S. suis serotype 2 strain 10 (wildtype, wt), kindly provided by Hilde Smith (Wageningen University and Research, Lelystad, Netherlands), was grown in Todd-Hewitt-Broth (THB) (Bacto, BD, Heidelberg, Germany, catalogue 249240) or on Columbia blood agar plates (Thermo Oxoid, Schwerte, Germany, catalogue PB5039A), at 37°C with 5% CO₂. The isogenic in frame deletion mutant S. suis 10Δide _Ssuis ∇ide _Ssuis _C195S (∇ide _Ssuis _C195S) was generated in a previous study (16) and cultivated the same way as the wildtype. Escherichia coli BL21 carrying either pETide _Ssuis _homologue (12) or pETide _Ssuis _homologue_C195S (16) were cultured in Luria-Bertani (LB) medium with 100 μg/mL ampicillin.

Recombinant Ide _Ssuis _homologue (rIde _Ssuis _h) and recombinant Ide _Ssuis _homologue_C195S (rIde _Ssuis _h_C195S) were expressed in Escherichia coli BL21 pETide _Ssuis _homologue and BL21 pETide _Ssuis _homologue_C195S, respectively, after addition of isopropyl-β-D-thiogalactopyranoside (IPTG) during the exponential growth phase and lastly purified through Ni-affinity chromatography as described previously (12, 16).

S. suis 10 (wildtype, wt) and 10Δide _Ssuis ∇ide _Ssuis _C195S were grown in Iscove′s Modified Dulbecco′s Medium (IMDM, PAN Biotech, Aidenbach, Germany, catalogue P04-20150) supplemented with 5% heat-inactivated fetal calf serum (FCS) until the late exponential growth phase (OD_600nm 1.0 – 1.3) and then centrifuged at 3,400 x g for 10 min. The culture supernatant was collected and sterile filtered (0.45 µm, PES-Membrane, Sarstedt via Schubert Laborfachhandel, Leipzig, Germany, catalogue 1826-001). 24-fold concentration of Ide _Ssuis or Ide _Ssuis _C195S, respectively, was performed using Vivaspin 20^® filters with a 100 kDa cutoff (Sartorius, Göttingen, Germany, catalogue VS2042) according to manufacturer information. Activity of Ide _Ssuis and Ide _Ssuis _C195S in concentrated supernatants was verified by Western blot analysis.

Western blot analysis was conducted to detect Ide _Ssuis or its derivatives as well as immunoglobulins and respective cleavage products. To verify cleavage activity of Ide _Ssuis of S. suis on soluble IgM, 100 µl 24-fold concentrated culture supernatant was preincubated with 1:100 diluted porcine serum for 2 h at 37°C on a rotator. To detect secreted IgM in B cell culture supernatants, porcine mandibular lymph node cells and PBMCs were incubated with rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S for 45 min at 37°C, followed by removal of the recombinant proteins and cleavage products by extensive washing. Receptor cleavage was confirmed by flow cytometry as described below. Cells were resuspended in IMDM culture medium and incubated for 21 h. Cell culture supernatant was collected directly, 1 h and 21 h after washing and investigated in an anti-IgM Western blot.

Proteins were separated in SDS-Page under reducing conditions using 4% stacking gels and 10% separating gels, followed by semi-dry blotting onto a nitrocellulose membrane (Roth, Arlesheim, Switzerland, catalogue HP40.1). The membrane was blocked for 1 h at room temperature (RT) or overnight at 4°C under constant shaking in Tris-buffered saline with 0.05% Tween (TBST) supplemented with 5% skim milk powder. All primary and secondary antibodies and corresponding dilutions used are specified in Supplementary Table 1 . Antibody incubations were performed in TBST with 1% skim milk powder at RT under constant shaking. Proteins were analysed using WesternBright™ chemiluminescence substrate (Biozym, Hessisch Oldendorf, Germany, catalogue 541014) according to manufacturer’s instructions. Detection was carried out using Fusion SL imaging system and FusionCapt Advance Solo 4 software (both by Vilber Lourmat, Eberhardzell, Germany). Marker bands were overlaid automatically.

All chromatographic purification procedures were carried out with the ÄCTA prime plus chromatography system (GE Healthcare, Little Chalfont, United Kingdom). Size exclusion chromatography (SEC) was performed with the HiLoad 16/600 Superdex 200pg column (GE Healthcare via VWR, Darmstadt, Germany, catalogue 28-9893-35) under the following conditions: flow rate 1 mL/min, 2 mL sample load, collected eluate 2 mL/tube and phosphate buffered saline (PBS, pH 7.35) as running buffer. For affinity chromatography and pre-absorption steps, proteins were covalently bound to CNBr-activated Sepharose 4B (GE Healthcare, Little Chalfont, United Kingdom, catalogue 17-0430-01) according to manufacturer instructions (instruction #71-7086-00 AF). Elution was performed with 0.1 M citric acid/sodium phosphate, pH 2.5 and the eluates were neutralized with 0.5 M trisodium phosphate.

For purification of IgM, porcine serum was precipitated with 6% polyethylenglycole 4000 (PEG4000, Carl Roth, Arlesheim, Switzerland, catalogue 0156), the precipitate was then solubilized in PBS and purified by affinity chromatography with Sepharose columns covalently coupled to Protein G (HiTrap protein G column, GE Healthcare via VWR, Darmstadt, Germany, catalogue 29–0485-81), followed by size exclusion chromatography. The eluates with corresponding size to IgM were pooled and concentrated with 10 kDa cutoff centrifugal filters (VIVASPIN 20^®, Sartorius, Göttingen, Germany, catalogue VS2001).

Anti-porcine IgM F(ab')₂ polyclonal antibodies were generated as described in (16). Briefly, purified porcine IgM was cleaved by rIde _Ssuis (20 µg rIde _Ssuis /mg IgM) and thereby generated IgM F(ab')₂ fragments were purified by SEC (see above) and covalently coupled to CNBr-activated Sepharose 4B columns. Within a previous study (16), the purified IgM F(ab')₂ fragments were also used as antigen for rabbit immunization consisting of one priming and two booster immunizations. A part of the generated rabbit antiserum was precipitated with 40% saturated ammonium sulfate solution and the precipitate was solubilized in 0.1 M sodium acetate buffer, pH 4.5. One milligram pepsin (3829 U/mg, Fluka Analytical via Sigma-Aldrich, Darmstadt, Germany, catalogue 77152) was added to 50 mg gammaglobulin, followed by incubation for 24h at 37°C. Resulting rabbit anti-IgM F(ab')₂ antibodies were first purified by affinity chromatography with IgM-F(ab')₂-coupled Sepharose 4B columns (see above). After neutralization (pH 7.5), the eluate was preabsorbed with porcine IgG-coupled Sepharose 4B columns (10 mg IgG/mL Sepharose) and the flow through was concentrated with 10 kDa cutoff centrifugal filters to 2 mL (VIVASPIN 20^®, Sartorius, Göttingen, Germany, catalogue VS2001). After a second purification step via SEC, a part of the rabbit anti-IgM F(ab')₂ antibody preparation was conjugated wit NHS-fluorescein (ThermoFisher Scientific, Schwerte, Germany, catalogue 46410) and purified following manufacturer instructions. The degree of labeling was calculated from spectrophotometric measurements as specified in the manufacturer’s instructions. The other part of the rabbit anti-IgM F(ab')₂ antibody preparation remained pure and was used in the stimulation experiments. The final purified polyclonal rabbit anti-porcine IgM F(ab')₂ antibody preparation (following referred to as anti-IgM F(ab')₂) was stored at 4°C in PBS supplemented with 0.1% ProClin™.

Mandibular lymph nodes and whole blood was collected from commercially farmed pigs (German Landrace x Pietrain or German saddleback pig) 11 months of age (data shown in Figures 1 , 2 , and Supplementary Figures 3, 6A ) or between 5 and 16 weeks of age (data shown in Figures 3 - 5 , and Supplementary Figures 2, 6B, 7–10 ).

For the preparation of peripheral blood mononuclear cells (PBMC) the blood was centrifuged at 400 x g for 10 min. Plasma was removed and the blood was diluted with the same amount of PBS. Density gradient centrifugation was performed over Ficoll (PAN Biotech, Aidenbach, Germany, catalogue P04-601000) at 900 x g for 35 min, followed by gently collecting the interphase with a pipette. Cells were washed with IMDM + 10% heat-inactivated FCS (500 x g, 12 min, RT) and again with PBS (400 x g, 10 min, RT). If necessary, erythrocyte lysis was performed by resuspending the cells in 1-2 mL lysis buffer (PBS supplemented with ammonium chloride, potassium hydrogen carbonate and ethylenediaminetetraacetic acid (EDTA)) and incubating for 2-3 min. Lysis was stopped by adding 1 mL heat-inactivated FCS and filling up with 10 mL IMDM + 10% FCS. After two washing steps (400 x g, 6 min, RT), cells were used directly or resuspended in heat-inactivated FCS supplemented with 10% dimethyl sulfoxide (DMSO, Sigma, Darmstadt, Germany, catalogue D2650-100ML) (freezing medium) and frozen at - 80°C.

Lymph nodes were minced and transferred into digestion medium consisting of IMDM supplemented with 1% penicillin/streptomycin, 50 µM gentamicin and DNase (final concentration 111.1 U/mL). After incubation for 15 min at 37°C, samples were homogenized using gentleMACS™ Dissociator according to manufacturer information. This step was repeated, followed by filtration through sterilized cotton wool and rinsing with PBS. Cells were washed three times with PBS at 400 x g and 4°C for 12 min and then resuspended in freezing medium and stored at -80°C.

Cells were thawed, counted with Trypan Blue and adjusted to 1 x 10⁷ cells/mL in IMDM medium supplemented with 10% heat-inactivated FCS and 1% penicillin-streptomycin (PAN Biotech, Aidenbach, Germany, catalogue P06-07100) (IMDM⁺⁺). Additionally, 10 ng/mL recombinant porcine IL-2 (R&D Systems, Minneapolis, US, catalogue 652-P2) was added to increase cell viability. Four million cells were incubated with 16 µg rIde _Ssuis _homologue, 16 µg rIde _Ssuis _homologue_C195S or 100 µl 24-fold concentrated culture supernatant from S. suis wt or 10Δide _{Ssui s}∇ide _Ssuis _C195S, respectively, for 45 min, 2 h, 3 h, or 4 h at 37°C with 5% CO₂. Then, cells were transferred into a 96 well plate with 4x10⁵ cells/well on ice, washed two times with cold PBS and stained 20 min at 4°C in 100 µl 1:500 prediluted viability dye. Cells were washed one time with cold PBS supplemented with 3% heat-inactivated FCS and 0.1% sodium acid (FACS buffer) and two times with cold PBS, followed by fixing for 30 min at 4°C with 100 µl 2% paraformaldehyde (PFA) and three times washing with cold FACS buffer.

Next, cells were stained for 15 min at 4°C in 20 µl of staining mix 1 containing anti-IgM Fc and anti-IgM F(ab')₂, F(ab')₂ control mix containing anti-IgM Fc and anti-rabbit IgG isotype control or IgM Fc control mix containing mouse IgG1 isotype control and anti-IgM F(ab')₂, respectively, diluted in FACS buffer. After two washing steps in FACS buffer, cells were blocked for 10 min at 4°C in 20 µl FACS buffer containing 30% porcine heat-inactivated serum (Fc block I). The cells were then stained in 20 µl of staining mix 2 containing anti-mouse IgG1 and anti-CD3 or CD3 control mix containing anti-mouse IgG1 and mouse IgG2a isotype control, respectively, diluted in Fc block I. Following 15 min incubation at 4°C, cells were washed once with FACS buffer and twice with FACS buffer containing 0.5% saponin (saponin buffer) to prepare for intracellular staining. After blocking with saponin buffer containing 30% porcine heat-inactivated serum (Fc block II) for 5 min at 4°C, cells were stained for 30 min at 4°C in 20 µl of staining mix 3 containing anti-CD79α diluted in Fc block II or CD79α fluorescence minus one (FMO) control mix without anti-CD79α. After washing once with saponin buffer and once with FACS buffer, cells were resuspended in 100 µl FACS buffer and measured by flow cytometry (LSRFortessa, BD Biosciences, Heidelberg, Germany). Analyses were carried out using FlowJo™ 10 software (BD Biosciences, Heidelberg, Germany) and gates were placed based on specific isotype controls. First, lymphocytes were gated forward scatter area (FSC-A) versus side scatter area (SSC-A) based on size and granularity. Next, single cells were gated forward scatter height (FSC-H) versus forward scatter area (FSC-A) to exclude doublets. Viable cells were distinguished using CD79α as B cell marker and CD3 as T cell marker in B cells (CD79α⁺CD3^-) and T cells (CD79α^-CD3⁺). IgM Fc⁺ CD79α⁺ B cells were analyzed for an intact (IgM Fc⁺ F(ab')₂ ⁺) or cleaved (IgM Fc⁺ F(ab')₂ ⁻) IgM B cell receptor.

For investigation of receptor cleavage on CD21⁺ or CD21^- cells, staining mix 2 contained anti-mouse IgG1 and anti-CD21 in Fc block I and staining mix 3 contained Streptavidin in Fc block I. All other steps were performed as described above.

All antibodies used are specified in Supplementary Table 2 . All washing steps were performed for 4 min at 400 x g or 500 x g after fixation, respectively.

Cells were thawed, counted and adjusted to 1 x 10⁷ cells/mL in IMDM⁺⁺ + poIL2 as described above. Four million cells were incubated with 16 µg rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S, respectively, for 45 min at 37°C with 5% CO₂. After removal of the recombinant proteins by two washing steps with IMDM⁺⁺, cells were incubated in IMDM⁺⁺ supplemented with porcine IL-2 to promote cell survival for up to 20 hours. For each investigated time point, 4 x 10⁵ cells/well were transferred into a 96 well plate onto ice and washed two times with cold PBS. All following steps were performed analogous to BCR cleavage experiments.

PBMC were thawed, counted and adjusted to 1 x 10⁷ cells/mL in IMDM⁺⁺ + poIL2 as described above. Four million cells were incubated with 0.5 mg/mL IgM for 30 min at 37°C or remained untreated, respectively. After washing with IMDM⁺⁺, cells were transferred into a 96 well plate with 4 x 10⁵ cells/well and washed two times with cold PBS. Staining for viability and fixating with 2% PFA were performed as described above. Next, cells were stained for 15 min at 4°C in 20 µl of staining mix 1 containing anti-IgM Fc or IgM Fc control mix containing mouse IgG1 isotype control, respectively, diluted in FACS buffer. After two washing steps in FACS buffer, cells were blocked for 10 min at 4°C in 20 µl FACS buffer containing 30% porcine heat-inactivated serum (Fc block I). The cells were then stained in 20 µl of staining mix 2 containing anti-mouse IgG1 and anti-CD172a or CD172a control mix containing anti-mouse IgG1 and mouse IgG2b isotype control, respectively, diluted in Fc block I. Following 15 min incubation at 4°C, cells were washed once with FACS buffer and twice with saponin buffer. After blocking with saponin buffer containing 30% porcine heat-inactivated serum (Fc block II) for 5 min at 4°C, cells were stained for 30 min at 4°C in 20 µl of intracellular staining mix containing anti-CD79α diluted in Fc block II or CD79α FMO mix without anti-CD79α. All antibodies used are specified in Supplementary Table 2 . Cells were measured by flow cytometry as described above.

We investigated the impact of BCR cleavage by rIde _Ssuis on the stimulatory capacity of anti-IgM F(ab')₂ and pervanadate. PBMC were used freshly isolated or thawed and adjusted to 1 x 10⁷ cells/mL in IMDM⁺⁺ + poIL2 as described above. Four million cells were incubated with 16 µg rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S, respectively, for 45 min at 37°C with 5% CO₂. After two washing steps with cold PBS, cells were stained for 20 min at 4°C in prediluted viability dye with anti-IgM Fc or mouse IgG1 isotype control, respectively. Then, cells were washed two times with cold IMDM⁺⁺ medium and resuspended in IMDM⁺⁺ + poIL-2. Cells were seeded into a 96 well plate with a concentration of 4 x 10⁵ cells/well and incubated for 20 min at 37°C.

To prepare pervanadate, 10 µl 0.1 M sodium orthovanadate (Sigma-Aldrich, Darmstadt, Germany, catalogue 450243-10G) was incubated with the same amount of 1:50 prediluted (0.6% final conc.) hydrogen peroxide (Carl Roth, Arlesheim, Switzerland, catalogue 8070.3) for 15 min at RT, followed by inactivation of surplus H₂O₂ by adding 30 µl catalase (Sigma-Aldrich, Darmstadt, Germany, catalogue C9322-1G) (20 mg/mL catalase in KH₂PO₄ buffer). The solution was centrifuged at 5000 x g for 5 min and the supernatant diluted 1:20 in buffer A (20 mM HEPES-NaOH, 120 mM NaCl, pH 7.4).

Cells were stimulated with anti-IgM F(ab')₂ (10 µg/mL final concentration) for 1, 2, or 4 min, respectively. Stimulation with 10 µl pervanadate (0.1 mM final concentration) was performed for 5, 10, 15, or 20 min. All agents used for stimulation are specified in Supplementary Table 3 . Stimulation was stopped all together by adding the same amount 4% PFA and incubation for 10 min at RT, followed by two washing steps with cold PBS and fixation with 2% PFA for 30 min at 4°C. After two times washing with cold FACS buffer, cells were stained for 15 min at 4°C in 20 µl of staining mix 1 containing anti-porcine IgG or the specific isotype control, diluted in FACS buffer. Next, cells were washed two times with cold FACS buffer, followed by 10 min incubation at 4°C in 20 µl Fc block I. Cells were then stained for 15 min at 4°C in 20 µl of staining mix 2 containing anti-mouse IgG1 and streptavidine, diluted in Fc block I. After two washing steps with cold PBS, cells were permeabilized with 100 µl ice-cold 99% methanol for 30 min on ice, followed by centrifugation, removal of supernatant and two times washing with cold FACS buffer. Next, cells were incubated in 20 µl FACS buffer with 30% murine serum (Fc block III) for 10 min at 4°C and then stained for 30 min at RT in 20 µl staining mix 3 containing anti-phospholipase C-γ2 or mouse IgG1 isotype control, respectively, diluted in Fc block III. After washing with FACS buffer twice, cells were incubated for 15 min at 4°C in 20 µl of staining mix 4 containing anti-CD21 or the respective isotype control, diluted in Fc block I. Cells were measured by flow cytometry as described above.

Cells for rIde _Ssuis cleavage control as well as IgM Fc and F(ab')₂ isotype controls were stained as described for BCR cleavage. All antibodies used for staining are specified in Supplementary Table 2 . All washing steps were performed for 4 min at 400 x g or 500 x g after fixation, respectively.

All data represent at least three independent experiments. All statistical analyses were performed with GraphPad Prism 9 (Dotmatics, Boston, Massachusetts, US, www.graphpad.com). Distribution of data was tested both with Shapiro-Wilk and Kolmogorov-Smirnov test. If data achieve normal distribution, unpaired two-tailed t-test was applied. In case of no normal distribution, Mann-Whitney test was conducted. For comparison of more than two measurements, mixed analysis of variance (ANOVA) with subsequent Turkey´s multiple comparisons test was used. Figures show mean and standard deviation or median. Probabilities below 0.05 were considered significant (p< 0.05 *, p< 0.01 **, p< 0.001 ***, p< 0.0001 ****). Flow cytometric data was analyzed with FlowJo™ 10 software (BD Biosciences, Heidelberg, Germany, www.flowjo.com).

Ide _Ssuis cleaves soluble porcine IgM and leads to a reduction of porcine IgM bound to the bacterial surface (12, 16). The cysteine protease IdeS of Streptococcus pyogenes is known to cleave soluble IgG as well as the IgG B cell receptor (32). Thus, we hypothesized that Ide _Ssuis is also able to cleave porcine IgM as a transmembrane protein within the B cell receptor complex. This hypothesis was first investigated with the following recombinant derivatives of Ide _Ssuis : the intact IgM protease domain (rIde _Ssuis _homologue), the point-mutated, inactive IgM protease domain (rIde _Ssuis _homologue_C195S) or the intact IgM protease domain together with the large C-terminus of still unknown function (rIde _Ssuis ) ( Supplementary Figure 1 ). Specifically, porcine mandibular lymph node cells and PBMCs were treated with these recombinant proteins and stained for IgM Fc⁺ B cells. Representative plots of flow cytometry analyses of mandibular lymph node cells are shown in Figure 1A . Viable lymphocytes were distinguished in B cells (CD79α⁺CD3^-) and T cells (CD79α^-CD3⁺). As Ide _Ssuis cleaves IgM between constant domain C2 and C3 (16), we chose a monoclonal anti-porcine IgM Fc antibody (K52 1C3), putatively recognizing the membrane bound C3-C4 IgM fragment after Ide _Ssuis treatment, to detect IgM⁺ B cells. Polyclonal rabbit anti-IgM F(ab')₂- specific antibodies were used to detect the F(ab) part of IgM and therefore an intact IgM BCR (IgM Fc⁺ F(ab')₂ ⁺). Lower panels of Figure 1A show representative pseudocolor plots of IgM⁺ B cells after treatment with the indicated recombinant proteins. We verified the cleavage activity of rIde _Ssuis containing the large C-Terminus in a representative number of experiments and found no significant difference in IgM Fc⁺ F(ab')₂ ⁺ B cells in comparison to the cleavage activity of rIde _Ssuis _homologue (data not shown). Therefore, we continued the following experiments only using rIde _Ssuis _homologue.

Indeed, flow cytometry analysis revealed a significant reduction of approximately sixty percent in IgM Fc⁺ F(ab')₂ ⁺ cells after treatment with rIde _Ssuis _homologue for 45 min, whereas non-functional rIde _Ssuis _homologue_C195S had no significant effect on the number of IgM Fc⁺ F(ab')₂ ⁺ cells ( Figure 1B ). After continuous incubation with rIde _Ssuis _homologue for 4 h, the reduction in IgM Fc⁺ F(ab')₂ ⁺ cells increased to over seventy percent. These findings for the 45 min incubation time period could be reproduced in PBMCs as well as in mLN cells of piglets of a different age ( Supplementary Figure 2 ).

Taken together, these data indicate that rIde _Ssuis _homologue as well as rIde _Ssuis are able to cleave membrane-bound IgM and that cysteine at position 195 is crucial for this cleavage activity, analogous to cleavage of soluble IgM (16). The large C-terminal domain of Ide _Ssuis is dispensable for this cleavage activity on the IgM BCR. Both active recombinant variants of Ide _Ssuis efficiently remove the F(ab')₂ part of the IgM BCR. The Fc part remained unaffected and was still detectable on B cells after up to 4 h treatment with rIde _Ssuis _homologue ( Supplementary Figure 3 ).

We further assessed whether Ide _Ssuis secreted by S. suis is able to cleave the IgM BCR. S. suis cps2 strain 10 (wt) and the isogenic S. suis mutant 10Δide_Ssuis∇ide_Ssuis_C195S, which expresses a point-mutated, inactive full-length variant of Ide _Ssuis (16), were grown in IMDM medium. The supernatant was 24-fold concentrated via 100 kDa cutoff centrifugal filters to retain Ide_Ssuis or Ide_Ssuis_C195S (both 124 kDa) while eliminating the largest possible proportion of other secreted bacterial components, especially potential B cell damaging factors like suilysin. Cleavage of soluble IgM by those culture supernatants was confirmed in anti-IgM Western blot analyses under reducing conditions ( Supplementary Figure 4 ).

Concentrated bacterial culture supernatants were incubated for 4 h with mandibular lymph node cells. Anti-Ide _Ssuis Western blot analysis revealed characteristic Ide_Ssuis protein bands between 175 and 270 kDa as described by Seele et al. ( Supplementary Figure 5 ) (12), which suggests that Ide_Ssuis and Ide_Ssuis_C195S are fairly stable under these conditions. For comparison, recombinant homologue proteins were also incubated with mandibular lymph node cells and examined in parallel. Typical protein bands were detectable between 52 and 66 kDa as described by Rungelrath et al. (16).

Analogous to BCR cleavage with recombinant enzymes, mandibular lymph node cells were incubated with concentrated culture supernatant and analyzed in flow cytometry for an intact (IgM Fc⁺ F(ab')₂ ⁺) or cleaved (IgM Fc⁺ F(ab')₂ ⁻) IgM BCR (gating described in Figure 1A ). A significant reduction of IgM Fc⁺ F(ab')₂ ⁺ B cells was detectable after treatment with S. suis wt supernatant for 45 min and increased to approximately thirty percent when incubated for 4 h ( Figure 1C ). In contrast, there was no detectable effect on IgM Fc⁺ F(ab')₂ ⁺ B cells after incubation with concentrated S. suis 10Δide _Ssuis ∇ide _Ssuis _C195S culture supernatant. These results demonstrate that not only rIde _Ssuis _homologue (as shown above) but also Ide _Ssuis expressed by S. suis is able to cleave the porcine IgM BCR.

To verify that the measured IgM signals originate from the IgM BCR and not secreted and then surface-bound soluble IgM, i.e. bound to FcµR (18, 33, 34), we investigated B cell culture supernatants in anti-IgM Western blot analysis. Cells were treated with rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S, followed by removal of the recombinant proteins and cleavage products and incubation for 21 h. Cell culture supernatant was collected directly, 1 h and 21 h after washing and investigated for soluble IgM. Neither cells preincubated with rIde _Ssuis _homologue nor rIde _Ssuis _homologue_C195S showed IgM-specific protein bands at any time point ( Supplementary Figure 6 ). These findings indicate that the cells do not produce soluble IgM that could bind to the cell surface. This is in accordance with the interpretation that the detected IgM F(ab')₂ signals rely on the expression of the IgM BCR and not on receptor-bound secreted IgM. Additionally, we investigated the capacity of PBMCs to bind soluble IgM to further exclude that soluble and subsequently surface-bound IgM was contributing to the IgM signals measured by flow cytometry. To this end, porcine PBMC were incubated with 0.5 mg/mL soluble porcine IgM and analyzed for IgM Fc⁺ cells by flow cytometry. Cells positive for myeloid marker CD172a, known to express IgM receptors on the cell surface (18), were used as positive control. No significant effect on the percentage of IgM Fc⁺ porcine B cells (IgM Fc⁺ CD79α⁺) could be measured after IgM preincubation ( Supplementary Figure 7 ). In contrast, CD172a⁺ cells showed a significant increase in the percentage of IgM Fc⁺ cells as well as in the median fluorescence intensity of the IgM Fc signal, indicating that porcine CD172a⁺ cells efficiently bind soluble IgM on the cell surface in contrast to B cells. This is also consistent with our interpretation that the IgM signal of B cells derives from transmembrane IgM as part of the IgM BCR.

In conclusion, these results show for the first time that rIde _Ssuis _homologue cleaves the IgM BCR as present on the surface of B cells (CD79α⁺ CD3^-) in porcine mandibular lymph node cells and PBMC.

After treatment of human PBMC with IgG-degrading enzyme IdeS of S. pyogenes, the number of IgG BCRs takes 16 h to return to pretreatment levels (32). To verify if the porcine IgM BCR is recovered in a similar time course, mandibular lymph node cells were incubated with rIde _Ssuis _homologue or non-functional rIde _Ssuis _homologue_C195S, followed by incubation for up to 20 h. Cells were stained for CD79α, IgM Fc and IgM F(ab')₂ and analyzed in flow cytometry for intact or cleaved BCR as described above ( Figure 1A ). The number of IgM Fc⁺ F(ab')₂ ⁺ B cells in mLN cells increased over time from approximately 20% at 0 h to almost 80% at 20 h after treatment with rIde _Ssuis _homologue. After 20 h, numbers of IgM Fc⁺ F(ab')₂ ⁺ cells were comparable between rIde _Ssuis _homologue and rIde _Ssuis _homologue_C195S treated cells ( Figure 2 ). These findings demonstrate that the IgM BCR signal takes approximately 20 h to fully recover. The full recovery of the IgM signal on CD79α⁺CD3^- B cells after rIde _Ssuis _homologue cleavage further supports the interpretation that this signal is due to the porcine IgM BCR and not to surface-bound soluble IgM.

BCR signaling plays a key role in B cell activation and differentiation. BCR ligand-dependent activation results in the recruitment of protein tyrosine kinases Lyn and Syk and downstream tyrosine phosphorylation of phospholipase C-γ2 (PLC-γ2) (35).

We hypothesized that IgM BCR cleavage leads to impaired B cell activation. Therefore, we incubated PBMC with rIde _Ssuis _homologue to cleave the BCR or non-functional rIde _Ssuis _homologue_C195S for comparison as described above. Cleavage of the IgM BCR was verified by flow cytometry as described in Figure 1A . Next, the cells were treated with different stimuli and analyzed by flow cytometry for tyrosine-phosphorylated PLC-γ2 (pPLC-γ2) as readout parameter for activation of the intracellular downstream signal cascade. IgM BCR specific stimulation was achieved by anti-IgM F(ab')₂. The gating strategy of IgM⁺ and IgG⁺ cells and representative histograms of IgM⁺ cells after 2 min stimulation with anti-IgM F(ab')₂ in comparison to medium control are shown in Figure 3A . Protein tyrosine phosphatase inhibitor pervanadate intervenes at the level of downstream signal transduction (31) and was used for BCR-independent increase of phosphorylation.

We investigated the stimulatory capacity and the consequence of BCR cleavage by rIde _Ssuis _homologue on the percentage of IgM Fc⁺ pPLC-γ2⁺ cells in PBMC of 7 to 8-week-old piglets and 16-week-old pigs. The cells of the different age groups behaved similarly ( Supplementary Figure 8 ). Hence, we thereupon focused on PBMC from 7 to 8-week old piglets as the typical age period for S. suis disease.

Stimulated with anti-IgM F(ab')₂, IgM Fc⁺ cells preincubated with rIde _Ssuis _homologue_C195S showed a significant increase in the percentage of pPLC-γ2⁺ cells in comparison to IgM⁺ cells preincubated with rIde _Ssuis _homologue ( Figure 3B ). These findings demonstrate that IgM BCR cleavage prevents recognition and ligand-dependent activation via the F(ab) part and therefore impairs intracellular signaling. As expected, IgG⁺ cells showed no increased phosphorylation of PLC-γ2 after stimulation with anti-IgM F(ab')₂ as well as no effect of preincubation with functional or non-functional rIde _Ssuis constructs. In contrast, the median fluorescence intensity (MFI) of pPLC-γ2⁺ did not increase after stimulation of IgM⁺ cells preincubated with rIde _Ssuis _homologue_C195S in comparison to IgM⁺ cells preincubated with rIde _Ssuis _homologue ( Supplementary Figure 9 ). Taken together, these data indicate that IgM BCR-specific stimulation induces a higher frequency of pPLC-γ2⁺ IgM⁺ cells, while the level of phosphorylation per single cell remains restricted to a certain limit when all PTKs are already recruited.

When treated with pervanadate, both IgM⁺ and IgG⁺ cells showed a time-dependent increase in pPLC-γ2⁺ cells independent of enzyme preincubation ( Figure 3B ). This demonstrates that rIde _Ssuis _homologue cleavage activity only affects the IgM BCR-mediated activation. Following 45 min preincubation with either rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S, the intracellular downstream signaling cascade appears to be still intact. In contrast to BCR-mediated stimulation, treatment with pervanadate led to a time-dependent increase in frequency as well as in MFI in both IgM⁺ and IgG⁺ cells ( Supplementary Figure 9 ). However, IgM⁺ cells do not reach quite the same high levels of phosphorylation as IgG⁺ cells. In conclusion, pervanadate stimulation increased the number of pPLC-γ2⁺ cells as well as phosphorylation levels per cell in the investigated IgM⁺ and IgG⁺ cells with no difference between IgM⁺ cells with intact or cleaved BCR.

Additionally, after stimulation with anti-IgM F(ab')₂ or pervanadate, cells were gated for IgM^- cells and analyzed for the percentage of pPLC-γ2⁺ cells as described above. In addition to IgG⁺ B cells, the IgM^- cell population also includes IgA⁺ and IgD⁺ B cells. As expected, IgM^- cells showed no increase in pPLC-γ2⁺ cells after IgM-specific stimulation, whereas pPLC-γ2⁺ cells increased significantly after treatment with pervanadate, independent from pretreatment with rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S ( Supplementary Figure 10 ). These results underline the specificity of polyclonal anti-IgM F(ab')₂ used for stimulation of IgM⁺ B cells via the IgM F(ab) part of the BCR.

To investigate Ide _Ssuis cleavage activity on different IgM-expressing B cell subpopulations, i.e. a CD21⁺ B2 population and a CD21^- population, containing putative B1-like cells (36), PBMC were treated with rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S, stained for IgM Fc and CD21 and analyzed in flow cytometry. Both IgM⁺CD21⁺ and IgM⁺CD21^- cells showed a significant reduction in IgM F(ab')₂ ⁺ cells after incubation with rIde _Ssuis _homologue in contrast to rIde _Ssuis _homologue_C195S treated cells ( Figure 4 ). In conclusion, rIde _Ssuis _homologue cleaves the IgM BCR present on B1-like and B2 cell subsets differentiated by CD21 expression.

We further assessed the consequence of IgM BCR cleavage on the signaling of CD21⁺ B2 cells compared to CD21^- B cells. Cells were pretreated with rIde _Ssuis _homologue or rIde _Ssuis _homologue_C195S before stimulation with specific anti-IgM F(ab')₂ or protein tyrosine phosphatase inhibitor pervanadate (as described above). Stimulation of both CD21 populations within IgM⁺ cells with anti-IgM F(ab')₂ after preincubation with rIde _Ssuis _homologue_C195S resulted in a distinct increase in pPLC-γ2⁺ cells in contrast to IgM⁺ cells preincubated with rIde _Ssuis _homologue ( Figure 5 ). Significant differences were found comparing CD21⁺ and CD21^- cells preincubated with non-functional rIde _Ssuis _homologue_C195S after a stimulation time of 1 min and 2 min, with higher levels of pPLC-γ2⁺cells within CD21^- cells. Phosphorylation of PLC-γ2 was completely abolished by preincubation with rIde _Ssuis _homologue, with no significant differences between CD21⁺ and CD21^- cells, indicating that both IgM⁺ cell subsets are equally impaired in downstream BCR signaling. After stimulation with pervanadate, both cell types showed an increase in pPLC-γ2⁺ cells independent from preincubation. Statistical analysis found higher percentages of pPLC-γ2⁺ cells within rIde _Ssuis _homologue_C195S-preincubated CD21^- cells after a stimulation period of 5 min as well as within rIde _Ssuis _homologue-preincubated CD21^- cells after a stimulation period of 5, 10, and 15 min in comparison to CD21⁺ cells. Taken together, significantly more CD21^- cells were positive for pPLC-γ2 at specific time points. In summary, these data demonstrate IgM BCR cleavage by rIde _Ssuis _homologue and subsequent impairment of intracellular signaling after ligand-dependent receptor activation on both naïve and innate B cell subsets.