Thymus-Brain Connections in T-Cell Acute Lymphoblastic Leukemia Open Access

T-cell acute lymphoblastic leukemia (T-ALL) is a neoplasm of lymphoblasts committed to the T-cell lineage. It is currently classified as a form of T-cell lymphoblastic leukemia/lymphoma (T-ALL/LBL). T-ALL involves peripheral blood and bone marrow infiltration. In contrast, T-cell lymphoblastic lymphoma (T-LBL) presents with primary involvement of the thymus, nodal, or extranodal sites, with no or minimal evidence of peripheral blood and bone marrow involvement [1]. Thus, T-LBL and T-ALL are distinguished based on a 20% bone marrow infiltration cutoff point. When bone marrow infiltration levels are below 20%, it characterizes T-LBL, whereas levels above this threshold are classified as T-ALL [2].

T-ALL accounts for about 15% of childhood and 25% of adult ALL cases. It is still considered a high-risk disease associated with a higher risk of induction failure, early relapse, and central nervous system (CNS) infiltration and relapse compared to the most common B-cell acute lymphoblastic leukemia. Among T-ALL subtypes, the early T-cell precursor lymphoblastic leukemia (ETP-ALL) accounts for approximately 10–13% of T-ALL cases in childhood and 5–10% of adult ALL cases, presenting a unique immunophenotypic and genomic profile. This profile seems to represent a higher risk disease subtype than non-ETP-ALL in adults [1], being associated with higher initial refractoriness to treatment, higher relapse risk, and lower overall survival [3]. However, other studies have shown no difference in overall survival between ETP-ALL and T-ALL patients [4‒6].

Although the most common site of relapse of T-ALL is the bone marrow, infiltration of the CNS also occurs, which is particularly important regarding the risk of leukemic relapse and disease prognosis. Disseminated T-ALL is typically isolated to the leptomeninges and rarely invades the brain parenchyma. Among the entry routes for ALL cells in the CNS, the blood-cerebrospinal fluid barrier has been proposed to serve as a primary entry point for the blasts [7]. Indeed, CNS involvement is measured by the detection of leukemic cells in the cerebrospinal fluid (CSF) after a lumbar puncture. However, some studies have shown that B-ALL cells migrate directly from the bone marrow to the CNS, bypassing the need to enter and exit the CNS vasculature [8, 9], suggesting this possibility for T-ALL cells.

Different strategies for therapies involve the search for biomarkers and the use of drugs that target CNS invasion and survival mechanisms employed by leukemia cells in the leptomeninges [10]. Some of these drugs target adhesion molecules such as integrins (e.g., VLA-4) and chemokine receptors (e.g., CXCR4) which may be involved in the preferential location of T-ALL cells in the CSF and are also crucial for the survival, migration, and egress of T-ALL cells and normal thymocytes from the thymus [10, 11].

The thymus-brain connection can be mediated by innervation, through the release of neurotransmitters, and by hormones and other soluble molecules produced by both organs. Notably, various receptors for neurotransmitters, neuropeptides, and hormones are expressed in developing thymocytes and T-ALL blasts [12, 13]. Their activation can affect cell survival, proliferation, and migration, which are key biological responses impacting intrathymic T-cell development (for a comprehensive overview of human T-cell development, we refer to a recent review by Bosticardo and Notarangelo [14]). Conversely, lymphoblasts produce cytokines and chemokines that could have local and systemic effects on CNS activity once infiltrated, for example, activating microglia and promoting or facilitating CNS infiltration. In this review, we discuss the molecules possibly involved in the thymus-brain crosstalk in T-ALL, including those commonly expressed by the two systems, and their possible roles in blast migration and infiltration of the CNS.

Neurotransmitters, comprising amino acids, monoamines, and peptides, are constitutively produced in the thymus by both autonomic innervation and thymic cells [15, 16]. Accordingly, thymic microenvironmental cells and developing thymocytes express related receptors [12], which, upon activation, can modulate thymic physiology and thymocyte differentiation.

Thymic innervation is characterized by the presence of autonomic nervous system fibers, providing sympathetic noradrenergic fibers within the parenchyma [15, 17, 18]. The primary source of these nerves is postganglionic neurons in the superior cervical and stellate ganglia [19]. Using pseudorabies virus (PRV) to trace neural connections, a study also identified the preganglionic sympathetic neurons responsible for innervating the thymus in rats. These neurons were found in the intermediolateral cell column of the T1–T7 spinal cord segments [20]. Interestingly, this study observed considerable overlap with CNS cell groups connected to the sympathetic innervation of non-immune organs, such as the spleen and bone marrow, suggesting that all sympathetically innervated organs have some neuronal populations in common.

Nerve fibers are observed in all thymic compartments, including the capsule/subcapsular region, cortex, cortico-medullary junction, and medulla, in close association with thymocytes and thymic epithelial cells (TECs), as well as with dendritic cells, macrophages, and B cells [18]. The widespread presence of nerve fibers strongly suggests the possible link between the nervous system and all stages of T-cell development, which depends on the interaction of developing thymocytes with other cells in the thymic microenvironment. However, the direct role of thymus innervation in T-lymphocyte differentiation is still in need of further confirmation. The thymus also seems to be innervated by the parasympathetic fibers from branches of the vagus nerve [21]. However, this is still a matter of debate [16]. Nevertheless, there is an intrathymic non-neuronal cholinergic system represented by the presence of acetylcholine (ACh), which is synthesized and released by TECs and thymocytes, and the expression of the corresponding receptors by both cell compartments [22‒24].

Besides normal developing lymphocytes, T-ALL primary cells and cell lines express neurotransmitters, neuropeptides, and cognate receptors, as summarized in Table 1. This expression may influence the survival and function of T-ALL blasts, as well as the adjacent microenvironment, affecting nearby cells including CNS and other immune cells. In terms of CNS infiltration, the production of neurotransmitters by T-ALL could change normal brain function. This could potentially lead to neurological symptoms or complications in patients with T-ALL. The expression and importance of neurotransmitters and their receptors in T-ALL blasts will be discussed in more detail in the following sections.

Epinephrine and norepinephrine (adrenaline and noradrenaline, respectively) are hormones and neurotransmitters released from the adrenal glands, whereas norepinephrine is released from the sympathetic nerve terminals. They act through adrenergic receptors (ARs) present on many cell types, including developing T lymphocytes [51].

Different studies have found that epinephrine and norepinephrine exert stimulatory effects on cancer development and progression, and β-adrenergic receptor (β-AR) blockers have anti-tumor effects, inhibiting melanoma, rectal, breast, and gastric cancer cell proliferation [52‒55]. The expression and role of β-adrenergic receptors (β-ARs) were observed in T-ALL primary cells and cell lines such as Jurkat, CEM-C1, and CEM-C7. Xu and collaborators [25] recently showed that newly diagnosed T-ALL patients express higher levels of β1, β2, and β3-ARs than the control group. Using Carvedilol, a third-generation non-selective β-AR blocker, they observed inhibition of T-ALL viability in vitro and in vivo by the induction of apoptosis and arrest of the cell cycle. Treatment with Carvedilol reduced β1, β2, and β3-ARs levels in CEM-C1 and CEM-C7 cells and inhibited their downstream signaling pathways, reducing PI3K, p-PI3K, AKT, p-AKT, mTOR, JAK2, STAT3, and p-STAT3 levels. In vivo, Carvedilol treatment notably reduced both the size and weight of xenograft tumors in mice. Additionally, it triggered apoptosis in the tumors while reducing the presence of CEM-C1 cells in murine peripheral blood and bone marrow.

Propranolol, another non-selective β-AR blocker, induced cytotoxicity in vitro, in Jurkat and MOLT-4 cells. Also, propranolol significantly decreased vascular endothelial growth factor (VEGF) production and MMP-2 (metalloprotease-2) activity when the cell lines were activated with PMA [56, 57]. VEGF is a co-mitogen for the vascular endothelial cells, promoting vascular permeability, regeneration, and angiogenesis. The importance of angiogenic factors in solid tumors is well described, but it also seems to be correlated with the treatment, relapse, and prognosis of hematolymphoid tumors such as T-ALL [58]. It was indeed shown in B-ALL in vivo, by transplanting leukemia samples into NOD/SCID mice. CNS ALL cells were characterized by high expression of VEGF transcripts when compared with corresponding bone marrow-derived samples. Importantly, a clear reduction in meningeal infiltration was seen in animals treated with bevacizumab (which specifically binds to human but not to murine VEGF) as compared with controls. No effect of leukemia reduction was observed in the bone marrow or spleen, indicating that ALL invasion of the CNS is mediated by human ALL-derived VEGF and can be reduced by VEGF inhibition [59].

MMPs are zinc-dependent proteases involved in the degradation of extracellular matrix components, cleavage of membrane receptors such as integrins, as well as chemokine and cytokine processing, which modulate cell migration and spread [60]. Yet, whether β-AR blockers modulate T-ALL migratory patterns and invasion needs further examination. In any case, these data show the potential role of β-AR activation or inhibition in T-ALL development.

As mentioned above, there is an intrathymic cholinergic system represented by ACh, which is synthesized and released by TECs and thymocytes that also express cholinergic receptors [22‒24]. Cholinergic receptors (AChRs) consist of two major subtypes: the metabotropic G-protein-coupled muscarinic receptors (mAChRs) and the ionotropic nicotinic receptors (nAChRs). Multiple nAChRs subunits have been described as being expressed by normal thymocytes and TECs (e.g., α3, α5, α7, β2, and β4 in thymocytes; and α2, α3, α5, α7, β4, and ε in TECs) [12]. The expression of mAChRs and nAChRs in thymocytes was also confirmed in binding studies using radiolabeled ligands [61].

The cholinergic machinery is also present in T-ALL blasts, and a more detailed overview of this subject has been reviewed in detail by Dobrovinskaya and colleagues [26]. ACh production by T-ALL cell lines (CEM, HSB-2, Jurkat, MOLT-3, and MOLT-4) is considerably higher when compared to B-ALL cell lines and unstimulated peripheral blood mononuclear leukocytes obtained from healthy donors [23, 27]. One of the main mechanisms responsible for appropriate cholinergic function is performed by the acetylcholinesterase (AChE) enzyme. Interestingly, a study has shown that blood and lymphocyte AChE activity was increased in the newly diagnosed T- and B-ALL samples and decreased in the remission induction group in relation to the controls, reinforcing the possible role of the cholinergic machinery in leukemia [62].

AChRs have been shown to play an important role in leukemic cell proliferation, differentiation, and apoptosis. The presence of the mRNA-encoding nAChRs α3, α5, α6, α9, and α10 subunits was reported in T-ALL cell lines as CEM and Jurkat [63, 64]. Jurkat, but not CEM, also expresses the α7 subunit [64]. Regarding mAChRs, Jurkat cells present high M3 expression, lower M4 and M5, and no M1 and M2 expression [65]. CEM cells also express functional mAChRs linked to NO synthesis by nNOS and/or iNOS [28]. Also, both mAChRs and nAChRs were described in T-ALL primary cells [66], suggesting that these cells can also respond to ACh or ACh agonists.

Since T-ALL blasts produce a large quantity of ACh, it is conceivable that it may also exert a significant paracrine effect on neighboring stromal cells in their local niche. Based on that, Dobrovinskaya and colleagues [26] suggested a hypothetical model of non-neuronal cholinergic system involvement in T leukemogenesis in the bone marrow. In this scenario, ACh produced by T-ALL blasts would induce an autocrine activation leading to proliferation and a paracrine effect on osteoblasts and endothelial cells that express nAChRs and mAChRs, playing an important role in the remodeling of microenvironmental niches. In this same line, one can argue that a similar phenomenon could occur in other niches or sites of blast invasion, such as the CNS. The concept of a molecule that influences both leukemic cells and the different niches where these cells are located may be of great interest for a better understanding of the pathogenesis of the disease. Furthermore, the role of ACh released by human T lymphocytes on endothelial cells was recently shown [67]. T-cell-derived ACh increased endothelial nitric oxide synthase activity, promoted vasorelaxation, reduced vascular endothelial activation, and promoted barrier integrity, suggesting a cholinergic immune regulation of vascular endothelial function in human inflammation. Thus, it is also conceivable that local ACh release by T-ALL cells impacts cell transmigration and access to different tissues.

Glutamate is the most abundant excitatory neurotransmitter in the CNS, involved in different aspects of the brain’s physiology and function. Glutamate acts via two classes of receptors: GluRs that are ligand-gated ion channels and called ionotropic glutamate receptors (iGluRs), and GluRs that are G-protein-coupled and called metabotropic glutamate receptors, which activate intracellular signal transduction pathways following binding of glutamate. GluRs are expressed by various immune cells, including dendritic cells, T cells, T-LBL and T-ALL blasts, and cell lines such as Jurkat and FRO [68].

Studies by Ganor et al. [29, 69] showed that Jurkat cells express high levels of iGluR3, and glutamate treatment increases the membrane expression of CD147 in these cells, which is an inducer of MMPs and could, consequently, modulate cell adhesion and migration. Employing an experimental model of leukemia engraftment by injecting Jurkat cells into chick embryos, they have shown that glutamate treatment enhanced Jurkat engraftment and pro-metastatic effect [29].

More recently, chimeric antigen receptor (CAR) T-cell therapy targeting human CD147 was used for treating T-ALL [70]. CD147-CAR T cells showed anti-tumor activity against T-ALL primary and Jurkat cells by cytotoxicity mediated by granzyme B and perforin and by releasing a high level of cytokines. CD147-CAR T cells also controlled the progression of human T-ALL in a mouse xenograft model, showing their promising potential as a target for T-ALL therapy. These data reinforce the potential role of glutamate in T-ALL since it modulates CD147 membrane expression in T-ALL cells in vivo.

Besides that, glutamate increases integrin-mediated T-cell adhesion to laminin and fibronectin and CXCR4-dependent chemotactic migration toward CXCL12 [29]. Those are also critical events for the migration and spread of T-ALL blasts. Whether glutamate increases T-ALL cell adhesion and migration and CNS infiltration still needs further investigation.

Dopamine (DA) is a catecholamine neurotransmitter extensively distributed in the CNS and involved in regulating many vital physiological processes [71, 72]. The interaction of DA with one of the five DA G-protein-coupled receptors (DR 1-5) activates a complex signaling network involved in the regulation of motor neurons, modulation of emotional states, cognitive function, and hormonal and sympathetic regulation [71, 73, 74].

DA can interact with cells from the immune system, regulating immune functions such as cell adhesion, cytokine secretion, and chemotaxis [32, 75, 76]. Mignini et al. [16] observed the presence of DA and correlated receptors in the thymus of rodents, mainly in the medulla and cortico-medullary junction. In addition, an increase in the number of thymic regulatory T cells (tTregs) was observed in conditions of prenatal catecholamine deficiency, suggesting the participation of DA and its receptors in tTreg development [77].

T-ALL cells also express DRs, and the possible role of DA in T-ALL development was suggested through in vitro experiments. Basu et al. [33] showed that Jurkat cells express DR1 and DR2, like normal T cells, but diverge in the low expression of DR3, DR4, and DR5. Specific DR1 and DR2 agonists inhibited the proliferation of normal T cells. Still, this effect was not observed in Jurkat cells due to the failure of DR1-mediated activation of cyclic AMP signaling and missense mutation of DR2 that affects the inhibition of phosphorylation of ZAP-70, an essential downstream protein transducing signal from the T-cell receptor. These results suggest the involvement of DA and alterations in DR1 and DR2 receptors in the uncontrolled proliferation of leukemic cells. In addition, polymorphisms in genes encoding enzymes involved in the production and degradation of DA and DR4 were associated with CSF biomarkers of CNS injury in ALL primary cells of children under treatment [34]. In this study, the distinction between ALL patients was not described, but 15% of patients had T-ALL.

Semaphorins (SEMAs) are axon-guiding glycoproteins with repulsive properties that can bind neuropilins (NRPs) and plexins, initially described in the nervous system. SEMAs are classified into eight classes: classes 1 and 2 are found in invertebrates, classes 3 to 7 comprise vertebrate SEMAs, and class V represents viral-encoded SEMAs. Vertebrate membrane-bound SEMAs (classes 4–7) bind directly to plexin subclasses A–D, whereas secreted SEMAs (class 3) bind NRPs (1 and 2). As NRPs have a short cytoplasmic tail, they bind to class A plexins as co-receptors able to activate signaling pathways. The expression of SEMAs, NRPs, and plexins is well described in the immune system. Some are found in the human thymus, expressed by TEC and thymocytes, mediating thymocyte migration and thymocyte-TEC interactions [78].

NRPs and SEMAs have been described in different pathologies, mainly neurodevelopmental disorders and cancer. The expression levels of NRPs and SEMAs are also related to the progression of solid tumors. Besides the expression of class 3 SEMAs A and F, NRP1 and NRP2, and plexin A in the human thymuses in TECs and developing thymocytes, their expression was also found in the thymus in T-ALL and T-LBL conditions. Interestingly, NRP2 and SEMA3F are highly expressed in T-LBL compared to T-ALL samples [35]. In addition, the SEMA3F protein levels in T-LBL biopsies were higher in intrathymic malignant cells when compared with the normal thymus or with T-LBL cells localized in other peripheral sites such as lymph nodes and non-lymphoid organs. Functionally, SEMA3F can inhibit T-ALL and LBL migration toward CXCL12 and sphingosine-1-phosphate (S1P), two important chemotactic factors related to blast localization and infiltration. Malignant T-cell precursors respond to SEMA3F similarly to normal thymocytes, meaning that T-ALL blasts expressing lower levels of NRP2 and SEMA3F would respond and migrate toward CXCL12 and S1P. In contrast, blasts expressing higher levels of NRP2 and SEMA3F would not respond to CXCL12 and S1P appropriately, being less able to migrate and infiltrate [35].

NRPs can also bind to different members of the VEGF family, playing an important role not only in vascular and neuronal development but also in the proliferation and migration of leukemic cells [36, 58, 79]. As discussed above, samples from patients with ALL have elevated levels of VEGF-A in the CSF compared to the serum. Importantly, elevated VEGF-A levels were observed in the CSF of patients with CNS infiltration compared with patients without CNS infiltration, suggesting the participation of VEGF in CNS metastasis [80]. Since NRPs are also receptors for VEGFs, they might participate in VEGF-mediated CNS invasion.

S1P is a phospholipid involved in different cell processes, including thymocyte survival, egress from the thymus, and peripheral T-lymphocyte migration and egress from peripheral lymphoid organs. Primary T-ALL and T-LBL cells and different cell lines respond to S1P through the activation of S1P1 [35, 38]. Interestingly, S1P activation can result in two different migratory responses in vitro: lower S1P concentrations induce chemotaxis, whereas higher concentrations induce the fugetaxis of cell lines such as Jurkat, CEM, MOLT-4, and HPB-ALL. The analysis of signaling pathways following low or high doses of S1P suggests that Rac1 activity is possibly involved in regulating these specific migratory responses [38]. Moreover, using a T-LBL model in zebrafish, a study has shown that the inhibition of S1P1 induces lymphoma cells to disaggregate and intravasate into the vasculature in vivo, thus implicating high S1P1 levels in the blockade of T-LBL dissemination [81].

Besides the role of S1P in T-ALL and T-LBL migration and thymocyte egress, it can also mediate neuronal signaling and function, suggesting that it could mediate the thymus-brain crosstalk. First, directly, by possibly inducing CNS infiltration of T-ALL blasts. Second, indirectly, by triggering glutamate secretion in the CNS and by potentiating depolarization-evoked glutamate secretion. Exogenously added S1P at a nanomolar concentration elicited glutamate secretion from hippocampal cells, and S1P at a picomolar level potentiated depolarization-evoked secretion in neurons [39]. These data suggest that variations in circulating or tissue levels of S1P could impact both thymic and CNS function, including T-ALL localization and infiltration (Fig. 1). The role of S1P/S1P1 in this context needs further investigation. Furthermore, S1P/S1P1 migratory responses of T-ALL primary cells (as well as T-LBL samples) can be controlled by SEMA3F/NRP2 [35], showing a more complex mechanism of regulation of S1P1 signaling. Since SEMAS are also present in both leukemia and CNS cells, one can imagine that they can mediate the crosstalk and impact organ invasion.

Chemokines and chemokine receptors are constitutively expressed in the immune and nervous systems, being essential for the thymus and brain development and homeostasis. These proteins play a role in the migration, differentiation, and proliferation of thymocytes, TECs, glial, and neuronal cells [82, 83], and protection at CNS barriers [84]. They are also involved in cancer cell survival, proliferation, and migration [85], as well as in T-ALL development and migration [86].

Different studies have shown that blasts from patients with T-ALL express altered levels of CXCR4 [87‒89]. Similarly, the expression of CXCR4 was observed in T-ALL cell lines such as Jurkat, CEM, and MOLT-4 [43, 89]. The CXCR4-CXCL12 axis has been implicated as a potential pathway that drives T-ALL invasion into the CNS. Using a T-ALL in vivo model, where alymphoid NGC mice were injected with Jurkat cells expressing CXCR4 or deficient for this receptor, it was shown that the absence of CXCR4 increased survival and minimal signs of CNS pathology [44].

Another study by Jost et al. [43] showed that Jurkat and CEM cells injected into mice invaded the CNS and were located in the meninges. However, treatment with AMD3100, a CXCR4 antagonist, reduced the meninge infiltration [43]. Together, these data support the participation of the CXCR4/CXCL12 axis in targeting T-ALL cells in the CNS.

Most primary T-ALL samples and cell lines also overexpress CCR7, which is a Notch1-induced adhesion regulator. It is estimated that Notch1 signaling pathway activation occurs in at least 80% of all T-ALL cases, and a study has shown that CCR7 was expressed in 80% of T-ALL lines and 73% of primary T-ALL samples analyzed [47]. In this study, it was shown that transplantation of human T-ALL cells into alymphoid Rag2^−/−Il2rg^−/− mice led to distinct survival patterns depending on the CCR7 expression, where hosts that received CCR7⁺ cells succumbed to the disease earlier than those receiving CCR7 cells. CCR7 expression was sufficient for CNS infiltration in response to CCL19, detected mainly in brain venules near infiltrating lymphocytes. Interestingly, CCR7 function seems specific for Notch1-induced T-cell malignancies since deletion of this receptor in two models of B-ALL failed to suppress CNS infiltration.

CXCR4 and CCR7 expression levels in patient samples with CNS involvement were significantly increased compared to patients without CNS involvement. In addition, in NSG mice transplanted with Jurkat cells, leukemic cells infiltrated the bone marrow and the CNS. CCR7 blockade significantly reduced Jurkat infiltration into the CNS but not into the bone marrow, reinforcing the role of CCR7 in T-ALL infiltration into the CNS [45].

Some endogenous components produced in the thymus, such as thymulin, thymosin, and thymopoietin hormones, play a role in thymocyte differentiation [90, 91]. Thymic hormones are also found in the CNS since thymulin can be produced by astrocytes and glial cells in the brain of rodents, and thymosin can be produced in different regions of the CNS, mainly in the hippocampus [48, 49]. Studies by Safieh-Garabedian et al. [92, 93] showed that pre-treatment with thymulin inhibited the induction of pro-inflammatory cytokines and reversed local inflammation in vivo in rat models using an intracerebroventricular (i.c.v.) injection of LPS or endotoxin. Moreover, early treatment with thymosin B4 was anti-inflammatory and reduced cortical lesions and loss of hippocampal cells in traumatic brain injury, suggesting a neuroprotective role of thymulin and thymosin in the CNS [94].

Significantly lower plasma levels of active thymulin have been observed in samples from patients with initial T-ALL or during relapse when compared to healthy controls, apparently due to a zinc deficiency [50]. Interestingly, active thymulin in physiologic doses (corresponding to the amount in the blood of young healthy individuals) reduced the proliferation of primary and T-ALL cell lines. Although there is evidence for the role of thymic hormones, such as thymulin, in T-ALL development and the CNS separately, further studies are necessary to explore the role of these hormones in the context of the thymus-brain connections in T-ALL.

Different aspects of the possible connections between the thymus (and T lymphoblasts) and the brain are explored herein, focusing on soluble molecules and their receptors. Among them, neurotransmitters, thymic hormones, chemokines, and other molecules shared between the thymus and the brain may participate in the crosstalk between these two organs and, consequently, in T-ALL development and invasion in the CNS.

Understanding the specific mechanisms and routes by which T-ALL cells infiltrate the CNS remains an important question and would be of great value to leukemia treatment. For example, understanding whether T-ALL cells reach the CNS by first invading the bone marrow or the CSF, and identifying the molecules involved in different migration patterns could help to identify biomarkers able to predict or detect CNS infiltration in T-ALL at an early stage, thereby helping to monitor disease progression. Understanding how leukemia cells interact with neural cells, immune cells, and other components of the CNS may also provide insights into disease progression and potential therapeutic targets. Therefore, a deeper understanding of the role of different mediators of thymus-brain connections could shed light on the mechanisms involved in the pathogenesis of the disease including lymphoblast dissemination and CNS infiltration, given their clinical importance.

The authors have no conflicts of interest to declare.

This work was supported by Fiocruz, Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) – Rio de Janeiro, and the Mercosur Fund for Structural Convergence (FOCEM). It was developed in the Brazilian National Institute of Science and Technology frameworks on neuroimmunomodulation (CNPq) and the Rio de Janeiro Research Network on Neuroinflammation (FAPERJ).

D.A.M.C. and V.C.A. contributed to the conception, organization, and writing of the manuscript. E.P.B. contributed to the conception and writing of the manuscript and organized references.