Fig 1.
Construction and verification of a glycoprotein C (gC) deletion mutant of Ab4.
A: An illustrative overview of generation of the deletion mutant in an infectious artificial bacterial chromosome (BAC) containing plaque purified Ab4 (Ab4). A PCR product that harbored the kanR gene instead of nucleotides 98–1315 of the ORF16 gene, which encodes gC, was electroporated into E. coli GS1783 for en passant (two-step) Red recombination. Intermediate colonies were screened and used for the second recombination step. Removal of KanR was induced by 1% arabinose resulting in the final construct. B: Identity of the mutant (Ab4GΔgC) was confirmed by restriction enzyme digests. BAC DNA from the parental, intermediate and final constructs was digested with HindIII and separated by 0.8% agarose gel electrophoresis. The intermediate construct harboring KanR had an additional band of 6.8 kbp (lower arrow). After resolution, a fragment of 8.6 kbp (upper arrow) containing the ORF16 gene disappeared and a fragment of 7.4 kbp in the final mutant appeared. This apparent reduction is in line with the deletion introduced by mutagenesis. C: BAC DNA from Ab4 (wild-type) or Ab4GΔgC was used as the template for PCR using gC primers that span the deleted region (forward: 5’- CTCCGACCAGTGGAGTTATTAT-3’; reverse: 5’- CTACTGTTTTTACCAGCGCTTC-3’). Bands of the predicted size were amplified in either case, thus confirming deletion of gC-encoding sequences in the mutant Ab4GΔgC.
Table 1.
Synthetic oligonucleotides for engineering and sequencing of the gC deletion mutant of Ab4.
Fig 2.
Flow cytometry gating strategy for quantification of platelet-derived microparticles in equine platelet samples.
A: Platelet events in citrate-anticoagulated platelet-rich plasma were identified and gated as CD41-positive cells (R1 region) in a CD41 fluorescence versus forward scatter (FSC) dotplot. The R1 region or gate was established on an isotype control for the CD41 antibody. Representative image from platelets exposed to the RacL11 strain of EHV-1 at 1 plaque forming unit (PFU)/cell. B: Platelet-derived microparticles (PDMPs) were defined as small events (<101 log FSC units) positive for Annexin V and CD41. The PDMP percentage was obtained from the lower right quadrant of an Annexin V fluorescence versus FSC dotplot of the R1 gate (CD41-positive events), with the quadrants being defined on a negative sample in which 1 mM EDTA was added to the buffer with Annexin V. The PDMP percentage was 0.1% in this representative image of platelets exposed to rabbit kidney 13 (RK) cell lysate at an equivalent volume to 1 PFU/cell (mock-infected negative control). The events in the upper left and right quadrants are platelets that are negative (94.0%) and positive for Annexin V (0.9%), respectively. C: Representative image of PDMP quantification in platelets exposed to RacL11 at 1 PFU/cell. In this sample, there are 12.1% PDMP (lower right quadrant) and 22.1% of platelets are weakly positive for Annexin V (upper right quadrant).
Fig 3.
EHV-1 induces platelet P-selectin expression and shedding of platelet-derived microparticles in equine platelet-rich plasma.
Platelets were exposed for 10 minutes to RacL11 or Ab4 EHV-1 strains at increasing PFU/cell (0.01, 0.1, 0.5, or 1) with PBS and RK lysate as negative controls and thrombin-convulxin (TC, 0.15 U/mL-0.05 ug/mL) as a positive control. Then the mean percentage ± SD of platelets positive for P-selectin (A) and PDMPs (B) were quantified (n = 8). At the higher PFU/cell of 0.5 and 1, both strains induced P-selectin expression and microvesiculation, along with compaction and narrowing of the platelet event cloud and increased small events on forward and side scatter dotplots (C). Microvesiculation was more prominent with RacL11 (B, C arrow). Exposure to both viruses at 5 PFU/cell replicated these findings, demonstrating a dose-dependent platelet activation response (S1 Fig). * p ≤ 0.01 versus PBS or RK negative controls for each strain. ** p < 0.001 versus TC or Ab4 at 0.5 or 1 PFU/cell.
Fig 4.
EHV-1-induced platelet P-selectin expression requires plasma and exogenous calcium.
A: Effect of plasma: Equine platelets in ACD-anticoagulated platelet-rich plasma (PRP), washed platelets (WP) or washed platelets with added ACD-anticoagulated platelet-derived microparticle-depleted plasma (WP+MDP) were exposed for 10 minutes to RacL11 at 1 PFU/cell with PBS and RK lysate negative controls and a thrombin (T, 0.15 U/mL) positive control. Washed platelets did not express P-selectin when exposed to virus unless plasma was present. In contrast, thrombin-induced P-selectin expression was independent of plasma (n = 5). Data represents mean ± SD. * p < 0.001 versus PRP or WP + MDP for RacL11-exposed platelets. B: Effect of calcium: Equine citrate-anticoagulated PRP was exposed to RacL11 at 1 PFU/cell for 10 minutes with increasing calcium concentrations (0 to 2.5 mM), with the above controls. EHV-1-induced P-selectin expression required at least 1 mM of exogenous calcium (n = 3). Data represents mean ± SD. * p < 0.001 versus 1.0, 2.0 or 2.5 mM calcium.
Fig 5.
The RacL11 strain of EHV-1 induces P-selectin exteriorization through the extrinsic pathway of coagulation.
The mean ± SD percentages of P-selectin-positive platelets and platelet-derived microparticles (PDMPs) were quantified in equine platelets in citrate-anticoagulated PRP with or without hirudin (10 units), after exposure to RacL11 at 1 PFU/cell or rabbit kidney (RK) cell lysate negative and thrombin (T, 1 U/mL) positive controls. Hirudin significantly reduced P-selectin expression (A, n = 4) and microvesiculation (B, n = 6) in response to RacL11 or thrombin. * p < 0.05 versus untreated PRP. In RacL11-exposed washed ACD-anticoagulated platelets, P-selectin expression did not occur in the absence of plasma, but was still present in the thrombin-convulxin (TC, 0.15 U/mL-0.05 ug/mL) control. Addition of citrate-anticoagulated equine (E) or human (H) PDMP-depleted plasma (MDP) containing all coagulation factors re-established P-selectin expression in washed platelets (C, n = 3). In contrast, microvesiculation was still induced by RacL11 in washed platelets without MDP, but addition of equine or human MDP boosted the percentage of PDMPs, with human MDP having a weaker response in virus-exposed platelets. (D, n = 6). * p <0.05 versus washed platelets with no added plasma. ** p < 0.05 versus equine plasma. Addition of FVII- or FX-deficient human MDP, instead of human MDP replete in all coagulation factors (Full), did not result in P-selectin expression in RacL11-exposed washed platelets. However, supplementation of FVII-deficient MDP with purified human FVIIa (1 nM, FVII- + FVIIa) re-established P-selectin expression induced by RacL11, indicating FX generation was secondary to extrinsic pathway activation (E, n = 3 to 7).Microvesiculation was also significantly reduced in FVII- and FX-deficient MDP, with supplemental FVIIa boosting the response in FVII-deficient MDP (F, n = 4). * p <0.05 versus Full MDP. In contrast, addition of human FIX-, FXI- or FXII-deficient MDP to washed platelets did not significantly affect P selectin expression (G) or PDMP release (H) (n = 3 to 7). Similar results were seen with Ab4 (S2 Fig).
Table 2.
Equid herpesvirus type 1 generates factor Xa (FXa) in a dose-dependent manner.
Fig 6.
Tissue factor mediates EHV-1-induced platelet activation, with no role for virus glycoprotein C.
A goat polyclonal anti-rabbit tissue factor (TF) antibody (anti-TF, 152 ug/mL) abolished P-selectin expression (A) and significantly decreased the numbers of platelet-derived microparticles (PDMP) (B) in equine platelets exposed to RacL11 and Ab4 EHV-1 strains at 1 PFU/cell (n = 6). Data shown is mean ± SD. * p = 0.033 for RacL11 and p = 0.011 for Ab4 versus goat IgG controls. In contrast, platelets still expressed P-selectin (C) and shed PDMPs (D) when exposed to an Ab4-based envelope glycoprotein C deletion mutant (ΔgC) at 1 PFU/cell. The deletion mutant induced stronger vesiculation and lower P-selection translocation than wild type Ab4 (WT) (n = 8). Data shown is mean ± SD. * versus ΔgC mutant.
Fig 7.
EHV-1 generates thrombin in equine plasma.
Thrombin generation by purified virus (at amounts equivalent to 1 PFU/cell) was measured using calibrated automated thrombography in corn trypsin inhibitor (CTI)-containing citrate-anticoagulated platelet-derived microparticle-depleted equine plasma (CTI-MDP) with or without autologous platelets (1 x 105/reaction). Data shown is lag time (A), total thrombin generated (nM*minute, derived from the area under the thrombogram curve) (B) and compilation thrombogram curves (C, mean of 3 separate experiments). Addition of platelets shortened the lag time and boosted the total thrombin generated by both virus strains. Ab4, but not RacL11, generated small amounts of thrombin in CTI plasma alone. No thrombin was generated (lag time >30 minutes, total thrombin generated 0 nM*minute) in negative controls (CTI-MDP with or without autologous platelets and CTI-MDP with RK lysate with or without platelets).
Fig 8.
Model for EHV-1-induced activation of platelets.
In our model of EHV-1-induced platelet activation, we propose that the virus first binds to platelets, which triggers platelet fragmentation or microvesiculation in a TF- and thrombin-independent manner. Concurrently, factor VII (FVII) forms a complex with TF in the viral envelope and activates factor X (FX), with FXa generating thrombin in the vicinity of or on the surface of platelets. Thrombin generation may be propagated by phosphatidylserine (PS) in the viral envelope. Thrombin binds to protease-activated receptors (PAR) on platelet surfaces, inducing release of α-granules with subsequent surface P-selectin and PS expression and additional microvesiculation. If marked, microvesiculation results in loss of membrane-anchored P-selectin on microparticles. Depending on the viral strain or load, thrombin may be generated on the surface of virus not attached to platelets.