An inhalable microbe-oncolytic virus consortium for lung cancer treatment

Summary

Oncolytic viruses (OVs) represent a promising cancer therapy due to their selective cytotoxicity, yet clinical success in lung cancer is hindered by biological barriers and immunosuppressive microenvironments. To address these barriers, an inhalable microbe-OVs consortium is developed by conjugating tumor-responsive PEGylated adenoviruses (Ads) to motile algae (Synechococcus WH8102) via click chemistry. Driven by algal motility, the consortium penetrates respiratory mucus and epithelial barriers to reach lung tumor sites. Intratumoral calcium ion deprivation by the algae disrupts cellular tight junctions and facilitates deep penetration into solid tumors, thereby potentiating Ads-mediated immunogenic cell death (ICD) and boosting microbial and viral co-activated antitumor immunity. Inhalation administration demonstrates robust antitumor effectiveness in murine lung tumor. This inhalable microbe-OV strategy provides a promising scheme for clinical oncolytic microorganism biotherapeutics.

Graphical abstract

Highlights

• •An inhalable microbe-oncolytic virus consortium is developed for lung cancer
• •Mucus and alveolar epithelial barriers are penetrated via algal motility
• •Tight junctions are disrupted by algal calcium competition for deep targeting
• •Robust antitumor immunity is promoted by microbial and viral co-activation

Ling et al. develop an inhalable microbe-oncolytic virus consortium for lung cancer. Through algal motility and calcium ion competition, the consortium opens tumor tight junctions to facilitate the targeted delivery of oncolytic viruses into deep-seated tumors, providing a promising scheme for eliciting potent oncolytic viral-microbial co-activated antitumor immune responses.

Introduction

Lung cancer is the most frequently diagnosed malignancy and the leading cause of cancer-related death globally.¹ It is also characterized by a high tendency to develop multifocal lesions.² Surgical resection is only suitable for a subset of early-stage patients, whereas locally advanced and unresectable cases largely rely on chemotherapy. However, due to insufficient drug accumulation and poor targeting within the lung, therapeutic outcomes remain suboptimal.³ Recent breakthroughs in immunotherapy, particularly immune checkpoint inhibitors (ICIs), have become integral components of the first-line and adjuvant treatments for non-small cell lung cancer.⁴ Nevertheless, a substantial number of patients with lung cancer, especially those with highly heterogeneous and immunologically “cold” tumors, respond poorly to single-agent ICI therapy.^5,6 Therefore, developing immune strategies capable of effectively stimulating antigen release, enhancing antigen presentation, and recruiting immune cells is critically important.

Oncolytic viruses (OVs) have attracted considerable attention for their dual functions: lysing tumor cells to release antigens and subsequently activating anti-tumor immunity.^7,8 This makes them particularly promising for treating lung tumors with high heterogeneity.⁹ At present, at least six OVs are undergoing phase I/II clinical trials for lung cancer, though no phase III studies have yet confirmed definitive efficacy.¹⁰ Clinical translation of OVs faces multiple delivery challenges: systemically administered naked viruses are rapidly eliminated by neutralizing antibodies and host immunity, which limits efficacy and repeat dosing.^11,12 The anatomical structure and microenvironment of the lung make it an ideal target for local OV delivery. The high specific surface area of the alveoli facilitates efficient viral infection.¹³ Local inhalable administration not only increases drug concentration at tumor sites but also minimizes systemic exposure and associated toxicities.¹⁴ However, even when free OVs reach the tumor, uptake occurs mainly in superficial tumor cells, with limited penetration into the dense intratumoral core.¹⁵ Thus, developing inhalable OV delivery systems that shield viral surface antigens, reduce immune clearance, and precisely target deep tumor regions is crucial for enhancing lung cancer immunotherapy.

Recently, active-motility carriers have emerged as powerful alternatives to traditional delivery carriers, offering advantages in targeting, tissue penetration, and intratumoral drug distribution.^16,17,18 Microalgae possess favorable biocompatibility and have demonstrated significant potential as active-motility biocarriers in biomedical applications.¹⁹ In lung cancer, tumor invasion and desmoplastic stroma disrupted alveolar architecture, accompanied by E-cadherin down-regulation.^20,21,22,23 Within this barrier, the intrinsic motility of microalgae enabled broad distribution and tissue penetration.²⁴ Cyanobacteria were highly adaptable to the hypoxic and metabolic-enriched tumor microenvironment (TME).^25,26,27,28 Specifically, the nitrogen-sensing apparatus of this strain directed active chemotactic accumulation toward metabolite-rich regions.²⁹ Unlike traditional flagellated swimmers prone to mechanical entanglement,³⁰ its non-flagellated motility exhibited minimal interference with the stroma, facilitating traversal through narrow interstitial pores.^31,32,33 Crucially, the dense Ca²⁺-binding motifs on SwmA (Hill coefficient = 7.4) and the induced PhoX phosphatase collectively sequestered extracellular Ca²⁺.^31,34 This competitive depletion triggered the conformational collapse of Ca²⁺-dependent E-cadherin domains, reducing intercellular adhesion and promoting deep penetration.

Building on above properties, we engineered an inhalable cyanobacteria-virus consortium by covalently linking PEGylated adenovirus (Ad) type 11 (Ad11) to the surface of the cyanobacterium Synechococcus WH8102 via click chemistry (Figures 1A and 1B). This stable microbe-virus consortium integrates motility, tumor penetration, direct oncolysis, and immune activation into a single system. The particle size of the system meets aerodynamic requirements for lung deposition.²⁴ Upon inhalation, the consortium navigates effectively to lung lesions, synergistically achieving tumor penetration, localized release, and immune activation (Figures 1C–1F). Ad-mediated tumor lysis releases antigens locally, while pathogen-associated molecular patterns (PAMPs) on microalgal surfaces induce M1 macrophage and dendritic cell (DC) maturation, enhancing antigen presentation and initiating robust anti-tumor immunity.³⁵ In both orthotopic and distant metastatic lung tumor, as well as human tumor-bearing and humanized mouse models, the system demonstrated significant tumor inhibition efficacy and favorable biosafety profile, highlighting its clinical potential as a platform for lung cancer therapy.

Figure 1.

Construction and characterization of microbe-OVs consortium

(A) Schematic diagram of pulmonary inhalation therapy.

(B) Preparation process of microbe-OVs consortium via click chemistry-based bioconjugation.

(C) In vivo process post-inhalation of the consortium.

(D) Migration through the bronchial mucus layer.

(E) Penetration across the alveolar epithelial barrier.

(F) Deep tumor penetration.

(G) CLSM images showing colocalization of FITC-labeled Ads and DAPI-stained algae (representative image from one of three independent experiments; DAPI channel was pseudo-colored in red).

(H) TEM images of free Ads (spike proteins indicated by arrows), PEG-Ads, algae, and the consortium (representative image from one of three independent experiments).

(I) Dynamic light scattering analysis showing the particle size of each component of the formulation (n = 3).

(J) Zeta potential of each component of the formulation (n = 3).

(K) Real-time fluorescence quantitative PCR analysis showing the connection rate of Ads at different Ad-to-algae ratios (n = 3).

(L) Flow cytometry analysis showing algae loading efficiency at different Ad-to-algae ratios (representative image from one of three independent experiments).

(M) Release kinetics of PEG-Ads containing disulfide bonds (purple) and non-disulfide bonds (pink) in a simulated tumor microenvironment (glutathione concentration = 0.2 mM) (n = 3).

(N) Cytotoxicity evaluation of PEG-Ads and virus-algae consortium (n = 3).

Statistical analysis: data are presented as mean ± SD. Statistical significance was calculated using two-tailed Student’s t test (ns: no significance; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001).

Results

Preparation and characterization

The consortium was synthesized using click chemistry for precise microbial integration. OVs were PEGylated to mask native antigens and introduced to 11,12-didehydro-5,6-dihydrodibenz (b, f) azocine (DBCO) using the DBCO-PEG₂₀₀₀-S-S-NHS ester, while the surface of algae were decorated with Azido-PEG₂₀₀₀-NHS. After surface modifications, the unreacted DBCO-PEG was removed from Ads by ultrafiltration, and the unreacted Azido-PEG was separated from algae by centrifugation. The purified components were then mixed to form the consortium through the bioorthogonal click reaction (Figure 1B).

The consortium was characterized to evaluate conjugation efficacy. The confocal laser scanning microscopy (CLSM) and transmission electron microscope (TEM) imaging confirmed the formation of consortium (Figures 1G and 1H). Dynamic light scattering revealed an increase in particle size after PEGylation and conjugation, indicating successful crosslinking (Figure 1I). Furthermore, changes in zeta potential indicated that the bioconjugations were mainly due to click chemistry rather than electrostatic interactions (Figure 1J). Ad to algae ratios of 20:1, 10:1, and 5:1 (PFU:CFU) were tested to optimize stoichiometry of the consortium (Figure 1K). Real-time fluorescent PCR analysis indicated Ad connection rate of 53.43%, 77.83%, and 85.26%, respectively. Flow cytometry analysis revealed that at 5:1 ratio, 71.0% of algae were conjugated with Ads, while loading efficiency was increased to 95.8% at 10:1 ratio, confirming the optimal stoichiometry (Figure 1I). Next-generation impactor (NGI) analysis showed that the consortium exhibited a favorable distribution, with 57.53% of particles in the 4.46 to 0.94 μm range, which are ideal for alveolar and respiratory tract deposition (Figure S1).

In a lung-fluid-mimicking environment, responsive PEGylated Ads were completely released from the consortium within 12 h, ensuring that Ads would efficiently infect tumor cells. In comparison, the non-responsive release of Ads released only ∼40% in 12 h, which coincided with bacterial division and membrane component renewal (Figure 1M). Under serum-free DMEM conditions with a maintained glutathione concentration of 0.2 mM, the virus-algae conjugates demonstrated optimal cytotoxicity. Notably, the cytotoxicity evaluation revealed no significant differences in the effects of PEG-Ads and free Ads on the viability of TC-1-hCD46 cells, indicating that PEGylation did not appreciably compromise the infectivity of the viruses (Figure 1N).

In vitro penetration evaluation

These key attributes of algae in the targeting, retention, and penetration were characterized in detail. Real-time confocal observations showed that Synechococcus WH8102 moved at over twice the speed of Synechococcus PCC7942, which served as the negative control due to the absence of the specific swimming ability^32,36 (Figures 2A and 2B; Figure S2). Notably, the loading of Ads did not significantly impede the motility of WH8102. To investigate mucus penetration post-inhalation, an artificial mucus penetration model was established using a gelatin gel coated with artificial mucus to simulate the respiratory barrier.³⁷ WH8102 and PCC7942 labeled with DiI were utilized to the mucus layer and assessed by fluorescence intensity within the gelatin layer after 12 h. WH8102-contained consortium retained 34% in the gel compared to 11% for PCC7942-contained consortium (Figures 2C and 2D). The WH8102, WH8102-contained consortium, and PCC7942 were further co-incubated with macrophages to evaluate the clearance of algae over 48 h. PCC7942 was rapidly cleared within 4 h, whereas both WH8102 and WH8102-contained consortium still remained 56.04% and 47.10% after 48 h, respectively (Figure 2E). These results suggested that the motility of WH8102 could not only enhance penetration efficiency across the mucus barrier but also reduce susceptibility to mucociliary clearance and phagocytosis, thereby promoting deep tissue access and prolonging retention.^29,38 Moreover, Matrigel containing TC-1-hCD46 was conducted to partially simulate the alveolar epithelium and observe the ability of consortium to penetrate this barrier. Three microliters of WH8102 or PCC7942, containing 1 × 10⁷ CFU, was injected with a microinjector to Matrigel. Confocal microscopy 3D scanning was used to assess dispersion at 2 and 12 h post-injection. WH8102 was dispersed evenly throughout Matrigel, while PCC7942 aggregated mostly (Figure 2F). These emphasized the essential role of motility in the survival of microorganisms, particularly in active innate immuno-microenvironments.³⁹

Figure 2.

In vitro penetration evaluation of the consortium

(A) CLSM images showing movement trajectories of WH8102, WH8102-contained consortium, and PCC7942 within 20 s (representative image from one of three independent experiments).

(B) Relative movement rates of WH8102, WH8102-contained consortium, and PCC7942 (n = 3).

(C) 3D CLSM images of an artificial mucus penetration by WH8102-contained consortium and PCC792-contained consortium, simulating lung mucus and tissue (representative image from one of three independent experiments).

(D) Quantification of relative fluorescence intensity using enzyme-linked immunoassay (n = 3).

(E) Retention of WH8102, WH8102-contained consortium, and PCC7942 after co-culture with MH-S macrophages for 48 h, showing macrophage clearance resistance (n = 3).

(F) Dispersion and targeting of WH8102 and PCC7942 in a matrix gel model containing TC-1-hCD46 cells, visualized with 3D CLSM. TC-1-hCD46 cells were stained with DiI, while WH8102 and PCC7942 were marked with DiR (pseudo-colored in green) (representative image from one of three independent experiments).

(G) Adhesion capabilities of Ads, PEG-Ads, and the consortium in TC-1-hCD46 cells, imaged using CLSM after 1 h of co-culture followed by PBS washes (representative image from one of three independent experiments).

(H) Schematic illustration showing algae competitively absorbing Ca²⁺ from tumor cells to disrupt tight junctions.

(I) CLSM images of intracellular Ca²⁺ content in TC-1-hCD46 cells after 12 h of co-culture with different formulations, using Fluo-4, AM fluorescent probe (green for Ca²⁺ and red for cell membrane) (representative image from one of three independent experiments).

(J) Flow cytometry analysis of intracellular Ca²⁺ content in TC-1-hCD46 cells after 12 h of co-culture, using Fluo-4, AM fluorescent probe (representative image from one of three independent experiments).

(K) Western blot analysis of E-cadherin expression in TC-1-hCD46 cells after 12 h of co-culture with PCC7942, WH8102, and WH8102-contained consortium (representative image from one of three independent experiments).

(L) Penetration behavior in a tumor spheroid model using FITC-labeled Ads and DAPI-stained algae (representative image from one of three independent experiments).

Statistical analysis: data are presented as mean ± SD. Statistical significance was calculated using two-tailed Student’s t test (∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001).

The adhesion capabilities of algae to tumor cells were evaluated by co-incubating free Ads, PEG-Ads, and the consortium with TC-1-hCD46 cells for 1 h. After removing unattached formulations, cells were fixed and imaged using CLSM. The retention of the consortium was significantly increased compared with other groups (Figure 2G; Figure S3). This adhesion behavior was intended to simulate the physiological conditions following intratracheal nebulization, where respiratory reflexes, such as coughing, could lead to premature expulsion of inhaled agents. The intrinsic adhesion of the consortium could facilitate enhanced pulmonary retention and mitigate dosage loss during administration. The algae and Ads were conjugated through reduction-sensitive disulfide bonds, which could be gradually cleaved under the lung TME. In addition, the continuous renewal and remodeling of algae membrane components during pulmonary residence could further promote the release of Ads. These synergistic mechanisms collectively ensured the timely and effective release of Ads without impeding receptor-mediated infection or internalization.

Ca²⁺, as the signaling, not only drive the motility of Synechococcus WH8102 but also play a crucial role in the formation of Ca²⁺-dependent cellular adhesion molecules like E-cadherin that are essential for maintaining tight intercellular connections.^31,40 Herein, the algae from the consortium would competitively take up Ca²⁺ with tumor cells.⁴¹ The calcium uptake capability of algae is mediated by specific membrane proteins, including mechanosensitive channels and Ca²⁺/H⁺ antiporters, which promote the unidirectional influx of extracellular calcium ions into the cells with negligible reverse transport.^32,42 Previous studies have demonstrated that calcium deprivation restricts algal growth, whereas calcium supplementation under acidic conditions enhances survival and proliferation. Based on these biological characteristics, it was inferred that algae could competitively uptake calcium ions against tumor cells within the TME. To confirm whether WH8102 is capable of competitively absorbing Ca²⁺ from tumor cells, we suspended WH8102 in calcium/magnesium-free D-PBS and incubated it either with TC-1-hCD46 tumor cells or alone in D-PBS. WH8102 co-cultured with tumor cells exhibited markedly enhanced intracellular calcium fluorescence, while those incubated in D-PBS alone showed negligible signal (Figure S4). These results suggested that calcium uptake occurs preferentially in the presence of tumor cells, likely sourced from the tumor-associated microenvironment, supporting the hypothesis that WH8102 competes with tumor cells for Ca²⁺ in situ.

Furthermore, WH8102 or PCC7942 was co-incubated with TC-1-hCD46 cells for 12 h, and the intracellular Ca²⁺ content was then assessed using a Fluo-4, AM fluorescent probe (Figures 2H and 2I). The results indicated that WH8102 absorbed more amount of Ca²⁺ than PCC7942, which were consistent with the results of flow cytometry analysis (Figure 2J). Notably, Ca²⁺ uptake was unaffected by the presence of Ads. Western blot analysis further revealed the decreased E-cadherin expression in TC-1-hCD46 cells co-cultured with WH8102, suggesting that the reduction of Ca²⁺ within the tumor can decrease the expression of E-cadherin (Figure 2K). Additionally, penetration behavior was investigated in a tumor spheroid model using FITC-labeled Ads and DAPI-stained algae. The results showed that the consortium could enhance tumor penetration by competing Ca²⁺ and disrupting tight cancer intercellular connections (Figure 2L).

Validation of Ad infection and expression in TC-1-hCD46 cells

To verify that the human CD46 receptor introduced into TC-1 cells enables effective adenovirus type 11 (Ad11) infection, we conducted viral uptake and transfection studies in both TC-1-hCD46 and wild-type TC-1 (TC-1-WT) cells. In a 3-h uptake assay, confocal microscopy revealed that FITC-labeled Ad11 were effectively internalized by TC-1-hCD46 cells, whereas minimal fluorescence was detected in TC-1-WT cells, indicating poor virus uptake (Figure S5). Furthermore, a 10-h transfection experiment using Ad11-GFP demonstrated strong GFP expression in TC-1-hCD46 cells, while TC-1-WT cells showed weak signals (Figure S6A). Quantitative PCR further revealed that the Ad11 content in TC-1-hCD46 cells was 434.01 times higher than in TC-1-WT cells (Figure S6B), indicating that CD46 expression substantially enhances Ad11 entry and intracellular replication. These results confirm that the engineered TC-1-hCD46 cells are permissive to Ad11 infection and support productive viral replication, validating their use in this study.

In vivo distribution and antitumor efficacy

Encouraged by the potential of the consortium to effectively overcome in vitro physical barriers, the in vivo tissue distribution studies were conducted in a TC-1-hCD46-Luc orthotopic lung cancer mouse model. The consortium labeled with Cy7 were administered through intratracheal nebulization at a dose of 1 × 10⁷ PFU of Ads, and their distribution was monitored using the in vivo imaging system (IVIS) (Figures 3A and 3B). Two hours post-administration, the consortium exhibited the highest lung deposition rate than those of PEG-Ads and Ads. At 12 h, lung retention rate of the consortium remained at 75.18%, while Ads from other groups were nearly eliminated. This suggested that the inherent motility of algae in the consortium facilitated OVs mucus penetration and reduced OVs clearance. At 24 h, mice were euthanized, and the distribution in major organs was analyzed (Figures 3C and 3D). The inhalable consortium primarily retained in the lungs, whereas systemic consortium predominantly accumulated in liver. The quantification of the Ad content in lung was also performed using real-time fluorescence quantitative PCR (real-time qPCR) (Figure 3E). The results showed significantly high retention (638.59%) for inhalable consortium compared to inhalable Ads alone, while systemic administration had minimal lung retention (29.03%). To further validate these observations, lung sections were analyzed, revealing significant green fluorescence concentrated at tumor sites, which confirmed the effective accumulation of the consortium in lung tumors (Figure 3F; Figure S7).^43,44

Figure 3.

In vivo distribution and efficacy assessment in orthotopic lung cancer mice model

(A) In vivo imaging of the internal distribution of the consortium, PEG-Ads, and Ads at tumor sites, with comparison to intravenous administration (n = 3).

(B) Quantitative analysis of average fluorescence intensity per unit area corresponding to (A), showing lung deposition over time (n = 3).

(C) Distribution of formulations in excised organs, observed via IVIS at 24 h post-administration following euthanasia (n = 3).

(D) Quantitative chart of fluorescence intensity per unit area for each organ (n = 3).

(E) Quantitative fluorescence PCR analysis of Ad retention in lung tissues 12 h post-administration.

(F) CLSM images of lung sections showing distribution of the consortium, PEG-Ads, and Ads, with lung tissues stained by DAPI and Ads labeled with FITC-NHS ester (one representative section from three mice is shown).

(G) Tumor growth and body weight monitoring in TC-1-hCD46-Luc tumor-bearing mice treated with intratracheal nebulization of formulations. Terminal procedures, including organ collection for immune analysis, were conducted on day 10.

(H) IVIS analysis of tumor bioluminescence and fluorescence intensity per unit area in different treatment groups, demonstrating treatment efficacy (n = 6).

Statistical analysis: data are presented as mean ± SD. Statistical significance was calculated using two-tailed Student’s t test (∗∗∗∗p < 0.0001).

The pharmacodynamics studies were conducted in the orthotopic lung cancer mice model, following the treated protocol in Figure 3G. Tumor bioluminescence area and fluorescence intensity were measured during the treatment (Figure 3H). The consortium nearly achieved complete tumor eradication, whereas PEG-Ads and Ads exhibited moderate tumor suppression, and algae alone showed no significant antitumor effect. The consortium-treated mice maintained stable body weight, suggesting minimal systemic toxicity and good preliminary safety (Figure S8). An additional control group without responsive release of Ads showed poor efficacy (Figure S9). This underscores the critical role of responsive Ad release in ensuring effective tumor cell infection and therapeutic success.

Long-term monitoring validates the sustained efficacy of the consortium

To further address the long-term therapeutic potential and the possibility of achieving complete remission, we extended the monitoring period of tumor-bearing mice to 22 days (Figure S10A). Bioluminescence imaging demonstrated that throughout the entire observation period (day 0–day 21), the signals in the Consortium group remained at a minimal level near the baseline, indicating robust and sustained tumor suppression (Figures S10B and S10C). In stark contrast, the tumor signals in the PBS and Algae groups exhibited uncontrolled exponential growth after day 14. This prolonged tumor inhibition directly translated into a significant survival advantage. The Consortium group maintained a 100% survival rate until the experimental endpoint, whereas the PBS group suffered a 40% mortality rate by day 21 due to excessive tumor burden (Figure S10C). Furthermore, stable body weight profiles were recorded across all groups, suggesting no significant systemic toxicity of the Consortium group during long-term treatment (Figure S10D). Ex vivo gross observation of the lungs on day 22 further substantiated these findings. While minimal residual nodules remained visible on the lung surfaces of the Consortium group (indicated by red arrows in Figure S10E), the lungs appeared significantly cleaner than the lungs of control groups, which were entirely occupied by massive, confluent tumor masses. Collectively, these data confirmed that the inhalable consortium therapy not only achieves rapid tumor inhibition but also possesses the clinical potential to induce long-term remission.

Immunological studies

The immune response was evaluated in orthotopic lung cancer mice model, using blood and lung tissue samples collected after treatment (Figure 4). Additional protein immunization experiments were performed in wild-type TC-1 tumor-bearing mice to exclude background immune activation (Figure S11).

Figure 4.

Comprehensive analysis of immune activation and immunosuppression

(A–I) Flow cytometry analysis of immune cell populations in blood and lung tissues (n = 3).

(A and B) CD8⁺ T cells in blood and lung; (C and D) regulatory T cells (Tregs) in blood and lung; (E) dendritic cells (DCs) in lung; (F) memory T cells in blood; (G and H) myeloid-derived suppressor cells (MDSCs) in lung and blood; (I) and polarization of macrophages in lung.

(J–O) Quantification of cytokines in blood and lung tissues. (J) IFN-γ, (K) IL-6, and (L) TNF-α in blood; (M) IFN-γ, (N) IL-6, and (O) TNF-α in lung. Cytokine levels were measured via ELISA (n = 3).

(P) Immunofluorescence analysis of immune infiltration in advanced orthotopic lung tumors. Mice bearing TC-1-hCD46 lung metastases were treated via intratracheal nebulization on day 18, and lungs were harvested on day 20 for sectioning. CD8⁺ (green), CD4⁺ (red), and nuclei (blue) were stained (one representative image).

Statistical analysis: data are presented as mean ± SD. Statistical significance was calculated using two-tailed Student’s t test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001).

The consortium treatment elicited a robust antitumor immune response characterized by significantly elevated CD8⁺ T cell proportions in both peripheral blood and tumor tissues. Specifically, blood CD8⁺ T cells reached 32.9% in the consortium group, markedly higher than those in the PEG-Ads (26.6%), Ads (23.3%), algae (19.4%), and PBS (14.4%) groups (Figures 4A and 4B). Concurrently, regulatory T cells (Tregs) were substantially reduced in both compartments (Figures 4C and 4D), indicating an alleviation of immunosuppressive pressure. This shift was accompanied by enhanced activation of DCs in the lung (Figure 4E) and an expansion of memory T cells in circulation (Figure 4F), suggesting efficient antigen presentation and long-term immune priming.

Importantly, the consortium also suppressed key immunosuppressive populations. Both monocytic and granulocytic myeloid-derived suppressor cells (MDSCs) were significantly decreased in blood and tumor tissues compared to other groups (Figures 4G and 4H). Moreover, macrophage polarization analysis revealed a significant reduction in M2-like phenotypes and a relative increase in M1-like populations (Figure 4I), indicating a shift in the TME toward a pro-inflammatory state. This phenotypic transition was primarily attributed to the immunomodulatory effect of algae, which reprogrammed macrophage function by activating pattern recognition receptors and inducing the secretion of inflammatory cytokines, thereby promoting M1 polarization and strengthening the antitumor immune response.

Cytokine profiling showed that levels of interferon gamma (IFN-γ), interleukin-6 (IL-6), and tumor necrosis factor alpha (TNF-α) in blood and lung tissues exhibited distinct patterns across groups (Figures 4J–4O). IFN-γ levels in blood were highest in the Ad group, indicating potent systemic antiviral immune activation. In contrast, the consortium group showed moderate IFN-γ elevation, suggesting effective shielding from premature antiviral responses. IL-6 and TNF-α levels were markedly elevated in the consortium group, both in blood and lung tissues, supporting enhanced T cell recruitment, macrophage activation, and pro-inflammatory TME remodeling.⁴⁵

To further investigate the immune activation potential in advanced tumors, immunofluorescence experiments were conducted in advanced tumor-bearing mice using the TC-1-hCD46 orthotopic model. Eighteen days post-tumor inoculation, intratracheal nebulization of the consortium, PEG-Ads, Ads, algae, or PBS was administered. Forty-eight hours later, mice were euthanized, and lung tissue sections were stained to detect CD4⁺ (green) and CD8⁺ (red) T cells (Figure 4P). The results indicated that the consortium significantly increased immune cell recruitment and CD8⁺ T cell infiltration. These effects were particularly prominent in the immunosuppressive microenvironment of advanced tumors, emphasizing the ability of consortium to activate antitumor immunity even under challenging conditions.⁴⁶

Consortium-triggered adoptive T cell transfer validates systemic anti-tumor immune function

Since previous immunophenotyping revealed that the consortium treatment significantly increased the infiltration of CD8⁺ T cells in both peripheral blood and lung tissues, these effector T cells serve as the primary drivers of tumor regression and systemic immunosurveillance. To directly evaluate whether the CD8⁺ T cells induced by the consortium possess independent functional anti-tumor efficacy, we designed and executed an adoptive T cell transfer (ACT) assay.

In this experimental scheme, donor mice were first pre-stimulated via intratracheal nebulization of the consortium, followed by the isolation of CD8⁺ T cells from their peripheral blood. These cells were then adoptively transferred into recipient mice that had been pre-established with TC-1 lung tumors (Figure 5A). Subsequent dynamic monitoring via IVIS demonstrated that recipient mice receiving T cells from the consortium-treated group exhibited the weakest fluorescence signals on days 7 and 14 post-transfer compared to the PBS and monotherapy control groups (Figure 5B). Quantitative analysis further confirmed that the average fluorescence intensity in this group increased most slowly (Figure S12), indicating a significant inhibition of malignant tumor expansion in the lungs. Furthermore, the body weights of recipient mice in all groups remained stable throughout the 14-day observation period (Figure S13), suggesting a favorable preliminary biosafety profile for this cellular therapy.

Figure 5.

Study of therapeutic effects in the lung distal metastatic tumor model and safety assessments

(A) Schematic illustration of the ACT experimental design.

(B) Representative in vivo bioluminescence images of recipient mice showing lung tumor progression on days 0, 7, and 14 after receiving CD8⁺ T cells from differently treated donors (n = 5).

(C) Experimental design for the distal metastatic tumor model.

(D) Lung tissues collected on day 21, fixed in Bouin’s solution to visualize pulmonary tumor nodules in different treatment groups (n = 5).

(E) Images of tumors harvested from mice on day 21 post-treatment (n = 5).

(F) Quantification of lung tumor nodules on day 21 for each group (n = 5).

(G and H) Size measurements and tumor weight of distal tumors in different treatment groups on day 21 (n = 5).

(I) Survival curves of the remaining half of the mice (n = 5).

(J) CT images of the axial, coronal, and sagittal planes of healthy mouse lungs 4 days after intervention with different treatment groups (one representative image).

Statistical analysis: data are presented as mean ± SD. Statistical significance was calculated using two-tailed Student’s t test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001).

Collectively, these findings provide compelling evidence that CD8⁺ T cells elicited by the consortium are not only expanded in number but also endowed with potent anti-tumor cytotoxic activity, providing a mechanistic foundation for the sustained systemic immune protection observed in subsequent models.

In vivo antitumor efficacy in lung distal metastatic tumor model

Based on the promising outcomes of the pharmacodynamics studies, a distal metastatic tumor model was further investigated to determine whether treatment-induced antitumor immune activation could result in long-term suppression of distant metastases. The experimental design was shown in Figure 5C. On day 21, half of the mice were euthanized, and their lungs were extracted and fixed in Bouin’s solution to assess pulmonary tumor nodules in different treatment groups. Quantification showed few and small primary nodules in the consortium-treated group, indicating effective inhibition of primary tumor growth and spread (Figures 5D and 5F). The consortium-treated distal tumor sizes and weights were distinctly suppressed, demonstrating the strong immune memory generated by in situ treatment (Figures 5E, 5G, and 5H). Survival analysis of the remaining mice further showed the consortium-treated mice had the highest survival rate (Figure 5I).

H&E staining was conducted to evaluate the safety profile of each treatment group (Figure S14). Additionally, the PBS group showed intact organ structures without any significant pathological changes, serving as a reliable control. The consortium group displayed no toxicity or structural damage to major organs, indicating good tissue compatibility at administered dosages. Minimal fluctuations in body weight further supported the safety of the consortium treatments, indicating no significant systemic toxicity (Figure S15). On day 4, computed tomography (CT) scans showed no significant areas of lung consolidation (Figure 5J). Lung tissues appeared uniformly, with no high-density regions or ground-glass opacities. Coronal and sagittal images showed no inflammatory changes, confirming the absence of lung inflammation and underscoring the safety profile of the consortium treatments.

Antitumor efficacy in human-derived xenograft model (BALB/c nude mice)

A human A549-Luc pulmonary orthotopic model was established in BALB/c nude mice by tail-vein inoculation of A549-Luc lung cancer cells. The inhalable consortium, PEG-Ads, Ads, or algae alone were subsequently administered via intratracheal nebulization following the same dosing schedule as in the murine C57 model (Figure 6A).

Figure 6.

Cross-species validation of the inhalable consortium in human-derived tumor-bearing and humanized mouse models

(A) Experimental design of the A549-Luc pulmonary orthotopic model in BALB/c nude mice.

(B) In vivo bioluminescence imaging showing tumor progression in each treatment group at different time points; representative images of three randomly selected mice per group (n = 5).

(C) Images of excised lungs from each group on day 10.

(D) Quantitative analysis of lung bioluminescence intensity (n = 5).

(E) Statistics of the number of visible pulmonary tumor nodules in each group (n = 5).

(F) Schematic of the human T cell-humanized NOG mouse model.

(G and H) Flow cytometry analysis of human CD8⁺ T cells in lung and blood (n = 3).

Statistical analysis: data are presented as mean ± SD. Statistical significance was calculated using two-tailed Student’s t test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001).

Bioluminescence imaging revealed that the consortium exhibited the most potent inhibition of lung tumor burden, with luminescent intensity markedly reduced compared with the PBS group (Figures 6B–6E). Quantitative analysis further showed a tumor inhibition rate of approximately 95%, confirming near-complete suppression of pulmonary tumor growth. Mice in the consortium group showed markedly fewer metastatic nodules, whereas those in the algae and PBS groups exhibited extensive tumor dissemination or mortality. Because BALB/c nude mice lack functional T cells, this result primarily reflects the direct oncolytic and local cytotoxic effects of the inhalable consortium rather than adaptive immune activation. No significant body-weight loss was observed in the consortium-treated mice (Figure S16), confirming the biosafety of the inhalable formulation.

Collectively, these data verify that the inhalable consortium maintains strong antitumor efficacy against human-derived lung cancer, supporting its translational applicability across species for human.

Immune activation in human T cell-humanized NOG mice

To further evaluate the role of adaptive immunity, a human T cell-humanized NOG mouse model was established. Human A549-Luc lung cancer cells were first injected via the tail vein, followed by intravenous infusion of PBMC-derived, IL-2–expanded human T cells to reconstruct a functional human-derived immune system. Treatments and sampling were performed as illustrated in Figure 6F.

Flow cytometry analysis demonstrated that the consortium elicited the strongest cytotoxic T cell-mediated antitumor immune activation in both lung and peripheral blood, consistent with results observed in the C57BL/6 tumor model. This finding confirms that the consortium effectively promotes human CD8⁺ T cell activation and infiltration across species, validating its translational potential as a clinically relevant oncolytic microorganism platform (Figures 6G and 6H).

These results demonstrated that the inhalable consortium effectively recruited and activated human cytotoxic T cells across species, validating that its therapeutic mechanism extends beyond direct oncolysis to include robust human T cell-mediated immune responses. This cross-species immune activation underscored the translational potential of the consortium as a clinically relevant oncolytic microorganism platform.

Discussion

OV therapies are a promising form of cancer immunotherapy that selectively infect and lyse tumor cells while sparing normal cells. This process leads to the release of tumor antigens and the activation of antitumor immunity. Given the challenges in developing treatment strategies for solid tumors, OV therapies have advantages as one multimodal cancer immunotherapy. However, they are also facing obstacles in clinical development. Currently available OVs formulations are administered via intratumoral injection, making them difficult to apply for deep-seated tumors, especially lung cancer.

This study aimed to overcome these limitations by developing an inhalable microbe-OVs consortium, which synergistically combined the benefits of microbe and OVs. A major breakthrough was achieved in targeting capability, as the consortium utilized their motility of algae to evade respiratory mucosal clearance and enhance deposition in alveoli. The motility further allowed the consortium to traverse the alveolar barrier. Another advantage of the consortium was their ability to retain the competitive absorption of Ca²⁺ from tumor cells. E-cadherin was degraded through this competition, facilitating the penetration of the consortium into the tumor interior. Then Ads were responsively released within the tumor, avoiding insufficient Ads penetration. Furthermore, the tumor antigens released after OV-induced lysis were presented by DCs that were coactivated by algae and OVs. Overall, the inhalable consortium showed great potential for expanding the applications of OVs, particularly in the treatment of deep-seated tumors like lung cancer.

In contrast to conventional OVs formulations such as PEGylated viruses or encapsulation systems, the microbe-assisted strategy developed in this study introduces a behaviorally regulated delivery paradigm, enabling a shift from passive diffusion to active targeting. The algae in the consortium carry PAMPs, which, in combination with tumor antigen release from virus-mediated oncolysis, promote the activation and maturation of M1-type macrophages and DCs. This facilitates efficient antigen presentation and drives CD8⁺ T cell infiltration and long-term immune memory. Such dual synergy—physical penetration and immune activation—provides a framework for improving therapeutic response rates in deep-seated tumors.

This system also offers broad potential for future development. Its modular design allows for substitution or co-delivery of various viruses, therapeutic proteins, or immunomodulators. It could also be combined with immune checkpoint inhibitors (e.g., anti-PD-1/PD-L1 antibodies), STING agonists, or tumor neoantigen vaccines to further augment systemic immune responses. From a translational perspective, the platform demonstrated good tunability and programmability, supporting personalized design based on lesion type and immune context. However, further investigation is required to assess long-term biosafety, biodegradability, and host-microbe interactions. Synthetic biology tools such as controllable lysis or retrieval circuits may be introduced in future versions to improve clinical predictability and control.

Limitations of the study

A key limitation is the absence of patient-derived xenograft (PDX) models, which more accurately reflect the heterogeneity of human lung tumors. Additionally, while rodent models provided essential proof of concept, physiological differences necessitate further validation in large animal models or clinical trials to evaluate long-term safety and bio-distribution. Future research will prioritize these advanced systems to further substantiate the clinical translatability of this therapeutic platform.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jin Sun (sunjin@syphu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •This paper does not report original (genomic/sequencing) data.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

This work was financially supported by National Key Research and Development Program of China (no. 2021YFA0909901), the National Natural Science Foundation of China (nos. 82073777, 81803442, 82372111, and 82404557), the Liaoning Revitalization Talents Program (no. XLYC22202019), the Liaoning Province Youth Science Foundation Class B (formerly Provincial Outstanding Youth Fund) (no. 2025-YQ-16), and the Outstanding Youth lifting Program in the Shenyang Pharmaceutical University (YQ202301).

Author contributions

Conceptualization, M.S., J.S., and F.L.; methodology, H.L., J.S., F.L., Z.C., X.S., M.N., and W.W.; performed experiments, H.L., S.Y., M.N., and W.W.; data analysis, H.L. and S.Y.; writing – review and editing, all authors.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-mouse CD45 antibody (APC-Cy7)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 147717; RRID: AB_2888795
Anti-mouse CD3 antibody (FITC)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 100204; RRID: AB_312661
Anti-mouse CD4 antibody (APC)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 100412; RRID: AB_312697
Anti-mouse CD8 antibody (PerCP-Cy5.5)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 100734; RRID: AB_2075238
Anti-mouse CD25 antibody (BV421)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 102043: RRID: AB_2562611
Anti-mouse F4/80 antibody (PerCP-Cy5.5)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 157318; RRID: AB_2894430
Anti-mouse CD11b antibody (FITC)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 101206; RRID: AB_312789
Anti-mouse CD206 antibody (APC)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 141712; RRID: AB_10900420
Anti-mouse CD80 antibody (PE)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 104708; RRID: AB_313129
Anti-mouse CD11c antibody (FITC)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 117306; RRID: AB_313775
Anti-mouse CD86 antibody (APC)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 105114; RRID: AB_313167
Anti-mouse GR-1 antibody (BV421)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 108445; RRID: AB_2562903
Anti-mouse CD62L antibody (APC)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 104412; RRID: AB_313099
Anti-mouse CD44 antibody (PE)	BioLegend (USA), distributed by Univ (Shanghai, China)	Cat# 103008; RRID: AB_312959
Anti-mouse FOXP3 (PE, Intracellular)	BD Biosciences (New Jersey, USA)	Cat# 560401; RRID: AB_1645201
Anti-mouse CD206 (AF647, Intracellular)	BD Biosciences (New Jersey, USA)	Cat# 565250; RRID: AB_2739133
Anti-E-cadherin	ABclonal (Wuhan, China)	Cat# A20798; Clone Number: ARC51012
Anti-β-actin antibodies	ABclonal (Wuhan, China)	Cat# AC026; Clone Number: ARC5115-01
Anti-Human CD45 (FITC)	Starter Biotech (Hangzhou, China)	Cat#S0B1515; Clone Number: S-839-3
Anti-Human CD3 (Alexa Fluor 647),	Starter Biotech (Hangzhou, China)	Cat# S0B8251; Clone Number: S-R476
Anti-Human CD8a (PE)	Starter Biotech (Hangzhou, China)	Cat#S0B1634; Clone Number: S-1036-34
Bacterial and virus strains
Oncolytic Adenovirus Ad11-5ETel-GFP	Beijing Bio-Targeting Therapeutics	N/A
Synechococcus WH8102	Roscoff Culture Collection	Cat# RCC539
Synechococcus PCC7942	Institute of Hydrobiology, CAS	Cat# FACHB-805
Cell lines
TC-1-hCD46	Zhengzhou University (Zhengzhou, China)	N/A
TC-1-hCD46-Luc	Zhengzhou University (Zhengzhou, China)	N/A
JH-293	Zhengzhou University (Zhengzhou, China)	N/A
Wild-type TC-1	ICell Bioscience (Shanghai, China)	N/A
A549-Luc	Wuhan Sunncell (Wuhan, China)	N/A
Human PBMCs	Wuhan Sunncell (Wuhan, China)	N/A
MH-S (Murine alveolar macrophages)	Beijing ittabio (Bei Jing, China)	N/A
Chemicals, peptides, and recombinant proteins
DMEM	Meilunbio (Dalian, China)	Cat# MA0212
RPMI-1640 Medium	Meilunbio (Dalian, China)	Cat# MA0215
T cell expansion medium	Meilunbio (Dalian, China)	Cat# PWL098
Fetal Bovine Serum (FBS)	Meilunbio (Dalian, China)	Cat# PWL001
Recombinant human IL-2	Meilunbio (Dalian, China)	Cat# MB5951
D-luciferin	Meilunbio (Dalian, China)	Cat# MB1834
PCR-S11-Red Sea medium	(Self-prepared, various components)	N/A
BG11 medium	(Self-prepared, various components)	N/A
Matrigel	Meilunbio (Dalian, China)	Cat# PWL300
Methylcellulose	Meilunbio (Dalian, China)	Cat# MB3040
RIPA Lysis Buffer	Meilunbio (Dalian, China)	Cat# MA0151
DBCO-PEG2000-S-S-NHS	Beijing ittabio (Bei Jing, China)	Custom order
DBCO-PEG2000-NHS	Beijing ittabio (Bei Jing, China)	Custom order
Azido-PEG2000-NHS	Beijing ittabio (Bei Jing, China)	Custom order
Human CD3/CD28 T cell activation beads	BEAVER (Suzhou, China)	Cat# 70910-1000
Collagenase IV	Dingguo (Beijing, China)	Cat# DH074
BCA Protein Assay Kit	Dingguo (Beijing, China)	Cat# BCA01
SDS-PAGE Protein Marker	Meilunbio (Dalian, China)	Cat# MA0342
MTT Solution	Meilunbio (Dalian, China)	Cat# MB4698
Fluo-4 AM	Beyotime (China)	Cat# S1060
Recombinant Human CD46 Protein	Aladdin Biochemical (Shanghai, China)	Cat# P15529-11
Recombinant Photinus pyralis Luciferase Protein	Aladdin Biochemical (Shanghai, China)	Cat# P08659
c-Ha-Ras proteins	Aladdin Biochemical (Shanghai, China)	Custom order
Simulated Lung Fluid Components (DPPC, DPPG, GSH, etc.)	(Self-prepared, various components)	This paper
Artificial Mucus Components (Mucin, DNA, DTPA, Egg yolk)	(Self-prepared, various components)	This paper
Triton X-100	Aladdin Biochemical (Shanghai, China)	Cat# T434386
Tween 80	Aladdin Biochemical (Shanghai, China)	Cat# T485985
FITC-NHS ester	Macklin (Shanghai, China)	Cat# C804900
Cy7-NHS ester	Meilunbio (Dalian, China)	Cat# MB12199
DAPI	Meilunbio (Dalian, China)	Cat# MA0128
DiI	Aladdin Biochemical (Shanghai, China)	Cat# D266337
4% Paraformaldehyde	Meilunbio (Dalian, China)	Cat# MA0192
SYBR Green PCR Master Mix	Vazyme Biotech (Nanjing, China)	Cat# Q311-02/03
Tert-butanol	Aladdin Biochemical (Shanghai, China)	Cat# T110352
Gelatin	Aladdin Biochemical (Shanghai, China)	Cat# 232-554-6
DMSO	Aladdin Biochemical (Shanghai, China)	Cat# D670381
Tribromoethanol	Meilunbio (Dalian, China)	Cat# MA0478
Ferric oxide nanoparticles (Fe3O4)	Aladdin Biochemical (Shanghai, China)	Cat# I109514
Critical commercial assays
IFN-γ ELISA Kits	Jianglai Bio (Shanghai, China)	JLW10967
IL-6 ELISA Kits	Jianglai Bio (Shanghai, China)	JLW20268
TNF-α ELISA Kits	Jianglai Bio (Shanghai, China)	JLW10484
Mouse lymphocyte separation kit	Solarbio (Beijing, China)	Cat# P8620
Mouse CD8+ T cell Isolation Kit	BEAVER (Suzhou, China)	Cat# 70902-50
Total DNA Extraction Kit	Dingguo (Beijing, China)	Cat# NEP062-1
Experimental models
C57BL/6JOlaHsd mice	Liaoning Changsheng Biotech	N/A
BALB/c nude mice	Liaoning Changsheng Biotech	N/A
NOG mice	Liaoning Changsheng Biotech	N/A
Software and algorithms
Living Image 4.4	PerkinElmer	N/A
Slide Viewer 2.6	3DHISTECH	N/A
ImageJ 1.53	NIH (USA)	N/A
FlowJo 10.8.1	BD Biosciences (New Jersey, USA)	N/A
GraphPad Prism 10.1	GraphPad	N/A
NIS-Elements Viewer 4.0	Nikon	N/A
Hardware and others
Dynamic Light Scattering	Malvern Instruments (UK)	Zetasizer Nano ZS
Transmission Electron Microscope	JEOL (Japan)	JEM-2100
Carbon-coated copper grid	EMCN (Beijing, China)	Cat# BZ110215
Visualized small-animal microspraying device	YSKD Biotechnology (Beijing, China)	N/A
Next-Generation Impactor (NGI)	Mingshi Yang group, SYPU	N/A
Flow cytometer	BD Biosciences (New Jersey, USA)	FACSCelesta
Microplate Reader	Thermo Scientific (USA)	Multiskan FC
Confocal laser scanning microscope	Nikon (Tokyo, Japan)	A1R
IVIS Imaging System	PerkinElmer	IVIS Lumina III
100 kDa, 30 kDa Ultrafiltration tubes	(Various, e.g., Amicon/Millipore)	N/A
PVDF membranes ( 0.45 μm)	Merck (Beijing, China)	Cat# IPFL00010
70 μm nylon mesh	Biosharp (Beijing, China)	BS-70-XBS
Oligonucleotides
Primer: Ad11 E1A Forward/Reverse	BGI Genomics (Shenzhen, China)	Forward:TTGGACGGCTCCTGGAATAG; Reverse:TGTGACGGAAACAACCCTGACT

Experimental model and study participant details

Cell lines

TC-1-hCD46 and JH-293: The murine non-small cell lung cancer (TC-1) cell line expressing human CD46 and the JH-293 cells were provided by the Sino-British Research Center at Zhengzhou University (Zhengzhou, China).

Wild-type TC-1: Purchased from ICell Bioscience Inc. (Shanghai, China).

A549-Luc and Human PBMCs: Human A549-Luc cells and primary human PBMCs were purchased from Wuhan Sunncell Co., Ltd. (Wuhan, China).

Cultivation: All TC-1 variants and A549-Luc cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS). Cells were maintained in a humidified incubator at 37°C with 5% CO₂.

T cell Expansion: Primary human T cells derived from PBMCs were expanded in serum-free T cell culture medium. Expansion was facilitated using human CD3/CD28 T cell activation magnetic beads supplemented with 150 IU/mL recombinant human IL-2. Activated T cells were harvested after 5 days of expansion.

Genetic Authentication: Short Tandem Repeat (STR) genotype analysis was performed on all cells involved in the experiments to verify their identity.

Mycoplasma Detection: All cell lines were subjected to mycoplasma contamination testing to ensure that the experimental process was not compromised by microbial interference.

Viruses and bacteria

Oncolytic Adenovirus Ad11-5ETel-GFP (Ads): The recombinant virus was obtained from Beijing Bio-Targeting Therapeutics Technology Inc. (Beijing, China). All viral experiments followed the supplier’s standard operating protocols to ensure reproducibility.

Synechococcus WH8102: This motile marine cyanobacteria strain was purchased from the Roscoff Culture Collection (Roscoff, France).

Synechococcus PCC7942: This non-motile strain was sourced from the Institute of Hydrobiology, Chinese Academy of Sciences (Wuhan, China).

Cultivation: Synechococcus WH8102 was cultured in PCR-S11-Red Sea medium at 20°C under a 12 h light/dark cycle. Synechococcus PCC7942 was maintained in BG11 medium at 25°C with the same 12 h light/dark cycle.

Experimental animals

Strains: Male C57BL/6JOlaHsd, BALB/c nude, and NOG mice (6–8 weeks old, weighing 18–22 g) were used.

Source: All animals were purchased from Liaoning Changsheng Biotechnology Co., Ltd. (Benxi, Liaoning, China).

Housing: Mice were housed in a specific pathogen-free (SPF) facility with a 12 h light/dark cycle and provided with ad libitum access to food and water.

All animal procedures were approved by the Ethics Committee of the First Affiliated Hospital of China Medical University (CMU2021722, CMU20253292) and Shenyang Pharmaceutical University (SYPU-IACUC-S2024-0621-102). No evidence suggested that the gender or sex of the animals influenced the experimental outcome.

Method details

Construction of recombinant adenovirus Ad11-5ETel-GFP

The recombinant adenovirus Ad11-5ETel-GFP (Ad11) used in this study is based on the Ad11-5EP backbone. This backbone was constructed by substituting a 365 bp fragment comprising the enhancer and promoter of the Ad5 E1A gene (specifically bp 195–559) for the corresponding upstream region of the Ad11 E1A coding sequence to enhance early gene expression. To enable tumor-specific replication, the Ad5 E1A promoter within this chimeric backbone was subsequently replaced by a human telomerase reverse transcriptase (hTERT) promoter (−714 to 0 bp). In addition, an enhanced green fluorescent protein (EGFP) expression cassette was inserted into the E3 region, specifically replacing the 18.5K gene (ΔE3 18.5K), to facilitate visualization of viral infection. A detailed schematic of the virus construction logic and genome structure was illustrated in Figure S17. The illustrations and construction methods were cited from European Patent EP2716764A1.

Virus titration

The titration of Ads was conducted according to the cytopathic effect (CPE) method.⁴⁷ Briefly, JH-293 cells were seeded in 96-well plates at a density of 3,000 cells per well in DMEM medium containing 2% FBS and incubated overnight for adherence. The adenovirus samples were first diluted in DMEM medium containing 2% FBS and added to the first row of wells. Serial 10-fold dilutions were then performed across the plate, and the last row was left without virus as a blank control. The plates were incubated for 7–10 days. After incubation, each well was examined under a microscope. Wells exhibiting cytopathic effects were recorded as positive, while those without CPE were recorded as negative. The viral titer was calculated using the Reed–Muench method.

Preparation of Microbe-OVs consortium

The microbe-oncolytic virus consortium was synthesized via a click reaction. Briefly, 1.2 mg of DBCO-PEG₂₀₀₀-S-S-NHS ester was reacted with 1.124 × 10⁸ PFU of Ad11-5ETel-GFP in 1 mL of PBS (pH 7.2) for 1 h. Concurrently, 2.4 mg of Azido-PEG2000-NHS ester was incubated with 1.124 × 10⁷ PFU of Synechococcus (WH8102 or PCC7942) under identical conditions. Unreacted reagents were removed using 100 kDa ultrafiltration tubes 10 via centrifugation (3000 rpm, 5 min). The modified Ads and algae were then mixed under gentle stirring for 1 h. The resulting consortium was collected by centrifugation (3000 rpm, 5 min) and resuspended to remove unbound components.

Characterization of the consortium

The particle size distribution, surface charge, and structural morphology of the microbe-oncolytic virus consortium were characterized using dynamic light scattering (DLS), TEM, and CLSM.

DLS Measurements: Measurements were performed on a Malvern Zetasizer Nano ZS. Samples were diluted in PBS (pH 7.4) to a final concentration of approximately 10⁷ PFU/mL and measured at 25°C with a scattering angle of 173°. Each sample was measured in triplicate, and the zeta potential was recorded simultaneously.

TEM Imaging: Imaging was conducted using a JEM-2100 transmission electron microscope operated at an acceleration voltage of 200 kV. 10 μLs of the diluted sample were dropped onto a carbon-coated copper grid, negatively stained with 2% (w/v) phosphotungstic acid (pH 7.0), and air-dried before imaging.

CLSM Observation: For CLSM observation, algae were labeled with DAPI and DiI, while Ads were labeled with FITC-NHS ester. The dual-labeled samples were dropped onto coverslips, fixed with 4% paraformaldehyde for 10 min, and gently rinsed with PBS. Fluorescence images were captured under a Nikon A1R confocal laser scanning microscope using a 100× oil immersion objective. Images were analyzed using NIS-Elements Viewer.

Real-time qPCR assay for Ads detection

Real-time Quantitative PCR (qPCR) was employed to quantify adenoviral DNA by targeting the E1A gene of Ads. The following primer sequences, synthesized by BGI Genomics, were used:

Forward primer: TTGGACGGCTCCTGGAATAG.

Reverse primer: TGTGACGGAAACAACCCTGACT.

Each 20 μL PCR reaction contained 10 μL of SYBR Green PCR Mix, 0.5 μL of each primer (10 μM), 2 μL of template DNA, and 7 μL of nuclease-free water. The thermocycling protocol was as follows:

Initial denaturation: 95°C for 5 min.

40 cycles of: 95°C for 15 s, 60°C for 60 s.

Melting curve analysis: 95°C for 15 s, 60°C for 60 s, 95°C for 10 s. A melting curve was used to verify the specificity of amplification.

Quantification of Ads conjugation efficiency via real-time qPCR

To determine the conjugation efficiency of Ad11 within the Microbe-OVs consortium, viral DNA content in both the conjugated fraction (pellet) and the unbound fraction (supernatant) was quantified. During the preparation, the consortium was isolated by centrifugation, and the pellet and supernatant were collected separately. Each sample was treated with 0.3% Triton X-100, diluted to a final volume of 1 mL with nuclease-free water, and boiled at 100°C for 15 min to lyse viral particles and release DNA. Following centrifugation at 12,000 rpm for 10 min, the resulting supernatants were used as templates for qPCR analysis. The conjugation rate was calculated using the following formula: Conjugation Rate (%) = (Viral DNA in pellet/Total viral DNA in pellet and supernatant) × 100%

Assessment of algae loading rate by flow cytometry

To evaluate the loading efficiency of Ad11-5ETel-GFP onto algae, flow cytometry was performed using fluorescently labeled virus. Ad11 was conjugated with FITC-NHS and incubated with algae for 15 min at room temperature. To remove unbound virus, the mixture was washed three times via centrifugation at 4000 rpm for 5 min using a 30 kDa ultrafiltration tube. Unmodified algae were used as a negative control to establish baseline autofluorescence. Samples were analyzed using a BD flow cytometer (BB515 channel, PMT voltage: 555 V). Each group was tested in triplicate (n = 3). The loading rate was calculated as the percentage of FITC-positive algae relative to the total population using FlowJo software.

Evaluation of virus release in simulated lung fluid by real-time qPCR

The release kinetics of Ad11 from the consortium were evaluated in simulated lung fluid (SLF). The consortium was prepared using DBCO-PEG₂₀₀₀-S-S-NHS as the redox-responsive linker. A parallel group using DBCO-PEG₂₀₀₀-NHS (non-cleavable) served as a control. Both groups were incubated in SLF at 37°C. The SLF was self-prepared with the following concentrations: 4.8 mg/mL DPPC, 0.5 mg/mL DPPG, 0.1 mg/mL cholesterol, 8.8 mg/mL albumin, 2.6 mg/mL IgG, 1.5 mg/mL transferrin, 140 μM ascorbate, 95 μM urate, 200 μM glutathione, 1 μL/mL gentamicin, and 77.5 μL/mL HBSS (pH 7.4). At predetermined time points, the supernatant and pellet were separated by centrifugation. Viral DNA in each fraction was quantified by Real-time qPCR. The released fraction was defined as the percentage of Ad11 detected in the supernatant relative to the total viral content. All conditions were tested in triplicate (n = 3).

NGI analysis

The aerodynamic particle size distribution (APSD) was compared among three groups: Ads, PEG-Ads, and the consortium. Tests were conducted using a Next-Generation Impactor (NGI). NGI collection plates were pre-coated with a 10% (v/v) Tween 80-ethanol solution to prevent particle bounce. All samples were subjected to gradient elution with 25%, 50%, 75%, and 100% tert-butanol, followed by lyophilization. During the assay, the inlet flow rate was set to 60 mL/min under 4 kPa pressure, maintaining a pressure ratio (P3/P2) of 0.5 or less. After a 4-s aerosolization, stages were rinsed with ultrapure water, concentrated to 1 mL using 30 kDa ultrafiltration tubes, and analyzed via Real-time qPCR.

Analysis of movement rate and trajectories of the consortium

To visualize and quantify the motility of the Microbe-OVs consortium, algae were pre-labeled with DiI. A hanging drop technique was employed for sample preparation. Briefly, 5 μL of the labeled consortium suspension was placed at the center of a clean coverslip. Vaseline was applied to the four corners of the coverslip to prevent evaporation and ensure chamber stability. A glass slide was then gently placed over the coverslip to seal the drop and form a viewing chamber.

Fluorescent images and time-lapse videos were acquired using a Nikon A1R confocal laser scanning microscope equipped with a 100× oil immersion objective. Videos were recorded for a duration of 20 s to capture the dynamic swimming behavior of the consortium.

Post-acquisition, the videos were processed using ImageJ software. Each video was split into 21 consecutive frames at 1-s intervals. Movement trajectories were analyzed using the MTrackJ plugin. The relative movement rate was calculated based on the displacement per unit time. Each condition was analyzed in triplicate (n = 3).

Mucus layer permeation assay

The mucus layer permeation assay was by creating a bio-mimetic respiratory barrier as detailed below. Briefly, artificial mucus was prepared by dissolving 250 mg mucin, 500 mg DNA, 0.295 mg DTPA, 250 μL egg yolk emulsion, 250 mg NaCl, 110 mg KCl, and 1 mL RPMI 1640 medium in DNase-free water to a final volume of 50 mL. The mixture was equilibrated at 25°C for 2 h.

To prepare the assay chamber, a layer of 10% gelatin gel (4 mm thickness) was cast in the bottom of each well of a 24-well plate and solidified at 37°C. Subsequently, a 2 mm thick layer of artificial mucus was overlaid on the gelatin. DiI-labeled algae (1 × 10⁸ CFU in 10 μL) were gently applied to the center of the mucus surface. After incubation at 37°C for 12 h, the fluorescence intensity in the gelatin layer was measured using a multifunctional enzyme labeler (excitation wavelength 549 nm/emission wavelength 565 nm).

Blank gelatin served as the baseline. A group where 1 × 10⁸ CFU algae were directly deposited onto the gelatin layer (without mucus) was defined as 100% permeation. Relative penetration rates were calculated accordingly. For qualitative visualization, z stack confocal laser scanning microscopy was performed (25 micrometers per slice, total depth 500 micrometers). 3D reconstruction and fluorescence signal analysis were conducted using NIS-Elements Viewer software.

Assessment of macrophage clearance evasion by the consortium

The macrophage clearance evasion ability of the consortium was evaluated by co-culturing the consortium with MH-S murine alveolar macrophages.³⁹ MH-S murine alveolar macrophages were cultured in RPMI-1640 medium supplemented with 10% FBS. Cells were maintained at 37°C in a humidified incubator with 5% CO2. For the assay, macrophages were seeded into 24-well plates at a density of 1×10⁵ cells per well and allowed to adhere overnight.

To initiate the experiment, algae and the consortium were added to the respective wells at a 1:1 volume ratio with the culture medium. The co-cultures were incubated at 37°C in the dark to simulate physiological conditions. At predetermined time points (0, 4, 12, 24, and 48 h), the optical density at 680 nm (OD680) of the culture supernatant was measured using a Multiskan FC microplate reader to determine the concentration of residual algae.

Wells containing algae only (without macrophages) served as control groups to account for spontaneous sedimentation or decay. The clearance rate was calculated by comparing the OD680 values between the experimental groups and their corresponding controls. All conditions were tested in triplicate(n = 3).

Characterization of algae-induced diffusion in matrix gel

To assess the motility-driven diffusion behavior of algae, a matrix gel model was constructed. Matrigel was mixed with DMEM medium at a 1:1 ratio (v/v), and DiI-labeled TC-1-hCD46 cells were suspended within the mixture at a density of 1×10⁶ cells/mL. The cell–gel mixture (200 μL per well) was dispensed into a 24-well plate and maintained at room temperature during operation to prevent premature gelation.

Prior to solidification, 4 μL of PBS containing 1 × 10⁷ CFU of Cy7-labeled WH8102 algae was gently injected into the center of the matrix using a micropipette. After gelation, the plate was incubated at room temperature, and the distribution of algae was assessed at 2 and 12 h post-injection.

Three-dimensional imaging was performed using a Nikon A1R confocal laser scanning microscope (Nikon, Japan). z stack acquisition was conducted at 10 μm intervals for a total of 21 steps (210 μm depth), and Cy7 fluorescence was used to trace algae movement. Image reconstruction and fluorescence distribution analysis were performed using NIS-Elements Viewer software.

Evaluation of competitive calcium uptake and disruption of tumor cell tight junctions

To investigate the competitive absorption of calcium ions by algae and its impact on tumor cell intercellular junctions, a co-culture assay was performed using TC-1-hCD46 cells. Sterile 14 mm glass coverslips were placed in 24-well plates, and 1×10⁵ cells were seeded per well. After 12 h of incubation, the medium was replaced with serum-free DMEM, and 1 × 10⁶ CFU of WH8102 algae were added. The cells were co-cultured at 37°C with 5% CO₂ for 12 h without agitation.

For calcium imaging, cells were washed three times with PBS and stained with 2 microM Fluo-4, AM in serum-free medium for 30 min at 37°C under light-protected conditions. Intracellular calcium fluorescence was immediately captured using a Nikon A1R confocal laser scanning microscope.

In parallel, tight junction disruption was assessed via Western blot. Cells were lysed using RIPA buffer supplemented with protease inhibitors. Total protein concentration was determined using a BCA Protein Assay Kit. Protein samples (20 microg each) were separated by SDS-PAGE at a constant voltage of 120 V. Proteins were transferred onto PVDF membranes using a semi-dry blotting system for 90 min. E-cadherin levels were analyzed using specific antibodies diluted according to the manufacturer’s instructions. Signals were detected using ECL substrate and visualized on a chemiluminescence imaging system.

Cytotoxicity assessment on TC-1-hCD46 cells

To assess the cytotoxicity of Ads and PEG-Ads, TC-1-hCD46 cells were seeded at a density of 1×10⁴ cells per well in 96-well plates (100 μL per well). After 12 h of incubation, the medium was replaced with serum-free DMEM, and 1 × 107 PFU of either Ads or PEG-Ads (per well) was added. The formulations were diluted in serum-free DMEM prior to addition. Cells were co-cultured for 12 h at 37°C in a humidified incubator with 5% CO₂.

Following treatment, the culture medium was removed, and 20 μL of MTT solution (5 mg/mL in PBS) was added to each well. After incubation at 37°C for 4 h, the supernatant was discarded, and 150 μL of DMSO was added to dissolve the formazan crystals. Absorbance was measured at 570 nm using a Multiskan FC microplate reader. Each group was tested in triplicate (n = 3), and cell viability was expressed relative to untreated control wells.

Assessment of tumor cell adhesion

To evaluate the adhesion efficiency of different formulations to tumor cells, TC-1-hCD46 cells were seeded at a density of 1×10⁵ cells per well onto sterile 24-well glass-bottom plates. After 12 h of incubation, the medium was replaced with serum-free DMEM. Three experimental groups were added at a dose equivalent to 1 × 10⁷ PFU per well (based on Ads content): FITC-NHS-labeled Ads, FITC-NHS-labeled PEG-Ads and DAPI-prestained algae formulated within the Microbe-OVs consortium.

After 30 min of incubation at 37°C in a 5% CO₂ incubator, the wells were gently washed three times with PBS to remove unbound materials. Cells were fixed with 4% paraformaldehyde for 10 min at room temperature and then stained with ready-to-use DAPI for 15 min in the dark.

Fluorescence imaging and morphological observation were performed using a Nikon A1R confocal laser scanning microscope equipped with a 100× oil immersion objective. Fluorescence signals (FITC and DAPI channels) and cell morphology (bright-field, TD channel) were captured simultaneously using NIS-Elements software. Image analysis was conducted using ImageJ v1.53t. The total fluorescence intensity of the FITC channel was measured and normalized to the number of DAPI-stained nuclei to calculate the mean fluorescence signal per cell. All conditions were tested in triplicate (n = 3).

Evaluation of tumor spheroid penetration

Tumor spheroids were prepared using a hanging drop method. TC-1-hCD46 cells were enzymatically dissociated and resuspended in complete culture medium. A 0.48% (w/v) methylcellulose solution (Dalian Meilun Biotech, China) was pre-prepared and mixed 1:1 (v/v) with the cell suspension to enhance surface tension. The final cell concentration was adjusted to 10,000 cells per 20 μL.

A 20 μL aliquot of the resulting high surface tension cell suspension was carefully pipetted onto the inner side of a 96-well round-bottom plate lid. Drops self-aggregated into spherical structures due to surface tension. To maintain humidity and minimize edge effects, PBS was added to the outer ring of wells on the plate. The lid was then gently placed onto the bottom plate, inverted, and incubated at 37°C in a 5% CO₂ incubator for 24 h to allow spheroid formation.

After 24 h, the lid was returned to upright position, and 200 μL of complete medium was added to each well. The PBS in the peripheral wells was removed, and the plate was sealed with parafilm. Centrifugation was performed at 2800 rpm for 3 min to collect spheroids into the wells. The spheroids were then cultured for an additional 12 h before treatment.

To evaluate penetration, spheroids were treated with free Ads (1 × 10⁷ PFU per well), free algae (1 × 10⁶ CFU per well), or the consortium containing both components at equivalent doses. Penetration behavior was observed at 6 and 12 h post-treatment using a Nikon A1R confocal laser scanning microscope (Nikon, Japan). FITC-NHS ester-labeled Ads and DAPI-stained algae were used to visualize the components. z stack images were acquired at 10 μm intervals for a total depth of approximately 200–250 μm.

Establishment of TC-1-hCD46-Luc orthotopic lung cancer model

To establish an orthotopic lung cancer model, 1.66 × 10⁶ TC-1-hCD46-Luc cells were resuspended in 125 μL of sterile PBS and injected into the tail vein of C57BL/6 mice. Following intravenous injection, the tumor cells spontaneously colonized the lungs and formed pulmonary tumors. This approach mimics clinical lung tumor development and enables noninvasive monitoring through bioluminescence imaging, as the engineered TC-1-hCD46-Luc cells co-express human CD46 and firefly luciferase.

Given that CD46 is the primary cellular receptor for Ad11 and the endogenous murine ortholog cannot mediate efficient viral entry, the introduction of human CD46 enables species-crossing infection, better simulating the clinical Ad11–CD46 interactions observed in patients. This model allows for the evaluation of Ad11 infection specificity via human CD46 and facilitates real-time in vivo tracking of tumor burden. The tail vein injection modeling approach is consistent with previously reported orthotopic lung cancer protocols.⁴⁸

In vivo distribution and retention analysis

The orthotopic lung tumor model was established via tail vein injection of 1.66×10⁶ TC-1-hCD46-Luc cells per mice. On day 7 post-inoculation, tumor-bearing C57BL/6 mice were randomly assigned to four groups (n = 3) and treated via pulmonary inhalation with free Ads (1 × 10⁷ PFU), free algae (WH8102, 1 × 10⁶ CFU), or the consortium containing both components at the same respective doses, while a negative control group received intravenous administration of the consortium. For fluorescence-based biodistribution analysis, Ads were labeled with Cy7-NHS ester prior to formulation. Intratracheal nebulization was performed using a visualized microspraying device (YSKD Biotechnology). In vivo fluorescence images were acquired at 2, 6, and 12 h post-administration using an IVIS Lumina III imaging system, and the average fluorescence intensity per unit area was calculated using Living Image software to evaluate lung deposition and clearance kinetics. At 24 h, mice were euthanized, and major organs including the heart, liver, spleen, lungs, and kidneys were excised for ex vivo imaging and fluorescence quantification.

In a separate experiment to quantify viral retention, mice received the same treatments via inhalation. At 12 h post-administration, lung tissues were harvested and subjected to total DNA extraction. Quantification of Ads in lung tissues was performed by Real-time qPCR targeting the E1A gene.

Analysis of pulmonary drug distribution and mucosal permeation

To evaluate pulmonary distribution and mucosal permeation of different formulations, orthotopic lung cancer models were established as described in the section “Establishment of TC-1-hCD46-Luc orthotopic lung cancer model.” On day 7 post-inoculation, tumor-bearing C57BL/6 mice (n = 3 per group) were treated with FITC-NHS ester-labeled Ads, PEG-Ads, or the consortium via intratracheal nebulization, using the same small-animal nebulization device (YSKD Biotechnology Co., Ltd., Beijing, China) as in previous experiments. The administered dose for each group was standardized as follows: 1 × 10⁷ PFU of Ads for the free Ads group; 1 × 10⁶ CFU of WH8102 algae for the free algae group; and a combination of 1 × 10⁷ PFU Ads and 1 × 10⁶ CFU algae for the consortium group, maintaining the same component ratios across all comparisons. All doses were calculated on a per-mouse basis and delivered in a total volume of 25 μL via intratracheal nebulization.

After 24 h, mice were euthanized, and whole lungs were harvested. Samples were immersed and stored in 4% paraformaldehyde (PFA) at 4°C until further processing. Paraffin embedding, tissue sectioning (4–6 μm), and fluorescence scanning were performed by a commercial service provider (Servicebio, Wuhan, China). Sections were stained with DAPI to label nuclei, and images were acquired using a slide scanner system. Lung distribution of FITC-labeled formulations was visualized and analyzed using SlideViewer 2.6 software.

To compare permeability across normal and tumor-bearing pulmonary environments, a parallel batch of healthy mice (n = 3) was treated with the same formulations and analyzed under identical conditions. This healthy control served to highlight the enhanced alveolar permeability associated with the pathological tumor microenvironment.

Therapeutic efficacy assessment in an orthotopic lung cancer model

For the therapeutic efficacy assessment, an orthotopic lung tumor model was established via tail vein injection of 1.66×10⁶ TC-1-hCD46-Luc cells per mice. On day 7 post-inoculation (designated as treatment day 0), male C57BL/6 mice (6–8 weeks old, 18–22 g) were randomly assigned into five groups (n = 6 per group): PBS control, free algae (1 × 10⁶ CFU per mouse), free Ads (1 × 10⁷ PFU per mouse), PEG-Ads (1 × 10⁷ PFU per mouse), and the consortium (1 × 10⁷ PFU Ads and 1 × 10⁶ CFU algae per mouse). All formulations were diluted in PBS to a final volume of 25 μL and administered via intratracheal nebulization using a visualized small-animal microspraying device (YSKD Biotechnology).

Tumor progression was monitored via an IVIS Lumina III in vivo imaging system on days 0, 3, 6, and 9 post-treatment. Bioluminescence imaging was performed 10 min after intraperitoneal injection of 250 μL of D-luciferin solution (15 mg/mL). Images were captured once fluorescence signals stabilized and were analyzed using Living Image software. Tumor burden was quantified as the average fluorescence intensity per unit area within defined regions of interest encompassing the thoracic cavity. On day 10, all mice were euthanized. For each group, three mice were randomly selected for downstream immunological analysis of blood and lung samples, while the remaining three mice were utilized for histology, imaging, and safety assessments. To ensure experimental rigor, tumor evaluation and imaging analyses were conducted in a blinded manner.

Analysis of immune cell populations and cytokine levels

Ten days after treatment, three mice per group (n = 3) were anesthetized with 2.5% tribromoethanol, and whole blood was collected via cardiac puncture into EDTA-coated tubes. Following euthanasia, lungs were excised, and tumor-bearing lobes were dissected after removing the main bronchi. Peripheral blood lymphocytes were isolated using a mouse lymphocyte separation kit (Solarbio; Cat. No. P8620). Lung tissues were minced and digested in 2 mL of 1.8 mg/mL collagenase IV (Dingguo, Cat# DH074) at 37°C and 220 rpm for 1 h. Digestion was terminated by adding complete medium with 10% FBS, and the resulting suspensions were filtered through a 70 micrometer nylon mesh. Cells were centrifuged at 500 g for 5 min, and pellets were resuspended in 200 μL of Cell Staining Buffer (BioLegend) for antibody staining.

Single-cell suspensions from blood and lung tissues were stained with fluorochrome-conjugated antibodies (BioLegend) targeting surface markers: CD45, CD3, CD4, CD8, CD25, F4/80, CD11b, CD206, CD80, CD11c, CD86, GR-1, CD62L, and CD44. All antibodies were used at 1 mL per tube in a total volume of 200 μL, and staining was performed at 4°C in the dark for 20 min. For intracellular staining of FOXP3 and CD206, cells were fixed and permeabilized using a Fixation/Permeabilization Solution Kit (BD Biosciences; Cat. No. 554714). Intracellular antibodies were incubated under the same conditions for 20 min. All samples were filtered through a 70 micrometer nylon strainer prior to analysis on a BD FACSCelesta cytometer (BD Biosciences). Data were analyzed using FlowJo software (version 10.8.1), following the gating strategies provided in the Supplementary Information. The representative gating strategies used for flow cytometry analysis are shown in Figures S18–S26.

Cytokine levels of IFN-gamma, IL-6, and TNF-alpha were measured using ELISA kits (Jianglai Bio) according to the manufacturer’s protocols. Serum and lung protein extracts were obtained from three mice per group. In cases where animals showed distress on day 9 (PBS and algae groups), they were humanely euthanized in accordance with institutional ethical guidelines, and blood was collected for analysis. Lung tissues were homogenized in RIPA lysis buffer supplemented with protease inhibitors (Meilunbio), and total protein concentration was quantified using a BCA assay (Dingguo). Absorbance was measured at 450 nm using a Multiskan FC microplate reader. Cytokine concentrations were normalized to sample volume for serum and to total protein content for lung tissues.

Immunofluorescence analysis of a terminal lung cancer model

An orthotopic lung cancer model (n = 3 per group) was established via tail vein injection of 1.66×10⁶ TC-1-hCD46-Luc cells per mice. On day 18 post-inoculation (representing a terminal stage), tumor-bearing C57BL/6 mice received treatments via pulmonary inhalation in the following groups: PBS, free algae (1 × 10⁶ CFU), free Ads (1 × 10⁷ PFU), PEG-Ads (1 × 10⁷ PFU), or the consortium (1 × 10⁷ PFU Ads and 1 × 10⁶ CFU algae).

On day 20, mice were euthanized, and lung tissues were harvested and fixed in 4% paraformaldehyde at 4°C for 24 h. Tissue paraffin embedding, sectioning, and immunofluorescence staining for CD4⁺ and CD8⁺ T cells were performed by Servicebio (Wuhan, China). Images were acquired using a digital slide scanner and analyzed with SlideViewer 2.6 software to evaluate the intratumoral distribution and infiltration density of T cell populations across different treatment groups.

Long-term tumor monitoring

To evaluate the long-term therapeutic potential and survival benefits of the inhalable consortium, an extended orthotopic lung tumor monitoring experiment was conducted. The orthotopic lung cancer model was established in C57BL/6 mice (n = 5 per group) by tail vein injection of 1.66×10⁶ TC-1-hCD46-Luc cells on Day −7. Following a baseline bioluminescence measurement on Day 0, the mice were treated with PBS, Algae (1 × 10⁶ CFU), Ads (1 × 10⁷ PFU), PEG-Ads (1 × 10⁷ PFU), or the Microbe-OVs Consortium (1 × 10⁷ PFU Ads and 1 × 10⁶ CFU algae). Treatments were administered via intratracheal nebulization using a visualized small-animal microspraying device twice, specifically on Day 1 and Day 8 post-baseline.

Tumor progression and therapeutic efficacy were dynamically monitored through in vivo bioluminescence imaging on Days 0, 7, 14, and 21 using the IVIS Lumina III imaging system. Prior to each imaging session, mice were intraperitoneally injected with 250 μL of D-luciferin solution (15 mg/mL). Body weights were recorded throughout the 21-day observation period to assess treatment-associated systemic toxicity and survival benefits. On Day 22, all surviving mice were humanely euthanized. The lungs were subsequently excised and fixed in Bouin’s solution for gross morphological assessment and photography to evaluate the presence of residual pulmonary tumor nodules.

Assessment of background CD8⁺ T cell activation induced by exogenous proteins of TC-1-hCD46-Luc cell

To evaluate whether exogenous proteins used in the TC-1-hCD46-Luc tumor model—namely human CD46, firefly luciferase, and activated human c-Ha-Ras—could induce nonspecific immune activation, a separate validation experiment was conducted. Wild-type TC-1 tumor-bearing mice were established following the same procedure as described in the “Establishment of TC-1-hCD46-Luc orthotopic lung cancer model” section. A total of 12 mice were randomly divided into four groups (n = 3 each) and intraperitoneally injected on days 7 and 9 post-tumor inoculation with either PBS (negative control) or one of the recombinant proteins (10 μg per mouse, dissolved in 100 μL PBS). All proteins were obtained from Shanghai Aladdin Biochemical Technology Co., Ltd.

In parallel, TC-1-hCD46-Luc tumor-bearing mice (n = 3) were injected with PBS and served as a positive control for antigen-specific immune activation. On day 10, all mice were euthanized. Peripheral blood and lung tissues were collected and processed as described in the section “Analysis of immune cell populations and cytokine levels” CD8⁺ T cell populations were analyzed via flow cytometry to assess antigen-specific T cell activation. The representative gating strategies used for flow cytometry analysis are shown in Figures S27 and S28.

Consortium-triggered adoptive T cell transfer assay

To evaluate the functional anti-tumor capacity of CD8⁺ T cells elicited by the consortium, an ACT experiment was performed. Donor mice (n = 5 per group) were established with orthotopic TC-1-hCD46-Luc lung tumors on Day −14 via tail vein injection of 1.66×10⁶ TC-1-hCD46-Luc cells per mice. These donor mice were randomly assigned into four groups and received the following interventions via intratracheal nebulization on Day −7 and Day −3: free PBS (control), free Ads (1 × 10⁷ PFU), free PEG-Ads (1 × 10⁷ PFU), or the Microbe-OVs Consortium (1 × 10⁷ PFU Ads and 1 × 10⁶ CFU algae).

Simultaneously, a second cohort of recipient mice (n = 5 per group) was inoculated with 1.66×10⁶ TC-1-hCD46-Luc cells per mice via tail vein injection on Day −7 to establish pulmonary tumors. On Day 1, the treated donor mice were euthanized, and peripheral blood was collected. CD8⁺ T cells were isolated using a Mouse CD8⁺ T cell Isolation Kit (BEAVER, Suzhou, China) according to the manufacturer’s instructions. The isolated T cells from each donor group were equally distributed and administered to the corresponding recipient mice via tail vein injection. Tumor progression in the recipient mice was monitored via the IVIS Lumina III imaging system on Days 0, 7, and 14.

Evaluation of immune activation in the distal metastatic tumor model

An orthotopic lung tumor model was established via tail vein injection of 1.66×10⁶ TC-1-hCD46-Luc cells into C57BL/6 mice. On day 0 post-inoculation, mice were randomized into five treatment groups (n = 10 per group): PBS control, algae, Ads, PEG-Ads, and the consortium. Drug administration was conducted via intratracheal nebulization on days 0, 2, 4, and 6 using the same device and dosage as described above (1 × 10⁷ PFU Ads in 25 μL per dose). On day 9, all mice were subcutaneously inoculated with 4 × 10⁶ TC-1 cells in the right flank to establish a distal ectopic tumor model.

Five mice per group (n = 5) were designated for endpoint analysis. From day 15 to day 21, ectopic tumor volume and body weight were recorded every other day. Tumor dimensions were measured using digital calipers, and volume was calculated using the formula V = 0.5 × L × W².

On day 21, mice were euthanized, and lung tissues were fixed in Bouin’s solution for 24 h to facilitate visual enumeration of pulmonary nodules. Ectopic tumors were dissected, weighed, and measured to evaluate systemic immune activation.

The remaining five mice per group (n = 5) were assigned to long-term survival monitoring. Survival was tracked continuously without further intervention. Survival data were plotted using GraphPad Prism. Tumor burden and survival data were analyzed independently, and statistical comparisons were performed using unpaired two-tailed t-tests. Immune activation in this model was primarily evaluated through suppression of distal tumor growth and peripheral CD8⁺ T cell responses, as presented in Figure 4. No additional immunological assays were conducted during the survival follow-up phase.

CT imaging-based safety assessment

To evaluate pulmonary safety, healthy C57BL/6 mice (n = 3 per group, without tumor inoculation) were treated via intratracheal nebulization with either 1 × 10⁷ PFU of the consortium (based on Ads), 1 × 10⁶ CFU of algae, or an equivalent volume of PBS (25 μL). On day 4 after administration, mice underwent micro-CT scanning at the Imaging Center of the First Affiliated Hospital of China Medical University. Ferric oxide nanoparticles (Fe₃O₄) were used as a contrast agent to enhance visualization of pulmonary structures. CT image acquisition and analysis were performed by imaging physicians at the Animal Center of China Medical University. Axial, coronal, and sagittal images of lung tissue were obtained. Representative scans from one mouse per group were selected for presentation.

Histopathological analysis of the distal metastatic tumor model via H&E staining

In the distal metastatic tumor model, five mice per group were euthanized on day 21. Heart, spleen, lungs, and kidneys were harvested and fixed in 4% paraformaldehyde at 4°C for 24 h. Tissues from one representative mouse per group were selected for paraffin embedding, sectioning, and hematoxylin and eosin (H&E) staining, which were performed by Wuhan Servicebio Technology Co., Ltd. Fully scanned images of stained sections were acquired and analyzed using SlideViewer 2.6 to evaluate histopathological changes across treatment groups.

Antitumor efficacy in human-derived xenograft model (BALB/c nude mice)

To establish a human-derived lung orthotopic model, male BALB/c nude mice (6–8 weeks old, 18–22 g) were injected via the tail vein with 2 × 10⁶ A549-Luc cells suspended in 100 μL of sterile PBS. On days 0, 3, 6, and 9 post-tumor inoculation, mice received treatment via intratracheal nebulization with free PBS, free algae (1 × 10⁶ CFU), free Ads (1 × 10⁷ PFU), free PEG-Ads (1 × 10⁷ PFU), or the Microbe-OVs consortium (1 × 10⁷ PFU Ads and 1 × 10⁶ CFU algae) in a total volume of 25 μL per dose using a visualized small-animal microspraying device.

Tumor progression was monitored on days 0, 3, 6, and 9 using an IVIS Lumina III imaging system. Bioluminescence imaging was performed 15 min after intraperitoneal injection of D-luciferin solution. On day 10, all mice were euthanized. Lungs were excised and fixed in Bouin’s solution for 24 h to facilitate the visual enumeration and photography of tumor nodules.

Immune activation in human T cell–humanized NOG mice

To evaluate immune activation in a humanized environment, male NOG mice (6–8 weeks old, 18–22 g) were utilized for human immune reconstitution. On day 0, mice were injected via the tail vein with 2 × 10⁶ A549-Luc cells suspended in 100 μL PBS to establish a human-derived lung orthotopic model. On day 1, each mouse received an intravenous infusion of 2 × 10⁶ IL-2–expanded human T cells, which were derived from PBMCs and cultured for 5 days in T cell expansion medium containing 200 IU/mL IL-2.

On day 5, mice were treated via intratracheal nebulization with PBS, algae (1 × 10⁶ CFU), Ads (1 × 10⁷ PFU), PEG-Ads (1 × 10⁷ PFU), or the consortium (1 × 10⁷ PFU Ads and 1 × 10⁶ CFU algae) using a small-animal microsprayer (YSKD Biotechnology, Beijing, China). On day 7, mice were euthanized, and blood and lung tissues were collected for flow cytometry analysis (n = 3 per group).

Single-cell suspensions were prepared by mechanical dissociation and enzymatic digestion as described for the C57BL/6 model. Cells were stained with the following human-specific antibodies (all purchased from Starter Biotech Co., Ltd., China): FITC Mouse Anti-Human CD45 Antibody, Alexa Fluor 647 Mouse Anti-Human CD3 Antibody and PE Mouse Anti-Human CD8α Antibody. Samples were analyzed on a BD FACSCelesta flow cytometer (BD Biosciences, USA), and data were processed using FlowJo v10.8.1. The percentages of human CD8⁺ T cells were calculated within the CD45⁺ CD3⁺ population. Other consumables and staining procedures were consistent with those used in the C57BL/6 immune-profiling experiments. The representative gating strategies used for flow cytometry analysis are shown in Figures S29 and S30.

Quantification and statistical analysis

Data are presented as means ± SD. For in vitro experiments, n represents the number of independent biological replicates. For in vivo studies, n indicates the number of animals per group. The student’s t test (two-tailed) and one-way analysis of variance were used to analyze the comparison between groups. Statistical significance was considered when: ns: no significance, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. All data were analyzed by GraphPad Prism 10.1 and the excel software version 2021.

Ethical approval for animal experiments

All animal experiments were conducted in strict accordance with the ARRIVE guidelines and the Guide for the Care and Use of Laboratory Animals. The ethical oversight was partitioned based on the specific experimental models and institutional facilities used:

Murine Models at Shenyang Pharmaceutical University: Procedures involving C57BL/6 mice for pulmonary distribution, orthotopic tumor efficacy, and the distal metastatic tumor model were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Shenyang Pharmaceutical University (approval number: SYPU-IACUC-S2024-0621-102).

Specialized Murine Assays at China Medical University: The long-term monitoring and adoptive T cell transfer (ACT) assays were conducted under protocols approved by the Ethics Committee of China Medical University (approval number: CMU20253292).

Humanized and Xenograft Models: All animal studies involving BALB/c nude mice bearing A549-Luc tumors and human PBMC-reconstituted NOG mice were approved by the Ethics Committee of the First Affiliated Hospital of China Medical University (approval number: CMU2021722).

Regarding human-derived materials, human A549-Luc cells and PBMCs were purchased from Wuhan Sunncell Biotechnology. These samples were fully anonymized and certified for research use only by the provider. As they cannot be traced back to individual donors, they were exempt from additional institutional ethical approval for donor identification.

Published: April 24, 2026