Pentamidine niosomes thwart S100B effects in human colon carcinoma biopsies favouring wtp53 rescue

Abstract S100B protein bridges chronic mucosal inflammation and colorectal cancer given its ability to activate NF‐kappaB transcription via RAGE signalling and sequestrate pro‐apoptotic wtp53. Being an S100B inhibitor, pentamidine antagonizes S100B‐wtp53 interaction, restoring wtp53‐mediated pro‐apoptotic control in cancer cells in several types of tumours. The expression of S100B, pro‐inflammatory molecules and wtp53 protein was evaluated in human biopsies deriving from controls, ulcerative colitis and colon cancer patients at baseline (a) and (b) following S100B targeting with niosomal PENtamidine VEhiculation (PENVE), to maximize drug permeabilization in the tissue. Cultured biopsies underwent immunoblot, EMSA, ELISA and biochemical assays for S100B and related pro‐inflammatory/pro‐apoptotic proteins. Exogenous S100B (0.005‐5 μmol/L) alone, or in the presence of PENVE (0.005‐5 μmol/L), was tested in control biopsies while PENVE (5 μmol/L) was evaluated on control, peritumoral, ulcerative colitis and colon cancer biopsies. Our data show that S100B level progressively increases in control, peritumoral, ulcerative colitis and colon cancer enabling a pro‐inflammatory/angiogenic and antiapoptotic environment, featured by iNOS, VEGF and IL‐6 up‐regulation and wtp53 and Bax inhibition. PENVE inhibited S100B activity, reducing its capability to activate RAGE/phosphor‐p38 MAPK/NF‐kappaB and favouring its disengagement with wtp53. PENVE blocks S100B activity and rescues wtp53 expression determining pro‐apoptotic control in colon cancer, suggesting pentamidine as a potential anticancer drug.


| INTRODUC TI ON
By promoting the release of cytokines, interleukins and other pro-inflammatory signalling molecules, chronic intestinal inflammation significantly contributes to the carcinogenic microenvironment 1,2 and genomic instability able to escape the control of tumour suppressor factors, such as wtp53. 3,4 In this context, enteric glial S100B protein overexpression has been linked to the typical features of reactive gliosis, driving the progression from chronic intestinal inflammation to colonic neoplastic lesions. S100B is a neurotrophin, constitutionally and specifically expressed by enteric glial cells (EGCs) in the enteric nervous system, which belongs to a multigene family of diffusible Ca 2+ /Zn 2+ -binding proteins. [5][6][7] It is known that S100B protein overexpression correlates with poor prognosis in melanoma 8,9 and glioma, 10 and early relapse following curative resection in colorectal carcinoma patients, 11 suggesting its direct involvement in the perpetuation of a tumour-promoting microenvironment. At micromolar concentrations, S100B accumulates at the receptor for advanced glycation end products (RAGE) site and such interaction leads to mitogen-activated protein kinase (MAPK) phosphorylation and nuclear factor-κB (NF-κB) activation. [12][13][14] This event, in turn, promotes the downstream release of pro-inflammatory cytokines and the transcription of inducible nitric oxide (iNOS). Interestingly, S100B has also been proposed as an inhibitor of wtp53, 15 a key pro-apoptotic protein linked to colon carcinogenesis. By interacting with the C-terminus of wtp53, S100B prevents its tetramerization and protein kinase C-mediated phosphorylation, inhibiting the transcriptional and tumour suppressor activity of wtp53. 15,16 The targeting of enteric glial S100B protein and wtp53/S100B interaction might thus represent a new strategy in colorectal carcinoma therapy. 13 In this context, the well-known anti-protozoal drug pentamidine 17 has been shown to disrupt S100B-wtp53 interaction in melanoma and glioma cells, acting as an inhibitor of S100B pro-cancerogenic activity. 8,18,19 This evidence has clinically translated into ongoing clinical trials, aiming at testing the efficacy of pentamidine as an anticancer drug, in melanoma patients and in patients with metastatic colon cancer undergoing standard chemotherapy as second-line and/or third-line treatment (ClinicalTrials.gov Identifier: NCT00729807 and NCT00809796, respectively). We have previously demonstrated that pentamidine exerts marked anti-inflammatory effects in a mice model of ulcerative colitis, by likely targeting S100B-wtp53 interaction. 20 However, the effects and the mechanisms of pentamidine on S100B-mediated inflammation in colon cancer have not been explored yet.
In the present study, we explored (a) the endogenous expression of glial S100B protein in human colonic mucosa from healthy, peritumoral, ulcerative colitis (UC) and cancer biopsies and its correlation with the expression of pro-inflammatory markers and pro-apoptotic factors. We also assessed (b) the effects of increasing concentrations of exogenous S100B protein (0.005-5 µmol/L) in control biopsies. To maximize pentamidine efficacy to block glial S100B-induced pro-inflammatory and pro-cancerogenic effects, we tested (c) increasing concentrations (0.005-5 µmol/L) of chitosan-coated vesicular formulation of pentamidine (PENVE, PENtamidineVEhiculation) 21,22 in human healthy mucosal biopsies after the challenge with exogenous S100B proteins (5 µmol/L). Finally, we investigated (d) whether pentamidine vehiculation by PENVE (5 µmol/L) was able to rescue mucosal pro-inflammatory markers and pro-apoptotic factors in order to evaluate whether S100B/wtp53 targeting by PENVE might potentially represent an alternative strategy to the current colon cancer chemotherapy.

| Experimental design
We collected surgical specimens of peritumoral and tumoral areas from ten patients diagnosed with colon cancer (6 females; mean age 47 ± 0.5 years). None of the patients had a familial history of colon cancer; hence, all cancers were considered sporadic. Patients were diagnosed with left (5 patients, 2 females) and sigmoid (5 patients, 4 females) sporadic colon cancer with no evidence of lymph nodes or distant metastasis and/or local invasion at pre-operative staging (T1 or T2, N0, M0). As positive controls, we collected four mucosal biopsies from the recto-sigmoid region of eight UC patients (5 females; mean age 47 ± 0.7 years) undergoing a colonoscopy for relapse of rectal bleeding. All patients had a proven histological diagnosis of UC with a Mayo 2 score at endoscopy and none of them had dysplastic modifications at routine histopathological examinations. Four rectosigmoid biopsies form eight otherwise healthy individuals (2 females; mean age 50 ± 1.1 years) undergoing colonoscopy for colon cancer screening served as controls.
Human colonic mucosal samples were used for the experiments and were divided into four groups, as follows: (a) control group comprising colonic specimens collected from eight controls undergoing colonoscopy for colon cancer screening (6 males; mean age 50 years); (b) peritumoral group comprising surgical specimens of cancer-free peritumoral areas (distance >5 cm from tumour site) collected from eight patients diagnosed with colon cancer (3 males; mean age 47 years); (c) ulcerative colitis group comprising surgical colonic specimens collected from eight UC patients (3 males; mean age 47 years); (d) colon cancer group comprising surgical specimens of tumoral areas collected from ten patients diagnosed with colon cancer (2 males; mean age 47 years).
All patients received and signed an informed consent, and all procedures were approved by the ethical committee of the University of Naples 'Federico II'. Mucosal biopsies were cut in thin slices (400 µm) using a Vibratome VT1200 (Leica Microsystem) to get organotypic culture according to the procedure described. 23 Specimens were rapidly washed in ice-cold sterile PBS 1×, orientated and immobilized using cyanoacrylate glue. The vibration amplitude was set at 2.95-3.00 mm. The slices were then cultured in 6-well plates in FBS-supplemented Dulbecco Modified Eagle's Medium (DMEM) at 37°C in 5% CO 2 /95%. In the first set of experiments, specimens were cultured for 24 hours to assess the basal expression of S100B and other pro-inflammatory and pro-apoptotic protein expression (see Scheme 1). In parallel, mucosal biopsies from the control group were challenged with increasing concentrations of exogenous S100B protein (0.005, 0.05, 0.5 and 5 µmol/L) orco-incubated with S100B (5 µmol/L) and PENVE at increasing concentrations (0.005, 0.05, 0.5 and 5 µmol/L) for 24 hours. After the incubation, supernatants were isolated, and specimens underwent homogenization for biochemical and molecular studies. In other experiments, mucosal biopsies from controls, peritumoral, ulcerative colitis and colon cancer patients were cultured as above described, treated with PENVE (5 µmol/L) for 24 hours and subsequently processed for both biochemical and histological procedures (see Scheme 1).

| Preparation and characterization of chitosancoated niosome of pentamidine (PENVE)
Niosomes were prepared using thin-film hydration method. 21 Tween 20 (7.5 mmol/L), cholesterol (15 mmol/L) and DCP (7.5 mmol/L) were dissolved in an organic solvent mixture (chloroform/methanol 3:1 v/v) that was then evaporated using rotary evaporator (VV2000, Heidolph) to form a thin 'film'. The film was hydrated using 5 mL of pentamidine solution (5 mg/mL), vortexed and sonicated at 60°C and 16% amplitude for 5 minutes using ultrasonic microprobe (Vibra-Cell VCX-400, Sonics & Materials). The unilamellar vesicle suspension was purified by gel filtration chromatography using Sephadex G75 with HEPES buffer as eluent. The chitosan (low molecular weight) solution was obtained solubilizing chitosan in acetate buffer (0.2 mol/L, pH 4.4) after overnight stirring. The chitosan-coated niosomes were obtained by adding 1 mL of chitosan solution to an equal volume of uncoated niosomes. The suspension was stirred for 3 hours in a thermostatic water bath at 10°C. Release profiles of pentamidine from PENVE were evaluated in vitro using cellulose membranes. 24
Subsequently, the relative bands were quantified by densitometric scanning with Versadoc (Bio-Rad Laboratories) and a computer program (Quantity One Software, Bio-Rad Laboratories). 32 P-γ-ATP was from Amersham. Poly dI-dC was from Boehringer-Mannheim.
Oligonucleotide synthesis was performed to our specifications by Tib Molbiol (Boehringer-Mannheim).

| NO quantification
NO was measured as nitrite (NO2-) accumulation in human biopsies supernatants by a spectrophotometer assay based on the Griess reaction as previously described. 25

| Enzyme-linked immunosorbent assay for S100B, IL-6 and VEGF
Enzyme-linked immunosorbent assay (ELISA) for S100B (Biovendor R&D) IL-6 and VEGF (all from Thermo Fisher Scientific) was carried out on human biopsies supernatants according to the manufacturer's protocol.

| Myeloperoxidase assay
Myeloperoxidase (MPO), a marker of polymorphonuclear leucocyte accumulation and general inflammation occurring in colonic tissues, was determined as previously described. 26 After removal, human colonic tissues were rinsed with a cold saline solution, opened and deprived of the mucosa using a glass slide. The resulting layer was then homogenized in a solution containing 0.5% hexadecyltrimethylammonium bromide (Sigma-Aldrich) dissolved in 10 mmol/L potassium phosphate buffer and centrifuged for 30 minutes at 20 000 g at 37°C. An aliquot of the supernatant was mixed with a solution of tetramethylbenzidine (1.6 mmol/L; Sigma-Aldrich) and 0.1 mmol/L hydrogen peroxide (Sigma-Aldrich). The absorbance was then spectrophotometrically measured at 650 nm. MPO activity was determined as the amount of enzyme degrading 1 mmol/min of peroxide at 37°C and was expressed in milli units per 100 mg of wet tissue weight.

| Malondialdehyde (MDA) quantification
Malondialdehyde (MDA) was measured with the thiobarbituric acid colorimetric assay in the tissues. 27 Briefly, 1 mL 10% (w/v) trichloroacetic acid was added to 450 μL of tissue lysate. After centrifugation, 1.3 mL 0.5% (w/v) thiobarbituric acid was added and the mixture was heated at 80°C for 20 minutes. After cooling, MDA formation was recorded (absorbance 530 nm and absorbance 550 nm) in a Perkin Elmer spectrofluorometer and the results were presented as ng MDA/mL.

| Immunohistochemistry
After the treatments, mucosal biopsies were fixed in buffered formalin, embedded in paraffin and cut into 5 μm-thick serial sections.
According to the manufacturer's instructions, after heat-mediated antigen retrieval, the tissue was formaldehyde fixed and blocked

| Statistical analysis
Results were expressed as mean ± SEM of n experiments. Statistical analysis was performed using parametric one-way analysis of variance (ANOVA) and multiple comparisons were performed by Bonferroni's post hoc test; P values <.05 were considered significant.

| D ISCUSS I ON
Even though genetic mutations in epithelial stem cells are considered the main step in colon cancer development, 28,29 compelling recent data indicate that chronic dysfunctions of the surrounding cell microenvironment are critical for cancer development and progression. 30,31 Among the secreted factors in the pro-cancerogenic environment, S100B protein is an EGC-derived neurotrophic protein, commonly up-regulated in intestinal inflammatory conditions and associated with tumour progression and prognosis. 15,19,32 Our results have shown a progressive increase in S100B levels human peritumoral, UC and tumour mucosal biopsies, which was correlated with the downstream activation of the proliferative signalling pathway RAGE/MAPK/NF-kappaB and a parallel reduction of proapoptotic protein wtp53 expression. Besides its effects on cells proliferation and survival factors, the overexpression of S100B protein was also linked to up-regulation of several well-known effectors involved in neo-angiogenesis and invasion of tumour cells, as attested by VEGF and IL-6 increase and AQP4 up-regulation, a member of water channel proteins responsible for tumour cells proliferation and invasion. 33 We also demonstrated that pentamidine is able of inhibiting the S100B-mediated sequestration of wtp53, restoring its 'genome guardian' functions in cancer tissue and promoting apoptosis, guarantee prevention of tumorigenic drift through its delivery by PENtamidine loaded VEsicles (PENVE). Non-tumoral neighbouring cells may promote tumour growth by secreting paracrine signals from the tumour microenvironment. 34,35 Previous results from our group have demonstrated that S100B protein is a key neurotrophic factor involved in colon inflammation and its targeting by pentamidine is accompanied by significant amelioration of colitis in mice. 20 Despite chronic inflammatory conditions are well-established risk factors in colon cancer, there is still limited evidence supporting the role of S100B in the pathophysiology of colon cancer. Our results support that, by fuelling intestinal inflammation, S100B can build a pro-malignant microenvironment that initiates tumour growth, as supported by the evidence that long-standing UC is one of the bestrecognized predisposing factors to colon carcinogenesis. 2,36 In our study, the up-regulation of S100B orchestrates an increase in oxidative stress and nitric oxide (NO) production, through the increased expression of iNOS. It is well established that lipid peroxidationreactive oxygen species (ROS) are among the initiating factors in colon cancer and that pro-inflammatory cytokines, particularly IL-6, have been related to tumour progression during chronic inflammatory injury. 37 In this context, the observation that PENVE significantly reduced the number of tissue-infiltrating macrophages ( Figure S1) suggests that this drug may have a beneficial effect also by reducing inflammatory cytokines from infiltrating immune cells. Interestingly, we have found that the expression profiles of proliferative and pro-apoptotic proteins, following the exogenous administration of S100B, closely resemble those observed in UC specimens, even in the absence of dysplastic modification of the epithelium. In line with previously reported data, we observed that the function of wtp53 and pro-apoptotic Bax protein were significantly inhibited in UC specimens, further supporting the evidence that the impairment of p53 is an early modification in chronic inflammatory conditions. This evidence is also supported by other studies, showing that in colon cancer patients, an increased immunoreactivity for S100B protein is a reliable prognostic factor of recurrence after curative resection. 38 S100B-driven pro-malignant microenvironment may thus precede cancer development, rather than being its consequence.
As stated before, our major finding was that S100B induced inhibition of wtp53 protein was effectively counteracted by the concomitant administration of pentamidine in the PENVE formulation. Pentamidine is a small-molecule allosteric inhibitor of S100B/wtp53 crosstalk, currently tested in phase II clinical trial for malignant melanoma (ClinicalTrials.gov Identifier: NCT00729807).
In line with this trend, Oncozyme started a new clinical study aiming at testing pentamidine treatment in patients with metastatic colon cancer undergoing standard chemotherapy as second-line and/or third-line treatment (ClinicalTrials.gov Identifier: NCT00809796).
However, no results have been published yet and the molecular mechanism(s) responsible for pentamidine anticancer activity remains largely unclear, except for the observation that pentamidine can inhibit oncogenic PRL phosphatases. 39 This study demonstrates that through the direct inhibition of wtp53 function, S100B may contribute to cancer cells' genomic instability, thus increasing the chance to escape to standard anticancer therapies. Although preliminary, our results provide evidence for the first time that pentamidine, by restoring wtp53 function in ex vivo culture of colorectal carcinoma specimens, might represent a novel therapeutic strategy in colon cancer. Hence, targeting the tumoral environment and the S100B/wtp53 crosstalk might represent an effective complementary strategy in counteracting tumour microenvironment (TME) (Scheme 2).
One of the main limitations to pentamidine use in vivo is its unfavourable colonic bioavailability. Since the drug has been originally developed for the treatment of protozoal lung infections, 17 it is indeed only available for aerosol or IV injections. 40,41 The required doses, in order to successfully deliver the compound in colon cancer cells, could, therefore, lead to severe side effects related to its renal and pancreatic toxicity in vivo. 42 In the PENVE niosomal delivery system, the pentamidine is entrapped into a gastro-resistant vehicle, easily degraded by bile salts, pancreatic enzymes and the low pH in the gastrointestinal tract. To improve vesicular membrane integrity, niosomal vesicles were coated with polymers including chitosan, pectin and polyethylenglycol (PEG) that act as a bio-adhesive, biodegradable and hydrophilic polymer, enhancing mucoadhesion with mucosal tissues and promoting interpenetration. 22 In a future perspective, this improved pharmacokinetic profile could also allow for reducing the orally administered doses in humans, ensuring the best therapeutic results and minimizing the risk of side effects. Although indirectly, this study suggests that enteric glia-derived S100B promotes colon carcinogenic drift. However, only future investigations will more in-depth characterize the role played by enteric glia in the pathophysiology of colon cancer. Overcoming the poor colonic bioavailability via this new PENVE formulation, our data also highlight a possible use of pentamidine as a chemotherapy drug, enhancing the efficacy of available tools for colon cancer therapy.

ACK N OWLED G EM ENTS
The assistance of the staff is gratefully appreciated.

CO N FLI C T O F I NTE R E S T
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.