Transcript M. tuberculosis - Denise Kirschner

Dynamics of the Immune Response
during human infection with
M.tuberculosis
Denise Kirschner, Ph.D.
Dept. of Microbiology/Immunology
Univ. of Michigan Medical School
Outline of Presentation
•
•
•
•
Introduction to TB immunobiology
Modeling the host-pathogen interaction
Experimental Method- temporal model
Results:
• dynamics of infection
• depletion/deletion experiments
• Spatio-temporal models
• granuloma formation
Mycobacterium
tuberculosis
1/3 of the world infected
3 million+ die each year
no clear understanding of distinction
between different disease trajectories:
70%
Exposure
No infection
30%
Infection
5%
95%
Acute
disease
Latent
disease
5-10%
Reactivation
HUMAN GRANULOMA- snap shot
Cell mediated immunity in
M. tuberculosis infection
What elements of the host-mycobacterial
dynamical system contribute to different
disease outcomes once exposed?
Hypothesis: components of the cell
mediated immune response determine
either latency or active disease (primary
or reactivation)
 Wigginton and Kirschner J Immunology 166:1951-1976,
2001
Cellmediated
Immunity:
Activated
MFs
Humoralmediated
immunity
Complex interactions between
cytokines and T cells: black=production,
green=upregulation, red=downregulation
Experimental Approach
Build a virtual model of human TB
describing temporal changes in
broncoalveolar lavage fluid (BAL) to
predict mechanisms underlying different
disease outcomes
Use model to ask questions about the
system
Methodology for TB Model
Describe separate cellular and cytokine
interactions
Translate into mathematical expressions
nonlinear ordinary differential equations
Estimate rates of interactions from data
(parameter estimation)
Simulate model and validate with data
Perform experiments
Variables tracked in our
model:
Macrophages: resting, activated,
chronically infected
T cells: Th0, Th1, Th2
Cytokines: IFN-g,IL-4, IL-10, IL-12
Bacteria: both extracellular and
intracellular
Define 4 submodels
Parameter Estimation: inclusion
of experimental data
Estimated from literature giving weight to
humans or human cells and to M.
tuberculosis over other mycobacteria
species
Units are cells/ml or pg/ml of BAL
Sensitivity and Uncertainty analyses can
be performed to test these values or
estimate values for unknown parameters
Example: estimating growth
rate of M. tuberculosis
in vitro estimates for doubling times of
H37Rv lab strain within macrophages
ranged from 28 hours to 96 hours
In mouse lung tissue, H37Rv estimated to
have a doubling time of 63.2 hours
We can estimate the growth rates of
intracellular vs. extracellular growth rates
from these values (rate=ln2/doub. time )
Model Outcomes: Virtual infection
within humans over 500 days
No infection - resting macrophages are at
their average value in lung (3x105/ml)
(negative control)
Clearance - a small amount of bacteria
are introduced and infection is cleared
(PPD-)
latent TB (a few macrophages harbor all may miss them in biopsy)
Active, primary TB
What determines these
different outcomes?
Detailed Uncertainty and Sensitivity
Analyses on all parameters in the system
Total T cells
Varying T
cell killing of
infected
macrophages
Total bacteria
different disease
outcomes
 Production of IL-4
Rates of macrophage activation and
infection
Rate t cells lyse infected macrophages
Rate extracellular bacteria are killed by
activated macrophages
Production of IFN-g from NK and CD8 cells
Virtual Deletion and
Depletion Experiments:
Deletion: mimic knockout (disruption)
experiments where the element is
removed from the system at day 0. D
Depletion: mimic depletion of an element
by setting it to zero after latency is
achieved.
Summary of Deletion Experiments:
IFN-g: Active disease within 100 days
IL-12: Active disease within 100 days
IL-10: oscillations around latent state –
thus it is needed to maintain stability of
latent state
Depletion Experiments
IFN-g: progress to active disease within
500 days
IL-12: still able to maintain latency; much
higher bacterial load
IL-10:
IL-10 Depletion
Present Work- cellular level
Include in the temporal BAL model:
CD8+ T cells and TNF-a
(D. Sud)
Develop a spatio-temporal model of
infection
** Granuloma Formation and Function
 3 approaches
Role of Dendritic cells in priming of T cells
2-compartment model: lymph nodes + lung
(Dr. S. Marino)
Present Work: intracellular level
Temporal specificity by M. tuberculosis
inhibiting antigen presentation in
macrophages
(S. Chang)
The balance of activation, killing and iron
homeostasis in determining M. tuberculosis
survival within a macrophage
(J. Christian Ray)
Spatio-temporal models of
granuloma formation
Metapopulation Model
(Drs. S. Ganguli & D. Gammack)
Agent based model
(Drs. J. S-Juarez & S. Ganguli)
PDE model
(Dr. D. Gammack)
Metapopulation
Modeling
Discrete Spatial Model
of Granuloma Development
 Partition space: nxn
lattice of compartments
 Model diffusion between
compartments
movement based on local
differences (gradient)
Probabilistic movement
 Model interactions within
compartments
Existing temporal model
n2 Systems of ODEs
Modeling diffusion
Example:
Chemokine C diffuses
out from a source
C
Modeling diffusion
Example:
Chemokine C diffuses
out from a source
Diffusion of
macrophages M is
biased towards higher
concentrations of C
C
M
Model: series of ODE
systems
Generate ODEs for C, M, … within each
compartment: terms for source, decay,
diffusion, etc.
Solve ODE system over short time interval
Generate new diffusion patterns based on
updated values; generate new ODEs
Iterate…
Discrete spatial model:
simulations
Agent Based Modeling
Model Agents
DISCRETE ENTITIES
Cells
Macrophages in different states: Activated,
Resting, Infected and Chronically infected
Effector T cells
CONTINUOUS ENTITIES
Chemokine
Extracellular mycobacteria
Model Framework: lattice with
agents and continuous entities
Rules: an example
Resting macrophage phagocytosis
Rules: an example
Macrophage activation by T cells
Granuloma formation- solid
Resting macrophages
Infected macrophages
Chronically infected m.
Activated macrophage
Bacteria
T cells
Necrosis
2x2 mm sq.
Granuloma formation-necrotic
Resting macrophages
Infected macrophages
Chronically infected m.
Activated macrophage
Bacteria
T cells
Necrosis
Kirschner Group Acknowledgments
past &present
Jose S.-Juarez, PhD
David Gammack, PhD
Simeone Marino, PhD
Suman Ganguli, PhD
Ping Ye, PhD
Seema Bajaria, MS
Ian Joseph
Christian Ray
Stewart Chang
Dhruv Sud
Joe Waliga
NIH and The Whitaker Foundation
Collaborators: JoAnne Flynn (Pitt)
John Chan (Albert Einstein)