Overexpression of the Cell Cycle Inhibitor p16^INK4a Promotes a Prothrombotic Phenotype Following Vascular Injury in Mice

Abstract

Objective

Age-associated cellular senescence is thought to promote vascular dysfunction. p16^INK4a is a cell cycle inhibitor that promotes senescence and is upregulated during normal aging. In this study, we examine the contribution of p16^INK4a overexpression on venous thrombosis.

Methods and Results

Mice overexpressing p16^INK4a were studied with four different vascular injury models: (1) ferric chloride (FeCl₃) and (2) Rose Bengal to induce saphenous vein thrombus formation; (3) FeCl₃ and vascular ligation to examine thrombus resolution; and (4) LPS administration to initiate inflammation-induced vascular dysfunction. p16^INK4a transgenic mice had accelerated occlusion times (13.1 ± 0.4 min) compared to normal controls (19.7 ± 1.1 min) in the FeCl₃ model and 12.7 ± 2.0 and 18.6 ± 1.9, respectively in the Rose Bengal model. Moreover, overexpression of p16^INK4a delayed thrombus resolution compared to normal controls. In response to LPS treatment, the p16^INK4a transgenic mice showed enhanced thrombin generation in plasma-based calibrated automated thrombography (CAT) assays. Finally, bone marrow transplantation studies suggested increased p16^INK4a expression in hematopoietic cells contributes to thrombosis, demonstrating a role for p16^INK4a expression in venous thrombosis.

Conclusions

Venous thrombosis is augmented by overexpression of the cellular senescence gene p16^INK4a.

Introduction

Aging is an important risk factor for developing cardiovascular disease, and also the least understood^1,2. Venous thromboembolism (VTE) is characterized by the development of thrombi in the deep veins of the legs, which are prone to dislodging and embolizing to the lungs. This condition accounts for 140,000–200,000 deaths each year in the U.S^3,4. The risk of developing VTE substantially increases with age and individuals over the age of 55 years have an annual incidence 5–7 times higher than young adults⁵. While VTE in the younger population is often explained by mutations in hemostatic genes, mechanisms behind the increased risk of VTE in the elderly are less well understood.

Senescence is one cellular phenomenon known to be associated with aging. Cellular senescence is a stress-induced process, controlled by cell cycle inhibitors, which promotes an irreversible growth arrest^6–9. p16^INK4a, a cell cycle inhibitor that promotes senescence, binds to cyclin-dependent kinases 4 and 6 to disrupt phosphorylation of the retinoblastoma protein, causing a G1 cell cycle arrest¹⁰. Expression of p16^INK4a increases with age in many tissues and is a biomarker of aging^11–14. Furthermore, p16^INK4a expression correlates with biomarkers of senescence, such as senescence-associated β-galactosidase expression, and expression is associated with gerontogenic activities such as smoking, physical inactivity and ad lib feeding in humans or mice^11,14. In some tissues such as pancreatic beta cells, neural stem cells, and hematopoietic stem or progenitor cells, the age-induced increase in p16^INK4a expression is associated with reduced cellular proliferation coupled with an impaired tissue response to injury^15–17. Additionally, senescent cells are thought to contribute to aging pathology through the production of cytokines (IL-6) that further promote inflammation and cellular dysfunction¹⁸.

The contribution of senescence to disease in the venous circulation, and how this may be involved in age-related VTE or a possible prothrombotic phenotype, remains largely uncharacterized. The aim of this study was to ascertain whether overexpression of p16^INK4a modified venous thrombus formation in several well-defined animal models. Our results demonstrate that p16^INK4a overexpression augments vascular occlusion and delayed thrombus resolution relative to wild-type controls. Furthermore, p16^INK4a transgenic mice display enhanced thrombin generation, increased thrombin-antithrombin (TAT), and increased PAI-1 levels when exposed to low-dose lipopolysaccharide (LPS). Additionally, bone marrow transplantation between wild-type and p16^INK4a transgenic mice demonstrated a substantial contribution of hematopoietic cells to this phenotype. Overall, these results show that expression of p16^INK4a is involved in promoting a prothrombotic environment in the venous vasculature.

Methods

A detailed description of the methods are presented in the supplemental materials.

Mice

All animal procedures were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee, UNC-Chapel Hill. The bacterial artificial chromosome (BAC) transgenic mice overexpressing p16^INK4a used in this study have been previously described¹⁵.

Hemostasis Model

Hemostasis was assessed as previously described¹⁹. Briefly, wild-type and p16^INK4a transgenic mice at 8 weeks of age were anesthetized with 2.5% tribromoethanol (Sigma Aldrich – St. Louis, MO, T48402) at 0.1 mL/g body weight. The saphenous vein of anesthetized mice was exposed and transected with a 23-G needle. Once bleeding stopped, a longitudinal cut was made in the vessel and the blood gently wiped away with kimwipes (Kimberly – Clark - Roswell, GA) until clotted. The blood clot was disrupted using a 30-G needle and the blood gently wiped away. Clot disruption was repeated every time hemostasis occurred and each hemostatic event was recorded using Chart software for 20 minutes.

FeCl₃ Vascular Injury

The saphenous vein thrombosis model was performed as previously described¹⁹. Briefly, the saphenous veins of anesthetized mice was exposed and dissected away from the saphenous artery. A 0.5 × 2 mm piece of filter paper was soaked in 2.5% (N=4), 5% (N=5) or 10% (N=4) FeCl₃ (Sigma Aldrich – F7134), and laid over the saphenous vein for 3 minutes. The filter paper was then removed and the tissue was washed 3 times with warm saline. Blood flow was monitored using a 20-MHz Doppler flow probe (Indus Instruments – Webster, TX). Occlusion is defined as the absence of blood flow for one minute. The time to flow restriction was defined as the time after injury at first cessation of blood flow.

Rose Bengal Photochemical Vascular Injury

Photochemical injury was performed as previously described²⁰. Briefly, both right and left saphenous veins of anesthetized mice were exposed. The left saphenous vein was catheterized using catheters made in-house using pulled PE-10 tubing (Braintree Scientific - Braintree, MA). Rose Bengal (Sigma Aldrich – R-3877), diluted to 30 mg/mL in normal saline, was infused through the catheter at a dose of 75 mg/kg through a gastight syringe (Hamilton Co. - Reno,NV). Prior to infusion of Rose Bengal, a 1.75 mW green light (540 nm) (Prizmatix – Southfield, MI) was directed 0.5 cm over the injury site on the right saphenous vein. Light was applied to the vessel until a stable thrombus (defined as the absence of blood flow for 1 minute) was achieved.

Thrombus resolution

We developed a new method to measure thrombus resolution using the saphenous vein. Wild-type and p16^INK4a transgenic mice (n=3 per genotype at each time point) were subjected to 10% FeCl₃ injury to the saphenous vein. The tissue was then washed 3 times with warm saline and a single ligature was placed upstream of the thrombus using a 8-0 monofilament polypropylene suture to prevent embolization and the leg was sutured closed. Mice were sacrificed at various time points and the saphenous neurovascular bundle was removed and fixed overnight in 4% paraformaldehyde and paraffin embedded. Five micron sections were cut and hematoxylin and eosin (H&E) stained to visualize the presence of a thrombus under light microscopy. Vessels were sectioned through and those sections showing the greatest area of occlusion were chosen for analysis. Such sections typically occurred near the center of the injured vessel. Images were analyzed using ImageJ software to calculate the percent of the vessel lumen that remained occluded by a thrombus.

Low Dose Lipopolysaccharide (LPS) Treatment

Wild-type and p16^INK4a transgenic mice were treated with 2mg/kg intraperitoneal injection of LPS (Sigma L3012). At various times (1, 3, and 5 hours), the mice (n=5 per genotype each time point) were anesthetized and 1mL of blood was collected from the inferior vena cava (IVC) into 3.8% sodium citrate at a ratio of 1:9 using a 25 gauge needle. Whole blood was spun at 4,000 × g for 15 minutes and the platelet poor plasma was collected and stored at −80°C until analyzed.

Bone marrow transplantation

This procedure was performed as described previously²¹. Briefly, mice were irradiated using a Cesium¹³⁷ irradiator (JL Shepherd, San Fernando, CA) with a total of 11 Gy (two doses of 550 rad, with a 4 hour rest) to abolish endogenous hematopoietic cells. Bone marrow cells were isolated from donor mice²¹ and 1 × 10⁷ cells were injected (100 µL) into the retro-orbital sinus. Four weeks after irradiation, recipient mice underwent FeCl₃ injury to the saphenous vein to determine vascular occlusion times. At termination, recipient mouse bone marrow was genotyped to verify successful repopulation of donor cells by polymerase chain reaction. Expression of p16^INK4a was compared to an IL-2 loading control.

Statistics

All statistical analyses were performed with GraphPad Prism. All measurements are respresented as the mean ± standard error of the mean (SEM). One-way ANOVA or Students T-test were performed where indicated. Values of p<0.05 were considered statistically significant.

Results

p16^INK4a transgenic and wild-type mice respond similarly in a hemostasis model

To determine the contribution of p16^INK4a overexpression on potential hemostatic defects, mice initially underwent a model of saphenous vein hemostasis. No difference was observed in the number of hemostatic clots formed over 20 minutes between transgenic (25.8 ± 3.4) and wild-type mice (25.8 ± 2.1, Figure 1A) or in the average time to hemostasis (33.5 ± 3.7 sec and 36.6 ± 2.6 sec, respectively, Figure 1B). Furthermore, no significant differences were observed in body weight, venous blood flow velocity, plasma prothrombin time (PT) and complete blood count (CBC) between the two groups of mice (Supplemental Table 1). These results suggest there is no obvious physical or hematologic phenotype in the p16^INK4a transgenic mice at the ages studied.

Figure 1. Hemostatic Parameters in Saphenous Vein Hemostasis Model.

Hemostatic measurements were compared between wild-type and p16^INK4a transgenic mice following blunt injury to the saphenous vein. (A) The number of hemostatic clots formed and (B) the average time to stoppage of bleeding following serial clot disruption over 30 minutes. Data not statistically significant.

p16^INK4a transgenic mice display a prothrombotic phenotype in a FeCl₃ injury model

FeCl₃ injury is a well-established mechanism for inducing thrombus formation in vivo^22–24. We first demonstrated a dose-dependent effect of FeCl₃ on the occlusion time in the saphenous vein, exposing wild-type mice to 2.5, 5, and 10% FeCl₃ injuries to the saphenous vein (Figure 2A).

Figure 2. Overexpression of p16^INK4a Decreases Time to Occlusion in FeCl₃ Vascular Injury Model.

(A) Wild-type mice were subjected to increasing doses of FeCl₃ to the saphenous vein to determine a dose-dependent effect on the occlusion time as described in the Materials and Methods *p<0.01 versus 2.5% FeCl₃. (B) Vascular occlusion times were compared between wild-type and p16^INK4a transgenic mice after 5% FeCl₃ injury to the saphenous vein. The occlusion time represents the amount of time required to form an occlusive thrombus as described in the Materials and Methods. * denotes p<0.01 versus wild-type control by student t-test.

Wild-type and p16^INK4a transgenic mice were then subjected to FeCl₃ (5%) injury to the saphenous vein. The p16^INK4a transgenic mice showed a significantly faster time to occlusion (13.1 ± 0.4 min) compared to wild-type mice (19.7 ± 1.1 min, Figure 2B). Furthermore, the time to flow restriction was also measured. The p16^INK4a transgenic mice demonstrated faster times to flow restriction (6.4 ± 0.91) compared to wild-type controls (8.7 ± 0.54, p<0.05, data not shown). These results indicate overexpression of p16^INK4a results in a prothrombotic phenotype following vascular injury.

p16^INK4a transgenic mice display a prothrombotic phenotype in a photochemical injury model

The excitation of Rose Bengal to induce photochemical injury is another well-established mechanism for inducing thrombus formation in vivo^25–27. Upon photochemical injury to the saphenous vein, p16^INK4a transgenic mice displayed a significantly faster time to occlusion (12.7 ± 2.0 min) when compared to wild-type mice (18.6 ± 1.9 min, Figure 3). These results suggest the prothrombotic phenotype in mice overexpressing p16^INK4a can be recapitulated in other vascular injury models.

Figure 3. p16INK4a Transgenic Mice have Decreased Time to Occlusion in Rose Bengal Photochemical Vascular Injury Model.

Vascular occlusion times between wild-type and p16^INK4a transgenic mice were compared following saphenous vein injury with 75 mg/kg Rose Bengal excited with 1.75 mW green light at 540 nm. Occlusion times represent the amount of time required to form an occlusive thrombus. *denotes p<0.05 versus wild-type control by student t-test.

p16^INK4a transgenic mice exhibit delayed thrombus resolution

Venous thrombosis is characterized by the presence of unresolved thrombi in the lower extremities. In order to study the effect of overexpressing p16^INK4a on thrombus resolution, thrombi formed post-FeCl₃ injury in wild-type and p16^INK4a transgenic mice were monitored over time. Mice were euthanized from 1 hour to 15 days post-FeCl₃ injury and vascular ligation. No significant differences in thrombus resolution were observed until 7 days after vascular injury. By 10 days post-injury, all wild-type mice exhibited complete thrombus resolution, whereas p16^INK4a transgenic mice maintained an average of 60% vessel occlusion. p16^INK4a transgenic mice required additional time post-injury for thrombus resolution relative to wild-type controls (Figure 4A). Representative images show little difference in percent occlusion at one day (Figure 4B) between p16^INK4a transgenic and wild-type mice. Black staining represents FeCl₃ trapped within the thrombus. At 10 days, we observed that residual FeCl₃ was mostly contained within inflammatory macrophages and was present in the perivascular space of wild-type mice. However, residual FeCl₃ contained within macrophages was still present in the intravascular space of p16^INK4a transgenic mice at 10days (Figure 4B). These results demonstrate a defect in thrombus resolution with p16^INK4a overexpression.

Figure 4. p16^INK4a Transgenic Mice Display Defective Thrombus Resolution.

(A) Thrombus resolution was measured over time after 10% FeCl₃ injury to the saphenous vein and stasis induced by ligation as described in the Materials and Methods. *denotes p<0.05 versus respective wild-type control by student t-test. (B) Representative histologic images were analyzed using ImageJ software to determine percent occlusion (plotted in A). (a) Wild-type at 1 day, (b) p16 Transgenic at 1 day, (c) Wild-type at 10 days, (d) p16 Transgenic at 10 days.

ν - Wild-Type, σ - p16^INK4a Transgenic

p16^INK4a transgenic mice display enhanced thrombin generation in response to LPS challenge

Chronic inflammation and endothelial dysfunction have been linked to enhanced thrombin generation and the risk of venous thrombosis^28–30. LPS is known to activate the vascular endothelium and promote the formation of spontaneous thrombi^31–34. To study the effects of inflammation-induced coagulation, we exposed p16^INK4a transgenic and wild-type mice to low dose LPS. When analyzed by calibrated automated thrombography (CAT), plasma from the p16^INK4a transgenic mice showed significantly shorter lagtime to initiation of thrombin generation and time to peak amount of thrombin generated at all time points post LPS treatment (Table 1). The peak amount of thrombin generated was significantly higher in p16^INK4a transgenic mice 3 and 5 hours after LPS treatment (Table 1). The observed differences in thrombin generation demonstrates p16^INK4a transgenic mice are able to generate more thrombin and have a prothrombotic phenotype when challenged with LPS.

Table 1.

Thrombin Generation in Wild-type (WT) vs p16^INK4a Transgenic (Tg) Mouse Plasma after LPS Treatment^*

Time (hrs)	Lagtime (min)		Peak Height (nM)		Time to Peak (min)
	WT	p16INK4a Tg	WT	p16INK4a Tg	WT	p16INK4a Tg
0	2.17 ± 0.19	2.33 ± 0.24	46.16 ± 0.53	49.15 ± 1.98	5.33 ± 0.87	5.21 ± 0.62
1	3.4 ± 0.16	1.85 ± 0.12*	45.44 ± 9.45	44.21 ± 1.3	6.19 ± 0.27	4.52 ± 0.14*
3	2.85 ± 0.11	1.85 ± 0.09*	36.44 ± 0.43	51.26 ± 0.63*	6.07 ± 0.16	4.41 ± 0.16*
5	2.01 ± 0.17	2.51 ± 0.19*	32.06 ± 0.46	39.55 ± 0.67*	5.18 ± 0.23	5.52 ± 0.11*

^*Thrombin generation in plasma from mice treated with 2mg/kg LPS was measured using CAT, as described in the Supplemental Materials and Methods. Data represents experiments performed in duplicate with 5 mice per group per time point.

^*p<0.05

p16^INK4a expression in hematopoietic cells contributes to the observed prothrombotic phenotype

In order to determine the relative contribution of p16^INK4a expression in the hematopoietic cell compartment to the observed prothrombotic phenotype, bone marrow transplants were performed between transgenic and wild-type mice. Following transplantation and recovery, mice were subjected to 10% FeCl₃ injury to the saphenous vein. Consistent with our previous results (Figure 2B), transgenic mice receiving transgenic bone marrow had retained their significantly reduced occlusion time (8.4 ± 0.48 min) when compared to wild-type mice receiving wild-type bone marrow (13.0 ± 0.79 min, Figure 5). Interestingly, transgenic mice receiving wild-type bone marrow displayed occlusion times similar to wild-type mice (12.8 ± 1.3 min), whereas wild-type mice receiving transgenic bone marrow displayed occlusion times similar to transgenic mice (9.5 ± 0.61 min, Figure 5). PCR results confirmed successful reconstitution by donor bone marrow cells (Supplemental Figure 3). These results demonstrate that the effects of p16^INK4a overexpression are, at least in part, mediated by hematopoietic cells.

Figure 5. Vascular Occlusion Times are Altered by Bone Marrow Transplantation.

Vascular occlusion times were compared between cohorts of bone marrow transplanted mice following injury of the saphenous vein with 10% FeCl₃. Occlusion time represent the amount of time required to form an occlusive thrombus. *denotes p<0.05 versus wild-type control cohort. ** denotes p<0.05 versus transgenic control cohort. # denotes p=0.08 versus wild type control cohort. Statistical relevance was determined by one-way ANOVA with Tukey’s post-hoc analysis.

Discussion

Senescence is a complex process that is thought to contribute to cardiovascular pathologies associated with aging. Despite several reviews describing prothrombotic changes in senescent vascular endothelial cells^35,36, no studies have described a venous thrombotic phenotype in mice overexpressing senescence-promoting genes. In the current study, we examined parameters that define venous thrombotic potential in a mouse model of premature senescence through transgenic overexpression of the cell cycle inhibitor p16^INK4a. As expected, the p16^INK4a transgenic mouse exhibited increased expression of p16^INK4a mRNA by real-time PCR analysis in all tissues tested. Mice overexpressing p16^INK4a exhibit normal basal hemostatic parameters as tested by CBC, PT, and an in vivo hemostasis model. This suggests that in the absence of vascular injury, overexpression of p16^INK4a has no overt hemostatic consequences. However, upon challenge in various vascular injury models, the p16^INK4a transgenic mice displayed an obvious prothrombotic response.

We have demonstrated a prothrombotic phenotype using two different vascular injury models. Exposure of the vessel to FeCl₃ is a type of biochemical injury that results in endothelial denudation and exposure of the subendothelium following lipid peroxidation³⁷. This type of oxidative damage produces thrombi that are rich in platelets, but also contain red blood cells both encased in a dense fibrin meshwork indicating a role for soluble plasma factors driving thrombus formation^38–40. Rose Bengal is a fluorescein-based chemical that is excited to produce reactive oxygen species when exposed to green light at 540 nm. This results in endothelial activation, although there is very little denudation, and is accompanied by rapid platelet adhesion. Thrombi in this model are primarily composed of platelets and contain less fibrin implying this process is mostly platelet driven³⁸. The observation that p16^INK4a transgenic mice exhibit faster occlusion times in both of these models suggests there is likely a contribution by both soluble plasma factors and circulating cells. The differences observed in the time to flow restriction may reflect altered rates of thrombus growth between wild-type and p16^INK4a transgenic mice, which could be indicative of the potential to produce larger thrombi.

In addition to more rapid rates of venous occlusion, p16^INK4a transgenic mice also display impaired thrombus resolution. The percent occlusion appears to be correlated with the sustaining of inflammatory infiltration. It is possible that the inability to clear residual FeCl₃ from the intravascular space could be involved in further promoting thrombus formation. The increased production of PAI-1 observed in p16^INK4a transgenic mice could also partly explain the thrombus resolution defect (Supplemental Figure 2). Evidence in the literature suggests increased circulating PAI-1 could have a negative impact on wound healing and fibrinolysis. Originally, Farrehi et al demonstrated enhanced fibrinolysis in PAI-1 deficient mice⁴¹. Eitzman et al found that transgenic mice overexpressing PAI-1 have more severe fibrosis following bleomycin-induced lung injury⁴². Zaman et al showed a profibrotic effect of PAI-1 overexpression in the heart following myocardial infarction⁴³. Recently, McDonald et al demonstrated that aged mice display impaired thrombus resolution following stasis induced by inferior vena cava (IVC) ligation⁴⁴. In addition, they reported differences in various plasma and venous endothelium-associated proteins between aged and young wild-type mice⁴⁴. While an exact mechanism to account for the observed thrombus resolution defect in the aged mice is not yet known⁴⁴, it is possible that changes in both the vessel wall and soluble plasma factors contribute, which may also be true of p16^INK4a overexpressing mice.

To better understand differences in thrombus formation between p16^INK4a transgenic and wild-type mice, coagulation parameters in mouse plasma samples were analyzed by CAT after inducing endothelial dysfunction with LPS. Plasma analysis by CAT is sensitive to changes in coagulation factor levels⁴⁵ and able to detect differences in thrombin generation parameters following a thrombotic event in human patients⁴⁶. Our results show that p16^INK4a transgenic mice are able to initiate thrombin generation faster, achieve a higher peak amount of thrombin, and peak at a faster rate than wild-type controls. Therefore, p16^INK4a transgenic mice exhibit greater thrombin generation after LPS challenge compared to wild-type controls. To complement these data, p16^INK4a transgenic mice also showed elevated plasma levels of TAT and PAI-1 following LPS challenge. These markers are commonly used to measure activation of coagulation (TAT)^47,48 and endothelial activation (PAI-1)^49,50. Yamamoto et al showed that aged mice had elevated induction of PAI-1 compared to young mice after LPS treatment, suggesting PAI-1 is important in endotoxin-induced thrombosis³¹. Since PAI-1 is both a marker of endothelial cell senescence and a potent fibrinolysis inhibitor,⁵¹ it could also participate in the delayed thrombus resolution seen in p16^INK4a transgenic mice.

To begin establishing a mechanism for the observed differences between wild-type and p16^INK4a transgenic mice, bone marrow transplants were performed to determine the contribution of hematopoietic cells to the prothrombotic phenotype. We found that wild-type mice given p16^INK4a transgenic bone marrow had occlusion times very similar to that of transgenic controls. Similarly, p16^INK4a transgenic mice given wild-type bone marrow had occlusion times very similar to wild-type controls. These data show that the prothrombotic phenotype observed in mice overexpressing this gene is attributed to p16^INK4a expression in hematopoietic cells.

A growing body of in vitro evidence suggests senescence in the vascular endothelium may also participate in the transition to a procoagulant state during aging. Senescence in the vascular endothelium is associated with an array of phenotypic changes with pathological consequences^7,8. These changes include upregulation of PAI-1, inflammatory cytokines (including interleukin-1α and interleukin-6), matrix metalloproteinases, and the down-regulation of endothelial nitric oxide synthase^35,36,52. Thus, a role for the endothelium cannot be discounted and may warrant further investigation in this model.

In contradistinction to the present data supporting a role for p16^INK4a in venous thrombosis, a differing role for the expression of p16^INK4a in arterial vascular diseases has been suggested. Through genome-wide association studies, several groups have found a link between single nucleotide polymorphisms (SNPs) on chromosome 9p21.3 close to the p16^INK4a open-reading frame and several atherosclerotic diseases (coronary artery disease, ischemic stroke, abdominal aortic aneurysm)⁵³. Liu et al. have recently shown that individuals harboring the SNP genotypes associated with increased atherosclerotic risk exhibit decreased expression of p16^INK4a and other INK4/ARF transcripts⁵⁴. Individuals at increased risk appear to differ in the expression and splicing of linear and circular forms of ANRIL, a long, non-coding RNA emanating from the INK4a/ARF locus thought to participate in INK4a/ARF expression⁵⁵. This observation suggests that decreased production of p16^INK4a is associated with an increased risk of atherosclerosis, likely through limiting aberrant or excess proliferation of cellular components of atheromatous plaques. This suggests expression of anti-proliferative molecules at the INK4a/ARF locus protects individuals from atherosclerosis^53,56. In accord with this view, mice lacking p16^INK4a have been shown to be more prone to vessel occlusion in a carotid artery injury model⁵⁷. Our current data, combined with prior work in the venous system^35,36, suggest the intriguing possibility that age-induced p16^INK4a expression and cellular senescence might play opposing roles with regard to thrombosis and atherosclerosis in the venous and arterial systems, respectively.

Characterizing the link between age-related genetic changes and age-related cardiovascular diseases such as venous thrombosis is of paramount importance. Overexpression of proteins such as p16^INK4a, which promote senescence and vascular dysfunction, could be the key age-related genetic change explaining cardiovascular maladies. Together, our results demonstrate that p16^INK4a overexpression and cellular senescence contribute to a prothrombotic phenotype and defective thrombus resolution. The results of this study provide the foundation for research on the effects of vascular senescence on venous thrombosis.