Skip to main content
Download PDF

TIE1 and TEK signalling, intraocular pressure, and primary open-angle glaucoma: a Mendelian randomization study

Journal of Translational Medicine volume 21, Article number: 847 (2023) Cite this article

Abstract

Background

In primary open-angle glaucoma (POAG), lowering intraocular pressure (IOP) is the only proven way of slowing vision loss. Schlemm’s canal (SC) is a hybrid vascular and lymphatic vessel that mediates aqueous humour drainage from the anterior ocular chamber. Animal studies support the importance of SC endothelial angiopoietin-TEK signalling, and more recently TIE1 signalling, in maintaining normal IOP. However, human genetic support for a causal role of TIE1 and TEK signalling in lowering IOP is currently lacking.

Methods

GWAS summary statistics were obtained for plasma soluble TIE1 (sTIE1) protein levels (N = 35,559), soluble TEK (sTEK) protein levels (N = 35,559), IOP (N = 139,555) and POAG (Ncases = 16,677, Ncontrols = 199,580). Mendelian randomization (MR) was performed to estimate the association of genetically proxied TIE1 and TEK protein levels with IOP and POAG liability. Where significant MR estimates were obtained, genetic colocalization was performed to assess the probability of a shared causal variant (PPshared) versus distinct (PPdistinct) causal variants underlying TIE1/TEK signalling and the outcome. Publicly available single-nucleus RNA-sequencing data were leveraged to investigate differential expression of TIE1 and TEK in the human ocular anterior segment.

Results

Increased genetically proxied TIE1 signalling and TEK signalling associated with a reduction in IOP (− 0.21 mmHg per SD increase in sTIE1, 95% CI = − 0.09 to − 0.33 mmHg, P = 6.57 × 10–4, and − 0.14 mmHg per SD decrease in sTEK, 95% CI = − 0.03 to − 0.25 mmHg, P = 0.011), but not with POAG liability. Colocalization analysis found that the probability of a shared causal variant was greater for TIE1 and IOP than for TEK and IOP (PPshared/(PPdistinct + PPshared) = 0.98 for TIE1 and 0.30 for TEK). In the anterior segment, TIE1 and TEK were preferentially expressed in SC, lymphatic, and vascular endothelium.

Conclusions

This study provides novel human genetic support for a causal role of both TIE1 and TEK signalling in regulating IOP. Here, combined evidence from cis-MR and colocalization analyses provide stronger support for TIE1 than TEK as a potential IOP-lowering therapeutic target.

Introduction

Primary open-angle glaucoma is a chronic, degenerative optic neuropathy and represents the leading cause of irreversible blindness worldwide [1, 2]. No curative therapies exist for POAG and so there is a clinical imperative to identify novel efficacious drug targets for its treatment. Pathophysiologically, POAG is characterised by a rise in IOP following increased resistance to aqueous humour drainage from the anterior chamber of the eye, primarily via the trabecular meshwork and Schlemm’s canal (SC), a large lymphatic-like vessel [1]. IOP is an established causal risk factor for POAG and lowering IOP remains the only proven way of slowing visual decline in POAG. Although the involvement of SC in regulating IOP is established, pharmacological therapies targeting SC have yet to be approved for use in clinical practice.

TIE1/TEK signalling in SC has been proposed as a potential therapeutic target in the treatment of elevated IOP [3, 4]. TIE1 and TEK (also known as TIE2) are transmembrane protein receptor tyrosine kinases that are highly expressed in SC endothelial cells [5]. Angiopoietin 1 (ANGPT1) and Angiopoietin 2 (ANGPT2) act as strong and weak agonists, respectively, for the TEK receptor in SC [6]. No direct ligands have been discovered for TIE1 and so it is considered an orphan receptor that interacts functionally with TEK receptor signalling [7, 8]. Historically, animal studies have highlighted the importance of TEK signalling [9,10,11], but recent evidence suggests that TIE1 signalling may also be critical in the development of SC and the maintenance of normal IOP [3]. In humans, previous GWAS studies have identified significant associations between variants in the ANGPT1 and ANGPT2 gene regions with both IOP [12] and POAG [13] but no GWAS associations have been identified in the TIE1 or TEK gene regions. One study highlighted the importance of rare loss-of-function TEK mutations in the development of primary congenital glaucoma but no human genetic evidence for TIE1 signalling in the regulation of IOP has yet been shown. Given the high false discovery rate of animal studies in predicting drug target efficacy in humans, human-centric genetic support for TIE1/TEK signalling should be considered critical in its prioritization as a drug target worth pursuing in large-scale randomized controlled trials (RCTs) [14].

Drug target Mendelian randomization leverages naturally arising human genetic variation to infer the causal effect of a putative drug target on an outcome [15, 16]. Here, ‘cis’ genetic variants located in the vicinity of a gene encoding a protein drug target of interest are used to proxy the pharmacological perturbation of this same drug target. The random allocation of genetic variants from parents to offspring is analogous to random treatment allocation in an RCT and thus, the phenotypic effect of a particular genetic variant should not systematically relate to environmental confounding. Germline genetic variation is non-modifiable and temporally precedes the onset of disease, thus the phenotypic effect of a particular genetic variant is also less susceptible to reverse causation. Given the inherent vulnerability of conventional observational studies to unmeasured confounding and reverse causation, MR can strengthen causal inferences made from observational data. Empirical studies show that supportive genetic evidence increases the probability of successful approval of novel drug targets by twofold [17] and that the impact of prior genetic support is strongest when genetic variants from protein-coding regions of the genome are used [18]. Accordingly, leveraging cis variants that associate with protein expression levels, so-called protein quantitative trait loci (pQTLs), strengthens the use of MR as a tool to investigate putative drug target efficacy.

Genetic variants associated with increased or decreased protein expression of a given receptor can be considered analogous to lifelong exposure to a drug stimulating or inhibiting this same receptor [19]. We sought to mimic the effect of pharmacologically perturbing membrane-bound TIE1 and TEK receptor signalling (i.e., the effect of this signalling pathway in SC) using summary genetic association data for circulating protein levels of sTIE1 and sTEK. sTIE1 is generated upon proteolytic cleavage of the extracellular domain of membrane-bound TIE1 receptor [20] and such TIE1 ectodomain shedding amplifies membrane-bound angiopoietin-TEK signalling [7, 8]. We therefore assumed a positive relationship between circulating sTIE1 levels and membrane-bound TIE1/TEK signalling. Similarly, sTEK is generated from proteolytic cleavage of the extracellular domain of membrane-bound TEK receptors [21, 22]. However, sTEK contains the ligand-binding domain of the receptor and so it binds angiopoietins, inhibiting the activation and phosphorylation of membrane-bound TEK receptors [22, 23]. We therefore assumed a negative relationship between circulating sTEK levels and membrane-bound TEK receptor signalling. Thus, leveraging large-scale summary-based genetic association data for plasma sTIE1 and sTEK protein levels, we used cis-MR and colocalization to infer the causal effect of TIE1 and TEK signalling perturbation on IOP and liability to POAG.

Methods

Study design

A flowchart illustrating the statistical analysis plan is shown in Fig. 1. First, two-sample cis-Mendelian randomization was performed to investigate associations of increased genetically proxied TIE1 and TEK signalling with IOP and liability to POAG. Where a significant MR association was identified, Bayesian colocalization analysis was performed to assess the likelihood that this MR estimate was attributable to genetic confounding through a variant in linkage disequilibrium (LD) with the genetic instrument. Finally, we investigated differential cellular mRNA expression of TIE1 and TEK in the human ocular anterior segment by analysing publicly available single-nucleus transcriptomic data.

Fig. 1
figure 1

Flowchart outlining the study design. Study design and statistical analysis plan. Gene coordinates for drug-target gene region obtained from Ensembl (version 108, genome build GRCh38). MR Mendelian Randomization, GWAS genome-wide association study, LD linkage disequilibrium

Full size image

Data sources

We leveraged GWAS data in European ancestry individuals for all exposures and outcomes. GWAS data for circulating protein levels of sTIE1 and sTEK and corresponding pQTL data were obtained from the deCODE Genetics GWAS of the plasma proteome of 35,559 Icelandic individuals [24]. For IOP, summary statistics were obtained from the largest GWAS of IOP performed to date (N = 139,555), combining data from the UK Biobank (UKBB), EPIC-Norfolk, and the International Glaucoma Genetics Consortium (IGGC) [12]. For POAG, summary statistics were obtained from the largest GWAS meta-analysis of POAG performed in European ancestry individuals (Ncases = 16,677, Ncontrols = 199,580) [13]. A summary of these GWASs is provided in Additional file 1: Table S1. We also leveraged single-nucleus RNA-seq data in six tissues from the van Zyl et al. [25] cell atlas of the human ocular anterior segment. The study was reported in line with the ‘strengthening the reporting of observational studies in epidemiology using mendelian randomization’ (STROBE-MR) guidelines (Additional file 15: Table S15) [26].

Genetic instruments

To proxy TIE1 and TEK signalling perturbation, we selected near-uncorrelated (pairwise LD threshold of r2 < 0.1) variants associated with circulating sTIE1 or sTEK protein levels at genome-wide significance (P < 5 × 10–8) located within the TIE1 and TEK gene regions ± 100kB gene regions (Fig. 1). Publicly available genomic coordinates from Ensembl version 108 with reference to the appropriate reference genome panel for this GWAS (GRCh38/hg38) were used to identify the TIE1 (chr1: 43,300,982–43,323,108) and TEK (chr9: 27,109,141–27,230,176) gene coordinates. The use of a ± 100 kB window reduces the risk of horizontal pleiotropy whilst still allowing variants outside the protein coding region that may regulate protein expression to be sampled. PLINK v2.0 [27] and phase 3, version 5 of the 1000 Genomes European reference panel [28] was used to perform LD clumping, which helps ensure that each instrumental variant represents independent biological signals, thus avoiding overestimation of instrument strength. Further details, including a table of the drug target gene, genomic coordinates, number of instrumental variants, R2 and F-statistics, are reported in Additional file 2: Table S2. The F-statistic quantifies the strength of the relationship between the genetic instrument and the exposure and a value > 10 indicates a strong instrument [29]. The R2 value quantifies the proportion of the variance in the exposure explained by the genetic instrument.

Mendelian randomization

Drug target MR relies on three core assumptions. First, the genetic instrument associates robustly with the drug target (relevance). Second, the genetic instrument shares no common cause with the outcome (independence). Third, the genetic instrument influences the outcome only via its effect on the drug target (exclusion-restriction) [30]. These are illustrated in Fig. 2. MR estimates for genetically proxied TIE1 signalling and TEK signalling were expressed per standard deviation (SD) increase in circulating sTIE1 and per SD decrease in circulating sTEK, respectively.

Fig. 2
figure 2

Direct Acyclic Graph illustrating core MR assumptions. The three core MR assumptions are: (i) the genetic instrument is robustly associated with the drug target, (ii) the genetic instrument is independent of any confounders, (iii) the genetic instrument influences IOP/POAG exclusively via the drug target (or factors downstream of the drug target). In the Wald ratio method, the instrument-outcome association (BGY) is divided by the instrument-exposure association (BGX). IOP intraocular pressure, POAG primary open-angle glaucoma. The broken lines represent forbidden pathways that constitute violations of the MR assumptions

Full size image

Genetic associations were harmonised by aligning effect alleles in both exposure and outcome datasets (Additional file 3: Table S3 and Additional file 4: Table S4). Using Ensembl’s variant effect predictor (VEP) annotation (https://genetics.opentargets.org/), information on how individual pQTLs may affect sTIE1 and sTEK levels based on genomic location are provided in Additional file 5: Table S5 and Additional file 6: Table S6. MR estimates were generated by calculating the Wald-ratio = the variant-outcome regression coefficient divided by the variant-exposure regression coefficient [31]. Where multiple variants comprised the instrument, the inverse-variance weighted (IVW) method was used to generate the MR effect estimate. The IVW approach assumes the absence of any horizontal pleiotropy, and so methods robust to the influence of pleiotropy and violation of the third MR assumption were used as sensitivity analyses [30]. The weighted-median [32], Contamination Mixture [33], MR-Egger [34] and MR-PRESSO [35] methods were applied. We also reported the P-values for the Egger-intercept test and the MR-PRESSO global heterogeneity test. Further details of these methods are provided in Additional file 23. Mendelian randomization analyses were performed using the TwoSampleMR [36] and MendelianRandomization [37] packages in R (version 4.1.2).

Colocalization analysis

Coloc evaluates the likelihood of a shared vs distinct causal variants underlying the drug target gene region and outcome. Approximate Bayes Factors are calculated based on three prior probabilities: p1, p2 and p12, which represent the prior probabilities that any given variant in the drug target gene region (TIE1 and TEK) is associated with either trait 1 (circulating levels of sTIE1/ sTEK), trait 2 (IOP), or both traits, respectively [38]. No single value of p12 is appropriate for all datasets and so we used Giambartolomei et al.’s recommended prior of p12 = 1 × 10–5, as proposed in the original methods paper [38], and p12 = 5 × 10–5, which subsequent studies have shown represents a reasonable balance between false positives and false negatives [39].

A high posterior probability of a shared causal variant (PPshared) relative to the posterior probability of distinct causal variants (PPdistinct) supports a common causal pathway underlying TIE1 or TEK signalling and IOP. In contrast, a higher PPdistinct relative to the PPshared supports distinct causal pathway underlying TIE1 or TEK signalling and IOP, indicating potential confounding by LD in the corresponding cis-MR result. In confounding by LD, two variants in the drug target gene region of interest are in LD or correlated with one another, one associating with the exposure and the other associating with the outcome [40]. The probability of colocalization conditional on the presence of a causal variant can be calculated as PPshared/(PPdistinct + PPshared). Colocalization analyses were performed using the coloc package in R (version 4.1.2). Further details are provided in Additional file 23.

Replication

We replicated our cis-MR and coloc analyses using an independent GWAS for both sTIE1 and sTEK protein levels, Sun et al.’s GWAS of the human plasma proteome [41], which consisted of 3,301 healthy participants from the UK INTERVAL study.

Single cell transcriptomic and differential gene expression analysis

No protein or gene expression quantitative trait loci are currently available for human ocular anterior segment tissues and hence our cis-MR and colocalization analyses could not use tissue-specific genetic association data. We therefore inspected the expression of TIE1 and TEK in the single-nucleus RNA-seq dataset of six tissues (central cornea, corneoscleral wedge, trabecular meshwork/Schlemm’s canal, iris, ciliary body, and lens) from the van Zyl et al. [5, 25] cell atlas of the human ocular anterior segment. We plotted gene expression values (log(TPK + 1)) for TIE1 and TEK across the 39 cell types detected and examined in which cell types TIE1 and TEK were most highly expressed, based on differential gene expression analysis across cell types performed in van Zyl et al. [25]. Only genes expressed in more than 25% of cells in any cell type and with a log2(fold-change) above 0.25 were included in the analysis.

Results

cis-Mendelian randomization

TIE1, TEK and IOP

Increased genetically proxied TIE1 signalling was associated with a 0.21 mmHg lower IOP per SD increase in circulating sTIE1 (95% CI = 0.09 to 0.33 mmHg, P = 6.57 × 10–4) (Fig. 3 and Additional file 7: Table S7). Increased genetically proxied TEK signalling was associated with a 0.14 mmHg lower IOP per SD decrease in circulating sTEK (95% CI = 0.03 to 0.25 mmHg, P = 0.011) (Fig. 3 and Additional file 8: Table S8). Results were consistent across pleiotropy-robust sensitivity analyses (Fig. 3) and were not driven by any one single variant (Additional file 16: Figure S1, Additional file 17: Figure S2, Additional file 18: Figure S3, Additional file 19: Figure S4). Replication using an independent GWAS for circulating sTIE1 and sTEK yielded similar results (Additional file 7: Tables S7 and Additional file 10: Table S10).

Fig. 3
figure 3

Forest plot of cis-MR estimates for increased TIE1 and TEK signalling on IOP. Forest plot illustrating MR estimates for the effect of increased genetically proxied TIE1 and TEK signalling on intraocular pressure, expressed as the mmHg change in IOP per SD increase in circulating sTIE1 and a SD decrease in circulating sTEK, respectively. Dot represents the point estimate and the lines the 95% confidence intervals. Dotted line represents line of null effect

Full size image

TIE1, TEK and POAG

We did not identify an association between increased genetically proxied TIE1 signalling and liability to POAG (OR = 1.04 per SD increase in circulating sTIE1, 95% CI = 0.94 to 1.15, P = 0.479) (Fig. 4 and Additional file 7: Table S7). Results were consistent across pleiotropy-robust sensitivity analyses for TIE1. Similarly, we did not identify an association between increased genetically proxied TEK signalling and liability to POAG (OR = 1.05 per SD decrease in circulating sTEK, 95% CI = 0.97 to 1.14, P = 0.243) (Fig. 4 and Additional file 8: Table S8). However, results were inconsistent across pleiotropy-robust sensitivity analyses (Fig. 4) for TEK, and the Egger intercept test indicated the presence of horizontal pleiotropy (P = 0.042).

Fig. 4
figure 4

Forest plot of cis-MR estimates for increased TIE1 and TEK signalling on liability to POAG. Forest plot illustrating MR estimates for the effect of increased genetically proxied TIE1 and TEK signalling on liability to POAG, expressed as the odds ratio (OR) for POAG per SD increase in circulating sTIE1 and a SD decrease in circulating sTEK, respectively. Dot represents the point estimate and the lines the 95% confidence intervals. Dotted line represents line of null effect

Full size image

Colocalization analysis

TIE1 and IOP

Colocalization analysis supported the presence of a shared causal variant underlying TIE1 receptor signalling and IOP (Table 1 and Additional file 11: Table S11). At both prior probabilities of p12 = 1 × 10–5 and p12 = 5 × 10–5, the posterior probability of sTIE1 and IOP sharing a causal variant, PPshared, was orders of magnitude greater than the posterior probability of distinct causal variants underlying sTIE1 and IOP, PPdistinct, at 53- and 262-fold greater, respectively. The posterior probability of a shared causal variant, conditional on the presence of one or more causal variants, PPshared / (PPdistinct + PPshared), was 0.98 in both cases. Coloc findings were replicated using an independent GWAS for circulating sTIE1 protein levels (Additional file 12: Table S12). Consistent with this, LocusZoom and LocusCompare plots in Figs. 5 and 6, respectively, show a concordant distribution of genetic associations in the TIE1 gene region with both sTIE1 and IOP. Of the variants in the gene region, the variant with the highest PPshared exhibits relatively strong associations with both sTIE1 and IOP, with highly correlated variants exhibiting similarly strong associations and less correlated variants exhibiting progressively weaker associations.

Table 1 Coloc results for TIE1 or TEK signalling and IOP
Full size table
Fig. 5
figure 5

LocusZoom plot for sTIE1 protein levels and IOP in TIE1 gene region. LocusZoom plot of genetic associations of variants within TIE1 gene region with circulating sTIE1 protein levels (A) and IOP (B). The purple diamond is the variant with the highest posterior probability of being the shared causal variant underlying the two traits (PPshared), as determined by coloc. Points are color-coded based on their LD (r2) relative to the variant with the highest PPshared. LocusZoom plot of sTIE1 genetic associations using a 100kB window is shown in Additional file 22: Figure S7

Full size image
Fig. 6
figure 6

LocusCompare plot for sTIE1 protein levels and IOP in the TIE1 gene region. LocusCompare plot illustrating genetic associations within TIE1 gene region between circulating sTIE1 protein levels (x-axis) and IOP (y-axis). Each data point represents a genetic variant. The purple diamond is the genetic variant with the strongest association with sTIE1 protein levels. Points are color-coded based on each variant’s LD (r2) relative to the variant with the highest colocalization posterior probability in the gene region

Full size image

TEK and IOP

Colocalization analysis yielded mixed evidence in support of a shared causal variant underlying TEK receptor signalling and IOP (Table 1 and Additional file 11: Table S11). At p12 = 1 × 10–5, PPdistinct was approximately 2.5-fold that of PPshared; conversely, at p12 = 5 × 10–5, PPshared was approximately twice that of PPdistinct. At these two priors, PPshared/(PPdistinct + PPshared) was 0.30 and 0.68, respectively. Replication using an independent GWAS for circulating sTEK yielded similar results (Additional file 12: Table S12). Consistent with this, LocusZoom and LocusCompare plots do not show a concordance of genetic associations in the TEK gene region for sTEK and IOP (Additional file 20: Figure S5 and Additional file 21: Figure S6).

Single cell expression in six tissues in the anterior segment of the eye

Upon inspection of TIE1 and TEK single nucleus RNA-sequencing expression levels and differential gene expression across 39 cell types in six tissues in the ocular anterior segment (cornea, corneoscleral wedge, trabecular meshwork/Schlemm’s canal, ciliary body, iris, and lens) [25], we found that TIE1 and TEK are preferentially and most highly expressed in SC endothelium, vascular endothelium, and lymphatic endothelium cells (P < 1 × 10–53 for associations) (Fig. 7 and Additional file 13: Table S13).

Fig. 7
figure 7

Single-nucleus RNA sequencing expression of TIE1 and TEK in anterior segment of human ocular tissue. Dot plots illustrating single-nucleus RNA-sequencing expression levels of TIE1 and TEK in the anterior segment of non-diseased human eyes. The size of each dot reflects the fraction of cells expressing mRNA for each gene. The color intensity depicts the average normalized gene expression levels (log(TPK + 1)) in expressing cells for the given cell type. A full list of abbreviations is provided in Additional file 14: Table S14

Full size image

Discussion

We performed cis-MR and colocalization to investigate the causal effect of increased TIE1 and TEK receptor signalling on IOP and liability to POAG. We find novel, human-centric genetic evidence to support a causal effect of increased TIE1/TEK signalling in lowering IOP, with stronger support for TIE1 than TEK.

TIE1 signalling and IOP

We found that increased genetically proxied TIE1 signalling associated with lower IOP. Recent evidence shows that knocking out the full-length TIE1 receptor in mice results in abnormal SC development and raised IOP [3], pointing to an indispensable role for TIE1 signalling in IOP regulation. TIE1 is an orphan receptor that does not directly bind ANGPT1 or ANGPT2 but interacts with TEK via heterodimerization [8]. Molecular studies investigating the interaction between TIE1 and TEK have shown that ectodomain shedding and the release of sTIE1 facilitates ANGPT1-mediated activation of membrane-bound TEK receptors [7, 8]. The significant MR associations and aligned single cell expression patterns reported here for both TIE1 and TEK signalling are consistent with the idea that both receptors act cooperatively to regulate IOP in humans, thereby providing further clarity on the role of TIE1 receptors in relation to that of TEK receptors. Moreover, our cis-MR results were supported by colocalization at both conservative (p12 = 10–5) and liberal priors (p12 = 5 × 10–5), suggesting that our TIE1 MR result is unlikely to be driven by confounding by LD. Even at the more conservative prior, there was a 52-fold higher probability of a shared causal variant as opposed to distinct causal variants underlying both TIE1 signalling and IOP. Thus, building on recent animal data [3], our results provide novel human genetic support for a causal effect of increased TIE1 signalling on IOP-lowering in humans. This is supported by emerging data implicating heterozygous TIE1 mutations in the development of childhood glaucoma [42]. Taken together, these findings raise the possibility that perturbation of TIE1 signalling represents a therapeutic target in addition to, or independently of, TEK receptor perturbation. It should be noted that the absence of a natural endogenous ligand for TIE1 need not preclude its perturbation pharmacologically. For instance, the recreational drug, phenylcyclohexyl piperidine (PCP), binds to an orphan receptor in the NMDA receptor to induce potent dissociative and hallucinogenic effects [43].

TEK signalling and IOP

We found that increased genetically proxied TEK signalling associated with lower IOP, though to a lesser extent than TIE1 signalling. Rare mutations in the TEK gene that lead to loss of TEK protein function were shown to be present in a subset of families with primary congenital glaucoma (PGC) [9]. Previous animal studies suggest a protective role for TEK receptor signalling in regulating IOP and reducing liability to POAG phenotypes. For example, adult mice deficient in both ANGPT1/ANGPT2 or TEK develop impaired SC integrity and retinal ganglion cell dysfunction [10, 44] that can be rescued by TEK reactivation [10]. In humans, previous large-scale GWAS data identified significant associations for variants in the ANGPT1 and ANGPT2 gene regions with IOP [12] and POAG [13], and eQTLs targeting ANGPT1 and ANGPT2 have recently been shown to colocalize with IOP [45]. However, no GWAS associations have previously been observed in the TIE1 or TEK gene regions. Our cis-MR results provide novel human-centric genetic support for a protective effect of TEK receptor signalling in reducing IOP and consistent with this, single cell RNA expression data shows that TEK is preferentially expressed in SC, vascular and lymphatic endothelial cells in the human ocular anterior segment. However, genetic support from colocalization analysis was mixed, with a comparatively higher probability of a shared causal variant at p12 = 5 × 10–5, but a comparatively higher probability of distinct causal variants at p12 = 10–5, indicating potential confounding by LD in our TEK MR result. Interestingly, a recent phase II double-blind RCT in patients with ocular hypertension or POAG found that twice daily application of topical Razuprotafib, which activates TEK receptors via vascular endothelial protein tyrosine phosphatase (VE-PTP) inhibition, reduced IOP when given in addition to latanoprost [46]. The comparatively stronger genetic support for TIE1 perturbation identified in this study, raises the possibility that an IOP-lowering effect may additionally be achieved by drugs stimulating TIE1 receptors.

TIE1, TEK signalling and POAG

In this study, we did not identify an association between genetically proxied TIE1 or TEK signalling with liability to POAG. One possible reason is insufficient power in the present study. The dichotomous nature of POAG outcome data and scope for phenotypic misclassification and heterogeneity amongst POAG cases in the underlying GWAS can reduce power to detect genetic associations with POAG liability. Furthermore, in cis-MR, variant selection is restricted to the vicinity of the drug target gene region and so fewer variants comprise the genetic instrument than in standard MR, where variants are sampled from throughout the entire genome. Whilst the inclusion of trans variants distal to the drug target gene region may increase power, it also significantly increases the risk of horizontal pleiotropy and in turn undermines causal inference [16, 19]. Considering the overall null result, one also cannot rule out the possibility that any beneficial effect of IOP reduction is concurrently opposed by as yet poorly understood deleterious effects of increased TIE1/TEK signalling.

Strengths

This study has several strengths. Prior evidence for functional consequences of altered TIE1/TEK signalling is primarily derived from animal studies. Although such studies are invaluable in understanding the molecular mechanisms via which TIE1/TEK signalling may affect IOP and POAG liability, they exhibit a high false discovery rate—estimated as high as 92.6% [14]—in yielding efficacious drug targets in humans. Furthermore, conventional observational methods investigating the effect of TIE1/TEK signalling are vulnerable to residual confounding and reverse causation. MR leverages the randomly allocated and non-modifiable nature of genetic variants to reduce unmeasured confounding and reverse causation, and thus produce human-centric causal association estimates. Given that individuals are blinded to their genotype, the results are also robust to the influence of ascertainment bias [15]. A growing body of empirical evidence supports the integration of genetics into drug discovery efforts, with prior genetic evidence for a drug target significantly reducing the risk of late-stage failure in subsequent clinical trials in humans [17, 47]. Indeed, genetic associations with complex diseases often correspond to genomic regions enriched for known drug targets of FDA-approved medications for the given disease [48, 49].

The strength of causal inference with MR is contingent on the validity of the core MR assumptions. In this respect, the use of cis-pQTLs at stringent inclusion thresholds of P < 5 × 10–8 increases the likelihood that the instrumental variants functionally relate to TIE1 and TEK signalling, thereby strengthening the biological plausibility of the relevance assumption. With respect to the exclusion-restriction assumption, we leveraged variants within 100kB of TIE1 and TEK gene regions, thereby reducing the likelihood of direct effects of our genetic instrument on IOP or POAG liability not exclusively via TIE1/TEK signalling. The use of cis-pQTLs is also important because the impact of prior genetic support for drug target success is greatest when genetic variants from protein-coding regions of the genome are used [18]. Finally, we replicated our cis-MR results using an independent GWAS for sTIE1 and sTEK protein levels.

Limitations

Our study has some limitations. We assume that plasma levels of sTIE1 and sTEK directly relate to the degree of membrane-bound TIE1/TEK receptor signalling. However, this relationship may be confounded by variability in the levels of ANGPT1 and ANGPT2, the levels of TIE1 and TEK receptor cleavage, and the duration that cleavage products remain in the plasma (as determined by the stability of the cleaved protein and processes removing them from the circulation). Here, we used TIE1/TEK receptor levels as surrogates for TIE1/TEK receptor signalling but future studies could explore perturbation of ANGPT1, the main ligand for TEK receptors. Furthermore, the molecular mechanisms surrounding TIE1 and TEK ectodomain shedding, TIE1/TEK heterodimerization, and their resultant effects on angiopoietin-TEK signalling are still being understood. Consequently, the stronger genetic signal for TIE1 observed here could in part reflect sTIE1 levels being more representative of membrane-bound TIE1 signalling than sTEK levels are of membrane-bound TEK signalling. There is also uncertainty surrounding the extent to which TIE1 perturbation can be modelled as distinct from TEK perturbation, through the use of soluble TIE1 and TEK or otherwise.

MR estimates represent small lifelong genetic differences in TIE1/TEK signalling and so the magnitude of MR estimates are not interpretable on the same scale as those of a discrete clinical intervention, e.g., estimates derived from an RCT investigating TIE1/TEK signalling perturbation. Genetic data for exposures and outcomes were obtained exclusively from individuals of European ancestry to avoid confounding by ancestry and so the extent to which these findings can be generalised to other ancestries remains uncertain. The high combined values for PPH0, PPH1 and PPH2 (see Additional file 9: Table S9 and Additional file 10: Table S10) suggest our coloc analyses may be underpowered and so replication of these analyses once larger GWASs for TIE1, TEK, and IOP are available, is warranted. Nevertheless, replication of colocalization results using two independent plasma proteome GWASs suggests that these results are robust. It should also be noted that the cis-MR and colocalization analyses performed here are not specific to ocular tissue. Neither pQTL nor eQTL data are currently available for human ocular anterior segment tissue. Preferential TIE1 and TEK mRNA expression in cells from aqueous humour outflow pathways in human ocular tissue suggests that it is biologically plausible that the genetic associations for IOP observed here reflect TIE1/TEK signalling in the anterior segment. Nevertheless, our study assumes that pQTLs in the TIE1 and TEK gene regions affect SC protein levels as well systemic (plasma) protein levels, which further studies using pQTLs or eQTLs from ocular anterior segment tissue will be required to confirm.

Conclusion

In this study, we find novel human genetic support for a causal role of both TIE1 and TEK signalling in regulating IOP. Furthermore, combined evidence from Mendelian randomization and colocalization analyses provide stronger support for TIE1 than TEK as a potential IOP-lowering therapeutic target. Further clinical studies investigating this prospect are warranted.

Availability of data and materials

All data sources used in this study are reported in Additional file 1: Table S1. POAG summary data can be downloaded at http://ftp.ebi.ac.uk/pub/databases/gwas/summary_statistics/GCST90011001-GCST90012000/GCST90011766/. sTIE1 and sTEK summary data can be downloaded at https://www.decode.com/summarydata/. IOP summary data may be available on reasonable request.

References

  1. Weinreb RN, Leung CK, Crowston JG, et al. Primary open-angle glaucoma. Nat Rev Dis Primers. 2016;2:16067.

    Article  PubMed  Google Scholar 

  2. Sun Y, Chen A, Zou M, et al. Time trends, associations and prevalence of blindness and vision loss due to glaucoma: an analysis of observational data from the Global Burden of Disease Study 2017. BMJ Open. 2022;12(1): e053805.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Du J, Thomson BR, Onay T, Quaggin SE. endothelial tyrosine kinase tie1 is required for normal schlemm’s canal development-brief report. Arterioscler Thromb Vasc Biol. 2022;42(3):348–51.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Thomson BR, Liu P, Onay T, et al. Cellular crosstalk regulates the aqueous humor outflow pathway and provides new targets for glaucoma therapies. Nat Commun. 2021;12(1):6072.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. van Zyl T, Yan W, McAdams A, et al. Cell atlas of aqueous humor outflow pathways in eyes of humans and four model species provides insight into glaucoma pathogenesis. Proc Natl Acad Sci U S A. 2020;117(19):10339–49.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Thomson BR, Heinen S, Jeansson M, et al. A lymphatic defect causes ocular hypertension and glaucoma in mice. J Clin Invest. 2014;124(10):4320–4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Marron MB, Singh H, Tahir TA, et al. Regulated proteolytic processing of Tie1 modulates ligand responsiveness of the receptor-tyrosine kinase Tie2. J Biol Chem. 2007;282(42):30509–17.

    Article  CAS  PubMed  Google Scholar 

  8. Leppänen VM, Saharinen P, Alitalo K. Structural basis of Tie2 activation and Tie2/Tie1 heterodimerization. Proc Natl Acad Sci U S A. 2017;114(17):4376–81.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Souma T, Tompson SW, Thomson BR, et al. Angiopoietin receptor TEK mutations underlie primary congenital glaucoma with variable expressivity. J Clin Invest. 2016;126(7):2575–87.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Kim J, Park DY, Bae H, et al. Impaired angiopoietin/Tie2 signaling compromises Schlemm’s canal integrity and induces glaucoma. J Clin Invest. 2017;127(10):3877–96.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Thomson BR, Carota IA, Souma T, Soman S, Vestweber D, Quaggin SE. Targeting the vascular-specific phosphatase PTPRB protects against retinal ganglion cell loss in a pre-clinical model of glaucoma. Elife. 2019;8: e48474.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Khawaja AP, Cooke Bailey JN, Wareham NJ, et al. Genome-wide analyses identify 68 new loci associated with intraocular pressure and improve risk prediction for primary open-angle glaucoma. Nat Genet. 2018;50(6):778–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Gharahkhani P, Jorgenson E, Hysi P, et al. Genome-wide meta-analysis identifies 127 open-angle glaucoma loci with consistent effect across ancestries. Nat Commun. 2021;12(1):1258.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Hingorani AD, Kuan V, Finan C, et al. Improving the odds of drug development success through human genomics: modelling study. Sci Rep. 2019;9(1):18911.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Smith GD, Ebrahim S. “Mendelian randomization”: can genetic epidemiology contribute to understanding environmental determinants of disease? Int J Epidemiol. 2003;32(1):1–22.

    Article  PubMed  Google Scholar 

  16. Gill D, Georgakis MK, Walker VM, et al. Mendelian randomization for studying the effects of perturbing drug targets. Wellcome Open Res. 2021;6:16.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Nelson MR, Tipney H, Painter JL, et al. The support of human genetic evidence for approved drug indications. Nat Genet. 2015;47(8):856–60.

    Article  CAS  PubMed  Google Scholar 

  18. King EA, Davis JW, Degner JF. Are drug targets with genetic support twice as likely to be approved? Revised estimates of the impact of genetic support for drug mechanisms on the probability of drug approval. PLoS Genet. 2019;15(12): e1008489.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Schmidt AF, Finan C, Gordillo-Marañón M, et al. Genetic drug target validation using Mendelian randomisation. Nat Commun. 2020;11(1):3255.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Korhonen EA, Lampinen A, Giri H, et al. Tie1 controls angiopoietin function in vascular remodeling and inflammation. J Clin Invest. 2016;126(9):3495–510.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Reusch P, Barleon B, Weindel K, et al. Identification of a soluble form of the angiopoietin receptor TIE-2 released from endothelial cells and present in human blood. Angiogenesis. 2001;4(2):123–31.

    Article  CAS  PubMed  Google Scholar 

  22. Alawo DOA, Tahir TA, Fischer M, Bates DG, Amirova SR, Brindle NPJ. Regulation of angiopoietin signalling by soluble tie2 ectodomain and engineered ligand trap. Sci Rep. 2017;7(1):3658.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Findley CM, Cudmore MJ, Ahmed A, Kontos CD. VEGF induces Tie2 shedding via a phosphoinositide 3-kinase/Akt dependent pathway to modulate Tie2 signaling. Arterioscler Thromb Vasc Biol. 2007;27(12):2619–26.

    Article  CAS  PubMed  Google Scholar 

  24. Ferkingstad E, Sulem P, Atlason BA, et al. Large-scale integration of the plasma proteome with genetics and disease. Nat Genet. 2021;53(12):1712–21.

    Article  CAS  PubMed  Google Scholar 

  25. van Zyl T, Yan W, McAdams AM, Monavarfeshani A, Hageman GS, Sanes JR. Cell atlas of the human ocular anterior segment: Tissue-specific and shared cell types. Proc Natl Acad Sci U S A. 2022;119(29): e2200914119.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Skrivankova VW, Richmond RC, Woolf BAR, et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ. 2021;375: n2233.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559–75.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. 1000 Genomes Project Consortium, Auton A, Brooks LD, et al. A global reference for human genetic variation. Nature. 2015;526(7571):68–74.

    Article  Google Scholar 

  29. Burgess S, Thompson SG, CRP CHD Genetics Collaboration. Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol. 2011;40(3):755–64.

    Article  PubMed  Google Scholar 

  30. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–65.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Burgess S, Small DS, Thompson SG. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res. 2017;26(5):2333–55.

    Article  PubMed  Google Scholar 

  32. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304–14.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Burgess S, Foley CN, Allara E, Staley JR, Howson JMM. A robust and efficient method for Mendelian randomization with hundreds of genetic variants. Nat Commun. 2020;11(1):376.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(8):1196.

    Article  CAS  PubMed  Google Scholar 

  36. Hemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7: e34408.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Yavorska OO, Burgess S. MendelianRandomization: an R package for performing Mendelian randomization analyses using summarized data. Int J Epidemiol. 2017;46(6):1734–9.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Giambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, Plagnol V. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10(5): e1004383.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Wallace C. Eliciting priors and relaxing the single causal variant assumption in colocalisation analyses. PLoS Genet. 2020;16(4): e1008720.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Zuber V, Grinberg NF, Gill D, et al. Combining evidence from Mendelian randomization and colocalization: Review and comparison of approaches. Am J Hum Genet. 2022;109(5):767–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Sun BB, Maranville JC, Peters JE, et al. Genomic atlas of the human plasma proteome. Nature. 2018;558(7708):73–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Young TL, Whisenhunt K, Martin S, et al. TIE1 gene variants identified in patients with childhood glaucoma. Invest Ophthalmol Vis Sci. 2023;64(8):4352.

    Google Scholar 

  43. Fagg GE. Phencyclidine and related drugs bind to the activated N-methyl-D-aspartate receptor-channel complex in rat brain membranes. Neurosci Lett. 1987;76(2):221–7.

    Article  CAS  PubMed  Google Scholar 

  44. Thomson BR, Souma T, Tompson SW, et al. Angiopoietin-1 is required for Schlemm’s canal development in mice and humans. J Clin Invest. 2017;127(12):4421–36.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Hamel AR, Rouhana JM, Yan W, et al. Using expression and splicing QTLs to identify new regulatory mechanisms and causal genes for Primary Open Angle Glaucoma. medRxiv. 2022. https://doi.org/10.1101/2022.05.14.22275022.

    Article  Google Scholar 

  46. Brigell M, Withers B, Buch A, Peters KG. Tie2 activation via VE-PTP inhibition with razuprotafib as an adjunct to latanoprost in patients with open angle glaucoma or ocular hypertension. Transl Vis Sci Technol. 2022;11(1):7.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Ochoa D, Karim M, Ghoussaini M, Hulcoop DG, McDonagh EM, Dunham I. Human genetics evidence supports two-thirds of the 2021 FDA-approved drugs. Nat Rev Drug Discov. 2022;21(8):551.

    Article  CAS  PubMed  Google Scholar 

  48. Segrè AV, Wei N, Altshuler D, Florez JC, DIAGRAM Consortium, MAGIC Investigators. Pathways targeted by antidiabetes drugs are enriched for multiple genes associated with type 2 diabetes risk. Diabetes. 2015;64(4):1470–83.

    Article  PubMed  Google Scholar 

  49. Plenge RM, Scolnick EM, Altshuler D. Validating therapeutic targets through human genetics. Nat Rev Drug Discov. 2013;12(8):581–94.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

SR is supported by the National Institute for Health and Care Research (NIHR) Academic Foundation Programme. AVS is supported by the National Eye Institute (NEI) (R01EY0314240, P30EY014104). DG is supported by the British Heart Foundation Centre of Research Excellence at Imperial College London (RE/18/4/34215). APK is supported by a UK Research and Innovation Future Leaders Fellowship, an Alcon Research Institute Young Investigator Award, and a Lister Institute Fellowship. NZ is supported by a National Institutes of Health (NIH) K23 Career Development Award (K23EY132634), NIH R21 Exploratory. Developmental Research Grant Award (1R21EY032953) and Research to Prevent Blindness Career Development Award. JLW is supported in part by the NEI (R01EY02928, R01EY022305, R01EY031820, R01EY032559 and P30EY032559). The funding organizations had no role in the design or conduct of the study; data collection, management, analysis, or interpretation; manuscript preparation, review, or approval; the decision to submit the manuscript for publication.

Author information

Author notes

  1. Ayellet V. Segrè and Janey Wiggs—Joint senior authors.

Authors and Affiliations

  1. Faculty of Medicine, Imperial College London, London, UK

    Skanda Rajasundaram & Ayellet V. Segrè

  2. Department of Ophthalmology, Harvard Medical School, Boston, MA, USA

    Skanda Rajasundaram, Nazlee Zebardast, Katherine Duchinski, Ayellet V. Segrè & Janey Wiggs

  3. Ocular Genomics Institute, Massachusetts Eye and Ear, Boston, MA, USA

    Nazlee Zebardast, Katherine Duchinski, Ayellet V. Segrè & Janey Wiggs

  4. Broad Institute of MIT and Harvard, Cambridge, MA, USA

    Puja Mehta, Katherine Duchinski & Janey Wiggs

  5. NIHR Biomedical Research Centre, Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, London, UK

    Puja Mehta

  6. UCL Institute of Cardiovascular Science, London, UK

    Puja Mehta

  7. MRC Biostatistics Unit, University of Cambridge, Cambridge, UK

    Anthony P. Khawaja, Alasdair Warwick & Stephen Burgess

  8. Department of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London, UK

    Alasdair Warwick & Dipender Gill

Authors
  1. Skanda Rajasundaram
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Nazlee Zebardast
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Puja Mehta
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Anthony P. Khawaja
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Alasdair Warwick
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Katherine Duchinski
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Stephen Burgess
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Dipender Gill
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Ayellet V. Segrè
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Janey Wiggs
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Concept and design: SR, AVS, JLW; Analysis: SR, PM, AVS; Interpretation: SR, AVS, DG, APK, NZ and JLW; Drafting of the manuscript: SR; Critical revision of the manuscript for intellectual content: all authors; Supervision: AVS, NZ, DG and JLW.

Corresponding author

Correspondence to Skanda Rajasundaram.

Ethics declarations

Ethics approval and consent to participate

Written informed consent was obtained for all participants in the original GWASs and the studies were granted ethical approval from the relevant ethics review board.

Consent for publication

Not applicable.

Competing interests

DG is employed by Novo Nordisk. APK has acted as a consultant to Abbvie, Aerie, Google Health, Novartis, Reichert, Santen, and Thea, and has received lecturing fees from Heidelberg Engineering. JLW has consulted for Allergan, Editas, Maze, Regnexbio and Broadwing Bio. No other disclosures are reported.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Table S1.

Summary of Data Sources.

Additional file 2: Table S2.

Drug target gene regions, number of instrumental variants, R2 and F-statistics.

Additional file 3: Table S3.

Table of Mendelian randomization instrumental variants for TIE1 signalling perturbation.

Additional file 4: Table S4.

Table of Mendelian randomization instrumental variants for TEK signalling perturbation.

Additional file 5: Table S5.

Table of genomic location of instrumental variants for TIE1 pQTLs.

Additional file 6: Table S6.

Table of genomic location of instrumental variants for TEK pQTLs.

Additional file 7: Table S7.

Mendelian randomization estimates for the effect of increased genetically predicted TIE1 signalling (using deCODE Genetics GWAS of plasma proteome (N = 35,559)) on IOP and POAG.

Additional file 8: Table S8.

Mendelian randomization estimates for the effect of increased genetically predicted TEK signalling (using deCODE Genetics GWAS of plasma proteome (N = 35,559)) on IOP and POAG.

Additional file 9: Table S9.

Mendelian randomization estimates for the effect of increased genetically predicted TIE1 signalling (using Sun et al. [41] GWAS of plasma proteome (N = 3,301)) on IOP and POAG.

Additional file 10: Table S10.

Mendelian randomization estimates for the effect of increased genetically predicted TEK signalling (using Sun et al. [41] GWAS of plasma proteome (N = 3,301)) on IOP and POAG.

Additional file 11: Table S11.

Coloc analysis for Tie1, TEK and IOP using deCODE GWAS of plasma proteome (N = 35,559).

Additional file 12: Table S12.

Colocalization analysis for Tie1, TEK and IOP using Sun et al. [41] GWAS of plasma proteome (N = 3,301).

Additional file 13: Table S13.

Cell types with significant cell type specific expression for TIE1 and TIE2 in single-nucleus RNA sequencing data of the anterior segment of health human ocular tissues.

Additional file 14: Table S14.

List of cell types identified in six anterior segment tissues from healthy human eyes based on clustering of single-nucleus RNA-sequencing data.

Additional file 15: Table S15.

STROBE-MR Checklist.

Additional file 16: Figure S1.

Leave-one-out plot for MR estimate of genetically proxied TIE1 and IOP.

Additional file 17: Figure S2.

Leave-one-out plot for MR estimate of genetically proxied TIE1 and IOP.

Additional file 18: Figure S3.

Scatter plot of the genetic associations of instrumental variants with sTIE1 and intraocular pressure.

Additional file 19: Figure S4.

Scatter plot of the genetic associations of instrumental variants with sTEK and intraocular pressure.

Additional file 20: Figure S5.

LocusZoom plot of genetic associations with sTEK protein levels and IOP in TEK gene region.

Additional file 21: Figure S6.

LocusCompare plot of genetic associations with sTEK protein levels and IOP in the TEK gene region.

Additional file 22: Figure S7.

LocusZoom plot of genetic associations with sTIE1 protein levels in the TIE1 gene region with 100kB window.

Additional file 23: Methods S1.

Supplementary Methods.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Rajasundaram, S., Zebardast, N., Mehta, P. et al. TIE1 and TEK signalling, intraocular pressure, and primary open-angle glaucoma: a Mendelian randomization study. J Transl Med 21, 847 (2023). https://doi.org/10.1186/s12967-023-04737-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12967-023-04737-9

Keywords

Journal of Translational Medicine

ISSN: 1479-5876

Contact us