Extracellular milieu grossly alters pathogen-specific immune response of mammary epithelial cells

Abstract

Background: Considerably divergent data have been published from attempts to model the E. coli vs. S. aureus specific immune reaction of the udder using primary cultures of bovine mammary epithelial cells from cows (pbMEC). Some groups reported a swift, strong and transient inflammatory response against challenges with E. coli and only a weak and retarded response against S. aureus, in agreement with the respective reaction of the udder. Others found almost the reverse. Presence or absence of fetal calf serum distinguished the experimental setting between both groups. We examined here if this causes the divergent reaction of the pbMEC towards both pathogen species. We challenged pbMEC with proteins from heat killed E. coli or S. aureus pathogens or purified TLR2 and TLR4 ligands. The stimuli were applied in normal growth medium with (SM10) or without (SM0) 10% fetal calf serum, or in the basal medium supplemented with 10 mg/ml milk proteins (SM Milk).

Results: Withdrawal of FCS slowed down and decreased the extent by which E. coli or LPS enhanced the expression of cyto- and chemokine encoding genes through impaired TLR4 signalling but enforced their expression during stimulation with S. aureus. SM Milk strongly quenched the induction of those genes. S. aureus strain specific differences in the reaction of the pbMEC could only be recorded in SM0. NF-κB factors were activated by E. coli in all stimulation media, but only to a small extent by S. aureus, solely in SM0. Purified ligands for TLR2 stimulated expression of those genes and activated NF-κB equally well in SM10 and SM0. The mRNA destabilizing factor tristetraproline was only induced by E. coli in SM10 and by purified PAMPs.

Conclusions: Our data cross validate the correctness of previously published divergent data on the pathogen-specific induction of key immune genes in pbMEC. The differences are due to the presence of FCS, modulating signalling through TLR4 and TLR-unrelated pathogen receptors. S. aureus does not substantially activate any TLR signalling in MEC. Rather, receptors distinct from TLRs perceive the presence of S. aureus and control the immune response against this pathogen in MEC.

Fig. 1

Composition of the challenge medium influences extent and kinetic of pathogen-specific induction of TNF and NOS2A expression. a Influence of the different stimulation media upon extent and kinetics of the pathogen specific induction of TNF and NOS2A expression in pbMEC. a) Effect of absence (SM0) or presence (SM10) of 10 % FCS in the challenge medium to which 30 μg/ml of heat killed E.coli ₁₃₀₃ or S. aureus ₁₀₂₇ had been added. The mRNA concentration (ordinate) was measured at different times after challenge (abscissa). b) Same as a), but the challenge medium was supplemented with 10 mg/ml of total milk protein (SM milk) and S. aureus strain N305 was also included. Note that the addition milk increased the base line expression (control). c) Same as a), but the challenge was done with two S. aureus strains. Values are means and SEM. (error bars) from two independent biological replica, each sampled in duplicate. b Data are from the same experiments as shown in (a), but the EXPANDER software was used to display the data for several genes. Each line displays the fold changes (compared to the unstimulated control) in the expression of the respective gene as indicated over the time of the challenge (1, 3, 24 [h]), normalized across all conditions to the average of 0 and variance 1. Data have been taken from Additional file 2: Table S1A. Group A, immediate early genes, Group B, secondary immune response genes. c Same as (b), but the analysis included only the data from the two different S. aureus strains (indicated). Data have been taken from Additional file 2: Table S1B *, p < 0.05 vs. unstimulated control (t-test, with Bonferroni’s correction)

Fig. 2

Influence of medium composition on pathogen or PAMP mediated NF-κB activation in pbMEC. a E. coli, but not S. aureus activates NF-κB in all media. pbMEC were transiently transfected with the NF-κB driven luciferase reporter gene and subsequently challenged for 24 h with 30 μg/ of heat killed pathogens in the media as indicated and the luciferase activity was measured from lysates (ordinate). Mean values (± S.E.M.) from five experiments (four biological replica) for SM0 and SM10, and two biological replica experiments for SM Milk. Each experiment was assayed in triplicate (*, p < 0.05; ***, p < 0.001 vs. control). b Only the E. coli challenge degrades IκBα. Degradation of IκBα was visualized in Western-blots after challenging the pbMEC in SM0 or SM10 with 30 μg/ml of heat killed E. coli or S. aureus ₁₀₂₇ particles. Thirty μg of cell lysates having been collected from unstimulated control cultures (c) or at the times as indicated after the challenge had been resolved in any slot of a 12 % SDS-gel and blotted onto nitrocellulose. The primary antibody was specific for IκBα. Data are representative for two independent experiments

Fig. 3

Dissection of the influence of FCS upon TLR2 or TLR4 mediated pathogen recognition. a TLR2, but not TLR4 signalling activates immune gene expression irrespective of FCS presence. pbMEC were challenges in SM10 or SM0 with synthetic lipopetides (Pam2CSK4, 10 ng/ml; Pam3CSK4, 100 ng/ml) as TLR2 ligands or purified LPS (10 ng/ml) as TLR4 ligand. Same experimental setting as in Fig. 1. The mRNA concentration (ordinate) of TNF or NOS2A was measured at different times after challenge (abscissa). Data are mean values and SEM from two biological replica experiments, each assayed in duplicate; *, p < 0.05 vs. unstimulated control (t-test, with Bonferroni’s correction). b Purified PAMPS activate NF-κB in pbMEC irrespective of FCS supplementation. pbMEC were transfected with the NF-κB reporter expressing vector (same experimental setting as in Fig. 2a) and challenged for 24 h with purified LPS or Pam2CSK4 (10 ng/ml each) or 100 ng/ml of Pam3CSK4. Mean values of increased luciferase activity over that of unstimulated controls (ordinate) from two (LPS, Pam2CSK4) independent experiments, each assayed in triplicate. Different superscript letters indicate significant (t-test, p < 0.05) difference from each other and from the unstimulated control. c Lysates from shattered S. aureus do not strongly activate NF-κB in pbMEC. Same as b, but comparing in pbMEC the NF-κB activation through heat killed bacteria (particles) with that caused by supernatants (lysates) from RiboLyser crushed heat killed pathogens from S. aureus strains N305 and RF122 to that of E. coli. Protein concentrations applied for the respective challenge are given below the columns. Each condition was assayed in triplicate. Superscript letters (a, b) indicate significant (p < 0.05) difference between the column means and from the unstimulated controls

Fig. 4

Modulation of TTP expression through FCS. a TTP expression is only activated by E. coli particles in SM10. pbMEC were stimulated with 30 μg/ml of heat killed pathogens media with or without 10 % FCS (SM10 or SM0), for the time as indicated. The fold increase in the TTP mRNA concentration above the control value is indicated. Mean values ± SEM from two biological replica experiments, each assayed in duplicate. Only values from E. coli stimulation at 1 h and 3 h in SM10 resulted in significantly increased mean values (p < 0.001). b Addition of milk prevents induction of TTP expression. Same experimental setting as above, but stimulation was in SM Milk. No significant induction of TTP mRNA concentration, but a significant (p < 0.05) decrease compared to the unstimulated control at t 1 h. c Pam2CSK4, but not LPS triggers TTP expression irrespective of the presence of FCS. Same experimental setting as in A) but stimulation was with 10 ng/ml of either Pam2CSK4 or highly purified LPS. All values, but that for LPS at 1 h after the challenge differed significantly (p < 0.01) from the control value