Analysis of differentially expressed proteins involved in hand, foot and mouth disease and normal sera


Corresponding author: L. Zhang, School of Basic Medical Sciences, Southern Medical University, Guangzhou 510515, Guangdong, China
E-mail:; L. Zhu, School of Public Health and Tropical Medicine, Southern Medical University, Guangzhou 510515, Guangdong, China


Clin Microbiol Infect 2012; 18: E188–E196


We implemented 2-D DIGE technology on proteins prepared from serum obtained from children with hand, foot and mouth disease (HFMD) and controls, to study the differentially expressed proteins in control and HFMD serum samples. Proteins found to be differentially expressed were identified with matrix-assisted laser desorption/ionization time-of-flight/ time-of-flight mass spectrometry (MALDI-TOF/TOF MS) analysis. We identified 30 proteins from mild HFMD samples and 39 proteins from severe HFMD samples, compared with the normal controls. 25 proteins among them (14 up-regulated and 11 down-regulated proteins) are found in both HFMD groups. Classification analysis and protein–protein interaction map showed that they associate with multiple functional groups, including transporter activity and atalytic activity. These findings build up a comprehensive profile of the HFMD proteome and provide a useful basis for further analysis of the pathogenic mechanism and the regulatory network of HFMD.


Hand, foot and mouth disease (HFMD), a highly infectious disease that mainly affects children <5 years old, is characterized by fever, papules on the hands and feet, and mouth ulcers. It can be caused by viruses that belong to the enterovirus genus (group), including polioviruses, coxsackieviruses, echoviruses and enteroviruses [1]. Among them, Coxsackieviruses 16 (CA16) and enterovirus 71 (EV71) have been identified as the most frequent pathogens of HFMD [2]. The virus transmits through a direct contact with mucous, oral or nasal secretions or faeces of an infected person. HFMD patients with or without symptoms are the major infectious sources [3]. The disease causes fever, tetter on the hands and feet and ulceration in the mouth, and may further develop into myocarditis, pulmonary oedema, aseptic meningoencephalitis and other complications [4,5].

Since the first HFMD case was reported in New Zealand in 1957, subsequent outbreaks have been reported worldwide. Historically, the outbreaks of HFMD were reported spontaneously and at a small scale in European countries and the United States [6]. Since 1997, large epidemic outbreaks of HFMD associated with severe neurological complications and high mortality rates have occurred in Malaysia, Taiwan, Singapore, Japan and other Asian-Pacific areas. Large outbreaks of HFMD have been reported since 2008 in mainland China, resulting in millions of infections and hundreds of deaths in children [2]. Noticeably, both the number of severe cases and mortality rate for HFMD have significantly increased in recent years and the epidemics of HFMD pose serious public health threats to children throughout the world [7,8]. Although treatment has been improved with the application of new technologies, there is currently no specific antiviral therapy to cure HFMD and no vaccine to prevent severe HFMD infection [2].

Usually the enterovirus exhibits strong transitivity; moreover, the recent infection rate is high and the transmission paths are complicated, causing large epidemics in a short period of time. Screening and identification of HFMD play important roles in preventing its occurrence and development. Therefore rapid and accurate diagnosis is very important. Serum is rich in protein composition which changes in abundance during different stages. Discovering molecular markers from serum may be valuable in the early diagnosis of HFMD.

Proteomics analysis is currently considered to be a powerful tool for global evaluation of protein expression, and has been widely applied in studying biomarkers for diseases [9]. Two-dimensional fluorescence difference gel electrophoresis (2D-DIGE), a new proteomics technology, has been developed in recent years. It shows significantly improved sensitivity and repeatability when compared with the traditional two-dimensional electrophoresis. It is a new means to provide an overall view of the proteome of the disease.

A recent study carried out a comparative proteome analysis between EV71 and CA16, the two most causative agents of HFMD in cultured human rhabdomyosarcoma (RD) host cells [10]. The study showed the effect of the viral infection (e.g. active shutting down of the protein synthesis of the host cell), as well as cellular response to the viral infection, such as the intracellular immune response. The study also provided interesting molecular explanations of the disease phenotype, such as loss of muscle tone caused by down-regulated HSP27 and desmin. However, the study focused on the intracellular differentially expressed proteins using cancer cell lines, which deviates from the in vivo situation. Moreover, the infection may cause a series of responses by multiple cell types in the human body, which is far more complex than the cell line model. Also, there is no analysis of the extracellular proteins in infected cells. In this study, we aim to identify potential proteins as biomarkers in HFMD serum samples using 2D-DIGE and matrix-assisted laser desorption/ionization time-of-flight/ mass spectrometry (MALDI-TOF MS) methodologies to identify differentially expressed proteins in HFMD compared with normal serum samples and further to determine the possible molecular markers for HFMD. Classification analysis of the proteins identified based on the PANTHER and protein–protein interaction analysis indicated that these proteins might have important roles in a variety of cellular processes and structures, including spermatogenesis, cell signalling, cell skeleton and metabolism.

Materials and Methods

Preparation of serum samples

Blood samples from 16 HFMD children (eight mild and eight severe) according to the ‘foot and mouth disease prevention control guide’ (2008 Edition) issued by the Ministry of Health, China (, were randomly selected to be analysed. These children were proven to have no other disease after a series of checks in the hospital. Clinical symptoms and laboratory test (Enterovirus 71 (EV71) nucleic acid detection kit) confirmed that the EV71 infections caused all these HFMD cases. Another eight blood samples from normal children were collected as controls. A pooled sample consisting of equal amounts of each of 16 experimental samples (mild HFMD compared with normal control; severe HFMD compared with normal control) was used as a pooled internal standard. Serum was separated by centrifugation at 800 g for 30 min. Aliquots of serum were collected and stored at −80°C until ready for use. Serum samples were processed using the ProteoPrep Blue Albumin Depletion Kit (Sigma, St. Louis, MO, USA), which selectively removes albumin and IgG from the serum sample according to the manufacturer’s instructions. To depurate the protein extraction and determine the final protein concentration, the 2-D Clean-up Kit (GE Healthcare, UK) and 2-D Quant Kit (GE Healthcare, London, UK) were used sequentially, following the manufacturer’s instructions.

Protein labelling with CyDye DIGE fluor

Protein extracts were labelled with three CyDye DIGE fluors, Cy2, Cy3 and Cy5, for 2-D DIGE technology according to the manufacturer’s recommended protocols. A total of 50 μg of protein from a normal sample was labelled with 400 pmol of Cy3, and 50 μg of protein from a mild HFMD sample or severe HFMD sample was labeled with 400 pmol of Cy5. An internal reference standard, consisting of two mixed samples used in the experiment, was labelled with Cy2. The labelling reaction was incubated on ice for 30 min under dark conditions. The reactions were then quenched with the addition of 10 mM lysine for 10 min on ice in the dark. The labelled samples were pooled and prepared for the subsequent steps of the experiment.

2-D electrophoresis

2-D electrophoresis was performed as described earlier with a few minor modifications [11]. Immobilized pH gradient (IPG) strips (24 cm, pH 4–7) were hydrated in hydration buffer (7 M urea, 2 M thiourea, 4% w/v CHAPS, 40 mM DTT, 1% IPG buffer, pH 4–7 and 0.002% w/v Bromophenol blue) and isoelectric focusing (IEF) was performed according to the manufacturer’s instructions using pH 4–7 IPG strips with an Ettan IPGphor II System (GE Healthcare). After the isoelectric focusing (IEF), the proteins were reduced and alkylated by successive treatment for 15 min with equilibration buffer containing 2% w/v DTT followed by 2.5% w/v IAA. Second-dimension separation was performed on a 12.5% SDS-PAGE gel using an Ettan DALT six instrument (GE Healthcare).

Gel image acquisition and analysis

Labelled proteins were visualized using the Typhoon 9400 imager (GE Healthcare) and analysed using DeCyder Software (V6.0; GE Healthcare) as described previously [12,13]. Briefly, the Cy2, Cy3 and Cy5 labelled images for each gel were scanned with excitation/emission wavelengths of 488/520, 532/580 and 633/670 nm, respectively. After the CyDye labelling, signals were imaged, and the gels were stained using Deep Purple total protein stain (GE Healthcare) according to the standard protocol and scanned with excitation/emission wavelengths of 532/560 nm. Then, 2-D DIGE gel images were analysed with DeCyder software. The protein expression patterns of mild HFMD or severe HFMD serum samples were compared with the control (control vs. mild HFMD or control vs. severe HFMD). Protein spots with significant differences in abundance (more than 1.5-fold) [14] were selected in the stained gels for spot picking.

Spot picking and enzymatic digestion

Protein spot-features that were significantly (p <0.05) increased or decreased in all HFMD samples compared with control samples were chosen for further analysis. Three preparative gels were performed to enable recovery of sufficient protein within the individual spot feature to allow protein identification. In the automated procedure, the selected protein spots were picked up, washed with 25 mM NH4HCO3 and 50% v/v methanol and then digested with 12.5 ng/mL trypsin (Sequencing-grade; Promega, Madison, WI, USA) in 25 mM NH4HCO3 for 12 h at 37°C. The tryptic peptides were extracted with 60%v/v ACN and 0.1% v/v TFA, then dissolved in 4 mg/mL CHCA matrix in 70% (v/v) ACN and 0.1% (v/v) TFA and spotted on the mass spectrometry (MS) sample plate.


MS analysis was performed on an Utraflex III MALDI-TOF/TOF MS (Bruker, Karlsruhe, Baden-Württemberg, Germany) operating in positive ion reflector mode. Monoisotopic peak masses were acquired in a mass range of 700–4000 Da with a signal-to-noise ratio (S/N) >200. Four of the most intense ion signals, excluding common trypsin autolysis peaks and matrix ion signals, were automatically selected as precursors for MS/MS acquisition. The peptide mass fingerprint combined MS/MS spectra were searched against the NCBInr database using the Biotools software (Version 3.2; Bruker) and MASCOT Version 2.2 (Matrix Science, Boston, MA, USA) with the following parameters: Homo sapiens, trypsin cleavage, one missed cleavage allowed, carbamidomethylation as fixed modification, oxidation of methionines allowed as variable modification, peptide mass tolerance at 100 ppm, fragment tolerance at 0.7 Da. The criteria for successfully identified proteins include: ion score confidence interval (CI %) for PMF, MS/MS data ≥95%; a minimum of four peptides hits required in PMF data-based identification; and at least two peptides of distinct sequence identified.

Database searching

The MS/MS spectra were searched using MASCOT to search human and EV71 protein databases to identify proteins. The following peptide search criteria were employed: species: Homo sapiens and EV71 virus, trypsin digest (one missed cleavage allowed), peptide tolerance: 100 ppm, mass value: monoisotropic, protein mass ranged from 1 to 100 kDa. NCBI databases were searched. Protein identification by PMF was performed using the MASCOT search engine.

Protein categorization

The identified proteins were classified according to the PANTHER (Protein ANalysis THrough Evolutionary Relationships) system (, which classifies genes and proteins by their functions. The PANTHER ontology, a highly controlled vocabulary (ontology terms) by molecular function, biological process and molecular pathway, was employed to categorize proteins into families and subfamilies with shared functions.

Predication of protein–protein interaction

The identified proteins were preceded by the STRING (functional Protein Association Networks) system ( STRING is a database of known and predicted protein interactions, including direct (physical) and indirect (functional) associations.


2-D DIGE analysis of differential protein expression

The goal of this study was to look for some proteins that were considered potential biomarkers for early diagnosis of HFMD. Characterization of HFMD children’s samples from which serum was obtained for DIGE analysis is presented in Fig. 1. Samples were processed with the ProteoPrep Blue Albumin Depletion kit to remove the two most abundant serum proteins from serum for higher sensitivity to visualize less abundant proteins. As previously described [15], differential protein analysis was then performed using 2-D DIGE. After 2-D DIGE analysis, differential protein spots were detected using DeCyder analysis software. Protein profiles were compared between the HFMD samples and the control samples. We selected the numbered spots that were at least 1.5-fold differentially expressed proteins.

Figure 1.

 Analysis of hand, foot and mouth disease (HFMD) samples as revealed by 2-D DIGE analysis. Proteins extracted from HFMD and control samples were labelled with Cy3 and Cy5, respectively. An internal standard protein sample (a mixture of HFMD and control samples) was labelled with the Cy2 dye. The Cy3-labelled HFMD samples and the Cy5-labelled control samples were mixed and separated on a 2-D DIGE gel (IEF on pH 4–7 strips followed by 12.5% second dimension SDS-PAGE). The green spots represent down-regulated proteins, while the red spots represent up-regulated proteins in HFMD samples compared with the control samples. The yellow arrow represents the identified proteins that showed significantly differential expression in the HFMD and control samples. The number in the figure corresponds to that presented in Tables 1 and 2. (a) Mild HFMD vs control samples. (b) Severe HFMD vs control samples.

MALDI-TOF PMF analysis

A total of 30 differentially expressed proteins (18 up-regulated and 12 down-regulated) were identified in the mild HFMD samples compared with the level in control samples. Thirty-nine differentially expressed proteins (22 up-regulated and 17 down-regulated) were identified in the severe HFMD samples compared with the level in control samples (Table 1). Table 2 summarizes the 25 proteins that all appeared in both the mild and severe HFMD samples compared with the level in control samples. Among these, 14 proteins had increased and 11 proteins had decreased in the HFMD samples compared with the level in control samples.

Table 1.   Differential proteins identified by mass spectrometry (MS) after 2-D DIGE of the mild and severe hand, foot and mouth disease (HFMD) samples vs. control samples
Master No.Accession No.MrpIMASCOT scoreCoverage (%)Av. ratiot-testMwProteins
Mild HFMDSevere HFMDMild HFMDSevere HFMD
  1. ‘—’ not detected.

272gi|896277117764.93728−1.510.0036130This CDS feature is included to show the translation of the corresponding V_region. Presently translation qualifiers on the V_region features are illegal
274gi|559580631056064.81349−1.96−1.660.000210.0064130Inter-alpha (globulin) inhibitor H2
279gi|119581148577544.8234420−1.89−1.550.000760.0085130Keratin 9 (epidermolytic palmoplantar keratoderma)
422gi|134105218855836.32339201.630.0067111Chain A, human complement factor B
442gi|2240531620725.452345−2.24−2.250.00230.0023109Macroglobulin alpha2
480gi|1778701646006.2658213−2.1−1.70.000470.0028107Alpha-2-macroglobulin precursor
552gi|4501987709635.333227−1.62−1.650.00830.01297Afamin precursor
716gi|2258128617285.26304201.70.005768Complement 9
912gi|6573461374996.1212217−1.7−1.520.00750.003462Chain A, crystal structure of human beta-2-glycoprotein-I (apolipoprotein-H)
1083gi|16418467383824.54324141.761.590.010.01251Leucine-rich alpha-2-glycoprotein 1 precursor
1103gi|113197968481714.4911561.690.01349KRT9 protein
1137gi|27692693486415.9910424−1.640.004946ALB protein
1163gi|169214179456425.09304311.750.03842Predicted: similar to complement component 3
1192gi|78101271402044.82220372.022.410.0140.01239Chain C, human complement component C3c
1203gi|15277503405365.38467461.530.03338ACTB protein
1230gi|47124562316475.61551341.90.02837HP protein
1237gi|21961605590205.54346242.080.03236Keratin 10
1242gi|38305346170324.2237222.642.740.030.004136Antigen MLAA-42
1257gi|119579599513885.41265142.772.750.00620.008135Haptoglobin, isoform CRA_b
1258gi|47124562316475.48386232.072.610.0140.02936HP protein
1265gi|1181380171049134.99628262.162.980.0380.04634Chain B, human complement component C3b
1279gi|1181379651049124.97468151.563.40.0440.03333Chain B, structure of complement C3b: insights into complement activation and regulation
1291gi|157831596442805.9413626−1.590.02533Chain A, Alpha1-Antitrypsin
1340gi|27574252159285.2339291.62.020.00850.002331Chain B, deoxy haemoglobin (A-Gly-C:v1m; B,D:v1m,C93a,N108k)
1347gi|27754776333956.21400501.510.03831Ficolin 3 isoform 1 precursor
1378gi|122920512684086.0619415−1.950.008730Chain A, human serum albumin complexed with myristate and aspirin
1390gi|119390093236896.41539522.170.04930Chain B, Crig bound to C3c
1437gi|170684576244266.49226452.821.950.0480.0329Immunoglobulin kappa 4 light chain
1493gi|576259233585.44344261.51.580.0160.03928Chain A, the structure of pentameric human serum amyloid P component
1531gi|55961583133275.6811231−1.60.02526Apolipoprotein M
1561gi|178779433585.2511116−2.37−2.920.00910.005623Apolipoprotein A-IV precursor
1566gi|563320281415.1426423−2.64−2.960.0110.00421Apolipoprotein A-IV
1591gi|223092186725.3153142.94.310.0310.02919Ceruloplasmin CpF5 fragment
1617gi|224510585129965.6846196−1.75−1.970.00630.0008216Chain A, crystal structure of the apo form of human wild-type transthyretin
1620gi|189339646127405.424656−1.610.01116Chain A, crystal structure of the F87mL110M mutant of human Transthyretin At Ph 4.6.
1664gi|18697273687594.339924−2.2−2.320.00130.002812Chain A, structure and dynamics of human apolipoprotein C-Iii
1669gi|225986116756.29945324.2629.280.0160.01210Amyloid-related serum protein SAA
1674gi|1427777089094.68673−2.15−2.760.0140.003810Chain A, Nmr structure of human apolipoprotein C-Ii in the presence of SDS
Table 2.   Differential proteins identified all appeared by mass spectrometry (MS) after 2-D DIGE in both the mild and severe hand, foot and mouth disease samples vs. control samples
Master noAccession No.MrpIMwProteins
274gi|559580631056064.8130Inter-alpha (globulin) inhibitor H2
279gi|119581148577544.82130Keratin 9 (epidermolytic palmoplantar keratoderma)
442gi|2240531620725.45109Macroglobulin alpha2
480gi|1778701646006.26107Alpha-2-macroglobulin precursor
552gi|4501987709635.397Afamin precursor
912gi|6573461374996.1262Chain A, crystal structure of human beta-2-glycoprotein-I (Apolipoprotein-H)
1083gi|16418467383824.5451Leucine-rich alpha-2-glycoprotein 1 precursor
1192gi|78101271402044.8239Chain C, human complement component C3c
1242gi|38305346170324.2236Antigen MLAA-42
1257gi|119579599513885.4135Haptoglobin, isoform CRA_b
1258gi|47124562316475.4836HP protein
1265gi|1181380171049134.9934Chain B, human complement component C3b
1279gi|1181379651049124.9733Chain B, structure of complement C3b: insights into complement activation and regulation
1340gi|27574252159285.2331Chain B, deoxy haemoglobin (A-Gly-C:v1m; B,D:v1m,C93a,N108k)
1437gi|170684576244266.4929Immunoglobulin kappa 4 light chain
1493gi|576259233585.4428Chain A, the structure of pentameric human serum amyloid P component
1561gi|178779433585.2523Apolipoprotein A-IV precursor
1566gi|563320281415.1421Apolipoprotein A-IV
1591gi|223092186725.3119Ceruloplasmin CpF5 fragment
1617gi|224510585129965.6816Chain A, crystal structure of the apo form of human wild-type transthyretin
1664gi|18697273687594.3312Chain A, structure and dynamics of human apolipoprotein C-Iii
1669gi|225986116756.2910Amyloid related serum protein SAA
1674gi|1427777089094.610Chain A, Nmr structure of human apolipoprotein C-Ii in the presence of SDS

Classification of identified proteins

According to cellular functions and processes, the proteins were distributed into categories based on molecular function, biological processes involved and protein classes using the PANTHER classification system. The results are shown in Fig. 2. The most dominant function that the identified protein was involved in was transporter activity (28.0%) followed by binding (24.0%), enzyme regulator activity (16.0%), catalytic activity (16.0%) and receptor activity (12.0%). Most of the identified proteins played roles in the metabolic process (17.3%) and immune system process (15.4%). Other biological processes that the proteins were involved in included response to stimulus (11.5%), cellular process (11.5%), transport (11.5%), etc. As for protein class, most of the proteins belonged to transporter (40.0%), transfer/carrier protein (13.3%) or signalling molecule (13.3%); others belonged to receptor (6.7%), defence/immunity protein (6.7%), enzyme modulator (6.7%), protease (6.7%) or structural protein (6.7%) (Fig. 2a–c).

Figure 2.

 Classification analysis of proteins found in hand, foot and mouth disease (HFMD) samples. Categorization was based on information that was obtained from the online PANTHER classification system. (a) Classification of the identified proteins of HFMD samples according to the molecular function. (b) Classification of the identified proteins of HFMD samples according to the biological processes. (c) Classification of the identified proteins of HFMD samples according to the protein class.

Protein–protein interaction network

The protein interaction network for the identified proteins was constructed by STRING (Fig. 3) in order to better understand the pathogenic mechanisms in HFMD. Biological systems can be modelled as complex network systems with many interactions among the components in different pathways. These interactions provide us with essential information about the function and behaviour of the analysed system.

Figure 3.

 The network of protein–protein interactions of HFMD.


Treatment of HFMD is currently symptomatic, and mild HFMD resolves spontaneously without complications within 7–10 days. However, severe HFMD can cause severe neurological symptoms and complications, such as acute flaccid paralysis, aseptic meningitis and neurogenic pulmonary oedema. In recent years, although treatment has been updated with the application of new technologies, there is currently no specific antiviral therapy to cure and no vaccine to prevent severe HFMD infection. Therefore, early detection of severe HFMD is important and precise staging is necessary for assessing the risk of progression and planning the appropriate treatment.

Due to the limited practicability of clinical markers for early diagnosis, the focus of HFMD investigation has shifted to molecular markers. Protein profiles of specific stages can be more accurate in reflecting the status of disease progression. In the present study, we took advantage of DIGE-based proteomics to identify stage-specific markers that reflect the status of HFMD. DIGE-based proteomics has the advantages of adequate sensitivity, high reproducibility, wide linear dynamic range and minimized experimental variation over traditional proteomics because of internal standards and fluorescence labelling. Although no protein identified was reported to have direct connection to HFMD the majority of the differentially expressed proteins are involved in the host immune response system, inflammation and protein aggregation.

ITIH2 (inter-alpha (globulin) inhibitor H2), A2M (macroglobulin alpha2), etc. (protease inhibitors), were suggested to be involved in the viral infection. ITIH2 belongs to the inter-alpha-trypsin inhibitors (ITI) family, which is involved in extracellular matrix stabilization and the host-virus interaction process. A2M, a multifunctional protease inhibitor and cytokine transporter, inhibits various proteases, including trypsin, thrombin and collagenase, and is an essential component in salivary innate immunity against haemaglutination mediated by S-OIV [16]. LRG1 (leucine-rich alpha-2-glycoprotein 1 precursor) belongs to the leucine-rich repeat (LRR) family of proteins, which is expressed during granulocyte differentiation [17], indicating that LRG1 might play a role in the immune response when the host clears infections. The up-regulation of LRG1 and down-regulation of ITIH2 and A2M suggests the immune response to infection with HFMD.

Interestingly, A2M is also involved in the amyloid response in Alzheimer’s disease together with apolipoproteins [18]. These proteins bind the misfolded Aβ and the complex is then cleared by LRP1 (low density lipoprotein receptor-related protein 1). Down-regulation of A2M in HFMD decreases the amyloid clearance. APCS (chain A, the structure of pentameric human serum amyloid P component) and TTR (chain A, crystal structure of the apo form of human wild-type transthyretin), which were found to be up-regulated in HFMD patients, are also related to protein aggregation. APCS is a universal constituent of the abnormal tissue deposits in amyloidosis, and its glycosylated form is resistant to degradation, causing the amyloid deposits to remain [19]. Extracellular deposition of wild-type TTR caused by transthyretin amyloidosis leads to amyloid fibril formation in various tissues and organs, and may further result in failure of organs (e.g. heart failure) [20,21]. SAA, which was found to be the most up-regulated (more than 20-fold) in both mild and severe HFMD, was reported to be closely related to tissue damage or inflammation (may increase 1000-fold) [22]. Interestingly, SAA is also found to be elevated in Alzheimer’s disease and Down’s syndrome, especially in the prolonged stage [23]. As an acute phase protein, SAA is found in only some patients with early-stage AD; however, the inflammation may impair Aβ clearance due to the loss of apoE from CSF-HDL [23,24]. These findings strongly suggested that the amyloidosis, including increasing of the amyloidgenic material (TTR), increasing of the amyloid-stabilizing factor (APCS) and decreasing of the amyloid clearance (A2M and SAA), may be an important molecular event in HFMD.

When comparing with the previous study using a cultured cancer cell line [10], β-actin is the only protein among their findings which appears in our study. In our study, we found that β-actin was up-regulated in the serum of the severe HFMD patients. β-Actin is an intracellular protein. Its increase in serum may indicate cell death caused by serious infection.

In the current study, although none of the above-mentioned differentially expressed proteins is specifically correlated with HFMD, the combination of their variations may still serve as a useful hint for diagnosis of early-stage HFMD prior to the appearance of severe symptoms.


This work was supported by Guangdong Natural Science Foundation (9151008901000033), National clinical key specialty project Foundation, the Program for New Century Excellent Talents in University (NCET-08-0647 and NCET-09-0088), the Key Program of Natural Science Foundation of Guangdong Province (92510515000008) and the Foundation for Sci & Tech Research Project of Guangdong (2009A030200015).

Transparency Declaration

The authors declare that no conflicting of interests exist.