Generation of Col4α5 deficient rats by the rGONAD method
On the basis of human mutations as described previously, we introduced novel Col4α5 deficient rats with CRISPR/Cas9 and the rGONAD technology28,30. Tandem STOP codons were integrated into 27 bases after the first ATG in the rat Col4α5 gene (Supplementary Fig. 1a). The mutation was verified by PCR followed by DNA sequencing, resulting to be expressed only 9 amino acids of COL4A5 at its N-terminus (Supplementary Fig. 1b,c). We detected no COL4A5 protein expression in Col4α5 deficient male rats by immunofluorescence and Western blot analyses with monoclonal antibodies against the NC1 (C-terminus) domains of type IV collagens (See below; Figs. 7,8). In addition, the mutant rats which were integrated the other flame-shift tandem STOP codons (Col4α5 15 aa stop; Supplementary Fig. 2) and were deleted 56 bp including the first ATG (Col4α5 56 bp deletion; Supplementary Fig. 3), also generated, and revealed the same phenotype as Alport syndrome rats (Supplementary Fig. 2, 3). These data collectively indicate the successful generation of Col4α5 deficient rats.
Physiological analyses in Col4α5 mutants
The following data are from the mutants rats shown in Supplementary Fig. 1a,b. Col4α5 deficient rats were viable, fertile, and expected Mendelian ratios. However, all hemizygous mutant males died from 18 to 28 weeks of age (n = 58; Fig. 1a). Heterozygous mutant females died at 35 to 100 weeks of age, among them, approximately 30% of females survived over 100 weeks (n = 33, Fig. 1a). To assess of functional and histological abnormalities of the kidney, we measured the hematuria and proteinuria in Col4α5 deficient rats. Hematuria was found from postnatal 21 days in hemizygous mutant males, and from 4 weeks of age in heterozygous mutant females (data not shown). Proteinuria was observed (about 6.0 mg/16 h) by 6 weeks of age in hemizygous mutant males (see Supplementary Table 1), then, increased > 20.0 mg/16 h in all of mutant males after 12 weeks of age (n = 35), and 62% of mutant females in 16 week of age (n = 28/45) (Fig. 1b). On the contrary, these phenomena were not detected in wildtype male and female rats in any week of age (Fig. 1b).
Physiological analyses in Col4α5 mutant rats. (a) Estimated survival functions in wildtype (WT) and Col4α5 mutant (Hemi; hemizygous males, het; heterozygous females) rats. (b) Proteinuria in wildtype and Col4α5 deficient rats from 4 to 52 weeks of age. (c–f) Measurements of body weight (c), urine volume (d), blood urea nitrogen (BUN) of serum (e), serum creatinine (Cre) (f), protein creatinine ratio (g), and creatinine clearance ratio (h) in wildtype and Col4α5 mutant males from 8 to 20 weeks of age.
The mutant males showed decrease of body weight and increase of urine volume, compared with those of wildtype males (Fig. 1c,d). The levels of blood urea nitrogen (BUN) and creatinine (Cre) in serum highly increased in 20 weeks of age of mutant males (Fig. 1e,f). The protein/creatinine ratio was elevated in mutant males from 8 weeks of age, and creatinine clearance ratio was significantly lower in mutant males by 20 weeks of age (Fig. 1g,h).
Renal histology in Col4α5 deficient rats
The kidney sections from wildtype and Col4α5 deficient rats were stained with hematoxylin and eosin (HE), periodic acid Schiff (PAS), periodic acid methenamine silver (PAM), and Masson trichrome (MT) (Fig. 2). At 8 weeks of age, the glomeruli exhibited capillary tuft collapse in hemizygous mutant males (Fig. 2a). However, the kidney displayed overall sparing of the tubulointerstitium (Fig. 2a). By the 20 weeks of age in hemizygous mutant males, substantial numbers of glomeruli revealed the abnormalities, including parietal cell hyperplasia mimicking crescent formation, focal sclerosis, and with overall sparing of the tubulointerstitium (Fig. 2b, Supplementary Fig. 4). The kidney sections from heterozygous mutant females at 8–20 weeks of age displayed focal abnormalities of the glomeruli and tubulointerstitium (Fig. 2a,b, Supplementary Fig. 4).
Histological analyses of Col4α5 deficient kidneys. (a,b) Representative microscopic images in wildtype (WT) and Col4α5 mutant (Hemi; hemizygous males, Het; heterozygous females) rats at 8 weeks (a) and 20 weeks (b) of age. These tissue sections were prepared and stained with hematoxylin and eosin (HE), periodic acid Schiff (PAS), periodic acid methenamine silver (PAM), and Masson trichrome (MT). Scale bars, 100 µm.
To analyze the 3 dimensional ultrastructure of GBM, low-vacuum scanning electron microscopy (LVSEM) was performed, as described previously31. LVSEM revealed the coarse meshwork structure of the GBM with numerous pin-holes in hemizygous mutant males, in contrast to the smoothly arranged surface in wildtype males (Fig. 3, Supplementary Fig. 5). The transmission electron microscopy (TEM) image of the GBM in mutant males exhibited partial thickening from 8 weeks of age, and thickening of the GBM became accentuated and widely spread with increasing age (Fig. 3, Supplementary Fig. 5). Moreover, TEM also revealed substantial thinning of the GBM and split or fragmentation of the lamina densa by 20 weeks of age (Fig. 3, Supplementary Fig. 5).
Electron photomicrographs of glomerular basement membranes in Col4α5 mutant rats. (a,b) Representative low-vacuum scanning electron microscopy (LVSEM, left) and transmission electron microscopy (TEM; right) images in wildtype (WT) and Col4α5 mutant (Hemi) males at 8 weeks (a) and 20 weeks (b) of age. Black and white arrowheads indicate the coarse meshwork structure of the GBM. Black arrows indicate cut side of the capillary walls. Red arrows indicate thickening of the GBM. Green arrows indicate thin patterns of the GBM. Yellow arrows indicate the splitting or fragmenting of the lamina densa. Red insets are revealed the higher magnification of left panels. E: Endothelial cells, M: Mesangial cells, P: Podocytes, R: Red blood cells. Scale bars, 5 µm (left), 1 µm (right).
Renal glomerular and tubulointerstitial fibrosis in Col4α5 deficient rats
To evaluate fibrosis of the glomeruli and tubulointerstitium in Col4α5 deficient rats, we examined kidney sections from wildtype and Col4α5 deficient rats stained with antibodies specific for α-smooth muscle actin (α-SMA) and fibronectin (Fig. 4). Immunostaining showed that expression level of α-SMA, a maker of renal glomerular and tubulointerstitial fibrosis, was increased around the glomeruli from hemizygous mutant males when compared with those of wildtype littermates at 20 weeks of age, but it was not detected in the renal tubular epithelia (Fig. 4a). The α-SMA expression of kidney sections from the heterozygous females was rarely detectable at 20 weeks of age (Fig. 4a, Supplementary Fig. 6). In contrast, the expression level of fibronectin, a marker of pathological deposition of the extracellular matrix (ECM), was increased in 12 weeks of age of hemizygous mutant males (Fig. 4b). We then observed fibronectin expression in heterozygous mutant females at 12 weeks of age (Fig. 4b, Supplementary Fig. 7).
Renal fibrosis in Col4α5 deficient rats. (a,b) Immunostaining of kidney sections with α-SMA (a) or fibronectin (b) (red), nestin (green; glomeruli), and DAPI (blue; nuclei) in wildtype (WT) and Col4α5 mutant (Hemi; hemizygous males, Het; heterozygous females) rats from 8 to 20 weeks of age. Scale bars, 50 µm.
Next, it has previously been reported that human, dog, and mouse Alport kidneys showed the progressive deposition in the GBM such as laminins32,33, and the progressive induction in glomeruli such as profibrotic cytokine TGFβ132 and matrix metalloproteinase (MMPs)34. To characterize such possibilities in Alport model rats, we performed immunostaining and Western blotting of Col4α5 deficient male kidneys for measuring the levels of these proteins. We first observed that an antibody to the laminin β2 chain exhibited intense staining of the GBM in Col4α5 deficient males from 8 weeks of age (Fig. 5a, Supplementary Fig. 8a). To explore whether the TGFβ pathway was affected in Col4α5 deficient rats, we examined levels of phospho-Smad3 which are the most critical mediators in TGFβ signaling pathway by Western blotting. We found that phosphorylation levels were increased in mutant male kidneys in comparison with those in wildtype (Fig. 5d). We then observed the protein expression of TGFβ downstream mediator CTGF35 in Col4α5 deficient rats. Immunostaining analyses also showed the level of CTGF was increased in the GBM of mutant males by 12 weeks of age (Fig. 5b, Supplementary Fig. 8b). Moreover, we observed that an antibody to the MMP3/10 revealed intense staining of the GBM in Col4α5 deficient males by 12 weeks of age (Fig. 5c, Supplementary Fig. 8c). In addition, we also detected increased MMP12 levels by Western blotting (Fig. 5d).
Immunostaining and Western blotting of the GBM in Col4α5 deficient kidneys. (a–c) Immunofluorescence analyses of rat kidney sections with antibodies against (a): Laminin β2 (red), Nestin (green; glomeruli); (b): CTGF (red), Nephrin (green); (c): MMP3/10 (red), Nephrin (green); and DAPI (blue; nuclei) in wildtype (WT) and Col4α5 mutant (Hemi) male rats at 8 weeks (left) and 20 weeks (right) of age. Scale bars, 50 µm. (d) Detection of phospho- and total Smad3, and MMP12 by Western blot analyses of the kidneys in wildtype (W) and Col4α5 mutant (H) male rats from 8, 12, 16, to 20 weeks of age. Tubulin was evaluated as an internal control.
In the recently report, osteopontin (OPN) and LDL receptor (LDLR) are highly expressed in renal tubules of Alport mice36. To investigate the expression of OPN and LDLR in Alport model rats, we performed immunostaining and Western blotting of Col4α5 deficient male kidneys and/or plasma samples for measuring the levels of these proteins. Immunostaining analyses revealed the level of OPN was elevated in the renal tubules of mutant males from 8 weeks of age (Fig. 6a, Supplementary Fig. 9a). In addition, Western blotting also showed an increase of OPN expression in kidneys (Fig. 6c) and plasma (Fig. 6d) of Col4α5 deficient males. In addition, LDLR expression was increased in the renal tubules of mutant males by 16 weeks of age by immunostaining (Fig. 6b, Supplementary Fig. 9b).
OPN and LDLR expressions in Col4α5 deficient rats. (a,b) Immunostaining of kidney sections with (a) OPN or (b) LDLR (red), Nestin (green; glomeruli), and DAPI (blue; nuclei) in wildtype (WT) and Col4α5 mutant (Hemi) male rats in wildtype (WT) and Col4α5 mutant (Hemi) male rats at 8 weeks (left) and 20 weeks (right) of age. Scale bars, 50 µm. (c,d) Detection of OPN by Western blotting of the kidneys (c) or plasma (d) in wildtype (W) and Col4α5 mutant (H) male rats from 8, 12, 16, to 20 weeks of age. Tubulin and CBB staining were evaluated as an internal control. Asterisk in (d) indicates non-specific signal.
The expression of type IV collagen α1-6 in Col4α5 deficient rats
The type IV collagen networks comprised of α1/α1/α2 (IV) (all basement membranes), α3/α4/α5 (IV) (GBM), and α5/α5/α6 (IV) (Bowman’s capsule) protomers were observed in the glomeruli. To examine whether localizations and/or levels of the proteins were changed in Col4α5 deficient rats, we investigated Col4α5 deficient rats with graded levels of the other’s type IV collagen α1, α2, α3, α4, and α6 (IV). First, we produced monoclonal antibodies that specifically recognize rat COL4A6 protein. We obtained rCol4A6 antibodies by injecting the antigen in Col4a5 deficient males, but we could not obtain any specific antibody injected to wildtype rats at all. Immunoblot analysis revealed that the monoclonal anti-rCOL4A6, reacted specifically with a band corresponding to the position of the similar molecular weights in recombinant rCOL4A6 proteins, and only reacted with rCOL4A6, but not with other rCOL4 proteins (Supplementary Fig. 10).
To determine whether localization of type IV collagens of α1-6 (IV) in the kidney of Col4α5 deficient rats changed, we performed the immunofluorescence analyses. At 8–20 weeks of age, COL4A5 expression was absent in hemizygous mutant males, and present in a mosaic pattern in heterozygous mutant females (Fig. 7a,b, Supplementary Fig. 11–14). The expressions of COL4A3, COL4A4, and COL4A6 were also absent in hemizygous mutant males, and present in a mosaic pattern in heterozygous mutant females (Fig. 7a,b, Supplementary Fig. 11–14). At 8 weeks of age, in contrast, expressions of COL4A1 and COL4A2 were present in Col4α5 deficient rats (Fig. 7a, Supplementary Fig. 11). At 20 weeks of age, the kidney showed a strong accumulation of COL4A1 and COL4A2 proteins in both the GBM and Bowman’s capsule (Fig. 7b, Supplementary Fig. 14). These data suggest that α3/α4/α5 (IV) and α5/α5/α6 (IV) chains of type IV collagen disrupted in Col4α5 deficient rats.
Type IV collagen distributions in Col4α5 deficient kidneys. (a,b) Immunofluorescence analyses of kidney sections with antibodies against α1-6 (IV) (red), nestin (green; glomeruli), and DAPI (blue; nuclei) in wildtype (WT) and Col4α5 mutant (Hemi; hemizygous males, Het; heterozygous females) rats at 8 weeks (a) and 20 weeks (b) of age. Scale bars, 50 µm.
To verify the results of immunostaining, we analyzed COL4 protein expressions of noncollagenous domains 1 (NC1) by Western blot analyses that were prepared from kidneys of wildtype and hemizygous mutant males (Fig. 8). There were no bands detectable NC1 domains of type IV collagen α3, α4, and α5 (IV), derived from α3/α4/α5 (IV) complexes, in kidney of Col4α5 deficient males (Fig. 8). Then, the NC1 domains of type IV collagen α5 and α6 (IV), derived from α5/α5/α6 (IV) complexes, were also absent, whereas the blotting for the NC1 domains of type IV collagen α1 and α2 (IV), derived from α1/α1/α2 (IV) complexes, were present in hemizygous mutant males as well as wild type (Fig. 8). Therefore, these findings fully corroborate the results of immunofluorescence analyses.
Western blot analyses of type IV collagen in Col4α5 mutant kidneys. Collagenase-solubilized renal basement membranes from wildtype (W) and Col4α5 mutant (H) males at 8 weeks of age were separated by SDS-PAGE and blotted with antibodies that the specifically recognized α1-6 (IV) NC1 domains. D: NC1 dimer, M: NC1 monomer.
