Skip to main content
NCBI home page
Molecular Biology and Evolution logoLink to Molecular Biology and Evolution
. 2016 Jun 10;33(9):2337–2344. doi: 10.1093/molbev/msw104

Characterization of Somatically‐Eliminated Genes During Development of the Sea Lamprey (Petromyzon marinus)

Stephanie A Bryant 1, Joseph R Herdy 1, Chris T Amemiya 2,3, Jeramiah J Smith 1,*
PMCID: PMC4989109  PMID: 27288344

Abstract

The sea lamprey (Petromyzon marinus) is a basal vertebrate that undergoes developmentally programmed genome rearrangements (PGRs) during early development. These events facilitate the elimination of ∼20% of the genome from the somatic cell lineage, resulting in distinct somatic and germline genomes. Thus far only a handful of germline-specific genes have been definitively identified within the estimated 500 Mb of DNA that is deleted during PGR, although a few thousand germline-specific genes are thought to exist. To improve our understanding of the evolutionary/developmental logic of PGR, we generated computational predictions to identify candidate germline-specific genes within a new transcriptomic dataset derived from adult germline and the early embryonic stages during which PGR occurs. Follow-up validation studies identified 44 germline-specific genes and further characterized patterns of transcription and DNA loss during early embryogenesis. Expression analyses reveal that many of these genes are differentially expressed during early embryogenesis and presumably function in the early development of the germline. Ontology analyses indicate that many of these germline-specific genes play known roles in germline development, pluripotency, and oncogenesis (when misexpressed). These studies provide support for the theory that PGR serves to segregate molecular functions related to germline development/pluripotency in order to prevent their potential misexpression in somatic cells. This larger set of eliminated genes also allows us to extend the evolutionary/developmental breadth of this theory, as some deleted genes (or their gnathostome homologs) appear to be associated with the early development of somatic lineages, perhaps through the evolution of novel functions within gnathostome lineages.

Keywords: lamprey, development, genome, rearrangement, vertebrate.

Introduction

It was once believed that all cell types within an organism have identical genomes, with cells of the immune system and cancer cells being the notable exceptions. Yet some organisms have been found to undergo extensive DNA elimination events, resulting in germ cells with genomes distinct from that of somatic cells. These events typically occur during early development and can result in elimination of up to ∼90% of the germline DNA from somatic cell lineages in some organisms (Wang and Davis 2014). Though the mechanism and biological function of these events are still debated, over 100 organisms across a broad phylogenetic distribution are known to undergo some form of DNA elimination, including nematodes, insects, copepods, and vertebrates such as hagfish, lamprey, and finches (Pigozzi and Solari 1998; Goday and Esteban 2001; Kubota et al. 2001; Bachmann-Waldmann et al. 2004; Drouin 2006; Smith et al. 2009).

Early studies on nematodes (Ascaris suum and Parascaris univalens) and hagfish (Eptatretus cirrhatus) suggested that eliminated DNA was highly repetitive, and more recent studies in zebra finch and copepods seem to corroborate these findings (Muller et al. 1982; Goto et al. 1998; Degtyarev et al. 2004; Drouin 2006; Itoh et al. 2009; Niedermaier and Moritz 2000). Yet the identification of highly repetitive sequences within eliminated DNA does not exclude the possibility that these events also facilitate the elimination of genes that are actively transcribed in developing or mature germ cells. In sea lamprey (Petromyzon marinus), studies have revealed that DNA eliminated during programmed genome rearrangement (PGR) contains large amounts of repetitive DNA (Smith et al. 2009, 2010). However, somatically retained sequences are also highly repetitive and a small number of validated genes have been identified within the eliminated fraction of the lamprey genome (Smith et al. 2010, 2013). Efforts aimed at identifying eliminated genes have the capacity to provide critical insight into the evolutionary/developmental logic of DNA elimination by identifying biological functions that are withheld from somatic cell lineages but retained by the germline.

Studies with the resolution to parse relatively rare single-copy sequences from large fractions of repetitive DNA, such as those utilizing high‐throughput sequencing, have already enabled deeper genomic characterization of DNA elimination events. Deep sequencing of A. suum germline and somatic genomes revealed that, in addition to the removal of repetitive DNA, ∼12.7 Mb of single-copy sequences are eliminated from the somatic genome (Wang et al. 2012). Similarly, analysis of low coverage genomic shotgun sequencing of the sea lamprey germline genome identified a handful of well-validated genes that are eliminated during PGR and several other candidate loci (Smith et al. 2013). Consistent with evidence that eliminated genes in A. suum are expressed in the germline and during early embryogenesis, ontology analyses indicate that eliminated protein-coding sequences in lamprey likely play a role in germ cell development and in the adult germline (Smith et al. 2012; Wang et al. 2012).

To gain further insight into the functional role of eliminated genes, we sought to validate candidate deletions and examine their transcriptional patterns in various developmental stages, including those encompassing PGR. Here, we present new data revealing patterns of gene expression, DNA loss, and comparative functional information from 44 genes that can be definitively classified as eliminated from somatic cell lineages during PGR. Analysis of transcriptional data indicates that a large fraction of eliminated genes are differentially expressed throughout lamprey embryogenesis, particularly during the stages surrounding PGR and the maternal-to-zygotic transition (MZT). Altogether, these analyses provide insight into the patterns of DNA loss over the time course of PGR and the evolutionary/developmental logic underlying the programmatic elimination of DNA from somatic cell lineages. Moreover, these data corroborate previous studies proposing that eliminated genes have conserved roles in germline development and oncogenesis (when misexpressed), and expand these studies by identifying several new genes that likely contribute to these critical biological processes (Smith et al. 2012).

Results and Discussion

To characterize transcripts that are expressed in germline tissues and during the time course of PGR, RNAseq data were generated for pools of embryos that were collected at D1 (day 1), D2, D2.5, D3, D4, and D5 post-fertilization. These time points correspond to the 24–32 cell stage, the blastula stage, the midblastula transition, dorsal cone formation, gastrulation, and neural groove formation, respectively, and coincide with major changes in gene expression during vertebrate development. These sequencing runs yielded ∼130.4 million 100-bp paired-end reads, which were relatively evenly distributed across time points (D1: 25 million, D2: 19.9 million, D2.5: 20.7 million, D3: 24 million, D4: 22.9 million, D5: 17.9 million). These reads and previously published data from testes (SRX104180) and later embryonic stages (SRX110029-35) were used to assemble a reference transcriptome. The final transcriptome assembly consists of 74,938 unique transcripts and 70,975 additional subsequences (i.e., isoforms or assembly variants). Estimates of transcript abundance in testes and individual embryonic stages were recalculated using RSEM v1.2.9 (Li and Dewey 2011). Alignment of assembled transcripts to human RefSeq proteins generated a total of 66,948 blast hits (E < 1e−4), though several transcripts aligned to the same human homolog. In total, these transcripts yielded alignments to 13,159 unique human genes.

To identify transcripts within eliminated regions, assembled contigs were filtered based on two criteria: (1) computational evidence for somatic elimination (i.e., transcript presence in lamprey germline assembly and absence in the somatic WGS assembly), and (2) average expression values for each transcript in testes and D1–5 embryos as estimated by RSEM (Li and Dewey 2011). These criteria identified a total of 576 transcribed sequences that were considered candidates for somatic elimination during PGR. PCR-validation of these candidate genomic intervals identified 44 genes that are present in lamprey germline DNA (testes) but absent or depleted in DNA that was extracted from a panel of somatic tissues (liver, kidney, fin, muscle, and blood) (fig. 1, supplementary fig. S1, Supplementary Material online). The remaining genes yielded similar amplification patterns in germline and somatic tissues, and thus did not meet our operational criteria for classification as somatically eliminated. These may represent portions of genes that are present in the soma but were not sampled by somatic genome shotgun sequencing due to random sampling processes, genes with highly similar paralogs in germline and soma, or biases in the shotgun cloning/sequencing approach.

Fig. 1.

Fig. 1

Open in a new tab

Examples of secondary validation of somatically depleted genes. Validated sequences show amplification in testes DNA but no or reduced amplification in DNA from a panel of somatic tissues (liver, kidney, fin, muscle, blood) from two animals (males 12 and 13). T = Testes, L = Liver, K = Kidney, F = Fin, M = Muscle, B = Blood, m = 100 bp DNA ladder. The lowest band under “m” corresponds to the 100-bp band of the DNA ladder in all panels except for that of HMF1, which corresponds to the 200-bp band. Amplified loci are labeled with their corresponding human homolog.

While levels of amplification between germline and somatic tissues as a whole are strikingly different, background amplification was observed for some primers and tissues. In particular, several faint bands were observed in amplified samples from male 12 muscle DNA and male 13 blood DNA. To test whether these background bands represent amplification of the targeted germline candidate region, paralogous sequence, or amplification artifacts, we sequenced PCR products corresponding to all background bands from these two samples and their matched germline amplicon. Of the 78 amplicons targeted for sequencing, 25 yielded sequence data of sufficient quality to permit comparison to their corresponding germline sequence. For all of these samples, the sequences for blood and muscle were identical to their corresponding germline bands (supplementary table S1, Supplementary Material online). This observation suggests that background amplicons are likely derived from rare cells that retain some fraction of the germline genome yet reside in somatic tissues, rather than representing inefficient amplification of paralogous sequences or amplification artifacts. While characterization of these specific cell types is likely to be challenging, it is interesting to speculate that these might represent cells that are capable of adopting a wider range of cell fates later in development (i.e., resident stem cells). Alternately, amplification of germline-specific sequences might reflect the presence of cell-free DNA that traces its origin to the differentiated germline.

Of our 44 validated germline-specific sequences, 20 aligned to human RefSeq proteins. These 20 genes, in addition to three other genes that exhibited dynamic patterns of expression during development, were chosen for qPCR analysis to characterize the pattern of DNA elimination and the time course of PGR. The abundance of germline DNA sequences relative to retained sequences was estimated by quantitative real-time PCR in genomic DNA from embryos (D1–5), testes, and blood. Results of these analyses reveal a relatively uniform pattern of elimination across tissues, consistent with previous findings that indicate that DNA loss coincides with the MZT (fig. 2) (Smith et al. 2009). At D1, the majority of surveyed sequences are at their highest relative abundance, closely resembling estimates in the germline (testes). By D2.5, the approximate timing of the MZT, the abundance of all germline-specific DNAs has substantially decreased, and by D5 most germline-specific DNAs are essentially absent, similar to estimates in somatic tissue (blood).

Fig. 2.

Fig. 2

Open in a new tab

Relative abundance of a subset (n = 23) of somatically depleted genes based on qPCR in genomic DNA. Relative abundance estimates were calculated using the ΔΔct method of relative quantification using two sequences that amplify consistently in germline, somatic, and embryonic tissues as controls (supplementary table S1, Supplementary Material online, comp4257946_c0_seq1 and comp4249214_c0_seq1). The ΔΔct values for each gene are standardized to ΔΔct values from testes. D1, D2, D2.5, D3, D4, and D5 correspond to days 1–5 post-fertilization.

While this pattern of deletion largely correlates with that of the previously described germline-limited fragment Germ1, the presence of a slight excess of residual DNA between D2.5 and D5 suggests the existence of a transitional period between the PGR event at D2.5 and the physical removal of targeted DNA apparent by D5 (Smith et al. 2012). This observation is particularly interesting in light of previous studies suggesting the presence of abundant DNA breaks in post-gastrulation embryonic stages (Smith et al. 2009).

Closer examination of our RNAseq analysis revealed that the effective read counts of the majority of our validated genes is ∼1 across most embryonic time points, presumably due to low transcription levels in early embryonic stages. However, patterns of gene expression appeared to be generally consistent with the timing of the activation of zygotic transcription, with the number of transcriptionally active genes increasing after D2.5 (supplementary fig. S2, Supplementary Material online). Differential expression analysis identified 33 genes within the total validated subset that show significant differential expression between tissues (supplementary fig. S2, Supplementary Material online). Because eliminated genes were generally characterized by low effective read counts in our RNAseq datasets, we cross-validated our analyses using the Nanostring nCounter Gene Expression Assay (Nanostring Technologies, Seattle, USA).

Nanostring probes were designed to target transcripts with coding sequences that overlap validated germline genomic intervals (supplementary table S1, Supplementary Material online). Because two germline reads aligned to variants of the same transcript, for which only one unique probe could be designed, only 43 unique transcript sequences were used for probe design. Nanostring abundance estimates generated average normalized transcript counts of ∼87 across embryonic samples. Temporal clustering of transcript counts reveals that most genes exhibit an abrupt increase in transcription after D2 (fig. 3). This broad-scale pattern presumably reflects major changes in gene expression that accompany the onset of zygotic gene expression at the MZT (Yartseva and Giraldez 2015). Fold change analyses identified 42 of 43 genes that display significant (P < 0.05) differential expression between tissues. Of these, 42 are differentially expressed relative to testes, 25 are differentially expressed between D1 and later embryonic time points (D2–5), and three are differentially expressed between adjacent embryonic time points (fig. 3). Together, results of both RNAseq and Nanostring experiments reveal that eliminated genes are differentially transcribed during lamprey development, suggesting that many of the genes targeted for elimination by PGR may play a functional role in embryogenesis.

Fig. 3.

Fig. 3

Open in a new tab

Transcription and differential expression of somatically depleted genes based on Nanostring estimates. (Left) Transcript abundance as estimated by Nanostring count values. Data are presented using a spectral color theme to highlight the broad range of expression values across genes, tissues, and embryonic time points. Purple indicates low expression and red indicates high expression. Transcripts are labeled with their corresponding human homolog. (Center) Log2-fold changes in expression relative to D1. (Right) Red asterisks indicate genes that are differentially expressed between testes and any embryonic time point (D1–5), blue asterisks indicate genes that are differentially expressed between D1 and any subsequent time point (D1 vs. D2, D1 vs. D2.5, etc.), and green asterisks indicate genes that are differentially expressed between adjacent time points (D1 vs. D2, D2 vs. D2.5, etc.). D1, D2, D2.5, D3, D4, and D5 correspond to days 1–5 post-fertilization.

To shed light on the possible function of eliminated sequences, we performed a statistical overrepresentation test using PANTHER (Huang da et al. 2009a, b; Maglott et al. 2011; Mi et al. 2016). Within the full set of genes screened for germline-specificity (n = 576), 311 had nonredundant hits to the human RefSeq protein database and were used as the reference list in our overrepresentation test. Within the validated subset of germline-specific genes (n = 44), 20 had hits to the human RefSeq protein database, 17 of which were nonredundant and were used as the analyzed list in our overrepresentation test. 294 protein IDs from the reference list and 15 protein IDs from the analyzed list were mapped to the PANTHER database.

Analysis of these lists using the PANTHER Pathways annotation data set indicates that genes in the Wnt and Cadherin signaling pathways are the most overrepresented in our validated subset of genes, accounting for 20% and 13.33% of our validated genes, respectively. However, the majority of the genes in each list were unclassified (n = 244 reference genes and n = 9 validated genes). As such, overrepresentation estimates are based on only six unique protein IDs from the analyzed list, and several of these appear in multiple categories. For instance, the two genes that are assigned to the Cadherin signaling pathway are also assigned to the Wnt signaling pathway, resulting in only three unique genes falling under these annotation categories (WNT-5A, Myosin-7B, and Protocadherin Fat 3). Other pathways that this analysis indicates are overrepresented among our validated genes include the p53 pathway and p53 pathway feedback loop 2, Inflammation mediated by chemokine and cytokine signaling, asparagine and aspartate biosynthesis, and nicotinic acetylcholine receptor signaling. However, these overrepresentation values are based on the presence of one gene in each category, many of which occur in multiple categories.

To gain further insight on the function of eliminated genes, including those that were unclassified in our PANTHER Pathways analysis, we manually examined gene-specific datasets curated at NCBI gene to gain further perspective on the known functions of their human homologs (Maglott et al. 2011). Consistent with the results of our statistical overrepresentation test, this broad-based functional information suggests that several of our validated germline-specific genes are involved in Wnt and Cadherin signaling pathways, both of which are strongly implicated in oncogenesis (Polakis 2000; Nelson and Nusse 2004). Among our validated deletions are sequences with homology to HFM1 ATP-dependent DNA helicase homolog (HFM1), which is expressed primarily in germ cells, and human cancer-testis antigen synaptonemal complex protein 1 (Tanaka et al. 2006; Yi et al. 2007). Notably, expression of cancer-testis antigens is normally limited to the germline, and is only observed within somatic tissues in the context of oncogenesis (Fratta et al. 2011). Moreover, previous experiments have shown that ectopic expression of germline-limited genes contributes to tumorigenesis in other species (Drosophila and Hydractinia) (Janic et al. 2010; Fratta et al. 2011; Millane et al. 2011). Other deleted genes include protein phosphatase Mg2+/Mn2+ dependent 1D (PPM1D), hydroxylysine kinase (HYKK), wingless-type MMTV integration site family member 5A (WNT-5a), and MAD2L1 binding protein (MAD2L1BP), all of which are implicated in oncogenesis or tumorigenesis either in their normal state or when misexpressed (Date et al. 2013; Lin et al. 2014; Wang and Liu 2014; Zhang et al. 2014). These results are consistent with homology information for previously validated deletions, which include sequences homologous to cancer-testis antigen 68 and WNT7a/b (Smith et al. 2012).

Other genes in our validated subset tend to fall into categories related to apoptosis and development. In this context, it seems likely that apoptosis regulators such as baculoviral IAP repeat containing 1 (BIRC1) and baculoviral IAP repeat containing 3 (BIRC3) may be targeted for deletion to prevent possible misregulation of apoptosis and subsequent tumorigenesis in somatic cells (Davoodi et al. 2010; Bai et al. 2014; Allam et al. 2015). Yet a small number of deleted genes have no overt annotations related to oncogenesis or other disease states. These include retinitis pigmentosa GTPase regulator (RPGR: eye photoreceptor development), myosin, heavy chain 7B, cardiac muscle, beta (MYHB: skeletal and cardiac muscle development), glutamic–oxaloacetic transaminase 1 (amino acid metabolism), and leucine-rich repeat containing eight family member C (LRRC8C: adipocyte and immune cell differentiation) (Tominaga et al. 2004; Murga-Zamalloa et al. 2010; Gakovic et al. 2011; Shen et al. 2011; Esposito et al. 2013;). It seems plausible that this relatively small fraction of genes might have evolved germline-specific functions in the lamprey lineage or that gnathostome genomes have evolved to deploy this subset of genes during the development and maintenance of somatic cell lineages. In this regard, it is notable that some eliminated genes have human homologs that play roles in nervous system development, such as FAT atypical cadherin 3 (FAT3: similar to the Drosophila tumor suppressor gene FAT), cordon-bleu WH2 repeat protein like 1, neurofascin, RPGR, and trinucleotide repeat containing 18 (Margolis et al. 1997; Mitsui et al. 2002; Carroll et al. 2003; Koticha et al. 2005; Murga-Zamalloa et al. 2010; Gakovic et al. 2011; Buttermore et al. 2012). As these genes are known to be involved in neurogenesis in many other vertebrate and nonvertebrate species, it is unclear how genes with these functions may have become dispensable in the lamprey somatic genome.

While these analyses provide insight into possible functions and expression patterns of germline-specific genes over early embryonic development, it is important to note that transcription at the whole-embryo level is a function of both transcriptional regulation and DNA elimination. We reasoned that comparing the pattern of genomic DNA elimination to our transcriptional data might provide greater insight into the relationship between gene abundance and transcription throughout development. While our qPCR analyses show that the predicted pattern of elimination in genomic DNA holds true for most surveyed genes, a small number of sequences show greater variation in abundance across embryogenesis. These include genes with homology to FAT3, BIRC1, and BIRC3. The elimination of these genes seems to occur over a more protracted period than the majority of other eliminated genes. Comparing this pattern of elimination to our estimates of gene expression, we find that their transcript abundance is highest at D3–5. We speculate that the delayed elimination of these genes might reflect their function, perhaps indicating that they play some role in the early differentiation of somatic lineages. As such, it may be more constructive to conceptualize eliminated genes as somatically depleted, rather than “germline-specific” per se.

In depth characterization of lamprey embryonic transcriptomes has permitted the identification of a larger catalog of somatically depleted genes and provides critical evidence supporting the implicit hypothesis that eliminated genes are biologically functional (i.e., they are transcribed). These data corroborate previous findings in studies of the nematode A. suum that identify single-copy transcribed genes related to germline maintenance and development in eliminated DNA, providing further evidence that PGR may be a bona fide mechanism of germline gene regulation. Somatically depleted genes identified here may be useful markers of eliminated material in future experiments designed to further characterize PGR. Studies aimed at identifying more contiguous eliminated chromosomal fragments and their sequence content will likely provide greater perspective on the content of eliminated DNA and the biological function of PGR. Altogether, these studies contribute to a growing body of evidence that suggests PGR serves as a permanent gene silencing mechanism that prevents misexpression of several genes related to germline development, pluripotency, and oncogenesis, including genes for which such functions have yet to be defined in human or other mammalian models.

Materials and Methods

Animals

All animals were obtained from the Lake Michigan population via the Great Lakes Fisheries Commission and maintained under University of Kentucky IACUC protocol number 2011-0848. Animals were euthanized by immersion in MS-222 (150 μg/ml), dissected, and tissues were immediately snap-frozen for the isolation of DNA from adult germline and somatic tissues.

Lamprey Embryos

In vitro fertilizations were performed with sexually mature adult animals. Eggs and sperm were collected in crystallization dishes and allowed to incubate in 10% Holtfreter’s solution for 10 min to permit fertilization (Nikitina et al. 2009). After visually confirming activation, embryos were rinsed in distilled water to remove excess sperm and maintained in 10% Holtfreter’s solution at 18°C throughout development. At days 1, 2, 2.5, 3, 4, and 5 post-fertilization, embryos were collected in 1.7 ml centrifuge tubes and snap frozen for subsequent RNA extractions. Pools of embryos used for these analyses were obtained from several independent in vitro fertilizations. Pools used for RNAsequencing (fertilizations generated in 2012) were also independent of those used in Nanostring and qPCR experiments (fertilizations generated in 2015).

RNA Sequencing

Total RNA was isolated from lamprey embryos using Trizol extraction. RNA quality was assessed on the Agilent 2100 Bioanalyzer (Agilent Technologies) and samples with RNA Integrity Number (RIN) >8 were sent to the HudsonAlpha Genomic Services Lab (HudsonAlpha, Huntsville, AL) for paired-end sequencing on the Illumina platform. Raw reads were assembled using Trinity (trinityrnaseq_r2013-02-25) using default parameters and integrated quality clipping with Trimmomatic (Bolger et al. 2014).

Screening for Somatically Eliminated Fragments

To identify sequences eliminated during PGR, transcripts were filtered based on computational evidence for somatic elimination. Candidate germline-specific transcripts were identified from the transcriptome assembly by aligning contigs to shotgun sequence datasets from germline (SRX025555) and somatic (liver, AEFG01) whole genome shotgun datasets using Blast (Altschul et al. 1990). To provide homology information, reads were aligned to human RefSeq proteins using BLASTx (Altschul et al. 1990). Transcripts that aligned to the germline sequence dataset (>98% identical over >100 bp) and failed to align to the somatic dataset at these same thresholds were selected for secondary screening. Secondary screening was performed to filter transcriptome assembly artifacts based on transcript abundance estimates generated by RSEM v1.2.9 (Li and Dewey 2011). To target germline-specific transcripts with a broad range of predicted expression during development, 480 transcripts with the highest combined average expression in testes and embryos and 96 transcripts with high expression in testes and low expression in embryos were selected for PCR validation.

PCR validation assays were performed in DNA from blood and testes in order to further evaluate predicted germline-specific regions. To assess whether patterns of deletion were consistent across somatic tissues, PCR amplifications were also performed using DNA from liver, kidney, fin, and muscle. Oligonucleotide primer pairs were designed for the genomic sequences corresponding to candidate germline-specific reads using Primer3 and used to prime PCR reactions under the following amplification conditions [1 ng of DNA, 50 ng of each primer, 3 U Taq polymerase, 5× PCR buffer, and 200 mM each of dATP, dCTP, dGTP, and dTTP; oligonucleotide sequences and optimal thermal cycling conditions are provided in supplementary table S1, Supplementary Material online] (Rozen and Skaletsky 2000). The DNAs used in these reactions were extracted from testes, liver, kidney, fin, muscle, and blood collected from two individuals (Males 12 and 13) using standard phenol/chloroform extraction (Sambrook et al. 2006).

Sequencing PCR Products

For all 44 validated genes, PCR products corresponding to testes DNA from males 12 and 13 were sequenced. For sequences that also amplified in somatic tissues, PCR products corresponding to muscle DNA from male 12 and blood DNA from male 13 were sequenced. Muscle and blood DNA were chosen due to their consistent levels of relatively high background amplification. Samples for which there was no apparent background amplification in muscle or blood were not sequenced. In total, sequence data was generated for PCR products from 88 testes bands, 41 muscle bands, and 37 blood bands. Prior to sequencing, PCR products were purified with ExoSAP-IT (Affymatrix) according to the manufacturer’s instructions. 4 μl of purified template was added to 1 μl of primer mix (5 pM/μl) and sequenced on the ABI3730 at the Advanced Genetic Technologies Center at the University of Kentucky. Sequence data were assembled using SeqMan Pro v13.0 (DNASTAR) with the ProAssembler algorithm. SNP Discovery parameters were as follows: Minimum Score at SNP: 20; Minimum Neighborhood Score: 0 (default); Minimum Neighborhood Window: 0 (default); Heterozygous Peak Threshold: 50%; without Strict Base Matching. These data are included in supplementary table S1, Supplementary Material online.

Nanostring Gene Expression Analysis

The Nanostring nCounter Gene Expression Assay (Nanostring Technologies, Seattle, USA) was used to estimate expression of the 44 validated germline-specific genes throughout embryogenesis and in the adult germline. The transcript sequences to which the validated germline-specific genes aligned and the sequences for six control genes were submitted to Nanostring Technologies (Seattle, USA) for custom CodeSet design. Due to sequence similarity between comp266794_c0_seq1 and its variant, comp266794_c0_seq2, only a shared probe could be designed for these targets. The final codeset included 49 capture probes; 43 designed from the transcript sequences corresponding to validated germline-specific genes and six designed from control genes (supplementary table S1, Supplementary Material online).

The Direct-zol RNA MiniPrep (Zymo Research) was used to isolate RNA from pools of snap-frozen embryos collected at days 1–5 post-fertilization (D1, D2, D2.5, D3, D4, and D5; n = 3 biological replicates each) and adult testes (n = 3 technical replicates). RNA was quantified on the Nanodrop 2000 Spectrophotometer (Thermo Scientific) and RNA integrity was assessed using the Agilent 2100 BioAnalyzer (Agilent Technologies). RNA samples with RNA Integrity Number (RIN) ≥8 were used in the Nanostring nCounter Assay according to the manufacturer’s protocol.

Count data for all genes were analyzed by the nSolver Analysis software v2.5 (Nanostring Technologies, Seattle, USA). Background subtraction was performed using the geometric mean of internal negative controls. Normalization was performed using the geometric mean of internal positive controls and the mean of two endogenous housekeeping genes (homologs of EF1A_A and EF1A_B), which show consistent expression in embryos and testes, respectively.

Real-Time PCR

Quantitative real-time PCR was used to measure the relative abundance of a subset (n = 23) of validated germline-specific genes during embryogenesis and in germline (testes) and somatic (blood) tissues. This subset of genes consisted of 20 sequences that aligned to human RefSeq proteins and three other genes that exhibited patterns of dynamic expression during development based on RNAseq estimates. DNA was extracted from blood, testes, and D1, D2, D2.5, D3, D4, and D5 embryos using standard phenol-chloroform extraction (Sambrook et al. 2006). Real-time PCR was performed on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad) using Sso Advanced Universal SYBR Green Super Mix, ∼1 ng DNA, and 50 ng of each primer. Thermal cycling conditions were 3 min initial denaturation at 95°C, followed by 50 cycles of 95°C for 10 s and 60°C for 30 s. Relative expression estimates were calculated using the ΔΔct method of relative quantification using two control sequences shown to amplify consistently in germline, somatic, and embryonic tissues (supplementary table S1, Supplementary Material online). Final ΔΔct values were standardized to testes values.

Differential Expression Analysis

Differential expression analysis of RNAseq data was performed on the full RNAseq datasets for embryos and testes using EBseq v1.1.5 (Leng et al. 2013). EBseq employs an empirical Bayes method to identify differentially expressed genes and isoforms across two or more conditions in an RNAseq experiment (Leng et al. 2013). Fold change was calculated between testes and embryos, between D1 embryos and each subsequent embryonic stage (D1 vs. D2, D1 vs. D2.5, etc.), and between adjacent embryonic time points (D1 vs. D2, D2 vs. D2.5, etc.). False discovery rate was controlled at 0.05. Differential expression analysis of Nanostring data was performed using the nSolver Analysis software v2.5 (Nanostring Technologies, Seattle, USA). Following background subtraction and normalization, fold change estimates were calculated between testes and embryos, between D1 embryos and each subsequent embryonic stage, and between adjacent embryonic time points. Those with P-values <0.05 were classified as differentially expressed. Hierarchical clustering of RNAseq and Nanostring data was performed using Ward multivariate two-way clustering (SAS Institute Inc. 1989–2007).

Gene Ontology Analysis

To assess whether any functional categories were overrepresented among deleted genes, we performed a PANTHER Overrepresentation Test (release 20160321; PANTHER version 10.0 Released 2015-05-15) (Mi et al. 2016). A nonredundant list of the human RefSeq proteins identified for our validated genes (n = 17) was compared with the list of nonredundant human RefSeq proteins (n = 311) in the entire set of screened candidates using the PANTHER Pathways Annotation dataset.

Supplementary Material

Supplementary table S1 and figures S1 and S2 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).

Supplementary Data
supp_33_9_2337__index.html (1KB, html)

Acknowledgments

We thank Brett Spear and Shirley Qui for granting access to real-time PCR resources used in this project. This study was funded by the National Institute of General Medical Sciences of the National Institutes of Health under award number R01GM104123 to J.S. and under award number R24GM095471 to C.T.A.

References

  1. Allam R, Maillard MH, Tardivel A, Chennupati V, Bega H, Yu CW, Velin D, Schneider P, Maslowski KM. 2015. Epithelial NAIPs protect against colonic tumorigenesis. J Exp Med. 212:369–383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol. 215:403–410. [DOI] [PubMed] [Google Scholar]
  3. Bachmann-Waldmann C, Jentsch S, Tobler H, Muller F. 2004. Chromatin diminution leads to rapid evolutionary changes in the organization of the germ line genomes of the parasitic nematodes A. suum and P. univalens. Mol Biochem Parasitol. 134:53–64. [DOI] [PubMed] [Google Scholar]
  4. Bai L, Smith DC, Wang S. 2014. Small-molecule SMAC mimetics as new cancer therapeutics. Pharmacol Ther. 144:82–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30:2114–2120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Buttermore ED, Piochon C, Wallace ML, Philpot BD, Hansel C, Bhat MA. 2012. Pinceau organization in the cerebellum requires distinct functions of neurofascin in Purkinje and basket neurons during postnatal development. J Neurosci. 32:4724–4742. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Carroll EA, Gerrelli D, Gasca S, Berg E, Beier DR, Copp AJ, Klingensmith J. 2003. Cordon-bleu is a conserved gene involved in neural tube formation. Dev Biol. 262:16–31. [DOI] [PubMed] [Google Scholar]
  8. Date DA, Burrows AC, Venere M, Jackson MW, Summers MK. 2013. Coordinated regulation of p31(Comet) and Mad2 expression is required for cellular proliferation. Cell Cycle 12:3824–3832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Davoodi J, Ghahremani MH, Es-Haghi A, Mohammad-Gholi A, Mackenzie A. 2010. Neuronal apoptosis inhibitory protein, NAIP, is an inhibitor of procaspase-9. Int J Biochem Cell Biol. 42:958–964. [DOI] [PubMed] [Google Scholar]
  10. Degtyarev S, Boykova T, Grishanin A, Belyakin S, Rubtsov N, Karamysheva T, Makarevich G, Akifyev A, Zhimulev I. 2004. The molecular structure of the DNA fragments eliminated during chromatin diminution in Cyclops kolensis. Genome Res. 14:2287–2294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Drouin G. 2006. Chromatin diminution in the copepod Mesocyclops edax: diminution of tandemly repeated DNA families from somatic cells. Genome 49:657–665. [DOI] [PubMed] [Google Scholar]
  12. Esposito T, Sampaolo S, Limongelli G, Varone A, Formicola D, Diodato D, Farina O, Napolitano F, Pacileo G, Gianfrancesco F, et al. 2013. Digenic mutational inheritance of the integrin alpha 7 and the myosin heavy chain 7B genes causes congenital myopathy with left ventricular non-compact cardiomyopathy. Orphanet J Rare Dis. 8:91.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fratta E, Coral S, Covre A, Parisi G, Colizzi F, Danielli R, Nicolay HJ, Sigalotti L, Maio M. 2011. The biology of cancer testis antigens: putative function, regulation and therapeutic potential. Mol Oncol. 5:164–182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gakovic M, Shu X, Kasioulis I, Carpanini S, Moraga I, Wright AF. 2011. The role of RPGR in cilia formation and actin stability. Hum Mol Genet. 20:4840–4850. [DOI] [PubMed] [Google Scholar]
  15. Goday C, Esteban MR. 2001. Chromosome elimination in sciarid flies. Bioessays 23:242–250. [DOI] [PubMed] [Google Scholar]
  16. Goto Y, Kubota S, Kohno S. 1998. Highly repetitive DNA sequences that are restricted to the germ line in the hagfish Eptatretus cirrhatus: a mosaic of eliminated elements. Chromosoma 107:17–32. [DOI] [PubMed] [Google Scholar]
  17. Huang da W, Sherman BT, Lempicki RA. 2009a. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 37:1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Huang da W, Sherman BT, Lempicki RA. 2009b. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 4:44–57. [DOI] [PubMed] [Google Scholar]
  19. Itoh Y, Kampf K, Pigozzi MI, Arnold AP. 2009. Molecular cloning and characterization of the germline-restricted chromosome sequence in the zebra finch. Chromosoma 118:527–536. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Janic A, Mendizabal L, Llamazares S, Rossell D, Gonzalez C. 2010. Ectopic expression of germline genes drives malignant brain tumor growth in Drosophila. Science 330:1824–1827. [DOI] [PubMed] [Google Scholar]
  21. Koticha D, Babiarz J, Kane-Goldsmith N, Jacob J, Raju K, Grumet M. 2005. Cell adhesion and neurite outgrowth are promoted by neurofascin NF155 and inhibited by NF186. Mol Cell Neurosci. 30:137–148. [DOI] [PubMed] [Google Scholar]
  22. Kubota S, Takano J, Tsuneishi R, Kobayakawa S, Fujikawa N, Nabeyama M, Kohno S. 2001. Highly repetitive DNA families restricted to germ cells in a Japanese hagfish (Eptatretus burgeri): a hierarchical and mosaic structure in eliminated chromosomes. Genetica 111:319–328. [DOI] [PubMed] [Google Scholar]
  23. Leng N, Dawson JA, Thomson JA, Ruotti V, Rissman AI, Smits BM, Haag JD, Gould MN, Stewart RM, Kendziorski C. 2013. EBSeq: an empirical Bayes hierarchical model for inference in RNA-seq experiments. Bioinformatics 29:1035–1043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Li B, Dewey CN. 2011. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics 12:323.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lin L, Liu Y, Zhao W, Sun B, Chen Q. 2014. Wnt5A expression is associated with the tumor metastasis and clinical survival in cervical cancer. Int J Clin Exp Pathol. 7:6072–6078. [PMC free article] [PubMed] [Google Scholar]
  26. Maglott D, Ostell J, Pruitt KD, Tatusova T. 2011. Entrez Gene: gene-centered information at NCBI. Nucleic Acids Res. 39:D52–D57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Margolis RL, Abraham MR, Gatchell SB, Li SH, Kidwai AS, Breschel TS, Stine OC, Callahan C, McInnis MG, Ross CA. 1997. cDNAs with long CAG trinucleotide repeats from human brain. Hum Genet. 100:114–122. [DOI] [PubMed] [Google Scholar]
  28. Mi H, Poudel S, Muruganujan A, Casagrande JT, Thomas PD. 2016. PANTHER version 10: expanded protein families and functions, and analysis tools. Nucleic Acids Res. 44:D336–D342. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Millane RC, Kanska J, Duffy DJ, Seoighe C, Cunningham S, Plickert G, Frank U. 2011. Induced stem cell neoplasia in a cnidarian by ectopic expression of a POU domain transcription factor. Development 138:2429–2439. [DOI] [PubMed] [Google Scholar]
  30. Mitsui K, Nakajima D, Ohara O, Nakayama M. 2002. Mammalian fat3: a large protein that contains multiple cadherin and EGF-like motifs. Biochem Biophys Res Commun. 290:1260–1266. [DOI] [PubMed] [Google Scholar]
  31. Muller F, Walker P, Aeby P, Neuhaus H, Felder H, Back E, Tobler H. 1982. Nucleotide sequence of satellite DNA contained in the eliminated genome of Ascaris lumbricoides. Nucleic Acids Res. 10:7493–7510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Murga-Zamalloa C, Swaroop A, Khanna H. 2010. Multiprotein complexes of Retinitis Pigmentosa GTPase regulator (RPGR), a ciliary protein mutated in X-linked Retinitis Pigmentosa (XLRP). Adv Exp Med Biol. 664:105–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nelson WJ, Nusse R. 2004. Convergence of Wnt, beta-catenin, and cadherin pathways. Science 303:1483–1487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Niedermaier J, Moritz KB. 2000. Organization and dynamics of satellite and telomere DNAs in Ascaris: implications for formation and programmed breakdown of compound chromosomes. Chromosoma 109:439–452. [DOI] [PubMed] [Google Scholar]
  35. Nikitina N, Bronner-Fraser M, Sauka-Spengler T. 2009. Culturing lamprey embryos. Cold Spring Harb Protoc. 2009:pdb prot5122. [DOI] [PubMed] [Google Scholar]
  36. Pigozzi MI, Solari AJ. 1998. Germ cell restriction and regular transmission of an accessory chromosome that mimics a sex body in the zebra finch, Taeniopygia guttata. Chromosome Res. 6:105–113. [DOI] [PubMed] [Google Scholar]
  37. Polakis P. 2000. Wnt signaling and cancer. Genes Dev. 14:1837–1851. [PubMed] [Google Scholar]
  38. Rozen S, Skaletsky H. 2000. Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol. 132:365–386. [DOI] [PubMed] [Google Scholar]
  39. Sambrook J, Russell DW, Sambrook J. 2006. The condensed protocols from molecular cloning: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. [Google Scholar]
  40. SAS Institute Inc. C, NC. 1989–2007. JMP®. Version 11.
  41. Shen H, Damcott C, Shuldiner SR, Chai S, Yang R, Hu H, Gibson Q, Ryan KA, Mitchell BD, Gong DW. 2011. Genome-wide association study identifies genetic variants in GOT1 determining serum aspartate aminotransferase levels. J Hum Genet. 56:801–805. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Smith JJ, Antonacci F, Eichler EE, Amemiya CT. 2009. Programmed loss of millions of base pairs from a vertebrate genome. Proc Natl Acad Sci U S A. 106:11212–11217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Smith JJ, Baker C, Eichler EE, Amemiya CT. 2012. Genetic consequences of programmed genome rearrangement. Curr Biol. 22:1524–1529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Smith JJ, Kuraku S, Holt C, Sauka-Spengler T, Jiang N, Campbell MS, Yandell MD, Manousaki T, Meyer A, Bloom OE, et al. 2013. Sequencing of the sea lamprey (Petromyzon marinus) genome provides insights into vertebrate evolution. Nat Genet. 45:415–421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Smith JJ, Stuart AB, Sauka-Spengler T, Clifton SW, Amemiya CT. 2010. Development and analysis of a germline BAC resource for the sea lamprey, a vertebrate that undergoes substantial chromatin diminution. Chromosoma 119:381–389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Tanaka K, Miyamoto N, Shouguchi-Miyata J, Ikeda JE. 2006. HFM1, the human homologue of yeast Mer3, encodes a putative DNA helicase expressed specifically in germ-line cells. DNA Seq 17:242–246. [DOI] [PubMed] [Google Scholar]
  47. Tominaga K, Kondo C, Kagata T, Hishida T, Nishizuka M, Imagawa M. 2004. The novel gene fad158, having a transmembrane domain and leucine-rich repeat, stimulates adipocyte differentiation. J Biol Chem. 279:34840–34848. [DOI] [PubMed] [Google Scholar]
  48. Wang J, Davis RE. 2014. Programmed DNA elimination in multicellular organisms. Curr Opin Genet Dev. 27:26–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wang H, Liu ZD. 2014. Genetic association between AGPHD1 variant and lung cancer risk. Cell Biochem Biophys. 70:1963–1968. [DOI] [PubMed] [Google Scholar]
  50. Wang J, Mitreva M, Berriman M, Thorne A, Magrini V, Koutsovoulos G, Kumar S, Blaxter ML, Davis RE. 2012. Silencing of germline-expressed genes by DNA elimination in somatic cells. Dev Cell 23:1072–1080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Yartseva V, Giraldez AJ. 2015. The Maternal-to-Zygotic transition during vertebrate development: a model for reprogramming. Curr Top Dev Biol. 113:191–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Yi M, Liao H, Huang S, Xie X, Luo G. 2007. Expression of cancer-testis antigen SCP-1 mRNA in human nasopharyngeal carcinoma. Lin Chung Er Bi Yan Hou Tou Jing Wai Ke Za Zhi. 21:343–345. [PubMed] [Google Scholar]
  53. Zhang C, Chen Y, Wang M, Chen X, Li Y, Song E, Liu X, Kim S, Peng H. 2014. PPM1D silencing by RNA interference inhibits the proliferation of lung cancer cells. World J Surg Oncol. 12:258.. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_33_9_2337__index.html (1KB, html)
supp_msw104_Supplementary_Table_S1_042716.xlsx (125.4KB, xlsx)
supp_msw104_Supplementary_Figures_SABR_MBE.pdf (16.1MB, pdf)

Articles from Molecular Biology and Evolution are provided here courtesy of Oxford University Press

RESOURCES