Aspartate-induced neuronal necrosis in infant mice: protective effect of carbohydrate and insulin.

Infant mice given large doses of glutamate or aspartate develop hypothalamic neuronal necrosis. Studies by others demonstrated that simultaneous administration of carbohydrate or prior injection with insulin markedly decreased glutamate-induced neuronal damage. We investigated whether carbohydrate and insulin exert a similar protective effect against aspartate-induced neuronal necrosis. Eight-day-old mice administered aspartate at 750 and 1000 mg/kg body weight developed neuronal necrosis (45.9 +/- 7.2 and 80.8 +/- 17.3 necrotic neurons/section, respectively). When carbohydrate (1 g/kg body weight) was administered simultaneously no lesions were detected in mice administered 750 mg/kg body weight aspartate, while 30.1 +/- 14.2 necrotic neurons/section were noted at 1000 mg aspartate/kg body weight. Mice administered 1000 mg/kg body weight aspartate with prior injection of insulin had 28.4 +/- 12.6 necrotic neurons/section, while 4.2 +/- 1.4 necrotic neurons/section were noted in insulin treated mice given 750 mg aspartate/kg body weight. Carbohydrate and insulin treatments has only minimal effects on plasma aspartate concentrations.

Utilization of intravenously administered glycogen by young pigs.

Previous studies evaluated solutions of small oligosaccharides as potential sources of carbohydrate-derived energy for patients fed intravenously. Although results with these solutions were disappointing, the data suggested that very large oligosaccharides were potential sources of intravenous carbohydrate. To test this hypothesis, four young pigs (3.6 +/- 0.2 kg; mean +/- SD) were infused with sterile solutions for a 6-d period. On days 1 and 6, a balanced isotonic electrolyte solution was infused. On days 2-5 a 9% solution of glycogen was infused at a rate providing 17.7 +/- 0.77 g/d. For each study day the remaining portion of the energy, protein, essential fatty acids and micronutrients was supplied enterally. No adverse reactions were noted during glycogen infusion, and the animals continued to grow. Glycogen utilization was 66.4 +/- 4.3%. Of the total carbohydrate excreted, 85.4% was composed of oligosaccharides of maltotetraose size or larger. Free glucose accounted for 3.5% of the total excreted, while maltose plus maltotriose accounted for 11.1%. Plasma concentrations of oligosaccharide-bound glucose increased during glycogen infusion, rising from a base-line value of 11.0 +/- 14 mg/dL to an overall mean value for the 4-d period of 100.3 +/- 31.6 mg/dL.

Correlation of glutamate plus aspartate dose, plasma amino acid concentration and neuronal necrosis in infant mice.

Eight-day-old mice were given by gavage glutamate and aspartate mixtures providing each amino acid at 125, 250 or 500 mg/kg body weight (250, 500 and 1000 mg total dicarboxylic amino acids/kg) and the degree and extent of neuronal necrosis were determined. Similar studies were carried out in mice given monosodium L-glutamate at 250 or 500 mg/kg body weight. Plasma aspartate and glutamate concentrations were determined at each dose level. No animal given either glutamate or the glutamate plus aspartate mixture at 250 mg/kg developed neuronal necrosis. However, neuronal necrosis developed in 30% of animals given glutamate at 500 mg/kg (12+/-2 necrotic neurons/section in the region of maximal damage) and in 17% of animals given 250 mg glutamate/kg plus 250 mg aspartate/kg (11-13 necrotic neurons/section in the region of maximal damage). The threshold mean peak plasma glutamate plus aspartate concentration associated with neuronal necrosis was 128+/-24 mumol/dl. Using these data, and previously published data for aspartate-induced neurotoxicity (Finkelstein et al. Toxicology 1983, 29, 109), the individual threshold plasma glutamate and aspartate concentrations associated with neuronal necrosis were calculated to be 110 mumol/dl for aspartate and 75 mumol/dl for glutamate.

Effect of carbohydrate ingestion on plasma aspartate concentrations in infant mice administered sodium L-aspartate.

Previous studies of infant pigs have demonstrated that simultaneous ingestion of carbohydrate (1 g/kg body weight) and glutamate (300 mg/kg body weight) resulted in markedly lower mean peak plasma glutamate concentration and a smaller area under the plasma glutamate concentration-time curve (AUC) than ingestion of the equivalent amount of glutamate alone. This study was carried out to investigate whether a similar carbohydrate-induced effect occurred with the other dicarboxylic amino acid, aspartate. Eight-day-old mice were administered sodium L-aspartate at 250, 500 and 1000 mg/kg body weight with and without 1.0 g/kg body weight of partially hydrolyzed cornstarch. Mean peak plasma aspartate concentration and plasma aspartate AUC values increased in proportion to the aspartate dose. The addition of carbohydrate to the aspartate solution had no significant effect on either mean peak plasma aspartate concentrations or AUC values at aspartate doses of 250 and 500 mg/kg body weight. A modest, but significant effect of carbohydrate was noted on the mean peak plasma aspartate levels in animals administered 1000 mg/kg body weight aspartate (P less than 0.05, Student's t-test). However, analysis of variance showed no significant carbohydrate effect and plasma AUC values were not significantly affected. These data indicate that carbohydrate affects the metabolism of aspartate and glutamate differently.

Correlation of aspartate dose, plasma dicarboxylic amino acid concentration, and neuronal necrosis in infant mice.

Eight-day-old mice were administered aspartate at 0, 1.88, 3.76, 4.89, 5.64 and 7.52 mmol/kg body wt and the degree and extent of neuronal necrosis determined. In addition, plasma aspartate and glutamate concentrations were determined at each aspartate dose. Animals administered aspartate at 0, 1.88 and 3.76 mmol/kg body wt did not develop neuronal necrosis. Hypothalamic neuronal necrosis (7.33 +/- 1.52 necrotic neurons/section of maximal damage) was found in 3 of 10 animals administered aspartate at 4.89 mmol/kg body wt. The extent of neuronal necrosis was proportional to dose once a neurotoxic dose of aspartate was reached. All 12 animals administered aspartate at 5.64 mmol/kg body wt developed lesions (49.5 +/- 7.2 necrotic neurons/section of maximal damage). Similarly, all 18 mice administered aspartate at 7.52 mmol/kg developed hypothalamic lesions (80.8 +/- 17.8 necrotic neurons/section of maximal damage). Infant mice administered the highest dose of aspartate not producing neuronal necrosis (3.76 mmol/kg) had a mean (+/- S.D.) peak plasma aspartate concentration of 87 +/- 23 mumol/dl and a mean peak plasma glutamate concentration of 64 +/- 22 mumol/dl. Thus, the toxic threshold for these amino acids must be greater than those values.

Utilization of intravenously administered beta-cellobiose and maltose by young pigs.

Intravenous solutions of glucose oligosaccharides are potential sources of carbohydrate-derived energy for patients requiring intravenous feeding. Relatively little is known about utilization of glucose oligosaccharides linked by beta-glucosidic bonds. We compared the utilization of maltose (alpha-D-glucosyl-1,4-D-glucose) and beta-cellobiose (beta-D-glucosyl-1,4-D-glucose) when administered intravenously (19 g per day) to young pigs for a 5-day period. Animals infused with maltose excreted 15% of the infused disaccharide over the 5-day infusion period. No evidence of maltose accumulation was noted in plasma, and kidney morphology was normal. Animals infused with beta-cellobiose excreted 95% of the infused disaccharide in the urine. The mean (+/- SD) plasma total glucose concentration increased significantly over base-line values of 114 +/- 39 mg/dl to a value of 180 +/- 28 mg/dl during cellobiose infusion, indicating accumulation of cellobiose in body water. Kidney morphology in cellobiose-infused animals was normal. Intravenously infused beta-cellobiose is poorly utilized by the pig when compared with the utilization of its alpha-1,4 linked isomer, maltose.

Segmental localization of sensory cells that innervate the bladder.

The present study labels the neuronal cell bodies that give rise to afferent fibers that innervate the bladder of cat and rat. The method used was the retrograde transport of horseradish peroxidase (HRP) from its injection site in the bladder to cells in various dorsal root ganglia. In the rat, the labelled cells are located in the L1-L2 and L6-S1 dorsal root ganglia. In the cat, the labelled cells are located in the L2-L5 and S1-S4 dorsal root ganglia. This confirms older clinical findings, and for the first time directly demonstrates the afferent cell bodies for the bladder. The bladder afferents are small ganglion cells in both rat and cat, and because there is a correlation between the size of axon and the cell body from which it originates, we conclude that the great majority of bladder afferents are small myelinated or unmyelinated axons. In addition, by restricting the HRP to one side of the bladder, we are able to show that some afferent cell bodies send their distal processes across the midline. These results will be useful in considerations of the neural control of bladder function.

Nuclei in which functionally identified spinothalamic tract neurons terminate.

The approximate level of termination of the axons of individual, functionally characterized spinothalamic tract neurons within the monkey thalmus was mapped by antidromic activation using a monopolar electrode which was moved in a systematic grid of tracks through the thalamus. The course of individual axons could be followed through several thalamic levels, and in a few cases branches to both the VPL nucleus and to the intralaminar nuclei were demonstrated. Most of the axons studied, however, projected just to the VPLc or VPLo nuclei. The spinothalamic tract cells that projected to the VPLc nucleus included representative of all known functional categories: low threshold, wide dynamic range, high threshold and "deep." It is speculated that these different classes of spinothalamic projections could make contributions to such sensory modalities as touch, proprioception and pain.

Primary afferent fibers in the tract of Lissauer in the rat.

More than two-thirds of the axons in the tract of Lissauer at mid-thoracic and lumbosacral levels of the rat spinal cord are primary afferent fibers. The proportions of primary afferents in the tract are approximately the same at the two spinal levels. A slightly higher percentage of the unmyelinated, as opposed to the myelinated, fibers are primary afferents. There is a somewhat greater percentage of primary afferent axons in medial parts of the mid-thoracic levels, but all areas of the tract that were examined contain a majority of primary afferent fibers. The primary afferent axons appear to travel less than a segment in the tract at mid-thoracic levels but for several segments in the tract at lumbo-sacral levels. These data indicate that the tract of Lissauer is predominately a primary afferent fiber system in these segments of the rat.

Spinal cord potentials evoked by cutaneous afferents in the monkey.

1. Negative intermediary cord potentials and the equivalent field potentials were recorded from the surface or within the monkey lumbosacral spinal cord in response to stimulation of myelinated afferent fibers in cutaneous or mixed nerves of the hindlimb. 2. Cord potentials resembling the N1 and N2 potentials described in the cat spinal cord were found but, in addition, activation of small myelinated fibers produced a later potential named here the N3-wave. By use of a subtraction technique, it is estimated that the N3-wave has a latency of 11.4 (+/- 3.5 SD) ms from the time of arrival of the volley in the largest affs at 9 (+/- 3) ms after its onset, and the wave lasts 23 (+/- 5.7) ms. 3. The N3-wave is not lost following spinal cord transection, but may instead be enhanced. It is thus due to neural circuitry intrinsic to the lumbosacral spinal cord. 4. The longitudinal distribution of the N3-wave is similar to that of the N1- and N2-waves. 5. The field potential associated with the N3-wave and recorded from within the spinal cord has two negative foci in some animals: near the dorsalmost part of the dorsal horn and in an area equivalent to Rexed's laminae IV-VI. The field potential reverses in sign in the ventral horn. 6. The N3-wave is evoked by Adelta fibers. This was shown by grading the stimulus strength, by measuring the conduction delay for producing the wave when stimuli are applied either proximally or distally on the sural nerve, and by showing that the N3-wave persists when the Aalphabeta fibers are anodally blocked. 7. There is often a late burst discharge in spinal neurons, including spinothalamic tract neurons, which can be attributed to Adelta fibers and which corresponds in time to the N3-wave. 8. It is proposed that the N3-wave can be used as a monitor of the central effects of Adelta fibers in the spinal cord.

Organization and receptive fields of primate spinothalamic tract neurons.

A technique is described for recording from axons belonging to the spinothalamic tract of the monkey. The axons arose from cell bodies located within the spinal cord since the latency of orthodromic activation by afferents within the dorsal funiculus was short. The axons were antidromically activated from the ipsilateral diencephalon. The spectrum of conduction velocities indicates that the recordings favored large-diamter axons. However, all of the classes of spinothalamic tract units described from soma-dendritic recordings were represented in the sample. When the locations of the axons in the ventrolateral white matter were mapped, there was virtually complete overlap in the distributions of hair-activated, low-, and high-threshold spinothalamic tract axons, suggesting that the "lateral spinothalamic tract" conveys tactile, as well as pain and temperature, information. The only segregated population of axons were those belonging to units activated by receptors in deep tissues, including muscle. These were in a band along the ventral surface of the cord. The stimulus points for antidromically activating spinothalamic cells of axons were in the known diencephalic course of the spinothalamic tract, including the ventral posterior lateral nucleus. Stimulus point locations were similar for high-threshold and other categories of units. Receptive-field sizes were smaller for high-threshold spinothalamic cells or axons than for hair-activated or low-threshold units. Receptive-field size was correlated with position on the hindlimb. The smallest fields belonged to cells in lamina I, with progressively larger sizes for cells in laminae IV and V. Receptive-field shape was evaluated by the length/width ratio, which was smallest for high-threshold units and progressively larger for low-threshold and hair-activated units. The receptive-field positions of spinothalamic tract axons were related to the locations of the axons. There was a rough somatotopic representation in the tract, with the most caudal dermatomes represented dorsolaterally, and the most rostral ventromedially.

Responses of primate spinothalamic tract neurons to electrical stimulation of hindlimb peripheral nerves.

The responses of spinothalamic tract neurons were studied by extra- and intracellular recordings from the lumbosacral spinal cord in anesthetized rhesus monkeys (Macaca mulatta). The neurons were identified by antidromic activation from the contralateral diencephalon. They were then classified by the mildest form of mechanical stimulation applied to the ipsilateral hindlimb. The effects of electrical stimulation of the nerve(s) supplying the receptive field were investigated. Graded electrical stimulation revealed that the threshold responses of spinothalamic tract neurons excited by weak mechanical stimuli occurred when the largest afferent fibers were activated. On the other hand, neurons that required intense mechanical stimulation for their excitation tended to have higher thresholds to electrical stimulation. Some spinothalamic tract cells were shown to receive monosynaptic excitatory connections from peripheral nerve fibers, although polysynaptic connections may generally be more important. An input from unmyelinated afferent fibers was demonstrated. It is concluded the primate spinothalamic tract neurons receive a rich convergent input from a variety of cutaneous receptors. The experiments provide some evidence for the most likely types of receptors.