Morphological analysis on the distribution of membrane lipids and a membrane protein, NAP-22, during neuronal development in vitro

Interaction of dynamin I with NAP-22, a neuronal protein enriched in the presynaptic region

Neurons have well-developed membrane microdomains called "rafts" that are recovered as a detergent-resistant low-density membrane microdomain fraction (DRM). NAP-22 is one of the major protein components of neuronal DRM and localizes in the presynaptic region. In order to know the role of NAP-22 in the synaptic transmission, NAP-22 binding proteins in the cytosol were searched with an affinity screening with NAP-22 as a bait and several protein bands were detected. Using mass-analysis and western blotting, one of the main band of ∼90 kDa was identified as dynamin I. The GTPase activity of dynamin I was partly inhibited by NAP-22 expressed in bacteria and this inhibition was recovered by the addition of calmodulin, a NAP-22 binding protein. The GTPase activity of dynamin was known to be activated with acidic membrane lipids such as phosphatidylserine and the addition of NAP-22, a phosphatidylserine binding protein, inhibited the activation of the GTPase by this lipid. Since NAP-22 localizes on the presynaptic plasma membrane and on synaptic vesicles, these results suggest the participation of NAP-22 in the membrane cycling through binding to dynamin and acidic membrane lipids at the presynaptic region.

A proteomic analysis of contextual fear conditioned rats reveals dynamic modifications in neuron and oligodendrocyte protein expression in the dentate gyrus

Contextual memory is an intricate process involving synaptic plasticity and network rearrangement. Both are governed by many molecular processes including phosphorylation and modulation of protein expression. However, little is known about the molecules involved in it. Here, we exploited the advantages of a quantitative proteomic approach to identify a great number of molecules in the rat dentate gyrus after a contextual fear conditioning session. Our results allowed us to highlight protein expression patterns, not only related to neuroplasticity, but also to myelin structure, such as myelin basic protein and myelin proteolipid protein showing a decrease in expression. Validation of the modification in protein expression reveals a dynamic profile during the 48 h following the fear conditioning session. The expression of proteins involved in neurite outgrowth, such as BASP-1 and calcineurin B1, and in synaptic structure and function, VAMP2 and RAB3C, was increased in the dentate gyrus of rats submitted to fear conditioning compared to controls. We showed that the increase in BASP-1 protein was specific to fear conditioning learning as it was not present in immediate-shock rats, neither in rats exposed to a novel environment without being shocked. As myelin is known to stabilise synaptic network, the decrease in myelin proteins suggests a neuroglia interactive process taking place in the dentate gyrus in the 24 h following contextual fear learning, which has never been demonstrated before. These results therefore open the way to the study of new plasticity mechanisms underlying learning and memory.

Ganglioside contained in the neuronal tissue-enriched acidic protein of 22 kDa (NAP-22) fraction prepared from the detergent-resistant membrane microdomain of rat brain inhibits the phosphatase activity of calcineurin

Neurons have well-developed membrane microdomains called "rafts" that are recovered as a detergent-resistant membrane microdomain fraction (DRM). Neuronal tissue-enriched acidic protein of 22 kDa (NAP-22) is one of the major protein components of neuronal DRM. To determine the cellular function of NAP-22, interacting proteins were screened with an immunoprecipitation assay, and calcineurin (CaN) was detected. Further studies with NAP-22 prepared from DRM and CaN expressed in bacteria showed the binding of these proteins and a dose-dependent inhibitory effect of the NAP-22 fraction on the phosphatase activity of CaN. On the other hand, NAP-22 expressed in bacteria showed low binding to CaN and a weak inhibitory effect on phosphatase activity. To solve this discrepancy, identification of a nonprotein component that modulates CaN activity in the DRM-derived NAP-22 fraction was attempted. After lyophilization, a lipid fraction was extracted with chloroform/methanol. The lipid fraction showed an inhibitory effect on CaN without NAP-22, and further fractionation of the extract with thin-layer chromatography showed the presence of several lipid bands having an inhibitory effect on CaN. The mobility of these bands coincided with that of authentic ganglioside (GM1a, GD1a, GD1b, and GT1b), and authentic ganglioside showed an inhibitory effect on CaN. Treatment of lipid with endoglycoceramidase, which degrades ganglioside to glycochain and ceramide, caused a diminution of the inhibitory effect. These results show that DRM-derived NAP-22 binds several lipids, including ganglioside, and that ganglioside inhibits the phosphatase activity of CaN.

Interaction of NAP-22 with brain glutamic acid decarboxylase (GAD)

NAP-22 (also called BASP1 or CAP-23) is a neuron-enriched protein localized mainly in the synaptic vesicles and the synaptic plasma membrane. Biochemically, it is recovered in the lipid raft fraction. In order to understand the physiological function of the neuronal lipid raft, NAP-22 binding proteins were screened with a pull-down assay. Glutamic acid decarboxylase (GAD) was detected through LC-MS/MS, and Western blotting using a specific antibody confirmed the result. Two isoforms of GAD, GAD65 and GAD67, were expressed in bacteria as GST-fusion forms and the interaction with NAP-22 was confirmed in vitro. Partial co-localization of NAP-22 with GAD65 and GAD67 was also observed in cultured neurons. The binding showed no effect on the enzymatic activity of GAD65 and GAD67. These results hence suggest that NAP-22 could participate in the transport of GAD65 and GAD67 to the presynaptic termini and their retention on the synaptic vesicles as an anchoring protein.

Identification of dynamin as a septin-binding protein

Lipid rafts (detergent-resistant low-density membrane microdomain: DRM) are signal-transducing membrane platforms. In a previous study, we showed maturation-dependent localization of septin in the DRM fraction of rat brain. Mammalian septin is composed with 13-14 isoforms and these isoforms assemble to form rod-shaped hetero-oligomeric complexes. End-to-end polymerization of these complexes results in the formation of higher order structures such as filamentous sheets or bundles of filaments that restrict the fluid-like diffusion of the membrane proteins and lipids. Considering the function of septin as the membrane scaffold, elucidation of the molecular interaction of septin in DRM could be a breakthrough to understand another role of lipid rafts. In order to identify septin-binding proteins in DRM, solubilization and fractionation of septin from DRM was attempted. Several proteins were co-fractionated with septin and LC-MS/MS analysis identified one of these proteins as dynamin and Western blotting using anti-dynamin confirmed this result. Immunoprecipitation of septin11 in a crude supernatant showed co-precipitation of dynamin and dynamin fraction prepared from brain contained several septin isoforms. Within bacterially expressed septin isoforms, septin5 and septin11 bound dynamin but septin9 did not. These results suggest that some septin isoforms participate in the dynamin-related membrane dynamics.

Subcellular and regional location of "brain" proteins BASP1 and MARCKS in kidney and testis

Proteins BASP1 and MARCKS are abundant in axonal endings of neurons. Similarly to brain-specific protein GAP-43, BASP1 and MARCKS are reversibly bound to the plasma membrane. These proteins control both actin polymerization and actin cytoskeleton binding to the membrane. Performing these functions, BASP1 and MARCKS take part in growth cone guidance during development and in neurotransmitter secretion in adults. These activities predetermine the pivotal role of BASP1 and MARCKS in learning and memory. BASP1 and MARCKS were also found in non-nerve tissues, in particular, in the kidney and testis. Evidently, the physiological roles of these proteins differ in different tissues. Correspondingly, their intracellular location and activities may not be similar to those in neurons. In this paper, we analyze subcellular fractions (cytoplasm and nuclei) of rat kidney and testis with the purpose of determining the intracellular location of BASP1 and MARCKS. Western blots demonstrated that in these tissues, as in the brain, both proteins are present in the cytoplasm of the cell. According to our immunohistochemical study, BASP1 and MARCKS are specifically distributed in the tissues studied. In kidney, both proteins are present in cells located in glomeruli. In the testicular tubules, BASP1 is mainly expressed at the late stage of spermatogenesis (in spermatids) and is preserved in mature spermatozoa, while MARCKS appears equally during all stages of spermatogenesis. MARCKS is not found in mature spermatozoa. The results indicate that study of functions of BASP1 and MARCKS in the kidney and in the reproduction system holds much promise.