Urokinase expression in course of benign and malignant mammary lesions: comparison between nodular and healthy tissues

Abstract

Purpose

Increased expression of urokinase (uPA), a member of the serine protease family, is an effector of metastatic cascade and has been reported in various malignancies, including breast cancer. uPA overexpression in cancer tissues was correlated with a more aggressive phenotype and it is considered a strong and independent unfavorable prognostic factor in breast cancer.

Methods

Using real-time PCR assay, we analyzed uPA expression of malignant and benign breast nodular lesions versus healthy tissues (normal breast and lymphocytes).

Results

We found that besides breast cancer nodule, normal mammary gland and lymphocytes overexpressed uPA too. Tissues obtained from women with benign lesions expressed homogeneous and lower uPA.

Conclusions

In conclusion, although uPA overexpression is typical of cancer tissues, it could be considered as a feature of the whole organism affected by cancer. On the basis of these first results, uPA could be considered for further studies as a possible useful therapeutic target in breast cancer.

Introduction

Despite recent progress in treatment and prevention strategies, breast cancer is still associated with a high rate of morbidity and mortality due to high incidence of developing metastasis. Urokinase (uPA) is a member of the serine protease family, known to be involved in different steps of the metastatic cascade (Ghiso et al. 1999; Allgayer 2006). Urokinase catalyzes conversion of inactive plasminogen to its active form plasmin (Choong and Nadesapillai 2003). When activated, plasmin degrades most components of the extracellular matrix (ECM), allowing tumor cells to have access to the systemic circulation and colonize various organs (Rabbani and Xing 1998). Moreover, independently by catalytic function, binding of uPA with its specific receptor uPAR triggers intracellular signals promoting chemotaxis, angiogenesis, mitogenesis and growth factors activation; these events are considered determinant features of malignant phenotype (Politis 1996; Bajou et al. 2002). Because uPA system is a fundamental regulator of tumor behavior and metastasis, its expression was measured in several malignancies (Morita et al. 1998; Kim et al. 2002; Shiomi et al. 2000; Usher 2005), including breast cancer (Costantini et al. 1996; Dublin et al. 2000). Overexpression of urokinase-plasminogen activation cascade was detected in cancer tissues and correlated with a more aggressive phenotype (Borgfeldt et al. 2001; Foca et al. 2000). Urokinase is universally considered a strong and independent unfavorable prognostic factor in breast cancer (Leissner et al. 2006; Harbeck et al. 2002; Look et al. 2002; Jänicke et al. 2001 ; Harris et al. 2007).

In the present pilot study, we investigated uPA expression in course of malignant and benign breast lesions. For the first time, uPA expression of malignant nodules was compared with that obtained from healthy tissues (normal breast and lymphocytes) of breast cancer patients.

Finally, HER-2, an important prognostic factor in breast cancer (Gonzales-Angulo et al. 2006), was analyzed as tumor marker not epigenetically controlled to be compared with uPA.

Materials and methods

Study design

Tissue specimens originated from a population of women taking part at the breast-screening program promoted by the Breast Unit of “Sapienza”, University of Rome. This project was approved by the local Ethic Committee; written informed consent was obtained from participating women. From September 2006 to May 2008, we recruited a group of 28 women who had to be submitted to diagnostic biopsy, because of nodular imaging detected during the screening program. Clinical specimens were obtained at time of biopsies. Biopsies were performed using Mammotome procedure or by surgery. Each of recruited patients was invited to participate in the study giving one or more of following tissues: a representative part of breast lesion (benign or malign) and/or a part of normal breast around the nodular lesion and/or a specimen of circulating lymphocytes extracted from venous blood. Tissues were immediately snap-frozen and stored at −80°C within 2 h. Blood were placed in ice immediately after collection and lymphocytes were separated through Lympholyte-H reagent within 2 h. According to histopathology, nodular specimens were classified as benign or malignant. At the end of recruitment, we obtained the following specimens: 19 cancer tissues, 10 specimens of healthful breast near to cancer lesion, 15 specimens of lymphocytes from cancer patients, 5 benign lesion, 4 specimens of healthful breast near to benign lesion, 9 specimens of lymphocytes from women with benign lesion. Finally, we obtained 10 specimens of lymphocytes from women free from breast lesion after screening; we considered this population of healthful women as control group. From ten patients with breast cancer and three women with benign lesion, we obtained all three tissues: lesion, healthful breast and lymphocytes.

Study population

No women received chemotherapy or radiotherapy before recruitment. Breast cancer patients and tumor features are listed in Table 1. Table 2 describes the characteristics of women with benign lesions, analyzed for uPA and HER-2. Breast cancer women had average age of 58 years, ranging from 32 to 81 years; 12 (63%) were in premenopausal status at time of biopsy. Tumors were prevalently invasive (68%), ductal type (68%), stage I (52%), ER and/or PGR positive (79%), HER2 negative (84%); Ki67 value was low/medium in 79% of cases.

Table 1.

Characteristics of subjects with breast cancer (N = 19)

Characteristics	No. of subjects (%)
Age (years)
Median	58
Range	32–81
Stage at diagnosis
In situ	4 (21.06)
I	10 (52.63)
II	2 (10.52)
III	2 (10.52)
IV	1 (5.27)
Carcinoma
Ductal	13 (68.43)
Lobular	6 (31.57)
Lymph node status
N0	16 (84.21)
N+	2 (10.52)
N1 mic	1 (5.27)
Hormone-receptor status
ER and/or PR positive	15 (78.94)
ER and PR negative	4 (21.05)
HER-2 status (IHC)a
+++	3 (15.79)
Negative	16 (84.21)
Triple negative breast cancer	2 (10.52)
Ki67 value
Low	9 (47.38)
Medium	6 (31.57)
High	4 (21.05)

Percentage in parenthesis

ER estrogen receptor, PR progesterone receptor, HER-2 Human epidermal growth factor 2, IHC immunohistochemistry

^aHER-2 status was assessed using the Hercept (Dako). Any tumors were classified as ++ reanalyzed using dual-color fluorescent in situ hybridization

Table 2.

Characteristics of women with benign lesions, analyzed for uPA and HER-2 (N = 5)

Characteristics	No. of subjects (%)
Age (years)
Median	49
Range	30–77
Histology
Fibroadenoma (NOS)	3 (60)
Other benign lesions	2 (40)
Dimensions	<20 mm (100)

Percentage in parenthesis

RNA extraction and real-time PCR analysis

Total RNA was purified from fresh biopsies and lymphocytes with the RNeasy Kit (Qiagen) following the manufacturer’s instructions. 0.5 μg of total RNA was used for cDNA synthesis, with 50 pmol of oligo-d(T)₁₆ and 50 U of Stratascript reverse transcriptase (Stratagene, La Jolla, CA, USA) at 42°C for 1 h, as indicated by the manufacturer. 0.01 μg of total cDNA was used for each real-time reaction; analysis was performed in triplicate for each sample. cDNA was mixed with 0.3 μM of each forward and reverse primers and 0.25 μM of each Taqman probe with 10 μl of 2× Master mix (FluoCycle Mix, Euroclone) to a final reaction volume of 20 μl. Reactions were performed on Opticon2 DNA Engine (MJ Research) with a hot start step of 95°C for 10 min followed by 40 amplification cycles (94°C, 10 s; 62°C, 20 s; 72°C, 30 s), and by a final 72°C extension for 10 min. Amplification efficiency for each primer pairs and relative probe was previously determined by amplification of a linear standard curve (from 0.32 to 200 ng) of total cDNA assessed by ultraviolet spectrophotometer. Standard curves showed good linearity and amplification efficiency (98–99%) for each primer set of experimental genes and reference (18S) gene. Primer and probe sets for uPA, HER2 and reference gene 18S are shown in Table 3. As negative controls, each sample was previously run with 18S primers without reverse transcription to detect genomic DNA contamination; moreover, blank controls were assayed in each reaction and for each primer set to detect DNA contamination of reagents. Total cDNA levels were standardized by normalizing to the 18S control and presented as no. of fold increase versus the value of the control sample mRNA.

Table 3.

Characteristics of oligonucleotides used as real-time PCR primers and probes

Gene name	Primer name	Primer sequence (5′–3′)
18S	HS18SF1	CGGCGACGACCCATTCGAAC
	HS18SR1	GAATCGAACCCTGATTCCCCGTC
	HS18SPR	[Texas Red]CCTATCAACTTTCGATGGTAGTCGCCGTGCC[BHQ2]
uPA	HSUPAF1	GTGGATGTGCCCTGAAGGA
	HSUPAR1	TGCGGATCCAGGGTAAGAAG
	HSUPAPR	[FAM]AAGCCAGGCGTCTACACGAGAGTCTCAC[BHQ1]
HER-2	HSHER2F1	GGCCTGCGGGAGCTG
	HSHER2R1	TCCGCTGGATCAAGACCC
	HSHER2PR	[FAM]TCCTTTCAAGATCTCTGTGAGGCTTCGAAG[BHQ1]

Statistical analysis

One-way ANOVA was computed and Bonferroni post-test was used to evaluate any significant (p < 0.05) difference reported in this paper. All assays were repeated at least three times; all histograms show the mean value ± SEM. Asterisks in figures evidence the statistically significant differences; differences lacking of remarks are to be considered non-significant.

Results

For each sample, mRNA expression of uPA was determined. Specimens where divided into three groups: samples obtained from breast cancer patients, samples obtained from women with benign tumors and control (healthy women’s lymphocytes). All analyzed tissues expressed uPA, in different quantities.

Urokinase expression in breast cancer patients and in women with benign lesions

Compared to control signal, the amount of uPA mRNA was significantly higher in cancer nodules than in benign breast tumors (5.3-fold vs. 3.6-fold control; p < 0.001) (Fig. 1a). Urokinase expression in breast cancer patients group was significantly higher than in women with benign lesions in all the analyzed tissues. Moreover, in women with breast cancer it was possible to observe a significant increase in uPA mRNA in cancer tissue versus normal breast around cancer and lymphocytes (5.3-fold vs. 4.8-fold p < 0.01). No significant differences were detected in tissues (nodule, normal breast and lymphocytes) obtained from women with benign lesions. Lymphocytes from control subjects showed significantly lower uPA expression than all the other tested samples (p < 0.001).

Fig. 1.

uPA mRNA expression in breast cancers. a Real-time PCR analysis of uPA expression in nodules, normal breast around cancer and lymphocytes from women with malignant (breast cancer) and benign lesions and in lymphocytes from healthy volunteers. Results are expressed as fold number increase versus control (normal lymphocytes) assumed as 1, ±standard deviation. uPA values were previously normalized to 18S RNA values. b Real-time PCR analysis of uPA expression in nodules and lymphocytes from each single patient with cancer or benign lesions. Results are expressed as fold number increase versus control (normal lymphocytes, not shown). c Mean values of samples shown in b

Comparison of uPA expression between nodules and lymphocytes in each single patient with malignant breast cancer confirmed that except for few subjects, lymphocytes and nodule uPA expression was similar (Fig. 1b). This analysis was performed only on patients for which we had both nodules and lymphocytes; average values for this analysis are shown in Fig. 1c.

HER-2 expression in breast cancer patients and in women with benign lesions

We decided to investigate HER-2 expression in order to compare uPA and HER-2 expressions in experimental tissues. HER-2 results overexpressed prevalently in malignant tissues (Fig. 2). Normal breast around cancer and lymphocytes from women with breast cancer showed significantly lower HER-2 expression versus malignant nodules (p < 0.001). Nodules, breast tissue around nodules and lymphocytes obtained from women with benign lesion expressed HER-2 at similar levels; moreover, uPA expression in these tissues was similar to lymphocytes from control subjects. Their values were significantly lower than that measured in malignant tissues. Lymphocytes and normal breast tissues, from women with malignant lesions, showed slightly, although significant (p < 0.05), higher HER-2 expression with respect to women with benign lesion and control subjects. Analysis of nodules and lymphocytes in single patients was performed for HER-2 as described above for uPA and revealed that increased HER-2 expression was significantly higher (p < 0.001) in nodules in each patient (data not shown).

Fig. 2.

Real-time PCR analysis of HER-2 expression in nodules, normal breast around cancer and lymphocytes from women with malignant (breast cancer) and benign lesions and in lymphocytes from healthy volunteers. Results are expressed as fold number increase versus control (normal lymphocytes) assumed as 1, ±standard deviation. HER-2 values were previously normalized to 18S RNA values

Urokinase expression in breast cancer tissue specimen and their association with clinical and histomorphological parameters

Relationship between uPA levels and clinical or histomorphological variables was analyzed in the group of 19 breast cancer women recruited (Fig. 3). Patients were grouped considering menopausal status, histological type, Ki67 value, hormonal and HER-2 receptor status, in order to evidence eventual links between these features and uPA values. The levels of uPA mRNA expression did not differ significantly between tumors in relation to clinical and pathological parameters, except for locally advanced triple negative tumors, in which uPA expression was very high (11.9-fold vs. 5.5-fold; p < 0.001).

Fig. 3.

Urokinase expression in breast cancer group. Patients were grouped considering clinical ad histomorphological parameters commonly used in clinical practice. Results are expressed as fold number increase versus control (normal lymphocytes, not shown)

Discussion

The aim of our study was to characterize uPA expression in the complex tumor-host. As shown in Fig. 1a, although the higher expression in malignant nodules (according to its crucial role in tumor spread), uPA results overexpressed in all the tissues obtained from breast cancer patients (i.e., in normal breast around cancer and in lymphocytes). Our data suggest that uPA overexpression is not a prerogative of malignant tissues and it could be considered as a feature of the whole organism affected by cancer. Besides, data confronting nodular and lymphocytic uPA expression in single patients showed that, except for few subjects, lymphocytes reflected uPA expression of malignant nodules (Fig. 1b, c). If these results will be confirmed by larger studies, lymphocytic uPA could be considered as marker of uPA expression of cancer lesions and they may potentially indicate uPA prognostic value. In order to gain more insight into modulation of uPA expression in breast cancer, we decided to compare uPA mRNA with HER-2 mRNA (Xing and Rabbani 1999), obtained from same groups of tissues. As shown in Fig. 2, HER-2 overexpression is preeminent in malignant nodules, whereas uPA overexpression seems to involve the whole patient-cancer complex. HER-2 and uPA have different mechanisms of transcriptional control. HER-2 overexpression results from the amplification of its oncogene, that, implicating a change in copy number, is almost exclusive for tumor cells (although a slight overexpression is observed in breast tissue and lymphocytes from cancer patients). On the other side, uPA expression is under control of epigenetic changes, i.e., related to the hypomethylation status of its promoter (Xing and Rabbani 1999; Pakneshan et al. 2004; Guo et al. 2002). Hypomethylation of DNA seems to be not exclusive of one tissue in the same organism: Piyathilake and Johanning (2002), investigating the status of global DNA methylation in lung cancer patients, found no significant difference in global hypomethylation between malignant tissue and normal buccal mucosal cells. Undoubtedly, the most important elements able to modulate DNA methylation machinery (inadequate dietary intake of methyl groups or impaired activity of enzymes acting on one-carbon metabolism) are systemic disorders. Inadequate dietary intake of methyl groups (due to deficiency of choline, folate, methionine and vitamin B12) (Kim 2004) induces proto-oncogene (c-myc, c-fos and c-Ha-ras) DNA hypomethylation and consequent overexpression of corresponding mRNA in rat liver; these changes are related to the development of hepatocellular carcinoma in rat (Ghoshal et al. 2006). Interestingly, lymphocytes are described as good markers of the methylation status of organism (Rogers 1995; Jacob et al. 1998; Rampersaud et al. 2000), showing a decrease of global DNA methylation, in response to folate depletion diet. Our results, showing similar uPA expression between cancer tissues and lymphocytes from cancer patients, support the hypothesis that multiple systemic elements, perturbing the genomic methylation efficiency, may reflect on uPA amounts expressed by several tissues in the organism. On the other side, coherent with the different mechanism of transcriptional control, HER-2 in lymphocytes from cancer patients was more similar to healthy tissues, indicating a cancer-specific overexpression of HER-2.

In benign lesions, we found a moderate expression of uPA mRNA (Fig. 1a). Urokinase, aging on ECM proteolysis, could make the spread of benign nodular lesions easier. As described in breast cancer patients, uPA is measured not only in nodules, but also in normal breast and lymphocytes. We supposed that systemic expression of uPA is linked to the epigenetic control of uPA expression.

In the analyzed malignant tissues, uPA mRNA levels did not differ significantly between tumors in relation to common clinical–pathological parameters (hormone receptors, HER-2 status, Ki67 value, menopausal status), except for two subjects, that expressed very high levels of uPA mRNA (Fig. 3). These cancers were negative for expression of estrogen, progesterone and HER-2 receptors. This feature, called “triple negative status” (Tian et al. 2008; Rakha and Ellis 2009), distinguishes a group of cancers with a very aggressive behavior and poor prognosis. From our knowledge, this is the first study that evidenced a relationship between “triple negative status” and uPA overexpression. If these data were confirmed by larger studies, uPA could become a possible useful therapeutic target in triple negative tumors, in which hormonal treatment and therapies against HER-2 cannot be used, because the lack of their corresponding receptors. In conclusion, our experimental study provides bases for future investigations on epigenetic alterations in breast cancer and on the possible control of methylation machinery for the inhibition of uPA expression.