In Nature Genetics, Klarin et al reported a genome-wide association study (GWAS) of venous thromboembolism
(VTE).[1] The authors identified 33 VTE risk loci, of which 22 were novel. Of special interest
is the genetic overlap with arterial vascular disease and especially peripheral artery
disease (PAD).[1] The genetic overlap between VTE with stroke and coronary artery disease (CAD) was
weaker. Using linkage disequilibrium score regression, a stronger positive correlation
between VTE and PAD (r = 0.47, p = 2.0 × 10−15) than for VTE and CAD (r = 0.27, p = 1.2 × 10−7) or VTE and large artery stroke (r = 0.35, p = 0.02) was observed throughout the genome.[1] This is in agreement with another GWAS study by Klarin et al that identified factor
V Leiden (rs6025) to be a graded risk factor for PAD severity.[2] No association between factor Leiden and CAD or stroke was observed.[2] Interestingly, these findings are also in agreement with previous nationwide epidemiological
register-based family studies in Sweden.[3]
[4]
[5] Family history of VTE is a strong risk factor for VTE in Sweden as determined using
the world's largest family database—the Swedish Multi-generation Register.[6] However, family history of VTE is not a strong risk factor for CAD or ischemic stroke
in Sweden.[3]
[4] Instead, a weak to moderate association is observed between family history of VTE
and PAD in Sweden.[5] Among Swedish families the association between family history of VTE and PAD is
especially strong for arterial thrombosis and embolism in the peripheral vasculature
in agreement with the GWAS studies by Klarin et al.[1]
[2]
[5]
To better understand the genetic overlap reported by Klarin et al[1] for VTE with arterial disease we have performed a Reactome pathway analysis (https://reactome.org) for VTE, PAD, stroke, and CAD.[7] Reactome pathway analyses are used to interpret high-throughput experimental results.
It could also be used to formulate predictive models of normal and disease pathways.[7] We used risk loci (or nearest loci) as reported in GWAS for VTE,[1] PAD,[2] stroke,[8] and CAD.[9] In [Fig. 1] are presented the 25 most relevant pathways identified for VTE, PAD, stroke, and
CAD by a Reactome pathway analysis based on reported genes from GWAS of VTE, PAD,
stroke, and CAD.[1]
[2]
[8]
[9] VTE shares 10 of the 25 (40%) most relevant pathways with PAD. VTE shares two pathways
(8%) with stroke and one (4%) with CAD. PAD shares three (12%) pathways with stroke
and two (8%) with CAD. Stroke shares two (8%) pathways with CAD. This is in agreement
with a previous Swedish nationwide epidemiological register-based study that indicated
that although PAD partially shares familial susceptibility with CAD and stroke, unique
inherited site-specific factors are likely to exist.[10]
Fig. 1 The 25 most relevant Reactome pathways (https://reactome.org)[7] according to genome-wide association study (GWAS) associated genes for venous thromboembolism
(VTE),[1] peripheral artery disease (PAD),[2] stroke,[8] and coronary artery disease (CAD).[9] VTE and PAD share most pathways (10/25 = 40%). Color is graded according to strength
of association.
Of special interest is that all 10 shared pathways for VTE and PAD involve the immune
system and inflammation ([Fig. 1]). Nine of the VTE Reactome pathways are involved in coagulation and three VTE Reactome
pathways involve platelets. Two VTE Reactome pathways involve the complement system
and one VTE Reactome pathway involves mitogen-activated protein kinase signaling for
integrins.
Reactome pathway analysis has not been performed for VTE. Baaten et al performed a
Reactome pathway analysis of experimental arterial thrombosis in mouse models.[11] A limitation of the mouse study was the inclusion of a large set of human genes
that are linked to inherited bleeding disorders, whereas genes related to the vascular
component of thrombosis were limited. Thus, mainly hemostatic pathways were detected.
However, as in the present study cancer pathways were identified ([Fig. 1]).[11] Many studies have linked cardiovascular disorders and cancer.[12]
[13] Reactome pathway analyses have also been performed on gene expression data from
samples of peripheral blood from patients.[14]
[15]
[16] The expression of vascular and liver plasma proteins are not reflected in these
analysis.[14]
[15]
[16] Nair et al focused on expression of inflammatory genes obtained by linking PolySearch
and CADgene databases.[16] A study by Ghosh et al used GWAS data for Reactome pathway analysis of CAD.[17] Several CAD pathways identified by Ghosh et al involve mechanism also observed in
the present study, that is, metabolism (lipids), signal transduction, extracellular
matrix organization, immune system, and transport of small molecules ([Fig. 1]).[17]
A limitation of the Reactome analysis is that the analysis does not take into account
the effect size and the minor allele frequency of respective GWAS variant. Only the
genome-wide significant gene loci were entered directly in the Reactome analysis.
Moreover, in many cases we do not know which gene locus harbors the functional variant
causing the association with VTE, PAD, stroke, and CAD. Still, the Reactome pathway
analysis appears to be fairly valid as coagulation Reactome pathways are mainly related
to VTE and Lipid Reactome pathways to arterial disease (especially CAD) ([Fig. 1]). The following coagulation-related pathways are only associated with VTE: formation
of fibrin clot; gamma-carboxylation of protein precursors; transport of gamma-carboxylated
protein precursors from the endoplasmic reticulum to the Golgi apparatus; intrinsic
pathway of fibrin clot formation; common pathway of fibrin clot formation, gamma carboxylation,
hypusine formation, and arylsulfatase activation; and extrinsic pathway of fibrin
clot formation ([Fig. 1]). Several lipid pathways are only related to CAD: plasma lipoprotein assembly, remodeling,
and clearance; plasma lipoprotein remodeling; NR1H3 and NR1H2 regulate gene expression
linked to cholesterol transport and efflux; plasma lipoprotein clearance; low-density
lipoprotein remodeling; and plasma lipoprotein clearance ([Fig. 1]).
In conclusion, two GWAS by Klarin et al genetically link VTE to arterial disease (especially
PAD).[1]
[2] Our Reactome pathway analysis indicates that especially VTE and PAD share many Reactome
pathways, while less Reactome pathways are shared for VTE with stroke or CAD. The
Reactome pathway analysis not only confirms the involvement of the coagulation system
and platelets in VTE but also indicates the importance of the immune system and inflammation
in both VTE and PAD. The immune system might be a novel target for the development
of new therapeutic agents for VTE and PAD.
