Keywords
Alzheimer's disease - bioinformatics - Chinese herbs - Chinese herbal compound - biomarkers
- therapeutic targets
Introduction
Alzheimer's disease (AD) is a degenerative disease of the central nervous system,
primarily characterized by memory, cognitive, and language impairments, as well as
behavioral changes, accounting for about 60% of all cases of dementia. Its pathological
features mainly include the deposition of β-amyloid protein, neurofibrillary tangles,
synaptic damage, and neuronal loss.[1] According to statistics, there are currently over 50 million people with dementia
worldwide, and this number is expected to reach 152 million by 2050, with medical
care and other related expenditures surpassing 1.1 trillion US dollars.[2] The pathogenesis of AD is not entirely clear, involving hypotheses related to oxidative
stress, mitochondrial dysfunction, neuroinflammation, vascular changes, and abnormalities
in metabolic pathways.[1] Moreover, no drugs with significant efficacy have been developed yet, making AD
a focus and challenge in modern medical research.
Bioinformatics primarily focuses on biological data related to nucleic acids, proteins,
analyzing and mining these data for applications in studying the origins of life,
biological evolution, disease occurrence, and development patterns.[3] In recent years, the application of bioinformatics in researching the mechanisms,
potential therapeutic targets, and drug design of AD has become increasingly widespread
and important.[4] This article provides an overview of the applications of bioinformatics in the diagnosis,
identification of potential targets, and drug development of AD.
The Application of Bioinformatics in the Diagnosis of Alzheimer's Disease
The Application of Bioinformatics in the Diagnosis of Alzheimer's Disease
Currently, the diagnosis of AD in clinical settings is mainly through the collection
of medical history, psychological assessments, and various imaging tests.[5] However, these indicators largely depend on the experience of physicians and carry
a certain degree of subjectivity. Biomarkers can diagnose AD more rapidly, objectively,
and accurately and are also crucial for monitoring the progression of AD and evaluating
the effectiveness of treatments. Commonly used biomarkers for AD mainly include amyloid-β
(Aβ), total microtubule-associated protein tau (t-tau), and phospho-tau (p-tau) proteins
in cerebrospinal fluid (CSF).[6] The pathogenesis of AD is complex, and identifying new and effective biomarkers
can not only improve the accuracy of AD diagnosis and understand its pathological
mechanisms but also aid in drug development. The rapidly developing bioinformatics
technologies provide a convenient and effective means for identifying biomarkers,
and the combined use of various bioinformatics methods can uncover new biomarkers
for AD and reveal the biological processes of the disease.
AD generally cannot be diagnosed through biopsies to obtain tissue samples, and postmortem
tissue samples cannot reflect the early pathology of the disease. Therefore, body
fluids are an ideal source for detecting AD biomarkers. CSF is located in the subarachnoid
space and ventricular system of the brain and spinal cord. It carries the brain's
interstitial fluid through the ventricular membranes, and the changes in proteins
in the CSF can directly reflect the neuropathology of the brain. Therefore, CSF biomarkers
are a fundamental neuropathological indicator, widely used in the diagnosis of neurodegenerative
diseases.[7] In recent years, many proteins in the CSF have been identified as potential biomarkers
for AD, such as neuroflament light (NFL)-related to neuronal damage, neurogranin and
synaptosome associated protein 25 (SNAP-25) reflecting synaptic dysfunction unsoluble
triggering receptor expressed on myeloid cells 2 (sTrem2) and chitinase-3-like protein
(YKL-40) for neuroinflammation, etc.[8] In a proteomics study on AD CSF biomarkers, Bader et al[9] developed a highly reproducible mass spectrometry proteomics workflow called the
“rectangular strategy,” which allows for deep analysis with minimal volumes of CSF.
This study not only identified known AD biomarkers, such as tau protein, but also
screened 40 new potential biomarkers, including Parkinson protein 7 (PARK7) and superoxide
dismutase 1 (SOD1) related to neurodegeneration, tyrosine 3-monooxygenase (YWHAZ)
related to AD genetics, chitinase 3 like 1 (CHI3L1) reflecting astrocyte activation,
and many proteins related to glucose metabolism. Li et al[10] identified 29 significantly differentially expressed proteins by analyzing the differential
protein expression in the CSF of AD patients compared to healthy individuals. Through
functional enrichment analysis, they found that these proteins were mostly concentrated
in metabolic-related pathways. To further identify central proteins in the CSF of
AD patients, the team used the Least Absolute Shrinkage and Selection Operator regression
and random forest feature selection algorithms for further data processing, screening
out six central proteins: YWHAZ, SPARC-related modular calcium binding 1 (SMOC1),
CH3L1, aldolase A (ALDOA), secreted phosphoprotein 1 (SPP1), and pyruvate kinase M
(PKM). Among them, SMOC1, ALDOA, and PKM were specifically upregulated in AD patients
and highly correlated with Aβ and tau protein pathology, making them potential biomarkers
for diagnosing AD.
At the same time, blood also contains a large number of disease-related proteins and
can be obtained noninvasively, with simple sampling, high efficiency and low cost.
Yao et al[11] combined computational prediction with experimental validation for the first time
to identify blood protein biomarkers for AD. They first collected tissue transcriptome
data from AD patients and healthy control groups from the Gene Expression Omnibus
database, identifying 2,754 differentially expressed genes. Then, they used a blood
secretion protein prediction program to predict these genes, finding 296 genes encoding
AD-related blood secretory proteins. Based on the expression levels of these proteins'
corresponding genes, their functions, and their relevance to AD, they selected 10
proteins as potential biomarkers for AD. Finally, they collected blood samples from
AD patients and healthy controls and conducted experimental validation through the
enzyme-linked immunosorbent assay, where gelsolin, brain-derived neurotrophic factor,
tissue inhibitor of metalloproteinases 1 (TIMP1), very low-density lipoprotein receptor
(VLDLR), and amyloid beta precursor like protein 2 were consistent with the prediction
results. Receiver operating characteristic curve analysis found that TIMP1 and VLDLR
had the strongest ability to differentiate AD patients from healthy controls, which
was also confirmed by subsequent western blot experiments. TIMP1 is a cytokine with
neuroprotective effects, capable of improving cognitive dysfunction in AD by clearing
Aβ protein deposits and maintaining synaptic integrity.[12] VLDLR is an aolipoprotein E (ApoE) receptor related to the risk factors for AD,
distributed in synaptic regions. It participates in neuronal synaptic plasticity and
affects learning and memory abilities by regulating the renin-angiotensin system (Ras)
signaling pathway associated with neurodegenerative changes such as AD, altering the
formation of presynaptic and postsynaptic dendritic spines.[13] In the blood of AD patients, TIMP1 significantly increases and VLDLR significantly
decreases,[11] suggesting that TIMP1 and VLDLR could potentially become new blood biomarkers for
AD. The abovementioned research provides an effective method for finding AD-related
biomarkers in blood. However, a limitation is that gene expression changes do not
accurately reflect protein expression changes. Therefore, the proteins predicted by
this method need further validation in large-scale blood samples.
In addition, body fluids such as saliva and urine can also be used as noninvasive
samples for the diagnosis of AD. For example, Guo et al[14] used sequencing technology to detect saliva and gingival crevicular fluid of subjects.
β diversity analysis showed a significant difference in the periodontal microbiota
between AD patients and control group, and the main species of the microbiota changed
with the severity of AD.
Although many body fluids are easily obtainable, different sample collection procedures,
such as serum and plasma, quantitative platforms, etc., can have a certain impact
on research results. Ideal biomarkers should meet conditions such as disease specificity,
result reproducibility, and translatability from the laboratory to clinical settings.
Therefore, large-scale prospective multicenter studies are needed for these identified
potential AD biomarkers. Compared to traditional research methods, bioinformatics
technology has the advantages of high throughput and big data and can achieve automated
analysis and processing of data through computer programs and algorithms, effectively
avoiding human intervention. It has now become an important approach for further exploration
of the pathogenesis of AD and diagnosis based on biomarkers.
The Application of Bioinformatics in Exploring Therapeutic Targets for Alzheimer's
Disease
The Application of Bioinformatics in Exploring Therapeutic Targets for Alzheimer's
Disease
Due to the complex pathological process of AD, identifying effective therapeutic targets
is particularly challenging. As mentioned above, there are multiple pathogenic hypotheses
in the occurrence and progression of AD, each of which provides a possible strategy
for treating AD.
Bioinformatics has become an important tool for identifying therapeutic targets for
AD. Some transcriptomic studies on AD have reported that stress and immune responses
are closely related to the pathogenesis of AD,[15]
[16]
[17] which could serve as potential therapeutic targets for AD. However, these studies
have analyzed small sample sizes and only assessed differential expression at the
single gene level, which has certain limitations. Park et al[18] first used two independent cohorts, Alzheimer's disease neuroimaging initiative
(ADNI) and AddNeuroMed, to conduct weighted gene coexpression network analysis based
on blood samples from AD patients. They identified AD-related modules, determined
the biological pathways enriched in AD-related modules through enrichment analysis,
and conducted correlation analysis with known AD biomarkers such as Aβ42 and p-tau. Studies have found significant dysregulation of the Fc gamma (Fcγ) receptor-mediated
phagocytosis pathway, osteoclast differentiation pathway, and tuberculosis pathway
in ADNI and AddNeuroMed. Key genes ankyrin repeat and PH domain 1 (ASAP1) and protein
kinase C delta (PRKCD) in the Fcγreceptor-mediated phagocytosis pathway show abnormal
expression, and PRKCD is strongly correlated with cognitive function, Aβ42, and p-tau. This suggests that blocking the Fcγ receptor and its pathway may alleviate
AD pathology. Phitthayaphong et al[19] provided experimental evidence for this viewpoint through in vitro experiments.
In recent years, emerging single-cell sequencing technology has become an effective
approach to deeply understand the molecular mechanisms and to identify therapeutic
targets in the pathophysiological process of AD. Some single-cell sequencing studies
on AD brain tissue have identified a series of brain cell clusters related to AD,[20]
[21] but these studies have mainly focused on clustering and differential analysis, without
fully utilizing the single-cell sequencing data. Lau et al[22] conducted single-nucleus transcriptome analysis on 169,496 nuclei from cortical
samples of AD patients and normal control group, identifying 43 specific cell clusters.
Differential analysis showed that cell type-specific transcriptional changes in AD
were associated with disruptions in biological processes such as angiogenesis, immune
activation, synaptic signaling, and myelination. Subcluster analysis indicated that
AD brains contain fewer neuroprotective astrocytes and oligodendrocytes compared to
normal brains and induced a subpopulation of angiogenic endothelial cells.
These angiogenic endothelial cells exhibit increased expression of angiogenic growth
factor and its receptors EGF-like domain multiple 7 (EGFL7), Fms-related receptor
tyrosine kinase 1, von willebrand factor, and antigen presentation mechanism beta-2-microglobulin
and major histocompatibility complex, indicating that the pathogenesis of AD may be
related to dysregulation of angiogenesis and antigen presentation in endothelial cells.
These studies have revealed previously unknown pathological molecular changes and
cell targets, providing an important theoretical basis for the treatment or improvement
of AD pathological progress.
Bioinformatics technology has provided some new ideas for exploring the treatment
of AD, but how to reasonably and accurately analyze a large amount of research data,
explore key signaling targets, and promote clinical translation remains a research
hotspot in the fields of bioinformatics and neuroscience.
The Application of Bioinformatics in Drug Research for Alzheimer's Disease Treatment
The Application of Bioinformatics in Drug Research for Alzheimer's Disease Treatment
Current Status of Drugs for Alzheimer's Disease Treatment
The treatment methods for AD so far mainly fall into two directions: (1) Preventing
or delaying the onset and progression of AD to reduce or even repair neuronal damage;
(2) Symptomatic treatment aimed at improving cognitive impairment and controlling
psychiatric symptoms. Currently, the AD treatment drugs approved by the Food and Drug
Administration (FDA) mainly focus on symptomatic treatment, including acetylcholinesterase
inhibitors (donepezil, galantamine, rivastigmine, and huperzine A), and N-methyl-D-aspartic
acid receptor antagonists (memantine hydrochloride). These drugs can only alleviate
patient symptoms, have a single therapeutic effect and limited duration, unable to
produce significant disease improvement effects.[23]
[24] The Aβ-targeting monoclonal antibody Aducanumab, which was newly approved for marketing
in 2021, also failed to prevent or delay the progression of AD, and its efficacy has
been controversial.[25] Therefore, there is an urgent need for new treatment methods that can prevent or
delay clinical symptoms of AD, slow down or even terminate the pathological progression
of AD.
The Application of Bioinformatics in the Research on Chinese Herbs in Treating Alzheimer's
Disease
Compared to FDA-approved Western medicine, Chinese herbs have the characteristics
of multiple components and targets, as well as unique advantages such as overall regulation,
reliable efficacy, and minimal adverse reactions. Bioinformatics plays an important
role in the research and development of new Chinese herbs in treating AD. By utilizing
bioinformatics technology, rich information related to drugs can be effectively managed,
such as the molecular targets of drugs, the association between targets and diseases,
and the interactions of targets in cellular network environments. Based on the analysis
of these data, the biological activity and pharmacokinetic characteristics of Chinese
herbs can be predicted, thereby identifying potential drugs for AD. So far, many research
works have proven that Chinese herbs have significant anti-AD effects.
The Application of Bioinformatics in the Research on Single Chinese Herb in Treating
Alzheimer's Disease
Many databases provide the basis for bioinformatics research on Chinese herbs in treating
AD. Sun et al[26] conducted large-scale literature data mining on PubMed and clinical trial databases
(www.Clinicaltrials.gov), proving that various Chinese herbs and their active ingredients, such as Yinxingye
(Ginkgo Folium), Shishan (Huperzia), and Danshen (Salviae Miltiorrhizae Radix et Rhizoma),
can exert anti-AD effects by inhibiting AD-related pathways, such as down-regulating
intracellular Ca2+ homeostasis and inhibiting inflammatory cytokines. Ginsenoside Rd is an active ingredient
with broad pharmacological effects in Renshen (Ginseng Radix et Rhizoma). Chen et
al[27] conducted literature searches on major scientific databases such as China national
knowledge infrastructure, Elsevier, ScienceDirect, and PubMed and found that ginsenoside
Rd is a neuroprotective agent with anti-inflammatory, antioxidative, antiapoptotic,
and mitochondrial protective effects. It can inhibit neurotoxicity and regulate nerve
growth factors to promote nerve regeneration through these pathways, making it a multifunctional
lead compound for the treatment of AD and even neurodegenerative diseases.
In recent years, network pharmacology has become one of the hot topics in the field
of traditional Chinese medicine (TCM) research. Wang et al[28] used a combination of network pharmacology and comparative metabolomics to study
the main active ingredients of the Chinese herb Huangjing (Polygonati Rhizoma) and
its anti-AD target effects. They further used molecular docking to identify the binding
ability of these active ingredients with the AD-related target AChE. Finally, they
verified the actual therapeutic effects of several key active ingredients in Huangjing
(Polygonati Rhizoma) on AD through in vivo and in vitro experiments. The results indicated
that the active ingredients in Huangjing (Polygonati Rhizoma), such as cafestol, rutin,
and isorhamnetin, can significantly inhibit AChE activity, increase neuronal cell
vitality, exhibit anti-inflammatory properties, and reduce oxidative stress damage,
thereby demonstrating a multilayered anti-AD effects.
In addition, various molecular simulation methods have been widely used in AD drug
development. Shinzato et al[29] conducted studies on the binding characteristics of some curcumin derivatives with
Aβ using protein-ligand docking simulations and fragment molecular orbital methods.
They found that when the COH3 group in the aromatic ring of curcumin-Ib was substituted with OH, it can strongly
bind to Aβ and effectively inhibit Aβ aggregation, thus playing a therapeutic role
in AD treatment.
The Application of Bioinformatics in the Study of Chinese Herbal Compound Treatment
for Alzheimer's Disease
Chinese herbal compound refers to a prescription composed of two or more Chinese medicinal
ingredients. It has complex chemical compositions and exerts pharmacological effects
on multiple components and targets. The synergistic effects among various drugs often
result in better therapeutic outcomes compared to single herbal medicine. Traditional
pharmacology, molecular biology, and other experimental methods have limitations in
fully elucidating the characteristics of drug effects. Therefore, the application
of bioinformatics techniques in the study of Chinese herbal compound is more advantageous
for the development of therapeutic drugs.
Research has shown that Liuwei Dihuang Pill, a classic Chinese herbal formula, has
the effects of regulating immune cell infiltration to alleviate neuroinflammation,
reducing Aβ aggregation, and delaying cognitive impairment.[30] However, the specific molecular mechanisms of its action have not been fully understood.
Zhao et al[31] used bioinformatics methods to explore the molecular immune mechanism of Liuwei
Dihuang Pill in the treatment of AD by studying the infiltration patterns of different
types of immune cells in AD. The study found that in hippocampal tissue samples from
AD patients, the infiltration levels of M2 macrophages and quiescent CD4+ memory T
cells were higher compared to healthy individuals, and there were significant differences
in the infiltration levels of M1 macrophages as well. Protein–protein interaction
(PPI) network analysis revealed that Liuwei Dihuang pill could regulate two core immune
targets in AD, nuclear factor kappa B inhibitor alpha, and protein kinase C beta (PRKCB),
which were involved in multiple biological signaling pathways. Among them, PRKCB showed
the best molecular docking effect with quercetin, a key active component of the herbal
formula. Therefore, through bioinformatics, the study discovered and confirmed that
Liuwei Dihuang Pill can regulate immune cell infiltration through multiple components,
targets, and pathways, providing a new research direction for the immune therapy of
AD.
Li et al[32] developed a new Chinese medicine formula called Nao Tan Qing based on the theory
of resolving phlegm and opening orifices in TCM. The formula is composed of Dannanxing
(Arisaema cum Bile), Huangqin (Scutellariae Radix), Huanglian (Coptidis Rhizoma),
Banxia (Pinelliae Rhizoma), Tianma (Gastrodiae Rhizoma), Ganjiang (Zingiberis Rhizoma),
Shichangpu (Acori Tatarinowii Rhizoma), and Gancao (Glycyrrhizae Radix et Rhizoma).
This study used network pharmacology and omics analysis to discover that Nao Tan Qing
can inhibit neuroinflammation in AD mice by regulating the NF-κB and Toll-like receptor
pathways while also regulating their glucose and lipid metabolism. The study demonstrated
that Nao Tan Qing may improve AD by modulating signal pathways associated with neuroinflammation
and metabolism and is a potential drug for the treatment of AD.
Shen et al[33] applied network pharmacology to explore the action targets, pathways, and mechanisms
of Chuanxiong Renshen Decoction (CRD) in the treatment of AD. The study first identified
the main effective components of CRD using mass spectrometry analysis, retrieved potential
targets of CRD components from the SwissTargetPrediction database, and screened AD-related
targets from the Disgenet and Genacards databases. By taking the intersection of these
two sets of targets, they obtained 65 potential key target genes for CRD in the treatment
of AD. PPI network analysis identified the core targets for the treatment of AD as
caspase 3, epidermal growth factor receptor, amyloid beta precursor protein, cannabinoid
receptor 1, prostaglandin-endoperoxide synthase 2, and glutamate metabotropic receptor
5. Molecular docking results showed good binding of each component with potential
core targets. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment
analysis indicated a concentration on the tumor necrosis factor (TNF) signaling pathway,
and mitogen-activated protein kinase 1 (MAPK) signaling pathway.
The Application of Bioinformatics in the Research on Combined Chinese Medicine and
Western Medicine Treatment for Alzheimer's Disease
Research has confirmed that the combined use of Chinese medicine and Western medicine
showed a stronger anti-AD effect compared to using them individually.[34] Wang et al[35] conducted a network meta-analysis on the effect of Kidney-tonifying Decoction combined
with donepezil in the treatment of AD. A total of 56 related studies were retrieved,
and 30 randomized clinical trials were eventually included. The analysis revealed
that the total effective rate of the treatment of AD with Kidney-tonifying Decoction
and donepezil in combination was the highest at 88%, followed by Kidney-tonifying
Decoction alone at 58.5% and donepezil alone at 3.5%. Compared to single-drug therapy,
combination therapy can significantly improve clinical symptoms of AD such as memory
decline and motor dysfunction, resulting in better clinical efficacy. However, the
specific action mechanisms still require further investigation.
The pure bioinformatics prediction research on Chinese herbs and compound formulas
for treating AD is still somewhat limited (see [Table 1]). In future research, a combination of bioinformatics and experimental verification
can be used to more deeply explore the mechanism of action and pharmacological substance
basis of Chinese herbs and compound formulas, and better serve clinical practice.
Table 1
The application of bioinformatics in the treatment of AD with Chinese herbs
Chinese herbs
|
Research method
|
Action mechanism
|
Yinxingye (Ginkgo Folium), Shishan (Huperzia), and Danshen (Salviae Miltiorrhizae
Radix et Rhizoma)
|
Data mining
|
Downregulation of intracellular Ca2+ homeostasis, inhibition of inflammatory cytokines, and inhibition of AD-related pathways[26]
|
Ginsenoside Rd
|
Data mining
|
Inhibition of neurotoxicity, regulation of nerve growth factor to promote nerve regeneration[27]
|
Huangjing (polygonati rhizoma)
|
Network pharmacology combined with metabolomics
|
Inhibition of AChE activity, enhancement of neuronal cell vitality, anti-inflammatory
effects, and reduction of oxidative stress damage[28]
|
Curcumin-Ib
|
Protein-ligand docking simulation and fragment molecular orbital
|
Effective inhibition of Aβ aggregation[29]
|
Liuwei Dihuang Pill
|
Data mining combined with molecular docking
|
Multifaceted regulation of immune cell infiltration[31]
|
Nao Tan Qing
|
Network pharmacology combined with metabolomics
|
Inhibition of neuroinflammation through the NF-κB and Toll-like receptor pathways,
while also regulating glucose and lipid metabolism[32]
|
Chuanxiong Renshen Decoction
|
Network pharmacology combined with molecular docking
|
Regulation of the TNF signal pathway and MAPK signal pathway[33]
|
Kidney-tonifying Decoction combined with donepezil
|
Meta-analysis
|
Improvement of clinical symptoms of AD such as memory decline and motor dysfunction[35]
|
Abbreviations: Aβ, amyloid-β; AChE, acetylcholinesterase; AD, Alzheimer's disease;
MAPK, mitogen-activated protein kinase 1; TNF, tumor necrosis factor.
Conclusion
The pathological process of AD is complex, and the pathogenesis is not yet clear.
Currently, there is a lack of simple and effective diagnostic methods in clinical
practice, and it also faces many challenges of anti-AD drugs. The emergence of bioinformatics
provides new ideas for screening key pathological biomarkers of AD, analyzing differentially
expressed genes in AD, exploring new targets, and discovering new anti-AD drugs. It
is worth noting that in bioinformatics research, data preprocessing (such as detection
of abnormal values, data normalization, etc.) and statistical analysis methods will
have a significant impact on the results. In actual research, the lack of consistent
coordination in analysis methods across platforms and laboratories makes it difficult
to compare data.[36] Therefore, it is crucial to establish standard operating procedures and quality
control protocols, improve the repeatability of methods across platforms, and promote
data comparison between studies. In addition, there are certain limitations in bioinformatics
prediction research. Experimental validation at the biological level and in-depth
interdisciplinary research should be conducted based on this foundation to determine
the role of Chinese herbs and Chinese herbal compound in the complex mechanism of
AD. Furthermore, these predictions must undergo clinical validation before they can
further provide new theoretical basis for the comprehensive diagnosis and treatment
of AD with Chinese medicine.