**4. Discussion**

Chagas heart disease (CHD) is characterized by dilated cardiomyopathy that affects the atria, ventricles, conduction system, and autonomic nervous system (ANS) [16]. In the pathogenesis of CHD, various mechanisms are implicated such as: the persistence of the parasite; denervation; both microvascular and endothelial dysfunction; a persistent and exacerbated immune response (via the imbalance between pro-inflammatory and anti-inflammatory cytokines) or even autoimmunity [17]; as well as the induction of a permanent oxidative stress caused by both reactive nitrogen species (RNS) and reactive oxygen species (ROS), which in turn directly affects the structure and function of the respiratory chain of cardiac tissue [18]. All of these factors combine to prevent the correct remodeling of heart tissue (functional tissue is replaced by non-functional fibrotic tissue), ultimately leading to heart failure, for which the only treatment option is transplantation [3].

Hence, it is urgent that new therapeutic alternatives be found to eliminate the parasite, control inflammation, and ameliorate cardiac damage, as well as minimizing the harmful side effects in human treatment posed by the current therapeutic options. In this sense, phytoderivatives obtained from aromatic plants represent a highly diverse platform for the discovery of bioactive compounds. In our group, promising in vitro trypanocidal activity was reported for the major terpenes of *L. alba* EOs [8], limonene and caryophyllene oxide; with limonene being the most effective compound against all parasite forms (IC50 of 9.0, 28.7, and 41.8 μg/mL for amastigotes, trypomastigotes, and epimastigotes, respectively) [8]. This activity was also exhibited in in vivo models applied to Wistar rats with chronic *T. cruzi* infection, which were treated with synergistic fractions of these EOs enriched in limonene and caryophyllene oxide [10]. Recently, in vitro immunomodulatory properties on macrophages infected with *T. cruzi* were also ascribed to these phytoderivatives [9].

In order to clarify the potential role of certain relevant anti- and pro-inflammatory cytokines and oxidative markers in the trypanocidal and cardioprotective activities observed for *L. alba* EO, the present study assessed the behavior of these markers in an in vivo model of *T. cruzi* chronic infection induced in Wistar rats. In infected and untreated animals (positive control), CHD was verified by the appearance of a statistically significant enlargement of cardiac silhouette parameters (MD and ML) at 67 d.a.i (*p* < 0.05, compared to the uninfected group). Histological findings observed in the animals belonging to the positive control group confirmed the success of CCC induction; findings such as: damage to cardiac tissue, mainly in the form of loose, elongated, and sinusoidal fibers in the myocardium; the presence of multifocal and diffuse mixed inflammatory infiltrates (predominantly of the lymphohistiocytic type); and foci of lymphohistiocytic infiltrate in cardiac plexus neurons [3,19]. At that time, continuous once-daily oral treatments with the three studied schemes (LIMOX, LIMOXBNZ, and BNZ), were administrated for 31 days.

The results showed that the treatment comprised of a mixture of limonene-enriched fraction of *L. alba* EOs and caryophyllene oxide (LIMOX) demonstrated the best performance in restoring the normal shape of the heart compared to the other therapies trialed (LIMOXBNZ and BNZ). As such, the heart dimensions of animals belonging to the LIMOX group exhibited the most reduced ML and MD measurements among infected animals, even to the point of regression to values similar to those of the uninfected control (negative). Not unexpectedly, positive control group rats (infected and untreated) presented the most notable bulging and dilated shape, followed closely those of the BNZ and LIMOXBNZ groups.

Histologically, animals treated with LIMOX evidenced very few inflammatory foci and minimal fibrosis in the heart tissue, with similar characteristics to those of the negative control, in the whole-slide scanning analysis. These results could be attributable to the trypanocidal and cardioprotective effects previously ascribed to limonene and caryophyllene oxide [8,10]. In contrast, somewhat more inflammatory foci were discovered in the tissues of the animals treated with LIMOXBNZ, in whom the signs of fibrosis were also more evident. These adverse effects could be due to the subtherapeutical BNZ doses present in the mixture, which may promote a pro-inflammatory response [10,15]. In a similar manner, animals treated with the conventional intervention (BNZ) exhibited multiple inflammatory foci in the heart, with greater cell variety (particularly lymphocytes, histiocytes, plasma cells, and monocytes), a greater area of fibrous tissue, larger cardiomyocytes, and necrosis.

In order to assess, in a histological context, the cardiac immune response modulated by chronic *T. cruzi* infection and by the therapies trialed, the present study applied an immunohistochemical technique targeting a variety of antigens relevant to CCC pathogenesis using Qupath software; an open-source whole-slide image analysis tool [13]. The results showed

that the highest percentages of immunoreactivity for TNF-α and iNOS were present in the infected and untreated animals of the positive control group; as were significantly increased levels of IFN-G. These findings were accompanied by the impairment of the anti-inflammatory IL-4. In this regard, IFN-G regulates over a thousand genes by activating Janus tyrosine kinase (JAK) and phosphorylation of the transducer, and serving as an activator of the transcription 1 (STAT-1) pathway. This latter induces the transcription of TNF-α, interferon-inducible factor 1 (IRF1), and iNOS, among other inflammatory cytokines and chemokines [20]. Likewise, in the context of Chagas disease, IFN-γ acts synergistically with TNF-α through the activation of the nuclear transcription factor NF-kB for the positive regulation of iNOS expression; producing high levels of nitric oxide and RNS [21]. This phenomenon represents the activation of general trypanocidal mechanisms, such as the production of reactive species, through the induction of mitochondrial ROS and NADP oxidases [20,22]. These reactive species promote the production of the peroxynitrite anion, a strong oxidant mechanism that arises as a strategy for the elimination of the parasite; causing morphological disruption, severe alterations in its metabolism, calcium homeostasis, and trypanothione depletion [23,24].

Coherently, in this study, the most trypanocidal therapies LIMOX and BNZ (with 83% of parasitological cure in cardiac tissue assessed by qPCR), were accompanied by the highest IFN-G immunoreactivity. In this regard, it was hypothesized that in *T. cruzi* infection, the elevated levels of this cytokine could be a double-edged sword; since despite its recognized role in parasite tissue clearance and as an antifibrotic agent [25], an excess of IFN-G production could cause serious damage to cardiac tissue [25]. Nevertheless, an apparent regulatory mechanism was observed in animals subjected to LIMOX therapy (though not in the case of BNZ), represented by an increase in IL-4 production accompanied by restoration of IL-10 levels and lower percentages of TNF-α (compared to the other experimental groups, *p* < 0.05) and iNOS (whose levels did not exhibit statistically significant differences from the negative control, *p* > 0.05).

These differences in the tissue profile of pro- and anti-inflammatory mediators could explain the significant macroscopic and microscopic differences observed in the cardiac architecture between the LIMOX and BNZ groups; in which only animals treated with the terpene mixture presented regression in diameter measurements of the heart silhouette and significant reduction in inflammatory infiltrates or foci with fibrosis. In this context, it is known that TNF-α can induce collagen synthesis and fibrosis, thus contributing to the loss of cardiomyocyte contractility and its replacement by fibrotic tissue [25]. In addition, the same substance can be involved in the development of heart failure through apoptosis and induction of iNOS, with the consequent production of nitric oxide, which exerts a very strong inotropic effect [25]. These findings have been consistently documented in patients who expired from CCC [26]. Likewise, the expression of IL-10 and IL-4 has been linked to the improvement of cardiac function, as determined by the values of the left ventricular ejection fraction and the ventricular diastolic diameter [26]; factors which constitute a possible immunomodulatory tolerance mechanism which could potentially prevent cardiac damage. In this respect, IL-10 is considered a very important regulatory cytokine, and its production is associated with a better prognosis in chronic CHD, suggesting a protective effect for the Th1 response [26–29].

It is worth mentioning that these compensatory mechanisms represented by the stimulation of IL-4 and IL-10 secretion were not observed in the combined therapy of LIMOX and BNZ (LIMOXBNZ), in which both IFN-γ and TNF-α were slightly increased without a significant anti-inflammatory response. These findings suggest that BNZ, even in subtherapeutical doses, causes an antagonist effect on the Th1 response triggered by LIMOX therapy. These results align with those of Quintero et al. [9] who reported that therapies composed of BNZ, alone or in combination with *L. alba* fractions (rich in limonene and citral/caryophyllene oxide), impaired IL-4 secretion by *T. cruzi*-infected macrophages.

In the same work, Quintero et al. [9] showed that the synergistic combination of *L. alba* fractions rich in limonene and citral/caryophyllene oxide produced a significant reduction of the pro-inflammatory cytokines (IFN-γ, IL-2, and TNF-α), with a concomitant increase of the anti-inflammatory cytokines (IL-4 and IL-10), in the extracellular media of the infected macrophages [9]. Although both studies (the aforementioned as well as the current work) show evidence of an immunomodulatory effect (reduction of all proinflammatory cytokines with significant elevation of IL-4) by fractions derived from *L. alba* EOs enriched in limonene and caryophyllene oxide, a significant disagreement exists with respect to the behavior of IFN-γ [9]. In this context, the elevated IFN-γ levels observed only in the present in vivo study could be explained by additional compensatory cellular mechanisms converging in the global response of the innate immune system, which can be uniquely perceived in a context of cardiac tissue [25].

Regarding the toxicity of the therapies trialed, slight hepatomegaly accompanied by mild to moderate microscopic signs of periportal inflammation were reported in a percentage (16.6%) of animals belonging to the groups treated with both studied phytotherapies (LIMOX and LIMOXBNZ). These morphological features were correlated with elevated levels of AST transaminase (*p* < 0.05) in these same rats. On the other hand, kidney function assessed by biochemical, morphological, and histopathological analysis did not exhibit any significant alterations among control and experimental groups. Nevertheless, a statistically significant elevation of BUN levels was found in infected rats treated with LIMOX when compared to the reference treatment (BNZ) (*p* ≤ 0.005). In addition, 16% of the animals treated with BNZ presented a macroscopic alteration in the color of the kidney (pale brown).

With reference to an effect on spleen architecture, differential responses were observed among therapies. Specifically, splenomegaly was present in the positive control animals and in those treated with both LIMOX and LIMOXBNZ, while BNZ caused a reduction in the size of this organ. This hyposplenism could be correlated with the deleterious effect on adequate balancing of the defense response, as observed herein, in animals treated with the reference therapy [30].

In hemogram analysis, a significant thrombocytosis (*p* < 0.005) was reported for both the LIMOX and LIMOXBNZ therapies. Interestingly, platelet counts in peripheral blood have been inversely associated with disease severity in patients with CCC [31]. In this regard, platelets play an important role in immune response, including protective functions via the release of chemokines that attract and activate leukocytes and, at the same time, platelet surface molecules such as P-selectin and GPIIB/IIIa (Glycoprotein IIb/IIIa) [31]. Likewise, atypical lymphocytes were observed in the peripheral blood smear analysis of all infected animals (experimental groups and positive control) with the highest percentage of these reactive cells in BNZ-treated rats (10.8% for BNZ vs. 3.8% for positive control and 2% for LIMOXBNZ and LIMOX). These cells are produced after a strong process of antigenic stimulation and stress, being classified as a nonspecific response to stimulus or as precursors of memory T and B cells [31]. However, a low percentage of these cells can normally be found in peripheral blood (2–6%) [32], thus the abnormally elevated percentages observed in BNZ therapy could be a reflex to its toxicity.

#### **5. Conclusions**

This research contributes to clarifying, in a chronic CHD model in Wistar rats, immunomodulation as a possible trypanocidal and cardioprotective mechanism of LIMOX, a therapy based on a synergistic mixture composed of caryophyllene oxide and a limoneneenriched fraction derived from *L. alba* EO. This therapy showed clear benefits in controlling parasite load, apparently through a mechanism related to the enhancement of the nonspecific immune response mediated by high levels of IFN-γ. Remarkably, LIMOX also demonstrated high performance in controlling the progression of cardiac involvement in vivo, and even reversing the progression of dilated cardiomyopathy to levels similar to those found in uninfected animals. Correspondingly, a significant reduction in the severity of inflammatory foci and tissue damage was also confirmed by histopathological analysis, as well as greater tissue remodeling function. The cardioprotective effect observed via

LIMOX treatment was correlated with a protective mechanism derived from an increase in levels of the anti-inflammatory interleukin, IL-4, with a concomitant decrease of TNF-α and a reestablishment of IL-10. Thus, LIMOX becomes an interesting compound for the development of a holistic alternative therapy for the treatment of the chronic phase of Chagas disease.

**Author Contributions:** Conceptualization, D.X.E.-M. and L.T.G.S.; methodology, D.X.E.-M., L.T.G.S., C.I.G.R., E.E.S., C.A.V.-L., J.J.Q.R. and J.C.M.H.; formal analysis, D.X.E.-M. and E.E.S.; investigation, D.X.E.-M.; resources, L.T.G.S., C.A.V.-L., C.I.G.R. and E.E.S.; writing—original draft preparation, D.X.E.-M.; writing—review and editing, L.T.G.S.; visualization, D.X.E.-M.; supervision, L.T.G.S., E.E.S. and C.I.G.R.; project administration, D.X.E.-M.; funding acquisition, L.T.G.S. and E.E.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Ministry of National Education, the Ministry of Industry, Commerce, and Tourism, and ICETEX, call for scientific ecosystem—Scientific Colombia. Francisco José de Caldas Fund, Contract RC-FP44842-212-2018; and also supported by the Vicerrectoría de Investigaciones—Universidad de Santander UDES, under grant 001-18.

**Institutional Review Board Statement:** The study was approved by the the Universidad Industrial de Santander and Universidad de Santander (Agreement Number 010-VII, 15 and 16 May 2017 approved).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data is contained within the article.

**Acknowledgments:** The authors would like to express their appreciation to Camilo Durán for his support during the collection and characterization of EOs samples; to Jorge Luis Fernández Alonso for the specimen identifications; to Martha Lucía Díaz Galvis for her assistance in the *Trypanosoma cruzi* differentiation technique; to Hector Martínez for the slide scanning; and finally, we would like to kindly thank Juan Pablo Mejía Cupajita for his support with the standardization of immunohistochemical and hematoxylin-eosin techniques.

**Conflicts of Interest:** The authors declare no conflict of interest.
