Evolution of Glutamate Metabolism via GLUD2 Enhances Lactate-Dependent Synaptic Plasticity and Complex Cognition

Abstract

Human evolution is characterized by rapid brain enlargement and the emergence of unique cognitive abilities. Besides its distinctive cytoarchitectural organization and extensive inter-neuronal connectivity, the human brain is also defined by high rates of synaptic, mainly glutamatergic, transmission, and energy utilization. While these adaptations' origins remain elusive, evolutionary changes occurred in synaptic glutamate metabolism in the common ancestor of humans and apes via the emergence of GLUD2, a gene encoding the human glutamate dehydrogenase 2 (hGDH2) isoenzyme. Driven by positive selection, hGDH2 became adapted to function upon intense excitatory firing, a process central to the long-term strengthening of synaptic connections. It also gained expression in brain astrocytes and cortical pyramidal neurons, including the CA1-CA3 hippocampal cells, neurons crucial to cognition. In mice transgenic for GLUD2, theta-burst-evoked long-term potentiation (LTP) is markedly enhanced in hippocampal CA3-CA1 synapses, with patch-clamp recordings from CA1 pyramidal neurons revealing increased sNMDA receptor currents. D-lactate blocked LTP enhancement, implying that glutamate metabolism via hGDH2 potentiates L-lactate-dependent glia-neuron interaction, a process essential to memory consolidation. The transgenic (Tg) mice exhibited increased dendritic spine density/synaptogenesis in the hippocampus and improved complex cognitive functions. Hence, enhancement of neuron-glia communication, via GLUD2 evolution, likely contributed to human cognitive advancement by potentiating synaptic plasticity and inter-neuronal connectivity.

Keywords: CA1/CA3 LTP; GLUD2; glutamate; human brain evolution; lactate; synaptic plasticity.

Figure 3

Schematic representation of a tripartite glutamatergic synapse in the hippocampi of transgenic (GLUD2) and wild-type (Wt) mice. Glutamate (GLU), released from presynaptic nerve terminals during neurotransmission, acts on post-synaptic NMDA and AMPA receptors. Synaptic glutamate is then rapidly removed from the synaptic cleft by uptake into the surrounding astrocytes (small arrow), where it is in part transported into the mitochondria. In the Wt mice, glutamate is converted to α-ketoglutarate via the endogenous mGDH1, whereas in the GLUD2 Tg mice, this reaction is catalyzed by both the expressed hGDH2 and the mGDH1. α-ketoglutarate is subsequently metabolized through the TCA cycle, giving rise to lactate in the cytoplasm [66]. Indeed, metabolic studies using [U-13 C] glutamate have shown substantial incorporation of the label into lactate in a pattern that could only arise via metabolism of [U-13C] glutamate through the GDH-TCA cycle pathway [66]. Also, the observed labeling pattern of TCA cycle intermediates, such as citrate, permits the conclusion that part of citrate is exported to the cytoplasm, where it is catabolized by the ATP citrate lyase (ACLY) to oxaloacetate (OXAA) and acetyl-CoA [66]. Subsequently, OXAA is converted via the cytosolic malic enzyme (ME) to pyruvate, which gives rise to lactate by the action of LDH. Upon neuronal excitation, astrocytes release increased amounts of lactate (large arrow), which facilitates NMDA receptor signaling. These lactate-mediated effects are enhanced in the GLUD2 Tg mice, leading to increased synaptic plasticity and synaptogenesis [20].

Figure 1

Divergence of the GLUD2 gene among members of the hominoid radiation. Phylogenetic tree (adapted from references [22] and [34]) showing GLUD2 evolution in primates. It has been estimated that GLUD2 emerged about −23 Mya and evolved along the human and ape lineages. Depicted here are amino acid substitutions that occurred on the different branches (A,B, B,C, C,D, and D,E) of the phylogenetic tree. Amino acid replacements thought to have evolved under positive selection are green-colored, whereas those that differ between the GDH2 of the human and the GDH2 of non-human apes are blue-colored. Besides the amino acid replacements, shown here to have occurred in the gibbon lineage after its separation from that of the human and great apes, additional changes have been recently detected in Hylobates moloch, Symphalangus, and Nomascus [33]. Divergence times in millions of years (Mya) are shown at the bottom of the figure (not to scale). These time estimates are based on References [35,36,37,38,39]. In contrast to GLUD2, its ancestor, the GLUD1 gene, remained conserved [22].

Figure 2

Structural model of human GDH2. Graphic representation of the 3D structure of the hGDH2 hexamer, comprising two trimers (1A,B) (PDB code: 6G2U). Each color corresponds to one of the six identical subunits. In (2), one monomer is depicted. The NAD+-binding domain, the glutamate-binding domain, the antenna, and the pivot helix are identified. In (3), the precise locations of residues (Ser-443 and Ala-456) that provide hGDH2 with unique properties are shown in green. Ser-443 is located in the small C-terminal α-helix of the antenna and Ala-456 is in the pivot helix. Also, the precise location of the hGDH2 mutations Glu441Arg, Ser445Leu, Ser445Ala, Ser448Pro, Lys450Glu, and Hist454Tyr that affect the basal activity and/or regulation (see text and Ref [35]) are shown in blue, except for the Ser445Ala change, which modifies Parkinson’s disease onset. This is shown in red. Interestingly, except for Ser445Ala, all these mutations occur in hGDH1, attenuating GTP inhibition and causing the HI/HA syndrome. These observations imply that the same mutations in hGDH1 and hGDH2 can have diverse functional consequences. The PyMOL Molecular Graphics System, Version 2.5 Schrödinger, LLC was used to create the graphics. This was obtained from Schrödinger, Inc., New York, NY, USA.