Novel animal pole-enriched maternal mRNAs are preferentially expressed in neural ectoderm

Authors

  • Paaqua A. Grant,

    1. Department of Biological Sciences, The George Washington University, Washington
    Search for more papers by this author
  • Bo Yan,

    1. Department of Anatomy and Regenerative Biology, The George Washington University, School of Medicine and Health Sciences, Washington
    Search for more papers by this author
  • Michael A. Johnson,

    1. Department of Biological Sciences, The George Washington University, Washington
    Search for more papers by this author
  • Diana L.E. Johnson,

    1. Department of Biological Sciences, The George Washington University, Washington
    Search for more papers by this author
  • Sally A. Moody

    Corresponding author
    1. Department of Anatomy and Regenerative Biology, The George Washington University, School of Medicine and Health Sciences, Washington
    • Correspondence to: Sally A. Moody, PhD, Department of Anatomy and Regenerative Biology, The George Washington University, School of Medicine and Health Sciences, 2300 I Street, N.W., Washington, DC 20037. E-mail: samoody@gwu.edu

    Search for more papers by this author

Abstract

Background: Many animals utilize maternal mRNAs to pre-pattern the embryo before the onset of zygotic transcription. In Xenopus laevis, vegetal factors specify the germ line, endoderm, and dorsal axis, but there are few studies demonstrating roles for animal-enriched maternal mRNAs. Therefore, we carried out a microarray analysis to identify novel maternal transcripts enriched in 8-cell-stage animal blastomeres. Results: We identified 39 mRNAs isolated from 8-cell animal blastomeres that are >4-fold enriched compared to vegetal pole mRNAs. We characterized 14 of these that are of unknown function. We validated the microarray results for 8/14 genes by qRT-PCR and for 14/14 genes by in situ hybridization assays. Because no developmental functions are reported yet, we provide the expression patterns for each of the 14 genes. Each is expressed in the animal hemisphere of unfertilized eggs, 8-cell animal blastomeres, and diffusely in blastula animal cap ectoderm, gastrula ectoderm and neural ectoderm, neural crest (and derivatives) and cranial placodes (and derivatives). They have varying later expression in some mesodermal and endodermal tissues in tail bud through larval stages. Conclusions: Novel animal-enriched maternal mRNAs are preferentially expressed in ectodermal derivatives, particularly neural ectoderm. However, they are later expressed in derivatives of other germ layers. Developmental Dynamics 243:478–496, 2014. © 2013 Wiley Periodicals, Inc.

INTRODUCTION

In many invertebrates, RNAs and proteins that are synthesized during oogenesis are sequestered to different cytoplasmic domains in the oocyte. After fertilization, some of these molecules specify which cells will enter the germ cell lineage (Bowerman, 1998; Nishida, 2002; Becalska and Gavis, 2009), whereas others determine the anterior-posterior (A/P) and dorsal-ventral (D/V) axes of the embryo (Haston and Reijo-Pera, 2007; Kenyon, 2007; Ratnaparkhi and Courey, 2007). In amphibians and fish, there also is evidence for maternal, localized factors that influence early developmental processes (Heasman, 2006; White and Heasman, 2008; Langdon and Mullins, 2011). For example, in Xenopus laevis, several mRNAs that are required for germ line, endoderm, and dorsal axis formation are localized to the vegetal region (Houston and King, 2000; Kloc et al., 2001; Venkataraman et al., 2004; Zhou and King, 2004; Cha et al., 2011; Lai et al., 2012; Mei et al., 2013). Maternal mRNAs also have been identified that are localized to the animal region, which later gives rise to the embryonic ectoderm (reviewed in Sullivan et al., 1998; King et al., 2005). However, there are few studies of the developmental roles of animal localized transcripts in patterning the embryo.

Although the functions of animal-enriched mRNAs in Xenopus are not well characterized, there is convincing evidence that some of them bias dorsal animal blastomeres to form neural tissue (reviewed in Sullivan et al., 1998). Removal of the dorsal animal blastomeres from 8-cell and 16-cell embryos resulted in significant deficits in the neural plate and neural tube (Kageura, 1995). When dorsal animal blastomeres were transplanted to sites that normally give rise to epidermis, they retain their original fate and differentiate into ectopic neural tissue (Gallagher et al., 1991; Moody et al., 2000). Dorsal animal blastomeres autonomously express neural marker genes when cultured as explants, whereas ventral animal and vegetal cells do not (Gallagher et al., 1991; Kageura, 1995). Injecting total RNA isolated from the dorsal animal blastomeres into ventral animal blastomeres caused neural tissue rather than epidermis to form (Hainski and Moody, 1992). Maternal Wnt signaling from ventral blastomeres, in tandem with dorsal localized activation of maternal mRNAs, is required for segregating dorsal and ventral fate (Pandur et al., 2002). Finally, animal-enriched maternal Sox3 and maternal Zic2 mRNAs appear to maintain ectodermal fate by down-regulating Nodal expression, thus limiting the mesodermal and endodermal domains (Zhang et al., 2003, 2004; Zhang and Klymkowsky, 2007; Houston and Wiley 2005). Because embryonic gene transcription does not begin until after these manipulations were performed (Yang et al., 2002), the autonomous ability of the dorsal animal blastomeres to form neural tissue is likely due to maternal molecules. Despite the evidence that animal-localized maternal transcripts can bias blastomeres toward a neural fate, the identities of the involved mRNAs have not been explored in detail.

To generate a more complete list of transcripts localized to the animal hemisphere of the early Xenopus laevis embryo, microarray expression analyses were performed comparing mRNAs contained in animal blastomeres (AN) and those in the vegetal poles of vegetal blastomeres (VP) from 8-cell embryos. We identified 39 mRNAs that are >4-fold enriched in animal blastomeres compared to vegetal pole mRNAs and characterized the developmental expression patterns of 14 that are of unknown function. The temporal and spatial expression patterns of these 14 genes were analyzed by qRT-PCR and in situ hybridization (ISH). From these data, we predict that the heretofore uncharacterized genes may have roles in establishing early ectodermal fate, particularly of neural derivatives, and also have later roles in some mesodermal and endodermal organs.

RESULTS AND DISCUSSION

Several animal-enriched mRNAs have been identified in Xenopus laevis oocytes by candidate gene approaches (reviewed in King et al., 2005; Sindelka et al., 2008, 2010). To identify a more complete list of animal-enriched mRNAs that may regulate the earliest steps of ectodermal specification, we dissected the four animal blastomeres (AN) of the 8-cell embryo, which mostly give rise to ectodermal lineages, and the vegetal poles of the four vegetal blastomeres (VP), which mostly give rise to endodermal and germ cell lineages (Moody and Kline, 1990) (Fig. 1). Many of the previously identified animal-enriched mRNAs were identified in the oocyte animal hemisphere, and could be subject to relocalization after the cytoplasmic reorganizations that occur at fertilization. Therefore, to ensure that we would detect mRNAs that remain in the presumptive ectodermal lineage after fertilization, we isolated cleavage stage blastomeres. Further, we collected samples at the 8-cell stage because it precedes the earliest detection of zygotic transcription (Yang et al., 2002), thus ensuring that any localized mRNAs were transcribed during oogenesis. To control for accurate dissections, each sample was tested for the presence or absence of a known AN transcript (foxD4L1.1, aka foxD5; Sullivan et al., 2001), and a known VP transcript (Vg1; Weeks and Melton, 1987).

Figure 1.

Embryo dissection at the 8-cell stage. An 8-cell Xenopus laevis embryo was fixed in EtOH-acetic acid, and manually separated into three portions: the 4 animal blastomeres (AN, view from blastocoel); the 4 vegetal blastomeres were cut into a vegetal pole (VP) portion and an equatorial (EQ) portion. AN and VP samples were collected for microarray analyses and EQ samples were discarded.

Numerous mRNAs Are Enriched in Animal Blastomeres

Of the ∼29,900 genes represented on the Affymetrix Xenopus laevis GeneChip v2, 96 were >3-fold enriched in the AN samples compared to the VP samples (see Supp. Table S1, which is available online). This agrees with a previous report that animal-localized mRNAs are not more than 3–10-fold enriched in the oocyte (King et al., 2005). Of these, more than half are of unknown function (Fig. 2); the remaining identified genes are associated with metabolism, transcription, transport, and general cellular processes. Two of these >3-fold enriched transcripts were identified previously as animal pole localized: Coco and Grhl1. Coco is a member of the Cerberus/Dan family of secreted BMP inhibitors, and it is proposed to block BMP and TGF-β signaling to modulate ectodermal competence and specification prior to neural induction (Bell et al., 2003). Grhl1 is a member of the Grainyhead-like transcription factor family, and is required for terminal differentiation of epidermis, including being necessary and sufficient to activate expression of epidermis specific keratin (Wilanowski et al., 2002; Tao et al., 2005). Additionally, one of the transcripts identified as >2-fold enriched, foxI2, was identified previously as an animal-enriched mRNA that is required to activate the zygotic ectoderm determinant, foxI1e (Cha et al., 2012). Thus, previously described animal-localized mRNAs are present in our microarray dataset.

Figure 2.

The majority of animal transcripts enriched >3-fold over vegetal pole encode unknown genes. A pie chart identifies the proportion of genes within different gene ontologies.

It was surprising, however, that 16 other mRNAs that were reported to be enriched in the Xenopus oocyte animal hemisphere (reviewed in King et al., 2005) were not detected in our >3-fold enriched animal blastomere transcript dataset. Nine of these were at higher concentrations in our AN samples, but at levels <3-fold enriched (An4, 1.53-fold; XlPOU60 [akapou5f3.3], 1.55-fold; PABP, 2.15-fold; Tcf-1 [aka tcf7], 1.63-fold; Vera, 1.25-fold; xArHβ, 1.22-fold; xl21, 1.32-fold; XGβ1, 2.24-fold; xlan4, 1.2-fold). β-TrCP was not animal enriched in our dataset, most likely because it is transcribed as three mRNAs, one of which is animal and two of which are vegetal enriched (Hudson et al., 1996). Ets1/2 were not animal enriched in our dataset most likely because they localize to both the animal region and the vegetally-located germ plasm (Meyer et al., 1997). Although there is a report of fibronectin mRNA in the animal hemisphere of the oocyte (Oberman and Yisraeli, 1995), these authors did not address whether it is also in the vegetal pole, and our microarray analysis of the 8-cell embryo found it to be slightly enriched vegetally. Four transcripts were expected to be in our dataset based on strong animal expression at cleavage stages, as detected by RNAse protection (An1, An2, An3; Rebagliati et al., 1985) or ISH (XPar-1; Ossipova et al., 2002), but the values from our microarray analysis did not reach statistical significance. We assume that different sensitivities in the different methods of detection likely contribute to our negative findings. Another contributing factor may be that some of these transcripts (An1, An2, XPar1) are most highly concentrated in the animal-central region rather than animal pole of the oocyte when sampled by highly sensitive qPCR tomography (Sindelka et al., 2008, 2010).

Animal-enriched maternal mRNAs also have been identified in zebrafish (gdf6a [aka radar], pou2, runx2b; reviewed in Langdon and Mullins, 2011), and ectoderm progenitor-enriched maternal mRNAs have been identified in sea urchin (Hmx, Tbx2/3; Ben-Tabou de Leon et al., 2013). Interestingly, in Ciona intestinalis a number of maternal transcripts are localized to the posterior-most region of eggs and early embryos, but this report failed to identify any localization to the animal-pole (Yamada et al., 2005). These reports from Xenopus, zebrafish, and marine chordates indicate that identifying animal-enriched maternal transcripts varies widely with species.

It is interesting to note that the subcellular mechanism by which animal transcripts are localized has yet to be revealed. This is in contrast to vegetal mRNA localization, which relies on either association with the vegetally positioned mitochondrial cloud or with microtubular transport (reviewed in King et al., 2005). Measurements of RNA concentrations in 90-μm bins across the animal-vegetal axis of the oocyte reveal that while animal mRNAs are at their highest levels in the animal hemisphere, they are not restricted to this region, suggesting that they are not actively localized (Sindelka et al., 2008, 2010). However, there is one report of cortical localization of an animal mRNA (Schroeder and Yost, 1996), and PHAX and An1 mRNAs each contain cis-acting motifs that are required for vegetal localization but they are not bound to Staufen protein, which is required for vegetally transported mRNAs (Snedden et al., 2013). How animal transcripts acquire and maintain their animal location has yet to be fully explored.

We also identified 177 of the genes represented on the GeneChip that were >3-fold enriched in VP samples compared to the AN samples (Supp. Table S2). Of these, six mRNAs among the most highly enriched (>4 fold) in the VP samples were identified previously as vegetal pole, and/or vegetal cortex localized: Vg1, DEADSouth, Otx1 homolog 1, NIF, acsl1, and brachyury- and Tbx-related (King et al., 2005; Cuykendall and Houston, 2010). Vg1 is a member of the TGF-β family of signaling factors that is involved in the specification of endoderm and mesoderm (Weeks and Melton, 1987; Birsoy et al., 2006; White and Heasman, 2008). DEADSouth is a member of the DEAD-box family of helicases and is likely involved in specifying the germ-line, as it is localized exclusively to the germ plasm of the oocyte and primordial germ cells, and is later expressed in male germ cells (MacArthur et al., 2000). Otx1 homolog 1 is a vertebrate homeodomain-containing transcription factor gene related to Drosophila ocelliless (oc; formerly orthodenticle) that, while present at the vegetal pole of X. laevis oocytes, is later expressed in the brain and developing sensory organs of the head (Kablar et al., 1996). NIF (also known as CTDSPL CTD [carboxy-terminal domain, RNA polymerase II, polypeptide A] small phosphatase-like) has not been well characterized; it may play a role as a tumor suppressor (Kashuba et al., 2004). acsl1 (acyl-CoA synthetase long-chain family member 1) is a member of a large family of enzymes responsible for lipid biosynthesis and fatty-acid degradation, and is required for maintaining meiotic arrest in Xenopus laevis (Wang et al., 2012). brachyury and Tbx-related (aka VegT) is a T-box transcription factor required for both endoderm and mesoderm formation (Stennard et al., 1996; Zhang and King, 1996; Horb and Thomsen, 1997; Zhang et al., 1998). It should be noted that the enrichment values reported in Cuykendal and Houston (2010) differ from ours (Supp. Table S2); this is most likely because we compared animal blastomeres to vegetal pole regions of cleavage stage embryos whereas they compared oocyte vegetal cortex to the whole oocyte. Nonetheless, finding these same localized transcripts in our VP dataset supports the validity of our microarray approach.

Validation of Microarray Results

Of the 96 transcripts observed to be >3-fold enriched in the AN samples compared to the VP samples, 39 were >4-fold enriched (Table 1); of these 56.4% are of unknown function, 10.3% are associated with metabolism, 7.7% are transcription factors, 2.6% are transporters, and the remaining 23% are involved in a variety of general cellular processes. We chose 12 of these (Table 1) and two that were >3.39-fold enriched (Supp. Table S1) for further analysis. Of these, 11/14 were analyzed by qRT-PCR comparison of AN and VP mRNAs (Fig. 3). We were unable to design primers that successfully amplified Xl.16112, Xl.14311, and Xl.52642, perhaps due to the short length of their EST sequences that are currently available. We used a vegetal-localized transcript, Vg1, as an internal control for accurate separation of AN and VP mRNA and found it was significantly (>1.5 fold; P < 0.05) enriched in VP samples as expected (Fig. 3). Of the 11 AN enriched transcripts we could analyze, eight (WBP2NL, SMCR7L, Xl.11373, srebf2, Xl.54800, Xl.15460, Xl.29665, TNFAIP1) were significantly enriched in AN samples and three (Xl.86248, aurka-b, gskip) were not. In addition, the spatial expression patterns of all 14 mRNAs were analyzed at the 8-cell stage by ISH (Fig. 4). Each is enriched in the animal blastomeres of 8-cell embryos, and there is no detectable signal in the vegetal poles. These results demonstrate that the majority of the novel transcripts identified by microarray are animal-enriched, and therefore may have a role in ectoderm development. However, for some (aurka-b, Xl.16112, srebf2, Xl.54800, Xl.15460, Xl.29665, gskip) transcripts were detected in the equatorial portion of the vegetal blastomere (Fig. 4; Table 2), a region that was excluded from the microarray samples (Fig. 1). Consistent with the 8-cell fate maps (Moody and Kline, 1990), five out of seven of these genes also are expressed in the mesoderm at neural plate stages (Table 2).

Table 1. Maternal Transcripts That Showed >4-Fold Enrichment in AN Samples Versus VP Samples at P < 0.05a
Affymetrix probeset IDFold changeGene title or unigene clusterGO
  1. a

    Affymetrix Probeset ID from GeneChip® Xenopus laevis Genome 2.0 Array. GO, gene ontology; highlighted rows, genes selected for validation and further study.

Xl2.14775.1.A1_at9.88Xl.86248Unknown
Xl2.16255.1.S1_at9.21Xl.16255Unknown
Xl2.51884.2.S2_at8.34DAN domain family, member 5 (Coco)signal transduction
Xl2.49110.1.S1_at8.05WBP2 N-terminal likeUnknown
Xl2.10324.1.A1_at7.60Similar to selenoprotein T (MGC53056)Cellular component
Xl2.19681.1.S1_at7.32Hypothetical protein MGC68803Associated with metabolism
Xl2.53404.1.S1_x_at7.09Xl.78855Cellular component
Xl2.10732.1.A1_at6.91Xl.10732Transcription regulator
Xl2.4421.1.S1_at6.80p46XlEg22Kinase
Xl2.25185.1.A1_at6.77Core promoter element binding protein.Transcription regulator
Xl2.7099.1.S1_at6.67Similar to Mannose-1-phosphate guanylyltransferaseAssociated with metabolism
Xl2.16112.1.A1_at6.24Similar to tropicalis gene of unknown function.Unknown
Xl2.3045.3.S1_at6.04Xl.76657Unknown
Xl2.30215.1.S1_at5.68Smith-Magenis syndrome chromosome region, candidate 7-likeUnknown
Xl2.15167.1.A1_at5.62Xl.15167Unknown
Xl2.54204.1.A1_at5.48Xl.54204Unknown
Xl2.45255.1.S1_at5.31CDK2-associated protein 1Cell-cycle
Xl2.11373.1.A1_at5.30Xl.11373Unknown
Xl2.39370.1.A1_at5.18Xl.83466Unknown
Xl2.31931.3.A1_at5.12Similar to ribonucleoprotein H3aCellular component
Xl2.19528.1.S1_at5.06Transmembrane protein 192Transmembrane
Xl2.18680.1.S1_at5.02Xl.6394Kinase
Xl2.40141.1.S1_at4.98Xl.40141 
Xl2.39370.2.S1_at4.88Xl.83466Unknown
Xl2.14311.1.A1_at4.76Xl.14311Unknown
Xl2.7766.1.A1_at4.65Similar to ubiquitin carboxyl-terminal hydrolase 54-likeAssociated with metabolism
Xl2.48217.1.S1_at4.54Heat shock protein 105Other cellular process
Xl2.53477.3.A1_at4.51Xl.62682Unknown
Xl2.30815.1.S1_at4.45smoothelinCellular component
Xl2.25691.1.A1_at4.37Xl.25691Unknown
Xl2.52642.1.A1_at4.37Xl.52642Unknown
Xl2.16570.1.S2_at4.33Phosphoribosyl pyrophosphate synthetase 1Associated with metabolism
Xl2.19746.1.S1_at4.31Solute carrier family 48 (heme transporter), member 1Transporter
Xl2.6819.1.S1_at4.23PTK2 protein tyrosine kinase 2Kinase
Xl2.47355.1.S1_at4.17CTP synthase IIAssociated with metabolism
Xl2.34538.1.A1_at4.16Sterol regulatory element binding transcription factor 2 (srebf2)Factor, transcription
Xl2.43012.1.A1_at4.11Xl.54800Unknown
Xl2.21754.1.A1_at4.06Tumor necrosis factor, alpha-induced protein 1 (endothelial)Signal Transduction
Xl2.19657.1.S1_at4.05Xl.19657Unknown
Xl2.29665.1.S1_at4.02Xl.29665Unknown
Figure 3.

qRT-PCR comparisons of AN and VP samples. Of the 11 genes identified by microarray to be animal-enriched for which successful primers were designed, 8 were detected at significantly higher levels (*P < 0.05) in AN (blue) versus VP (red) samples. Numbers on Y-axis indicate fold differences relative to vegetal value (mean values from a minimum of three independent samples replicated twice). Bars represent standard error of the mean. As a control, a known VP mRNA, Vg1, also was tested, and as expected from the literature is highly enriched in the VP samples.

Figure 4.

Location of maternal transcripts of the 14 characterized genes assayed by ISH at the 8-cell-stage cleavage embryo. All 14 are enriched in the animal blastomeres, whose equatorial cleavage furrow is indicated by an arrow. Animal pole is up and vegetal pole is down. For some embryos, ISH signal is also detected in the equatorial region (see Table 2).

Table 2. Developmental Expression Patterns of Animal-Enriched mRNAs Determined by ISHa
Gene namesMaternalBlastula st 8–9Gastrula st 10–11.5Neural Plate E: st 12–13 L: st 14–18Closed Neural Tube/ Early Tail Bud st 23–25Late Tail bud st 32Larvae st 35–38 Same as tail bud plus:
  1. a

    AC, Animal Cap; An, Animal; Ant, anterior; BA, branchial arches; CG, cement gland; CrGa, cranial ganglia; E, early; ecto, ectoderm; Egg, unfertilized egg; epi, epidermis; Eq, equatorial; L, late; meso, mesoderm; ms, muscle; NC, neural crest; ND, not determined; NP, neural plate; NT, neural tube; Olf, olfactory placode; Oto, otocyst; OV, optic vesicle; phar, anterior pharynx; PPE, pre-placodal ectoderm; Ret, retina; SpCd, Spinal cord; st, stage; 8CS, 8-cell stage.

Xl.86248Egg: An 8CS: AnAC ectoecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, epi, paraxial mesoNT, OV, Olf, Oto, BA, epi, somitesBrain, Ret, Olf, Lens, Oto, BA, somites, hypaxial ms, nephric meso, liver, ventral gutCrGa, SpCd, trunk NC, heart
WBP2NLEgg: An 8CS: AnAC ecto, Eqecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, paraxial mesoNT, OV, Olf, Oto, BA, somitesBrain, Ret, SpCd, Olf, Lens, Oto, BA, somites, hypaxial ms, nephric meso, liverCrGa, trunk NC, heart, phar
aurka-bEgg: An 8CS: An/EqAC ecto, Eqecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, CG, epi, paraxial mesoNT, Oto, BA, CG, somitesBrain, SpCd, Oto, BA, CG, somitesRet, Olf, Lens, CrGa, trunk NC, nephric meso, phar
Xl.16112Egg: An 8CS: An/EqAC ectoectoE: neural ecto, epi L: NP, NC, PPE, epiNT, OV, Olf, BA, epiBrain, Ret, Olf, Oto, BA, epi, somitesCrGa, SpCd, heart, phar
SMCR7LEgg: An 8CS: AnAC ecto, EqectoE: neural ecto, epi L: NP, NC, PPE, epi, paraxial mesoNT, OV, Olf, Lens, Oto, BA, nephric mesoBrain, Ret, Olf, Oto, BA, loss of lens, somites, nephric meso, ventral gutCrGa, SpCd, trunk NC, heart, blood islands
Xl.11373Egg: An 8CS: AnAC ectoectoE: neural ecto, epi L: NP, NC, PPE, epiNT, OV, Olf, Oto, BA, CG, somitesBrain, Ret, SpCd, Oto, BA, somites, nephric mesoLens, trunk NC, heart, phar
Xl.14311Egg: An 8CS: AnAC ectoecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, epiNT, OV, Oto, BA, epiBrain, Ret, Lens, Oto, BA, nephric mesoCrGa, SpCd, trunk NC, somites, hypaxial ms, heart, phar
Xl.52642Egg: An 8CS: AnAC ectoecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, epiNT, OV, Olf, Oto, BA, CG, epi, somitesBrain, Ret, SpCd, Olf, Lens, Oto, BA, epi, somites, nephric meso, heartnoto, phar
srebf2Egg: An 8CS: An/EqAC ectoecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, epi, paraxial mesoNT, OV, Olf, Oto, BA, somitesBrain, Ret, SpCd, Olf, Lens, Oto, BA, somites, nephric mesoCrGa, noto
Xl.54800Egg: An 8CS: An/EqAC ectoecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, epi, paraxial mesoNT, OV, Lens, BA, epi, somitesBrain, Ret, Lens, Oto, BA, epi, somites, nephric meso, liverdorsal SpCd, trunk NC, heart, phar
Xl.15460Egg: ND 8CS: An/EqAC ectoecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, epi, paraxial mesoNT, OV, Lens, BABrain, Ret, SpCd, Oto, BA, livertrunk NC, nephric meso, phar
Xl.29665Egg: An 8CS: An/EqAC ecto, Eqecto, involuting mesoE: neural ecto, epi L: NP, NC, PPE, epi, paraxial mesoNT, OV, Lens, Oto, BA, epi, somitesBrain, Ret, SpCd, Olf, Lens, Oto, BA, epi, somites, nephric mesoheart, liver, phar
TNFAIP1Egg: ND 8CS: AnAC ectoectoE: neural ecto, epi L: NP, NC, PPE, epiNT, Oto, BA, CG, somitesBrain, Ret, SpCd, Olf, Lens, Oto, BA, somites, nephric mesoCrGa, trunk NC, heart, phar
gskipEgg: An 8CS: An/EqAC ectoectoE: neural ecto, epi L: NP, NC, PPE, epiNT, OV, Oto, BA, epi, CGBrain, Ret, SpCd, Lens, Oto, BA, somites, nephric mesoCrGa, trunk NC, heart, liver, phar

Identification and Expression Patterns of Novel Animal-Enriched Transcripts

Many of the genes identified as animal-enriched have not been annotated and/or fully sequenced, and therefore are of unknown function (Fig. 2; Table 1). To predict gene functions, we fully sequenced the clones of all 14 genes of unknown function, performed BLAST searches, and compared sequences to the Xenopus tropicalis (genome build 4.2) and Xenopus laevis (genome build 6.11) genomes. We were able to identify six of the fourteen gene products: WBP2NL, Aurka-b, SMCR7L, Sreb2, TNFAlP1, and Gskip. The remaining eight genes (Xl.86248, Xl.16112, Xl.11373, Xl.14311, Xl.52642, Xl.54800, Xl.15460, Xl.29665) are currently only known as expressed sequences by virtue of their presence in cDNA libraries.

As a first step in characterizing these 14 genes of unknown function, we determined their developmental expression patterns, which have not been previously reported. First, to demonstrate expression levels over developmental time, qRT-PCR was performed on whole embryos for the 11 genes for which we could design primers. We observed 4 general patterns of expression (Fig. 5). Two mRNAs (aurka-b, Xl.54800) are expressed at their highest levels maternally, with significantly decreased levels by blastula stages, and expression remains relatively low through late larval stages (Fig. 5A). Four mRNAs (srebf2, WBP2NL, SMCR7L, Xl.86248) are enriched maternally with reduced expression levels during blastula and gastrula stages, followed by increased expression after neurulation that remains relatively high through late larval stages (Fig. 5B). Two mRNAs (Xl.15460, Xl.29665) are expressed at their lowest levels maternally, followed by increased expression after gastrulation or neurulation, which remains relatively high through late larval stages (Fig. 5C). Three mRNAs (Xl.11373, TNFAIP1, gskip) have strong maternal expression that is gradually reduced during gastrula to neural plate stages, followed by increased expression at larval stages (Fig. 5D). These patterns suggest that all 14 genes have early developmental roles, either before or during the maternal-to-zygotic transition at blastula stages. Furthermore, because these genes continue to be expressed throughout later development, they may play important roles in early tissue specification, differentiation, and organogenesis. However, because the dynamic patterns seen in whole embryo qRT-PCR analyses likely reflect changes in tissue-specific patterns of expression, spatial expression patterns were analyzed by ISH.

Figure 5.

Developmental time course of expression in whole embryos as detected by qRT-PCR. A: Two genes (aurka-b, Xl.54800) are most highly expressed during the maternal phase of development (st 4), gradually decrease during the onset of zygotic transcription (st 8–10), and remain at relatively low levels during the remaining stages analyzed. B: Four genes (srebf2, WBP2NL, SMCR7L, Xl.86248) are highly expressed during the maternal phase of development and gradually decrease during the onset of zygotic transcription, but rebound to higher expression levels at neurulation (st 19) through larval (st 35–38) stages. C: Two genes (Xl.15460, Xl.29665) are expressed at the lowest level maternally and at relatively high levels after neural plate stages. D: Three genes (Xl.11373, TNFAIP1, gskip) are strongly expressed maternally, with decreasing expression at early zygotic stages and strong larval expression. Y-axis indicates relative expression compared to lowest expression for that transcript (mean values from a minimum of three independent samples replicated twice). X-axis indicates developmental stage. Bars represent standard error of the mean.

For all 14 genes, maternal transcripts were animal enriched in unfertilized eggs (Table 2) and in 8-cell embryos (Fig. 4), indicating that the cytoplasmic movements that occur at fertilization do not significantly reorganize them. Therefore, it is likely that these mRNAs also are animal enriched in late stage oocytes; this conclusion is corroborated by the observation that mRNAs that are animal enriched in oocytes continue to be animal enriched after in vitro fertilization (Sindelka et al., 2008). This animal-enriched location of transcripts was maintained in the animal cap ectoderm of the blastula (Fig. 6); four were additionally expressed in equatorial regions, coinciding with their later expression in paraxial mesoderm. All 14 genes were expressed diffusely throughout the ectoderm at gastrula stages, and five were weakly expressed in the non-involuted mesoderm (Table 2). At neural plate stages, all 14 genes were expressed throughout the dorsal ectoderm with enrichment in the neural plate, neural crest, and pre-placodal ectoderm (PPE) (Fig. 7A, B; Table 2). These early, broadly ectodermal distributions of transcripts, which are presumed maternal, suggest an early role in ectodermal germ layer formation, but this needs to be tested experimentally by knock-down of maternal mRNAs and/or proteins in the oocyte. However, after the onset of zygotic transcription as the embryo begins to neurulate and differentiate, the expression patterns of these 14 genes diverge and become distinct in several organs, as described below and detailed in Table 2.

Figure 6.

Expression of the 14 characterized genes assayed by ISH in stage 8–9 blastula embryos. All 14 genes are strongly expressed in the animal cap ectoderm (top), but for some ISH signal is also detected in the equatorial region (*, also see Table 2). Animal pole is up and vegetal pole is down.

Figure 7.

Expression patterns of animal-enriched genes from neural plate to late tail bud stages. A: Expression patterns of the seven highest AN-enriched genes that were characterized. B: Expression patterns of the remaining seven AN-enriched genes that were characterized. Early (left) and late (third from right) neural plate stages are anterior views with dorsal to the top. Early (second from right) and late (far right) tail bud stages are side views with anterior to the right and dorsal to the top. ba, branchial arch; br, brain; cg, cement gland; epi, epidermis; h, heart; le, lens; li, liver; nc, neural crest; ne, nephric mesoderm; np, neural plate; olf, olfactory, oto, otocyst; ov, optic vesicle; pm, paraxial mesoderm; ppe, pre-placodal ectoderm; ret; retina; sc, spinal cord; so, somites; vg, ventral gut.

Transcribed locus Xl.86248. At late neural plate stages, expression is enriched in the neural plate, neural crest, and PPE, but also is expressed in the paraxial mesoderm (Fig. 7A). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, with declining expression in the epidermis, and begins to be expressed in the posterior somites. By stage 32, it is also expressed in hypaxial muscle precursors, nephric mesoderm, liver, and the ventral gut. By stage 38, there is additional expression in neural derivatives and heart (Fig. 8A). These spatial patterns suggest that the increasing expression levels at later stages detected by qRT-PCR (Fig. 5A) reflect the gradual broadening of Xl.86248 expression domains over developmental time.

Figure 8.

Expression patterns of animal-enriched genes at late larval stages. A: Expression patterns of the five highest AN-enriched genes that were characterized. B: Expression patterns of the five next highest AN-enriched genes that were characterized. C: Expression patterns of the remaining four AN-enriched genes that were characterized. Left: Side views of embryos with anterior to the right and dorsal to the top. Right: Transverse vibratome sections of whole mount ISH preparations showing internal histology. ba, branchial arch; br, brain; cg, cement gland; epi, epidermis; h, heart; le, lens; li, liver; nc, neural crest; ne, nephric mesoderm; np, neural plate; olf, olfactory, oto, otocyst; ov, optic vesicle; pm, paraxial mesoderm; ppe, pre-placodal ectoderm; ret; retina; sc, spinal cord; so, somites; vg, ventral gut.

WBP2NL. WBP2NL, also known as PostAcrosomal sheath WW domain-binding Protein (PAWP), promotes meiotic resumption and pronuclear formation in multiple vertebrate species and has been shown in Xenopus to do this through inducing intracellular calcium release (Wu et al., 2007). At late neural plate stages, WBP2NL expression is enriched in the neural plate, neural crest, PPE and paraxial mesoderm (Fig. 7A). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, with declining expression in the epidermis, and begins to be expressed in the posterior somites. By stage 32, it is also expressed in the hypaxial muscle precursors, nephric mesoderm, and liver. By stage 38, there is additional expression in neural derivatives, heart, and anterior pharynx (Fig. 8A). These spatial patterns suggest that the increasing expression levels at later stages detected by qRT-PCR (Fig. 5B) reflect the gradual broadening of WBP2NL expression domains over developmental time.

Aurka-b. Aurka-b, previously known as P46XlEg22, is a serine/threonine kinase related to the Drosophila gene aurora, and is predicted to be involved in controlling chromosome segregation during meiosis (Yanai et al., 1997). At late neural plate stages, aurka-b expression is enriched in the neural plate, neural crest, PPE, cement gland, and paraxial mesoderm (Fig. 7A). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, with declining expression in the epidermis. It is strongly expressed in the cement gland. By stage 32, the strongest expression is in the brain, retina, spinal cord, and otocyst, with weak expression in the branchial arches and somites. By stage 38, there is additional expression in the neural derivatives, nephric mesoderm and anterior pharynx (Fig. 8A). These spatial patterns suggest that the lower levels of expression at later stages detected by qRT-PCR (Fig. 5A) reflect the gradual restriction of aurka-b expression to smaller tissue domains from its initial broad pattern.

Transcribed locus Xl.16112. At late neural plate stages, this gene is expressed in the neural plate, neural crest, PPE, and epidermis (Fig. 7A). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, with strong expression in the epidermis. By stage 32, there is additional expression in the somites. By stage 38, there is additional expression in neural derivatives, heart, and anterior pharynx (Fig. 8A).

SMCR7L. SMCR7L, also known as MItochondrial Elongation Factor 1 (MIEF1), has been shown to be localized to the mitochondrial outer membrane and involved in controlling the dynamics of mitochondrial fission and fusion (Zhao et al., 2011, Palmer et al., 2011). At late neural plate stages, SMCR7L expression is enhanced in the neural plate, neural crest, PPE, and paraxial mesoderm (Fig. 7A). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes and begins to be expressed in nephric mesoderm. By stage 32, there is additional expression in the somites and ventral gut, and a loss of lens staining. By stage 38, there is additional expression in neural derivatives, heart, and blood islands (Fig. 8A). These spatial patterns suggest that the increasing expression levels at later stages detected by qRT-PCR (Fig. 5B) reflect the gradual broadening of SMCR7L expression domains over developmental time.

Transcribed locus Xl.11373. At late neural plate stages, this gene is expressed in the neural plate, neural crest, PPE, and epidermis (Fig. 7A). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, and begins to be expressed in cement gland and somites. By stage 32, there is additional expression in nephric mesoderm. By stage 38, there is additional expression in lens, heart, and anterior pharynx (Fig. 8B). These spatial patterns are consistent with the dynamic expression levels detected by qRT-PCR (Fig. 5D).

Transcribed locus Xl.14311. At late neural plate stages, this gene is expressed in the neural plate, neural crest, PPE, and epidermis (Fig. 7A). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes and in the epidermis. By stage 32, there is additional expression in the nephric mesoderm. By stage 38, there is additional expression in neural and placode derivatives, dorsal somites, heart, and anterior pharynx (Fig. 8B).

Transcribed locus Xl.52642. At late neural plate stages, this gene is expressed in the neural plate, neural crest, PPE, and epidermis (Fig. 7B). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, with strong expression in the epidermis and cement gland. It also begins to be expressed in the somites. By stage 32, there is additional expression in the nephric mesoderm and heart. By stage 38, there is additional expression in notochord and anterior pharynx (Fig. 8B).

Srebf2. Srebf2 is an activating transcription factor containing a basic-Helix-Loop-Helix-Leucine-Zipper DNA binding domain that binds to sterol-regulatory elements (Hua et al., 1993). At late neural plate stages, srebf2 is expressed in the neural plate, neural crest, PPE, epidermis, and paraxial mesoderm (Fig. 7B). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, and it begins to be expressed in the somites. By stage 32, there is additional expression in nephric mesoderm. By stage 38, there is additional expression in neural and placode derivatives and notochord (Fig. 8B). These spatial patterns suggest that the increasing expression levels at later stages detected by qRT-PCR (Fig. 5B) reflect the gradual broadening of srebf2 expression domains over developmental time.

Transcribed locus Xl.54800. At late neural plate stages, this gene is expressed in the neural plate, neural crest, strongly in the PPE, and in the epidermis and paraxial mesoderm (Fig. 7B). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, in the epidermis, and begins to be expressed in the somites. By stage 32, it is additionally expressed in the nephric mesoderm and liver. By stage 38, it is additionally expressed in neural derivatives, heart, and anterior pharynx (Fig. 8B). These spatial patterns suggest that the lower levels of expression at later stages detected by qRT-PCR (Fig. 5A) reflect the gradual restriction of Xl.54800 expression to smaller tissue domains from its initial broad pattern.

Transcribed locus Xl.15460. At late neural plate stages, this gene is expressed in the neural plate, neural crest, PPE, epidermis, and paraxial mesoderm (Fig. 7B). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes. By stage 32, there is additional expression in the liver. By stage 38, there is additional expression in neural crest derivatives and nephric mesoderm (Fig. 8C). These spatial patterns suggest that the high levels of expression at later stages as detected by qRT-PCR (Fig. 5C) reflect the broadening of Xl.15460 expression domains over developmental time.

Transcribed locus Xl.29665. At late neural plate stages, this gene is expressed in the neural plate, neural crest, PPE, epidermis, and paraxial mesoderm (Fig. 7B). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, in the epidermis, and begins to be expressed in the somites. By stage 32, there is additional expression in nephric mesoderm. By stage 38, there is additional expression in heart, liver, and anterior pharynx (Fig. 8C). These spatial patterns suggest that the high levels of expression at later stages as detected by qRT-PCR (Fig. 5C) reflect the broadening of Xl.29665 expression domains over developmental time.

TNFAIP1. TNFAIP1, also known as BTB/POZ domain-containing adapter for CUL3-mediated RhoA degradation protein 2, appears to be involved in regulation of inflammation, apoptosis, and development, likely through its effects on the actin cytoskeleton (Dixit et al., 1990; Chen et al., 2009). At late neural plate stages, TNFAIP1 is expressed in the neural plate, neural crest, PPE, and epidermis (Fig. 7B). At closed neural tube stages, it continues to be expressed in derivatives of the neural tube, neural crest, and placodes, and begins to be expressed in the cement gland and somites. By stage 32, there is additional expression in the nephric mesoderm. By stage 38, there is additional expression in neural and placode derivatives, heart and anterior pharynx (Fig. 8C). These spatial patterns are consistent with the dynamic expression levels detected by qRT-PCR (Fig. 5D).

Gskip. Gskip contains a GSK3-beta interaction domain and an RII-binding domain that binds to protein kinase-A; it likely mediates the phosphorylation of GSK3-beta and may integrate PKA and GSK3-beta signaling pathways (Chou et al., 2006; Lin et al., 2009; Hundsrucker et al., 2010). At late neural plate stages, gskip is expressed in the neural plate, neural crest, PPE, and epidermis (Fig. 7B). At closed neural tube stages, it continues to be expressed in ectodermal and neural derivatives. By stage 32, there is additional expression in somites and nephric mesoderm. By stage 38, there is additional expression in the neural and placode derivatives, heart and liver (Fig. 8C). These spatial patterns are consistent with the dynamic expression levels detected by qRT-PCR (Fig. 5D).

In Summary

All 14 of these maternally deposited transcripts initially are enriched in the animal hemisphere of Xenopus laevis unfertilized eggs and animal blastomeres of cleavage stage embryos. They remain enriched in the blastula animal cap ectoderm and gastrula ectoderm. Broad and diffuse ectodermal enrichment, particularly in the neural ectoderm, neural crest, and PPE, is a common feature at neural plate stages. However, as zygotic transcription is initiated and the neural tube closes, expression patterns and levels of expression become more discrete and begin to diverge. These data demonstrate that although the animal-enriched transcripts are preferentially and broadly expressed in ectodermally-derived tissues at early stages, most of the transcripts also are expressed in limited mesodermal and endodermal tissues at tail bud to larval stages. Thus, these transcripts are not strictly restricted to derivatives of the ectodermal germ layer. Nonetheless, all 14 animal-enriched genes share a common feature: they are all expressed diffusely and broadly in early ectoderm, both neural and non-neural, and zygotically continue to be expressed in neural tissues throughout early development. It will now be important to determine their specific developmental roles, both maternally and zygotically.

EXPERIMENTAL PROCEDURES

Embryos

Fertilized Xenopus laevis eggs were obtained by either in vitro fertilization or gonadotropin-induced natural mating of adult frogs as described elsewhere (Moody, 2000). For microarray and qRT-PCR assays, embryos were collected at the 8-cell stage, fixed in 98% ethanol, 2% acetic acid, and stored at −20°C (Sive et al., 2000). The four animal blastomeres (AN) were manually dissected from the vegetal blastomeres (Fig. 1). The vegetal blastomeres were divided approximately in half across the animal-vegetal axis, and the vegetal pole half (VP) was collected (Fig. 1). Dissected pieces from 30–40 embryos were pooled as a single sample and snap frozen. For qRT-PCR and whole-mount ISH analyses, embryos were cultured in 10% Steinberg's solution and collected from cleavage through late larval stages, as detailed below.

RT-PCR

To ensure that the AN and VP samples were not cross-contaminated, prior to microarray analyses RT-PCR was performed on each pooled sample with primers for maternal mRNAs known to be localized to either the animal (foxD4L1.1, aka foxD5; Sullivan et al., 2001) or vegetal pole (Vg1; Melton, 1987) regions. RNA was extracted with the RNeasy mini kit (Qiagen, Chatsworth, CA). Semi-quantitative RT-PCR within linear ranges was performed as previously described (Yan et al., 2009). Primers are listed in Table 3; some primer sequences were obtained from published papers, whereas the others were designed with the Primer3 software (Rozen and Skaletsky, 2000). All RT-PCR assays were repeated at least 3 times. Bands were visualized with a Storm 860 phosphorimager (Molecular Dynamics, Sunnyvale, CA), the band intensities were measured with the ImageQuant analysis program and compared to an internal control (H4).

Table 3. Primer Sequencesa
  1. a

    All primers are listed from 5′ to 3′.

qRT-PCR primers 
qXl.29665-FAAAGCTTCCCATTCTGTTCCA
qXl.29665-RAAGTGGAATTCTTGGCAATGG
qXl.34538-FTACTGCTATGGCAGCCTCCT
qXl.34538-RCCCATTGAGCCATGAAGTCT
qXl.52642-FGGAAATTTGGGGAGATTGGT
qXl.52642-RCTTCGGAAAACGAAGCACTC
qXl.43012-FCGTACCACTGGGGGTTAGAA
qXl.43012-RCCTCAACAGAGGAGCAGAGG
qXl.4421-FTTTCTAGTGGAGGCGGCTAA
qXl.4421-RAAAAGGCACCAGGTTTTGTG
RT-PCR primers 
Xl.16112-FAGGCTGATGCCAGGTCACTCCA
Xl.16112-RTCCAAGGTTCTCCATTGGCTCCGA
Xl.14311-FAGAGGGTATGCAGGGGCGGA
Xl.14311-RCGAAGCCAACGGAGAACCTTGGA
Xl.11373-FTGGCCTGAGACATCCCCCACA
Xl.11373-RTGGTGCTCAGTGATGTCATGTGTC
Xl.29665-FTGAGGCAACATCGCATTGCCG
Xl.29665-RTCTTCCCCTGGTCAAGGCCG
Xl.34538-FGCCCTTGCTTCTGCAGGCTCT
Xl.34538-RAGCCTCTACACCAACTGCACCA
Xl-52642-FGGCTCGCGGGTCTGGGTGTA
Xl.52642-RTCAGCGGGTCTGGGTGTCCA
Xl.43012-FAGGGTCCAGGCAAGGGATGGA
Xl.43012-RCCTCTCGTGCGAGCGAAGGC
Xl.21754-FAAGGGCAGACACACACGGCG
Xl.21754-RGGTGGTGGAGCCTGTATGAAAGGA
Xl.4421-FCCCAAACCACAGGCTGCCCC
Xl.4421-RTGCCGGGTCTGGGGAAAGGT
SMCR7L-FGTCGGTTGGGCTTGGGAGGC
SMCR7L-RCCGACACCCAACCGCTGCAT
Xl.14775-FTGGGCAGACCTGCCGAGGTT
Xl.14775-RTCCGAATCGTACAATTTTGCCCCC
WBP2NL-FGGTGGCTGGGAGGGTCAAGC
WBP2NL-RGCCTGATGGTTGGGGTGGGG

Microarray and Statistical Analyses

Three samples of AN and three samples of VP were collected from independent matings of different parents. AN and VP samples from the same mating, i.e., containing sibling embryos, were compared in the microarray analysis. Furthermore, all samples were processed in parallel for cDNA labeling and GeneChip hybridization to reduce inter-sample variations. RNA integrity was assessed using the Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA) and only samples with an integrity number > 7.0 were used. Total RNAs were labeled and fragmented with the Ovation Biotin RNA Amplification and Labeling System (NuGEN Technologies, Inc., San Carlos, CA). Briefly, 50 ng total RNA were used for first- and second-strand cDNA syntheses. The synthesized cDNAs were amplified, purified with a PCR purification kit (Qiagen), then labeled and fragmented following the Ovation kit instructions. The labeled cDNAs were purified with the Dye Ex kit (Qiagen), and shipped to the NINDS-NIMH Microarray Consortium at The Translational Genomics Research Institute (T-GEN, Tempe, AZ).

The Affymetrix GeneChip® Xenopus laevis Genome Array (v2.0) contains more than 32,400 probe sets representing approximately 29,900 transcripts. The Affymetrix Chip hybridization and initial statistical analyses were performed at T-GEN. The chips were washed and scanned as recommended by the Ovation Biotin RNA Amplification and Labeling System User Guide (version 1.0). GCOS software was used to determine signal intensities and detection calls for each gene. Using GeneSpring software the replicate correlation values were determined to be >0.99 for the three AN samples and >0.99 for the three VP samples, indicating consistency between replicates despite random genetic variance between parental frogs. Genes were filtered for at least 1 present call out of 6 calls. For those remaining mRNAs expressed >2-fold higher in AN than VP samples, statistical significance was tested using a paired t-test (P < 0.05). The raw data are presented in Supplemental Tables S1 and S2, and available for download from GEO (GSE48659).

qRT-PCR

qRT-PCR analyses were performed on total RNA extracted from unfertilized eggs, and embryos collected at stages 4–38. First-strand cDNA synthesis was performed as above. Gene expression levels were assayed by EvaGreen (Bio-Rad, Hercules, CA) incorporation detected using the ABI Prism 384XL Detection System (Applied Biosystems, Foster City, CA) using the SsoFast EvaGreen Supermix for amplification and fluorescence (Bio-Rad) in a two-step cycle with an annealing/extension temperature of 60°C for 30 s and denaturizing step at 95°C for 15 s. Real-time PCR results were analyzed using the comparative Ct, 2-ΔΔCt, method (Livak and Schmittgen, 2001; Schmittgen and Livak, 2008). Relative expression levels were normalized to ornithine decarboxylase (ODC) levels. All primer sequences are listed in Table 3.

Sequencing and Whole Mount In Situ Hybridization

Plasmids encoding for a subset of the mRNAs enriched >4-fold in animal blastomeres in the microarray assay were purchased from Open Biosystems (Thermo Scientific, Huntsville, AL). These were fully sequenced by standard techniques in both directions to assist in identification of the encoding genes. Anti-sense digoxigenin-labeled RNA probes were synthesized by in vitro transcription (Ambion, Austin, TX; Megascript kit). Albino or wild type unfertilized eggs and embryos were fixed in MEMFA and processed for whole mount ISH according to standard protocols (Sive et al., 2000). To visualize internal expression, embryos already processed for whole-mount ISH were embedded in 4% agarose and sectioned at 70 μm with a vibratome. Sections were mounted on glass slides, coverslipped, and photographed.

ACKNOWLEDGMENTS

We thank Mr. Newt Moore for photography and sectioning, and Himani Datta Majumdar and Karen Neilson for assistance in sequencing. This work was funded in part by the Cosmos Club Foundation Scholars Award, and the Dilthey Award and CCFF Award from George Washington University.

Ancillary