SEARCH

SEARCH BY CITATION

Keywords:

  • Alzheimer's disease;
  • DNA oxidation;
  • Oxidative stress;
  • Ventricular CSF;
  • Gas chromatography/mass spectrometry with selective ion monitoring

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

Abstract : One of the leading etiologic hypotheses regarding Alzheimer's disease (AD) is the involvement of free radical-mediated oxidative stress in neuronal degeneration. Although several recent studies show an increase in levels of brain DNA oxidation in both aging and AD, there have been no studies of levels of markers of DNA oxidation in ventricular CSF. This is a study of levels of 8-hydroxy-2′-deoxyguanosine (8-OHdG), the predominant marker of oxidative DNA damage, in intact DNA and as the “free” repair product that results from repair mechanisms. Free 8-OHdG was isolated from CSF from nine AD and five age-matched control subjects using solidphase extraction columns and measured using gas chromatography/mass spectrometry with selective ion monitoring. Intact DNA was isolated from the same samples and the levels of 8-OHdG determined in the intact structures. Quantification of results was carried out using stable isotope-labeled 8-OHdG. By using this sensitive methodology, statistically significant elevations (p < 0.05) of 8-OHdG were observed in intact DNA in AD subjects compared with age-matched control subjects. In contrast, levels of free 8-OHdG, removed via repair mechanisms, were depleted significantly in AD samples (p < 0.05). Our results demonstrate an increase in unrepaired oxygen radical-mediated damage in AD DNA as evidenced by the increased presence of 8-OHdG in intact DNA and decreased concentrations of the free repair product. These data suggest that the brain in AD may be subject to the double insult of increased oxidative stress, as well as deficiencies in repair mechanisms responsible for removal of oxidized bases.

Recent evidence suggests that oxidative stress may contribute to neuronal degeneration in several neurological disorders, including stroke, amyotrophic lateral sclerosis, Parkinson's disease, head trauma, and Alzheimer's disease (AD). Evidence for a role of oxidative stress in AD includes studies that demonstrate an increase in levels of brain iron (Ehmann et al., 1986 ; Thompson et al., 1988 ; Good et al., 1992 ; Samudralwar et al., 1995), the presence of redox active iron (Smith et al., 1997), increased levels of lipid peroxidation (Lovell et al., 1995), protein oxidation (Smith et al., 1991 ; Hensley et al., 1995), 4-hydroxynonenal (HNE), a marker of lipid peroxidation, in AD brain (Markesbery and Lovell, 1998) and ventricular CSF (Lovell et al., 1997), as well as a concomitant decline in polyunsaturated fatty acids (Svennerholm and Gottfries, 1994 ; Prasad et al., 1998). Markers of oxidative stress are present in neurofibrillary tangles and senile plaques in AD brain (Smith et al., 1994, 1995, 1996 ; Yan et al., 1994 ; Good et al., 1996 ; Montine et al., 1997 ; Sayre et al., 1997). Transgenic mice that develop amyloid-β-peptide deposits in the brain have the features of oxidative stress found in AD (Smith et al., 1998).

Markers of oxidative stress in nuclear and mitochondrial DNA are increased in AD brain (Mecocci et al., 1994 ; Lyras et al., 1997 ; Gabbita et al., 1998) and in normal aging (Loft and Poulsen, 1996). Attack of DNA by reactive oxygen species leads to the hydroxylation of DNA bases (Adelman et al., 1988 ; Wagner et al., 1992), the most prominent of which is 8-hydroxy-2′-deoxyguanosine (8-OHdG). Levels of 8-OHdG are elevated in mitochondrial DNA in AD cerebral cortex compared with controls (Mecocci et al., 1994), but not in nuclear DNA (TeKoppele et al., 1996). In a recent study from our laboratory, Gabbita et al. (1998), using gas chromatography/mass spectrometry with selective ion monitoring (GC/MS-SIM) to measure six oxidatively modified bases, observed a statistically significant increase of 8-OHdG, 8-hydroxy-2′-dexyadenine, and 5-hydroxyuracil, a product of cytosine oxidation, in parietal, temporal, and frontal lobes in AD compared with age-matched control subjects. In addition, we found a statistically significant increase in 5-hydroxycytosine (another product of cytosine oxidation) in AD parietal and temporal lobes compared with age-matched control subjects. Of the six base adducts analyzed, 8-OHdG demonstrated the largest absolute levels, indicating its prominence as a marker of oxidative stress in DNA. Although 8-OHdG levels have been relatively well studied in intact DNA extracted from brain, there have been no studies of levels of the modified base in intact DNA or free modified base in CSF, which serves to filter and eventually dispose of degraded cellular material, and thus may offer a means of assessing levels of DNA oxidation in living subjects.

This study provides a quantitative measurement of concentrations of 8-OHdG both in intact DNA and as the free repair product in ventricular CSF of autopsied AD and age-matched control subjects using stable isotope dilution GC/MS-SIM.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

Reagents

N,O-bis(Trimethylsilyl)trifluoroacetamide (BSTFA) plus 1% trimethylsilylchlorosilane and anhydrous pyridine were from Aldrich Chemicals (Milwaukee, WI, U.S.A.). Molecular biology grade phenol, chloroform, isoamyl alcohol, proteinase K, and other standard chemicals were from Sigma (St. Louis, MO, U.S.A.). Stable labeled [8-13C,7,9-15N2]8-OHdG and guanine were purchased from Cambridge Isotope Laboratories (Andover, MA, U.S.A.). Cutoff filters of 5,000 moleduclar weight were purchased from Fisher, and C18 solid-phase extraction columns were from Waters.

Ventricular CSF sampling

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

Approximately 20-25 ml of ventricular CSF was removed from the lateral ventricle at autopsy using an 18-gauge spinal needle and virgin polyethylene syringes from nine AD (four men, five women) and five control (one man, four women) subjects. The demographic data are presented in Table 1. The smaples were centrifuged immediately to remove particulate matter, and the supernatant was placed in fresh polyethylene tubes and stored at -80°C until used for analysis. The mean ± SEM age was 78.2 ± 2.9 years for AD subjects and 80.8 ± 2.9 years for control subjects. The mean postmortem interval (PMI) was 2.5 ± 0.3 h for AD subjects and 3.2 ± 0.4 h for control subjects. All AD subjects demonstrated progressive intellectual decline and met the NINCDS-ADRDA Work Group criteria (McKhann et al., 1984) for the clinical diagnosis of probable AD. Histopathologic diagnosis was based on study of multiple sections of neocortex, hippocampus, amygdala, entorhinal cortex, basal ganglia, brainstem, and cerebellum stained with hematoxylin and eosin, the modified Bielschowsky silver stain, 10D-5 (for β-amyloid, Athena Neurosciences) (Hyman et al., 1992), and ubiquitin immunochemistry. All AD subjects met accepted criteria for histopathologic diagnosis of AD (Mirra et al., 1991 ; NIA-Reagan Institute, 1997). Control subjects were individuals without a history of dementia or other neurological disorders who underwent annual neuropsychologic testing as a part of our normal volunteer control group study. All control subjects had test scores within the normal range. Neuropathologic evaluation of control brains revealed only age-associated gross and histopathologic alterations. Mean brain weight in AD subjects was significantly lower (1,065 ± 52 g) than that of controls (1,241 ± 62 g) (p < 0.05). Subjects were excluded from this study who had been on a respirator or who had suffered prolonged terminal hypoxia, recent or old infarcts, intracranial hemorrhages, drug intoxication, alcoholism, or metastatic diseases of the central nervous system. Hypoxic changes were not found in any brain region on microscopic examination in any of the subjects used in this study.

Table 1.  Subject demographic dataAge, PMI, DNA content, Braak staging, and protein content for AD and control subjects are presented as means ± SEM.
 Age (yr)PMI (h) DNA content (μg/ml of CSF) SexBraak stageProtein content (mg/ml)
AD78.2 ± 2.92.53 ± 0.295.0 ± 1.34 M/5 F61.43 ± 0.10
Control80.8 ± 2.93.2 ± 0.45.6 ± 1.71 M/4 F1.0 ± 0.411.36 ± 0.15

Isolation of free 8-OHdG

Ventricular CSF samples were thawed at room temperature, and free 8-OHdG was purified using a modification of the procedure of Shigenaga et al. (1994). Samples were thoroughly vortexed for 2 min and 20-μl aliquots removed for protein content determinations using the Pierce bicinchoninic acid method (Sigma). Accurately measured volumes (9.9-33.4 ml) of CSF were then passed through C18 solid-phase extraction columns that were preconditioned with 10 ml of HPLC grade methanol, 10 ml of 18 MΩ distilled/deionized water, and 10 ml of 50 mM KH2PO4 (pH 7.4) under slight vacuum. The eluent CSF was collected in fresh side-arm test tubes to prevent cross-contamination between samples and reserved for intact DNA isolation. After addition of the CSF samples, an aliquot of stable labeled 8-OHdG as an internal standard was added, and the columns were washed with 4 ml of 50 mM KH2PO4 followed by 2 × 2 ml washes of 5% methanol/KH2PO4. The 8-OHdG was eluted from the column with 3 ml of 15% methanol/KH2PO4. The internal standard was added to the Sep-Pak column after addition of the sample rather than directly to the sample, because variable separation efficiencies could contribute to errors in the quantification of 8-OHdG in intact DNA. The eluent was added to new C18 columns preconditioned as described above. The columns were washed with 1 ml of distilled/deionized water and dried for 15 min under slight vacuum. Purified 8-HdG was eluted with 2 ml of HPLC grade methanol, placed in 5-ml conical glass tubes, and lyophilized. Using standard solutions of DNA and nonlabeled 8-OHdG, the purification was found to allow passage of ~97% of DNA while retaining nearly all 8-OHdG as determined by UV-Vis spectrometry.

Isolation of DNA from ventricular CSF

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

Intact DNA was isolated from ventricular CSF using a modification of the procedure of Mecocci et al. (1993). The CSF was mixed with a 1/10 volume of 1 M Na2EDTA, 5% sodium dodecyl sulfate, and 50 mM Tris-HCl (pH 8.0) along with 400 μl of 10 mg/ml proteinase K. The samples were incubated for 2 h at 37°C to digest any cellular material. After heating, 2.4 ml of 5 M NaCl was added and the solution extracted three times with buffer-saturated phenol containing 5.5 mM-hydroxyquinoline to prevent artifactual oxidation of DNA. The samples was then extracted three times with 24 : 1 chloroform/isoamyl alcohol. An additional 2.4 ml of 5 M NaCl was added and the solution centrifuged through a 5,000 molecular weight cutoff filter at a speed of 2,000 g for ~14 h at 4°C. The resulting DNA was resuspended in 1 ml of distilled/deionized water and the concentration determined at 260 nm with 50 μg of DNA equal to an absorbance of 1.0 using a Gynesys 5 UV-VIS spectrophotometer. Absorbances were also measured at 280 nm, with the mean ± SEM 260\280 ratio of 1.5 ± 0.04 indicating a relatively pure DNA preparation with some slight protein contamination. Polyacrylamide gel electrophoresis of a representative DNA sample indicated the presence of a single band of ~400 bp.

Stable labeled 8-OHdG was then added and the samples lyophilized. After lyophilization, the DNA and free 8-OHdG samples were subjected to formic acid hydrolysis and BSTFA derivatization as previously described (Gabbita et al., 1998). The samples were placed in 5-ml conical glass tubes with Teflon-backed septa. Five hundred microliters of 90% formic acid was added, and the tubes were evacuated completely by interesting an 18-gauge needle and withdrawing the air. The samples were then heated at 140°C overnight to cleave the glycosidic bonds between bases and sugar moieties in DNA, releasing the bases. After hydrolysis, the samples were lyophilized and the DNA samples extracted with 2 ml of absolute ethanol dried over sodium sulfate by vortexing for 2 min. After centrifugation at 800 g for 10 min to pellet sugars and any other nonhydrolyzed material, 110 μl was removed and placed in fresh glass tubes. The samples were again lyophilized. For derivatization, the residue was vortexed with 200 μl of 1 : 1 BSTFA/anhydrous pyridine at 140°C for 16 h. The sample tubes were evacuated to prevent antifactual oxidation of guanine during hydrolysis or derivatization. After derivatization, the samples were again lyophilized (~5 min) and the residue resuspended in 20 μl of BSTFA/pyridine immediately before analysis.

GC/MS-SIM analysis

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

Analysis of samples was carried out on a Hewlett-Packard model HP6890 gas chromatograph equipped with a mass spectrometer operated in selective ion-monitoring mode. Chromatographic separations were on a 30-m, 0.25-mm i.d., 0.25-μm film thickness phenylmethylsiloxane column using 99.999% helium as a carrier gas at an inlet pressure of 11.8 psi and a constant flow. The injector port was maintained at 250°C and was lined with silanized glass wool to provide homogeneous vaporization of samples. Initial temperature in the column was held at 100°C for 2 min, followed by the following temperature ramps : 100-178°C at 3°C/min, 178-181°C at 0.3°C/min, 181-208°C at 3°C/min, and 208-280°C at 10°C/min, where the final temperature was held for 2 min. The temperature of the ion source in the mass spectrometer was maintained at 200°C. Identification of 8-OHdG was via selective ion monitoring. Retention times for 8-OHdG were essentially constant. Shot-to-shot variation was <2% for standard 8-OHdG samples. Levels of 8-OHdG were quantified using stable labeled 8-OHdG as an internal standard as described by Dizdaroglu (1993). For quantification, peaks of interest were integrated and the area of the 8-OHdG peak compared with the internal standard peak, which has a known concentration. Because GC/MS uses mass spectrometry as the detector, the signals from the labeled and unlabeled compounds may be separated. The advantage of using stable labeled 8-OHdG as an internal standard is that it responds in the same way as the compound of interest during hydrolysis and derivatization.

Statistical analyses

Statistical analyses of the data were carried out using a one-tailed Student's t test, because it provided sufficient statistical power for the relatively small number of samples analyzed. The analyses were performed using the commercially available ABSTAT software.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

Statistical comparison of mean age, PMI, DNA content (micrograms per milliliter), and protein content (milligrams per milliliter) indicated no statistically significant differences between AD and control subjects (Table 1). The means ± SEM of free and DNA-bound 8-OHdG are shown in Fig. 1. The results are expressed as nanomoles of derivatized 8-OHdG per microgram of DNA for DNA-bound material and nanomoles of derivatized 8-OHdG per milligram of protein for free material. Protein content was determined using the Pierce bicinchoninic acid method. Levels of 8-OHdG in intact DNA were elevated significantly in ventricular CSF samples from the eight AD subjects (0.021 ± 0.007 nmol/μg of DNA) compared with samples from five age-matched control subjects (0.004 ± 0.0008 nmol/μg of DNA) (p = 0.04) (Table 2). Calculation of correlation coefficients between Braak staging (an index of neuropathologic severity) and levels of 8-OHdG indicated a statistically significant (p < 0.05) positive correlation of 0.8911.

image

Figure 1. Mean ± SEM levels of free 8-OHdG and 8-OHdG in intact DNA in ventricular CSF of AD and control subjects. *p < 0.05, one-sided Student's t test.

Download figure to PowerPoint

Table 2.  Levels of 8-OHdG in intact DNA and as the free repair product in individual AD and control subjects
Alzheimer's diseaseControls 
Subject no.nmol of 8-OHdG in intact DNAnmol of free 8-OHdGnmol of 8-OHdG in intact DNAnmol of free 8-OHdG
  1. a p < 0.05, significant elevation in AD CSF, one-sided Student's t test.

  2. b p < 0.05, significant elevation in control CSF, one-sided Student's t test.

10.01780.00480.258
20.05150.00470.157
30.0540.2350.00160.296
40.00280.0360.00560.0064
50.0320.1620.00220.116
60.00320.002  
70.0470.007  
80.00460.002  
90.00700.003  
Mean ± SEM0.021 ± 0.007a0.062 ± 0.0310.004 ± 0.00080.167 ± 0.052b

In contrast, free 8-OHdG levels, which reflect modified bases removed from DNA via repair mechanisms, demonstrated opposite trends. Free 8-OHdG levels were statistically significantly (p = 0.05) depleted in AD subjects (0.062 ± 0.031 nmol/mg of protein) compared with control subjects (0.167 ± 0.052 nmol/mg of protein) (Table 2). Calculation of the correlation coefficient for Braak staging and free 8-OHdG values demonstrated a negative correlation (r = -0.40), which was not statistically significant. Calculation of correlation coefficients for levels of CSF 8-OHdG in intact DNA and levels of 8-OHdG in nuclear DNA in frontal, temporal, and parietal lobes in a subset of two AD and two control subjects indicated a strong positive correlation between the two (r = +0.97). There were no meaningful correlations between levels of 8-OHdG in brain nuclear DNA and levels of free 8-OHdG in ventricular CSF. Calculation of correlation coefficients for brain weight and CSF free 8-OHdG revealed a significant positive correlation (r = +0.73 ; p < 0.05) and an insignificant weak positive correlation between brain weight and 8-OHdG in intact DNA in CSF (r = +0.46). These correlations between brain weight and 8-OHdG in control subjects were not meaningful.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

This is the first study to demonstrate an increase in levels of 8-OHdG in intact DNA while simultaneously demonstrating a significant decrease in levels of free 8-OHdG in ventricular CSF from short PMI AD patients compared with control subjects. Ventricular CSF serves to filter and dispose of degraded cellular material from the brain and would serve as a better index of brain DNA oxidation compared with urine or blood, which would reflect total body DNA damage. The differences observed reached statistical significance (p± 0.05) for both free 8-OHdG and 8-OHdG in intact DNA, with 8-OHdG in intact DNA reaching levels 5.3 times those in control subjects. Free 8-OHdG, which results from its excision from damaged DNA presumably by base-specific glycosylases, was 2.4 times lower in AD CSF compared with levels in control subjects. These findings correlate with the index of neuropathologic severity of AD for 8-OHdG in intact DNA. Free 8-OHdG demonstrated a negative correlation with Braak staging, although it was not statistically significant.

This study used stable labeled 8-OHdG and GC/MSSIM to allow for unequivocal identification of peaks of interest based not only on chromatographic retention time, but also on mass spectra. The results of our current study agree well with a previous study from our laboratory that showed a statistically significant increase in levels of 8-OHdG in nuclear DNA samples isolated from frontal, temporal, and parietal lobe structures from AD subjects (Gabbita et al., 1998) compared with age-matched control subjects. The concentrations of 8-OHdG observed for DNA from ventricular CSF are relatively comparable to those observed in our nuclear DNA study (AD temporal lobe DNA, 0.008 nmol/μg of DNA ; AD CSF, 0.0210 nmol/μg of DNA ; control CSF, 0.004 nmol/μg of DNA). Our current results also agree in principle with those of Mecocci et al. (1994) and Lyras et al. (1997), which showed increased DNA oxidation in brain in AD. In addition, these findings correlate well with our previous studies showing increased lipid peroxidation and increased levels of HNE, a neurotoxic marker of lipid preoxidation in AD hippocampus and amygdala (Markesbery and Lovell, 1998) and decreased polyunsaturated fatty acid levels, particularly arachidonic and docosahexenoic acids in AD (Svennerholm and Gottfries, 1994 ; Prasad et al., 1998). Also, we have observed increased levels of HNE in AD CSF compared with age-matched control subjects (Lovell et al., 1997). These data, together with our current data, strengthen the hypothesis of increased oxidative stress in the pathogenesis of AD.

To our knowledge, this is the first report of levels of free 8-OHdG in CSF samples from which 8-OHdG levels were also measured in intact DNA. Levels of free 8-OHdG are important in that they reflect the extent of repair of oxidized DNA. The measurement of both free and intact levels of 8-OHdG reflects the rate of damage and balance between damage and repair.

The attack of DNA by reactive oxygen species occurs continuously as a consequence of metabolic and other biochemical reactions (Loft and Poulsen, 1996) and leads to the hydroxylation of DNA bases (Adelman et al., 1988 ; Wagner et al., 1992), the most prominent of which is 8-OHdG. Repair of 8-OHdG in damaged DNA is through a base-specific glycosylase (Chung et al., 1991 a,b ; Tchou et al., 1991), which functions to cleave the altered base allowing subsequent transport to the CSF and blood, and eventual excretion in urine (Shigenaga et al., 1989 ; Fraga et al., 1990). Thus, 8-OHdG has been suggested to serve as an efficient biomarker of oxidative DNA damage in vivo (Shigenaga et al., 1989). The study of free 8-OHdG levels in urine indicates that excretion decreases with age (Loft and Poulsen, 1996). Studies of Mecocci et al. (1994) indicate that 8-OHdG in intact DNA increases with aging. In a study of rats, Fraga et al. (1990) showed that the rate of oxidative damage to DNA decreased with age, along with a decreased rate of metabolism, leading to an increase in steady-state levels of 8-OHdG due to failing repair mechanisms. Oxidative damage to DNA that escapes repair may be a major cause of the age-dependent decline in normal cell function, and a major contributor to aging and other age-related degenerative diseases, such as AD and cancer (Adelman et al., 1988 ; Ames, 1989 ; Fraga et al., 1990). Although the presence of 8-OHdG has been shown to induce GC to GG or At substitutions in DNA replication (Kuchino et al., 1987 ; Shibutani et al., 1991 ; Kamiya et al., 1995), the precise functional significance of increased levels of 8-OHdG in postmitotic cells, such as neurons, is unknown. It is conceivable that these altered bases may interfere with interaction of normal binding of transcription factors. Preliminary experiments in our laboratory suggest that the presence of 8-OHdG in the nuclear factor-κB binding region significantly alters its binding.

There is no information in the literature about 8-OHdG levels in CSF. Due to the variation in data points and overlap between AD and control subjects, elevations of 8-OHdG in DNA or decreases in free 8-OHdG cannot serve as a diagnostic test for AD. However, these data, along with our previous studies demonstrating an increase in HNE and decreased glutathione transferase (an enzyme capable of detoxifying HNE) in ventricular CSF in AD (Lovell et al., 1997), coupled with reports of Montine et al. (1998) of elevations of isoprostane F2 and neuroprostane F4 (markers of lipid peroxidation) in AD CSF, suggest that CSF may have utility in determining oxidative status of the brain, disease progression, or response to therapy in living subjects. Studies of CSF from living patients in various stages of AD, as well as control subjects, are needed to evaluate this possibility.

Overall, this study demonstrating an elevation of 8-OHdG in DNA and a decrease in free 8-OHdG in ventricular CSF in AD that correlate with disease severity not only supports the concept that the AD brain is subject to increased oxidative stress, but also suggests that it may be subjected to deficiencies in essential DNA repair mechanisms.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES

This work was supported by NIH grants 5P50-AG05144 and 1PO1-AG05119 and grants from the Abercrombie Foundation and the Kleburg Foundation. The authors thank Drs. Daron Davis and David Wekstein for CSF procurement, Ms. Jane Meara for technical assistance, Ms. Paula Thomason for editorial assistance, and Mr. Cecil Runyons for providing subject demographic data.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. Ventricular CSF sampling
  5. Isolation of DNA from ventricular CSF
  6. GC/MS-SIM analysis
  7. RESULTS
  8. DISCUSSION
  9. Acknowledgements
  10. REFERENCES