This study shows discovered 35 DEPs between OCD patients and healthy controls, among which 8 are up-regulated and 27 are down-regulated in OCD. Two proteins (DBH, GRN) were previously thought to be possibly related to OCD. Three proteins were widely discussed in other mental disorders but not reported in OCD before (A2M, APOD, IGFBP1). Among these proteins, DBH functions as a vital molecule in sympathetic system, others are important part in immuno-inflammation system. However, we didn’t detect significantly enriched KEGG pathways, which limited our power to illuminate the mechanism of OCD. Nevertheless, results of GO enrichment still provided some clues. Several immuno-inflammatory related GO terms were enriched in our results. Concerning the protein functions and PPI network, our results are in favor of the sympathetic alteration and the existence of immune-inflammatory mechanism in the pathophysiology in OCD.
We found dopamine beta-hydroxylase (DBH) protein expressed higher in OCD. DBH is an enzyme critical in the process of synthesizing norepinephrine (NE) from dopamine, of which the serum level was thought to be an indicator of sympathetic activity (17). Studies proved that DBH played a meaningful role in schizophrenia, depression, attention deficit hyperactivity disorder and so on (18, 19). There was a study in mice suggesting that Dbh-/- mice virtually lacked obsessive-compulsive behaviors seen in control mice, indicating a potential role of DBH and NE in stress-induced compulsive behaviors(20). Though the abnormal sympathetic activity could be interpreted as the heightened autonomic arousal symptoms in OCD, previous researches regarding sympathetic activity in OCD patients yielded inconsistent results (21–23). Gender differences may be a source of the heterogeneity(24). Therefore, though our study suggested an overall elevation of serum DBH level in OCD, stratification by sex will be needed in future study. Also, the role of norepinephrine in OCD pathology and clinical treatment strategy using anti-adrenergic drugs needs further ascertainment.
Other proteins in our study are mostly related to immuno-inflammation. In fact, researchers have been paying more and more attention on the inflammation mechanism in mental disorders. It was suggested that inflammatory inducers induced by environmental triggers activate the nervous system inflammation. Exposure to inflammatory cytokines may lead to CNS inflammation, affecting key brain areas and driving to clinical symptoms. It may be a vicious spiral that clinical symptoms are also the risk factors of inflammation(25). Inspired by the discovery of the Pediatric Autoimmune Neuropsychiatric Disorder Associated with Streptococcal infection/ Pediatric Acute onset Neuropsychiatric Syndrome (PANDAS/PANS), the inflammation and autoimmunity hypothesis of OCD has attracted researchers’ attention (26). Evidence of serology(27), radiology (28) and therapeusis(26) cumulated. Previous serology clues focused on the cytokines including TNF-α, IL-6, IL-1β and so on(27). There was also a proteomic study in OCD highlighting haptoglobin and hemopexin, which were both immuno-inflammation related proteins. However, the small sample size of that study posed challenge to its power(14). In our study, we provided support for the immuno-inflammatory hypothesis of OCD again.
Progranulins (GRN, or PGRN) is a protein we found decreased in OCD. GRN is a key regulator of inflammation widely expressed in microglia and neurons(29), and also a neurotrophic factor whose deficiency may result in frontotemporal dementia (FTD), a kind of dementia frequently manifesting obsessive-compulsive behaviors(30). In previous work, we supposed some shared pathophysiological factors between OCD and FTD and made a gene*environment analysis, revealing the interaction between GRN gene and the early trauma in OCD (31). There was also a study reported that Grn−/- mice could develop excessive self-grooming behavior (32). Our result was in accordance with previous studies, revealing a complementarity between genomic and proteomic studies and suggesting that GRN may be an important molecule and a promising biomarker in OCD, worthy of further exploration.
Besides GRN, we also found A2M, APOD, IGFBP1 and IGFBP4 significantly altered in serum level of OCD. These proteins are connected intactly in the PPI network that may imply the existence of immuno-inflammatory mechanism.
Alpha-2-macroglobulin (A2M), one of the most important acute phase proteins, was found increased in OCD. A2M was mostly studied in Alzheimer’s disease as a biomarker indicating systematic inflammation(33). There were literatures reporting the elevated serum level of A2M in schizophrenia(34) and depression(35). However, there was also another study reporting its lower serum level in drug-naïve schizophrenia and even lower level in those at-risk-of-psychosis(36). It’s uncertain whether drug usage have any effect on the altered level of A2M. A2M was also a candidate biomarker for treatment response(37, 38). Furthermore, there were researches studying about A2M at genetic or transcriptive level, both were in favor of the role A2M playing in mental disorders(39, 40). Our study reported the elevated level of A2M as a possible biomarker in OCD patients for the first time. Nevertheless, inflammation influenced by other factors such as depressive symptoms was not ruled out in our study. Further studies taking other confounding factors into consideration will be needed. Longitudinal cohort study is also required to outline the trajectory of A2M level during the development of disease, which is vital to tell the inflammatory mechanism of OCD.
We found the elevated serum level of Apolipoprotein D (APOD) in OCD patients for the first time. APOD is an atypical apolipoprotein transporting extracellular lipid, which highly expresses in the nervous system(41) and is the major protein in cerebrospinal fluid(42). Besides, APOD is associated with membrane phospholipid signal transduction and metabolism(43). It has been reported as a potential biomarker in several mental disorders. However, the results were mixed. There was a study found APOD plasma level higher in drug-naive schizophrenia patients(44). This alteration may be explained by the compensatory neuroprotective effects of APOD by attenuating lipid peroxidation. Conversely, other studies found APOD serum level lower in schizophrenia patients(43, 45–47). This contradiction may be owing to the effect of atypical antipsychotics usage. Studies of major depressive disorder (MDD) also met with mixed results (48–51). Nevertheless, it was noteworthy that APOD level was found decreased in serum but elevated in central nervous system in schizophrenia and bipolar disorder patients (43, 45). It may suggest the region-specific effect of APOD compensatory to the systemic lipid metabolism insufficiencies in mental disorder. In our study, we found serum level of APOD higher in OCD. As previous studies also reported the elevation of lipid peroxidation in OCD(52, 53), it is possible that APOD alteration represents a nonspecific response to the disturbed lipid metabolism shared by OCD and other mental disorders. As our samples were from drug-free patients, the effects of drug usage can be ruled out. Though few studies focus on it, the abnormal lipid peroxidation induced area-specific neuronal cell damage may play a part in OCD pathophysiology(52). Due to the essential role of lipid in brain, lipid metabolism will be a promising direction in OCD research. However, the cross-sectional design limited our power to detect the causal relationship. Longitudinal studies and multi-omics analysis will help to understand the pathophysiological role of APOD in OCD.
Insulin-like growth factor-binding protein 1(IGFBP1) and Insulin-like growth factor-binding protein 4 (IGFBP4) we found down-regulated in OCD were both important members in the insulin-like growth factor (IGF) system, which consists of growth peptides (IGF1/2), receptors (IGF1R/2R), and circulating binding proteins (IGFBPs 1–6). IGF1 plays a critical regulatory role in many human tissues. Furthermore, it is famous for its far-reaching effect on immunological system(54). When it comes to mental disorders, IGF1 has been documented as a mood regulator(55) of which the serum level was altered in affective disorders and schizophrenia (56). IGFBPs bind to IGF1 in circulation (57), acting as dynamic regulators of IGF bioactivity by transporting IGF1 and regulating IGF metabolism (58). IGFBP1 was found inversely correlated with the level of free IGF1 in healthy participants(59). Furthermore, IGFBP1 also took part in the immunological system by manifesting positive bilateral regulatory effect with proinflammatory cytokines (60). IGFBP4 mRNA expression was also found increased in hippocampus of schizophrenia patients, indicating the activation of immuno-inflammation-related pathways(61). Studies about IGF system in OCD were scarce. There was one study suggesting a higher serum level of IGF1 in OCD(62), while higher serum level in OCD patients was found correlated with better SSRI responses in another study(63). None of literatures about IGFBPs in OCD were available. In our study, we found IGFBP1 and IGFBP4 downregulated in OCD, but none of significant difference was observed between groups with regards to IGF1, which was different from previous studies. IGFBP1 is rapidly regulated by metabolic status (57). Therefore, we can’t rule out the possibility that the different expression of IGFBP1 was misted by metabolic difference between groups. Also, it was uncertain that whether IGFBP1 and IGFBP4 function in IGF-dependent pathways in OCD. Alternatively, it may also imply a dynamic cooperation mechanism to maintain the balance of IGF system. As shown in interaction network, A2M may also interact with IGFBPs in IGF regulation via IGFBP1/A2M combination, though the exact mechanism was unknown(64). Longitudinal and targeted studies towards IGF systems in OCD will be needed to clarify the pathophysiological role of IGF system molecules.
Furthermore, our study also reported the elevated level of Ig delta chain C (IGHD) and immunoglobulin J chain (JCHAIN). There was a study reporting that Ig kappa chain C was decreased in drug-naïve OCD but was elevated after treatment(13). These results may imply some alteration of secretory immunology in OCD, but the exact pathological meaning was uncertain. There are still other promising proteins in our study, whilst the lack of literatures made it hard to tell their pathophysiological roles. For example, ribonuclease 4 (RNASE4) and angiogenin (ANG) that promote neurogenesis by stimulating angiogenesis(65, 66), which were closely related and both were found downregulated in our samples; syntaxin-binding protein 5-like (STXBP5L) that regulates the neurotransmission presynaptically and may play an important role in mental disorders(67, 68), which was detected with extremely low expression and extremely small p in our result; pyrophosphatase/phosphodiesterase family member 2 (ENPP2) that regulates myelin formation and maintenance, which was detected decreased in the postmortem interval delay of MDD(69).