Hyaluronic acid (MW = 1.0 × 10⁵ Da) was purchased from Freda Biochem Co., Ltd (Shandong, China). Doxorubicin hydrochloride (DOX HCl) was purchased from Dalian Meilun Biotechnology Co., Ltd. (Purity 98%, Dalian, China). CDDP was purchased from Shandong Boyuan Pharmaceutical Co., Ltd. (Purity 99%, Jinan, China). Poly (ADP-ribose) polymerase (PARP) and survivin antibodies were purchased from Abcam (Cambridge, UK). Purified deionized water was prepared using a Milli-Q plus system (Millipore Co., Billerica, MA, USA).

Murine CD44⁺ mammary carcinoma cells (4T1) and normal CD44^- fibroblast cells (NIH-3T3) were purchased from American Type Culture Collection (ATCC, Rockefeller, Maryland, USA), and cultured at 37°C in DMEM supplemented with 10% FBS, penicillin (50 U/ml) and streptomycin (50 U/ml).

Female BALB/c mice (4–5 weeks old) were purchased from the Experimental Animal Center of Jilin University, and housed in a standard environment with access to normal chow diet until they reached a weight of 18–20 g. A tumor-bearing model was generated by inoculating 1.5 × 10⁶ 4T1 cells into the abdominal mammary glands of female BALB/c mice. All animal procedures were designed to minimize animal suffering and were approved by the Institutional Animal Care and Use Committee of Jilin University.

Hyaluronic acid-DOX-CDDP nanoparticles were prepared according to previous studies (Zhang et al., 2018; Zhao et al., 2018; Wang et al., 2019). Hyaluronic acid and phosphate buffer (pH=7.4, 1:9 volume ratio) were blended in deionized water until the HA was completely dissolved. Following dissolution, the pH was adjusted to 7.0 using 0.05 M NaOH. An aqueous DOX solution (6.0mg/ml) was added dropwise into the polymer solution, then stirred overnight in the dark at room temperature. Next, a predetermined amount of CDDP was dissolved and added into the reaction mixture mentioned above, and then, incubated at 37°C for 72 h. Finally, HA-DOX-CDDP was obtained using a dialysis method with deionized water as the releasing medium for 10 h.

Dynamic light scanning (DLS) (Wyatt Technology, Santa Barbara, CA, USA) and transmission electron microscopy (TEM) (JEM-1011, JEOL, Japan) were performed to determine HA-DOX-CDDP nanoparticle size and morphology, respectively. Five milliliters of drug-loaded nanoparticles (0.1 mg/ml) were added to a clean glass container and dissolved prior to DLS detection. Ten microliters of drug-loaded nanoparticles (at the same concentration of 0.1 mg/ml) were added to a copper wire and completely dried for TEM analysis.

To determine the release characteristics of HA-DOX-CDDP in PBS (pH 5.5, pH 6.8, and pH 7.4), 1.0 mg of freeze-dried HA-DOX-CDDP micelles was used to prepare release diluent (100.0 mg/ml). The micelles were sealed with 10.0 ml of diluted solution in a dialysis bag (MW cut-off 3,500 Da) and placed in 100.0 ml of release medium, and the solution containing the dialysis bag was shaken at 70 rpm at 37°C. At 0.5, 1, 2, 4, 8, 12, 24, and 48 h, 2.0 ml of release diluent was removed and replaced with fresh diluent. The amounts of released DOX and CDDP were determined using a fluorescence spectrophotometer (UV-1800, Shimadzu, Japan) (λex = 480 nm and λem = 590 nm) and an inductively coupled plasmae mass spctrometer (ICP-MS, Xseries II, Thermoscientific, USA), respectively.

Cellular uptake and intracellular DOX release of HA-DOX-CDDP micelles, free DOX, and free CDDP in 4T1 and 3T3 cells were assessed using flow cytometry (FCM) (Beckman, California, USA). Cells were seeded in 6-well plates at a density of 2×10⁵ cells per well, then incubated for 12 h at 37°C. The original culture medium was replaced with medium containing free DOX+CDDP or HA-DOX-CDDP at a final DOX HCl concentration of 10.0μmg/ml. In the HA-pretreated groups, cells were incubated with HA micelles for 1 h prior to treatment. Following pretreatment, the medium was replaced with medium containing HA-DOX-CDDP at the same concentration. After 6 h of incubation in a thermostatic incubator, the cells were harvested, washed in cold PBS, and centrifuged twice at 4°C for 5 min at 1,500 rpm. Finally, 500.0 μl of cell suspension was subjected to FCM analysis.

The cellular localization of DOX in 4T1 and 3T3 cells was determined using an LSM 780 confocal laser scanning microscope (CLSM) (Carl Zeiss, Jena, Germany). Cells (1.5 × 10⁵) were incubated on cover slips in 6-well plates for 12 h, and the medium was replaced with medium containing HA-DOX-CDDP micelles, or free DOX+CDDP in DMEM at a final DOX HCl concentration of 10.0 μg/ml. Hyaluronic acid pretreatment groups were pretreated with HA, and then incubated with HA-DOX-CDDP. After incubation for 6 h, cells were fixed in 4% (W/V) PBS-buffered formaldehyde for 30 min at 15–20°C. The cells were then stained with DAPI (blue) and Alexa 488 (green) at 37°C to visualize the cell nucleus and cytoskeleton, respectively.

The cytotoxicity of HA-DOX-CDDP was evaluated using the MTT assay. Briefly, 7 × 10³ 4T1 or NIH-3T3 cells per well were seeded in 96-well plates and incubated for 12 h. Two hundred microliters of culture medium was replaced with fresh medium containing free DOX+CDDP, HA-DOX-CDDP, or HA micelles at various concentrations, and each incubated for 24 or 48 h. Then, 20.0 μl of MTT (5.0 mg/ml) was added to each well, and the cells were incubated for 4 h at 37°C. Next, the medium was replaced with 150.0 μl of DMSO to dissolve the formazan crystals, and the cells were shaken for 10 min prior to analysis. Absorbance was measured at 490 nm using a microplate reader (Bio-Rad 680, Hercules, CA, USA). Cell viability was calculated using equation 1:

In Eq. 1, A_sample and A_control represent the absorbances of sample and control wells, respectively.

A 3D cell suspension cultures model was established using HDP 1096 Perfecta 3D^®96-well hanging drop plates (3D Biomatrix, USA) as described in a previous study (Xiao et al., 2018). Briefly, suspensions of 4T1 and NIH-3T3 cells were added into agarose solution preprocessed 96-well plates at 2.5 × 10⁴ cells per well, and the cells were incubated at 37°C. After 5 days, multicellular spheroids were randomly divided into HA (control), free DOX+CDDP, and HA-DOX-CDDP groups with a final DOX HCl concentration of 10.0 μg/ml. Morphologic changes of the multicellular spheroids were visualized using CLSM after treatment for 24 h.

The biodistribution of DOX in major internal organs and tumor tissues following intravenous injections was assessed using fluorescence imaging in vivo. When tumors grew to 60–80 mm³, the 4T1 tumor-bearing mice were injected with 0.1% normal saline (NS) as a control, HA-DOX-CDDP micelles, or DOX+CDDP via the lateral tail vein, at a DOX HCl dose of 5.0 mg/kg. Major organs (heart, liver, spleen, lung, and kidney) and tumors were excised 6 or 12 h post-injection, washed with NS three times, and analyzed for DOX-related fluorescence using an in vivo imaging system (Maestro 500 FL, Cambridge Research & Instrumentation, Inc., USA). A 150 W halogen lamp and a 450–500 nm excitation filter were used for DOX fluorescence analysis. Averaged signals with autofluorescence excluded were quantitatively analyzed using Maestro™2.4 software. Tumor volume (V) was calculated as 0.5 × length × width × height.

In vivo antitumor efficacies of free drugs and HA-DOX-CDDP micelle were evaluated using the 4T1-xenografted Balb/c mice. Similarly to the detections of tissue distributions, 0.1 ml of cell suspension containing 1.5 × 10⁶ 4T1 cells in NS was injected subcutaneously into the abdominal mammary gland of 4–5 weeks old mice weighing of 18–20 g. When the tumor volume increased to about 60–100 mm³, the mice were treated with NS, combination of free DOX and CDDP, or HA-DOX-CDDP micelle at a DOX concentration of 5.0 mg/kg⁻¹ body weight by the tail-veil injections on 0, 5, 10, 15, and 20 days. In the course of treatment, tumor volumes and body weights were monitored every other day. Tumor volume (V) was calculated as 0.5 × length × width × height.

Tumor-bearing mice were randomly divided into three groups (n=5) that received NS, DOX + CDDP, or HA-DOX-CDDP micelles via injection. The mice were sacrificed 8 days after the final injection. Mouse body and tumor weights were recorded at 1–2-day intervals. Tumors, major organs, and lymph nodes were washed with NS, fixed in 4% paraformaldehyde, and embedded in paraffin. Tissues in paraffin blocks were cut into ~5-μm slices for hematoxylin and eosin (H&E) staining and ~3-μm slices for immunofluorescence (PARP and survivin) analyses. Alterations in pathological histology and fluorescence intensity were detected using a Nikon microscope (Eclipse Ti, Optical Apparatus Co., Ardmore, USA) and CLSM, respectively. All image data collected were analyzed using ImageJ software (National Institutes of Health, Bethesda, Maryland).

All experiments were performed in triplicate and results were analyzed for statistical significance using IBM SPSS 17.0 (IBM Company, Armonk, NY, USA). Differences between the groups were estimated using one-way ANOVA or t-tests, and reported as means ± standard deviations (SD). Two-tailed P values <0.05 were considered to indicate statistically significant results. When P values <0.01 or <0.001, results were considered to have statistically obvious difference.

As shown in Figure 1, HA-DOX and HA-DOX-CDDP were spherical (Figures 1A, B). The diameters from TEM microimages were about 100 nm. The hydrodynamic radius (Rh) was about 80 nm. The relative smaller size from TEM detection should be attributed shrinking of nanoparticles during preparing the samples for TEM detection (Figure 1C). The FT-IR spectra of HA, HA-DOX, and HA-DOX-CDDP are shown in in Figure 1D. The absorption peaks of HA-DOX and HA-DOX-CDDP in the FT-IR spectra were 1300–1000 cm^-1 duo to the stretching vibration of C-O bond in DOX, but in the spectra of HA-DOX-CDDP, the characteristic absorption peak at 3300–3200 cm^-1 was also present duo to the amines (NH2) in CDDP. The waveforms of DOX-loaded HA micelles were nearly identical, and slightly different from those of HA, which indicated that the three types of micelles had similar chemical construction.

To determine the release characteristics of HA-DOX-CDDP under physiological conditions, and in the intracellular microenvironment, in vitro drug release of DOX and CDDP from micelles was evaluated at pH 5.5, 6.8, and 7.4 (Figures 1E, F). The micelles exhibited a burst release pattern during the first 4–6 h, and then showed sustained release at different acidic conditions. Doxorubicin and CDDP release from cross-linked micelles was less than 35% at pH 7.4. However, as pH decreased to 6.8 or 5.5, DOX and CDDP release from cross-linked micelles increased to approximately 50% and 80%, respectively, after 72 h.

Flow cytometry and CLSM were used to investigate the internalization profiles of HA-DOX-CDDP micelles in 3T3 and 4T1 cells. After incubation with free drugs (DOX+CDDP) or HA-DOX-CDDP for 6 h, the highest levels of free drug internalization were observed in 3T3 cells. In CD44-positive 4T1 cells, HA-DOX-CDDP-treated groups exhibited significantly higher internalization than groups treated with free drugs or pretreated with HA (Figure 2A). In addition, the fluorescence intensity induced by HA-pretreatment was almost similar to that induced by HA-DOX-CDDP treatment in 3T3 cells, but less than that following treatment with free drugs in both 3T3 and 4T1 cells. The results obtained from FCM analysis were consistent with those obtained using CLSM (Figure 2B).

Inhibition of cell proliferation in response to DOX-loaded HA micelles was evaluated using the MTT assay in 3T3 and 4T1 cells following 24 or 48 h of exposure (Figure 3). Groups treated with the free drugs (DOX+CDDP) and HA-DOX-CDDP micelles both exhibited time- and dose-dependent cytotoxicity. However, treatment with HA groups induced slightly smaller effects on cell viability, indicating low cytotoxicity against the two cell types. Compared with groups treated with free drugs, HA-DOX-CDDP-treated groups exhibited significantly more cytotoxicity toward 4T1 cells, which express high levels of CD44 receptors, following treatment at 2.5, 5, and 10 μg/ml for 24 h, or treatment at a concentration of 0.65 μg/ml or higher for 48 h (Figure 3B). In contrast, no significant differences in cell viability were observed following treatment with free drugs or HA-DOX-CDDP in 3T3 cells, which express low levels of CD44. HA-DOX-CDDP against the 3T3 and 4T1 cells with the IC50s of 3.07 and 1.10 μg/ml for 24 h exposure, but 2.04 and 0.34 μg/ml for 48 h exposure. Moreover, free DOX plus CDDP against the 3T3 and 4T1 cells with the IC50s of 3.45 and 1.16 μg/ml for 24 h exposure, but 3.58 and 0.56 μg/ml for 48 h exposure.

Three-dimensional multicellular spheroid models of 3T3 and 4T1 cells were established to further evaluate the tumor inhibition effects and cellular uptake of HA-DOX-CDDP micelles (see Figure 4). The 4T1 spheroid volumes in the HA-DOX-CDDP treatment group were significantly smaller than those in the free drugs group (P < 0.05). In contrast, no significant differences in 3T3 spheroid volume were observed between the HA-DOX-CDDP and DOX+CDDP groups (Figures 4A, B). In addition, drug penetration was evaluated following co-incubation with drugs for 24 h. Compared with the free drugs group, the cores of the multicellular 4T1 spheroids in the HA-DOX-CDDP group showed much stronger fluorescence intensity (P<0.05) (Figures 4A, C). Consistent with the proliferation results, no significant differences in fluorescence intensity were observed between the HA-DOX-CDDP and DOX+CDDP groups in the 3T3 cell model. These results demonstrate that the nanoparticulate drug formulation showed enhanced inhibition of cell proliferation and cellular uptake via CD44 receptors on the surfaces of 4T1 tumor cells.

Drug distribution in major organs and isolated tumors was evaluated 6 or 12 h after injection of free or nanoparticulate drugs (Figure 5). The livers and kidneys showed strong fluorescence in the HA-DOX-CDDP and free DOX+CDDP groups 6 h post-injection, and the fluorescence intensity was higher at 12 h (Figure 5A). These results may be attributed to the high rate of metabolism of small-molecule anticancer drugs in the liver and kidney (Wicki et al., 2015; Huang et al., 2017). The fluorescence signals in major organs were significantly weaker for the HA-DOX-CDDP group than for the free drug group at both 6 and 12 h post-injection, except in the spleen, suggesting lower levels of drug distribution in the spleen relative to the other organs and tissues. In contrast, the fluorescence intensity in tumors was 1.3- and 2.1-fold higher in the HA-DOX-CDDP group than in the free drug groups after 6 and 12 h exposure, respectively (Figure 5B).

The tumor suppression effect of NPs and free drugs in vivo was evaluated in BALB/c mice injected with 4T1 cells in the abdominal fat pad. Tumor volumes and body weights were calculated at the end of the experiment. As shown in Figure 6A, the tumors progressed rapidly with an average volume of 1600 ± 487 mm³ in the control group. In contrast, the tumors were significantly smaller in the HA-DOX-CDDP- and DOX+CDDP-treated groups. Treatment with HA-DOX-CDDP resulted in 66% tumor inhibition compared with that in the other treatment groups. In addition, body weight decreased rapidly in the free DOX+CDDP group during the initial 10 days of treatment, then stabilized gradually. Overall, the animals in this group experienced a 19% loss in body weight (Figure 6B). In the HA-DOX-CDDP group, body weight decreased at a rate similar to that in the control group during the early treatment stage, and total body weight loss was 3.7%. Body weight did not significantly change in the control group. As shown in Figure 6C, organ coefficients were calculated to evaluate side effects of drug treatment. In the free DOX+CDDP group, the organ coefficients of the liver and spleen were obviously lower than in the other two groups (P<0.01). There were no significant differences in organ coefficients between the HA-DOX-CDDP and control groups.

Pathological analyses of tumor tissues were performed to confirm the anticancer efficacies of the different drug formulations. As shown in Figure 7A, most tumor cells exhibited integrated cellular morphologies in the control group. In contrast, tumor cell necrosis was observed in the HA-DOX-CDDP and DOX+CDDP groups. Tissues in the HA-DOX-CDDP group had larger necrotic areas (56.3% ± 5.8%) than those in the free DOX+CDDP (35.65 ± 6.5%) and control groups (1.5% ± 0.8%) (Figure 7B).

To investigate mechanisms of apoptosis in tumor cells following treatment, immunofluorescence of PARP and survivin was evaluated in tumor tissue sections. As shown in Figure 7C, the fluorescence density of PARP in the control group was lower than that in the other two groups. Furthermore, PARP fluorescence intensity in the HA-DOX-CDDP group was approximately 4.15- and 2.23-fold higher than that in the free drugs and control groups, respectively (Figure 7D). Compared with the control group, the fluorescence intensity of survivin was lower, to differing extents, in the HA-DOX-CDDP and DOX + CDDP groups. Furthermore, the HA-DOX-CDDP groups showed significantly weaker survivin fluorescence than the free drugs group (Figure 7E).