Imunogenetica bolii Alzheimer: antigenul leucocitar uman

Introduction

Alzheimer’s disease (AD), the most frequent major neurocognitive disorder, doubled in prevalence in the past 25 years, and it is predicted to become the sixth leading cause of early death globally by 2040⁽¹⁾. Being a major public health problem, it needs new predictive, preventive and curative measures. Since the currently available treatment options are considered rather symptomatic (i.e., cholinesterase inhibitors, NMDA receptor blockers)⁽²⁾, disease-modifying agents, mirroring a more in-depth understanding of the pathogenic mechanisms, are needed^(3,4).

Some well-established pathogenic theories are neuronal death and synaptic loss that result from the formation and accumulation of the extracellular amyloid plaques and the intraneuronal neurofibrillary tangles through the amyloid cascade and the hyperphosphorylation of tau-protein, starting from the amyloid precursor protein (APP). Mutations of the genes coding for the APP (chr 21), or for the proteases involved in the process (PRES1 – chr 14 and PRES2 – chr 1) proved to be disease-causing mutations^(5,6), but are considered to be rather rare, accounting for the familial, autosomal dominant form of AD – i.e., the early-onset form, that represents less than 1% of all cases of AD⁽⁸⁾.

APOE e4 allele (chr 19) was found to be the most important genetic risk factor for late-onset AD (LOAD), although the pathophysiological pathway remains largely unknown. APOE is a major apolipoprotein expressed in the brain, playing a role in mediating synaptogenesis, synaptic plasticity and neuroinflammation⁽⁷⁾; however, mutations can be found only in 50% of patients⁽⁸⁾.

We can therefore conclude that none of these already established theories are necessary, nor sufficient for AD pathogenesis^(9,10).

The extracellular amyloid plaques and the intraneuronal neurofibrillary tangles deposits in the brain act as antigens. They activate microglia, the brain’s resident macrophage, and astrocytes, initiating the inflammation cascade⁽¹¹⁾. This results in the invasion of peripheral monocytes, neutrophils and CD8 T cells disrupting the brain blood barrier, a disruption that has been found to be a critical component of AD pathogenesis⁽¹²⁾. The macrophage phagocytic dysfunction results in the apoptosis of the macrophage, further accumulating the amyloid deposits⁽¹³⁾. The released proinflammatory cytokines (IL-18, IL-1b, TNF-a, ROS, NO etc.⁽¹⁵⁾) can increase the production of APP and amplify the amyloid cascade, thus entering a self-reinforcing vicious circle^(14,16,17) of chronic inflammation, referred to as abnormal inflammatory status⁽¹³⁾, that leads to synapse loss⁽¹⁸⁾.

The inflammation pathogenic theory of the Alzheimer’s disease has been studied for the last three decades to try to fill the gap in the state of the art^(19,20), some authors even stating that it begins decades before AD clinical onset⁽²¹⁾.

It is supported by observational studies that found increased infiltrations of monocytes/macrophages, lymphocytes and T cells in the brain of AD animal models and humans^(22,23). It is also supported by epidemiological studies showing a low incidence of AD in patients with long-term anti-inflammatory treatment^(14,15,24), although it could not be proven in clinical trials⁽¹⁴⁾. Furthermore, it is supported by shared genetic variants between some autoimmune disease and AD^(25,26). Recent genome-wide association studies identified more than 30 genetic loci for AD, many related to immune response (e.g., CD33, INPP5D, CLU, TREM2, CR1, SPI1, ABCA7, EPHA1, MS4As, INPP5D, MEF2C, HLA-DRB5-DRB1)^(27,28).

One key player involved in the immune response of this chronic neuroinflammation is the human leukocyte antigen (HLA), the human equivalent of the major histocompatibility class (MHC). The amyloid plaques that accumulate in the AD patients’ brain are internalized by the microglia and processed into pieces that, in combination with a specific HLA class I/II molecule, are presented to the T lymphocytes⁽²⁹⁾. The upregulation of HLA class II antigens is now accepted as a marker of the activated microglia⁽³⁰⁾. The resulted combination also stimulates B lymphocytes to secrete antibodies against amyloid peptides. Thus, MHC is a co-stimulatory molecule in the neuroinflammation cascade, one branch of the AD pathophysiology⁽²⁹⁾.

This paper reviews the current evidence for the involvement of the HLA system in the pathogenesis of Alzheimer’s disease. We searched the PubMed database using the MeSH terms “HLA” and “Alzheimer’s disease” for anytime published, full-text English written articles. Out of the 375 results given, we included the relevant once for the aim of this review.

The human leukocyte antigen system

The human leukocyte antigen (chr 6p21), first mapped in 1993⁽³¹⁾ and 1999⁽³²⁾, is the densest region of the human genome, with more than 200 genes^(33,34) and approximately 3.5 million base pairs⁽³⁵⁾. Its main function is the presentation of short peptides to specialized immune cells, playing a key role in innate and adaptative immunity.

It is divided into five subregions with different functions (Table 1)^(26,33,35).
 

It is one of the most polymorphic regions in the human genome, up to date, 26,889 HLA alleles being named (19,578 class I and 7,302 class II)⁽³⁶⁾.

These code for a and b polypeptide chains that form an antigen binding groove: 2 a for HLA class I, a and b for HLA class II⁽³⁴⁾. Change in even one amino acid results in different antigen binding groove, thus a different HLA antigen that can bind a different pathogen⁽³⁴⁾. The match between the pathogen and the HLA groove is essential for the immune response^(37,38). In the absence of a match, the pathogen can persist, causing inflammation, cell damage and autoimmunity⁽³⁷⁾. Therefore, it is likely that the polymorphism of this region represents an evolutionary adaption to the multitude of pathogens faced⁽³⁴⁾.

HLA class I and AD

In the central nervous system, HLA class I are highly expressed in the microglia and endothelial cells, lower levels being expressed in astrocytes, oligodendrocytes and neurons (where they play a role in synaptic plasticity)⁽³⁹⁾.

A great amount of research focused on HLA-A, especially on its most common allele HLA-A2, trying to associate it with AD pathogenesis or onset age. The results were highly inconsistent over more than two decades^(40-50), probably due to small sample size, different HLA typing methods, participants’ ethnicity or study designs. More recent studies focusing on HLA-A SNPs tried to associate them with atrophy of cortical brain structures (rs9260168 and rs3823342 with left para-hippocampus, as seen on MRI), or subcortical structures (rs76475517 with amygdala). The study sample was obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and included 812 participants (281 health controls, 483 participants in the mild cognitive impairment stage, and 48 AD participants), but the results were at a marginal significance (Pc=0.054)⁽⁵¹⁾.

HLA-B and HLA-C are relatively understudied. HLA-B7 and HLA-Cw*0702 were associated with an increased risk for AD (OR 2.3; 95% CI; 1.4-3.7; p=0.001), but not with AD onset age, in a case-control Caucasian population study, annually followed-up for up to 15 years. HLA-B7 and HLA-Cw*0702 being in tight linkage disequilibrium, the true risk loci could not be determined⁽⁵²⁾.

HLA-B8 was found to be increased in volunteers with impaired cognition. HLA-B8 is part of the so-called “autoimmune” ancestral haplotype (HLA-DQ2, HLA-DR3, HLA-B8, HLA-Cw7 and HLA-A1) which is carried by most Caucasians and known to be positively associated with autoimmune disorders and by elevated circulating levels of inflammatory cytokines, such as IL-1 and TNF‑a⁽⁵³⁾. HLA B16 was found to be associated with AD in a Japanese study⁽⁵⁴⁾. Newer studies found HLA B*07:00/x genotype being a risk factor for sporadic LOAD, in an Iranian study⁽⁵⁵⁾. Furthermore, cases with this genotype were found to have worse clinical response at two-year follow-up after rivastigmine treatment than in non-carriers⁽⁵⁵⁾.

Regarding non-classical HLA, HFE (homeostatic iron regulator) SNP rs1800562 correlated with slower atrophy rate of the right middle temporal lobe in a study sample obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) that included 812 participants (281 health controls, 483 participants in the mild cognitive impairment stage, and 48 AD participants; Pc=0.003)⁽⁵⁶⁾.

HLA class II and AD

In the central nervous system, HLA class II are highly expressed in the microglia, and with lower levels in astrocytes and endothelial cells⁽³⁹⁾. Both their predisposing and protective role in the AD pathophysiology have been investigated in the recent years.

The first gene of this class that was found to be a risk factor for AD by a large genome-wide association study (GWAS). European ancestry based meta-analysis was the HLA-DRB5-DRB1 region⁽²⁸⁾. HLA-DRB1 and HLA-DRB5 expressed in the microglia was found to be positively correlated with measures of AD pathology⁽⁵⁷⁾.

HLA-DR1, HLA-DR2 and HLA-DR3 were associated with a high risk of AD in two studies: one case-control study of brain tissues of 78 patients with late-onset AD and 50 controls, that found even higher significance in ApoE4 (-)⁽⁵⁸⁾, respectively one study based on ADNI data⁽⁵⁹⁾.

HLA-DPB1⁽⁶⁰⁾ as well as HLA-DRA⁽²⁵⁾ were associated with a higher risk of AD in two GWAS.

HLA-DRB1*15 was found to be a risk factor for AD (OR 5.4; 95% CI; 2.7-10.8) in a Tunisian population study, as well as DRB1*04 (OR 1.9; 95% CI; 1-3.4), DQB1*05 (OR 0.36; 95% CI; 0.18-0.76), DQB1*06 (OR 3.8; 95% CI; 2.2-6.8), and haplotypes DRB1*1501/DQB1*0602 (OR 5.4; 95% CI; 2.7-10.9) and DRB1*0402/DQB1*0302 (OR 2.9; 95% CI; 1.4-5.7)⁽⁶¹⁾.  HLA-DRB1*03 was found to be a risk factor for late-onset AD in a German population study⁽⁶²⁾.

HLA-DRB1/DQB1 SNPs rs35445101, rs1130399 and rs28746809 were associated with a smaller volume of the left posterior cingulate cortex, as seen on MRI, and rs2854275 was associated with a larger volume. These were found in mild cognitive impairment (MCI) or cognitively normal stage too, suggesting that these SNPs might play a role in disease development from early stages – i.e., MCI or normal cognitive state⁽⁵⁹⁾.

The haplotype A*03:01~B*07:02~DRB1*15:01*DQA1*01:02~DQB1*06:02 was also significantly associated with AD in a 11,690 white individuals multicentric study. Class I and class II components separately remained at a significant association⁽⁶³⁾.

Microglia HLA-DR immunostaining was found to increase with the clinical progression of AD, in the gray matter of entorhinal cortex and in the hippocampus, based on one postmortem study of brain sections of AD and control cases⁽⁶⁴⁾.

On the other hand, some HLA class II have been found to offer protection for AD development. HLA-DR4, DR6 and DR9 were associated with a low risk of AD^(33,59). HLA-DR4 proved to protect against high levels of glial fibrillary acidic protein, a specific astrocyte marker in AD hippocampus, in a small sample size post-mortem study⁽⁶⁵⁾.

HLA DRB1*13:02 proved to protect against total, cortical and subcortical gray matter volume reduction, but the study sample was rather small, and it consisted of only female participants⁽³⁷⁾. HLA DRB1*13:02 also proved to protect against neural network variability, that is considered to reflect brain dysfunction, even in the presence of ApoE4 risk allele, but, again, the study sample consisted of only females⁽⁶⁶⁾. HLA DRB1*13:01 that differs from HLA DRB1*13:02 by a single nucleotide had mixed results⁽⁶⁶⁾. Patients carrying HLA-DRB1*04:00/X genotype had a better clinical outcome at the two-year follow-up after rivastigmine treatment than non-carriers⁽⁵⁵⁾.

HLA DQB1*06:00/x genotype proved to be protective against LOAD, the cases with this genotype having a milder severity that non-carriers, and had a better clinical response at the two-year follow-up after rivastigmine treatment⁽⁵⁵⁾.

Moreover, it was observed that longer the AD progressed, the more decreased were the levels of HLA class II⁽⁶⁷⁾.

Some common HLA associations between AD and autoimmune diseases have been found, proving an overlap of pathogenic pathways. HLA-DRB5 SNPs rs2516049 was found both in AD and psoriasis, rs12679874, rs2570088, rs16980051 and rs2298428 were found in Crohn disease, while rs2280231, rs8055533 and rs7258465 were found in type 1 diabetes, all in large-sample data from GWAS⁽²⁵⁾. Rs2516049 was also associated with higher neurofibrillary tangle accumulation⁽²⁵⁾.

HLA class III and AD

TNF-a, one of the main proinflammatory cytokines, proved be a risk factor for AD. It was associated with a higher atrophy of both total brain and hippocampal volume than expected for age⁽⁵⁶⁾.

On the other hand, in a study based on ADNI data, TNF-a rs2534672 and rs2395488 were associated with a larger volume of the middle temporal lobe, thus offering protection for AD⁽⁵⁶⁾. 

RAGE (receptor for advanced glycation end products) rs2070600 was associated with progressive atrophy of the right hippocampus-CA1⁽⁵⁶⁾.

Discussion and conclusions

Alzheimer’s disease is the result of normal aging plus abnormal pathological injuries⁽⁵⁹⁾. Some of these pathological injuries were found to be inflammation-related, being referred to as “inflammaging”⁽⁶⁸⁾ or immunosenescence⁽⁶⁸⁾. Increasing evidence shows that neuroinflammation is the primary cause of neurodegeneration⁽⁶⁹⁾.

Although it is uncertain whether this inflammatory mechanism can cause neurodegeneration or it is present to remove the debris associated with already ongoing neurodegenerative mechanisms, up to date data state that the former is more likely.

The HLA role of this inflammaging is relatively understudied up to date, but data could fill the gap in the state of the art. Although, because of the polymorphisms of HLA genes, different levels of expressions and differences between ethnic and geographical populations exist, it is difficult to find strong associations⁽²⁹⁾, some HLA gene variants being established as intermediate risk factors for Alzheimer’s disease ⁽³⁹⁾.

A better understanding of neuroinflammation processes could open the path to new therapeutic approaches, such as inflammation resolution agents^(14,70) or immunotherapy^(71,72).