Abstract
T cell lines derived in low concentrations of recombinant IL-2 (rIL-2) from TIL of patients with epithelial ovarian carcinoma (EOC) often exhibit specific cytotoxicity against autologous tumour cells. However, the ability of T cells at the tumour site to respond to ovarian carcinoma cells may be affected by the production of cytokines by the various cell types present. Using reverse transcriptase-polymerase chain reaction (RT-PCR) we investigated cytokine transcripts in: (i) established EOC tumour cell lines; (ii) solid tumour specimens or peritoneal exudate cells (PEC) from ascites or peritoneal washings of patients with EOC; and (iii) CD4+ TCRαβ+ and CD8+ TCRαβ+ TIL-derived T cell lines developed in rIL-2. We have found that (i) established EOC tumour cell lines expressed transcripts for transforming growth factor-beta 2 (TGF-β2) (7/7), but not IL-10 (0/7) or interferon-gamma (IFN-γ) (0/7) and rarely IL-2 (1/7); (ii) PEC expressed transcripts for IL-2 (12/13), IL-10 (9/13), and TGF-β2 (12/13), and less often, IFN-γ (3/13), whereas solid tumour specimens from eight patients with EOC expressed transcripts for IL-2 (4/8), TGF-β2 (4/8), and IL-10 (5/8), but not for IFN-γ (0/8); (iii) CD4+ TCRαβ+ T cell lines expressed transcripts for IFN-γ (4/4), IL-2 (4/4) and IL-10 (3/4), whereas CD8+ TCRαβ+ T cell lines expressed transcripts for IFN-γ (5/5), IL-2 (1/5) and IL-10 (2/5). None of these T cell lines expressed TGF-β2 transcripts. The frequency of IL-2 and TGF-β2 transcripts in solid tumours was significantly lower than in the PEC (P = 0.0475). CD4+ or CD8+ T cell lines expressing IFN-γ, IL-2 and IL-10 transcripts were derived in culture with rIL-2 from the TIL of specimens that did not necessarily express these cytokines in the absence of rIL-2. The frequency of cytokine transcripts in T cell lines compared with these same transcripts in the PEC was significantly higher for IFN-γ (P = 0.0005) and lower for TGF-β2 (P = 0.0001). An association was observed between the expression of cytokine transcripts in vivo or by TIL-derived cell lines and functions exhibited by either production of cytokines or in vitro cytotoxicity.
Keywords: ovarian cancer, cytokines, RT-PCR
INTRODUCTION
Advanced stage epithelial ovarian carcinoma (EOC), which comprises 80% of patients with EOC, has a 5-year survival < 20%. The introduction of platinum and paclitaxel chemotherapy has led to an improvement in progression-free survival [1], although new therapy approaches are still needed. Results of studies that we [2–4] and others [5] have carried out suggest that TIL from EOC may represent an in vivo active immune response of the host to the tumour in certain patients with EOC. Furthermore, results from several clinical trials performed with i.p. cytokines both with and without activated lymphocytes [6–9] suggest that certain patients with EOC may be responsive to immunotherapy. The detection of certain cytokines in the microenvironment of the tumour, such as interferon-gamma (IFN-γ) and IL-2, could suggest that T cell activation has occurred in vivo.
Cytokines produced at the tumour site may be produced either constitutively or by cells in response to specific signals. Producing cells may be T cells or other mononuclear leucocytes that have been activated in vivo [2,3]. Activated subsets of CD4+ TCRαβ+ and CD8+ TCRαβ+ T cells may produce a number of cytokines, including IFN-γ, granulocyte-macrophage colony-stimulating factor (GM-CSF), tumour necrosis factor-alpha (TNF-α), IL-10, IL-4 and IL-2. A subset of activated CD4+ T cells may produce Th1-type cytokines, such as IFN-γ, TNF-α and GM-CSF, which could facilitate the development of cytotoxic CD8+ TCRαβ+ T cells from precytotoxic precursors. Different subsets of cells may produce Th2-type cytokines such as IL-10, IL-4, IL-5 and IL-6. Other growth-regulating cytokines such as transforming growth factor-beta (TGF-β) can be produced by tumour cells and other cells in the microenvironment in a constitutive manner. TGF-β has been shown to interfere with the activation of T cells [10]. Both TGF-β and IL-10 may result in decreased expression of MHC on tumour cells, lymphocytes and monocytes [11–18]. TGF-β through its blockade of cell division in G1/S is also an effective inhibitor of cell growth [19] and has been shown to induce apoptosis in T cells [20]. IL-10 inhibits the growth of activated monocytes [21], and also certain T lymphocytes of the Th1 type [22]. Other inhibitory effects of certain cytokines include reduction of IFN-γ production by IL-6 [23], activation of natural suppressor cells by GM-CSF [24], and inhibition by IL-4 of IL-1β, IL-6, TNF-α, TNF-β and IFN-γ production by lymphoid cells [25].
A number of cytokines including IL-1α, IL-10, IL-6, TNF-α, GM-CSF and IFN-γ have been identified in tissue specimens obtained from patients with EOC [17,26]. In only a few instances has expression been detected by the RNase protection assay [27] or by the more sensitive reverse transcriptase-polymerase chain reaction (RT-PCR) assay [26]. Cultured tumour cells of EOC have been shown to express RNA and/or protein of IL-1, GM-CSF, TNF-α, IL-6 and TGF-β [15,17,27–29].
In this study we investigated and compared cytokine transcripts in solid tumour specimens and peritoneal exudate cells (PEC) (washings or ascites) containing tumour from patients with EOC. We also studied cytokine transcripts in established EOC tumour cell lines and CD4+ TCRαβ+ and CD8+αβ+ TIL-derived T cell lines.
MATERIALS AND METHODS
Fresh solid tumour and PEC specimens
All specimens from patients that were acquired for this study were obtained under IRB-approved protocols. PEC containing tumour were obtained either from ascites or from peritoneal washings and prepared as described previously [30]. Briefly, heparinized malignant ascites or saline washings were centrifuged, the pellet was suspended in calcium- and magnesium-free PBS and layered onto a Ficoll–Hypaque density cushion. After centrifugation, cells were removed from the interface and washed twice in PBS. One aliquot of ≈ 107 cells was centrifuged and RNA was extracted from the cell pellet as described below. RNA from solid tumour specimens was extracted from ≈ 0.5 g of tumour that was minced with a scalpel and subjected to the extraction procedure described below.
Tumour cell lines
Tumour cell lines SK-OV-3, CA-OV-3, and 2774 were maintained at 37°C in L-15 medium (Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS). The EOC tumour cell line 2780 was obtained from Dr Plunkett and the 2008 and HEY tumour lines were provided by Dr R. Bast. The short-term ovarian tumour cell line (MDA113) was developed by culturing peritoneal exudate cells that were primarily tumour cells from a patient with EOC in L-15 medium supplemented with 10% FBS.
TIL-derived T cell lines
TIL-derived T cell lines were established and maintained in AIM V medium (Gibco-BRL, Gaithersburg, MD) supplemented with 600 U/ml rIL-2 (Proleukin IL-2, 18 × 106 U/ml protein; Chiron, Emeryville, CA), as we have previously described [30]. This concentration of IL-2 is one log lower than has been reported by other investigators [31]. In certain cases a CELLector CD8 device (Applied Immune Sciences Inc., Menlo Park, CA) was utilized to obtain CD8+ or CD4+ T cell lines of higher purity [32]. Using this procedure T cell lines were derived from 9 of 16 patients [33]. Cell surface phenotypes of the T lymphocyte cell lines were determined by cell surface immunofluorescence, as we have described previously [30]. The following MoAbs were used (Becton Dickinson, San Jose, CA): anti-CD3, anti-TCRαβ, anti-CD4, anti-CD8, anti-CD16 and anti-CD56. Cells were fixed and analysed using a fluorescence-activated cell sorter (Epics Profile Analyser, Coulter Electronics, Hialeah, FL). Some of these data have been reported previously [30]. T cells and PEC were cryopreserved in liquid nitrogen in 95% FBS/5% dimethyl sulfoxide.
RNA extraction
RNA was extracted from cells using guanidine thiocyanate-phenol [34,35]. Cells (≥ 2 × 106 T cells, 1–10 × 106 PEC cells, or 3–10 × 106 cells from ovarian tumour cell lines or 0.5 g minced solid tumour) were lysed with 1 ml solution A (guanidine thiocynate-phenol), mixed with 100 μl chloroform, allowed to sit in ice for 10–15 min and then centrifuged for 15–20 min in a microfuge. The upper phase was collected and RNA was precipitated with an equal volume of isopropanol for 1 h (or overnight) at −20°C. The RNA was centrifuged for 5–10 min in the microfuge, and the RNA pellet was washed twice with 70% ethanol. The RNA was then air-dried, suspended in diethyl pyrocarbonate (DEPC)-treated water and stored at −80°C. The RNA was quantified by absorbance at 260 nm. In some experiments, RNA from T cell lines and fresh TIL was extracted directly from cryopreserved cells. Briefly, cells stored in liquid nitrogen were thawed, suspended in PBS, centrifuged and the cell pellets were processed as above to extract RNA. No significant differences were seen in the results obtained from cryopreserved cells compared with those obtained from fresh cells.
cDNA synthesis
Total RNA (1–2 μg) was transcribed with Moloney murine leukaemia virus reverse transcriptase in 25 μl reaction mixtures containing 50 mm Tris–HCl (pH 8.3 at 37°C), 40 mm KCl, 6 mm MgCl2, 10 mm DTT, 20–40 U RNase-inhibitor (Boehringer Mannheim, Indianapolis, IN), 0.01 OD260 units of oligo (dT)12–18, 4 mm each of deoxynucleotide triphosphate (Pharmacia, Piscataway, NJ) and 200 U reverse transcriptase (Gibco-BRL). The reaction was carried out overnight at 37°C. The cDNA was used immediately or stored at −20°C.
PCR amplification of cDNA
cDNA aliquots (5–10%) were amplified in a 50 μl reaction mixture containing 0.5 U of Amplitaq (Perkin Elmer Cetus, Norwalk, CT) according to the instructions of the manufacturer. Amplification consisted of 35 cycles of 1 min at 95°C, 1 min at 55°C and 1.5 min at 72°C, followed by 6 min at 72°C in a DNA thermal cycler (Perkin Elmer Cetus). Each reaction contained a pair of primers specific for each cytokine [26,36,37]. The following primers were used: IL-2, 5′ ATGTACAGGATGCAACTCCTGTCTT 3′ and 5′ GTCAGTGTTGAGATGATGCTTTGAC 3′, product size 458 base pairs (bp); IL-4, 5′ CCTCTGTTCTTCCTGCTAGC 3′ and CCGTTTCAGGAATCGGATCA 3′, product size 300 bp; IL-10, 5′ CTGAGAACCAAGACCCAGACATCAAGG 3′ and 5′ CAATAAGGTTTCTCAAGGGGCTGGGTC 3′, product size 351 bp; IFN-γ, 5′ ATGAAATATACAAGTTATATCTTGGCTTT 3′ and 5′ GATGCTCTTCGACCTCGAAACAGCAT 3′, product size 494 bp; TGF-β2, 5′ AAATGGATACACGAACCCAA 3′ and 5′ GCTGCATTTGCAAGACTTTAC 3′, product size 247 bp; CD3δ, 5′ CTGGACCTGGGAAAACGCATC 3′ and 5′ GTACTGAGCATCATCTCTCGATC 3′, product size 309 bp; β-actin, 5′ TGACGGGGTCACCCACACTGTGCCCATCTA 3′ and 5′ CTAGAAGCATTGCGGTGGACGATGGAGGG 3′, product size 661 bp. All of the primers were designed in a manner so that the correct sized product could only be amplified from cDNA and not from genomic DNA sequences. A fraction (15 μl) of each PCR product was displayed on a 1.5% agarose/0.1 × TBE gel (44.6 mm Tris, 44.6 mm boric acid, 1 mm EDTA, pH 8.0) containing 0.5 μg/ml ethidium bromide (EtBr) and visualized with UV light. Molecular weight markers consisted of a commercially prepared 100-bp ladder (Gibco-BRL) or pBR328 plasmid DNA cut with restriction enzymes Bgl I and Hinf I (Boehringer Mannheim). In certain instances, PCR product was purified with a Wizard PCR purification kit (Promega) and digested with restriction enzymes (Hinf I (New England Biolabs, Inc., Beverly, MA) or Xba I (Promega Corp., Madison, WI)) following the manufacturer's directions. The restriction enzyme digests were purified with the Ultra Clean DNA purification kit (Mo Bio Labs, Inc., Solana Beach, CA) and displayed on a 5% polyacrylamide gel.
Determination by ELISA of cytokines produced by T cell lines or in peritoneal fluids obtained from EOC patients
Cryopreserved T cell lines were thawed and grown in AIM V medium supplemented with 10% FBS. The growth medium was replaced with AIM V without FBS, and samples were collected after 6, 24 and 48 h incubation at 37°C and stored at −80°C. The samples were assayed for cytokine concentration using IFN-γ, IL-10 assay kits (Biosource International, Camarillo, CA), TGF-β2 (R&D Systems, Minneapolis, MN), and IL-2 (Amersham Life Sciences Inc., Arlington Heights, IL) following the manufacturer's directions. Peritoneal fluid specimens were examined for IL-2, IL-10, IFN-γ and TGF-β, as we have previously described [33,38].
Determinaton of in vitro cytotoxicity by TIL-derived T cell lines
In vitro cytotoxicity against autologous tumour cells and natural killer (NK)-resistant Daudi tumour cells was determined in a 51Cr release assay using a method that we have previously described [30].
Statistical analysis
Comparisons were made by two-sided exact χ2 analysis using the StatXact software package.
RESULTS
Representative results of cytokine transcript analysis using RT-PCR and the cytokine-specific amplification primers described in Materials and Methods are shown in Fig. 1. These representative results are from TIL-derived CD4+ (a) and CD8+ (b) T cell lines, a PEC tumour specimen (c), a solid tumour specimen (d), and an established ovarian carcinoma tumour cell line (e).
Fig. 1.
Representative reverse transcriptase-polymerase chain reaction (RT-PCR) results from different specimens. Shown are RT-PCR results for the indicated cytokines obtained from: (a) a CD4+ T cell line (012) in which transcripts were observed for IL-2, IL-10, IFN-γ, CD3 and β-actin; (b) a CD8+ T cell line (077) in which transcripts were observed for IL-2, IFN-γ, CD3 and β-actin; (c) a peritoneal exudate tumour specimen (135) in which transcripts were observed for IL-2, IL-10, IFN-γ, transforming growth factor-beta 2 (TGF-β2) (weak), CD3 and β-actin; (d) a solid tumour specimen (R24p) in which transcripts were observed for IL-2, IL-10, TGF-β2, CD3 and β-actin; and (e) an epithelial ovarian carcinoma (EOC) tumour cell line (2008) in which transcripts were observed for TGF-β2 and β-actin. Molecular weight markers (mol. wt) were developed from a commercially prepared 100 base pair (bp) DNA ladder. The 600 bp marker is shown by the brightest band.
Cytokine transcripts of EOC tumour cell lines
We determined the expression of cytokine transcripts in six established EOC tumour cell lines (SK-OV-3, CA-OV-3, 2774, 2780, Hey, and 2008) and one short-term ovarian tumour cell line (MDA 113). TGF-β2 transcripts were found in all seven cell lines (representative result shown in Fig. 1e). Five of six ovarian tumour cell lines (SK-OV-3, CA-OV-3, 2774, Hey, and 2008) also expressed IL-4 transcripts (the MDA 113 line was not tested for IL-4), demonstrating that 80% of the EOC tumour cell lines expressed IL-4 transcripts. In contrast, none of the seven ovarian tumour cell lines expressed IL-10 or IFN-γ (representative result shown in Fig. 1e). One of seven lines, the MDA113, expressed message for IL-2. However, CD3δ transcripts were not detected in this culture, suggesting that T cells were an unlikely source for these transcripts. RNA extracts of the 2774 tumour cell line repeatedly expressed CD3δ-specific transcripts. However, analysis of 2774 cells by cell surface immunofluorescence using an anti-CD3 MoAb and the fluorescence-activated cell sorter (FACS) was negative, suggesting that CD3 protein was not expressed (data not shown). IL-4 transcripts are expressed in many tissues and further analyses were not performed with these primers.
Cell surface leucocyte phenotypes in PEC specimens
Proportions of mononuclear leucocytes in PEC specimens ranged from 1% to 100% (mean ± s.d., 55 ± 11.1%) of the total cells present. Seven PEC specimens were phenotyped for lymphocyte surface antigens using FACS analysis, and demonstrated that the live gated lymphocyte population was primarily CD3+ cells. The results are shown as mean ± s.d. and are as follows: CD3+ (n = 5), 77 ± 4.8% (range 52–90.5%); TCRαβ+ (n = 6), 73 ± 5.2% (range 46–86%); CD4+ cells (n = 6), 51.6 ± 8.2% (range 32–87%); CD8+ cells (n = 6), 18.5 ± 2.3% (range 12–27%); CD4/CD8 ratio (n = 6), 3.27 ± 0.9 (range 1.4–7.9). CD16+ cells (n = 7), 4 ± 0.6% (range 3–7%) and CD56+ cells (n = 6), 7 ± 3.9% (range 3.5–16%) were present only in small proportions.
Cytokine transcripts in RNA extracts of PEC
Specimens of PEC from 13 patients with EOC (Table 1) were examined by a cytopathologist, and tumour cells were detected in all but one specimen (012). RT-PCR was performed on RNA extracted from these PEC (Table 1). Transcripts of the following cytokines were found: IL-10, nine of 13 specimens; IL-2, 12 of 13 specimens; TGF-β2, 12 of 13 specimens; IFN-γ, three of 13 specimens.
Table 1.
Cytokine transcripts in peritoneal fluid cell extracts and protein in peritoneal fluids of patients with epithelial ovarian carcinoma
Cytokine proteins in cell-free fluids associated with PEC
Cell-free peritoneal fluid specimens obtained from six to eight of 13 patients were tested for IL-2, IL-10, IFN-γ and TGF-β2 (active) by cytokine-specific ELISA (Table 1). Proportions of peritoneal fluids with detectable concentrations of the four cytokines were, respectively, as follows: IL-2, one of six; IL-10, three of eight; IFN-γ, one of six; TGF-β2 (after acidification), three of seven. The proportions of peritoneal fluids in which cytokine proteins were detected relative to the numbers of RNA extracts from these specimens that expressed cytokine transcripts by RT-PCR, were as follows: IL-2, one of six; IL-10, three of six; IFN-γ, none of one; TGF-β2, three of seven. Thus, the presence of cytokines was detected by transcript expression more frequently than the protein product, with the exception of one specimen of fluid in which a low concentration of IFN-γ was detected, in the absence of detectable IFN-γ transcript.
Cytokine transcripts in solid tumour specimens from patients with EOC
RNA extracts from specimens of primary and/or metastases obtained from eight patients were examined for the presence of the cytokine transcripts of IL-10, IL-2, IFN-γ and TGF-β2. The following cytokine transcripts were detected in either primary or metastatic tumour specimens; IL-10, five of eight; IL-2, four of eight; IFN-γ, none of eight; TGF-β2, four of eight (Table 2). In three patients transcripts were examined from both primary and metastatic specimens (R17, R24 and R25) (Table 2). Differences were observed in the presence of IL-2, IL-10 and TGF-β2 transcripts between primary and metastatic tumour specimens from patient R24 only.
Table 2.
Cytokine-specific transcripts detected by reverse transcriptase-polymerase chain reaction (RT-PCR) in solid tumour specimens from patients with epithelial ovarian carcinoma
Comparisons of the frequency of cytokine transcripts IL-2, IFN-γ, IL-10 and TGF-β2 in PEC and in solid specimens showed significant differences: for IL-2, 12/13 versus 4/8 (P = 0.0475), and for TGF-β2, 12/13 versus 4/8 (P = 0.0475). The comparisons of transcripts in PEC and solid tumours for IL-10 (9/13 versus 5/8) and for IFN-γ (3/13 versus 0/8) were not significant. There were also no significant differences in the expression of these transcripts between primary and metastatic tumours.
Cytokine transcripts detected in ovarian TIL-derived T cell lines
We have developed in low concentrations of rIL-2 (600 U/ml) T cell lines from the same solid tumour or PEC specimens that were used for cytokine analysis by RT-PCR [30]. Phenotypic analysis of these T cell lines demonstrated that all contained ≥ 93% CD3+ and ≥ 88.1% cells. T cell lines from four of these patients were of the CD4+ CD8− phenotype and contained ≥ 92.6% CD4+ cells (range 92.2–96.2%) and < 9.7% CD8+ cells (range 3.4–9.7%). These lines were designated CD4+ (Table 3). T cell lines from the remaining five patients were of the CD4− CD8+ phenotype and contained ≥ 88.5% CD8+ cells (range 88.5–99.7%) and ≤ 4% CD4+ cells (range 0.5–4.0%). These T cell lines were designated CD8+ (Table 3). CD16+ cells were absent or present in very low proportions in these T cell lines (≤ 3.7%; range 0–3.7%). CD56+ cells were also present in low proportions in six T cell lines (≤ 9.7%; range 0–9.7%). However, the T cell line from patient 123 (CD4+ line) contained 21% CD56+ cells and the T cell line from patient 070 (CD8+ line) contained 40.4% CD56+ cells. RNA extracts from these T cell lines were analysed for the presence of cytokine-specific transcripts (Table 3). All nine T cell lines exhibited CD3 transcripts and IFN-γ transcripts, but none exhibited transcripts for TGF-β2. All four of the CD4+ CD8− T cell lines expressed IL-2 transcripts, and three of four expressed IL-10 transcripts. Of the five CD8+ CD4− T cell lines, two expressed IL-10 transcripts and one expressed IL-2 transcripts.
Table 3.
Detection of cytokine transcripts by reverse transcriptase-polymerase chain reaction (RT-PCR) in RNA extracts of nine ovarian TIL-derived T cell lines
A comparison of frequencies of transcripts for IL-2, IL-10, IFN-γ and TGF-β2 in the tumour cell lines and in the TIL-derived T cell lines (Table 3) revealed significant differences in the expression of IFN-γ (0/7 versus 9/9, respectively; P = 0.0001) and TGF-β2 (7/7 versus 0/9, P = 0.0001). Frequencies in the expression of IL-2 (1/7 versus 5/9) and IL-10 (0/7 versus 5/9) in extracts from the tumour cell lines versus the extracts from the T cell lines were not statistically different. However, comparison of the frequencies of cytokine transcripts in CD4+ and CD8+ T cells was statistically different for IL-2: 4/4 for CD4+versus 1/5 for the CD8+ T cell lines (P = 0.05).
A comparison of the frequency of cytokine transcripts in the PEC versus TIL-derived T cell lines revealed a higher frequency of IFN-γ transcripts in the TIL-derived T cell lines versus the PEC (9/9 versus 3/13, P = 0.0005), and a higher frequency of TGF-β2 transcripts in the PEC versus the T cell lines (12/13 versus 0/9, P < 0.0001). The frequencies of IL-2 and IL-10, respectively, were not statistically different: IL-2, 12/13 versus 5/9; IL-10, 9/13 versus 5/9.
Results comparing transcripts in TIL (PEC or solid specimen) with transcripts in six T cell lines derived from the TIL were statistically different for TGF-β2 expression, 5/6 in TIL versus 0/6 in T cell lines (P = 0.0152), and approached significance for IFN-γ expression, 2/6 in TIL versus 6/6 in T cell lines (P = 0.0606).
Cytokine production by ELISA and in vivo cytotoxicity by 51Cr release exhibited by TIL-derived T cell lines
TIL-derived T cell lines were tested for cytokine proteins in culture experiments, or for in vitro cytotoxicity against autologous tumours (Table 4). These studies showed that 3/3 T cell lines tested produced IFN-γ (two of these had no detectable transcript in the PEC from which they were derived) and 1/4 produced substantial amounts of IL-10 (although IL-10 transcript was not detected in the PEC). Six of six T cell lines tested exhibited autologous tumour cell cytotoxicity. One T cell line (O12) also exhibited substantial cytotoxicity against Daudi cells.
Table 4.
Cytokine production and in vitro cytotoxicity measured by 51Cr release exhibited by TIL-derived T cell lines
Because other investigators have failed to detect IL-2 transcripts in solid EOC tumour specimens [26], we digested the RT-PCR products from a solid primary tumour (patient R24) and from a CD4+ T cell line (patient 046) with Hinf I and Xba I to verify that IL-2 transcripts were being detected. The sequence for IL-2 mRNA (Genbank accession no. E00978) contains one site each for Hinf I and Xba I. Restriction enzyme digestion of the PCR products from both specimens yielded DNA fragments of the expected size, confirming that IL-2 transcripts had been detected with our assay (Fig. 2).
Fig. 2.
Restriction enzyme digestion analysis of IL-2 polymerase chain reaction (PCR) products. IL-2 primer-specific PCR products (100 μl reaction for each digestion) from a solid tumour specimen and a CD4+ cell line were digested with Hinf I or Xha I. The restriction digests were displayed on a 5% polyacrylamide/0.1 × TBE gel. Molecular markers (M) were pBR 328 DNA cut with Bgl I and Hinf I restriction enzymes.
DISCUSSION
We have shown that T cell lines, either CD8+ TCRαβ+ or CD4+ TCRαβ+ or mixtures of these, can be developed in ≈ 50% of tumour specimens from certain patients with EOC, by expanding single-cell mixtures of TIL and tumour cells in low concentrations of IL-2 [30]. Certain ovarian TIL-derived T cell lines with the CD8+ TCRαβ+ phenotype exhibited preferential cytotoxicity against autologous tumour cells, and the TCR and MHC class I molecules were shown to be involved in this activity [2,3,30,32]. In other experiments we have shown that certain TIL-derived T cell lines produced IFN-γ and TNF-α in response to autologous, but not allogeneic, tumour cells [4]. In addition, fresh (uncultured) ovarian TIL contain substantial proportions of T cells utilizing identical T cell receptors, suggesting that these T cells have been expanded in vivo at the site of the tumour in response to particular tumour peptide/MHC epitope(s) [39]. We have shown that HLA class I expression on ovarian tumour cells correlates with T cell infiltration in vivo and with T cell expansion in vitro, in low concentrations of rIL-2 [40]. Cytokines that are produced by infiltrating mononuclear cells or tumour cells at the site of the tumour may play a significant role in regulation of HLA class I and II expression on tumour cells and in the development and regulation of T cell responses against the tumour cells. TGF-β2 transcripts (as well as TGF-β1, data not shown) were found in all seven EOC tumour cell lines examined, whereas others detected TGF-β1 transcripts in two of three EOC cell lines examined, and TGF-β2 transcripts in one of three EOC tumour lines using the RNase protection assay [27]. In a recent study we showed by quantitative image analysis that anti-TGF-β1 and anti-TGF-β2 MoAbs resulted in a high density of staining of matched pathologic sections of primary and metastatic tumours of the ovary (Gordinier et al., submitted). IL-10 [41,42] and IL-2 [43] were reported in malignant cells of other types of cancer, but not in ovarian cancer cells. In our study IL-2 transcript was detected in only 1/7 ovarian tumour cell lines and transcripts for IL-10 and IFN-γ were absent.
Our findings on cytokine transcripts in solid tumour specimens appear to be at variance with those of others [26] who found IL-2 transcripts in 0 of 11 ovarian solid tumour specimens and IFN-γ transcripts in eight of 11 solid tumour specimens [26]. The most likely explanation for these differences may be the different types of EOC tumours studied. Pisa et al. [26] employed tumour specimens primarily from endometrioid tumours, whereas in our study specimens were obtained primarily from serous and undifferentiated carcinoma. Endometrioid tumours may have more lymphocytic infiltrates than other types of EOC. Serous carcinomas are the most frequent histologies found in patients with EOC.
The higher frequency of detection of IL-2 transcripts in PEC (12/13) in comparison with solid tumour (4/8) suggested that T cell activation might be more frequent in the PEC in comparison with solid tumours. In support of this finding, IFN-γ transcript was detected more frequently in the PEC (3/13 cases), but none in solid tumour specimens. However, the fact that IFN-γ transcript was detected in only three of a total of 21 patient specimens (13 from PEC and eight from solid tumour) also suggests that T cell anergy is present overall in most tumour specimens either from the peritoneal cavity or from solid tumour specimens. This effect could be due to soluble factors produced in the tumour environment, or due to the absence of costimulation. We have shown that dendritic cells, which are necessary cells for costimulation in vivo, are present in the peritoneal cavity tumour environment of patients with peritoneal carcinomatosis [44]. However, these peritoneal dendritic cells, which we have characterized as lineage negative DR+, have low or absent expression of CD80 (B7.1), and lower expression of CD86 (B7.2) in comparison with matched peripheral blood specimens [44]. Along the same lines we have shown absent or significantly down-regulated CD3ζ transcripts and proteins in most tumour specimens from patients with EOC and, in contrast, have detected CD3ζ transcripts and protein in most malignant ascites specimens from patients with EOC (Pappas et al. unpublished results). It is interesting, moreover, that CD4+ or CD8+ T cell lines that produce transcripts for IFN-γ, IL-2 and IL-10, and in certain instances the protein as well, have been generated from specimens that do not always express transcripts for these cytokines before culture in vitro with IL-2, which has obviously induced expression of these transcripts. These T cell lines also exhibit preferential autologous killing in comparison with their effects on K562-resistant Daudi cell lines.
The finding that none of our T cell line samples expressed transcripts for TGF-β2, in contrast to the detection of TGF-β2 transcripts in extracts of all tumour cell lines, suggests that tumour cells could be the main source of TGF-β2 production in tumour specimens. In contrast, TGF-β1 is produced by TIL-derived T cell lines (data not shown). It remains to be determined whether the finding of TGF-β2 expression in EOC is a marker for tumour aggressiveness, as has been reported in melanoma [45].
We conclude the following: (i) that transcripts of IL-2 are present in substantially higher proportions of PEC specimens than in solid tumour specimens, and that IFN-γ transcript is expressed at low frequency, and (ii) that TGF-β2 message is expressed preferentially in established EOC tumour cell lines and that ovarian tumours are important producers of TGF-β2. In contrast, IL-10 transcripts were not produced by EOC tumour cell lines.
Acknowledgments
We acknowledge the excellent technical assistance provided by B. Tomasovic and S. Templin. This work was supported by Public Health Service grants CA-64943, MOI-RR02558, UO1CA-62461 from the National Cancer Institute and grant EDT-56 from the American Cancer Society to R.S.F. Oligonucleotides were provided by the Macromolecular Synthesis Facility at UTMDACC supported by CA-16672 from the National Cancer Institute. Statistical Analyses were performed by Edward N. Atkinson, Department of Biomathematics, UTMDACC.
References
- 1.McGuire WP, Hoskins WJ, Brady MF. Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Eng J Med. 1996;334:1–6. doi: 10.1056/NEJM199601043340101. [DOI] [PubMed] [Google Scholar]
- 2.Ioannides CG, Freedman RS, Platsoucas CD, Rasheed S, Kim Y-P. Cytotoxic T cell clones isolated from ovarian tumor-infiltrating lymphocytes recognize multiple antigenic epitopes on autologous tumor cells. J Immunol. 1991;146:1700–7. [PubMed] [Google Scholar]
- 3.Ioannides CG, Platsoucas CD, Rashed S, Wharton JT, Edwards CL, Freedman RS. Tumor cytolysis by lymphocytes infiltrating ovarian malignant ascites. Cancer Res. 1991;51:4257–65. [PubMed] [Google Scholar]
- 4.Kooi S, Freedman RS, Rodriquez-Villanueva J, Platsoucas CD. Cytokine production by T-cell lines derived from tumor-infiltrating lymphocytes (TIL) from patients with ovarian carcinoma: tumor-specific immune responses and inhibition of antigen-independent cytokine production by ovarian tumor cells. Lymph Cyto Res. 1993;12:429–37. [PubMed] [Google Scholar]
- 5.Peoples GE, Goedegebuure PS, Andrews JV, Schoof DD, Eberlein TJ. HLA-A2 presents shared tumor-associated antigens derived from endogenous proteins in ovarian cancer. J Immunol. 1993;151:5481–91. [PubMed] [Google Scholar]
- 6.Willemse PHB, DeVries EGE, Mulder NH, Alders JG, Bouma J, Sleijfer DT. Intraperitoneal human recombinant interferon alpha-2b in minimum residual ovarian cancer. Eur J Cancer. 1990;26:353–8. doi: 10.1016/0277-5379(90)90233-j. [DOI] [PubMed] [Google Scholar]
- 7.Pujade-Lauraine E, Guastella JP, Colombo N, et al. Intraperitoneal recombinant interferon-gamma in ovarian cancer patients with residual disease at second-look laparotomy. J Clin Oncol. 1996;14:343–50. doi: 10.1200/JCO.1996.14.2.343. [DOI] [PubMed] [Google Scholar]
- 8.Canevari S, Stoter G, Arienti F, et al. Regression of advanced ovarian carcinoma by intraperitoneal treatment with autologous T lymphocytes retargeted by a bispecific monoclonal antibody. J Natl Cancer Inst. 1995;87:1463–9. doi: 10.1093/jnci/87.19.1463. [DOI] [PubMed] [Google Scholar]
- 9.Urba WJ, Clark JW, Steis RG, et al. Intraperitoneal lymphokine-activated killer cell/interleukin-2 therapy in patients with intra-abdominal cancer: immunologic considerations. J Natl Cancer Inst. 1989;81:602–11. doi: 10.1093/jnci/81.8.602. [DOI] [PubMed] [Google Scholar]
- 10.Mule JJ, Schwarz AB, Roberts AB, Sporn MB, Rosenberg SA. Transforming growth factor-beta inhibits the in vivo generation of lymphokine-activated killer cells and cytotoxic T cells. Cancer Immunol Immunother. 1988;26:95–100. doi: 10.1007/BF00205600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Czarniecki CW, Chiu HH, Wong GHW, McCabe SM, Palladino MA. Transforming growth factor-β1 modulates the expression of class II histocompatibility antigens on human cells. J Immunol. 1988;140:4217–23. [PubMed] [Google Scholar]
- 12.de Waal Malefyt R, Haanen J, Spits H, et al. Interleukin 10 (IL-10) and viral IL-10 strongly reduce antigen-specific human T cell proliferation by diminishing the antigen-presenting capacity of monocytes via downregulation of class II major histocompatibility complex expression. J Exp Med. 1991;174:915–24. doi: 10.1084/jem.174.4.915. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Donnet-Hughes A, Schiffrin EJ, Huggett AC. Expression of MHC antigens by intestinal epithelial cells. Effect of transforming growth factor-beta 2 (TGF-β2) Clin Exp Immunol. 1995;99:240–4. doi: 10.1111/j.1365-2249.1995.tb05539.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Fiorentino DF, Zlotnik A, Vieira P, Mosmann TR, Howard M, Moore KW, O'Garra A. IL-10 acts on the antigen-presenting cell to inhibit cytokine production by Th1 cells. J Immunol. 1991;146:3444–51. [PubMed] [Google Scholar]
- 15.Marth C, Lang T, Koza A, Mayer I, Daxenbichler G. Transforming growth factor-beta and ovarian carcinoma cells: regulation of proliferation and surface antigen expression. Cancer Lett. 1990;51:221–5. doi: 10.1016/0304-3835(90)90106-8. [DOI] [PubMed] [Google Scholar]
- 16.Matsuda M, Salazar F, Petersson M, et al. Interleukin 10 pretreatment protects target cells from tumor- and allo-specific cytotoxic T cells and downregulates HLA class I expression. J Exp Med. 1994;180:2371–6. doi: 10.1084/jem.180.6.2371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Moradi MM, Carson LF, Weinberg JB, Haney AF, Twiggs LB, Ramakrishnan S. Serum and ascitic fluid levels of interleukin-1, interleukin-6, and tumor necrosis factor-alpha in patients with ovarian epithelial cancer. Cancer. 1993;72:2433–40. doi: 10.1002/1097-0142(19931015)72:8<2433::aid-cncr2820720822>3.0.co;2-l. [DOI] [PubMed] [Google Scholar]
- 18.Panek RB, Lee Y-J, Benveniste EN. TGF-β suppression of IFN-γ-induced class II MHC gene expression does not involve inhibition of phosphorylation of JAK1, JAK2, or signal transducers and activators of transcription, or modification of IFN-γ enhanced factor X expression. J Immunol. 1995;154:610–9. [PubMed] [Google Scholar]
- 19.Polyak K. Negative regulation of cell growth by TGFbeta. Biochim Biophys Acta Rev Cancer. 1996;1242:185–99. doi: 10.1016/0304-419x(95)00009-5. [DOI] [PubMed] [Google Scholar]
- 20.Weller M, Malipiero U, Groscurth P, Fontana A. T cell apoptosis induced by interleukin-2 deprivation or transforming growth factor-beta2: modulation by the phosphatase inhibitors okadaic acid and calyculin A. Exp Cell Res. 1995;221:395–403. doi: 10.1006/excr.1995.1390. [DOI] [PubMed] [Google Scholar]
- 21.Geng Y, Gulbins E, Altman A, Lotz M. Monocyte deactivation by interleukin 10 via inhibition of tyrosine kinase activity and the Ras signaling pathway. Proc Natl Acad Sci USA. 1994;1:8602–6. doi: 10.1073/pnas.91.18.8602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Romano MF, Lamberti A, Petrella A, Bisogni R, Tassone P, Formisano S, Venuta S, Turco MC. IL-10 inhibits nuclear factor-κB/Rel nuclear activity in CD3-stimulated human peripheral T lymphocytes. J Immunol. 1996;156:2119–23. [PubMed] [Google Scholar]
- 23.Yamamoto N, Zou J-P, Li X-F, et al. Regulatory mechanisms for production of IFN-γ and TNF by antitumor T cells or macrophages in the tumor-bearing state. J Immunol. 1995;154:2281–90. [PubMed] [Google Scholar]
- 24.Young MRI, Wright MA, Matthews JP, Malik I, Prechel M. Suppression of T cell proliferation by tumor-induced granulocyte-macrophage progenitor cells producing transforming growth factor-beta and nitric oxide. J Immunol. 1996;156:1916–22. [PubMed] [Google Scholar]
- 25.Lee JD, Rhoades K, Economou JS. Interleukin-4 inhibits the expression of tumour necrosis factors α and β, interleukins 1β and -6, and interferon-gamma. Immunol Cell Biol. 1994;73:57–61. doi: 10.1038/icb.1995.9. [DOI] [PubMed] [Google Scholar]
- 26.Pisa P, Halapi E, Pisa EK, Gerdin E, Hising C, Bucht A, Gerdin B, Kiessling R. Selective expression of interleukin 10, interferon γ, and granulocyte-macrophage colony stimulating factor in ovarian cancer biopsies. Proc Natl Acad Sci USA. 1992;89:7708–12. doi: 10.1073/pnas.89.16.7708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bartlett JMS, Rabiasz GJ, Scott WN, Langdon SP, Smyth JF, Miller WR. Transforming growth factor-β mRNA expression and growth control of human ovarian carcinoma cells. Brit J Cancer. 1992;65:655–60. doi: 10.1038/bjc.1992.140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Watson JM, Sensintaffar JL, Berek JS, Martinez-Maza O. Constitutive production of interleukin 6 by ovarian cancer cell lines and by primary ovarian tumor cultures. Cancer Res. 1990;50:6959–65. [PubMed] [Google Scholar]
- 29.Malik STA, Knowles RG, East N, Lando D, Stamp G, Balkwill FR. Antitumor activity of γ-interferon in ascitic and solid tumor models of human ovarian cancer. Cancer Res. 1991;51:6643–9. [PubMed] [Google Scholar]
- 30.Freedman RS, Tomasovic B, Templin S, Atkinson EN, Kudelka A, Edwards CL, Platsoucas CD. Large-scale expansion in interleukin-2 of tumor-infiltrating lymphocytes from patients with ovarian carcinoma for adoptive immunotherapy. J Immunol Methods. 1994;167:145–60. doi: 10.1016/0022-1759(94)90084-1. [DOI] [PubMed] [Google Scholar]
- 31.Topalian SL, Solomon D, Avis FP, et al. Immunotherapy of patients with advanced cancer using tumor-infiltrating lymphocytes and recombinant interleukin-2: a pilot study. J Clin Oncol. 1988;6:839–53. doi: 10.1200/JCO.1988.6.5.839. [DOI] [PubMed] [Google Scholar]
- 32.Freedman RS, Platsoucas CD. Immunotherapy for peritoneal ovarian carcinoma metastasis using ex vivo expanded tumor infiltrating lymphocytes (TIL) Cancer Treat Res. 1996;82:115–46. doi: 10.1007/978-1-4613-1247-5_8. [DOI] [PubMed] [Google Scholar]
- 33.Freedman RS, Edwards CL, Kavanagh JJ, et al. Intraperitoneal adoptive immunotherapy of ovarian carcinoma with tumor-infiltrating lymphocytes and low-dose recombinant interleukin-2: a pilot trial. J Immunother. 1994;16:198–210. doi: 10.1097/00002371-199410000-00004. [DOI] [PubMed] [Google Scholar]
- 34.Xie WQ, Rothblum LI. Rapid, small-scale RNA isolation from tissue culture cells. Biotechniques. 1991;11:324, 326–7. [PubMed] [Google Scholar]
- 35.Chomczynski P, Sacchi N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156–9. doi: 10.1006/abio.1987.9999. [DOI] [PubMed] [Google Scholar]
- 36.Spagnoli GC, Juretic A, Schultz-Thater E, et al. On the relative roles of interleukin-2 and interleukin-10 in the generation of lymphokine-activated killer cell activity. Cell Immunol. 1993;146:391–405. doi: 10.1006/cimm.1993.1035. [DOI] [PubMed] [Google Scholar]
- 37.Lüscher U, Filgueira L, Juretic A, Zuber M, Lüscher NJ, Heberer M, Spagnoli GC. The pattern of cytokine gene expression in freshly excised human metastatic melanoma suggests a state of reversible anergy of tumor-infiltrating lymphocytes. Int J Cancer. 1994;57:612–9. doi: 10.1002/ijc.2910570428. [DOI] [PubMed] [Google Scholar]
- 38.Freedman RS, Gibbons JA, Giedlin M, et al. Immunopharmacology and cytokine production of a low dose schedule of intraperitoneal administered human recombinant interleukin-2 in patients with advanced epithelial ovarian carcinoma. J Immunother. 1997;19:443–51. doi: 10.1097/00002371-199611000-00009. [DOI] [PubMed] [Google Scholar]
- 39.Lin WL, Chen P-F, Freedman RS, Platsoucas CD. Fresh (uncultured) tumor-infiltrating lymphocytes (TIL) from patients with malignant melanoma or ovarian carcinoma are comprised of oligoclonal T cells. J Immunol. 1993;150:176A. [Google Scholar]
- 40.Kooi S, Zhang H-Z, Patenia R, Edwards CL, Platsoucas CD, Freedman RS. HLA class I expression on human ovarian carcinoma cells correlates with T-cell infiltration in vivo and T-cell expansion in vitro in low concentrations of recombinant interleukin-2. Cell Immunol. 1996;174:116–28. doi: 10.1006/cimm.1996.0301. [DOI] [PubMed] [Google Scholar]
- 41.Bost KL, Bieligk SC, Jaffe BM. Lymphokine mRNA expression by transplantable murine B lymphocytic malignancies: tumor-derived IL-10 as a possible mechanism for modulating the anti-tumor response. J Immunol. 1995;154:718–29. [PubMed] [Google Scholar]
- 42.Smith DR, Kunkel SL, Burdick MD, Wilke CA, Orringer MB, Whyte RI, Strieter RM. Production of interleukin-10 by human bronchogenic carcinoma. Am J Pathol. 1994;145:18–25. [PMC free article] [PubMed] [Google Scholar]
- 43.Lin WC, Yasumura S, Suminami Y, Sung MW, Nagashima S, Stanson J, Whiteside TL. Constitutive production of IL-2 by human carcinoma cells, expression of IL-2 receptor, and tumor cell growth. J Immunol. 1995;155:4805–16. [PubMed] [Google Scholar]
- 44.Melichar B, Savary C, Kudelka AA, Verschraegen C, Kavanagh JJ, Edwards CL, Platsoucas CD, Freedman RL. Lineage negative HLA-DR positive cells with the phenotype of undifferentiated dendritic cells in patients with carcinoma of the abdomen and pelvis. Clin Cancer Res. in press. [PubMed]
- 45.Reed JA, McNutt NS, Prieto VG, Albino AP. Expression of transforming growth factor-β2 in malignant melanoma correlates with the depth of tumor invasion. Am J Pathol. 1994;145:97–104. [PMC free article] [PubMed] [Google Scholar]