1.2. Schizophrenia
The genetic architecture of schizophrenia is highly complex and heterogeneous. It is characterized by rare mutations that recently emerged with relatively high risk and common variants with individually minor effects on the disease
[10][10]. Genes implicated by both common and rare alleles operate in crucial pathways for brain development, including histone modification, neuronal migration, transcriptional regulation, immune function, and synaptic plasticity
[11][11].
People living with this disease have a significantly reduced average life expectancy, ~20 years lower than the general population. Nonetheless, the mortality rates are high across all age groups
[8][12][8,12]. The current diagnosis of schizophrenia is mainly based on phenomenological observation and clinical descriptions using the standard operational criteria defined in systematic classifications, namely the Diagnostic and Statistical Manual of Mental Disorders, edition five (DSM-5), and International Classification of Diseases, version 11 (ICD-11), published by the American Psychiatric Association and WHO, respectively
[3][13][14][3,13,14]. The main problem is that these diagnostic definitions have relatively good reliability but no established validity
[15][15].
Epidemiologic studies show that it can take up to several years between symptom onset and diagnosis; evidence suggests that the earlier the diagnosis, the better the prognosis, by decreasing the duration of untreated psychosis
[16][17][16,17].
The symptoms, which typically arise during adolescence or early adulthood, are defined as: (i) positive, such as hallucinations, delusions, and thought disorder; (ii) negative, such as poverty of speech or alogia, lack of motivation and social withdrawal; and (iii) cognitive symptoms, such as attention and learning deficits. While positive symptoms can stabilize throughout the course of illness, negative symptoms tend to increase and become chronic along with cognitive impairments
[18][19][20][18,
19,20], although currently available interventions, such as antipsychotics and cognitive remediation, can reduce negative and cognitive symptomatology
[21][22][21,22].
Psychotic symptoms, which integrate positive symptoms, are a defining feature of SCZ spectrum disorders, and their onset defines the first episode of psychosis
[23][24][23,24]. Despite being considered the main feature for disease onset and diagnostic recognition, psychotic disorders are characterized by an earlier stage, a pre-psychotic stage termed prodrome, which is usually missed by clinicians
[25][26][25,26].
The pathophysiology of SCZ remains unclear, lacking a comprehensive view of the underlying neurobiological mechanisms, although some aspects are beginning to be clarified. Dopaminergic dysfunction has been one of the pathophysiological hypotheses defended for decades, under various formulations, and is supported by genetic findings
[27][31].
Hypo and hyperactivities of the dopaminergic system are seen in SCZ patients, and both are linked to the symptoms previously described
[28][29][32,33]. Additionally, other dysfunctions underlying the pathophysiology of SCZ, such as neurotransmitter signaling of glutamate, hypothalamic-pituitary-axonal (HPA) axis signaling, immune system dysregulation and synaptic plasticity anomalies have been reported
[19][29][30][19,33,34]. Changes in brain structures, which have also been proposed as etiologically relevant, are correlated with some of these alterations
[30][34].
1.3. The Search for Biomarkers
To improve knowledge about these complex disorders, “omics” approaches have emerged to shed light on disease pathogenesis and support a trustworthy way of predicting and diagnosing PD
[20][31][20,35]. With a vast potential associated, high-throughput omics technologies can be a solution to predict clinical endpoints, with the improvement of patient care and outcomes as the ultimate goal. However, the translation from research to a successful clinical omics-based test is far from the great potential of these approaches
[32][33][36,37].
The search for candidate biomarkers is one of the outputs of -omic studies. According to the National Institute of Health (NIH), a biological marker, generally just termed as a biomarker, is a “characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention”
[34][38]. The study of the brain and the associated disorders is complex since it presents a high degree of inter- and intra-cellular heterogeneity; so, different locations may have a distinct proteome due to modifications in different cell types and cellular networks. The CNS proteome can change even with minimal alterations in the normal course of its development and/or function
[35][36][39,40]. Proteomics can be a powerful tool since it can give a real-time evaluation of an individual state, health vs. disease, and, in an ideal scenario, predict the susceptibility to develop a specific mental disorder
[4][35][4,39]. The possibility of identifying and quantifying the proteins makes the proteomic approach more reliable for evaluating psychiatric diseases at different stages. Moreover, protein-based tests can offer the nearest view of the pathophysiological process behind PD since their expression and function are the results of what happens during post-transcriptional (e.g., alternative mRNA splicing) and post-translational events (e.g., phosphorylation, glycosylation, oxidation), as well as the interactions between them
[3][4][37][3,4,42].
2. Mass Spectrometry
Since its development, mass spectrometry (MS)-based technologies have been improved and, in recent decades, became a well-suited method for biomarker discovery, supporting the expansion of the proteomics field
[46,47][38][39]. The success of MS in proteomics is due to its specificity and sensitivity, which are mainly attributable to advances in liquid chromatography coupled to tandem MS (LC-MS/MS) approaches. This type of technology can reveal proteome insights at the composition, structure, and function level. Proteomics tools make it possible to evaluate the proteins in complex biological samples qualitatively and quantitatively (either relative or absolute)
[48,49][40][41].
In the beginning, successes in proteomics approaches were supported by two-dimensional gel electrophoresis (2-DE), with complex protein mixtures being separated by their molecular charge (isoelectric point) and mass (molecular weight) in the first and second dimensions, respectively. This approach calculates protein abundances based on stained protein spots’ intensities, followed by MS analysis for protein identification
[23,50,51][23][42][43]. Although improvements were made, other methodologies emerged to circumvent some of the previous technical drawbacks, namely to face the dynamic range limitations and the unsuitable separation and detection of some protein subtypes, such as membrane proteins
[23,48][23][40]. Throughout the years, improvements in proteomics approaches were achieved, and a variety of more in-depth MS-based methods were quickly applied to compare protein profiles, usually between control versus disease states. Considering this, there are two main groups within quantitative proteomics methods: (i) labeling techniques, which involve different isotopic labeling of samples, including chemical, enzymatic or metabolic labeling, followed by MS analysis; and (ii) label-free techniques, where the sample is individually analyzed without the addition of any other chemical compound. The newest quantitative approaches are regarded as versatile and cost-effective alternatives to labeled quantitation, having gained significant interest in recent years, mainly due to the development of more sensitive and reliable methods. Additionally, some techniques capable of detecting either relative or absolute peptide levels can provide a targeted MS approach and be used as a validation method
[51,52,53][43][44][45].
The absence of molecular biomarkers being used in the clinical environment and the increasing use of large proteomics screenings to search for SCZ biomarkers, allowed
uscholars to perform
this work, by providing a systematic review and meta-analysis on the use of MS-based methods in proteomic studies to assess biomarkers or a panel of biomarkers associated with SCZ based only on the analysis of peripheral fluids.
3. Mass Spectrometry Proteomics in Neuropsychiatric Disorder Biomarkers Assessment
A total of 217 proteins were identified as altered between SCZ and healthy control groups in peripheral fluids, including serum, plasma, PBMCs, sweat, and saliva.
Apolipoproteins (APOs) were the group of proteins mostly reported in SCZ vs. control studies as differentially expressed. In fact, ten studies reported the dysregulation of apolipoproteins [63,65,66,67,69,71,75,76,79,81][46][47][48][49][50][51][52][53][54][55]. APOs are very important in lipid homeostasis by transporting cholesterol and lipids between cells, having a well-established role in the transport and metabolism of lipids, and in inflammatory and immune response regulation [91,92][56][57]. This group of compounds has been indicated as potential candidates for psychiatric biomarkers, with several studies reporting altered levels of cholesterol and APOs in psychiatric disorders [57][58][59][92,93,94]. Accordingly, APOs alterations were associated with inflammatory response [67 [49][54][55],79,81], immune system [63[46][53],76], lipid metabolism [67[49][53],76], cardiovascular system [66][48], retinoid transport [81][55], and cognitive decline and underlying morphological changes [69][50].
APOA1 is the major protein component of the HDL fraction in plasma. Together with APOA2, APOA4, APOC1, and APOD, APOA1 is recognized for regulating the plasma levels of free fatty acids, having an important role in HDL and triglyceride-rich lipoprotein metabolism in the reverse cholesterol transport pathway [95][60]. APOA1 is also reported as having pro-immune and anti-inflammatory potential [91][56]. In all selected studies where it was identified as altered, ApoA1 level was reported to be reduced in schizophrenia patients compared to healthy subjects [69,76,79,81][50][53][54][55].
APOA2, the second most abundant protein in HDL fraction, is a key regulator of HDL metabolism
[95][60], although its inflammation role is not clearly defined, with different studies reporting it as having pro- and anti-inflammatory effects
[96][61]. APOA2 was identified as differentially expressed in four studies, being downregulated in SCZ patients in all studies
[67,69,76,79][49][50][53][54].
APOA4, a lipid-binding protein, is known to be involved in a broad spectrum of biological processes, including lipid metabolism, reverse cholesterol transport, atherosclerosis protection, and glucose hemostasis
[97][62]. APOA4 was identified as differentially expressed in four studies; however, it showed a heterogeneous behavior: downregulated in three studies
[67,69,79][49][50][54] and upregulated in only one study
[71][51].
The apolipoproteins APOC1, APOC2, APOC3, APOD, and APOE were identified in three studies as differently expressed, showing a general tendency of downregulation in SCZ patients except for APOE, which has a trend for upregulation. Of these, only for APOD, a soluble carrier protein of lipophilic molecules that is mostly expressed in neurons and glial cells within the central and peripheral nervous system
[98][63], the results were consistent in all three studies, and it was identified as decreased in SCZ patients
[58,65,72][64][47][65]. A trend of downregulated behavior was identified for APOC1 (the smallest of all APOs, participating in lipid transport and metabolism)
[67[49][54],
79], APOC2 (a small exchangeable apolipoprotein found on triglyceride-rich lipoprotein particles)
[71[51][54],
79], and APOC3 (an APO capable of inhibiting lipoprotein lipase and hepatic lipase)
[67[49][51],
71], in two out of three studies.
APOF
[69[50][52],
75], APOH
[65[47][49],
67], and APOL1
[69,75][50][52] had a similar behavior: upregulated in the two studies. For APOB, no clear trend was observed, with one study reporting its increase
[66][48] and another a decrease
[69][50] in SCZ patients.
RET4 is mainly expressed in the liver with a primary function to transport retinol (vitamin A) from the liver to peripheral tissues, with retinol being essential for the brain to facilitate learning, memory, and cognition
[99][66]. Retinoid signaling plays a vital role in immune cell function. Accordingly, it is suggested that factors that affect this system could have important implications for SCZ and other psychiatric disorders-associated inflammatory stress
[100][67].
ANT3, a glycoprotein anticoagulant mainly produced in the liver that exerts anticoagulant and anti-inflammatory effects by targeting activated thrombin and other blood coagulation factors
[102][68], was identified as increased in SCZ patients
[67,69,81][49][50][55].
FCN3 is a ficolin, a protein containing both a collagen-like domain and a fibrinogen-like domain with a specific binding affinity for N-acetylglucosamine. FCN3 can complex with mannose-associated serine proteases to activate the complement pathway
[103][69], being ficolins’ activation already reported as a potential biomarker of the severity of schizophrenia
[104][70]. In the selected studies, FCN3 was also identified in three studies as upregulated FC
[67,69,76][49][50][71].
The immune system and inflammatory response were the most identified biological processes altered in SCZ patients
[63,67,68,74,76,79,80][46][49][72][73][71][54][74]. These results agree with current knowledge about SCZ, associating the immune system and inflammatory response with the SCZ pathophysiology
[106,107,108][75][76][77]. In fact, a wide range of immune alterations has been reported in SCZ patients, such as elevated levels of cytokines and inflammation markers, abnormalities of the blood-brain barrier, CNS inflammation, and increased autoantibody reactivity
[107][76].
Several other mechanisms have also been linked to SCZ, including mitochondrial dysfunction, energy metabolism processes, complement and coagulation cascades, oxidative stress, transport, morphological changes, cognitive impairment, lipid metabolism, and hypothalamic–pituitary–adrenal (HPA) axis over-activation
[87,108][77][78].