INTRODUCTION
In addition to their primary complex role in hemostasis, the anucleate platelets play a crucial role in the regulation of innate and adaptive immunities.1,2 It has also been shown that platelets have a decisive part in processes such as angiogenesis, vasculopathy, and in tumor growth and metastasis.3-5 Especially the interactions among endothelial cells, platelets, and different leukocytes are characteristic for the pathogenesis of these inflammatory diseases.6 Activated platelets secrete proinflammatory mediators such as cytokines, chemokines, and growth factors from their granules and additionally express a series of surface adhesion receptors leading to interactions between platelets and endothelial cells.7,8 Consequently, endothelial cells enhance the expression of adhesion receptors.9 Our recent interest has focused on the involvement of platelets in the pathogenesis of cardiac allograft vasculopathy (CAV). Here, we could previously show that activation of the platelet P2Y12-receptor by ADP plays a key role in facilitation of neointima formation in the experimental murine aortic allograft model. These results demonstrated that the inhibition of platelet aggregation by monotherapy with clopidogrel effectively leads to a significant reduction of transplant arteriosclerosis as experimental analog of CAV and to decreased infiltration of macrophages and dendritic cells in murine aortic allografts.10,11
Small animal models such as immunodeficient mice harboring human cells or tissues (“humanized mice”) are a promising tool for studying complex mechanisms in human biology and may partly overcome species-specific differences.12-17 Lorber et al18 initially developed a humanized mouse model of vascular rejection by orthotopically transplanting human artery segments into immunodeficient mice that were later reconstituted with human peripheral blood lymphocytes and subsequently rejected their xenograft. We have recently shown, using this xenotransplantation model, that C57/Bl6-Rag2−/−γc−/− mice are a more suitable recipient than C.B-17-SCID/beige mice as they exhibit an improved reconstitution rate with higher levels of CD45+ human cells and more reliably reject human arterial xenografts.19,20
However, it is still not clear whether human platelets (hPlts) alone are sufficient to cause intimal proliferation of a transplanted vascularized organ or vessel. Therefore, the aim of this study was to investigate whether human platelets are able to interact with and activate human endothelial cells and to induce the development of transplant arteriosclerosis in a humanized C57/Bl6-Rag2−/−γc−/− mouse arterial xenograft model in the absence of adaptive immunity and natural killer cells (NK cells).
MATERIALS AND METHODS
Animals
Original breeding pairs for C57/Bl6-Rag2/−γc−/− mice were obtained from Taconic (Germantown, NY). This mouse strain is characterized by its deficiency for B and T lymphocytes and NK cells. Mice were aged between 6 and 12 wk at the time of experimental use and were bred and maintained at the central animal facility of the University of Erlangen-Nürnberg (Preclinical Experimental Animal Center) in isolated ventilated cages with sterilized food, water, and bedding. Mice were maintained on trimethoprim–sulfamethoxazole (8/140 mg/kg) in their drinking water. All experiments were conducted in accordance with institutional and state guidelines (Regierung von Mittelfranken, AZ 54-2532.1-31/12).
Abdominal Transplantation of Human Mammary Artery Segments
The procedure of transplanting human vessels in the abdominal aortic position of mice was performed using the technique initially described by Abele-Ohl et al.20Figure 1 graphically summarizes the design of the experimental model.
;)
FIGURE 1.: “Central illustration”: Experimental design of the xenogeneic mouse model. The chart illustrates the process of human vessel-to-mouse xenotransplantation, continuous hPlts injection and the analytic evaluation. hPlt, human platelet; i.v., intravenous; TRAP-6, thrombin receptor activating peptide-6.
Human Platelets
HPlts were isolated and ex vivo activated as described in the Materials and Methods (SDC, https://links.lww.com/TP/C277).
Histological Analysis
Mammary artery grafts were analyzed on day 30 after transplantation for Elastica van Gieson staining and immunofluorescence examinations (Materials and Methods, SDC, https://links.lww.com/TP/C277).
Statistical Analysis
Results are given as the mean per group ± SEM. The data were analyzed using 1-way ANOVA followed by a Bonferroni-Holm correction with the significance set at P values of *≤ 0.05, **≤ 0.01, and ***≤ 0.001.
Ethics Statement
The infrarenal C57/Bl6-Rag2−/ −γc−/− artery was replaced by transplanting size-matched human arterial segments obtained from surgical specimens; informed consent was obtained from all patients and all data were analyzed anonymously (institutional review board approval obtained by the Ethics Committee of the University of Erlangen-Nürnberg).
RESULTS
HPlts Collected by Apheresis Show a Decreased Activation Status but Can Be Activated Effectively In Vitro by Thrombin Receptor Activating Peptide-6
Isolation of pure platelets without affecting the activation status is crucial for platelet research studies. To ensure that the majority of hPlts occur in a nonactivated state, we compared the activation status of hPlts with allophycocyanin-conjugated antibodies against the platelet activation makers CD62p or CD63 from 2 different isolation methods. As a positive control to detect the hPlts population overall, we used fluorescein isothiocyanate–conjugated CD42b, which was highly expressed in nonactivated and activated hPlts. Nonspecific binding was detected by isotype controls (anti–immunoglobulin G). Figure 2A shows the flow cytometric analysis of hPlts isolated by centrifugation using a modified technique described by Cazenave et al21 and Dickfeld et al22 compared with hPlts collected by apheresis at the Institute of Transfusion Medicine Erlangen.23 The hPlt fraction was 98.71% ± 0.67% (hPlts collected by apheresis) versus 92.28% ± 1.18% (hPlts collected by centrifugation and washing, n = 6 per group). HPlts isolated by apheresis showed significantly less expression of both platelet activation markers compared with hPlts collected by centrifugation and washing (17.32% ± 2.19% versus 48.38% ± 3.00% [CD62p] and 3.72% ± 0.84% versus 41.38% ± 3.00% [CD63], P ≤ 0.001, n = 6 per group) (Figure 2A). Therefore, we used the leukocyte-poor hPlt concentrate collected by apheresis for further experiments.
FIGURE 2.: Flow cytometric analysis of hPlt activation status using the established platelet activation markers antihuman CD62p and antihuman CD63. Results are given in percentage of positive measured hPlts. For negative or positive controls, the corresponding IgG-isotype controls, respectively, the hPlt detection marker antihuman CD42b were used (A and B). The effects of different concentrations of the reagents TRAP-6 (10, 20 µM/L) and ADP (5, 20 µM/L) on the activation status of hPlts collected by apheresis (B). Data are represented as mean ± SEM (n = 6 per group; ***P ≤ 0.001, hPlts collected by apheresis vs hPlts collected by centrifugation and washing, in vitro–activated hPlts vs nonactivated hPlts). hPlt, human platelet; IgG, immunoglobulin G; TRAP-6, thrombin receptor activating peptide-6.
Furthermore, we analyzed different hPlt concentrations and the efficacy of the platelet activation reagents thrombin receptor activating peptide-6 (TRAP-6) and ADP on hPlts. Results are shown in Figure 2B. Stimulation with 20 µM/L TRAP-6 caused the most successful activation of the marker of CD62p and CD63 in comparison with nonactivated hPlts (94.40% ± 1.44% versus 15.08% ± 2.47% [CD62p], 85.74% ± 2.72% versus 3.72% ± 0.84% [CD63], ***P ≤ 0.001, n = 6 per group]. TRAP-6 (20 µM/L) was therefore chosen for hPlt activation in the consecutive experiments.
In Vivo Recovery of Nonactivated and In Vitro–activated hPlts in C57/Bl6-Rag2−/−γc−/− Mice
First, we determined the necessary concentration of hPlts for their recovery in the Rag2−/−γc−/−mouse circulation. Therefore, using the percentage of antihuman CD41a-positive flow cytometry triggered events at 5 min after intravenous (IV) injection of nonactivated hPlts as 100% recovery according to Xu et al.24 The circulation of hPlts in mouse whole blood was evaluated for 2 different hPlt concentrations (4 × 108/2 × 108) at 5, 30, 60, and 120 min after injection. Compared to 4 × 108 of nonactivated hPlts, the in vivo recovery of 2 × 108 was significantly reduced from 36.71% ± 3.04% versus 15.03% ± 2.58% (60 min) (P ≤ 0.01), and 25.73% ± 3.81% versus 10.49% ± 3.07% (120 min) (P ≤ 0.05) (Figure 3A). Thus, in the following experiments, we injected 4 × 108 hPlts. In addition, we investigate whether the recovery of hPlts in the circulation was affected by the activation status of hPlts or the site of injection.
FIGURE 3.: Recovery of CD41a+ hPlts in C57/Bl6-Rag2−/−γc−/− mouse whole blood. Mouse whole blood was collected 5, 30, 60, and 120 min after hPlt intravenous injection. The hPlts in a total Plt population (human and mouse) are represented by a CD41a-positive flow cytometry gate. Recovery of nonactivated hPlts 5 min after injection was defined as 100% recovery. A, Flow cytometric analysis and comparison of mouse whole blood after lateral tail vein injection of 4 × 108 and 2 × 108 nonactivated hPlts. B, Flow cytometric analysis and comparison of mouse whole blood after injection of 4 × 108 nonactivated or in vitro–activated hPlts into the lateral tail vein or retrobulbar venous plexus. C, Flow cytometric analysis and comparison of mouse whole blood after tail vein injection of 4 × 108 nonactivated or in vitro–activated hPlts. Data are represented as mean ± SEM (n = 4 per group; ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, 4 × 108 nonactivated hPlts vs 2 × 108 nonactivated hPlts, tail vein injected hPlts vs retrobulbar-injected hPlts, in vitro–activated hPlts vs nonactivated hPlts). hPlt, human platelet.
Thus, we compared the in vivo recoveries of tail vein and retrobulbar-injected nonactivated or in vitro–activated hPlts. Figure 3B shows very similar (P > 0.05) hPlt recoveries for the time points 30, 60, and 120 min after injection.
Only 5 min after retrobulbar injection of activated hPlts significantly less (P ≤ 0.001) of these platelets were detected in mouse peripheral blood (11.60% ± 7.70%) compared with injection via the tail vein (59.47% ± 4.18%). For the following experiments tail vein injection was chosen as the route for hPlt application. In this group, we also found reduced circulating hPlts after injection of 4 × 108 in vitro–activated versus nonactivated hPlts in all point of times (59.47% ± 4.18% versus 100% ± 4.05% [5 min], P ≤ 0.01; 11.09% ± 1.84% versus 65.96% ± 5.17% [30 min], P ≤ 0.001; 4.53% ± 0.32% versus 36.71% ± 3.04% [60 min], P ≤ 0.001; 1.00% ± 0.43% versus 27.73% ± 3.81% [120 min], P ≤ 0.01; n = 4 per groups; Figure 3C).
hPlts Significantly Shifted Their Activation Status When Circulating in C57/Bl6-Rag2−/−γc−/− Mice
To ascertain the activation status of hPlts in the mouse blood circulation, we analyzed the level of CD62p and CD63 after hPlt tail vein injection. Nonactivated hPlts significantly increased their CD62p and CD63 expressions 60 and 120 min after injection compared with 5 min after (for CD62p: 24.35% ± 4.59% [60 min] and 29.05% ± 7.70% [120 min] versus 9.93% ± 1.45% [5 min]; P ≤ 0.05) (for CD63: 14.55% ± 2.38% [60 min], P ≤ 0.01 and 18.23% ± 4.52% [120 min], P ≤ 0.05 versus 3.80% ± 1.60% [5 min]) (Figure 4A and B). They decreased them after 30 and 60 min when using ex vivo–activated hPlts compared with 5 min after injection (for CD62p: 46.50% ± 2.20% [30 min], P ≤ 0.001 and 18.00% ± 3.04% [60 min], P ≤ 0.001 versus 74.65% ± 3.58% [5 min]) (for CD63: 54.50% ± 11.78% [30 min], P = 0.269 and 39.63% ± 7.20% [60 min] versus 69.93% ± 6.46% [5 min], P ≤ 0.05; Figure 4C and D).
FIGURE 4.: Both nonactivated and in vitro–activated hPlts significantly shifted their activation status when circulating in C57/Bl6-Rag2−/−γc−/− mice. At 5, 30, 60, and 120 min after tail vein injection of nonactivated (A and B) or in vitro–activated (C and D) hPlts, mouse whole blood was collected and analyzed by flow cytometry. Expression of the platelet activation markers antihuman CD62p (A and C) and antihuman CD63 (B and D) was measured. Data are represented as mean ± SEM (n = 4 per group; ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, 5 min after injection vs 30, 60, and 120 min after injection). hPlt, human platelet.
Both Dichlorodihydrofluorescein and Qtracker 800 Did Not Influence the Activation Status of hPlts
The reliable visualization of different cell types in vivo is a key factor for successful implementation of experimental studies. Accordingly, the aim in the present analyses was the evaluation of in vivo localization and dissemination of human platelets injected in C57/Bl6-Rag2−/−γc−/− mice lacking T-, B-, and NK cells using an in vivo fluorescence imaging system and a confocal fluorescence laser endomicroscope. The labeling of hPlts with the fluorescence dyes dichlorodihydrofluorescein (DCF) and Qtracker 800 was determined. Initially, the question was if the labeling affects the activation status or the activation capacity of hPlts. Flow cytometric measurements showed slightly but not significantly less of the fluorescent dye Qtracker 800 on nonactivated hPlts (78.40% ± 2.55%) versus ex vivo–activated hPlts (84.90% ± 0.51%) (n = 4 per group) and in contrast similar DCF-labeled hPlts between the 2 platelet groups (99.10% ± 0.36% nonactivated hPlts versus 98.36% ± 0.64% activated hPlts) (Figure S1A, SDC, https://links.lww.com/TP/C277). The percentage of Qtracker 800-labeled versus DCF hPlts in both platelet “types” was significantly decreased (P ≤ 0.001) (Figure S1A, SDC, https://links.lww.com/TP/C277). DCF and Qtracker do not influence the activation capacity of hPlts seen in only slight expression of CD62p and CD63 on nonactivated versus ex vivo–activated hPlts (for DCF: 10.13% ± 2.28% [CD62p] and 4.24% ± 0.68% [CD63] nonactivated versus 76.06% ± 6.63% [CD62] and 36.52% ± 5.13% [CD63] ex vivo–activated hPlts) (Figure S1B, SDC, https://links.lww.com/TP/C277) (for Qtracker: 19.60% ± 1.24% [CD62] and 8.30% ± 0.71% [CD63] versus 72.02% ± 1.02% 8 [CD62p] and 43.13% ± 1.07% [CD63]) (Figure S1C, SDC, https://links.lww.com/TP/C277). We labeled hPlts for in vivo studies with DCF.
In Vivo Monitoring of Nonactivated and Ex Vivo–activated hPlts in Rag2−/−γc−/− Mice Using the Multispectral Imaging System Maestro
Using DCF-labeled nonactivated and in vitro–activated hPlts for injection in the tail vein of C57/Bl6-Rag2−/−γc−/− mice, we observed their localization, distribution, and the temporal respectively local degradation at 5, 30, 60, 120 min, and also 24 h after injection using the multispectral imaging system Maestro. The background fluorescence spectrum of mouse tissues or cells was represented by pink and white pseudo-colors, whereas the fluorescence signal of DCF-labeled hPlts was indicated by a green pseudo-color. Over time, the fluorescence intensity decreased in both treatment groups; meanwhile, no DCF could be detected during the observed dates in the negative control (Figure 5). When comparing activated versus nonactivated hPlts injected in mice, the ex vivo–activated hPlts vanished much faster than the nonactivated hPlts seen in less green pseudo-color from date 60 min (n = 4 per group; Figure 5 shows representative images).
FIGURE 5.: In vivo monitoring of 4 × 108 nonactivated and ex vivo–activated DCF-labeled hPlts in Rag2−/−γc−/− mice using the multispectral imaging system Maestro in the temporal course at 5 min, 30 min, 60 min, 120 min, and 24 h after injection. Shown are representative mice in each group of nonactivated hPlts, of activated hPlts as well as of the untreated control group (n = 4 per group). The fluorescence signal of DCF-stained hPlts is represented in green. The spectrum of background fluorescence is illustrated by the colors pink and white. A, Images of the depilated mice bodies in a supine position. The bladder has been marked with an arrow. B, Images of the euthanized and laparotomised mice with an exteriorized intestine to view the organs accessible. DCF, dichlorodihydrofluorescein; hPlt, human platelet.
Furthermore, we regarded the in vivo localization of activated and nonactivated DCF-labeled hPlts in the lung and liver over time. For quantification of the mean fluorescence intensity, we generated (manually) regions of interest across the organs and evaluated the appropriate fluorescence intensity (Materials and Methods, SDC, https://links.lww.com/TP/C277). In both analyzed organs, fluorescence intensity dropped over time independent of the activation status of the injected hPlts (Figure 6A), but was much faster using activated compared with nonactivated hPlts (Figure 6A). The highest fluorescence intensity in both liver approaches was reached first after 30 min. This was confirmed by the mean fluorescence intensity quantification at different time points after tail vein injection (Material and Methods, SDC, https://links.lww.com/TP/C277; Figure 6B).
FIGURE 6.: After tail vein injection of DCF-labeled nonactivated and activated hPlts in Rag2−/− γc−/− mice, it can be seen that the fluorescence intensity in the lung and liver decreases over time, starting 5 and 30 min after injection, respectively. A, At the time points 5, 30, 60, and 120 min after injection, lungs and livers were removed and DCF-stained hPlts were detected using the multispectral imaging system Maestro. B, The diagrams show the quantification of the MFI in the lung and the liver. Data are represented as mean ± SEM (n = 4 per group; **P ≤ 0.01, *P ≤ 0.05, nonactivated hPlts vs activated hPlts). DCF, dichlorodihydrofluorescein; hPlt, human platelet; MFI, mean fluorescence intensity.
Interaction of hPlts With the Human Endothelium in C57/Bl6-Rag2−/−γc−/− Mice In Vivo
Using the confocal fluorescence laser endomicroscope we could detect the in vivo interaction of IV-injected DCF-labeled hPlts with the human endothelium of transplanted mammary artery xenografts on day 1 after transplantation in C57/Bl6-Rag2−/−γc−/− mice. The results show that both the nonactivated and in vitro–activated hPlts firmly adhere to the human endothelium of the mammary artery segment. Figure 7A–Cand for activated and Figure 7D–F for nonactivated hPlts each represent the same section of the human artery in the appropriate animal, which was recorded every 5 s. The pictures in Figure 7A and D were shot approximately 2 h after injection of hPlts. The circles indicate the firmly adherent hPlts. The sections of the human xenograft whose recipient was treated with in vitro–activated hPlts demonstrate that a higher number of hPlts firmly adhere to the human endothelium compared with the mammary artery segment of the recipient mouse, which was injected with nonactivated hPlts. As a control, according to the time course previously described, sections of a murine vessel after injection of DCF-labeled hPlts were analyzed, showing only a few hPlts firmly adhered to murine vessels (n = 3 animals per group).
FIGURE 7.: Both in vitro activated (A–C) and nonactivated (D–F) hPlts firmly adhere to the human endothelium of the mammary artery xenograft on d 1 after transplantation approximately 2 h after injection. As a control (G–I), only a few hPlts adhere to murine vessels. A–F, The serial images pictured every 5 s show the interaction of DCF-labeled hPlts with human endothelium in a segment of the human mammary artery xenograft on d 1 postsurgery after intravenous injection of in vitro–activated (A–C) and nonactivated (D–F) hPlts using the confocal fluorescence laser endomicroscope OptiScan FIVE 1. For comparison, a segment of murine vessels after injection of DCF-labeled hPlts is shown over time (G–I). Circles indicate adherent hPlts and arrows point out the hPlts present in circulation (n = 3 animals per group). DCF, dichlorodihydrofluorescein; hPlt, human platelet.
hPlts Injection Results in Significant Development of Transplant Arteriosclerosis
Mammary artery xenografts from C57/Bl6-Rag2−/−γc−/− recipients were analyzed on day 30 after transplantation, the time point at which previous studies showed a sufficiently high level of rejection.19,25 Recipient mice received daily 4 × 108 nonactivated or activated hPlts, respectively (from day 1 to 29 postoperatively; see also Figure 1). Previous experiments showed that the transplantation of human artery segments from diabetic patients resulted in a high rate of postoperative thrombosis [Ensminger et al, unpublished data, 2021]. Therefore, we excluded this type of human artery in our present experimental setup. Morphometric analysis of transplanted human artery xenografts showed that daily IV injection of both nonactivated and in vitro–activated hPlts results in a significant and similarly pronounced intimal proliferation (47.0% ± 5.67% intima proliferation [nonactivated hPlts], 43.20% ± 6.52% [activated hPlts], n = 5 per group, P ≤ 0.01) as compared to untreated controls with rare signs of vascular lesions (12.26% ± 5.21% intima proliferation [without hPlts], n = 5; Figure 8). This indicates that platelets alone trigger chronic inflammatory changes in transplanted vessels even in the absence of the adaptive immune system. None of the recipient mice showed any kind of hemorrhage or major postoperative bleeding.
FIGURE 8.: Morphometric analysis of the degree of intimal thickening using Elastica van Gieson-stained section. Areas between the lumen and the internal elastica lamina were circumscribed manually and measured. The pictures show human xenografts of C57/Bl6-Rag2−/−γc−/− mice explanted on d 30 after surgery (original magnification ×100). Transplanted arteries injected with nonactivated (B) and in vitro–activated hPlts (C) showed significantly elevated neointima formation compared with untreated controls (A). D, Quantification of intimal proliferation performed using ANAlysis Image Analysis software (Olympus Germany) (n = 5 animals per group, **P ≤ 0.01 vs untreated controls, ***P ≤ 0.001 vs untreated controls). hPlt, human platelet.
Injection of Activated hPlts Showed Significantly Increased Intercellular Adhesion Molecule-1, Platelet-derived Growth Factor Beta, and Alpha-Smooth Muscle Actin Protein Expression in Mammary Artery Xenografts
After daily IV injection of in vitro–activated hPlts, we could detect a significantly increased expression of human intercellular adhesion molecule-1 (ICAM-1; 11.64% ± 2.17%, P ≤ 0.05) and human platelet-derived growth factor receptor beta (PDGFRβ; 7.72% ± 1.00%, P ≤ 0.01) in the intima of the transplanted xenografts as compared to transplanted artery segments treated with nonactivated hPlts (3.69% ± 1.71% [ICAM-1], 2.87% ± 1.01% [PDGFRβ]; Figure 9A–C). In the adventitia and media of implanted xenografts treated with in vitro–activated hPlts, expression of human ICAM-1 and PDGFRβ was also higher than observed in the group of nonactivated hPlts, but did not reach significance (adventitia: 8.93% ± 3.50% versus 5.79% ± 2.12% [ICAM-1], 20.01% ± 6.02% versus 10.23% ± 2.30% [PDGFRβ]) (media: 7.47% ± 3.97% versus 2.20% ± 0.47% [ICAM-1], 10.38% ± 3.80% versus 4.05% ± 1.09% [PDGFRβ]; Figure 9A–C). To detect an intimal proliferation as a result of smooth muscle cell (SMC) immigration, mammary artery segments were marked with a monoclonal antibody against human alpha-smooth muscle actin. Quantification revealed a significant increase of human alpha-smooth muscle actin within the intima (33.80% ± 6.74% [activated hPlts], 11.10% ± 3.15% [nonactivated hPlts], P ≤ 0.01) and consequently a significantly decreased expression in the media (21.80% ± 4.83% [activated hPlts], 38.28% ± 4.11% [nonactivated hPlts], P ≤ 0.01) of mammary arteries after continuous treatment with in vitro–activated hPlts compared with the group of nonactivated hPlts (Figure 9A, subpanels e and f, and D) (all analysis were done with n = 5 per group).
FIGURE 9.: Immunofluorescence staining of cryostat sections of human mammary arteries on d 30 after transplantation. A significantly increased expression of human ICAM-1 (A, subpanels a and b) and human PDGFRβ (A, subpanels c and d) can be seen in xenografts treated with in vitro–activated hPlts. Staining with an antibody against human α-SMA revealed strong SMC migration into the developing neointima in xenografts treated with in vitro–activated hPlts (A, subpanels e and f). The adventitia is shown on the left and the intima and the lumen of the transplanted artery are shown on the right of each individual section. One representative section is shown of 5 experiments (original magnification ×100). B–D, Quantification of human ICAM-1 (B), human PDGFRβ (C), and human α-SMA (D) per artery wall layer of the xenograft on d 30 after transplantation was performed with computerized image analysis using CellSense Dimension. The positive stained area in relation to the total area of each section [in %] was analyzed using an original magnification of ×100 (n = 5 animals per group; **P ≤ 0.01, *P ≤ 0.05; adventitia, media, and intima of the group + nonactivated hPlts vs adventitia, media, and intima of the group + activated hPlts). α-SMA, alpha-smooth muscle cell actin; SMC, smooth muscle cell; ICAM-1, intracellular adhesion molecule-1; PDGFR, platelet-derived growth factor receptor.
DISCUSSION
Platelets are mainly known for their central role as cellular mediators of primary and secondary hemostasis. However, experimental studies have shown that platelets are also crucially involved in the regulation of innate and adaptive immunities, especially in the initiation and development of atherosclerosis.26,27 Data of this study demonstrate for the first time that hPlts are able to interact with the endothelium of human arterial xenografts in mice in the absence of adaptive immunity and thereby induce neointima formation. This effect was more pronounced using ex vivo–activated compared with nonactivated hPlts. Further analysis revealed a significantly increased expression of ICAM-1 and PDGFRβ with progressing SMC migration from the vessel media into the developing neointima as underlying mechanisms.
To validate our experimental model, we sought to ensure that vital human platelets circulate in the bloodstream of transplanted C57/Bl6-Rag2−/−γc−/− recipients throughout the entire trial period. Injection via the lateral tail vein proofed to be the most suitable access route as previously described.28 Furthermore, flow cytometric analysis demonstrated nonactivated hPlts remaining within the mouse circulation for a significantly longer time than ex vivo–activated hPlts but becoming activated over time. Additional investigations using multispectral imaging and confocal fluorescence laser endomicroscope for in vivo localization and distribution of DCF-labeled hPlts in C57/Bl6-Rag2−/−γc−/− mice confirmed that signal strength decreases in a time-dependent manner after injection of both nonactivated and ex vivo–activated hPlts. A previous study by Hu and Yang concluded that increased hPlt decomposition in a humanized mouse model may have been the consequence of a mouse macrophage-mediated rejection. It was also demonstrated that depletion of mouse macrophages using clodronate liposomes resulted in an improved hPlt reconstitution of NOD/SCID or NOD/SCID/γc−/− mice.29 However, in our experimental model, a sufficiently high hPlt reconstitution was achieved by a daily hPlt tail vein injection in transplanted C57/Bl6-Rag2−/−γc−/− mice. Therefore, macrophages were not depleted in our model to exclude potential side effects of clodronate on the development of vascular lesions.
The next question was if the injected hPlts were able to adhere to and activate human endothelial cells in our humanized transplant model. Using the high-resolution imaging method of confocal fluorescence laser endomicroscopy, it was shown that up to 2 h after injection, hPlts stably adhere to and activate human endothelial cells as shown by an increased surface expression of ICAM-1.
Our study demonstrates that an isolated application of both nonactivated and in vitro–activated hPlts in the absence of lymphocytes results in a similarly pronounced intimal proliferation of human artery xenografts. In contrast, nontreated human mammary artery grafts showed only discrete signs of intima proliferation in the absence of hPlts as in previous studies.19,20 Endogenous mouse platelets alone do not significantly lead to intimal proliferation of human vessels and, as described previously, do not adhere on xenogeneic endothelium due to species incompatibility.30
Both in vitro– and in vivo–activated hPlts transiently adhere to the structurally and functionally intact but activated endothelium of the human xenograft, initiated by the interaction between platelet glycoprotein (GP) Ib-IX-V receptor complex (von Willebrand factor receptor complex) and is secreted and expressed by endothelial cells.6 In addition, platelets detect endothelial P-selectin via P-selectin GP ligand 1 or GPIbα, followed by subsequent platelet adhesion.6,7 We could show, using the high-resolution imaging method of OptiScan FIVE 1, that 2 h after IV injection of TRAP-6–activated hPlts, a higher number of hPlts adhere to the endothelium of transplanted human mammary artery grafts than after injection of nonactivated hPlts, possibly due to the amplification effect of TRAP-6 as a prothrombotic agonist of the protease-activating receptor 1. The activation of protease-activating receptors is a key mechanism during cardiac surgery with extracorporeal circulation.31 Another explanation for the notably slower kinetics of the adhesion process of primary nonactivated hPlts may be the required time for an in vivo activation within the mouse blood circulation. This in vivo activation process is most likely also initiated by shear stress32 as well as by the initial, transient contact of the platelet GPIbα receptor with human endothelial secreted von Willebrand factor.6,33 Therefore, the firm adhesion of hPlts to the human endothelium of the xenograft may be delayed. In addition, as a result of the injection of 4 × 108 hPlts in C57/Bl6-Rag2−/−γc−/− mice, the total platelet amount of the recipient was increased by 24%,34 which may also have caused higher shear stress and thereby further activation of hPlts.35
Although the transplanted C57/Bl6-Rag2−/−γc−/− mice of both treated groups showed a similar intimal proliferation, we detected differences in the composition of the protein expression in the vessel wall of human mammary artery grafts. An increased expression of human ICAM-1 and human PDGFRβ within the intima, media, and adventitia of human arteries treated with in vitro–activated hPlts was observed in contrast to the group treated with nonactivated hPlts. This could be explained as TRAPs are thrombin receptor agonists and thrombin is one of the most potent platelet activation factor.36,37 Our results show TRAP-6 effectively activating hPlts and hence initiating a strong proinflammatory response on the human endothelium and consequently an increased migration of vascular SMCs toward the intima compared with only in vivo–activated hPlts. This was supported by increased PDGFRβ expression in grafts of mice injected with TRAP-6–activated hPlts. PDGFRβ is an important receptor involved in stimulation of cell growth and cell motility and responds to stimuli like thrombin.38,39 Further studies demonstrated that neointimal proliferation is associated with both a significant migration of SMCs toward the intima and a significantly increased expression of PDGF or PDGFRβ.40-43 Blocking the PDGFRβ results in hampered SMC migration.44 Other potential mechanisms include CD40L on activated platelets inducing endothelial cells to secrete signature chemokines and to express adhesion molecules, such as ICAM-1 and VCAM-1,45 and have been investigated in previous experimental mouse studies.10,11
In summary, our findings affirm both in vitro– and in vivo–activated hPlts as an isolated factor responsible for the development of transplant arteriosclerosis in a humanized mouse artery xenograft model and support the therapeutic usage of the antiplatelet agents such as clopidogrel for effectively reducing transplant arteriosclerosis in experimental vascular grafts.10,11 This humanized C57/Bl6-Rag2−/−γc−/− mouse xenograft model may be useful to investigate further cell subpopulations required for the development of transplant arteriosclerosis and to test novel therapeutic strategies46,47 as well as drug interactions altering the platelet response and ultimately to prevent the development of CAV in human patients.
ACKNOWLEDGMENTS
The authors would like to thank the staff of the transfusion medicine of the University of Erlangen-Nürnberg for the provided human platelet concentrate. They would also like to thank Nina Wollin from their experimental laboratory for her technical assistance and the staff of the animal facility of the University of Erlangen-Nürnberg for their expert care of animals used for this study.
REFERENCES
1. Semple JW, Freedman J. Platelets and innate immunity. Cell Mol Life Sci. 2010;67:499–511.
2. Semple JW, Italiano JE Jr, Freedman J. Platelets and the immune continuum. Nat Rev Immunol. 2011;11:264–274.
3. Jurk K, Kehrel BE. Pathophysiology and biochemistry of platelets. Internist (Berl). 2010;51:1086–1094.
4. Modjeski KL, Morrell CN. Small cells, big effects: the role of platelets in transplant vasculopathy. J Thromb Thrombolysis. 2014;37:17–23.
5. Smyth SS, McEver RP, Weyrich AS, et al.; 2009 Platelet Colloquium Participants. Platelet functions beyond hemostasis. J Thromb Haemost. 2009;7:1759–1766.
6. Gawaz M, Langer H, May AE. Platelets in inflammation and atherogenesis. J Clin Invest. 2005;115:3378–3384.
7. Siegel-Axel DI, Gawaz M. Platelets and endothelial cells. Semin Thromb Hemost. 2007;33:128–135.
8. Yashiro K, Matsumoto Y, Ihara H, et al. Involvement of platelet activation by P2Y12 receptor in the development of transplant arteriosclerosis in mice. Transplantation. 2009;87:660–667.
9. Gawaz M, Brand K, Dickfeld T, et al. Platelets induce alterations of chemotactic and adhesive properties of endothelial cells mediated through an interleukin-1-dependent mechanism. Implications for atherogenesis. Atherosclerosis. 2000;148:75–85.
10. Abele S, Weyand M, Wollin M, et al. Clopidogrel reduces the development of transplant arteriosclerosis. J Thorac Cardiovasc Surg. 2006;131:1161–1166.
11. Abele S, Spriewald BM, Ramsperger-Gleixner M, et al. Attenuation of transplant arteriosclerosis with clopidogrel is associated with a reduction of infiltrating dendritic cells and macrophages in murine aortic allografts. Transplantation. 2009;87:207–216.
12. Adigbli G, Ménoret S, Cross AR, et al. Humanization of immunodeficient animals for the modeling of transplantation, graft versus host disease, and regenerative medicine. Transplantation. 2020;104:2290–2306.
13. Banuelos SJ, Shultz LD, Greiner DL, et al. Rejection of human islets and human HLA-A2.1 transgenic mouse islets by alloreactive human lymphocytes in immunodeficient NOD-scid and NOD-Rag1(null)Prf1(null) mice. Clin Immunol. 2004;112:273–283.
14. Goto R, Issa F, Heidt S, et al. Ischemia-reperfusion injury accelerates human antibody-mediated transplant vasculopathy. Transplantation. 2013;96:139–145.
15. McCune JM, Namikawa R, Kaneshima H, et al. The SCID-hu mouse: murine model for the analysis of human hematolymphoid differentiation and function. Science. 1988;241:1632–1639.
16. Melkus MW, Estes JD, Padgett-Thomas A, et al. Humanized mice mount specific adaptive and innate immune responses to EBV and TSST-1. Nat Med. 2006;12:1316–1322.
17. Pober JS, Bothwell AL, Lorber MI, et al. Immunopathology of human T cell responses to skin, artery and endothelial cell grafts in the human peripheral blood lymphocyte/severe combined immunodeficient mouse. Springer Semin Immunopathol. 2003;25:167–180.
18. Lorber MI, Wilson JH, Robert ME, et al. Human allogeneic vascular rejection after arterial transplantation and peripheral lymphoid reconstitution in severe combined immunodeficient mice. Transplantation. 1999;67:897–903.
19. Abele-Ohl S, Leis M, Mahmoudian S, et al. Rag2
-/- gamma-chain
-/- mice as hosts for human vessel transplantation and allogeneic human leukocyte reconstitution. Transpl Immunol. 2010;23:59–64.
20. Abele-Ohl S, Leis M, Wollin M, et al. Human cytomegalovirus infection leads to elevated levels of transplant arteriosclerosis in a humanized mouse aortic xenograft model. Am J Transplant. 2012;12:1720–1729.
21. Cazenave JP, Mulvihill JN, Sutter-Bay A, et al. A Centrifugation Technique for the Preparation of Suspensions of Nonactivated Human Platelets. Kluwer Academic Publishers; 1993:431–439.
22. Dickfeld T, Lengyel E, May AE, et al. Transient interaction of activated platelets with endothelial cells induces expression of monocyte-chemoattractant protein-1 via a p38 mitogen-activated protein kinase mediated pathway. Implications for atherogenesis. Cardiovasc Res. 2001;49:189–199.
23. Zimmermann R, Reske S, Metzler P, et al. Preparation of highly concentrated and white cell-poor platelet-rich plasma by plateletpheresis. Vox Sang. 2008;95:20–25.
24. Xu F, Gelderman MP, Farrell J, et al. Temperature cycling improves in vivo recovery of cold-stored human platelets in a mouse model of transfusion. Transfusion. 2013;53:1178–1186.
25. Ensminger SM, Spriewald BM, Witzke O, et al. Kinetics of transplant arteriosclerosis in MHC-class I mismatched and fully allogeneic mouse aortic allografts. Transplantation. 2002;73:1068–1074.
26. Huo Y, Ley KF. Role of platelets in the development of atherosclerosis. Trends Cardiovasc Med. 2004;14:18–22.
27. Lievens D, von Hundelshausen P. Platelets in atherosclerosis. Thromb Haemost. 2011;106:827–838.
28. Fuhrmann J, Jouni R, Alex J, et al. Assessment of human platelet survival in the NOD/SCID mouse model: technical considerations. Transfusion. 2016;56:1370–1376.
29. Hu Z, Yang YG. Full reconstitution of human platelets in humanized mice after macrophage depletion. Blood. 2012;120:1713–1716.
30. Tellides G, Tereb DA, Kirkiles-Smith NC, et al. Interferon-gamma elicits arteriosclerosis in the absence of leukocytes. Nature. 2000;403:207–211.
31. Landis C. Why thrombin PAR1 receptors are important to the cardiac surgical patient. J Extra Corpor Technol. 2007;39:305–307.
32. Jesty J, Yin W, Perrotta P, et al. Platelet activation in a circulating flow loop: combined effects of shear stress and exposure time. Platelets. 2003;14:143–149.
33. Dayananda KM, Singh I, Mondal N, et al. von Willebrand factor self-association on platelet GpIbalpha under hydrodynamic shear: effect on shear-induced platelet activation. Blood. 2010;116:3990–3998.
34. Schmitt A, Guichard J, Massé JM, et al. Of mice and men: comparison of the ultrastructure of megakaryocytes and platelets. Exp Hematol. 2001;29:1295–1302.
35. Schulz-Heik K, Ramachandran J, Bluestein D, et al. The extent of platelet activation under shear depends on platelet count: differential expression of anionic phospholipid and factor Va. Pathophysiol Haemost Thromb. 2005;34:255–262.
36. Dardik R, Savion N, Kaufmann Y, et al. Thrombin promotes platelet-mediated melanoma cell adhesion to endothelial cells under flow conditions: role of platelet glycoproteins P-selectin and GPIIb-IIIA. Br J Cancer. 1998;77:2069–2075.
37. Montalescot G. Platelet biology and implications for antiplatelet therapy in atherothrombotic disease. Clin Appl Thromb Hemost. 2011;17:371–380.
38. Fager G. Thrombin and proliferation of vascular smooth muscle cells. Circ Res. 1995;77:645–650.
39. Heldin CH, Westermark B. Mechanism of action and in vivo role of platelet-derived growth factor. Physiol Rev. 1999;79:1283–1316.
40. Buetow BS, Tappan KA, Crosby JR, et al. Chimera analysis supports a predominant role of PDGFRbeta in promoting smooth-muscle cell chemotaxis after arterial injury. Am J Pathol. 2003;163:979–984.
41. Heim C, Eckl S, Preidl R, et al. Delayed therapy with clopidogrel and everolimus prevents progression of transplant arteriosclerosis and impairs humoral alloimmunity in murine aortic allografts. Eur J Cardiothorac Surg. 2015;47:180–187.
42. Jansen MA, Otten HG, de Weger RA, et al. Immunological and fibrotic mechanisms in cardiac allograft vasculopathy. Transplantation. 2015;99:2467–2475.
43. Mancini MC, Evans JT. Role of platelet-derived growth factor in allograft vasculopathy. Ann Surg. 2000;231:682–688.
44. Gocht A, Spriewald B, Distler JHW, et al. Small molecule tyrosine kinase inhibitor nintedanib reduces development of cardiac allograft vasculopathy in murine aortic allografts. Transplant Direct. 2018;4:e367.
45. Henn V, Slupsky JR, Gräfe M, et al. CD40 ligand on activated platelets triggers an inflammatory reaction of endothelial cells. Nature. 1998;391:591–594.
46. Heim C, Gocht A, Weyand M, et al. New targets for the prevention of chronic rejection after thoracic organ transplantation. Thorac Cardiovasc Surg. 2018;66:20–30.
47. Nikolova AP, Kobashigawa JA. Cardiac allograft vasculopathy: the enduring enemy of cardiac transplantation. Transplantation. 2019;103:1338–1348.
