Sage Journals: Discover world-class research

Abstract

It is known that high-risk human papillomavirus infection is the main etiological factor in cervical carcinogenesis. However, human papillomavirus screening is not sufficient for early diagnosis. In this study, we aimed to identify potential biomarkers common to cervical carcinoma and human papillomavirus infection by proteomics for human papillomavirus–based early diagnosis and prognosis. To this end, we collected 76 cases of fresh cervical tissues and 116 cases of paraffin-embedded tissue slices, diagnosed as cervical squamous cell carcinoma, cervical intraepithelial neoplasia II–III, or normal cervix from ethnic Uighur and Han women. Human papillomavirus infection by eight oncogenic human papillomavirus types was detected in tissue DNA samples using a quantitative polymerase chain reaction. The protein profile of cervical specimens from human papillomavirus 16–positive squamous cell carcinoma and human papillomavirus–negative normal controls was analyzed by proteomics and bioinformatics. The expression of candidate proteins was further determined by quantitative reverse transcriptase-polymerase chain reaction and immunohistochemistry. We identified 67 proteins that were differentially expressed in human papillomavirus 16–positive squamous cell carcinoma compared to normal cervix. The quantitative reverse transcriptase-polymerase chain reaction analysis verified the upregulation of ASAH1, PCBP2, DDX5, MCM5, TAGLN2, hnRNPA1, ENO1, TYPH, CYC, and MCM4 in squamous cell carcinoma compared to normal cervix (p < 0.05). In addition, the transcription of PCBP2, MCM5, hnRNPA1, TYPH, and CYC was also significantly increased in cervical intraepithelial neoplasia II–III compared to normal cervix. Immunohistochemistry staining further confirmed the overexpression of PCBP2, hnRNPA1, ASAH1, and DDX5 in squamous cell carcinoma and cervical intraepithelial neoplasia II–III compared to normal controls (p < 0.05). Our data suggest that the expression of ASAH1, PCBP2, DDX5, and hnRNPA1, and possibly MCM4, MCM5, CYC, ENO1, and TYPH, is upregulated during cervical carcinogenesis and potentially associated with human papillomavirus infection. Further validation studies of the profile will contribute to establishing auxiliary diagnostic markers for human papillomavirus–based cancer prognosis.

Keywords

Cervical carcinoma isobaric tags for relative and absolute quantitation proteomics human papillomavirus infection cervical specimen

Introduction

Cervical cancer is one of the leading causes of cancer death among women in developing countries, with an estimated 450,000 new cases and 250,000 deaths each year.^1,2 Invasive cancer accounts for more than 70% of patients diagnosed with cervical cancer, among which squamous cell carcinoma (SCC) is the most frequent type.^2,3 Detection of cervical cancer at an early stage is crucial for effective treatment and associated with excellent survival rates, but women diagnosed with the advanced disease are often untreatable and have poor outcomes.⁴ The Pap smear test is widely adopted as a cytological screening method for cervical dysplasia and precancerous lesions and may greatly reduce the incidence of cervical cancer in organized screening programs.^5,6 Genital infection by human papillomaviruses (HPVs), in particular, by oncogenic high-risk HPV types (such as HPV16 and 18) is generally known as the main etiological factor in cervical carcinogenesis.^7,8 Accordingly, HPV genotyping alone or with a Pap test is available as an alternative to cytology screening. However, most of the HPV infections are transient and harmless or eliminated by the host immunity. Even persistent infections have a relatively long latent period before inducing cancer, and thus, HPV testing may not improve cancer prognosis.^9,10 The p16INK4a/Ki-67 dual staining cytology is introduced as a high-sensitive alternative for the screening of cervical intraepithelial neoplasia (CIN) that are high-risk HPV positive and Pap negative.^11,12 However, since the expression of p16INK4a and Ki-67 is not dependent on HPV infection or progression to cervical carcinoma, the specificity of this dual staining is quite limited.¹³ Thus, HPV-based screening or prognostication of cervical carcinoma may require the use of auxiliary diagnostic biomarkers that are highly associated with the infection by high-risk HPVs.

Proteins are executors of biological functions and important targets in cancer research in the post-genomic era. The advances in emerging proteomic techniques provide options for the screening of cancer-associated biomarkers and drug targets.^14,15 The two-dimensional gel electrophoresis (2DE) is widely used in the profiling of cellular proteins, but it has low sensitivity.¹⁶ With the development of a two-dimensional difference in gel electrophoresis (2D-DIGE), this technology is greatly improved in accuracy and sensitivity and successful in the identification of low-abundance proteins, such as potential tumor biomarkers.¹⁷ The proteomics strategy based on isotopic labeling with isobaric tags for relative and absolute quantitation (iTRAQ) reagents currently allows for the analysis of up to eight different samples simultaneously and quantitative detection of very low–abundance proteins, such as HPV-derived proteins from cervical smears.^18–20

Previous proteomics studies have identified a number of biomarkers for early diagnosis or prognostication of cervical carcinoma and its precursors.²¹ Because high-risk HPV types are frequently detected in most of the cases of cervical carcinoma and precancerous lesions, these biomarkers for cervical carcinoma are possibly associated with the HPV infection. Nevertheless, protein profile changes common to cervical carcinoma and high-risk HPV infection have not been intensively studied.

In this study, the profile of differentially expressed proteins between HPV16-positive SCC and HPV-negative normal controls (NC) were investigated by iTRAQ-based proteomics followed by quantitative verification. The aim of this study was to further reveal the regulatory network of protein expression in cervical carcinoma that is associated with the HPV infection and identify auxiliary diagnostic markers for HPV-based cancer prognostication.

Materials and methods

Tissue specimens

The experimental protocol of this study was approved and monitored by the Ethics Committee of the Xinjiang Medical University, China. All procedures performed in this study were followed in accordance with the Helsinki Declaration of 1975, as revised in 2000.⁵ The data were analyzed for all patients and healthy individuals anonymously throughout the study. Written informed consent was obtained from all individual participants included in the study. Patients with cervical SCC, CIN, and controls with normal cervix (NC) were enrolled in the study, according to the diagnostic criteria of the World Health Organization and the Chinese Medical Association. The median age of the patients was 43 (range: 26–63) years. In total, 76 fresh cervical lesions and normal cervix tissue from ethnic Han and Uighur women were collected by routine biopsies and surgical operations at the Department of Gynecology, the First Affiliated Hospital of Xinjiang Medical University. These specimens were pathologically classified as 36 cases of SCC, 15 cases of CIN II–III, and 25 cases of NC. Tissue specimens were snap frozen in liquid nitrogen after resection and stored in a −80°C freezer. For immunohistochemical (IHC) analysis, 116 cases of paraffin-embedded cervical specimens were obtained from the specimen bank in the Pathology Department of the First Affiliated Hospital of Xinjiang Medical University, China. These specimens were selected after a case review by two experienced pathologists and classified as 47 cases of SCC, 37 cases of CIN II–III, and 32 cases of NC.

The clinical staging of the patients was based on the guideline established by the International Federation of Gynecology and Obstetrics (FIGO) of 1994–1997, as revised in 1999 by Pecorelli et al. Tumor specimens were collected from cancer patients who underwent radical surgery for cervical SCC at clinical stages I–IIa. Precancerous lesions were collected from patients with CIN II–III as biopsies or tissue specimens from cervical conization. The tissue specimens classified as NC were obtained from patients without a history of cervical lesions or any form of cancer who underwent a hysterectomy for nonmalignant reasons during the same time period. Indications for hysterectomy were fibroids, prolapsed uterus, or adenomyosis, primarily a combination of fibroids with prolapse.

DNA extraction and HPV detection

Genomic DNA was extracted from fresh cervical specimens using a QIAamp DNA Mini Kit for tissue DNA (51306; Qiagen, Valencia, CA, USA) and quantified using a spectrophotometer (NanoDrop 2000; Thermo Scientific, Wilmington, DE, USA). The HPV infection was analyzed by a multi-fluorescent quantitative polymerase chain reaction (qPCR) assay using a Real Quality RQ-HPV HR kit (RQ-22-120A; AB Analitica s.r.l., Padova, Italy) with four probes for HPV genotyping of HPV16, HPV31, 18/45, and HPV33/52/58/67, respectively, on the ABI-7500 PCR system (Applied Biosystems, Carlsbad, CA, USA). The probe (primers) for human β-globin gene was used as an internal control for DNA quality.

Protein extraction

For proteomic analysis with 8-plex iTRAQ labeling, two groups of protein samples were prepared to separately analyze tissue specimens from ethnic Uighur or Han women. Each group contained four cases of HPV16-positive tumor specimens and four cases of HPV-negative normal cervix, respectively. A frozen tissue specimen was homogenized by grinding and transferred into a 1.5-mL tube that contained pre-chilled tricarboxylic acid (TCA)–acetone (500 mL acetone, 50 g TCA). After centrifugation (12,000 RCF, 4°C, 10 min) and discarding the supernatant, the precipitate was dissolved in a protein lysis buffer (8 M urea, 30 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 1 mM phenylmethylsulfonyl fluoride (PMSF), 2 mM ethylenediaminetetraacetic acid (EDTA), and 10 mM dithiothreitol (DTT)). The protein sample (lysate) was then centrifuged and transferred into a new tube for removing the undissolved tissue residue. The DTT was added to the protein sample in a final concentration of 10 mM for reduction. After incubation for 45 min at 56°C, iodoacetamide buffer (55 mM IAA and 100 mM NH₄HCO₃) was added to the sample and incubated for 45 min for alkylation. The mixture was further mixed with pre-chilled 80% acetone at −20°C overnight and subsequently centrifuged for precipitation. The precipitate was dissolved in 300 µL dissociation buffer containing 50% triethylamine bicarbonate (TEAB) and 0.1% sodium dodecyl sulfate (SDS).

Tryptic digestion and iTRAQ reagent labeling

Each protein sample (100 µg) was digested with trypsin (protein:trypsin at 30:1) in a dissociation buffer (0.1% in TEAB; 37°C for 24 h). Eight peptide samples in each of the two groups (for ethnic Uighur or Han women) were labeled with 8-plex iTRAQ reagents (AB Sciex, Foster City, CA, USA) according to the manufacturer’s protocol (iTRAQ113, 114, 115, 116, 117, 118, 119, and 121). After removing the reaction solvent by speed vacuum, the labeled peptides were dissolved in 20 mM NH₄FA (pH 10) for further analysis.

Peptide fractionation by strong cation exchange chromatography and reversed-phase chromatography

High-resolution strong cation exchange chromatography was used to remove the redundant iTRAQ reagents and any interfering substances that might affect the mass spectrometry analysis. Labeled peptides were loaded onto an SCX column (Luna SCX, 4.6 mm × 250 mm; Phenomenex, Torrance, CA, USA) and eluted by a stepwise linear elution program as follows: 0–10 min equilibration in Buffer A (25% acetonitrile (ACN), 20 mM KCl, and 10 mM KH₂PO₄, pH 3.0), 10–15 min fast elution in 0%–5% Buffer B (25% ACN, 1 M KCl, and 10 mM KH₂PO₄, pH 3.0), 15–50 min linear elution with 5%–30% Buffer B, and 50–55 min washing elution with 30%–80% Buffer B. For desalting and further fractionation, the peptide fractions were loaded onto an reversed-phase (RP) column (Luna C18, 4.6 mm inner diameter × 250 mm length; Phenomenex) and eluted by a step linear elution program as follows: 0–10 min equilibration in a 100% solution A (2% ACN and 20 mM NH4FA, pH 10), 10–15 min fast elution in 0%–12% solution B (80% ACN and 20 mM NH₄FA, pH 10), 15–50 min linear elution with 12%–56% solution B, and 50–55 min washing elution with 56%–80% solution B. All of the procedures were performed using a Prominence HPLC system (Shimadzu, Kyoto, Japan) with a flow rate of 1.0 mL/min, and the peptides were monitored at 214 nm. The fractioned peptides were collected at a rate of one tube per minute during the linear elution period.

Peptide analysis by nano-liquid chromatography coupled with Q-Exactive mass spectrometry

The peptide fractions were loaded onto a nano RP column (5 µm Hypersil C18, 75 µm × 100 mm, Thermo Fisher Scientific Inc., Waltham, MA, USA) mounted in a Prominence nano HPLC system (Shimadzu). The peptides were eluted with an ACN gradient from 5% to 40% containing 0.1% formic acid for 65 min at 400 mL/min. The eluents were transferred to a Q-Exactive MS (Thermo Fisher Scientific Inc.), which was run in positive ion mode and a data-dependent manner with full MS scan from 350 to 6000 m/z, resolution at 70,000, tandem mass spectrometric (MS/MS) scan with a minimum signal threshold 17,500 and isolation at 2 Da. To evaluate the performance of the mass spectrometry on the iTRAQ-labelled samples, two MS/MS acquisition modes, higher collision energy dissociation (HCD) and collision-induced dissociation were employed. To optimize the MS/MS acquisition efficiency of HCD, the normalized collision energy was systemically examined from 25% to 70%.

Database search and quantitative data analysis

The raw MS/MS data were converted into Mascot Generic Format (MGF) format using Proteome Discoverer 1.3 (Thermo Fisher Scientific Inc.). The exported MGF files were searched by Mascot 2.3 (Matrix Science, Boston, MA, USA) against the Uniprot Human 2009–12 database with a precursor mass tolerance set at 15 ppm and production tolerance of 0.02 Da. An automatic decoy database search was performed. Carbamidomethylation of cysteines was set as a fixed modification (C) and oxidation of methionines (M), Gln to pyro-Glu (N-term Q), and 8-plex iTRAQ modifications of N-term, K, and Y were considered variable modifications. A maximum of one miscleavage was accepted. A protein with at least one unique peptide and a false discovery rate (FDR) < 0.01 qualified for further quantification analysis. The fold change in protein abundance was defined as the median ratio of all significantly matched spectra with tag signals. Based on analysis of Proteome Discoverer software, the coefficients of variation (CV) values of all quantified proteins and the quantitative results derived from duplicating injections were compared in parallel. The differential expression of all proteins was presented as the fold change in iTRAQ ratios. The upregulation of a protein was presented by the fold change of at least or more than 1.2 times and the downregulation by fold change of at least or less than 0.83 (1.0/1.2).

Bioinformatics analysis of MetaCore™ software

The differentially expressed proteins were further characterized by bioinformatic analysis using the software package MetaCore™ 6.18 (http://lsresearch.thomsonreuters.com/pages/solutions/1/metacore, Thomson Reuters) and its online database (https://portal.genego.com) to reveal the underlying pathways and protein–protein interaction networks and evaluate the candidate proteins as potential biomarkers.

RNA extraction and quantitative reverse transcription polymerase chain reaction analysis

The frozen tissue (~50 mg) was packed in aluminum foil and pulverized by grinding under liquid nitrogen. Total RNA was isolated from the powder by dissolving with TRIzol lysis buffer (Invitrogen, Carlsbad, CA, USA) followed by phenol/chloroform extraction and ethanol precipitation. The complementary DNA (cDNA) was synthesized from 1 µg of total messenger RNA (mRNA) by reverse transcription (RT) using a RevertAid Kit (MBI Fermentas, Burlington, ON, Canada) at 42°C for 60 min. For subsequent analysis, primer pairs specific to target mRNAs that were suitable for quantitative reverse transcription polymerase chain reaction (qRT-PCR) were designed and synthesized by the Takara Bio company (Takara Bio, Dalian, China). The cDNA (20 ng) was analyzed in a 25 µL mixture by qRT-PCR using an iQ5™ PCR system (Bio-Rad Laboratories Inc., Irvine, CA, USA) and SYBR Premix Ex Taq™ Kit (Ti RNase H Plus, Takara Bio, Tokyo, Japan). The qPCR reaction conditions were as follows: initial denaturation at 95°C for 30 s, followed by 40 cycles of denaturation at 95°C for 5 s and synthesis at 60°C for 30 s, and a final incubation at 4°C for 10 min. The expression level of target genes was quantified by the 2^−ΔΔCt method using the internal software of the PCR system, setting β-actin as an internal control.

IHC analysis

Paraffin-embedded tissues were sectioned into 3 µm slices. The tissue slices were conventionally stained by the streptavidin peroxidase–conjugated (S-P) method using primary antibodies (Abcam, Bristol, England; Thermo Fisher Scientific Inc.) recognizing the target proteins and IHC Kits containing biotin-labeled secondary antibodies (Zhongshan Golden Bridge Biotechnology Co., Peking, China) in accordance with the manufacturer’s recommended procedures. The negative controls were stained with phosphate-buffered saline (PBS) instead of primary antibodies. All tissue slices were counterstained with hematoxylin.

For evaluation, stained tissue sections were scored under a light microscope by two experienced pathologists. The positive-stained area and intensity were scored using the following criteria. Positively stained area: score 0 for negative signal, 1 for 0%–30% positive, 2 for 31%–60%, and 3 for more than 60%. Positive intensity: score 0 for negative signal, 1 for weak signal intensity, 2 for moderate intensity, and 3 for high intensity. A consensus number between the two investigators was reached for each tissue slice. An overall score (0–6) was calculated by the addition of both scores to evaluate the expression level of a protein in three categories: an overall score of 0–2 as the loss or weak expression, 3–4 as moderate expression, and 5–6 as a strong expression of the protein.

Statistical analysis

Statistical analysis was performed with SPSS 17.0 for Windows (SPSS Inc., Chicago, IL, USA). All p values were two-sided, and the significance level was p < 0.05. The data derived from the qRT-PCR and IHC analysis were compared by one-way analysis of variance tests or paired samples t-tests. The specificity and sensitivity of potential biomarkers were determined by receiver operating characteristic (ROC) curve analysis.

Results

Identification of the protein expression profile associated with cervical carcinoma and HPV16 infection by nano-LC coupled with Q-Exactive MS

Among the 76 cases with fresh tissue specimens, HPV infection was detected in 52 cases by real-time PCR using genotyping probes for HPV16, 31, 18/45, and 33/52/58/67 (Table 1). Of these, 28 cases of SCC, 11 cases of CIN II–III, and 5 cases of NC were positive for HPV16 infection, respectively. For proteomics analysis by 8-plex iTRAQ labeling, eight cases of tissue specimens were selected as a group, including four cases of HPV16-positive SCC and four cases of HPV-negative NC. To exclude etiological factors possibly associated with the genetic background, two different groups of protein samples were prepared to separately analyze tissue specimens from the ethnic Uighur and Han women, respectively. The protein profile was investigated by nano-LC coupled with Q-Exactive MS.

Table 1.

HPV genotyping of cervical specimens.

	n	HPV16	HPV16/18	HPV16/33/52/58/67
NC	25	5	n.d.	n.d.
CIN II–III	15	11	n.d.	n.d.
SCC	36	28	1	7

HPV: human papillomavirus; NC: normal controls; CIN: cervical intraepithelial neoplasia; SCC: squamous cell carcinoma; n.d.: not detected.

To compare the overall proteomic profile distinguishing SCC from NC, we integrated peptide spectra that corresponded to four protein samples from SCC and NC in each of the two groups (for ethnic Uighur and Han women), respectively, using a functional module of the software Proteome Discoverer (version 1.3). Because the tissue specimens from SCC were almost identical in histopathologies, such as clinical stage (before stage IIa), differential state (low or moderate differentiation), and HPV infection, it was reasonable to consider either tumor specimens or NC in each group as biological repeats (data not shown). Subsequent analysis identified 67 differentially expressed proteins between SCC and NC, which was identical to the proteomics data for ethnic Uighur and Han women, setting a fold change of 1.2 times as the cut-off value for the differential expression. Of these, 33 proteins were upregulated (fold change > 1.2) and 34 proteins downregulated (fold change < 0.83 or 1.0/1.2) in SCC compared with NC, respectively (Table 2).

Table 2.

Differentially expressed proteins in HPV-positive SCC compared with HPV-negative normal controls.

No.	Protein information				Peptide detection			Fold change^a
No.	Rec. protein name	Uniprot ID	Rec. symbol	MW	Peptide score	Peptide coverage	Unique peptides	Han^b	Uighur^c
1	Acid ceramidase	Q13510	ASAH1	44.6	52.42	3.04	1	2.523	1.296
2	Minichromosome maintenance complex component 5	P33992	MCM5	82.2	20.98	1.77	1	2.382	1.443
3	Cytochrome c	P99999	CYC	11.7	138.7	43.81	4	2.093	1.473
4	Transmembrane 9 superfamily member 4	Q92544	TM9S4	74.5	45.97	2.96	1	2.008	1.372
5	Acidic leucine-rich nuclear phosphoprotein 32 A	P39687	AN32A	28.6	102.7	19.68	1	2.000	1.483
6	Serine/arginine-rich splicing factor 7	Q16629	SRSF7	27.4	59.54	13.87	3	1.725	1.289
7	RNA-binding protein 8A	Q9Y5S9	RBM8A	19.9	39.44	6.32	1	1.611	1.202
8	Serine/arginine-rich splicing factor 10	O75494	SRS10	31.3	39.75	6.11	1	1.609	1.535
9	Poly(rC)-binding protein 2	Q15366	PCBP2	38.6	151.52	18.63	3	1.608	1.274
10	Osteoclast-stimulating factor 1	Q92882	OSTF1	23.8	60.92	14.49	2	1.594	2.332
11	Transgelin-2	P37802	TAGLN2/TAGL2	22.4	362.32	70.35	9	1.546	1.371
12	Proliferating cell nuclear antigen	P12004	PCNA	28.8	239.24	35.25	6	1.508	1.239
13	Minichromosome maintenance complex component 2	P49736	MCM2	101.8	85.23	6.53	5	1.497	1.274
14	60 kDa heat shock protein, mitochondrial	P10809	CH60	62	723.15	48.69	21	1.466	1.26
15	Heterogeneous nuclear ribonucleoproteins C1/C2	P07910	HNRPC	33.6	353.5	27.12	10	1.464	1.272
16	Cytochrome b-245 heavy chain	P04839	CY24B	65.3	77.69	4.39	2	1.429	1.53
17	Heterogeneous nuclear ribonucleoprotein A1	P09651	hnRNPA1	38.7	319.73	31.99	8	1.405	1.21
18	Alpha-enolase	P06733	ENO1	47.1	837.96	64.06	16	1.386	1.288
19	Nucleolin	P19338	NUCL	76.6	556.38	27.75	18	1.382	1.273
20	Thymidine phosphorylase	P19971	TYPH	49.9	700.18	49.38	17	1.349	1.396
21	Profilin-1	P07737	PROF1	15	276.21	52.86	6	1.315	1.294
22	V-type proton ATPase subunit B, brain isoform	P21281	VATB2	56.5	149.63	13.11	4	1.301	1.305
23	Probable ATP-dependent RNA helicase DDX5	P17844	DDX5	69.1	344.55	19.54	4	1.270	1.805
24	Acidic leucine-rich nuclear phosphoprotein 32 B	Q92688	AN32B	28.8	120.58	23.51	3	1.267	1.715
25	Retinoid-inducible serine carboxypeptidase	Q9HB40	RISC	50.8	25.73	1.99	1	1.259	1.466
26	AP-3 complex subunit delta-1	O14617	AP3D1	130.1	59.67	1.91	1	1.250	1.246
27	Small ubiquitin-related modifier 2	P61956	SUMO2	10.9	80.25	18.95	2	1.249	1.405
28	Malate dehydrogenase, mitochondrial	P40926	MDHM	35.5	616.42	52.66	13	1.246	1.325
29	Signal recognition particle 9 kDa protein	P49458	SRP09	10.1	136.15	25.58	2	1.242	1.304
30	Voltage-dependent anion-selective channel protein 1	P21796	VDAC1	30.8	365.31	46.29	9	1.228	1.344
31	Minichromosome maintenance complex component 4	P33991	MCM4	96.5	60.81	3.48	2	1.220	1.794
32	Stress-70 protein, mitochondrial	P38646	GRP75	73.6	560.83	29.16	15	1.220	1.257
33	Glutathione S-transferase kappa 1	Q9Y2Q3	GSTK1	25.5	149.52	27.43	5	1.218	1.320
34	Prolargin	P51888	PRELP	43.8	353.42	32.98	10	0.752	0.203
35	Lumican	P51884	LUM	38.4	391.75	37.57	11	0.728	0.173
36	Putative beta-actin-like protein 3	Q9BYX7	ACTBM	42	228.24	16.8	1	0.712	0.239
37	Collagen type V alpha 1 chain	P20908	COL5A1	183.4	75.72	1.96	3	0.655	0.105
38	Prostaglandin-H2 d-isomerase	P41222	PTGDS	21	63.91	8.95	1	0.654	0.203
39	Extracellular superoxide dismutase	P08294	SODE	25.8	65.1	16.67	3	0.649	0.162
40	Protein-glutamine gamma-glutamyltransferase 2	P21980	TGM2	77.3	458.7	23.29	13	0.571	0.509
41	Decorin	P07585	PGS2/ DCN	39.7	391.15	35.65	11	0.561	0.145
41	Fibrillin-1	P35555	FBN1	312	1023.85	16.86	32	0.550	0.235
43	Fibulin-1	P23142	FBLN1	77.2	325.65	22.48	9	0.529	0.229
44	Olfactomedin-like protein 3	Q9NRN5	OLFL3	46	114.09	18.97	5	0.517	0.256
45	Tenascin-X	P22105	TENX	464	228.94	5.53	10	0.510	0.269
46	KN motif and ankyrin repeat domain–containing protein 2	Q63ZY3	KANK2	91.1	48.98	1.53	1	0.501	0.293
47	Angiotensinogen	P01019	ANGT	53.1	84.3	8.66	3	0.495	0.408
48	Collagen alpha-2(VI)	P12110	CO6A2	108.5	702.44	22.57	17	0.487	0.279
49	Alpha-1-antitrypsin	P01009	A1AT	46.7	445.26	45.69	15	0.480	0.270
50	Protein AMBP	P02760	AMBP	39	143.03	21.31	5	0.479	0.446
51	Collagen alpha-3(VI)	P12111	CO6A3	343.5	2891.26	32.89	73	0.479	0.243
52	Collagen type VI alpha 1 chain	P12109	COL6A1	108.5	963.74	38.23	25	0.478	0.299
53	Apolipoprotein A-II	P02652	APOA2	11.2	137.32	60	4	0.468	0.140
54	Clusterin	P10909	CLUS	52.5	80.81	12.47	4	0.465	0.404
55	Cornifin-A	P35321	SPR1A	9.9	73.67	34.83	1	0.440	0.377
56	Dermatopontin	Q07507	DERM	24	151.01	21.89	4	0.436	0.240
57	Collagen type V alpha 2 chain	P05997	COL5A2	144.8	53.29	2.8	2	0.422	0.168
58	Collagen alpha-1(I) chain	P02452	CO1A1	138.9	781.82	17.28	17	0.420	0.122
59	Collagen type I alpha 2 chain	P08123	COL1A2	129.2	525.09	13.18	14	0.416	0.120
60	Collagen type XIV alpha 1 chain	Q05707	COL14A1	193.4	1843.61	39.2	50	0.414	0.164
61	Alpha-1-antichymotrypsin	P01011	AACT	47.6	128.76	13.48	4	0.409	0.413
62	Collagen type XXI alpha 1 chain	Q96P44	COL21A1	99.3	22.95	1.15	1	0.399	0.154
63	Creatine kinase B-type	P12277	KCRB	42.6	215.99	24.67	6	0.362	0.405
64	Caveolin-1	Q03135	CAV1	20.5	63.51	22.47	3	0.350	0.164
65	Keratin, type II cytoskeletal 6B	P04259	K2C6B	60	1345.58	51.24	1	0.345	0.149
66	Collagen type III alpha 1 chain	P02461	COL3A1	138.5	162.93	4.3	4	0.330	0.126
67	Cornulin	Q9UBG3	CRNN	53.5	305.51	27.88	8	0.310	0.117

HPV: human papillomavirus; SCC: squamous cell carcinoma; MW: molecular weight.

For a candidate protein, the fold change (iTRAQ ratio) of at least or more than 1.20 and at least or less than 0.83 (1/1.20) in cervical SCC compared with controls was considered and reported as up- and downregulation, respectively.

b,c

Samples from ethnic Han and Uighur women, respectively.

Bioinformatics analysis of the identified proteins and the prediction of potential biomarkers

The role of the identified proteins in disease development or carcinogenesis was further evaluated by bioinformatics analysis using MetaCore™ software (version 6.16) and its ontology database (http://www.genego.com). The gene ontology analysis showed that most of these proteins are localized in extracellular parts, intracellular organelles, or membrane-enclosed lumen (Figure 1) and involved in diverse cellular functions, such as glycolysis, gluconeogenesis, protein folding and maturation, or mitochondrial functions (Figure 2), and different biological processes, such as responses to injury repair, acute inflammatory reactions, or blood coagulation (Figure 3). The biomarker assessment analysis using the disease ontology database discovered that 16 of these proteins had been previously identified as potential biomarkers for cervical carcinoma. In accordance with our proteomics data, 11 of these proteins (Q03135, P07585, P08123, P20908, P02461, Q05707, Q9UBG3, P35555, P51884, P22105, and P21980) were downregulated and 5 proteins (P17844, Q9Y2Q3, P49736, P19338, and P12004) were upregulated in cervical carcinoma compared with NC. In addition, three proteins (Q03135, P19971, and P10909) were associated with SCC.

Figure 1.

The network profiling of cellular signaling and gene expression for 67 candidate proteins by MetaCore™ software. The network includes all candidate proteins identified by proteomic analysis of tissue specimens from HPV-positive SCC and HPV-negative normal controls of the uterine cervix, as described in Table 2.

Figure 2.

The distribution analysis of signaling pathways associated with 67 candidate proteins by MetaCore™ software. All candidate proteins described in Table 2 were included in the analysis.

Figure 3.

Major biological processes associated with 67 candidate proteins as displayed by MetaCore™ software analysis. The description of all candidate proteins was shown in Table 2.

Verification of the profile of protein expression by quantitative RT-PCR and IHC

To verify the result of the proteomics analysis and identify novel biomarkers for cervical carcinoma, we selected 10 candidate proteins that were upregulated in the SCC: ASAH1, MCM5, CYC, MCM4, DDX5, PCBP2, TAGLN2, hnRNPA1, ENO1, and TYPH (see Table 2). Bioinformatics analysis and literature information indicated that the selected proteins might be associated with various cancers but had not been intensively studied in cervical carcinoma and its precursor lesions so far, with the exception of DDX5. The mRNA expression of these proteins was determined in 76 cases of cervical lesions and NC by qRT-PCR using primer pairs specific to mRNAs coding for the target proteins (Table 3). The results showed a significantly increased transcription of these genes in SCC compared with NC (p < 0.05), which was consistent with the proteomics data. In addition, significant upregulation was also found for PCBP2, MCM5, hnRNPA1, TYPH, CYC, and MCM4 in SCC as compared with CIN II–III, and for ASAH1, DDX5, and ENO1 in CIN II–III compared with NC (p < 0.05) (Table 4, Figure 4).

Table 3.

List of primers used in quantitative RT-PCR analysis.

Gene	Access	Forward primer (5′–3′)	Reverse primer (5′–3′)	Product (bp)
ASAH1	NM_177924.3	GAAACCATGTATGATGTCCTGTCAA	GGTCACATGGAGATGGTGTCTTC	196
PCBP2	NM_005016.5	AGGTTACCATCACTGGATCTGCTG	GGATGAATCTGCATTGTTCTAGCTG	116
DDX5	NM_004396.3	AAGAGCACCCTGATTTGGCTA	TGGGTTCAGTGAAATTCTGTCTTG	174
MCM4	NM_005914.3	ACAGGTGCTCTTGTCCTGAGTG	CTCAATGGGATTTGCTGCTG	186
MCM5	NM_006739.3	CTATGCCCTGCCCAGGAAGT	AAGTCCACGCATTTGCATTTGT	102
hnRNPA1	NM_002136.2	CGAAGGTAGGCTGGCAGATA	CAGGCTCTCATCAGTTGTTTCA	187
ENO1	NM_001201483.1	CTCTGAAGTCATCCTGCCAGTCC	CAGCTTGTTGCCAGCATGAGA	70
TYPH	NM_001113755.2	GGCTGCTGTATCGTGGGTCA	GAACTTAACGTCCACCACCAGAG	168
CYC	NM_018947.5	ACACCTGACCAGAAACTTTGTCTCC	GCCAAAGCAGCAGCTCAGTATGTA	79
TAGLN2	NM_001277224.1	TTCCAGAACTGGCTCAAGGATG	CCAGAGGTCCACAGTTTGGAAG	198
β-actin	NM_001101.3	CATCCGTAAAGACCTCTATGCCAAC	ATGGAGCCACCGATCCACA	171

RT-PCR: reverse transcription polymerase chain reaction.

Table 4.

Quantitative RT-PCR analysis of 10 proteins in cervical lesions.

	mRNA expression level ( $\bar{X}$ ±s)			ANOVA test (p value)
	SCC (n = 36)	CIN (n = 15)	NC (n = 25)	F	p	SCC vs NC	CIN vs NC	SCC vs CIN
ASAH1	0.388 ± 0.376	0.430 ± 0.380	0.048 ± 0.046	7.802	0.001	0.003	0.001	0.718
PCBP2	17.273 ± 16.852	6.448 ± 4.188	3.491 ± 3.375	7.843	0.001	4.424E−4	0.514	0.015
DDX5	1.251 ± 1.214	1.536 ± 1.482	0.133 ± 0.131	8.043	0.001	0.003	0.001	0.487
MCM5	0.504 ± 0.496	0.139 ± 0.138	0.026 ± 0.024	7.979	0.001	0.001	0.444	0.007
TAGLN2	1.488 ± 1.469	0.960 ± 0.340	0.119 ± 0.116	6.68	0.003	0.001	0.058	0.156
hnRNPA1	0.053 ± 0.052	0.021 ± 0.020	0.016 ± 0.016	4.082	0.025	0.015	0.741	0.036
ENO1	0.788 ± 0.761	0.536 ± 0.354	0.074 ± 0.072	9.289	3.797E−4	8.527E−5	0.018	0.176
TYPH	0.427 ± 0.423	0.192 ± 0.191	0.030 ± 0.028	9.684	2.857E−4	6.436E−5	0.146	0.033
CYC	0.778 ± 0.776	0.170 ± 0.114	0.181 ± 0.180	6.519	0.003	0.005	0.962	0.005
MCM4	0.508 ± 0.507	0.124 ± 0.120	0.025 ± 0.024	6.392	0.006	0.003	0.521	0.013

HPV: human papillomavirus; NC: normal controls; CIN: cervical intraepithelial neoplasia; SCC: squamous cell carcinoma; ANOVA: analysis of variance; RT-PCR: reverse transcription polymerase chain reaction; mRNA: messenger RNA.

SCC, CIN, or NC refers to patients with cervical squamous cell carcinoma, cervical intraepithelial neoplasia, or normal cervix, respectively; SCC/CIN, SCC/NC, CIN/NC, comparative analysis of two groups indicated, with statistical significance at p < 0.05.

Figure 4.

The transcription levels of 10 candidate genes in cervical lesions. (a–j) transcription levels of ASAH1, PCBP2, DDX5, MCM, TAGLN2, hnRNPA1, EON1, TYPH, CYC, and MCM4 in NC, CIN II-III, and SCC, as shown in Table 4, were displayed in bar graphs; *statistical significance by one-way ANOVA test at p < 0.05. The analysis included 10 candidate proteins identified by proteomics as described in Table 2.

We analyzed the expression of five candidate proteins, ASAH1, PCBP2, DDX5, hnRNPA1, and TAGLN2, by IHC to further verify the data of proteomics and qRT-PCR. The strong or moderate staining of PCBP2, hnRNPA1, and DDX5 was localized in the nucleus and ASAH1 in the cytoplasm of epithelial cells of cervical lesions, whereas TAGLN2 was exclusively found in vascular endothelial cells in the mesenchyme (Figure 5). The expression of PCBP2, hnRNPA1, DDX5, and ASAH1 was gradually increased in cervical specimens from NC to CIN II–III and SCC with significant differences (Table 5 and Figure 6) (p < 0.05). In the case of TAGLN2, no reliable statistical result was available because this protein was expressed in vascular endothelial cells instead of cervical epithelial cells in CIN II–III and controls and thus not comparable with SCC. Based on the data in Table 5, we evaluated the diagnostic potential of these proteins for cervical carcinoma by ROC curve analysis (Figure 7). The detection of ASAH1 in SCC, CIN II–III, and NC had a relatively high specificity (75.4%) but low sensitivity (57.4%), while PCBP2, DDX5, and hnRNPA1 were detected with relatively high sensitivity (83%–93.6%) and low specificity (42%–63.8%) (Table 6).

Figure 5.

IHC staining pattern of five candidate proteins in cervical lesions. (a–c) typical expression patterns of ASAH1, PCPB2, DDX5, hnRNPA1, and TAGLN2 as weak expression in normal controls (a1–a5), moderate expression in CIN II–III (b1–b5), and strong expression in SCC (c1–c5), respectively. The description of candidate proteins was shown in Table 2.

Table 5.

Immunohistochemical detection of four proteins in cervical specimens.

	Relative expression level ( $\bar{X}$ ±s)			ANOVA test (p value)
	SCC (n = 47)	CIN (n = 37)	NC (n = 32)	F	p	SCC vs NC	CIN vs NC	SCC vs CIN
ASAH1	4.617 ± 0.709	4.243 ± 0.925	3.094 ± 0.734	36.527	5.81E−13	1.44E−13	2.19E−8	0.034
PCBP2	5.040 ± 0.624	4.700 ± 0.661	3.530 ± 0.879	44.476	5.66E−15	1.80E−15	5.32E−10	0.032
DDX5	5.020 ± 0.608	4.680 ± 0.709	3.810 ± 0.998	24.275	1.70E−9	3.13E−10	8.02E−6	0.042
hnRNPA1	5.213 ± 0.550	4.892 ± 0.774	4.219 ± 0.870	18.203	1.40E−7	2.35E−8	1.88E−4	0.046

HPV: human papillomavirus; NC: normal controls; CIN: cervical intraepithelial neoplasia; SCC: squamous cell carcinoma; ANOVA: analysis of variance.

SCC, CIN, or NC refers to patients with cervical squamous cell carcinoma, cervical intraepithelial neoplasia, or normal cervix, respectively; SCC/CIN, SCC/NC, CIN/NC, comparative analysis between two groups, with statistical significance at p < 0.05.

Figure 6.

IHC-based analysis of protein expression for four candidate genes in cervical lesions. Protein expression levels of ASAH1, PCPB2, DDX5, and hnRNPA1 in NC, CIN II-III, and SCC, as shown in Table 5, were displayed in bar graphs (a–d); *statistical significance by one-way ANOVA test at p < 0.05. The analysis included four candidate proteins described in Table 2.

Figure 7.

IHC-based ROC curves for four candidate proteins. The ROC curve analysis for ASAH1, PCPB2, DDX5, and hnRNPA1 (a–d), respectively, in the detection of NC, CIN II-III, and SCC by immunohistochemistry.

Table 6.

The sensitivity and specificity of four candidate proteins in the detection of cervical carcinoma.

	Sensitivity (%)	Specificity (%)	AUC (area under the curve)
ASAH1	57.4	75.4	0.751
PCBP2	83.0	63.8	0.764
DDX5	83.0	53.6	0.723
hnRNPA1	93.6	42.0	0.707

Based on the analysis of HPV infection by detection of eight high-risk HPV types, all of the cases of SCC and most of the cases of CIN II–III were positive for HPV infection (100% and 73%), and most of the NC were HPV negative (80%) (see Table 1). Accordingly, we suggested that a possible correlation of protein expression with HPV infection may already be integrated into the results of the qRT-PCR and proteomics analysis. However, independent analyses are required to verify these results in the future.

Discussion

Proteomics strategies are frequently used to profile the protein expression of cervical carcinoma and its precursor lesions to identify tissue-based tumor biomarkers. Early studies showed the differential protein pattern between cervical SCC specimens and controls from the uterine cervix using 2DE technology and identified a number of proteins as potential biomarkers for SCC, such as HSP70, HSP27 (HSPB1), transgelin-2, eIF3B, NCF-2, ANXA6, TYK2, S100A9, and ZNF-217.^22–25 The development of 2D-DIGE has improved the profiling sensitivity and accuracy for low-abundance proteins in the proteomics research of cancer. Recent studies based on 2D-DIGE have reported another set of proteins as possible tissue-based biomarkers for cervical SCC and precancerous lesions, such as S100A9, HSPB1, eEF1A1, FABP5, MnSOD, and cornulin.^26,27 With the improved detection capacity for large numbers of low-abundance proteins by iTRAQ-based or shotgun proteomics, proteomics studies have revealed the complexity of the regulation network of protein expression during cervical carcinogenesis and identified novel biomarkers for SCC, including HSPB, G6PD, ALDH3A1, STAT1, KRT7, and KR17.^28,29 In addition, redox proteomics identified the differential expression of ERp57, GST, TrxR2, and iNOS as a biomarker profile for HPV-positive invasive cervical carcinoma.³⁰ However, HPV genotyping was often not performed or considered in previous proteomics studies of cervical carcinoma. In our study, the protein expression pattern of HPV16-positive tumor specimens was compared with that of HPV-negative NC from the uterine cervix by 8-plex iTRAQ proteomics. We identified a profile of 67 differentially expressed proteins that were identical among two independent analyses of protein samples from the ethnic Uighur and Han women. Of these, 33 proteins were upregulated and 34 downregulated in SCC compared with controls. Subsequently, the profile was evaluated by bioinformatics analysis and verified by detection of 10 candidate proteins using quantitative RT-PCR and IHC.

As a result of bioinformatics evaluation, most of the identified proteins are functionally involved in glycolysis, gluconeogenesis, protein folding and maturation, mitochondrial functions, the cell cycle, or apoptosis. The biomarker assessment analysis predicted 16 proteins that are known biomarkers for cervical carcinoma or its precursors. Based on the proteomics data of this study, 5 of these proteins (GSTK1, MCM2, NUCL (NUNCL), DDX5, and PCNA) were upregulated and 11 proteins (PGS2, CAV1, CRNN, FBN1, COL1A2, COL5A1, COL3A1, COL14A1, LUM, TENX, and TGM2) downregulated in SCC compared with NC.

The upregulation of four predicted biomarkers, such as GSTK1, MCM2, NUCL, and PCNA, in cervical lesions or cells is reported by other studies. The expression of GSTK1 is downregulated in KB-8-5-11 cervical carcinoma cells in response to cisplatin treatment.³¹ MCM2 is overexpressed in CIN and may serve as a predictive marker for cervical carcinoma.³² NUCL may bind to the promoter of HPV18 in the S-phase and enhance the expression of viral-coding oncogenes in HPV18-positive cervical carcinoma cells.³³ PCNA is a coordinator of essential cellular functions for cell growth and maintenance. PCNA expression is upregulated in CIN and associated with high-risk HPV infections and has a predictive value for cervical carcinoma.^34,35

The downregulation of 11 predicted biomarkers by bioinformatics was partially described by previous studies. PGS2/DCN is a newly identified soluble and matrix-derived tumor repressor that antagonizes signaling of various receptor tyrosine kinases (RTKs).³⁶ The overexpression of PGS2 may induce cell cycle arrest of SiHa cervical carcinoma cells.³⁷ CAV1 is an integral membrane protein of lipid rafts and a putative tumor suppressor gene that is downregulated in cervical lesions and cells transformed by HPV infection.^38,39 CRNN is a marker of late epidermal differentiation and is downregulated in cervical lesions.^25,30 FBN/LTPB is a calcium-binding protein of the extracellular matrix, and overexpression of FBN2 but not FBN1 has been reported in cervical adenocarcinoma.⁴⁰ LUM is a small leucine-rich proteoglycan expressed in the periphery of cancer nests of cervical SCC.⁴¹ LUM expression is downregulated in prostate cancer, and its stromal expression is associated with reduced metastasis and prolonged survival of patients with pancreatic cancer.^42,43 Contrary to the results of this study, the expression of TGM2, a multifunctional enzyme with transglutaminase crosslinking, G protein signaling, and kinase activities, is upregulated in cervical SCC and is associated with invasiveness.⁴⁴ Among the four tenascins (C, X, R, and W), glycoproteins that are located in the extracellular matrix, tenascin-C is known as an “oncofetal” protein controlled by many stimuli relevant to embryonal and fetal development, whereas tenascin-X (TENX) expression is barely affected by external factors.⁴⁵ Tenascin-C expression is markedly increased in invasive cervical carcinoma and oral SCC, but the role of TENX in tumorigenesis is largely unknown.^46,47 Collagens are triple-helix proteins formed by the association of three polypeptide chains that are coded by different genes of the same collagen type.⁴⁸ Vertebrates have 28 different types of collagens that provide tissue-specific structure and function by assembling into a variety of supramolecular structures of the extracellular matrix.⁴⁹ Thus, the downregulation of several collagens, as shown in this study, indicates the dysfunction of extracellular matrices during tumor invasion and metastasis. COL5A2 is downregulated in cervical SCC and adenocarcinoma.⁵⁰ The coding genes for COL1A2, COL14A1, and COL15A1 are aberrantly methylated at the promoter region in a number of cancers, but not in cervical carcinoma.^51–53 The COL3A1 expression is suppressed in nasopharyngeal carcinoma and upregulated by inhibition of miR-29 a/b. Its upregulation in renal cell carcinoma by ectopic expression of Let-7d micro RNA (miRNA) results in enhanced cell proliferation and migration.^54,55 Based on the biomarker predictions from bioinformatics and previous findings, the differential expression of NUNCL, PCNA, and CAV1 is mostly associated with high-risk HPV infection in cervical SCC. Future studies are required to validate these markers. In particular, future studies need to investigate the association between downregulation of multiple proteins and promoter methylation of their coding genes to study possible novel mechanisms of HPV-induced carcinogenesis.

To verify the accuracy and specificity of our proteomics data, we determined the expression of 10 candidate proteins, namely, ASAH1, MCM5, CYC, MCM4, DDX5, PCBP2, TAGLN2, hnRNPA1, ENO1, and TYPH at both the transcriptional and protein expression levels. The transcripts coding for all of these proteins were significantly upregulated in cervical SCC as shown by qRT-PCR analysis. This was further confirmed by IHC analysis of four out of the five proteins, ASAH1, DDX5, PCBP2, and hnRNPA1.

TAGLN2 (TAGL2) is an actin-binding protein involved in cell morphology and transformation.⁵⁶ Concordant with our data from quantitative RT-PCR analysis, TAGLN2 expression is increased in cervical SCC and gastric adenocarcinoma but is downregulated in lung and breast carcinoma.^22,57–59 Suppression of TAGLN2 inhibits proliferation and migration of cervical carcinoma cells.⁶⁰ Intriguingly, we found this protein to be located exclusively in the vascular endothelium by IHC and therefore could not verify the data of proteomics and RT-PCR analysis.

DDX5/p68 is an ATP-dependent RNA helicase and co-activator of transcription factors and thereby plays a role in RNA metabolism.⁶¹ p68 is upregulated in SiHa cervical carcinoma cells in response to treatment with calcitriol (vitamin D3).⁶² p68 overexpression may enhance β-catenin and androgen receptor signaling to promote cell proliferation in prostate and non-small-cell lung cancer cells.^63,64 p68 acts as a co-activator of p53 and upregulates the expression of the oncogenic protein kinase, polo-like kinase 1 (PLK1) in the absence of functional p53, which in turn promotes mitosis in various cancers.^65,66 However, not much is known about the regulation of DDX5 in cervical SCC.

ASAH1/AC is a lipid hydrolase that converts ceramides into sphingosine (SPH) and free fatty acids in different tissues. ASAH1 regulates the dynamic homeostasis of ceramides with sphingosine-1-phosphate (S1P), an anti-apoptotic signal that regulates cell maintenance and proliferation.⁶⁷ Accordingly, the elevation of ASAH1 results in an imbalance of this homeostasis between ceramide and S1P, which in turn promotes tumorigenesis.⁶⁸ ASAH1 is upregulated in epithelial ovarian cancer and breast cancer and is associated with a relatively good prognosis.^69,70

Poly-cytosine-binding proteins (PCPBs) are RNA-binding proteins that regulate gene expression at multiple levels.⁷¹ PCPBs have two main forms, the PCBP1 and PCBP2, with 82% homology in amino acid sequence.⁷² PCBP1 is downregulated in various tumors;⁷³ in contrast, PCBP2 is upregulated in leukemia and glioma.^74,75 In glioma and gastric carcinoma, PCBP2 interacts with tumor suppressors, such as FHL3, SIRT6, and miR-34a, and thus, its elevation supports tumor cell growth.^76,77

HNRPA1 is a member of a family of RNA-binding proteins that are involved in a variety of functions, such as the regulation of transcription, mRNA metabolism, and translation.⁷⁸ hnRNPA1 may activate epithelial-to-mesenchymal transitions by interaction with the serine–arginine-rich splicing factor 1 (SRSF1) that regulates oncogenic alternative splicing.⁷⁹ Concordant with our results, increased hnRNPA1 expression is detected in cervical lesions, with the highest level in SCC cases that are positive for HPV infection.⁸⁰ In cervical epithelial cells immortalized with HPV16 E6 and E7, the hnRNPA1 expression is upregulated in response to hypoxia and glucose deprivation, typical features of the tumor microenvironment.⁸¹

MCM proteins play important roles in cell cycle progression in mediating DNA replication initiation and elongation only once per cell cycle. MCM 4, 5, and 6 are differentially expressed in different types of cervical lesions, while MCM 4, 5, 6, and 10 are upregulated in SCC. In SCC, the overexpression of MCM 4, 6, and 10 is correlated with the advancement of tumor stages.³² MCM detection in cervical smears has been suggested as an alternative to cytology screening of cervical lesions.⁸²

CYC is essential in mitochondrial electron transport and intrinsic type II apoptosis. This kind of apoptotic event involves changes in CYC phosphorylation, increased reactive oxygen species via increased mitochondrial membrane potentials, and oxidation of cardiolipin followed by CYC release from the mitochondria.⁸³ CYC expression is downregulated in response to radiation or treatment of cervical carcinoma cells with anti-cancer drugs.^84,85 CYC expression is induced by ectopic expression of HPV16 E7 in HPV-negative cervical carcinoma cells, indicating that its expression is probably associated with HPV16 infection.⁸⁶ However, the upregulation of CYC in cervical lesions and SCC has not been reported previously.

ENO1 (enolase alpha) is a key glycolytic enzyme in the cytoplasm and on the surface of several types of cells, where it concentrates proteolytic activity by acting as a plasminogen receptor.⁸⁷ The expression of ENO1 is upregulated in gynecological cancers of the uterine cervix, ovary, and endometrium, and highest expression is frequently associated with brain metastases.^88,89

TYPH (TP, ECGF1) is involved in thymidine metabolism and homeostasis by catalyzing thymidine dephosphorylation with broad substrate specificity.⁹⁰ TYPH is elevated in various tumors and plays a role in tumor progression and angiogenesis.⁹¹ TYPH is highly expressed in microvessels of CIN and SCC and appears to be involved in angiogenesis in CIN.⁹² TYPH activity is increased in invasive cervical SCC, with the highest level in lymph node metastases.⁹³

Based on these analyses, we demonstrated that the data of iTRAQ-based proteomics were not only supported by the qRT-PCR and IHC verification, and bioinformatic evaluation, but also highly consistent with earlier findings. However, due to the high HPV positivity of cervical lesions, namely, almost all specimens of SCC and most of the CIN II-III are positive for high-risk HPV genotypes, the analysis of gene expression for candidate proteins associated with HPV infection was very limited. Nevertheless, the proteomic identification of candidate proteins based on HPV genotyping was, to a very high extent, informative for understanding the mechanisms of HPV-driven cervical carcinogenesis.

As an alternative to HPV genotyping, the detection of viral oncogenic E6/E7 transcripts is considered as a potential marker of productive viral infection.⁹⁴ The high HPV E6/E7 mRNA expression is related to the severity of cervical lesions and a poor prognosis of cervical carcinoma, in contrast to the HPV DNA positivity that is not always associated with the histological grade of cervical lesions.^95,96 Although both HPV DNA and E6/E7 mRNA can be detected at equally high rates in cervical carcinoma, the E6/E7 mRNA positivity is lower than HPV DNA in intraepithelial lesions.^97,98 In addition, HPV E6/E7 mRNA assay may have higher specificity in the analysis of ThinPrep Pap Test samples, but lower sensitivity than HPV genotyping.^99,100 These studies suggest that not all viral DNA copies are transcriptionally active in cervical lesions, and the E6/E7 mRNA assay may discriminate potentially progressive cervical lesions from transient dysplasias and provide useful information for early cervical cancer screenings.

Another limitation of this study was the analysis of HPV infection by only eight oncogenic HPV types (HPV16, 18, 45, 31, 33, 52, 58, and 67) using a method for clinical diagnosis. In accordance with the definition of International Agency for Research of Cancer (IARC) in 2009, HPV16, 18, 31, 33, 35, 45, 52, and 58 are carcinogenic to humans and most frequently found in cervical cancer, among which HPV16 is the most potent HPV type to cause human cancers, and HPV39, 51, 56, 59, and 68 are also carcinogenic and less constantly found in cervical cancer, while HPV26, 53, 66, 67, 70, 73, and 82 are considered “possibly carcinogenic” to humans in case of cervical cancer.¹⁰¹ Therefore, the application of novel methods is required to detect other HPV genotypes and validate the result of this study in the future.

Conclusion

In this study, we show a profile of 67 proteins that are differentially expressed in HPV16-positive cervical SCC compared with HPV-negative NC. Since our proteomics data were confirmed by quantitative verification of nine candidate proteins and were mostly supported by previous findings of other studies, we suggest that the profiles, in particular, the upregulation of ASAH1, PCBP2, DDX5, and hnRNPA1, and possibly MCM4, MCM5, CYC, ENO1, and TYPH, are potential biomarkers for cervical SCC and high-risk HPV infection. However, the role of most of these proteins in HPV-induced carcinogenesis is largely unknown. Further validation of the profile, and the association between the altered expression of these proteins and HPV infection by other approaches, will contribute to establishing auxiliary diagnostic markers for HPV-based cancer prognosis.

Footnotes

Acknowledgements

S.Q. and W.T. contributed equally to this work.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

Funding

This study was supported by Natural Science Foundation of China (81360321). The funders had no role in the study design, data collection, analysis, decision to publish, or preparation of the manuscript.

Informed consent

Written informed consent was obtained from all individual participants included in the study.

References

Torre

Bray

Siegel

. Global cancer statistics, 2012. CA Cancer J Clin 2015; 65(2): 87–108.

Ferlay

Shin

Bray

. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010; 127(12): 2893–2917.

Silverberg

Ioffe

. Pathology of cervical cancer. Cancer J 2003; 9(5): 335–347.

Denny

. Cervical cancer: prevention and treatment. Discov Med 2012; 14(75): 125–131.

Lokamani

Looi

Md Ali

. Gelsolin and ceruloplasmin as potential predictive biomarkers for cervical cancer by 2D-DIGE proteomics analysis. Pathol Oncol Res 2014; 20(1): 119–129.

Richart

. Screening. The next century. Cancer 1995; 76: 1919–1927.

Echelman

Feldman

. Management of cervical precancers: a global perspective. Hematol Oncol Clin North Am 2012; 26: 31–44.

Kim

. Prevalence and seroprevalence of low-risk human papillomavirus in Korean women. J Korean Med Sci 2012; 27: 922–928.

Tindle

. Immune evasion in human papillomavirus-associated cervical cancer. Nat Rev Cancer 2002; 2: 59–65.

10.

Ferenczy

Franco

. Persistent human papillomavirus infection and cervical neoplasia. Lancet Oncol 2002; 3(1): 11–16.

11.

Chen

Lei

. Evaluation of p16/Ki-67 dual staining in detection of cervical precancer and cancers: a multicenter study in China. Oncotarget 2016; 7(16): 21181–21189.

12.

Petry

Schmidt

Scherbring

. Triaging Pap cytology negative, HPV positive cervical cancer screening results with p16/Ki-67 Dual-stained cytology. Gynecol Oncol 2011; 121(3): 505–509.

13.

Wentzensen

Schiffman

Palmer

. Triage of HPV positive women in cervical cancer screening. J Clin Virol 2016; 76(Suppl. 1): S49–S55.

14.

Cho

Cheng

. Oncoproteomics: current trends and future perspectives. Expert Rev Proteomics 2007; 4(3): 401–410.

15.

Rai

Zhang

Rosenzweig

. Proteomic approaches to tumor marker discovery. Arch Pathol Lab Med 2002; 126: 1518–1526.

16.

Blackstock

Weir

. Proteomics: quantitative and physical mapping of cellular proteins. Trends Biotechnol 1999; 17(3): 121–127.

17.

Arentz

Weiland

Oehler

. State of the art of 2D DIGE. Proteomics Clin Appl 2015; 9(3–4): 277–288.

18.

Zieske

. A perspective on the use of iTRAQ reagent technology for protein complex and profiling studies. J Exp Bot 2006; 57(7): 1501–1508.

19.

Latterich

Abramovitz

Leyland-Jones

. Proteomics: new technologies and clinical applications. Eur J Cancer 2008; 44(18): 2737–2741.

20.

Papachristou

Roumeliotis

Chrysagi

. The shotgun proteomic study of the human ThinPrep cervical smear using iTRAQ mass-tagging and 2D LC-FT-Orbitrap-MS: the detection of the human papillomavirus at the protein level. J Proteome Res 2013; 12(5): 2078–2089.

21.

Di Domenico

De Marco

Perluigi

. Proteomics strategies to analyze HPV-transformed cells: relevance to cervical cancer. Expert Rev Proteomics 2013; 10(5): 461–472.

22.

Zhu

. Proteomic identification of differentially-expressed proteins in squamous cervical cancer. Gynecol Oncol 2009; 112(1): 248–256.

23.

Arnouk

Merkley

Podolsky

. Characterization of molecular markers indicative of cervical cancer progression. Proteomics Clin Appl 2009; 3(5): 516–527.

24.

Lomnytska

Becker

Hellman

. Diagnostic protein marker patterns in squamous cervical cancer. Proteomics Clin Appl 2010; 4(1): 17–31.

25.

Fukushima

Murakami

Yoshitomi

. Comparative proteomic profiling in squamous cell carcinoma of the uterine cervix. Proteomics Clin Appl 2011; 5(3–4): 133–140.

26.

Wang

Jia

Huang

. Identification of biomarkers for lymph node metastasis in early-stage cervical cancer by tissue-based proteomics. Br J Cancer 2014; 110(7): 1748–1758.

27.

Zhao

Wang

. Differentially expressed proteins among normal cervix, cervical intraepithelial neoplasia and cervical squamous cell carcinoma. Clin Transl Oncol 2015; 17(8): 620–631.

28.

Ding

Yang

She

. iTRAQ-based quantitative proteomic analysis of cervical cancer. Int J Oncol 2015; 46(4): 1748–1758.

29.

Escobar-Hoyos

Yang

Zhu

. Keratin 17 in premalignant and malignant squamous lesions of the cervix: proteomic discovery and immunohistochemical validation as a diagnostic and prognostic biomarker. Mod Pathol 2014; 27(4): 621–630.

30.

De Marco

Bucaj

Foppoli

. Oxidative stress in HPV-driven viral carcinogenesis: redox proteomics analysis of HPV-16 dysplastic and neoplastic tissues. PLoS ONE 2012; 7(3): e34366.

31.

Doherty

Lawlor

Gillet

. Collateral sensitivity to cisplatin in KB-8-5-11 drug-resistant cancer cells. Anticancer Res 2014; 34(1): 503–507.

32.

Das

Prasad

Yadav

. Over expression of minichromosome maintenance genes is clinically correlated to cervical carcinogenesis. PLoS ONE 2013; 8(7): e69607.

33.

Grinstein

Wernet

Snijders

. Nucleolin as activator of human papillomavirus type 18 oncogene transcription in cervical cancer. J Exp Med 2002; 196(8): 1067–1078.

34.

Wang

Zheng

. Predictive significance of the alterations of p16INK4A, p14ARF, p53, and proliferating cell nuclear antigen expression in the progression of cervical cancer. Clin Cancer Res 2004; 10(7): 2407–2414.

35.

Branca

Ciotti

Giorgi

. Up-regulation of proliferating cell nuclear antigen (PCNA) is closely associated with high-risk human papillomavirus (HPV) and progression of cervical intraepithelial neoplasia (CIN), but does not predict disease outcome in cervical cancer. Eur J Obstet Gynecol Reprod Biol 2007; 130(2): 223–231.

36.

Neill

Schaefer

Iozzo

. Oncosuppressive functions of decorin. Mol Cell Oncol 2015; 2(3): e975645.

37.

Zhao

Wang

. Influences of recombinant adeno-associated virus-mediated Decorin gene transfection on cell cycle and apoptosis of SiHa cells. Ai Zheng 2005; 24(1): 28–32 (in Chinese).

38.

Razani

Altschuler

Zhu

. Caveolin-1 expression is down-regulated in cells transformed by the human papilloma virus in a p53-dependent manner. Replacement of caveolin-1 expression suppresses HPV-mediated cell transformation. Biochemistry 2000; 39(45): 13916–13924.

39.

Chan

Yeh

. Mutational, epigenetic and expressional analyses of caveolin-1 gene in cervical cancers. Int J Oncol 2003; 23(3): 599–604.

40.

Ren

Zhao

. LTPB2 acts as a prognostic factor and promotes progression of cervical adenocarcinoma. Am J Transl Res 2015; 7(6): 1095–1105.

41.

Naito

Ishiwata

Kurban

. Expression and accumulation of lumican protein in uterine cervical cancer cells at the periphery of cancer nests. Int J Oncol 2002; 20(5): 943–948.

42.

Truty

Kang

. Extracellular lumican inhibits pancreatic cancer cell growth and is associated with prolonged survival after surgery. Clin Cancer Res 2014; 20(24): 6529–6540.

43.

Suhovskih

Mostovich

Kunin

. Proteoglycan expression in normal human prostate tissue and prostate cancer. ISRN Oncol 2013; 2013: 680136.

44.

Lai

Greenberg

. TGM2 and implications for human disease: role of alternative splicing. Front Biosci. 2013; 18: 504–519.

45.

Chiovaro

Chiquet-Ehrismann

Chiquet

. Transcriptional regulation of tenascin genes. Cell Adh Migr 2015; 9(1–2): 34–47.

46.

Buyukbayram

Arslan

. Value of tenascin-C content and association with clinicopathological parameters in uterine cervical lesions. Int J Cancer 2002; 100(6): 719–722.

47.

Nagata

Fujita

Ida

. Identification of potential biomarkers of lymph node metastasis in oral squamous cell carcinoma by cDNA microarray analysis. Int J Cancer 2003; 106(5): 683–689.

48.

Bella

. Collagen structure: new tricks from a very old dog. Biochem J 2016; 473(8): 1001–1025.

49.

Mienaltowski

Birk

. Structure, physiology, and biochemistry of collagens. Adv Exp Med Biol 2014; 802: 5–29.

50.

Contag

Gostout

Clayton

. Comparison of gene expression in squamous cell carcinoma and adenocarcinoma of the uterine cervix. Gynecol Oncol 2004; 95(3): 610–617.

51.

Misawa

Kanazawa

Misawa

. Hypermethylation of collagen α2 (I) gene (COL1A2) is an independent predictor of survival in head and neck cancer. Cancer Biomark 2011–2012; 10(3–4): 135–144.

52.

Zhou

Jiang

. Identification of a DNA methylome profile of esophageal squamous cell carcinoma and potential plasma epigenetic biomarkers for early diagnosis. PLoS ONE 2014; 9(7): e103162.

53.

Morris

Ricketts

Gentle

. Identification of candidate tumour suppressor genes frequently methylated in renal cell carcinoma. Oncogene 2010; 29(14): 2104–2117.

54.

Zhao

Shi

. Let-7d suppresses growth, metastasis, and tumor macrophage infiltration in renal cell carcinoma by targeting COL3A1 and CCL7. Mol Cancer 2014; 13: 206.

55.

Qiu

Sun

Deng

. miR-29a/b enhances cell migration and invasion in nasopharyngeal carcinoma progression by regulating SPARC and COL3A1 gene expression. PLoS ONE 2015; 10(3): e0120969.

56.

Shapland

Hsuan

Totty

. Purification and properties of transgelin: a transformation and shape change sensitive actin-gelling protein. J Cell Biol 1993; 121: 1065–1073.

57.

Huang

Chen

Wang

. Identification of transgelin as a potential novel biomarker for gastric adenocarcinoma based on proteomics technology. J Cancer Res Clin Oncol 2008; 134: 1219–1227.

58.

Kim

Rhee

. Proteomic analysis distinguishes basaloid carcinoma as a distinct subtype of nonsmall cell lung carcinoma. Proteomics 2004; 4: 3394–3400.

59.

Shields

Rogers-Graham

Der

. Loss of transgelin in breast and colon tumors and in RIE-1 cells by Ras deregulation of gene expression through Raf-independent pathways. J Biol Chem 2002; 277: 9790–9799.

60.

Yakabe

Murakami

Kajimura

. Functional significance of transgelin-2 in uterine cervical squamous cell carcinoma. J Obstet Gynaecol Res 2016; 42(5): 566–572.

61.

Dai

Cao

Yang

. P68 RNA helicase as a molecular target for cancer therapy. J Exp Clin Cancer Res 2014; 33: 64.

62.

González-Duarte

Cázares-Ordoñez

Díaz

. The expression of RNA helicase DDX5 is transcriptionally upregulated by calcitriol through a vitamin D response element in the proximal promoter in SiHa cervical cells. Mol Cell Biochem 2015; 410(1–2): 65–73.

63.

Wang

Luo

Zhou

. DDX5 promotes proliferation and tumorigenesis of non-small-cell lung cancer cells by activating β-catenin signaling pathway. Cancer Sci 2015; 106(10): 1303–1312.

64.

Clark

Hadjimichael

Temperley

. p68/Ddx5 supports β-catenin & RNAP II during androgen receptor mediated transcription in prostate cancer. PLoS ONE 2013; 8(1): e54150.

65.

Lyer

Nicol

Quinlan

. The RNA helicase/transcriptional co-regulator, p68 (DDX5), stimulates expression of oncogenic protein kinase, Polo-like kinase-1 (PLK1), and is associated with elevated PLK1 levels in human breast cancers. Cell Cycle 2014; 13(9): 1413–1423.

66.

Takai

Hamanaka

Yoshimatsu

. Polo-like kinases (Plks) and cancer. Oncogene 2005; 24(2): 287–291.

67.

Ponnusamy

Meyers-Needham

Senkal

. Sphingolipids and cancer: ceramide and sphingosine-1-phosphate in the regulation of cell death and drug resistance. Future Oncol 2010; 6(10): 1603–1624.

68.

Canals

Perry

Jenkins

. Drug targeting of sphingolipid metabolism: sphingomyelinases and ceramidases. Br J Pharmacol 2011; 163(4): 694–712.

69.

Sanger

Ruckhaberleb

Gyorffy

. Acid ceramidase is associated with an improved prognosis in both DCIS and invasive breast cancer. Mol Oncol 2015; 9(1): 58–67.

70.

Hanker

Karn

Holtrich

. Acid ceramidase (AC)—a key enzyme of sphingolipid metabolism—correlates with better prognosis in epithelial ovarian cancer. Int J Gynecol Pathol 2013; 32(3): 249–257.

71.

Evans

Mitchell

Spriqqs

. Members of the poly (rC) binding protein family stimulate the activity of the c-myc internal ribosome entry segment in vitro and in vivo. Oncogene 2003; 22(39): 8012–8020.

72.

Blyn

Towner

Semler

. Requirement of poly(rC) binding protein 2 for translation of poliovirus RNA. J Virol 1997; 71(8): 6243–6246.

73.

Wang

Vardy

Tan

. PCBP1 suppresses the translation of metastasis-associated PRL-3 phosphatase. Cancer Cell 2010; 18(1): 52–62.

74.

Han

Xin

Zhao

. RNA-binding protein PCBP2 modulates glioma growth by regulating FHL3. J Clin Invest 2013; 123(5): 2103–2118.

75.

Perrotti

Calabretta

. Post-transcriptional mechanisms in BCR/ABL leukemogenesis: role of shuttling RNA-binding proteins. Oncogene 2002; 21(56): 8577–8583.

76.

Chen

Hao

Liu

. The histone deacetylase SIRT6 suppresses the expression of the RNA-binding protein PCBP2 in glioma. Biochem Biophys Res Commun 2014; 446(1): 364–369.

77.

Liu

Zhang

. The RNA-binding protein PCBP2 facilitates gastric carcinoma growth by targeting miR-34a. Biochem Biophys Res Commun 2014; 448(4): 437–442.

78.

Jean-Philippe

Paz

Caputi

. hnRNP A1: the Swiss army knife of gene expression. Int J Mol Sci 2013; 14(9): 18999–19024.

79.

Bonomi

di Matteo

Buratti

. HnRNP A1 controls a splicing regulatory circuit promoting mesenchymal-to-epithelial transition. Nucleic Acids Res 2013; 41(18): 8665–8679.

80.

Fay

Kelehan

Lambkin

. Increased expression of cellular RNA-binding proteins in HPV-induced neoplasia and cervical cancer. J Med Virol 2009; 81(5): 897–907.

81.

Denko

Schindler

Koong

. Epigenetic regulation of gene expression in cervical cancer cells by the tumor microenvironment. Clin Cancer Res 2000; 6(2): 480–487.

82.

Mukherjee

Muralidhar

Bafna

. MCM immunocytochemistry as a first line cervical screening test in developing countries: a prospective cohort study in a regional cancer centre in India. Br J Cancer 2007; 96(7): 1107–1111.

83.

Hüttemann

Pecina

Rainbolt

. The multiple functions of cytochrome c and their regulation in life and death decisions of the mammalian cell: from respiration to apoptosis. Mitochondrion 2011; 11(3): 369–381.

84.

Lee

Yim

Kim

. Proteomic analysis of anti-cancer effects by paclitaxel treatment in cervical cancer cells. Gynecol Oncol 2005; 98(1): 45–53.

85.

Yashar

Spanos

Taylor

. Potentiation of the radiation effect with genistein in cervical cancer cells. Gynecol Oncol 2005; 99(1): 199–205.

86.

Lee

Shim

Kho

. Protein profiling and identification of modulators regulated by the E7 oncogene in the C33A cell line by proteomics and genomics. Proteomics 2004; 4(3): 839–848.

87.

Díaz-Ramos

Roig-Borrellas

García-Melero

. α-Enolase, a multifunctional protein: its role on pathophysiological situations. J Biomed Biotechnol 2012; 2012: 156795.

88.

Yoshida

Okamoto

Tozawa-Ono

. Proteomic analysis of differential protein expression by brain metastases of gynecological malignancies. Hum Cell 2013; 26(2): 56–66.

89.

Feo

Arcuri

Piddini

. ENO1 gene product binds to the c-myc promoter and acts as a transcriptional repressor: relationship with Myc promoter-binding protein 1 (MBP-1). FEBS Lett 2000; 473(1): 47–52.

90.

Focher

Spadari

. Thymidine phosphorylase: a two-face Janus in anticancer chemotherapy. Curr Cancer Drug Targets 2001; 1(2): 141–153.

91.

Akiyama

Furukawa

Sumizawa

. The role of thymidine phosphorylase, an angiogenic enzyme, in tumor progression. Cancer Sci 2004; 95(11): 851–857.

92.

Isaka

Sawai

Tomiie

. Expression of platelet-derived endothelial cell growth factor/thymidine phosphorylase in cervical intraepithelial neoplasia. Int J Oncol 2002; 21(2): 281–287.

93.

Kobayashi

Wada

Ohara

. Enzymatic activities of uridine and thymidine phosphorylase in normal and cancerous uterine cervical tissues. Hum Cell 2007; 20(4): 107–110.

94.

Grce

Matovina

Milutin-Gasperov

. Advances in cervical cancer control and future perspectives. Coll Antropol 2010; 34(2): 731–736.

95.

De Boer

Jordanova

Kenter

. High human papillomavirus oncogene mRNA expression and not viral DNA load is associated with poor prognosis in cervical cancer patients. Clin Cancer Res 2007; 13(1): 132–138.

96.

Tropé

Sjøborg

Eskild

. Performance of human papillomavirus DNA and mRNA testing strategies for women with and without cervical neoplasia. J Clin Microbiol 2009; 47(8): 2458–2464.

97.

Möckel

Quaas

Meisel

. Human papillomavirus E6/E7 mRNA testing has higher specificity than liquid-based DNA testing in the evaluation of cervical intraepithelial neoplasia. Anal Quant Cytol Histol 2011; 33(6): 311–315.

98.

Basu

Banerjee

Mittal

. Sensitivity of APTIMA HPV E6/E7 mRNA test in comparison with hybrid capture 2 HPV DNA test for detection of high risk oncogenic human papillomavirus in 396 biopsy confirmed cervical cancers. J Med Virol 2016; 88(7): 1271–1278.

99.

Munkhdelger

Kim

Wang

. Performance of HPV E6/E7 mRNA RT-qPCR for screening and diagnosis of cervical cancer with ThinPrep Pap test samples. Exp Mol Pathol 2014; 97(2): 279–284.

100.

Wang

Lee

Park

. Diagnostic performance of HPV E6/E7 mRNA and HPV DNA assays for the detection and screening of oncogenic human papillomavirus infection among woman with cervical lesions in China. Asian Pac J Cancer Prev 2015; 16(17): 7633–7640.

101.

Bouvard

Baan

Straif

. A review of human carcinogens—part B: biological agents. Lancet Oncol 2009; 10(4): 321–322.

Proteomic identification of potential biomarkers for cervical squamous cell carcinoma and human papillomavirus infection

Abstract

Keywords

Introduction

Materials and methods

Tissue specimens

DNA extraction and HPV detection

Protein extraction

Tryptic digestion and iTRAQ reagent labeling

Peptide fractionation by strong cation exchange chromatography and reversed-phase chromatography

Peptide analysis by nano-liquid chromatography coupled with Q-Exactive mass spectrometry

Database search and quantitative data analysis

Bioinformatics analysis of MetaCore™ software

RNA extraction and quantitative reverse transcription polymerase chain reaction analysis

IHC analysis

Statistical analysis

Results

Identification of the protein expression profile associated with cervical carcinoma and HPV16 infection by nano-LC coupled with Q-Exactive MS

Bioinformatics analysis of the identified proteins and the prediction of potential biomarkers

Verification of the profile of protein expression by quantitative RT-PCR and IHC

Discussion

Conclusion

Footnotes

Acknowledgements

Declaration of conflicting interests

Ethical approval

Funding

Informed consent

References