Sage Journals: Discover world-class research

Abstract

Background:

In recent years, driven by the rapid advancement of proteomics research, numerous scholars have investigated the intricate associations between plasma proteins and various diseases. Thus, this study aimed to identify novel therapeutic targets for preventing and treating metabolic-related diseases through Mendelian randomization (MR).

Methods:

This study primarily utilized the MR method, leveraging genetic data from multiple large-scale publicly available genome-wide association studies. We employed two-sample MR within this framework to assess the associations between 1001 plasma proteins and 5 metabolism-related diseases. Finally, we strengthen the robustness and reliability of the MR results by conducting a series of sensitivity analyses, including bidirectional MR, colocalization analysis, Cochran’s Q test, and the MR-Egger intercept test.

Results:

The results from the inverse variance weighted method revealed that, following false discovery rate correction, many plasma proteins are significantly associated with metabolic-related diseases. Genetically predicted risks vary across diseases: for coronary artery disease, from 0.82 FGR proto-oncogene, Src family tyrosine kinase (FGR) to 1.13 (interleukin-6); for obesity, from 0.992 (POLR2F) to 1.005 (PRKAB1); for osteoporosis, from 0.998 (AIF1) to 1.001 (CLC); for stroke, from 0.71 (TNFRSF1A) to 1.47 (TGM2); and for type 2 diabetes, from 0.79 (KRT18) to 1.47 (RAB37).

Conclusion:

Our findings reveal numerous plasma proteins linked to metabolic-related diseases. These findings offer fresh insights into the etiology, diagnostics, and treatment of these conditions.

Keywords

Mendelian randomization metabolic-related diseases plasma protein proteomics type 2 diabetes

Introduction

The endocrine metabolic system is recognized as one of the most vital systems in the human body and plays a pivotal role in maintaining homeostasis.¹ Metabolic disorders frequently contribute to the onset of diseases such as diabetes, obesity, hyperuricemia, osteoporosis, stroke, and coronary heart disease. As research into metabolic-related diseases progresses, numerous studies have indicated that multiple metabolic-related diseases frequently co-occur, implying potential internal correlations.² Therefore, in 1999, the World Health Organization introduced the concept of metabolic syndrome, characterized by central obesity and concurrent conditions such as dyslipidemia and hyperglycemia.³ Furthermore, stroke, coronary heart disease, and osteoporosis are also recognized as components of metabolic-related diseases. Studies indicate that their occurrence and development are interconnected, sharing common underlying causes, mechanisms, and characteristics.⁴ As the primary driver of non-communicable diseases, metabolic-related conditions have emerged as a leading cause of death worldwide, affecting a growing number of patients.⁵

Plasma proteins have been studied for disease diagnosis and treatment for more than a century, yet traditional proteomics techniques have limited their progress.⁶ Advances in high-throughput detection and quantitative techniques for serum proteins have spurred further research into the relationship between proteins and disease risk.^7,8 Plasma proteins can also act as binding sites for clinical drugs, modifying downstream protein activities to impede disease progression.⁹ Previous studies have categorized plasma proteins into five distinct groups according to their potential pharmacological roles: substitution of deficient or aberrant proteins, enhancement of existing channels, provision of novel functions or activities, interference with molecular or organismal processes, and delivery of other compounds or proteins.¹⁰ To date, numerous plasma proteins have been identified as potential therapeutic targets for various diseases.^11,12 A large-scale population study identified 47 plasma proteins linked to type 2 diabetes.¹³ Another study identified three plasma proteins linked to stroke.¹⁴ Research suggests that choosing drug targets supported by direct genetic evidence in drug development trials can double the success rate of clinical drug development.¹⁵ However, existing studies often face challenges such as small sample sizes, numerous confounding factors, and uncertainties about factors influencing proteomics, potentially leading to biased results.

Mendelian randomization (MR) was initially regarded as a promising alternative to randomized controlled trials, employing genetic tools as credible proxies for exposure and disease. According to Mendelian genetics, genetic information is randomly distributed at conception, occurring before disease onset, thereby minimizing confounding biases from various factors.¹⁶ Furthermore, MR in conjunction with proteomics is increasingly recognized as a powerful approach for identifying potential therapeutic targets in the context of various diseases.^17–19 For example, the study conducted by Bourgault et al.²⁰ identified 29 plasma proteins that may serve as potential therapeutic targets for acute pancreatitis, whereas Rasooly et al.²¹ highlighted seven plasma proteins that could offer new therapeutic opportunities for heart failure.

This study explored the relationship between plasma protein levels and five metabolic-related diseases—coronary artery disease (CAD), obesity, osteoporosis, stroke, and type 2 diabetes—through MR. In previous studies, we utilized the most recent plasma proteomics data to identify additional drug targets for treating metabolic-related diseases.

Methods

Study design

This study utilized the MR method to investigate the causal relationships between the levels of 1001 plasma proteins and 5 metabolic-related diseases—CAD, obesity, osteoporosis, stroke, and type 2 diabetes—across multiple independent large-scale human genome-wide association studies (GWAS). The reliability of the findings was subsequently assessed via various sensitivity analysis methods and bidirectional MR. In all MR analyses, three fundamental assumptions must be rigorously followed: first, the selected genetic instruments must strongly correlate with the exposure factor; second, they must remain unaffected by confounding variables; and third, their effect on the outcome must be entirely mediated by the specified exposure factor.²² Figure 1 depicts the foundational design framework of this study.

Figure 1.

Study flow chart.

Data sources

All the data utilized in this study were obtained from global GWASs conducted within European populations, which were exclusively sourced from publicly accessible aggregated datasets (see Table 1).

Table 1.

Data sources for studied phenotypes.

Phenotype	Cases	Controls	Sample size	PMID/GWAS ID
Plasma proteins	—	—	54,000	38412862
Coronary artery disease	—	—	352,063	34017140/ebi-a-GCST90013864
Obesity	4688	458,322	463,010	ukb-b-15541
Osteoporosis	1976	461,034	463,010	ukb-b-17796
Stroke	11,929	472,192	484,121	34594039/ebi-a-GCST90018864
T2DM	38,841	451,248	490,089	34594039/ebi-a-GCST90018926

GWAS, genome-wide association study.

Genetic data for plasma proteins were obtained from a recent GWAS, in which where researchers utilized Olink proteomics data from 54,000 samples in the UK Biobank to identify genetic loci associated with 1001 different plasma protein levels.²³ Genetic data for CAD were sourced from a GWAS involving 407,746 samples.²⁴ Data on obesity and osteoporosis were obtained from publicly available studies conducted by the MRC-IEU on the UK Biobank.²⁵ Genetic data for stroke and type 2 diabetes were obtained from a comprehensive study involving the UK Biobank and FinnGen research.²⁶ Importantly, minimal overlap in samples was observed when the datasets of plasma proteins and the five metabolic-related diseases were compared.

Instrument selection

To ensure a sufficient number of single-nucleotide polymorphisms (SNPs) for subsequent statistical analysis, we established a significance threshold for plasma protein levels of 1 × 10⁻⁵. We then utilized the European 1000 Genomes panel to select SNPs exhibiting relative independence, specifically within a genomic span of 10,000 base pairs, with a linkage disequilibrium threshold (r²) less than 0.01. Furthermore, we evaluated the statistical robustness of all the SNPs via F-statistics, excluding those with an F-statistic less than 10 to mitigate bias arising from weak instrumental variables.²⁷

Statistical analysis

This study primarily employs the inverse variance weighting (IVW) method to evaluate the association between plasma protein levels and type 2 diabetes. It is complemented by four sensitivity analysis techniques—weighted median, simple mode, weighted mode, and MR-Egger tests—to ensure the robustness of the IVW analysis findings. The study proceeded as follows: first, the relationships between the levels of 1001 plasma proteins and 5 metabolic-related diseases were evaluated, and the overall effect size was ultimately calculated. Cochran’s Q test was used to assess the heterogeneity of genetic instruments (p < 0.05), whereas the MR-Egger intercept test was used to assess directional pleiotropy (p < 0.05).²⁸ The false discovery rate (FDR) was applied for multiple testing corrections, with IVW-FDR < 0.05 set as the significance threshold. The inverse associations of identified plasma proteins linked to metabolic-related diseases were subsequently investigated. Finally, Bayesian colocalization analysis was employed to assess the probability of two traits sharing causal variants. Genes with a posterior probability (PP₄) of ⩾0.8 were classified as having strong evidence of colocalization, whereas those with a PP₄ of ⩾0.5 were considered to have moderate evidence supporting colocalization.²⁹

All the statistical analyses were conducted via the TwoSampleMR package (version 0.5.6) and the coloc package within the R software environment (version 4.3.3). Significance was evaluated via rigorous statistical criteria for two-tailed tests.

Results

Instrument selection

As indicated in Table S1, 4–573 SNPs were chosen on the basis of their respective selection criteria to investigate the associations between 1001 plasma proteins and metabolic-related diseases. All the selected genetic instruments presented F-statistics greater than 10, demonstrating the study’s ability to minimize bias from weak instrumental variables.

Proteome-wide MR analysis identified eight proteins that circulate in CAD

Table S1 displays the IVW results revealing significant associations of 138 plasma proteins with CAD (p < 0.05). However, following FDR correction, only 34 plasma proteins were consistently associated with CAD (IVW-FDR < 0.05). Among them, 18 plasma proteins were negatively associated with CAD risk, whereas 16 plasma proteins were positively associated with CAD risk. The genetic predictions of CAD risk varied from 0.82 for FGR to 1.13 for interleukin-6 (IL6; Figure 2). Despite Cochran’s Q test identifying heterogeneity in certain results, the consistent directionality observed with IVW by MR-Egger, weighted median, simple mode, weighted mode, and other sensitivity analysis methods indicates that heterogeneity can be ignored. Furthermore, the pleiotropy test revealed horizontal pleiotropy in the outcomes for CD5, LAG3, and PROC (Table S2). During the replication phase, following FDR correction and the exclusion of results exhibiting horizontal pleiotropy, we successfully validated eight plasma proteins associated with CAD (Tables S3 and S4 and Figure 3(a)). IL6 (odds ratio (OR): 1.13; 95% confidence interval (CI): 1.06, 1.21), LAYN (1.06; 1.03, 1.10), and CDH1 (1.06; 1.03, 1.09) were positively associated with CAD, whereas CEACAM8 (0.91; 0.87, 0.96), HYAL1 (0.87; 0.83, 0.91), LGALS7_LGALS7B (0.94; 0.91, 0.97), NCR1 (0.91; 0.88, 0.95), and NRP2 (0.92; 0.88, 0.97) were negatively associated with CAD.

Figure 2.

Volcano plot of the MR analysis results.

Figure 3.

Forest plot of MR analysis. (a) Forest plot of MR analysis in CAD. (b) Forest plot of MR analysis in obesity. (c) Forest plot of MR analysis in osteoporosis. (d) Forest plot of MR analysis in stroke. (e) Forest plot of MR analysis in T2DM.

Proteome-wide MR analysis identified one protein associated with obesity

Table S1 shows the IVW results, which revealed significant associations of 245 plasma proteins with obesity (p < 0.05). After FDR correction, 146 plasma proteins were consistently associated with obesity (IVW-FDR < 0.05). Among them, 64 plasma proteins were negatively associated with obesity risk, whereas 82 plasma proteins were positively associated with obesity risk. The genetic prediction of obesity risk ranged from 0.992 for POLR2F to 1.005 for PRKAB1 (Figure 2). Despite Cochran’s Q test identifying heterogeneity in certain results, the consistency in directionality observed with IVW by MR-Egger, weighted median, simple mode, weighted mode, and other sensitivity analysis methods indicates that heterogeneity can be ignored. Furthermore, the pleiotropy test revealed horizontal pleiotropy in the outcomes for ADA2, ADM, AIF1, and 46 other plasma proteins (Table S2). During the replication phase, following FDR correction and the exclusion of results exhibiting horizontal pleiotropy, we successfully validated only one plasma protein associated with obesity (Tables S3 and S4 and Figure 3(b)). Specifically, HYAL1 (0.998; 0.997, 0.999) was negatively associated with the risk of obesity.

Proteome-wide MR analysis identified three proteins that circulate in osteoporosis

Table S1 shows the IVW results revealed significant associations of 106 plasma proteins with osteoporosis (p < 0.05). Following FDR correction, only nine plasma proteins were consistently associated with osteoporosis (IVW-FDR < 0.05). Among them, seven plasma proteins were negatively associated with osteoporosis risk, whereas two plasma proteins were positively associated with osteoporosis risk. The genetic prediction of osteoporosis risk ranged from 0.998 for AIF1 to 1.001 for CLC (Figures 2 and 3(c)). Despite Cochran’s Q test identifying heterogeneity in certain results, the consistency in directionality observed with IVW by MR-Egger, weighted median, simple mode, weighted mode, and other sensitivity analysis methods indicates that heterogeneity can be ignored. Furthermore, the pleiotropy test revealed horizontal pleiotropy in the outcomes for PRSS27 only (Table S2). During the replication phase, following FDR correction and the exclusion of results exhibiting horizontal pleiotropy, we successfully validated three plasma proteins associated with osteoporosis (Tables S3 and S4 and Figure 3(c)). Among these, only CLC (1.001; 1.000, 1.002) was positively associated with osteoporosis, whereas C2 (0.999; 0.998, 1.000) and AIF1 (0.998; 0.997, 0.999) demonstrated negative associations.

Proteome-wide MR analysis identified 42 proteins associated with stroke

Table S1 shows the IVW results revealed significant associations of 238 plasma proteins with stroke (p < 0.05). After FDR correction, 138 plasma proteins were consistently associated with stroke (IVW-FDR < 0.05). Among them, 85 plasma proteins were negatively associated with stroke risk, whereas 53 plasma proteins were positively associated with stroke risk. The genetic prediction of stroke risk ranged from 0.71 for TNFRSF1A to 1.47 for TGM2 (Figures 2 and 3(d)). Despite Cochran’s Q test identifying heterogeneity in certain results, the consistency in directionality observed with IVW by MR-Egger, weighted median, simple mode, weighted mode, and other sensitivity analysis methods indicates that heterogeneity can be ignored. Furthermore, the pleiotropy test revealed horizontal pleiotropy in the outcomes for 47 plasma proteins (Table S2). During the replication phase, following FDR correction and the exclusion of results exhibiting horizontal pleiotropy, we successfully validated 42 plasma proteins associated with stroke (Tables S3 and S4 and Figure 3(d)). The genetically predicted stroke risk ranged from 0.78 for FSTL3 to 1.22 for SEZ6L2.

Proteome-wide MR analysis identified 11 proteins associated with type 2 diabetes

Table S1 displays the IVW results, which revealed significant associations of 160 plasma proteins with type 2 diabetes (p < 0.05). After FDR correction, 55 plasma proteins were consistently associated with type 2 diabetes (IVW-FDR < 0.05). Among them, 26 plasma proteins were negatively associated with type 2 diabetes risk, whereas 29 plasma proteins were positively associated with type 2 diabetes risk. The genetic prediction of type 2 diabetes risk ranged from 0.79 for KRT18 to 1.47 for RAB37 (Figures 2 and 3(e)). Despite Cochran’s Q test identifying heterogeneity in certain results, the consistency in directionality observed with IVW by MR-Egger, weighted median, simple mode, weighted mode, and other sensitivity analysis methods indicates that heterogeneity can be ignored. Furthermore, the pleiotropy test revealed horizontal pleiotropy in the outcomes for 18 plasma proteins, including ABHD14B and ADA2 (Table S2). During the replication phase, following FDR correction and the exclusion of results exhibiting horizontal pleiotropy, we successfully validated 11 plasma proteins associated with type 2 diabetes (Tables S3 and S4 and Figure 3(e)). Specifically, CCL27 (1.18; 1.11, 1.26), SEZ6L2 (1.15; 1.08, 1.22), OMD (1.13; 1.07, 1.20), and SPON1 (1.09; 1.03, 1.15) were positively associated with type 2 diabetes. In contrast, CSF1 (0.87; 0.80, 0.94), FLT3LG (OR: 0.94; 95% CI: 0.90, 0.98), IL15 (0.91; 0.87, 0.95), IL18BP (0.88; 0.83, 0.94), KIR2DL3 (0.96; 0.95, 0.97), MSR1 (0.90; 0.86, 0.94), and TNFRSF21 (0.88; 0.82, 0.94) were negatively associated with type 2 diabetes.

Reverse MR analysis of metabolic-related diseases with the levels of the 56 proteins

As depicted in Table S5 and Figure 4, CAD has inverse causal associations with 9 plasma proteins such as CCL27 and DLL1; obesity has inverse causal associations with 10 plasma proteins including FSTL3 and AIF1; stroke has inverse causal associations with 11 plasma proteins such as CD46; and type 2 diabetes has inverse causal associations with 6 plasma proteins, including GZMA and C2. Owing to a shortage of suitable SNPs for MR regarding osteoporosis, we refrained from exploring their reverse causal relationships with these 56 plasma proteins.

Figure 4.

MR analysis of the associations of metabolic-related diseases with metabolic-related disease-related plasma proteins.

Colocalization analysis supported the causality of plasma proteins with metabolic-related diseases

As indicated in Table S6, among the plasma proteins identified through proteome-wide MR related to metabolic-related diseases, only the association between CCL27 and type 2 diabetes received robust genetic colocalization support (PP₄ ⩾ 0.8). In contrast, HYAL1 and CAD presented moderate support (0.8 ⩾ PP₄ ⩾ 0.5). Regrettably, other associations between plasma proteins and their respective diseases did not demonstrate significant colocalization support in the MR analysis.

Selected plasma proteins currently under investigation as drug targets

Table 2 presents a selection of plasma proteins that are either in development or already available on the market, on the basis of data from DrugBank. IL6-related drugs are the most commonly used primarily for treating conditions such as rheumatic polymyalgia and rheumatoid arthritis. Notably, the majority of potential plasma proteins are currently being developed as therapeutic targets for cancer treatment.

Table 2.

Selected plasma proteins currently under investigation as drug targets.

Plasma protein	Drug	Research and development status	Indication
IL6	Levilimab	Approved for listing	Rheumatoid arthritis, COVID-19
	Sarilumab	Approved for listing	Polyarticular juvenile idiopathic arthritis, Rheumatic polymyalgia, Rheumatoid arthritis
	Siltuximab	Approved for listing	Multicentre Castleman disease
LAYN	WO2022246085	Discoveries	Tumor
CDH1	19A11	Preclinical	Tumor
	KBA-1636	Preclinical	Colorectal cancer, Lynch syndrome type II
CEACAM8	Tc 99m besilesomab	Approved for listing	Osteomyelitis
	Yttrium-90 besilesomab	Clinical phase II	Hematopoietic stem cell transplantation
NRP2	Efzofitimod	Clinical phase III	Pulmonary nodule, pulmonary fibrosis
	aNRP2-10	Preclinical	Triple-negative breast cancer
CLC	ARGX-118	Clinical phase I	Airway pathology
	WO2023139090	Discoveries	Diseases of the digestive system
IL15	BGB-R046	Clinical phase I	Metastatic solid tumor
	GT201	Clinical phase II	Advanced malignant solid tumor
IL18BP	COM503	Clinical phase I	Advanced malignant solid tumor

Discussion

This MR study assessed the associations between 1001 plasma proteins and the risk of 5 metabolic-related diseases via multiple sensitivity analysis methods. Thirty-one, 100, 8, 91, and 37 plasma proteins were significantly associated with CAD, obesity, osteoporosis, stroke, and type 2 diabetes, respectively. It also explored the inverse associations of these plasma proteins with the respective diseases. These findings provide new insights into the treatment of these diseases.

Our study initially validated associations between several plasma proteins and diseases previously identified in the literature. For example, a meta-analysis underscored the association between IL6 and CAD, suggesting that inhibiting IL6 could effectively target cardiovascular disease prevention.³⁰ IL6 is recognized as a proinflammatory cytokine that induces the liver to produce acute-phase proteins in response to tissue damage and infection.³¹ Multiple prospective studies have demonstrated that elevated plasma IL6 levels are associated with increased CAD risk,³² which aligns with our study findings. In our study, we observed a negative association between CCL4 and CAD risk. However, mouse studies have suggested that inhibiting CCL4 can lower the circulating levels of inflammatory factors such as IL6 and TNF-α, thus reducing atherosclerotic plaque size, slowing progression, and improving plaque stability.^33,34 PPP3R1, a critical calcineurin subunit, was shown to increase energy expenditure and mitigate diet-induced obesity in genetically modified mice.³⁵ While our study identified pleiotropy in its association, given the involvement of various calcineurin structures in energy metabolism,³⁶ there remains reason to suspect a potential link between PPP3R1 and obesity.

Our study identified eight plasma proteins linked to osteoporosis. An animal study suggested that miR-181a-5p inhibits the differentiation of preosteoblastic MC3T3-E1 cells in vitro, partially through disrupting Runx1-dependent suppression of AIF1 transcription.³⁷ AIF1, a calcium-binding protein originating from monocytes and macrophages, may be implicated in diverse metabolic disorders, including atherosclerosis and diabetes.³⁸ MMPs constitute a diverse group of endopeptidases that facilitate extracellular matrix degradation or remodeling, which are pivotal in processes such as wound healing and angiogenesis.³⁹ Multiple MR studies have established a link between elevated plasma MMP levels and reduced stroke risk.⁴⁰ The associations of TYRO3, ARG1, and MICB with type 2 diabetes have been previously documented in certain studies,⁴¹ underscoring the relative reliability of the data sources employed in this study.

Furthermore, we identified numerous other plasma proteins potentially linked to metabolic disorders. For example, in epithelial ovarian cancer, DDK4 might enhance tumor invasion by activating the JNK pathway to promote actin filament formation.⁴² In addition, FAP is linked to poor prognosis in multiple cancers, including gastric cancer.⁴³ The overexpression of FAP can remodel the extracellular matrix by cleaving substrate peptides or proteins, thereby enhancing tumor cell migration.⁴⁴ Furthermore, FAP is involved in various signaling pathways including the PI3K-AKT and TGF-β pathways,⁴⁵ and dysregulation of the PI3K-AKT pathway is closely associated with obesity and type 2 diabetes.⁴⁶

This study has several significant strengths. First, genetic data on plasma proteins and five metabolic-related diseases originate from extensive population-based GWASs, which are meticulously structured to minimize sample overlap and thereby mitigate the potential confounding effects arising from shared samples across datasets. Second, all the chosen genetic instruments present F-statistics greater than 10, indicating robust instrument strength that effectively mitigates bias from weaker instruments and bolsters the credibility of the instrumental variables utilized in the analysis. Third, employing diverse sensitivity analysis methods bolstered the reliability of the study outcomes. Finally, several disease risk factors identified through routine observational studies have been shown to lack a causal association with disease in randomized controlled trials, leading to costly trials that fail to produce the anticipated positive outcomes. In contrast to conventional observational studies and GWASs, MR associations are immune to causal inversion, are less prone to confounding effects, and incur significantly lower costs, thus preserving valuable resources for more impactful research.⁴⁷

While this study is relatively comprehensive, it is important to acknowledge several limitations that may impact our interpretation of the research findings. First, MR inherently faces limitations, such as trait heterogeneity and developmental compensation issues, which could impact the accuracy and applicability of our research findings.⁴⁸ Second, our reliance on summary-level data restricted our capacity to perform stratified analysis or a comprehensive investigation into individual-level data. Third, given that our study participants predominantly had European ancestry, caution is warranted in extrapolating the research findings to more ethnically diverse populations, such as Asians. Subsequent studies are crucial for validating the relevance of our research findings across diverse ethnic groups. Fourth, while uncovering causal relationships between numerous plasma proteins and metabolic-related diseases, our understanding of their underlying mechanisms remains incomplete, underscoring the need for additional basic research to elucidate these intricate pathways. Finally, we did not include all metabolism-related diseases, such as hypertension and metabolic-associated fatty liver disease, in our analyses because similar studies using the same methods and datasets have already been reported, or relevant data were unavailable. Nevertheless, by investigating five metabolism-related diseases as outcome events, our study helps to address existing research gaps to the greatest extent possible.

Conclusion

In conclusion, this comprehensive proteome-wide MR study identified numerous plasma proteins strongly linked to metabolic-related diseases, thus identifying promising targets for understanding their pathogenesis, diagnostics, and therapeutic interventions. Further fundamental experiments are essential to evaluate the efficacy of these candidate targets.

Supplemental Material

sj-docx-1-tae-10.1177_20420188251343140 – Supplemental material for Novel therapeutic targets for metabolism-related diseases: proteomic Mendelian randomization and colocalization analyses

Supplemental material, sj-docx-1-tae-10.1177_20420188251343140 for Novel therapeutic targets for metabolism-related diseases: proteomic Mendelian randomization and colocalization analyses by Yue-Yang Zhang, Bin-Lu Wang, Bing-Xue Chen and Qin Wan in Therapeutic Advances in Endocrinology and Metabolism

Supplemental Material

sj-xlsx-2-tae-10.1177_20420188251343140 – Supplemental material for Novel therapeutic targets for metabolism-related diseases: proteomic Mendelian randomization and colocalization analyses

Supplemental material, sj-xlsx-2-tae-10.1177_20420188251343140 for Novel therapeutic targets for metabolism-related diseases: proteomic Mendelian randomization and colocalization analyses by Yue-Yang Zhang, Bin-Lu Wang, Bing-Xue Chen and Qin Wan in Therapeutic Advances in Endocrinology and Metabolism

Footnotes

Appendix

Acknowledgements

We are grateful to the authors and participants of all GWASs for the summary statistics used.

Declarations

ORCID iDs

Yue-Yang Zhang

Qin Wan

Supplemental material

Supplemental material for this article is available online.

References

Heindel

Blumberg

Cave

, et al. Metabolism disrupting chemicals and metabolic disorders. Reprod Toxicol 2017; 68: 3–33.

Guerra

JVS

Dias

MMG

Brilhante

AJVC

, et al. Multifactorial basis and therapeutic strategies in metabolism-related diseases. Nutrients 2021; 13(8): 2830.

Fahed

Aoun

Bou Zerdan

, et al. Metabolic syndrome: updates on pathophysiology and management in 2021. Int J Mol Sci 2022; 23(2): 786.

Grundy

Cleeman

Daniels

, et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005; 112(17): 2735–2752.

Saklayen

MG.

The global epidemic of the metabolic syndrome. Curr Hypertens Rep 2018; 20(2): 12.

Anderson

NG.

The human plasma proteome: history, character, and diagnostic prospects. Mol Cell Proteomics 2002; 1(11): 845–867.

Elhadad

Jonasson

Huth

, et al. Deciphering the plasma proteome of type 2 diabetes. Diabetes 2020; 69(12): 2766–2778.

Rooney

Chen

Echouffo-Tcheugui

, et al. Proteomic predictors of incident diabetes: results from the atherosclerosis risk in communities (ARIC) study. Diabetes Care 2023; 46(4): 733–741.

Roth

Mensah

Johnson

, et al. Global burden of cardiovascular diseases and risk factors, 1990–2019: update from the GBD 2019 study. J Am Coll Cardiol 2020; 76(25): 2982–3021.

10.

Dimitrov

DS.

Therapeutic proteins. Methods Mol Biol 2012; 899: 1–26.

11.

Parayath

Amiji

MM.

Therapeutic targeting strategies using endogenous cells and proteins. J Control Release 2017; 258: 81–94.

12.

Zheng

Lin

Wang

, et al. Effects of circulating inflammatory proteins on spinal degenerative diseases: evidence from genetic correlations and Mendelian randomization study. JOR Spine 2024; 7(2): e1346.

13.

Sattar

Taheri

Astling

, et al. Prediction of cardiometabolic health through changes in plasma proteins with intentional weight loss in the DiRECT and DIADEM-I randomized clinical trials of type 2 diabetes remission. Diabetes Care 2023; 46(11): 1949–1957.

14.

Zou

Wang

, et al. Promising therapeutic targets for ischemic stroke identified from plasma and cerebrospinal fluid proteomes: a multicenter Mendelian randomization study. Int J Surg 2023; 110(2): 766–776.

15.

Nelson

Tipney

Painter

, et al. The support of human genetic evidence for approved drug indications. Nat Genet 2015; 47(8): 856–860.

16.

Lawlor

Harbord

Sterne

JAC

, et al. Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat Med 2008; 27(8): 1133–1163.

17.

Sun

Yun

Lin

, et al. Comprehensive Mendelian randomization analysis of plasma proteomics to identify new therapeutic targets for the treatment of coronary heart disease and myocardial infarction. J Transl Med 2024; 22(1): 404.

18.

Zheng

Wang

Lin

, et al. Mendelian randomization analysis suggests no associations of human herpes viruses with amyotrophic lateral sclerosis. Front Neurosci 2023; 17: 1299122.

19.

Zheng

Wang

Lin

, et al. Effects of circulating inflammatory proteins on osteoporosis and fractures: evidence from genetic correlation and Mendelian randomization study. Front Endocrinol 2024; 15: 1386556.

20.

Bourgault

Abner

Manikpurage

, et al. Proteome-wide Mendelian randomization identifies causal links between blood proteins and acute pancreatitis. Gastroenterology 2023; 164(6): 953–965.e3.

21.

Rasooly

Peloso

Pereira

, et al. Genome-wide association analysis and Mendelian randomization proteomics identify drug targets for heart failure. Nat Commun 2023; 14(1): 3826.

22.

Levin

Burgess

Mendelian randomization as a tool for cardiovascular research: a review. JAMA Cardiol 2024; 9(1): 79–89.

23.

Suhre

Genetic associations with ratios between protein levels detect new pQTLs and reveal protein-protein interactions. Cell Genom 2024; 4(3): 100506.

24.

Mbatchou

Barnard

Backman

, et al. Computationally efficient whole-genome regression for quantitative and binary traits. Nat Genet 2021; 53(7): 1097–1103.

25.

Mitchell

Elsworth

Mitchell

, et al. MRC IEU UK Biobank GWAS pipeline version 2. data.bris, https://doi.org/10.5523/bris.pnoat8cxo0u52p6ynfaekeigi

26.

Sakaue

Kanai

Tanigawa

, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet 2021; 53(10): 1415–1424.

27.

Papadimitriou

Dimou

Tsilidis

, et al. Physical activity and risks of breast and colorectal cancer: a Mendelian randomisation analysis. Nat Commun 2020; 11(1): 597.

28.

Bowden

Davey Smith

Haycock

, et al. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol 2016; 40(4): 304–314.

29.

Lin

Zhou

Potential drug targets for multiple sclerosis identified through Mendelian randomization analysis. Brain J Neurol 2023; 146(8): 3364–3372.

30.

Interleukin-6 Receptor Mendelian Randomisation Analysis (IL6R MR) Consortium; Swerdlow

Holmes

Kuchenbaecker

, et al. The interleukin-6 receptor as a target for prevention of coronary heart disease: a Mendelian randomisation analysis. Lancet 2012; 379(9822): 1214–1224.

31.

Kerr

Stirling

Ludlam

CA.

Interleukin 6 and haemostasis. Br J Haematol 2001; 115(1): 3–12.

32.

Danesh

Kaptoge

Mann

, et al. Long-term interleukin-6 levels and subsequent risk of coronary heart disease: two new prospective studies and a systematic review. PLoS Med 2008; 5(4): e78.

33.

Chang

T-T

Yang

H-Y

Chen

, et al. CCL4 inhibition in atherosclerosis: effects on plaque stability, endothelial cell adhesiveness, and macrophages activation. Int J Mol Sci 2020; 21(18): 6567.

34.

Sprague

Khalil

RA.

Inflammatory cytokines in vascular dysfunction and vascular disease. Biochem Pharmacol 2009; 78(6): 539–552.

35.

Pfluger

Kabra

Aichler

, et al. Calcineurin links mitochondrial elongation with energy metabolism. Cell Metab 2015; 22(5): 838–850.

36.

Cereghetti

Costa

Scorrano

Inhibition of Drp1-dependent mitochondrial fragmentation and apoptosis by a polypeptide antagonist of calcineurin. Cell Death Differ 2010; 17(11): 1785–1794.

37.

Liu

Chang

Dong

MicroRNA-181a-5p curbs osteogenic differentiation and bone formation partially through impairing Runx1-dependent inhibition of AIF-1 transcription. Endocrinol Metab (Seoul) 2023; 38(1): 156–173.

38.

Sikora

Kopeć

Piotrowska

, et al. Role of allograft inflammatory factor-1 in pathogenesis of diseases. Immunol Lett 2020; 218: 1–4.

39.

Rempe

Hartz

AMS

Bauer

Matrix metalloproteinases in the brain and blood-brain barrier: versatile breakers and makers. J Cereb Blood Flow Metab 2016; 36(9): 1481–1507.

40.

Cárcel-Márquez

Cullell

Muiño

, et al. Causal effect of MMP-1 (matrix metalloproteinase-1), MMP-8, and MMP-12 levels on ischemic stroke: a Mendelian randomization study. Stroke 2021; 52(7): e316–e320.

41.

Ghanbari

Yazdanpanah

, et al. Connecting genomics and proteomics to identify protein biomarkers for adult and youth-onset type 2 diabetes: a two-sample Mendelian randomization study. Diabetes 2022; 71(6): 1324–1337.

42.

Wang

Wei

Zhang

Dickkopf-4 is frequently overexpressed in epithelial ovarian carcinoma and promotes tumor invasion. BMC Cancer 2017; 17(1): 455.

43.

Zhang

Yue

Duan

, et al. Joint analysis identified FAP as a prognostic and diagnostic biomarker correlated immune infiltration in gastric cancer. Pathol Res Pract 2023; 245: 154462.

44.

Liu

Han

Wang

, et al. Cancer-associated fibroblasts: an emerging target of anti-cancer immunotherapy. J Hematol Oncol 2019; 12(1): 86.

45.

Tulley

Chen

W-T.

Transcriptional regulation of seprase in invasive melanoma cells by transforming growth factor-β signaling. J Biol Chem 2014; 289(22): 15280–15296.

46.

Huang

Liu

Guo

, et al. The PI3K/AKT pathway in obesity and type 2 diabetes. Int J Biol Sci 2018; 14(11): 1483–1496.

47.

Larsson

Butterworth

Burgess

Mendelian randomization for cardiovascular diseases: principles and applications. Eur Heart J 2023; 44(47): 4913–4924.

48.

Haycock

Burgess

Wade

, et al. Best (but oft-forgotten) practices: the design, analysis, and interpretation of Mendelian randomization studies. Am J Clin Nutr 2016; 103(4): 965–978.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB

0.47 MB