Severity-Related Increase and Cognitive Correlates of Serum VEGF Levels in Alzheimer’s Disease ApoE4 Carriers

Abstract

Vascular endothelial growth factor (VEGF) is an angioneurin involved in the regulation of vascular and neural functions relevant for the pathophysiology of Alzheimer’s disease (AD), but the influence of AD severity and ApoE4 status on circulating VEGF and its relationship with cognition has not been investigated. We assessed serum VEGF levels and cognitive performance in AD, amnestic mild cognitive impairment (MCI), and control subjects. VEGF levels were higher in AD patients than in MCI cases and controls (p < 0.05) and showed a progressive increase with clinical severity in the whole study population (p < 0.01). Among AD patients, severity-related VEGF elevations were significant in ApoE4 carriers (p < 0.05), but not in non-carriers. Increased VEGF levels were associated with disease severity and showed mild correlations with cognitive impairment that were only consistent for the ADAS-cog+ items remembering test instructions (memory) and maze task (executive functions) in the group of AD patients (p < 0.05). On the other hand, higher VEGF values were related to better memory and language performance in ApoE4 carriers with moderately-severe AD. According to these results showing severity- and ApoE4-related differences in serum VEGF and its cognitive correlates, it is suggested that increases in VEGF levels might represent an endogenous response driven by pathological factors and could entail cognitive benefits in AD patients, particularly in ApoE4 carriers. Our findings support the notion that VEGF constitutes a relevant molecular target to be further explored in AD pathology and therapy.

Keywords

Alzheimer’s disease apolipoprotein E epsilon-4 allele clinical severity cognition serum vascular endothelial growth factor

INTRODUCTION

Vascular endothelial growth factor (VEGF) is a hypoxia-inducible angioneurin regulating vascular, endothelial/blood-brain barrier, and neural functions that has been involved in the pathophysiology of Alzheimer’s disease (AD) [1 –3]. Several studies reported VEGF alterations in the brain, cerebrospinal fluid (CSF) and plasma/serum of AD patients. VEGF expression was found to be enhanced in vascular and parenchymal compartments of AD brains compared to nondemented subjects; to co-localize with amyloid-β (Aβ) plaques and with clusters of glial cells; and to correlate with Aβ and tau pathology [4 –11]. Increased levels of VEGF in the CSF of AD and vascular dementia patients have been reported too [12], although other authors observed no variations [13] or reduced CSF VEGF values in mild dementia stages [14]. Investigations on plasma/serum VEGF yielded controversial results showing increased [15 –17], unchanged [18, 19] or even decreased VEGF levels in AD compared to controls [20 –22]. Apart from discrepancies of results, all these studies included small samples of AD cases, and none of them evaluated variations in circulating VEGF according to stages of clinical severity of the disease.

VEGF demonstrated relevant neuroprotective, neurotrophic and cognitive effects in experimental conditions [23 –25]. The accumulation of VEGF within Aβ plaques [11] and the inverse correlation of VEGF-positive microvessels with tau and Aβ deposits found in AD brains [26] suggest that an enhanced VEGF activity could reflect a protective brain response counteracting AD pathology, whereas a reduced VEGF signaling might contribute to enhance it. In fact, elevated CSF levels of VEGF were recently reported to be associated with more optimal brain aging outcomes (less decline in hippocampal volume, episodic memory, and executive function), particularly in individuals showing early AD biomarkers [27]. However, the relationships of serum VEGF concentrations with measures of cognitive performance were not properly investigated in previous studies.

Based on findings of reduced CSF VEGF in moderately-severe but not mild AD cases [28] and of enhanced plasma VEGF in ApoE4 (apolipoprotein epsilon-4 allele) carriers with a fast deterioration rate [15], we intended to investigate the influence of AD clinical severity (mild, moderate and moderately-severe stages) and ApoE4 status (ApoE4 carriers and non-carriers) on the serum levels of VEGF and its relationship with cognition. In the present study we evaluated VEGF serum levels in elderly controls, MCI subjects and AD patients. Severity- and ApoE4-related differences in VEGF levels, as well as in the correlations of VEGF with cognitive performance, clinical and lab parameters wereinvestigated.

MATERIALS AND METHODS

Subjects

The study sample consisted of 355 Caucasian subjects, including 245 patients with AD (193 women; mean age: 74.99±7.40 years), 48 amnestic MCI patients (35 women; mean age: 73.46±7.57 years) and sixty two healthy cognitive controls (43 women; mean age: 68.44±7.20 years). AD patients met DSM-IV criteria [29] and NINCDS-ADRDA criteria for probable AD [30]. Amnesic MCI patients were selected according to Petersen criteria revised [31]. Subjects having any other significant neurological or psychiatric disease, active allergies, unstable medical conditions or clinically significant laboratory abnormalities were not included in the study. Patients and controls were not taking systemic corticosteroids, anti-parkinsonian agents, narcotics or cholinesterase inhibitors for at least one month prior to blood sampling. Patients showing clinically significant depression in the medical evaluation and/or scores higher than fifteen in the 17-item subscale of the Hamilton Depression Scale [32], were not included in the study. None of the participants reported changes in the general level of physical activity in the month prior study evaluations. The study was conducted according to Good Clinical Practice guidelines and written informed consent was obtained from all participants.

Measurement of serum VEGF levels

A butterfly-21 INT (Venisystems, Abbott Ireland Ltd., Sligo, Ireland) was inserted into the antecubital vein and baseline blood samples were taken during the morning using evacuated blood collecting tubes (Venojet, Terumo Europe N.V., Leuven, Belgium). Then, serum samples were extracted and stored at –40°C until assays. Serum VEGF levels were determined by using a solid phase enzyme-linked immunosorbent assay (ELISA) kit specific for the quantitative determination of human VEGF¹⁶⁵ levels in cell culture supernates, serum and plasma (R&D Systems, Inc., Minneapolis, MN, USA) provided by Vitro SA (Spain). The minimum detection limit of the assay was <9.5 pg/ml, and the intra- and inter-assay coefficients of variation were <10%.

Clinical and laboratory evaluations

Cognitive performance was evaluated in all participants by using the Mini-Mental State Examination (MMSE) [33] and the Alzheimer’s Disease Assessment Scale-Cognitive-Plus (ADAS-cog+) [34], a 14-item extended version of the AD Assessment Scale-cognitive subscale with an increased sensitivity to detect cognitive changes in milder patients. Based on the procedure previously reported by Farlow et al. [35] for the 11-item ADAS-cog, the 14 individual items of the ADAS-cog+ were distributed into four cognitive domains (memory, language, praxis, and attention-executive function) from which we calculated two main composite scores. The memory and language composite score included the sum of scores of the ADAS-cog+ items assigned to memory and language domains (word recall, word recognition, naming objects and fingers, orientation, remembering test instructions, following commands, spoken language ability, comprehension, word finding difficulties); whereas the praxis and executive function composite consisted of the sum of scores of items related to the praxis and attention-executive function domains (ideational praxis, constructional praxis, attention/distractibility, digit cancellation, and maze task). AD severity was assessed by using the Clinical Interview Based Impression of Severity with Caregiver Input (CIBIS+) [36] and rated as mild (AD-3; CIBIS+ score = 3), moderate (AD-4; CIBIS+ score = 4), and moderately-severe (AD-5; CIBIS+ score = 5). CIBIS+ scores in all controls and MCI cases included in this study were 1 and 2, respectively.

Other clinical and lab evaluations included biological parameters (age, sex, age of disease onset, disease duration, weight, height, body mass index, systolic blood pressure, diastolic blood pressure, heart rate, and temperature), APOE genotype, and blood analytes (platelets, red and white blood cell counts, glucose, total cholesterol, LDL and HDL cholesterol, creatinine, total proteins, and albumin).

Statistical analysis

Data are presented as means plus/minus standard deviations (X (SD)). VEGF data did not follow a normal distribution as evaluated with the Kolmogorov-Smirnov test, but log-transformed VEGF data were normally distributed in all study groups. Therefore, parametric statistics were used for analysis of VEGF natural log data. Group comparisons were done by Chi-Square and ANOVA analyses as appropriate. The influence of disease severity (CIBIS+ score) and ApoE4 status on circulating VEGF were evaluated by ANCOVA using age, sex, platelet counts, and ApoE4 or CIBIS+, respectively, as covariates. Relationships between VEGF natural log values and scores of cognitive performance (MMSE score, total ADAS-cog+ score, scores of individual ADAS-cog+ items and ADAS-cog+ composite scores), disease severity, and clinical and lab parameters were evaluated by partial correlation analysis with corrections for age, gender, ApoE4 and/or CIBIS+. Statistical analyses were conducted in the whole study population by diagnosis (controls, MCI and AD) and by clinical severity (controls, MCI, AD-3, AD-4 and AD-5), and in subgroups of AD patients stratified by disease severity and ApoE4 status. Probability values lower than 0.05 were considered statistically significant.

RESULTS

Baseline VEGF serum levels were significantly higher in AD patients than in MCI subjects and in controls (p < 0.05) (Table 1). VEGF levels were elevated in AD patients (344 pg/ml) by approximately 30% and 50% compared to MCI cases (268 pg/ml) and controls (228 pg/ml), respectively. As depicted in Table 1, sex distribution and platelet counts were similar in control, MCI and AD groups; while average age and ApoE4 frequency were significantly higher in AD and MCI cases than in controls. Age, sex and ApoE4 status had no significant effects on VEGF concentrations neither in the total study sample nor in any diagnostic group. However, VEGF values were significantly influenced by platelet counts in the whole study population [F(3,351) = 20.81; p < 0.001] and in the group of AD patients [F(3,241) = 15.06; p < 0.001].

Table 1

Baseline VEGF serum levels and clinical characteristics in controls, MCI subjects and AD patients

	Controls (n = 62)	MCI (N = 48)	AD (n = 245)	Analysis
	N (%)	N (%)	N (%)	χ ²	df	p
Female gender	43 (69.4)	35 (72.9)	193 (78.8)	2.79	2	ns
APOE ɛ4 allele	8 (12.9)	20 (41.7)	116 (47.3)	24.38	2	<0.001
	Mean (SD)	Mean (SD)	Mean (SD)	F	df	p
VEGF natural log	5.18 (0.78)	5.20 (0.95)	5.52 (0.88)^*	4.65	3, 351	0.010
Age (y)	68.44 (7.20)	73.46 (7.57)^†	74.99 (7.40)^†	19.48	2, 352	<0.001
Platelets (×10⁹/L)	218,76 (33.62)	220,38 (40.12)	226,17 (57.17)	0.65	2, 352	ns
MMSE (score)	27.94 (1.56)^#	24.83 (1.34)^#	17.36 (4.13)^#	269.51	2, 352	<0.001

MCI, mild cognitive impairment; AD, Alzheimer’s disease, MMSE, Mini-Mental State Examiation. VEGF analysis by ANCOVA: Significant effect of platelet counts on VEGF [F(3,351) = 20.81; p < 0.001]; ^*p < 0.05 versus controls and MCI. ^†p < 0.01 versus controls; ^#p < 0.001 versus the rest of the groups (Bonferroni test).

VEGF levels increased progressively with clinical severity as evaluated by the CIBIS+ (Table 2). Consequently, VEGF natural log values showed a significant positive correlation with CIBIS+ scores (r = +0.167, p < 0.01) in the whole study sample after corrections for age, sex, ApoE4 status and platelet counts. VEGF concentrations were almost similar in MCI and mild AD cases; and were significantly higher (p < 0.05) in moderately-severe AD patients than in controls, MCI cases, and mild AD patients (Table 2).

Table 2

VEGF natural log serum levels and clinical characteristics in controls, MCI subjects and subgroups of AD patients stratified by disease severity

	Controls (n = 62)	MCI (N = 48)	AD-3 (n = 70)	AD-4 (n = 116)	AD-5 (n = 59)	Analysis
	N (%)	N (%)	N (%)	N (%)	N (%)	χ ²	df	p
Female gender	43 (69.4)	35 (72.9)	48 (68.6)	97 (83.6)	48 (81.4)	8.55	4	ns
APOE ɛ4 allele	8 (12.9)	20 (41.7)	26 (37.1)	65 (56.0)	25 (42.4)	31.64	4	<0.001
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	F	df	p
VEGF natural log	5.18 (0.78)	5.20 (0.95)	5.39 (0.77)	5.48 (0.97)	5.75 (0.81)^*	3.67	5, 349	0.006
Age (y)	68.44 (7.20)	73.46 (7.57)^‡	72.09 (7.95)^†	76.13 (6.87)^‡	76.19 (6.87)^‡	14.05	4, 350	<0.001
Platelets (×10⁹/L)	218.76 (33.62)	220.38 (40.12)	221.86 (39.24)	226.86 (61.69)	229.93 (65.84)	0.53	4, 350	ns
MMSE (score)	27.94 (1.56)^#	24.83 (1.34)^#	22.16 (2.34)^#	16.79 (2.62)^#	12.76 (1.28)^#	562.80	4, 350	<0.001

MCI, mild cognitive impairment; AD, Alzheimer’s disease; MMSE, Mini-Mental State Examination. VEGF analysis by ANCOVA: Significant effect of platelet counts on VEGF [F(5,349) = 20.10; p < 0.001]; ^*p < 0.05 versus controls, MCI and AD-3. ^†p < 0.05 versus controls, AD-4 and AD-5; ^‡p < 0.01 versus controls; ^#p < 0.001 versus the rest of the groups (Bonferroni test).

Clinical characteristics and baseline VEGF levels in subgroups of AD patients stratified by severity stage and ApoE4 status are summarized in Table 3. Although ApoE4 did not have an overall influence on VEGF levels in AD patients, the severity-related increase of VEGF was significant (p < 0.05) in the subgroup of ApoE4 carriers, but not in ApoE4 non-carriers (Table 3). Among ApoE4 carriers, VEGF concentrations were increased in moderately-severe patients (457 pg/ml) with respect to moderate (342 pg/ml) and mild (272 pg/ml) AD subgroups by about 35% and 70%, respectively (Table 3). Gender distribution, age of onset of the disease, and platelet counts were similar among subgroups of AD patients. Average age was lower (p < 0.01) in mild than in moderate and moderately-severe no-ApoE4 cases, but was similar for all severity-related subgroups of ApoE4 patients (Table 3). Disease duration increased significantly (p < 0.001) with AD severity in ApoE4 carriers and non-carriers, and was significantly shorter (p < 0.01) in ApoE4 than in no-ApoE4 cases of the moderately-severe subgroup (Table 3). Cognitive performance as measured by MMSE and ADAS-cog+ showed the expected progressive decline with increasing AD severity (p < 0.001) in ApoE4 and in no-ApoE4 patients (Table 3). However, ApoE4 carriers showed lower MMSE scores than non-carriers in mild (p < 0.01), moderate (p < 0.05) and moderately-severe (p < 0.05) subgroups (Table 3). Age, age of disease onset and disease duration were not significantly associated with VEGF values in the group of AD patients, but VEGF natural log levels showed significant negative correlations with disease duration in moderately-severe ApoE4 cases (r = –0.421; p < 0.05), as well as with age (r = –0.502; p < 0.01) and with age of disease onset (r = –0.476; p < 0.01) in moderately-severe no-ApoE4 patients.

Table 3

Baseline VEGF serum levels and clinical characteristics in subgroups of AD patients stratified by clinical severity and ApoE4 status

Parameter	ApoE4	AD Severity Subgroup			Analysis
		AD-3	AD-4	AD-5
n	No/Yes	44/26	51/65	34/25
		N (%)	N (%)	N (%)	χ ²	df	p
Female gender (females)	No	28 (63.6)	41 (80.4)	27 (79.4)	4.09	2	ns
	Yes	20 (76.9)	56 (86.2)	21 (84.0)	1.16	2	ns
		Mean (SD)	Mean (SD)	Mean (SD)	F	df	p
VEGF Natural log	No	5.38 (0.79)	5.53 (0.94)	5.62 (0.89)	0.71	4, 124	ns
	Yes	5.39 (0.74)	5.44 (0.99)	5.92 (0.68)^#	3.24	4, 111	0.043
Age (y)	No	70.93 (7.92)	76.16 (8.32)^†	77.38 (7.50)^†	7.68	2, 126	<0.01
	Yes	74.04 (7.76)	76.11 (5.53)	74.56 (5.64)	1.30	2, 113	ns
Age of onset (y)	No	68.59 (7.63)	71.94 (7.83)	71.41 (7.99)	2.39	2, 126	ns
	Yes	71.58 (7.97)	71.45 (6.53)	69.72 (6.26)	0.66	2, 113	ns
Disease duration (y)	No	2.30 (1.29)	4.25 (2.45)	5.85 (1.83)	32.46	2, 126	<0.001
	Yes	2.50 (1.63)	4.42 (2.38)	4.60 (1.58)^**	9.10	2, 113	<0.001
Platelets (×10⁹/L)	No	225.02 (37.19)	233.22 (62.83)	225.18 (79.34)	0.28	2, 126	ns
	Yes	216.50 (42.70)	221.88 (60.81)	236.40 (41.77)	0.98	2, 113	ns
MMSE (score)	No	22.86 (2.05)	17.43 (2.57)	13.12 (1.55)	198.73	2, 126	<0.001
	Yes	20.96 (2.36)^**	16.29 (2.57)^*	12.28 (0.46)^*	96.47	2, 113	<0.001
ADAS-cog+ (score)	No	24.20 (5.92)	38.67 (8.96)	62.02 (7.54)	234.74	2, 126	<0.001
	Yes	25.44 (6.02)	39.15 (7.65)	60.12 (8.16)	141.27	2, 113	<0.001

AD-3, AD-4, and AD-5: Subgroups of mild, moderate and moderately-severe AD patients, respectively. VEGF analysis by ANCOVA: Significant effect of platelet counts on VEGF [F(3,112) = 25.13; p < 0.001] in ApoE4 patients; ^#p < 0.05 versus AD-3 and AD-4 subgroups. ^†p < 0.01 versus AD-3 subgroup. ^*p < 0.05 and ^**p < 0.01 versus No ApoE4.

In the AD population, there were no ApoE4-related differences for none of the other clinical parameters and analytes evaluated, but respiratory rate increased [F(2,242) = 5.81, p = 0.003] and diastolic blood pressure decreased [F(2,242) = 3.64, p = 0.028] progressively with clinical severity. In the subgroup of patients with moderately-severe AD, the levels of total cholesterol and LDL cholesterol were higher (p < 0.05) in ApoE4 carriers (248.52±62.66 mg/dl; and 170.24±59.20 mg/dl) than in non-carriers (211.00±53.14 mg/dl; and 136.06±39.67 mg/dl); and VEGF natural log showed significant positive correlations with total cholesterol in the whole subgroup (p < 0.05) and in ApoE4 cases (p < 0.05).

The correlations between VEGF natural log values and cognitive performance measures found in each diagnostic group and in subgroups of AD patients are depicted in Table 4. VEGF showed no significant correlations with cognition assessments in the control group and correlated significantly only with naming objects and fingers scores in MCI cases (Table 4). No ApoE4-related differences were found in controls and MCI for the cognitive correlations of VEGF (data not shown). In the total AD sample, serum VEGF natural log values showed significant positive correlations with ADAS-cog+ total scores (p < 0.05); with ADAS-cog+ composite scores for memory and language (p < 0.05), and for praxis and executive function (p < 0.01); as well as with particular scores of the ADAS-cog+ items naming objects/fingers, following commands, remembering test instructions, and maze task (Table 4). The only two of these correlations remaining significant after corrections for disease severity (CIBIS+ score) were those of VEGF with remembering test instructions (r = +0.136; p < 0.05) and with maze task (r = +0.137; p < 0.05). Correlations between VEGF natural log values and cognitive performance scores were significant only for ADAS-cog+ items related with memory and language (following commands and remembering test instructions) in no-ApoE4 AD patients; and were significant for memory and language composite, naming objects and fingers, praxis and executive function composite, its associated items (constructional praxis, digit cancellation, and maze task), MMSE, and total ADAS-cog+ in ApoE4 AD cases (Table 4). However, the only correlation remaining significant after correction for the disease severity effect in ApoE4-related AD subgroups was the positive correlation between VEGF and remembering test instructions (r = +0.204; p < 0.05) found in no-ApoE4 AD cases. On the other hand, serum VEGF correlated negatively and significantly (p < 0.05) with total ADAS-cog+, memory and language composite, naming objects/fingers, and following commands scores in the subgroup of moderately-severe ApoE4 patients (Table 4, Fig. 1). These correlations were also negative, but nonsignificant in no-ApoE4 moderately-severe AD cases (Table 4).

Fig.1

Correlations of serum VEGF natural log values with cognitive performance scores in moderately-severe ApoE4 AD cases. Higher VEGF levels were associated with better cognitive performance (lower scores) in the naming objects and fingers item (a), the following commands task (b), the memory and language composite (c), and the total ADAS-cog+ (d). Partial correlation analysis with corrections for age and sex was employed.

Table 4

Relationships between serum VEGF natural log values and scores of cognitive performance in controls, MCI, and AD patients

Parameter	Controls	MCI	AD patients	AD NoApoE4	AD ApoE4	AD-5 NoApoE4	AD-5 ApoE4
n	62	48	245	129	116	34	25
MEMORY &LANGUAGE	r = +0.092	r = –0.054	r = +0.141	r = +0.117	r = + 0.184	r = –0.126	r = –0.421
COMPOSITE	p = 0.490	p = 0.	p = 0.029	p = 0.192	p = 0.050	p = 0.492	p = 0.045
Naming objects/fingers	r = +0.165	r = +0.392	r = +0.180	r = +0.148	r = +0.225	r = –0.177	r = –0.476
	p = 0.212	p = 0.008	p = 0.005	p = 0.096	p = 0.016	p = 0.331	p = 0.022
Following commands	r = +0.100	r = +0.135	r = +0.145	r = +0.176	r = +0.114	r = –0.099	r = –0.442
	p = 0.452	p = 0.376	p = 0.024	p = 0.048	p = 0.226	p = 0.589	p = 0.034
Remembering test instructions	r = +0.064	r = –0.086	r = +0.198^*	r = +0.219^*	r = +0.176	r = –0.210	r = –0.323
	p = 0.632	p = 0.574	p = 0.002	p = 0.013	p = 0.060	p = 0.248	p = 0.132
PRAXIS &EXECUTIVE	r = +0.243	r = –0.016	r = +0.185	r = +0.128	r = +0.254	r = –0.174	r = –0.217
FUNCTION COMPOSITE	p = 0.064	p = 0.918	p = 0.004	p = 0.153	p = 0.006	p = 0.340	p = 0.319
Constructional praxis	r = +0.135	r = –0.039	r = +0.107	r = +0.047	r = +0.186	r = +0.032	r = +0.057
	p = 0.308	p = 0.797	p = 0.096	p = 0.601	p = 0.048	p = 0.862	p = 0.796
Digit cancellation	r = +0.196	r = +0.070	r = +0.122	r = –0.002	r = +0.237	r = –0.032	r = –0.347
	p = 0.136	p = 0.646	p = 0.058	p = 0.979	p = 0.011	p = 0.863	p = 0.105
Maze task	r = +0.064	r = +0.074	r = +0.196 ^*	r = +0.155	r = +0.253	r = –0.213	r = –0.019
	p = 0.631	p = 0.629	p = 0.002	p = 0.082	p = 0.007	p = 0.242	p = 0.933
MMSE	r = –0.081	r = +0.100	r = –0.123	r = –0.085	r = –0.191	r = +0.196	r = +0.353
	p = 0.544	p = 0.515	p = 0.057	p = 0.340	p = 0.041	p = 0.283	p = 0.098
ADAS-cog+	r = +0.189	r = –0.027	r = +0.163	r = +0.126	r = +0.220	r = –0.160	r = –0.419
	p = 0.151	p = 0.860	p = 0.011	p = 0.157	p = 0.019	p = 0.383	p = 0.047

MCI, mild cognitive impairment; AD, Alzheimer’s disease; MMSE, Mini-Mental State Examination; ADAS-Cog+, Alzheimer’s Disease Assessment Scale-Cognitive-Plus. AD-5: Subgroup of moderately-severe AD patients. ApoE4, no-ApoE4: Patients with and without the APOE ɛ4 allele, respectively. Partial correlation analysis with corrections for age, sex, and ApoE4 status (when appropriate) was employed. Asterisks indicate the VEGF correlations that remain significant after corrections for disease severity (CIBIS+ score) in the group of AD patients (Remembering test instructions: r = +0.136; p = 0.035. Maze task: r = +0.137; p = 0.034) and in the subgroup of noApoE4 AD cases (Remembering test instructions: r = +0.204; p = 0.022).

DISCUSSION

Results of the present investigation demonstrated enhanced VEGF serum levels in AD patients and a progressive increase in circulating VEGF with increasing clinical severity (Tables 1 and 2). Our results are consistent with the elevations of plasma/serum VEGF previously found in AD patients [15 –17] but are in contrast with the lack of changes and the reductions in peripheral VEGF reported by other authors [18 –22]. An important limitation of all the previous studies on plasma/serum VEGF was the small number of AD patients evaluated that ranged from 20 to 96 cases. Average levels and the increase of VEGF in AD with respect to controls observed in the present study are of the same magnitude as those recently reported for mild-to-moderate AD patients with depression [19]. However, the influence of depression on VEGF levels was ruled out in our study because patients with significant depressive symptoms were excluded and average scores of the Hamilton depression scale were similar in all diagnostic groups and AD subgroups (data not shown). Studies showing elevated VEGF levels in the brains [4 –11] and in the CSF of AD patients [12] aresupporting our findings too.

The severity-related increase of serum VEGF levels found in AD patients was significant in ApoE4 carriers, but not in non-carriers (Table 3). Associations of VEGF levels with AD severity as evaluated with the Clinical Dementia Rating (CDR) were reported to be nonsignificant in two previous publications. Mateo et al. [21] detected no differences for serum VEGF concentrations according to the disease severity in a sample of 51 AD patients, including 22 mild, 19 moderate, and 10 severe cases, but data are not shown. In a sample of 49 mild-to-moderate AD patients, Jung et al. [19] observed a nonsignificant positive correlation between serum VEGF and CRD scores of similar slope to that found in our study. In good accordance with our results, it was recently found that plasma levels of other angiogenesis-related factors such as angiogenin and tissue inhibitor of matrix metalloproteinase-4 were increased in AD patients; were higher in severe AD cases relative to mild AD cases; tended to be elevated in ApoE4 carriers compared with non-carriers; and correlated significantly with MMSE and CDR scores [37]. The progressive elevation of serum VEGF along the course of the disease is also consistent with findings of previous studies showing that the increase of VEGF levels in the CSF of dementia patients was related to the clinical severity [12]; that the levels of VEGF associated to microglia clusters were enhanced in the brains of severe AD patients [4]; and that VEGF expression in the frontal and parahippocampal cortex increased with AD severity as measured by Braak tangle stage [10]. Severity-related reductions of CSF and brain VEGF were also reported in AD. CSF levels of VEGF were found to be unchanged in mild AD and reduced in moderately-severe AD cases [28]; while the expression of VEGF189, a non-diffusible VEGF isoform that remains bound to the cell surface and the extracellular matrix, was reported to be markedly increased in the hippocampus and entorhinal cortex of early AD cases and to decrease progressively with ADseverity [8].

Moderately-severe ApoE4 patients showed the highest levels of serum VEGF and a negative correlation of VEGF values with the duration of the disease. Nonsignificant negative correlations between VEGF and disease duration were also found in ApoE4 cases with mild and moderate AD (data not shown). These findings indicate that VEGF is more elevated in patients with a faster progression to the moderately-severe disease stage and are in line with the increased levels of VEGF reported by Chiappelli et al. [15] in APOE4 patients with a fast cognitive deterioration compared to those with a slow rate of cognitive decline. In moderately-severe no-ApoE4 patients, VEGF levels were higher in cases with an early onset of the disease as indicated by the negative correlation between VEGF levels and age of onset of the disease. These negative correlations of VEGF with disease duration and age of disease onset in moderately-severe ApoE4 carriers and non-carriers, respectively, might reflect a more marked increase of serum VEGF in patients with a more aggressive AD phenotype (earlier onset and faster progression of the disease). The associations reported for VEGF with CSF and brain contents/deposition of Aβ and tau [10 , 27], and with markers of neuroinflammation as microgliosis and tumor necrosis factor alpha (TNF-alpha) levels [4, 6] indicate that VEGF increases in conditions of enhanced AD pathology and seem to support our results. These observations, together with findings that CSF VEGF is negatively associated with whole brain atrophy in Aβ-positive individuals [38], interacts with Aβ and tau in protecting against hippocampal atrophy and cognitive decline, and exerts strongest neuroprotective effects in the presence of prominent AD biomarkers [27], are suggesting that VEGF elevations might represent a compensatory response trying to counteract the clinicopathologic manifestations of AD. Although interactions of VEGF with APOE have not been investigated in AD patients, the recent demonstrations of reduced VEGF levels in the hippocampus of ApoE4 mice and of decreases in the apoE4-related accumulations of brain Aβ42 and hyperphosphorylated tau induced by intra-hippocampal injections of VEGF-expressing adeno-associated-virus [39] are suggesting the involvement of VEGF-mediated mechanisms in the expression of AD pathology driven by Apoe4. Such VEGF-APOE interactions could account for the elevations of circulating VEGF found in moderately-severe ApoE4 patients and are in support of a protective role of VEGF on ApoE4-related AD pathology.

Hypoxia, hypoperfusion and Aβ have been proposed as factors that are likely to cause upregulation of VEGF in AD brains [6 , 40]. Mechanisms responsible for the increase of serum VEGF in AD have to be elucidated, and its investigation was beyond the scope of this study; but the rise in respiratory rate and the reduction in diastolic blood pressure we found in association with AD severity could translate into a state of relative hypoxia/hypoperfusion, contributing to enhance VEGF in advanced disease stages. However, the lack of significant correlations does not support a direct influence of these parameters on the elevations of VEGF in the population of this study. Platelet counts showed consistent positive correlations with VEGF levels across study groups, but these associations were independent of group differences in circulating VEGF. Hypercholesterolemia was associated with increased VEGF serum levels in moderately-severe AD patients, particularly in ApoE4 carriers, and might contribute to the VEGF increase in advanced AD. Positive correlations of serum VEGF concentrations with total cholesterol and LDL cholesterol were also reported in adult controls and in patients with hypercholesterolemia [41, 42]. Other factors not evaluated in this study that might be involved in the elevations of serum VEGF in AD patients include TNF-alpha, a proinflammatory cytokine overexpressed in AD [43] that stimulates VEGF production [44]; differences in the distribution of vegf gene polymorphisms [15]; reductions in the expression or sensitivity of VEGF receptors[45, 46]; and changes in the circulating levels of soluble forms of VEGF receptors (sVEGFR-1 and sVEGFR-2) which trap VEGF and compete with its binding activity [2, 47].

Overall, with the exception of the significant positive correlation between VEGF and naming objects and fingers found in MCI cases, circulating VEGF did not correlate with measures of cognitive performance in groups (Table 4) nor in ApoE4-related subgroups of controls and MCI. In AD patients we found that higher VEGF serum values were associated with a worse performance (higher ADAS-cog+ scores) in all the cognitive domains evaluated (memory, language, praxis, and executive function). Associations of serum VEGF with cognitive impairment observed in our study are reflecting the progressive increase with AD severity of both VEGF levels and cognitive deterioration, as indicated by the fact that only mild correlations with remembering test instructions and maze task items remained significant after corrections for disease severity (Table 4). Some authors failed to detect significant correlations between serum VEGF levels and global measures of cognition (MMSE) and severity (CDR) in small samples of AD cases [17 , 21], but relationships of serum VEGF with specific cognitive domains were not addressed by previous studies. In ApoE4-related AD subgroups, although associations of VEGF with cognitive impairments were more prominent in ApoE4 carriers, the only significant correlation found after adjustment for AD severity was a mild positive one between VEGF and remembering test instructions in no-ApoE4 cases (Table 4). On the other hand, elevated serum VEGF was significantly associated with better memory and language functioning in moderately-severe ApoE4 patients (Table 4, Fig. 1), and non-significantly in subgroups of advanced no-ApoE4 (Table 4) and mild ApoE4 cases (data not shown). This finding is in line with the attenuated decline of memory found in individuals with elevated CSF VEGF and AD biomarkers [27], as well as with experimental evidence that VEGF reduces cognitive impairment and improves memory in AD animal models [24 , 48–50]. The reversion of cognitive deficits recently found in ApoE4 mice receiving VEGF-expressing-lentivirus [39] indicates a positive effect of VEGF on cognition in ApoE4 carriers and supports the associations between VEGF and cognitive functions observed in our study. In addition, the elevated VEGF levels and short disease duration shown in our ApoE4 cases with advanced AD are resembling the increased VEGF concentrations and fast cognitive decline previously reported in thirteen ApoE4 patients [15] and are suggesting the convenience of assessing long-term relationships of VEGF and cognition. In this regard, the associations between VEGF levels and changes over time in ADAS-cog+ scores shown in both clinical trials and follow-up studies [51, 52] could contribute to elucidate the influence of circulating VEGF on AD progression and on drug effects on cognitive decline [51]. Although the influence of circulating VEGF on brain pathology and functioning is unknown, all the previous observations suggest that the upregulation of serum VEGF could be a neuroprotective response helping to reduce cognitive deterioration in AD, and particularly in ApoE4 patients.

In summary, and taking into consideration that VEGF is increased and correlates with cognitive functioning in advanced ApoE4 AD patients, but not in MCI cases; that results of studies on the serum levels of VEGF in AD are conflicting; that VEGF levels were reported to be decreased in the CSF of mild AD cases [28] and in the hippocampus of ApoE4 mice [39]; and that enhanced expression of VEGF in the CSF and in the brain were found to be associated, respectively, with reduced brain atrophy and memory decline in humans [38] and with a rescue of the ApoE4 pathological phenotype in mice [39], it seems that elevations in circulating VEGF evolve with increasing AD severity as a secondary compensatory response that could attenuate the cognitive decline in advanced ApoE4 AD; whereas reduced VEGF expression in early AD stages might have a primary role in the development of AD pathology.

According to results of the present investigation, VEGF serum levels are elevated in association with cognitive impairment in AD patients; show a significant severity-related increase in ApoE4 carriers; and are associated to better memory and language performance in moderately-severe ApoE4 cases. These findings pointing to VEGF as a relevantmolecular target in AD pathology and therapy need to be confirmed in future studies because the influence of disease severity and ApoE4 on serum VEGF and its cognitive correlates has not been previously investigated. The absence of longitudinal assessments, the lack of data on pathological factors involved in the regulation of VEGF, and the small sample size of controls, MCI cases, and AD subgroups are limiting the interpretation of our results. Long-term correlations between changes in serum VEGF and cognitive functions, interactions of VEGF with its soluble receptors according to AD severity and ApoE4 status, associations of circulating VEGF with its CSF levels, as well as with peripheral and cerebral vascular pathology, and the potential influence of VEGF polymorphisms on the relationships of VEGF and cognition are topics that deserve further investigation in AD.

Footnotes

ACKNOWLEDGMENTS

We thank Olalla Iglesias for technical support. This study was supported by Research Grants from the Fundación Antidemencia Al-Andalus (Córdoba, Spain), from the Society for the Study of Neuroprotection and Neuroplasticity (Cluj-Napoca, Romania), and from Ever NeuroPharma (Unterach, Austria). AA was also supported by the Sixth Framework Programme of the European Union (LSHB-CT-2006-037702).

Authors’ disclosures available online ().

References

Iadecola

(2010) The overlap between neurodegenerative and vascular factors in the pathogenesis of dementia. Acta Neuropathol 120, 287–296.

Ruiz de Almodovar

, Lambrechts

, Mazzone

, Carmeliet

(2009) Role and therapeutic potential of VEGF in the nervous system. Physiol Rev 89, 607–648.

Zlokovic

(2011) Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nat Rev Neurosci 12, 723–738.

Jantaratnotai

, Ling

, Cheng

, Schwab

, McGeer

, McLarnon

(2013) Upregulation and expression patterns of the angiogenic transcription factor ets-1 in Alzheimer’s disease brain. J Alzheimers Dis 37, 367–377.

Kalaria

, Cohen

, Premkumar

, Nag

, LaManna

, Lust

(1998) Vascular endothelial growth factor in Alzheimer’s disease and experimental cerebral ischemia. Brain Res Mol Brain Res 62, 101–105.

Ryu

, Cho

, Choi

, Wang

, McLarnon

(2009) Microglial VEGF receptor response is an integral chemotactic component in Alzheimer’s disease pathology. J Neurosci 29, 3–13.

Sanchez

, Tripathy

, Luo

, Yin

, Martinez

, Grammas

(2013) Neurovascular unit and the effects of dosage in VEGF toxicity: Role for oxidative stress and thrombin. J Alzheimers Dis 34, 281–291.

Tang

, Mao

, Xie

, Greenberg

, Jin

(2013) Expression level of vascular endothelial growth factor in hippocampus is associated with cognitive impairment in patients with Alzheimer’s disease. Neurobiol Aging 34, 1412–1415.

Thirumangalakudi

, Samany

, Owoso

, Wiskar

, Grammas

(2006) Angiogenic proteins are expressed by brain blood vessels in Alzheimer’s disease. J Alzheimers Dis 10, 111–118.

10.

Thomas

, Miners

, Love

(2015) Post-mortem assessment of hypoperfusion of cerebral cortex in Alzheimer’s disease and vascular dementia. Brain 138, 1059–1069.

11.

Yang

, Bae

, Kang

, Gwag

, Gho

, Chae

(2004) Co-accumulation of vascular endothelial growth factor with beta-amyloid in the brain of patients with Alzheimer’s disease. Neurobiol Aging 25, 283–290.

12.

Tarkowski

, Issa

, Sjögren

, Wallin

, Blennow

, Tarkowski

, Kumar

(2002) Increased intrathecal levels of the angiogenic factors VEGF and TGF-beta in Alzheimer’s disease and vascular dementia. Neurobiol Aging 23, 237–243.

13.

Blasko

, Lederer

, Oberbauer

, Walch

, Kemmler

, Hinterhuber

, Marksteiner

, Humpel

(2006) Measurement of thirteen biological markers in CSF of patients with Alzheimer’s disease and other dementias. Dement Geriatr Cogn Disord 21, 9–15.

14.

Guo

, Alexopoulos

, Perneczky

(2013) Heart-type fatty acid binding protein and vascular endothelial growth factor: Cerebrospinal fluid biomarker candidates for Alzheimer’s disease. Eur Arch Psychiatry Clin Neurosci 263, 553–560.

15.

Chiappelli

, Borroni

, Archetti

, Calabrese

, Corsi

, Franceschi

, Padovani

, Licastro

(2006) VEGF gene and phenotype relation with Alzheimer’s disease and mild cognitive impairment. Rejuvenation Res 9, 485–493.

16.

Corsi

, Licastro

, Porcellini

, Dogliotti

, Galliera

, Lamont

, Innocenzi

, Fitzgerald

(2011) Reduced plasma levels of P-selectin and L-selectin in a pilot study from Alzheimer disease: Relationship with neuro-degeneration. Biogerontology 12, 451–454.

17.

Kim

, Kim do

(2012) Decreased serum angiogenin level in Alzheimer’s disease. Prog Neuropsychopharmacol Biol Psychiatry 38, 116–120.

18.

Del Bo

, Scarlato

, Ghezzi

, Martinelli Boneschi

, Fenoglio

, Galbiati

, Virgilio

, Galimberti

, Crimi

, Ferrarese

, Scarpini

, Bresolin

, Comi

(2005) Vascular endothelial growth factor gene variability is associated with increased risk for AD. Ann Neurol 57, 373–380.

19.

Jung

, Kim

, Yoon

, Moon

, Roh

, Lee

, Choi

, Jung

, Kim

(2015) The effect of depression on serum VEGF level in Alzheimer’s disease. Dis Marker 2015, 742612.

20.

Huang

, Jia

, Liu

(2013) Decreased serum levels of the angiogenic factors VEGF and TGF-β1 in Alzheimer’s disease and amnestic mild cognitive impairment. Neurosci Lett 550, 60–63.

21.

Mateo

, Llorca

, Infante

, Rodríguez-Rodríguez

, Fernández-Viadero

, Peña

, Berciano

, Combarros

(2007) Low serum VEGF levels are associated with Alzheimer’sdisease. Acta Neurol Scand 116, 56–58.

22.

Zhu

, Li

, Sun

, Wang

, Zhou

(2015) Quantification of microRNA-210 in the cerebrospinal fluid and serum: Implications for Alzheimer’s disease. Exp Ther Med 9, 1013–1017.

23.

Antequera

, Portero

, Bolos

, Orive

, Hernández

, Pedraz

, Carro

(2012) Encapsulated VEGF-secreting cells enhance proliferation of neuronal progenitors in the hippocampus of AβPP/Ps1 mice. J Alzheimers Dis 29, 187–200.

24.

Religa

, Cao

, Religa

, Xue

, Bogdanovic

, Westaway

, Marti

, Winblad

, Cao

(2013) VEGF significantly restores impaired memory behavior in Alzheimer’s mice by improvement of vascular survival. Sci Rep 3, 2053.

25.

Yang

, Kwon

, Gho

, Chae

(2005) Specific interaction of VEGF165 with beta-amyloid, and its protective effect on beta-amyloid-induced neurotoxicity. J Neurochem 93, 118–127.

26.

Provias

, Jeynes

(2008) Neurofibrillary tangles and senile plaques in Alzheimer’s brains are associated with reduced capillary expression of vascular endothelial growth factor and endothelial nitric oxide synthase. Curr Neurovasc Res 5, 199–205.

27.

Hohman

, Bell

, Jefferson

(2015) Alzheimer’s Disease Neuroimaging Initiative. The role of vascular endothelial growth factor in neurodegeneration and cognitive decline: Exploring interactions with biomarkers of Alzheimer disease. JAMA Neurol 72, 520–529.

28.

Ohrfelt

, Andreasson

, Simon

, Zetterberg

, Edman

, Potter

, Holder

, Devanarayan

, Seeburger

, Smith

, Blennow

, Wallin

(2011) Screening for new biomarkers for subcortical vascular dementia and Alzheimer’s disease. Dement Geriatr Cogn Dis Extra 1, 31–42.

29.

American Psychiatric Association (1984) Diagnostic and Statistical Manual of Mental Disorders, 4th Ed, American Psychiatric Association, Washington DC.

30.

McKhann

, Drachman

, Folstein

, Katzman

, Price

, Stadlan

(1984) Clinical diagnosis of Alzheimer’s disease: Report of NINCDS-ADRDA Work Group. Neurology 34, 939–944.

31.

Artero

, Petersen

, Touchon

, Ritchie

(2006) Revised criteria for mild cognitive impairment: Validation within a longitudinal population study. Dement Geriatr Cogn Disord 22, 465–470.

32.

Hamilton

(1967) Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 6, 278–296.

33.

Folstein

, Folstein

, McHugh

(1975) Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

34.

Mohs

, Knopman

, Petersen

, Ferris

, Ernesto

, Grundman

(1997) Development of cognitive instruments for use in clinical trials of antidementia drugs: Additions to the Alzheimer’s Disease Assessment Scale that broaden its scope: The Alzheimer’s Disease Cooperative Study. Alzheimer Dis Assoc Disord 11, S13–S21.

35.

Farlow

, Schmitt

, Aarsland

, Grossberg

, Somogyi

, Meng

(2013) Comparing clinical profiles in Alzheimer’s disease and Parkinson’s disease dementia. Dement Geriatr Cogn Dis Extra 3, 281–290.

36.

Knopman

, Knapp

, Gracon

, Davis

(1994) The Clinician Interview-Based Impression (CIBI): A clinician’s global change rating scale in Alzheimer’s disease. Neurology 44, 2315–2321.

37.

Qin

, Jia

, Wang

, Zuo

, Wu

, Zhou

, Li

, Min

, Wei

, Tang

, Xing

, Dong

, Wang

, Gao

, Li

, Jia

(2015) Elevated plasma angiogenesis factors in Alzheimer’s disease. J Alzheimers Dis 45, 245–252.

38.

Paterson

, Bartlett

, Blennow

, Fox

, Alzheimer’s Disease Neuroimaging Initiative , Shaw

, Trojanowski

, Zetterberg

, Schott

(2014) Cerebrospinal fluid markers including trefoil factor 3 are associated with neurodegeneration in amyloid-positive individuals. Transl Psychiatry 4, e419.

39.

Salomon-Zimri

, Glat

, Barhum

, Luz

, Boehm-Cagan

, Liraz

, Ben-Zur

, Offen

, Michaelson

(2016) Reversal of ApoE4-driven brain pathology by vascular endothelial growth factor treatment. J Alzheimers Dis 53, 1443–1458.

40.

Barker

, Ashby

, Wellington

, Barrow

, Palmer

, Kehoe

, Esiri

, Love

(2014) Pathophysiology of white matter perfusion in Alzheimer’s disease and vascular dementia. Brain 137, 1524–1532.

41.

, Zheng

, Wang

, Zhong

, Yu

, Bao

, Cao

, Li

, Si-Tu

(2012) Vascular endothelial growth factor downregulates apolipoprotein M expression by inhibiting Foxa2 in a Nur77-dependent manner. Rejuvenation Res 15, 423–434.

42.

Trapé

, Morales

, Molina

, Filella

, Marcos

, Salinas

, Franquesa

(2006) Vascular endothelial growth factor serum concentrations in hypercholesterolemic patients. Scand J Clin Lab Invest 66, 261–267.

43.

Alvarez

, Cacabelos

, Sampedro

, Garcia-Fantini

, Aleixandre

(2007) Serum TNF-alpha levels are increased and correlate negatively with free IGF-I in Alzheimer disease. Neurobiol Aging 18, 533–536.

44.

Haneda

, Hasegawa

, Hirano

, Hashimoto

, Ohsaki

, Ichiyama

(2011) Leukotriene D4 enhances tumor necrosis factor-α-induced vascular endothelial growth factor production in human monocytes/macrophages. Cytokine 55, 24–28.

45.

Gao

, Shen

, Gabriel

, Law

, Yang

, Young

, Su

(2009) Attenuation of brain response to vascular endothelial growth factor-mediated angiogenesis and neurogenesis in aged mice. Stroke 40, 3596–3600.

46.

Petcu

, Smith

, Miroiu

, Opris

(2010) Angiogenesis in old-aged subjects after ischemic stroke: A cautionary note for investigators. J Angiogenes Res 2, 26.

47.

Shibuya

(2013) Vascular endothelial growth factor and its receptor system: Physiological functions in angiogenesis and pathological roles in various diseases. J Biochem 153, 13–19.

48.

Garcia

, Ornellas

, Martin

, Patti

, Mello

, Frussa-Filho

, Han

, Longo

(2014) Therapeutic effects of the transplantation of VEGF overexpressing bone marrow mesenchymal stem cells in the hippocampus of murine model of Alzheimer’s disease. Front Aging Neurosci 6, 30.

49.

Spuch

, Antequera

, Portero

, Orive

, Hernández

, Molina

, Bermejo-Pareja

, Pedraz

, Carro

(2010) The effect of encapsulated VEGF-secreting cells on brain amyloid load and behavioral impairment in a mouse model of Alzheimer’s disease. Biomaterials 31, 5608–5618.

50.

Wang

, Xie

, Guo

, Zhao

, Jiang

, Song

, Zhu

, Lai

, Xu

, Bi

(2011) VEGF-induced angiogenesis ameliorates the memory impairment in APP transgenic mouse model of Alzheimer’s disease. Biochem Biophys Res Commun 411, 620–626.

51.

Alvarez

, Alvarez

, Iglesias

, Crespo

, Figueroa

, Aleixandre

, Linares

, Granizo

, Garcia-Fantini

, Marey

, Masliah

, Winter

, Muresanu

, Moessler

(2016) Synergistic increase of serum BDNF in Alzheimer patients treated with cerebrolysin and donepezil: Association with cognitive improvement in ApoE4 cases. Int J Neuropsychopharmacol 19, 1–6.

52.

Cortes

, Portet

, Touchon

, Vellas

(2007) Six and 18-month changes in mild to moderate Alzheimer’s patients treated with cetylcholinesterase inhibitors: What can we learn for clinical outcomes of therapeutic trials? J Nutr Health Aging 11, 330–337.