B. R. Groveman, Department of Biomedical Sciences, Florida State University, Tallahassee, FL 32306, USA Fax: +1 406 363 9286 Tel: +1 406 363 9341 E-mail: email@example.com
Src family kinases (SFKs) play critical roles in the regulation of many cellular functions by growth factors, G-protein-coupled receptors and ligand-gated ion channels. Recent data have shown that SFKs serve as a convergent point of multiple signaling pathways regulating N-methyl-d-aspartate (NMDA) receptors in the central nervous system. Multiple SFK molecules, such as Src and Fyn, closely associate with their substrate, NMDA receptors, via indirect and direct binding mechanisms. The NMDA receptor is associated with an SFK signaling complex consisting of SFKs; the SFK-activating phosphatase, protein tyrosine phosphatase α; and the SFK-inactivating kinase, C-terminal Src kinase. Early studies have demonstrated that intramolecular interactions with the SH2 or SH3 domain lock SFKs in a closed conformation. Disruption of the interdomain interactions can induce the activation of SFKs with multiple signaling pathways involved in regulation of this process. The enzyme activity of SFKs appears ‘graded’, exhibiting different levels coinciding with activation states. It has also been proposed that the SH2 and SH3 domains may stimulate catalytic activity of protein tyrosine kinases, such as Abl. Recently, it has been found that the enzyme activity of neuronal Src protein is associated with its stability, and that the SH2 and SH3 domain interactions may act not only to constrain the activation of neuronal Src, but also to regulate the enzyme activity of active neuronal Src. Collectively, these findings demonstrate novel mechanisms underlying the regulation of SFKs.
Src family kinases (SFKs) associate with multiple types of voltage- or ligand-gated ion channels expressed in the peripheral and central nervous system (CNS). Through regulating the activity of voltage-gated ion channels, SFKs such as Src and Fyn are involved in the regulation of neuronal excitability and activity [1–6]. The regulation of nicotinic acetylcholine receptors by Src is reported to play a critical role in the regulation of neuronal survival . The interaction mediated by the proline-rich region of tau protein and the SH3 domain of Fyn or Src represents an important mechanism underlying a role of SFKs in neuropathogenesis . Src, Fyn and Lck are found to play roles in neuronal growth and oligodendrocyte maturation [9–11]. Mice lacking Fyn show severe myelin deficits [10,11].
Activation of protein kinase C (PKC) correlates with all phases of learning, including acquisition, consolidation and reconsolidation . Detailed mechanistic studies show that PKC activates SFKs and thereby induces tyrosine phosphorylation of N-methyl-d-aspartate (NMDA) receptors , leading to a rapid increase in surface expression of NMDA receptors . This PKC–SFK pathway has also been found to be important in the regulation of NMDA receptor functions by G-protein-coupled receptors . NMDA receptors are known for their roles in synaptogenesis and synaptic plasticity. Vascular endothelial growth factor stimulation activates SFKs, which increases tyrosine phosphorylation of the NMDA receptor NR2B subunit and promotes neuronal migration and synaptogenesis . Chemical transmitters released from astrocytes (termed gliotransmitters) modulate synaptic transmission and neuronal function. Blocking gliotransmission may inhibit adenosine A1 receptors and impair sleep homeostasis [17–19]. Furthermore, adenosine A1 receptor antagonists reduce the tyrosine phosphorylation of Src and the NMDA receptor NR2B subunit, decrease the surface expression of the NR2B subunit, and inhibit NMDA receptor-mediated miniature excitatory postsynaptic currents . Mice lacking Fyn show impairment of various behaviors, such as spatial learning, suckling, emotional behaviors and ethanol sensitivity . These mice display defects in synaptic functions mediated by both glutamate and γ-aminobutyrate .
γ-aminobutyrate is the principal inhibitory neurotransmitter in the CNS. γ-aminobutyrate type A (GABAA) receptors, which are composed of α, β and γ2 subunits, are the major γ-aminobutyrate receptors responsible for fast synaptic inhibition in the brain [22,23]. Activation of SFKs may increase the tyrosine phopshorylation of the β2/3 and γ2 subunits and potentiate GABAA receptor-mediated whole-cell and synaptic responses in central neurons [22–24]. Furthermore, Fyn-induced phosphorylation of tyrosine residues 365 and 367 (Y365/7) within the GABAA receptor γ2 subunit negatively regulates the endocytosis of GABAA receptors and enhances synaptic inhibition [22,25].
Fyn interacts with the γ2 subunit in a phosphorylation-dependent manner. Tyrosine phosphorylation of the γ2 subunit is significantly reduced in the hippocampus of mice lacking Fyn, suggesting that Fyn is an important kinase that contributes to the phosphorylation of this subunit in vivo . A recent study  shows that cerebral ischemia induces both the increase of excitation and the decrease of inhibition, which results in neuronal excitotoxicity. Activation of γ-aminobutyrate receptors attenuates neuronal death through inhibiting the tyrosine phosphorylation of the NMDA receptor NR2A subunit by Src after cerebral ischemia and reperfusion .
Interactions between dopamine and glutamate in the prefrontal cortex are known to be essential for cognitive functions, such as working memory [27,28]. Activation of dopamine D1 receptors may significantly increase the expression of NR2B subunits . This effect can be blocked by knockdown of Fyn . Administration of l-3,4-dihydroxyphenylalanine in rats with unilateral dopamine denervation induces a progressive increase of contraversive circling behavior and modulates the expression of Src, Lyn and PKC . Lyn, a member of the SFKs, negatively regulates the release of dopamine in the mesolimbic system and the rewarding properties of alcohol . Regrettably, because of space limitations, this minireview is not able to encompass all of the interesting and undoubtedly important data regarding SFKs in the nervous system. Thus, the focus is on the association of Src with NMDA receptors and the regulation of stability and activity of neuronal Src (n-Src) protein.
The association of Src, the Src activator [protein tyrosine phosphatase α (PTPα)] and inhibitor [C-terminal Src kinase (Csk)] with NMDA receptors
Recent studies have shown that SFKs serve as a convergent point of multiple, diverse signaling pathways that regulate NMDA receptor function in the CNS [32,33]. More detailed studies have shown that Src may form a complex with NMDA receptors via the binding of its SH2 domain to a region N-terminal of the PDZ1 domain of postsynaptic density protein 95 (PSD-95) , whereas Fyn participates via the binding of its SH2 domain to the PSD-95 PDZ3 domain . The PDZ domain is a common structural domain consisting of 80–90 amino acids [36,37]. PSD-95 comprises three PDZ domains, an SH3 domain and a guanylate kinase-like domain [36,37]. PSD-95 is almost exclusively located in the postsynaptic density of neurons. Both the PSD-95 PDZ1 and PDZ2 domains may bind to the ESDV motif at the C terminus of NMDA receptor NR2 subunits [38,39]. PSD-95 protein has been found to be a negative regulator of Src in the regulation of NMDA receptors . Src has also been shown to interact with the NMDA receptor complex through its unique domain binding to NADH dehydrogenase subunit 2 . Preventing this unique domain interaction does not alter the catalytic activity or the SH2 or SH3 binding of Src but, instead, blocks Src-mediated NMDA receptor up-regulation .
The interaction of n-Src and the C terminus of the NMDA receptor NR2A subunit in vitro was examined recently using surface plasmon resonance . Wild-type n-Src protein specifically binds to the NR2A C terminus protein fragment with a 100 nm range binding affinity, and such binding does not appear to be affected by disruption of the kinase, SH2 and/or SH3 domains of n-Src . Although the minimum sequence(s) required for NMDA receptor-Src binding remain to be clarified, it can now be concluded that, via multiple (direct and indirect) mechanisms, numerous SFK molecules are closely linked and ready to act on their substrate: NMDA receptors (Fig. 1).
Protein tyrosine phosphatase α (PTPα) is one of the enzymes that can dephosphorylate C-terminal tyrosine [Y527 in chicken cellular Src (c-Src)] of SFKs and thereby increase their activity [43–48]. This phosphatase is highly expressed in the CNS and may associate with NMDA receptors through the binding of its distal phosphatase domain (D2) to the PDZ2 domain of PSD-95  (Fig. 1). Removal of PTPα abolishes the constitutive regulation of NMDA receptors by Src without affecting the Src association with NMDA receptors. Furthermore, through SFKs, the intracellular application of PTPα catalytic domains (D1 + D2) into neurons enhances NMDA receptor-mediated synaptic responses and induces a long-term potentiation in CA1 hippocampal neurons . Consistently, PTPα knockout mice display both reduced long-term potentiation in CA1 hippocampal neurons  and impaired spatial learning [50,51]. Thus, the SFK activator PTPα is also closely associated with NMDA receptors and acts endogenously as a positive regulator that is necessary for the initiation and maintenance of NMDA receptor up-regulation by SFKs (Fig. 1).
Conversely, phosphorylation of the C terminus tyrosine of SFKs carried out by Csks results in a reduction of SFK activity [52,53]. So far, two members have been identified in the Csk family: Csk and Csk homologous kinase (Chk) [52,53]. It has been reported that, by inhibiting SFK activity, Csk regulates neural development and differentiation [54–57], synaptic plasticity [58,59], toxicity  and nociceptor function . The binding of Csk to the phosphotyrosine of the substrates of SFKs has been found to be a key mechanism allowing Csk to move into proximity of (and thereby phosphorylate and inhibit) SFKs [59,62,63]. Csk associates with both the NR1 and NR2 subunits of NMDA receptors in a Src activity-dependent manner [59,64] and serves as a ‘brake’ blocking the up-regulation of NMDA receptors and the induction of long-term potentiation [59,64]. By contrast to the NR2A and NR2B subunits, no apparent tyrosine phosphorylation is found in the NR1 subunit of NMDA receptors in the brain [65,66]. However, truncation of the NR1 subunit C terminus, which contains only one tyrosine (Y837), significantly reduces the Csk association with the NR1-1a/NR2A receptor complex expressed in HEK-293 cells . Furthermore, no effect from Csk application can be produced on NR1-1a/NR2A receptors with either truncation of the NR2A C terminus at residue 857 or mutation of Y837 in the NR1-1a subunit to phenylalanine [59,64]. These studies indicate that both the NR1 and NR2 subunits are critically involved in the regulation of NMDA receptor functions by Csk [59,64] (Fig. 1). Taken together, it has been demonstrated that NMDA receptors are closely associated with a SFK signaling complex consisting of SFKs; the SFK-activating phosphatase, PTPα; and the SFK-inactivating kinase, Csk (Fig. 1). In this SFK signaling complex, SFKs are the principal effectors regulating NMDA receptor activity. PTPα acts as driving force for the initiation and maintenance of the SFK activity required for the constitutive regulation of NMDA receptors . The dynamically recruited Csk is responsible for depressing activated SFKs and thereby down-regulating NMDA receptors [59,64] (Fig. 1). This well controlled molecular apparatus ensures that SFK activity and the enhancement of NMDA receptor function are under control, yet remain responsive to appropriate stimuli (Fig. 1).
The regulation of Src protein stability and activity
Early studies have demonstrated that interdomain interactions of SFKs, stabilized by binding of the SH2 domain to C-terminal tail containing phosphorylated tyrosine (Y527 in chicken c-Src) and/or the binding of SH3 domain to a polyproline type II helix in the linker region between the SH2 domain and the catalytic domain, lock the molecules in a closed conformation, disrupt the kinase active site and inactivate SFKs [43,47,67–69] (Fig. 2A). In this auto-inhibited state, the SH2 and SH3 domains turn inward to help lock the structure in an inactive conformation . Extrinsic ligands for the SH2 or SH3 domains can induce activation of SFKs by disrupting these intramolecular interactions [71–74]. These studies have not only demonstrated that multiple mechanisms are involved in the regulation of SFK activation, but also revealed that the enzyme activity of SFKs is ‘graded’, exhibiting different levels coinciding with activation states .
Protein unfolding is an important regulatory mechanism of protein functions in vivo, including protein translocation and degradation [75–77]. Although the resistance of proteins to unraveling in vivo may not be the same as that determined by the thermodynamic or kinetic stability measured in vitro by heat or solvent denaturation , the active form of Src (e.g. v-Src or c-Src/Y527F) is also found to be more vulnerable to ubiquitin-dependent degradation  and has a shorter half-life in cells compared to wild-type or inactive Src . v-Src expressed and purified from Escherichia coli is minimally phosphorylated at tyrosine 416, has a different secondary structure content, and is less stable compared to c-Src . In n-Src purified from bacteria, both Y424 in the activation loop and Y535 in the C-terminus can be autophosphorylated in vitro , which is consistent with the findings of earlier studies [43,47,67,68,82]. The secondary structure of n-Src mutants, including constitutively active n-Src and inactive n-Src, is almost identical to that of the wild-type n-Src .
It has been found that destabilizing adenylate kinase by urea enhances the catalytic activity without changing the secondary and tertiary structure of the enzyme . Similarly, reducing n-Src stability by adding urea at low concentrations produces no detectable change in the overall structure, although it potentiates the catalytic activity of n-Src . Detailed studies have shown that altering n-Src enzyme activity either pharmacologically or genetically is consistently associated with changes in n-Src stability, as determined by far-UV CD and light scattering . This decreased stability–increased activity scenario follows the premise of the ‘stability–function theory’, which predicts an inverse relationship between activity and stability when residues important for function are replaced . Such a stability–function relationship has been found in a number of other proteins, including barnase , barstar , staphylococcal nuclease [87,88], T4 lysozyme , thioredoxin  and AmpC β-lactamases [90,91]. A similar trend has also been observed for c-Src upon myristoylation, which decreased the intracellular stability of c-Src at the same time as up-regulating its activity . Although the underlying mechanisms remain to be clarified, these findings strongly suggest that Src activity is closely associated with its stability and that modulating Src stability might prove to be an effective way of regulating Src activity  (Fig. 2).
The SH3 and SH2 domains in Src have been recognized to be involved in the negative regulation of Src by stabilizing the protein through internal substrate interactions [43,47,67–69]. However, based on findings in Fps, Abl and ZAP-70 kinases, it has been proposed that the SH2 domains of cytoplasmic tyrosine kinases can have two general types of regulatory effects [93–96]. In the kinase-active state, the SH2 domain promotes an active conformation of the adjacent catalytic domain. Conversely, the SH2 domain can act in conjunction with an additional SH2 or SH3 domain to maintain an inactive state through intramolecular interactions with the catalytic domain [95,96]. For example, upon activation of Abl, the Abl SH2 and SH3 domains may adopt an extended conformation, with the SH2 domain contacting the tip of the kinase N-terminal lobe through a conserved isoleucine residue [95,96]. This latter interface allows the SH2 domain to stimulate the catalytic activity of Abl [95,96].
Within purified recombinant n-Src protein in an active state (n-Src/Y535F), disruption of either the SH2 or SH3 domain significantly reduces the autophosphorylation of the activation loop in the kinase domain, as well as the enzyme activity measured in vitro and in cells expressing these constructs . Compared to the SH3 domain, the SH2 domain appears to play a greater role in the activity regulation of active n-Src . Consistently, incubation with the constitutively active n-Src protein produces higher phosphorylation in the the NR2A C terminus protein fragment than that produced by incubation with wild-type n-Src. Nevertheless, the SH3 and/or SH2 domain disruptions in active Src significantly reduce NR2A C terminus protein phosphorylation . The disruption of the SH3 and SH2 domains also significantly impacts the Src regulation of NR1-1a/NR2A receptor activity expressed in HEK-293 cells . Thus, the SH2 and SH3 domain interactions are found to act not only to constrain the activation of n-Src, but also to promote the enzyme activity of the active n-Src (Fig. 2B).
Research efforts conducted over more than a century have rendered Src as one of the best understood signaling molecules in a biological system [43,47,82,97–99]. SFKs play critical roles in the regulation of many cellular functions. Dysfunctions of SFKs can lead to memory impairment, osteoporosis, immune deficiencies and oncogenesis [21,99,100]. SFKs have become important targets for the development of therapeutic approaches to many diseases, including several types of cancer [99,101–103]; inflammatory and neuropathic pain ; and noise-induced hearing loss [105,106].
However, many fundamental questions remain to be addressed. For example, it is still not fully understood how SFKs interact with their substrates, such as NMDA receptors; how each domain of SFKs functions when SFKs act on their substrates; and what mechanisms may underlie the association of enzyme activity and stability of SFKs. Further investigations aiming to answer these questions are critical for understanding the regulation of activity-dependent neuroplasticity in the nervous system and for developing effective approaches to treat clinical problems, such as cancer, chronic pain and mental disorders. We expect that the systematic in vitro and in vivo (in neurons) characterization of the structure and functions of various Src mutant proteins will provide novel insights for our understanding of the regulation of SFK functions in the nervous system.
This work was supported by a grant from NIH (5R01 NS053567-04) to X.M.Y. S.F. was supported by the State Scholarship Fund of China (2009845013).