Figure 1.
Bacterial induced TLR2/6 activation.
TLR2/6 expressing HEK transfectants were co-incubated with different strains of bifidobacteria or lactobacilli. (A) Dose response experiment indicating the effects of the different supplements (FBS = fetal bovine serum, HS = human serum, plasma = human plasma, MS = mouse serum) on TLR2/6 activity. TLR2/6 expressing cells were stimulated with B. breve NutRes 200 in a ratio of 15∶1 (bacteria:cell) for 16H. NFκB activity measured as Relative Light Units (RLU) was determined using a Luciferase reporting system as described in materials and methods. Relative NFκB activity was determined calculating the ratios between medium activity and the activity in the different serum fractions. (B) TLR2/6 activity induced by 3 Bifidobacteria and 3 Lactobacilli strains. TLR2/6-expressing cells were stimulated at a ratio of 15∶1 (bacteria:cell) in serum-free or 5% human serum (HS) containing medium. NFκB activity measured as RLU was determined as described above.
Figure 2.
EVs inhibit Bifidobacterium breve induced TLR2/6 activation.
TLR2/6-expressing HEK transfectants were stimulated with B. breve NutRes 200 in a ratio of 1∶15 for 16H. (A) Dose response experiment indicating the effects of the different human serum fractions on TLR2/6 activity (HS = intact human serum, HS-D = EV depleted human serum, HS-EV = human serum EVs in medium). NFκB activity was determined as described in Figure 1. (B) Data presented in Figure 2a were used to calculate Area Under the Curve (AUC) values. HS and HS-EVs dose-dependently inhibited TLR2/6 activity. Hash-signs indicate a significant difference (##P<0.01) compared to medium. EV depletion dose dependently rescued TLR2/6 activity. Asterisks indicate a significant difference (**P<0.01) compared to HS-D.
Figure 3.
EVs differentially affect synthetic ligand-induced TLR2 activation.
THP-1 reporter cells were incubated with synthetic TLR2 ligands addressing different TLR2 heterodimers in serum free medium, or medium supplemented with 5% of the indicated serum fractions (HS = intact human serum, HS-D = EV depleted human serum, HS-EV = human serum EVs in medium). (A, C, E, G) Dose response experiments indicating the effects of respectively Pam3CSK4, Pam2CSK4, FSL-1 or LTA stimulation on THP-1 activation. NFκB activity measured as OD values was determined using an alkaline phosphatase reporting system as described in materials and methods. (B, D, F, H) Data presented in Figure 3A,C,E,G were used to calculate Area Under the Curve (AUC) values. (B) Pam3CSK4 induced TLR2/1 activity was increased in medium supplemented with serum fractions compared to serum-free medium, depletion of EVs reduced TLR2/1 activation compared to HS ($$P<0.01) and HS-D ($P<0.05). (D) No modulatory effect by serum fractions were observed on Pam2SK4 stimulation. (E,F) HS and HS-EVs significantly inhibited FSL-1 and LTA induced TLR2/6 activity compared to medium (***P<0.001, **P<0.01, *P<0.05).
Figure 4.
EV-TLR2 expression is not required for the suppressive effect of EVs.
TLR2/6 expressing HEK transfectants were stimulated with B. breve NutRes 200 in a ratio of 1∶15 for 16H. (A) Dose response experiment indicating the effects of the different mouse serum fractions on TLR2/6 activity (WT = wild type, WT-D = EV depleted wild type serum, KO = TLR2 deficient mouse serum, KO-D = EVs depleted TLR2 deficient mouse serum in medium. NFκB activity was determined as described in Figure 1. Relative activity was determined calculating the ratios between medium activity and the activity in the different serum fractions. (B) Data presented in Figure 4a were used to calculate Area Under the Curve (AUC) values. EV depletion of both WT and KO sera significantly increased TLR2/6 activity (***P<0.001).
Figure 5.
EV mediated bacterial aggregation.
Picture A and C respectively represent a typical example of L. rhamnosus NutRes 1 and B. breve Nutres 200 co-incubated with medium, medium with ExoQuick or medium depleted for EVs. Picture B and D respectively represent a typical example of L. rhamnosus NutRes 1 and B. breve Nutres 200 co-incubated with HS or HS-EVs.
Figure 6.
EVs induce bacterial aggregation.
B. breve NutRes 200 and L. rhamnosus NutRes 1 were seeded in flat-bottom 96-wells plates at 2.5×106 bacteria/well in medium, medium with ExoQuick (EQ)(control), HS, HS-D or HS-EV and incubated at 37°C. After 16H cultures were analyzed for aggregation using microscopy and pictures were taken. Pictures were digitally processed and analyzed using ImageJ software, calculating the area covered by objects. Intact serum as well as HS-EVs induced bacterial aggregation, reducing the area covered by objects by approximately 50%. Aggregation was not observed upon EV depletion. No differences between strains could be observed.
Figure 7.
EVs enhance DC phagocytosis of bacteria.
1×105 DCs were co-incubated with 2.5×106 pHrodo-labeled B. breve NutRes 200, L. rhamnosus NutRes 1 or pHrodo Red Dextran as a control for DC phagocytic capacity at 37°C. (A) Medium was supplemented with increasing concentrations of HS, HS-D or HS-D reconstituted with the serum EV fraction (HS-recon). After 3H, DCs were collected and analyzed for fluorescence using flow cytometery. Degree of phagocytosis is determined as an increase in mean fluorescence index (MFI). EV depletion significantly reduced B. breve phagocytosis at 1 and 5% supplementation compared to HS (*P<0.05). Phagocytosis of L. rhamnosus was reduced to background level (dotted line) upon EV depletion at 1%, 5% and 10% supplementation compared to HS (**P<0.01, *P<0.05, *P<0.05, respectively). Reconstitution of HS-D with EVs normalized the phagocytic response compared to HS. Data are represented as mean ± SEM n = 2. Experiment was repeated at least twice with similar results. (B) Medium was supplemented with 5% of the indicated serum fractions and 30 µg/ml pHrodo Red Dextran. After 1 hour DCs were harvested and MFI determined. Medium as well as all serum fractions increased dextran uptake compared to DCs alone (DC). No differences in dextran uptake could be measured between medium and the different serum fractions. Data are represented as mean ± SEM n = 4.
Figure 8.
B. breve but not L. rhamnosus ligate DC-expressed TLR2.
1×105 DCs were co-incubated with 1×106 B. breve NutRes 200 or L. rhamnosus NutRes 1 at 37°C in medium, HS, HS-D or HS-EVs. TLR2 activity was inhibited by preincubating cells with a specific anti TLR2 antibody or isotype control. After 16H, supernatants were collected and analyzed for IL-6 and TNFα release. Relative cytokine levels were calculated according to the ratio between responses at serum-free medium and serum fraction supplemented medium. (A) Blocking TLR2 significantly inhibited DC IL-6 release after ligation by B. breve but not L. rhamnosus irrespective of the serum fractions. (B) Blocking TLR2 significantly inhibited DC TNFα release after ligation of B. breve in the presence of medium and HS-D but not intact HS or HS-EVs. DC TNFα release in response to ligation of L. rhamnosus was not affected. Data are represented as mean ± SEM n = 4 (***P<0.001)(**P<0.01)(*P<0.05).
Figure 9.
EVs differentially modulate bacterial induced DC cytokine release.
2×105 DCs were co-incubated with 2×106 B. breve NutRes 200 or L. rhamnosus NutRes 1 at 37°C in medium, HS, HS-D or HS-EVs. After 16H, supernatants were collected and analyzed for IL-6 and TNFα release. In another set of similar experiments, DCs were first pretreated with 10 µg/ml cytochalasine D, blocking bacterial phagocytosis. Relative cytokine levels were calculated according to the ratio between responses at serum-free medium and serum fraction supplemented medium. (A) HS and HS-EVs significantly inhibit B. Breve NutRes 200 induced DC IL-6 release compared to HS-D (***P<0.001)(**P<0.01). TNFα release was not affected but upon blocking phagocytosis a significant different TNFα release between HS, HS-EVs and HS-D could be measured (B) (*P<0.05). L. rhamnosus NutRes 1 stimulated DCs release significantly more IL-6 (C) and TNFα (D) in the presence of HS or HS-EVs compared to HS-D. Blocking L. rhamnosus NutRes 1 phagocytosis inhibited DC IL-6 and TNFα release below detection level (ND). Data are represented as mean ± SEM n = 4.