Zinc-finger nuclease-mediated targeted insertion of reporter genes for quantitative imaging of gene expression in sea urchin embryos

Abstract

To understand complex biological systems, such as the development of multicellular organisms, it is important to characterize the gene expression dynamics. However, there is currently no universal technique for targeted insertion of reporter genes and quantitative imaging in multicellular model systems. Recently, genome editing using zinc-finger nucleases (ZFNs) has been reported in several models. ZFNs consist of a zinc-finger DNA-binding array with the nuclease domain of the restriction enzyme FokI and facilitate targeted transgene insertion. In this study, we successfully inserted a GFP reporter cassette into the HpEts1 gene locus of the sea urchin, Hemicentrotus pulcherrimus. We achieved this insertion by injecting eggs with a pair of ZFNs for HpEts1 with a targeting donor construct that contained ∼1-kb homology arms and a 2A-histone H2B–GFP cassette. We increased the efficiency of the ZFN-mediated targeted transgene insertion by in situ linearization of the targeting donor construct and cointroduction of an mRNA for a dominant-negative form of HpLig4, which encodes the H. pulcherrimus homolog of DNA ligase IV required for error-prone nonhomologous end joining. We measured the fluorescence intensity of GFP at the single-cell level in living embryos during development and found that there was variation in HpEts1 expression among the primary mesenchyme cells. These findings demonstrate the feasibility of ZFN-mediated targeted transgene insertion to enable quantification of the expression levels of endogenous genes during development in living sea urchin embryos.

The phenotype and behavior of cells are largely determined by the expression levels of thousands of genes (1, 2). In some cases, the expression dynamics of specific genes also affect cell behavior (3). Therefore, to understand the molecular mechanisms of cellular events, it is necessary to quantify the expression of genes and their dynamics at the single-cell level. Techniques for insertion of a reporter gene into a genomic locus of interest and quantitative imaging of the reporter activities have been developed for this purpose (1, 4). By using these techniques in bacteria, yeast, and mammalian cells, it has been reported that a population of genetically identical cells can exhibit extensive cell-to-cell variability in the expression levels of many genes (3–6). However, for multicellular organisms, a universal technique for gene targeting and quantitative imaging has not yet been established in living animals.

Recently, a method for targeted gene editing using engineered zinc-finger nucleases (ZFNs) has been used in Drosophila (7), sea urchins (8), zebrafish (9, 10), plants (11), and human cultured cells (12, 13). ZFNs consist of a customized array of zinc-finger domains that bind to a specific DNA sequence and the nuclease domain of the restriction enzyme FokI. When two ZFNs bind to their associated target sequences in the appropriate direction, the nuclease domains dimerize, and a double-stranded break (DSB) is introduced. The ZFN-induced DSB can then be repaired with high efficiency by either homology-directed repair (HDR) or error-prone nonhomologous end joining (NHEJ) repair independently of a DNA template. Therefore, ZFNs can introduce a site-specific insertion or deletion at the DSB site after NHEJ repair (9, 11). Alternatively, ZFNs can produce defined genetic modifications, including the insertion of a reporter gene, near the site of the DSB by HDR using an exogenous targeting donor construct (7, 11). In animals, it has been noted that ZFN-induced DSBs are mainly repaired by NHEJ, and therefore ZFN-mediated targeted gene correction and transgene insertion are considered to be challenging (7, 14).

In the present study, we performed targeted insertion of a GFP reporter cassette into the endogenous HpEts1 locus of the sea urchin, Hemicentrotus pulcherrimus, by injecting a pair of ZFNs with a targeting donor construct. The sea urchin embryo, which is transparent, simple, and readily accessible to experimental perturbations, offers a unique opportunity to study the regulation of morphogenesis during early development. In addition, in vivo quantitative imaging methodology at the cellular level has been established by using confocal laser scanning microscopy (CLSM) (15). Using CLSM, we measured the fluorescence intensity of GFP at single-cell resolution in reporter knock-in sea urchin embryos and found that there was variation in HpEts1 expression among primary mesenchyme cells (PMCs). These findings suggest that ZFN-mediated targeted transgene insertion can be used to quantify the expression levels of endogenous genes during development in sea urchin embryos.

Results

Targeted Transgene Insertion into the HpEts1 Locus Using ZFNs.

To explore the possibility of inserting a reporter cassette into a genomic site of interest in the sea urchin, H. pulcherrimus, we selected the HpEts1L and HpEts1R ZFNs, whose target sites (5′-GGGGTTGACG-3′ and 5′-GATGATGACT-3′, respectively) are located upstream of the stop codon of the HpEts1 gene responsible for PMC differentiation (16), by bacterial one-hybrid (B1H) and single-strand annealing (SSA) screenings (8) (Fig. 1A and Fig. S1). Zygotic expression of the HpEts1 transcription factor, which is encoded by the HpEts1 gene, is detected in the nuclei of presumptive PMCs at the hatched blastula stage and in PMCs at the mesenchyme blastula stage (17, 18). To examine the activity of the HpEts1 ZFNs in sea urchin embryos, we amplified the DNA fragments around the target sites for the HpEts1 ZFNs by PCR using genomic DNA extracted from HpEts1 ZFN mRNA-injected embryos and control embryos at 2.5 and 3.5 h postfertilization (hpf) and digested with the restriction enzyme AciI (Fig. 1B). An AciI-resistant fragment showing the introduction of mutagenesis was observed among the DNA fragments from the HpEts1 ZFN mRNA-injected embryos at 2.5 hpf (two- to four-cell stage), and the amount of this fragment was slightly increased at 3.5 hpf (four- to eight-cell stage), indicating that mutagenic NHEJ events occurred after the injection of the HpEts1 ZFN mRNAs. These findings suggest that the HpEts1 ZFNs introduced a DSB at their target site as early as 2.5 hpf.

Fig. 1.

The HpEts1 ZFNs introduce a DSB at a target site in sea urchin embryos. (A) The HpEts1 gene showing the ZFN target site. A schematic representation of the H. pulcherrimus homolog of the Ets1 gene (HpEts1) is shown. Exons are indicated by boxes. The gray and black boxes represent untranslated and coding regions, respectively. The bent arrow depicts the transcription start site. The ZFN-targeted sequence and the interaction site of the pair of ZFNs used in this study are shown. The red bar indicates the stop codon of the HpEts1 gene. (B) Analysis of the mutations induced by ZFNs. A schematic representation of the HpEts1 genomic region used for the PCR-based analysis is shown in Upper. The primer sites are indicated by arrows. The amplified region contains a target site for the HpEts1 ZFNs and two AciI sites. One of the AciI sites is within the HpEts1 ZFN target site. Lower shows a representative analysis of the PCR products. The PCR products amplified from genomic DNA extracted at 2.5 and 3.5 hpf from sea urchin embryos injected with the HpEts1 ZFN mRNAs (ZFNs) or noninjected control embryos (Noninj.) were purified, digested with AciI, and analyzed by agarose gel electrophoresis.

Next, to examine the availability of ZFN-mediated targeted transgene insertion in sea urchin embryos, we prepared two targeting donor constructs, designated Ets-HRD (Fig. 2A) and Ets-HRD+T (Fig. 2B). The first targeting donor construct, Ets-HRD, contained ∼1-kb homology arms and a 2A-H2B–GFP cassette (2A is a self-cleaving peptide sequence) (19). Therefore, insertion of the reporter cassette into the HpEts1 locus was expected to result in the expression of two polypeptides: full-length HpEts1 fused with the 17-amino acid sequence of the 2A peptide and H2B–GFP, which localizes to the nucleus where its fluorescence can be accurately quantified (20) (Fig. 2A). We confirmed that the 2A peptide mediated protein cleavage in sea urchin embryos by injecting mRNAs for 2A-linked constructs and performing Western blot analyses (Fig. S2). The other targeting donor vector, Ets-HRD+T, contained HpEts1 ZFN target sites at both ends of an Ets-HRD donor cassette and thus generated a linearized targeting donor in the embryos (Fig. 2B). In Drosophila, it was reported that the efficiency of ZFN-mediated targeted gene modification was increased by using an extrachromosomal linear donor, which is linearized in situ, compared with a circular donor (21). In addition, we planned to repress NHEJ repair in the sea urchin embryos. For this purpose, we cloned a cDNA for the carboxyl-terminal tandem BRCT repeat of DNA ligase IV and prepared its mRNA (DN-lig4) with the expectation that overexpression of this mRNA would induce a dominant-negative effect, as reported in human cultured cells (22).

Fig. 2.

ZFN-mediated targeted gene insertion. (A) Targeting donor construct (Ets-HRD) for insertion of the 2A-H2B–GFP cassette into the HpEts1 locus. The structure of the HpEts1 locus and the targeted HpEts1 allele are shown. The gray and black boxes represent coding and noncoding exons, respectively. Schematic representations of the proteins derived from the targeted HpEts1 allele are also shown. These proteins are separated into the full-length HpEts1 protein and H2B–GFP during translation by the 2A self-cleaving peptide. The primer sites for the genomic PCR analysis are indicated. (B) Structure of the targeting donor construct Ets-HRD+T, which contains HpEts1 ZFN target sites at both ends of the Ets-HRD cassette. (C) Representative results of PCR-based genotyping analysis of the HpEts1 locus. PCR was performed on genomic DNA extracted from embryos, which had been injected immediately after fertilization at 24 hpf. The primers used were either primers 1 and 2 (as shown in A) or primers to amplify the control gene HpArs. The PCR products were separated by gel electrophoresis. H and T represent injection of the Ets-HRD and Ets-HRD+T targeting donor constructs, respectively. (D–F) GFP-expressing embryo at 30 hpf injected with HpEts1 ZFN mRNAs, DN-lig4 mRNA, and Ets-HRD+T. A representative embryo expressing GFP in the PMCs is viewed from the vegetal pole. (G–I) Noninjected control embryo at 30 hpf. (D and G) Bright field images. (E and H) Fluorescent images. (F and I) Merged images of D and E and of G and H, respectively. The arrowheads indicate GFP fluorescence in the nuclei of PMCs. Background autofluorescence is denoted by asterisks. (Scale bars, 20 μm.)

To validate the utility of ZFN-mediated targeted transgene insertion in the sea urchin, we injected several combinations of the targeting donor constructs, HpEts1 ZFN mRNAs, and DN-lig4 mRNA, and we then performed PCR analyses using genomic DNA extracted from the embryos at 24 hpf (Fig. 2C). As expected, no PCR products were detected in the noninjected and donor-injected samples. In contrast, PCR products of the expected size were observed in the donor/ZFN-coinjected samples and were significantly increased in the Ets-HRD+T/ZFN-coinjected samples compared with the Ets-HRD/ZFN-coinjected samples. Furthermore, the amount of the PCR product was slightly increased in the embryos coinjected with HpEts1 ZFN mRNAs, Ets-HRD+T, and DN-lig4 mRNA compared with those that did not receive DN-lig4 mRNA. Sequencing of the PCR products confirmed the occurrence of the targeted insertion using the donor constructs (Fig. S3). Next, we examined H2B–GFP expression in the injected embryos using epifluorescence microscopy at 30 hpf (gastrula stage) and counted the numbers of H2B–GFP-expressing embryos (Fig. 2 D–I and Table 1). In some embryos, GFP fluorescence was observed in the nuclei of PMCs, in which HpEts1 is expressed (16, 17), and was never detected in other cell types. Consistent with the genomic PCR analysis, GFP-expressing embryos were more frequently observed after injection of Ets-HRD+T, HpEts1 ZFNs, and DN-lig4 mRNAs than after injections with other combinations. These findings suggest that targeted transgene insertion using ZFNs is feasible in sea urchin embryos and that the combined use of donor constructs containing the target sites and DN-lig4 increases the insertion efficiencies.

Table 1.

Frequencies of GFP-expressing embryos

Targeting donor construct	HpEts1 ZFN mRNAs, each, pg	DN-lig4 mRNA, pg	Mean GFP-expressing embryos, %*	Mean abnormal embryos, %*
Ets-HRD (40 fg)	—	—	0 (±0)	6.10 (±0.98)
Ets-HRD (40 fg)	1	—	1.86 (±0.60)	10.01 (±1.16)
Ets-HRD+T (40 fg)	—	—	0 (±0)	6.97 (±0.26)
Ets-HRD+T (40 fg)	1	—	12.01 (±0.77)	7.71 (±1.82)
Ets-HRD+T (40 fg)	1	5	15.92 (±1.55)	11.67 (±1.84)

There were three experiments for each targeting donor construct, and in each experiment, 200 embryos were injected.

*Values are mean percentages (averages of all trials) of GFP-expressing and abnormal embryos, respectively, with the SEM in parentheses.

Quantitative Imaging of Endogenous Gene Expression in Living Sea Urchin Embryos.

In sea urchin embryos, in vivo quantification of GFP reporter gene expression at the single-cell level has been established by using CLSM (15). We predicted that application of this technique to embryos in which the GFP gene was knocked into an endogenous genomic locus would enable real-time quantification of endogenous gene expression. To explore this hypothesis, we imaged embryos injected with Ets-HRD+T, HpEts1 ZFNs, and DN-lig4 mRNAs at 16, 18, 20, 22, and 24 hpf using CLSM (Fig. 3 A–E). GFP fluorescence was observed in the nuclei of migrating PMCs at 16 and 18 hpf and in those of PMCs from 18 to 24 hpf. The numbers of GFP-expressing PMCs in the individual embryos ranged from 3 to 43 at 24 hpf (Table S1). Considering that an H. pulcherrimus embryo contains 55 ± 10 PMCs at this stage (23), the result suggests that ZFN-mediated targeted transgene insertion occurred at the developmental stage when there were 2–16 PMC-generating cells (Table S1).

Fig. 3.

Quantification of the mean fluorescence intensities in the nuclei of GFP-expressing cells during development. (A–E) Projections of z-stack images of GFP-expressing embryos. Representative GFP-expressing embryos injected with HpEts1 ZFN mRNAs, DN-lig4 mRNA, and Ets-HRD+T were imaged at 16 (A), 18 (B), 20 (C), 22 (D), and 24 (E) hpf by using CLSM. The double-headed arrows indicate the animal pole (represented as a red A) and vegetal pole (represented as a green V). The white arrowheads indicate GFP fluorescence in the nuclei of PMCs. The white dotted lines indicate the outside cell surfaces of the embryos. (Scale bars, 10 μm.) (F–J) Distributions of the fluorescence intensity of GFP-expressing cells during development. The scatter plots on the left show the distributions of the GFP fluorescence intensities in individual embryos at 16 (F), 18 (G), 20 (H), 22 (I), and 24 (J) hpf. The orange lines represent the mean fluorescence intensities of GFP-expressing cells at each time point. The histograms on the right show the distributions of the fluorescence intensities of GFP-expressing cells at each time point.

As shown in Fig. 3, the mean fluorescence intensity in the cells increased during development. In addition, variation in the fluorescence intensity among cells increased after 18 h (Fig. 3 F–J). There are several possible reasons for this observation. First, it may have been caused by variation in the GFP copy number, because some GFP-expressing cells may only have had one GFP reporter gene monoallelically inserted into the HpEts1 locus, whereas other GFP-expressing cells may have had two GFP reporter genes biallelically inserted into the HpEts1 locus. To explore this possibility, we coinjected more than 200 fertilized eggs with HpEts1 ZFNs and DN-lig4 mRNA together with Ets-HRD+T and Ets-HRD+mT containing an mCherry cDNA instead of the GFP cDNA (Fig. S4A). We obtained fluorescence images at 24 hpf by CLSM (Fig. S4B). In most fluorescent protein-expressing embryos, only one type of fluorescence was observed; 7% and 8% of injected embryos expressed GFP and mCherry, respectively. Although 2% of the embryos possessed both GFP- and mCherry-expressing PMCs, there were no embryos with GFP/mCherry double-positive PMCs. These findings suggest that the reporter construct is monoallelically inserted into the HpEts1 locus in most fluorescent protein-expressing PMCs. Another possibility is that the variation in fluorescence intensity may have originated from differences in the endogenous HpEts1 expression levels between PMCs. To explore this possibility, we examined the fluorescence intensity of GFP-expressing PMCs and the ingression of PMCs in embryos injected with HpEts1 ZFN mRNAs, DN-lig4 mRNA, Ets-HRD+T, and a moderate dose of an mRNA for a dominant-negative form of HpEts1 (ΔHpEts) (ref. 16; Fig. 4). We expected that the threshold level at which HpEts1 activates PMC differentiation would be raised by the expression of ΔHpEts, which lacks the activation domain and antagonizes native HpEts1 function by blocking its binding to the target site. If the GFP fluorescence intensity is related to the expression level of endogenous HpEts1, we expected that some presumptive PMCs expressing above-threshold levels of HpEts1 and H2B–GFP would differentiate into PMCs, whereas others with below-threshold levels of HpEts1 and H2B–GFP would not migrate into the blastocoel. Consistent with this hypothesis, significantly higher fluorescence intensity was observed in the nuclei of the PMC population than in the other GFP-positive cell population in the blastoderm (Fig. 4). These findings suggest that there is variation in zygotic HpEts1 expression among PMCs.

Fig. 4.

Variation in fluorescence intensity among GFP-expressing cells. HpEts1 ZFNs, DN-lig4 mRNA, Ets-HRD+T, and ΔHpEts mRNA were injected into fertilized eggs, and the fluorescence intensities of GFP-positive cells were quantified at 24 hpf. (A) Distribution of the GFP fluorescence intensities in migratory, nonmigratory, or total GFP-positive (GP) cells in embryos injected with a low dose of ΔHpEts mRNA. (B–E) Representative projections of z-stack images of a GFP-expressing embryo coinjected with ΔHpEts mRNA. B, C, and D show GFP, Texas Red, and merged images, respectively. (E) Pseudocolored and enlarged image of the area marked by the white square in B. Inset represents the pixel intensity profile. The filled and open arrowheads indicate GFP-positive cells ingressed and not ingressed into the blastocoel, respectively. (Scale bars, 20 μm.)

Discussion

In this study, we have demonstrated that targeted transgene insertion into the HpEts1 locus can be efficiently achieved by cointroduction of ZFNs and a targeting donor construct in sea urchin embryos. Moreover, by combining this technique with fluorescence quantification, we were able to measure the expression levels of endogenous genes in living sea urchin embryos.

Targeted Transgene Insertion in Sea Urchin Embryos Using ZFNs.

Insertion of the reporter gene into the target site was detected in embryos coinjected with the targeting donor construct and ZFN mRNAs, but not in embryos injected with the targeting donor alone. These findings indicate that spontaneous homologous recombination is a rare event in the sea urchin embryo and that a ZFN-induced DSB at the target site stimulates HDR, resulting in the induction of targeted transgene insertion at a detectable level. This result is in agreement with earlier work using Drosophila and mice (7, 14), suggesting that ZFN-mediated targeted insertion is feasible in model animals, although the efficiency of the insertion may depend on the species. Addition of ZFN target sites at both ends of the targeting donor cassette significantly increased the efficiency of ZFN-mediated targeted transgene insertion. In Drosophila, it was reported that an inserted donor was quite inefficient, an excised circular donor was better, and an extrachromosomal linear donor was best for HDR-mediated targeted gene modification (21). Therefore, it is considered that linearization of the targeting donor by ZFNs facilitates DNA strand invasion, which is required for HDR, or stimulates SSA, which involves extensive end processing to reveal complementary single strands in each repeat.

Cointroduction of DN-lig4 mRNA increased the efficiency of ZFN-mediated targeted transgene insertion. It has been reported that DNA ligase IV, a major component of the NHEJ pathway, forms a complex with XRCC4 and seals DNA ends (22). Moreover, DSB repair can be biased toward HDR by disrupting the function of DNA ligase IV, resulting in an increase in the efficiency of targeted gene modification (7). Therefore, we hypothesized that in the sea urchin, inhibition of DNA ligase IV through the introduction of DN-lig4 mRNA increases the propensity to repair ZFN-induced DSBs through HDR, resulting in enhanced efficiency of ZFN-mediated targeted transgene insertion. However, in the Lig4 mutant of Drosophila, DSB repair was almost completely biased toward HDR, whereas in the sea urchin, the DN-lig4–mediated enhancement of the efficiency of ZFN-mediated targeted transgene insertion was modest (7). This result may arise because when DSBs are introduced by ZFNs, the DN-lig4 protein is not translated at a sufficiently high rate to completely inhibit the function of the endogenous DNA ligase IV. Therefore, the introduction of recombinant DN-lig4 and ZFN proteins into fertilized sea urchin eggs may enhance the efficiency of ZFN-mediated targeted transgene insertion.

Unintentional off-target cleavages are a potential problem with genome editing using ZFNs (24–26). Potential off-target sites are typically defined by scanning the genome for sites similar to the ZFN recognition sequences. However, because we could not directly search for putative off-target sites, owing to the lack of availability of the H. pulcherrimus genomic sequence, we searched for putative off-target sites that contained zero or one mismatches relative to each HpEts1 ZFN target site in the genome of Strongylocentrotus purpuratus, which is closely related to H. pulcherrimus. We found only one sequence containing one mismatch on Sp-Ets1/2, the ortholog of HpEts1. This finding implies that the HpEts1 ZFN target site on the HpEts1 gene is a unique on-target site in the H. pulcherrimus genome and that insertion of the reporter gene into other genomic sites is a rare event when HpEts1 ZFNs are used in H. pulcherrimus embryos.

Quantitative Imaging of Endogenous Gene Expression in Living Sea Urchin Embryos.

We successfully demonstrated visualization of endogenous gene expression in living sea urchin embryos by ZFN-mediated insertion of a 2A-H2B–GFP cassette into the HpEts1 locus. In this study, we used the 2A peptide, which is useful for balanced coexpression of multiple proteins from a single promoter, to avoid the synthesis of a fusion protein (19). Although a GFP fusion protein provides subcellular localization information as well as indicates the expression level of the gene of interest (1, 3), GFP may inhibit the function of its fusion partner. In our study, most of the reporter knock-in embryos exhibited normal development, and fluorescent signals were clearly detected in the nuclei of PMCs, suggesting that ZFN-mediated insertion of the 2A-H2B–GFP cassette is a useful technique for quantitative imaging of gene expression at the single-cell level in developing embryos.

In the present study, we found variation in the expression levels of zygotic HpEts1 among PMCs. However, variation in the amounts of HpEts1 mRNA and HpEts1 protein has not been detected by standard whole-mount in situ hybridization and immunostaining (16, 18). This variation may arise because these relatively low-sensitivity techniques cannot distinguish subtle variation in the expression levels. Another possibility is that substantial amounts of the HpEts1 gene product, which was reported to be maternally expressed at abundant levels in whole sea urchin embryos (16), mask the variation in the amounts of the zygotic HpEts1 gene product. In the latter case, it is difficult to distinguish between maternal and zygotic expression by conventional methods. In fact, although HpEts1 mRNA is maternally expressed at abundant levels in the whole embryo during early development (16) and the mRNA and HpEts1 protein can be detected in presumptive PMCs and PMCs from the hatching blastula stage, it remains unclear whether this expression pattern is because of zygotic expression of HpEts1, selective degradation of maternal gene products except in presumptive PMCs and PMCs, or both. Therefore, ZFN-mediated reporter insertion is a useful technique for detecting the zygotic expression of an endogenous gene of interest. Our findings also show that PMC differentiation depends on the expression levels of zygotic HpEts1, suggesting that if maternal HpEts1 mRNA and maternal HpEts1 protein remain in PMCs, they are functionally negligible. Further investigations are needed to understand the biological relevance of the variation in the zygotic HpEts1 expression levels among PMCs.

In most model animals, including the sea urchin, conventional transgenic methods, such as introduction of a reporter construct containing a cis-regulatory element of genes into embryos, are carried out to analyze the regulatory mechanism for the spatiotemporal expression of genes (27). However, in these methods, the reporter gene is randomly integrated into the genome, and its expression in embryos may be affected by positional effects, depending on the integration site (28). It was reported that the expression levels of some genes are regulated by distal regulatory elements (29). Therefore, it is uncertain whether sufficient regulatory elements required for endogenous expression are contained in the reporter constructs. In addition, the exogenous DNAs introduced into the sea urchin embryo are joined into concatemers, and multiple copies of them become incorporated into the chromosome (30), meaning that the reporter expression level is not always consistent with the endogenous expression level. In contrast, the combination of the ZFN-mediated targeted gene insertion and in vivo quantitative imaging techniques avoids the above-mentioned problems and enables quantification of the endogenous gene expression level in each cell in real time. The combination of ZFN-mediated targeted gene insertion and live imaging can potentially be applied to examine the relationship between the gene expression level in a cell and its fate. For example, it may be useful for elucidating the transfating mechanisms of presumptive blastocoelar cells to PMCs in embryos whose PMCs were depleted at the mesenchyme stage (31). The combined approach described here using sea urchin embryos might be extended to other multicellular model systems, such as nematodes and zebrafish, in which not only fluorescence imaging techniques but also ZFN-mediated genome modification techniques are available (9, 10, 32).

Materials and Methods

ZFNs targeting HpEts1 were selected by B1H and SSA screenings as described (8); further information is included in SI Materials and Methods. Full descriptions of the constructions of the plasmids used in this study and the sea urchin culture conditions, as well as descriptions of the mRNA synthesis and microinjection, PCR-based genotyping assay, and imaging analysis, are detailed in SI Materials and Methods and Fig. S5. The sequences of oligonucleotides used in this study are listed in Tables S2 and S3.