Developing neurons undergo significant changes in morphology resulting from cell soma translocation and the rapid extension of dendritic and axonal branches. Exocytosis is a critical regulator of these morphological changes, allowing developing neurons to increase cell surface area in growing neurites with the repeated insertion of cell membrane (Kee et al., 1997; Hazuka et al., 1999). Some components of exocytotic machinery have been shown to be required for proper neuron outgrowth in cultured neurons (Vega and Hsu, 2001) and normal nervous system development in vivo (Friedrich et al., 1997).
Whereas many of these exocytotic regulators are conserved from yeast to mammals, some specialized molecules have evolved to regulate neurotransmitter and neuropeptide release at mature neuronal synapses. Mutants for several synaptic vesicle-associated proteins molecules display neurotransmission phenotypes with no apparent disturbance in neuronal development (Janz et al., 1999; Verhage et al., 2000; Varoqueaux et al., 2002; Murthy et al., 2003). This finding is true for the Synaptic Vesicle 2 (SV2) genes, a family of vertebrate-specific genes that encode synaptic vesicle-associated proteins present at virtually all sites of synaptic vesicle and neuropeptide release. During embryogenesis, SV2 is only weakly detected in discrete locations of the developing nervous system (Janz et al., 1998; Wang et al., 2003). This finding is consistent with the fact that SV2 knockout mice display normal brain morphology and synapse structure at birth, although seizure activity in these mutants ultimately results in death (Janz et al., 1999). The question remains whether there are synaptic vesicle-associated proteins that are expressed in the developing nervous system and potentially play a role in neural development.
SVtwO-related Protein (SVOP) is distantly related to members of the SV2 family, and all share a predicted 12 transmembrane domain structure, indicating these proteins may function as transporters. Recent evidence suggests that SV2 proteins may function as calcium transporters in some capacity, but no function for SVOP has yet been identified (Sudhof, 2004). Unlike the SV2 proteins, rat SVOP is detectable at significant levels in the brain before birth, although a developmental expression analysis has not been performed in any species (Janz et al., 1998). In addition, we identified a fragment of the Xenopus SVOP gene in a screen for immediate transcriptional targets for the proneural basic helix–loop–helix (bHLH) factors Xath5 and XNeuroD, suggesting a role for this gene at early stages of neural differentiation (Logan et al., in press). Here, we describe the cloning of the Xenopus laevis ortholog of SVOP and show that it is expressed throughout neurogenesis in the developing nervous system, including the spinal cord, brain, and trigeminal ganglion. Our observations suggest that SVOP may make a contribution to neuronal development that is conserved across species.
RESULTS AND DISCUSSION
We first identified a fragment of the Xenopus laevis SVOP gene in a differential screen for immediate transcriptional targets for the proneural bHLH factors Xath5 and XNeuroD, suggesting that it may be expressed at early stages of neuronal differentiation downstream of proneural bHLH factors (Logan et al., in press). We, therefore, performed a cDNA library screen using a stage 28–30 Xenopus head library and isolated a 3,041-basepair cDNA clone that contained a predicted open reading frame encoding a predicted 548 amino acid protein. BLAST analysis revealed that this predicted protein matched the SVOP protein from multiple species with high identity (Fig. 1). BLAST analysis of the Xenopus tropicalis genome (v4.1 assembly) identified a genomic scaffold containing regions at least 88% identical to the Xenopus laevis SVOP cDNA sequence (Scaffold_17, 188881:212507). However, these regions did not generate a complete putative coding sequence for Xenopus tropicalis and, thus, were not included in our subsequent sequence analysis. Using both vertebrate and invertebrate SVOP ortholog sequences, a dendrogram was constructed with a ClustalW neighbor-joining program and this analysis showed high identity between Xenopus laevis SVOP and SVOP from other vertebrates (Fig. 1). Bootstrap analysis confirmed a high fidelity (>97%) of predicted branch points within the tree.
A ClustalW sequence alignment of Xenopus SVOP with SVOP proteins from Caenorhabditis elegans, Drosophila, chick, mouse, rat, and human is presented in Figure 2. Twelve putative transmembrane domains, which are based on predictions by Janz et al. (1998), are indicated. These putative transmembrane domains are also found in related SV2 proteins and suggest homology with bacterial carbohydrate transporters (Feany et al., 1992; Gingrich et al., 1992; Janz et al., 1998). Both SVOP and the SV2 family members also contain a conserved pair of negatively charged amino acids within the putative first transmembrane domain, suggesting that they may transport a positively charged molecule (Janz et al., 1998, 1999). We found that these residues are also conserved in the Xenopus SVOP sequence (asterisks in Fig. 2).
Because there has been no published analysis of developmental expression for SVOP, we sought to determine the developmental time course of SVOP expression by performing reverse transcriptase-polymerase chain reaction (RT-PCR) on cDNA generated from whole embryos collected at various developmental time points, and these results were confirmed by performing real-time quantitative RT-PCR (qRT-PCR) on independent samples. qRT-PCR revealed no significant expression at cleavage stages (16-cell), suggesting a lack of maternal mRNA contribution (Fig. 3B), and no SVOP expression was detectable at stage 10–10.5 (gastrula stage), indicating no expression before formation of the neural plate (Fig. 3A,B). SVOP expression was detectable beginning at stage 12 (early neurula stage) and expression increased throughout neurulation (st. 12–22) and into tail bud stages (st. 25–36) as the nervous system developed and matured (Fig. 3A,B). The weak SVOP expression at stages 12 through 16 correlates with the onset in expression of several proneural bHLH factors during early neurogenesis (Ferreiro et al., 1993; Lee et al., 1995; Ma et al., 1996). Robust expression at later stages (stage 19–20 and older) may be due to increased numbers of differentiating neurons or to increased expression associated with neuronal maturation.
To better elucidate the spatial and temporal expression pattern of SVOP during development, we performed whole-mount in situ hybridizations using a digoxigenin (DIG) -labeled antisense RNA probe. Consistent with the RT-PCR analysis, SVOP was first detectable by whole-mount in situ hybridization at stage 19 and was expressed within the developing nervous system, specifically in the neural tube and the trigeminal ganglia (Fig. 4A). SVOP expression was maintained in these regions of the developing nervous system through late neurula and early tail bud stages (stage 32; Fig. 4B–F).
Sections through labeled embryos at stage 20 revealed two dorsolateral domains of SVOP expression in the neural tube (Fig. 5A), consistent with expression in early differentiating Rohon–Beard neurons, as labeled by the markers Xaml (Fig. 5B; Tracey et al., 1998; Perron et al., 1999) and XHox11L2 (Fig. 5C; Patterson and Krieg, 1999). By stage 22, SVOP expression was detectable in two additional ventral domains (Fig. 5D), likely differentiating motoneurons, and this expression was stronger by stage 25 (Fig. 5E). This pattern of SVOP expression is comparable to the expression of NeuroD in the neural tube at stage 22 (Fig. 5F), consistent with regulation of SVOP expression by proneural bHLH factors. As development progressed, SVOP was expressed in ventrolateral regions of the brain and spinal cord at stage 30 (Fig. 5G–I), but by stage 42 was expressed throughout the dorsal–ventral extent of the neural tube at both cranial (Fig. 5J) and caudal (Fig. 5K) levels. Within the developing neural retina, SVOP expression was not detectable at stage 30 (Fig. 5I), even though there are early-born postmitotic neurons present at this stage. However, by stage 42, SVOP was expressed in the inner layers of the neural retina, which contain early-born neurons such as retinal ganglion cells and amacrine cells, suggesting that SVOP is expressed as retinal neurons mature. Therefore, SVOP expression appears to follow the pattern of neuronal maturation in the developing Xenopus nervous system.
In this study, we present the first developmental expression of SVOP and show that in Xenopus SVOP expression is restricted to the developing nervous system and is first detectable in early differentiating Rohon–Beard neurons. As development progresses SVOP becomes broadly expressed throughout the central nervous system, consistent with the fact that in the adult rat SVOP is expressed throughout the entire brain (Janz et al., 1998).
The function of SVOP is still undetermined, but based on sequence similarity, it is most closely related to the SV2 family of synaptic vesicle proteins. Although SVOP and the SV2 proteins share several traits, there are several notable differences between them. SVOP is evolutionarily conserved, whereas the SV2 genes are found only in vertebrates. Also, although both SVOP and SV2 colocalize with synaptic vesicles in the adult brain, SVOP is also detected within the cell bodies of certain mature neuronal cell types (Janz et al., 1998), suggesting that it may not function exclusively as a regulator of neurotransmission. It is interesting to note that our analysis reveals that Xenopus SVOP is first detected earlier than the neuronal-specific synaptobrevin (sybII; Knecht et al., 1995), a regulator of synaptic exocytosis, which is first detected at significant levels by in situ hybridization at stage 24. Future experiments will be required to determine whether SVOP contributes to neuronal differentiation or maturation during neural development.
Isolation of Xenopus laevis SVOP
A 245-bp fragment corresponding to Xenopus laevis SVOP was obtained from a Differential Display screen for transcriptional targets for the proneural bHLH factors Xath5 and XNeuroD (Logan et al., in press). This fragment was labeled by Klenow using 32P-dCTP and used to screen a stage 28–30 Xenopus head cDNA library in lambda ZAP II (gift from R. Harland). A cDNA clone was recovered from positive phage using a Rapid Excision Kit (Stratagene). The clone was fully sequenced by the University of Utah Sequencing Core Facility, and sequence analysis was performed using Sequencher (Gene Code Corp). Full-length sequence for SVOP was deposited into the GenBank database with accession no. DQ167576.
RT-PCR analysis of SVOP
Total RNA was isolated from various stage embryos using Trizol (Invitrogen) followed by DNase (Ambion) treatment. First-strand cDNA was synthesized by reverse transcribing with oligo-dT primer (Invitrogen) and Superscript II RNase H-reverse transcriptase (Invitrogen). PCR was performed using the following primers: SVOP-F, 5′-CGTGTTGTTTACTAACTGATCTACAACC-3′; SVOP-R, 5′-TTATACATGGGGGCAAACCAG-3′; EF1α-F, 5′-CAGATTGGTGCTGGATATGC-3′; EF1α-R 5′-ACTGCCTTGATGACTCCTAG-3′. Results were confirmed by performing real-time PCR reactions using SYBR Green PCR Master Mix (Applied Biosystems). All qRT-PCR was performed in triplicate in an Applied Biosystems Prism 7700 Sequence Detection System. Cycling conditions were the same for all primer sets: 95°C for 10 min and cycles of 95°C for 15 sec, 55°C for 20 sec, and 72°C for 40 sec. SDS2.1 software (Applied Biosystems) was used to visualize the data. The Comparative Critical Threshold Method described in Applied Biosystems User Bulletin #2 was used to determine relative changes in gene expression levels using Histone 4 as a reference, which did not change significantly in our samples. Primers were HistoneH4-F 5′-CTCAAGGTTTTCCTGGAGAATGTC-3′ and HistoneH4-R 5′-TTTATCCGCCGAAGCCGTAGAG-3′.
In Situ Hybridizations
Full-length antisense RNA probe against Xenopus laevis SVOP was generated by in vitro transcription using an AmpliScribe kit (Epicentre) and incorporating digoxigenin-11-UTP (Boehringer Mannheim). Embryos were collected at the appropriate developmental stages (Nieuwkoop and Faber, 1994) and fixed in MEMFA, and whole-mount in situ hybridization was performed as previously described (Hutcheson and Vetter, 2001). For section analysis of SVOP expression, embryos were paraffin embedded and serial sectioned at 12.5 microns after whole-mount in situ hybridization.
We thank Dr. Richard Harland for the Xenopus laevis head cDNA library. We thank Erin Callahan and Kaitlyn Le for technical assistance with in situ hybridizations and paraffin sectioning. We also thank Dr. Sean Speese for comments on the manuscript. M.A.L. received a NIH Developmental Biology Training Grant and a NIH Genetics Training Grant, and M.L.V. received a NIH grant.