Identification Of Diagnostic Panel For Major Depressive Disorder In Plasma:a Microarray Proteomics Study

Background:Major depressive disorder (MDD) is a serious mental disorder, affecting up to15%of the general population. Currently, there is a great difficulty to diagnose and treat this disease. Several hypotheses had explained MDD’s etiology previously, including decrements in neurotrophic factors, disturbances of monoamine neurotransmitters and increases of oxidative stress. However, few of them had been widely acknowledged. Additionally, the diagnosis of MDD merely relies on the subjective identification of symptom clusters. The current diagnostic modality leads to a considerable inaccurate rate, as a small percentage of MDD patients having hyperactivity, euphoria, hallucination and even delusion, which give rise to a significant overlap of symptoms to other psychiatric diseases such as bipolar disorder (BPD) and schizophrenia (SCZ). In light of these situations, identification of the candidate biomarkers of MDD facilitates to not only unravel the physiopathologic mechanism, also develop an empirical laboratory-based diagnostic and differential diagnosis approach of MDD.Reverse phrase protein microarray (RPPM), a novel proteomic approach, is a high throughput quantitative platform targeted assessing biological samples (blood, urine and cerebrospinal fluid etc.) simultaneously, witch have been widely applied to research projects of neuropsychiatry. Previous researches focused on characterizing biochemical alterations in depressed animal models, and uncovering the disorder of the peripheral and central system, witch implicated the onset of depression. However, animal model cannot completely copy the heterogeneity of clinical depression, it is greatly essential to perform the blood analysis of depressed patients.Purpose1. PRMA based proteomic method was used to compare the levels of nineteen plasma proteins between depressed patients and healthy controls. This step was aimed to identify the differentially expressed proteins in depressed patients relative to healthy controls.2. By comprehensively analyzing the function of perturbed plasma proteins, we aimed to uncover the molecular basis of MDD.3. We sought to identify and independently validate the diagnostic performance of the potential biomarkers for MDD.Method Samples:The training set samples were used to identify the candidate panel; the test set samples were used to independently validate the diagnostic performance of the identified panel.1. Plasma:Fasting blood samples were collected. The training set, including162first-episode drug-naive MDD subjects and203healthy controls; test sets including74MDD,58BPD,100SCZ and52healthy controls subjects.2. Proteins:Nineteen candidate biological biomarkers resulted from previous researches were selected to detect, including MMP7、 RANTES、MCP-1、MIF、Eotaxin、MMP-9、A2M、IL-8、 Apo-B100、VCAM-1、Leptin、CRP、Apo-Al、SDF-1、MIP-1beta、 Apo-H、VEGF、PON-1、Apo-J.3. Acquisition of the plasma spectra:PRMA based proteomic platform.4. Data analysis:machine-learning algorithms (RF and SVM) coupled with LR were employed to develop a discriminative panel that associated to neuropsychiatry in training set.5. Independently validate the diagnostic performance of plasma panel: The classifier model constructed by the differential proteins was applied to separate MDD from CON, BPD and SCZ subjects in the test set. A ROC curve analysis was carried out to quantify the ability of this biomarker panel.Results: RPPM-based proteomic method was used to compare nineteen protein levels profiling of plasma sampled from162depressed patients and203healthy controls. A five-marker panel consisted of MIP-1beta、Eotaxin、 MMP-9、Apo-B100、Apo-H was yielded by BIC, and RF、SVM and LR algorithms indicate it has the capacity to separate MDD and CON subjects. As compared to healthy controls, depressed patients were characterized by MIP-1beta and Eotaxin were decreased, MMP-9、Apo-B100and Apo-H. Functional analysis of these differential proteins suggested that depression was mainly involved in inflammation reaction disturbances of lipid metabolism and enzyme system disorder. Independent validation test showed that this differential panel was capable of diagnosing blinded samples in MDD and CON with a sensitivity of77.0%and a specificity of88.3%, meanwhile it could classify MDD from BPD with a sensitivity of77.0%and a specificity of63.0%and MDD from SCZ with a sensitivity of77.0%and a specificity of62.0%as well.ConclusionUsing the RPPM detection of plasma obtained from the depressed patients and healthy controls, we successfully identified a panel of differential proteins of depression in plasma. By analyzing the underlying the molecular function of these differential proteins, we provided valuable cures for uncovering the underlying molecular mechanisms of depression, In addition, we performed the independent tests to validate the diagnostic performance of the discriminative panel, which facilitated to develop an objective diagnostic assessment for depression.