The DNA Content of Purkinje Cells in Mammals D. M. A. MANN,' P. 0. YATES AND C. M. BARTON Department of Pathology, The University, Manchester, MI3 9PT, England
ABSTRACT Nerve cells have generally been assumed to have a diploid DNA content, typical of non-dividing somatic cells. However several reports have suggested that certain nerve cell types, notably Purkinje cells of the cerebellum, are polyploid. Other studies have contradicted these findings, stating Purkinje cells to be diploid. In this paper we reinvestigate the DNA status of Purkinje cells, in a variety of mammalian species. Cell DNA content is measured on tissue smears by Feulgen microspectrophotometry. Results show that for all species examined by us, Purkinje cells have, without exception, a DNA content comparable to that of somatic cells. A critical appraisal of the techniques used in those studies claiming a tetraploid DNA content for Purkinje cells leads us to believe our findings to be correct. Nerve cells are typical non-dividing somatic cells and as such have generally been assumed to have a diploid DNA content. Most of the early studies and, indeed, some recent reports have suggested t h a t in various species t h e DNA content of certain large nerve cells, notably Purkinje cells, exceeds diploid and may amount to tetraploid or even octoploid levels. The species examined were r a t (Sandritter et al., '67; Lentz and Lapham, '69, '70; Novakova et al., '70; Bohn and Mitchell, '761, cat (Herman and Lapham, '68, '691, rabbit (Kusch and Yarygin, '65) and human (Lapham, '65, '68). However, later studies have questioned the validity of these findings. Morselt et al. ('72) demonstrated rat Purkinje cell nuclei to have DNA and histone content appropriate to that of a diploid level. Furthermore, Fujita et al. ('74) found no evidence to suggest DNA synthesis in rat Purkinje cells and concluded that such neurones remained diploid throughout postnatal life. Cohen et al. ('73) measured by cytophotometric means, a greater than diploid DNA content in mouse Purkinje cells, but were unable to confirm this by biochemical assay. These authors stated t h a t no evidence could therefore be found for a tetraploid DNA content. We also have been unable to corroborate previous reports of tetraploidy in human Purkinje cells (Mann and Yates, "73). Autoradiographic studies (Mares e t al., '73; Manuelidis and Manuelidis, '74) have also failed to support a finding of tetraploidy in human Purkinje cells. J. COMP. NEUR. (1978) 180.' 345-348.
In this paper therefore the DNA status of Purkinje cells of a variety of species is re-examined. MATERIALS AND METHODS
The brains of adult animals were removed as quickly as possible after death. Smears made of the fresh cerebellar tissue, by squashing between glass slides and spreading by drawing one across the other, were fixed in 5% acetic ethanol, and subsequently stained for DNA by the Feulgen technique using a 6minute hydrolysis in 1N HC1 a t 60°C. Microscopic determinations of DNA content were carried out using a Barr & Stroud Integrating Microdensitometer, a t a wavelength of 550 +- 10 nm using 100 x objective. The smallest masking hole was selected such that a minimal amount of cytoplasm was included with the nucleus in the measured field. Every DNA measurement was the mean of three recordings and a minimum of 30 nerve cell nuclei of each type were measured. Estimations of non-specific light loss were made through an area of unstained cytoplasm adjacent to the measured nucleus. DNA content was expressed as the difference between the measured nuclear and non-specific cytoplasmic absorbance for the corresponding cell. These values of DNA content represent nuclear absorption of light per cell and are expressed in arbitrary units of summated light absorption.
' All correspondence to: Dr. D. M. A. Mann, Department of Pathology, The
University, Manchester M13 9PT, England.
345
346
D. M. A. MA",
P. 0. YATES AND C. M. BARTON
RESULTS
Measurements of mean DNA content of Purkinje and granule cell nuclei for each species examined are shown in table 1. Variation in mean DNA content within and between species reflect differences in staining intensity arising from fluctuations in t h e extent of hydrolysis of smears stained on separate occasions. Results clearly demonstrate t h a t in all t h e species examined by us in no case did Purkinje cells have a mean DNA content significantly different (p > 0.05) to t h a t of t h e corresponding granular neurone, which has been previously shown to be diploid (Mann and Yates, '73). DISCUSSION
The results of this study show t h a t for all those species examined by us, t h e adult Purkinje cell has a DNA content comparable to t h a t of normal diploid somatic cells. These findings a r e in accordance with those of Cohen e t al. ('731, Morselt e t al. ('72) and Fujita e t al. ('741, but fail to confirm those reports of Purkinje cell polyploidy in rat (Sandritter et al., '67; Lentz and Lapham, '69, '70; Novakova et al., '70; Bohn and Mitchell, '76), cat (Herm a n a n d Lapham, '68, '69) a n d h u m a n (Lapham, '65, '68). I n a previous paper (Mann and Yates, '73) a thorough investigation of t h e causes and magnitude of non-specific light loss, in thick tissue sections was carried out. From this study we concluded t h a t unless appropriate compensation for non-specific light loss is made, gross inaccuracies of measurement of DNA content by transmission densitometry a r e likely to result. By the use of tissue smears, wherever
practical for this type of study, the cell cytoplasm is thinly dispersed over t h e slide surface and non-specific light loss is minimized. In t h e previously mentioned reports, either paraffin sections of 16-25 p m thickness (Lentz and Lapham, '69, '70; Herman and Lapham, '69; Bohn and Mitchell, '76) or crushed cryostat sections of 80-100 p m (Novakova et al., '70) were used. I n both these types of preparation non-specific light loss through cytoplasm adjacent to the nucleus of neurones must be considerable, yet was either stated as negligible or was not reported a s being necessary for compensation. These studies yielded values of ratio of Purkinje to glial cell DNA ranging from 1.6 to 2.1, which were considered to be a n indication of a tetraploid state. Although mouse Purkinje cells were measured as diploid by a biochemical method (Cohen e t al., '73), a parallel cytophotometric study produced a ratio of DNA contents in Purkinje to glial cell nuclei of 1.40. This discrepancy, these authors explained, resulted from a n unreliability of t h e Feulgen stain due to interference by other nuclear macromolecules, or arose because of the differences in DNA distribution in each cell type. However, in their methods, fresh tissue was squashed without smearing and i t is not stated whether any correction was made for this reduced, but not insubstantial, non-specific component. I t is possible therefore t h a t a DNA ratio of 1.40 represents t h e inclusion of the more optically dense P u r k i n j e cell cytoplasm w i t h t h e stained DNA in t h e nuclear measurement. We have stressed t h e importance of making appropriate corrections for non-specific light loss, whether in tissue, smears, squashes or
TABLE 1
Mean DNA contents (tS.D.)for Purkinje and granule cells of those species shown DNA content
DNA content Species
Mouse
Rat
Guinea pig
Rabbit
Purkinje cell
Granule cell
10.72 0.4 10.7%0.5 10.62 0.5 10.8% 0.5 10.82 0.3 10.7% 0.4 10.82 0.5 12.8%0.8 13.62 1.2 13.3%0.7 13.220.7 13.520.6
10.52 0.4 10.4%0.6 10.4%0.5 10.32 0.5 10.5k0.5 10.42 0.5 10.4% 0.4 12.42 0.6 13.22 0.4 13.52 0.5 13.52 0.6 13.62 0.6
Species
Cat Dog Pig cow Chimpanzee Human
Purklnje cell
Granule cell
10.62 0.6 9.92 0.6 10.02 0.6 9.9k 0.6 10.8'0.7 10.62 0.5 13.22 0.8 8.32 0.6 11.52 0.8 10.72 0.5 11.8% 0.8 8.52 0.4
10.120.6 10.02 0.7 10.12 0.6 10.0%0.7 11.62 0.4 10.7+ 0.5 13.1* 0.7 9.0%0.4 11.520.4 10.32 0.4 12.02 0.5 8.32 0.3
PURKINJE CELL DNA CONTENT
sections, if reliable measurements of DNA content are to be made. This point of view has also been expressed by Morselt e t al. ('72) and Fujita et al. ('74). In every study where a thorough consideration of non-specific factors has been made (Mann and Yates, '73; Morselt et al., '72; Fujita et al., '74), Purkinje and other large nerve cells have been shown to have a diploid DNA content. It has only been in those reports where this problem has been apparently inadequately treated that findings of a tetraploid DNA content have been reported. We conclude therefore that in the variety of species examined by ourselves, Purkinje and other nerve cells are, without exception, diploid. Although these findings contradict certain previous reports, a critical appraisal of the techniques used in these studies leads us to believe that our results are correct. ACKNOWLEDGMENTS
We thank the National Fund for Research into Crippling Diseases who generously supported this work. LITERATURE CITED Bohn, R. C.. and R. B. Mitchell 1976 Cytophotometric identification of tetraploid purkinje cells in young and aged rats. J. Neurobiol., 7: 255-258. Cohen, J., V. Mares and 2. Lodin 1973 DNA content of purified preparations of mouse Purkinje neurones isolated hy a velocity sedimentation technique. J. Neurochem., 20; 651-657. Fujita, S., T. Hattori, M.Fukudu and T. Kitamura 1974 DNA contents in Purkinje cells and inner granule neu-
347
rones in the developing r a t cerebellum. Dev. Growth 81 Diff., 16: 205-211. Herman, C. J., and L. W. Lapham 1968 DNA content of neurones in the cat hippocampus. Science, 160: 537. 1969 Neuronal polyploidy and nuclear volume in cat central nervous system. Brain Res., 15: 35-48. Kusch, A. A., and V. N. Yarygin 1965 Polyploidy of monoand binucleate neurones in upper cervical ganglion of rabbit. Tsitologiia, 7: 228-233. Lapham, L. W. 1965 The tetraploid DNA content of normal human Purkinje cells and its development during the perinatal period. A quantitative cytochemical study. Excerpta. Med (Amst.) Cong. Ser., 100: 445-449. 1968 Tetraploid DNA content of Purkinje neurones of human cerebellar cortex. Science, 159: 310-312. Lentz, R. D., and L. W. Lapham 1969 A quantitative cytochemical study of the DNA content of neurones of rat cerebellar cortex. J. Neurochem., 16: 379-384. 1970 Postnatal development of tetraploid DNA content in r a t Purkinje cells. A quantitative cytochemical study. J. Neuropath. exp. Neurol., 29: 43-56. Mann, D. M. A,, and P. 0. Yates 1973 Polyploidy in the human nervous system. Part I. The DNA content of neurones and glia of the cerebellum. J. Neurol. Sci., 18: 183-196. Manuelidis, L., and E. E. Manuelidis 1974 On the DNA content of cerebellar Purkinje cells in vivo and in vitro. Exp. Neurol., 43: 192-206. Mares, V., 2. Lodin and S. Sacha 1973 A cytochemical and autoradiographic study of nuclear DNA i n mouse purkinje cells. Brain Res., 53: 273-289. Morselt, A. F. W., D. J. Braakman and J. James 1972 Feulgen-DNA and fast green histone estimations in individual cell nuclei of the cerebellum of young and old rats. Acta. Histochem (Jena), 43: 281-286. Novakova, V., W. Sandritter and G. Schlueter 1970 DNA content of neurones in r a t central nervous system. Exptl. Cell. Res., 60: 454-456. Sandritter, W., V. Novakova, J. Pilny and G. Kiefer 1967 Cytophotometrische Messungen des Nucleinsaure und Proteingehaltes von Ganglienzellen der Ratte wahrend der Postnatalen Entwicklungund im Alter. 2. Zellforsch., 80: 145-152.
ABSTRACT Nerve cells have generally been assumed to have a diploid DNA content, typical of non-dividing somatic cells. However several reports have suggested that certain nerve cell types, notably Purkinje cells of the cerebellum, are polyploid. Other studies have contradicted these findings, stating Purkinje cells to be diploid. In this paper we reinvestigate the DNA status of Purkinje cells, in a variety of mammalian species. Cell DNA content is measured on tissue smears by Feulgen microspectrophotometry. Results show that for all species examined by us, Purkinje cells have, without exception, a DNA content comparable to that of somatic cells. A critical appraisal of the techniques used in those studies claiming a tetraploid DNA content for Purkinje cells leads us to believe our findings to be correct. Nerve cells are typical non-dividing somatic cells and as such have generally been assumed to have a diploid DNA content. Most of the early studies and, indeed, some recent reports have suggested t h a t in various species t h e DNA content of certain large nerve cells, notably Purkinje cells, exceeds diploid and may amount to tetraploid or even octoploid levels. The species examined were r a t (Sandritter et al., '67; Lentz and Lapham, '69, '70; Novakova et al., '70; Bohn and Mitchell, '761, cat (Herman and Lapham, '68, '691, rabbit (Kusch and Yarygin, '65) and human (Lapham, '65, '68). However, later studies have questioned the validity of these findings. Morselt et al. ('72) demonstrated rat Purkinje cell nuclei to have DNA and histone content appropriate to that of a diploid level. Furthermore, Fujita et al. ('74) found no evidence to suggest DNA synthesis in rat Purkinje cells and concluded that such neurones remained diploid throughout postnatal life. Cohen et al. ('73) measured by cytophotometric means, a greater than diploid DNA content in mouse Purkinje cells, but were unable to confirm this by biochemical assay. These authors stated t h a t no evidence could therefore be found for a tetraploid DNA content. We also have been unable to corroborate previous reports of tetraploidy in human Purkinje cells (Mann and Yates, "73). Autoradiographic studies (Mares e t al., '73; Manuelidis and Manuelidis, '74) have also failed to support a finding of tetraploidy in human Purkinje cells. J. COMP. NEUR. (1978) 180.' 345-348.
In this paper therefore the DNA status of Purkinje cells of a variety of species is re-examined. MATERIALS AND METHODS
The brains of adult animals were removed as quickly as possible after death. Smears made of the fresh cerebellar tissue, by squashing between glass slides and spreading by drawing one across the other, were fixed in 5% acetic ethanol, and subsequently stained for DNA by the Feulgen technique using a 6minute hydrolysis in 1N HC1 a t 60°C. Microscopic determinations of DNA content were carried out using a Barr & Stroud Integrating Microdensitometer, a t a wavelength of 550 +- 10 nm using 100 x objective. The smallest masking hole was selected such that a minimal amount of cytoplasm was included with the nucleus in the measured field. Every DNA measurement was the mean of three recordings and a minimum of 30 nerve cell nuclei of each type were measured. Estimations of non-specific light loss were made through an area of unstained cytoplasm adjacent to the measured nucleus. DNA content was expressed as the difference between the measured nuclear and non-specific cytoplasmic absorbance for the corresponding cell. These values of DNA content represent nuclear absorption of light per cell and are expressed in arbitrary units of summated light absorption.
' All correspondence to: Dr. D. M. A. Mann, Department of Pathology, The
University, Manchester M13 9PT, England.
345
346
D. M. A. MA",
P. 0. YATES AND C. M. BARTON
RESULTS
Measurements of mean DNA content of Purkinje and granule cell nuclei for each species examined are shown in table 1. Variation in mean DNA content within and between species reflect differences in staining intensity arising from fluctuations in t h e extent of hydrolysis of smears stained on separate occasions. Results clearly demonstrate t h a t in all t h e species examined by us in no case did Purkinje cells have a mean DNA content significantly different (p > 0.05) to t h a t of t h e corresponding granular neurone, which has been previously shown to be diploid (Mann and Yates, '73). DISCUSSION
The results of this study show t h a t for all those species examined by us, t h e adult Purkinje cell has a DNA content comparable to t h a t of normal diploid somatic cells. These findings a r e in accordance with those of Cohen e t al. ('731, Morselt e t al. ('72) and Fujita e t al. ('741, but fail to confirm those reports of Purkinje cell polyploidy in rat (Sandritter et al., '67; Lentz and Lapham, '69, '70; Novakova et al., '70; Bohn and Mitchell, '76), cat (Herm a n a n d Lapham, '68, '69) a n d h u m a n (Lapham, '65, '68). I n a previous paper (Mann and Yates, '73) a thorough investigation of t h e causes and magnitude of non-specific light loss, in thick tissue sections was carried out. From this study we concluded t h a t unless appropriate compensation for non-specific light loss is made, gross inaccuracies of measurement of DNA content by transmission densitometry a r e likely to result. By the use of tissue smears, wherever
practical for this type of study, the cell cytoplasm is thinly dispersed over t h e slide surface and non-specific light loss is minimized. In t h e previously mentioned reports, either paraffin sections of 16-25 p m thickness (Lentz and Lapham, '69, '70; Herman and Lapham, '69; Bohn and Mitchell, '76) or crushed cryostat sections of 80-100 p m (Novakova et al., '70) were used. I n both these types of preparation non-specific light loss through cytoplasm adjacent to the nucleus of neurones must be considerable, yet was either stated as negligible or was not reported a s being necessary for compensation. These studies yielded values of ratio of Purkinje to glial cell DNA ranging from 1.6 to 2.1, which were considered to be a n indication of a tetraploid state. Although mouse Purkinje cells were measured as diploid by a biochemical method (Cohen e t al., '73), a parallel cytophotometric study produced a ratio of DNA contents in Purkinje to glial cell nuclei of 1.40. This discrepancy, these authors explained, resulted from a n unreliability of t h e Feulgen stain due to interference by other nuclear macromolecules, or arose because of the differences in DNA distribution in each cell type. However, in their methods, fresh tissue was squashed without smearing and i t is not stated whether any correction was made for this reduced, but not insubstantial, non-specific component. I t is possible therefore t h a t a DNA ratio of 1.40 represents t h e inclusion of the more optically dense P u r k i n j e cell cytoplasm w i t h t h e stained DNA in t h e nuclear measurement. We have stressed t h e importance of making appropriate corrections for non-specific light loss, whether in tissue, smears, squashes or
TABLE 1
Mean DNA contents (tS.D.)for Purkinje and granule cells of those species shown DNA content
DNA content Species
Mouse
Rat
Guinea pig
Rabbit
Purkinje cell
Granule cell
10.72 0.4 10.7%0.5 10.62 0.5 10.8% 0.5 10.82 0.3 10.7% 0.4 10.82 0.5 12.8%0.8 13.62 1.2 13.3%0.7 13.220.7 13.520.6
10.52 0.4 10.4%0.6 10.4%0.5 10.32 0.5 10.5k0.5 10.42 0.5 10.4% 0.4 12.42 0.6 13.22 0.4 13.52 0.5 13.52 0.6 13.62 0.6
Species
Cat Dog Pig cow Chimpanzee Human
Purklnje cell
Granule cell
10.62 0.6 9.92 0.6 10.02 0.6 9.9k 0.6 10.8'0.7 10.62 0.5 13.22 0.8 8.32 0.6 11.52 0.8 10.72 0.5 11.8% 0.8 8.52 0.4
10.120.6 10.02 0.7 10.12 0.6 10.0%0.7 11.62 0.4 10.7+ 0.5 13.1* 0.7 9.0%0.4 11.520.4 10.32 0.4 12.02 0.5 8.32 0.3
PURKINJE CELL DNA CONTENT
sections, if reliable measurements of DNA content are to be made. This point of view has also been expressed by Morselt e t al. ('72) and Fujita et al. ('74). In every study where a thorough consideration of non-specific factors has been made (Mann and Yates, '73; Morselt et al., '72; Fujita et al., '74), Purkinje and other large nerve cells have been shown to have a diploid DNA content. It has only been in those reports where this problem has been apparently inadequately treated that findings of a tetraploid DNA content have been reported. We conclude therefore that in the variety of species examined by ourselves, Purkinje and other nerve cells are, without exception, diploid. Although these findings contradict certain previous reports, a critical appraisal of the techniques used in these studies leads us to believe that our results are correct. ACKNOWLEDGMENTS
We thank the National Fund for Research into Crippling Diseases who generously supported this work. LITERATURE CITED Bohn, R. C.. and R. B. Mitchell 1976 Cytophotometric identification of tetraploid purkinje cells in young and aged rats. J. Neurobiol., 7: 255-258. Cohen, J., V. Mares and 2. Lodin 1973 DNA content of purified preparations of mouse Purkinje neurones isolated hy a velocity sedimentation technique. J. Neurochem., 20; 651-657. Fujita, S., T. Hattori, M.Fukudu and T. Kitamura 1974 DNA contents in Purkinje cells and inner granule neu-
347
rones in the developing r a t cerebellum. Dev. Growth 81 Diff., 16: 205-211. Herman, C. J., and L. W. Lapham 1968 DNA content of neurones in the cat hippocampus. Science, 160: 537. 1969 Neuronal polyploidy and nuclear volume in cat central nervous system. Brain Res., 15: 35-48. Kusch, A. A., and V. N. Yarygin 1965 Polyploidy of monoand binucleate neurones in upper cervical ganglion of rabbit. Tsitologiia, 7: 228-233. Lapham, L. W. 1965 The tetraploid DNA content of normal human Purkinje cells and its development during the perinatal period. A quantitative cytochemical study. Excerpta. Med (Amst.) Cong. Ser., 100: 445-449. 1968 Tetraploid DNA content of Purkinje neurones of human cerebellar cortex. Science, 159: 310-312. Lentz, R. D., and L. W. Lapham 1969 A quantitative cytochemical study of the DNA content of neurones of rat cerebellar cortex. J. Neurochem., 16: 379-384. 1970 Postnatal development of tetraploid DNA content in r a t Purkinje cells. A quantitative cytochemical study. J. Neuropath. exp. Neurol., 29: 43-56. Mann, D. M. A,, and P. 0. Yates 1973 Polyploidy in the human nervous system. Part I. The DNA content of neurones and glia of the cerebellum. J. Neurol. Sci., 18: 183-196. Manuelidis, L., and E. E. Manuelidis 1974 On the DNA content of cerebellar Purkinje cells in vivo and in vitro. Exp. Neurol., 43: 192-206. Mares, V., 2. Lodin and S. Sacha 1973 A cytochemical and autoradiographic study of nuclear DNA i n mouse purkinje cells. Brain Res., 53: 273-289. Morselt, A. F. W., D. J. Braakman and J. James 1972 Feulgen-DNA and fast green histone estimations in individual cell nuclei of the cerebellum of young and old rats. Acta. Histochem (Jena), 43: 281-286. Novakova, V., W. Sandritter and G. Schlueter 1970 DNA content of neurones in r a t central nervous system. Exptl. Cell. Res., 60: 454-456. Sandritter, W., V. Novakova, J. Pilny and G. Kiefer 1967 Cytophotometrische Messungen des Nucleinsaure und Proteingehaltes von Ganglienzellen der Ratte wahrend der Postnatalen Entwicklungund im Alter. 2. Zellforsch., 80: 145-152.
