Acid sphingomyelinase deficiency protects from cisplatin ... - Nature

7 downloads 84 Views 447KB Size Report
Aug 4, 2008 - Cisplatin is one of the most effectively used chemother- apeutic agents for cancer treatment. However, in humans, important cytotoxic side ...
Oncogene (2008) 27, 6590–6595

& 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $32.00 www.nature.com/onc

SHORT COMMUNICATION

Acid sphingomyelinase deficiency protects from cisplatin-induced gastrointestinal damage A Rebillard1,2, N Rioux-Leclercq3,4, C Muller1, P Bellaud3, F Jouan4, O Meurette1,2, E Jouan1,2, L Vernhet1,2, C Le Que´ment1,2, A Carpinteiro5, M Schenck5, D Lagadic-Gossmann1,2, E Gulbins5 and MT Dimanche-Boitrel1,2 1

Inserm, UMR620, IFR 140 GFAS, Rennes, France; 2UPRES EA SeRAIC, University of Rennes, Rennes, France; Department of Pathology, Hospital, Rennes University, Rennes, France; 4CNRS/UMR6061, Rennes University, Rennes, France and 5Department of Molecular Biology, University of Duisburg-Essen, Essen, Germany 3

Cisplatin is one of the most effectively used chemotherapeutic agents for cancer treatment. However, in humans, important cytotoxic side effects are observed including dose-limiting renal damage and profound gastrointestinal symptomatology. The toxic responses to cisplatin in mice are similar to those in human patients. Here, we evaluated whether the acid sphingomyelinase (Asm) mediates at least some of the toxic in vivo effects of cisplatin. To this end, we determined the toxic effects of a single intraperitoneal dose of cisplatin (27 mg/kg) in wild type (Asm þ / þ ) and Asm-deficient mice (Asm/). Tissue injury and apoptosis were determined histologically on hematoxylin–eosin and TUNEL (terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling) stainings 3, 12, 36 and 72 h after treatment. Our results revealed severe toxicity of cisplatin in Asm þ / þ mice with increased numbers of apoptotic cells in the thymus and small intestine. In marked contrast, Asm/ mice were resistant to cisplatin and no apoptosis was observed in these organs after treatment. Moreover, cisplatin treatment primarily triggered apoptosis of endothelial cells in microvessels of intestine and thymus, an effect that was absent in mice lacking Asm. The data thus suggest that at least some toxic effects of cisplatin are mediated by the Asm in vivo resulting in early death of endothelial cells and consecutive organ damage. Oncogene (2008) 27, 6590–6595; doi:10.1038/onc.2008.257; published online 4 August 2008 Keywords: acid sphingomyelinase; cisplatin; small intestine; endothelial cells; p53 Cisplatin (cis-diamminedichloroplatinum or CDDP) is an anticancer agent widely used in the treatment of solid tumors (Loehrer and Einhorn, 1984), although its use is limited by severe toxic side effects. Indeed, cisplatin

Correspondence: Dr MT Dimanche-Boitrel, INSERM UMR620, University of Rennes 1, Faculty of Pharmacy, 2 Av du Pr Le´on Bernard, Rennes 35043, France. E-mail: [email protected] Received 29 May 2007; revised 10 June 2008; accepted 1 July 2008; published online 4 August 2008

exerts toxic effects on normal tissues, including renal tubular necrosis and dose-dependent damage of the gastrointestinal epithelium in mice. These damages are very similar to those observed in humans, thus providing evidence of the usefulness of the mouse model to predict toxicological responses to anticancer drugs (Harrison, 1981). Definition of the molecular mechanisms of cisplatin-induced toxicity is important for the development of adjunctive therapies to prevent cisplatin toxicity and to understand mechanisms of cisplatin resistance. Although DNA is considered as the primary target of cisplatin, this anticancer agent has also been shown to interact with the plasma membrane, thereby reducing the activity of certain ion channels (Grunicke and Hofmann, 1992). Recently, we have demonstrated that cisplatin induces apoptosis through the activation of acid sphingomyelinase (Asm) (Lacour et al., 2004; Rebillard et al., 2007). This enzyme, which converts sphingomyelin to ceramide, has also been shown to be activated by many proapoptotic receptors (CD95/Fas and TNF-R), stress stimuli (irradiation and ultraviolet light) and antiproliferative agents (rituximab) (Santana et al., 1996; Huang et al., 1997; Gulbins and Grassme, 2002; Gulbins and Kolesnick, 2003; Bezombes et al., 2004). At present, the role of Asm in the toxicity of anticancer drugs in vivo requires definition. It has been shown that ex vivo Asm-deficient mice (Asm/) oocytes were resistant to doxorubicin (Morita et al., 2000). Furthermore, Asm/ mice were protected from radiation-induced damage in the lung (Santana et al., 1996), small intestine (Paris et al., 2001) and brain (Pena et al., 2000). These results led us to study the role of Asm in cisplatin toxicity in vivo by using Asm/ mice in comparison with syngenic C57BL/6 wild-type mice (Horinouchi et al., 1995). Earlier publications revealed an LD50 of cisplatin (18 mg/kg) in mice (Prestayko et al., 1979). At a dose of 14 mg/kg, cisplatin induced reversible gastrointestinal alterations (Harrison, 1981). This fact led us to choose a higher dose of cisplatin (27 mg/kg), which caused severe toxicity with acute renal failure and death of the mice on days 4–5 after cisplatin injection. We observed a tissue-specific activity of cisplatin toxicity in wild-type (Asm þ / þ ) mice. The most dramatic

Asm-deficient mice and cisplatin toxicity A Rebillard et al

6591

histological changes were observed in the small intestine (ileum) and thymus of these mice. Villi in ileum were shortened with a reduction in the number of goblet cells, and an increase of apoptotic bodies following 72 h of cisplatin treatment (Figure 1a), whereas intestinal villi of mice treated with vehicle remained intact. Similar cisplatin toxic effects were previously observed in the small intestine of B6D2F1 (Harrison, 1981), CBA (Allan and Smyth, 1986) and BALB/c mice (Tamaki et al., 2003). We also detected an increase in the number of apoptotic bodies with a reduction of lymphocytic and thymic cells in the thymus (Figure 1b). No histological changes were observed in the liver, lung and bone marrow tissues of Asm þ / þ mice treated with a single intraperitoneal dose of cisplatin (27 mg/kg) for 72 h (data not shown). Next, we studied the involvement of Asm in cisplatininduced toxicity in the small intestine and thymus by comparing the toxic effect of a single intraperitoneal injection of cisplatin (27 mg/kg) in Asm þ / þ and Asm/ mice. To characterize apoptosis in tissues, a TUNEL (terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling) assay was performed in the small intestine sections. Gastrointestinal toxicity of CDDP was blunted in Asm/ mice and intestinal villi in these mice displayed an almost normal histology (Figure 2a, upper panel) with very few apoptotic cells 72 h after CDDP treatment (Figure 2a, lower panel). This contrasted with a near-total villi loss (Figure 2a, upper panel) with massive apoptosis of surface and crypt epithelial cells in Asm þ / þ mice treated for 72 h with CDDP (Figure 2a, lower panel). Similarly, whereas numerous apoptotic cells and apoptotic bodies were histologically observed in the thymus of CDDP-treated Asm þ / þ mice, no significant increase of apoptotic cells was detected in the thymus of CDDP-treated Asm/ mice (Figure 2b, upper panel). TUNEL assay showed many apoptotic nuclei in the thymus of Asm þ / þ mice treated with CDDP for 72 h, an effect that was absent in the thymus of similarly treated Asm/ mice (Figure 2b, lower panel). To quantify the percentage of cell death, apoptotic cells (TUNEL-positive cells in brown color) were scored on thymic sections by counting 25 fields. Although almost no apoptosis was detected in the thymus of vehicle-treated Asm þ / þ mice, CDDP treatment for 72 h resulted in severe apoptosis in the thymus with 20% of fields with 10–20 apoptotic cells, 33% of fields with 20–50 apoptotic cells and 18% of fields with more than 50 apoptotic cells (Figure 2c). Acid sphingomyelinase deficiency protected from CDDP toxicity following 72 h of treatment and led to a 70% decrease in the percentage of fields with 10–20 apoptotic cells (Po0.05) and an 80% reduction in the percentage of the fields with 20–50 apoptotic cells (Po0.05) (Figure 2c). In summary, our data show for the first time that cisplatin treatment caused injuries in the small intestine and thymus that are absent in Asm/ mice. These effects are not because of a decrease in platin level in serum of Asm/ mice (Supplementary Data). These findings are reminiscent of the resistance of the

Asm+/+ vehicle 72 h

Asm+/+ vehicle 72 h

Asm+/+ CDDP 72 h

Asm+/+ CDDP 72 h

Figure 1 Cisplatin-induced toxicity in the small intestine (a) and thymus (b). Freshly prepared cisplatin (1 mg/ml in normal saline; CDDP, Merck, Lyon, France) was administered intraperitoneally at a single dose of 27 mg/kg. Control mice received saline vehicle. To collect tissues, mice were killed by cervical dislocation, the organs removed, fixed in 4% paraformaldehyde and embedded in paraffin. Transverse tissue sections were obtained, attached to polylysine-treated slides, deparaffinized by heating at 90 1C for 10 min and at 60 1C for 5 min, followed by two xylene washes for 5 min, and staining with hematoxylin–eosin according to a standard protocol. One representative of hematoxylin–eosin staining of tissue sections is shown (  400 magnification).

intestinal tract in Asm/ mice to irradiation (Paris et al., 2001). To achieve a detailed characterization of apoptosis in the small intestine, we performed time course experiments following cisplatin treatment. TUNEL staining revealed many apoptotic cells mainly in the lamina propria of villi as early as 3 h after CDDP treatment of Asm þ / þ mice (Figure 3a, upper panel). Extensive apoptosis in epithelial cells of intestinal crypts and villi, respectively, was observed 12 or 36 h after CDDP treatment of Asm þ / þ mice (Figure 3a, upper panel). In contrast, intestinal villi were preserved in Asm/ mice treated with cisplatin for 3, 12 or 36 h (Figure 3a, lower panel). Injection of the vehicle was without effect in either Asm þ / þ or Asm/ mice (Figure 3a). The positive TUNEL staining in cells of the lamina propria 3 h after CDDP treatment suggested that endothelial cells might be involved in the toxicity of cisplatin, similar to the observation previously published Oncogene

Asm-deficient mice and cisplatin toxicity A Rebillard et al

6592

Asm+/+ CDDP 72 h

Asm-/CDDP 72 h

>50 50< >20 20< >10 10 6 to 9 (+/+) (+/+) (-/-) (+/+) (-/-) (+/+) (-/-) Vehicle CDDP CDDP CDDP CDDP CDDP CDDP 12h 3h 3h 12h 12h 36h 36h

Asm+/+ CDDP 36 h Percent crypt/villus units with apoptotic epithelial cells

Asm+/+ vehicle 36 h

100 90 80 70 60 50 40 30 20 10 0

3 to 5 1 to 2 0

100 90 80 70 60 50 40 30 20 10 0 (+/+) (+/+) (-/-) (+/+) (-/-) (+/+) (-/-) Vehicle CDDP CDDP CDDP CDDP CDDP CDDP 12h 3h 3h 12h 12h 36h 36h

-/-

Asm CDDP 36 h

Asm+/+ Vehicle 72 h

Asm+/+vehicle 3h

Asm+/+CDDP 3 h

Asm+/+ CDDP 3 h

Asm+/+ CDDP 12 h

Asm+/+ CDDP 36 h

Asm+/+ CDDP 72 h

Figure 3 Asm expression is required for cisplatin-induced early microvascular endothelial apoptosis. (a) Apoptosis in the lamina propria precedes apoptosis in the epithelium. Small intestine (ileum) was collected 3, 12 or 36 h after cisplatin treatment of each nine Asm þ / þ and Asm/ mice and stained by terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling. Representative terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling stainings are shown (  400 magnification). (b) Double stainings for factor VIII Von Willebrand staining (red) (  400 magnification) and by terminal deoxynucleotidyl transferase (TdT)mediated nick end labeling assay (green) identified apoptotic cells in the lamina propria as endothelial cells. After three washes in phosphate-buffered saline, tissue sections were blocked in horse serum solution for 20 min and stained with a rabbit anti-factor VIII Von Willebrand antibody (1:300 diluted in horse serum, Dako Cytomation SA, Dako France SAS, Trappes, France) for 30 min. The sections were then incubated for 30 min with a TRITC-coupled anti-rabbit IgG (1:500, Sigma-Aldrich, St Quentin Fallavier, France) and washed three times in phosphate-buffered saline (endothelial cells in red). The signal was visualized using a fluorescent microscope (DMRXA Leica microscope, COHU high-performance CCD camera using Metavue software). (c) Frequency histograms of apoptotic endothelial and epithelial cells in crypt/villus units following 3, 12 or 36 h after treatment with cis-diamminedichloroplatinum are shown. Apoptotic endothelial and epithelial cells were scored in 200 crypt/villus units on the small intestine sections after terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling assay. Data represent mean scores from three experiments. Statistical analyses were performed as described in Figure 3c (Supplementary Data). (d) Apoptotic endothelial cells and apoptotic thymocytes, respectively, identified in thymic sections by a co-staining with factor VIII Von Willebrand (in red) and terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling assay are shown (in green) (  400 magnification).

CDDP treatment, indicating that apoptosis of endothelial cells occurred before apoptosis in the epithelium of crypts and villi (Figure 3a, upper panel). We quantified apoptosis by counting the percentage of crypts/villi units with apoptotic endothelial or epithelial cells 3, 12 and 36 h after CDDP treatment in Asm þ / þ or Asm/ mice. In vehicle-treated Asm þ / þ mice, only a few apoptotic endothelial or epithelial cells were detected in crypts/villi units (Figure 3c). Asm þ / þ mice treated for 3 h with

CDDP exhibited massive endothelial apoptosis with 54.5% of crypts/villi units showing six to nine apoptotic cells (CDDP 3 h versus vehicle, Po0.001) and 27.5% of crypts/villi units showing more than 10 apoptotic endothelial cells (CDDP 3 h versus vehicle, Po0.01). The percentage of crypts/villi units with apoptotic endothelial cells decreased 12 h after CDDP treatment of Asm þ / þ mice (a 75% decrease in the percentage of units with more than 10 apoptotic cells, CDDP 12 h Oncogene

Asm-deficient mice and cisplatin toxicity A Rebillard et al

6594

p53+/+ CDDP 36 h

p53+/+ CDDP 3 h

p53-/CDDP 36 h

p53-/CDDP 3 h

p53+/+ CDDP 36 h

p53-/CDDP 36 h

p53+/+ CDDP 3 h

p53-/CDDP 3 h

p53+/+ CDDP 36 h

p53-/CDDP 36 h

Figure 4 Cisplatin-induced damages are dependent on p53 protein expression (a) Small intestine was collected 36 h after cisplatin treatment of each six p53 þ / þ and p53/ mice and stained by terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling assay. Representative terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling stainings are shown (  400 magnification). (b) Small intestine was collected 3 h after cisplatin treatment of each five p53 þ / þ and p53/ mice and co-stained with factor VIII Von Willebrand (in red) and terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling assay (in green) (  400 magnification). (c) Thymus was collected 36 h after cisplatin treatment of each six p53 þ / þ and p53/ mice and stained by terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling. Representative terminal deoxynucleotidyl transferase (TdT)mediated nick end labeling stainings are shown (  400 magnification). (d) Thymus was collected 3 and 36 h after cisplatin treatment of each p53 þ / þ and p53/ mice and co-stained with factor VIII Von Willebrand (in red) and terminal deoxynucleotidyl transferase (TdT)mediated nick end labeling assay (in green) (  400 magnification).

versus CDDP 3 h, Po0.001), although the percentage of crypts/villi units with apoptotic epithelial cells increased (a 60% increase in the percentage of units with 3–5 apoptotic cells, CDDP 12 h versus 3 h, Po0.001) Oncogene

(Figure 3c). In contrast, both endothelial and epithelial cells of the small intestine were almost completely protected in Asm/ mice from CDDP-induced cytotoxicity 36 h after CDDP treatment (Figure 3c).

Asm-deficient mice and cisplatin toxicity A Rebillard et al

6595

These data suggest that cisplatin induces small intestine toxicity through apoptosis of endothelial cells before columnar epithelial cells. Thus, the alteration of the microcirculation in the crypt/villus units may permit CDDP-mediated apoptosis in epithelial cells and/or may directly trigger death of these cells. Similar results were observed in the thymus of Asm þ / þ mice with massive apoptosis of endothelial cells 3 or 12 h after CDDP treatment (Figure 3d, indicated by the co-staining in yellow), followed by massive apoptosis of thymocytes 36 or 72 h after CDDP treatment (Figure 3d, fragmented nuclei stained in green). Cisplatin cytotoxicity has been previously shown to be p53-dependent (Vasey et al., 1996). To study the involvement of p53 protein in cisplatin toxicity in the small intestine and thymus, we treated p53 þ / þ and p53/ mice with a 27 mg/kg dose of cisplatin. Interestingly, co-stainings with factor VIII Von Willebrand and TUNEL revealed apoptosis of endothelial cells in the small intestine (Figure 4b) and thymus (Figure 4d) of p53 þ / þ mice treated with cisplatin for 3 h, which was absent in p53/ mice. After 36 h of cisplatin treatment, no lesion in the small intestine and thymus of p53/ mice was observed (Figures 4a and c). Recently, we have shown that cisplatin treatment rapidly activates Asm independently of DNA damage in the course of cisplatin-induced apoptosis in p53-mutated HT29 human colon cancer cells (Rebillard et al., 2007).

Our data demonstrate for the first time that another cell death pathway dependent on Asm could account for cisplatin toxicity in vivo and suggest that mechanisms leading to cisplatin resistance in tumor cells may also arise through alterations in Asm-dependent apoptotic pathway. Moreover, p53 signaling is also involved in cisplatin endothelial damage in the small intestine and thymus. Our results are in agreement with those of Sansom and Clarke (2002) showing that p53 deficiency protected the intestinal stem cell compartment against cisplatin-mediated lethality. They also suggest that Asm deficiency may result in the same protection against cisplatin toxicity. In summary, our data suggest for the first time that Asm/p53 dually regulate cisplatin-induced damage in the small intestine and thymus. It is intriguing to speculate whether cisplatin sequentially activates p53 and Asm to trigger death or whether both pathways are independently required to mediate death by cisplatin. However, a detailed analysis of the interactions of the two pathways is beyond the focus of this paper and requires further study. Acknowledgements We thank F Paris, M Le Ve´e and I Gue´non. This study was supported by grants from the Ligue Nationale Contre le Cancer (the Coˆte d’Armor, Ille et Vilaine and Loire-Atlantique Committees), Rennes Me´tropole, the Re´gion Bretagne and DFG-10-2238-Gu2.

References Allan SG, Smyth JF. (1986). Small intestinal mucosal toxicity of cis-platinum—comparison of toxicity with platinum analogues and dexamethasone. Br J Cancer 53: 355–360. Bezombes C, Grazide S, Garret C, Fabre C, Quillet-Mary A, Muller S et al. (2004). Rituximab antiproliferative effect in B-lymphoma cells is associated with acid-sphingomyelinase activation in raft microdomains. Blood 104: 1166–1173. Grunicke H, Hofmann J. (1992). Cytotoxic and cytostatic effects of antitumor agents induced at the plasma membrane level. Pharmacol Ther 55: 1–30. Gulbins E, Grassme H. (2002). Ceramide and cell death receptor clustering. Biochim Biophys Acta 1585: 139–145. Gulbins E, Kolesnick R. (2003). Raft ceramide in molecular medicine. Oncogene 22: 7070–7077. Harrison Jr SD. (1981). Toxicologic evaluation of cis-diamminedichloroplatinum II in B6D2F1 mice. Fundam Appl Toxicol 1: 382–385. Horinouchi K, Erlich S, Perl DP, Ferlinz K, Bisgaier CL, Sandhoff K et al. (1995). Acid sphingomyelinase deficient mice: a model of types A and B Niemann–Pick disease. Nat Genet 10: 288–293. Huang C, Ma W, Ding M, Bowden GT, Dong Z. (1997). Direct evidence for an important role of sphingomyelinase in ultravioletinduced activation of c-Jun N-terminal kinase. J Biol Chem 272: 27753–27757. Lacour S, Hammann A, Grazide S, Lagadic-Gossmann D, Athias A, Sergent O et al. (2004). Cisplatin-induced CD95 redistribution into membrane lipid rafts of HT29 human colon cancer cells. Cancer Res 64: 3593–3598. Loehrer PJ, Einhorn LH. (1984). Drugs five years later. Cisplatin. Ann Intern Med 100: 704–713. Morita Y, Perez GI, Paris F, Miranda SR, Ehleiter D, HaimovitzFriedman A et al. (2000). Oocyte apoptosis is suppressed by

disruption of the acid sphingomyelinase gene or by sphingosine1-phosphate therapy. Nat Med 6: 1109–1114. Paris F, Fuks Z, Kang A, Capodieci P, Juan G, Ehleiter D et al. (2001). Endothelial apoptosis as the primary lesion initiating intestinal radiation damage in mice. Science 293: 293–297. Pena LA, Fuks Z, Kolesnick RN. (2000). Radiation-induced apoptosis of endothelial cells in the murine central nervous system: protection by fibroblast growth factor and sphingomyelinase deficiency. Cancer Res 60: 321–327. Prestayko AW, Bradner WT, Huftalen JB, Rose WC, Schurig JE, Cleare MJ et al. (1979). Antileukemia (L1210) activity and toxicity of cis-dichlorodiammineplatinum(II) analogs. Cancer Treat Rep 63: 1503–1508. Rebillard A, Tekpli X, Meurette O, Sergent O, Le Moigne-Muller G, Vernhet L et al. (2007). Cisplatin-induced apoptosis involves membrane fluidification via inhibition of NHE1 in human colon cancer cells. Cancer Res 67: 7865–7874. Sansom OJ, Clarke AR. (2002). The ability to engage enterocyte apoptosis does not predict long-term crypt survival in p53 and Msh2 deficient mice. Oncogene 21: 5934–5939. Santana P, Pena LA, Haimovitz-Friedman A, Martin S, Green D, McLoughlin M et al. (1996). Acid sphingomyelinase-deficient human lymphoblasts and mice are defective in radiation-induced apoptosis. Cell 86: 189–199. Tamaki T, Naomoto Y, Kimura S, Kawashima R, Shirakawa Y, Shigemitsu K et al. (2003). Apoptosis in normal tissues induced by anti-cancer drugs. J Int Med Res 31: 6–16. Vasey PA, Jones NA, Jenkins S, Dive C, Brown R. (1996). Cisplatin, camptothecin, and taxol sensitivities of cells with p53-associated multidrug resistance. Mol Pharmacol 50: 1536–1540.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc) Oncogene