genomewide characterization of host-pathogen interactions by ...

Maren Depke
Results
Gene Expression Pattern of Bone-Marrow Derived Macrophages after Interferon-gamma Treatment
Fig. R.3.1:
Transcriptome data visualization using Principal
Component Analysis (PCA).
Log-intensity data were used to derive a PCA
plot containing 12 arrays (analyzing 3 biological
replicates of 2 strains and 2 treatment groups).
Resulting principal components were set to
cover 95 % of total variation, while the total
number of principal components was not
limited. The analysis was restricted to noncontrol
probe sets that were not absent on all
12 arrays, when absence of expression is
defined via a p-value > 0.01 on profile level in
Rosetta Resolver. Differences on the axis for
principal component 1 are by a factor of
approximately 250 smaller than differences on
the axis for principal components 2 and 3,
which is shown in the small insert that depicts
the same PCA but is turned by 90° to the right
around the axis of principle component 2.
Arrays of BALB/c BMM samples are
represented by dots (•), arrays of C57BL/6
BMM samples by rhombi (). IFN-γ treated
samples are indicated in gray, and non-treated
control level samples are colored in black.
As transcriptome data demonstrated a high reproducibility of biological replicates in the
model system, selected samples were used for different proteome applications rather than
analyzing different biological replicates with only one proteomics approach (Dinh Hoang Dang
Khoa).
For a general comparison, gelfree proteome analysis and transcriptome analysis are more
comparable than 2-DE proteome analysis and transcriptome analysis. Therefore, further data
examination focused on that.
Overall comparison of transcriptome [Maren Depke] and LC-MS/MS proteome [Dinh Hoang
Dang Khoa] analyses
900 of 946 identified proteins from gel-free LC-MS/MS approach could be mapped to the
20074 EntrezGene records that were specified to be represented on the Affymetrix GeneChip
Mouse Gene 1.0 ST array by the Rosetta Resolver software (Table R.3.1). As expected, a much
bigger number of genes was accessible by transcriptome analysis than proteins by gelfree
proteome analysis. Nevertheless, for almost all identified proteins the gene expression pattern
could be recorded.
Table R.3.1: Overall comparison transcriptome [Maren Depke] and LC-MS/MS proteome [Dinh Hoang Dang Khoa] analyses and
numbers of differentially regulated genes and proteins with differing abundance resulting from LC-MS/MS analysis.
total number of
detected genes or
proteins
in BALB/c BMM
IFN-γ effects a
in C57BL/6 BMM
at non-treated
control level
strain differences a
at IFN-γ
treated level
microarray 20074 442 (2.2 %) 396 (2.0 %) 222 (1.1 %) 230 (1.1 %)
LC-MS/MS 946 45 (4.8 %) 52 (5.5 %) 218 (23.0 %) 308 (32.6 %)
a Percentage values indicate the proportion of regulation in both approaches. Values were calculated by dividing the number of
regulated genes by the total number of genes available on the array or the number of regulated proteins by the total number of
identified proteins.
92