Research article

Transcriptome architecture across tissues in the pig

André LJ Ferraz^1,2 , Ana Ojeda¹ , Manel López-Béjar³ , Lana T Fernandes⁴ , Anna Castelló¹ , Josep M Folch¹ and Miguel Pérez-Enciso^1,5

¹Departament de Ciència Animal i dels Aliments, Facultat de Veterinària, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain

²Faculdade de Ciências Agrárias e Veterinária, Universidade Estadual Paulista (UNESP), 14884-900 Jaboticabal – SP, Brazil

³Departament de Sanitat i d'Anatomia Animals, Facultat de Veterinària, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain

⁴Centre de Recerca en Sanitat Animal (CReSA), Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain

⁵Institut Català de Recerca i Estudis Avançats (ICREA), C/Lluis Companys 23, 08010 Barcelona, Spain

author email corresponding author email

BMC Genomics 2008, 9:173doi:10.1186/1471-2164-9-173

Published:	16 April 2008

Abstract

Background

Artificial selection has resulted in animal breeds with extreme phenotypes. As an organism is made up of many different tissues and organs, each with its own genetic programme, it is pertinent to ask: How relevant is tissue in terms of total transcriptome variability? Which are the genes most distinctly expressed between tissues? Does breed or sex equally affect the transcriptome across tissues?

Results

In order to gain insight on these issues, we conducted microarray expression profiling of 16 different tissues from four animals of two extreme pig breeds, Large White and Iberian, two males and two females. Mixed model analysis and neighbor – joining trees showed that tissues with similar developmental origin clustered closer than those with different embryonic origins. Often a sound biological interpretation was possible for overrepresented gene ontology categories within differentially expressed genes between groups of tissues. For instance, an excess of nervous system or muscle development genes were found among tissues of ectoderm or mesoderm origins, respectively. Tissue accounted for ~11 times more variability than sex or breed. Nevertheless, we were able to confidently identify genes with differential expression across tissues between breeds (33 genes) and between sexes (19 genes). The genes primarily affected by sex were overall different than those affected by breed or tissue. Interaction with tissue can be important for differentially expressed genes between breeds but not so much for genes whose expression differ between sexes.

Conclusion

Embryonic development leaves an enduring footprint on the transcriptome. The interaction in gene × tissue for differentially expressed genes between breeds suggests that animal breeding has targeted differentially each tissue's transcriptome.