Cytokine responses of bovine macrophages to diverse clinical Mycobacterium avium subspecies paratuberculosis strains

Background Mycobacterium avium subsp. paratuberculosis (MAP), the causative agent of Johne's disease (JD) persistently infects and survives within the host macrophages. While it is established that substantial genotypic variation exists among MAP, evidence for the correlates that associate specific MAP genotypes with clinical or sub-clinical disease phenotypes is presently unknown. Thus we studied strain differences in intracellular MAP survival and host responses in a bovine monocyte derived macrophage (MDM) system. Results Intracellular survival studies showed that a bovine MAP isolate (B1018) and a human MAP isolate (Hu6) persisted in relatively higher numbers when compared with a sheep MAP isolate (S7565) at 24-hr, 48-hr and 96-hr post infection (PI). MDMs stimulated with B1018 up-regulated IL-10 at the transcript level and down-regulated TNFα at the protein and transcript levels compared with stimulations by the S7565 and Hu6. MDMs infected with Hu6 showed a down regulatory pattern of IL-10 and TNFα compared to stimulations by S7565. Cells stimulated with B1018 and Hu6 had low levels of matrix metalloprotease-3 (MMP3) and high levels of tissue inhibitor of metalloprotease-1 (TIMP1) at 96-hr PI relative to MDMs stimulated by S7565. Conclusion Taken together, results suggest that the bovine (B1018) and the human (Hu6) MAP isolates lead to anti-inflammatory and anti-invasive pathways in the macrophage environment whereas the sheep (S7565) MAP isolate induces a pro-inflammatory pathway. Thus the infecting strain genotype may play a role in polarizing the host immune responses and dictate the clinicopathological outcomes in this economically important disease.


Background
Johne's disease (JD) is caused by the intracellular pathogen, Mycobacterium avium subsp. paratuberculosis (MAP). Several molecular techniques have been applied to differentiate and characterize MAP isolates from diverse hosts and geographic locations [1][2][3]. A recent study applied highly discriminatory molecular markers termed short sequence repeats (SSR) to analyze the diversity among MAP isolates from a variety of hosts [4]. The results provided evidence for interspecies transmission of several MAP genotypes with some showing host specificity. Intriguingly, all genotyping studies addressing diversity using a variety of methods show that MAP isolated from human Crohn's disease cases are a subset of MAP genotypes widespread in distribution among animal populations. These findings raise several questions regarding the association of specific genotypes with human disease and/ or chronic sub-clinical versus overt clinical disease in animals. Since no information on disease phenotypes was obtained when the isolates were acquired for genotyping in our laboratory, logical interpretation of genotype-phenotype associations was not possible. In the absence of clinical data associated with genotypes, and a suitable animal model to rapidly identify strain-associated differ-ences, studying MAP interactions in a cellular (macrophage) interphase provides an indirect tool to help dissect the early molecular events that occur during hostpathogen interactions.
Despite the fact that hosts have only a limited number of pathways in which they respond to pathogens, macrophages show both pathogen-specific gene expression profiles and a shared gene expression pattern when infected with diverse bacterial pathogens [5]. An in vitro study of human macrophage responses to a repertoire of genotypically and epidemiologically well defined clinical isolates of Mycobacterium tuberculosis (MTB) showed a strain dependent host response [6]. A more recent study has IL-10 mRNA expressed by MDM cells exposed to M. paratuberculosis strains over time was measured by Real time RT PCR and fold change in gene expression relative to β actin was calculated by 2 -∆∆CT method Figure 1 IL-10 mRNA expressed by MDM cells exposed to M. paratuberculosis strains over time was measured by Real time RT PCR and fold change in gene expression relative to β actin was calculated by 2 -∆∆CT method. Mean fold change in gene expression is plotted on Y-axis (note that the Y1-axis scales have been optimized for magnitude of IL-10 expression for each strain). Intracellular bacterial numbers based on the amplification of hsp65 were calculated based on genome size of MAP and represented as genome equivalents (GE) on second Y-axis. MDMs stimulated with B1018 and Hu6 MAP isolates increased IL-10 mRNA by 96 hrs PI relative to cell stimulations by S7565 MAP isolate and M. avium.
shown a shared and a unique gene expression signature by human macrophages stimulated with four isolates of M. avium that varied in growth characteristics [7]. Significant differences in cytokine-chemokine profiles or global gene expression profiles in either well-established cell lines (THP-1 or U937) or peripheral blood mononuclear cells (PBMCs) in response to diverse pathogenic and non pathogenic mycobacteria, have also been documented by several recent studies [8][9][10][11]. Taken together, available data in the current literature strongly suggests that macrophages infected with mycobacteria have differential gene expression profiles and that the infecting genotypes may dictate down stream host responses. Surprisingly, there have been no reports so far about comparative analyses of diverse clinical isolates of MAP within a host/host macrophage, which is a well established fact in other mycobac-teria. We believe that this crucial piece of evidence is important in order to understand complex mechanisms underlying the virulence of this economically important pathogen.
Towards these long term goals, in this study we asked if genotypically distinct strains of MAP derived from different host species elicit differential responses in bovine monocyte derived macrophages. To test our hypothesis that there would be no strain-specific variation in host responses, we studied the growth kinetics of genotypically distinct strains of MAP in both BOMAC cells [12] and bovine monocyte derived macrophages (MDM). We compared the modulation of cytokines such as IL-10, TNFα and matrix metalloproteinase (MMP) genes such as MMP3 and MMP9 as a function of infecting genotype. The IL-10 protein secreted by MDM cells exposed to MAP over time was measured by ELISA Figure 2 IL-10 protein secreted by MDM cells exposed to MAP over time was measured by ELISA. Total amount of protein (pg/ml) is plotted on Y-axis (note that the Y1-axis scales have been optimized for magnitude of IL-10 expression for each strain). Intracellular bacterial numbers were calculated based on genome size of MAP and represented as genome equivalents (GE) on second Y-axis. MDMs stimulated with B1018 and Hu6 MAP isolates gradually up regulated IL-10 secretion from 2-hr until 96-hr PI.
importance of these cytokines in JD has been reported elsewhere [13,14]. Cytokines IL-10 and TNFα were evaluated because the relative balance in expression of these cytokines indicates macrophage activation. PBMCs isolated from cattle infected with JD have been shown to up regulate MMP9 and TIMP after stimulation with MAP [15]. MMPs when secreted in lower quantities help in leukocyte migration but when secreted in larger quantities cause tissue destruction [16]. A balance between MMP and TIMP is important in the extent of tissue degradation [16]. In summary, in this study we report a differential response of bovine monocyte derived macrophages to a variety of MAP isolates.

Intracellular survival kinetics of MAP strains
B1018 (bovine MAP isolate) was efficiently phagocytosed by bovine MDMs and persisted at fairly high numbers when compared to other isolates at all time points (Figures 1, 2, 3, 4). S7565 (sheep MAP isolate) decreased in bacterial numbers until 24-hr PI, started to multiply by 48-hr PI and dropped in total intracellular bacterial num-TNFα mRNA expressed by MDM cells exposed to MAP over time was measured by Real time RT PCR and fold change in gene expression relative to β actin was calculated by 2 -∆∆CT method Figure 3 TNFα mRNA expressed by MDM cells exposed to MAP over time was measured by Real time RT PCR and fold change in gene expression relative to β actin was calculated by 2 -∆∆CT method. Mean fold change in gene expression is plotted on Y-axis (note that the Y1-axis scales have been optimized for magnitude of IL-10 expression for each strain). Intracellular bacterial numbers based on the amplification of hsp65 were calculated based on the genome size of MAP and represented as genome equivalents (GE) on second Y-axis. MDMs stimulated with B1018 expressed lower levels of TNFα mRNA relative to other cell stimulations. In BOMAC cells S7565 multiplied more rapidly and stayed in higher numbers at 96 hr PI relative to B1018. However, in bovine MDM cells B1018 was efficiently internalized and stayed in higher numbers relative to S7565 and Hu6 MAP isolates (data not shown).

TNFα expression profile
MDMs stimulated with B1018 and Hu6 isolates showed a down regulatory trend in TNFα mRNA expression from 2-TNFα protein secreted by MDM cells exposed to MAP over time was measured by ELISA Figure 4 TNFα protein secreted by MDM cells exposed to MAP over time was measured by ELISA. Total amount of protein (pg/ml) is plotted on Y-axis (note that the Y1-axis scales have been optimized for magnitude of TNFα expression for each strain). Intracellular bacterial numbers were calculated based on the genome size of MAP and represented as genome equivalents (GE) on second Y-axis. MDMs stimulated with B1018 secreted low amounts of TNFα protein relative to other cell stimulations.

MMP3 mRNA expression
MDMs stimulated with B1018 increased the production of MMP3 mRNA at 16-hr which was followed by a rapid decline until 96 hrs PI (table 1). Cells stimulated with S7565 produced a peak level of MMP3 at 24-hr that declined by 96-hr. While MMP3 mRNA production by cells stimulated with B1018 was high at 16-hr relative to cells stimulated by S7565 and Hu6, the data was not statistically significant (P = 1.0). At 24-hr there was a significant increase (P < 0.05) in the production of MMP3 mRNA by cells stimulated with S7565 relative to other stimulations. There was a gradual up-regulation of MMP3 mRNA production by MDMs stimulated with Hu6 until 48 hrs and declined by 96-hr.

MMP9 mRNA expression
Cells stimulated with B1018 showed lower levels of MMP9 mRNA production at 48-hr and 96-hr PI relative to cell stimulations with other MAP isolates (table 2). Cells stimulated with S7565 had a peak MMP9 mRNA production at 48-hr after infection, which was significantly higher (P < 0.05) when compared with stimulations by other MAP isolates (table 2). Cells stimulated with Hu6 also showed an up-regulatory trend in MMP9 mRNA production until 96-hr (table 2). Cells stimulated with Hu6

TIMP1 mRNA expression
LSMean values of TIMP1 mRNA levels suggested that MDMs stimulated with B1018 showed higher levels relative to cell stimulations by S7565 (data not statistically significant) (p = 1.0). There was a peak production of TIMP1 mRNA observed at 96-hr in MDMs stimulated with B1018 (table 3).

Discussion
Macrophages are the first line of host defense against any invading bacteria. Despite the fact that macrophages offer a hostile environment to several pathogenic bacteria, MAP is able to persistently survive and replicate within the phagosome environment of host macrophages. Studying the biochemical processes operating at the host-pathogen interface will help elucidate the mechanisms by which MAP has developed expertise to survive and replicate inside macrophages.

Conclusion
The present findings provide key insights into the MAP genotype-disease phenotype associations. Further analysis of this complex "ancient dialogue" between MAP and macrophages derived from its natural host (bovines) will help elucidate the pathogenesis associated with different genotypes isolated from diverse host species. These studies will aid in understanding the proximal events involved in the progression of JD and the virulence of MAP isolates thus enabling design of early intervention strategies. Future studies with a broader range of MAP isolates with common and unique genotypes associated with JD are warranted.

Preparation of monocyte derived macrophages (MDM)
Two colostrum-deprived Holstein bull calves obtained from a Johne's disease free herd served as a source for peripheral blood and MDMs. The calves were tested 4 and 8 weeks after birth by fecal culture and serum ELISA and were confirmed to be JD free. The protocol used for the preparation of MDMs is described elsewhere [40]. Briefly, peripheral blood was drawn from jugular vein into acidcitrate dextrose containing vacutainers (BD Vacutainer, Rutherford, NJ). Blood was centrifuged at 2000 rpm for 20 minutes to obtain a clean buffy coat. Buffy coats were re-suspended in sterile 1× PBS (1:10 dilution) and overlaid on Histopaque (Sigma Aldrich, St. Louis, Mo) following manufacturer's recommendations. The tubes were centrifuged at 400 × g for 30 minutes at room temperature to separate mononuclear cells from other polymorphonuclear leukocytes. Mononuclear cells from the interphase were harvested carefully using a sterile Pasteur pipette and transferred to a second sterile 50-ml conical tube. The cells were washed with 10-ml of 1× PBS at 250 × g for 10 minutes. Supernatant was discarded and the cell pellet was resuspended in a small volume of 1× PBS. The mononuclear cells were transferred to TEFLON jars (Savillex Corporation, Minnetonka, MN) containing RPMI 1640 medium supplemented with L-glutamine, HEPES and 20% autologous serum. The jars were incubated at 37°C, 95% air and 5% CO 2 for 4 days. After four days monocytes differentiated and became larger in size. Differentiated monocytes were counted using a hemacytometer and seeded onto the tissue culture plates at appropriate dilutions and incubated at 37°C, 95% air and 5% CO 2 for 2 hours. Plates were washed twice with sterile 1× PBS to remove the nonadherent cells. The adherent cells were used for all in-vitro infections.

M. paratuberculosis isolates
The selected MAP strains included B1018 (SSR fingerprint: 7G4GGT9nG), S7565 (SSR fingerprint: 15G3GGT), Hu6 (SSR fingerprint: 7G5GGT11nG) and Ma6043 (IS900 positive isolate identified as M. avium intracellulare by multiple target analyses; no SSR data) (11). B1018 was isolated from a cow with clinical JD and carried a fingerprint which is common to about 12% of bovine M. paratuberculosis strains in a national collection (Harris and Sreevatsan, Unpublished) and in greater than 45% of isolates from Ohio (21), S7565 was isolated from sheep and Hu6 was isolated from a Crohn's disease patient.

Bacterial cultures
All the MAP cultures were incubated at 37°C on MB7H9 plates supplemented with OADC enrichment medium and Mycobactin J. After 3-4 weeks of growth on MB7H9 plates, cultures were confirmed to be free of other contaminating organisms as determined by nil growth on BHI or blood agar plates incubated overnight at 37°C. Few colonies from MB7H9 plate cultures were inoculated into MB7H9 broth culture supplemented with OADC enrichment medium and Mycobactin J and incubated at 37°C for 3 days to achieve actively growing MAP. Three day-old cultures were used to obtain an optical density at 600 nm (OD 600 ) to determine the colony forming units (cfu) of bacteria using the formula: 0.3 at OD600 = 10 9 cfu/ml, and applied in all in vitro infections of bovine MDM. Bacteria were used at a 5:1 multiplicity of infection. Based on the genome size of MAP one genome equivalent was calculated to be equal to about 9.9 fmg of MAP DNA.

Development of hsp65 standard for quantification purposes
DNA extracted from the broth cultures of MAP was pooled and the concentrations were determined. Five, ten-fold dilutions of the DNA was made and used as template in the real time PCR for amplification of hsp65 gene.
Obtained ct values were imported onto an Excel spread sheet. Ct vales were plotted against genome equivalents (Y axis) and regression analysis was performed. This regression equation was used to estimate the bacterial numbers (as genome equivalents) in all treatment samples.

Quantitation of extracellular cytokine production by ELISA
Cytokine sandwich ELISA was used to detect IL-10 and TNFα from the culture supernatants. The protocol was adopted and modified to our conditions as described [43,44]. Briefly, 96 well micro titer plates (NUNC, Rochester, NY) were coated with mouse anti bovine IL-10 (Serotec Inc, Raleigh, NC) or mouse anti bovine TNFα (generous gift from Dr. Paape, USDA, Ames, IA) for overnight at 4°C. Next day plates were washed with PBS Tween20 (0.05%) and blocked with PBS/BSA (0.5%) for one hour at room temperatures. Culture supernatants were added to the plates. The standard protein was serially diluted and added to the corner wells. Standard protein for TNFα was purchased from Endogen where as standard protein for IL-10 was a generous gift from Dr. Howard, Animal Research Center, UK. The plates were incubated at 4°C overnight. The plates were then washed briefly with PBS Tween20 (0.05%) and mouse anti bovine IL-10 labeled with biotin (Serotec Inc, Raleigh, NC) or rabbit anti 1 bovine TNFα (generous gift from Dr. Paape, USDA, Ames, IA) was added and incubated at 37°C for 3 hours.
Plates were then washed and streptavidin HRP (Serotec Inc, Raleigh, NC) was added to plates that were used for IL-10 detection, incubated for an hour at room temperature. Anti rabbit IgG1 labeled with HRP (Biorad, CA) was added to plates for detecting TNFα and incubated at room temperature for 2 hours. All the plates were then added with color developing solution (Biorad, CA) and plates were read using ELISA plate reader at 413 nm.

Estimation of protein concentration of IL-10 standard protein
The standard protein used for quantitating IL-10 in culture supernatants was gifted by Dr. Howard, UK. The source of recombinant bovine IL-10 was culture supernatant obtained from Cos-7 cells transfected with a plasmid encoding bovine IL-10. The IL-10 supplied had 3000 biological units per ml. One biological unit of IL-10 supplied to us corresponded to an equivalent of 600 ng of protein.