[1] 6w Platelet Membrane Abnormahty in Alzheimer's Disease

[1] 53w The sex ratio of t h e control group (20 m e d l 6 women) did not differ significantly (chi-square) from that of the demented g r o u p (9 m e d l 5 women). In addition, the mean age o f patients in the control group, 63.4 2 10.5

[2] 118w Table I. Fluorescence Data far Platelet Membranes Labeled with DPH Rotation Fluorescence Rate Steady-State Lifetime ( x 10*/lsec) Anisotropy (phase, nsec) Control subjects 1.93 (0.13)" 0.2000 (0.0060)' 8.5 (0.3) Alzheimer's 2.15 (0.24)b 0.1887 (0.0085)" 8.5 (0.2) (n = 36) patients (n = 24) 0.22 J Rotation rates were calculated from steady-state anisotropy and fluorescence lifetime (phase) measurements made at 37.0 -+ 0.1"C. Fluorescence lifetimes were calculated for measurements made with a modulation frequency of 18 MHz. Standard deviations are in parentheses. Statistically significant differences are: a versus h: p = 3.8 x two-tailed t test;p = 8.9 X Mann-Whitney U test. versus ' : p = 1.3 X lo-', two-tailed t cest;p = 1.8 X Mann-Whitney U test.

[3] 150w DPH = 1,6-diphenyl-l,3,5-hexatriene years, SD, was similar to that of the Alzheimer's disease group, 64.5 k 9.8 years. The mean duration of illness for the demented group was 4.9 t 2.2 years and the mean Mini-Mental State score for this group was 15.3 5 4.3. All laboratory analyses were done using coded samples and performed by personnel who were blind to the demographic and clinical description of the subjects. Spectra characteristic of DPH and TMA-DPH were observed in each case [so]. The excitation and emission maxima for the DPH-labeled preparations occurred at 359.5 t 0.5 nm and 427.5 ? 0.5 nm, respectively, while those for TMA-DPH-labeled preparations occurred at 362.0 t 0.0 nm and 427.5 ? 0.5 nm, respectively. Spectra obtained from the labeled platelet membranes prepared from the demented patients did not differ from those prepared from controls. Similar results were obtained with intact platelets and red cell ghosts.

[4] 303w Apparent rotational rates calculated from steadystate and fluorescence lifetime measurements made at 37.0 t 0.1"C (Tables 1 and 2) revealed the mean rotation rate of D P H (Table 1) in platelet membranes from the demented group was 2.15 x 108/sec. This was significantly increased when compared to the respective value for the normal control group, 1.93 X lO*/sec ( p = 3.8 x two-tailed t test;p = 8.9 x Mann-Whitney U test). The increase in the apparent rotation rates of D P H in platelet membranes from the Alzheimer's disease group was due solely to a corresponding decrease in the mean steady-state anisotropy of labeled specimens, while the mean fluorescence lifetime o[ DPH remained unchanged. As shown in Table 1 and Figure 1, the mean steady-state anisotropy of DPH in platelet membranes from the demented group, 0.1887, was significantly decreased when compared to the respective value for the control group, 0.2000 ( p = 1.3 x lo-', two-tailed t test;p = ,-\ In contrast to the findings with DPH, the mean rotation rate of TMA-DPH in platelet membranes from the demented group, 2.05 x 108/sec, did not differ significantly from the respective value for the control group, 2.04 x 108/sec (Table 2). Moreover, the mean steady-state anisotropy and mean fluorescence lifetime (phase, 18 MHz) of TMA-DPH in platelet membranes from the demented group, 0.2503 and 4.7 nsec, respectively, did not differ from the corresponding values for platelet membranes from the control group, 0.2507 and 4.7 nsec, respectively. Since DPH localizes to the hydrocarbon region of lipid membranes while TMA-DPH is anchored to the membrane lipidaqueous interface, these results imply a selective alteration in molecular dynamics at the hydrocarbon region of platelet membranes from patients in the Alzheimer's group that does not appear to extend superficially to affect the lipid-aqueous interface at 37°C.

[5] 157w The observed increase in platelet membrane fluidity associated with the Alzheimer's group, as reflected by a reduction in the steady-state anisotropy of DPHlabeled membranes, was significantly correlated with clinical severity as measured by the Mini-Mental State score. As shown in Figure 2 . / / 14 l7I 0.16 0.17 0.18 0.19 0.20 0.21 IDPH ANISOTROPY Fig 2. ReIationship of steady-state anisotropy (37°C) of 1,6diphenyl-l,3,S-bexatriene (DPHi-labeled platelet membranes from patients with proba6le Alzheimer's disease to dementia severity, as indicated by the Mini-Mental State score. Decreasing anisotropy values and Mini-Mental State score rt$ect increasing membrane juidity and increasing severity of dementia, respectively. The variables were fit to a reciprocal model of the form 1IY = a + bX, where Y = Mini Mentalstate score, X = DPH anisotropy, a = 0.36 + 0.06, SE, and 6 = -1.55 5 0.33. The standard error of this estimate was 0.01. The corresponding correlation coefiicient was -0.73 (p = 1.2 x 10-').

[6] 60w labeled platelet membranes from patients with Alzheimer's disease. Moreover, the mean DPH steadystate anisotropy for labeled platelets from patients with a presenile onset of symptoms (<65 years, n = ll), 0.1894 f 0.0101, SD, was similar to the corresponding mean value for those patients whose onset of symptoms occurred in the senium (265 years, n = 13), 0.1897 ? 0.0109.

[7] 100w While we have previously reported that the steadystate anisotropy of D P H in labeled platelets had a statistically significant positive correlation with the age of healthy subjects during the second to ninth decade 171, the correlation of these variables for the normal control subjects in the present study (r, = +0.31) did not reach statistical significance. This apparent discrepancy most likely resulted from the narrower age range of the normal controls. The mean rotation rate, steadystate anisotropy, and fluorescence lifetime of DPH or TMA-DPH in labeled platelet membranes were not significantly affected by patient gender within any diagnostic group.

[8] 42w We further considered the possibility that the observed increase in platelet membrane fluidity associ- x 4,200; B and D, x .3.3,000. Platelets rontaining an werabundant system of trabeculated cisternae appear in the micrographs of the pathological sample (C and D ) .

[9] 115w ated with Alzheimer's disease resulted from a decrease in the mean age of circulating platelets in this pathological group. Since platelets decrease in volume with increasing cell age [lo, ll}, we determined the mean volume of the platelets isolated from study subjects as a relative index of mean cell age. The mean volume of platelets prepared from the demented group, 4.56 p,m3 t 0.49 km3, SD, did not differ significantly from that of the control group, 4.32 km3 t 0.61 km3. This observation suggests that the increased platelet membrane fluidity observed for the demented group did not result from a reduction in the mean age of circulating platelets compared to the neurologically healthy control group.

[10] 171w To investigate the cellular basis for the observed alteration in platelet membrane fluidity associated with Alzheimer's disease, we used electron microscopy to examine coded platelet preparations from 5 demented patients and 5 neurologically healthy controls. A prominent cytological difference emerged between the diagnostic groups, as shown in Figure 3. The specimens from patients with Alzheimer's disease contained many (approximately 10 to 25%) unusual platelets that exhibited an extensive system of trabeculated cisternae, bounded by ribosome-free membrane. This system did not appear to be contiguous with the plasma membrane in any of the photomicrographs, and therefore seems unlikely to represent an overdevelopment of the canalicular system. Instead, this membrane system appears to lie within the affected platelets. By comparison, cells of this type accounted for only approximately 5% of the total platelets in blood samples from healthy controls. Since internal platelet membranes have been reported to be more fluid (lower DPH anisotropy, 37°C) than external platelet membranes 1261, the ultrastructural difference suggested by these preliminary studies is consistent with the fluorescence data.

[1] 188w These results confirm and extend the findings of our previous preliminary report of a platelet membrane abnormality associated with Alzheimer's disease. An increase in platelet membrane fluidity, as reflected by a significant decrease in the steady-state anisotropy of DPH-labeled membranes, has now been reported for groups of patients with Alzheimer's disease in Boston C43, 471, Pittsburgh, and London f17). Initial assessments of the specificity of this platelet membrane phenomenon are promising. The increased platelet membrane fluidity associated with Alzheimer's disease was not found in platelets from patients with depression (483, a common cause of reversible dementia in the elderly [14,30); mania [44}, which may also be accompanied by a secondary dementia [35}; or multi-infarct dementia [17}. One of the 24 patients with probable Alzheimer's disease has been autopsied since initiation of this study, and the diagnosis of Alzheimer's disease was confirmed. The steady-state anisotropy of DPHlabeled platelets from this individual, 0.1886, was in the aberrant range (described by the 0.1716 cutoff value. Abnormal ccll membrane composition and structure have also been found in brain tissue obtain& at autopsy from patients who died with neuropathologically confirmed Alzheimer's disease [3,7,28,41).

[2] 108w None of several possible sources of artifact considered could account for the observed increase in platelet membrane fluidity associated with Alzheimer's disease. Investigator bias was eliminated by ensuring that laboratory staff remained blind to clinical data and, conversely, that clinicians remained blind to the biophysical data. Fasting blood samples were used to minimize the possible effects of eating, even though we have found no effect on the platelet membrane characteristics measured resulting from the ingestion of a single meal. While the isolation of different subpopulations of platelets from the comparison populations was a potential source of spurious results, platelet yields for the samples exceeded 70% in all cases.

[3] 87w None of the patients were being treated with medications that affect platelet membrane fluidity, including neuroleptics and antidepressants, and none of the demented patients or controls had any history of exposure to such medications. Furthermore, it is unlikely that our findings result from unreported exposure to these medications, based on the effects these agents have on the physical properties of cell membranes. Phenothiazines increase the steady-state anisotropy of DPH-labeled platelet membranes while haloperidol and tricyclic antidepressants have no effect on this parameter at clinically relevant concentrations [40][41][42][43][44][45][46].

[4] 23w Finally, the increase in platelet membrane fluidity associated with Alzheimer's disease does not appear to result from nonspecific effects of chronic illness, since:

[5] 42w (1) the steady-state anisotropy of DPH-labeled platelet membranes from the demented group was not significantly correlated with duration of illness; and (2) none of the patients in the study were suffering from malnutrition or vitamin deficiency syndromes that often accompany chronic debilitation.

[6] 91w If our conclusions ,are valid, the finding of an abnormal platelet membrane fluidity, as reflected by the steady-state anisotropy of DPH at 37"C, has considerable potential both cilinically and as a research tool to study dementia in the elderly. The data support the contention that Alzheimer's disease may be a systemic disorder, rather than ,in isolated brain disease f4, 61. In contrast, our results do not support the hypothesis that Alzheimer's disease results from a pathological acceleration of the normal aging process. We have previously demonstrated that the steady-state anisotropy of

[7] 218w DPH-labeled platelet membranes increases with age in healthy individuals, reflecting an age-dependent decrease in membrane fluidity from the second to ninth decades {9}. Similar age-related changes in membrane fluidity have been reported for human lymphocytes 1321 and hepatic tissue [361, as well as for brain cells in rodents [S, 161. Based on these observations, the accelerated aging hypothesis would predict platelet membranes from patients with Alzheimer's disease to be overly rigid. In contrast, we found that Alzheimer's disease was associated with increased platelet membrane fluidity, as reflected by a reduction in the mean steady-state anisotropy of DPH-labeled platelet membranes. Further assessment of platelet membrane fluidity with DPH in patients with Alzheimer's disease, as well as further investigation of the physiological basis of the abnormality, is clearly warranted. Knowledge of the underlying cause of membrane fluidity changes may serve as a clue to the etiology of Alzheimer's disease. At the same time, the abnormality itself, even in the absence of further knowledge of pathophysiology, may prove useful in the differential diagnosis of dementia. It remains to be seen if abnormal platelet membrane fluidity precedes the onset of symptoms of dementia for patients with Alzheimer's disease. If so, the findings would be extremely useful in genetic and epidemiologic studies of Alzheimer's disease by identifying individuals at risk before symptoms occur.

[1] 121w The study population consisted of 60 subjects who were at least 45 years of age. A complete history and physical examination, including detailed neurological, psychiatric, and mental status assessments, were performed for each of the subjects. All subjects had normal complete blood counts, urinalysis, blood chemistry screens, serum folate and BIZ levels, and thyroid function tests, and none were suffering from malnutrition or vitamin deficiency syndromes, as shown by clinical and laboratory results. Subjects with disorders that affect lipid composition in blood cell membranes or lipid profiles in serum, including alcohol or other substance abuse, generalized atherosclerosis, familial lipidoses, diabetes mellitus, uremia, and nephrosis, as well as subjects who were on restrictive diets, were excluded from the study C2, 18, 331.

[2] 220w The control group consisted of 36 neurologically healthy volunteers between the ages of 45 and 85 years (20 men/ 16 women). The Alzheimer's disease group consisted of 24 outpatients (9 m e d l 5 women) between the ages of 47 and 87 with "probable" Alzheimer's disease, as defined by NINCDS-ADRDA criteria [25]. Individuals who had any history of treatment with neuroleptic or antidepressant drugs, or were currently being treated with any medication that affects platelet membrane fluidity in vitro or in vivo, were excluded from these two groups E42-461. In addition, all patients were required to remain free of any drug ingestion for at least 24 hours before blood was taken. The clinical diagnosis of probable Alzheimer's disease was made conjointly by a neurologist and a psychiatrist on the basis of the insidious onset of dementia with progression bur no other systemic or brain diseases that may cause dementia El, 12, 15, 251. The duration of dementia symptoms was determined to the nearesr year according to history provided by the patient, the patient's family, and the treating physician. The degree of cognitive impairment was graded by the Mini-Mental State score (0, worst; 30, best) [13]. All subjects provided written informed consent before admission to the study. For demented patients, written informed consent was also obrained from a family member.

[3] 72w Cell membranes and inta,ct platelets were diluted to a final optical density of 0.03 at 1600 nm. Ten-milliliter suspensions of each were labeled for 30 minutes at 37°C by the addition of 10 p,l of 1 mM solutions of D P H or TMA-DPH in dimethylformamide [ 2 1, 311, [42][43][44][45][46][47][48][49][50]. Fluorescence measurements were performed on an SLM 4800 spectrofluorometer at a cuvette temperature of 37.0 t O.l"C, as previously described [SO].

[4] 99w Apparent rotation rates were calculated from steady-state anisotropy and fluorescence lifetime measurements (phase, 18 MHz) according to the Perrin equation {37], using 0.335 as the limiting anisotropy for both D P H and TMA-DPH 120, 31, 34, 381. A decrease in steady-state anisotropy in the presence of a constant fluorescence lifetime reflects an increase in probe rotational rate. Because the rotation of D P H and TMA-DPH in biological membranes is anisotropic, the apparent rotational rates are the result of a prominent component reflecting the range of probe motion as well as a component reflecting rate of motion [5,19,20,22,23,241.

[5] 105w Platelets were fixed in pellets at 37°C for 30 minutes with 3% glutaraldehyde in 125 mM sodium cacodylate buffer with 75 mM sucrose at p H 7.3. Samples were rinsed at least 1 hour in several changes of buffer, then postfixed for 1 hour at room temperature with. 1% osmium tetroxide in the same buffer. Samples were dehydrated with ethanol and embedded in Epon araldite. Seccions cut with a diamond knife on a Sorvall Porter-Blum MT-1 ultramicrotome were stained for 15 minutes with 3% aqueous uranyl acetate and 10 minutes in Reynold's lead citrate, then examined at 60 kV in a Philips 300 electron microscope.

[6] 93w For continuous variables comparisons of respective mean values between groups were made using two-tailed t tests. Between-group comparisons of categorical variables were made using the chi-square statistic. Relationships among continuous variables were explored using the parametric Pearson correlation coefficient. All comparisons that revealed statistically significant dififerences according to the t test remained significant when examined by the nonparametric Mann-Whitney u test, Likewise, those pairs of variables for which the pearson corrtllation coefficient was significantly different from zero were also related by a statistically significant Spearman rank correlation coefficient (nonparametric) of the Same sign,

[1] 110w Fluorescence studies of membrane fluidity were conducted double-blind using platelet and red cell membranes prepared from 24 demented patients with probable Alzheimer's disease and 36 neurologically healthy subjects. The fluidity of the hydrocarbon and lipid-aqueous interface regions of cell membranes was determined at 37°C by fluorescence spectroscopy using the lipid probes I,Gdiphenyl-1,3,5-hexatriene (DPH) and l-{4-(trimethylamino)phenyl]-G phenyl-1,3,5-hexatriene (TMA-DPH), respectively. The rotation rate of TMA-DPH in labeled platelet membranes did not differ between the groups. In contrast, the rotation rate of DPH in labeled platelet membranes from the demented patients (2.15 f 0.24 x 10%ec, SD) was greater than that for the normal controls (1.93 +. 0.13 x 108/sec, SD,p = 3.8

[2] 104w This difference was reflected by a reduction in the steady-state anisotropy of DPH in labeled platelets from the demented group (0.1887 f 0.0085, SD) when compared to the respective mean for the controls (0.2000 f 0.0060, SD; p = 1.3 x lo-'). Abnormal membrane fluidity was significantly correlated with severity of dementia, but not with duration of illness or apparent age of onset. The findings do not support the hypothesis that Alzheimer's disease results from a pathological acceleration of the normal aging process, since normal aging is associated with decreased fluidity of cell membranes from platelets, as well as from lymphocytes, hepatocytes, and neurons.

[3] 213w Zubenko GS, Cohen BM, Boller F, Malinakova I, Keefe N, Chojnacki B: Platelet membrane abnormality in Alzheimer's disease. Ann Neurol 22237-244, 1987 Dementia is a major and growing public health problem affecting approximately one in twenty persons over age 65 and one in four over age 80. In the United States, Alzheimer's disease is the most common form of dementia in the elderly 129, 391. The formulation of an adequate solution to this problem is likely to require considerable insight into the etiology and pathophysiology of Alzheimer's disease and dementia in geqeral. A considerable obstacle to pharmacological approaches in the treatment of Alzheimer's disease is that a diagnosis is usually not made until late in the illness, when substantial neurodegeneration has already occurred. Methods that would reveal the pathophysiological processes of Alzheimer's disease in the early stages, before dementia is evident, would substantially improve the likelihood of developing successful therapeutic interventions. Similarly, identifying patients who are in the early stages of Alzheimer's disease would aid in the detection of pathophysiological changes specific to the disorder, rather than indirect, nonspecific signs of nervous system degeneration. Finally, early diagnosis would substantially advance ge-netic and environmental studies that are now hampered because it is not possible to identify affected individuals who do not have clinical symptoms.

[4] 143w The present double-blind studies confirm and extend our reports suggesting increased platelet membrane fluidity is associated with Alzheimer's disease 143, [47][48][49][50]. Platelet and red blood cell preparations were isolated from 24 demented patients with probable Alzheimer's disease and from 36 neurologically healthy controls. Analyses of membrane fluidity were performed at 37°C with the fluorescent probes 1,6diphenyl-1,3,5-hexatriene (DPH) and 1-{4-(trimethylamino)phenyl]-6-phenyl-1,3,5-hexatriene (TMA-DPH), which label the membrane hydrocarbon region and lipid-aqueous interface, respectively 13 1, 331. Together these probes provide information on the physical characteristics of synthetic and biological membranes at deep and superficial regions. The resulting biophysical parameters for corresponding membrane preparations were compared among the drug-free, ageand sex-matched diagnostic groups, and their relationships to demographic variables and clinical severity were examined. We also describe initial investigations o f the cellular basis for the increased platelet membrane fluidity in t h e Alzheimer's disease group.

[5] 17w Platelet membranes were prepared at 4°C by hypotonic lysis, homogenization, and differential centrifugation, as previously described [SO].

[6] 123w Blood drawing and platelet isolation were performed according to a minor modification of the method of Corash and coworkers, which typically results in Platelet Yields Of 95 to 99% [i 11, as previously described [5Of. A 20-ml fasting blood sample was obtained from each subject between 9 AM and 12 noon by antecubital venipuncture through a 19-gauge needle into a plastic syringe containing EDTA as an anticoagulant and protease inhibitor. Platelets were quantitatively isolated by differential centrifugation at 23°C; final yields exceeded 90% in all cases. Platelet counts and mean cell volumes were determined with a Model ZBI Coulter Counter and C: 1000 Channelyzer [so]. Transmission electron micrographs of the platelet preparations revealed less than 0.5% contamination by red blood cells and leukocytes.