Figure 1.
Differentially expressed gene-lists obtained in the 36 comparisons.
The samples have been grouped based on the treatment state (all samples (n = 68), samples with the same treatment in remission and relapse (n = 39) and samples with no treatment at all (n = 12)) and gender (all samples, female samples and male samples) and two gene expression comparisons (relapse vs. remission and remission vs. controls) have been performed within each group using two algorithms: t-test (uncorrected p<0.01) and Rank Products (FDR <0.1). Paired comparisons have been done for relapse/remission comparisons as the samples were obtained from the very same patients. The length of the differentially expressed gene-list for each comparison is shown in bold. The distribution of over- and underexpressed genes is also showed (all genes/overexpressed genes/underexpressed genes). A: lists obtained with the t-test algorithm; B: lists obtained with the Rank Products algorithm.
Table 1.
Average gene number and upregulated/downregulated gene ratio per list by comparison type (based on algorithm, sex, treatment and disease state).
Table 2.
Number of genes differentially expressed in both conditions by specificity (female, male, common), proportion of genes altered in the opposite direction in relapse and remission and significance of this proportion (p value of chi-square test).
Figure 2.
Results of the coexpression analysis performed on the 174 differentially expressed genes.
The main component of the network is shown with a visualization of the specificities (female/male/common) previously assigned to the genes. Blue nodes: female genes; red nodes: male genes; green genes: common genes. Red edges: positive correlations; green edges: negative correlations.
Figure 3.
Main component of the coexpression network with the fold-changes obtained in remission and relapse.
The color of the nodes codes for the differential expression of the genes (overexpression/underexpression), the node shapes for the type of gene (protein-coding/non-coding) and the color of the edges for the sign of the correlation (positive correlation/negative correlation).
Figure 4.
Percentage of genes obtained from “treatment-independent” comparisons for the main coexpression modules.
The percentage of the genes that appeared in the comparisons made with the samples from the patients with the same treatment in remission and relapse and from patients with no treatment at all is visualized and used as a measure for how truly a module reflects the underlying pathophysiology.
Figure 5.
Graphical representation of the “mirror pattern” of the genes altered both in relapse and remission.
A: direction of the alteration (over/underexpression) of the 174 differentially expressed genes in the different comparisons. REL: relapse vs. remission; REM: remission vs. controls. B: box-plot representation of the fold-change values vs. controls in remission and relapse for each of the four main modules of the coexpression network. _REM: remission vs. controls; _REL: relapse vs. controls.
Figure 6.
Serum concentration of Tcn1 of patients in remission, relapse and CIS and healthy controls.
A: box-plot of the protein concentration values in the four groups of the first sample set (49 samples). B: box-plot of the protein concentration values of the samples of the second set (19 samples). *p<0.05; **p<0.01; ***p<0.001.
Figure 7.
Viral gene expression of the EBV in peripheral blood leukocytes.
-ddCts of lytic vs. latency gene expression were obtained from –dCts of BZLF1 expression (lytic expression) and the mean –dCts of the 4 latency genes (EBNA1, EBNA2, LMP1 and LMP2; latency expression). -ddCt >0: higher lytic expression; -ddCt <0: higher latency expression. None of the differences between conditions (controls, patients in remission, patients in relapse) were statistically significant at p<0.05 in none of the groups (all samples, female subjects and male subjects).
Figure 8.
Functional relations between some of the genes of the male module and processes, cells and events known to be involved in multiple sclerosis as described by literature.
Figure 9.
Functional relations between vitamin B12 and MS-related elements as described by literature along with the relevant gene expression regulation relations deduced from the coexpression network.
Figure 10.
Theoretical model for the relapsing-remitting molecular dynamics responsible for disease progression of RRMS in female patients.
Table 3.
Main characteristics of patients and controls included in the study.
Figure 11.
Distribution of samples across the subgroups made based on treatment, sex and condition.
The code, sex and treatment-state are displayed for each subject (treatment is not displayed for controls). SAME TREATMENT SAMPLES: samples of patients with the same treatment state in relapse and remission and matched controls; NO TREATMENT SAMPLES: samples of patients with no treatment in relapse and remission and matched controls; AZA: azathioprine; GA: glatiramer acetate; IFN-β: interferon β; NA: natalizumab.