Selective delipidation of Mycobacterium bovis BCG retains antitumor efficacy against non-muscle invasive bladder cancer

Abstract

Purpose: Repeated instillations of bacillus Calmette et Guérin (BCG) are the gold standard immunotherapeutic treatment for reducing recurrence for patients with high-grade papillary non-muscle invasive bladder cancer (NMIBC) and for eradicating bladder carcinoma-in situ. Unfortunately, some patients are unable to tolerate BCG due to treatment-associated toxicity and bladder removal is sometimes performed for BCG-intolerance. Prior studies suggest that selectively delipidated BCG (dBCG) improves tolerability of intrapulmonary delivery reducing tissue damage and increasing efficacy in preventing Mycobacterium tuberculosis infection in mice. To address the lack of treatment options for NMIBC with BCG-intolerance, we examined if selective delipidation would compromise BCG's antitumor efficacy and at the same time increase tolerability to the treatment.

Materials and methods: Murine syngeneic MB49 bladder cancer models and in vitro human innate effector cell cytotoxicity assays were used to evaluate efficacy and immune impact of selective delipidation in Tokyo and TICE BCG strains.

Results: Both dBCG-Tokyo and dBCG-TICE effectively treated subcutaneous MB49 tumors in mice and enhanced tumor-infiltrating CD8⁺ T and natural killer cells, similar to conventional BCG. However, when compared to conventional BCG, only dBCG-Tokyo retained a significant effect on intratumoral tumor-specific CD8⁺ and γδ T cells by increasing their frequencies in tumor tissue and their production of antitumoral function-related cytokines, i.e., IFN-γ and granzyme B. Further, dBCG-Tokyo but not dBCG-TICE enhanced the function and cytotoxicity of innate effector cells against human bladder cancer T24 in vitro.

Conclusions: These data support clinical investigation of dBCG-Tokyo as a treatment for patients with BCG-intolerant NMIBC.

Keywords: BCG; Bladder cancer; Delipidation; Treatment.

Fig. 1

Selective removal of lipids in BCG does not compromise the treatment efficacy in the mouse bladder tumor model. a As indicated in the scheme, WT C57BL/6 mice were subcutaneously challenged with MB49 tumor and treated with either b BCG-Tokyo and dBCG-Tokyo or c BCG-TICE and dBCG-TICE intratumorally weekly for 3 weeks starting on day 5. Shown are b pooled data from 3 independent experiments (n = number of tumors) or c pooled data from 2 independent experiments (n = number of tumors). b, c Two-way ANOVA. Mean ± SEM. d, e WT female C57BL/6 mice were orthotopically challenged with MB49 tumor and treated with either BCG-Tokyo or dBCG-Tokyo intravesically through bladder instillation via urethral catheter, weekly for 5 weeks starting on day 1. IVES, intravesical (bladder administration). Shown are (d) study scheme and (e) bladder weight data pooled from 3 independent orthotopic survival experiments (each dot represents one bladder). Mann–Whitney test were applied after normality test. Mean ± SEM

Fig. 2

Selective removal of lipids in BCG-Tokyo can reduce inflammatory toxicity without affecting anti-tumor cytokines. a–c Urine collected at various time points (baseline or post-BCG) of orthotopic study were measured for mouse cytokines by Luminex assay (pooled from 2 independent orthotopic experiments; each dot represents one urine sample). Shown are ratios of a IL-6 to IL-10, b IL-1β to IL-10 by d15 and c TNF-α to IL-10 by d1. Student t-test or Mann–Whitney test was applied after normality test. Mean ± SEM. d–k At the end of s.c. MB49 tumor study, d–g 24-h supernatant from tumor tissue or h–k d22 serum were collected for cytokine profiling by Luminex assay. d–g one representative experiment (each symbol represents one tumor); h–i pooled data from two independent experiments (each symbol represents one mouse); j–k one representative experiment (each symbol represents one mouse). d–k Student t-test or Mann–Whitney test was applied accordingly after normality test. Mean ± SEM

Fig. 3

Delipidation does not compromise the effect of BCG-Tokyo in promoting tumor-infiltrating CD8⁺ T cells and γδ T cells. s.c. MB49-challenged mice, as described in Fig. 1 legend, were sacrificed on day 22. Tumors were collected and processed for analysis of immune infiltrates by flow cytometry. Shown are: a gating strategy and example for tumor-infiltrating CD8⁺ T cells (gated under live CD45⁺CD3⁺ cells) and subsequent markers; b–e pooled data from 3 independent experiments of dBCG-Tokyo study (left panels), or one representative experiment of dBCG-TICE study (right panels) for AN of cells per 0.1 g of tumor; f gating strategy and example for tumor-infiltrating γδ T cells (gated under live CD45⁺CD3⁻ cells) and subsequent markers; g–j pooled data from 3 independent experiments of dBCG-Tokyo study for AN of cells per 0.1 g of tumor. One symbol represents one tumor. Student t-test or Mann–Whitney test were applied accordingly after normality test. Mean ± SEM

Fig. 4

Delipidated Tokyo strain retains BCG’s effect on increasing cytotoxicity and function of expanded human innate effector cells. a As indicated in the diagram, PBMC-derived innate effector cells from 3 different bladder cancer patients were expanded for 14 days with or without BCG-Tokyo or dBCG-Tokyo treatment. b On day 14, cytotoxicity assays were performed in triplicates against T24 cells at effector to target ratio of 25:1 (pooled results from 3 patients); meanwhile, innate effector cells grew from replicated wells were also analyzed for c CD3⁺γ9δ2⁺ T cells and d CD3⁻CD56⁺ NK cells by flow cytometry (representative of 3 patients; AN of cells grew from 2 × 10⁶ PBMCs on day 0). Student t-test was applied accordantly after normality test. Mean ± SEM

Fig. 5

Strain-specific profile of apolar lipids from PE-extracts of BCG-Tokyo and BCG-TICE. a Brief scheme (created using Biorender.com) showing apolar lipids being removed from the surface of BCG by sequential 100% petroleum ether (PE) extractions without compromising the cell envelope integrity leaving the dBCG bacteria viable. b Apolar lipids PDIMs and TAGs were analyzed by thin layer chromatography (TLC) using a solvent system of PE/acetone (94:4, v/v). Left: Lipids obtained from each sequential PE-extraction (numbers indicate extraction times). Right: Comparison between PE-extracted lipids (PE fraction, pooled from all 3 PE extractions/PE-ext.), and lipids remaining on the cell envelope of dBCG strains (C/M fraction/C:M-ext.). c Apolar lipids MycB, PGLs and TDM were analyzed by TLC using a solvent system of C/M (95:5, v/v). Left: Lipids obtained from each sequential PE-extraction (Numbers indicate extraction times). Right: Comparison between PE-extracted lipids (PE fraction, pooled from all 3 PE extractions/PE-ext.), and the lipids remaining on the cell envelope of dBCG strains (C/M fraction/C:M-ext.) d Densitometry analysis and Log₂-fold changes of the PE-extracted lipids of BCG-Tokyo compared to those of BCG-TICE. Abbreviations: ext., extracts of; a.u., arbitrary units; PE, petroleum ether; C:M, chloroform/methanol (2:1, v/v, lipid extracts); MycB, mycoside B; PDIMs, phthiocerol dimycocerosates; PGL, phenolic glycolipid; TAGs, triacylglycerols; TDM, trehalose dimycolate