Developmental Biology 353 (2011) 147–159

Contents lists available at ScienceDirect

Developmental Biology

journal homepage: www.elsevier.com/developmentalbiology

Evolution of Developmental Control Mechanisms Asymmetric localization of germline markers Vasa and Nanos during early development in the amphioxus Branchiostoma floridae

Hui-Ru Wu a,1, Yen-Ta Chen a,1, Yi-Hsien Su a, Yi-Jyun Luo a, Linda Z. Holland b, Jr-Kai Yu a,c,⁎ a Institute of Cellular and Organismic Biology, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei, 11529, Taiwan b Marine Biology Research Division, Scripps Institution of Oceanography, UCSD, La Jolla, CA 92093-0202, USA c Institute of Oceanography, National Taiwan University, No. 1, Section 4, Roosevelt Road, Taipei, 10617, Taiwan article info abstract

Article history: The origin of germline cells was a crucial step in animal evolution. Therefore, in both developmental biology Received for publication 20 April 2010 and evolutionary biology, the mechanisms of germline specification have been extensively studied over the Revised 15 February 2011 past two centuries. However, in many animals, the process of germline specification remains unclear. Here, Accepted 15 February 2011 we show that in the cephalochordate amphioxus Branchiostoma floridae, the germ cell-specific molecular Available online 24 February 2011 markers Vasa and Nanos become localized to the vegetal pole cytoplasm during oogenesis and are inherited asymmetrically by a single blastomere during cleavage. After gastrulation, this founder cell gives rise to a Keywords: Germline cell cluster of progeny that display typical characters of primordial germ cells (PGCs). Blastomeres separated at Vasa the two-cell stage grow into twin embryos, but one of the twins fails to develop this Vasa-positive cell Nanos population, suggesting that the vegetal pole cytoplasm is required for the formation of putative PGCs in Cephalochordate amphioxus embryos. Contrary to the hypothesis that cephalochordates may form their PGCs by epigenesis, Evolution our data strongly support a preformation mode of germ cell specification in amphioxus. In addition to the early localization of their maternal transcripts in the putative PGCs, amphioxus Vasa and Nanos are also expressed zygotically in the tail bud, which is the posterior growth zone of amphioxus. Thus, in addition to PGC specification, amphioxus Vasa and Nanos may also function in highly proliferating somatic stem cells. © 2011 Elsevier Inc. All rights reserved.

Introduction 2006; Swartz et al., 2008; Voronina et al., 2008; Yoon et al., 1997); however, many important questions remain unanswered: for exam- The segregation of germline cells from the somatic cell lineages ple, how did mechanisms for PGC specification evolve during during animal development is considered an effective way to ensure metazoan radiation; and can comparative studies reveal the ancestral/ early protection of the genome from somatic mutations during the life derived modes of PGC specification in specific animal groups? history of multicellular animals (Extavour, 2007). There are two major In studying the evolution of germ cell specification mechanisms, we modes for the specification of PGCs (Extavour and Akam, 2003). In focused on cephalochordates (commonly known as amphioxus or some animals, PGCs are specified by the localization of maternal lancelets) because of their basal phylogenetic position within the cytoplasmic determinants into specific blastomeres during early chordates (Bourlat et al., 2006; Putnam et al., 2008). The mode of germ development (preformation); however, in other species, PGCs are cell specification in amphioxus is controversial. The earliest recogniz- determined later in development by inductive signals from neigh- able PGCs are found in clusters at the ventral extremities of the somites boring tissues (epigenesis). Comparative studies based on both in late larvae (Nieuwkoop and Sutasurya, 1979; Ruppert, 1997). Due to morphological criteria and conserved germ cell-specific markers, their segmental distribution and rather late origin during development, such as the Vasa and Nanos gene products (Ewen-Campen et al., it has been argued that PGCs are formed epigenetically in amphioxus 2010), have provided new insights into the evolution of germ cell (Extavour, 2007; Nieuwkoop and Sutasurya, 1979). specification mechanisms (Brown and Swalla, 2007; Brown et al., However, an electron microscopic study identified a region of 2009; Chang et al., 2002; Dill and Seaver, 2008; Extavour and Akam, subcortical cytoplasm near the vegetal pole of amphioxus eggs that 2003; Extavour et al., 2005; Extavour, 2007; Fabioux et al., 2004; contains whorls of endoplasmic reticulum (ER) with associated Ikenishi and Tanaka, 2000; Ozhan-Kizil et al., 2009; Pfister et al., 2008; electron-dense particles and mitochondria (Holland and Holland, Rebscher et al., 2007; Rosner et al., 2009; Shirae-Kurabayashi et al., 1992). Based on morphological comparisons, this vegetal pole plasm was suggested to be the amphioxus equivalent of the germ plasm (Holland and Holland, 1992); however, the fate of this pole plasm was ⁎ Corresponding author. Fax: +886 2 2785 8059. E-mail address: [email protected] (J.-K. Yu). not analyzed beyond the mid-blastula stage, and its precise molecular 1 These authors contributed equally to this work. components were not determined. In this study, we demonstrate that

0012-1606/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2011.02.014 148 H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 the gene products of the germ cell-specific markers Vasa and Nanos levelateachspecific stage. The sequences of Q-PCR primers were as co-localize with the vegetal pole plasm during amphioxus develop- follows: BfVasaQF, 5′-CCAATACCATGCCCAAGACT-3′ and BfVasaQR, 5′- ment, and the cells containing these markers display typical ACAAACCAAGTGCCTTCACC-3′ for Bf-Vasa;BfNanosQF,5′-CCCTGTCCTT- characters of PGCs. In addition, blastomere separation experiments ATGGCCTACA-3′ and BfNanosQR, 5′-GTTGGGTAGCTGGTTGGTGT-3′ for show that this localized pole plasm is required for the formation of Bf-Nanos; Bf18SQF, 5′-CCTGCGGCTTAATTTGACTC-3′ and Bf18SQR, 5′- putative PGCs in amphioxus embryos. Thus, we propose that these AACTAAGAACGGCCATGCAC-3′ for 18S rRNA. cells are the presumptive amphioxus PGCs. These results provide strong evidence for the preformation mode of germ cell determina- In situ hybridization tion in amphioxus. We used a primer matching the vector sequence adjacent to the 3′ Materials and methods end of the cDNA insert with a T7 promoter sequence added to its 5′ end as a reverse primer (pDONR222-T7-Reverse, 5′-TAATACGACTCACTA- Cloning of Bf-Vasa and Bf-Nanos TAGGGAGGGGATATCAGCTGGATG-3′) and a primer matching the vector sequence adjacent to the 5′ end of the cDNA insert with an SP6 A BLAST search using zebrafish Vasa (NP_571132) as the query promoter sequence added as a forward primer (pDONR222-SP6- against the assembled Branchiostoma floridae draft genome (Putnam Forward, 5′-ATTTAGGTGACACTATAGAAGACGGCCAGTCTTAAGCTC-3′) et al., 2008) led to the identification of two predicted gene models for PCR amplification of the full-length cDNA for riboprobe synthesis. (estExt_GenewiseH_1.C_5230018 and e_gw.58.215.1) coding for a The following PCR protocol was used: 2 min 94 °C; 5 cycles of 30 s 94 °C, Vasa-like protein. Detailed analysis of the sequence and synteny structure 30 s 45 °C, 3 min 68 °C; 25 cycles of 30 s 94 °C, 30 s 58 °C, 3 min 68 °C; 1 around these two gene models confirmed that they represent two cycle of 7 min 68 °C. DIG-labeled Bf-Vasa antisense riboprobes and different haplotypes of the same Vasa locus in the amphioxus genome. fluorescein-labeled Bf-Nanos antisense riboprobes were synthesized We used the better annotated gene model (estExt_GenewiseH_ 1. using T7 RNA polymerase. Control sense riboprobes were synthesized C_5230018) to search the amphioxus cDNA/EST database (Yu et al., by SP6 RNA polymerase. The procedure for in situ hybridization in 2008b)(http://amphioxus.icob.sinica.edu.tw/) and identified one cDNA amphioxus embryos was performed as previously described (Yu et al., cluster (00124) representing Vasa cDNA. Sequencing the longest cDNA 2008a; Yu and Holland, 2009b). For color detection, alkaline phospha- clone (bfne054i20, 3165 bp in length) from this cluster indicated that it tase-conjugated anti-DIG or anti-fluorescein were used contained most of the coding sequence of Vasa, along with its 3′UTR; (Roche). Alkaline phosphatase reaction products were visualized with however, part of the 5′ coding sequence and the 5′UTR were missing. We, NBT-BCIP (dark-purple color for DIG-labeled probes) or Fast Red tablets therefore, performed 5′RACE to obtain the full-length cDNA for Vasa using (Sigma) (red color for fluorescein-labeled probes). For double fluores- an outer gene-specificprimer(5′-AATTCTCGGGTGTCGGGATT-3′)andthe cent in situ hybridization, we used anti-DIG-POD and anti-fluorescein- 5′RACE Outer Primer provided in the FirstChoice RLM-RACE Kit (Ambion) POD antibodies (Roche) to detect the anti-sense riboprobes and then with cDNA prepared from unfertilized eggs. Next, we performed nested used the TSA Plus Cyanine 3 & Fluorescein system (PerkinElmer) to PCR using the inner gene-specificprimer(5′-TCCTCTCCGCATCTGAAACA- amplify the fluorescent signal. In some cases, DAPI (Invitrogen, 1 μg/ml 3′)andthe5′RACE Inner Primer. We obtained a 946-bp cDNA fragment in PBST) was used for nuclear staining, and the plasma membrane stain that overlapped with the cDNA clone bfne054i20. Assembly of the CellMask Orange (Invitrogen, 2.5 μg/ml in PBST) was used to mark the isolated cDNA clone and the 5′RACE fragment revealed a cDNA sequence general outline of the cells. The samples were visualized and photo- of 3939 bp with a putative open reading frame encoding a protein of 785 graphed using either a Zeiss Axio Imager.A1 microscope with DIC optics amino acids. We designated the assembled cDNA Bf-Vasa (NCBI GenBank or a Leica TCS-SP5-AOBS confocal microscope. The TEM methods used accession number HM004550). A similar BLAST search using human were as described previously (Holland and Holland, 1992). NANOS1 (NP_955631) as the query led to the identification of one predicted amphioxus gene model (fgenesh2_pg.scaffold_142000014) for generation and immunoblot amphioxus Nanos. Subsequently, we identified the corresponding cDNA cluster (01416) in the amphioxus cDNA/EST database. The longest cDNA A polyclonal antibody was produced using the COOH-terminal clone (bfad001o17) from this cluster was 2684 bp, with an open reading region of the BfVasa protein (amino acid 271 to the stop sequence) for frame encoding a putative Nanos protein of 294 amino acids, and we rabbit immunization. The recombinant BfVasa protein was expressed designated this sequence Bf-Nanos (NCBI GenBank accession number using the pET system (Novagen). After the initial immunization, the HM004549). These orthology assignments were further confirmed by rabbit was boosted and bled in two-week intervals. Batches of sequence alignments and/or molecular phylogenetic analysis of the antisera were checked for antibody titer by western blots. For western translated amino acid sequences from the amphioxus Vasa and Nanos blot analysis, 5 μg of amphioxus proteins from each developmental cDNA sequences (Suppl. Figs. S1–S3). The translated amino acid stage was resolved by SDS-PAGE, and the proteins were then sequences were aligned using the Clustal X program (Jeanmougin et al., transferred onto a PVDF membrane. The membrane was probed 1998), and the alignment was confirmed manually. After removing gaps, with primary antibodies (anti-BfVasa serum 1:50,000, or an anti- the verified alignments were used to construct phylogenetic trees with alpha-tubulin antibody, Sigma, 1:50,000), detected with an HRP- the MEGA program, version 4.0, based on the neighbor-joining method conjugated goat anti-rabbit secondary antibody (Jackson, 1:50,000), (Tamura et al., 2007). Bootstrap support values were calculated by 1000 and developed with the SuperSignal West Dura Extended Duration pseudoreplications. Substrate (Pierce).

Quantitative PCR (Q-PCR) analysis Antibody staining

Messenger RNA from each embryonic stage was isolated using the For BfVasa protein staining, samples were fixed in 4% PFA-MOPS- RNeasy Micro kit (Qiagen) and then reverse transcribed using the iScript EGTA fixation buffer (Yu and Holland, 2009a) for 1 h at room cDNA synthesis kit (Bio-Rad) as previously described (Yu and Holland, temperature or overnight at 4 °C, and samples were then stored in 2009c). The resulting cDNA was used as the template for Q-PCR analysis. 70% ethanol (EtOH) at −20 °C until use. Fixed samples were The Q-PCR analysis was performed on a Roche LightCycler 480 machine rehydrated in a stepwise fashion (50% EtOH/PBS; 30% EtOH/PBS; using the LightCycler 480 SYBR Green I Master system (Roche). The and PBST washes three times). After rehydration, samples were expression levels of Vasa and Nanos were normalized to the 18S rRNA permeabilized in 5 μg proteinase K/ml PBST for 3–5 min. Proteinase K H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 149 digestion was stopped using two washes with 2 mg/ml glycine in removal of the fertilization envelopes, as described previously (Yu and PBST, and samples were re-fixed with 4% PFA in PBST for 1 h at room Holland, 2009b), and the intact embryos were selected. Individual temperature. After three washes in PBST, samples were washed with blastomeres were separated by micro-surgery under a dissecting

1% H2O2 in PBS for 10 min to quench the endogenous peroxidase microscope using a gelatin-coated glass needle drawn from a Pasteur activity. After blocking with blocking solution (PBS with 5% goat pipette. We cultured isolated blastomeres individually and monitored serum, 0.1% Triton X-100, and 0.1% BSA) for 1 h at room temperature, twin embryo pairs produced from the same two-cell stage embryos. samples were incubated with anti-BfVasa serum at a 1:20,000 dilution Each of the isolated blastomeres was cultured in an individual well in in blocking solution overnight at 4 °C. Samples were then washed six a 24-well cell culture plate (also coated with gelatin on the bottom of times in PBST before the secondary antibody (HRP-conjugated goat the wells) until the wild-type control embryos from the same batch of anti-rabbit antibody, Jackson, 1:1,000 in blocking solution) was fertilized eggs developed into early neurula stage embryos (7–8h added. After 2 h at room temperature, the samples were washed six post-insemination). Twin embryos that developed from the same times in PBST, and the HRP color reaction was performed using DAB two-cell stage embryo were fixed and saved in individual micro- tablets (Sigma). For double fluorescent staining of BfVasa RNA and centrifuge tubes for further in situ hybridization with a Bf-Vasa probe; BfVasa protein, in situ hybridization was performed as described wild-type control embryos were also collected for the same analysis. above using the anti-DIG-POD antibody and the fluorescein substrate Additionally, we allowed some of the twin embryos to develop until in the TSA Plus system to detect the mRNA. Next, the immunodetec- early larval stage (two days post-insemination) for further investiga- tion of BfVasa protein was performed as described above, with the tion with antibody staining using anti-BfVasa serum. exception of using the anti-BfVasa serum at a 1:10,000 dilution and using an Alexa Fluor 594 goat anti-rabbit antibody (Invitrogen, 1:400 Results and discussion in blocking solution) as secondary antibody. DAPI (Invitrogen, 1 μg/ml in PBST) was used for nuclear staining. Localization of Vasa and Nanos during amphioxus oogenesis

Blastomere separations Using in situ hybridization and Q-PCR analysis, we demonstrated the presence of both Vasa and Nanos mRNA in the gonads of adult Collection of amphioxus gametes and fertilization were performed female and male amphioxus (Fig. 1). Vasa mRNA was uniformly as described previously (Yu and Holland, 2009a); fertilized eggs were distributed throughout the cytoplasm of oocytes b100 μm in diameter cultured at room temperature until the first cleavage furrow was (Figs. 1A and B), with the intensity of the signal being inversely just completed (approximately 60 min post-insemination). Two-cell proportional to the size of the oocyte. The signal was barely detectable stage embryos were transferred to a gelatin-coated dish, followed by in the cytoplasm of oocytes N100 μm in diameter, which is most likely

Fig. 1. Localization of Vasa (A–C) and Nanos (D–F) mRNA in the developing gonads in adult amphioxus. B and E are higher magnification images of developing oocytes. Scale bars are 50 μm. White arrows indicate the expression domains of Vasa and Nanos in cross-sections of adult female (A, D) and male (C, F) gonads. In B and E, black arrowheads indicate transcripts in the cytoplasm of small oocytes, while black arrows indicate RNA particles in the vegetal portion of large oocytes. In A, C, D, and F, gonads are encircled by dashed lines. o, ovary; t, testis; GV, germinal vesicle (encircled by dashed lines in B and E). (G) Transcript levels of amphioxus Vasa and Nanos as percentages of those of 18S rRNA at different developmental stages. 150 H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159

Fig. 2. (A–D) Changes in the distribution of Vasa RNA after fertilization. Times after insemination indicated at upper right of images. Zygotes viewed from the vegetal pole. (E–L) Co- localization of Vasa and Nanos RNA in unfertilized (E–H) and fertilized (I–L) amphioxus eggs. E and I, lateral views with the animal side up and vegetal side down. White arrowheads indicate polar bodies (pb); white arrows indicate localization of Vasa and Nanos signals. F–H, and J–L are vegetal views showing aggregates of Vasa/Nanos mRNA granules after fertilization. (M–O) Ultrastructural analysis of amphioxus egg vegetal pole plasm by TEM. M. Sheets of electron-dense particles with endoplasmic reticulum (ER, red arrows) are localized near the vegetal cortex in the unfertilized egg. N is higher magnification, showing electron-dense particles (red arrowheads) with the associated ER and mitochondria (mt). O shows the assembly and composition of vegetal pole plasm (red arrows) after fertilization. cg, cortical granule; pvs, perivitelline space; vl, vitelline layer; yg, yolk granule. due to dilution by the accumulating yolk (Holland and Holland, 1991). expressed at lower levels than Nanos in both sexes, and both genes are This result is in contrast to a previous report demonstrating uniform expressed at lower levels in males than females (Fig. 1G). In addition, distribution of Vasa transcripts throughout the cytoplasm of all Vasa and Nanos transcripts were present at all of the developmental oocytes (Bridgham et al., 2008). Interestingly, in the largest oocytes stages examined (Fig. 1G). Therefore, we conclude not only that Vasa (~140 μm in diameter), we found a punctate pattern of Vasa mRNA and Nanos mRNA transcripts are reliable germline markers in near the vegetal cortex, which is at the opposite side from the amphioxus, but also that in females, their maternal transcripts germinal vesicle (Fig. 1B, black arrows), suggesting that the accumulate in the vegetal cortex of unovulated oocytes. segregation of maternal Vasa mRNA occurs in late oogenesis. In male amphioxus, a low level of Vasa RNA was present in a thin layer of Aggregation of Vasa and Nanos mRNA after fertilization cells at the periphery of the testis (Fig. 1C) (Bridgham et al., 2008). In adult gonads of both sexes, Nanos was expressed in the same patterns We then examined the distribution patterns of Vasa and Nanos as Vasa (Figs. 1D–F). Q-PCR analysis showed that Vasa mRNA is transcripts during amphioxus fertilization. Amphioxus oocytes undergo

Fig. 3. (A-N) Localization of Vasa and Nanos transcripts during cleavage. All images are side views with the animal side up and vegetal side down, except in panels C, D, F, J, and K, which are shown in vegetal views. Aggregated Vasa or Nanos transcripts are indicated by white arrows in each panel. A, animal side; V, vegetal side. Scale bars, 50 μm. (O–Q) Examples of aggregated Vasa RNA in either the left or the right blastomere in two-cell stage embryos. Panel O illustrates the orientation of the major axes of the dividing two-cell stage embryos. The embryos in P and Q are viewed from the vegetal side. A, animal side; V, vegetal side; L, left side; R, right side. The proportions of the two cases are indicated in the upper-right area of panels P and Q. White arrows, Vasa RNA; white arrowheads, DAPI staining of dividing nuclei. Scale bars, 50 μm. H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 151 the first meiotic division and arrest at the second meiotic metaphase hybridization assay demonstrated that Vasa and Nanos transcripts stage before being spawned and fertilized (Holland and Yu, 2004). In were exactly co-localized during this process near the vegetal pole these spawned oocytes, which are typically referred to as unfertilized (Figs. 2E–L). The timing of this RNA aggregation closely corresponds to eggs, Vasa transcripts were broadly distributed within the cytoplasm, the timing of the compaction of the vegetal pole plasm (Figs. 2MandN) with higher concentrations localized in small patches near the vegetal during early amphioxus development (Holland and Holland, 1992). pole (Fig. 2A). After insemination, these RNA-containing patches began After fertilization, the sub-cortical sheets of endoplasmic reticulum (ER) to aggregate. By approximately three minutes after insemination, a and associated particles and mitochondria form into a compact single compact aggregation of RNA was clearly visible in most fertilized structure with whorls of ER near the vegetal cortex (Fig. 2O). Based on eggs (Figs. 2B–D; Suppl. Fig. S4). The double fluorescent in situ these observations, we conclude that Vasa/Nanos RNA transcripts represent components of the vegetal pole plasm in amphioxus eggs.

Asymmetric inheritance of vegetal pole plasm during early embryogenesis

In fertilized eggs over 10 min post-insemination, a clear aggregate containing Vasa/Nanos mRNA was localized at the vegetal cortex (Figs. 3A and H), while some diffuse transcripts were also present within the cytoplasm (Figs. 3A and H; Suppl. Fig. S5). After the first cleavage, this aggregate was segregated into only one of the two blastomeres (Figs. 3B and I). Because the first cleavage divides the amphioxus egg equally into left and right blastomeres (Conklin, 1932), we determined which of the first two blastomeres inherits this aggregate. To orient the two-cell stage amphioxus embryo, we used DAPI nuclear staining of the second polar body to identify the animal- vegetal axis. DAPI staining of the dividing nuclei during second cleavage was also used to determine the second cleavage axis (Figs. 3O–Q). The aggregated Vasa RNA is always inherited by one of the two slightly larger posterior blastomeres in four-cell stage embryos (Figs. 3C, J, P and Q); therefore, the polarity of the 2nd cleavage (anterior vs. posterior) can be identified by the cell-size and the location of the aggregated Vasa RNA. Of the twenty-four embryos to which we could assign an orientation, sixteen had this Vasa RNA aggregate in their left blastomere (Fig. 3P), whereas eight had it in their right blastomere (Fig. 3Q), suggesting that the segregation of the pole plasm between the left or right blastomere during the first cleavage is variable. During subsequent cleavage stages, the pole plasm containing the Vasa/Nanos RNA was inherited by a single vegetal blastomere until the early gastrula stage (Figs. 3C–G, J–N). At the early gastrula stage, the cell containing the Vasa RNA aggregate began to invaginate with the vegetal plate, which becomes the future mesendoderm (Figs. 4A and D, and the corresponding drawing in Fig. 4H). When invagination was complete, resulting in a cap-shaped gastrula (approximately 5 h post-fertilization), we ob- served that the Vasa RNA dispersed in the cytoplasm but remained localized to the base of the cell (Figs. 4E and I). Then, this cell divided into two cells with roughly equal amounts of Vasa mRNA still localized to the basal cytoplasm (Figs. 4B, F, and J). Each of these two cells then divided asymmetrically, with only the two basal daughter cells inheriting Vasa RNA (Figs. 4C, G, and K). We also observed that by the mid-late gastrula stage, zygotic Vasa expression could be detected around the closing blastopore (Figs. 4B and C). In the early neurula stage, the ventral Vasa-containing cells were situated between the mesendoderm and ectoderm (Figs. 4L–N). Judging from their relative

Fig. 4. Expression of Vasa and Nanos transcripts during amphioxus development. Whole-mount embryos are side views with anterior towards the left. Scale bars, 50 μm. (A–C) Distribution of Vasa RNA during amphioxus gastrulation. Arrows indicate the aggregated maternal Vasa RNA. Arrowheads indicate zygotic Vasa expression around the blastopore. (D–G) Confocal microscopic images showing the invagination of Vasa- containing cells during gastrulation. Green, Vasa RNA; blue, DAPI staining of nuclei; red, CellMask staining. (H–K) Schematic drawings corresponding to the images in D–G. N, nucleus; ec, ectoderm; men, mesendoderm. (L–N) During the neurula stage, Vasa- containing cells (arrows) seem to migrate between the ectoderm and mesendoderm towards the posterior end of the embryo expressing zygotic Vasa (arrowheads). (O) In the 36 h larval stage, Vasa-positive cells are restricted to the tail bud. (P–V) Distribution of Nanos mRNA in the gastrula (P–R), neurula (S–U), and larva (V). The cells with maternal Nanos mRNA are indicated by black arrows. These cells are presumptive PGCs and seem to migrate posteriorly during neurulation to the vicinity of the most posterior ventral endoderm adjacent to a zygotic Nanos expression domain (black arrowheads). 152 H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 position to the blastopore at different time points during neurulation, we suggest that these ventral Vasa-positive cells seem to move as a cluster toward the posterior end of the embryo. By the late neurula stage, this cluster of Vasa-positive cells was located in the most posterior ventral endoderm adjacent to the zygotic Vasa expression domain near the blastopore (Fig. 4N). Later in the larval stage, these Vasa-positive cells remained in the most posterior ventral endoderm and became indistinguishable from the posterior tail bud cells, which were also expressing Vasa mRNA (Fig. 4O). The Nanos expression domain always coincided with that of Vasa at all developmental stages that we examined (Figs. 4P–V). A Nanos- positive cell is located in the middle of the invaginating vegetal plate in the early gastrula stage (Fig. 4P); subsequently, it divides into a cell cluster during the late gastrula and early neurula stages (Figs. 4Q–T). Similar to Vasa, a zygotic Nanos expression domain was also present around the closing blastopore during the gastrula stage (Fig. 4R), and the ventrally Nanos-positive cell cluster eventually merged into this domain near the most posterior end of the embryo (Figs. 4S–V). Based on the inheritance of the maternal germ cell-specific markers Vasa and Nanos in these ventral cells and their seemingly migratory behavior compared to the primordial germ cells in other organisms (Richardson and Lehmann, 2010), we propose that these cells are the presumptive PGCs in amphioxus.

Co-localization of Vasa RNA and protein

In many species, maternal Vasa protein is localized to the germ plasm and has essential functions for germ cell specification (Ewen- Campen et al., 2010). The observation that in amphioxus, maternal Vasa RNA is segregated in early embryogenesis into the presumptive PGCs raises the question of whether Vasa protein is also a germ plasm component in this species. We produced a rabbit anti-BfVasa polyclonal antibody to address this question. In western blot assays, the antiserum detected a single band consistent with the predicted mass of BfVasa protein (80 kDa) (Fig. 5A). The pre-immune rabbit serum did not react with amphioxus proteins of similar sizes. Western blotting also revealed that BfVasa protein was present at all developmental stages examined (Fig. 5B). Anti-BfVasa staining of sections of adult amphioxus ovaries showed that Vasa protein was most conspicuous in the cytoplasm of immature oocytes (Fig. 5D, black arrowhead). During oocyte growth, the signal became weaker, and the distribution was localized near the nuclear envelope in a punctate pattern (Fig. 5D, white arrows). This distribution pattern resembles the nuage-like structure in amphioxus and other animals (Carre et al., 2002; Holland and Holland, 1991; Knaut et al., 2000; Komiya et al., 1994; Tsunekawa et al., 2000). Antibody labeling of sections of adult testis revealed localized Vasa in the cytoplasm of spermatogenic cells (Fig. 5G). Strong staining was detected in the spermatogonia near the base of the germinal epithelium, but it was absent in the more differentiated spermatids in the center of the testis. We also investigated Vasa protein distribution during amphi- oxus embryogenesis by whole-mount immunostaining (Figs. 5I–O). In

Fig. 5. (A) Western blots of proteins from lysates of unfertilized (UF) eggs probed with pre-immune, which recognized a band of ~30 kDa (arrowhead) or with anti-BfVasa serum, which recognized a single band of approximately 80 kDa. (B) Western blots of proteins from lysates of different developmental stages probed with anti-BfVasa serum. The same blot was stripped and re-probed with anti-α-Tubulin as a control for loading. (C–H) Immunohistochemical staining of sections of adult amphioxus (C–E, female; F–H, male) with anti-BfVasa serum (C, D, F, G) and control pre-immune serum (E, H). Small oocytes in ovary (o) are indicated by the black arrowheads in D and E. White arrows indicate germinal vesicles (GV). (G, H) Spermatogonia in adult testes are indicated by black arrows. No specific staining was detected with pre-immune serum. Scale bars, 50 μm. (I–O) Whole-mounts of amphioxus embryos labeled with anti-BfVasa serum. (I) Fertilized egg; (J) blastula; (K) mid-gastrula; (L) early neurula; (M) mid-neurula; (N) late-neurula; (O) early larva. Black arrows indicate labeling in presumptive PGCs. Black arrowheads indicate zygotic protein signals around blastopore. Scale bars, 50 μm. H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 153 fertilized eggs and embryos from early cleavages to the early gastrula hybridization with the Vasa RNA probe after the mid-gastrula stage stage, Vasa protein appeared to be present ubiquitously in the (Figs. 5K–O), suggesting co-localization of Vasa protein and RNA to cytoplasm, and we did not detect clear aggregation of Vasa protein the same cell population: the presumptive PGCs in amphioxus. with chromogenic substrates (Figs. 5I and J). Clear localized protein Protein signals were also present around the closing blastopore after signals were observed in the same cell cluster shown by in situ mid-gastrula stage.

Fig. 6. Fluorescent double-labeling of Vasa mRNA (green) and Vasa protein (red). (A–T) Confocal microscopic images of eggs and cleavage stage embryos. (A, E, I, M, Q) Side views of whole embryos; animal pole at the top. High magnification views of the boxed zone in each stage are shown below each whole mount. After fertilization, Vasa mRNA became localized to the periphery of the pole plasm (white arrowheads in the merged images), whereas Vasa protein appeared to occupy the central position of the pole plasm (white arrows). (U–I′) High-magnification images of Vasa-positive cells showing the dynamic distribution of Vasa mRNA and protein from the blastula stage to the neurula stage. Embryos were counterstained with DAPI. Vasa protein began to accumulate in perinuclear spots (white arrows). The cell inheriting the maternal Vasa mRNA (white arrowheads) divided into multiple cells. Numbers in Z, C′,F′, and I′ indicate the nuclei associated with Vasa mRNA and protein. Scale bars, 10 μm. 154 H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159

Additionally, we combined more sensitive fluorescent double- We also found that from the mid-gastrula stage to the late neurula staining methods with confocal microscopy to detect Vasa mRNA and stage, the number of the ventral Vasa-positive cells steadily increased Vasa protein distributions during early amphioxus development. In to at least eight cells in a single cluster before they merge into the unfertilized eggs, Vasa mRNA was clearly localized in patches near the most posterior ventral endoderm (Figs. 6C′,F′,I′; and 7B–D, I). vegetal cortex, whereas diffuse Vasa mRNA signals were present Although we could recognize a patch of cells with slightly higher Vasa throughout the entire egg (Fig. 6A). Although the protein signals were protein signals at the posterior end of the hindgut endoderm in the mostly diffuse and cytoplasmic, at a higher magnification, we observed a larval stage (Figs. 7F and H, white arrows), to what extent these cells slightly higher level of Vasa protein near the Vasa RNA patches in the continue to increase at this site is uncertain as it becomes difficult to pole plasm (Figs. 6B–D). However, the mRNA and protein signals did not distinguish them from the posterior tail bud cells, which are also precisely coincide, suggesting that they are most likely in different expressing Vasa zygotically (Figs. 7E–H). The amphioxus tail bud, cytoplasmic structures. In fertilized eggs and cleavage stage embryos, a located at the junction of the posterior developing neural tube higher concentration of Vasa protein was increasingly associated with (neurenteric canal) and mesendoderm, has been proposed to be the the aggregated Vasa RNA (Figs. 6E–T). The Vasa protein was also more posterior growth zone that supplies new cells for elongation of the centrally located within the pole plasm, whereas Vasa RNA localized neural tube, notochord, somites, and hindgut endoderm (Schubert primarily to its periphery (Figs. 6H, L, P and T). In addition, diffuse et al., 2001). Therefore, it is highly likely that these Vasa-positive tail cytoplasmic Vasa protein was still detectable in cleavage stage embryos bud cells represent undifferentiated somatic stem cells in amphioxus. (Figs. 6E, I and M). Beginning at the blastula stage, Vasa protein was Thus, in amphioxus, Vasa/Nanos might function not only in the detected as faint perinuclear staining in a punctate pattern, and the specification of PGCs, but also in the formation of multipotent somatic protein signals that co-localized with Vasa mRNA decreased consider- stem cells. This is in line with recent evidence that many conserved ably (Figs. 6U–W). At the mid-gastrula stage, when the Vasa mRNA- “germline genes”, including Vasa and Nanos, are required for the containing pole plasm dispersed to the cytoplasm, the perinuclear Vasa formation of multipotent progenitor cell types in a wide variety of protein was most conspicuous in the cell harboring Vasa mRNA animals (Gustafson and Wessel, 2010; Juliano and Wessel, 2010). (Figs. 6X–Z); however, faint immunostaining of perinuclear protein was also detected in other cells throughout the embryo (Figs. 6YandZ). Blastomere separations of two-cell stage embryos cause the disappearance Beginning at the late gastrula stage, when the Vasa RNA-positive cell of ventral Vasa-positive cells from one of the resulting twin embryos divides into multiple cells, the perinuclear protein signals were co- localized with Vasa RNA in the same migrating cells (Figs. 6A′–I′)In To test the hypothesis that vegetal pole plasm is required for the addition, Vasa proteins were also present in a domain surrounding the production of presumptive PGCs in amphioxus, we performed blastopore with zygotic Vasa expression (Fig. 7A, arrowhead). In the rest blastomere separations in two-cell stage embryos and allowed the of the embryo, the Vasa protein becomes mostly undetectable (Fig. 7A). separated blastomeres to develop until the control embryos from the The dynamic redistribution pattern of amphioxus Vasa protein in the same batch of embryos reached the early neurula stage. This end point presumptive PGCs and somatic cells during early amphioxus develop- was initially chosen because the presumptive PGCs can be recognized ment is highly reminiscent of the situation observed in zebrafish (Knaut easily by Vasa-positive staining at this time as a cell cluster located et al., 2000) and ascidians (Shirae-Kurabayashi et al., 2006), which between the posterior ventral ectoderm and mesendoderm, just suggests that the preferential expression of Vasa gene products in PGCs, before they merge into the posterior zygotic Vasa-expressing domain. along with the fast turnover of Vasa protein in somatic cells, are The aggregated vegetal pole plasm segregates asymmetrically into conserved mechanisms for PGC specification. only one of the blastomeres at the two-cell stage (Fig. 3). If vegetal

Fig. 7. (A–D) A late neurula labeled with Bf-Vasa anti-serum to show presumptive PGCs. B–D are high-magnification serial z-stack views of the boxed area in A. Eight Vasa-positive cells are visible between the ectoderm (ec) and mesendoderm (men) before they merge into the posterior zygotic Vasa-expressing domain (white arrowhead). Scale bars, 10 μm. (E–H) Vasa protein localization in early larval stages. White arrow indicates cells in the posterior growth zone containing Vasa protein. Labeling with acetylated alpha-tubulin antibody (green) reveals cilia on the epidermis and inside the neural tube. Scale bar, 50 μm. (I) Percentages of the observed number of Vasa-positive cells derived from the blastomere containing pole plasm at different stages of amphioxus development. Sample size for each developmental stage is shown on top. Numbers of Vasa-positive cells are denoted by different colors. B, blastula; EG, early gastrula; MG, mid gastrula; LG, late gastrula; EN, early neurula; MN, mid neurula; LN, late neurula. H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 155 pole plasm is required to produce PGCs, then only one of the resulting (Figs. 8A–C; Suppl. Fig. S6). This observation is consistent with the twins would contain PGCs. Indeed, in 9 out of 10 cases, a ventral Vasa- hypothesis that the vegetal pole plasm is required for the formation of positive cell cluster was only present in one of the two twin embryos presumptive PGCs in amphioxus embryos. In one case, we could not resulting from blastomeres dissociated at the of two-cell stage recognize the presence of Vasa-positive cells between the posterior

Fig. 8. Vasa mRNA and protein localization after blastomere separation. (A) Control embryos. Green represents the Vasa mRNA expression domain. Vasa-positive presumptive PGCs (black arrow in right-hand panel) were present in all of the 25 wild-type embryos examined. Scale bar, 50 μm. (B) Diagram of the blastomere separation experiments. (C) Vasa expression in early neurula stage embryos derived from single blastomeres separated at the two-cell stage. Black arrow indicates the Vasa-positive presumptive PGCs. Star denotes absence of ventral Vasa-positive PGCs. Twins-1a and Twins-1b are derived from the same two-cell stage embryo; similarly, Twins-2a and Twins-2b are derived from one two-cell stage embryo. (D) Vasa protein distribution in early larvae derived from single blastomeres separated at the two-cell stage. Double arrowhead in bright field images at left indicates the normal tail bud, and triple arrowhead indicates the corresponding area in the curly-tail phenotype. White arrow indicates Vasa protein (red) in the tail bud. Star denotes the absence of Vasa protein and the disorganized posterior end in the curly-tailed larva. 156 H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 ventral mesendoderm and ectoderm, apart from the zygotic expres- (summarized in Fig. 9). Thus, our results strongly suggest that amphioxus sion domain around the closing blastopore in both twin embryos specify their presumptive PGCs by localized maternal cytoplasm [i.e. (Fig. 8C, Twins-2a and Twins-2b). It is possible that in this pair, the vegetal pole plasm (Holland and Holland, 1992)] and, therefore, Vasa-positive cells had fused into the zygotic Vasa-expressing domain. amphioxus can be placed in a short list of animals that specify their We also observed that the zygotic Vasa-expressing domain near the germ cells by preformation (Extavour and Akam, 2003; Extavour, 2007). posterior closing blastopore was mostly unaffected after blastomere In addition to amphioxus, many species in the other two chordate separation compared to the control embryos, suggesting that this lineages, the urochordates and , also have maternally localized zygotic Vasa expression is independently controlled and does not rely germ plasm (Brown et al., 2009; Fujimura and Takamura, 2000; Komiya on the presence of maternal vegetal pole plasm. et al., 1994; Shirae-Kurabayashi et al., 2006; Tsunekawa et al., 2000; Yoon To further test the function of vegetal pole plasm during amphioxus et al., 1997); thus, using localized cytoplasm for germ cell specification development, we performed additional blastomere separation experi- seems to be a widespread mechanism among chordates. However, in ments as described above in two-cell stage embryos but allowed the vertebrates, it has been suggested that an epigenetic induction separated blastomeres develop to the larval stage (Fig. 8D). Surprisingly, mechanism is the basal mode of PGC formation, and germ plasm arose although all of these embryos expressed Vasa mRNA zygotically near the independently in several lineages (Bachvarova et al., 2009; tailbud during development, there was always one larva (most likely Johnson et al., 2003). This raises the question of whether the presence of derived from the blastomere lacking the maternal vegetal plasm) that germ plasm in the amphioxus B. floridae is an ancestral or derived feature developed a distinct curly-tail phenotype with a much lower number of within the cephalochordates. Additional work on the molecular Vasa protein-positive cells near the posterior end by the L2 stage (2–3 mechanisms underlying germ cell formation using species from the days post-fertilization) (Fig. 8D and Suppl. Fig. S7). Close examination other major group of amphioxus, Asymmetron,suchasAsymmetron revealed that the posterior structure of these larvae was disorganized; lucayanum (Holland and Holland, 2010), should help to answer this the posterior notochord, neural tube (especially the neurenteric canal), question. and the opening of the hindgut (anus) were not properly formed Although localized germ plasm can also be found in some (Fig. 8D and Suppl. Fig. S7), indicating that the loss of maternal vegetal protostome species, such as C. elegans, Drosophila, and chaetognaths, plasm also causes serious defects in the posterior growth zone of an epigenetic mechanism of PGC formation appears to be more common amphioxus larvae. Therefore, our results suggest a model in which among most of the bilaterian phyla studied to date and is widely localized maternal vegetal pole plasm is necessary for specifying considered the ancestral mechanism (Extavour and Akam, 2003; presumptive PGCs in amphioxus embryos, and the proper production Extavour, 2007; Voronina et al., 2008). Intriguingly, chaetognaths, of presumptive PGCs is instrumental in the organization of the posterior which exhibit deuterostome-like development, also have early aggre- growth zone in amphioxus larvae. We hypothesize that after these gated vegetal pole germ plasm that is highly similar to that in presumptive PGCs reach the tail bud, they may send out signals that are amphioxus eggs. Furthermore, this germ plasm is inherited by a single necessary to maintain zygotic Vasa protein accumulation in the blastomere at each cleavage until the gastrula stage (Carre et al., 2002; multipotent somatic stem cells in the amphioxus tail bud. Alternatively, Hyman, 1951). Chaetognaths are fast-evolving, and their phylogenetic these migratory Vasa protein-positive cells may not be obligate PGCs position is uncertain, with different molecular analyses placing them in and may directly contribute to the tail bud to become multipotent various positions within the Bilateria (Helmkampf et al., 2008; Mallatt progenitor cells in the posterior growth zone. Subsequently, these and Winchell, 2002; Telford and Holland, 1993), although recent progenitor cells could give rise not only to germ cells but also to somatic phylogenomic studies favor a protostome affiliation for this group tissues in amphioxus. It will require precise cell labeling or ablation (Dunn et al., 2008; Marletaz et al., 2006; Matus et al., 2006). Moreover, studies on both presumptive PGCs and tail bud cells to further address these recent studies appear to support the chaetognaths branching out this question. close to the base of the protostome tree. The remarkable similarity In the sea urchin Strongylocentrotus purpuratus and the ascidian between amphioxus (a basal chordate in the deuterostomes) and Ciona intestinalis, embryos depleted of Vasa-positive cells during early chaetognaths (a putative basal protostome phylum) in their mode of embryogenesis are capable of re-establishing localized Vasa distribu- germ plasm formation may thus add another interesting example of the tion during later developmental stages and growing into fertile adults, convergence of developmental mechanisms in animals belonging to suggesting that compensatory mechanisms for germ cell formation disparate animal phyla. exist in these organisms (Takamura et al., 2002; Voronina et al., 2008). It has been demonstrated that in many animal taxa, primordial germ We attempted to raise the larvae derived from single blastomeres for cells and somatic stem cells share a common origin; moreover, a prolonged period, but unfortunately, these deformed larvae suffered developmental studies have shown that conserved genes, such as from high mortality and did not survive beyond 10 days post- Vasa, Nanos,andPiwi,functioninthespecification and maintenance of fertilization (data not shown). As this precluded us from investigating both cell types across different metazoan phyla (reviewed in Juliano and whether there is a compensatory mechanism that can induce PGCs Wessel, 2010). For example, in sponges, the archeocytes are the epigenetically during later development in amphioxus, we cannot pluripotent stem cells that can give rise to both somatic and germ exclude the possibility that in juveniles or adults, other multipotent cells, and Piwi homologs are expressed in this cell type (Funayama et al., somatic stem cells may also contribute to the germline. Further 2010). The adult multipotent interstitial stem cells in hydra (expressing experiments will be required to test this possibility. Nevertheless, our Vasa and Nanos; Mochizuki et al., 2000; Mochizuki et al., 2001) and the observation strongly suggests that the cell population specified by the neoblasts in planarians (expressing Vasa, Nanos, and Piwi; Pfister et al., maternal vegetal pole plasm is necessary for the function of the 2008; Shibata et al., 2010) also have similar stem cell properties and can posterior growth zone in amphioxus larvae; the loss of this cell differentiate into germ cells and somatic cells. In the polychaete annelid population is detrimental and causes mortality at larval stages. Platynereis dumerilii, Vasa, Nanos,andPiwi are found in the highly proliferating mesodermal posterior growth zone (MPGZ), which can Conclusions also give rise to both germ cells and somatic tissues (Rebscher et al., 2007). In various sea urchin species, Vasa, Nanos,andPiwi are expressed Our data indicate that the gene products of Vasa and Nanos are reliable in descendants of the small micromeres and subsequently become germline markers in amphioxus. Contrary to the previous hypothesis that restricted to the coelomic pouches, from which the entire adult amphioxus may form germ cells epigenetically in late larval develop- rudiment will form, suggesting that these conserved molecular factors ment, our results clearly demonstrate that maternal germline determi- are involved in the formation of multipotent progenitor cells that nants are asymmetrically localized very early in amphioxus development contribute to the generation of the entire adult body, including both H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 157

Fig. 9. Schematic model of germ plasm distribution and presumptive PGC formation in amphioxus. The distribution of Vasa mRNA is shown in green and Vasa protein in red. UF egg, unfertilized egg; A, animal side; V, vegetal side; GV, germinal vesicle; 1st PB, first polar body; BC, blastocoel; EG, early gastrula; MG, mid-gastrula stage; BP, blastopore; LG, late gastrula stage; EN, early neurula stage; MN, mid-neurula stage; FBC, fin box coelom; NC, nerve cord; MYT, myotome; NOT, notochord; PH, pharynx; HC, hepatic cecum; MFD, metapleural fold; ESO, early stage oocyte; SPC, spermatocyte.

somatic cells and germ cells (Fujii et al., 2006; Juliano et al., 2006, 2010; mingle and interact with each other in the posterior growth zone in Voronina et al., 2008). In addition, species lacking small amphioxus and whether they are obligate germ cells and somatic stem micromeres, such as sea stars, also have Vasa protein and/or transcripts cells, respectively, remain open questions for further studies. Detailed enriched in the larval coelomic pouches, suggesting a conserved cell labeling experiments and additional studies on the roles of other mechanism for the formation of multipotent progenitor cells at the common “germline genes”, such as Piwi, Tudor, Boule,andPumilio, coelomic pouch to produce an adult rudiment within the during amphioxus development may help to resolve these issues. (Juliano and Wessel, 2009). Our data also indicate a close relationship Finally, despite the highly regulative nature of early amphioxus between presumptive PGCs and multipotent somatic stem cells in development that had been demonstrated by early embryologists amphioxus. However, we identified two distinct Vasa protein-positive (Conklin, 1933; Tung et al., 1958; Wilson, 1893;andreviewedin cell populations during early amphioxus embryogenesis: the first cell Holland and Holland, 2007), our data raise a surprising issue that the population inherits maternal vegetal pole plasm containing Vasa/Nanos first two blastomeres in amphioxus embryos are different. Although gene products and migrates posteriorly during neurulation to merge Wilson mentioned that many larvae derived from isolated blastomeres with the second cell population, which expresses Vasa/Nanos zygotically of the two-cell stage embryos sometimes had abnormal tails (Wilson, in the posterior growth zone. We hypothesize that these two cell 1893), he and the following researchers appeared to overlook this populations may represent the primordial germ cells and somatic stem phenomenon and made the general conclusion that blastomeres cells in amphioxus, respectively. Whether these two populations of cells isolated at two-cell stage can develop into typical embryos and larvae. 158 H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159

We repeated these classic experiments in the present study with Extavour, C.G., Akam, M., 2003. Mechanisms of germ cell specification across the metazoans: epigenesis and preformation. Development 130, 5869–5884. modern molecular markers, which enabled us to reveal the differences Extavour, C.G., Pang, K., Matus, D.Q., Martindale, M.Q., 2005. vasa and nanos expression between those seemingly identical twin dwarf neurulae, and subse- patterns in a sea anemone and the evolution of bilaterian germ cell specification quently let us recognize that the curly-tail phenotype in those larvae mechanisms. Evol. Dev. 7, 201–215. fi Fabioux, C., Huvet, A., Lelong, C., Robert, R., Pouvreau, S., Daniel, J.Y., Minguant, C., Le may be the result from lacking vegetal pole plasm in one of the rst two Pennec, M., 2004. Oyster vasa-like gene as a marker of the germline cell blastomeres. It is tempting to speculate that in addition to the germ cell development in Crassostrea gigas. Biochem. Biophys. Res. Commun. 320, 592–598. determinants, the vegetal pole plasm of amphioxus eggs may also Fujii, T., et al., 2006. Developmental expression of HpNanos,theHemicentrotus – contain certain somatic determinants, similar to those in the polarized pulcherrimus homologue of nanos. Gene Expr. Patterns 6, 572 577. Fujimura, M., Takamura, K., 2000. Characterization of an ascidian DEAD-box gene, Ci- cytoplasms of Xenopus and ascidian eggs (Sardet et al., 2007; White and DEAD1: specific expression in the germ cells and its mRNA localization in the Heasman, 2008). Therefore, although the first two blastomeres of posterior-most blastomeres in early embryos. Dev. Genes Evol. 210, 64–72. ascidian embryos never give rise to complete embryos if separated, Funayama, N., Nakatsukasa, M., Mohri, K., Masuda, Y., Agata, K., 2010. Piwi expression in archeocytes and choanocytes in demosponges: insights into the stem cell system in amphioxus is not entirely unlike ascidians and frogs in the cytoplasmic demosponges. Evol. Dev. 12, 275–287. localization of determinants after fertilization. More work is needed to Gustafson, E.A., Wessel, G.M., 2010. Vasa genes: emerging roles in the germ line and in further characterize the molecular components and functions of vegetal multipotent cells. Bioessays 32, 626–637. Helmkampf, M., Bruchhaus, I., Hausdorf, B., 2008. Multigene analysis of lophophorate pole plasm in amphioxus to shed more light on the evolution of early and chaetognath phylogenetic relationships. Mol. Phylogenet. Evol. 46, 206–214. patterning mechanisms in chordates. Holland, L.Z., Holland, N.D., 1992. Early development in the lancelet (=amphioxus) Branchiostoma floridae from sperm entry through pronuclear fusion — presence of vegetal pole plasm and lack of conspicuous ooplasmic segregation. Biol. Bull. 182, Accession numbers 77–96. Holland, L.Z., Holland, N.D., 2007. A revised fate map for amphioxus and the evolution of – Amphioxus (B. floridae) Nanos and Vasa sequences were deposited axial patterning in chordates. Integr. Comp. Biol. 47, 360 372. Holland, L.Z., Yu, J.K., 2004. Cephalochordate (amphioxus) embryos: procurement, in the NCBI GenBank database with the accession numbers HM004549 culture, and basic methods. Methods Cell. Biol. 74, 195–215. (Bf-Nanos) and HM004550 (Bf-Vasa). Holland, L.Z., Holland, N.D., 1991. The fine-structure of the growth stage oocytes of a lancelet (=amphioxus), Branchiostoma lanceolatum. Invertebr. Reprod. Dev. 19, 107–122. Acknowledgments Holland, L.Z., Holland, N.D., 2010. Laboratory spawning and development of the Bahama lancelet, Asymmetron lucayanum (cephalochordata): fertilization through feeding larvae. Biol. Bull. 219, 132–141. We would like to thank Dr. Li-Tzu Li and the personnel of the ICOB Hyman, L.H., 1951. The Invertebrates: Smaller Coelomate Groups. McGraw-Hill, core facility for excellent technical assistance. We would also like to New York. thank Dr. John Lawrence and Dr. Susan Bell at the University of South Ikenishi, K., Tanaka, T.S., 2000. Spatio-temporal expression of Xenopus vasa homolog, Florida for providing laboratory facilities during the summer breeding XVLG1, in oocytes and embryos: the presence of XVLG1 RNA in somatic cells as well as germline cells. Dev. Growth Differ. 42, 95–103. season of amphioxus. We thank the two anonymous reviewers for Jeanmougin, F., Thompson, J.D., Gouy, M., Higgins, D.G., Gibson, T.J., 1998. Multiple their helpful comments on the manuscript. This work was supported sequence alignment with Clustal X. Trends Biochem. Sci. 23, 403–405. by intramural funding from the Institute of Cellular and Organismic Johnson, A.D., Drum, M., Bachvarova, R.F., Masi, T., White, M.E., Crother, B.I., 2003. Evolution of predetermined germ cells in vertebrate embryos: implications for Biology, Academia Sinica, Taiwan (to Y.H.S. and J.K.Y.), grants from the macroevolution. Evol. Dev. 5, 414–431. National Science Foundation of the USA (to L.Z.H.), and the Career Juliano, C.E., Wessel, G.M., 2009. An evolutionary transition of Vasa regulation in Development Award from Academia Sinica, Taiwan (to J.K.Y.). echinoderms. Evol. Dev. 11, 560–573. Juliano, C., Wessel, G., 2010. Versatile germline genes. Science 329, 640–641. Juliano, C.E., Voronina, E., Stack, C., Aldrich, M., Cameron, A.R., Wessel, G.M., 2006. Germ Appendix A. Supplementary data line determinants are not localized early in sea urchin development, but do accumulate in the small micromere lineage. Dev. Biol. 300, 406–415. Juliano, C.E., Yajima, M., Wessel, G.M., 2010. Nanos functions to maintain the fate of the Supplementary data to this article can be found online at small micromere lineage in the sea urchin embryo. Dev. Biol. 337, 220–232. doi:10.1016/j.ydbio.2011.02.014. Knaut, H., Pelegri, F., Bohmann, K., Schwarz, H., Nusslein-Volhard, C., 2000. Zebrafish vasa RNA but not its protein is a component of the germ plasm and segregates asymmetrically before germline specification. J. Cell Biol. 149, 875–888. References Komiya, T., Itoh, K., Ikenishi, K., Furusawa, M., 1994. Isolation and characterization of a novel gene of the DEAD box protein family which is specifically expressed in germ Bachvarova, R.F., Crother, B.I., Johnson, A.D., 2009. Evolution of germ cell development cells of Xenopus laevis. Dev. Biol. 162, 354–363. in tetrapods: comparison of urodeles and amniotes. Evol. Dev. 11, 603–609. Mallatt, J., Winchell, C.J., 2002. Testing the new animal phylogeny: first use of combined Bourlat, S.J., et al., 2006. Deuterostome phylogeny reveals monophyletic chordates and large-subunit and small-subunit rRNA gene sequences to classify the protostomes. the new phylum Xenoturbellida. Nature 444, 85–88. Mol. Biol. Evol. 19, 289–301. Bridgham, J.T., Brown, J.E., Rodriguez-Mari, A., Catchen, J.M., Thornton, J.W., 2008. Marletaz, F., et al., 2006. Chaetognath phylogenomics: a protostome with deutero- Evolution of a new function by degenerative mutation in cephalochordate steroid stome-like development. Curr. Biol. 16, R577–R578. receptors. PLoS Genet. 4, e1000191. Matus, D.Q., Copley, R.R., Dunn, C.W., Hejnol, A., Eccleston, H., Halanych, K.M., Brown, F.D., Swalla, B.J., 2007. Vasa expression in a colonial ascidian, Botrylloides Martindale, M.Q., Telford, M.J., 2006. Broad taxon and gene sampling indicate violaceus. Evol. Dev. 9, 165–177. that chaetognaths are protostomes. Curr. Biol. 16, R575–R576. Brown, F.D., Tiozzo, S., Roux, M.M., Ishizuka, K., Swalla, B.J., De Tomaso, A.W., 2009. Early Mochizuki, K., Sano, H., Kobayashi, S., Nishimiya-Fujisawa, C., Fujisawa, T., 2000. lineage specification of long-lived germline precursors in the colonial ascidian Expression and evolutionary conservation of nanos-related genes in Hydra. Dev. Botryllus schlosseri. Development 136, 3485–3494. Genes Evol. 210, 591–602. Carre, D., Djediat, C., Sardet, C., 2002. Formation of a large Vasa-positive germ granule Mochizuki, K., Nishimiya-Fujisawa, C., Fujisawa, T., 2001. Universal occurrence of the and its inheritance by germ cells in the enigmatic Chaetognaths. Development 129, vasa-related genes among metazoans and their germline expression in Hydra. Dev. 661–670. Genes Evol. 211, 299–308. Chang, C.C., Dearden, P., Akam, M., 2002. Germ line development in the grasshopper Nieuwkoop, P.D., Sutasurya, L.A., 1979. Primordial Germ Cells in the Chordates. Schistocerca gregaria: vasa as a marker. Dev. Biol. 252, 100–118. Cambridge Univ. Press, Cambridge. Conklin, E.G., 1932. The embryology of amphioxus. J. Morphol. 54, 69–140. Ozhan-Kizil, G., Havemann, J., Gerberding, M., 2009. Germ cells in the crustacean Conklin, E.G., 1933. The development of isolated and partially separated blastomeres of Parhyale hawaiensis depend on Vasa protein for their maintenance but not for their amphioxus. J. Exp. Zool. 64, 303–375. formation. Dev. Biol. 327, 230–239. Dill, K.K., Seaver, E.C., 2008. Vasa and nanos are coexpressed in somatic and germ line Pfister, D., De Mulder, K., Hartenstein, V., Kuales, G., Borgonie, G., Marx, F., Morris, J., tissue from early embryonic cleavage stages through adulthood in the polychaete Ladurner, P., 2008. stem cells and the germ line: developmental and Capitella sp. I. Dev. Genes Evol. 218, 453–463. evolutionary implications of macvasa expression in Macrostomum lignano. Dev. Dunn, C.W., et al., 2008. Broad phylogenomic sampling improves resolution of the Biol. 319, 146–159. animal tree of life. Nature 452, 745–749. Putnam, N.H., et al., 2008. The amphioxus genome and the evolution of the chordate Ewen-Campen, B., Schwager, E.E., Extavour, C.G., 2010. The molecular machinery of karyotype. Nature 453, 1064–1071. germ line specification. Mol. Reprod. Dev. 77, 3–18. Rebscher, N., Zelada-Gonzalez, F., Banisch, T.U., Raible, F., Arendt, D., 2007. Vasa unveils Extavour, C.G.M., 2007. Evolution of the bilaterian germ line: lineage origin and a common origin of germ cells and of somatic stem cells from the posterior growth modulation of specification mechanisms. Integr. Comp. Biol. 47, 770–785. zone in the polychaete Platynereis dumerilii. Dev. Biol. 306, 599–611. H.-R. Wu et al. / Developmental Biology 353 (2011) 147–159 159

Richardson, B.E., Lehmann, R., 2010. Mechanisms guiding primordial germ cell Tsunekawa, N., Naito, M., Sakai, Y., Nishida, T., Noce, T., 2000. Isolation of chicken vasa migration: strategies from different organisms. Nat. Rev. Mol. Cell Biol. 11, 37–49. homolog gene and tracing the origin of primordial germ cells. Development 127, Rosner, A., Moiseeva, E., Rinkevich, Y., Lapidot, Z., Rinkevich, B., 2009. Vasa and the germ 2741–2750. line lineage in a colonial urochordate. Dev. Biol. 331, 113–128. Tung, T.C., Wu, S.C., Tung, Y.F., 1958. The development of isolated blastomeres of Ruppert, E.E., 1997. Cephalochordata (Acrania). In: Harrison, F.W., Ruppert, E.E. (Eds.), Amphioxus. Sci. Sin. 7, 1280–1320. Microscopic Anatomy of Invertebrates, Vol. 15. Wiley-Liss, New York, pp. 349–504. Voronina, E., et al., 2008. Vasa protein expression is restricted to the small micromeres Sardet, C., Paix, A., Prodon, F., Dru, P., Chenevert, J., 2007. From oocyte to 16-cell of the sea urchin, but is inducible in other lineages early in development. Dev. Biol. stage: cytoplasmic and cortical reorganizations that pattern the ascidian embryo. 314, 276–286. Dev. Dyn. 236, 1716–1731. White, J.A., Heasman, J., 2008. Maternal control of pattern formation in Xenopus laevis. Schubert, M., Holland, L.Z., Stokes, M.D., Holland, N.D., 2001. Three amphioxus Wnt J. Exp. Zool. B 310B, 73–84. genes (AmphiWnt3, AmphiWnt5, and AmphiWnt6) associated with the tail bud: the Wilson, E.B., 1893. Amphioxus, and the mosaic theory of development. J. Morphol. 8, evolution of somitogenesis in chordates. Dev. Biol. 240, 262–273. 579–639. Shibata, N., Rouhana, L., Agata, K., 2010. Cellular and molecular dissection of pluripotent Yoon, C., Kawakami, K., Hopkins, N., 1997. Zebrafish vasa homologue RNA is localized to adult somatic stem cells in planarians. Dev. Growth Differ. 52, 27–41. the cleavage planes of 2- and 4-cell-stage embryos and is expressed in the Shirae-Kurabayashi, M., Nishikata, T., Takamura, K., Tanaka, K.J., Nakamoto, C., Nakamura, primordial germ cells. Development 124, 3157–3165. A., 2006. Dynamic redistribution of vasa homolog and exclusion of somatic cell Yu, J.K.S., Holland, L.Z., 2009a. Amphioxus (Branchiostoma floridae) spawning and determinants during germ cell specification in Ciona intestinalis. Development 133, embryo collection. CSH Protoc. doi:10.1101/pdb.prot5285. 2683–2693. Yu, J.K.S., Holland, L.Z., 2009b. Amphioxus whole-mount in situ hybridization. CSH Swartz, S.Z., Chan, X.Y., Lambert, J.D., 2008. Localization of Vasa mRNA during early Protoc. doi:10.1101/pdb.prot5286. cleavage of the snail Ilyanassa. Dev. Genes Evol. 218, 107–113. Yu, J.K.S., Holland, L.Z., 2009c. Extraction of RNA from small amounts of amphioxus Takamura, K., Fujimura, M., Yamaguchi, Y., 2002. Primordial germ cells originate from embryos for cDNA synthesis by RT-PCR. CSH Protoc. doi:10.1101/pdb.prot5289. the endodermal strand cells in the ascidian Ciona intestinalis. Dev. Genes Evol. 212, Yu, J.K., Meulemans, D., McKeown, S.J., Bronner-Fraser, M., 2008a. Insights from the 11–18. amphioxus genome on the origin of vertebrate neural crest. Genome Res. 18, Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: molecular evolutionary genetics 1127–1132. analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599. Yu, J.K., Wang, M.C., Shin, I.T., Kohara, Y., Holland, L.Z., Satoh, N., Satou, Y., 2008b. A Telford, M.J., Holland, P.W., 1993. The phylogenetic affinities of the chaetognaths: a cDNA resource for the cephalochordate amphioxus Branchiostoma floridae. Dev. molecular analysis. Mol. Biol. Evol. 10, 660–676. Genes Evol. 218, 723–727.