lgM-deficiency leads to reduced airway hyperresponsiveness and class switching to IgE in HDM-induced asthma.

(A) Schematic diagram showing sensitisation and challenge protocol where mice (IgM KO) and wild type littermate control (WT) were sensitised with HDM 1 μg intra-tracheally on days 0 and challenged with HDM 3 μg on days 8-12. Analysis was done on day 15. (B) Airway resistance and elastance were measured with increasing doses of acetyl methacholine (0 −40 mg/mL). (C) Total lung eosinophil numbers (live+Siglec-F+CD11c-) and B cells (live+B220+CD19+MHCII+) were stained and analysed by Flow cytometry and enumerated from % of live cells. (D) Muc5a gene expression in whole lung tissue. (E) Total IgE and HDM-specific IgE in serum. (F) IgG1 and IgM surface expression in mediastinal lymph node B cells of WT, IgM KO and μMT KO mice. (G) Marginal Zone (live+B220+CD19+MHCII+CD21/CD35+CD23-), follicular (live+B220+CD19+MHCII+CD23+CD21/CD35+) and Germinal Centre (live+B220+CD19+MHCII+GL7+FAS+) B cells in the mediastinal lymph node of WT and IgM KO mice challenged with HDM. Shown is mean ± SEM from two pooled experiments (n=7 - 10). Significant differences between groups were performed by Student t-test (Mann-Whitney) (c, d, e) or by Two-Way ANOVA with Benforroni post-test (b) and are described as: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

lgM-deficiency does not lead to reduced Th2 allergic airway inflammation and serum transfer restores IgE, but not AHR.

Fig. a-c, mice treated as in Figure 1A. (A) Total mediastinal lymph node CD4 T cell numbers (live+CD3+CD4+) and % of Follicular Helper T cells (live+CD3+CD4+PD-1+CXCR5+) of CD4 T cells were stained and analysed by Flow cytometry and enumerated from % of live cells. (B) Mediastinal lymph nodes were stimulated with anti-CD3 (10μg/mL) for 5 days and supernatants were used to measure levels of IL-4 and IL-13. Cytokines were not detected in unstimulated or HDM (30μg) stimulated mLN. (C) Representative FACS plots and frequencies of lung CD4 T cells (live+CD3+CD4+) producing IL-4 and IL-13 after 5 hr stimulation with PMA/ionomycin in the presence of monensin. (D) Schematic diagram showing serum transfer from WT to IgM KO which were then sensitised as shown in Figure 1,A. (E) Airway resistance was measured with increasing doses of acetyl methacholine (0-40 mg/mL). (F) Total IgE in serum of mice either transferred with WT serum, IgM KO serum or no serum. Shown is the mean ± SEM from two pooled experiments (n=5 - 8). Significant differences between groups were performed by Student t-test (Mann-Whitney) (C, D, E) or by Two-Way ANOVA with Benforroni post-test (B) and are described as: *p<0.05, **p<0.01, ***p<0.001, ****p< 0.0001.

Partial wild type bone marrow replenishment restores AHR in IgM-deficient mice (Figure A-C, mice treated as in Figure 1A).

(A) Schematic diagram showing WT and IgM-deficient mice being chemically irradiated with busulfan (25 mg per day for 3 days) and adoptively transferred with congenic bone marrow (10×106 per mouse intravenously) at day 4. Mice (WT to IgM KO plus bone marrow) were then rested for 8 weeks before being sensitised as shown in Figure 1,A. (B) Airway resistance and elastance were measured with increasing doses of acetyl methacholine (0 −40 mg/mL). (C) Total lung neutrophils (live+CD11b+Ly6G+), eosinophils (live+Siglec-F+CD11c-), inflammatory macrophages (live+Ly6G-CD11b+F4/80+), CD4 T cells (live+CD3+CD4+CD8-) and B cells (live+B220+CD19+MHCII+) were stained and analysed by Flow cytometry and enumerated from % of live cells. Shown is the mean ± SD from one experiments (n=5 – 6 per group). Significant differences between groups were performed by Student t-test (Mann-Whitney) (C) or by Two-Way ANOVA with Benforroni post-test (B) and are described as: *p<0.05, **p<0.01, ***p<0.001, ****p< 0.0001.

Genes associated with muscle contraction are downregulated in IgM-deficient mice.

WT and IgM KO mice were treated as in Figure 1A and RNA was collected from the whole lung for RNA sequencing. (A) Principal-component (PC) analysis showing variation in the global gene expression profiles across the different groups. PC1 (60%) and PC2 (18%), which capture the greatest variation in gene expression, are shown. Orange colour represents WT HDM, green colour represents WT PBS, Blue colour represents IgM KO HDM and purple crosses represent IgM KO PBS. Each dot represents an individual mouse. (B) Number of differentially expressed genes between groups. (C) Heatmaps depicting the differently expressed genes between WT and IgM KO samples from HDM-treated and PBS mice ranked based on hierarchical clustering. (D-E) Volcano plots: numbers and colour relate to genes that have an adjusted p value <0.05. Blue, significantly downregulated; red, significantly up regulated; grey, non-differentially expressed. P values were adjusted for multiple testing using the Benjamini-Hochberg method. (D) represent changes between WT and IgM KO treated with HDM and (E) represents changes between WT and IgM KO treated with saline. (F) Gene set enrichment analysis (GSEA) of hallmark gene sets from the Molecular Signatures Database of the Broad Institute, showing the normalized enrichment scores (NES) for lung RNA-Seq data from WT mice.

BAIAP2L1 is expressed in close contact to smooth muscle.

(A) Mouse lungs were homogenised in RIPA buffer and blotted on nitrocellulose. Rabbit anti-human BAIAP2L1 and mouse anti-GAPDH was used as primary antibody. Lines 1-3 is WT mice, lines 4-6 is IgM KO mice sensitised and challenged with HDM. (B) Lung sections from WT and IgM KO mice sensitised and challenged with HDM were immunostained for nuclei stained DAPI (Blue), anti-BAIAP2L1 (red), α smooth muscle actin (green) and merged images (Magenta). Insert below shows zoomed in image of merged BAIAP2L1 and α smooth muscle actin. (C) Representative flow cytometry plots showing BAIAP2L1 expression (Live+Singlets+CD45-α SMA+BAIAP2L1+) in WT and IgM KO treated with HDM or PBS. Quantification of % BAIAP2L1 in α smooth muscle actin is shown. Shown are representative images from 2 independent experiments (n= 3 mice per group).

CRISPR-based deletion of BAIAP2L1 leads to reduced smooth muscle contraction at a single-cell level.

(A) Bronchial smooth muscle cells (1.6×105 cells/well) were transfected with CRISPR-Cas9 single guide RNAs (scramble, BAIAP2L1 and ERDR1), stimulated with recombinant human IL-13 (100 ng/mL) and TNF-α (10 ng/mL) for 48 h. Cells were then transferred to elastomeric micropatterns, stimulated again with rIL-13 and rTNF-α and fixed before imaging on a StellarVision microscope. (B) Representative images of single BSMCs on micropatterns from scramble, BAIAP2L1 and ERDR1 stimulated with 10 ng/mL TNF-α. DNA was stained with DAPI, actin fibers with Phaloidin-565 and elastomeric micropatterns are coated in Fibronectin-488. Merged images are shown on the right. (C) Violin plots showing contraction of 50-100 cells/condition stimulated with 10 ng/mL TNF-α, individual dots represent a single cell contraction, where blue is scramble sgRNA, red is BAIAP2L1 sgRNA and green is ERDR1 is sgRNA. (D) Violin plots showing contraction of 50-100 cells/condition stimulated with 100 ng/mL IL-13, individual dots represent a single cell contraction, where blue is scramble sgRNA, red is BAIAP2L1 sgRNA and green is ERDR1 is sgRNA. Shown is mean ±SEMs from two pooled experiment (n=50 - 100). Significant differences between groups were performed by student t-test (Mann-Whitney) and p value is shown.

CRISPR-based deletion of BAIAP2L1 reduces smooth muscle contraction upon stimulation with acetylcholine.

(A) Bronchial smooth muscle cells (1.6×105 cells/well) were transfected with CRISPR-Cas9 single guide RNAs (scramble, BAIAP2L1 and ERDR1), stimulated with Acetylcholine (10µM) for 48 h. Cells were then transferred to elastomeric micropatterns, stimulated again with ACh (10µM) and fixed before imaging on a StellarVision microscope. (B) Representative images of single BSMCs on a single micropattern from unstimulated, scramble, BAIAP2L1 and ERDR1 stimulated with ACh (10µM). Shown are DAPI, actin (red), a green fluorescent micropattern and merged images. (C) Violin plots showing contraction of 50-100 cells/condition stimulated with ACh (10µM) individual dots represent a single cell contraction, where blue is scramble sgRNA, red is BAIAP2L1 sgRNA and green is ERDR1 is sgRNA. Shown is mean ±SD from 1 representative experiment of 2 independent experiments (n=50 - 100). Significant differences between groups were performed by student t-test (Mann-Whitney) and p value is shown.

Working model showing how Baiap2l1 could influence muscle contraction through influencing myosin and actin filament interaction.

© 2021, BioRender. This image was created using BioRender.com. It is not available under a CC-BY 4.0 licence and further reproduction of this panel would need permission from the copyright holder.

IgM deficiency leads to reduced airway hyperresponsiveness, but eosinophils and B cells are unaffected.

(A) Schematic diagram showing sensitisation and challenge protocol where mice (IgM KO) and wild-type littermate control (WT) were sensitised with HDM 100 μg intratracheally on days 0 and challenged with HDM 10 μg on days 8-12. Analysis was done on day 15. (B) Airway resistance and elastance were measured with increasing doses of acetyl methacholine (0 −40 mg/mL). (C) Frequencies of lung and mediastinal lymph node B cells (live+B220+CD19+MHCII+) and were stained and analysed by Flow cytometry. (D) Frequencies of lung eosinophils (live+Siglec-F+CD11c-) were stained and analysed by Flow cytometry. (E) HDM-specific IgM in serum. (F) Histology analyses of lung sections (scale 200μm), stained with periodic acid-Schiff. A.U., Arbitrary units (G) Quantification of follicular B cells, Marginal Zone B cells and Germinal Centre B cells. Shown is mean ±SEMs from two pooled experiment (n= 10-12). Significant differences between groups were performed by Two way ANOVA with Benforroni post-test and are described as: *p<0.05, **p<0.01, ***p<0.001, ****p< 0.0001.

Reduced airway hyperresponsiveness in lgM-deficient mice is independent of allergen

IgM KO and wild type (WT) were sensitised intraperitoneally with (50µg in 200µl) of ovalbumin (OVA) adsorbed to 0.65% alum on days 0, 7, 14. On days 23, 24, 25, mice were intranasally challenged with 100µg of OVA. AHR was measured on day 26. (A) Airway resistance and elastance were measured with increasing doses of acetyl methacholine (0 −40 mg/mL). (B) Frequencies of lung eosinophils (live+Siglec-F+CD11c-) were stained and analysed by Flow cytometry. (C) IgM KO and wild type (WT) were challenged with 25 μg of papain on days 1, 2 and 3 or PBS. AHR was measured on day 4. Airway resistance and elastance were measured with increasing doses of acetyl methacholine (0 −40 mg/mL). (D) Frequencies of lung B cells (live+B220+CD19+MHCII+) and eosinophils (live+Siglec-F+CD11c-) were stained and analysed by Flow cytometry. Shown is mean ±SEMs from two pooled experiment (n= 10-12). Significant differences between groups were performed by Two way ANOVA with Benforroni post-test and are described as: **p<0.01, ***p<0.001, ****p< 0.0001.

Reduced airway hyperresponsiveness in lgM-deficient mice is independent of mouse background in HDM-induced asthma.

(A) Mice (IgM KO and WT) were sensitised and challenged as in Figure 1, A and Airway resistance and elastance were measured with increasing doses of acetyl methacholine (0 −320 mg/mL). (B) Histology analyses of lung sections (magnification x20), stained with periodic acid-Schiff. A.U., Arbitrary units. (C) Total IgE, HDM-specific IgM, HDM-specific IgG1, HDM-specific IgG2a and HDM-specific IgE in serum. Shown is mean ±SD from 1 representative experiment of 2 independent experiments (n= 4-6). Significant differences between groups were performed by student t-test (Mann-Whitney) (C) or Two way ANOVA with Benforroni post-test and are described as: *p<0.05, **p<0.01, ***p<0.001, ****p< 0.0001.

B cell development is not impacted by lack of lgM, however, IgD expression is upregulated in all tissues.

(A) B cell expression of IgM and IgD in IgM KO and WT mice treated with HDM in mediastinal lymph nodes, (B) Peritoneal and (C) Spleen. (D) Bone marrow was flushed out IgM KO and WT mice treated with HDM and stained for B1 cells (CD5+CD43+) and pre-B cells (CD43+BP-1+).

B cell and bone marrow reconstitution do not restore airway hyperresponsiveness in lgM-deficient mice in HDM-induced allergic asthma.

(A) Schematic diagram showing sorted B cells (live+B220+CD19+) (2-5×106) or bone marrow cells (10×106) transferred from congenic WT (CD45.1) to IgM KO a day before being sensitised as shown in Fig. 1,A. (B) Airway resistance was measured with increasing doses of acetyl methacholine (0-40 mg/mL). (C) Total IgE and IgM in serum of mice reconstituted with WT B cells or bone marrow. (D) House Dust Mite specific IgE and IgG1 in IgM KO mice transferred with WT serum (mice treated as in Figure 2D). Shown is mean ±SD from 1 experiment (n= 4-6). Significant differences between groups were performed by Two-way ANOVA with Benforroni post-test and are described as: ***p<0.001.

Bone marrow chimaera generation using busulfan to restore IgM function.

(A) Schematic diagram showing WT and IgM-deficient mice being chemically irradiated with busulfan (25 mg per day for 3 days) and adoptively transferred with congenic WT or syngeneic IgM KO bone marrow (10×106 per mouse intravenously) 24 hours post-last busulfan treatment. Recipient mice (WT to IgM KO) were then rested for 8 weeks before being sensitised as shown in Fig 1a. (B) FACS plots of bone marrow cells from either vehicle (10% DMSO) or busulfan (75 mg/kg) treated mice, showing live cells (BV605 live/dead dye), Lineage positive (Lin+) and negative cells (Lin-), frequencies of Sca-1+c-Kit+ cells within Lin+ and frequencies of CD45+ cells within Sca-1+c-Kit+ cells. (C) Quantification of frequencies of lineage negative cells (Lin-) and frequencies of CD45+ cells within Sca-1+c-Kit+ cells. (D) Quantification of frequencies of CD45.1+ cells within live lymphocytes in lung cells of WT or IgM KO busulfan treated recipient mice that received donor WT CD45.1 bone marrow, 8 weeks post busulfan treatment. Shown is the mean ± SD from one experiments (n=5 – 6 per group). Significant differences between groups were performed by Student t-test (Mann-Whitney) (C) and are described as: *p<0.05.

Reduced airway hyperresponsiveness in lgM-deficient mice is independent of microbial influence in HDM-induced asthma.

(A) Schematic diagram showing treatment of IgM KO and WT mice with antibiotic mixture 3 times a week for 2 weeks via oral gavage. These mice were then sensitised and challenged as in Figure 1A. Depicted are the number of colonies from TSA agar plates grown for 16 hrs at 37°C from faecal samples of WT mice treated (Abx) or not treated with an antibiotic cocktail (No Abx). (B) Airway resistance and elastance were measured with increasing doses of acetyl methacholine (0 −40 mg/mL). (C) Total IgE, HDM-specific IgG2a, HDM-specific IgG1, HDM-specific IgE and HDM-specific IgM in serum. Shown is mean ±SDs from 1 experiment (n= 4-6). Significant differences between groups were performed by student t-test (Mann-Whitney) and are described as: *p<0.05, **p<0.01, ***p<0.001, ****p< 0.0001. Abx, Antibiotic.

Genes associated with muscle contraction are downregulated in IgM-deficient mice.

Gene set enrichment analysis (GSEA) showing gene ratio of activated and suppressed pathways from lung RNA-seq data from WT mice and IgM KO mice. Related to Figure 4, F.

BAIAP2L1 is mainly expressed by alpha-smooth muscle cells.

(A) Flow cytometry gating strategy showing expression of BAIAP2L1 in CD45+ cells (live+Singlets1+Singlets2+CD45+BAIAP2L1+), alpha-smooth muscle positive cells (live+Singlets1+Singlets2+CD45-αSMA+BAIAP2L1+) and alpha-smooth muscle negative cells (live+Singlets1+Singlets2+CD45-αSMA-BAIAP2L1+). (B) Quantification of expression of BAIAP2L1 in CD45+ cells (live+Singlets1+Singlets2+CD45+BAIAP2L1+) and alpha-smooth muscle negative cells (live+Singlets1+Singlets2+CD45-αSMA-BAIAP2L1+).

Targeting and validation of BAIAP2L1 deletion by CRISPR.

(A) Schematic of human BAIAP2L1 coding region showing conserved N-terminal I-BAR domain and C-terminal SH3-like domain. (B) Genomic sequence of show target sgRNA1 region (yellow) and primer 2 regions (grey). (C) DNA gel (1.6%) from PCR showing DNA ladder and a 1000bp product in stimulated scramble transfected sgRNA (lanes 3-4) and reduced expression of the band in stimulated BAIAP2L1 transfected sgRNA (lanes 5-6) and unstimulated cells (lanes 1-2) using primer 2 region for PCR. (D) Lactate dehydrogenase (LDH) assay showing absorbance at 490 nm in supernatants from sgRNA transfected cells treated with media alone, TNF-α or acetylcholine. Positive control are dead cells killed by 1% Triton X and control is media alone.