Abstract

Introduction

   Patients with the multibacillary form of leprosy can develop reactional
   episodes of acute inflammation, known as erythema nodosum leprosum
   (ENL), which are characterized by the appearance of painful cutaneous
   nodules and systemic symptoms. Neutrophils have been recognized to play
   a role in the pathogenesis of ENL, and recent global transcriptomic
   analysis revealed neutrophil-related processes as a signature of ENL
   skin lesions.

Methods

   In this study, we expanded this analysis to the blood compartment,
   comparing whole blood transcriptomics of patients with non-reactional
   lepromatous leprosy at diagnosis (LL, n=7) and patients with ENL before
   administration of anti-reactional treatment (ENL, n=15). Furthermore, a
   follow-up study was performed with patients experiencing an ENL episode
   at the time of diagnosis and after 7 days of thalidomide treatment
   (THAL, n=10). Validation in an independent cohort (ENL=8; LL=7) was
   performed by RT-qPCR.

Results

   An enrichment of neutrophil activation and degranulation-related genes
   was observed in the ENL group, with the gene for the neutrophil
   activation marker CD177 being the most enriched gene of ENL episode
   when compared to its expression in the LL group. A more
   pro-inflammatory transcriptome was also observed, with increased
   expression of genes related to innate immunity. Validation in an
   independent cohort indicated that S100A8 expression could discriminate
   ENL from LL. Supernatants of blood cells stimulated in vitro with
   Mycobacterium leprae sonicate showed higher levels of CD177 compared to
   the level of untreated cells, indicating that the leprosy bacillus can
   activate neutrophils expressing CD177. Of note, suggestive higher CD177
   protein levels were found in the sera of patients with severe/moderate
   ENL episodes when compared with patients with mild episodes and LL
   patients, highlighting CD177 as a potential systemic marker of ENL
   severity that deserves future confirmation. Furthermore, a follow-up
   study was performed with patients at the time of ENL diagnosis and
   after 7 days of thalidomide treatment (THAL, n=10). Enrichment of
   neutrophil pathways was sustained in the transcriptomic profile of
   patients undergoing treatment; however, important immune targets that
   might be relevant to the effect of thalidomide at a systemic level,
   particularly NLRP6 and IL5RA, were revealed.

Discussion

   In conclusion, our study reinforces the key role played by neutrophils
   in ENL pathogenesis and shed lights on potential diagnostic candidates
   and novel therapeutic targets that could benefit patients with leprosy.

   Keywords: leprosy, erythema nodosum leprosum, neutrophils, blood
   transcriptomics, RNAseq, thalidomide

1. Introduction

   The ancient disease leprosy is an ongoing global health issue,
   especially in countries such as India, Brazil, and Nepal, where its
   endemic status is constantly reinforced by the high numbers of new
   cases each year ([57]1). This chronic infectious disease, caused by
   Mycobacterium leprae and M. lepromatosis, can manifest as skin and
   peripheral nerve lesions, which without treatment may develop into
   serious and irreversible deformities. The clinical manifestation of the
   disease is highly linked to the host immune response, resulting in a
   broad spectrum of clinical forms ([58]2). The Ridley and Jopling
   classification, which considers histopathological aspects together with
   clinical, bacteriological, and immunological criteria, has defined the
   clinical manifestations from the least severe, contained form, known as
   the tuberculoid polar form (TT), to the most disseminated form,
   characterized by a high bacillary load, known as the lepromatous polar
   form (LL). Between these polar forms lie forms presenting graduations
   of symptoms, which are recognized as the borderline tuberculoid (BT),
   borderline borderline (BB), and borderline lepromatous (BL) forms
   ([59]3).

   The long-lasting course of leprosy can be interrupted by episodes of
   acute inflammatory responses, known as leprosy reactions. Leprosy
   reactions are linked to the high morbidity associated with the disease
   and the aggravation of neuropathic symptoms ([60]4, [61]5). About
   30–50% of multibacillary patients, represented by the BL and LL forms,
   develop erythema nodosum leprosum (ENL) ([62]5), and it is more
   frequently observed during leprosy treatment with multidrug therapy
   (MDT). However, ENL can also occur before the start of MDT or even
   years after treatment conclusion ([63]6–[64]8). ENL is characterized by
   the appearance of new nodular inflammatory cutaneous lesions, which may
   occur with systemic symptoms, such as fever, arthralgia, myalgia, and
   malaise ([65]4)

   Management of ENL episodes involves the use of immunosuppressive drugs
   to control the exacerbated inflammatory response, often requiring
   long-term use of corticosteroid and/or thalidomide, causing serious
   side effects ([66]5, [67]9). Thalidomide is extremely efficient in
   ameliorating ENL symptoms ([68]10, [69]11); however, the drug
   teratogenic properties limit its use to men and women beyond
   reproductive age. Therefore, it has only been approved for use in a few
   countries, such as India and Brazil ([70]12). Although it is believed
   that the efficiency of thalidomide is due to its immunomodulatory
   effects, its exact molecular targets in ENL are not yet fully
   understood. The definition of the inflammatory pathways inhibited by
   thalidomide in ENL could help improve therapy by the use of alternative
   drugs with less side effects.

   Although the mechanisms that trigger the reaction and the crucial
   pathways linked to the inflammatory response are still elusive,
   neutrophils are recognized as key players during ENL ([71]13, [72]14).
   Early histological studies revealed a rich neutrophilic infiltrate in
   cutaneous ENL lesions ([73]15, [74]16). In recent years, alterations in
   neutrophil physiology are described to occur during ENL. These include
   increased neutrophil degranulation ([75]17), the presence of neutrophil
   extracellular traps (NETs) in ENL skin lesions ([76]18), and the
   association of CD64 expression in the circulating neutrophil population
   with the severity of the episode ([77]19). Moreover, after 7 days of
   thalidomide treatment the neutrophilic infiltrate decreases in
   cutaneous ENL lesions ([78]18–[79]21), suggesting that limiting
   neutrophil participation during ENL is important for its remission, and
   that the effectiveness of the drug might be, in part, linked to this
   phenomenon.

   Given the countless questions that still require attention regarding
   prevention, immunopathogenesis, and therapeutic interventions in ENL,
   high throughput techniques are a resourceful investigative tool that
   can be utilized to unveil novel targets and pathways of interest. At
   present, most transcriptomic studies in leprosy have focused on skin
   samples and have reported increased expression of neutrophil genes in
   the context of ENL ([80]22–[81]24). While cutaneous manifestations are
   central in ENL, the systemic alterations that occur are poorly explored
   and understood. While investigating blood cells expression profile in
   leprosy, most studies explore differences between polar forms ([82]25,
   [83]26) and the sole transcriptomic study of peripheral blood
   mononucleated cells (PBMCs) from ENL patients strengthened the
   importance of inflammatory pathways ([84]27). Nevertheless, the absence
   of neutrophils in this particular study limited the overall delineation
   of important pathways to the systemic inflammatory process.

   A goal of the study reported here was to provide a whole blood
   transcriptomic profile of patients with active ENL and compare it with
   that of patients with non-reactional LL for a better understanding of
   the systemic pathogenic events associated with ENL. Secondly, whole
   blood transcriptomics analyses of patients with ENL before and after 7
   days of thalidomide treatment was conducted as a strategy to identify
   thalidomide-modulated pathways, and thus, identify potential targets
   for efficient ENL treatment.

2. Methods

2.1. Patients

   Patients with leprosy enrolled in this study were assisted at the Souza
   Araujo Outpatient Unit (Fundação Oswaldo Cruz, Rio de Janeiro, Brazil),
   a reference center for leprosy diagnosis and treatment. The study
   population was divided into two groups: 1) newly diagnosed lepromatous
   leprosy (LL) patients, treatment naïve, with no signs of reaction; and
   2) patients with reaction diagnostics (ENL) before starting
   anti-reactional treatment, previously diagnosed with BL/LL forms of
   leprosy. Patients were enrolled in three main different assays: RNAseq
   {ENL (n=15), LL (n=7)}; RT-qPCR assay {ENL (n=8), LL (n=7)}; CD177
   ELISA {ENL (n=28), LL (n=21)}. Patients with ENL were also followed up
   on the seventh day of thalidomide therapy.

   Leprosy diagnosis was based on the presence of hypopigmented,
   anesthetic skin patches, thickened nerves, histopathological skin
   biopsies analysis, and acid-fast bacilli in skin smears ([85]3).
   Patients with leprosy received MDT as recommended by the World Health
   Organization (WHO). ENL diagnosis was based on the detection of an
   acute appearance of crops of tender cutaneous or subcutaneous lesions
   with or without any systemic symptoms. Patients with ENL were enrolled
   in the longitudinal study when treated with thalidomide (100–300 mg
   daily) {RNAseq (n=10); RT-qPCR assay (n=8)}, in compliance with
   Brazilian Ministry of Health guidelines. Patients with ENL included in
   this study did not receive corticosteroids or immunomodulatory drugs at
   least 6 months before sample collection. Healthy donors (HD) were also
   included in the study for in vitro analysis (n=7). None of the patients
   or HD presented co-morbidities, such as co-infections, diabetes, or
   cancer.

   The study was approved by the Instituto Oswaldo Cruz ethics committee,
   CAAE number 56113716.5.0000.5248.

2.2. Blood samples

   Blood samples were collected after the informed consent form was signed
   by the participant. Peripheral whole blood samples were collected from
   all leprosy patients and HD enrolled in the study. A second point of
   collection was included in leprosy patients with ENL 7 days after
   starting thalidomide treatment. Approximately 2.5 mL of peripheral
   blood was collected in PAXgene Blood RNA Tubes (Qiagen, Germany) for
   transcriptomic analysis and quantitative reverse transcription
   polymerase chain reaction (RT-qPCR). In addition, blood samples were
   also collected in heparin-coated vacutainer collection tubes (BD, USA)
   and tubes with no additives (BD, USA), which were used, respectively,
   for whole blood culture or to obtain serum for CD177 protein
   quantification. Serum was obtained by incubating samples for 15 minutes
   at room temperature, followed by centrifugation at 300 x g for 10 min
   at 4°C for serum collection at the supernatant and storage at -80°C.

2.3. ENL severity

   The clinical severity of the ENL cases was determined according to the
   ENL severity scale defined by ENLIST ([86]12). This is a 10-item scale
   comprising: visual scale of pain; fever; the number, inflammation, and
   extent of ENL skin lesions; peripheral edema; bone pain; inflammation
   of joints and/or digits; lymphadenopathy; and nerve tenderness owing to
   ENL. Patients whose clinical records at the time of sample collection
   encapsulated the aforementioned information were categorized.
   Conversely, patients with incomplete records were disregarded from
   classification and were designated as “not classified.”

2.4. RNA extraction and sequencing

   Whole blood transcriptomics were performed with blood samples collected
   in the PAXgene Blood RNA Tubes. Total RNA was obtained using the
   PAXgene Blood RNA kit (Qiagen, Germany) following the manufacturer’s
   protocol, including the kit’s DNAse treatment, as instructed. Agilent
   TapeStation 2200 (Agilent, USA) was used to assess RNA quality and all
   RNAs presented an RNA integrity number above 8. mRNA was enriched using
   the NEBNext Poly(A) mRNA Magnetic Isolation Module (New England
   Bioscience, USA). cDNA library preparation was performed using the
   NEBNext Ultra II Directional RNA Library Prep kit for Illumina (New
   England Bioscience, USA). The libraries were sequenced at the
   Next-generation sequencing core facility (Colorado State University,
   USA) using the NextSeq 500/550 High Output kit v2.5 (75 cycles)
   (Illumina, USA) on the Illumina NextSeq 550 platform. The raw data were
   deposited in Gene Expression Omnibus ([87]GSE198609). [88]Supplementary
   Table S1 summarizes the baseline characteristics of the patients
   included in the RNAseq analysis.

2.5. RNAseq analysis

   Initially, quality control (QC) of reads obtained after sequencing was
   accomplished using FastQC v.0.11.8
   ([89]https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) and
   MultiQC v.1.9 ([90]28), followed by removal of adapters using fastp
   v.0.21.0. In this step, poly(A) sequences were also removed, followed
   by trimming the initial 10 bases at the 5’-end. Sequence reads were
   mapped and quantified using the reference human transcriptome
   (GRCh38p.12,
   [91]http://refgenomes.databio.org/v2/asset/fa159612d40b1bedea9a279eb249
   99b3d27145f9dd70dcca/salmon_index/splash?tag=default) according to the
   Salmon v.1.4. pipeline using the –gcBias and –seqBias flags, and
   transcript quantification was summarized into ENSEMBLE genes with
   tximport v.1.12.0 ([92]29) and biomaRt v.2.40.5 ([93]30). Differential
   gene expression (DGE) analysis using DESeq2 v.1.24.0 ([94]31) was
   performed with ENL versus LL (adjusted for sex) groups, and paired
   samples of patients with ENL at day 7 of thalidomide treatment versus
   at the time of diagnosis (THAL vs. ENL). P-values were adjusted for
   multiple testing using the Benjamini and Hochberg procedure to control
   the false discovery rate (FDR) ([95]32), and fold-changes (FC) were
   moderated with the “ashr” adaptive estimator ([96]33). Genes considered
   differentially expressed complied with the following: |log[2]FC| ≥
   0.585 (FC ≥ 1.5) and FDR ≤ 0.1. The normalized expression matrix was
   transformed using shifted logarithm (base 2) and standardized to mean
   zero and unit standard deviation for the heatmap and unsupervised
   hierarchical clustering. Hierarchical clustering of genes and samples
   was done using the R package pheatmap v.1.0.12 with the Pearson
   correlation coefficient as the distance metric ([97]34). Pathway
   enrichment analysis was performed using “biological processes” from
   Gene Ontology database through over-representation analysis (ORA) using
   the package cluterProfiler v.4.10.0 ([98]35–[99]38). All P-values
   reported are two-tailed. For principal component analysis (PCA), the
   top 500 most variable genes after normalization were used, applying
   DESeq2’s variance stabilization procedure (blinded concerning groups).

2.6. Quantitative reverse transcription PCR

   Reverse transcription was performed on whole blood RNA using
   Superscript IV VILO Master Mix (Invitrogen-Thermo Fisher Scientific,
   USA) and RT-qPCR reactions were performed in duplicates on a ViiA7
   Real-Time PCR System (Applied Biosystems, USA) using PowerUp SYBR Green
   Master Mix (Applied Biosystems-Thermo Fisher Scientific, USA). For
   reactions, 10 ng of cDNA and 400 nM of oligonucleotide primers were
   used (except for CD177 primers, which were used at 800 nM). The
   sequences of primers are available in [100]Supplementary Table S2 . The
   genes RPL13a and RPS16 were used as reference genes for normalization.
   The relative gene expression analysis was performed using the
   efficiency-adjusted N0 method as previously described ([101]39,
   [102]40). Briefly, raw data were exported from ViiA7 proprietary
   software and imported into LinRegPCR v.2022 for obtaining N0 values.
   The N0 values were normalized by taking their ratio to the
   normalization factor calculated using the geometric mean of N0 values
   for the reference genes. Normalized expression values were converted
   into log[2] before statistical analyses and visualization in R v.4.1.
   Statistical significance of the difference between the two means was
   assessed using Welch’s t-test, while the paired t-test was used for the
   follow-up design. Additionally, Spearman’s correlation tests were
   performed between CD177 gene expression levels and the remaining subset
   of genes analyzed in patient samples of the ENL group using GraphPad
   Prism v.8.0.0.

2.7. Whole blood cell culture with M. leprae sonicate

   We collected two milliliters of whole blood from a group of seven
   individuals, consisting of two males and five females, with an average
   age of 31 years (referred to as “HD”). The blood was placed directly
   into 24-well plates with no additional medium, and then incubated at
   37°C and 5% CO[2] for 24 hours in the presence or absence of M. leprae
   sonicate (NR19329, BEI Resources, USA) at a concentration of 10 μg/mL.
   Following incubation, the whole blood samples were subjected to the
   previously described centrifugation process ([103]17), and the plasma
   was analyzed for CD177 measurement.

2.8. CD177 protein measurement

   Concentrations of CD177 protein in the serum samples of patients
   (diluted 1:2) or plasma of whole blood cell cultures (diluted 1:4) were
   determined via a commercial enzyme-linked immunosorbent assay kit
   (ELISA), according to the manufacturer’s protocol (Cat#: EH80RB,
   Invitrogen-Thermo Fisher Scientific, USA). Readings were performed in
   the EON Microplate spectrophotometer (BioTek, USA) and data were
   analyzed using Gen5 software (BioTek, USA). Mean absorbances of
   duplicate readings for each sample were compared with those of the
   standard curve to give the concentrations, given in ng/mL. The minimum
   detectable concentration of human CD177 was 0.21 ng/mL.

2.9. Statistical analysis

   Statistical analysis of CD177 levels was carried out using GraphPad
   Prism v.8.0.0. For the analysis of the plasma levels after whole blood
   cell stimulation with M. leprae sonicate, a two-tailed paired t-test
   was performed after the normal distribution was confirmed by the
   Shapiro–Wilk test. For analysis of the patient serum levels, normal
   distribution was first determined using the Shapiro–Wilk test then the
   two-tailed Welch’s t-test was applied. For all statistical analyses, a
   P-value was considered statistically significant when lower than 0.05.

3. Results

3.1. Patient clinical background

   Characteristics of the patients with leprosy enrolled in this study are
   shown in [104]Supplementary Tables S1, S3, S4 . Patients were divided
   into either treatment-naïve with the LL form without any sign of
   reaction at leprosy diagnosis (LL group) or those with active ENL
   before starting anti-reactional treatment (ENL group). The median age
   of the LL group was 38 years (interquartile range, 30–51), and 70% were
   men. The ENL group comprised 75% of men and the median age was 40 years
   (interquartile range, 31–55). None of the patients of the LL group
   developed reaction during the first 4 months of MDT treatment. Blood
   samples of all patients were collected for analysis, and a second
   sampling was performed for the patients of the ENL group after 7 days
   of thalidomide treatment (THAL group).

3.2. Genes associated with neutrophil activation and degranulation are
enriched in the blood transcriptomics of patients with ENL

   Genome-wide transcriptional profiles were generated from the blood of
   patients from ENL group (n = 15) versus patients from the LL group (n =
   7). A dataset of 16,525 expressed genes was generated, and the PCA of
   the 500 most differentially expressed genes was unable to separate
   samples into two distinct groups ([105] Figure 1A ). Although LL
   patient samples were tightly grouped, ENL patient samples were
   scattered between the two first components, highlighting their greater
   heterogeneity. Thirty-four statistically significant differentially
   expressed genes were identified, all of which were enriched in ENL
   compared to their levels in LL ([106] Figure 1B ; [107]Supplementary
   Table S5 ). The neutrophil activation marker CD177 ([108]41) was the
   most enriched gene in ENL, increased by about 8-fold compared to the
   level in LL (log[2]FC = 2.96, adj. p = 0.0003), followed by SOCS3
   (log[2]FC = 2.05, adj. p = 0.001), a master regulator of the immune
   response ([109]42), and FCGR1A (log[2]FC = 1.97, adj. p = 0.003) and
   FCGR1B (log[2]FC = 1.90, adj. p = 0.003), encoding the Fc gamma
   receptors 1A (CD64A) and 1B (CD64B), respectively ([110]43, [111]44). A
   heatmap comprising all 34 differentially expressed genes was drawn
   ([112] Figure 1C ). A homogeneous profile was observed for LL patients
   in agreement with the PCA pattern. In contrast, ENL patients clearly
   showed a more heterogeneous profile, with a few (n = 4) presenting an
   expression profile more similar to that of LL patients, aligned with
   the diverse symptomatology observed clinically. In fact, the
   categorization of severity of the ENL episodes according to the ENLIST
   criteria exhibited significant heterogeneity ([113] Supplementary Table
   S1 ). However, a tentative to correlate the RNAseq profile with the
   severity and/or chronicity of the reaction failed, with no
   identification of discernible patterns according to the ENLIST severity
   status.

Figure 1.

   [114]Figure 1
   [115]Open in a new tab

   Patients with lepromatous leprosy experiencing ENL display a distinct
   blood transcriptomic profile to that of non-reactional patients. (A)
   Principal component analysis of the 500 most variable genes in patients
   with lepromatous leprosy and active erythema nodosum leprosum (ENL;
   red, n=15) compared to those that were non-reactional (LL; yellow,
   n=7). (B) Volcano plot analysis of differentially expressed genes in
   ENL vs. LL. Differentially expressed genes are depicted in blue. (C)
   Heatmap showing the differentially expressed genes in the blood of ENL
   patients over LL patients. Expression levels are represented by a scale
   of log[2]FC from -3 (dark blue) to 3 (bright red).

   A functional analysis of the genes significantly enriched in the blood
   of ENL patients by over-representation analysis (ORA) showed an
   abundance of genes involved in innate immunity and inflammatory
   pathways ([116] Figures 2A, B ). Notably, 9 out of the 34 upregulated
   genes (26.5%) were associated with neutrophil/granulocyte activation,
   degranulation, and migration pathways ([117] Figure 2B ). These genes
   were CD177, CYSTM1, HP, LRG1, MCEMP1, MMP8, OLFM4, S100A12, and S100A9.
   This was an expected observation based on previous literature reports
   of neutrophilia during ENL ([118]45–[119]47). Indeed, we also observed
   higher neutrophil counts in the blood of a subset of ENL patients from
   the RNAseq cohort with available blood counts by the time they were
   recruited for the study (LL= 7; ENL = 5) ([120] Supplementary Figure S1
   ). Other pathways with enrichment included “response to bacterium” with
   5 associated genes including TLR5, and “interferon-gamma-mediated
   signaling pathway” with 3 associated genes, FCGR1A, FCGR1B, and SOCS3.

Figure 2.

   [121]Figure 2
   [122]Open in a new tab

   Biological processes and related genes enriched in the blood
   transcriptomics of patients experiencing ENL. (A) Dotplot and (B)
   heatplot of enriched biological processes associated with positively
   regulated genes in patients with erythema nodosum leprosum (ENL)
   obtained by the over-representation analysis (ORA). In (A), pathways
   are depicted in a hierarchy according to the enrichment ratio. Circle
   size and color indicate, respectively, the number of genes associated
   with the pathway and the P-value according to the false discovery rate
   (FDR). In (B), the identity of the genes associated to each enriched
   biological process is given. The square color indicates the log[2]FC
   value according to the scale.

   To further explore neutrophil associated genes, the genes CD177, CHIT1,
   OLFM4, S100A12, S100A9, and S100A8 were selected based on their
   extensive use as markers of neutrophilic processes ([123]48–[124]53).
   Of note, the inclusion of the genes S100A8 and CHIT1 in this analysis,
   despite their absence from differentially expressed genes, was based on
   the intimate relation between S100A8 and S100A9, which often act as a
   heterodimer ([125]54), and the presence of CHIT1 among top positively
   regulated genes in the subsequent longitudinal study. The expression
   levels of neutrophil-related genes in the RNAseq for each patient
   showed an overall heterogeneous profile within the ENL group and some
   patients had markedly increased expression compared to the levels in
   non-reactional LL patients ([126] Figure 3A ). A similar pattern of
   enrichment was also observed with the CHIT1 and S100A8 genes that,
   although not among the differentially expressed genes from the RNAseq
   data, had higher average expression levels in ENL patients compared to
   the levels in LL patients ([127] Figure 3A ). Replication of RNAseq
   results by RT-qPCR in an independent patient cohort (LL=7; ENL=8)
   revealed a similar distribution pattern of the expression levels for
   the chosen neutrophil genes ([128] Figure 3B ). Although the mean
   expression values obtained by RT-qPCR were higher in ENL patients than
   in LL patients for four genes among six tested, a significant
   difference was only observed for S100A8 and CHIT1 expression.
   Nevertheless, a positive correlation was observed between the
   normalized expression levels of CD177 and all three S100 genes analyzed
   in ENL patients ([129] Figure 3C ), where the strongest correlation was
   observed with S100A12 (r = 0.8332; p = 0.0154), followed respectively
   by correlation with S100A8 (r = 0.8095; p = 0.0218) and S100A9
   expression (r = 0.7381; p = 0.0458), while no correlation was observed
   with the remaining set of genes (data not shown).

Figure 3.

   [130]Figure 3
   [131]Open in a new tab

   Neutrophil genes are enriched in the blood transcriptomics of patients
   experiencing ENL. (A) Normalized expression levels from RNAseq
   differential gene expression (DGE) analysis of patients with
   lepromatous leprosy who were non-reactional (LL; purple, n=7) or had
   active erythema nodosum leprosum (ENL; red, n=15) for neutrophil genes:
   CD177, CHIT1, OLFM4, S100A12, S100A8, and S100A9. (B) Log normalized
   expression levels of the neutrophil genes CD177 (LL=6; ENL=8), CHIT1
   (LL=5; ENL=6), OLFM4 (LL=4; ENL=8), S100A12 (LL=6; ENL=8); S100A8
   (LL=7; ENL=8) and S100A9 (LL=7; ENL=8) obtained by RT-qPCR in a new
   cohort of LL (purple) and ENL patients (red). The genes RPL13 and RPS16
   were used as reference controls for gene expression normalization.
   Statistical analysis was performed by applying Welch’s t-test. (C)
   Spearman’s correlation test of normalized expression log[2] data from
   RT-qPCR of ENL patients between CD177 and S100A8 (left panel), CD177
   and S100A9 (middle panel), and CD177 and S100A12 (right panel).
   Positive correlation and significance are considered when the r-value
   is > 0.7 and the p-value is < 0.05, respectively.

   A suggestive increase in protein levels of CD177 was found in serum
   samples of ENL patients with the median value about twice that of the
   levels in LL patients (ENL median: 4.016 ng/mL; LL median: 2.094 ng/mL,
   p = 0.1210) ([132] Supplementary Figure S2 ). Notably, subsequent ENL
   severity stratification according to ENLIST parameters revealed
   suggestive higher serum CD177 levels in patients with moderate or
   severe episodes compared to those of patients with mild episodes (p =
   0.0651), as well as comparing with LL patients (p = 0.0524) ([133]
   Figure 4A ). Moreover, in vitro stimulation of whole blood cells from
   HD with M. leprae sonicate significantly increased the release of this
   protein into the plasma ([134] Figure 4B ), suggesting that
   mycobacterial components may activate neutrophils expressing CD177 in
   vivo. Taken together, these results reinforce the key role played by
   neutrophils in ENL immunopathogenesis and suggest that the severity of
   the reactional episode may be linked to the degree of neutrophil
   activation.

Figure 4.

   Figure 4
   [135]Open in a new tab

   The CD177 molecule may be related to ENL severity. (A) CD177 levels in
   sera from non-reactional lepromatous leprosy (LL; n=21) and ENL
   patients classified according to the ENLIST ENL Severity Scale: mild
   ENL reaction (n=5) and moderate-severe reaction (n=16). Statistical
   analysis was performed using a two-tailed Welch’s t-test. (B) Whole
   blood cells from healthy donors (n=7) were stimulated with 10 µg/mL of
   M. leprae sonicate (MLs) or not (NS, non-stimulated) for 24 h and the
   CD177 protein levels in the plasma were determined by ELISA.
   Statistical analysis was performed using a two-tailed paired t-test
   (*p<0.05).

3.3. Effect of thalidomide treatment on the blood transcriptomics of patients
with ENL

   Together with establishing a blood transcriptomic profile of patients
   experiencing ENL, this study sought to determine the profile of these
   patients after starting thalidomide treatment (THAL). A segment of the
   patients with ENL who were administered thalidomide therapy and met the
   specified criteria delineated in the methodology section were recruited
   for participation in the longitudinal investigation (n=10). The
   timepoint of choice for sample collection was at the seventh day of
   treatment, the time at which several indications of treatment efficacy
   have previously been reported, especially for cutaneous symptoms
   ([136]10, [137]11). Owing to the heterogeneous pattern of gene
   expression among patients with ENL, pairwise comparisons for each
   patient were performed to determine the transcriptional changes. As
   expected, the PCA of the 500 most variable genes was unable to separate
   samples into two distinct groups highlighting their heterogeneity
   ([138] Figure 5A ). In this RNAseq analysis, 16,527 transcripts were
   detected. However, only 64 genes were found to be significantly
   differentially expressed ([139] Figure 5B ). Hierarchical clustering
   shows less prominent regulation of these 64 genes after 7 days of
   treatment in two patients, presenting a transcriptional profile more
   similar to patients before treatment. The hierarchical clustering also
   showed a more homogeneous gene expression pattern after thalidomide
   treatment in comparison with that before treatment, with the upper
   third of the heatmap displaying the downregulated genes, and the lower
   two-thirds displaying the upregulated genes in the THAL samples ([140]
   Figure 5C ).

Figure 5.

   [141]Figure 5
   [142]Open in a new tab

   Pairwise comparisons for patients with ENL before and after 7 days of
   thalidomide treatment to determine transcriptional changes in the
   blood. (A) Principal component analysis of the 500 most variable genes
   in patients with erythema nodosum leprosum (n=10) at diagnosis (ENL;
   red) and after 7 days of thalidomide treatment (THAL; blue). (B)
   Volcano plot analysis of differentially expressed genes in the ENL and
   THAL groups. Significantly expressed genes are represented in blue. (C)
   Heatmap of differentially expressed genes in the blood cells of the ENL
   (red) and THAL (blue) groups. The scale on the right represents the
   expression levels of each gene ranging from a log[2]FC value of -4
   (dark blue) to 4 (bright red).

   DGE analysis of patients with ENL before and after thalidomide
   treatment found a total of 41 positively regulated genes and 23
   negatively regulated genes in the THAL group when compared to the
   levels in the ENL group ([143] Supplementary Table S6 ). The three most
   upregulated genes were ATP2C2 (log[2]FC = 2.09, adj. p < 0.001), TMEM92
   (log[2]FC = 2.07, adj. p < 0.001), and IGFB2 (log[2]FC = 2.02, adj. p <
   0.001). Among the upregulated genes, some immunity-related genes were
   observed, such as CAMP and IL5RA ([144] Figure 5B ; [145]Supplementary
   Table S6 ).

   Although no significantly enriched pathways were found for the set of
   downregulated genes, some of these genes are associated with important
   immune regulation processes, such as the inflammasome-related gene
   NLRP6, and MRAS, which encodes a regulatory protein from the RAS family
   GTPases ([146] Supplementary Table S6 ). For the upregulated genes,
   over 20 enriched pathways were found; however, 41% of the genes (17 out
   of 41) were still associated with neutrophil activation/degranulation
   pathways, suggesting the persistence of neutrophil activation despite
   the initiation of thalidomide treatment ([147] Figures 6A, B ). Besides
   the pathways linked to neutrophil activation, several unique biological
   processes, unseen in the ENL vs. LL transcriptomic analysis, were found
   enriched during thalidomide treatment, such as “humoral immune
   response” and “antimicrobial humoral response” with 8 genes each, and
   “mucosal immune response” and “antibacterial humoral immune response
   mediated by antimicrobial peptide” with 4 genes each ([148] Figures 6A,
   B ).

Figure 6.

   [149]Figure 6
   [150]Open in a new tab

   Biological processes represented in the blood transcriptomic profiles
   of patients with ENL before and after 7 days of thalidomide treatment.
   (A) Dotplot portraying the enriched pathways in blood cells of patients
   with erythema nodosum leprosum (ENL) at day 7 of thalidomide treatment
   (THAL group), compared to the moment of ENL diagnosis (ENL group). The
   pathways are hierarchically positioned according to the enrichment
   ratio. Circle size indicates the number of genes associated with the
   pathway and the color indicates the P-value according to the false
   discovery rate (FDR). (B) Heatplot of enriched biological processes
   associated with genes positively regulated in the THAL group obtained
   by over-representation analysis (ORA). Values of log[2]FC are
   color-graded according to the scale.

   The majority of upregulated genes associated with neutrophilic pathways
   during thalidomide treatment, including CAMP, CHIT1, LTF, and LCN2
   genes, are broadly different from those observed in the ENL vs. LL
   comparison ([151] Figure 7A ) and are strongly associated with
   neutrophil granules ([152]55, [153]56). This sustained neutrophil
   participation even during thalidomide treatment can be inferred not
   only by the pattern of increased expression levels of the neutrophilic
   markers, CHIT1 and OLFM4, in the RNAseq analysis ([154] Figure 7B ),
   but also by the increased OLFM4 expression in the RT-qPCR replication
   analysis using a different patient cohort ([155] Figure 7C ).

Figure 7.

   [156]Figure 7
   [157]Open in a new tab

   Neutrophil genes and pathways are still overrepresented in
   thalidomide-treated patients with ENL. (A) Dotplot representing the
   neutrophil genes differentially expressed in the ENL vs. LL comparison
   and in the THAL vs. ENL paired comparison. The circle size represents
   the log of the adjusted p-values. The circle color represents the
   log[2]FC values according to the scale. (B) Normalized expression
   levels from RNAseq differential gene expression (DGE) analysis of the
   longitudinal analysis of patients with ENL (n=10) at diagnosis (ENL
   group; red) and after 7 days of thalidomide treatment (THAL group;
   blue) for CHIT1, OLFM4, IL5RA, PI3, NLRP6, and VSIG4 genes (C)
   Expression levels of IL5RA (n=8), PI3 (n=7), NLRP6 (n=8), VSIG4 (n=7)
   and OLFM4 (n=8) genes obtained by RT-qPCR in a new cohort of patients
   with ENL at diagnosis (ENL; red) and after 7 days of thalidomide
   treatment (THAL; blue). The genes RPL13 and RPS16 were used as
   reference controls for gene expression normalization. Statistical
   analysis was performed using a paired t-test.

   A detailed inspection of the positively regulated genes revealed
   interesting targets that might contribute to subsiding the inflammatory
   process during an ENL episode, such as the genes VSIG4, PI3, and IL5RA
   ([158] Figure 6B ). Indeed, the pairwise comparison of the expression
   levels of these genes revealed higher levels after thalidomide
   treatment in most patients ([159] Figure 7B ). The RT-qPCR replication
   analysis identified the same expression pattern for IL5RA (log[2]FC =
   0.48; p = 0.0259). The overall gene expression behavior of VSIG4 by
   RT-qPCR was also like that seen in the RNAseq. While the difference
   between VSIG4 averages was not statistically different, the higher
   levels of expression during thalidomide treatment may be worth future
   investigation with larger cohort of samples. In contrast, RT-qPCR
   analysis of PI3 showed high heterogeneity, underpowering the experiment
   to detect average differences ([160] Figure 7C ). NLRP6 was also
   selected for analysis because of its association with the inflammasome.
   In most patients, the expression levels of NLRP6 were reduced by
   thalidomide treatment in the RNAseq analysis ([161] Figure 7B ). A
   similar result was observed in the RT-qPCR analysis with a distinct
   cohort of patients (log[2]FC = -1.41; p = 0.0549) ([162] Figure 7C ),
   confirming the differences observed in the RNAseq.

4. Discussion

   Whole blood transcriptomic analysis as a tool to better understand
   diseases has been increasingly used over the last decade ([163]57,
   [164]58). Our study applied this approach to analyze patients with
   active ENL, as well as thalidomide-induced changes in ENL patients
   under treatment to better understand the pathways important for ENL
   management. Our RNAseq data demonstrated a highly inflammatory profile
   in the blood of patients with active ENL compared to non-reactional
   lepromatous patients. The most prominent enriched genes in ENL patients
   were those associated with neutrophil activation and degranulation.
   This was an expected result based on the neutrophilia previously
   described and herein confirmed in ENL patients, and the increased
   levels of activation and degranulation already reported in circulating
   ENL neutrophils ([165]13, [166]14, [167]17–[168]20). The data obtained
   in blood align with the neutrophilic infiltrate in ENL skin lesions
   ([169]15, [170]16), and the recent meta-analysis of available
   transcriptomic data of skin lesions highlighting the upregulation of
   pathways associated with a neutrophil response in ENL ([171]24). A
   neutrophil-enriched blood transcriptomic profile was described in
   tuberculosis ([172]59), rheumatoid arthritis and SLE ([173]60–[174]62).
   Interestingly, the neutrophilic signature in SLE has been linked to an
   increased risk of vascular inflammation ([175]63), which can also be
   observed in a portion of ENL skin lesions ([176]64). In conclusion, our
   findings strengthen the importance of neutrophils during ENL,
   suggesting that they are a linchpin of ENL immunopathogenesis.

   Although PBMC preparations are devoid of the normal-density
   neutrophils, Dupnik et al. (2015) reported neutrophil-related genes
   among the ones upregulated in the PBMCs of ENL patients ([177]27), such
   as OLFM4, LTF, and LCN2. This observation could have been related to an
   increased low-density neutrophil population reported in ENL patients
   ([178]17). A major hypothesis formulated by the authors was that
   low-density neutrophils are formed because of degranulation, which
   would indicate the augmented degranulation and activation in the
   neutrophil population ([179]17). Our study corroborates this previous
   finding by not only demonstrating a neutrophilic signature profile in
   blood cells of patients with ENL, but also by observing the enrichment
   of the biological process from gene ontology of “neutrophil
   degranulation”.

   In our RNA-seq analysis, CD177 appeared as the most differently
   expressed gene in ENL when compared to non-reactional LL patients.
   CD177 is commonly acknowledged as a marker of neutrophil activation
   ([180]65, [181]66) and it has been associated with neutrophil migration
   ([182]67), degranulation ([183]68) and NET formation ([184]69)
   Increased numbers of circulating CD177^+ neutrophils have also been
   reported in other systemic inflammatory diseases, such as sepsis
   ([185]70), ANCA-associated systemic vasculitis (AAV), and SLE
   ([186]71). Of note, CD177 expression levels in our study were very
   heterogeneous. This observation was not unexpected as CD177 variability
   in the human population is markedly linked to an individual’s genetic
   background ([187]72). We found suggestive increased CD177 serum levels
   in moderate-severe reactional episodes compared to those found in mild
   episodes, pointing CD177 as a potential biomarker candidate for ENL
   severity, which deserves confirmation in future studies. Interestingly,
   CD177 has also been recently linked to COVID-19 severity ([188]48,
   [189]65, [190]73).

   Moreover, several genes identified in our study with higher expression
   levels in whole blood during the active episode of ENL corroborate
   previous studies. Notably, this includes genes such as S100A8, S100A9,
   and S100A12, whose respective encoded proteins are reported to be
   elevated in the serum of ENL patients ([191]74, [192]75). The S100
   protein family is intimately linked to neutrophil activation ([193]54,
   [194]76) and inflammatory processes, and diseases such as sepsis,
   rheumatoid arthritis, psoriasis and SLE ([195]54, [196]77). Moreover,
   S100A8 and S100A9 can stimulate transendothelial migration of
   neutrophils into affected tissue ([197]78–[198]80) and recently were
   defined as markers of NETosis ([199]76). Increased expression of
   S100A8/S100A9 was also observed in ENL skin lesions ([200]22),
   reinforcing their role as key players of ENL inflammation. Remarkably,
   in the RT-qPCR cohort, we also observed a positive correlation of
   expression of CD177 with those of S100A8/A9/A12, where the pattern of
   increased expression level of one gene matches the other, suggesting an
   interplay of these neutrophil-associated genes during ENL.
   Interestingly, several studies have observed a combined increased
   expression of these genes in different inflammatory contexts, such as
   sepsis ([201]81) and COVID-19 ([202]82), strengthening the correlation
   among these genes. Nevertheless, the precise mechanistic role of S100
   proteins during ENL requires further study.

   The increased expression of FCGR1A and FCGR1B, which code for distinct
   forms of the high affinity FCγ receptor (CD64), in ENL patients agrees
   with previous descriptions of increased CD64 on both circulating
   neutrophils ([203]19) and PBMCs from ENL patients ([204]27). Moreover,
   increased levels of CD64^+ neutrophils have been associated with the
   severity of the ENL episode ([205]19). The FCGR1B gene is induced in
   monocytes and neutrophils by IFN-γ ([206]83). The upregulation of
   FCGR1B expression is, therefore, substantiated by the extensive number
   of studies reporting the IFN-γ contribution during ENL, such as the
   observations of increased IFN-γ levels in patient sera ([207]84,
   [208]85), overexpression of IFN-γ and related genes in PBMCs ([209]27,
   [210]86, [211]87), and in ENL skin lesions ([212]22, [213]86).

   Several reports demonstrate the capacity of thalidomide to modulate
   aspects of neutrophil biology ([214]88), such as decreasing neutrophil
   phagocytosis ([215]89), transendothelial migration to sites of
   inflammation ([216]90, [217]91), and downregulation of the neutrophil
   activation marker CD11b ([218]92). Previous studies on ENL report a
   reduction of the neutrophilic inflammatory infiltrate ([219]18–[220]20)
   and NET formation ([221]18) in ENL skin lesions after 7 days of
   thalidomide treatment. However, our blood transcriptomics analysis
   revealed the persistence of inflammatory pathways, including those
   linked to neutrophil activation and degranulation, at day 7 of
   treatment with thalidomide. A prominent neutrophil gene during
   thalidomide treatment was CHIT1. Augmented gene expression of CHIT1 was
   reported in ENL skin lesions ([222]93) and increased chitotriosidase
   activity in the serum of ENL patients in comparison to those of healthy
   individuals have been reported ([223]94). CHIT1 encodes the enzyme
   chitotriosidase (or chitinase 1), which is linked to neutrophil
   granules ([224]95) and associated with an intense inflammatory profile
   ([225]96–[226]98). Altogether, these data indicate that a longer period
   of treatment with the drug is required to observe reduction of
   neutrophil enrichment and activation systemically.

   Despite the persistence of an inflammatory profile in the blood after a
   short period of thalidomide administration, several new upregulated
   genes/pathways with anti-inflammatory effects were detected. These
   include the gene V-set immunoglobulin-domain-containing 4 (VSIG4), that
   codes for a membrane protein belonging to complement receptor of the
   immunoglobulin superfamily (CRig). VSIG4 has been implicated in NLRP3
   inflammasome inhibition ([227]99), suppression of T cell activity
   ([228]100) and inhibition of LPS-induced pro-inflammatory macrophage
   activity ([229]101). Interestingly, successful administration of VSIG4
   results in remission of rheumatoid arthritis ([230]102) and decreases
   kidney deposition of immunocomplexes to ameliorate SLE-associated
   glomerulonephritis ([231]103). This points to VSIG4 as a potential
   candidate for ENL therapy. The study of VSIG4 upregulation in ENL
   patients under thalidomide treatment needs further confirmation with a
   larger cohort of samples and with longer periods of treatment.

   Another important change in the transcriptomic profile observed during
   thalidomide treatment was the appearance of genes associated with
   humoral and Th2 immunity, such as IL5RA. This is in line with the
   observation that thalidomide treatment of PBMCs in vitro stimulates the
   production of Th2-related cytokines ([232]104). Based on these data, we
   posit that the emergence of a Th2 response during thalidomide treatment
   counteracts the involvement of a T-cell mediated response in ENL
   pathology ([233]105).

   NLRP6, a gene that codes for the cytoplasmic receptor of the NLRP6
   inflammasome, was also downregulated after thalidomide treatment. This
   gene is also linked to neutrophil migration and induction of NETosis
   during sepsis ([234]106). Thus, based on this, we speculate that NLRP6
   inflammasome participates in ENL pathogenesis and represents a
   potential inflammatory pathway targeted by thalidomide.

   It is noteworthy to discuss the lack of modulation of genes directly
   associated with the TNF pathway, which is tightly associated with the
   action of thalidomide and its analogs as inhibitors of this cytokine
   ([235]21, [236]107–[237]111). In fact, our data are in line with a
   previous study where ENL patients were followed up during 21 days of
   thalidomide treatment. In that study, a modest increase in TNF serum
   levels was observed during treatment, even with evident improvement of
   skin lesions. These data together with observations of increased
   circulating levels of IL-12 and IL-2 and IFN-γ expression by T cells,
   suggests that thalidomide promotes a complex immunoregulatory response
   ([238]112).

   Finally, it is important to address the limitations of our study, which
   includes a modest sample size and intrinsic sex bias in the
   longitudinal study, due to restrictions of thalidomide administration
   to women within reproductive age. Moreover, the experimental design did
   not account for possible implications of MDT treatment in the
   transcriptional profile or CD177 serum levels of patients with ENL,
   although it was not found a clear distinguishable pattern among ENL
   patients that could be linked to different MDT status. Yet, further
   studies are advised to overcome these limitations and expand the
   current findings.

   In summary, despite the limitations, our data, together with published
   reports, support the protagonist role of neutrophils during ENL
   immunopathogenesis and highlight their participation in a systemic
   context. Although immunomodulation by thalidomide was found to be
   modest in the blood compartment after a short period of treatment in
   patients with ENL, the expression of genes with potential
   anti-inflammatory roles was detected. This suggests a systemic path to
   the resolution of ENL associated inflammation. Our data also revealed
   potential diagnostic markers and new targets for therapeutic
   intervention for ENL management.

Data availability statement

   The datasets presented in this study can be found in online
   repositories. The names of the repository/repositories and accession
   number(s) can be found below: [239]https://www.ncbi.nlm.nih.gov/geo/,
   [240]GSE198609.

Ethics statement

   The studies involving humans were approved by Instituto Oswaldo Cruz
   ethics committee (CEP Fiocruz/IOC). The studies were conducted in
   accordance with the local legislation and institutional requirements.
   The participants provided their written informed consent to participate
   in this study.

Author contributions

   TR: Writing – original draft, Methodology, Investigation, Formal
   analysis. TL-C: Writing – original draft, Methodology, Investigation,
   Formal analysis. IT: Writing – review & editing, Methodology,
   Investigation. MAM: Writing – review & editing, Investigation, Formal
   analysis. AD: Writing – review & editing, Methodology, Investigation.
   MS: Writing – review & editing, Methodology, Investigation. MF: Writing
   – review & editing, Investigation, Formal analysis. MK: Writing –
   review & editing, Investigation. FC: Writing – review & editing,
   Methodology, Investigation. MMM: Writing – review & editing, Resources,
   Project administration, Funding acquisition. AF: Writing – review &
   editing, Resources. AS: Writing – review & editing, Investigation,
   Formal analysis. MH: Writing – review & editing, Formal analysis. MBP:
   Writing – review & editing, Supervision, Project administration. JB:
   Writing – review & editing, Supervision, Resources, Funding
   acquisition. MOM: Writing – review & editing, Supervision, Funding
   acquisition. MCVP: Writing – review & editing, Supervision, Resources,
   Funding acquisition, Conceptualization. VS: Project administration,
   Writing – review & editing, Supervision, Resources, Funding
   acquisition, Conceptualization.

Acknowledgments