Mice harboring an MCTO mutation exhibit renal failure resembling nephropathy in human patients.

Multicentric carpotarsal osteolysis (MCTO) is a condition involving progressive osteolysis of the carpal and tarsal bones that is associated with glomerular sclerosis and renal failure (MCTO nephropathy). Previous work identified an autosomal dominant missense mutation in the transactivation domain of the transcription factor MAFB as the cause of MCTO. Several methods are currently used for MCTO nephropathy treatment, but these methods are invasive and lead to severe side effects, limiting their use. Therefore, the development of alternative treatments for MCTO nephropathy is required; however, the pathogenesis of MCTO in vivo is unclear without access to a mouse model. Here, we report the generation of an MCTO mouse model using the CRISPR/Cas9 system. These mice exhibit nephropathy symptoms that are similar to those observed in MCTO patients. MafbMCTO/MCTO mice show developmental defects in body weight from postnatal day 0, which persist as they age. They also exhibit high urine albumin creatinine levels from a young age, mimicking the nephropathic symptoms of MCTO patients. Characteristics of glomerular sclerosis reported in human patients are also observed, such as histological evidence of focal segmental glomerulosclerosis (FSGS), podocyte foot process microvillus transformation and podocyte foot process effacement. Therefore, this study contributes to the development of an alternative treatment for MCTO nephropathy by providing a viable mouse model.

Introduction

Multicentric carpotarsal osteolysis (MCTO) is a rare osteolytic disease in which affected patients suffer from progressive osteolysis of the carpal and tarsal bones and may also develop renal failure in childhood [14]. Zankl et al. used exome capture and next-generation sequencing to reveal that an autosomal dominant missense mutation in the transactivation domain of v-maf musculoaponeurotic fibrosarcoma oncogene ortholog B (MAFB) causes MCTO in human patients [26].

MafB is a member of the large Maf transcription factor family, which regulates gene expression by binding as dimers to Maf recognition elements (MAREs). MafB is expressed in various cell types, such as pancreatic α and β cells; cells of the thymus, parathyroid gland, hair follicles, epidermal keratinocytes, and renal glomerulus; and cells of myeloid lineages [4, 5, 8,9,10,11, 14,15,16, 19, 22]. In cells of myeloid lineages, MafB has been reported to silence the self-renewal ability of macrophages and induce the expression of F4/80 to support cell differentiation [1, 17]. With regard to pathology, conditional knock-out of Mafb in animal models specifically in hematopoietic stem cells (Tie2-Cre::Mafb) and myeloid cells (LysM-Cre::Mafb) accelerates obesity and leads to autoimmune phenotypes due to defects in the production of the complement component C1q in macrophages, respectively [23, 24]. Additionally, MafB promotes the development of atherosclerosis by inducing the expression of apoptosis inhibitor of macrophages (AIM) to prevent foam-cell apoptosis as well as excess inflammation after ischemic stroke [7, 21]. Cuevas et al. showed that MCTO-mutated macrophages displayed increased anti-inflammatory gene expression compared with controls, suggesting a contribution of MCTO-mutated MafB to disease progression [3].

MafB also regulates foot process development during podocyte differentiation, with Mafb-deficient embryos showing kidney hypoplasia and podocyte foot process effacement [17]. Fan et al. defined Mafb as a locus of susceptibility for albuminuria in the diabetic KK/Ta mice strain [6]. Morito et al. proved its relevance by overexpression of Mafb specifically in podocytes, which rescued albuminuria and nephrin depletion in mice treated with streptozotocin (STZ) [18], highlighting the importance of MafB expression in renal diseases, including MCTO. Although cyclosporine A has been successful in treating MCTO-associated nephropathy (MCTO nephropathy), patients usually suffer from side effects such as hirsutism, gingivitis, and hypertension with long-term use [2, 9]. Thus, clarification of the pathogenesis of MCTO nephropathy and the development of alternative treatments using an MCTO mouse model are needed.

The delay in producing safer treatments results from an unclear understanding of MCTO nephropathy, as insights from the use of animal models are needed. In this study, we generated a mouse harboring a human MCTO mutation c.176C>T, (p.Pro59Leu) by using the CRISPR/Cas9 system. Based on a previous analysis of Mafb-deficient embryos [17], we hypothesized that the MCTO missense mutation causes glomerular sclerosis via dysfunction of podocyte foot process maintenance and development.

Materials and Methods

Animals

C57BL/6J and Jcl:CD1 (ICR) mice were purchased from Charles River Laboratories Japan (Kanagawa, Japan) and CLEA Japan (Tokyo, Japan), respectively. The mice were kept in plastic cages under specific-pathogen-free conditions in a room maintained at 23.5 ± 2.5°C and 52.5 ± 12.5% relative humidity under a 14-h light:10-h dark cycle. The mice had free access to commercial chow (Milk Fat diet, Oriental Yeast, Tokyo, Japan) and filtered water. All experiments were performed in compliance with relevant Japanese and institutional laws and guidelines and were approved by the University of Tsukuba Animal Ethics Committee (authorization number 17-154).

Production of p.Pro59Leu Mafb mutant mice with CRISPR/Cas9 technology

To induce the c.176C>T mutation in the Mafb gene via the CRISPR/Cas9 technique, we selected the sequence 5’-GCACGGAGTGCTGAGCGGGG-3’ as the guide RNA (gRNA) target. This sequence was inserted into the entry site of pX330-U6-Chimeric_BB-CBh-hSpCas9, which was a gift from Feng Zhang (Addgene plasmid #42230), and this plasmid was designated as pX330-MafB. In addition, we designed a 200-nt single-stranded DNA oligonucleotide (ssODN) donor to induce c.[176C>T;177G>T] (p.Pro59Leu) and c.174C>T (p.Thr58Thr: synonymous substitution). This synonymous substitution was designed to increase the induction efficiency of the p.Pro59Leu mutation by preventing recutting of the edited target by active Cas9 [12]. These mutated nucleotides were placed between 98-nt 5’- and 3’-homology arms derived from positions 76–173 and 178–275 of the Mafb coding sequence, respectively. This MafB-P59L-ssODN was obtained from Integrated DNA Technologies (Skokie, IL, USA).

Microinjection

Pregnant mare serum gonadotropin (5 units) and human chorionic gonadotropin (5 units) were intraperitoneally injected into female C57BL/6J mice (Charles River Laboratories) with a 48-h interval between injections, and the mice were naturally mated with male C57BL/6J mice. Embryos were collected from the oviducts of the female mice, and a mixture of pX330-MafB(5 ng/µl) and MafB-P59L-ssODN (10 ng/µl) was injected into each male pronucleus of 292 embryos. Shortly thereafter (15 min to 2 h), 283 living embryos were transferred to the oviducts of pseudopregnant ICR mice (Charles River Laboratories, Japan), and 81 newborn mice were obtained. Of the 81 newborn mice, 6 carried the p.Pro59Leu mutation, which was confirmed by PCR with AmpliTaq Gold 360 Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) and specific primers (Mafb mut genotype mutF, 5’-TCGGTGTCGTCCACTCTT-3’; Mafb mut genotype wtF, 5’-TCGGTGTCGTCCACCCCG-3’; Mafb mut Genotype R, 5’-CATCGTGAGTCACACCTGCT-3’). Sequencing of the PCR products was conducted using a BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Waltham, MA, USA) and the MafB Mut Genotype F Primer (5’-ACGTCAACGACTTCGACCTT-3’).

Urinary creatinine and albumin measurements

The urinary albumin and creatinine levels of the mice at specific ages were measured using a Hitachi 7170 automated analyzer (Hitachi High-Technologies Corporation, Tokyo, Japan). Albumin was determined via turbidimetric immunoassays.

Histopathological analysis of murine renal tissues

Each mouse was exsanguinated while under ether anesthesia. Its organs were then fixed with 10% formalin in 0.01 mol/l phosphate buffer (pH 7.2) and embedded in paraffin. Sections were assessed by periodic acid-Schiff (PAS) staining and Masson trichrome staining for histopathological examination under light microscopy. The ratio of sclerotic glomeruli was calculated according to a previous study [25]. In brief, PAS-stained glomeruli were counted, and sclerotic glomeruli were divided by the total number of glomeruli counted for each sample. Transmission electron microscopy was performed using standard methods.

Statistical analysis

All results are expressed as the means ± SEM. Significant differences between two groups were analyzed using Student’s t-test. Differences were considered statistically significant at P<0.05.

Results

Generation of mice with the MCTO mutation with the CRISPR/Cas9 system

Zankl et al. reported that an autosomal dominant cystine (C) to thymine (T) missense mutation in the acidic transactivation domain of the MAFB gene at position 176 causes MCTO nephropathy in humans [26]. This mutation, which has also been reported by Mumm et al. and Mehawej et al., causes both nephropathy and osteolysis in MCTO patients [14, 19]. To introduce this point mutation into the mouse genome with the CRISPR/Cas9 system, we first selected the ideal CRISPR cleavage target site within the transactivation domain of Mafb. As the Cas9 nuclease cleaves specific sites that are followed by a protospacer adjacent motif (PAM) sequence, we selected the site between positions 169 and 189 of the Mafb gene as the target sequence (Fig. 1A). We also designed a 200-nt ssODN donor for homology-directed repair (HDR)-mediated genome mutation, in which nucleotides 99, 101, and 102 from the 5’ end were replaced with T. The 101st nucleotide T was placed between the 100-nt 5’ and 102-nt 3’ homology arms, which are equivalent to nucleotide 175 from the 5’ end and nucleotide 177 from the 3’ end of the Mafb gene, respectively. The pX330-MafB vector was used to express the Mafb-targeting gRNA under the control of the U6 promoter in zygotes. The pX330-MafB vector also allowed the expression of Cas9 cDNA controlled by a CBh promoter, causing cleavage of the target site of Mafb guided by the 20-bp double-stranded DNA derived from the Mafb gene. The ssODN and the pX330-MafB vector were comicroinjected into mouse zygotes, followed by PCR analysis and sequencing analysis to screen for complete and/or mosaic pups with the MCTO mutation. Mosaic mice were further crossed with Mafb mice to obtain Mafb mice (F1 generation), and Mafb mice were obtained by crossing Mafb mice (F2 generation). Sequencing analysis was performed on the F2 generation to confirm the integration of the MCTO mutation (Fig. 1B), and PCR assays were performed with primer pairs for the MafbWT and MafbMCTO sequences (Fig. 1C, upper sequences). The MafbWT PCR product of 301 bp was observed only in Mafb and Mafb mice, whereas Mafb mice only exhibited a nonspecific PCR product of approximately 600 bp (Fig. 1C, lower left). A MafbMCTO PCR product was observed only in Mafb and Mafb mice, without any nonspecific PCR products (Fig. 1C, lower right).

Fig. 1.

Generation of Mafb (A) The upper sequence shows the Mafb domains within the genome. Red box, transactivation domain; green box, extended homology region; blue box, basic region; gray box, leucine zipper domain. The middle sequence shows the CRISPR target sequence at position 169–189 bp of the Mafb gene (boxed). The 200-nt oligo DNA sequence is shown in the lower sequence. The red bold face C/G in the middle sequence and T in the lower sequence indicate the target nucleotide mutation. (B) Genomic sequence of the CRISPR target site in Mafb and Mafb mice. Bold letters in the black box indicate the target mutation. (C) Genotyping methods for Mafb mice. In the upper sequence, the MafbWT allele and its associated primer are illustrated (Mafb mut genotype wtF). The 176C>T mutation and its associated primer are shown in the lower sequence (Mafb mut genotype mutF). The genotype of each mouse was determined via PCR. Representative results of PCR are shown in the lower images.

Growth deficiency in MafbMCTO/MCTO mice is observed from P0 and continues after birth

Previous reports have shown that MCTO patients exhibit nephropathy and osteolysis from early childhood [14]. To apply this finding to our Mafb mice, we examined whether Mafb and Mafb mice show phenotypes from postnatal day 0 (P0) to 2 weeks. At P0, Mafb mice showed body weights similar to Mafb mice, while Mafb mice exhibited a 10% lower body weight than Mafb mice (Fig. 2A, Table 1). The differences between Mafb and Mafb mice were still observed at 2 weeks, with Mafb mice presenting a 23% lower body weight than Mafb mice (Fig. 2B, Table 1). Body weight was 40% lower in Mafb mice than both Mafb and Mafb at 4 weeks (Table 1), and body weight remained lower until 38 weeks (Fig. 2C, Table 1). Although the results were not significant due to the limited sample numbers, a tendency for body weight to be lower by 24%, was observed in Mafb mice compared with Mafb mice at 38 weeks (Table 1). All the individual body weights are shown in Supplementary Table 1–3. These results indicate that the MCTO mutation not only affects development at P0 but also influences the subsequent growth pattern in vivo.

Fig. 2.

The MCTO mutation influences subsequent body growth. (A) Representative photographs of Mafb, Mafb, and Mafb littermates at P0. Scale bar: 1 cm. (B) Representative photographs of Mafb, Mafb, and Mafb littermates at 2 weeks. Scale bar: 1 cm. (C) Representative photographs of Mafb, Mafb , and Mafb littermates at 38 weeks. Scale bar: 1 cm. Table 1 provides the average weights of each genotype from P0 to 38 weeks. Supplementary Table 1, 2 and 3 provides the body weights of individual mice. The results for P0 and 2 weeks are from both male and female mice. Measurements from 2 weeks onwards were performed only in female mice. The data are presented as the mean ± SEM.; *P<0.05 (Student’s t-test); **P<0.01 (Student’s t-test).

Table 1.

Body weight of each genotype over 38 weeks

Week	MafbWT/WT			MafbMCTO/WT			MafbMCTO/MCTO			P value
	Average (g)	Standard error(average ±)	Sample number	Average (g)	Standard error(average ±)	Sample number	Average (g)	Standard error(average ±)	Sample number	MafbMCTO/MCTO, MafbWT/WT	MafbMCTO/MCTO, MafbMCTO/WT	MafbMCTO/MCTO, MafbMCTO/MCTO
0 (P0)	1.41	0.04	7	1.33	0.02	10	1.29	0.49	5	0.02	0.22	0.08
2	6.96	0.54	5	6.93	0.13	10	5.40	2.41	3	0.03	0.0001	0.94
4	12.37	0.59	3	12.50	0.57	6	7.27	4.20	4	0.01	0.001	0.89
8	18.00	1.35	3	17.20	0.58	4	12.33	6.17	6	0.02	0.004	0.57
10	19.60	1.83	3	18.00	0.26	4	14.08	5.75	6	0.01	0.01	0.35
12	20.77	2.44	3	18.65	0.46	4	15.20	6.21	6	0.03	0.02	0.36
14	22.17	2.30	3	19.38	0.37	4	15.97	6.52	6	0.01	0.003	0.22
16	22.97	2.25	3	19.93	0.50	4	15.28	6.24	6	0.002	0.0003	0.18
18	23.63	1.97	3	20.18	1.01	4	16.50	7.38	5	0.02	0.03	0.24
20	24.27	1.37	6	20.60	0.50	4	17.36	5.79	9	0.0001	0.001	0.07
24	27.73	4.08	3	22.58	0.74	4	17.88	7.30	6	0.01	0.003	0.20
26	28.93	4.29	3	22.55	0.74	4	18.15	7.41	6	0.01	0.004	0.14
28	28.27	4.22	3	22.58	0.85	4	18.25	7.45	6	0.01	0.004	0.18
38	31.53	5.67	3	24.03	0.94	4	18.57	7.58	6	0.01	0.001	0.18

Note: P value calculated with Student’s t-test.

MafbMCTO/MCTO mice exhibit an increased urine albumin creatinine ratio and develop FSGS lesions

Because renal symptoms appear in childhood in some MCTO patients with the 176C>T mutation [14], we assessed whether the Mafb mice show a similar phenotype to clinical cases by analyzing the urine albumin creatinine ratio (ACR) at 4 weeks. As expected, Mafb mice already exhibited urine ACR levels 6 times higher than those of Mafb mice at 4 weeks, which continued until 38 weeks (Fig. 3A, Supplementary Table 4–6). Similar to body weight, 3 out of 4 samples from the Mafb mice showed a trend of increased ACRs compared with Mafb mice at 38 weeks of age (urinary albumin levels of Mafb versus Mafb: 0.23, 0.24, and 0.29 versus 0.11, 0.77, 0.65, and 2.48; Fig. 3A, Supplementary Table 4–6).

Fig. 3.

Mafb (A) Urinary albumin creatinine ratios measured in Mafb, Mafb, and Mafb mice from 4 weeks to 38 weeks. The sample numbers of each genotype were as follows (for Mafb, Mafb, and Mafb, respectively): 3, 4, and 3 at 4 weeks; 3, 4, and 4 at 8 to 12 weeks; 3, 4, and 6 at 14 to 18 weeks; 6, 4, and 9 at 20 and 26 weeks; and 3, 4, and 6 at 24, 28, and 38 weeks. Measurements at 4 weeks were performed in both male and female mice, while measurements at other times were performed only in female mice. (B) Kidneys of Mafb and Mafb mice subjected to PAS staining. Representative images are shown. Scale bar: 100 µm (left). The ratio of sclerotic glomeruli to total glomeruli was calculated in a blinded test (Mafb, n=4; Mafb, n=3). Solid gray circles, Mafb; solid white circles, Mafb (right). (C) Kidneys of Mafb and Mafb mice stained with Masson’s trichrome staining. Representative images are shown. Scale bar: 100 µm. Yellow arrowhead, sclerotic regions of a single glomerulus; blue area, fibrotic regions. Histological analyses were conducted using 25- and 26-week-old mice. Supplementary Table 4, 5, and 6 provides the urine albumin creatinine ratios of individual mice. The data are presented as the mean ± SEM.; *P<0.05 (Student’s t-test); **P<0.01 (Student’s t-test).

Renal histopathology was also performed using PAS staining, and it was found that female Mafb mice showed a significantly higher percentage of sclerotic glomeruli than Mafb mice (Fig. 3B). To assess the sclerotic phenotype of Mafb and Mafb mice, we performed Masson’s trichrome staining, which revealed that Mafb mice showed more sclerotic glomeruli and tubulointerstitial lesions (blue regions) than Mafb mice (Fig. 3C).

Since previous reports on Mafb mice have shown podocyte foot process defects, which were also found in biopsy studies of an MCTO patient [2, 17], we performed transmission electron microscopic analysis of mature glomeruli in 25- and 26-week-old Mafb and Mafb female mice. Microvillous transformation of the foot processes of podocytes was frequently observed in the Bowman’s space in Mafb glomeruli, which differed from the observations for Mafb glomeruli, and this is a known characteristic of FSGS (Fig. 4A). Higher magnification images revealed Mafb podocytes with clear discrete foot processes, whereas Mafb podocytes exhibited foot process effacement, similar to that seen in MCTO nephropathy patients (Fig. 4B). Thus, Mafb mice display clear evidence of podocyte defects, which resemble MCTO nephropathy symptoms, from early stages of life.

Fig. 4.

Mafb. (A) Electron microscopy image of 26-week-old female Mafb and Mafb glomeruli. Representative image are shown. White arrowhead, glomerular basement membrane; black arrowhead, Bowman’s space; scale bar: 2 µm (Mafb, n=3; Mafb, n=3). (B) Electron microscopy images of 26-week-old female Mafb and Mafb podocyte foot processes at 5,000× magnification. Representative image are shown. PC, podocyte; GBM, glomerular basement membrane.

Discussion

MCTO is caused by an autosomal dominant missense mutation in the transactivation domain of the MAFB gene and is characterized by osteolysis of the carpal/tarsal bones and nephropathy. The development of a safer treatment for MCTO nephropathy remains a challenge because mouse model resources are limited. To overcome this issue, we generated a mouse model of MCTO via the CRISPR/Cas9 system to study the pathogenesis of MCTO nephropathy and contribute to the development of alternative treatments.

Among several mutations in MAFB reported to cause MCTO, we chose to generate mice with the 176C>T mutation reported by Zankl et al. [26]. The 176C>T mutation has been reported by several researchers, providing plausibility for comparing clinical reports and our Mafb mice [14, 19]. In addition, Mehawej et al. reported the parental relationships of patients with the 176C>T mutation, suggesting the possibility that this mutation will be transmitted to further generations of Mafb mice [14]. Moreover, since nephropathy may occur from osteolysis or vice versa, the generation of a model based on a mutation with a phenotype that includes these symptoms will provide new directions for developing treatments that take both factors into consideration.

Although MCTO nephropathy has been reported to occur from an autosomal dominant missense mutation of the MAFB gene [26], a milder phenotype was observed in Mafb mice than in Mafb mice. This result could be explained by previous reports on resistance against kidney defects in disease model mice with a C57BL/6J mouse background [4, 8] as well as the general differences between mice and humans [20]. Nevertheless, Mafb mice tended to show higher ACRs than Mafb mice, and this might be more evident under aged conditions. In contrast, Mafb mice showed permanently higher ACRs from 4 to 38 weeks than both Mafb and Mafb mice. Such results imitate the proteinuria observed in human patients during childhood, which arises between 0 and 6 years of age, making the Mafb model superior to the Mafb model. Histological and transmission electron microscopy analyses of the glomeruli of Mafb mice showed evidence of FSGS and both podocyte foot process effacement and microvillous transformation. These last characteristics have both been reported in previous clinical reports on MCTO nephropathy and are typical hallmarks of FSGS, suggesting the usefulness of Mafb mice as a model organism [2, 13, 14].

While our results indicate that Mafb mice show MCTO nephropathic phenotypes from 4 weeks of age, the reduced body weight of Mafb mice observed at P0 may arise as a skeletal defect, considering that symptoms also appear in the skeletal systems of MCTO patients. Therefore, further analyses are required to distinguish the tissue-specific effect of MCTO mutation in bone mineralization and remodeling. Mehawej et al. also reported clinical variability among patients, such as an absence of nephropathy associated with some mutations, the onset of renal symptoms associated with the same mutations, and incomplete penetrance within a family, all of which occurred for unknown reasons [14]. The generation of MCTO model mice with different mutations and further investigation of the modifier genes surrounding MAFB will be necessary to resolve the meaning behind these findings. Moreover, although renal failure is frequently reported in MCTO patients, diagnosis usually occurs at the later stages of the disease. Studying signs of renal failure using Mafb mice at earlier stages may lead to the development of diagnostic measures to prevent MCTO nephropathy before chronic stages.

In conclusion, we used CRISPR/Cas9 technology to generate a mouse model of MCTO nephropathy with symptoms that resemble those of human MCTO patients. These mice showed a permanently reduced weight from birth and an excessive urine albumin creatinine ratio from an early age. Consistent with these characteristics, histological analysis showed features of podocyte defects in Mafb mice, which may be the cause of the MCTO nephropathy. Therefore, Mafb mice could be useful for the establishment of new treatments for MCTO nephropathy as ideal model organisms.