Peripheral blood samples were collected by jugular venipuncture from Holstein cows in lactation (n = 12) bred in the Unité Expérimentale de Physiologie Animale (UEPAO INRAE, France). No diseases (infectious, metabolic, or reproductive) were recorded within 30 days prior to sampling for all the selected animals. For antigen presentation assays, animals showing memory T-cell response against Staphylococcus aureus were selected. Blood samples were collected into 10 ml vacuum tubes containing ethylenediaminetetraacetic acid (EDTA) (Vacutainer K2-EDTA, BD) and kept at room temperature until processing (<1 h). Animal handling and blood sampling were conducted with the approval of Ethics Committee of Val de Loire (France, DGRIs agreement APAFIS#29498-2021020410061759 v2) in strict accordance with all the applicable provisions established by the European directive 2010/63/UE.

Blood samples were centrifuged at 1,000 g for 30 min at 20°C with gentle deceleration. Buffy coat cells were collected, resuspended with ACK Lysing Buffer (Thermo Fisher Scientific), and incubated for 5 min at room temperature for the lysis of residual erythrocytes. Afterwards, the cells were washed twice with wash buffer [Dulbecco's phosphate-buffered saline (DPBS) without Ca²⁺ and Mg²⁺ supplemented with 2% (v/v) fetal bovine serum (FBS) and 2 mM EDTA], resuspended in FACS buffer (DPBS without Ca²⁺ and Mg²⁺ supplemented with 2% (v/v) normal goat serum and 2 mM EDTA), and counted using the LUNA FL Automated Cell Counter (Logos).

Simplified protocol 1—Buffy coat cells were resuspended in moDC medium (RPMI 10% FBS, 2 mM glutamine, 1 mM sodium pyruvate, 100 nM nonessential amino acids, and 50 μM β-mercaptoethanol) supplemented with premium grade human recombinant GM-CSF and IL-4 (20 ng/ml of each, Miltenyi Biotec) at a concentration of 3 × 10⁶ cells/ml and seeded in polystyrene tissue culture gas plasma-treated 6 well-plates (Falcon) (3 ml/well). After 12–14 h of incubation at 38.5°C with 5% CO₂, non-adherent cells were discarded, adherent cells were gently washed twice with warm DPBS without Ca²⁺ and Mg²⁺, and 3 ml of warm moDC medium were added per well. At day 3 of culture, 50% of medium was replaced with fresh moDC medium containing 40 ng/ml of GM-CSF and IL-4. At day 6 of culture, non-adherent and adherent moDCs were harvested after treatment with 0.05% (m/v) trypsin-EDTA (Gibco).

Simplified protocol 2—Buffy coat cells were used for the separation of PBMC using density gradient centrifugation as previously described (14). PBMC was resuspended in moDC medium supplemented with 20 ng/ml of human recombinant GM-CSF and IL-4 (Miltenyi Biotec) at a concentration of 3 x 10⁶ cells/ml and treated as described in simplified protocol 1.

Standard protocol—CD14⁺ cells were isolated from 3 × 10⁷ PBMC by magnetic sorting using CD14 microbeads (human, Miltenyi Biotec) according to manufacturer's conditions. Afterwards, 3 × 10⁶ cells were resuspended in 3 ml of moDC medium and seeded in tissue culture polystyrene gas plasma-treated 6 well-plates (Falcon). After 3 days, 50% of medium was replaced with moDC medium containing 40 ng/ml of GM-CSF and IL-4 and at day 6, non-adherent and adherent moDCs were harvested after treatment with 0.05% trypsin-EDTA (Gibco). A diagram illustrating the isolation of blood cells and the protocols for the production of moDC is given in Figure 1.

Serum-free media—Buffy coat cells were processed as described in simplified protocol 1 using AIM V Serum-free Medium (Gibco) supplemented with AlbuMAX (Gibco) or X-VIVO 15 serum-free hematopoietic cell medium (Lonza).

Cells were harvested by treatment with 0.05% trypsin-EDTA or by gentle scraping in cold PBS containing 2 mM EDTA (for CD80 and CD86 labeling) and washed twice with FACS buffer. For the labeling of surface markers, the following mouse primary antibodies were used: anti-MHCII (CAT82A, IgG1, Kingfisher), anti-CD11c (BAQ153A, IgM, Biorad), anti-CD209 (209MD26A, IgG2a, Kingfisher), anti-CD11b (hybridome IAH CC104, IgG2b, kindly provided by Dirk Werling), anti-CD172a (CC149 conjugated to RPE-Cy5, Biorad), anti-CD14 (TUK4 conjugated to Alexa Fluor 647), anti-CD80 (IL-A159 conjugated to FITC), and anti-CD86 (IL-A190 conjugated to RPE). The following antimouse secondary antibodies were used when necessary: anti-IgG1 conjugated to Alexa 488 (A211121, Invitrogen), anti-IgM conjugated to PE-Cy7 (406513, Biolegend), anti-IgG2a conjugated to RPE (115-115-206, Jackson ImmunoResearch), and anti-IgG2b conjugated to BV510 (743175, BD Biosciences). After the antibody incubations, dead cells were labeled using the Fixable Viability Dye eFluor 450 (eBioscience). Cells were examined by flow cytometry using a BD LSR Fortessa cytometer and data were analyzed with the Kaluza software (Beckman Coulter). Dead cells were excluded from the analysis and gates were set according to appropriate isotype/control staining (Supplementary Figures 1, 3).

Monocyte-derived DCs harvested at day 6 were treated with LPS at 50 ng/ml (Sigma-Aldrich) or a cytokine cocktail composed of recombinant human IL-6 (Peprotech), recombinant human interleukin-1β (IL-1β) (Peprotech), and recombinant bovine interferon-γ (IFN-γ) (Biorad), each of them at 50 ng/ml, in moDC medium for 14 h at 38.5°C with 5% CO₂. Treated cells were gently harvested with rubber scrapers in cold PBS with 2 mM EDTA. For stimulations with bacteria, moDC were harvested at day 9 and infected with a S. aureus strain (#1403) isolated from a bovine clinical mastitis at a multiplicity of infection (MOI) of 50 in moDC medium for 2 h under standard conditions. Afterwards, the cells were washed with DPBS and processed for reverse transcription-quantitative PCR (RT-qPCR) as follows.

Cells were lysed in RA1 buffer supplemented with 1% (v/v) β-mercaptoethanol (Macherey-Nagel) and stored at −80°C until processing. Total RNA was extracted by using the NucleoSpin RNA Extraction Kit (Macherey-Nagel) and the residual genomic DNA was removed by DNase digestion with RNase-free DNase (Macherey-Nagel). The total RNA quantifications were carried out using a NanoDrop spectrophotometer (NanoDrop Technologies). For cDNA synthesis, total RNA (100 ng) was reverse transcribed using the 5X iScript reverse transcription supermix (Biorad) according to the manufacturer's instructions. cDNA samples were stored at −20°C until use. Reverse transcription-quantitative PCR assays were performed in a LightCycler 480 instrument (Roche). Four microliter of 10-fold diluted cDNA were added to a mixture of (2X) iTaq Universal SYBR Green Supermix (Biorad) and 0.25 μM of each primer in a total volume of 10 μl. Thermal protocol was 95°C for 5 min followed by 40 cycles of 95°C for 10 s, 60°C for 30 s, and acquisition of a melting curve at the end of the run. The specificity of primer pairs was checked via melting curve analysis. Primers used in this study (Supplementary Table 1) were designed and tested as previously described (14). Fold changes were calculated by the ΔΔCt method using ACTB, PPIA, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as reference genes (15).

Monocyte-derived DCs harvested at day 6 were resuspended in moDC medium at a concentration of 2 x 10⁶ cells/ml and seeded in a tissue culture polystyrene vacuum gas plasma-treated 48 well-plates (Falcon) (100 μl/well). Cells were allowed to settle and adhere to the plate for 1 h at 38.5°C in a humidified incubator with 5% CO₂. Afterwards, 100 μl of warm moDC medium supplemented or not with 1 μg of pHrodo Red Escherichia coli BioParticles Conjugate (Life Technologies) were added to treated and mock wells, respectively. After 1.5 h incubation at 38.5°C with 5% CO₂, moDCs was collected with 0.05% trypsin-EDTA, washed twice with FACS buffer, and analyzed by flow cytometry. The phagocytic activity of cells was estimated based on the acidification of ingested particles, as the pHrodo dye increases in fluorescence in acidic pH conditions. The rate of phagocytosis was deduced from the percentage of cells showing high fluorescence emission at 590 nm, taking moDCs and pHrodo beads alone as reference.

Antigen pulsing—MoDCs harvested at day 6 were seeded in 48 well-plates (10⁵ cells/well in 500 μl of moDC medium) and incubated for 18 h at 38.5°C with 5% CO₂. Then, cells were pulsed with a heat-inactivated (30 min at 70°C in PBS) S. aureus strain (#1403) isolated from a bovine clinical mastitis at a MOI of 50 for 2 h.

T-cell isolation and CellTrace labeling—At day 7, PBMC were isolated from autologous blood samples as described in the simplified protocol 2 and used for the separation of T cells by magnetic sorting. Briefly, 3 × 10⁷ PBMC were incubated with 3 μg of an anti-bovine CD3⁺ antibody (MM1A, Biorad) in 300 μl of FACS buffer for 30 min at 4°C. After 1 wash with FACS buffer, cells were resuspended in 240 μl of FACS buffer and 60 μl of antimouse immunoglobulin G (IgG) microbeads (Miltenyi Biotec), incubated for 30 min at 4°C, and washed with FACS buffer. CD3⁺ cells were sorted using MS columns (Miltenyi Biotec) and an octoMACS separator (Miltenyi Biotec) according to the manufacturer's conditions. T cells were subsequently labeled with 2.5 μM of CellTrace Violet (Invitrogen) using the manufacturer's protocol for cells in suspension and resuspended in X-VIVO medium at 5 × 10⁵ cells/ml.

Co-culture and analysis by flow cytometry—After antigen pulsing, moDC medium was aspirated and replaced with 500 μl/well of the labeled T-cell suspension X-VIVO. After 3 days of co-culture (day 10), T cells were harvested from the supernatant and washed with FACS buffer before and after the staining of dead cells with the Fixable Viability Dye eFluor 660 (eBioscience). The percentage of T-lymphocytes proliferation was analyzed by flow cytometry based on the loss of CellTrace staining upon cell division using a BD LSR Fortessa cytometer. Data were analyzed with the Kaluza software (Beckman Coulter). Dead cells were excluded from the analysis and gates were set according to negative (lymphocytes alone) and positive (lymphocytes alone under polyclonal stimulation; 5 μg/ml of anti-bovine CD3 MM1A and 1 μg/ml of anti-bovine CD28 CC220 antibodies, Biorad) controls (Supplementary Figure 2). The levels of antigen presentation were estimated as a ratio between the percentage of T-cell proliferation upon co-culture with infected and control moDCs. A diagram illustrating each step of this assay is shown in Figure 2.

Frozen buffy coat cells were resuspended in moDC medium supplemented with 20 ng/ml of GM-CSF and IL-4 and seeded in tissue culture polystyrene gas plasma-treated 6 well-plates (10⁷ cells in 3 ml/well). After 3 days at 38.5°C with 5% CO₂, non-adherent cells were collected, centrifuged, and resuspended in Xvivo medium supplemented with recombinant human IL-2 (Peprotech, 10 ng/ml) at 10⁶ cells/ml and seeded in 12 well-plates. Adherent cells were washed three times with warm DPBS and 3 ml of moDC medium supplemented with 20 ng/ml of GM-CSF and IL-4 were added per well. At day 5, 50% of medium from adherent and non-adherent cell cultures were replaced with fresh medium containing the doubled concentration of the corresponding cytokines. From day 6 to day 10, moDC seeding and pulsing, as well as T-cell isolation and labeling were performed as described in the protocol for fresh cells keeping the same proportion of cells and reagents, taking into consideration the cell yield. A diagram illustrating each step of this assay is given in Figure 3.

Data were analyzed and plotted using GraphPad Prism, version 6.0 (GraphPad Software Incorporation). The Kruskal–Wallis test was used to compare the independent groups and pairwise comparisons were carried out using the Dunn's test. Data shown represent the mean and the SD from at least three individual replicates (different cows).

Monocyte-derived dendritic cells obtained by using two simplified methods were evaluated for their phenotype and maturation levels. A standard protocol based on magnetically-sorted CD14⁺ cells, as previously described for bovine samples (11), was used as a reference control. First, microscopic analyses showed no morphological differences among moDC produced using any of the three tested approaches (Figures 1, 4A). No significant differences were observed neither in the total number nor in the viability of the obtained cells (Figures 4B,C). However, as shown in Figure 4D, the simplified protocols 1 and 2 showed a higher yield when the number of derived moDC was estimated taking into consideration the number of precursor cells (buffy coat cells, PBMC, and sorted CD14⁺ cells, respectively). The analysis of moDC surface markers at day 6 by flow cytometry uncovered similar levels of CD11b-, MHC-II-, and CD11c-positive cells produced using both the simplified 1 and standard protocols. Nevertheless, a lower percentage of cells positive for these markers was observed when simplified 2 and standard protocols were compared (Figure 4E). Of note, intermediate levels of CD11b expression were found out for moDC produced by employing any of the simplified protocols (Figure 4E, upper histograms).

The evaluation of CD80 and CD86 expression, two dendritic cells maturation markers, in immature moDCs at day 6, revealed a higher proportion of positive cells when the standard protocol was employed (Figures 5A,B). Then, the maturation of these cells was induced upon treatment with LPS or a cocktail of cytokines containing IL-6, IL-1β, and IFN-γ. As shown in Figure 5C, treatments induced an increase of CD80 levels in moDCs produced by employing any of the tested protocols. However, no increase of CD86 levels was observed for any of the tested groups upon cells stimulation.

We next complemented our analysis by evaluating the expression of CD14, CD209, and CD172a after 9 days of culture. In all the groups, about 90% of CD11c⁺ MHC-II⁺ cells co-expressed CD11b and CD14. Of note, the percentage of CD11c⁺ MHC-II⁺ cells co-expressing CD172a and CD209 was significantly lower in moDCs produced using the standard protocol (Figures 6A,B). Additionally, the maturation of moDCs upon an infection with S. aureus was verified by qPCR. As shown in Figure 6C, the exposure to bacteria induced the upregulation of CD80, CD83, CD86, and CD205 in all the moDC groups. Curiously, CD209 was less overexpressed in cells produced using the standard protocol.

In order to compare the functional capacities of moDCs obtained by the tested approaches, we initially evaluated their capacity to uptake antigens. For this, we took advantage of the pHrodo-conjugated particles technology. Because pHrodo dye has its fluorescence increased under acidic conditions, the rate of phagocytosis was measured by flow cytometry upon the pulse of moDC with pHrodo Red Escherichia coli BioParticles Conjugate (Figures 7A,B). As shown in Figure 7C, moDCs produced by using any of the evaluated protocols proved to have a high antigen uptake capacity. Indeed, about 80% of cells were able to phagocyte particle conjugates after a 1.5-h exposure and no significant differences were observed between the tested groups.

We also tested the capacity of moDC to present antigens by carrying out an assay given in Figure 2. Briefly, moDC was co-cultured with autologous T cells labeled to trace cell proliferation by flow cytometry. The percentage of T-cell proliferation in co-cultures with mock moDC and S. aureus-pulsed moDC was deduced based on the loss of cell trace staining upon cell divisions (Figure 8A). Then, the levels of antigen presentation were estimated as a ratio between the percentage of T-cell proliferation in co-cultures with infected and control moDC. As shown in Figure 8B, moDC produced using the simplified protocol 1 showed a better capacity to present bacterial antigens.

Based on the results of our initial screening, we next set out to adapt the simplified protocol 1 to study the antigen presentation using a single sample of frozen bovine blood cells. For this, adherent and non-adherent cells were separately kept in culture for the production of moDCs and isolation of CD3⁺ T cells (Figure 3). This approach enabled us to obtain from 10⁷ frozen buffy coat cells about 1.8 and 0.5 x 10⁶ moDC and T cells, respectively (Figure 9A). In both the cases, the viability levels observed were of about 90% (Figure 9B). These cells were used in the antigen presentation assay previously described and as shown in Figure 9C, the results obtained with frozen cells were comparable to the ones observed using fresh T cells. However, a higher variation of the S. aureus-specific T-cell response was observed using frozen cells (2.8 ± 1.9; mean ± SD; n = 3).

Since the nonspecific activation of in-vitro-derived dendritic cells has been associated to the presence of FBS in the medium (16), we tested if the simplified protocol 1 could be used for the production of bovine moDCs under serum-free conditions. For this, cells from the same animals were processed using the serum-free media AIM V and X-VIVO 15 or RPMI supplemented with FBS (moDC medium) as a control. No differences were observed in the viability nor in the total number of cells obtained using any of the tested media (Figures 10A,B). Nevertheless, the phenotype of moDC produced in serum-free or FBS-supplemented media was different. As given in Figure 10C (left), less cells showing typical moDC size and internal complexity were observed in serum-free media. Additionally, moDC produced in absence of serum exhibited a lower expression of all the surface markers analyzed (Figure 10C middle and right). In fact, the percentage of cells co-expressing MHC-II, CD11c, CD172, and CD209 was significantly lower in both the serum-free media when compared to RPMI-FBS (Figure 10D).

Subsequently, the functional activity of these cells was evaluated by using the antigen presentation assay given in Figure 2. As expected, CD3⁺ lymphocytes co-cultured with S. aureus-pulsed moDC proliferated at higher levels when compared to mock moDC, independently of the used medium (Figures 11A,B). However, higher rates of bacterial antigen presentation were observed in the FBS-supplemented medium.