Doyle G. Graham

Pathogenetic Studies of Hexane and Carbon Disulfide Neurotoxicity

Two commonly employed solvents, n-hexane and carbon disulfide (CS2), although chemically dissimilar, result in identical neurofilament-filled swellings of the distal axon in both the central and peripheral nervous systems. Whereas CS2 is itself a neurotoxicant, hexane requires metabolism to the gamma-diketone, 2,5-hexanedione (HD). Both HD and CS2 react with protein amino functions to yield initial adducts (pyrrolyl or dithiocarbamate derivatives, respectively), which then undergo oxidation or decomposition to an electrophile (oxidized pyrrole ring or isothiocyanate), that then reacts with protein nucleophiles to result in protein cross-linking. It is postulated that progressive cross-linking of the stable neurofilament during its anterograde transport in the longest axons ultimately results in the accumulation of neurofilaments within axonal swellings. Reaction with additional targets appears to be responsible for the degeneration of the axon distal to the swellings.

Studies of the molecular pathogenesis of hexane neuropathy: II. Evidence that pyrrole derivatization of lysyl residues leads to protein crosslinking☆☆☆

Abstract In the reaction between ethanolamine and 2,5-hexanedione, 1-(2-hydroxyethyl)-2,5-dimethylpyrrole was formed, and the pyrrole was found to autoxidize to form an orange chromophore. Similar orange chromophores were observed in the reaction of 2,5-hexanedione, 2,5-heptanedione, and 3,6-octanedione with a variety of primary amines and with proteins. The development of the orange chromophore in the reaction of 2,5-hexanedione with proteins was attended by a proportional derivatization of lysyl residues and by extensive intramolecular and intermolecular crosslinking. These observations suggest that the sequence of events in the crosslinking of neurofilaments during chronic n -hexane intoxication may be metabolism to 2,5-hexanedione, formation of an imine with lysyl residues, cyclization to form a pyrrole, autoxidation of the pyrrole, and finally covalent crosslinking involving pyrrole rings.

The effect of 3,4-dimethyl substitution on the neurotoxicity of 2,5-hexanedione: II. Dimethyl substitution accelerates pyrrole formation and protein crosslinking

3,4-Dimethyl-2,5-hexanedione and 2,5-hexanedione were reacted with model amines to yield N-substituted 2,3,4,5-tetramethylpyrroles and 2,5-dimethylpyrroles, respectively. When compared to the unsubstituted parent compound 2,5-hexanedione, 3,4-dimethyl-2,5-hexanedione was found to cyclize approximately eight times as rapidly on a molar basis at 37 degrees C, with an activation energy of 3290 cal/mole less than 2,5-hexanedione. In addition, 1-benzyl-2,3,4,5-tetramethylpyrrole oxidized more readily than 1-benzyl-2,5-dimethylpyrrole with a difference in the half-wave potentials of 0.29 V. Both gamma-diketones led to progressive crosslinking of proteins in vitro, with the dimethyl substitution accelerating this process by a factor of 40. The formation of pyrrolyl derivatives in vivo was demonstrated by the characteristic absorption spectra obtained following reaction of erythrocyte proteins from intoxicated rats with Ehrlichs reagent. There was progressive formation of protein-bound dimethylpyrroles following exposure to 2,5-hexanedione and formation of tetramethylpyrroles following exposure to 3,4-dimethyl-2,5-hexanedione in vivo. Preparations of axonal pads also demonstrated pyrrole derivatization in vivo. In addition, spectrin preparations of erythrocytes from intoxicated rats showed a large amount of high molecular weight protein (400,000 Da), corresponding to dimerized spectrin. Thus, 3,4-dimethyl-2,5-hexanedione, which is 20 to 30 times more potent on a molar basis than 2,5-hexanedione in leading to a neurofilamentous neuropathy, is associated with more rapid pyrrole formation and protein crosslinking in vitro, and it has been demonstrated that these processes occur in vivo. These observations support the hypothesis that pyrrole formation and autoxidation occur following exposure to gamma-diketones, leading to covalent crosslinking of proteins in vivo, a process which may explain the pathogenesis of neurofilament accumulation in these neuropathies.

Crosslinking of Apolipoprotein E by Products of Lipid Peroxidation

Apolipoprotein E (APOE) genotype and advancing aging are interacting risk factors in the expression of lateonset and sporadic Alzheimers disease (AD). We tested the hypothesis that 2 products of lipid peroxidation, malondialdehyde (MDA) and 4-hydroxy-2-nonenal (HNE), covalently modify APOE and alter its metabolism. In vitro, both HNE and MDA crosslinked purified APOE3 and APOE4. HNE was a more potent crosslinker than MDA, and purified APOE3 was more susceptible to crosslinking by HNE than was purified APOE4. In P19 neuroglial cultures, oxidative stress with lipid peroxidation led to increased intracellular accumulation of anti-HNE and anti-APOE immunoreactive proteins of approximately SO kDa. Intracellular accumulation of the 50 kDa APOE-immunoreactive protein (APOE-50) was not prevented by cyclohexamide, suggesting formation by post-translational mechanisms. In CSF, a 50 kDa APOE-immunoreactive protein co-migrated with proteins most immunoreactive for HNE and MDA adducts, and containing NaB3H4-reducible bonds. These proteins were in CSF from adult subjects (with or without dementia), and in AD patients homozygous for APOE3 or APOE4 alleles. These data suggest that HNE covalently crosslinks APOE in P19 neuroglial cultures to form a 50 kDa protein, and that similar modifications of APOE appear to occur in vivo.

Neurotoxicity of Endogenous Cysteinylcatechols

Progression of Parkinsons disease has been associated with several biochemical changes in the substantia nigra including increased oxidative challenge, catechol oxidation, and inhibition of mitochondrial complex I activity. Cysteinylcatechols, formed by nucleophilic addition of cysteine to oxidized catechols, have been identified as markers of catechol oxidation in brain tissue. We have examined the neurotoxicity of a series of cysteinylcatechols. Of the compounds examined, only 5-S-cysteinyl-3,4-dihydroxyphenylacetate (cysdopac) was specifically cytotoxic to differentiated P19 neuroglial cultures. Cysdopac also was neurotoxic to pyramidal neurons in organotypic cultures of hippocampus, and this effect was ablated by selective N-methyl-D-aspartate (NMDA) receptor antagonists. In vitro, cysdopac was a potent inhibitor of mitochondrial complex I activity. However, electrophysiologic experiments failed to demonstrate NMDA receptor agonist activity for cysdopac, nor did cysdopac inhibit glutamate uptake. These results showed that cysdopac was the most potent neurotoxin of this series of cysteinylcatechols and suggest that cysdopac may function as an indirect excitotoxin, potentially via inhibition of mitochondrial respiration.

Evidence that pyrrole formation is a pathogenetic step in γ-diketone neuropathy☆

Abstract Previous studies from this laboratory have demonstrated that the addition of methyl groups at the 3 and 4 positions of the 2,5-hexanedione (2,5-HD) molecule results both in more rapid pyrrole formation and in enhanced neurotoxicity. In order to define more clearly the relationship between rates of pyrrole formation and neurotoxicity, the dl and meso diastereomers of 3,4-dimethyl-2,5-hexanedione (DMHD), 3,4-diethyl-2,5-hexanedione (DEHD), and 3,4-diisopropyl-2,5-hexanedione (DiPHD) were synthesized and purified. The rates of pyrrole formation were compared with that of unsubstituted 2,5-HD, and rates of in vitro crosslinking were determined. Each of the compounds was administered to rats to determine relative neurotoxicity. Hindlimb paralysis was reached after a total administered dose of 1.6 mmol/kg of dl -DMHD, while 5.9 mmol/kg of meso -DMHD was required. Paralysis was not achieved with either diastereomer of DEHD or DiPHD, although both produced systemic toxicity. Histologic sections of spinal cords and anterior roots from rats treated with DMHD revealed large neurofilament-filled axonal swellings, while more distal sections contained axons undergoing Wallerian-type degeneration. Neither axonal swellings nor Wallerian-type degeneration were seen in sections from spinal cord or peripheral nerve of rats treated with DEHD or DiPHD. The rates of pyrrole formation were in the order dl -DMHD > meso -DMHD > 2,5-HD > dl -DEHD > meso -DEHD > dl -DiPHD . meso -DiPHD, while in vitro crosslinking rates were in the order dl -DMHD > meso -DMHD . dl -DEHD > meso -DEHD > 2,5-HD > dl -DiPHD > meso -DiPHD. Cyclic voltammetry showed that the autoxidation of pyrroles derived from DMHD, DEHD, and DiPHD occurred more readily than that derived from 2,5-HD. In addition, we report for the first time the segregation of axoplasmic organelles in animals treated with DMHD, providing further evidence that the neurofilamentous axonopathies caused by such compounds as β,β′-iminodipropionitrile (IDPN), 2,5-HD and CS 2 share a common underlying mechanism. The strong correlations between rates of pyrrole formation, rates of in vitro crosslinking and relative neurotoxicity are seen as evidence that pyrrole formation is a step in the pathogenetic sequence of γ-diketone neuropathy.

Covalent Crosslinking of Neurofilament Proteins by Oxidized Catechols as a Potential Mechanism of Lewy Body Formation

Brainstem Lewy bodies (LB) are neuronal inclusions that are closely related to Parkinsons disease (PD). The filamentous component of LB from patients with PD contains biochemically altered neurofilaments (NF). Herein we have tested the hypothesis that the oxidized products of catechols may covalently crosslink NF. Neurofilaments were incubated in the presence of oxidized L-dopa, dopamine, or dopac and then analyzed by SDS-PAGE and protein staining or immunoblotting with monoclonal antibodies specific for neurolfilamcnt subunit proteins. Oxidized catechols yielded the same pattern of NF protein crosslinking as known covalent crosslinking agents. Coincubation of NF and catechols with Na-acetyl-L-lysine (NAL) produced strong reactivity on immunoblots probed with a polyclonal antiserum specific for NAL crosslinked to protein (antiserum 1400/3). Crosslinking of NAL to model proteins by oxidized dopac was followed by antibody capture assays using antiserum 1400/3. Increasing immunoreactivity was observed for 0.01 to 1.0 mM dopac and was augmented by Cu2+, Fe2+, Fe3+, Mn2+, or Mn3+ up to 0.1 mM. These results show that the products of catechol oxidation can covalently crosslink neurofilaments, that the crosslinking mechanism can involve lysine, and that copper, iron, and manganese ions can accelerate catechol-mediated protein crosslinking.

The relative neurotoxicities of n-hexane, methyl n-butyl ketone, 2,5-hexanediol, and 2,5-hexanedione following oral or intraperitoneal administration in hens

Abstract The sensitivity of the hen to neurotoxicity produced by po and ip administration of n -hexane, methyl n -butyl ketone (M n BK), 2,5-hexanediol (2,5-HDOH), and 2,5-hexanedione (2,5-HD) was investigated. While po administration of one or two doses of these chemicals at a 21-day interval caused acute effects, it did not induce neuropathy in treated hens. Subchronic po or ip administration of n -hexane caused only weakness, which subsided after cessation of administration. By contrast, subchronic administration of the other three related compounds caused neurotoxicity characterized by ataxia, which progressed to paralysis in some hens. Severity of the neurotoxic effect was dependent on both the test compound and its route of administration of a similar dosage. Generally, ip injection caused more severe effects than po administration. Pathological examination of nervous system tissues of hens treated with the 2,5-HD, 2,5-HDOH, and M n BK showed giant paranodal axonal swelling followed by Wallerian degeneration of axons and myelin in peripheral nerve and spinal cord. Wallerian degeneration in the spinal cord was observed almost exclusively in the ventral columns of the lower spinal cord. n -Hexane failed to produce the characteristic pathological features produced by related compounds. The neurotoxic potency of these chemicals which considers onset and magnitude of clinical signs and severity of histopathologic changes was in descending order: 2,5-HD > 2,5-HDOH > M n BK > n -hexane when given by either method.

Studies of the molecular pathogenesis of hexane neuropathy. I. Evaluation of the inhibition of glyceraldehyde‐3‐phosphate dehydrogenase by 2,5‐hexanedione

Inhibition of the sulfhydryl enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) by 2,5-hexanedione (2,5-HD) was found to be irreversible, proceeding via a reversible enzyme-inhibitor intermediate, while acetone was a weak reversible inhibitor. Comparison of 2,5-HD and acetone with p-chloromercuribenzoate (PCMB) and N-ethylmaleimide (NEM) demonstrated that the former are not significant sulfhydryl reagents, since they must be present at more than 10(4) times higher concentrations than PCMB or NEM to effect measurable inhibition of this enzyme. Thus it is unlikely that inhibition of GAPDH by 2,5-HD has any significance in the molecular pathogenesis of hexane neuropathy. The irreversibility of 2,5-HD inhibition, on the othe hand, suggests that 2,5-HD reacts with amino groups rather than sulfhydryl groups on proteins. This reaction is proposed as the molecular lesion in hexane neuropathy.

Delayed neurotoxicity of O-ethyl O-4-nitrophenyl phenylphosphonothioate: Subchronic (90 days) oral administration in hens

Abstract Delayed neurotoxicity in hens was produced following daily oral administration of 0.1, 0.5, 1.0, 2.5, 5.0, and 10.0 mg/kg of technical (85%) O -ethyl O -4-nitrophenyl phenylphosphonothioate (EPN) in gelatin capsules for 90 days. Daily, three groups of hens were given empty gelatin capsules, 10 mg/kg of tri- o -cresyl phosphate (TOCP), or 1 mg/kg of parathion ( O,O -diethyl O -4-nitrophenyl phenylphosphorothioate) and served as gelatin capsule controls, positive controls, and negative controls, respectively. TOCP-Treated hens developed delayed neurotoxicity, and those given parathion showed leg weakness with subsequent recovery when the administration of this agent had stopped. The clinical condition of most ataxic hens deteriorated during the 30-day observation period following the end of the oral administration of EPN. Severity of the clinical condition depended on the size of the daily ingested dose, i.e., while hens given small doses showed only ataxia, those treated with large doses progressed to paralysis and died. Days of administration and “total administered dose” before onset of ataxia depended on the daily dose. Degeneration of myelin and axons in the spinal cord was the most consistent histologic change and was identical to that found in TOCP control hens. Only one hen showed sciatic nerve degeneration. Livers from two hens given the highest dose of EPN manifested a moderate degree of hemosiderosis. Plasma cholin esterase was significantly inhibited in all surviving hens given EPN or TOCP at the end of the observation period. A group of hens treated daily with 0.01 mg/kg of EPN showed no abnormality in gait or behavior, and its plasma cholinesterase activity was not significantly different from that of the control. Hens treated with parathion had plasma cholinesterase activity comparable to that of the control 30 days after the administration of the last dose.