In Tuberculous Cattle Transiting Early Peripartum Period Specific In Vitro PBMC Stimulation Induces an Increase of CD14 Expressing Cells, B B2 Cells, and CD25 Cells with Half IFNγ Production

Mammalian tuberculosis is a chronic granulomatous infectious disease caused by members of the Mycobacterium tuberculosis complex that affects cattle and a wide range of other mammalian species, including human beings (WOAH, 2022). Within this complex Mycobacterium bovis is the major causative agent of bovine tuberculosis (TB) (Blanco et al., 2021). TB poses a public health threat because of its zoonotic nature (de Macedo Couto et al., 2022) and is worldwide reported in cattle generating significant financial loss (Blanco et al., 2021). Zoonotic TB has been associated with the extrapulmonary form in children, usually caused by the consumption of unpasteurized milk from infected cows (de Macedo Couto et al., 2022). World Health Organization points out that human tuberculosis surveys need adjustments to include children and extrapulmonary tuberculosis because they focus only on bacteriologically confirmed tuberculosis in adults (WHO, 2019, 2020). In Argentina between 2018 and 2019, Garrahan Pediatric Hospital reported three cases of pediatric zoonotic TB in immunocompetent patients. Two cases exhibited extrapulmonary disease and referred ingestion of dairy products purchased in informal markets (Highton et al., 2018; Vega Saldaña et al., 2019). The reduction of M. bovis infection in cattle should be the pillar of disease prevention in humans (de Macedo Couto et al., 2022).

M. bovis is an acid-fast intracellular pathogen, and the cell-mediated immune response is essential in its control (Maggioli et al., 2015; Guerra-Maupome et al., 2019). In M. bovis infected host, immune protection and diagnosis depends on cell-mediated immunity (Pollock et al., 2005). The tuberculin intradermal diagnostic test detects cell-mediated immunity in M. bovis infected cattle, either as the caudal fold test or the single cervical test (Schiller et al., 2010; Roperto et al., 2017). The former showed 68–96.8% sensitivity and 96–98.8% specificity, and the latter 80–91% and 75.5–96.8%, respectively (Schiller et al., 2010). The single tuberculin test occasionally presents false positive and false negative reactions. Cross reactivity to other mycobacteria or sensitization by other allergens can cause non-specific responses that lead to false positive reactions. Anergy present during the late stage of infection, the pre-allergic period in early cases (until 3–6 weeks post-infection), desensitized animals by PPD administration during the preceding 8 to 60 days, old cattle, postparturient desensitization, low potency tuberculin, subcutaneous injection (rather than intradermal), or bacterial contamination of the tuberculin can lead to false negative reactions (Borham et al., 2022).

T helper (Th) 1 cell-mediated immune response in infected cattle is characterized by the production of IFNγ capable of activating the microbicidal pathways of macrophages. LTCD4 appears to be the dominant population producing IFNγ while LTγδ releases it in lower levels, and LTCD8 has a greater involvement in the apoptosis of infected cells (Pollock et al., 2005). After TB infection, changes in LT subpopulations occur and comprise three phases. First LTγδ decreases, and then increases (suggesting the recruitment at the site of infection and clonal expansion), second LTCD4:LTCD8 ratio increases, and third this ratio decreases (Pollock et al., 1996). TB progression also triggers a shift from a Th1 response towards a Th2 response with associated anergy of cell mediated immunity and the development of humoral immune response (Waters et al., 2012). A 91% of the bovines that show humoral specific immune response present macroscopic TB lesions, and in 73% of the cases, lesions match generalized TB (Garbaccio et al., 2019).

During mammalian reproduction, the mother fails to reject the fetal allograft because some immunological mechanisms allow maternal fetal tolerance (Skarzynski et al., 2022). The most critical phase of the productive life of high yielding dairy cows occurs from 3 weeks before to 3 weeks after calving. This lapse is known as the early peripartum period (EPPp) and is also called transition (Van Kampen and Mallard, 1997). During the EPPp, healthy cows go from a non-lactating to a lactating state and from a pregnant to a non-pregnant condition; thereby fundamental changes occur. This situation could be considered a physiological adaptation however, when changes are dramatic and long lasting, adaptations are difficult. Dairy EPPp cows struggle to regain homeostasis, but some adaptive mechanisms may be dysregulated. For example, reduced immunological competence and overt systemic proinflammatory response are present. Transcriptomic studies described increased activities in the circulatory cells that belong to the immune system, indicating that its functions are not suppressed but are deregulated. In addition, in this lapse of multiple aggressions, the overt and systemic proinflammatory response occurs with a release of proinflammatory cytokines capable of attenuating the cellular immune response (Trevisi and Minuti, 2018).

Kerr et al. (1946) reported adverse effects of pregnancy on TB diagnosis because they showed that of 20 bovines positive to the tuberculin test, seven lost the capacity for immunological reaction after parturition, and four to six weeks after parturition those bovines recovered this capacity. Buddle et al. (1994) described that pregnancy did not appear to affect the susceptibility to M. bovis infection. Recently, a cross-sectional study surveyed 1865 farmed cattle from 79 herds in selected dairy-intensive districts of Bangladesh. This study identified pregnancy as a risk factor associated with TB infection at cattle level, a risk factor that also increased the odds of TB infection by 1.7 times (Shaheenur et al., 2020). As there is little information about the role of pregnancy in the ongoing of TB immune response in naturally infected cattle, the objective of this study was to evaluate activation, expansion, and IFNγ production after M. bovis in vitro specific stimulation of PBMC leukocyte subpopulations from Argentinian Holstein cows that reacted to the tuberculin skin test and were transiting the EPP period.

This study enrolled 19 Argentinian Holstein cows older than two years housed on three private dairy farms. Two farms had endemic TB confirmed by the single caudal fold test with PPDb (SENASA, 2012) and by the presence of disseminated granulomatous lesions. The third farm had never presented PPDb reactors for 12 years. The EPPp was established between three weeks before birth and three weeks after, according to Van Kampen and Mallard (1997). Ten of the 19 cows reacted to the single PPDb caudal fold test and came from dairy farms with endemic TB. PPDb cows were sub-classified into two study groups based on EPP period establishment. One group consisted of PPDb reactors that transited the EPP period (PPDbEPPp) (n=5). The second group consisted of PPDb reactors that did not transit the EPP period (PPDbNoEPPp) (n=5). The remaining nine cows did not react to the PPDb caudal fold test and came from a dairy farm that had not reported TB for 12 years. PPDb negative cows were also sub-classified into two groups based on EPP period establishment. The third group consisted of non PPDb reactors that transited the EPP period (NoPPDbEPPp) (n=5). The last group consisted of non PPDb reactors that did not transit the EPP period (NoPPDbNoEPPp) (n=4).

Noncoagulated blood (15 mL) was extracted by jugular venipuncture using ethylenediaminetetraacetic acid (EDTA) (ANTICOAGULANT W, WIENER, Rosario, Argentina) from all cows in this study. Samples were collected once morning milking ended. Immobilization was done according to welfare rules with a nontraumatic halter. Absolute blood cell populations and relative leukocyte differential counts were rated with a coulter (BC 3000 PLUS MINDRAY, Shenzhen, Popular Republic of China) and Giemsa stained blood smears (MERCK, Saint Paul MN, EEUU), respectively.

To perform the specific stimulation assay PBMC were separated by gradient centrifugation from bovine noncoagulated blood. Blood was diluted in PBS (1:3) and layered onto Histopaque 1077 (SIGMA-ALDRICH, Saint Louis, EEUU). Diluted blood was centrifuged under 400 g for 30 minutes at room temperature with a swinging bucket without a brake (SORVALL RC-3C, SORVALL THERMO SCIENTIFIC ™, Waltham, USA). PBMC were collected, and washed twice in PBS, and Trypan blue vital staining assessed viability. PBMC were resuspended, at a concentration of 1x106 PBMC/mL, in RPMI 1640 media (SIGMA-ALDRICH) supplemented with 0.3 g/L glutamine (SIGMA-ALDRICH), 2g/L sodium bicarbonate (SIGMA-ALDRICH), 50 mg/L gentamicin (SIGMA-ALDRICH), and 10% bovine fetal serum (PAA LABORATORIES GmbH, Cölbe, Germany).

To determine in-vitro PBMC specific activity 500 µL (1x106) of suspension were set in duplicated tubes; one was to define resting initial values and the other to define values after specific stimulation or final values. Specific stimulation was with PPDb (CDV Serie 044, Ciudad Autónoma de Buenos Aires, Argentina) at a final concentration of 20 µg/mL (Joardan et al., 2002; Hodgkin, 2005). PPDb stimuli were certified by the Argentinian National Animal Health Authority (SENASA) under WOAH standards, derived from inactivated AN 5 strain, and the concentration was 1mg/ml containing 32.500 UI/ml. Incubation lasted six days at 37°C in a 5% CO2 chamber (Joardan et al., 2002; Waters et al., 2000). To measure PBMC metabolic activity and viability the colorimetric experiment with thiazolyl blue tetrazolium bromide (MTT) was performed (Ramayo et al., 2005).

Dual-color indirect immunolabelling of duplicates and FC defined resting and stimulated PBMC sets and subsets percentages. Autofluorescence and nonspecific reaction to secondary antibody (Ab) controls were included. For dual-color indirect immunolabelling primary monoclonal antibodies (MAb) and secondary Ab were in cocktail of two. Each MAb was at 15 µg/mL, and each secondary Ab was in a 1:200 dilution (Traversa et al., 2010). MAb were mouse IgG1 and IgM isotypes; secondary Ab were goat antimouseIgG1 conjugated to phycoerythrin (PE), and goat antimouse IgM conjugated to fluorescein isothiocyanate (FITC) (Jackson IMMUNO RESEARCH LABORATORIES INC, Bar Harbor, USA). Table 1 details MAb used during dual indirect immunolabelling, MAb specificity and cells recognized, and fluorescence channel. During primary and secondary PBMC immunolabeling incubation periods lasted 15 minutes under darkness and refrigeration. PBMC were washed, fixed with formalin solution, and stored under darkness. A flow cytometer BD FACSCanto™ (BD™, Franklin Lakes, USA) acquired ten thousand fixed PBMC. FCS EXPRESS 3 trial version (De Novo Software, Los Angeles, USA) processed acquisition data of resting and stimulated PBMC populations

Table 1: MAb and secondary Ab applied during secondary immunolabeling of bovine PBMC.

Leukocytes sets and subsets were indirectly immunolabeled with specific MAb and secondary Ab. MAb were mouse IgG1(MIgG1) and IgM(MIgM) isotypes. Ab were goat anti-mouseIgG1 conjugated to phycoerythrin (PE) or goat anti-mouseIgM conjugated to fluorescein isothiocyanate (FITC). Minor leukocyte populations were labeled with PE and major ones with FITC so that minority populations were labeled with the fluorochromes that exhibit the highest emission capability.

and subpopulations. PBMC gates were defined with a side scatter (SSC) and forward scatter (FSC) dot plot. Then, a second dot plot was performed to define fluorescence in PE and FITC channels. To corroborate gates in fluorescence channels a histogram determining fluorescence peaks was set, and data from those peaks were backgated to the second dot plot. Statistical information to obtain percentages of each population or subpopulation was requested.

IFNγ production was quantified in duplicates with a sandwich ELISA (BOVIGAM, PRIONICS GmbH, Zurich, Switzerland) in PBMC stimulation assay culture media, following the manufacturer protocol. The colorimetric signal was read with a microplate reader under 450 nm. The results were the mean optical density (OD) of duplicate supernatants plus the standard deviation (Rhodes et al., 2001). To complement the cytoquine assay PBMC metabolic activity was measured with the colorimetric experiment with MTT. Plates were read with a microplate reader under a 570 nm filter and stimulation index (SI) was expressed as the ratio between PPDb treated PBMC and concanavalin-A treated PBMC (Ramayo et al., 2005).

Statistical analyses were performed with GRAPHPAD PRISM trial version 9.0.0 for Windows, GraphPad Software, San Diego, California USA, www.graphpad.com. Stimulated PBMC subsets were statistically compared against resting PBMC subsets with a multiple paired t test. p adjusted values were chosen, and the significance level was p<0.05. Statistical analyses of the other variables were performed with one-way ANOVA followed by Dunett’s multiple comparisons test, and the significance level was p<0.05. If normality could not be assumed Kruskal and Wallis test replaced the one-way ANOVA, and the significance level was p<0.05.

The average hematology parameters of cows under study are summarized in Table 2. Average absolute counts of red blood cells and platelets (data not shown) were within bovine hematological reference ranges for healthy cattle. The average absolute leukocyte counts were higher than the reported ranges for healthy cattle in all study groups, and there were no statistically significant differences between groups. The differential leukocyte counts (data not shown) were also within normal ranges for healthy cattle (Roland et al., 2014), and there were no statistically significant differences between groups.

After six days in culture with PPDb, PBMC were alive, metabolically active, and produced IFNγ. Figure 1A shows PBMC activation through stimulation indexes from the MTT assay. The difference between MTT assay average stimulation indexes between study groups was not statistically significant. Figure 1B displays cytokine production through optical densities (OD) from the IFN-γ assay. PBMC produced higher levels of IFNγ in both PPDb reactor groups. PPDbNoEPPp cows and PPDbEPPp cows presented a mean OD value of 0.86±0.96 and 0.46±0.57, respectively. Even though IFNγ mean production in PPDbNoEPPp cows almost doubled the production of PPDbEPPp cows, differences were not statistically significant (p=0.06).

WC1 γδ TCR cells responsible for recognizing PPDb antigens did not display statistically significant increase response in cows from the groups under study with stimulation assay measured by flow cytometry (Figure 2A).

Regarding the processing and presentation of the antigen, CD14 expressing cells increased significantly in both PPDb study groups and B B2 cells increased significantly in PPDbEPPp study group. PPDbEPPp cows presented CD14 mean resting values of 1.44% and mean stimulated values of 11.69% (p=0.008), and PPDbNoEPPp cows showed values of 2.95% and 9.06% (Table 3) (p=0.005), respectively (Figure 2B). B B2 cells from cows in PPDbEPPp displayed a statistically significant response to PPDb stimuli (p=0.004) (Figure 2C) from a mean resting value of 8.97% to a mean stimulated of 24.92% (Table 3).

After stimuli, PBMC from PPDb reactor cows showed the highest percentages of cells expressing CD25 activation marker (interleukin 2 receptor α chain). The increase was statistically significant in PPDbEPPp study group (p=0.03) (Figure 2D). PPDbEPPp cows presented a mean stimulated value of 17.41% from a mean resting value of 5.45% (Table 3).

Respecting the adaptive cellular immune response, the mean stimulated LT values were lower than resting values in all the experimental groups (Figure 2E). In the PPDbEPPp study group, the resting LT mean value was 66.86%, while the NoPPDbNoEPPp study group presented 62.08% (Table 3). In NoPPDbEPPp and PPDbNoEPPp study groups those values were 54.62% and 48.61% respectively (Table 3).

CD8 cytotoxic T cells presented a weak response in the two study groups that did not transit the EPP period, and in the two groups that transited the EPP period, CD8 T cells response was negative (Figure 2F). In the PPDbNoEPPp group, CD8 T cells mean resting percentage was 12.92%, and the mean stimulated percentage was 16.79%. In the NoPPDbNoEPPp group, these percentages were 20.41% and 21.35%, respectively (Table 3).

CD4 helper T cells survived but presented a negative response in all study groups (Figure 2G). However, in PPDb negative cattle resting CD4 T cells were higher than in PPDb positive cattle. CD4 T cells resting values were 32.78% in the NoPPDbEPPp group and 32.32% in the NoPPDbNoEPPp group (Table 3).

CD45Ro recall activated T cells presented a positive response in the four study groups (Figure 2H). The mean resting values of CD45Ro T cells in both groups of animals that went through the EPP period were lower than those presented by both groups that did not go through the EPP period. In PPDbEPPp cows, resting CD45Ro T cells was 1.22%, and in the NoPPDbEPPp cows, it was 1.85% (Table 3).

Specific in vitro immune response of TB natural host was characterized, quantitatively and functionally because limited research has been conducted in M. bovis naturally infected dairy cows transiting the early peripartum period. Quantitative characterization was carried out with differential leukocyte counts and functional characterization with specific stimulation assays based on flow cytometry counts, IFNγ production, and color based MTT assay.

All study groups showed higher absolute leukocyte and lymphocyte counts than reported values for healthy cattle (Roland et al., 2014). According to leukogram patterns interpretation and statistical analyses, leukocytosis and lymphocytosis can be considered physiologic (Webb and Latimer, 2011; Eclinpath Com, 2021). NoPPDbEPPp study group showed a physiologic neutrophilic profile with no increment in band neutrophils and without lymphopenia (Webb and Latimer, 2011; Eclinpath Com, 2021). Even though leukocytosis and lymphocytosis were physiologic in this study, some authors describe in tuberculin

Table 2: Absolute differential leukocytes counts in PPDbEPPp, PPDbNoEPPp and NoPPDbEPPp cattle (cells/µL).

reactor cattle leukocytosis (Mohankumar et al., 2011; Quevillon et al., 2013) and lymphocytosis (Mohankumar et al., 2011; Quevillon et al., 2013; Javed et al., 2010; Killick et al., 2011).

In the present study, all study groups showed physiological leukograms, but when FC assessed PBMC characterization, differences between leukocyte subsets became detectable. NoPPDbNoEPPp study group showed resting percentages of LTCD4, LTCD8, and LTγδ within reported values for healthy cattle that did not transit early peripartum period and lowered CD14 cells percentage (Pastoret et al., 1998). NoPPDbEPPp study group presented resting LT, LTCD4, and LTCD8 values within reported ones for healthy cattle that transited the early peripartum period (Harp et al., 2004). Both study groups that were PPDb reactor showed LTCD4:LTCD8 ratios (data not shown) according to those described in tuberculous cattle (Pollock et al., 1996), and LTCD4 and LTCD8 percentages accorded to reported values for tuberculous cattle (Manzo-Sandoval et al., 2023). The PPDbEPPp group showed the lowest LTCD4:LTCD8 ratio, with the highest LTCD8 absolute count and percentage, but the PPDbNoEPPp group had the lowest LTCD8 counts. Furthermore, both PPDb reactor study groups showed lower resting CD14 expressing cells percentages than percentages reported for healthy cattle that did not transit EPPp (Pastoret et al., 1998), and the lowest one belonged to the PPDbEPPp study group. Regarding resting LTγδ counts in PPDb reactor study groups, values agreed with those reported in tuberculous cattle (Pollock et al., 1996).

In our four study groups, specific lymphocyte stimulation with PPDb caused increased expression of the activation marker CD25, IFNγ production, and metabolically active PBMC survival until the stimulation assay ended. After PPDb stimulation, the proportion of CD14 expressing cells increased significantly in both study groups of PPDb reactors, although they had lower resting values. In cattle, the CD14 molecule is expressed mainly on macrophages and monocytes (Sohn et al., 2004). Macrophages and monocytes are antigen presenting cells that play a central role in innate immunity against infection and help initiate cell mediated adaptative immunity (Blanco et al., 2021). Our results suggest that in naturally M. bovis infected dairy cows, the early peripartum period would not influence the response to PPDb by CD14 expressing cells.

In the current study, B B2 lymphocytes increased significantly after specific stimuli in tuberculin reactor cattle that transited the early peripartum period. The B B2 marker is present in most ruminants B lymphocytes (with equivalence to CD19 marker) (Stabel et al., 2022), has no known human orthologue (Foote et al., 2007; Davis and Hamilton, 2008), and may play a role in immunoglobulin production in vaccinated cattle (Foote et al., 2007). The B B 2 subset demonstrated a trend towards a higher number in Mycobacterium avium subsp. paratuberculosis naturally infected cows in the clinical stage of disease in specifically stimulated cells (Stabel et al., 2022). Our results suggest that the early peripartum period in naturally M. bovis infected dairy cows can induce the expansion of B B2 lymphocytes after specific stimuli; this finding might be part of a Th2 immune profile.

CD25 lymphocytes expanded significantly in PPDbEPPp study group. CD25 is the interleukin 2 receptor α chain (Maue et al., 2005), and interleukin 2 (IL2) is a cytokine that, by interacting with CD25 can assume pleiotropic functions (Ross and Cattrell, 2018). In human tuberculosis CD25 is a phenotype marker of LTCD4 regulatory subsets (Roberts et al., 2007; Ahmed and Vyakarnam, 2020). Unlike human tuberculosis, in bovine tuberculosis, LTγδ perform regulatory functions (Waters et al., 2011). In bovine, LTγδ coexpress low levels of CD25 (Baldwin et al., 2021) and releases lower levels of INFγ than LTCD4 (Pollock et al., 2005). In M. bovis experimentally infected cattle, CD25 is also coexpressed by activated LTCD8 that produces lowered levels of INFγ (Liébana et al., 1999).

In our study, IFNγ production carried out by PPDb stimulated PBMC from PPDbNoEPPp study group almost doubled the production of PPDbEPPp study group. Relating the IFNγ detection and the reproductive stage in cattle, Buddle et al. (1994) reported that in the first test after calving, the IFNγ production by PPDb stimulated PBMC was lower than the production before calving.

Results from this study suggest that in naturally M. bovis infected dairy cows, the early peripartum period might not influence the presentation of PPDb by CD14 expressing cells, and might upregulate the expansion of B B2 and CD25 lymphocytes and lower IFNγ production when PBMC are in vitro specifically stimulated. The expansion of B B2 cells might indicate antibody synthesis that can be detected with serological diagnostic tests, complementing the single tuberculin test in infected tuberculin false negative cows during the early peripartum period. The same happens with the Bovigam assay because the expansion of CD25 lymphocytes in M. bovis infected cows during the early peripartum period might point out the activation of some subsets that might have lowered, but still detectable, IFNγ production. This study provides information on the influence of the transition period in bovine tuberculosis in vitro immune response, a field not extensively studied.

This work was supported by Secretaría de Ciencia Arte y Tecnología Universidad Nacional del Centro de la Provincia de Buenos Aires, Instituto Nacional de Tecnología Agropecuaria Estación Experimental Agropecuaria Balcarce, Instituto de Investigaciones Biomédicas en Retrovirus y SIDA and Monoclonal Antibody Center Washington State University. We thank Claudia Morsella for performing the sandwich ELISA for IFN-γ determination.

1. In tuberculous cows, CD14 cells specifically stimulated increased significantly, suggesting that early peripartum period might not influence antigen acquisition and presentation.
2. In naturally M. bovis infected dairy cows transiting early peripartum period, specifically stimulated B B2 (BAQ44A+) lymphocytes expanded significantly.
3. After stimuli, PBMC from PPDb reactor cows showed increased percentages of cells expressing the CD25 activation marker (interleukin-2 receptor α chain), but in PPDb reactor cows transiting early peripartum period, the increase was statistically significant.
4. IFNγ production was half in specifically stimulated PBMC from tuberculin reactor cattle transiting early peripartum period. 
5. In PPDb reactor dairy cows that go through early peripartum period and are false negative to the single intradermal tuberculin test, serological tests and IFNγ detection assays could be helpful as complementary tests.

MJT conceived and conducted the research and the experiments and wrote the manuscript. MS acquired flow cytometry. MAC Statistically analyzed stimulation assay results. ER statistically designed the study. FP, SE and MCJ supervised and oversaw the research. WCD developed immunolabeling protocol. MCJ and WCD conceptualized the manuscript. MCJ and WCD contributed equally to the research.

We confirm that no competing interests exist with any company or institution mentioned in the manuscript.