2,5-Hexanedione-induced accumulations of neurofilament-immunoreactive material throughout the rat autonomic nervous system

Brain Research, 444 (1988) 383-388

383

Elsevier BRE 22798

2,5-Hexanedione-induced accumulations of neurofilamentimmunoreactive material throughout the rat autonomic nervous system G i a n - L u c a Ferri I , S o u s s a n Z a r e h 1, A n n a A m a d o r i I , A n n a B a s t o n e 2, Maria S b r a c c i a 2, D o r i s D a h l 3 a n d N o r a Frontali 2 t Department oJ'Anatomy, 'T or Vergata' University, Rome (Italy), Zlstituto Superiore di Sanit~, Rome (Italy) and 3Department of Neuropathology, Harvard Medical School, Boston, MA (U.S.A.), and Spinal Cord Injury Research, West Roxbury Veterans Administration Medical Center, West Roxbury, MA 02132 (U.S.A)

(Accepted 8 December 1987) Key words: 2,5-Hexanedione; Neurofilament; Neuronal cytoskeleton; Autonomic nerve; Enteric nervous system

In rats intoxicated with 2,5-hexanedione, nerve fibres supplyingvirtually all visceral organs showed large numbers of densely immunoreactive accumulations of neurofilament-like material, of fusiform, elongated, smoothlytapering morphology. In the gut, round t o oval, morphologicallydifferent lesions were also present, and abnormal neurofilament-immunoreactive accumulations were revealed in oesophageal terminal end-plates. An extensive damage to autonomic nerve fibres, which are largely non-myelinated, was thus revealed in 2,5-hexanedione intoxication. The observed diversity in lesion morphologymay suggest heterogeneity in cytoskeletal and/or associated proteins among autonomic neurones.

The toxic effects of hexacarbon solvents on the peripheral nervous system are well established, nhexane and methyl n-butyl ketohe producing a typical symmetrical sensorimotor polyneuropathy 24'25. In addition, signs and symptoms suggestive of autonomic dysfunction were also noteo, such as blurred vision, constipation, hyperhydrosis, or impotence (see refs. 24, 25). In rat and other animals treatment with hexacarbons, or their major toxic derivative 2,5hexanedione, induces a closely comparable sensorimotor neuropathy, showing, like human cases, neurofilament-containing giant axonal swellings, starting in distal, pre-terminal axons 2a. Interestingly, a decreased pupillomotor i'esponse was revealed in 2,5-hexanedione-treated rats ~, while altered neurotransmission and certain axonal changes were shcwa in the superior cervical ganglion 19. In spite of such clinical and experimental indications of vegetative dysfunction, however, little is known concerning the neuropathology of autonomic changes in hexacarbon

intoxication. Using neurofilament-protein immunocytochemistry, we have thus studied the innervation of visceral organs in rats exposed to 2,5-hexanedione. Male Sprague-Dawley rats (n = 9, Charles River, Italy) were injected intraperitoneally with 2,5hexanedione (400 mg/kg b. wt., 5 days/week for 3.5 weeks). Normal male rats (n = 9, of the same strain, source, age and initial b. wt. about 300 g) were used as controls. Since major reductions in food intake are observed in 2,5-hexanedione intoxication 3, 6 control animals were pair-fed with an equal number of treated ones. At sacrifice, tissue samples were taken from iris, respiratory tract (trachea and bronchi), gut (oesophagus, stomach, small and large bowel), cardiovascular system (including renal and iliac arteries and veins), urogenital tract (ureter, bladder, penis, testis, epididymis, vas deferens, seminal vesicles) and diaphragm. Muscularis externa and submucosa were layer-separated from samples of gut and blad-

Correspondence: G.-L. Ferri, Department of Anatomy, 'Tor Vergata' University, via O. Raimondo, 00173 Rome, Italy.

384 tier, as described s'9, stretched onto glass slides and allowed to dry. The membranous portion of trachea and/or main bronchi, the whole wall of ureter, arteries and veins (opened), as well as samples of mesentery and diaphragm (full-thickness), were similarly flattened and dried on slides. Preparations were then rehydrated (about 1 h) in phosphate-buffered saline (0.05 tool/I, pH 7.2) containing Triton X-100 (1%: PBS-Triton), detached from slides, pre-incubated (2-4 h) in PBS-Triton containing normal rat and sheep sera (3% each), bovi:e serum albumin (0.1%) and sodium azide (0.02%) and stained by immunofluorescence 5. The antiserum R39 which recognises the three subunits of the neurofilament triplet 6'7 was used. Crossreactivffy of the antiserum with tubulin and high molecular weight microtubule-associated proteins (MAPs) could be excluded histologically, as Purkinje cells in the cerebellar cortex were not stained 6. To exclude cross-reactivity with tau MAPs, tubulin isolated from calf brain by 1-3 cycles of polymerisationdepolymerisation (tau MAPs copurify with tubulin in this procedure) was run on sodium dodecyl sulphatepolyacrylamide gel electrophoresis ( S D S - P A G E ) and reacted with the R39 antiserum in the immunoblotting procedure. No staining was thus obtained in the tau range tool. wt. 68 kDa, pI 5.6: unpublished data). Furthermore, proteins in the tau range were not among those isolated by immunoaffinity chromatography with the neurofilament antibodies from rat spinal cord, nor from the translation products of rat spin~:! cord free polysomes 26'27 (proteins in the tau range were prominent in the fractions applied to the columns).

For the primary incubation (30-50 h, at room temperature) the antiserum R39 was diluted (1:250500) with the same medium as used for pre-incubation. After a 6- to 8-h washing (in 6 - 1 0 changes of PBS-Triton), preparations were incubated overnight with an FITC-conjugated goat anti-rabbit serum (IgG fraction), washed (about 24 h, in 8 - 1 0 changes of PBS-Triton) and mounted on slides with PBS-glycerine. Before immunostaining, preparatk, ns could be conveniently kept at room temperature, in nitrogen atmosphere, at reduced pressure, no change in staining quality being found for storage times as long as 8 weeks. In view of the immunochemical methods used, results reported will refer to neurofilament-like immunoreactivity throughout. Cryostat blocks were also made with fresh unfixed tissues, which were oriented in aluminum foil moulds l°, covered with O.C.T. compound (Miles), frozen (at - 3 0 °C) and stored at low temperature (-70 °C). Sections were collected on poly-L-iysinecoated slides 14, allowed to dry (20-60 rain) and postfixed with acetone (3-5 rain, at room temperature). Neurofilament proteins immunofluorescence was carried out as above, using shorter incubations and washings, and omitting Triton. For both sections and whole-mounts, routine immunocytochemical controls, including substitution of each step in turn with PBS, as well as the use of non-immune rabbit serum, were used to confirm specificity of immunostaining. In all contrc~l rats, a dense network on neurofilament-immanoreactive nerves was revealed in all areas studied, as previously described for the iris 22, colon and ileum 2a5 and other organ systems ~3'21. As r:loted in the above studies, the vast majority of neu-

Figs. 1 and 2. Neurofilament protein immunofluorescence on whole-mounts of jejunal submucosa. Nerve fibres supplying small arteries (v) sh:3w numerous elongated, progressively-tapering dilations in a 2,5-hexanedione-treated rat (1, x80), as opposed to their smooth course it1 a control (2, x70). Figs. 3 and 4. Whole-mount preparations of ureter: note fusiform accumulations of neurofilament-like material along nerve fibres in a 2,5-hexanedione-treated rat (4, x 130), which are absent in a control animal (3, x60). Fig. 5. Whole-mount preparation of iris from an intoxicated rat. Neurofilament-immunoreactive lesion- are very numerous (x 85). Figs. 6-9. Neuromuscular junctions as seen in whole-mount preparations of diaphragm (6,7) and oeso0hageal muscularis (8,9). Accumulations of neurofilament-like material are seen in terminal end-plates of intoxicated rats (6,8), while normal animals (7,9) display only a delicate neurofilament framework at this level (x300, 200, 230, 150, respectively). Figs. 10 and 11. Whole-mount of caecal submucosa: numerous neurofilament-immunoreactive cell bodies can be seen in ganglia, their unreactive nuclei being clearly depicted. In a treated rat (11 x 120), various small, round lesions (open arrows), are associated with one of elongated, fusiform shape (solid arrow). Note the smooth morphology of fibres in a control rat (10, x 100). Fig. 12. Severely damaged oesophageal ganglion from an int,~xicated rat. Various enlarged fibres (bottom), as well as striking round accumulations of neurofilament-like material can be seen (a pale-stained neuronal cell body is arrowed for comparison; muscularis, immunofluorescence, xS0).

385

386 rofilament-immunostained nerve fibres, though running in many cases a complex, whirled course, showed a smooth profile, devoid of major variations in diameter (Figs. 2, 3 and 10). A remarkably different picture was revealed in 2,5-hexanedione-intoxicated rats, autonomic fibres showing large numbers of strongly reactive accumulations of neurofilament-like material in all organs studied. As expected, whole-mounts proved very helpful in correctly defining lesional pattern and distribution. In such preparations, in fact, long tracts of the three-dimensional course of nerve fibres were revealed, thus providing a comprehensive overview comparable to that obtained with teased preparations of peripheral nerve fibres. A consistent pattern was demonstrated in most locations, neurofilament accumulations showing a fusiform, elongated and smoothly tapering shape in perivascular nerves (both supplying large arteries and veins, and running in association with smaller intramural vessels, such as those of the gut: Fig. 1), in the respiratory and urogenital tract (Fig. 4) and in the iris (Fig. 5). It was not uncommon to see many neurofilament accumulations along the course of single nerve fascicles (Fig. 5), thus probably indicating the presence of multiple lesions along the same fibres. In the gastrointestinal tract, densely neurofiiament-immunoreactive lesions of neatly round shape were also revealed (Fig. 11), both in the main plexuses and in nerve fibres supplying the other layers. On the basis of their shape and size, such neurofilamerit accumulations were easily distinguished from the elongated ones described above. It may be of interest to note that the latter fusiform lesions were not as numerous in the gut wall as elsewhere, but were striking around intramural blood vessels (Fig. 1). A complex pattern was revealed in the oesophagus, very large accumulations of neutofilament-like material being found (Fig. 12), while obvious changes were revealed in the terminal innervation of striated muscle. Neuromuscular junctions, in fact, were difficult to see in neurofilament-immunostained preparations of the normal oesophageal muscularis, only a delicate, faintly stained neurofilament component being revealed at this level (Fig. 9). In 2,5-hexanedione-treated rats, however, distinct neurofilament accumulations were present in motor end-plates of this organ (Fig. 8), thus making neuromuscular junc-

tions apparent and easy to recognise. In both normal (Fig. 7) and treated animals (Fig. 6), laminar preparations of diaphragm showed patterns closely comparable to those observed in the oesophagus. Extensive changes were thus demonstrated throughout the autonomic nervous system of rats treated with 2,5-hexanedione, indicating a very widespread action of this chemical on neuronal intermediate filaments. The evolution of such neurofilament accumulations and their impact on autonomic fibre function remain to be ascertained. However, it is conceivable that certain signs and symptoms, as well as functional changes observed in n-hexane, methyl n-butyl ketone, or 2,5-hexanedione intoxication and suggestive of autonomic dysfunction 1'19'24'25, due to visceral nerve damage. It is relevant to note that visceral nerve fibres are largely non-myelinated, in both the enteric nervous system 12 and the intramural innervation of other organs 2s. So far, in fact, most studies on 2,5-hexanedione intoxication have been almost exclusively dedicated to the investigation of myelinated fibres, changes in non-myelinated ones being almost incidentally noted 23'2s. In view of the apparent correlation between giant neurofilament accumulations and Ranvier's nodes in myelinated fibres 23, a detailed investigation of the evolution of structural changes in non-myelinated fibres will be of interest. Alterations in neurofilament-microtubule interactions, possibly due to crossqinking of neurofilament proteins 4a6, may be a useful way to think about neurofilament accumulations 29. Morphological heterogeneity of lesions observed in the present study might thus reflect diversity in proteins associated with the main cytoskeletal elements, i.e. microtubules and neurofilaments. It may be of relevance that dih~rences in neurofilament aggregation have been proposed to exist between certain somatic and parasympathetic motor axons ~7. The gut is provided with a complex innervation, which has long been known as the 'enteric nervous system' and contains many heterogeneous populations of neurons (see ref. 11). Enteric glia, too, have been shown to be heterogeneous 12 and to display immunureactivity for various glial proteins, including glial fibrillary acid (GFA) 8'ls. A unique characteristic, which distinguishes enteric from all other auto~ nomic ganglia, is their capacity of developing a vari-

387 ety of coordinated responses on the exclusive basis of i n t r a m u r a l ' c i r c u i t r y u. Interestingly, nerve projections to the gut originating from extrinsic sources are comparatively few in n u m b e r and enter the wall with blood vessels t2. In the present study, 'fusiform' neurofilament accumulations were conspicuous along gut perivascular nerves, but ' r o u n d ' lesions were not noted in the same location. Such f n d i n g m a y be taken to indicate the intrinsic nature of most fibres displaying the latter type of lesion. As regards the oesophagus, some of the patients suffering of a specific n e u r o m o t o r disturbance of this

organ, d e n o m i n a t e d achalasia, d e m o n s t r a t e d typical Lewy bodies in oesophageal myenteric neurons TM. It is suggestive to note that neurofilament antigens were revealed in L¢wy bodies 29, while a n o t h e r toxic c o m p o u n d severely affecting neurofilaments, i.e. acrylamide, was shown to produce in dogs an oesophageai dysfunction similar to the h u m a n condition 2°.

1 Abdel-Rahman, M.S., Saladin, J.J., Bohman, C.E. and

13 Hacker, G.W., Polak, J.M., Springall, D.R., Ballesta, J., Cadieux, A., Gu, J., Trojanowski, J.Q., Dahl, D. and Marangos, P.J., Antibodies to neurofilament protein and other brain proteins reveal the innervation of peripheral organs, Histochemistry, 82 (1985) 581-593. 14 Huang, W.M., Gibson, S.J., Facer, P., Gu, J. and Polak, J.M., Improved section adhesion for immunocytochemistry using high molecular weight polymers of L-lysine as a slide coating, Histochemistry, 77 (1983) 275-279. 15 Jessen, K.R. and Mirsky, R., Glial cells in the enteric nervous system contain glial fibriilary acidic protein, Nature (Lond.), 286 (1980) 736-737. 16 Lapadula, D.M., Irwin, R.D., Suwita, E. and Abou-Donia, M.B., Cross-linking of neurofilament proteins of rat spinal cord in vivo after administration of 2,5-hexanedione, J. Neurochem., 86 (1986) 1843-1850. 17 Price, R.L., Paggi, P. and Lasek, R.J., The density of neurofilaments differs in the somatic motor and parasympathetic oculomotor axons of chickens. In G.W. Bailey (Ed.), Proceedings of the 44th Annual Meeting of the Electron Microscopy Society of America, San Francisco Press, San Francisco, 1986, pp. 276-277. 18 Qualman, S.J., Haupt, H.M., Yang, 1~'. ar:d !-larailton, S.R., Esophageal Lewy bodies ~,ssociatcd with ganglion cell loss in achalasia. Similarity to Par[ imon's disease, Gastroenterol~gy, 87 (1984) 848-856. 19 Rossi, A., Simonati, A., Rizzuto, N. md Toschi, G., Neurotoxic action of 2,5-hexanedione on the:, autonomic nervous system: ultrastructural and functioLal alterations in the rat sympathetic superior cervical ganglion, Brain Research, 243 (1982) 373-377. 20 SatcheU, P.M. and McLeod, J.G., Megaoesophagus due to acrylamide neuropathy, J. Neurol. Neurosurg. Psychiatry, 44 (1981) 906-913. 21 Schlaepfer, W.W. and Lynch, R.G., Immunofluorescence studies of neurofilaments in the rat and human peripheral an:2 ,.entral nervous system, J. Cell Biol., 74 (1977) 241-250. 22 Seiger, A., Dahi, D., Ayer-LeLievre, C. and Bj6rklund, H., Appearance and distribution of neurofilament immunoreactivity in iris nerves, J. Comp. Neurol., 223 (1984) 457-470. 23 Spencer, P.S. and Schaumburg, H.H., Ultrastructural studies of the dying-back process. III. The evolution of experi-

Couri, D., The effect of 2-hexanone and 2-hexanone metabolites on pupillomotor activity and growth, Am. Ind. Hyg. Assoc. J., 39 (1978) 94-97. 2 Bj6rklund, H., Dahi, D. and Seiger, A., Neurofilament and glial fibrillary acid protein-related immunoreactivity in rodent enteric nervous system, Neuroscience, 12 (1984) 277-287. 3 Cangiano, A., Lutzemberger, L., Rizzuto, N., Simonati, A., Rossi, A. and Toschi, G., Neurotoxic effects of 2,5hexanedione in rats: early morphological and functional changes in nerve fibres and neuromuscular junctions, Neurotoxicology, 2 (1980) 25-32. 4 Carden, M.J., Lee, V.M.-Y. and Schlaepfer, W.W., 2,5Hexanedione neuropathy is associated with the covalent crosslinking of neurofilament proteins, Neurochem. Pathol., 5 (1986) 25-36. 5 Coons, A.H., Leduc, E.H. and Connolly, J., Studies of antibody production. I. A method for ,'he histochemical demonstration of specific antibody and i!s appiicauon to a study of the hyperimmune rabbit, J. Exp. Med., 102 (1955) 49-60. 6 Dahl, D. and Bigi~ami, A., Preparation of antisera to neurofilament protein from chicken brain and human sciatic nerve, J. Comp. Neurol., 176 (1977) 645-657. 7 Dahl, D. and Bignami, A., Intermediate filaments in nervous tissue. In J.W. Shay (Ed.), Cell and Muscle Motility, Vol. 6, Plenum, New York, 1985, pp. 75-96. 8 Ferri, G.-L., Probert, L., Cocchia, D., Mi~,hetti, F., Marangos, P.J. and Polak, J.M., Evidence for the presence of S-100 protein in the gliai component of the enteric nervo~,s system, Nature (Lond.), 297 (1982) 409-410. 9 Ferri, G.-L., Adrian, T.E., Ghatei, M.A., O'Shaughnessy, D.J., Probert, L., Lee, Y.C., Buchan, A.M.J., Polak, J.M. ancl Bloom, S.R., Tissue localisation and ~elative distribution of regulatory peptides in separated latyers from the human bowel, Gastroenterology, 84 (1983) 777-786. 10 Ferri, G.-L., Papadia, C., Cocchia, D. and Polak, J.M., Aluminum foil moulds for cryostat blocks, Stain Technol., 62 (1987) 59-60. 11 Furness, J.B. and Costa, M., fhe Enteri~ Nervous ~,ystem, Churchill Livingstone, Edinburgh, 1987, 2~8 pp. ""~zuao~,,,,," ~,~u,~ G,, lnnervation ...... of "he gastrointe~t~ial tract, Int. Rev. Cytol., 59 (1979) 129-193.

G. Siracusa and A. Riva are t h a n k e d for encouragement and stimulation during the progress of this work, which was partly supported by a National Research Council ( C N R ) R e s e a r c h Contract.

388 mental peripheral giant axonal degeneration, J. Neuropathol. Exp. NeuroL, 36 (1977) 276-299. 24 Spencer, P.S., Co~:ri, D. and Schaumburg, H.H., n. Hexane and methyl n-butyl ketorie. In P.S. Spencer and H.H. Schaumburg (Eds.), Experimental and Clinical Neurotoxicology, Williams and Wilkins, Baltimore, 1980, pp. 456-475. 25 Spencer, P.S., Schaumburg, H.H., Sabri, M.I. and Veronesi, B., The enlarging view of hexacarbon neurotoxicity, CRC Crit. Rev. Toxicol., 7 (1980) 279-356. 26 Strocchi, P., Dahl, D. and Gilbert, J.M., Studies on the bio-

synthesis of intermediate filament proteins in the rat CNS, J. Neurochem., 39 (1982) 1132-1~41. 27 Strocchi, P., Gilbert, J.M., Benowitz, L.I., Dahl, D. and Lewis, E.R., Cellular origin and biosynthesis of rat optic nerve proteins: a two-dimensional gel analysis, J. Neurochem., 43 (1984) 349-357. 28 Williams, P.L. and Warwick, R., Grays Anatomy, Churchill Livingstone, Edinburgh, 1980, 1578 pp. 29 Yen, S.-H., Dickson, D.W., Peterson, C. and Goldman, J.E., Cytoskeletal abnormalities in neuropathology, Prog. Neuropathol., 6 (1986) 63-90.