Accepted Manuscript

Targeting colorectal cancer cells by a novel sphingosine kinase 1 inhibitor PF-543

TongFa Ju, DaQuan Gao, Zheng-yu Fang, M.D. Ph.D

PII: S0006-291X(16)30053-5
DOI: 10.1016/j.bbrc.2016.01.053
Reference: YBBRC 35170

To appear in: Biochemical and Biophysical Research Communications

Received Date: 4 January 2016

Accepted Date: 9 January 2016

Please cite this article as: T. Ju, D. Gao, Z.-y. Fang, Targeting colorectal cancer cells by a novel
sphingosine kinase 1 inhibitor PF-543, Biochemical and Biophysical Research Communications (2016),
doi: 10.1016/j.bbrc.2016.01.053.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
1 Targeting colorectal cancer cells by a novel sphingosine kinase 1
2 inhibitor PF-543
3
4 TongFa Ju 1, DaQuan Gao 1 and Zheng-yu Fang 2*
5
1
6 Department of Anal-colorectal Surgery, HangZhou First Peoples Hospital, HangZhou, China
2

PT
7 Department of Anal-colorectal Surgery, the First Affiliated Hospital of Zhejiang Chinese Medical
8 University, Hangzhou, China
9

RI
10 * Corresponding author
11 Dr. Fang Zheng-yu M.D. Ph.D

SC
12 54 YouDian Road, HangZhou 310006, China
13 Tel: +8613858112772
14 Email: [email protected]

U
15
AN
16 Abstract. In this study, we showed that PF-543, a novel sphingosine kinase 1 (SphK1) inhibitor, exerted
17 potent anti-proliferative and cytotoxic effects against a panel of established (HCT-116, HT-29 and DLD-1)
18 and primary human colorectal cancer (CRC) cells. Its sensitivity was negatively associated with SphK1
M

19 expression level in the CRC cells. Surprisingly, PF-543 mainly induced programmed necrosis, but not
20 apoptosis, in the CRC cells. CRC cell necrotic death was detected by lactate dehydrogenase (LDH)
D

21 release, mitochondrial membrane potential (MMP) collapse and mitochondrial P53-cyclophilin-D (Cyp-D)
22 complexation. Correspondingly, the necrosis inhibitor necrostatin-1 largely attenuated PF-543-induced
TE

23 cytotoxicity against CRC cells. Meanwhile, the Cyp-D inhibitors (sanglifehrin A and cyclosporin A), or
24 shRNA-mediated knockdown of Cyp-D, remarkably alleviated PF-543-induced CRC cell necrotic death.
25 Reversely, over-expression of wild-type Cyp-D in HCT-116 cells significantly increased PF-543s
EP

26 sensitivity. In vivo, PF-543 intravenous injection significantly suppressed HCT-116 xenograft growth in
27 severe combined immunodeficient (SCID) mice, whiling remarkably improving the mice survival. The in
C

28 vivo activity by PF-543 was largely attenuated when combined with the Cyp-D inhibitor cyclosporin A.
29 Collectively, our results demonstrate that PF-543 exerts potent anti-CRC activity in vitro and in vivo.
AC

30 Mitochondrial programmed necrosis pathway is likely the key mechanism responsible for PF-543s
31 actions in CRC cells.
32

33

1
 ACCEPTED MANUSCRIPT
34 1. Introduction
35
36 The colorectal cancer (CRC) is the second most common cancer among humans, causing significant
37 cancer-related mortalities worldwide [1,2,3]. Surgical resection of the primary tumor is still the curative
38 treatment option of the affected patients [1,2,3]. Yet, any remaining tumors as well as metastasis and
39 recurrence tumors are often resistant to conventional chemotherapies [1,2,3]. Thus, it is necessary and
40 vital to develop alternative treatment options for CRC [1,2,3].

PT
41
42 Existing evidences have shown that sphingosine kinase 1 (SphK1) is over-expressed and/or
43 over-activated in CRC and other human cancers, which is correlated with patients poor prognosis [4,5].

RI
44 SphK1 catalyses the phosphorylation of sphingosine to sphingosine-1-phosphate (S1P), thus decreasing
45 the level of pro-death ceramide/sphingosine, while increasing the level of pro-survival S1P [4,5].

SC
46 Preclinical cancer studies, using strategies of overexpression, mutation, small interfering RNA
47 (siRNA)/microRNA knockdown, have indicated that SphK1 promotes cancer cell survival, cell proliferation
and resistance to apoptosis and necrosis [4,5]. Consequently, studies have investigated the potential

U
48
49 anti-cancer activity of several SphK1 inhibitors [4,5]. Some of these inhibitors have demonstrated
AN
50 promising results against CRC and other cancer cells [4,5].
51
52 Recent studies have characterized a novel and extremely potent SphK1 inhibitor, namely PF-543 [6].
M

53 It showed 100-fold selectivity for SphK1 over SphK2 [6]. In the current study, we investigated the effects
54 of PF-543 against CRC cells both in vitro and in vivo.
D

55
56 2. Materials and methods
TE

57
58 2.1. Chemicals, reagents and antibodies. PF-543 was provided by Selleck (Shanghai, China). The
EP

59 cyclophilin-D (Cyp-D) inhibitors sanglifehrin A (SfA) [7] and cyclosporin A (CsA) [7,8] were purchased from
60 Sigma-Aldrich (Shanghai, China). The caspase inhibitor z-VAD-fmk and the histone deacetylase inhibitor
61 (HDACi) suberoylanilide hydroxamic acid (SAHA) were also from Sigma. The necrosis inhibitor
C

62 necrostatin-1 (Nec-1) was obtained from Cayman Chemical (Beijing, China). All antibodies utilized in this
AC

63 study were from Santa Cruz Biotech (Santa Cruz, CA). The cell culture reagents were from Invitrogen
64 (Shanghai, China).
65
66 2.2. Culture of CRC cell lines. CRC cell lines, including HCT-116, HT-29 and DLD-1, were provided
67 by American Type Cell Collection Company (Manassas, VA). Cells were maintained in RPMI-1640/DMEM
68 medium plus 6-10% FBS, and necessary antibiotics.
69
70 2.3. Culture of primary colon cancer cells. Primary colon cancer cells were derived from three
71 informed male patients (46/62/58 years old). Following surgical resection, colon cancer tissues were

2
 ACCEPTED MANUSCRIPT
72 minced into 1 mm3 in size, and washed. The tissue fragments were then placed in DMEM medium with
73 1 mg/mL Collagenase IV (Sigma) for 1.5 hrs at 37C for digestion. Every 15 min, the solution was
74 vigorously shaken for 15 seconds to force dissociation. Cells were then sieved through a 40 m filter, and
75 resuspended in primary culture medium. The protocol was approved by the Ethics Review Board of
76 authors institutions, and was in agreement with the collection of clinical samples (Declaration of Helsinki).
77
78 2.4. MTT cell survival assay. Cell survival was evaluated by via

PT
79 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium (MTT) assay. Cells (0.5 104 cells/well) were seeded
80 onto 96-well plates. Following treatment, twenty L per well of MTT (5 mg/mL, Sigma) solution was added,
81 the colorimetric assay was measured at the absorbance of 570 nm via a Tecan Infinite 200 Pro microplate

RI
82 reader.
83

SC
84 2.5. LDH cell necrosis assay. Lactate dehydrogenase (LDH) release in the media is an early marker
85 of cell necrosis, which was carried out via Pierce LDH Cytotoxicity Assay Kit (Thermo Scientific,
Pittsburgh, PA) in this study. % LDH release = LDH released in conditional medium/(LDH released in

U
86
87 conditional medium + LDH in cell lysates).
AN
88
89 2.6. Colony formation assay. CRC cells (200 cells/well) were seeded onto a top layer of 0.3%
90 low-melting agarose (Sigma) in 24-well plates with a bottom layer of 1.0% agarose in complete culture
M

91 medium. After 12 days incubation, colonies containing > 20 cells were visualized and counted.
92
D

93 2.6. ssDNA ELISA assay of apoptosis. DNA denature is one characteristic sign of cell apoptosis.
94 Following treatment, denatured single strand DNA (ssDNA) was detected through a nucleosomal
TE

95 monoclonal antibody in an ELISA format. Cells (0.5104/well) were seeded onto 96-well plates. Cell
96 apoptosis was analyzed by the ssDNA ELISA kit from Chemicon International (Temecula, CA).
EP

97
98 2.7. Annexin-V apoptosis assay. After treatment, cells were washed and re-suspended in binding
99 buffer (BD Biosciences, San Jose, CA) and stained with Annexin-V-FITC (BD, 5 L/sample), followed by
C

100 BD FACSCanto II flow cytometer (BD Biosciences) detection of Annexin-V positive cells.
AC

101
102 2.8. Assay of caspase-3/-9 activity. CRC cells were seeded onto 96-well plats. Following treatment,
103 caspase-Glo reagent (100 L/well) was added. Caspase-3/-9 activity was determined via the
104 caspase-Glo 3/9 kits (Promega, Shanghai, China).
105
106 2.9. The mitochondrial membrane potential (MMP) detection. As previously described [9,10], the
107 MMP was detected via JC-10 dye (Invitrogen). When the MMP is decreasing, monomeric JC-10 will form
108 in the cytosol and exhibit a green fluorescence. Briefly, cells were stained with 5 g/mL of JC-10 for 10
109 min under the dark. Afterwards, cells were washed and detected immediately on a microplate reader with

3
 ACCEPTED MANUSCRIPT
110 an excitation filter of 485 nm [9].
111
112 2.10. Western blots. Cells were harvested and incubated in lysis buffer described [11]. Twenty g
113 extracts per sample were separated by 10% SDS gel and transferred onto PVDF membranes (Millipore,
114 Bedford, MA), which were then probed with indicated primary antibodies and corresponding second
115 antibodies [11]. Blots were visualized using an enhanced chemiluminescence (ECL) detection system.
116 The intensity of each band was quantified via ImageJ software. For detecting mitochondrial proteins,

PT
117 intact mitochondria were separated through Mitochondria Isolation Kit for Cultured Cells (Pierce,
118 Rockford, IL).
119

RI
120 2.11. Mitochondrial immunoprecipitation (mito-IP). As reported [11], 600 g mitochondrial protein
121 lysates were pre-cleared with 20 L of protein A/G PLUS-agarose (Santa Cruz) for 1 hr. The supernatant

SC
122 was then incubated with anti-Cyp-D antibody (Santa Cruz) overnight at 4 C. The protein A/G
123 PLUS-agarose (35 L) was then added to the supernatants for 4 hrs. Pellets were washed six times with
PBS, resuspended in lysis buffer, and then assayed by Western blots.

U
124
125
AN
126 2.12. Cyp-D shRNA stable knockdown. The two sets of lentiviral particles with human Cyp-D shRNAs
127 were purchased separately from Santa Cruz Biotech (sc-44892-V, Cyp-D shRNA sequence-1 [12]) and
128 Genechem (Shanghai, China, Cyp-D shRNA sequence-2, designed in [13] ). Ten L/mL of lentiviral
M

129 particles were added to HCT-116 cells for 12 hrs. Afterwards, fresh growth medium was added, and cells
130 were further cultured for another 24 hrs. Puromycin (5.0 g/mL) was added to select resistant stable
D

131 HCT-116 cells for 2-3 weeks. Expression of Cyp-D in stable HCT-116 cells was detected by Western
132 blots.
TE

133
134 2.14. Cyp-D expressing construct and transfection. The full-length Cyp-D cDNA (provided by Dr.
EP

135 Jiangs lab [11]) was sub-cloned into pSuper-puro-Flag vector (YRgene, Shanghai, China). The empty
136 vector (pSuper-puro-Flag) or Cyp-D expressing construct was transfected into HCT-116 cells via
137 Lipofectamine 2000 (Invitrogen). The stable clones expressing Cyp-D construct or the empty vector were
C

138 selected via puromycin (5 g/mL) for 12-14 days. Stable cells were subjected to Western blot assay of
AC

139 Cyp-D.
140
141 2.15. HT-116 xenograft. The animal protocols were in line with national and international regulations,
142 and were approved by authors Institutional Animal Care and Use Committee (IACUC). Female severe
143 combined immunodeficiency (SCID) mice of 6-7 weeks old were injected subcutaneously with HCT-116
144 cells (5 million cells/mice) in 30% Matrigel (BD Biosciences). Twenty days later, when tumors were
145 approximately 300 mm3 in volume, mice were randomized and treated with either vehicle (Saline); PF-543
146 (25 mg/kg, i.v., once every two days, for 30 days), and/or CsA (5 mg/kg, i.v., once every two days, for 30
147 days) [10] (n=10 per group). PF-543 and CsA were freshly prepared. Tumor volumes were measured

4
 ACCEPTED MANUSCRIPT
148 weekly by the modified ellipsoid formula: (/6) AB2 , and A is the longest and B is the shortest
149 perpendicular axis of an assumed ellipsoid corresponding to tumor mass [10]. Mice body weights were
150 also recorded weekly as an index of toxicity.
151
152 2.16.Statistical analysis. All data were presented as mean standard error (SE).
153 Repeated-measures analysis of variance (RMANOVA) followed by Dunnetts post hoc test for multiple
154 comparisons (SPSS 16.0) were applied to evaluate statistical significance of observed differences.

PT
155
156 3. Results
157

RI
158 3.1. PF-543 is anti-proliferative and cytotoxic to CRC cells
159

SC
160 We first counted the number of viable (trypan blue exclusive) HCT-116 cells, cultured in regular or
161 PF-543-containing medium. Results in Fig 1a demonstrated that PF-543 at 2.5 and 10 M significantly
inhibited HCT-116 cell proliferation, as the number of viable HCT-116 cells was dramatically decreased

U
162
163 with PF-543 (2.5 and 10 M) treatment (Fig 1a). PF-543 at 0.1 M had no significant effect on HCT-116
AN
164 cell proliferation (Fig 1a). Colony formation assay results in Fig 1b demonstrated that PF-543, at 2.5/10
165 M, significantly reduced the number of viably HCT-116 cell colonies. These results indicate that PF-543
166 inhibits HCT-116 cell proliferation in a titration-dependent manner.
M

167
168 Next, MTT assay was performed to test the effect of PF-543 on CRC cell survival. As shown in Fig 1c,
D

169 HCT-116 cell survival, reflected by the MTT OD value, was significantly inhibited following treatment with
170 PF-543 (96 hrs). The anti-survival effect by PF-543 was again dose-dependent (Fig 1c). Meanwhile, a
TE

171 time-dependent response by PF-543 was also noticed (Fig 1d). It took at least 48 hrs for PF-543 (10 M)
172 to exert significant anti-survival activity (Fig 1d). We also tested the potential role of PF-543 on two other
EP

173 CRC cell lines: DLD-1 and HT-29. MTT assay results in Fig 1e showed that PF-543 was cytotoxic to both
174 cell lines. Notably, HT-29 cells, expressing high-level of SphK1, were more vulnerable to PF-543 than
175 DLD-1 cells, the latter showed low SphK1 expression (Fig 1e, upper panel).
C

176
AC

177 The potential activity of PF-543 on ex-vivo cultured primary human colon cancer cells was also tested.
178 As described, three primary human colon cancer cells (namely patient 1 or P1, P2 and P3) were
179 established from affected patients. PF-543 (2.5 and 10 M, 96 hrs) was cytotoxic to all three lines (Fig 1f).
180 Its activity was most potent in P2 primary cancer cells (Fig 1f), where SphK1 expression was high (Fig 1f,
181 upper panel). Reversely, P1 cells expressed lowest level of SphK1, and were relatively resistant to
182 PF-543 (Fig 1f). These results show that PF-543 is anti-proliferative and cytotoxic against a panel of
183 established and primary human CRC cells, and its activity appears dependent on SphK1 expression
184 level.
185

5
 ACCEPTED MANUSCRIPT
186 3.2. PF-543 mainly induces CRC cell necrotic death
187
188 Above studies demonstrated the cytotoxic effects by PF-543 in CRC cells, we next studied the role of
189 apoptosis in the process. A total of four independent apoptosis assays were carried, including Annexin V
190 FACS assay, ssDNA ELISA assay, caspase-3/-9 activity assay, and Western blot detecting cleaved- poly
191 (ADP-ribose) polymerase (PARP). To our surprise, we failed to observe significant apoptosis induction in
192 PF-543-treated HCT-116 cells (Fig 2a-d). Here, we applied PF-543 at 10 M, which was obviously

PT
193 cytotoxic (Fig 1), and apoptosis was detected at different time points (24, 48, 72 and 96 hrs) (Fig 2a-d).
194 We also tested PF-543 at other concentrations (2.5 and 50 M), and similar no apoptosis induction results
195 were observed (Data not shown). On the other hand, the histone deacetylase inhibitor SAHA

RI
196 (suberoylanilide hydroxamic acid), serving as a positive control, induced significant apoptosis activation in
197 HCT-116 cells (Fig 2a-d). The above experiments were also repeated in two other CRC cell lines and in

SC
198 the three primary human colon cancer lines, same no-apoptosis induction results were obtained (Data not
199 shown).

U
200
201 Next, we wanted to know if PF-543-mediated cytotoxicity was due to necrosis in HCT-116 cells. LDH
AN
202 assay results in Fig 2e showed that PF-543 dose-dependently increased LDH release (an early marker of
203 cell necrosis), which was accompanied with mitochondrial membrane potential (MMP) reduction (Fig 2f).
204 In addition, mito-IB (Western blots analyzing mitochondrial proteins) and mito-IP (mitochondrial
M

205 immunoprecipitation) assay results showed that PF-543 induced P53 translocation to mitochondria (Fig
206 2g), it also formed a complex with Cyp-D (Fig 2g). The latter is the initial step for programmed necrosis
D

207 induction [14,15]. Voltage-dependent anion channel (VDAC), another key functional component of mPTP,
208 was not in the Cyp-D-P53 complex (Fig 2g). Further studies showed that PF-543-induced HCT-116
TE

209 cytotoxicity, or MTT OD reduction, was largely attenuated with co-treatment of the necrosis inhibitor
210 Necrostatin-1 (Nec-1) [15] (Fig 2h), but not by the apoptosis inhibitor z-VAD-fmk (Fig 2h). Same
EP

211 PF-543-induced pro-necrosis, but not pro-apoptosis, results were also seen in two other CRC cell lines
212 (Fig 2i), and in primary human colon cancer cells (Fig 2j). These results indicate that PF-543 mainly
213 induces programmed necrosis, but not apoptosis, in CRC cells.
C

214
AC

215 3.3. Modulation of mitochondrial protein Cyp-D alters PF-543s sensitivity in CRC cells
216
217 Above results indicated that PF-543 induced P53 translocation to mitochondria, and complexation
218 with the mPTP component Cyp-D, which then regulated mPTP opening and cell necrosis. To further
219 support this hypothesis, the Cyp-D inhibitors, including sanglifehrin A (SfA) and cyclosporin A (CsA) [7,8],
220 were utilized. As demonstrated, the two Cyp-D inhibitors significantly attenuated PF-543-induced MMP
221 reduction in HCT-116 cells (Fig 3a), indicating mPTP blockage. As a result, PF-543-induced HCT-116
222 cytotoxicity (MTT OD reduction) (Fig 3b) and necrosis (LDH release) (Fig 3c) were dramatically inhibited
223 with co-treatment of SfA/CsA. Similar results were also seen in two other CRC cell lines (Data not shown).

6
 ACCEPTED MANUSCRIPT
224 Note that SfA or CsA alone showed almost no effect on HCT-116 cell survival nor necrosis (Fig 3b and c).
225
226 To rule out the possible off-target effects of the Cyp-D inhibitors, shRNA method was applied to stably
227 knockdown Cyp-D in HCT-116 cells. Western blot results verified Cyp-D downregulation in stable
228 HCT-116 cells expressing Cyp-D-shRNAs (Fig 3d). Here, two shRNAs targeting non-overlapping Cyp-D
229 cDNA sequences (seq-1 and seq-2) were applied (see methods), each displayed high efficiency in
230 silencing Cyp-D (Fig 3d). Notably, we showed that stable HCT-116 cells bearing Cyp-D-shRNAs (seq-1/-2)

PT
231 were resistant to PF-543 (Fig 3e and f). PF-543-induced cytotoxicity and necrosis were largely inhibited in
232 Cyp-D-silenced HCT-116 cells (Fig 3e and f).
233

RI
234 Based on the above results, we would speculate that PF-543-induced activity could be further
235 enhanced with Cyp-D over-expression. Using the method described, we were able to establish two stable

SC
236 HCT-116 cell lines expressing wild-type Cyp-D (Flag-tagged) (Fig 3g). MTT assay (Fig 3h) and LDH
237 assay (Fig 3i) results demonstrated that PF-543-induced cytotoxicity was augmented in
Cyp-D-overexpressed HCT-116 cells. In another word, PF-543s sensitivity was significantly increased

U
238
239 with Cyp-D overexpression. The above Cyp-D shRNA and over-expression experiments were also
AN
240 repeated in two other CRC cell lines (DLD-1 and HT-29), and similar results were obtained. These results
241 together support a key role of Cyp-D-mediated necrosis pathway in mediating PF-543s actions against
242 CRC cells.
M

243
244 3.4. PF-543 inhibits HCT-116 tumor growth in SCID mice
D

245
246 Finally, we studied the potential anti-cancer effects by PF-543 in vivo. A significant number of
TE

247 HCT-116 cells were inoculated to SCID mice to establish tumor xenograft model. Tumor growth curve
248 results in Fig 4a demonstrated that PF-543 administration (25 mg/kg, i.v., once every two days)
EP

249 remarkably suppressed HCT-116 xenograft growth in SCID mice, resulting in significant tumor volume
250 reduction (Fig 4a). In addition, the survival PF-543-administrated mice was remarkably improved, as
251 compared to the vehicle (Saline) mice (Fig 4b). Significantly, the in vivo anti-cancer activity by PF-543
C

252 was largely attenuated with co-administration of the Cyp-D inhibitor CsA (Fig 4a and b), suggesting that
AC

253 Cyp-D-mediated programmed-necrosis pathway also participated in PF-543-exerted activity in vivo. In

254 line with previous studies [10], CsA alone at tested concentration (25 mg/kg, i.v.) failed to affect tumor
255 growth or mice survival (Fig 4a and b). Notably, the mice body weights were not significantly affected by
256 the tested regimens (Fig 4c). Neither did we notice any signs of other apparent toxicities in the animals.
257 These results together demonstrated that PF-543 inhibited HCT-116 xenografts growth in SCID mice, and
258 its activity in vivo was significantly attenuated with co-administration of CsA.
259
260 4. Discussion
261

7
 ACCEPTED MANUSCRIPT
262 SphK1 inhibitors have demonstrated promising anti-cancer results in preclinical CRC models, either
263 alone or in combination with other conventional anti-cancer agents [16,17]. In the current study, we
264 showed that PF-543, a novel, specific and extremely potent SphK1 inhibitor, exerted significant anti-CRC
265 activity in vitro and in vivo. We provided evidence to show that programmed necrosis pathway, but not
266 apoptosis, likely mediated PF-543s activity against CRC cells.
267
268 Necrotic cell death, or cell necrosis, is previously considered to be a passive cell death process. Yet,

PT
269 recent studies have implied that necrosis is also a precisely-regulated event [18,19]. In particularly,
270 stress-induced P53 mitochondrial translocation and subsequent mPTP opening are vital in regulating the
271 programmed necrosis activation [14,18,19,20]. mPTP is composed of three primary components,

RI
272 including VDAC, the adenine nucleotide translocator-1 (ANT-1) and Cyp-D [18,19]. A number of stimuli,
273 including Ca2+ overload, sustained hypoxia, ultraviolent (UV) radiation, as well as several anti-cancer

SC
274 drugs were shown to induce P53 mitochondrial translocation and Cyp-D association, leading to mPTP
275 opening [12,19,21,22,23]. The mitochondria will then swell to mediate cell necrosis [12,19,21,22,23]. In
the current study, our results suggest that this mitochondrial programmed necrosis pathway also

U
276
277 mediates PF-543-induced cytotoxicity against CRC cells.
AN
278
279 CsA and SfA suppress mPTP-dependent programmed necrosis pathway via binding to Cyp-D, thus
280 inhibiting its peptidyl-prolylcis-transisomerase (PPIase) activity [7,8]. These two Cyp-D inhibitors were
M

281 shown to block programmed necrosis by a number of cytotoxic agents or stresses. For example,
282 salinomycin-induced glioma cell necrosis was inhibited by the Cyp-D inhibitors [15]. In addition, CsA was
D

283 shown to alleviate gemcitabine-induced cytotoxicity in pancreatic cancer cells [11]. Meanwhile,
284 doxorubicin-induced non-apoptotic death of lung cancer cells was attenuated with pretreatment of CsA
TE

285 [24]. In line with these studies, we showed that CsA and SfA remarkably inhibited PF-543-induced CRC
286 cell necrosis. At the meantime, PF-543 cytotoxicity was significantly alleviated in Cyp-D-silenced (by
EP

287 shRNA) CRC cells. On the other hand, Cyp-D over-expression dramatically enhanced PF-543 sensitivity
288 and cytotoxicity in CRC cells. These results indicate that Cyp-D-dependent programmed necrosis is
289 responsible PF-543 cytotoxicity against CRC cells. In summary, our results indicate that PF-543 exerts
C

290 potent anti-CRC activity in vitro and in vivo, possibly via inducing cancer cell programmed necrosis.
AC

291
292 5. Conflict of interests
293
294 The authors have no conflict of interests
295
296

297

8
 ACCEPTED MANUSCRIPT
298 6. References
299

300 [1] H.J. Schmoll, A. Stein, Colorectal cancer in 2013: Towards improved drugs, combinations and patient selection, Nat

301 Rev Clin Oncol 11 (2014) 79-80.

302 [2] E.J. Kuipers, T. Rosch, M. Bretthauer, Colorectal cancer screening--optimizing current strategies and new directions,

303 Nat Rev Clin Oncol 10 (2013) 130-142.

PT
304 [3] A. Brouquet, B. Nordlinger, Metastatic colorectal cancer outcome and fatty liver disease, Nat Rev Gastroenterol

305 Hepatol 10 (2013) 266-267.

RI
306 [4] D. Shida, K. Takabe, D. Kapitonov, S. Milstien, S. Spiegel, Targeting SphK1 as a new strategy against cancer, Curr

SC
307 Drug Targets 9 (2008) 662-673.

308 [5] M. Vadas, P. Xia, G. McCaughan, J. Gamble, The role of sphingosine kinase 1 in cancer: oncogene or non-oncogene

309 addiction?, Biochim Biophys Acta 1781 (2008) 442-447.

310
U
[6] M.E. Schnute, M.D. McReynolds, T. Kasten, M. Yates, G. Jerome, J.W. Rains, T. Hall, J. Chrencik, M. Kraus, C.N.
AN
311 Cronin, M. Saabye, M.K. Highkin, R. Broadus, S. Ogawa, K. Cukyne, L.E. Zawadzke, V. Peterkin, K. Iyanar, J.A.

312 Scholten, J. Wendling, H. Fujiwara, O. Nemirovskiy, A.J. Wittwer, M.M. Nagiec, Modulation of cellular S1P
M

313 levels with a novel, potent and specific inhibitor of sphingosine kinase-1, Biochem J (2012).

314 [7] S.J. Clarke, G.P. McStay, A.P. Halestrap, Sanglifehrin A acts as a potent inhibitor of the mitochondrial permeability
D

315 transition and reperfusion injury of the heart by binding to cyclophilin-D at a different site from cyclosporin
TE

316 A, J Biol Chem 277 (2002) 34793-34799.

317 [8] P.G. Sullivan, M.B. Thompson, S.W. Scheff, Cyclosporin A attenuates acute mitochondrial dysfunction following
EP

318 traumatic brain injury, Exp Neurol 160 (1999) 226-234.

319 [9] Y. Qiu, T. Yu, W. Wang, K. Pan, D. Shi, H. Sun, Curcumin-induced melanoma cell death is associated with
C

320 mitochondrial permeability transition pore (mPTP) opening, Biochem Biophys Res Commun 448 (2014)
AC

321 15-21.

322 [10] C. Zhou, Z. Chen, X. Lu, H. Wu, Q. Yang, D. Xu, Icaritin activates JNK-dependent mPTP necrosis pathway in

323 colorectal cancer cells, Tumour Biol (2015).

324 [11] S.H. Chen, D.L. Li, F. Yang, Z. Wu, Y.Y. Zhao, Y. Jiang, Gemcitabine-induced pancreatic cancer cell death is

325 associated with MST1/cyclophilin D mitochondrial complexation, Biochimie 103 (2014) 71-79.

326 [12] C. Ji, B. Yang, Z. Yang, Y. Tu, Y.L. Yang, L. He, Z.G. Bi, Ultra-violet B (UVB)-induced skin cell death occurs through a

327 cyclophilin D intrinsic signaling pathway, Biochem Biophys Res Commun 425 (2012) 825-829.

9
 ACCEPTED MANUSCRIPT

328 [13] S. Zamorano, D. Rojas-Rivera, F. Lisbona, V. Parra, F.A. Court, R. Villegas, E.H. Cheng, S.J. Korsmeyer, S. Lavandero,

329 C. Hetz, A BAX/BAK and cyclophilin D-independent intrinsic apoptosis pathway, PLoS One 7 (2012) e37782.

330 [14] A.V. Vaseva, N.D. Marchenko, K. Ji, S.E. Tsirka, S. Holzmann, U.M. Moll, p53 opens the mitochondrial

331 permeability transition pore to trigger necrosis, Cell 149 (2012) 1536-1548.

332 [15] L.S. Qin, P.F. Jia, Z.Q. Zhang, S.M. Zhang, ROS-p53-cyclophilin-D signaling mediates salinomycin-induced glioma

PT
333 cell necrosis, J Exp Clin Cancer Res 34 (2015) 57.

334 [16] T. Kawamori, T. Kaneshiro, M. Okumura, S. Maalouf, A. Uflacker, J. Bielawski, Y.A. Hannun, L.M. Obeid, Role for

RI
335 sphingosine kinase 1 in colon carcinogenesis, FASEB J 23 (2009) 405-414.

336 [17] M. Xiao, Y. Liu, F. Zou, Sensitization of human colon cancer cells to sodium butyrate-induced apoptosis by

SC
337 modulation of sphingosine kinase 2 and protein kinase D, Exp Cell Res 318 (2012) 43-52.

338 [18] J.W. Elrod, J.D. Molkentin, Physiologic functions of cyclophilin D and the mitochondrial permeability transition

U
339 pore, Circ J 77 (2013) 1111-1122.
AN
340 [19] A.P. Halestrap, Calcium, mitochondria and reperfusion injury: a pore way to die, Biochem Soc Trans 34 (2006)

341 232-237.
M

342 [20] H. Du, L. Guo, F. Fang, D. Chen, A.A. Sosunov, G.M. McKhann, Y. Yan, C. Wang, H. Zhang, J.D. Molkentin, F.J.

343 Gunn-Moore, J.P. Vonsattel, O. Arancio, J.X. Chen, S.D. Yan, Cyclophilin D deficiency attenuates mitochondrial
D

344 and neuronal perturbation and ameliorates learning and memory in Alzheimer's disease, Nat Med 14 (2008)
TE

345 1097-1105.

346 [21] B. Chen, M. Xu, H. Zhang, J.X. Wang, P. Zheng, L. Gong, G.J. Wu, T. Dai, Cisplatin-induced non-apoptotic death of
EP

347 pancreatic cancer cells requires mitochondrial cyclophilin-D-p53 signaling, Biochem Biophys Res Commun

348 437 (2013) 526-531.

C

349 [22] S.H. Chen, D.L. Li, F. Yang, Z. Wu, Y.Y. Zhao, Y. Jiang, Gemcitabine-induced pancreatic cancer cell death is

350 associated with MST1/Cyclophilin D mitochondrial complexation, Biochimie 103 (2014) 71-79.
AC

351 [23] L.P. Zhao, C. Ji, P.H. Lu, C. Li, B. Xu, H. Gao, Oxygen glucose deprivation (OGD)/re-oxygenation-induced in vitro

352 neuronal cell death involves mitochondrial cyclophilin-D/P53 signaling axis, Neurochem Res 38 (2013)

353 705-713.

354 [24] J.H. Lu, Z.F. Shi, H. Xu, The mitochondrial cyclophilin D/p53 complexation mediates doxorubicin-induced

355 non-apoptotic death of A549 lung cancer cells, Mol Cell Biochem 389 (2014) 17-24.

356
357
10
 ACCEPTED MANUSCRIPT
358 Figure legends
359
360 Fig 1. PF-543 is anti-proliferative and cytotoxic to CRC cells. Human CRC cell lines, including
361 HCT-116 (a-d), DLD-1 and HT-29 (e), as well as patient-derived primary colon cancer cells (P1, P2 and
362 P3, f), were incubated with regular or PF-543-containg medium for indicated time, cell proliferation was
363 tested by viable cell counting assay (a) and colony formation assay (b); Cell survival was examined by
364 MTT assay (c-f). Expression of SphK1 and Tubulin (the equal loading) was examined by Western blots (e

PT
365 and f, upper panels). Relative SphK1 expression (verse Tubulin) was quantified. C stands for untreated
366 control cells (For all Figs). Values represented the mean SEM (n=5). Experiments in this figure were
367 repeated three times, with similar results achieved. * P < 0.05 versus C group.

RI
368
369 Fig 2. PF-543 mainly induces necrotic death of CRC cells. HCT-116 cells were treated with PF-543

SC
370 (10 M) or SAHA (2.5 M) for applied time, cell apoptosis was tested by listed assays (a-c); Expression of
371 cleaved PARP (C-PARP) and Tubulin (loading) was tested (d); LDH content in the conditional medium
was shown (e); The MMP reduction was reflected by the JC-10 dye assay (f); Mitochondrial expression

U
372
373 (Mito-IB) and association (Mito-IP) of Cyp-D, P53 and VDAC were tested using described methods (g,
AN
374 left), and expression of the proteins in whole cell lysates (WCL-IB) was tested as inputs (g, right under).
375 HCT-116 cells, pretreated with the apoptosis inhibitor z-VAD-fmk (50 M) or the necrosis inhibitor Nec-1
376 (10 M) for 1 hr, were treated with PF-543 (2.5/10 M) for 96 hrs, cell survival was tested by MTT assay
M

377 (h). DLD-1 and HT-29 cells (i), or the primary human colon cancer cells (P2/P3, f), were incubated in
378 PF-543 (10 M)-containing medium for 72 hrs, cell apoptosis and necrosis were tested by Annexin V
D

379 FACS assay and LDH release assay, respectively. Cyp-D bound P53 was quantified (n=4, g). Values
380 represented the mean SEM (n=5). Experiments in this figure were repeated four times, with similar
TE

381 results achieved. * P < 0.05 versus C group. # P < 0.05 versus PF-543 only group.
382
EP

383 Fig 3. Modulation of mitochondrial protein Cyp-D alters PF-543s sensitivity. HCT-116 cells,
384 pre-added with sanglifehrin A (SfA, 2.5 M) or cyclosporin A (CsA, 0.25 M) for 1 hr, were treated with
385 PF-543 (2.5/10 M) for applied time, MMP reduction, cell survival and necrosis were tested by JC-10 dye
C

386 assay (a), MTT assay (b) and LDH release assay (c), respectively. Expression of Cyp-D and Tubulin (the
AC

387 equal loading) in stable HCT-116 cells expressing scramble control shRNA (shScr) (d), Cyp-D shRNAs
388 (shCyp-D, seq-1/2) (d), the empty vector (pSuper-puro) (g) or pSuper-puro-Cyp-D-Flag (Cyp-D-Flag,
389 two lines) (g) was shown. Above cells were treated with PF-543 (2.5/10 M) for applied time, cell survival
390 (e and h) and necrosis (f and i) were tested. Parental stands for parental cells (d and g). Transfection
391 stands for the transfection reagent (Lipofectamine 2000) alone (d). Values represented the mean SEM
392 (n=5). Experiments in this figure were repeated three times, with similar results achieved. * P < 0.05
393 versus C group (a-c). # P < 0.05 versus PF-543 only group (a-c). * P < 0.05 versus C of shScr group
394 or Vector group (d-i) .# P < 0.05 versus PF-543 only group of shScr group or Vector group (e, f, h
395 and i).

11
 ACCEPTED MANUSCRIPT
396
397 Fig 4. PF-543 inhibits HCT-116 xenografts in SCID mice. SCID mice (n = 10 per group) bearing
398 HCT-116 xenografts were administrated with: Saline (Vehicle), PF-543 (25 mg/kg, i.v., once every two
399 days, 30 days), and/or CsA (5 mg/kg, i.v. once every two days, 30 days), tumor volumes (a) and mice
400 body weights (c) were recorded every week. Mice survival at week-6 was also shown (b). In vivo
401 experiments were repeated twice, with similar results obtained. * P < 0.05 versus Vehicle group. # P <
402 0.05 versus PF-543 only group.

PT
403
404

RI
U SC
AN
M
D
TE
C EP
AC

12
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP

Fig 1
a. c. d. 120
C

HCT-116 120
6000 C
PF-543 (0.1 M) 100
100
5000 PF-543 (2.5 M)
AC

Survival (% vs. C)

*
Survival (% vs. C)
Viable cell count

PF-543 (10 M) 80 80
4000

3000
* 60 * 60 *
** * * 40
*
2000 40 *
*
1000 20 20
*
0 0 0
Day 0 Day 2 Day 4 C 0.1 1 2.5 10 50 C 24 48 72 96 120 hrs
PF-543 (M), 96 hrs PF-543 (10 M)
b. e. HCT-29
f. Primary human colon cancer cells
HCT-116

120 120
DLD-1
HT-29

120 P1 P2 P3
Viable colony count (per dish)

DLD-1 45kD- -SphK1

45kD- -SphK1
0.11 0.70 0.31
100 100 1.64 0.10 0.96
100
55kD- -Tubulin 55kD- -Tubulin
*
Survival (% vs. C)
Survival (% vs. C)

80 80 80
* * *
60 60 * * 60 *
40 40 40
*
*
HCT-116

*
* *
DLD-1

20
* * 20 20
*
P1 P2 P3 P1 P2 P3 P1 P2 P3
0 0 0
C 0.1 2.5 10 C 0.1 1 2.5 10 50 C 2.5 10
PF-543 (M), 12 days PF-543 (M), 96 hrs PF-543 (M), 96 hrs
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M

Fig 2
a. b. c. d.
HCT-116
20 0.9
*
D

14 PF-543 (10 M), hrs

18
16
0.8
* 12
Activity of Caspase-3
Activity of Caspase-9 * C 24 48 72 SAHA
0.7 116kD- -PARP
ssDNA ELISA OD
Annexin V ratio

14
TE

0.6 10 89kD-
-C-PARP
Folds (vs. C)

12
0.5 8 -Tubulin
10 55kD-
0.4 6 1.0
8 C-PARP/Tubulin
6 0.3
4
0.8 * *
0.2 0.6
4
EP

2 0.4
2 0.1 0.2
0 0 0 0.0
C 24 48 72 96 48 C 24 48 7296 48, hrs C 24 48 72 96 48, hrs C 24 48 72 48, hrs
PF-543 (10 M), hrs SAHA PF-543 (10 M), hrs SAHA PF-543 (10 M), hrs SAHA PF-543 (10 M) SAHA

e. f. g.
C

IgG Mito-IP: CypD Cyp-D bound p53

0.8
* *
2
45 PF-543 10 C 2.5 10 M, 6 hrs
* 53kD- -p53
0.6 *
MMP reduction (Folds vs. C)

40
*
AC

-Cyp-D 0.4
35 1.5 40kD-
*
Medium LDH (%)

30 * * 32kD- -VDAC 0.2

25 * 1 50kD- -IgG 0.0

C
2.5 10 M
20 PF-543, 6 hrs
15 Mito-IB PF-543, 6 hrs WCL-IB PF-543, 6 hrs
0.5 C 2.5 10 M C 2.5 10 M
10 * 53kD- -p53 53kD- -p53
5
37kD- -Cyp-D 37kD- -Cyp-D
0 0
C 0.1 1 2.5 10 50 C 0.1 1 2.5 10 50 32kD- -VDAC 32kD- -VDAC
PF-543 (M), 96 hrs PF-543 (M), 6 hrs 55kD- -Tubulin -Tubulin
55kD-

h. 120 Veh
z-VAD-fmk
i. 40 j. 30

100 Nec-1 * *
#
#
* 30
Survival (% vs. C)

80 C
* C C C
Percentage

20
Percentage

PF-543 PF-543 PF-543 PF-543

60 ** 20 *
40
** 10
10

20 *
0 0 0
C 2.5 10 Annexin V LDH Annexin V LDH Annexin V LDH Annexin V LDH
DLD-1 HT-29 Primary human colon Primary human colon
PF-543 (M), 96 hrs
cancer cells (P2) cancer cells (P3)
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
Fig 3
M

a. b. c.
2 120 Veh 25
HCT-116
D

SfA Veh
MMP reduction (Folds vs. C)

Veh
SfA * 100 #
CsA SfA
*
CsA *# # # 20 CsA
**
Survival (% vs. C)

**
# #
Medium LDH (%)

#
# 80 *
TE

** ** 15
#
1 60 * #
* *#
40 * 10 ** #

5
EP

20

0 0 0
C 2.5 10 C 2.5 10 C 2.5 10
PF-543 (M), 6 hrs PF-543 (M), 96 hrs PF-543 (M), 96 hrs
d. e.120 f.25
C

shScr
shScr
(s )
2)
-D q1

shCyp-D (seq1)
eq

shCyp-D (seq1)
yp (se
on

shCyp-D (seq2)
100 *
Sc cti

shCyp-D (seq2)
sh -D

#
an l

20
Tr nta
sh sfe

AC

Medium LDH (%)

#
yp

**
sh r

*
Survival (% vs. C)
re

#
C
C
Pa

80 #
55kD- -Tubulin
* ** 15
#
#

**
60
37kD- -Cyp-D # #
10
1.5
40 * **
20 5
1.0
0.5 0 0
0.0 * * C 2.5 10 C 2.5 10
CypD/Tubulin PF-543 (M), 96 hrs PF-543 (M), 96 hrs
g. h. i.
)

)
(1

(2

120 Vector 45 Vector

ag

ag

#
l

-fl

-fl

Cyp-D-flag (1)
ta

Cyp-D-flag (1)
or

40
-D

-D
en

*
ct

Cyp-D-flag (2)
yp

yp

100
r

#
Pa

Ve

Cyp-D-flag (2)
C

35 #
*
Medium LDH (%)

-Cyp-D-Flag
80 30 # *
Survival (% vs. C)

37kD- -Cyp-D
* 25 *
55kD- -Tubulin 60
20 *
4.0 #
3.0 * 40
*# * 15 *
* # 10
2.0 20 * * *# 5
1.0
0 0
0.0 C 2.5 10 C 2.5 10
CypD/Tubulin PF-543 (M), 96 hrs PF-543 (M), 96 hrs
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
C EP
AC

Fig 4

a. 3
b. c.
HCT-116 tumor volume (mm ) Mice survival (at week-6) Mice body weights (g)
120%
2500 24 Vehicle
Vehicle
100%
*
10/10 CsA
CsA
2000 PF-543
PF-543
# 80% PF-543+CsA
PF-543+CsA 22
1500 * 60% #
1000 40%
*
4/10
20

500 * 20%
1/10 1/10

0 0% 18
e
sA

43 3
sA

1 2 3 4 5 6 7 8 weeks 1 2 3 4 5 6 7 8 weeks
cl

4
PF F-5
C

+C
hi
Ve

P
-5
 ACCEPTED MANUSCRIPT
Highlights of paper by TongFa Ju et al.,

1. PF-543 is anti-proliferative and cytotoxic to established and primary CRC cells

2. PF-543 induces programmed necrosis, but not apoptosis, in CRC cells

3. Modulation of mitochondrial protein cyclophilin-D alters PF-543s sensitivity

PT
4. PF-543 inhibits HCT-116 xenograft growth in SCID mice, improving mice survival

5. Co-administration of cyclophilin-D inhibitor CsA inhibits PF-543s activity in vivo

RI
U SC
AN
M
D
TE
C EP
AC