HIF-α Activation Impacts Macrophage Function during Murine Leishmania major Infection

Leishmanial skin lesions are characterized by inflammatory hypoxia alongside the activation of hypoxia-inducible factors, HIF-1α and HIF-2α, and subsequent expression of the HIF-α target VEGF-A during Leishmania major infection. However, the factors responsible for HIF-α activation are not known. We hypothesize that hypoxia and proinflammatory stimuli contribute to HIF-α activation during infection. RNA-Seq of leishmanial lesions revealed that transcripts associated with HIF-1α signaling were induced. To determine whether hypoxia contributes to HIF-α activation, we followed the fate of myeloid cells infiltrating from the blood and into hypoxic lesions. Recruited myeloid cells experienced hypoxia when they entered inflamed lesions, and the length of time in lesions increased their hypoxic signature. To determine whether proinflammatory stimuli in the inflamed tissue can also influence HIF-α activation, we subjected macrophages to various proinflammatory stimuli and measured VEGF-A. While parasites alone did not induce VEGF-A, and proinflammatory stimuli only modestly induced VEGF-A, HIF-α stabilization increased VEGF-A during infection. HIF-α stabilization did not impact parasite entry, growth, or killing. Conversely, the absence of ARNT/HIF-α signaling enhanced parasite internalization. Altogether, these findings suggest that HIF-α is active during infection, and while macrophage HIF-α activation promotes lymphatic remodeling through VEGF-A production, HIF-α activation does not impact parasite internalization or control.


Introduction
Leishmaniasis is an inflammatory disease caused by vector-borne, obligate intracellular protozoan parasites of the genus Leishmania. Leishmaniasis is endemic in many developing countries in tropical and subtropical regions of the globe, where up to 1.7 million new cases in 98 countries occur annually [1,2]. Leishmania manifests in three forms: (1) visceral leishmaniasis (VL), which is systemic and fatal if untreated, (2) cutaneous leishmaniasis (CL), which causes nodules, papules, and lesions on the surface of the skin, and (3) mucocutaneous leishmaniasis (MCL), wherein parasites from primary cutaneous lesions spread to other parts of the body [3]. Infections with different species of Leishmania manifest in different clinical forms. In particular, Leishmania major is responsible for CL in the Eastern Mediterranean and North Africa, where the disease is considered a major public health problem and accounts for 70% of CL cases worldwide [4]. The burden on these countries, already high due to current socioeconomic conditions, limited resources and medical infrastructure, and civil strife, is further exacerbated by the lack of a vaccine and ineffective chemotherapeutic treatment for the disease [3,4]. At 4 weeks p.i., when mice were presented with lesions, we performed RNA-Seq o fected and control uninfected ears of naïve mice. Differentially expressed genes (D between infected and naïve ears were identified by limma voomWithQualityWeigh R, and pathway enrichment analysis of DEGs was performed using EGSEA agains KEGG database. The gene expression profiles derived from the RNA-Seq data were sidered statistically significant if fold change >2 and p < 0.05. Hierarchical clustering ysis revealed that many transcripts that were significantly induced following L. majo fection are associated with the HIF-1α signaling pathway, including Egln1 (Phd2), E (Phd3), Eno2, Hif1a, Hk2, Hk3, Ldha, Nos2, Vegfa, and Vhl ( Figure 1). Altogether, our g transcriptomic analysis suggests that HIF-1α signaling is increased during L. major i tion in vivo, consistent with findings that leishmanial lesions are hypoxic. Figure 1. The HIF-1α signaling pathway is elevated following infection with L. major parasi vivo. C57BL/6 mice were infected with 2 × 10 6 L. major metacyclic promastigote parasites intr mally in the ear. At 4 weeks p.i., naïve control (N = 4) and infected (N = 6) ears were subject RNA-Seq analysis. Hierarchical clustering of the expression profile was performed. Heat map dicate fold changes in gene expression in L. major-infected ears with >2-fold (red) or <2-fold ( compared to naïve controls. The HIF-1α signaling pathway was obtained using the KEGG pat analysis database. Relative expression was normalized with the Z score.

Myeloid Cells Experience Inflammatory Hypoxia in Leishmanial Lesions
We hypothesize that inflammatory hypoxia further potentiates chronic inflamm in leishmanial lesions. We previously reported that inflammatory monocytes, whic massively recruited to the site of infection following L. major inoculation, are less hyp at the cellular level compared to macrophages in the skin during infection [9]. How it was not clear whether this was a cell-intrinsic feature that distinguishes monocytes macrophages, or if the tissue microenvironment imprints a hypoxic signature upon t cells. Therefore, we utilized an adoptive transfer system in which CD11b + cells iso from the bone marrow and spleens of CD45.1 donor mice were injected into CD45. cipient mice that were infected 4 weeks prior to the cell transfer with L. major para Infected mice were then treated with pimonidazole to determine which cells experi Figure 1. The HIF-1α signaling pathway is elevated following infection with L. major parasites in vivo. C57BL/6 mice were infected with 2 × 10 6 L. major metacyclic promastigote parasites intradermally in the ear. At 4 weeks p.i., naïve control (N = 4) and infected (N = 6) ears were subjected to RNA-Seq analysis. Hierarchical clustering of the expression profile was performed. Heat maps indicate fold changes in gene expression in L. major-infected ears with >2-fold (red) or <2-fold (blue) compared to naïve controls. The HIF-1α signaling pathway was obtained using the KEGG pathway analysis database. Relative expression was normalized with the Z score.

Myeloid Cells Experience Inflammatory Hypoxia in Leishmanial Lesions
We hypothesize that inflammatory hypoxia further potentiates chronic inflammation in leishmanial lesions. We previously reported that inflammatory monocytes, which are massively recruited to the site of infection following L. major inoculation, are less hypoxic at the cellular level compared to macrophages in the skin during infection [9]. However, it was not clear whether this was a cell-intrinsic feature that distinguishes monocytes from macrophages, or if the tissue microenvironment imprints a hypoxic signature upon these cells. Therefore, we utilized an adoptive transfer system in which CD11b + cells isolated from the bone marrow and spleens of CD45.1 donor mice were injected into CD45.2 recipient mice that were infected 4 weeks prior to the cell transfer with L. major parasites. Infected mice were then treated with pimonidazole to determine which cells experience hypoxic conditions at the cellular level ( Figure 2A). In most studies, hypoxia is detected at the tissue level; however, this Pathogens 2021, 10, 1584 4 of 18 adoptive transfer model examines hypoxia at a higher resolution and determines if monocytes can become hypoxic as they enter inflamed tissues at the cellular level. Using this system, coupled with flow cytometry, we found that CD45.1 + CD11b + cells were detectable in the lesions of infected CD45.2 + mice, suggesting that myeloid cells are actively recruited to the site of infection ( Figure 2B). Further analysis of the transferred myeloid cells from the donor mouse present at the site of infection showed that CD45.1 + cells displayed a CD11b + Ly6G + neutrophil or CD11b + Ly6G − CD64 + macrophage phenotype. Confirming previous results, these data suggest that monocytes differentiate into macrophages following Leishmania infection [19][20][21]. The amount of hypoxia staining based on the pimonidazole median fluorescence intensity (MFI) for each myeloid cell type from donor CD45.1 + cells as well as the endogenous host CD45.2 + cells was investigated ( Figure 2C). These studies revealed that CD11b + Ly6G + neutrophils exhibited lower levels of pimonidazole staining compared to macrophages in both the CD45.2 + host and CD45.1 + transferred cells. Importantly, macrophages derived from CD45.1 + donor mice possessed similar pimonidazole MFIs compared to endogenous CD45.2 + macrophages ( Figure 2D). These data suggest that monocyte-derived macrophages recruited from the circulation experience hypoxic conditions in inflamed lesions, which may impact their metabolic reprogramming and thus the differentiation or function of these cells.
sues at the cellular level. Using this system, coupled with flow cytom CD45.1 + CD11b + cells were detectable in the lesions of infected CD45. that myeloid cells are actively recruited to the site of infection ( Figure 2 of the transferred myeloid cells from the donor mouse present at t showed that CD45.1 + cells displayed a CD11b + Ly6G + neutrophil or C macrophage phenotype. Confirming previous results, these data sugg differentiate into macrophages following Leishmania infection [19][20][21]. poxia staining based on the pimonidazole median fluorescence inten myeloid cell type from donor CD45.1 + cells as well as the endogenou was investigated ( Figure 2C). These studies revealed that CD11b + Ly hibited lower levels of pimonidazole staining compared to macrop CD45.2 + host and CD45.1 + transferred cells. Importantly, macroph CD45.1 + donor mice possessed similar pimonidazole MFIs compar CD45.2 + macrophages ( Figure 2D). These data suggest that monoc phages recruited from the circulation experience hypoxic conditions i which may impact their metabolic reprogramming and thus the differe of these cells.  Following L. major infection, monocytes migrate into lesions and exhibit a hypoxic signature, but the length of time in the inflamed tissue may influence the hypoxic signature of the cell, further impacting myeloid cell differentiation and function. To determine if increased time in the lesion enhances hypoxia at the cellular level, we investigated myeloid cells at different stages of their transition from monocytes to macrophages based on the downregulation of Ly6C (an inflammatory monocyte marker) and upregulation of CD64 (a macrophage marker) ( Figure 3A). This strategy of examining myeloid cells that are actively differentiating from Ly6C hi CD64 − inflammatory monocytes to Ly6C − CD64 + macrophages, coupled with pimonidazole staining, determined that hypoxia occurred at the cellular level across the stages of transition. Confirming previous work [9], we found that Ly6C hi CD64 − inflammatory monocytes exhibited less pimonidazole staining than Ly6C − CD64 + fully differentiated macrophages ( Figure 3B,C). Moreover, as monocytes transitioned to Ly6C hi CD64 + cells, their pimonidazole stain positivity also increased. The trend continued, where Ly6C lo CD64 + cells had higher levels of pimonidazole compared to Ly6C hi CD64 + cells that eventually peaked in the Ly6C − CD64 + macrophages, which had the greatest pimonidazole MFI ( Figure 3B,C). These data suggest that inflammatory monocytes that enter the inflamed tissue experience hypoxic conditions, and the length of time in the lesion imprints a hypoxic signature upon these cells as they differentiate into macrophages.

Proinflammatory Stimuli and HIF-α Stabilization Induce Macrophage V in an ARNT/HIF-Dependent Manner during L. major Infection
To determine whether additional proinflammatory stimuli prese fection could also activate HIF-α in addition to hypoxia during infect an in vitro culture system using BMDMs. Macrophages were cultur panel of proinflammatory stimuli, including LPS and IFNγ, in the pre DMOG to stabilize HIF-1α and HIF-2α. These conditions in conjuncti

Proinflammatory Stimuli and HIF-α Stabilization Induce Macrophage VEGF-A P in an ARNT/HIF-Dependent Manner during L. major Infection
To determine whether additional proinflammatory stimuli present at the fection could also activate HIF-α in addition to hypoxia during infection, we e an in vitro culture system using BMDMs. Macrophages were cultured alone panel of proinflammatory stimuli, including LPS and IFNγ, in the presence or a DMOG to stabilize HIF-1α and HIF-2α. These conditions in conjunction with L

Proinflammatory Stimuli and HIF-α Stabilization Induce Macrophage VEGF-A Production in an ARNT/HIF-Dependent Manner during L. major Infection
To determine whether additional proinflammatory stimuli present at the site of infection could also activate HIF-α in addition to hypoxia during infection, we established an in vitro culture system using BMDMs. Macrophages were cultured alone or with a panel of proinflammatory stimuli, including LPS and IFNγ, in the presence or absence of DMOG to stabilize HIF-1α and HIF-2α. These conditions in conjunction with L. major infection in macrophages mimic the hypoxic environment and the elevated HIF-1α and HIF-2α expression present at the site of infection in leishmanial lesions [9]. Specifically, macrophages were first infected (or not) with L. major parasites (5:1 MOI) and then cultured with LPS alone, IFNγ alone, or both LPS and IFNγ. Compared to uninfected control macrophages, infection with L. major parasites alone did not induce VEGF-A production after 24 h ( Figure 4A). Amongst uninfected macrophages, only the combination of LPS/IFNγ resulted in significant VEGF-A production ( Figure 4A). In infected macrophages, the addition of LPS or LPS/IFNγ led to significant VEGF-A production ( Figure 4A). Importantly, VEGF-A production was significantly elevated in all conditions with the addition of DMOG when comparing the same condition with and without DMOG ( Figure 4A). athogens 2021, 10, x FOR PEER REVIEW Figure 4. Macrophages produce VEGF-A in response to proinflammatory stimu lization during L. major infection in an ARNT/HIF-dependent manner. BMDMs LPS (100 ng/mL) alone, IFNγ (10 ng/mL) alone, or both LPS (100 ng/mL) and I macrophages that were infected (or not) with L. major parasites (5:1 MOI). BMD tured under these conditions with and without 0.2 mM DMOG, which stabiliz were generated from (A) LysM Cre ARNT f/+ or (B) LysM Cre ARNT f/f to determine ARNT/HIF signaling for VEGF-A production. Supernatants were collected after duction was quantified by ELISA. Data are presented as mean +SEM. ** p < 0. comparing proinflammatory stimuli to media alone; ### p < 0.005, ## p < 0.01, tsame treatment condition with and without DMOG. Although not depicted in t be noted that all LysM Cre ARNT f/+ BMDMs treated with DMOG produced signifi A than LysM Cre ARNT f/f BMDMs treated with DMOG (p < 0.01 by t-test compari strains exposed to the same condition). Macrophages produce VEGF-A in response to proinflammatory stimuli and HIF-α stabilization during L. major infection in an ARNT/HIF-dependent manner. BMDMs were cultured with LPS (100 ng/mL) alone, IFNγ (10 ng/mL) alone, or both LPS (100 ng/mL) and IFNγ (10 ng/mL) in macrophages that were infected (or not) with L. major parasites (5:1 MOI). BMDMs were also cultured under these conditions with and without 0.2 mM DMOG, which stabilizes HIF-α. BMDMs were generated from (A) LysM Cre ARNT f/+ or (B) LysM Cre ARNT f/f to determine the contribution of ARNT/HIF signaling for VEGF-A production. Supernatants were collected after 24 h. VEGF-A production was quantified by ELISA. Data are presented as mean +SEM. ** p < 0.01, * p < 0.05, t-test comparing proinflammatory stimuli to media alone; ### p < 0.005, ## p < 0.01, # p < 0.05, t-test comparing the same treatment condition with and without DMOG. Although not depicted in the figure, it should be noted that all LysM Cre ARNT f/+ BMDMs treated with DMOG produced significantly more VEGF-A than LysM Cre ARNT f/f BMDMs treated with DMOG (p < 0.01 by t-test comparing different mouse strains exposed to the same condition).
To confirm that DMOG stabilizes HIF-α, leading to HIF-α activation during infection, VEGF-A production was measured from BMDMs generated from LysM Cre ARNT f/f , which are deficient in myeloid ARNT/HIF-α signaling, and LysM Cre ARNT f/+ control mice, which exhibit intact ARNT/HIF-α signaling. Using these genetic mouse models demonstrates that DMOG treatment directly affects ARNT/HIF-α signaling and does not act as an off-target effect. Compared to macrophages from LysM Cre ARNT f/+ mice, LysM Cre ARNT f/f macrophages produced significantly less basal VEGF-A ( Figure 4A,B).
In LysM Cre ARNT f/f macrophages, only LPS plus parasites led to significant VEGF-A production that did not increase with DMOG, suggesting that some VEGF-A production during infection can be independent of ARNT/HIF-α signaling ( Figure 4B). Importantly, LysM Cre ARNT f/f macrophages produced significantly less VEGF-A compared to LysM Cre ARNT f/+ macrophages in the presence of DMOG for all conditions ( Figure 4A,B). Altogether, these results suggest that some proinflammatory stimuli present in leishmanial lesions could activate HIF-α, but the overall effects of the proinflammatory stimuli tested are modest compared to chemical stabilization of HIF-1α and HIF-2α.

DMOG Induces HIF-1α and HIF-2α Activation during L. major Infection
During L. major infection in vivo, HIF-1α and HIF-2α, as well as multiple HIF-1α and HIF-2α targets, including Vegfa, Nos2, and Arg1, are elevated at the site of infection [7][8][9]. To determine the effects of pan-HIF-α stabilization during L. major infection, we stimulated BMDMs in the presence of DMOG to mimic HIF-1α and HIF-2α stabilization in vitro. Macrophages were generated from LysM Cre ARNT f/f and LysM Cre ARNT f/+ mice to determine the requirement for myeloid HIF-α signaling in response to DMOG administration. Macrophages were exposed to 0.1 or 0.2 mM DMOG prior to infection, and the expression of VEGF-A, which can result from HIF-1α or HIF-2α activation, was examined by real-time PCR. Our results show that L. major parasites alone do not induce VEGF-A expression in LysM Cre ARNT f/+ macrophages ( Figure 5A), confirming previous findings [9]. However, DMOG exposure in infected and uninfected macrophages upregulated Vegfa, confirming the results in Figure 4. Importantly, VEGF-A expression was lower in LysM Cre ARNT f/f macrophages compared to LysM Cre ARNT f/+ control BMDMs during L. major infection in response to DMOG, suggesting that HIF-α mediates Vegfa expression during infection. In addition, HIF-1α-specific and HIF-2αspecific target genes were examined with and without infection in BMDMs treated (or not) with DMOG. As with Vegfa, the expression of HIF-1α-specific targets Nos2 and Pgk1 was significantly higher in uninfected DMOG-treated LysM Cre ARNT f/+ macrophages compared to uninfected untreated LysM Cre ARNT f/+ controls, but the same trend was not seen in LysM Cre ARNT f/f macrophages ( Figure 5B,C). While DMOG increased both HIF-1α-specific targets Nos2 and Pgk1 in LysM Cre ARNT f/+ macrophages compared to untreated cells during infection, this effect was only significant for Nos2 expression ( Figure 5B,C). Importantly, there was no significant difference detected in HIF-1α targets Nos2 and Pgk1 in LysM Cre ARNT f/f macrophages with DMOG treatment compared to untreated cells ( Figure 5B,C). The expression of the HIF-2αspecific target Arg1 was significantly higher in both uninfected and infected LysM Cre ARNT f/+ macrophages with DMOG treatment compared to untreated LysM Cre ARNT f/+ macrophages ( Figure 5D). While Arg1 was not significantly elevated in LysM Cre ARNT f/f macrophages with DMOG treatment, some increased Arg1 was detected in uninfected LysM Cre ARNT f/f macrophages, suggesting that some DMOG-induced Arg1 is independent of ARNT/HIF-α signaling ( Figure 5D). The expression of the HIF-2α-specific target Epo was higher in both uninfected and infected LysM Cre ARNT f/+ macrophages with DMOG treatment compared to untreated LysM Cre ARNT f/+ macrophages, but this result was not significant ( Figure 5E). In summary, the findings from this figure show: (1) Vegfa and both HIF-1α-and HIF-2αspecific target genes are increased with DMOG treatment, and (2) the infection status of the macrophages does not influence HIF-α activation in response to DMOG, suggesting that the parasite does not directly play a role in macrophage HIF-α activation and, when HIF-α is activated, the parasite does not modulate that response. Altogether, these data indicate that DMOG is an effective tool to simulate Vegfa and other HIF-α targets during L. major infection.

Macrophage HIF-α Deletion and Stabilization during L. major Internalization and Killing
Given that DMOG simultaneously induces transcripts involved in parasite persistence such as Arg1, indicative of an M2 macrophage, and parasite killing such as Nos2, indicative of an M1 proinflammatory macrophage, we analyzed parasite persistence in macrophages by monitoring parasite internalization as well as their growth and survival in macrophages with and without DMOG treatment. To determine if HIF-α stabilization impacts the ability of macrophages to phagocytose L. major parasites, C57BL/6 BMDMs were treated with DMOG prior to infection. We found an equal number of parasites per macrophage after 2 h of infection in DMOG-treated cells compared to untreated macrophages, suggesting that HIF-α stabilization does not impact macrophage phagocytosis of L. major parasites ( Figure 6A). To determine if HIF-α stabilization impacts parasite killing, BMDMs were infected and then treated with DMOG for the duration of infection. As we previously reported, DMOG treatment does not impact parasite growth in macrophages for the first 72 h of infection ( Figure 6B). Although LPS and IFNγ lead to parasite killing, resulting in lower numbers of parasites after 72 h, DMOG treatment does not impact the effects of LPS and IFNγ ( Figure 6B).  While HIF-α stabilization did not affect host-parasite interactions in the first 72 h, we wanted to determine if the absence of HIF-α signaling affects host-parasite interactions early during infection. BMDMs from LysM Cre ARNT f/f and LysM Cre ARNT f/+ mice were infected, and internalized L. major parasites were quantified after 2 h. Surprisingly, LysM Cre ARNT f/f macrophages possessed a significantly higher number of parasites per macrophage compared to LysM Cre ARNT f/+ macrophages, independent of DMOG treatment ( Figure 6C). These data show that macrophages deficient in ARNT/HIF-α signaling exhibit higher numbers of phagocytosed parasites at 2 h. After 72 h p.i., parasite numbers were higher in LysM Cre ARNT f/f macrophages compared to LysM Cre ARNT f/+ controls, irrespective of DMOG treatment ( Figure 6D). LysM Cre ARNT f/f macrophages exposed to LPS While HIF-α stabilization did not affect host-parasite interactions in the first 72 h, we wanted to determine if the absence of HIF-α signaling affects host-parasite interactions early during infection. BMDMs from LysM Cre ARNT f/f and LysM Cre ARNT f/+ mice were infected, and internalized L. major parasites were quantified after 2 h. Surprisingly, LysM Cre ARNT f/f macrophages possessed a significantly higher number of parasites per macrophage compared to LysM Cre ARNT f/+ macrophages, independent of DMOG treatment ( Figure 6C). These data show that macrophages deficient in ARNT/HIF-α signaling exhibit higher numbers of phagocytosed parasites at 2 h. After 72 h p.i., parasite numbers were higher in LysM Cre ARNT f/f macrophages compared to LysM Cre ARNT f/+ controls, irrespective of DMOG treatment ( Figure 6D). LysM Cre ARNT f/f macrophages exposed to LPS and IFNγ resulted in lower numbers of parasites when compared to untreated LysM Cre ARNT f/f macrophages, suggesting that macrophages deficient in ARNT/HIF-α signaling can control parasites upon proinflammatory stimulation ( Figure 6D). However, LysM Cre ARNT f/f macrophages possessed a higher number of parasites per macrophage compared to LysM Cre ARNT f/+ controls following LPS and IFNγ stimulation, indicating that LysM Cre ARNT f/f macrophages do exhibit a slight, yet significant, defect in their ability to kill parasites ( Figure 6D, significant by t-test but not depicted on the graph). Taken together, our findings show that macrophage HIF-α stabilization does not impact parasite phagocytosis or killing, but ARNT/HIF-α signaling restricts parasite entry into macrophages.

Discussion
During L. major infection, myeloid cells, including monocytes, are recruited to dermal lesions, where they experience hypoxic conditions. Our data show that the longevity of myeloid cells in the inflamed tissue enhances their hypoxic state. We find that the hypoxic environment in leishmanial lesions is associated with the activation of HIF-α signaling. Given that monocytes differentiate into macrophages at the site of infection, and HIF-α activation can impact macrophage function, we investigated the role of HIFα signaling using a combination of strategies to augment and delete HIF-α activation during L. major infection. However, a caveat to our approach is the use of a high-dose model of Leishmania infection, which may induce more tissue damage and hypoxia than a more physiological low dose of parasites; therefore, the relevance of hypoxia and HIF-α during a low-dose infection needs to be confirmed. Upon infection, macrophages become activated to phagocytize and kill parasites by NO and ROS. However, macrophages also serve as the host cell and replicative niche for parasites. Macrophages also orchestrate lymphatic remodeling for lesion resolution. Therefore, macrophages play multiple roles in both parasite control and persistence as well as wound healing in CL. Altogether, we found that pharmacological HIF-α activation promoted the ability of macrophages to drive lymphatic remodeling through VEGF-A production during infection, but HIF-α activation did not impact parasite phagocytosis or killing. Alternatively, basal HIF-α signaling restricted macrophage parasite phagocytosis. While HIF-α stabilization did not enhance parasite killing, our findings suggest that pharmacological activation of HIF-α could induce VEGF-A, which would be beneficial for promoting lymphangiogenesis to improve lesion resolution during CL.
Hypoxia promotes macrophage phagocytosis in a HIF-1α-dependent manner, while HIF-2α is not involved in phagocytosis under hypoxic conditions [22]. In this study, we detected higher numbers of internalized parasites in macrophages deficient in myeloid ARNT/HIF-α signaling compared to macrophages with intact ARNT/HIF-α signaling in normoxic conditions. These data suggest that HIF-α inhibits parasite phagocytosis. While HIF-2α may not participate in phagocytosis under hypoxic conditions, basal HIF-2α is necessary and sufficient to suppress phagocytosis and efferocytosis under normoxic conditions [23]. The uptake of apoptotic cells or Staphylococcus aureus is higher in LysM Cre HIF-2α f/f macrophages compared to controls [23]. Moreover, LysM Cre ARNT f/f macrophages (missing both HIF-1α and HIF-2α signaling, such as the macrophages used in our experiments) show enhanced phagocytosis and efferocytosis compared to control macrophages [23]. As a result, we hypothesize that the absence of HIF-2α in our LysM Cre ARNT f/f macrophages is responsible for the enhanced parasite uptake during L. major infection. Therefore, we propose that HIF-2α acts as a phagocytic repressor during Leishmania infection.
HIF-α stabilization leads to lower bacterial and fungal burdens within macrophages [24,25]. Complementing the work of others, our data show that inflammatory stabilization of HIF-α by LPS and IFNγ leads to lower parasite numbers, but chemical HIF-α stabilization by DMOG treatment does not lead to lower parasite numbers [6,9]. HIF-α stabilization can occur by two major mechanisms, including inhibiting prolyl hydroxylase (PHD) enzymes or factor inhibiting HIF (FIH) [26]. While previous studies stabilized HIF-α by inhibiting PHD enzymes without affecting FIH [6], we stabilized HIF-α by DMOG treatment, which inhibits both PHD enzymes and FIH, and showed that chemical HIF-α stabilization alone does not lead to lower parasite numbers in the absence of additional proinflammatory stimuli. Therefore, our body of work continues to support the hypothesis that multiple factors from the tissue microenvironment, such as hypoxia and proinflammatory cytokines, contribute to parasite control. However, the requirement for HIF-α signaling in parasite control is not completely clear. While it has been reported that myeloid HIF-1α is required for the robust killing of L. major parasites in a NO-dependent manner, we found that ARNT/HIF-α signaling only plays a minor, although significant, role in parasite control. Given that LysM Cre ARNT f/f macrophages could efficiently control parasites in response to LPS and IFNγ, our data suggest that L. major parasites are controlled in an ARNT/HIF-α-independent mechanism in vitro. Mice deficient in myeloid HIF-1α exhibited higher parasite numbers following L. major infection in vivo, but we did not detect differences in parasite burdens in infected LysM Cre ARNT f/f mice, which are deficient in both myeloid HIF-1α and HIF-2α signaling [6,9]. Our in vitro findings here showing that LysM Cre ARNT f/f macrophages can control parasites in response to proinflammatory stimuli are consistent with our findings in vivo showing that LysM Cre ARNT f/f mice can control parasites at similar levels to LysM Cre ARNT f/+ controls [8,9]. It should be noted that although LysM Cre ARNT f/f macrophages exhibited a slightly higher number of parasites at 72 h following LPS and IFNγ treatment compared to LysM Cre ARNT f/+ macrophages, suggesting that HIF-α contributes to parasite control, LysM Cre ARNT f/f macrophages also internalized more parasites at 2 h, which could be responsible for the slightly higher numbers of parasites in LysM Cre ARNT f/f macrophages at 72 h, simultaneously refuting a role for HIF-α-signaling in parasite control. Altogether, these data suggest that parasite control occurs predominantly through a HIF-α-independent mechanism in vitro.
While it is clear that HIF-α subunits are active during infection, the specific factors responsible for HIF-α activation have not been defined and appear to be context-dependent. Hypoxia is hypothesized to be a major driver of HIF-α activation, but proinflammatory signals can also activate HIF-1α and HIF-2α [27,28]. Here, we used VEGF-A production as a surrogate for HIF-α activation during infection. L. major parasites can activate TLRs, and TLR activation can lead to HIF-α activation [29][30][31][32][33]. Similarly, L. major induces reactive oxygen species (ROS) production, and ROS can activate HIF-α [34][35][36][37], but parasites alone did not induce VEGF-A production by macrophages. These data suggest that parasites, parasite TLR ligation, and parasite-induced ROS production do not contribute to HIF-α activation during infection in vitro. A caveat to our study is that all experiments were performed under normoxic conditions, as we attempted to uncouple hypoxia from HIF-α activation, but it will be important to determine if similar results occur under hypoxic conditions. One might speculate that hypoxia along with DMOG treatment or the addition of proinflammatory stimuli might intensify the expression of HIF-α target genes or VEGF-A production. We tested LPS and IFNγ alone and in combination, given their known roles in stimulating proinflammatory M1 macrophages. We also predict that macrophages are exposed to these molecules during L. major infection. LPS may activate skin macrophages during CL, given the presence of the microbiome and the compromised integrity of the skin [38]. IFNγ is also elevated at the site of infection due to the host Th1 immune response [3]. While proinflammatory stimuli have been hypothesized to induce HIF-α activation, we did not find these factors to be robust drivers of macrophage VEGF-A production. However, other inflammatory stimuli such as TLR ligands, e.g., lipoteichoic acids and peptidoglycans from Gram-positive bacteria that are known to dominate in the skin microbiome, should be tested for their ability to induce HIF-α activation and/or VEGF-A production. Alternatively, hypoxia or other soluble mediators in the inflamed skin may drive HIF-α activation and VEGF-A production by macrophages during L. major infection. VEGF-A can be induced by a variety of growth factors and cytokines, including FGF2, PDGF, TGF-β, IL-1β, IL-6, IL-8/CXCL8, and TNFα, and some of these soluble mediators act synergistically with hypoxia [39][40][41][42][43][44][45][46]. Importantly, many of these factors are present in leishmanial lesions and may induce VEGF-A during L. major infection. As a result, the identification of the factors responsible for VEGF-A production by macrophages is the focus of ongoing investigation in the lab, given the critical role of VEGF-A in lesion resolution during CL.

Mice
Female and male animals used in these experiments were either purchased from the National Cancer Institute or bred in a vivarium on campus. CD45.2 + C57BL/6 and CD45.1 + C57BL/6 mice were purchased from the National Cancer Institute. Mice with a myeloid-specific ARNT conditional knockout were bred by crossing a strain expressing the LysM Cre allele with a strain with a floxed ARNT conditional allele [47,48]. The LysM Cre ARNT f/f and LysM Cre ARNT f/+ control mice were a gift from M. Celeste Simon (University of Pennsylvania, Philadelphia, PA). LysM Cre ARNT f/f mice were infected alongside LysM Cre ARNT f/+ controls for experiments. All mice were housed in vivariums under pathogen-free conditions at the University of Arkansas for Medical Sciences (UAMS). All infections were induced in mice between 6 and 8 weeks of age. All procedures performed were approved by UAMS IACUC and followed institutional guidelines.

In Vivo Infections
Infections were performed by injecting 2 × 10 6 parasites in 10 µL of PBS intradermally into the right ear of mice. Mice were anesthetized with ketamine and xylazine prior to infection. Lesion development was monitored weekly by measuring ear thickness, lesion diameter, and pathology to calculate lesion volume. Ears were digested for 90 min at 37 • C with 0.25 mg/mL liberase TL (Roche) with 10 µg/mL DNase I (Sigma) in RPMI 1640 media (Gibco).

Isolation of CD11b + Cells and Adoptive Transfer
Spleens, femurs, tibias, and fibulas were taken from C57BL/6 mice to obtain CD11b + cells. Single-cell suspensions were enriched using a CD11b MicroBead isolation kit (Miltenyi Biotec) to obtain myeloid cells. For positive selection, the autoMACS Pro Separator (Miltenyi) was used. Pooled CD11b + cells from the bone marrow and spleens (purity > 97%) were resuspended in PBS, and 4 × 10 6 cells were injected into the retroorbital sinus of the recipient C57BL/6 mouse.

Generation of Bone Marrow-Derived Macrophages
Femurs were removed from mice and soaked in 70% ethanol for 2 min and then flushed with 10 mL of cDMEM. The cells were resuspended, counted, and plated at 5 × 10 6 /mL in a 100 mm Petri dish in 10 mL of conditioned macrophage media (cDMEM with 25% of L929 cell supernatants) for 7 days. At day 3 of 7, an additional 10 mL of conditioned macrophage media was added to each Petri dish. At day 7, the cells were washed with ice-cold PBS and harvested by dislodging by pipetting and a cell scraper. The cells were then pooled and counted to plate into either 24-well plates (1 × 10 6 cells in 1 mL) or 48-well plates (5 × 10 5 cells in 500 µL).

In Vitro Infections with DMOG Treatment
Bone marrow-derived macrophages (BMDMs) were plated as described above into 24-well plates overnight. Macrophages were cultured with L. major parasites (MOI 5:1), and extracellular parasites were washed away after 2 h. Following the washes, cells were cultured in triplicate with 100 ng/mL LPS (Sigma), 10 ng/mL IFNγ (Peprotech), with or without 0.1, 0.2, or 0.3 mM DMOG (Sigma) for 2, 24, or 72 h. Pretreated macrophages were incubated overnight with DMOG prior to infection.

Microscopy for Parasite Quantification In Vitro
After infection with DsRed L. major parasites, BMDMs were stained by removing conditioned media from each well and washing with PBS before fixation by methanol at -20 • C for 3 min. After fixation, the methanol was removed, and the wells were washed twice with PBS before staining with DAPI in PBS (Invitrogen) for 5 min in the dark at room temperature. After staining, the DAPI solution was removed, and the wells were again washed twice with PBS. After washing, 500 µL of PBS was added to each well to keep the cells hydrated. The plate was then wrapped in aluminum foil and stored in a 4 • C refrigerator. Fluorescence imaging was performed on a Keyence BZ-X810 using the 20X Plan-Fluor NA 0.45 objective in high-resolution mode. Five images were taken at random locations in each well in both the DAPI and DsRed/TxRed channels. Cell counts were recorded using the BZ Image Analyzer (Keyence). Parasite counts were normalized to macrophage number by dividing the average number of parasites by the average number of macrophages for an individual well.

VEGF-A Production
Cell-free supernatants were collected at 24 h following parasite infection to measure VEGF-A production using the Mouse VEGF-A ELISA kit according to the manufacturer's instructions (R and D Systems).

Statistics
All data were analyzed using GraphPad Prism 8 or 9, and p ≤ 0.05 was considered statistically significant. Statistical significance was calculated using a 2-tailed Student's unpaired or paired t-test for a single comparison between groups. Grubbs' test was used to identify and mathematically remove outlier data points. For multiple-comparison analysis, statistical significance was determined by a one-way analysis of variance (ANOVA) followed by the post hoc Tukey's test with no designated control group or a Mann-Whitney test with a designated control group.