Transcript Rossetti C BrucRes Conf 07 v2 LGA

Temporal Transcriptional Analysis of
the Early in vivo Initial Interactions of
both Brucella & Bovine Host
Carlos A. Rossetti, Brian Kamery, Sara Lawhon, Jairo Nunes,
Tamara Gull, Cristi Galindo, Sangeeta Khare, Robin Everts,
Mitch Magee, Harris Lewin, Stephen Johnston, Harold Garner,
Ken Drake & L. Garry Adams
Department of Veterinary Pathobiology
College of Veterinary Medicine
Texas A&M University
Brucellosis Research Conference
1 XII 07 Chicago
Pathogenesis of initial infection
Natural infections occur primarily
through adhesion and penetration
of mucous membranes
- B. abortus & melitensis:
alimentary tract
-B. canis, B. ovis & B. suis:
genital tract
Macrophages, dendritic cells, &
neutrophils phagocytose free
Brucella in the submucosal
interstitium
Persistence of infection in
reticuloendotelial system
Metastasis to regional
(primary) lymph nodes
Objective
The goal of this study was to analyze the
transcriptome of host and Brucella
during their early in vivo interactions in
an effort to understand how this
interaction modulates the outcome of
the infectious process.
METHODOLOGY
• Four - 12 hour ligated ileal loop
non-survival surgeries in 3week old male brucellosis-free
beef calves under BSL3
condition
• Abdominal wall incised under
general anesthesia
• Twenty one – 6 cm ligated ileal
loops
- 7 inoculated with 3 ml of Live
1x109 Wild Type (WT) Bmel 16M
- 7 inoculated with 3 ml of Heat
Inactivated (H-I)1x109 Bmel
- 7 injected with 3 ml of media
(control loops)
METHODOLOGY
• Three loops (1 WT, 1
H-I and 1 control)
excised at 0.25, 0.5,
1, 2, 4, 8 & 12 h PI
• Sampled for tissue
associated bacteria,
histopathology, TEM,
SEM & RNA
• Calves euthanatized
at 12 h PI
BACTERIOLOGICAL RESULTS
Peyer's patches colonization
CFU/g of tissue (log 10)
6.5
6.4
6.3
6.2
6.1
6
5.9
5.8
5.7
0.25
0.5
1
2
4
Time post-inoculation (h)
8
12
BACTERIOLOGICAL RESULTS
• B. melitensis was recovered from
– Systemic blood within 30 min PI
– Mesenteric LN & liver at 12 h PI
– Control & HI loops at 8 & 12 h PI
Conclusions: Rapid penetration of B.
melitensis through Peyer’s patch & metastasis
via lymphatic vessels followed by systemic
bacteremia & organ colonization
Intracellular B. melitensis gene
expression profile
• 4 biological replicas of B. melitensis RNA were enriched &
amplified from total RNA extracted from PP at 15 min to 4 h PI
& co-hybridized against B. melitensis gDNA on 3.2 K
B. melitensis oligo-arrays
• Original total RNA from WT B. melitensis-infected PP (non
enriched, non amplified) was also co-hybridized against
B. melitensis gDNA to B. melitensis oligo-arrays - Oligospots
with signal were considered non-specific & eliminated from
all analyses to reduce false positive gene detection
• The intracellular in vivo B. melitensis gene expression was
compared to the gene expression in the inoculum (cultures
of B. melitensis at late-log growth phase)
In Vivo Intracellular B. melitensis
expression profile
70
618 genes differentially expressed
60
Down reg
- 365 up-regulated
- 253 down-regulated
50
# of genes
Up reg
40
30
20
10
0
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
Protein
biosynthesis
Energy
production and
conversion
Q
R
Unknown
function
COGs functional categories
Amino acid
transport and
metabolism
P
Cell wall/
membrane
biogenesis
In Vivo intracellular profile of
B. melitensis expression
• B. melitensis had a common in vivo
transcriptional profile in the first 4 h PI
• 618 genes (19.3 % of B. melitensis genome)
were identified as differentially expressed in
at least 4 of 5 time points evaluated
• Most of the functional categories were over
expressed, except transcription, defense,
motility, intracellular trafficking & secretion
• 37.5% of the genes differentially expressed
lacked functional annotation
In vivo Bovine Peyer’s patch gene
expression profile
• Total RNA was extracted from 4 calves
at 15 min to 4 h post-infection from WT,
H-I and control loops (n=60)
• RNA was co-hybridized against bovine
reference RNA & 13K custom bovine
arrays (UIUC)
B. melitensis-infected Bovine Peyer’s
patch transcriptional profile
70
Up-regulated
Down-regulated
60
224 genes differentially expressed
- 196 Up-regulated
# of genes
50
15 m - 1 h PI
40
- 28 Down-regulated
30
20
10
0
A
B
C
D
E
F
G
H
I
J
GO biologica l proce sse s
50
45
1 h - 4 h PI
40
Down-reg
- 459 Up-regulated
- 704 Down-regulated
35
# of genes
Up-reg
1163 genes differentially expressed
30
25
20
15
10
5
0
A
B
C
D
E
F
G
H
I
J
K
L
M
N
GO biological processes
O
P
Q
R
S
T
U
V
W
B. melitensis infected bovine Peyer’s
patch transcriptional profile
•
Two different expression profiles were observed in
the bovine Peyer’s patch during the first 4 h PI
– Up-regulation of the transcriptome in the first hour PI (86%)
– Down-regulation of the transcriptome between the 1 and
the 4 h PI (63%)
•
Interesting findings
–
–
–
–
Anti-chemo attractant PMN and monocyte response
Pro-apoptotic response
Pro-abortive transcriptional response
Arresting of the cell cycle and inhibition of cell proliferation
and differentiation
H-I B. melitensis-inoculated bovine Peyer’s
patch expression profile
45
140 genes differentially expressed
Up-regulated
40
- 78 Up-regulated
- 62 Down-regulated
Dow n-regulated
35
# of genes
30
25
20
15
10
5
0
A
B
Inflammatory and
immune response
C
D
E
GO biological processes
F
G
Unknown
function
Mathematical modeling predictive analysis
framework for mechanistic discovery
•
Data fusion
– Prior biological knowledge (qualitative data)
– Metadata
– Data extracted from the actual experiment (quantitative data)
Multi-conditional analysis
Biosystem
modeling and
discovery
Mathematical modeling predictive analysis
framework for mechanistic discovery
•
Data fusion
– Prior biological knowledge (qualitative data)
– Metadata
– Data extracted from the actual experiment (quantitative data)
•
Pathways & GO comparative analysis: Identify significantly expressed
genes associated with known metabolic and regulatory pathways &
map significant changed genes to GO categories
Multi-conditional analysis
Biosystem
modeling and
discovery
B. melitensis bio-signature mechanistic
candidate genes in vivo
Dynamic Bayesian modeling:
- 11 highly activated GO
biological groups containing
78 mechanistic candidate
genes in the first 4 h PI
- 17 top pathways analyzed
containing 119 mechanistic
candidate genes
B. melitensis bio-signature mechanistic
candidate genes in vivo
Dynamic Bayesian modeling:
- 11 highly activated GO
biological groups containing
78 mechanistic candidate
genes in the first 4 h PI
- 17 top pathways analyzed
containing 119 mechanistic
candidate genes
Top Scoring GO
Biological Groups
Top Scoring
Pathways
1. Amino acid
biosynthesis
2. Adaptation to atypical
conditions
3. Cell envelope
4. Regulation of
transcription
5. Transcription
6. Transport: Cations
and iron carrying
compounds
1. Oxidative
phosphorylation
2. Protein export
3. ABC transporters
4. Pyruvate
metabolism
5. TCA cycle
6. Two component
system
Bovine bio-signature mechanistic
candidate genes in vivo
Dynamic Bayesian modeling:
- 47 highly activated GO biological
groups containing 52 mechanistic
candidate genes in the first 4 h PI
- 16 top pathways analyzed
containing 37 mechanistic candidate
genes
Bovine bio-signature mechanistic
candidate genes in vivo
Dynamic Bayesian modeling:
- 47 highly activated GO biological
groups containing 52 mechanistic
candidate genes in the first 4 h PI
- 16 top pathways analyzed
containing 37 mechanistic candidate
genes
Top Scoring GO
Biological Groups
1. Positive regulation of
TNFa biosynthesis
2. Proteolysis and
peptidolysis
3. MHC II presentation
4. Inflammatory response
5. Coupled receptor
protein signaling
pathway
6. Regulation of cell cycle
Top Scoring
Pathways
1. NK cells mediated
cytotoxicity
2. Cytokine-cytokine
receptor interaction
3. MAPK signaling
4. Cell adhesion
5. Ca signaling
pathway
6. Leukocyte
transendotelial
migration
Conclusions
•
Brucella invade the host via intestinal Peyer’s patches followed
•
NO histopathological changes in early infected tissues
•
A common transcriptional profile was identified in B. melitensis
in the first 4 h PI in vivo
•
Two different transcriptional profiles were observed in the
bovine host early in the infection
•
This study provides specific genes & pathways to further
elucidate how both host and Brucella interact in vivo during the
early infectious process to the eventual benefit of the pathogen
and to the detriment of the naïve host
by metastasis & systemic distribution and organ colonization via
blood and lymphatic vessels
Future steps
• Develop of modern software and modeling
approaches that help connect Brucella effectors
with host targets
• Laser Capture Micro-dissection (LCM) analysis to
study the temporal expression profile of both,
Brucella and the host more precisely, providing an
approach of how Brucella modify their
transcriptome inside different cell types & how
these cells respond to Brucella infection
• Discovery of novel genes & important pathways
critical to the host response in the pathogen
virulence to determine potential targets for
subsequent therapeutic and vaccine research
Acknowledgements
•
•
Dr. Adams’ lab
- Tiffany Fausett
- Josely Figueiredo
- Tamara Gull
- Doris Hunter
- Sangeeta Khare
- Sara Lawhon
- Jairo Nunes
- Alan Patranella
- Roberta Pugh
- Quynhtien Tran
B. melitensis microarray printing
•
Microarray analysis
- Dr. Cristi L. Galindo and Dr.
Harold Garner (UTSWMS – Dallas)
- Dr. Bryan Kamery and Dr. Ken
Drake (Seralogix, Inc.)
•
Financial support: I.N.T.A.Fulbright Argentina scholarship,
NIH/NIAID Western Regional
Center of Excellence, & DHS –
FAZD grants
- Dr. Mitchell McGee and Dr.
Stephen A. Johnston from the
Center for Innovations in Medicine,
A.S.U
•
Bovine microarray printing
- Dr. Robin Everts and Dr. Harris
Lewin from UIUC