Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Lung cancer is one of the most commonly diagnosed malignancies throughout the world. The prognosis for lung cancer is generally poor. Only 20 to 30% of the patients are operable at diagnosis, and only half of these can be successfully resected. The five-year survival rate of lung cancer is less than 15%. About 80% of the patients die of tumor recurrence following surgical resection (1). Although many studies have focused on the carcinogenic and metastatic process of lung cancer, most have failed to contribute to a significant breakthrough.

Recently, aberrant expression of mucins and mucin-related antigens have been noted to be poor survival factors in carcinomas arising from various organs, such as colon and breast cancer (2, 3). Experimentally, mucins are known to promote tumor cell invasion and metastasis and to modulate the immune recognition phenomenon of cancer cells (4). So far, few studies have addressed the prognostic implications of mucin and mucin-associated antigens in lung cancer, yielding conflicting results (8, 9).

The dysregulation of mucin expression in cancer includes aberrant glycosylation, underglycosylation, and overexpression of mucin peptides (10). Evidence through in vitro study has demonstrated that the expression of mucin in cancer cells can decrease tumor cell aggregation, promote tumor cell invasion, block lymphocyte targeting, and facilitate metastasis by escaping surveillance of immune systems. Several studies have investigated the expression status of mucins in lung cancer, some of which focused on the prognostic implications of mucin-associated carbohydrate antigens in lung cancer. The results have pointed out that expression of some carbohydrate antigens, e.g., sialyl Le^x and sialyl Le^a antigens, in lung cancers was correlated with shorter survival (11).

In one of our previous studies, we demonstrated the clinical significance of sialomucin expression in non-small cell lung cancer (NSCLC). Patients bearing tumors with sialomucin expression tended to have postoperative relapse and poor prognoses despite curative resection (12). It was not clear whether the expression of sialomucin in lung cancer is caused by an abnormal glycosylation process or by the expression of a specific mucin gene product. In searching for a possible mucin gene candidate to explain this phenomenon, we performed another study, using slot blot analysis with mucin gene oligonucleotide probes against the tandem repeat (VNTR) region of various mucin genes (MUC1, -2, -3, -4, -5B, and -5AC). The results showed a tendency of NSCLCs with overexpression of MUC5 genes (MUC5B and MUC5AC) to relapse after curative operation (13). Could the increase of MUC5 gene products lead to increased sialylation of mucins? The slot blot analysis using mucin gene VNTR probes had several shortcomings in comparing the expression amounts among different mucin genes and among different specimens. The signal intensities obtained from mucin gene VNTR probes may not consistently equal the exact copy numbers of mucin gene transcripts, because of the unpredictable number of repeats for mucin genes in different individuals (14). One solution is to develop a method that can precisely quantitate the mucin gene transcripts within and among individuals, through quantitation of the copy numbers of the nontandem repeat region of mucin genes. We therefore established a modified quantitative competitive polymerase chain reaction (QC-PCR) analysis to address this problem. We found that MUC5AC gene expression did correlate with sialomucin expression.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Tissues and RNA Extraction

Twenty-five tumor specimens of NSCLC (12 squamous cell carcinomas and 13 adenocarcinomas) and 15 paired nontumor lung portions were randomly selected from the tissue bank at National Taiwan University Hospital. The standard procedures for treating the resected specimens are summarized below. Immediately after excision, tumor samples are snap-frozen in isopentane at 60°C, placed in a sterile jar, and stored at 70°C until processed. Total cellular RNA is extracted from tissues using the guanidinium thiocyanate-phenol-chloroform extraction method (15). After verification of specimen histology by cryostat sectioning, the frozen tissue specimens (100 to 500 mg) are placed in guanidinium thiocyanate solution and subjected to RNA extraction. The paraffin blocks of these tumors are sliced and stained histochemically for sialomucin and sulfomucin. The clinical data of the patients are collected by chart review and regular outpatient follow-up.

Histochemistry

High-iron diamine/alcian blue pH 2.5 (HID/AB) stain was used to study the expression of sialomucin in NSCLC  tissues (16). After dewaxing and hydration, the paraffin  sections were treated with diamine solutions (containing N-N'-dimethyl-m-phenylene diamine and N-N'-dimethyl- p-phenylene diamine) for 18 h at room temperature. The sections were washed briefly with distilled water and treated with 1% alcian blue in 3% aqueous acetic acid for 30 min and were then washed thoroughly in 80% (vol/vol) alcohol. The sections were dehydrated and mounted. Sialomucin was stained blue, and sulfomucin was stained black/ dark brown.

Immediately after staining, the slides were evaluated using a method previously described (17). The histochemically stained tumors were examined by two separate investigators without their knowing the clinical information. The mucin content was graded from 0 to 4: 0, absence; 1, trace amounts; 2, focal presence; 3, moderately extensive; 4, widespread. Blocks were considered as positive when mucin content was rated at least 2 by two investigators. Any block given a different score by the examiners was reviewed using a video monitor for final grading by both reviewers.

Construction of the Competitive Fragments Used for QC-PCR

The competitive templates for amplification of specific mucin genes (MUC1, MUC2, MUC5AC) were complementary RNA (cRNA) fragments produced by in vitro transcription of mutated DNA fragments constructed by using the PCR technique developed by Celi and associates (18) and Deng and colleagues (19). (For illustration, see Figure 1A.) Table 1 summarizes the primer sets used for generating both mutated and native PCR fragments of each mucin gene and G-like gene (a housekeeping gene used as an internal control for RNA quantity [20, 21]). The mutated PCR fragments were ligated into PGEM-T vectors, and in vitro transcription was carried out by T7 RNA polymerase. The details of the procedures are described as follows.

View larger version (16K): [in this window] [in a new window]   View larger version (48K): [in this window] [in a new window]   View larger version (26K): [in this window] [in a new window]   Figure 1.   (A) Example of construction of RNA internal standards. Mutated MUC2 fragments were first amplified by PCR, then ligated into PGEM-T vector, cloned, sequenced, and transcribed in vitro into RNA internal standards after large-scale plasmid preparation. (B) Examples of the QC-PCR products of MUC1, MUC2, MUC5AC, and G-like genes after electrophoresis in 6% nondenaturing polyacrylamide gel and staining with ethidium bromide. (C ) Example of using linear regression to calculate the amount of MUC1 gene transcripts in a tumor specimen. The linear regression equation was y = a + bx, where the values on the y-axis are natural log transformations (intensity of native fragment × size of internal standard/intensity of internal standard × size of native fragment), and the values on the x-axis are natural log transformations of the quantity of internal standards in each tube of PCR reaction. e(a/b) equals the amount of target gene transcript in picograms.

After electrophoresis in 6% polyacrylamide gel and staining with ethidium bromide, the bands of mutated PCR fragments were cut out, sliced, and put into a microfuge tube. Elution buffer with 1 to 2 times the volume of the slice was added to the microfuge tube, and the mixture was incubated overnight at 37°C on a rotary platform. After centrifugation, the supernatant was eluted with equal volume of 1:1 phenol:chloroform, centrifuged, and precipitated overnight at 20°C with 95% alcohol 2.5 times the volume of the aqueous layer.

The eluted DNA fragment was subjected to ligation reaction with PGEM-T vector in a molar ratio of 1:20 (50 ng PGEM-T:100 ng DNA fragment). Three microliters of ligation mixture was mixed with 200 µl of JM-109 competent cells, incubated in ice for 30 min, put at 42°C for 40 s, and reincubated in ice for 10 min. The transformation mixture was then plated on Ampicillin-LB agar for blue-white selection. At least 20 white colonies were selected for plasmid minipreparation and DNA sequencing. The positive clones were applied to large-scale preparation for plasmids which were used as the templates for RNA in vitro transcription.

In Vitro Transcription of RNA Templates for QC-PCR

Ten micrograms of PGEM-T vector cloned with each specific mutated PCR fragment were linearized with Sac1 restriction enzyme, followed by Klegnow fragment for 3' overhang conversion. After elution with 1% agarose electrophoresis, phenol-chloroform elution and alcohol precipitation, the linearized cloned PGEM-T vectors were dissolved in water treated with diethyl pyrocarbonate (DEPC-H₂O). A total of 3.75 µg of the vector was added with 20 µl of first strand reaction mixture (200 mM Tris-HCl, 30 mM MgCl₂, 10 mM spermicide, 10 mM dithiothreitol), 100 U of RNasin (RNase inhibitor 400 U/µl; Promega, Madison, WI), NTPs (50 mM ATP, CTP, GTP, UTP), and 40 U of T7 RNA polymerase in a 100-µl reaction. After incubation at 39°C for 2 h, 4 U of RNase-free DNase RQ-1 was added. The mixture was incubated at 37°C for 15 min to degrade the DNA template. After phenol-chloroform-isoamylalcohol elution and alcohol precipitation, the RNA transcript was dissolved in 20 µl of DEPC-H₂O. After spectrophotometric measurement of the RNA concentration, the RNA was serially diluted and stored at 70°C. The cRNA transcripts were named as the internal standards for each native (target) gene studied in QC-PCR.

QC-PCR

Six micrograms of tissue total RNA with 8 µg of random hexamer was mixed to a final volume of 10 µl. A total of 1.25 µl of this mixture was aliquoted into eight tubes, each individually containing 1,000, 300, 100, 30, 10, 3, 1, and 0.3 pg of target gene RNA internal standards. Reverse transcription (RT) reaction was carried out in a final volume of 10 µl, with the addition of a reaction mixture containing Superscript RTase (BRL, Gaithersburg, MD), RNasin, dNTP, dithiothreitol, and 5× reaction buffer. After reaction at 37°C for 90 min, the RT products were diluted by 12 µl of deionized, distilled water and heated at 85°C for 5 min. Two microliters of RT products from each tube were used for PCR in a volume of 20 µl. The thermocycle conditions for each mucin gene and G-like gene are listed in Table 1.

After amplification, the PCR products were electrophoresed at 80 V for 1 to 2 h in 6% nondenaturing polyacrylamide gel, stained in 5 µg/ml ethidium bromide solution for 30 min, and then destained. The signal intensity of amplified native and mutated products was directly measured and digitized by IS-1000 digital imaging system (Alpha Innotech Incorp., San Leandro, CA). A representative set of QC-PCR is shown in Figure 1C. With the aid of quantitative densitometry, the intensities of the native mucin gene products from tissue RNA templates, and intensities from mutant templates were plotted as a function of known amounts of mutant templates. The point where the intensities (of native and mutated mucin gene products) were equal to each other represented that the molecules of native mucin gene transcripts approximated those of the mutant templates. The quantity of the RNA transcripts of each mucin genes (MUC1, MUC2, MUC5AC) determined from QC-PCR were further corrected by the quantity of G-like gene transcripts. Expression ratios of the various mucin genes were defined as the quantity of mucin gene RNA transcripts divided by the quantity of G-like gene transcripts.

Statistical Analysis

Mann-Whitney rank sum test (22) was used to clarify the associations between mucin genes expression and pathologic stage, histologic type, and sialomucin expression of NSCLC.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

The epidemiologic data of 25 NSCLC patients are provided in Table 2. Seventeen were male, and eight were female. The mean age was 59.9 ± 14.0 yr. Sixteen were smokers, and nine were nonsmokers. Fourteen had Stage I, three had Stage II, and eight had stage IIIa disease.

Examples for the QC-PCR of mucin genes and G-like gene are shown in Figure 1B. The internal standard RNA and native RNA had generated PCR products of predicted size, with picture of competition by ethidium bromide staining, and the example of the application of linear regression to calculate the amount of mucin gene transcripts is shown in Figure 1C. The r value of the linear regression in QC-PCR assay ranged from 1.000 to 0.951. And the intra-assay and inter-assay coefficients of variations of the QC-PCR assay were 15.1% and 20.9%, respectively. The expression ratios (to G internal control gene) of MUC1, MUC2, and MUC5AC in paired nontumor lung tissues and NSCLCs are listed in Tables 3 and 4.

NSCLCs had a random expression of the three mucin genes studied, and no specific mucin gene expression could be demonstrated in lung cancer tissues. By comparing the expression amounts of mucin genes with the histologic type, both squamous cell carcinoma and adenocarcinoma can express high amounts of mucin gene transcripts. Adenocarcinomas expressed more MUC5AC gene transcripts than did squamous cell carcinomas (P = 0.03). There was no correlation between mucin gene expression and pathologic stages nor was there any association between cigarette smoking and any mucin gene product in NSCLC (P > 0.06).

The tumor tissues tended to express higher amounts of mucin genes than did their nontumor counterparts. Among the 15 paired surgical samples, mucin genes were overexpressed (arbitrarily defined as twice the expression ratio of nontumor lung portions) in 10 of the tissue pairs, five of the 10 had MUC1 overexpression, six had MUC2 overexpression, and five had MUC5AC overexpression (Table 5). Eight of the 10 adenocarcinomas had at least one mucin gene overexpression, as compared with two of the five squamous cell carcinomas.

Ten of the 13 adenocarcinomas expressed sialomucin (Table 4). In contrast, none of the 12 squamous cell carcinomas expressed sialomucin. Among the three mucin genes studied, the expression amounts of MUC5AC were higher in tumors with sialomucin expression than in those without (P = 0.012) (Figure 2).

View larger version (25K): [in this window] [in a new window]   Figure 2.   The expression ratios of MUC1 (A), MUC2 (B), and MUC5AC (C ) in NSCLC based on sialomucin expression.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In this study, we used QC-PCR analysis to evaluate the expression of mRNA transcripts encoding three mucin genes (MUC1, MUC2, and MUC5AC) in NSCLC. Among the three mucin genes studied, only MUC5AC gene expression was found to be associated with adenocarcinoma histologic subtype and sialomucin expression.

Several techniques have been applied to investigate the expression status of mucin genes in normal or neoplastic tissues, such as Northern or slot blot analysis, in situ hybridization, Western blot analysis, or immunohistochemistry (2, 23, 24). All these analyses, except one reported by Voynow and Rose (14), are mainly qualitative, or at best semiquantitative. Similar to our present study, the technique developed by Voynow and Rose was based on the construction of cRNA standards of the non-tandem repeat regions of mucin genes (MUC1, MUC2, MUC5AC), and create a cRNA standard curve for mucin genes. This method seemed to provide a satisfactory result in evaluating mucin gene transcripts in both cell lines and tissues. However, the method required a relatively high amount of good-quality RNA and may not be suitable for clinical application because of the limited size of the available specimen and the degradation of tissue RNA during processing.

The application of RT-PCR can increase the detection capability of tumor micro-metastasis and the subtle change of gene expression in cells or tissues (25). Because of the PCR amplification, only a scanty amount (much less than Northern blot analysis and slot blot analysis) of RNA is required to analyze the expression of a specific gene. However, tube-to-tube variation in reaction efficiency may cause as much as sixfold of difference among reactions in both reverse transcription and polymerase chain reaction, which make precise quantitation impossible in RT-PCR (27, 29). Several modifications have been carried out to overcome this problem, such as kinetic PCR assay (using serial dilution of RNA templates and cycle number restriction to control the PCR reaction at the exponential phase), and quantitative PCR (by adding external or internal standards of target gene for competition throughout the reaction of RT-PCR). Among these modifications, constructing a mutant RNA fragment of the target gene (internal standard) to compete with the native target gene transcripts during RT-PCR was the best choice for eliminating many factors that interfere with the data interpretation. This method overcomes uncertain factors in many steps that affect efficiency, including RNA extraction, cDNA synthesis, and PCR amplification. It also has a low test variability, usually less than 15% (30). In this study, we constructed cRNA internal standards from non-tandem repeat coding sequences of MUC1, MUC2, and MUC5AC and showed an acceptable variability in quantitation: the intra-assay and inter-assay coefficients of variance are 15.1% and 20.9%, respectively. Because of the low data variability, the low number of templates needed, and no need of radioactive material, QC-PCR is an ideal method to evaluate gene expression in clinical specimens.

There are, however, several disadvantages to the QC-PCR method. First, reactions must be carried out in multiple tubes (eight tubes in the present study) containing serial dilution of competitive fragments, a design developed to establish a competition curve and to assure the quality of reaction. Both the cost and time consumed in examining gene expression are high, since at least two sets of RT-PCR must be performed for the target gene in each case: one set is for the gene of interest, and one for the internal control gene. Second, the contamination of DNA, particularly by plasmid DNA, during RNA extraction is a problem of concern. Third, the uncertainty of potential differences in the kinetics of amplification between mutant and native fragments is also a problem frequently encountered.

Glycosylation of mucin peptide is a major process that determines the influence of mucoprotein on cell behavior. Among the many steps of glycosylation, sialylation is a major one. The abundance of sialic acid residue contributes to anti-adhesion properties of mucins (31). Alterations of sialylation influence the binding of cancer cells to reticuloendothelial cells and extracellular matrix and promote metastasis (32). Tumor cells with abundant mucin expression, especially with heavy sialyl acid residues, may have a better chance of survival and cause a poor prognosis (33, 34). Treatment of mouse and human colon adenocarcinoma cell lines with a specific inhibitor of sialic acid incorporation reduce the incidence of lung or liver metastases (35).

Whether mucin core peptide can be a major determinant of specific steps of the glycosylation process is unclear. Few studies have addressed this question. Cancer cells may have altered sets of glycosyltransferases, leading to a change of carbohydrate antigens regardless of which apomucins have been expressed. One example is the alteration of glycosyltransferases activity, such as 2,6-sialyltransferase (an enzyme related to formation of sialyl Tn antigen), in colon cancer cell lines (36). Alternatively, the expression of certain apomucins may have a linkage with specific carbohydrate antigens. A study investigating the expression of both apomucin mRNA and mucin-associated carbohydrate antigens has demonstrated a mucin peptide-specific glycosylation pattern in gastric mucosa epithelial cells (37). Superficial epithelial cells with MUC5AC gene expression coexpressed sialyl Le^a antigen, while in antral mucous cells expressing MUC6 gene, only Le^y antigen was expressed. Although not performing an immunohistochemical assay of sialylated related antigens, the present study shows a correlation between sialomucin and MUC5AC gene expression in lung adenocarcinomas. It is possible that an apomucin may instruct its own pattern of glycosylation, with the overexpression of certain apomucins, so that the associated carbohydrate antigens may influence the biologic behavior of the cancer cells. However, it is also possible that the expression of sialomucin and MUC5AC mucin gene in lung adenocarcinoma may be independent biologic events. Further investigations are necessary to verify this hypothesis.

In conclusion, we established a precise method to study the expression of mucin genes in surgical specimens of lung cancer and demonstrated an association between MUC5AC gene expression and sialylation of mucins in NSCLC. Further studies are anticipated to elucidate the underlying mechanism contributing to this phenomenon.