Pde1b controls cyclic nucleotide signaling in cDC1 and macrophages to regulate immune responses

Pde1b is selectively expressed in cDC1 dendritic cells, macrophages and neutrophils, but not in cDC2, B cells or T cells

To determine the effects of Pde1b loss, global knockout mice (Pde1b−/−) were utilized. No differences in health or viability were observed among WT, Pde1b+/− and Pde1b−/− (KO) genotypes, consistent with prior reports (Reed et al, 2002). Deletion of exon 3 in the Pde1b gene generated a ~150 bp smaller band in Pde1b−/− mice compared to WT upon PCR genotyping (Fig. EV1A). This Pde1b−/− strain produces unstable transcripts as measured by qPCR (Fig. EV1B). However, no protein can be detected by Western blot (Fig. EV1C,D).

Single-cell RNA sequencing (scRNA-seq) of splenic cells from Pde1b−/− and WT mice analyzed an average of ~25,000 high-quality cells per genotype across four biological replicates (n = 4 each), with rigorous batch correction, cell-type annotation and confirmation of major findings in independently sorted and flow cytometry-validated cohorts. Canonical correlation analysis and clustering identified 14 major immune cell populations (Fig. 1A). Comparing Pde1b–/− to WT controls revealed differences in the frequency and distribution of mature B cells, macrophages, memory T cells and transitional B cells (Fig. 1B). While Pde1b−/− mice showed minor detection of residual 5′ transcript fragments due to disruption of the gene by the exon 3 insertion cassette (Fig. 1C), overall detected transcript counts were reduced compared to WT. Pde1b transcript expression levels in WT confirmed a strong presence in macrophages (cluster 8) and DC (cluster 9), and to a lesser extent in neutrophils (cluster 12) and granulocytes (cluster 13), with expression largely abolished in the KO (Fig. 1D).

To corroborate gene expression profiling with protein assessment, a Western blot was performed to confirm levels in the spleen of C57BL/6 wild-type (WT) mice, with only marginal levels in the thymus (Fig. 1E); brain tissue was included as a reference sample based on prior reports of Pde1b expression to confirm expected tissue-specific expression (Reed et al, 2002). For deeper probing, FACS-sorted analysis of splenic immune populations revealed Pde1b protein expression to be in cDC1, macrophages and neutrophils, with no expression measured in cDC2 subsets, B cells or T cells, as evaluated by qPCR (Fig. 1F). Flow cytometric analysis of spleen confirmed Pde1b expression in cDC1, defined as CD24+CD172a^negIRF8+ within the CD11+ MHC class II+ subsets (Fig. 1G) and in the F4/80+ CD64+ macrophages (Fig. 1H).

Pde1b is essential for effective clearance of Listeria bacterial infection in vivo

Given select expression of Pde1b in myeloid APC populations, in vivo challenge experiments were performed to assess the capacity to mount functional immune responses against bacterial challenge. To assess bacterial burden, mice were infected with live sublethal levels of Listeria monocytogenes (LM). Bacterial burden in the spleen remained elevated in Pde1b−/− mice compared to WT controls at days 4 and 6 post-infection (Fig. 2A). To assess in vivo antigen-specific T cell responses, studies employed the attenuated strain of LM expressing the 2W antigen (ΔactA-LM-2W), which has reduced virulence but is still capable of eliciting a strong adaptive immune response that can be well quantified. Using 2W:I-Ab MHC class II tetramers, a markedly lower frequency of LM-antigen-specific CD44^hiCD4 + T cells were measured on both days 4 and 6 post-infection in Pde1b−/− mice compared to WT controls (Fig. 2B,C). As an internal control, CD3+CD4^neg T cells (representing the CD8+ T cell compartment) showed no binding to the tetramer, consistent with MHC class I restriction. Assessment of total spleen by qPCR at day 6 post-infection revealed increased expression of cytokines Il10, Il17a and Tgfb1 and the chemokine Ccr9 (Fig. 2D).

Loss of Pde1b impairs macrophage inflammatory function and cDC1 antigen presentation

Given the reduced priming of LM-specific CD4 + T cells observed in Pde1b−/− mice following in vivo pathogen challenge, studies then investigated the specific role of macrophages and cDC1 in inflammation and adaptive immunity. Pde1b−/− bone marrow-derived macrophages showed impaired LM clearance upon infection (Fig. 3A), concomitant with elevated levels of Arg1 and Il10 and depressed expression of Il12a (Fig. 3B,C). Secreted cytokine profiles confirmed increased IL-10 and reduced IL-12p70 (Fig. 3D).

Flt3L-generated bone marrow DC (Flt3L-cDC) from Pde1b−/− mice infected with LM encoding ovalbumin (LM-OVA) had reduced capacity to stimulate proliferation of OT-I CD8+ T cells (OVA-specific T cell receptor-transgenic CD8+ T cells recognizing OVA peptide presented on MHC class I) (Fig. 3E), consistent with impaired cross-presentation capacity required for priming responses to intracellular pathogens. Given the selective expression of Pde1b in cDC1, studies next assessed whether this defect reflects a cDC1-intrinsic impairment. FACS-sorted cDC1 from the spleens of Pde1b−/− mice infected with LM-OVA at varying concentrations had reduced capacity to drive CD8+ T cell responses, whereas cDC2 had poor priming capacity regardless of genotype (Fig. 3F). Supernatants from LM-infected Pde1b−/− cDC1 and OT-I CD8+ T cell cocultures contained lower levels of TNFα, IFNγ and IL-12p70 compared to WT controls (Fig. 3G). These results demonstrate that Pde1b expression is important for effective inflammatory responses and antigen-specific T cell priming in myeloid APC.

Rescue by adoptive transfer of WT cDC, but not Pde1b-deficient cDC, is able to partially restore priming of LM-specific CD8+ T cell responses in Pde1b−/− recipient mice in vivo

To corroborate whether impaired CD8+ T cell priming in Pde1b−/− mice reflects a defect intrinsic to dendritic cells in vivo, splenic-derived WT or Pde1b-deficient cDC were adoptively transferred into Pde1b−/− mice prior to infection with LM-OVA. Transfer of WT cDC partially restored LM-specific CD8+ T cell expansion, whereas transfer of Pde1b-deficient cDC failed to rescue this defect and remained comparable to Pde1b−/− mice without transfer (Fig. 4A,B). As an internal control, MHC class II-restricted CD4+ T cells showed no tetramer staining, confirming antigen- and MHC-specificity. Successful in vivo adoptive cell transfer was validated by measuring the frequency of CD11c+ MHC class II+ cells in the spleen (Fig. 4C).

Pde1b deficiency induces modest transcriptional remodeling in cDC1 without altering subset development

To determine how deficiency in Pde1b alters cDC functional capacities, transcriptional profiling and immunophenotyping were performed using scRNA-seq and flow cytometric analyses. Within the DC cluster, a limited number of major transcriptional changes were observed, reflecting modest global alterations (Fig. 5A). These included increased expression of Batf1, Ide, Cd8b1, and Mrc1, alongside reduced expression of Pde1b, Smad5, Adgrg6, Il4i1, and Tank in Pde1b-deficient mice. Deeper analysis of the DC compartment enabled identification of distinct subsets through integration of UMAP-based cluster segregation and canonical lineage marker expression, identifying classical DCs (cDCs, spanning both cDC1 and cDC2) via Batf3 and Zbtb46 expression to distinguish them from plasmacytoid DCs (pDCs), with cDC1 further defined by Xcr1, Irf8, Aif1, Cd8a, and Cd24; cDC2 defined by Itgam (CD11b) and Sirpa (CD172a); and pDCs defined by Siglech and Irf8 (Figs. 5B,C and EV2). Pde1b expression was largely restricted to the cDC1 cluster (Fig. 5D). Notably, both Pde4b and Pde7a phosphodiesterase genes were also present within cDC1 (Fig. EV3); no other PDE family members or isoforms were detected.

Differential gene expression analysis was performed in cDC1 cells comparing WT and Pde1b⁻/⁻ mice. At steady state, loss of Pde1b resulted in modest but coordinated transcriptional changes. Functional annotation of differentially expressed genes identified alterations in transcripts associated with inflammatory, cytokine-responsive, interferon-related, migratory and metabolic programs (Fig. 5E). Among these changes, expression of NFκB-associated genes, including Nfkbiz, Relb, Traf1, and Rel was reduced, while genes linked to purinergic and cyclic nucleotide signaling, including P2ry14, Dgka, and Grk3, were increased. Additional decreases were observed in cytokine-responsive and interferon-associated genes, including Il15ra, Tyk2, Jak2, Il21r, Irf7, Nlrc5, and Isg15. Examination of the broader set of top differentially expressed genes revealed that these changes were not restricted to a select number, but instead reflect a distributed transcriptional signature characterized by numerous modest shifts in cDC1 gene expression profile in absence of Pde1b at steady state (Fig. 5F). Importantly, flow cytometric analysis showed no significant differences in the frequencies of cDC1 or cDC2 populations (Fig. 5G), indicating that Pde1b is dispensable for DC subset development and homeostasis at steady state.

Pde1b marks splenic macrophage heterogeneity and is associated with distinct transcriptional programs

Comparative transcriptional profiling of splenic macrophages revealed increased expression of Fn1, Ide, Alcam, Lrk2, Cxcl12, and Fgfr2 in Pde1b-deficient macrophages, concomitant with decreased expression of Apoc1, Apol7c, Marcksl1, and Il11ra1 (Fig. 6A). Deeper analyses identified two subclusters in macrophages, with Pde1b expression restricted to subcluster 1 (Pde1b^pos) (Fig. 6B,C). Several genes were differentially expressed between Pde1b^pos and Pde1b^neg in WT spleen, indicating transcriptional heterogeneity (Fig. EV4A,B). Differential expression analysis within Pde1b^pos macrophages identified genes altered between WT and Pde1b−/− cells, which were subsequently examined across macrophage subclusters 0 and 1. Analysis of biological replicates confirmed reproducible identification of the Pde1b^pos macrophage subset, with Pde1b expression consistently detected in the Pde1b^pos cluster but absent from the Pde1b^neg population (Fig. EV4C). Notably, the Pde1b^pos subcluster remained identifiable in Pde1b−/− mice, indicating that Pde1b expression itself was not required for the establishment of this macrophage subpopulation. Examination of WT and Pde1b−/− macrophages revealed increased expression of Cebpb, Folr2, Stat4, Arntl, Fgfr2, and Nars2 within the Pde1b^pos subset of knockout mice (Fig. 6D), consistent with a potential shift in inflammatory programming. Curated differential expression analysis of the Pde1b^pos macrophage subcluster revealed modest transcriptional changes, including increased expression of antigen presentation and regulatory genes (Nlrc5, H2-DMb2, Cebpb, and Fosl2) and selective reduction of the environmental sensing receptor Fpr2, a G-protein-coupled receptor involved in chemotactic responses, tissue-derived danger sensing and pro-resolving inflammatory signaling (Fig. 6E). Subset-specific analyses by flow cytometry showed reduced frequency of F4/80+CD11b^lo splenic macrophages (Fig. 6F). Within the F4/80+CD11b^loCD64+CD11c^neg compartment, Pde1b−/− macrophages exhibited reduced MHC class II expression (Fig. 6G). Collectively, Pde1b deficiency reduces the frequency of a defined splenic macrophage subset at steady state and is associated with modest but coordinated remodeling of immune sensing and antigen presentation programs within this compartment.

Pde1b loss elevates cyclic nucleotide and kinase activities

Studies next evaluated the downstream changes in cAMP-PKA and cGMP-PKG in cDC and macrophages in the absence of Pde1b. Under steady-state, splenic-derived FACS-purified cDC1 and macrophages from Pde1b-deficient mice exhibited significantly elevated intracellular levels of both cAMP and cGMP compared to WT controls (Fig. 7A,B). By contrast, cDC2 subsets showed no change in cyclic nucleotide concentration, demonstrating a selective role for Pde1b in specific myeloid APC lineages. Correspondingly, loss of Pde1b resulted in increased activities of both PKA and PKG kinase in cDC1 and macrophages, but not in cDC2 populations, as quantified by kinase phosphorylation assays (Fig. 7C,D). Comparison of WT and Pde1b−/− cDC1 and macrophages revealed no significant changes in expression of adenylate cyclase, soluble guanylate cyclase or PKA and PKG subunits, indicating that elevated signaling is not accompanied by compensatory transcriptional remodeling of upstream or downstream pathway components (Fig. 7E). Taken together, these data reveal that Pde1b limits cyclic nucleotide signaling and downstream kinase activation.

Acute inhibition of PDE1, but not PDE4 or PDE7, phenocopies the impaired immune priming observed in Pde1b-deficient cells in vitro

To determine whether this phenotype reflects specificity for PDE1 or redundancy with other phosphodiesterases, we first assessed expression of candidate PDE family members. Pde4b and Pde7a genes were expressed in both cDC1 and macrophages, with no evidence of compensatory upregulation in either population or distinct subsets in Pde1b−/− mice (Figs. 8A,B and EV5A,B). We next evaluated the functional contribution of these enzymes using pharmacological inhibitors in primary DC. These included a PDE1-selective inhibitor (DSR-141562; preferential for PDE1B, IC50 = 43.9 nM vs PDE1A = 97.6 nM and PDE1C = 431.8 nM), a pan-PDE1 inhibitor (ITI-214), a PDE4 inhibitor (roflumilast; APTA-2217) and a PDE7 inhibitor (PDE7-IN-2). Concentrations were selected based on published IC50 values where available. As these compounds differ in isoform selectivity, results were interpreted as pathway-level inhibition. PDE1 inhibition markedly reduced inflammatory cytokine production following TLR2 stimulation with heat-killed Listeria monocytogenes (HKLM), whereas PDE4 inhibition produced only partial effects, and PDE7 inhibition had no detectable impact (Fig. 8C). To relate these functional effects to cyclic nucleotide signaling, intracellular cAMP was measured under basal and HKLM-stimulated conditions across increasing inhibitor concentrations. PDE1 inhibition produced a larger increase in cAMP than PDE4 inhibition across all doses tested, although both inhibitors elevated cAMP in a dose-dependent manner (Fig. 8D,E).

Absence of Pde1b results in increased phosphorylation of CREB and GSK3 in DC and macrophages

To determine downstream consequences of heightened cyclic nucleotide-driven protein kinase activity, phosphorylation states of CREB and GSK3 were assessed in sorted DC and macrophage populations. Single-cell transcriptomics showed no significant changes in levels of total Gsk3a, Gsk3b, Stat3 or Creb1 expression in either DC or macrophages (Fig. 9A,B). Phospho-flow cytometric analysis of DC subsets revealed increased levels of phosphorylated CREB at Serine 133 (pCREB_S133) and GSK3β at serine 9 (pGSK3β_S9) in the cDC1 subsets (Fig. 9C). Parallel analysis of the F4/80+CD11b^lo macrophages also observed increased levels of pCREB_S133 and moderately elevated levels of pGSK3β_S9 (Fig. 9D). These post-translational modifications correlate with suppressed inflammatory gene signatures and are consistent with reprogramming toward immunoregulatory transcriptional states, providing a mechanistic basis for reduced inflammatory and antigen presentation capacity.

Animals

Mouse models used in the study included C57BL/6 (B6; WT), Pde1b−/−(B6.129P2-Pde1btm1Dgen/Mmnc), OT-I (C57BL/6-Tg(TcraTcrb)1100Mjb/J), and OT-II (B6.Cg-Tg(TcraTcrb)425Cbn/J). B6, OT-I and OT-II animals were purchased from The Jackson Laboratory. Pde1b knockout mice (Pde1b−/−, KO; strain name: B6.129P2-Pde1btm1Dgen/Mmnc) were procured from MMRC through the University of North Carolina, Chapel Hill. Knockout was achieved via insertion of a neomycin resistance cassette into exon 3, disrupting the coding region. Genotyping was performed by PCR using the following primers (5’ to 3’): Forward: AGTTTTCTGGAGGCCTTGGAGACAG; Reverse: CCCCATTGCTCACTTGGTCTGAATG; Neomycin: GGGTGGGATTAGATAAATGCCTGCTCT. Both male and female mice were used for in vivo studies and as a source of bone marrow, spleen and lymph nodes for in vitro and ex vivo experiments. All animal experiments were conducted in accordance with institutional guidelines and were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Minnesota under approved protocol no. IACUC_2502-42757. All procedures complied with national and institutional regulations governing the care and use of laboratory animals.

Antibodies and bioreagents

The following antibodies and respective clones purchased from BioLegend were used for staining and/or depletion: CD11c (N418), MHC class II (M5/114/15.2), CD24 (M1/69, 30-F1), CD172a (P84), IRF8 (V3GYWCH), F4/80 (BM8), IRF4 (3E4), CX3CR1 (SA011F11), CD11b (M70), CD64 (X54-5/7.1), Mertk (2b10c42), B220 (RA3-6B2), CD19 (4B12), IgD (11-26 c.2a), CD38 (90), CD44 (IM7), TCRβ (H57-597), CD3 (17A2), CD4 (RM4-5), CD8 (53-6.7), SiglecH (eBio440C), Ly6G (IA8), phospho-CREB1 (Ser133) (A22026B). The following antibodies were purchased from Abcam: Pde1b (EPR12293(2)). The following antibodies were purchased from Thermo Fisher Scientific (Waltham, MA, USA): phospho-GSK3β (1C9E2). 2W-I-Ab tetramer was received as a gift from Marc Jenkins and T. Dileepan. H2kb_SIINFEKLpMHC tetramer received from NIH Tetramer Core Facility. CFSE labeling dye (Thermo Fisher Scientific, Waltham, MA, USA) was utilized to stain T cells for proliferation studies.

Single-cell RNA sequencing

Single-cell RNA sequencing (scRNA-seq) was performed on spleen samples from 10-week-old Pde1b−/− and WT mice (n = 4 per group). Tissues were dissociated into single-cell suspensions using the GentleMACS Dissociator (Miltenyi, Germany), with red blood cells removed by ACK lysis. Live cells were processed for 5’ scRNA-seq, targeting ~25,000 cells per sample at a sequencing depth of 50,000 reads per cell using the Chromium Single Cell Controller (10x Genomics) at the University of Minnesota Genome Center (UMGC) Single Cell Platform. Libraries were prepared using Chromium GEM-X Single Cell 5’ chemistry (10x Genomics), pooled, and sequenced on an Illumina NovaSeq 6000 platform. Raw base call files were processed, demultiplexed, and aligned to the mouse reference genome (mm10-2020-A) using Cell Ranger (10x Genomics). Resulting gene-cell matrices were analyzed in Seurat v5 (R v4.4.1) (Hao et al, 2024). Cells were subjected to standard quality control filtering to remove low-quality cells and potential doublets based on transcript abundance, gene complexity and mitochondrial content. Data were normalized using SCTransform with glmGamPoi, and variable features were used for principal component analysis (PCA). To minimize confounding from adaptive immune receptor diversity, immunoglobulin and T-cell receptor genes were excluded from dimensionality reduction. Batch effects across biological replicates were corrected using Seurat integration based on canonical correlation analysis (CCA), and integrated embeddings were used for downstream clustering and visualization. Cell populations were identified using shared nearest-neighbor graph construction followed by Louvain clustering and visualized using Uniform Manifold Approximation and Projection (UMAP). Automated cell type annotation was performed using SingleR with the ImmGen reference dataset, and assignments were refined by manual curation using canonical marker gene expression. Differential gene expression analyses were performed using a pseudobulk framework in which counts were aggregated by biological replicate and condition, followed by DESeq2-based testing implemented in Seurat (Love et al, 2014). Cluster-specific marker genes were identified using FindAllMarkers (Wilcoxon rank-sum test). Visualization of gene expression was performed using SCTransform-normalized values. RNA-seq data have been deposited in GEO under accession number GSE297611 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE297611).

In vitro generation of bone marrow-derived macrophages and Flt3L-generated dendritic cells

Bone marrow cells derived from murine tibias and femurs were cultured with 10 ng/mL of recombinant mouse Flt3L in IMDM supplemented with Pen/Strep and 10% fetal bovine serum (FBS) to generate DC. After 8–10 days of in vitro culture, cells were harvested for in vitro studies or further labeled with fluorochrome-labeled antibodies to CD24, CD172a, CD11c, and MHC class II to purify by FACS prior to in vitro culture studies or magnetic sorting using CD11c ultrapure magnetic beads by Miltenyi. Bone marrow cells cultured with 25 ng/mL of murine M-CSF in RPMI supplemented with Pen/Strep and 10% FBS generated macrophages.

Isolation of splenic-derived dendritic cells and macrophages

Mouse spleens were harvested prior to dissociation into a single-cell suspension using the GentleMACS dissociator (Miltenyi, Germany). Erythrocytes were then removed through the use of ACK lysis buffer. FACS-sorting using the BD FACS Jazz (BD Biosciences, San Jose, USA) was employed to isolate CD11c+MHC class II + CD24+CD172a^neg cDC1, CD11c+MHC class II+CD24^negCD172a + cDC2, or F4/80+CD11b^loCD64+CD11c^neg subsets from total splenocytes. The sorting approach utilized FSC-H vs FSC-A to discriminate out doublets and Live/Dead fixable dyes to gate on live populations. The procedure yielded greater than ~97% purity, as corroborated by flow cytometric analyses. Cells were then utilized for in vitro T cell priming, ex vivo pathogen-infection studies and measurements of cyclic nucleotides and protein kinase activities.

Pathogen challenge in vivo experiments

Listeria monocytogenes strain ΔactA-LM-2W and the 2W:I-Ab tetramers were kindly provided by Drs. Marc Jenkins and Thamotharampillai Dileepan (University of Minnesota, Minneapolis, MN). This LM strain has been shown to cause transient infection after intravenous (i.v.) injection, strong proliferation of responder CD4+ T cells, and subsequent rapid clearance from the system. Pde1b−/− mice at 8–12 weeks of age were injected intravenously with 1 × 10⁷ LM resuspended in 150 µL of PBS. This dose corresponds to a sublethal systemic infection that permits quantification of bacterial clearance and adaptive immune responses without mortality during the experimental window. Inocula were prepared from log-phase cultures and confirmed by plating to verify input CFU. Four to six days post-infection, animals were sacrificed to harvest the spleen and draining lymph nodes (dLN). Tissues were then processed into a single-cell suspension, and erythrocytes were removed prior to staining with 2W:I-Ab tetramers and antibodies. In some experiments, magnetic bead isolation using antibodies targeting the tetramer fluorochrome was employed to enrich tetramer-bound LM-specific T cells, as previously described (Dileepan et al, 2021; Kotov and Jenkins, 2019). Cells were then analyzed by standard flow cytometric analysis for frequency of 2W:I-Ab specific CD4+ T cells.

Adoptive transfer of dendritic cells in vivo

DC were isolated from spleens of WT or Pde1b−/− mice by fluorescence-activated cell sorting (FACS) of CD11c+MHC class II+ cells, with further enrichment of cDC1 populations defined as CD24+CD172a^neg subsets where indicated. Sorted populations had >95% purity. Recipient Pde1b−/− mice (8–12 weeks of age) received 2 × 10⁵ sorted cDC in 100 µl via intravenous (i.v.) injection. 24 h after adoptive transfer, mice were infected intravenously with LM expressing ovalbumin (LM-OVA). Transferred cDC were not pre-loaded with antigen ex vivo, allowing antigen acquisition and presentation to occur in vivo during infection. Spleens were harvested 4–6 days post-infection, and antigen-specific CD8+ T cell responses were quantified by flow cytometry.

Bacterial burden quantification

Spleens were aseptically harvested from infected mice and homogenized in sterile PBS. Serial dilutions of homogenates were plated on brain heart infusion (BHI) agar plates and incubated at 37 °C overnight. Colony-forming units (CFU) were counted the following day and normalized to whole spleen homogenate per animal.

Isolation, activation and proliferation of T cells by dendritic cells in vitro

Naive T cells derived from OT mice were isolated using negative depletion approaches through magnetic columns (Miltenyi, Germany). Briefly, spleen and lymph nodes from either OT-I or OT-II mice were isolated. Tissues were subjected to the GentleMACS dissociator (Miltenyi, Germany) to generate single-cell suspensions. Next, OT-II-derived cells were labeled with antibodies to CD8 and MHC class II to deplete the subsets to isolate CD4 + T cells. Similarly, antibodies to CD4 and MHC class II were used to purify the CD8+ T cells in the OT-I model. These cells then served as responders. To prime these responder naive T cells, Flt3L-generated DC were cultured with HKLM encoding ovalbumin. In select studies, DC were FACS-sorted to isolate the CD24+CD172a^neg cDC1 and CD24^negCD172a + cDC2 subsets. After 5 h of in vitro culture and successive washing, HKLM-fed DC were used to prime CFSE-labeled T cells for 3–4 days in vitro. Proliferation capacity was then measured by monitoring fluorescence intensity dilutions by flow cytometric analysis. In parallel experiments, RNA was isolated for gene expression analysis, and the supernatant was collected for cytokine assessment by ELISA and Luminex Multiplex.

Pharmacologic inhibition of PDE

Total splenocytes, DC or macrophages were isolated from the mice spleen and immediately plated in 96-well plates at equal density in complete RPMI. To assess isoform-specific contributions of phosphodiesterases to cyclic nucleotide signaling, cells were pretreated for 30–60 min with pharmacologic inhibitors prior to stimulation. The following inhibitors were used: DSR-141562 (Pde1b-preferring; 45 and 150 nM), ITI-214 (pan-PDE1; 10 and 30 nM), roflumilast (PDE4; 1 and 3 nM) and PDE7-IN-2 (PDE7; 1 and 3 µM). Concentrations were based on reported IC₅₀ values. Final DMSO concentration was ≤0.1% in all conditions. Cells were subsequently stimulated with high-dose HKLM or infected with live LM. Cells were harvested 6-8 h post-stimulation for RNA isolation or 18–24 h for supernatant collection and subsequent ELISA measurements.

Flow cytometric analysis and fluorescence-activated cell sorting (FACS)

Cells were prepared for staining in 0.5% BSA and 2 mM EDTA in PBS with NaN3. Fluorochrome-labeled antibodies were utilized for extracellular staining. For intracellular antibody labeling, cells were fixed in 3% PFA prior to permeabilizing with 0.15% saponin in PBS and staining with primary and secondary antibodies. For phosphoprotein staining, cells were fixed and permeabilized using True-Phos Perm Buffer (BioLegend), following manufacturer instructions to preserve phosphorylation states prior to antibody staining. Cells were analyzed using the BD FACS Lyric flow cytometric analyzer (BD Biosciences, USA) and datasets evaluated using FlowJo (FlowJo LLC, BD Biosciences, USA). All gates were established using unstained and isotype controls. For fluorescence-activated cell sorting (FACS), isolated single cell suspensions were prepared for staining with 1% BSA and 2 mM EDTA in PBS. Cells were stained with antibodies for 20 min prior to washing. Labeled cells were then sorted using the BD FACSJazz flow cytometric sorter (BD Biosciences, USA).

Real-time quantitative PCR (qPCR)

To analyze gene expression, total RNA was isolated from cells using Trizol (Thermo Fisher Scientific, Waltham, USA) and reverse transcribed into cDNA using Veriti Thermocycler (Thermo Fisher Scientific, Waltham, USA). For qPCR, TaqMan Gene Expression Fast Advanced Master Mix and Gene Expression Assay kits were used to quantify expression levels. Taqman probes to Pde1b (Mm00476643_m1), Irf8 (Mm00492567_m1), Irf4 (Mm00516431_m1), Adgre1 (Mm00802529_m1), Ly6c (Mm03009946_m1), Listeria monocytogenes (Ba07921936_s1, Ba07922252_s1), Nos2 (Mm00440502_m1), Arg1 (Mm00475988_m1), Il6 (Mm00446190_m1), Tnfa (Mm00443258_m1), Il12a (Mm00434169_m1), Il10 (Mm01288386_m1), Tgfb1 (Mm01178820_m1), Il17a (Mm00439618_m1), Ccr9 (Mm02528165_s1), Cd74 (Mm00658576_m1), and Creb1 (Mm00501607_m1). For normalization, housekeeping probes used in the qPCR multiplex assays included GAPDH (Mm99999915_g1) and ACTB (Mm02619580_g1). All probes ordered from Thermo Fisher Scientific (Waltham, MA, USA). Samples were acquired and analyzed using the QuantStudio 6 qPCR instrument (Thermo Fisher Scientific, Waltham, MA). Expression levels of target transcripts were calculated by the comparative Ct method (2–ΔΔCt formula).

Luminex and ELISA assays

Luminex was used for multiplex quantitation of supernatants from cell cultures. For long-term storage, supernatants were collected and stored at −80 °C prior to analyses. The Multiplex Luminex Protein Assay (Thermo Fisher Scientific, USA) kit for detecting IFNγ, IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12p70, IL-17A (CTLA-8), IL-22, TGFβ1, and TNFɑ. Procedures followed manufacturer-recommended protocols. To corroborate multiplex analyses, singleplex ELISA assays were performed to detect IL-6, IL1β, IFNγ, IL-12p70, IL-10, TGFβ, and TNFɑ (R&D Systems; BD Biosciences).

Cytokine, protein kinase, and cGMP/cAMP immunoassays

Levels of cAMP and cGMP were measured using an ELISA kit in accordance with the manufacturer's instructions (Enzo Life Science, Farmingdale, NY, USA; Abcam, Boston, MA, USA). For steady-state analyses, FACS-sorted splenic cDC1, cDC2, and macrophage populations were used. For pharmacologic inhibition studies, splenocytes or macrophages were treated with indicated inhibitors in the presence or absence of HKLM stimulation. Intracellular cAMP levels were then measured at defined time points following treatment. PKA and PKG activities were quantified using colorimetric kinase activity assays (Thermo Fisher Scientific, USA) and normalized to total protein content in each sample to account for variation in cell number and extraction efficiency.

Western blotting

Protein lysates were generated using NP-40 lysis buffer supplemented with protease inhibitor cocktail (Thermo Fisher, USA) for protein extraction prior to running on a 4–12% SDS-PAGE gel. Protein was then wet transferred to a nitrocellulose blot, blocked in 5% dry milk and subsequently probed with primary and secondary fluorescent-tagged antibodies diluted in TBS-Tween. Blot images were captured using the Licor Odyssey (Licor Biosciences, Lincoln, Nebraska, USA). Quantitative analyses were performed using the ImageStudio software (Licor Biosciences). Band intensities were quantified using the densitometric analysis function by measuring the integrated pixel intensity of target protein and loading control bands in separate fluorescence channels, and target protein expression was normalized to the loading control (i.e. GAPDH, Vinculin) by calculating the ratio of target band intensity to the corresponding loading control intensity.

Statistics and reproducibility analyses

All experiments were performed using biological replicates across at least three separate animal cohorts, unless otherwise indicated in the figure legends. Where applicable, technical replicates (such as qPCR duplicates or triplicates) were performed within assays to ensure measurement accuracy. Statistical analyses were performed using two-tailed unpaired Student’s t-tests. These tests assume approximate normal distribution of the data and similar variance between groups; given the sample sizes and consistency with standard practice in the field, these assumptions were not formally tested. Variance between groups was comparable based on the similar dispersion of data points. Data were presented as mean ± SEM, as indicated in the figure legends. Statistical significance is denoted as p < 0.05 (*), p < 0.01 (**), and ns (not significant); exact p values are provided in the figures. No formal sample size calculations were performed. Sample sizes were selected based on prior experience with similar experimental systems and are consistent with standard practices in the field. Predefined exclusion criteria included failed infection, unsuccessful intravenous (i.v.) cell transfer, or technical issues during tissue processing. All exclusions are reported in the corresponding figure legends. No data points were excluded unless meeting these criteria. Investigators were not blinded to group allocation during experiments or outcome assessment. No formal randomization procedure was used; animals were assigned to experimental groups based on genotype and handled identically across conditions to minimize potential bias. This study was conducted and reported in accordance with the ARRIVE guidelines for animal research.

Use of AI-assisted technologies

During the preparation of this work, the authors used AI-assisted technologies to assist with condensing and refining the discussion section and the figure legends for clarity and conciseness. Following this assistance, all content was carefully reviewed and edited to ensure accuracy and consistency. Authors take full responsibility for all content of the publication.

The authors declare that they have no competing interests that could be perceived to influence the research, presentation, or interpretation of the data in this study. The authors report no financial, personal, or intellectual conflicts of interest.