In a recent study published in the eBioMedicine, researchers examined the effect of serum levels of 20 amino acids in a mother during pregnancy on offspring birthweight in a two-sample summary data Mendelian randomization (MR) framework.
Studies have estimated that adequate fetal growth requires between 10 and 60 grams (g)/day per kilogram (kg) fetus weight of amino acids. In vivo studies in humans have shown that amino acids are essential for protein synthesis and the modulation of multiple cell signaling pathways.
Since the interactions between maternal, placental, and fetal mechanisms are complex, there is no insight into how the fetus receives different amino acids from the mother. Moreover, data are scarce from human studies on how maternal amino acids influence fetal growth and development. All previous studies have fetched inconsistent results of amino acid supplementation in high-risk pregnancies.
It is also noteworthy that previous studies have found that branched-chain amino acids (BCCAs), including valine, leucine, and isoleucine, cross the placenta more rapidly using sodium-independent L system, and their high concentrations in maternal serum lead to a higher risk of intrauterine growth-restricted pregnancies. Previous studies have used the MR approach to confirm the causal effect of maternal smoking during pregnancy on slower fetal growth.
About the study
In the present MR study, researchers used maternal single nucleotide polymorphisms (SNPs) as instrumental variables (IVs) to infer the cause-and-effect relationship between genetically influenced intrauterine exposures (in this case, maternal circulating amino acids) and offspring birthweight.
They used data from recent metabolites and offspring birthweight genome-wide association studies (GWAS) encompassing up to 86,507 and 406,063 participants, respectively. Further, the study dataset comprised 20 amino acids that make proteins in the human body. Examples include glutamate, glutamine, leucine, and serine.
The team used the weighted linear model adjusted (WLM-adjusted) analyses, an approximation of the structural equation modeling (SEM) approach developed by Warrington et al., to adjust for offspring genetic effects in the summary data used to evaluate the association of maternal genetic variants on offspring birthweight. Likewise, they estimated associations between genetic variants and amino acids.
The researchers selected maternal SNPs strongly correlated with 20 different circulating amino acids from these GWASs validated in 2966 pregnant women enrolled in the Born in Bradford (BiB) study and 4407 women in the Fenland study.
In the BiB study, they measured amino acids as part of a nuclear magnetic resonance (NMR) metabolomics analysis at 24 to 28 weeks of gestation. It yielded data for nine amino acids in 2966 European women. Further, they selected SNPs from another GWAS conducted among women and men.
Next, the team natural log-transformed amino acid levels, winsorized at five standard deviations (SDs) and transformed to Z scores, then adjusted for maternal age and top 10 principal components (PCs) from genomic data, e.g., minor allele frequency (MAF). Then, they regressed each resulting residual against the corresponding SNP used in the primary MR analysis. This exercise yielded 89 SNP-amino acid associations.
The study results suggested that while maternal serum glutamine and serine levels positively affected offspring birthweight, leucine, and phenylalanine had a negative effect. Several sensitivity analyses accounting for bias due to violation of MR assumptions also supported these findings, although, for some amino acids, these estimates might be imprecise.
Further, MR analyses suggested that the positive effect of maternal glutamine, a non-essential amino acid that becomes conditionally essential as fetal demand surpasses maternal synthesis, on offspring birthweight might be liver-type isoenzyme dependent, specifically, when instrumented by the missense variant rs2657879 in glutaminase 2 (GLS2).
While the genetic variant rs2657879 had a strong positive effect, the effect instrumented by rs7587672, an expression quantitative trait loci (eQTL) for GLS, this effect turned inverse. Note that GLS2 encodes the catalyst that helps convert glutamine to ammonia and glutamate in the liver, whereas GLS encodes the kidney-type isozyme.
During the late phase of pregnancy, glycine amino acid supplements one carbon for the synthesis and methylation of deoxyribonucleic acid (DNA) necessary for fetal growth. However, there is evidence that glycine supply remains lower than fetal demand because of its inadequate transportation across the human placenta. Thankfully, maternal circulating serine, not transported to the fetal circulation, is then used within the uteroplacental tissues to synthesize glycine, thus, contributing to the fetal glycine supply.
The current study findings support the above hypothesis that maternal circulating serine has a causal effect on offspring birthweight while glycine does not. However, results based on two SNPs, rs561931 of phosphoglycerate dehydrogenase (PHGDH) and rs4947534 of phosphoserine phosphatase (PSPH), suggested positive effects. Note these two gene loci encode enzymes involved in serine de novo biosynthesis. Further exploration of this interlinked serine and glycine metabolism using multivariable MR might be valuable if performed using large sample sizes.
Studies have proposed that leucine amino acid modulates fetal muscle protein synthesis through mammalian target of the rapamycin (mTOR) signaling pathway. Accordingly, maternal circulating leucine had an inverse effect on offspring birthweight in the primary MR analysis. In contrast to findings of previous studies, researchers found higher maternal circulating BCAAs had a negative association with offspring birthweight.
However, owing to the close link between BCAA metabolism and insulin resistance, there is a need for more MR studies in larger sample sizes to delineate the impact of maternal fasting insulin and BCAAs circulating in the mother’s serum on offspring birthweight. Finally, the primary MR analysis demonstrated an inverse effect of phenylalanine on offspring birthweight, supported by similar though imprecise effect estimates of sensitivity analyses.
To summarize, the study results indicated that genetically predicted levels of glutamine and serine in maternal serum elevate offspring birthweight, while leucine and phenylalanine decrease it.
According to the authors, they used the largest GWAS available. Yet, they failed to robustly estimate several potentially significant clinical effects, such as, of the alanine amino acid. Thus, they emphasized conducting larger GWAS of amino acids and offspring birthweight to replicate their findings.
In addition, future studies should explore mechanisms underlying these effects, especially how amino acids get transferred across the placenta and the role of fetal genotypes in placental transmission. Most importantly, randomized controlled trials should establish whether supplementing maternal circulating amino acids during pregnancy could help optimize fetal growth.