Specific ligands of the peripheral benzodiazepine receptor induce apoptosis and cell cycle arrest in human colorectal cancer cells

The peripheral benzodiazepine receptor (PBR) has been implicated in growth control of various tumour models. Although colorectal cancers were found to overexpress PBR, the functional role of PBR in colorectal cancer growth has not been addressed to date. Using primary cell cultures of human colorectal cancers and the human colorectal carcinoma cell lines HT29, LS174T, and Colo320 DM we studied the involvement of PBR in the growth control and apoptosis of colorectal cancers. Both mRNA and protein expression of PBR were detected by RT-PCR and flow cytometry. Using confocal laser scanning microscopy and immunohistochemistry the PBR was localized in the mitochondria. The specific PBR ligands FGIN-1-27, PK 11195, or Ro5-4864 inhibited cell proliferation dose-dependently. FGIN-1-27 decreased the mitochondrial membrane potential, which indicates an early event in apoptosis. Furthermore, FGIN-1-27, PK 11195 or Ro5-4864 increased caspase-3 activity. In addition to their apoptosis-inducing effects, PBR ligands induced cell cycle arrest in the G 1/G 0-phase. Thus, our data demonstrate a functional involvement of PBR in colorectal cancer growth and qualify the PBR as a possible target for innovative therapeutic approaches in colorectal cancer. © 2001 Cancer Research Campaign http://www.bjcancer.com

Tumour cell growth is characterized by an imbalance between cell division and cell death. A common step in tumorigenesis is the loss of the capability of cells to undergo apoptosis. The localization of PBR in the mitochondrial membrane and its implication in the permeability transition pore (Zorov, 1996;Fennell et al, 2001) suggest that the PBR takes part in the regulation of the mitochondrial permeability and induction of apoptosis. However, the role of PBR in the regulation of apoptosis is not yet understood. Specific PBR ligands were shown to either induce apoptosis directly (Marchetti et al, 1996a;Tanimoto et al, 1999;Fischer et al, 2001), or to facilitate apoptosis by inhibiting the antiapoptotic effects of BCL-2 (Hirsch et al, 1998;Larochette et al, 1999;Ravagnan et al, 1999). In contrast, apoptosis-protective effects of PBR ligands have been reported as well (Bono et al, 1999). Besides their putative role in the regulation of apoptosis,

Specific ligands of the peripheral benzodiazepine receptor induce apoptosis and cell cycle arrest in human colorectal cancer cells
PBR ligands were shown to induce cell cycle arrest in the G 1 /G 0and G 2 /M-phase in breast carcinoma and melanoma cell lines (Landau et al, 1998;Carmel et al, 1999). Similarly, the new benzazepine BBL22, classified as a PBR-specific ligand, induced arrest in the G 2 /M-phase in different tumour cell lines of epithelial or haematopoietic origin followed by an induction of apoptosis (Xia et al, 2000).
As PBR was shown to be overexpressed in colorectal tumours (Katz et al, 1990b), PBR may well play a functional role in colorectal carcinogenesis. However, an involvement of PBR in colorectal cancer growth has not been investigated so far. In this study we show that specific PBR ligands inhibit proliferation of colorectal cancer cells, which is associated with an induction of apoptosis and cell cycle arrest.

Cell culture
The human colorectal adenocarcinoma cell lines HT29 and Colo320 DM (neurorendocrine-differentiated) were grown in RPMI 1640 medium supplemented with 10% fetal calf serum. The human colorectal adenocarcinoma cell line LS174T was grown in Dulbecco's minimal essential medium supplemented with 10% fetal calf serum. Cell lines were cultured in a humidified atmosphere containing 5% CO 2 at 37˚C .
Surgically resected specimens of primary colorectal carcinomas (3 rectum, 3 colon) were obtained from 4 female and 2 male patients who underwent surgery in the Department of Surgery, Benjamin Franklin University Hospital, Free University Berlin. The human tumour material was used according to the standards set by the Ethical Committee of the Benjamin Franklin University Hospital, Free University of Berlin. The age of the patients ranged from 23 to 78 years. Primary cell cultures were prepared by mechanical dissection using a Medimachine with 50 µm Medicons (Becton Dickinson, Heidelberg, Germany) according to the manufacturer's instructions. Isolated cells were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, penicillin (100 U ml -1 ), and streptomycin (100 µg ml -1 ) and cultured in a humidified atmosphere containing 5% CO 2 at 37˚C .

Reverse transcriptase chain reaction (RT-PCR)
RNA isolation, reverse transcription, and PCR reactions were carried out as described . Amplification consisted of 38 cycles with the following conditions for denaturation, annealing and extension: 94˚C for 45 s, 60˚C for 45 s and 72˚C for 2 min. The primers for amplification of cDNA were designed using the PRIMER program (Whitehead Institute for Biomedical Research, Cambridge, MA, USA) based on the cDNA sequences of human 18 kDa PBR subunit obtained from the GenBank (accession number: M36035). Primers for amplification were: Forward: 5′-CACGCTCTACTCAGCCATGG-3′ Reverse: 5′-GCAGTAGTTGAGTGTGGTCGC-3′ The expected PCR product size was 298 bp. PCR products were sequenced on an ABI 310 sequencer (Applied Biosystems, Foster City, CA, USA) and specificity of the transcripts was confirmed by the NCBI (National Center for Biotechnology Information, Bethesda, MD, USA) BLASTIN 2.0 search program (Altschul et al, 1997).

Immunofluorescence labelling and confocal microscopy
Cells were plated on glass cover slips for 24 h. Cells were stained with the mitochondrial dye CMTMRos (100 nM in medium, 30 min, 37˚C) (Molecular Probes, Eugene, OR, USA). After washing with PBS, the samples were fixed by incubation with 4% paraformaldehyde in PBS for 30 min at room temperature, washed with PBS, and permeabilized with cold methanol (20 min, -20˚C) followed by acetone (10 s, -20˚C). Cells were incubated for 1 h at room temperature with primary anti-PBR monoclonal antibody 8D7 (5 µg ml -1 in PBS) (Dussossoy et al, 1996) or isotypic control mouse IgG1 (DAKO, Hamburg, Germany), respectively. Thereafter, cells were washed with PBS twice and incubated with 4 µg ml -1 secondary Alexa TM 488-labelled goat-anti-mouse IgG antibody (Molecular Probes, Eugene, OR, USA) for 1 h at room temperature. Fluorescence and transmission images were obtained using the inverted confocal microscope LSM 410 with a 63x /1,2 W Korr objective (Zeiss, Oberkochen, Germany).
Immunohistochemistry 23 paraffin-embedded colorectal carcinomas were studied. These were 1 well differentiated, 13 moderately differentiated, and 9 poorly differentiated cancers. 2 µm sections of the tumours were deparaffinized and rehydrated in a series of alcohol solutions of decreasing concentrations (Grabowski et al, 2000(Grabowski et al, , 2001. Then sections were transferred into a robotic machine (Chemo-mate, DAKO, Heidelberg, Germany) and staining procedure was automatically performed under following standard conditions. Sections were incubated with the anti-PBR antibody 8D7 (0.5 µg ml -1 ) for 30 min at room temperature. After washing samples were incubated with the second anti-mouse IgG (1:20 dilution) (DAKO, Heidelberg, Germany) for 30 min at room temperature. The APAAP complex (DAKO, Heidelberg, Germany) was incubated for 30 min at a dilution 1:50 in RPMI medium containing 10% fetal calf serum and 1% sodium azide. Staining was detected using fast-red system (DAKO, Heidelberg, Germany) and samples were counterstained with haemalaun.

Cell proliferation assay
The ability of the PBR ligands to modulate cell proliferation was studied using the crystal violet method (Gillies et al, 1986). In brief, cells were seeded on 96-well plates at a density of 1000 cells well -1 (HT29 and Colo320 DM) or 2000 cells well -1 (LS174T). After 72 h the PBR ligands FGIN-1-27, PK 11195 (Tocris, Bristol, UK), or Ro5-4685 (Sigma, Deisenhofen, Germany) were added at concentrations of 1 nM to 100 µM. The FGIN-1-52 (Kozikowski et al, 1993) or the GABA A receptor ligand clonazepam (Sigma, Deisenhofen, Germany) were used as controls. Each concentration group consisted of 10 wells. The incubation medium was changed every day. Cell quantification was performed after 0, 24, 48, 72 and 96 h of incubation. Cells of each well were washed (200 µl PBS) and fixed (100 µl, 1% glutaraldehyde in PBS for 15 min at room temperature). After another washing step (200 µl PBS), cells were stained with 0.1% crystal violet in PBS for 30 min at room temperature. The unbound dye was removed by washing with H 2 O for 30 min. Crystal violet which had absorbed onto the cells was solubilized with 100 µl 0.2% Triton-X-100 in PBS for at least 24 h at 37˚C. The crystal violet containing solution was measured spectrometrically at 570 nm using an ELISA-Reader. In the range of 500 to 20 000 cells well -1 , the measured extinction was a linear function of the cell number .

Cell cycle analysis
Cell cycle analysis was performed by the method of Vindelov and Christensen (Vindelov and Christensen, 1990). 5 × 10 5 cells per well were cultured for 24 h and then exposed to PBR ligands for another 24 h. Cells were trypsinized, washed and the nuclei were isolated using CycleTest PLUS DNA Reagent Kit (Becton Dickinson, Heidelberg, Germany). According to the manufacturer's instructions DNA was stained with propidium iodide. The DNA content of the nuclei was detected by flow cytometry and analysed using CellFit software (Becton Dickinson, Heidelberg, Germany).

Statistical analysis
Comparison of multiple means was performed with nonparametric ANOVA. Comparison of individual drug treatments to control treatments was performed with the unpaired, 2-tailed Mann-Whitney U test for proliferation, ∆Ψ M measurements, and caspase-3 activity experiments. Data are expressed as mean percentage of control ± SEM. For cell cycle analysis the unpaired student's t-test was used. P values were considered to be significant at < 0.05.

PBR expression in colorectal cancer cells
The mRNA expression of PBR was investigated by RT-PCR. Both in HT29 cells and in primary cultures of colorectal carcinomas, PBR mRNA was detected ( Figure 1A). The specificity of the obtained PCR products was confirmed by direct sequencing of cDNAs. HT29 cells did not express any β-chain of the GABA Areceptor (data not shown), which are essential for benzodiazepines binding to the GABA A -receptor. To investigate the expression of the PBR protein, cells were stained with the monoclonal anti-PBR antibody 8D7 and fluorescence was analysed by flow cytometry. The expression of PBR protein was detected in primary cell cultures of colorectal cancer cells ( Figure 1B), as well as in the cell lines HT29, LS174T and Colo320 DM ( Figure 1C-E). In nonpermeabilized cells, no specific 8D7 fluorescence was observed indicating an intracellular localization of PBR ( Figure 1C-E).

PBR localization in mitochondria
To further characterize the subcellular localization of PBR, HT29 cells were simultaneously stained with the mitochondrial dye CMTMRos and the anti-PBR antibody 8D7. Confocal laser scanning microscopy imaged the subcellular distribution of fluorescence. Green fluorescence of 8D7-labelled PBR was detected within the cytoplasm but neither in the cell membrane nor in the nucleus (Figure 2A, E). The staining pattern was comparable to the one obtained by the mitochondrial dye CMTMRos ( Figure 2B, I). Superposition of the images of 8D7-labelled PBR and CMTMRos stained mitochondria resulted in a yellow colour ( Figure 2C) indicating a co-localization of PBR and mitochondria. Likewise, PBR was localized in the mitochondria in the colorectal cell lines LS174T and Colo320 DM as well as in primary cultures of 6 resected colorectal cancers (data not shown). To further verify the mitochondrial localization of PBR, subcellular PBR expression was immunohistochemically determined in 23 other colorectal cancers. The differentiation grade of these tumours ranged from well (n = 1), moderately (n = 13), to poorly (n = 9) differentiated.
In all 23 cancers, the specific PBR staining was observed unevenly distributed within the cytoplasm ( Figure 2K, L). No specific PBR staining was observed in the plasma membrane nor in the nuclei of cells, indicating that in colorectal cancers PBR is located in the mitochondria.

Induction of mitochondrial alterations by specific PBR ligands
PBR is located mainly in the outer mitochondrial membrane and is thought to form the permeability transition pore, which plays an important role in the regulation of ∆ψ M . Therefore, we investigated if specific PBR ligands can modulate the ∆ψ M and mitochondrial volume. FGIN-1-27 significantly depolarized the mitochondria of HT29 cells in a dose-dependent manner as assayed by the  Figure 4A). The results were confirmed by using the mitochondrial dye CMTMRos. CMTMRos stained HT29 cells displayed mean fluorescence of 83 ± 6% (10 µM), 68 ± 4% (50 µM), and 64 ± 6% (100 µM) of control, respectively, when incubated for 6 h with FGIN-1-27. Simultaneously, an increase of mitochondrial volume of HT29 cells and primary cultures was detected ( Figure 4A), indicating that FGIN-1-27 not only induced a mitochondrial membrane depolarization but also an increase in mitochondrial volume. In contrast to FGIN-1-27, PK 11195 or Ro5-4864 showed no effects on HT29 cells after 1 h to 16 h of incubation. Moreover, the non-PBR-specific compound FGIN-1-52 did not alter the ∆ψ M (data not shown). In control experiments the K + ionophore valinomycin was used to reduce the JC-1 590 nm fluorescence (Cossarizza et al, 1993) resulting in a mean 590 nm fluorescence of one third of corresponding control.

DISCUSSION
A tightly regulated balance between cell division and cell death is a prerequisite for normal tissue development. Impaired regulation of either process can lead to tumorigenesis. In this study we show that specific PBR ligands induce both an arrest of the cell cycle and an increase in apoptosis in human colorectal cancer cells. Our data indicate that PBR might be involved in both the regulation of cell division and cell death, suggesting an important role of PBR    (Kozikowski et al, 1993) in colorectal cancer growth. Moreover, we provide evidence that the cell cycle interfering and proapoptotic effects of specific PBR ligands are associated with an inhibition of proliferation of colorectal cancer cells.
The PBR is mainly located in the outer membranes of mitochondria (Gavish et al, 1989). Recently, a perinuclear or nuclear localization has been described in breast cancer and glioma cells, which was shown to be associated with an aggressive tumour type (Hardwick et al, 1999;Brown et al, 2000). Using confocal laser scanning microscopy and immunohistochemistry, we studied PBR expression in human colorectal cancers of distinct grades of differentiation, and in several colorectal cancer cell lines. In all 29 cancers and in the 3 cell lines studied, PBR was found to be localized in the mitochondria. Thus, PBR is primarily targeted to mitochondrial membranes of either well, moderately or poorly differentiated colorectal cancers.
To investigate the functional involvement of PBR in colorectal cancer growth we used specific exogenous ligands. These ligands were shown to interfere with the regulation of both the cell cycle and apoptosis. The specific PBR ligands induced cell cycle arrest. Upon ligand treatment, the proportion of colorectal cancer cells in the G 1 /G 0 -phase increased markedly, indicating that the cells Means as percentage of control ± SEM of 4 independent experiments are shown. In addition, caspase-3 activity was determined of primary tumour cells of 6 patients after incubation with 50 µM FGIN-1-27. Mean ± SEM are shown. *P < 0.05 versus control without agent stayed longer in this phase of the cell cycle. This suggests that PBR ligands act at the classical G 1 checkpoint, preventing cells from entering the S-phase. All 3 PBR ligands studied induced cell cycle arrest at this restriction point, suggesting a common signalling pathway. Cell cycle interfering effects of PBR ligands have been shown previously. In breast cancer PBR ligands induced a cell cycle arrest at both major restriction points, the G 1 /S-and the G 2 /M-junction (Carmel et al, 1999), whereas in lung and melanoma cells an accumulation in the G 2 /M-phase was observed (Camins et al, 1995;Landau et al, 1998). These differences may reflect tissue-specific PBR signal transduction.
In addition to their cell cycle-arresting effects, we here showed for the first time that specific PBR ligands can induce apoptosis in non-haematopoietic cancer cells. The PBR ligands FGIN-1-27, PK 11195 and Ro5-4864 stimulated caspase-3 activity in the 3 cell lines and in all 6 primary cultures of colorectal cancers. Recently, mitochondrial permeability transition has been described as initiating event in the process of apoptosis . Within the mitochondria, PBR is known to be located in the membrane forming the permeability transition pore (Zorov, 1996). This pore is considered to play a central role in the initiation of apoptosis (Zamzami et al, 1995). In this study, the PBR ligand FGIN-1-27 decreased ∆ψ M and caused mitochondrial swelling. Whether FGIN-1-27 induced mitochondrial alterations are a prerequisite for the induction of apoptosis in colorectal cancer cells is not yet understood. It has been reported that mitochondrial permeability transition occurs independently of the initiation of the apoptotic cascade (Lemasters et al, 1998). The actual role of the mitochondrial transition in PBR ligand-induced apoptosis needs to be further investigated.
In this study, we showed that the cell cycle-arresting and proapoptotic effects resulted in an inhibition of cell proliferation.
In line with previous studies in a variety of tumour models (Garnier et al, 1993;Neary et al, 1995;Landau et al, 1998;Carmel et al, 1999;Beinlich et al, 1999), the specific PBR ligands FGIN-1-27, PK 11195 and Ro5-4864 exhibited antiproliferative effects in colorectal cancer cells at micromolar concentrations. To clarify the PBR specificity of the observed effects, we used the benzodiazepine clonazepam or the indoleacetamide FGIN-1-52. Despite their very similar structure to Ro5-4864 or FGIN-1-27, respectively, neither clonazepam nor FGIN-1-52 affected apoptosis, cell cycle or the proliferation of human colorectal cancer cells. This indicates that the effects are specific for PBR. Nevertheless, there is a quantitative discrepancy between the micromolar ligand concentrations necessary to inhibit cell proliferation and the nanomolar-binding affinities. This discrepancy might be due to the existence of a putative 'low-affinity PBR' which has previously been suggested to exist in other tissues (Pawlikowski et al, 1988;Kunert-Radek et al, 1994). This putative low-affinity binding site has been implied in the regulation of calcium channels (Cantor et al, 1984;Taft and DeLorenzo, 1984), suggesting that PBR ligands might affect calcium influx responsible for growth inhibition. However, other factors including cellular absorption, ligand metabolism, as well as differences in the conditions used for the cell proliferation and binding assays (Wang et al, 1984b) have yet to be ruled out.
Even though the exact targets and mechanisms of the proapoptotic and antiproliferative effects of the specific PBR ligands remain to be elucidated, the ability of these ligands to modulate both apoptosis and cell cycle regulation qualify them as promising agents for innovative treatment strategies in colorectal cancer disease.