Engineered macrophages as near-infrared light activated drug vectors for chemo-photodynamic therapy of primary and bone metastatic breast cancer

Patients with primary and bone metastatic breast cancer have significantly reduced survival and life quality. Due to the poor drug delivery efficiency of anti-metastasis therapy and the limited response rate of immunotherapy for breast cancer, effective treatment remains a formidable challenge. In this work, engineered macrophages (Oxa(IV)@ZnPc@M) carrying nanomedicine containing oxaliplatin prodrug and photosensitizer are designed as near-infrared (NIR) light-activated drug vectors, aiming to achieve enhanced chemo/photo/immunotherapy of primary and bone metastatic tumors. Oxa(IV)@ZnPc@M exhibits an anti-tumor M1 phenotype polarization and can efficiently home to primary and bone metastatic tumors. Additionally, therapeutics inside Oxa(IV)@ZnPc@M undergo NIR triggered release, which can kill primary tumors via combined chemo-photodynamic therapy and induce immunogenic cell death simultaneously. Oxa(IV)@ZnPc@M combined with anti-PD-L1 can eliminate primary and bone metastatic tumors, activate tumor-specific antitumor immune response, and improve overall survival with limited systemic toxicity. Therefore, this all-in-one macrophage provides a treatment platform for effective therapy of primary and bone metastatic tumors.


Response to Reviewers
In recently years, cell-based systems, such as erythrocytes, platelets, stem cells, 218 monocytes/macrophages, lymphocytes, dendritic cells, neutrophils, have emerged as 219 interesting alternative to biological drug carriers 1 . These cell systems possess numerous 220 advantages including prolonged blood circulation time, high biocompatibility, high mobility 221 and inherent biodegradability 2 . For example, erythrocytes are also named red blood cells 222 (RBCs), in addition to the above advantages, they also have a high surface to volume ratio, 223 which is beneficial for drug loading and surface modification, and they are no nucleus, nucleus, in DiD positive macrophages of Oxa(IV)@Lip@M were counted to be about 70%, which was 250 similar to that of LPS treated M1 cells, while blank BMMs only counted to be about 37%, 251 indicating that the drug loaded macrophages could retain its M1 phenotype in the tumor tissue.    Response: Thanks for reviewer's comment. According to Fig. 4c and e (Fig. R3b and d), 287 though Oxa(IV)@ZnPc@M had similar contents in the bone tumor and the primary tumor, it 288 is no doubt that the drug loaded macrophages have stronger tumor targeting ability than 289 nanoparticles in both primary tumor and bone tumors. According to the results, is indeed not 290 reasonable to define the bone metastatic tumor as EPR-deficient, we changed the description 291 in the revised manuscript, and EPR-deficient was deleted. Actually, in addition to the EPR effect, there are so many factors that can affect the tumor distribution behaviors no matter the 293 nano-based or cell-based systems. We know that using a better EPR-deficient model can more 294 perfectly support our paper, we tried to use mice that have been inoculated with tumor cell but 295 did not form visible tumors to serve as the EPR-deficient bone metastatic tumor model, but we 296 failed to get obvious imaging results, so we used mice that have established visible bone 297 metastatic tumors for in vivo fluorescence imaging, but it is not strictly a EPR-deficient tumor 298 model, thus the result may reasonable. We will try our best in building a generally accepted intensity of primary tumors and bone metastatic tumors. *p <0.05, **p <0.01, ***p < 0.001 (n=3). 307 308 3. Comment: The mechanism of the drug (Oxa and ZnPc) release from Oxa(IV)@ZnPc should 309 be further investigated. Besides, for Oxa(IV)@ZnPc@M without irradiation, is the drug (Oxa 310 and ZnPc) release from the system due to their exocytosis from the macrophages? Or the drug 311 was first released from Oxa(IV)@ZnPc and then from macrophages.   The release of platinum from Oxa(IV)@ZnPc was in a pH dependent manner, with less 325 than 20% was released at pH 7.4 within 12 h, while a significantly increased release at pH 5.5 326 was observed (Fig. R4a). This was attributed to the formation of the characteristic acid-327 cleavable coordination bonds 8, 9 in the core of Oxa(IV)@ZnPc. Free ZnPc is an extremely 328 hydrophobic photosensitizer, it can easily accumulate in the aqueous, resulting in UV-vis and 329 fluorescence quenching (Fig. R4c-d). Therefore, it is hard to investigate the in vitro drug 330 release behavior of ZnPc. After a series of screening, we tried to use PBS buffer containing 2% 331 tween 80 as release medium. The release of ZnPc from Oxa(IV)@ZnPc is slow under the 332 current release medium. The drug release rate at pH 5.5 is slightly higher than pH 6.5 and pH 333 7.4, this might be due to that the acid-cleavable coordination bonds in the core of 334 Oxa(IV)@ZnPc could fracture in acidic conditions, leading to the decomposition of 335 nanoparticles and the re-assembly of the outer lipid.  and ZnPc respectively (Release 1-free drug). The sediment was dispersed in 2 mL H2O for 352 fluorescence detection of calcein and ZnPc respectively (Release 2-nanoparticle).

353
The detectable fluorescence intensity of calcein and ZnPc in Release 1 (Fig. R5a), 354 indicated that the drug within BMMs can be released in the form of free molecule. The 355 detectable fluorescence intensity of calcein and ZnPc in Release 2 (Fig. R5b), indicated that the drug within BMMs can be partly released in the form of nanoparticles, consistent with the 357 result in Fig. 3b and 3e. According to the above analysis, the release of drug from 358 Oxa(IV)@ZnPc@M could be exist in two patterns, one part released by exocytosis in the form 359 of nanoparticles, the other part was first released from Oxa(IV)@ZnPc and then from  Oxa(IV)@ZnPc@M is carried out in a static environment in vitro, which can ensure a higher 370 incubation concentration, thus the present Oxa(IV)@ZnPc@M has high drug loading. As 371 nanoparticles can be distributed in various tissues throughout the body after intravenous 372 injection, the amount of nanoparticles that can really bind to the monocytes in the blood can be 373 very small, and since the blood is in a flowing state, which affect the uptake of nanoparticles 374 by the blood monocytes as well. As for nanoparticles captured by the reticuloendothelial The long-term stability of the nanoparticles should be evaluated. Cal@ZnPc (using fluorescent dye calcein to replace Oxa prodrug) can be more directly explain 399 the problem. Therefore, we further performed the experiment using Cal@ZnPc as the model 400 ( Fig. R7), the results was consistent with that of using Cal@DiI. In the revised manuscript, we 401 replaced Fig. 1e with Fig. R7, and all the changes were highlighted in red.    for assessing IFN-γ and TNF-α in CD8 + T cells. Live/dead cell discrimination was conducted 481 using eFluor TM 780 according to manufacturer's instructions.

482
As shown in Fig. R12, mice treated with Oxa(IV)@ZnPc@M (+) + anti-PD-L1 exhibited 483 the highest effect in promoting intratumoral infiltration of CD8 + and CD4 + T cells , where 9.99% 484 CD8 + T cells and 19.0% CD4 + T cells were induced, which were 5.9-fold and 6.9-fold higher   SPP@ZnPc, Cal@ZnPc and Oxa(IV)@Lip were defined in the main text for better 563 understanding. Cal@DiI was deleted in the revised manuscript. Some missing unit is "℃", this 564 may be due to a problem with the submission system when converting word to PDF. Besides, 565 the entire manuscript was carefully checked and some writing and grammatical mistakes were 566 corrected in the revised manuscript, and all the changes made in the manuscript were 567 highlighted in red. We appreciate the reviewer so much to help us improve our manuscript. This manuscript addresses an interesting approach to deliver therapies to breast tumors in the 572 primary and bone sites. They showed strong potential therapy using this approach, but there 573 are a few improvements that need to be made to this manuscript. Overall, it describes a novel 574 approach to drug delivery that could be of interest to a general audience and potentially 575 applicable to other diseases.  were highlighted in red. We appreciate the reviewer so much to help us improve our manuscript.

599
The sentence"Despite promising, immunotherapy is restricted to only a fraction of cases 600 and tumor types, and causes specific immunotoxicity in some patients"was revised as "Despite     Response: Thanks for reviewer's comment. We tried our best to condense the introduction, the sentence that discuss the results in the introduction were deleted or moved to 621 the part of discussion. The detected contents were listed as follows. We appreciate the reviewer 622 so much to help us improve our manuscript.

919
All of these results indicated the cell-based system had higher effect to elicit the immune 920 system by the combined chemo/photo/immuo-therapy than that of nano-based system. This 921 part of experiment were added to the revised manuscript, the related pictures have been 922 reformatted in the revised manuscript, and all the changes were highlighted in red.  dramatically up-regulated compared with other treatment groups (Fig. R26), indicating the best 975 effect on maturating and activating of DCs, which showed great potential to stimulate the 976 subsequent tumor-specific immune response. This section of experiment was added to the 977 revised manuscript, and all the changes were highlighted in red.

994
All of these results indicated the cell-based system had higher effect to elicit the immune 995 system by the combined chemo/photo/immuo-therapy than that of nano-based system. This 996 part of experiment were added to the revised manuscript, the related pictures have been 997 reformatted in the revised manuscript, and all the changes were highlighted in red. tumor-draining lymph nodes on day 2 after various treatment. *p <0.05, **p <0.01, ***p < 0.001 (n=4). 1005   Fig.1j, what does the white arrows and the blank dots represent? Does