CD38, CD157, and RAGE as Molecular Determinants for Social Behavior

Abstract

Recent studies provide evidence to support that cluster of differentiation 38 (CD38) and CD157 meaningfully act in the brain as neuroregulators. They primarily affect social behaviors. Social behaviors are impaired in Cd38 and Cd157 knockout mice. Single-nucleotide polymorphisms of the CD38 and CD157/BST1 genes are associated with multiple neurological and psychiatric conditions, including autism spectrum disorder, Parkinson's disease, and schizophrenia. In addition, both antigens are related to infectious and immunoregulational processes. The most important clues to demonstrate how these molecules play a role in the brain are oxytocin (OT) and the OT system. OT is axo-dendritically secreted into the brain from OT-containing neurons and causes activation of OT receptors mainly on hypothalamic neurons. Here, we overview the CD38/CD157-dependent OT release mechanism as the initiation step for social behavior. The receptor for advanced glycation end-products (RAGE) is a newly identified molecule as an OT binding protein and serves as a transporter of OT to the brain, crossing over the blood-brain barrier, resulting in the regulation of brain OT levels. We point out new roles of CD38 and CD157 during neuronal development and aging in relation to nicotinamide adenine dinucleotide⁺ levels in embryonic and adult nervous systems. Finally, we discuss how CD38, CD157, and RAGE are crucial for social recognition and behavior in daily life.

Keywords: CD157; CD38; NAD; RAGE; anxiety; autism; cyclic ADP-ribose (cADPR); oxytocin transporter; social memory.

Figure 1

Cluster of differentiation 157 (CD157) expression analyzed by immunofluorescence staining in the embryonic mouse brain. Representative images were obtained from the E17 embryonic hypothalamus near the third ventricle (3V). CD157 is stained in green, nestin in red, and the nucleus in blue. Two merged images show the expression of CD157 in neural stem cells. Scale bar, 100 and 30 μm for the four left and enlarged images, respectively.

Figure 2

Identification of Ca²⁺-calmodulin-dependent protein kinase II (CAMKIIβ) as the human CD38 (hCD38)-associated protein by MALDI-TOF. The domain structures of the canonical sequence of human CAMKIIβ (isoform-4: UniProtKB/Swiss-Prot, Q13554) are shown as squares in the middle. The kinase domain (14–272) is shown in blue. The regulatory tail (273–315), linker region/insertions, and the association domain (532–664) are also depicted. The corresponding portions of “Chain A, crystal structure of CAMKIIβ isoform 1” (CAMK IiB isoform 1) and isoform CRA_A are marked as blue and green bars, respectively. The numbers 1–3 and 4 indicate the domains of CAMKIIβ at which CD38 binds, and sequences of four tryptic peptides of CAMKIIβ identified by MALDI-TOF/MS, respectively.

Figure 3

Effect of a CAMKII inhibitor on wound-induced migration facilitated by the expression of human influenza hemagglutinin (HA)-hCD38. (A,B) Human embryonic kidney 293T (HEK293T) cells grown on plastic dishes were transfected with HA-hCD38 complementary DNA (cDNA) constructs. One day after transfection, a cell-free area was formed by scratching the plastic dishes, and they were treated with 0.1% dimethyl sulfoxide (DMSO) solution (control, A) or with 1 µM KN-62, a CaM kinase inhibitor, dissolved in 0.1% DMSO solution (B). At 24 h after scratching, they were fixed, permeabilized, and stained with HA monoclonal antibody (mAb). The right panels are merged images of fluorescent (Left) and phase-contrast (Middle) pictures. Scale bar, 100 µm. (C) Quantification of cell migration using the monolayer wound healing assay. The degree of migration was expressed as a reduction of the cell-free area by occupied cells. Each bar corresponds to 293T cells transfected with (green bars) or without (yellow bars) 1 µM of KN-62 (+KN). Each bar is the mean ± standard error of the mean (SEM) in seven experiments. Two-way ANOVA, F_3,7 = 7.98, p = 0.0001. A significant difference was seen in the conditions of CD38 transfection and KN treatment, but no significant interaction between transfection of CD38 and treatment with KN-62 was observed (p = 0.18). Bonferroni’s post hoc analysis revealed a significant difference from the KN-free control value (*p < 0.05).

Figure 4

Hypotheses of CD38 and CaMKII interaction. Two models represent the possible means of activation of CaMKII by binding of calmodulin (CaM) and Ca²⁺ to the regulatory domain of CaMKII. Ca²⁺ is increased from ryanodine receptors bound by cyclic ADP-ribose (cADPR). cADPR is produced by either type II or type III CD38.

Figure 5

Scheme of CD38- and CD157-mediated pathways related to social behavior. The scheme shows the possible molecular mechanisms for increases in brain oxytocin (OT). One is OT release due to CD38 and CD157. Another is owing to receptor for advanced glycation end-products (RAGE)-dependent transport of OT into the brain. CD38- and CD157-dependent social behaviors are listed. The increased OT may trigger prosocial behavior.

Figure 6

Scheme of the recruitment of OT in the brain extracellular space. OT in the blood is transported by RAGE on the endothelial cells of the blood–brain barrier (BBB). The OT in oxytocinergic neurons is secreted into the extracellular space via CD38 or CD157 from dendrites and axon collaterals. At the axon terminals, OT is released into the portal vein at the posterior pituitary. The turnover of OT may be a basis of social behavior.