The expression of both Regnase-1 and -3 is induced in acutely stimulated mast cells

To determine how the expression of RBP-encoding genes changes in mast cells upon stimulation with IgE and antigen complexes, we first re-analyzed RNA-seq datasets of murine bone marrow–derived mast cells (BMMCs) stimulated for 2 h with IgE and antigen complexes from Li et al (2021), focusing only on RBP-encoding genes as defined in the RBP2go database (Caudron-Herger et al, 2021) and including only RBPs detected in more than one dataset (Fig S1A and Table S1). This resulted in overall 827 differentially expressed transcripts, of which 435 were associated with general RNA metabolic processes at gene ontology (GO) analysis. Within this category, RBP transcripts divided into further functional categories linked primarily to RNA processing, translation, and stability (Fig S1B). Next, we found that Zc3h12c (Regnase-3) was among the most highly induced RBP-encoding transcripts after stimulation (Fig S1A). Two other Regnase family members, Zc3h12a (Regnase-1) and Zc3h12d (Regnase-4), were also induced, whereas Zc3h12b (Regnase-2) was downmodulated. Re-analysis of ATAC-seq data from Li et al (2021) also showed increased accessibility around the Zc3h12a and Zc3h12c promoters upon IgE crosslinking, suggesting transcriptional activation after acute stimulation (Fig S1C). Analysis of the expression of Zc3h12a-d in mast cells extracted ex vivo from different mouse tissues (ImmGen database) (Dwyer et al, 2016) also revealed the high expression of Zc3h12c in most tissue-derived mast cells (notably from the skin, less prominently from the peritoneal cavity), followed by the modest expression of Zc3h12a and very low levels of Zc3h12b and Zc3h12d (Fig S1D). Finally, data mining of published RNA-seq data from human mast cells revealed induction of ZC3H12A and ZC3H12C expression in stimulated mast cells obtained from the human skin (Gao et al, 2023). Similar results were obtained by RNA-seq of stimulated human peripheral blood–derived mast cells (Cildir et al, 2019) (Fig S1E). Overall, these observations point toward a role of selected Regnase family members in modulating mast cell functions, and we therefore focused our attention on this family of RBPs.

To validate and extend these findings, we first measured the expression of the Regnase-encoding transcripts in BMMCs (∼99% pure populations, determined by FcεRIα and c-Kit staining; Fig S1F) activated with IgE and antigen complexes for the indicated times. Zc3h12b and Zc3h12d were expressed at very low levels and were not significantly induced upon stimulation, whereas both Zc3h12a and Zc3h12c were expressed and highly induced at early time points (within the first 1–2 h) in activated mast cells (Fig 1A). Similar significant induction specifically of Zc3h12a and Zc3h12c was observed in mast cells separated ex vivo from the peritoneal cavity and stimulated with PMA and ionomycin (Figs 1B and S1G). Regnase-1 and -3 were also induced at the protein level upon activation of BMMCs and peritoneal cavity–derived, in vitro–expanded mast cells (Fig 1C and D), pointing toward a role of these two Regnase family members in regulating mast cell responses to stimuli. We also detected the expected proteolytic cleavage of Regnase-1 upon activation, consistent with Malt-1 activation downstream of the IgE receptor complex (Klemm et al, 2006; Uehata et al, 2013). Whether Regnase-3 is regulated post-translationally remains to be understood, especially considering the complex pattern revealed by Western blot (Fig 1C and D), as reported also in the literature (von Gamm et al, 2019). In terms of subcellular localization, both Regnase-1 and -3 were detected within punctuated structures in the cytoplasm of acutely activated mast cells, with a limited overlap (Fig S2). Because mast cells can be activated by a large variety of signals, we determined whether stimuli other than IgE and antigen complexes affected Regnase expression. We found that both Zc3h12a and Zc3h12c were induced, albeit at different extents, by inflammatory signals that induce cytokine expression in mast cells, including IL-33 and LPS, but not by the c-Kit ligand stem cell factor (SCF), which by itself did not induce cytokine expression (Figs 1E and S3). Overall, these data point toward a role of Regnase-1 and -3 in modulating mast cell activation and/or effector functions that we set out to investigate.

Regnase-1 and -3 restrain the expression of inflammatory cytokines in mast cells

To gain a broad understanding of the impact of Regnase-1 and -3 on the expression inflammatory transcripts, we first used siRNAs to deplete both factors (schematics of siRNA targeting strategy in Fig 2A). Depletion of both Regnase-1 and -3, alone and in combination, was highly effective at the mRNA and protein levels (Fig 2B and C). Notably, depletion of Regnase-3 led to the significantly increased expression of Regnase-1, pointing toward a role of Regnase-3 in restraining Regnase-1 expression (Fig 2C), consistent with previous data described in macrophages (Liu et al, 2021; von Gamm et al, 2019). Next, we stimulated the cells for 2 h with IgE and antigen, and we measured the expression of 734 immune transcripts by NanoString digital profiling. We chose this method because we were primarily interested in immune-related inflammatory transcripts. We found that 325 of these transcripts were expressed by mast cells and that the depletion of Regnase-1, alone or in combination with Regnase-3, led to the increased expression of mRNAs encoding inflammatory mediators, including established Regnase-1 targets such as Nfkbiz and Il2 (Uehata et al, 2013; Jeltsch et al, 2014) (Fig 2D and Table S2). Differently from Regnase-1, depletion of Regnase-3 alone had only a very modest impact on the inflammatory transcriptome of mast cells, with only nine genes significantly up-regulated (log₂ fold change >0.5 and P ≤ 0.05) (Fig 2D). Depletion of both proteins at the same time led to an overall combined phenotype that enhanced the effect of the individual depletions, with transcripts such as Tnf, Il1b, Il2, and Nfkbiz becoming even more prominently dysregulated (Fig 2D), suggesting redundant functions at least on specific targets. Overall, we found that Regnase-1, and to a lesser extent Regnase-3, restrains the expression of inflammatory transcripts in mast cells. The down-regulated transcripts were likely to be the result of indirect effects of Regnase depletion, and no previously described Regnase-bound, direct targets (Mino et al, 2015) were identified among the down-regulated transcripts (Fig 2E). Vice versa, among the significant up-regulated genes several direct Regnase-1 targets could be observed (red asterisks, Fig 2E), suggesting that our analysis identified transcripts at least in part directly modulated by Regnase proteins. Interestingly, Il2 and Tnf were among the few inflammatory transcripts significantly increased upon depletion of both Regnase-1 and -3, alone or in combination. Because Il2 was expressed at low levels in mast cells, we further investigated the impact of Regnases on TNF expression. First, we confirmed the increased expression of the Tnf transcript in an independent set of experiments (Fig 3A). Next, we found that, concordant with the mRNA data, TNF protein expression was significantly increased upon depletion of Regnase-3, alone or in combination with Regnase-1, whereas the impact of Regnase-1 depletion alone was modest (Fig 3B). Next, to confirm these findings using an independent experimental setup, we used CRISPR/Cas9-mediated gene editing to deplete the expression of Zc3h12a or Zc3h12c (Fig 3C). Depletion of either protein was highly effective in mast cells, as shown by Western blot (Fig 3D). Concordant with our siRNA data, depletion of both Regnase-1 and -3 led to significantly increased Tnf mRNA expression (Fig 3E) and to increased protein expression, measured by intracellular staining, both in response to IgE and antigen complexes or PMA and ionomycin stimulation (Fig 3F). We also measured the expression of other cytokines prominently produced by mast cells. We found that IL-6 expression was modestly affected by Regnase-1 deletion (Fig S4A), consistent with our NanoString data, showing mildly increased expression, and it was unaffected by Regnase-3 deletion. Although at the mRNA level we could not detect any effect on the Il13 transcript, IL-13 expression was mostly increased upon deletion of either Regnase, suggesting indirect regulation.

To confirm these data, we sought to force the expression of Regnase enzymes in mast cells. However, we found that most overexpression methods, including inducible vectors, failed to provide stable Regnase-1 and -3 protein expression, because of deleterious effects, especially of Regnase-1 on mast cell viability. We therefore optimized an in vitro transcription (IVT) system for the direct, short-term delivery of mRNAs encoding Regnase-1 and -3. The T7-driven transcripts were stabilized by the addition of the 5′UTR and 3′UTR from the HBB gene, followed by 5′ capping and 3′ poly(A) tailing. We generated transcripts leading to the expression of HA-tagged WT Regnase-1 and -3 and versions carrying point mutations in the RNase enzymatic domain (D141N and D252N, respectively), leading to disruption of the catalytic activity (Matsushita et al, 2009). mRNA delivery of Regnase-1 and -3, either WT or RNase mutant, proved to be feasible and efficient, and all proteins were highly expressed in transfected mast cells, as measured by intracellular staining using an antibody against the HA-tag (Figs 3G and S4B). Cell viability was comparable across conditions (Fig S4B), and the overall subcellular localization of the IVT mRNA-derived proteins was comparable to that of endogenously expressed Regnase-1 and -3 (Fig S5). Analysis of cytokine expression by these cells upon stimulation with either IgE and antigen complexes or PMA and ionomycin revealed that both Regnase-1 and -3 significantly diminished the ability of mast cells to produce TNF in an RNase-dependent manner (Fig 3H). Similar results were obtained by measuring TNF release in the culture supernatant by ELISA (Fig S4C). Compared with Regnase-1, the somewhat reduced capability of Regnase-3 to limit TNF expression was likely linked to the extent of overexpression that we could achieve in this experimental system, which was less pronounced for Regnase-3 (Fig 3G). Mostly consistent with our CRISPR/Cas9 knockout data, the overexpression of Regnase-3 had no detectable effects on IL-6 production, whereas Regnase-1 limited both IL-6 and IL-13 expression by mast cells (Fig S6A). Next, we used the human mast cell lines HMC-1.1 and HMC-1.2 (Sundström et al, 2003) to assess the impact of Regnase-1 on TNF expression in the human system. We found that IVT transfection of human Regnase-1, but not a catalytically inactive version (D141N), significantly reduced TNF expression by human mast cells (Fig S6B and C), indicating that Regnase-dependent regulation of TNF expression is relevant also in the context of human mast cells. To determine whether the effect of Regnase-1 and -3 on murine TNF expression was mediated by the direct targeting of the Tnf transcript, we cloned the full-length 3′UTR of Tnf into a luciferase reporter vector. We then co-transfected it with Regnase-1– or Regnase-3–expressing vectors, either WT or RNase-impaired. We found that the Tnf 3′UTR was strongly affected by both Regnases and that luciferase expression was restored upon mutation of their RNase enzymatic sites, pointing toward a direct effect on Tnf that required the RNase enzymatic activity of Regnase proteins (Fig 3I). Finally, actinomycin D treatment of cells transfected with IVT mRNA encoding Regnase-1 or Regnase-3 showed reduced stability over time of the Tnf mRNA (Fig 3J). Again, this effect was more prominent for Regnase-1, and potentially linked to the higher level of expression that we could achieve in the IVT transfection system, compared with Regnase-3. Overall, both Regnase-1 and -3 are required to effectively limit TNF expression by mast cells, although their effect on other inflammatory cytokines appears to be more selective and predominantly linked to Regnase-1.

Regnase-3 is required to restrain the expression of Regnase-1 in mast cells

Depletion of Regnase-3 by RNAi led to significantly increased Regnase-1 expression in mast cells (Fig 2B and C), indicating that understanding the phenotype of either knockdown or knockout is complicated by the presence of direct and indirect effects and cross-regulation between these two factors. To better dissect the role of Regnase-3 in regulating Regnase-1 expression, we first confirmed the effect of Regnase-3 on the Zc3h12a mRNA by deleting Zc3h12c by CRISPR/Cas9, showing that Zc3h12a mRNA expression was indeed significantly increased (Fig 4A). Concordant with this result, we found that upon overexpression of Regnase-3 by IVT mRNA transfection, Regnase-1 protein expression was significantly reduced (Fig 4B). Regnase-1 is also known to diminish the stability of its own transcript through stem–loop structures in the 3′UTR (Iwasaki et al, 2011; Tse et al, 2022). To assess the impact of the individual Regnases on Zc3h12a expression, we cloned the 3′UTR of Zc3h12a in a luciferase reporter plasmid and measured its response upon co-transfection of Regnase-1 or Regnase-3. Both proteins were able to strongly affect luciferase expression in an RNase activity–dependent manner, suggesting a direct effect on the 3′UTR of Zc3h12a (Fig 4C). Finally, to determine whether Regnase-3 was required to modulate Zc3h12a transcript stability in mast cells, we treated the cells with actinomycin D, followed by the measurement of Zc3h12a mRNA levels over time. Deletion of Regnase-3 was sufficient to increase the half-life of the Zc3h12a transcript in mast cells (Fig 4D), whereas the overexpression of Regnase-3 by IVT mRNA transfection reduced it (Fig 4E). On the contrary, we found no evidence of any impact of Regnase-1 depletion or deletion on Regnase-3 (Fig 4F).

To gain more insights into the crosstalk between Regnase-1 and -3, we performed immunoprecipitation and mass spectrometry analyses of the endogenously expressed proteins in mast cells activated with PMA and ionomycin (Figs 4G and H and S7A, and Table S3). We found that immunoprecipitation of Regnase-1 recovered Regnase-3 as its primary endogenous interactor (Fig 4G). This interaction was confirmed in HEK293T cells upon co-transfection of Regnase-1 and Regnase-3 (Figs 4I and S7B), and in BMMCs sequentially transduced to overexpress catalytically inactive Regnase-1 D141N and Regnase-3 D252N (Figs 4J and S7C). Similar results were obtained using an anti-FLAG antibody to immunoprecipitate FLAG-tagged Regnase-1 or Regnase-3 (Fig S7D). Furthermore, the interaction between Regnase-1 and -3 was unaffected by RNase treatment, suggesting that it does not require RNA binding to occur (Fig S7E). On the contrary, Regnase-3 interacted primarily with ribosomal proteins, suggesting an involvement in translation, and with the 14-3-3 family of proteins (Fig 4H). 14-3-3 proteins were previously shown to interact with FLAG-HA-tagged Regnase-1 when transfected into HeLa cells and to contribute to Regnase-1 protein stability (Akaki et al, 2021). Because in that system Regnase-3 was not expressed, it could not be recovered as an interactor of Regnase-1. However, in our experimental setting in which both Regnase family members are expressed, we found that complex formation involves primarily Regnase-1 and -3 on the one side and Regnase-3 and 14-3-3 proteins on the other, although cell- and stimulus-specific differences may also apply. The fact that in mast cells Regnase-3 was identified as an interactor of Regnase-1, but not the opposite, suggests that Regnase-3 is predominantly interacting with other protein partners, and only a relatively small proportion remains available for Regnase-1 interaction. To determine which region(s) of the Regnase-3 protein are important for its interaction with Regnase-1, we generated truncation mutants lacking different domains of Regnase-3 (Fig 4K). We then co-transfected HEK293T cells and confirmed protein expression by immunoblot (Fig 4L). Next, immunoprecipitation of Regnase-1 recovered Regnase-3 as an interacting partner of all Regnase-3 constructs except for the shortest (1–245aa) protein, lacking the PIN domain (Fig 4M). This is consistent with the PIN domain being important for protein oligomerization, as shown previously for Regnase-1 (Yokogawa et al, 2016). In sum, our data reveal that Regnase-3 is a direct interactor and regulator of Regnase-1 expression in mast cells.

Regnase-1, but not Regnase-3, is required to maintain mast cell proliferation and viability

Having established the role of Regnase-1 and -3 in regulating inflammatory responses in mast cells, we investigated their role in modulating mast cell growth and survival. First, depletion of Regnase-1, but not Regnase-3, was sufficient to reduce cell viability 72 h after siRNA transfection, measured by LIVE/DEAD staining (Fig 5A). Consistent with this observation, measuring the extent of apoptosis by Annexin V staining revealed a significant increase in the percentage of early apoptotic cells upon Regnase-1 depletion (Fig 5B), as well as diminished ability of the cells to proliferate, measured by BrdU incorporation (Fig 5C). Cell cycle analysis confirmed an increased percentage of cells in G₁ and a concomitant decrease in the percentage of cells in the S phase (Fig 5D). A similar increase in cell death was observed upon deletion (by CRISPR/Cas9) of Zc3h12a, which remained significant regardless of the presence of IL-3 (an essential survival factor for mast cells) in the culture medium (Figs 5E and S8A) and was accompanied by reduced overall proliferation (Figs 5F and S8B). Meanwhile, deletion of Regnase-1 did not affect the capacity of the cells to degranulate in response to IgE and antigen stimulation (Fig S8C).

Because the limited gene set in the NanoString panel provided mainly information relative to inflammatory genes, to gain insights into the mechanism(s) that may lead to such pronounced impairment in the ability of mast cells to survive in the absence of Regnase-1, we deleted Regnase-1 by CRISPR/Cas9 and performed RNA sequencing. We found that 131 genes were up-regulated, whereas 187 genes were down-regulated in the absence of Regnase-1, pointing toward both direct and indirect effects (Fig 6A and Table S4). Among the up-regulated genes, the basal expression of Tnf was increased, even in the absence of stimulation. Some previously reported direct targets of Regnase-1 (Mino et al, 2015), including Nfkbid and Ran, were also modestly up-regulated (Table S4). Transcripts encoding the mast cell protease Mcpt2 and the metalloprotease Adam8 were increased, whereas some genes involved in regulating mast cell degranulation such as Hdc (Nakazawa et al, 2014) were down-regulated. GO analysis of the down-regulated genes clustered around signal transduction, as well as cholesterol and lipid metabolism (Fig 6B). These genes include Scd1, Scd2, Acsl3, and Fads2, which are important for lipid synthesis, and Abcg1, Abca1, Pcsk9, and Fabp5, which are involved in lipid transport, pointing toward overall dysregulation in lipid homeostasis. Metabolism-related genes downmodulated in the absence of Regnase-1 also included Aldoc, Pdk1, and Hk2, which play a role in glucose metabolism. These results point toward involvement of Regnase-1 in regulating metabolic processes required for cell maintenance, as observed also in other cell types (Younce et al, 2009; Nagahama et al, 2018; Behrens et al, 2021; Reina-Campos et al, 2023). On the contrary, categories associated with the up-regulated genes revealed many GO terms associated with the control of cell cycle, and DNA replication and repair (Fig 6B). Indeed, some of the most differentially expressed genes included factors important for cell cycle progression, such as Cdca7, Cks1b, and Cdk1. Although classical anti- and pro-apoptosis–related genes, such as Bcl2, Bax, and Bcl2l1, were unchanged, DNA damage and replication markers such as Aunip, Ung, Pold1, and H2ax were up-regulated, suggesting that Regnase-1 knockout cells were more susceptible to cellular stress, in addition to cell cycle defects.

To verify this phenotype using an independent experimental approach, we took advantage of mast cells derived from the bone marrow of Zc3h12a^fl/fl mice (Li et al, 2017). Briefly, we transduced bone marrow cells with a lentiviral vector expressing an EGFP-Cre fusion protein (Leoni et al, 2023) reaching at least 50% efficiency of transduction (Fig 6C). After sorting of EGFP⁺ cells, the expression of Regnase-1 became undetectable by Western blot (Fig 6D). Similar to cells treated with RNAi or Cas9 RNPs, Zc3h12a^fl/fl-Cre mast cells showed reduced viability in different stress conditions, including the withdrawal of the survival factor IL-3 and short-term UV irradiation (Figs 6E and F and S8D and E). Proliferation was similarly reduced in these cells even in conditions that optimally sustain mast cell growth (addition of IL-3, with or without SCF) (Figs 6G and S8F). Overall, using numerous independent experimental approaches, we found that Regnase-1 is central to the maintenance of mast cell survival and basal proliferation, and that Regnase-1 deletion leads to widespread transcriptome changes associated with cell cycle defects.

Mice

All mice used in this study were on a C57BL/6 background, housed in a specific pathogen-free barrier facility under 12-h dark/light cycles, 20–24°C temperature, and 50–65% humidity conditions. All animal studies were performed in accordance with Swiss Federal Veterinary Office guidelines and with approval from the Cantonal animal experimentation committee, Dipartimento della Sanità e della Socialità Cantone Ticino (authorization number TI10/19).

Cell cultures

BMMCs were in vitro–differentiated from the bone marrow of 6- to 8-wk-old mice. Whole bone marrow cells were cultured in IMDM (Gibco) containing 10% heat-inactivated FBS (Gibco) and IL-3 (recombinant, prepared in-house), and supplemented with 0.1 mM non-essential amino acids (Gibco), 2 mM L-alanyl-L-glutamine dipeptide (Gibco), 100 U/ml penicillin, 100 mg/ml streptomycin (Gibco), and 50 μM β-mercaptoethanol, as previously described (Montagner et al, 2016; Leoni et al, 2017). After 4 wk, ∼99% of the cells were FcεRIα⁺ c-Kit⁺ and were used for downstream analyses. For all experiments, BMMCs between week 4 and week 8 of age were used. Peritoneal cavity mast cells were isolated ex vivo from mouse peritoneal cavity by intraperitoneal lavage followed by magnetic bead–based enrichment for c-Kit⁺ cells using anti-CD117 (c-Kit)-APC antibody (BioLegend) and APC microbeads (Miltenyi Biotec). Cells were expanded for at least 1 wk in the presence of IL-3 and 30 ng/ml recombinant SCF (Peprotech) to obtain peritoneal cavity–derived mast cells. To generate mast cells lacking Regnase-1, BMMCs from Zc3h12a^fl/fl mice (Li et al, 2017) was transduced with a lentiviral vector expressing the Cre recombinase fused to an EGFP reporter (Leoni et al, 2023). Downstream analyses were performed by gating on EGFP⁺ cells (for flow cytometry–based assays) or by sorting for EGFP⁺ cells using BD FACSymphony S6 Cell Sorter (BD Biosciences) (for Western blot experiments). HEK293T cells were maintained in DMEM (Gibco) supplemented with 10% FBS (Gibco), 1 mM sodium pyruvate (Gibco), 100 U/ml penicillin, 100 mg/ml streptomycin (Gibco), and 50 μM β-mercaptoethanol. The HMC-1.1 and HMC-1.2 human mast cell lines were kindly provided by Joseph Butterfield and were cultured as described previously (Sundström et al, 2003).

Mast cell activation

Murine mast cell cultures were stimulated with the following for the indicated time periods: 1 μg/ml IgE/anti-DNP antibody (Sigma-Aldrich) and 0.2 μg/ml HSA-DNP (Sigma-Aldrich), 20 nM PMA (Sigma-Aldrich) and 2 μM ionomycin (Sigma-Aldrich), 10 ng/ml recombinant IL-33 (BioLegend), 1 μg/ml LPS (InvivoGen), and 10 ng/ml SCF (Peprotech). Human HMC-1.1 and HMC-1.2 cells were stimulated with 20 nM PMA (Sigma-Aldrich) and 2 μM ionomycin (Sigma-Aldrich) for 3 h.

Reverse transcription–quantitative PCR (RT–qPCR)

Mast cells were lysed in TRI Reagent RT (Molecular Research Center), and total RNA was extracted using Direct-zol RNA Microprep Kit (Zymo Research) according to the manufacturer’s protocol. cDNA was synthesized by reverse transcription using qScript cDNA SuperMix (Quanta Bioscience). Target genes were amplified using PerfeCTa SYBR Green FastMix (Quanta Bioscience) and QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific). All primer sequences used are listed in Table S5. Data analysis was performed using the 2^−ΔΔCt or 2^−ΔCt method.

Western blots

Mast cells were lysed in RIPA buffer (10 mM Tris–HCl, pH 8.0, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl) supplemented with protease inhibitor cocktail (Sigma-Aldrich) for protein detection. Protein concentration was quantified using Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Samples (50–100 μg) were run on SDS–polyacrylamide gels and blotted on a PVDF membrane using a wet transfer system. Membranes were blocked in 5% milk in TBST (5 mM Tris, pH 7.3, 150 mM NaCl, 0.1% Tween-20) for 30 min at RT, then incubated with primary antibodies overnight at 4°C, followed by the corresponding HRP-conjugated secondary antibody for 1 h at RT. All antibodies used are listed in Table S5. Chemiluminescence detection was performed using Clarity Western ECL Substrate (Bio-Rad) and Fusion FX7 EDGE Imaging System (WITec). Protein band intensity was quantified using ImageJ software version 1.53 h (Schindelin et al, 2012).

Surface marker and intracellular protein stainings

For surface marker staining, mast cells were stained using CD117 (c-Kit)-APC or APC/Cy7 and FcεRIα-PE (all from BioLegend) at 1:200 dilution for 20 min on ice. For all intracellular staining experiments, cells were washed with ice-cold PBS and stained with LIVE/DEAD Fixable Aqua or Blue Dead Cell Stain (Thermo Fisher Scientific) for 20 min at RT before cell fixation. For intracellular cytokine staining, cells were stimulated with different stimuli for 4 h. Brefeldin A (Sigma-Aldrich) was added in the last 2 h of stimulation. Cells were then fixed in 4% PFA for 10 min at RT, followed by permeabilization with 1% BSA/0.5% saponin solution in PBS, and staining with the following antibodies for 20 min on ice: TNF-α-PE/Cy7, IL-6-PE or APC (both from BioLegend) and IL-13-PE or eFluor 450 (eBioscience) for BMMCs, and human TNF-α-PE (BioLegend) for HMC-1.1 and HMC-1.2 cells. For intracellular staining of HA-tagged proteins, cells were fixed and permeabilized using eBioscience Foxp3/Transcription Factor Staining Buffer Set (Thermo Fisher Scientific) according to the manufacturer’s instructions. Cells were then stained with an anti-HA.11 epitope tag (BioLegend) at 1:200 dilution for 1 h at RT, followed by anti-mouse IgG (H+L)-Alexa Fluor 647–conjugated secondary antibody at 1:500 dilution for 30 min at RT. All antibodies used are listed in Table S5. Flow cytometry data were collected using FACSymphony A5 or LSRFortessa (BD Biosciences) and analyzed by FlowJo v10.6.0 (BD Biosciences).

siRNA transfection

BMMCs were transfected with siRNA cocktails targeting Zc3h12a and/or Zc3h12c (Horizon Discovery/Dharmacon) using 100 μl Neon Transfection System Kit (Thermo Fisher Scientific) according to the manufacturer’s protocol. siGLO Green Transfection Indicator (Horizon Discovery/Dharmacon) was used as a control. Briefly, cells were washed with PBS and resuspended in 100 μl of buffer R. siRNAs (200 pmol) were then added to the cell suspension, and cells were electroporated with one pulse at 1,600 V and 30 ms of width. Transfected cells were kept in antibiotic-free medium for 24 h. Downstream experiments were performed after 48–72 h. All siRNA sequences used are listed in Table S5.

NanoString profiling

BMMCs transfected with siRNAs against Zc3h12a and/or Zc3h12c or siGLO control were stimulated with 1 μg/ml IgE/anti-DNP antibody and 0.2 μg/ml HSA-DNP antigen for 2 h, then lysed in TRI Reagent RT. Total RNA was extracted using Direct-zol RNA Microprep Kit and quantified using Qubit RNA High Sensitivity Assay Kit and Qubit 3 Fluorometer (Thermo Fisher Scientific). Purified RNA (50 ng) was hybridized to the nCounter Myeloid Innate Immunity Panel v2 codeset (NanoString Technologies) for 16 h at 65°C. After hybridization, 30–35 μl of sample was added to the nCounter cartridge and analyzed using nCounter SPRINT Profiler (NanoString Technologies) according to the manufacturer’s instructions. Data analysis was carried out using nSolver Analysis Software v4.0 (NanoString Technologies) and the Omics Playground web-based platform (Akhmedov et al, 2020). Genes with log₂ fold change >0.5 and <−0.5 and P ≤ 0.05 were considered as differentially expressed.

CRISPR/Cas9 gene editing

BMMCs were transfected with Cas9-gRNA RNPs targeting Zc3h12a or Zc3h12c using 10 μl Neon Transfection System Kit (Thermo Fisher Scientific) according to the manufacturer’s protocol, and as previously described (Leoni et al, 2023). To generate gRNAs, equal amounts (400 pmol) of crRNA and tracrRNA were mixed with nuclease-free duplex buffer and annealed by boiling at 95°C, then cooling down to RT. Three different crRNAs were selected for Zc3h12a and Zc3h12c. A non-targeting crRNA (scrambled) was used as a control (all from Integrated DNA Technologies). Cas9-gRNA RNPs were prepared by incubating either 0.5 μl of each gRNA with 1.5 μl TrueCut Cas9 Protein v2 (5 μg/μl; Thermo Fisher Scientific) or 1 μl of each gRNA with 1.5 μl recombinant Cas9-NLS (5 μg/μl, in-house) for 20 min at RT. To increase transfection efficiency, Alt-R Electroporation Enhancer (Integrated DNA Technologies) was added to the transfection mix. Cells were resuspended in 10 μl of buffer R, and electroporated with the RNPs with one pulse at 1,600 V and 30 ms of width for BMMCs, and one pulse at 1,700 V and 20 ms of width for HMC-1.1 and HMC-1.2 cells. Transfected cells were kept in antibiotic-free medium for 24 h. Downstream experiments were performed after 1 wk. All oligonucleotides used are listed in Table S5.

Expression plasmids

Regnase-1 and Regnase-3 expression plasmids were generated using standard cloning techniques. The murine Zc3h12a coding sequence was amplified from the pRetro-Xtight-Myc-GFP-Regnase-1 plasmid (Behrens et al, 2021), whereas the Zc3h12c coding sequence was amplified from cDNA obtained from BMMCs. These were either cloned into the pCDNA3 vector (Thermo Fisher Scientific) and pCDH-EF1α-T2A-copGFP (System Biosciences) for expression in HEK293T or tagged with FLAG-HA, then cloned into the pUC57-mini vector (synthesized by GenScript) to be used as a template for in vitro mRNA transcription. The human ZC3H12A coding sequence was amplified from the pEXPR-IBA105-MCPIP1 plasmid (Behrens et al, 2018), then cloned into the pUC57-mini vector. To generate the RNase-inactive mutants, residues D141 of Regnase-1 (either human or mouse) and D252 of Regnase-3 were mutated to asparagine using QuikChange II XL Site-Directed Mutagenesis Kit (Agilent) following the manufacturer’s protocol. To generate C-terminal truncations of Regnase-3, a stop codon was incorporated at the indicated sites using QuikChange II XL Site-Directed Mutagenesis Kit (Agilent) following the manufacturer’s protocol. To generate the luciferase reporter plasmid, the full-length 3′UTR of Zc3h12a was amplified from genomic DNA obtained from BMMCs and cloned downstream of the luciferase reporter gene in pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega).

In vitro mRNA transcription

To express Regnase-1 and -3, pUC57-mini plasmids harboring the WT or mutated coding sequence downstream of the T7 promoter were first transcribed into mRNA using HiScribe T7 ARCA mRNA Kit (New England BioLabs) according to the manufacturer’s protocol. Briefly, the plasmids were linearized by digestion with SpeI (New England BioLabs), followed by purification using NucleoSpin Gel and PCR Clean-up Kit (Macherey-Nagel). Linearized DNA templates (1 μg) were then assembled together with ARCA/NTP mix, T7 RNA polymerase mix (both included in the kit), and pseudo-UTP (Jena Bioscience) for the IVT reaction. mRNA synthesis was completed by removing the DNA template through DNase I treatment, followed by poly(A) tailing by adding poly(A) polymerase and buffer (all included in the kit) to the reaction mix. All reactions were each done at 37°C for 30 min. Resulting mRNA products were purified using Monarch RNA Cleanup Kit (New England BioLabs), visualized on 1% TBE gel, and quantified using NanoDrop 2000. BMMCs were transfected with Regnase-1 or Regnase-3 (WT or RNase-inactive mutant D141N or D252N) IVT mRNA using 10 μl Neon Transfection System Kit (Thermo Fisher Scientific) according to the manufacturer’s protocol. IVT mRNA from the pUC57-mini plasmid expressing ZsGreen alone was used as a control. Briefly, cells were washed with PBS and resuspended in 10 μl of buffer R. IVT mRNA (3 pmol for BMMCs and 0.25 pmol for HMC-1.1 and HMC-1.2 cells) and 10U RNase inhibitor (Promega) were then added to the cell suspension. Cell electroporation was performed with one pulse at 1,600 V and 30 ms of width for BMMCs, and one pulse at 1,700 V and 20 ms of width for HMC-1.1 and HMC-1.2 cells. Transfected cells were kept in antibiotic-free medium, and downstream experiments were performed within 24 h.

Luciferase reporter assays

HEK293T cells were transfected with 3 μg of a pmirGLO plasmid containing the 3′UTR of Zc3h12a or Tnf (plasmid 207127; Addgene) (Leoni et al, 2023) and 1 μg of pCDNA3-Regnase-1 WT/D141N or pCDH-Regnase-3 WT/D252N or the corresponding empty control plasmids using polyethylenimine (PEI). After 48 h, luciferase activity was analyzed using Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s protocol and measured using GloMax Discover (Promega).

mRNA stability assay

BMMCs were treated with 1 μg/ml IgE/anti-DNP (Sigma-Aldrich) for 30 min followed by stimulation with 0.2 μg/ml HSA-DNP (Sigma-Aldrich) for 30 min. Transcription was then blocked by incubating the cells with 10 μg/ml actinomycin D (Sigma-Aldrich). Cells were harvested at different time points (15, 30, and 60 min), and RNA isolation and RT–qPCR were performed as mentioned above. mRNA expression was calculated using the 2^−ΔCt and 2^−ΔΔCt method, normalizing against Gapdh expression. When possible, the mRNA decay rate (half-life, t^1/2) was calculated by non-linear regression curve fitting (one phase decay) using GraphPad Prism version 9 (GraphPad).

Cell viability and apoptosis assays

Cell viability was measured by staining cells with LIVE/DEAD Fixable Aqua Dead Cell Stain or Blue Dead Cell Stain (Thermo Fisher Scientific) for 20 min at RT. Cell apoptosis was measured by co-staining cells with Annexin V and 7-AAD using PE Annexin V Apoptosis Detection Kit I (BD Biosciences), according to the manufacturer’s protocol. All experiments were performed without acute stimulation.

Cell proliferation and cell cycle analysis

Cell proliferation was measured by BrdU incorporation for 12–16 h at 37°C, followed by BrdU staining using APC BrdU Flow Kit (BD Biosciences), according to the manufacturer’s protocol. Cell cycle progression was analyzed by co-staining cells with 7-AAD. All experiments were performed without acute stimulation.

Cell degranulation assay

BMMCs were treated with 1 μg/ml IgE/anti-DNP (Sigma-Aldrich) and 0.2 μg/ml HSA-DNP (Sigma-Aldrich) for 30 min. Cell degranulation was measured by Annexin V staining using PE Annexin V Apoptosis Detection Kit I (BD Biosciences), according to the manufacturer’s protocol.

RNA sequencing

BMMCs were transfected with Zc3h12a RNPs or scrambled RNPs as a control (four independent biological replicates). After 1 wk in culture, cells were lysed in TRI Reagent RT (Molecular Research Center) and RNA was extracted using Direct-zol RNA Microprep Kit (Zymo Research). After poly(A) mRNA enrichment and Tecan Revelo mRNA Library Preparation, mRNA sequencing using Illumina NovaSeq 6000 (2 × 50-bp reads) was outsourced to Next Generation Sequencing Platform at the University of Bern (Switzerland). Read quality control was assessed using fastqc v.0.11.9 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) and RSeQC v.4.0.0 (Wang et al, 2012), followed by read mapping to the reference genome Mus_musculus.GRCm39.107 using HiSat2 v.2.2.1 (Kim et al, 2015). Counts were generated and corrected for batch effects using featureCounts v.2.0.1 (Liao et al, 2014) and removeBatchEffect function in R (Ritchie et al, 2015), respectively. Differential expression was analyzed by combining three DE methods: DESeq2 (Wald), edgeR (QLF), and limma (trend), using the Omics Playground platform (Akhmedov et al, 2020). Genes with log₂ fold change >0.5 and <−0.5 and P ≤ 0.05 were considered differentially expressed and included in the gene ontology (GO) analysis performed using DAVID Bioinformatics Resource v2023q2 (Huang da et al, 2009; Sherman et al, 2022). The top 10 GO enriched terms in up-regulated and down-regulated genes were graphically represented (for up-regulated genes, only categories with >10 genes were considered). Data visualization was performed with RStudio version 4.1.

Meta-analysis of RBP expression

Re-analysis of RNA-seq data from Li et al (2021) was performed using the DESeq2 package in R (Love et al, 2014). Genes encoding for RBPs identified in more than one RBPome listed in the RBP2GO database (Caudron-Herger et al, 2021) and with the sum of counts in all samples ≥10 were included in the differential expression analysis. RBP genes with log₂ fold change >0.5 and <−0.5 and P ≤ 0.05 were considered differentially expressed and included in functional categorization analysis using PANTHER 17.0 (Mi et al, 2019). First, differentially expressed RBP genes broadly categorized under the GO term RNA metabolic process (GO:0016070) were identified. This was followed by more specific categorization using the following GO terms: regulation of RNA stability (GO:0043487), RNA modification (GO:0009451), RNA processing (GO:0006396), RNA localization (GO:0006403), and translation (GO:0006412). Data visualization was performed with RStudio version 4.1. ATAC-seq data from Li et al (2021) were visualized using Integrative Genomics Viewer (IGV) 2.16.0 (Robinson et al, 2011).

Lentiviral preparation and cell transduction

To generate lentivirus, HEK293T cells were transfected with lentiviral vectors together with the packaging vectors psPAX2 and pMD2.G (plasmids 12260 and 12259 by Didier Trono; Addgene) using PEI. After 48 h, lentiviral particles in the supernatant were collected and concentrated using a PEG-8000 solution as previously described (Lo & Yee, 2007). For co-expression experiments, BMMCs were first transduced with pScalps-Regnase-1 D141N lentivirus and treated with puromycin for 48 h, followed by a second transduction with pScalps-Regnase-3 D252N lentivirus. For Cre recombinase expression, the pScalps-EGFP-Cre recombinase vector (Leoni et al, 2023) (plasmid 207132; Addgene) was used to transduce WT and Zc3h12a^fl/fl BMMCs, for at least 96 h. Flow cytometry–based analyses were done by gating on EGFP⁺ cells, whereas Western blot experiments were performed on EGFP⁺ cell population sorted using BD FACSymphony S6 Cell Sorter (BD Biosciences).

Immunoprecipitation and enzymatic digestion

BMMCs (2 × 10⁷) were activated with 2 nM PMA and 200 nM ionomycin for 1 h. Cells were washed with ice-cold PBS and lysed in IP buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.25% NP-40, 1X protease inhibitor [Sigma-Aldrich], 1X phosphatase inhibitor [Sigma-Aldrich]). Lysates were cleared of cell debris by centrifugation at 10,000g for 10 min at 4°C, followed by pre-clearing with 20 μl Dynabeads Protein G (30 mg/ml; Invitrogen) at 4°C with rotation for 2 h. 1 mg of pre-cleared lysates was incubated with 50 μl Dynabeads Protein G (30 mg/ml) and 5 μg Regnase-1 (R&D Systems) or Regnase-3 (Helmholtz Munich Monoclonal Antibody Core Facility) antibody at 4°C with rotation for 16 h. Mouse IgG (Sigma-Aldrich) or rat IgG (Thermo Fisher Scientific) antibodies were used as isotype controls, respectively. Immunoprecipitated proteins were pulled down using DynaMag-2 Magnet (Thermo Fisher Scientific) and washed two times with IP buffer with 0.1% NP-40, followed by two washes with detergent-free wash buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1X protease inhibitor, 1X phosphatase inhibitor).

On-bead digestion of immunoprecipitated proteins was performed as follows. Beads were resuspended in 8 M urea, 50 mM ammonium bicarbonate buffer. Proteins were then reduced with 10 mM dithiothreitol for 60 min at 37°C and alkylated with 50 mM iodoacetamide for 30 min at RT. Digestion was carried out in 8 M urea, 50 mM ammonium bicarbonate buffer for 2 h at 37°C with 1 μg of Lys-C (FUJIFILM Wako Chemicals), after which the digestion buffer was diluted to final 2 M urea with 50 mM ammonium bicarbonate. 1 μg of trypsin (Promega) was added for overnight digestion at 37°C. All steps were performed in agitation to avoid bead precipitation. Digestion was stopped by adding acetonitrile to 2% and trifluoroacetic acid to 0.3%, and the beads were collected with DynaMag-2 Magnet (Thermo Fisher Scientific). Digested peptides were purified by loading the supernatant into C18 StageTips (Rappsilber et al, 2007), and eluted with 80% acetonitrile, 0.5% acetic acid. Finally, the elution buffer was evaporated by vacuum centrifugation and purified peptides were resolved in 2% acetonitrile, 0.5% acetic acid, 0.1% trifluoroacetic acid for single-shot LC-MS/MS measurements.

LC-MS/MS

Peptides were separated on an EASY-nLC 1200 HPLC system (Thermo Fisher Scientific) coupled online via a nanoelectrospray source (Thermo Fisher Scientific) to a Q Exactive HF mass spectrometer (Thermo Fisher Scientific). Peptides were loaded in buffer A (0.1% formic acid) into a 75 μm inner diameter, 50 cm length column, packed in-house with ReproSil-Pur 120 C18-AQ 1.9 μm resin (Dr. Maisch HPLC GmbH), and eluted over a 150-min linear gradient of 5–30% buffer B (80% acetonitrile, 0.1% formic acid) at a flow rate of 250 nl/min. Xcalibur software (Thermo Fisher Scientific) operated the Q Exactive HF in a data-dependent mode with a survey scan range of 300–1,650 m/z, resolution of 60,000 at 200 m/z, maximum injection time of 20 ms, and AGC target of 3 × 10⁶. The top 10 most abundant ions with charge 2–5 were isolated with a 1.8 m/z isolation window and fragmented by higher energy collisional dissociation (HCD) at a normalized collision energy of 27. MS/MS spectra were acquired with a resolution of 15,000 at 200 m/z, maximum injection time of 55 ms, and AGC target of 1 × 10⁵. Dynamic exclusion was set to 30 s to avoid repeated sequencing.

LC-MS/MS data analysis

MS raw files were processed using MaxQuant software v.1.6.7.0 (Cox & Mann, 2008). Peptides and proteins were identified with a 0.01 false discovery rate by employing the integrated Andromeda search engine (Cox et al, 2011) to search spectra against the mouse UniProt database (July 2019) and a common contaminants database (247 entries). Enzyme specificity was set as “Trypsin/P” with a maximum of two missed cleavages and minimum length of seven amino acids. N-terminal protein acetylation and methionine oxidation were set as variable modifications, and cysteine carbamidomethylation as a fixed modification. Match between runs was enabled to transfer identifications across samples based on mass and normalized retention times, with a matching time window of 0.7 min and an alignment time window of 20 min. Label-free protein quantification (LFQ) was performed with the MaxLFQ algorithm (Cox et al, 2014) with a minimum required peptide ratio count of 1. Data analysis was performed using Perseus software v.1.6.2.3 (Tyanova et al, 2016). Data were pre-processed by removing proteins only identified by site, reverse hits, and potential contaminants. After log₂ transformation of LFQ intensities, biological replicates of each experimental condition were grouped and proteins were filtered for a minimum of three valid values in at least one group. Missing data points were then replaced by imputation from a normal distribution with 0.3 width and 1.8 downshift, and a two-sided two-samples t test (0.05 false discovery rate, 250 randomizations) was used to identify significant changes in protein intensity between each immunoprecipitation experiment and its corresponding isotype control.

Regnase-1 and -3 co-transfection in HEK293T cells

HEK293T cells were co-transfected with pCDNA3-Regnase-1 and pCDNA3-Regnase-3 plasmids (10 μg each, full-length or C-term truncated) using polyethylenimine (PEI). After 48 h, cells were harvested for immunoprecipitation experiments using 5 μg anti-Regnase-1 (R&D Systems), anti-Regnase-3 (Helmholtz Munich Monoclonal Antibody Core Facility), anti-HA (BioLegend), or anti-FLAG (Sigma-Aldrich) antibodies. For experiments with RNase treatment, pre-cleared lysates were treated with 100 μg/ml RNase A (Zymo Research) and RNase I (Thermo Fisher Scientific) at 37°C for 30 min before incubation with the antibody–bead complex.

Immunofluorescence

Cells were fixed with 4% PFA in PBS for 10 min, then cytospun on a coverslip. Cells were permeabilized with 0.2% Triton X-100 in PBS for 15 min and blocked with 5% BSA in PBS for 30 min. Cells were then stained with anti-Regnase-1 (R&D Systems) or anti-Regnase-3 (Helmholtz Munich Monoclonal Antibody Core Facility) at 1:100 dilution for 1.5 h, followed by anti-mouse IgG (H+L)-Alexa Fluor 568– or anti-rat IgG (H+L)-Alexa Fluor 647–conjugated secondary antibody (Thermo Fisher Scientific) at 1:500 dilution for 30 min. All incubations were done at RT. Stained cells were washed three times with PBS and once with water, followed by nucleus counterstaining and mounting on microscopy slides using Vectashield (Vector Laboratories) with DAPI. All antibodies used are listed in Table S5. Confocal microscopy imaging was performed using Leica Stellaris SP8 confocal laser scanning microscope with Leica HCX PL APO 40 × 1.30 oil or 63 × 1.40 oil objectives (Leica Microsystems). Image acquisition and processing were done using Leica Application Suite X (Leica Microsystems) and ImageJ version 1.53 h (Schindelin et al, 2012) software.

ELISA

1 × 10⁵ cells were stimulated with 1 μg/ml IgE/anti-DNP and 0.2 μg/ml HSA-DNP for 6 h in a 96-well plate before collection of the supernatant and measurement of TNF release using Mouse TNF-α High Sensitivity ELISA (eBioscience), following the manufacturer’s instructions.

Data analysis

All data shown consist of biological (not technical) replicates. Data representation and statistical analysis were performed using GraphPad Prism v9. All source data for all figures are provided in the Source Data Files.