1. Introduction
Alterations in immune function and reactivity in schizophrenia (SZ) and mood disorders have been found and extensively reviewed [1,2,3]. Reports of immune dysfunction in transcriptomic and proteomic studies of brain tissue have been reported primarily in SZ (reviewed in [4,5]). The majority of studies have analyzed antibodies, cytokines, and blood cell types, and indexed these markers as either an indicator of pro- or anti-inflammatory reactions or secondary responses. These studies have shown a remarkable diversity of findings and some gains in understanding relationships between psychopathology and immune function. For example, in schizophrenia, a meta-analysis of cytokine studies showed some consensus in acute relapse and first-episode psychosis for some cytokines (IL-1β, IL-6, and TGF-β), which might be state markers for acute exacerbations, whereas others (IL-12, IFN-γ, TNF-α, and sIL-2R) may be trait markers. One overview of these findings is that perhaps a subset of patients with psychopathology can be found with alterations in immune response [6]. Cell-type specificity was also analyzed and there were increases in markers for natural killer cells and CD4+ lymphocytes in schizophrenia [1]. More strikingly, some forms of autoimmune illness involve antibodies to glutamatergic receptors in brain (for review [7]) and are associated with both encephalitis and psychiatric disorders.
Epidemiological studies implicated immune mechanisms as risk factors for developing schizophrenia in offspring (recently reviewed in [8]). Some of the epidemiological studies have found a large odds ratio for factors such as season of birth, maternal viral infection, cytokine alterations, and antibody/viral titer for developing schizophrenia or other psychotic disorders. Other studies point to endogenous biomolecules, such as human endogenous retroviruses (HERV) products, to be possible underlying antigens that could be made in brain cells and trigger immune responses [9,10,11].

1.1. Genetic Loci Related to Major Histocompatibility (MHC) Region
Multiple genome-wide association (GWAS) studies have reported significant MHC single-nucleotide polymorphisms (SNPs) associated with schizophrenia, bipolar disorder, cognition, or hippocampal volume [12,13,14,15,16,17,18]. Finding highly associated SNPs within this region is a first step towards locating causative gene(s). Currently, these associations with schizophrenia do not necessarily imply immunity as the causative mechanism underlying the association, as the linkage disequilibrium is complicated across different populations. This highly replicated region encompasses hundreds of genes, some that are non-immune, in long-range linkage disequilibrium (LD), with extensive allelic heterogeneity. The GWAS studies to date have not found the gene(s) responsible or causative variants in the MHC locus, although the HLA-B and HLA-DRB1 loci were protective, and the HLA-C increased risk for schizophrenia [12]. One recent study of the C4 gene in the MHC locus showed an association with increased brain expression and risk of schizophrenia [19]. Thus, additional evidence of gene expression in the MHC locus needs to be pursued to directly examine some of the immune genes from persons with psychiatric disorders.

1.2. Gene Expression in Schizophrenia, Major Depression, and Bipolar Disorder
The correlated expression of genes in brain or in peripheral tissues forms a co-expression network. Gene−gene expression correlation matrices can be built into coherent network connections. An expression quantitative trait locus (eQTL) study of a developing brain series mapped the significant loci to co-expression networks that were highly enriched for significant GWAS findings in schizophrenia [20]. The co-expression network was formed around a hub of several interconnected MHC genes, increasing interest to find the truly causative genes, and not merely candidate genes.
A second gene expression co-network of a module of genes that was differentially expressed between SZ and controls in whole blood also formed a network involving the MHC locus [20]. This co-expression network module was independent of antipsychotic medication in SZ subjects, and included multiple members of the MHC locus: heat shock protein 90 kDa α (cytosolic), class B member 1 (HSP90AB1); ring finger protein 1 (RING1); casein kinase 2, β polypeptide (CSNK2B); tubulin, β (TUBB); and ATP-binding cassette, sub-family F, member 1 (ABCF1). These two studies show the feasibility of expression studies to inform existing GWASes of differential regulation in brain and whole blood.
Microarray, quantitative PCR (qPCR) and RNA (cDNA) sequencing are useful techniques that examine gene expression in biological samples. The experiments that are reported begin with an inquiry into bipolar disorder using Affymetrix human exon 1.0 ST microarrays, to compare a matched cohort of bipolar type I subjects to controls using anterior cingulate cortex postmortem samples. The anterior cingulate is a region profoundly implicated in mood dysregulation. To validate the exon microarray results, qPCR experiments were performed in a larger cohort using five different brain regions, cell lines, and blood samples.

1.3. MHC Class II Genes
The purpose of the current study was to examine immune gene expression focusing on CD74, HLA-DRB1, and HLA-DPA1, which are integral components of the MHC Class II pathway (Figure 1). Using a cohort of postmortem brain tissue, we compared three diagnoses (schizophrenia, bipolar disorder type I, and major depressive disorder) to controls.
The three MHC Class II genes of relevance for this paper share well-described molecular functions within the immune system of antigen processing and presentation (Figure 1). CD74 has multiple roles in the immune system: it is a transmembrane protein that associates to MHC Class II α and β chains, and is a receptor of the proinflammatory cytokine macrophage migration inhibitory factor which modulates macrophage and monocyte activation. HLA-DPA1 dimers associate with CD74 trimers forming a heterononamer from NCBI RefSeq. The level of HLA-DRB1 gene expression is complex due to the multiple alleles and involvement of polymorphisms of the HLA-DRB1 promoter. Susceptibility to rheumatoid arthritis (RA) is associated with defined HLA-DRB1 alleles that are related to expression, and there is thought to be a negative correlation between incidences of RA and schizophrenia [21,22].
HLA-DPA1 SNPs, in particular rs3077, are associated with risk for persistent infection with the hepatitis B and C viruses, first noted in [23] and then subsequently replicated in multiple reports. Presumably, carriers of HLA-DPA1 coding variants along with variants in HLA-DPB and HLA-DRB have different sensitivities to viral antigen binding, clearing the virus by altering the antigen-presenting capabilities, and building effective immune response.
Taken together, the MHC Class II genes have multiple alleles and tissue-specific expression patterns which have not been described, and these transcripts have been quantitatively assessed in brain. Previously, MHC Class I genes were shown to be localized to neurons [24], while the expression of MHC Class II molecules in human brain has been found on microglia cells using OX-6 immunohistochemistry. Thus, we focused an investigation on members of the MHC Class II genes in multiple brain regions and disorders.
The MHC Class II are heterodimers composed of α and β transmembrane chains, such as HLA-DPA1 and HLA-DPB1, the latter a dimer that associates with the invariant chain of CD74 which forms a trimer in the endoplasmic reticulum. The trimer traffics to the Golgi compartment and then is transported to the late endosomic compartments, thereby loading the antigen peptide [25]. CD74 acts as a chaperone that regulates antigen presentation. Invariant chain of CD74 involves binding to the HLA-DPA1 and HLA-DPB1 dimers. After trafficking and targeting antigen, the invariant chain is released from MHC-II and then the antigen peptide binds [26,27]. The antigen-bound MHC-II complex moves to the plasma membrane. The plasma membrane-bound antigen-MHC II (pMHCII) interacts with T-cell receptor (TCR). TCR recognizes the antigen only when presented by foreign MHC molecules. TCR activates the pMHCII complex, and facilitate CD4 T-cell stimulation on the cell surface [28]. Crystal structure shows the ternary model of CD4-pMHCII-TCR that appears V-shaped with pMHC II at apex, but CD4 and TCR do not have direct contact. The signaling of CD4 and TCR is coordinated around the pMHCII [29]. Therefore, MHCII genes play an essential role in immune responses.