Influence of rimonabant treatment on peripheral blood mononuclear cells; flow cytometry analysis and gene expression profiling

Introduction

The endocannabinoid system consists of the cannabinoid type 1 (CB1) and cannabinoid type 2 (CB2) receptors, their endogenous ligands anandamide and 2-arachidonoyl glycerol and the enzymes involved in their biosynthesis and metabolism (Di Marzo, Bifulco & De Petrocellis, 2004). CB1 is involved in the regulation of food intake, energy balance and metabolism of glucose and lipids (Di Marzo & Matias, 2005). In clinical studies, CB1 receptor blockage by the selective CB1 antagonist rimonabant (SR141716A) induced weight loss and improvement in serum lipid, glucose and insulin levels by targeting central and peripheral CB1 receptors (Van Gaal et al., 2008). However, some patients experienced depression, and this was considered an unacceptable side effect for treating obesity/metabolic syndrome. Hence, the drug was withdrawn from clinical use but there is remaining interest in some of its many potential medical applications (Cooper & Regnell, 2014; Zhou et al., 2012) including treatment of various malignancies. It is therefore of interest to investigate possible adverse effects on blood cells in patients treated with rimonabant.

The endocannabinoid system is regulating various aspects of lymphocyte proliferation, maturation and immune response (Klein, 2005; Muppidi et al., 2011; Pandey et al., 2009; Pereira et al., 2009; Sido, Nagarkatti & Nagarkatti, 2014). Targeting the endocannabinoid system may therefore be a possible new treatment option in various lymphoproliferative disorders. CB1 receptors are expressed on cells of the immune system, but generally at lower levels than CB2 (Bouaboula et al., 1993; Galiegue et al., 1995). We and others have found that CB1 and CB2 are highly expressed on neoplastic lymphocytes in malignant lymphoma (Gustafsson et al., 2008; Islam et al., 2003; McKallip et al., 2002; Wasik et al., 2014). Targeting of CB1 and CB2 with endogenous or synthetic agonists reduced cell proliferation in vitro and in vivo and induced programmed cell death selectively in tumor cells of mantle cell lymphoma (Flygare et al., 2005; Gustafsson et al., 2006; Gustafsson et al., 2008; Wasik et al., 2011). Similarly, CB2 agonists induced cell death in T cell lymphoblastic leukemia (McKallip et al., 2002). Also the CB1 antagonist rimonabant impaired proliferation and induced cell death in ex vivo isolated mantle cell lymphoma cells, alone, or in combination with anandamide (Flygare et al., 2005). Others have reported antiproliferative effects of rimonabant on in vitro activated PBMC but not on freshly isolated, non-activated PBMC (Gallotta et al., 2010; Malfitano et al., 2008). These results show that CB1 blockade may have immunomodulatory and antiproliferative effects on malignant lymphoma and on activated normal lymphocytes in vitro but seems to spare resting lymphocytes. Very little is published on the effects of rimonabant on human PBMC in vivo, and the aim of this study was to investigate how treatment with rimonabant affected blood leukocytes. We therefore collected blood cells from obese patients treated with rimonabant and analyzed blood leukocytes by flow cytometry before and during treatment. To investigate which genes were differentially expressed in PBMC during rimonabant treatment, we used oligonucleotide arrays to compare gene expression profiles in PBMC before and at 4 weeks of treatment. This pilot study shows that rimonabant treatment induces expression of genes involved in immune responses but have only marginal effects on leukocyte subset frequencies in blood.

Materials and Methods

Results

Analysis of PBMC by flow cytometry before and during treatment with rimonabant

Blood levels of mononuclear cells on eight obese patients were analyzed by flow cytometry before and during treatment with rimonabant. There were no significant changes in the relative frequencies of total CD3+ T cells, CD4+ T cells, CD8+ T cells, B cells or CD3–CD4+ cells (monocytes and dendritic cells) in the patients during the treatment period (Table 1, graphically presented in Fig. 1) There was however a trend towards an increase in percentage of CD3−, CD16+/and or CD56+ cells (before treatment median 9% range 7–14%; after treatment median 12% range 9–15% p = 0.049) (Table 1 and Fig. 1).

Discussion

In this study we investigated the possible effects on PBMC of treatment with the CB1 antagonist rimonabant in patients taking the drug for obesity. We found that the distribution of leukocyte subsets remained rather constant, as analyzed by flow cytometry before treatment and after 4 weeks of treatment with rimonabant with the exception of CD3−, CD16+ and/or CD56+ cells that increased after treatment. This subset includes NK cells (CD3−, CD56+ and/or CD16+) and also subsets of monocytes (CD3–CD16+). There were no significant changes in expression levels of cannabinoid receptors or enzymes involved in synthesis and metabolism of endocannabinoids. However gene expression analysis suggested that genes involved in metabolism, cell death and the innate immune system were up regulated during treatment.

Rimonabant is the first selective CB1 antagonist registered for clinical use and was clinically developed for treatment of obesity and the metabolic syndrome. Beside the effect on food intake, anti-proliferative actions on normal and malignant cells have been reported. Cannabinoid receptors are often more highly expressed on malignant cells than on their normal counterparts and cancer cells are usually more sensitive to the action of cannabinoids than normal cells (reviewed in Flygare & Sander, 2008; Sido, Nagarkatti & Nagarkatti, 2014; Wasik, Christensson & Sander, 2011). Rimonabant has been reported to induce growth inhibition or apoptosis on several malignancies including breast, thyroid and colon cancer (Bifulco et al., 2004; De Petrocellis et al., 1998; Santoro et al., 2009; Sarnataro et al., 2006). We have previously demonstrated that mantle cell lymphoma and other B cell lymphomas have higher expression of CB1 and CB2 than normal lymphocytes (Gustafsson et al., 2008; Islam et al., 2003; Wasik et al., 2014). Cannabinoid receptor agonists, at 1–10 µM levels, reduced proliferation and induced programmed cell death in mantle cell lymphoma in vitro and in a xenotransplant model (Flygare et al., 2005; Gustafsson et al., 2006; Gustafsson et al., 2008; Schatz et al., 1997; Wasik et al., 2011). Interestingly, similar concentrations of rimonabant induced cell death in ex vivo isolated mantle cell lymphoma cells (Flygare et al., 2005). While these studies suggest that targeting of CB1 may be of use in cancer therapy, concern may be raised since anti-proliferative effects have been reported in PBMC (Malfitano et al., 2008). In these studies, rimonabant inhibited mitogen induced cell proliferation in vitro via G1/S phase arrest without induction of cell death (Malfitano et al., 2008). In contrast, Gallotta et al. (2010) reported that freshly isolated PBMC are highly resistant to the cytotoxic and cytostatic effects of rimonabant compared to leukemia-derived cell lines. It is possible that the different sensitivity to CB1 antagonism in freshly isolated, compared to mitogen activated, PBMC may reflect differences in expression levels of CB1. Resting leukocytes express very low levels of CB1 (Bouaboula et al., 1993; Galiegue et al., 1995; Kaminski et al., 1992) while receptor levels may increase upon activation by mitogens, cytokines or exposure to CB1 agonists (Borner et al., 2007; Nong et al., 2002; Schatz et al., 1997). We did not detect any significant differences in expression levels of CB1 or other components of the endocannabinoid system during rimonabant treatment for 4 weeks. Furthermore, our studies on ex vivo isolated PBMC from rimonabant treated patients demonstrated very minor changes in frequencies of T cells, B cells, CD3−, CD16+ and/or CD56+ cells or CD3–CD4+ cells or on total lymphocyte counts, in line with the results of Gallotta et al. (2010).

Global gene expression analysis demonstrated significant changes in genes coding for components of the innate immune system. The study design does not make it possible to discriminate if the differences in gene expression can be ascribed to certain subsets of leukocytes or if it is a general process, seen in all PBMC. However, many of the genes that were more highly expressed after treatment with rimonabant are expressed in NK cells (such as KLRF1 and CD160) and monocytes, which imply that the treatment is associated with the activation of certain inflammatory and immunological functions of the innate immune system. Interestingly, rimonabant has been shown to directly activate human and mouse macrophages and thereby inhibit the development of the intracellular pathogen Brucella suis (Gross et al., 2000). Furthermore, studies on lipopolysaccharide activated human macrophages showed that CB1 receptor blockade by rimonabant suppressed production of inflammatory cytokines (IL-1β, IL-6, IL-8, TNF-α) and matrix metalloproteinase-9 (Sugamura et al., 2009).

Conclusions

In conclusion our results show that rimonabant treatment induces expression of genes involved in immune responses but have only marginal effects on leukocyte subset frequencies in blood. This is in marked contrast to previous studies in which rimonabant induced cell death in malignant B lymphocytes that express high levels of CB1 (Flygare et al., 2005). The relatively small effects on normal leukocytes suggest that CB1 targeting may be further investigated as a therapeutic approach in lymphoma treatment, enabling selective effects of tumor cells.

Supplemental Information