Streptococcus pneumoniae Rapidly Translocate from the Nasopharynx through the Cribriform Plate to Invade the Outer Meninges.

The entry routes and translocation mechanisms of microorganisms or particulate materials into the central nervous system remain obscure We report here that Streptococcus pneumoniae (pneumococcus), or polystyrene microspheres of similar size, appear in the meninges of the dorsal cortex of mice within minutes of inhaled delivery. Recovery of viable bacteria from dissected tissue and fluorescence microscopy show that up to at least 72 h, pneumococci and microspheres were predominantly found in the outer of the two meninges: the pachymeninx. No pneumococci were found in blood or cerebrospinal fluid. Intravital imaging through the skull, aligned with flow cytometry showed recruitment and activation of LysM+ cells in the dorsal pachymeninx at 5 and 10 hours following intranasal infection. Imaging of the cribriform plate suggested that both pneumococci and microspheres entered through the foramina via an inward flow of fluid connecting the nose to the pachymeninx. Our findings bring new insight into the varied mechanisms of pneumococcal invasion of the central nervous system, but they are also pertinent to the delivery of drugs to the brain and the entry of airborne particulate matter into the cranium. IMPORTANCE Using two-photon imaging, we show that pneumococci translocate from the nasopharynx to the dorsal meninges of a mouse in the absence of any bacteria found in blood or cerebrospinal fluid. Strikingly, this takes place within minutes of inhaled delivery of pneumococci, suggesting the existence of an inward flow of fluid connecting the nasopharynx to the meninges, rather than a receptor-mediated mechanism. We also show that this process is size dependent, as microspheres of the same size as pneumococci can translocate along the same pathway, while larger size microspheres cannot. Furthermore, we describe the host response to invasion of the outer meninges. Our study provides a completely new insight into the key initial events that occur during the translocation of pneumococci directly from the nasal cavity to the meninges, with relevance to the development of intranasal drug delivery systems and the investigations of brain damage caused by inhaled air pollutants.

Where are we? The anatomy of the murine cortical meninges revisited for intravital imaging, immunology, and clearance of waste from the brain.

Rapid progress is being made in understanding the roles of the cerebral meninges in the maintenance of normal brain function, in immune surveillance, and as a site of disease. Most basic research on the meninges and the neural brain is now done on mice, major attractions being the availability of reporter mice with fluorescent cells, and of a huge range of antibodies useful for immunocytochemistry and the characterization of isolated cells. In addition, two-photon microscopy through the unperforated calvaria allows intravital imaging of the undisturbed meninges with sub-micron resolution. The anatomy of the dorsal meninges of the mouse (and, indeed, of all mammals) differs considerably from that shown in many published diagrams: over cortical convexities, the outer layer, the dura, is usually thicker than the inner layer, the leptomeninx, and both layers are richly vascularized and innervated, and communicate with the lymphatic system. A membrane barrier separates them and, in disease, inflammation can be localized to one layer or the other, so experimentalists must be able to identify the compartment they are studying. Here, we present current knowledge of the functional anatomy of the meninges, particularly as it appears in intravital imaging, and review their role as a gateway between the brain, blood, and lymphatics, drawing on information that is scattered among works on different pathologies.

The mouse cortical meninges are the site of immune responses to many different pathogens, and are accessible to intravital imaging.

A wide range of viral and microbial infections are known to cause meningitis, and there is evidence that the meninges are the gateway to pathogenic invasion of the brain parenchyma. Hence observation of these regions has wide application to understanding host-pathogen interactions. Interactions between pathogens and cells of the immune response can be modified by changes in their environment, such as suppression of the flow of blood and lymph, and, particularly in the case of the meninges, with their unsupported membranes, invasive dissection can alter the tissue architecture. For these reasons, intravital imaging through the unperforated skull is the method of choice. We give a protocol for a simple method of two-photon microscopy through the thinned cortical skull of the anesthetized mouse to enable real-time imaging with sub-micron resolution through the meninges and into the superficial brain parenchyma. In reporter mice in which selected cell types express fluorescent proteins, imaging after infection with fluorescent pathogens (lymphocytic choriomeningitis virus, Trypanosoma brucei or Plasmodium berghei) has shown strong recruitment to the cortical meninges of immune cells, including neutrophils, T cells, and putative dendritic cells and macrophages. Without special labeling, the boundaries between the dura mater, the leptomeninx, and the parenchyma are not directly visualized in intravital two-photon microscopy, but other landmarks and characteristics, which we illustrate, allow the researcher to identify the compartment being imaged. While most infectious meningitides are localized mainly in the dura mater, others involve recruitment of immune cells to the leptomeninx.

Perivascular Arrest of CD8+ T Cells Is a Signature of Experimental Cerebral Malaria.

There is significant evidence that brain-infiltrating CD8+ T cells play a central role in the development of experimental cerebral malaria (ECM) during Plasmodium berghei ANKA infection of C57BL/6 mice. However, the mechanisms through which they mediate their pathogenic activity during malaria infection remain poorly understood. Utilizing intravital two-photon microscopy combined with detailed ex vivo flow cytometric analysis, we show that brain-infiltrating T cells accumulate within the perivascular spaces of brains of mice infected with both ECM-inducing (P. berghei ANKA) and non-inducing (P. berghei NK65) infections. However, perivascular T cells displayed an arrested behavior specifically during P. berghei ANKA infection, despite the brain-accumulating CD8+ T cells exhibiting comparable activation phenotypes during both infections. We observed T cells forming long-term cognate interactions with CX3CR1-bearing antigen presenting cells within the brains during P. berghei ANKA infection, but abrogation of this interaction by targeted depletion of the APC cells failed to prevent ECM development. Pathogenic CD8+ T cells were found to colocalize with rare apoptotic cells expressing CD31, a marker of endothelial cells, within the brain during ECM. However, cellular apoptosis was a rare event and did not result in loss of cerebral vasculature or correspond with the extensive disruption to its integrity observed during ECM. In summary, our data show that the arrest of T cells in the perivascular compartments of the brain is a unique signature of ECM-inducing malaria infection and implies an important role for this event in the development of the ECM-syndrome.

Spatial profiles of markers of glycolysis, mitochondria, and proton pumps in a rat glioma suggest coordinated programming for proliferation.

BACKGROUND: In cancer cells in vitro, the glycolytic pathway and the mitochondrial tricarboxylic acid (TCA) cycle are programmed to produce more precursor molecules, and relatively less ATP, than in differentiated cells. We address the questions of whether and where these changes occur in vivo in glioblastomas grown from C6 cells in rat brain. These gliomas show some spatial organization, notably in the upregulation of membrane proton transporters near the rim. RESULTS: We immunolabeled pairs of proteins (as well as DNA) on sections of rat brains containing gliomas, measured the profiles of fluorescence intensity on strips 200 µm wide and at least 3 mm long running perpendicular to the tumor rim, and expressed the intensity in the glioma relative to that outside. On averaged profiles, labeling of a marker of the glycolytic pathway, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), was, as expected, greater in the glioma. Over distances up to 2.5 mm into the glioma, expression of a marker of the TCA cycle, Tom20, a pre-protein receptor on the translocation complex of the mitochondrial outer membrane, was also upregulated. The ratio of upregulation of Tom20 to upregulation of GAPDH was, on average, slightly greater than one. Near the rim (0.4-0.8 mm), GAPDH was expressed less and there was a peak in the mean ratio of 1.16, SEM = 0.001, N = 16 pairs of profiles. An antibody to V-ATPase, which, by pumping protons into vacuoles contributes to cell growth, also indicated upregulation by about 40%. When compared directly with GAPDH, upregulation of V-ATPase was only 0.764, SD = 0.016 of GAPDH upregulation. CONCLUSIONS: Although there was considerable variation between individual measured profiles, on average, markers of the glycolytic pathway, of mitochondria, and of cell proliferation showed coherent upregulation in C6 gliomas. There is a zone, close to the rim, where mitochondrial presence is upregulated more than the glycolytic pathway, in agreement with earlier suggestions that lactate is taken up by cells near the rim.

Intravital imaging of a massive lymphocyte response in the cortical dura of mice after peripheral infection by trypanosomes.

Peripheral infection by Trypanosoma brucei, the protozoan responsible for sleeping sickness, activates lymphocytes, and, at later stages, causes meningoencephalitis. We have videoed the cortical meninges and superficial parenchyma of C56BL/6 reporter mice infected with T.b.brucei. By use of a two-photon microscope to image through the thinned skull, the integrity of the tissues was maintained. We observed a 47-fold increase in CD2+ T cells in the meninges by 12 days post infection (dpi). CD11c+ dendritic cells also increased, and extravascular trypanosomes, made visible either by expression of a fluorescent protein, or by intravenous injection of furamidine, appeared. The likelihood that invasion will spread from the meninges to the parenchyma will depend strongly on whether the trypanosomes are below the arachnoid membrane, or above it, in the dura. Making use of optical signals from the skull bone, blood vessels and dural cells, we conclude that up to 40 dpi, the extravascular trypanosomes were essentially confined to the dura, as were the great majority of the T cells. Inhibition of T cell activation by intraperitoneal injection of abatacept reduced the numbers of meningeal T cells at 12 dpi and their mean speed fell from 11.64 ± 0.34 µm/min (mean ± SEM) to 5.2 ± 1.2 µm/min (p = 0.007). The T cells occasionally made contact lasting tens of minutes with dendritic cells, indicative of antigen presentation. The population and motility of the trypanosomes tended to decline after about 30 dpi. We suggest that the lymphocyte infiltration of the meninges may later contribute to encephalitis, but have no evidence that the dural trypanosomes invade the parenchyma.

In vivo imaging of trypanosome-brain interactions and development of a rapid screening test for drugs against CNS stage trypanosomiasis.

HUMAN AFRICAN TRYPANOSOMIASIS (HAT) MANIFESTS IN TWO STAGES OF DISEASE: firstly, haemolymphatic, and secondly, an encephalitic phase involving the central nervous system (CNS). New drugs to treat the second-stage disease are urgently needed, yet testing of novel drug candidates is a slow process because the established animal model relies on detecting parasitemia in the blood as late as 180 days after treatment. To expedite compound screening, we have modified the GVR35 strain of Trypanosoma brucei brucei to express luciferase, and have monitored parasite distribution in infected mice following treatment with trypanocidal compounds using serial, non-invasive, bioluminescence imaging. Parasites were detected in the brains of infected mice following treatment with diminazene, a drug which cures stage 1 but not stage 2 disease. Intravital multi-photon microscopy revealed that trypanosomes enter the brain meninges as early as day 5 post-infection but can be killed by diminazene, whereas those that cross the blood-brain barrier and enter the parenchyma by day 21 survived treatment and later caused bloodstream recrudescence. In contrast, all bioluminescent parasites were permanently eliminated by treatment with melarsoprol and DB829, compounds known to cure stage 2 disease. We show that this use of imaging reduces by two thirds the time taken to assess drug efficacy and provides a dual-modal imaging platform for monitoring trypanosome infection in different areas of the brain.

The spatial organization of proton and lactate transport in a rat brain tumor.

Tumors create a heterogeneous acidic microenvironment which assists their growth and which must be taken into account in the design of drugs and their delivery. In addition, the acidic extracellular pH (pHe) is itself exploited in several experimental techniques for drug delivery. The way the acidity is created is not clear. We report here the spatial organization of key proton-handling proteins in C6 gliomas in rat brain. The mean profiles across the tumor rim of the Na+/H+ exchanger NHE1, and the lactate-H+ cotransporter MCT1, both showed peaks. NHE1, which is important for extension and migration of cells in vitro, showed a peak 1.55 times higher than in extratumoural tissue at 0.33 mm from the edge. MCT1 had a broader peak, further into the tumor (maximum 1.76 fold at 1.0 mm from the edge). In contrast, MCT4 and the carbonic anhydrase CAIX, which are associated with hypoxia, were not significantly upregulated in the rim. The spatial distribution of MCT4 was highly correlated with that of CAIX, suggesting that their expression is regulated by the same factors. Since protons extruded by NHE1 diffuse away through extracellular clefts, NHE1 requires a continuous source of intracellular protons. From the stoichiometries of metabolic pathways that produce or consume H+, and the greater availability of glucose compared to oxygen in most parts of a tumor, we support the classic view that most of the net proton efflux from C6 gliomas originates in glycolytic formation of lactate and H+ inside the tumor, but add that some lactate is taken up into cells in the rim on MCT1, and some lactate diffuses away, leaving its associated protons available to re-enter cells for extrusion on NHE1. Therapeutic inhibition of NHE1, MCT1 or CAIX is predicted to affect different parts of a tumor.

In vivo real-time multiphoton imaging of T lymphocytes in the mouse brain after experimental stroke.

BACKGROUND AND PURPOSE: To gain a better understanding of T cell behavior after stroke, we have developed real-time in vivo brain imaging of T cells by multiphoton microscopy after middle cerebral artery occlusion. METHODS: Adult male hCD2-GFP transgenic mice that exhibit green fluorescent protein-labeled T cells underwent permanent left distal middle cerebral artery occlusion by electrocoagulation (n=6) or sham surgery (n=6) and then multiphoton laser imaging 72 hours later. RESULTS: Extravasated T cell number significantly increased after middle cerebral artery occlusion versus sham. Two T cell populations existed after middle cerebral artery occlusion, possibly driven by 2 T cell subpopulations; 1 had significantly lower and the other significantly higher track velocity and displacement rate than sham. CONCLUSIONS: The different motilities and behaviors of T cells observed using our imaging approach after stroke could reveal important mechanisms of immune surveillance for future therapeutic exploitations.

Subtraction method for intravital two-photon microscopy: intraparenchymal imaging and quantification of extravasation in mouse brain cortex.

Brain pathologies, including stroke and tumors, are associated with a variable degree of breakdown of the blood-brain barrier (BBB), which can usefully be studied in animal models. We describe a new optical technique for quantifying extravasation in the cortex of the living mouse and for imaging intraparenchymal tissue. Leakiness of the BBB was induced by microbeam x-irradiation. Two fluorescent dyes were simultaneously infused intravenously, one of high molecular weight (fluorescein-labeled dextran, 70 kDa, green fluorescence) and one of low molecular weight (sulforhodamine B, 559 Da, red fluorescence). A two-photon microscope, directed through a cranial window, obtained separate images of the two dyes in the cortex. The gains of the two channels were adjusted so that the signals coming from within the vessels were equal. Subtraction of the image of the fluorescein-dextran from that of the sulforhodamine B gave images in which the vasculature was invisible and the sulforhodamine B in the parenchyma could be imaged with high resolution. Algorithms are presented for rapidly quantifying the extravasation without the need for shape recognition and for calculating the permeability of the BBB. Sulforhodamine B labeled certain intraparenchymal cells; these cells, and other details, were best observed using the subtraction method.

A glia-neuron alanine/ammonium shuttle is central to energy metabolism in bee retina.

It has been proposed that glial cells may supply carbon fuel to neurons and also that there are fluxes of ammonium from neurons to glia. We have investigated both these proposals in Apis retinal slices, in which virtually all the mitochondria are in the photoreceptor neurons. Normally the superfusate contained no substrate of energy metabolism; addition of glucose or alanine did not increase oxygen consumption (QO2), confirming that the neurons received adequate substrate from glycogen in the glia. 1,4-Dideoxy-1,4-imino-D-arabinitol (DAB, 100 microm), an inhibitor of glycogen phosphorylase, progressively decreased QO2. This decrease was reversed by alanine but not glucose. Ammonium-sensitive microelectrodes did not detect significant extracellular [NH(4)(+)] ([NH(4)(+)](e)) in slices superfused with normal superfusate. Removal of Cl(-), necessary for cotransport of NH(4)(+) into the glia, increased [NH(4)(+)](e) so that at the end of 2 min photostimulation mean [NH(4)(+)](e) was 0.442 mM (S.E.M. = 0.082 mM, n = 16). In 0 Cl(-), [NH(4)(+)](e) was reduced by 2-(methylamino)isobutyrate (MeAIB) an inhibitor of alanine transport. MeAIB also blocked oxidation of alanine in the presence of DAB, but did not decrease QO2 in normal superfusate. Lactate (l and d) and pyruvate (but not glucose) increased QO2 in DAB and decreased [NH(4)(+)](e) in 0 Cl(-). These results strengthen the evidence that in superfused retinal slices, glucose is metabolized exclusively in the glia, which supply alanine to the neurons, and that ammonium returns to the glia. They also show that another fuel (perhaps lactate) can be supplied by the glia to the neurons.

Serial in vivo spectroscopic nuclear magnetic resonance imaging of lactate and extracellular pH in rat gliomas shows redistribution of protons away from sites of glycolysis.

The acidity of the tumor microenvironment aids tumor growth, and mechanisms causing it are targets for potential therapies. We have imaged extracellular pH (pHe) in C6 cell gliomas in rat brain using 1H magnetic resonance spectroscopy in vivo. We used a new probe molecule, ISUCA [(+/-)2-(imidazol-1-yl)succinic acid], and fast imaging techniques, with spiral acquisition in k-space. We obtained a map of metabolites [136 ms echo time (TE)] and then infused ISUCA in a femoral vein (25 mmol/kg body weight over 110 min) and obtained two consecutive images of pHe within the tumor (40 ms TE, each acquisition taking 25 min). pHe (where ISUCA was present) ranged from 6.5 to 7.5 in voxels of 0.75 microL and did not change detectably when [ISUCA] increased. Infusion of glucose (0.2 mmol/kg.min) decreased tumor pHe by, on average, 0.150 (SE, 0.007; P < 0.0001, 524 voxels in four rats) and increased the mean area of measurable lactate peaks by 54.4 +/- 3.4% (P < 0.0001, 287 voxels). However, voxel-by-voxel analysis showed that, both before and during glucose infusion, the distributions of lactate and extracellular acidity were very different. In tumor voxels where both could be measured, the glucose-induced increase in lactate showed no spatial correlation with the decrease in pHe. We suggest that, although glycolysis is the main source of protons, distributed sites of proton influx and efflux cause pHe to be acidic at sites remote from lactate production.

Cellular pathways of energy metabolism in the brain: is glucose used by neurons or astrocytes?

Most techniques presently available to measure cerebral activity in humans and animals, i.e. positron emission tomography (PET), autoradiography, and functional magnetic resonance imaging, do not record the activity of neurons directly. Furthermore, they do not allow the investigator to discriminate which cell type is using glucose, the predominant fuel provided to the brain by the blood. Here, we review the experimental approaches aimed at determining the percentage of glucose that is taken up by neurons and by astrocytes. This review is integrated in an overview of the current concepts on compartmentation and substrate trafficking between astrocytes and neurons. In the brain in vivo, about half of the glucose leaving the capillaries crosses the extracellular space and directly enters neurons. The other half is taken up by astrocytes. Calculations suggest that neurons consume more energy than do astrocytes, implying that astrocytes transfer an intermediate substrate to neurons. Experimental approaches in vitro on the honeybee drone retina and on the isolated vagus nerve also point to a continuous transfer of intermediate metabolites from glial cells to neurons in these tissues. Solid direct evidence of such transfer in the mammalian brain in vivo is still lacking. PET using [(18)F]fluorodeoxyglucose reflects in part glucose uptake by astrocytes but does not indicate to which step the glucose taken up is metabolized within this cell type. Finally, the sequence of metabolic changes occurring during a transient increase of electrical activity in specific regions of the brain remains to be clarified.

The ammonium-induced increase in rat brain lactate concentration is rapid and reversible and is compatible with trafficking and signaling roles for ammonium.

The glutamate-glutamine shuttle requires a flux of fixed N from neurons to astrocytes. The suggestion that some or all of this N is ammonium has received support from reports that ammonium (as NH(4)(+)) rapidly enters astrocytes. Ammonium might also help control astrocyte energy metabolism by increasing lactate production. If ammonium has these functions, then its effect on brain metabolism must be rapid and reversible. To make a minimal test of this requirement, we have followed the time courses of the changes induced by a 4 min venous infusion of 1 mol/L NH(4)Cl, 2.5 mmol/kg body weight, in rat. Extracellular [NH(4)(+)] in cortex, monitored with ion-selective microelectrodes, reached a peak of approximately 0.7 mmol/L 1.65 mins after the end of the infusion, then recovered. Brain metabolites were monitored non-invasively every 4 mins by (1)H magnetic resonance spectroscopy. Lactate peak area during the 3.2 min acquisition starting at the end of the infusion was 1.84+/-0.24 times baseline (+/-s.e.m., P=0.009, n=9). Lactate increased until 13.2+/-2.1 mins after the end of the infusion and recovered halfway to baseline by 31.2 mins. Glutamate decreased by at least 7.1% (P=0.0026). Infusion of NaCl caused no change in lactate signal. Cerebral blood flow, measured by arterial magnetization labeling, more than doubled, suggesting that the lactate increase was not caused by hypoxia. At least three consecutive ammonium-induced increases in lactate signal could be evoked. The results are compatible with an intercellular trafficking/signaling function for ammonium.

Immunocytochemical analysis of rat vagus nerve by antibodies against glycogen phosphorylase isozymes.

Glycogen is an endogenous store of glucose equivalents for energy metabolism in many tissues. The brain contains a significant amount of glycogen the role of which as an energy reserve is currently under debate. Apparently little is known concerning a possible role of glycogen in peripheral nerves. We have demonstrated immunocytochemically the presence of glycogen phosphorylase (GP), a key enzyme in glycogen metabolism, in large and small axons of the rat vagus nerve, but not in Schwann cells. Furthermore, the isozyme-specific antibodies applied detected only the presence of the brain isoform BB of GP, but not the muscle isoform MM. This is in agreement with the occurrence of solely the BB isoform in the few brain and spinal cord neurons that contain GP. In contrast, astroglial cells in brain and spinal cord have previously been shown to contain both isoforms. Since GP isozymes are regulated differentially, the expression of isoform BB may provide hints to possible functions of glycogen in the vagus nerve.

Neuron-glial trafficking of NH4+ and K+: separate routes of uptake into glial cells of bee retina.

Ammonium (NH4+ and/or NH3) and K+ are released from active neurons and taken up by glial cells, and can modify glial cell behaviour. Study of these fluxes is most advanced in the retina of the honeybee drone, which consists essentially of identical neurons (photoreceptors) and identical glial cells (outer pigment cells). In isolated bee retinal glial cells, ammonium crosses the membrane as NH4+ on a Cl- cotransporter. We have now investigated, in the more physiological conditions of a retinal slice, whether the NH4+-Cl- cotransporter can transport K+ and whether the major K+ conductance can transport NH4+. We increased [NH4+] or [K+] in the superfusate and monitored uptake by recording from the glial cell syncytium or from interstitial space with microelectrodes selective for H+ or K+. In normal superfusate solution, ammonium acidified the glial cells but, after 6 min superfusion in low [Cl-] solution, ammonium alkalinized them. In the same low [Cl-] conditions, the rise in intraglial [K+] induced by an increase in superfusate [K+] was unchanged, i.e. no K+ flux on the Cl- cotransporter was detected. Ba2+ (5 mm) abolished the glial depolarization induced by K+ released from photoreceptors but did not reduce NH4+uptake. We estimate that when extracellular [NH4+] is increased, 62-100% is taken up by the NH4+-Cl- cotransporter and that when K+ is increased, 77-100% is taken up by routes selective for K+. This separation makes it possible that the glial uptake of NH4+ and of K+, and hence their signalling roles, might be regulated separately.

Uptake of locally applied deoxyglucose, glucose and lactate by axons and Schwann cells of rat vagus nerve.

We asked whether, in a steady state, neurons and glial cells both take up glucose sufficient for their energy requirements, or whether glial cells take up a disproportionate amount and transfer metabolic substrate to neurons. A desheathed rat vagus nerve was held crossways in a laminar flow perfusion chamber and stimulated at 2 Hz. (14)C-labelled substrate was applied from a micropipette for 5 min over a < 0.6 mm band of the surface of the nerve. After 10-55 min incubation, the nerve was lyophilized and the longitudinal distribution of radioactivity measured. When the weakly metabolizable analogue of glucose, 2-deoxy-[U-(14)C]D-glucose (*DG), was applied, the profiles of the radioactivity broadened with time, reaching distances several times the mean length of the Schwann cells (0.32 mm; most of the Schwann cells are non-myelinating). The profiles were well fitted by curves calculated for diffusion in a single compartment, the mean diffusion coefficient being 463 +/- 34 microm(2) s(-1) (+/- S.E.M., n = 16). Applications of *DG were repeated in the presence of the gap junction blocker, carbenoxolone (100 microM). The profiles were now narrower and better fitted with two compartments. One compartment had a coefficient not significantly different from that in the absence of the gap junction blocker (axons), the other compartment had a coefficient of 204 +/- 24 microm(2) s(-1), n = 4. Addition of the gap junction blocker 18-alpha-glycyrrhetinic acid, or blocking electrical activity with TTX, also reduced longitudinal diffusion. Ascribing the compartment in which diffusion was reduced by these treatments to non-myelinating Schwann cells, we conclude that 78.0 +/- 3.6 % (n = 9) of the uptake of *DG was into Schwann cells. This suggests that there was transfer of metabolic substrate from Schwann cells to axons. Local application of [(14)C]glucose or [(14)C]lactate led to variable labelling along the length of the nerve, but with both substrates narrow peaks were often present at the application site; these were greatly reduced by subsequent treatment with amylase, a glycogen-degrading enzyme.