Directed evolution of three-finger toxin to produce serine protease inhibitors.

Directed evolution is a very popular strategy for improving biophysical properties and even for generating proteins with novel functions. Recent advances in combinatorial protein engineering mean it is now possible to develop protein scaffolds that could substitute for whole antibody-associated properties as emerging therapeutic proteins. In particular, disulfide-rich proteins are attractive templates for directed evolution in the search for novel molecules because they can regulate the activities of receptors, enzymes, and other molecules. Previously, we demonstrated that functional regulatory molecules against interleukin-6 receptor (IL-6R) could be obtained by directed evolution of the three-finger toxin (3F) scaffold. In the present study, trypsin was selected as a target for directed evolution to further explore the potential use of the 3F cDNA display library. After seven rounds of selection, the DNA sequences converged. The recombinant proteins produced by the selected candidates had inhibitory activity against trypsin (Ki of 33-450 nM). Three of the six groups had Ki values that were comparable to bovine pancreatic trypsin inhibitor and soybean trypsin inhibitor. Two of the candidates also had inhibitory effects against chymotrypsin and kallikrein. This study suggests that 3F protein is suitable for the preparation of high-diversity libraries that can be utilized to obtain protease inhibitors. In addition to our previous successful targeting of IL-6R, the technique developed in our studies may have wide applications in the generation of regulatory molecules for targets of interest, such as receptors, enzymes for research, diagnostic applications, and therapeutic uses.

The C2A domain in dysferlin is important for association with MG53 (TRIM72).

In skeletal muscle, Mitsugumin 53 (MG53), also known as muscle-specific tripartite motif 72, reportedly interacts with dysferlin to regulate membrane repair. To better understand the interactions between dysferlin and MG53, we conducted immunoprecipitation (IP) and pull-down assays. Based on IP assays, the C2A domain in dysferlin associated with MG53. MG53 reportedly exists as a monomer, a homodimer, or an oligomer, depending on the redox state. Based on pull-down assays, wild-type dysferlin associated with MG53 dimers in a Ca2+-dependent manner, but MG53 oligomers associated with both wild-type and C2A-mutant dysferlin in a Ca2+-independent manner. In pull-down assays, a pathogenic missense mutation in the C2A domain (W52R-C2A) inhibited the association between dysferlin and MG53 dimers, but another missense mutation (V67D-C2A) altered the calcium sensitivity of the association between the C2A domain and MG53 dimers. In contrast to the multimers, the MG53 monomers did not interact with wild-type or C2A mutant dysferlin in pull-down assays. These results indicated that the C2A domain in dysferlin is important for the Ca2+-dependent association with MG53 dimers and that dysferlin may associate with MG53 dimers in response to the influx of Ca2+ that occurs during membrane injury. To examine the biological role of the association between dysferlin and MG53, we co-expressed EGFP-dysferlin with RFP-tagged wild-type MG53 or RFP-tagged mutant MG53 (RFP-C242A-MG53) in mouse skeletal muscle, and observed molecular behavior during sarcolemmal repair; it has been reported that the C242A-MG53 mutant forms dimers, but not oligomers. In response to membrane wounding, dysferlin accumulated at the injury site within 1 second; this dysferlin accumulation was followed by the accumulation of wild-type MG53. However, accumulation of RFP-C242A MG53 at the wounded site was impaired relative to that of RFP-wild-type MG53. Co-transfection of RFP-C242A MG53 inhibited the recruitment of dysferlin to the sarcolemmal injury site. We also examined the molecular behavior of GFP-wild-type MG53 during sarcolemmal repair in dysferlin-deficient mice which show progressive muscular dystrophy, and found that GFP-MG53 accumulated at the wound similar to wild-type mice. Our data indicate that the coordination between dysferlin and MG53 plays an important role in efficient sarcolemmal repair.

Caenorhabditis elegans Rab escort protein (REP-1) differently regulates each Rab protein function and localization in a tissue-dependent manner.

Rab proteins play a critical role in intracellular vesicle trafficking and require post-translational modification by adding lipids at the C-terminus for proper functions. This modification is preceded by the formation of a trimeric protein complex with the Rab escort protein (REP) and the Rab geranylgeranyltransferase (RabGGTase). However, the genetic hierarchy among these proteins and the tissue-specificity of each protein function are not yet clearly understood. Here we identified the Caenorhabditis elegans rep-1 gene and found that a rep-1 mutant showed a mild defect in synaptic transmission and defecation behaviors. Genetic analyses using the exocytic Rab mutants rab-3 or rab-27 suggested that rep-1 functions only in the RAB-27 pathway, and not in the RAB-3 pathway, for synaptic transmission at neuromuscular junctions. However, the disruption of REP-1 did not cause defecation defects compared to severe defects in either RAB-27 or RabGGTase disruption, suggesting that REP-1 is not essential for RAB-27 signaling in defection. Some Rab proteins did not physically interact with REP-1, and localization of these Rab proteins was not severely affected by REP-1 disruption. These findings suggest that REP-1 functions are required in specific Rab pathways and in specific tissues, and that some Rab proteins are functionally prenylated without REP-1.

Regulation of AMPA receptor trafficking by delta-catenin.

delta-catenin is a protein that binds to the classical cadherins and to synaptic scaffolding proteins in a manner which allows the protein to serve as a link between the adherens junction and the postsynaptic complex. Here we show the regulatory role of delta-catenin on synaptic transmission. delta-catenin increased the AMPA receptor-mediated EPSC, but had no significant effect on the NMDA receptor-mediated EPSC. The effect of delta-catenin on the AMPAR EPSC was mediated by its PDZ ligand. delta-catenin regulates the surface expression of GluR2 in the dendritic spines of neurons. Immunoprecipitation revealed that delta-catenin bound to GRIP-1. In COS7 cells, co-transfection of delta-catenin, GRIP and GluR2 showed that delta-catenin regulates the membrane localization of GRIP through its PDZ ligand and increased the surface expression of GluR2. This study directly shows that delta-catenin is essential for the trafficking and positioning GluR2 in the spine and thus further links delta-catenin to neuronal plasticity.

Affixin activates Rac1 via betaPIX in C2C12 myoblast.

Affixin/beta-parvin is an integrin-linked kinase (ILK)-binding focal adhesion protein highly expressed in skeletal muscle and heart. To elucidate the possible role of affixin in skeletal muscle, we established stable C2C12 cell line expressing T7-tagged human affixin (C2C12-affixin cells). Exogenous expression of affixin promotes lamellipodium formation where affixin, ILK alphap21-activated kinase (PAK)-interactive exchange factor (PIX) and betaPIX accumulate. The association of affixin and betaPIX was confirmed by immunoprecipitation and pull down assay. In C2C12-affixin cells, an increased level of activated Rac1 but not Cdc42 was observed, and mutant betaPIX lacking guanine nucleotide exchange factor activity inhibited lamellipodium formation. These results suggest that affixin is involved in reorganization of subsarcolemmal cytoskeletal actin by activation of Rac1 through alpha and betaPIXs in skeletal muscle.

Identification and characterization of a novel phosphorylation site on the GluR1 subunit of AMPA receptors.

Phosphorylation of various AMPA receptor subunits can alter synaptic transmission and plasticity at excitatory glutamatergic synapses in the central nervous system. Here, we identified threonine-840 (T840) on the GluR1 subunit of AMPA receptors as a novel phosphorylation site. T840 is phosphorylated by protein kinase C (PKC) in vitro and is a highly turned-over phosphorylation site in the hippocampus. Interestingly, the high basal phosphorylation of T840 in the hippocampus is maintained by a persistent activity of a protein kinase, which is counter-balanced by a basal protein phosphatase activity. To study the function of T840, we generated a line of mutant mice lacking this phosphorylation site using a gene knock-in technique. The mice generated lack T840, in addition to two previously identified phosphorylation sites S831 and S845. Using this mouse, we demonstrate that T840 may regulate synaptic plasticity in an age-dependent manner.

Melatonin signaling regulates locomotion behavior and homeostatic states through distinct receptor pathways in Caenorhabditis elegans.

Melatonin is a hormone that controls circadian rhythms and seasonal behavioral changes in vertebrates. Recent studies indicate that melatonin participates in diverse physiological functions including the modulation of neural activities. Melatonin is also detected in many other organisms that do not exhibit obvious circadian rhythms, but their precise functions are not known. To understand the role of melatonin and its genetic pathway in vivo, we examined the effects of melatonin and its receptor antagonists on various behaviors in Caenorhabditis elegans. Exogenously applied melatonin specifically decreased locomotion rates in 15-min treatments, suggesting that melatonin directly regulates neural activities for locomotion. This melatonin signaling functions through MT1-like melatonin receptors, because the MT1/2 receptor antagonist luzindole effectively blocked the effect of melatonin on locomotion, while MT2-specific antagonist 4-phenyl-2-propionamidotetralin (4-P-PDOT) and MT3-selective antagonist prazosin had no effect. Alternatively, long-term treatment with prazosin specifically altered homeostatic states of the worm, suggesting another melatonin-signaling pathway through MT3-like receptors. We also found that two G-protein alpha subunit mutants and newly isolated five mutants exhibited defects in response to melatonin. Our findings imply that melatonin acts as a neuromodulator by regulating locomotion behavior and as a ligand for homeostatic control through distinct receptor pathways in C. elegans.

Dysferlin interacts with affixin (beta-parvin) at the sarcolemma.

The dysferlin gene is defective in Miyoshi myopathy (MM) and limb girdle muscular dystrophy type 2B (LGMD2B). Dysferlin is a sarcolemmal protein that is implicated in calcium-dependent membrane repair. Affixin (beta-parvin) is a novel, integrin-linked kinase-binding protein that is involved in the linkage between integrin and the cytoskeleton. Here we show that affixin is a dysferlin binding protein that colocalizes with dysferlin at the sarcolemma of normal human skeletal muscle. The immunoreactivity of affixin was reduced in sarcolemma of MM and LGMD2B muscles, although the total amount of the affixin protein was normal. Altered immunoreactivity of affixin was also observed in other muscle diseases including LGMD1C, where both affixin and dysferlin showed quite similar changes with a reduction of sarcolemmal staining with or without cytoplasmic accumulations. Colocalization of dysferlin and affixin was confirmed by immunofluorescence analysis using dysferlin-expressing C2 myoblasts. Wild-type and mutant dysferlin colocalized with endogenous affixin. The interaction of dysferlin and affixin was confirmed by immunoprecipitation study using normal human and mouse skeletal muscles. Using immunoprecipitation with deletion mutants of dysferlin, we have identified that C-terminal region of dysferlin is an apparent binding site for affixin. We also found N-terminal calponin homology domain of affixin as a binding site for dysferlin. Our results suggest that affixin may participate in membrane repair with dysferlin.

Input- and subunit-specific AMPA receptor trafficking underlying long-term potentiation at hippocampal CA3 synapses.

Hippocampal CA3 pyramidal neurons receive synaptic inputs from both mossy fibres (MFs) and associational fibres (AFs). Long-term potentiation (LTP) at these synapses differs in its induction sites and N-methyl-D-aspartate receptor (NMDAR) dependence. Most evidence favours the presynaptic and postsynaptic mechanisms for induction of MF LTP and AF LTP, respectively. This implies that molecular and functional properties differ between MF and AF synapses at both presynaptic and postsynaptic sites. In this study, we focused on the difference in the postsynaptic trafficking of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs) between these synapses. To trace the subunit-specific trafficking of AMPARs at each synapse, GluR1 and GluR2 subunits were introduced into CA3 pyramidal neurons in hippocampal organotypic cultures using the Sindbis viral expression system. The electrophysiologically-tagged GluR2 AMPARs, produced by the viral-mediated transfer of the unedited form of GluR2 (GluR2Q), were inserted into both MF and AF postsynaptic sites in a neuronal activity-independent manner. Endogenous Ca(2+)-impermeable AMPARs at these synapses were replaced with exogenous Ca(2+)-permeable receptors, and Ca(2+) influx via the newly expressed postsynaptic AMPARs induced NMDAR-independent LTP at AF synapses. In contrast, no GluR1 AMPAR produced by the gene transfer was constitutively incorporated into AF postsynaptic sites, and only a small amount into MF postsynaptic sites. The synaptic trafficking of GluR1 AMPARs was triggered by the activity of Ca(2+)/calmodulin-dependent kinase II or high-frequency stimulation to induce LTP at AF synapses, but not at MF synapses. These results indicate that MF and AF postsynaptic sites possess distinct properties for AMPAR trafficking in CA3 pyramidal neurons.

Recent advances in the study of AMPA receptors.

As glutamate is a dominant excitatory neurotransmitter in the central nervous system, glutamate receptors, and especially AMPA receptors, are located ubiquitously in all brain areas. In this paper, we reviewed recent advances of studies on AMPA receptor functions. AMPA receptors are cation-conducting complexes composed of various combinations of four subunits (GluR1 to GluR4). The glutamine residue located in the pore-forming segment of GluR2 subunit (Q/R site) is changed to arginine by RNA editing at the pre mRNA stage in normal adult mammalian animal. The edited GluR2 subunit is a major determination of Ca(2+) permeability of the AMPA receptor; only edited GluR2-lacking receptor shows high-Ca(2+) permeability. The assembly of glutamate AMPA receptor subunit is not completely according to the stochastic theory. The heteromeric subunits assembly is more rapid than the homomeric assembly is. The transfer of AMPA receptor subunit to the plasma membrane is conducted in multiple ways. Many molecules that interact with the intracellular domain of AMPA receptor subunits are reported as the modulators of AMPA receptor subunit transfer. In the motoneuron of sporadic amyotrophic lateral sclerosis (ALS) patients, the efficiency of RNA editing at the GluR2 Q/R site is significantly decreased. Relative low level of edited GluR2 subunit expression is likely responsible for motoneuronal death in ALS. Recently, AMPA receptors in glial cells have been studied. Bergmann glial cells in cerebellum express Ca(2+)-permeable AMPA receptors. Conversion of these AMPA receptors to Ca(2+)-impermeable type receptors induces morphological and functional changes. Glioblastoma cells also express Ca(2+)-permeable AMPA receptors, and their conversion to Ca(2+)-impermeable receptors inhibits cell locomotion and induces apoptosis.

Dopamine D1 receptors mediate CREB phosphorylation via phosphorylation of the NMDA receptor at Ser897-NR1.

Addictive drugs such as amphetamine and cocaine stimulate the dopaminergic system, activate dopamine receptors and induce gene expression throughout the striatum. The signal transduction pathway leading from dopamine receptor stimulation at the synapse to gene expression in the nucleus has not been fully elucidated. Here, we present evidence that D1 receptor stimulation leads to phosphorylation of the transcription factor Ca2+ and cyclic AMP response element binding protein (CREB) in the nucleus by means of NMDA receptor-mediated Ca2+ signaling. Stimulation of D1 receptors induces the phosphorylation of Ser897 on the NR1 subunit by protein kinase A (PKA). This phosphorylation event is crucial for D1 receptor-mediated CREB phosphorylation. Dopamine cannot induce CRE-mediated gene expression in neurons transfected with a phosphorylation-deficient NR1 construct. Moreover, stimulation of D1 receptors or increase in cyclic AMP levels leads to an increase in cytosolic Ca2+ in the presence of glutamate, but not in the absence of glutamate, indicating the ability of dopamine and cyclic AMP to facilitate NMDA channel activity. The recruitment of the NMDA receptor signal transduction pathway by D1 receptors may provide a general mechanism for gene regulation that is fundamental for mechanisms of drug addiction and long-term memory.

Overexpression of Ca2+-permeable AMPA receptor promotes delayed cell death of hippocampal CA1 neurons following transient forebrain ischemia.

To examine the role of Ca(2+) entry through AMPA receptors in the pathogenesis of the ischemia-induced cell death of hippocampal neurons, we delivered cDNA of Q/R site-unedited form (GluR2Q) of AMPA receptor subunit GluR2 in the hippocampus by using an HVJ-liposome-mediated gene transfer technique. Two days prior to transient forebrain ischemia, we injected an HVJ-liposome containing cDNA of the GluR2Q-myc fusion gene into a rat unilateral hippocampus. In the absence of ischemic insult, overexpression of Ca(2+)-permeable GluR2Q did not cause any neurodegeneration in the cDNA-injected hippocampus. In ischemic rats, overexpression of Ca(2+)-permeable GluR2Q markedly promoted ischemic cell death of CA1 pyramidal neurons, while complete rescue of CA1 pyramidal neurons from ischemic damage occurred in the hippocampal hemisphere opposite the GluR2Q expression. Overexpression of the Q/R-site edited form (GluR2R) of subunit GluR2 did not affect the ischemia-induced damage of CA1 pyramidal neurons. From these results, we suggest that the Ca(2+)-permeability of AMPA receptors does not have a direct contribution to glutamate receptor-mediated neurotoxicity but has a promotive action in the evolution of ischemia-induced neurodegeneration of vulnerable neurons.

Postsynaptic expression of a new calcium pathway in hippocampal CA3 neurons and its influence on mossy fiber long-term potentiation.

Long-term potentiation (LTP) in the CA1 region of the hippocampus is induced by postsynaptic Ca(2+) influx via NMDA receptors (NMDARs). However, this synaptic plasticity occurs independently of NMDARs when Ca(2+)-permeable AMPA receptors (AMPARs) are expressed at postsynaptic sites using various genetic techniques, indicating that an increase in Ca(2+) level at critical postsynaptic sites, regardless of its entry pathway, triggers the induction of LTP at CA1 synapses. In contrast, NMDARs are sparsely distributed on mossy fiber (MF) synapses in CA3 hippocampal neurons, and most evidence favors the presynaptic mechanism for LTP induction, although some reports suggested a postsynaptic mechanism. In this study, we examined whether Ca(2+) influx through the newly produced postsynaptic receptors during high-frequency stimulation affects the induction of MF LTP. For this purpose, we expressed Ca(2+)-permeable AMPARs in CA3 pyramidal neurons by Sindbis viral-mediated gene transfer of the unedited form of the glutamate receptor 2 (GluR2Q) subunit, as a new pathway for postsynaptic Ca(2+) entry, in rat hippocampal organotypic cultures. Virally expressed myc-tagged GluR2Q was detected at the complex spines known as the thorny excrescences, which serve as postsynaptic targets for MF synaptic input, on the proximal apical dendrites of CA3 pyramidal cells. Furthermore, endogenous Ca(2+)-impermeable AMPARs at MF synapses were converted into Ca(2+)-permeable receptors by GluR2Q expression. However, the postsynaptic expression of Ca(2+)-permeable AMPARs had no significant influence on the two types of MF LTP induced by different stimulus protocols. These results supported the notion that MF LTP is independent of postsynaptic Ca(2+).