MDD patients fulfilling the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria for major depressive episode who were treated with selective serotonin reuptake inhibitors (SSRIs: escitalopram, fluoxetine, paroxetine, sertraline), or non-SSRI (mirtazapine) were recruited. Diagnosis was confirmed by board certified psychiatrists based on Samsung Psychiatric Evaluation Schedule, case review notes, and SCID (structured clinical interview for DSM-IV) to diagnose depression. A minimum baseline score of 15 for 17-item Hamilton Rating Scale for Depression (HAM-D) was required . Study participants were excluded if they had pregnancy, significant medical conditions, abnormal laboratory baseline values, unstable psychiatric features (e.g., suicidal), history of alcohol or drug dependence, seizures, head trauma with loss of consciousness, neurological illness, or concomitant Axis I psychiatric disorder.
Patients received monotherapy for 6 weeks with one of four commonly used SSRI or mirtazapine antidepressants by clinician’s choice. In this study, choice of drug was driven by the preference of the physician, with consideration of anticipated side effects in at-risk individuals. Dose titration was completed within two weeks. Trough plasma samples were drawn at the end of week 6 for plasma drug concentrations. Lorazepam 0.5–1 mg was allowed at bedtime for insomnia. Patients were seen by a psychiatrist, who monitored their adverse events and severity of depression. The HAM-D was administered by a single trained rater every two weeks . The rater and personnel of protein quantitation were blinded to the hypotheses, drug assignment, and HAM-D data. Therapeutic response was defined as 50% or more reduction of HAM-D score by 6 weeks after initiation of antidepressant treatment.
Clinical data of depressant patients including age, gender, HAM-D score, antidepressant, family history, onset age, number of episode, and duration of current episode were collected for each individual. Healthy individuals without known past medical or psychiatric history or family history of MDD were included as controls. This study was approved by Samsung Medical Center Institutional Review Board. Informed consent was obtained from all participants.
Peripheral blood specimens were obtained from patients with MDD at baseline and at 6 weeks after initiation of antidepressant treatment. Patients were categorized into four groups according to antidepressant type used and treatment responsiveness: SSRIs responder, SSRIs nonresponder, mirtazapine responder, and mirtazapine nonresponder. For marker prioritization, we prepared pooled plasma by pooling equal amounts of plasma specimen from five individuals for each group. To investigate whether proteins detected in pooled plasma were detectable in each plasma, we used these 45 specimens individually. For protein quantitation, we used additional 166 plasma specimens at baseline and at 6 weeks after initiation of either SSRIs or mirtazapine treatment from 78 MDD patients and 10 healthy controls (Fig. 1).
NuPAGE gels (4~12%) were acquired from Invitrogen and PIERCE (Rockford, IL, USA). Sequencing-grade modified trypsin was purchased from Promega (Madison, WI, USA). Acetonitrile (ACN; MS grade), water (MS grade), and formic acid (FA; ACS regent grade) were acquired from Aldrich (Milwaukee, WI, USA). Seven isotope-labeled peptides were synthesized as internal standards for MRM assay (AnyGen Co., Gwangju, Korea): TEDTIFL*R, VANYVDWI*NDR, ITLPDFTGDL*R, ELLESYI*DGR, VSLATV*DK, NALALFVLP*K, and AADDTWEP*FASGK for alpha-1-acid glycoprotein 1 (AGP1), hepatocyte growth factor activator (HGFA), lipopolysaccharide-binding protein (LBP), prothrombin, selenoprotein P (SeP), thyroxine-binding globulin (TBG), and transthyretin, respectively. * represents amino acid labeled with 13C15N heavy isotope.
Marker surrogate selection
We searched proteins that were either related to depressive disorders or located on antidepressant action sites or metabolic pathways in previous individual studies and meta-analysis. In addition to protein markers, genetic markers, DNA, and mRNA were also selected. We manually reviewed them and obtained proteins encoded by genes as well as proteins described in previous literatures. Among selected proteins, we gave priorities to proteins present in serum or plasma based on public databases including Healthy Human Individuals Integrated Plasma Proteome Database and Sys-Body Fluid database as they were studied much more in literatures. Finally, a total of 111 proteins were selected for MRM-based marker prioritization (Additional file 1: Table S1).
For marker prioritization, 50 μL of plasma was resolved on 4~12% NuPAGE gel which was cut into 10 bands and subjected to in-gel tryptic digestion prior to MRM analysis. In-gel digestion was achieved at a 50:1 ratio for 16 h at 37°C. Tryptic digests were recovered by extraction with 50% ACN/0.1% FA and purified using an OMIX C-18 tip (Agilent, Technologies, Santa Clara, CA, USA).
For protein quantitation,1 μL of plasma was mixed with 40 μL of 10 mM of dithiothreitol and incubated at 60°C for 45 min for protein reduction. Next, 5 mM iodoacetamide was added. Samples were incubated at room temperature for 30 min in the dark to induce alkylation followed by digestion with trypsin (1:50) overnight at 37°C, after which digested samples were dried using a centrifugal evaporator. Each tryptic digest sample was spiked with stable isotope labeled peptide. Samples were dried and reconstituted with 20 μL of 0.1 % FA in water.
MRM-based marker prioritization
We generated MRM transitions using MRMPilotTM v2.0 (AB Sciex, Framingham, MA, USA) against 111 proteins selected by literature review. We then monitored these MRM methods using pooled plasma. We performed MRM analysis using a QTRAP 5500 hybrid triple quadrupole/linear ion trap mass spectrometer (AB Sciex) equipped with a nano-electrospray ion source. MRM mode setting was as follows: curtain gas and spray gas at 20 and 25 psi, respectively; collision gas set on high level; and declustering potential set at 100 V. Among 111 proteins, 33 proteins showed signal-to-noise ratio (S/N) above 8 in pooled plasma specimen. For these 33 proteins, we performed MRM analysis using individual plasma specimens to select marker candidates for protein quantitation using stable isotope labeled peptide. The MRM method was identical to that was used for protein measurement of pooled plasma.
MRM-based protein quantitation
Through marker prioritization, seven proteins were detectable (S/N > 8) in individual plasma. For these seven proteins, we performed protein quantitation using stable isotope labeled peptide. Each tryptic digest sample was spiked with stable isotope labeled peptide. Samples were dried and reconstituted with 20 μL of 0.1 % FA in water. MRM analysis was performed using Agilent 6490 Triple Quadrupole mass spectrometer equipped with Agilent 1260 Infinity LC system (Agilent Technologies Inc., Table 1). Tryptic peptides were loaded onto a reversed phase analytical column (150 mm × 0.2 mm i.d., Agilent ZORBAX Eclipse Plus, 1.6 μm particle size) that was maintained at a column temperature of 40°C. Sample separations were achieved using mobile phase A consisting of 0.05% FA and 0.2% methanol in water and mobile phase B consisting of 0.05% FA and 0.2% methanol in ACN. The gradient method was composed of multiple linear gradients as follows (time, % B, flow rate): 3 min, 2% B, 0.25 mL/min; 43 min, 30% B, 0.25 mL/min; 47 min, 90% B, 0.27 mL/min; 53.1 min, 1% B, 0.25 mL/min. Separated peptides were ionized using positive electrospray ionization: 3,500 V capillary voltage, 150 V (high pressure RF) and 60 V nozzle voltage (low pressure RF), a sheath gas flow of 11 L/min at a temperature of 200°C, a drying gas flow rate of 16 L/min at a temperature of 150°C, and 30 psi nebulizer gas flow.
To identify proteins differentially expressed between MDD patients and healthy individuals and between responders and nonresponders at baseline, we performed Fisher’s exact test for categorical variables and t-test or Mann-Whitney U test for continuous variables. Comparisons between patients at baseline and at 6 weeks after initiation of antidepressants in total and subset of patients were performed using Wilcoxon signed rank test or paired t-test. Each comparison was independently performed for SSRIs and mirtazapine groups. Proteins and clinical variables with univariate p-values less than 0.200 were included in multivariate analysis using partial Spearman correlation analysis. P-value of less than 0.050 was regarded as statistically significant. SAS version 9.3 (SAS Institute, Cary, NC, USA) was used for all statistical analyses.