Fetal growth is determined by interactions between fetal genes and the maternal uterine environment. We are performing a genome wide association study (GWAS) to address the hypothesis that gene-environment interactions in the context of the maternal-fetal unit impact fetal size at birth and maternal metabolism. This is being accomplished using Caucasian, Hispanic, and Afro-Caribbean DNA samples collected from mothers and offspring as part of the NIH-funded Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study. HAPO is a multicenter, international study in which high quality phenotypic data related to fetal growth and maternal glucose metabolism was collected from 25,000 pregnant women of varied racial backgrounds using standardized protocols that were uniform across centers. A major strength of HAPO is the large number of subjects available for replication. In addition to the samples analyzed in the initial GWAS, there are additional cohorts of Caucasian, Asian and Israeli Arab/Bedouin and Jewish mothers and babies which are available for replication of the initial GWAS findings. Importantly, phenotype data related to fetal growth and maternal metabolism was collected from these mothers and babies using standardized protocols that were uniform across centers and identical to those used to phenotype subjects in the initial GWAS. In this initial replication study, we intend to focus efforts on a phenotype for which: (i) there are not other ongoing, large GWAS efforts and (ii) there are large replication cohorts from the same ethnic background available. Thus, this study aims to replicate the contribution of maternal SNPs to offspring birth weight and the contribution of fetal SNPs before and after correction for maternal genotype to offspring birth weight. Importantly, our proposal takes advantage of one of the unique strengths of HAPO, access to GWAS data from both mothers and their offspring, and we will be able to consider the contribution of both offspring and maternal genetic variation to the regulation of birth weight. We intend to use a two-stage replication approach by including an additional large cohort of Caucasian subjects from the UK who were recruited under the auspices of the Exeter Family Study of Childhood Health (EFSOCH) and ALSPAC. This project has the following specific aims.

1. To perform a two stage replication of the GWAS results from maternal SNP vs. offspring birth weight associations. We will take 115 SNPs into 2,556 Stage 1 maternal replication samples and a further 15 SNPs into 11,000 Stage 2 maternal replication samples and perform a joint analysis on greater than 15,000 mothers.

2. To perform a two stage replication of the GWAS results from fetal SNPs vs. offspring birth weight associations, including analyses corrected for maternal genotype. We will take 115 SNPs into 2,669 Stage 1 offspring replication samples and a further 15 SNPs into 11,000 Stage 2 offspring replication samples and perform a joint analysis on greater than 15,000 offspring.

The first aim of our study is to investigate, genome-wide, the association between maternal genotype and offspring birth weight. This aim does not overlap with current aims of the EGG consortium.

Since we are also interested in the association between fetal genotype and offspring birth weight, there is some overlap between our proposed analyses and the current GWAS meta-analysis (including 1500 ALSPAC offspring) being carried out by the EGG consortium. However, our proposed analysis is distinct from these efforts because we will adjust for maternal genotype at the genome-wide stage and follow up hits based on those data. To our knowledge, other than in HAPO, there are no existing genome-wide association studies with both maternal and fetal genotype data on large numbers of subjects. It is important to point out that whereas the proposal amendment for B768 asks for the genotyping of top hits from a fetal-only genome-wide study in EGG to test in replication samples whether these top hit are independent of maternal genotype, our proposal is to follow up hits from a genome-wide study that is already adjusted for maternal genotype.

Another feature of HAPO that is distinct from the current EGG collaboration is that genome-wide data are available on different ethnic groups: Afro-Caribbean and Hispanic, in addition to European. This may lead to the prioritisation of different top hits for follow-up, as was seen with the recent discovery of the KCNQ1 gene for type 2 diabetes (Unoki et al and Yasuda st al, 2008 Nat Genet).

Recent successes from genome-wide association studies have highlighted the need for collaboration between centers to provide sufficient power to detect modest effects of common genetic variants. For early growth traits, efforts are already underway in the EGG consortium to ahieve this for fetal genotype vs. birth weight. We acknowledge that the current combined genome-wide fetal genotype sample in EGG exceeds our own fetal genotype sample, and that the contribution of HAPO may be better placed in the wider consortium in regard to this aim. Like ALSPAC, the HAPO group is keen to participate, where possible, in the EGG consortium and the emerging EAGLE consortium once we have examined our own genome-wide data.

In Stage 1 of the planned replication study, we will genotype 115 offspring SNPs for size at birth in 2,669 additional Caucasian subjects from the HAPO study as well as 115 maternal SNPs for size at birth in 2556 Caucasian HAPO mothers. The genotypes from the Stage 1 replication study will be analyzed together with the genotypes from the 1500 mothers and offspring analyzed in the original GWAS. From these analyses, we will identify the 15 most significant offspring SNPs for birth weight and 15 most significant maternal SNPs for offspring birth weight. These will then be typed in the ALSPAC mothers and offspring. Thus, from ALSPAC, we will need offspring birth weight data and additional data that will need to be controlled for in the analyses (gestational age of the neonate at the time of delivery, neonatal gender, parity, maternal age, BMI, and blood pressure). We will also need genotype data. As noted, the SNPs to be genotyped will be determined from analysis of the GWAS and Stage 1 replication study. We are planning that the ALSPAC samples will be genotyped at Kbiosciences. We will pay for the genotyping using the resources of the grant from the NIH.