TUNEL assay showed apoptosis increased by PLD+pUH, not by CQ

TUNEL assay was performed to assess the effects of different therapeutics on apoptosis of the treated tumors. Figure 3-9A showed that there were apparently stronger signals of apoptosis (green dots) in the PLD+pUH+CQ and PLD+pUH groups than those in the CQ and control groups, and the differences were highly significant (Figure 3-9B). The intensities of apoptosis signal were statistically similar in the CQ group and the control group. The difference between PLD+pUH+CQ group and PLD+pUH group was not significant as well. These results reflected that cancer cell apoptosis was blotting targeting LC3. The expression of LC3-II was raised in the PLD+pUH+CQ and CQ groups, but not in the PLD+pUH and control groups. These findings coincided with the autophagy inhibition caused by CQ administration observed in

immunohistochemical stain.

3.4 Discussion

Previous studies [29,61] showed that PLD in combination with pUH significantly inhibited the growth of 4T1 murine breast cancer in an in vivo brain metastatic tumor model. It was proved that pUH on the tumor after injection of PLD significantly increased the accumulation of PLD in the sonicated tumor tissue and cancer cells. Based on these results, we investigated if inhibition of tumor growth and tumor relapse could be further improved by combining PLD+pUH with CQ. To have a better observation, a subcutaneous tumor model was used instead. The experimental results shown in Figure 3-5 displayed that both PLD+pUH with and without CQ could successfully retard 4T1 tumor growth right after the treatment. Stronger retarding effect on tumor growth was obtained with the administration of CQ. The tumor inhibitory effect in the PLD+pUH group did not persist long, as the treated tumor began to grow again since Day 17.

Treatment with PLD+pUH destroyed most cancer cells and resulted in tumor shrinkage, but still a small fraction of damaged cancer cells survived and regained the tumor growth several days after the treatment. The story tremendously differed in the case with PLD+pUH treatment followed by daily CQ administration. Treatment with PLD+pUH+CQ not only retarded cancer tumor growth more effectively, but also postponed the tumor regrowth until Day 32 and the regrowth rate of tumor was much slower as compared with the PLD+pUH group. This significant difference suggested that CQ, by blocking the escape mechanism of autophagy, rendered the cancer cells damaged by PLD+pUH treatment more difficult to survive. The prolonged retardation on tumor growth resembled the results obtained by Hirsch et al. [63,64], in which they

when co-administered with doxorubicin. It implied that the recurrence-postponing effect of CQ observed in our study might be partially attributed to similar anti-CSC activity.

Several researchers claimed that CQ potentiates anti-cancer therapeutics through mechanisms outside autophagy inhibition. Maes et al. found that CQ normalized tumor vessel structure, and the effect could not be mimicked by genetic inhibition of autophagy [41]. Balic et al. indicated that CQ suppressed CSC via inhibition of CXCR4 and Hedgehog signaling [40]. King et al. proposed that CQ synergized mTORi through mechanism related to cholesterol metabolism [65]. Nonetheless, the majority of researches still attributed the anti-tumor potency of CQ to its lysosomal acidifying/autophagy inhibition ability. Choi et al. concluded the anti-CSC activity of CQ was through autophagy inhibition despite their finding that Jak2-STAT3 pathway might also play a role [66]. Wei et al. showed in their study that autophagy inhibition rendered CSC susceptible to photodynamic therapy, regardless pharmacologic inhibition with CQ or genetic silencing [67]. Likewise, Lee et al. also found that CQ sensitized glioma cells to temozolomide, and the sensitizing effect was observed with other autophagy inhibitors as well [68]. They also demonstrated the sensitization by autophagy inhibition was p53 dependent. Similarly, Maycotte et al. identified certain subtypes of triple-negative breast cancer more responsive to autophagy-inhibiting treatment, and this susceptibility could be predicted by high STAT3 expression [69].

Furthermore, it is indicated that some activity of CQ shares the same underlying mechanism with autophagy inhibition: both are consequences of lysosome disruption.

Elliot et al. found that lysosomal inhibition by CQ impaired de novo nucleotide biosynthesis and depleted aspartate in pancreatic ductal adenocarcinoma [70]. Even the aforementioned vessel-normalizing ability of CQ was proposed in a recent study to be related to lysosomal dysfunction [71].

In immunohistochemical studies and Western blotting, the accumulation of LC3-II was markedly increased in PLD+pUH+CQ group and in CQ group. These findings reflected the late-stage inhibition of autophagy by CQ. On the other hand, it was observed that the expression of LC3-II was slightly decreased in PLD+pUH group comparing to control group. Autophagy is usually induced by hyperthermic treatment as a response to protect cell from metabolic stress [72,73]. Nevertheless, there are controversy about how doxorubicin affects autophagy [74,75]. It is suggested that doxorubicin stimulates the initiation of autophagy but interferes the lysosomal function, therefore resulting in overall decrease in autophagy flux [75]. Park et al. demonstrated that doxorubicin reduces autophagosome formation via increase in mTOR expression [76]. The slightly reduced expression in LC3-II observed in PLD+pUH group could be explained by the summation of two opposite effects: upregulation by hyperthermia and downregulation by doxorubicin.

In our study, we used PLD+pUH to treat the tumors and employed CQ to prevent the remnant cancer cells escaping through autophagy. The results showed that when CQ used in combination with PLD+pUH, it assisted tumor suppression and postponed or even prevented tumor relapse. But CQ monotherapy has little impact on tumor growth as compared to the control group (as shown in Figure 3-5B). Autophagy works as a surviving mechanism when cancer cells face strong stresses. CQ blocks autophagy in cancer cells damaged by anti-tumor therapy, and therefore aggravates the severe condition remaining cancer cells confront, and eventually results in decreasing the probability of tumor relapse. However, this effect is not prominent by CQ monotherapy, and hence in the absence of antitumor therapeutics CQ does not alter tumor growth response.

Hyperthermia and CQ are both reported to activate anti-tumor immune system [12,77]. The expression of tumor antigens, including MHC class I, heat shock proteins, and exosomes, are observed in hyperthermia-treated tumors. Hyperthermia also stimulates NK cells, CD8+ T cells, and dendritic cells. Along with these, the trafficking of immune cells between lymphoid organs and tumor are improved [12]. CQ is shown to function as an anti-tumor immune modulator by resetting tumor-associated macrophages [77]. The immunomodulatory effects of hyperthermia and CQ may synergize each other and underlie the persistent suppression on tumor growth in our study.

The survival rate in the PLD+pUH+CQ group was better than that in the PLD+pUH group. Observed up to Day 60, the survival rate is 75% (6 among 8 treated mice) and 56% (5 among 9 treated mice) for the PLD+pUH+CQ group and the PLD+pUH group, respectively. There was no observable tumor (complete remission) for these surviving mice after the treatment of PLD+pUH with or without CQ. Despite the great success in achieving prolonged remission, the survival in the PLD+pUH+CQ group was not significantly better than PLD+pUH (p-value: 0.2). It might be due to that the therapeutic efficacy of PLD+pUH was sufficiently potent in suppressing tumor growth in the subcutaneous 4T1 tumor model, so the marginal benefit of adding CQ into treatment were harder to be demonstrated since PLD+pUH already had a good outcome.

The difference of survival owing to CQ might be more apparent in a more lethal tumor model refreactory to PLD+pUH treatment. The survival in CQ group was nearly identical to that in control group, consistent with the observation in tumor growth that CQ monotherapy had little benefit.

CQ in cooperation with nanomedicine has been suggested as a promising therapeutic strategy to treat cancer [78]. Pelt et al. highlighted in their review the

advantages of CQ to complement nanomedicine for cancer therapy, including autophagy inhibition, normalization of tumor vasculature, and reducing the hepatic clearance of nanoparticles [78]. There are increasing studies that practically exploit this strategy in cancer therapeutics. Sun et al. used CQ to reduce the ‘stemness’ of breast cancer stem cell to increase their susceptibility to chemotherapeutics such as doxorubicin and docetaxel [79]. Shao et al. designed a MPEG-PLA nanoparticle co-delivering CQ and doxorubicin to kill ovarian cancer, utilizing the lysosome-interfering property of CQ to hinder drug sequestration [80]. Wolfram et al. pretreated mice with CQ to reduce nanoparticle uptake by macrophage [81]. Lv et al. took advantage of the vessel-normalizing ability of CQ to improve microcirculation in tumor and promoted nanodrug delivery [82]. These studies demonstrated the versatile powerfulness of CQ in combination use with nanomedicine in cancer therapeutics. Our study further extended the strategy by adding pulsed-wave ultrasound hyperthermia, which assisted the delivery of nanodrug into tumor tissue and created a vicious microenvironment so that autophagy inhibition by CQ became crucial.

3.5 Conclusions

We demonstrated that pulsed-wave ultrasound hyperthermia (pUH) enhanced PLD delivery in combination with chloroquine (CQ) could persistently suppress 4T1 tumor growth and postpone its recurrence. These results may pave the way to develop new combinatorial strategy for treatment-refractory cancer.

Figure 3-1. The scheme of PEGylated Liposomal Doxorubicin (PLD) + pulsed-wave Ultrasound Hyperthermia (pUH) + chloroquine (CQ) in cancer treatment.

Figure 3-2. Time schedule of treatment experiment. PEGylated Liposomal Doxorubicin (PLD) was given intravenously on Day5 after tumor implantation. Pulsed-wave

ultrasound hyperthermia (pUH) was administered 10~15 minutes after PLD

administration. Then mice were orally fed chloroquine (CQ) dissolved in drink water daily till experiment end.

Figure 3-3. (A) Fluorescent microscopic images of 4T1 murine breast cancer cells in vitro treated with PLD+CQ+H or PLD+H. Doxorubicin (red) distribution with respect

to nuclei (blue, stained with Hoechst 33342 dye) were shown. (B) Mean fluorescent intensity of doxorubicin with respect to nucleus region area. Abbreviation: H:

hyperthermia. ns: not significant.

Figure 3-4. MTT Cytotoxicity Assay. The cell viability was reduced by PLD with hyperthermia in a dose-dependent manner. The addition of CQ (10μM) further potentiated the cytotoxicity of PLD+H comparing to the counterpart without CQ. **:

p<0.01, ***: p<0.001. Abbreviation: H=hyperthermia

Figure 3-5. (A) Representative photographs of tumor for each group. Region encircled by dashed line indicated tumor. Scale bar=1cm. (B) The response of subcutaneous 4T1

murine breast cancer to different treatment: PLD+pUH+CQ, PLD+pUH, CQ, and control groups. * denotes p<0.05, and ** denotes p<0.01 between PLD+pUH+CQ and PLD+pUH, respectively.

Figure 3-6. The Kaplan-Meier survival plot for PLD+pUH+CQ, PLD+pUH, CQ, and control groups.

Figure 3-7. Histological examinations with hematoxylin-eosin staining for each experimental group. Scale bar = 100 μm.

Figure 3-8. Immunohistochemical stain for LC3 (brown stain) for each experimental group. LC3 accumulation reflects late-stage inhibition of autophagy. Greatly increased accumulation of LC3 was observed in both PLD+pUH+CQ and CQ groups, whereas slightly increase in the PLD+pUH group, and nearly no accumulation in the control group. Scale bar = 100 μm.

Figure 3-9. (A) Fluorescent microscopic images of TUNEL assay for each experimental

group. Apoptotic signals (green) were much more enhanced in PLD+pUH+CQ group and PLD+pUH group. Scale bar = 200 μm. (B) The fluorescent intensities for each experimental group were quantified and analyzed for statistical significance. **: p<0.01.

***: p<0.001. ns: not significant.

Figure 3-10. Western blot for LC3 for each experimental group. Increased expression of LC3-II was observed in PLD+pUH+CQ group and CQ group, reflecting the late-stage autohphagy inhibition by CQ. LC3 expression was slightly reduced in PLD+pUH group comparing to control group.