Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Jan;11(1):e9540.
doi: 10.15252/emmm.201809540.

Murine MPDZ-linked Hydrocephalus Is Caused by Hyperpermeability of the Choroid Plexus

Affiliations
Free PMC article

Murine MPDZ-linked Hydrocephalus Is Caused by Hyperpermeability of the Choroid Plexus

Junning Yang et al. EMBO Mol Med. .
Free PMC article

Abstract

Though congenital hydrocephalus is heritable, it has been linked only to eight genes, one of which is MPDZ Humans and mice that carry a truncated version of MPDZ incur severe hydrocephalus resulting in acute morbidity and lethality. We show by magnetic resonance imaging that contrast medium penetrates into the brain ventricles of mice carrying a Mpdz loss-of-function mutation, whereas none is detected in the ventricles of normal mice, implying that the permeability of the choroid plexus epithelial cell monolayer is abnormally high. Comparative proteomic analysis of the cerebrospinal fluid of normal and hydrocephalic mice revealed up to a 53-fold increase in protein concentration, suggesting that transcytosis through the choroid plexus epithelial cells of Mpdz KO mice is substantially higher than in normal mice. These conclusions are supported by ultrastructural evidence, and by immunohistochemistry and cytology data. Our results provide a straightforward and concise explanation for the pathophysiology of Mpdz-linked hydrocephalus.

Keywords: cerebrospinal fluid; choroid plexus; hydrocephalus; magnetic resonance imaging; proteomics.

Figures

Figure 1
Figure 1. Hydrocephalus was detected by PET in Mpdz −/− mice

Images of P4 (top) and P21 (bottom) Mpdz +/+ and Mpdz −/− mice. Arrowheads point to the domed foreheads of the latter.

Mean weights of P18‐P21 Mpdz +/+ and Mpdz −/− mice (n = 8–12, mean ± SD; the value of P was determined by two‐tailed Student's t‐test).

Coronal, axial, and sagittal (top to bottom) PET images of P18‐P21 Mpdz +/+ and Mpdz −/− mice. The emission intensity is shown as a 7‐point temperature scale from black (0) to white (7).

Mean PET emission intensities in the indicated brain sections (n = 4, mean ± SD; the values of P were determined by two‐tailed Student's t‐test).

Figure 2
Figure 2. Severe hydrocephalus and leakage of contrast medium were detected in Mpdz −/− mice by MRI

Coronal, sagittal, and axial (clockwise from top left corner image of each genotype) MR images of Mpdz +/+ and Mpdz −/− P18‐P21 mice. Ventricles are pseudo‐colored in red and 3D‐reconstructed in the lower left panels.

Means of total ventricle and brain volumes of Mpdz +/+ and Mpdz −/− mice (n = 8, mean ± SD; the values of P were determined by two‐tailed Student's t‐test).

A coronal HE‐stained section (center) flanked by anatomically corresponding coronal T2‐weighted MR images of Mpdz −/− P18‐P21 mice. Arrows or arrowheads show the match between the lateral ventricle CP in the HE section and in the MR images.

Duplicate rows of T2‐weighted coronal images and anatomically corresponding T1‐weighted coronal images at 1 min and at the peak‐signal time point after contrast medium injection; the bottom row shows a similar set of axial images. Areas surrounded by squares are magnified in the insets; arrows mark the contrast medium signal in the T1‐weighted images. Each row corresponds to one mouse aged 18–21 days.

Time courses of the normalized T1‐weighted image intensities corresponding to the areas surrounded by numbered squares on the T2‐weighted images.

Data information: Scale bars, 1 mm; insets, 0.5 mm.
Figure EV1
Figure EV1. No contrast medium is detected in the brain ventricles of Mpdz +/+ mice

Coronal T2‐weighted MR image and an anatomically corresponding HE‐stained section. Scale bar, 1 mm.

Triplicate rows of T2‐weighted and T1‐weighted coronal images 1 and 10 min post‐contrast medium injection. The squares surround the locations of the lateral ventricles. Each row corresponds to one P18‐P21 mouse. Scale bar, 1 mm.

Time courses of the normalized T1‐weighted image intensities corresponding to the locations of the areas surrounded by numbered squares in the T2‐weighted images.

Figure 3
Figure 3. The Sylvian aqueduct of the Mpdz −/− mouse is stenotic

Coronal HE‐stained brain sections of a P18 Mpdz +/+ mouse (out of a total of three) showing the 3rd ventricle, and the proximal and distal sections of the aqueduct of Sylvius; these features are marked by arrows in the insets.

The same, for P18 Mpdz −/− mice.

Top views of brains of P7 Mpdz +/+ and Mpdz −/− mice that show the sites of Evans blue injection, and the midline planes (dashed line) sectioned to produce the sagittal images below. They show the extent of Evans blue spread in the ventricles and surrounding tissue (one out of three experiments). In, injection; LV, lateral ventricle; SA, Sylvian aqueduct; V3, third ventricle; V4, fourth ventricle.

Data information: Scale bars, 1 mm; insets, 0.25 mm.
Figure 4
Figure 4. Mpdz was localized proximal to the apical surface of CPECs

Immunofluorescence images of 10‐μm‐thick sections of CP from third ventricle villi of Mpdz +/+ and Mpdz −/− mice were immunolabeled as shown. The areas in the square frames are magnified in the bottom panels. Scale bars, top panels, 100 μm; bottom panels, 10 μm.

CP sections from lateral ventricle villi immunolabeled for ZO1 (B) or E‐cadherin (E‐cad; C). The images are representative of two Mpdz +/+ mice and three Mpdz −/− mice. Scale bars, top panels, 100 μm; bottom panels, 10 μm.

Immunoblots with the indicated antibodies (top) and their quantifications. The densitometry measurements were normalized relative to the signal of the samples transduced by non‐targeting shRNA (Ctrl). Note that the same β‐actin immunoblot was used twice because the Jam‐C and E‐cadherin samples were immunoblotted on the same membrane (mean ± SD, n = 3; the values of P were determined by two‐tailed Student's t‐test).

Source data are available online for this figure.
Figure 5
Figure 5. CPECs of Mpdz −/− mice harbored structural defects

TEM images of longitudinal sections of lateral ventricle CP villi of P18‐P21 Mpdz +/+ and Mpdz −/− mice. Scale bar, 2 μm.

A large number of voids of varying sizes were evident in the CPECs of Mpdz −/− mice. Higher magnification images of the areas in the rectangles are shown underneath the panels. Adherens junctions are denoted by horizontal lines and arrows; voids are indicated by arrows. Scale bar, 1 μm; insets, 200 nm.

Triplicate images of tight junctions proximal to the apical faces of CPECs in Mpdz +/+ and Mpdz −/− mice. Scale bar, 100 nm.

Mitochondria were smaller and frequently lacked large portions of their cristae. Autophagosomes are present in some of them (arrows). Scale bar, 200 nm. The images are representative of two Mpdz +/+ mice and two Mpdz −/− mice.

Time course and standard deviations of the impedance of confluent hpCPEC monolayers that were transduced by either MPDZ or non‐targeting (Ctrl) shRNA. Each record represents four wells (mean ± SD). MPDZ immunoblot of each cell group is shown in the inset.

Source data are available online for this figure.
Figure EV2
Figure EV2. The structure of CPEC tight junctions of P12‐14 Mpdz −/− mice is defective
A gallery of three TEM images of tight junctions between the CPECs from lateral ventricle CP villi of Mpdz +/+ and Mpdz −/− P12‐14 mice. Scale bars, 100 nm.
Figure EV3
Figure EV3. Capillaries between CPECs of Mpdz −/− mice do not harbor structural defects
TEM images of sections of capillaries from lateral ventricle CP villi of Mpdz +/+ and Mpdz −/− P15‐P21 mice. The magnified fields show intercellular junctions between endothelial cells or fenestrae. Scale bars, 1 μm; insets, 250 nm.
Figure 6
Figure 6. Fluid‐phase uptake by CPECs is higher in Mpdz −/− relative to Mpdz +/+ mice

TEM images of CPEC sections from lateral ventricle villi of two Mpdz +/+ and two Mpdz −/− P14‐P16 mice injected with HRP. The CPs were reacted with hydrogen peroxide and DAB ex vivo. The dark particles are DAB deposits internalized by micropinocytosis. The magnified fields to the right show individual particles engulfed in macropinosomes. Note the layered structure of the particles, and the macropinosome that is open to the ventricular space in the Mpdz −/− section. Scale bars, 1 μm; insets, 100 nm.

Mean number of engulfed DAB particles per cell in Mpdz +/+ and Mpdz −/− mice (mean ± SD, n = 22; the value of P was determined by two‐tailed Student's t‐test).

A CPEC section showing the engulfment of a DAB particle by the cell's basal ruffles in the magnified field. Scale bars, 1 μm; insets, 200 nm.

A CPEC section showing a preponderance of macropinosomes close to the apical face of the cell, and a magnified field that contains several macropinosomes. Scale bars, 1 μm; insets, 200 nm.

Mean numbers of DAB‐containing macropinosomes close to the apical or basal sides of CPECs from Mpdz +/+ or Mpdz −/− mice (mean ± SD, n = 20; the values of P were determined by two‐tailed Student's t‐test).

Figure 7
Figure 7. LDLR abundance and endocytosis by CPECs are higher in Mpdz −/− relative to Mpdz +/+ mice

Immunofluorescence images of 10‐μm‐thick sections from lateral ventricle CP villi of Mpdz +/+ and Mpdz −/− P14‐P16 mice immunolabeled by anti‐LDLR. Scale bars, 50 (top) and 25 (bottom) μm.

Mean fluorescence intensities (normalized relative to the highest recorded intensity) per cell in several LDLR‐immunolabeled CP sections (mean ± SD, n = 101; the value of P was determined by two‐tailed Student's t‐test).

Immunoblots showing the abundances of LDLR and MPDZ in hCPECs transduced by MPDZ‐targeting or non‐targeting shRNA. Mean abundances were quantified by densitometry of the LDLR bands, normalized relative to the β‐actin bands (mean ± SD, n = 3; the values of P were determined by two‐tailed Student's t‐test).

Fluorescence images of hCPECs transduced by either MPDZ or non‐targeting (Ctrl) shRNA and immunolabeled by anti‐LDLR either before (0 min) or after 8 min of constitutive endocytosis of LDLR. Scale bar, 10 μm. The histograms below show the mean fluorescence intensities per hCPEC in each cell group, normalized relative to the highest recorded intensity (mean ± SD, n = 41–53; the values of P were determined by two‐tailed Student's t‐test).

Source data are available online for this figure.
Figure 8
Figure 8. The protein content in the CSF of Mpdz −/− mice was higher than in the CSF of Mpdz +/+ mice

Mean protein concentration in the CSF of Mpdz +/+ and Mpdz −/− P14‐P21 mice (mean ± SD, n = 2–3; the values of P were determined by two‐tailed Student's t‐test).

Volcano plot showing that the protein contents in sera of P15‐P21 Mpdz +/+ and Mpdz −/− mice are similar to each other.

A volcano plot showing that 23 proteins are at least twofold significantly more abundant in the CSF of P15‐21 Mpdz −/− mice (red circles, protein names are indicated), whereas only two proteins were more abundant in the CSF of Mpdz +/+ mice (green circles). The CSF composition was analyzed in three Mpdz −/− and three Mpdz +/+ mice.

Images of 10‐μm‐thick sections from lateral ventricle CP villi of Mpdz +/+ and Mpdz −/− mice, immunolabeled for Nkcc1. The areas in the marked squares are magnified below. Scale bars, top, 50 μm, bottom 20 μm. The histograms show the mean fluorescence intensities per hCPEC in each cell group, normalized relative to the highest recorded intensity (mean ± SD, n = 103; the value of P was determined by two‐tailed Student's t‐test).

Immunoblots of LDLR and Mpdz, with β‐actin as loading control. The MPDZ immunoblot shows the efficiency of the shRNA‐mediated knockdown. Bands from three immunoblots were quantified by normalization to β‐actin bands (mean ± SD; the value of P was determined by two‐tailed Student's t‐test).

Source data are available online for this figure.

Similar articles

See all similar articles

Cited by 1 article

References

    1. Adachi M, Hamazaki Y, Kobayashi Y, Itoh M, Tsukita S, Furuse M, Tsukita S (2009) Similar and distinct properties of MUPP1 and Patj, two homologous PDZ domain‐containing tight‐junction proteins. Mol Cell Biol 29: 2372–2389 - PMC - PubMed
    1. Al‐Dosari MS, Al‐Owain M, Tulbah M, Kurdi W, Adly N, Al‐Hemidan A, Masoodi TA, Albash B, Alkuraya FS (2013) Mutation in MPDZ causes severe congenital hydrocephalus. J Med Genet 50: 54–58 - PubMed
    1. Al‐Jezawi NK, Al‐Shamsi AM, Suleiman J, Ben‐Salem S, John A, Vijayan R, Ali BR, Al‐Gazali L (2018) Compound heterozygous variants in the multiple PDZ domain protein (MPDZ) cause a case of mild non‐progressive communicating hydrocephalus. BMC Med Genet 19: 34 - PMC - PubMed
    1. Annenkov A (2009) The insulin‐like growth factor (IGF) receptor type 1 (IGF1R) as an essential component of the signalling network regulating neurogenesis. Mol Neurobiol 40: 195–215 - PubMed
    1. Balin BJ, Broadwell RD (1988) Transcytosis of protein through the mammalian cerebral epithelium and endothelium. I. Choroid plexus and the blood‐cerebrospinal fluid barrier. J Neurocytol 17: 809–826 - PubMed

Publication types

LinkOut - more resources

Feedback