{lo}, which suggests a mechanism for accelerated loss
DNA Cvtosine Methylaiion in Brain of Patients with Alzheimer’s Disease Netanel G. Schwob, MD,” Josephine Nalbantoglu, PhD,t Kenneth E. M. Hastings, PhD,” Tom Mikkelsen, MD,” and Neil R. Cashman, MD” We developed a novel quantitative assay to test the hypothesis that defects in DNA cytosine methylation might be responsible for the brain chromatin abnormalities and transcriptional alterations observed in patients with Alzheimer’s disease (AD). We found no significant difference in percent methylation of CCGG sites from brain DNA of 44 patients with AD compared with 20 normal subjects. These results, however, would not exclude genomic redistribution of methylcytosine in AD, or disturbed methylation of a limited population of critical brain-specific genes. Schwob NG, Nalbantoglu J, Hastings KEM, Mikkelsen T, Cashman NR. DNA cytosine methyladon in brain of patients with Alzheimer’s disease. Ann Neurol 1990;28:91-94
Approximately 5% of DNA cytosine residues are methylated in the vertebrate genome, usually in the sequence CpG [l, 21. Much investigation suggests that cytosine methylation participates in the regulation of gene expression. Hypomethylated “CpG islands” are associated with the 5’ regions of many mammalian genes {3]. Tissue-specific genes and their control sequences are often highly methylated in organs in which they are not expressed, such as the erythrocyte globin genes in brain {47. In cell transfection experiments and transgenic animals, transgene expression can be modified by methylation of the exogenous sequence IS, 61. In view of the importance of DNA methylation in gene expression, it is of interest that the percentage of methylated cytosines in the genome appears to diminish with aging [7, 87, particularly in brain [9].Maintenance of methylcytosine during DNA excision repair is more efficient in proliferating than in quiescent cells
From the ‘Department of Neurology and Neurosurgery, McGill University, Montreal Neurological Institute, Montreal, and the ?University of Quebec, Poinre-Claire, Quebec, Canada. Received Aug 25, 1989, and in revised form Jan 4,1990. Accepted for publication Jan 15, 1990. Address correspondence to Dr Cashman, Department of Neurology and Neurosurgery, McGill University, Montreal Neurological Insticute, 3801 University, Montreal, Quebec H3A 2B4, Canada.
of methylcytosine in end-mitotic neurons. In contradistinction to the general trend of dernethylation, DNA repetitive sequences in rat brain undergo increased methylation with aging Ill}. Considering that some repetitive sequence families in rats (BC-1) and primates (BC-200) may be involved in the regulation of brain-specific genes [12, 131, increased methylation of these putative brain-specific gene regulatory elements could have global effects on brain-specific transcription. We hypothesized that gradual alteration in DNA methylation underlies the molecular and cellular derangements of the late-onset neurodegenerative disorders such as Alzheimer’s disease (AD). Prominent alterations in DNA methylation could contribute to the “inactive” chromatin structure 114- 161 and profound global disruption in gene expression C17, IS] observed in brains of patients with AD. We developed a novel assay to measure CCGG methylation in bulk DNA, based on the quantitation of end-labeled DNA fragments generated by the endonuclease isoschizomer pair Msp I and Hpa 11. We believe that our assay has simplified and improved quantitation of isoschizomer endonuclease digestions, in comparison with conventional gel analysis of generated fragments. We found no apparent difference in CCGG methylation between brains of patients with AD and brains of control subjects, but we observed a readily detectable difference between human brain DNA and hypomethylated lambda phage DNA. This study, previously partially presented in abstract form {l9], is to our knowledge the first investigation of DNA methylation in postmortem human brain.
Methods DNA was extracted from 44 brains of patients with AD and 20 control frozen brain slices provided by Drs R. Quirion and Y. Robitaille at the Montreal Douglas Hospital Brain Bank. Fifteen (34%) of 44 patients with AD were male, and 6 (30%) of 20 control subjects were male. The mean age of patients with AD and control subjects was 76.5 years and 59.8 years, respectively. Control brains were from patients dying of nondementing medical illnesses with no neuropathological features of AD. We have incomplete records for the delay between death and autopsy, but the delay between autopsy and DNA extraction was not significantly different in brains from patients with AD and control brains (23.4 + 5.35 months in control subjects and 33.4 + 5.55 months in brains from patients with AD [T = 1.27, NS]). Frontal cortex (approximately 2 gm) dissected from underlying white matter was homogenized in 0.15 M sodium chloride and 0.1 M ethylenediamine tetraacetate, pH 8.0. After addition of sodium dodecyl sulfate to 2% and sodium perchlorate to 1 M, brain DNA was extracted from homogenates with phenol according to the method of Maniatis and colleagues [20]. Lambda phage used as hypomethylated control DNA was purchased from Pharmacia (Dorval, Quebec).
Copyright 0 1790 by the American Neurological Association 91
I
1 lor 100
-
90 80 (3
Msp I
0 70Y
or
x
I
Hpa II
8
5' -NC 3' -NGGC Klenow Fragment Polymerase
I
Pig 1. CCGG methylation assay. Brain D N A is digested by Msp I (recognizingall CCGG) or Hpa I I (cleaving only CCGG nnmethylated at the internal C). Endonwlease-generated fragments are overhang-labeled with phosphorus 32 (32P) deoxycytidine 5'-triphosphate (dCTP) and Klenow fragment D N A polymerase, precipitated by tricbloroacetic acid with excess salmon spem DNA, and deposited on filter discs by a cell harvester; incorporated 32P is quantitated by Cerenkov radiation in water.
40-
::f 10
+ dCTP*
5'-NCC* 3 -NGGC
6050-
OL
L
HPA- II I
92 Annals of Neurology Vol 28 No 1 July 1990
I
1
140
130 120
110 100
c3
90
O 80 F X
70
a
60
0
50 40
30
MSP - I
10
Quadruplicate aliquots of 0.5 pg DNA in 50 pl were digested in round-bottom 96-well microtiter plates (Gibco, Burlington, Ontario) with 2 U of either Hpa I1 (which cleaves only CCGG sequences containing nonmethylated internal cytosines) or Msp I (which cleaves CCGG regardless of whether the internal cytosine is methylated) in buffers recommended by the supplier (Pharmacia). For each patient or control, quadruplicates were also prepared without endonucleases to determine the extent of fragmentation of D N A samples. Digests were incubated overnight at 37C, and the fragments generated were quantitated through an end-filling reaction with 2 U Klenow fragment DNA polymerase (Pharmacia) and phosphorus 32-labeled deoxycytidine 5'-triphosphate (dCTPj (ICN Radiochemicals, Montreal, Quebec) diluted to a specific activity of 300 Cdmmol Thus, the radiolabel incorporated in the by unlabeled dC". DNA samples was proportional to the number of restriction sites recognized by the two enzymes (Fig 1). Labeled DNA was trichloroacetic acid-precipitated with 5 pg of salmon sperm DNA, collected, and washed on filter mats (Skatron, Sterling, VA) with an MIT cell harvester (Cambridge, MA), and filter discs were placed in water-filled vials for quantitation of Cerenkov radiation in a scintillation counter. The percentage of methylated internal cytosines in the
I
0
I
I
I
I
1
2
3
4
Units of enzyme Fig 2. Control brain DNA, restriction enzyme titration: quadruplicate samples (0.5 & digested with Hpa 11 or Msp I and radiolabeled with phosphorus 32 deoxycytidine S'-tPiphosphate, as detailed in Figure 1. Similar digestion Kinetics were observed in brain D N A from patients with Alzheimer's disease (AD). Enzymesfor quantitative comparison of brain D N A from patients with A D and control brain DNA were usedat 2 Ulwell (4UltL&/,within the saturation plateau of each enzyme. restriction site CCGG was calculated by the following forI and CPMH,, 11 refer to the mean mula, where CPMM,~ counts per minute of Klenow end-filled quadruplicates of the two restriction enzymes:
Some labeling of DNA was observed without prior endonuclease digestion (Fig 2j, probably due to Klenow exonuclease-polymerase activity upon D N A fragments gener-
ated by extraction procedures. Restriction enzymes were used at saturating conditions as determined by establishing dose-response curves for each (see Fig 2), and for the Klenow fragment (not shown).
Results To establish the reproducibility and sensitivity of our assay, 10 of our 20 normal brain slices were extracted and assayed in two separate experiments; the mean percent CCGG methylation obtained in the two experiments was 55.7% and 49.0%, respectively. In contrast, phage D N A (assayed with brain D N A as a hypomethylated control) was ?.3% CCGG methylated (mean of two experiments). All quadruplicates varied by 15% or less. The mean CCGG methylation of brain D N A in all control subjects was 52.9 t 1.79% SEM, similar to values obtained in previous studies of human cultured fibroblasts [21). The mean percent CCGG methylation of brain D N A from patients with A D (54.1 & 2.26%) did not significantly differ from that of normal controls (Fig 3). We found no correlation between either sex or age and percent CCGG methylation in either group of brain DNAs, or when all subjects were pooled (not shown). Our sample, however, included only patients aged from 45 to 92 years; a larger sample, including young subjects, may be necessary to detect age-related alteration of D N A methylation. Unexpectedly, with our initial sample (12 AD and 11 controls), we found that D N A from brains of patients with A D prior to enzyme digestion was more fragmented than identically prepared control DNA, as assessed by Klenow exonuclease-polymerase labeling prior to endonuclease digestion ( p < 0.05, unpaired t test E2 I)). Subsequent experiments with the larger sample now reported revealed a similar trend that did not achieve statistical signhcance ( p < 0.2).
Discussion We describe a simple quantitative assay for determining percent methylation of internal cytosines CCGG sites in genomic DNA, and report that brain D N A from patients with A D and control brain D N A have similar mean values of percent CCGG methylation. Degenerating neurons, the cells most likely to reflect the molecular derangements of AD, probably contributed little to the bulk-extracted D N A from brain slices from patients with AD. Nonetheless, an overall reduction in messenger R N A (-A) levels {l?}and in brain-specific mRNAs El81, as well as chromatin abnormalities [14, 151, have been detected in similarly bulk-extracted R N A and D N A samples from AD brain. Thus, although we cannot conclude that degenerating neurons in A D have normal levels of CCGG methylation, we can conclude that global RNA and D N A defects observed in A D are not associated
100
9c 80
c 70
0 .+
60
5
2
50
c3
8 0
40 30 20 10
0
-
Control
x = 52.92
AD x = 54.1 2
-
Fig 3. Percent CCGG methylation of brain D N A from patients with Alzheimer’s disease and control brain D N A , using the formula for methyiation shown i n the text. Dashed lines denote the mean and two standard deviations ofpercent CCGG metbylation ofthe 20 control brain samples.
with significant overall changes in CCGG methylation. Additionally, if the extent of D N A methylation can be regarded as an index of senescence [7-91, our results suggest that brains from patients with A D is not significantly more senescent than aged-matched controls. Our assay detects total CCGG methylation. Although only about 6% of methylated cytosines occur in this sequence, 21% of CCGG sites occur in CpGrich regions 5’ to many vertebrate genes E22). Thus, our assay may be sensitive to methylation status of CpG islands (constituting approximately 1% of total genomic D N A [3)), which contain some upstream elements critical in gene regulation. Our assay, however, would have limited sensitivity to defects in D N A methylation at certain restricted sites such as the BC200 repeat sequence, or specific genes critical in AD. In addition, loss of methylation in single-copy genes occurring in parallel with increased methylation of repeat sequences, as probably occurs in rats 1111, would not be detected by o m assay. Further investigation is
Brief Communication: Schwob et al: DNA Methylation in Alzheimer’s Disease 93
warranted to determine the extent and role of DNA methylation in neurodegenerative disease. This work was supported by a Chercher Boursier grant from Fonds de la Recherche en Sante du Quebec (NRC). We are grateful to Drs R. Quirion and Y . Robitaille for providing the brain samples used in these studies, to S. Boulet for technical assistance, to L. Cetola and R. Cashman for preparing the manuscript, and to Dr J. Antel for helpful comments.
References I. Razin A, Riggs AD. DNA methylation and gene function. Science 1980;210:604-610 2. Adams u p , Burdon m.Molecular biology of DNA methylation. New York: Springer-Verlag, 1985 3. Bird AP. CpG-rich islands and the function of DNA methylation. Nature 1986;321:209-213 4. Shen CKJ, Maniatis T. Tissue-specific DNA methylation in a cluster of rabbit beta-like globin genes. Proc Natl Acad Sci USA 1980;77:6634-6638 5. Busslinger M, Hurst J, Flavell RA. DNA methylation and the regulation of globin gene expression. Cell 1983;34:197-206 6. Keshet I, Yisraeli J, Cedar H. Effect of regional DNA methylation on gene expression. Proc Natl Acad Sci USA 1985;82: 2560-2564 7. Wilson VL, Jones PA. DNA methylation decreases in aging but not in immortal cells. Science 1983;220:1055-1057 8. Wilson VL, Smith RA, Ma S, Cutler RG. Genomic 5methyldeoxycytidine decreases with age. J Biol Chem 1987; 262:9948-995 1 9. Vanyushin BF, Nemirovsky LE,Klimenki VV, et al. The 5methylcytosine in DNA of rats. Tissue and age specificity and changes induced by hydrocortisone and other agents. Gerontologia 1973;19:138- 152 10. Kastan MB, Gowans BJ, Lieberman MW. Methylation of deoxycytidine incorporated by excision-repair synthesis of DNA. Cell 1982;30:509-516 11. Rath PC, Kanungo MS. Methylation of repetitive DNA sequences in the brain during aging of the rat. FEBS Lett 1989; 244~193-198 12. Watson JB, SutcliffeJG. Primate brain-specific cytoplasmic transcript of the Alu repeat family. Mol Cell Biol 1987;7:33243327 13. McKinnon RD, Danielson P, Brow MA, et al. The neuronal identifier sequence as a positive regulatory element for neuronal gene expression. In: Easter SS, Barald KF, Carlson BM, eds. From message to mind. Sunderland: Sinauer Associates, 1988;6:78-89 14. Lewis PN, Lukiw WJ, De Boni U, Crapper-McIachlan DR. Changes in chromatin structure associated with Alzheimer’s disease. J Neurochem 1981;37:1193-1202 15. Crapper-McLachlanDRC, Lewis PN, Lukiw WJ, et al. Chromatin structure in dementia. Ann Neurol 1984;15:329-334 16. Ball DJ, Gross DS, Garrard WT. 5-methylcytosine is localized in nucleosomes that contain histone HI. Proc Natl Acad Sci USA 1983;80:5490-5494 17. Guillernette JG, Wong L, Crapper-McLachlan DR, Lewis PN. Characterization of messenger RNA from cerebral cortex of control and Alzheimer-afflicted brain. J Neurochem 1986;47: 987-997 18. Clark AW, Krekosski CA, Parhad IM, et al. Altered expression of genes for amyloid and cytoskeletal proteins in Alzheimer cortex. Ann Neurol 1989;25:331-339 19. Schwob NG, Nalbantoglu J, Hastings K, et al. Brain DNA
methylation in patients with Alzheimer’s disease. Neurology 1989;39(Suppl 1):251 20. Maniatis T, Fritsch EF, Sambrook J. Molecular cloning-a laboratory manual. New York: Cold Spring Harbor Laboratories, 1982 21. Reis RJS, Goldstein S. Interclonal variation in methylation patterns for expressed and non-expressed genes. Nucleic Acids Res 1982;10:293-304 22. Lindsay S, Bird AP. Use of restriction enzymes to detect potential gene sequences in mammalian DNA. Nature 7987;327: 336-338
Tissue Distribution and Transmission of Mitochondrial DNA Deletions in Mitochondrial Myopathies M. Zeviani, MD,* C. Gellera, BSc,” M. Pannacci, BSc,* G. Uziel, MD,” A. Prelle, MD,? S. Servidei, MD,$ and S. DiDonato, MD*
By using a combination of Southern blot hybridization analysis, polymerase-chain reaction amplification, and direct nucleotide sequencing, we studied deletions of mitochondrial DNA (mtDNA) in several nonfamilial patients with progressive external ophthalmoplegia and Kearns-Sayre syndrome, and in some of their direct relatives. Results suggest that the heteroplasmic mtDNA populations are already present at a very early stage of development, and that there is no direct transmission of mtDNA heteroplasmy by maternal inheritance. Zeviani M, Gellera C, Pannacci M, Uziel G, Prelle A, Servidei S, DiDonato S. Tissue distribution and transmission of mitochondrial DNA deletions in mitochondrial myopathies. Ann Neurol 1990;28:94-97
Heteroplasmy of the human mitochondrial DNA (mtDNA) [l] due to deletions 12-41 or insertions I51 accounts for clinical entities as different as KearnsSayre syndrome (KSS) 12, 31, progressive external ophthalmoplegia (PEO) 141, and Pearson’s syndrome
[a.
From the “Department of Biochemistry and Genetics, Istituto Nazionale Neurologico “Car10 Besta,” Milan, the ‘rlstituto di Clinica Neurologica, Centro “Dino Ferrari,” UniversitP Statale, Milan, and the SClinica Neurologica, Universiti Cattolica, Rome, Italy. Received Oct 31, 1989, and in revised form Dec 27. Accepted for publication Jan 16, 1990. Address correspondence to Dr DiDonato, Department of Biochemistry and Genetics, Istituto Nazionale Neurologico “C. Besta,” via Celoria 11, 20133, Milan, Italy.
94 Copyright 0 1990 by the American Neurological Association
DNA Cvtosine Methylaiion in Brain of Patients with Alzheimer’s Disease Netanel G. Schwob, MD,” Josephine Nalbantoglu, PhD,t Kenneth E. M. Hastings, PhD,” Tom Mikkelsen, MD,” and Neil R. Cashman, MD” We developed a novel quantitative assay to test the hypothesis that defects in DNA cytosine methylation might be responsible for the brain chromatin abnormalities and transcriptional alterations observed in patients with Alzheimer’s disease (AD). We found no significant difference in percent methylation of CCGG sites from brain DNA of 44 patients with AD compared with 20 normal subjects. These results, however, would not exclude genomic redistribution of methylcytosine in AD, or disturbed methylation of a limited population of critical brain-specific genes. Schwob NG, Nalbantoglu J, Hastings KEM, Mikkelsen T, Cashman NR. DNA cytosine methyladon in brain of patients with Alzheimer’s disease. Ann Neurol 1990;28:91-94
Approximately 5% of DNA cytosine residues are methylated in the vertebrate genome, usually in the sequence CpG [l, 21. Much investigation suggests that cytosine methylation participates in the regulation of gene expression. Hypomethylated “CpG islands” are associated with the 5’ regions of many mammalian genes {3]. Tissue-specific genes and their control sequences are often highly methylated in organs in which they are not expressed, such as the erythrocyte globin genes in brain {47. In cell transfection experiments and transgenic animals, transgene expression can be modified by methylation of the exogenous sequence IS, 61. In view of the importance of DNA methylation in gene expression, it is of interest that the percentage of methylated cytosines in the genome appears to diminish with aging [7, 87, particularly in brain [9].Maintenance of methylcytosine during DNA excision repair is more efficient in proliferating than in quiescent cells
From the ‘Department of Neurology and Neurosurgery, McGill University, Montreal Neurological Institute, Montreal, and the ?University of Quebec, Poinre-Claire, Quebec, Canada. Received Aug 25, 1989, and in revised form Jan 4,1990. Accepted for publication Jan 15, 1990. Address correspondence to Dr Cashman, Department of Neurology and Neurosurgery, McGill University, Montreal Neurological Insticute, 3801 University, Montreal, Quebec H3A 2B4, Canada.
of methylcytosine in end-mitotic neurons. In contradistinction to the general trend of dernethylation, DNA repetitive sequences in rat brain undergo increased methylation with aging Ill}. Considering that some repetitive sequence families in rats (BC-1) and primates (BC-200) may be involved in the regulation of brain-specific genes [12, 131, increased methylation of these putative brain-specific gene regulatory elements could have global effects on brain-specific transcription. We hypothesized that gradual alteration in DNA methylation underlies the molecular and cellular derangements of the late-onset neurodegenerative disorders such as Alzheimer’s disease (AD). Prominent alterations in DNA methylation could contribute to the “inactive” chromatin structure 114- 161 and profound global disruption in gene expression C17, IS] observed in brains of patients with AD. We developed a novel assay to measure CCGG methylation in bulk DNA, based on the quantitation of end-labeled DNA fragments generated by the endonuclease isoschizomer pair Msp I and Hpa 11. We believe that our assay has simplified and improved quantitation of isoschizomer endonuclease digestions, in comparison with conventional gel analysis of generated fragments. We found no apparent difference in CCGG methylation between brains of patients with AD and brains of control subjects, but we observed a readily detectable difference between human brain DNA and hypomethylated lambda phage DNA. This study, previously partially presented in abstract form {l9], is to our knowledge the first investigation of DNA methylation in postmortem human brain.
Methods DNA was extracted from 44 brains of patients with AD and 20 control frozen brain slices provided by Drs R. Quirion and Y. Robitaille at the Montreal Douglas Hospital Brain Bank. Fifteen (34%) of 44 patients with AD were male, and 6 (30%) of 20 control subjects were male. The mean age of patients with AD and control subjects was 76.5 years and 59.8 years, respectively. Control brains were from patients dying of nondementing medical illnesses with no neuropathological features of AD. We have incomplete records for the delay between death and autopsy, but the delay between autopsy and DNA extraction was not significantly different in brains from patients with AD and control brains (23.4 + 5.35 months in control subjects and 33.4 + 5.55 months in brains from patients with AD [T = 1.27, NS]). Frontal cortex (approximately 2 gm) dissected from underlying white matter was homogenized in 0.15 M sodium chloride and 0.1 M ethylenediamine tetraacetate, pH 8.0. After addition of sodium dodecyl sulfate to 2% and sodium perchlorate to 1 M, brain DNA was extracted from homogenates with phenol according to the method of Maniatis and colleagues [20]. Lambda phage used as hypomethylated control DNA was purchased from Pharmacia (Dorval, Quebec).
Copyright 0 1790 by the American Neurological Association 91
I
1 lor 100
-
90 80 (3
Msp I
0 70Y
or
x
I
Hpa II
8
5' -NC 3' -NGGC Klenow Fragment Polymerase
I
Pig 1. CCGG methylation assay. Brain D N A is digested by Msp I (recognizingall CCGG) or Hpa I I (cleaving only CCGG nnmethylated at the internal C). Endonwlease-generated fragments are overhang-labeled with phosphorus 32 (32P) deoxycytidine 5'-triphosphate (dCTP) and Klenow fragment D N A polymerase, precipitated by tricbloroacetic acid with excess salmon spem DNA, and deposited on filter discs by a cell harvester; incorporated 32P is quantitated by Cerenkov radiation in water.
40-
::f 10
+ dCTP*
5'-NCC* 3 -NGGC
6050-
OL
L
HPA- II I
92 Annals of Neurology Vol 28 No 1 July 1990
I
1
140
130 120
110 100
c3
90
O 80 F X
70
a
60
0
50 40
30
MSP - I
10
Quadruplicate aliquots of 0.5 pg DNA in 50 pl were digested in round-bottom 96-well microtiter plates (Gibco, Burlington, Ontario) with 2 U of either Hpa I1 (which cleaves only CCGG sequences containing nonmethylated internal cytosines) or Msp I (which cleaves CCGG regardless of whether the internal cytosine is methylated) in buffers recommended by the supplier (Pharmacia). For each patient or control, quadruplicates were also prepared without endonucleases to determine the extent of fragmentation of D N A samples. Digests were incubated overnight at 37C, and the fragments generated were quantitated through an end-filling reaction with 2 U Klenow fragment DNA polymerase (Pharmacia) and phosphorus 32-labeled deoxycytidine 5'-triphosphate (dCTPj (ICN Radiochemicals, Montreal, Quebec) diluted to a specific activity of 300 Cdmmol Thus, the radiolabel incorporated in the by unlabeled dC". DNA samples was proportional to the number of restriction sites recognized by the two enzymes (Fig 1). Labeled DNA was trichloroacetic acid-precipitated with 5 pg of salmon sperm DNA, collected, and washed on filter mats (Skatron, Sterling, VA) with an MIT cell harvester (Cambridge, MA), and filter discs were placed in water-filled vials for quantitation of Cerenkov radiation in a scintillation counter. The percentage of methylated internal cytosines in the
I
0
I
I
I
I
1
2
3
4
Units of enzyme Fig 2. Control brain DNA, restriction enzyme titration: quadruplicate samples (0.5 & digested with Hpa 11 or Msp I and radiolabeled with phosphorus 32 deoxycytidine S'-tPiphosphate, as detailed in Figure 1. Similar digestion Kinetics were observed in brain D N A from patients with Alzheimer's disease (AD). Enzymesfor quantitative comparison of brain D N A from patients with A D and control brain DNA were usedat 2 Ulwell (4UltL&/,within the saturation plateau of each enzyme. restriction site CCGG was calculated by the following forI and CPMH,, 11 refer to the mean mula, where CPMM,~ counts per minute of Klenow end-filled quadruplicates of the two restriction enzymes:
Some labeling of DNA was observed without prior endonuclease digestion (Fig 2j, probably due to Klenow exonuclease-polymerase activity upon D N A fragments gener-
ated by extraction procedures. Restriction enzymes were used at saturating conditions as determined by establishing dose-response curves for each (see Fig 2), and for the Klenow fragment (not shown).
Results To establish the reproducibility and sensitivity of our assay, 10 of our 20 normal brain slices were extracted and assayed in two separate experiments; the mean percent CCGG methylation obtained in the two experiments was 55.7% and 49.0%, respectively. In contrast, phage D N A (assayed with brain D N A as a hypomethylated control) was ?.3% CCGG methylated (mean of two experiments). All quadruplicates varied by 15% or less. The mean CCGG methylation of brain D N A in all control subjects was 52.9 t 1.79% SEM, similar to values obtained in previous studies of human cultured fibroblasts [21). The mean percent CCGG methylation of brain D N A from patients with A D (54.1 & 2.26%) did not significantly differ from that of normal controls (Fig 3). We found no correlation between either sex or age and percent CCGG methylation in either group of brain DNAs, or when all subjects were pooled (not shown). Our sample, however, included only patients aged from 45 to 92 years; a larger sample, including young subjects, may be necessary to detect age-related alteration of D N A methylation. Unexpectedly, with our initial sample (12 AD and 11 controls), we found that D N A from brains of patients with A D prior to enzyme digestion was more fragmented than identically prepared control DNA, as assessed by Klenow exonuclease-polymerase labeling prior to endonuclease digestion ( p < 0.05, unpaired t test E2 I)). Subsequent experiments with the larger sample now reported revealed a similar trend that did not achieve statistical signhcance ( p < 0.2).
Discussion We describe a simple quantitative assay for determining percent methylation of internal cytosines CCGG sites in genomic DNA, and report that brain D N A from patients with A D and control brain D N A have similar mean values of percent CCGG methylation. Degenerating neurons, the cells most likely to reflect the molecular derangements of AD, probably contributed little to the bulk-extracted D N A from brain slices from patients with AD. Nonetheless, an overall reduction in messenger R N A (-A) levels {l?}and in brain-specific mRNAs El81, as well as chromatin abnormalities [14, 151, have been detected in similarly bulk-extracted R N A and D N A samples from AD brain. Thus, although we cannot conclude that degenerating neurons in A D have normal levels of CCGG methylation, we can conclude that global RNA and D N A defects observed in A D are not associated
100
9c 80
c 70
0 .+
60
5
2
50
c3
8 0
40 30 20 10
0
-
Control
x = 52.92
AD x = 54.1 2
-
Fig 3. Percent CCGG methylation of brain D N A from patients with Alzheimer’s disease and control brain D N A , using the formula for methyiation shown i n the text. Dashed lines denote the mean and two standard deviations ofpercent CCGG metbylation ofthe 20 control brain samples.
with significant overall changes in CCGG methylation. Additionally, if the extent of D N A methylation can be regarded as an index of senescence [7-91, our results suggest that brains from patients with A D is not significantly more senescent than aged-matched controls. Our assay detects total CCGG methylation. Although only about 6% of methylated cytosines occur in this sequence, 21% of CCGG sites occur in CpGrich regions 5’ to many vertebrate genes E22). Thus, our assay may be sensitive to methylation status of CpG islands (constituting approximately 1% of total genomic D N A [3)), which contain some upstream elements critical in gene regulation. Our assay, however, would have limited sensitivity to defects in D N A methylation at certain restricted sites such as the BC200 repeat sequence, or specific genes critical in AD. In addition, loss of methylation in single-copy genes occurring in parallel with increased methylation of repeat sequences, as probably occurs in rats 1111, would not be detected by o m assay. Further investigation is
Brief Communication: Schwob et al: DNA Methylation in Alzheimer’s Disease 93
warranted to determine the extent and role of DNA methylation in neurodegenerative disease. This work was supported by a Chercher Boursier grant from Fonds de la Recherche en Sante du Quebec (NRC). We are grateful to Drs R. Quirion and Y . Robitaille for providing the brain samples used in these studies, to S. Boulet for technical assistance, to L. Cetola and R. Cashman for preparing the manuscript, and to Dr J. Antel for helpful comments.
References I. Razin A, Riggs AD. DNA methylation and gene function. Science 1980;210:604-610 2. Adams u p , Burdon m.Molecular biology of DNA methylation. New York: Springer-Verlag, 1985 3. Bird AP. CpG-rich islands and the function of DNA methylation. Nature 1986;321:209-213 4. Shen CKJ, Maniatis T. Tissue-specific DNA methylation in a cluster of rabbit beta-like globin genes. Proc Natl Acad Sci USA 1980;77:6634-6638 5. Busslinger M, Hurst J, Flavell RA. DNA methylation and the regulation of globin gene expression. Cell 1983;34:197-206 6. Keshet I, Yisraeli J, Cedar H. Effect of regional DNA methylation on gene expression. Proc Natl Acad Sci USA 1985;82: 2560-2564 7. Wilson VL, Jones PA. DNA methylation decreases in aging but not in immortal cells. Science 1983;220:1055-1057 8. Wilson VL, Smith RA, Ma S, Cutler RG. Genomic 5methyldeoxycytidine decreases with age. J Biol Chem 1987; 262:9948-995 1 9. Vanyushin BF, Nemirovsky LE,Klimenki VV, et al. The 5methylcytosine in DNA of rats. Tissue and age specificity and changes induced by hydrocortisone and other agents. Gerontologia 1973;19:138- 152 10. Kastan MB, Gowans BJ, Lieberman MW. Methylation of deoxycytidine incorporated by excision-repair synthesis of DNA. Cell 1982;30:509-516 11. Rath PC, Kanungo MS. Methylation of repetitive DNA sequences in the brain during aging of the rat. FEBS Lett 1989; 244~193-198 12. Watson JB, SutcliffeJG. Primate brain-specific cytoplasmic transcript of the Alu repeat family. Mol Cell Biol 1987;7:33243327 13. McKinnon RD, Danielson P, Brow MA, et al. The neuronal identifier sequence as a positive regulatory element for neuronal gene expression. In: Easter SS, Barald KF, Carlson BM, eds. From message to mind. Sunderland: Sinauer Associates, 1988;6:78-89 14. Lewis PN, Lukiw WJ, De Boni U, Crapper-McIachlan DR. Changes in chromatin structure associated with Alzheimer’s disease. J Neurochem 1981;37:1193-1202 15. Crapper-McLachlanDRC, Lewis PN, Lukiw WJ, et al. Chromatin structure in dementia. Ann Neurol 1984;15:329-334 16. Ball DJ, Gross DS, Garrard WT. 5-methylcytosine is localized in nucleosomes that contain histone HI. Proc Natl Acad Sci USA 1983;80:5490-5494 17. Guillernette JG, Wong L, Crapper-McLachlan DR, Lewis PN. Characterization of messenger RNA from cerebral cortex of control and Alzheimer-afflicted brain. J Neurochem 1986;47: 987-997 18. Clark AW, Krekosski CA, Parhad IM, et al. Altered expression of genes for amyloid and cytoskeletal proteins in Alzheimer cortex. Ann Neurol 1989;25:331-339 19. Schwob NG, Nalbantoglu J, Hastings K, et al. Brain DNA
methylation in patients with Alzheimer’s disease. Neurology 1989;39(Suppl 1):251 20. Maniatis T, Fritsch EF, Sambrook J. Molecular cloning-a laboratory manual. New York: Cold Spring Harbor Laboratories, 1982 21. Reis RJS, Goldstein S. Interclonal variation in methylation patterns for expressed and non-expressed genes. Nucleic Acids Res 1982;10:293-304 22. Lindsay S, Bird AP. Use of restriction enzymes to detect potential gene sequences in mammalian DNA. Nature 7987;327: 336-338
Tissue Distribution and Transmission of Mitochondrial DNA Deletions in Mitochondrial Myopathies M. Zeviani, MD,* C. Gellera, BSc,” M. Pannacci, BSc,* G. Uziel, MD,” A. Prelle, MD,? S. Servidei, MD,$ and S. DiDonato, MD*
By using a combination of Southern blot hybridization analysis, polymerase-chain reaction amplification, and direct nucleotide sequencing, we studied deletions of mitochondrial DNA (mtDNA) in several nonfamilial patients with progressive external ophthalmoplegia and Kearns-Sayre syndrome, and in some of their direct relatives. Results suggest that the heteroplasmic mtDNA populations are already present at a very early stage of development, and that there is no direct transmission of mtDNA heteroplasmy by maternal inheritance. Zeviani M, Gellera C, Pannacci M, Uziel G, Prelle A, Servidei S, DiDonato S. Tissue distribution and transmission of mitochondrial DNA deletions in mitochondrial myopathies. Ann Neurol 1990;28:94-97
Heteroplasmy of the human mitochondrial DNA (mtDNA) [l] due to deletions 12-41 or insertions I51 accounts for clinical entities as different as KearnsSayre syndrome (KSS) 12, 31, progressive external ophthalmoplegia (PEO) 141, and Pearson’s syndrome
[a.
From the “Department of Biochemistry and Genetics, Istituto Nazionale Neurologico “Car10 Besta,” Milan, the ‘rlstituto di Clinica Neurologica, Centro “Dino Ferrari,” UniversitP Statale, Milan, and the SClinica Neurologica, Universiti Cattolica, Rome, Italy. Received Oct 31, 1989, and in revised form Dec 27. Accepted for publication Jan 16, 1990. Address correspondence to Dr DiDonato, Department of Biochemistry and Genetics, Istituto Nazionale Neurologico “C. Besta,” via Celoria 11, 20133, Milan, Italy.
94 Copyright 0 1990 by the American Neurological Association
