Col4a1^+/SVC mice (SVC PBA) defects (cuboidal appearance, formation of multiple cell layers) of parietal epithelial cells of Bowman’s capsule (black arrow) and glomerulocystic kidney disease (retraction of vascular tuft, green arrow). SVC: untreated Col4a1^+/SVC. (n = 4). (C) Image analysis of Bowman’s capsules in WT (n = 75 capsules) and Col4a1^+/SVC (n = 102) (χ² = 42.52; 1 df). (D) Image analysis of glomerulocystic kidney disease in WT (n = 75 glomeruli) and Col4a1^+/SVC (n = 102) (χ² = 15.49; 1 df). (E) Image analysis reveals similar frequency of Bowman’s capsule defects in treated (n = 84) and untreated (n = 102) Col4a1^+/SVC. (χ² = 0.009; 1 df) (F) Image analysis reveals similar frequency of cystic defects in treated (n = 103 glomeruli) and untreated (n = 94) Col4a1^+/SVC (χ² = 0.44; 1 df). (G) Haematoxylin and eosin staining of eyes revealed persistence of ASD defects in treated mice (SVC PBA) including iridocorneal adhesion (arrow), atrophy of the ciliary body (^*) size bar = 50 μm. (H) Scoring of anterior segment (presence/absence of iridocorneal adhesion, ciliary body morphology, presence/absence of open Shlemm’s canal) in WT, untreated (SVC) and treated mutant (SVC + PBA) mice. (WT n = 4, SVC n = 4, SVC PBA n = 6 mice). (I) Slit lamp analysis revealed persistence of opacity of the cornea (scoring system applies only to corneal opacity) and iris hypoplasia in treated mice. Corneal neovascularization (see `severe’ panel) is also observed. The right panel is the graphical representation of the scoring of corneal opacity. (SVC n = 6, SVC PBA n = 4 animals). (J) Image analysis of corneal opacity scoring. (C)–(F): Two-sided chi-square test; (A), (H): One-way ANOVA post hoc Sidak test ^*P-value<0.05, ^****P-value <0.0001.

Figure 3

PBA treatment of established disease. (A) Diagram of 1-month oral PBA treatment for the treatment of established phenotypes in 5-month-old mice. (B) MRI image of a mutant animal pre and post treatment reveals ICH (black arrow). (C) ImageJ analysis of MRI data based on bleed number (top) and bleed volume (bottom) (SVC n = 11, SVC PBA n = 8). (D) Perl’s staining of brains sections reveals ICH (blue staining) untreated and treated mutant mice (SVC and SVC 1 M PBA). (E) Image analysis indicating percentage of tissue that stained positive for ICH (n = 3). (F) Slit lamp analysis reveals persistence of corneal opacity and iris hypoplasia. The images are of the same eye of the animal pre and post treatment (n = 4 animals). (G) Scoring of anterior segment (presence/absence of iridocorneal adhesion, ciliary body morphology, presence/absence of open Shlemm’s canal) in untreated (SVC) and treated Col4a1^+/SVC (SVC 1M PBA). (SVC n = 4, SVC 1M PBA n = 4 mice) (H) Image analysis of Bowman’s capsule defects in untreated and treated mice (n = 116). (I) Image analysis of glomerulocystic defects (χ² = 9.506; 1 df, P = 0.002). (Histopathology of the eye and kidney is provided in Supplementary Material, Figs S2 and S3). (D), (E) One-way ANOVA with Bonferonni post hoc test; (F) Two-sided chi-square ^*P-value < 0.05, ^***P-value < 0.001.

Figure 4

Chemical chaperone activity increases deposition of collagen IV in vivo. (A) Western blotting showed increased levels of Bip (∼2.9-fold increase) and ATF4 (∼4.6-fold increase) in untreated Col4a1^+/Svc (SVC unt) which is reduced following PBA treatment. Representative band of total protein stain is given as loading control (the entire gel is provided in Supplementary Material, Fig. S9). (B) ImageJ densitometry analysis of Bip and ATF4 (n = 3). (C) Immunostaining against Col4a1 in PBA-treated and untreated Col4a1^+/Svc mice (WT: wild-type; SVC: untreated; SVC 1M PBA: 1-month treatment; SVC LL PBA: lifelong chronic treatment from conception) on kidney sections revealed increased deposition of Col4a1 following PBA treatment. (n = 3–5 mice, Bowman’s capsule: SVC n = 30, SVC 1 M PBA n = 42, SVC LL PBA n = 16; Tubular BM SVC n = 89, SVC 1M PBA n = 77, SVC LL PBA n = 47; Vascular BM SVC n = 36, SVC 1M PBA n = 23, SVC LL PBA n = 15). (D) ImageJ analysis of fluorescence staining. (B), (D) One-way ANOVA post hoc Sidak test ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Figure 5

Effect of PBA on BM structure. (A) Normal appearance of BM of tubules (white arrow) in untreated Col4a1^+/SVC (SVC) and mice treated for 1 month at 4 months of age (SVC 1M PBA) or from conception for 5 months (SVC LL PBA). (B) Severe defects in BM of Bowman’s capsule in all mice including bulges (white arrow SVC), basket weave appearance (white arrow SVC 1M PBA) and multiple layers (white arrow SVC LL PBA). Evidence of enlarged ER (red arrow SVC 1M PBA) and increased vesicles (red arrow SVC LL PBA) is also observed. (C) Irregular thickening (white arrow) of GBM in treated and untreated mice. Thinner BM areas are also observed (black arrow). (D) VBM defects include interruption (white arrow SVC, 1M PBA), presence of collagen fibrils (white arrow bottom panel 1 M PBA) and more fuzzy but continuous BM (bottom panel LL PBA) black size bar 1 μm, white size bar 5 μm. One-way ANOVA post hoc Tukey test ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Figure 6

PBA does not improve BM strength. (A) PAS stain of tissue section from Col4a1^+/SVC and 1-month treated Col4a1^+/SVC (SVC PBA), which shows separation of dermis from epidermis (circle). (B) ImageJ analysis revealed increased separation in treated mice (SVC n = 7, SVC PBA n = 4) unpaired t-test ^*P < 0.05 (Stain of WT littermate is provided in Supplementary Material, Fig. S8).