Elevated Levels of Multiple Biomarkers of Alzheimer's Disease in the Aqueous Humor of Eyes With Open-Angle Glaucoma

Abstract

Purpose.: To investigate levels of biomarkers of Alzheimer's disease in the aqueous humor in patients with open-angle glaucoma (OAG).

Methods.: Aqueous humor samples were collected from 38 patients with cataracts, 20 patients with POAG, and 32 patients with exfoliation glaucoma. Aqueous levels of apolipoprotein (Apo) AI, ApoCIII, ApoE, transthyretin (TTR), complement factor H, complement C3, and α2-macroglobulin (α2M) were determined by multiplex bead immunoassay. Clinical data were obtained from medical charts.

Results.: Age and sex did not differ significantly among groups. POAG patients had significantly higher aqueous levels of Apo AI, ApoCIII, ApoE, TTR, and α2M compared with cataract patients (P < 0.001, P = 0.036, P < 0.001, P < 0.001, and P < 0.001, respectively). Corresponding values for exfoliation glaucoma patients were also significantly higher compared with those for cataract patients (P = 0.003, P = 0.009, P < 0.001, P < 0.001, and P < 0.001), and those levels correlated positively with each other in eyes with OAG (P < 0.001 for all combinations). Complement factor H and complement C3 levels did not differ among groups. Mean deviation values for the Humphrey visual field test correlated positively with levels of Apo AI, ApoE, TTR, and complement factor H in OAG patients (P = 0.026, 0.012, 0.008, and 0.027, respectively), but age and IOP values were not correlated with protein levels.

Conclusions.: OAG patients had elevated levels of multiple biomarkers of Alzheimer's disease in the aqueous humor.

Introduction

Glaucoma is an ocular neurodegenerative disease and is the leading cause of blindness throughout the world. ¹ Various factors accelerate the progression of glaucomatous optic neuropathy, and a poor response to treatments aimed at reducing IOP is one of the most important risk factors for blindness in patients with glaucoma. IOP is determined by the balance between the inflow and outflow of the aqueous humor, which includes numerous biologically active factors. We previously reported that the aqueous humor of eyes with open-angle glaucoma (OAG) contained higher levels of multiple cytokines and growth factors compared with nonglaucomatous eyes, as assessed by using the multiplex bead immunoassay. ² Data suggest that the aqueous humor may be a useful sample for detecting biomarkers and that these factors may be involved in the development and/or progression of glaucomatous optic neuropathy. 

Alzheimer's disease (AD) is one of the most common causes of dementia in elderly patients, who may have progressive neurodegenerative disorders. Given the limitations of AD diagnosis and therapy, potential fluid biomarkers of AD, especially in the cerebrospinal fluid (CSF), have been sought, and some of the biomarkers that have been identified were expected to be useful as aids in the diagnosis and treatment of AD. ³ Some of those biomarkers are related to the pathogenesis of AD. For instance, the genotype of apolipoprotein (Apo) E is the well-known risk factor for progression of AD, and Apo E is involved in the aggregation and clearance of amyloid-β, the major component of amyloid plaques which initiate a pathogenic cascade leading to AD. ⁴ Transthyretin (TTR), an another biomarker of AD, directly interacts with amyloid-β and inhibits its ability to aggregate. ^5,6  

Because both AD and glaucoma are age-related progressive neurodegenerative diseases, possible pathologic characteristics that the two diseases may share have been investigated. For instance, some reports found that patients with AD had an increased occurrence of glaucoma. ^7–9 Another report indicated that the retinal ganglion cell layer was thinner and that cupping of the optic nerve head was larger in patients with AD compared with age-matched controls. ¹⁰ Thus, available evidence suggests a link between AD and glaucoma. Also, aqueous levels of some AD-related biomarkers, such as apolipoproteins and TTR, were reportedly high in eyes with POAG. ^11–14 Previous reports thus suggested shared characteristics between AD and glaucoma with regard to fluid biomarkers. However, multiple quantitative analyses of AD-related biomarkers in POAG eyes have not been completed, and such correlations remain to be confirmed. In addition, fewer reports about aqueous levels of these biomarkers in eyes with exfoliation glaucoma (ExG) were published compared with reports about eyes with POAG. 

Here, we report simultaneous increases in the levels of AD-related biomarkers in the aqueous humor samples obtained from eyes with OAG (POAG and ExG). The levels of the analytes correlated strongly with each other and correlated partly with the severity of visual field defects in OAG eyes. 

Materials and Methods

Patients

This cross-sectional study was approved by the institutional review board of Kumamoto University, Kumamoto, Japan. All procedures conformed to the Declaration of Helsinki, and each patient gave informed consent for participation in the study. Included in this study were glaucomatous patients with ExG or POAG who underwent cataract surgery and/or trabeculectomy. The control group was comprised of cataractous patients without glaucoma who underwent phacoemulsification. Exclusion criteria were as follows: (1) eyes with ocular diseases other than cataract and/or glaucoma and (2) eyes with a history of intraocular surgery including laser treatment. When both eyes of a patient met the inclusion criteria, we included in our analysis only the eye that was treated first. 

IOP values were measured by using a Goldmann tonometer during 1 PM and 4 PM, and the mean of three continuous measurements before surgery was said to be the preoperative IOP. In patients with OAG, visual fields were measured with the Humphrey Visual Field Analyzer within 3 months of the objective surgery except for one patient, whose data were unreliable because of severe visual loss. 

Collection of the Aqueous Humor Samples

Aqueous humor samples were collected at the beginning of surgery, before any incisional procedures, from 38 cataractous (nonglaucomatous) eyes (patients) and 52 eyes (patients) with phakic OAG (20 patients with POAG and 32 patients with ExG). The Table gives the characteristics of the patients. Of those, 47, 48, and 46 patients received eye drops of β-blockers, prostaglandin analogues, and carbonic anhydrase inhibitors at the time of surgery, respectively. Aqueous humor (70–100 μL) was withdrawn by using a limbal paracentesis with a 30-gauge needle and a tuberculin syringe. Contact with intraocular tissues was carefully avoided, and contamination of samples with blood was prevented. Samples were immediately frozen on dry ice and were stored in a freezer at −80°C until measurements were made by means of a multiplex bead-based immunoassay. 

Table

 

Characteristics of the Study Patients

Characteristic	Cataract (Nonglaucomaus) Patients	POAG Patients	ExG Patients
Number of patients	38	20	32
Male/female	20/18	15/5	19/13
Age, y
Mean ± SD	75.5 ± 5.7	71.6 ± 7.3	73.9 ± 7.1
Range	61−83	57−83	56−83
Preoperative IOP, mm Hg
Mean ± SD	12.4 ± 2.7	26.8 ± 7.3*	26.9 ± 7.4*
Range	7.7−21.0	14.0−45.0	15.0−43.7
MD in Humphrey visual field analysis, dB
Mean ± SD	Untested	−16.8 ± 9.4	−18.7 ± 8.5
Range	Untested	−29.1 to −2.0	−31.0 to −2.8
Number of glaucoma eye drops
Mean ± SD	0	2.7 ± 0.7	2.9 ± 0.6
Range	0	1−4	1−4
Duration of glaucoma therapy, mo
Mean ± SD	0	98.8 ± 83.9	53.1 ± 48.4†
Range	0	13.8−367.1	2.5−185.4

*   P < 0.01 compared with control (cataract) patients.

†   P < 0.05 compared with POAG patients.

Multiple Immunoassay Analyses

Analyses utilized a multiplex bead-based immunoassay, xMAP, and Human Neurodegenerative Disease Panel 1 (Luminex, Austin, TX), and levels of seven neurodegenerative disease-related proteins in the aqueous humor, Apo AI, ApoCIII, ApoE, TTR, complement factor H, complement component C3, and α2-macroglobulin (α2M), were determined as described previously. ² Briefly, a 25-μL aliquot of the aqueous humor sample was transferred to a plate, and part of each aliquot was placed into one of the capture microsphere multiplexes. After incubation at 4°C for 18 hours, multiplexed cocktails of biotinylated reporter antibodies were mixed and then incubated at room temperature for 1 hour. Multiplexes were developed by using an excess of streptavidin–phycoerythrin solution. The solution was mixed with each multiplex, after which incubation at room temperature proceeded for 30 minutes. Vacuum filtration was used to reduce the volumes of the multiplexed reactions, and then the volumes were increased by dilution with a matrix buffer. A Luminex 200 instrument (Luminex) was used for analysis, and data were interpreted via proprietary data analysis software (DNASIS Plex version 2.5; Hitachi Software Engineering, Tokyo, Japan). 

Data Analysis

Data were analyzed by using the JMP Version 8 statistical package program (SAS Institute, Cary, NC). The unpaired two-tailed t-test and χ² test were used for comparison of characteristics of patients in the groups. The Tukey-Kramer honestly significant difference test was utilized to compare the levels of aqueous proteins in the groups. Correlations among analytes were assessed by calculating Spearman correlation coefficients. A P value of less than 0.05 was considered statistically significant. 

Results

Increased Multiple AD-Related Protein Levels in the Aqueous Humor Obtained From OAG Eyes

Figure 1 presents the results of AD-related protein measurements. In control (nonglaucomatous) samples, mean values ± SD of Apo AI, ApoCIII, ApoE, TTR, complement factor H, complement component C3, and α2M were 1.65 ± 2.03, 0.01 ± 0.03, 0.25 ± 0.22, 7.67 ± 3.97, 1.15 ± 0.06, 0.76 ± 1.02, and 0.33 ± 0.57 ng/mL, respectively. Compared with the control cases, ExG cases had significantly higher levels of Apo AI (2.86-fold, P = 0.0028), ApoCIII (6.51-fold, P = 0.0088), ApoE (3.37-fold, P < 0.0001), TTR (2.27-fold, P < 0.0001), and α2M (7.51-fold, P < 0.0001). Corresponding values in POAG cases were also significantly higher than those in control cases (3.55-fold, P = 0.0003; 6.26-fold, P = 0.0360; 3.58-fold, P < 0.0001; 2.07-fold, P = 0.0002; and 6.99-fold, P < 0.0001, respectively). In contrast, levels of complement factor H and complement component C3 were not different among the groups. Levels of any measured protein in the POAG and ExG groups did not differ significantly. 

Figure 1. 

 

Comparison of levels of analytes in eyes of control patients (cataractous; left columns), and in eyes with ExG (middle columns), or POAG (right columns). Distributions of the concentrations of Apo AI (A), ApoCIII (B), ApoE (C), TTR (D), complement factor H (E), complement component C3 (F), and α2-macroglobulin (α2M) (G) are presented with the quantile box plots. The mean of the total sample is indicated by a horizontal line across each plot. The median is indicated by a horizontal line in the body of a box. The 75th and 25th quantiles are represented in the top and bottom of a box. The box covers the interquantile range of the data. The 10th and 90th quantiles are the lines above and below each box. **P < 0.01, *P < 0.05.

Multiple Correlations Among AD-Related Protein Levels in the Aqueous Humor of OAG Eyes

Figure 2 shows multiple correlations among proteins measured in the aqueous humor of 52 OAG eyes. Statistical analysis revealed high correlations among levels of Apo AI, ApoCIII, ApoE, TTR, complement component C3, and α2M in these cases (P < 0.0001 in all combinations except for complement component C3-ApoCIII, P = 0.0026). The level of complement factor H, however, was weakly correlated only with levels of Apo AI and α2M (P = 0.0281 and 0.0452, respectively). 

Figure 2. 

 

Multivariate correlations between the analyte concentrations. Scatterplot matrices are on the right, and statistical values appear on the left. Concentrations of Apo AI, ApoCIII, ApoE, TTR, complement component C3, and α2M were highly correlated with each other (P < 0.0001). The level of complement factor H was correlated only with the levels of Apo AI and α2M (P < 0.05).

Relationships of IOP and Visual Field Severity to Levels of AD-Related Proteins in the Aqueous Humor of OAG Eyes

Additional analyses were performed to determine clinical characteristics that were associated with levels of AD-related proteins in the aqueous humor obtained from OAG patients. Statistical analysis revealed that age, preoperative IOP value, number of glaucoma eye drops, and duration of glaucoma therapy were not significantly correlated with the level of any protein measured (data not shown). The mean deviation (MD) value in the Humphrey visual field analysis correlated positively with levels of Apo AI, ApoE, TTR, and complement factor H (P = 0.0258, 0.0117, 0.0081, and 0.0271, respectively), although the correlations were not strong (coefficients of correlation were 0.31, 0.35, 0.37, and 0.31, respectively) (Fig. 3). 

Figure 3. 

 

Correlations of the MD in Humphrey visual field analysis with aqueous Apo A1 (A), ApoE (B), TTR (C), and complement factor H (D). The x-axes represent the levels of analytes, and the y-axes represent the MD values (dB). Scatterplots show that the MD values correlated positively with the levels of Apo AI, ApoE, TTR, and complement factor H (P = 0.0258, 0.0117, 0.0081, and 0.0271, respectively), and their coefficients of correlation were 0.31, 0.35, 0.37, and 0.31, respectively.

Discussion

The aqueous humor provides nutrients to avascular tissues, such as the lens, cornea, and trabecular meshwork, in the anterior ocular segment and, in humans, drains mainly through the conventional outflow pathway. The aqueous humor contains various biologically active factors, such as cytokines and growth factors, and some of these factors may be associated with the pathophysiology of glaucoma. ^15–19 The multiplex bead immunoassay is a recently developed technique that is highly sensitive compared with the enzyme-linked immunosorbent assay, ²⁰ and it allows simultaneous measurement of the levels of numerous cytokines and growth factors in a small aqueous humor sample. ^21,22 Past studies, including ours, reported that the new technique can be a useful and reliable method for assessing biologically active factors in the aqueous humor samples. ^2,23–25 Our present investigation revealed, via the multiplex bead immunoassay, detectable levels of AD-related biomarkers, Apo AI, ApoCIII, ApoE, TTR, complement factor H, complement component C3, and α2M, in human aqueous humor samples obtained from patients with cataract or OAG. 

In the present study, human aqueous humor samples obtained from both ExG and POAG patients had elevated levels of Apo AI, ApoCIII, ApoE, TTR, and α2M. Elevated levels of these biomarkers cannot be explained simply by an impaired blood–aqueous barrier by following reasons. Firstly, there were no significant differences in any of the factors measured in the present study between POAG and ExG, though ExG eyes were thought to have more disrupted blood–aqueous barrier compared with POAG. Secondly, levels of complement factor H and complement component C3 were not elevated in OAG patients compared with the controls. Finally, the levels of α2-macroglobulin and complement factor H were negatively correlated with each other (P = 0.0452). Increased local production in the glaucomatous anterior ocular segment may be another explanation. Although the ciliary epithelium reportedly expressed some of the analytes, ²⁶ their physiologic roles in the anterior ocular segment remain elusive. 

Statistical analysis revealed that number of glaucoma eye drops and duration of glaucoma therapy were not significantly correlated with the level of any protein measured. However, the effects of pre-operative glaucoma eye drops might be masked, because many of glaucoma patients had received full medication before surgery. For instance, 47 (90.0%), 48 (92.3%), and 46 (88.5%) of 52 glaucoma patients received eye drops of β-blockers, prostaglandin analogues, and carbonic anhydrase inhibitors at the time of surgery, respectively. Thus, we did not have enough sample number of glaucoma patients without eye drops, and therefore we cannot exclude the possibility that glaucoma eye drops affect the aqueous levels of AD-related biomarkers. 

The present study is the first to provide evidence of the correlation between aqueous levels of AD-related biomarkers and the severity of visual field defects in OAG patients. The MD value correlated positively with the aqueous levels of Apo AI, ApoE, TTR, and complement factor H (P = 0.0258, 0.0117, 0.0081, and 0.0271, respectively). However, coefficients of correlation were relatively low (<0.4), and therefore additional studies with large sample sizes are needed to conclude that aqueous levels of those biomarkers truly reflect the severity of the visual field defect. 

Apos make lipids water soluble by binding, thereby forming lipoproteins transporting lipids. Although the APOE genotype was reportedly related to AD-associated risk, ⁴ the level of ApoE was not elevated in the CSF. ^27–29 In contrast, previous results about the association of the APOE genotype with OAG were controversial, ^30,31 and in the present study and a past study, the aqueous humor of OAG eyes demonstrated an increased level of ApoE. ¹¹ Other Apos were also found, by using proteomics techniques, to be related to AD. ³² One interesting result was that ApoE-containing lipoproteins had the potential to protect retinal ganglion cells against glutamate-induced apoptosis. ³³ In that study, α2M, another AD-related biomarker whose level was increased in OAG eyes in our study here, interfered with the neuroprotective effects of ApoE-containing lipoproteins, because ApoE and α2M utilize a common receptor, the low density lipoprotein receptor-related protein 1. In addition, levels of α2M were high in both human POAG eyes and rat glaucoma model eyes, ¹⁴ and α2M had neurotoxic effects on retinal ganglion cells. ^34,35 Although the exact pathophysiologic roles of ApoE and α2M in glaucomatous eyes have not been fully clarified, increased levels of ApoE and α2M in the aqueous humor of glaucomatous eyes suggest a link between these proteins and glaucoma pathophysiology. 

TTR (which was previously called prealbumin) transports thyroxine and retinol and was reported to be a potential biomarker of AD, although past reports about AD-related changes in TTR in CSF have been controversial. ^27,36 In the eye, TTR is expressed in the ciliary epithelium and RPE, ^26,37,38 and eyes with POAG contain more TTR in the aqueous humor compared with cataractous eyes. ^12,13 In the present study, the aqueous level of TTR increased not only in eyes with POAG, but also in eyes with ExG. In contrast, Bouhenni et al. reported that the level of TTR decreased in the aqueous humor of eyes with primary congenital glaucoma. ³⁹ A well-known TTR-related disease is familial amyloid polyneuropathy, ^40–42 and patients with this disease often develop secondary glaucoma. ⁴³ Data therefore suggest that, although the role of aqueous TTR in glaucoma pathophysiology is unclear, TTR might be related to glaucomatous changes in the anterior chamber. 

In conclusion, the findings of our study described here revealed that both ExG patients and POAG patients manifested elevated levels of biomarkers of AD, such as apolipoproteins, α2M, and TTR, in the aqueous humor. Levels of those analytes correlated with each other and may reflect the severity of glaucoma. 

Acknowledgments

Supported by the Japan Society for the Promotion of Science KAKENHI Grant numbers 23390403, 23659814, 23791993, and 23791994. 

Disclosure: T. Inoue, None; T. Kawaji, None; H. Tanihara, None