**3. Discussion**

As shown in our study, three SVMP toxins, Jar, Jar-C, and BnP1, isolated from *Bothrops* venoms, induce the expression of adhesion molecules on the microvasculature of murine cremaster muscle. Each of these toxins shows essential differences in the compositions of their structural domains. In general, PIII SVMPs (Jar) are more hemorrhagic than PI SVMPs (BnP1), although both degrade extracellular matrix molecules [29,30]. The stronger hemorrhagic activity of PIII SVMPs has been attributed to the presence of a disintegrin-like domain adjacent to the catalytic domain. This disintegrin domain favors the anchoring of the molecule to the basement membrane, amplifying the hemorrhagic potential [30]. Some SVMPs of *Bothrops* venoms degrade components of the basement membrane, increasing vascular permeability, and induce the expression of adhesion molecules [30,31]. Although Jar-C does not contain the catalytic domain, the ECD-disintegrin and cysteine-rich domains have also been shown to be essential for the inflammatory function of these toxins [9,16].

Unlike the hemorrhagic action of the SVMPs, the toxins Jar, Jar-C, and BnP1 exerted similar effects on inducing the expression of adhesion molecules responsible for changes in the leukocyte-endothelium interaction, regardless of the composition of their domains. Our results reveal the responses of endothelial and leukocyte cells to the presence of toxins through the expression of the key molecules ICAM-1, CD11a, and CD11b in the early stages of cellular migration and the expression of PECAM-1 in the late stages studied (Figures 1–4).

Leukocyte recruitment is regulated by factors that alter microvessels to promote firm adhesion and cell migration to the site of infection or tissue injury, inducing inflammation and subsequent tissue repair [25]. A cascade of events occurs to achieve this effect, including endothelial cell activation, the release of inflammatory mediators and the expression of adhesion molecules. These processes are often amplified by pathogen-associated molecular patterns (PAMPs) [32], damage-associated molecular patterns (DAMPs) [33] or venom-associated molecular patterns (VAMPs) [34] that function as alarmins stimulating the innate immune response alone or through interactions with cytokines and chemicals.

*Bothrops* venoms induce inflammatory responses and activate signaling pathways that culminate in the transcription of inflammatory genes such as cytokines and eicosanoids [35,36], inducing endothelial activation to promote capture, rolling, firm adhesion and cell migration [37,38].

Previous studies by our group have identified the critical role for SVMPs in the proinflammatory cytokine action induced by *Bothrops jararaca* and *Philodryas patagoniensis* venoms; both venoms contain a large amount of metalloproteases [8,39]. Additionally, we recently reported that Jar, Jar-C, and BnP1 induce cell adhesion and migration in postcapillary venules in the cremaster muscle of mice, as observed using intravital microscopy [9]. These data corroborate the increase in the expression of adhesion molecules observed in the present study.

Other PIII SMVPs, namely HF3 isolated from *Bothrops jararaca* venom and patagonfibrase isolated from the Dipsadidae snake *Philodryas patagoniensis*, also induce changes in leukocyte-endothelial interactions [39,40].

The recruitment of leukocytes to the injured tissues is a characteristic event of the inflammatory response, and the leukocyte-endothelium interaction is promoted by the expression of cell adhesion glycoproteins on the surface of leukocytes and endothelial cells. This event occurs mainly when integrins on leukocytes bind to adhesion molecules on the endothelium [23–25,41].

Increases in the expression of ICAM-2, PECAM-1, and Jam-A have been described in the literature after the injection of TNF-α and IL-1β, and the interaction between these adhesion molecules facilitates the migration of leukocytes to extravascular tissue [25,41]. Notably, the proinflammatory role of Jar might be attributed to its ability to process the tumor necrosis factor (TNF-α) precursor [7], and it is still capable of inducing the mRNA expression of proinflammatory cytokines such as TNF-α, IL-6, and IL-1β [42]. Additionally, once the leukocyte adheres to ICAM-1, this binding increases its avidity through interactions with integrins, facilitating the transmigration process on the leukocyte surface [43]. This observation might explain the increased expression of ICAM-1, CD11a and CD11b observed in the first hours evaluated after the injection of the different toxins (Figures 1, 2A, 3 and 4).

In general, leukocytes are recruited during an inflammatory insult, and this mobility is essential for the defense of the organism. Neutrophils are more frequently observed and are quickly recruited and activated [41], although other circulating leukocytes also participate in this process, depending on the inflammatory stimulus. Based on our results, both LFA-1 (CD11a/CD18) and Mac-1 (CD11b/CD18) are expressed after the injection of toxins, suggesting that both adhesion molecules participate in this leukocyte-endothelium interaction. However, the differences between these molecules should be highlighted. Both LFA-1 and MAC-1 are members of the β2 integrin family but are expressed in different cell types. MAC-1 is present on neutrophils and monocytes [25,42], identifying the two types of cell populations present in the peritoneal exudate from mice 4 h after the injection of the different toxins (Figure 4), whereas LFA-1 is expressed on all effector leukocytes (Figure 3).

Phillipson et al. [44] described that the adherence of neutrophils to the activated endothelium is mainly mediated by LFA-1, but intraluminal crawling, which follows adherence and spreading, depends on MAC-1. Sumagin et al. [45] observed an increase in CD11b expression in monocytes and a decrease in this expression in neutrophils in postcapillary venules of the cremaster muscle of mice without inflammatory stimulation. Some authors have also shown that the MAC-1 blockade prevents leukocyte recruitment [46,47]. The balance and regulation of this signaling via integrins contribute to leukocyte transmigration guided by endothelial cells.

PECAM-1 is the central adhesion molecule related to cellular transmigration. This protein is expressed at high levels in endothelial cells. It accumulates in intercellular contacts and is associated with the junctional regulation of endothelial complexes such as β-catenin, VE-cadherin, and adherent endothelial junctions, signaling remodeling of the cytoskeleton during the transmigration process [44,48]. We observed a significant increase in the expression of the PECAM-1 mRNA (Figure 2B) and protein (Figure 1E) in the cremaster muscle of mice injected with all toxins at the later time points compared with the ICAM-1 molecule that is responsible for firm adhesion, which was expressed in the first studied periods (Figures 1D and 2A).

These molecules have also been shown to play important roles in other studies examining *Bothrops* venoms. L-selectin, ICAM-1, and PECAM-1 molecules participate in the leukocyte recruitment induced by *Bothrops asper* venom, or BaP1, a P1 SVMP [31,35]. This inflammatory effect was related to the actions of inflammatory cytokines and leukotrienes released by the venom that act on endothelial cells and leukocytes [49].

The results obtained in the present study confirm the importance of metalloproteases in inducing the local inflammatory response observed in previous studies [49–51], particularly in microcirculatory changes [8,9]. Nevertheless, our results reveal the initial participation of the molecules ICAM-1, CD11a, and CD11b and the late expression of PECAM-1. This expression follows the pattern of cell adhesion and migration in vivo and is compatible with acute inflammation and consistent with results observed in other studies [9,38]. The participation of other adhesion molecules in the observed changes should not be excluded.

#### **4. Conclusions**

In the present study, we compared different SVMPs: Jar, a PIII with high hemorrhagic activity, Jar-C, which is devoid of hemorrhagic activity, and BnP1, a PI SVMP with low hemorrhagic activity. Despite the differences in the composition and hemorrhagic actions of these toxins, they similarly modulate the expression of adhesion molecules that promote leukocyte recruitment. In addition, their expression followed a sequential pattern, consistent with the recruitment and adhesion processes that were observed in previous in vivo studies performed using intravital microscopy. Therefore, we suggest a possible mechanism underlying the leukocyte-endothelium interaction after exposure to snake venom toxins. A similar pattern was observed even in other inflammatory pathologies. These data contribute to the understanding of local reactions observed after envenomation induced by *Bothrops* snakebites. The same approach can be used to evaluate if other toxins, such as PLA2 and LAAO present in other *Bothrops* venoms, contribute to the inflammatory process observed in this envenoming.
