Association between the expression of IL-10 and T cell activation proteins loss in early breast cancer patients

Abstract

Breast cancer patients may express abnormal cellular immune responses affecting their immunological competence. The analysis of immunological parameters may be useful as indicators of T cell function. To determine the expression of lymphocyte activation proteins and cytokines in tumor and non-metastatic axillary lymph nodes, 30 breast cancer patients were monitored. CD3 polypeptides, PTKs (protein tyrosine kinases) and phosphorylated tyrosines were studied by Western Blot and cytokines mRNA expression was determined by RT-PCR (reverse transcription-polymerase chain reaction). This group of patient had shown high immunohistochemistry expression of IL-10 in tumors. Activation proteins were mainly expressed in involved lymph nodes comparing with their expression in tumors. The differences in expression of CD3 polypeptides and p56^lck between both locations were significant. There was no statistical association between PTKs and IL-10 in the tumor but more than 50% of cases who express IL-10 lost p56^lck, p59^fyn. A direct association between IL-10 and CD3-polypeptides was observed, however 52.2% of patients who express IL-10 did not express 41 kDa CD3-ζ form. IL-10 mRNA was detected in more than 50% of tumors contrary to the prevalence of type 1 cytokines in regional nodes (40%). The lack of expression of lymphocyte activations proteins and the high expression of IL-10 suggest a downregulation on T cells function in the tumors. These results are useful in order to understand the local immune response that would be key in the control of the tumor progression.

Introduction

Breast cancer is the most common cancer in women and it ranks as the fifth cause of death from cancer overall. It is the leading cause of cancer mortality in women (Parkin et al. 2005; Parkin and Fernández 2006).

Patients with advanced breast cancer could have depressed cellular immunity affecting their immunological competence against the tumor (Hadden 1999).

The local production of cytokines within the tumor microenvironment is crucial in mounting an immune response against tumor cells, and the presence of suppressive cytokines might thwart the effector’s responses. Failure to stimulate the production of particular cytokines may inhibit the functions of tumor-associated lymphocytes directed to the tumor cells (Wong et al. 1998; Venetsanakos et al. 1997).

Mononuclear cells of patients affected by breast cancer have a defective IL-12 production capability, while generating higher amounts of IL-10 (Merendino et al. 1999). A significant type 2 response (IL-10 and IL-4) has been observed in the peripheral blood lymphocytes of patients with malignant breast cancer (Rosen et al. 1998). Several studies indicate that the Th2 cytokine pattern is present at the tumor site and suggest that these cytokines may mediate immunosuppression (Huang et al. 1995). Immunohistochemical analyses have demonstrated that the presence of IL-1α and IL-1β in tumor cells in patients with invasive cancer and ductal carcinoma in situ (Kurtzman et al. 1999; Pantschenko et al. 2003). Significant expression of IL-8, TNF-α, IL-10, TGF-β1, and IL-2 transcripts in neoplastic human breast tissue have also been observed (Green et al. 1997; Marrogi et al. 1997; Benoy et al. 2004).

The immune system is capable of responding to cancer as evidenced by systemic, regional and intratumoral lymphocyte activation (Wong et al. 1998). Lymphocyte infiltration may correlate, in some cases, with prognosis; but these lymphocytes have shown suppression of their function in cancer patiens (Venetsanakos et al. 1997).

The earliest biochemical response elicited by T cell receptor (TCR) after binding MHC (major histocompatibility complex)-associated peptide is the rapid phosphorylation of tyrosine residues on a variety of substrates including the immunoreceptor tyrosine-based activation motifs (ITAMs) in ζ chain, and other CD3 chains via non-receptor protein tyrosine kinases (PTKs) such as lymphocyte-specific src-family PTKs, p56^lck and p59^fyn and T cell-specific syk-family PTK, ZAP-70 (Hermiston et al. 2002; Krishnan et al. 2003; Palacios and Weiss 2004; Chakraborty et al. 2006). In cancer, alterations of tyrosine kinases p56^lck, p59^fyn and CD3-ζ and CD3-ε polypeptides cause deficiencies in the T cells activation, suggesting a downregulation in the immune response (Nakagomi et al. 1993; Reichert et al. 1998, 2002; Salvadori et al. 2000; Mizoguchi et al. 1992; Gratama et al. 1999; Nieland et al. 1998).

We have studied previously the immunohistochemistry expression of IL-10 in early stage breast cancer patients. IL-10 was associated with some poor prognostic tumor markers (Llanes-Fernández et al. 2006). In this sense, we thought it was important to study the relationship of this cytokine with other immunological factors associated to the T lymphocyte activation process in this group of patients. In this work, we compare the expression of IL-10 with CD3-ζ and CD3-ε polypeptides and tyrosine kinases p56^lck and p59^fyn, in the tumor microenvironment in breast cancer patients. These factors were also determined in the non-metastatic lymph nodes from patients. Few studies have investigated some of these parameters in non-metastatic lymph nodes (Ishigami et al. 2004; Kohrt et al. 2005). The immune profile analysis of tumor-draining nodes may provide an additional information on clinical outcome (Kohrt et al. 2005). We also studied the cytokine pattern in both the locations in order to know the predominant immune response. This study may be important to understand the local immunosuppression process and to define targets that could be modified in tumor microenvironment in order to increase the effectiveness of biological therapies in early breast cancer patients.

Materials and methods

Patients and samples

Thirty patients from The National Institute of Oncology and Radiobiology, Havana, Cuba were studied. The patients were confirmed as having breast cancer and had no previous history of cancer therapy. The cancers were in the early clinical stages (I and II). Patients ranged from 27 to 85 years of age at the time of surgery (51 as median age).

Tumor and proximal regional lymph nodes from the breast cancer patients were collected aseptically during surgical removal. A pathologist performed histological diagnoses of the tumors and regional lymph nodes. Among the 30 breast tumors examined, 23 were infiltrating ductal carcinomas of different grades, three were invasive lobular carcinoma, two were medullary carcinoma, one was a papillary intracystic carcinoma, and one was a mucinous carcinoma. The tumor areas were analyzed and uniformly selected from the center to the periphery of the tumor, where the activity is higher. The frozen and paraffin lymph nodes sections were appropriately analyzed to determine the status of every lymph node harvested, thus excluding the presence of metastatic lymph node material. The study was carried out on only non-metastatic lymph nodes from level I. The halves of lymph nodes were used to determine the markers. We had a representation of all lymph node architecture in each sample.

Normal tissues samples were collected aseptically from patients undergoing cosmetic breast surgery.

All samples were maintained at −80°C until used. Samples were cut in pieces of 100 mg to be processing.

Western blot

CD3-ζ, CD3-ε, p56^lck, and p59^fyn were detected by Western blot analysis. T cells activation proteins were determined in 30 breast cancer patients, for p59^fyn we just had 28 samples available for assays. Proteins were extracted from the tumor samples and regional lymph nodes of breast cancer patients by Trizol method (Chomczynski 1993). Phosphoproteins were also determined in proteins extracted by Trizol (Yang et al. 2006). The cell lysates were centrifugated at 20,000×g to remove nuclei and cell debris and then dialyzed against 0.1% SDS (sodium dodecyl sulphate). The protein concentration of samples was evaluated using the Bio-Rad Protein Assay (Bio-Rad Laboratories). Twenty micrograms of protein from each sample was mixed with 6× SDS sample buffer, and separated on 12% SDS-PAGE (SDS-polyacrylamide gel electrophoresis). Later, the proteins were electroblotted to PROTRAN BA83, pure nitrocellulose transfer and immobilization membrane (Schleicher & Schuell) in a semi-dry transfer (Owl Panther Semidray Electroblotter) for 90 min at 120 constant MA. Nonspecific binding sites were blocked by incubating the nitrocellulose membrane with washing buffer [PBS (phosphate-buffered saline)/Tween20 to 0.05%] containing 5% skim milk, during 1 h at room temperature. The immunoblots were then incubated overnight at 4°C with the first specific mouse-monoclonal antibody [anti-CD3-ζ (6B10.2) 1:1000, anti-CD3-ε (UCH-T1) 1:50, anti-Lck (3A5) 1:100, anti-Fyn (15) 1:100, anti-p-Tyr (PY99) 1:100, anti-actine (C-2) 1:100 or anti-CD2 (Huly-m1) 1:100], supplied by Santa Cruz Biotechnology (Santa Cruz, CA, USA). The blots were washed four times for 15 min before incubating with horseradish peroxidase-conjugated monoclonal anti-mouse IgG (Promega) for 2 h at room temperature. After washing, the blots were developed by the chemiluminescence method according to the manufacturer’s specifications (SuperSignal West Pico Chemiluminiscent, Pierce).

The assayed activation proteins were compared with the immunohistochemistry IL-10 expression obtained in a previous work (Llanes-Fernández et al. 2006), in 27 tumor samples of patients from our study group.

RT-PCR analysis

The mRNA levels of cytokines were analyzed using RT-PCR (reverse transcription-polymerase chain reaction) as described elsewhere (Favre et al. 1997). Total RNA was extracted from 100 mg of tissue as previously described (Chomczynski 1993). Sections of the frozen tissues were homogenized with a polytron in 1 ml of Trizol (Gibco, Gaitherburg, MD), and then 200 μl of chloroform were added and whirled in a vortex. Each sample was centrifuged at 10,000×g for 15 min. The upper aqueous phase was transferred to another tube for RNA precipitation by adding 500 μl of isopropanol. Then, the sample was rinsed with 75% ethanol and resuspended in 25 μl of DEPC (diethylpyrocarbonate)–water. The RNA was quantified in a DU-40 spectrophotometer, and 3 μg of each sample were used to analyze its integrity in urea–polyacrylamide gel. The RNA samples were stored at −80°C until used. First strand cDNA was generated using Oligo d(T)₁₈ and reverse transcriptase M-MLV in a reaction volume of 20 μl at 37°C for 2 h and inactivated at 65°C for 10 min. PCR amplification of the resultant cDNA was made in a Peltier PTC200 thermocycler, and it was carried out in a final volume of 50 μl containing 5 μl of 10× PCR buffer, 3 μl of MgCl₂ (25 mM), 1 μl of dNTP (10 mM each), 1 μl (100 ng/μl) of cytokine or G3PDH (glyceraldehyde-3-phosphate dehydrogenase) specific primers (control gene), 0.5 μl (1 U/μl) of Taq polymerase (Gibco, BRL), 1 μl of first-strand cDNA and PCR water. PCR was carried out for 35 cycles with 45 s for denaturation (94°C); 45 s for IL-2 (65°C), for IFN-γ (60°C), and for IL-10 (55°C) annealing; and a final extension of 2 min (72°C). The PCR products were electrophoresed on 1% of agarose gel and stained with ethidium bromide before visualization in UV light, DNA bands were photographed. The oligonucleotide primers used were:

IL-10a

5′-ATGCCCCAAGCTGAGAACCAAGACCCA (sense)

IL-10b

5′-TCTCAAGGGGCTGGGTCAGCTATCCCA (antisense)

IL-2a

5′-CATTGCACTAAGTCTTGCACTTGTCA (sense)

IL-2b

5′-CGTTGATATTGCTGATTAAGTCCCTG (antisense)

IFN-γa

5′-GCATAGTTTTGGGTTCTCTTGGCTGTTACTGC (sense)

IFN-γb

5′-CTCCTTTTTCGCTTCCCTGTTTT (antisense)

G3PDHa

5′-ACCACAGTCCATGCATCAC (sense)

G3PDHb

5′-TCCACCACCCTGTTGCTGTCA (antisense)

The resultant PCR products were: 426 bp for IL-10, 305 bp for IL-2, 427 bp for IFN-γ, and 452 bp for G3PDH.

Phytohemaglutinin (PHA)- or Concanavalin A (ConA)-stimulated and non-stimulated peripheral blood mononuclear cells (PBMC) were used as positive controls.

Statistical analysis

A descriptive analysis of the data was performed using frequency distributions and percentages. The association between variables and the difference in expression of variables between the tumor microenvironment and lymph nodes were tested using the independence χ ²-test (a significance level of 0.05 was considered). The statistical analysis was carried out using the statistical package SPSS, 11.0 for Windows (Microsoft Corp).

Results

Expression of CD3-ζ and CD3-ε polypeptides in the microenvironment of tumors and regional non-metastatic lymph nodes in breast cancer patients

The presence of ε and ζ CD3 polypeptides in tumors and axillary non-metastatic lymph nodes of patients were analyzed by Western Blot. Ten out of 30 patients (33.3%) did not express CD3-ζ in the tumor samples, but 90% (27/30) of them expressed in their lymph nodes. This difference was statistically significant. The difference in expression of CD3-ε was more evident, and also significant when comparing those two microenvironments. A total of 53.3% (16/30) of patients did not express CD3-ε in the tumor samples, while 83.3% (25/30) of them expressed in the lymph nodes. Figure 1a and b illustrates the CD3-ζ and CD3-ε expression of some representative patients in both locations, and Table 1 summarizes these expressions. It is noteworthy that in some patients who expressed these polypeptides in their lymph nodes was not possible to detect them in their tumors (i.e. patients 21–25 and patient 27, Fig. 1a, b). Some patients expressed CD3-ζ (i.e. patients 26, 28–31 and patient 33, Fig. 1a) and CD3-ε (i.e. patients 26 and 30; Fig. 1b) in both the locations. Only in a few patients the expression of the polypeptides was confined to the tumors (i.e. patient 34, Fig. 1b). In ten out of 30 patients, the expression of CD3-ζ coincides with CD3-ε expression, in both microenvironments. In all tumor samples the presence of lymphocytes was verified using an anti-CD2 antibody (Fig. 1c). The expression of these two polypeptides was not detected in any of the normal breast tissues (Fig. 1e).

Fig. 1.

Expression of the lymphocyte activation proteins CD3-ζ and CD3-ε in tumor and regional non-metastatic lymph nodes of breast cancer patients. CD3-ζ and CD3-ε were determined by Western blot. Lines 1–13 represent patients 21, 22, 24–34, line 14 represent PBMC PHA-stimulated. a CD3-ζ in tumors (T) and lymph nodes (LN) samples. b CD3-ε in tumors (T) and lymph nodes (LN), c expression of CD2 as lymphocyte control in tumor samples, d expression of actin control in tumor samples. e CD3-ζ and CD3-ε in normal breast tissues (the immunoblots were incubated with anti-CD3-ε and anti-CD3-ζ at the same time). Lines 1–4 normal tissues; line 5 PBMC PHA-stimulated. The actin control for each normal tissue is shown

Table 1.

Expression of cytokines, CD3-polypeptides, and PTKs in breast cancer patients

	Patients				p
	Tumor		Lymph nodes
	n/total	%	n/total	%
Protein expression
CD3-ζ	20/30	66.7	27/30	90	0.028*
CD3-ε	14/30	46.7	25/30	83.3	0.003*
p56lck	15/30	50	29/30	96.7	0.000*
p59fyn	12/28	42.8	18/28	64.3	0.108
mRNA cytokine expression
IL-10	16/30	53.3	6/30	20	0.007*
IL-2	10/30	33.3	8/30	26.7	0.573
IFN-γ	0/30	0	4/30	13.3	ND

ND not determined

* Statistically significant difference between the expression of variables in the tumor microenvironment and non-metastatic lymph nodes

Some patients expressed different forms of CD3-ζ polypeptide at the tumor site and lymph nodes by Western blot analysis. Figure 2a and b shows the results obtained in some representative patients (15–20, 36 and 37) expressing different forms of CD3-ζ in both microenvironment. In tumors, 14, 6, and 14 patients expressed bands of CD3-ζ in range of 41, 16, and 31 kDa, respectively. In lymph nodes, the expression was more heterogeneous showing CD3-ζ in range of 16, 21, 32, and 41 kDa in 20, 8, 14, and 11 patients, respectively. The molecular weights of these bands correspond with the molecular weights of CD3-ζ phosphorylated and non-phosphorylated forms, even monomer and dimer (Furukawa et al. 1994). Furukawa et al. (1994) immunoprecipitated a 32 kDa protein as the nonphosphorylated ζ dimer, the 34–42 kDa cluster representing the tyrosine-phosphorylated ζ dimer and the 20–22 kD protein as the phosphorylated monomer. Then, in order to corroborate the presence of phosphoprotein, we measured tyrosines phosphorylated by Western blot analysis. Phosphoproteins in the range of 21, 41, and 75 kDa were detected in lymph nodes and tumor microenvironment of some patients. The tyrosine-phosphorylated pattern in lymph nodes of patients 16–18, 20, and 37 and tumors of patients 9, 12, 14, 18–20, 36, and 37 are shown in Fig. 2c. Patient 19 (line 6, Fig. 2c) acted as a negative control as well as normal tissue (line 7, Fig. 2c). This result suggests a previous activation state of T cell in both the microenvironments.

Fig. 2.

Expression of different molecular forms of CD3-ζ and phosphorylated tyrosines in tumors and non-metastatic lymph nodes of breast cancer patient. CD3-ζ and phosphoproteins were determined by Western blot. a CD3-ζ in lymph nodes. Lines 1–6 patients 15–20; lines 7 and 8 patients 36 and 37; and line 9 PBMC ConA-stimulated. b CD3-ζ in tumors. Line 1 normal tissue; line 2 PBMC ConA-stimulated; lines 3–7 patients 16–20; lines 8 and 9 patients 36 and 37. c Phosphorylated tyrosines in tumors and lymph nodes. Lines 1–6 non-metastatic lymph node samples (lines 1–3 patients 16–18; line 4 patient 20; line 5 patient 37; line 6 patient 19); line 7 normal tissue; lines 8–15 tumor samples (lines 8–10 patients 9, 12, 14; lines 11–13 patients 18–20; lines 14 and 15 patients 36 and 37), and line 16 PBMC ConA-stimulated. The actin control for each line is shown

Expression of p56^lck and p59^fyn in the microenvironment of tumors and regional non-metastatic lymph nodes in breast cancer patients

The tyrosine kinases, p56^lck and p59^fyn, were detected in the tumors and in axillary non-metastatic lymph nodes of patients by Western blot analyses; p56^lck were observed in the 50% (15/30) of patients tumor samples, while 96.7% (29/30) of the patients expressed this protein in the lymph nodes. This difference in expression was statistically significant. In the case of p59^fyn, difference was not statistically significant, however 12 out of 28 (42.8%) patients expressed it at the tumor site and 18 out of 28 (64.3%) at their axillary lymph nodes. Table 1 shows the expression of p56^lck and p59^fyn in tumors and lymph nodes samples. Figure 3 illustrates the expression of these proteins in both microenvironments in some representative patients. Some patients did not express PTKs at the tumor site (i.e. patients 21, 22, 24, 29; Fig. 3a); but they expressed p56^lck (i.e. patients 21–27; Fig. 3b) and p59^fyn (i.e. patients 24, 25, 27; Fig. 3b) in their nodes. In contrast; the expression of both tyrosine kinases was not detected in any of the normal breast tissues (line 1, Fig. 3a, b).

Fig. 3.

Expression of PTKs p56^lck and p59^fyn in tumor and regional non-metastatic lymph nodes of breast cancer patients. p56^lck and p59^fyn were determined by Western blot (the immunoblots were incubated with anti-Lck and anti-Fyn at the same time). a Expression of PTKs p56^lck and p59^fyn in tumors. Line 1 normal tissue; line 2 PBMC PHA-stimulated; lines 3 and 4 patients 21 and 22; lines 5 and 6 patients 24 and 25; lines 7 and 8 patients 27 and 29. b Expression of PTKs p56^lck and p59^fyn in lymph nodes. Line 1 normal tissue; line 2 PBMC PHA-stimulated; lines 3 and 4 patients 21 and 22; lines 5–8 patients 24–27. The actin control for each line is shown

Association between IL-10 and lymphocyte activation proteins in breast cancer patients

The immunohistochemistry expression of IL-10 was positive in 23 out of 27 patients, 85% of the tumor samples (Llanes-Fernández et al. 2006). The presence of IL-10 was corroborated by determining its concentration in all of IL-10 positives samples (data not shown). Table 2 summarizes the relationship between IL-10 and activation proteins assayed (CD3-ζ, CD3-ζ (41 kDa), CD3-ε, p56^lck and p59^fyn) in tumors. The 63% of patients were positives for IL-10 and CD3-ζ in a statistically significant way, and just 44.4% were positive for IL-10 and CD3-ε. On the contrary, if we compare the association of IL-10 in tumors with the expression of 41 kDa molecular form of CD3-ζ, 52.2% of patients who express IL-10 do not express this molecular form. As we analyzed before, this form is similar to the CD3-ζ-phosphorylated dimer, and its presence could be indicative of a previous activation state.

Table 2.

Relationship between IL-10 and CD3-ζ, CD3-ζ 41 kDa, CD3-ε, p56^lck, and p59^fyn in tumors

		CD3-ζ (T)a		CD3-ζ 41kD (T)		CD3-ε (T)		p56lck (T)		Total	p59fyn (T)		Total
		Neg	Pos	Neg	Pos	Neg	Pos	Neg	Pos		Neg	Pos
IL-10b (T)	Neg	4	0	4	0	4	0	2	2	4	2	2	4
	Pos	6	17	12	11	11	12	12	11	23	13	8	21
	Total	10	17	16	11	15	12	14	13	27c	15	10	25d

T tumor

^aStatistically significant association

^bIL-10 expression was obtained in a previous work in 27 tumor samples of patients from our study group (Llanes-Fernández et al. 2006)

^cStatistical associations were determined in 27 tumor samples we had previously determined IL-10

^dSamples from 25 out of 28 patients studied for p59^fyn were comparable with the results obtained for IL-10 in tumors

The associations of IL-10 with protein tyrosine kinases were not significant but 52.2 and 62% of the patients who expressed IL-10 were negatives for p56^lck and p59^fyn, respectively.

Expression of a cytokine pattern in the microenvironment of tumors and regional non-metastatic lymph nodes in breast cancer patients

To analyze the mRNA expression of cytokines, we amplified mRNA of IL-2, IFN-γ and IL-10 by RT-PCR in 30 tumors samples, and in regional non-metastatic lymph nodes from each patient. Table 1 summarizes the expression of cytokines at the tumor site and lymph nodes. IL-10 was expressed in 16/30 patients in the tumors (53.3% of the sample); only in 20% (6/30) of the patients this cytokine was expressed at their lymph nodes. Thus, 68.7% of the patients who express IL-10 at the tumor site did not express this cytokine at the axillary lymph nodes. This difference in expression was statistically significant. In IL-2, no significant difference occurred. However, IL-2 expression was observed in 10/30 patients in the microenvironment of the tumor (33.3%) and 8/30 patients expressed in the lymph nodes (26.7%). IFN-γ was not detected in tumors, but it was expressed in 13.3% (4/30) of lymph nodes. Figure 4a and b illustrates the amplified products of cytokines, from some representative patients, at the tumor site and non-metastatic lymph nodes, respectively. In contrast to the cytokine expression in the tumor tissues, IL-10, IL-2 and IFN-γ were not detected in any of the normal breast tissues (Fig. 4c). Amplification of G3PDH mRNA was successful in all tissue samples. IL-4 was studied in 15 tumors and it was not detected in any of these cases (data not shown).

Fig. 4.

Amplified cytokine products of IL-10, IL-2, IFN-γ and G3PDH detected by RT-PCR. a mRNA expression of cytokine in breast tumors. Lines 1–9 patients 12–20; line 10 PBMC PHA-stimulated. b mRNA expression of cytokine in lymph nodes. Lines 1–9 patients 12–20; line 10 PBMC PHA-stimulated. c Normal tissues (lines 1–5); line 6 PBMC PHA-stimulated and line 7 non-stimulated PBMC

Discussion

Cytokines and growth factors are powerful modulators of the immune response, their expression at the tumor site could confer advantage to the tumor to grow and suppress the cytotoxic activity of the infiltrating lymphocytes. Therefore, the analysis of these soluble factors in the tumor microenvironment provides an insight into the understanding of the tumor behavior and may be used as a prognostic factor (Marrogi et al. 1997; Rao et al. 2006). The existence of IL-10 in tumors has been associated with immune suppression and increased tumorigenicity (Poindexter et al. 2004). IL-10 is known to have inhibitory effects on T cell proliferation and function (Halak et al. 1999). IL-10 has also been shown to inhibit antigen presentation by antigen presenting cells (APC); as well as the presentation of tumor-associated antigens by tumor cells (Poindexter et al. 2004; Beissert et al. 1995; Toomey et al. 1999). Production of IL-10 is associated with the induction of anergy in T lymphocytes (Becker et al. 1993).

Other factor that may cause a decrease in the local immune response in cancer patients is a downregulation on T cells activation proteins. The lack of expression of CD3-ζ and CD3-ε chains affects not only the TCR-activated signaling pathway but also the expresssion of TCR/CD3 on lymphocytes surface (Arase et al. 2001; La Gruta et al. 2004). Abnormal expression of PTKs: p56^lck, p59^fyn have also been implicated in immune suppression in tumor-bearing hosts (Salvadori et al. 2000; Nervi et al. 2000). These defects contribute to a cellular immune dysfunction.

Decreasing of ζ chain in tumor-associated and infiltrating lymphocytes on patients with head and neck, ovary, colorectal, hepatocellular, and gastric cancer has been reported by other authors (Nakagomi et al. 1993; Kuss et al. 1999, 2002; Matsuda et al. 1995; Ishigami et al. 2002; Muller et al. 2002; Taylor et al. 2004; Maki et al. 2003). These results correlate with our findings in the breast tumor site. We observed a decrease in the expression of the CD3-ζ and CD3-ε chains in tumor samples in 33.3 and 53.3% of the patients, respectively, compared with the corresponding non-metastatic lymph nodes. It is now widely accepted the immunological changes of tumor-draining lymph nodes may provide additional biological and prognostic information in breast cancer (Kohrt et al. 2005). Few studies have investigated CD3-ζ expression in regional lymph node in cancer patients. Ishigami et al. (2004) found impaired CD3-ζ expression in regional lymph node and peripheral blood lymphocytes in gastric cancer patients. However, in our work, a high percent of the patients expressed both chains in the regional non-metastatic lymph nodes. Different forms of CD3-ζ in tumors and lymph nodes related with phosphorylated and non-phosphorylated forms of monomer and dimer were found in our work (Furukawa et al. 1994). The tyrosine-phosphorylated pattern obtained could indicate a previous immunological competence in both locations.

Alterations of p56^lck have been found not only in cancer, but also in other pathological situations; such as, leprosy, rheumatoid arthritis, lupus erythematosus and type 1 diabetes (Mizoguchi et al. 1992; Gratama et al. 1999; Nervi et al. 2000; Zea et al. 1998; Matsuda et al. 1998; Matache et al. 1999). In our work, a lack of expression of p56^lck and p59^fyn at the tumor site in comparison with their expression in the axillary lymph nodes of breast cancer patients was observed.

The association of ζ chain with the expression of IL-10 in the tumor microenvironment is still unclear. Nieland et al. (1998) reported evidences of association between ζ chain and cytokines in early breast carcinoma. Pancreatic cancer patients have shown elevated concentrations of IL-10 and little correlation with a CD3-ζ loss (von Bernstorff et al. 2001). We analyzed the association between the expression of IL-10 and the presence of CD3-ζ and CD3-ε in tumors. High percentages of patients who express CD3-polypeptides did also express IL-10 in tumors. On the contrary, more than 50% of patients who express IL-10 in tumors did not expressed 41 kDa molecular form of CD3-ζ. The diminution of the phosphorylated form in the presence of a supressive factor could suggest a local affectation of immune response in tumor microenvironment in early breast cancer patients.

Otherwise, some authors have found a kind of relationship between IL-10 and p56^lck. Lck-deficient cells have been capable to express constitutively elevated levels of lymphokine mRNA, including IL-4, IL-5, and IL-10 (al-Ramadi et al. 1996). A low expression of p56^lck has been associated with a Th2/Th1 cell ratio unbalance (Nervi et al. 2000). We have not observed a statistical association between the immunohistochemistry expression of IL-10 and PTKs in early breast cancer patients. However, more than 50% of patients who express IL-10 did not express p56^lck and p59^fyn at the tumor site.

In this study, we found heterogeneous cytokine mRNA profiles as it has been reported by other authors (Venetsanakos et al. 1997; Poindexter et al. 2004). A predominance of IL-10 was observed at the tumor site that corroborated the presence of a suppressive microenvironment in tumors of early breast cancer patients. Contrarily, a prevalence of type 1 cytokines in the regional lymph nodes occurred, as a consequence of the expression of IL-2 and IFN-γ. The relative lack of IFN-γ expression, observed in tumors, could also suggest impairment of anti-tumor immune response to breast cancer (Zanussi et al. 2003).

Lymphocytes from the tumor site and involved lymph nodes showed immunological competence as evidenced by expression of cytokines mRNA, activation proteins and a phosphotyrosine pattern. However the finding that 52.2% of patients expressing IL-10 did not express 41 kDa molecular form of CD3-ζ suggests that the lymphocytes from the tumor microenvironment could be affected by suppressive mechanisms that may cause a decrease of lymphocyte activation in the tumor or it also may cause locally post-activation shutdown of the immune system. The presence of IL-10 as a suppressive factor could affect the local immune response, even in the case of patients expressing the ζ chain and PTKs. This hypothesis could be confirmed carrying out functional studies. The expression of IL-10 and activation proteins as potential prognostic factors will be further studied as well as their relationship with other important classic prognostic factors in the breast tumor.

Conclusion

An important finding of this work is the significant downregulation of CD3 polypeptides and PTKs in the tumor site which is in contrast with their expression in the regional lymph nodes of patients. The presence of IL-10 and the loss of T cell activating molecules in the microenvironment of the breast tumors may reflect an inhibitory role on T cells function, even in the early stage of the neoplasia.

Abbreviations

ITAMs

Immunoreceptor tyrosine-based activation motifs

PTKs

Protein tyrosine kinases

MHC

Major histocompatibility complex

PBS

Phosphate-buffered saline

PHA

Phytohemaglutinin

TCR

T cell receptor

RT-PCR

Reverse transcription-polymerase chain reaction

SDS

Sodium dodecyl sulphate

SDS-PAGE

SDS-polyacrylamide gel electrophoresis

ConA

Concanavalin A

DEPC

Diethylpyrocarbonate

G3PDH

Glyceraldehydes-3-phosphate dehydrogenase

PBMC

Peripheral blood mononuclear cells

APC

Antigen presenting cells