CT-M8 Mice: A New Mouse Model Demonstrates That Basophils Have a Nonredundant Role in Lupus-Like Disease Development

Tissue-specific mouse models are essential tools to decipher the role of each cell compartment and/or their expressed genes in the pathophysiology of diseases. Here, we describe a new knock-in mouse model allowing expression of both the fluorescent protein tdTomato and the CRE recombinase selectively in the basophil compartment under the control of the Mcpt8 gene. These “CT-M8” mice did not show any abnormalities in their peripheral distribution of major immune cell populations nor their basophil function. CT-M8 mice allowed the identification of basophils by immunofluorescence and flow cytometry and basophil-specific Cre-mediated floxed gene deletion. Breeding of our CT-M8 mice with the ROSA26flox-stop-DTA mice led to the generation of basophil-deficient mice with no detectable abnormalities in other cell compartments. These mice were then used to document basophil involvement in systemic lupus erythematosus (SLE) pathophysiology since we previously reported by transient depletion of these cells during the course of an ongoing disease that they support and amplify autoantibody production in two distinct lupus-like mouse models (Lyn−/− and pristane-induced). Here, constitutive basophil deficiency prevented pristane-induced lupus-like disease development by limiting autoantibody titers and renal damages. Therefore, basophils have a nonredundant role in pristane-induced lupus-like disease and are involved in both its induction and amplification. This CT-M8 new mouse model will allow us to finely decipher the role of basophils and their expressed genes in health and disease.


INTRODUCTION
Systemic lupus erythematosus (SLE) is an autoimmune disease (AID) affecting mainly women of child-bearing age. SLE can affect different organs, such as skin, lungs, joints, or kidneys. Lupus nephritis is one of the most severe organ manifestations of SLE. It affects 25% to 60% of SLE patients, and about 20% of these patients develop a life-threatening end-stage renal disease (1). SLE is characterized by the presence in the patient's blood of autoreactive antibodies of several isotypes, mainly raised against nuclear antigens such as double-stranded DNA (dsDNA) or ribonucleoproteins (RNPs) (2). Once conjugated to autoantigens and complement factors, the formed immune complexes (ICs) can deposit in target organs where they induce inflammation, leading to organ injury and consequent organ dysfunction. ICs also induce Fc receptor-dependent activation of innate immune cells such as plasmacytoid dendritic cells (pDCs), neutrophils, macrophages, and basophils, which all contribute to an amplification loop of autoantibody production and disease activity (1,3).
We previously reported that basophils accumulate in secondary lymphoid organs (SLOs) during lupus pathogenesis through mechanisms dependent on IgE, IL-4, prostaglandin D2 (PGD2), and CXCR4. There, they promote autoreactive plasma cell accumulation and consequent autoantibody production (4)(5)(6). These findings were validated in several human SLE patient cohorts and multiple lupus-like mouse models. Among these models, basophils were shown to contribute to the pristaneinduced lupus-like disease (PIL) (7). In mice with established disease, basophil depletion dampened disease activity by reducing short-lived plasma cell number, serum autoantibody titers, IC deposition in glomeruli, and the proinflammatory environment in the kidneys (4,6,7), identifying basophils as promising therapeutic targets in SLE. However, whether basophils are involved in the primary development of the disease remains unknown.
Here, we report the generation of a new mouse model called CT-M8 (Mcpt8 tm1Ics ) that allows the expression of the CRE recombinase (CRE) and the tandem Tomato (tdT) fluorescent protein downstream of the gene of the basophil-specific protease mast cell protease 8 (Mcpt8) (8). This model enabled convenient basophil detection by fluorescent microscopy and flow cytometry as well as basophil-specific floxed gene deletion. Breeding of these CT-M8 mice with the ROSA-DTA mice (expressing a loxP site-flanked stop cassette upstream of the gene encoding diphtheria toxin A under the control of the Rosa26 promoter [B6.129P2-Gt(ROSA)26Sortm1(DTA)Lky/J)] (9) allowed the generation of constitutive basophil-deficient mice (Mcpt8 CT/+ Rosa26 DTA/+ ). In the PIL model, basophil deficiency prevented the accumulation of short-lived plasma cells in SLOs as well as the detection of anti-RNP autoantibodies and IC deposits in the glomeruli of the mice 8 weeks after pristane injection. Overall, this new model of basophil-deficient mice allowed showing that basophils have a nonredundant role in PIL development and might be mandatory in the SLE-like disease course to reach a pathogenic threshold.

Generation of Mcpt8 tm1Ics or CT-M8 Mice
The C57BL/6 Mcpt8-Cre-tdTomato or Mcpt8 tm1Ics or "CT-M8" mice were generated by the Institut Clinique de la Souris (Illkirch, France) from the Phenomin consortium. Briefly, a cassette containing an internal ribosome entry site (IRES), the cDNA encoding the CRE recombinase, a viral Thosea asigna virus 2A (T2A) sequence, and the fluorescent protein tandem Tomato or "tdT" was knocked in together with a Neo resistance (Neo r ) cassette flanked by two FRT sites after the stop codon of the Mcpt8 gene. Chimeric mice were generated by the injection of mutant ES cells into C57BL/6 blastocysts and cross-bred with Flp deleter C57BL/6N females (10) to obtain Neo r -deleted Mcpt8 tm1lcs mice ( Figure 1A). For phenotyping experiments, mice from both sexes in equal proportions were used between 8 and 15 weeks of age. For pristane-induced lupus-like (PIL) disease experiments, 8-weekold female mice of the indicated genotype received a single intraperitoneal injection of 500 µl of pristane (Sigma-Aldrich, Burlington, MA, USA) or phosphate-buffered saline (PBS, Gibco) as a control and were sacrificed 8 weeks later. Mice were maintained under specific pathogen-free conditions in our animal facilities. The study was conducted in accordance with the French and European guidelines and approved by the local ethics committee, comitéd'ethique Paris Nord N°121 and the Ministère de l'enseignement supeŕieur, de la recherche et de l'innovation under the authorization number APAFIS#14115.

Mouse Sample Processing
Mice were euthanized in a regulated CO 2 chamber. Blood was harvested with a heparinized syringe with a 25-G needle by cardiac puncture. Peritoneal lavages were realized by injecting 5 ml of PBS and 2 ml of air into the peritoneal cavity. After 50 rotations, the peritoneal wall was opened and peritoneal lavage was harvested with a pipette. Cervical, axillary, and inguinal lymph nodes were harvested and pooled. Bone marrow was extracted from both femurs, and the spleen was harvested. The left kidney was embedded in OCT (CellPath, Newton, UK) and snap-frozen in liquid nitrogen and kept at −80°C until immunofluorescence analysis. Single-cell suspensions from blood, bone marrow, peritoneal lavage, spleen, or peripheral lymph nodes (LNs) (pooling inguinal, axillary, and cervical LNs) were prepared by mechanical disruption over a 40-mm cell strainer (Falcon, Corning, Corning, NY, USA) when necessary. Blood was centrifuged at 300×g for 10 min at room temperature, and plasma was harvested for further analysis. The specimens were incubated for 5 min at room temperature in ACK lysing buffer (150 mM of NH 4 Cl, 12 mM of NaHCO 3 , 1 mM of EDTA, pH 7.4) and neutralized with twice the volume of PBS, and cells were centrifuged (500×g, 5 min). This step was repeated three times for blood samples. Cell concentrations were determined with a hemacytometer, and 1-5 million cells per point were used for fluorescence-activated cell sorting (FACS) staining (see below).

Flow Cytometry
Unspecific antibody-binding sites were saturated with a blocking buffer containing 10 µg/ml of anti-CD16/CD32 antibody clone

Ex Vivo Primary Cell Stimulation
Mouse splenocytes were harvested through mechanical disruption over a 40-mm of cell strainer, and red blood cells were lysed as

TH2 Polarization of Spleen CD4+ T Cells Ex Vivo
F(ab')2 anti-CD3 (clone145-2C11 at 10 µg/ml) and anti-CD28 (clone PV-1 at 2 µg/ml) antibodies (BioXcell, Lebanon, NH, USA) were coated overnight at 4°C in 96-well plates (Costar) in sterilefiltered 20-mM carbonate buffer (pH 9.6) containing 2 mM of MgCl 2 and 0.01% NaN 3 . Wells were washed in PBS before plating the cells. Total spleen CD4 + T cells were sorted by magnetic negative selection following the manufacturer's instructions (STEMCELL Technologies, Vancouver, Canada). CD4 + T cells were then resuspended in a culture medium (same as above) at 0.25 million cells/ml. Two hundred microliters of cell suspension were then added to wells and supplemented with 10 ng/ml of IL-2 and 50 ng/mL of IL-4 (both from BioLegend, San Diego, CA, USA). After 2 days in the culture at 37°C and 5% CO 2 , IL-2 and IL-4 were added again at the same concentrations for 3 more days. TH2-polarized CD4 + T cells were then stimulated for 4 h with 40 nM of PMA and 800 nM of ionomycin in the presence of 2 µg/ ml of brefeldin A (Sigma-Aldrich, Merck). Flow cytometry analysis was then performed as described above.

Detection of Anti-RNP Autoantibodies
Maxisorp plates (Thermo Scientific) were coated overnight at 4°C with 10 µg/ml of purified RNP complexes (Immunovision, Springdale, AR, USA) diluted in carbonate buffer (100 mM of NaHCO 3 and 30 mM of Na 2 CO 3 , pH 9.6). Plates were washed 3 times in 300 µl of PBS containing 0.05% Tween-20 (PBS-T, Bio-Rad Laboratories) and then saturated for 1 h at room temperature with 200 µl of PBS containing 5% FCS (Gibco, Waltham, MA, USA). Plates were then washed 3 times in 300 µl of PBS-T. All plasma samples were diluted 1:100 in PBS-T containing 5% FCS and 100 µl added to the wells and incubated for 2 h at room temperature. After 5 washes in PBS-T, a secondary antibody goat anti-mouse IgG conjugated to horseradish peroxidase (HRP) (Invitrogen) was diluted in PBS-T at 5% FCS at a final concentration of 500 ng/ml, and 100 µl was added to the wells for 1 h at room temperature. The plates were finally washed 5 times in 300 µl of PBS-T. A total of 100 µl of tetramethylbenzidine (TMB) substrate (ThermoFisher Scientific, Waltham, MA, USA) was added to the wells, incubated at room temperature for at least 20 min, and the reaction was stopped by adding 100 µl of 0.2 N of sulfuric acid solution. Optical density at 450 nm was measured by spectrophotometry (Infinite 200 Pro plate reader, Tecan, Männedorf, Germany). On each plate, similar negative and positive controls were run. Optical density (OD) values were first normalized to the negative controls, and then, the presented results were normalized to the mean of the PBS-injected control mice values and expressed in arbitrary units.

Statistics
Student's unpaired t-tests were used to compare the differences of one variable between two groups when distributions were Gaussian and Mann-Whitney U tests for nonparametric distributions. When more than two groups were compared, oneway ANOVA coupled with Tukey's multiple comparisons test was used. Statistics were done using Prism v9 software (Graphpad).

CT-M8: A New Basophil-Specific Mouse Model
We aimed at developing a basophil-specific mouse model in which we could both follow basophil localization by immunofluorescence and flow cytometry and express the CRE recombinase selectively in the basophil compartment. To achieve that goal, we took advantage of the gene encoding for the basophil-specific protease Mcpt8 (8). The strategy described in the Methods section and Figure 1A led to the generation of Mcpt8 tm1Ics mice named here "CT-M8" (Mcpt8 CT/CT ), standing for CRE recombinase and tdTomato expressions (CT) under the control of Mcpt8 gene transcription (M8). CT-M8 mice had normal fertility and did not show any dysregulation of their immune cell distribution in lymphoid organs or blood ( Table 1). As expected, the construct allowed straightforward basophil detection by flow cytometry ( Figure 1B). Direct tdTomato + basophil gating on living singlet CD45 + cells was as efficient to detect basophils as a conventional gating strategy for basophils (CD45 lo CD49b + FcϵRIa + CD123 + CD200R3 + ) ( Figure 1B; Table 1). By confocal microscopy, basophil detection was evident in CT-M8 mice, and tdTomato expression in the spleen was restricted to IgE-bearing cells ( Figure 1C; Supplementary Figure S1). Of note, tdTomato expression in Mcpt8 CT/+ basophils was half that in Mcpt8 CT/CT basophils ( Figures 1B, D). Although inserted after the stop codon, the construct inserted into the Mcpt8 locus led to a drastic reduction of Mcpt8 mRNA expression as measured by qPCR in bone marrow (BM) basophils isolated from Mcpt8 CT/CT mice. However, Mcpt8 mRNA expression was still significantly retained in Mcpt8 CT/+ basophils ( Figure 1E).
tdTomato expression was strictly restricted to the basophil population in all tested organs (blood, peritoneal lavage, lymph nodes, bone marrow, and spleen) ( Figure 1D; Table 1). Spleen basophils from Mcpt8 CT/CT and Mcpt8 CT/+ mice produced IL-4 and IL-6 to the same extent as WT (Mcpt8 +/+ ) basophils after phorbol myristate acetate (PMA) and ionomycin stimulation, demonstrating that construct expression did not alter the maximal abilities of the basophils from CT-M8 mice to produce these cytokines (Figures 2A-C). CD200R1 is a recognized mouse basophil activation marker (13,14). Anti-IgE-mediated basophil activation resulted in similar CD200R1 upregulation on spleen basophils from Mcpt8 +/+ and Mcpt8 CT/CT mice and to similar IL-4 and IL-6 productions ( Figures 2D-J). Furthermore, IL-3-mediated activation was as efficient in Mcpt8 +/+ as in Mcpt8 CT/CT basophils to induce IL-4 and IL-6 productions ( Figures 2K, L).
Overall, these data demonstrated that the CT-M8 mice (i) allowed convenient and selective basophil detection by microscopy and flow cytometry, (ii) did not show any abnormalities in their immune cell populations, and (iii) had a tdTomato expression restricted to basophils in all the compartments analyzed, and that basophils from these mice, despite an alteration of Mcpt8 expression, exhibited normal

CRE Expression in CT-M8 Mice Is Basophil-Specific and Functional
We next sought to verify that functional CRE recombinase expression occurred only in basophils and allowed induction of basophil-specific floxed gene recombination. To that end, Mcpt8 CT/+ mice were bred with Il4 fl/fl mice (11). Unlike IL-6 production, which was not altered, PMA-ionomycin-induced IL-4 production by basophils from Mcpt8 CT/+ Il4 fl/fl mice was completely abrogated as compared to Mcpt8 CT/+ Il4 +/+ and Mcpt8 +/+ Il4 fl/fl mice ( Figure 3A). Thus, CRE recombinase was functional, and its expression was restricted to the basophil compartment since IL-4 production in in vitro polarized T helper type 2 (TH2) CD4 + T cells was not impaired in Mcpt8 CT/+ Il4 fl/fl mice as compared to Mcpt8 CT/+ Il4 +/+ mice ( Figure 3B). Next, we bred the CT-M8 mice with the ROSA-DTA mice (9). Mcpt8 CT/+ Rosa26 DTA/+ mice were constitutively basophildeficient and did not show any other immune cell population defects in homeostatic conditions as compared to Mcpt8 CT/+ Rosa26 +/+ and Mcpt8 +/+ Rosa26 DTA/DTA mice in all analyzed compartments ( Table 2; Figures 4A-C).
Overall, these results confirmed the basophil-restricted expression of a functional CRE recombinase in CT-M8 mice that allowed generating basophil-specific ablation of selected genes and a constitutive basophil-deficient mouse model.

Basophils Have a Nonredundant Role in Lupus-Like Nephritis Amplification
By accumulating in SLOs during lupus pathogenesis, we previously showed that basophils support the survival of autoreactive antibody-secreting cells, leading to the amplification of autoantibody titers both in lupus-like mouse models and in SLE patient cohorts (4, 6, 7). Pristane (TMPD; 2, 6, 10, 14-tetramethylpentadecane) is a mineral oil able to induce the development of a lupus-like phenotype when injected intraperitoneally (ip) into C57BL/6 mice (15). The resulting PIL was associated with peripheral basophilia and an accumulation of basophils in SLOs. When depleting basophils using different approaches over 14 days in C57BL/6 female mice in established PIL (24 weeks after pristane injection), the accumulation of autoantibody-secreting plasma cells and amplification of autoantibody production were dampened (7). However, these approaches did not allow for evaluating the contribution of basophils to the onset of the disease.
We, therefore, took advantage of our new model of constitutive basophil-deficient mice described above to address this question. Basophil-sufficient (Mcpt8 CT/CT Rosa26 +/+ ) and basophil-deficient (Mcpt8 CT/+ Rosa26 DTA/+ ) female mice were injected ip with pristane at around 8 weeks of age and euthanized for analysis 8 weeks after injection. This time point corresponded to an early stage of PIL since neither proteinuria nor other biological parameters evidencing kidney dysfunction (blood urea nitrogen and creatininemia) were detectable (data not shown). However, 8 weeks after pristane injection in basophil-sufficient animals, basophil and short-lived plasma cell (CD19 + CD138 + B cells) accumulation in SLOs (spleen and lymph nodes), peripheral basophilia, and autoantibodies in plasma were detected ( Figures 4A-F). This was associated with the detection of complement component C3 and IgG deposits in the kidney glomeruli of pristane-injected mice, unlike the glomeruli of PBSinjected control mice, confirming the early development stages of the pristane-induced lupus-like nephritis (Figures 4G, H).
As expected, no basophil was detected in basophil-deficient mice (Mcpt8 CT/+ Rosa26 DTA/+ ) 8 weeks after pristane injection in any of the compartments analyzed ( Figures 4A-C). No accumulation of CD19 + CD138 + short-lived plasma cells was detected in SLOs of pristane-injected basophil-deficient mice, suggesting that basophil deficiency prevented PIL development ( Figures 4D, E). Indeed, no significant increase in antiribonucleoprotein (anti-RNP) antibodies, autoantibodies present in PIL in C57BL/6 mice (6, 7), was measurable in the blood of basophil-deficient mice ( Figure 4F). Moreover, no significant C3 and IgG deposits in the glomeruli of basophildeficient mice were detected 8 weeks after pristane injection ( Figures 4G, H). ) mice were injected ip with PBS or pristane (PIL). Eight weeks later, the mice were euthanized and analyzed.
(A-C) Basophil accumulation in the spleen (A) and peripheral lymph nodes (LNs) (cervical, axillary, and inguinal) (B) and blood basophils (C) were quantified by flow cytometry, as shown in Figure 1 and described in the Methods section. Basophils were defined as CD45 lo CD3 − CD19 − CD117 − CD49b + FcϵRIa + tdTomato + cells among CD45 + viable singlets. (D, E) CD19 + CD138 + plasma cells were quantified in the spleen (D) and peripheral lymph nodes (E) of the mice described above. (F) Anti-RNP IgG autoantibody titers were quantified by ELISA from plasma samples of indicated mice, as described in the Methods section. The O.D. values at 450 nm were determined, and data were normalized to the mean of PBS-injected Mcpt8 CT/CT Rosa26 +/+ values. A.U., arbitrary units. (G, H) Four-micrometer cryosections of kidneys were analyzed by immunofluorescence for C3 (G) and IgG (H) staining as shown on images (scale bar, 60 µm) and quantified in adjacent bar graphs as A.U., corresponding to the ratio between fluorescence intensity measured in glomeruli to that in interstitial background. (A-H) Results are from at least three independent experiments and are presented as individual values in bars representing the mean values. Statistical analyses were done by one-way ANOVA followed by Tukey's multiple comparison tests between the indicated groups. NS, not significant, p > 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
Overall, these results demonstrate that basophils have a nonredundant role in PIL development as basophil deficiency sufficed to prevent the onset of autoantibody production and lupus-like nephritis induced by pristane injection.

DISCUSSION
Beyond its role in allergic and parasitic diseases, type 2 immunity is now identified as a key player in the pathophysiology of various autoimmune diseases (16,17). Several studies have demonstrated its relevance to SLE pathogenesis and identified IgE and basophils as promising therapeutic targets due to their role in amplifying both autoantibody production and tissue damage. Most of these studies are based on a translational approach that allowed us to study and decipher in lupus-like mouse models what was observed in human patients (4,6,(18)(19)(20). The use of a novel mouse model allowing basophil-specific short-term depletion (Mcpt8 DTR mice (21)) demonstrated that basophil targeting efficiently dampened lupus-like activity in mice with established disease (6). Here, using a new basophildeficient mouse model, we show that basophils were required to reach a pathogenic threshold in PIL, supporting their nonredundant role in early lupus-like disease development.
Besides specific depletion of basophils, the new CT-M8 mouse model also enables convenient detection of basophils by flow cytometry and confocal microscopy as well as the selective depletion of a floxed gene in the basophil compartment. In particular, our analysis showed that constitutive basophil deficiency in this model did not lead to any identified immunological bias. Likewise, mice rendered IL-4 deficient in the basophil compartment did not impact the ability of TH2polarized CD4 + T cells from the same mice to produce IL-4. Unexpectedly, Mcpt8 mRNA expression was dramatically dampened by the insertion of the IRES-based construct after the stop codon site of the Mcpt8 gene. If a knock-in approach would be expected to induce Mcpt8 deficiency (22), the reason explaining this alteration of Mcpt8 expression in our model will need further investigation. Fortunately, Mcpt8 expression was still detected in Mcpt8 CT/+ heterozygous mice, allowing for the generation of Mcpt8-sufficient mice expressing both tdTomato and CRE recombinase only in the basophil compartment. Of note, Mcpt8-reduced expression is known not to induce any significant alteration in basophil function in mouse models of parasite infection (22) or sepsis (12).
To our knowledge, three other mouse models expressing the CRE recombinase selectively in basophils are currently available. Generated in 2010 by Voehringer and colleagues, the "Mcpt8CRE" mouse is a BAC transgenic mouse strain expressing the Cre recombinase inserted behind the start codon of the Mcpt8 gene. Unfortunately, these mice showed a constitutive 90% basophil deficiency, probably due to CRE overexpression preventing its use for floxed gene deletion selectively in basophils (23). The "Basoph8" mice were developed by Locksley and colleagues, who replaced the Mcpt8 gene at the start site by inserting a reporter cassette containing a sequence encoding the yellow fluorescent protein (YFP), followed by an IRES and a sequence encoding humanized Cre recombinase. This mouse strain is effective at basophil-specific Cre-mediated gene deletion and expresses constitutively the YFP in basophils (22). The third one was recently developed by Karasuyama and colleagues and called Mcpt8 iCre/+ . This mouse strain expresses the improved Cre (iCre) recombinase inserted to replace the first exon of the Mcpt8 gene and is efficient in basophil-specific Cre-mediated gene deletion but does not express constitutively any fluorescent reporter gene (11).
The advantages of the CT-M8 mice are then similar to the ones of the Basoph8 mouse model (22), but they express another fluorescent protein (tdTomato), allowing distinct breeding strategies for the generation of multiple gene/cell-specific reporter mouse strains. In addition, the CT-M8 mice (in their heterozygous Mcpt8 CT/+ strain) retain significant expression of the Mcpt8 gene while allowing fully sufficient functional expression of the CRE recombinase.
Mast cells and basophils share immediate common progenitors, and Mcpt8 gene-controlled CRE or DTR expression may be detected at low levels in other cell types including mast cells, granulocyte-macrophage progenitors, T cells, B cells, NK cells, eosinophils, and dendritic cells (23)(24)(25). Therefore, both the construct used to generate CT-M8 mice and DTa expression in Mcpt8-expressing cells could have affected immune cell populations other than basophils. However, analysis of CT-M8 mice showed that tdTomato expression was fully restricted to the basophil compartment ( Table 1). As well, Mcpt8-dependent DTa expression in Mcpt8 CT/+ R 26 -Stop fl/WT -DTA mice did not impact any immune cell populations other than basophils ( Table 2). The very low Mcpt8 mRNA expression levels in the former cell populations, or their progenitors, as compared to mature basophils, may explain the basophil-selectivity observed in the various Mcpt8 gene-based mouse models.
In the present study, basophil deficiency prevented the development of the pristane-induced lupus-like disease. Indeed, our data suggest that basophils are necessary to reach a pathogenic threshold in early PIL. This is evidenced by the ablation of plasmablast accumulation in SLOs, autoantibody production, and glomerular deposition of circulating immune complexes in basophil-deficient mice.
Further investigations will be required to determine the mechanisms by which basophils contribute to autoimmune evolution in the PIL. In this context, our new basophil-specific mouse model will allow selective depletion in the basophil compartment of candidate genes. For instance, it will allow deciphering of the axis of autoreactive IgE and FcϵRI-mediated basophil activation in the nonredundant role of basophils in lupuslike pathogenesis. Indeed, autoreactive IgE is highly prevalent in SLE patients with active disease (26), and the basophil activation status correlates strongly with disease activity (4,6). Furthermore, IgE itself is known to exert an immunoregulatory role in various lupus-like mouse models (4, 18). We will now be able to deplete specifically FcϵRI in basophils by breeding CT-M8 mice with FcϵRIg fl/fl mice (C57BL/6J-Fcer1g em1Gfng/J , Jax mice #036592) and demonstrate whether basophil involvement in lupus-like disease amplification depends on FcϵRI-mediated signals. The role of factors produced by basophils during the course of the disease, such as IL-4 or IL-6, could be assessed as well. Although technically challenging, basophil reconstitution of basophil-deficient animals through frequent adoptive transfer of purified basophils during the whole procedure may further help to decipher the non-redundant role of basophils in PIL onset.
The nonredundant role of basophils in lupus-like pathogenesis demonstrated in PIL will need to be validated in other lupus-like mouse models, including spontaneous models. This preclinical development is necessary to determine whether targeting basophils or basophil-activating factors in recently diagnosed SLE patients may be a promising therapeutic strategy to prevent the development of lesions in targeted organs. Several approaches may be envisaged to alter the basophil contribution to lupus nephritis development, including targeting IgE, PTGDRs, or directly basophils through a basophil-specific surface marker that still needs to be identified. Basophils have a deleterious role in various allergic, inflammatory, and autoimmune diseases (16,27). Identifying therapeutic strategies targeting these cells, their mediators, or the factors involved in their contribution to these diseases may represent a promising area of clinical development.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by the comited 'ethique Paris Nord N°121 and the Ministère de l'enseignement supeŕieur, de la recherche et de l'innovation under the authorization number APAFIS#14115.

AUTHOR CONTRIBUTIONS
JT designed the experiments, conducted the experiments, and wrote the manuscript. NC conceived the project, designed the experiments, conducted the experiments, wrote the manuscript, and directed the project. QS, LC, CP, UB, MB, and ED conducted the experiments, analyzed the data, and/or edited the manuscript. HK and KM provided the Il4 fl/fl mice and edited the manuscript. NC had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.