SEARCH

SEARCH BY CITATION

Keywords:

  • hepatocarcinogenesis;
  • proliferation;
  • immunohistochemistry;
  • cell cycle

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

BACKGROUND

Increased expression of the proto-oncogene c-myc is a common phenomenon in hepatocellular carcinoma (HCC). The proto-oncogenes c-fos and c-jun are involved in cell cycle progression and cellular proliferation.

METHODS

The objective of this study was to elucidate the mechanism of hepatocarcinogenesis with regard to the expressions of c-myc, c-fos, and c-jun. One hundred fifty biopsied HCC specimens were stained immunohistochemically for the above phenotypic markers both in tumor tissue and in adjacent nontumor tissue.

RESULTS

Although the expression of c-myc was high (74%) in tumor tissue, it was significantly less compared with the expression in nontumor tissue (100%; P = 0.0002). The expression of c-myc was inversely proportional to the grade of differentiation in tumor tissue (P = 0.0108; correlation coefficient [r] = −0.244); that is, tissue with poorer histologic differentiation had a lower level of c-myc expression. There were inverse associations between the expression of c-myc and the expression of mutated p53 (P = 0.0017; r = −0.285) as well as the expression of Ki67 (P = 0.057; r = −0.147). There was significantly high expression of c-fos in tumor tissue compared with the expression in nontumor tissue (91% vs. 0%; P < 0.0001). Both the tumor tissue and the nontumor tissue had high levels of expression of c-jun (96.53% and 100%, respectively). There was a trend toward a positive association between the expression of c-fos and the expression of c-jun in tumor tissue (P = 0.07; r = 0.162).

CONCLUSIONS

Because c-myc is a known inducer of wild type p53, decreased c-myc expression may lead to uncontrolled cell growth because of the lack of p53 expression that normally induces apoptosis. The coordinated expression of c-fos and c-jun in HCC may reflect the coordinated tumor cell cycle of progression and proliferation; however, future studies are required to elucidate this possibility. Cancer 2001;91:106–12. © 2001 American Cancer Society.

Hepatocellular carcinoma (HCC) is one of the most common cancers in the world.1, 2 It causes significant mortality, especially in countries where there is a high prevalence of chronic hepatitis B virus (HBV) and hepatitis C virus infection. Various proto-oncogenes, tumor suppressor genes, growth factor genes, and virologic factors have been implicated in hepatocarcinogenesis.3, 4 The mechanism of hepatocarcinogenesis probably is multifactorial, involving multiple oncogenes.5 However, the sequence of events in the process of hepatocarcinogenesis has not been characterized fully.

The proto-oncogene c-myc stimulates a pattern of cellular gene expression by its regulatory elements.4 Its encoded proteins (62,000-dalton, DNA-binding, transcriptional regulatory proteins) are involved in gene expression during cell growth and differentiation.6 It has been shown to be important in the process of hepatocarcinogenesis in woodchucks,7 and increased expression of c-myc after the integration of the woodchuck hepatitis virus DNA has been documented.8 However, its role in human HCC remains unclear, although the over-expression of c-myc is a common phenomenon.9, 10

The oncogene c-fos, which is required for quiescent cells to enter the cell cycle,11 also is up-regulated in HCC.12 HBV X peptide has been shown to activate the c-fos gene, which is postulated to contribute to hepatocarcinogenesis.4 HBV X peptide also has been reported to activate the expression of c-jun.13

The objective of this study was to evaluate the expression of these oncogenes in a large cohort of Chinese patients with HCC. In patients with nontumor tissue in the biopsied tissue block, the expression of these genes in nontumor tissue also was examined.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

Patients

We previously reported the expression of phenotypic differentiation markers in 290 Chinese patients with HCC.14 One hundred fifty liver specimens from that study were available for the current study. There were no differences in the distribution of gender, age, survival, or other histologic features between these 150 patients and the data base of the original 290 cases. All of the patients studied were seen at Queen Mary Hospital, Hong Kong, during the period of 1986 and 1993. All liver specimens were fixed with buffered formalin and embedded in paraffin. The clinical and histologic diagnosis of HCC in all patients fulfilled the diagnostic criteria recommended by the World Health Organization.15 Sera from these patients were tested for hepatitis B surface antigen (HBsAg) by using an enzyme-linked immunosorbent assay (Abbott Laboratories, Chicago, IL). The survival data were analyzed only for patients who had not received any form of medical or surgical anticancer treatment. Survival time was defined as the time from the diagnostic liver biopsy to the time of death.

Histologic Assessment

Tissue sections that showed features of fibrolamellar type HCC were excluded. Cholangiocarcinoma also was excluded by performing hematoxylin and eosin staining together with reticulin and mucin staining by Gordon and Sweet silver and mucicarmine, respectively. For those sections with adequate surrounding tissue for assessment, the presence or absence of cirrhosis was noted. The histologic differentiation was graded from 1 to 3 according to whether the tumors were well differentiated, moderately differentiated, or poorly differentiated, respectively.16 Cellular and nuclear pleomorphisms as well as mitotic figures were assessed and graded in order of severity from 0 to 4.

Immunohistochemistry

Immunohistochemical staining was performed on 5-μm-thick, formalin fixed, paraffin embedded tissue sections based on the standard, indirect, streptavidin-biotin horseradish peroxidase method (Amersham, Chicago, IL) using diaminobenzidine (Sigma Chemical Co., St. Louis, MO) as a substrate. Three proto-oncogenes (c-myc, c-fos, and c-jun) were traced. Monoclonal primary antibodies against these gene products were used (anti-c-myc: dilution, 1:100, Novocastra Laboratories, Newcastle, United Kingdom; anti-c-fos: dilution, 1:20; Oncogene Science, Cambridge, MA; anti-c-jun: dilution, 1:500; Oncogene Science). Bcl-2 also was traced (anti-bcl-2: dilution, 1:40; Dako Corporation, Carpinteria, CA). All antibodies were diluted in Tris-buffered saline, pH 7.4, supplemented with 1% bovine serum albumin (BSA) (Sigma Chemical Co.).

Briefly, the tissue sections were deparaffinized with xylene and rehydrated through serial graded alcohol solutions. Nonspecific adsorption was blocked by 10% normal serum from the second-layer antibody species in 3% BSA for c-myc, c-jun, and bcl-2. For c-fos, the blocking scheme recommended by the manufacturer was used (30 minutes in 0.1 M TNB buffer, pH 7.5; Tris HCl; and 0.15 M NaCl with 0.5% Dupoint blocking reagent; Tyramide Signal Amplication Kit; New Life Science, Boston, MA). Primary antibodies in appropriate dilutions were added and incubated for 1 hour at 37 °C (for c-myc) or overnight at 4 °C (for c-fos, c-jun, and bcl-2). Secondary antibodies were applied according to the manufacturer's recommendations (Amersham). Biotin peroxidase-conjugated streptavidin was used as the tracer. Diaminobenzidine was used as the substrate. Tyramide signal amplification was used for c-fos tracing (New Life Science). The slides were counterstained with hematoxylin before mounting in Permount.

Nonliver tissues and tumors that are known to express the oncogene products described above were used as positive controls, and an unrelated primary antibody was used as a negative control. The methods for the detection of two biliary cell phenotypic markers (AE1–AE3 and cytokeratin 19 [CK19]), proliferative markers (proliferative cell nuclear antigen [PCNA] and, Ki67), the gene for regulating apoptosis (p53), and α-fetoprotein (AFP) were described in our previous study.14 Detection of the expression of p53 signified mutated p53 protein, which had a long half-life that enabled it to be detected by monoclonal antibody.17

All sections that were stained with the markers described above were graded from 0 to 4 according to the following assessment: 0, no positive cells; 1, < 1 –24% positive cells; 2, 25–49% positive cells; 3, 50–74% positive cells; and 4, ≥ 75% positive cells. The grading was assessed by two reviewers (P.-C.W. and J.Y.-N.L.) independently and blindly.

Statistical Analysis

The association of two categoric variables was tested by chi-square test or Fisher exact test (two-tailed). A Mann–Whitney nonparametric test was used for the comparison of variables with skewed distribution. The correlation between nominal, ordered variables (i.e., graded expression of phenotypic markers) was tested by Mantel–Haenszel tests for linear association. A Kendall correlation was used to determine the association of ordinal data and nominal ordered data. A comparison of the natural survival time of patients with and without c-myc expression was performed by Kaplan–Meier survival analysis.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

The clinical and pathologic features of the studied patients are summarized in Table 1. The majority of the patients were HBsAg carriers (86.3%) and had clinical or histologic evidence of cirrhosis (77.17%). The characteristics of the patients in this study were similar to those reported previously in the Chinese population.18, 19 The expression levels of the various phenotypic markers are listed in Table 2.

Table 1. The Clinical and Pathologic Features of the Patients in This Study
CharacteristicsData
  1. M: male; F: female; HBsAg: hepatitis B surface antigen.

No. of patients150
Median age in yrs (range)56 (14–88)
Gender (M:F)128:22
Median survival in weeks (range) (n = 62)8 (1–212)
No. of patients with HBsAg positive serum (%)63 of 73 (86.3)
Histologic evidence of cirrhosis (%)71 of 92 (77.17)
Tumor differentiation-well:moderate:poor16:75:39
Median no. of mitotic figures (range)2 (0–4)
Median no. of cellular pleomorphism (range)2 (1–4)
Table 2. The Expressions of Various Phenotypic Markers of HCCs in This Study
Phenotypic markerPositive (%)Median grade of expression (range)
c-myc
 Tumor tissue741+ (0–4+)
 Nontumor tissue1003+ (1–4+)
c-fos
 Tumor tissue912+ (0–4+)
 Nontumor tissue0
c-jun
 Tumor tissue972+ (0–4+)
 Nontumor tissue1002+ (1–4+)

Expression of c-myc, c-fos, and c-jun in Tumor and Nontumor Tissues

The immunohistochemical staining for c-myc, c-fos and c-jun is illustrated in Figure 1. Expression of c-myc was detected in 74% of the tumor tissue samples (median, Grade 1+; range, Grade 0–4+). In 44 samples in which nontumor tissue was present in the section for assessment, all 44 samples (100%) had a high level of c-myc expression (median, Grade 3+; range, Grade 1–4+). The overall expression of c-myc in nontumor tissue was significantly greater than that in tumor tissue (P = 0.0002). If the analysis is performed based on the 44 samples with both tumor tissue and nontumor tissue in the same section to avoid nonpairing assessment bias, then the difference in c-myc expression in tumor tissue was 68% compared with 100% in nontumor tissue (P < 0.0001).

With the sensitive detection assay for c-fos, expression of c-fos in tumor tissue was detected in 91% (median, 2+; range, 0–4+) compared with 0% in nontumor tissue (n = 44 samples; P < 0.0001). There was no association between the expression of c-myc and c-fos in the tumor tissue (P = not significant [NS]).

thumbnail image

Figure 1. Sections demonstrating the expression of c-myc (cytoplasmic distribution) (a) (original magnification, ×200), c-fos (nuclear distribution) (b) (original magnification, ×100), and c-jun (nuclear distribution) (c) (original magnification, ×200) in hepatocellular carcinoma. Positive signals were traced immunohistochemically by using diminobenzadine, which appeared brown. The sections were counterstained with hematoxylin.

Download figure to PowerPoint

c-Jun was detected at high levels in both tumor tissue (96.53% of samples; median, Grade 2+; range, Grade 0–4+) and nontumor tissue (100% of samples; median, Grade 2+; range Grade 1–4+; P = NS). There was no association between the expression of c-myc and c-jun in tumor tissue (P = NS). However, there was a borderline positive association between the expression of c-fos and c-jun in tumor tissue (correlation coefficient [r] = 0.162; P = 0.07).

Significance of c-myc Expression in HCC Tumor Tissue

The expression of c-myc was not associated with the clinical demographics (gender, age), clinical or histologic evidence of cirrhosis, serum HBsAg positivity, or patient survival (for patients who did not received any treatment). There was no association between the expression of c-myc in tumor tissue and the number of mitotic figures and the cellular pleomorphisms in the HCC samples (P = NS). However, the expression of c-myc was inversely proportional to the grade of differentiation of the HCC samples (P = 0.0108; r = −0.244), that is, poorer histologic differentiation of HCC had a lower level of c-myc expression.

There was no association between the expressions of c-myc and AFP (91% of samples positive; median, Grade 2+; range, Grade 0–4+; P = NS). The expression of c-myc was associated negatively with the expression of p53 (43% of samples positive; median, Grade 0; range, Grade 0–4+; P = 0.0017; r = −0.285). There was also a borderline negative association between the expression of c-myc and the expression of Ki67 (95% of samples positive; median, Grade 2+, range, Grade 0–4+; P = 0.057; r = −0.147). However, no association was observed between the expression of c-myc and PCNA (98% of samples positive; median, Grade 2+; range, Grade 1–4+) in the tumor tissue. There was also no association between the expression of c-myc and bcl-2 (11% of samples positive; median, Grade 0+; range, Grade 0–1+) or biliary markers (AE1/AE3: 33% of samples positive; median, Grade 0; range, Grade 0–3+; CK19: 29% of samples positive; median, Grade 0; range, Grades 0–3+) in the tumor tissue.

Significance of c-fos and c-jun Expression in HCC Tumor Tissue

The expression levels of c-fos and c-jun were not associated with the clinical demographics (gender, age), clinical or histologic evidence of cirrhosis, serum HBsAg positivity, or patient survival. They also were not associated with the histologic features of HCC (the number of mitotic figures, cellular pleomorphisms, cellular differentiation) (P = NS). The expression levels of the various phenotypic markers (AFP, p53, bcl-2, Ki67, PCNA, AE1/AE3, and CK19) were not found to be associated with the expressions of c-fos and c-jun respectively (P = NS for all).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

This study demonstrated a few important points. First, c-myc was expressed at high levels in nontumor liver tissue adjacent to HCC tissue. Second, the expression of c-myc in HCC tissue was reduced compared with the adjacent tissue, and the reduction was more marked with poorer differentiation of the tumor and greater expression of mutated p53. Third, c-fos was detected in HCC tissue and not in tumor tissue based on a highly sensitive detection assay. Fourth, c-jun was detected at high levels in both HCC tissue and nontumor tissue. Finally, we were able to show a trend for the coordinated expression of c-fos and c-jun in HCC tissue.

Previous studies have shown that there is no enhancement of expression of c-myc in normal liver tissue.20, 21 Other studies also have shown that c-myc expression is increased in HCC tissue and may participate in hepatocarcinogenesis.10, 12, 22, 23 Kawate et al.,10 by using the differential polymerase chain reaction method, showed that 14 of 42 HCC patients (33.3%) had c-myc amplification that was associated with less differentiated tumors and shorter survival. However, the c-myc expression was not studied in the adjacent nontumor tissues. The current study showed three interesting observations concerning c-myc. There was a high level of expression of c-myc in adjacent nontumor tissue. However, c-myc expression in HCC samples, although it was increased, as reported by Kawate et al., was less than that in the adjacent nontumor tissue and was related inversely to the differentiation of tumor cells. The decrease in c-myc expression also was associated with an increase in expression of mutated p53 in the tumor cells. It has been shown that the expression of the proliferation markers Ki67 and PCNA is increased in cirrhotic tissue, possibly related to liver cell regeneration.24 Also, c-myc is a known inducer of wild type p53 and may serve as one of the check points to ensure coordinated cellular proliferation.25–27 The current findings suggest that this check point may not be functional as part of the cellular control mechanism during the evolution of hepatocarcinogenesis because of the decrease in c-myc expression in the tumor cells. Conversely, the apoptosis of tumor cells was jeopardized further by the increased expression of mutated p53 accompanying the decreased expression of c-myc.

The short half-life of p53 and the high incidence of p53 mutations preclude the demonstration of the relations between c-myc and wide type p53. Also, mcl-1, a member of the bcl-2 family, is believed to be a regulator of the effect of c-myc on p53.28, 29 It will be interesting to determine the expression of mcl-1 in HCC when such reagents are available. The observation that c-myc still was detected in the majority of HCC samples, albeit at lower levels, suggests that the reduced expression of c-myc may be secondary to the dedifferentiation of tumor cells, as evidenced by the association between the poorer differentiation of the HCC and the decreased expression of c-myc in this study. Therefore, it seems that reduced c-myc expression is not the primary trigger for tumor cell transformation. Our findings are in accord with the findings of Su et al.30 The lack of a correlation between c-myc and c-fos/c-jun in tumor tissue suggests that c-myc is not playing a role in triggering the AP1 pathway in tumor cell proliferation.

It is noteworthy that similar patterns of expression (increased expression in adjacent nontumor tissue but decreased in tumor tissue) also have been reported in the expression of the ras oncogene.23, 31 The expression of c-ras is believed to be activated by growth factors through a series of mitogen-activated protein kinases (i.e., Mos, Raf, Tpl2, MEK1, MEK2, ERK1, and ERK2). This may induce the expression of c-myc,32, 33 which may serve as one of the check points for uncontrolled growth. The parallel observation for c-ras and c-myc suggests that, although growth factors triggering cellular growth may play a role in the process of malignant transformation (as reflected by the high expression levels of c-ras and c-myc in adjacent nontumor tissue), they may not play a significant role in sustaining the growth of the tumor cells (as reflected by the reduced expression levels of c-ras and c-myc).

The results of this study suggest that c-fos and c-jun were expressed in HCC tissue in a coordinated manner. Also, c-fos and c-jun form part of the AP1 complex and are involved in cell cycle progression and cellular proliferation.34, 35 c-Fos is detected only with tyramide amplification, which may either reflect a technical issue (altered epitope for detection in formalin fixed tissue or poor affinity of the antibody tracer) or a low level of c-fos expression. However, the detection of c-fos in HCC tissue compared with the lack of detection of c-fos in adjacent nontumor tissue suggests that c-fos expression is definitely greater in HCC tissue compared with nontumor tissue.

A technical limitation of immunohistochemistry on formalin fixed, paraffin embedded tissue may also affect the study of c-jun expression. The expression level of c-jun is high in both nontumor tissue and tumor tissue. The expression levels of c-fos and c-jun usually are coregulated.34, 35 However, c-jun can be involved in both cell cycle progression/proliferation and apoptosis.36 If the uncoordinated expression of c-fos and c-jun in the adjacent nontumor is confirmed in the future through molecular studies, then the increased expression of c-jun may reflect more of the cell growth checkpoint through apoptosis, similar to c-myc. The coordinated expression of c-fos and c-jun in tumor tissue may reflect coordinated tumor cell cycle progression and proliferation, a feature of HCC (rapid tumor growth), but it remains to be elucidated by further studies.

The strength of the current study is the attempt to evaluate the role of these proto-oncogene products in HCC. The weakness of this study is the fact that we were studying patients who already had developed disease at a single point in time. In addition, the technical limitations of immunohistochemistry also limited the strength of our interpretation. Nevertheless, this study provided some possible new insights on hepatocarcinogenesis and a number of new questions. Further studies of the events that trigger c-myc, c-fos, and c-jun expression through molecular studies and the functional role of these proteins in various stages of hepatocarcinogenesis may help to define further the molecular events that lead to the development of HCC.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

The uncoordinated control of cellular growth as a result of decreased apoptosis secondary to the relative diminished expression of c-myc in tumor cells may be one of the contributory mechanisms of hepatocarcinogenesis. The role of increased expressions of c-fos and c-jun in the process of hepatocarcinogenesis remains to be elucidated in future studies.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

The authors thank the staff of the Hong Kong Births and Deaths General Registry Office for verifying the dates of death of some of the patients studied.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES