rs13412535 — SERPINE2

SERPINE2 — The Platelet Thrombin Brake Hiding Outside Your Coagulation Panel

Most genetic thrombophilia testing focuses on the coagulation cascade: Factor V Leiden, prothrombin G20210A, antithrombin deficiency. These variants are clinically established but together account for fewer than half of all heritable VTE cases. rs13412535, a regulatory variant in the SERPINE2 gene, points to a parallel fibrinolytic control circuit — one involving Protease Nexin-1 (PN-1), a [serpin | serine protease inhibitor — a family of proteins that regulate proteolytic enzyme activity by acting as decoy substrates] expressed in platelets, vascular cells, and most tissues, but barely detectable in free plasma.

The Mechanism

SERPINE2 encodes Protease Nexin-1 (PN-1)¹¹ Protease Nexin-1 (PN-1)
a 50 kDa glycoprotein in the serpin superfamily, phylogenetically closest to PAI-1 (SERPINE1), the canonical fibrinolysis inhibitor. PN-1 does not simply mimic PAI-1 — it operates through a distinct and broader inhibitory spectrum. PN-1 inhibits thrombin with kinetics enhanced 1,000-fold by heparan sulfate proteoglycans on cell membranes and in extracellular matrix, making it the most efficient tissue-bound thrombin inhibitor known. Beyond thrombin, PN-1 suppresses urokinase plasminogen activator (uPA), tissue plasminogen activator (tPA), and plasmin — the very enzymes responsible for dissolving fibrin clots.

The consequence is profound: platelet PN-1 inhibits both the generation of plasmin by fibrin-bound tPA and the enzymatic activity of fibrin-bound plasmin itself, reducing the rate of fibrin degradation approximately 10-fold²² platelet PN-1 inhibits both the generation of plasmin by fibrin-bound tPA and the enzymatic activity of fibrin-bound plasmin itself, reducing the rate of fibrin degradation approximately 10-fold
Boulaftali et al. 2011 showed that antibody blockade of PN-1 in platelet-rich plasma fully restores tPA-mediated fibrinolysis that PN-1 had suppressed. PN-1 is thus a "lysis-resistance factor" for platelet-rich thrombi — the same type that form in veins under low-flow conditions.

rs13412535 is an intronic regulatory variant in SERPINE2. The A allele is the effect allele identified in the Thibord 2022 cross-ancestry GWAS; the most likely mechanism is altered SERPINE2 transcriptional regulation shifting PN-1 expression in platelets or vascular cells, though the exact cis-regulatory mechanism has not yet been resolved in functional studies.

The Evidence

Thibord et al. 2022 performed a cross-ancestry meta-analysis of 81,669 VTE cases across 30 studies (European, African, and Hispanic ancestry), identifying 135 independent genomic loci associated with VTE risk³³ Thibord et al. 2022 performed a cross-ancestry meta-analysis of 81,669 VTE cases across 30 studies (European, African, and Hispanic ancestry), identifying 135 independent genomic loci associated with VTE risk
The SERPINE2 locus was a novel replicated finding: rs13412535 A allele showed discovery OR 1.06 (p=3.05×10⁻¹⁰) and replication OR 1.08 (p=1.10×10⁻³⁶), with concordant effect direction across ancestry groups. The effect allele frequency was approximately 0.20 globally — meaning this variant is common, not rare, and contributes meaningfully to population-level VTE burden.

Ghouse et al. 2023 extended this to a genome-wide meta-analysis of 81,190 VTE cases and 1,419,671 controls across six cohorts, identifying 93 risk loci (62 previously unreported)⁴⁴ Ghouse et al. 2023 extended this to a genome-wide meta-analysis of 81,190 VTE cases and 1,419,671 controls across six cohorts, identifying 93 risk loci (62 previously unreported)
The study demonstrated that individuals in the top 0.1% of polygenic risk score — built partly on loci like SERPINE2 — face VTE risk comparable to carriers of homozygous or compound heterozygous monogenic thrombophilia mutations. This underscores that common low-effect variants like rs13412535 aggregate into clinically meaningful risk when combined.

The biological plausibility is supported by mouse knockout experiments: PN-1-deficient mice show dramatically accelerated vascular recanalization after tPA treatment — 92% recanalization within one hour versus 15% in wild-type mice — and spontaneously enhanced clot lysis without PN-1⁵⁵ PN-1-deficient mice show dramatically accelerated vascular recanalization after tPA treatment — 92% recanalization within one hour versus 15% in wild-type mice — and spontaneously enhanced clot lysis without PN-1
Boulaftali et al. 2011 used a dorsal skinfold chamber model with real-time intravital microscopy to quantify thrombolysis kinetics; recanalization in PN-1-null mice took 13±2 minutes vs over 60 minutes in wild-type. Increased PN-1 activity — the likely consequence of the A risk allele — mirrors the wild-type phenotype of enhanced clot lysis resistance.

Evidence level is strong: the GWAS finding has been replicated across ancestries at genome-wide significance with concordant effects, and the mechanistic biology of PN-1 in fibrinolytic control is extensively characterized. The specific cis-regulatory mechanism of rs13412535 has not been resolved by eQTL studies, placing it one step below established.

Practical Actions

The per-allele OR of ~1.07 represents a modest individual effect, but the A allele frequency of ~17–23% in Europeans means a large proportion of the population carries at least one copy. For AG heterozygotes, this variant is best understood as a mild modifier of baseline thrombotic risk — most relevant when combined with other thrombophilic variants, traditional VTE risk factors (obesity, immobility, oral contraceptives, surgery, cancer), or family history.

AA homozygotes carry approximately doubled allele dosage. The additive architecture of the association means two copies confer meaningfully higher risk than one — though the absolute risk increment remains smaller than classical thrombophilia mutations like Factor V Leiden or prothrombin G20210A.

This variant is not included on standard thrombophilia panels. Clinicians performing thrombophilia workup after unprovoked VTE will not find it on coagulation assays; awareness requires genetic testing. The fibrinolytic mechanism — PN-1 suppressing tPA and plasmin activity — also implies potentially different implications for thrombolytic therapy response compared to coagulation-pathway variants.

Interactions

rs13412535 converges on the same fibrinolytic arm targeted by Factor V Leiden (rs6025) and prothrombin G20210A (rs1799963), which act through the coagulation cascade upstream. Concurrent Factor V Leiden plus a SERPINE2 fibrinolysis variant creates a mechanistic double-hit: more thrombin generated AND slower clot dissolution. Antithrombin variants (rs121909547 SERPINC1) that reduce thrombin inhibition further compound this picture.

The fibrinolytic suppression mechanism may also interact with SERPINE1 (PAI-1) variants — if both PAI-1 and PN-1 activity are elevated, the combined resistance to plasmin-mediated clot dissolution would be additive.