Genetically engineered bio-composite mediated dual-modality imaging-guided synergistic chemo-FUAS for tumor therapy

Abstract

Focused ultrasound ablation surgery (FUAS) combined with bacterial synergistic therapy has significant potential in solid tumors, however, this combination therapy has been limited mainly due to its weak anti-tumor effects, single imaging modality, and exhibition of tumor recurrence. Herein, we aim to develop a novel multifunctional bio-composite comprising the coating of polydopamine (pDA) and doxorubicin (DOX) on the surface of genetically engineered bacteria (GVs-E. coli) for the ultrasound imaging (USI) and photoacoustic imaging (PAI)-guided synergistic chemo-FUAS therapy of tumor. The bio-composite preferentially colonize into hypoxic tumor tissues, the gas vesicles (GVs) expressed in GVs-E. coli can effectively synergize to FUAS through their cavitation effect and perform USI, and the pDA coating on the surface can produce a uniform PAI signal. Simultaneously, FUAS can trigger the controlled release of chemotherapy drug DOX. Our results confirmed that a combination of biotherapy synergistic chemo-FUAS therapy resulted in significant tumor inhibition and extended survival of 4T1 breast tumor model. This strategy has the potential to solve the inherent limitations of traditional FUAS treatments, while improving the therapeutic effect in conjunction with chemotherapy. It provides a new paradigm for image-guided, multi-level collaborative anti-cancer therapy.

Graphical abstract

1. Introduction

Cancer therapeutic remains a significant challenge in clinic practice due to its tumor heterogeneity and complex microenvironment [1,2]. Traditional treatment methods include surgery, chemotherapy, and radiotherapy, but its high toxicity, drug resistance, and low treatment efficiency hinder its further clinical application [3]. Focused ultrasound ablation surgery (FUAS) is an emerging non-invasive treatment technology, has gained widespread application in the treatment of various solid tumors [4,5]. However, it still remains challenging for deep and large solid tumors in complete ablation. While increasing the ablation power and duration may enhance the therapeutic effect, it also elevates the risk to surrounding normal tissues [6]. Therefore, it's urgent to explore some new ways to improve the efficiency of the FUAS ablation. Synergistic agents (SAs) can enhance the ablation efficiency to improve cancer therapy, has attracted increasing attention. Various conventional SAs such as hydroxyapatite, gas-filled nanoparticles, and lipiodol can enhance the FUAS ablation efficiency through cavitation effect or change the acoustic properties of the tumor tissue, but limited by inefficient tumor targeting, complex preparation process, and the inevitable rapid metabolism [[7], [8], [9], [10]].

To address these limitations, our team has the development of bacteria-based biological SAs [[11], [12], [13]]. Bacteria, such as Escherichia coli, Bifidobacterium, Salmonella, and Clostridium, have been found to demonstrate tumor-targeting capabilities due to their inherent tropism towards the anoxic tumor microenvironment [[14], [15], [16]]. Research has demonstrated that genetic engineering techniques have the potential to significantly diminish the virulence of bacteria [17]. The acoustic reporter gene (ARG) encoding gas vesicle proteins was successfully introduced into Escherichia coli and GVs-E. coli was successfully prepared. These unique protein shell nanoparticles (NPs) as the cavitation nucleus can be used as gaseous FUAS SAs to enhance the treatment of FUAS, which can avoid the deficiency of introducing exogenous synergism to a certain extent [18,19]. Initial experimental results indicate that viable bacterial cells exhibit significant potential as tumor-specific delivery vectors, capable of being conjugated with SAs, imaging contrast agents, and anticancer drugs via diverse bioconjugation strategies such as electrostatic interactions, aptamer-mediated targeting, and carbodiimide-based covalent coupling. [[20], [21], [22], [23]]. These bacterial-based biological SAs significantly potentiate FUAS efficacy, enabling precise cancer therapy. However, certain challenges remain to be addressed, particularly regarding the manufacturing complexity and in vivo stability of bacterial-drug conjugates.

Polydopamine (pDA), a bioinspired polymer mimicking mussel adhesive protein, has emerged as a versatile surface coating material due to the exceptional adhesive properties of its catechol functional groups [24,25]. The material's unique binding capability stems from its ability to form multiple interactions with cell surfaces, including covalent bonds with amine/thiol groups and hydrogen bonds with hydroxyl groups [26]. The multifunctional adhesion mechanism of pDA, which involves hydrogen bonding, π-π stacking, Michael addition, and Schiff base reactions, enables robust bacterial surface modification while potentially overcoming traditional limitations of bacterial-nanoparticle conjugation by providing more stable interfacial bonding through synergistic interactions and better preservation of bacterial viability and functionality compared to conventional conjugation methods [27,28].

Herein, in this study, we developed a novel multifunctional bio-composite using genetic engineering bacteria and surface modification to integrate bacteria with chemotherapy agents and explore their potential for anti-tumor therapy with synergistic FUAS (Scheme 1). E. coli was genetically modified to produce gas vesicle (GVs) inside and then coated with pDA and doxorubicin (DOX) on the surface of GVs-E. coli through in situ interfacial polymerization, rendering GVs-Ec@p/D, which were injected intravenously into the tumor bearing mice (Scheme 1a). The obtained GVs-Ec@p/D could rapidly accumulate and penetration into the tumor region by taking advantage of their tumor-targeting and colonization capability. Upon irradiation with FUAS, GVs can utilize their cavitated nucleus as SAs to produce coagulative necrosis of the tumor for enhancing FUAS efficacy. At the same time, the pDA coating increased the acoustic impedance difference between the modified bacteria and the tumor tissue, which further enhanced the ablation ability of FUAS on the basis of GVs-E. coli. Furthermore, DOX released from GVs-Ec@p/D further eradicates residual tumor cells post-FUAS ablation, thus achieving a synergistic effect between FUAS and chemotherapy (Scheme 1b). By the way, GVs-Ec@p/D can not only enable visual monitoring of bacterial gene expression in vivo through USI, but also make use of the photothermal effect of pDA in PAI, which can provide the distribution of the bio-composite in the tumor region for monitoring in real time (Scheme 1c). Therefore, the bio-composite avoids the deficiency of introducing exogenous SAs and contrast agents, providing an intelligible and utility strategy that enhances the efficacy of FUAS combined with chemotherapy guided by modality imaging to achieve effective antitumor synergy.

Scheme 1.

Schematic illustration of multifunctional bio-composite for anti-tumor therapy with synergistic FUAS (a) The construction of GVs-Ec@p/D and injected it intravenously into mice. (b) Upon intravenous injection, the obtained GVs-Ec@p/D in the tumor target area for enhancing FUAS combined with chemotherapy. (c) Guidance of dual-modality imaging with the irradiation of FUAS.

2. Materials and methods

2.1. Materials

The plasmid pET28a_T7-ARG1 (Addgene plasmid # 106473) was stored by the State Key Laboratory of Ultrasound in Medicine and Engineering. E. coli BL21(AI) were purchased from Weidi Biotechnology Co., Ltd. (Shanghai, China). Dopamine, ICG and DOX·HCL were purchased from Aladdin. Tris-HCl buffer (pH 8.5, 1.0 mM), fluorescein isothiocyanate (FITC) and 4^’,6-Diamidino-2-phenylindole (DAPI) were supplied by Beyotime Biotechnology (Shanghai, China). 2,3,5-triphenyltetrazolium chloride (TTC) and Cell Counting Kit-8 (CCK-8 kit) was obtained from Damas. All other chemical reagents were of analytical grade and used as received without further purification unless indicated.

2.2. Culture of genetically engineered bacteria

The GVs-E. coli was constructed according to our previous study [20,21,29,30]. Briefly, the plasmid ARG1 was transferred into E. coli BL21(AI) and cultured on Luria broth (LB) agar plates containing kanamycin and glucose. The obtained colonies were transferred to 5 mL LB medium (50 μg/mL kanamycin, 1 % glucose) and incubated for 16 h at 37 °C in a shaking incubator. Then the bacterial solution was transferred to LB medium (50 μg/mL kanamycin, 0.2 % glucose) at a ratio of 1:100 and cultured at 37 °C until OD₆₀₀ = 0.5, induced with L-arabinose (0.5 %) and isopropyl β-D1-thiogalactopyranoside (0.4 mM) for a duration of 22 h at 30 °C to express GVs. Finally, GVs-E. coli was resuspended in 1 mL of phosphate buffered saline (PBS) and diluted to 10⁻⁶ before spreading onto LB agar plates for counting.

2.3. Synthesis of GVs-Ec@p/D

GVs-Ec@p was obtained by adding GVs-E. coli (1 × 10⁸ CFU) to 1 mL of Tris-HCl buffer (pH 8.5, 10 mM) containing dopamine (1 mg/mL) and stirred at 200 rpm for 30 min. At the base of GVs-Ec@p for further adding 0.2 mL of DOX (1 mg/mL) with stirring for 2 h, the resultant GVs-Ec@p/D was obtained and washed by PBS to remove free DOX. The collected GVs-Ec@p and GVs-Ec@p/D were resuspend in PBS and store at 4 °C for further use.

2.4. Characterization of GVs-Ec@p/D

The morphology of GVs-E. coli was characterized using transmission electron microscopy (TEM Hitachi H-7600, Japan). The hydrodynamic diameter, zeta potential and absorption spectra of GVs-Ec@p/D were determined using Malvern nanoscale potentiometer (Zetasizer Pro, UK) and UV–VIS–NIR spectrophotometer (Shimadzu, Japan), respectively. The viability of GVs-Ec@p/D at different concentrations of pDA and DOX as determined using bacterial plate counts. Equal amounts of GVs-E. coli, GVs-Ec@p and GVs-Ec@p/D were diluted in LB medium at OD₆₀₀ and incubated at 37 °C with gentle shaking. OD values of the cultures were recorded once per hour using a microplate absorbance reader (Tecan Infinite 200pro), while bacterial counts were determined every 2 h through standard plate counting methodology (serial dilution and plating on LB agar plates) during the 10 h cultivation period. The effect of the GVs-Ec@p/D was evaluated using a LIVE/DEAD bacterial viability kit according to the manufacturer's protocol. PDA was replaced with pDA-FITC during the preparation of GVs-Ec@p/D and the coating of pDA and DOX on the surface of GVs-E. coli was observed by confocal laser scanning microscopy (CLSM) (Nikon A1, Japan). To determine the percentage of pDA and DOX coating on the bacterial surface, GVs-E. coli, GVs-Ec@p and GVs-Ec@p/D were analyzed by flow cytometry (FCM, BD, USA), all quantitative experiments were performed in triplicate using independently prepared samples.

2.5. Cell culture and animal model

Murine breast cancer 4T1 cells and human umbilical vein endothelial cells (HUVECs) were purchased from the Chinese Academy of Science Cell Bank. As recommended by the American Type Culture Collection (ATCC), the cells were cultured at 37 °C under 5 % CO₂.

The animals received appropriate care in compliance with the Guidance Suggestions for the Care and Use of Laboratory Animals. The procedures were approved by the Chongqing Medical University Animal Ethics Committee. Five-to six-week-old female BALB/c mice (Hunan SJA Laboratory Animal Co.,Ltd.) were subcutaneously injected with 0.1 mL 4T1 cells (1 × 10⁷ cells/mL) on the right side to establish a tumor model. Tumor volume was calculated according to the following formula: Tumor volume = 0.5 × length × (width) ².

2.6. In vivo tumor-targeting capacity and biodistribution of GVs-Ec@p/D

In order to detect the tumor-targeting ability of GVs-Ec@p/D, the DOX deposited on the bacterial indication was replaced with ICG to construct the GVs-Ec@p/I using the same method [31]. ICG was used as a fluorescent tracer to visualize and quantify biodistribution patterns. Following intravenous injection of GVs-Ec@p/I, the fluorescence imaging (FLI) was conducted using the Xenogen IVIS Spectrum imaging system (Perkin Elmer, USA) at various time points (0 h, 6 h, 24 h, 48 h, 56 h, and 72 h) and the FL average value was calculated. Additionally, at 72 h post-administration, tumors and major organs (heart, liver, spleen, lung, and kidneys) were harvested for ex vivo FLI.

To verify the biodistribution of each group, fifteen tumor-bearing mice were randomly divided into control group, GVs-E. coli group and GVs-Ec@p/D group (n = 5). PBS (100 μL), GVs-E. coli (1 × 10⁸ CFU, 100 μL) and GVs-Ec@p/D (the same dose of GVs-E. coli) were injected intravenously for three consecutive days, respectively. The major organs (heart, liver, spleen, lung, kidney) and tumor tissues were harvested and homogenized in sterile PBS at 3 days post-injection. After serial dilutions of the homogenate, 100 μL was plated on LB solid agar plates containing resistance, and bacterial colonies were counted and analyzed 24 h later.

To confirm the tumor-targeting delivery of DOX loaded on the surface of GVs-Ec@p/D, ten tumor-bearing mice were divided into DOX group and GVs-Ec@p/D group, free DOX solution (2 mg/mL, 100 μL) and GVs-Ec@p/D (1 × 10⁸ CFU, 100 μL) were injected intravenously for three consecutive days. The major organs and tumors were collected to stained with DAPI and observed by CLSM.

2.7. Cellular uptake of DOX

To examine the release of DOX, the GVs-Ec@p/D at 1 × 10⁸ CFU were incubated in 1 mL RPMI 1640 medium at 37 °C for 4 h, one group was exposed to FUAS (120 W, 3S) ablation, and the other group was left untreated. The supernatant was collected by centrifugation, and 100 μL were added to 5 × 10⁴ 4T1 cells cultivated in confocal dishes (150 μL each well) for 30min. The cell nuclei were stained with DAPI for 30 min, and then they were purified with PBS three times. The fluorescence signals were observed by CLSM.

2.8. Drug release assay

To investigate the effect of FUSA irradiation and the tumor microenvironment on the DOX release performance, the samples were divided into groups GVs-Ec@p/D and GVs-Ec@p/D + FUAS. These samples were then transferred into dialysis bags and dialyzed against 20 mL of the corresponding buffer (pH = 7.4 or 6.5) under dark conditions with gentle shaking. At predetermined time intervals, 2 mL of the release medium was collected and replaced with an equal volume of fresh PBS. All drug release results were averaged from three independent measurements. The DOX release concentration was determined using ultraviolet–visible–near-infrared (UV–Vis–NIR) spectroscopy.

2.9. In vitro cytotoxicity assays

To evaluate the cellular toxicity of GVs-E. coli and GVs-Ec@p/D with or without FUAS irradiation, the GVs-E. coli and GVs-Ec@p/D at the concentration of 5 × 10⁶ CFU/mL and the PBS were added to the 4T1 cells (1 × 10⁴ cells per well) and incubated for 4 h, respectively. Otherwise, the same dose of GVs-E. coli and GVs-Ec@p/D were exposed to FUAS irradiation (120w) for 3s, the supernatants obtained were processed as described above. Finally, the cell viability was determined by CCK-8 assay. In addition, the cells were seeded in confocal dishes (1 × 10⁴ cells per well) co-stained with calcein-AM/PI for 30 min and imaged by CLSM to observe the living cells and dead cells. Meanwhile, flow cytometry was used to quantify the apoptosis rate in different groups.

2.10. Dual-modality imaging of GVs-Ec@p/D

GVs-E. coli and E. coli at the concentrations of OD₆₀₀ = 1 were added to 1 % agar gel wells to evaluate USI in vitro by ultrasound diagnostic imaging equipment (Mylab 90, Esaote, Italy) in B-mode and CEUS-mode, and the mean echo intensity were analyzed by DFY soft-ware (Chongqing Medical Ultrasound Imaging Institute). For in vivo USI, ten tumor-bearing mice were divided into E. coli group and GVs- E. coli group (n = 5), in which all tumors were maintained at a standardized volume of 100 ± 2 mm³ throughout imaging procedures, and US screening was performed to exclude tumors with pre-existing necrosis (identified as hypoechoic areas), thereby ensuring accurate gray value measurements. Then they were injected with E. coli (1 × 10⁸ CFU, 100 μL) and GVs-E. coli (1 × 10⁸ CFU, 100 μL) for three consecutive days. The US images of the tumor location were taken and the gray value of the target area were quantitatively analyzed.

To assess the PAI potential of GVs-Ec@p/D, which was prepared with different concentrations of pDA (0.5, 1.0, 1.5, 2.0, 2.5 mg/mL) and then introduced into the agar gel model. PA images were obtained by the Vevo LAZR Photoacoustic Imaging System (Visual Sonics Inc., Toronto, Canada) with an LZ-250 probe to analyze the PAI signals. For in vivo PAI, the mice were injected intravenously with GVs-Ec@p/D (1 × 10⁸ CFU, 100 μL, 1 mg/mL based on pDA) for three consecutive days. The PA images were collected at different time points, and the PA signals were analyzed.

2.11. Synergistic effect of GVs-Ec@p/D for FUAS

The focused ultrasound tumor therapeutic system (Model-JC200 Chongqing Haifu Medical Technology Co., LTD., Chongqing, China) was used for all FUAS experiments. The in vitro synergistic effect of GVs-Ec@p/D for FUAS was evaluated by ex vivo bovine liver. The bovine liver in each group were injected with saline (100 μL), E. coli (100 μL, 1 × 10⁸ CFU, the same dose for the following group), GVs-E. coli, and GVs-Ec@p/D, respectively. Immediately after that, the injection site was treated with FUAS ablation at different power (90 W and 120 W) for 3 s. The gray value and coagulative necrosis volume of the ablation area were quantitatively analyzed, and the formula is as follows:V (mm³) = (π/6) × length × width × depth.

To evaluate the synergistic effect in vivo FUAS ablation, fifteen tumor-bearing mice were randomly divided into PBS, GVs-E. coli, and GVs-Ec@p/D group. The injection procedure of different groups was similar to that imaging injections described above. All mice were euthanized after 24 h FUAS treatment (120 W, 3 S), and tumor tissues were sliced into sections for TTC staining. The energy efficiency factor (EEF) can be calculated as follows: EEF (J/mm³ ) = ηPt/V (η = 0.7).

2.12. In vivo synergistic chemo-FUAS therapy efficacy

For evaluating the in vivo synergistic chemo-FUAS therapy efficacy of GVs-Ec@p/D, upon the volume of the tumor reaching 100 mm³, the tumor-bearing mice were randomly divided into five groups: control, FUAS, GVs-E. coli + FUAS, pDA/DOX + FUAS and GVs-Ec@p/D + FUAS. The mice were intravenously administered with PBS (100 μL), GVs-E. coli (100 μL, 1 × 10⁸ CFU of GVs-E. coli), pDA/DOX (100 μL, containing 100 μg pDA and 200 μg DOX), GVs-Ec@p/D (100 μL, containing concentration equal to GVs-E. coli and pDA/DOX), this time-point was defined as day 0. For the groups designated for FUAS ablation, the tumor areas were irradiated at 3 d post-injection using FUAS (120 W) for 3 s. For treatment process, the body weights and tumor volumes of mice were recorded every two days, while the survival of mice was monitored daily to obtain survival rate curves. The tumor volume was calculated by the formula: Volume = (Length × Width × Width)/2. For assessing the relative tumor volume (RTV) pre and post-treatment, the initial tumor volume before the first injection was used as standardized. 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 mm^3. One day after FUAS treatment, tumor tissues were stained with hematoxylin-eosin (H&E) for necrosis assessment. On the 16th day after treatment, tumor tissues were weighed and photographed, and then stained with proliferating cell nuclear antigen (PCNA) and terminal-deoxynucleotidyl transferase-mediated nick end labelling (TUNEL) for further analysis to assess the effect of antitumor therapy in vivo.

2.13. Biosafety assay

The cell viability of GVs-Ec@p/D towards normal HUVEC cells and 4T1 cells was detected by the CCK-8 assay. To evaluate the therapeutic safety, the whole blood of each mouse was drawn to test the blood routine indexes and the important blood biochemical values (white blood cell count, WBC; red blood cell count, RBC; hemoglobin levels, HGB; platelet count, PLT; alanine aminotransferase, ALT; blood urea nitrogen, BUN; aspartate aminotransferase, AST; creatinine, CREA; Creatine Kinase, CK; lactate dehydrogenase, LDH) on days 3, 7 and 14 after injection of GVs-Ec@p/D. All the mice were euthanized at 14th day post-administration, major organs (heart, liver, spleen, lung and kidney) were collected, fixed with 4 % paraformaldehyde, and subjected to H&E staining, and compared with a control group of healthy mice.

2.14. Statistical analysis

All data in this work were presented as mean ± standard deviation and statistically analyzed using GraphPad Prism software 9.5.0. Student t-test was used to identify the statistical significance between the two groups, and one-way analysis of variance (ANOVA) was used for multiple comparisons. The threshold for statistical significance was ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

3. Results and discussions

3.1. Characterization of GVs-Ec@p/D

After induction and repeated centrifugal washing, a lot of white foamy cells could be seen floating on the upper layer of the EP tube, confirming the successful expression of the GVs (Fig. 1a). Gram stain reveals pale pink, long rod-shaped bacteria, and the results show typical features consistent with gram-negative bacteria (Fig. 1b). TEM reveals that GVs-E. coli contains a large number of bubble structures (Fig. 1c), aligning with previous literature reports [29,30], demonstrating the successful construction of GVs-E. coli, which is essential for subsequent targeting of tumors, USI, and ablation of FUAS.

Fig. 1.

Characterization of GVs-Ec@p/D. (a) GVs-E.coli dispersed in PBS. (b) Gram staining of GVs-E.coli. Scale bar: 50 μm. (c) TEM image of GVs-E.coli. Scale bar: 200 nm. (d) Zeta potential and (e) Size distribution of GVs-E.coli, GVs-Ec@p and GVs-Ec@p/D. (f) Absorbance spectra of DOX, GVs-E.coli, GVs-Ec@p and GVs-Ec@p/D. (g) Plate counts of GVs-Ec@p at different pDA concentrations (0, 0.5, 1, 2, and 4 mg/mL). (h) Plate counts of GVs-Ec@p/D after incubation with 0, 50, 100, 200, and 400 μg/mL DOX. (i) Growth curve of GVs-E.coli, GVs-Ec@p and GVs-Ec@p/D. (j) CLSM images of GVs-Ec@p/D. Green and red indicate pDA-FITC and DOX, respectively. Scale bar: 25 μm. (k) Representative flow cytometry histograms of GVs-E.coli, GVs-Ec@p and GVs-Ec@p/D, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

After decorating with pDA (GVs-Ec@p) and pDA/DOX(GVs-Ec@p/D), the hydrodynamic diameter of GVs-E. coli increased slightly from 1062 ± 38 to 1320 ± 33 and 1240 ± 19 nm (Fig. 1d), respectively. Due to the co-deposition of DOX, GVs-Ec@p/D showed a more homogeneous distribution along with a decrease in hydrodynamic size, and this is consistent with previous reports [32,33]. As shown in Fig. 1e, the zeta potentials of GVs-E. coli, GVs-Ec@p and GVs-Ec@p/D were −33.3 mV, −27.9 mV and −22.5 mV, respectively. As we expected, after coating with pDA, the surface potential of GVs-Ec@p exhibited slightly increase, along with the integration of positively charged DOX into the pDA coating, the zeta potential of GVs-Ec@p/D became more positive [34,35]. Simultaneously, SEM imaging of GVs-Ec@p clearly reveals numerous uniformly-sized pDA nanoparticles deposited on the bacterial surface, confirming the successful formation of the pDA coating (Fig. S1). Moreover, GVs-Ec@p/D showed a characteristic absorption peak of DOX at approximately 490 nm. By contrast, GVs-E. coli and GVs-Ec@p had no such characteristic absorption peaks (Fig. 1f), indicating successful loading of DOX.

To investigate the effect of the varying concentrations of dopamine coating and DOX loading on the viability of GVs-E. coli, we observed that cell viability remained largely unaffected by dopamine at a concentration of 1 mg/mL and DOX at a concentration of 200 μg/mL (Fig. 1g and h). Importantly, compared to naked GVs-E. coli, GVs-Ec@p and GVs-Ec@p/D prepared according to the concentrations determined above showed similar bacterial growth curve, and the modified bacteria still maintained comparable viability to GVs-E. coli as shown by the plate count results (Fig. 1i–S2). The result of the live/dead bacterial double stain showed that the loading of the pDA/DOX under appropriate concentration will not affect the vitality of GVs-E. coli (Fig. S3). All of the above evidence shows that indicated the negligible impact of modification on bacterial vitality. As shown in Fig. 1j, CLSM showed that FITC-labeled pDA emitted visible green fluorescence that colocalized with the red fluorescence of DOX after co-deposition, confirming that pDA and DOX were efficiently decorated on the bacterium [32]. Furthermore, flow cytometry's quantitative assessment indicated that approximately 79.54 % ± 0.44 % of GVs-E. coli were decorated with pDA/DOX in an optimized state (Fig. 1k–S4).

3.2. In vivo tumor-targeting capacity and biodistribution of GVs-Ec@p/D

We first evaluated the tumor-targeting capacity of GVs-Ec@p/D, the in vivo imaging results and the fluorescence intensity of mice injected with ICG and GVs-Ec@p/I are shown in Fig. 2a and b. In GVs-Ec@p/I group, a significant enhancement in fluorescence signals was observed at the tumor sites, and the fluorescence signals gradually increased over time and peaked at 56 h post-injection. By contrast, following the intravenous injection of free ICG, weak fluorescence signals were observed in tumors of mice, and the fluorescence signal of ICG group was significantly lower than that of GVs-Ec@p/I group, indicating that free ICG showed limited tumor-targeting ability. In vitro FLI of tumor tissues and organs showed that obvious fluorescence signals were observed in the tumor tissue in GVs-Ec@p/I group, only weak fluorescence signals in ICG group were observed in various organs (Fig. 2c). To further confirm this, we collected the major organs and tumors of the mice that received PBS, GVs-E. coli and GVs-Ec@p/D and homogenized them, followed by plating counts. In vivo tissue homogenization experiments (Fig. 2d) showed that significant bacterial colonies were observed in the tumor region collected from GVs-E. coli and GVs-Ec@p/D treated group, while few bacterial colonies were detected in other major organs (Fig. 2e), further indicating that the biodistribution of the modified bacteria were no different from GVs-E. coli and could selectively colonize in the tumor site [36,37].

Fig. 2.

In vivo tumor-targeting capacity and biodistribution of GVs-Ec@p/D (a) FLI and (b) Quantitative FL intensity of tumor tissue at different time points (n = 5). (c) Ex vivo FLI of the tumors and major organs collected from the tumor-bearing mice treated as indicated at 72 h. (d) Homogenates of tumor tissues of GVs-E.coli and GVs-Ec@p/D were cultured on solid LB agar at 37 °C. (e) Quantification of bacterial colonization in various organs of tumor-bearing mice after injection of GVs-E.coli and GVs-Ec@p/D. (f) Fluorescent imaging of the tumor and major organ sections after intravenous injection of GVs-Ec@p/D, and free DOX at 72 h. Scale bar: 100 μm.

To confirmed that DOX loaded on the surface of GVs-Ec@p/D successfully reached the tumor site by observing the DOX signal of frozen sections. Fig. 2f showed that in the GVs-Ec@p/D group, there were more red fluorescence signals of DOX could be observed in the tumor sites but only few DOX signals in the other major organs. By contrast, free DOX distributed in main organs such as liver, spleen and lung showed more red signals than tumor tissue. Thus, the GVs-Ec@p/D could efficiently carry DOX to the tumor site and reduce the accumulation in major organs. These results indicated that GVs-Ec@p/D possessed the ability to selectively tumor target area and effectively enhance the retention of loaded drugs.

3.3. In vitro FUAS therapy efficacy and cytotoxicity assays

For cancer therapy, accurate and controlled drug release via FUAS irradiation is critical for subsequent treatment. The ability of DOX release was examined through exposing GVs-Ec@p/D to FUAS conditions. By adding equivalent release medium to 4T1 breast cancer cells, a significantly stronger red fluorescence from DOX appeared in 4T1 cells incubated with FUAS-irradiated release medium (Fig. 3a). GVs-E. coli is an extracellular bacterium; therefore, GVs-Ec@p/D, as an extracellular vector, does not rely on the classical endocytic pathway but instead releases its payload through FUAS triggering and the tumor microenvironment [38]. As shown in Fig. 3b, under tumor-mimetic acidic conditions (pH 6.5), GVs-Ec@p/D demonstrated a DOX release rate of 44.18 % within 1 h, representing a statistically significant increase (∗p < 0.05) compared to release at physiological pH (7.4). Notably, when combined with FUAS irradiation at pH 6.5, the release of DOX were dramatically accelerated, achieving 73.42 % ± 1.72 % release within 1 h and reaching near-complete release (98.26 % ± 0.83 %) by 12 h. These results demonstrate that FUAS irradiation serves as the predominant trigger for controlled drug release, enabling near-complete DOX liberation under tumor-mimetic conditions through synergistic pH-responsive activation.

Fig. 3.

In vitro FUAS therapy efficacy and cytotoxicity assays. (a) DOX uptake of 4T1 cells incubated with the 4 h bacterial with or without FUAS irradiation. The nuclei were stained by DAPI. Red fluorescence: DOX. Scale bar: 50 μm. (b) Cumulative DOX release from GVs-Ec@p/D under different pH conditions with/without FUAS irradiation (n = 3). (c) Cell viability of 4T1 cells treated with bacterial supernatants. The equivalent bacterial supernatants were from GVs-E. coli, GVs-E. coli + FUAS, GVs-Ec@p/D, and GVs-Ec@p/D + FUAS. (d) Quantitative analysis of apoptosis rates of 4T1 cells. (e) Fluorescence microscopy images of 4 T1 tumor cells treated with the bacterial supernatants stated above stained with calcein-AM (green fluorescence) and PI (red fluorescence), Scale bar: 100 μm. (f) Tumor cell apoptosis analysis by flow cytometry after different treatments (n = 3). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Next, we investigated the tumor cytotoxicity of GVs-E. coli and GVs-Ec@p/D with FUAS irradiation by the CCK-8 assay. As depicted in Fig. 3c, 70.9 % of tumor cells were killed in the GVs-Ec@p/D + FUAS group, and the cellular viability in FUAS group was significantly lower than the group not exposed to FUAS (∗∗∗∗p < 0.0001). The live and dead cell staining assay showed that numerous red dead cells could be seen in tumor cells incubated with FUAS-treated GVs-Ec@p/D supernatant relative to bacterial supernatants from control or GVs-E. coli, GVs-E. coli + FUAS, and GVs-Ec@p/D groups. In contrast, the strong green fluorescence of the group without FUAS irradiation indicates that most of the 4T1 cells remained alive, which emphasized that the FUAS-induced tumor cytotoxicity of GVs-Ec@p/D by the release of DOX (Fig. 3e). Furthermore, flow cytometry results and quantitative analysis are consistent with results in CLSM characterization (Fig. 3d–f).

These results indicate that FUAS irradiation can significantly increase the transient release of DOX at tumor sites, thereby facilitating controlled drug release and producing more potent tumor cytotoxicity. The combination of GVs-Ec@p/D and FUAS can maximize the synergistic effect of “FUAS + chemotherapy”, ensuring the efficacy of subsequent in vivo treatment.

3.4. Dual-modality imaging in vitro and in vivo

Encouraged by the in vitro FUAS-activated the tumor cytotoxicity of GVs-Ec@p/D, we proceeded to explore their potential for dual-modality imaging of hypoxic tumors in vitro and in vivo. The precise treatment of FUAS is inseparable from the real-time guidance of USI, as exhibited in Fig. 4a, compared with group E. coli, group GVs-E. coli showed obvious echo intensity in B-mode and CEUS-mode. Quantitative analysis shows that the EI index of group GVs-E. coli is 3.08- and 2.60-fold higher than that of group E. coli in B-mode and CEUS-mode, respectively (Fig. 4b). In vivo USI, no significant change in the gray value of the tumor image was observed on 3 days after E. coli injection, while that of the tumor target area was enhanced in both modes when a certain amount of GVs-E. coli accumulated in the tumor of mice (Fig. 4d). Compared to clinically used microbubble contrast agents that undergo rapid clearance (typically within 30–60 min), our bio-composite demonstrates significantly prolonged tumor retention (>72 h) [39]. This extended duration enables continuous imaging guidance throughout FUAS procedures, addressing a critical limitation of current clinical agents. The quantitative analysis showed that the change of gray value after GVs-E. coli injection in both modes of USI was statistically significant (Fig. 4e and f, ∗∗∗∗p < 0.0001). This bacterial imaging visualization method can provide more detailed information about the tumor after intravenous injection and effectively guide tumor therapy.

Fig. 4.

Dual modality imaging in vitro and in vivo (a) In vitro US images in B-mode and CEUS-mode (yellow dotted circle marks the US images of the bacterial solution in the AGAR model). (b) Quantitative analysis of EI in B-mode and CEUS-mode (n = 3, ∗∗∗∗p < 0.0001). (c) PA intensities and images of various concentrations of pDA (n = 5). (d) In vivo USI of tumors in B-mode and CEUS-mode. (e), (f) Quantitative analysis of EI in B-mode and CEUS-mode (n = 5, ∗p < 0.1, ∗∗p < 0.01, ∗∗∗∗p < 0.0001). (g) In vivo PAI of tumors after intravenous injection of GVs-Ec@p/D at various time points (white dotted circles marks the regions of signal enhancement). (h) Quantitative PA intensity (n = 5, ∗∗∗∗p < 0.0001). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

We further investigated the ability of GVs-E. coli to PAI. In vitro PAI showed that the photoacoustic signal of GVs-Ec@p/D enhanced linearly with its concentration (Fig. 4c), which makes it a good PA contrast agent. Within the tumor location, PAI can provide functional information about the tissue vasculature [40,41]. The PAI trend in vivo was consistent with the FLI during the observation period, and the injection of GVs-Ec@p/D showed a continuous accumulation of signal intensity within the tumor tissue (Fig. 4g and h), which peaked at 56 h after injection, and the photoacoustic signals were still visible after 3 days. These results indicated that GVs-Ec@p/D has a good tumor-targeting ability and PAI ability, which can make up for the shortage of the existing monitoring modalities in the treatment of FUAS, providing more complete diagnosis and treatment information.

Integrating PAI tracking with USI enhances the precision of joint treatments, offering extensive information for optimal therapeutic outcomes at the most opportune time, simultaneously boosting treatment efficacy and safety, and amalgamating the advantages of dual-modality imaging [[40], [41], [42]].

3.5. Synergistic effect of GVs-Ec@p/D to FUAS

The synergistic tumor ablation effect of the bio-composite with FUAS in vitro was systematically evaluated by using ex vivo bovine liver experiments. After FUAS ablation, a gray-white coagulative necrosis area with clear boundary and regular shape appeared in the target area (Fig. 5a). According to Fig. 5b and d, we can see that as the FUAS power increased, there was a notable rise in both the gray value and the coagulative necrosis volume in each group. Under the same irradiation power, the gray values and necrosis volumes of the GVs-E. coli group and GVs-Ec@p/D group were significantly higher than the PBS group (∗∗∗∗p < 0.0001). This phenomenon may be attributed to GVs-E. coli serving as inertial cavitation nuclei that enhance FUAS ablation efficacy, which has been well-documented in previous studies [43,44]. The gray value and necrotic volume of the GVs-Ec@p/D group were significantly higher than those of the GVs-E. coli group, suggesting that the pDA coating on the surface of engineered bacteria can be used as a FUAS synergist, and this may be because the pDA coating causes a large acoustic impedance difference between the modified bacteria and the tumor tissue, which further enhances the ablation ability of FUAS and is beneficial to the subsequent tumor ablation in vivo [45,46]. Based on the results of in vitro treatment, the parameters of 120 W and 3 s were selected for in vivo treatment.

Fig. 5.

Synergistic effect of GVs-Ec@p/D by FUAS. (a) Representative photographs of ablated bovine livers after FUAS irradiation. The red dotted circle marks the coagulative necrosis of bovine livers. (b) Quantitative analysis of the gray value of bovine livers after FUAS irradiation (n = 3, ∗∗∗∗p < 0.0001). (c) Schematic diagram of the treatment process of FUAS. (d) Quantitative analysis of necrotic volume after FUAS ablation of bovine livers (n = 3, ∗∗∗∗p < 0.0001). (e) Coagulative necrosis of tumors by TTC staining after FUAS, the necrotic tissue appears gray and the normal tumor tissue is red (yellow dotted circle marks the necrotic tissue). (f) The gray values of tumor tissues in each group after FUAS (n = 5, ∗∗p < 0.01, -∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). (g) In vivo USI of tumor tissues before and after FUAS ablation in different groups. The yellow dotted circle marks the tumor nodules. (h) The coagulative necrosis of tumor tissues in each group after FUAS (n = 5, ∗∗∗∗p < 0.0001). (i) The EEF of tumor tissues in each group after FUAS (n = 5, ∗∗∗p < 0.001, -∗∗∗∗p < 0.0001). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5c shows a schematic of the in vivo treatment process. On the second day after FUAS ablation, tumor tissue was taken for TTC staining, with red areas representing unablated tissue and white areas representing coagulated necrotic tissue. A large number of white coagulative necrotic areas were visible in the tumor tissues of the GVs-Ec@p/D group, as can be clearly seen in Fig. 5e. The specific volumes of coagulative necrosis in the PBS, GVs-E. coli, and GVs-Ec@p/D groups were (14.57 ± 0.72), (46.62 ± 1.25) and (81.73 ± 2.10) mm³, respectively. Ultrasound comparison images were collected at the tumor before and after FUAS ablation (Fig. 5g), and also by quantitatively analyzing the gray value and coagulative necrosis volume after FUAS ablation, it was shown that the gray value and coagulative necrosis volume of group GVs-Ec@p/D was higher than that of group GVs-E. coli, and both of them were significantly higher than that of group PBS (Fig. 5f–h, ∗∗∗∗p < 0.0001, -∗∗p < 0.01, ∗∗∗p < 0.001), which was consistent with the results of the in vitro bovine liver experiments. Meanwhile, the EEF values in the GVs-Ec@p/D group were significantly lower than those in the other two groups (Fig. 5i), indicating that less FUAS energy was required per unit volume of tumor ablation. These results suggest that the bio-composite GVs-Ec@p/D superimposes the FUAS ablation effect through the binding effect of the airbag protein GVs, which act as cavitation nuclei, with the pDA coating, improving the efficacy of FUAS and maximizing the efficiency of tumor treatment [30,47].

3.6. In vivo synergistic chemo-FUAS therapy efficacy

Encouraged by the effective tumor targeting ability and tumor cell-killing effect of bacterial biohybrid, we subsequently assessed the in vivo anti-tumor therapeutic efficacy of GVs-Ec@p/D in 4T1 tumor-bearing mice. Fig. 6a depicts the protocol of the treatment regimen. The in vivo anti-tumor activity across all groups was evaluated following systemic treatment through the quantification of tumor volume, measured every two days over a 16-day monitoring period, after which the tumors were dissected out and weighed. As depicted in Fig. 6c, the tumor of mice in FUAS group grew rapidly, and the growth of tumors in mice treated with GVs-E. coli + FUAS and pDA/DOX + FUAS were inhibited to some extent, which may be attributed to the limited combination of chemotherapy and targeted drug delivery (degree of synergy > 0.1) [48]. As expected, the growth of tumor was further inhibited in the GVs-Ec@p/D + FUAS treated group, suggesting that the improved accumulation of DOX in the tumor with the help of bio-composite contributed to the enhanced FUAS. Consistent with this, the GVs-Ec@p/D + FUAS group showed minimal tumor volume and the lightest tumor weight (Fig. 6b–d). As shown in Fig. 6e, GVs-Ec@p/D + FUAS significantly suppressed tumor growth and greatly extended the survival of treated mice, demonstrating the potent antitumor efficacy. In addition, there was no significant difference in the body weight among the treatment groups (Fig. 6f). The synergistic effect of FUAS was further confirmed by H&E, TUNEL and PCNA staining of tumor sections after treatment (Fig. 6g). Pathological examination via H&E staining revealed that all FUAS-treated tumors exhibited partial necrotic regions compared to the control group. Notably, in group GVs-Ec@p/D + FUAS, there was significant nuclear fragmentation, lysis, and even complete dissolution of cellular structures, as well as liquefaction and vacuolization. PCNA staining of the tumor tissues further showed that GVs-Ec@p/D + FUAS decreased the proliferation rates of the tumor cells. TUNEL staining showed that mice treated in the GVs-Ec@p/D + FUAS group had the highest number of apoptotic tumor cells (green fluorescence), and the area of apoptosis and necrosis in their tumors was much larger than that in the other groups.

Fig. 6.

In vivo synergistic chemo-FUAS therapy efficacy. (a) The protocol of the treatment regimen. (b)Photos of tumors at the 16th day after treatment. (c) Relative tumor volume of different groups (n = 5, ∗∗∗∗P < 0.0001). (d) Tumor weight on the 16th day (n = 5, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001). The survival curves (e) and body weight (f) of mice in each group (n = 5). (g) Optical microscope images of tumor sections from each group were stained for H&E staining, PCNA, and TUNEL testing. Scale bar: 100 μm.

In conclusion, GVs-Ec@p/D + FUAS effectively inhibited tumor growth and extended the survival of tumor-bearing mice. By leveraging targeted and synergistic FUAS ablation, chemotherapy drugs were successfully delivered to the hypoxic regions of the tumor, ensuring treatment safety and enhancing the cytotoxic effects on tumor cells. Additionally, the dual-modality imaging capabilities of GVs-Ec@p/D facilitated optimal treatment timing and precise image guidance.

3.7. Biosafety assay

To evaluate the cytotoxicity of GVs-Ec@p/D, cell cytotoxicity using the CCK-8 assay were analyzed, the cell viabilities of 4T1 and HUVECs treated with different concentrations of GVs-Ec@p/D showed good biosafety (Fig. S5a). Furthermore, to evaluate the in vivo safety of GVs-Ec@p/D, the blood biochemical indexes and H&E staining in mice were examined. The blood indexes of the treated mice fluctuated within the normal range both in the short and long term after injection, indicating that no toxicity was detected in the mice during the test (Fig. S5b–S5f). Additionally, the H&E staining analysis of the heart, liver, spleen, lung, and kidney revealed no noticeable damage with the use of GVs-Ec@p/D (Fig. S5g). These results confirm the biosafety of multifunctional bacterial bio-composite in tumor therapy.

4. Conclusions and future directions

We have successfully developed a novel multifunctional bio-composite (GVs-Ec@p/D) that integrates chemo-FUAS synergistic therapy under dual-modal imaging guidance for solid tumor treatment. By conjugating DOX to genetically engineered GVs-E. coli through pDA self-polymerization, we constructed GVs-Ec@p/D with minimal impact on bacterial viability. The innate tumor-targeting capability and active motility of bacteria enabled successful tumor colonization and intratumoral co-localization. The developed bio-composite demonstrates concurrent USI/PAI capabilities while significantly enhancing therapeutic efficacy through the synergistic combination of FUAS-triggered precise DOX release, GVs-mediated cavitation effects, and pDA-enhanced acoustic impedance contrast, collectively achieving potent antitumor outcomes.

This breakthrough addresses critical limitations in current combination therapies: while chemo-immunotherapy suffers from immune-status dependency and systemic toxicity [49], and dual-drug nanoplatforms face penetration/distribution challenges [50]. Our system uniquely combines: (i) focused ultrasound ablation, (ii) spatially controlled drug release, and (iii) cavitation-enhanced permeability. The integrated design resolves the longstanding trade-offs in conventional systems [[51], [52], [53]] by simultaneously achieving targeting precision and real-time treatment monitoring - establishing a new paradigm for oncology treatment.

While our findings are promising, additional research is necessary to fully investigate the broader potential and to achieve a more comprehensive understanding of the underlying mechanisms: 1)The exploration of the mechanism of DOX release triggered by FUAS irradiation; 2) The photothermal effect of pDA coating can be combined with photothermal therapy to achieve multi-means combined anti-tumor therapy; 3) evaluate imaging and treatment efficacy in other tumor models; 4) Beyond the targeted integration of chemotherapeutic agents, this strategy is also applicable for the combination with other functional agents for bacterium-mediated biomedical applications; 5) The systemic immunologic consequences of bio-composite mediated FUAS therapy remain to be fully elucidated. It is anticipated that the bacterial bio-composite will undergo further refinement through the optimization of its comprehensive properties. This will establish a solid foundation for the subsequent clinical improvement of FUAS, while integrating multifunctional therapeutic strategies for cancer treatment.

CRediT authorship contribution statement

Yijun Zhou: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Conceptualization. Mingyang Zhang: Visualization, Methodology, Investigation, Formal analysis, Conceptualization. Haiyan Yang: Writing – review & editing, Funding acquisition, Formal analysis, Data curation, Conceptualization. Guangrong Zheng: Visualization, Software, Methodology. Li Ren: Methodology, Conceptualization. Zhong Zhang: Validation, Methodology. Li Lin: Validation, Methodology. Youqian He: Validation. Qi Wang: Methodology. Jianzhong Zou: Writing – review & editing, Validation, Supervision, Resources, Project administration.

Ethics approval and consent to participate

All animal procedures were performed in accordance with the Guidelines for Care and Use of Laboratory Animals of Chongqing Medical University and approved by the Animal Ethics Committee of Chongqing Medical University (IACUC-CQMU-2024-12092).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

No data was used for the research described in the article.