Intramuscular Boosting with hIFN-Alpha 2b Enhances BCGphipps-Induced Protection in a Murine Model of Leprosy

: Host immunity to Mycobacterium leprae encompasses a spectrum of mechanisms that range from cellular immunity-driven protection to damage associated with humoral immunity as in type-2 leprosy reactions. Although type I interferons (IFNs) participate in eliminating intracellular pathogens, their contribution to the production of antibodies and CD3 + FOXP3 + regulatory T cells (Tregs) in BCG vaccine-mediated protection in leprosy is unknown. BCGphipps (BCGph) priming followed by intramuscular hIFN- α 2b boost signiﬁcantly reduced lesion size and Mycobacterium lepraemurium growth in the skin. T follicular regulatory cells (TFR), a subset of Tregs induced by immunization or infection, reside in the germinal centers (GCs) and modulate antibody production. We found impaired Treg induction and improved GCs in draining lymph nodes of BCGph primed and hIFN- α 2b boosted mice. Moreover, these mice elicited signiﬁcant amounts of IL-4 and IL-10 in serum. Thus, our results support the adjuvant properties of hIFN- α 2b in the context of BCGph priming to enhance protective immunity against skin leprosy.


Introduction
Mycobacterium leprae and Mycobacterium lepromatosis, the etiological agents of human leprosy [1,2], affect the skin and peripheral nerves [3]. Although multidrug treatments (MDT) for leprosy are available, the number of new cases remains stable (700,000 per year). Furthermore, the lack of an effective diagnostic procedure makes tracking of M. leprae transmission difficult [4][5][6][7]. The genetic predisposition to early leprosy [8] emphasizes the need for an opportune diagnosis and development of low-cost immunotherapies to reduce leprosy incidence in developed countries [9,10]. The spectrum of clinical manifestations in leprosy correlates with the type of immune response elicited. For example, tuberculoid leprosy (TT), characterized by a dominant Th1 response, restricts bacilli growth via IFN-γdriven activation of microbicidal functions of macrophages. In contrast, the production of classic Th2 cytokines (IL-4 and IL-10) has been associated with loss of protection and enhanced bacillary growth in disseminated lepromatous leprosy (LL) [10]. There are also intermediate subtypes that represent a mixture of Th1 and Th2 responses [10].
Infection of mice with M. lepraemurium (MLM) [11,12] has been used to study immunological mechanisms of human leprosy [11][12][13], despite the fact that MLM does not show tropism for peripheral nerves [11,13] as M. leprae does [14]. Nevertheless, both mycobacteria are non-cultivable and grow slowly in the host, doubling their numbers in 1-3 weeks. Experimental subcutaneous infection with small doses of bacteria closely mimics the natural disease that usually occurs when disruption of barriers facilitates mycobacterial colonization, triggering an immune reaction at a peripheral level [13,[15][16][17]. While cellmediated immunity (CMI) is impaired, humoral immunity is practically unaffected in human and murine leprosy [18]. In addition, Th1-or Th2-cells, other T lymphocytes, shape the host immunity to M. leprae infection [18,19].
It is intriguing how M. leprae influences host immunity in the different subtypes of mycobacterial disease. One hypothesis is that the pathogen downregulates host humoral and cellular immunity to infect the skin and peripheral nerves [3,14,15]. There is also evidence that the bacillus impairs macrophage activation. It also decreases the expression of TNF-alpha and molecules such as prostaglandins which induce the production of IL-1β and reactive oxygen and nitrogen intermediates (ROI and RNI), critical in the mycobactericidal functions of macrophages [20][21][22].
Humoral responses generated in the GCs of secondary and tertiary lymphoid tissues participate in the protective immunity against pathogens [23][24][25][26]. A recent report on leprosy highlighted the critical role of Tregs in the inhibition of Th1 and Th17 cell responses [19]. Indeed, the balance between these subsets of T cells at the GCs [27][28][29] dictates the protective or pathological outcomes of infectious and autoimmune diseases [30,31] and is critical for the development of therapies to induce protective immunity [26,27]. In particular, T helper follicular cells (TFH) and especially T follicular regulatory cells (TFR) modulate the GC response to induce long-lasting humoral immune responses while preventing the production of autoantibodies [32,33]. Given that type I IFN is associated with the generation of both TFH and Tregs, it emerges as a promising adjuvant to stimulate TFH responses in the GCs in the context of BCG vaccination [34].
We previously reported the protection conferred by successive intramuscular IFN boosting in mice vaccinated with Mycobacterium bovis BCG and challenged with MLM [42] or against aerosolized M. tuberculosis challenge [41]. In mice challenged with MLM, protection correlated with enhanced production of iNOS, nitric oxide, and IgG3 antibodies [42]. In the context of the M. tuberculosis challenge, the protected mice elicited a Th1 response [41]. In a more recent study, type I IFNs modulated autophagy and metabolism in leprosy [50]. We decided in this study to further investigate the protection against murine leprosy induced by vaccination with M. bovis BCGphipps (BCGph) and a pharmaceutical presentation of human IFN-α 2b [51]. In recent studies, a recombinant BCG that expresses human IFN-α 2b was used to successfully treat bladder cancer [51][52][53][54][55]. Consistent with these data, our goal in this study was to analyze the protection induced by BCGph priming and hIFN-α 2b boosting in a murine model of leprosy. Altogether, our results envisage the potential of BCGph subcutaneous priming and intramuscular hIFN-α 2b boosting to induce protective immunity against human leprosy.

Animals
Specific pathogen-free Balb/c mice were bred and housed in the Animal facilities of the Escuela Nacional de Ciencias Biológicas, Instituto Politécnico Nacional in Mexico City.
All animal experiments were performed after approval (07 March 2014) by the Institutional Animal Care and Management Committee (CIE-ENCB ZOO 016/2014).

Microorganisms
The BCG Phipps strain (ATCC ® 35744™) (BCGph) is a variant of the BCG strain that has lost the region RD1 which encodes ESAT-6 and CFP10, with the lowest number of T cell epitopes, 316, versus BCG Tokyo, 359, and BCG Pasteur, 331, (Zhang et al., 2013) [51] and was provided by Dr. Hernández-Pando (INCMNSZ, Mexico City, MX). Vaccine suspensions were then stored at -80 • C until use. Mycobacterium lepraemurium (MLM) non-cultivable strain was isolated from the spleen of mice bearing a four-month infection via a procedure described by Prabhakaran et al., 1976 [56], followed by purification using the percoll isolation procedure described by Draper P., 1980 [57]. Briefly, 4 g of splenic tissue were suspended in 20 mL of 0.2 M sucrose and ground in a glass Potter-Elvehjem type homogenizer. The resulting suspension was centrifuged for 20 min at 150× g to remove cell debris (Sorvall RV5B, rotor HB4) (Sorvall Instruments, Wilmington, DE, USA). Then, 9 mL of the supernatant was overlaid on 12 mL of 0.3 M sucrose, and the tubes were centrifuged at 3500× g for 10 min at 4-10 • C (Sorvall RC5B). The resulting bacilli-rich pellet was resuspended in 20 mL of 0.2 M sucrose, and 9 mL aliquots were overlaid on 12 mL of 1.5 M KCl. Then, the tube was centrifuged at 4 • C for 10 min at 3500× g. The bacilli were collected, washed three times with phosphate-buffered-saline (PBS) (0.01 M Na/K phosphate, 0.15 M NaCl, pH 7.4), and resuspended in 40 mL of a solution containing a mixture of Percoll (3 parts) and 0.1% Tween 80 (7 parts). The suspension was centrifuged at 23,000× g. The bacillary pellet was collected and washed five times with PBS until eliminating the Percoll. The purity of the bacillary preparation was confirmed by Ziehl-Neelsen staining. The purified bacillary suspension was prepared in synthetic 7H9 Middlebrook broth-OADC medium (DIFCO, Detroit, MI, USA) (7H9-OADC medium) and quantified via a nephelometric reference curve prepared with known quantities of bacteria. The bacillary suspension was aliquoted and frozen at −20 • C, until ready for use.

M. lepraemurium Challenge
For the challenge experiments, mice were inoculated intradermally with 2 × 10 6 viable MLM in 20 µL of PBS. Then, the mice were housed in standard conditions and supplied with purified water and Purina rodent chow chips (Cuautitlan Izcalli, Mexico) ad libitum. The infected mice were monitored for eight weeks, the time points when lesions were evident in the skin of infected animals. Disease progression was variable in the mice and correlated with experimental therapies (PBS, BCGph, or hIFN-α 2b).

Measurements of NO in the Sera of Mice
Briefly, sera were treated with ethanol (1:1) to precipitate proteins and other interfering factors. After centrifugation at 3000 rpm for 20 min, 20 µL of the recovered serum was placed into wells of an ELISA plate together with 100 µL of the Griess reagent, and 80 µL of distilled water. The Griess reagent is a volume-to-volume mixture of solution A (2% sulfanilamide in 5% phosphoric acid) and solution B (0.1% N-(1-naphthyl) ethylenediamine, dihydrochloride. After a 30 min incubation at 25 • C, the pink color developed was registered in an ELISA reader at 595 nm. The values were extrapolated to a standard curve of sodium nitrite in the range 1-100 µM. Five-µm formalin-fixed paraffinized sections were incubated at 60 • C overnight for deparaffinization. Tissue sections were quickly transferred to xylenes (247642-2.5L, SIGMA) and hydrated by sequentially immersing slides into absolute alcohol, 96% alcohol, 70% alcohol, and water. Slides were immersed in an antigen retrieval solution, boiled for 30 min, and cooled down for 10 min at room temperature (RT). Slides were rinsed several times in water and transferred to PBS. Non-specific binding was blocked with 5% normal donkey serum in PBS containing 0.1% Tween 20 and 0.1% Triton-X-100, for 30 min at RT in a humid chamber. Primary antibodies were added to slides and incubated overnight in a moist chamber at RT. Slides were gently washed in PBS, and fluorescently labeled secondary antibodies were incubated for 2 h, at RT, or overnight in a humid chamber. Finally, slides were rinsed for 1 h in PBS and mounted with Vectashield antifade mounting media with DAPI (H-1200, Vector Laboratories, Burlingame, CA, USA). Pictures were taken with a Zeiss Axioplan microscope. All germinal centers (GCs) in a lymph node section were measured with an automated tool of the Carl Zeiss microscope, and the average GC size was calculated using the Prism software. CD3 + FoxP3 + Tregs were counted in 5 random 200 × fields per tissue to calculate the average number of positive cells per field.

Statistical Analysis
Statistically significant differences among groups were calculated by one-way ANOVA, Newman-Keuls multiple comparisons, unpaired/paired two-tailed Student's t-test, pvalues < 0.05 were considered to be statistically significant. Statistical differences were calculated with Graph Pad Prism.

BCGphipps (BCGph)-Priming Followed by Successive Intramuscular hIFN-α 2b Boosting Protects Balb/c Mice from Intradermal Challenge with M. lepraemurium
We previously developed a prime-boost protocol based on successive intramuscular boosting of IFNα after BCG vaccination of mice [42] with promising protection results. Recent studies used a BCG recombinant that expresses human IFN-α 2b to treat bladder cancer [51,53]. Therefore, in the present study, our goal was to analyze the protection induced by BCGph priming followed by hIFN-α 2b boosting in a murine model of leprosy, following the scheme depicted in Figure 1.
ANOVA, Newman-Keuls multiple comparisons, unpaired/paired two-tailed Student's ttest, p-values < 0.05 were considered to be statistically significant. Statistical differences were calculated with Graph Pad Prism.

BCGphipps (BCGph)-Priming Followed by Successive Intramuscular hIFN-α 2b Boosting Protects Balb/c Mice from Intradermal Challenge with M. lepraemurium
We previously developed a prime-boost protocol based on successive intramuscular boosting of IFNα after BCG vaccination of mice [42] with promising protection results. Recent studies used a BCG recombinant that expresses human IFN-α 2b to treat bladder cancer [51,53]. Therefore, in the present study, our goal was to analyze the protection induced by BCGph priming followed by hIFN-α 2b boosting in a murine model of leprosy, following the scheme depicted in Figure 1.
Briefly, mice were intradermally challenged with MLM, eight weeks after the third boost with 300 IU of hIFN-α 2b. Disease progression was monitored during the next eight weeks. At the end of this period, we collected inguinal lymph nodes to detect GCs and enumerate Tregs and other cell populations by immunofluorescence. We also used immunofluorescence to visualize iNOS production by activated macrophages, location of nitrotyrosine (NT) residues, and accumulation of Gr-1+ myeloid cells in the skin of our experimental mice. Additionally, we collected the serum to measure the cytokines and antibodies to glycolipids of MLM by ELISA (Figure 1). Groups of Balb/c mice (n = 6) were subcutaneously vaccinated with a low dose of BCGph (8 × 10 3 CFU). Mice from each group were intramuscularly boosted on consecutive days (30, 31, and 32 days after vaccination) with PBS or hIFN-α 2b. Mice were challenged with 2 × 10 6 viable MLM eight weeks after the last boosting. After that, mice were sacrificed, and biospecimens were collected eight weeks post-challenge. The development of skin lesions was monitored and measured as described in Material and Methods. Groups of Balb/c mice (n = 6) were subcutaneously vaccinated with a low dose of BCGph (8 × 10 3 CFU). Mice from each group were intramuscularly boosted on consecutive days (30, 31, and 32 days after vaccination) with PBS or hIFN-α 2b. Mice were challenged with 2 × 10 6 viable MLM eight weeks after the last boosting. After that, mice were sacrificed, and biospecimens were collected eight weeks post-challenge. The development of skin lesions was monitored and measured as described in Material and Methods.
Briefly, mice were intradermally challenged with MLM, eight weeks after the third boost with 300 IU of hIFN-α 2b. Disease progression was monitored during the next eight weeks. At the end of this period, we collected inguinal lymph nodes to detect GCs and enumerate Tregs and other cell populations by immunofluorescence. We also used immunofluorescence to visualize iNOS production by activated macrophages, location of nitrotyrosine (NT) residues, and accumulation of Gr-1+ myeloid cells in the skin of our experimental mice. Additionally, we collected the serum to measure the cytokines and antibodies to glycolipids of MLM by ELISA (Figure 1).
First, to evaluate the protective effect of BCGph priming and hIFN-α 2b boosting against the intradermal challenge of mice with MLM, we measured the skin lesions and visualized bacilli colonization of the skin by the Ziehl-Neelsen stain (Figure 2A-E). The skin of unprimed mice that received only vehicle (PBS), without MLM challenge, showed no bacilli (Figure 2A). The mice inoculated with vehicle and challenged with MLM showed extensive amounts of MLM bacilli in the skin ( Figure 2B), while priming mice with BCGph led to a reduction in the mycobacterial growth ( Figure 2C). Successive hIFN-α 2b administration also reduced mycobacterial growth in the skin as compared with mice challenged with MLM ( Figure 2D). The mice primed with BCGph and boosted with hIFN-α 2b had the lowest mycobacterial growth ( Figure 2E). Consistent with these data, the healthy mice challenged with MLM that received only vehicle (PBS) developed the most extensive skin lesions (35 ± 9.9 mm 2 ), while the mice that were BCGph-primed had skin lesions of similar size to the ones developed in the mice inoculated with the vehicle and challenged with MLM (13.5 ± 4.24 mm 2 ) ( Figure 2F). The mice subcutaneously injected with hIFN-α 2b showed smaller skin lesions (20 ± 1.77 µm 2 ). Of note, mice that were BCGph primed and boosted with hIFN-α 2b showed a significant reduction in skin lesion size (5 ± 7.07 mm 2 ) (p < 0.05) ( Figure 2F). visualized bacilli colonization of the skin by the Ziehl-Neelsen stain (Figure 2A-E). The skin of unprimed mice that received only vehicle (PBS), without MLM challenge, showed no bacilli (Figure 2A). The mice inoculated with vehicle and challenged with MLM showed extensive amounts of MLM bacilli in the skin ( Figure 2B), while priming mice with BCGph led to a reduction in the mycobacterial growth ( Figure 2C). Successive hIFNα 2b administration also reduced mycobacterial growth in the skin as compared with mice challenged with MLM ( Figure 2D). The mice primed with BCGph and boosted with hIFNα 2b had the lowest mycobacterial growth ( Figure 2E). Consistent with these data, the healthy mice challenged with MLM that received only vehicle (PBS) developed the most extensive skin lesions (35 ± 9.9 mm 2 ), while the mice that were BCGph-primed had skin lesions of similar size to the ones developed in the mice inoculated with the vehicle and challenged with MLM (13.5 ± 4.24 mm 2 ) ( Figure F). The mice subcutaneously injected with hIFN-α 2b showed smaller skin lesions (20 ± 1.77 µm 2 ). Of note, mice that were BCGph primed and boosted with hIFN-α 2b showed a significant reduction in skin lesion size (5 ± 7.07 mm 2 ) (p < 0.05) ( Figure 2F). (F) Efficient protection is equivalent to either the absence of skin lesions or a significant reduction in skin lesion size, expressed in mm 2 , n = 6 mice/group. * p < 0.05; ** p < 0.005; *** p < 0.0005 was considered to be statistically significant. with BCGph, boosted with hIFN-α 2b, and challenged with MLM; (F) Efficient protection is equivalent to either the absence of skin lesions or a significant reduction in skin lesion size, expressed in mm 2 , n = 6 mice/group. * p < 0.05; ** p < 0.005; *** p < 0.0005 was considered to be statistically significant.

Minimal Skin Pathology in Mice Primed with BCGphipps and Boosted with hIFNα 2b Correlates with Decreased Nitrotyrosine Accumulation and Impaired Neutrophil Infiltration
Nitric oxide is critical for the killing of intracellular pathogens. However, excessive production of nitric oxide, superoxide radical (O 2 − , and hydrogen peroxide leads to the generation of toxic and unstable peroxynitrite, a molecule that reacts with tyrosine residues on proteins to form the stable NT [56][57][58][59][60]. Although peroxynitrite kills mycobacteria, it also damages immune cells [57][58][59]. Thus, we decided to search for iNOS producing cells, NT, and neutrophils in the skin of our experimental mice by immunofluorescence. We saw background signals for iNOS and NT and very few Gr-1 + cells in the skin of mice receiving only PBS ( Figure 3A). Increased mycobacterial load and enlarged necrotic lesions in the skin correlated with the accumulation of NT residues in the necrotic skin area [58,59] of unvaccinated ( Figure 3B) and BCGph primed mice challenged with MLM ( Figure 3C). Consistent with a more balanced local inflammatory response and amelioration of skin pathology, NT and neutrophil infiltration were decreased in mice receiving successive inoculations with hIFN-α 2b ( Figure 3D) or mice BCGph primed and boosted with hIFNα 2b ( Figure 3E). Impaired infiltration with Gr-1 + neutrophils and reduced NT areas coincided with the smallest inflammatory area in histological skin sections of mice primed with BCGph and boosted with hIFN-α 2b ( Figure 3F). [58,59] of unvaccinated ( Figure 3B) and BCGph primed mice challenged with MLM (Figure 3C). Consistent with a more balanced local inflammatory response and amelioration of skin pathology, NT and neutrophil infiltration were decreased in mice receiving successive inoculations with hIFN-α 2b ( Figure 3D) or mice BCGph primed and boosted with hIFN-α 2b ( Figure 3E). Impaired infiltration with Gr-1 + neutrophils and reduced NT areas coincided with the smallest inflammatory area in histological skin sections of mice primed with BCGph and boosted with hIFN-α 2b ( Figure 3F).

Anti-Inflammatory Cytokines Are Increased in the Serum of Mice Primed with BCGphipps and Boosted with hIFN-α 2b
Pro-and anti-inflammatory cytokines were measured in the serum of Balb/c mice after challenge with MLM. Consistent with the immunosuppressive state caused by MLM, the concentrations of proinflammatory cytokines were low in the serum of Balb/c mice inoculated with PBS and challenged with MLM (IFN-γ, 625 ± 442 pg/mL; TNF-α, 697 ± 436 pg/mL; IL-6, 68 ± 48 pg/mL; IL-17, 95 ± 67 pg/mL). In contrast, they had moderate levels of anti-inflammatory cytokines in the serum (IL-10, 1763 ± 124 pg/mL and IL-4, 1250 ± 173 pg/mL). Although mice were protected by BCGph priming and successive boosting with hIFN-α 2b, the systemic amounts of proinflammatory cytokines were not statistically different from those in mice inoculated with PBS and challenged with MLM. Unexpectedly, the concentration for IL-4 (1875 ± 352 pg/mL, p < 0.05), and IL-10 (4250 ± 271 pg/mL) were significantly increased in the serum of these mice as compared with mice inoculated with PBS and challenged with MLM (Table 1). These results suggest that our vaccination protocol might be activating anti-inflammatory pathways to prevent tissue damage.

BCGphipps Priming and hIFN-α 2b Boosting Influences GCs Responses in Mouse DLNs
There are special Tregs located in the GCs known as T follicular regulatory cells (TFR). They modulate both the GCs responses and prevent the generation of autoreactive plasma cells [28,33]. The decreased Tregs in the DLNs suggested that GCs responses could be unleashed in mice primed with BCGph and boosted with hIFN-α 2b. Consistent with the absence of antigenic stimulation, we did not detect GCs in the DLNs of uninfected mice ( Figure 5A). Small GCs were present in the DLNs of unprimed mice, challenged with MLM ( Figure 5B).
Interestingly, mice primed with BCGph had significantly larger GCs than mice challenged with MLM ( Figure 5C). The mice immunized with hIFN-α 2b ( Figure 5D) had GCs with a comparable size to those in mice challenged with MLM. The mice primed with BCGph and boosted with hIFN-α 2b had GCs with an intermediate size ( Figure 5E). To have a more quantitative assessment of the GCs responses in the DLNs, we calculated the average GC size in individual DLN sections. Consistent with our observations, we found a significant reduction in the average GC size in the DLNs after hIFN-α 2b injection and in the BCGph priming/hIFN-α 2b boosting protocol ( Figure 5E). (F) morphometric analysis shows a significant reduction in Tregs in the DLNs of BCGph primed/hIFN-α 2b boosted mice. Scale bar = 100 µm, n = 6 mice/group * p < 0.05; ** p < 0.005; *** p < 0.0005 was considered to be statistically significant.

BCGphipps Priming and hIFN-α 2b Boosting Influences GCs Responses in Mouse DLNs
There are special Tregs located in the GCs known as T follicular regulatory cells (TFR). They modulate both the GCs responses and prevent the generation of autoreactive plasma cells [28,33]. The decreased Tregs in the DLNs suggested that GCs responses could be unleashed in mice primed with BCGph and boosted with hIFN-α 2b. Consistent with the absence of antigenic stimulation, we did not detect GCs in the DLNs of uninfected mice ( Figure 5A). Small GCs were present in the DLNs of unprimed mice, challenged with MLM ( Figure 5B). (F) GC were measured in individual DLNs of experimental mice to calculate the average GC size. Scale bar = 100 µm. n = 6 mice/group. * p < 0.05; ** p < 0.005; **** p < 0.0005 was considered to be statistically significant.

Comparable Levels of Glycolipid-Specific Antibodies in Mice Challenged with M. lepraemurium after BCGphipps Priming and hIFN-α 2b Boosting
According to the differences observed in GC formation and organization, we measured the concentration of antibodies specific for glycolipids extracted from MLM. Globally, our results show that the systemic amounts of different classes and isotypes of glycolipid-specific antibodies were not affected by vaccination, successive administration of hIFN-α2b, or the BCGph priming/hIFN-α 2b protocol (Table 2). Mice boosted only with hIFN-α 2b elicited a higher amount of IgG and IgG1 antibodies in the serum. However, IgG and Ig G1 antibody levels were not statistically significant as compared with the other experimental groups (Table 2). Interestingly, mice primed with BCGph had significantly larger GCs than mice challenged with MLM ( Figure 5C). The mice immunized with hIFN-α 2b ( Figure 5D) had GCs with a comparable size to those in mice challenged with MLM. The mice primed with BCGph and boosted with hIFN-α 2b had GCs with an intermediate size ( Figure 5E). To have a more quantitative assessment of the GCs responses in the DLNs, we calculated the average GC size in individual DLN sections. Consistent with our observations, we found a significant reduction in the average GC size in the DLNs after hIFN-α 2b injection and in the BCGph priming/hIFN-α 2b boosting protocol ( Figure 5E).

Comparable Levels of Glycolipid-Specific Antibodies in Mice Challenged with M. lepraemurium after BCGphipps Priming and hIFN-α 2b Boosting
According to the differences observed in GC formation and organization, we measured the concentration of antibodies specific for glycolipids extracted from MLM. Globally, our results show that the systemic amounts of different classes and isotypes of glycolipidspecific antibodies were not affected by vaccination, successive administration of hIFN-α2b, or the BCGph priming/hIFN-α 2b protocol (Table 2). Mice boosted only with hIFN-α 2b elicited a higher amount of IgG and IgG1 antibodies in the serum. However, IgG and Ig G1 antibody levels were not statistically significant as compared with the other experimental groups ( Table 2).

Discussion
In the present study, we have shown that our prime-boosting protocol based on subcutaneous BCGphipps priming and successive intramuscular boosting with increasing amounts of hIFN-α 2b protected mice from intradermal challenge with MLM.
The prime-boosting protocol prevented MLM growth in the skin and ameliorated local pathology by decreasing NT production and neutrophilic inflammation ( Figure 3E). It also resulted in the induction of anti-inflammatory cytokines (IL-4 and IL-10) that likely counteracted excessive inflammation in the skin and manifested systemically. Unexpectedly, although the therapeutic regimen decreased CD3 + FOXP3 + Tregs in the DLNs ( Figure 4E), GC responses were not drastically enhanced ( Figure 5E), and the serum concentrations of glycolipid-specific antibodies were similar in all the experimental groups except in the mice inoculated only with hIFN-α 2b (Table 2). Nevertheless, our experimental data support the potential use of BCG priming adjuvanted with intramuscular IFNα 2b boosting, as a promising immunotherapy that deserves exploration in human leprosy.
The clinical manifestations of leprosy correlate with the immunological spectrum detected in patients and range from LL to TT [2,4,14,18]. In general, host immunity to M. leprae is characterized mainly by a polarized T helper response (Th1/Th2). On one side of the spectrum, although LL is characterized by anergy, there is still some reactivity to mycobacterial antigens and the production of IL-4 and IL-10. In contrast, in TT, a predominant Th1 response leads to the IFN-γ driven activation of macrophages, which accelerates the elimination of the bacilli. Mice infected with MLM typically develop lepromatous-like leprosy without a clear dominance for humoral or cellular immunity [18][19][20].
Protective antibodies elicited by infection or vaccination [63] are produced by a process called affinity maturation, which takes place in germinal centers formed in secondary (spleen, lymph nodes) or tertiary lymphoid organs (TLO) generated in peripheral tissues [23,25]. It is precisely in the GCs where the production of memory B cells and long-lived antibody-secreting plasma cells contribute to the generation of long-term immunity to future encounters with pathogens. The efficient induction of GC responses depends on the collaboration of multiple cell types, including B cells, T follicular helper cells (TFH), tingible body macrophages, and follicular dendritic cells (FDCs) [64]. Indeed, vaccines trigger successful GC responses, and ectopic germinal centers in lymph nodes [24,25,31,49]. Moreover, T follicular helper cells and T follicular regulatory cells [27][28][29][30] maintain the delicate balance between the induction of effective antibody responses and autoimmunity. Although we did not rule out autoimmunity induction in our study, we still focused on the effects of the prime-boost protocol with BCGph and hIFN-α 2b on the Tregs from the DLNs. Protected mice had significantly fewer CD3 + FOXP3 + Tregs in their DLNs. However, further studies are required to determine if some of these Tregs are T follicular helper cells that regulate GC reactions [27][28][29][30]. These results might imply that efficient induction of protective immunity combined with a slight decrease of Tregs in the DLNs is required to control mycobacterial growth and to prevent pathologic inflammation in the skin.
Previous studies have reported the systemic and mucosal effects of type I IFN on T follicular helper cells and T follicular regulatory cells [19,32,64]. Indeed, type I IFN indirectly increases TFH in lymph nodes by stimulating T cell priming by dendritic cells [34]. Type I IFN also promotes isotype switch in GCs. In addition, a recent study showed that type I IFN triggers the CXCL13-driven formation of TLO in the lungs [49]. Thus, the induction of CXCL13 by type I IFN could enhance the recruitment of naïve B cells to secondary lymphoid organs and increase the production of antibody-secreting plasma cells, globally improving long-lasting memory and plasma cells immunological responses [23][24][25]64], particularly in the presence of adjuvants like poly IC [65]. In addition, type I IFN administration induces GC formation in the lungs of patients afflicted by autoimmune diseases [31] or tuberculosis [66,67]. Thus, CXCL13 blockade could help treat these chronic inflammatory disorders [49]. In this context, we observed a very mild decrease in the GC size in DLNs of protected prime-boosted mice ( Figure 5E). However, additional research is needed to better define the correlation between GC formation and the protection induced in those prime-boosted mice. Under the experimental settings of priming with BCGph and intramuscular hIFN-α 2b boosting (51), the mice in our study experienced higher protection by the BCGph-hIFN-α 2b (80%) administration than by BCGph alone (60%), as determined by the absence or the small size of the skin lesions ( Figure 2) and a comparable quantity of IgG and IgG1 antibodies specific for glycolipids of MLM (Table 2). Our results contrast with the increased amount of IgG in mice protected by vaccination with a different strain of BCG (Pasteur) [41]. These data suggest that the type of BCG strain used to formulate the vaccine could influence the outcome of the humoral immune response against MLM or M. leprae [51]. Furthermore, several studies in tuberculosis have shown the potential and beneficial effects of the heterologous prime-boost strategy in the induction of protective immune responses [68][69][70]. Moreover, it is well known that the type and the route of immunization influence the magnitude of the antibody and cellular immune responses. Surprisingly, protected mice show significantly higher IL-4 and IL-10 in the serum as compared with mice treated only with PBS, BCGph, or hIFN-α 2b alone (Table 1). How anti-inflammatory cytokines influence protective immunity in leprosy is currently under investigation in our laboratory. However, several studies have shown that efficient cellular immunity to mycobacteria relies on a delicate balance between protection and prevention of pathology [71]. In addition, there is evidence of induction of both cytokine types in other experimental models [72][73][74][75].
Our collective results suggest that the success of our prime-boost protocol depends on the induction of a balanced immune response that efficiently eliminates MLM in the skin and limits tissue damage by inducing anti-inflammatory cytokines (IL-4 and IL-10) that likely protect immune cells [76,77]. Indeed, IL-10 produced by induced Tregs modulates inflammation at mucosal surfaces, specifically those related to inflammatory bowel disease [77,78]. These findings warrant a detailed kinetic study of B cell activation, GC formation and regulation, T follicular regulatory cells cells generation, and NF-κB activation in our prime-boost system with BCGph and hIFN-α 2b followed by MLM infection to have a more mechanistic understanding of the roles of immune cells and NF-κB pathways in the induction of protective immunity in the skin and lymph nodes after MLM challenge.

Conclusions
From these data, we can conclude that the successive intramuscular boosting with hIFN-α 2b enhances M. bovis BCG vaccination, protecting against intradermal challenge with MLM and revealing the promising adjuvant properties of human IFN-a 2b in the context of BCG vaccination for preventing leprosy infections in endemic areas. To the best of our knowledge, this is the first report of the therapeutic efficacy of BCGphipps in combination with a pharmaceutical presentation of hIFN-α 2b in a preclinical model of leprosy-like disease.