Milk Protein Gene Expression is Related to Chromatin Conformation, Chromatin Loop Attachment to Nuclear Structure and Nuclear Localization of Genes
Maria Ballester1, Clémence Kress1, Mohammad Bagher Montazer Torbati1, Cathy Hue-Beauvais1, Kiên Kiêu3, Gaëtan Lehmann4, Stéphanie Droineau1, Monique Rijnkels2, Eve Devinoy1 1UR1196-Génomique et Physiologie de la Lactation, INRA, 78352 Jouy en josas Cedex, France ; E-mail : email@example.com, Phone : 33 1 34 65 29 21, Fax : 33 1 34 65 22 41 2Children's Nutrition Research Center (CNRC) - Baylor College of Medicine, 1100 Bates Street, Houston TX, USA 3UR341-Mathématiques et Informatique Appliquées, INRA, 78352 Jouy en Josas Cedex, France 4UR1198-Biologie du Développement et de la Reproduction, INRA, 78352 Jouy en Josas Cedex, France
Milk production is partly regulated through milk protein gene
expression, which is specific to the mammary gland and peaks during
lactation. This regulation is controlled by lactogenic hormones which
activate the Jak2 Stat5 transduction pathway leading to the interaction
of Stat5 with response elements located upstream of these genes.
For one milk protein gene chosen as a model, the Whey Acidic Protein (WAP) gene, our results show that the expression level of the gene during the pregnancy – lactation cycle, in the mouse and the rabbit, is related to a more or less open chromatin conformation at three regions located around 1, 3 and 6 kb upstream from the gene (-1, -3 and -6kb, respectively), and that the non-expression of the gene in the liver is related to a closed conformation. The chromatin is in an open conformation further upstream and downstream in both tissues,. This may allow the expression of the two genes (ramp3 and Tbrg4) which flank the WAP and are constitutively expressed in the mouse. The DNA methylation profile from -6 to +4 kb, depends on the tissue, DNA being methylated to a low level in the mammary gland but highly methylated in the liver.
At the border of this -6 to +4 kb region, two regions of attachment to the nuclear matrix have been predicted. Their sequences exhibit similarity with the consensus sequence of SATB1, which has not been detected in the mammary gland. These two regions might explain the organization of the chromatin in different domains, forming different loops according to the tissue. A third predicted MAR located near the first exon of Tbrg4 interacts with the topoisomerase II, independently of hormonal treatment in the HC11 mouse mammary cells. This region is therefore always in interaction with nuclear structure. On the contrary the attachment of the chromatin loops surrounding the WAP gene as well as casein gene to nuclear structure varies, in HC11 cells, with hormonal treatments and is therefore also related to gene expression.
In HC11 cells, as well as in non expressing cells (mouse embryonic fibroblasts: 3T6), the analysis of nuclei having with intact three dimensional structure shows that the WAP gene is located more within the inside of the nucleus, whereas the casein genes are located close to the nuclear periphery. However, in mouse mammary cells a significant relocation of these genes occurs during lactogenic hormone treatments, with a movement of the casein gene towards the interior and an inverse movement of the WAP gene towards the nuclear periphery. Furthermore, the position of the casein genes relative to their chromosome territory is modified, the casein genes being found more frequently outside of its territory in casein gene expressing cells. These results are new breakthroughs in understanding how milk component expression is controlled. They will be discussed.