ReviewProteomics as a reliable approach for discovery of blood-based Alzheimer’s disease biomarkers: A systematic review and meta-analysis
Graphical abstract
Introduction
The pathophysiology of Alzheimer’s disease (AD), which is known to begin years before abnormal changes in the brain occur, raises the need for early detection. Identification of AD biomarkers during its asymptomatic stage is crucial for effective disease modification (Schneider et al., 2009), which is currently hindered by the various constraints in diagnosis and the nature of the pathogenesis of the disease. At present, validated AD biomarkers are predominantly derived from the cerebrospinal fluid (CSF) or measured using neuroimaging (Leinenbach et al., 2014). Validated CSF-based biomarkers include reduced amyloid beta 42 fragment (Aß42) and increased phosphorylated tau (p-tau). Potential imaging biomarkers that have been proposed are glial inflammation, epigenomic alterations, structural and functional alterations as well as synaptic and cellular degeneration which can be accessed via positron emission tomography (PET) or functional magnetic resonance imaging (fMRI) (Aël Chetelat and Baron, 2003; Márquez and Yassa, 2019; Sperling et al., 2011). Although these biomarkers are sensitive and specific for diagnosis of AD at an early stage, their applicability at clinical setting is impeded by limitations like invasive procedure, affordability and unavailability in most clinics.
The disrupted blood brain barrier (BBB) in AD allows small molecules associated with neurodegeneration to leak into the periphery (Sweeney et al., 2018; Zipser et al., 2007), making blood-based AD biomarkers possible. The feasibility of blood-based biomarkers is associated with better patient acceptance and cost-effectiveness. This would enable population-based screening as the first step in a multistage diagnostic procedure, which would in turn aid early detection (Henriksen et al., 2014; O’Bryant et al., 2017). Earlier efforts in the search for AD blood-based biomarkers involved targeted discovery that is focused on specific candidate biological molecules in pathways known to be related with AD pathogenesis, namely Aβ, amyloid precursor protein (APP), cholesterol metabolism, oxidative stress and inflammation (Chen-Plotkin, 2014; Schneider et al., 2009). Yet, the usefulness of the respective biological molecules as single biomarker that discriminate AD and MCI from healthy aging individuals remains inconclusive (Fukumoto et al., 2003; Mayeux and Schupf, 2011; Schneider et al., 2009; Tamaoka et al., 1996). This has called for an untargeted or unbiased discovery whereby multiple to several hundreds or even thousands of biological molecules are studied simultaneously with advanced bioinformatics and statistics that can facilitate vast data analysis.
Proteome, the set of proteins in a given time and space with varied composition in different cells or tissues, is commonly studied in untargeted biomarker discovery (Tambor et al., 2010). Proteomic approach, which is clinically relevant, has been actively used to search for diagnostic biomarkers of various diseases via platforms like mass spectrometry [Matrix-Assisted Laser Desorption/Ionization-Time of Flight (MALDI-TOF)], shotgun proteomics (Surface-Enhanced Laser Desorption/Ionization-Time of Flight (SELDI-TOF), Liquid Chromatography coupled with Tandem Mass Spectrometry (LC–MS/ MS) and Isobaric Tags for Relative and Absolute Quantification (iTRAQ), electrophoresis (sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), 2D-PAGE and 2D-differential gel electrophoresis (DIGE)] and immunoassays (ELISA, western blot) (Tambor et al., 2010). The utilization of untargeted proteomic technology, in particular, has already resulted in promising discoveries of biomarker panels with good diagnostic values and reflective of AD pathology (Henkel et al., 2012; Kitamura et al., 2017; Kumar et al., 2018; Ray et al., 2007; Soares et al., 2009, 2012; Thambisetty et al., 2010; Zhao et al., 2015). Nevertheless, blood-based AD biomarkers remain to be established. The main challenge appears to be failure to replicate results of similar accuracy during the validation phase (Henriksen et al., 2014; Snyder et al., 2014). As part of the effort in examining the replicability of discovered biomarkers, Kiddle et al. (2014) found α-1-antitrypsin, α2M, ApoE and complement C3 (C3) to be reported in five independent study cohorts. The authors, however, did not report the regulation level of the biomarkers. The discovery of similar AD biomarkers from multiple studies could be promising but their utility would remain questionable when there were inconsistencies at regulation level. Such inconsistencies were prevalent even amongst widely scrutinized biomarkers in blood proteome like Aβ40, Aβ42 and ApoE, making them unreliable AD blood-based biomarkers (Rembach et al., 2013; Song et al., 2011; Wang et al., 2014). The present systematic review and meta-analysis were performed in the context of blood-based AD CB, focusing on untargeted discovery approach (to avoid being biased to any CB) and proteomic platforms (proteins play ubiquitous role in diseases) (Rifai et al., 2006). The replicability and regulation of AD CB were assessed in terms of study cohorts, blood in general, blood fractions, proteomic platforms and continents.
Section snippets
Literature search strategy
Articles published between July 1984 (when the first AD diagnostic criteria were published) and July 2019 were searched through five electronic databases, namely PubMed/ Medline (based on medical subject headings (MESH terms) and free text strings), EMBASE, Web of Science, Science Direct and Springer Link. The Boolean search involved the combination use of terms like “Alzheimer”, “Alzheimer’s disease”, “biological marker”, “biomarker”, “plasma”, “serum”, “blood”, “blood-based”, “proteomics”,
Literature search results and characteristics of included studies
A total of 1651 articles were identified through PubMed/ Medline, EMBASE, Web of Science, Science Direct and Springer. Further to removal of duplicates, 1419 articles remained. In addition, six articles were identified from relevant review articles. Out of 1425 articles that were screened based on titles and abstracts, only 75 full text articles were deemed relevant and further assessed for eligibility based on the pre-specified inclusion criteria. Finally, a total of 22 studies were included
Discussion
The current study had presented a systematic review and meta-analysis of untargeted discovery of highly replicable, blood-based CB that are able to differentiate between AD and control subjects across various proteomic platforms and blood fractions. Overall, this study had identified six CB that are replicated with consistent regulation in at least three independent cohorts. They are α2Mps, PPps, ApoA-1ps, afaminp, IGFBP-2ps and fibrinogen-γ-chainp. It was found that the majority of
Conclusion
Altogether, the majority of replicated CB were found to be involved in systemic inflammation. Besides, the present systematic review had also identified six CB that were consistently regulated in at least three independent cohorts at the discovery stage level via unbiased study design that involved the use of various proteomics platforms and blood fractions. These CB are known to be involved in AD pathogenesis via Aβ formation, neurofibrillary tangles, diabetes and CVD. Amongst the highly
Declaration of competing interest
None.
CRediT authorship contribution statement
Siti Hajar Rehiman: Conceptualization, Writing - original draft, Formal analysis. Siong Meng Lim: Conceptualization, Writing - review & editing, Formal analysis, Supervision. Chin Fen Neoh: Conceptualization, Formal analysis, Writing - review & editing. Abu Bakar Abdul Majeed: Conceptualization, Writing - review & editing, Supervision, Funding acquisition. Ai-Vyrn Chin: Conceptualization, Writing - review & editing. Maw Pin Tan: Conceptualization, Writing - review & editing. Shahrul Bahyah
Acknowledgments
We acknowledge receipt of financial grant support from the Ministry of Higher Education (MOHE) Malaysia (reference no. 600-RMI/LRGS 5/3 [3/2012]).
References (123)
- et al.
Complement activation as a biomarker for Alzheimer’s disease
Immunobiology
(2012) - et al.
Pancreatic polypeptide regulates glucagon release through PPYR1 receptors expressed in mouse and human alpha-cells
Biochim. Biophys. Acta
(2015) - et al.
Interleukin‐6 and α‐2‐macroglobulin indicate an acute‐phase state in Alzheimer’s disease cortices
FEBS Lett.
(1991) - et al.
Temporal evolution of biomarkers and cognitive markers in the asymptomatic, MCI, and dementia stage of Alzheimer’s disease
Alzheimers Dement.
(2015) - et al.
CSF markers for incipient Alzheimer’s disease
Lancet Neurol.
(2003) - et al.
Neuropeptide Y: an anti-aging player?
Trends Neurosci.
(2015) - et al.
Neuroinflammation and disruption in working memory in aged mice after acute stimulation of the peripheral innate immune system
Brain Behav. Immun.
(2008) Unbiased approaches to biomarker discovery in neurodegenerative diseases
Neuron
(2014)- et al.
Decreased expression and increased oxidation of plasma haptoglobin in Alzheimer disease: insights from redox proteomics
Free Radic. Biol. Med.
(2012) - et al.
Fibrinogen and β-amyloid association alters thrombosis and fibrinolysis: a possible contributing factor to Alzheimer’s disease
Neuron
(2010)