IL-1β, IL-6, IL-10, and TNFα single nucleotide polymorphisms are associated with cerebrospinal fluid levels of biomarkers of Alzheimer’s disease

Background Neuroinflammation plays an important role in Alzheimer’s disease (AD). During this process, activated microglia release pro-inflammatory cytokines such as interleukin (IL)-1α, IL-1β, IL-6 and tumor necrosis factor α (TNFα) that participate in neuron damage. However, anti-inflammatory cytokines (such as IL-10), which maintain homeostasis of immune response, are also released. Previous studies showed the association of IL-1α -889C/T (rs1800587), IL-1β -1473G/C (rs1143623), IL-6 -174C/G (rs1800795), IL-10 -1082G/A (rs1800896) and TNFα -308A/G (rs1800629) polymorphisms with AD. Methods In this study, we assessed whether people carrying certain genotypes in these polymorphisms are more prone to develop AD-related pathology, reflected by pathological levels of cerebrospinal fluid (CSF) AD biomarkers including amyloid β1–42 (Aβ1–42), total tau (t‐tau), tau phosphorylated at Thr 181 (p‐tau181), Ser 199 (p‐tau199), and Thr 231 (p‐ tau231), and visinin‐like protein 1 (VILIP‐1). The study included 115 AD patients, 53 patients with mild cognitive impairment (MCI), 11 healthy controls, and 54 patients with other causes of dementia. Results A significant increase in p-tau CSF levels was found in patients with the AA IL-10 -1082G/A and GG TNFα -308A/G genotypes, and in carriers of a G allele in IL-1β -1473C/G and IL-6 -174C/G polymorphisms. T-tau levels were increased while Aβ1-42 levels were decreased in carriers of a G allele in IL-1β -1473C/G polymorphism. An increase in VILIP-1 levels was observed in patients with CG and GG IL-1β whether levels among with different IL-10 -1082G/A, IL-1β -1473C/G, IL-1α -889C/T, IL-6 -174C/G and TNFα -308A/G genotypes that were previously associated with AD [12,13]. We compared the levels of six AD CSF biomarkers (Aβ 1-42 , t-tau, p-tau 181 , p-tau 199 , p-tau 231 and VILIP-1) among patients with aforementioned genotypes. This study gave several notable findings. Levels of Aβ 1-42 were decreased, while levels of t-tau were increased in carriers of G allele in IL-1β -1473C/G polymorphism. T-tau levels were also significantly increased in patients with CG IL-1β -1473C/G genotype. P-tau levels were significantly increased in patients with AA IL-10 -1082G/A and GG TNFα -308A/G genotype, and in carriers of G allele in IL-1β -1473C/G and IL-6 -174C/G polymorphisms. Levels of VILIP-1 were increased in patients with CG and GG IL-1β -1473C/G, GC IL-6 -174C/G and GG TNFα -308A/G genotype.


Abstract
Background Neuroinflammation plays an important role in Alzheimer's disease (AD).
Results A significant increase in p-tau CSF levels was found in patients with the AA IL-10 -1082G/A and GG TNFα -308A/G genotypes, and in carriers of a G allele in IL-1β -1473C/G and IL-6 -174C/G polymorphisms. T-tau levels were increased while Aβ1-42 levels were decreased in carriers of a G allele in IL-1β -1473C/G polymorphism. An increase in VILIP-1 levels was observed in patients with CG and GG IL-1β -1473C/G, GC IL-6 -174C/G and GG TNFα -308A/G genotype.
Conclusions These results suggest that persons carrying certain genotypes in IL10 (-1082G/A), IL1β (1473C/G), IL6 (-174C/G) and TNFα (-308A/G) could be more vulnerable to development of neuroinglammation, and consequently of AD. 3 Background Inflammatory processes are enhanced in the brain of Alzheimer's disease (AD) patients [1,2]. Microglial cells become activated and produce high levels of cytokines. In early stages of AD, activated microglia phagocytose amyloid β (Aβ) peptide, but when they are activated over extended periods [3], they can no longer clear Aβ, and the proinflammatory cytokines they release participate in propagation of pathological tau proteins and neuron damage [4,5]. The main pro-inflammatory cytokines released from activated microglia are interleukin (IL)-1α, IL-1β, IL-6 and tumor necrosis factor α (TNFα) [6]. During sustained inflammation, anti-inflammatory cytokines (such as IL-10) are also released and maintain homeostasis of the immune response [6]. Single nucleotide polymorphisms (SNPs) in genes for IL-1α, IL-1β, IL-6, IL-10 and TNFα were previously associated with AD [7,8]. It was shown that certain SNPs can influence gene transcription and consequently the amount of the produced cytokines [9-11]. The association of these SNPs with AD has been mostly tested in epidemiological studies by comparison of genotype distribution between AD patients and healthy controls (HC). Only a few studies measured levels of cerebrospinal fluid (CSF) AD biomarkers in patients with IL-10 -1082G/A, IL-1β -1473C/G, IL-1α -889C/T, IL-6 -174C/G and TNFα -308A/G genotypes [12,13]. CSF AD biomarkers such as amyloid β 1-42 (Aβ 1-42 ), total tau (t-tau), tau phosphorylated at amino acids Thr 181 (p-tau 181 ), Ser 199 (p-tau 199 ), and Thr 231 (ptau 231 ), and visinin-like protein 1 (VILIP-1) serve as endophenotypes of AD, as they reflect AD-related pathology [14]. CSF Aβ 1-42 [15] and phosphorylated tau proteins [16] are indicators of senile plaques and neurofibrillary tangles in the brain, respectively, while CSF VILIP-1 and t-tau reflect neurodegeneration [17,18]. Here we assessed possible differences in the levels of CSF AD biomarkers (Aβ 1-42 , t-tau, p-tau 181 , p-tau 199 , p-tau 231 and VILIP-1) among patients with different IL-10 -1082G/A, IL-1β -1473C/G, IL-1α -889C/T, IL-6 -174C/G and TNFα -308A/G genotypes to test whether people carrying certain genotypes are more prone to develop AD-related pathology as reflected by their levels of CSF biomarkers.  [19]. In addition to neuropsychological testing, complete blood tests (levels of folic acid (B9), vitamin B12, thyroid function test, serology for Lyme's disease and syphilis) and a full neurological examination were done. Dementia due to AD was diagnosed by using the National Institutes on Aging -Alzheimer's Association (NIA-AA) criteria of McKhann et al. [20], while for MCI diagnosis the criteria of Albert et al. [21] were used. FTD was diagnosed according toNeary et al. [22], while VaD was diagnosed using the criteria of National Institute for Neurological Disorders and Stroke-Association Internationale pour la Recherche et l'Enseignement en Neurosciences (NINCDS-AIREN) [23], and the Hachinski Ischemic Score (HIS) [24]. All procedures were in accord with the Helsinki Declaration [25]

Analysis of CSF biomarkers
CSF was collected by lumbar puncture between intervertebral spaces L3/L4 or L4/L5. CSF was centrifuged at 2,000 g for 10 min, aliquoted and stored at −80°C in polypropylene tubes. Levels of CSF biomarkers were determined by following enzyme-linked

Statistical analysis
Data normality was tested using the Kolmogorov-Smirnov test. However, because of the small number of subjects in some groups, non-parametric statistics were used regardless of the results of the test for normality. Levels of CSF biomarkers were compared among groups using the non-parametric Kruskal-Wallis test. No association between IL-1α -889C/T (rs1800587) polymorphism and CSF biomarkers was detected in any of the analyzed groups.

TNFα -308A/G (rs1800629)
As only three AD patients were carriers of AA TNFα -308 genotype ( Table 1), these patients were grouped together with carriers of AG TNFα -308 genotype. P-tau 231 (U=805.5, Z=-2.220, p=0.026) and VILIP-1 (U=762.5, Z=-2.517, p=0.012) levels were significantly increased in AD patients with GG compared to AA and AG TNFα -308 genotype  Table 2). VILIP-1 levels were also significantly increased in AD patients with GG compared to AG TNFα -308 genotype (K-W post hoc p=0.002; Figure 7, Table 2). Levels of t-tau, p-tau 181 , p-tau 199 , p-tau 231 and VILIP-1 were significantly increased in patients with AA compared to AG TNFα -308 genotype in all patients (when all subjects were grouped together, in AD, MCI patients and HC combined, in AD and MCI patients combined, and in AD patients), while levels of t-tau and VILIP-1 were increased in patients with AA compared to GG TNFα -308 genotype (when all subjects were grouped together and in AD, MCI patients and HC; Table 2, Figure 8). The three AD patients carriers of AA TNFα -308 genotype, could not be evaluated separately and should be validated in a larger of population.
Levels of Aβ 1-42 were decreased, while levels of t-tau were increased in carriers of G allele in IL-1β -1473C/G polymorphism. T-tau levels were also significantly increased in patients with CG IL-1β -1473C/G genotype. P-tau levels were significantly increased in patients with AA IL-10 -1082G/A and GG TNFα -308A/G genotype, and in carriers of G allele in IL-1β -1473C/G and IL-6 -174C/G polymorphisms. Levels of VILIP-1 were increased in patients with CG and GG IL-1β -1473C/G, GC IL-6 -174C/G and GG TNFα -308A/G genotype.
SNPs in genes for IL-1α, IL-1β, IL-6, IL-10 and TNFα can influence transcription and consequently the amount of the produced cytokines [9-11]. Decrease in the amount of anti-inflammatory cytokines and increase in pro-inflammatory cytokines results in increased inflammation, favouring the development of AD [27]. In that way certain genotypes in these SNPs (IL-10 -1082G/A, IL-1β -1473C/G, IL-1α -889C/T, IL-6 -174C/G and TNFα -308A/G) can make some people more vulnerable to the development of neuroinflammation and consequently the development of AD. Given that the production of IL-10 is significantly decreased in carriers of the IL-10 -1082 A genotype [28,29], a decrease in anti-inflammatory cytokine IL-10 levels could result in increased inflammation, favouring the development of AD [27]. It was found that the C IL-6 -174 allele is associated with decrease in IL-6 plasma levels [10] so this genotype could be protective against AD. TNFα being a main pro-inflammatory cytokine, its higher production is associated with increased inflammation and AD progression. TNFα inhibitors have been suggested as potential therapeutics for AD [30]. The influence of TNFα -308 polymorphism on TNFα protein production remains however unclear. Most studies reported that the A TNFα -308 allele is associated with increased production of TNFα [9,31,32], while some studies did not find differences in TNFα protein levels in patients with different TNFα -308 genotypes [33]. Regarding polymorphisms in additional pro-inflammatory cytokines IL-1α and IL-1β that were also tested in this study, it was showed that T allele in the IL-1α -889 polymorphism was associated with increased transcriptional activity in IL-1α gene and  [37][38][39][40][41][42][43]. However, other investigators found no association between IL-10 -1082 polymorphism and AD [44][45][46][47][48][49][50][51][52] or showed GG IL-10 -1082 genotype to be significantly increased in AD patients [53] and AA IL-10 -1082 genotype to decrease the risk for AD [54]. Meta-analyses revealed an association between IL-10 -1082 AA and AG genotype and increased risk for AD [55], and an association between IL-10 -1082 GG genotype and reduced risk for AD [56]. However, the meta-analysis of Mun et al. found no association between IL-10 -1082 polymorphism and AD risk [8]. Our results agree with studies showing association between IL-10 -1082 A genotype and increased risk for AD [11,[37][38][39][40][41][42][43].
Cytokine IL-1β is likely involved in cognitive decline related to inflammation [57]. As such, polymorphisms in IL-1β were studied to assess possible association with AD (for example, IL-1β -511, IL-1β -31 and IL-1β +3953 polymorphisms [8, [58][59][60]). Association of IL-1β -1473G/C polymorphism with AD was assessed in only two studies. There was no significant difference in distribution of IL-1β -1473 genotypes between AD patients and controls [61,62]. In contrast to these studies, we observed levels of various CSF AD biomarkers to be altered in subjects with different IL-1β -1473 genotypes. Our results indicate that IL-1β -1473 polymorphism may represent a consistent marker of AD and that the frequency of IL-1β -1473 genotypes should be further tested on larger AD and MCI cohorts.
Variable results were also obtained from investigations of the association between the TNFα -308A/G polymorphism and AD. Several confirmed that presence of the A allele in the TNFα -308 polymorphism increases the risk for AD [46, [120][121][122], while others found no association between this polymorphism and AD [12,33,47,70,[123][124][125][126][127][128]. Other authors suggested that the A allele in the TNFα -308 polymorphism is protective against AD [13, 129,130]. Meta-analyses also gave inconsistent results. Furthermore, Di Bona et al. [131] did not confirm the association between TNFα -308 polymorphism and AD. The meta-analysis of Lee et al. [7] showed that the A allele in the TNFα -308 polymorphism may be a risk factor for AD in East Asians, but not in Middle Easterners and Europeans.
Wang [132] confirmed that the A allele increases risk for AD in Asians but decreases risk in Northern Europeans. Our study included only three AD patients with the AA TNFα -308 genotype. These three patients had pathological levels of all examined CSF AD biomarkers, except for Aβ   (Table 2). This result remains however inconclusive due to the small sample. We also detected pathological levels of CSF AD biomarkers in patients with the GG TNFα -308 genotype The levels of CSF AD biomarkers in patients with different TNFα -308 genotypes were also investigated by Sarajärvi et al. [13] and Laws et al. [12] Although the genetic analysis of Sarajärvi et al. [13] showed that A allele carriers are less susceptible for AD than GG homozygotes, their analysis of biomarkers in patients with different TNFα -308 genotypes revealed that levels of Aβ 1-42 were pathological in carriers of an A TNFα -308 allele compared to GG homozygotes [13]. This contrasts with our study as we detected pathological CSF levels of p-tau 231 and VILIP-1 in GG homozygotes in comparison to carriers of an A TNFα -308 allele, and we found no differences in CSF Aβ  levels between patients with different TNFα -308 genotype. The findings of Laws et al. support our results [12]. Although the results of our previous genetic study [128] showed no significant difference in distribution of TNFα -308 genotypes between AD patients and HC, in the present study we detected pathological levels of CSF p-tau 231 and VILIP-1 in AD patients with the GG compared to AG TNFα -308 genotypes. Other groups also did not detect a difference in distribution of TNFα -308 genotypes between AD patients and HC, but observed difference in distribution of haplotypes (that include the TNFα -308 polymorphism) between AD patients and HC [130,133]. Thus, the scope of our next study should be analysis of TNFα haplotypes' distribution between AD patients and HC. Our study suggest that heterozygosity in TNFα -308 polymorphism could be protective against AD, as pathological levels of CSF AD biomarkers were detected in both AA and GG TNFα -308 homozygotes. This deserves further validation.
However, the results on measurement of these and other inflammatory markers in body fluids were inconsistent [134]. Thus, recently a lot of meta-analyses were conducted with purpose to determine the potential of inflammatory markers as biomarkers of AD. The increase in IL-6 was associated with all-cause dementia, but not AD in meta-analyses of Darewwsh et al. [135] and Koyama et al. [136] Additional meta-analyses observed increase in peripheral IL-6, IL-1β [137][138][139] and TNF-α [139] in AD patients compared to HC. However, meta-analyses of Saleem et al. [140] and Su et al. [138] observed no significant difference in inflammatory markers between MCI patients and HC. Brosseron et al. [134] divided inflammatory markers measured in body fluids into three groups by involvement in the disease; 1) cytokines unchanged during disease (like IL-1α), 2) cytokines that increase slightly but steadily during disease (like IL-1β, IL-6, and TNF-α) and 3) cytokines that have a peak when MCI converses to AD.

Conclusions
In conclusion, our study reveals altered levels of CSF AD biomarkers in carriers of

Consent for publication
All patients gave consent for publication.

Availability of data and materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.