Transcriptional Regulators with Broad Expression in the Zebrafish Spinal Cord

Abstract

Background

The spinal cord is a crucial part of the vertebrate CNS, controlling movements and receiving and processing sensory information from the trunk and limbs. However, there is much we do not know about how this essential organ develops. Here, we describe expression of 21 transcription factors and one transcriptional regulator in zebrafish spinal cord.

Results

We analyzed the expression of aurkb, foxb1a, foxb1b, her8a, homeza, ivns1abpb, mybl2b, myt1a, nr2f1b, onecut1, sall1a, sall3a, sall3b, sall4, sox2, sox19b, sp8b, tsc22d1, wdhd1, zfhx3b, znf804a, and znf1032 in wild-type and MIB E3 ubiquitin protein ligase 1 zebrafish embryos. While all of these genes are broadly expressed in spinal cord, they have distinct expression patterns from one another. Some are predominantly expressed in progenitor domains, and others in subsets of post-mitotic cells. Given the conservation of spinal cord development, and the transcription factors and transcriptional regulators that orchestrate it, we expect that these genes will have similar spinal cord expression patterns in other vertebrates, including mammals and humans.

Conclusions

Our data identify 22 different transcriptional regulators that are strong candidates for playing different roles in spinal cord development. For several of these genes, this is the first published description of their spinal cord expression.

Introduction

The spinal cord is a crucial part of the Central Nervous System (CNS), responsible for controlling movements and receiving and processing sensory information from the trunk and the limbs. Even though the spinal cord is relatively simple compared to the brain, there are still fundamental gaps in our knowledge of how it develops and appropriate neuronal and glial spinal cell types are made, maintained and connected into appropriate neuronal circuitry. Transcription factors, proteins that bind DNA and regulate gene expression, play crucial roles in these processes ^1–3. In addition, regulatory proteins that modulate the complex organization of chromatin to permit transcription factor activity are also essential for the correct development and differentiation of these cells and tissues. However, we have no idea how many transcription factors and transcriptional regulators are expressed in the spinal cord during development and the temporo-spatial expression patterns of many transcription factors and transcriptional regulators are currently unknown. This significantly impedes our ability to identify the regulatory gene networks that orchestrate spinal cord development. To begin addressing this, as part of a larger gene expression screen to identify transcription factor and transcriptional regulator genes that are expressed in the developing spinal cord ^4,5 we identified 22 genes that are expressed broadly in the zebrafish embryonic spinal cord. When we started this study, the expression of many of these genes was unknown. Since then, some of them have been studied by other groups, although often in other tissues. In most cases, the published expression data is still limited, particularly in the spinal cord and for some of the genes that we analyze in this paper, there are currently no other published reports of their expression in zebrafish embryos. For example, expression of znf1032, has, to our knowledge, never been examined in any vertebrate before. There is also no published zebrafish expression data for homeza, ivns1abpb, mybl2b, onecut1, tsc22d1 or wdhd1, although there are some online photographs available from large scale expression screens ^6–12.

In this report, we describe the spinal cord expression of all 22 of these genes in zebrafish embryos at key developmental stages, and we also show their expression in the brain and other trunk tissues. In addition, we examine whether these genes are expressed in progenitor and/or post-mitotic spinal cells. We analyze their expression in both whole-mount preparations and lateral cross-sections in wild-type (WT) embryos and MIB E3 ubiquitin protein ligase 1 (mib1, formerly called mind bomb) mutants. mib1 encodes a ubiquitin ligase that is part of the Notch pathway, and in mib1^ta52b mutants the vast majority of spinal cord progenitor cells precociously differentiate as early-forming populations of spinal cord post-mitotic neurons, at the expense of later forming neurons and glia ^13–17 This enables us to distinguish between progenitor domain expression (which should be lost in mib1 mutants) and post-mitotic expression (which is often, although not always, expanded into additional cells in mib1 mutants). In combination with analysing WT trunk cross-sections, this allows us to discriminate between expression in progenitor cells and post-mitotic cells. Interestingly, we find that while all of these genes are broadly expressed in the spinal cord, they have distinct expression patterns from one another. Some are expressed in progenitor cells, others in both progenitor cells and different subsets of post-mitotic cells, and others are predominantly expressed in distinct subtypes of post-mitotic cells. This identification of transcription factor and transcriptional regulator genes with distinct expression patterns in progenitor and/or post-mitotic cells is an essential first step in determining the gene regulatory networks that specify the correct development of spinal cord circuitry and function.

Results

We have performed several different high-throughput gene expression screens to identify transcription factor genes that are expressed in the developing spinal cord ^{e.g. 4,5}. In previous studies, we analyzed transcription factors that are expressed by specific types of spinal cord interneurons ^4,5,18–23 However, we have also identified a subset of 21 transcription factor genes and 1 transcriptional regulator gene that our initial analyses suggested were broadly expressed in the zebrafish embryonic spinal cord, but are not ubiquitious throughout the embryo. Each of the transcription factor genes encodes a protein containing a verified DNA-binding domain, according to functional classification by InterPro (Table 1; ²⁴). In addition, we analysed expression of the putative transcriptional regulator gene, aurkb - which is involved in chromatin silencing of differentiated post-mitotic mesenchymal stem cells ²⁵. As discussed in the introduction, spinal cord expression of some of these genes has not been reported before, while for most of the others, there are only limited descriptions of their expression in the published literature (see Discussion for more detail).

Table 1. InterPro Identifier Numbers and Domain Names for Transcription Factor and Transcriptional Regulator Proteins Identified in this Study.

Column 1 lists proteins analyzed in this study. Column 2 indicates whether the protein has chromatin-remodelling or DNA-binding activity. Columns 3 and 4 contain the InterPro identifier numbers and protein domain names for the transcription-related functional domains contained within each protein ²⁴ Column 5 provides references for the papers that document the transcriptional activity of these protein domains and/or the proteins themselves.

Protein Name	Chromatin or DNA-Binding Activity	InterPro Identifier Number	InterPro Protein Domain Name	Reference(s) Describing Transcriptional Activity
Aurkb	Chromatin	IPR000719	Protein kinase domain	1
Foxb1a	DNA	IPR001766	Fork head domain	2
Foxb1b	DNA	IPR001766	Fork head domain	2
Her8a	DNA	IPR001598	Myc-type, basic helix-loop-helix domain	3
Homeza	DNA	IPR001356	Homeobox domain	4
Ivns1abpb	DNA	IPR000210	BTB/POZ domain	5–7
Mybl2b	DNA	IPR001005	SANT/Myb domain	8,9
Myt1a	DNA	IPR002515	Zinc finger, C2H2C-type domain	10–13
Nr2f1b	DNA	IPR001628	Zinc finger, nuclear hormone receptor-type	14,15
Onecut1	DNA	IPR001356	Homeobox domain	4
Sall1a	DNA	IPR013087	Zinc finger C2H2-type domain	10,16
Sall3a	DNA	IPR013087	Zinc finger C2H2-type domain	10,16
Sall3b	DNA	IPR013087	Zinc finger C2H2-type domain	10,16
Sall4	DNA	IPR013087	Zinc finger C2H2-type domain	10,16
Sox2	DNA	IPR009071	High mobility group box domain	17
Sox19b	DNA	IPR009071	High mobility group box domain	17
Sp8b	DNA	IPR013087	Zinc finger C2H2-type domain	10,16
Tsc22d1	DNA	IPR000580	TSC22/Bun domain	18
Wdhd1	DNA	IPR009071	High mobility group box domain	17
Zfhx3b	DNA	IPR001356	Homeobox domain	4
Znf804a	DNA	IPR013087	Zinc finger C2H2-type domain	10,16
Znf1032	DNA	IPR013087	Zinc finger C2H2-type domain	10,16

As these genes may play important roles in spinal cord development, we analyzed their expression in more detail. At 24 hours post fertilization (h), aurkb, foxb1a, foxb1b, her8a, homeza, ivns1abpb, mybl2b, myt1a, nr2f1b, onecut1, sall1a, sall3a, sall3b, sall4, sox2, sox19b, sp8b, tsc22d1, wdhd1, zfhx3b, znf804a, and znf1032 are all expressed throughout the rostro-caudal and dorso-ventral spinal cord (Fig. 1). All of these genes are also broadly expressed throughout the hindbrain (Figs. 1, 2, 3 and 4, for a schematic of anatomical locations see Fig. 1W) and all of them except znf804a and znf1032 (Fig. 1U, V) have distinct, spatially-restricted expression patterns within the fore- and midbrain. In contrast, znf804a and znf1032 are more broadly expressed in these brain regions (Figs. 1 and 3). In addition, aurkb, ivns1abpb, mybl2b, sall4, wdhd1, znf804a, and znf1032 are variably expressed in the blood (located in the dorsal aorta and cardinal vein) beneath the notochord (Fig. 1A, F, G, N, S, U, V, W).

Figure 1. Broad Expression of Transcription Factor and Transcriptional Regulator Genes in Wholemount Zebrafish Embryos at 24 h.

(A-V) Lateral views of wholemount WT zebrafish embryos at 24 h. A minimum of 5 embryos were analysed in detail per gene to determine the representative expression pattern (see Experimental Procedures for further details). (W) Schematic of a lateral view of a 24 h wholemount zebrafish embryo. F = forebrain, M = midbrain, H = hindbrain, V = mid- and hindbrain ventricles, SC = spinal cord, N = notochord, B = blood (black dotted line indicates boundary between dorsal aorta and cardinal vein), Y = yolk and the eye and lens are indicated with concentric blue dotted circles. In all panels, rostral is left and dorsal is up. Transcriptional regulator gene (A) aurkb, and transcription factor genes (B) foxb1a, (C) foxb1b, (D) her8a, (E) homeza, (F) ivns1abpb, (G) mybl2b, (H) myt1a, (I) nr2f1b, (J) onecut1, (K) sall1a, (L) sall3a, (M) sall3b, (N) sall4, (O) sox2 (P), sox19b, (Q) sp8b, (R) tsc22d1, (S) wdhd1, (T) zfhx3b, (U) znf804a, and (V) znf1032 are all broadly expressed throughout the rostro-caudal and dorso-ventral spinal cord and all 22 of these genes are also variably expressed in the brain. (A) aurkb, (F) ivns1abpb, (G) mybl2b, (N) sall4, (S) wdhd1, (U) znf804a, and (V) znf1032 are also variably expressed in the blood beneath the notochord. (N) sall4 in situ hybridization experiments were performed with the molecular crowding reagent Dextran Sulfate (see Experimental Procedures for rationale). All other in situ hybridization experiments in this figure were performed without this reagent. Scale bar: 200 μm.

Figure 2. A Subset of Transcription Factor and Transcriptional Regulator Genes Lose Expression in the Spinal Cord of Zebrafish mib1^ta52b Mutant Embryos at 24 h.

(A-Y) Lateral views of (A, C, F, H, K, M, P, R, U, W) head, (B, D, G, I, L, N, Q, S, V, X) spinal cord, and (E, J, O, T, Y) tail in (A, B, F, G, K, L, P, Q, U, V) sibling and (C-E, H-J, M-O, R-T, W-Y) mib1^ta52b mutant embryos at 24 h. Rostral, left. Dorsal, up. A minimum of 5 embryos were analysed per gene for each genotype to determine representative expression patterns (see Experimental Procedures). None of the in situ hybridization experiments in this figure were performed with the molecular crowding reagent Dextran Sulfate. Scale bar: (A, C, F, H, K, M, P, R, U, W) 50 μm, (B, D, G, I, L, N, Q, S, V, X) 20 μm, and (E, J, O, T, Y) 25 μm.

Figure 3. A Subset of Transcription Factor Genes Show Reduced Expression in the Spinal Cord of Zebrafish mib1^ta52b Mutant Embryos at 24 h.

Lateral views of (A, C, E, G, I, K, M, O, Q, S, U, W, Y, A’, C’, E’, G’, I’, K’, M’, O’, Q’) head, and (B, D, F, H, J, L, N, P, R, T, V, X, Z, B’, D’, F’, H’, J’, L’, N’, P’, R’) spinal cord in (A, B, E, F, I, J, M, N, Q, R, U, V, Y, Z, C’, D’, G’, H’, K’, L’, O’, P’) sibling and (C, D, G, H, K, L, O, P, S, T, W, X, A’, B’, E’, F’, I’, J’, M’, N’, Q’, R’) mib1^ta52b mutant embryos at 24 h. Rostral, left. Dorsal, up. A minimum of 5 embryos were analysed per gene for each genotype to determine representative expression patterns (see Experimental Procedures). (Y-B’) sall4 in situ hybridization experiments were performed with the molecular crowding reagent Dextran Sulfate (see Experimental Procedures for rationale). All other in situ hybridization experiments in this figure were performed without this reagent. Scale bar: (A, C, E, G, I, K, M, O, Q, S, U, W, Y, A’, C’, E’, G’, I’, K’, M’, O’, Q’) 50 μm, (B, D, F, H, J, L, N, P, R, T, V, X, Z, B’, D’, F’, H’, J’, L’, N’, P’, R’) 20 μm.

Figure 4. A Subset of Transcription Factor Genes Show Expanded Expression in the Spinal Cord of Zebrafish mib1^ta52b Mutant Embryos at 24 h.

Lateral views of (A, C, E, G, I, K, M, O, Q, S, U, W) head, and (B, D, F, H, J, L, N, P, R, T, V, X) spinal cord in (A, B, E, F, I, J, M, N, Q, R, U, V) sibling and (C, D, G, H, K, L, O, P, S, T, W, X) mib1^ta52b mutant embryos at 24 h. Rostral, left. Dorsal, up. A minimum of 5 embryos were analysed per gene for each genotype to determine representative expression patterns (see Experimental Procedures). In most of these cases, gene expression can be observed in additional spinal cord regions in mib1^ta52b mutants compared to WT and sibling embryos, because of the expansion of specific populations of post-mitotic cells into neighboring locations along both dorsal-ventral and medial-lateral axes (see Figure 8). None of the in situ hybridization experiments in this figure were performed with the molecular crowding reagent Dextran Sulfate. Scale bar: (A, C, E, G, I, K, M, O, Q, S, U, W) 50 μm, (B, D, F, H, J, L, N, P, R, T, V, X) 20 μm.

The spinal cord expression of aurkb, foxb1a, homeza, myt1a, onecut1, sall1a, sall3b, sall4, sox2, sox19b, sp8b, zfhx3b, and znf1032 persists at 36 h (Fig. 5A, B, E, H, J, K, M–Q, T, V). In contrast, the spinal cord expression of foxb1b, her8a, ivns1abpb, mybl2b, nr2f1b, sall3a, tsc22d1, wdhd1, and znf804a decreases by this stage, either persisting most strongly in ventral spinal cord foxb1b, her8a, ivns1abpb, mybl2b, sall3a, and znf804a), only very weakly in the spinal cord (nr2f1b and tsc22d1), or only in very few spinal cells (wdhd1) (Fig. 5C, D, F, G, I, L, R, S, U). her8a,sox2 and znf804a are now also expressed in neuromasts, deposited at intervals along the length of the embryo by the migrating lateral line primordium (black arrows, Fig. 5W, X, E’). The variable expression of aurkb, ivns1abpb, mybl2b, sall4, wdhd1, znf804a and znf1032 in the blood (dorsal aorta and cardinal vein immediately beneath the notochord) persists at this stage (Fig. 5Z–F’). Interestingly, wdhd1 is expressed in individual blood cells at this stage (Fig. 5D’).

Figure 5. Transcription Factor and Transcriptional Regulator Gene Expression in Zebrafish Spinal Cord and Mesendodermal Tissues at 36 h.

Lateral (A-F’) views of spinal cord (A-V), neuromasts of the lateral line primordium (W, X, E’), the hypochord (Y), and the blood (Z-F’) in WT zebrafish embryos at 36 h. Rostral, left. Dorsal, up. (G’-J’) Dorsal wholemount views of WT zebrafish embryos at 36 h. Rostral, left. A minimum of 5 embryos were analysed per gene to determine the representative expression pattern (see Experimental Procedures). (A-V) In the spinal cord views, the morphological boundary between the ventral spinal cord and the notochord is visible towards the bottom of the panel and the spinal cord is in focus. In panels (W, X, E’) the focal plane is more lateral, and the somites and lateral line primordium are in focus. Black arrows (W, X, E’) indicate neuromasts, deposited at intervals along the length of the embryo by the migrating lateral line primordium. White arrowheads (Y) indicate the hypochord, at the ventral interface between the notochord and the blood. (Z-F’) Expression in the blood is observed in the dorsal aorta (DA, white arrow) and cardinal vein (CV, white arrow), beneath the notochord. (G’-J’) Expression in the pectoral fin buds is indicated (blue arrows). in situ hybridization experiments with (L) sall3a, (M, I’) sall3b, (N, C’) sall4, and (T, J’) zfhx3b were performed with the molecular crowding reagent Dextran Sulfate (see Experimental Procedures for rationale). All other in situ hybridization experiments in this figure were performed without this reagent. Scale bar: (A-F’) 30 μm, (G’-J’) 130 μm.

While we concentrated predominantly on spinal cord expression of these genes, at 36 h we also, as at 24 h, observed different brain expression patterns (Figure 6, for a schematic of anatomical locations within the brain, see Fig. 6W). myt1a, sox19b, sp8b and zfhx3b have distinct expression patterns in the pallium and subpallium of the telencephalon (dorsal forebrain) (Fig. 6H, P, Q, T, W). In the diencephalon (ventral forebrain), homeza, onecut1, sox2 and wdhd1 are expressed in the habenula, and there is variable expression of nr2f1b, onecut1, sox19b, sp8b, wdhd1 and zfhx3b in the preoptic region, sox19b and myt1a in the thalamic region, and aurkb, foxb1a,foxb1b, myt1a, wdhd1 and znf1032 in the hypothalamus (Fig. 6A, B, C, E, H, I, J, O, P, Q, S, T, V, W). We also observed distinct expression patterns within the hindbrain. For example, aurkb, foxb1a and foxb1b are expressed in both the tectum and tegmentum, at the boundary with the hindbrain (Fig. 6A, B, C, W). In contrast, ivns1abpb, mybl2b, myt1a, nr2f1b, sox2, tsc22d1, wdhd1 and zfhx3b are variably expressed in the tectum (dorsal midbrain) (Fig. 6F, G, H, I, O, R, S, T, W). Within the tegmentum (ventral midbrain), myt1a, nr2f1b, sox2 and zfhx3b are more broadly expressed than they are in the tectum, whilst wdhd1 expression is more spatially defined in the tegmentum compared to the tectum (Fig. 6H, I, O, T, W). We also observed expression of onecut1, sox19b, sp8b and znf1032 in the tegmentum (Fig. 6J, P, Q, R, W). Within the hindbrain (cerebellum), myt1a, nr2f1b, sox2, sox19b, sp8b, zfhx3b and znf1032 are widely expressed throughout the structure. In contrast, her8a, homeza, ivns1abpb, mybl2b and wdhd1 are restricted to expression in the rhombomeres (white arrowheads) (Fig. 6D, E, F, G, H, I, O, P, Q, S, T, V, W). In comparison to the genes described above, we observed extensive expression of sall1a, sall3a, sall3b, sall4, tsc22d1 and znf804a throughout all brain regions (Fig. 6K, L, M, N, R, U, W). Lateral to the brain, we also observed expression of aurkb and tsc22d1 in the otic vesicles (white asterisks, Fig. 6A, R), aurkb, onecut1, sall3b and sox2 in the branchial (gill) mesenchyme adjacent to the hindbrain (black arrows) (Fig. 6A, J, M, O), and onecut1 and sp8b in the olfactory bulbs (blue arrowheads) (Fig. 6J and data not shown).

Figure 6. Transcription Factor and Transcriptional Regulator Gene Expression in Zebrafish Brain at 36 h.

(A-V) Lateral views of heads in WT zebrafish embryos at 36 h. A minimum of 5 embryos were analysed per gene (see Experimental Procedures for further details). (W) Schematic of a lateral view of the head of a 36 h zebrafish embryo. The telencephalon (dorsal forebrain, yellow) consists dorsally and ventrally of the pallium (P) and subpallium (SP) respectively. The diencephalon (ventral forebrain, orange) consists of the habenula (H), the hypothalamus (Hyp), the Thalamic region (Th), the Posterior Tuberculum (PT) and the preoptic region (PO). The midbrain (blue) consists dorsally and ventrally of the tectum (Te) and the tegmentum (Tg) resepectively. The hindbrain (green) consists of the cerebellum (C). V = ventricle in the hindbrain. Black dotted line = midbrain-hindbrain boundary. Blue dotted lines = otic vesicle. Transcriptional regulator gene (A) aurkb, and transcription factor genes (B) foxb1a, (C) foxb1b, (D) her8a, (E) homeza, (F) ivns1abpb, (G) mybl2b, (H) myt1a, (I) nr2f1b, (J) onecut1, (K) sall1a, (L) sall3a, (M) sall3b, (N) sall4, (O) sox2 (P), sox19b, (Q) sp8b, (R) tsc22d1, (S) wdhd1, (T) zfhx3b, (U) znf804a, and (V) znf1032 are all all variably expressed in the brain. (D, E, F, G, S) White arrowheads depict weak expression in the rhomobomeres of the hindbrain. (A, R) aurkb and tsc22d1 are expressed in the otic vesicles (white asterisks). (A, J, M, O) aurkb, onecut1, sall3b and sox2 are expressed in the branchial (gill) mesenchyme (black arrow). (J and data not shown) onecut1 and sp8b are expressed in the olfactory bulbs (blue arrowhead). (N) sall4 in situ hybridization experiments were performed with the molecular crowding reagent Dextran Sulfate (see Experimental Procedures for rationale). All other in situ hybridization experiments in this figure were performed without this reagent. Scale bar: 100 μm.

To determine whether each of these genes is expressed in spinal cord progenitor cells and/or post-mitotic cells we analyzed their expression in spinal cord cross-sections and mib mutants at 24 h. Our WT spinal cord cross-sections show that most of these genes, aurkb, foxb1a, foxb1b, her8a, homeza, ivns1abpb, mybl2b, nr2f1b, sall1a, sall3b, sall4, sox2, sox19b, sp8b, tsc22d1, wdhd1, zfhx3b, znf804a, and znf1032, are expressed in both the medial (progenitor), and lateral (postmitotic) domains of the spinal cord (Fig. 7). These genes are also expressed throughout the dorso-ventral axis of the spinal cord, except for sall3b and sp8b, which are only expressed in the ventral two-thirds of the spinal cord (Fig. 7M, Q), andfoxb1b and wdhd1, which are expressed throughout the dorso-ventral medial (progenitor) domains of the spinal cord, but are only expressed in the ventral two-thirds of the lateral (post-mitotic) domains of the spinal cord (Fig. 7C, S). In contrast, myt1a and onecut1 are only expressed in the lateral (post-mitotic) domains of the spinal cord (Fig. 7H, J), and if sall3a is expressed in the medial spinal cord, that expression is much weaker than the lateral expression (Fig. 7L). myt1a and onecut1 expression extends to all dorso-ventral domains of the spinal cord, but sall3a expression is absent in the dorsal-most spinal cord (Fig. 7H, J, L). Consistent with the whole-mount analyses (Fig. 1), aurkb, ivns1abpb, mybl2b, sall4, wdhd1, znf804a, and znf1032 expression is also visible in the blood immediately beneath the notochord (Fig. 7A, F, G, N, S, U, V, W). In addition, sox2 is also expressed in the hypochord, at the ventral interface between the notochord and the blood (Fig. 7O, for a schematic of anatomical locations see Fig. 7W) and tsc22d1 is expressed in the pronephros, which are tubular structures paired ventrally, lateral to the blood, on each side of the cross-section (Fig. 7R, and see Fig. 7W).

Figure 7. Broad Expression of Transcription Factor and Transcriptional Regulator Genes in Zebrafish Spinal Cord at 24 h.

(A-V) Cross-section views of trunk expression of transcriptional regulator gene (A) aurkb, and transcription factor genes (B) foxb1a, (C) foxb1b, (D) her8a, (E) homeza, (F) ivns1abpb, (G) mybl2b, (H) myt1a, (I) nr2f1b, (J) onecut1, (K) sall1a, (L) sall3a, (M) sall3b, (N) sall4, (O) sox2, (P) sox19b, (Q) sp8b, (R) tsc22d1, (S) wdhd1, (T) zfhx3b, (U) znf804a, and (V) znf1032 in WT zebrafish embryos at 24 h. Dorsal, up. A minimum of 5 embryos were analysed per gene to determine the representative expression pattern (see Experimental Procedures). As indicated in the schematic cross-section (W), the spinal cord (SC), is located above the notochord (N), which is above the hypochord (H, indicated with red arrow), dorsal aorta (DA), and the cardinal vein (CV). The somites (S) can be seen on both sides of these tissues and the pronephros tubes (P, indicated with red arrows) are ventral, either side of the cardinal vein. Within the spinal cord, the dotted line indicates the midline, the small oval indicates the central canal and the small black triangles indicate the roof plate and floor plate. (N) sall4 in situ hybridization experiments were performed with the molecular crowding reagent Dextran Sulfate (see Experimental Procedures for rationale). All other in situ hybridization experiments in this figure were performed without this reagent. Scale bar: 30 μm.

To further confirm whether these genes are expressed in progenitor cells or post-mitotic cells, we also examined their expression in 24 h mib1^ta52b mutants. As described in the introduction, the vast majority of spinal cord progenitor cells precociously differentiate as early-forming populations of post-mitotic neurons in these mutants, at the expense of later forming neurons and glia ^13–17. Therefore, spinal cord progenitor domain expression is lost, and post-mitotic expression is usually expanded to additional cells, in these mutants (e.g. see Fig. 8A–D).

Figure 8. Expression of a Subset of Transcription Factor Genes is Either Lost in Progenitor Domains and/or Expanded in Post-Mitotic Domains in the Spinal Cord of Zebrafish mib1^ta52b Mutant Embryos at 24 h.

(A-L) Cross-section views of trunk expression of transcription factor genes (A, B) dbx2 (expressed in a spinal cord progenitor domain), (C-D) elavl3 (expressed by all spinal cord post-mitotic cells), (E) homeza, (F) ivns1abpb, (G) myt1a, (H) onecut1, (I) sall1a, (J) sall3a, (K) sall3b, and (L) tsc22d1 in (A, C) sibling and (B, D, E, F, G, H, I, J, K, L) mib1^ta52b mutant embryos at 24 h. Dorsal, up. A minimum of 5 embryos were analysed per gene per genotype to determine representative expression patterns (see Experimental Procedures). For schematic, please see Fig. 7W. The expression patterns of homeza, ivns1abpb, myt1a, onecut1, sall1a, sall3a, sall3b and tsc22d1 in sibling spinal cords at 24 h (data not shown) is identical to that in 24 h WT embryos shown in Fig. 7. None of the in situ hybridization experiments in this figure were performed with the molecular crowding reagent Dextran Sulfate. Scale bar: 30 μm.

mib1^ta52b mutant embryos and their siblings were processed and stained identically and we analyzed at least 5 embryos per gene per genotype (see Experimental Procedures). We found that expression of five of the genes, aurkb, foxb1b, her8a, wdhd1 and zfhx3b, is lost from the spinal cord in all but the very caudal tail in mib1^ta52b mutants (Fig. 2), suggesting that these genes are all expressed in spinal cord progenitor cells. Interestingly though, brain expression of these genes is largely unchanged in these mutants, with the exception that her8a expression is slightly less intense, and zfhx3b expression is slightly expanded in the telencephalon (Fig. 2M, W). aurkb and wdhd expression in the blood is also unaffected (Fig. 2E, T).

Eleven other genes also lose a lot of their normal spinal cord expression in mib1^ta52b mutants, but in contrast to the five genes discussed above, they are still expressed in distinct subsets of spinal cord cells (Fig. 3). Expression of mybl2b, nr2f1b, sox19b, and foxb1a is lost in all but a few dorsal spinal cord cells, whereas expression of homeza and ivns1abpb persists in more cells but only in dorso-caudal spinal cord (Fig. 3D, H, L, P, T, X). In contrast, small clusters of cells in both dorsal and ventral spinal cord still express sall4, sox2, sp8b, and znf1032, whereas znf804a is only expressed in small clusters of ventral spinal cord cells (Fig. 3B’, F’, J’, N’, R’). These data suggest that these genes are expressed broadly in spinal cord progenitor cells, and also in a small number of post-mitotic spinal cord cells. Interestingly, and in contrast to the five genes discussed above (Fig. 2), all eleven of these genes also have a dramatic reduction in hindbrain expression in mib1^ta52b mutants. However, consistent with the five genes discussed above, their expression in the fore- and midbrain regions is largely unchanged in mib1^ta52b mutants, with the exception of sall4, which is expanded into a slightly larger territory of the dorsal midbrain, and sox2 and sp8b, which are almost lost in the dorsal forebrain (Fig. 3A’, E’, I’). Expression of mybl2b, ivns1abpb, sall4, znf1032, and znf804a in the blood is unaffected in mib1^ta52b mutants (Fig. 3D, X, B’, N’, R’).

Unlike the 16 genes discussed above, expression of the remaining six genes, myt1a, onecut1, sall1a, sall3a, sall3b, and tsc22d1, persists and is expanded into additional cells in the spinal cord of mib1^ta52b mutant embryos (Fig. 4). This suggests that these genes are expressed by post-mitotic spinal cord cells. myt1a, onecut1, sall1a, and sall3a expression is also expanded to include additional cells throughout the brain in mib1^ta52b mutants, including the hindbrain (Fig. 4C, G, K, O). In contrast, sall3b and tsc22d1 expression is only expanded into additional cells in the hindbrain and largely unchanged elsewhere in the brain (Fig. 4S, W).

We also performed cross-sectional analyses for some of the genes that still had spinal cord expression in mib1^ta52b mutants (Fig. 8). As a comparison, we first analyzed the expression of a known spinal cord progenitor domain marker, dbx2, and post-mitotic marker, elavl3, in spinal cord cross-sections (Fig. 8A–D). In WT zebrafish embryos, dbx2 is expressed in the spinal cord progenitor domains pd5, pd6, p0 and p1 at 24 h ^26,27. Our cross-sections show that this progenitor domain expression occupies a medial mid-dorsoventral position in the spinal cord of sibling embryos but is lost completely in the spinal cord of mib1^ta52b mutants (Fig. 8A, B). In contrast, elavl3 is expressed by all post-mitotic neurons in the zebrafish spinal cord ^28–30. In sibling embryo cross sections, elavl3 is expressed by post-mitotic cells in the lateral spinal cord. It is not expressed by progenitor cells in the medial spinal cord. In mib1^ta52b mutant cross-sections, elavl3 expression persists laterally but also expands medially to include every cell in the spinal cord (Fig. 8C, D), consistent with spinal progenitor cells precociously differentiating and becoming post-mitotic in these mutants.

The spinal cord expression of homeza and ivns1abpb in mib1^ta52b mutants appears, in lateral view, to be in specific post-mitotic dorsal spinal cord domains. To further confirm this we analyzed spinal cord cross sections for these two genes. Consistent with the lateral views, the only remaining expression of homeza and ivns1abpb is in the very dorsal spinal cord, where it extends across the medio-lateral axis, as we would expect for a specific population of post-mitotic cells (Fig. 8E, F). This is in contrast to the expression of these genes in WT and sibling embryos, where they are expressed in both medial progenitor, and lateral post-mitotic cells along most of the dorsoventral axis of the spinal cord (Fig. 7E, F and data not shown). This suggests that all progenitor domain expression of these genes has been lost and that post-mitotic expression only persists in the dorsal spinal cord. We also performed cross-sectional analyzes for some of the genes that, from the lateral views, appear to be expressed in post-mitotic cells throughout the dorsal-ventral axis of the spinal cord in mib1^ta52b mutants (Fig. 7). As predicted from the elavl3 results, for each of these genes, we see expression throughout most of the spinal cord, both medially and laterally (Fig. 8 G–L), although, consistent with the lateral view (Fig. 4P), the expression for sall3a (Fig. 8J) is less extensive than that of the other genes.

Discussion and Conclusions

In this paper, while we, for completeness, include some expression data for other tissues, we primarily describe the expression of 21 different transcription factor genes and 1 transcriptional regulator gene in the zebrafish spinal cord. Given the high conservation of spinal cord development and transcription factor expression in vertebrates ^31,32, these data are relevant not only to zebrafish development, but also to other vertebrates, including mammals and humans. All of these genes have broad, but not ubiquitious, expression patterns in the spinal cord. This suggests that they have specific functions in particular subsets of spinal cells, rather than general house-keeping roles in all cells. In order to hypothesize about the possible functions of these genes, we first need to know whether they are expressed in progenitor and/or post-mitotic cells in the spinal cord. At the stages that we examined, in WT embryos, spinal cord progenitor cells are located medially and as cells become post-mitotic, they move laterally. In contrast, spinal cord progenitor domain expression is lost (e.g. dbx2, Fig. 8A, B) and post-mitotic expression (e.g. elavl3, Fig. 8C, D) is usually expanded into additional cells in mib1^ta52b mutants Based on our cross-sectional analyses and the dramatic reduction of their spinal cord expression in mib1^ta52b mutants, we conclude that aurkb, foxb1b, her8a, wdhd1, and zfhx3b are all expressed in spinal cord progenitor cells (Figs. 2 and 7). mybl2b, nr2f1b, sox19b, foxb1a, homeza, ivns1abpb, sall4, sox2, sp8b, znf1032 and znf804a are also predominantly expressed in progenitor cells, but mybl2b, nr2f1b, sox19b, foxb1a, homeza and ivns1abpb are also expressed in some dorsal post-mitotic cells, whereas sall4, sox2, sp8b, and znf1032 are expressed in some post-mitotic cells in both the dorsal and ventral spinal cord and znf804a is expressed in only some ventral spinal cord post-mitotic cells (Figs. 3 and 7). In contrast, myt1a, onecut1, sall1a, sall3a, sall3b, and tsc22d1 are predominantly expressed by different subsets of post-mitotic spinal cord cells, although our cross-sectional analyses suggest that sall1a, sall3b and tsc22d1 are also expressed, at variable extents, in spinal cord progenitor cells (Figs. 4 and 7).

This is the first detailed description of spinal cord expression for most of these genes, in any vertebrate. The exceptions are her8a ^33,34, sall1a ³⁵, sall3a ³⁵, sox2 ^36–39, sox19b ^37,40–42 and sp8b ^43,44 There is also some limited spinal expression data at 24 h for aurkb ^{45,46 supplementary data}, foxb1b ^47,48, nr2f1b ^{49 supplementary data}, sall4 ⁵⁰, zfhx3b ⁵¹, as well as at stages earlier than those we examine in this paper for foxb1a ⁴⁷, foxb1b ⁵², myt1a ⁵³ and sall1a ³⁵. Finally, there are some data at www.ZFIN.org or https://ibcs-bip-web1.ibcs.kit.edu/ffdb/ from large scale zebrafish expression screens, that show spinal cord expression of aurkb, foxb1a, foxb1b, her8a, homeza, ivns1abpb, mybl2b, nr2f1b, onecut1, sall1a, sall3b, sall4, sox2, sox19b, sp8b, tsc22d1, wdhd1 and zfhx3b ^6–11. For homeza, ivns1abpb, mybl2b, onecut1, tsc22d1, and wdhd1, these online database photos are the only spinal cord expression data that we are aware of, outside this paper. To our knowledge, there is no other expression data, in any tissue, in any vertebrate, for znf1032 and no spinal cord expression data in any vertebrate for znf804a, not even in online databases of large-scale expression screens.

For all but one of the genes where there is some additional spinal cord expression data, our results are consistent with these other reports. The one possible exception is sall3b, where a photo from a large scale expression screen shows unrestricted expression, although no lateral views are provided so it is possible that this experiment just had higher levels of background expression than our data ⁷. However, for several genes, the data from the other sources shows weaker / less apparent spinal cord expression than our results and / or does not show spinal cord expression at stages older than 24 h. This is probably because many of these studies were not concentrating specifically on the spinal cord and, therefore, developed their in situ hybridization staining reactions to a level more appropriate for examining expression in other tissues. Spinal cord expression is often weaker than, for example, brain expression, particularly at later stages ^22,23.

Notably, even in cases where some expression in the spinal cord has previously been shown, with the exception of sall1a ³⁵, sall3a ³⁵, sox2 ^39,54 and sp8b ⁴⁴ there are no cross-sectional analyses of spinal expression and, with the exception of her8a ⁵⁵, there is no analysis of spinal cord expression in mib1 mutants, which means that it is often not clear whether the reported spinal cord expression is in progenitor and / or post-mitotic cells. This is a crucial piece of information for considering the functional roles of these genes in the spinal cord. In this paper, we have analyzed these aspects of expression for all 22 genes.

Interestingly, as described above, we find that while all of the genes that we examined are broadly expressed in the spinal cord, most of them have distinct expression patterns from one another. This suggests that they have different, specific functions in the spinal cord. However, for most of these genes, there is currently no data to suggest what their function(s) may be in the spinal cord. The exceptions are aurkb, which may be important for axonal outgrowth of spinal motor neurons ⁵⁶, her8a, which is important for Notch signaling and, hence, neurogenesis e.g. ^34,55,57, onecut1, which data from mouse and chick suggest may be important for motoneuron and V1 interneuron (Renshaw cell) development ^58–60, sox2, which is required for correct differentiation of spinal motoneurons and oligodendrocytes ^39,61,62 and sp8b, which is required for correct specification of the pMN/p3 progenitor domain boundary and hence the correct development of the cells that develop from these domains ^44,63. Interestingly, at earlier stages of spinal cord development, sall4 and sp8b, might belong to a common gene regulatory network as sall4 is thought to regulate pou5f3 expression, which in turn regulates hoxb1a/b expression in the posterior neurectoderm that forms the spinal cord, and hoxb1a/b over-expression can induce expression of sp8b ^64–67. Further eludidation of these functions and the roles of the other genes in spinal cord development await further studies.

In conclusion, this study identifies 21 transcription factor genes and 1 transcriptional regulator gene with specific spinal cord expression patterns, that may have important roles in spinal cord development. In this way, it provides key new knowledge for Developmental Biologists, especially those interested in CNS development, and / or specific transcription factors, regulatory genes, or gene regulatory networks.

Experimental Procedures

Ethics statement

All zebrafish experiments in this research were carried out in accordance with the recommendations and approval of Syracuse University Institutional Animal Care and Use (IACUC) committee.

Zebrafish husbandry and fish lines

Zebrafish (Danio rerio) were maintained on a 14-h light/10-h dark cycle at 28.5°C. Embryos were obtained from natural paired and/or grouped spawnings of WT (AB, TL or AB/TL hybrid) fish, or heterozygous mib1^ta52b mutants ¹⁴.

in situ hybridization

We fixed embryos in 4% paraformaldehyde / phosphate-buffered saline (PBS) and performed single in situ hybridization experiments as previously described ^17,68. For WT experiments, 30 embryos were processed per tube. For mib1^ta52b experiments, 30 embryos from an incross of heterozygous mib1^ta52b mutant parents were also processed per tube. Mendelian genetics predicts that we would, therefore, have about 7-8 mutant embryos per tube. For the experiments in this paper, we had a minimum of 6, and a maximum of 10 mib1^ta52b mutant embryos per tube. Both mutant and sibling embryos were treated identically and stained for the exact same length of time to enable the comparison of expression patterns between genotypes. Since in situ hybridization is not quantitative, when comparing mib mutants and sibling embryos, we concentrated mainly on whether gene expression was expanded into additional cells or lost from specific domains. In some cases, we also saw stronger or weaker staining in the mutants. As the embryos were processed identically, this probably reflects changes in the level of expression, but to be sure of this, more quantitative methods, such as spatial transcriptomics, would need to be used, which are beyond the scope of this study.

For all genes except dbx2, we synthesized in situ RNA riboprobes using the methods described in ¹⁸ using the primers and annealing temperatures shown in Table 2. To avoid cross-reactivity, whenever possible, riboprobes were designed against 3’UTR or coding sequence lacking all conserved protein domains in Pfam ⁶⁹. Primers were designed using Primer3 web version 4.1.0 at https://primer3.ut.ee ^70,71 and the following design parameters: optimum primer size: 22 bp (minimum: 20 bp, maximum: 25 bp), optimum annealing temperature: 58.0°C (minimum: 57.0°C, maximum: 60.0°C), and optimum GC content: 50% (minimum: 40%, maximum: 60%). The preferred product size range was 800-1100 bp. This was not always possible, if there was little or no novel coding and/or 3’ UTR sequence available (see Table 2). The PCR conditions were: 98.0°C for 30 seconds, 35 cycles of: 98.0°C for 10 seconds; annealing temperature in Table 2 for 30 seconds and 72.0°C for 30 seconds, followed by a final extension for 5 minutes at 72.0°C. The PCR product was assessed on a 1% TAE gel, before purifying through phenol:chloroform:isoamyl alcohol extraction and precipitation with 0.2 M NaCl and ice-cold ethanol. If non-specific banding was generated in addition to the desired PCR product, the specific product was purified from the agarose gel using the Monarch DNA Gel Extraction Kit (NEB, T1020S). Each reverse primer contains the T3 RNA Polymerase minimal promoter sequence (shown in bold and underlined in Table 2). The template for the dbx2 in situ RNA riboprobe was prepared from plasmid DNA kindly provided by Gribble and colleagues ²⁶. For all RNA riboprobes except dbx2, in situ probe synthesis was performed using 1 μg purified PCR product, T3 RNA Polymerase (11031171001, Roche) and DIG RNA Labeling Mix (11277073910, Roche). To synthesize the dbx2 RNA riboprobe, 1 μg of linearized plasmid DNA was used, together with T7 RNA Polymerase (10881775001, Roche) and DIG RNA Labeling Mix (11277073910, Roche).

Table 2. Gene Names, ZFIN Identifiers, Primer Sequences and PCR Conditions for in situ Hybridization Riboprobe Synthesis.

Column 1 lists genes analyzed in this study. Column 2 provides the unique ZFIN identification number for each gene. Columns 3 and 4 contain the primer sequences and expected product sizes (in base pairs (bp)) respectively, used to generate templates for anti-sense RNA riboprobe synthesis from 27 h WT cDNA. Bold and underlined text in column 3 shows the T3 RNA Polymerase minimal promoter sequence added to the reverse primers. Column 5 indicates the annealing temperature for the primer pairs shown in each row of column 3. For further conditions for riboprobe synthesis, please see Experimental Procedures.

Gene Name	ZFIN Gene ID	Primer Sequences	PCR Product Size (bp)	Annealing Temperature (°C)
aurkb	ZDB-GENE-020419-40	Forward: CTGGATTAAACAGCCGCCATG	1082	65.0
		Reverse: AATTAACCCTCACTAAAGGGACAGAAGAACAGCTATAATACAGTGGAG
elavl3	ZDB-GENE-980526-76	Forward: GAAGTCCAACCTAACAAACCATGC	1064	65.0
		Reverse: AATTAACCCTCACTAAAGGGATCTTAGTCCCCAACCCTCCTATC
foxb1a	ZDB-GENE-990616-47	Forward: CAGACGCTGCCCACTACCTC	429	67.0
		Reverse: AATTAACCCTCACTAAAGGGAGGCTCAGAGATTGCGGAGAG
foxb1b	ZDB-GENE-990415-77	Forward: AACAACAAGCAAAACTCCGAATG	1029	63.0
		Reverse: AATTAACCCTCACTAAAGGGAGGTAGCCTAAATTCATGACATGC
her8a	ZDB-GENE-030131-2376	Forward: ACCAATGCGTCACTCAGATTTG	853	64.0
		Reverse: AATTAACCCTCACTAAAGGGACCATGTGGCATAAGGAACATTGTC
homeza	ZDB-GENE-030616-592	Forward: GGCATAGCCGCTACAAAAAG	1069	62.0
		Reverse: AATTAACCCTCACTAAAGGGATCTTGCAGTGGTCCTGTCTG
ivns1abpb	ZDB-GENE-030131-6266	Forward: ATTATGCCCGTTCTGGACTG	846	63.0
		Reverse: AATTAACCCTCACTAAAGGGATAAATGGGCTCCACTCGTTC
mybl2b	ZDB-GENE-041007-1	Forward: AAAACCCCGCTCACACAGAAG	507	66.0
		Reverse: AATTAACCCTCACTAAAGGGACAGTGGAGAGATAGTTTTGGGTG
myt1a	ZDB-GENE-030131-3885	Forward: GGCTACACCAAAAGCAGCTC	1065	64.0
		Reverse: AATTAACCCTCACTAAAGGGAAGCTCTAGGGCAACCTGACA
nr2f1b	ZDB-GENE-040426-1438	Forward: AGCTAAATTTGGGCTGTCAGAC	1028	64.0
		Reverse: AATTAACCCTCACTAAAGGGAGCCCCAAATAATCAACCTGCAG
onecut1	ZDB-GENE-040426-1469	Forward: GTGGACACTGCTCCGCTATATC	856	65.0
		Reverse: AATTAACCCTCACTAAAGGGAGGAACGGAGGAGTGGTTCTG
sall1a	ZDB-GENE-020228-2	Forward: CTCTTTAACTGGCACACACACATG	1041	65.0
		Reverse: AATTAACCCTCACTAAAGGGAGAGAAACTGGGCCCTAAAAGATC
sall3a	ZDB-GENE-020228-4	Forward: CGGACATCCCTTCTCCAGATTC	1026	65.0
		Reverse: AATTAACCCTCACTAAAGGGAATTCCTGTGCTACATCCAGAAC
sall3b	ZDB GENE-030131-9140	Forward: CAGTGAAATCAACTCCGAGCTTG	957	64.0
		Reverse: AATTAACCCTCACTAAAGGGAGGCTTGCTGTCAAATATGGAGAC
sall4	ZDB GENE-060328-2	Forward: CTCCGCTCTATGACCCTCAG	1189	64.0
		Reverse: AATTAACCCTCACTAAAGGGAGTGCAAAGTGTCTGGCGATA
sox2	ZDB-GENE-030909-1	Forward: CCCACCTACAGCATGTCCTATTC	970	65.0
		Reverse: AATTAACCCTCACTAAAGGGATCTAACAGATGAAGAGTGGGAGAC
sox19b	ZDB-GENE-010111-1	Forward: TTAAACCAGAACCCCTGTCG	1112	62.0
		Reverse: AATTAACCCTCACTAAAGGGACTGACAGCGAAGATCAGTGC
sp8b	ZDB-GENE-030131-3654	Forward: CGACGTGTAACAAAATCGGGAG	683	64.0
		Reverse: AATTAACCCTCACTAAAGGGAAGACGAGAGAACCGGTTTGAATC
tsc22d1	ZDB-GENE-030131-7785	Forward: GACCTTGGAGTTTCTTTGTGCTG	347	65.0
		Reverse: AATTAACCCTCACTAAAGGGATTCACCAGATCCATGGCTTGTTC
wdhd1	ZDB-GENE-030131-1665	Forward: TTATGGTCACTCTGAAGGCCAC	748	65.0
		Reverse: AATTAACCCTCACTAAAGGGAATCCCAAATCAGCAGGTTTCCTC
zfhx3b	ZDB-GENE-030131-7577	Forward: AGCCCATCCTCATGTGTTTC	1197	63.0
		Reverse: AATTAACCCTCACTAAAGGGAGGTTCTGGGTGCATTCACTT
znf804a	ZDB-GENE-070912-632	Forward: TCATGCAAGCAAGAGAGTGG	1172	63.0
		Reverse: AATTAACCCTCACTAAAGGGACAGTCCTCTGCAGGGCTAAC
znf1032	ZDB-GENE-080305-4	Forward: TATCGTGGTCCCGTAATCTACAC	366	64.0
		Reverse: AATTAACCCTCACTAAAGGGATTCTCGTGTGGTGTTTAAGGTG

Embryos older than 24 h were usually incubated in 0.003% 1-phenyl-2-thiourea (PTU) to prevent pigment formation. For some experiments, we added 5% of Dextran Sulfate to the hybridization buffer. In cases where this was done it is indicated in the relevant figure legend. Dextran sulfate can increase specific staining in in situ hybridization experiments as it facilitates molecular crowding ^72,73.

In cases of low riboprobe hybridization efficiency, we exposed embryos to prolonged staining. In some cases, this produced higher background (diffuse, non-specific staining), especially in the hindbrain, where ventricles can sometimes trap anti-sense riboprobes ²³.

Imaging and Expression Analysis

Prior to imaging, all embryos for each gene were placed in a small dish of 70% glycerol / 30% sterile water, and examined using a stereo microscope (SMZ1000, Nikon), equipped with a gooseneck LED microscope light. We examined the expression pattern of each gene in each embryo to check if it was reproducible. If we observed variable staining within a genotype (for example, due to microclimates being formed during the in situ hybridization process which can cause differential exposure of the embryos to experimental reagents), we repeated the experiment. Once satisfied that we had reproducible staining, a minium of 5 embryos per gene, for each genotype, were analysed further as described below. mib1^ta52b mutant embryos have an obvious morphological phenotype of curled tails and large heads and, in addition, the expression patterns of all of the genes that we examined were clearly changed in these mutants compared to their siblings. Therefore, we could not examine expression patterns blind to genotype.

All embryos were deyolked in 70% glycerol / 30% sterile water using mounting pins. For lateral and dorsal views of the embryo, whole embryos were mounted in 70% glycerol between coverslip sandwiches (24 mm x 60 mm coverslips; VWR, 48393-106), with 2-4 coverslips (22 mm x 22 mm; VWR, 16004-094) on either side of the sample to avoid sample compression. Cross-sections were cut by hand using a razor blade mounted in a 12 cm blade holder (World Precision Instruments, Cat. # 14134). Differential interference contrast (DIC) pictures were taken using an AxioCam MRc5 camera mounted on a Zeiss Axio Imager M1 compound microscope. All images were processed for brightness-contrast and colour balance using Adobe Photoshop software ⁷⁴ Images of control and mib1^ta52b mutant embryos are from the same experiment and they were processed identically. Figures were assembled using Adobe Photoshop ⁷⁴.