Peritoneal tumor carcinomatosis: pharmacological targeting with hyaluronan-based bioconjugates overcomes therapeutic indications of current drugs

Abstract

Peritoneal carcinomatosis still lacks reliable therapeutic options. We aimed at testing a drug delivery strategy allowing a controlled release of cytotoxic molecules and selective targeting of tumor cells. We comparatively assessed the efficacy of a loco-regional intraperitoneal treatment in immunocompromised mice with bioconjugates formed by chemical linking of paclitaxel or SN-38 to hyaluronan, against three models of peritoneal carcinomatosis derived from human colorectal, gastric and esophageal tumor cell xenografts. In vitro, bioconjugates were selectively internalized through mechanisms largely dependent on interaction with the CD44 receptor and caveolin-mediated endocytosis, which led to accumulation of compounds into lysosomes of tumor cells. Moreover, they inhibited tumor growth comparably to free drugs. In vivo, efficacy of bioconjugates or free drugs against luciferase-transduced tumor cells was assessed by bioluminescence optical imaging, and by recording mice survival. The intraperitoneal administration of bioconjugates in tumor-bearing mice exerted overlapping or improved therapeutic efficacy compared with unconjugated drugs. Overall, drug conjugation to hyaluronan significantly improved the profiles of in vivo tolerability and widened the field of application of existing drugs, over their formal approval or current use. Therefore, this approach can be envisaged as a promising therapeutic strategy for loco-regional treatment of peritoneal carcinomatosis.

Figure 1. Interaction of bioconjugates with cancer cell lines.

A, BODIPY-labeled ONCOFID-P (50 µg/mL in paclitaxel equivalents) or ONCOFID-S (50 µg/mL in SN-38 equivalents) were added to tumor cells and flow cytometry analysis was performed at different time points thereafter (0.5, 1, 2, 5, 10, 15, 30 or 60 minutes). Panels illustrate cytometry profiles at 3 representative time points. B, whole kinetics of interaction at all time points tested. C, kinetics of the fluorescence intensity (geo mean) detected on tumor cells at the same time points analysed as in B. Panels B and C report mean ± SD of 3 independent experiments.

Figure 2. Endocytosis pathways involved in bioconjugate cell entry.

HT-29, MKN-45 and OE-21 tumor cells were left untreated (solid line) or treated (dashed line) for 1 hour with selective chemical inhibitors of different pathways involved in endocytosis (amiloride, chlorpromazine, cytochalasin D and filipin III). Subsequently, cells were exposed for 30 minutes to ONCOFID-P and then treated with hyaluronidase for 4 hours, to be finally analyzed by flow cytometry. Data at the upper-right corner of each panel report the respective geo mean values, and the percentage of reduction induced by treatment.

Figure 3. Confocal microscopy analysis and co-localization studies.

A, accumulation of bioconjugates in HT-29, MKN-45 and KYSE-30. Cells were incubated with BODIPY-labeled ONCOFID-P (50 µg/mL in paclitaxel equivalents) or ONCOFID-S (50 µg/mL in SN-38 equivalents) for 1 hour, washed and fixed before analysis. B, co-localization analysis of bioconjugates in lysosomes. HT-29, MKN-45 and OE-33 cells were treated with LysoTracker green, incubated with BODIPY-labeled compounds and finally analyzed by confocal microscopy. Left pictures show the fluorescence of the labeled bioconjugates (red) in single cells, while central pictures illustrate signals (green) from lysosomes. The merging of the 2 components is visible in right pictures. Lysosomes were occupied by bioconjugates by ∼90% to 100%, as assessed by the Zeiss’profile software tool. Experiments were repeated at least twice with consistent results.

Figure 4. Assessment of bioconjugate mechanism of action.

A, rearrangement of tumor cell microtubular architecture after drug treatment. HT-29, MKN-45 and OE-21 cells were treated with ONCOFID-P or free paclitaxel for 4 hours at 37°C. After treatment, cells were fixed, permeabilized, and stained with an anti-β-tubulin mAb and anti-mouse Ig Alexa 546-conjugated antiserum. Cells treated with free drug or bioconjugate disclosed the same interferences on the microtubular mesh. B, inhibition of Topo I activity after ONCOFID-S or SN-38 treatment in HT-29, MKN-45 and OE-21 cells. Gels show the supercoiled or relaxed forms of pBR322 plasmid after incubation with a 1∶50 dilution of nuclear protein neat extracts obtained from tumor cells treated with conjugated or free drug for 4 hours. Lane 1, marker; lane 2, relaxed pBR322 plasmid (positive control); lane 3, supercoiled plasmid (negative control); lane 4, supercoiled plasmid in the presence of nuclear protein neat extract from drug-untreated cells; lane 5, supercoiled pBR322 admixed with nuclear protein neat extract from ONCOFID-S treated cells; lane 6, supercoiled pBR322 admixed with nuclear protein neat extract from SN-38-treated cells. C, quantification of the reactions shown in B. Figure reports mean ± SD of 3 independent experiments.

Figure 5. Assessment of in vivo tumor growth and response to therapy.

A, bioluminescence imaging of pharmacologically treated or untreated mice with peritoneal carcinomatosis induced by luciferase-transduced tumors. Panels show three representative mice per group at one month after tumor injection. B, cumulative results. Each box plot reports mean ± SD of total photon emission from 6 mice per group at one month from peritoneal carcinomatosis induction. Statistical analysis (Kruskal-Wallis test) is reported in tables at the right of each panel.

Figure 6. In vivo therapeutic activity of bioconjugates.

Kaplan-Meier survival curves of mice with peritoneal carcinomatosis from HT-29, MKN-45 or OE-21 tumor cells. Animals were randomly assigned to an experimental group and drug treatment was initiated according to therapeutic schedule reported in Materials and Methods. All experimental groups were statistically compared each other, but only significant values are reported in each panel.