This article is a US Government work and, as such, is in the public domain in the United States of America.
Expression of FasL in squamous cell carcinomas of the cervix and cervical intraepithelial neoplasia and its role in tumor escape mechanism†
Version of Record online: 2 FEB 2006
Copyright © 2006 American Cancer Society
Volume 106, Issue 5, pages 1065–1077, 1 March 2006
How to Cite
Ibrahim, R., Frederickson, H., Parr, A., Ward, Y., Moncur, J. and Khleif, S. N. (2006), Expression of FasL in squamous cell carcinomas of the cervix and cervical intraepithelial neoplasia and its role in tumor escape mechanism. Cancer, 106: 1065–1077. doi: 10.1002/cncr.21697
- Issue online: 17 FEB 2006
- Version of Record online: 2 FEB 2006
- Manuscript Accepted: 25 AUG 2005
- Manuscript Revised: 22 JUL 2005
- Manuscript Received: 10 NOV 2004
- tumor escape mechanisms;
- cervical carcinoma;
To date, several mechanisms have been described by which malignant cells escape from the immune system. One of these is through the expression of FasL. The authors hypothesized that the Fas/FasL interaction enables cervical carcinoma cells to induce apoptosis of the cells of the immune system and thereby escape from them.
The authors tested the expression of FASL on the surface of cervical carcinoma tissues. Next, they stained the same cervical tissues with anti-human leukocyte common antigen and TUNEL to identify apoptotic cells. An in vitro functional assay was then done to test if the FASL expressed on the surface of cervical carcinoma cell lines was or was not responsible for inducing apoptosis in T-cells. Finally, they compared the expression of FASL on normal and dysplastic cervical tissues.
Ninety-four percent of the cervical carcinoma tissues the authors tested expressed FasL and the majority of the apoptotic cells in the specimens were leukocytes with very few tumor cells. In the in vitro functional assay, only the Fasl expressing cell line and not the Fasl negative cell line was able to induce apoptosis of the Fas-expressing Jurkat cells. On examining the normal cervical tissues, the authors found that the expression of Fasl was confined to the basal cell layer with loss of expression observed in the suprabasal layers, which made it an immune privileged site. Conversely, there was persistent expression of FasL in the dysplastic layers in cervical dysplasia and squamous cell carcinoma specimens.
The findings of the current study support the authors' hypothesis that persistent expression of FasL plays a role in the ability of cervical carcinoma cells to escape from the immune system. Cancer 2006. Published 2006 by the American Cancer Society.
The CD95 ligand (FasL) is a 37-kilodalton (kD), type 2 transmembrane protein. On binding with its receptor, Fas (CD95), it activates a death domain extending into the intracytoplasmic region. This results in the activation of a death-inducing signaling complex that leads to the DNA fragmentation associated with apoptosis in the Fas-expressing cells.1 The CD95 ligand is highly expressed in activated mature lymphocytes and in many nonlymphoid tissues (e.g., the liver, heart, lung, kidney, skin, and ovary),2, 3 in which it plays a role in the general regulation of cell death.4 In addition, this pathway also is involved in several regulatory processes of the immune system including T-cell selection in the thymus for the development of self-tolerance,5 and the clonal deletion of activated T cells to down-regulate the inflammatory processes. In areas in which inflammation will lead to permanent damage (e.g., the retina and the testicles), there is an overexpression of FasL to escape specific killing by the cytotoxic T lymphocytes.6
Tumor-associated antigens (TAAs) have been shown to induce tumor-specific immune responses in the host.7 Tumors develop by escaping the generation of an effective immune response against TAAs. Several mechanisms have been proposed that account for tumor escape, including defective antigen presentation, interference with tumor-T cell interaction, and production of immunosuppressive factors.8–10 One other possible mechanism of immune escape is T cell death resulting from the activation of the Fas-mediated T-cell apoptosis. Activated T cells increase Fas expression on their cell surface; therefore, they become more susceptible to apoptosis when they interact with cells that express the FasL molecule.11 Increased Fas expression on T cells is one of the control mechanisms that limit the immune response. However, this control mechanism means that T cells become susceptible to apoptosis induced by cells expressing FasL. Many tumors have been shown to express FasL, including head and neck, breast, hepatocellular, colon, lung, and pancreatic carcinomas.12–15 Some of these malignancies also have been shown to be functional as tumor cell lines.16
Cervical carcinoma is a tumor that possesses distinct TAAs. Human papillomaviruses (HPVs) have been shown to cause progressive changes in the cervical epithelium, leading to cervical carcinoma. Greater than 99% of cervical malignancies harbor HPV (mainly HPV-16 and HPV-18). E6 and E7 are the two proteins expressed by the virus that are necessary for the initiation and maintenance of carcinoma. These two proteins have been found to act as tumor antigens, and yet many women develop preinvasive cervical intraepithelial neoplasia (CIN) and cervical carcinoma. A FasL-expressing cervical carcinoma cell line has been shown to induce apoptosis in Fas-sensitive, cytotoxic T-lymphocytes (CTLs) specific for that cell line, suggesting that the Fas-FasL system may be a mechanism of immune escape in cervical carcinoma.17
In the current study, we characterized the expression of Fas-FasL in cervical carcinoma and premalignant cervical lesions to explore their role in the ability of tumor cells to escape from the immune system.
MATERIALS AND METHODS
Paraffin blocks of human cervical carcinoma and normal cervical tissues were obtained from archival material at the National Cancer Institute (NCI) and from the Gynecologic Oncology Group Tissue Bank. Blocks from the tumors of these patients were cut in 5-μm sections. Unstained slides of cervical biopsies with dysplasia were generously provided by Dr. Michael Birrer (NCI, Bethesda, MD).
Cervical carcinoma cell lines used in the current study were obtained from American TypeCulture Collection (ATCC; Manassas, VA) (HeLa, Mississippi751, HS588T, SW756, HT-3, SiHa, DoTo, and Me180) or had been originated and prepared at the NCI (H3344, H3199, H3365, H3342, and H3274) from cervical carcinoma specimens taken from patients as part of an Institutional review Board-approved clinical trial. The cells from ATCC were maintained according to ATCC recommendations (available at URL: http://www.atcc.org [accessed May 2004]). The cell lines established at the NCI were maintained in Dulbecco modified Eagle medium/F-12 (at a dilution of 1:1) supplemented with 10% fetal bovine serum with penicillin (at a dose of 100 U/mL) and streptomycin (at a dose of 100 μg/mL). Jurkat A3 cells (a FasL-sensitive ATCC lymphoid cell line) were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum with penicillin, streptomycin, and HEPES buffer. The CaSki cell lines were a generous gift from Dr. Liang Qiao (Loyola University, Maywood, IL).
Immunohistochemical Detection of FasL in Cervical Carcinoma Cell Lines and Cervical Tissues
Slides prepared from paraffin-embedded tissues were deparaffinized in xylene and rehydrated and endogenous peroxidase was quenched in 3% hydrogen peroxide. Antigen retrieval was facilitated by heating the slides 3 times in 10 mM of citric acid buffer at 1100 watts for 3 minutes each time. Distilled water was added to keep the concentration of citric acid constant. The slides were cooled and transferred to phosphate-buffered saline (PBS). Cell lines were harvested with a combination of trypsin and ethylenediamine tetraacetic acid (EDTA) and washed in PBS, and the slides were prepared using a standard cytospin technique. Endogenous peroxidases were quenched with 3% hydrogen peroxide. Slides then were fixed in HistoCHOICE® fixative (AMRESCO, Solon, OH) for 1 hour and washed in PBS. Fixed slides from paraffin-embedded tissues or cytospin preparations were blocked with 1.5% goat serum (or horse serum for mouse monoclonal antibodies) for 30 minutes, washed, and incubated with primary antibody for 30 minutes at room temperature. The primary antibodies used were rabbit anti-Fas (CD95) ligand (C20) (Santa Cruz Biotechnology, Santa Cruz, CA) at a dilution of 1:200 and anti-Fas (CD95) ligand (monoclonal antibody 33 [MoAb33]) (Transduction Laboratories, Lexington, KY) at a dilution of 1:250. After washing, the slides were incubated with the appropriate secondary antibody (mouse or rabbit) for 30 minutes, washed, and treated with the avidin–biotin complex for 30 minutes. The peroxidase reaction was developed with diaminobenzidine tetrahydrochloride solution using the ABC Elite peroxidase kit (Vector Laboratories, Burlingame, CA). The sections were counterstained with hematoxylin. Negative controls were established for MoAb33 (Transduction Laboratories) with mouse immunoglobulin (Ig) G (Vector Laboratories) instead of the primary antibody. In addition, a negative control for C20 (Santa Cruz Biotechnology) was established by incubating the primary antibody with the blocking peptide (FasL peptide C20; (Santa Cruz Biotechnology) for 2 hours before exposing the antibody to the slides. To further test the specificity of immunostaining for FasL using the C20 and MoAb33 antibodies, we performed immunohistochemistry (IHC) on a randomly selected subset of paraffin-embedded cervical carcinoma specimens using the G247–4 antihuman FasL antibody (BD Biosciences, San Diego, CA), which was previously described for its specificity and low background staining.18 The antibody was used according to the manufacturer's instructions, at a dilution of 1:100. Slides were randomly reviewed and IHC staining of dysplasia or tumor was graded according to the intensity of the staining observed at a magnification of ×100 (0: negative; 1+: weak staining to 3+: strong staining), the percentage of cells staining (0–100%), and the subcellular localization of the staining (cytoplasmic and/or nuclear). In slides with cervical dysplasia, the thickness of the positively stained cells also was described.
Immunofluorescent Terminal Deoxynucleotidyl Transferase Biotin-dUTP Nick End Labeling Staining of Cervical Tissues
Paraffin sections were deparaffinized and antigen retrieval was performed as described for IHC. The slides then were washed in 0.2% bovine serum albumin (BSA)/PBS, and were blocked with 20% goat serum in 2% BSA/PBS for 30 minutes at room temperature. Antihuman leukocyte common antigen (LCA) (Dako Corporation, Carpinteria, CA), the primary antibody, was diluted in 2% goat serum and 2% BSA/PBS to a dilution of 1:70. The slides were incubated with the primary antibody overnight at 4 °C. The specimens then were washed in 0.2% BSA/PBS. Negative control slides were incubated without the primary antibody. Goat antimouse IgG conjugated with Texas Red dye (Jackson ImmunoResearch, West Grove, PA) diluted with 2% BSA/PBS to a dilution of 1:100 was incubated on the slides at room temperature for 1 hour. The slides were washed with 0.2% BSA/PBS and apoptosis was detected using the In Situ Cell Death Detection Assay (Roche Molecular Biochemicals, Mannheim, Germany) according to the manufacturer's instructions. The labeling reaction was performed using a solution containing terminal deoxynucleotidyl transferase and fluorescein-dUTP. The slides were coverslipped and incubated at 37 °C for 60 minutes in a humidity chamber. Terminal deoxynucleotidyl transferase was omitted from negative control slides, which were included in each run. Cells were visualized on a Zeiss Axioplan 2 fluorescent microscope equipped with a ×100 (1.40 oil immersion) Plan Apochromat lens (we used a ×100 objective lens and a ×10 ocular lens, for a total magnification of ×1000) (Carl Zeiss, Inc., Thornwood, NY). Confocal images were generated using a LSM 510 laser-scanning microscope (Carl Zeiss, Inc.).
Reverse Transcriptase-Polymerase Chain Reaction Detection of Fas-L mRNA
Reverse transcriptase–polymerase chain reaction (RT-PCR) was performed as previously described by Chappell et al.19 Briefly, total RNA from all cervical carcinoma cell lines (all 14 cell lines listed previously) and Jurkat cells was obtained using the TRIZOL method (Life Technologies, Inc., Bethesda, MD). RT with gene-specific intron-spanning primers was used to synthesize cDNA followed by PCR amplification. FasL intron spanning forward and reverse primers (FASLIGS 38 and 39; Sigma-GenoSys, The Woodlands, TX) 5′GGTTCTGGTTGCCTTGGTAGGATTG 3′ and 3′AGCCGAAAAACGTCTGAGATTCCTC 5′, respectively, generated a 566-base pair (bp) product. Primers used to amplify β-actin generated a 540-bp product that was used as a positive control. The cDNA synthesis step was performed at 45 °C for 20 minutes followed by 94 °C for 2 minutes. After cDNA synthesis, PCR amplification was performed with 40 cycles of the following: 94 °C for 15 seconds, a gradient from 42–52 °C for 30 seconds, and 72 °C for 1 minute. PCR products were analyzed on a 2% agarose gel using ethidium bromide and ultraviolet illumination for the detection of DNA fragments. To confirm the identity of the bands, the National Cancer Institute-I (NCI-I) restriction enzyme (GIBCO BRL, Gaithersburg, MD) was used because it is known to have only one restriction site (5′-CC() GG-3′, 3′-GG()CC-5′)and it cuts the PCR product at the appropriate site, giving 2 bands. All the cell lines that gave a positive band with RT-PCR were tested with the NCI-I restriction enzyme. One μL of the buffer was mixed with a 1 μL of the enzyme and they were added to 8 μL of the PCR product from the second round. After an incubation of 1 hour at 37 °C, the mixture was run on a 1% agarose gel for 35 minutes at 75 volts.
Detection of FasL by Enzyme-Linked Immunoadsorbent Assay
To further confirm the presence of the FasL protein that was detected by IHC, CaSki carcinoma cell lines used in the functional assay were tested for the protein using a FasL enzyme-linked immunoadsorbent assay (ELISA) kit (Oncogene, San Diego, CA) following the manufacturer's recommended protocol. Briefly, 50 μL of biotinylated detector antibody was put in every well, to which 100 μL of FasL antibody was added. The cell lysate was added for 3 hours and the wells then were washed with the ELISA wash buffer 3 times. Then, 100 μL of the 400X conjugate were added for 30 minutes at room temperature. This was followed by three washes with the wash buffer and the addition of tetramethylbenzidine (TMB) substrate for another 30 minutes at room temperature in the dark. The reaction was stopped using the stop solution and the absorbance was measured (wave length 450/595 nanometers).
Detection of FasL by Flow Cytometry
Cervical carcinoma cell lines were stained with the anti-FasL antibody, NOK-1 (Pharmingen, San Diego, CA) at 10 μg/mL and incubated for 30 minutes at 4 °C followed by incubation with fluorescein isothiocyanate-labeled goat antimouse IgG for 30 minutes at 4 °C. After washing, the cells were analyzed with flow cytometry using FACScan (Becton Dickinson, Franklin Lakes, NJ).
Functional Assay for Apoptosis
The functional assay for apoptosis was performed as described by Shiraki et al.20 The FasL-positive cervical carcinoma cell line, CaSki, was plated in 24-well tissue culture plates and grown to confluence. The cells then were washed with PBS and fixed with 2% paraformaldehyde at 4 °C for 30 minutes. The cells were washed with serum-free RPMI-1640 medium three times. Jurkat cells (20 × 105) were added to each well in suspension. After 48 hours of coculture, the media in each well, along with the floating Jurkat cells, was removed and centrifuged. The cell pellets then were fixed with 70% methanol overnight at -20 °C. After centrifugation, the cells were resuspended in 150 μL of propidium iodide solution that was RNase and DNase free (Roche Molecular Biochemicals) and incubated at room temperature in the dark for 30 minutes. A total of 20,000 events were measured per sample, using a FACScan flow cytometer (Becton Dickinson). The percentage of apoptotic cells was determined by evaluating the sub-2N DNA content. Jurkat cells grown in 24-well culture plates without CaSki cells were used for measuring the rate of spontaneous apoptosis in the Jurkat cells. In addition, Jurkat cells incubated with 500 ng of the agonistic anti-Fas antibody clone CH-11 (MBL International, Watertown, MA) were used as a positive control. FasL-negative SW756 cells were grown in a 24-well plate and used as a negative control to FasL-positive CaSki cells. To demonstrate that this apoptosis was Fas/FasL mediated, Jurkat cells were cocultured with fixed CaSki cells in the presence of a neutralizing anti-FasL monoclonal antibody (NOK-1 and 4H9; BD Pharmingen, and MBL International, respectively) or an isotype IgG control.
To confirm the results we obtained from the functional assay, we repeated the experiment using an APO-BRDU flow cytometry kit (Sigma-Aldrich Corporation St. Louis, MO) for apoptosis. Briefly, the cells were cocultured as mentioned earlier and then fixed using 1% paraformaldehyde (PFA), washed, and refixed with 70% ethanol. The DNA labeling solution was prepared (reaction buffer, TdT enzyme, Br-dUTP, and distilled water) and added to the fixed cells for 60 minutes in a 37 °C water bath. After washing the cells, they were resuspended using the anti-BrDU-fluorescein in rinsing buffer and incubated in the dark for 30 minutes. Finally, propidium iodide/RNase A solution was added and the cells were analyzed within 3 hours of staining. The results were compared with the positive and negative control cells provided with the kit by the manufacturer.
Detection of Fas-L in Cervical Tumor Tissues by IHC
To check for FasL expression in cervical carcinoma specimens, 87 paraffin-embedded cervical carcinoma tissues were stained with an anti-FasL monoclonal antibody (MoAb33); 80 of the 87 sections (92%) stained positive (Table 1). FasL staining was observed in the cytoplasm, but not in the nucleus of all the positive samples. The tumor samples demonstrated 50–80% staining of the tumor cells (Fig. 1A). The surrounding stromal tissues stained negative for FasL and therefore served as an internal negative control. Mouse IgG purified from pooled serum and containing a spectrum of the IgG subclasses was used as a negative control (Fig. 1B). Similar results were obtained when the polyclonal antibody was used (Fig. 1C) and preincubation with the immunizing peptide was found to successfully block the antibody and therefore served as a negative control to demonstrate the antibody's specificity (Fig. 1D). We also observed increased staining in the nonneoplastic cells when using the C20 antibody (Fig. 1C) compared with MoAb33 (Fig. 1A). As mentioned previously, some of the stromal immunoreactivity was the result of lymphocytes. However, there was some nonspecific stromal staining as well that exceeded that observed with the MoAb33. We attribute this finding to differences between the antibodies. Although the C20 antibody demonstrated strong immunoreactivity in neoplastic cells, it also consistently demonstrated more nonspecific background staining than the MoAb33 antibody. We believed the preincubation studies helped to show this by demonstrating that tumor and lymphocyte-associated staining were attenuated, whereas background stromal (nonspecific) staining remained.
|Specimen no.||FasL intensity;||% tumor cells staineda||Cytoplasmicb||Nuclearb||Primary vs. metastatic sitec|
To further confirm the specificity of immunostaining for FasL using the C20 and MoAb33 antibodies, IHC was performed on a subset of paraffin-embedded cervical carcinoma specimens using the G247–4 antihuman FasL antibody.18 A comparison of the pattern and intensity of staining demonstrated no significant differences between the C20, MoAb33, and G247–4 antibodies in the subset of cervical carcinoma tissues tested.
Detection of Apoptosis in Tumor-Infiltrating Leukocytes in Tumor Specimens
We next wanted to evaluate whether tumor-infiltrating leukocytes in cervical tumors expressing FasL undergo apoptosis. Immunofluorescent staining using the TUNEL assay was used to test for apoptosis and anti-LCA antibody was used to detect leukocytes in tumor specimens. Seven of 10 FasL-positive tumors (70%) demonstrated apoptosis in tumor-infiltrating leukocytes using double staining with both TUNEL and anti-LCA (Fig. 2). Conversely, only one of four tumors found to be negative for Fas-L demonstrated double staining in tumor-infiltrating leukocytes. Although we believe this represents a trend, the small number of tumors that were not found to express FasL prevents adequate statistical power for comparison.
Expression of Fas L in Cervical Carcinoma Cell Lines
To test the functional correlation between the expression of FasL in cervical carcinoma and the induction of apoptosis in T cells, we divided the carcinoma cell lines used into FasL-positive and FasL-negative cell lines. We conducted an in vitro functional assay after testing the 14 available, established cervical carcinoma cell lines for the presence of FasL mRNA expression by RT-PCR.19 An expected, a 566-bp product was found to be present in 10 of the 14 cell lines tested, thereby confirming the presence of FasL expression in those carcinoma cell lines (only the results of the CaSki and SW756 cervical carcinoma cell lines are shown in Figure 3). β-actin oligonucleotides were used to detect a 540-bp RNA band and were used as a control. To confirm that the 566-bp band is the appropriate FasL band, we performed restriction digestion using the NCI-I restriction enzyme. Because FasL has only one restriction site, digesting that band with the NCI-I should result in two bands, as we observed. Testing the expression of FasL on the surface of the cervical carcinoma cell lines was performed using flow cytometry after staining with the NOK-1 MoAb. Figure 3B shows that FasL is expressed on the surface of CaSki carcinoma cells but not the SW756 carcinoma cells. These results correlate with the results of RT-PCR shown in Figure 3A.
Induction of Lymphocyte Apoptosis by Cervical Carcinoma Cells
To test whether the expression of FasL contributes to the induction of apoptosis shown in T cells, we used an in vitro apoptosis assay. In this assay, we used the CaSki cervical carcinoma cell line as a representative of a FasL-expressing group and the SW756 cervical carcinoma cell line was used as an example of a FasL-negative group. The two cell lines were used to test the induction of apoptosis in the Jurkat cell line. Only those Jurkat cells cultured with the CaSki cervical carcinoma cells demonstrated a 3–8.5 fold increase in the number of apoptotic cells compared with the Jurkat cells cultured alone or with the SW756 cell line. This was comparable to the amount of apoptosis induced by adding the agonistic anti- Fas antibody to Jurkat cells. This increase in apoptotic cells was found to be blocked efficiently by the addition of the neutralizing anti-FasL antibodies NOK-1 and 4H9 (Fig. 3c). Using the TdT Br-dUTP for labeling DNA breaks and the propidium iodide (PI) for counterstaining the total DNA, we were able to confirm the same findings. When the Jurkat cells were cocultured with the CaSki cervical carcinoma cells for 48 hours, the percentage of apoptotic cells increased from 2% (noted in Jurkat cells alone) to 17.4%. This increase was blocked by the addition of the antagonistic anti-FasL antibody NOK-1 and not with the addition of an isotype control. Using the FasL-negative cell line (SW756) instead of the CaSki cervical carcinoma cell line resulted in no apparent increase in the apoptotic cells (Fig. 4).
Fas-L is Expressed in Normal Cervical Tissue and CIN
Because FasL expression was found in the majority of invasive cervical carcinomas, the question arose as to whether it was expressed in normal cervical tissue and cervical dysplasia. Fifty specimens of both low and high grade squamous intraepithelial lesions were examined with IHC for FasL expression, using MoAb33 (Table 2). Forty of the 50 samples (80%) expressed FasL in the areas of dysplastic epithelium. The immunostaining was cytoplasmic and approximately 50–80% of the dysplastic cells were found to be positive (Fig. 5).
|Pathology no.||FasL intensity||% of tumor cells stained||Cytoplasmica||Nucleara||Cell layers in CIN staining positive||CIN grade|
|12||++||80%||+||Basal area of CIN||1|
|13||+||50%||+||Basal area of CIN||1|
|18||+/++||50%||+||Basal to full thickness||2|
|19||+||50%||+||Basal to 50%||2|
|22||+/++||50%||+||Basal to 75%||3|
|25||+/++||80%||+||Areas of CIN unstained||3|
|26||+||50%||+||Included in CIN||3|
|29||++||80%||+||Basal to level of CIN||3|
|33||+/++||50%||+||CIN-1 1+, CIN-2/3 1–2+||3|
|36||+||< 50%||+||CIN-3 unstained, CIN-2 stained||3|
|37||++/+++||80% basal||+||Included in CIN||3|
|43||+/++||+||Basal to higher areas with CIN||1–2|
|44||++||All CIN||+||All CIN||1–3|
|45||+/++||70%||+||Basal to 75%||1–3|
|47||++||80%||+||Basal to level of CIN||1–3|
|48||+/++||80%||+||Included in CIN||2–3|
|49||+/++||50%||+||Less staining but basal||2–3|
|50||+/++||50%||+||Basal to 50%||2–3|
The pattern of staining described by the manufacturers of the FasL antibodies used in the current study was membranous and cytoplasmic. Although both patterns were observed, we did find that the predominant staining pattern was cytoplasmic. We agree that a stronger membranous pattern might be expected. However, on reviewing other similar publications regarding the expression of FasL in neoplastic cells, we are confident that the relative abundance of membranous versus cytoplasmic staining noted in the current study is similar to that described in other publications.21 Mouse IgG was used as a negative control and demonstrated no staining (data not shown). The areas of normal cervical tissues in biopsies demonstrating dysplasia demonstrated staining for FasL that was confined only to the basal cell layer of the epithelium; however, mature squamous cells were found to demonstrate no FasL expression.
FasL is a membrane protein capable of inducing apoptosis in target cells on coupling with its Fas receptor. The Fas-FasL system plays an important role in activated T cell apoptosis. Through the induction of apoptosis in T cells, the Fas-FasL system has been shown to play a role in eliminating harmful immunologic responses.22 The identification of FasL in normal tissues (testis, anterior chamber of the eye and brain) has suggested a role in immune-privileged sites.23
The Fas-FasL pathway may, at least in part, play a role in tumor evasion of the immune system by inducing Tcell apoptosis in the tumor-infiltrating lymphocytes. Expression of FasL has been reported in various malignancies, including ovarian, endometrial, breast, lung, gastrointestinal (esophageal, gastric, and colon), head and neck, renal cell, prostrate, and skin tumors.12, 14, 15, 24–26 The majority of malignant tissues examined were found to express FasL and some studies reported FasL-induced apoptosis of lymphocytes in vitro by malignant cell lines.13 It has been reported that activated human T cells and natural killer cells release bioactive Fas ligand in microvesicles.27, 28 Newly synthesized FasL is stored in lysosomes and the delivery to the surface is controlled by polarized degranulation. The microvesicles (measuring 100–200nm in greatest dimension) then are secreted by the activated T cells and are detected in the supernatant fluid.27, 28 Those microvesicles also were observed in melanoma cells and Western blot analysis demonstrated that melanosomes isolated from these cells expressed FasL.29 The expression of FasL was tested in six cervical tumors by Contreras et al., and the authors found that all six tumors expressed FasL.17 We investigated the expression of FasL in cervical carcinoma specimens and in biopsies with varying degrees of cervical dysplasia to determine whether it can play a role in the mechanism of tumor escape from the immune response. In addition, we tested FasL expression in normal cervical tissues to determine the normal distribution of FasL expression. We also investigated whether FasL expression in a cervical carcinoma cell line was capable of inducing apoptosis of cocultured lymphocytes. The current study data showed that FasL is expressed in the majority of invasive cervical tumors (94%). In the majority of tumors, large percentages (approximately 50–80 %) of individual tumor cells were found to stain positively for FasL. The cytoplasmic staining, the intensity of the stain, and the percentage of tumor cells stained in cervical carcinoma specimens tested in the current study were found to be well correlated with previously described IHC studies of other malignancies.13, 20, 30–32 In the current study, we also stained a total of 22 tumor specimens using the C20 rabbit polyclonal antihuman FasL antibody (Santa Cruz Biotechnology), which was reported to be more specific than MoAb33.18 These 2 antibodies correlated in 21 of the 22 samples tested. The 3 negative samples were negative using both stains and 17/18 positive samples correlated. The C20 antibody was found to result in more background staining (Figs. 1A, C). The immunizing blocking peptide (FasL C20, Santa Cruz Biotechnology) competed effectively with the C20 antibody, demonstrating the specificity of that polyclonal antibody to FasL (Fig. 1D).
In their study, Bennett et al. demonstrated that leukocytes undergo apoptosis in areas of FasL-positive esophageal tumors.32 Apoptosis of leukocytes detected by immunofluorescence as reported herein for cervical carcinoma is consistent with and further supports this hypothesis. More FasL-positive tumors demonstrated leukocytic apoptosis in comparison to FasL-negative tumors (7 of 10 tumors vs.1 of 4 tumors). Only five tumors were found to be negative for FasL, which prevented adequate power for statistical analysis to compare FasL-positive with FasL-negative tumors. However, there was a trend in the association between leukocytic apoptosis and FasL expression. The presence of leukocytes undergoing apoptosis in FasL-negative tumors further suggests that other mechanisms of escape from the immune system are functioning in these tumors. In addition, in those tumors that express FasL but demonstrate no apoptotic leukocytes, the reason might be that the leukocytes do not express Fas on their surface, or Fas might not be active or “functional,” which warrants further studying. On further investigation of the role of FasL expression in the development of cervical carcinoma, extensive expression of FasL also was found in CIN specimens. Only the dysplastic cells, not the differentiated cells, were found to stain positively for FasL. The positive staining was found to parallel the grade of CIN, with FasL-positive cells essentially replacing the entire epithelium in high-grade CIN or carcinoma in situ. In addition, we also found that in normal portions of squamous epithelium (stained in the same specimens as the CIN) and in stained normal cervical specimens, the basal cell layer stained positive for FasL, but the higher, more differentiated cell layers did not. The selective staining for FasL of the basal cell layer, in contrast to the rest of the normal cells in the cervix, suggests a selective down-regulation of FasL in differentiated cells. This indicated that the basal layer of the epithelium may form an “immune sanctuary” or “immune-privileged site” to keep inflammatory cells from destroying the cervical epithelial “stem cells” during recurrent episodes of cervicitis. FasL expression is observed in all cells of the basal layer, not only the stem cells. We believe this occurs because the stem cells require an intact basal layer to function properly.
The data from the current study also indicate that dysplastic cells may either lose the ability to down-regulate the FasL expression that usually occurs in more differentiated cells of the squamous epithelium or FasL might be a unique property of the basal cells and, because basaloid cells expand in the epithelium during the neoplastic process, FasL expression is maintained. As preinvasive lesions develop, there is loss of down-regulation of FasL that may facilitate an escape from the immune response. The mechanism of this down-regulation will be explored in future studies.
A previous study suggested the Fas-FasL pathway induced apoptosis in CTLs, which were specific for a cultured cervical carcinoma cell line (HeLa).17 In the current study, we demonstrated that culturing Fas-sensitive Jurkat cells with FasL-positive CaSki cervical carcinoma cells led to the induction of apoptosis in Jurkat cells comparable to the level of induction obtained when using the agonistic anti-Fas antibody clone CH11. These results were not noted when we used the FasL-negative SW756 cervical carcinoma cell line or when we blocked FasL with the blocking anti-FasL antibody. These results are consistent with data reported in renal cell carcinoma specimens,25 and suggest that FasL may play a role in tumor cell induction of lymphocyte apoptosis. Zeytun et al.33 postulated that the presence of the constitutive expression of both Fas and FasL in tumor-specific CTLs and tumor cells suggests that the survival of tumors in the host depends on which cell type can more efficiently induce apoptosis. The underlying consistency in studies regarding the Fas-FasL system in malignant tumors is that it is a normally occurring system that undergoes modification in the malignant counterpart. These modifications, which include an increase in the levels of FasL, allow an advantage toward the development of tumor growth and spread by escaping from immune surveillance and the facilitation of tissue destruction. FasL expression in CIN as demonstrated in the current study supports the hypothesis of Zeytun et al.33 The extensive expression of FasL noted in CIN suggests that the Fas-FasL system plays a role in tumor cells' escape from immune surveillance in cervical carcinoma, and its involvement occurs early in the disease process. CIN may escape immune surveillance when the transformed cells fail to down-regulate their expression of FasL.
The expression of FasL in CIN, as well as in the majority of cervical tumor samples, implies that the Fas-FasL system plays a role in the escape of tumor cells from the immune system. Initially, this may allow a preinvasive lesion (e.g., CIN) to develop into an invasive tumor by overwhelming the host's immune system. The extensive staining noted in the invasive tumors in the current study also suggests that the same system may be one of the mechanisms that immunotherapy needs to overcome.