Background
Thymidine phosphorylase (TP: EC 2.4.2.4.) catalyses the reversible phosphorolysis of thymidine to thymine and 2-deoxyribose-1-phosphate [
1]. TP is identical with platelet-derived endothelial cell growth factor (PD-ECGF) which promotes angiogenesis [
1‐
6]. TP activity, PD-ECGF/TP protein and PD-ECGF/TP gene expression are significantly increased in tumor tissues as compared to the adjacent non-neoplastic tissues in a variety of human carcinomas [
7‐
9]. In certain tumors a high TP activity, as well as a high of PD-ECGF/TP protein and PD-ECGF/TP gene expression are associated with unfavorable prognosis [
10‐
12]. On the other hand, the high TP activity and the PD-ECGF/TP protein expression are suggested to be useful prognostic factors for tumors treated with chemotherapy or radiotherapy [
13‐
15].
Endometrial carcinoma is one of the most common malignancies of the female genital tract. The increase of cytosol TP activity and a higher PD-ECGF/TP protein and PD-ECGF/TP gene expression as compared to the adjacent non-neoplastic tissues were reported in other gynecologic malignancies [
9,
16‐
18]. The data on PD-ECGF/TP protein expression and its correlation with microvessel density (MD) in endometrial cancer are controversial [
19‐
21]. The cytosol TP activity and its correlation with MD in endometrial cancer has not been carried out so far.
The aim of the study was also to evaluate the influence of the cytosol TP activity and PD-ECGF/TP protein expression on the intensity of angiogenesis in malignant tumors enabling prognosis and choice of the proper therapy for patients with endometrial cancer after surgery.
Discussion
Thymidine phosphorylase (TP) is identical with platelet-derived endothelial cell growth factor (PD-ECGF) which promotes angiogenesis [
2‐
6]. The aim of our study was to evaluate the cytosol activity of TP together with its protein level (PD-ECGF/TP protein expression) in relation to the intensity of angiogenesis in endometrial cancer.
Fourty three patients with endometrial malignant tumor were included in the study. The control was the normal endometrium obtained from sixteen patients. The enzyme activity was measured in the cytosol from tumor and normal tissues using the spectrophotometric method and expressed in micromoles of thymine released during 1 h per milligram of protein. A significantly higher mean cytosol TP activity was stated in these malignant tumors as compared to the control. We obtained similar observations for cytosol TP activity from ovarian [
17] and breast carcinoma [
16]. In benign and mixed uterine tumors [
9] we also found a significantly higher mean TP activity (in the cytoplasmatic soluble fraction obtained at 50 000 × g and in partially purified fraction after DEAE-Sepharose) from only those women who were in the follicular phase of the cycle during surgery as compared to the control (myometrium), although the obtained TP activity values did not differ in the normal myometrium from women who were in luteal and in follicular phase of the menstrual cycle during surgery [
9]. In our present data no statistically significant dependence has been found between the cytosol TP activities in normal endometrium in both phases of the menstrual cycle, either (data not published).
Our results are in agreement with these obtained by Takebayashi et al. [
8] for several human solid tumors. TP activity was measured in tissue homogenate using the same spectrophotometric method as ours and expressed in micromoles of thymine released during 1 h per milligram of protein and was significantly higher in carcinomas of oesophagus, stomach, colorectum, pancreas and lung than in the adjacent non-neoplastic tissue. On the contrary, there was no significant difference between the TP activity in carcinoma of liver and thyroid and than in the control materials [
8]. The high TP activity in tumors was associated with unfavorable prognosis [
10‐
12], but there were published data that in patients treated with chemotherapy this high TP activity might contribute to better prognosis [
13,
14].
The enzyme was measured using the spectrophotometric method and expressed in nanomoles of 5-fluorouracil (5-FU) released during 1 minute per mg protein in the cytosol from cervical cancer (n = 20), leiomyoma (n = 23), ovarian cancer (n = 46), ovarian endometriosis (n = 21) benign epithelial ovarian tumor (n = 27) and also from endometrial cancer tissues (n = 26) [
27]. This study indicated that he mean activity in endometrial cancer tissues was significantly higher (64,5 nanomoles of 5-FU × mg protein
-1 × minute
-1) than in the corresponding normal tissues (27,5 nanomoles of 5-FU × mg protein
-1 × minute
-1[
27].
We have also measured the expression of PD-ECGF/TP protein in the endometrial cancer and normal endometrium using an immunohistochemical technique with the P-GF-44C monoclonal antibody. In our study the staining was mainly found in cytoplasm, sometimes endothelial nuclei and part of stroma. Our observations are similar with these obtained by Fujiwaki et al. for endometrial cancer [
28]. In our study the expression of PD-ECGF/TP protein was statistically higher in endometrial cancer as compared to the control. No expression was detected in coexisting normal endometrial stroma or stroma of endometriosis either [
29]. Immunohistochemical staining with a monoclonal antibody against TP was also used in the study of the cancerous oesophagus, stomach, colon, bladder, pancreas and lung, and the proportion of TP-postive tumors was also significantly higher than that of the TP-positive adjacent normal tissues [
8]. However our results remain in controversy with those obtained by Sivridis et al. [
30] who found a poor expression of TP in endometrial carcinoma. They reported, that immunopositiveness for PD-ECGF/TP was increased in a small proportion of endometrial carcinomas (approximately 5%) and up-regulation occurred predominantly at the invading tumor front which is probably induced by cytokines of histocystic and lymphocytic origin [
30].
Takebayashi et al. [
8] assayed the expression of PD-ECGF/TP protein and the activity of TP in the same human solid tumor specimens and found a correlation between them [
8]. Little is known of TP activity in endometrial cancers except for the works of Kamoi et al. [
31] (article and abstract in Japanase) and Suzuki [
27]. Thus we have studied spectrophotometrically the cytosol TP activity in tumor samples and tested if it is correlated with immunohistochemically studied expression of PD-ECGF/TP protein in the same cases of well-differentiated endometrial carcinomas. Our study revealed that the expression of PD-ECGF/TP protein was correlated with the cytosol activity of TP in the same endometrial cancer, and this correlation was statistically significant.
In endometrial cancer a high microvessel count is an independent prognostic factor [
32‐
34] as it has relation to metastases [
35]. However, Seki et al. [
36] reported, that MD was not associated with lymph node metastasis or surgical stage and therefore angiogenesis does not necessarily contribute to metastasis in endometrial cancer [
36]. Abulafia et al. [
37] reported that angiogenesis was increased in complex endometrial hyperplasia as compared to controls of simple hyperplasia. The angiogenic capability of complex hyperplasia was comparable to the FIGO stage Ia endometrial cancer. An increased angiogenesis was found in cases of invasive (stages Ib and Ic) endometrial cancer as compared to complex hyperplasia or stage Ia endometrial cancer [
37]. Fujiwaki at al. [
38] found that the immunohistochemical expression of TP was associated with the increase of MD as an index of angiogenesis in endometrial cancer, which suggested that TP might play an important role in angiogenesis [
38]. In our study MD attained a significantly higher value in endometrial cancer in comparison with normal endometrium. Our data confirmed the correlation between MD and the cytosol TP activity and revealed the weaker of such a correlation between MD and PD-ECGF/TP protein expression in endometrial cancer, indicating that the active enzyme but not its protein influences in some way the vessel development i.e. that variety of factors may influence the activation of enzymatic protein. A high PD-ECGF/TP protein expression accompanying the high cytosol TP activity may concern also the cases of its localization in stroma. Also Sakamoto et al. [
29] showed that immunohistochemical PD-ECGF/TP expression was predominantly observed in tumor stroma, and that TP-positive cells in the stroma are probably macrophages and plasma cells [
29]. Seki et al. [
36] found, that PD-ECGF production within cancer cells was not so strong as in the stroma cells and there was no correlation between the expression of PD-ECGF in cancer cells and any clinicopathological variables including microvessel count [
36]. Another study failed to find a relationship between PD-ECGF expression and angiogenesis either [
39]. In the study of Sivridis et al. [
40] the expression of PD-ECGF in tumor cells did not effectively contribute to the angiogenic process and was deprived of any practical significance since no relationship was found between this expression and the histopathological factor examined, but was associated with aggressive histological features endometrial carcinomas [
40]. In our study in endometrial cancer revealed no significant correlation between the cytosol TP activity, PD-ECGF/TP protein expression, MD and histopathological factors. Our results were in agreement with these obtained by Suzuki at al. [
27]. There were no significant differences in the cytosol enzyme activity according to tumor stage classification. When assessed according to histopathologic degree of differentiation, the enzyme activity appeared to be higher in pathologic grade G – 3 than in grade G – 1 and G – 2, but this difference did not show significance [
27]. We found that TP activity in tumor was higher in grade G – 2 than that obtained in G – 1 and G – 3 and this difference did not show significance either. The same concerns PD-ECGF protein expression. Our results are in conformity with these obtained by Mazurek et al. [
41] who found no statistically significant correlation between TP immunohistochemical expression and histopatological grade and FIGO stage, but particular FIGO stages showed a significant trend of increasing TP endometrial cancer overexpression [
41]. We also found that PD-ECGF/TP protein expression was higher in FIGO stage II and III vs I, but there were no significant differences, either.
The mechanism of endometrial angiogenesis involves stimulation by ovarian steroids of production of angiogenic regulators by endometrial epithelium and stroma which then act on the endothelium [
42]. Endothelial cells are the principal actors in angiogenesis, which after effecting the proteolytic degradation of the basemant membranes, migrate through the membrane to form sprouts [
43]. 2-deoxy-D-ribose generated by PD-ECGF/TP catalyzing phosphorolysis of thymidine, stimulates endothelial cell migration [
1,
44] via a mechanistic pathway similar to that described for glucose that attracts endothelial cells along its concentration gradient [
45]. However, Brown et al. suppose, that 2-deoxy-D-ribose induces cellular oxidative stress, which generates oxygen radicals during early stages of protein glycation and promotes secretion of angiogenic factors [
46,
47]. Moreover, it was suggested that, beside 2-deoxy-D-ribose, thymine might also exhibit angiogenic properties stimulating the effects of other growth factor [
48]. The extent to which TP is implicated in neoangiogenesis is far from being fully elucidated [
49]. Although the mechanism of TP induction is not yet completely clear, but TNF, IL10 and other cytokines have been clearly shown to induce its expression. The various complex interactions of TP give it an essential role in cellular functioning and, hence, it is an ideal target in cancer therapy [
50].
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
EMZ conceived the study, performed the biochemical experiments and drafted the manuscript, RK performed the immunohistochemical experiments, helped to write the manuscript, AB collected tissue specimen, clinical records. JB conceived the idea, provided helpful comments. All authors read and approved the final manuscript.