Amyloid β Levels in Human Red Blood Cells

Amyloid β-peptide (Aβ) is hypothesized to play a key role by oxidatively impairing the capacity of red blood cells (RBCs) to deliver oxygen to the brain. These processes are implicated in the pathogenesis of Alzheimer's disease (AD). Although plasma Aβ has been investigated thoroughly, the presence and distribution of Aβ in human RBCs are still unclear. In this study, we quantitated Aβ40 and Aβ42 in human RBCs with ELISA assays, and provided evidence that significant amounts of Aβ could be detected in RBCs and that the RBC Aβ levels increased with aging. The RBC Aβ levels increased with aging. On the other hand, providing an antioxidant supplement (astaxanthin, a polar carotenoid) to humans was found to decrease RBC Aβ as well as oxidative stress marker levels. These results suggest that plasma Aβ40 and Aβ42 bind to RBCs (possibly with aging), implying a pathogenic role of RBC Aβ. Moreover, the data indicate that RBC Aβ40 and Aβ42 may constitute biomarkers of AD. As a preventive strategy, therapeutic application of astaxanthin as an Aβ-lowering agent in RBCs could be considered as a possible anti-dementia agent. Trial Registration Controlled-Trials.com ISRCTN42483402


Introduction
Alzheimer's disease (AD) is the most common form of dementia. Since AD is associated with the progressive accumulation of amyloid b-peptide (Ab) in the human brain, a pathogenic role of Ab in the brain has been widely recognized [1,2].
Over the last decade or so, the presence of Amyloid b-peptide (Ab) in peripheral blood plasma has received increasing attention [3][4][5][6], and plasma Ab is hypothesized to readily contact red blood cells (RBCs) and impair the capacity of RBCs in circulating human blood [7,8]. Our group and other researchers have investigated the hypothesis, and found that Ab induces oxidative injury to RBC by binding to them, and causing accumulation of phospholipid hydroperoxides (PLOOH), a specific marker for RBC membrane oxidative injury [9,10]. Ab also induces the binding of erythrocytes to endothelial cells and decreases endothelial viability, perhaps by the generation of oxidative and inflammatory stress [11]. These studies [9][10][11] provide a possibility that Ab plays a key role in blood and oxidatively impairs RBC function (e.g., oxygen delivery to the brain), thereby potentially facilitating AD. However, to the best of our knowledge, no extensive study of the presence and distribution of Ab in human RBC has been undertaken.
The aim of this study was to ascertain the distribution of Ab in the RBCs of young and senior subjects by applying a commercial ELISA assay. The RBC Ab concentrations were compared between young and senior subjects and also compared to plasma Ab levels. In addition, we previously conducted a randomized, double-blind, placebo-controlled human study to evaluate whether nutritional supplementation with the antioxidant astaxanthin (a polar carotenoid) affected RBC PLOOH [12]. Thus, RBCs that had been obtained from the human study [12] were subjected to Ab determination in order to evaluate the relationship between RBC Ab and the antioxidant/oxidant profile.

Ethics Statement
The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Anti-Aging Science (Tokyo, Japan; ethics No. I030807). All of the subjects provided written informed consent to the experimental protocol before participating in the study.

Blood Samples from Young and Senior Volunteers
Twenty-four young healthy human volunteers [12 men and 12 women, between 22 and 29 years of age (mean 6 SE, 24.260.6)] and 38 senior healthy volunteers [20 men and 18 women, between 48 and 69 years of age (mean 6 SE, 56.261.0)] participated in this study. Blood was collected into a tube containing EDTA-2Na as an anticoagulant and was subjected to centrifugation at 1,000 g for ten min at 4uC. After the plasma and buffy coat were removed, RBCs were washed three times with phosphate buffered saline (PBS, pH 7.4) to prepare packed cells.

Measurement of Ab40 and Ab42 in RBCs and Plasma
For determination of Ab40 and Ab42 in RBCs, human b Amyloid (1-40) ELISA kits (WAKO, Osaka, Japan) and human b Amyloid (1-42) ELISA kits (WAKO) were used, respectively. These kits are commercially available and used worldwide. We tested conditions for measurement of RBC Ab, and the optimized protocol is as follows. Packed cells (200 mL) were mixed with 200 mL of water and one mL of 70% formic acid. A 40 mL aliquot was collected and mixed with 760 mL of 1 mol/L Tris-HCl with protease inhibitors, and the mixture was diluted two-fold with the standard diluent present in each Ab40 and Ab42 ELISA kit. A 100 mL aliquot was subjected to either the Ab40 or the Ab42 ELISA kit (in triplicate). Briefly, aliquots (100 mL) were transferred to BAN50-coated (specific for the N-terminal portion of human Ab1-16) 96 well microplates and incubated overnight at 4uC. After washing five times with the kit's wash solution, HRP-conjugated BA27 (specific for the C-terminal portion of human Ab40) or HRP-conjugated BC05 (specific for the C-terminal portion of human Ab42) was added and incubated at 4uC for one h. After washing in the same manner, TMB solution from the ELISA kit was added and incubated for 30 min. Stop solution in each Ab40 and Ab42 ELISA kit was added and absorbance was read at 450 nm using a microplate reader (GENios, TECAN Co., Ltd.). For plasma Ab40 and Ab42, these were determined according to the ELISA kit protocol.

Previous Human Study Protocol
As described above, we previously conducted a randomized, double-blind, placebo controlled human study to evaluate whether antioxidant supplementation (astaxanthin) affected RBC phospholipid peroxidation [12]. In this study, blood samples (RBC and plasma) that had been obtained from the human study [12] were subjected to Ab determination. That previous human study protocol was organized as follows. A total of 30 healthy subjects (15 men and 15 women), between 50 and 69 years of age (mean 6 SE, 56.361.0), randomly received zero mg (placebo), six mg, or 12 mg antioxidant (astaxanthin, PurestaH, Yamaha Motor Co., Ltd.; Shizuoka, Japan). During the 12-week trial, subjects ingested one of the three astaxanthin doses (zero, six, or 12 mg) capsules with an appropriate amount of water once daily after breakfast. Before and after the supplementation period (weeks zero and 12), blood samples were collected from the subjects. From the blood samples, RBC and plasma were prepared, and their PLOOH and antioxidants (carotenoids and tocopherols) were measured by HPLC techniques [9,[12][13][14][15][16].

Statistical Analyses
Data are presented as means 6 SE. Differences in Ab concentrations between young and senior subjects were compared using Student's t-test or Welch's t-test for equal or unequal variances; the Mann-Whitney U test was used when the distribution was skewed. For correlation analysis, Pearson's correlation coefficient test for normal data or Spearman's rank correlation coefficient test for nonparametric data was used. A difference was considered significant at P,0.05.

RBC Ab in Young Human Volunteers
The ELISA assay kit is designed to be used for the quantitative determination of Ab40 and Ab42 in human fluid samples including plasma, CSF, tissue homogenate, and tissue culture media [3,[17][18][19]. The ELISA kit was developed by Suzuki et al. and shows extremely highly sensitivity and reproducibility [20]. Using the kit, we found that Ab could be detected in RBC packed cells. Therefore, conditions for the measurement of RBC Ab40 and Ab42 were optimized, and Ab40 and Ab42 levels in RBCs of young healthy volunteers were determined. As a result, RBC Ab concentrations were calculated to be 5.3260.21 pmol/g hemoglobin for Ab40 and 2.0960.06 pmol/g hemoglobin for Ab42 (Table 1). If Ab levels in RBC were compared to plasma, Ab40 and Ab42 levels in RBC were about 8 and 14 times higher than those of plasma, respectively. The experiments also confirmed that the RBC Ab42/Ab40 ratio was about 1.8 times higher than plasma ( Table 1).

Comparison of RBC Ab Concentrations between Young and Senior Subjects
It was found that the concentrations of RBC Ab40 and Ab42 in senior healthy volunteers were significantly higher than those of young healthy volunteers ( Table 1). The plasma Ab40 and Ab42 in elderly subjects were also higher than those of young volunteers. When we analyzed the relationship between Ab in RBC and plasma, significant positive correlations were found (Fig. 1). On the other hand, RBC PLOOH levels in senior subjects were also higher than those of young volunteers.

Astaxanthin Antioxidant Supplementation Trial: RBC Ab and the PLOOH Profile
As already reported in our former study [12], after human volunteers (senior subjects) received supplementation with the antioxidant astaxanthin, RBC astaxanthin concentrations were significantly increased and the RBC PLOOH concentration decreased. For other parameters, astaxanthin supplementation showed a safety profile with no side effects (e.g., death of RBC, leukocytosis, and inflammatory conditions) [12]. In the present study, we measured RBC Ab40 and Ab42 concentrations, and these levels were found to be significantly decreased after supplementation (Fig. 1A, Table 2). In addition, there were inverse relationships between RBC Ab and astaxanthin concentrations (Fig. 2). On the other hand, astaxanthin supplementation did not affect the levels of plasma Ab40 and Ab42 (Fig. 1B, Table 2). When we analyzed the relationship between RBC Ab and an oxidative stress marker (PLOOH), we found a significant positive correlation between RBC PLOOH and Ab concentrations (Fig. 3).

Discussion
The brain is generally regarded as the origin of the Ab that is deposited in plaques of AD patients [5,21]. Plasma Ab and CSF concentrations are believed to be in a dynamic equilibrium [22,23], suggesting that increased Ab production in the brain could be associated with increased Ab concentrations in blood plasma. This means that brain Ab is transferred across the bloodbrain barrier to the plasma [24,25]. As proof of the transfer, peripheral administration of anti-Ab antibody (m266) to PDAPP transgenic mice (AD-model mice) increased plasma Ab up to 1000-fold [26]. Additional evidence for the presence of Ab in blood plasma was obtained in studies of platelets [27]. As mentioned in the Introduction, our group and other researchers found that Ab is capable of binding to RBC in vitro as well as in vivo animal studies [9]. Thus, it is likely that Ab in peripheral blood plasma may readily contact RBC in circulating human blood.   . X-axis is the concentration of RBC Ab. Y-axis is concentration of RBC astaxanthin that had been measured in our former human study [12]. doi:10.1371/journal.pone.0049620.g002 Although plasma Ab has been investigated thoroughly in previous studies [1,5,21,28], little attention has been paid to RBC Ab. In the present study, we provide evidence that Ab is indeed present in human RBC. We found that RBC Ab40 and Ab42 levels in healthy human volunteers were about 8-and 14-times higher than plasma Ab40 and Ab42, respectively (Table 1), when RBC Ab levels per g hemoglobin were compared to plasma Ab levels per g protein. These results suggested that plasma Ab42 and Ab40 readily bind to RBC, and this may provide an explanation for lower concentrations of ''unbound'' Ab42 and Ab40 in human plasma [3]. In addition, the experiments confirmed that the RBC Ab42/Ab40 ratio was about 1.8-times higher than in plasma (Table 1). This may be related to the fact that Ab42 interacts with RBC more avidly than Ab40 [9,29], probably because two additional hydrophobic amino acids at the C-terminus of Ab42 increase the rate of Ab insertion into the RBC bilayer [30].
We also found that RBC Ab40 and Ab42 levels in healthy elderly subjects were higher than in young volunteers. It was reported that brain as well as plasma Ab (Ab40 and Ab42) levels tend to increase according to age [31]. The age-related increase in plasma Ab could be connected to increases in Ab production or a reduction in Ab clearance in the brain. These shifts may be related to changes in the central or peripheral activity of Ab synthetic enzymes (e.g., b-secretase or c-secretase) or Ab catabolic enzymes (e.g., insulin-degrading enzyme or neprilysin) associated with aging. Therefore, the age-related enhanced binding of Ab to RBC may reflect age-dependent changes in Ab metabolism. Since significant positive correlations were observed between RBC and plasma Ab concentrations (Fig. 1), this may strengthen the hypothesis.
In order to supply oxygen to the brain, RBCs must deform as they pass through the narrow pores of capillaries. However, RBC deformability reportedly decreases when Ab adheres to RBC [32]. Thus, the interaction of Ab with RBC may decrease blood flow, impair oxygen delivery to the brain and contribute to brain hypoxia [32]. These processes are implicated in the pathogenesis of AD. In support of these notions, a relationship between blood (plasma) Ab and AD was observed in Down syndrome patients, among whom those with elevated Ab levels in plasma were reported to have a greater risk of developing AD [33]. Additional research (e.g., measurement of Ab levels in RBCs of AD patients) would be necessary to confirm these hypotheses.
On the other hand, as a preventive strategy, compounds that are capable of minimizing the accumulation of Ab in blood might be useful therapeutically. In this study, we showed that after astaxanthin supplementation, Ab40 and Ab42 concentrations in RBC (but not plasma) were significantly decreased (Fig. 1, Table 2). . X-axis is concentration of RBC Ab. Y-axis is concentration of RBC PLOOH [phosphatidylcholine hydroperoxide (PCOOH) and phosphatidylethanolamine hydroperoxide (PEOOH)] that had been measured in our former human study [12]. doi:10.1371/journal.pone.0049620.g003 Table 2. Changes in Amyloid b levels in RBC and plasma before and after a 12 week administration of 0, 6 or 12 mg astaxanthin. Significantly different between before and after astaxanthin administration: *P,0.05. Blood samples (RBC and plasma) that had been obtained from our former human study [12] were subjected to Ab determination. doi:10.1371/journal.pone.0049620.t002 In addition, inverse relationships between RBC Ab and astaxanthin levels were found (Fig. 2). Our previous in vitro and in vivo murine studies also indicated that carotenoid supplementation, especially astaxanthin, could attenuate Ab-induced oxidative stress in RBCs [9]. It is therefore likely that carotenoids (astaxanthin) act as antioxidants and/or reduce the binding of Ab to RBCs, thereby improving the resistance of RBCs to Ab-induced oxidative damage. For other carotenoid, b-carotene reportedly inhibited fibrillation and oligomerization of Ab [34,35], indicating a possibility that carotenoid moieties may bind to C-terminal portion of Ab, thereby inhibiting the binding of Ab to RBC. On the other hand, for currently unknown reasons, astaxanthin changed the levels of Ab in RBC but not in plasma. This may be related to Ab clearance from plasma, since excessive plasma Ab is reportedly cleared from the circulation by mainly hepatic Ab uptake through the interactions with liver low-density lipoprotein receptor-related protein (LRP-1) [36][37][38]. Further studies are needed to evaluate the effectiveness and mechanisms by which carotenoid (astaxanthin) could be beneficial for the treatment of dementia.
Studies have reported that Ab elicits neurotoxic activity via generation of reactive oxygen species (ROS) [39]. The mechanism by which Ab generates ROS is not fully understood, although one study implicates involvement of the methionine residue at position 35 of Ab [40]. If, indeed, Ab induces ROS, it could in turn trigger membrane oxidative injury in RBCs. Because Ab seems to cause RBC aggregation and hemolysis [9], it is plausible that Abinduced hemolysis enhances a cascade of oxidative reactions in RBC. These reactions produce superoxide, which dismutates to form hydrogen peroxide. These ROS cause formation and accumulation of RBC PLOOH, and this could increase membrane rigidity and decrease the deformability of RBCs. In concordance with these considerations, positive correlations between RBC Ab and PLOOH were found in the present study (Fig. 3).
In conclusion, we provided evidence that Ab40 and Ab42 concentrations were much higher in RBCs than in plasma and that RBC Ab levels increased with aging. We also found that after astaxanthin supplementation, there was a decrease in RBC Ab concentrations. The RBC Ab levels were positively correlated with RBC PLOOH, and inversely correlated with RBC astaxanthin. Based on the present findings, we are currently collecting the blood samples from living AD subjects to investigate the pathogenic roles of RBC Ab and usability of RBC Ab as biomarkers of AD. Also, we are investigating therapeutic application of carotenoids (astaxanthin) for possible anti-dementia agents. These results will be presented in the near future as a different study.