Neurobiology of Aging 21 (2000) 271–281 www.elsevier.com/locate/neuaging Hemorheological changes and overproduction of cytokines from immune cells in mild to moderate dementia of the Alzheimer’s type: adverse effects on cerebromicrovascular system. S.B. Solertea,*, G. Ceresinib, E. Ferraria, M. Fioravantia a Department of Internal Medicine, Geriatrics and Gerontology Clinic, School of Geriatrics, University of Pavia, Ospedale S.Margherita, Piazza Borromeo 2, 27100, Pavia, Italy b Department of Internal Medicine and Biomedical Sciences, Section of Geriatrics, University of Parma, Parma, Italy Received 15 September 1999; received in revised form 13 January 2000; accepted 20 January 2000 Abstract An association between hemorheological alterations (i.e., whole-blood and plasma hyperviscosity, reduced erythrocyte deformability, increased red cell aggregation, hyperfibrinogenemia and increased acute-phase protein levels) and the mild stage of senile dementia of the Alzheimer’s type (DAT) was suggested in the present study. In particular, hyperfibrinogenemia and the increase of erytrhocyte aggregation were correlated with the increased generation and release of TNF-a and IFN-g (spontaneous release and IL-2-modulated release) from natural killer (NK) lymphocytes (CD161, CD561, CD3- cells) of patients with DAT; whereas a normal cytokine release from NK cells was found in healthy old subjects and in patients with vascular dementia (VaD). The in vitro and in vivo administration of the hemorheologic drug pentoxifylline (PTX) significantly reduced spontaneous and IL-2-modulated cytokine overproduction from NK cells (in vitro effects with 500 U/ml and 1000 U/ml/NK cells) and improved all the hemorheological parameters. Taken together, these data suggest that disturbances of cerebrovascular flow and of hemorheology could be considered a negative component related to the pathogenesis and progression of DAT neurodegeneration. The association between hemorheological changes and alterations of TNF-a and IFN-g release from NK may indicate a potential immunorheologic mechanism associated with cerebrovascular damage in DAT and could suggest the use of vascular protective drugs as support of the main pharmacological and non-pharmacological therapy of AD. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Alzheimer’s disease; Dementia; Aging; Cerebrovascular disease; Immunity; Natural killer cells; Tumor necrosis factor (TNF)-a; Interferon-g; Interleukin-2 (IL-2); Cytokines; Hemorheology; Acute-phase proteins; Fibrinogen; Pentoxifylline 1. Introduction Several important neuropathological changes found in the brain of patients with Alzheimer’s disease (AD) have been associated with disorders of cerebral blood perfusion and of cerebrovascular hemodynamics and with the presence of cerebromicrovascular pathology [11]. These changes can actively contribute to impair the delivery of metabolic substrates (e.g., glucose) and of oxygen in brain areas affected by AD [28,36,40]. Within this context, risk factors related to cerebral hemodynamic alterations and to cerebrovascular changes should be too early recognized in * Corresponding author. Tel.: 139-0382-27769; fax: 139-038224270. E-mail address: bsolerte@libero.it (S.B. Solerte). an attempt to delay the progression of neurodegeneration and to improve both clinical and prognostic aspects of patients with probable or possible AD. In a wide variety of clinical conditions, hemorheological disorders (i.e., whole-blood, plasma and serum hyperviscosity, reduced erythrocyte deformability, increased red cell aggregation, and hyperfibrinogenemia) can contribute to induce cerebrovascular complications and brain angiopathy by means of direct atherothrombotic mechanisms or by worsening cerebral blood flow and cerebrovascular hemodynamics [7,16,17,35]. In fact, hemorheological derangement can potentially impair micro- and macrovascular blood flow in the brain of AD subjects, generating a passage from laminar to turbulent blood flow in microcapillary segments as well as in arteriolar districts of cerebrovascular system. 0197-4580/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 1 9 7 - 4 5 8 0 ( 0 0 ) 0 0 1 0 5 - 6 272 S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 In our preliminary investigations, we reported specific hemorheological alterations already in elderly in-patients with cerebrovascular disease [58] and blood rheology disorders in old patients with DAT [20]. Concerning hemorheological derangement, our hypothesis suggest that the association of the impaired cerebral blood flow with altered hemorheology in DAT could be mediated by an immune mechanism arising from cytokines produced by mononuclear cells such as natural killer (NK) cells and by monocytes and activated T-lymphocytes. The origin of cerebrovascular disturbances could arise from an abnormal interaction among the endothelial and the astrocytic components of the blood– brain barrier and peripheral immune effectors such as the NK cells. Within this context, a neuroimmune-inflammatory reaction has been suggested in the pathophysiology of AD [14,29,38,50], where a molecular and functional dysregulation of natural killer (NK) cell compartment has been previously reported in patients with severe DAT [51,53,55–57]. In fact, these patients exhibited an increased NK cell cytotoxic activity (CD 161, CD 561, CD3- cell activity) after cytokine-modulation with IL-2, IFN b, and IFN-g, whereas the cytotoxic function (NKCC) was not physiologically suppressed during exposure with glucocorticoids. These changes would seem to be linked to alterations of the transduction system that regulates NKCC in DAT and in particular to protein kinase C-system activity (PKC bII isoform activity) [53]. Alterations of PKC-system activity were also demonstrated in other peripheral non neuronal cells up to characterize PKC and NK cell dysregulation as potential peripheral markers in diagnosing AD [23]. Moreover the increased spontaneous and cytokinemodulated NK cytotoxicity was related with the impairment of cognitive function in patients with moderate to severe clinical stage of DAT, so characterizing the immune dysregulation of NK cells after cytokine modulation as a pathogenetic component involved in the cognitive derangement of AD subjects [53,56,57]. The NK molecular and functional disregulation found in DAT could be also responsible for changes of generation and release of cytokines from these cells and in particular of the inflammatory cytokines TNF-a and IFN-g [31]. This hypothesis may be of a certain pathophysiological interest because these cytokines could induce alterations of vascular and endothelial functions and of blood rheology pattern by increasing plasma fibrinogen levels and whole-blood viscosity [43,61]. In light of these considerations, the aim of the present study was to evaluate: (a) the patterns of TNF-a and IFN-g generation and release from NK cells (spontaneous release and IL-2-modulated release) of old patients with mild to moderate DAT; (b) the hemorheological pattern in patients with DAT and the potential link between NK-derived TNF-a and IFN-g and hemorheology in an attempt to suggest an immunorheologic basis in the pathogenesis of cerebromicrovascular damage; (c) the possibility to antagonize these mechanisms by using the immunorheologic drug (i.e., pentoxifylline: PTX), that could reduce the NK-dependent overproduction of cytokines and hemorheological alterations associated with cerebromicrovascular disorders of patients with DAT. 2. Methods 2.1. Study population Twenty-six healthy subjects (12 of young and 14 of old age), 12 old patients with vascular dementia (VaD), and 12 old patients with dementia of the Alzheimer’s type (DAT) were recruited for the study. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethical Committee of the Department of Internal Medicine of the University of Pavia. Written informed consent was also obtained from all subjects and patients or, where appropriate, from their caregivers. The healthy old subjects fulfilled the strict health criteria of the SENIEUR protocol to exclude clinical and immunological alterations [34]. The diagnosis of DAT and VaD in old patients was conducted in agreement to the criteria of DSM III-R [2] and confirmed for probable DAT according to the diagnostic standards of NINCDS-ADRDA criteria [39] and for VaD by using NINDS-AIREN criteria [48] and the Hachinski iscemic score [24]. Diagnosis of probable or possible DAT was also supported by clinical and neurological evaluations and by brain imaging (MRI or CT scan). Patients with DAT were also analyzed for APOE genotype by means of the DNA genomic extraction of whole blood, amplified by polymerase chain reaction (PCR) [65]: two patients had the e3/e3 type, six had the e3/e4 type, and four had the e4/e4 type. Cognitive status and the severity of dementia were assessed by using the Minimum Mental State Examination test (MMSE) [21]; the score of MMSE ranged from 21 to 27 in patients with DAT ad from 20 to 26 in patients with VaD, therefore, dementia was classified as mild to moderate in both groups. All patients with DAT were free of any medication (for at least 3 weeks) and no diseases (e.g., respiratory and urinary tract infections) known to affect immune function were observed 2 months before the inclusion in the study. 2.2. Study design Hemorheological parameters were evaluated together with the generation and release of TNF-a and IFN-g from NK cells (spontaneous and IL-2 modulated release) in healthy subjects and in patients with DAT and VaD. The xanthine derivative pentoxifylline (3,7-dimethyl-1-[5-oxohexyl]-xanthine; Trental™), was employed for in vitro study (incubation of 500 ml/cells and 1000 ml/cells with NK) and for in vivo study (4 weeks of treatment with 0.4 g orally, 23 daily) in old patients with DAT. S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 2.3. Biochemical determinations Whole-blood was drawn at 8:00 a.m. and anticoagulated with 2Na-EDTA for blood count of total lymphocytes (small cells), hemoglobin and hematocrit; all these parameters were measured by an automatic procedure (Dasit–Ise autoanalyzer and Sysmex Toa F800 Microcell Counter, Dasit, Bareggio, Italy). Circulating serum proteins (albumin, pre-albumin, transferrin, retinol binding protein, a-1 antitrypsin, a-1 acid glycoprotein) and plasma fibrinogen levels were determined by computerized immunonephelometry; specific antisera, calibrators, and controls were used (BNA-Nephelometer; Dade Behring, S.p.A., Milano, Italy). 2.4. Hemorheological measurements The hemorheological study was conducted according to the guidelines of the International Committee for Standardization in Hemorheology [30]. Blood, plasma and serum viscosity were measured with a rotational viscometer (Rotovisco CV100, Rheocontroller RC20, HAAKE AG, Karlsrhue, Germany); measurements, expressed in milliPascal 3 seconds (mPas), were made at shear-rates of 300 (for plasma and serum viscosity), 200 and 1 s (for whole-blood viscosity). The erythrocyte aggregability was derived from the ratio between the measurements of blood viscosity at shear-rates of 1 and 200 s (shear-rate 1 : 200 ratio) [32]. The erythrocyte deformability was measured by a filtration technique of anticoagulated whole blood filtered through polycarbonate filters (Hemafil®, Bio–Rad, Richmond, CA, USA) at 37°C under a negative pressure of 20 cm water [47]; results were expressed as volume of blood per minute (ml/min) without correction for hematocrit. 2.5. Separation of NK cells Immunological procedures were conducted in a completely sterile manner in a class II biological safety cabinet (Microflow 51426, MDH Ltd., Andover, UK). After an overnight fast peripheral blood mononuclear cells (PBMC) were obtained from heparinized venous blood samples (Vacutainer, Hemogard, lithium heparin, Becton Dickinson, Myelan, France) of healthy subjects and elderly patients with VaD and DAT. The separation of NK cells was obtained as previously described [53,59]. In summary, PBMC were isolated by Ficoll–Hypaque density gradient (Lympholyte-H, Cedarlane Laboratories Ltd., Ornby, Ontario, Canada) and plastic adherent cells and B cells were removed by incubation at 37°C in Petri-culture dishes for 1 h. The remaining non-adherent cell population was passed through nylon wool columns preincubated for 1 h with RPMI 1640 supplemented with 10% of heat-inactivated autologous serum (RPMI/AS) at 37°C (% CO2 in air). T/NK cells were obtained by rinsing the columns with tissue culture medium, which leaves B cells and remaining monocytes attached to the nylon wool. As previously reported 273 [53,59], the enriched fraction of PBMC containing T/NK cells was used for the magnetic cell separation procedure (MACS, Miltenyi Biotech GmbH, Bergisch, Gladbach, Germany). The negative unlabeled fraction represents the enriched non magnetic NK cell fraction (CD 161, CD 561, CD 3- cells); whereas the magnetic labeled fraction recognized the non-NK cell fraction (CD31, CD41, CD191, CD331 cells). The efficiency of the immunomagnetic separation was evaluated with flow cytometry by using AntiLeu 11b (anti-CD161) and Anti-Leu 19 (anti-CD561) antibodies purchased from Becton Dickinson and FACScan (Becton Dickinson, Mountain View, CA, USA). The purity of NK cell population was of 97 6 1%, whereas viability was more than 95% (Trypan blue uptake). After magnetic separation, NK cells were washed three times (with 0.9% saline and complete RPMI medium) and finally resuspended in complete RPMI 1640 medium (with 10% of heat inactivated fetal calf serum (FCS), 1% of glutamine and 100 mg/ml of gentamycin) and concentrated to a measured density (Sysmex Toa F800 Microcell Counter, Dasit, Bareggio, Italy) of 7.75 3 106 cells/ml of complete medium. These cells were incubated for 20 h at 37°C, in a humidified atmosphere of 95% air and 5% CO2 (Heraeus incubator BB 6220, Hanau, Germany), without modulators, with IL-2 (100 U/ml/cells of recombinant human IL-2, Proleukin, Chiron Corporation, Emeryville, USA), pentoxifylline (PTX: 500 and 1000 mg/ml/cells of Trental™, Hoechst Marion Roussel, Frankfurt AM, Germany) and PTX (500 and 1000 mg/ml/cells) co-incubated with IL-2 (100 U/ml/ cells) to measure the concentrations of TNF-a and of IFN-g released by NK cells in the supernatant fluids. 2.6. ELISA for TNF-a and IFN-g assay A 500-ml volume of the supernatant fluids of cultured NK cells was centrifuged and 300 ml of volume was frozen at 280°C until the assay of cytokines. The fluid was resuspended at 4°C and analyzed for the TNF-a and IFN-g content by using a standard ELISA procedure and following the manufacture’s instructions (Bender MedSystems Diagnostics GmbH, Wien, Austria). The sensitivity of the method was 0.5 pg/ml for TNF-a (standard curve points at 0, 3.2, 6.4, 12.8, 25.6, 64, 160, 400, 1000 pg/ml) and 0.3 pg/ml for IFN-g (standard curve points at 0, 0.8, 1.5, 3.1, 6.25, 12.5, 25, 50, 100 pg/ml). The intra-assay precision was always below 6%, and the inter-assay precision was within 5% to 8% for both the determinations. 2.7. Statistical analysis One-way ANOVA, F-test was employed to evaluate differences among healthy subjects and old patients with DAT and VaD. Parametric paired Student’s t-test was used to evaluate variations of cytokine release after the incubation of NK cells with IL-2, PTX and PTX1IL-2 (in vitro study) 274 S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 Table 1 Clinical and biochemical characteristics of healthy young and old subjects and of old patients with VaD and DAT n. women : men age, years BMI,a kg/m2 Hb, g/dl total lymphocytes, cells/mm3 blood glucose, mmol/l albumin, g/l transferrin, g/l pre-albumin, g/l retinol-binding protein, g/l Healthy young subjects Healthy old subjects Old patients with VaD Old patients with DAT ANOVA (F test) 12 6:6 32.3 (4) 21.4 (1.6) 14 (1.3) 1992 (55) 4.4 (0.4) 43.3 (3) 3.18 (0.3) 0.33 (0.03) 0.044 (0.008) 14 8:6 77 (6) 22.2 (1.5) 13.7 (1.2) 1949 (64) 4.37 (0.58) 42.7 (2.9) 3.12 (0.3) 0.33 (0.05) 0.042 (0.007) 12 6:6 75.8 (4) 22 (1.7) 13.9 (1.3) 1940 (50) 4.61 (0.5) 42.1 (2) 3.08 (0.5) 0.28 (0.04) 0.042 (0.009) 12 7:5 76.5 (5.5) 21.6 (1.2) 13.5 (1.2) 1953 (56) 4.79 (0.4) 41.9 (2.8) 3.13 (0.4) 0.31 (0.03) 0.040 (0.006) — — — ns ns ns ns ns ns ns ns a BMI 5 body mass index. * data are expressed as mean value (standard deviaiton between brackets). and to determine differences of hemorheological parameters and circulating protein levels before and after 4-weeks treatment with orally PTX (in vivo study). Linear regression analysis and parametric Pearson’s correlation coefficients were also employed to analyze correlations among the spontaneous cytokine release from NK, fibrinogen, erythrocyte aggregability and MMSE score in old patients with DAT and VaD. We accepted p , 0.05 as the threshold of statistical significance (two-tailed). All the analyses were run with the SPSS, Inc./PC1 version 3.0 statistical package (SPSS, Inc., Chicago, IL, USA). 3. Results Table 1 summarizes the clinical and the biochemical characteristics of healthy subjects and of old patients with VaD and DAT. No differences were demonstrated within the groups. In particular, no changes of body weight, of the total number of lymphocytes and of nutritional parameters were found between healthy subjects and patients with dementia. Figs. 1-3 show hemorheological parameters and circulating protein levels in healthy subjects and in patients with Fig. 1. Mean variations (6 SD) of blood, plasma, and serum viscosity in healthy subjects of young and old age (open bars), and in old patients with dementia of Alzheimer’s type (closed bars) in basal conditions (B) and after pentoxifylline in vivo (PTX). *p , 0.01, **p , 0.001 versus healthy young and old subjects; °p , 0.05, °°p , 0.01, °°°p , 0.001 versus B. S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 275 Fig. 2. Mean variations (6 SD) of erythrocyte deformability and erythrocyte aggregability in healthy subjects of young and old age (open bars), and in old patients with dementia of Alzheimer’s type (closed bars) in basal conditions (B) and after pentoxifylline in vivo (PTX). **p , 0.001 versus healthy young and old subjects; °°p , 0.01, °°°p , 0.001 versus B. DAT before and after treatment with PTX (after PTX). Significant differences were found in basal conditions (B) for each parameter between DAT patients and healthy subjects of young and old age. In particular, higher viscosity (whole-blood, plasma and serum), fibrinogen and erytrhocyte aggregability, lower erythrocyte deformability and higher serum a1-antitrypsin and a1-acid glycoprotein levels were found in patients with DAT than in healthy subjects of young and old age. Concerning whole-blood viscosity, increased levels were found at both high shear-rates (200 s, p , 0.01) and low shear-rates (1 s, p , 0.001) in patients with DAT compared to healthy subjects of young and old age. Treatment with PTX (4 weeks of therapy with 0.4 g orally, 23 daily in DAT patients) significantly reduced blood, plasma and serum viscosity (Fig. 1) and erythrocyte aggregability (Fig. 2), whereas erythrocyte deformability was significantly increased after PTX (Fig. 2), reaching the normal values found in healthy subjects. Treatment with PTX also reduced plasma fibrinogen and serum protein levels in patients with DAT (Fig. 3). As concerns the hemorheological parameters and circulating protein levels, no differences were found between patients with VaD and healthy young and old subjects (data not shown). Fig. 4 shows the mean variations of TNF-a levels in the supernatant fluids of NK cells in healthy subjects and in patients with VaD and DAT. Spontaneous and IL-2-modu- Fig. 3. Mean variations (6 SD) of hematocrit, plasma fibrinogen, serum a1 antitrypsin and a1 acid glycoprotein levels in healthy subjects of young and old age (open bars), and in old patients with dementia of Alzheimer’s type (closed bars) in basal conditions (B) and after pentoxifylline in vivo (PTX). *p , 0.01, **p , 0.001 versus healthy young and old subjects; °p , 0.05, °°p , 0.01 versus B. 276 S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 Fig. 4. Mean variations (6 SD) of spontaneous and IL-2-modulated release of TNF-a in the supernatant fluids of NK cells, in healthy old subjects (open bars), and in old patients with vascular dementia (VaD: lightly shaded bars) and with dementia of Alzheimer’s type (DAT: closed bars). Variations measured in DAT patients were evaluated in basal conditions (B), after pentoxifylline in vitro (PTX) and after co-incubation of PTX with IL-2. *p , 0.001 versus healthy subjects and patients with VaD; °p , 0.01, °°p , 0.001 versus B. lated release of TNF-a were similar in healthy subjects and patients with VaD. On the contrary, higher TNF-a levels were found in spontaneous conditions (B) and after IL-2 exposure in patients with DAT than in the other two groups of subjects (p , 0.001 for both spontaneous and IL-2 mediated release). The incubation of NK cells with PTX significantly reduced TNF-a levels in spontaneous conditions and after incubation with IL-2. Fig. 5 shows the mean variations of IFN-g levels in the supernatant fluids of NK cells in healthy subjects and in patients with VaD and DAT. Spontaneous and IL-2-modulated release of IFN-g were similar in healthy subjects and patients with VaD, whereas higher IFN-g levels were demonstrated in spontaneous conditions (B) and after IL-2 exposure in DAT patients than in healthy subjects and patients with VaD (p , 0.001 for both spontaneous and IL-2-mod- ulated release). The incubation of NK cells with PTX significantly reduced IFN-g levels in spontaneous conditions and after IL-2 modulation. It is important to observe that the reduction of TNF-a and IFN-g released from NK cells after incubation with PTX was dose dependent (from 500 to 1000 mg/ml/cells) and that in spontaneous conditions the suppression of TNF-a was more pronounced than the suppression of IFN-g. Fig. 6 shows the linear regression analysis and the correlations among the spontaneous generation of TNF-a and of IFN-g from NK cells, fibrinogen and erythrocyte aggregability in patients with DAT. Significant positive correlations were found among TNF-a and IFN-g levels in the supernatants fluids of NK cells and the increase of plasma fibrinogen concentrations and of erythrocyte aggregability (p , 0.001 for TNF-a and p , 0.01 for IFN-g). On the Fig. 5. Mean variations (6 SD) of spontaneous and IL-2-modulated release of IFN-g in the supernatant fluids of NK cells, in healthy old subjects (open bars), and in old patients with vascular dementia (VaD: lightly shaded bars) and with dementia of Alzheimer’s type (DAT: closed bars). Variations measured in DAT patients were evaluated in basal conditions (B), after pentoxifylline in vitro (PTX) and after co-incubation of PTX with IL-2. *p , 0.001 versus healthy subjects and patients with VaD; °p , 0.01, °°p , 0.001 versus B. S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 277 Fig. 6. Linear regression analysis and correlations among the spontaneous release of TNF-a and of IFN-g in the supernatants of NK cells, plasma fibrinogen and erythrocyte aggregability (E.A.) in old patients with dementia of the Alzheimer’s type. BV 5 blood viscosity. contrary, no significant correlations were demonstrated among these parameters in patients with VaD (data not shown). Finally, significant negative correlations between the spontaneous cytokine release from NK cells and MMSE score (y 5 33.6 – 0.4x, r 5 20.89, p , 0.001 for TNF-a and y 5 30.4 – 0.2x, r 5 20.64, p , 0.05 for IFN-g) and between hemorheological parameters and MMSE score (y 5 151– 0.8x, r 5 20.77, p , 0.01 for plasma fibrinogen and y 5 2.1– 0.06x, r 5 20.79 p , 0.01 for erythrocyte aggregability) were found in patients with DAT. 4. Discussion A growing body of information has pointed to a role of vascular factors, that influence cerebral blood flow and cerebrovascular hemodynamics (at both microcirculatory and macrocirculatory levels), in the pathogenesis of the neurodegenerative mechanisms of AD [11,26,27,44,60]. In fact, common physiopathological elements involving the cerebrovascular system can be demonstrable between AD and vascular dementia (VaD). On the other hand, AD may represent a risk for stroke and for the overall cerebrovascular mortality [18,22]. Within this context, the demonstration of important hemorheological alterations in DAT could contribute to explain the abnormalities of cerebral blood flow (CBF) and of the functional activity of brain capillary network brain [11]. The worsening of CBF and of morphology of cerebral microvessels could be suggested to have a role on alterations of neuronal homeostatic functions also determining the loss of blood-brain barrier selectivity and the activation of microglia [8,32,45,46]. The possibility of a hemorheological-dependent mechanism leading to impair CBF and cerebromicrovascular system could therefore introduce new relevant aspects to understand the development of pathogenetic inflammatory processes linked to AD. Our study has demonstrated hemorheological disorders and increased acute-phase protein levels in patients with mild to moderate DAT. Because hemorheology could exert a strong influence on the integrity of cerebrovascualr system several important consequences may be expected in brain microcirculation of patients with DAT. On the other hand, we can also suggest that hemorheological abnormalities could reduce CBF, particularly in ischemic areas of brain with low perfusion pressures and disturbed autoregulation. Blood hyperviscosity at low shear-rate (1 s) may induce a non-Newtonian behavior of blood, increasing erythrocyte aggregation and slowing down the CBF in capillary segments and venous post-capillary districts [49]. The persistence of low CBF may lead to a transfer of platelets from the axial region of flow toward vessel wall, increasing the risk of endothelial damage and of microthrombotic occlusions of brain microvessels [15]. In addition, whole-blood and plasma hyperviscosity at high shear-rates (200 and 300 s) and hyperfibrinogenemia may increase flow turbulence and stagnation in the arteriolar side of cerebral circulation, inducing alterations of structural viscosity, of thixotrophy and of the viscoelastic properties of blood [16,17,35,58]. On the other hand, changes of circulating fibrinogen, a1-antitrypsin, a1-acid glycoprotein levels (acute-phase reactants) and of serum viscosity can also contribute to worse ischemic damage and hemorheological pattern and to promote an extension of microvascular angiopathy in AD brain. As suggested in the purpose of the study, the hemorheo- 278 S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 logical derangement found in DAT might potentially be linked to immunomediated factors (the “immunorheologic hypothesis”) involving NK cell compartment as well as monocytes and activated T-cells. Our previous investigations have demonstrated an abnormal activation of NK cells and a dysregulation of the protein kinase C (PKC)-dependent mechanism involved in the regulation of NK cytolytic function in old patients with severe DAT [51,53,55–57]. PKC-system activity regulates NK cytotoxic response (NKCC) in spontaneous conditions and after exposure with cytokines (e.g., IL-2, IFN-b and IFN-g). The impairment of PKC downregulation found in NK has been also demonstrated in other cells, up to suggest a common molecular defect involving PKC system in the pathophysiological characterization of AD [23]. An increased NK cytotoxic reactivity to cytokines was found in patients with DAT, whereas these cells were less responsive to the immunosuppression induced by cortisol. The abnormal PKC-dependent regulation of NK cell function in AD could have also consequences on generation and release of inflammatory cytokines from these cells and in particular of TNF-a and IFN-g. In fact, activation of NK cell compartment has been recognized to trigger TNF-a and IFN-g production and to increase the release of cytokines with important implications on inflammation and on regulation of the antigen-independent cellular immune response [9,31,64]. This hypothesis would seem to be confirmed in the present study, since an increased production of TNF-a and IFN-g from pure NK cell population has been demonstrated either in basal conditions (spontaneous NKCC without exposure with cytokines) and after activation with IL-2 (cytokine-mediated NKCC). Besides to support the neuroimmune mechanism of AD neurodegeneration [37,42], these immunological changes could directly contribute to alter hemorheology in AD subjects. In fact, the spontaneous release of TNF-a and IFN-g from NK cells significantly correlated with the increase of plasma fibrinogen levels and of erythrocyte aggregability in patients with DAT. These hemorheological factors are considered two main determinants of structural viscosity and can be related to disorders of hemorheology and of blood perfusion in brain microvascular system [1,15–17,35,49,52,58]. The immunorheologic mechanism associated with cerebrovascular damage could therefore represent a crucial event in the progression of AD neuropathology. Furthermore, the abnormal release of TNF-a and IFN-g from NK cells could directly contribute to the progression of brain vascular disorders of AD subjects. In fact, these proinflammatory cytokines can alter brain vascular network promoting changes of blood– brain barrier (BBB) integrity [8,19,25,42] and the development of focal cerebral ischemia [10]. Further on, the cerebrovascular disturbances linked to BBB alterations could worse the already existing inflammatory reaction in the brain, so sustaining the immuno-vascular component of disease also related to the trans-endothelial migration across BBB of peripheral immune cells into the brain parenchyma. The increased traffic of activated NK cells releasing TNF-a and IFN-g, as well as of other immune cells, into the central nervous system can enhance the risk to develop neuropathological lesions and to induce amyloidogenesis in neuronal cells [6], finally decreasing cognitive functions of AD subjects. The correlations between TNF-a and IFN-g spontaneous release from NK cells and the reduction of cognitive pattern found in the present study would seem to be in agreement with this hypothesis. In summary, our concept incorporates the immunorheologic hypothesis of AD neuropathology (Fig. 7). This concept suggests that AD neurodegeneration could be in part dependent on abnormal cytokine generation and release from mononuclear NK (CD32, CD161, CD561) immune cells leading to hemorheological and cerebrovascular disturbances into AD brain. In combination with other well known vascular factors, changes of hemorheology may be of a certain importance in giving rise to local areas of brain tissue hypoxia finally increasing the risk of a neuroinflammatory reaction against the cellular components of AD brain. Otherwise, we cannot exclude the possibility that activated immune cells and neuroinflammatory reaction could be dependent on vascular alterations related to brain ischemic disorders and CBF abnormalities. The immunorheologic mechanism associated with cerebrovascular disorders and neurodegeneration may provide new perspectives in the clinical approach of AD subjects. In fact, taken together all these data can suggest the use of the methylxanthine derivative PTX in an attempt to antagonize TNF-a and IFN-g release from NK and to improve hemorheology in the cerebrovascular system of patients with DAT. The drug may act at the two levels in a separate manner, either by suppressing the spontaneous and IL-2modulated release of cytokines from NK [31,33,43,54,61] or by improving blood rheology pattern by means of specific effects on fibrinogen levels, blood viscosity and red cell deformability [4,41,52]. Our study demonstrated that in vitro utilization of PTX significantly normalized the TNF-a and IFN-g release from NK both in spontaneous conditions and during modulation with IL-2, so reducing the potential adverse effects of NK immunological hyperactivity on AD progression. On the one hand, the immunosuppressive effect of PTX on cytokine generation may contribute to antagonize the immunorheologic mechanism leading to the progression of cerebrovascular disorders. On the other hand, in vivo treatment with PTX (4 weeks with 0.4 g orally, 23 daily) may improve hemorheological parameters also decreasing circulating acute-phase protein levels in subjects with DAT. This combined immuno-hemorheologic effect of PTX could reduce the impact of altered hemorheology and immunity on CBF and on the progression of hypoxic events in brain areas affected by AD. These evidences are also in agreement with experimental and clinica studies that have demonstrated improved CBF and brain metabolism after PTX administration [62,63]. We can summarize that an immunorheologic mechanism S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 279 Fig. 7. Schematic representation of the immunorheologic hypothesis that could link the immunological and vascular factors with the pathogenesis and progression of AD neurodegeneration. leading to cerebromicrovascular damage could be present in mild to moderate DAT and hence in the early clinical stage of the disease. Both hemorheological and NK functional disorders can alter cerebral circulation and cerebrovascular hemodynamics, so determining vascular-dependent neurotoxic effects against AD brain [11,12,13]. The link between cerebrovascular risk factors and neuroautoimmune-inflammatory reaction [50] could indicate the potential benefit of drugs with both immunorheological and vascular effects, as the xanthine related molecules pentoxifylline and propentofylline, to support the main pharmacological and non-pharmacological treatment of AD and to antagonize or prevent the autodestructive process working in the brain. Acknowledgments We wish to thank Dr. Luca Cravello for excellent statistic and graphic supports, Drs. Gianni Cuzzoni and Riccardo Pagliaro for the clinical recruitment in the Alzheimer’s Unit of our Department and Drs. Silvia Severgnini and Nadia Cerutti for the technical assistance with the immunological procedures. The work was fully supported by two different grants approved by the University of Pavia: 1) Progetto di Ricerca Ateneo, anno finanziario 1999; 2) F.A.R.-Comitato 6, anno finanziario 1998/1999 related to Dr. Marisa Fioravanti. References [1] Ajmani RS, Rifkind JM. Hemorheology changes during human aging. Gerontology 1998;44:111–20. [2] American Psychiatric Association. Diagnostic and Statistical Manual fo Mental Disorders, 3rd edn, revised. Washington DC: American Psychiatric Association, 1987. [3] Anegon I, Cuturi MC, Trinchieri G, Perussia B. Interaction of Fc receptor (CD16) with ligands induces transcription of IL-2 receptor 280 [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 (CD25) and lymphokine genes and expression of their products in human natural killer cells. J Exp Med 1988;167:452–72. Aviado DM, Dettelbach HR. Pharmacology of pentoxifylline, a hemorheologic agent for the treatment of intermittent claudication. Angiology 1984;35:407–17. Belayev L, Busto R, Zhao W, Ginsberg MD. Quantitative evaluation of blood-brain barrier permeability following middle cerebral artery occlusion in rats. Brain Res 1996;739:88 –96. Blasko I, Marx F, Steiner F, Hartman T, Grubeck–Loebenstein B. TNFa plus IFNg induce the production of Alzheimer b-amyloid peptides and decrease the secretion of APPs. FASEB J 1999;13:63– 8. Brown MM, Marshall J. Regulation of cerebral blood flow in response to changes in blood viscosity. Lancet 1985;i:604 –9. Buèe L, Hof PR, Delacourte A. Brain microvascular changes in Alzheimer’s disease and other dementia’s. Ann NY Acad Sci 1997; 826:7–24. Cuturi MC, Anegon I, Sherman F, et al. Production of hematopoietic colony-stimulating factor by human natural killer cells. J Exp Med 1989;169:569 – 83. Dawson DA, Martin D, Hallenbeck JM. Inhibition of TNF-alpha reduces focal cerebral ischemic injury in the spontaneously hypertensive rat. Neurosci Lett 1996;218:41– 4. de la Torre JC. Cerebromicrovascular pathology in Alzheimer’s disease compared to normal aging. Gerontology 1997;43:26 – 43. de la Torre JC. Impaired brain microcirculation may trigger Alzheimer’s disease. Neurosci Behav Res 1994;18:397– 401. de la Torre JC, Mussivand T. Can disturbed brain microcirculation cause Alzheimer’ disease? Neurol Res 1993;15:146 –53. Dickinson DW, Rogers J. Neuroimmunology of Alzheimer’s disease: a conference report. Neurobiol Aging 1992;13:793–98. Dintenfass L. Viscosity factors in hypertensive and cardiovascular disease. Cardiovasc Med 1977;2:337–57. Dintenfass L. Some rheological factors in pathogenesis of thrombosis. Lancet 1965;ii:370 –71. Dormandy JA. Haemorheological aspects of thrombosis. Br J Haematol 1980;45:519 –22. Ferrucci L, Guralnik JM, Salive ME, et al. Cognitive impairment and risk of stroke in the older population. J Am Geriatr Soc 1996;44:237– 41. Fiala M, Zhang L, Gan X, et al. Amyloid-b induces chemokine secretion and monocyte migration across a blood– brain barrier model. Mol Med 1998;4:480 – 89. Fioravanti M, Ricciardi T, Cottinelli M, et al. Hemorheologic alterations and acute-phase reaction are related to recently-onset patients with senile dementia of the Alzheimer’s type (SDAT). Neurobiol Aging 1998;19:S246 –S247. Folstein M, Folstein S, McHugh PR. Mini Mental State. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189 –98. Gale CR, Martyn CN, Cooper C. Cognitive impairment and mortality in a cohort of elderly people. Br Med J 1996;ii:608 –11. Gasparini L, Racchi M, Binetti G, et al. Peripheral markers in testing pathophysiological hypotheses and diagnosing Alzheimer’s disease. FASEB J 1998;12:17–34. Hachinski VC, Lassen NA, Marshall J. Multiinfarct dementia: a cause of mental deterioration in the elderly. Lancet 1974;ii:207–10. Hardy JA, Mann DMA, Wester P, Winblad B. An integrative hypothesis concerning the pathogenesis and progression of Alzheimer’s disease. Neurobiol Aging 1986;7:489 –502. Henderson AS. The risk factors for AD: a review and a hypothesis. Acta Psychiatr Scand 1998;78:257–75. Hofman A, Breteler MMB, Bots ML, et al. Atherosclerosis, apolipoprotein E, and prevalence of dementia and Alzheimer’s disease. The Rotterdam Study. Lancet 1997;i:151– 4. Hoyer, S. Abnormalities of glucose metabolism in Alzheimer’s disease. Ann NY Acad Sci 1991;640:53– 8. [29] Huberman M, Shalit F, Roth–Deri I, et al. Correlation of cytokine secretion by mononuclear cells of Alzheimer patients and their disease stage. J Neuroimmunol 1994;52:147–52. [30] International Committee for Standardization in Hematology. Guidelines for measurement of blood viscosity and erythrocyte deformability. Clin Hemorheol 1986;6:439 –53. [31] Jewett A, Bonavida B. Pivotal role of endogenous TNF-a in the IL-2 driven activation and proliferation of the functionally immune NK “free” subset. Cell Immunol 1993;151:257– 63. [32] Kalaria RN, Harik SI. Reduced glucose transporter at the blood– brain barrier in cerebral cortex in Alzheimer’s disease. J Neurochem 1989; 53:1083– 88. [33] Liang L, Beshay E, Prud’homme GJ. The phosphodiesterase inhibitors pentoxifylline and rolipram prevent diabetes in NOD mice. Diabetes 1998;47:570 –5. [34] Lighthart GJ, Corberand JX, Fournier C, et al. Admission criteria for immunogerontological studies in man: the SENIEUR protocol. Mech Ageing Dev 1984;28:47–55. [35] Lowe GDO, Forbes CD. Blood rheology and thrombosis. Clin Haematol 1981;10:343– 67. [36] Marcus DL, de Leon M, Goldman J, et al. Altered glucose metabolism in microvessels from patients with Alzheimer’s disease. Ann Neurol 1989;26:91– 4. [37] McGeer PL, McGeer EG. The inflammatory response system of the brain: implications for therapy of Alzheimer and other neurodegenerative diseases. Brain Res Rev 1995;21:195–218. [38] McGeer PL, Rogers J, McGeer EG. Neuroimmune mechanisms in Alzheimer’s disease pathogenesis. Alzheimer Dis Assoc Disord 1994; 8:149 –58. [39] Mckhann G, Drachman D, Folstein M, Katzmann R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Workgroup under the auspices of Department of Healthy and Human Services Task Force on Alzheimer’s disease. Neurology 1984;34:939 – 44. [40] Meier-Ruge W, Bertoni–Freddari C, Iwangoff P. Changes in brain glucose metabolism as a key to the pathogenesis of Alzheimer’s disease. Gerontology 1994;40:246 –52. [41] Muller R. Hemorheology and peripheral vascular diseases: a new therapeutic approach. J Med 1981;12:209 –36. [42] Munoz–Fernandez MA, Fresno M. The role of TNF, interleukin 6, interferon-g and inducible nitric oxide synthase in the development and pathology of the nervous system. Progr Neurobiol 1998;56:307– 40. [43] Munro JM, Cotran RS. The pathogenesis of atherosclerosis: atherogenesis and inflammation. Lab Invest 1988;58:249 – 61. [44] Ott A, Breteler MMB, de Bruyne MC, Van Harskamp F, Grobbee DE. Hofman A. Atrial fibrillation and dementia in a population-based study. The Rotterdam Study. Stroke 1997;28:316 –21. [45] Petito CK. Early and late mechanisms of increased vascular permeability following experimental cerebral infarction. J Neuropathol Exp Neurol 1979;38:222–34. [46] Platell M, Teisser E, Cecchelli R. Hypoxia dramatically increases the non specific transport of blood-borne proteins to the brain. J Neurochem 1997;68:874 –77. [47] Reid H, Barnes AJ, Lock PI, Dormandy JA. A simple method for measuring erythrocyte deformability. J Clin Pathol 1976;29:855–58. [48] Roman GC, Tatemichi TK, Erkinjuntti T, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 1993;43:250 – 60. [49] Schmid Schonbein H-H. Microrheology of erythrocytes, blood viscosity and the distribution of blood flow in the microcirculation. Int Rev Physiol 1976;9:1– 62. [50] Singh VK. Neuroautoimmunity: pathogenetic implications for Alzheimer’s disease. Gerontology 1997;43:79 –94. [51] Solerte SB, Cerutti N, Severgnini S, Rondanelli M, Ferrari E, Fioravanti M. Decreased immunosuppressive effect of cortisol on natural S.B. Solerte et al. / Neurobiology of Aging 21 (2000) 271–281 [52] [53] [54] [55] [56] [57] killer cytotoxic activity in senile dementia of the Alzheimer’s type. Dement Geriatr Cogn Disord 1998;9:149 –56. Solerte SB, Fioravanti M, Cerutti N, et al. Retrospective analysis of long-term hemorheologic effects of pentoxifylline in diabetic patients with angiopathic complications. Acta Diabetol 1997;34:67–74. Solerte SB, Fioravanti M, Pascale A, Ferrari E, Govoni S, Battaini F. Increased natural killer cell cytotoxicity in Alzheimer’s disease may involve protein kinase C dysregulation. Neurobiol Aging 1998;19: 191–9. Solerte SB, Fioravanti M, Rondanelli M, Ferrari, E. Pentoxifylline reduces the spontaneous and IL-2 mediated reactivity of natural killer cell compartment in senile dementia of the Alzheimer’s type (SDAT). Neurobiol Aging 1998;19:S178 –S179. Solerte SB, Fioravanti M, Severgnini S, Cerutti N, Locatelli M, Ferrari, E. Variability of natural killer (NK) cell immune function in normal aging and senile dementia: pathophysiological implications. Aging Clin Exp Res 1997;9:32–3. Solerte SB, Fioravanti M, Severgnini S, et al. Excitatory pattern of g-interferon on natural killer cell activity in senile dementia of the Alzheimer’s type. Dement Geriatr Cogn Disord 1997;8:308 –13. Solerte SB, Fioravanti M, Severgnini S, et al. Enhanced cytotoxic response of natural killer cells to IL-2 in Alzheimer’s disease. Dementia 1996;7:343– 48. 281 [58] Solerte SB, Locatelli M, Pezza N, et al. Haemorheology and cerebrovascular disease in aging. Retrospective analysis of associated risk factors. Facts Res Gerontol 1996;S:7–22. [59] Solerte SB, Fioravanti M, Schifino N, et al. Dehydroepiandrosterone sulfate decreases the IL-2-mediated overactivity of the natural killer cell compartment in senile dementia of the Alzheimer type. Dement Geriatr Cogn Disord 1999;10:21–7. [60] Stewart R. Cardiovascular factors in Alzheimer’s disease. J Neurol Neurosurg Psychiatry 1998;65:143–7. [61] Strieter RM, Remick DG, Ward PA, Splenger RN. Cellular and molecular regulation of TNF-alpha production by pentoxifylline. Biomed Biochem Res Commun 1988;155:1230 –36. [62] Trigoe R, Hayashi T, Anegawa S, et al. Effect of propentofylline and pentoxifylline on cerebral blood flow using 133I-IMP SPECT in patients with cerebral arteriosclerosis. Clin Ther 1994;16:65–73. [63] Toung TJ, Kirsch JR, Maruki Y, Traystman RJ. Effects of pentoxifylline on cerebral blood flow, metabolism and evoked response after total cerebral ischemia in dogs. Crit Care Med 1994;22:273– 81. [64] Trinchieri G. Biology of natural killer cells. Adv Immunol 1989;47: 187–386. [65] Wenham PR, Price WH, Blundell G. Apolipoprotein E genotyping by one-stage PCR. Lancet 1991;ii:1158 –9.