Mcph1-deficient mice reveal a role for MCPH1 in otitis media

Abstract

Otitis media is a common reason for hearing loss, especially in children. Otitis media is a multifactorial disease and environmental factors, anatomic dysmorphology and genetic predisposition can all contribute to its pathogenesis. However, the reasons for the variable susceptibility to otitis media are elusive. MCPH1 mutations cause primary microcephaly in humans. So far, no hearing impairment has been reported either in the MCPH1 patients or mouse models with Mcph1 deficiency. In this study, Mcph1-deficient (Mcph1(tm1a) (/tm1a) ) mice were produced using embryonic stem cells with a targeted mutation by the Sanger Institute's Mouse Genetics Project. Auditory brainstem response measurements revealed that Mcph1(tm1a) (/tm1a) mice had mild to moderate hearing impairment with around 70% penetrance. We found otitis media with effusion in the hearing-impaired Mcph1(tm1a) (/tm1a) mice by anatomic and histological examinations. Expression of Mcph1 in the epithelial cells of middle ear cavities supported its involvement in the development of otitis media. Other defects of Mcph1(tm1a) (/tm1a) mice included small skull sizes, increased micronuclei in red blood cells, increased B cells and ocular abnormalities. These findings not only recapitulated the defects found in other Mcph1-deficient mice or MCPH1 patients, but also revealed an unexpected phenotype, otitis media with hearing impairment, which suggests Mcph1 is a new gene underlying genetic predisposition to otitis media.

Figure 1. Production of Mcph1-deficienct (Mcph1^tm1a/tm1a) mice.

(A) Schematic of knockout strategy for Mcph1 gene based on knockout-first design. A promoterless cassette including LacZ and neo genes was inserted in the third intron of Mcph1 gene flanked by FRT sites. LoxP sites flank the critical exon (exon4 of Mcph1 gene in knockout-first design). See http://www.knockoutmouse.org/martsearch/project/41705 for more details. (B) Short range PCR for genotyping. Wild type allele produces one band of 366bp. Due to the insertion of the cassette, primers designed for the wild type allele do not have product for the mutant allele using the short range PCR (illustrated as the left panel, schematic illustration is not in scale). The homozygous allele produces only one band of 185 bp. The heterozygotes produce two bands of 185 bp and 366 bp. (C) Quantitative real-time PCR showed largely reduced transcript of Mcph1 in Mcph1^{tm1a /tm1a} (n = 3) mice compared to wild type mice (n = 3) and the residual levels vary in different organs.

Figure 2. Mcph1-deficient mice have mild to moderate hearing impairment characterized with conductive hearing impairment.

(A) ABR measurement results of 14 week old Mcph1^tm1a/tm1a mice (n = 11) in MGP showed mild to moderate hearing impairment, or normal hearing compared to control mice. The green baseline area shows a reference range for the control wild type mice with the same genetic background, plotting the median and 2.5 to 97.5 percentile of the population (n = 440). (B) Input-output function (IOF) analysis. The peak-peak amplitude of wave 1 (P1-N1 amplitude) of click-evoked ABRs is plotted as a function of dB SL (Sensation Level, dB above threshold) for wild type (green) and Mcph1^tm1a/tm1a (red) mice. There was no significant difference of IOF slopes of Mcph1^tm1a/tm1a (n = 24, slope  = 0.144+/−0.066; mean +/− SD) and wild type mice (n = 36, slope  = 0.133+/−0.048) (t-test, p = 0.444).

Figure 3. Recurrent ABR measurement indicated the relation between the hearing profile and middle ear defects.

(A) Results of recurrent ABR measurement (click thresholds) with age. Hearing impairment can be detected as early as 3 weeks old in Mcph1^tm1a/tm1a mice (n = 13). Hearing profile of the Mcph1^tm1a/tm1a mice showed either a stable, progressive, or fluctuating pattern with age (three of them marked dark). All the wild type (n = 13) and heterozygous (n = 17) mice displayed normal click thresholds with age. (B) Auditory chain (incus-stapes joint) and oval window sound transduction was severely impeded. Normal incus-stapes joint of auditory chain in a Mcph1^{+ /+} mouse, and a clear oval window is necessary for sound vibration conduction. After removing some of the amorphous material in the middle ear cavity of a Mcph1^{tm1a /tm1a} mouse, the incus-stapes joint (arrow head) and the oval window (arrow) is present but embedded in the amorphous material. Scale bar, 1 mm. (C–F) Correlation between middle ear defects and hearing sensitivity change with time. (C) Normal ABR thresholds and middle ear structure in a wild type mouse: normal middle ear cavity is full of air, tympanic membrane is transparent and normal morphology of ossicles. (D) Progressively elevated ABR thresholds with age in a Mcph1^{tm1a /tm1a} mouse. Amorphous mass filled the middle ear cavity and outgrew into external ear canal. Ossicles were embedded in the amorphous mass and appeared to have thinner bony structure. (E) Fluctuating ABR thresholds in a Mcph1^{tm1a /tm1a} mouse. Watery effusion with bubbles was seen in the middle ear cavity and normal gross morphology of ossicles. (F) Stable and moderate hearing impairment in a Mcph1^{tm1a /tm1a} mouse. The middle cavity was filled with pus-like secretion. Normal gross morphology of ossicles but with rough surface. Scale bar, 1 mm.

Figure 4. Similar ABR waveforms in Mcph1^tm1a/tm1a and wild type mice at 9 weeks old.

A & B illustrate click-evoked ABR waveforms recorded in a wildtype (green) and aMcph1^{tm1a /tm1a} (red) mouse, respectively. The dB SPL of the click stimulus for each response is indicated on the abscissa. The scale bar indicates 5 µV amplitude of response. The heavy lines indicate the click-ABR threshold allocated to each mouse. C. (inset) Mean click-evoked ABR waveforms recorded at 21 dB sensation level for wildtype (n = 8, green) and Mcph1^{tm1a /tm1a} (n = 8, red).

Figure 5. Hematoxylin and eosin staining of the middle ear in adult mice indicated otitis media.

Clear middle ear cavity (MEC) and thin mucoperiosteum in wild type mice (A,C). MECs of Mcph1^{tm1a /tm1a} mice (B,D) were filled with exudate and lined with thickened mucoperiosteum. High magnification for mucoperiosteum of MEC framed in A and B (C,D). Inflammatory cells (E,F) in MECs. Scale bar, 200 µm (A,B), 20 µm (C–F).

Figure 6. Normal inner ear structure in Mcph1^tm1a/tm1a mice.

(A) Scanning electron microscope (SEM) showed normal development of hair cells at P4 in Mcph1^{tm1a /tm1a} mice (Mcph1^{+ /+}, n = 3; Mcph1^{tm1a /tm1a}, n = 3. scale bar, 10 µm). (B) HE slides displayed comparable structure of inner ears (basal turn) in Mcph1^{+ /+} and Mcph1^{tm1a /tm1a} mice at 4–5 weeks old (Mcph1^{+ /+}, n = 3; Mcph1^{tm1a /tm1a}, n = 3. scale bar, 50 µm).

Figure 7. Mcph1 is expressed in the middle ear.

Immunochemistry using an antibody shows that Mcph1 (brown labelling) is expressed in epithelial cells of the middle ear cavity at P7 (A,B) and P28 (C,D) wild type mice. Mcph1 is expressed in both ciliated (B,D) cells close to orifice of Eustachian tube and non-ciliated cells (A,C). Scale bar, 20 µm.

Figure 8. Mcph1^{tm1a /tm1a} mice have normal skull structure but smaller skull size.

(A) X-ray assay showed comparable structure of craniofacial skeleton between wild type and Mcph1^{tm1a /tm1a} mice at 14 weeks old. (B) Measurement showed that skull width and length in female Mcph1^{tm1a /tm1a} mice (n = 6) are significantly smaller than those of wild type mice (n = 7) (Rank Sum test). The weight of the brain was measured at 16 weeks of age showed that female and male Mcph1^{tm1a /tm1a} mice (n = 3, each sex) had lighter brain than the local control mice (n = 3, each sex) and MGP wild type mouse baseline (female, n = 10; male n = 187).

Figure 9. Mcph1^tm1a/tm1a mice show evidence of genomic instability.

Adult Mcph1^{tm1a /tm1a} (n = 7) mice showed increased genomic instability when compared to wild type mice (n = 34) as determined by an increased prevalence of micronucleated normochromatic erythrocytes (MN-NCE). Data is presented as the Mean ± SEM.

Figure 10. Mcph1^tm1a/tm1a mice had normal control of Salmonella Typhimurium and Citrobacter rodentium challenge.

(A) No difference of weight change (Mean ± SEM) between Mcph1^tm1a/tm1a (n = 4) and Mchp1^+/+ (n = 8) mice infected by Salmonella Typhimurium and monitored by weight loss for 21 days. (B) Viable bacterial counts from the spleen (i) and the liver (ii) from the same mice infected by Salmonella Typhimurium at 21 days post infection. There was only one Mcph1^tm1a/tm1a mouse out of 4 tested with any bacteria present in the liver. Mann Whitney U tests were used, p (two-tailed) is indicated in the figure. (C) Viable bacterial counts (Mean ± SEM) being shed in the stool from Mcph1^tm1a/tm1a (n = 5) and Mchp1^+/+ (n = 5) mice were similar over a 28 day infection of Citrobacter rodentium.

Figure 11. Measurement of anti-TetC specific antibodies in immunised mouse serum.

Mcph1^tm1a/tm1a and Mchp1^+/+ mice were immunised intranasally with fragment C of tetanus toxin on day 0, 7 and 21. Presence of antigen-specific Ig in serum isolated at day 28 was showed no difference in Mcph1^tm1a/tm1a (n = 6) and Mchp1^+/+ (n = 5). The solid bar represents the Mean ± SEM. Mann Whitney U tests were used, p (two-tailed) is indicated in the figure.

Figure 12. Mcph1^{tm1a /tm1a} mice have ocular abnormalities.

(A) Slit lamp images (12× magnification) revealed corneal (center) opacity and vascularisation (right) in Mcph1^{tm1a /tm1a} mice. The difference of ocular abnormality portion is significant between wild type (n = 23) and Mcph1^tm1/tm1a (n = 14) mice (Fischer's exact test: p = 0.002). (B) Wild type eye shows normal lens and retina. The anterior and posterior chamber spaces are well defined. ac  =  anterior chamber; pc  =  posterior chamber. (C) Mcph1^{tm1a /tm1a} eye shows cataractous lens and thin retina. The anterior and posterior chambers are collapsed. Scale bar, 500 µm (B,C). (D) Inset from (C) with solid line shows cataractous lens and loss of retinal cell layers. Scale bar, 100 µm.