1. Genetics and Genomics

Body mass index and adipose distribution have opposing genetic impacts on human blood traits

  1. Christopher S Thom  Is a corresponding author
  2. Madison B Wilken
  3. Stella T Chou
  4. Benjamin F Voight  Is a corresponding author
  1. Division of Neonatology, Children's Hospital of Philadelphia, United States
  2. Division of Hematology, Children's Hospital of Philadelphia, United States
  3. Department of Systems Pharmacology and Translational Therapeutics, University of Pennsylvania - Perelman School of Medicine, United States
  4. Department of Genetics, University of Pennsylvania - Perelman School of Medicine, United States
  5. Institute for Translational Medicine, University of Pennsylvania - Perelman School of Medicine, United States
https://doi.org/10.7554/eLife.75317
Share this article
Cite this article
  1. Christopher S Thom
  2. Madison B Wilken
  3. Stella T Chou
  4. Benjamin F Voight
(2022)
Body mass index and adipose distribution have opposing genetic impacts on human blood traits
eLife 11:e75317.
https://doi.org/10.7554/eLife.75317
  2. Download BibTeX
  3. Download .RIS

Abstract

Body mass index (BMI), hyperlipidemia, and truncal adipose distribution concordantly elevate cardiovascular disease risks, but have unknown genetic effects on blood trait variation. Using Mendelian randomization, we define unexpectedly opposing roles for increased BMI and truncal adipose distribution on blood traits. Elevated genetically determined BMI and lipid levels decreased hemoglobin and hematocrit levels, consistent with clinical observations associating obesity and anemia. We found that lipid-related effects were confined to erythroid traits. In contrast, BMI affected multiple blood lineages, indicating broad effects on hematopoiesis. Increased truncal adipose distribution opposed BMI effects, increasing hemoglobin and blood cell counts across lineages. Conditional analyses indicated genes, pathways, and cell types responsible for these effects, including Leptin Receptor and other blood cell-extrinsic factors in adipocytes and endothelium that regulate hematopoietic stem and progenitor cell biology. Our findings identify novel roles for obesity on hematopoiesis, including a previously underappreciated role for genetically determined adipose distribution in determining blood cell formation and function.

Editor's evaluation

The study shows that genetically determined adiposity plays a previously underappreciated role in determining blood cell formation and function. The authors have clearly spelled out their hypothesis and performed all the relevant and available analyses in the "Mendelian Randomization toolbox". The study will help understand the pathogenesis for clonal hematopoiesis.

https://doi.org/10.7554/eLife.75317.sa0

Introduction

Blood cell homeostasis is achieved through incompletely understood coordination of blood cell-intrinsic gene regulation and blood cell-extrinsic environmental mechanisms (Comazzetto et al., 2021; Ulirsch et al., 2019). The importance of blood cell formation and function in normal hematopoietic development, hematologic diseases, and clinical manifestations of systemic disorders has prompted extensive investigation of loci underlying human blood trait variation through genome-wide association studies (GWAS) (Astle et al., 2016; Chen et al., 2020; Vuckovic et al., 2020). One shortcoming has been an inability to identify extrinsic effects from these data.

Adipocytes and endothelial cells within the bone marrow environment regulate hematopoiesis (Comazzetto et al., 2021; Zhong et al., 2020). Discrete adipocyte populations differentially modulate systemic physiology and homeostasis (Hildreth et al., 2021). For example, white adipose tissue has a derogatory effect on hematopoiesis, whereas mesenchymal-derived bone marrow adipocyte populations support blood cell formation (Comazzetto et al., 2021; Cuminetti and Arranz, 2019; Wang et al., 2018; Zhong et al., 2020).

Single nucleotide polymorphisms (SNPs) that genetically increase body mass index (BMI) also raise metabolic and cardiovascular disease risks (Pulit et al., 2019). Some observational studies have linked obesity (BMI > 30 kg/m2; Aigner et al., 2014) or hypercholesterolemia (Shalev et al., 2007) with anemia. However, others observed apparent erythrocytosis in obese individuals (Keohane et al., 2013). Genetic relationships have not been elucidated for BMI on erythroid or other blood traits. A genetic predisposition to accumulate truncal adipose tissue elevates waist-to-hip ratio (WHR). Like BMI, WHR influences cardiovascular risks (Huang et al., 2021; Pulit et al., 2019), but genetic impacts of WHR on blood traits are unknown.

Our study was designed to test several hypotheses. First, we wanted to determine if increased BMI decreased hemoglobin (HGB) at the genetic level, consistent with higher anemia risk in obese individuals (Aigner et al., 2014). Second, we wanted to identify BMI-related traits and mechanisms responsible for effects on erythroid traits (e.g., HGB level and hematocrit [HCT]). We specifically examined effects of adipose distribution (WHR), which impacts cardiovascular disease risk along with BMI variation (Huang et al., 2021). We also analyzed genetically determined lipid fractions, since hyperlipidemia has been linked with anemia risk (Shalev et al., 2007) and since lipids can impact erythrocyte stability (Mohandas and Gallagher, 2008). Third, we asked if BMI and related traits impacted non-erythroid blood cell lineages. This led us to assess genetic impacts on a total of 15 quantitative blood traits, which generally reflect perturbations in blood cell formation and/or function (Chen et al., 2020; Vuckovic et al., 2020).

We used a Mendelian randomization (MR) framework for our study, anticipating that the results would help to clarify the complex interplay between cardiometabolic traits and hematopoiesis without necessarily revealing clinically apparent effects. MR leverages variants linked to an exposure trait to estimate causal genetic effects on an outcome (Hemani et al., 2018). Multivariable MR (MVMR) and causal mediation analyses can parse effects from multiple factors (Burgess et al., 2017). Interrogating causal effects of BMI and related factors on erythroid and other blood traits revealed unexpected associations between BMI, WHR, and hematopoietic variation. Conditional genome-wide analyses using mtCOJO (Zhu et al., 2018) highlighted blood loci that were substantially influenced by BMI and/or WHR, helping to reveal genes and pathways by which these physiological factors impact blood trait variation.

Results

Causal association between genetically determined BMI and lower HGB level

We hypothesized that BMI would decrease HGB at the genetic level, consistent with clinical observations. Using MR, we found that each standard deviation (SD) unit increase in BMI caused a 0.057 SD decrease in HGB levels by the inverse variance weighted (IVW) method (p = 1.0 × 10–5) that was directionally consistent across sensitivity analyses without evidence of horizontal pleiotropy or weak instrument bias (Figure 1a, Figure 1—figure supplement 1a, Supplementary file 1—Table 1). Similar effects were observed for HCT (Figure 1b and Figure 1—figure supplement 1b), suggesting BMI is genetically linked with reduced HGB.

Figure 1 with 9 supplements see all
Download asset Open asset
Body mass index (BMI) and waist-to-hip ratio (WHR) exert opposing effects on blood traits.

(a–f) Effects of BMI, WHR, WHRadjBMI on (a) hemoglobin (HGB), (b) hematocrit (HCT), (c) mean corpuscular volume (MCV), (d) red blood cell count (RBC), (e) platelet count (PLT), or (f) white blood cell count (WBC). Shown in top panel are effects of BMI, WHR, or WHRadjBMI on HGB in univariable Mendelian randomization (MR) experiments by inverse variance weighted (IVW) method. Underneath univariable MR results, effects of BMI or WHR at 639 LD-independent WHR-associated single nucleotide polymorphisms (SNPs) are shown. Bottom row of panels show effects of BMI or WHR at 1268 LD-independent BMI-associated SNPs. Effects are in SD units with 95% confidence intervals. *p < 0.05, **p < 0.003. (g) Schematic summarizing effects of indicated exposures on blood traits (created with https://BioRender.com).

Cholesterol levels impact erythroid traits independent of BMI

We next investigated previously proposed mechanisms to explain observational links between BMI and anemia. For example, we hypothesized that hypercholesterolemia may cause anemia (Shalev et al., 2007) through altered erythrocyte membrane formation and stability (Mohandas and Gallagher, 2008). We confirmed that increased total cholesterol (TC) or lipid fractions (low density lipoprotein [LDL] or high density lipoprotein [HDL]), but not triglyceride levels (TG), decreased HGB or HCT (Figure 1—figure supplements 2–3). However, multivariable and mediation experiments revealed that cholesterol levels alter erythroid traits via mechanisms independent from BMI (Figure 1—figure supplement 3b and Supplementary file 1—Table 2). Chronic inflammation and iron deficiency, which cause decreased erythrocyte size (microcytosis), have also been suggested to mediate obesity-related anemia (Aigner et al., 2014). However, BMI did not alter erythrocyte mean corpuscular volume (MCV) by MR (p = 0.12, Figure 1c, Figure 1—figure supplement 4a). In sum, these findings aligned with clinical observations linking BMI with anemia risk, but argued against prevailing mechanistic hypotheses at the genetic level.

Reverse causality experiments also identified inverse correlations between erythroid and metabolic traits (Figure 1—figure supplement 5). Directional MR Steiger (Hemani et al., 2017) analyses were inconsistent (Supplementary file 1—Table 3), perhaps limited by blood trait measurement variation or quantitative adjustments for individual characteristics (Vuckovic et al., 2020).

Genetic impacts of WHR oppose BMI effects on blood trait variation

We then considered an alternative hypothesis that the physiological distribution of adipose, as measured by WHR, could impact BMI-related anemia risk. Unexpectedly, and in contrast to BMI, higher WHR increased red blood cell (RBC) traits (Figure 1a and b and Figure 1—figure supplement 6a,b). WHR adjusted for BMI on an individual level (WHRadjBMI) exacerbated these positive effects (Figure 1a and b and Figure 1—figure supplement 6c,d) and also associated with decreased MCV (Figure 1c and Figure 1—figure supplement 4b,c). Multivariable analyses formally validated the opposing, cross-mediating effects of BMI and WHR on erythroid traits (Figure 1a, b and c).

Next, we asked whether BMI- and WHR-related effects impacted non-erythroid blood traits. Multivariable and mediation analyses on other blood traits identified cross-mediating opposing effects of WHR and BMI on quantitative blood counts across cell lineages (Figure 1d, e and f and Figure 1—figure supplements 7–8). These effects persisted after accounting for related blood traits (Figure 1—figure supplement 9). The directionally consistent effects across multiple lineages suggested that underlying mechanisms related to hematopoietic stem and progenitor cells (HSCs) common to these lineages (Thom and Voight, 2020; Figure 1g). These findings also argued against sex-related effects. Men generally have higher HGB (Vuckovic et al., 2020) and WHR (Pulit et al., 2019), but women can have higher platelet and neutrophil counts (Bain, 1996).

BMI- and WHR-adjusted conditional analyses clarify blood trait variation loci

Our findings suggested that effects from BMI and WHR were likely to have influenced some previously reported blood trait loci. To identify blood loci related to these factors, we applied mtCOJO (Zhu et al., 2018) to condition blood trait GWAS data on polygenic-measured BMI and/or WHR. BMI adjustment modestly changed SNP effect sizes and retained most (91%) unadjusted lead sentinel variants (Figure 2a and b and Supplementary file 1—Table 4). However, combined BMI and WHR adjustment resulted in substantial SNP effect size changes (>36-fold increased SD (SD = 7.6 × 10–3) vs. BMI adjustment alone (SD = 2.1 × 10–4), p < 0.0001 by F-test to compare variances), with a negative skew (–2.0) reflecting adjustment for the positive effect of WHR on HGB levels (Figure 2a and b). Thus, while prior GWAS adjusted for BMI (Astle et al., 2016), we identified more prominent effects for WHR across blood traits and lineages.

Figure 2 with 6 supplements see all
Download asset Open asset
Conditional blood trait analysis based on body mass index (BMI) and/or waist-to-hip ratio (WHR) modifies interpretation of genomic loci that impact blood trait variation.

(a) Violin plots showing the dispersion in effect size at genome-wide significant loci after adjusting erythroid traits (hemoglobin [HGB] or hematocrit [HCT]) for BMI, or WHR and BMI. **p < 0.0001 by F-test to compare variances. (b) Scatterplots depicting changes in effect sizes and p-values for all genome-wide significant sentinel loci before or after adjustment. Novel loci (red) had p < 5 × 10–8 only after adjustment and represent new loci (not in LD with genome-wide significant single nucleotide polymorphisms [SNPs] before adjustment). Clarified loci (pink) are sentinel SNPs with p < 5 × 10–8 after adjustment and are in linkage disequilibrium with significant pre-adjustment SNPs. Missing loci (blue) are those with adjusted p > 5 × 10−8, which were significant pre-adjustment. Shared SNPs (gray) are sentinel SNPs before and after adjustment for the indicated factors. (c) After adjustment for WHR and BMI, the common coding SNP (rs145878042) in RAPGEF3 significantly impacts HGB level. (d) Adjustment for WHR and BMI alters interpretation of SNP effects at the RSPO3 locus, including more significant effects for new sentinel variant rs72959041 (unadjusted p = 2.1 × 10–10, adjusted p = 3.4 × 10–40). (e) Gene-based Manhattan plots for HGB, before or after BMI/WHR adjustment. (f) Gene ontology analyses for hematopoietic, mesenchymal, and vascular biological processes for HGB loci before and after mtCOJO adjustment for BMI and WHR. Significance reflects Fisher’s exact test after multiple testing.

Combined BMI/WHR adjustment shifted 341 HGB-associated loci toward the null, supporting a key role for BMI- and WHR-mediated mechanisms at these sites, while also identifying 844 sites that either clarified interpretation of previously implicated HGB loci or tagged previously unreported regions (n = 242 novel loci, Figure 2b and Supplementary file 1—Table 5). For example, a missense coding variant in RAPGEF3 (rs145878042), previously linked to BMI (Pulit et al., 2019), WHRadjBMI (Pulit et al., 2019), and platelet distribution width (Vuckovic et al., 2020), did not meet genome-wide significance for HGB (p = 0.003) until BMI/WHR adjustment (p = 2.4 × 10–14, Figure 2b and c). Interpretation of SNPs at the RSPO3 locus also dramatically changed (Figure 2b and d). RSPO3, a Wnt pathway modulator that directs development of bone and other tissues (Nilsson et al., 2021), has been linked with adipose distribution (Pulit et al., 2019) and blood trait variation (Vuckovic et al., 2020). Similar effects were seen in adjusted HCT data (Figure 2a and b and Supplementary file 1—Tables 6–7) and quantitative traits across blood lineages (Figure 2—figure supplement 1). These conditional analyses presumably revealed sites where BMI and/or WHR biology most strongly impact blood trait variation, although it is possible that some pleiotropic loci independently regulate blood traits through shared or different gene regulation.

Functional enrichment analyses identified many consistent genes and processes in unadjusted vs. BMI/WHR-adjusted data across blood traits, with some notable changes (Figure 2e and f, Figure 2—figure supplements 2–6 and Supplementary file 1—Tables 8–23) (Mi et al., 2019; Watanabe et al., 2019). For example, at the gene level, association with LEPR in these adjusted analysis for HGB elevated to statistical attention (Figure 2e). LEPR perturbations cause obesity (Dubern and Clement, 2012), and LepR+ endothelial niches support HSC survival (Comazzetto et al., 2021). Further, adjusted HGB locus-related genes were enriched for some endothelial and mesenchymal development processes, albeit with diminished p-values due to power loss from limited SNP sets (Figure 2f and Supplementary file 1-Tables 8–17). Adjusted RBC, PLT, and WBC data also demonstrated enrichment of endothelial and cell adhesion pathways (Figure 2—figure supplement 1 and Supplementary file 1-Tables 18–23). These findings highlight the relevance for BMI, WHR, and related biology in regulating multilineage blood traits, including contributions from mesenchyme-derived adipocytes (Zhong et al., 2020) and stromal endothelial cells in bone marrow (Comazzetto et al., 2021).

Discussion

The obesity epidemic has increased the importance of understanding associated systemic comorbidities (Koenen et al., 2021), including complex physiology linking cardiometabolic and blood traits. While some clinical epidemiological studies have proposed iron deficiency and chronic inflammation to explain anemia observed in obese populations (Aigner et al., 2014; Benova and Tencerova, 2020; Koenen et al., 2021), confounders inherent to observational studies may limit interpretation. Consistent with most clinical observations (Aigner et al., 2014), genetically determined BMI is indeed causally associated with lower HGB and HCT levels.

We identified divergent genome-wide effects of BMI and WHR on erythroid traits. Whereas increased BMI or WHR are typically thought to concordantly raise cardiovascular risk, our results agree with recent findings showing that adipose distribution can influence obesity-related comorbidities (Huang et al., 2021). We were somewhat surprised to identify impacts for BMI and WHR on multilineage blood traits that extended beyond clinically reported erythroid effects, suggesting BMI and WHR may act through different mechanisms than previously proposed (Aigner et al., 2014; Koenen et al., 2021). While the absolute effect sizes are unlikely to impact patient management (e.g., a 1 SD unit increase in BMI [~4.8 kg/m2] decreases HGB by ~0.06 g/dL [Beutler and Waalen, 2006; Gharahkhani et al., 2019]), genetic mechanisms linking BMI, WHR, and blood traits may help elucidate how cardiovascular disease (Heyde et al., 2021; Rohde et al., 2021) and cardiometabolic derangements (Fuster et al., 2020; Jaiswal et al., 2014) are linked to normal or clonal hematopoiesis.

Directionally consistent effects for BMI and WHR across blood lineages may indicate influences on HSCs in the bone marrow (Figure 1g). For example, genetic predisposition to accumulate bone marrow white adipose tissue may underlie age- (Tuljapurkar et al., 2011) or obesity-related (Benova and Tencerova, 2020) cytopenias by regulating HSC self-renewal or differentiation (Wang et al., 2018). Alternatively, genetically determined differences in bone marrow stromal cell types (e.g., mesenchymal stem cell-derived bone marrow adipocytes; Zhong et al., 2020) could impact HSC biology. Finally, WHR-related mechanisms impacting blood trait variation may reflect inhibitory paracrine or endocrine effects from gluteal or truncal adipose depots (Comazzetto et al., 2021).

Strengths of this study include the use of the largest and most recent GWAS statistics available for all traits, as well as consistent directional effect trends across multiple analyses. However, these analyses were restricted to individuals of European descent, perhaps limiting generalizability. Our study is also subject to limitations of currently available MR methods, including potential MR assumption violations (see Materials and methods), unforeseen pleiotropy, or traits correlating with BMI or WHR that may confound direct causality (Burgess et al., 2019).

At minimum, this provides a rationale for concurrent BMI and WHR adjustments when analyzing blood trait GWAS loci to avoid directional bias. These adjustments also provide novel stratification criteria for blood trait GWAS fine mapping studies and candidate blood gene selection. This will be particularly important for studies aiming to explain metabolic or stromal effects on blood cells, which are notably distinct from cholesterol- or lipid-mediated peripheral effects on erythroid cells at the genetic level.

Materials and methods

GWAS summary statistics collection

Request a detailed protocol

We analyzed publicly available GWAS summary statistics for blood traits (n = 563,085) (Vuckovic et al., 2020), BMI (n = 484,680) (Pulit et al., 2019), WHR (n = 485,486) (Pulit et al., 2019), WHRadjBMI (n = 484,563) (Pulit et al., 2019), CAD (n = 547,261) (van der Harst and Verweij, 2018), and lipid traits including TC (n = 215,551), TG (n = 211,491), LDL (n = 215,196), and HDL (n = 210,967) (Klarin et al., 2018). A glossary of these traits, including unit measurements and descriptions, is available in Supplementary file 1—Table 24. Data were derived from individuals of European ancestry only and were analyzed using genome build hg19/GRCh37.

Instrumental variable creation

Request a detailed protocol

To construct instrumental variables (IVs), we identified all SNPs common to exposure and outcome data sets and clumped genome-wide significant SNPs for the exposure to identify linkage-independent SNPs (EUR r2 < 0.01) in 500 kb regions using TwoSample MR. IV strengths were estimated using F-statistics calculated as described (Burgess and Thompson, 2011). IVs used in this study can be found on GitHub (https://github.com/thomchr/ObesityAdiposityBloodMR) or obtained upon request.

MR and causal effect estimation

Request a detailed protocol

Univariable MR analyses (TwoSample MR package v0.5.5; Hemani et al., 2018) were conducted using R (v3.6.3). Random variant allele allocation at meiosis enables the MR approach to address confounding and reverse causality that can otherwise preclude causal inference from epidemiologic and cohort studies. Key assumptions must hold in order to make valid conclusions from MR studies. For example, independent genetic instruments (SNPs) must be specifically associated with the exposure trait. Weak instruments, the presence of horizontal pleiotropy, heterogeneity, and error in measured instrument-exposure associations can limit applicability or inferences gleaned from MR studies (Burgess et al., 2019).

Presented data show causal estimates from IVW (random effects model), weighted median, and MR Egger regression methods. We assessed pleiotropic bias using MR Egger regression intercepts, which if significantly non-zero can imply directional bias (Bowden et al., 2015). MVMR analyses utilized the MVMR package (Sanderson et al., 2019) in R. Results shown are IVW method-based causal estimates. Causal direction analyses utilized MR-Steiger and we report values for sensitivity, statistical significance, and inference of the ‘correct causal direction’ (Hemani et al., 2017).

For continuous outcomes (blood traits, lipid traits, BMI), results are presented as beta effect values representing changes in SD units for these traits, per SD unit change in exposure. SD unit estimations were previously calculated for BMI (Gharahkhani et al., 2019) and HGB (Beutler and Waalen, 2006). For dichotomous outcomes (CAD), causal effect estimates can be converted to odds ratios by exponentiating causal effect estimates ( = exp^[effect]) to calculate a value reflecting the change in outcome per SD unit increase in exposure (Burgess and Labrecque, 2018). However, CAD outcome values are presented as SD units to facilitate comparison with blood trait effects.

Mediation analysis

Request a detailed protocol

Mediation analysis estimates were calculated as described (Burgess et al., 2017). Total and direct effects are reported for the exposure and mediating trait on each outcome.

Conditional GWAS analysis

Request a detailed protocol

Conditional analyses of filtered SNP sets, containing SNPs found in BMI, WHR, and blood trait summary statistics, were analyzed using mtCOJO with a limit of r2 < 0.01 (Yang et al., 2011). Results were clumped using plink (v1.90 beta) (Purcell et al., 2007) to identify linkage-independent sentinel SNPs with r2 < 0.01 in 500 kb genomic regions (flagged parameters were --clump-p1 5E-8 --clump-p2 1 --clump-r2 0.01 --clump-kb 500). Separate experiments were performed on the same filtered SNP sets to adjust for BMI, or both BMI and WHR. To compare uncorrected with BMI- or BMI-and-WHR-adjusted results, we aggregated sentinel SNPs and clumped based on original GWAS p-values (--clump-p1 1 --clump-p2 1 --clump-r2 0.01 --clump-kb 500) to retrieve a complete set of linkage-independent loci. This second clump output allowed us to calculate how many regions were shared, nullified, or novel in the adjusted vs. unadjusted data sets. The gene nearest to each sentinel locus was identified using bedtools (Quinlan and Hall, 2010). Locus zoom plots were created through the online instrument (http://locuszoom.org, Pruim et al., 2010).

Gene-level analyses

Request a detailed protocol

We identified gene and tissue associations for blood trait summary statistics before and after adjustment using FUMA (Watanabe et al., 2019), which uses MAGMA for gene identification (de Leeuw et al., 2015).

Gene ontology

Request a detailed protocol

Gene lists were analyzed for significantly over- or under-enriched gene ontology biological processes. Statistical significance was assigned based on Fisher’s exact test p < 0.05 after Bonferroni correction for multiple testing (http://geneontology.org, Mi et al., 2019).

Statistical analyses and data presentation

Request a detailed protocol

Estimated effects from exposure(s) on outcome are presented from IVW, weighted median, and MR Egger regression measures. Because Cochran’s Q test (included in the TwoSample MR package; Hemani et al., 2018 found heterogeneity in some IVs, we utilized the random effect model when performing inverse variance weighted MR). Thus, we performed and report MR results using IVs that had not undergone pruning. Statistical significance was defined as p < 0.05 for all experiments. For experiments that analyzed 16 blood traits, we also report those that met a more stringent threshold of p < 0.003 (~0.05/16).

Statistics were calculated with GraphPad Prism 8. Figures were prepared using GraphPad Prism 8 and Inkscape (v1.1). Schematic cartoons were created using BioRender.

Coding scripts and data sets

Request a detailed protocol

Data availability

The current manuscript is a computational study based on publicly available data sets, so no primary data were generated for this manuscript. All relevant coding scripts and data sets can be found on GitHub (https://github.com/thomchr/ObesityAdiposityBloodMR) or by request.

The following previously published data sets were used
    1. Vuckovic et al
    (2020) GWAS Catalog
    ID GCST90002379-GCST90002407. Blood Trait Summary Statistics (UKBB).
    ftp://ftp.sanger.ac.uk/pub/project/humgen/summary_statistics/UKBB_blood_cell_traits/
    1. Vuckovic et al
    (2020) humangenetics
    ID Meta-analysis. Blood Trait Summary Statistics (Meta-analysis).
    http://www.mhi-humangenetics.org/en/resources
    1. Pulit et al
    (2019) Zenodo
    Obesity and Adiposity Summary Statistics.
    https://doi.org/10.5281/zenodo.1251813

References

Article and author information

Author details

  1. Christopher S Thom

    Division of Neonatology, Children's Hospital of Philadelphia, Philadelphia, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Writing – original draft, Writing – review and editing
    For correspondence
    thomc@chop.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-1830-9922

  2. Madison B Wilken

    Division of Neonatology, Children's Hospital of Philadelphia, Philadelphia, United States
    Contribution
    Formal analysis, Writing – review and editing
    Competing interests
    No competing interests declared

  3. Stella T Chou

    Division of Hematology, Children's Hospital of Philadelphia, Philadelphia, United States
    Contribution
    Investigation, Writing – review and editing
    Competing interests
    No competing interests declared

  4. Benjamin F Voight

    1. Department of Systems Pharmacology and Translational Therapeutics, University of Pennsylvania - Perelman School of Medicine, Philadelphia, United States
    2. Department of Genetics, University of Pennsylvania - Perelman School of Medicine, Philadelphia, United States
    3. Institute for Translational Medicine, University of Pennsylvania - Perelman School of Medicine, Philadelphia, United States
    Contribution
    Conceptualization, Data curation, Funding acquisition, Investigation, Supervision, Writing – original draft, Writing – review and editing
    For correspondence
    bvoight@pennmedicine.upenn.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6205-9994

Funding

Eunice Kennedy Shriver National Institute of Child Health and Human Development (T32HD043021)

  • Christopher S Thom

National Heart, Lung, and Blood Institute (K99HL156052)

  • Christopher S Thom

National Heart, Lung, and Blood Institute (U01HL124696)

  • Stella T Chou

National Institute of Diabetes and Digestive and Kidney Diseases (R56DK101478)

  • Benjamin F Voight

National Institute of Diabetes and Digestive and Kidney Diseases (UM1DK126194)

  • Benjamin F Voight

Linda Pechenik Montague (Investigator Award)

  • Benjamin F Voight

Children's Hospital of Philadelphia (K-readiness award)

  • Christopher S Thom

National Heart, Lung, and Blood Institute (U24HL134763)

  • Christopher S Thom

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

CST and BFV designed the study. CST, MW, STC, and BFV conducted, analyzed, and/or interpreted experimental data. CST and BFV wrote the paper. All authors approved the final version of the manuscript. This work was supported by National Institutes of Health (T32 HD043021, K99 HL156052, and U24 HL134763 to CST; U01 HL124696 to STC; R56 DK101478 and UM1 DK126194 to BFV), a Linda Pechenik Montague Investigator Award (BFV), and a Children’s Hospital of Philadelphia K-readiness award (CST).

Copyright

© 2022, Thom et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,088
    views
  • 160
    downloads
  • 6
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Christopher S Thom
  2. Madison B Wilken
  3. Stella T Chou
  4. Benjamin F Voight
(2022)
Body mass index and adipose distribution have opposing genetic impacts on human blood traits
eLife 11:e75317.
https://doi.org/10.7554/eLife.75317

Share this article

https://doi.org/10.7554/eLife.75317

Categories and tags

Research organism

Further reading

    1. Computational and Systems Biology
    2. Genetics and Genomics

    Functional characteristics and computational model of abundant hyperactive loci in the human genome

    Sanjarbek Hudaiberdiev, Ivan Ovcharenko
    Research Article

    Enhancers and promoters are classically considered to be bound by a small set of transcription factors (TFs) in a sequence-specific manner. This assumption has come under increasing skepticism as the datasets of ChIP-seq assays of TFs have expanded. In particular, high-occupancy target (HOT) loci attract hundreds of TFs with often no detectable correlation between ChIP-seq peaks and DNA-binding motif presence. Here, we used a set of 1003 TF ChIP-seq datasets (HepG2, K562, H1) to analyze the patterns of ChIP-seq peak co-occurrence in combination with functional genomics datasets. We identified 43,891 HOT loci forming at the promoter (53%) and enhancer (47%) regions. HOT promoters regulate housekeeping genes, whereas HOT enhancers are involved in tissue-specific process regulation. HOT loci form the foundation of human super-enhancers and evolve under strong negative selection, with some of these loci being located in ultraconserved regions. Sequence-based classification analysis of HOT loci suggested that their formation is driven by the sequence features, and the density of mapped ChIP-seq peaks across TF-bound loci correlates with sequence features and the expression level of flanking genes. Based on the affinities to bind to promoters and enhancers we detected five distinct clusters of TFs that form the core of the HOT loci. We report an abundance of HOT loci in the human genome and a commitment of 51% of all TF ChIP-seq binding events to HOT locus formation thus challenging the classical model of enhancer activity and propose a model of HOT locus formation based on the existence of large transcriptional condensates.

    1. Developmental Biology
    2. Genetics and Genomics

    Rescuable sleep and synaptogenesis phenotypes in a Drosophila model of O-GlcNAc transferase intellectual disability

    Ignacy Czajewski, Bijayalaxmi Swain ... Daan MF van Aalten
    Research Article

    O-GlcNAcylation is an essential intracellular protein modification mediated by O-GlcNAc transferase (OGT) and O-GlcNAcase (OGA). Recently, missense mutations in OGT have been linked to intellectual disability, indicating that this modification is important for the development and functioning of the nervous system. However, the processes that are most sensitive to perturbations in O-GlcNAcylation remain to be identified. Here, we uncover quantifiable phenotypes in the fruit fly Drosophila melanogaster carrying a patient-derived OGT mutation in the catalytic domain. Hypo-O-GlcNAcylation leads to defects in synaptogenesis and reduced sleep stability. Both these phenotypes can be partially rescued by genetically or chemically targeting OGA, suggesting that a balance of OGT/OGA activity is required for normal neuronal development and function.

    1. Evolutionary Biology
    2. Genetics and Genomics

    Systematic genetic characterization of the human PKR kinase domain highlights its functional malleability to escape a poxvirus substrate mimic

    Michael James Chambers, Sophia B Scobell, Meru J Sadhu
    Research Article

    Evolutionary arms races can arise at the contact surfaces between host and viral proteins, producing dynamic spaces in which genetic variants are continually pursued.  However, the sampling of genetic variation must be balanced with the need to maintain protein function. A striking case is given by protein kinase R (PKR), a member of the mammalian innate immune system. PKR detects viral replication within the host cell and halts protein synthesis to prevent viral replication by phosphorylating eIF2α, a component of the translation initiation machinery. PKR is targeted by many viral antagonists, including poxvirus pseudosubstrate antagonists that mimic the natural substrate, eIF2α, and inhibit PKR activity. Remarkably, PKR has several rapidly evolving residues at this interface, suggesting it is engaging in an evolutionary arms race, despite the surface’s critical role in phosphorylating eIF2α. To systematically explore the evolutionary opportunities available at this dynamic interface, we generated and characterized a library of 426 SNP-accessible nonsynonymous variants of human PKR for their ability to escape inhibition by the model pseudosubstrate inhibitor K3, encoded by the vaccinia virus gene K3L. We identified key sites in the PKR kinase domain that harbor K3-resistant variants, as well as critical sites where variation leads to loss of function. We find K3-resistant variants are readily available throughout the interface and are enriched at sites under positive selection. Moreover, variants beneficial against K3 were also beneficial against an enhanced variant of K3, indicating resilience to viral adaptation. Overall, we find that the eIF2α-binding surface of PKR is highly malleable, potentiating its evolutionary ability to combat viral inhibition.