Evidence-Based Complementary and Alternative Medicine

Evidence-Based Complementary and Alternative Medicine / 2016 / Article

Research Article | Open Access

Volume 2016 |Article ID 8239817 | https://doi.org/10.1155/2016/8239817

Hung-Tsu Cheng, Chaang-Ray Chen, Chia-Yang Li, Chao-Ying Huang, Wun-Yi Shu, Ian C. Hsu, "The Classification of Sini Decoction Pattern in Traditional Chinese Medicine by Gene Expression Profiling", Evidence-Based Complementary and Alternative Medicine, vol. 2016, Article ID 8239817, 13 pages, 2016. https://doi.org/10.1155/2016/8239817

The Classification of Sini Decoction Pattern in Traditional Chinese Medicine by Gene Expression Profiling

Academic Editor: Kamal D. Moudgil
Received18 Nov 2015
Revised21 Feb 2016
Accepted10 Mar 2016
Published20 Apr 2016

Abstract

We investigated the syndromes of the Sini decoction pattern (SDP), a common ZHENG in traditional Chinese medicine (TCM). The syndromes of SDP were correlated with various severe Yang deficiency related symptoms. To obtain a common profile for SDP, we distributed questionnaires to 300 senior clinical TCM practitioners. According to the survey, we concluded 2 sets of symptoms for SDP: (1) pulse feels deep or faint and (2) reversal cold of the extremities. Twenty-four individuals from Taipei City Hospital, Linsen Chinese Medicine Branch, Taiwan, were recruited. We extracted the total mRNA of peripheral blood mononuclear cells from the 24 individuals for microarray experiments. Twelve individuals (including 6 SDP patients and 6 non-SDP individuals) were used as the training set to identify biomarkers for distinguishing the SDP and non-SDP groups. The remaining 12 individuals were used as the test set. The test results indicated that the gene expression profiles of the identified biomarkers could effectively distinguish the 2 groups by adopting a hierarchical clustering algorithm. Our results suggest the feasibility of using the identified biomarkers in facilitating the diagnosis of TCM ZHENGs. Furthermore, the gene expression profiles of biomarker genes could provide a molecular explanation corresponding to the ZHENG of TCM.

1. Introduction

Traditional Chinese medicine (TCM) has been adopted to treat the Chinese for thousands of years on the basis of some classic Chinese medicine textbooks, including Huang Ti Nei Jing (“The Inner Cannon of Huangti”) [1], Shénnóng Běn Cǎo Jīng [2], and Shang Han Za Bing Lun (“Treatise on Cold Damage Disorders”) [3]. Shang Han Za Bing Lun is a Chinese medical treatise written by Zhang Zhongjing before AD 220, at the end of the Han dynasty. The treatise is the oldest comprehensive clinical textbook in the world containing principles, methods, formulas, and medicine and is the clinical literature of TCM theory with practice. To obtain the appropriate decoction for a particular ZHENG requires a diagnostic system, namely, the Decoction ZHENG proposed in Shang Han Za Bing Lun. Through the methods of inspection, listening, smelling, inquiry, and palpation, physicians collect information on patients to identify their TCM ZHENGs according to their personal experience and then treat these patients with herbs, which is called decoction. The methods applied in TCM have been criticized as insufficiently scientific. To overcome this weakness, we conducted systems biology research to differentiate TCM ZHENGs.

Systems biology-based diagnostic principles can be used to support the relationship between TCM and current biomedicine [4, 5]. From the perspective of gene expression profile, TCM ZHENGs differentiation is closely related to gene polymorphisms and differences in gene expression profile. Therefore, applying advanced sequencing techniques and cDNA microarray studies can help to clarify the biological basis of TCM ZHENGs [4]. For example, by using microarray, real-time polymerase chain reaction, and enzyme-linked immunosorbent assay technologies, Shen Yang Xu ZHENG has been determined to be involved with the gene level [6] and may be primarily attributed to the insufficient activity of the mitogen-activated protein kinase (MAPK) pathway [7]. The cold ZHENG was discovered to be possibly caused by a physiological imbalance and disorder of metabolite processes by using microarray technology [8]. Human mRNAs in peripheral blood mononuclear cells (PBMCs) were extracted for microarray experiments to analyze differently expressed gene expressions for TCM cold and hot ZHENG differentiation [911].

In this study, we investigated the syndromes of Sini decoction pattern (SDP). Sini decoction was first mentioned in Shang Han Za Bing Lun Sini decoction consisting of three herbs: fresh Aconitum carmichaelii, dry Zingiber officinale, and Glycyrrhiza uralensis, which was characterized as a remedy to have essential effect of recuperating the patients from collapse. It was used to treat the syndrome of displaying coldness on the extremities, pulse feels deep or faint, continuous diarrhea with undigested cereal, profuse perspiration, abdominal fullness and distention, vomiting, and lethargy [12]. Based on the clinical applications of Sini decoction, SDP was then developed by Sun [13]. The syndromes of SDP were correlated with numerous symptoms, such as a slow pulse, coldness on the extremities, continuous diarrhea with undigested cereal, and severe diarrhea, but no common profile exists. According to our survey of 300 senior clinical TCM practitioners, we obtained a common profile for SDP: (1) pulse feels deep or faint and (2) reversal cold of the extremities. We collected SDP patients with these 2 symptoms. In this study, 12 patients with SDP and 12 non-SDP individuals were recruited from Taipei City Hospital, Linsen Chinese Medicine Branch, Taiwan.

We extracted the total mRNA of PBMCs from the 24 individuals for microarray experiments. Twelve individuals (including 6 SDP patients and 6 non-SDP individuals) were used as the training set to identify the biomarkers that were screened using the hypothesis test for distinguishing the SDP and non-SDP groups. The remaining 12 individuals (including 6 SDP patients and 6 non-SDP individuals) were used as the test set to examine whether the identified biomarkers can distinguish the SDP and non-SDP groups. The test results indicated that the gene expression profiling by the biomarkers could effectively distinguish the 2 groups by using a hierarchical clustering algorithm.

Our results suggested the feasibility of using biomarkers extracted from the gene expressions of PBMCs in facilitating the diagnosis of TCM ZHENGs. Furthermore, the gene expression profiles of biomarkers could provide the possible molecular mechanisms of related TCM ZHENGs.

2. Materials and Methods

2.1. Chinese Medicine Terminology

All Chinese medicine terms used in this paper were in accordance with the International Acupuncture Nomenclature (IAN) proposed by WHO in 1991; if not available in IAN, the standard nomenclature proposed in the WHO International Standard Terminologies on Traditional Medicine in the Western Pacific Region [14] was adapted. These translated Chinese medicine terms are presented in italics in this paper.

2.2. Human Participants

The syndromes of SDP were associated with a slow, deep, and faint pulse that disappeared, coldness on the extremities, severe diarrhea, continuous diarrhea with undigested cereal, generalized pain, abdominal fullness, cold-fluid retention on the diaphragm, profuse sweating, severe spasms in the feet, tight clamping of the extremities, and reversal cold in both the hands and feet; however, no common profile exists. To obtain a common profile for SDP, we distributed questionnaires to 300 senior clinical TCM practitioners. Our questionnaire listed 41 symptoms related to SDP. In the survey, each senior clinical TCM practitioner assigned a score to each symptom, indicating the correlation between the symptom and SDP from high to low (5, 4, 3, 2, 1, and 0). We identified symptoms that collectively constituted a common SDP profile according to the frequency with which participating TCM practitioners assigned a score of 5 to the symptoms; the frequency threshold was set at higher than 35% of the returned surveys. We recruited SDP patients with the aforementioned symptoms. In this study, 12 SDP patients and 12 non-SDP individuals from Taipei City Hospital, Linsen Chinese Medicine Branch, Taiwan, were recruited. All of the individuals provided informed consent. The definition of healthy subject was diagnosed by western medicine rather different from TCM. It is difficult to diagnose healthy subject. Therefore, we consider substituting non-SDP patients for healthy subject. We try our best to collect non-SDP patients who are alike healthy subjects.

2.3. Blood Sample Collection

We phlebotomized 30 mL of human peripheral blood from each donor and then separated the blood into three 10 mL plastic whole blood tubes with spray-coated K2EDTA (Becton Dickenson, Franklin Lakes, NJ, USA). The blood was transferred to a 50 mL centrifuge tube. The PBMCs were then prepared from the 30 mL of peripheral blood through Ficoll-Hypaque density gradient centrifugation. The PBMCs were added to an RNAlater RNA Stabilization Reagent (Qiagen, Valencia, CA, USA) and then placed into a bottle with liquid nitrogen. All of these procedures were performed immediately after the blood was collected at the laboratory of Taipei City Hospital, Linsen Chinese Medicine Branch, Taiwan. The isolated PBMCs were frozen, stored in a liquid nitrogen freezer, and transported to the microarray experimental laboratory at National Tsing Hua University, Hsinchu, Taiwan, for further microarray experiments.

2.4. RNA Extraction and Microarray Hybridization

We extracted total RNA by using the RNeasy Mini Kit according to the manufacturer’s protocol (Qiagen, Valencia, CA, USA). Total RNA quantity and quality were assessed using an ultraviolet absorption spectrophotometer at 260 and 280 nm. Subsequently, α- and β-globin mRNA were reduced from a portion of the total RNA samples by using the GLOBINclear human kit (Ambion, Austin, TX, USA), according to the manufacturer’s instructions, at the recommended start quantity of 10 μg of total RNA. The quality of total RNA was evaluated using the Agilent 2100 bioanalyzer with the RNA 6000 Nano LabChip kit (Agilent Technologies, Palo Alto, CA, USA). Before reverse transcription, each RNA sample was spiked with a mixture of Arabidopsis mRNAs, which we called doping control. Fluorescence-labeled (Cy3, Cy5-3DNA Capture Reagent) cDNA was conducted using the 3DNA Array 900 labeling kit according to the manufacturer’s protocols (Genisphere, Hatfield, PA, USA). Reverse transcription was performed using SuperScript II (Invitrogen, Carlsbad, CA, USA). Hybridization was performed at 65°C in a water bath for 16 to 18 h, and arrays were washed according to the manufacturer’s protocol (Corning Life Sciences, New York, NY, USA). The arrays were scanned using GenePix 4000B scanner (Axon Instruments, Foster City, CA, USA).

2.5. Microarray Fabrication

A total of 9600 human cDNA clones were purchased from Incyte Genomics (Wilmington, DE, USA). These cDNAs were randomly selected from the IMAGE library. Resequencing was performed on these clones, and 7334 cDNAs were verified to have corrected DNA sequences. These 7334 sequence-verified human cDNAs (Incyte Genomics) and 10 Arabidopsis cDNAs (SpotReport cDNA Array Validation System; Stratagene, La Jolla, CA, USA), serving as spike-in controls, and one housekeeping gene (β-actin), serving as a positive control, were arrayed on Corning UltraGAPS slides (Corning, Corning, NY, USA). These 7334 human cDNAs and 96 Arabidopsis cDNAs and housekeeping genes were spotted in quadruplicate on each array making a total of 32248 spots to enhance the statistical confidence in the gene expression data. Each array had 32,448 spots. The arrays were postprocessed according to the Corning UltraGAPS Coated Slides instruction manual.

2.6. Experimental Design

Loop design, a type of statistical microarray experimental design, was adopted to construct an optimal design. The basic principles of optimal design are a balance among the factors, approximately equal sampling of varieties, and minimal distance between pairs of varieties [15]. Samples from 24 individuals were hybridized on 48 arrays based on the 2 loop designs presented in Figure 1. The individuals of samples L1A1, L1A2, L1A3, L1C1, L1C2, L1C3, L2A1, L2A2, L2A3, L2C1, L2C2, and L2C3 were SDP patients, and those of samples L1B1, L1B2, L1B3, L1D1, L1D2, L1D3, L2B1, L2B2, L2B3, L2D1, L2D2, and L2D3 were non-SDP individuals. Each arrow in the figure indicates a microarray hybridization experiment. The arrow heads and tails represent samples labeled with Cy5 (3DNA Capture Reagent) and Cy3 (3DNA Capture Reagent), respectively. The black solid circles stand for the SDP samples and the open circles stand for non-SDP samples. We used 12 individuals in Loop 1 experiments as the training set and the remaining 12 individuals in Loop 2 experiments as the test set.

2.7. Microarray Data Analysis and Statistical Analysis

Microarray data preprocessing, normalization, and statistical analysis were performed using the bioinformatics software suite Tsing Hua Engine for Microarray Experiment [16].

Spot-screening rules were applied to screen invalid spots on these arrays. The spot-screening rules for the data of Loop 1 (the training set) were as follows: (1) exclude the spots defined as “flag bad” or “absent” in all GenePix Results (GPR), (2) exclude the spots with diameters less than 100 μm, (3) exclude the spots with a coefficient of variation (CV) of pixel intensity of over 100% in both channels, and (4) exclude the spots with signal to noise ratios (SNRs) in both channels were less than 2 in the training set microarray experiments. The SNR is defined as (: mean of pixel intensities of the signal; : median pixel intensity of the background). The logarithm of the ratios for all valid spots on each array was normalized using global lowess normalization [17]. After the data were preprocessed, the normalized log ratios of the fluorescence intensity of each cDNA spot were analyzed using a log linear model, which was described in our previous study [18], to obtain the least-squares estimates , , , , , , , , , , , and for the relative expressions among all of the samples. The relative expressions of sample are defined as (expression of sample )/(geometric average of the expression of all samples), for each gene, where runs from 1 to 12. For the training set microarray data, 4951 genes satisfied the selection criteria of the spot-screening rules and log linear model. The selection criteria for the data of Loop 2 (the test set) are described in the subsequent subsection. The microarray data is available at GEO (GSE67090).

2.8. Searching Biomarkers to Distinguish SDP

We used Loop 1 experiments as the training set and Loop 2 experiments as the test set. For the 4951 genes of Loop 1 (the training set), 192 genes were identified using an test with null hypothesis : (, at a Bonferroni-adjusted significance level of 0.05/4951, where , , , , , and were the least-squares estimates for the relative expressions of the SDP samples and , , , , , and were the least-squares estimates for the relative expressions of the non-SDP samples.

The 192 biomarker genes were then used to distinguish the SDP and non-SDP groups of the test set. We applied the selection criteria to filter bad spots to perform an effective normalization and a log linear model. In order to include additional available genes to be tested, we relaxed the selection criteria of the spot-screening rules for the data of Loop 2 (the test set) as follows: (1) exclude the spots defined as “flag bad” or “absent” in all GPR files, (2) exclude the spots with diameters less than 50 μm, (3) exclude the spots with a CV of pixel standard deviation of over 150% in both channels, and (4) exclude the spots with SNRs in either channel were less than 1 in the test microarray experiments. The logarithm of the ratios for all valid spots on each array was normalized under the same normalization conditions as those for the training set. We then used hierarchical clustering to verify whether the gene expression profiles of the 192 biomarker genes could classify the 12 test samples into 2 groups and how well they could distinguish the SDP patients and non-SDP individuals.

2.9. Functional Annotation and Enrichment Analysis

Functional annotation and enrichment analysis were accessed through the database for annotation, visualization, and integrated discovery (DAVID) [19], which performed a modified Fisher’s exact test to identify overrepresented functions. DAVID is a widely employed web tool that provides a rapid means to understand the biological meaning of a long list of genes of interest. For any gene list, DAVID can map and convert identifiers (IDs) and accessions among different gene IDs. A long gene list can be reduced into biological theme-related groups of genes, for example, gene ontology (GO) [20] and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway [21].

The Entrez gene IDs of 192 biomarker genes were used as the gene ID for bioinformatics analysis. Biological processes of GO terms were used for functional annotation and enrichment analysis of these biomarker genes.

3. Results

3.1. Questionnaires to Obtain a Common Profile for SDP

We distributed questionnaires to 300 senior clinical TCM practitioners, and 43 questionnaires were returned. Our questionnaire listed 41 symptoms related to SDP. In the survey, each senior clinical TCM practitioner assigned a score to each symptom, indicating the correlation between the symptom and SDP, from high to low (5, 4, 3, 2, 1, and 0). Based on the frequency with which participating TCM practitioners assigned a score of 5 to the symptoms, we identified symptoms that collectively constituted a common SDP profile (Figure 2). The frequency threshold was set at higher than 35% of the 43 returned surveys (i.e., symptoms assigned a score of 5 by at least 16 surveys). This process revealed 7 symptoms to form our symptom profile of SDP patients. By grouping related symptoms, we obtained a common profile for SDP as 2 sets of symptoms: (1) pulse feels deep or faint and (2) reversal cold of the extremities. We recruited SDP patients with the 2 sets of symptoms. In this study, 12 SDP patients and 12 non-SDP individuals were recruited from Taipei City Hospital, Linsen Chinese Medicine Branch, Taiwan. All of the participants provided informed consent.

3.2. Identifying Biomarker Genes from the Training Set Microarray Experiment

On the basis of the results for the 4951 genes of Loop 1 (the training set), 192 genes were identified (Table 1) using an test with null hypothesis : , at a Bonferroni-adjusted significance level of 0.05/4951. By applying hierarchical clustering with Spearman distance and average linkage, the 12 training samples were appropriately separated into 2 groups, as shown in Figure 3. In Figure 3, the gene expression value used to do the clustering is the relative expressions of the labeled sample.


Gene name

ABHD3ABI2ACSM2BACTBAGFG1
AKAP9ARG1ARID5AATP1B3ATP6AP2
ATP6V0BBLZF1BPGMC11orf79C12orf24
C12orf32C17orf91C6orf106CAPN1CBARA1
CCL18CCL5CD1DCD24CD40
CD44CD83CD8ACD93CDC42
CDC42SE2CLDN1COMTCOX10CP
CREMCST7CXCL9DEFA1DEFA3
DEFA4DENND2DDGCR6LDLG2DUSP6
EEF2EIF3EEIF3IEPAS1EXOSC9
FAM113BFAM175BFCER1AFLNBFMOD
FOLR3FOSFOSBFYNGCA
GIMAP1GLRXGNG11GNG2GNLY
GPR56GSTA2GTF2E1GZMAH3F3B
HERPUD1HIST1H2ACHIST1H2BDHIST1H2BGHIST1H2BJ
HIST2H2AA3HLA-AHLA-DMAHLA-DQB1HLA-DRB1
HLA-GHs.282050Hs.467411Hs.534061Hs.586682
Hs.591807Hs.599359Hs.655903Hs.674438Hs.88605
HSP90B1HSPA1AHTATIP2HTR7P1ID2
IFIT1IL10RAIL15RAIL1BIL32
JARID1CKIF3CKLF10KLF6LCP1
LNP1LOC400027LOC729776LTBLYRM1
MAPK13MBPMCM3APMESTMLH3
MNDAMRPL3MRPL49MTMR6MYL6
NCRNA00084NHEDC2NPYNUCB1NUDT16
NUP93ORM1PACS1PCDH8PDE4B
PEA15PF4PLAURPNLIPRP2POLR1C
POLR2DPOU2F2PPBPPPP1R12BPPP1R7
PPP2R2APPRC1PPT1PRLPSMC6
RAB4ARARRES3RGS18RGS2RHOQ
RNF40RNPEPRPL41RPS11RPS24
RPS27RPS4XRPS4Y1RUNX3S1PR5
SEC22BSELLSERPINE2SFRS1SH2B3
SH3D19SHCBP1SLC2A3P1SOD2SPTY2D1
SRGNSTX8SYTL2TADA3LTARP
THBS1TIMP1TMEM176ATNFAIP8L2TPCN1
TRA@TRAPPC1TRMT11VIMVPS13D
WDR42AWIZXPNPEP1YARSUnknown
UnknownZNF286A

The sequence of the clone was not yet identified in the BLAST.
3.3. Distinguishing Power of the Biomarker Genes to the Test Set Microarray Experiment

The 192 biomarker genes identified in the training set experiments were then used to test the ability to distinguish the SDP patients and non-SDP individuals in the test set. To include additional available genes to be tested, we relaxed the selection criteria of the spot-screening rules for the data of the test set (Loop 2). The logarithm of the ratios for all valid spots on each array was normalized under the same normalization conditions as those for the training set. We then used hierarchical clustering to verify whether the gene expression profiles of the 192 biomarker genes could classify the 12 test samples into 2 groups and how well they could distinguish the SDP patients and non-SDP individuals. Under the same clustering conditions (i.e., hierarchical clustering with Spearman distance and average linkage), Figure 4 showed that the 192 biomarker genes can classify the 12 test samples into 2 groups. One group comprised L2A2, L2A1, L2A3, L2C3, L2C2, and L2D1 samples and the other comprised L2D2, L2B3, L2D3, L2B2, L2C1, and L2B1 samples. Except for L2D1 and L2C1 samples, all of the samples were correctly grouped into SDP and non-SDP groups. In Figure 4, the gene expression value used to do the clustering is the relative expressions of the labeled sample. The microarray data of 4 biomarker genes from 7 samples were confirmed by quantitative PCR (qPCR). The correlation coefficients between microarray data and qPCR for each gene are 0.73, 0.83, 0.93, and 0.76 (Appendix A).

3.4. Functional Annotation and Enrichment Analysis Results for the 192 Biomarker Genes

The functional annotation and enrichment analysis were accessed using DAVID. Seventy-eight GO terms were enriched with a value less than 0.05. All enriched GO terms are listed in Appendix B. 78 GO terms were enriched with a value of less than 0.05. We further applied 2 stringent criteria to filter the high-ranked GO terms: (1) GO terms containing more than 10 genes and (2) a value of less than 0.01. Consequently, 11 GO terms were identified (Table 2).


GO termDescription# gene value

GO:0006952Defense response240.000000
GO:0006955Immune response240.000002
GO:0009617Response to bacterium110.000088
GO:0001817Regulation of cytokine production110.000172
GO:0048534Hemopoietic or lymphoid organ development130.000562
GO:0030097Hemopoiesis120.000999
GO:0002520Immune system development130.001183
GO:0007610Behavior140.004162
GO:0006412Translation120.004210
GO:0001775Cell activation110.008090
GO:0019725Cellular homeostasis130.009574

Table 2 showed that the 11 GO terms are substantially correlated with immune, inflammation, and hemopoiesis functions. In detail, the immune function includes defense response, immune response, response to bacterium, and immune system development. The inflammation function includes cytokine production regulation, behavior, cell activation, and cellular homeostasis. The hematopoiesis function includes hemopoietic or lymphoid organ development and hemopoiesis.

4. Discussion

Of the 4951 genes of the training set, 192 genes were identified using an test with null hypothesis. By applying hierarchical clustering with Spearman distance and average linkage, the 12 training samples were appropriately separated into SDP and non-SDP groups. The 192 biomarker genes were then used to distinguish the SDP patients and non-SDP individuals in the test set. Under the same clustering conditions, the 12 test samples were clustered into 2 groups. One group comprised L2A2, L2A1, L2A3, L2C3, L2C2, and L2D1 samples and the other comprised L2D2, L2B3, L2D3, L2B2, L2C1, and L2B1 samples. Except for L2D1 and L2C1 samples, all of the samples were correctly grouped into the SDP and non-SDP groups. Thus, we reviewed the original information of these 2 samples. L2C1 sample (classified as SDP according to TCM but clustered into the non-SDP group by using biomarkers) had a deep, rough, and rapid pulse and reversal cold of the extremities. Among the symptoms of L2C1 sample, reversal cold of the extremities matched the selected symptom of SDP; however, a deep, rough, and rapid pulse was inconsistent with a deep or faint pulse, a selected symptom for SDP. L2D1 sample (categorized as non-SDP according to TCM but clustered into the SDP group by using biomarkers) had a deep and tight pulse, which was particularly deep at the guan site (the guan is immediately central to the radial styloid at the wrist, where the tip of the physician’s middle finger is placed), and reversal cold of the extremities. Among the symptoms of L2D1 sample, a deep and tight pulse, particularly at the guan site, did not meet the symptom of a deep or faint pulse, and the sample was thus classified as non-SDP according to TCM. However, reversal cold of the extremities matched the selected symptom of SDP. These findings indicate that, in classifying these 2 samples into SDP and non-SDP groups, uncertain areas exist. Nevertheless, determining biomarkers by using the mRNA profiling method can facilitate identifying the SDP group. Therefore, the traditional dialectical method of TCM is quite trustworthy.

The results of the functional annotation and enrichment analysis of 192 biomarker genes revealed that the 11 GO terms were substantially correlated with immune, inflammation, and hemopoiesis functions. Among the symptoms of SDP, a deep or faint pulse implied the insufficiency of (in the field of medicine, refers both to the refined nutritive substance that flows within the human body and to its functional activities). Insufficiency of indicates the body’s resistance against disease is weak. Pathogens can invade the human body easily, and viruses and bacteria could then cause inflammation easily.

5. Conclusion

Our results provided the biomarker genes of SDP, which was first mentioned in Shang Han Za Bing Lun. The results provide molecular evidence for TCM in the ZHENGs differentiation of SDP. Using mRNA profiling experiments on patients’ PBMCs to discover biomarker genes to test an unknown group of individuals can effectively identify a specific ZHENG. We hope that the mRNA profiling method will become a model for TCM physicians in diagnosis and assist TCM practitioners in evaluating the curative effect of the medication.

Appendix

A.

The microarray data of 4 biomarker genes from 7 samples were confirmed by quantitative PCR (qPCR). The correlation plots between microarray data and qPCR data are shown in Table 3. The correlation coefficients for each gene are 0.73 (GNG2), 0.83 (RUNX3), 0.93 (CST7), and 0.76 (MNDA), shown in Figure 5.


Genes SamplesMicroarray dataqPCR data
log2 of microarray relative gene expression data: log2 ratio of two samples’ microarray relative gene expression data: , that is, log2 of qPCR relative gene expression data: Ctlog2 ratio of two samples’ qPCR relative gene expression data: Ct, that is,

GNG2(a) L2A1 
(b) L2B2
0.572 
0.097
0.476−0.35 
0.14
0.49
(a) L2B2 
(b) L2A2
0.097 
0.235
−0.1390.14 
0.80
0.66
(a) L2A2 
(b) L2D1
0.235 
−0.389
0.6250.80 
1.75
0.95
(a) L2D1 
(b) L2C2
−0.389 
−0.222
−0.1671.75 
0.50
−1.25
(a) L2C2 
(b) L2D2
−0.222 
−0.258
0.0350.50 
1.21
0.71
(a) L2D2 
(b) L2D3
−0.258 
0.131
−0.3881.21 
0.67
−0.54
(a) L2D3 
(b) L2A1
0.131 
0.572
−0.4420.67 
−0.35
−1.02

RUNX3(a) L2A1 
(b) L2B2
0.835 
−0.022
0.8581.39 
2.49
1.10
(a) L2B2 
(b) L2A2
−0.022 
−0.042
0.0202.49 
2.81
0.32
(a) L2A2 
(b) L2D1
−0.042 
−0.099
0.0572.81 
2.85
0.04
(a) L2D1 
(b) L2C2
−0.099 
−0.258
0.1592.85 
2.37
−0.48
(a) L2C2 
(b) L2D2
−0.258 
−0.190
−0.0682.37 
2.86
0.49
(a) L2D2 
(b) L2D3
−0.190 
0.196
−0.3852.86 
2.37
−0.49
(a) L2D3 
(b) L2A1
0.196 
0.835
−0.6402.37 
1.39
−0.98

CST7(a) L2A1 
(b) L2B2
1.301 
0.224
1.076−1.65 
−0.90
0.75
(a) L2B2 
(b) L2A2
0.224 
0.493
−0.268−0.90 
−0.72
0.18
(a) L2A2 
(b) L2D1
0.493 
−0.454
0.946−0.72 
0.04
0.76
(a) L2D1 
(b) L2C2
−0.454 
−0.414
−0.0390.04 
−0.01
−0.05
(a) L2C2 
(b) L2D2
−0.414 
−0.325
−0.089−0.01 
0.23
0.24
(a) L2D2 
(b) L2D3
−0.325 
0.300
−0.6260.23 
−0.62
−0.85
(a) L2D3 
(b) L2A1
0.300 
1.301
−1.000−0.62 
−1.65
−1.03

MNDA(a) L2A1 
(b) L2B2
−1.136 
0.313
−1.4492.29 
0.76
−1.53
(a) L2B2 
(b) L2A2
0.313 
−0.229
0.5420.76 
1.88
1.12
(a) L2A2 
(b) L2D1
−0.229 
−0.185
−0.0441.88 
1.31
−0.57
(a) L2D1 
(b) L2C2
−0.185 
−0.015
−0.1701.31 
−0.08
−1.39
(a) L2C2 
(b) L2D2
−0.015 
0.349
−0.365−0.08 
0.17
0.25
(a) L2D2 
(b) L2D3
0.349 
0.357
−0.0080.17 
1.10
0.93
(a) L2D3 
(b) L2A1
0.357 
−1.136
1.4931.10 
2.29
1.19

B.

The complete list of all enriched GO terms for 192 biomarker genes (sorted according to value) is shown in Table 4.


GO termDescription# gene value

GO:0006952Defense response240.000000
GO:0006955Immune response240.000002
GO:0019882Antigen processing and presentation70.000073
GO:0009617Response to bacterium110.000088
GO:0006968Cellular defense response70.000149
GO:0001817Regulation of cytokine production110.000172
GO:0048534Hemopoietic or lymphoid organ development130.000562
GO:0030097Hemopoiesis120.000999
GO:0002520Immune system development130.001183
GO:0002696Positive regulation of leukocyte activation80.001499
GO:0001819Positive regulation of cytokine production70.001573
GO:0050867Positive regulation of cell activation80.002053
GO:0007610Behavior140.004162
GO:0002763Positive regulation of myeloid leukocyte differentiation40.004167
GO:0006412Translation120.004210
GO:0002684Positive regulation of immune system process100.006451
GO:0006414Translational elongation70.006467
GO:0002822Regulation of adaptive immune response based on somatic recombination of immune receptors built from immunoglobulin superfamily domains50.007075
GO:0048584Positive regulation of response to stimulus100.007697
GO:0001775Cell activation110.008090
GO:0051240Positive regulation of multicellular organism process100.008151
GO:0002819Regulation of adaptive immune response50.008161
GO:0030162Regulation of proteolysis50.008161
GO:0001818Negative regulation of cytokine production40.008578
GO:0002824Positive regulation of adaptive immune response based on somatic recombination of immune receptors built from immunoglobulin superfamily domains40.008578
GO:0002544Chronic inflammatory response30.009217
GO:0010829Negative regulation of glucose transport30.009217
GO:0050832Defense response to fungus30.009217
GO:0051043Regulation of membrane protein ectodomain proteolysis30.009217
GO:0002694Regulation of leukocyte activation80.009343
GO:0002706Regulation of lymphocyte mediated immunity50.009351
GO:0019725Cellular homeostasis130.009574
GO:0002821Positive regulation of adaptive immune response40.010479
GO:0042742Defense response to bacterium50.010649
GO:0042592Homeostatic process180.011731
GO:0006873Cellular ion homeostasis110.012021
GO:0050865Regulation of cell activation80.012239
GO:0009611Response to wounding150.012528
GO:0007267Cell-cell signaling150.013447
GO:0050778Positive regulation of immune response70.013518
GO:0031640Killing of cells of another organism30.013542
GO:0055082Cellular chemical homeostasis110.013819
GO:0002705Positive regulation of leukocyte mediated immunity40.014959
GO:0002708Positive regulation of lymphocyte mediated immunity40.014959
GO:0010827Regulation of glucose transport40.014959
GO:0034101Erythrocyte homeostasis50.015220
GO:0030098Lymphocyte differentiation60.016523
GO:0051251Positive regulation of lymphocyte activation60.016523
GO:0050801Ion homeostasis110.017255
GO:0045639Positive regulation of myeloid cell differentiation40.017545
GO:0002703Regulation of leukocyte mediated immunity50.018862
GO:0006935Chemotaxis70.020473
GO:0042330Taxis70.020473
GO:0045765Regulation of angiogenesis50.020868
GO:0006954Inflammatory response100.021243
GO:0046649Lymphocyte activation80.024760
GO:0045321Leukocyte activation90.026088
GO:0002761Regulation of myeloid leukocyte differentiation40.026706
GO:0050870Positive regulation of T cell activation50.030175
GO:0048872Homeostasis of number of cells60.030221
GO:0009620Response to fungus30.030551
GO:0032655Regulation of interleukin-12 production30.030551
GO:0031349Positive regulation of defense response50.032829
GO:0002699Positive regulation of immune effectors process40.033982
GO:0032103Positive regulation of response to external stimulus50.035615
GO:0048878Chemical homeostasis120.036704
GO:0002504Antigen processing and presentation of peptide or polysaccharide antigen via MHC class II30.037413
GO:0016525Negative regulation of angiogenesis30.037413
GO:0050850Positive regulation of calcium-mediated signaling30.037413
GO:0006928Cell motion130.037784
GO:0006334Nucleosome assembly40.037967
GO:0031497Chromatin assembly40.037967
GO:0048002Antigen processing and presentation of peptide antigen30.044801
GO:0010033Response to organic substance180.044896
GO:0030218Erythrocyte differentiation40.046616
GO:0065004Protein-DNA complex assembly40.046616
GO:0002521Leukocyte differentiation60.046793
GO:0014070Response to organic cyclic substance60.049526
GO:0002237Response to molecule of bacterial origin50.051544
GO:0007626Locomotory behavior80.052288
GO:0055080Cation homeostasis80.052288
GO:0050848Regulation of calcium-mediated signaling30.052676
GO:0051249Regulation of lymphocyte activation60.055270
GO:0034728Nucleosome organization40.056149
GO:0045619Regulation of lymphocyte differentiation40.056149
GO:0002697Regulation of immune effectors process50.058846
GO:0002712Regulation of B cell mediated immunity30.060999
GO:0002889Regulation of immunoglobulin mediated immune response30.060999
GO:0051235Maintenance of location40.061237
GO:0032914Positive regulation of transforming growth factor-beta 1 production20.061881
GO:0045651Positive regulation of macrophage differentiation20.061881
GO:0007611Learning or memory50.062693
GO:0008285Negative regulation of cell proliferation100.063328
GO:0051241Negative regulation of multicellular organism process60.064581
GO:0051789Response to protein stimulus50.066670
GO:0016192Vesicle-mediated transport130.067653
GO:0042060Wound healing70.068209
GO:0001906Cell killing30.069736
GO:0046635Positive regulation of alpha-beta T cell activation30.069736
GO:0051094Positive regulation of developmental process90.072195
GO:0050863Regulation of T cell activation50.075009
GO:0050900Leukocyte migration40.077725
GO:0009895Negative regulation of catabolic process30.078852
GO:0046634Regulation of alpha-beta T cell activation30.078852
GO:0030099Myeloid cell differentiation50.079368
GO:0055066Di-, tri-valent inorganic cation homeostasis70.080251
GO:0006323DNA packaging40.083611
GO:0030217T cell differentiation40.083611
GO:0043066Negative regulation of apoptosis100.085435
GO:0046324Regulation of glucose import30.088316
GO:0002863Positive regulation of inflammatory response to antigenic stimulus20.091380
GO:0032891Negative regulation of organic acid transport20.091380
GO:0032908Regulation of transforming growth factor-beta 1 production20.091380
GO:0051918Negative regulation of fibrinolysis20.091380
GO:0043069Negative regulation of programmed cell death100.092873
GO:0060548Negative regulation of cell death100.095435
GO:0045637Regulation of myeloid cell differentiation40.095934
GO:0030003Cellular cation homeostasis70.096877
GO:0045582Positive regulation of T cell differentiation30.098097

Competing Interests

The authors declare that they have no competing interests.

References

  1. P. U. Unschuld, Huang Di Nei Jing Su Wen, University of California Press, Los Angeles, Calif, USA, 2003.
  2. K.-H. Lee, S. Morris-Natschke, X. Yang et al., “Recent progress of research on medicinal mushrooms, foods, and other herbal products used in traditional chinese medicine,” Journal of Traditional and Complementary Medicine, vol. 2, no. 2, pp. 84–95, 2012. View at: Google Scholar
  3. F. Yu, T. Takahashi, J. Moriya et al., “Traditional Chinese medicine and kampo: a review from the distant past for the future,” Journal of International Medical Research, vol. 34, no. 3, pp. 231–239, 2006. View at: Publisher Site | Google Scholar
  4. M. Xue, H. J. Yin, and K. J. Chen, “Some thoughts on studying essence of syndromes based on genomics,” Zhongguo Zhong Xi Yi Jie He Za Zhi, vol. 26, no. 1, pp. 88–90, 2006. View at: Google Scholar
  5. J. van der Greef, H. van Wietmarschen, J. Schroën, M. Wang, T. Hankemeier, and G. W. Xu, “Systems biology-based diagnostic principles as pillars of the bridge between Chinese and Western medicine,” Planta Medica, vol. 76, no. 17, pp. 2036–2047, 2010. View at: Publisher Site | Google Scholar
  6. W. J. Ding, Y. Z. Zeng, W. H. Li et al., “Identification of linkage disequilibrium SNPs from a kidney-yang deficiency syndrome pedigree,” American Journal of Chinese Medicine, vol. 37, no. 3, pp. 427–438, 2009. View at: Publisher Site | Google Scholar
  7. W. J. Ding, S. L. Yan, Y. Z. Zeng et al., “Insufficient activity of MAPK pathway is a key monitor of Kidney-Yang Deficiency Syndrome,” Journal of Alternative and Complementary Medicine, vol. 15, no. 6, pp. 653–660, 2009. View at: Publisher Site | Google Scholar
  8. L. Yang, M. Wang, W. Wu, and L. Zhang, “Transcriptome analysis of cold syndrome using microarray,” American Journal of Chinese Medicine, vol. 35, no. 4, pp. 609–620, 2007. View at: Publisher Site | Google Scholar
  9. G. Chen, C. Lu, Q. Zha et al., “A network-based analysis of traditional Chinese medicine cold and hot patterns in rheumatoid arthritis,” Complementary Therapies in Medicine, vol. 20, no. 1-2, pp. 23–30, 2012. View at: Publisher Site | Google Scholar
  10. Q. Pan-Hammarström, S. Wen, and L. Hammarström, “Cytokine gene expression profiles in human lymphocytes induced by a formula of traditional Chinese medicine, vigconic VI-28,” Journal of Interferon and Cytokine Research, vol. 26, no. 9, pp. 628–636, 2006. View at: Publisher Site | Google Scholar
  11. C. Lu, C. Xiao, G. Chen et al., “Cold and heat pattern of rheumatoid arthritis in traditional Chinese medicine: distinct molecular signatures indentified by microarray expression profiles in CD4-positive T cell,” Rheumatology International, vol. 32, no. 1, pp. 61–68, 2012. View at: Publisher Site | Google Scholar
  12. J. Liu, K. Peter, D. Shi et al., “Traditional formula, modern application: Chinese medicine formula Sini tang improves early ventricular remodeling and cardiac function after myocardial infarction in rats,” Evidence-Based Complementary and Alternative Medicine, vol. 2014, Article ID 141938, 10 pages, 2014. View at: Publisher Site | Google Scholar
  13. S. M. Sun, Qian Jin Yi Fang, People's Medical Publishing House, Beijing, China, 1955.
  14. World Health Organization and Regional Office for the Western Pacific, WHO International Standard Terminologies on Traditional Medicine in the Western Pacific Region, World Health Organization Regional Office for the Western Pacific, Manila, Philippines, 2007.
  15. M. K. Kerr and G. A. Churchill, “Experimental design for gene expression microarrays,” Biostatistics, vol. 2, no. 2, pp. 183–201, 2001. View at: Publisher Site | Google Scholar
  16. C.-R. Chen, W.-Y. Shu, M.-L. Tsai, W.-C. Cheng, and I. C. Hsu, “THEME: a web tool for loop-design microarray data analysis,” Computers in Biology and Medicine, vol. 42, no. 2, pp. 228–234, 2012. View at: Publisher Site | Google Scholar
  17. J. Schuchhardt, D. Beule, A. Malik et al., “Normalization strategies for cDNA microarrays,” Nucleic Acids Research, vol. 28, no. 10, p. E47, 2000. View at: Publisher Site | Google Scholar
  18. W.-C. Cheng, W.-Y. Shu, C.-Y. Li et al., “Intra- and inter-individual variance of gene expression in clinical studies,” PLoS ONE, vol. 7, no. 6, Article ID e38650, 2012. View at: Publisher Site | Google Scholar
  19. D. W. da Huang, B. T. Sherman, and R. A. Lempicki, “Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists,” Nucleic Acids Research, vol. 37, no. 1, pp. 1–13, 2009. View at: Publisher Site | Google Scholar
  20. M. Ashburner, C. A. Ball, J. A. Blake et al., “Gene ontology: tool for the unification of biology. The Gene Ontology Consortium,” Nature Genetics, vol. 25, no. 1, pp. 25–29, 2000. View at: Publisher Site | Google Scholar
  21. M. Kanehisa and S. Goto, “KEGG: kyoto encyclopedia of genes and genomes,” Nucleic Acids Research, vol. 28, no. 1, pp. 27–30, 2000. View at: Publisher Site | Google Scholar

Copyright © 2016 Hung-Tsu Cheng et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


More related articles

 PDF Download Citation Citation
 Download other formatsMore
 Order printed copiesOrder
Views2676
Downloads596
Citations

Related articles

We are committed to sharing findings related to COVID-19 as quickly as possible. We will be providing unlimited waivers of publication charges for accepted research articles as well as case reports and case series related to COVID-19. Review articles are excluded from this waiver policy. Sign up here as a reviewer to help fast-track new submissions.