This article is mentioned in:

Introduction

Colorectal cancer is the third most frequently diagnosed cancer in males and the second most commonly diagnosed cancer in females, with over 1.2 million new cases and 608,700 deaths estimated to have occurred in 2008 worldwide (1). Early detection is essential for the successful diagnosis and treatment of CRC as well as for increasing the survival rate of CRC patients (2–4). Unfortunately, there are no significant clinical symptoms at an early stage making it difficult to provide effective diagnostic proposals for CRC patients (5). The existing diagnostic methods for CRC include fecal occult blood test, colonoscopy and carcinoembryonic antigen, but their sensitivity and accuracy is inadequate (6–9). Therefore, it is essential to identify novel serum biomarkers for CRC with high sensitivity and specificity.

Surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF-MS) and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) are high throughput techniques for the analysis of complex biological specimens with high sensitivity and specificity (10,11). SELDI-TOF-MS combined with ProteinChip technology has been used to identify novel biomarkers for CRC, breast cancer, thyroid cancer, endometriosis and other diseases (12–16). This technology can detect biomarkers in serum samples effectively, but it can only obtain the molecular weight data of biomarkers without protein sequences. In the present study, several technologies were combined to purify and identify a CRC-related protein. The SELDI-TOF-MS combined with ProteinChip technology was first used to detect the 3.9-kDa protein in serum samples. HPLC and MALDI-TOF-MS were then applied to purify and track the 3.9-kDa protein, which provided a pure 3.9-kDa protein biomarker. Liquid chromatography/mass spectometry (LC-MS/MS) was finally adopted to identify and decode the structural information of the 3.9-kDa protein biomarker.

Materials and methods

Patients

A total of 99 serum samples were collected upon approval by the Department of Surgical Oncology, The Second Affiliated School of Zhejiang University. All the serum samples used in this paper were collected with the agreement of the patients. Another 73 serum samples were collected from CRC patients, along with an additional 26 samples from healthy individuals. Patients with CRC had a median age of 56 years (ranging from 41 to 75; 46 male and 27 female) and consisted of 9 stage I, 21 stage II, 32 stage III and 11 stage IV patients, according to the TNM stage-system of UICC. The healthy control group was age- and gender-matched with the CRC group. Diagnoses of all the CRCs were confirmed pathological by two pathologists. All blood samples were collected preoperatively in the morning without food intake. The blood samples were placed at room temperature for 1–2 h, and centrifuged at 3,000 rpm for 10 min. Finally, the serum samples were stored at −80°C for further research.

Prochip array analysis

All serum specimens were thawed on ice and centrifuged at 10,000 rpm for 4 min at 4°C. An amount of 5 μl of each supernatant was mixed with 10 μl U9 buffer (9 M Urea, 2% CHAPS, 1% DTT) in a 96-well plate, and then the plate was agitated on a platform shaker for 30 min at 4°C. The U9/serum mixture was further diluted with 185 μl binding buffer (50 mM sodium acetate, pH 4.0). At the same time, the weak cation exchange (CM10) chips were activated by adding 200 μl binding buffer and agitated for 5 min twice. A 100 μl of diluted serum sample was added into each well of the activated Prochip arrays and agitated for 60 min. The chips were washed three times with 200 μl binding buffer and twice with 200 μl deionized water. After being air-dried, 1 μl of saturated solution of sinapinic acid in 50% acetonitrile and 0.5% trifluoroacetic acid was added to each well twice. Finally, the chips were detected using the Protein Biological System II (PBS-II), a mass spectrometer reader (Ciphergen Biosystems, USA).

Date were obtained by an average of 100 laser shots with an intensity of 170, a detector sensitivity of 6, a highest mass of 50 kDa and an optimized range of 2–20 kDa. The instrument was calibrated using the all-in-one peptide molecular mass standard before detecting samples.

Bioinformatic analysis

After baseline subtraction, the spectra intensities of all samples were normalized to the total ion current. Noise was filtered from the spectra, and then markers were detected with an automatic peak detection pass. Peak clusters were completed to cluster the markers in different samples with similar masses, which were defined by a mass window of 0.3% mass error. All the above analysis was performed using ProteinChip software 3.2.

Serum fractionation

After freeze thawing on ice, 100 μl of the serum sample was mixed with 200 μl U9 buffer (9 M Urea, 1% DTT 2% CHAPS) and vibrated on ice for 30 min. Then the sample was diluted with 5 ml WCX binding buffer (50 mM NaAc, pH 4.0) and loaded on the Oasis WCX SPE column (6 cc/150 mg, 30 μm) (Waters, USA). After washing with 5 ml WCX binding buffer, the column was eluted with 3 ml WCX elution buffer (1 M NaCl, 50 mM NaAc, pH 4.0). Every eluted fraction was concentrated to 100 μl using SpeedVac for further purification.

Purification of candidate protein biomarker

HPLC separation was performed using SCL-10AVP (Shimadzu, Japan) with a Sunchrom C18 column (250×4.6 mm, 5-μm particle size, 300 Å) (Waters, USA) and a C18 guard column (10 × 3 mm)(Shimadzu). The mobile phase consisted of solvent A (0.1% TFA/H2O) and solvent B (0.1% TFA/ACN). The HPLC separation was achieved with a linear solvent gradient: 100% A (0 min - 100% A (10 min) - 70% B (60 min) - 100% B (65 min) - 100% B (90 min) at a flow-rate of 0.5 ml/min. Multiple wavelengths of 214, 254 and 280 nm were used to detect the eluate. Each peak fraction was collected and concentrated using SpeedVac. The concentrated samples were analyzed using AXIMA-CFR™ Plus MALDI-TOF mass spectrometer (Shimadzu/Kratos, Manchester, UK) in linear mode to trace the protein biomarkers with α-cyano-4-hydroxycinnamic acid as matrix.

Identification of the candidate protein marker using LC-MS/MS

The fraction containing the candidate protein biomarker was in-solution digested following a standard protocol. Briefly, each fraction was dissolved in 25 mM NH4HCO3, and reduced with 10 mM DTT for 1 h at room temperature, alkylated by using 40 mM iodoacetamide in the dark for 1 h at room temperature. Then 40 mM DTT was added to quench the iodacetamide for 30 min at room temperature. An amount of 0.1 μg Trypsin (Promega Corporation, USA) was then added into the sample solution and incubated at 37°C overnight. A total amount of 1 μl formic acid was added to stop the digestion. The digested sample was loaded onto a home-made C18 column (100 mm × 100 μm) packed with Sunchron packing-material (SP-120-3-ODS-A, 3 μm) followed by nano-LC-ESI-MS/MS analysis. The LTQ mass spectrometer was operated in the data-dependent mode, in which the initial MS scan recorded the mass to charge (m/z) ratios of ions over the mass range from 400–2,000 Da, and then the 5 most abundant ions were automatically selected for subsequent collision-activated dissociation. All MS/MS data were searched in a human protein database downloaded from NCBI using the SEQUEST program (Thermo, USA). The p-values were calculated with the SPSS version 11.0 software and a p<0.05 was required for the results to be considered statistically significant.

Results

Discovery analysis

The data derived from the CRC and the healthy control groups were analyzed. The qualified peaks were ranked by the p-value of the non-parametric tests, and the top 10 peaks with obvious differences were selected and listed in Table I. Results showed that the intensity of the 3.9-kDa peak in the healthy controls (mean intensity value 6127.9) was much higher than that of the CRC patients (mean intensity value 2404.3) (Fig. 1).

Figure 1 The 3.9-kDa protein in colorectal cancer patients and healthy controls detected using SELDI-TOF-MS. Group 0, healthy control group; group 1, colorectal cancer patient group. Expression of the 3.9-kDa protein in group 0 was higher than that in group 1.

Table I Comparison of the protein peak intensities between the CRC patients and the healthy controls.

Table I

Comparison of the protein peak intensities between the CRC patients and the healthy controls.

Biomarker (m/z)	p-value	Mean in group 0	Mean in group 1	Std in group 0	Std in group 1
5639.535	1.64E-11	2007.398	7988.462	877.703	3904.474
5864.255	2.05E-11	60.212	263.025	38.586	161.735
5846.277	2.37E-11	38.989	347.398	23.900	234.303
5661.475	6.52E-11	176.875	529.646	72.959	242.287
8942.126	3.76E-10	2028.114	279.755	1333.727	663.509
5086.084	2.16E-09	62.566	284.565	56.215	176.938
5910.981	2.46E-09	1481.202	5753.641	994.956	3391.010
3938.931	5.37E-09	6127.917	2404.337	2567.623	1644.808
8146.209	5.37E-09	993.202	239.002	557.253	297.599
5070.722	8.94E-09	131.457	645.025	153.796	522.170

[i] Groups 0 and 1 represent the healthy control and the CRC patients, respectively.

Purification of the 3.9-kDa biomarker

The separation and purification of all the 10 biomarkers were carried out, and it was found that the purification of the 3.9-kDa protein was much easier than the others. As shown in Table I, the difference in the concentration of the 3.9-kDa protein in the serum of CRC patients and health controls was also very distinct. Therefore, the 3.9-kDa protein was selected as a biomarker for CRC and was further separated and purified. Serum samples were separated by HPLC and the eluate was sampled every 2 min. All of the eluates were analyzed by MALDI-TOF-MS. The 3.9-kDa biomarker was present in the eluate between 47 and 48 min (Fig. 2A). Then, we focused on the elutate from 47 and 48 min during HPLC separation, and the sampling frequency during this period was found to be even higher. Finally, it was confirmed that the 3.9-kDa protein flowed out from 47.62 to 47.85 min. Fig. 2B shows the result of the MALDI-TOF-MS analysis data of the 3.9-kDa biomarker.

Figure 2 Data of the purification and detection of the 3.9-kDa protein. (A) HPLC chromatogram of the candidate protein marker. The biomarker (3.9-kDa) is indicated. (B) MALDI-TOF-MS data of the 3.9-kDa biomarker.

Identification of the 3.9-kDa biomarker

The purified 3.9-kDa protein was identified by HPLC-ESI-MS/MS, with the data shown in Fig. 3. After trypsin digestion, the protein was divided into several protein segments. After searching the database, two peptide fragments were found including R. DLINEAMDVKL and R. EVDQDDEENSEEDE (Table II). Therefore, the sequence of the 3.9-kDa protein was decoded as DLINEAMDVKLKRQESQQKEVDQDD-EENSEEDE, which is a peptide fragment of serine/theonine kinase 4.

Figure 3 Results of the identification of the 3.9-kDa protein by LC-MS/MS. (A) Chromatogram of the peptide mixture. MS/MS spectra of (B) R. DLINEAMDVKL and (C) MS/MS spectra of R. EVDQDDEENSEEDE.

Table II Structural data of the 3.9 kDa protein biomarker.

Table II

Structural data of the 3.9 kDa protein biomarker.

Peptide	p-value	Sf	Score
R. DLINEAMDVKL	3.8E-005	0.87	2.207
R. EVDQDDEENSEEDE	1.1E-007	0.97	6.254

[i] P-value displays the probability value for the protein. Sf displays the final score that indicates the quality of the protein match. Score displays the value that is based upon the probability that the peptide is a random match to the spectral data.

Discussion

Early diagnosis is important to increase the survival rate for CRC patients. Carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) are tumor biomarkers for the diagnosis of CRC (17,18). However, these tumor markers either lack sensitivity or have a poor positive predictive value. Therefore, it is necessary to identify novel biomarkers with high sensitivity and specificity for the accurate diagnosis of CRC.

There are many different types of methods to identify protein biomarkers for various diseases. Two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) has been widely used to identify the differences in protein expression for the identification of useful biomarkers (19). However, this technique has poor reproducibility and limited resolution for proteins with molecular weights less than 20,000 Da. The SELDI-TOF-MS and MALDI-TOF-MS technologies are rapid and have sensitivity and high throughput. They can identify low-molecular-weight proteins, even with a molecular weight less than 20 kDa (20,21). These two technologies have identified a great number of small proteins, and the profiles of such proteins could be used to indicate the presence or absence of a disease with a variety of tumor types (22–25).

The protein biomarkers detected by SELDI-TOF-MS simply are assigned an M/Z value, but no protein name and structural information are provided. In this study, SELDI-TOF-MS combined with CM10 ProteinChip was used to identify the 3.9-kDa protein biomarker for CRC. We applied several biochemical technologies to perform the separation and purification of this biomarker. First, we used WCX-SPE column to enrich the proteins, which can be enriched by a CM10 protein chip. Then, HPLC was applied to separate and purify the 3.9-kDa protein with the eluates and was monitored by MALDI-TOF-MS. Finally, the biomarker was identified and decoded as a peptide fragment of MST1/STK4.

Serine/theonine kinase 4 (MST1/STK4) can phosphorylate H2B at S14 depending upon the cleavage of MST1 by caspase-3 and the phosphorylation is uniquely associated with apoptotic chromatin in humans (26). Researchers have reported that MST1/STK4 is a CRC-related protein according to the results of protein micro-arrays (27,28). It was also observed that the higher expression of MST1/STK4 was associated with an improved survival rate of colon cancer patients (29). Therefore, as a peptide fragment of MST1/STK4, the 3.9-kDa protein in serum is more meaningful for the diagnosis and prognosis for CRC patients.

In summary, we successfully found a 3.9-kDa protein biomarker, which has the clinical potential for the early diagnosis of CRC. The structural information of the 3.9-kDa protein biomarker also indicates that it is a peptide fragment of MST1/STK4. Further studies involving larger patient population or using pre-diagnostic serum are needed to confirm the importance of the 3.9-kDa protein biomarker as a diagnostic marker for CRC.

Acknowledgements

The authors appreciate the support from the National Natural Science Foundation of China (30901731), the National High Technology Research and Development Program of China (2012AA02A506) and the Fundamental Research Funds for the Central Universities (2012FZA7004).

References