Blockage of CX3CL1 Attenuates Platelet and Leukocyte Recruitment in Murine Hepatic I/R

Introduction: The chemokine fractalkine (CX3CL1) is critically involved in the pathophysiology of different inflammatory diseases and myocardial ischemia-reperfusion (I/R). This study aimed to analyze the role of CX3CL1 in the activation of platelets and leukocytes during hepatic I/R. Methods: Under inhalation anesthesia, C57BL6 mice were subjected to warm hepatic I/R (90 min/240 min). The animals were pretreated either with a function-blocking anti-mouse CX3CL1 antibody or IgG control administered systemically before ischemia. Sham-operated animals served as controls (n = 7 each group). The inflammatory response and sinusoidal perfusion failure were evaluated by intravital microscopy. Hepatic transaminases plasma levels and histopathological tissue damage were determined as markers of hepatocellular injury. Results: Sinusoidal perfusion failure, leukocyte recruitment to the liver, and transaminase activities were sharply increased upon I/R compared to sham-operated mice. Firm adhesion of platelets and concordantly leukocytes to endothelial cells is reduced significantly by a function-blocking anti-CX3CL1 antibody. We demonstrate that inhibition of CX3CL1 signaling attenuates leukocyte adhesion in the postischemic liver but does not significantly ameliorate overall perfusion failure and hepatocellular injury. Discussion/Conclusion: Our in vivo data demonstrate a mild attenuating effect of CX3CL1 blockade on platelet and leukocyte, but not CD4+ T cell accumulation and activation in hepatic I/R injury. We report a significant effect of blocking chemokine CX3CL1 on sinusoidal perfusion failure without considerably improving overall hepatocellular injury during early reperfusion.


Introduction
Ischemia-reperfusion (I/R) injury causes significant morbidity and mortality in organ transplantation, trauma, and septic or hemorrhagic shock. It is one of the main reasons for early organ failure after liver transplantation. The complex cascade of I/R-induced injury includes different mechanisms of cell death and the activation of im- mune cells in a sterile inflammatory response. Hepatic microcirculation is considered one of the primary targets of I/R injury of the liver. The injured endothelial lining upregulates adhesion molecules, allowing the interaction of neutrophils, CD4+ T cells, and platelets [1]. The accumulation and interaction of these leukocytes initiate a multistep inflammatory cascade.
Endothelial leukocyte capture with subsequent rolling and firm adhesion is followed by extravasation into the perivascular tissue [2,3]. Small chemoattractant proteins attract and facilitate leukocyte recruitment to the site of inflammation. Therefore, these chemokines and their receptors are effective therapeutic targets in inflammatory processes. Fractalkine (CX3CL1) is a chemokine of the CX3C family [4]. Increasing evidence indicates its critical involvement in the pathophysiology of different inflammatory diseases like atherosclerosis [5,6]. The CX3CR1 is profusely expressed on various cell types, such as circulating leukocytes [7,8].
Interaction between endothelial CX3CL1 and its receptor mediates chemoattraction and promotes leukocyte adhesion in an integrin-independent way [7,9,10]. Correspondingly, leukocytes are recruited to CX3CL1coated surfaces in venules or sinusoids [9]. Interestingly, platelets, which express CX3CR1, are strictly required for CX3CL1-induced leukocyte adhesion [8]. Endothelial CX3CL1 also induces activation of surface-adherent platelets with consecutive platelet degranulation and exposure of the adhesion molecule P-selectin. In the liver, CX3CR1 is most abundantly expressed on Kupffer cells (KCs) [11] and endothelial cells, while hepatic stellate cells (HSCs) are the main source of CX3CL1 [12]. In primary biliary cirrhosis, CX3CL1 has been shown to mediate lymphocyte recruitment to the liver [13]. In this study, we test the hypothesis that blocking CX3CL1/CX3CR1 axis attenuates postischemic recruitment of leukocytes, CD4+ T cells, and platelets, resulting in protection from I/R injury without increasing the risk for hemorrhage. Fig. 1. Experimental setup. a Female 5-to 7-week-old C57BL/6 mice were injected with function-blocking anti-CX3CL1 monoclonal antibody TP233AF or isotype control 24 h and 30 min before ischemia. b Reversible ischemia of the left liver lobe was achieved by clamping the supplying nerve vessel bundle. A jugular catheter was placed for administration of saline, fluorophores, and ex vivo labeled cells from donor mice. A catheter in the ipsilateral carotid artery was used for invasive monitoring of blood pressure. c Modified Leitz-Orthoplan microscope was used for intravital fluorescence microscopy before collecting (d) the postischemic liver and whole blood for further analysis.

Experimental Research on Vertebrates
Female 5-to 7-week-old C57BL/6 mice were used in all experiments. The mice were purchased from Charles River ® (Sulzfeld, Germany) and were housed for at least 1 week in our animal facility to recover from transportation before experiments. All experiments were carried out according to German animal safety regulations and the ARRIVE guidelines.

Intravital Fluorescence Microscopy
A modified Leitz-Orthoplan microscope was used for intravital fluorescence microscopy, as described previously (17). Leukocytes were stained in vivo by rhodamine 6G (0.05%, 100 μL, i.v., Sigma, Taufkirchen, Germany). Fluorescein isothiocyanate-conjugated dextran (MW 150000; 100 μL, 5%, Sigma) was used as plasma marker for sinusoidal perfusion. Platelets or CD4+ T cells were isolated from syngeneic mice with a magnetic cell separation kit according to the manufacturer's instructions and labeled ex vivo with rhodamine 6G before intravenous application and intravital microscopy. For detailed protocol of intravital microscopy, see

Hepatocellular Injury
Serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) activity was determined with an automated analyzer using standardized test systems, see online supplementary protocol 2. The degree of liver damage in the postischemic left lobe was assessed by a semiquantitative score described elsewhere [15] in hematoxylin/eosin stained slides.

Statistics
ANOVA on ranks followed by the Student-Newman-Keuls test was used to estimate stochastic probability in intergroup comparison (SigmaPlot 12, Jandel Scientific, Erkrath, Germany). Mean values ± SEM are given. p values less than 0.05 were considered significant.

Leukocyte-Endothelial Cell Interactions
Leukocyte-endothelial cell interactions were analyzed in postsinusoidal venules as an indicator of microvascular injury and hepatic inflammation after I/R. As shown in Figure 2, the numbers of rolling and firmly adherent leukocytes (1.9 ± 0.3/mm/s; 14 ± 1.4/mm 2 ) in postsinusoidal venules were very low in sham-operated mice in contrast to 90 min of ischemia followed by 120 min of reperfusion in vehicle-treated mice (7.9 ± 0.7/mm/s; 241.4 ± 17.1/ mm 2 ). Blocking CX3CL1 by using function-blocking anti-mouse CX3CL1 antibody reduced the postischemic number of rolling and adherent leukocytes by about ∼20% respective ∼50% (Fig. 3), compared to the isotype IgG-treated I/R group (4.4 ± 0.5/mm/s; 105 ± 8.2/mm 2 ), (data not shown).

Migration of CD4+ T Cells
As shown in Figure 3, the number of CD4+ T cells accumulated in sinusoids did not differ between the I/R groups undergoing treatment with anti-CX3CL1 antibody and control antibody. Similar findings were observed after both 30 min and 120 min of reperfusion.

Sinusoidal Perfusion Failure
Sinusoidal perfusion failure was determined using in vivo microscopy as a recognized parameter of microvascular I/R injury. In the vehicle-treated I/R group, about 29 ± 1% of all sinusoids were not per-fused (Fig. 4). Anti-CX3CL1 treatment significantly improved post-I/R perfusion failure (14 ± 2% nonperfused sinusoids).

Hepatocellular Injury
Plasma activity of hepatic transaminases was measured as a marker of hepatocellular necrotic injury. Hepatic I/R (90 min/240 min) increased the activity of AST and ALT considerably in the vehicle-treated group compared to the sham-operated group. In contrast to the data on sinusoidal perfusion, blocking CX3CL1 did not have a protective effect (Fig. 4).
Histopathologic analysis showed hydropic degeneration and necrotic cells after I/R compared to the shamoperated mice. Like the liver enzyme activity data, treatment with anti-CX3CL1 antibody did not improve the postischemic tissue injury (histology score 1.9 ± 0.5 vs. 1.9 ± 0.8, Fig. 4).

Discussion
Experimental and clinical studies suggest a crucial role of platelets in the formation of hepatic I/R injury. Thrombocytopenia markedly reduces edema formation and leukocyte infiltration in multiple acute or chronic inflammation models. Blocking the chemokine CX3CL1 seems a promising approach to attenuate proinflammatory leu- CX3CL1 is mainly expressed and secreted by HSCs and sinusoidal endothelium [12]. It has been reported to be involved in lymphocyte recruitment to the liver [13]. CX3CR1 on KCs in the liver recognizes CX3CL1 expression [11] of inflamed endothelial cells, as do activated platelets, which are consecutively recruited in large numbers to the surface of the inflamed endothelium [16].

Platelet Recruitment to the Postischemic Liver
The initial contact with activated endothelial cells leads to platelet activation and permanent adherence to intercellular adhesion molecule 1 [1]. Fibrinogen serves as a bridging molecule [2]. Schulz et al. [16] show that CX3CL1 on inflamed endothelium activates recruited platelets and thereby initiates leukocyte accumulation.
Moreover, data from numerous cardiovascular studies suggest that the CX3CR1/CX3CL1 axis affects the recruitment and activation of platelets and inflammatory monocytes to activated endothelial cells [17]. Our in vivo results demonstrate that inhibition of CX3CL1 axis with function-blocking anti-mouse CX3CL1 antibody significantly attenuates platelet-endothelium interactions after hepatic I/R. Both I/R-induced platelet rolling and firm adherence were reduced after treatment. These findings are consistent with the function of CX3CL1 as a platelet-activating chemokine and explain our observation of fewer platelets being recruited to the postischemic liver. The marked reduction of platelet adhesion during early reperfusion despite hepatocellular disruption might be partly attributed to the blocked chemotactic capacities of HSCs, who are the main hepatic source of CX3CL1 and respond rapidly to liver damage. Leukocyte Recruitment in Hepatic I/R Independently of its functions in chemotaxis, CX3CL1 directly induces capture and firm adhesion of flowing leukocytes in sinusoids and venules [9,10]. As CX3CL1 is expressed on endothelial cells and readily produced by HSCs, we assumed that interrupting CX3CL1-dependent platelet activation by function-blocking anti-mouse CX-3CL1 antibody TP233AF would also attenuate leukocyte recruitment. P-selectin contributes to the initial recruitment of leukocytes to the injury site during inflammation. However, it is not expressed on the endothelium of hepatic sinusoids, and recent data show that leukocytes adhere to liver sinusoids in a platelet-dependent manner. Platelets bind to hyaluronic acid on liver sinusoid endothelial cells via CD44 and pave the way for the adhesion of immune cells [18]. Therefore, we hypothesized that CX3CR1/CX3CL1 axis can alter both the classical and nonclassical way of leukocyte adhesion. CX3CL1-induced P-selectin release from platelets contributes to plateletleukocyte interaction and facilitates leukocyte adhesion [16]. Suitable for this assumption, we observe a significant decrease in leukocyte migration in the postischemic liver during early reperfusion. A possible explanation for this observation is the CX3CL1-induced change in the recruiting process of leukocytes. When CX3CL1 is present, the mechanism for leukocyte adhesion is altered. The chemokine domain acts as an adhesion molecule making the association with proteoglycans and other adhesion molecules unnecessary [16]. The association between CX3CR1 and integrins through the co-expression of CX-3CL1 and integrin ligands, such as intercellular adhesion molecule-1 and vascular adhesion molecule-1, potentiates cell adhesion even further [9].
Migration of CD4+ T Cells to the Postischemic Liver CX3CL1 expressed on inflamed endothelium is thought to act as a vascular gateway for (CX3CR1-expressing) effector T cells by rapidly capturing them from the blood and promoting tissue migration [9]. T cells, especially CD4+ T cells, play a major role in the pathogenesis of hepatic I/R [14]. CD4+ T cells are not cytotoxic but contribute to I/R injury by modulating activation and function of other cells, such as platelets, endothelial cells, and HSCs. The ability of CX3CL1 to act as a chemoattractant for CD4+ T cells suggests an important role in the recruitment of these cells to the postischemic liver. Moreover, allograft rejection, a process also characterized by an intense cellular immune response with an influx of circulating leukocytes into the transplanted organ [19], could be significantly reduced by treatment with CX-3CR1-specific blocking antibodies. However, the accumulation of CD4+ T cells in the postischemic liver was not significantly affected by anti-CX3CL1 antibody treatment. Recent findings of our group show that T cells directly interact with hepatic dendritic cells [15]. Therefore, CD4+ T cells might not depend on the chemotactic effect of CX3CL1 in their activation, especially in the early phase of hepatic I/R analyzed in this study.

Organ Damage and Hepatic Reperfusion Injury
We aimed to investigate whether the treatment with a chemokine inhibitor is protective against I/R injury. Such an effect seemed consequential as platelets, neutrophils, and T cells are known to mediate necrotic and apoptotic injury in the postischemic liver and play a critical role in sinusoidal perfusion failure [15,[20][21][22][23]. Platelets interact with other inflammatory cells, especially neutrophils, and can directly induce tissue injury by releasing free oxygen radicals and inflammatory mediators. Our results show a significant effect of inhibition of CX3CL1 on platelet and neutrophil activation and sinusoidal failure. However, AST/ALT levels and histopathologic damage as markers of overall hepatocellular integrity and liver necrosis were not significantly affected. The decisive mechanism why platelet and leukocyte interaction are blocked but migration and liver injury are not affected may be found in the activity of KCs. Experimental data from a murine model of liver injury shows that KCs as tissue-resident liver macrophages play an essential role in initiating and maintaining inflammatory responses. They release proinflammatory cytokines and chemokines and activate other nonparenchymal liver cells, such as endothelial or HSCs [24]. Therefore, KCs have strong detrimental or regenerative capacity in liver injury. While downregulating KC activity has been shown to be beneficial, blocking KCs obliterates their immunomodulatory function [25,26]. As KCs strongly react to the CX-3CL1/CX3CR-axis [11], we hypothesize that the beneficial effect of TP233AF on platelet and leukocyte recruitment is completely antagonized by the harmful blockade of KC function [27]. Another explanation for our findings is the beneficial effect of CX3CR1/CX3CL1 on the inflammatory activity of dendritic cells. Sutti et al. [28] show that the CX3CL1/CX3CR1 axis affects the IL-10 mediated anti-inflammatory activity of type 2 myeloid liver DCs. Moreover, blockade of CX3CL1 reduced monocyte-derived proinflammatory DC-mediated liver damage in a hepatotoxic model of inflammatory liver injury [29]. Contrary to hepatocytes undergoing hypoxic cell death, ischemic tissue potentially destined for ischemic infarction but not irreversibly injured might be susceptible to immunomodulatory Eur Surg Res 2023;64:185-192 DOI: 10.1159/000524024 rescue and benefit from an ameliorated microperfusion during late reperfusion. Inhibiting the anti-inflammatory DC response and the regenerative capacity of KCs might cancel this road to tissue rescue. Further studies with a longer reperfusion time should be carried out to validate this hypothesis.

Conclusion
Taken together, our in vivo data show a mild attenuating effect of CX3CL1 blockade on platelet and leukocyte, but not CD4+ T cell accumulation and activation in hepatic I/R injury. We report a significant effect of blocking chemokine CX3CL1 on sinusoidal perfusion failure without considerably improving overall hepatocellular injury during early reperfusion.

Statement of Ethics
Our experiments were reviewed by the Government of Upper Bavaria before they were performed and approved as applications 8-13 and 23-13.