The authors have declared that no competing interests exist.
Conceived and designed the experiments: HB LG JF BD. Performed the experiments: AS. Analyzed the data: AS LL PS HB. Contributed reagents/materials/analysis tools: RR DC WF RF. Wrote the paper: AS BD HB.
Alzheimer’s disease (AD) is the most common cause of dementia in the human population, characterized by a spectrum of neuropathological abnormalities that results in memory impairment and loss of other cognitive processes as well as the presence of non-cognitive symptoms. Transcriptomic analyses provide an important approach to elucidating the pathogenesis of complex diseases like AD, helping to figure out both pre-clinical markers to identify susceptible patients and the early pathogenic mechanisms to serve as therapeutic targets. This study provides the gene expression profile of postmortem brain tissue from subjects with clinic-pathological AD (Braak IV, V, or V and CERAD B or C; and CDR ≥1), preclinical AD (Braak IV, V, or VI and CERAD B or C; and CDR = 0), and healthy older individuals (Braak ≤ II and CERAD 0 or A; and CDR = 0) in order to establish genes related to both AD neuropathology and clinical emergence of dementia. Based on differential gene expression, hierarchical clustering and network analysis, genes involved in energy metabolism, oxidative stress, DNA damage/repair, senescence, and transcriptional regulation were implicated with the neuropathology of AD; a transcriptional profile related to clinical manifestation of AD could not be detected with reliability using differential gene expression analysis, although genes involved in synaptic plasticity, and cell cycle seems to have a role revealed by gene classifier. In conclusion, the present data suggest gene expression profile changes secondary to the development of AD-related pathology and some genes that appear to be related to the clinical manifestation of dementia in subjects with significant AD pathology, making necessary further investigations to better understand these transcriptional findings on the pathogenesis and clinical emergence of AD.
Alzheimer’s disease (AD) is the most frequent dementing disorder in the elderly and is characterized by a progressive decline in memory and other cognitive domains
However, there is a growing body of evidence that suggest the accumulation of Aβ in the brain in cognitively healthy older subjects. Post-mortem studies showed that a significant proportion of subjects with neuropathological diagnosis of AD did not have any evidence of cognitive impairment in the last assessment prior to death
The study of brain mRNA expression profile may help to disentangle some of the molecular mechanisms related to the development of AD-related pathology and the manifestation of clinical dementia in these subjects. Previous studies showed a significant deregulation in the expression of genes related to the energy metabolism
Nevertheless, no study to date have addressed whether specific patterns of gene expression is associated to the development of AD-related pathology and to the clinical manifestation of dementia in subjects with significant AD pathology. Therefore, the aims of this study were: (1) to determine which genes and their biological functions are related to the development of AD pathology; (2) to determine which genes and their biological functions are related to the clinical manifestation of dementia in subjects with significant AD pathology.
After death, a trained gerontologist interviewed a knowledgeable informant who had at least weekly contact with deceased subjects. Past medical history, cognitive performance and functional status was determined for each subject
At autopsy, hippocampus was dissected and frozen at −80°C. Hippocampal specimens were obtained from the Brain Bank of the Brazilian Aging Brain Study Group
Neuropathological examinations were performed using immunohistochemistry according to internationally accepted criteria
Based on pathological and clinical criteria, subjects were categorized into three groups: I) 9 subjects with neuropathological AD (Braak ≥ IV and CERAD = B or C), and clinical dementia (CDR ≥1), termed “
Sample ID | Gender | Age | Braak | CERAD | CDR | PMI |
CP-AD1 | F | 99 | V | B | 3 | 18.3 |
CP-AD2 | F | 82 | IV | B | 2 | 13.5 |
CP-AD3 | F | 86 | IV | C | 1 | 10.1 |
CP-AD4 | F | 83 | V | A | 3 | 12.1 |
CP-AD5 | M | 69 | VI | C | 2 | 15.0 |
CP-AD6 | F | 87 | V | B | 3 | 17.7 |
CP-AD7 | F | 82 | V | C | 2 | 11.8 |
CP-AD8 | F | 77 | IV | A | 3 | 16.0 |
CP-AD9 | F | 83 | VI | C | 3 | 10.8 |
P-AD1 | F | 87 | V | C | 0 | 11.1 |
P-AD2 | F | 85 | V | C | 0 | 9.6 |
P-AD3 | M | 72 | VI | C | 0 | 11.7 |
P-AD4 | F | 86 | VI | C | 0 | 16.0 |
N1 | F | 71 | 0 | 0 | 0 | 16.1 |
N2 | M | 79 | I | 0 | 0 | 8.3 |
N3 | F | 81 | I | 0 | 0 | 11.9 |
N4 | M | 77 | I | A | 0 | 6.5 |
N5 | M | 57 | 0 | 0 | 0 | 9.8 |
N6 | F | 65 | 0 | 0 | 0 | 12.7 |
N7 | F | 59 | 0 | 0 | 0 | 14.0 |
N8 | M | 89 | II | 0 | 0 | 14.2 |
N9 | F | 82 | 0 | 0 | 0 | 14.8 |
N10 | F | 94 | II | 0 | 0 | 12.3 |
Subjects were divided in three groups according to neuropathological and clinical criteria: clinic-pathological Alzheimer’s disease (CP-AD), pathological/preclinical Alzheimer’s disease (P-AD), and normal older individuals (N). Sample ID, sample identification; Age, age at death in years; F, female; M, male; Braak, Braak stage; CERAD, Consortium to Establish a Registry for Alzheimeŕs Disease score; CDR, Clinical Dementia Ratio score; PMI, post-mortem interval in hours.
This study was approved by the Ethical Board for Research Project Analysis (CAPPesq) of the University of Sao Paulo Medical School (research protocol 285/04) and was conducted in accordance to the Helsinki Declaration.
Total RNA was isolated from the frozen hippocampus by the RNeasy Mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. RNA purity and yield were determined by UV spectrophotometry for all RNA samples (
Reverse transcription was performed by adding 1× first strand buffer and 0.01 mol/l dithiothrectol (Invitrogen Life Technology, Carlsbad, CA, USA), 2 µl diethylpyrocarbonate (DEPC; Sigma, St Louis, MO, USA) treated water, 40 U rRNasin (Promega, Madison, WI, USA), 1 mmol/l dNTP (Amersham Biosciences, Piscataway, NJ, USA), and 400 U SuperScriptTM II Reverse Transcriptase (Invitrogen Life Technology) to a final volume of 20 µl. The reaction was incubated for 120 minutes at 42°C. Second-strand synthesis was performed by adding 53 µl of DEPC-treated water, 20 µl of 5× second strand buffer (Invitrogen Life Technology), 1 mmol/l dNTP, 1 U RNase H (Invitrogen Life Technology), 10 U
The double strand cDNA (dscDNA) was stopped by adding 0.05 mol/l EDTA. UltraPureTM Phenol (Invitrogen, Carlsbad, CA, USA):chloroform:isoamyl alcohol (Merck), at a ratio of 25∶24∶1 and a pH of 8.0, was used for cDNA purification. The dscDNA was precipitated with absolute ETOH (Merck) and resuspended in 10 µl DEPC-treated water. The dscDNAs were subjected to in vitro transcription using reagents from RibomaxTM Large scale RNA production system T7 kit (Promega), in accordance with the manufacturer’s recommendation. The amplified RNA (aRNA) was reverse transcribed into cDNA using 9 µg random hexamer (dN6; Amersham Bio- sciences, Little Chalfont, UK). cDNA synthesis was continued with the same conditions used in the first strand of the first round. The second strand was synthesized using Advantage® cDNA Polymerase (Clontech, Mountain View, CA, USA), and purification was performed in accordance with the methodology cited above.
The aRNA quality, in terms of purity and integrity (
Labeled cDNA was generated in a reverse transcriptase reaction in the presence of 7 µg of amplified RNA (aRNA), 9 µg of a random hexamer primer (Invitrogen Life Technologies, Carlsbad, CA), Cy3- or Cy5-labeled dCTP (Amersham, Biosciences, Little Chalfont, UK), and 400 U SuperScriptTM II Reverse Transcriptase (Invitrogen Life Technology). The residual dye was removed using illustra AutoSeq™ G-50 (GE Healthcare, Little Chalfont, UK
After hybridization, slides were washed as follows: 2× Saline Sodium Citrate (SSC) for 10 minutes, 0.1× SSC/0.1% SDS for 10 minutes (two times), and 0.1× SSC for 10 minutes (two times) at 37°C. All solutions were pre-heated to 42°C. Hybridized arrays were scanned on the ScanArrayTM Express (Packard BioScience Biochip Technologies, Billerica, MA, USA), and Cy5/Cy3 signals were quantified using the histogram method with ScanArray Express software (Perkin-Elmer Life Sciences, Boston, MA, USA). Fluorescent intensities of Cy5 and Cy3 channels on each slide were subjected to spot filtering and normalization. We first eliminated all saturated points (≥63,000; approximately 16 bits) and performed a local background subtraction, considering for analysis only those spots with positives values. Normalization was performed using locally weighted linear regression within and across arrays for inter-slide normalization. After normalization, data for each gene were reported as the logarithm of the expression ratio used to represent the relative gene expression levels in the experimental samples. The raw data from hybridizations and experimental conditions can be obtained at the Gene Expression Omnibus under accession number GSE13214. A detailed description of the platform array is available in accession number GPL1930.
For analysis of genes related to pathological changes, individuals with AD pathology (CP-AD and P-AD) were compared to individuals without pathology (N). To identify genes implicated with clinical manifestation of dementia, individuals who present AD neuropathology but differ on the clinical status were compared (CP-AD versus P-AD). For both analyses, we first used Student’s
To see more properties implicated with the differentially expressed genes and their partners, we used a network approach. By querying three human interactome databases (HPRD
Classifiers to separate the CP
To address which genes were related to the neuropathological processes of AD (“neuropathological AD-related genes”, npADGs), we compared the gene expression profile (Student’s
regulation of transcription, DNA-dependent (P-value 0.0317) | C-terminal protein amino acid modification (P-value 0.007) |
zinc finger protein 266 ( |
isoprenylcysteine carboxyl methyltransferase ( |
general transcription factor IIH, polypeptide 1, 62 kDa ( |
plasminogen activator, urokinase receptor ( |
AF4/FMR2 family, member 3 ( |
|
zinc finger and BTB domain containing 7B ( |
GLE1 RNA export mediator homolog (yeast) ( |
helicase-like transcription factor ( |
RAE1 RNA export 1 homolog (S. pombe) ( |
zinc finger protein 84 ( |
|
zinc finger protein 576 ( |
synemin, intermediate filament protein ( |
transformation/transcription domain-associated protein ( |
|
zinc finger protein 394 ( |
glutamate receptor, ionotropic, kainate 5 ( |
zinc finger protein 559 ( |
|
lysine (K)-specific demethylase 2B ( |
v-myc myelocytomatosis viral oncogene homolog (avian) ( |
|
|
helicase-like transcription factor ( |
carbohydrate (N-acetylgalactosamine 4-sulfate 6-O) sulfotransferase 15 ( |
ubiquitin-conjugating enzyme E2A ( |
|
transformation/transcription domain-associated protein ( |
Werner helicase interacting protein 1 ( |
lysine (K)-specific demethylase 2B ( |
|
|
acid phosphatase 5, tartrate resistant ( |
nucleoporin 50 kDa ( |
|
GLE1 RNA export mediator homolog (yeast) ( |
nicotinamide nucleotide transhydrogenase ( |
RAE1 RNA export 1 homolog (S. pombe) ( |
|
|
glutamate receptor, ionotropic, kainate 5 ( |
nucleoporin 50 kDa ( |
|
RAE1 RNA export 1 homolog (S. pombe) ( |
acid phosphatase 5, tartrate resistant ( |
|
|
isoprenylcysteine carboxyl methyltransferase ( |
|
translocase of inner mitochondrial membrane 9 homolog (yeast) ( |
Biological process categories significantly overrepresented by npADGs (P<0.05, Fisher’s exact test). Other similar significant categories are not included to reduce redundancy.
Transcription factor (TF) processes were among the largest categories of npADGs. In addition, one of the hallmarks of AD, reduced energy metabolism, was reflected by categories of down-regulated npADGs (
Searching for an expression pattern that could differentiate individuals with AD histopathology from normal individuals, a hierarchical clustering was carried out for npADGs (77 genes). However, it was not possible to obtain a clear separation pattern between (CP-AD + P-AD) and N groups (
When the hierarchical clustering was performed using npADGs with
Hierarchical clustering was performed by using the expression values from the genes related to AD neuropathology with
Aiming to explore the 47 genes identified in the clustering analysis and their connectors an interaction network was constructed. Genes were mapped on the human “interactome” and then only those in our array platform were selected (301 genes). We looked for significant differences in the average PCC of a gene and their interacting partners in subjects who presented AD pathology (CP-AD and P-AD groups) and those who were free of such histopathology. This metric gives an estimate of the difference in correlation of each interaction around a gene between the two groups (AD pathology vs. controls). This revealed 25 genes that displayed altered PCC as a function of presence or absence of AD neuropathology (
(
Of the 25 significant genes identified in the network,
Analysis of interactions between the 25 significant genes and their partners revealed that they form an interconnected network, and a functional analysis of these genes demonstrated overrepresentation of some GO categories involved with transcriptional regulation, DNA damage, inflammatory signaling, cell adhesion, neuron differentiation, and neuron apoptosis (
When individuals establish substantial neuropathological changes of AD, some of them develop the clinical dementia syndrome, while others remain asymptomatic for a long period, i.e. the preclinical stages of AD. Thus, to assess which genes might be related to the clinical expression of AD (“clinical AD-related genes”, cADGs), we compared the gene expression profile (Student
As our aim was to find out transcriptional differences between CP-AD and P-AD individuals that could not be achieved by differential gene expressions in our study based on our small sample size, we searched for classifiers that were able to completely distinguish these groups utilizing another approach, which did not consider the DEGs, but started using all the genes in the microarray platform and applying a mathematical model design for small sample sizes.
Six genes selected by feature selection were used to generate the classifiers between those groups:
(
Gene expression profile changes related to AD pathology are implicated with energy metabolism, oxidative stress, DNA damage and transcriptional regulation. Once established of significant AD pathology, some genes involved with synaptic plasticity, and cell cycle appear to be involved with the clinical outcome of the illness and might represent the molecular mechanisms that underlie the cognitive reserve. CP-AD, clinic-pathological Alzheimer’s disease; P-AD, pathological/preclinical Alzheimer’s disease.
Microarray studies comparing AD subjects with normal elderly individuals have uncovered multiple pathophysiological processes that have been implicated in AD, including energy metabolism
The AD neuropathology-related gene set was involved in some important functions suspected of a role in AD and brain aging, in particular energy metabolism and oxidative stress (hexose transport, hexose biosynthetic process, acetyl-CoA catabolic process, negative regulation of reactive oxygen species metabolic process), immune function (negative regulation of interleukin-1 production), DNA repair (DNA synthesis involved in DNA repair), senescence (regulation of telomere maintenance) and transcriptional regulation (regulation of transcription, DNA-dependent, and chromatin modification). We have to consider that these functions/gene expression might be disrupted as an early response to the increased accumulation of Aβ peptide and tau observed in the CP-AD and P-AD individuals
Glucose metabolism is impaired in AD brain
From the AD pathology-related DEGs (77 genes), we also identified a gene set (47 genes) providing two clear patterns between individuals with neuropathology and normal subjects. These 47 genes were utilized to construct an interaction network. Network science deals with complexity by “simplifying” complex systems, summarizing them merely as components (nodes) and interactions (edges) between them. The resulting “interactome”, the networks of interactions between cellular components, can serve as scaffold information to extract global or local graph theory properties. Once shown to be statistically different from randomized networks, such properties can then be related back to a better understanding of biological processes
Regarding to genes related to clinical manifestation of dementia in brains with substantial AD histopathology, we compared individuals that have been clinically and histopathologically confirmed to have AD (CP-AD) with individuals who did not fulfill clinical criteria for AD but demonstrate high levels of AD-related pathology (P-AD). With the bias of a limited statistical power, a reliable expression change could not be detected, and so, CP-AD and P-AD were transcriptionally indistinguishable using a statistical test with this small sample size.
However, we utilized a classification approach, which does not consider the differentially expressed genes, but it starts with all genes of the array platform, to discriminate CP-AD and P-AD individuals. Disease classification is another approach already used in the molecular diagnosis and classification of several illnesses, including AD
Therefore, we have added some evidence to a hypothesis model in which relatively independent processes contribute to the AD pathology and AD clinical manifestation (
Limitations of this work are comprised by both small sample size and gender unbalance. Searching the DEGs taking out men or women, as in leave-one-out analysis, the final list of genes presents variations independently of gender (data not shown). To overcome these limitations, we used different approaches to find relevant genes: 1) hierarchical clustering and network analysis using DEGs, and 2) classification analysis, not using the DEGs, but starting with all genes in the array.
Further studies with larger sample sizes are necessary to better understand the pathogenic mechanisms of early stages of AD, and to discover pre-clinical biomarkers and rational therapeutic targets. To this end, studies with pre-symptomatic animal models could be of extreme importance on developing of time or stage-dependent interventions to achieve optimal results in delaying the progression of AD-related pathological changes or clinical symptoms of dementia.
(TIF)
(PDF)
(PDF)
(PDF)
(PDF)
(PDF)
We thank Louise Danielle Mota, Waleska Martins, and Elen Bastos for technical assistance with the experimental procedures.