Disruption of Fuz in mouse embryos generates hypoplastic hindbrain development and reduced cranial nerve ganglia

Abstract

Background:

The brain and spinal cord formation is initiated in the earliest stages of mammalian pregnancy in a highly organized process known as neurulation. Environmental or genetic interferences can impair neurulation, resulting in clinically significant birth defects known collectively as neural tube defects. The Fuz gene encodes a subunit of the CPLANE complex, a macromolecular planar polarity effector required for ciliogenesis. Ablation of Fuz in mouse embryos results in exencephaly and spina bifida, including dysmorphic craniofacial structures due to defective cilia formation and impaired Sonic Hedgehog signaling.

Results:

We demonstrate that knocking Fuz out during embryonic mouse development results in a hypoplastic hindbrain phenotype, displaying abnormal rhombomeres with reduced length and width. This phenotype is associated with persistent reduction of ventral neuroepithelial stiffness in a notochord adjacent area at the level of the rhombomere 5. The formation of cranial and paravertebral ganglia is also impaired in these embryos.

Conclusions:

This study reveals that hypoplastic hindbrain development, identified by abnormal rhombomere morphology and persistent loss of ventral neuroepithelial stiffness, precedes exencephaly in Fuz ablated murine mutants, indicating that the gene Fuz has a critical function sustaining normal neural tube development and neuronal differentiation.

INTRODUCTION

The World Health Organization estimates that every year, 8 million babies are born worldwide with a birth defect. Cardiac defects, neural tube defects (NTDs), and Down syndrome are described as the most common severe birth defects, which are primarily caused by either genetic mutations, maternal infections, nutritional imbalances, environmental factors, or an interaction between two or more of these factors^{1, 2, 3}.

NTDs are a consequence of disturbances during neurulation, one of the earliest and most critical ontogenetic milestones, set right before organogenesis and continuing until the embryonic structures of the brain and spinal cord are developed⁴. The genetic network required for neurulation has been rigorously investigated in vertebrate models of embryonic development. From these studies, 340 genes were identified as being associated with abnormal neural tube closure phenotypes in different mouse models (MGI mouse phenotype annotation - MP:0003720). The vast majority of studied NTD mutants present with a neural tube phenotype, including exencephaly (225 annotated genes, MP:0000914) and spina bifida (60 annotated genes, MP:0003054), while a minority of the described mutants can be characterized as having craniorachischisis (17 annotated genes, MP:0008784). Remarkably, the craniorachischisis phenotype observed in mouse embryos is associated with genes coding for the “core” Planar Cell Polarity regulators (including Celsr1; Dact1; Dvl1; Dvl2; Dvl3; Med12; Ptk7; Scrib; Sfrp1; Sfrp2; Sfrp5; Vangl1; Vangl2).

The Fuz gene (fuzzy planar cell polarity gene) encodes a planar cell polarity effector protein required for ciliogenesis and the organization of directional cellular movement. Ablation of Fuz in mouse embryos generates several embryonic defects, including defective cilia, abnormal neural tube closure, and dysmorphic craniofacial structures, and variants in this gene have been previously associated with NTDs and craniosynostosis in humans^{5, 6, 7, 8}. Ablation of the Fuz gene in mouse embryos does not generate craniorachischisis instead, the embryos analyzed during organogenesis present with exencephaly and/or spina bifida, which are also found as known phenotypes for mutants in the Sonic Hedgehog (Shh) pathway and in mouse models for ciliopathies. In fact, considering both Shh mutants and ciliopathy models, there are 32 genes annotated on MGI for the development of exencephaly, including Fuz, Gli2, Gli3, Gpr161, Intu, Ptch1, Rab23, Sufu and Zic2 (MP:0000914).

In vertebrates, the evolutionarily conserved abilities of the primary cilia for the transduction of signals that will modulate cellular behavior, perception, and motility after receiving biomechanical and chemical extracellular stimuli, are fundamental requirements for normal development of several anatomical structures, including the brain and the spinal cord^{6, 9, 10}, inner ear¹¹, kidneys¹², larynx¹³, and the eyes¹⁴.

The FUZ protein physically interacts with INTU and WDPCP¹⁵ as individual monomeric subunits to form the CPLANE macromolecular complex (ciliogenesis and planar polarity effector complex), which is required for cilia formation via regulation of apical actin assembly¹⁶, septin modulation¹⁷, assembly of intraflagellar transport complex IFT-A¹⁸ and binding to RAB23 and RSG1 during the final steps in ciliogenesis^{19, 20}. Genetic ablation experiments in the mouse revealed that both Fuz and Intu genes are required for neural tube and craniofacial formation. Abnormalities in these genes are also critical in generating phenotypes linked to cilia defects, including polydactyly, skeletal defects, abnormal Sonic Hedgehog signaling, and GLI3 processing^{21, 22}. Besides WDPCP participation as a core interaction partner of the constituted CPLANE complex, neural tube closure was not yet systematically studied in Wdpcp murine ablated embryos, although they do share a similar craniofacial phenotype, including facial and palate clefts accompanied by polydactyly, abnormal truncus arteriosus, and atrioventricular septal defects when compared with those described as Fuz and Intu mutant phenotypes²³.

Observations of the abnormal neural tube closure phenotypes reported in mouse mutant embryos for the CPLANE proteins and also for the downstream components of the Sonic Hedgehog pathway (Gli3, Gpr161, Rab23, and Sufu), which are not linked to a holoprosencephaly phenotype (Shh, Smo, Stil, Hhat, Cdon, Gas1 – upstream on Shh pathway), strongly suggest that Fuz, Intu, and most probably, Wdpcp, are required during ciliogenesis to sustain proper transduction of the Sonic Hedgehog signaling pathway within neuroepithelial cells.

Fuz is a highly conserved gene, comprehensively studied in different model organisms such as mice, Zebrafish, and Xenopus^{6, 7, 10, 24}. As the ablation of Fuz is not linked to craniorachischisis, the typical PCP mutant phenotype, instead it has been associated with exencephaly and abnormal regulation of ciliogenesis and ciliary function, we aimed to understand the function of Fuz regulating the structural properties of neuroepithelial tissues during neural tube closure, and most importantly the influences of tissue stiffness during hindbrain morphogenesis.

Using optical coherence tomography (OCT), Brillouin light scattering²⁵, and confocal microscopy, we observed that the disruption of Fuz generates abnormal rhombomeres resulting in a hypoplastic hindbrain phenotype. This trait is accompanied by persistent loss of ventral neuroepithelial stiffness, specifically at the rhombomere 5 (r5). Concomitantly, the interconnected nature of adjacent neuronal structures derived from the neural tube is impaired during cranial ganglia formation, emphasizing the crucial function of Fuz sustaining neuroepithelial integrity during neural tube closure and development.

RESULTS

Disruption of Fuz generates hypoplastic hindbrain and abnormal rhombomeres with diminished length and width.

The rostral region of the neural plate will establish the forebrain, midbrain, and hindbrain. The hindbrain is segmented into rhombomeres, which will guide the organization of neuronal projections throughout the adjacent craniofacial tissues²⁶. In this study, we observed the hindbrain development of Fuz ablated mouse embryos at E9.5 and E10.5 using confocal microscopy and optical coherence tomography (OCT) images (Figure 1 b–j, Supplemental File Figures S1 and S2). Image quantification of the distances between the rhombomere borders indicates that the hindbrain of Fuz mutant embryos does not develop normally, with abnormal rhombomeres due to their diminished length and width. Considering the most anterior even-numbered rhombomeres, r2, and r4, this difference in the measured length, from left to right borders, can be detected as early as E9.5 and becomes more accentuated at E10.5 (Figure 1 m, o, q - Rhombomere Length Shift).

Figure 1. Fuz mutant embryos develop abnormal rhombomeres with diminished length and width.

(a) Illustration of the mouse embryonic hindbrain showing the identification of wildtype and mutant rhombomeres. (b-j) Confocal imaging representing the hindbrains of Fuz embryos at E10.5; (b, c, d) PAX6 and TUBB3 merged channels; (e, f, g) PAX6 immunostaining; (h, i, j) TUBB3 immunostaining. (k) Correlation of phenotype and genotype of offspring from Fuz heterozygous intercross. (l) Rhombomeres 1 to rhombomeres 3 width shift measured in pixels, (m) Rhombomeres 2 Length shift measured in pixels, (n) Rhombomeres 3 to rhombomeres 5 width shift (pixels), (o) Rhombomeres 4 Length shift (pixels), (p) Rhombomeres 6 to rhombomeres 8 width shift (pixels), (q) Rhombomeres 6 Length shift (pixels). The rhombomeres width and length were measured from at least four embryos per genotype (Supplemental File, Figure S2). * and **** indicate P < 0.05 and P <0.0001, respectively, by One-Way ANOVA and Bonferroni’s comparisons test. The yellow arrows indicate the rhombomeres 1 to rhombomeres 3 width (b, c, d). The orange arrows indicate the right side of rhombomeres 2 (b, c, d). The black arrows (e-j) indicate PAX6 and TUBB3 positive cells at the level of the rhombomere 6. The scale bar is 0.5 mm.

The distance between boundaries of the odd-numbered rhombomeres r1 to r3, and r3 to r5, and between the more posterior even-numbered rhombomeres, r6 to r8, was also measured. The decreased boundary distance between the odd rhombomeres r1 to r3 can be noticed at E9.5, while decreased boundary distances from r1 to r3, r3 to r5 and r6 to r8 become more accentuated at E10.5, indicating that the overall tissue structure of the hindbrain does not develop to its full potential in the absence of the Fuz gene product (Figure 1 l, n, p - Rhombomere Width Shift).

Fuz mutant embryos exhibit persistent loss of ventral neural fold stiffness.

In vertebrate embryos, neurons descend from three different lineages arising from the neural tube neuroepithelium, neural crest, and cranial placodes, all of which are derived from the ectodermal germinal layer^{27, 28, 29}. Defective cilium impairs SHH signaling, and as a consequence, disruption of normal development of ventral neuroepithelial cells is observed during neurulation^{30, 31, 32, 33}.

We focused on measuring hindbrain neuroepithelial stiffness at the level of rhombomere 5, in an area adjacent to the otic pits, a primordial embryonic structure derived from the ectodermal placodes, which give rise to the inner ear and is located in a hindbrain region that does not provide a significant contribution of neural crest cells to the mesenchyme³⁴. Our stiffness measurements during neurulation indicate that tissue stiffness decreases in the neural folds at E9.5 in all of the mutants analyzed (Figure 2), while at E10.5 the dorsal areas of the neural folds partially recover stiffness (Figure 3 r), indicating a persistent loss of ventral neural fold stiffness throughout neural tube development in Fuz ablated embryos (Figure 3 p–q).

Figure 2. Fuz mutant embryos present with decreased neural fold stiffness at E9.5.

(a) Illustration of the Experimental Scheme for Tissue Stiffness Measurements using the Brillouin-OCT Multimodal Imaging System. (b-f) 3D-OCT images show the hindbrains of Fuz embryos at E 9.5. (g-k) 2D-OCT optical sections show the neural folds at the rhombomere 5. (l-p) Brillouin frequency shift images represent tissue stiffness in the anatomical areas identified by 2D-OCT optical sections. Region-wise average Brillouin frequency shift of wildtype (+/+), heterozygous (+/−), and nullizygous (−/−) embryos at the (q) Medial Hingepoint, (r) Ventral Neural Fold, (s) Dorsal Neural Fold, (t) Surface Ectoderm and (u) Otic Pit regions, (v) Brillouin frequency shift Color Bar. The stiffness measurements by OCT-Brillouin were repeated at least five times per genotype. * and **** indicate P < 0.05 and P <0.0001, respectively, by One-Way ANOVA and Bonferroni’s comparisons test. The scale bar is 0.25 mm.

Figure 3. Fuz mutant embryos exhibit persistent loss of ventral neural fold stiffness at E10.5.

(a-e) 3D-OCT images show the hindbrains of Fuz embryos at E10.5. (f-j) 2D-OCT optical sections show the neural folds at the rhombomere 5. (k-o) Brillouin frequency shift images represent tissue stiffness in the anatomical areas identified by 2D-OCT optical sections. Region-wise average Brillouin frequency shift of wildtype (+/+), heterozygous (+/−), and nullizygous (−/−) embryos at the (p) Medial Hingepoint, (q) Ventral Neural Fold, (r) Dorsal Neural Fold, (s) Surface Ectoderm and (t) Otic Pit regions. (u) Brillouin frequency shift Color Bar. The stiffness measurements by OCT-Brillouin were repeated at least five times per genotype. * and **** indicate P < 0.05 and P <0.0001, respectively, by One-Way ANOVA and Bonferroni’s comparisons test. The scale bar is 0.5 mm.

Cranial and paravertebral ganglia development is reduced in Fuz ablated embryos.

Given the importance of Fuz during neuroepithelial development, we hypothesized that neuronal differentiation would decrease in Fuz ablated embryos. Therefore, the neural fate from PAX6 positive derived cells would be expected to be disrupted in the cranial ganglia and posterior areas along the rostral caudal axis. In support of this idea, we performed immunostaining to detect PAX6 and TUBB3 (neurofilament - Tuj1 monoclonal serum). Fuz ablated embryos present a decrease in Pax6 positive cells and reduced cranial ganglia development consequently the pattern of differentiated neurons along the body axis is significantly decreased (Figure1 b–j; Figure 4 a–i), indicating that the function of FUZ in ciliogenesis is a critical requirement preceding neurogenesis that originated from neural plate derived tissues.

Figure 4. Fuz mutant embryos exhibit reduced cranial and paravertebral ganglia.

Confocal imaging representing the cranial view of Fuz embryos at E10.5. (a, b, c) Cranial merged channels; (d, e, f) Cranial PAX6 immunostaining; (g, h, i) Cranial TUBB3 immunostaining. The immunostainings were repeated at least three times per genotype. The white arrows (a, b, c) indicate PAX6 positive cells in the eyes. The black arrows (d, e, f) indicate PAX6 positive cells in the maxillary prominences. The orange arrows (a, b, c) and black arrows (g, h, i) indicate TUBB3 positive cells in the cranial ganglia. The scale bar is 0.5 mm.

SOX10 reveals abnormalities during cranial ganglia patterning in Fuz ablated embryos.

The cranial ganglia, including the trigeminal ganglia, is derived from cranial neural crest and ectodermal placode cell lineages³⁵. After observing the persistent loss of ventral neural fold stiffness at the level of the rhombomere 5 in Fuz ablated embryos (Figure 3q), we decided to investigate trigeminal ganglia development. Cranial neural crest cells, in their early migration from rhombomeres 2 and 4 toward the trigeminal ganglia, express a number of genes, including Sox10^{36, 37, 38}. We hypothesized that sensory neurons and gliogenic derivatives from SOX10 positive neural crest cells would be compromised in the posterior areas along the rostral caudal axis. We observed an abnormal diffuse pattern of SOX10 positive trigeminal ganglia in an area adjacent to the first pharyngeal arch, where the neural crest cells accumulate in a region close to the maxillary prominences during trigeminal ganglia development (arrows in Figure 5 d–f).

Figure 5. Fuz mutant embryos exhibit abnormal SOX10 positive cranial neural crest migration.

Confocal imaging representing the cranial view of Fuz embryos at E11. (a, b, c) Cranial view of PAX3 and SOX10 merged channels. (d, e, f) Cranial SOX10 immunostaining. (g, h, i) Cranial PAX3 immunostaining. The immunostainings were repeated at least three times per genotype. The white arrows (a, b, c) and black arrows (d, e, f) indicate SOX10 positive cells in the maxillary prominences. The orange arrows (a, b, c) and black arrows (g, h, i) indicate PAX3 positive cells at the midbrain-hindbrain boundary. The scale bar is 0.5 mm.

PAX3 exposes malformations within the midbrain-hindbrain boundary in the absence of Fuz

At the most anterior level of the hindbrain, in an area adjacent to rhombomere 1 (r1), the isthmic organizer is formed at the midbrain-hindbrain boundary³⁹. Crucial molecular components such as Shh, Fgf8, Pax3, and Wnt1 are expressed in this transient signaling center to delineate the tectum and cerebellum^{40, 41}. Malformations of the isthmic organizer result in anterior NTDs, such as exencephaly, a severe anomaly characterized by the inversion of cranial neural folds⁴². To gain insight into the midbrain-hindbrain boundary formation in the absence of Fuz, we performed PAX3 immunostainings. The typical PAX3 staining pattern observed in the wildtype decreases in the anterior regions of the neural tube, indicating that the crucial role of PAX3 delineating the midbrain-hindbrain boundary during isthmus organizer formation is disrupted in Fuz ablated embryos at E10.5 (arrows in Figure 5 g–i).

DISCUSSION

Primary neurulation begins when the ectoderm gives rise to a flat sheet of cells forming a neural plate. The midline cells along the neural plate will then express Planar Cell Polarity (PCP) genes to coordinate convergent and extension movements, following a cranial-caudal orientation, which makes them form an elongated but, at the same time, narrower neural groove. Consistent with observations reported in previous studies^{6, 13, 24} the Fuz ablated embryos do not present with craniorachischisis, which is the characteristic phenotype observed in mutant mouse embryos for PCP regulatory genes (Celsr1; Dact1; Dvl1; Dvl2; Dvl3; Ptk7; Scrib; Vangl1; Vangl2). Instead, Fuz ablated embryos present with exencephaly and/or spina bifida, which is associated with mutant mouse models for genes involved in ciliogenesis or transduction of Sonic Hedgehog signaling (Fuz, Intu, Gli3, Gpr161, Rab23, and Sufu). Our observations of Fuz ablated mouse embryos during hindbrain development at E9.5 and E10.5 indicate that the transient division of the hindbrain, defining eight transverse segments termed rhombomeres (r1 to r8), is abnormal, distinguished by diminished length and width between rhombomere borders (Figure 1 a–q, Supplemental Files Figures S1 and S2). This observation suggests a hypoplastic development of the hindbrain in the absence of Fuz, where the unique molecular cues expressed by each rhombomere would be disturbed, disrupting the generation of cranial nerves and the spatial organization of neural crest cell migration.

In most vertebrate models of neurulation, ventral sources of Sonic Hedgehog signaling stimulate the proliferation of the left and right neural progenitors, elevating two apposed neuroepithelial folds from the floor plate. The specification of the roof plate is then mainly driven by BMP signaling inhibition and WNT signaling activation until the two layers of apposed neuroepithelia grow enough to reach each other dorsally, forming closure points denominated neuropores. As the notochord provides the initial source of secreted SHH, the neural fold cells positioned ventrally will be exposed to higher concentrations of SHH than the dorsally located cells, establishing a morphogenetic gradient during neural fold growth^{24, 43, 44}.

To better understand how biomechanical forces can be altered on the verge of causing neural tube defects, it is fundamental to develop technologies capable of measuring and mapping mechanical forces guiding morphogenesis. The most traditional techniques providing high-resolution maps, such as atomic force microscopy, magnetic-bead twisting, optical tweezers, and micropipette aspiration⁴⁵, require contact and, therefore, are limited to live measurements. Non-contact methods, including acoustic microscopy⁴⁶ or elastography⁴⁷, regardless of significant advances in optical coherence elastography technology^{48, 49}, are limited for embryonic studies due to their lack of spatial resolution and mechanical accuracy.

The images represented in Figures 2 and 3 exhibit the use of a novel, multimodal imaging technique combining high-resolution optical coherence tomography (OCT) and Brillouin microscopy²⁵ to evaluate tissue structure and stiffness during hindbrain neural fold development in Fuz ablated embryos. In advance of real-time Brillouin imaging, OCT was used to detect three-dimensional details (~10 μm spatial resolution) along rhombomere 5 (r5) and the otic pits (auditory pit), defining a transverse axis to measure tissue stiffness based on the frequency shift of Brillouin light scattering (GHz). Our data shows that neuroepithelial stiffness is decreased in the absence of Fuz at E9.5 (Figure 2 q–s), while the most dorsal areas of the neural folds partially recover stiffness at E10.5 (Figure 3 r), indicating persistent loss of ventral neural fold stiffness throughout neurulation (Figure 2 q–r; Figure 3 p–q).

Previous studies in mouse embryos have shown that ablation of genes with a function upstream of the SHH signaling pathway reduces the strength of SHH signaling, being associated with the midline defect holoprosencephaly⁴⁴, including Shh⁵⁰, Smo⁵¹, Stil⁵², Hhat⁵³, Cdon⁵⁴ and Gas1⁵⁵ (MP:0005157), while ablation of genes with an inhibitory action downstream of SHH, including Gli3⁵⁶, Gpr161⁵⁷, Rab23⁵⁸ and Sufu⁵¹, can lead to an increase in SHH signaling and are associated with exencephaly and spina bifida phenotypes (MP:0000914 and MP:0003054). Interestingly, spina bifida and exencephaly are also observed after ablation of genes required for the last steps during ciliogenesis including Dync2i2⁵⁹, Fuz⁶, Ift122⁶⁰, Ift140⁶¹, Ift57⁶², Intu²² and Kif7⁶³ (MP:0013202).

It is well described that ciliopathies are associated with the disruption of genes downstream of the SHH signaling pathway^{32, 64}. Primary cilia function as cellular antennas, transducing extracellular signals as well as SHH signaling. Genes acting downstream to SHH play a role in ciliogenesis and ciliary function, establishing a connection between SHH signaling and the normal structure and function of cilia within cells. As downstream mediators of planar cell polarity within the CPLANE are basally located at the ciliary bodies, disruptions in this complex contribute to phenotypes associated with ciliopathies^{18, 64, 65}.

Our stiffness measurements at E9.5 and E10.5 point towards a direction where, in the absence of Fuz, impaired cilia development disrupts the transduction of Sonic Hedgehog signaling. Considering our initial identification of a hypoplastic hindbrain phenotype (Figure 1) coupled with impaired neuroepithelial floor plate development (Figures 2 and 3), the distribution of mechanical forces governing neurulation undergoes significant alterations in the absence of Fuz in this mouse line (Figures 1, 2 and 3).

When studying embryonic mouse mutants, it is essential to recognize that mid-gestation lethality can mask NTDs, and careful observation is needed to ensure that a specific phenotype is responsible for disturbing normal neurulation⁴. Neuroepithelial cells rely on the primary cilium for transduction of SHH signaling, which is required to maintain neuroepithelial cells as cycling neural progenitors. During neuronal differentiation, the cell bodies of neuroepithelial cells are translocated toward their basal poles until they delaminate from the neuroepithelium in a process involving loss of apical polarity and primary cilium disassembly⁶⁶. Our observations using PAX6 and TUBB3 staining to map neuronal differentiation indicate that normal development of the cranial ganglia is impaired in Fuz mutant embryos, and the pattern of differentiated neurons along the body axis is significantly decreased, indicating that the function of Fuz in neuroepithelial cells is required for neurogenesis (Figure 4 d–i).

Immunostaining for SOX10 also indicates impaired development of cranial ganglia, including more ventral areas along the dorsal-ventral axis (Figure 5 d–f). Cranial neural crest cells contribute to the cranial ganglia, primarily delaminating from the even-numbered rhombomeres 2 and 4 towards the first and second pharyngeal arches, respectively, while the population of neural crest cells derived from the rhombomeres 6 until 8 will migrate posteriorly, towards the third to sixth pharyngeal arches^{67, 68, 69}. Considering the hypoplastic hindbrain development in Fuz ablated embryos, presenting abnormal rhombomere segmentation, we have also performed PAX3 immunostaining. The typical PAX3 pattern is disrupted at the midbrain-hindbrain boundary in the Fuz ablated embryos, indicating abnormalities during isthmic organizer formation (Figure 5 g–i).

As the neural plate elongates during neurulation, the diencephalon (forebrain), mesencephalon (midbrain), and rhombencephalon (hindbrain) are defined anteriorly at the neuraxis. The boundary established between the mesencephalon and rhombencephalon forms the isthmic organizer, a transient signaling center, which works to regulate the development of anatomical structures derived from the midbrain-hindbrain boundaries. Malformations of the isthmic organizer lead to anterior NTDs, including exencephaly, a severe type of NTD in which the cranial neural folds are overturned, also presenting unfused skull bones, which results in brain exposure and early lethality⁴². This work shows for the first time that hypoplastic hindbrain development, presenting abnormal rhombomere morphology and persistent loss of ventral neuroepithelial stiffness, precedes exencephaly in Fuz ablated murine embryos.

EXPERIMENTAL PROCEDURES

Generation and Genotyping the Knockout Fuz^Gt1(neo) Mice

The Fuz^Gt1(neo) mutant mice were generated by Lexicon Genetics Inc. (The Woodlands, TX, USA) using embryonic stem cells (ES) derived from clone OST180427 identified by an OmniBank sequence tag. Gene trapping involved the strategic insertion of a vector into the second exon of the Fuz gene within ES cells. These modified ES cells were selected for blastocyst microinjection into C57BL/6 mice, leading to the development of chimeric mice. To assess the germline transmission of the targeted allele, chimeric males were mated with C57BL/6 females. Subsequently, heterozygous Fuz^Gt1(neo) males were bred with heterozygous Fuz^Gt1(neo) females, resulting in the generation of homozygous Fuz^Gt1(neo) mice. All experimental mice in this study exhibited a mixed background, combining genetic elements from strains 129 and C57.

Genotyping

Genomic DNA extraction from tail samples was performed using the DirectPCR-tail kit (Viagen Biotech Inc., Los Angeles, CA). The extracted DNA was used for PCR genotyping. The amplification of the wildtype (wt) allele yielded a product of 295 base pairs, utilizing the Fuz forward primer (5’-AGTAGAGGCTCG GAGCCTTTAGG-3’) and Fuz reverse primer (5’- TCACCTAAGCCAGGAACCACTGC-3’). The mutant allele (Gt1) resulted in a 220 base pair product, employing the Fuz forward primer (5’-AGTAGAGG CTCGGAGCCTTTAGG-3’) and LTR reverse primer (3’-ATAAACCCTCTTGCAGTTGCATC-3’).

Generation of Experimental Animals

The Fuz^Gt1(neo) mutant mice were housed within the Baylor College of Medicine Vivarium, fully accredited by AAALAC. The animals were sustained in a carefully controlled environment, utilizing clear polycarbonate microisolator cages and ad libitum access to both food and water (Harlan Teklad Rodent Diet no. 8606, Ralston Purina, St. Louis, MO). The animals were maintained in a 12-hour light-dark cycle. Nulligravid heterozygous Fuz^Gt1(neo) females aged 50–70 days were mated overnight with heterozygous Fuz^Gt1(neo) males. The following morning, these females were examined for the presence of vaginal plugs. The initiation of gestation was designated as 10 p.m. of the previous night, aligning with the midpoint of the dark cycle (Snell GD, 1948), and the identification of a vaginal plug was demarcated as embryonic day E0.5. Pregnant females were euthanized by CO2 asphyxiation on gestational days E8.5, E9.5, E10.5, and E11.0, and the embryos were collected by the caesarian section. Embryos were kept in 100% rat serum at Baylor College of Medicine. The fresh embryos were transferred to the University of Houston within 30 minutes. The yolk sac was dissected for genotyping, and the embryos were mounted to expose the hindbrain neuropore for Brillouin-OCT measurements.

Multimodal Brillouin-OCT system

The multimodal Brillouin-OCT system has been described in a previous publication²⁵. Briefly, the home-built system consists of a swept-source OCT sub-system, a dual VIPA-based Brillouin microscopy sub-system, and a combined scanning arm. The Brillouin microscopy system consisted of a 660 nm single-mode laser source (Torus, Laser Quantum., Inc, CA). The incident power on the sample was 35 mW. The collected backscattered light from the sample was transferred to the dual VIPA spectrometer⁷⁰, and an electron-multiplying charge-coupled device camera was used to detect the Brillouin frequency shift of the sample. The camera acquisition time was 0.2 s per spectrum acquisition. Before every measurement, the system was calibrated with standard materials such as water, acetone, and methanol. The sample was imaged with a microscope objective with 0.25 NA, resulting in an axial resolution of ~36 μm and lateral resolution of ~3.8 μm. The swept source OCT system had a central wavelength of ~1310 nm, scan rate of ~50 kHz, scan range of ~105 nm, and ~8 mW incident power on the sample. The lateral and axial resolutions were ~17.5 μm and ~10 μm in air, respectively. Light from both systems was combined using a dichroic mirror, and galvanometer-mounted mirrors scanned the beam across the sample. For Brillouin imaging, the sample was stepped by a motorized vertical stage. A custom instrumentation software that utilizes the OCT structural image to guide Brillouin imaging was developed.

Immunohistochemistry

Pax6 (42–6600, ThermoFisher) and Tubb3 (2H3-Tuj1, DSHB); Sox10 and Pax3 (DSHB) whole-mount immunostaining was performed on E10.5 embryos, after fixation for 30 minutes in 4% PFA. To remove the fixative agent, the embryos were washed three times in PBS for five minutes and a consecutive dehydration step to 100% methanol. Endogenous peroxidase enzymatic activity was inhibited by incubation in Dent’s bleach (MeOH:DMSO:30%H₂O₂; 4:1:1) for 1 hour at room temperature. Embryos were rehydrated in TBS containing 0.1% Tween-20 (TBST) for 30 minutes and incubated in a 1:500 dilution of anti-Pax6 and anti-Tubb3 overnight at room temperature. Embryos were washed five times in PBST for five minutes and then incubated overnight at room temperature, with a 1:1000 dilution of fluorescent secondary antibodies in TBST. The embryos were washed five times in TBST the next day for five minutes. Before general cell visualization, the embryos were incubated overnight in DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride; 2 μg/ml) to label all the nuclei. Before visualization on the microscope, the embryos were cleared in 25% glycerol and imaged using a Nikon CSU-W1 Yokogawa spinning disc confocal. The obtained Z-stacks were projected at maximum intensity using the Nikon NIS-Elements software and exported as TIFF files⁶⁹. Single-channel images were converted to black and white and then inverted to represent embryos with white backgrounds and the respective immunostaining in black.

STATISTICAL ANALYSIS

Statistics

Significance levels were assessed using One-Way ANOVA and Bonferroni’s comparisons test for statistical analysis. * and **** indicate P < 0.05 and P <0.0001, respectively. Differences were considered statistically significant at * and ****, indicating P < 0.05 and P <0.0001. All calculations were achieved using the GraphPad Prism Software.