Survival and persistence of Mycobacterium avium subsp. paratuberculosis (MAP) in the intestinal mucosa is associated with host immune tolerance. However, the initial events during MAP interaction with its host that lead to pathogen survival, granulomatous inflammation, and clinical disease progression are poorly defined. We hypothesize that immune tolerance is initiated upon initial contact of MAP with the intestinal Peyer's patch. To test our hypothesis, ligated ileal loops in neonatal calves were infected with MAP. Intestinal tissue RNAs were collected (0.5, 1, 2, 4, 8 and 12 hrs post-infection), processed, and hybridized to bovine gene expression microarrays. By comparing the gene transcription responses of calves infected with the MAP, informative complex patterns of expression were clearly visible. To interpret these complex data, changes in the gene expression were further analyzed by dynamic Bayesian analysis, and genes were grouped into the specific pathways and gene ontology categories to create a holistic model. This model revealed three different phases of responses: i) early (30 min and 1 hr post-infection), ii) intermediate (2, 4 and 8 hrs post-infection), and iii) late (12 hrs post-infection). We describe here the data that include expression profiles for perturbed pathways, as well as, mechanistic genes (genes predicted to have regulatory influence) that are associated with immune tolerance. In the Early Phase of MAP infection, multiple pathways were initiated in response to MAP invasion via receptor mediated endocytosis and changes in intestinal permeability. During the Intermediate Phase, perturbed pathways involved the inflammatory responses, cytokine-cytokine receptor interaction, and cell-cell signaling. During the Late Phase of infection, gene responses associated with immune tolerance were initiated at the level of T-cell signaling. Our study provides evidence that MAP infection resulted in differentially regulated genes, perturbed pathways and specifically modified mechanistic genes contributing to the colonization of Peyer's patch.
Citation: Khare S, Lawhon SD, Drake KL, Nunes JES, Figueiredo JF, et al. (2012) Systems Biology Analysis of Gene Expression during In Vivo Mycobacterium avium paratuberculosis Enteric Colonization Reveals Role for Immune Tolerance. PLoS ONE 7(8): e42127. doi:10.1371/journal.pone.0042127
Editor: Niyaz Ahmed, University of Hyderabad, India
Received: December 12, 2011; Accepted: July 3, 2012; Published: August 17, 2012
Copyright: © Khare et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was supported by funds to Sangeeta Khare from the U.S. Department of Agriculture Cooperative State Research, Education and Extension Service-National Research Initiative-Coordinated Agricultural Project-Johne's Disease Integrated Program (2004-35605-14243 sub-award Q6286224133) and Animal Health formula funds through Texas A&M University. This study was also supported by funds awarded to Leslie Garry Adams from the Texas A&M University Life Sciences Task Force. Sara D. Lawhon was supported by National Institutes of Health (U.S.) NIH-NIAID AI060933. The computational analysis work was supported in part by NIH NIAID SBIR grants 2R44AI058362-02 and R43AI084223-01 awarded to Kenneth L. Drake. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have the following interest: Kenneth L. Drake is an employee of Seralogix, LLC and has an equity position in this company. Seralogix, LLC is a private business developing the software analysis and modeling software for commercial purposes. Robin E. Everts is currently employed by Sequenom, Inc., San Diego, CA, however all activities (except reviewing manuscript) by Robin E. Everts related to this manuscript were performed while employed at the University of Illinois, Urbana-Champaign, IL. There are no patents, products in development or marketed products to declare. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials, as detailed online in the guide for authors. Leslie Garry Adams has served as an ad hoc guest editor for PLOS ONE.
Mycobacterium avium subsp. paratuberculosis (MAP) causes a chronic enteric infection (Johne's disease) in cattle and other ruminants that is established after ingestion of bacteria followed by invasion and colonization of the intestinal mucosa. The major hurdle in understanding MAP infection is its chronic nature and delayed onset of clinical symptoms. Much is known regarding the host response of chronically infected cattle, but the understanding of the early events in the host is limited. In the jejunal-ileal Peyer's patches, MAP gain entry in intestinal mucosa via interaction with M cells, goblet cells, epithelial cells, dendritic cells or macrophages –. Lesions that are characterized by aggregates of macrophages, epithelioid cells, and giant cells develop in the intestinal mucosa of experimentally infected neonatal calves within 5 months . Moreover, systemic humoral and cellular immune responses develop within months after instillation of MAP into the tonsillar crypts of neonatal calves . Mononuclear cells isolated from the intestine of cows subclinically infected with MAP showed a state of tolerance , . Several studies have focused on the role of circulating mononuclear phagocytes during MAP infection –. It appears that mucosal immunological tolerance is required for persistent infection. However, the outcome of the host-pathogen interaction depends on the collective response of various cell types that are present at the mucosal surface. Additionally, knowledge of which components of the host response are involved in the activation of innate immunity is poorly understood in chronic infections. Thus, more comprehensive knowledge is needed regarding the pathogen interaction within the host milieu of the natural site of infection. Toward this goal, we have successfully established the perinatal calf ligated ileal loop model for studying early changes in the mucosa during MAP infection .
About 65% of U.S. herds are infected with MAP . This level of infection definitely leads to contamination of the environment via contaminated water supply, dust bio-aerosol, milk and food supply. This contaminated environment not only affects the spread of MAP among animals, it may also be associated with the human intestinal disorder, Crohn's disease –. Interestingly, MAP has been isolated from intestinal tissue, as well as, peripheral blood of human patients suffering with a similar granulomatous inflammatory disease known as Crohn's disease . Several pathogens, including MAP, have been claimed to be associated with the Crohn's disease in human , . Like Johne's disease, Crohn's disease also affects a pediatric population. In a recent study, identification of MAP in gut tissue and blood from pediatric inflammatory bowel disease patients suggests the possible involvement of MAP in the early stages of development of Crohn's disease in children . Crohn's disease is likely to be more than one disease, which complicates research. Furthermore, microbial contact or invasion may be confined to parts of the alimentary tract that are relatively inaccessible to tissue sampling, such as the ileum or jejunum. The in vivo perinatal calf jejunal-ileal loop model provides an ideal animal model to more precisely study the early pathogenesis of MAP or other microbial pathogens in Crohn's disease.
In the present study, we hypothesized that MAP induces an immune tolerance in Peyer's patch very soon after invasion. To test our hypothesis, we analyzed the temporal gene expression in detail during the early stages of colonization of Peyer's patch by MAP. We applied a systems biology approach to analyze the complex microarray data enabling us to identify specific cellular pathway perturbations and predicted cell type involvement during the early infection and colonization. Furthermore, we identified innate response signatures (mechanistic genes) adequate to envisage the subsequent adaptive immune responses leading to persistent MAP infection. The aims of this study were two fold, 1) to utilize a systems biology analytical approach to gain new insights regarding the most perturbed cellular pathways; and 2) to provide evidence for tolerance related components of mucosal immunity during early MAP infection.
Materials and Methods
Culture of MAP
Mycobacterium avium subsp. paratuberculosis (ATCC 19698) from American Type Culture Collection (ATCC), Manassas, VA, was grown aerobically in 7H9 broth (Difco Laboratories, Detroit, MI) supplemented with 2.5% (vol/vol) glycerol (Sigma Chemical Co., St. Louis, MO), oleic acid-albumin-dextrose-catalase (Difco Laboratories, Detroit, MI), 0.05% Tween 80 (Sigma Chemical Co., MO), and 2 mg/liter of Mycobactin J (Allied Monitor, Inc., Fayette, MO). Single-cell suspensions and enumeration of MAP were done as described earlier .
Four clinically healthy male, unrelated Holstein-Friesian calves, 3–4 weeks of age and weighing 45–55 kg, were used in the experiment under an approved animal use protocol in accordance with animal use policy under the supervision of the Texas A & M University Institutional Animal Care and Research Advisory Committee (AUP 2007-70). The calves were fed antibiotic-free milk replacer twice daily and water ad libitum. All of the calves were tested for the presence of Salmonella spp. and MAP. Fecal specimens and rectal swabs were collected from calves two weeks prior and immediately before the experiment. Fecal specimens were prepared for the PCR for detection of MAP  and culture based detection of Salmonella . Only calves with negative tests for these pathogens were used in these experiments.
Bovine Ligated Jejunal-Ileal Loop Surgery
The calves were fasted for 24 hrs prior to the non-survival surgery, anesthetized and maintained analgesic for the course of the 12 hrs experiment. In brief, anesthesia was induced with Propofol (Abbot Laboratories, Chicago, IL) followed by placement of an endotracheal tube and maintenance with isoflurane (Abbot Laboratories, Chicago, IL) for the duration of the experiment. The detailed bovine ligated ileal loop surgery procedure is described elsewhere . Loops were prepared exclusively in the 1.0–1.2 meter long Peyer's patch that is proximal to the ileocecal junction and included jejunum and ileum. Loops were inoculated with 3.0 ml of PBS or 3×109 cfu of MAP in 3.0 ml of PBS. At 0.5, 1, 2, 4, 8 and 12 hrs after inoculation, one each of control (PBS inoculated) and experimental (MAP inoculated) loops were excised. Samples for bacteriologic culture and RNA extraction were collected as described below. Throughout the experimental procedure, the calves were monitored for vital signs (blood pressure, heart rate, hydration status, anesthesia depth and temperature). The calves were euthanized with a rapid overdose (single bolus at 60 mg/lb IV) of pentobarbital sodium after the final samples were collected at 12 hrs post-inoculation.
A 6 mm biopsy punch was used to collect two intestinal mucosal samples from each loop of Peyer's patch for bacteriology. Intestinal biopsy samples were washed three times in PBS, weighed, placed in the Whirl-Pak™ bag, homogenized in a Colworth-Stomacher blender in PBS, and serially diluted. The tissue extracts were plated onto Herrold egg yolk media containing amphotericin, nalidixic acid, vancomycin (Becton Dickinson and Company, Sparks, MD) with or without Mycobactin J and incubated at 37°C. The cultures were observed visually weekly for any contamination, and the final counts of colony forming units were recorded on week 16.
Extraction and Quality Analysis of RNA
A 6 mm biopsy punch was used to collect 8 intestinal mucosal samples (at each time) of Peyer's patch from the excised PBS control and MAP infected loops for the extraction of RNA at 0.5 1, 2, 4, 8, and 12 hrs post-infection. The tissue was immediately minced with a scalpel blade and transferred to TRI Reagent™ (Molecular Research Center, Cincinnati, OH). Two biopsy punches (approximately 0.1 mg of tissue) were placed into 0.5 ml of TRI reagent™. Tissues were further disrupted with hand-held mechanical tissue grinder equipped with a RNase, DNase free plastic disposable pestle. The RNA extraction was done using the recommended protocol from the manufacturer (Molecular Research Center, Cincinnati, OH). The resultant RNA pellet was re-suspended in DEPC-treated water (Ambion, Austin, TX). Genomic DNA was removed by RNase-free DNase I treatment (DNA-free, Ambion) according to the manufacturer's instructions, and samples were stored at −80°C until used. RNA concentration was quantified by measuring absorbance at λ260 nm using a NanoDrop® ND-1000 (NanoDrop, Wilmington, DW). RNA quality was evaluated by measuring ratio of absorbance at λ260 nm to absorbance at λ280 nm, agarose gel electrophoresis, and using a Nano-Chip® on an Agilent 2100 Bioanalyzer (Agilent, Palo Alto, CA). All the RNA samples used in this study were of good to excellent quality (results not shown), and had distinct 18S and 28S rRNA peaks and RNA size distribution. Ten µg of this high quality experimental RNA was reverse transcribed and used to create indirectly labeled Cy5 cDNA as starting material for each microarray.
Bovine Reference RNA
Bovine reference RNA was prepared in-house and constituted of equal proportions of total RNA from Madin-Darby bovine kidney and bovine B lymphocyte cell lines, and fresh bovine brain cortex and cerebellum , . This reference RNA has been shown to hybridize to the great majority of the open reading frames (ORFs) represented on the microarray. The reference RNA was treated in the same way as the experimental and control RNA for the co-hybridization with each sample on the microarray.
Bovine Microarrays, Sample Preparation and Hybridization
Bovine microarrays were obtained from the W. M. Keck Center (University of Illinois at Urbana-Champaign). These custom microarrays consisted of 70-mer oligonucleotides representing 13,258 unique oligos with 12,220 cattle ORFs. A detailed description of the design and development of microarray has been published elsewhere . Labeling of cDNA and hybridization to microarray have been described previously . Briefly, 10 µg of RNA was reverse transcribed using Superscript III reverse transcriptase (Invitrogen, Carlsbad, CA) and labeled with amino-allyl-UTP (Ambion, Austin, TX). Cy3 and Cy5 dye esters were covalently linked to the amino-allyl group by incubating the samples with the dye esters in 0.1 M sodium carbonate buffer. cDNA from bovine experimental samples (i.e. from MAP infected and PBS control loops) were labeled with Cy5 and co-hybridized against Cy3 labeled cDNA generated from the bovine reference RNA sample 13K bovine 70-mer oligonucleotides array. Prior to hybridization, the microarrays were denatured by exposing to steam from boiling water for three seconds, cross-linked ultraviolet light and then immersed in pre-hybridization buffer (5× sodium chloride, sodium citrate buffer (SSC), 0.1% sodium dodecyl sulfate (SDS) (Ambion, Austin, TX), 1% bovine serum albumin (BSA) at 42°C for a minimum of 45 min followed by four washes in RNase-, DNase-free, distilled water, immersion in 100% isopropanol for 10 seconds, and dried by centrifugation. Slides were hybridized at 42°C for approximately 40 hrs in a dark humid chamber (Corning, Corning, NY) and washed for 10 min at 42°C with low stringency buffer (1×SSC, 0.2% SDS), followed by two 5 min washes in a higher stringency buffer (0.1×SSC, 0.2% SDS and 0.1×SSC) at room temperature in the dark with mild agitation.
Data Acquisition and Normalization of Hybridized Spots on Microarrays
Slides were scanned using a GenePix 4100 laser scanner (Axon Instruments Inc., Foster City, CA). The spots with fluorescent signal representing genes on the arrays were adjusted for background and normalized to internal controls using image analysis software. If the fluorescent signal of any spot was below the background, they were disregarded in all analyses. An average of 56% of spots in experimental samples (Cy5 channel; laser excitation = 635), 77% of spots in reference samples (Cy3 channel; laser excitation = 532), and 54% of spots in the combined samples (Cy5/Cy3 channels) had a signal to noise ratio greater than three.
Microarray Data Analysis
Microarray data analysis was performed by two different methods. In the first method, arrays were normalized by scaling against the average reference intensity value (i.e., average across all microarrays), normalized by the global mean, and then log transformed before statistical analyses were performed. Signals flagged as “non-acceptable" by GenePix (Axon Instruments Inc. Foster City, CA) were removed across all arrays in order to ensure that subsequent analyses for each time point were comparable. Pairwise comparisons of averaged signal values and Student's t test were performed using GeneSifter software (VizX Labs, Seattle, WA). Normalization of each sample was performed against the bovine reference RNA signals across slides and within each slide (across the duplicate spots). Before normalization, duplicate spots were separated and treated as technical replicates. The two spots representing a single gene were therefore required to “agree," based on subsequent analysis steps. A fold-change of at least 1.5-fold and P<0.05 was expected for a difference in signal to be considered statistically significant. All possible individual pairwise comparisons between controls and infections were also performed using Spotfire DecisionSite software (Spotfire, Inc., Somerville, MA). Genes were further filtered using these various comparisons in order to ensure biological relevance (i.e., that observed differences were not the result of random variation between uninfected animals) and consistency (i.e., reproducibility across experiments).
In the second method, more recent computational tools termed the BioSignature Discovery System (BioSignatureDS) (Seralogix, LLC, Austin, TX) were employed to conduct comparative pathogenicity analysis and modeling. This approach for genomic data analysis and modeling at the system biology level offers an integrated view of biological mechanisms and networks of interactions. Specifically for the analysis reported herein, the tools were used to: 1) determine significant gene modulations via a z-score sliding window threshold technique and fold change; 2) conduct biological system level analysis employing Bayesian network models for scoring and ranking of metabolic pathways, signaling pathways and gene ontology (GO) groups; 3) conduct Bayesian candidate mechanistic gene analysis to identify genes within the network models that are most responsible for causing pathway and GO group perturbations; and 4) create a genetic network system model derived from the candidate mechanistic genes and their genetic interactions. More detailed description of the computational techniques employed by BioSignatureDS was described in our previous publication .
Microarray Data Deposited in the Gene Expression Omnibus
The microarray data were deposited in the Gene Expression Omnibus at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/geo/) Accession # GSE13888.
Induction of RNAi in HeLa Cells
HeLa cells (ATCC CCL-2.2, from ATCC, Manassas, VA) were cultured in F12K medium supplemented with L-glutamine and 10% heat inactivated fetal bovine serum (HI-FBS). On day 1, approximately 4×104 HeLa cells were placed into 24 well cell culture Plates (Corning, Corning, NY) in 0.4 ml of culture medium and placed in a 37°C incubator with 5% CO2. The following day, for each 24-well culture to be transfected, 50 µl of serum-free cell growth medium (OPTI-MEM, Invitrogen, Carlsbad, CA) was mixed in separate compartments with Silencer® validated siRNAs (Ambion, Austin, TX). The siRNAs were used at final concentrations according to the manufacturer's protocol: mitogen-activated protein kinase 1 (MAPK1) (ID 1449) 30 nM, MAPK1 (ID 1544) 100 nM, epidermal growth factor EGF (ID 645) 50 nM, and negative control siRNA (AM4635) 50 nM. Simultaneously, 1 µl of TransFectin® lipid reagent (Bio-Rad, Hercules, CA) was diluted into 50 µl of serum-free cell growth medium for each 24-well culture to be transfected. The diluted siRNA was combined and mixed with the diluted TransFectin® reagent. Additional negative controls consisted of cells to which either 100 µl of OPTI-MEM or 100 µl of OPTI-MEM with 1 µl of TransFectin® were added. After 20 min incubation at RT, 100 µl of the siRNA-TransFectin mixture were added to the 400 µl of F12K cell culture media on the cells. On day 3, one ml of F12K medium with 10% FBS was added to the each well. On day 4, transfected cells were infected with MAP at a multiplicity of infection (MOI) of 10 bacteria per HeLa cell, and bacterial invasion was determined as described below. For validation of RNAi efficiency, RNA from transfected cells was extracted at the same time of infection (i.e. 48 hrs post-transfection) using Tri Reagent (MRC, Cincinnati, OH) according to the manufacturer's protocol. Contaminant genomic DNA was removed by RNase-free DNase I treatment (Ambion) according to the manufacturer's instructions, and samples were stored at −80°C until used. RNA concentration was quantitated by NanoDrop® ND-1000 (NanoDrop, Wilmington, DW). Target mRNA levels were measured by qRT-PCR as previously described for microarray validation .
Functional Assessment of siRNA Transfected HeLa Cells For Invasion of MAP
HeLa cells were cultured as described above. Before adding MAP, the HeLa cells from 2 wells were detached and counted. Invasion assays were performed by removing 1.2 ml of the medium overlying the HeLa cells monolayers and adding 100 µl of a bacterial inoculum re-suspended in cell culture media, at a MOI of 10:1. Bacteria were centrifuged onto the cells at 800×g for 2 min followed by 1 hr of incubation at 37°C. Then, cells were washed three times with PBS to remove extracellular bacteria and re-incubated with F12K media supplemented with 100 µg ml−1 of gentamicin solution (Sigma, St. Louis, MO) for 2 hrs. After antibiotic treatment, infected cultures were washed three times with PBS and then lysed with 0.1% Triton X-100 (Sigma). Lysates were serially diluted and cultured on Herrold egg yolk media supplemented with Mycobactin J and amphotericin, nalidixic acid, and vancomycin for quantification of colony-forming units (CFU). Duplicate wells were used for each experiment, and the experiments were performed three times.
Kinetics of Trans-Epithelial Resistance of Polarized Epithelial Cells During Invasion of pathogen
T84 cells (ATCC 14028, from ATCC, Manassas, VA) were grown in DMEM/F12 medium (Gibco, Life Technologies) containing 1.2 g of sodium bicarbonate per liter, 2.5 mM L-glutamine, 15 mM HEPES, and 0.5 mM sodium pyruvate (Gibco, Life Technologies) supplemented with 10% fetal calf serum. T84 cells were polarized by seeding 4×105cells/well on the apical compartment of 12-mm-diameter Transwell plates (polycarbonate membrane with a pore size of 0.4 µm; Corning Costar) and 1.5 ml of media was added to the basolateral compartment. The medium was changed every other day, and the transepithelial electrical resistance (TER) was measured after 7–8 days. After the cells reached a TER of at least1,500 µ/cm2, they were incubated overnight in fresh medium, and the invasion assay was performed on the following day using a MOI of 10:1. The change in the TER was measured at various time intervals, starting at 2 mins post-infection to 24 hrs post-infection, with a voltmeter (Millipore-ERS resistance meter; Millipore, Bedford, Mass.).
Statistical Analysis of Bacteriology
Colonization of tissue samples and transfected HeLa cells was considered to be positive when MAP was detected by bacteriological culture. Tissue burden was defined as the number of colony forming units per milligram of tissue. The statistical significance of differences was calculated using two-tailed Student's t test.
Invasion of Ileal Mucosa by MAP
MAP was recovered from the MAP-inoculated ileal tissues at all the time points post-infection (data not shown). No bacteria were detected in the PBS inoculated loops. Among the infected loops, no significant changes in the number of MAP were detected at any times post-inoculation (0.5–12 hrs).
Bovine Peyer's Patches Inoculated with MAP Reveals a Complex Temporal Pattern of Transcriptional Profile
In order to gain detailed insight into the changes in the transcriptional profile of genes in bovine intestinal Peyer's patch mucosa inoculated with 3×109 cfu of MAP (30, 60, 120, 240, 480, and 720 min post-infection), initially the microarray data analysis was performed by using GeneSifter software (where a fold-change of at least 1.5-fold and P<0.05 was required for a difference in signal to be considered statistically significant, Table 1). Classical analysis of the altered gene expression by GeneSifter, provides the static changes in the experimental conditions with a profound spectrum of data; however, after filtering the data into the biological relevant and significant genes, a limited number of genes were found to have statistically significant expression.
As shown in Table 1, no genes were significantly up-regulated in MAP-infected animals at the earliest time point tested (30 min), and only modest numbers of genes (less than 30 by 8 hr) were increased in expression over the experimental time course. In contrast, at the earliest time points, MAP-infected loops had down-regulation in gene expression. This observation clearly reflected that measurement of subtle changes may be undetected if: 1) the quantity of the change is very small to measure, or 2) if experiments are conducted using out bred populations, or 3) if the sample used is a heterogeneous population of several cell types expressing different levels of expression of same gene. To overcome these challenges, we extended the analysis to create a temporal relationship of various genes using BiosignatureDS tools that employed a method termed the Dynamic Bayesian Gene Group Activation (DBGGA)(Seralogix, Austin, TX). Bayesian network models were created for all the known signaling and metabolic pathways and for Gene Ontology (GO) biological processes. The models were trained with the control group data (uninfected) with the experimental data (infected) used as evidence to test how different experimental data are applied in fitting the control model. This difference is determined by measuring the negative log-likelihood that, in turn, was transformed to a z-score test statistics that is referred to, here after, as the Bayesian z-score. This method ranked groups of genes at each time point and across all time points to determine differences between experimental conditions , . Similarly, how well the results of individual genes fit a model was also determined, producing Bayesian z-scores for each gene within a pathway or GO category. This method used a less stringent spot quality filtering technique; a more sophisticated universal reference normalization method in conjunction with Lowess correction; and a Bayesian variance estimator that infers a better prediction of the standard deviation for genes which have a low number of biological replicates , . Using this approach, the contribution of small changes in key regulatory genes was taken into account.
Classification of Perturbed Pathways using Bayesian z-scoring Divides the Host Transcriptional Response into Three Phases
As indicated by the host system level pathway analyses that identified significantly perturbed pathways over the experimental time course, there were three logical classification phases (Early, Intermediate and Late) in which it is proposed that invasion occurs in the Early Phase and longer term evasion and host immune tolerance occurs in all three phases. The Early Phase of infection consisted of the 30 and 60 min time periods post-infection. The Intermediate Phase consisted of the 120, 240, and 480 min time periods post-infection, and the Late Phase infection at 720 minutes post-infection. A comprehensive Bayesian z-score list of significantly perturbed pathways identified for MAP inoculated loop, passing the 97.5% confidence threshold is provided in Tables 2, 3, and 4, organized into Early, Intermediate, and Late Phase responses, respectively. The rank order of pathways in Tables 2, 3 and 4 are based on the Bayesian z-scores ranging from largest to smallest for the first time point in each of the phases (i.e., t = 30 minutes for Early, t = 120 minutes for Intermediate, and t = 720 minutes for Late). These tables organize the pathways in terms of their state of activation or suppression. Of 220 pathways scored by the DBGGA method, the Early Phase response had 82 significantly perturbed signaling and metabolic pathways, the Intermediate Phase had 70, and the Late Phase had 117. There were 30 pathways that were significantly perturbed in common to all three phases and are highlighted in italics in Tables 2, 3 and 4. The Early Phase had 23 pathways that were uniquely perturbed compared to the other phases while the Intermediate Phase had only 11, and the Late Phase had 45 uniquely perturbed pathways. These uniquely perturbed pathways are indicated in these tables with an “*" before the name of the pathway. As a supplement to Tables 2, 3 and 4, a heat map of all pathway scores is provided in Figure S1 to better visualize the temporal patterns and the degree of perturbation at each time point post infection.
System Level Pathway Results and Immune Response Phases
There were 30 common pathways significantly perturbed in all three phases. These common pathways may be important to both short term and long term host tolerance to MAP. In the “common" pathways, there were several pathways involved in the host immune response including: Complement and Coagulation Cascades indicating a non-specific defense mechanism; Hematopoietic Cell Lineage indicating immune cell differentiation; CD40L Signaling indicating T cell activation; Cytokine-Cytokine Receptor indicating immune cell communication, PPAR Signaling indicating inflammatory response of immune cells; and Toll-Like Receptor Signaling that signifies triggering the innate immune response. These immune related pathways all had basically strong activation as shown in Table 2 for the Early Phase in Figure 1.
The darker red gradients indicate higher activation scores (more up-regulated gene expression within the pathway) while the darker green gradients indicate more suppressed pathway activity (more down-regulated gene expression) of MAP infected Peyer's patch. The pathway threshold score was selected for a 97.5% confidence.
A few pathways that were apparently being manipulated by MAP infection showed a reversal from suppression in the Early Phase to activation in the Late Phase that include Cell Communication, One Carbon Pool by Folate, and Long-term Potentiation while the Microtubule-associated Protein 1 pathway reversed from activated to a suppressed state (Figure 1). Other pathways that were activated in all three phases include Complement and Coagulation Cascade pathway, Adipocytokine Signaling Pathway, Hematopoietic Cell Lineage Pathway, and Neuroactive ligand-receptor interaction. The genes and networks involved in these pathways are discussed in more detail in the Discussion section.
There are 23 pathways that are uniquely perturbed during the Early Phase of MAP infection and 26 non-unique pathways that are in common with either the Intermediate or Late Phases that include a number of immune and metabolic pathways as annotated in Tables 2, 3 and 4. These included for example, Tyrosine Metabolism, Histidine Metabolism, Phosphatidylinositol Signaling System, Tryptophan Metabolism, Selenoamino Acid Metabolism, Androgen and Estrogen Metabolism, Glycosphingolipid Biosynthesis, Aminophosphonate Metabolism, Glycerophospholipid Metabolism, GnRH Signaling Pathway, Ether Lipid Metabolishm, and Glycolysis/Gluconeogenesis.
The temporal perturbation of these pathways illustrates the complexity of MAP's pathogenicity in the host. The functional roles of these pathways with regard to host invasion and evasion are presented in more detail In the Discussion section. The analysis resulted in the development of a biological systems level model.
Gene Ontology (GO) Biological Process Results
The DBGGA scoring method was applied to 2,254 GO biological process categories. Each category had to contain at least 10 measured genes to be scored. For GO scoring, the DBGGA method employs a naive Bayesian classifier network model. Scoring results indicated that the Early Phase response had 467 significantly perturbed GO categories (Table S1), the Intermediate Phase had 105 (Table S2), and the Late Phase had 691 (Table S3). There were 27 GO categories that were significantly perturbed in common to all three phases. In the Early Phase there were 272 uniquely perturbed GO categories while the Intermediate Phase had 31 and the Late Phase had 459 that were uniquely perturbed. There is a broad range of strongly activated and suppressed biological processes. Several GO categories of interest were selected from the Early Phase for heat map plotting shown in Figure 2. Of the top 50 GO activated categories, early Phase activation may be associated with MAP host invasion and early immune defense processes. In contrast, of the top 50 suppressed GO categories, the suppressed categories may be a result of MAP's manipulation of these host's processes to facilitate invasion into the host cells and to subvert the host's immune defenses.
The darker red gradients indicate higher activation scores (more up-regulated gene expression within the pathway) while the darker green gradients indicate more suppressed GO activity (more down-regulated gene expression) of MAP infected Peyer's patch. The pathway threshold score was selected for a 97.5% confidence.
Several activated GO categories of interest include processes involved in: 1) early to late endosome transport; 2) cellular calcium ion homeostasis; 3) induction of apoptosis; 4) positive regulation of NF-κB transcription factor activity; 5) inactivation of MAPK activity; 6) vacuolar transport; 7) Wnt receptor signaling; 8) actin filament bundle formation; 9) defense response to Gram-positive bacterium; and 10) activation of phospholipase C activity. These categories all had an activated state in the Early Phase of the immune response (Figure 2). Interestingly, there were six GO categories that reversed from activated state in the Early Phase to a suppressed state in the Late Phase that included: 1) positive regulation of vasodilation; 2) T-helper 1 type immune response; 3) complement activation, alternative pathway; 4) regulation of Rab GTPase activity; 5) response to toxin; and 6) activation of transmembrane receptor protein tyrosine kinase activity.
Several interesting GO categories that were suppressed in the Early Phase and then reversed to an active state in the Late Phase include processes involved in: 1) I-κB kinase/NF-κB cascade; 2) actin cytoskeleton reorganization; 3) Wnt receptor signaling pathway through beta-catenin; 4) T cell differentiation; 5) microtubule-based movement; 6) positive regulation of endocytosis; 7) positive regulation of apoptosis; 8) cytoskeletal anchoring at plasma membrane; 9) response to calcium ion; 10) negative regulation of NF-κB transcription factor activity; and 11) negative regulation of axon extension. Other categories that were suppressed in the Early Phase and remained suppressed or neutral included: 1) Rho protein signal transduction; 2) positive regulation of Wnt receptor signaling pathway; 3) T cell proliferation; 4) regulation of actin filament polymerization; 5) cell-cell junction assembly; 6) post-Golgi vesicle-mediated transport; and 7) regulation of lipoprotein lipase activity.
There were 554 GO categories that were strongly activated in the Late Phase (Table S3). These activated categories provided evidence of the host mounting a more effective immune response that included: 1) neutrophil activation; 2) positive regulation of inflammatory response; 3) innate immune response; 4) positive regulation of B cell proliferation; 5) positive regulation of T cell mediated immunity; 6) endothelial cell proliferation; 7) positive regulation of cytokine production; and 8) positive regulation of cell adhesion. In contrast, there were only 130 GO categories that were significantly suppressed in the Late Phase that included: 1) immunoglobulin mediated immune response; and 2) natural killer cell activation.
Discovery of Mechanistic Genes during MAP Infection Reveals Importance of the Cross-Talk via Inter-Pathway Interactions
The DBGGA analysis identifies mechanistic genes by Bayesian modeling of the genes in the context of their upstream and downstream relationships over the complete time course. Mechanistic genes significantly influenced disease progression and contributed most to the discrepancies between the MAP infected vs. PBS control tissue. We propose that these mechanistic genes may play an important role in the outcome of the host-pathogen interaction. A complete list of all the mechanistic genes is provided in the Table S4. We further focused on only 43 pathways involved in signaling and immune response. Our intent was to identify mechanistic genes that are associated in multiple pathways which may be the source of cross-talk and thus have more significant influence governing the host immune tolerance to MAP and illustrating the importance of cross-talk. Of the 43 pathways analyzed, 36 pathways had at least one overlapping mechanistic gene. It was found that 141 mechanistic genes had overlaps within the 36 pathways examined. These genes are listed in Table S5. Of highest interest were those genes that had influence (overlap) across numerous pathways. The mechanistic gene, AKT2 (v-akt murine thymoma viral oncogene homolog 2), had overlap with 11 important pathways that include: 1) Adipocytokine signaling; 2) Insulin signaling; 3) Fc epsilon RI signaling; 4) T cell receptor signaling; 5) Toll-like receptor signaling; 6) Tight junction; 7) Integrin-mediated cell adhesion; 8) VEGF signaling; 9) mTOR signaling; 10) ErbB signaling; and 11) MAPK signaling. These pathways are involved in a wide variety of biological processes including, but not limited to cell proliferation, differentiation, apoptosis, tumorogenesis, as well as glycogen synthesis and glucose uptake. The protein encoded by AKT2 is a member of the AKT, also called PKB, serine/threonine protein kinase family. AKT kinases play a key role in regulating cell survival, insulin signaling, angiogenesis and tumor formation. Only 14 other mechanistic genes had overlap with 6 or more pathways as listed in Table S5. Several cytokines (IL1a, IL1ß, IL4R, IL5, IL7, IL15, IL23A, IFN-γ) were also overlapping mechanistic genes in analyzed pathways. These genes will be addressed in more detail in the Discussion section.
MAP has the uncanny ability to persist within the host for an indefinite period of time that can last several years. Hence, MAP must have efficient host invasion and host immune evasion processes that should be evident by MAP's manipulation of certain host immune response and metabolic pathways. We utilized the perinatal calf ligated jejunal-ileal loop model to study the sequential changes in the host intestine immediately after infection with MAP. A key role of intestinal mucosal epithelia is barrier function, which prevents colonization or invasion by foreign microorganisms. However, in Johne's disease, MAP invade M cells, enterocytes, dendritic cells and macrophages, and are capable of resisting host defenses and multiply to reach very high intracellular numbers leading to chronic granulomatous lesions , . In infected subclinical and clinically affected animals, systemic immune response is achieved. Persistence of the organism in the intestinal Peyer's patch in the presence of a systemic immune response suggests that the immune response in the intestine may be fundamentally different from the systemic response. In fact, a state of immune tolerance was detected at the mucosal level during subclinical Johne's disease . Furthermore, it has been shown earlier that the human intestinal macrophages display profound inflammatory anergy despite avid phagocytic and bacteriological activity . The aim of our study was to discover if the immune tolerance is initiated, and if so, how quickly after the pathogen comes in contact with the intestinal mucosa. We hypothesized from a biological system perspective that MAP pathogenicity should show evidence of: 1) host invasion by manipulating host cellular functions related to mucosal immune barrier; and 2) subversion of host immune response that permits MAP uptake, survival and proliferation.
Host Invasion through Compromising the Mucosal Immune Barrier
In the pathway scores listed in Tables 2, 3 and 4, there are several suppressed pathways that may be associated with MAP host invasion by impeding mucosal epithelial barrier function that include Cell Communication (CC), Tight Junction (TJ), Integrin-mediated Cell Adhesion (IMCA), and Trefoil Factors Initiated Mucosal Healing (TFIMH) pathways. A key observation is the suppressed state of the Cell Communication Pathway, which interestingly was suppressed in the Early Phase and became activated in the Late Phase (Figure 1). The CC pathway includes the genes from the TJ, IMCA as well as the Gap Junction (GJ) and Adherens Junction (AJ) pathways. These pathways form the intercellular junction complexes between adjacent intestinal epithelial cells that are critical components of the intestinal mucosal barrier that creates a semi-permeable diffusion barrier. Studies have shown that activation (increased gene expressions) of these junction pathways may lead to strengthening the intestinal barrier while suppression (disrupted gene expression) may result in weakening of the immune barrier. As shown in the heat map scores of Figure 3(a), the AJ, TJ, and TFIMH pathways are suppressed in the Early Phase while the state of the Gap Junction pathway was activated. This suggests that MAP host invasion may be disrupting critical cell communication processes in a complex manner. This complex nature of cell disruption was also analyzed by measuring the Trans-Epithelial Resistance (TER) of an in vitro model polarized epithelial cells during MAP interaction (Figure 3(b)). MAP infection caused a marked decrease in the TER, adding credibility that increased permeability of in vivo host intestinal epithelium may facilitate bacterial invasion through the intestinal epithelium.
The set of genes within the gap junction, tight junction, adherens junction and integrin-mediated cell adhesion are contained in the cell communication pathway. (3a). Pathway scores for Adherens Junction (AJ), Cell Communication (CC), Gap Junction (GJ), Integrin-mediated Cell Adhesion (IMCA), Tight Junction (TJ) and Trefoil Factors Initiated Mucosal Healing (TFIMH). Red indicates an activated state while green indicates suppression. Note the reversal of the cell communication pathway from suppressed to activated from the Phase 1 (30–60 min.) to the Phase 3 time period (720 min.). (3b). Changes in the Trans-Epithelial Resistance of polarized T-84 cells exposed to MAP for 24 h. Data represent mean ± SD from 3 measurements at each time point and three independent experiments.
Cell adhesion serves to facilitate trafficking and migration of T lymphocytes into sites of inflammation, movement of lymphocytes within the rich environment found in extravascular tissue, and the physical interaction between antigen-reactive T cells and antigen-presenting cells that is required for efficient T-cell activation . As shown in Figure 3(a), the IMCA and TFIMH pathways are suppressed in the early and Late Phases which suggest that MAP may disrupt T lymphocyte recruitment that helps explain the lack of chronic inflammation observed in the MAP infected ileal loops and subvert mucosal healing. Over time the trend is for the TJ, IMCA and TFIMH pathways to remain suppressed, but GJ and AJ pathways become activated. This suggests that MAP may need to suppress important host cell communication, adhesion and healing processes for penetrating the mucosal immune barrier, but activate cell adhesion mechanisms for longer term survival in the Late Phase. It has been proposed by others  that some bacteria survival mechanism in mucosal epithelial cells is for the bacteria to hijack integrin-linked kinase to stabilize focal adhesions and block cell detachment of infected cells. The rapid turnover and exfoliation of mucosal epithelial cells provides an innate defense system against bacterial infection. Nevertheless, bacteria such as MAP may be able to subvert this immune defense mechanism and colonize the epithelium efficiently and survive. Furthermore, M cells are unique among cells of the intestinal epithelium as they display a high density of Beta1 integrins on their luminal surface. In a recent study, we documented that the early host response was evident by the presence of MAP in the vicinity of M cells and goblet cells . Integrins have affinity for the fibronectin attachment protein of MAP. Thus, M cells are thought to play a role in the host defense by down-regulating integrins and thus avoiding the fibronectin bridge formation for the entry of MAP into the ileal mucosa .
Junction (Gap, Tight, Adherens) Pathways.
The junction related mechanistic genes (significant differential gene expressions determined by DBGGA analysis) are shown in the heat map of Figure 4. The key down-regulated genes of high interest in the Early Phase include: MAPK1, CTNNB1, ERBB2, PARD3 ACTN2, CLDN7, ACTB, CSNK2B, CSNK2B, GNAI3, MAP2K1, TCF7L1, SRC and whose biological roles are described in Table 5. Many of these genes are involved with maintaining the integrity of the epithelial layer. According to the Adherens Junction Bayesian network model (not shown), SRC has strong correlated relationships with other downstream genes, i.e., the gene relationship SRC->RAC1. RAC1 (ras-related C3 botulinum toxin substrate 1 rho family small GTP binding protein Rac1) gene expression is suppressed across all three phases. RAC1 encodes a GTPase protein belonging to the RAS superfamily of small GTP-binding proteins that regulate a diverse array of cellular events including the control of cell growth, cytoskeletal reorganization, and the activation of protein kinases.
The heat map shows a dominance of down regulated junction related gene expression occurring in the Early Phase as indicated by the darker green boxes. Genes listed surpassed the |Bayesian z-score|>2.24 at any of the time points. Red indicates up regulation while green indicates down regulation. Time is minutes post-infection.
Cell Adhesion Molecules (CAM) and Integrin-Mediated Cell Adhesion (IMCA) Pathway.
The impairment of cell adhesion may be an important mechanism for MAP invasion in the Early Phase as evident by the IMCA pathway suppression, while the strong Late Phase activation of CAM pathway may be a MAP survival mechanism which prevents infected cell detachment. To explore this in more detail, the gene level activities in the Early, Intermediate, and Late Phases were examined (Figure 5). An important gene, CDH5 (cadherin 1, type 1, E-cadherin (epithelial)), of epithelial cells that form an adhesion point for many subtypes of lymphocytes including intraepithelial lymphocytes lacked expression or was slightly down-regulated in all phases. Integrins function in neutrophil adherence but the majority of integrins was down regulated or not expressed in the Early Phase. The Intermediate and Late Phases had a greater number of up regulated integrins that may support the strengthening of the immune barrier. In the CAM and IMCA pathway there were nine strongly down-regulated genes in the Early Phase that supports impaired cell adhesion, i.e. mucosal barrier weakening. These genes include ITGB1, PTK2, MAP2K1, SELL, MAPK1, Mpzl1, CD99, ITGA4 and CLDN7 and are described in Table 6. Note that CLDN7 was described above as an integral membrane protein and component of tight junction as was the role of MAPK1. The key Intermediate and Late Phase up-regulated genes in the CAM and IMCA pathways, in support of MAP survival mechanism (mucosal barrier strengthening) are PDPK1, CNTN1, NRXN3, SPN, CSPG2, HLA-DOB, SELP, PTPRC, ITGAM, TLN1, NCAM1, and RHOC. Detailed description of these genes is provided in Table 7.
The heat map shows a dominance of higher perturbed cell adhesion gene expression (both up- and down-regulated) occurring in the Early Phase as indicated by the darker red and green boxes for time t = 30 and t = 60 minutes. Genes listed surpassed the |Bayesian z-score|>2.24 at any of the time points. Red indicates an activated state while green indicates suppression. Time is minutes post-infection.
Trefoil Factors Initiated Mucosal Healing (TFIMH) Pathway.
Epithelial continuity can also depend on a family of small, yet abundant, secreted proteins–the trefoil factors. The immune related TFIMH pathway is suppressed in the Early Phase (Figure 3a). The trefoil factors maintain the integrity of the gastrointestinal tract, despite the continual presence of microbial flora and injurious agents . Unfortunately, the trefoil factors gene probes were not included on the bovine microarray employed during this study. However, the TFIMH pathway suppression (as determined by other observed gene expressions) could imply impaired trefoil factors gene expression, and consequently, a possible invasion mechanism of MAP by subverting mucosal healing. Genes that dominate the suppressed pathway activity are PTK2, ITGB1, MAPK1 and CTNNB1. The biological roles of these genes are described in Table 8.
Subversion of Host Immune Response Processes
Host Cellular Uptake of MAP and Phagocytosis Arrest.
A new perspective in the pathogenesis on mycobacterial diseases (M. tuberculosis) is the exploitation of host cell signaling pathways by the pathogen. Upon infection, the phosphatases and kinases of several pathogenic bacteria modify host proteins and help in the establishment of the disease. The uptake of M. tuberculosis by macrophages is associated with a number of Early Phase signaling events, such as the recruitment and activation of members of the Src family of protein tyrosine kinases . These kinases result in the increased tyrosine phosphorylation of multiple macrophage proteins and the activation of phospholipase D . Phospholipase products have been linked to phagocytosis mechanisms of bacteria uptake . Examination of the pathways that include CSK (c-src tyrosine kinase) indicated that this gene is significantly up-regulated in the Early Phase and transitioned to a moderately down-regulated state in the Intermediate and Late Phases. CSK is associated with the Regulation of Actin Cytoskeleton, Epithelial Cell Signaling, Integrin-mediated Cell Adhesion, and Activation of Csk Through T-Cell Receptor pathways, all of these pathways were highly activated in the Early Phase and transitioned to suppressed states in the Intermediate and Late Phases. In this study, several classes of phospholipases were significantly up-regulated in the Early Phase that included PLA2G1B (phospholipase A2, group IB (pancreas)), PLCD1 (phosholipase C, delta 1), PLCB4 (phospholipase C, beta 4), and PLD1 (phospholipase D1, phosphatidylcholine-specific). Table 9 lists the pathways in which these genes are considered mechanistic. Phospholipases are a group of enzymes that hydrolyze phospholipids into fatty acids and other lipophilic molecules and have been implicated in numerous cellular pathways, including signal transduction, membrane trafficking, and the regulation of mitosis. Elevated levels of phospholipases have been linked to intracellular calcium elevations during bacteria invasion .
It has been shown that M. tuberculosis is able to hi-jack lipid metabolism to drive the progression of the disease , , . The Phosphatidylinosital Signaling System (PSS) is of interest, because phosphatidylinositol lipids have been identified as key signaling mediators for random cell migration as well as chemoattractant-induced directional migration. The PSS was initially highly activated and trended to be suppressed in the Late Phase.
Phosphatidylinosital Signaling System (PSS) Pathway.
The significantly up-regulated genes involved in this signaling event were PLCD1, PLCB4, INPP4A, ITPR2, ITPR3. The genes, PLCD1 and PLCB4 genes encode phospholipases that are ubiquitously expressed and have diverse biological functions including roles in inflammation, cell growth, signaling and death and maintenance of membrane phospholipids. Significantly down-regulated in all immune response phases in PSS is the gene CALM2 (calmodulin 2) that is known to mediate the control of a large number of enzymes and other proteins by Ca++. The biological roles of these genes are described in Table 10.
It has been observed elsewhere  that pathogenic mycobacteria (human macrophages infected with Mycobacterium avium subsp. hominissuis) have been shown to interfere with Ca++ and PI3K signaling pathways which are essential pathways for phagosomal maturation that requires CALM2 activation. The CALM2 gene expression data, from the referenced human macrophage study, was consistently down-regulated at all measured time points along with markedly reduced STX3 (syntaxin 3) expression.
Microtubule-Associated Protein 1 (M-AP1) Pathway.
Syntaxins are included in the M-AP1 pathway that was highly activated in the Early Phase. The protein encoded by STX3 is a member of the syntaxin family of cellular receptors for transport vesicles that participate in exocytosis in neutrophils. Other members of the syntaxin family have been associated with M. tuberculosis phagosome maturation arrest . This pathway reversed from a highly activated state to a highly suppressed state in Late Phase (Figure 1). It has been observed in murine macrophages that mycobacteria arrest the maturation of the early endosome to a phagolysosome by inhibiting fusion of the mycobacterium-containing phagosome with lysosomes –. The M-AP1 pathway activation reversal may suggest an important mechanism for MAP host immune evasion. In the M-AP1 pathway genes, SNAP23 and Vamp2, were highly up-regulated, while the genes, Vti1a and YKT6, were strongly down-regulated in the Early Phase. In the Late Phase, there were five strongly down-regulated genes that dominate the suppression of M-AP1 pathway. These down-regulated mechanistic genes include Vti1b, STX8, STX10, YKT6, STX6 and GOSR2. The STX genes are members of the syntaxin family involved in protein trafficking from early to late endosomes via vesicle fusion and exocytosis. The biological roles of these genes are described in Table 11.
Calcium Signaling (CS) Pathway.
The CS pathway was strongly activated in all three phases suggesting MAP infection has influence on this process during invasion and possibly related to MAP survival long term. Calcium signaling plays an important role in a broad range of regulatory effects on enzymes and proteins and influence on other major pathways including MAPK Signaling, Apoptosis, Long-term Potentiation, Long-term Depression, Phosphatidylinositol Signaling and others. Across all three immune phases, there were three significantly up-regulated genes that include NFATC4, CAMK2A, and PLCB3 while there were another four genes that are significantly up-regulated only in the Early Phase that include ADCY8, EDNRB, NFKBIB, and TACR2. There were four significantly down-regulated genes that included CALM2, PPID, GNAQ, and ATP2B4. The biological roles of these genes are described in Table 12. NFATC4 plays a role in the inducible expression of cytokine genes in T cells, especially in the induction of the IL-2 and IL-4. However, there was no evidence of IL-2 or IL-4 expression in any phase of the host immune response of our study.
Host Immune Tolerance Subversion of Activated Immune Related Pathways.
Defective sensing and killing of bacteria may drive the onset of chronic diseases like Johne's and Crohn's . Although there are signs that the host is sensing the presence of MAP by generating an immune response in all phases, MAP successfully invades and evades the host immune processes. More specifically, the major immune related pathways that were activated in the Early Phase included the Toll-like Receptor Signaling, Hematopoietic Cell Linage, Adipocytokine Signaling Pathway, CD40L Signaling, Wnt Signaling, Cytokine-Cytokine Receptor Interactions, Complement and Coagulation Cascades, and Lectin Induced Complement pathways. Further examination of several of these pathways at the network level provided evidence that MAP was potentially interfering with their immune response functionalities.
Toll-like ReceptorSignaling (TLRS) Pathway Subversion.
With the triggering of the TLRS pathway, it could be presumed that the host had initiated an effective immune response. Examining this pathway at the network and gene expression level indicates that the source of pathway perturbation comes from genes that are both highly up-regulated and down-regulated over the complete time course. Figure 6 shows the Bayesian network for TLRS pathway and the Bayesian z-score gene expression temporal heat map for all genes on this pathway. The toll-like receptor signaling appears defective in that it is not producing the expected expression patterns for proinflammatory cytokines. The key cytokines, IL-1β, TNF, IL-6, and IL-12 are not significantly expressed, although IL-1β is eventually up-regulated in the Late Phase. Also of interest are the chemokines CCL3 (MIP-1α), CCL5(RANTES), CXCL9, CXCL10, and CXCL11 which are not significantly expressed and suggests a potential disruption of monocyte and natural killer cell stimulation and T-cell migration that could explain, in part, the host immune tolerance for MAP. In the early and intermediate phase of MAP invasion, there is significant expression of TLR4, TLR3 and TLR9, but no TLR2 at any phase. TLR4 is expressed on the cell surface of enterocytes and numerous cells of the immune system such as dendritic cells, B lymphocytes and NK cells. On the other hand, MAP can also interact with TLR9 located within the endosomal compartments of phagocytic cells and B lymphocytes and functions to alert the immune system of MAP infections. The lack of TLR2 expression appears contrary to published results for in vitro M. paratuberculosis infected murine macrophages in which it was concluded that TLR2 is one of the key recognition receptors , . This could imply that the in vivo pathogenesis of MAP has differing invasion mechanism than in vitro, or the mechanisms are different between host species. Also there is no significant expression for MYD88 or NFkβ1 until the late phase in which NFkβ1 eventually becomes significantly expressed.
Toll-like receptor signaling pathway Bayesian network representation at 30 min post-infection (left), and the DBGGA scores for the gene expression of MAP infected host Peyer's patch versus non-infected controls (right). Gene nodes with orange circles on the network are those defined as mechanistic genes that surpass a threshold |Bayesian z-score|>2.24. The network shows gene nodes with gradient colors representing the level of expression (deeper red for higher up-regulated genes and deeper green for down-regulated). The heat map is colorized and corresponds to the gene node expression levels. Grey color represents little to no expression difference between MAP infected and controls. The heat map columns are by time post-infection in minutes.
Hematopoietic Cell Lineage (HCL) Pathway Subversion.
The activation of the HCL pathway may also be an indicator of host immune response to MAP. The key genes that dominate the activation of HCL pathway are IL-4R, CD14, CD59, GYPA, FLT3 and CSF1R. The biological roles of these genes are described in Table 13. Interleukin-4R is a receptor for both IL-4 and IL-13 and couples to the JAK1/2/3-STAT6 pathway. The IL-4 response is involved in promoting Th2 cell differentiation. CD14 is a surface antigen that is preferentially expressed on monocytes/macrophages. It cooperates with other proteins to mediate the innate immune response to bacterial lipopolysaccharide. CD59 regulates complement-mediated cell lysis, and it is involved in lymphocyte signal transduction and is a potent inhibitor of the complement membrane attack complex while also playing a role in signal transduction pathways in the activation of T cells. GYPA is a major sialoglycoprotein of the erythrocyte membrane. Interestingly this protein has been linked to receptor-ligand interactions involved in the invasion of erythrocytes by malarial parasite  and may suggest a similar MAP influence. FLT3 and its ligand FLT3LG play an important role in the immune response by regulating the functions of granulocytes/macrophage. As observed in our study, the FLT3LG gene expression was significantly down-regulated in the early phase and then up-regulated in the late phase. Inhibition of FLT3LG has been shown to significantly impair the immune system, as well as cause a reduction in myeloid progenitor cells. The number of B-cell progenitors, dendritic cells and natural killer cells have been reported to be significantly reduced in in vivo murine studies . This suggests another evasion mechanism of MAP during the Early Phase that subverts the host immune response. CSF1R is the receptor for colony stimulating factor 1, a cytokine that controls the production, differentiation, and function of macrophages. This gene was up-regulated only at 30 minutes post-infection and was down-regulated in the intermediate and late phase suggesting a longer term mechanism of host immune tolerance to MAP. Our study clearly indicates that the host responses to MAP starts immediately after sensing the microbial interaction with the intestinal mucosa that in turn normally releases signals to stimulate recruitment of pro-inflammatory leucocytes, immune cells, or both.
CD40L Signaling Pathway Subversion.
Interaction between CD40L on activated T-cells and CD40 receptors on macrophages is crucial for maintaining a Th1 response and activation of macrophages . The CD40L (ligand) signaling pathway was activated in the Early Phase, tended to be suppressed in the Intermediate Phase and strongly activated in the Late Stage as shown in Figure 1. CD40 relies on interaction with TRAF proteins to mediate an intracellular signal in response to CD40L binding. The pathway downstream of TRAFs activates the transcription factor NF-κB through 3 different kinase pathways involving MAP-kinases, NIK (NF-κB inducing kinase) and I-κB kinases . All three of these kinases were down-regulated during MAP infection. Thus, during MAP infection, the antigen receptors of T-cells were stimulated; however, due to the lack of co-stimulator molecules from APCs, further T-cell activation apparently was greatly reduced, reducing the host response of immune activation to a level approaching an anergic state at the level of the intestine and influence the disease progression from paucibacillary form to the multibacillary form of the disease.
Cytokine-Cytokine Receptor Interactions (CCRI) Pathway Activation.
The CCRI pathway was strongly activated in all three phases with only a few receptors dominating the activation. These genes were involved in extracellular membrane receptor interaction that included chemokines (CC and CXC), interleukins (ILs), platelet-derived growth factors (PDGFs), and tumor necrosis factors (TNFs). Chemokines and their receptors are important for the migration of various cell types into inflammatory sites. Only the genes CCR4, CXCL9, BLR1 and CCR8 were highly up-regulated in the Early Phase, while the remaining chemokines where moderately down-regulated or not expressed in this Phase. In the Intermediate and Late Phases, the previous chemokines were not expressed and the following chemokines become strongly up-regulated: CCL24, CX3CL1, CCL8, and CCL20 while CXCL11 become strongly down-regulated. CCR4 is a receptor for CCL5, and CXCL11 is chemotactic for activated T cells. BLR1 also known as CXCR5, has a role in Peyer's patch primary follicles relating to B cell migration . The biological roles of these genes are described in Table 14. Studies of the CCL8 receptor and its ligands suggested its role in regulation of monocyte chemotaxis and thymic cell apoptosis. More specifically, this receptor may contribute to the proper positioning of activated T cells within the antigenic challenge sites and specialized areas of lymphoid tissues . The gene CCL20 may be involved in formation and function of the mucosal lymphoid tissues by attracting lymphocytes and dendritic cells towards epithelial cells.
The function of the immune system depends in a large part on Interleukins that are predominately synthesized by helper CD4+ T lymphocytes, as well as through monocytes, macrophages, and endothelial cells. Interleukins promote the development and differentiation of T, B, and hematopoietic cells. The strongly expressed interleukins in the Early Phase include IL-4R, IL-18RAP and IL-17RB. IL-4R reverses expression direction and was strongly down regulated in the Intermediate and Late Phase. The strongly up-regulated ILs in the Intermediate and Late Phase include IL-1B, IL-18RAP, IL-7, IL-8RB, and IL-6 while the strongly down-regulated genes include IL-13, IL-15, IL-1A, and IL-4R. The biological roles of these genes are described in Table 15. The soluble epithelial factors (IL-7 and IL-15) differentially regulate homeostasis of intraepithelial lymphocytes and other mucosal leukocytes. IL7 can be produced locally by intestinal epithelial and epithelial goblet cells, and may serve as a regulatory factor for intestinal mucosal lymphocytes. The IL8RB is a receptor for IL8 and mediates neutrophil migration to sites of inflammation. The angiogenic effects of IL8 in intestinal microvascular endothelial cells are mediated by this receptor. Leukocytes play an important role in the maintenance of epithelial barrier. Interestingly, the gene, IL-13, is known to be critical in regulating immune response, but it was strongly down-regulated in the host response and may be important to MAP survival long term.
The elevated expression of platelet-derived growth factors (PDGF) have been linked to early signaling events for infection by intracellular pathogens . PDGF genes and their receptors that were strongly up-regulated in the Early Phase include VEGFB, FLT3, FLT3LG, and CSF1R. VEGFB signals via the endothelial receptor FLT1 and is a regulator of blood vessel physiology, with a role in endothelial targeting of lipids to peripheral tissues. FLT3 encodes a class III receptor tyrosine kinase that regulates hematopoiesis. CSF1R encodes a tyrosine kinase transmembrane receptor and is involved in the functions of macrophages. Expression levels subsided for VEGFB, FLT2LG, and CSF1R in the Intermediate and Late Phase, but FLT3 was triphasic, in that it was highly up-regulated in the Early Phase, highly down-regulated in the Intermediate Phase and becomes highly up-regulated in the Late Phase. Only FLT1 and VEGFC were up-regulated in the Intermediate and Late Phases while KDR and PDGFC were strongly down-regulated. FLT1 encodes a receptor tyrosine-kinase and plays a key role in vascular development and regulation of vascular permeability. VEGFC encodes a PDGF that has a role in endothelial cell growth, stimulating their proliferation and migration and also has effects on the permeability of blood vessels. The down-regulated KDR gene encodes one of the two receptors of the VEGF and is a main mediator of VEGF-induced endothelial proliferation, survival, migration, tubular morphogenesis and sprouting. The down-regulated gene PDGFC is a receptor with tyrosine-kinase activity that has roles in the regulation of many biological processes including embryonic development, angiogenesis, cell proliferation and differentiation, and contribute to the pathophysiology of some diseases, including cancer. Biological roles of these genes are provided in Table 16.
Adipocytokine Signaling (AS) Pathway Manipulation.
The activated AS pathway may be a novel pathway associated with MAP invasion. Several adipocytokines have been found to have a central role in the regulation of inflammation and immunity  and may be important as another MAP host evasion strategy. Adipocytokines exert different effects on the innate immune system and either suppress or activate the monocyte–macrophage system. ADIPOQ (gene encoding adiponectin) through interaction with its receptor ADIPOR1 (and ADIPOR2) suppresses the NF-kB-dependent synthesis of tumour-necrosis factor (TNF) and interferon (IFN). Adiponectin also induces apoptosis of monocytes and inhibits phagocytosis by macrophages.
Wnt Signaling (WS) Pathway Activation.
In our study, Wnt signaling, which was a highly scored pathway during the Early Phase, is known to be associated with regeneration of nervous system cells using an integrative computational model for intestinal tissue renewal . Planar cell polarity signaling is one of the downstream pathways in the Wnt signaling, and it leads to the activation of the small GTPases and Rac-1. Rac-1, one of the mechanistic genes identified in our study, may regulate cell adhesion and epithelial cell motility in response to MAP entry. In neurons, Rac-1 acts through the protein kinase cdk5 and p35 to phosphorylate and down-regulate Pak1, increasing neuronal migration. Rac-1 also interacts with several other factors to regulate enteric neuronal network. Neurexophilins (neuropeptide-like proteins) and neurexins also participate in a neuron signaling pathway –. Neuronactin and contactin transmembrane proteins are also known to mediate cell-cell interactions in nervous system . Neurexins contain epidermal growth factor-like sequences and domains homologous to the G domain repeats of laminin A, as related to its function in ileal mucosa and cell-cell interactions. Intestinal motility is affected by the invading enteric pathogens. A number of gastrointestinal hormones appear to affect intestinal motility . Interestingly, it was shown earlier that the enteric nervous system is involved in inflammatory bowel disease in which MAP is associated , . In cattle and sheep with Johne's disease, myenteric ganglionitis with cellular infiltration occurs . During the experimental inoculation of sheep with MAP, some sheep developed aggregations of mononuclear cells around enteric nerves in the ileal submucosa and myenteric plexus . However, such lesions were not detected in sheep that did not subsequently develop classical disease manifestations. At this juncture, it is not clear if the Johne's disease is also an outcome of the enteric neuropathy that starts when MAP colonizes in the intestine. Thus, further studies are warranted to understand the correlation between MAP colonization and enteric neuropathy. Furthermore, neurotrophins are known to activate two different classes of receptors, the Trk family of receptor tyrosine kinases and p75NTR, a member of the TNF receptor superfamily. Our gene expression data indicate that neurotrophins may be activating TNF receptor superfamily. This interaction may in turn activate many signaling pathways, including those mediated by ras and members of the cdc-42/ras/rho G protein families, and by the MAP kinase, PI-3 kinase, and Jun kinase cascades.
Mitogen-activated Protein Kinase 1 (MAPK1) Influence on MAP invasion.
In our study we focused on the MAPK1 gene. The protein encoded by MAPK1 is a member of the MAP kinase family. MAP kinases, also known as extracellular signal-regulated kinases (ERKs), act as an integration point for multiple biochemical signals, and are involved in a wide variety of cellular processes such as proliferation, differentiation, transcription regulation and development. Unlike other pathogens, MAP did not exhibit increased invasion or replication during the 12 hr post-infection, though MAP was in continuous contact with host Peyer's patch in the ligated ileal loop . To provide evidence that inhibition in the entry of MAP is MAPK1 dependent, we specifically knocked down in vitro MAPK1 gene expression in HeLa cells by siRNA. The invasion of MAP in HeLa cells was highly significantly reduced when we silenced MAP kinase by introducing siRNA (Figure 7). Thus, MAP kinase is probably one of the key genes influencing invasion of MAP.
HeLa cells were transfected with Glyceraldehyde 3 phosphate dehydrogenase (GAPDH), mitogen activated protein kinase 1 (MAPK1), and Epithelial Growth Factor Receptor (EGFR) siRNAs, infected with MAP, and measured for the effects on invasion. Invasion is given as a percentage relative to the negative control that consisted of HeLa cells incubated with transfecting reagent in serum free medium.
Host Immune Tolerance: Suppressed Immune Related Pathways
Pathways that are suppressed may be assumed to be an indicator of MAP host processes that are hi-jacked, but in a way to subvert the host's defensive response. The One Carbon Pool by Folate, Long-term Potentiation, Long-term Depression, and CCR3 Signaling in Eosinophils pathways are potentially hi-jacked processes suppressed in the Early Phase. Further examination of these pathways at the network level provided evidence that MAP was potentially interfering with their immune response functionality.
One Carbon Pool By Folate (OCPF) Pathway Suppression.
The OCPF pathway is suppressed in the Early Phase, was tri-phasic (suppressed, activated, and suppressed) in the Intermediate Phase and was strongly activated in the Late Phase as shown in the pathway heat map of Figure 1. This pathway may be novel to MAP pathogenicity and its impairment may adversely impact genome integrity, disrupt establishment of other metabolic pathways and mechanisms that underlie folate-associated pathologies. A previous study  found that folate deficiency inhibits the proliferation of primary CD8+ T Lymphocytes, another possible mechanism underlying host immune tolerance of MAP. Within the One Carbon Pool by Folate pathway are three strongly down-regulated genes (Early Phase) that include MTHFD1, MTHFD2, and GART. In the Late phase, GART reversed to become strongly up-regulated while MTHFD1 and MTHFD2 reduced expression levels to a moderate insignificant level. A review of the literature found MTHFD1 and MTHFD2 to have some association with immune response. The MTHFD1 encodes a protein that possesses three distinct enzymatic activities that are essential cofactors for thymidylate and purine synthesis . Disorders of purine metabolism lead to immunodeficiency having marked susceptibility to infection . Interestingly, the protein encoded by MTHFD2 was found to have highly fluctuating protein abundance levels over time in mouse macrophages infected with Salmonella enterica . This suggests that MTHFD1 and MTHFD2 may be novel to the MAP invasion mechanisms and may warrant further examination in future studies. More details of the biological role of these genes are provided in Table 17.
Long-term Potentiation (LTP) and Long-term Depression (LTD) Pathway Suppressions.
Other novel pathways suppressed in the early phase and reversed to an activated state in the late phase are the LTP and LTD pathways. MAP pathogenicity appears to have interaction with neuronal activity, the mechanisms of which are not well understood. The dominating genes causing the pathways' suppressed scores are PPP1CA, PPP1CB, MAPK1, GNAI3, GNAO1, IGFR1, and Gucy2c. The biological roles of these genes are provided in Table 18. Recently, it was found that Gucy2c is involved in regulating AKT-dependent intestinal barrier integrity . GNAI3 has been linked as an important participant in lymphocyte position and chemokine receptor signaling in B cells .
CCR3 Signaling in Eosinophils (CSE) Pathway Suppression.
In our study, CSE pathway was strongly suppressed in the Early Phase, inactive in the Intermediate Phase, and moderately suppressed in the Late Phase. CSE pathway suppression may be a key mechanism that supports the host tolerance to MAP. Eosinophils are a key class of leukocytes involved in inflammatory responses. Blocking eosinophil activation and the signaling pathways that lead to chemotaxis, degranulation and reactive oxygen release may alleviate inflammatory conditions and inflammation-associated tissue damage which may be a longer term survival mechanism of MAP. A number of genes are strongly down-regulated or not differentially expressed at all during the Early Phase and the majority of genes are lowly expressed in the Intermediate and Late Phases. In fact, the gene CCR3 has a low differential expression across all phases. The protein encoded by CCR3 is a receptor for C-C type chemokines and belongs to family 1 of the G protein-coupled receptors. This receptor binds and responds to a variety of chemokines, including eotaxin (CCL11), eotaxin-3 (CCL26), MCP-3 (CCL7), MCP-4 (CCL13), and RANTES (CCL5). It is highly expressed in eosinophils and basophils, and is also detected in TH1 and TH2 cells. The key genes dominating the suppression of CCR3 Signaling in the Early Phase are GNAQ, MAP2K1, and MYL2 although there are several other genes that had a moderate down regulation. Biological roles of these genes are provided in Table 19. GNAQ was described previously in the Calcium Signaling pathway. MAP2K1 is an essential component of MAP kinase signal transduction pathway, involved in many cellular processes such as proliferation, differentiation, transcription regulation and development and be also be important for eosinophil chemotaxis. The gene MYL2 encodes the myosin, light chain 2 protein. In CSE pathway, MYL2 is associated with the assembly of actomysin filaments. It is also involved in muscle contraction through cyclic interactions with actin-rich thin filaments, creating acontractile force. CSE pathway suppression may disrupt immune defenses by alleviating inflammatory conditions. This may signify another contributing mechanism to host immune tolerance of MAP.
During the Intermediate Phase of MAP infection, higher levels of perturbation of signaling and immune response related pathways reflect active host-pathogen interactions. Autocrine and paracrine cell-cell signaling are very important for the interaction and maintenance of homeostasis within the diverse cell population (enterocytes, dendritic cells, macrophages, eosinophils, mast cells, and natural killer cells) of the intestinal mucosa , .
Host Immune Tolerance: Non-Responsive Immune Related Pathways
Pathways that are non-responsive (not activated) may be an indicator of disrupted host processes by MAP. An interesting finding was that interaction of MAP with the host failed to induce several key immune related pathways during all three phases of host response. These pathways included Fc Epsilon RI Signaling, B cell Receptor (BCR) Signaling, Activation of CSK Through T Cell Receptor Signaling, Natural Killer Cell Mediated Cytotoxicity, and T Cell Receptor Signaling. Fc epsilon RI signaling is exclusive to the mast cells . Suppression of mast cells activity thus affect the innate responses of the host to release several activated molecules, such as biogenic amines (histamines), proteoglycans (heparin), lipid mediators such as leukotrienes (LTC4, LTD4 and LTE4), prostaglandins (especially PDG2) and secretion of cytokines, the most important of which are TNF-α, IL4 and IL5. The suppression of these mediators, cytokines and T-cell receptors signaling along with the up-regulation in the epithelial repair mechanisms and reduced inflammation may enhance MAP intracellular survival and facilitate persistent infection. BCR Signaling inactivity may imply that any signaling pathways emanating from the B cell antigens is likely not stimulating any B lymphocyte immune response.
To further understand the mechanistic events that are suppressing T-cell activation, the CD40L Signaling (CS) and T-cell Signaling (TCS) pathways were examined at the gene expression and network level. Table 20 indicates the DBGGA gene Bayesian z-score results across the Early, Intermediate and Late Phases of host immune response for CS pathway that shows a majority of important genes as not significantly expressed. Suppression of genes in the CS pathway may have a further negative regulation on a large number of genes implicated in host defense against pathogens. For the TCS pathway, the DBGGA analysis clearly indicated a defective antigen processing and presentation by MHC class II molecule as shown in Figure 8a. In this graphical representation of T-cell signaling, several genes encoding MHC molecules were not differentially expressed. MAP infection did not change the expression level of co-stimulatory molecules (CD28, CD24, CD40LG and CD80) that are known to be involved in the activation of PI3K and GRB2 that finally activate NFκB. Interestingly, during MAP infection, LCK and NFATC1 (nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 1), NFATC4 (nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 4) are the mechanistic genes in T-cell signaling pathway and activation of these genes leads to ubiquitin-mediated proteolysis. The NFATC1 and NFATC4 genes were common mechanistic genes in VEGF signaling pathway and T-cell receptor signaling pathway. NFATC1 was strongly down-regulated in the Early Phase and up-regulated in the Late Phase while NFATC4 was only strongly up-regulated in the Early Phase. The products of NFATC1 and NFATC4 genes play a role in the inducible expression of cytokine genes in T-cells, especially in the induction of the IL-2 or IL-4 gene transcription that, in our study, were not differentially expressed. These gene products are also involved in regulation, activation, proliferation and differentiation of T-cells as well as lymphoid and non-lymphoid cells. Several of these genes are involved in maintaining a fine balance between immunity and tolerance.
The model (a) and the heat map (b) indicate an overall lack of differential expression that reflects a defective host immune response state to MAP in bovine Peyer's patch. A defective antigen processing and presentation by MHC class II molecule and lack of immune response is readily evident by examining the network (a) and the Bayesian z-score heat map (b).
Moreover, the expression of all the genes related to MHC molecules (HLA-DMA, HLA-A, HLA-DQB2, HLA-DRA, HLA-DQA1, HLA-DMB, HLA-DOA, HLA-DOB) were not differentially expressed or tended to be down-regulated the entire period of our experiments (Figure 8b). Full T-cell activation requires: 1) binding of the T-cell receptor to the antigen-MHC complex on the antigen presenting cell, and 2) a co-stimulatory signal provided by the binding of the T-cell's CD28 protein to the B7 protein on the Antigen Presenting Cell (APC). In our study, MAP infection resulted in down-regulation of gene expression of MHC molecules at all the time points post-inoculation. Table 21 summarizes the responses or expression of TLR and MHC genes upon interaction of MAP with several host systems –. The MHC down-regulation has also been demonstrated to occur in in vitro grown macrophages (as early as 12 hr post-infection), as well as, in macrophages isolated from subclinical and clinical phase of infection , , . Thus, it is plausible that the down-regulation of the MHC molecules observed in the present study at the Early Phase of infection initiates and facilitates permanently persistent MAP infections. Given the role of MHC molecules in triggering the Th cells, the down-regulation of MHC during the establishment of MAP infection may block the effector arms of immune system. This irreversible down-regulation of MHC expression may contribute at some level to the paucity of T-cell infiltrates and tubercle formation observed in Johne's disease lesions .
Other Mechanistic Genes of Interest
Expression levels of keratinocyte growth factor (KGF), insulin-like growth factor-1 (IGF secreted by intraepithelial lymphocytes) and macrophage-derived factors were highly up-regulated during the Late Phase of infection. Several genes that encode for antimicrobial mechanisms against any invading pathogens were down-regulated in the MAP infection. These included nitric oxide synthase (endothelial as well as hepatic), dipepdidyl-peptidase, and dihydrofolate reductase (DHFR) at all the time points. DHFR is critical for nitric oxide bioavailability in bovine aortic endothelium cells .
Host Immune Tolerance by Antigen Mimicry
Immune tolerance can also be induced via antigen mimicry. In Crohn's patients, amino acid similarities between MAP and intestinal proteins was examined in detail . Auto-reactive lymphocytes specific for glutathione peroxidase participate in the decreased activity of this enzyme observed in Crohn's disease patients. This in turn could lead to an imbalanced and inefficient endogenous antioxidant response in the intestinal mucosa of Crohn's disease patients. Further studies are warranted to understand if a similar kind of antigen mimicry occurs in Johne's disease.
Biological System Level Modeling
A biological system model of the host response to MAP infection was created from the merger of 14 overlapping pathways that were considered to be major players in the host immune tolerance as identified and described above. Figure 9 depicts our conceptual holistic model of the interplay between pathways. This figure illustrates only a portion of the complex interplay that may be occurring as MAP subverts and hi-jacks different host biological processes. The actual systems Bayesian network is comprised of 433 genes constructed from known biological relationships contained in the overlapping pathways and resulted in a very dense network model as illustrated in Figure 10. This system-level network was interrogated to identify genes and key regulatory points (hubs) that are purported to be governing the host response to MAP. Since the model is trained by the host-pathogen response data, the computational nature of the dynamic Bayesian networks permits interrogation of the model both computationally and visually to identify correlated relationships and candidate regulator hubs that are potential targets for immune and/or therapeutic intervention. Table 22 illustrates the interrogation of the model for highly correlated downstream gene relationships for the important regulatory gene AKT3. AKT3 was identified as a key mechanistic gene and a gene with high overlap with multiple pathways. Ultimately, the diversion of the host gene response to benefit the pathogen depends on the activation of different genes in a particular pathway. This system model enabled a broader examination of the interrelated pathway-host response that we could not have otherwise identified from traditional statistical analysis methods. Further interrogation of the system model led to the identification of several mechanistic genes that have high positive correlated relationships and influence on downstream genes that included: LEPR->Ppara, LEP->Stk11, SOCS3->JAK1, SOCS3->Irs2, STAT1->CXCL10, STAT3-Prkag2, MAP3K8->MAP3K14, NFKB1->IFNA13, NFKB2->TNF, JAK1->PTPN11, IFNAR2->STAT1, MAP3K14->IKBKB, TNF->TNFRSF1A, and NFKBIA->NFKBIB. Some of the more dominating negative correlated relationships included: NFKB1->IL8, STAT3->POMC, NFKB1->SOCS3, NFKB1->IL12B, NFKB1->CXCL10, and LEPR->CAMKK2. The identification of these relationships provides important insight into possible points of interaction between MAP and the host and novel points of intervention.
Each box that defines specific pathways has its temporal heat map scores to document the dynamic state of the pathway (see figure legend). The arrows connecting the pathways are drawn to represent possible causal relationships between the pathways.
Fourteen pathways were used to construct the network developed from temporal in vivo host transcriptome data of MAP infected bovine Peyer's patch. The full network model is shown on the bottom figure. A magnified section of the model is shown in the top figure. Mechanistic genes are shown with orange rings. A few important mechanistic genes and arcs (gene-to-gene relationships) are indicated on the magnified view. The thickness of the arcs and their color indicates the strength of positive or negative correlation between the interconnected genes. Some of the pathway arcs are labled: +p = phosphorylation; −p = dephosphorylation; +u = ubiquitination; +m = methylation; e = expression;c = compound; and b = binding and unlabeled connecting arc implies activation.
The temporal in vivo host global gene expression analysis of the MAP infected major target organ in the target animal species provides unique opportunities to systematically identify and define the complexities of major pathways influencing the pathogenesis of Johne's Disease, particularly during the early, intermediate and late phase responses in the first 12 hours post-infection. Our Bayesian analysis and modeling of host gene expression data considerably strengthen the hypothesis that MAP subverts the bovine host innate and adaptive immune responses toward immune tolerance. More specifically, we identified no less than ten major cellular pathways that were subverted to reduce host cellular uptake and phagocytosis of MAP one of which is supported by our in vitro RNAi silencing of the mechanistic MAPK1 gene resulting in highly significant reduced invasion of MAP. Furthermore, our analyses disclosed that MAP compromised the host mucosal immune barrier by manipulating the major mechanistic genes of the junction (gap, tight, adherens), cell adhesion molecules – intergrin mediated pathways, and the trefoil factor initiated mucosal healing pathway, adding credibility that the MAP induced decreased trans-epithelial resistance as revealed in our in vitro model and likely has considerable in vivo significance. Finally, we created a robust biological system model of the bovine host response to MAP infection facilitating computational and visual interrogation of the model to identify several potential targets for intervention. We demonstrated that the systems biology approach not only facilitated observations of a holistic functional picture of early responses to MAP, but also uncovered new pathways reinforcing immune tolerance while identifying mechanistic pathways compromising the enteric mucosal immune barrier during colonization of Peyer's patch by MAP.
Significantly Perturbed Pathways of All Phases. This heat map figure shows all scored pathways meeting a 97.5% confidence threshold at any time point and contains the same pathways as listed in Tables 2, 3 and 4. Starting at the left hand column top pathway, the order of the pathways are from the highest activated pathway score (i.e., Parkinson's Disease highest score at t = 30 min.) to the most suppressed pathway score (i.e., Thiamine metabolism lowest score at t = 720 min.) in the bottom right column. In this figure, the pathway scores are shown as darker red gradients indicating higher activation scores (more up-regulated gene expression within the pathway gene set) while the darker green gradients indicating more suppressed pathway activity (more down-regulated gene expression). Grey is near a zero score and black is equal to a zero score.
Significantly Perturbed GO Categories for Early Phase Response. List of significantly perturbed gene ontology categories of biological processes for the Early Phase time period (30 and 60 minutes post-infection). Scores are determined by Dynamic Bayesian Gene Group Activation technique as explained in the text.
Significantly Perturbed GO Categories for Intermediate Phase Response. List of significantly perturbed gene ontology categories of biological processes for the Intermediate Phase time period (120, 240, and 480 minutes post-infection). Scores are determined by Dynamic Bayesian Gene Group Activation technique as explained in the main text.
Significantly Perturbed GO Categories for Late Phase Response. List of significantly perturbed gene ontology categories of biological processes for the Late Phase time period (720 minutes post infection). Scores are determined by Dynamic Bayesian Gene Group Activation technique as explained in the main text.
Mechanistic Genes Identified by Dynamic Bayesian Modeling. State of modulation is indicated by the “+" for up-regulation and “−" for down-regulation.
Mechanistic Gene Cross-Talk. This Table provides list of mechanistic genes for 43 pathways involved in signaling and immune response. Several mechanistic genes are associated in multiple pathways that may be the source of cross-talk, and thus, have more significant influence governing the host immune tolerance to MAP. Of the 43 pathways analyzed, 36 pathways had at least one overlapping mechanistic gene. It was found that 141 mechanistic genes had overlaps within the 36 pathways examined.
Conceived and designed the experiments: SK LGA. Performed the experiments: SK SDL JESN JFF CAR TG LGA. Analyzed the data: SK KLD CLG HRG LGA. Contributed reagents/materials/analysis tools: KLD REE HAL. Wrote the paper: SK KLD LGA.
- 1. Hines ME 2nd, Stabel JR, Sweeney RW, Griffin F, Talaat AM, et al. (2007) Experimental challenge models for Johne's disease: A review and proposed international guidelines. Vet Microbiol 122: 197–222. doi: 10.1016/j.vetmic.2007.03.009
- 2. Khare S, Nunes JS, Figueiredo JF, Lawhon SD, Rossetti CA, et al. (2009) Early phase morphological lesions and transcriptional responses of bovine ileum infected with Mycobacterium avium subsp. paratuberculosis. Vet Pathol 46: 717–728. doi: 10.1354/vp.08-vp-0187-g-fl
- 3. Momotani E, Whipple DL, Thiermann AB, Cheville NF (1988) Role of M cells and macrophages in the entrance of Mycobacterium paratuberculosis into domes of ileal Peyer's patches in calves. Vet Pathol 25: 131–137. doi: 10.1177/030098588802500205
- 4. Larsen AB, Merkal RS, Cutlip RC (1975) Age of cattle as related to resistance to infection with Mycobacterium paratuberculosis. Am J Vet Res 36: 255–257.
- 5. Waters WR, Miller JM, Palmer MV, Stabel JR, Jones DE, et al. (2003) Early induction of humoral and cellular immune responses during experimental Mycobacterium avium subsp. paratuberculosis infection of calves. Infect Immun 71: 5130–5138. doi: 10.1128/iai.71.9.5130-5138.2003
- 6. Bendixen PH (1977) Application of the direct leukocyte-migration agarose test in cattle from a Mycobacterium paratuberculosis-infected herd. Am J Vet Res 38: 2027–2028.
- 7. Weiss DJ, Evanson OA, Souza CD (2006) Mucosal immune response in cattle with subclinical Johne's disease. Vet Pathol 43: 127–135. doi: 10.1354/vp.43-2-127
- 8. Crawford GC, Ziccardi MH, Gonzales BJ, Woods LM, Fischer JK, et al. (2006) Mycobacterium avium subspecies paratuberculosis and Mycobacterium avium subsp. avium infections in a tule elk (Cervus elaphus nannodes) herd. J Wildl Dis 42: 715–723. doi: 10.7589/0090-3558-42.4.715
- 9. Murphy JT, Sommer S, Kabara EA, Verman N, Kuelbs MA, et al. (2006) Gene expression profiling of monocyte-derived macrophages following infection with Mycobacterium avium subspecies avium and Mycobacterium avium subspecies paratuberculosis. Physiol Genomics 28: 67–75. doi: 10.1152/physiolgenomics.00098.2006
- 10. Souza CD, Evanson OA, Weiss DJ (2007) Role of the mitogen-activated protein kinase pathway in the differential response of bovine monocytes to Mycobacterium avium subsp. paratuberculosis and Mycobacterium avium subsp. avium. Microbes Infect 9: 1545–1552. doi: 10.1016/j.micinf.2007.08.008
- 11. Weiss DJ, Souza CD, Evanson OA, Sanders M, Rutherford M (2008) Bovine monocyte TLR2 receptors differentially regulate the intracellular fate of Mycobacterium avium subsp. paratuberculosis and Mycobacterium avium subsp. avium. J Leukoc Biol 83: 48–55. doi: 10.1189/jlb.0707490
- 12. Souza CD, Evanson OA, Weiss DJ (2006) Mitogen activated protein kinase p38 pathway is an important component of the anti-inflammatory response in Mycobacterium avium subsp. paratuberculosis-infected bovine monocytes. Microb Pathog 41: 59–66. doi: 10.1016/j.micpath.2006.04.002
- 13. United States Department of Agriculture, Animal and Plant Health Inspection Service web site. Available: http://www.aphis.usda.gov/animal_health/nahms/dairy/downloads/dairy07/Dairy07_is_Johnes.pdf. Accessed 2012 July 25. Johne's Disease on U.S. Dairies, 1991–2007, (info sheet, 4p, 4/08).
- 14. Pierce ES (2009) Possible transmission of Mycobacterium avium subspecies paratuberculosis through potable water: lessons from an urban cluster of Crohn's disease. Gut Pathog 1: 17. doi: 10.1186/1757-4749-1-17
- 15. Chamberlin WM, Naser SA (2006) Integrating theories of the etiology of Crohn's disease. On the etiology of Crohn's disease: questioning the hypotheses. Med Sci Monit 12: RA27–33.
- 16. Lidar M, Langevitz P, Shoenfeld Y (2009) The role of infection in inflammatory bowel disease: initiation, exacerbation and protection. Isr Med Assoc J 11: 558–563.
- 17. Pineton de Chambrun G, Colombel JF, Poulain D, Darfeuille-Michaud A (2008) Pathogenic agents in inflammatory bowel diseases. Curr Opin Gastroenterol 24: 440–447. doi: 10.1097/mog.0b013e3283023be5
- 18. Sartor RB (2003) Targeting enteric bacteria in treatment of inflammatory bowel diseases: why, how, and when. Curr Opin Gastroenterol 19: 358–365. doi: 10.1097/00001574-200307000-00006
- 19. Sartor RB (2005) Does Mycobacterium avium subspecies paratuberculosis cause Crohn's disease? Gut 54: 896–898. doi: 10.1136/gut.2004.055889
- 20. Knosel T, Schewe C, Petersen N, Dietel M, Petersen I (2009) Prevalence of infectious pathogens in Crohn's disease. Pathol Res Pract 205: 223–230. doi: 10.1016/j.prp.2008.04.018
- 21. Waddell LA, Rajic A, Sargeant J, Harris J, Amezcua R, et al. (2008) The zoonotic potential of Mycobacterium avium spp. paratuberculosis: a systematic review. Can J Public Health 99: 145–155.
- 22. Kirkwood CD, Wagner J, Boniface K, Vaughan J, Michalski WP, et al. (2009) Mycobacterium avium subspecies paratuberculosis in children with early-onset Crohn's disease. Inflamm Bowel Dis 15: 1643–1655. doi: 10.1002/ibd.20967
- 23. Khare S, Ficht TA, Santos RL, Romano J, Ficht AR, et al. (2004) Rapid and sensitive detection of Mycobacterium avium subsp. paratuberculosis in bovine milk and feces by a combination of immunomagnetic bead separation-conventional PCR and real-time PCR. J Clin Microbiol 42: 1075–1081. doi: 10.1128/jcm.42.3.1075-1081.2004
- 24. Lawhon SD, Khare S, Rossetti CA, Everts RE, Galindo CL, et al. (2011) Role of SPI-1 Secreted Effectors in Acute Bovine Response to Salmonella enterica Serovar Typhimurium: A Systems Biology Analysis Approach. PLoS ONE e26869. doi: 10.1371/journal.pone.0026869
- 25. Rossetti CA, Galindo CL, Everts RE, Lewin HA, Garner HR, et al. (2011) Comparative analysis of the early transcriptome of Brucella abortus–infected monocyte-derived macrophages from cattle naturally resistant or susceptible to brucellosis. Res Vet Sci 91: 40–51. doi: 10.1016/j.rvsc.2010.09.002
- 26. Loor JJ, Everts RE, Bionaz M, Dann HM, Morin DE, et al. (2007) Nutrition-induced ketosis alters metabolic and signaling gene networks in liver of periparturient dairy cows. Physiol Genomics 32: 105–116. doi: 10.1152/physiolgenomics.00188.2007
- 27. Everts RE, Band MR, Liu ZL, Kumar CG, Liu L, et al. (2005) A 7872 cDNA microarray and its use in bovine functional genomics. Vet Immunol Immunopathol 105: 235–245. doi: 10.1016/j.vetimm.2005.02.003
- 28. Adams LG, Khare S, Lawhon SD, Rossetti CA, Lewin HA, et al. (2011) Multi-comparative systems biology analysis reveals time-course biosignatures of in vivo bovine pathway responses to B.melitensis, S.enterica Typhimurium and M.avium paratuberculosis. BMC Proc 5 Suppl 4: S6. doi: 10.1186/1753-6561-5-s4-s6
- 29. Baldi P, Long AD (2001) A Bayesian framework for the analysis of microarray expression data: regularized t -test and statistical inferences of gene changes. Bioinformatics 17: 509–519. doi: 10.1093/bioinformatics/17.6.509
- 30. Long AD, Mangalam HJ, Chan BY, Tolleri L, Hatfield GW, et al. (2001) Improved statistical inference from DNA microarray data using analysis of variance and a Bayesian statistical framework. Analysis of global gene expression in Escherichia coli K12. J Biol Chem 276: 19937–19944. doi: 10.1074/jbc.m010192200
- 31. Plattner BL, Chiang YW, Roth JA, Platt R, Huffman E, et al. (2011) Direct inoculation of Mycobacterium avium subspecies Paratuberculosis into ileocecal Peyer's patches results in colonization of the intestine in a calf model. Vet Pathol 48: 584–592. doi: 10.1177/0300985810383874
- 32. Bermudez LE, Petrofsky M, Sommer S, Barletta RG (2010) Peyer's patch-deficient mice demonstrate that Mycobacterium avium subsp. paratuberculosis translocates across the mucosal barrier via both M cells and enterocytes but has inefficient dissemination. Infect Immun 78: 3570–3577. doi: 10.1128/iai.01411-09
- 33. Smythies LE, Sellers M, Clements RH, Mosteller-Barnum M, Meng G, et al. (2005) Human intestinal macrophages display profound inflammatory anergy despite avid phagocytic and bacteriocidal activity. J Clin Invest 115: 66–75. doi: 10.1172/jci19229
- 34. Fooksman DR, Vardhana S, Vasiliver-Shamis G, Liese J, Michael L Dustin (2011) Functional Anatomy of T Cell Activation and Synapse Formation. Annual Review of Immunology doi: 10.1146/annurev-immunol-030409-101308
- 35. Kim M, Ogawa M, Fujita Y, Yoshikawa Y, Nagai T, et al. (2009) Bacteria hijack integrin-linked kinase to stabilize focal adhesions and block cell detachment. CNature 459: 578–582. doi: 10.1038/nature07952
- 36. Secott TE, Lin TL, Wu CC (2004) Mycobacterium avium subsp. paratuberculosis fibronectin attachment protein facilitates M-cell targeting and invasion through a fibronectin bridge with host integrins. Infect Immun 72: 3724–3732. doi: 10.1128/iai.72.7.3724-3732.2004
- 37. Taupin D, Podolsky DK (2003) Trefoil factors: initiators of mucosal healing. Nat Rev Mol Cell Biol 4: 721–732. doi: 10.1038/nrm1203
- 38. Meena LS, Rajin (2010) Survival mechanisms of pathogenic Mycobacterium tuberculosis H37Rv. The FEBS Journal 277: 2416–2427. doi: 10.1111/j.1742-4658.2010.07666.x
- 39. Kusner DJ, Hall CF, Schlesinger LS (1996) Activation of phospholipase D is tightly coupled to the phagocytosis of Mycobacterium tuberculosis or opsonized zymosan by human macrophage. J Exp Med 184: 585–595. doi: 10.1084/jem.184.2.585
- 40. Rosales C (2005) Molecular Mechanisms of Phagocytosis: Landes Biosciences and Springer. 164 p.
- 41. Goldfine H, Wadsworth SJ, Johnston NC (2000) Activation of Host Phospholipases C and D in Macrophages after Infection with Listeria monocytogenes. Infect Immun 68: 5735–41. doi: 10.1128/iai.68.10.5735-5741.2000
- 42. Russell DG, Cardona P-J, Kim M-J, Allain S, Altare F (2009) Foamy macrophages and the progression of the human TB granuloma. Nat Immunol 10: 943–948. doi: 10.1038/ni.1781
- 43. Russell DG, VanderVen BC, Lee W, Abramovitch RB, Kim M, et al. (2010) Mycobacterium tuberculosis wears what it eats. Cell Host Microbe 8: 68–76. doi: 10.1016/j.chom.2010.06.002
- 44. Sharbati J, Lewin A, Kutz-Lohroff B, Kamal E, Einspanier R, et al. (2011) Integrated MicroRNA-mRNA-Analysis of Human Monocyte Derived Macrophages upon Mycobacterium avium subsp. hominissuis Infection. PLoS ONE 6: e20258. doi: 10.1371/journal.pone.0020258
- 45. Fratti RA, Chua J, Vergne I, Deretic V (2003) Mycobacterium tuberculosis glycosylated phosphatidylinositol causes phagosome maturation arrest. PNAS 100: 5437–42. doi: 10.1073/pnas.0737613100
- 46. Armstrong JA, Hart PD (1971) Response of cultured macrophages to mycobacterium tuberculosis with observation on fusion of lysosomes with phagosomes. J Exp Med 143: 713–740. doi: 10.1084/jem.134.3.713
- 47. Clemens DL, Horwitz MA (1995) Characterization of the Mycobacterium tuberculosis phagosome and evidence that phagosomal maturation is inhibited. J Exp Med 181: 257–270. doi: 10.1084/jem.181.1.257
- 48. Crowle AJ, Dahl R, Ross E, May MH (1991) Evidence that vesicles containing living, virulent Mycobacterium tuberculosis or Mycobacterium avium in cultured human macrophages are not acidic. Infect Immun 59: 1823–1831.
- 49. Man SM, Kaakoush NO, Mitchell HM (2011) The role of bacteria and pattern-recognition receptors in Crohn's disease. Nat Rev Gastroenterol Hepatol 8: 152–168. doi: 10.1038/nrgastro.2011.3
- 50. Ferwerda G, Kullberg BJ, Jong DJd, Girardin SE, Langenberg DML, et al. (2007) Mycobacterium paratuberculosis is recognized by Toll-like receptors and NOD2. Journal of Leukocyte Biology 82: 1011–8. doi: 10.1189/jlb.0307147
- 51. Pai RK, Pennini ME, Tobian AAR, Canaday DH, Boom WH, et al. (2004) Prolonged Toll-Like Receptor Signaling by Mycobacterium tuberculosis and Its 19-Kilodalton Lipoprotein Inhibits Gamma Genes in Macrophages Interferon-Induced Regulation of Selected Genes in Macrophages. Infect Immun 72: 6603–14. doi: 10.1128/iai.72.11.6603-6614.2004
- 52. Baum J, Pinder M, Conway DJ (2003) Erythrocyte Invasion Phenotypes of Plasmodium falciparum in the Gambia. Infect Immun 71: 1856–63. doi: 10.1128/iai.71.4.1856-1863.2003
- 53. McKenna HJ, Stocking KL, Miller RE, Brasel K, Smedt TD, et al. (2009) Mice lacking flt3 ligand have deficient hematopoiesis affecting hematopoietic progenitor cells, dendritic cells, and natural killer cells. Blood 95: 3489–3497.
- 54. Sommer S, Pudrith CB, Colvin CJ, Coussens PM (2009) Mycobacterium avium subspecies paratuberculosis suppresses expression of IL-12p40 and iNOS genes induced by signalling through CD40 in bovine monocyte-derived macrophages. Vet Immunol Immunopathol 128: 44–52. doi: 10.1016/j.vetimm.2008.10.294
- 55. Young LS, Eliopoulos AG, Gallagher NJ, Dawson CW (1998) CD40 and epithelial cells: across the great divide. Immunol Today 19: 502–506. doi: 10.1016/s0167-5699(98)01340-1
- 56. Legler DF, Loetscher M, Roos RS, Clark-Lewis I, Baggiolini M, et al. (1998) B Cell–attracting Chemokine 1, a Human CXC Chemokine Expressed in Lymphoid Tissues, Selectively Attracts B Lymphocytes via BLR1/CXCR5. J Exp Med 187: 655–660. doi: 10.1084/jem.187.4.655
- 57. Tiffany HL, Lautens LL, Gao J-L, Pease J, Locati M, et al. (1997) Identification of CCR8: A Human Monocyte and Thymus Receptor for the CC Chemokine I-309. J Exp Med 186: 165–170. doi: 10.1084/jem.186.1.165
- 58. Greenwell-Wild T, Vázquez N, Sim D, Schito M, Chatterjee D, et al. (2002) Mycobacterium avium Infection and Modulation of Human Macrophage Gene Expression. The Journal of Immunology 169: 6286–6297.
- 59. Tilg H, Moschen AR (2006) Adipocytokines: mediators linking adipose tissue, inflammation and immunity. Nature Reviews Immunology 6: 772–783. doi: 10.1038/nri1937
- 60. van Leeuwen IM, Mirams GR, Walter A, Fletcher A, Murray P, et al. (2009) An integrative computational model for intestinal tissue renewal. Cell Prolif 42: 617–636. doi: 10.1111/j.1365-2184.2009.00627.x
- 61. Missler M, Fernandez-Chacon R, Sudhof TC (1998) The making of neurexins. J Neurochem 71: 1339–1347. doi: 10.1046/j.1471-4159.1998.71041339.x
- 62. Missler M, Hammer RE, Sudhof TC (1998) Neurexophilin binding to alpha-neurexins. A single LNS domain functions as an independently folding ligand-binding unit. J Biol Chem 273: 34716–34723. doi: 10.1074/jbc.273.52.34716
- 63. Missler M, Sudhof TC (1998) Neurexophilins form a conserved family of neuropeptide-like glycoproteins. J Neurosci 18: 3630–3638.
- 64. Barthalay Y, Hipeau-Jacquotte R, de la Escalera S, Jimenez F, Piovant M (1990) Drosophila neurotactin mediates heterophilic cell adhesion. Embo J 9: 3603–3609.
- 65. Furness J (2008) The enteric nervous system: normal functions and enteric neuropathies. Neurogastroenterol Motil 20 Suppl 1: 32–38. doi: 10.1111/j.1365-2982.2008.01094.x
- 66. Scanu AM, Bull TJ, Cannas S, Sanderson JD, Sechi LA, et al. (2007) Mycobacterium avium subspecies paratuberculosis infection in cases of irritable bowel syndrome and comparison with Crohn's disease and Johne's disease: common neural and immune pathogenicities. J Clin Microbiol 45: 3883–3890. doi: 10.1128/jcm.01371-07
- 67. Sechi LA, Ruehl A, Ahmed N, Usai D, Paccagnini D, et al. (2007) Mycobacterium avium subspecies paratuberculosis infects and multiplies in enteric glial cells. World J Gastroenterol 13: 5731–5735.
- 68. Buergelt CD, Hall C, McEntee K, Duncan JR (1978) Pathological evaluation of paratuberculosis in naturally infected cattle. Vet Pathol 15: 196–207. doi: 10.1177/030098587801500206
- 69. Gwozdz JM, Thompson KG, Manktelow BW (2001) Lymphocytic neuritis of the ileum in sheep with naturally acquired and experimental paratuberculosis. J Comp Pathol 124: 317–320. doi: 10.1053/jcpa.2001.0468
- 70. Courtemanche C, Elson-Schwab I, Mashiyama ST, Kerry N, Ames BN (2004) Folate deficiency inhibits the proliferation of primary human CD8+ T lymphocytes in vitro. J Immunol 173: 3186–3192.
- 71. Krajinovic M (2004) MTHFD1 gene: role in disease susceptibility and pharmacogenetics. J Immunol 173: 3186–3192. doi: 10.2217/146224220.127.116.119
- 72. Nyhan WL (2005) Disorders of purine and pyrimidine metabolism. Mol Genet Metab 86: 25–33. doi: 10.1016/j.ymgme.2005.07.027
- 73. Shi L, Chowdhury SM, Smallwood HS, Yoon H, Mottaz-Brewer HM, et al. (2009) Proteomic Investigation of the Time Course Responses of RAW 264.7 Macrophages to Infection with Salmonella enterica. Infect Immun 77: 3227–3233. doi: 10.1128/iai.00063-09
- 74. Lin JE, Snook AE, Li P, Stoecker BA, Kim GW, et al. (2012) GUCY2C Opposes Systemic Genotoxic Tumorigenesis by Regulating AKT-Dependent Intestinal Barrier Integrity. PLoS One 7: e31686. doi: 10.1371/journal.pone.0031686
- 75. Hwang I-Y, Park C, Harrision K, Huang N-N, Kehrl JH (2011) Variations in Gnai2 and Rgs1 Expression Affect Chemokine Receptor Signaling and the Organization of Secondary Lymphoid Organs. Genes Immun 11: 384–396. doi: 10.1038/gene.2010.27
- 76. Olsen I, Boysen P, Kulberg S, Hope JC, Jungersen G, et al. (2005) Bovine NK cells can produce gamma interferon in response to the secreted mycobacterial proteins ESAT-6 and MPP14 but not in response to MPB70. Infect Immun 73: 5628–5635. doi: 10.1128/iai.73.9.5628-5635.2005
- 77. Singh UP, Singh R, Singh S, Karls RK, Quinn FD, et al. (2008) CXCL10+ T cells and NK cells assist in the recruitment and activation of CXCR3+ and CXCL11+ leukocytes during Mycobacteria-enhanced colitis. BMC Immunol 9: 25–32. doi: 10.1186/1471-2172-9-25
- 78. Boyce JA (2003) Mast cells: beyond IgE. J Allergy Clin Immunol 111: 24–32. doi: 10.1067/mai.2003.60
- 79. Alzuherri HMLD, Clarke CJ (1997) Altered intestinal macrophage phenotype in ovine paratuberculosis. Res Vet Sci 63: 139–143. doi: 10.1016/s0034-5288(97)90007-2
- 80. Berger STGF (2006) A comparison of ovine monocyte-derived macrophage function following infection with Mycobacterium avium ssp. avium and Mycobacterium avium ssp. paratuberculosis. Immunol Cell Biol 84: 349–356. doi: 10.1111/j.1440-1711.2006.01431.x
- 81. Ferwerda GKB, de Jong DJ, Girardin SE, Langenberg DM, van Crevel R, et al. (2007) Mycobacterium paratuberculosis is recognized by Toll-like receptors and NOD2. J Leukoc Biol 82: 1011–1018. doi: 10.1189/jlb.0307147
- 82. Lei LHJ (2007) Limited phenotypic and functional maturation of bovine monocyte-derived dendritic cells following Mycobacterium avium subspecies paratuberculosis infection in vitro. Vet Immunol Immunopathol 120: 177–186. doi: 10.1016/j.vetimm.2007.06.031
- 83. Purdie ACPK, Begg DJ, de Silva K, Whittington RJ (2012) Expression of genes associated with the antigen presentation and processing pathway are consistently regulated in early Mycobacterium avium subsp. paratuberculosis infection. Comp Immunol Microbiol Infect Dis 35: 151–162. doi: 10.1016/j.cimid.2011.12.007
- 84. Ruiz-Larrañaga OMC, Iriondo M, Garrido JM, Molina E, Vazquez P, et al. (2011) Genetic variation of toll-like receptor genes and infection by Mycobacterium avium ssp. paratuberculosis in Holstein-Friesian cattle. J Dairy Sci 3635–3641.
- 85. Taylor DLZL, Begg DJ, de Silva K, Whittington RJ (2008) Toll-like receptor genes are differentially expressed at the sites of infection during the progression of Johne's disease in outbred sheep. Vet Immunol Immunopathol 124: 132–151. doi: 10.1016/j.vetimm.2008.02.021
- 86. Weiss DJEO, McClenahan DJ, Abrahamsen MS, Walcheck BK (2001) Regulation of expression of major histocompatibility antigens by bovine macrophages infected with Mycobacterium avium subsp. paratuberculosis or Mycobacterium avium subsp. avium. Infect Immun 96: 1002–1008. doi: 10.1128/iai.69.2.1002-1008.2001
- 87. Berger ST, Griffin FT (2006) A comparison of ovine monocyte-derived macrophage function following infection with Mycobacterium avium ssp. avium and Mycobacterium avium ssp. paratuberculosis. Immunol Cell Biol 84: 349–356. doi: 10.1111/j.1440-1711.2006.01431.x
- 88. Weiss DJ, Evanson OA, McClenahan DJ, Abrahamsen MS, Walcheck BK (2001) Regulation of expression of major histocompatibility antigens by bovine macrophages infected with Mycobacterium avium subsp. paratuberculosis or Mycobacterium avium subsp. avium. Infect Immun 69: 1002–1008. doi: 10.1128/iai.69.2.1002-1008.2001
- 89. Weiss DJ, Souza CD (2008) Review paper: modulation of mononuclear phagocyte function by Mycobacterium avium subsp. paratuberculosis. Vet Pathol 45: 829–841. doi: 10.1354/vp.45-6-829
- 90. Chalupsky K, Cai H (2005) Endothelial dihydrofolate reductase: critical for nitric oxide bioavailability and role in angiotensin II uncoupling of endothelial nitric oxide synthase. Proc Natl Acad Sci U S A 102: 9056–9061. doi: 10.1073/pnas.0409594102
- 91. Polymeros D, Bogdanos DP, Day R, Arioli D, Vergani D, et al. (2006) Does cross-reactivity between mycobacterium avium paratuberculosis and human intestinal antigens characterize Crohn's disease? Gastroenterology 131: 85–96. doi: 10.1053/j.gastro.2006.04.021