Vol. 49 No. 2/2002 341–350 QUARTERLY

Searching for ways to improve the characterization of breast cancer we examined the relationship between the status of the FHIT gene transcript and amplification of c-myc and the c-erbB2 oncogene. Abnormal FHIT transcript was detected in 32 of 79 cancers examined. The presence of Fhit protein estimated by Western blots was evident only in cancers exhibiting a normal-sized FHIT transcript. This indicates that abnormal FHIT transcripts observed in our study did not encode any Fhit protein or the amount of such protein was very low. There was no association between the presence of aberrant FHIT gene transcript with age, tumor size, estrogen and progesterone receptor status, local metastases and histological grading. However, the abnormalities in FHIT gene transcripts were observed with different frequency depending on the histopathological type of the tumor. The aberrant FHIT transcript was detected in 60% of lobular cancers and only in 28% of ductal cancers. Analyzing the occurrence of c-myc and c-erbB2 amplification and the presence of aberrant FHIT gene transcripts we found that the aberrant FHIT transcript more frequently occurred in tissues with c-myc amplification. There was a significant (P < 0.05) correlation between the occurrence of the aberrant FHIT gene transcript with accompanying c-myc amplification and positive lymph node status. However, in order to evaluate the predictive value of these findings in breast cancer, an extended clinical follow up will be necessary.

The FHIT gene on chromosome 3 positioned at 3p14 encodes a protein with the activity of dinucleoside 5¢,5¢¢¢-P 1 ,P 3 -triphosphate (Ap 3 A) hydrolase (Barnes et al., 1996).Since its discovery in 1996, alterations of the FHIT gene have been frequently observed in a variety of tumors (Ohta et al., 1996;Negrini et al., 1996;Fong et al., 1997;Ingvarson et al., 2001;Yuan et al., 2000).Gathered evidence suggests that FHIT is a tumor suppressor gene in many cancers.This assumption is supported by the study of Siprashvilli et al. (1997), which has showed that transfection of FHIT cDNA in cancer cell lines from gastric, large-cell lung, nasopharyngeal and renal cell cancer significantly suppresses tumorigenicity in nude mice.However, in another study, Otterson et al. (1998) demonstrated that introduction of wild-type FHIT cDNA into a cervical carcinoma cell line lacking endogenous Fhit expression did not change the rate of cell proliferation or altered tumorigenicity in animals.Those contradictory results may suggest the possibility that FHIT may exert a tumor suppressor activity in a cell lineage-specific manner or that the FHIT gene is indirectly linked with tumor growth.In breast cancer abnormal transcripts of the FHIT gene have been observed with various frequency depending on size and histological grade (Hayashi et al., 1997;Pandis et al., 1997;Campiglio et al., 1999).However, it has been reported that clinicopathological analysis of 61 Japanese primary breast cancer specimens revealed no correlation between abnormal FHIT transcripts and tumor-node metastasis classification, tumor size, estrogen receptor and progesterone receptor status (Hayashi et al., 1997).
Studies on oncogene amplification in breast cancer have shown prognostic significance of c-erbB2 and c-myc amplification (Champeme et al., 1994;Berns et al., 1995;Deming et al., 2000).It has been observed that c-myc amplification can occur at an early stage of tumor formation and it is not often detectable in nodal metastasis (Watson et al., 1993).In most re-ports c-erbB2 amplification is negatively correlated with estrogen receptor and progesterone receptor status, whereas c-myc amplification is more prevalent in the steroid receptor-positive subpopulation.Overall, breast cancer patients with c-myc or c-erbB2 amplification in their primary tumors have a shorter relapse-free survival and reduced survival (Lonn et al., 1995).In order to evaluate the clinical relevance of FHIT gene alterations in breast cancer, we have analyzed FHIT transcripts and the amplification of c-erbB2 and c-myc in primary breast cancer.

Patients and collection of tissues.
Seventy-nine patients with primary breast cancer who attended the Department of Surgical Oncology at the Medical University of Gdañsk (in 1997Gdañsk (in -1998) ) were randomly selected for the study.Patients had either an excision biopsy or modified radical mastectomy.All specimens were measured and serially sectioned.All tissues were reviewed by the same pathologist (A.K.).Each specimen was dissected in a 5 mm bread-loaf fashion and measured in all three depicted dimensions.The largest gross dimension was included in the computer analysis.Infiltrating cancers were considered ductal or lobular or tubular based upon their predominant cell type.Grading of invasive tumors followed the modified Bloom and Richardson method (1957).Quantitative estrogen and progesterone receptor analysis (immunohistochemical) was performed on fresh-frozen tissue from the original specimen.Samples of cancerous tissues were dissected by pathomorphologists immediately after surgery, frozen in liquid nitrogen and stored at -80°C.RNA extraction and reverse transcription.Total RNA was extracted from tissues frozen in liquid nitrogen and stored at -80°C using Total RNA Prep Plus Kit (A&A Biotechnology, Gdañsk, Poland).RNA was stored as a pellet under ethanol at -40°C.Reverse transcription was performed in 20 mL final volume of 50 mM Tris/HCl, pH 8.3, 75 mM KCl, 3 mM MgCl 2 , 10 mM dithiothreitol, 1 mM dNTPs, 250 ng oligo(dT) (Gibco BRL, Paisley, England), 12.5 U of MMLV-RT (Epicentre Technologies, Madison, WI, U.S.A.), 10 U of RNasin (Promega, Madison, WI, U.S.A.), and 1 mg of RNA.The reaction was incubated for 10 min at 65°C, then run for 90 min at 42°C and boiled for 5 min.In order to control the integrity of the obtained cDNA, a 511-bp fragment of b-actin cDNA was amplified with primers as described (Negrini et al., 1996).The PCR consisted of an initial denaturation at 95°C for 3 min and 35 cycles of 30 s at 95°C, 30 s at 53°C, 1 min at 72°C, and a final extension of 10 min at 72°C.The PCR reaction was performed in a Perkin-Elmer thermal cycler 480.
Nested PCR.To assess the status of the FHIT gene transcript we used the nested PCR reaction reported by Ohta et al. (1996).The primers used were: UR4, (forward); 7D, (reverse) for the first round and UR5, (forward); 7B, (reverse) for the second round of amplification (Druck et al., 1997).DNA extraction.DNA extraction from the tissues examined was carried out using Genomic DNA Prep Plus Kit (A&A Biotechnology).The DNA content was measured by light absorption at 260 nm.The purity of DNA was assessed based on the calculated ratio A 260 /A 280 .
Amplification of c-myc and c-erbB2.Amplification of c-myc and c-erbB2 was assessed by a semi-quantitative multiplex PCR assay, which in our hands proved to be useful in evaluating changes in template copy number (Sakowicz et al., 2001;Chrzan et al., 2001).The reaction mixture contained 50 mM Tris/HCl, pH 9.0, 20 mM ammonium sulfate, 100-500 ng of template, 0.50 mM each of 5¢ and 3¢ primers, 0.25 mM of each dNTP, 2.5 mM MgCl 2 and 1 U of Tfl DNA polymerase (Epicentre Technologies).The PCR reaction was performed for 30 cycles of 94°C (1 min), 65°C (1 min), 72°C (2 min) and 94°C (1 min), 56°C (1 min), 72°C (2 min) for c-myc and c-erbB2 amplification, respectively.For c-myc amplification 5¢-CTCGAATTCCTTCCAGATA-TCCTCGCTG-3¢ and 5¢-CACTGCGCGCTGC-GCCAGGTTT-3¢ primers were used.This defines a DNA fragment of 258 bp.The c-myc primers were based on the nucleic acid sequence of the gene (GeneBank, accession No. J00120).Amplification of c-erbB2 was performed with the primers described by Lonn et al. (1995).The PCR products were separated by agarose gel electrophoresis and the ethidium bromide-stained bands were quantified with the use of a Gel Doc 2000 system (Bio-Rad) and compared using computer program Quantity One (Bio-Rad).To produce a titration curve of gene amplification we used the cell lines MCF7 and SKBR3 with a defined c-myc and c-erbB2 gene copy number, respectively.Tissue plasminogen activator (TPA) and thymidine kinase (TK) were the reference genes for c-myc and c-erbB2 amplification, respectively.The primers for TPA and TK amplification were as described by Lonn et al. (1995).

Statistical analysis.
The analysis of statistical significance of the correlations between clinicopathological parameters and FHIT expression, c-myc and c-erbB2 amplification was performed using Pearson's chi-square test.

RESULTS
The status of the FHIT gene transcript was analyzed in 79 breast cancers by performing RT-PCR.A PCR product of 763 bp, corresponding to the fragment of the FHIT gene transcript encompassing exons 2 to 9, was observed in 58% of cancerous tissues (Table I).In addition to the normal-sized product, aberrant size products as well as no products were observed (Fig. 1).The overall occurrence of an aberrant FHIT transcript (no transcripts plus abnormal-sized transcripts) was 40.5% (Table 1).The 511 bp fragment of b-actin transcript was successfully amplified in all ana-lyzed tissues (not shown), suggesting a good quality of the cDNA template.Thus, it could be assumed that the observed absence of a PCR product for the FHIT gene transcript indicates a real loss of Fhit mRNA expression in these tissues.This assumption was confirmed by Western blots performed on tissue extracts (Fig. 2).The Fhit protein was detected in all cancerous tissues that exhibited the normal FHIT transcript.On the other hand, the Fhit protein could not be detected in any of the tissues lacking the FHIT gene transcript (compare Fig. 1 and Fig. 2).In all cases in which, besides the normal also the aberrant FHIT transcript was present, the Western blot showed only a normal-sized band.This indicates that abnormal FHIT transcripts observed in our study did not encode any Fhit protein or the amount of such a protein was very low.
In order to assess the role of FHIT alterations in the prognosis of patients with breast 344 R. Kowara and others 2002 In our studies we also analyzed the dependence of FHIT transcript status on the histopathological type of tumor.We found that the abnormalities in FHIT gene transcripts occurred more frequently in lobular cancers (60%) than in ductal cancers (28%) (Table 2).The same high occurrence of abnormal FHIT transcripts was observed in tubular cancers (60%), but we had no sufficient number of patients within this group to perform statistical analysis.
Data from studies on oncogene amplification indicate that in breast cancer amplification of c-erbB2 or c-myc has prognostic significance (Champeme et al., 1994;Berns et al., 1995;Deming et al., 2000).In order to evaluate the clinical relevance of FHIT gene alterations in breast cancer, we analyzed the FHIT transcripts and the amplification of c-erbB2 and c-myc.To determine the gene copy number of the c-erbB2 and c-myc genes we used multiplex PCR.In the same tube two fragments of DNA were amplified, one representing the studied gene and the other a control gene.As a control gene for c-myc we chose the gene for tissue plasminogen activator (TPA).Both genes are located on chromosome 8.The gene for thymidine kinase (TK) located on 17q23.2 was the control gene for c-erbB2, placed on the same chromosome.Representative results of multiplex PCR are presented in Fig. 3.We found amplification of the c-myc  gene in 16% of the tumors studied.Amplification of the c-erbB2 oncogene was observed in 21% of cancerous tissues.The percentage of c-myc and c-erbB2 amplification was of the same magnitude as reported by others in breast cancer.
Analyzing the occurrence of c-myc and c-erbB2 amplification and alteration of FHIT gene transcripts we found that normal FHIT transcript occurred more frequently (43 of 65) in cancers with the normal number of c-myc gene copies (Table 3).On the other hand, aberrant FHIT transcripts were found predominantly (64%) in cancers with c-myc amplification.Since the aberrant FHIT transcripts were frequently accompanied by c-myc amplification, we analyzed relation of the coexisting abnormalities to some clinical and epidemiological prognostic factors.We found that cancers having an aberrant FHIT gene transcript with accompanying c-myc amplification occurred more frequently (P < 0.05) in pa-tients with positive lymph node status (Table 4).

DISCUSSION
In breast cancer as well as in carcinomas of the digestive tract in the most frequently observed aberrant FHIT transcripts exon 8 is missing (Ohta et al., 1996;Negrini et al., 1996).Exon 8 contains the histidine triad domain, which is essential for the anticancer activity of Fhit protein (Siprashvili et al., 1997).Thus it is possible that in tumor cells the Fhit protein is inactivated by the appearance of FHIT gene transcripts lacking coding domain sequences essential for Fhit activity.However, we were unable to detect any abnormal-sized Fhit protein in tissues that displayed, besides the normal, also an aberrant FHIT transcript.This indicates that the abnormal transcripts do not encode any Fhit The presence of any abnormal transcript with or without accompanying normal-sized transcript was counted as "aberrant".*P < 0.05.protein or the amount of such a protein is very low.The simultaneous presence of normal and aberrant FHIT transcripts within the same tumor cells has been reported (Negrini et al., 1996;Guo et al., 2000).However, data from studies on various cancer cell lines indicate that the cells in which normal and aberrant FHIT transcripts are present express only the normal-sized Fhit protein but not the truncated form (Druck et al., 1997;Otterson et al., 1998).On the other hand, the expression level of Fhit is reduced when this protein is translated from transcripts lacking the non-coding exons 3 or 4 (Sozzi et al., 1997).In our studies we observed a reduced level of the Fhit protein in some cancers displaying both normal and aberrant FHIT transcripts and in cancers with the normal FHIT transcript only (not shown).Since primary cancers are polyclonal the interpretation of these observations is difficult.In several types of cancer a strong association of impaired Fhit protein expression with the disruption of FHIT transcript has been observed (Baffa et al., 1998;Guo et al., 2000;Huiping et al., 2000;Yoshino et al., 2000).It has been postulated that in some cancers, including breast cancer, alterations in the FHIT gene represent an early event in carcinogenesis (Baffa et al., 1998;Zou et al., 1999;Guo et al., 2000;Huiping et al., 2000), whereas in other cancers aberrant FHIT transcripts occur at later stages of cancer development (Yoshino et al., 2000).
In order to assess the clinical relevance of FHIT transcript alteration in breast cancer, we have analyzed some clinical and epidemiological characteristics of patients having normal and aberrant FHIT transcripts.There was no association of FHIT transcript alteration with age, tumor size, estrogen and progesterone receptor status, local metastases and histological grading.Similar results have been reported by Hayashi et al. (1997) who analyzed a group of 61 Japanese patients.However, these investigators have found an association of the occurrence of aberrant FHIT transcripts with bilateral breast cancer and experience of childbirth.Since in our group of patients there were no bilateral breast cancer cases we could not make such an analysis.In A B our study we did not observe any correlation of FHIT transcript aberration with childbirth experience (not shown).On the other hand, we did observe a relation between the presence of aberrant FHIT transcripts and the histopathological type of tumor.The aberrant FHIT transcripts were more frequently observed in lobular cancers (60%) than in ductal cancers (28%), P < 0.05 (Table 2).Since ductal cancers are less differentiated than lobular cancers it could be assumed that abnormalities in FHIT mRNA depend on the stage of tumor cells differentiation.These tumors would be regarded as having a relatively good prognosis on the basis of conventional clinicopathological diagnosis, although lobular cancers have uncertain malignant potential.
Numerous studies on oncogene amplification in breast cancer have shown prognostic significance of c-erbB2 and c-myc amplification.The frequency of amplification of both oncogenes observed by us in the breast cancers examined was comparable to that reported by others (Borg et al., 1992;Berns et al., 1992;Chen et al., 1995;Mark et al., 1999;Deming et al., 2000).We did not see any rela-tions between c-erbB2 amplification and aberration of FHIT gene status.On the other hand, our study revealed a more frequent incidence of abnormal FHIT gene transcript occurrence in breast cancer displaying amplification of c-myc.The observed association of abnormalities in FHIT transcript with c-myc amplification would indicate that in breast cancer c-myc is somehow involved in inactivation of the FHIT gene.Since FHIT is regarded as a tumor suppressor gene, the observed association of aberrant FHIT gene transcript occurrence with c-myc amplification suggests that abnormal expression of c-myc may facilitate progression of breast neoplasia.
To date axillary lymph node status is the best prognostic indicator in breast cancer.Numerous clinical trials have established that involvement of axillary lymph nodes is related to high recurrence and early death rate.We found that aberrant FHIT transcript with accompanying c-myc amplification occurred more frequently in breast cancers with positive lymph node status (Table 4).This may indicate that alteration of FHIT transcript with accompanying c-myc amplification could have Tumor size (mm) 20.4 ± 2.8 25.5 ± 7.9 Age (years) 62.1 ± 10.9 58.1 ± 13.1 a predictive value in breast cancer.However, to determine whether FHIT transcript aberration has a prognostic utility in breast cancer, extended clinical follow up will be necessary.

Figure 1 .
Figure 1.RT-PCR analysis of FHIT gene transcript in breast cancer tissues.PCR products were separated on 2% agarose gel.Arrow indicates the normal-sized product of the FHIT gene (763 bp).Lane M, DNA molecular size markers; lane 0, product of control PCR performed without the reverse transcriptase product.Numbers above the lanes indicate the patient's number.

Figure 2 .
Figure 2. Immunoblot analysis of samples derived from breast cancer tissues.Tissue extracts (100 ng of protein) were subjected to SDS/PAGE (12%) and transferred to an Immobilon transfer membrane.The membranes were immunobloted with anti-Fhit polyclonal antibodies.On lanes St recombinant Fhit protein (10 ng) was loaded.Numbers above the lanes indicate the patient's number.For FHIT transcript status in these tissues see Fig. 1.
of c-myc and c-erbB2 amplification with the status of FHIT gene transcript in cancerous tissues.The presence of any abnormal transcript with or without accompanying normal-sized transcript was counted as "aberrant".*P < 0.05.

Figure 3 .
Figure 3. Analysis of c-myc and c-erbB2 amplification in breast cancer.A: PCR analysis of c-myc amplification with the gene for tissue plasminogen activator (TPA) as a reference gene.B: PCR analysis of c-erbB2 amplification with the gene for thymidine kinase (TK) as a reference gene.Lane M, DNA molecular size markers; lane 0, product of control PCR performed without the template; lane W, PCR product with DNA isolated from a healthy person; lanes MC and SK, products of PCR performed with DNA isolated from cells with defined c-myc (MCF 7) and c-erbB2 (SKBR 3) amplification, respectively.Numbers above the lanes indicate the patient's number.

Table 4 .
Comparison of c-myc amplification with FHIT gene transcript status in breast cancer.Relation with clinical and epidemiological characteristics of patients.*P < 0.05.

Table 1 . Comparison of FHIT gene transcript status in breast cancer tissues with some clinical and epidemiological characteristics of patients.
cancer we analyzed the relation between the FHIT gene transcript status and other prognostic factors.Clinical and epidemiological characteristics of patients with normal and aberrant FHIT transcripts are summarized in Table1.There was no association of the presence of the aberrant FHIT gene transcript with age, tumor size, estrogen and progesterone receptor status, local metastases and histological grading.