1. Developmental Biology
  2. Neuroscience

The microtubule-binding protein EML3 is required for mammalian embryonic growth and cerebral cortical development, and Eml3 null mice are a model of cobblestone brain malformation

  1. Isabelle Carrier  Is a corresponding author
  2. Eduardo Diez  Is a corresponding author
  3. Valerio E Piscopo
  4. Susanne Bechstedt
  5. Hans van Bokhoven
  6. Myriam Srour
  7. Albert M Berghuis
  8. Stefano Stifani
  9. Yojiro Yamanaka
  10. Roderick McInnes
  1. Lady Davis Research Institute for Medical Research of the Montreal Jewish General Hospital, McGill University, Canada
  2. Montreal Neurological Institute-Hospital, McGill University, Canada
  3. Department of Anatomy and Cell Biology, McGill University, Canada
  4. Centre de recherche en biologie structurale, McGill University, Canada
  5. Department of Human Genetics, Radboud University Medical Center, Donders Institute for Brain, Cognition, and Behaviour, Netherlands
  6. Montreal Children’s Hospital, MUHC-RI, Canada
  7. Department of Pediatrics, McGill University, Canada
  8. Department of Neurology and Neurosurgery, McGill University, Canada
  9. Department of Biochemistry, McGill University, Canada
  10. Department of Microbiology and Immunology, McGill University, Canada
  11. Goodman Cancer Institute, McGill University, Canada
  12. Department of Human Genetics, McGill University, Canada
https://doi.org/10.7554/eLife.107102.3
Share this article
Cite this article
  1. Isabelle Carrier
  2. Eduardo Diez
  3. Valerio E Piscopo
  4. Susanne Bechstedt
  5. Hans van Bokhoven
  6. Myriam Srour
  7. Albert M Berghuis
  8. Stefano Stifani
  9. Yojiro Yamanaka
  10. Roderick McInnes
(2026)
The microtubule-binding protein EML3 is required for mammalian embryonic growth and cerebral cortical development, and Eml3 null mice are a model of cobblestone brain malformation
eLife 14:RP107102.
https://doi.org/10.7554/eLife.107102.3
  2. Download BibTeX
  3. Download .RIS
7 figures, 3 tables and 2 additional files

Figures

Figure 1
Download asset Open asset
Constitutive and tissue-specific targeting of Eml3.

(A) Eml3 has 22 coding exons in mouse chromosome 19. (B) Constitutive Eml3 null mice (Eml3tm1e(EUCOMM)Wtsi or Eml3gt) with a gene trap allele from a EUCOMM ES cell line. (C) The gene trap-floxed allele (Eml3em1Mci or Eml3gt-fl) with a LoxP site inserted into intron 19 (floxed exons 11–19). (D) The floxed allele (Eml3em1.1Mci or Eml3fl) with the gene trap cassette excised. (E) The tissue-specific Eml3 null alleles, (Eml3fl;Cre) in the presence of Cre recombinases under the control of tissue-specific promoters. The use of a germline-expressed Cre driver (E2a-Cre) yielded a global Eml3 null allele named Eml3em1.2Mci. Immunoblotting of E15.5 whole embryo lysates with EML3 antibody (EML3-C884) detected no EML3 protein (~90 kDa) in homozygotes for either (F) the gene trap allele (Eml3tm1e(EUCOMM)Wtsi or Eml3gt), or (G) the exon 11–19 deletion allele (Eml3em1.2Mci or Eml3−). Heterozygous embryos yielded approximately half the amount of protein as wild-type controls in both mouse lines. GAPDH immunoblotting was used as the protein loading control.

Figure 1—source data 1

PDF file containing original TIF images used to prepare Figure 1, panels F and G, indicating the relevant bands and genotypes.

https://cdn.elifesciences.org/articles/107102/elife-107102-fig1-data1-v1.zip
Figure 1—source data 2

TIF files of images obtained from Bio-Rad ChemiDoc for immunoblotting displayed in Figure 1F, G.

https://cdn.elifesciences.org/articles/107102/elife-107102-fig1-data2-v1.zip
Figure 2
Download asset Open asset
Eml3−/− embryos are smaller than control littermates and are developmentally delayed.

At each gestational age between E8.5 and E18.5, the Eml3−/− embryos (A) and their placentas (B) (E10.5–E18.5) are significantly smaller than their Eml3+/+ and Eml3+/− control littermates. The size was measured either by area (E7.5–E10.5 embryos; E10.5–E18.5 placentas) or weight (E11.5–E18.5 embryos); for each litter analyzed, the size of each embryo or placenta was measured relative to the average of the control embryos (Eml3+/+ and Eml3+/−) and then normalized to the average of the Eml3+/+ embryos. The average percent size difference is shown below each image (n = 5–32). Dunn’s pairwise comparison, *p < 0.05, ns = not significant. Scale bars, 100 µm for E7.5, 500 µm for E8.5, and 1 mm for E9.5–E18.5. (C) At each gestational age between E8.5 and E12.5, the Eml3−/− embryos have significantly fewer somite pairs than Eml3+/+ and Eml3+/− control littermates; average difference in number of somite pairs ± SD shown (n = 20–50). (D) Weight of embryos was plotted against the number of somite pairs counted for Eml3−/− and control embryos. Linear regression was used to fit the datapoints; Eml3−/− embryos are significantly smaller (p < 0.0001) than control embryos with the same number of somite pairs.

Figure 3
Download asset Open asset
Focal neuronal ectopias (FNEs) in Eml3−/− embryos.

(A) Nissl (cresyl-violet) staining of coronal sections of forebrains from E18.5 Eml3+/+ and Eml3−/− embryos. The Eml3+/+ sections in the top panels show normal brain architecture with a close-up view in the panel on the right. An FNE is present in the section of the Eml3−/− cortex shown in the bottom panels. The close-up in the bottom-right panel shows over-migrating neurons (dark cresyl-violet staining) extending past the marginal zone and into the subarachnoid space. Ca, calvarium; CP, cortical plate; MZ, marginal zone; SAS, subarachnoid space; SVZ, subventricular zone; V, ventricle; VZ, ventricular zone. Scale bars, 100 µm. (B) Immunolabeling of an E18.5 Eml3−/− mouse brain section for laminin (green), a major component of the extracellular matrix of the pial basement membrane, and for neuron-specific tubulin beta-3 (TUBB3, red). The pial basement membrane is disrupted at the FNEs, with neurons (TUBB3+) over-migrating into the subarachnoid space. Laminin speckles are observed (arrow). CP, cortical plate; PBM, pial basement membrane; SAS, subarachnoid space. Scale bar, 50 µm. (C) Heads of E15.5 Eml3+/+ and Eml3−/− embryos were immuno-stained and cleared before imaging with a fluorescence stereomicroscope (a–c) or with a confocal microscope (a’–c”). Immunostaining was performed with antibodies against the ECM protein laminin (green in a–c and a’–c’, white in a”–c”) and a marker of terminally differentiated neurons, tubulin beta-3 (TUBB3, red). The asterisks point to the same area in the high magnification confocal microscope images as in the low magnification stereomicroscope images. Scale bars, 1 mm (a–c) and 0.1 mm (a’–c”).

Figure 4
Download asset Open asset
Eml3−/− embryos are developmentally delayed vs littermate controls, but the onset and progression of neurogenesis are equal in somite-matched Eml3−/− and control forebrains; additionally, neurons of all cortical plate layers are present in Eml3−/− focal neuronal ectopias (FNEs).

(A) The onset and progression of neurogenesis are delayed in Eml3−/− vs littermate control forebrains. Forebrains from embryos collected at E11.5 were cryosectioned and processed for indirect immunofluorescence with markers for cell populations whose relative abundances indicate the onset and progression of neurogenesis. The percentage of TBR2-positive intermediate progenitors was determined. Eml3+/+ and Eml3+/− data are pooled as controls (Ctrls) vs Eml3−/− embryos. The total number of cells in the embryonic brain section was determined with the nuclear stain Hoechst. Each dot on the graph represents the mean percentage of cells positive for TBR2 from analysis of 2–4 sections per embryo. Error bars represent SEM. Pairwise comparison using Student’s t-test, *p < 0.05. (B) The onset and progression of neurogenesis are equal in somite-matched Eml3−/− and control forebrains. Forebrains from embryos collected at E10.5 and E11.5 were cryosectioned and processed for indirect immunofluorescence with markers for cell populations whose relative abundances indicate the onset and progression of neurogenesis. The percentage of TBR2-positive intermediate progenitors was determined and is plotted against the number of somite pairs counted in each embryo. The total number of cells in the embryonic brain section was determined with the nuclear stain Hoechst. Each dot on the graph represents the mean percentage of cells positive for TBR2 in an individual embryo. Error bars represent SD from analysis of 2–4 sections per embryo. Eml3+/+ and Eml3+/− data are pooled as controls (Ctrls) vs Eml3−/− embryos. Shown below the graphs is the Theiler stage of development that corresponds to the number of somite pairs counted. Also indicated is the age of the embryos in days post-coitum. (C) Neurons of all cortical plate layers are present in Eml3−/− FNEs at E18.5. (a, b) Nissl (cresyl-violet) staining of serial coronal sections through a single FNE. (a’, b’) Immunolabeling of adjacent sections of the same FNE with markers of cortical plate layers. (a’) CUX1 immunostaining of late-born upper layer neurons (II–IV) and CTIP2 immunostaining of intermediate layer neurons (V). (b’) TLE4 immunostaining of early-born deeper layer neurons (VI) and CTIP2 immunostaining. Scale bars, 500 µm (a, b) and 100 µm (a’, b’).

Figure 5 with 1 supplement
Download asset Open asset
Onset of neuronal over-migration and structural integrity of pial basement membrane (PBM) ECM.

(A) Double immunostaining of laminin (green) and tubulin beta-3 (TUBB3, red) at 39 somite pair (sp) embryonic stage (TS17; collected at E11.5). Scale bars, 20 µm. (a) In a heterozygous brain, the PBM is continuous and migrating neuroblasts are observed underneath, whereas in an Eml3−/− embryo (b) a gap in the ECM and ectopic neuroblasts (arrowhead) were detected. (B, C) Coronal head sections of control and Eml3−/− embryos at different developmental stages were stained for laminin and tubulin beta-3. The length of gaps in the ECM was measured and the presence of over-migrating neuroblasts (NBs) was noted. The total length of PBM evaluated was measured for normalization. The green arrow points to the Eml3−/− embryo that had a focal neuronal ectopia (FNE) at the earliest developmental stage, that is, 39 somite pairs. (B) For each embryo, the total length of gaps per mm of PBM analyzed was plotted against the number of somite pairs counted. The Eml3−/− embryos with and without FNEs were plotted separately. (C) For each embryo, the length of the largest gap observed was plotted against the number of somite pairs counted. For the Eml3−/− embryos, the largest gaps with and without over-migrating NBs through the gap were plotted separately. No over-migrating NBs were ever observed in gaps of control embryos. (D) Transmission electron microscopy was performed on sections of Eml3−/− and control embryos (Ctrls; Eml3+/+ and Eml3+/−) matched by developmental stage. Three stages were analyzed (31–33, 39–41, and 46–47 somite pairs; sp). The embryo ages in days post-coitum are indicated. The total length of PBM was measured in each TEM micrograph. The ECM length was then subdivided into regions with different structural integrity. Normal, dense: sharp, straight, electron-dense ECM. Slightly altered: thicker, diffuse, less dense ECM. Severely altered: delaminating ECM, with electron-dense material shedding off a diffuse PBM. Absent: regions with no ECM overlaying the NEP basal end-feet. Representative micrographs shown. At the 31–33 sp stage, control ECM was 8.6 ± 7.8% normal (dense), 85.4 ± 8.0% slightly altered, 4.3 ± 3.6% severely altered, and 1.6 ± 1.4% absent, whereas Eml3−/− ECM was 8.5 ± 7.0% normal (dense), 81.4 ± 2.9% slightly altered, 9.3 ± 8.1% severely altered, and 0.7 ± 1.1% absent. At the 39–41 sp stage, control ECM was 66.8 ± 6.3% normal, 19.3 ± 9.0% slightly altered, and 13.9 ± 15.2% severely altered, whereas Eml3−/− ECM was 14.4 ± 6.7% normal, 55.1 ± 23.8% slightly altered, 29.3 ± 30.2% severely altered, and 1.2 ± 0.4% absent. At the 46–47 sp stage, control ECM was 56.8 ± 9.1% normal, 28.2 ± 4.0% slightly altered, and 15.0 ± 5.1% severely altered, whereas Eml3−/− ECM was 13.9 ± 5.9% normal, 34.1 ± 7.2% slightly altered, 50.8 ± 11.6% severely altered, and 1.2 ± 1.6% absent. Scale bars, 100 nm.

Figure 5—figure supplement 1
Download asset Open asset
Components of the PBM in stage-matched E9.5 Eml3+/+ and Eml3−/− embryos.

Immunolabeling of the pial basement membrane on sections of developing brains from stage-matched E9.5 Eml3+/+ (top) and Eml3−/− (bottom) mouse embryos. Antibodies against the major components of the extracellular matrix of the PBM: laminin (green, left) and collagen IV (red, left), as well as for the major ECM component receptors expressed on the end-feet of neuroepithelial cells: α-dystroglycan (green, right) and integrin α6 (red, right) were used. MM, meningeal mesenchyme; NE, neuroepithelium; PBM, pial basement membrane. Scale bar, 10 µm.

Figure 6
Download asset Open asset
EML3 is expressed in the tissues that form and maintain the pial basement membrane (PBM).

(A) EML3 protein expression (green) as determined by indirect immunofluorescence on a coronal cryosection of an E10.5 Eml3+/+ embryo head. This developmental timepoint, 37 somite pairs, immediately precedes the appearance of the first focal neuronal ectopia (FNE) in Eml3−/− embryos (at 39 somite pairs). An Eml3−/− littermate with 35 somite pairs was included as a control for antibody staining specificity. Ctx, cerebral cortex; Me, mesenchyme; PP, preplate; SVZ, subventricular zone; VZ, ventricular zone; V, ventricle. Hoechst nuclear stain (blue). Scale bars, 100 µm. (B) EML3 protein expression (green) during forebrain development as determined by indirect immunofluorescence on E9.5–E18.5 Eml3+/+ embryo cryosections. In the bottom panels, the same EML3-stained section is shown with additional immunostaining for markers of neural progenitor cells (SOX1, white) and of post-mitotic neuronal cells (TUBB3, red), as well as Hoechst nuclear stain (blue). The dotted line indicates the location of the PBM, which separates mesenchyme (Me) and cerebral cortex whose cell composition changes during development. At E9.5, the cortex is spanned by neuroepithelial cells (NE). At E10.5 and E11.5, the neuroepithelial progenitors are differentiating into radial glial cells, and the cortex can now be divided into ventricular zone (VZ), subventricular zone (SVZ), and the post-mitotic neurons of the preplate (PP). At E12.5–E18.5, radially migrating neurons are populating the region below the PBM and forming the developing cortical plate (CP). V, ventricle. Scale bars, 25 µm.

Figure 7
Download asset Open asset
Verification of EML3 protein interactions in co-transfected cells.

HEK293T cells were transfected with plasmids encoding for the indicated full-length proteins. A fraction of the cell lysates was immunoblotted for verification of protein expression (INPUT lanes, 2% or 4% of the total lysate), and the remainder was used for immunoprecipitations with EML3 rabbit polyclonal antibody (EML3 IP lanes). HEK293T cells express low amounts of endogenous EML3 protein. (A) EML3 interacts with DYNLL1 and YWHAE proteins in co-transfection experiments. DYNLL1 co-immunoprecipitates with co-expressed EML3 protein in lane A’. YWHAE co-immunoprecipitates with co-expressed EML3 protein in lane B’. ROM1 does not co-immunoprecipitate with co-expressed EML3 protein in lane C’. DYNLL1 and YWHAE are not immunoprecipitated by the EML3 antibody when no EML3-expressing construct is co-transfected into the cells in lane D’. (B) Mutating the TQT86 motif of the EML3 protein to AAA abolishes binding to DYNLL1. DYNLL1 co-immunoprecipitates with co-expressed EML3 protein in lane A’. DYNLL1 does not co-immunoprecipitate with co-expressed EML3TQT86AAA mutant protein in lane B’.

Figure 7—source data 1

PDF file containing scans of original films used to prepare Figure 7A, B, indicating the relevant bands and coIP conditions.

https://cdn.elifesciences.org/articles/107102/elife-107102-fig7-data1-v1.zip
Figure 7—source data 2

Scans of original films corresponding to Figure 7A.

https://cdn.elifesciences.org/articles/107102/elife-107102-fig7-data2-v1.zip
Figure 7—source data 3

Scans of original films corresponding to Figure 7A.

https://cdn.elifesciences.org/articles/107102/elife-107102-fig7-data3-v1.zip
Figure 7—source data 4

Scans of original films corresponding to Figure 7B.

https://cdn.elifesciences.org/articles/107102/elife-107102-fig7-data4-v1.zip

Tables

Table 1
Phenotypes observed in Eml3 null mice.
PhenotypeSub-phenotypeAge when observed*Penetrance of phenotype vs Eml3 genotype
Eml3+/+Eml3+/−Eml3−/−
Embryo development delaySmall embryo and placenta size†E8.5–E18.52/41 (4.9%)1/75 (1.3%)23/23 (100%)
Fewer somite pairs‡E8.5–E12.53/25 (12%)4/51 (8%)11/20 (55%)
Delayed neurogenesis§E10.5–E11.50/10 (0%)0/1 (0%)8/11 (72.7%)
Thin corpus callosumE18.50/8 (0%)0/1 (0%)10/10 (100%)
Brain defectFocal neuronal ectopiasE11.5–E18.50/10 (0%)n.d.10/14 (71.4%)
Perinatal lethalityFailure to inflate lungs**E18.55/24 (20.8%)5/33 (15.2%)6/10 (60%)
Survival up to weaning age††PN18–24n.d.355 Het
vs 191 WT
(~7% Het lethality)		11 KO
vs 191 WT
(~94% KO lethality)
  1. *

    Embryonic (E) or postnatal (PN) ages at which the phenotype was observed. When a range of ages was studied, the penetrance data shown in the table is from the underlined age.

  2. †

    Embryo sizes were measured as area of the section corresponding to the sagittally bisected E8.5 and E10.5 embryos or as weights for E11.5 and E18.5 embryos. E18.5 weight analyses shown, with 15% lighter weight vs average of heterozygote littermates as the threshold for small size. 100% penetrance for low weight of Eml3 null embryos was observed at E12.5–E18.5. Eml3 null placentas are small at E10.5–E18.5.

  3. ‡

    Somite pairs were counted under a dissection microscope for E8.5–E12.5 embryos. E12.5 data shown, with a minimum difference of 1.5 somite pairs vs the average for heterozygous littermates as the threshold for delay.

  4. §

    Delayed neurogenesis was determined by staining E10.5–E11.5 frontal cortices with markers for neuroepithelial progenitors (PAX6 and SOX1), intermediate neuronal progenitors (TBR2), and postmitotic projection neurons (TBR1). E11.5 data shown as the number of embryos with at least a 20% lower percentage of TBR2-positive cells in their frontal cortex as compared to a control littermate.

  5. Corpus callosum (CC) thicknesses were measured in Nissl-stained coronal sections of E18.5 whole brains. Pairwise comparisons among littermates determined that all Eml3 null embryos had a disproportionally thinner CC than their littermates even when normalizing for the overall smaller size of the embryos.

  6. **

    Air breathing and lung floating assays were performed on E18.5 embryos collected by cesarean section. The pups were observed and filmed for 90 min. At the end of the observation period, all mice were euthanized, and the lungs were removed and placed in PBS for the lung floating assay.

  7. ††

    We estimate the number of missing mice at weaning age by comparing the number of Eml3 null and heterozygous mice to the number of weaned wild-type mice. Given that we analyze litters from heterozygote intercrosses and that we expect a ratio of 1 WT:2Het:1KO, and that we obtained a total of 191 WT mice, we were expected to recover 382 Het and 191 KO mice.

Table 2
Lethality in Eml3 null mice.
Mouse ageNo. of animalsEml3 genotypep value*
+/++/−−/−
Embryos (dpc)†
11.533982167900.7980
12.51132661260.6988
13.51514767370.1981
18.5144437724‡0.0576
Weaned mice§55719135511<0.0001
  1. *

    A χ2 goodness of fit test was performed at each stage for genotype ratios obtained from heterozygote intercrosses.

  2. †

    Days post coitum.

  3. ‡

    Some advanced resorptions observed but not collected.

  4. §

    Mice were weaned at about three weeks of age.

Table 3
EML3 protein coIP-MS data from three different tissues.
E15.5 whole embryoE12.5 embryo headE12.5 placenta
Eml3−/−Eml3+/+Eml3−/−Eml3+/+Eml3+/+Eml3+/+Eml3AAA/AAA
αEML3αEML3αEML3αEML3αIgGαEML3αEML3
Protein nameSymbolUniProt IDMW (kDa)n = 3n = 3n = 2n = 2n = 2n = 2n = 1
Echinoderm microtubule-associated protein-like 3EML3EMAL3_MOUSE; Q8VC039602880229123380
Tubulin beta-3 chainTUBB3TBB3_MOUSE; Q9ERD75000050000
14-3-3 protein thetaYWHAQ1433T_MOUSE; P682542818000246
14-3-3 protein epsilonYWHAE1433E_MOUSE; P6225929000150120
14-3-3 protein etaYWHAH1433F_MOUSE; P685102800000267
14-3-3 protein zeta/deltaYWHAZ1433Z_MOUSE; P631012800000247
14-3-3 protein gammaYWHAG1433G_MOUSE; P619822800000190
Dynein light chain 1, cytoplasmicDYNLL1DYL1_MOUSE; P6316810011000180
Myosin-10MYH10MYH10_MOUSE; Q6187922910000240
Fibrillin-2FBN2FBN2_MOUSE; Q6155531400020000
Gap junction alpha-1 proteinCJA1CXA1_MOUSE; P2324243000001310
THO complex 4ALYREFQ0VBL5_MOUSE; O085832701300000
60S ribosomal protein L7RPL7RL7_MOUSE; P141483100011000
Integrin alpha-5ITGA5ITA5_MOUSE; P116881150900000
Sfrs10 proteinTRA2BQ5PR75_MOUSE; Q5PR75270800000
NucleolinNCLNUCL_MOUSE; P09405770007000

Additional files

MDAR checklist
https://cdn.elifesciences.org/articles/107102/elife-107102-mdarchecklist1-v1.docx
Reporting standard 1

ARRIVE checklist.

https://cdn.elifesciences.org/articles/107102/elife-107102-repstand1-v1.pdf

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Isabelle Carrier
  2. Eduardo Diez
  3. Valerio E Piscopo
  4. Susanne Bechstedt
  5. Hans van Bokhoven
  6. Myriam Srour
  7. Albert M Berghuis
  8. Stefano Stifani
  9. Yojiro Yamanaka
  10. Roderick McInnes
(2026)
The microtubule-binding protein EML3 is required for mammalian embryonic growth and cerebral cortical development, and Eml3 null mice are a model of cobblestone brain malformation
eLife 14:RP107102.
https://doi.org/10.7554/eLife.107102.3