Cytoprotective 3K3A-activated protein C and plasma: A comparison of therapeutics for the endotheliopathy of trauma

Abstract

BACKGROUND:

Both healthy plasma and cytoprotective aPC (3K3A-aPC) have been shown to mitigate the endotheliopathy of trauma (EoT), but optimal therapeutics remain unknown. Our aim was therefore to determine optimal therapies to mitigate EoT by investigating the effectiveness of 3K3A-aPC with and without plasma-based resuscitation strategies.

METHODS:

Electric cell-substrate impedance sensing (ECIS) was used to measure real-time permeability changes in endothelial cells. Cells were treated with a 2 μg/mL solution of aPC 30 minutes prior to stimulation with plasma taken from severely injured trauma patients (ISS>15 and BD <−6) (TP). Healthy plasma, or plasma frozen within 24 hours (FP24), was added concomitantly with TP. Cells treated with thrombin and untreated cells were included in this study as control groups.

RESULTS:

A dose-dependent difference was found between the 5% and 10% plasma-treated groups when HUVECs were simultaneously stimulated with TP (μd 7.346 95%CI 4.574 to 10.12). There was no difference when compared to TP alone in the 5% (μd 5.713 95%CI −1.751 to 13.18) or 10% group (μd −1.633 95%CI −9.097 to 5.832). When 3K3A-aPC was added to plasma and TP, the 5% group showed improvement in permeability compared to TP alone (μd 10.11 95%CI 2.642 to 17.57), but there was no difference in the 10% group (μd −1.394 95%CI −8.859 to 6.070). The combination of 3K3A-aPC, plasma, and TP at both the 5% plasma (μd −28.52 95%CI-34.72 to −22.32) and 10% plasma concentrations (μd −40.02 95%CI −46.22 to −33.82) had higher inter-cellular permeability than the 3K3A-aPC pre-incubation group.

CONCLUSION:

Our data shows that FP24, in a post-trauma environment, pre-treatment with 3K3A-aPC can potentially mitigate the EoT to a greater degree than FP24 with or without 3K3A-aPC. Although further exploration is needed, this represents a potentially ideal and perhaps superior therapeutic treatment for the dysregulated thromboinflammation of injured patients.

LEVEL OF EVIDENCE:

Prognostic/Epidemiological, Therapeutic/Care Management, Level III.

Noncompressible hemorrhage is the most common cause of preventable death following trauma, and largely attributable to trauma induced coagulopathy (TIC)^1–16. Although stopping the bleeding is crucial and garners much of our initial attention, many patients do not exsanguinate and fortunately survive past their initial injury and resulting hemorrhagic shock manifest as a systemic thromboinflammatory response that disrupts the endothelium. We encounter this thromboinflammation or endotheliopathy of trauma (EoT) clinically as the vascular leak-induced tissue edema that contributes to the end organ dysfunction and late morbidity and mortality in the injured patient^17–20. Therefore, restoring this perturbed inflammation and endothelial barrier integrity is integral to further reducing trauma-related morbidity and mortality.

There have been numerous strategies to mitigate this problem. Previously, plasma has been shown to have cytoprotective effects on endothelial cells treated with thrombin or vascular endothelial growth factor (VEGF)^21–23. However, these models are somewhat artificial in that they use specific pro-inflammatory ligands which may not represent the post-trauma milieu. Because of this, the therapeutic effect of plasma is unknown when primary endothelial cells are stimulated with ex vivo plasma from highly injured patients in shock (TP). On the other hand, cytoprotective aPC (3K3A-aPC) has also been shown to mitigate the EoT, particularly when treated with ex vivo plasma from TP^{24, 25}. It’s imperative to note that this is not recombinant aPC²⁶, but a new bioengineered form of the protein with reduced anticoagulant activity but retains normal cytoprotective activity²⁷. Furthermore, although both plasma and 3K3A-aPC have been shown to mitigate the EoT, optimal therapeutics are currently unknown.

Mechanistically, repair of the vascular endothelium via plasma resuscitation has been linked to both restoration of the endothelial glycocalyx in addition to mitigating EoT via direct signaling on the endothelium^{21–23, 28}. Conversely, the mechanisms behind cytoprotective aPC’s therapeutic effect include decreased thrombin-mediated protein hyperpermeability and restoration of cellular RhoA GTPase activity via protease-activated receptor-1 (PAR-1) signaling on endothelial cells^{25, 29}. However, the spectrum of direct mechanisms behind the beneficial effects of plasma resuscitation and 3K3A-aPC’s overall endothelial cytoprotectivity in trauma remain unknown. Our ultimate aim was to investigate the optimal therapies to mitigate EoT by testing the effectiveness of plasma-based resuscitation strategies in endothelial cells treated with trauma plasma with and without 3K3A-aPC. Since the effects of plasma and 3K3A-aPC on the vascular endothelium is mechanistically different, we therefore hypothesized that there will be distinct effects of plasma and cytoprotective 3K3A-aPC in the setting of the post-trauma milieu.

MATERIALS AND METHODS

This dataset is comprised of clinical patient data and patient samples obtained as part of a prospectively maintained, observational database of adults meeting the highest-level trauma activation criteria at an urban level 1 trauma center. Institutional Review Board (IRB) approval of our research protocol was obtained under exemption from informed consent requirements for emergency research (21 CFR 50.24). The STrengthening the Reporting of OBservational Studies in Epidemiology (STROBE) checklist was used to ensure abidance to the Enhancing the QUAlity and Transparency Of Health Research (EQUATOR) guidelines for reporting observational studies (SDC 1: STROBE Checklist).

Patient Selection and Sample Collection

For injured plasma (TP), selected patients were 18 years of age or older who met criteria for the highest level of trauma activation at an urban, level 1 trauma center. Plasma samples were collected upon arrival to the emergency department (ED), prior to the infusion of any blood products. Further inclusion criteria for this study included either injury severity score (ISS) > 15 with concurrent base excess (BE) < – 6 (Supplemental Table S1). For healthy donor plasma, we used frozen plasma, FP24, which is frozen within 24 hours of collection from a pool of healthy plasma donors.

Assessment of Permeability

Electric cell-substrate impedance sensing (ECIS^®) measures permeability changes between cells in real time in primary human umbilical vein endothelial cells (HUVECs) by measuring resistance between cells, and therefore permeability as the inverse of resistance, via gold electrodes located at the bottom of each well. HUVECs were cultured and grown to confluence prior to transfer to a standard, 96-well, 10idf plate. The gold electrodes ran an alternating current at 4,000 Hz, where the impedance of cell membranes is high, allowing most of the current to flow under the cells and through the tight spaces between the cells. Growth curves, based on the increased resistance in each ECIS plate well, were then obtained prior to commencement of any experiment.

Samples of 10% TP, as previously described²⁵, and either 5% or 10% plasma were applied to each well, confluent with HUVECs, containing the corresponding volume un-supplemented, VascuLife ^® media to a total volume of 200 μL solution in each well, with or without 30-minute pre-incubation with cytoprotective 3K3A-aPC in a final concentration of 2 μg/mL per well. For TP, results of preliminary ECIS data were used to categorize samples by phenotype based on the degree of permeability induced on cells and then pooled to decrease the variability of the response induced on HUVECs. For this stage, pooled plasma samples originating in patients with an ISS >15 and BE < –6 were used. Resistance, and therefore permeability changes between cells, were then recorded for 45 minutes. Recombinant thrombin was used as a positive control. A group of untreated cells served as a negative control. We quantified permeability changes in the control and treatment groups by calculating the mean difference (μd) of the area under their curves (AUC). Statistical significance was determined by one-way analysis of variance (ANOVA) followed by a post hoc analysis using Tukey’s multiple comparison’s test.

RESULTS

A total of 9 healthy plasma donor samples were collected and used for this study. The donors were 57.33 ± 20.65 years of age, two-thirds were male (66%), and were well representative of all blood types (Table 1). Recall that permeability changes are quantified based on the area under the curves of normalized resistance and therefore increases in resistance translates to decreases in permeability and vice versa. As previously shown, TP induced statistically significant increases in inter-cellular permeability (μd −36.03 95%CI −46.22 to −25.84) that were mitigated by pre-incubation with 3K3A-aPC (μd 38.62%CI 29.32 to 47.92). The positive control group, treated with thrombin, also induced a statistically significant increase in the inter-cellular permeability of HUVECs (μd −54.75 95%CI −64.94 to −44.56) when compared to the untreated control group (Figure 1). All comparisons of resistance, which are inversely related to inter-cellular permeability, are reported in Table 2.

Table 1.

FP24 Donor Characteristics

FP24 Donor Characteristics
Age (m, SD)	57.33 (20.65)
Gender (Female)	33% (N=3)
Race (White)	89% (N=8)
Blood Type
B+	N=3 (33%)
AB+	N=2 (22%)
A+	N=2 (22%)
O+	N=2 (22%)

Figure 1.

Tracings of untreated control group (control), positive control groups thrombin and TP, and 3K3A-aPC pre-incubation prior to stimulation with TP with the corresponding AUC.

Table 2.

Bivariable Comparisons of All Groups by Resistance Quantified by Area Under the Curves

Group Comparisons	Mean Diff.	95% CI of diff.
5% FP24 vs. 10% FP24	−1.255	−4.028 to 1.517
5% FP24 vs. 5% FP24 +TP	11.08	8.304 to 13.85
5% FP24 vs. 10% FP24 +TP	18.42	15.65 to 21.19
5% FP24 vs. 3K3A-aPC + 5% FP24 + TP	6.683	3.910 to 9.455
5% FP24 vs. 3K3A-aPC + 10% FP24 + TP	18.18	15.41 to 20.96
5% FP24 vs. 3K3A-aPC	−21.83	−28.03 to −15.63
5% FP24 vs. TP	16.79	9.324 to 24.25
5% FP24 vs. Thrombin	35.51	28.04 to 42.97
5% FP24 vs. CTRL	−19.24	−26.71 to −11.78
10% FP24 vs. 5% FP24 +TP	12.33	9.559 to 15.10
10% FP24 vs. 10% FP24 +TP	19.68	16.90 to 22.45
10% FP24 vs. 3K3A-aPC + 5% FP24 + TP	7.938	5.165 to 10.71
10% FP24 vs. 3K3A-aPC + 10% FP24 + TP	19.44	16.67 to 22.21
10% FP24 vs. 3K3A-aPC	−20.58	−26.78 to −14.38
10% FP24 vs. TP	18.04	10.58 to 25.51
10% FP24 vs. Thrombin	36.76	29.30 to 44.23
10% FP24 vs. CTRL	−17.99	−25.45 to −10.52
5% FP24 +TP vs. 10% FP24 +TP	7.346	4.574 to 10.12
5% FP24 +TP vs. 3K3A-aPC + 5% FP24 + TP	−4.393	−7.166 to −1.621
5% FP24 +TP vs. 3K3A-aPC + 10% FP24 + TP	7.108	4.335 to 9.880
5% FP24 +TP vs. 3K3A-aPC	−32.91	−39.11 to −26.71
5% FP24 +TP vs. TP	5.713	−1.751 to 13.18
5% FP24 +TP vs. Thrombin	24.43	16.97 to 31.90
5% FP24 +TP vs. CTRL	−30.32	−37.78 to −22.85
10% FP24 +TP vs. 3K3A-aPC + 5% FP24 + TP	−11.74	−14.51 to −8.967
10% FP24 +TP vs. 3K3A-aPC + 10% FP24 + TP	−0.2381	−3.011 to 2.534
10% FP24 +TP vs. 3K3A-aPC	−40.26	−46.46 to −34.06
10% FP24 +TP vs. TP	−1.633	−9.097 to 5.832
10% FP24 +TP vs. Thrombin	17.09	9.623 to 24.55
10% FP24 +TP vs. CTRL	−37.66	−45.13 to −30.20
3K3A-aPC + 5% FP24 + TP vs. 3K3A-aPC + 10% FP24 + TP	11.5	8.729 to 14.27
3K3A-aPC + 5% FP24 + TP vs. 3K3A-aPC	−28.52	−34.72 to −22.32
3K3A-aPC + 5% FP24 + TP vs. TP	10.11	2.642 to 17.57
3K3A-aPC + 5% FP24 + TP vs. Thrombin	28.83	21.36 to 36.29
3K3A-aPC + 5% FP24 + TP vs. CTRL	−25.92	−33.39 to −18.46
3K3A-aPC + 10% FP24 + TP vs. 3K3A-aPC	−40.02	−46.22 to −33.82
3K3A-aPC + 10% FP24 + TP vs. TP	−1.394	−8.859 to 6.070
3K3A-aPC + 10% FP24 + TP vs. Thrombin	17.33	9.861 to 24.79
3K3A-aPC + 10% FP24 + TP vs. CTRL	−37.42	−44.89 to −29.96
3K3A-aPC vs. TP	38.62	29.32 to 47.92
3K3A-aPC vs. Thrombin	57.34	48.04 to 66.64
3K3A-aPC vs. CTRL	2.593	−6.705 to 11.89
TP vs. Thrombin	18.72	8.534 to 28.91
TP vs. CTRL	−36.03	−46.22 to −25.84
Thrombin vs. CTRL	−54.75	−64.94 to −44.56

As an initial baseline, primary endothelial cells were treated with healthy plasma alone as shown in Figure 2. The 5% plasma treated HUVECs (μd −19.24 95%CI −26.71 to −11.78) and 10% plasma-treated groups (μd −17.99 95%CI −25.45 to −10.52) induced increases in permeability within 45 minutes when compared to untreated controls. Both concentrations of healthy plasma, however, induced similar permeability changes (μd −1.255 95%CI −4.027 to 1.517) when compared to one another. These permeability changes were lesser in degree than TP-induced changes (μd 16.79 95%CI 9.324 to 24.25), but greater than the permeability change caused by TP with 30-minute pre-incubation with 3K3A-aPC (μd −21.83 95%CI −28.03 to −15.63) for the 5% plasma. Similarly, the inter-cellular permeability induced by the 10% healthy plasma alone was smaller than the one induced by TP (μd 18.04 95%CI 10.58 to 25.51), but greater than the permeability change induced by TP with 30-minute pre-incubation with 3K3A-aPC (μd −20.58 95%CI −26.78 to −14.38).

Figure 2.

Tracings of positive control group TP and 3K3A-aPC pre-incubation prior to stimulation with TP with tracings of 5% and 10% plasma alone with the corresponding AUC.

When HUVECs were simultaneously stimulated with plasma and TP, there was now a difference between the 5% and 10% plasma-treated groups in which the lower concentration of healthy plasma had a better permeability profile when compared to the higher concentration (μd 7.346 95%CI 4.574 to 10.12) (Figure 3). Notably, when comparing simultaneous treatment of TP with plasma, not only is there still a statistically significant increase in permeability in the 5% plasma+TP (μd −30.32 95%CI −37.78 to −22.85) and 10% plasma+TP groups (μd −37.66 95%CI −45.13 to −30.20) after 45 minutes when compared to untreated cells, but the increase in permeability is not statistically different from that induced by TP alone when compared to the 5% plasma (μd 5.713 95%CI −1.751 to 13.18) or the 10% plasma group (μd −1.633 95%CI −9.097 to 5.832).

Figure 3.

Tracings of positive control group TP and 3K3A-aPC pre-incubation prior to stimulation with TP with tracings of 5% and 10% plasma stimulation simultaneously with TP with the corresponding AUC.

However, when 3K3A-aPC is added to the simultaneous addition of plasma and TP, the 5% plasma combination with aPC pre-incubation showed a statistically significant improvement in the inter-cellular permeability of HUVECs when compared to TP alone (μd 10.11 95%CI 2.642 to 17.57), whereas the 10% plasma group continued to have no difference when compared to the TP alone group (μd −1.394 95%CI −8.859 to 6.070) as seen in Figure 4. The 5% plasma-3K3A-aPC combination also had a significantly better permeability profile than the 10% plasma-3K3A-aPC combination (μd 11.5 95%CI 8.729 to 14.27). Nevertheless, the 5% plasma-3K3A-aPC combination still induced a statistically significant increase in inter-cellular permeability in 45 minutes when compared to the untreated control group (μd −25.92 95%CI −33.39 to −18.46). Also, the combination of 3K3A-aPC, plasma, and TP at both the 5% plasma (μd −28.52 95%CI −34.72 to −22.32) and 10% plasma concentrations (μd −40.02 95%CI −46.22 to −33.82) had higher inter-cellular permeability than the 3K3A-aPC pre-incubation group, which was also treated with TP.

Figure 4.

Tracings of positive control group TP and 3K3A-aPC pre-incubation prior to stimulation with TP with TP with tracings of 5% and 10% plasma with APC combinations simultaneously with TP stimulation with the corresponding AUC.

DISCUSSION

While plasma-based resuscitation has become the standard of care for trauma patients, there remains considerable controversy about the biologic mechanisms that drive its clinical benefit. Indeed, there remains a paucity of data to demonstrate how plasma modulates the thromboinflammatory milieu after injury. While of potential clinical benefit, plasma or whole blood-based therapies are logistically problematic especially in resource constrained or austere settings. To combat this, many groups, including ours, continue to develop cytoprotective inflammomodulatory therapeutics which could represent a shelf-stable medication or resuscitation in a syringe for the inured patient. The presented data suggest that 3K3A-aPC provides better restoration and protection of endothelial barrier function than plasma in an in vitro trauma model. In fact, our data shows that plasma, while often thought to be protective, in fact causes an increase in permeability. There were no differences in the permeability profile of plasma at different doses alone on HUVECs. Also, plasma did not appear to mitigate the endothelial permeability induced by TP, which contrasts with the barrier protection that 3K3A-aPC appears to provide. Only when cytoprotective 3K3A-aPC is added to plasma was there a dose-dependent protection of barrier function manifest by an elimination of TP-induced permeability. Taken together, plasma is deleterious alone and non-protective in a trauma plasma-induced endotheliopathy model.

Though the presented data may differ from prior data in the existing literature which relies on pro inflammatory ligands to induce permeability (e.g. VEGF, thrombin), a different methodology was employed in our experiments through the incubation of HUVECs with TP which approximates the post-trauma vascular microenvironment of injured patients in shock^21–23. We achieved this by using a pool of 10% trauma plasma, by total volume, collected from highly injured patients (ISS >15) in shock (BE < −6). This pool was phenotyped via ECIS based on their permeability profiles prior to experimental stimulation of primary HUVECs in each corresponding group as described^{21, 25}. Not surprisingly, there was a dose-dependent response when trauma plasma is added to different concentrations of healthy resuscitative plasma, where higher doses produced higher permeability. Similarly, there was a dose-dependent response with 3K3A-aPC pre-incubation prior to stimulation with trauma plasma and different concentrations of healthy resuscitative plasma.

Mechanistically, it is possible that plasma may actually blunt the cytoprotective milieu after trauma. Our data shows that the addition of plasma caused permeability by itself and separately attenuated the protective effects of 3K3A-aPC. This could be possibly happening via pre-circulating proinflammatory cytokines or components that act as damage-associated molecular patterns (DAMPs) in donors or build-up of thrombin during storage³⁰. Healthy resuscitative plasma, both in this study and in the real world, is drawn from a variety of individuals who appear healthy but may have different sets of co-morbidities and underlying inflammation, like smoking, which may affect its potential therapeutic utility. For example, circulating complement proteins in the post-injury plasma microenvironment are associated with adverse outcomes, have been shown to be increased by plasma resuscitation³¹. Although further exploration is needed, 3K3A-aPC represents a potentially ideal and perhaps superior therapeutic treatment for dysregulated thromboinflammation for injured patients.

This study has several limitations. Like similar studies, ex vivo trauma plasma samples were obtained on arrival to the emergency department, and therefore excludes the remainder of the inevitably variable patient’s course. There is also a degree of selection bias that was introduced when TP samples were phenotyped by their degree of permeability. However, pooling TP helped reduce the variability within our experimental groups. With respect to 3K3A-aPC therapy, the results of our study are limited to pre-incubation therapy only due to inherent limitations of ECIS methodology. It is important to consider that when cells are pre-incubated with 3K3A-aPC, the effect of trauma plasma is highly mitigated or negated. However, when healthy resuscitative plasma is added to HUVECs in a post-trauma environment, their inter-cellular electrical resistance does not recover. We believe expanding the results of our study to flow conditions and/or animal models will further advance this topic by more accurately simulating the vascular endothelial microenvironment.

In summary, although early hemostasis and optimization of blood volume is imperative, plasma resuscitation alone is unlikely the ideal therapeutic to combat post-injury thromboinflammation which contributes to late morbidity and mortality in these patients. This is especially true in the era of personalized resuscitation where a one size fits all approach is no longer sufficient. 3K3A-aPC, which continues to represent a potentially ideal therapeutic for thromboinflammation in trauma patients who survive the initial hemorrhagic shock at the level of their vascular endothelium. 3K3A-aPC may also have attenuated effects in the setting of treatment with plasma. Further exploration is needed for this potentially ideal and perhaps superior therapeutic treatment for dysregulated thromboinflammation for injured patients.