Transcription in pronuclei and one- to four-cell embryos drives early development in a nematode

Summary

Background

A long-standing view of development is that transcription is silenced in the oocyte until early divisions in the embryo. The point at which major transcription is reactivated varies between organisms, but is usually after the 2-cell stage. However, this model may not be universal.

Results

We used RNA-seq and exploited the protracted development of the parasitic nematode Ascaris suum, to provide a comprehensive time course of mRNA expression, degradation, and translation during early development. Surprisingly, we find that ~4,000 genes are transcribed prior to pronuclear fusion and in the 1–4 cell embryos. Intriguingly, we do not detect maternal contribution of many orthologs of maternal C. elegans mRNAs, but instead find these are newly transcribed in the A. suum zygote prior to pronuclear fusion. Ribosome profiling demonstrates that, in general, early embryonic mRNAs are not stored for subsequent translation, but are directly translated following their synthesis. The role of maternally contributed and zygotically transcribed genes differs between the nematodes A. suum and C. elegans despite the fact that the two nematodes appear to exhibit highly similar morphological patterns during early development.

Conclusions

Our study indicates that major transcription can occur immediately after fertilization and prior to pronuclear fusion in metazoa, suggesting that newly transcribed genes appear to drive A. suum early development. Furthermore, the mechanisms used for controlling the timing of the expression of key conserved genes has been altered between the two nematodes, illustrating significant plasticity in the regulatory networks that play important roles in developmental outcomes in nematodes.

Introduction

The transition from the oocyte to embryo development involves transcriptional silencing followed by reactivation of transcription during early cell division, known as zygotic genome, or gene activation. During this period of transcriptional silence, regulation of gene expression is considered primarily post-transcriptional, relying on the deposition of maternal mRNAs and proteins from the oocyte [1, 2]. The timing of transcriptional reactivation varies between organisms and has been called the maternal to zygotic transition, a period where maternal mRNAs are degraded and development is dependent on transcription from the zygote. While this model of early development is generally accepted, it remains to be determined if transcription is silenced during this period in all organisms.

The maternal to zygotic transition has been characterized in many animals including mice, frogs, fish, fruit flies, sea urchins, and nematodes [1, 2]. Zygotic genome activation is often associated first with a minor wave of transcription, followed by a major wave of transcription. In vertebrates, the first wave of transcription begins as early as the 1-cell stage in mice, whereas it occurs at the ~246 cell stage in frogs. The second major wave of transcription in mice and frogs occurs at the 2-cell and ~4,000 cell stage, respectively. In invertebrates, the first wave of transcription occurs as early as the 1-cell stage in sea urchins and as late as the 64-cell stage in fruit flies. Transcription has been reported to occur in the male pronucleus in zygotes of both sea urchins and mice, but this appears to be limited to only one or a few genes [1, 2]. Thus, in general, major transcription in animals does not occur prior to fusion of male and female pronuclei in the zygote.

The parasitic worm Ascaris is a nematode particularly amenable for the analysis of early development. Studies on Ascaris have contributed to our understanding of key processes in biology, including the first determinate lineage, the centrosome cycle, meiosis, mitosis, and the continuity of chromosomes [3–7]. Older studies suggested that some transcription might occur in Ascaris suum zygotes prior to pronuclear fusion during their maturation in the uterus [8, 9]. In support of this, we recently observed that A. suum microRNAs are transcribed during the maturation of the zygote as it passes through the uterus [10].

The prevailing view in metazoa is that significant RNA polymerase II transcription is absent in zygotes and does not occur until after cleavage initiates [2, 11]. Here, we exploited the protracted and synchronous early development of A. suum to obtain samples that include the extended maturation of the zygote prior to pronuclear fusion and the first embryonic divisions. We used RNA-seq to examine mRNA dynamics during these stages to define and characterize maternal mRNAs, mRNA transcription in zygotes prior to pronuclear fusion and in early embryos, and the maternal to zygotic transition. Our study provides a comprehensive analysis of mRNA dynamics during very early stages of development. In addition, it provides a unique molecular and comparative perspective of development between two nematodes, A. suum and C. elegans, that exhibit striking similarity in cell lineage and early development, yet diverged ~400 million years ago [12, 13]. Surprisingly, we find that extensive A. suum mRNA transcription occurs during zygote maturation prior to pronuclear fusion and in the 1–4 cell embryo. Thus, major transcription can start at a stage much earlier than previously thought in metazoa, prior to pronuclear fusion, and before what has historically been considered the time of reactivation of nematode transcription [14–17]. Our data suggest that ongoing transcription, rather than maternal mRNA deposition, drives A. suum early development. These data open the question as to whether significant early transcription also occurs in other organisms during these early periods of development. Our data also indicate that the expression of conserved and key groups of genes required for early development in related organisms can be rewired to be regulated either by transcriptional or posttranscriptional mechanisms.

Results

Ascaris suum early development

The unique biology of A. suum allows us to collect significant amounts of oocytes, zygotes, and discrete stages of early embryos. In particular, the protracted development enables sampling of mRNA levels during zygote maturation (prior to pronuclear fusion) and very early development (Fig. 1A). A comparison of RNAs present in mature oocytes ready for fertilization and zygotes as they mature in the uterus enabled us to distinguish maternally contributed mRNAs, characterize their degradation, and identify newly transcribed RNAs in zygotes. As the oocytes (arrested in prophase I of meiosis) pass through the seminal receptacle (spermatheca in C. elegans) and enter the uterus, they are fertilized (Fig. 1A). Following fertilization, the eggs become impermeable due to the formation of a complex inner layer and an outer chitinous shell, and then undergo meiosis I and II as in C. elegans. The zygote, without pronuclear fusion or cell division, moves through 7–10 inches of the uterus over a 12–24 hr period (Fig. 1A). During this time, the zygote matures, and this maturation period is required for subsequent development [18]. Importantly, the two pronuclei remain separated during zygote maturation in the uterus. The long uterus allows for the isolation of zygotes from different regions representing different stages of zygote maturation (zygote1–4, see Fig. 1A). Zygote4 eggs (0 hr, eggs released from the uterus) will not develop until the appropriate environmental conditions are encountered upon passing from the pig host. These environmental conditions can be reproduced in the lab allowing for embryo development and subsequent collection of synchronously staged embryos (see Fig. 1A). After 24 hr of development at 30°C, the two pronuclei fuse and the first division occurs at around 36 hr [19]. Subsequent cell divisions occur every ~13 hr on average and development of the L1 and L2 occurs over an extended period of 10 and 21 days, respectively.

Figure 1. mRNA profiles during A. suum early development.

A. Female reproductive system and timeline for A. suum zygote, embryo, and larval development. Samples used for RNA-seq are illustrated. Zygote maturation (red) occurs in the female uterus as the zygotes move toward the vulva. This process takes ~18 h (12 – 24 h) to complete and is required for subsequent development. During this process, the male and female pronuclei remain distinct. Early embryonic development (blue) initiates only after the zygote is released from the worm and pig host into a favorable environment. Incubation of zygote4 (0 hr) embryos at 30°C in high humidity leads to relatively synchronous development (~85%) as illustrated. Early embryonic development includes chromatin diminution occurring between 46 and 96 hr of development (4–16 cells). A. suum L2 development (green) occurs within the egg.

B. Gene expression profiles during A. suum development. Using RNA-seq data from 13 developmental stages, we classified 7,626 expressed genes into four super-groups (A-D) with 12 group gene expression patterns represented by lines of different color (see Fig. S2 and Table S2). The overall expression profiles for these 12 groups of genes are illustrated as a standard score (Z-score), with similar colors (as in Fig. 1A) indicating the developmental stages where the genes are primarily expressed. The main plots show changes in mRNA levels as a function of developmental time (x-axis, shown in days), while the inserts for super-group A and B show the mRNA changes as a function of developmental stage (x-axis). Each gene group is numbered with the number in brackets indicating how many genes are associated with that group.

C. RT-PCR profiles corroborate the RNA-seq data. RT-PCR was performed for selected transcripts corresponding to different gene expression groups (color coding matches transcripts from different stages: pink for oocyte/maternal, red for 1-cell zygotic, blue for early embryonic and green for late embryonic and larval). A strong correlation between the RNA-seq and RT-PCR data was observed (see Figure S1F and Table S1).

D. Differentially expressed genes at five major developmental transitions in A. suum. Differentially regulated genes were defined as described in Supplemental Experimental Procedure using a cutoff of P-value < 0.01.

See also Figure S1 and S2 and Table S1 and S2

A. suum early developmental mRNA expression patterns and differentially expressed genes

RNA-seq data were generated from mature oocytes ready for fertilization and 12 discrete stages of A. suum development (Fig. 1A). We defined mRNA levels in reads per kilobase of mRNA per million mapped reads (RPKM). Of the 15,399 A. suum known and predicted genes [20], 7,626 (49.5%) genes are expressed at significant levels during these 13 stages (Fig. 1B, see Supplemental Experimental Procedures).

mRNA poly(A)-tail lengths can change in early development and contribute to differential gene expression [21]. To ensure that our analysis of mRNAs using poly(A)-selected RNA was not biased by short poly(A)-tail length or changes in poly(A)-tail length, we compared RNA-seq data from total RNA to polyA-selected RNA. We also carried out qRT-PCR on random primed total RNA (Fig. S1A–C, Table S1). These analyses suggested our mRNA profiles are not biased by the poly(A) selection. We further corroborated the RNA-seq data by comparing the developmental expression profiles of 28 genes using RT-PCR (Fig. 1C, Fig. S1F, and Table S1). The RT-PCR results strongly support the RNA-seq data, with 21 (75%) genes demonstrating PCR expression profiles similar to those observed by RNA-seq. The RNA-seq data from five stages were each compared to a corresponding biological replicate used for polysome analysis (RNA-seq data on polysome fractions; see below). An average correlation of R = 0.923 was observed between these replicates (Fig. S1D). Furthermore, the expression patterns for 81.5% of the genes are consistent across these five stages (Fig. S1E). Although we only have biological replicates for five developmental stages, these five stages all show very high level of consistency, even after a different sample processing procedure (polysome fractionation). Overall, these data illustrate that the RNA-seq expression patterns we obtained are reproducible and unbiased.

The mRNA expression patterns from 7,626 genes were organized into four “super-groups” (A-D) and their groups (1–12) (Fig. 1B, Fig. S2A–C, and Table S2). Super-group A consists of 1,177 genes whose mRNAs are present at the higher levels in oocytes than at any other stage. These maternal mRNAs are rapidly degraded following fertilization. Some of these maternal genes are then transcribed again in zygotes (zygote4) prior to pronuclear fusion, though most are gone before the 4-cell stage (Fig. 1B, insert). Super-group B consists of 1,662 genes that are specifically transcribed and expressed in zygotes prior to pronuclear fusion (zygote1–4) and degraded before the 10-cell stage (see Fig. 1B insert). Super-group C consists of 2,962 genes that are transcribed prior to pronuclear fusion, in 1–4 cell embryos, and in 16–256 cell embryos (Fig. 1B). Super-group D corresponds to 1,825 genes that are transcribed in the late embryo, during larval development, and in the larvae (Fig. 1B). Gene lists for all the groups, their expression levels, and biological functions are available in the Supplemental Information (Fig. S2 and Table S2). Further analysis of these gene expression patterns suggested that there were five major developmental transitions (Fig 1D, Fig S1G, Table S2 and see Supplemental Experimental Procedures) including the oocyte to zygote, zygote to early embryo, early to late embryo, late embryo to L1, and L1 to L2 larval stage transitions. Overall, these data demonstrate dynamic expression patterns of mRNA transcription and degradation during A. suum development.

The mRNA expression data demonstrate that gene transcription starts right after fertilization, prior to pronuclear fusion, and remains active through early embryo cleavage. Concordant with the accumulation of new mRNAs during zygote maturation in the uterus, we observed that a number of components of the pol II transcription machinery and a variety of transcription factor mRNAs were synthesized (Table S3). These include the RNA polymerase II subunits, rpb1 (ama-1 in C. elegans) and rbp9. Overall, the significant number of upregulated genes during this very early time-point of development provides evidence for some of the earliest, robust, and diverse transcription observed in metazoa.

Our high-resolution RNA-seq data of oocyte and early zygote stages enabled us to define maternal mRNA contributions and transcription prior to pronuclear fusion. mRNA present at a higher level in oocytes than in zygote1 were considered to be maternally contributed (super-group A). Significant transcription was observed in the 1-cell zygotes prior to pronuclear fusion (super-group B) and early embryos (super-group C). Thus, in A. suum, many mRNAs that would traditionally be considered maternal in C. elegans (present in zygote4) are newly transcribed immediately after fertilization and prior to pronuclear fusion (group 4–6; ~ 2,500 mRNAs). The transcription observed in A. suum zygotes prior to pronuclear fusion appears to correspond to zygotic gene activation described in other organisms. In addition to transcription in pronuclei, we also observed significant transcription of diverse mRNAs in 1–4 cell embryos, a developmental stage prior to what is traditionally considered the maternal to zygotic transition in nematodes (groups 7–8, ~ 1,500 mRNAs) [2, 14].

Our data also demonstrate several distinct periods of mRNA degradation. One period occurs soon after fertilization and leads to rapid degradation of mRNAs (Fig, 1B, see groups 1–3). Thus, in A. suum, both maternal mRNA turnover and early zygotic transcription begins immediately following fertilization (Fig. 1B). In contrast, this is thought to occur at the ~ 4–32 cell stage in C. elegans [14, 22]. Interestingly, we did not identify any A. suum maternally contributed RNAs equivalent to the C. elegans class I maternal RNAs that remain stable throughout early development [23]. At least four distinct additional groups of mRNAs undergo rapid degradation (Fig, 1B, see groups 4–7). This suggests discrete control in the specificity and timing of degradation of these mRNAs. Overall, >3,000 mRNA are degraded during the 4–64 cell stage in A. suum.

Early transcription occurs in both pronuclei and in germline and somatic cells in A. suum early development

We carried out immunolocalization studies for phosphorylated RNA polymerase II to determine if the transcription we observed by RNA-seq in zygotes prior to pronuclear fusion was derived from one or both A. suum pronuclei. We observed significant nuclear staining with antibodies to the unphosphorylated C-terminal domain (CTD, data not shown) and to the phosphorylated CTD ser-2 and ser-5 epitopes (Fig. 2), demonstrating that activated and elongating RNA polymerase II (ser-2-P) were present in both pronuclei. Thus, the transcripts identified by RNA-seq are likely derived from both male and female pronuclei. Furthermore, activated and elongated RNA pol II staining (ser-2-P) was also observed in both germ cell and somatic precursor nuclei in early embryos (Fig. 2). This differs from what has generally observed in C. elegans, where the germ cell nuclei are transcriptionally inactive and elongating RNA polymerase II is not observed until at least the 4-cell stage of development [24].

Figure 2. Immunolocalization of phosphorylated RNA polymerase II in early A. suum embryos.

1–8 cell embryos were stained with DAPI (blue) and immunohistochemistry carried out with affinity purified antibodies against the phosphorylated CTD of RNA polymerase II, where Ser2 indicates elongating polymerase and Ser5 indicates activated, promoter-associated polymerase. 1-cell embryos exhibit staining in both pronuclei prior to fusion indicating that early transcription occurs in both pronuclei. Both somatic and germ cell nuclei have significant pol II staining during early development. Other CTD and H5 (Ser2) antibodies produced similar staining patterns (data not shown).

Orthologs of C. elegans maternal mRNAs are transcribed in A. suum zygotes and 1–4 cell embryos

In C. elegans, a large number of maternal mRNAs and proteins are contributed to the oocyte [25]. Following oocyte formation, C. elegans pol II transcription is generally not thought to occur until about the 4-cell stage and development can continue without transcription as far as at least the 120-cell stage [16, 17]. In A. suum, maturation of the zygotes within the uterus is required for subsequent development [18]. During this period, we identified diverse and significant levels of transcription. We found that the steady state levels of mRNAs corresponding to A. suum genes orthologous to C. elegans maternal genes were significantly higher in developmental stages after oocytes and that transcription of many of these genes occurred during zygote maturation or in the 1–4 cell embryo (Table 1). These include orthologs of C. elegans car-1, mei-1, pal-1, lin-12/glp-1, gld-1, pie-1/oma/pos-1-like, and members of the puf and par families [25]. These genes are involved in transcription, signaling, polarity, and RNA regulation and are known to play key roles in C. elegans early development. Overall, these data suggest that many orthologs of C. elegans maternal mRNAs are synthesized in A. suum zygotes prior to pronuclear fusion and in early embryos. A more comprehensive comparison of oogenesis-related genes derived from a C. elegans microarray analysis [26] further indicates that for a large number of C. elegans oogenesis/maternal genes, their orthologs in A. suum are not maternally deposited; instead, they are synthesized when needed after fertilization and during early embryogenesis (see Supplemental Text).

Table 1.

Expression of Orthologs of C. elegans Maternal Genes during A. suum Development

A. suum Gene ID	C. elegans Gene Name	ovary	oocyte	zygote1	zygote2	zygote3	zygote4	1-cell (24hr)	2-cell (46hr)	2–4-cell (64hr)	10–26-cell (96hr)	32–64 (116hr)	~256-cell (7Day)	L1	L2	C. elegans Function
ASU_00037	lin-12/glp-1	2	3	6	10	14	26	41	77	150	41	9	9	5	3	Signaling
ASU_09625	car-1	244	105	315	264	179	356	409	362	183	136	116	91	58	82	RNA regulation
ASU_00286	crb-1	8	0	2	3	5	7	17	26	19	16	32	75	38	4	Signaling
ASU_03524	mei-1	7	3	4	6	6	8	9	13	9	10	12	10	8	7	Meiosis regulation
ASU_04180	pal-1	5	1	6	11	29	30	39	64	99	39	37	30	7	1	Transcription
ASU_12687	par-1	59	14	8	15	29	18	17	22	38	14	18	35	46	67	Cellular Polarity
ASU_09475	par-3	11	7	2	3	6	5	24	29	33	27	20	29	16	5	Cellular Polarity
ASU_03534	par-4	18	14	6	7	6	10	20	33	18	29	26	25	15	28	Cellular Polarity
ASU_05134	gld-1	15	7	12	14	14	19	26	25	13	3	1	1	1	1	RNA regulation
ASU_06726	gld-1	607	45	212	203	285	208	66	30	14	2	1	1	1	0	RNA regulation
ASU_05173	pie-1	126	232	364	501	723	851	869	984	705	142	16	11	9	0	Transcription and translation
ASU_08918	puf-9	70	95	131	177	424	247	185	199	208	65	85	102	51	56	RNA regulation
ASU_11218	puf-12	7	6	7	5	7	13	18	22	20	23	18	17	16	5	RNA regulation
ASU_14399	puf-11	23	10	26	30	39	47	27	19	24	5	6	7	6	9	RNA regulation

values are in RPKM

see also Table S3

A. suum mRNAs translation is coupled to newly transcribed RNAs rather than stored RNAs

During early embryonic development, gene expression is often regulated through differential mRNA translation. We examined the translational regulation of A. suum mRNAs to determine if they were developmentally and differentially recruited to ribosomes for translation. We assessed whether an mRNA is actively translated based on the distribution of the mRNA within a sucrose gradient following centrifugation. Cytoplasmic extracts of 0, 1, 3, 5, and 7 day A. suum embryos were subjected to sucrose density fractionation to distinguish mRNAs associated with no (Free fraction), a few (Middle fraction), or many ribosomes (polysomes, Heavy fraction) as a measure of their translation (Fig. 3A). As the embryos develop, we observed an increase in rRNA subunits in the Heavy fractions of the gradients corresponding to polysomes and increased translation from 1 to 7 day embryos (Fig. 3A). Treatment of extracts with EDTA or puromycin led to a significant shift in 80S ribosomes to lighter fractions, suggesting that ribosomes in heavier fractions are associated with active mRNA translation (Fig. S3A).

Figure 3. mRNA association with ribosomes during A. suum embryo development.

A. RNA profiles from A. suum embryo sucrose gradient fractionation. Cytoplasmic extracts from 0, 1, 3, 5, and 7 day embryos were centrifuged through 10–60% sucrose gradients and fractions collected. Total RNA from each fraction was separated on denaturing agarose gels and visualized with ethidium bromide staining. The prominent bands correspond to rRNAs and tRNAs as labeled. The fractions were pooled into a Free fraction (1–7), Middle fraction (8–13) and Heavy fraction (14–19) for RNA-seq and RT-PCR analyses.

B. RT-PCR data support the sucrose gradient RNA-seq results. RT-PCR of genes selected to represent different groups identified are shown with their corresponding color (as in Fig. 1C). F: Free fraction, M: Middle fraction and H: Heavy fraction. A strong correlation between RNA-seq and RT-PCR is observed (see Figure S1F and Table S1).

C. Newly transcribed embryonic mRNAs are readily translated during embryo development. RNA-seq was used to analyze the association of transcripts in fractions of sucrose gradients for 1, 3, 5, and 7 day A. suum embryos. Box-and-whisker plots illustrate and compare the location of transcripts in Free and Heavy fractions of the gradient by comparing the amount of each transcript as a percentage of total transcript level. In these and all subsequent box-and-whisker plots (in Fig. S3), the top and bottom ends of each box represent the 75th and 25th percentile, respectively; the middle line represents the median value; the extended lines illustrate the range (highest and lowest value); and dots represent values beyond the extremes of the whisker. *P < 0.05, **P < 0.01, t-test on the five-number summary.

See also Figure S3 and Table S4

RNA-seq libraries corresponding to 0, 1, 3, 5, and 7 day A. suum embryo mRNAs isolated from Free, Middle, and Heavy (polysomes) regions of sucrose gradients were prepared and sequenced. To confirm the polysome RNA-seq data, we carried out RT-PCR on >20 mRNAs (Fig. 3B, Fig. S1F, and Table S1) and 95% of the expression patterns were consistent with the polysome RNA-seq data. A global analysis of the data indicates that the majority of RNAs are present in the Middle fraction at all stages (Fig. S3B). The major differences in mRNA distributions are seen in those mRNAs that move to the heavy fraction or polysomes. Unexpectedly, the majority of maternal/1-cell zygote mRNAs are not primarily associated with heavy fraction polysomes and are not developmentally or differentially recruited to ribosomes. Most of these mRNAs are present in the Middle fraction and for most of the developmental stages, less than 25% are found in the Heavy fraction (Fig. 3C). In contrast, newly transcribed embryo/larva RNAs are more readily recruited to the Heavy fraction (Fig. 3C and Fig. S3B). On average, zygotic genes have ~30% or more of their mRNAs in polysomes. These data suggest that maternal/1-cell zygote mRNAs are, in general, not developmentally and differentially recruited to polysomes as has been observed for maternal mRNA translation in other organisms [2, 27]. Furthermore, transcripts made during early development appear to be recruited directly to polysomes and are not made and then deposited for future use. This data along with the RNA-seq transcriptome data indicate that the maternal contribution of RNAs and their translation is limited in A. suum; instead zygotically transcribed mRNAs are made and directly recruited to polysomes as needed during development.

We observed a cluster of 85 genes actively transcribed throughout early development that are poorly recruited to polysomes in 5-day embryos (Fig. S3C–D and Table S4). Over 80% of these genes encode ribosomal proteins. We identified a UUGUUGU motif present in the 3’ UTR of the majority of these A. suum ribosomal protein mRNAs (Fig. S3F). However, further analyzes indicate that this sequence motif alone cannot account for the translational regulation of these A. suum ribosomal protein mRNAs (see Supplemental Text).

We also evaluated the role of miRNAs in the regulation of gene expression during A. suum early development. Overall, our analyses (see Fig. S4D–F, Table S6, and Supplemental Text) suggest that mRNA levels during this period of development, in general, are not primarily regulated by miRNAs alone.

Conservation of embryonic gene expression patterns between A. suum and C. elegans

To further understand nematode development and gene expression conservation across nematode evolution, we compared the developmental transcriptomes of A. suum to those from three independent studies on C. elegans [14, 22, 28] (Fig. 4A). Baugh et al. and Levin et al. used handpicked embryos with precise staging whereas Hillier et al. used either early or late mixed stages embryos (Fig. 4A). For expression comparison, 5,765 orthologous gene pairs (OrthoMCL 1:1 pairs) from A. suum and C. elegans were used. Data were converted into standard Z-scores and expression correlations were evaluated using Pearson’s correlation (see Supplemental Experimental Procedures).

Figure 4. Conservation of embryonic gene expression patterns between A. suum and C. elegans.

A. Early development gene expression studies from C. elegans compared to A. suum developmental stages examined in the current study. Data from Baugh et al [14] and Levin et al [22] are derived from microarray experiments using handpicked embryos. Data from Hillier et al [28] are part of the modENCODE project that used RNA-seq to characterize C. elegans transcriptomes from various mixed developmental stages.

B. Genes with similar expression patterns in both A. suum and C. elegans. Genes with shared expression patterns between A. suum and the other three studies (with Pearson’s correlation coefficient R > 0.3) are indicated. Overall, 682 genes have conserved expression patterns with the two sets of handpicked embryos, and 297 genes are conserved from all three studies (see Table S5).

C. Conservation of gene expression patterns varies during development. Genes with conserved expression between staged A. suum and handpicked C. elegans embryos (682, see Fig. 4B) were compared to the total number of orthologous pairs from the same datasets. Shown are the percentages of the conserved genes (y-axis) as a function of the gene groups (see Fig. 1B) with distinct expression profiles (x-axis).

See also Table S5

When all these studies are compared, 4,531 (45%) orthologous gene pairs from A. suum and C. elegans show conserved embryonic expression patterns (see Fig. 4B and Table S5). This conservation varies for the 12 groups of genes (see Fig. 1B). Taking all three studies into consideration, we see an overlap of 297 gene pairs that have very consistent expression patterns. This number increased to 682 when data from the better-matched, handpicked stages of C. elegans (Fig. 4B) were used. These numbers are likely underrepresented as 1) different platforms (microarray and RNA-seq) were used for the analysis, 2) synchronicity of matched stages exhibits some variability, and 3) the analysis is based on stringent matching of 1:1 orthologous pairs. However, this set of 682 overlapping genes likely represents a core set of evolutionarily conserved genes necessary for nematode embryonic development and are mainly from gene groups 2, 5, 7, and 11 (Fig. 4C). The genes with the most conserved expression between the two nematodes correspond to group 11 (genes that peak at morphogenesis and L1 larvae). Levin et al. [22] reported a similar phenomenon in a study that compared five closely related Caenorhabditis species, where the conserved expression patterns were described as punctuated developmental milestones. Our data extend these observations in embryo development (4-cell and onward) between nematodes that have diverged ~400 million years ago [12].

Discussion

Here, we describe the early developmental transcriptomes and polysomal RNAs of the parasitic nematode A. suum. The availability of distinct stages of zygote maturation and early embryo development (Fig. 1A) enabled us to generate a comprehensive time course of mRNA expression, degradation, and translation in A. suum. Our study provides a unique, high-resolution analysis of maternal mRNAs, the onset of zygotic genome activation, maternal mRNA degradation, and early transcription and mRNA turnover in an organism with protracted early development. Although a variety of previous transcriptome studies have examined early development in model organisms, to our knowledge, none have comprehensively examined the period following fertilization through the 2-cell stage at this temporal resolution. Our RNA-seq data demonstrate that significant transcription occurs during zygote maturation prior to pronuclear fusion and in the 1–4 cell embryos of A. suum. This contrasts with the general view in metazoa that transcription is largely quiescent from the oocyte until at least the 2-cell stage. Notably, significant transcription occurs prior to pronuclear fusion. In addition, fewer maternally contributed mRNAs are observed in A. suum than in the related nematode, C. elegans, and the timing of the maternal to zygotic transition occurs much earlier. We found that many orthologs of maternal C. elegans mRNAs are not maternally contributed in A. suum, but are transcribed during A. suum zygote maturation and in the early embryo. Consistent with this, mRNAs made during A. suum early development, in general, do not appear to be stored for subsequent translation, but are directly translated following their synthesis. Overall, these data suggest that the timing for the maternal to zygotic transition in A. suum appears to occur soon after fertilization (Fig. 5). Our data provide a new perspective for the roles of maternally contributed and zygotically transcribed genes in metazoan early development. In A. suum, the maternal contribution is more limited, and newly transcribed genes appear to drive early development. This predominant use of newly transcribed genes may have been enabled by the protracted early development in A. suum.

Figure 5. Comparison of mRNA expression patterns during A. suum and C. elegans early development.

Note that maternal mRNAs are degraded and transcription initiates immediately after fertilization and continues prior to pronuclear fusion in A. suum (shaded area). A. suum major zygotic transcription also occurs much earlier than in C. elegans. A. suum mRNA levels are derived from Fig. 1B, with pink, red, blue, and sky blue representing genes from group 1–3, 4–5, 6–7, and 8–9, respectively. C. elegans data are obtained from several studies [2, 14, 23]. See also Figure S4 and Table S6

mRNA transcription and turnover during early A. suum development

Previous studies in sea urchins and mice suggested that there is some, but very limited, transcription in the paternal pronucleus of the zygote [1, 2]. In A. suum, ~ 2,500 genes are transcribed prior to pronuclear fusion (Fig. 5). This represents an unprecedented level of diverse transcription at this early stage of development. These transcripts are likely derived from both male and female pronuclei, as seen from immunohistochemical staining using antibodies against active, elongating RNA polymerase II (ser-2-P-CTD, Fig. 2). At least 222 genes show a significant increase in expression from the oocyte to zygote1 (Table S2), indicating transcription occurs immediately following fertilization. We also observed significant miRNA transcription in these stages in a previous study [10]. A number of mRNAs encoding core transcription machinery or specific transcription factors are upregulated during this period, consistent with their requirement for transcription. Many maternal mRNAs are rapidly degraded following fertilization, and all maternal mRNAs and most zygotically transcribed mRNAs are degraded before the 4-cell stage. These data suggest that A. suum zygotic gene activation and maternal mRNA degradation occurs very soon after fertilization and prior to pronuclear fusion (Fig. 5). Additional waves of transcription occur during early development (Fig. 1B, 1D and Fig. 5).

Major mRNA degradation occurs in both A. suum and C. elegans at the 4–32 cell stage (Fig. 5). In C. elegans, this developmental period is considered the maternal to zygotic transition. The degraded C. elegans mRNAs appear to be exclusively maternal while in A. suum, they are mixed populations of maternal and zygotic mRNAs. Degradation of mRNAs at this time in A. suum, long after the first onset of A. suum transcription and mRNA degradation prior to pronuclear fusion, may suggest that the high mRNA turnover at these stages may not be associated with the maternal to zygotic transition per se, but is an integral phenomenon of this developmental transition in nematodes. In A. suum, programmed DNA elimination (chromatin diminution) also occurs during this same time period [20]. Both of these processes may contribute to the mRNA changes we observed. Additional and significant periods of mRNA degradation also occur during the 1–4 cell stage (Fig. 5) and between the L1 and L2 larvae (Fig. 1B and 1D), suggesting additional dynamic gene regulation during these developmental stages.

Altered gene expression patterns in A. suum early development illustrate molecular plasticity in nematodes

Although early cleavages and cell lineage between A. suum and C. elegans are identical [13], we observed significant differences associated with the timing of early transcription and maternal contributions in early development. Significant A. suum transcription occurs during zygote maturation prior to pronuclear fusion and through 4-cell development. In contrast, transcription is thought to be absent or very limited during these periods in C. elegans development (Fig 5). Over 60% the mRNAs associated with early C. elegans development are maternally contributed [14, 23]. These maternal mRNAs are divided into two classes [23]. Levels of Class I maternal mRNA (70% of the maternal mRNAs) remain the same throughout early development, while Class II maternal mRNAs decrease with development (Fig. 5). The presence of this large number of maternal mRNAs is thought to drive many aspects of early C. elegans development.

We hypothesized that genes normally transcribed and stored maternally in oocytes and important for early C. elegans development, might be synthesized in A. suum during the extended period of zygote maturation in the uterus and 1–4 cells stage of the early embryo. In support of this hypothesis, we found that the orthologs of a variety of well-known C. elegans maternal mRNAs that play key roles in development were transcribed in A. suum either in maturing zygotes or 1–4 cell embryos (Table 1 and Table S3). Interestingly, several studies suggest that transcription in some nematodes (Acrobeloides nanus and Rhabditis sp.) is required during the first cleavages [29, 30]. Exposure to α-amanitin to inhibit pol II transcription stops development of A. nanus and Rhabditis sp. embryos at the 5-cell stage, whereas similar treatment of C. elegans 1-cell embryos results in 120–150 cell embryos. These differences in the requirement for transcription in early cleavages have been suggested to be a function of variation in maternal contributions to the embryos in different nematodes [29]. The duration of the C. elegans cell cycle during early development is on average 15–20 min [29, 31], whereas it is ~ 150 min in A. nanus [29], and ~13 hr in A. suum [19]. In addition, A. suum zygotes undergo an extended period of maturation in the uterus (~ 12 – 24 hr). Rapid development and cell division has been suggested to be potentially incompatible with significant transcription, necessitating a reliance on maternal contributions and post-trancriptional gene expression under those circumstances [32–34]. The very slow development in A. suum may have allowed a shift away from the use and reliance on maternal contributions for early development and enabled transcriptional programs to drive differential gene expression in early development. This was previously suggested for the nematode A. nanus [29], and here we provide molecular evidence to support this hypothesis. Therefore, even though C. elegans and A. suum share identical early developmental patterns, the regulatory mechanisms used for controlling the timing of the expression of key conserved genes has been altered. Our data illustrate significant plasticity in the regulatory networks that play important roles in developmental outcomes in nematodes.

Onset of zygotic gene activation prior to cleavage in A. suum

The transition from the oocyte to development involves transcriptional silencing and then reactivation. Zygotic genome activation in animals has been historically observed during early division. Our data indicate that this can occur in the pronuclei of the zygote and prior to pronuclear fusion. This opens the possibility that this strategy of early gene expression may be used in other animals. It is possible that transcription in some other animals may begin as early as we have observed but has not been detected due to the lack of temporal resolution at the appropriate stages or the depth of the analysis. Recent studies in Drosophila suggest some transcription occurs in the pre-blastoderm, earlier than transcription was previously thought to occur [35–38]. In addition, recent studies in mice and humans have described additional transcription at very early stages of development [39].

Molecular conservation in A. suum and C. elegans early development

The nematode C. elegans is a widely used genetic model for general developmental studies and is considered a prototypical nematode. However, nematodes are a diverse phylum and studies on early nematode development have shown significant variations [40, 41]. The early cleavages and cell lineage between A. suum and C. elegans are identical despite ~400 million years of evolutionary divergence [12, 13]. A comparison of developmental transcriptomes (4-cell stage to L2) between the two demonstrated that ~40% of orthologous genes have similar expression patterns, and we found 682 genes whose expression patterns are highly conserved (Fig. 4 and Table S5). While our study defines major differences in the timing of expression of genes involved in very early development between A. suum and C. elegans, our comparison also revealed conserved nematode gene expression patterns in early development. A similar observation was recently described by Levin et al. [22] for closely related Caenorhabditis sp. and the conserved patterns were defined as developmental milestones. Our data expands this concept to nematodes as divergent as C. elegans and A. suum, suggesting the expression of these genes represents a core set of important genes for early nematode development.

Conclusions

In summary, we show that the maternal mRNA contribution to early nematode development in A. suum is significantly less than in C. elegans. In A. suum, significant transcription initiates immediately following fertilization, continues prior to pronuclear fusion, and proceeds in early cell divisions. Overall, these data indicate that significant levels and diversity of zygotic transcription occurs in Ascaris much earlier than previously thought in nematodes and metazoa. While there are numerous developmental processes and orthologous genes conserved between C. elegans and A. suum, our data indicate that there is also flexibility in the mode of regulation for gene expression, and that rewiring of the gene expression programs from a primarily post-transcriptional to a transcriptional mechanism, can readily be used to generate the appropriate timing of gene expression necessary for early nematode development.

Highlights.

~4,000 genes are transcribed in pronuclei and 1–4 cell embryos in nematode A. suum

Maternally contributed mRNAs are degraded right after fertilization in A. suum

Post-transcriptional regulation is rewired to transcriptional in A. suum development