14. Systems biologic approach of diseases
14.11. Application of systems biological approaches
In Chapter 9 it was discussed that in complex diseases most variants identified until now conferred relatively small increments in risk, and explained only a small proportion of familial clustering. Gao et al. tried to alleviate this problem in a systems biological study, in which efforts have been made to prioritize positional candidate genes for complex diseases utilize the protein-protein interaction (PPI) information. 266 known disease genes, and 983 positional candidate genes from the 18 established linkage loci of T1DM, were compiled from the T1Dbase (http://t1dbase.org). It was found that the PPI network of known T1DM genes has distinct topological features from others, with significantly higher number of interactions among themselves. Then those positional candidates were defined to be new candidate disease genes that were first degree PPI neighbours of the 266 known disease genes. This led to a list of 68 genes.
Then it was investigated whether the characteristics of these correspond to those of disease genes. Disease genes have more interactions, and are cited more often in
scientific papers. For the predicted genes, one may argue that their appearance in T1DM publications could be a result of their interactions with the known disease genes, as interacting genes often appear in the same publications. To address this issue, all PubMed records were excluded from the analysis of predicted genes that have cited the known T1DM genes. Out of the 68 new candidates 13 (~20%) were cited significantly more often than random in T1DM publications, as compared to only ~6.9% of the Human Protein Reference Database genes. This was a ~3-fold enrichment. As a group, members of the 68 list were significantly (p<10-7) more likely to appear in T1DM-related publications than members of a random set of 68 genes. It shows that there is a high possibility that these genes play a role in T1DM. Out of the 68 novel candidates, more than a third (24) interact with at least two known disease genes, and about a sixth (12) interacts with at least three. It shows the connection between disease modules and provides further proof that they are really disease genes. This is intuitive, as subsets of genes having much more interactions with each other than with others are likely to be from a same functional network module, and consequently to be involved in the same physiological processes and disease phenotypes.
Figure 14.7. Protein-protein interaction network of the top 5 predictions (ellipse) in T1DM among the 68 proteins and their corresponding baits (round rectangle; interacting known
T1DM genes). Bright magenta nodes represent genes with significant citation in T1DM-related publications (p<0.01). Source: Gao et al. 2009; 18/02/2013.
The number of independent baits (known T1DM genes) for each gene was also determined. Figure 14.7 shows the PPI network of the top 5 candidates in terms of number of baits. On the top are ESR1 and VIL2, each with 6 baits. Interestingly, they are also among the top in terms of independent citations in T1D-related publications and network degrees. ESR1, or estrogen receptor 1, has been cited in 139 (124, after
194 Genetics and genomics
removing co-citation with known disease genes) T1DM-related publications, which ranked number 1 (1) out of the 68 candidates; the number for VIL2 is 30 (29), ranked number 8. The odds ratios to random genes are all greater than 1, at 9.6 for VIL2 and 6.2 for ESR1. Both have abundant interactions with other proteins, with k=163, #1 of the 68 for ESR1; and k=43, #11 for VIL2. These are within the top 2% of all genes, and both can be considered hubs. It must be added that in this case the disease is not caused by mutations in these genes, and thus here the term ‘disease gene’ has a little different meaning as previously used. But they have hub function in protein interaction networks, and part of the disease network. From all of these top 5 genes in Figure 14.7 can be shown that they have significantly more connections compared to the average and all are in pathways connected to T1DM.
New results were gained in a study, where interaction network was deduced from GWAS results of 5 complex diseases (Menon R et al., 2011). Five neurodegenerative and/or autoimmune complex human diseases (Parkinson's disease-Park, Alzheimer's disease-Alz, multiple sclerosis-MS, rheumatoid arthritis-RA and Type 1 diabetes-T1DM) were included. Pathway enrichment analyses were performed on each disease interactome independently. Several issues related to immune function and growth factor signalling pathways appeared in all autoimmune diseases, and, surprisingly, in Alzheimer's disease. Furthermore, the paired analyses of disease interactomes revealed significant molecular and functional relatedness among autoimmune diseases, and, unexpectedly, between T1DM and Alz. T1DM-Alz pair had the highest rank, followed by the autoimmune disease pairs MS-RA and T1DM-RA, and then by MS-Alz and RA-Alz. All Park pairs scored very low. These results are shown in Figure 14.8 displaying a network summary of relationships among the five diseases. This systems biological approach revealed some new and interesting results, from which some are shown below.
Figure 14.8. Overall disease relatedness based on shared pathways in the Panther, KEGG and CGAP-BioCarta databases. Green nodes indicate the neurodegenerative disorders, whereas pink nodes highlight the autoimmune diseases. The color of the edges connecting
the nodes reflects the shared pathway rank ranging from 3 (highest relatedness) to 30 (lowest relatedness). Parkinson's disease-Park, Alzheimer's disease-Alz, multiple
sclerosis-MS, rheumatoid arthritis-RA and Type 1 diabetes-T1D.
Source: Menon R et al., 2011; 18/02/2013.
Numerous immune related pathways were enriched in autoimmune interactomes.
This result was expected as many of the susceptibility genes in RA, T1D and MS were immune related. Notably, the pathways B-cell activation and T-cell activation appeared in all autoimmune diseases. It is well known that both arms of adaptive immunity greatly contribute to autoimmunity. Surprisingly, the same pathways appeared in Alzheimer's disease. The role of adaptive immunity in Alz so far has remained under-explored, however some studies suggest altered T cell phenotypes and responses in such patients. Interestingly, regular use of anti-inflammatory drugs reduces the odds of developing Alz. This observation that the Alz genetic framework may have an impact on immune function, questions the classical distinction between inflammatory and non-inflammatory diseases, and supports the hypothesis that, even though the primary insult is not inflammation but neurodegeneration, immunological pathways play a role in the etiopathogenesis of Alzheimer's disease.
Owing to the often unknown interactions between drug targets and other cellular components, drugs whose efficacy was predicted by specific target-binding experiments may not have the same effect in different clinical settings, in which that target is of modified contextual importance (e.g. tissue-specific isoform compensates for the loss of function of the inhibited protein). Furthermore, single-target drugs may, perhaps, correct some dysfunctional aspects of the disease module, but could alter the activity of other network neighbourhoods, leading to detectable side effects. This network-based view of drug action implies that most disease phenotypes are difficult to reverse through the use of a single ‘magic bullet,’ i.e., an intervention that affects a single node in the network. While network-based approaches represent a relatively recent trend in drug discovery, given the intricate network effects drug development must face, the nascent field of network pharmacology, at the intersection of network medicine and polypharmacology, is poised to become an essential component of drug development strategies. The efficacy of this approach has been demonstrated by combinatorial therapies of AIDS, cancer, or depression, raising an important question:
can one systematically identify multiple drug targets with optimal impact on the disease phenotype? This is an archetypical network problem, leading to methods to identify optimal drug combinations starting either from the metabolic network, or from the bipartite network linking compounds to their drug-response phenotypes. Research in this direction has led to potentially safer multi-target combinations for inflammatory conditions, or to the identification of 14 optimal anti-cancer drug combinations.
Equally important, drug-target networks that link approved or experimental drugs to their protein targets have helped organize the considerable knowledge base encoding the interplay between diseases and drugs. Its analysis demonstrated the preponderance of palliative drugs, i.e., drugs that do not target the actual source of the disease (i.e., the disease-associated proteins) but proteins in the network neighbourhood of the disease proteins.
The first step of rational drug design is an understanding of the cellular dysfunction caused by a disease. By definition, this dysfunction is limited to the disease module, which means that one can reduce the search for therapeutic agents to those that induce detectable changes in the particular disease module. This represents a significant reduction of the search space, also aiding the development of biomarkers for disease detection, as changes in the activity of the disease module components are expected to show the strongest correlations with disease progression (Barabási AL et al., 2011).
196 Genetics and genomics