Hyaluronic acid–doxorubicin nanoparticles for targeted treatment of colorectal cancer

Abstract Colorectal cancer, common in both men and women, occurs when tumors form in the linings of the colon. Common treatments of colorectal cancer include surgery, chemotherapy, and radiation therapy; however, many colorectal cancer treatments often damage healthy tissues and cells, inducing severe side effects. Conventional chemotherapeutic agents such as doxorubicin (Dox) can be potentially used for the treatment of colorectal cancer; however, they suffer from limited targeting and lack of selectivity. Here, we report that doxorubicin complexed to hyaluronic acid (HA) (HA‐Dox) exhibits an unusual behavior of high accumulation in the intestines for at least 24 hr when injected intravenously. Intravenous administrations of HA‐Dox effectively preserved the mucosal epithelial intestinal integrity in a chemical induced colon cancer model in mice. Moreover, treatment with HA‐Dox decreased the expression of intestinal apoptotic and inflammatory markers. The results suggest that HA‐Dox could effectively inhibit the development of colorectal cancer in a safe manner, which potentially be used a promising therapeutic option.

Doxorubicin encapsulated in a liposome (Doxil), administrated intravenously, has been used to treat breast cancer, ovarian cancer, Kaposi's sarcoma, and other solid tumors. [69][70][71][72][73] However, administration of Doxil induces low blood counts and increasing risk for anemia. To our knowledge, there have been no reports of free Dox, injected intravenously, treating colorectal cancers.
Hyaluronic acid (HA), a naturally found polymer found in the body, has been shown to be useful in cancer treatment as a therapeutic. HA is able to bind to receptors on the cell membrane of tumor cells, which slows metastasis and cell migration [74][75][76][77] Additionally, HA can enhance the effect of conventional anticancer drugs. In a dose-dependent manner, HA inhibits the cellular proliferation of various cancer cells including prostate, bladder, breast, melanoma, and fibrosarcoma cells. 78 Even in the presence of escape mechanisms associated with cancer progression, the ability of HA to slow cancer growth was unaffected. 78 HA, coupled to a conventional anticancer drug, could reduce the therapeutic dose along with reduced toxicity to healthy cells.
A combination of Dox and HA can be used as treatment.
Recently, a Dox-loaded hyaluronic acid ceramide nanoassembly releasing microspheres, administered intra-arterially, was used to treat tumors. 79 In this study, we assessed whether intravenous injection of Dox, complexed to hyaluronic acid (HA), can target the intestines without any affinity moieties. Our data show that HA-complexed Dox (HA-Dox) exhibits surprisingly high accumulation in the intestine and inhibits the progression of AOM/DSS-induced colon cancer as well as provides a potential therapeutic effect in a safe manner.

| Synthesis
The drug doxorubicin (Dox), was chemically linked to 50 KDa hyaluronic acid (HA) polymer via nucleophilic acyl substitution reactions. Briefly, the polymer was dissolved in a mixture of deionized water and DMSO at a ratio of 1:1 volume. The catalyst 4-Dimethylaminopyridine (DMAP) and the activator 1-ethyl-3-3dimethylaminopropyl) carbodiimide hydrochloride (EDC) were added to the solvent mixture at a molar ratio of 1:1 relative to the HA monomers. Following 30 min of activation, Dox was added into the reaction mixture at molar ratios 0.4:1 relative to the HA monomer and left stirring for 24 hr. Following reaction, the product was purified by size exclusion chromatography via Sephadex G-25 PD-10 desalting columns (5,000 MW exclusion limit) followed by overnight dialysis (3,500 Molecular Weight Cut Off (MWCO)) against DI water. The samples were then lyophilized, stored at 4 C and reconstituted in PBS before use. The amount of Dox incorporated was measured using its respective fluorescence spectra at Ex/Em 470/590.

| Characterization
The size and morphology of HA-Dox were examined by transmission electron microscopy. Briefly, 2.0 μl of HA-Dox suspensions were allowed to air-dry on Formvar carbon-coated cupper grids. Transmission electron microscopy (TEM) was performed on a JEOL JEM-1400 TEM instrument, operating at a voltage of 100 kV (JEOL USA, Inc.).
Particle zeta potential was measured by dynamic light scattering (DLS) on Malvern Zetasizer (Malvern). The mean particle size of HA-Dox was estimated with a NanoSight NS3000 (Malvern Panalytical Inc., Westborough, MA).

| In vitro hydrolysis assay
Lyophilized HA-Dox was dissolved in PBS at 1 mg/mL DOX concentration (pH 7.4) and incubated for 5 days at 37 C in Slide-A-Lyzer MINI dialysis devices (10,000 MWCO). The devices were inserted in microcentrifuge tubes holding 1 ml of PBS. The 100 μl of Dox-release medium in the microcentrifuge tubes was collected at the indicated time points and its concentration was measured via fluorescence using the TECAN plate reader. The amount of Dox released at each time point was divided with the initial amount loaded to obtain the cumulative release. All measurements were carried out in triplicate, and the results were indicated as the mean ± SD.

| Ethics statement
All animal studies were carried out in strict accordance with the Guide for the Care and Use of Laboratory Animals as adopted by National Institute of Health, approved by Harvard University IACUC. Mice were housed in cages with free access to water and food, located in a wellventilated temperature-controlled room between 18 and 23 C with relative humidity ranging from 40 to 60% under a 12-hour light/dark period.

| Serum and RBC collection
Healthy female balb/c mice (50-56 days) were purchased from the Charles River Laboratory (Wilmington, MA). Whole blood, collected in EDTA coated tubes (BD Microtainer) or serum collecting tubes (BD Microtainer), was spun at 1,000 g for 10 min at 4 C. Plasma and well as the buffy layer containing white blood cells and platelets, was removed; serum was stored at 4 C for 1 hr. Isolated erythrocytes (RBCs) were washed by adding ice cold 1x Dulbecco's-phosphatebuffered-saline (DPBS) pH 7.4 up to 12 ml total volume and pipetting gently up and down to mix RBC extensively. RBC suspensions were centrifuged at 600g, 15 min, 4 C; supernatant was removed and this wash step was repeated three times.  Intestines were harvested and far red fluorescence signals were imaged using Perkin Elmer IVIS small animal imaging system.

| In vitro hydrolysis
In vitro release rates of HA-Dox indicated a slow and steady release of the drug from HA. Approximately, 20.6 ± 0.8 wt% Dox was released at the end of 120 hr (Figure 2). The in vitro release profile is similar to the pattern obtained previously attributing the successful incorporation of Dox and its slow release via amide hydrolysis. 80 However, the slow release does not affect the therapeutic efficacy of HA-Dox compared to the free Dox.

| Pharmacokinetics of HA-Dox
Pharmacokinetics of HA-Dox was evaluated in healthy Balb/c mice  F I G U R E 3 Pharmacokinetics of HA-Dox: Pharmacokinetics both free Dox and HA-Dox represented as %ID in blood at 0.08 hr, 0.5 hr, 6 hr, and 24 hr. Each data point represent means ± SEM (n = 3).* p < .05; non paired two-tailed t-test The mechanisms of enhanced persistence of HA-Dox in blood were studied. Hyaluronic acid by itself has a blood half-life of 3-5 min in the blood. Hence, simple association of dox with HA is not expected to render long circulation. We assessed whether association of dox-HA with red blood cells (RBCs) may be responsible for extended circulation.
Prior literature has shown that RBC-association enhances the circulation of polymeric nanoparticles. [81][82][83] HA-Dox bound to murine RBC with high efficiency; nearly 100% of RBC exhibited attachment HA-Dox on the surface in the presence of 55% serum (Figure 4a,b).
Adsorption of HA-Dox onto RBCs did not induce agglutination (Figure 4c, inset), implying that HA-Dox may not be detrimental to RBC. Furthermore, we investigated whether HA-Dox induces exposure to phosphatidylserine, a signal released from senesced and damaged RBC that facilitates their clearance from the blood circulation.

| Biodistribution of HA-Dox
Biodistribution of Dox and HA-Dox in mice was assessed. Significant accumulation of HA-Dox was found in intestine (45% ID), while only 7% ID was found in the liver 0.5 hr after administration (Figure 5a).
Minimal amounts were detected in the heart, lung, spleen, kidney, brain, stomach, pancreas, and the spleen. Free Dox accumulated in the liver (8% ID) 0.5 hr after administration and 30% ID of free Dox accumulated in the intestine. 5% ID of free Dox remained in circulation. There was a significant increase in HA-Dox accumulation in the intestine compared to its free Dox counterpart (45% ID vs 30% ID respectively). The amount of HA-Dox in the intestine persisted at 45% ID for 24 hr. Similar trend was found for free Dox except that the magnitude of accumulation was at 15% ID. (Figure 5b). There was nearly a threefold increase in the intestine:blood ratio of HA-Dox compared to free Dox.
Within the intestines, most HA-Dox as well as free Dox was found to accumulate in the duodenum and jejunum after 0.08 hr.
Unlike free Dox, a significant accumulation of HA-Dox was also observed in the ileum, colon, and cecum (Figure 6a). Compared to 0.08 hr, the amount of HA-Dox in the duodenum and jejunum remained fairly constant after 0.5 hr (Figure 6b Furthermore, HA-Dox accumulation (displayed as bright green dots) in each intestinal tissue, 0.05 hr and 6 hr after intravenous administration, was also confirmed using confocal microscopy ( Figure S3).
To assess the role of HA in targeting of HA-Dox in the intestine, localization of AlexaFluro647-labeled HA was measured. Signal intensities were observed in the stomach 0.08 hr after administration. Some signal was observed in all parts of the intestine (duodenum, jejunum, ileum, and the colon), suggesting that HA accumulates in the intestines ( Figure S4a).
Although at a reduced level, signal could still be observed in the duodenum, colon, and different parts of the jejunum and ileum 0.5 hr postinjection ( Figure S4b). Similar to HA-AlexaFluro647, strong signal was observed in the stomach for HA-Dox-AlexaFluro647 and low signal were

| The effect of Dox and HA-Dox on intestinal inflammation in mice with chemical-induced colon cancer
The effect of HA-Dox on the inflammation in AOM/DSS mice was also assessed in terms of (cyclooxygenase-2) COX-2 and inducible nitric oxide synthase (iNOS) expression levels by immunohistochemistry ( Figure S6   In addition, there have been studies focusing on the inter-and intra-cellular interactions between hepatic cells and nanoparticles as well as the elimination of nanoparticles through the hepatobiliary system. 95 In our study, HA-Dox was found in the liver. Since no HA-Dox There is evidence that a link between inflammation and colorectal cancer exists. [97][98][99][100][101][102][103][104][105][106][107] Many studies have shown that pro-inflammatory mediators such as cyclooxygenase-2 (COX-2) and lipoxygenase pathways may lead to tumor cell proliferation, growth, thus promoting colorectal cancers. Anti-inflammatory agents such as COX2 inhibitors, as well as iNOS inhibitors, suppress colorectal cancer by inhibiting inflammatory pathways. 108-124 COX-2 inhibitors (e.g., rofexoxib, celecoxib, and valdecoxib), subclass of nonsteroidal anti-inflammatory drugs, reduce the production of prostaglandins, chemical that promotes inflammation. 114,125,126 By providing anti-inflammatory benefits, it allows the COX-1 enzyme to retain its gastroprotectivity functions. Our findings suggest that HA-Dox functions similarly to COX-2 inhibitors in this regard.
There are many reports that show that chemotherapy causes extensive damage to the DNA present in the intestinal cell wall.
Researchers have been looking into reducing the damage done to the intestinal cell walls from chemotherapy, as it would render the treatment more bearable and allow for a higher rate of implementation of chemotherapy. [127][128][129][130][131][132][133][134][135][136] Considering this, in our study, it is unclear whether HA-Dox induces apoptosis in the intestine due to the continuous shedding of the intestinal epithelium. As this is the most rapidly renewing tissue in the body, undergoing almost complete cellular turnover in as little as a few days, it is difficult to distinguish between normal cell death and apoptosis that is symptomatic of intestinal cancer. In IBD, excessive cell death and apoptosis is observed in the colon and ileum epithelium. The development of IBD-related colorectal cancer (CRC), chronic inflammation is a major risk factor for gastrointestinal malignancies development in IBD patients. There has been evidence over the decade that show a link between chronic intestinal inflammation and CRC thorough a series of events; from the development of early dysplasia to low grade dysplasia to high-grade dysplasia to eventually converting to invasive adenocarcinoma. [137][138][139][140] The development of extra-intestinal malignancies has also been shown in IBD patients. Currently, therapies to treat IBD diminish the mucosal inflammatory response. 138 Our findings suggest HA-Dox lowers inflammation levels in all parts of the intestine, stopping the further development of CRC. Future studies should focus on assessing toxicity of HA-Dox in detail including cardiac toxicity. Future studies should also include a detailed evaluation of the mechanisms of antiinflammatory properties of HA-Dox.
In conclusion, our study revealed that HA-Dox may be an effective therapeutic agent to prevent or reduce the risk of colorectal cancer development, particularly for people who already suffer from IBD.