Ultrasound-visible engineered bacteria for tumor chemo-immunotherapy

Yaozhang Yang; Yuanyuan Wang; Fengyi Zeng; Yuhao Chen; Zhiyi Chen; Fei Yan

doi:10.1016/j.xcrm.2024.101512

. 2024 Apr 18;5(5):101512. doi: 10.1016/j.xcrm.2024.101512

Ultrasound-visible engineered bacteria for tumor chemo-immunotherapy

Yaozhang Yang ^1,^2,³, Yuanyuan Wang ³, Fengyi Zeng ^1,², Yuhao Chen ⁵, Zhiyi Chen ^1,^2,^4,^∗, Fei Yan ^3,^6,^∗∗

PMCID: PMC11148858 PMID: 38640931

Summary

Our previous work developed acoustic response bacteria, which enable the precise tuning of transgene expression through ultrasound. However, it is still difficult to visualize these bacteria in order to guide the sound wave to precisely irradiate them. Here, we develop ultrasound-visible engineered bacteria and chemically modify them with doxorubicin (DOX) on their surfaces. These engineered bacteria (Ec@DIG-GVs) can produce gas vesicles (GVs), providing a real-time imaging guide for remote hyperthermia high-intensity focused ultrasound (hHIFU) to induce the expression of the interferon (IFN)-γ gene. The production of IFN-γ can kill tumor cells, induce macrophage polarization from the M2 to the M1 phenotype, and promote the maturation of dendritic cells. DOX can be released in the acidic tumor microenvironment, resulting in immunogenic cell death of tumor cells. The concurrent effects of IFN-γ and DOX activate a tumor-specific T cell response, producing the synergistic anti-tumor efficacy. Our study provides a promising strategy for bacteria-mediated tumor chemo-immunotherapy.

Keywords: gas vesicles, ultrasound imaging, engineered bacteria, drug delivery, tumor treatment, tumor microenvironment, recurrence and metastasis, doxorubicin, interferon-gamma, protein expression

Graphical abstract

Highlights

•
Ultrasound-visible engineered bacteria can guide therapeutic gene expression
•
The tumor-targeting bacteria can deliver therapeutic genes and chemical drugs
•
The strategy of chemo-immunotherapy exerts strong synergistic anti-tumor effects

Yang et al. genetically engineer bacteria, with an acoustic reporter gene and the hyperthermia-responsive IFN-γ gene, and chemically modify them with DOX. These GV-containing bacteria can guide hHIFU to precisely induce the expression of the IFN-γ gene. The concurrent effects of IFN-γ and DOX activate a tumor-specific T cell response, improving the anti-tumor efficacy.

Introduction

The integration of synthetic biology technologies into the genetically engineered bacteria greatly improves their ability to response to internal tumor microenvironment or external physical stimulation, making it possible to finely regulate expression of therapeutic genes. Nowadays, many chemical, physical, and biological methods combined with various promoter elements have been developed for regulating the gene expression in these genetically engineered bacteria.¹^,² For example, the quorum sensing system is used to initiate therapeutic gene expression only when their population density reaches a threshold level.³^,⁴ Ray irradiation⁵ and light stimulation⁶^,⁷ are also used for modeling gene expression, providing high spatiotemporal precision in live animals. However, ray irradiation is ionizing, which may cause damage to normal tissues in the irradiation path,⁸ and the light approach is limited by poor penetration of the deep-seated tumors.⁹ Our previous study¹⁰ developed a mild hyperthermia-responsive gene circuit, which was integrated into bacteria, realizing spatiotemporally controllable therapeutic gene expression and largely improving their safety and efficacy. However, there is still a lack of a suitable approach to visualize these genetically engineered bacteria, which inhabit tumor sites for imaging-guided acoustic stimulation.

To date, many imaging techniques, including fluorescence imaging (FI), photoacoustic imaging (PAI), radionuclide imaging (RNI), and magnetic resonance imaging (MRI), were applied to visualize bacteria through labeling with exogenous contrast agents or endogenous reporter genes. Generally, the exogenous contrast agents that are used for labeling of genetically engineered bacteria cannot proliferate along with bacterial growth, resulting in difficulty monitoring these bacteria in the long term. In the past decades, many genetically encoded reporter genes, such as phytochrome-based reporter genes for PAI,¹¹ red fluorescent protein (RFP) genes for FI,¹²^,¹³ and magnetosomes for MRI,¹⁴ have been developed, making it possible to track these genetically engineered bacteria over a relatively long time. Despite all this, these imaging modalities are not still the best choice for guiding acoustic stimulation to activate the mild hyperthermia-responsive gene circuit in these genetically engineered bacteria in vivo. FI and PAI have limited tissue penetration capability.¹⁵ Both RNI and MRI are not able to achieve real-time visualization.¹⁶ By contrast, ultrasound is a widely used imaging modality that has many advantages including non-invasiveness, non-ionizing radiation, high tissue penetration, and real-time imaging capability.¹⁷ Especially, when applied to activate the mild hyperthermia-responsive gene circuit, it is a significantly more suitable than other modalities for guiding the high-energy sound wave to precisely position deep-seated tumors, since it may reduce the mismatch between two different modalities (imaging guide and acoustic stimulation). Recently, an acoustic reporter gene (ARG1), which can produce gas vesicles (GVs) in bacteria, has been developed, endowing ultrasound with the capability to image these genetically engineered bacteria in deep-seated tumors.¹⁸

In this study, we developed a kind of genetically engineered bacteria through introducing two compatible plasmids in the tumor-targeting bacterial strain Escherichia coli MG1655, with one plasmid harboring the ARG1 gene and the other carrying the mild hyperthermia-responsive gene circuit (named Ec@IG). ARG1 was encoded by 13 Gvp genes, which can produce GVs as ultrasound contrast agents for imaging of these tumor-targeting bacteria. The mild hyperthermia-responsive circuit was constructed through inserting the therapeutic interferon (IFN)-γ gene under the leftward (pL) and rightward (pR) phage lambda promoters, which can respond to heat at mild hyperthermia (42°C–45°C) from high-intensity-focused acoustic waves, making it possible to spatiotemporally control the expression of therapeutic genes. In order to enhance the anti-tumor efficacy,¹⁹ we coated the doxorubicin (DOX) drugs onto the surface of genetically engineered bacteria with GVs (Ec@IG-GVs) by chemical linkage and obtained DOX-loaded Ec@IG-GVs (named Ec@DIG-GVs). After systematic administration of Ec@DIG-GV bacteria into the tumor-bearing mice, the position of Ec@DIG-GV bacteria in the tumor could be observed by ultrasound contrast imaging, providing the precise mapping for guiding the high-intensity-focused acoustic waves to irradiate these intratumoral engineered bacteria. Upon exposure to hyperthermia high-intensity focused ultrasound (hHIFU) irradiation at 42°C–45°C, the Ec@DIG-GV bacteria would express and secrete IFN-γ to kill tumor cells and stimulate the immune cells, including macrophage polarization from the M2 to the M1 phenotype and maturation of dendritic cells (DCs). Meanwhile, DOX can also be released from the Ec@DIG-GV bacteria due to the acidic tumor microenvironment leading to the killing of tumor cells and immunogenic cell death (ICD). The macrophage polarization and maturation of DCs greatly favor the presentation of tumor antigens from ICD and the activation of tumor-killing CD4⁺ and CD8⁺ T cells, achieving synergistic anti-tumor efficacy.

Results

Construction and characterization of Ec@DIG-GVs

We firstly constructed the genetically engineered bacteria through transforming two compatible plasmids in the tumor-targeting Escherichia coli MG1655 (named Ec@IG). Compared with N-(β-ketocaproyl)-l-homoserine lactone (AHL)-uninduced Ec@IG, numerous GVs could be observed in the AHL-induced Ec@IG-GVs under the phase contrast microscope and transmission electron microscope (Figures 1A and 1B). The presence of GVs enables them to float in the upper water layers after centrifugation at 350g for 3 h, greatly facilitating the isolation process of Ec@IG-GVs (Figures S1A and S1B). DOX was further conjugated onto the surface of Ec@IG-GVs, producing red fluorescent signals under a fluorescence microscope (Figure 1C). In addition, compared with the DOX-unloaded Ec@IG-GVs, Ec@DIG-GVs showed strong absorption peaks at 490 nm, similar to DOX (Figure 1D). The fluorescence absorption intensity gradually increased with the increase of DOX doses used for conjugation with Ec@IG-GVs (Figure S2A). Meanwhile, Figure 1E showed that the drug amount loaded onto bacteria reached about 20.6 μg when the DOX concentration was 1.8 mg/mL, with 93.3% bacteria loading with DOX (Figure S2B). Interestingly, the conjugation of DOX on the surface of Ec@IG-GVs did not result in significant cytotoxicity to these engineered bacteria (Figures S3A and S3B).

Next, we further detected the ultrasound contrast imaging performance in vitro. For these AHL-uninduced Ec@DIG, there were hardly detectable ultrasound contrast signals. By contrast, apparent contrast signals could be detected in the AHL-induced Ec@DIG-GVs, showing concentration-dependent signal enhancement (Figures 1F and 1G). Upon exposure to hHIFU irradiation, these contrast signals dramatically disappeared due to the collapse of GVs in these engineered bacteria (Figures S4A and S4B). In order to examine the feasibility that the ultrasound contrast imaging-guided hHIFU stimulation could activate the gene circuit to express exogenous genes, these engineered bacteria with GVs (named Ec@MG-GVs) were obtained through replacing IFN-γ with the mCherry gene and buried into the agar phantom. Apparent contrast signals could be seen only on the side of the agar phantom, which was buried with Ec@MG-GVs. Under the guide of ultrasound contrast image, the acoustic focus of hHIFU was positioned at these Ec@MG-GVs and irradiated for 25 min at 45°C. Figure 1H clearly shows that the contrast signals in the hHIFU-irradiated region significantly decreased. As expected, red fluorescence signals were observed only in the irradiated region (Figure 1I), confirming successful mCherry gene expression under ultrasound contrast imaging-guided hHIFU irradiation. Time-dependent expression levels of the mCherry gene were also confirmed in Ec@MG through exposing Ec@MG to 45°C hHIFU irradiation for different amounts of time, from 0 to 25 min, with the highest fluorescence signal intensity at 25 min hHIFU exposure (Figure S5). SDS-PAGE analysis further confirmed the successful expression of the IFN-γ gene in the two-plasmid Ec@DIG exposed to 45°C hHIFU exposure (Figure 1J). Interestingly, the concentrations of IFN-γ protein secreted from the 45°C-hHIFU-irradiated Ec@DIG gradually increased with the irradiation time, achieving 5.4-fold more than the non-irradiated control at 30 min irradiation (Figure 1K). No significant damage to their bacterial activity was observed (Figures S6A and S6B). The ability of DOX release at the acidic condition was examined through exposing Ec@DIG-GVs to pH 5.4–7.4 conditions. Figures 1L and 1M revealed that DOX could effectively release from Ec@DIG-GVs bacteria, with significantly higher DOX release at pH 5.4 and 6.5 than that at pH 7.4. The results were further confirmed in the cell uptake experiment through adding the equivalent released media into the 4T1 breast cancer cells, revealing that significantly stronger red fluorescence from DOX appeared in the 4T1 cells incubated with these released media at pH 5.4 (Figure 1N). These results indicated that the acidic environment was beneficial to releasing DOX from the surface of engineered bacteria.

Ec@DIG-GVs for tumor cell killing and activation of immune responsive in vitro

Next, the in vitro tumor cell-killing effect was tested by incubating the 4T1 cells with PBS (pH5.4) or equivalent bacterial supernatants from the acidically incubated Ec@G-GVs, Ec@DIG-GVs, Ec@IG-GVs (exposure to hHIFU for 25 min), or Ec@DIG-GVs (exposure to hHIFU for 25 min), followed by the live- and dead-cell staining assay. As shown in Figure 2A, numerous red dead cells could be seen in the tumor cells that were incubated with the supernatant of Ec@DIG-GVs treated with hHIFU and pH 5.4 relative to PBS control or the bacterial supernatants of Ec@G-GVs, Ec@DIG-GVs, and Ec@IG-GVs+hHIFU groups. The quantitative analysis of viability of 4T1 cells by CCK8 test exhibited that 64.2% tumor cells were killed in the Ec@DIG-GVs+hHIFU group, which is 6.42-, 5.11-, 1.95-, and 1.98-fold higher than PBS, Ec@G-GVs, Ec@DIG-GVs, and Ec@IG-GVs+hHIFU, respectively (Figure 2B), and the combination index of the Ec@DIG-GVs+hHIFU group was about 0.56, indicating a synergistic effect of IFN-γ in combination with DOX in treating 4T1 tumors (Table S1). Flow cytometry analysis confirmed that these killed tumor cells were mainly attributed to apoptosis (Figures 2C and 2D). All these results indicate that the presence of both IFN-γ protein and DOX may produce significantly stronger tumor cell-killing effects.

Cytotoxicity of bacterial supernatants to 4T1 tumor cells and activation of immune response *in vitro*

(A) Fluorescence microscopy images of 4T1 tumor cells stained with Calcein-AM (green fluorescence) and PI (red fluorescence). 4T1 tumor cells were exposed to different bacterial supernatants for 12 h. The equivalent bacterial supernatants were from Ec@G-GVs, Ec@DIG-GVs, Ec@IG-GVs (exposure to hHIFU for 25 min), or Ec@DIG-GVs (exposure to hHIFU for 25 min). All bacteria were incubated at pH 5.4 for 10 h, and PBS at pH 5.4 was used for the control. Scale bar: 100 μm.

(B) Cell viability of 4T1 cells treated with bacterial supernatants stated above (n = 4 per group).

(C) Tumor cell apoptosis analysis by flow cytometry after different treatments.

(D) Quantitative analysis of apoptosis rates of 4T1 cells (n = 3 per group).

(E) Fluorescence microscopy images of subcellular localization of CRT exposure and HMGB1 distribution of 4T1 cells treated with bacterial supernatants. Scale bar: 25 μm.

(F) Flow cytometry analysis of the CRT expression of positive 4T1 cells (n = 3 per group).

(G) ELISA detected the amount of HMGB1 released to the cell culture media by 4T1 cells treated with various bacterial supernatants for 12 h (n = 3 per group).

(H and I) Flow cytometry analysis and quantification of M1 phenotype macrophages (CD80⁺, MHC class II⁺) after the RAW264.7 macrophages were treated by different groups as stated above (n = 3 per group).

(J and K) Flow cytometry analysis and quantification the expression of F4/80⁺ and CD206⁺ on macrophages (n = 3 per group).

(L) Flow cytometry analysis of the proportion of mature DCs (CD80⁺CD86⁺) after incubation with bacterial supernatants for 24 h.

(M) Quantitative analysis of the level of mature DCs (n = 3 per group).

(N–Q) IL-6 (N), IL-1β (O), IL-12 (P), and TGF-β (Q) secreted by DCs were analyzed by ELISA after different treatments (n = 4 per group).

For (B)–(D) and (F)–(Q), data are presented as mean ± SD. Statistical analysis was implemented by using one-way analysis of variance with Tukey’s test. ns, not significant; ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, and ∗p < 0.05. (+) denotes treatment with ultrasound contrast imaging-guided hHIFU irradiation.

Apart from tumor cell-killing activity, IFN-γ and DOX have many immune-activating effects. IFN-γ can induce macrophage polarization from the M2 to the M1 phenotype and induce the maturation of DCs. There is evidence that DOX can induce the immunogenic death (ICD) of tumor cells, release damps, and promote the transfer of tumor-associated antigens to DCs.²⁰ This evidence inspires us to assume that the combination of IFN-γ with DOX could produce stronger anti-tumor immune responses. To test it, we first detected the expression levels of calreticulin (CRT) and high-mobility group box 1 (HMGB1) proteins, indispensable markers for ICD of tumor cells, on these treated tumors cells. Data from the immunofluorescence staining clearly demonstrated that strong fluorescence of CRT appeared on the membrane of tumor cells treated with the supernatants from the Ec@DIG-GVs and Ec@DIG-GVs+hHIFU groups, while hardly any fluorescence of HMGB1 appeared in their nuclei (Figure 2E). The proportions of CRT-positive cells and HMGB1 release in the Ec@DIG-GVs and Ec@DIG-GVs+hHIFU groups achieved were 4- and 3-fold higher than those of the control, Ec@G-GVs, and Ec@IG-GVs+hHIFU groups, respectively (Figures 2F and 2G). These results indicate that DOX, but not IFN-γ, plays a crucial role in inducing the ICD of tumor cells. Next, we further detected the polarization of macrophages and the maturation of DCs through incubating RAW264.7 macrophages and 2.4 DCs with bacterial supernatants for 24 h. The results demonstrated that the CD80⁺/major histocompatibility complex class II (MHC class II)⁺ M1 macrophages significantly increased, but CD206⁺ M2 macrophages dramatically decreased, in the Ec@DIG-GVs+hHIFU group (Figures 2H–2K). Significantly more CD80⁺/CD86⁺ DCs appeared in the Ec@DIG-GVs+hHIFU group, revealing a stronger effect on promoting the maturation of DCs when they were exposed to the supernatants containing IFN-γ and DOX (Figures 2L and 2M). Moreover, immune factors secreted by DCs were further evaluated by ELISA kits, showing higher levels of interleukin (IL)-6, IL-1β, and IL-12 and lower levels of transforming growth factor β (TGF-β) in the cell culture medium of the Ec@DIG-GVs+hHIFU group (Figures 2N–2Q). These results provide strong evidence that IFN-γ and DOX from acidically incubated Ec@DIG-GVs+hHIFU have a powerful function in activating immune responses.

Local chemo-immunotherapy to active immune response in 4T1 subcutaneous tumor

To explore the feasibility of ultrasound contrast imaging-guided hHIFU irradiation to induce the exogenous gene expression of genetically engineered bacteria in vivo, we intratumorally injected the DiO-labeled Ec@MG or Ec@MG-GVs into the 4T1 tumor. Significantly enhanced acoustic signals could be observed in the tumor received with Ec@MG-GVs but not those treated with Ec@MG, which is attributed to the presence of GVs in the Ec@MG-GVs (Figures 3A and 3B). Upon receiving hHIFU irradiation under ultrasound contrast imaging guidance, these enhanced acoustic signals in the irradiated region disappeared due to the collapse of GVs in the Ec@MG-GVs, giving clear feedback information (Figure 3C). To examine the expression of the mCherry gene, the tumors that received Ec@MG-GVs+hHIFU were removed and observed by the IVIS. Figure 3D clearly shows that only the part of the tumor that received hHIFU irradiation produced fluorescence signals. Further observation of tumor sections by fluorescence microscopy revealed that obvious red fluorescence signals of mCherry proteins appeared in the left irradiated tumor but not in the right non-irradiated tumor (Figures 3E and 3F). Meanwhile, the concentrations of IFN-γ in tumor were higher in the Ec@DIG-GVs+hHIFU group than in the Ec@DIG-GVs group (Figure S7), confirming the feasibility of precisely activating the expression of target genes by ultrasound contrast imaging-guided hHIFU irradiation.

*In vivo* anti-tumor and immune activation effects after intratumoral injection of Ec@DIG-GVs combined with hHIFU

(A and B) Ultrasound imaging of tumor before and after intratumoral injection of Ec@MG or Ec@MG-GVs. The white circles represents the tumor area. Scale bar: 2 mm.

(C) Ultrasound imaging of tumor that was injected with Ec@MG-GVs and then received hHIFU irradiation. The white circles represent the tumor area, and the red circle represents the radiation area. Scale bar: 2 mm.

(D) IVIS imaging of mCherry expression in tumor injected with Ec@MG-GVs before and after ultrasound contrast imaging-guided hHIFU irradiation. Scale bar: 4 mm.

(E and F) Fluorescent imaging of the tumor sections that either received hHIFU irradiation or not after Ec@MG-GV injection. The white circles represent that the tumoral areas were affected by acoustic wave. The nuclei were stained by DAPI, bacteria were labeled by DiO dye, and the red fluorescence represents the production of the mCherry protein. Scale bar: 100 μm.

(G) Schematic illustration of tumor treatment procedure of Ec@DIG-GVs combined with ultrasound contrast imaging-guided hHIFU irradiation.

(H) Tumor growth curves of tumor-bearing mice treated with control, Ec@G-GVs+hHIFU, Ec@DIG-GVs, Ec@IG-GVs+hHIFU, and Ec@DIG-GVs+hHIFU (n = 7 per group, two-way analysis of variance with Tukey’s test). Data are presented as mean ± SD.

(I) Survival curves of mice in different treatment groups (n = 7 per group).

(J) H&E, TUNEL, and Ki67 staining of tumor tissues from different treatment groups. Scale bar: 50 μm.

(K and L) Flow cytometric analysis and relative quantification of mature DCs (CD80⁺CD86⁺) in the treated tumors (n = 3 per group).

(M–P) Flow cytometric analysis and relative quantification of CD4⁺ T cells (CD3⁺CD4⁺) and CD8⁺ T cells (CD3⁺CD8⁺) (n = 3 per group).

(Q–T) Flow cytometric analysis and quantification of the proportion of Tim3⁺CD4⁺CD8⁺ (Q), TNF-α⁺CD4⁺CD8⁺ (R), IFN-γ⁺CD4⁺CD8⁺ (S), and GZB⁺CD4⁺CD8⁺ (T) in the treated tumors (n = 3 per group).

For (K)–(T), data are presented as mean ± SD. Statistical analysis was implemented by using one-way analysis of variance with Tukey’s test. ns, not significant; ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, and ∗p < 0.05. (+) denotes treatment with ultrasound contrast imaging-guided hHIFU irradiation.

Next, the anti-tumor efficacy of Ec@DIG-GVs combined with ultrasound contrast imaging-guided hHIFU irradiation was investigated by using the 4T1 subcutaneous tumor model. Figure 3G presents the treatment scheme showing the tumors receiving once-weekly injections of PBS, Ec@G-GVs+hHIFU, Ec@DIG-GVs, Ec@IG-GVs+hHIFU, or Ec@DIG-GVs+hHIFU, respectively. Ultrasound contrast imaging-guided hHIFU at 45°C was used for 25 min to make sure all these bacteria were irradiated. Figures 3H and S8 clearly demonstrate that a significantly stronger anti-tumor effect was seen in the Ec@DIG-GVs+hHIFU group. Compared to other groups, the tumor-bearing mice in the Ec@DIG-GVs+hHIFU group achieved the longest mean survival time of more than 53 days (Figure 3I). Notably, Ec@DIG-GVs+hHIFU produced a significantly better tumor growth inhibition effect than the Ec@DIG-GVs and Ec@IG-GVs+hHIFU groups, which may contribute to the synergetic anti-tumor effects of IFN-γ and DOX. H&E histochemical assays revealed that the obvious necrosis could be observed in the Ec@DIG-GVs+hHIFU group, and numerous TUNEL-positive cells but fewer Ki67-positive tumor cells could be observed in the tumors treated with Ec@DIG-GVs+hHIFU (Figure 3J), indicating that Ec@DIG-GVs combined with ultrasound contrast imaging-guided hHIFU irradiation could greatly suppress tumor growth and result in greater apoptosis/necrosis of tumor cells. No significant weight loss was found in these treated mice (Figure S9).

To investigate the underlying mechanisms of Ec@DIG-GVs+hHIFU in the anti-tumor immune response, we prepared single-cell suspensions of tumors and analyzed the changes of immune cells by flow cytometry. As shown in Figures 3K, 3L, and S10A, the percentage of mature DCs (CD80⁺CD86⁺) significantly increased in the Ec@DIG-GVs+hHIFU group in comparison to the control, Ec@G-GVs+hHIFU, Ec@DIG-GVs, and Ec@IG-GVs+hHIFU groups. The highest level of M1 phenotype macrophages (CD80⁺, 48.9%) and the lowest level of M2 phenotype macrophages (CD206⁺, 8.65%) also appeared in the Ec@DIG-GVs+hHIFU group (Figures S10B and S11A–S11D), showing macrophage polarization from the M2 to the M1 phenotype and the maturation of DCs, which would greatly facilitate tumor antigen presentation and elimination. Moreover, the number of helper T cells (CD3⁺CD4⁺) and cytotoxic T cells (CD3⁺CD8⁺) substantially increased, achieving 30.0% and 34.8% proportion, respectively (Figures 3M–3P). Consistently, the regulatory T cells (CD4⁺FOXP3⁺) in the Ec@DIG-GVs+hHIFU group were significantly fewer than other groups (Figures S10C, S11E, and S11F). Remarkably, fewer Tim3⁺CD4⁺ and Tim3⁺CD8⁺ T cells, indicating less T cell exhaustion, were found in the Ec@DIG-GVs+hHIFU group relative to the other groups (Figures 3Q and S10C). We also confirmed the activation of T cells, revealing a significantly higher proportion of tumor necrosis factor α (TNF-α)⁺CD4⁺ and TNF-α⁺CD8⁺ T cells (Figures 3R and S10C), IFN-γ⁺CD4⁺ and IFN-γ⁺CD8⁺ T cells (Figures 3S and S10C), and GZB⁺CD4⁺ and GZB⁺CD8⁺ T cells (Figures 3T and S10C) in the Ec@DIG-GVs+hHIFU group. Collectively, these results proved that Ec@DIG-GVs+hHIFU could remodel the tumor immunosuppressive microenvironment and boost the anti-tumor immune responses.

In vivo ultrasound tracking of Ec@DIG-GV bacteria homing into tumors

Since GVs endow Ec@DIG-GVs with ultrasound contrast imaging capability, we wonder whether ultrasound can image and track their systemic homing of Ec@DIG-GV bacteria into tumors. Before testing this, we first evaluated their tumor-targeting performance; live or heat-killed indocyanine green N-succinimidyl ester (ICG)-labeled Ec@IG-GVs (Ec@IIG-GVs) bacteria were intravenously administrated into 4T1 tumor-bearing mice and observed by IVIS at different times. As shown in Figures 4A and 4B, gradually enhanced fluorescence signals appeared in the tumor of the mice that received live Ec@IIG-GVs but not the mice that received dead Ec@IIG-GV bacteria or free ICG. To further confirm this, we collected the major organs and tumors that received live or dead Ec@IIG-GV bacteria and homogenized them, followed by plating the tissue homogenate on Luria-Bertani agar plates with 100 μg/mL ampicillin and 50 μg/mL kanamycin for 24 h. The result demonstrated that there were the most bacterial colonies in the tumors of the mice that received live Ec@IIG-GV bacteria, while barely any bacteria grew on the plates for the tumors that received dead Ec@IIG-GV bacteria (Figures 4C and 4D), confirming that these engineered bacteria had excellent tumor-targeting performance.

*In vivo* tumor homing of Ec@DIG-GVs after intravenous injection

(A) IVIS imaging of tumor-bearing mice that received systemic administration of live or dead ICG-labeled Ec@IG-GV (Ec@IIG-GV) bacteria or free ICG (n = 3 per group).

(B) IVIS imaging of tumors and major organs from the tumor-bearing mice treated as stated above after 48 h (n = 3 per group).

(C and D) Representative photographs of solid agar plates and quantification of bacterial colonization in tumors and major organs as stated above (n = 3 per group, one-way analysis of variance with Tukey’s test).

(E) Ultrasound imaging of Ec@DIG or Ec@DIG-GVs in tumor at 0, 4, 12, 24, 36, and 48 h after intravenous injection. The white circles represent tumor areas (n = 3 for each time point). Scale bar: 2 mm.

(F) Fluorescent imaging of the tumor and major organ sections after intravenous injection of live Ec@DIG-GVs, dead Ec@DIG-GVs, and free DOX at 48 h. Scale bar: 50 μm.

(G) Immunofluorescence staining of tumor ischemic area to locate live Ec@DIG-GVs, dead Ec@DIG-GVs, and free DOX. Red fluorescence represents DOX. Blue fluorescence represents the nuclei of tumor cells stained by DAPI. Green fluorescence represents HIF-1α highly expressed cells. Scale bar: 50 μm.

(H) Detection of the concentrations of DOX in tumor tissue homogenate from different groups (n = 3 per group, one-way analysis of variance with Tukey’s test).

(I) The concentrations of IFN-γ protein in tumor tissue were tested by ELISA kit after 12, 24, and 48 h intravenous injection of Ec@DIG-GVs combined with or without hHIFU irradiation (n = 3 per group, two-way analysis of variance with Sidak’s test).

For (D), (H), and (I), data are presented as mean ± SD. ∗∗∗∗p < 0.0001 and ∗∗∗p < 0.01. (+) denotes treatment with ultrasound contrast imaging-guided hHIFU irradiation.

Next, we tested the feasibility of in vivo ultrasound tracking of Ec@DIG-GV bacteria homing into tumors. The tumor-bearing mice were intravenously injected with Ec@DIG or Ec@DIG-GV bacteria, and we imaged the bacteria in the tumors at different times. The results demonstrated that there were no detectable acoustic contrast signals in the tumors that received Ec@DIG bacteria due to the absence of GVs. By contrast, apparent acoustic contrast signals were seen in the tumors that received Ec@DIG-GV bacteria after 12 h. As time went on, these acoustic signals in the tumors gradually became stronger and more concentrated in the tumor center (Figures 4E and S12). These results were consistent with FI data, proving that the tumor homing of Ec@DIG-GV bacteria could be tracked by ultrasound contrast imaging.

To confirm the tumor-targeting delivery of DOX conjugated on the surface of Ec@DIG-GV bacteria, we systemically administrated free DOX or live or heat-killed Ec@DIG-GV bacteria with equivalent DOX into the tumor-bearing mice. We can see in Figure 4F that lots of red fluorescence signals of DOX could be observed in the tumor that received live Ec@DIG-GV bacteria, but there were only with a few DOX signals in the liver and barely any DOX signals in the heart, spleen, lung, and kidney. By contrast, there were more DOX signals in the liver and less in the tumor for the mice that received dead Ec@DIG-GV bacteria. Free DOX was distributed in all main organs including the heart, where DOX causes life-threatening cardiomyopathy. These results mainly contribute to the ability of live Ec@DIG-GV bacteria to home into the hypoxic areas of the tumor (Figure 4G). Furthermore, for the live Ec@DIG-GVs group, the concentration of DOX in the tumor was 2.5 times more than those of the dead Ec@DIG-GVs group and the free DOX group (Figure 4H). Moreover, the levels of IFN-γ were detected by ELISA kit at 12, 24, or 48 h after hHIFU irradiation, and the results showed that the Ec@DIG-GVs+hHIFU group produced more IFN-γ protein in the tumor than the Ec@DIG-GVs group (Figure 4I).

Systemic administration of Ec@DIG-GVs combined with hHIFU mediated anti-tumor chemo-immunotherapy

Inspired by the excellent tumor-targeting, drug delivery, and ultrasound contrast imaging abilities of Ec@DIG-GVs, we further investigated their anti-tumor efficacy by systemic administration of Ec@DIG-GVs combined with ultrasound contrast imaging-guided hHIFU irradiation. The tumor treatment procedure is illustrated in Figure 5A. The results showed that the Ec@DIG-GVs+hHIFU group produced the strongest anti-tumor efficacy, with only a 271 mm³ average tumor volume after 21 days. The limited anti-tumor effect could be observed in the mice treated with Ec@DIG-GVs or in the Ec@IG-GVs+hHIFU groups, while the PBS control and Ec@G-GVs+hHIFU groups reached 1,421.9 and 1,299 mm³, respectively (Figures 5B and S13). The treatment of Ec@DIG-GVs+hHIFU remarkably prolonged the survival time of tumor-bearing mice in comparison with other groups (Figure 5C). The immunofluorescence staining assays revealed that there were significantly more apoptotic cells but less Ki67-positive tumor cells in the Ec@DIG-GVs+hHIFU group than other groups (Figures 5D and 5E). Interestingly, obviously less pulmonary nodules were also observed in the Ec@DIG-GVs+hHIFU group (Figures S14A and S14B), indicating that systemic administration of Ec@DIG-GVs combined with ultrasound contrast imaging-guided hHIFU irradiation could exert a prominent function of preventing pulmonary metastasis. To further understand synergistic anti-tumor mechanisms, we first detected the dominant biomarkers of CRT and HMGB1 (ICD markers) in these treated tumors. The data from immunofluorescence staining revealed significantly more CRT and less HMGB1 immunofluorescent signals in the Ec@DIG-GVs and Ec@DIG-GVs+hHIFU groups (Figures 5F and 5G), indicating that DOX, but not IFN-γ, accelerated the production of CRT and the release of HMGB1 in these tumor cells. Next, we also examined the maturation of DCs in the tumors by flow cytometry, revealing 46.9% mature CD80⁺CD86⁺ DCs in the tumor treated with Ec@DIG-GVs+hHIFU, which is significantly higher than the Ec@DIG-GVs (24.3%) and Ec@IG-GVs+hHIFU (36.4%) groups (Figures 5H, 5I, and S15A). The ratio of M1 (F4/80⁺CD80⁺) and M2 (F4/80⁺CD206⁺) macrophages changed dramatically in Ec@DIG-GVs+hHIFU-treated tumors, with the proportion of M1 macrophages increasing to 42.2% but the proportion of M2 macrophages decreasing to 14.0%, versus 16.8% M1 macrophages and 35.3% M2 macrophages in the control group (Figures 5J–5M and S15B). Collectively, these results supported that Ec@DIG-GVs+hHIFU could substantially elicit ICD of tumor cells and promote the maturation of DCs and macrophage polarization from the M2 to the M1 phenotype, making it possible to produce more tumor-associated antigens and more effective antigen presentation.

*In vivo* anti-tumor and immune response after systemic administration of Ec@DIG-GVs combined with hHIFU

(A) Schematic illustration of tumor treatment procedure through systemic administration of Ec@DIG-GVs combined with ultrasound contrast imaging-guided hHIFU.

(B) Tumor growth curves of tumor-bearing mice treated with different treatments (n = 7 per group, two-way analysis of variance with Tukey’s test). Data are presented as mean ± SD.

(C) Survival time of tumor-bearing mice in each group (n = 7 per group).

(D and E) Tumor sections from different treatment groups were TUNEL stained (green fluorescence) and Ki67 stained (red fluorescence), respectively. Cell nuclei were labeled by DAPI. Scale bar: 100 μm.

(F and G) Immunofluorescence staining of tumor slices to reveal the level of CRT exposure and HMGB1 distribution. Blue fluorescence stands for nuclei. Green fluorescence stands for CRT or HMGB1 protein. Scale bar: 50 μm.

(H and I) Flow cytometry analysis and quantification of the proportion of mature DCs (CD80⁺CD86⁺) in the tumors at 7 days after treatment.

(J–M) Flow cytometry analysis and quantification of the ratios of M1 (F4/80⁺CD80⁺) and M2 (F4/80⁺CD206⁺) macrophages in the tumors.

(N and O) The percentages of CD4⁺ T cells (CD3⁺CD4⁺) (N) and CD8⁺ T cells (CD3⁺CD8⁺) (O) in the treated tumors.

(P) The percentage of suppressor T cells (CD4⁺FOXP3⁺) in the tumors.

(Q) Flow cytometry quantification of PD-1⁺ expression in T cells.

(R) The expression of MHC class I on the tumor cells after being treated with different groups.

(S–V) IL-1β (S), IL-12 (T), IL-6 (U), and TGF-β (V) in the tumors on the 7th day post-treatment (n = 4 per group, one-way analysis of variance with Tukey’s test). Data are presented as mean ± SD. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, and ∗p < 0.05. (+) denotes treatment with ultrasound contrast imaging-guided hHIFU irradiation.

For (H)–(R), n = 3 per group; data are presented as mean ± SD. Statistical analysis was implemented by using one-way analysis of variance with Tukey’s test.

DCs, as the critical players in the initiation and development of immune response, can induce the differentiation of naive T cells into helper and effector T cells, which allows DCs to effectively process antigen, strongly express costimulatory molecules, secrete cytokines, and migrate to tissues or lymphoid organs to prime T cells.²¹ In order to confirm this, we first analyzed the primary cytotoxic T lymphocytes (CD8⁺ T cells) and helper T cells (CD4⁺ T cells) in these treated tumors. As shown in Figures 5N and 5O, there were significantly increased CD3⁺CD4⁺ T cells and CD3⁺CD8⁺ T cells in the tumors treated with Ec@DIG-GVs+hHIFU relative to the other groups. The proportion of suppressor T cells significantly decreased in the tumors treated with Ec@DIG-GVs+hHIFU, as shown in Figure 5P. Notably, significantly upregulated expression of PD-1 on T cells and MHC class I on tumor cells was also observed in Ec@DIG-GVs+hHIFU-treated tumors (Figures 5Q and 5R), increasing the ability of immune cells to recognize tumor cells. Moreover, the number of GZB⁺CD4⁺CD8⁺ and IFN-γ⁺CD4⁺CD8⁺ T cells significantly increased and Tim-3⁺CD4⁺CD8⁺ T cells largely decreased in the tumors treated with Ec@DIG-GVs+hHIFU (Figures S15C and S16A–S16L), thus producing obviously higher levels of IL-1β, IL-12, and IL-6 cytokines (Figures 5S–5U). Interestingly, the level of TGF-β significantly decreased in the tumors after treatment with Ec@DIG-GVs+hHIFU in comparison with the other groups (Figure 5V). Collectively, these data indicated that treatment with Ec@DIG-GVs combined with hHIFU not only activated immune responses in tumor but also abrogated those negative immune inhibitors, which could establish a stable tumor immunity niche and achieve an effective anti-tumor effect.

Next, the immune microenvironment in the spleen was also detected to clarify the mechanism of inhibition effects against pulmonary metastasis of tumors. Flow cytometry analysis showed the increased proration of M1 phenotype macrophages (CD80⁺) and decreased M2 phenotype macrophages (CD206⁺) in the spleen of tumor-bearing mice treated with Ec@DIG-GVs+hHIFU, with 25.4% M1 phenotype macrophages and 11.9% M2 phenotype macrophages (Figures S17A and S18A–S18D). The proliferative CD4⁺ and CD8⁺ T cell populations increased 2.74 and 2.84 times higher than the control group after the treatment of Ec@DIG-GVs+hHIFU, reaching 29.0% and 30.5% cell ratios, respectively (Figures S18E–S18H). Interestingly, as shown in Figures S17B and S18I–S18L, the number of Tim-3⁺CD4⁺ and Tim-3⁺CD8⁺ T cell populations drastically decreased, which indicates that less T cells were fatigued. All data indicated that a strong immune response had developed in the spleen of tumor-bearing mice by the treatment of Ec@DIG-GVs+hHIFU. Next, the microenvironment of tumor-draining lymph nodes was further examined, revealing that the proportions of activated macrophages, T cells, and mature DCs were markedly increased in the Ec@DIG-GVs+hHIFU group (Figures S19A–S19C and S20A–S20H). Notably, no obvious side effects, including body weight of mice (Figure S21); histopathologic feature of major organs (Figure S22); blood routine indices of white blood cells, lymphocytes, monocytes, and neutrophilic granulocytes; or the blood biochemical indicators of alanine transaminase, aspartate transaminase, and γ-glutamyl transpeptidase, were observed in the Ec@DIG-GVs+hHIFU group, similar to the other groups (Figures S23A and S23B).

PD-L1 blockade augmented the anti-tumor efficacy of Ec@DIG-GVs+hHIFU through activating long-term immune memory

Immunological memory is an important hallmark of adaptive immunity, which could provide efficient protection to prevent previous pathogens from attacking organisms. PD-1/PD-L1 checkpoint blockade, combined with other treatment strategies, has achieved some success in tumor treatment. Considering that treatment with Ec@DIG-GVs+hHIFU could significantly increase the expression of PD-1 on T cells, we further assumed that PD-L1 blockade may enhance the anti-tumor efficacy of Ec@DIG-GVs+hHIFU through activating the long-term immune memory in tumor. To test this, we established 4T1 tumor-bearing mice and randomly divided them into four groups: the PBS control group, anti-PD-L1 (aPD-L1) group, Ec@DIG-GVs+hHIFU group, and Ec@DIG-GVs+hHIFU+aPD-L1 group. The Ec@DIG-GV bacteria were intravenously administrated into the tumor-bearing mice (day 0), followed by hHIFU irradiation after 48 h (day 2). The aPD-L1 antibodies were systemically administrated at 48 h post-hHIFU irradiation (day 4), and the total treatment procedure was repeated once after 7 days (day 7). After 31 days, the secondary 4T1 tumors were inoculated in the mice treated with Ec@DIG-GVs+hHIFU+aPD-L1 to evaluate the rechallenged tumor growth (Figure 6A). We can clearly see in Figure 6B that aPD-L1 alone did not produce an obvious anti-tumor effect, similar to the control group, leading to the fast death of the tumor-bearing mice. Treatment with Ec@DIG-GVs+hHIFU could produce significant anti-tumor effects in the first 21 days, but recurrence could be observed in the primary tumors. Notably, the primary tumors were completely eradicated when using Ec@DIG-GVs+hHIFU combined with aPD-L1. No abnormal weight change was founded in the mice treated with Ec@DIG-GVs+hHIFU+aPD-L1 (Figure 6C).

PD-L1 blockade enhanced anti-tumor efficacy and augmented immune memory

(A) Schematic illustration of Ec@DIG-GVs+hHIFU in combination with anti-PD-L1 (aPD-L1) for inhibiting primary and rechallenged tumor growth.

(B) Growth curves of primary tumors with different treatment strategies (n = 8 per group, two-way analysis of variance with Sidak’s test).

(C) The weight changes of 4T1 tumor-bearing mice during treatment process (n = 8 per group).

(D) Photographs of tumor growth on the 20th day after the inoculation of secondary tumors.

(E) Growth curves of rechallenged tumors in each group (n = 5 per group, two-way analysis of variance with Sidak’s test).

(F) Survival time of the treated mice (n = 5 per group).

(G) Photographs and H&E staining (scale bar: 4 mm) of tumor nodules in lungs of naive mice and mice treated with Ec@DIG-GVs+hHIFU+aPD-L1. Black circle: metastatic nodules.

(H) Quantification analysis of tumor nodules in the lungs in the different treatment groups (n = 5 per group, t test).

(I–L) Flow cytometry analysis and quantification of the ratios of central memory T cells (Tcms; CD62L⁺CD44⁺) and effector memory T cells (Tems; CD62L⁻CD44⁺) in the spleen of mice that received rechallenged tumors. Tcms and Tems were gated from CD4⁺ (I and J) and CD8⁺ (K and L) T cells (n = 3 per group, one-way analysis of variance with Tukey’s test).

(M–O) Detection of the levels of IL-6 (M), IL-12 (N), and IL-1β (O) in serum from mice rechallenged with secondary tumors after 7 days (n = 4 per group, one-way analysis of variance with Sidak’s test).

For (B), (C), (E), and (I)–(O), data are presented as mean ± S.D. ns, not significant; ∗∗∗∗p < 0.0001, ∗∗p < 0.01, and ∗p < 0.05. (+) denotes treatment with ultrasound contrast imaging-guided hHIFU irradiation.

Next, we rechallenged the mice treated with Ec@DIG-GVs+hHIFU+aPD-L1 by inoculating the 4T1 tumor cells on another side on day 31, with the naive mice as the control. The results showed that all naive mice had rapid tumor progression. By contrast, 60% of mice did not show tumor growth after tumor rechallenge with 4T1 cells in the Ec@DIG-GVs+hHIFU+aPD-L1 group (Figures 6D and 6E), achieving over 120 days of survival (Figure 6F). As shown in Figures 6G and 6H, numerous pulmonary nodules were observed in naive mice, but there were hardly any pulmonary metastatic nodules in the Ec@DIG-GVs+hHIFU+aPD-L1 group. These outcomes illustrated that the treatment with Ec@DIG-GVs+hHIFU+aPD-L1 could provide long-term immune protective effects against tumor recurrence and metastasis.

To explore the underlying mechanisms of Ec@DIG-GVs+hHIFU+aPD-L1 in activating the anti-tumor immune memory, the spleens and tumors of treated mice were collected to determine the proportion of central memory T cells (Tcms) and effector memory T cells (Tems) after 14 days post-secondary tumor inoculation. Interestingly, compared to the naive group, an obvious shift of the Tcm phenotype (CD62L⁺CD44⁺) to the Tem phenotype (CD62L⁻CD44⁺) was found in the CD4⁺ and CD8⁺ T cells of spleen after treatment with Ec@DIG-GVs+hHIFU+aPD-L1 (Figures 6I–6L and S24A). Similar results were observed in these Ec@DIG-GVs+hHIFU+aPD-L-treated tumors (Figures S24B and S25A–S25D). The levels of IL-6, IL-12, and IL-1β, important cytokines against tumors, remarkably increased in the mice treated with Ec@DIG-GVs+hHIFU+aPD-L1 (Figures 6M–6O). Tcms are mainly located in secondary lymphoid organs, which can provide an immune response only when they experience complex biologic evolutions. By contrast, Tems can exert strong immune functions by secreting various anti-tumor cytokines when they are stimulated by antigens.²² Since more Tems could be produced in the mice that received Ec@DIG-GVs+hHIFU+aPD-L1, it is reasonable for these mice to produce long-term anti-tumor immune effects. All these results proved that Ec@DIG-GVs+hHIFU+aPD-L1 could produce long-term anti-tumor effects by establishing long-term immune memory in these mice.

Discussion

With increasing evidence of tumor-colonizing bacteria, synthetic biology tools are being leveraged to repurpose bacteria as a tumor-targeting delivery system. Bacteria can be genetically engineered to encode toxins, immunomodulators, pro-drug-converting enzymes, or nanobodies and chemically modified with small-molecule drugs or nanoparticles, helping these payloads get into the tumor with low side effects.²³^,²⁴^,²⁵ By using a combination of bioactive molecule production and local payload release, these engineered bacteria can modulate the tumor microenvironment to improve their safety and efficacy. Evidence demonstrated that genetic circuits combined with physical stimulations such as light,⁶^,²⁶ magnetic field,²⁷ and focused ultrasound¹⁰ can enhance spatial and temporal control of therapeutic bacteria, enabling the precise tuning of this bacterial gene expression or drug release. However, it is still difficult for real-time tracking of these engineered bacteria to execute imaging-guided remote physical stimulation. Recently, the development of acoustic reporter genes has made it possible to image these engineered bacteria in vivo.²⁸

In this study, we developed the ultrasound-visible engineered bacteria (Ec@DIG-GVs) through expressing acoustic reporter genes (GVs) in the tumor-targeting bacteria, which were genetically engineered with a mild hyperthermia-responsive gene circuit harboring the therapeutic IFN-γ gene and chemically modified with DOX on their surface. The presence of GVs makes it possible to track Ec@DIG-GVs by ultrasound contrast imaging in a real-time manner, giving the spatiotemporal information in the tumor and thus guiding the operator to precisely position the sound wave of hHIFU at these engineered bacteria. More importantly, upon exposure to hHIFU irradiation, the contrast signals would disappear due to the collapse of GVs in the high acoustic pressure, producing immediate feedback for the operator. Notably, only 25 min hHIFU irradiation at 42°C–45°C was enough to induce the expression of the IFN-γ gene and produce therapeutic effects, which greatly improves safety in future clinic applications. In fact, ultrasound imaging has been proposed to guide HIFU treatments and has shown promise in various organs at preclinical and clinical levels.²⁹^,³⁰^,³¹^,³²^,³³ Compared to MRI-guided HIFU, ultrasound-guided HIFU has significantly more advantages, including offering practicalities in anesthesia and considerable cost savings.

It is important to confine the expression of the IFN-γ gene and the delivery of DOX to the local tumor since both IFN-γ and DOX may bring about obvious side effects when they are systemically administrated. IFN-γ is a very potent proinflammatory cytokine and plays a key role in the activation of cellular immunity and stimulation of the anti-tumor immune response. However, its short half-life and undesirable systemic side effects clearly limit clinical use of IFN-γ.³⁴ DOX is a powerful anti-tumor anthracycline drug, but its clinical use is limited due to the side effect of cardiotoxicity.³⁵^,³⁶ In this study, we used the tumor-targeting bacteria E. coli MG1655, with a mild hyperthermia-responsive IFN-γ gene circuit in the cytoplasm and DOX chemically conjugated on the bacterial surface, achieving effective intratumoral delivery of the IFN-γ gene and DOX payloads. Upon exposure to ultrasound contrast imaging-guided hHIFU irradiation, IFN-γ could be locally expressed in the tumor due to the mild hyperthermia-responsive gene circuit harboring the therapeutic IFN-γ gene under the leftward (pL) and rightward (pR) phage lambda promoters, which can respond to heat at 42°C–45°C from high-intensity-focused acoustic waves. Meanwhile, the acidic tumor microenvironment also greatly promotes the release of DOX from the bacterial surface thanks to acid-labile linkers of cis-aconitic anhydride. Under the optimized conjugation process, bacteria maintain around 87.1% of the cell viability after drug conjugation. The in vitro drug release indicates that DOX-bacteria conjugates were stable under the pH 7.4 physiological condition, whereas DOX could be rapidly released in pH 6.5 and 5.4 acidic conditions, indicating the local release capability of DOX in the tumor.

Recently, more and more evidence has demonstrated that DOX can also tackle tumors through the induction of ICD in tumor cells, releasing damps and promoting the transfer of tumor-associated antigens to DCs.³⁷^,³⁸ Numerous literature show that IFN-γ can induce macrophage polarization from the M2 to the M1 phenotype and induce the maturation of DCs,³⁹^,⁴⁰ which can improve the ability to recognize and present tumor-associated antigens from ICD. It is reasonable for us to speculate that IFN-γ and DOX can probably produce synergistic anti-tumor effects. Our results showed that Ec@DIG-GVs+hHIFU could substantially elicit ICD of tumor cells and promote the maturation of DCs, activating anti-tumor immune responses. Notably, significantly upregulated expression of PD-1 on T cells and MHC class I on tumor cells was also observed in Ec@DIG-GVs+hHIFU-treated tumors. PD-L1 blockade may further enhance the anti-tumor efficacy of Ec@DIG-GVs+hHIFU by activating the long-term immunological memory functions.

In summary, we developed ultrasound-visible engineered bacteria genetically modified with two compatible plasmids (one for the ARG1 gene and another for the mild hyperthermia-responsive IFN-γ gene circuit) and chemically modified with DOX on the surface of these bacteria. The ARG1 gene can produce GVs, making these bacteria visible by ultrasound imaging. The ultrasound-visible engineered bacteria may help precisely position the sound wave from hHIFU to activate the mild hyperthermia-responsive IFN-γ gene circuit when they target into the tumor, expressing IFN-γ to induce macrophage polarization from the M2 to the M1 phenotype and inducing the maturation of DCs. Meanwhile, DOX can be released in the acidic tumor microenvironment, killing the tumor cells and inducing the ICD of tumor cells, which promotes the transfer of tumor-associated antigens to DCs. Both IFN-γ and DOX collectively enhanced the activation anti-tumor immune responses and achieved synergistic anti-tumor effects. Our study provides a novel strategy for bacteria-based anti-tumor chemo-immunotherapy through remotely controllable gene expression by ultrasound-guided hHIFU and intratumoral delivery of DOX by using genetically and chemically engineered bacteria as carriers.

Limitations of the study

Our initial assessment is that Ec@DIG-GVs have negligible side effects in the treatment process. However, further toxicological analysis is important for the comprehensive test of toxicity effects of Ec@DIG-GVs. It is necessary to improve the compatibility of the two plasmids within Ec@DIG-GVs for increasing the expression of therapeutic proteins. Thus, a smaller bacterial dose can be used to ensure safety. Furthermore, future studies are needed to examine the anti-tumor efficacy of Ec@DIG-GVs in the patient-derived xenograft model, which will provide more valuable experience for clinical translation.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies

CRT antibody	Beyotime Biotechnology Co.(Shanghai, China)	Cat#AF1666
HMGB1 antibody	Beyotime Biotechnology Co.(Shanghai, China)	Cat#AF0180
Anti-mouse PD-L1 antibody	Biolegend	Cat#124338; RRID: AB_2800599
Anti-mouse CD45 antibody	Biolegend	Cat#103154; RRID: AB_2572115
Anti-mouse CD3 antibody	Biolegend	Cat#100204; RRID: AB_312660
Anti-mouse CD4 antibody	Biolegend	Cat#100429; RRID: AB_493698
Anti-mouse CD8 antibody	Biolegend	Cat#100722; RRID: AB_312760
Anti-mouse CD11b antibody	Biolegend	Cat#101205; RRID: AB_312788
Anti-mouse Tim-3 antibody	Biolegend	Cat#119703; RRID: AB_345378
Anti-mouse F4/80 antibody	Biolegend	Cat#123127; RRID: AB_893496
Anti-mouse CD11c antibody	Biolegend	Cat#117307; RRID: AB_313776
Anti-mouse CD80 antibody	Biolegend	Cat#104713; RRID: AB_313135
Anti-mouse CD86 antibody	Biolegend	Cat#105028; RRID: AB_2074994
Anti-mouse MHCI antibody	Biolegend	Cat#114613; RRID: AB_2750194
Anti-mouse CD206 antibody	Biolegend	Cat#141706; RRID: AB_10896421
Anti-mouse PD-1 antibody	Biolegend	Cat#109103; RRID: AB_313420
Anti-mouse CD44 antibody	Biolegend	Cat#103008; RRID: AB_493687
Anti-mouse CD62L antibody	Biolegend	Cat#104435; RRID: AB_2562560
Anti-mouse CD69 antibody	Biolegend	Cat#104521; RRID: AB_2260065
Anti-mouse IFN-γ antibody	Biolegend	Cat#505831; RRID: AB_2734492
Anti-mouse TNF-α antibody	Biolegend	Cat#506307; RRID: AB_315428
Anti-mouse Granzyme B antibody	Biolegend	Cat#372207; RRID: AB_2687031
Anti-mouse FOXP3 antibody	Biolegend	Cat#126419; RRID: AB_2565933
Anti-mouse CD16/32 antibody	Biolegend	Cat#101320; RRID: AB_1574975
HIF-1α antibody	Proteintech	Cat#66730-1-Ig; RRID: AB_2882080

Bacterial and virus strains

Escherichia coli MG1655	Angyubio Biotechnology Co. (Shanghai, China)	Cat#G6057

Biological samples

The main organs of mice for biodistribution	BALB/c	Isolated from mice raised in SPF environment
The Blood sample of mice for biochemical analysis	BALB/c	Isolated from mice raised in SPF environment
The main organs of mice for antitumor efficacy and safety assay	BALB/c	Isolated from mice raised in SPF environment

Chemicals, peptides, and recombinant proteins

Doxorubicin	Ruixi Biotechnology Co. (Xian, China)	N/A
ICG-NHS	Ruixi Biotechnology Co. (Xian, China)	N/A
3,3′-dioctadecyloxacarbocyanine perchlorate (DiO)	Thermo Fisher Scientific Co.	Cat#V22886
1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC)	Aladdin	Cat#E106172
N-Hydroxysuccinimide (NHS)	Aladdin	Cat#H109330
3 nM N-(β -ketocaproyl)-l-homoserine lactone (AHL)	Sigma	Cat#K0037
Dulbecco’s modified eagle medium (DMEM)	Gibco	Cat#C11995500BT
Fetal bovine serum (FBS)	Gibco	Cat#10270-106
Penicillin/streptomycin	Gibco	Cat#15140122
RPMI 1640 medium	Gibco	Cat#C11875500BT
Kanamycin	Aladdin	Cat#K103026
Ampicillin	Aladdin	Cat#A102050
HBSS solution	Procell	Cat#PB180321
collagenase V	YEASEN	Cat#40511ES60
FACS buffer	Biolegend	Cat#420201
RBC Lysis Buffer	Biolegend	Cat#420301
True-NuclearTM Transcription Factor Buffer Set	Biolegend	Cat#424401
RIPA lysis buffer	CWBIO	Cat#CW2333
DAPI	Beyotime Biotechnology Co.(Shanghai, China)	Cat#C1005
4% paraformaldehyde	Biosharp	Cat#BL539A

Critical commercial assays

Annexin V-FITC/PI apoptosis detection kit	Beyotime Biotechnology Co.(Shanghai, China)	Cat#C1062M
IFN-γ assay kit	Dakewe Bioengineering Co. (Shenzhen, China)	Cat#1210002
IL-12 assay kit	Dakewe Bioengineering Co. (Shenzhen, China)	Cat#1211202
IL-6 assay kit	Dakewe Bioengineering Co. (Shenzhen, China)	Cat#1210602
IL-1β assay kit	Dakewe Bioengineering Co. (Shenzhen, China)	Cat#1210122
TGF-β assay kit	Dakewe Bioengineering Co. (Shenzhen, China)	Cat#1217102
HMGB1 assay kit	Saipei Biotechnology Co. (Wuhan, China)	Cat#SP14752

Deposited data

In vitro experiments	This paper, Mendeley Data	https://doi.org/10.17632/hzjzmbjx26.1
In vivo experiments	This paper, Mendeley Data	https://doi.org/10.17632/pv5twz8zmn.1

Experimental models: Cell lines

4T1 cells	Pricella Biotechnology Co. (Wuhan China)	Cat#CL-0007
RAW264.7 cells	Pricella Biotechnology Co. (Wuhan China)	Cat#CL-0190
DC2.4 cells	Fuheng Biotechnology Co. (Shanghai China)	Cat#FH0510

Experimental models: Organisms/strains

Mouse: BALB/c.	Zhuhai Bes Test Bio-Tech Co. (Zhuhai, China)	N/A

Recombinant DNA

pBV220 plasmid	Miaoling Bioscience & Technology Co. (Wuhan, China)	N/A
pTD103-ARG1 plasmid	Bourdeau et al.¹⁸	Addgene Plasmid #106475; RRID: Addgene_106475

Software and algorithms

GraphPad Prism 8.0	GraphPad Software Inc.	Version 8.0
ImageJ	Image, Java	Version 1.8
FlowJo™	FlowJo, LLC	Version 10.0

Open in a new tab

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to Prof. Fei Yan (fei.yan@siat.ac.cn).

Materials availability

All reagents generated in this study are available from the lead contact with a completed materials transfer agreement (MTA).

Data and code availability

•
Original data has been deposited at website (https://data.mendeley.com/) and is publicly available as of the date of publication. DOI is available in the key resources table.
•
This paper does not report original code.
•
Any additional information required to reanalyze the data reported in this study is available from lead contact upon request.

Experimental model and subject details

Cell types and culture

The 4T1 breast tumor cells (Catalog No. CL-0007) and RAW264.7 cells (Catalog No. CL-0190) were obtained from Pricella Biotechnology Co. (Wuhan China) and DC2.4 cells (Catalog No. FH0510) were obtained from Fuheng Biotechnology Co. (Shanghai China). 4T1 breast tumor cells were cultured in the dulbecco’s modified eagle medium (DMEM, Gibco, Catalog No. C11995500BT), containing 10% fetal bovine serum (FBS, Gibco, Catalog No. 10270-106) and 1% penicillin/streptomycin (PS, Gibco, Catalog No. 15140122). RAW264.7 cells and DC2.4 cells were cultured RPMI 1640 medium (Gibco, Catalog No. C11875500BT) with 10% FBS and 1% penicillin/streptomycin. The cells were incubated at 37°C with 5% CO₂.

Animal model and care

Female BALB/c mice in 4–6 weeks were purchased from Zhuhai Bes Test Bio-Tech Co. (Zhuhai, China). Animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of the Animal Experiment Center of Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences (SIAT-IACUC-210831-HCS-YF-A2043). All mice were housed in SPF-grade pathogen-free environment with 12 h light/dark cycle at 20 ± 3°C and a relative humidity of 40%–70%.

Method details

Materials

The IFN-γ gene with N-terminal OmpA signal peptide sequence was synthesized by Sangon bioengineering Co. (Shanghai, China). pBV220 plasmid was purchased from Miaoling Bioscience & Technology Co. (Wuhan, China). The pTD103-ARG1 plasmid containing the ARG1 gene was purchased Addgene Co. (Catalog No. 106475). DOX modified with cis-aconitic anhydride and ICG-NHS were purchased from Ruixi Biotechnology Co. (Xian, China). E. coli MG1655 strain was obtained from Angyubio Biotechnology Co. (Catalog No. G6057). Anti-PD-L1 antibody was obtained from Biolegend (Clone: 10 F.9G2, Catalog No. 124338). Anti-CRT (Catalog No. AF1666) and Anti-HMGB1 (Catalog No. AF0180) antibodies were purchased from Beyotime Biotechnology Co. Murine IFN-γ (Catalog No. 1210002), IL-12 (Catalog No. 1211202), IL-6 (Catalog No. 1210602), IL-1β (Catalog No. 1210122) and TGF-β (Catalog No. 1217102) enzyme-linked immunosorbent assay (ELISA) kits were obtained from Dakewe Bioengineering Co. (Shenzhen, China). 3,3′-dioctadecyloxacarbocyanine perchlorate (DiO, Catalog No. V22886) were obtained from Thermo Fisher Scientific Co.

Preparation of engineered bacteria with a two-plasmid system(Ec@IG/Ec@MG)

The construction of pBV-IFN-γ or pBV-mCherry was referred to our previous study.¹⁰ The pTD103-ARG1 plasmid and the pBV-IFN-γ (Data S1) or pBV-mCherry (Data S2) plasmid were transformed into E. coli MG1655 competent cells by electroporation transformation methods. Then, these bacteria were cultured in 1 mL LB medium for 1 h at 37°C. Subsequently appropriate amount of bacterial solution was plated on LB solid plate with 50 μg mL⁻¹ kanamycin (Aladdin, Catalog No. K103026) and 100 μg mL⁻¹ ampicillin (Aladdin, Catalog No. A102050) for 16 h at 37°C. The bacterium colonies were considered as Ec@IG/Ec@MG bacteria, which were amplified in LB medium with 50 μg mL⁻¹ kanamycin and 100 μg mL⁻¹ ampicillin at 37°C overnight for further experiments. Ec@G was obtained through transforming the pTD103-ARG1 plasmid into E. coli MG1655 according to the similar method but only being amplified in LB with 50 μg mL⁻¹ kanamycin.

Induction of GVs and construction of Ec@DIG-GVs bacteria

The Ec@IG or Ec@MG bacteria were cultured at 37°C until OD_600nm = 0.4–0.5 and then 3 nM N-(β -ketocaproyl)-l-homoserine lactone (AHL, Sigma, Catalog No. K0037) was added to the media to induce the expression of GVs for 22 h at 30°C in order to obtain Ec@IG-GVs and Ec@MG-GVs bacteria. To obtain Ec@DIG-GVs bacteria, 1 mg cis-aconitic anhydride-modified DOX was activated with 3.3 mg N-Hydroxysuccinimide (NHS, Aladdin, Catalog No. H109330) and 5.5 mg 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC, Aladdin, Catalog No. E106172) in PBS solution at room temperature for 1 h. Then, Ec@IG-GVs bacteria (8 × 10⁸ CFU) were collected and reacted with DOX at room temperature for 10 min. The mixed solution was centrifuged to remove the supernatant and washed twice with PBS. In order to evaluate whether DOX binds to the bacterial surface, the Ec@IG-GVs bacteria were labeled by Hoechst staining at room temperature for 15 min. The Hoechst-stained Ec@IG-GVs bacteria were washed by PBS for three times. Then, the resulting Ec@IG-GVs bacteria were conjugated with DOX in the same way as stated above. The fluorescence colocalization of Ec@DIG-GVs bacteria was observed by inverted fluorescence microscope (Nikon, E−600, Japan). For detecting the number of drugs loaded onto the bacterial surface, the drug-loaded bacteria were collected and lysed with hydrochloric acid, and the total DOX were detected by microplate reader at 490 nm.

In vitro ultrasound contrast imaging of Ec@DIG-GVs bacteria

The Ec@DIG and Ec@DIG-GVs bacteria at the concentrations of OD_600nm = 0.5, 1, 2, 2.5 were added to 1% agar gel wells. Ultrasound contrast imaging of these bacteria were detected by VisualSonics Vevo2100 (VisualSonics, Inc, USA). The ultrasound parameters were 21 MHz, 20% ultrasound power, 1.18 mechanical index and a one-cycle pulse. Image signals were analyzed by ImageJ.

Cellular uptake of DOX

The Ec@DIG-GVs bacteria at 8 × 10⁸ CFU were incubated in LB medium with different pH (7.4, 6.5 or 5.4) at 37°C for 10 h to detect the release of DOX. Then, the solutions were centrifuged to collect the supernatant. The bacterial supernatants were added to 5×10⁴ 4T1 cells cultivated in 24-well plate (150 μL each well) for 6 h. After that, the cell nuclei were stained by DAPI (4′,6-diamidino-2-phenylindole, Beyotime, Catalog No. C1005) for 1 min at room temperature and then rinsed for three times with PBS. The fluorescence signals were observed by inverted fluorescence microscope.

Detection of CRT expression and HMGB1 release

To examine the ICD of tumor cells, 5×10⁴ 4T1 tumor cells per well were plated into a 24-well plate for 24 h. These cells were incubated with the bacterial supernatant as mentioned above for 12 h. Then, the cell supernatant was removed, and the tumor cells were washed by PBS for three times. The primary CRT or HMGB1 antibodies were used to incubate with these cells overnight at 4°C. The FITC-labeled secondary antibodies were added into these cells for 1 h. The CRT expression was analyzed by Inverted fluorescence microscope and flow cytometry, and the fluorescence signals of HMGB1 was observed by Inverted fluorescence microscope. For quantitative analysis of the extracellular release of HMGB1, the cell supernatant was collected and detected by ELASA kits (Saipei Biotechnology Co., Catalog No. SP14752).

In vitro stimulation of DC cells

To detect the maturation of DC cells, 1×10⁵ immature DC 2.4 cells were seeded on 6-well plates for each well. These cells were incubated with the bacterial supernatant as mentioned above for 24 h. After that, the supernatants were collected to determine the release of IL-1β, IL-6, TGF-β cytokines by ELISA kit. Furthermore, the collected DC 2.4 cells were resuspended in the flow cytometry buffer and stained with anti-CD45, anti-CD11C, anti-CD80, anti-CD86 antibodies for 15 min on ice according to the manufacturer’s instructions. The activation and maturation of DC 2.4 cells were analyzed by flow cytometry.

In vitro precise control of mCherry expression by ultrasound contrast imaging-guided hHIFU irradiation

1% (w/v) solid agarose gel was dissolved in PBS by heating and mixed with Ec@MG-GVs bacteria when the temperature of agarose solution decreased to 42°C. Subsequently, the mixed solution was poured into the mold and place it at 4°C for cooling. The agar phantom buried with Ec@MG-GVs bacteria was imaged as stated above. Then, the focus of the hHIFU was fixed at the Ec@MG-GVs bacteria for 25 min. The temperature was recorded by thermal imaging system and maintained at 45°C. The expression of mCherry protein was detected by IVIS imaging system (VIS Spectrum, PerkinElmer, USA) after 6 h. The hHIFU parameters were as follows: frequency: 1 MHz, duty cycle: 66.7%, pulse period: 150 ms, pulse duration: 100 ms, pulse frequency: 6.67Hz, peak negative acoustic pressure: 4.93 Mpa.

In vivo precise control of mCherry expression by ultrasound contrast imaging-guided hHIFU irradiation

The 4T1 murine subcutaneous tumor model was established. DiO-labeled Ec@MG-GVs bacteria (2.8 × 10⁷ CFU) were intratumorally injected into the mice, and the tumor was imaged by ultrasound to determine the location of bacteria. Then, the target area was demarcated and the hHIFU was applied (irradiation time: 25 min, temperature: 45°C). Subsequently, tumor tissue was collected and the expression of mCherry protein was evaluated by IVIS imaging system. After that, this tumor was cut into two parts, one part with hHIFU irradiation and the another without irradiation. These two parts of tumor tissue were immobilized overnight with 4% paraformaldehyde (Biosharp, Catalog No. BL539A), sectioned and then stained with DAPI to further examine bacterial location and mCherry protein expression under Inverted fluorescence microscope.

In vivo tumor-homing of Ec@DIG-GVs bacteria

In order to detect the tumor-homing ability of Ec@DIG-GVs bacteria, the DOX conjugated onto bacterial surface was replaced by ICG-NHS to construct the ICG-labeled Ec@IG-GVs (Ec@IIG-GVs) bacteria. Briefly, 2.8 × 10⁸ CFU Ec@IG-GVs were co-incubated with 1 mg/mL ICG-NHS in PBS solution for 1 h at room temperature. The bacterial solution was washed by PBS for three times and resuspend in 1 mL PBS. Subsequently, the resulting Ec@IIG-GVs (2.8 × 10⁷ CFU) in 100 μL PBS were injected intravenously into 4T1 subcutaneous tumor bearing mice. For comparison, the dead Ec@IG-GVs bacteria treated with 65°C for 2 h were also labeled with ICG-NHS. The free ICG-NHS at 4 μg/mL was used the control. The mice were imaged by the IVIS system (PekinEmer) at different times to observe the fluorescence signal distribution in the mice. Then these mice were sacrificed at 48 h and the major organs of heart, liver, spleen, lung, kidney and tumor were collected for the ex vivo fluorescence imaging. In addition, the organs and tumors were homogenized and plated on the solid LB plates with 50 μg mL⁻¹ kanamycin and 100 μg mL⁻¹ ampicillin after dilution of 10–1000 times. The bacterial colonies were counted and analyzed after 24 h. For the bio-distribution of Ec@DIG-GVs in the tumor-bearing mice, the similar procedure as Ec@IIG-GVs was applied. The major organs and tumors were collected for tissue slice, followed by staining with DAPI and FITC-labeled HIF-1α antibody (Proteintech, Catalog No. 66730-1-Ig). Meanwhile, the tumors were homogenized to detect the fluorescence intensity of DOX by Multiscan Spectrum at excitation wavelength: 490 nm, emission wavelength: 585 nm. For the detection of IFN-γ levels, the tumors were treated with Ec@DIG-GVs+hHIFU irradiation at 45°C for 25 min to induce the IFN-γ expression. After 48 h, the mice were sacrificed, and the tumors were collected, weighed and homogenized in RIPA lysis buffer (CWBIO, Catalog No. CW2333) on ice for 20 min. Then, the supernatant was centrifugated and detected by ELISA kit to evaluate the concentration of IFN-γ protein.

Ultrasound imaging of the tumor after intravenous administration of Ec@DIG-GVs

2.8×10⁷ Ec@DIG or Ec@DIG-GVs were intravenously injected into 4T1 tumor-bearing mice via the tail vein. For imaging of tumor, the mice were anesthetized by 2% isoflurane in oxygen, and the heating pad was used to keep the normal body temperature of mice. The ultrasonic gel was applied to the surface of the tumor in order to achieve the success of ultrasound transmission. First, the 1^st ultrasound contrast image of tumor was obtained by VisualSonics Vevo2100 and then a short high energy pulse was implemented to destroy the GVs in Ec@DIG-GVs. Subsequently, the second acoustic pressure collapse were implemented to acquire the 2^nd ultrasound contrast image of tumor. For image signal analysis, the first burst contrast signals represented the background signal of the tumor and the signal of GVs. The second burst contrast signals reflected the background signal. Therefore, the contrast signals of Ec@DIG-GVs could calculated by subtracted the 2^nd imaging signals from the 1^st imaging signals by ImageJ. Notably, the different tumor-bearing mice were used to obtain the ultrasound contrast signal at different time points since these GVs had been collapsed. The ultrasound imaging parameters included frequency of 21MHz, 20% ultrasound power, 1.18 mechanical index and a one-cycle pulse.

Immune response triggered by Ec@DIG-GVs combined with hHIFU

The tumors and spleens of each groups were obtained after 7 days post treatment. While for the rechallenged tumor-bearing mice, the tumors and spleens were collected after 14 days post the second tumor inoculation. Tumor and spleen tissues were digested into single cell suspension using HBSS solution (Procell, Catalog No. PB180321) with 2 mg/mL collagenase V (YEASEN, Catalog No. 40511ES60) for 6 h at 37°C. Subsequently, the resulting mixed solution was filtered to collect single-cell suspension and washed by FACS buffer (Biolegend, Cat No. 420201). For spleen samples, erythrocytes in the cell suspension were lysed by adding RBC Lysis Buffer (Biolegend, Cat No. 420301) for 15 min on ice and then filtered through 70 μm cell strainer. CD16/32 antibody (Biolegend, Cat No. 101320) was used to these single cell suspensions from tumor or spleen for 20 min on ice to pre-block the non-specific binding. After that, the cells were stained by surface antibodies, including CD45 (Biolegend, Cat No. 103154), CD3 (Biolegend, Cat No. 100204), CD4 (Biolegend, Cat No. 100429), CD8 (Biolegend, Cat No. 100722), Tim-3 (Biolegend, Cat No. 119703), CD11b (Biolegend, Cat No. 101205), F4/80 (Biolegend, Cat No. 123127), CD11c (Biolegend, Cat No. 117307), CD80 (Biolegend, Cat No. 104713), CD86 (Biolegend, Cat No. 105028), MHC I (Biolegend, Cat No. 114613), CD206 (Biolegend, Cat No. 141706), PD-1 (Biolegend, Cat No. 109103), CD44 (Biolegend, Cat No. 103008), CD62L (Biolegend, Cat No. 104435), CD69 (Biolegend, Cat No. 104521). For the intracellular and intranuclear staining, the cells were fixed and permeabilized by True-NuclearTM Transcription Factor Buffer Set (Biolegend, Cat No. 424401) after the surface antibody labeling. Then, these cells were further stained by intracellular or intranuclear antibodies of IFN-γ (Biolegend, Cat No. 505831), TNF-α (Biolegend, Cat No. 506307), Granzyme B (Biolegend, Cat No. 372207), FOXP3 (Biolegend, Cat No. 126419). The fluorescence signals were detected by flow cytometry and data were analyzed by FlowJo.

In vivo anti-tumor treatment

The 4T1 murine subcutaneous tumor model was established by injecting 1×10⁶ tumor cells into the right thigh of each mouse. The tumor-bearing mice with similar average tumor size were randomly divided into five groups, including PBS control, Ec@G-GVs+hHIFU, Ec@DIG-GVs, Ec@IG-GVs+hHIFU and Ec@DIG-GVs+hHIFU. 2.8 × 10⁷ CFU bacteria in 100 μL PBS were intratumorally injected into the tumor-bearing mice. 100 μL PBS was used as the control. The tumors of the mice in the Ec@G-GVs+hHIFU, Ec@IG-GVs+hHIFU and Ec@DIG-GVs+hHIFU groups were received with hHIFU irradiation for 25 min. The temperature of tumor was monitored by infrared imaging devices and maintained at 45°C. For hHIFU irradiation details, the ultrasound transducer was immersed in a tank filled with degassed distilled water. The tumor-bearing mouse was anesthetized and placed to a self-made platform with a hole in the center, and the tumor was fixed in the hole and contacted with water. The focal point of the transducer was positioned into the tumor. The degassed water acted as a coupling medium to deliver acoustic beams from the transducer to the tumor. hHIFU parameters were used as mentioned above. In the anti-tumor experiments of systemic administration of engineered bacteria, Ec@G-GVs, Ec@IG-GVs and Ec@DIG-GVs were intravenously injected into the tumor-bearing mice and received the hHIFU after 48 h. The treatment procedure was repeated once after 7 days. For treatment process, the tumor volume, survival time and weight of mice were recorded every 2 days by blinded manner. The tumor volume was calculated by the formula: tumor length × tumor width²/2. Mice with some severely symptoms from tumor development (e.g., breathing abnormalities, kyphosis) or weight loss of 20% or more were excluded. The survival curves were generated until the mice died or the tumor volume of the mice exceeded 2000 mm³.

The anti-tumor treatment of Ec@DIG-GVs+hHIFU combined with anti-PD-L1 antibody and tumor rechallenge study

The 4T1 tumor-bearing mice (with similar average tumor volume about 50 mm³) were randomly assigned divided into four groups, including PBS control group, aPD-L1 group, Ec@DIG-GVs+hHIFU group, Ec@DIG-GVs+hHIFU+aPD-L1 group. The control group and Ec@DIG-GVs+hHIFU group were performed as the same strategy with intravenous injection mentioned above. For aPD-L1 group, anti-PD-L1 antibodies at 100 μg dose per mouse were intravenously injected into tumor-bearing mice at day 4, day 11 and day 13. For Ec@DIG-GVs+hHIFU+aPD-L1 group, tumor-bearing mice were performed as the same treatment as Ec@DIG-GVs+hHIFU group but these mice were also received with anti-PD-L1 antibodies at the dose of 100 μg per mouse at day 4, day 11 and day 13. The tumor volume, weight and survival time of tumor-bearing mice were record by the same strategy mentioned above. After 30 days, the left thighs of the cured tumor-bearing mice with Ec@DIG-GVs+hHIFU+aPD-L1 and the right thighs of healthy mice were re-implanted with 5×10⁵ 4T1 tumor cells. The volume of rechallenged tumor, survival time and weight of mice were also monitored every 2 days by blinded manner. Mice with some severely symptoms from tumor development (e.g., breathing abnormalities, kyphosis) or weight loss of 20% or more were excluded. The survival curves were generated until the mice died or the tumor volume of the mice exceeded 2000 mm³.

Biosafety assay

The therapeutic safety was evaluated by blood routine indexes, blood biochemical values and H&E staining of heart, liver, spleen, lung, kidney sections of these treated mice. Briefly, the mice in each group were sacrificed after seven days post hHIFU irradiation (n = 3), and the major organs were collected and fixed by 4% paraformaldehyde overnight. Blood samples were obtained from murine eyes for detection of blood routine indexes. For examination of blood biochemical values, the blood serum was obtained through placing blood samples at room temperature for 2 h and then centrifuging at 350 g for 30 min.

Quantification and statistical analysis

Statistical analysis was disposed by GraphPad prism 8.0. Data were displayed for the mean ± SD. Two groups were analyzed by T test, and more than two groups were calculated by One/two-way ANOVA using the Tukey/Sidak’s test. The data were considered significant, when ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05; ns, no significant. The combination index (CI) was calculated by CI = A/a+B/b, (CI values <1, = 1, and >1 indicate synergistic, additive, and antagonistic effects, respectively).⁴¹

Acknowledgments

This work was supported by the National Key R&D Program of China (grant nos. 2020YFA0908800 and 2019YFE0110400), the National Natural Science Foundation of China (grant nos. 32171365, 81971621, and 82272028), the Key R&D Program of Hunan Province (grant no. 2021SK2035), the Natural Science Foundation of Hunan Province (grant nos. 2022JJ30039 and 2022JJ40392), the Natural Science Foundation of Guangdong Province (grant nos. 2021A1515011177 and 2020A1515110628), the Shenzhen Medical Research Fund (grant no. D2301012), and the Shenzhen Institute of Synthetic Biology Scientific Research Program (grant no. JCHZ20210002).

Author contributions

F.Y. and Z.C. designed and supported the project. Y.Y. performed the experiments and collected the data. Y.W., F.Z., and Y.C. analyzed and interpreted the data. F.Y. and Y.Y. wrote the paper.

Declaration of interests

The authors declare no competing interests.

Published: April 18, 2024

Footnotes

Supplemental information can be found online at https://doi.org/10.1016/j.xcrm.2024.101512.

Contributor Information

Zhiyi Chen, Email: zhiyi_chen@usc.edu.cn.

Fei Yan, Email: fei.yan@siat.ac.cn.

Supplemental information

Document S1. Figures S1–S25, Table S1, and Data S1 and S2

mmc1.pdf^{(8.3MB, pdf)}

Document S2. Article plus supplemental information

mmc2.pdf^{(18MB, pdf)}

References

1.Forbes N.S. Engineering the perfect (bacterial) cancer therapy. Nat. Rev. Cancer. 2010;10:785–794. doi: 10.1038/nrc2934. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Wu M.R., Jusiak B., Lu T.K. Engineering advanced cancer therapies with synthetic biology. Nat. Rev. Cancer. 2019;19:187–195. doi: 10.1038/s41568-019-0121-0. [DOI] [PubMed] [Google Scholar]
3.Din M.O., Danino T., Prindle A., Skalak M., Selimkhanov J., Allen K., Julio E., Atolia E., Tsimring L.S., Bhatia S.N., Hasty J. Synchronized cycles of bacterial lysis for in vivo delivery. Nature. 2016;536:81–85. doi: 10.1038/nature18930. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Chowdhury S., Castro S., Coker C., Hinchliffe T.E., Arpaia N., Danino T. Programmable bacteria induce durable tumor regression and systemic antitumor immunity. Nat. Med. 2019;25:1057–1063. doi: 10.1038/s41591-019-0498-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Nuyts S., Van Mellaert L., Theys J., Landuyt W., Bosmans E., Anné J., Lambin P. Radio-responsive recA promoter significantly increases TNFalpha production in recombinant clostridia after 2 Gy irradiation. Gene Ther. 2001;8:1197–1201. doi: 10.1038/sj.gt.3301499. [DOI] [PubMed] [Google Scholar]
6.Fu S., Zhang R., Gao Y., Xiong J., Li Y., Pu L., Xia A., Jin F. Programming the lifestyles of engineered bacteria for cancer therapy. Natl. Sci. Rev. 2023;10 doi: 10.1093/nsr/nwad031. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Zhu X., Chen S., Hu X., Zhao L., Wang Y., Huang J., Chen J., Qiu Y., Zhang X., Wang M., et al. Near-Infrared Nano-Optogenetic Activation of Cancer Immunotherapy via Engineered Bacteria. Adv. Mater. 2023;35 doi: 10.1002/adma.202207198. [DOI] [PubMed] [Google Scholar]
8.Nuyts S., Van Mellaert L., Theys J., Landuyt W., Lambin P., Anné J. The use of radiation-induced bacterial promoters in anaerobic conditions: A means to control gene expression in Clostridium-mediated therapy for cancer. Radiat. Res. 2001;155:716–723. doi: 10.1667/0033-7587(2001)155[0716:Tuorib]2.0.Co;2. [DOI] [PubMed] [Google Scholar]
9.Fan J.X., Li Z.H., Liu X.H., Zheng D.W., Chen Y., Zhang X.Z. Bacteria-Mediated Tumor Therapy Utilizing Photothermally-Controlled TNF-alpha Expression via Oral Administration. Nano Lett. 2018;18:2373–2380. doi: 10.1021/acs.nanolett.7b05323. [DOI] [PubMed] [Google Scholar]
10.Chen Y., Du M., Yuan Z., Chen Z., Yan F. Spatiotemporal control of engineered bacteria to express interferon-gamma by focused ultrasound for tumor immunotherapy. Nat. Commun. 2022;13:4468. doi: 10.1038/s41467-022-31932-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Gao R., Liu F., Liu W., Zeng S., Chen J., Gao R., Wang L., Fang C., Song L., Sedgwick A.C., et al. Background-suppressed tumor-targeted photoacoustic imaging using bacterial carriers. Proc. Natl. Acad. Sci. USA. 2022;119 doi: 10.1073/pnas.2121982119. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Zou Z.P., Du Y., Fang T.T., Zhou Y., Ye B.C. Biomarker-responsive engineered probiotic diagnoses, records, and ameliorates inflammatory bowel disease in mice. Cell Host Microbe. 2023;31:199–212.e5. doi: 10.1016/j.chom.2022.12.004. [DOI] [PubMed] [Google Scholar]
13.Wang L., Zhang X., Tang C., Li P., Zhu R., Sun J., Zhang Y., Cui H., Ma J., Song X., et al. Engineering consortia by polymeric microbial swarmbots. Nat. Commun. 2022;13:3879. doi: 10.1038/s41467-022-31467-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Ren G., Zhou X., Long R., Xie M., Kankala R.K., Wang S., Zhang Y.S., Liu Y. Biomedical applications of magnetosomes: State of the art and perspectives. Bioact. Mater. 2023;28:27–49. doi: 10.1016/j.bioactmat.2023.04.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Andreou C., Weissleder R., Kircher M.F. Multiplexed imaging in oncology. Nat. Biomed. Eng. 2022;6:527–540. doi: 10.1038/s41551-022-00891-5. [DOI] [PubMed] [Google Scholar]
16.Li P., Wang D., Hu J., Yang X. The role of imaging in targeted delivery of nanomedicine for cancer therapy. Adv. Drug Deliv. Rev. 2022;189 doi: 10.1016/j.addr.2022.114447. [DOI] [PubMed] [Google Scholar]
17.Farhadi A., Ho G.H., Sawyer D.P., Bourdeau R.W., Shapiro M.G. Ultrasound imaging of gene expression in mammalian cells. Science. 2019;365:1469–1475. doi: 10.1126/science.aax4804. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Bourdeau R.W., Lee-Gosselin A., Lakshmanan A., Farhadi A., Kumar S.R., Nety S.P., Shapiro M.G. Acoustic reporter genes for noninvasive imaging of microorganisms in mammalian hosts. Nature. 2018;553:86–90. doi: 10.1038/nature25021. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Jin S.M., Yoo Y.J., Shin H.S., Kim S., Lee S.N., Lee C.H., Kim H., Kim J.E., Bae Y.S., Hong J., et al. A nanoadjuvant that dynamically coordinates innate immune stimuli activation enhances cancer immunotherapy and reduces immune cell exhaustion. Nat. Nanotechnol. 2023;18:390–402. doi: 10.1038/s41565-022-01296-w. [DOI] [PubMed] [Google Scholar]
20.Qi J., Jin F., You Y., Du Y., Liu D., Xu X., Wang J., Zhu L., Chen M., Shu G., et al. Synergistic effect of tumor chemo-immunotherapy induced by leukocyte-hitchhiking thermal-sensitive micelles. Nat. Commun. 2021;12:4755. doi: 10.1038/s41467-021-24902-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Sabado R.L., Balan S., Bhardwaj N. Dendritic cell-based immunotherapy. Cell Res. 2017;27:74–95. doi: 10.1038/cr.2016.157. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Sallusto F., Lenig D., Förster R., Lipp M., Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature. 1999;401:708–712. doi: 10.1038/44385. [DOI] [PubMed] [Google Scholar]
23.Chen Q.W., Wang J.W., Wang X.N., Fan J.X., Liu X.H., Li B., Han Z.Y., Cheng S.X., Zhang X.Z. Inhibition of Tumor Progression through the Coupling of Bacterial Respiration with Tumor Metabolism. Angew. Chem. Int. Ed. Engl. 2020;59:21562–21570. doi: 10.1002/anie.202002649. [DOI] [PubMed] [Google Scholar]
24.Fan J.X., Peng M.Y., Wang H., Zheng H.R., Liu Z.L., Li C.X., Wang X.N., Liu X.H., Cheng S.X., Zhang X.Z. Engineered Bacterial Bioreactor for Tumor Therapy via Fenton-Like Reaction with Localized H(2) O(2) Generation. Adv. Mater. 2019;31 doi: 10.1002/adma.201808278. [DOI] [PubMed] [Google Scholar]
25.Liu L., He H., Luo Z., Zhou H., Liang R., Pan H., Ma Y., Cai L. In Situ Photocatalyzed Oxygen Generation with Photosynthetic Bacteria to Enable Robust Immunogenic Photodynamic Therapy in Triple-Negative Breast Cancer. Adv. Funct. Mater. 2020;30 doi: 10.1002/adfm.201910176. [DOI] [Google Scholar]
26.Zhang X., Pang G., Sun T., Liu X., Pan H., Zhang Y., Liu J., Chang J., Wang H., Liu D. A red light-controlled probiotic bio-system for in-situ gut-brain axis regulation. Biomaterials. 2023;294 doi: 10.1016/j.biomaterials.2023.122005. [DOI] [PubMed] [Google Scholar]
27.Ma X., Liang X., Li Y., Feng Q., Cheng K., Ma N., Zhu F., Guo X., Yue Y., Liu G., et al. Modular-designed engineered bacteria for precision tumor immunotherapy via spatiotemporal manipulation by magnetic field. Nat. Commun. 2023;14:1606. doi: 10.1038/s41467-023-37225-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Hurt R.C., Buss M.T., Duan M., Wong K., You M.Y., Sawyer D.P., Swift M.B., Dutka P., Barturen-Larrea P., Mittelstein D.R., et al. Genomically mined acoustic reporter genes for real-time in vivo monitoring of tumors and tumor-homing bacteria. Nat. Biotechnol. 2023;41:919–931. doi: 10.1038/s41587-022-01581-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Zhu J., Li Z., Zhang C., Lin L., Cao S., Che H., Shi X., Wang H., van Hest J.C.M. Single enzyme loaded nanoparticles for combinational ultrasound-guided focused ultrasound ablation and hypoxia-relieved chemotherapy. Theranostics. 2019;9:8048–8060. doi: 10.7150/thno.37054. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Liu T., Zhang N., Wang Z., Wu M., Chen Y., Ma M., Chen H., Shi J. Endogenous Catalytic Generation of O(2) Bubbles for In Situ Ultrasound-Guided High Intensity Focused Ultrasound Ablation. ACS Nano. 2017;11:9093–9102. doi: 10.1021/acsnano.7b03772. [DOI] [PubMed] [Google Scholar]
31.Crouzet S., Rebillard X., Chevallier D., Rischmann P., Pasticier G., Garcia G., Rouviere O., Chapelon J.Y., Gelet A. Multicentric oncologic outcomes of high-intensity focused ultrasound for localized prostate cancer in 803 patients. Eur. Urol. 2010;58:559–566. doi: 10.1016/j.eururo.2010.06.037. [DOI] [PubMed] [Google Scholar]
32.Illing R.O., Kennedy J.E., Wu F., ter Haar G.R., Protheroe A.S., Friend P.J., Gleeson F.V., Cranston D.W., Phillips R.R., Middleton M.R. The safety and feasibility of extracorporeal high-intensity focused ultrasound (HIFU) for the treatment of liver and kidney tumours in a Western population. Br. J. Cancer. 2005;93:890–895. doi: 10.1038/sj.bjc.6602803. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Ribault M., Chapelon J.Y., Cathignol D., Gelet A. Differential attenuation imaging for the characterization of high intensity focused ultrasound lesions. Ultrason. Imaging. 1998;20:160–177. doi: 10.1177/016173469802000302. [DOI] [PubMed] [Google Scholar]
34.De Benedetti F., Prencipe G., Bracaglia C., Marasco E., Grom A.A. Targeting interferon-γ in hyperinflammation: opportunities and challenges. Nat. Rev. Rheumatol. 2021;17:678–691. doi: 10.1038/s41584-021-00694-z. [DOI] [PubMed] [Google Scholar]
35.Burridge P.W., Li Y.F., Matsa E., Wu H., Ong S.G., Sharma A., Holmström A., Chang A.C., Coronado M.J., Ebert A.D., et al. Human induced pluripotent stem cell-derived cardiomyocytes recapitulate the predilection of breast cancer patients to doxorubicin-induced cardiotoxicity. Nat. Med. 2016;22:547–556. doi: 10.1038/nm.4087. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Wallace K.B., Sardão V.A., Oliveira P.J. Mitochondrial Determinants of Doxorubicin-Induced Cardiomyopathy. Circ. Res. 2020;126:926–941. doi: 10.1161/Circresaha.119.314681. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Sun K., Hu J., Meng X., Lei Y., Zhang X., Lu Z., Zhang L., Wang Z. Reinforcing the Induction of Immunogenic Cell Death Via Artificial Engineered Cascade Bioreactor-Enhanced Chemo-Immunotherapy for Optimizing Cancer Immunotherapy. Small. 2021;17 doi: 10.1002/smll.202101897. [DOI] [PubMed] [Google Scholar]
38.Zhan M., Qiu J., Fan Y., Chen L., Guo Y., Wang Z., Li J., Majoral J.P., Shi X. Phosphorous Dendron Micelles as a Nanomedicine Platform for Cooperative Tumor Chemoimmunotherapy via Synergistic Modulation of Immune Cells. Adv. Mater. 2023;35:e2208277. doi: 10.1002/adma.202208277. [DOI] [PubMed] [Google Scholar]
39.Du W., Frankel T.L., Green M., Zou W. IFNgamma signaling integrity in colorectal cancer immunity and immunotherapy. Cell. Mol. Immunol. 2022;19:23–32. doi: 10.1038/s41423-021-00735-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Gocher A.M., Workman C.J., Vignali D.A.A. Interferon-gamma: teammate or opponent in the tumour microenvironment? Nat. Rev. Immunol. 2022;22:158–172. doi: 10.1038/s41577-021-00566-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Robles A.J., Kurmasheva R.T., Bandyopadhyay A., Phelps D.A., Erickson S.W., Lai Z., Kurmashev D., Chen Y., Smith M.A., Houghton P.J. Evaluation of Eribulin Combined with Irinotecan for Treatment of Pediatric Cancer Xenografts. Clin. Cancer Res. 2020;26:3012–3023. doi: 10.1158/1078-0432.CCR-19-1822. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S25, Table S1, and Data S1 and S2

mmc1.pdf^{(8.3MB, pdf)}

Document S2. Article plus supplemental information

mmc2.pdf^{(18MB, pdf)}

Data Availability Statement

•
Original data has been deposited at website (https://data.mendeley.com/) and is publicly available as of the date of publication. DOI is available in the key resources table.
•
This paper does not report original code.
•
Any additional information required to reanalyze the data reported in this study is available from lead contact upon request.

[bib1] 1.Forbes N.S. Engineering the perfect (bacterial) cancer therapy. Nat. Rev. Cancer. 2010;10:785–794. doi: 10.1038/nrc2934. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] 2.Wu M.R., Jusiak B., Lu T.K. Engineering advanced cancer therapies with synthetic biology. Nat. Rev. Cancer. 2019;19:187–195. doi: 10.1038/s41568-019-0121-0. [DOI] [PubMed] [Google Scholar]

[bib3] 3.Din M.O., Danino T., Prindle A., Skalak M., Selimkhanov J., Allen K., Julio E., Atolia E., Tsimring L.S., Bhatia S.N., Hasty J. Synchronized cycles of bacterial lysis for in vivo delivery. Nature. 2016;536:81–85. doi: 10.1038/nature18930. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] 4.Chowdhury S., Castro S., Coker C., Hinchliffe T.E., Arpaia N., Danino T. Programmable bacteria induce durable tumor regression and systemic antitumor immunity. Nat. Med. 2019;25:1057–1063. doi: 10.1038/s41591-019-0498-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Nuyts S., Van Mellaert L., Theys J., Landuyt W., Bosmans E., Anné J., Lambin P. Radio-responsive recA promoter significantly increases TNFalpha production in recombinant clostridia after 2 Gy irradiation. Gene Ther. 2001;8:1197–1201. doi: 10.1038/sj.gt.3301499. [DOI] [PubMed] [Google Scholar]

[bib6] 6.Fu S., Zhang R., Gao Y., Xiong J., Li Y., Pu L., Xia A., Jin F. Programming the lifestyles of engineered bacteria for cancer therapy. Natl. Sci. Rev. 2023;10 doi: 10.1093/nsr/nwad031. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] 7.Zhu X., Chen S., Hu X., Zhao L., Wang Y., Huang J., Chen J., Qiu Y., Zhang X., Wang M., et al. Near-Infrared Nano-Optogenetic Activation of Cancer Immunotherapy via Engineered Bacteria. Adv. Mater. 2023;35 doi: 10.1002/adma.202207198. [DOI] [PubMed] [Google Scholar]

[bib8] 8.Nuyts S., Van Mellaert L., Theys J., Landuyt W., Lambin P., Anné J. The use of radiation-induced bacterial promoters in anaerobic conditions: A means to control gene expression in Clostridium-mediated therapy for cancer. Radiat. Res. 2001;155:716–723. doi: 10.1667/0033-7587(2001)155[0716:Tuorib]2.0.Co;2. [DOI] [PubMed] [Google Scholar]

[bib9] 9.Fan J.X., Li Z.H., Liu X.H., Zheng D.W., Chen Y., Zhang X.Z. Bacteria-Mediated Tumor Therapy Utilizing Photothermally-Controlled TNF-alpha Expression via Oral Administration. Nano Lett. 2018;18:2373–2380. doi: 10.1021/acs.nanolett.7b05323. [DOI] [PubMed] [Google Scholar]

[bib10] 10.Chen Y., Du M., Yuan Z., Chen Z., Yan F. Spatiotemporal control of engineered bacteria to express interferon-gamma by focused ultrasound for tumor immunotherapy. Nat. Commun. 2022;13:4468. doi: 10.1038/s41467-022-31932-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] 11.Gao R., Liu F., Liu W., Zeng S., Chen J., Gao R., Wang L., Fang C., Song L., Sedgwick A.C., et al. Background-suppressed tumor-targeted photoacoustic imaging using bacterial carriers. Proc. Natl. Acad. Sci. USA. 2022;119 doi: 10.1073/pnas.2121982119. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Zou Z.P., Du Y., Fang T.T., Zhou Y., Ye B.C. Biomarker-responsive engineered probiotic diagnoses, records, and ameliorates inflammatory bowel disease in mice. Cell Host Microbe. 2023;31:199–212.e5. doi: 10.1016/j.chom.2022.12.004. [DOI] [PubMed] [Google Scholar]

[bib13] 13.Wang L., Zhang X., Tang C., Li P., Zhu R., Sun J., Zhang Y., Cui H., Ma J., Song X., et al. Engineering consortia by polymeric microbial swarmbots. Nat. Commun. 2022;13:3879. doi: 10.1038/s41467-022-31467-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Ren G., Zhou X., Long R., Xie M., Kankala R.K., Wang S., Zhang Y.S., Liu Y. Biomedical applications of magnetosomes: State of the art and perspectives. Bioact. Mater. 2023;28:27–49. doi: 10.1016/j.bioactmat.2023.04.025. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] 15.Andreou C., Weissleder R., Kircher M.F. Multiplexed imaging in oncology. Nat. Biomed. Eng. 2022;6:527–540. doi: 10.1038/s41551-022-00891-5. [DOI] [PubMed] [Google Scholar]

[bib16] 16.Li P., Wang D., Hu J., Yang X. The role of imaging in targeted delivery of nanomedicine for cancer therapy. Adv. Drug Deliv. Rev. 2022;189 doi: 10.1016/j.addr.2022.114447. [DOI] [PubMed] [Google Scholar]

[bib17] 17.Farhadi A., Ho G.H., Sawyer D.P., Bourdeau R.W., Shapiro M.G. Ultrasound imaging of gene expression in mammalian cells. Science. 2019;365:1469–1475. doi: 10.1126/science.aax4804. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18.Bourdeau R.W., Lee-Gosselin A., Lakshmanan A., Farhadi A., Kumar S.R., Nety S.P., Shapiro M.G. Acoustic reporter genes for noninvasive imaging of microorganisms in mammalian hosts. Nature. 2018;553:86–90. doi: 10.1038/nature25021. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] 19.Jin S.M., Yoo Y.J., Shin H.S., Kim S., Lee S.N., Lee C.H., Kim H., Kim J.E., Bae Y.S., Hong J., et al. A nanoadjuvant that dynamically coordinates innate immune stimuli activation enhances cancer immunotherapy and reduces immune cell exhaustion. Nat. Nanotechnol. 2023;18:390–402. doi: 10.1038/s41565-022-01296-w. [DOI] [PubMed] [Google Scholar]

[bib20] 20.Qi J., Jin F., You Y., Du Y., Liu D., Xu X., Wang J., Zhu L., Chen M., Shu G., et al. Synergistic effect of tumor chemo-immunotherapy induced by leukocyte-hitchhiking thermal-sensitive micelles. Nat. Commun. 2021;12:4755. doi: 10.1038/s41467-021-24902-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] 21.Sabado R.L., Balan S., Bhardwaj N. Dendritic cell-based immunotherapy. Cell Res. 2017;27:74–95. doi: 10.1038/cr.2016.157. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] 22.Sallusto F., Lenig D., Förster R., Lipp M., Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature. 1999;401:708–712. doi: 10.1038/44385. [DOI] [PubMed] [Google Scholar]

[bib23] 23.Chen Q.W., Wang J.W., Wang X.N., Fan J.X., Liu X.H., Li B., Han Z.Y., Cheng S.X., Zhang X.Z. Inhibition of Tumor Progression through the Coupling of Bacterial Respiration with Tumor Metabolism. Angew. Chem. Int. Ed. Engl. 2020;59:21562–21570. doi: 10.1002/anie.202002649. [DOI] [PubMed] [Google Scholar]

[bib24] 24.Fan J.X., Peng M.Y., Wang H., Zheng H.R., Liu Z.L., Li C.X., Wang X.N., Liu X.H., Cheng S.X., Zhang X.Z. Engineered Bacterial Bioreactor for Tumor Therapy via Fenton-Like Reaction with Localized H(2) O(2) Generation. Adv. Mater. 2019;31 doi: 10.1002/adma.201808278. [DOI] [PubMed] [Google Scholar]

[bib25] 25.Liu L., He H., Luo Z., Zhou H., Liang R., Pan H., Ma Y., Cai L. In Situ Photocatalyzed Oxygen Generation with Photosynthetic Bacteria to Enable Robust Immunogenic Photodynamic Therapy in Triple-Negative Breast Cancer. Adv. Funct. Mater. 2020;30 doi: 10.1002/adfm.201910176. [DOI] [Google Scholar]

[bib26] 26.Zhang X., Pang G., Sun T., Liu X., Pan H., Zhang Y., Liu J., Chang J., Wang H., Liu D. A red light-controlled probiotic bio-system for in-situ gut-brain axis regulation. Biomaterials. 2023;294 doi: 10.1016/j.biomaterials.2023.122005. [DOI] [PubMed] [Google Scholar]

[bib27] 27.Ma X., Liang X., Li Y., Feng Q., Cheng K., Ma N., Zhu F., Guo X., Yue Y., Liu G., et al. Modular-designed engineered bacteria for precision tumor immunotherapy via spatiotemporal manipulation by magnetic field. Nat. Commun. 2023;14:1606. doi: 10.1038/s41467-023-37225-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28.Hurt R.C., Buss M.T., Duan M., Wong K., You M.Y., Sawyer D.P., Swift M.B., Dutka P., Barturen-Larrea P., Mittelstein D.R., et al. Genomically mined acoustic reporter genes for real-time in vivo monitoring of tumors and tumor-homing bacteria. Nat. Biotechnol. 2023;41:919–931. doi: 10.1038/s41587-022-01581-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] 29.Zhu J., Li Z., Zhang C., Lin L., Cao S., Che H., Shi X., Wang H., van Hest J.C.M. Single enzyme loaded nanoparticles for combinational ultrasound-guided focused ultrasound ablation and hypoxia-relieved chemotherapy. Theranostics. 2019;9:8048–8060. doi: 10.7150/thno.37054. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] 30.Liu T., Zhang N., Wang Z., Wu M., Chen Y., Ma M., Chen H., Shi J. Endogenous Catalytic Generation of O(2) Bubbles for In Situ Ultrasound-Guided High Intensity Focused Ultrasound Ablation. ACS Nano. 2017;11:9093–9102. doi: 10.1021/acsnano.7b03772. [DOI] [PubMed] [Google Scholar]

[bib31] 31.Crouzet S., Rebillard X., Chevallier D., Rischmann P., Pasticier G., Garcia G., Rouviere O., Chapelon J.Y., Gelet A. Multicentric oncologic outcomes of high-intensity focused ultrasound for localized prostate cancer in 803 patients. Eur. Urol. 2010;58:559–566. doi: 10.1016/j.eururo.2010.06.037. [DOI] [PubMed] [Google Scholar]

[bib32] 32.Illing R.O., Kennedy J.E., Wu F., ter Haar G.R., Protheroe A.S., Friend P.J., Gleeson F.V., Cranston D.W., Phillips R.R., Middleton M.R. The safety and feasibility of extracorporeal high-intensity focused ultrasound (HIFU) for the treatment of liver and kidney tumours in a Western population. Br. J. Cancer. 2005;93:890–895. doi: 10.1038/sj.bjc.6602803. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Ribault M., Chapelon J.Y., Cathignol D., Gelet A. Differential attenuation imaging for the characterization of high intensity focused ultrasound lesions. Ultrason. Imaging. 1998;20:160–177. doi: 10.1177/016173469802000302. [DOI] [PubMed] [Google Scholar]

[bib34] 34.De Benedetti F., Prencipe G., Bracaglia C., Marasco E., Grom A.A. Targeting interferon-γ in hyperinflammation: opportunities and challenges. Nat. Rev. Rheumatol. 2021;17:678–691. doi: 10.1038/s41584-021-00694-z. [DOI] [PubMed] [Google Scholar]

[bib35] 35.Burridge P.W., Li Y.F., Matsa E., Wu H., Ong S.G., Sharma A., Holmström A., Chang A.C., Coronado M.J., Ebert A.D., et al. Human induced pluripotent stem cell-derived cardiomyocytes recapitulate the predilection of breast cancer patients to doxorubicin-induced cardiotoxicity. Nat. Med. 2016;22:547–556. doi: 10.1038/nm.4087. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] 36.Wallace K.B., Sardão V.A., Oliveira P.J. Mitochondrial Determinants of Doxorubicin-Induced Cardiomyopathy. Circ. Res. 2020;126:926–941. doi: 10.1161/Circresaha.119.314681. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib37] 37.Sun K., Hu J., Meng X., Lei Y., Zhang X., Lu Z., Zhang L., Wang Z. Reinforcing the Induction of Immunogenic Cell Death Via Artificial Engineered Cascade Bioreactor-Enhanced Chemo-Immunotherapy for Optimizing Cancer Immunotherapy. Small. 2021;17 doi: 10.1002/smll.202101897. [DOI] [PubMed] [Google Scholar]

[bib38] 38.Zhan M., Qiu J., Fan Y., Chen L., Guo Y., Wang Z., Li J., Majoral J.P., Shi X. Phosphorous Dendron Micelles as a Nanomedicine Platform for Cooperative Tumor Chemoimmunotherapy via Synergistic Modulation of Immune Cells. Adv. Mater. 2023;35:e2208277. doi: 10.1002/adma.202208277. [DOI] [PubMed] [Google Scholar]

[bib39] 39.Du W., Frankel T.L., Green M., Zou W. IFNgamma signaling integrity in colorectal cancer immunity and immunotherapy. Cell. Mol. Immunol. 2022;19:23–32. doi: 10.1038/s41423-021-00735-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40.Gocher A.M., Workman C.J., Vignali D.A.A. Interferon-gamma: teammate or opponent in the tumour microenvironment? Nat. Rev. Immunol. 2022;22:158–172. doi: 10.1038/s41577-021-00566-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41.Robles A.J., Kurmasheva R.T., Bandyopadhyay A., Phelps D.A., Erickson S.W., Lai Z., Kurmashev D., Chen Y., Smith M.A., Houghton P.J. Evaluation of Eribulin Combined with Irinotecan for Treatment of Pediatric Cancer Xenografts. Clin. Cancer Res. 2020;26:3012–3023. doi: 10.1158/1078-0432.CCR-19-1822. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Ultrasound-visible engineered bacteria for tumor chemo-immunotherapy

Yaozhang Yang

Yuanyuan Wang

Fengyi Zeng

Yuhao Chen

Zhiyi Chen

Fei Yan

Summary

Graphical abstract

Highlights

Introduction

Results

Construction and characterization of Ec@DIG-GVs

Figure 1.

Ec@DIG-GVs for tumor cell killing and activation of immune responsive in vitro

Figure 2.

Local chemo-immunotherapy to active immune response in 4T1 subcutaneous tumor

Figure 3.

In vivo ultrasound tracking of Ec@DIG-GV bacteria homing into tumors

Figure 4.

Systemic administration of Ec@DIG-GVs combined with hHIFU mediated anti-tumor chemo-immunotherapy

Figure 5.

PD-L1 blockade augmented the anti-tumor efficacy of Ec@DIG-GVs+hHIFU through activating long-term immune memory

Figure 6.

Discussion

Limitations of the study

STAR★Methods

Key resources table

Resource availability

Lead contact

Materials availability

Data and code availability

Experimental model and subject details

Cell types and culture

Animal model and care

Method details

Materials

Preparation of engineered bacteria with a two-plasmid system(Ec@IG/Ec@MG)

Induction of GVs and construction of Ec@DIG-GVs bacteria

In vitro ultrasound contrast imaging of Ec@DIG-GVs bacteria

Cellular uptake of DOX

Detection of CRT expression and HMGB1 release

In vitro stimulation of DC cells

In vitro precise control of mCherry expression by ultrasound contrast imaging-guided hHIFU irradiation

In vivo precise control of mCherry expression by ultrasound contrast imaging-guided hHIFU irradiation

In vivo tumor-homing of Ec@DIG-GVs bacteria

Ultrasound imaging of the tumor after intravenous administration of Ec@DIG-GVs

Immune response triggered by Ec@DIG-GVs combined with hHIFU

In vivo anti-tumor treatment

The anti-tumor treatment of Ec@DIG-GVs+hHIFU combined with anti-PD-L1 antibody and tumor rechallenge study

Biosafety assay

Quantification and statistical analysis

Acknowledgments

Author contributions

Declaration of interests

Footnotes

Contributor Information

Supplemental information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases