Investigations were conducted on milk samples collected from 14 clinically healthy dromedary she-camels during their first 2 months of lactation. The animals were reared on a private camel farm in Al-Ahsa region in eastern Saudi Arabia. All camels were from Al-majaheem camel breed and in their third and fourth lactations. All animal procedures were approved by the Ethics Committee of King Faisal University (Approval No. KFU-REC-2021- DEC -EA000326).

Milk samples were collected from all mammary gland quarters during the evening milking. Before sampling, the udder was palpated and checked for signs of clinical mastitis, such as heat, swelling and pain in the infected quarters, and abnormal alteration in milk color and consistency (41). After discarding the first milk jets, teat ends were cleaned and disinfected and about 10 ml milk were collected into sterile glass tubes for the California Mastitis Test (CMT), microbiological analysis, and SCC. Another 100 ml of milk were milked into sterile glass bottles for cell separation and flow cytometry. The health status of the animals was classified as healthy or mastitic based on the results of the CMT (41, 42). For this, 3 ml of each quarter milk were added to an equal amount of CMT fluid and the mixture was rotated by circular movement. The reactions were graded according to the Scandinavian scoring system as previously described (41). A score of 1 was given if there was no visible thickening of the mixture; score 2 represented slight slime which tends to disappear with continued swirling; score 3 indicated distinct slime but without gel formation; score 4 represented the immediate formation of gel that moves as a mass during swirling; score 5 was given if the gel developed a convex surface and adhered to the bottom of the paddle. Animals with milk samples of a test score equal to or more than 3 in the absence of clinical signs of mastitis were classified as subclinical mastitis animals. For healthy she-camels (n = 8 animals) with a test score of <3 and no clinical signs of mastitis, pooled composite milk samples representing all four quarters were prepared for flow cytometry. In the affected group (n = 6 animals), only milk samples collected from the affected quarters were further processed for SCC and flow cytometry. Collected milk samples were kept in a cool box and were further processed in the lab within 4 h from the time of collection.

Milk SCC was performed after fat globule removal by the spin-wash method (43). Milk samples (500 μl) were diluted with 500 μl PBS in a 1.5 ml tube and the diluted samples were centrifuged at 1,000 × g for 2 min. The upper cream layer was removed using a cotton swab and the remaining skim layer was poured off. For the second wash, 1 ml PBS was added to the tube without resuspending the pellet. The washing step was repeated twice. After the final wash, the cell pellet was resuspended in 500 μl PBS by gently pipetting up and down. The washed cell suspension (100 μl) was stained to an equal volume of Turk solution, which stains the cell nuclei blue, and the SCC was performed using Neubauer counter and light microscopy (44).

For bacteriological analysis, 10 μl of milk were streaked on blood agar and MacConkey agar plates, and were incubated for 24-48 h at 37°C. The plates were then examined for growth colony morphology. Individual colonies were picked for microscopic identification using Gram staining (45). Briefly, thin smears were prepared from the plate cultures, allowed to air dry, and then fixed with heat. Smears were covered with crystal violet solution for 1 min followed by gentle rinsing with water. After that, the smears were covered with Gram iodine solution for 1 min followed by rinsing with water. After that, decolorizer solution was added to the smears for 20 s. Finally, counter-staining with safranin solution was performed for 1 min followed by rinsing with water. The smears were examined microscopically at 1,000 × magnification with oil immersion. The bacterial species were identified based on the shape, arrangement and gram reaction of the organisms as previously described (46).

Eight milk samples collected from eight healthy animals (each representing four quarter milk samples) and six milk samples collected from affected quarters of six affected animals were used for cell separation and flow cytometry. Separation of milk cells was performed according to a method previously described for caprine milk (47) with some modifications. Briefly, milk samples were diluted with cold PBS (25 ml milk and 25 ml PBS) in conical 50 ml polypropylene tubes and the tubes were centrifuged at 800 × g and 4°C for 20 min without brake. After removing the fat layer using a spatula, the supernatant was discarded. The cell pellet was resuspended with 30 ml cold PBS and washed twice at 600 × g and 4°C for 10 min. For parallel staining of blood leukocytes, leukocytes were separated from one EDTA blood sample collected from a healthy dromedary camel as previously described (48). Concisely, 1 ml blood was incubated with 5 ml distilled water for 20 s followed by the addition of 2 × PBS to restore tonicity. After centrifugation (1,000 × g, 15 min, 10°C, without brake), the supernatant was discarded and the washing step was repeated twice (at 500 × g and 250 × g for 10 min) until complete removal of red blood cells. Blood and milk cell pellets were suspended in 1 ml cell staining buffer (PBS containing 5 g/l BSA, 100 mg/l NaN3) at concentrations of 5 × 10⁶ cells/ml. Cell viability was determined after incubating the cells with the nucleic acid stain propidium iodide (PI; 2 μg/ml, BD Biosciences, Germany).

The monoclonal and polyclonal antibodies used are shown in Table 1 (44, 48, 49). Leukocytes were identified using a monoclonal mouse antibody against lama pan-leukocyte marker CD45 (clone LT12A). A monoclonal antibody to the myeloid cell marker signal-regulatory protein alpha (SIRPα), also called CD172a (clone DH59b), was used to differentiate myeloid cells from lymphoid cells. Macrophages and granulocytes were differentiated based on their differential expression of the lipopolysaccharide receptor CD14, which was detected using a mouse anti CD14 monoclonal antibody (clone Tuk4). The antibodies used for helper T cells and gamma delta T cells were mouse anti bovine CD4 (clone GC50A1), and WC1 (clone BAQ128A) antibodies. The monoclonal antibodies to major histocompatibility complex (MHC) class II (clone TH81A5) and the scavenger receptor CD163 (clone LND68A) were used for the analysis of macrophages phenotype. Furthermore, the monoclonal antibodies to CD18 (clone 6.7), CD11a (clone HUH73A), and CD44 (LT41A) were used to measure the expression of selected cell adhesion molecules. All antibodies were tested for reactivity against camel leukocytes in previous studies (35, 50–53). Secondary antibodies to mouse primary antibodies were goat anti-mouse IgM conjugated with Allophycocyanin (APC), goat anti-mouse IgG1 conjugated with fluorescein isothiocyanate (FITC), and goat anti-mouse IgG2a conjugated with phycoerythrin (PE).

Cell labeling was performed in a three steps staining procedure in a round-bottomed 96-well microtiter plate as previously described (47). For this, 100 μl of the leukocyte suspension (5 × 10⁵ leukocytes per well) of each sample were pipetted into the well of a microtiter plate. After short centrifugation of the plate for 3 min at 300 × g and 4°C, the supernatant was discarded and the primary antibodies (Table 1) diluted in staining buffer were added to the wells. After 15 min of incubation at 4°C, the cell suspension was washed with staining buffer for 3 min at 300 × g and 4°C. In the second staining step, fluorochrome-conjugated secondary antibodies (Table 1) to mouse primary antibodies were added to the wells and the plate was incubated for an additional 15 min at 4°C in the dark. Subsequently, the cells were washed twice with staining buffer for 3 min at 300 × g and 4°C. In the final staining step, an anti-CD14 monoclonal antibody conjugated with APC was added to selected wells followed by a 15 min incubation at 4°C in the dark. Finally, the cells were washed twice with staining buffer for 3 min at 300 × g and 4°C and resuspended in 150 μl staining buffer for flow cytometry. Staining with only antibody isotype controls was included. Labeled cells were analyzed on a FACSCalibur (BD Biosciences) by the acquisition of at least 100 000 total cells. Collected flow cytometric data were analyzed using the FCS Express Software (V3; BD Biosciences).

Bacterial phagocytosis by milk phagocytes was performed using heat-killed staphylococcus aureus (S. aureus) bacteria (Pansorbin, Calbiochem, Merck, Nottingham, UK) labeled with fluoresceinisothiocyanate (FITC, Sigma-Aldrich, St. Louis, Missouri, USA) (54). Separated milk cells (2 × 10⁵ in 100 μl RPMI medium) were incubated with S. aureus-FITC (50 bacteria/cell) for 45 min at 37°C in 96 well plates. After incubation, the plate was washed in RPMI medium (300 × g for 3 min) and the cells were resuspended in 150 μl of staining buffer and analyzed by flow cytometry. After the identification of the phagocyte population including granulocytes and macrophages, the percentage of phagocytosis-positive cells, as well as their mean green fluorescence intensity (MFI), were calculated.

Data were processed with the Microsoft office Excel^® program (version 2016 Microsoft) and statistical analysis was performed using the software program Prism (GraphPad software version 5, GraphPad Software, San Diego, USA). The Kolmogorov-Smirnov test (with the Dallal-Wilkinson-Lilliefor P-value) was performed to check the normal distribution of data. For normal-distributed data, the unpaired student's t-test was used to compare the mean of the two groups. For the data that failed to pass the normality test, the non-parametric Mann-Whitney test was used to compare the means. The comparison between milk granulocytes, macrophages, and lymphocytes regarding their phenotype was performed using a one-factorial analysis of variance (ANOVA) with Bonferroni's Multiple Comparison Test. The results for each analyzed parameter were presented graphically as means ± standard error of the mean (SEM). Results were considered statistically significant if the p-value was <0.05.

Using monoclonal antibodies to the leukocyte antigens CD172a and CD14, camel granulocytes, macrophages, and lymphocytes were identified in camel milk samples by flow cytometry. To confirm the expression pattern of the cell surface molecules, blood leukocytes were separated from one animal and were stained with the same combination of monoclonal antibodies (Figure 1). Camel milk granulocytes were identified based on their positive staining with the myeloid marker CD172a and low staining with CD14 (CD172a⁺CD14^low), while milk macrophages expressed both markers in high levels (CD172a⁺CD14^high). On the other hand, milk lymphocytes were identified based on their negative staining with the myeloid marker CD172a and the monocytic marker CD14 (CD172a⁻CD14⁻).

Milk macrophages showed the highest abundance (p < 0.05) of the myeloid marker CD172a (SIRP-alpha), the lipopolysaccharide receptor CD14, the scavenger receptor CD163, the antigen presentation receptor major histocompatibility complex (MHC) II molecules, the cell adhesion molecules CD18 (integrin beta chain), and CD11a (the alpha chain of the lymphocyte function-associated antigen 1; LFA-1) when compared to granulocytes and lymphocytes (Figure 2A). While granulocytes displayed an intermediate level of CD172a, CD14, CD163, MHCII, CD18, and CD11a, lymphocytes showed the lowest abundance of all those molecules, in comparison to granulocytes and macrophages (Figure 2A).

The analysis of light scatter characteristics of milk leukocyte subpopulations revealed similar (p > 0.05) cell size (as measured by the forward scatter; FSC) and granularity (as measured by side scatter; SSC) for milk granulocytes and macrophages (FSC^highSSC^high), while milk lymphocytes were identified as smaller cells with lower granularity (FSC^lowSSC^low), in comparison to granulocytes and macrophages (p < 0.05) (Figure 2B).

Subclinical mastitis (SCM) was diagnosed based on the California Mastitis Test (CMT) score and the absence of signs of clinical mastitis. SCM (CMT score of ≥3 in the absence of clinical signs of mastitis) was found in six of the 56 sampled quarters (10.7%). All affected camels had only 1 quarter with SCM. The bacteriological analysis identified bacterial cultures in all milk samples collected from the six SCM quarters and only in one sample collected from healthy animals. While two milk samples yielded single bacterial cultures with staphylococcus or streptococcus species, the other five samples yielded mixed bacterial cultures with streptococcus and staphylococcus (three samples) or streptococcus and coliform bacteria (two samples).

The SCC, the fraction of milk leukocytes (CD45+ cells), and the differential leukocyte composition were compared between the healthy and SCM animals. The SCC was significantly (p = 0.001) higher in milk samples from SCM camels (418.3 × 10³ cell/ ml) than healthy animals (103.8 × 10³ cell/ml) (Figure 3A). Milk samples with SCM contained higher percentages (p = 0.001) of total leukocytes (90.3 ± 3.1% of total cells) than healthy (48.5 ± 9.3% of total cells) milk samples (Figure 3B). The fraction of myeloid cells within the leukocyte population was also significantly elevated (p = 0.01) in SCM milk samples (97.8 ± 0.5% of leukocytes) compared to healthy milk samples (93.5 ± 1.3% of leukocytes) (Figure 3C), while the fraction of lymphoid cells was significantly (p = 0.01) lower in SCM than in healthy milk samples (Figure 3D). Similarly, the fraction of milk granulocytes was significantly (p = 0.02) expanded in SCM samples (79.1 ± 2.7% of leukocytes vs. 67.8 ± 3.7% of leukocytes in healthy milk) (Figure 3E), while the fraction of macrophages was decreased (16.2 ± 3.4% of leukocytes vs. 24.7 ± 3.6% of leukocytes in healthy milk; p = 0.04) (Figure 3F).

The percentage of viable, PI-negative (Figure 4A) milk leukocytes was significantly (p = 0.01) higher in milk samples from SCM animals (91.9 ± 1.9% of CD45+ cells) than healthy animals (85.8 ± 1.3% of CD45+ cells). In milk from SCM animals, the myeloid cell population contained a higher (p < 0.05) percentage of viable cells (91.9 ± 1.7% vs. 77.0 ± 3.3% of CD172a+ myeloid cells), while the percentage of viable cells under lymphoid cells was significantly lower (p < 0.05) in SCM milk (85.2 ± 4.3% vs. 91.0 ± 2.2% of CD172a-lymphoid cells) compared to healthy milk (Figure 4B).

Using monoclonal antibodies to camel CD4 and WC1, the percentages of helper T cells and γδ T cells were analyzed in milk from healthy and SCM animals (Figure 5A). While the fraction of WC1+ γδ T cells was comparable (p > 0.05) between milk samples from healthy and SCM animals, the percentage of CD4+ T helper cells was significantly (p = 0.002) lower in SCM (7.8 ± 1.4% of lymphocytes) than healthy milk (16.3 ± 1.8% of lymphocytes) (Figure 5B).

The analysis of forward scatter (FSC) and side scatter (SSC), which are indicators for cell size and granularity, respectively, revealed higher (p < 0.05) FSC values for granulocytes and macrophages from SCM milk compared to healthy milk (Figure 6A). Only for granulocytes, the SSC values were lower (p < 0.05) in SCM milk compared to healthy milk (Figure 6B).

The comparison between macrophages from SCM and healthy milk regarding their expression levels of several cell surface molecules revealed significant changes in their phenotype. While the abundance of cell surface CD172a, CD163, and CD18 did not differ (p < 0.05) between SCM and healthy milk samples, macrophages from SCM milk showed higher levels of MHCII but lower levels of CD14 and CD11a when compared (p < 0.05) to macrophages from healthy milk (Figure 7).

The antimicrobial function of milk phagocytes (granulocytes and macrophages) was analyzed by the evaluation of bacterial phagocytosis by flow cytometry (Figure 8A). The percentage of phagocytosis-positive cells was higher (p < 0.05) for phagocytes from SCM milk than healthy milk. The phagocytic capacity (the number of bacteria ingested by each cell as measured by MFI of phagocytosis-positive cells), however, did not differ (p > 0.05) between cells from SCM and healthy milk (Figure 8B).