2012 Annual Report
1a.Objectives (from AD-416):
Objective 1: Determine gene expression in human lactating mammary epithelium.
Subobjective 1A: Determine the pattern of mammary epithelial gene expression using milk fat globule mRNA from delivery through the first 4 weeks of lactation. Compare these results with those in mothers of premature infants and teenage mothers over a similar period of time.
Subobjective 1B: Characterize the mRNA response to exogenous lactogenic hormones.
Objective 2: Characterize inbred mouse strains for lactation performance, gene expression and weight gain among offspring in lean and obese animals, making use of a cross-fostering design where appropriate.
Subobjective 2A. Identify genes in which strain-dependent differences in mammary gland gene expression, and SNP haplotype, are correlated with strain-dependent differences in milk production, lactation persistence, mammary gland development, or milk composition.
Subobjective 2B. Determine the extent to which genes identified from the whole genome scan and microarray work described in 2A are responsible for the lactation defect in mice with maternal obesity.
Objective 3: Study the effect of nutrients on mammary gland development and function in mouse models. Define the critical window for effects on mammary gland development and function.
Subobjective 3A1. Determine effect of exposure to low protein diet by analyzing mammary gland development, milk production, and milk composition, as well as gene expression and gene promoter methylation in mammary gland tissue of dams exposed to diets with low protein content during gestation.
Subobjective 3A2. Use a mouse model for tissue-specific alteration of Dnmt1 levels to confirm role of DNA methylation in effects of low protein diet on mammary gland development.
Subobjective 3B. Define critical window for effect of low protein diet on mammary gland function by limiting nutritional intervention to specific developmental windows.
Subobjective 3C. Determine impact of low protein diet on genetic variants for mammary gland development and lactation capacity as identified in objective 2.
1b.Approach (from AD-416):
Children's Nutrition Research Center researchers will determine gene expression in human lactating mammary epithelium by isolating mRNA from human colostrum or milk over the first 4 weeks post partum and the expression arrays measured to determine the relative gene express over this period of time. Data from groups of mothers will be assessed to prove or disprove our hypotheses. A variety of potential lactogenic hormones will be administered short term (over 3 days) to normal women with established lactation between 6 and 12 weeks post partum. The hormones initially to be tested are prolactin, cortisol, and IGF-1. Breast milk will be collected every 3 hr and RNA isolated for measurement of expression of mRNA expression using microchip technology. The data will be compared to that already obtained from similar studies in women prior to and following the administration of recombinant human growth hormone. Additionally, a panel of lactation traits will be measured in 32 inbred strains of mice. The data from these measurements will be used as phenotype data in combination with whole genome SNP data to conduct a statistical association analysis across the entire mouse genome. The Viable yellow agouti (Avy) mouse will be used as a model of maternal obesity. Gene expression will be determined by microarray analysis of mammary tissue samples collected from obese and lean Avy females during early lactation. Genes that are differentially expressed between lean and obese females will then be compared to the list of genes identified to test for overlap. The lactation traits, as well as gene expression and epigenetic profiles will be measured in transgenic animals containing the conditional allele for Dnmt1 (dnmt1-lox2) and a mammary gland specific Cre recombinase to determine the effects of deletion in the mammary gland of Dnmt1. The data will be compared to those of low protein diets.
For Objective 1, this year we completed collection and initial analysis of the milk samples obtained from the normal mothers from delivery through 6 weeks post partum and have carried out the initial determination of the messenger RNA in the milk (mRNA carries the messages as to which proteins should be made, expression array analyses) but have only begun to do the more difficult job of interpreting massive amount of data generated from these 7 volunteers. We initially focused our attention on galactose (a component of milk sugar) and lactose (milk sugar). We observed that the genes involved in the synthesis of a compound UDP-galactose and its transport are low at birth and increase dramatically by 96 hours postpartum, and suggests that these pathways maybe rate limiting for milk sugar (lactose) production, a major determinant of milk volume. The collection of samples from mothers with premature infants, teenage mothers, and obese mothers may sound simple and straight forward but we have had difficulty in recruiting appropriate subjects. We have completed studies on 6 obese women, 2 teen mothers, and 1 mother with a premature infant. An additional teenager is in the midst of sample collection. We have begun processing RNA samples from the obese mothers in preparation to performing the expression array analyses. We recruited 22 additional women but at delivery they did not qualify for the study because of C-section, no milk production, dropped out, or had medical complications. We have 13 additional individuals who are awaiting delivery. To accomplish this we have reviewed over 815 charts at Ben Taub General and St. Luke's Hospitals and at two large obstetrical clinics in the Houston area.
We have determined that our objective 1B will not be possible to execute. Despite a publication demonstrating the use of human prolactin by outside researchers several years ago, we have found no company willing to provide us with rh-prolactin for human study since it has no profitable commercial use. Without prolactin, the comparisons with glucocorticosteriod and insulin become of lesser importance. We have moved to the measurement of micro RNAs in human milk. Mirco-RNAs are small pieces of RNA containing 18 to 30 nucleotides (or building blocks of RNA) and are thought to play a unique regulatory role in the turning on or off of genes and thus the production of specific protein. We have submitted a manuscript describing a number of unique and newly found miRNAs. Their function(s), if any, remain to be determined.
In objective 2, during the year we completed the analysis of litter weight gain data from 32 inbred strains of mice as a means of determining the variation in milk production in these strains over the first 8 days of lactation. This data was used to map regions in the mouse genome based on single nucleotide polymorphism data that we obtained from the mouse haplotype map resource (http://www.broadinstitute.org/mouse/hapmap/). We performed follow-up analysis of gene expression for our top candidates in the mammary tissue samples that were collected at day 10 postpartum. We collected additional samples of brain and mammary tissue from day 1 postpartum mice at the extremes of the litter gain distribution. These samples will be used in subsequent analysis of gene expression. We also measured the concentrations of 9 minerals in milk samples collected from the 32 strains. We used this data to map genomic regions in the mouse that are associated with variations in milk mineral concentrations. We collected whole mammary biopsies from inbred mice of 39 inbred strains at two key times during postnatal mammary gland development. Through image analysis we collected quantitative data describing mammary ductal development and then used this data to map genomic regions in the mouse that are associated with variations in mammary ductal development. We conducted a comparison of mammary gene expression among lean and obese mice during early lactation to determine if obesity-dependent lactational insufficiency could be linked to changes in specific pathways. For this analysis we used both micro-array analysis on whole tissue and laser-captured mammary epithelium, and real-time quantitative PCR. Lastly we completed an analysis or mammary tissue gene expression during a frequent sampling period of 3 days that was begun at mid lactation. In this analysis we followed expression patterns of known circadian pathway genes and genes necessary to galactose metabolism and demonstrated that some of these genes are regulated by the frequency of nursing.
In objective 3, we completed the phenotypic analysis of Dnmt-flox|WapCre mice. Unexpectedly cells that should harbor the deleted Dnmt1 allele do not persist in the mammary gland into old age (1.5 years old for mice). This brings to light questions about: cell homeostasis in the mammary gland and the role that Dnmt1 and DNA methylation might play in this; dependence on continued ovarian hormone or other signaling for certain cell populations to survive in the mammary gland; and the role of Dnmt1 and DNA methylation in the aging mammary gland. We are still working on conditions to determine accurately the deletion of Dnmt1 from Cre exposed tissues. We are currently breeding animals with the appropriate genotype into the genetic strain background needed for follow-up studies. These follow-up studies include serial transplantation of Dnmt1 deficient mammary tissue to determine effect of Dnmt1 deletion on stem, progenitor, and functionally differentiated cell populations, and hormonal manipulations to determine the role of ovarian hormones on Dnmt1 expression/activity and DNA methylation in the mammary gland. We are starting to analyze gene expression in the presumed Dnmt1deficient mammary glands by qPCR. Due to tissue heterogeneity, we anticipate we will have to perform laser capture to isolate mammary gland epithelium to accurately analyze gene expression in potential Dnmt1-defcient mammary epithelium.
Mapping of gene variants underlying maternal nurturing ability. Significant variation exists for maternal nurturing ability in inbred mice. Although classical gene mapping approaches have identified quantitative trait loci (QTL) that may account for this variation, the underlying genes are unknown. Children's Nutrition Research Center researchers in Houston, Texas, performed a study in which lactation performance data was used to map genomic regions associated with this maternal nurturing variation. Our work identified up to 15 regions in the mouse genome containing 13 genes that were associated variations in maternal nurturing ability. Among the strongest candidate genes were a growth factor receptor, the epidermal growth factor receptor, a steroid hormone receptor, the mineralocorticoid receptor, and a GTP binding protein, Guanine nucleotide binding protein G(q). A comparison of these results with genomic variation in other species such as cows and humans would be expected to yield insights into the regulation of lactation in these species as well.
Finding the genes to turn on lactation. A lack of knowledge exists on the human genes that are turned on in the mammary epithelium (cells) at the time of delivery that result in the synthesis of lactose, the primary factor responsible for determining milk volume. Scientists at the Children's Nutrition Research Center in Houston, Texas, conducted studies using unique techniques of molecular biology to determine that the rate- limiting gene involved in the synthesis of lactose are the enzymes responsible for the formation of galactose and the induction of the prolactin receptor, which is a key hormonal factor in the initiation and sustaining of lactation in humans. By knowing the rate- imiting events in the process we may be able to create new therapies that might make the initiation of lactation more uniformly successful in the early hours after delivery. These findings could have an impact in humans and even into the induction of lactation in commercial animals.