Increased expression of stathmin and elongation factor 1α in precancerous nodules with telomere dysfunction in hepatitis B viral cirrhotic patients

Background Telomere dysfunction is important in carcinogenesis, and recently, stathmin and elongation factor 1α (EF1α) were reported to be up-regulated in telomere dysfunctional mice. Methods In the present study, the expression levels of stathmin and EF1α in relation to telomere length, telomere dysfunction-induced foci (TIF), γ-H2AX, and p21WAF1/CIP1 expression were assessed in specimens of hepatitis B virus (HBV)-related multistep hepatocarcinogenesis, including 13 liver cirrhosis specimens, 14 low-grade dysplastic nodules (DN), 17 high-grade DNs, and 14 hepatocellular carcinomas (HCC). Five normal liver specimens were used as controls. TIF were analyzed by telomere fluorescent in situ hybridization (FISH) combined with immunostaining, while the protein expressions of stathmin, EF1α, γ-H2AX, and p21WAF1/CIP1 were detected by immunohistochemistry. Result The expressions of stathmin and EF1α gradually increased as multistep hepatocarcinogenesis progressed, showing the highest levels in HCC. Stathmin mRNA levels were higher in high-grade DNs than normal liver and liver cirrhosis, whereas EF1α mRNA expression did not show such a difference. The protein expressions of stathmin and EF1α were found in DNs of precancerous lesions, whereas they were absent or present at very low levels in normal liver and liver cirrhosis. Stathmin histoscores were higher in high-grade DNs and low-grade DNs than in normal liver (all, P < 0.05). EF1α histoscores were higher in high-grade DNs than in normal liver and liver cirrhosis (all, P < 0.05). Stathmin mRNA levels and histoscores, as well as EF1α histoscores (but not mRNA levels), were positively correlated with telomere shortening and γ-H2AX labeling index (all, P < 0.05). EF1α histoscores were also positively correlated with TIF (P < 0.001). Significantly greater inactivation of p21WAF1/CIP1 was observed in low-grade DNs, high-grade DNs, and HCC, compared to liver cirrhosis (all, P < 0.05). p21WAF1/CIP1 labeling index was inversely correlated with TIF, stathmin mRNA level, and EF1α histoscore (all, P < 0.05). Conclusion Stathmin and EF1α are suggested to be closely related to telomere dysfunction, DNA damage, and inactivation of p21WAF1/CIP1 in HBV-related multistep hepatocarcinogenesis. Accordingly, assessment of stathmin and EF1α levels as a reflection of telomere dysfunction may be helpful in evaluating the biological characteristics of precancerous hepatic nodules in hepatitis B viral cirrhotic patients.


Background
Hepatocellular carcinoma (HCC) is reported as the fifth most common cancer worldwide and the third most common cause of cancer-related deaths [1]. Hepatocarcinogenesis is a multistep process, and has been characterized as the progressive accumulation of genetic and molecular changes in chronic liver disease, which leads to the production of monoclonal populations of dysplastic or transformed hepatocytes. Mutations that impair the DNA damage response pathways facilitate the survival of dysplastic cells with shortened and dysfunctional telomeres, leading to the proliferation of cells with dicentric chromosomes and the accumulation of genomic instability due to breakage-fusion-bridge cycles. In fact, chromosomal instability is characteristic of hepatitis B virus (HBV)-related hepatocarcinogenesis [2]. Liver cirrhosis, a soil for HCC, is characterized by an increase in dysplastic hepatocytes that form dysplastic foci or dysplastic nodules (DNs). Nevertheless, while DNs are considered precancerous lesions of HCC [3], various natural outcomes of DNs have been reported [4,5].
Previously, expression of stathmin and elongation factor 1α (EF1α) was reported to be up-regulated in telomere dysfunctional mice [6]. Stathmin, encoded by the human STMN1 gene, is a major microtubule depolymerizing protein involved in cell cycle progression [7]. Stathmin directly interacts with soluble tubulin to form a complex that sequesters free tubulin and hinders the polymerization of microtubules [8]. In mitosis, stathmin is inactivated by phosphorylation to induce microtubule polymerization and the assembly of mitotic spindles; phosphorylated stathmin is reactivated by dephosphorylation, allowing cells to exit mitosis and enter a new interphase [9]. Stathmin, also known as oncoprotein 18, has been reported in several studies to be highly expressed in a wide variety of human cancers, including leukemia, breast, prostate, gastric, and liver cancer, suggesting that stathmin is a key molecule in tumorigenesis [10][11][12][13]. Moreover, stathmin overexpression in HCC reportedly increased the malignant potential thereof by regulating cell motility, cell migration, and cell proliferation [12]. Stathmin overexpression in HCC was also shown to be correlated with poor prognosis [14,15]. EF1α is a translational cofactor of eukaryotic protein synthesis that is important in the elongation of polypeptides. EF1α carries aminoacyl-tRNA to ribosomes and dissociates after correct codon-anticodon recognition by GTP hydrolysis [16]. Overexpression of EF1α in human and rodent cells was reported to be associated with increased cell proliferation, oncogenic transformation, and delayed cell senescence [17]. A previous functional study showed that EF1α interacts with p-Akt to control the activity of p-Akt and regulates the proliferation, survival, and motility of breast cancer cells [18]. Additionally, increased expression of EF1α was reported to be associated with the growth rate of an HCC cell line, but not with apoptosis [19].
The assessment of telomere dysfunction may be helpful in evaluating the biological characteristics of hepatic nodules in cirrhotic patients. However, systemic assessment of the expression levels of stathmin and EF1α in relation to telomere dysfunction, particularly telomere length and telomerase dysfunction induced foci (TIF), has not been conducted in defined lesions of human multistep hepatocarcinogenesis, including DNs. Accordingly, the present study attempted to evaluate and compare the expression levels of stathmin and EF1α in relation to telomere length, TIF, γ-H2AX, and p21 WAF1/CIP1 expression in human HBV-associated multistep hepatocarcinogenesis.

Tissue samples and pathological examination
A total of 58 liver specimens from 24 patients were investigated, including 13 cases of liver cirrhosis, 14 cases of lowgrade DNs, 17 cases of high-grade DNs, and 14 cases of HCC. The specimens were obtained from 19 men and five women whose ages ranged from 32 to 63 years (51 ± 7.8 years, mean ± SD). They were all serum HBsAgpositive and anti-HCV-negative. The clinicopathological features of the patients are summarized in Table 1. Large nodules, measuring at least 0.5 cm at their largest dimension, that were apparently distinct from the surrounding cirrhotic parenchyma, in terms of their color, texture, and degree of bulging at the cut surface; HCCs; and corresponding non-neoplastic liver tissue were collected. Half of each nodule was fixed in 10% buffered formalin, routinely processed, and embedded in paraffin for histological examination. The remaining half of each nodule was snapfrozen in liquid nitrogen and stored at −80°C until ready for use. Hepatocellular nodules were evaluated according to the criteria proposed by the International Consensus Group for Hepatocellular Neoplasia [3]. Differentiation of HCC was determined on the basis of Edmonson and Steiner classification [20]. For comparison, non-neoplastic liver tissues were obtained from 5 patients with metastatic carcinoma. Control subjects were negative for hepatitis virus and showed relatively normal liver histology. The ages of the controls ranged from 42 to 74 years (61 ± 13.9 years). Fresh frozen specimens were obtained from the Liver Cancer Specimen Bank (part of the National Research Bank Program, Korea Science and Engineering Foundation, Ministry of Science and Technology). This study was approved by the Institutional Review Board of Severance Hospital, Yonsei University College of Medicine, and the requirement for informed consent was waived.
Total RNA extraction and real-time quantitative RT-PCR Total RNA was isolated from human fresh frozen liver tissue samples using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The transcription thereof into cDNA was performed using a High Capacity RNA-to-cDNA kit (Applied Biosystems, Foster City, CA, USA). The following TaqMan Gene Expression Assays were purchased from Applied Biosystems: EF1α (Hs00265885_g1), stathmin (Hs01027 516_g1), and GAPDH (Hs99999905_m1). All measurements were performed in triplicate. The relative expression levels of target mRNAs were normalized to GAPDH mRNA levels.
PCR was performed in triplicate for each cDNA sample using the ABI PRISM 7300 Sequence Detection System (Applied Biosystems). Dual-labeled FAM probes containing a 5′-fluorescent reporter and a 3′-quencher were used to conduct the PCR experiments.

Telomere terminal restriction fragment length analysis
Telomere terminal restriction fragments were measured as previously described [2]. Briefly, telomere length was measured by Southern blotting. Two μg of digested DNA was separated on 0.7% agarose gel. Hybridization was carried out with 3′-end DIG-labeled d(TTAGGG) 4 (Roche Molecular Biochemicals, Mannheim, Germany) and detected as recommended by the manufacturer. The resulting X-ray film was scanned with a luminescent image analyzer (Fujifilm, Tokyo, Japan), and the telomere signals in each lane were quantified as a grid object, defined as a single column with 25 rows, using Image Gauge Software 2.54 (Fujifilm).
Telomere fluorescent in situ hybridization (FISH) combined immunostaining for γ-H2AX TIF were detected in representative sections of formalinfixed paraffin-embedded tissues by staining for γ-H2AX and by telomere in situ hybridization using a telomerespecific peptide nucleic acid probe (Cy3-(CCCTAA) 3
Similarly, the expression of stathmin protein also gradually increased as multistep hepatocarcinogenesis progressed towards HCC (P for trend < 0.001) ( Figure 1B, Figure 2B). Most normal liver specimens showed no expression of stathmin protein, except for one case that showed weak expression of histoscore grade 1. The mean histoscores for stathmin were 1.3 ± 0.63 in liver cirrhosis, 2.0 ± 1.57 in low-grade DNs, 2.6 ± 2.12 in high-grade DNs, and 4.1 ± 1.77 in HCC. Stathmin histoscores were significantly different between liver cirrhosis and HCC, as well as between low-grade DNs and HCC; moreover, stathmin histoscores were significantly greater in liver cirrhosis, low-grade DNs, high-grade DNs, and HCC than in normal liver (all, P < 0.05).
The associations between γ-H2AX labeling index and the expression levels of stathmin and EF1α were further analyzed. γ-H2AX labeling index showed a significant positive correlation with both mRNA levels and histoscores of stathmin (P < 0.001, R = 0.456 and P = 0.005, R = 0.351, respectively), indicating that an increase in stathmin expression is associated with DNA damage (Figure 4C and D). For EF1α expression, histoscores of EF1α protein showed a positive correlation with γ-H2AX index (P < 0.003, R = 0.373) ( Figure 4D); however, we observed no significant correlations between EF1α mRNA level and γ-H2AX labeling index (P = 0.555, R = −0.076) (Additional file 2: Figure S1B).

Telomere dysfunction induced foci (TIF) and their relationship with stathmin and EF1α expression in HBV-related multistep hepatocarcinogenesis
In order to determine the fractions of TIF during human multistep hepatocarcinogenesis, we performed telomere FISH combined with immunostaining for anti-γ-H2AX to detect colocalization of telomeres and γ-H2AX foci ( Figure 5A). A gradual increase of TIF was observed as multistep hepatocarcinogenesis progressed towards HCC with statistical significance (P for trend < 0.01) ( Figure 5B). Normal liver showed the lowest average number of TIF, 0.1 ± 0.30 (mean ± SD) (range, 0.06~0. 16), and the number of TIF gradually increased as multistep hepatocarcinogenesis progressed from normal liver, liver cirrhosis, low-grade DNs, and high-grade DNs to HCC (liver cirrhosis; 0. .56]). HCC showed a significantly higher number of TIF, compared to normal liver, liver cirrhosis, and low-grade DNs (all, P < 0.05); high-grade DNs showed a significantly higher number of TIF, compared to normal liver and liver cirrhosis (all, P < 0.05); and low-grade DNs showed a significantly higher number of TIF than normal liver (P = 0.048). We further analyzed the relationships between TIF and telomere length, as well as TIF and the protein expression levels of stathmin and EF1α. TIF showed a significant negative correlation with telomere length (P = 0.020, R = −0.415, Figure 5C). EF1α histoscore showed a significant positive correlation with TIF, suggesting that higher EF1α expression is associated with a greater number of TIF, indicating increased telomere dysfunction (P < 0.001, R = 0.602) ( Figure 5D). Stathmin histoscore, however, showed no significant correlation with TIF (P = 0.326, R = 0.182) (Additional file 2: Figure  S1C). We also observed no significant correlations between TIF and the mRNA levels of both stathmin and

Discussion
Telomere erosion occurs in proliferating cells with insufficient telomerase activity; thereby, critically short telomeres become dysfunctional and trigger apoptosis and/or senescence as a tumor suppressive mechanism. Meanwhile, mutations that impair the DNA damage response pathways allow for survival of cells with critically short telomeres. As a result, cells with dicentric chromosomes can proliferate and lead to the accumulation of genomic instability due to breakage-fusion-bridge cycles. In the present study, a gradual increase in shortened and dysfunctional telomeres was found as HBV-related multistep hepatocarcinogenesis progressed to HCC. Low-grade DNs showed a significantly higher number of TIF than normal liver; high-grade DNs exhibited a significantly greater number of TIF than liver cirrhosis and normal liver; and HCC showed the highest number of dysfunctional telomeres. Additionally, telomere length was inversely correlated with γ-H2AX labeling index, a DNA damage marker. These findings support that telomere dysfunction and DNA damage are important in the progression of HBV-related multistep hepatocarcinogenesis.
Stathmin was recently shown to be induced in response to telomere dysfunction and DNA damage in human aging and diseases [6]. Excessive stathmin activity was reported to generate chromosomal instability through blockage, subsequent to metaphase-to-anaphase transition, by decreasing the fidelity of chromosome segregation to spindle poles during anaphase in an in vitro study [21]. In the present study, stathmin expression at both the mRNA and protein level increased gradually as multistep hepatocarcinogenesis progressed from liver cirrhosis, lowgrade DNs, and high-grade DNs to HCC, which showed the highest levels of expression. Interestingly, stathmin mRNA levels were significantly higher in high-grade DNs than in liver cirrhosis and normal liver and the histoscores of stathmin protein were higher in high-grade DNs and low-grade DNs than normal liver; stathmin expression was not detected in normal liver. Additionally, the protein and mRNA expression levels of stathmin were well correlated with γ-H2AX labeling index and telomere shortening; moreover, fractions of TIF-positive cells were also well correlated with telomere length shortening.
The primary function of EF1α is to transport aminoacyl-tRNA to ribosomes during protein translation. The overexpression of EF1α was previously reported to be associated with cell proliferation, oncogenic transformation, and delayed cell senescence in human and rodent cells [16,17]. As well, previous functional studies demonstrated that EF1α regulates the proliferation, survival, and motility of breast cancer cells [18] and induces higher proliferation capacity in undifferentiated HCC cell lines [19]. In the present study, EF1α protein expression gradually increased in human multistep hepatocarcinogenesis, and was highest in HCC. Interestingly, EF1α histoscore in HGDN was significantly higher than that in normal liver and liver cirrhosis. The expression levels of EF1α protein were well correlated with γ-H2AX labeling index and TIF, whereas they exhibited a significant inverse correlation with telomere length. Meanwhile, EF1α mRNA level showed no significant correlation with telomere length, γ-H2AX labeling index, TIF, or p21 WAF1/CIP1 labeling index, and further investigation of a translational regulation mechanism for EF1α is needed. Notwithstanding, our data suggested that the expression of stathmin and EF1α arises in DNs as precancerous lesions and gradually increases along with progression of human hepatocarcinogenesis, suggesting that expression of stathmin and EF1α is induced by shortened and dysfunctional telomeres in B viral multistep hepatocarcinogenesis. p21 WAF1/CIP1 is a potent cyclin-dependent kinase inhibitor and the expression of this gene is controlled by the tumor suppressor protein p53, through which this protein mediates the p53-dependent cell cycle G1 phase arrest in response to a variety of stress stimuli [22]. In this study, p21 WAF1/CIP1 was inactivated in low-grade DNs, high-grade DNs, and HCC, in contrast to its high expression in liver cirrhosis. Interestingly, p21 WAF1/CIP1 labeling index showed a significant negative correlation with TIF, stathmin mRNA level, and the histoscores of EF1α protein in this study. Accordingly, the p21 WAF1/CIP1 cell cycle check point was discerned to be a defense mechanism against damaged and transformed cells, triggered by dysfunctional telomeres in multistep hepatocarcinogenesis.
The natural history of DNs has not been fully clarified, and prospective studies conducted in large series of histologically proven non-neoplastic nodules detected by ultrasonography during surveillance programs of cirrhosis have demonstrated a wide evolutionary fate for such lesions. The hazard ratios of high-grade DNs, low-grade DNs, and large regenerative nodules for transformation to HCC were reported to be 16.8, 2.96, and 1.0, respectively [4,5]. However, it is difficult to predict the biological behavior of individual hepatic nodules, as some high-grade DNs remain stable for a long time period and a few of them even disappear. Therefore, the assessment of shortened and dysfunctional telomeres may be helpful to evaluating biological characteristics of hepatic nodules, although direct measurement of telomere dysfunction using telomere FISH combined immunostaining is very time consuming and labor intensive. Nevertheless, this study revealed that the expressions of stathmin and EF1α are good indicators of shortened and dysfunctional telomeres, and the evaluation of the expression levels thereof in conjunction with p21 WAF1/CIP1 may prove helpful in characterizing the biological nature of hepatic nodules in B viral cirrhotic patients.
Here, we demonstrated for the first time that the expression of stathmin and EF1α are closely related to telomere dysfunction, DNA damage, and inactivation of p21 WAF1/CIP1 in the defined lesions of HBV-related multistep hepatocarcinogenesis, including liver cirrhosis, low-grade DNs, high-grade DNs, and HCC. Accordingly, assessment of stathmin and EF1α levels as a reflection of telomere dysfunction may be helpful in evaluating the biological characteristics of precancerous hepatic nodules in hepatitis B viral cirrhotic patients.

Additional files
Additional file 1: Table S1. Antibodies used in this study.
Additional file 2: Figure S1. Correlation of stathmin and elongation factor 1α (EF1α) expression with telomere dysfunction and DNA damage in HBV-related multistep hepatocarcinogenesis. A-B. Scatter plots reveal a correlation between EF1α mRNA expression and telomere terminal restriction fragment (TRF) length (A) and γ-H2AX labeling index (B). C. Correlation between telomere dysfunction induced foci (TIF) and stathmin protein level. D. Scatter plot of a correlation between p21 WAF1/CIP1 labeling index and EF1α mRNA expression.