Loss of heterozygosity and human telomerase reverse transcriptase (hTERT) expression in bronchial mucosa of heavy smokers




Lung carcinogenesis is a multistep process of accumulation of genetic changes, including loss of heterozygosity (LOH), and precedes phenotypic transformation of the bronchial mucosa. The activity of telomerase, correlating with the hTERT mRNA expression, is detectable in a majority of neoplasms. In this study, the frequency of LOH and hTERT expression in bronchial mucosa of heavy smokers in bronchoscopic biopsies was analyzed.


LOH was examined in 122 bronchial specimens from 81 smokers (67 normal mucosa/bronchitis, 12 squamous metaplasia, 28 dysplasia, 15 bronchogenic carcinoma specimens) by polymerase chain reaction (PCR) and capillary electrophoresis by using 7 fluorescence-labeled markers matching 5 chromosomal regions. hTERT expression was analyzed in 87 specimens (45 normal mucosa/bronchitis, 12 squamous metaplasia, 18 dysplasia, 12 bronchogenic carcinoma specimens) by real-time quantitative reverse-transcription PCR.


LOH was detected in at least 1 chromosomal region in 51 of 122 (41.8%) specimens; the incidence in normal bronchial mucosa and preneoplastic lesions was similar (20%–40%); a substantial rise (87%) occurred in carcinomas. The median normalized hTERTN values were 6.67 in normal epithelium/chronic bronchitis, 18.38 in squamous metaplasia, 13.31 in epithelial dysplasia, and 75.46 in carcinomas. These results were significantly different (P = .0036). With an increasing number of LOH, the median value of hTERTN expression rose, but hTERT was expressed also in tissue samples without any LOH detection.


Results indicated that hTERT expression, together with LOH, represent early events in lung carcinogenesis, as both were detected in precancerous lesions and in normal epithelium of heavy smokers. Cancer 2007. © 2007 American Cancer Society.

Bronchogenic carcinoma remains 1 of the leading causes of cancer death worldwide, with tobacco carcinogen exposure as the major risk factor for most of the lung cancers. With current standard diagnostic and therapeutic methods, the 5-year survival rate still remains around 15% (with slightly better results in developed countries).1 Carcinogenesis, in general, is believed to progress in multiple steps while accumulating genetic changes that comprise activation of oncogenes or inactivation of tumor-suppressor genes, eg, by mutation, loss of heterozygosity, or aberrant methylation.2, 3 Recent investigations in molecular biology of lung cancer have revealed several genetic changes in genes such as K-ras, in the family of myc oncogenes, c-erb-B2, p53, pRb, FHIT and other putative genes on the short arm of chromosomes 3 and 9. Some of these genetic changes are often detectable in early, preinvasive lung cancer or even in minimally altered bronchial epithelium in smokers,4, 5 and they have been shown to precede the morphological alteration into preneoplastic bronchial lesions, which are described as metaplasia, mild, moderate, and severe dysplasia, and carcinoma in situ (CiS).6, 7 Recently, new techniques that increase the detection rate of early or microinvasive lesions in the bronchial tree have become available, ie, fluorescence bronchoscopy, low-dose spiral and high-resolution computed tomography, or endobronchial ultrasonography.8 Microsatellites (single tandem repeat-STR loci) are small nucleotide repeats that are used in linkage studies as markers for detection of loss of heterozygosity (LOH) or genomic instability in human neoplasms to indicate the occurrence of tumor suppressor genes (TSGs) in these chromosomal regions.4, 5, 9 Defects in genes involved in DNA mismatch repair were detected in both hereditary and sporadic tumors of the colon, endometrium, ovary, and other sites, and have been implicated in tumorigenesis.10–12 Limited data concerning molecular detection of LOH at these regions in bronchial preneoplastic lesions are available.13–15 To investigate the role of mismatch-repair genes in carcinogenesis of bronchogenic carcinoma, we examined LOH from the regions of hMLH1 (3p21), hMSH2 (2p15-p22), and hPMS1 (2q31-q33) genes. To compare the frequency of LOH in mismatch repair genes with some LOHs frequently observed in lung carcinogenesis and malignant transformation in general, we investigated several STR loci on the short arm of chromosome 9, where 2 known tumor-suppressor genes p15 and p16 are located.16 Their protein products are cyclin-dependent kinase inhibitors (CKI) that target cyclin dependent kinase 4/6 (CDK4/6) and act as negative cell-cycle regulators by regulating Rb protein phosporylation.17 The presence of tumor-suppressor genes in telomeric region on chromosome 11 (11p15.5) is also of particular interest, considering the specific LOH in this region in many neoplasms, including breast carcinoma, or nonsmall cell lung carcinoma (NSCLC).18, 19

Telomeres are specialized structures found at chromosomal ends of eukaryotic cells that protect them from terminal fusion and degradation during cell division.20 The activity of telomerase, the enzyme elongating telomeric DNA, is thought to be responsible for the immortality of cells during embryonal development, stem cells, cells in reproductive tissues of the ovary and testes, and neoplastic cells as well.21, 22 This enzyme, with a very complex regulatory mechanism, consists of 3 major components: RNA telomerase component (hTERC), telomerase-associated protein (TEP1), and telomerase catalytic subunit (hTERT).23 hTERT mRNA expression was shown to be most closely linked with telomerase activity,24 and it was proposed as a novel biomarker of lung cancer.25 Studies reporting hTERT mRNA expression were almost exclusively performed on tumor samples in comparison with normal tissue samples in resection specimens. Shibuya et al.26 studied changes in telomerase activity and hTERT expression in bronchoscopically excised preneoplastic lesions in patients with suspicious or diagnosed lung malignancy. In the current study, we determined the hTERT mRNA expression by using real-time quantitative reverse transcription polymerase chain reaction (RQ-RT-PCR) in a cohort of heavy smokers who had decreased forced-expiration volume in a variety of preneoplastic bronchial lesions, invasive carcinomas, and morphologically normal bronchial epithelia, and we assessed the significance of hTERT mRNA expression compared with molecular changes in chromosomal regions of DNA-repair genes and cell-cycle regulators, p15 and p16.



We studied 122 bioptic specimens obtained from 81 patients (56 men and 25 women) by a combination of white-light and autofluorescence bronchoscopy. All patients were current or former smokers, with >30 pack-years of smoking exposure (range, 30 to 126 pack-years; median, 38 pack-years;) and with forced expiration volume in the first second of expiration (FEV1) <70% of predicted value. The mean age was 61.5 years (age range, 39 to 83 years). Autofluorescence bronchoscopy was performed under local or total anesthesia with the use of the Onco-LIFE system (Xillix Technologies Corporation, Richmond, British Columbia, Canada) after conventional white-light bronchoscopy. Specimens were taken from all areas suspicious for a pathological finding. In addition, 1 or 2 foci of bronchoscopically normal mucosa were sampled. From each tested area, 2 samples were obtained: 1 fixed in formalin, embedded in paraffin, and stained with hematoxylin and eosin (H & E) for histological study; the other was immediately frozen in liquid nitrogen, stored at −80°C and further used for molecular genetic studies. The H & E-stained specimens were examined by 2 independent observers and were classified according to WHO criteria.7 All tested samples were taken superficially from bronchial mucosa, and the percentage of epithelial cells in the sample was >70%. Additionally, peripheral blood from each patient was taken for the constitutional DNA analysis. Written informed consent was acquired from all participants before they were included in the study.

Sample Processing

Nucleic acids extractions from cryopreserved tissue and from peripheral blood (PB) samples

Total DNA and RNA was extracted from 8 to 10 5-μm thick sections of samples with histologically validated sufficient amounts of epithelial cells cut from fresh-frozen tissues by using Trizol reagent (Life Technologies, Merelbeke, Belgium). Peripheral blood samples were collected into sterile tubes containing anticoagulant and stored at −80°C before further analysis. Nucleated cells were separated from the samples by osmotic lysis and centrifugation. Total DNA and RNA were then extracted by using Trizol reagent (Life Technologies, Merelbeke, Belgium).

Polymorphic STR markers and genotyping using fluorescent primers

The microsatellite markers investigated in this study and their cytogenetic locations are displayed in Table 1. Primers for the polymerase chain reaction were designed according to sequences from the Genome Data Base (www.ncbi.nlm.nih.gov/Entrez) and labeled with fluorescent dye (6-FAM, HEX, TAMRA) in a final volume of 12.5 μl containing 10 to 50 ng of DNA template, MgCl2 (1 to 2 mM), primers (12 pM each), dNTPs (0.2 mM each), 1× reaction buffer, and Taq Polymerase (0.3 U, Top-Bio, Prague, Czech Republic). The amplification was performed in an automated thermal cycler with a 5 minute 95°C denaturation step followed by 35 cycles of 94°C denaturation for 30 seconds, annealing at the temperature 60°C to 66°C for 45 seconds, and extension at 72°C for 30 seconds. The last cycle was followed by a final extension at 72°C for 20 minutes. Polymerase chain reaction products were quantitatively detected by fragmentation analysis by using the automated genetic analyzer ABI PRISM 310 (Applied Biosystems, Foster City, Calif). Each fluorescent peak was quantified in terms of size (number of base pairs) and peak height (Fig. 1). For significant LOH, an arbitrary limit of 25% or higher decrease of the peak of 1 of the alleles at single STR locus in comparison to its expression in normal tissue, ie, constitutional DNA from peripheral blood leucocytes, was taken. This limit is in general concordance with other previous studies.27, 28 To evaluate the allelic loss, normalized allelic ratios (R) were calculated as the equation in which R = (A1/A2)/(N2/N1), where A1 and N1 are the peak heights of the smaller alleles (peak) and A2 and N2 of the larger alleles29 in the bronchial sample (A) and in the peripheral blood (N).

Figure 1.

Loss of heterozygosity (LOH) observed at STR loci. The analysis was performed by using quantitative detection by fragmentation analysis in an automated genetic analyzer. Examples shown are from 3 subjects: (a) a heterozygote patient for the analyzed allele without LOH; (b) noninformative homozygote; (c) a heterozygote patient with LOH in the larger allele.

Table 1. List of Analyzed Microsatellite Markers, Their Chromosomal Localization and Corresponding Primer Sequences
STR locusLocalizationRepeatHeterozygosity index*Sequences of primers
  • *

    Proportion of heterozygotes at individual loci among investigated patients.

D2S117 (hPMS1)2q31-q33(CA)n0.895′GAC ATC AGG TAT ATT CAA TCC AC
D2S123 (hMSH2)2p15-p22(CA)n0.765′AAA GAG GAT GCC TGC CTT TA
D3S1611 (hMLH1)3p21.3(CA)n0.655′CCC CAA GGC TGC ACT T
D9S2599p21(CA)n0.755′GGC ATC ATT GCN CCA T
D11S136311p15.5(CA)n0.575′GAA AAT GGT ATT TAG AAA CCA A


Real-time quantitative reverse transcription polymerase chain reaction (RQ-RT-PCR) analyses were performed by using IQCycler (Bio-Rad, Philadelphia, Pa). RQ-RT-PCR for housekeeping gene beta-2-microglobulin (β-2-M) by using hydrolyzation TaqMan probe (5′-TGA TGC TGC TTA CAT GTC TCG ATC CCA-3′) was used to evaluate the amount and amplificability of cDNA. The sequences of the primers β2-M-F (5′-TGA CTT TGT CAC AGC CCA AGA TA-3′) and β2-M-R (5′-AAT CCA AAT GCG GCA TCT TC-3′) were derived from those published by Bijwaard et al.30 For the hTERT gene expression measurement, we applied RQ-RT-PCR with LNA probe generated by means of the Universal Probe Library for Human found at the website www.universalprobelibrary.com (Roche, Mannheim, Germany). The region of hTERT was amplified with the primers hTERT F (5′-GCC TTC AAG AGC CAC GTC-3′), and hTERT R (5′-CCA CGA ACT GTC GCA TGT-3′), LNA probe number #19. The PCR mix was composed of Platinum Taq DNA Polymerase (1 unit in the PCR buffer provided by the manufacturer, Gibco BRL, Carlsbad, Tex), MgCl2 (3.5 mM), dNTPs (0.2 mM each), BSA (0.25 μg/μL), primers (0.2 μM each), probe (0.1 μM), and cDNA (1 μl) in a final volume of 20 μl. The IQcycler program consisted of incubation at 40°C for 2 minutes and initial denaturation at 95°C for 2 minutes, followed by 45 PCR cycles at 95°C for 10 seconds, and 60°C for 60 seconds (single fluorescence measurement). Standard curves were created by using plasmid DNA calibration for β-2-M and an hTERT calibrator prepared and provided by Roche Company (LightCycler TeloTAGGGhTERT Quantification Kit, Roche, Basel, Switzerland). The cDNA sequence of the whole β-2-M gene was cloned into a PCR 2.1-TOPO vector and transformed into a TOP10 E. coli strain by using the TOPO TA Cloning Kit (Gibco BRL). The selected clones were screened for presence of inserts by using PCR. After bulk production, the plasmid was extracted by using the Miniprep method and quantified by spectrophotometry. The plasmid was then serially diluted in a xeno DNA (Salmon sperm, Sigma, Saint Louis, Mo). Standard curves for the tested genes were prepared in the following final concentrations: 1) for the β-2-M housekeeping gene 1 × 107, 3 × 106, 1 × 106, 3 × 105, 1 × 105, 1 × 104 copies/μL; and 2) for the gene hTERT transcript 2 × 106, 2.41 × 105, 1.80 × 104, 2.42 × 103, 2.70 × 102 copies/μL. The β-2-M and hTERT copy numbers and hTERT normalized copy numbers (hTERTN) were calculated according to the standard curve method. hTERTN was calculated as a ratio between hTERT gene expression and β-2-M gene expression ×104.

Statistical Analysis

All statistical tests were performed with the statistical program StatView (Abacus Concept, Berkeley, Calif). Comparison among the morphological findings, the smoking status, and the presence of molecular changes was performed by using the Kruskal-Wallis H-test and the Mann-Whitney U test. All reported P values were 2-sided, and those <.05 were considered to be statistically significant.


Histology of Bronchial Biopsies

From 81 patients, a total number of 122 bronchial epithelial foci were evaluated (Table 2). Sixty-seven (54.9%) of the 122 biopsies displayed either no morphological changes or mild inflammation. Foci of squamous metaplasia were present in 12 (9.8%), and dysplasia (mild, moderate, or severe) was present in 28 (23%) biopsies. In 15 (12.3%) cases, invasive carcinoma was found (8 squamous cell carcinomas, 6 small cell carcinomas, and 1 large cell carcinoma).

Table 2. Summarized Loss of Heterozygosity (LOH) Frequency (%) in the Bronchial Epithelium of Smokers Correlated With Histological Findings
HistologySamples no. (%)LOH no. (%)
Normal bronchial mucosa/chronic inflammation67 (54.9)24 (35.8)
Squamous metaplasia12 (9.8)4 (33.3)
Mild dysplasia18 (14.8)7 (38.9)
Moderate dysplasia5 (4.1)1 (20)
Severe dysplasia5 (4.1)2 (40)
Carcinoma15 (12.3)13 (86.7)

Relation between Histological Changes in Bronchial Mucosa and the Presence of LOH

All of the markers with the exception of D11S1363 were highly informative (Table 1). All samples were informative for at least 2 markers. Altogether, 51 (41.8%) of the 122 samples exhibited LOH in 1 or more loci. Of the samples, 24.6% (30 of 122) demonstrated LOH at 1 of the investigated microsatellite loci, 8.2% (10 of 122) at 2, and 9% (11 of 122) at 3 or more loci. Frequencies of LOH for individual markers ranged from 4.4% to 28.9% (Table 3). The most frequently detected abnormality was LOH at 1 or more regions of 9p (with the region 9p21 marked with D9S156 as the most common), followed by changes in the mismatch-repair gene region of 3p (hMLH1) and 2q (hPMS1). LOH of 11p15.5 was detected just in 3 cases, but only 68 of 110 samples were informative for the 11p marker used. The relation between presence of LOH in general and the morphologic pattern is summarized in Table 2 and Figure 2. In the group of normal bronchial mucosa/bronchitis, 24 of 67 (35.8%) cases exhibited LOH in at least 1 of the loci; among biopsies with squamous metaplasia, 4 of 12 (33.3%) exhibited LOH; in the group of dysplasia, 10 of 28 (35.7%); and in the group of carcinoma, 13 of 15 (86.7%) showed molecular changes corresponding to LOH. The data indicate that the LOH incidence in normal bronchial mucosa and preneoplastic lesions of any type was similar, but a significantly increased rate of incidence occurred in the category of carcinomas. We also compared histological changes in the bronchial mucosa with the presence of LOH at each of the loci. Detailed results are demonstrated in Figure 3, where the percentage of informative samples exhibiting LOH at each locus in individual histological categories is given.

Figure 2.

Overall frequency (%) of loss of heterozygosity (LOH) in bronchial epithelium of smokers in correlation with histological findings.

Figure 3.

Relation between the loss of heterozygosity (LOH) at individual chromosomal loci according to histological categories in bronchial epithelium of heavy smokers.

Table 3. Loss of Heterozygosity (LOH) Frequency at Individual Chromosomal Loci
STR LocusNo.*LOH No. (%)
  • *

    Number of analyzed informative heterozygotes at individual chromosomal loci.

D2S117 (hPMS1)10810 (9.26)
D2S123 (hMSH2)885 (5.7)
D3S1611 (hMLH1)7713 (16.9)
D9S1569026 (28.9)
D9S17110117 (16.8)
D9S2599114 (15.4)
D11S1363683 (4.4)

Relation between Histological Changes in Bronchial Mucosa and hTERT mRNA Expression

A total number of 87 bronchial biopsies from which both RNA and DNA were available were examined by hTERT mRNA analysis. Bronchial biopsies with moderate or severe dysplasia and carcinoma were considered to be positive histological findings. In this group, 90% (18 of 20) of samples expressed the hTERT mRNA. Among the negative samples (normal mucosa, chronic inflammation, squamous metaplasia and mild dysplasia), 65.7% (43 of 67) showed hTERT expression. Detailed results are shown in Table 4. The normalized hTERT mRNA expression (hTERTN) (Table 4) in normal bronchial epithelium/chronic bronchitis was 0 to 972 copies with a median value of 6.67 copies. In squamous metaplasia, it was 0 to 125.4, with a median of 18.38; in dysplastic epithelium, 0 to 257.65, with a median of 13.3; and in carcinomas, 0 to 2270.25, with a median of 75.46. The levels of hTERTN mRNA expression were significantly different when compared with the histological grade of preinvasive lesions and carcinoma (Kruskal-Wallis H-test, P = .0036) (Fig. 4). When the groups considered negative (normal/inflammatory changed bronchial mucosa, squamous metaplasia, and mild dysplasia) versus positive (moderate and severe dysplasia and carcinoma) were compared, the difference of the hTERT mRNA expression was also significant (Mann-Whitney U test, P = .0023). The difference between individual groups was lower when dysplasia categories were considered separately, but it still remained significant (P = .0066).

Figure 4.

Normalized hTERT mRNA (hTERTN) expression in copies/μL. (a, b) Histological pattern versus hTERTN expression in bronchial mucosa of heavy smokers; the correlation was significant, ([a] Kruskal-Wallis H-test, P = .0036, [b] Mann-Whitney U test, P = .0023). (c) Correlation between the number of loci exhibiting LOH and hTERTN expression; the difference was significant, Kruskal-Wallis H-test, P < .013.

Table 4. Percentage of Cases With hTERTN mRNA Expression and Levels of mRNA hTERTN Expression According to Histological Changes In Preneoplastic and Invasive Bronchial Lesions In Smokers
HistologyNo.Positive no. (%)Median of hTERTN mRNA expression (copies/μL)
Normal mucosa/chronic inflammation4526 (57.8)6.67
Squamous metaplasia129 (75)18.38
Dysplasia1816 (88.9)13.31
 Mild108 (80)13.31
 Moderate44 (100)63.83
 Severe44 (100)2.68
Carcinoma1210 (83.3)75.46

Relation Between the Presence of LOH and hTERT mRNA Expression

Finally, we compared the number of LOH in all samples for which both DNA and RNA were available for the molecular analysis with the relative hTERT mRNA expression. In biopsies without LOH, regardless of their morphology, hTERTN mRNA expression was 0 to 257.65 (median 3.96). In samples with 1 LOH detected, hTERTN mRNA expression was 0 to 972.34 (median 8.39); with 2 LOH present, it was 0 to 54.94 (median 14.17); with 3 LOH present, it was 6.53 to 88.76 (median 31.65); and with 4 LOH detected, the value was 727.38 to 2270.25 (median 1094.05). These data suggest that the hTERT mRNA is expressed in many samples of bronchial mucosa without any LOH detection. With an increasing number of LOH present, the median value of hTERTN expression rises (Fig. 4). Nevertheless, the difference between the individual groups in general was of a lower statistical significance (P = .013).

Relation Between Presence of Molecular Changes and Smoking Status

The relation between smoking habits and frequency of LOH and hTERT mRNA expression was further investigated. The number of loci exhibiting LOH in 1 bioptic sample was compared with cumulative smoking exposure (pack-years index) of these patients. In samples without any LOH, the cigarette exposure was 30–126 (median 65.8); with 1 LOH, it was 30–67 (median 51.0); with 2 LOHs, 30–72 (median 53.7); with 3 LOHs, it was 30–55 (median 57.4); and with 4, it was 30–70 (median 83.5). These results were not statistically significant (P = .1847). Concerning dependence of hTERTN expression and smoking status (data not shown), no unambiguous conclusions can be drawn (P = .8579).


Multistep tumor development in lung carcinogenesis is deducted from numerous histological, immunohistochemical, and molecular genetic investigations. Because a high proportion of lung carcinomas is related to smoking, we investigated a defined population of heavy smokers for presence of molecular changes in normal bronchial mucosa, in the bronchial epithelium exhibiting preneoplastic bronchial lesions, and in samples with carcinoma. White-light and autofluorescence bronchoscopy were used to detect early macroscopically suspicious areas, from which biopsy specimens were obtained. The extracted DNA was analyzed for the loss of heterozygosity (LOH) by polymerase chain reaction (PCR) by using fluorescent primers with quantitative detection by fragmentation analysis. These analyses used 7 polymorphic STR markers targeted to 5 chromosomal regions, where genes guarding mismatch-repair and cell-cycle prosecution are located. The highest frequency of LOH occurred at the chromosomal region 9p21 with potentially important tumor-suppressor loci containing p15 and p16 genes. These frequently studied regions2, 4, 5 were compared with findings at loci of mismatch-repair genes on chromosomes 2p, 2q, and 3p and with the region on chromosome 11p15.5. As is shown in Figure 3, at least 1 of the STR loci comprising chromosome 9p exhibited LOH in >20% of biopsies in each category, except for squamous metaplasia. In the group of carcinomas, all of these loci showed LOH in >50% of samples. In contrast, LOH in mismatch-repair gene regions 2q and 3p tended to occur in the stage of moderate and severe dysplasia and in carcinoma. Concerning the LOH at 11p15.5, the number of noninformative homozygotes was too high to draw any further conclusions. Our results are in concordance with some previous studies, which clearly showed that LOH may be detected in histologically normal or metaplastic epithelium in smokers or exsmokers,4, 5 and that the number and frequency of LOH at specific gene loci increases with severity of bronchial preinvasive lesions observed morphologically.2 A lower detection rate of LOH in preinvasive lesions in comparison to previous reports may be caused by a quite low proportion of such lesions in investigated tissue samples. This is, in fact, related to a limited number of biopsies taken from 1 patient and from 1 bronchial division.

To our knowledge, only limited data dealing with hTERT mRNA expression in precancerous lesions occurring in heavy smokers are available. In the current study, we analyzed and quantified hTERT mRNA expression in the bronchial epithelium of heavy smokers who exhibited histological changes as described above. Although the hTERT mRNA expression RQ-RT-PCR detection method that we used represents only an indicator of the expression level of telomerase mRNA, not a direct assay for the presence of active telomerase, some recent studies have shown a close correlation between hTERT expression and telomerase activity, the expression of hTERT subunit being the limiting factor for telomerase activity.26, 31, 32 Moreover, the RQ-RT-PCR is the most sensitive, accurate, and adaptable RNA quantitative technique.30 We showed that elevation of hTERT mRNA expression correlates with severity of histological changes, which is in general concordance with previous studies performed on surgical resection specimens or in a group of patients with a previous malignancy and unknown smoking status.26, 33 In addition to these studies, our data encompass quite a large cohort of heavy smokers with a different morphological pattern of preinvasive lesions, invasive carcinoma, and bronchial epithelium exhibiting normal appearance or bronchitic changes. The vast majority of studies showed a direct association between telomerase activity and transformed cells in surgically resected lung carcinomas,25, 34, 35 with <0.5% of telomerase-positive normal cells.35 Recently, some other studies reported the absence of a connection between telomerase and cancer, demonstrating that telomere activity occurs in normal regenerative cells.36 In our study, telomerase expression was not restricted to invasive lesions exclusively. In contrast, >50% of normal bronchial epithelium of heavy smokers displayed hTERT expression, although the median values were slightly lower than those in dysplastic lesions. This fact may be associated with increased regenerative processes in damaged bronchial epithelium and may reflect an early and specific molecular change induced by chemical carcinogens present in tobacco smoke. Also, Yim et al.37 recently documented the association between telomerase activity in mid-passage normal bronchial epithelial cell cultures and tobacco exposure and suggested that tobacco carcinogen exposure may increase telomerase activity. Therefore, we assume that detection of hTERT in clinical samples of heavy smokers does not overtly imply presence of carcinoma or severe dysplastic lesions as was suggested by other authors.25 Another possibility of hTERT expression present in histologically normal or inflammatory changed bronchial mucosa could rest in the admixture of leucocytes in bronchial mucosa. However, the leucocyte number in normal mucosa observed during histological examination was very low, and even some bronchitis samples exhibited no hTERT mRNA expression. The negativity of hTERT mRNA expression in 2 patients with carcinoma may be explained by other telomerase-independent, recombination-based mechanisms, which prevent telomere shortening in neoplastic cells.

The timing of telomerase activation and correspondent hTERT mRNA expression in preneoplastic bronchial lesions is still not completely understood. We compared the telomerase expression with the occurrence of LOH, which is, together with microsatellite instability, regarded as being present already in normal bronchial epithelia of heavy smokers.5, 9 Our data suggest that also the telomerase activity measured by the expression of hTERT mRNA expression in bronchial mucosa of heavy smokers is present in approximately 58% of normal bronchial epithelium samples and in 83% of bronchial biopsies that exhibit metaplastic and dysplastic changes. The comparison of presence of investigated molecular changes and cumulative smoking exposure (pack-years index) of patients did not show any statistically significant relation, similarly demonstrated by other authors.28, 38

In conclusion, we systematically analyzed presence of LOH on various chromosomal loci and the hTERT mRNA expression at different morphological stages of carcinogenesis in bronchial epithelium of heavy smokers. We demonstrated the relation between molecular changes in chromosomal regions harboring mismatch-repair genes and cell-cycle regulators and telomerase expression, and between these molecular genetic events and histological changes in bronchial mucosa in of heavy smokers.

These data provide insight into hTERT mRNA expression in correlation with the presence of LOH in bronchial mucosa of heavy smokers and suggest that it represents 1 of the early events in the multistep process of pulmonary carcinogenesis.


The authors thank Dr. J. Tkaczyk for a valuable help with statistical analysis, and Bc. I. Hilska and Mgr. A. Augustinakova for technical assistance.