Brain Research, 173 (1979) 373-377 © Elsevier/North-Holland Biomedical Press
373
On the preparation of brain slices: morphology and cyclic nucleotides
J. GARTHWAITE, P. L. WOODHAMS, M..l. COLLINS and R. BALAZS MRC Developmental Neurobiology Unit, The Institute of Neurology, London I¥C1N 2NS ( U.K.)
(Accepted May 17th, 1979)
Slices of brain tissue incubated in physiological media are used extensively for a variety of neurochemical investigations. In adopting this approach it is of the utmost importance to know the extent to which the tissue in isolation resembles that in vivo. One of the least ambiguous ways of assessing the preservation of cells and their processes is by morphological examination, but such studies on incubated brain slices have, surprisingly, been scarce and have been restricted to electron microscopy of selected regions at high magnification2,4,10. In this paper we describe the ultrastructural appearance of slices of adult rat cerebellum prepared by the two most frequently used techniques: mechanical chopping and manual slicing. The results show that the two methods yielded tissue slices which, after incubation, differed substantially in their degree of preservation. We also compare a functional property of cerebellar cells in the two types of slices, namely their ability to generate cyclic G M P in response to excitatory amino acidsl,3,L Methods similar to those described by Mcllwain 6 were used. Cerebella were removed from rats (Porton strain) and cut sagittally into two through the centre of the vermis. Surface (and sometimes second) slices from each half were prepared at ambient temperature (20-23 °C) using a glass guide (recessed, 0.38 mm) and bow cutter. The cutting blade was a Gillette Techmatic ribbon cut to a width of about 1 mm and was moistened with Krebs solution before use. Alternatively, slices of comparable size (about 1 mg protein per slice) were prepared by chopping the tissue sagittally at room temperature into 0.38 mm slices using a Mcllwain tissue chopper 6. The slices were transferred to 50 ml conical flasks containing 20 ml Krebs solution (20-25 °C) pre-equilibrated with a mixture of 95 ~o 02-5 ~ COs (pH 7.4). Flasks were sealed and incubated at 37 °C in a shaking water bath whose speed was adjusted to keep the slices moving gently. The surface of the Krebs solution 3 was continuously gassed. The medium was usually replaced 2-3 times during the 2 h incubation period. Slices were removed using a small spatula, fixed for 30 min at room temperature in 4 ~ paraformaldehyde and 2 . 5 ~ glutaraldehyde in 0.1 M phosphate buffer, rinsed in veronal acetate buffer and osmicated for 1 h in 1 ~o OsO4 in veronal acetate buffer. After dehydration through ethanol and embedding in Spurr's resin, 0.5-1.0 # m sections
374
Fig. 1. Appearance at low magnification ( x 400) of (A) hand-cut (surface) and (B) chopped,cerebellar slices incubated for 2 h in Krebs solution, m: molecular layer, p : Purkinje cell layer, g: internal granular layer.
were cut and stained with toluidine blue. Ultrathin sections were contrasted with uranyl acetate and lead citrate and viewed in a J E O L 100S electron microscope. At a low survey magnification (Fig. 1A, B) and in more detail in the electron micrographs (Fig. 2A, B) it is evident that the morphological preservation of handsliced cerebella was much superior to that of chopped material. In the hand-cut tissue, apart from a damaged region extending up to 40 ffm from the cut edge of the slice (not shown), there was no evidence of pyknosis and the overall appearance was grossly similar to that obtained after perfusion fixation of the cerebellum in situ. Occasional vacuoles were seen in the internal granular layer but the cells and glomeruli appeared normal and there was no evidence of shrinkage or swelling of the Purkinje cells. By contrast, nearly all (1> 90 ~ ) of the total cell number in the chopped cerebellar slices appeared pyknotic and much vacuolation was seen in all regions. In the example shown in Fig. 1B, the Purkinje cells are still discernable but their nuclei are swollen and the cytoplasm is vacuolated. Often this was so marked that the cells seemed completely disrupted. Under the electron microscope it is seen that the swelling of the chopped tissue was both intra- and extra-cellular. Glial cells were unidentifiable and much of the internal granular layer (Fig. 2B) consisted simply of
375
Fig. 2. Electron micrographs showing regions of the internal granular layer of (A) hand-cut (B) chopped cerebellar slices, taken from regions > 80/~mfrom the cut edges of the slices. Magnification × 12,000. Arrows indicate mossy fibre terminals. pyknotic nuclei, empty spaces and membrane fragments. The elements showing the best preservation were synaptic structures, and mossy fibre terminals could usually be identified, although some other components of the glomeruli showed extensive swelling, giving the appearance of empty profiles containing one or two swollen mitochondria. The granule cell in Fig. 2B is also abnormal, with a pyknotic nucleus, swollen nuclear envelope, cytoplasm and mitochondria, and holes and vacuolation in the plasmalemma. The ultrastructural preservation of the molecular layer was likewise poor. In contrast, detailed examination of hand-sliced material confirmed the absence of degenerative changes in the internal granular layer (Fig. 2A) and all the usual subcellular organelles (nuclei, mitochondria, synaptic vesicles, microtubules, etc.) appeared normal; a degree of glial swelling was encountered in some regions, but this was not marked. Similarly, electron micrographs of Purkinje cells and the molecular layer showed their morphology to be largely normal (not shown). One of the widespread uses of brain slices is for measurement of changes in the levels of cyclic nucleotides in response to various transmittersT; chopped tissue is mostly used. When excitatory amino acids such as glutamate 5 or kainic acid 1 are injected into rat cerebellum in vivo, cyclic GMP levels markedly increase. A
376 TABLE I
Cyclic GMP generation by chopped and hand-cut slices in response to kainic acid and glutamate The slices were pre-incubated for 2 h and the amino acids (or saline) were then added from concentrated stock solutions (pH 7.4) ; 5 to 10 min later, slices were inactivated and cyclic G M P levels measured as described previouslyL Data represent mean d: S.E.M. from 4-8 experiments. Result in brackets is from second slices.
Additions
Cyclic GMP (pmol/mg protein) Chopped slices
Saline Kainate (300/~M) Saline Glutamate (10 mM)
3.4 4- 0.4 8.8 ± 0.5 4.0 4- 0.4 7.8 4- 0.2
Hand-cut slices 6.6 4- 0.6 258 4- 10 5.4 4- 0.6 78 4- 5 (67 4- 7)
comparison of the cyclic G M P response to these amino acids in the two types of slice reflected the morphological picture (Table I). On exposure of hand-cut surface slices to kainic acid (300 ktM) cyclic G M P levels rose from 6.6 to 258 pmol/mg protein. In chopped slices on the other hand the corresponding change was from 3.4 to a mere 8.8 pmol/mg protein. Analogous disparities were found using glutamate. The relative enrichment of cortex and depletion of white matter in the surface slices did not seem to be a contributory factor since when surface slices were chopped, the cyclic G M P response to glutamate was as small (2-fold increase) as in whole chopped cerebellum. In addition, the cause of the poor preservation in the chopped slices did not appear to be due solely to the presence of two cut surfaces, since with second hand-cut slices, apart from the additional damaged edge (up to 40 #m thick), the morphology was similar to that of surface slices, as was the cyclic G M P response to glutamate (Table I) although the latter was more variable than in the surface slices. The small changes observed with glutamate and kainate in the chopped slices were similar to those reported by othersS, 8,9 for chopped rat cerebellum. We have examined several variations of the chopping technique, such as prior immersion of the cerebellum in ice-cold Krebs solution, cutting at different thicknesses (0.5-0.25 mm), chopping in two directions (sagittally and coronally), and in a different plane (horizontally), but none resulted in a significantly improved morphology, nor in larger evoked changes in cyclic G M P levels. These results clearly indicate the importance of supplementing biochemical measurements with morphology. The extent to which the findings reported here apply to other regions of the adult brain, and indeed to other species, remains to be determined but in the absence of contrary evidence, slices cut by hand, though more laborious to prepare in quantity, are more likely to yield information of physiological relevance than those prepared by chopping. We would add here that, in contrast to the adult, slices of immature cerebellum (up to 14 days old) prepared by chopping show good morphological preservation (J. Garthwaite et al, in preparation).
377 1 Biggio, G., Corda, M. G., Casu, M., Salis, M. and Gessa, G. P., Disappearance of cerebellar cyclic GMP induced by kainic acid, Brain Research, 154 (1978) 203-208. 2 Cohen, M. M. and Hartmann, J. F., Biochemical and ultrastructural correlates of cerebral cortex slices metabolizing in vitro. In M. M. Cohen and R. S. Snider (Eds.), Morphological andBiochemical Correlates of Neural Activity, Harper and Row, New York, 1964, pp. 57-74. 3 Garthwaite, J'. and Balazs, R., Supersensitivity to the cyclic GMP response to glutamate during cerebellar maturation, Nature (Lond.), 275 (1978) 328-329. 4 Ibata, Y., Piccoli, F., Pappas, G. D. and Lajtha, A., An electron microscopic and biochemical study on the effect of cyanide and low Na ÷ on rat brain slices, Brain Research, 30(1971) 137-158. 5 Mao, C. C., Guidotti, A. and Costa, E., The regulation of cyclic guanosine monophosphate in rat cerebellum: possible involvement of putative amino acid neurotransmitters, Brain Research, 79 (1974) 510-514. 6 Mcllwain, H., Preparing neural tissues for metabolic study in isolation. In H. Mcllwain (Ed.), Practical Neurochemistry, Churchill Livingston, London, 1975, pp. 105-132. 7 Nathanson, .1. A., Cyclic nucleotides and nervous system function, Physiol. Rev., 51 (1977) 157-256. 8 Schmidt, M. J., Ryan, 2".J. and Molloy, B. B., Effects of kainic acid, a cyclic analogue of glutamic acid, on cyclic nucleotide accumulation in slices of rat cerebellum, Brain Research, 112 (1976) 113-126. 9 Schmidt, M. J., Thornberry, J. F. and Molloy, B. B., Effects of kainate and other glutamate analogues on cyclic nucleotide accumulation in slices of rat cerebellum, Brain Research, 121 (1977) 182-189. 10 Wanko, T. and Tower, D. B., Combined morphological and biochemical studies of incubated slices of cerebral cortex. In M. M. Cohen and R. S. Snider (Eds.), Morphological and Biochemical Correlates of Neural Activity, Harper and Row, New York, 1964, pp. 75-97.
373
On the preparation of brain slices: morphology and cyclic nucleotides
J. GARTHWAITE, P. L. WOODHAMS, M..l. COLLINS and R. BALAZS MRC Developmental Neurobiology Unit, The Institute of Neurology, London I¥C1N 2NS ( U.K.)
(Accepted May 17th, 1979)
Slices of brain tissue incubated in physiological media are used extensively for a variety of neurochemical investigations. In adopting this approach it is of the utmost importance to know the extent to which the tissue in isolation resembles that in vivo. One of the least ambiguous ways of assessing the preservation of cells and their processes is by morphological examination, but such studies on incubated brain slices have, surprisingly, been scarce and have been restricted to electron microscopy of selected regions at high magnification2,4,10. In this paper we describe the ultrastructural appearance of slices of adult rat cerebellum prepared by the two most frequently used techniques: mechanical chopping and manual slicing. The results show that the two methods yielded tissue slices which, after incubation, differed substantially in their degree of preservation. We also compare a functional property of cerebellar cells in the two types of slices, namely their ability to generate cyclic G M P in response to excitatory amino acidsl,3,L Methods similar to those described by Mcllwain 6 were used. Cerebella were removed from rats (Porton strain) and cut sagittally into two through the centre of the vermis. Surface (and sometimes second) slices from each half were prepared at ambient temperature (20-23 °C) using a glass guide (recessed, 0.38 mm) and bow cutter. The cutting blade was a Gillette Techmatic ribbon cut to a width of about 1 mm and was moistened with Krebs solution before use. Alternatively, slices of comparable size (about 1 mg protein per slice) were prepared by chopping the tissue sagittally at room temperature into 0.38 mm slices using a Mcllwain tissue chopper 6. The slices were transferred to 50 ml conical flasks containing 20 ml Krebs solution (20-25 °C) pre-equilibrated with a mixture of 95 ~o 02-5 ~ COs (pH 7.4). Flasks were sealed and incubated at 37 °C in a shaking water bath whose speed was adjusted to keep the slices moving gently. The surface of the Krebs solution 3 was continuously gassed. The medium was usually replaced 2-3 times during the 2 h incubation period. Slices were removed using a small spatula, fixed for 30 min at room temperature in 4 ~ paraformaldehyde and 2 . 5 ~ glutaraldehyde in 0.1 M phosphate buffer, rinsed in veronal acetate buffer and osmicated for 1 h in 1 ~o OsO4 in veronal acetate buffer. After dehydration through ethanol and embedding in Spurr's resin, 0.5-1.0 # m sections
374
Fig. 1. Appearance at low magnification ( x 400) of (A) hand-cut (surface) and (B) chopped,cerebellar slices incubated for 2 h in Krebs solution, m: molecular layer, p : Purkinje cell layer, g: internal granular layer.
were cut and stained with toluidine blue. Ultrathin sections were contrasted with uranyl acetate and lead citrate and viewed in a J E O L 100S electron microscope. At a low survey magnification (Fig. 1A, B) and in more detail in the electron micrographs (Fig. 2A, B) it is evident that the morphological preservation of handsliced cerebella was much superior to that of chopped material. In the hand-cut tissue, apart from a damaged region extending up to 40 ffm from the cut edge of the slice (not shown), there was no evidence of pyknosis and the overall appearance was grossly similar to that obtained after perfusion fixation of the cerebellum in situ. Occasional vacuoles were seen in the internal granular layer but the cells and glomeruli appeared normal and there was no evidence of shrinkage or swelling of the Purkinje cells. By contrast, nearly all (1> 90 ~ ) of the total cell number in the chopped cerebellar slices appeared pyknotic and much vacuolation was seen in all regions. In the example shown in Fig. 1B, the Purkinje cells are still discernable but their nuclei are swollen and the cytoplasm is vacuolated. Often this was so marked that the cells seemed completely disrupted. Under the electron microscope it is seen that the swelling of the chopped tissue was both intra- and extra-cellular. Glial cells were unidentifiable and much of the internal granular layer (Fig. 2B) consisted simply of
375
Fig. 2. Electron micrographs showing regions of the internal granular layer of (A) hand-cut (B) chopped cerebellar slices, taken from regions > 80/~mfrom the cut edges of the slices. Magnification × 12,000. Arrows indicate mossy fibre terminals. pyknotic nuclei, empty spaces and membrane fragments. The elements showing the best preservation were synaptic structures, and mossy fibre terminals could usually be identified, although some other components of the glomeruli showed extensive swelling, giving the appearance of empty profiles containing one or two swollen mitochondria. The granule cell in Fig. 2B is also abnormal, with a pyknotic nucleus, swollen nuclear envelope, cytoplasm and mitochondria, and holes and vacuolation in the plasmalemma. The ultrastructural preservation of the molecular layer was likewise poor. In contrast, detailed examination of hand-sliced material confirmed the absence of degenerative changes in the internal granular layer (Fig. 2A) and all the usual subcellular organelles (nuclei, mitochondria, synaptic vesicles, microtubules, etc.) appeared normal; a degree of glial swelling was encountered in some regions, but this was not marked. Similarly, electron micrographs of Purkinje cells and the molecular layer showed their morphology to be largely normal (not shown). One of the widespread uses of brain slices is for measurement of changes in the levels of cyclic nucleotides in response to various transmittersT; chopped tissue is mostly used. When excitatory amino acids such as glutamate 5 or kainic acid 1 are injected into rat cerebellum in vivo, cyclic GMP levels markedly increase. A
376 TABLE I
Cyclic GMP generation by chopped and hand-cut slices in response to kainic acid and glutamate The slices were pre-incubated for 2 h and the amino acids (or saline) were then added from concentrated stock solutions (pH 7.4) ; 5 to 10 min later, slices were inactivated and cyclic G M P levels measured as described previouslyL Data represent mean d: S.E.M. from 4-8 experiments. Result in brackets is from second slices.
Additions
Cyclic GMP (pmol/mg protein) Chopped slices
Saline Kainate (300/~M) Saline Glutamate (10 mM)
3.4 4- 0.4 8.8 ± 0.5 4.0 4- 0.4 7.8 4- 0.2
Hand-cut slices 6.6 4- 0.6 258 4- 10 5.4 4- 0.6 78 4- 5 (67 4- 7)
comparison of the cyclic G M P response to these amino acids in the two types of slice reflected the morphological picture (Table I). On exposure of hand-cut surface slices to kainic acid (300 ktM) cyclic G M P levels rose from 6.6 to 258 pmol/mg protein. In chopped slices on the other hand the corresponding change was from 3.4 to a mere 8.8 pmol/mg protein. Analogous disparities were found using glutamate. The relative enrichment of cortex and depletion of white matter in the surface slices did not seem to be a contributory factor since when surface slices were chopped, the cyclic G M P response to glutamate was as small (2-fold increase) as in whole chopped cerebellum. In addition, the cause of the poor preservation in the chopped slices did not appear to be due solely to the presence of two cut surfaces, since with second hand-cut slices, apart from the additional damaged edge (up to 40 #m thick), the morphology was similar to that of surface slices, as was the cyclic G M P response to glutamate (Table I) although the latter was more variable than in the surface slices. The small changes observed with glutamate and kainate in the chopped slices were similar to those reported by othersS, 8,9 for chopped rat cerebellum. We have examined several variations of the chopping technique, such as prior immersion of the cerebellum in ice-cold Krebs solution, cutting at different thicknesses (0.5-0.25 mm), chopping in two directions (sagittally and coronally), and in a different plane (horizontally), but none resulted in a significantly improved morphology, nor in larger evoked changes in cyclic G M P levels. These results clearly indicate the importance of supplementing biochemical measurements with morphology. The extent to which the findings reported here apply to other regions of the adult brain, and indeed to other species, remains to be determined but in the absence of contrary evidence, slices cut by hand, though more laborious to prepare in quantity, are more likely to yield information of physiological relevance than those prepared by chopping. We would add here that, in contrast to the adult, slices of immature cerebellum (up to 14 days old) prepared by chopping show good morphological preservation (J. Garthwaite et al, in preparation).
377 1 Biggio, G., Corda, M. G., Casu, M., Salis, M. and Gessa, G. P., Disappearance of cerebellar cyclic GMP induced by kainic acid, Brain Research, 154 (1978) 203-208. 2 Cohen, M. M. and Hartmann, J. F., Biochemical and ultrastructural correlates of cerebral cortex slices metabolizing in vitro. In M. M. Cohen and R. S. Snider (Eds.), Morphological andBiochemical Correlates of Neural Activity, Harper and Row, New York, 1964, pp. 57-74. 3 Garthwaite, J'. and Balazs, R., Supersensitivity to the cyclic GMP response to glutamate during cerebellar maturation, Nature (Lond.), 275 (1978) 328-329. 4 Ibata, Y., Piccoli, F., Pappas, G. D. and Lajtha, A., An electron microscopic and biochemical study on the effect of cyanide and low Na ÷ on rat brain slices, Brain Research, 30(1971) 137-158. 5 Mao, C. C., Guidotti, A. and Costa, E., The regulation of cyclic guanosine monophosphate in rat cerebellum: possible involvement of putative amino acid neurotransmitters, Brain Research, 79 (1974) 510-514. 6 Mcllwain, H., Preparing neural tissues for metabolic study in isolation. In H. Mcllwain (Ed.), Practical Neurochemistry, Churchill Livingston, London, 1975, pp. 105-132. 7 Nathanson, .1. A., Cyclic nucleotides and nervous system function, Physiol. Rev., 51 (1977) 157-256. 8 Schmidt, M. J., Ryan, 2".J. and Molloy, B. B., Effects of kainic acid, a cyclic analogue of glutamic acid, on cyclic nucleotide accumulation in slices of rat cerebellum, Brain Research, 112 (1976) 113-126. 9 Schmidt, M. J., Thornberry, J. F. and Molloy, B. B., Effects of kainate and other glutamate analogues on cyclic nucleotide accumulation in slices of rat cerebellum, Brain Research, 121 (1977) 182-189. 10 Wanko, T. and Tower, D. B., Combined morphological and biochemical studies of incubated slices of cerebral cortex. In M. M. Cohen and R. S. Snider (Eds.), Morphological and Biochemical Correlates of Neural Activity, Harper and Row, New York, 1964, pp. 75-97.
