Abstract
Background: Cancer cell growth has been described to depend on glucose utilization. Activation of Akt and up-regulation of pyruvate kinase M2 (M2-PK) are attributes of tumour glycolysis. Patients and Methods: In order to evaluate the prognostic relevance of glycolytic markers in breast cancer, the expression of pAkt and M2PK was analysed in 160 tissue samples. The staining results were compared with clinicopathological characteristics and survival data. Results: Overexpression of pAkt was detected in 58% and of M2PK in 70% of breast cancer samples. Increased pAkt-expression was accompanied with shorter survival time. In contrast, M2PK expression was significantly higher in patients surviving breast cancer for more than 13 years. Conclusion: Strong M2PK expression seems to be a favourable prognostic factor and its role in breast cancer progression should be further explored. Our data confirm previous observations of pAkt as a negative prognostic marker.
Breast cancer is the most frequently diagnosed malignancy among women and the leading cause of cancer death in Germany and worldwide (1, American Cancer Society, Cancer Statistics 2007). The probability of developing invasive breast cancer during a woman's lifetime is approximately 1 in 11 in Germany, according to the Robert Koch Institute, which represents 24% of all newly diagnosed malignant tumours (1). The chance of survival has increased significantly over the past 30 years. In 2007, the 5-year survival rate over all breast cancer stages was 89% in the U.S. and 76% in Europe (American Cancer Society, Cancer Statistics 2007). Currently applied prognostic factors for clinical use are age, nodal status, tumour size, histological grade, steroid receptor, and Her2neu status (2). Although breast cancer is a very well studied tumour entity, the influence of glucose metabolism on the ability to metastasize and on the survival time is unknown. For this study, a patient group was selected which had been treated for primary breast cancer more than 13 years ago. The patients' primary breast cancer tissues were analysed immunohistochemically for two selected key markers of tumour glycolysis. The staining results were compared with clinicopathological and survival data to gather information about the prognostic relevance of these markers.
Aerobic glycolysis in tumor cells was first described by Otto Warburg in 1924 (3), but there was only little interest in exploring the molecular fundamentals of this phenomenon. Among other factors, the successful application of 18FDG-PET (fluorodeoxyglucose positron-emission tomography) to detect malignant tumours by using a radio-labelled glucose analogue reawakened the interest in Warburg's theory. Tumour glycolysis is characterized by high glucose uptake, dependence on glucose supply, and lactate acidosis which is measurable in the tumour-surrounding tissue (3-5). In malignant cells, glucose withdrawal blocks glycolysis and causes apoptosis, whereas in normal cells it solely leads to cell cycle arrest (6). The dependence of malignant cells on aerobic glycolysis was already proven as well as the almost complete independence from mitochondrial ATP-production (7). Up-regulation of key enzymes and expression of different enzyme isotypes initiate metabolic changes up to tumour glycolysis (8).
Pyruvate kinase is a key enzyme of glycolysis and different tissue specific isoenzymes exist thereof. Malignant transformations result in a change of the isoenzyme composition. Usually, the tissue specific pyruvate kinase disappears and is increasingly substituted by M2PK, depending on the progress of the malignant disease (9, 10). The tetrameric form of M2PK can be found in normally proliferating cells, whereas the dimeric form is described as being dominant in tumour cells (9). The latter is enzymatically inactive and thus it blocks the glycolytic pathway and thereby leads to an accumulation of metabolites which serve as substrates for amino acid production or nucleic acid synthesis. The tetrameric M2PK catalyses the phosphoryl group transfer from phosphoenolpyruvate to ADP, yielding pyruvate and ATP. Contrary to previous reports on various tumour entities, breast cancer cells showed a lower expression of dimeric M2PK compared to the surrounding benign tissue (11). Thus, it appeared interesting to investigate the influence of M2PK expression on breast cancer progress.
The serine/threonine kinase Akt plays a leading role in tumour glycolysis (12-15). Additionally, Akt also promotes the immortality of tumour cells (14, 16). Human glioblastoma cells with high levels of phosphorylated and therefore activated Akt (pAkt) feature higher glucose consumption and lactate accumulation than the same cell line with low pAkt (5). Pharmacological inhibition of pAkt could be used in order to reverse this effect (5). It has been shown that the Akt pathway is activated in several tumour types and thereby it increases glucose uptake possibly by stimulating expression of GLUT1 and translocation of glucose transporters to the cell surface (17). Akt stimulates localisation of hexokinases at the outer mitochondrial membrane, which initiates phosphorylation of glucose as the first step of glycolysis (18). The metabolism of fatty acids is also regulated by Akt. Activated Akt enhances the synthesis of fatty acids and dephosphorylation of Akt decreases fatty acid synthesis and raises β-oxidation (catabolism) of fatty acids (19). In summary, the activation of Akt in tumour metabolism ensures energy supply in cancer cells by increasing glucose uptake, hexokinase activity and glycolytic processing. Akt also suppresses apoptosis and enables the assembly of new membranes in rapidly dividing cells by stimulating biosynthesis and suppressing β-oxidation of fatty acids.
Previous publications demonstrated M2PK and pAkt as essential factors of tumour specific up-regulation of glycolysis (5, 9-16). Thus, it appeared interesting to examine the effects of M2PK and pAkt expression on long-term survival in breast cancer.
Patients and Methods
Patients. For this study, a patient group of 160 breast cancer patients was selected who had been treated for primary breast cancer between 1985 and 1995. All investigations were approved by the Ethics Committee of the Medical Faculty of the University of Wuerzburg, Germany. During a patient follow-up of at least thirteen years, information about recurrences, metastases, secondary carcinomas and survival data was gathered. Clinicopathological data including oestrogen and progesterone receptor expression as well as survival data were compared with the immunohistochemical analysis of the primary breast tumours.
Immunohistochemical staining. Formalin-fixed and paraffin-embedded samples of 160 breast cancer specimens were cut in slices of 2 μm, placed on adhesive treated slides (Superfrost, Langenbrinck, Emmendingen, Germany) and deparaffinized with xylene. The slides were rinsed in decreasing concentrations of ethanol and subsequently antigens were unmasked in 10 mM sodium citrate buffer (pH=6.0) in a microwave oven at 600 W for 5 minutes. Sections were then rinsed in demineralised H2O, followed by incubation for 10 minutes in 3% H2O2 in methanol to inhibit endogenous peroxidase. Specimens were then washed with phosphate-buffered saline (PBS) and incubated with 1% goat serum in PBS for 15 minutes before they were incubated with two different primary antibodies. For the detection of M2PK, a mouse monoclonal antibody was used that only recognises the dimeric form of M2PK, which has been described as the characteristic form present in tumour tissue (14, DF4, ScheBo Biotech AG, Giessen, Germany) at a 1:250 dilution. For the detection of pAkt, a rabbit polyclonal antibody was applied (ab28821Abcam; Cambridge, UK) in attenuation of 1:200. The slides were incubated with the primary antibodies, thinned by antibody diluent (Dako, Hamburg, Germany), for 45-60 minutes at room temperature in a humidified chamber and then washed with PBS. For the detection reaction, the slides were incubated with an appropriate biotinylated secondary antibody and thereafter, with streptavidin-peroxidase according to the manufacturer's protocol (LSAB-kit; DAKO). The staining was developed by adding 3.3'-diaminobenzidine (DAB; DAKO) with subsequent counterstaining using haematoxylin. Sections were dehydrated in graded ethanol and embedded in Vitro Clud (Langenbrinck, Emmendingen, Germany). Staining protocols, positive and negative controls were used as previously described (10).
Statistical analysis and clinical correlation. Two independent researchers experienced in histopathology and immunohistochemistry determined the expression levels of M2PK and pAkt for each tissue specimen. A scoring scale, the Immunoreactive Score (IRS), as utilised for hormone receptor determination in breast cancer (21), was applied for semi-quantitative evaluation. The percentage of stained cells was classified as follows: score 1 indicating less than 10%, score 2 indicating 10-49%, score 3 indicating 50-79% and score 4 indicating staining of 80% or more of tumour cells. In the same way, staining degrees were determined: score 0 representing no staining, whereas score 1 stands for weak, score 2 for moderate and score 3 for strong staining intensity. By multiplying the score determined for the percentage of positively stained tumour cells by the score determined for staining intensity, the IRS was defined. For further statistical calculations, an IRS≤4 was preassigned as negative while an IRS>4 indicated positive marker expression. The statistical analysis was accomplished by a professional statistician applying SPSS for Windows, version 15.0 (SPSS Inc., USA).
Results
Expression of glucose metabolism markers in breast cancer tissue. This experimental study includes 160 patients who were treated for primary breast cancer between 1985 and 1995. The average age was 52, with a range from 19 to 79 years. Breast-preserving surgery and axillary dissection were performed as initial treatment for all the patients. When the immunohistochemical analysis was performed, the primary diagnosis for breast cancer was dated back between 13 and 23 years. By completion of this study, 61% (97 out of 160) of the patients were still alive and 39% (61 out of 160) had died. The estimated median survival time for the selected patient group was 16.4 years (confidence interval 15.2-17.6). The documented cause of death for 16 out of the 61 deceased patients was breast cancer. The primary breast cancer specimens of all patients were stained for expression of pAkt and M2PK. The IRS scores determined by each of the two observers showed no significant differences. In Table I, the clinicopathological and immunohistochemical characteristics are summarised. The Kolmogorov-Smirnov test confirmed that the selected patient sample does not follow Gaussian distribution and is stochastically independent (results not shown). Therefore, nonparametric tests were performed. The Mann-Whitney U-test was used to compare two independent samples.
The cellular staining pattern for M2PK and pAkt in the breast cancer tissue complied with previously investigated tumours. Staining for M2PK was found only in the cytoplasm, whereas for pAkt, also membranous and nuclear staining were detected in addition to predominantly cytoplasmic staining. Figure 1 shows examples of expression patterns of pAkt (c, d) and M2PK (a, b).
The chosen patients' samples featured an overexpression (IRS>4) of M2PK in 70% and an overexpression of pAkt in 58% of the breast tumours (Table I). Applying Spearman's Rho test, no correlation between M2PK, pAkt, oestrogen or progesterone receptor expression was found. Data about hormone receptor expression were available for patients who were surgically treated after 1989.
Glucose metabolism markers and clinical outcome. For survival analysis, all tumour samples from primary surgery with sufficient clinical data including follow-up were evaluated (n=160). Survival time was significantly shorter for patients without M2PK expression (IRS≤4) than for patients with strongly M2PK expressing tumours (IRS>4) (14.7 years [confidence interval 12.4-17] versus 17.1 years [confidence interval 15.8-18.5], log rank p=0.058) (Figure 2). When the patients for this clinical trial were selected, 62% of the patients with M2PK over expressing tumours were still alive, whereas only 42% of the patients without excessive M2PK expression had survived.
The IRS values for living and deceased patients were compared by applying the Mann-Whitney U-test. Significantly higher levels of M2PK expression in breast tumours of surviving patients were found (Figure 3). Age, tumour size, nodal status, oestrogen and progesterone receptor status did not correlate with M2PK expression (Mann-Whitney U-test, results not shown). Cox regression analysis was used to determine independent prognostic factors. Apart from age and nodal status, M2PK expression in breast cancer was identified as an independent prognostic factor for survival time (p=0.059).
Clinicopathological and immunohistochemical characteristics as well as survival data of 160 patients with breast cancer.
Patients with pAkt overexpressing tumours (IRS>4) had a slightly shorter survival time than patients with low pAkt expression (IRS ≤ 4) (Figure 4). This observation was not statistically significant (15.9 years [confidence interval 14.3-17.5] versus 17.1 years [confidence interval 15.3-18.8], log rank p=0.34). Tumour size, nodal status, and progesterone receptor status did not correlate with pAkt expression (Mann-Whitney U-test, results not shown). A significantly higher expression of pAkt was found in younger patients (Figure 5, Mann-Whitney U-test, p=0.038) and in patients with low oestrogen receptor status (Figure 6, Mann-Whitney U-test, p=0.106).
Immunohistochemical staining for M2PK and pAkt in breast cancer. Weak staining (a, c) versus strong staining pattern (b, d) for M2PK (a, b) and pAkt (c, d) in breast cancer are shown.
Discussion
This study shows that expression of M2PK in breast cancer is an independent prognostic factor for long-term survival (13 years or more). This is of significant clinical relevance since many breast cancer patients might be overtreated with aggressive adjuvant therapy. Routine diagnosis of M2PK in breast cancer might help in the decision finding process for individualised therapy concepts.
Recently, glycolytic markers as indicators of altered glucose metabolism in tumour cells attracted a growing interest. Until now, little is known about the prognostic relevance of the expression of glycolytic markers in breast cancer. The expression of M2PK and pAkt in breast cancer tissue of a group of patients with a very long follow-up time of 13 years or more was evaluated. The relevance of pAkt for tumour metabolism was described in several studies (22). High expression levels of pAkt were detected in younger patients and patients with low expression of oestrogen receptor. Patients with tumours overexpressing pAkt had a slightly shorter survival time than patients with low pAkt expression. This result matches previous reports describing pAkt overexpression as an indicator of poor outcome (12). To assay for M2PK expression, an antibody was used that specifically binds to the dimeric isoform of M2PK. The enzymatically inactive dimeric M2PK causes an accumulation of glycolytic metabolites upstream of phosphoenolpyruvate, whereby substrates for synthetic processes are provided. There was no evidence for a correlation between expression of M2PK and age, tumour size, nodal status, oestrogen, and progesterone receptor status. Survival time was significantly longer in patients expressing M2PK, so that M2PK was identified as an independent prognostic factor. This confirms previous findings that the M2PK level is higher in normal breast tissue and lower in ductal carcinoma in situ and malignant cells (11). However, these observations contradict previous reports, whereby the dimeric isoform of M2PK is dominant in tumour cells (9). According to the mentioned results, the role of M2PK may be different in slow growing breast tumours and rapidly progressing tumour types such as colon or ovarian carcinoma. In breast cancer, energy supply may be a more decisive factor than the assembly of new cells. This may explain the appearance of a glycolytic phenotype in breast cancer cells in which glycolysis is not blocked by dimeric M2PK, as observed for other tumour types (9-10).
Mortality rate for patients with M2PK overexpressing tumours (IRS>4) versus patients with low M2PK expression (IRS≤4).
The IRS for M2PK in breast cancer specimens of patients surviving more than 13 years after primary diagnosis versus deceased patients.
Mortality rate for patients with pAkt overexpressing tumours (IRS>4) versus patients with low pAkt expression (IRS≤4).
The IRS for pAkt in breast cancer patients below and above mean age.
The IRS for pAkt in breast cancer patients below and above mean IRS for oestrogen receptor.
For this experimental study, breast cancer samples of patients with a very long follow-up time of more than 13 years were analysed for M2PK and pAkt expression. The overexpression of pAkt was found to be indicative of poor outcome and M2PK overexpression was identified as an independent prognostic factor for long-term survival. These results may be useful in the prognosis of breast cancer patients and may suggest the use of glycolysis inhibitors for therapeutic purposes.
Acknowledgements
We thank Karin Benesch, André Schmidt and Kurt Weigand for valuable discussions, Alexander Benesch for the critical review of the manuscript, as well as Tanja Kottmann for the statistical analysis. This study was supported by the Medical Faculty of the University of Wuerzburg and Hilfe im Kampf gegen Krebs e.V., Wuerzburg.
Footnotes
- Received August 19, 2009.
- Revision received April 8, 2010.
- Accepted April 16, 2010.
- Copyright© 2010 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
