The Fas Counterattack In Vivo: Apoptotic Depletion of Tumor-Infiltrating Lymphocytes Associated with Fas Ligand Expression by Human Esophageal Carcinoma

Tissues

Human esophageal carcinomas (squamous, n = 6; adenocarcinoma, n = 2) were collected following surgical resection of esophageal carcinomas (n = 7) or investigative endoscopic biopsy (n = 1) at Mercy Hospital (Cork, Ireland), following a protocol approved by University Teaching Hospital’s Ethics Committee. None of the patients had received chemo-, radio-, or immunotherapy before tissue collection.

Immunohistochemical detection of FasL and FasR

Paraffin-embedded sections of esophageal tumors were deparaffinized in xylene and rehydrated before analysis. Slides were washed twice for 5 min in a wash buffer containing 50 mM Tris-Cl, pH 7.6, 50 mM NaCl, and 0.001% saponin, and endogenous peroxidase was quenched with 3% hydrogen peroxide in methanol for 5 min. Slides were washed as before, except that the wash buffer for this and all subsequent steps included 1% normal goat serum. Slides were then blocked for 1 h in wash buffer containing 5% normal goat serum. Slides were washed and incubated overnight at 4°C with a rabbit polyclonal anti-human FasL-specific IgG (Santa Cruz Biotechnology, Santa Cruz, CA) at 0.2 μg ml⁻¹ in wash buffer. Ab binding was localized using a biotinylated secondary Ab, avidin-conjugated horseradish peroxidase, and diaminobenzidine substrate, contained within the Vectastain ABC detection kit (Vector Laboratories, Burlingame, CA). The immunizing peptide (FasL, N-terminal amino acids 2–19; Santa Cruz Biotechnology) was included at 2 μg ml⁻¹ during primary Ab incubation in control staining. FasR was immunohistochemically detected using the same protocol, except that a rabbit polyclonal anti-human FasR-specific IgG (Santa Cruz Biotechnology) was used, along with its corresponding immunizing peptide (FasR, amino acids 316–335; Santa Cruz Biotechnology) in control staining. Slides were counterstained with hematoxylin.

Generation of a FasL-specific RNA probe (riboprobe)

A digoxigenin-labeled RNA hybridization probe (344 bp) was generated corresponding to codons 96–210 of the human FasL cDNA sequence. The riboprobe was synthesized by in vitro transcription using digoxygenin-11-UTP and T7 RNA polymerase (Boehringer Mannheim, Mannheim, Germany). Template for the in vitro transcription reaction was generated by PCR amplification of a fragment (codons 96–210) of FasL cDNA using a proofreading thermostable polymerase (UlTma DNA polymerase; Perkin-Elmer, Norwalk, CT) and an antisense primer to which a T7 promoter sequence was added.

RNA was isolated from FasL-expressing cells by homogenization in guanidine thiocyanate (Sigma, St. Louis, MO), followed by phenol extraction and ethanol precipitation. cDNA was synthesized using the AMV reverse transcriptase (Promega, Madison, WI) and random hexanucleotide primers (Boehringer Mannheim).

PCR was performed on the cDNA using the following sense and antisense primers, respectively: FasL, GGATTGGGCCTGGGGATGTTTCA and [pT7]-TTGTGGCTCAGGGGCAGGTTGTTG.

PCR primers were designed using the DNASTAR Lasergene Primerselect program (DNAstar, Madison, WI). Primer pairs were chosen to span introns in the FasL genomic sequence, thus ensuring mRNA-specific amplification. Primers were selected that showed no significant homology to any other genes in the EMBL DNA sequence database.

Thermal cycling (40 cycles) was as follows: denaturation at 96°C for 15 s, annealing at 55°C for 30 s, and extension at 72°C for 3 min. Primers were used at a final concentration of 0.1 μM each, dNTPs at 50 μM, and MgCl₂ at 1.5 mM. One unit of UlTma DNA polymerase was used per 50-μl reaction. PCR product specificity was confirmed by restriction mapping.

Using this PCR-amplified FasL cDNA fragment as template, a riboprobe was synthesized by in vitro transcription using digoxygenin-11-UTP and T7 RNA polymerase (Boehringer Mannheim), according to the manufacturer’s instructions. The nucleotide sequence of the FasL probe showed no significant homology to any other sequence in the EMBL DNA sequence database. An unlabeled riboprobe was also synthesized for use in competitive control hybridizations.

Localization of FasL mRNA expression by in situ hybridization

In situ hybridization was performed on paraffin-embedded human esophageal tumor sections (4 μm thick), mounted on aminopropylethoxysilane (APES)-treated slides. Prehybridization treatments involved washing twice 5 min each in 1) PBS, 2) PBS and 0.1 M glycine, 3) PBS and 0.3% Triton X-100, and 4) PBS again. Sections were digested for 30 min at 37°C with proteinase K (10 μg/ml in 100 mM Tris-HCl, 50 mM EDTA, pH 8), fixed for 5 min at 4°C in 4% paraformaldehyde and PBS, and then acetylated twice for 5 min in fresh 0.25% acetic anhydride and 0.1 M triethanolamine (pH 8). Sections were incubated at 37°C for 10 min in a prehybridization buffer consisting of 50% deionized formamide in 4× SSC. Hybridization was performed at 42°C overnight in hybridization buffer (50% formamide, 10% dextran sulfate, 1× Denhardt’s reagent, 4× SSC, 10 mM DTT, 500 μg/ml yeast tRNA, and 1 mg/ml heat-denatured herring sperm DNA) containing 1 ng/μl digoxigenin-labeled riboprobe. After hybridization, tissues were washed with increasing stringency to 0.1× SSC at 37°C. Hybridized probe was detected immunologically using alkaline phosphatase-conjugated sheep anti-digoxigenin Ab (Boehringer Mannheim) and visualized with nitroblue tetrazololium/5-bromo-4-chloro-3-indolyl phosphates (NBT/BCIP; purple/black precipitating product). Control slides involved competitive inhibition of hybridization by adding a 10-fold excess of unlabeled riboprobe to the digoxigenin-labeled riboprobe before hybridization. This resulted in a marked reduction of signal intensity, thus confirming the specificity of hybridization.

CD45 (leukocyte common Ag) staining

To identify tumor-infiltrating immunocytes, CD45 (leukocyte common Ag) staining was performed on paraffin-embedded esophageal tumor sections. Following deparaffinization and rehydration, sections were pretreated by microwave irradiation in 0.01 M citrate buffer for 5 min at 370W. The sections were cooled rapidly by immersing in 0.1 M PBS. The slides were then incubated with a mouse anti-human CD45 monoclonal IgG (Dako, Carpenteria, CA) at a dilution of 1/70 for 1 h. All incubations were conducted at room temperature. Next the slides were washed for 5 min in 0.1 M Tris-buffered saline (TBS), which was used for all washes. Following incubation with a secondary rabbit anti-mouse IgG (Dako) at a dilution of 1/25 for 30 min, washing was repeated. Sections were then incubated for 30 min with alkaline phosphatase-conjugated anti-alkaline phosphatase (APAAP) complex (Dako), at a dilution of 1/50. Following washing, the secondary Ab and APAAP complex incubations were repeated (10 min each), to enhance staining. Sections were then incubated for 10 min with an alkaline phosphatase substrate solution (fast blue; Sigma). CD45-positive immunocytes appeared blue when viewed under light microscopy, and were almost exclusively of lymphoid morphology (i.e., TIL).

Immunocyte cell death detection by CD45/TUNEL dual staining

Cell death was detected in situ in resected esophageal tumors by enzymatic labeling of DNA strand breaks using TUNEL (terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling) (Boehringer Mannheim), according to the manufacturer’s instructions. CD45/TUNEL dual staining was performed to allow apoptotic tumor-infiltrating immunocytes to be identified and enumerated. CD45 staining was performed first on esophageal tumor sections, as described. Following treatment with proteinase K (20 μg/ml in 10 mM Tris-Cl, pH 7.6) for 30 min, sections were washed in PBS and endogenous peroxidase activity was blocked in 3% hydrogen peroxide in methanol for 30 min. Next, the slides were washed. This and all subsequent washes were in PBS. The sections were treated with permeabilization solution (0.1% Triton X-100 in 0.1% sodium citrate) for 15 min. After washing, the labeling reaction was performed using a solution containing terminal deoxynucleotidyl transferase, its buffer, and fluorescein-dUTP. During this step, slides were coverslipped and incubated at 37°C for 60 min in a humidity chamber. Terminal deoxynucleotidyl transferase was omitted from negative control slides, which were included in each run. To localize cells containing labeled DNA strand breaks, sections were washed and incubated with a sheep anti-fluorescein Ab Fab fragment conjugated with horseradish peroxidase at 37°C in a humidity chamber for 30 min. Following washing, TUNEL-positive color development (brown) was obtained by incubating the sections with a diaminobenzidine substrate solution for 15 min. When viewed under light microscopy, CD45 single-positive cells stained blue, while CD45/TUNEL dual-positive cells appeared with brown nuclear staining and blue cytoplasmic/cell surface staining.

Cell counting and labeling indices

To quantify CD45-positive TIL infiltration of FasL-positive vs FasL-negative tumor regions, stained tumor sections were analyzed under light microscopy, as follows. FasL-positive and FasL-negative areas were located on a FasL-stained esophageal tumor section by one investigator. Tumor areas staining with high intensity similar to or greater than that of nearby FasL-positive lymphoid cells were regarded as strongly positive. Tumor areas staining with weak intensity close to the background staining observed in the corresponding peptide control slide were regarded as negligible/negative. A reference slide was used to assure consistent selection of FasL strongly positive or negligible/negative tumor areas. Areas of tumor ulceration were excluded. A consecutive, CD45-stained slide from the same tumor was superimposed on the FasL-stained slide. Using histologic landmarks, the corresponding FasL-positive and FasL-negative areas were located on this slide. The FasL-stained section was removed, and a second investigator, blinded as to the local status of FasL expression, counted the number of CD45-positive TIL per 2000 total nuclei in the area located by the first investigator. A similar approach was used to enumerate CD45/TUNEL dual-positive TIL within FasL-positive vs FasL-negative areas of esophageal tumors. For each type of staining, all slides were stained in a single experiment.

Labeling indices for TIL infiltration were expressed as the percentage of CD45-positive cells per 2000 total nuclei counted. Labeling indices for TIL apoptosis were expressed as the percentage of CD45/TUNEL dual-positive cells per 500 total CD45-positive cells counted.

Esophageal carcinomas express FasL

Using a FasL-specific rabbit polyclonal IgG, FasL-positive immunohistochemical staining was detected in all esophageal tumors (squamous, n = 6; adenocarcinoma, n = 2) (Fig. 1⇓). FasL-binding specificity was confirmed since staining was inhibited by inclusion of the immunizing peptide as a competitive inhibitor before primary Ab incubation. The staining within the tumors varied from strongly positive to negligible/negative. Although FasL-positive and FasL-negative areas usually occurred within the same tumor, in all specimens at least 70% (range 70–100%, n = 8) of the tumor tissue stained strongly positive. FasL expression was confirmed by detection of FasL mRNA in consecutive tumor sections using in situ hybridization. FasL mRNA detection corresponded closely with immunohistochemical staining for FasL protein (Fig. 1⇓). Hybridization was specific, as a 10-fold excess of unlabeled riboprobe competitively inhibited hybridization of labeled riboprobe in control sections.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Human esophageal carcinomas express Fas ligand. Immunoperoxidase staining using a FasL-specific rabbit polyclonal IgG was performed on paraffin-embedded esophageal carcinoma sections. Slides were counterstained with hematoxylin. A, FasL-positive immunohistochemical staining (brown) is shown in a representative squamous esophageal tumor island. C, As a control for specificity of Ab detection, the FasL-immunizing peptide was included during primary Ab incubation. Competitive inhibition of staining by the immunizing peptide confirms FasL specificity. Tumor sections sequential to those shown in A and C were used to detect FasL mRNA by in situ hybridization, using a digoxigenin-labeled FasL-specific riboprobe. B, Positive purple hybridization signal was obtained from the same tumor island that stained immunohistochemically positive for FasL. D, In a control hybridization, a 10-fold excess of unlabeled probe caused direct, competitive displacement of the labeled probe, confirming the specificity of hybridization. These results are representative of six squamous and two adenocarcinomas of the esophagus.

Using immunohistochemistry, all of the esophageal tumors were found to express FasR essentially homogenously throughout the tumor mass. Specificity of detection was confirmed since staining was inhibited by inclusion of the FasR-immunizing peptide as a competitive inhibitor before primary Ab incubation. Coexpression of FasL and FasR, which occurred over large areas of the tumors, did not result in an enhanced rate of tumor cell apoptosis, as assessed by TUNEL (not shown).

Reduced TIL infiltration of FasL-expressing esophageal tumor islands

All eight esophageal tumors showed infiltrates of cells staining immunohistochemically positive for CD45 (leukocyte common Ag), which were almost exclusively of lymphoid morphology, permitting identification of TIL (Fig. 2⇓). Tonsil sections included as positive controls stained strongly, and omission of the primary Ab abolished staining (data not shown). FasL expression in the eye (10) and testis (11) has been shown to preclude infiltration of these organs by proinflammatory, Fas-sensitive leukocytes, thereby maintaining a state of immune privilege at these locations. To evaluate whether FasL expression by esophageal tumor cells limited immune effector cell infiltration of the tumor, TIL numbers infiltrating FasL-positive vs FasL-negative tumor areas were enumerated and compared (Table I⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Apoptotic depletion of TIL in FasL-expressing regions of esophageal carcinoma. Immunoperoxidase staining (brown) using a FasL-specific Ab allowed identification of FasL-expressing (A) and FasL-negative (B) tumor regions in close proximity within individual esophageal carcinomas. FasL expression was confirmed by in situ hybridization (Fig. 1⇑). On a consecutive tumor section, apoptotic TIL were identified by CD45-TUNEL dual staining. C, A high proportion of TIL in contact with the FasL-expressing tumor tissue shown in A is apoptotic, exhibiting brown TUNEL-positive nuclear staining with blue CD45-positive cytoplasmic/cell surface staining. E, At higher magnification, condensation and fragmentation of TUNEL-positive nuclei can be seen against a blue, CD45-staining cytoplasmic background (arrow), confirming apoptotic TIL. D, Apoptosis is less frequent among TIL in contact with the FasL-negative tumor region shown in B. TUNEL staining is largely absent from these blue, CD45-stained TIL, seen at higher magnification in F. There is a striking lack of infiltration of the FasL-expressing tumor region by viable TIL (C) compared with the FasL-negative tumor region (D), which shows infiltration by blue, CD45 single-positive, nonapoptotic TIL (arrows). G and H show another consecutive section from the same tumor stained for CD45 alone. CD45 single staining (blue) of cells observed in the same areas as E and F, respectively, confirms the identity of TIL. These results are representative of five squamous and one adenocarcinoma of the esophagus. Quantitative results for both TIL infiltration and apoptosis in FasL-positive vs FasL-negative tumor regions are shown in Tables I and II, respectively.

Of the eight esophageal tumors examined, one squamous carcinoma expressed FasL over 100% of the tumor area, so comparison between FasL-positive vs FasL-negative areas was not possible. One biopsied adenocarcinoma did not contain sufficient tumor tissue to allow significant comparison between FasL-positive vs FasL-negative areas. In all six esophageal tumor sections that were amenable to analysis, a significant reduction of tumor infiltration by TIL was observed in FasL-expressing tumor regions (Fig. 2⇑). FasL expression by the tumors was associated with a statistically significant, mean fourfold reduction in TIL (range 1.6- to 12-fold, n = 6, p < 0.05; Wilcoxon Signed Ranks) when compared with FasL-negative areas within the same tumors. Thus, the expression of FasL by esophageal cancer cells appears to have a dramatic inhibitory effect on TIL infiltration of the tumor.

Increased cell death of TIL associated with FasL expression by esophageal cancers

FasL expressed in the eye (10) and testis (11) causes apoptosis of infiltrating Fas-sensitive immune effector cells, precluding proinflammatory immune responses and thus maintaining immune privilege in these organs. We wished to investigate whether the diminished TIL infiltration observed in FasL-expressing esophageal tumor tissue was a result of apoptotic depletion. CD45/TUNEL dual staining was used to identify apoptotic TIL in the esophageal cancer sections (Fig. 2⇑). CD45-positive TIL were first stained immunohistochemically (blue), followed by detection of DNA strand breaks on the same tumor section using TUNEL (brown). Dual-stained, apoptotic TIL therefore appeared as cells with brown TUNEL-positive, condensed, or fragmented nuclei surrounded by blue cytoplasmic/cell surface CD45 staining. TIL in immediate contact with tumor cells were examined, and the number that was CD45/TUNEL dual positive was expressed as a percentage per 500 total CD45-positive TIL counted in the same area. Levels of apoptotic TIL were quantified in FasL-positive vs FasL-negative areas of the same tumor (Table II⇓).

A statistically significant, mean twofold increase in TUNEL-positive cell death of CD45-positive TIL was observed in FasL-expressing areas of the esophageal tumors relative to FasL-negative areas (range 1.6- to 2.4-fold, n = 6, p < 0.05; Wilcoxon signed ranks). As a comparative control, the cell death levels of extratumoral leukocytes in the surrounding stroma were also quantified (data not shown). In the stroma, the levels of TUNEL-positive immunocytes (mean 17.1%, n = 4) were similar or slightly less than those in FasL-negative areas of tumor (mean 23.6%, n = 6). Therefore, infiltration of stromal lymphocytes into FasL-negative tumor regions resulted in only a marginal increase in cell death. The twofold increase in TIL apoptosis associated with FasL-expressing tumor tissue suggests that Fas-mediated apoptosis of TIL triggered by tumor-expressed FasL may contribute to the depletion of CD45-positive TIL observed in these tumor regions.