NADPH oxidase in B cells and macrophages protects against murine lupus by regulation of TLR7

Loss of NADPH oxidase (NOX2) exacerbates systemic lupus erythematosus (SLE) in mice and humans, but the mechanisms underlying this effect remain unclear. To identify the cell lineages in which NOX2 deficiency drives SLE, we employed conditional KO and chimeric approaches to delete Cybb in several hematopoietic cell lineages of MRL.Faslpr SLE-prone mice. Deletion of Cybb in macrophages/monocytes exacerbated SLE nephritis, though not to the degree observed in the Cybb global KOs. Unexpectedly, the absence of Cybb in B cells resulted in profound glomerulonephritis and interstitial nephritis, rivaling that seen with global deletion. Furthermore, we identified that NOX2 is a key regulator of TLR7, a driver of SLE pathology, both globally and specifically in B cells. This is mediated in part through suppression of TLR7-mediated NF-κB signaling in B cells. Thus, NOX2’s immunomodulatory effect in SLE is orchestrated not only by its function in the myeloid compartment, but through a pivotal role in B cells by selectively inhibiting TLR7 signaling.


Introduction
Systemic lupus erythematosus (SLE) is a multisystem autoimmune disease marked by loss of tolerance to self and rampant immune activation, resulting in damage to target organs (1).A hallmark feature of SLE is loss of tolerance to nuclear antigens and the formation of autoantibodies against nucleic acids and nucleoprotein complexes.The origins of autoantigens in SLE remain uncertain.Antigenic contents from dying neutrophils and/or a failure to dispose of cellular debris by macrophages are 2 possible sources of autoantigen in SLE (2).The NADPH oxidase complex, a group of transmembrane and cytosolic enzymes responsible for the respiratory burst crucial to eliminate microbes (3)(4)(5), is important for both of these processes (6)(7)(8)(9)(10).
Chronic granulomatous disease (CGD) results from loss-of-function mutations in key components of NADPH oxidase 2 (NOX2).Male patients with X-linked CGD have a propensity to develop a lupus-like disease (11,12).Susceptibility to systemic autoimmunity in patients with CGD is likely to be directly mediated by lack of NAPDH oxidase activity rather than indirect effects of having CGD per se, as heterozygous female carriers of the X-linked cytochrome b-245, β polypeptide (Cybb) null allele also have a higher likelihood of developing SLE, despite not having CGD (13,14).Moreover, loss-of-function polymorphisms in 2 other NOX2 components, neutrophil cytosolic factor 1 and 2 (NCF1 and -2), confer increased SLE susceptibility (15)(16)(17)(18).Others and we have consistently demonstrated that multiple rodent models of CGD exhibit the same increased susceptibility to autoimmunity as observed in human patients (19)(20)(21)(22)(23)(24)(25)(26)(27).Thus, NOX2 is a critical immune regulator in humans and mice.
Loss of NADPH oxidase (NOX2) exacerbates systemic lupus erythematosus (SLE) in mice and humans, but the mechanisms underlying this effect remain unclear.To identify the cell lineages in which NOX2 deficiency drives SLE, we employed conditional KO and chimeric approaches to delete Cybb in several hematopoietic cell lineages of MRL.Fas lpr SLE-prone mice.Deletion of Cybb in macrophages/monocytes exacerbated SLE nephritis, though not to the degree observed in the Cybb global KOs.Unexpectedly, the absence of Cybb in B cells resulted in profound glomerulonephritis and interstitial nephritis, rivaling that seen with global deletion.Furthermore, we identified that NOX2 is a key regulator of TLR7, a driver of SLE pathology, both globally and specifically in B cells.This is mediated in part through suppression of TLR7-mediated NF-κB signaling in B cells.Thus, NOX2's immunomodulatory effect in SLE is orchestrated not only by its function in the myeloid compartment, but through a pivotal role in B cells by selectively inhibiting TLR7 signaling.
In addition to cell specificity, it also remains unclear how NOX2 promotes its protective effect mechanistically.Interestingly, in our aforementioned study demonstrating exacerbated SLE in Cybb-deficient mice, we observed an increase in anti-RNA and anti-Smith (anti-Sm) autoantibody titers (20).The formation of these autoantibodies is TLR7 dependent in a B cell-intrinsic fashion (41).Thus, it is plausible that NOX2 and TLR7 may have a regulatory interaction in SLE, and that this relationship could be B cell intrinsic.
To investigate how NOX2 deficiency promotes autoimmunity in various cell subsets, we used in vivo chimeric and conditional KO approaches to delete Cybb, the gene encoding a critical subunit of the NOX2 complex, in specific cell subsets of MRL.Fas lpr SLE-prone mice, a strain in which we previously demonstrated strong regulatory activity of NOX2 via global KO of Cybb (20).The MRL.Fas lpr mouse is a leading spontaneous model for the study of SLE, developing nearly all features of the human disease (42).Moreover, the MRL.Fas lpr strain has accurately predicted responses in human translational studies, validating its utility (42)(43)(44)(45)(46)(47)(48).
Here, we found that Cybb deficiency in the hematopoietic compartment decreased survival, exacerbated SLE nephritis, and altered the autoantibody response in the MRL.Fas lpr model of SLE.Furthermore, the absence of CYBB in either B cells or macrophages/monocytes drove clinical and immunologic features of SLE, with the effect in B cells being particularly strong, whereas no effects were seen upon deleting Cybb in neutrophils or T cells.We also demonstrate that global Tlr7 deficiency is sufficient to suppress several hallmarks of SLE pathogenesis in Cybb-deficient MRL.Fas lpr mice.Finally, we show that NOX2 is a key negative regulator of TLR7 in B cells at both the genetic and signaling level, linking this mechanism to the in vivo phenotype of NOX2 deficiency.The identification of macrophages and, unexpectedly, B cells, as key sites of action for NOX2 in regulating autoimmunity represent important insights into how the NOX2 complex regulates multiple inflammatory states.

Results
Hematopoietic Cybb deficiency decreased survival, exacerbated kidney disease, and altered the anti-self response in SLE-prone MRL.Fas lpr mice.To determine whether NOX2 in the hematopoietic or stromal compartment is the primary driver of the exacerbated manifestations of autoimmunity that we and others observed in global NOX2-deficient mice (19-23, 26, 27), we first generated bone marrow (BM) chimeras in MRL.Fas lpr SLE-prone mice using Cybb-KO or control Cybb-sufficient donor BM.SLE pathology was assessed in chimeric mice 16-18 weeks after irradiation unless otherwise indicated.
Strikingly, mice receiving Cybb-deficient BM had reduced survival and increased urine protein compared with controls (Figure 1, A and B).Mice that received Cybb-deficient BM also had worse glomerular and interstitial nephritis (Figure 1, C and D).Spleen weight was increased in these mice (Figure 1E).Anti-nucleosome antibody titers were reduced in SLE-prone mice with a hematopoietic Cybb defect (Figure 1F), consistent with altered anti-nuclear antibody patterns seen in global Cybb-deficient mice (20).Hematopoietic Cybb deficiency did not alter anti-Sm titers (Figure 1G), unlike the global Cybb-KO mice.Chimeras that lacked CYBB in the hematopoietic compartment had elevated anti-RNA titers in male, but not female, mice (Figure 1H).
While the aforementioned data highlight a role for hematopoietic NOX2 in regulating SLE pathogenesis, these findings do not preclude a contribution from the stromal compartment.To address this consideration, we generated reciprocal BM chimeras.Stromal Cybb deficiency did not drive SLE kidney disease (Supplemental Figure 1, A-C) or alter the anti-self response (Supplemental Figure 1, D-F).
Global Tlr7 deficiency rescues exacerbated SLE in Cybb-deficient mice.NOX2-deficient animals have higher levels of TLR7-dependent autoantibodies.To test the hypothesis that NOX2 regulates SLE by dampening TLR7-mediated autoimmunity, we crossed globally Tlr7-deficient mice with global Cybb-KO mice on the MRL.Fas lpr SLE-prone background.At 15-16 weeks of age, compared with Cybb-KO mice, Tlr7 and Cybb double-KO mice had significantly reduced proteinuria and glomerulonephritis (Figure 2, A  and B).Interstitial nephritis was significantly decreased in male Tlr7 and Cybb double-KO mice (Figure 2C).Male Cybb -/Y Tlr7 -/Y mice had improved dermatitis (Figure 2D).Tlr7 and Cybb double-KO mice had reduced splenomegaly, but only female mice had improved lymphadenopathy (Figure 2, E and F).As expected, Tlr7 and Cybb double-KO mice had fewer TLR7-driven anti-RNA and anti-Sm autoantibodies, while female mice had increased titers of anti-nucleosome autoantibodies (Figure 2, G-I).
JCI Insight 2024;9(16):e178563 https://doi.org/10.1172/jci.insight.178563 Compared with Cybb-KO controls, Tlr7 and Cybb double-KO mice had reduced splenic B cell lymphopenia, a hallmark of severe SLE in both mice and humans; this was driven by an increase in the proportion of follicular B cells (Supplemental Table 1).The proportion of splenic CD11b + CD11c + age-associated B cell-like (ABC-like) B cells was increased in females, while the proportion of plasmablasts was unchanged.The proportion of naive CD4 + T cells trended upwards in male mice (P = 0.06).The percentage of plasmacytoid DCs (pDCs) was increased in both male and female Tlr7 and Cybb double-KO mice.The remainder of the myeloid compartment, including macrophages, monocytes, conventional DCs (cDCs), and neutrophils, was unchanged except for a decrease in macrophages in male mice.Cybb deletion in MRP8-expressing cells (neutrophils) does not impact murine SLE.To determine in which cell types NOX2 exerts a protective effect in systemic autoimmunity, we employed a conditional KO approach to selectively delete Cybb in immune cell types of interest.To accomplish this, we generated a Cybb conditional KO allele (Cybb fl/fl ), which was made directly on the MRL.Fas lpr background, using CRISPR/Cas9 technology (49).Deletion efficiency in all conditional KO strains was confirmed by qPCR of genomic DNA isolated from FACS-purified immune cell populations (Supplemental Table 2).
We did not observe any differences in proteinuria, nephritis, dermatitis, or splenomegaly in the Cybb fl/fl MRP8-Cre cohort (Figure 3, A-E).Axillary lymph node weights were increased in male controls, but this finding

R E S E A R C H A R T I C L E
JCI Insight 2024;9(16):e178563 https://doi.org/10.1172/jci.insight.178563was not seen in any of the other control cohorts (Figure 3F).Cybb deficiency in neutrophils did not impact titers of anti-nucleosome, -Sm, or -RNA autoantibodies (Figure 3, G-I).The percentages of splenic T cells, B cells, macrophages, neutrophils, cDCs, and pDCs were similar across genotypes (Supplemental Table 1).
Although we did not observe any significant differences in proteinuria among the groups, interstitial nephritis was exacerbated in Cybb fl/fl LysM-Cre +/-female mice and there was a trend toward worse glomerulonephritis (Figure 4, B-D).No differences in splenomegaly were observed (Figure 4E).While female Cybb fl/fl LysM-Cre +/-mice had reduced titers of anti-nucleosome antibodies (Figure 4F), no differences in anti-Sm or anti-RNA titers were observed (Figure 4, G and H).Cybb fl/fl LysM-Cre +/-females had elevated frequencies of CD11b + Gr1 + splenic neutrophils.The male Cybb fl/Y LysM-Cre +/-cohort had increased percentages of total B cells and antibody-forming cells (AFCs).No other differences in the myeloid, B cell, or T cell compartments were observed between the groups (Supplemental Table 1).
The mild exacerbation of disease seen in Cybb fl/fl LysM-Cre mice did not fully reproduce the exacerbated disease that we observed in the Cybb global KO (20), global heterozygote, or BM chimera cohorts.One explanation for this finding is that Cybb deletion efficiency was only 27% and 43% among splenic macrophages in male and female conditional KO cohorts, respectively (Supplemental Table 2), thus muting any potential phenotype.Therefore, to more thoroughly explore the role of Cybb in the myeloid compartment, we generated mixed BM chimeras with 80% of BM from the Rosa26-EGFP-DTA +/-LysM Cre +/-MRL.Fas lpr strain -which cannot form cells expressing LysM (i.e., almost all neutrophils and ~50% of splenic macrophages; Figure 4A and Supplemental Methods) -and 20% of BM from Cybb-KO MRL.Fas lpr mice.This strategy resulted in experimental mice in which Cybb was deleted in approximately 88% of splenic CD11b + Gr1 + neutrophils and absent in approximately 46% of CD11b + F4/80 + splenic macrophages (Supplemental Table 2).The experimental chimeric mice are referred to as ΔLysM Cybb -/-.SLE phenotypes were assessed in chimeric mice 16 weeks after irradiation.Proteinuria was increased in ΔLysM Cybb -/- MRL.Fas lpr female mice (Figure 4B).Glomerular and interstitial nephritis was exacerbated in both male and female ΔLysM Cybb -/-MRL.Fas lpr mice compared with controls (Figure 4, C and D).Spleen weights were increased in male, but not female, ΔLysM Cybb -/-MRL.Fas lpr mice (Figure 4E).
A potential limitation of the mixed BM chimera experiment is that 20% of BM-derived cells that are not LysM positive also lack Cybb.This could be relevant since Cybb heterozygous (Cybb +/-) female MRL.Fas lpr mice, in which 50% of cells lacked CYBB protein due to mosaic X inactivation, developed exacerbated SLE-like disease (20), indicating that the presence of a subset of cells that lacks CYBB can have a dominant effect in promoting the disease.To rule out that the absence of CYBB in 20% of cells could mediate the exacerbated disease phenotype seen in the ΔLysM Cybb -/-cohort, we generated an additional control of mixed complete 80:20 wild-type (WT)/Cybb-KO BM chimeras.Reconstitution of WT MRL.Fas lpr recipients with 20% of Cybb-KO BM did not worsen renal disease (Supplemental Figure 2, A-C) or alter the autoantibody response (Supplemental Figure 2, D-F).Thus, we conclude that in the ΔLysM Cybb -/-cohort, the 20% of non-LysM-expressing cells that are Cybb deficient are not responsible for the observed phenotypes.Cybb was efficiently deleted in both the CD4 + and CD8 + T cell compartments (Supplemental Table 2).We did not observe any significant differences in kidney disease (Supplemental Figure 3, A-C), dermatitis (Supplemental Figure 3D), or lymphoproliferation (Supplemental Figure 3, E and F).Autoantibody titers were similar between the groups, with the exception of anti-RNA levels which were decreased in female Cybb fl/fl CD4-Cre +/-mice (Supplemental Figure 3, G-I).
Anti-nucleosome titers were lower in both male and female Cybb fl/fl CD19-Cre +/-MRL.Fas lpr mice (Figure 5G).There were no differences in anti-RNA or anti-Sm titers among the groups (Figure 5, H and I).No differences were observed in the percentages of total CD19 + B cells or CD19 lo/int CD44 + C-D138 + intracellular κ hi AFCs in Cybb fl/Y CD19-Cre +/-versus control animals (Supplemental Table 1).Similarly, there were no differences in follicular, marginal zone, CD11b + CD11c + ABCs, or germinal center B cells among the groups (Supplemental Table 1).
B cell-intrinsic Tlr7 deficiency is sufficient to ameliorate severe nephritis in Cybb-deficient mice.Since global disease exacerbation mediated by the loss of NOX2 was ameliorated by global absence of TLR7, we wanted to determine in which cell types this effect was being mediated.TLR7 was previously shown to have a role in B cells in promoting disease, and here we showed (Figure 5) that NOX2 regulates disease in B cells.Hence, we hypothesized that B cell-intrinsic TLR7 would be important in driving NOX2-deregulated SLE.To test this, we deleted Tlr7 in B cells of globally Cybb-deficient mice MRL.Fas lpr mice by generating Cybb -/Y Tlr7 fl/Y CD19-Cre +/-(male) and Cybb -/-Tlr7 fl/fl CD19-Cre +/-(female) cohorts.SLE phenotypes were assessed at 15 weeks, comparing Cre-positive to Cre-negative littermate controls.Strikingly, all parameters of SLE, including proteinuria, glomerulonephritis, and interstitial nephritis were significantly improved in both males and females (Figure 6, A-C) when TLR7 was specifically and only deleted in B cells on a global Cybb-deficient background.Unlike in the cohort in which both genes were deleted globally, spleen weight and skin disease were not different from controls, while Cybb -/Y Tlr7 fl/Y CD19-Cre +/-males demonstrated an increase in lymphadenopathy (Figure 6, D-F).Similar to the global KO cohort, experimental mice had reduced anti-RNA and anti-Sm autoantibodies, with an increase in anti-nucleosome autoantibodies (Figure 6, G-I).
There were no differences in splenic frequencies of various immune cell types (Supplemental Table 1), except for a small increase in DC proportions and, notably, an increase in the percentage of naive CD4 + T cells in female mice.
Cybb-KO B cells have increased NF-κB pathway activation after TLR7 stimulation.While the results presented thus far indicate that the absence of NOX2 promotes B cell-specific, TLR7-dependent autoimmunity, whether this is a direct or indirect effect in the B cell could not be distinguished.To test for a direct effect of the absence of NOX2 in promoting TLR7 signaling, we measured NF-κB signaling downstream of TLR7 activation.We analyzed both upregulation of the phosphorylation of the NF-κB subunit p65 (p-p65) as well as p65 nuclear translocation in Cybb-sufficient and Cybb-deficient marginal zone B cells; marginal zone B cells have higher levels of TLR7 (41), making them more sensitive to TLR7 stimulation and thus enhancing the sensitivity of the approach.To demonstrate the fundamental biology of this relationship, we utilized ex vivo marginal zone B cells from WT and Cybb-KO B6 mice, as the excessive background inflammation in MRL.Fas lpr mice, which in turn can modulate TLR7 expression levels, could have confounded the conclusions.As BCR and TLR signals synergize to activate autoreactive B cells in SLE, we stimulated marginal zone B cells with a combination of anti-IgM and CL097, a TLR7 agonist.Fifteen minutes after stimulation, we observed an increase the percentage of p-p65 + B cells in Cybb-KO cells versus WT cells (Figure 7A).
As an additional and more definitive measure of TLR7-mediated activation of the NF-κB family, we assessed p65 nuclear translocation after stimulation with anti-IgM and CL097 or with CL097 alone.Consistent with a direct negative regulatory role of NOX2 on TLR7 signaling in B cells, we observed an increase in nuclear translocation in Cybb-KO B cells compared with Cybb-sufficient controls (Figure 7B) in cells treated with BCR plus TLR7 stimulation or with TLR7 stimulation alone.Based on the established importance of cysteine residues in TLR7's ectodomain (57,58), which theoretically could be modulated by NOX2-derived reactive oxygen species (ROS), we also pretreated selected samples with either catalase (to breakdown ROS) or hydrogen peroxide (as a source of ROS).Interestingly, hydrogen peroxide pretreatment quenched TLR7-initiated NF-κB signaling via CL097 and anti-IgM, while catalase pretreatment did not affect signaling in Cybb-KO cells but caused signaling in WT cells to significantly increase (Figure 7B).Importantly, Cybb genotype did not affect the response to stimulation of TLR9 with CpG alone or CpG and anti-IgM stimulation, indicating that this regulatory mechanism via NOX2 appears specific to activation through TLR7 (Figure 7C).

Discussion
The significance of NOX2 as a negative regulator of autoimmunity in multiple animal models, as well as human autoimmune syndromes, is becoming increasingly recognized (11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27).However, the mechanisms by which NOX2 mediates these effects have been largely unknown.Based on the totality of the cell-specific NOX2 deletion systems, encompassing 4 different Cre strains along with the mixed BM chimera approaches, here we have identified both B cell-and macrophage/monocyte-expressed NOX2 as fundamental negative regulators of SLE pathogenesis.In addition to identifying the cell-type-specific roles of NOX2 in our models -which by itself is important information -we have identified a B cell-intrinsic mechanism by which NOX2 operates.In B cells, NOX2 selectively regulates TLR7, but not TLR9, signaling.Thus, we have provided what we believe are new insights into major unaddressed questions regarding the emerging role of NOX2 in autoimmunity: in which cell types and by what mechanism does NOX2 mediate its regulatory functions?
A dominant regulatory role for NOX2 in the B cell was unexpected.Notably, the disease phenotype of B cell-specific Cybb deletion is more severe than that seen in any of the LysM-Cre-dependent systems, underscoring the relative importance of B cell-expressed NOX2 in the MRL.Fas lpr lupus model.In fact, B cell-specific NOX2 deletion nearly recapitulates the effects of global deletion.In line with our findings, contemporaneous work by Liu et al. showed that deletion of B cell-intrinsic Ncf1 (which encodes a cytosolic subunit of NADPH oxidase) results in spontaneous B cell activation and increased autoantibodies in aged WT mice (59).These phenotypes were partially reproduced in low-penetrant mouse models of systemic autoimmunity (59).However, disease endpoints were not assessed in these model systems and it is well known that autoantibodies and disease do not necessarily follow, including in the case of TLR9 deletion where anti-DNA antibodies are abrogated, while disease is actually exacerbated (60,61).Thus, our finding that NOX2 deficiency in the B cell compartment exacerbates both clinical disease (nephritis) and immunologic endpoints in a relevant spontaneous polygenic model of SLE is a distinctive and significant milestone in the understanding of NOX2 in the context of autoimmunity.
It is conceivable that NOX2 plays a crucial role in dampening the activity of a newly recognized subset of inflammatory B cells known as "atypical" or double-negative 2 memory B cells in humans, or ABCs in mice (62,63).ABCs are found at higher frequency in patients with SLE and in mouse models of lupus (62,63).TLR7 is important for the generation of ABCs (64)(65)(66)(67)(68)(69).Selective deletion of ABCs in the MRL.Fas lpr model improves nephritis and reduces T cell activation (70).Recently, Luo et al. have shown that expression of a conditional Ncf1-knockin allele in CD11c-expressing cells of NOX2-deficient mice alleviates lupus in both the pristane-induced lupus and spontaneous Yaa models.The authors attribute this phenotype to a regulatory function of NOX2 in pDCs (27).In light of our finding that NOX2 is an important regulator in B cells, and that CD11c is also expressed on ABCs, restoration of NOX2 in ABCs may well have contributed to the reduced disease burden observed by Luo et al.
It is perhaps expected that NOX2 in myeloid cells would regulate the pathogenesis of SLE, since the expression and function of NOX2 in macrophages, DCs, and neutrophils is well documented (7)(8)(9)(10)(27)(28)(29)(30).Indeed, others have shown macrophage activation in NOX2-deficient animals subjected to collagen-induced arthritis (71).As noted, pDCs have also been implicated in pristane-induced SLE (27).Cell-selective restoration of NOX2 function in the context of global deletion ameliorated disease in both models, establishing the importance of the target cells.Here, we took a complementary approach, using cell-selective deletion.As with both approaches, the limited efficiency and specificity of existing myeloid Cre strains in turn limits the ability to directly assess the differential contribution of myeloid subsets.In our case, we found an effect using LysM-Cre, which efficiently targets neutrophils but variably and partially targets macrophage populations (53).Cybb deficiency in the myeloid compartment, mediated by LysM-Cre in both our chimeric and conditional KO systems, does not fully recapitulate the Cybb global KO or total hematopoietic chimera phenotypes.The lack of a robust phenotype could be, in part, due to incomplete deletion of the allele.However, our combined results establish that NOX2 fulfills distinct, nonredundant regulatory functions across multiple cell types, which could also explain the partial phenotypes seen with macrophage/ monocyte-and neutrophil-specific deletion.
NOX2-dependent ROS are critical for the formation of classical neutrophil extracellular traps (NETs) (72,73).NETs were postulated to be a major source of autoantigen and were thought to consequently drive disease in SLE (74)(75)(76)(77)(78)(79)(80)(81)(82)(83).However, others and we have shown that abrogating NET formation by genetically deleting or pharmacologically inhibiting NOX2 (19)(20)(21)(22)(23)(24)(25)(26)(27), peptidyl arginine deiminase, type IV (23,84), and neutrophil elastase (85) exacerbated rather than ameliorated SLE phenotypes, or at best, had no effect.The data that we present here showing that conditional deletion of Cybb in neutrophils using MRP8 Cre did not impact SLE, along with that of the Homdahl group (27), strongly argues against the hypothesis that NET-derived autoantigens drive SLE.Rather, exacerbated SLE observed in the absence of NETs suggests that NETs may be immunomodulatory in SLE.Supporting a negative regulatory role for NETs, NOX2-dependent NETs can form aggregated structures that degrade proinflammatory cytokines (28).As targeting of Cybb in neutrophils had no effect on disease, it is unlikely that worse SLE observed in NOX2 deficiency can be explained by the absence of immunoregulatory NETs.
Recently, Liu et al. speculated that NCF1 deletion promotes increased TLR signaling via dysregulation of endolysosomal trafficking in B cells (59), based on in vitro studies.This mechanism is seemingly at odds with the selective regulation of TLR7 that we observe here.They reported that NADPH oxidase deficiency led to reduced noncanonical autophagy, characterized by diminished LC3 recruitment to the endosome and delayed lysosomal fusion, a process dependent on LC3 lipidation, and known as LC3-associated phagocytosis, or LAP (100).This in turn culminates in the reduced degradation of the TLR ligand CpG and sustained TLR9 signaling (59).These in vitro studies, however, were not linked in a causal way to the humoral activation and elevated autoantibody titers in their in vivo systems.Moreover, this interpretation hinges on a conceptual framework that was based on publications claiming that NOX2 (and RUN and cysteine-rich domain-containing Beclin 1 interacting protein [RUBICON]) are essential for LAP (101,102);

R E S E A R C H A R T I C L E
JCI Insight 2024;9(16):e178563 https://doi.org/10.1172/jci.insight.178563however, the original papers making these claims have been retracted.Indeed, our lab along with Anne Davidson's had already demonstrated that LC3 lipidation occurs normally without NOX2 or RUBICON (49), unlinking NOX2 from LC3-associated processes such as those discussed by Liu et al.
The model proposed by Liu et al. does not explain why TLR7 drives and TLR9 protects from SLE -in fact, it is incompatible.There is an emerging body of literature to suggest that TLR7 and TLR9 are differentially regulated in the endosomal compartment (103)(104)(105).Recently, the late endosomal biogenesis of lysosomal organelles complex 1-related complex (BORC) and GTPase adenosine 5-diphosphate ribosylation factor-like 8B (Arl8b) have been found to control intracellular TLR7, but not TLR9, turnover (104).It remains possible that NOX2 is also involved in this process.
In fact, how NOX2 selectively regulates TLR7 remains undetermined.One possible mechanism could be that NOX2-dependent ROS oxidizes essential cysteine residues in the TLR7 ectodomain; TLR9 lacks such residues (57, Supporting the significance of this observation, activation of NOX2 in mice undergoing infection with influenza, a pathogen known to activate TLR7, suppressed proinflammatory cytokine production and reduced antibody production (57).Consequently, we speculate that NOX2-derived ROS in B cells protects from systemic autoimmunity by inhibiting TLR7 signaling by oxidation of Cys98.This proposed mechanism would explain the selective impact of NOX2 on TLR7 and aligns with the opposing functions of TLR7 and TLR9 in SLE pathogenesis.Moreover, it is well established that environmental factors play a critical role in triggering the onset of autoimmune diseases (106).It is intriguing to speculate that the dysregulation of NOX2-dependent cysteine oxidation may act as an environmental catalyst that precipitates the onset of disease in genetically susceptible individuals.

Methods
Further information can be found in Supplemental Methods.
Sex as a biological variable.Our study examined male and female animals.Sex-dimorphic effects are reported.
Cybb fl/fl MRL.Fas lpr mice were generated by in vitro fertilization and CRISPR/Cas9 technology, as previously described (49,109).Two loxP sites were inserted in introns 3 and 4 flanking exon 4 of Cybb.To facilitate screening of founder mice and subsequent genotyping, an EcoR1 restriction site was added adjacent to each loxP site.
To generate the Tlr7-KO Cybb-KO cohort, Tlr7-deficient and Cybb-deficient mice were intercrossed until both alleles were homozygous.Because both Cybb and Tlr7 are on the X chromosome, crosses to obtain control and experimental female mice were done in parallel; control and experimental male mice were littermates.Disease was evaluated at 15-16 weeks of age.

JCI
Insight 2024;9(16):e178563 https://doi.org/10.1172/jci.insight.178563T cell Cybb deficiency does not affect murine SLE.To test the hypothesis that NOX2 deficiency in T cells contributed to autoimmunity in SLE, we crossed the Cybb fl/fl with CD4-Cre (54) MRL.Fas lpr strains to generate a cohort of experimental Cybb fl/fl CD4-Cre-positive mice and Cre-negative Cybb fl/fl controls.Male cohorts were analyzed at 22 weeks of age and female cohorts were analyzed at 19 weeks of age.