Peripheral Immune Cells and Cerebrospinal Fluid Biomarkers of Alzheimer’s Disease Pathology in Cognitively Intact Older Adults: The CABLE Study

Abstract

Background:

Inflammation plays a role in occurrence and progression of Alzheimer’s disease (AD). Whether peripheral immune cells are involved in major pathological processes including amyloid-β plaques and tau tangles is still controversial.

Objective:

We aimed to examine whether peripheral immune cells counts were associated with early changes in cerebrospinal fluid (CSF) biomarkers of AD pathology in cognitively intact older adults.

Methods:

This study included 738 objective cognitive normal participants from the Chinese Alzheimer’s Biomarker and Lifestyle (CABLE) database. Group comparisons of peripheral immune cells counts were tested by analysis of covariance. Multiple linear regression models were used to examine the associations of peripheral immune cells counts with CSF AD biomarkers.

Results:

In preclinical AD, peripheral lymphocytes and eosinophils changed dynamically along with disease progression. Consistently, regression analysis showed that lymphocytes and eosinophils were associated with Aβ pathology. There were no interaction effects of peripheral immune cells counts with APOE ɛ4, gender, age, and educate. Eosinophil to lymphocyte ratio were also significantly associated with Aβ-related biomarkers.

Conclusion:

Our findings showed the relationship between peripheral immune cells and Aβ pathological biomarkers, which indicated that peripheral immune might play a role in progression of AD pathology.

Keywords

Alzheimer’s disease cerebrospinal fluid inflammation pathology peripheral immune cells

INTRODUCTION

Alzheimer’s disease (AD), the most prevalent form of dementia, is one of the public health challenges in the 21st century [1]. The characteristic clinical symptoms of AD are difficulties with language, memory, and problem-solving. The accumulation of abnormally folded amyloid-β (Aβ) in amyloid plaques and tau proteins in neuronal tangles are typical pathologies of AD [1]. Basic studies find that alterations in neurons, microglia, and astroglia promote the insidious development of the disease prior to cognitive impairment [2]. Importantly, neuroinflammation triggered by immune cell responses in the brain is upstream or parallel to AD pathological proteins [3]. Thus, the importance of immunology in neurobiology should be appreciated and the immunological and neurobiological perspectives should be combined.

In addition to neuroinflammation, peripheral inflammation, a modifiable AD risk factor, has also been linked to AD [4]. Cortical inflammation is associated with upregulation of peripheral cytokines in animal models, which suggested that peripheral inflammation could trigger neuroinflammation [5, 6]. The mechanisms by which peripheral cytokines access the brain may involve crossing blood-brain-barrier (BBB) or crossing areas lacking the BBB [7]. Furthermore, inflammatory alterations of peripheral blood cells can promote and sustain neuroinflammation in AD. Peripheral immune cells can migrate to the central nervous system (CNS) and then interact with microglia [8, 9], and the interaction between peripheral immune cells and microglia can regulate the clearance of Aβ deposition in the mouse brain [10]. However, there is a shortage of evidence that peripheral inflammation affects the pathology of AD in the human body at present.

With the indepth understanding of AD, the development of biomarker provides possibilities to detect the dynamic changes of AD pathology. The 2018 National Institute on Aging-Alzheimer’s Association (NIA-AA) Research Framework divide biomarkers into Aβ deposition, pathologic tau, and neurodegeneration [AT(N)] framework [11]. In the cerebrospinal fluid (CSF), Aβ₄₂, phosphorylated tau (p-tau), and total tau (t-tau) have been selected to reflect A, T, and N respectively. In addition to Aβ₄₂, other biomarkers including Aβ_42/40, t-tau/Aβ₄₂, and p-tau/Aβ₄₂ are also regarded as sensitive indicators [12, 13]. Thus, these CSF biomarkers can reflect the dynamic changes of AD pathology with high accuracy. Considering that pathology has appeared in the preclinical stage of AD, AD biomarker researches are of great significance to early prevent disease.

Total and differential white blood cell (WBC) counts are common clinical tests characterized by low cost, high precision, and high standardization, and it can be used to monitor total WBC and specific WBC subpopulations. Therefore, we performed a cross-sectional analysis, evaluating the relationships of total and differential WBC counts to CSF pathological biomarkers in elderly adults without cognitive impairment.

METHODS

CABLE database

In this study, all included participants were selective from the Chinese Alzheimer’s Biomarker and Lifestyle (CABLE) study. CABLE, as an ongoing large-scale study, has been focused on AD risk factors and biomarkers in Chinese Northern Han population since 2017. All participants in CABLE were enrolled at Qingdao Municipal Hospital, Shandong Province, China. Participants were aged between 40 to 90 years and consisted of cognitively intact older adults and individuals with mild cognitively impaired (MCI) or AD. Comprehensive questionnaires and an electronic medical record system are reliable source of demographic information, AD risk factor profile, and medical history [14]. Professional medical doctors made a unanimous diagnosis through neuropsychological testing, CSF biomarkers, and brain magnetic resonance imaging examinations, which conform to the NIA-AA workgroup diagnostic criteria [15, 16].

Participants

This study included 738 normal cognitive participants from the CABLE study. The exclusion criteria included: 1) central nervous system infection, head trauma, epilepsy, multiple sclerosis, or other major neurological disorders; 2) major psychological disorders (e.g., depression); 3) severe systemic diseases (e.g., malignant tumors); 4) family history of genetic disease; 5) hematological system diseases; 6) recent history of acute infection; 7) history of autoimmune disease; 8) history of acute poisoning. General cognitive function was assessed by adapted Chinese-Modified Mini-Mental State Examination (CM-MMSE). Basic living ability was assessed by basic Activities of Daily Living score. Relevant information of each participant was obtained from the CABLE cohort, including age, gender, years of education, CM-MMSE, Apolipoprotein ɛ4 (APOE ɛ4) status, smoking history, hypertension, diabetes mellitus (DM) and levels of CSF biomarkers of AD pathology (Aβ₄₂, t-tau, p-tau, Aβ₄₂/Aβ₄₀, t-tau/Aβ₄₂, and p-tau/Aβ₄₂), as well as details about total and differential WBC counts.

CSF samples collection and measurements

CSF samples of participants were obtained by lumbar puncture after overnight fasting. Within 2 h, these samples were centrifuged at 2000×g for 10 min to eliminate impurities and snap frozen at –80°C until assay. Enzyme-free EP (Eppendorf) tube (AXYGEN; PCR-02-C) were used to store CSF in this study. The thaw/freezing cycle should not exceed two times. CSF Aβ₄₂, t-tau, and p-tau were detected by the enzyme-linked immunosorbent assay (ELISA) kit (Innotest-AMYLOID (1–42), PHOSPHO-TAU (181p), and hTAU-Ag; Fujirebio, Ghent, Belgium) on the microplate reader (Thermo Scientific™ Multiskan™ MK3). The standards and CSF samples were analyzed in duplicates, and the means values of the duplicates were used for subsequent statistical analyses. The mean inter-assay coefficient of variation was under 15% (8.69% for Aβ₄₂, 10.05% for Aβ₄₀, 10.11% for p-tau, and 13.11% for t-tau). The mean intra-assay coefficient of variation was under 10% (5.25% for Aβ₄₂, 3.73% for Aβ₄₀, 2.49% for p-tau, and 5.00% for t-tau). All analyses were operated by professional experimenters who were blind to clinical information.

Neuropathology [17] and neuroimaging [18] studies showed that approximately one-third of older adults with normal cognition had AD pathology in brains. Therefore, in this study, CSF biomarker positive participants were defined as having CSF Aβ₄₂ levels in the lower one-third of the distribution of participants (A+: ≤121.17 pg/mL) or having p-tau (T+: ≥39.02 pg/mL) or t-tau (N+: ≥181.19 pg/mL) levels in the upper one-third of the distribution. Similar methods were also used in other studies, which received reasonable results [19, 20]. According to the 2018 NIA-AA research framework, four different biomarker group combinations including stage 0, stage 1, stage 2, and suspected non-AD pathology (SNAP) were identified (Supplementary Table 1). Participants with normal measures of Aβ₄₂, p-tau, and t-tau (A–T–N–) were classified as stage 0. Participants with abnormal Aβ_1 - 42 but no abnormal p-tau or t-tau (A+T–N–) were classified as stage 1. Participants with abnormal Aβ_1 - 42, and abnormal p-tau or t-tau (A+T+N–, A+T–N+, A+T+N+) were classified as stage 2. According to the AD continuum category, we also classified participants into HC (stage 0) and preclinical AD (stage 1 and stage 2) [21].

Blood sample collection and measurements

After an overnight fast, trained nurse collected blood samples from the participants in the morning. Total and differential WBC counts were assayed by flow cytometry using a fully automatic hematology analyzer (Sysmex, Kobe, Japan) within two hours after collection. The within-batch CV was < 6%. Total and differential WBC counts included WBC count, neutrophil count (NEUT), neutrophil percentage (NEUT%), lymphocyte count (LY), lymphocyte percentage (LY%), monocyte count (MONO), monocyte percentage (MONO%), eosinophil count (EO), eosinophil percentage (EO%), basophil count (BASO), and basophil percentage (BASO%). Other blood specimens were centrifuged at 2000×g for 10 min and stored at –20°C until assay. The thaw-freezing cycle was limited not to surpass two times. DNA was extracted from these specimens using QIAamp®DNA Blood Mini Kit (250) and stored in an enzyme-free EP tube at –80°C until the APOE ɛ4 genotyping. Restriction fragment length polymorphism technology was applied for genotyping according to two specific loci, including rs7412 and rs429358. APOE ɛ4 carrier status was classified into APOE ɛ4 non-carriers or carriers with at least one ɛ4 allele.

Standard protocol approvals, registrations, and patient consents

The conduct of CABLE database complies with the Helsinki declaration, and the research plan was approved by the Institutional Ethics Committee of Qingdao Municipal Hospital. All study participants or their nursing staff directly provided written informed consent.

Statistical analysis

Sample characteristics were summarized using mean and standard deviation (SD) for continuous variables and count and percentage for categorical variables. The outlier values which situated outside three SD were excluded prior to subsequent analyses (Supplementary Table 2). All serum and CSF continuous variables were Box-Cox transformed to normalize the distributions. Continuous variables were standardized to z scores to facilitate comparison between models. Group comparisons of peripheral immune cells counts were tested by analysis of covariance (ANCOVA) with adjusting for age, gender, years of education, and APOE ɛ4 status.

To determine the associations between immune cells counts and CSF biomarkers of AD pathology, the multiple linear regression analyses were performed with CSF biomarkers as dependent variables and immune cells counts as independent variable. Each model was adjusted for age, gender (male or female), years of education, and APOE ɛ4 status (carriers or non-carriers). Sensitivity analyses were performed by 1) adding more covariates including smoking, DM, hypertension, MMSE, and body mass index (BMI); and 2) excluding SNAP participants. We examined interactions of immune cells counts with gender, age, education, APOE ɛ4 status, and Aβ status (A–or A+).

We used FDR corrections for multiple testing. For all regression models, we used the variance inflation factor (VIF) to explore multicollinearity. The VIF of all models did not exceed the value of 1.2. All statistical analyses were performed using R (version 4.0.2). A two-tailed p < 0.05 was considered statistically significant.

RESULTS

Characteristics of participants

The demographic characteristics participants in our study are summarized in Table 1. A total of 738 individuals who had both measurements of CSF AD biomarkers as well as total and differential WBC counts were included. The population was elderly (aged 61.11±10.90 years) and cognitively unimpaired (mean MMSE score = 27.90), and mean education of participants was 9.89 years (SD = 4.38 years). Female participants accounted for 42.55%. The APOE ɛ4 carriers accounted for roughly 14.91%. The prevalence of hypertension, DM, and smoking were respectively 37.94%, 14.36%, and 29.00%.

Table 1

Characteristics of study participants from CABLE database (N = 738)

Variable	Descriptive statistics
Age (y) mean (SD)	61.11 (10.90)
Gender (Female) N (%)	314 (42.55)
Education (y) mean (SD)	9.89 (4.38)
APOE ɛ4 N (%)	110 (14.91)
CSF Aβ₄₂ (pg/mL) mean (SD)	171.17 (81.31)
CSF p-tau (pg/mL) mean (SD)	36.73 (8.43)
CSF t-tau (pg/mL) mean (SD)	168.01 (64.88)
CSF Aβ₄₀ (pg/mL) mean (SD)	5976.12 (2615.50)
CSF Aβ_42/40 mean (SD)	0.03 (0.02)
CSF p-tau/Aβ₄₂ mean (SD)	0.24 (0.09)
CSF t-tau/Aβ₄₂ mean (SD)	1.07 (0.50)
CM-MMSE mean (SD)	27.90 (2.19)
WBC (10⁹/L) mean (SD)	5.99 (1.55)
NEUT (10⁹/L) mean (SD)	3.34 (1.21)
NEUT% (%) mean (SD)	55.16 (9.42)
LY (10⁹/L) mean (SD)	1.95 (0.57)
LY% (%) mean (SD)	33.51 (8.96)
MONO (10⁹/L) mean (SD)	0.47 (0.15)
MONO% (%) mean (SD)	7.92 (1.80)
EO (10⁹/L) mean (SD)	0.15 (0.10)
EO% (%) mean (SD)	2.53 (1.69)
BASO (10⁹/L) mean (SD)	0.03 (0.02)
BASO% (%) mean (SD)	0.52 (0.25)
BMI mean (SD)	25.38 (3.40)
Smoking N (%)	214 (29.00)
Hypertension N (%)	280 (37.94)
DM N (%)	106 (14.36)

CM-MMSE, China-Modified Mini Mental State Examination; APOE ɛ4, apolipoprotein epsilon ɛ4; CSF, cerebrospinal fluid; Aβ, amyloid-β; t-tau, total tau; p-tau, phosphorylated tau; WBC, white blood cells count; NEUT, neutrophil count; NEUT%, neutrophil percentage; LY, lymphocyte count; LY%, lymphocyte percentage; MONO, monocyte count; MONO%, monocyte percentage; EO, eosinophil count; EO%, eosinophil percentage; BASO, basophil count; BASO%, basophil percentage; BMI, body mass index; DM, diabetes mellitus; SD, standardized deviation.

Peripheral immune cells counts in different groups

To assess the associations of peripheral immune cells counts with Aβ deposition and the downstream processes of the amyloid cascade (tau pathology and neurodegeneration), we applied the ATN biological construct and biomarker classification framework. The results for group comparisons of peripheral immune cells counts were shown in Fig. 1 and Supplementary Tables 3–7. There were relationships between Aβ deposition and lymphocytes and eosinophils. In ATN biological construct, LY% were significantly decreased in A+ (p = 0.0093) and A+T+ (p = 0.0022) subgroup compared with A–and A–T–subgroup. In contrast, EO and EO% were significantly raised in A+T+ subgroup (EO, p = 0.0045; EO%, p = 0.0394). Furthermore, EO was significantly increased in A+ subgroup compared to A–subgroup (p = 0.0294). The differences in cells counts were also compared between stage 0, stage 1, and stage 2. There were decreased trends in LY% (p = 0.0061), and increased trends in EO (p = 0.0118) and EO% (p = 0.0436). However, peripheral immune cells had no significant change in T and N biological constructs.

Fig. 1

LY%, EO, and EO% across ATN biological constructs. In box-and-whisker plots the central horizontal bar shows the median, and the lower and upper boundaries show the 25th and 75th percentiles, respectively. Continuous variables were normalized and standardized to z scores. p-values derived from ANCOVA test. ^a Significant after FDR correction.

Associations between peripheral immune cells counts and CSF AD biomarkers in total participants

Figure 2 and Supplementary Table 8 summarized the linear regression results of associations between peripheral immune cells counts and CSF biomarkers in total participants. Results showed that decreased lymphocytes and increased eosinophils were significantly correlated with elevated brain Aβ burden. LY% was significantly associated with CSF Aβ₄₂ (β= 0.0990, p = 0.0092) and CSF p-tau/Aβ₄₂ (β= –0.0805, p = 0.0311). And LY% was marginally associated with CSF Aβ_42/40 (β= –0.0805, p =0.0610). There were negative associations of EO with Aβ₄₂ (β= –0.0900, p = 0.0186) and Aβ_42/40 (β= –0.0849, p = 0.0253), and positive associations with p-tau/Aβ₄₂ (β=0.0861, p = 0.0223) and t-tau/Aβ₄₂ (β=0.0894, p = 0.0164). EO% was significantly associated with p-tau/Aβ₄₂ (β=0.0766, p = 0.0415), and EO% was marginally associated with CSF Aβ₄₂ (β= –0.0718, p = 0.0598) and t-tau/Aβ₄₂ (β=0.0691, p = 0.0626). We failed to find the associations of other immune cells counts with any CSF biomarkers (Supplementary Table 8).

Fig. 2

Relations between CSF core biomarkers and peripheral immune cells counts. Heatmap (A) and scatter plots (B–M) show p values computed by multiple linear regression, after adjustment for age, gender, education, and APOE ɛ4 status. Continuous variables were normalized and standardized to z scores. ^a Significant after FDR correction.

Sensitivity and interaction analyses

Previous studies have found that underlying diseases (hypertension and DM) and lifestyle behaviors (smoking) could influence the peripheral immune cells [22 –24]. In addition, BMI was related to increased AD risk and associated with peripheral immune cells [25, 26]. Thus, we added smoking, DM, hypertension, MMSE, and BMI as additional covariates to do sensitivity analyses. After adding these covariates to multiple linear regression models, above associations in total participants still existed (Table 2). After excluding SNAP, we repeated regression analyses restricting to the 505 participants. The results did not significantly change in this analysis (Table 2).

Table 2

Results of sensitivity analyses

Variable	CSF Aβ₄₂		CSF Aβ_42/40		CSF p-tau/Aβ₄₂		CSF t-tau/Aβ₄₂
	β	p	β	p	β	p	β	p
adding more covariates ^b
LY%	0.1042	0.0074^a	0.0572	0.1342	–0.0781	0.0387	–0.0531	0.1552
EO	–0.0969	0.0124 ^a	–0.0884	0.0202 ^a	0.0924	0.0143 ^a	0.0925	0.0130 ^a
EO%	–0.0803	0.0382	–0.0713	0.0606	0.0867	0.0219	0.0764	0.0405
excluding SNAP ^c
LY%	0.0837	0.0656	0.0862	0.0571	–0.1178	0.0086 ^a	–0.1024	0.0236 ^a
EO	–0.1291	0.0047 ^a	–0.1254	0.0063 ^a	0.1091	0.0159 ^a	0.1239	0.0066 ^a
EO%	–0.1066	0.0187	–0.0954	0.0359	0.0847	0.0603	0.0902	0.0461

^aSignificant after FDR correction. ^badjusted for age, gender, education, APOE ɛ4 status, smoking, DM, hypertension, MMSE, and BMI. ^cadjusted for age, gender, education, and APOE ɛ4 status. CSF, cerebrospinal fluid; Aβ, amyloid-β; t-tau, total tau; p-tau, phosphorylated tau; LY%, lymphocyte percentage; EO, eosinophil count; EO%, eosinophil percentage; SNAP, suspected non-AD pathology.

We next determined whether there were interactions between peripheral immune cells counts and four covariables (APOE ɛ4 status, age, gender, and education) and Aβ status. Results revealed that there were no interaction effects of peripheral immune cells counts with APOE ɛ4 status, age, gender, education, and Aβ status (Supplementary Table 9).

Associations between eosinophil to lymphocyte ratio (ELR) and AD biomarkers

Considering that eosinophils and lymphocytes are related to the Aβ pathology, we used ELR (EO/LY or EO% /LY%) to reflect the overall effect of peripheral immune cells on Aβ pathology. Regression analysis showed that ELR were significantly associated with Aβ-related biomarkers (Fig. 3A–D; Supplementary Table 10). When we exclude SNAP, the associations between ELR and Aβ-related biomarkers were more significant (Fig. 3E-H; Supplementary Table 10).

Fig. 3

Associations of ELR with CSF core biomarkers. Scatter plots show p values in total participants (A–D) and in stage 0, 1, and 2 participants (E–H), adjusting for age, gender, education, and APOE ɛ4 status. Continuous variables were normalized and standardized to z scores. p-values were still significant after FDR corrections.

DISCUSSION

In this large-scale study, we revealed that peripheral lymphocytes and eosinophils changed dynamically along with disease progression. And peripheral lymphocytes and eosinophils were independently associated with Aβ pathology. Interestingly, we found that ELR were also significantly associated with Aβ-related biomarkers. These findings supported the potential role of peripheral immune cells in the pathogenesis of AD.

Previously, it was reported in AD subject alterations of peripheral immune cells, including dendritic cells, lymphocytes, neutrophils, monocytes, and basophils [27 –29]. However, there is a shortage of relevant study in preclinical AD. Herein, our findings extend the understanding of peripheral immune system to elderly individuals without objective cognitive impairment. In this study, we used CSF biomarkers to characterize AD pathophysiology and biologically define preclinical AD. Compared to CSF Aβ₄₂, Aβ_42/40, p-tau/Aβ₄₂, and t-tau/Aβ₄₂ all showed better concordance with Aβ PET and improved diagnostic accuracy for AD [12, 13]. By combining these four biomarkers for analysis, we could accurately capture the evidence that peripheral immune cells participated in the progression of AD at an early stage.

CNS has a specialized and dynamic system to fulfill its exceptional and vital tasks. In the past, the CNS was considered as an immune-privileged organ due to the presence of BBB. However, immune cells were present in both the subarachnoid space and the perivascular space [30], suggesting that CNS is also under regular immune surveillance like other organs. Consistently, our study also found that peripheral immune cells were related with the progression of Aβ pathology. Thus, peripheral immune cells may play important roles in both homeostasis and disease pathogenesis.

It was reported that AD patients had significantly lower levels of lymphocytes [28]. In line with this finding, our study showed decreased lymphocytes in preclinical AD. In the present study, we found that lymphocytes were negatively associated with Aβ pathology. Mounting evidence suggests that lymphocytes contribute to the removal of Aβ deposits in the brain. Immunoglobulin G (IgG) secreted by B lymphocytes were shown to bind microglial Fc receptor and induce microglial phagocytosis of Aβ in 5xFAD mice [31]. Aβ-reactive type 1 T helper (Th1) cells could migrate to the brain parenchyma and enhance the clearance of Aβ plaques by microglia in APP/PS1 model [32]. With respect to these reports, this human study successfully revealed an relationship between lymphocytes and Aβ pathology.

In addition to lymphocytes, our data showed that higher levels of eosinophils were related to increased Aβ pathology. It is important to note that the role of eosinophils in AD has not been reported before. However, multiple existing mechanisms may explain this finding. On one hand, eosinophils can disturb the balance of Th1/Th2. Eosinophils, participating in immune homeostasis, can release a wide variety of cytokines including tumor necrosis factor (TNF), interleukin (IL)-4, IL-6, and IL-10 [38]. IL-4 is crucial for Th2 cell differentiation. Increased Th2 cells can inhibit interferon (IFN)-γ production by Th1 [40]. Furthermore, indoleamine 2,3-dioxygenase (IDO), expressed by eosinophils, catalyzes the breakdown of tryptophan into kynurenine which can inhibit Th1 cell proliferation and promote its apoptosis [39]. Thus, increased eosinophils may inhibit Th1-mediated clearance of Aβ plaque in the brain parenchyma. On the other hand, inhibiting IL-10 selectively has been reported to improve microglial phagocytosis and reduce the Aβ burden [40]. Thus, eosinophils may inhibit the phagocytosis of microglia by secreting IL-10 and IL-4. With the development of detection technology, the role of eosinophils in immune homeostasis and immunity are increasingly recognized. Our new findings may provide a novel target for the interaction between peripheral inflammation and AD.

In this study, we used ELR to reflect overall peripheral immunity. As we expected, ELR showed a more significant correlation with CSF Aβ biomarkers. Our approach is supported by relevant studies. Blood levels of eosinophils may produce information about innate immunity, whereas lymphocyte levels are known biomarkers of the adaptive immunity [41]. Combining indicators of innate and adaptive immunity into separate ratios, such as platelet-to-lymphocyte and granulocyte-to-lymphocyte ratios, is thought to better reflect the relative balance of immunity [42]. In this study, only LY% showed relationships with CSF biomarkers. Considering that both LY and LY% are considered to reflect peripheral blood levels of lymphocytes, the results are confusing. This may be because LY% represents the number of immune cells relative to other immune cells, including eosinophils. In summary, all these evidences suggest that ELR can reflect the overall effect of peripheral immune cells on Aβ pathology.

Previous studies have showed that neutrophils and monocytes can enter the CNS and affect AD pathology [43, 44]. However, we did not find any association of Aβ pathology with neutrophils and monocytes. There are different interpretations of our results. For neutrophils, its infiltration in the brain is usually accompanied by the BBB breakdown. But our participants had no obvious BBB breakdown, indicating that neutrophils could not significantly affect the Aβ pathology in the brain. For monocytes, the recruitment of these cells to the CNS is achieved through C–C chemokine receptor 2 (CCR2) [45]. Thus, the influence of monocyte on the Aβ pathology mainly depends on the transport capacity mediated by CCR2 rather than the number of peripheral cells.

The harnessing of immune response to develop effective treatments for AD are being actively pursued. Aducanumab, humanized IgG, has used to reduce brain Aβ plaques in AD patients [46]. Indeed, targeting specific immune molecules involved in the pathogenesis of AD may be better than non-specific therapies. Modulating cytokines effectively reduced Aβ plaques and reversed cognitive dysfunction in AD models [47 –49]. However, it is necessary to conduct trials on large patient cohorts in the future to determine the effectiveness of these strategies. Considering that our study revealed the immune pattern at the beginning of AD pathology, it is of great significance for the application of immunity to prevent and treat diseases. Understanding the role of lymphocytes and eosinophils is beneficial to develop new therapeutic targets for AD.

Several limitations of our study should be mentioned. First, we used a cross-sectional design which did not show a causal relationship. Second, total and differential WBC counts could not reflect the inflammatory states of peripheral immune cells. Immune cells may play completely different roles in different inflammatory states. Finally, the relationship between peripheral immune cells and AD biomarkers were established in Chinese patients, so it needs to be verified in non-Chinese populations. Thus, it is meaningful for us to establish a validation cohort and find more specific biomarkers which can reflect the inflammatory states of peripheral immune cells in the future studies.

In summary, our study showed that peripheral lymphocytes and eosinophils were associated with Aβ pathology. The results of this study provided evidence that peripheral immune cells play roles in AD pathology at preclinical stages. This study may help us to understand the pathogenic and regulatory pathways in AD, and develop therapeutic strategies.

Footnotes

ACKNOWLEDGMENTS

This study was supported by grants from the National Natural Science Foundation of China (91849126, 81971032, and 81801274), the National Key R&D Program of China (2018YFC1314700), Tianqiao and Chrissy Chen Institute, and the State Key Laboratory of Neurobiology and Frontiers Center for Brain Science of Ministry of Education, Fudan University. The authors thank all participants of the present study as well as all members of staff of the CABLE study for their role in data collection.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

(2021) 2021 Alzheimer’s disease facts and figures. Alzheimers Dement 17, 327–406.

De Strooper

, Karran

(2016) The cellular phase of Alzheimer’s disease. Cell 164, 603–615.

Venegas

, Kumar

, Franklin

, Dierkes

, Brinkschulte

, Tejera

, Vieira-Saecker

, Schwartz

, Santarelli

, Kummer

, Griep

, Gelpi

, Beilharz

, Riedel

, Golenbock

, Geyer

, Walter

, Latz

, Heneka

(2017) Microglia-derived ASC specks cross-seed amyloid-β in Alzheimer’s disease. Nature 552, 355–361.

Jiang

, Jin

, Jiang

, Liao

, Feng

, Zhang

, Liu

, Hao

(2017) Alzheimer’s disease variants with the genome-wide significance are significantly enriched in immune pathways and active in immune cells. Mol Neurobiol 54, 594–600.

Noh

, Jeon

, Seo

(2014) Systemic injection of LPS induces region-specific neuroinflammation and mitochondrial dysfunction in normal mouse brain. Neurochem Int 69, 35–40.

, Yang

, Xiao

, Fan

, Wu

, Terrando

, Wang

(2014) Prolonged neuroinflammation after lipopolysaccharide exposure in aged rats. PLoS One 9, e106331.

Wang

, Ho

, Leung

, Goto

, Chang

(2019) Systemic inflammation linking chronic periodontitis to cognitive decline. Brain Behav Immun 81, 63–73.

Wyatt-Johnson

, Brutkiewicz

(2020) The complexity of microglial interactions with innate and adaptive immune cells in Alzheimer’s disease. Front Aging Neurosci 12, 592359.

Saresella

, La Rosa

, Piancone

, Zoppis

, Marventano

, Calabrese

, Rainone

, Nemni

, Mancuso

, Clerici

(2016) The NLRP3 and NLRP1 inflammasomes are activated in Alzheimer’s disease. Mol Neurodegener 11, 23.

10.

Tejera

, Mercan

, Sanchez-Caro

, Hanan

, Greenberg

, Soreq

, Latz

, Golenbock

, Heneka

(2019) Systemic inflammation impairs microglial Aβ clearance through NLRP3 inflammasome. EMBO J 38, e101064.

11.

Jack

Jr. , Bennett

, Blennow

, Carrillo

, Dunn

, Haeberlein

, Holtzman

, Jagust

, Jessen

, Karlawish

, Liu

, Molinuevo

, Montine

, Phelps

, Rankin

, Rowe

, Scheltens

, Siemers

, Snyder

, Sperling

(2018) NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 14, 535–562.

12.

Hansson

, Seibyl

, Stomrud

, Zetterberg

, Trojanowski

, Bittner

, Lifke

, Corradini

, Eichenlaub

, Batrla

, Buck

, Zink

, Rabe

, Blennow

, Shaw

(2018) CSF biomarkers of Alzheimer’s disease concord with amyloid-β PET and predict clinical progression: A study of fully automated immunoassays in BioFINDER and ADNI cohorts. Alzheimers Dement 14, 1470–1481.

13.

Lewczuk

, Matzen

, Blennow

, Parnetti

, Molinuevo

, Eusebi

, Kornhuber

, Morris

, Fagan

(2017) Cerebrospinal fluid Aβ42/40 corresponds better than Aβ42 to amyloid PET in Alzheimer’s disease. J Alzheimers Dis 55, 813–822.

14.

Hou

, Bi

, Tan

, Xu

, Li

, Shen

, Dou

, Tan

, Yu

(2019) Genome-wide association study identifies Alzheimer’s risk variant in MS4A6A influencing cerebrospinal fluid sTREM2 levels. Neurobiol Aging 84, 241.e213–241.e220.

15.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279.

16.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr. , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269.

17.

Reiman

, Chen

, Liu

, Bandy

, Yu

, Lee

, Ayutyanont

, Keppler

, Reeder

, Langbaum

, Alexander

, Klunk

, Mathis

, Price

, Aizenstein

, DeKosky

, Caselli

(2009) Fibrillar amyloid-beta burden in cognitively normal people at 3 levels of genetic risk for Alzheimer’s disease. Proc Natl Acad Sci U S A 106, 6820–6825.

18.

Morris

, Roe

, Xiong

, Fagan

, Goate

, Holtzman

, Mintun

(2010) APOE predicts amyloid-beta but not tau Alzheimer pathology in cognitively normal aging. Ann Neurol 67, 122–131.

19.

Soldan

, Pettigrew

, Cai

, Wang

, Moghekar

, O’Brien

, Selnes

, Albert

(2016) Hypothetical preclinical Alzheimer disease groups and longitudinal cognitive change. JAMA Neurol 73, 698–705.

20.

, Tan

, Bi

, Shen

, Xu

, Ma

, Li

, Dong

, Yu

(2020) Dynamic changes of CSF sTREM2 in preclinical Alzheimer’s disease: The CABLE study. Mol Neurodegener 15, 25.

21.

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

Jr. , Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wagster

, Phelps

(2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 280–292.

22.

Schwartz

, Weiss

(1994) Cigarette smoking and peripheral blood leukocyte differentials. Ann Epidemiol 4, 236–242.

23.

Ishida

, Kondo

, Funakoshi

, Satoh

, Maeda

, Kawazoe

, Yoshimura

, Tada

, Takahashi

, Ito

, Yasuno

, Masutani

, Nakashima

, Arima

(2021) White blood cell count and incidence of hypertension in the general Japanese population: ISSA-CKD study. PLoS One 16, e0246304.

24.

Liu

, Lai

, Guo

, Ma

, Li

, Fang

, Yang

, Cai

, Liu

, Zhang

, Yang

(2020) Total white blood cell count mediated the association between increased arterial stiffness and risk of type 2 diabetes mellitus in Chinese adults. Arterioscler Thromb Vasc Biol 40, 1009–1015.

25.

Womack

, Tien

, Feldman

, Shin

, Fennie

, Anastos

, Cohen

, Bacon

, Minkoff

(2007) Obesity and immune cell counts in women. Metabolism 56, 998–1004.

26.

Hassing

, Dahl

, Thorvaldsson

, Berg

, Gatz

, Pedersen

, Johansson

(2009) Overweight in midlife and risk of dementia: A 40-year follow-up study. Int J Obes (Lond) 33, 893–898.

27.

Vida

, Martinez de Toda

, Garrido

, Carro

, Molina

, De la Fuente

(2017) Impairment of several immune functions and redox state in blood cells of Alzheimer’s disease patients. Relevant role of neutrophils in oxidative stress. Front Immunol 8, 1974.

28.

Pellicanò

, Bulati

, Buffa

, Barbagallo

, Di Prima

, Misiano

, Picone

, Di Carlo

, Nuzzo

, Candore

, Vasto

, Lio

, Caruso

, Colonna-Romano

(2010) Systemic immune responses in Alzheimer’s disease: in vitro mononuclear cell activation and cytokine production. J Alzheimers Dis 21, 181–192.

29.

Chen

, Bu

, Jin

, Shen

, Wang

, Zhuang

, Zhang

, Zeng

, Yao

, Zhou

, Wang

(2017) Altered peripheral profile of blood cells in Alzheimer disease: A hospital-based case-control study. Medicine (Baltimore) 96, e6843.

30.

Ransohoff

, Engelhardt

(2012) The anatomical and cellular basis of immune surveillance in the central nervous system. Nat Rev Immunol 12, 623–635.

31.

Marsh

, Abud

, Lakatos

, Karimzadeh

, Yeung

, Davtyan

, Fote

, Lau

, Weinger

, Lane

, Inlay

, Poon

, Blurton-Jones

(2016) The adaptive immune system restrains Alzheimer’s disease pathogenesis by modulating microglial function. Proc Natl Acad Sci U S A 113, E1316–1325.

32.

Fisher

, Strominger

, Biton

, Nemirovsky

, Baron

, Monsonego

(2014) Th1 polarization of T cells injected into the cerebrospinal fluid induces brain immunosurveillance. J Immunol 192, 92–102.

33.

van der Willik

, Fani

, Rizopoulos

, Licher

, Fest

, Schagen

, Ikram

(2019) Balance between innate versus adaptive immune system and the risk of dementia: A population-based cohort study. J Neuroinflammation 16, 68.

34.

Begum

, Cunha

, Sidhu

, Alkam

, Scolnick

, Rosario

, Ethell

(2014) Women with the Alzheimer’s risk marker ApoE4 lose Aβ-specific CD4+ T cells 10-20 years before men. Transl Psychiatry 4, e414.

35.

Serrano-Pozo

, Qian

, Monsell

, Betensky

, Hyman

(2015) APOEɛ2 is associated with milder clinical and pathological Alzheimer disease. Ann Neurol 77, 917–929.

36.

Fernandez

, Hamby

, McReynolds

, Ray

(2019) The role of APOE4 in disrupting the homeostatic functions of astrocytes and microglia in aging and Alzheimer’s disease. Front Aging Neurosci 11, 14.

37.

Nortley

, Korte

, Izquierdo

, Hirunpattarasilp

, Mishra

, Jaunmuktane

, Kyrargyri

, Pfeiffer

, Khennouf

, Madry

, Gong

, Richard-Loendt

, Huang

, Saito

, Saido

, Brandner

, Sethi

, Attwell

(2019) Amyloid β oligomers constrict human capillaries in Alzheimer’s disease via signaling to pericytes.eaav. Science 365, 9518.

38.

Berek

(2016) Eosinophils: Important players in humoral immunity. Clin Exp Immunol 183, 57–64.

39.

Odemuyiwa

, Ghahary

, Li

, Puttagunta

, Lee

, Musat-Marcu

, Ghahary

, Moqbel

(2004) Cutting edge: Human eosinophils regulate T cell subset selection through indoleamine 2,3-dioxygenase. J Immunol 173, 5909–5913.

40.

Guillot-Sestier

, Doty

, Gate

, Rodriguez

Jr. , Leung

, Rezai-Zadeh

, Town

(2015) Il10 deficiency rebalances innate immunity to mitigate Alzheimer-like pathology. Neuron 85, 534–548.

41.

Weller

, Spencer

(2017) Functions of tissue-resident eosinophils. Nat Rev Immunol 17, 746–760.

42.

Fani

, Ahmad

, Ikram

, Ghanbari

, Ikram

(2021) Immunity and amyloid beta, total tau and neurofilament light chain: Findings from a community-based cohort study. Alzheimers Dement 17, 446–456.

43.

Zenaro

, Pietronigro

, Della Bianca

, Piacentino

, Marongiu

, Budui

, Turano

, Rossi

, Angiari

, Dusi

, Montresor

, Carlucci

, Nanì

, Tosadori

, Calciano

, Catalucci

, Berton

, Bonetti

, Constantin

(2015) Neutrophils promote Alzheimer’s disease-like pathology and cognitive decline via LFA-1 integrin. Nat Med 21, 880–886.

44.

Malm

, Koistinaho

, Pärepalo

, Vatanen

, Ooka

, Karlsson

, Koistinaho

(2005) Bone-marrow-derived cells contribute to the recruitment of microglial cells in response to beta-amyloid deposition in APP/PS1 double transgenic Alzheimer mice. Neurobiol Dis 18, 134–142.

45.

Colonna

, Wang

(2016) TREM2 variants: New keys to decipher Alzheimer disease pathogenesis. Nat Rev Neurosci 17, 201–207.

46.

Sevigny

, Chiao

, Bussière

, Weinreb

, Williams

, Maier

, Dunstan

, Salloway

, Chen

, Ling

, O’Gorman

, Qian

, Arastu

, Li

, Chollate

, Brennan

, Quintero-Monzon

, Scannevin

, Arnold

, Engber

, Rhodes

, Ferrero

, Hang

, Mikulskis

, Grimm

, Hock

, Nitsch

, Sandrock

(2016) The antibody aducanumab reduces Aβ plaques in Alzheimer’s disease. Nature 537, 50–56.

47.

MacPherson

, Sompol

, Kannarkat

, Chang

, Sniffen

, Wildner

, Norris

, Tansey

(2017) Peripheral administration of the soluble TNF inhibitor XPro1595 modifies brain immune cell profiles, decreases beta-amyloid plaque load, and rescues impaired long-term potentiation in 5xFAD mice. Neurobiol Dis 102, 81–95.

48.

Boyd

, Bennett

, Mori

, Governatori

, Runfeldt

, Norden

, Padmanabhan

, Neame

, Wefes

, Sanchez-Ramos

, Arendash

, Potter

(2010) GM-CSF upregulated in rheumatoid arthritis reverses cognitive impairment and amyloidosis in Alzheimer mice. J Alzheimers Dis 21, 507–518.

49.

Town

, Laouar

, Pittenger

, Mori

, Szekely

, Tan

, Duman

, Flavell

(2008) Blocking TGF-beta-Smad2/3 innate immune signaling mitigates Alzheimer-like pathology. Nat Med 14, 681–687.