https://orcid.org/0009-0003-1792-1180
Figures
Abstract
In recent years, immunotherapy of patients with higher-risk non-muscle invasive bladder cancer (NMIBC) in North America has relied on the use of the TICE strain of BCG. However, limitations in the supply chain have warranted investigation of the therapeutic benefit of other strains of BCG, such as BCG-Russia. Trained immunity, a form of innate immune memory, is now widely believed to be an important component of the therapeutic benefit of BCG. Therefore, in the present study we compared the effects of BCG-TICE and BCG-Russia on the acquisition of trained immunity and related secondary immune responses. C57BL/6 mice received a single intravenous injection of BCG-Russia or BCG-TICE. Four weeks later, bone marrow was collected for flow cytometric analysis of hematopoietic stem and progenitor cell (HSPC) populations, generation of bone marrow-derived macrophages, functional assessment of trained immunity, and transcriptomic profiling. Compared with BCG-Russia, BCG-TICE elicited increased levels of trained immunity, characterized by higher production of several proinflammatory cytokines upon secondary activation. BCG promoted the expansion of HSPCs independent of strain. BCG-TICE was linked to upregulation of key inflammation-related genes and enrichment of functionally relevant pathways. The results of this study reveal strain-dependent differences in the ability of BCG to induce innate immune memory and inflammatory pathways that could ultimately determine efficacy of immunotherapy of patients with NMIBC.
Citation: Nauman RW, Greer PA, Craig AW, Cotechini T, Siemens DR, Graham CH (2026) Differential effects of BCG-Russia and BCG-TICE on trained immunity: Potential implications for bladder cancer immunotherapy. PLoS One 21(9): e0357950. https://doi.org/10.1371/journal.pone.0357950
Editor: Odir Antonio Dellagostin, UFPL, BRAZIL
Received: May 22, 2026; Accepted: August 25, 2026; Published: September 11, 2026
Copyright: © 2026 Nauman et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data relevant to this study are included in the manuscript.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and publication of this article: This work was supported by: The Canadian Institutes of Health Research awarded to CHG, DRS and TC [Project Grant #173383 URL: https://cihr-irsc.gc.ca] Terry Fox Research Institute awarded to CHG, PAG, AWC and DRS [Program Project Grant #1138-05 URL: https://www.tfri.ca]. Funders did not play any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Since its isolation in 1921, Bacillus Calmette-Guérin (BCG) has been disseminated globally, leading to the emergence of distinct strains [1]. The TICE strain, preferred for intravesical therapy of patients with higher-risk non-muscle invasive bladder cancer (NMIBC) [2], is currently produced at a single facility in the United States. Prior to 2017, the Connaught strain (Sanofi Pasteur) was available as an alternative to BCG-TICE (OncoTICE®), but production has since been discontinued. The BCG-Russia strain produced by Verity Pharmaceuticals is approved as an alternative in Canada, but not in the United States. Ongoing shortages underscore the need for direct comparisons of the efficacy of different BCG strains, and several head-to-head clinical trials have explored this [3–5]. While major differences in clinical outcomes were not reported between most strains, BCG-TICE was shown to extend recurrence-free survival compared to BCG-Connaught [3,6], provided patients received maintenance therapy. This benefit did not extend to progression-free survival [3]. With BCG-Connaught no longer available to patients, the finding of decreased efficacy compared to BCG-TICE has limited relevance. The BCG-Russia strain has replaced Connaught as an approved substitute in some jurisdictions when BCG-TICE is unavailable, and there is some evidence that treatment with BCG-Russia provides a therapeutic benefit comparable with other strains [7]. However, it is crucial to conduct multi-arm head-to-head trials between BCG-Russia and BCG-TICE to more reliably determine whether efficacy differs between the two strains. The ongoing EVER trial (NCT05037279) aims to address this question.
While BCG is the standard immunotherapy for patients with higher-risk NMIBC, up to 50% of these patients suffer from early disease recurrence [8–10], with some progressing to the more lethal muscle-invasive disease [9,10]. Thus, it is important to improve our understanding of the mechanism of action of BCG so that better therapeutic approaches can be developed. Only recently has it been recognized that BCG induces long-term reprogramming of innate immune cells, a process known as trained immunity. This de facto state of immunological memory develops through epigenetic and metabolic reprogramming following recognition of mycobacterial pathogen-associated molecular patterns (PAMPs) by pattern recognition receptors (PRRs) [11]. Trained immunity is thought to contribute to the immunotherapeutic benefit of BCG [12] and studies have revealed that trained immunity in peripheral blood monocytes is associated with prolonged recurrence-free survival [13,14]. Reprogramming of hematopoietic stem cells and myeloid precursors is a key feature of BCG-induced trained immunity [15], and these systemic effects of locally administered BCG may be important for achieving maximal therapeutic efficacy [16]. Recent reports highlight that induction of trained immunity by different strains of BCG results in phenotypic variations [17]. Thus, strain-dependent effects on the development of trained immunity may have important implications for the management of NMIBC. However, the BCG strains commonly used in NMIBC immunotherapy have not been compared in terms of their ability to induce trained immunity and the trained phenotype elicited. Thus, in the present study we used a mouse model to assess differential induction of trained immunity and secondary responses between BCG-TICE and BCG-Russia. Our results indicate that functional reprogramming of macrophages differs between the two strains, with stronger secondary immune responses associated with BCG-TICE
Materials and methods
BCG
BCG-Russia was purchased from Verity Pharmaceuticals (Mississauga, ON, CA) as formulated for intravesical instillation (freeze-dried powder, 40 mg net weight per vial, 1–8 x 108 Colony Forming Units (CFU) per vial). BCG-TICE (Merck; Rahway, NJ, U.S.) was obtained from Investigational Drug Services at Kingston Health Sciences Centre (freeze-dried powder, 50 mg net weight per vial, 1–8 x 108 CFU per vial). BCG was aliquoted under aseptic conditions, with minimal exposure to light, and stored at 4°C. Both BCG strains were supplied in vials containing equivalent quantities as measured in CFUs, but differed in the net weight of freeze-dried powder. To control for this and standardize the dose to an average of 5.625 x 106 CFU (total amount injected), BCG was resuspended at a concentration of 10 mg/ml (Russian strain) or 12.5 mg/ml (TICE strain) in sterile PBS immediately before use.
Mouse models
C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME, U.S.) and housed in a pathogen-free facility with ad libitum access to food and water. All animal experiments were approved by the Queen’s University Animal Care Committee (Protocol Number: 2023-2457) in accordance with the Canadian Council on Animal Care (CCAC) guidelines. Researchers underwent technical training for all procedures performed as part of this study. Thirty-six mice were used across all studies (18 males and 18 females). Cumulatively, 12 mice served as controls, 13 mice received BCG-Russia, and 11 mice received BCG-TICE. Specific information concerning sample sizes, sex distribution, and treatment group allocation are specified in each figure caption.
To induce trained immunity, six- to eight-week-old female or male mice were intravenously injected with 5.625 x 106 CFU in 50 μl phosphate-buffered saline (PBS; ThermoFisher; Waltham, MA, U.S.). After four weeks, mice were euthanized via barbiturate overdose and tissues were collected aseptically. Euthanasia was performed by intraperitoneal (i.p.) injection of sodium pentobarbital (100 mg/kg; 10 mg/ml), followed by cardiac exsanguination or cervical dislocation.
For in vivo rechallenge, mice were injected with 10 μg LPS (E. coli O26:B6; MilliporeSigma; Burlington, MA, U.S.) i.p. four weeks after BCG administration. Four hours after the LPS injection, whole blood was collected via cardiac puncture in the presence of EDTA to prevent coagulation. Plasma was recovered from the supernatant of centrifuged whole blood and was stored at −80°C. Cytokine quantification was performed by Eve Technologies (Calgary, AB, CA) using the Mouse Focused 10-Plex Discovery Assay® (MDF10). Cytokines in this panel included GM-CSF, IFN-γ, IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12p70, CCL2, and TNF-α. Analytes with concentrations that fell outside the detectable range using the Eve Technologies assay were alternatively measured by ELISA (Invitrogen/ThermoFisher), or their raw fluorescence intensity (FI) values were reported to assess relative levels. Raw values of plasma cytokine concentrations are shown in S1 Table.
Animal welfare was monitored through daily visual checks, as well as weekly body weight monitoring and assessment of body condition score (BCS). Humane endpoints aligned with standard criteria, including >15% weight loss and BCS < 2. No animals died prematurely or were euthanized as a result of meeting humane endpoint criteria prior to the predefined study endpoint. All efforts were made to minimize suffering and distress.
Ex vivo stimulation assay
Bone marrow was harvested from femurs and tibiae in RPMI-1640 (ThermoFisher) supplemented with 2.5% heat-inactivated FBS (MilliporeSigma). To generate bone marrow-derived macrophages (BMDMs), 2×106 bone marrow cells were plated in 10-cm dishes with 10 mL of BMDM medium (2 mM L-glutamine, 10% FBS, 20 mM HEPES, 1% non-essential amino acids, 1% essential amino acids, 100 U/mL penicillin, 100 μg/mL streptomycin, 1 mM sodium pyruvate, 0.14% 5N NaOH (all ThermoFisher), in RPMI-1640 medium) supplemented with 20 ng/mL of recombinant murine macrophage colony stimulating factor (M-CSF; PeproTech®, ThermoFisher) (BMDM/M-CSF medium). Cells were incubated at 37°C and 5% CO2 for six days, with an additional 5 mL of BMDM/M-CSF medium added on days two and four. After macrophage differentiation was complete, each dish was washed of residual medium and non-adherent cells. To facilitate detachment from dishes, BMDMs were incubated with 5 mL of Cellstripper (Corning; Corning, NY, U.S.) for 20 minutes. Cells were resuspended in fresh BMDM/M-CSF medium and replated for stimulation assays.
BMDMs seeded in 96-well plates (105 cells/well) were allowed to adhere overnight (37°C, 5% CO2) prior to stimulation. After 15–20 hours, the old medium was discarded and fresh medium (containing 10 ng/ml M-CSF) and 100 ng/ml LPS (O26:B6) or 1 μg/ml Pam3CysSerLys4 (Pam3Cys; EMC Microcollections; Tübingen, BW, Germany) was added. After 24 hours, cell-free supernatants were collected and stored at −80°C and cells were lysed for RNA isolation. Cytokine quantification was performed as previously described (MDF10 or ELISA).
Flow cytometry
Bone marrow cells (2×106) in suspension were subjected to red blood cell lysis (Pharm Lyse™; BD Biosciences; San Diego, CA, U.S.), then incubated (4°C; 30 min.) with anti-CD16/CD32 (1:250; BD Biosciences) and fixable Aqua viability stain (1:500; Invitrogen) in PBS. Subsequently, cells were washed with PBS supplemented with 2 mM EDTA and 0.75% BSA and stained for lineage markers (4°C; 30 min.). Antibodies in the lineage marker cocktail included: anti-Ter-119 (BioLegend; San Diego, CA, U.S.), anti-CD11b (clone M1/70; BioLegend), anti-IgM (clone RMM-1; BioLegend), anti-CD3 (clone 17A2; BioLegend), anti-CD45R (clone RA3-6B2; BioLegend), and anti-Ly6G/C (Gr-1; clone RB6-8C5; BD Biosciences), all FITC-conjugated and at a 1:400 dilution (0.125 μg). Cells were then washed with PBS/EDTA/BSA and stained for HSPC markers (4°C; 30 min.) with 10% Brilliant Stain Buffer (BD Biosciences). The following fluorophore-conjugated antibodies were used for HSPC staining: anti-c-Kit–BV605 (1:400; clone 2B8; BD Biosciences), anti-Sca-1–PE-Cy7 (1:400; clone D7; BD Biosciences), anti-CD150–BV421 (1:200; clone Q38-480; BD Biosciences), anti-CD48–AF700 (1:200; clone HM48-1; BioLegend), anti-Flt3–PE (1:200; clone A2F10.1; BD Biosciences), and anti-CD34–AF647 (1:200; clone RAM34; BD Biosciences). Cells were subsequently washed in PBS/EDTA/BSA and resuspended in Fixation Buffer (BD Biosciences) containing 4.21% formaldehyde for 15 minutes (4°C). After a final wash, cells were resuspended in PBS/EDTA/BSA and stored at 4°C until they were analysed using a CytoFLEX S flow cytometer (Beckman Coulter; Brea, CA, U.S.). Data were analysed using FlowJo software (v10.10; BD).
Transcriptomic analysis
RNA was extracted from BMDMs using the PureLink Micro-to-Midi Total RNA Purification kit (Invitrogen). RNA integrity (all samples RIN = 10) was assessed with the Agilent 2100 Bioanalyzer (Agilent Technologies; Santa Clara, CA, U.S.) and quantified using a Nanodrop 2000 spectrophotometer (ThermoFisher). The nCounter assay was performed at the Ontario Institute for Cancer Research (OICR; Toronto, ON, CA) according to the manufacturer’s instructions, using the NanoString Myeloid Innate Immunity V2 Code Set (Bruker Spatial Biology; Bothell, WA, U.S.) and 50 ng RNA input. Data were analyzed with nSolver software (v4.0; Bruker Spatial Biology) and nCounter Advanced Analysis plugin (v2.0). Default parameters were used for normalization, with automated selection of housekeeping genes to minimize pairwise variation according to the geNorm algorithm. Adjusted p-values were generated using the Benjamini–Hochberg method of false discovery rate estimation. Figures were generated with R (v4.5.1).
Statistical analysis
Final data analysis was performed using GraphPad Prism (v10.4). One-way ANOVA followed by Tukey’s multiple comparisons test was used to determine statistical differences between three groups. Two-way ANOVA with Šídák’s post-hoc analysis was performed on ex vivo rechallenge data. The ROUT method (Q = 1%) was used where specified in figure captions to identify outliers, which were excluded from analysis. Data are presented as mean ± SEM, with p < 0.05 considered statistically significant.
Results
BCG-TICE and BCG-Russia induce expansion of hematopoietic stem cells
A known feature of BCG-induced trained immunity is induction of myelopoiesis, with expansion of HSPCs (LKS+ cells) peaking four weeks after BCG administration [15]. Four weeks after an intravenous injection of BCG, both strains induced a marked expansion of LKS+ cells compared with controls, with 13.2-fold and 11.2-fold mean increases following BCG-Russia and BCG-TICE vaccination, respectively (Figs 1A, B and S1). Bone marrow proportions of LKS+ cells showed a similar pattern, increasing significantly following administration of either strain (Fig 1C). No significant differences in the induction of myelopoiesis were detected between the two BCG strains.
(A) Representative fluorescence-activated cell sorting (FACS) plots of LKS+ (Lin–c-Kit+Sca-1+) frequencies four weeks after vaccination. Values on plots indicate LKS+ cells as proportion of all Lin– cells. (B) Absolute counts and (C) frequencies of LKS+ cells as proportion of all Lin– cells, four weeks after vaccination. Data points represent values from independent biological replicates (n = 4 per group; three males and one female per group; n = 12 total). Mean ± SEM, one-way ANOVA (ns = not significant, ***p < 0.001).
BCG-TICE is associated with elevated plasma cytokines following secondary in vivo challenge with LPS
To assess cytokine responses to heterologous stimulation, mice were challenged with i.p. LPS four weeks after BCG administration. Compared with unvaccinated controls, both BCG-Russia and BCG-TICE vaccination significantly increased IL-1β and IL-12 plasma levels four hours after intraperitoneal challenge with LPS (Fig 2A, B). Mice vaccinated with BCG-TICE exhibited significantly higher TNF-α, IL-1β, and IFN-γ plasma concentrations than mice vaccinated with BCG-Russia (Figs 2C and S2). However, there were no significant differences in plasma IL-10 and IL-12 levels in mice vaccinated with BCG-TICE versus BCG-Russia (Fig 2D, E).
(A-E) Plasma concentrations of IL-1β, IL-12, TNF-α, IL-10, and IL-2 four weeks after PBS or BCG vaccination and four hours after re-challenge (LPS, i.p.). TNF-α concentrations were assessed by ELISA, while all other reported cytokines were measured via the MDF10 multiplex assay. Data points represent values from individual mice (n = 4–6 per group; all female; n = 15 total). Concentration values from independent experimental repeats were normalized to the control group mean (set to 1.0) and expressed as fold changes. One outlier was removed from the TNF-α data set (BCG-Russia group) following an outlier test (ROUT method; Q = 1%). Mean ± SEM, one-way ANOVA (ns = not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).
BCG-TICE elicits stronger pro-inflammatory cytokine responses in macrophages
Acquisition of trained immunity in response to BCG is well-defined in cells of the monocyte/macrophage lineage. Moreover, results from in vivo challenge of BCG-trained mice were suggestive of a dominant role of monocytes and macrophages in the response to a secondary stimulus based on the heightened secretion of inflammatory mediators. To further investigate functional reprogramming of macrophages by the two BCG strains, we generated BMDMs and measured cytokine production following ex vivo restimulation. Compared with BCG-Russia or control groups, BMDMs from BCG-TICE–vaccinated mice exhibited significantly higher production of GM-CSF, IL-1β, IL-6, and TNF-α after a 24-hour stimulation with Pam3Cys (Fig 3A-D). Pam3Cys is a synthetic toll-like receptor (TLR) 1/2 agonist. No differences in CCL2, IL-10, or IL-12 levels were observed across groups after Pam3Cys stimulation (Fig 3E-G). In contrast, a 24-hour stimulation with LPS (a TLR4 agonist) resulted in significantly increased secretion of TNF-α by BMDMs isolated from mice vaccinated with BCG-TICE compared with BMDMs from mice vaccinated with BCG-Russia; however, the secretion of IL-1β, CCL2, and IL-10 was reduced in cultures of BMDMs from mice vaccinated with BCG-TICE relative to BCG-Russia. IL-12 levels in cultures of BMDMs from mice vaccinated with BCG-TICE were significantly increased compared with BMDMs from unvaccinated mice; however, levels of IL-12 were not significantly different in cultures of BMDMs obtained from mice vaccinated with BCG-TICE relative to BCG-Russia. Collectively, these findings indicate that BCG-TICE induces stronger cytokine responses than BCG-Russia after exposure certain stimuli, but also suppresses specific cytokines, indicating strain-dependent and stimulus-specific training effects.
(A-G) Concentrations of GM-CSF, IL-1β, IL-6, TNF-α, CCL2, IL-10, and IL-12 in supernatants of BMDMs from PBS- or BCG-vaccinated mice, 24 hours after restimulation. IL-6 concentrations were assessed by ELISA, while all other reported cytokines were measured via the MDF10 multiplex assay. Data points represent values from independent biological replicates (n = 3 per group; all male; n = 9 total). Mean ± SEM, two-way ANOVA (ns = not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).
Training with BCG-TICE triggers broad changes in the expression of myeloid-related genes and pathways upon reactivation
Given the divergent effects of the two strains of BCG on the release of inflammatory cytokines, we next determined whether transcriptional profiles in BMDMs were differentially regulated by BCG-TICE and BCG-Russia. We analyzed the expression of a targeted subset of myeloid genes in BMDMs using NanoString nCounter. In vivo training with BCG-Russia induced minimal transcriptional changes in LPS-restimulated macrophages, with just two genes (Ptprc, Gpr65) differentially expressed compared to controls (Fig 4A). Pathway analysis revealed limited enrichment, with single-gene representation in processes such as cell migration/adhesion and lymphocyte activation (Fig 4B). In contrast, BCG-TICE induced broad transcriptional reprogramming, with 127 genes differentially expressed relative to controls (Fig 4C). The most strongly upregulated genes (H2-Aa, CD74, H2-Ab1) were linked to processes such as antigen presentation and cell migration/adhesion. Several other pathways were highly enriched, with Th1 activation, angiogenesis, and cell migration/adhesion having the highest proportions of significantly differentially expressed genes (gene ratio > 0.38; Fig 4D). Notably, every pathway contained at least two differentially expressed genes, with some having over 30 significant genes.
(A) Volcano plot showing differentially expressed genes and (B) dot plot indicating pathway enrichment in BMDMs from mice trained with BCG-Russia compared with mice administered PBS (untrained controls). (C) Volcano plot showing differentially expressed genes and (D) dot plot indicating pathway enrichment in BMDMs from mice trained with BCG-TICE compared with mice administered PBS. Horizontal dashed line on volcano plots indicates Benjamini–Hochberg (BH) adjusted p-value of 0.05. Genes with log2FC < −1 are shaded blue; genes with log2FC > 1 are shaded red. In dot plots, colour scale is based on the minimum -log10(BH p-value) among genes represented in each pathway. Dot size is proportional to the total number of significant genes in each pathway. Gene ratio on the x-axis accounts for the number of genes that comprise each pathway, expressing the ratio of significant genes as a proportion of all pathway genes. Significant genes may be upregulated or downregulated. n = 3 independent biological replicates per group (all male; n = 9 total).
Directed global significance scores (dGSS) were low when comparing BCG-Russia with untrained macrophages, with only interferon signaling, Th1 activation, and complement activation showing modest increases (dGSS 2.1–2.6), none of which contained significantly upregulated genes (Fig 5A). In contrast, 84% of the pathways that emerged in our analysis were upregulated following BCG-TICE vaccination. The highest degree of upregulation was observed in T cell activation and checkpoint signalling (dGSS = 6.2), interferon signalling (dGSS = 5.1), and Th1 activation (dGSS = 4.8), all containing multiple significantly upregulated genes. Notably, several pathways showed opposing regulation, with pathways downregulated with BCG-Russia (e.g., Th2 activation) showing upregulation with BCG-TICE, whereas pathways suppressed by BCG-TICE (e.g., cell cycle and apoptosis) showing modest upregulation in BMDMs from mice vaccinated with BCG-Russia.
(A) Heatmap displaying directionality of pathway enrichment for each comparison. Values represent nSolver directed global significance scores. (B) Volcano plot showing differentially expressed genes and (C) dot plot indicating pathway enrichment in BMDMs from mice trained with BCG-TICE compared with mice trained with BCG-Russia. Horizontal dashed line on the volcano plot indicates Benjamini–Hochberg (BH) adjusted p-value of 0.05. Genes with log2FC < −1 are shaded blue; genes with log2FC > 1 are shaded red. In the dot plot, colour scale is based on the minimum -log10(BH p-value) among genes represented in each pathway. Dot size is proportional to the total number of significant genes in each pathway. Gene ratio on the x-axis accounts for the number of genes that comprise each pathway, expressing the ratio of significant genes as a proportion of all pathway genes. Significant genes may be upregulated or downregulated. n = 3 independent biological replicates per group (all male; n = 9 total).
Direct comparison of the two BCG strains revealed 51 significantly upregulated genes and 37 significantly downregulated genes in BMDMs from mice vaccinated with BCG-TICE (Fig 5B). Pathways related to Th2 activation, Th1 activation, and extracellular matrix (ECM) remodeling had the greatest representation of significantly differentially expressed genes (Fig 5C). Enrichment in ECM remodeling, cell migration/adhesion, growth factor signalling, and lymphocyte activation was driven by genes with the highest statistical significance. The pathways most strongly upregulated included ECM remodeling (dGSS = 10.4), Th1 (dGSS = 8.3) and Th2 (dGSS = 7.5) activation, and cell migration/adhesion (dGSS = 7.5; Fig 5A). These data indicate that vaccination with BCG-TICE elicits significantly greater transcriptional changes in immune-related genes of macrophages, representing more robust reactivation than vaccination with BCG-Russia.
Discussion
Divergent immunogenic properties among BCG strains are well characterized and recent studies have extended these differences to trained immunity [17]. In the present study, we investigated differences in trained immunity phenotypes induced by two strains of BCG used for the treatment of patients with NMIBC. The results revealed clear differences in the abilities of BCG-TICE and BCG-Russia to induce trained immunity, with BCG-TICE eliciting higher levels of trained immunity, demonstrated by enhanced inflammatory responses to subsequent stimuli. Following an in vivo heterologous inflammatory challenge, plasma cytokines were most elevated in mice that were vaccinated with BCG-TICE, despite both strains having similar effects on HSPC expansion. This is suggestive of independent mechanisms governing myelopoietic induction versus functional reprogramming during trained immunity acquisition. When macrophages from mice vaccinated with BCG-TICE were re-challenged ex vivo, they exhibited markedly enhanced secondary cytokine responses compared with those trained with BCG-Russia. Divergent transcriptional signatures further supported distinct trained immunity profiles between the two BCG strains.
Recent work has provided further evidence of a causal relationship between trained immunity and the therapeutic benefit of BCG [18], thereby supporting our suggestion that BCG-TICE may be linked to better patient outcomes based on our findings.
Our observation that both BCG-Russia and BCG-TICE trigger an expansion of hematopoietic progenitors in the bone marrow is concordant with previous studies [15,17]. Despite gross shifts in HSPC populations indicating non-specific measures of trained immunity, they are consistent with the systemic effects of intravesical BCG administration triggering myelopoiesis, a critical aspect of sustained innate immune memory [15]. Mature leukocytes that encounter BCG and readily acquire trained immunity may be short-lived, as is the case with circulating monocytes. Bone marrow HSPCs are known to undergo epigenetic and metabolic changes in response to BCG, leading to a long-lasting memory phenotype that is retained during cell division and differentiation. Consequently, trained bone marrow harbours a reservoir of myeloid cells, mostly monocytes, that enter the circulation and exert robust inflammatory responses upon reactivation. Our findings reveal that divergence in functional reprogramming, i.e., trained immunity, following exposure to different strains of BCG can be acquired after myeloid expansion has occurred.
Results of plasma cytokine analysis following in vivo LPS re-challenge provided further evidence that the two strains of BCG exert differential effects on functional reprogramming. The pattern of cytokines detected in elevated quantities following BCG-TICE–training is consistent with a monocyte/macrophage driven response. Significant differences were seen in the levels of certain acute-phase cytokines (IL-1β, TNF-α, IFN-γ) elicited between the two strains. IL-12 levels were significantly increased in BCG-treated mice compared with controls, but did not differ between the two strains, which could be explained by reduced susceptibility of dendritic cells to strain-dependent effects on trained immunity. Dendritic cells are known to be the main source of plasma IL-12 [19–21]. Moreover, we did not observe increased IL-12 production by trained macrophages that mirrored the heightened plasma concentrations, suggesting a minimal contribution of macrophages to systemic IL-12 concentrations. While it has been suggested that dendritic cells can acquire trained immunity [16,22], it remains to be determined whether this was the case in our model.
Ex vivo restimulation of BMDMs with LPS or Pam3Cys resulted in divergent cytokine profiles. This may be explained by the distinct PRRs that are selectively targeted by LPS (TLR4) versus Pam3Cys (TLR1/2). Notably, expression of Tlr2 was significantly upregulated in BCG-TICE– compared to BCG-Russia–trained BMDMs. It is hypothesized that, in the clinical setting, tumour-associated immunogens are responsible for triggering secondary responses in trained myeloid cells [12]. These molecules may act on a diverse repertoire of PRRs leading to robust reactivation, consistent with what was observed following TLR1/2 stimulation. Our findings were restricted to isolated stimulation of specific TLRs, whereas tumour damage-associated molecular patterns (DAMPs) can act on multiple PRRs concomitantly and chronically. However, TLR4 agonists are appropriate for this single-TLR approach, as TLR4 is engaged by several DAMPs (e.g., HMGB1, heat-shock proteins) that are present in the tumour microenvironment.
In contrast with a similar study by Xu et al. [17], our analysis revealed more limited transcriptional differences associated with BCG-Russia relative to untrained controls. A key methodological distinction between our study and the study of Xu et al. is that we assessed gene expression in re-challenged cells, whereas Xu et al, reported RNA sequencing of BMDMs trained ex vivo, without a secondary stimulus. We believe that assessing gene expression after restimulation is a strength of our study, as this reactivated state captures a more relevant trained immunity phenotype, rather than a trained, resting state.
In our study, macrophages from mice administered BCG-Russia displayed remarkably few alterations upon reactivation, with Ptprc and Gpr65 emerging as the only significantly differentially expressed genes, each with a fold-change < 0.25. The most differentially expressed genes when comparing BCG-TICE to control drove enrichment of several important pathways, with at least two of Sema4a, Cd74, H2-Ab1, and H2-Aa represented in cell migration/adhesion, Th1 activation, interferon signaling, antigen presentation, and T cell activation and checkpoint signaling. T cells play a central role in the therapeutic mechanism of BCG [23,24], with optimal efficacy relying on Th1 immunity [25,26]. Furthermore, interferon signalling in macrophages promotes polarization to a pro-inflammatory, M1-like, anti-tumoural phenotype and is responsible for directly stimulating cytotoxicity against bladder cancer cells [27–29]. This broadly supportive impact on anti-tumour immunity may not be limited to comparison against untrained cells, as pronounced upregulation of the Th1 and T cell activation pathways was also observed when directly comparing BCG-TICE to BCG-Russia. The high magnitude of ECM remodelling and cell migration/adhesion upregulation was primarily driven by high statistical significance of Itgal and Itga5 differential expression, each involved in both pathways. Itgal and Itga5 encode the alpha subunit (CD11a) of heterodimeric lymphocyte function-associated antigen 1 (LFA-1) and integrin alpha-5, respectively. Binding of CD11a to intercellular adhesion molecules is important for macrophage migration to lymph nodes [30]. It is possible, therefore, that Itgal upregulation enhances tumour antigen presentation by pro-inflammatory macrophages in secondary lymphoid tissue. The high-level consequences of ECM remodelling and cell migration by tumour-associated macrophages has been proposed to be a tumour-promoting effect because these cells often induce immune-evasive phenotypes in cancer cells [31]. However, adequate response to BCG in patients is associated with pro-inflammatory macrophages driving tumour clearance [31–33]. Taken together with upregulation of processes that promote Th1 immunity and T cell-mediated killing, we posit that enhanced macrophage infiltration in this context enhances anti-tumour immunity.
Overall, our study indicates that optimal induction of trained immunity is favoured by BCG-TICE, compared to BCG-Russia, as evidenced by more robust secondary responses. However, the precise mechanism by which BCG-TICE induces higher levels trained immunity is not known. Other groups have suggested a link to linoleic acid metabolism [17], but the genetic variations between strains that translate to stronger or weaker initial reprogramming have yet to be identified. Genomic divergence between BCG strains corresponds to molecular differences in key cell wall proteins, virulence factors, and PRR ligands [34], which are likely significant factors influencing differing trained immunity phenotypes induced by distinct BCG strains. For example, uniquely high expression of certain genes implicated in cell invasion and escape (e.g., mkl) have been attributed to BCG-TICE [34]. Duplication of the ESX-5 locus in BCG-TICE also leads to upregulation of several ESX-5–related genes [34]. In contrast, BCG-Russia has unusually low expression of the mce1R gene, which is a transcriptional regulator of the mce1 operon [34]. Further research is needed to determine whether expression of these immunogenic proteins directly influences the levels of trained immunity acquired.
While we delivered BCG intravenously for strong and reliable induction of trained immunity, this is not a viable approach in the clinical setting due to the high risk of sepsis. This model was favoured for strong, reliable induction of central trained immunity, eliminating the variable reprogramming that results when mice are administered intravesical BCG. As intravenous BCG administration provides direct access to the bone marrow for induction of reprogramming, it is possible that intravesical delivery produces differing trained immunity phenotypes. This may include weaker secondary responses. However, a recent study reported that treatment with intravesical versus intravenous BCG produces similar trained immunity phenotypes, provided that at least five weekly intravesical instillations are performed [18]. This suggests that the route of administration is unlikely to substantially affect the extent of central trained immunity acquisition, thereby supporting the translational relevance of our findings despite the fact that BCG was administered intravenously. Novel approaches to optimizing acquisition of central trained immunity are currently under investigation in pre-clinical studies. Such strategies do not involve systemic dissemination of live bacteria, hence are likely to have desirable safety profiles and be well-tolerated by patients. Delivery of purified components of the BCG cell wall and nanoparticle formulations are under investigation [35]. In addition, adjuvants for optimal trained immunity induction after intravesical BCG, such as β-glucan or linoleic acid [17,36], have been suggested. While still preserving the important local effects of BCG therapy, these novel therapeutic strategies have the potential to promote consistent and stronger trained immunity acquisition in NMIBC patients.
In conclusion, the present report provides novel evidence of superior trained immunity responses linked to BCG-TICE, compared to BCG-Russia. These findings could inform treatment of patients with high-risk NMIBC given the potential role of trained immunity in the mode of action of BCG. Systemic immune activation is critical for effective immunotherapy, including intravesical therapy. Further studies and clinical trials, like the ongoing EVER trial (NCT05037279), are needed to confirm the impact of BCG strain on patient outcomes.
Supporting information
S1 Fig. Flow cytometric gating strategy for identification of HSPCs.
Doublets and debris were excluded through gating by forward scatter (FSC-A) against FSC-H and side scatter (SSC-A). Dead cells were excluded based on positive staining for viability dye. Lineage-committed cells were excluded. Subsequent gate identifying HSPCs is shown in Fig 1a.
https://doi.org/10.1371/journal.pone.0357950.s001
(TIFF)
S2 Fig. Strain-dependent effects of BCG on IFN-γ plasma levels.
Fluorescence intensity (FI; arbitrary units) reflects plasma concentrations of IFN-γ four weeks after PBS or BCG vaccination and four hours after re-challenge (LPS, i.p.). Data points represent values from individual mice (all female). FI values from two independent experimental repeats are reported. Mean ± SEM, one-way ANOVA (ns = not significant, *p < 0.05, ***p < 0.001).
https://doi.org/10.1371/journal.pone.0357950.s002
(TIFF)
S1 Table. Raw values of plasma cytokine concentrations.
https://doi.org/10.1371/journal.pone.0357950.s003
(DOCX)
References
- 1. Abdallah AM, Behr MA. Evolution and strain variation in BCG. Adv Exp Med Biol. 2017;1019:155–69. pmid:29116634
- 2. Bhindi B, Kool R, Kulkarni GS, Siemens DR, Aprikian AG, Breau RH, et al. Canadian Urological Association guideline on the management of non-muscle-invasive bladder cancer - Full-text. Can Urol Assoc J. 2021;15(8):E424–60. pmid:33938798
- 3. Del Giudice F, Busetto GM, Gross MS, Maggi M, Sciarra A, Salciccia S, et al. Efficacy of three BCG strains (Connaught, TICE and RIVM) with or without secondary resection (re-TUR) for intermediate/high-risk non-muscle-invasive bladder cancers: results from a retrospective single-institution cohort analysis. J Cancer Res Clin Oncol. 2021;147(10):3073–80. pmid:33675400
- 4. Sengiku A, Ito M, Miyazaki Y, Sawazaki H, Takahashi T, Ogura K. A prospective comparative study of intravesical bacillus Calmette-Guérin therapy with the Tokyo or Connaught strain for nonmuscle invasive bladder cancer. J Urol. 2013;190(1):50–4. pmid:23376145
- 5. Vegt PD, Witjes JA, Witjes WP, Doesburg WH, Debruyne FM, van der Meijden AP. A randomized study of intravesical mitomycin C, bacillus Calmette-Guerin Tice and bacillus Calmette-Guerin RIVM treatment in pTa-pT1 papillary carcinoma and carcinoma in situ of the bladder. J Urol. 1995;153(3 Pt 2):929–33. pmid:7853577
- 6. Witjes JA, Dalbagni G, Karnes RJ, Shariat S, Joniau S, Palou J, et al. The efficacy of BCG TICE and BCG Connaught in a cohort of 2,099 patients with T1G3 non-muscle-invasive bladder cancer. Urol Oncol. 2016;34(11):484.e19–484.e25. pmid:27639776
- 7. Flury-Sutter S, Heuzeroth F, Arbelaez E, Bubendorf L, Püschel H, Hayoz S, et al. Efficacy and tolerability of bacillus Calmette-Guérin strain Russia for the treatment of non-muscle-invasive bladder cancer analysis of a prospective registry. Can Urol Assoc J. 2024;18(7):E233–9. pmid:38466867
- 8. Shelley MD, Court JB, Kynaston H, Wilt TJ, Fish RG, Mason M. Intravesical Bacillus Calmette-Guerin in Ta and T1 Bladder Cancer. Cochrane Database Syst Rev. 2000;2000(4):CD001986. pmid:11034738
- 9. Witjes JA. Management of BCG failures in superficial bladder cancer: a review. Eur Urol. 2006;49(5):790–7. pmid:16464532
- 10. Cambier S, Sylvester RJ, Collette L, Gontero P, Brausi MA, van Andel G, et al. EORTC nomograms and risk groups for predicting recurrence, progression, and disease-specific and overall survival in non-muscle-invasive stage Ta-T1 urothelial bladder cancer patients treated with 1-3 years of maintenance Bacillus Calmette-Guérin. Eur Urol. 2016;69(1):60–9. pmid:26210894
- 11. Netea MG, Domínguez-Andrés J, Barreiro LB, Chavakis T, Divangahi M, Fuchs E, et al. Defining trained immunity and its role in health and disease. Nat Rev Immunol. 2020;20(6):375–88. pmid:32132681
- 12. van Puffelen JH, Keating ST, Oosterwijk E, van der Heijden AG, Netea MG, Joosten LAB, et al. Trained immunity as a molecular mechanism for BCG immunotherapy in bladder cancer. Nat Rev Urol. 2020;17(9):513–25. pmid:32678343
- 13. Graham CH, Paré J-F, Cotechini T, Hopman W, Hindmarch CCT, Ghaffari A, et al. Innate immune memory is associated with increased disease-free survival in bladder cancer patients treated with bacillus Calmette-Guérin. Can Urol Assoc J. 2021;15(8):E412–7. pmid:33410746
- 14. Paré J-F, Tabasinezhad M, Grossman A, Atallah A, Hindmarch CCT, Tyryshkin K, et al. Association of Histone H3 trimethylation in circulating monocytes with lack of early recurrence in patients with bladder cancer following BCG induction therapy. Bladder Cancer. 2023;9(2):175–86. pmid:38993297
- 15. Kaufmann E, Sanz J, Dunn JL, Khan N, Mendonça LE, Pacis A, et al. BCG educates hematopoietic stem cells to generate protective innate immunity against tuberculosis. Cell. 2018;172(1–2):176–190.e19. pmid:29328912
- 16. Atallah A, Grossman A, Nauman RW, Paré J-F, Khan A, Siemens DR, et al. Systemic versus localized Bacillus Calmette Guérin immunotherapy of bladder cancer promotes an anti-tumoral microenvironment: Novel role of trained immunity. Int J Cancer. 2024;155(2):352–64. pmid:38483404
- 17. Xu J-C, Chen Z-Y, Huang X-J, Wu J, Huang H, Niu L-F, et al. Multi-omics analysis reveals that linoleic acid metabolism is associated with variations of trained immunity induced by distinct BCG strains. Sci Adv. 2024;10(14):eadk8093. pmid:38578989
- 18. Daman AW, Antonelli AC, Redelman-Sidi G, Paddock L, Khayat S, Ketavarapu M, et al. Microbial cancer immunotherapy reprograms hematopoiesis to enhance myeloid-driven anti-tumor immunity. Cancer Cell. 2025;43(8):1442–1459.e10. pmid:40446799
- 19. Hildreth AD, Padilla ET, Tafti RY, Legala AR, O’Sullivan TE. Sterile liver injury induces a protective tissue-resident cDC1-ILC1 circuit through cDC1-intrinsic cGAS-STING-dependent IL-12 production. Cell Rep. 2023;42(2):112141. pmid:36807146
- 20. Pompei L, Jang S, Zamlynny B, Ravikumar S, McBride A, Hickman SP, et al. Disparity in IL-12 release in dendritic cells and macrophages in response to Mycobacterium tuberculosis is due to use of distinct TLRs. J Immunol. 2007;178(8):5192–9. pmid:17404302
- 21. Siegemund S, Schütze N, Freudenberg MA, Lutz MB, Straubinger RK, Alber G. Production of IL-12, IL-23 and IL-27p28 by bone marrow-derived conventional dendritic cells rather than macrophages after LPS/TLR4-dependent induction by Salmonella Enteritidis. Immunobiology. 2007;212(9–10):739–50. pmid:18086375
- 22. Tepale-Segura A, Gajón JA, Muñoz-Cruz S, Castro-Escamilla O, Bonifaz LC. The cholera toxin B subunit induces trained immunity in dendritic cells and promotes CD8 T cell antitumor immunity. Front Immunol. 2024;15:1362289. pmid:38812523
- 23. Ratliff TL, Ritchey JK, Yuan JJ, Andriole GL, Catalona WJ. T-cell subsets required for intravesical BCG immunotherapy for bladder cancer. J Urol. 1993;150(3):1018–23. pmid:8102183
- 24. Biot C, Rentsch CA, Gsponer JR, Birkhäuser FD, Jusforgues-Saklani H, Lemaître F, et al. Preexisting BCG-specific T cells improve intravesical immunotherapy for bladder cancer. Sci Transl Med. 2012;4(137):137ra72. pmid:22674550
- 25. Saint F, Patard JJ, Groux Muscatelli B, Lefrere Belda MA, Gil Diez de Medina S, Abbou CC, et al. Evaluation of cellular tumour rejection mechanisms in the peritumoral bladder wall after bacillus Calmette-Guérin treatment. BJU Int. 2001;88(6):602–10. pmid:11678759
- 26. Pichler R, Gruenbacher G, Culig Z, Brunner A, Fuchs D, Fritz J, et al. Intratumoral Th2 predisposition combines with an increased Th1 functional phenotype in clinical response to intravesical BCG in bladder cancer. Cancer Immunol Immunother. 2017;66(4):427–40. pmid:28005163
- 27. Luo Y, Yamada H, Evanoff DP, Chen X. Role of Th1-stimulating cytokines in bacillus Calmette-Guérin (BCG)-induced macrophage cytotoxicity against mouse bladder cancer MBT-2 cells. Clin Exp Immunol. 2006;146(1):181–8. pmid:16968412
- 28. Vidyarthi A, Khan N, Agnihotri T, Negi S, Das DK, Aqdas M, et al. TLR-3 stimulation skews M2 macrophages to M1 through IFN-αβ signaling and restricts tumor progression. Front Immunol. 2018;9:1650. pmid:30072995
- 29. Kerneur C, Cano CE, Olive D. Major pathways involved in macrophage polarization in cancer. Front Immunol. 2022;13:1026954. pmid:36325334
- 30. Schittenhelm L, Hilkens CM, Morrison VL. Beta(2) integrins as regulators of dendritic cell, monocyte, and macrophage function. Front Immunol. 2017;8:1866.
- 31. Leblond MM, Zdimerova H, Desponds E, Verdeil G. Tumor-associated macrophages in bladder cancer: biological role, impact on therapeutic response and perspectives for immunotherapy. Cancers (Basel). 2021;13(18):4712. pmid:34572939
- 32. Martínez-López MF, de Almeida CR, Fontes M, Mendes RV, Kaufmann SHE, Fior R. Macrophages directly kill bladder cancer cells through TNF signaling as an early response to BCG therapy. Dis Model Mech. 2024;17(8):dmm050693. pmid:39114912
- 33. Tan Q-L, Zhou C-Y, Cheng L, Luo M, Liu C-P, Xu W-X, et al. Immunotherapy of Bacillus Calmette‑Guérin by targeting macrophages against bladder cancer in a NOD/scid IL2Rg‑/‑ mouse model. Mol Med Rep. 2020;22(1):362–70. pmid:32319653
- 34. Abdallah AM, Hill-Cawthorne GA, Otto TD, Coll F, Guerra-Assunção JA, Gao G, et al. Genomic expression catalogue of a global collection of BCG vaccine strains show evidence for highly diverged metabolic and cell-wall adaptations. Sci Rep. 2015;5:15443. pmid:26487098
- 35. Masuda H, Nakamura T, Noma Y, Harashima H. Application of BCG-CWS as a systemic adjuvant by using nanoparticulation technology. Mol Pharm. 2018;15(12):5762–71. pmid:30380885
- 36. Jurado LF, Daman AW, Li Z, Ross VMS, Nikolaou K, Tran KA, et al. A fungal-derived adjuvant amplifies the antitumoral potency of Bacillus Calmette-Guérin via reprogramming granulopoiesis. Immunity. 2025;58(8):1984–2001.e6. pmid:40555239