Combinatorial leaky probiotic for anticancer immunopotentiation and tumor eradication

Abstract

Combination therapies present a compelling therapeutic regimen against the immunosuppressive and heterogeneous microenvironment of solid tumors. However, incorporating separate therapeutic modalities in regimen designs can be encumbered by complex logistical, manufacturing, and pharmacokinetic considerations. Herein, we demonstrate a single-vector combinational anticancer therapy using an lpp gene knockout leaky probiotic for simultaneous secretion of immunotherapeutic and oncolytic effector molecules. Through fusion protein design and vector optimization, a Nissle1917 (EcN) bacteria vector is engineered to secrete Neoleukin-2/15 (Neo-2/15) cytokine-functionalized anti-PDL1 nanobody (aPDL1-Neo2/15) and anti-mesothelin-functionalized hemolysin E (HlyE-aMSLN). The multifunctional leaky probiotic enables synchronous immune activation and tumor-targeted cytolytic activity for effective tumor suppression, elevation of tumor immune cell infiltration, and establishment of anticancer immunological memory. lpp gene knockout is further shown to improve probiotic tolerability and intravenous applicability, offering a therapeutically viable approach for combination regimen development.

Keywords: anticancer immunity; bacterial therapy; chimeric protein; combinational cancer therapy; immune checkpoint inhibitors; leaky probiotic; oncolytic toxins.

Graphical abstract

Figure 1

Schematic of combinatorial immunotherapeutic-oncolytic probiotic for single-vector combinational anticancer treatment A loss-of-function strategy for spontaneous protein release from bacterial vectors is achieved through lpp gene knockout, giving rise to a leaky probiotic. Therapeutic components are then incorporated into the leaky probiotic through the insertion of a plasmid with a single promoter.

Figure 2

Characterization and optimization of immunotherapeutic-expressing leaky bacteria (A) Schematic illustrating immunomodulator-secreting bacteria prepared from EcB and EcB_Δlpp. (B) Quantification of secreted IL-2 variants (IL-2, Sk2, and Neo2/15) from EcB and EcB_Δlpp by western blot with supernatants from overnight (16 h) bacterial culture. Bacterial concentration following overnight (16 h) culture was measured through OD600. Wt denotes EcB_{wild type}; KO denotes EcB_Δlpp. Expression levels of IL-2 variants were detected using anti-His-HRP (BioLegend, #362613). (C) Viability of IL-2-dependent HT2 cells under treatment with bacterial supernatants derived from different bacterial vectors. IL-2_Positive denotes complete medium with commercial IL-2 (100 U/mL). IL-2_Negative denotes complete medium only. Value of representative OD600. Vector, 11.0; IL-2, 5.68; Sk2, 6.48; Neo2/15, 7.84. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA (∗∗∗∗p < 0.0001). Tumor progression and survival curves of MC38 tumor-bearing mice after intratumoral bacteria treatment when tumors reach (D and E) ∼100 mm³ and (F and G) ∼400 mm³. Inverted triangles indicate the date of intratumoral injection. Error bars represent mean ± SEM. Statistical analyses were performed by (D) unpaired t test; (F) paired t test. (E and G) Log rank (Mantel-Cox) test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001). (H) Examination of tumor sections following intratumoral treatment with PBS or cytokine-secreting EcB_Δlpp. FITC signal denotes CD8 TILs; AF647 signal denotes E. coli; and Hoechst signal denotes nucleus of cancer cells. Scale bar denotes 50 μm. (I) Quantification of secreted immunomodulator (aPDL1-Neo2/15, aCTLA4-Neo2/15, and aVSIG4-Neo2/15) from EcB_Δlpp. Bacterial concentration following overnight (16 h) culture was measured through OD600. Expression levels of anti-PD-L1, anti-CTLA4, or anti-VSIG4 nanobody-based Neo-2/15 were detected using anti-His-HRP (BioLegend, #362613). (J and K) Tumor progression and survival curves of LLC tumor-bearing mice after intratumoral bacteria treatment. Error bars represent mean ± SEM. Statistical analyses were performed by paired t test (∗p < 0.05). (C) The experiment is technical replicates, while (D–G, J, and K) these experiments are biological replicates.

Figure 3

Enable the utilization of leaky bacteria to establish a combining immunotherapeutic-oncolytic system (A) Schematic illustrating combinatorial EcB_Δlpp preparation. P represents promoter. SD represents Shine-Dalgarno sequence. (B) Viability of HEK293T_MSLN upon treatment with supernatants from HlyE- or oncolytic toxin (O, HlyE-aMSLN)-expressing bacteria. Error bars represent mean ± SD. Statistical analyses were performed by paired t test (∗∗p < 0.01). Quantification of secreted fusion protein was measured through ImageJ. (C and D) Quantification of secreted (C) oncolytic toxin (HlyE-aMSLN) and (D) immunomodulator (aPDL1-Neo2/15) from EcB_Δlpp by western blot with supernatants from overnight (16 h) bacterial culture. Bacterial concentration following overnight (16 h) culture was measured through OD600. Quantification of secreted fusion protein was measured through ImageJ. Expression levels of oncolytic toxins were detected using anti-Myc-HRP (Merck Millipore, #16–213), while expression levels of immunomodulators were detected using anti-His-HRP (BioLegend, #362613). (E and F) Viability of (E) HEK293T_{wild type} or (F) HEK293T_MSLN under treatment with supernatant containing immunomodulator with HlyE or oncolytic toxin. Error bars represent mean ± SD. Statistical analyses were performed by paired t test (∗∗p < 0.01). (G and H) Viability of (G) MC38 or (H) CT26_MSLN under treatment with bacterial supernatant containing immunomodulator and/or oncolytic toxin. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA (∗∗∗∗p < 0.0001). (I) Viability of IL-2-dependent HT2 cells under treatment with bacterial supernatant containing immunomodulator with/without oncolytic toxin. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA (∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). (B, E–I) These experiments are technical replicates.

Figure 4

Safety and functional characterization of combinatorial leaky probiotics (A) Schematic illustrating safety consideration and assessment for combinatorial immunotherapeutic-oncolytic probiotic (CIOP) development. (B–E) Weight of mice following treatment with PBS, EcN_{wild type}, or EcN_Δlpp. Each mouse received one intravenous (i.v.) injection with 5 × 10⁶ CFU probiotic. Error bars represent mean ± SEM. Statistical analyses were performed by two-way ANOVA (∗∗∗∗p < 0.0001). (C) CFU titers detected from the organs and resected tumors of mice bearing MC38 tumors 2 h following i.v. administration with 5 × 10⁶ CFU of either EcN_{wild type} or EcN_Δlpp. Error bars represent mean ± SD. Statistical analyses were performed by two-way ANOVA (∗∗∗p < 0.001). (D) EcN_{wild type} or EcN_Δlpp CFU upon 1 h culturing in the blood of mice or in PBS. Error bars represent mean ± SD. Statistical analyses were performed by two-way ANOVA (∗p < 0.05). (E) Survival curve of mice following i.v. injection with 5 × 10⁸ CFU EcN_{wild type} or EcN_Δlpp. Statistical analyses were performed by log rank (Mantel-Cox) test. (F) EcN_Δlpp CFU titers examined from tumor-bearing mouse organs 3 days following i.v. injection with 5 × 10⁸ CFU EcN_Δlpp. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA. (G) Survival curve of mice following i.v. injection with 5 × 10⁸ CFU of EcN_Δlpp or EcB_Δlpp expressing either HlyE-aMSLN or non-targeted HlyE. Statistical analyses were performed by log rank (Mantel-Cox) test. (H and I) Quantification of secreted (H) immunomodulator (aPDL1-Neo2/15) and (I) oncolytic toxin (HlyE-aMSLN) from EcN_Δlpp by western blot with supernatants from overnight (16 h) probiotic culture. Probiotic concentration following overnight (16 h) culture was measured through OD600. Quantification of secreted fusion protein was measured through ImageJ. Expression levels of immunomodulators were detected using anti-His-HRP (BioLegend, #362613), while expression levels of oncolytic toxins were detected using anti-Myc-HRP (Merck Millipore, #16–213). (J) Survival curve after 4 i.v. administrations with 5 × 10⁸ CFU of EcN_Δlpp and CIOP. A control group with combinatorial EcB_Δlpp co-expressing oncolytic and immunomodulatory chimeric proteins (CIOB) injections showed 100% mortality. Inverted triangles show the intravenous dosing regimen. Statistical analyses were performed by log rank (Mantel-Cox) test. (K) Viability of IL-2-dependent HT2 cells treated with supernatants from probiotic cultures. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA (∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). (L–N) Viability of (L) MC38, (M) B16F10, or (N) LLC cancer cells under treatment with supernatant from probiotic culture. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA (∗∗p < 0.01, ∗∗∗∗p < 0.0001). Experiments (D) and (K–N) are technical replicates, while experiments (B), (C), (E–G), and (J) are biological replicates.

Figure 5

Anticancer treatment with probiotics expressing immunotherapeutic, oncolytic, or combinatorial anticancer functionalities (A) Schematic illustrating assessment of intratumoral leaky probiotic therapy against established MC38 tumors. (B) Tumor progression and (C) survival curves of MC38 tumor-bearing mice after intratumoral leaky probiotic treatment. Error bars represent mean ± SEM. Statistical analyses were performed by (B) two-way ANOVA and (C) log rank (Mantel-Cox) test (∗∗p < 0.01, ∗∗∗∗p < 0.0001). (D) Schematic illustrating comparison of oncolytic probiotic (OP) and CIOP treatment against large established MC38 tumors and distal tumor growth. (E) Primary and (F) distal tumor progression curves of MC38 tumor-bearing mice after intratumoral leaky probiotic treatment when treated primary tumors reached ∼300 mm³. Error bars represent mean ± SEM. (E and F) Statistical analyses were performed by two-way ANOVA (∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). (G) Schematic illustrating assessment of intratumoral leaky probiotic treatments against mesothelin-expressing CT26 tumors. (H–J) Tumor progression and (I) survival curves of CT26_MSLN tumor-bearing mice after intratumoral leaky probiotic treatment. Tumor-bearing mice were considered deceased when the tumor size exceeded 1,200 mm³. Error bars represent mean ± SEM. Statistical analyses were performed by (H) paired t test and (I) log rank (Mantel-Cox) test (∗p < 0.05). (J) Schematic illustrating probiotic-based B16F10 tumor therapy and rechallenge test. For the rechallenge test, each naive or recovered mouse received 2.5 × 10⁵ MC38 subcutaneous inoculation. (K) Tumor progression and (L) survival curves of CT26_MSLN tumor-bearing mice after intratumoral leaky probiotic treatment. Error bars represent mean ± SEM. Statistical analyses were performed by (K) paired t test and (L) log rank (Mantel-Cox) test (∗p < 0.05, ∗∗p < 0.01). I, immunomodulator (aPDL1-Neo2/15); O, oncolytic toxin (HlyE-aMSLN). (M) Tumor progression curves of naive and recovered mice. All experiments are biological replicates.

Figure 6

Intravenous treatment with CIOP suppresses tumor growth and increases immune cell infiltration (A) Schematic illustrating probiotic-based MC38 tumor therapy. (B) Tumor progression and (C) survival curves of MC38 tumor-bearing mice after intravenous probiotic treatment. Inverted triangles indicate the date of intravenous injection. Error bars represent mean ± SEM. Statistical analyses were performed by (B) paired t test and (C) log rank (Mantel-Cox) test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). I, immunomodulator (aPDL1-Neo2/15); O, oncolytic toxin (HlyE-aMSLN). (D–J) MC38 tumor-bearing mice after four times intravenous injection. (D) Tumor weight. (E) Percentage of live CD45-positive TILs in MC38 solid tumors. (F, G) Percentage of live CD45⁻ and (F) CD8⁻ or (G) CD4-double-positive cytotoxic TILs. (H) Percentage of live CD45, MHCII, and CD19 triple-positive B TILs. (I) Percentage of live CD45, MHCII, and CD11c triple-positive dendritic cells in MC38 tumors. (J) Percentage of live CD45, MHCII, and F4/80 triple-positive macrophages in MC38 tumors. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA (∗p < 0.05, ∗∗p < 0.01). (K) Schematic illustrating probiotic-based LLC tumor therapy. (L) Tumor progression and (M) survival curves of LLC tumor-bearing mice after intravenous probiotic treatment. Error bars represent mean ± SEM. Statistical analyses were performed by (L) paired t test and (M) log rank (Mantel-Cox) test (∗∗p < 0.01). (N) Percentage of live CD45-positive TILs in LLC solid tumors. Error bars represent mean ± SD. Statistical analyses were performed by ordinary one-way ANOVA (∗p < 0.05, ∗∗p < 0.01). All experiments are biological replicates.