Intraperitoneal administration of nanoparticles containing tocopheryl succinate prevents peritoneal dissemination

Abstract Intraperitoneal administration of anticancer nanoparticles is a rational strategy for preventing peritoneal dissemination of colon cancer due to the prolonged retention of nanoparticles in the abdominal cavity. However, instability of nanoparticles in body fluids causes inefficient retention, reducing its anticancer effects. We have previously developed anticancer nanoparticles containing tocopheryl succinate, which showed high in vivo stability and multifunctional anticancer effects. In the present study, we have demonstrated that peritoneal dissemination derived from colon cancer was prevented by intraperitoneal administration of tocopheryl succinate nanoparticles. The biodistribution of tocopheryl succinate nanoparticles was evaluated using inductively coupled plasma mass spectroscopy and imaging analysis in mice administered quantum dot encapsulated tocopheryl succinate nanoparticles. Intraperitoneal administration of tocopheryl succinate nanoparticles showed longer retention in the abdominal cavity than by its intravenous (i.v.) administration. Moreover, due to effective biodistribution, tumor growth was prevented by intraperitoneal administration of tocopheryl succinate nanoparticles. Furthermore, the anticancer effect was attributed to the inhibition of cancer cell proliferation and improvement of the intraperitoneal microenvironment, such as decrease in the levels of vascular endothelial growth factor A, interleukin 10, and M2‐like phenotype of tumor‐associated macrophages. Collectively, intraperitoneal administration of tocopheryl succinate nanoparticles is expected to have multifaceted antitumor effects against colon cancer with peritoneal dissemination.


| INTRODUC TI ON
During the development of rational cancer therapy, nanoparticles such as liposomes, micelles, and lipid nanoparticles have been used for the delivery of hydrophobic and hydrophilic drugs and nucleic acids. [1][2][3] Tumor site-specific delivery along with improvement in solubility and avoidance of enzymatic degradation of various drugs can be achieved through nanoparticle formulations. [4][5][6] When nanoparticles are intravenously (i.v.) administered, they are passively delivered to tumor tissues via the enhanced permeability and retention (EPR) effect. 7 However, this delivery strategy using EPR effect requires the high circulation of nanoparticles in the blood and a specific physiological environment within the tumor, particularly hypervascularity. 8 Therefore, it is difficult to deliver nanoparticles to pancreatic cancer or peritoneal dissemination with hypovascularity by i.v. administration, even if the nanoparticles show excellent blood circulation. [9][10][11] Peritoneal dissemination is a major cause of death in abdominal cancers such as colon and gastric cancers rather than the peritoneal cancer itself. 12,13 During cancer progression, the cells disseminate into the peritoneal cavity and then metastasize to the peritoneum. 12,13 Therefore, chemotherapy is needed to prevent further spread of cancer cells in the peritoneum and peritoneal cavity, in addition to killing cancer cells directly. Alternatively, extensive formation of ascites occurs in an abnormal peritoneum, leading to decrease in the quality of life of patients. 14 In the potential mechanism underlying the formation of ascites it is considered that the leaky structure of tumor vessels alters the direction of oncotic pressure, leading to the flow of ascites into the peritoneal cavity. 15,16 A rational strategy to prevent the formation of ascites is to stabilize the vascular structure in addition to killing cancer cells.
Systemic administration of anticancer drugs via the i.v. route for the treatment of peritoneal dissemination is inefficient for achieving therapeutic efficacy due to the inferior transitivity of drugs into the peritoneum and peritoneal cavity. 17,18 To improve drug transitivity, intraperitoneal (i.p.) administration of anticancer drugs has a focus of study. 17,18 Anticancer drugs administered via the i.p. route can directly attack cancer cells in the peritoneum and peritoneal cavity at a high local concentration. 19,20 Furthermore, the i.p. retention time of drugs is affected by the size of the drug. Large-sized drug formulations such as anticancer nanoparticles remain in the peritoneal cavity for a long duration, while small-sized drug formulations enter the systemic circulation relatively rapidly. [21][22][23] Therefore, the i.p. administration of anticancer nanoparticles can effectively deliver the drugs into the peritoneal cavity, achieving a long retention time. The surface charge and size of the nanoparticles affect the i.p. retention time. 24,25 The i.p. administration of cationic nanoparticles demonstrated a longer i.p. retention time than that of anionic nanoparticles because cationic nanoparticles bind electrostatically to the negatively charged peritoneum. 26 However, cationic nanoparticles strongly interact with biogenic substances, and this affects the distribution of the nanoparticles in the peritoneal cavity, depending on their modified physicochemical properties. 27 Furthermore, cationic nanoparticles show potent cytotoxicity against not only cancer cells but also healthy cells. 28 Such undesirable effects hamper the use of cationic nanoparticles for the treatment of peritoneal dissemination.
Compared with cationic nanoparticles, anionic nanoparticles show high biocompatibility without serious cytotoxicity due to less interaction with biogenic substances. 29 Moreover, anionic nanoparticles show high stability without any change in assembly and particle size even in biofluids such as ascites and blood 27 ; therefore they are expected to maintain their size while in the disease environment of the peritoneal cavity. Therefore, we hypothesized that i.p. administration of anionic nanoparticles containing anticancer agents that act specifically on cancer cells would be an effective therapy for peritoneal dissemination without side effects.
α-Tocopheryl succinate (TS), a succinic acid ester of α-tocopherol (α-T), is a redox-silent analog of α-T. 30 Although physiological actions of α-tocopherol and its analogs are based on their antioxidative effects, 30 TS exerts multifaceted anticancer effects such as the induction of cancer cell-specific apoptosis and inhibition of tumor angiogenesis and multidrug-resistant protein. 30 Because esterase activity of healthy cells is higher than that of cancer cells, TS is easily hydrolyzed to α-T, a silent agent with anticancer effects, leading to mitigation of undesirable side effects. 30 Alternatively, TS with both hydrophilic and hydrophobic moieties is easily vesiculated, while nanoparticles consisting only of TS undergo structural changes and disassemble in the presence of divalent cations and serum in vivo. 31 We have previously developed nanoparticles (TS-NP) containing TS and egg phosphatidylcholine (EPC) to improve particle stability in vivo. 31 TS-NP exhibited anionic surface charges and high particle stability in vivo, 31 which is an advantageous physicochemical property for the treatment of peritoneal dissemination using i.p. administration of anticancer nanoparticles. When TS-NP were administered into tumor-bearing mice via the i.v. route, they existed as particles and were taken up by cancer cells at the tumor site, and they showed more potent anticancer effects than by TS itself. 31 However, it was unclear whether TS-NP showed long retention in the peritoneal cavity and exerts potential antitumor effects when administered via the i.p. route in a mouse model of peritoneal dissemination.
In the present study, we compared the retention time of TS-NP in the peritoneal cavity after its i.v. and i.p. administration; moreover,

| Preparation of nanoparticles containing TS (TS-NP and TS-NP-Qd)
TS-NP were prepared using a simple hydration method described in our previous study. 31 Briefly, to form a thin film, TS (50 mM) and EPC (32 mM) dissolved in ethanol were mixed in a roundbottomed glass tube at a molar ratio of 5:3.2 and then dried using nitrogen gas. The thin film was hydrated with PBS (−) containing 40 mM NaOH, and then nanoparticles were formed by sonication for 20 min in a bath-type sonicator (Yamato Scientific Co., Ltd., Tokyo, Japan). To encapsulate Qd into the TS-NP, a cationic core was prepared using surface modification of carboxyl-Qd using STR-R8. The anionic thin film consisting of TS and EPC was hydrated using this cationic core solution, followed by sonication for 20 min in a bath-type sonicator (Yamato Scientific Co., Ltd.). Corp. Oosaka, Japan). The particle size and surface charge of the nanoparticles were determined using Zetasizer Nano (Malvern Instruments Ltd, Worcestershire, UK).

| Stability of TS-NP in the presence of ascites
To obtain ascites, colon26 cells (2 × 10 5 ) suspended in PBS (−) were intraperitoneally (i.p.) administered to BALB/cCrSlc mice, followed by the collection of ascites at 2 weeks after the injection. The ascites sample at 4× dilution with PBS was incubated with TS-NP at room temperature for 0, 1, or 24 h. The change in particle size and surface charge indicated the stability of the TS-NP in the presence of ascites.

| Biodistribution of TS-NP-Qd in the mouse model of peritoneal dissemination by an in vivo imaging system (IVIS) and inductively coupled plasma mass spectrometry (ICP-MS)
The mouse model of peritoneal dissemination was prepared by i.p.

| Evaluation of tumor growth in mice using luminescence images
Colon26-Luc cells were injected i.p. into BALB/cCrSlc mice as

| Determination of cell viability using WST-1 assay
A cell viability assay was performed as previously described. 31 Colon26 cells were seeded on 96-well CellBIND plates (Corning) at a density of 5 × 10 3 cells/well. After incubation at 37°C for 24 h, the cells were treated with TS-NP or TS dissolved in ethanol (TS solution) for 48 h. Cell viability was determined using the WST-1 assay. Cell viability was estimated by dividing the absorbance (at 440 nm) of the sample by that of the nontreated group.

| Evaluation of the cellular uptake and intracellular trafficking of TS-NPs
The cellular uptake of nanoparticles was determined using flow cytometry, as previously described. 31 Colon26 cells were treated with EPC/DOTAP (7:3) liposomes and TS-NPs containing 1 mol% Rh-PE for 2 h. After washing twice with PBS (−) containing 2% FBS, the resuspended cells were subjected to flow cytometry analysis. The intracellular trafficking of TS-NPs was evaluated using confocal laser scanning microscopy, as previously described. 31 The cells were treated with Rh-PE-labeled TS-NP for 2 h. The endosomes/lysosomes and nuclei were stained with LysoTracker Green DND-26 and Hoechst 33342, respectively. The intracellular distribution of TS-NPs was observed using a Nikon AX confocal laser scanning microscope (Nikon Corporation, Tokyo, Japan).

| Determination of the type of TAM in ascites using flow cytometry
The ascites samples obtained (as described in Section 2.6.) were mixed with anti-CD163 antibody at 4°C for 1 h. After washing with PBS (−) containing 2% FBS, the samples were incubated with Alexa488-labeled anti-rabbit antibody at 4°C for 1 h. After washing with PBS (−) containing 2% FBS, CD163-positive macrophages were determined using flow cytometry.

| Statistical analysis
Statistical significance was determined using Student's t test or oneway ANOVA, followed by Tukey's honest significant difference test.
Statistical significance was set at p < 0.05.

| Effect of TS-NP on colon cancer cell proliferation
The particle size and ζ-potential of TS-NP prepared in this study were estimated at ~120 nm and −40 mV, respectively ( Table 1).
It had been previously reported that TS-NP induced potent cell death in B16-F1 cells, a mouse melanoma cell line, due to their homogenous cellular uptake attributed to their high particle stability. 31 To determine the potential application of TS-NP in the treatment of peritoneal dissemination derived from colon cancer cells, we determined the effect of TS-NP on the growth of colon26 cells. As shown in Figure 1B In general, the cellular uptake of anionic nanoparticles is ineffective. 35 Therefore, we examined the cellular uptake of rhodaminelabeled TS-NPs in colon26 cells. As shown in Figure 1C, flow cytometric analysis revealed that the rhodamine-labeled TS-NPs were taken up by the colon26 cells, although the surface charge was negative. Furthermore, we examined the intracellular trafficking of rhodamine-labeled TS-NPs using confocal laser scanning microscopy. As shown in Figure 1D effective intratumoral distribution mediated using the EPR effect of negatively charged TS-NP. 31 However, the stability of TS-NP in the abdominal cavity is unclear. Therefore, the physicochemical properties of TS-NP were evaluated in the presence of ascites. As shown in Figure 2A,C, the particle size distribution and ζ-potential of TS-NP in PBS were unchanged even after incubation for 24 h. The particle size in ascites showed a similar distribution to that in PBS, while smaller particles with size below 50 nm were observed ( Figure 2B).
As shown in Figure 2C, TS-NP has potent anionic charges even when incubated in ascites at 37°C for 24 h, although ζ-potential was slightly attenuated from −30 to −20 mV after incubation. These results suggested the high stability of TS-NP in the presence of ascites.

| Biodistribution of i.p. and i.v. administration of TS-NP in mice
When nanoparticles are administered i.p., the biodistribution depends on their particle size. [21][22][23] To evaluate the biodistribution of TS-NP with a constant particle size, TS-NP-encapsulating carboxyl-Qd was used in this study. To encapsulate negatively charged carboxyl-Qd in TS-NP, carboxyl-Qd was modified with positively charged STR-R8 (R8-Q d) ( Figure 3A). The particle size and ζ-potential of R8-Qd were observed as 55 nm and +20 mV, respectively ( Table 2). These positively charged R8-Qd were embedded into the negatively charged membranes of the TS-NP via electrostatic interaction (TS-NP-Qd) ( Figure 3A). As shown in Table 2, the particle size of TS-NP-Qd was enlarged, and ζ-potential shifted from positive to negative compared with R8-Qd. These results confirmed the successful incorporation of carboxyl-Qd in TS-NP. The encapsulation efficiency of the Qds into the TS-NPs was 63 ± 21%.
The physicochemical properties of TS-NP-Qd in ascites were comparable with those in PBS, and were similar to TS-NP without cationic Qd cores, suggesting that TS-NP-Qd was also stable in ascites ( Figure S1). Furthermore, using these TS-NP-Qd, the biodistribution of i.p. and i.v. administration of TS-NP was compared. TS-NP-Qd injected i.p. were observed in the peritoneal cavity after 1 h of administration, whereas TS-NP-Qd injected i.v. were not ( Figure 3B).
When carboxyl-Qd was administered, weak signals were observed only for i.p. injection ( Figure 3B). The retention time of TS-NP-Qd in the abdominal cavity was comparable with that of carboxyl-Qd, which was used as a control to evaluate the biodistribution of NPs with a constant particle size ( Figure 3C). As shown in Figure 3C,

| DISCUSS ION
In the present study, we evaluated the antitumor effects and biodistribution of i.p. administered TS-NP in a mouse model of peritoneal dissemination.
As shown in Figure 1B As shown in Figure 2, TS-NP showed high stability, even in the presence of ascites. Albumin is a major protein found in mouse peritoneal fluid, human ascites from patients with peritoneal carcinomatosis, and human serum from healthy donors. 27,40 It has been considered that albumin, a negatively charged protein, barely interacts with negatively charged TS-NP, which was supported by unchanged particle size of TS-NP in the presence of 50% FBS. 31 The stability of nanoparticles in body fluids depends on their surface charge. 27 In mice peritoneal fluid and human ascites, anionic nanoparticles showed high stability, whereas cationic nanoparticles strongly interacted with negatively charged proteins, such as albumin, due to their electrostatic interaction. 27 Therefore, negative surface charges on TS-NP contributed to their high stability in the presence of ascites.
As shown in Figure 3 the growth of colon cancer cells in the peritoneal cavity and peritoneum. As shown in Figure 3B-E, TS-NP showed efficient retention in the i.p. cavity that contributed to the potent antitumor effect.
We previously reported that the antitumor effect of TS was not observed in half of the melanoma-bearing mice population receiving i.p. administration of TS solution. 42 In tumor-bearing mice, antitumor effects were induced when TS molecules were translocated into the systemic blood circulation passing through the peritoneum.
Conversely, the assembled TS molecules remained in the peritoneal cavity and were not effective. Alternatively, TS in the peritoneal cavity was not efficiently hydrolyzed to α-T due to low esterase activity in ascites. 17 Therefore, the larger size of the TS-NP compared with that of the TS molecules led to effective i.p. retention, followed by induction of antitumor effects.
As shown in Figure 5A,B, i.p. administration of TS-NP inhibited ascites production and reduced VEGF-A levels. It is known that enhanced vascular permeability by angiogenesis-related factors, such as VEGF-A and angiopoietin-2 (Ang2), is one of the causes of ascites retention. 43,44 TS suppresses VEGF-A and Ang2. 44 Ang2 acts on Tie2, an angiopoietin receptor, to destabilize the vascular structure by attenuating the interaction between endothelial cells and pericytes, followed by enhanced vascular permeability. 45 Although further studies are needed, stabilization of vascular structure through suppression of Ang2 expression followed by normalization of vascular permeability is a possible mechanism underlying inhibition of ascites production by TS-NP.
In the ascites of mice administered TS-NP i.p., the M2-like phenotypes of TAM and IL-10 decreased ( Figure 5C, D). It has been reported that the M2-like polarization of TAM is mediated by IL- In summary, i.p. administered TS-NP showed potent retention in the peritoneal cavity. Consistent with the effective biodistribution, TS-NP inhibited tumor growth and improved the i.p. microenvironment by inhibiting ascites and VEGF-A production and decreasing the M2-like phenotype of TAM. Therefore, i.p. administration of TS-NP, which has multifaceted antitumor effects, is expected as a rational therapy for peritoneal dissemination.

ACK N OWLED G M ENTS
This work was supported in part by JSPS KAKENHI (Grant Number 18H03540). We would like to thank Editage (www.edita ge.com) for English language editing and Prof. Hiroyuki Yasui and Dr. Yuki Naito for their technical support during this study.

CO N FLI C T O F I NTE R E S T
The author declares no conflict of interest.