An investigation of herpes simplex virus type 1 latency in a novel mouse dorsal root ganglion model suggests a role for ICP34.5 in reactivation.

After a primary lytic infection at the epithelia, herpes simplex virus type 1 (HSV-1) enters the innervating sensory neurons and translocates to the nucleus, where it establishes a quiescent latent infection. Periodically, the virus can reactivate and the progeny viruses spread back to the epithelium. Here, we introduce an embryonic mouse dorsal root ganglion (DRG) culture system, which can be used to study the mechanisms that control the establishment, maintenance and reactivation from latency. Use of acyclovir is not necessary in our model. We examined different phases of the HSV-1 life cycle in DRG neurons, and showed that WT HSV-1 could establish both lytic and latent form of infection in the cells. After reactivating stimulus, the WT viruses showed all markers of true reactivation. In addition, we showed that deletion of the γ(1)34.5 gene rendered the virus incapable of reactivation, even though the virus was clearly able to replicate and persist in a quiescent form in the DRG neurons.

Myelin synthesis in the peripheral nervous system.

By imposing saltatory conduction on the nervous impulse, the principal role of the myelin sheath is to allow the faster propagation of action potentials along the axons which it surrounds. Peripheral nervous system (PNS) myelin is formed by the differentiation of the plasma membrane of Schwann cells. One of the biochemical characteristics that distinguishes myelin from other biological membranes is its high lipid-to-protein ratio. All the major lipid classes are represented in the myelin membrane, while several myelin-specific proteins have been identified. During development, the presence of axons is required for the initiation of myelination, but the nature of the axonal signal is still unknown. The only certainties are that this signal is synthesized by axons whose diameter is greater than 0.7 microm, and that the signal(s) include(s) a diffusible molecule. Morphological studies have provided us with information concerning the timing of myelination, the mechanism by which immature Schwann cells differentiate into a myelinating phenotype and lay down the myelin sheath around the axon, and the accumulation and the structure of the myelin membrane. The last 20 years have seen the identification and the cDNA and gene cloning of the major PNS myelin proteins, which signalled the beginning of the knock-out decade: transgenic null-mutant mice have been created for almost every protein gene. The study of these animals shows that the formation of myelin is considerably less sensitive to molecular alterations than the maintenance of myelin. During the same period, important data has been gathered concerning the synthesis and function of lipids in PNS myelin, although this field has received relatively little attention compared with that of their protein counterparts.

Crystal structure of non-fused glutathione S-transferase from Schistosoma japonicum in complex with glutathione.

Schistosoma japonicum glutathione S-transferase (SjGST) is a common fusion tag in recombinant protein production, and its 3-dimensional structure has been studied in the context of drug design. We have determined the crystal structure of non-fused SjGST complexed with glutathione, and compare it to complexes between glutathione and SjGST fusion proteins.

S100beta inhibits the phosphorylation of the L-MAG cytoplasmic domain by PKA.

The myelin-associated glycoprotein (MAG) is a cell adhesion molecule expressed by myelinating glia, existing as two isoforms that differ only by their cytoplasmic domains. We have studied the in vitro phosphorylation of recombinant rat MAG cytoplasmic domains by three kinases for which consensus sequences exist within this domain, revealing phosphorylation of the L-MAG-specific domain by protein kinase A (PKA). Phosphorylation of the L-MAG cytoplasmic domain by PKA was decreased in the presence of S100beta, providing a functional significance to the interaction between L-MAG and S100beta, and further indicating that L-MAG may play a role in myelinating glial cell signalling processes.

The small myelin-associated glycoprotein binds to tubulin and microtubules.

The myelin-associated glycoprotein (MAG) exists as two isoforms, differing only by their respective cytoplasmic domains, that have been suggested to function in the formation and maintenance of myelin. In the present study, a 50 kDa protein binding directly to the small MAG (S-MAG) cytoplasmic domain was detected and identified as tubulin, the core component of the microtubular cytoskeleton. In vitro, the S-MAG cytoplasmic domain slowed the polymerization rate of tubulin and co-purified with assembled microtubules. A significant sequence homology was found between the tau family tubulin-binding repeats and the carboxy-terminus of S-MAG. Our results indicate that S-MAG is the first member of the Ig superfamily that can be classified as a microtubule-associated protein, and place S-MAG in a dynamic structural complex that could participate in linking the axonal surface and the myelinating Schwann cell cytoskeleton.

Expression of the amino acid dimorphism in the small myelin-associated glycoprotein cytoplasmic domain in rat peripheral nerves during postnatal development.

The myelin-associated glycoprotein (MAG) is one of the proteins expressed during the period of myelin formation and is believed to play a major role in the initiation of myelination. It exists as two differentially expressed isoforms, L- and S-MAG, that are generated by alternative mRNA splicing. A nucleotide dimorphism at the mRNA level resulting in an Arg/Pro dimorphism in the cytoplasmic tail of the S-MAG protein has previously been detected in the rat brain. In this study, we show that this dimorphism is detectable in the rat peripheral nervous system. We propose an allelic origin for the dimorphism and demonstrate the differential expression of the S-MAG alleles in the sciatic nerves of heterozygous rats during the period of active myelination. We also present data on the properties of the two S-MAG cytoplasmic domains produced as GST fusion proteins. The importance of this differentially expressed amino acid dimorphism is discussed, taking into account both its probable effect on the S-MAG cytoplasmic domain function and its significance in functional and structural studies concerning the S-MAG protein.

High uptake and rapid metabolism of palmitate in peripheral nerves of normal and Trembler mice in vivo: Similarities and differences.

In vivo, short-term palmitate metabolism was studied in normal and Trembler mouse sciatic nerves, after intraneural injection of 6.34 pmol of [(3)H]palmitate. After 30 min of in vivo incubation, all the [(3)H]palmitate available for short-term lipid metabolism (90% of the injected substrate) had already been used up in both normal and Trembler sciatic nerves. Despite the normal metabolic level of the [(3)H]palmitate, the distribution of the label between the various polar lipids was abnormal in the mutant's nerves: sphingomyelin and phosphatidylcholine were more labelled than in normal nerves, while the other lipids, and particularly the cerebrosides, were less labelled. The labelled fatty acid distribution in four polar lipids, purified by thin-layer chromatography, was also abnormal, correlating with a decreased overall [(3)H]palmitate elongation (after 2 h, normal = 44%, Trembler = 24%), and a severely decreased synthesis (5- to 10-fold) of saturated very long-chain fatty acids. In normal nerves, saturated very long-chain fatty acids represented about 8% of the total labelled fatty acids and they were preferentially incorporated into the cerebrosides. The rapid depletion of the [(3)H]palmitate, the high level of its incorporation into polar lipids of both normal and Trembler mouse sciatic nerves and the excellent reproducibility of the results, have led us to propose a novel "pulsed wave" method for the in vivo study of intracellular lipid transport phenomena implicated in peripheral nerve myelin assembly.

A comparison of substrate-dependent fatty acyl-CoA synthetase activities and ATP hydrolysis in membrane pellets from normal and trembler mouse sciatic nerves.

Acyl-CoA synthetase activity was investigated in membrane fractions from normal and Trembler mouse sciatic nerves. Stearoyl-CoA synthesis in the Trembler assays was always about twice that observed in the control experiments under various conditions of CoA and ATP concentrations. In experiments using varying concentrations of fatty acid, the activation of stearate was very low in both systems (normal and Trembler), up to a substrate concentration of 20 mM, suggesting that fatty acid micelles are required for full activity. ATP, at concentrations of up to 1 mM, was totally (Trembler) or partially (control) hydrolyzed within 15 min, resulting in an apparently lower ligase activity in the mutant than in the control when an ATP-regenerating system was not added. The activation of saturated fatty acids decreased with increasing chain length. However, lignoceroyl-CoA formation was detected in the mutant but not in the control.

Increased acylation of lysophosphatidylcholine in microsomes of trembler mouse peripheral nerves.

We report here the presence of a Stearoyl-CoA: [1-palmitoyl]-glycerophosphorylcholine (LPC) transacylase activity in the microsomal fraction of normal and Trembler mouse sciatic nerves. Under the experimental conditions studied as a function of incubation time, protein concentration, acyl-CoA and LPC concentrations, the transacylase specific activity was 2-3 times higher in the microsomes of the mutant's nerves than in those of the control. The addition of 5 mM ATP-Mg to the incubation medium, in the absence of bovine serum albumin, leads to a 90% decrease of the stearoyl-CoA thioesterase activity, but increases the transacylation by only 10-20%. This may be due to the low value (10 microM) of the apparent K(m) for C(18)-CoA observed for the mutant's transacylase. In microsomes from control nerves, transacylation requires exogenous LPC, whereas in Trembler mouse sciatic nerve microsomes, the transacylase can use endogenous LPC.

Pathways of incorporation of fatty acid into glycerolipids of the murine peripheral nervous system in vivo: alterations in the dysmyelinating mutant trembler mouse.

In vivo glycerolipid metabolism was studied in sciatic nerves of normal and Trembler mice. The results showed that two kinetically independent pathways were implicated in the labeling of diacylglycerophospholipids from [3H]palmitate: the Kennedy pathway and a 'direct acylation' pathway. In normal nerves, 45% of the glycerophospholipids were labeled, with a rate constant k3 = 3.9 x 10(-3) min-1, from phosphatidic acid and diacylglycerol intermediates, themselves formed with a rate constant of k1 = 0.24 min-1 from a free 3H-fatty acid pool, FFA1, that represents 45% of the total injected label. The remaining 55% of the glycerophospholipids were labeled from a kinetically distinct free 3H-fatty acid pool, FFA2, with a rate constant of k4 = 9.8 x 10(-2) min-1, via a process that does not implicate a detectably labeled metabolic intermediate ('direct acylation'). Glycerophospholipid labeling via the Kennedy pathway in the Trembler mouse sciatic nerves was reduced to 75% of the normal level, while labeling via the 'direct acylation' pathway was increased 1.4-fold. The values of the rate constants for free 3H-fatty acid utilisation (k1 and k4) were both increased about 2.5-fold, while that of glycerophospholipid formation from diacylglycerol (k3) was close to normal.

Acylation of endogenous acyl acceptors by mouse sciatic nerve microsomes.

Phospholipid (chiefly phosphatidylcholine) labeling from radioactive acyl-CoAs by mouse sciatic nerve microsomes is observed in the absence of added acyl acceptors. The maximal acylation (ca 10% of administered) for 10 micrograms microsomal proteins is observed at relatively low amounts of oleoyl-CoA (0.2-0.3 nmol) and decreases as the acyl-CoA amount increases. Labeled lysophosphatidylcholine (almost exclusively esterified at position 2) is also observed, particularly when the [1-14C]oleoyl-CoA concentration is higher than 0.2-0.3 nmol/50 microliters. The labeled acyl group is mainly inserted in position 2] of the glycerophosphorylcholine. With 0.15 nmol labeled oleoyl-CoA, phosphatidylcholine acylation increases as a function of the protein amount and reaches 25% of the added label at 40 microgram proteins. It is evaluated that, in the presence of 10 microgram proteins, 2% of the microsomal phosphatidylcholine molecules are acylated from 0.1 nmol acyl-CoA. The acylation mechanism seems to involve an acyl exchange between acyl-CoA and phosphatidylcholine.

Trembler as a mouse model of CMT1A?

The Trembler mouse suffers from a dominantly inherited autosomal mutation that results in an abnormal myelination of the peripheral nervous system. Biochemical studies have shown that dysmyelination is the primary event, demyelination being a late-occurring process. The expression of myelin protein genes has been studied. The steady-state levels for PMP22 mRNA represent 10 and 5% of normal values in the nerves of heterozygous and homozygous Trembler, respectively. This is due to a reduced expression of the specific transcript driven by the promoter 1 of the PMP22 gene. Collective results indicate that Trembler dysmyelination is not necessarily the consequence of a large accumulation of the mutated PMP22 protein. Moreover, it appears that the situation in the Trembler is different from that encountered in most CMT1A patients, where an increased PMP22 gene dosage is responsible for the disease. Therefore, the Trembler mutant is perhaps not an ideal model for this human neuropathy.

Phospholipid and fatty acid composition of erythrocytes in type I and type II diabetes.

Different data have been reported concerning modifications of the erythrocyte lipid composition in the different types of diabetes. The heterogeneity of diabetes could be a cause for such differences. Ten type I and ten type II diabetics were carefully selected. The patients were poorly controlled (the mean glycosylated hemoglobin was 12.8% +/- 0.7%); their mean age was 54 +/- 5 years, with a mean duration of diabetes of 18 +/- 4 years. One half of them had severe diabetic complications (nephropathy, retinopathy, and/or polyneuropathy). The diabetics were compared with ten controls. The phospholipid composition was determined by HPTLC analysis, and the fatty acid moieties of the total phospholipids were measured by gas liquid chromatography associated with mass spectrometry. Under well-defined experimental conditions, these results demonstrated a slight, but significant (P less than .05), increase in the phosphatidylethanolamine (PE)/phosphatidylserine (PS) ratio using a Ninhydrin quantitation method; there was also an increase in two minor lipids content (phosphatidylinositol, phosphatidic acid) and the appearance of a lysolipid (lysoPE) in the patients. Whatever the type of diabetes, the red blood cells of diabetics showed no significant differences in their fatty acid contents.

A quantitative developmental study of neutral lipids during myelinogenesis in the peripheral nervous system of normal and trembler mice.

The quantitative accumulation of neutral lipids during the period of myelination in the peripheral nervous system was studied in normal and trembler mouse sciatic nerves, between the ages of 5 and 27 days. Neutral lipids were resolved by high-performance thin-layer chromatography, using the solvent mixture hexane/diethyl ether/acetic acid (90:15:2, v/v/v). The lipids were quantitated, after copper acetate/phosphoric acid charring, by densitometric scanning, using an external standard technique. Cholesterol and triacylglycerols accumulated in normal nerves throughout the period studied, while cholesteryl esters were not observed at any age. In trembler nerves, the accumulation of cholesterol took place at a much lower rate than in normal nerves and this lipid was deficient from the earliest stages of development. Triacylglycerols were not significantly deficient in trembler nerves during the first 2-3 weeks, but, after the age of 18 days, their quantity diminished significantly. Cholesteryl esters were first detected in the mutant nerves at the age of 18 days. These results, in agreement with those of a previous developmental study of the polar lipids, are strongly in favour of the view that the trembler mutation directly induces a process of dysmyelination and that demyelination is a secondary event.

A quantitative developmental study of the peripheral nerve lipid composition during myelinogenesis in normal and trembler mice.

The quantitative evolution of 10 polar lipids was examined in the sciatic nerves of normal and trembler mice between the ages of 3 days and 60 days. In normal nerves, the polar lipids accumulated slowly until the age of 9 days. A period of rapid accumulation then took place until 18 days of age, after which the phospholipids plateaued, while the glycolipid content continued to increase at a slower rate. The results obtained for the sciatic nerves of trembler mice show that the accumulation of all the polar lipids studied, except phosphatidylcholine and hydroxysulfatides, is abnormal from the earliest stages of postnatal development, and strongly support the view that the primary disorder in the trembler peripheral nervous system is one of dysmyelination. With the exception of cardiolipin, all the lipids in the trembler nerves stopped accumulating at the age of 18 days. The cerebrosides were the lipids the most affected severely at all ages.

Correlation between the morphology and the lipid and protein compositions in the peripheral nervous system of individual 8-day-old normal and trembler mice.

The hereditary, hypertrophic interstitial neuropathy which afflicts the trembler mouse manifests itself about two weeks after birth. Consequently, the identification of these mutant mice was not possible before this age, except when double mutants were available. We show that the trembler mice can be easily distinguished from their normal littermates before the clinical symptoms appear by using an HPTLC/densitometry technique that allows the simple and rapid analysis of the polar lipids extracted from one sciatic nerve. The results presented in this paper demonstrate important differences between the polar lipid compositions of sciatic nerves from 8-day-old normal and trembler littermates, whose phenotypes were confirmed by the morphological analysis of the contralateral sciatic nerves. The small amount of material that is needed for this identification makes it possible to use the remaining nerve material for other studies. Furthermore, important differences between the sciatic nerve protein compositions of normal and trembler mice, identified according to their polar lipid composition, were also observed and these differences can, therefore, also be employed for the identification of the mutants before the manifestation of the clinical symptoms of the trembler neuropathy.

Cerebroside formation in the peripheral nervous system of normal and Trembler mice.

The formation of cerebrosides by the galactosylation of ceramides was investigated in a microsomal fraction prepared from sciatic nerves of normal and Trembler mice. In the control, cerebroside synthesis is observed in the presence of uridine-diphosphate-galactose both from endogenously synthesized [1-14C]stearoyl-sphingosine (C18-ceramide), and from [1-14C]lignoceroyl-sphingosine (C24-ceramide). Cerebroside formation is also demonstrated by studying the galactosylation of exogenous ceramides with UDP[1-14C]-galactose. In the mutant, only trace amounts of labeled cerebrosides are formed from labeled stearoyl-sphingosine, whereas with lignoceroyl-sphingosine, no cerebroside synthesis is detected under conditions allowing their formation in the control. However, a higher rate of synthesis of short acyl chain-glucosyl ceramides is observed in the Trembler samples.

A developmental study of fatty acyl group contents in the peripheral nervous system of normal and trembler mice.

We quantitated the content of the fatty acyl groups of sciatic nerves from normal and Trembler mice between the postnatal ages of 5 and 60 d. Palmitoyl, stearoyl, and unsaturated 18-carbon groups increased normally in Trembler nerves during the first 9 d, after which their levels were notably lower than those observed for the normal littermates. In good agreement, the synthesis of palmitic acid by the fatty acid synthetase was normal in the PNS of 9-d-old Trembler mice. "Myelin-specific" saturated very long chain fatty acyl groups (VLCFAG) were deficient at all ages studied, the deficiency increasing from about 2.5-fold at 5 d, to 26-fold at 60 d, compared to the maximal 3- to 4-fold decrease observed for the more ubiquitous shorter chains. The results presented in this paper suggest that the highly abnormal VLCFAG content of the mutant nerves cannot be accounted for by an abnormal fatty acid synthetase activity. For the fatty acyl chains with 18 carbon atoms, however, the deficiency observed after 9 d of age can be almost entirely explained by the diminished levels of palmitic acid. The relatively late occurrence of the palmitic acid deficiency compared with that of the VLCFAG and the normal palmitate synthetase activity observed in young mutant mice indicates that the former is an indirect consequence of the Trembler mutation.

Lack of in vitro ceramide formation in the PNS of trembler mice.

The amidification of sphingosine by acyl donors has been investigated in a microsomal fraction prepared from sciatic nerves of normal and Trembler mice. In the control, a ceramide synthesis is observed in the presence of acyl-CoAs and not with free fatty acids. The synthesis increases as a function of the protein amount and the time and is dependent on acyl-CoA concentration. The level of synthesis is highly similar to that observed in vivo after palmitate injection into the sciatic nerves of normal mice. In the mutant, there is a major abnormality because a weak synthesis (20% of the control) is observed only with high acyl-CoA concentration (greater than 200 microM), whereas in the range of the physiological acyl-CoA concentrations (less than 20 microM), there is no ceramide formation from stearoyl-CoA or lignoceroyl-CoA.

Technique for injection into the sciatic nerve of the mouse for quantitative in vivo metabolic studies.

In this paper we describe a technique for intraneural injections, applicable to mouse peripheral nerves, which, compared with previous techniques, reduces trauma to the nerves and increases the level and reduces the variability of label recovery. Our technique employs glass needles (tip diameter, 50 micron) linked to a peristaltic pump by polyethylene tubing to inject small volumes (in the microliter range) of radiolabeled substrate solutions into mouse sciatic nerves, and allows the recovery of 20.9 +/- 1.9% (mean +/- standard deviation) and 30.5 +/- 4.8% of the injected radioactivity for 2 microliter [3H]acetate and 0.5 microliter of [3H]stearate, respectively.