DEGs of migraine and MS
The flow chart of this study was shown in Fig.1. Compared to the HC group, 2064 DEGs of migraine were identified in PRJEB40032 by differential gene analysis, including 1291 upregulated expressed genes and 773 downregulated expressed genes (Fig.2A). Similarly, 806 DEGs for MS were identified in GSE21942 in contrast to the HC group, including 461 upregulated expressed genes and 345 downregulated expressed genes (Fig.2B) (Supplementary Table 1, 2).
Common DEGs and their functions and pathways
As shown in Fig.2C, migraine and MS shared 112 DEGs, including 42 co-upregulated DEGs and 14 co-downregulated DEGs. In order to discover the shared mechanisms of migraine and MS from functions and pathways, we performed GO annotation and KEGG enrichment analysis for common DEGs. The results showed that myeloid leukocyte activation (5.34E-09), regulation of the inflammatory response (3.17E-08) and reactive oxygen metabolic process (5.62E-08) occupied the top three positions in biological processes (BP) (Fig.3A). Prostaglandin secretion (25.0%), regulation of release of cytochrome c from mitochondria (16.67%) and regulation of neuroinflammatory response (8.33%) played major roles by analysis of ClueGO. In addition, immune receptor activity (2.04E-04), peroxidase activity (2.19E-04) and other molecular functions (MF) as well as specific granule lumen (1.18E-12) and other cellular components (CC) also played an important role (Fig.3B, C). KEGG analysis yielded a total of 34 critical pathways, with the IL-17 signaling pathway (4.81E-05), TNF signaling pathway (1.28E-04) and NOD-like receptor signaling pathway (2.90E-04) in the top three positions (Fig.3D). These pathways were rich in types including infection (32%), immunity (20%), signal processing (15%), endocrine/metabolism (12%), growth/apoptosis (9%), cancer (6%), and cardiovascular (6%) (Fig.3E) (Supplementary Table 3, 4, 5).
Results of gene clusters
To further investigate highlights of the comorbidity of migraine and MS, a PPI network comprising 109 nodes and 224 edges was obtained for common DEGs using the STRING. Four gene clusters were obtained through the Cytoscape plug-in MCODE (Fig.4) and some of the results were displayed in Table 1. Among them, cluster1 was the most significant gene cluster containing 15 genes and involved in a variety of pathogenesis including immune, endocrine, and infection mediated by IL-17 signaling pathway (3.21E-06), TNF signalling pathway (6.47E-06), estrogen signaling pathway (1.16E-02) and Epstein-Barr virus infection (2.38E-02) and so on. The other three clusters were mainly involved in the pathways such as Staphylococcus aureus infection, Aldosterone synthesis and secretion and Osteoclast differentiation (Supplementary Table 6).
Table 1 | The KEGG pathways enrichment analysis of gene clusters.
|
Gene clusters
|
Genes
|
KEGG Pathways
|
Cluster1
Cluster2
Cluster3
Cluster4
|
IL1B,ICAM1,CXCL8,FOS
LCN2,CCR5,RETN,CD69
OSM,CD1E,CHI3L1,EGR1
CCR2,JUN,CD83
MMP8,LTF,DEFA4
EGR2,JUNB,NR4A2
MMP9,SRGN,TYROBP
|
hsa04657:IL-17 signaling pathway
hsa04668:TNF signaling pathway
hsa04064:NF-kappa B signaling pathway
hsa05323:Rheumatoid arthritis
hsa05321:Inflammatory bowel disease
hsa04915:Estrogen signaling pathway
hsa05169:Epstein-Barr virus infection
hsa05150:Staphylococcus aureus infection
hsa04621:NOD-like receptor signaling pathway
hsa04668:TNF signaling pathway
hsa04925:Aldosterone synthesis and secretion
hsa04380:Osteoclast differentiation
hsa04650:Natural killer cell mediated cytotoxicity
|
KEGG: The Kyoto Encyclopedia of Genes and Genomes.
|
Table 2 | The top 10 hub genes of migraine and MS through CytoHubba.
|
MNC
|
EPC
|
Degree
|
Stress
|
Closeness
|
IL1B
CXCL8
MMP9
ICAM1
JUN
LCN2
FOS
CAMP
MMP8
EGR1
|
IL1B
CXCL8
MMP9
ICAM1
JUN
LCN2
CAMP
MMP8
FOS
EGR1
|
IL1B
CXCL8
MMP9
JUN
ICAM1
LCN2
FOS
CAMP
MMP8
EGR1
|
IL1B
JUN
CXCL8
FOS
ICAM1
MMP9
EGR1
DUSP5
LCN2
MMP8
|
IL1B
CXCL8
MMP9
JUN
ICAM1
FOS
CAMP
MMP8
EGR1
LCN2
|
MS: multiple sclerosis; MNC, EPC, Degree, Stress,Closeness are the popular algorithms of CytoHubba.
|
Results of hub genes
In order to further probe the core risk genes of the comorbidity, the hub genes were screened using the Cytoscape plug-in CytoHubba. The results were shown in Fig.5A. Eventually, 9 hub genes (IL1B, JUN, CXCL8, FOS, ICAM1, MMP9, EGR1, LCN2, MMP8) were identified through 5 different algorithms, indicating that they play a vital role in the comorbidity of migraine and MS (Table 2).
Table 3 | Seventeen evaluated critical pathways of migraine and MS.
|
KEGG ID
|
Description
|
P-value
|
hsa04657
hsa04668
hsa04621
hsa05133
hsa04380
hsa05418
hsa05323
hsa04060
hsa04933
hsa05161
hsa04210
hsa04010
hsa05146
hsa04064
hsa04620
hsa05202
hsa05144
|
IL-17 signaling pathway
TNF signaling pathway
NOD-like receptor signaling pathway
Pertussis
Osteoclast differentiation
Fluid shear stress and atherosclerosis
Rheumatoid arthritis
Cytokine-cytokine receptor interaction
AGE-RAGE signaling pathway in diabetic complications
Hepatitis B
Apoptosis
MAPK signaling pathway
Amoebiasis
NF-kappa B signaling pathway
Toll-like receptor signaling pathway
Transcriptional misregulation in cancer
Malaria
|
4.81E-05
1.28E-04
2.90E-04
1.97E-03
2.02E-03
2.89E-03
4.11E-03
4.50E-03
5.32E-03
5.55E-03
1.53E-02
1.70E-02
3.51E-02
3.69E-02
3.69E-02
4.70E-02
4.84E-02
|
MS: multiple sclerosis; KEGG: The Kyoto Encyclopedia of Genes and Genomes.
|
Results of the evaluation of the hub genes and critical pathways
For further evaluation of hub genes and critical pathways, we did the following. Firstly, the hub genes were evaluated by ROC, and the results were shown in Fig.5B, C. In PRJEB40032 (migraine), AUC (IL1B) = 0.958, AUC (ICAM1) = 0.875, AUC (MMP9) = 0.883, AUC (CXCL8) = 0.950, AUC (FOS) = 0.850, AUC (JUN) = 0.900, AUC (EGR1) = 0.925, AUC (LCN2) = 0.933, AUC (MMP8) = 0.833. In GSE21942 (MS), AUC (IL1B) = 0.790, AUC (ICAM1) = 0.957, AUC (MMP9) = 0.995, AUC (CXCL8) = 0.914, AUC (FOS) = 0.714, AUC (JUN) = 0.862, AUC (EGR1) = 0.781, AUC (LCN2) = 0.792, AUC (MMP8) = 0.971. All 9 hub genes had an AUC >0.7 in migraine and MS, indicating a high diagnostic value for migraine and MS. Next, the hub genes were validated using DisGeNET and GeneCards, resulting in the presence of IL1B, ICAM1, MMP9, CXCL8 and FOS in both databases. By further verifying the literature, we determined that 5 hub genes have been shown to be statistically significant in migraine and MS by molecular biology experiments or animal experiments. Another important result was that 17 critical pathways were further validated in migraine and MS through CTD, and the results were shown in Table 3.
Table 4 | Drugs approved/being studied for migraine/MS.
|
Aprroved/being studied in migraine/MS Drugs
|
Drugs approved in migraine (5)
|
Aspirin*, Ibuprofen*, Naproxen*, Acetaminophen*
Nicardipine
|
Drugs being studied in migraine (13)
|
Baclofen**
Ibudilast***, Melatonin***, Gabapentin***, Pregabalin***
Fentanyl, Alprazolam, Cidofovir, Dipyridamole
Celecoxib, Verapamil, Zonisamide, enacarbil
|
Drugs approved in MS (5)
|
Baclofen**
Peginterferon beta-1a, Interferon beta-1a
Interferon beta-1b, Daclizumab
|
Drugs being studied in MS (30)
|
Aspirin*, Ibuprofen*, Naproxen*, Acetaminophen*
Ibudilast***, Melatonin***, Gabapentin***, Pregabalin***
Ustekinumab, Pravastatin, Omeprazole, Midazolam
Curcumin, Lithium carbonate, Interferon alfa, Interferon alfa-2b
Pirfenidone, Secukinumab, Plozalizumab, Filgrastim
Anakinra, Tabalumab, Imatinib, Masitinib, Fenebrutinib
Remibrutinib, Orelabrutinib, Tacrolimus, Abatacept, Otilimab
|
*: Drugs approved for migraine and being studied in MS. **: Drugs approved for MS and being studied in migraine. ***: Drugs that are being studied in both migraine and MS. MS: multiple sclerosis.
|
Potential drugs of migraine and MS
In order to predict therapeutic drugs of migraine and MS, we predicted drugs by targeting evaluated common hub genes and common critical pathways. The results were as follows: (1) 112 drugs targeting 5 hub genes were predicted by DGIdb. Among them, aspirin, ibuprofen, naproxen, and nicardipine were approved for migraine, of which aspirin, ibuprofen, naproxen (non-steroidal anti-inflammatory drugs (NSAIDs)) and acetaminophen were being studied in MS. Baclofen, a gamma-aminobutyric acid (GABA) receptor agonist, approved for MS, was being studied in migraine. There were 9 drugs studied in migraine and 12 drugs studied in MS, with 2 overlapping drugs (melatonin and ibudilast) (Fig.6A). (2) 535 drugs targeting 13 critical pathways were predicted through the CTD. Interferon beta-1a, interferon beta-1b, peginterferon interferon beta-1a, and daclizumab were approved for MS and were not studied in migraine. There were 5 drugs studied in migraine and 19 drugs studied in MS, including 2 overlapping drugs (pregabalin and gabapentin), which were classified as GABA analogs (Fig.6B). (3) 10 drugs targeting both hub genes and critical pathways were screened, mainly anti-inflammatory drugs, immunological agents and antineoplastic drugs, including Infliximab, rilonacept, canakinumab, gevokizumab targeting IL1B and three or more critical pathways. Infliximab was a TNF receptor blocker in clinical practice for rheumatoid arthritis. And rilonacept, canakinumab and gevokizumab were used as IL1β inhibitors for IL1-mediated inflammatory diseases. Also included verapamil studied in migraine and ustekinumab studied in MS ( Fig.6C, Table 4, Table 5). On the one hand, these results demonstrate the credibility of our findings, and on the other hand, they provide some important information on the treatment of drawing on from each other and the comorbidity. For example, NSAIDs may be used to treat MS, and GABA receptor agonists/analogs may work for migraine. It is important that there are 10 drugs with greater potential research value for the treatments of the comorbidity, especially IL1β inhibitor (Supplementary Table 7, 8).
Table 5 | Drugs targeting both hub genes and critical pathways.
|
Drugs
|
Targeting genes
|
Targeting pathways
|
Class and Effect
|
Infliximab
|
IL1B
|
Cytokine-cytokine receptor interaction
|
Anti-inflammatory
|
|
|
Rheumatoid arthritis
|
(TNF-α inhibitor)
|
|
|
MAPK signaling pathway
|
|
|
|
Apoptosis
|
|
Rilonacept
|
IL1B
|
Cytokine-cytokine receptor interaction
|
Anti-inflammatory
|
|
|
Toll-like receptor signaling pathway
|
(IL-1β inhibitor)
|
|
|
MAPK signaling pathway
|
|
|
|
NF-kappa B signaling pathway
|
|
Canakinumab
|
IL1B
|
Cytokine-cytokine receptor interaction
|
Immunological agent
|
|
|
NOD-like receptor signaling pathway
|
(IL-1β inhibitor)
|
|
|
MAPK signaling pathway
|
|
Gevokizumab
|
IL1B
|
Cytokine-cytokine receptor interaction
|
Immunological agent
|
|
|
NOD-like receptor signaling pathway
|
(IL-1β inhibitor)
|
|
|
MAPK signaling pathway
|
|
Verapamil①
|
IL1B, CXCL8
|
MAPK signaling pathway
|
Cardiovascular agent
(Calcium channel blocker)
|
Ustekinumab②
|
IL1B
|
Cytokine-cytokine receptor interaction
|
Immunological agent
(IL-12和IL-23 inhibitor)
|
Tretinoin
|
CXCL8
|
Transcriptional misregulation in cancer
|
Vitamin and mineral
(Vitamin A derivative)
|
Sunitinib
|
CXCL8
|
MAPK signaling pathway
|
Antineoplastic
(Tyrosine kinase inhibitor)
|
Marimastat
|
MMP9
|
TNF signaling pathway
|
Antineoplastic
(Angiogenesis inhibitor)
|
Prinomastat
|
MMP9
|
TNF signaling pathway
|
Antineoplastic
(MMPs inhibitor)
|
①: Drugs being studied in migraine. ②: Drugs being studied in multiple sclerosis.
|