Enhancing the Anticancer Activity of Attenuated Listeria monocytogenes by Cell Wall Functionalization with "Clickable" Doxorubicin

Abstract

Among bacteria used as anticancer vaccines, attenuated Listeria monocytogenes (Lm^at) stands out, because it spreads from one infected cancer cell to the next, induces a strong adaptive immune response, and is suitable for repeated injection cycles. Here, we use click chemistry to functionalize the Lm^at cell wall and turn the bacterium into an "intelligent carrier" of the chemotherapeutic drug doxorubicin. Doxorubicin-loaded Lm^at retains most of its biological properties and, compared to the control fluorophore-functionalized bacteria, shows enhanced cytotoxicity against melanoma cells both in vitro and in a xenograft model in zebrafish. Our results show that drugs can be covalently loaded on the Lm^at cell wall and pave the way to the development of new two-in-one therapeutic approaches combining immunotherapy with chemotherapy.

Figure 1

Generation of flu-Lm^at and characterization of its biological features in melanoma cell lines. (a) Schematic representation of the two-step approach used to functionalize the Lm^at cell wall. In the first step, Lm^at is incubated with an alkyne-d-alanine probe (alkDA, upper) or an alkyne-d-alanine-d-alanine probe (alkDADA, lower), which result in the metabolic functionalization of the fifth or fourth d-alanine of the PG stem pentapeptide with an alkyne group, respectively. In the second step, the azide-bearing AF488 green fluorophore (az-AF488) is attached to bacterial cell wall via click reaction (CuAAC reaction, using BTTP as the ligand), so that fluorescent Lm^at is obtained. When alkDA (but not alkDADA) is used, d,d-carboxypeptidases (CPs) and d,d-transpeptidases (TPs) can remove the alkyne- and/or fluorophore-bearing fifth d-alanine from the PG stem peptide, decreasing loading efficiency. TPs also cross-link the fourth d-alanine to meso-diaminopimelic acid (m-DAP), contributing to confer PG its characteristic 3D meshlike structure. (b) Fluorescence microscope images of bacteria incubated overnight (ON) with 1 mM alkDA probe (i, top) or alkDADA probe (ii, bottom), and MFI of bacteria populations incubated with increasing probe concentrations (middle). (c) Bacteria viability after ON incubation with 40 mM alkDA (dark gray bar) or alkDADA (gray bar) probe. (d) Viability of bacteria subjected to CuAAC reaction, after ON incubation with 40 mM alkDA probe (dark green bar) or alkDADA probe (green bar). For click reaction, the following optimized protocol was used: 25 μM az-AF488, 2.5 mM sodium ascorbate, 20 μM CuSO₄ and 160 μM BTTP, in PBS buffer, with incubation time set at 10 min (see Figure S4). (e) Proliferation of bacteria incubated ON with 40 mM alkDA or alkDADA probe and then subjected to CuAAC reaction with az-AF488 (flu-Lm^at-alkDA, dark green line; flu-Lm^at-alkDADA, green line). Unlabeled Lm^at (not incubated with a probe nor subjected to CR) is taken as control (black line). Panels (f)–(h) show infectivity assays of AF488-loaded Lm^at. (f) Representative confocal images of A375 melanoma cells after 3 h of infection with unlabeled Lm^at (left), flu-Lm^at-alkDA (middle), and flu-Lm^at-alkDADA (right). Blue denotes DAPI staining of cell nucleus; red denotes staining of actin filaments using Phalloidin 594. Green denotes flu-Lm^at-alkDA and flu-Lm^at-alkDADA detected through AF488 green fluorophore. Pink represents a rendering of Lm^at staining with primary anti-Listeria antibody and far-red secondary antibody. (g) A375 cells were infected at MOI 100 with bacteria incubated or not with alkDA probe or alkDADA probe and then subjected or not to click reaction. After 1 h of infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. At 3 h (left bars) and 6 h (right bars) post-infection, cells were lysed and intracellular Lm^at was quantified by plating for CFU. (h) A375 cells were infected at MOI 100 with flu-Lm^at-alkDA (dark green bars), flu-Lm^at-alkDADA (green bars) or unlabeled Lm^at (not incubated with a probe nor subjected to CR, black bars), in the presence of the indicated concentration of SMER28 inhibitor. After 2 h of infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. At 3 h post-infection, cells were lysed and intracellular Lm^at was quantified by plating for CFU. Panels (i) and (j) show intracellular replication of AF488-loaded Lm^at. (i) Bacteria doubling time between 3 h and 6 h was calculated based on the CFU obtained in panel (g). (j) Confocal microscope images of 501 Mel cells infected with flu-Lm^at-alkDADA at MOI 100. After 1 h of infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. The fluorescence images are representative of the increase in the number of intracellular bacteria over time (3, 6, and 12 h post-infection). Legend: blue, DAPI staining of cell nucleus; red, staining of actin filaments using Phalloidin 594; green, flu-Lm^at-alkDADA detected through AF488 green fluorophore. Panels (k) and (l) show cell-to-cell spreading assays. (k) Schematic representation of the experimental approach. 501 Mel melanoma cells were infected at MOI 50 with Cy5-loaded Lm^at-alkDA or Lm^at-alkDADA. After 2 h of infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. Then, infected cells were collected at 3 and 18 h post-infection, stained with anti-Listeria primary antibody and Alexa Fluor 488 secondary antibody, and analyzed by flow cytometry to determine the percentage of cells that acquire green fluorescence due to Lm^at spreading. (l) Percentage of green 501 Mel cells at 18 vs 3 h post-infection with flu-Lm^at-alkDA (dark green bars), flu-Lm^at-alkDADA (green bars), or unlabeled Lm^at (not incubated with a probe nor subjected to CR, black bars). (m) Kill rate assay. A375 melanoma cells were infected with AF488-loaded Lm^at-alkDA or Lm^at-alkDADA at MOI 2000. At 3 h post-infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. At 24 h post-infection, cells were fixed and stained with DAPI to count nuclei by fluorescence microscopy. [Legend: NP, no probe; PBS, no click reaction; CR, click reaction; CFU, colony forming units; MOI, multiplicity of infection; MFI, median fluorescence intensity. Graphs represent the mean ± SEM of at least three independent experiments, performed by using at least two independently functionalized stocks of Lm^at. Unpaired t-test. (*) p < 0.05, (**) p < 0.01, (***) p < 0.001. ns: not statistically significant.]

Figure 2

dox-Lm^at shows enhanced anticancer potential in melanoma cell lines. (a) Schematic representation of the experimental design used to functionalize Lm^at with doxorubicin and investigate whether dox-loaded Lm^at has enhanced cytotoxicity in vitro. Once preincubated with alkDADA, Lm^at is loaded with azide-bearing molecules (az-ATTO740 (az-flu), az-doxorubicin (az-dox) or az-VC-doxorubicin (az-VC-dox)) via CuAAC reaction to generate flu-Lm^at, dox-Lm^at, and dox-VC-Lm^at, respectively. For click reaction, the following optimized protocol was used: 5 μM az-ATTO740, 200 μM az-dox, or 200 μM az-VC-dox; 7.5 mM sodium ascorbate, 60 μM CuSO₄ and 480 μM BTTP; 0.9% w/v NaCl in water as reaction solvent; 25% DMSO as a cosolvent. After infection with flu-Lm^at, melanoma cells show decreased viability due to bacteria intrinsic cytotoxicity, which is enhanced when dox-Lm^at or dox-VC-Lm^at are used instead. (b) Pictures of bacterial cell pellets (top) and fluorescence microscope images (bottom) of: untreated Lm^at; Lm^at not metabolically labeled with the probe, but subjected to CuAAC reaction with az-dox (Lm^at + az-dox) or az-VC-dox (Lm^at + az-VC-dox); dox-Lm^at; dox-VC-Lm^at. (c) Quantification by flow cytometry of the MFI of the samples treated as in panel (b). (d) FLIM phasors plot of dox-Lm^at. The phasor populations of the different samples lie on different regions of the plot. From left to right: untreated Lm^at (blue teardrop); Lm^at not metabolically labeled with the probe but subjected to CuAAC reaction with az-dox (green teardrop); dox-Lm^at (yellow teardrop); az-dox (red teardrop). (e, f) Viability (panel (e)) and proliferation (panel (f)) of untreated Lm^at (black), flu-Lm^at (green), dox-Lm^at (red), and dox-VC-Lm^at (purple). Panels (g) and (h) show infectivity assays. (g) Representative confocal microscope images of A375 cells after 3 h of infection with dox-Lm^at. Blue denotes DAPI staining of cell nucleus. Green denotes staining of actin filaments using Phalloidin 488. Red denotes dox-Lm^at. (h) A375 cells were infected at MOI 100 with untreated Lm^at (black), flu-Lm^at (green) dox-Lm^at (red), and dox-VC-Lm^at (purple). After 2 h of infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. At 3 h (left bars) and 18 h (right bars) post-infection, cells were lysed and intracellular Lm^at was quantified by plating for CFU. (i) Intracellular replication of bacteria. Bacteria doubling time between 3 h and 18 h was calculated based on the CFU obtained in (h) for Lm^at (black), flu-Lm^at (green), dox-Lm^at (red), and dox-VC-Lm^at (purple). (j) Kill rate assay. A375 melanoma cells were infected with flu-Lm^at, dox-Lm^at or dox-VC-Lm^at at MOI 2000. At 3 h post-infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. At 48 h post-infection, cells were fixed and stained with DAPI to count nuclei by fluorescence microscopy. Panels (k) and (l) show the proliferation status of A375 melanoma cells infected with flu-Lm^at, dox-Lm^at, or dox-VC-Lm^at at MOI 1000. After 3 h post-infection, extracellular Lm^at was killed by medium replacement with fresh gentamycin-containing medium. At 48 h post-infection, cells were stained with anti-MCM7 antibody and proliferative vs nonproliferative cells were counted based on the presence vs absence of MCM7 nuclear staining. Representative microscope images (panel (k)) and quantification (panel (l)) of proliferative and nonproliferative A375 cells after infection with flu-Lm^at, dox-Lm^at, or dox-VC-Lm^at. Blue denotes DAPI; green denotes anti-MCM7 antibody. Panels (m) and (n) show the area of cancer cell mass developed in a xenograft model in zebrafish embryos. eGFP-expressing A375-PIG cells, previously infected with flu-Lm^at or dox-Lm^at at MOI 1000 for 2 h, were injected in 48 hpf embryos. [Here, and throughout, hpf stands for hours post-fertilization.] Then, embryos were allowed to grow for additional 48 h. At the end of this period, the area of green cancer cell mass was quantified. (m) Results of area quantification; at least 100 embryos were injected per experimental condition. (n) Representative pictures of 96 hpf embryos that, 48 h earlier, were injected with A375-PIG cells uninfected (left), infected with flu-Lm^at (middle), or infected with dox-Lm^at (right). The shape of the embryo and the perimeter of the injection site (yolk sac) are highlighted with a white dotted line, while the mass of cancer cells within the yolk sac (indicated with a white arrow) shows a green fluorescence signal. Scale bar = 300 μm. [Legend: NT, untreated cells; NP, no probe; PBS, no click reaction; CR, click reaction; CFU, colony forming units; MOI, multiplicity of infection; MFI, median fluorescence intensity. Graphs represent the mean ±SEM of at least three independent experiments, performed by using at least two independently functionalized stocks of Lm^at.] Unpaired t-test (in vitro assays), Kruskal–Wallis test (Dunn’s multiple comparisons test, xenograft assay). (*) p < 0.05, (**) p < 0.01, (***) p < 0.001, (****) p < 0.0001. ns, not statistically significant.]