Register      Login
Reproduction, Fertility and Development Reproduction, Fertility and Development Society
Vertebrate reproductive science and technology
RESEARCH ARTICLE

144 DOSAGE COMPENSATION AND X-LINKED GENE EXPRESSION IN BOVINE IN VIVO AND IN VITRO EMBRYOS

J. E. Duan A , N. K. Jue B , Z. Jiang A , R. O’Neill B , E. Wolf C , L. A. Blomberg D , H. Dong E , X. Zheng E , J. Chen E and X. C. Tian A
+ Author Affiliations
- Author Affiliations

A Department of Animal Science, University of Connecticut, Storrs, CT, USA;

B Department of Molecular and Cell Biology, University of Connecticut, Storrs, CT, USA;

C Laboratory for Functional Genome Analysis, Ludwig-Maximilians-Universität Munchen, Munich, Germany;

D Animal Biosciences and Biotechnology Laboratory, USDA, Beltsville, MD, USA;

E Institute of Animal Science, Xinjiang Academy of Animal Science, Urumqi, Xinjiang, P. R. China

Reproduction, Fertility and Development 28(2) 202-202 https://doi.org/10.1071/RDv28n2Ab144
Published: 3 December 2015

Abstract

In human and mouse diploid cells and gametes, expression levels of X-linked genes are hypothesised to balance with those of autosomal genes (Ohno’s “dosage compensation”). Such a phenomenon, however, has not been systematically studied in cattle or compared between in vivo and in vitro embryos. Using RNA-seq data, we compared dosage compensation and expression differences of X-linked genes in bovine in vitro and in vivo oocytes and embryos. RNA-seq datasets GSE59186 and GSE52415 were non-uniquely (paralogs included) mapped to the bovine reference genome assembly UMD3.1 using tophat2. Cufflinks v1.0.3 was used to estimate fragments per kb of exon per million fragments (FPKM), which were then log2-transformed. In order to assess overall patterns of chromosomal gene expression without bias, statistical outliers were removed. A total of 12 928 X-linked transcripts were used to calculate the relative X to autosomal gene (A) expression (RXE): log2 (X expression) – log2 (A expression) for dosage compensation. Values ≥0 indicate dosage compensation (or X : A ratio ≥ 1); values <0 indicate incomplete dosage compensation; value = –1 indicates no dosage compensation (or X : A ratio = 0.5). Cuffdiff was used to identify differentially expressed genes between stages and the 2 datasets. Cuffnorm normalized FPKM for expression patterns, which were further clustered and graphed in R. Expression pattern distributions across regions on X were calculated by merging Cuffnorm output genes.attr_table to the expression pattern lists. The RXE values were higher than –1 in all embryonic stages studied, with in vitro embryos having higher RXE, suggesting some but incomplete dosage compensation and in vitro embryos exhibiting higher levels of X-gene expression. In vitro-produced immature oocytes and 4-cell embryos had RXE of 0, suggesting that both may be completely compensated. Additionally, embryos of the 2 sources exhibited significant differences in levels of X-gene expression at 4-cell to 8-cell and 16-cell to blastocyst stages. All expressed X-linked genes fell in 5 groups (patterns 1–5): increased, decreased, increased and then decreased, constant, decreased and then increased. Patterns 1 to 4 were seen in both datasets, whereas pattern 5 was only present in vitro, which may be why in vitro embryos had higher dosage of X. We further found that these expression patterns correlated with the genes’ proximity to the X inactivation centre (Xic: ~82 Mb): the highly dynamic patterns 3 and 5 were associated with proximity to Xic, supporting the notion that the Xist RNA spreads along the X through the Xic. In vivo and in vitro bovine oocytes and embryos have undergone some degree but incomplete dosage compensation. Expression of X-linked genes is correlated with their proximity to Xic. Globally, in vivo and in vitro embryos exhibit major differences in levels of X-gene expression.