Dysregulation of Toll-Like Receptor Signaling-Associated Gene Expression in X-Linked Agammaglobulinemia: Implications for Correlations Genotype-Phenotype and Disease Expression

X-linked agammaglobulinemia (XLA) results from pathogenic genetic variants in Bruton’s tyrosine kinase (BTK), a key regulator in B-cell development [1]. This primary antibody immune deficiency is characterized by the failure to produce mature B-lymphocytes and unsuccessful Ig heavy chain rearrangement. Genetic variants in all BTK domains can cause XLA, with missense mutations being most common, and numerous variants of uncertain significance and polymorphisms have been identified. Most mutations result in enzyme truncation, with CpG dinucleotide mutations being the most common events in XLA [2]. Two comprehensive cohort studies of XLA patients have revealed a wide spectrum of BTK mutation types, including missense, nonsense, indels, splicing, and deletions [3, 4]. However, no significant genotype-phenotype correlations were observed between patients’ clinical manifestations and mutation types, either at diagnosis or during follow-up, underscoring the necessity for further research to fully elucidate this complex relationship [3, 4].

Toll-like receptors (TLRs) are an evolutionarily conserved receptor family involved in innate immunity by sensing structurally conserved molecules from microorganisms. TLRs are type I integral membrane glycoproteins with homology to interleukin (IL)-1R in the cytoplasmic region, an N-terminal extracellular domain with leucine-rich repeats, and a transmembrane domain activating a cytosolic signaling domain called TIR domain-containing adapter protein (TIRAP) [5]. This allows recognition of pathogen-associated molecular patterns, such as lipids, lipoproteins, and nucleic acids. TLRs are expressed in various cell types, including dendritic cells, phagocytes, B-lymphocytes, and endothelial cells [6, 7]. The mammalian TLR family has 13 members, with 10 identified in humans and 12 in mice [8]. Activation of pattern recognition receptors (PRRs) triggers intracellular pathways, inducing inflammatory cytokines, chemokines, and type I interferons, promoting the recruitment of leukocytes and other antigen-presenting cells [9].

BTK interacts with TIR domains of specific TLRs (4, 6, 8, and 9) and associates with signaling complex components like MYD88, TIRAP, and IRAK1 to activate nuclear factor kappa B (NF-kappaB), generating inflammation through proinflammatory cytokine transcription [10‒13]. In XLA patients, BTK impairment affects lipopolysaccharide (LPS)-induced cytokine production (TNF, IL-1 beta, IL-6, and IL-10) by TLRs [14‒17]. XLA patients are susceptible to encapsulated bacterial infections but typically not viral infections, except in cases such as echovirus-induced meningoencephalitis, which is often associated with a dermatomyositis-like syndrome and can be life-threatening [18]. Some patients develop autoimmune and/or inflammatory conditions like arthritis and inflammatory bowel disease (IBD) [19]. Low TLR activity may contribute to infection susceptibility, while hyperactivity is associated with chronic inflammatory diseases, including cardiovascular inflammation and arthritis [12, 20]. BTK inhibitors have been linked to rheumatoid arthritis prophylaxis, as BTK activates downstream of B-cell receptor, and Fc receptors in immunoglobulins [21]. However, BTK inhibitors do not fully mimic the XLA phenotype [22].

XLA-associated inflammatory conditions have been reported [19, 23, 24], with IBD/enteritis diagnosed in up to 10% of patients [23]. The risk of these complications is low, possibly due to immunoglobulin treatment’s immunomodulation. A recent Chinese study of 98 XLA patients found 45.9% had arthritis, correlating with a higher incidence of BTK missense mutations, suggesting a genotype-phenotype relationship [25]. Our pilot molecular study aimed to evaluate TLR signaling pathway changes, shedding light on the impact of different BTK mutations on XLA clinical manifestations, diversity, and severity.

Eight XLA patients were diagnosed based on the International Union of Immunology Societies criteria and the World Health Organization Scientific Group for primary immunodeficiency diseases [26, 27]: CD19+ B-lymphocytes below 2%, reduced or absent immunoglobulin subtypes in serum, and recurrent bacterial infection history. All patients received intravenous human immunoglobulin therapy. Demographic, laboratorial, and clinical characteristics are presented in Tables 1 and 2. Healthy male age-matched controls were used in all experiments. Total blood count and inflammatory markers (erythrocyte sedimentation rate, ferritin, and C-reactive protein) were tested for controls and patients at the Division of Clinical Pathology of the “Hospital de Clínicas da Unicamp.”

All reverse transcription-quantitative polymerase chain reaction (RT-qPCR) reagents and kits were acquired from Thermo Fisher Scientific Inc., Waltham, MA 02451, USA, and used according to manufacturers’ standards. Ten TLR signaling-related target genes were quantified by RT-qPCR in XLA patients with missense (n = 3) and nonsense (n = 5) BTK mutations, compared to healthy controls (n = 17). Table 3 lists selected genes and their main functions. Quantifications were performed using a StepOnePlus™ Real-Time PCR System.

Peripheral blood samples (6 mL) were collected in EDTA-containing tubes, centrifuged, and buffy coats were added to 1,200 μL RNAlater™ Stabilization Solution. PBMC total RNA was extracted using RiboPure™ Blood RNA Purification Kit and reverse transcribed with High-Capacity complementary DNA (cDNA) Reverse Transcription Kit with RNase Inhibitor. RNA quantification and purity were assessed by fluorometry (Qubit^® 2.0 Fluorometer) and spectrophotometry (NanoDrop 2000). RNA integrity numbers were determined by Bioanalyzer 2100 Expert (B.02.08. SI648) (Agilent Technologies, Santa Clara, CA, USA). Qualified RNA samples were converted into cDNA (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000540082). Absolute quantifications confirmed 10 ng cDNA was adequate for relative quantifications (online suppl. Table 2).

Relative quantification reactions were run in triplicates using TaqMan™ Universal Master Mix II, with UNG, respective primers, and purified deionized H₂O with 10 ng cDNA samples. Data were generated and analyzed by StepOne Software version 2.3. Quantifications with >0.3 CT value differences were excluded. GraphPad Prism 5 software (GraphPad Prism version 5.04 for Windows, GraphPad Software, San Diego California USA, (http://www.graphpad.com) was used for statistical analysis, with p values <0.05 considered significant. Wilcoxon matched-pairs signed rank test allowed comparison of relative quantification data to controls, calibrated to 1.0. For patients XLA04 and XLA05, five pairs were created; for others, 12 controls were considered. Spearman nonparametric test correlated all relative quantification data, including patients’ ages.

The BTK, IRAK2, PIK3R4, REL, TFRC, and UBE2N targets were predominantly hypo-expressed, whereas RIPK2, TLR3, and TLR10 did not show a specific predominance, with hypo- or hyper-expression. TLR6 was the only gene that was predominantly hyper-expressed. Patients XLA01-03, who harbor a BTK missense mutation, had a marked downregulation for IRAK2, PIK3R4, REL, and TFRC, accompanied by a marked hyper-expression of TLR3 and/or TLR6 (Fig. 1). With the exception of REL, XLA09 followed a similar pattern. In contrast to the nonsense group (the other 5 patients in our cohort), the gene expression results were quite similar for the patients in the missense group, providing initial evidence of a gene expression signature.

BTK mutation consequences analyzed in the XLA scenario revealed a complexity through numerous signaling pathways affecting the expression of hundreds of genes, proteins, and molecular regulators [29]. BTK functions in the immune system analyzed as a whole – which may differ in specific cell lines and subgroups [30‒32] – have not been completely elucidated notwithstanding its well-known role as a key regulator of B-cell development [33]. Accordingly, correlations between the XLA clinical manifestation diversity/severity and the genetic defect are inconclusive [3, 34]. Mutations in BTK are very miscellaneous with no single mutation accounting for more than 3% of patients. Missense mutations seem to impact more severely on XLA [23, 34]. Autoimmunity and/or chronic inflammation in XLA may be directly related to TLR signaling malfunction, leading to complications that are consistent with a diagnosis of arthritis, IBD, or other inflammatory conditions [19, 23]. Large studies described the high morbidity and mortality of XLA patients despite progress in diagnosis and treatment [4]. Regional differences were shown in the frequency of IBD and enteroviral meningoencephalitis complications that were found to be more prevalent in South American patients [24].

As anticipated, BTK downregulation was observed in all patients, except for XLA03 (Fig. 1, 2), who exhibited significant BTK hyper-expression (p = 0.001) despite known protein-level downregulation [28]. The mutation type may influence BTK expression, with nonsense mutations leading to RNA decay and BTK hypo-expression [28]. However, XLA03 (missense mutation) represents the first instance of BTK hyper-expression, possibly due to a compensatory mutational mechanism. Although inefficient since the protein remains defective, further studies on this patient could elucidate BTK expression regulators, as the molecular mechanisms behind this upregulation are unclear. It is crucial to note that the youngest patients, XLA04 and XLA05, were compared to only five control individuals instead of 12, potentially affecting the significance of our results. Consequently, we adjusted the threshold of significance (0.1 > p > 0.05) for some data (Fig. 1, 2).

TLR signaling-associated gene expression variability in XLA patients may influence clinical manifestations (Fig. 3). Despite individual differences and epigenetic factors, we assume that for each patient, dysregulation of TLR signaling gene expression results from mutations causing absent or defective BTK. BTK is involved in sensing microbes through multiple TLRs, including TLR2, TLR3, TLR4, TLR7/8, and TLR9 in human and mouse macrophages and dendritic cells [33]. However, it has been shown that the absence of BTK in XLA neutrophils does not impair functional responses to TLR signals (LPS for TLR4; imidazoquinoline compound for TLR7/8) [30]. Lee et al. [35] demonstrated that BTK directly phosphorylates TLR3, playing a critical role in initiating its signaling. Without BTK, TLR3-induced PI3K, AKT, and MAPK signaling, as well as activation of NF-kappaB, interferon regulatory factor 3, and AP-1 transcription factors, are defective, compromising the formation of the downstream TICAM1-RIPK1-TBK1 complex. Additionally, BTK-deficient macrophages fail to secrete inflammatory cytokines and interferon beta 1 (IFNB1) upon TLR3 stimulation [35].

XLA patients typically respond to various viral infections; however, their responses to enterovirus genus (e.g., echovirus), coxsackie, and live polio vaccine virus are compromised, potentially causing progressive or chronic enteroviral meningoencephalitis, arthritis, or enteritis [36, 37]. Recently, Hedin et al. [38] described a case of a 47-year-old man with XLA who developed severe tick-borne encephalitis (TBE) and was treated with plasma containing TBE virus (TBEV) IgG antibodies as rescue therapy. The patient carries a BTK splice-site mutation (c.1349+4A>G), which was predicted to cause skipping of exon 14 and a frameshift with a premature stop codon (p.V450Gfs*13). Despite the absence of B cells, the patient mounted a robust TBEV-specific CD8+ T cell response [38]. This finding suggests that XLA patients can clear acute viral infections and may benefit from future TBEV vaccination, highlighting the importance of cell-mediated immunity in controlling viral infections in patients with humoral immunodeficiencies [38].

BTK appears to positively regulate TLR3 signaling and antiviral responses through direct phosphorylation of TLR3, while other proteins like UBAC1 and TRIM8 provide negative regulation of the TLR3 pathway. The BTK-TLR3 interaction may have implications for antiviral immunity and potentially vaccine responses [33]. TLR3 hypo-expression may be linked to higher susceptibility to viral infections, as observed in patient XLA07. Conversely, TLR3 hyper-expression, found in patients XLA01 and XLA02, may contribute to their susceptibility to chronic inflammatory conditions. Our data on TLR3 expression dysregulation supports Barmetller et al.’s [23] hypothesis that XLA patients with missense BTK mutations exhibit more clinical phenotypes than those with nonsense mutations.

BTK deficiency and the resulting impairment of TLR3-induced PI3K and AKT activation likely affect the mechanistic target of rapamycin (mTOR) signaling pathway, a major downstream target of PI3K/AKT [39]. mTOR exists in two distinct multiprotein complexes, mTORC1 and mTORC2, which regulate protein and lipid syntheses, autophagy, and cell survival [39]. In XLA, impaired PI3K and AKT activation may reduce mTORC1 activity, as AKT phosphorylates and inhibits TSC1/2, a negative regulator of mTORC1 [39]. This could decrease phosphorylation of mTORC1 targets, RPS6KB1 (S6K1) and 4E-BPs, involved in protein synthesis and cell growth [39]. Additionally, mTORC2, regulated by PI3K, phosphorylates and activates AKT [39]. Thus, impaired PI3K activation due to BTK deficiency may affect mTORC2 activity, reducing AKT phosphorylation and potentially impacting cell survival and metabolism. This scenario may contribute to defective inflammatory cytokine and IFNB1 secretion in BTK-deficient macrophages upon TLR3 stimulation [35]. Further studies investigating the impact of BTK deficiency on mTORC1 and mTORC2 activity and their downstream targets in TLR3 signaling would provide more definitive insights into mTOR’s role in this cascade.

TLR6 showed the highest upregulation in XLA patients, with significant downregulation in XLA06, characterized by pneumonias, bacterial infections, and recurrent sinusitis. XLA09 had the highest TLR6 mRNA level (p = 0.0005) and shared some manifestations with XLA06, but experienced worsened symptoms due to a sigmoid inflammatory process (Table 1). TLR6 overregulation appears to promote inflammation, which becomes chronic over time. Conversely, TLR10 hyper-expression in XLA patients might indicate anti-inflammatory signaling. Blocking TLR10 with antagonistic antibodies enhances proinflammatory cytokine production, including IL-1 beta, specifically after exposure to TLR2 ligands [40, 41]. This modulatory property seems to follow TLR6 expression in our patients. Oosting et al. [40] emphasized that inhibitory effects of TLR10 are likely mediated through interaction with TLR2 and possibly TLR1 and TLR6. TLR6 and TLR10 expressions showed the highest correlation in our patients (p = 0.0004), with significant overregulation in XLA09 (p = 0.0093). TLR10 hyper-expression was observed in patients with chronic inflammatory conditions, while hypo-expression was seen in XLA06, prone to bacterial infections. Differently, patient XLA04 also had TLR10 downregulation without TLR6 upregulation. As the youngest patient in our cohort, his TLR10 quantification was in accordance with the positive correlation with patient’s age (p = 0.0218). His clinical manifestations were mild, mainly characterized by urinary tract infections.

The role of IRAK2 in inflammatory signaling pathways has been extensively studied, revealing important differences between humans and mice. In humans, IRAK2 is a critical positive regulator of TLR/IL-1R signaling, essential for early NF-kappaB and MAPK activation, cytokine production, and posttranscriptional regulation of inflammatory mediators like TNF [42]. However, in mice, IRAK2 appears to play a more nuanced role, with studies in Irak2^−/− cells showing intact early TLR responses but impaired late NF-kappaB activation and cytokine production [43, 44]. This discrepancy likely stems from the existence of inhibitory IRAK2 splice variants in mice, which are absent in humans [45].

Wang et al. [46] identified a functional IRAK2 variant in humans associated with increased colorectal cancer risk, potentially by impairing IRAK2-mediated TRAF6 ubiquitination and subsequent IL-6 transcription. Notably, patient XLA03 was diagnosed with hyperplastic rectum polyps. Considering IRAK2’s crucial role in propagating inflammatory signaling in human cells, its reduced expression may contribute to an aberration in this signaling cascade in patients with XLA [42]. However, the overall impact on the immune dysregulation and clinical manifestations in XLA likely depends on the intricate interplay between various inflammatory and anti-inflammatory pathways regulated by IRAK2 and its interacting partners. Moreover, a significant positive correlation was found between IRAK2 and PIK3R4 expressions (p = 0.0218), which suggests a potential crosstalk between IRAK2-mediated inflammatory pathways and PI3K signaling, which warrants further investigation in the context of XLA pathogenesis.

PIK3R4 plays a crucial role in autophagosome biogenesis and endosome trafficking [47]. In innate immunity, autophagy is responsible for eliminating intracellular microbes through autolysosomes, distributing cytosolic microbial antigens to PRRs, and acting as an antimicrobial effector for TLRs and other PRR signaling [48]. Deficient autophagy from aberrant variants, linked to inflammatory disorders like Crohn’s disease [48], may contribute to the molecular pathogenesis of gastrointestinal illnesses in the missense group (Table 1) if hypo-expression correlates with protein levels. Moreover, a study associated autosomal recessive agammaglobulinemia with a homozygous mutation in PIK3R1 [49], emphasizing the crucial role of class IA phosphoinositide-3 kinases in immune function and regulation.

All 3 patients harboring a BTK missense mutation exhibited significant downregulation of REL. Also, taking all 45 comparison pairs for all the 10 targets into account, the only meaningful negative correlation was between BTK and REL (p = 0.0368). Four other genes are members of the NF-kappaB complex subunits gene family: RELA (p65), RELB, NFKB1 (p50/p105), and NFKB2 (p52) [50]. Different dimer combinations act as transcriptional activators or repressors, with the RELA/REL heterodimer being an activator [51]. REL polymorphisms are associated with ulcerative colitis and RA, and variants in RELB and NFKB1 have been linked to primary immunodeficiencies [52‒55]. Low REL mRNA levels in patient XLA03, associated with rectum hyperplastic polyps, may be linked to increased susceptibility to disease observed in Rel^−/− knockout mice with higher intestinal polyp formation and proliferation rates [56, 57].

The IKK complex, composed of CHUK, IKBKB, and IKBKG (also known as IKK-alpha, IKK-beta, and IKK-gamma or NEMO, respectively), is a key player in the NF-kappaB signaling pathway downstream of IRAK/TRAF/TRADD [58]. TRAF6 activates the IKK complex through ubiquitination events [58], and impaired TRAF6 ubiquitination due to IRAK2 dysfunction could reduce IKK complex activation, affecting NF-kappaB signaling and target gene expression, such as IL-6. The significant downregulation of REL in patients with BTK missense variants suggests altered IKK complex activity, as the IKK complex phosphorylates and degrades NF-kappaB inhibitor proteins, which normally sequester NF-kappaB dimers (including REL-containing dimers) in the cytoplasm [58]. Moreover, autophagy deficiency resulting from aberrant PIK3R4 variants and autosomal recessive agammaglobulinemia associated with a homozygous PIK3R1 mutation may also be connected to altered IKK complex function, as autophagy regulates IKK activity and NF-kappaB signaling [59], and PI3K signaling can modulate IKK activity through AKT-mediated phosphorylation of CHUK [60].

Of note, BTK is a negative regulator of the NLRP3 inflammasome in human and murine myeloid cells [61]. BTK deficiency in XLA patients and mice or pharmacological BTK inhibition leads to enhanced NLRP3-mediated IL-1 beta processing and release, suggesting a potential role for dysregulated NLRP3 inflammasome activation in the inflammatory phenotype of some XLA patients [61]. Negative regulatory mechanisms may have a crucial role in controlling inflammatory responses [62]. Low levels of REL and IRAK2 could indicate reduced activation of the NF-kappaB signaling pathway and potentially lead to abnormal inflammatory responses, contributing to immune dysregulation and the development of inflammatory disorders or autoimmune conditions (Fig. 3). Proper regulation of inflammatory pathways, such as NF-kappaB and the NLRP3 inflammasome, is essential to prevent excessive inflammation, tissue damage, and the development of chronic inflammatory diseases [62].

Sanaei et al. [63] found increased transcripts of SOCS1 and RNF216 in XLA PBMCs compared to healthy PBMCs, with downregulation observed after TLR inductions in XLA patients. IRAK3 showed a significant increase after LPS treatment in healthy PBMCs but a decrease in XLA PBMCs, with lower transcripts in XLA PBMCs compared to healthy PBMCs after both LPS and CpG-A treatments. The authors mention conflicting results from studies on cells from XLA patients regarding impairments in TLR signaling and cytokine production. In PBMCs from XLA patients, phosphorylation and degradation of CHUK were normal, while phosphorylation levels of p38 (MAPK14) were lower. This suggests that BTK may activate the MAPK pathway rather than the canonical NF-kappaB pathway in these cells [14, 17, 63].

Furthermore, the X-box-binding protein 1 plays a pivotal role in the unfolded protein response (UPR) and is also activated by TLR signaling, linking the UPR to the inflammatory response [64]. This activation is distinct from its canonical role in the UPR, and it has been demonstrated to be implicated in the pathogenesis of rheumatoid arthritis [64]. Also, the crosstalk between the UPR and the NF-kappaB signaling pathway has been well-documented [65]. X-box-binding protein 1 upregulation in XLA patients may be associated with inadequate TLR activation and could potentially contribute to their chronic inflammatory disorders [28].

Together with IRAK2, PIK3R4, and REL, TFRC forms the tetrad of downregulated genes in our XLA missense group. Marked TFRC downregulation in XLA01-03 and XLA09 (p ≤ 0.0122) suggests a proinflammatory response and chronic inflammation in XLA patients, as TFRC is associated with combined immunodeficiency and adaptive immunity [66]. Our patients with TFRC downregulation displayed conditions related to chronic inflammation. Inflammatory macrophages (M1), acting as the first line of defense against pathogen invasion, showed low TFRC expression and high ferritin. On the other hand, alternately activated macrophages (M2), responsible for recycling Fe++ in the inflammatory process, expressed high levels of these proteins [67].

TFRC and RIPK2 were not correlated with any of the other genes. RIPK2 expression varied among patients (Fig. 2), with hypo-expression in XLA01 (p = 0.0010) and XLA07 (p = 0.0005) and hyper-expression in XLA06 (p = 0.0005) and XLA09 (p = 0.0049). No significant correlation was found, but RIPK2 upregulation in XLA06 and XLA09 may be linked to bacterial infection susceptibility (Table 1). Further studies on NOD2 expression in XLA could enrich results. Both wild-type NOD2 and NOD2 polymorphisms/mutations can independently cause aberrant RIPK2 signaling, leading to immune dysregulation and multiple inflammatory diseases [68]. The NOD2-RIPK2 complex’s role in inflammatory disease is contradictory. Loss-of-function NOD2 mutations cause Crohn’s disease, while gain-of-function mutations cause early-onset sarcoidosis. Wild-type NOD2 and RIPK2 activity have been implicated in asthma, inflammatory arthritis, and multiple sclerosis [68].

UBE2N, associated with various cancers and infections [69], contributes to proinflammatory scenarios when overregulated. Detected in XLA01, XLA03, and XLA09 (p = 0.0640), UBE2N overregulation contrasts with marked under-regulation in XLA04-06, potentially relating to their TLR expression and signaling. TLR genes and UBE2N were mostly under-regulated for XLA04-07, the youngest patients in our cohort. TLR3, TLR6, TLR10, and UBE2N expressions positively correlated with patient age (0.0218 ≤ p ≤ 0.0368). Additionally, all pairs composed of UBE2N and the three TLR genes had a significant positive correlation (0.0107 ≤ p ≤ 0.0279). STAT3, a pivotal transcriptional repressor on UBE2N, modulates NF-kappaB activity and restrains inflammation [69], making UBE2N a promising therapeutic target. Investigating this mechanism in XLA patients may provide valuable insights.

In conclusion, our study provides preliminary evidence for a distinct TLR pathway-gene expression signature in XLA patients harboring missense BTK mutations. This strongly suggests that pathogenic BTK variants may significantly impair TLR signaling, with complex functional impact predictions due to the vast spectrum of BTK mutations. The interplay between distinct TLR signaling alterations and impaired B-cell function likely contributes to the more severe clinical manifestations observed in XLA patients with missense BTK mutations. Despite the limited cohort size and absence of protein analyses, our findings have the potential to contribute to prognosis and guide treatment strategies for XLA patients. Elucidating the TLR-/NF-kappaB-negative regulatory mechanisms is crucial for developing therapeutic strategies to modulate inflammation in chronic inflammatory and autoimmune pathologies. Future functional studies on biological samples from XLA patients should consider the BTK mutation/variant type and the cell subpopulation to better understand the genotype-phenotype correlations and the cell-specific impact of BTK deficiency on TLR signaling. Further research is warranted to expand these results and elucidate the underlying molecular mechanisms in a larger patient cohort.

We thank all patients, their families, and the control volunteers for their generous cooperation; the Central Laboratory of High Performance Technologies (LaCTAD), where RNA integrity number detection was performed, and the Laboratory of Virus/FCM for generous provision of the StepOnePlus™ Real-Time PCR System.

Written informed consent was obtained from all adult participants and all underaged participants’ parent/legal guardian for both publication of the details of their medical case and participation in the study. The research project was authorized by the Research Ethics Committee of the University of Campinas Medical School (FCM-UNICAMP), Campinas, São Paulo, Brazil (CEP #1176/2011).

The authors declare that they have no conflicts of interest in regard to the research presented in this paper.

The research was supported by grants #2011/50589-5 and #2012/06194-9, São Paulo Research Foundation (FAPESP), and the Coordination for the Improvement of Higher Education Personnel (CAPES). M.T., T.A.E., and L.F.M. were recipient of scholarships provided by CAPES.

Conception and design of the study: M.T. and M.M.S.V. Acquisition of data: M.T., T.A.E., L.M., and M.M.S.V. Data analysis and interpretation: M.T. and T.A.E. Drafting and/or revising the manuscript: M.T., I.Q., and M.M.S.V.

Marcelo Teocchi and Thaís de Andrade Eugênio contributed equally to this work and share first authorship.

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author (M.T.).