Iwata 2025 — Ewingella americana as Novel Antitumor Bacterial Agent (Gut Microbes)

Iwata S, Yamasita N, Asukabe K, Sakari M, Miyako E · Gut Microbes · 2025 · Vol. 17, No. 1, article 2599562 · DOI: 10.1080/19490976.2025.2599562 · PMC: PMC12710904 · Open Access (CC-BY-NC)

N. Yamasita and K. Asukabe contributed equally. Corresponding author: Eijiro Miyako (e-miyako@jaist.ac.jp), JAIST, Nomi, Ishikawa, Japan. IACUC Approval No. 07-001 (JAIST).

TL;DR

A systematic screen of 45 gut bacterial strains isolated from three lower-vertebrate species (Japanese tree frog, fire-belly newt, Japanese grass lizard) identified nine strains safe enough for IV administration in immunocompetent BALB/c mice. Of these, Ewingella americana — a Gram-negative facultative anaerobe from Dryophytes japonicus — achieved 100% complete tumor regression (CR) in a syngeneic Colon-26 subcutaneous carcinoma model after a single intravenous dose, outperforming both anti-PD-L1 antibody (1/5 CR) and liposomal doxorubicin (0/5 CR) administered on standard multi-dose regimens. Tumor rechallenge experiments demonstrated durable immunological memory persisting beyond 60 days. Mechanistically, E. americana colonizes tumors with ~3,000-fold intratumoral expansion between 3 h and 24 h post-injection, induces direct bacterial cytotoxicity via secreted hemolysin and exotoxin, and activates T cells, B cells, and neutrophils within the tumor microenvironment. A comprehensive safety panel showed no hematological toxicity, no organ histopathology, and blood clearance by 24 h post-injection. needs-human-replication

Study Design and Cohort

Model: Syngeneic subcutaneous Colon-26 (murine colorectal carcinoma) model in female BALB/c mice (BALB/cCrSlc strain), 5–6 weeks old, 18–21 g at purchase (Japan SLC, Shizuoka, Japan). Pathogen-free, 22 ± 2 °C, 12-hour light/dark, ad libitum chow and water.

Tumor implantation: 1 × 10⁶ Colon-26 cells in 100 µL RPMI-1640, subcutaneous right flank. Treatment initiated when tumors reached ~200 mm³ (typically 7–10 days post-inoculation). Tumor volume measured daily by digital caliper: V = L × W²/2.

Group sizes:

Initial 9-strain screen: n = 3 per bacterial strain group, n = 3 PBS control (Figure 2)
Head-to-head comparative study (E. americana vs anti-PD-L1 vs DOX): n = 5 per group (Figure 3)
Safety / CBC / histology: n = 5 per group (Figure 6)
Tumor rechallenge: n = 10 E. americana-cured mice vs 10 naïve age-matched controls (Supp. Figure S5)
Intratumoral colonization kinetics: n = 3 independent experiments (Figure 4a)

Dose and route: All bacterial treatments: 200 µL suspension at 5 × 10⁹ CFU/mL (= 1 × 10⁹ CFU total) via tail-vein injection, single dose. Dose-finding pilot (n = 5 per dose group; Supp. Figure S1): three discrete doses tested — 2 × 10⁷, 2 × 10⁸, and 1 × 10⁹ CFU in 200 µL. 1 × 10⁹ CFU identified as the optimal dose achieving complete tumor regression while maintaining acceptable safety profiles. Doses exceeding 1 × 10⁹ CFU resulted in acute lethality — establishing this as the maximum tolerated dose. PBS negative control: equivalent volume. needs-supp-verification (dose-response curve detail is in Supp. Figure S1)

Comparator dosing (Figure 3): Anti-PD-L1 antibody and liposomal doxorubicin (DOX) each administered IV every other day × 4 total doses, 2.5 mg/kg per dose; total 10 mg/kg each. E. americana: single IV dose only.

Culture medium for E. americana: Pearl Core E-MC64 medium (Eiken Chemical Co., Ltd., Tokyo, Japan); bacteria suspended in their growth medium (not PBS) to preserve viability.

Observation period: Up to 40 days post-tumor implantation for the initial screen (Figure 2f); up to 60 days for the comparative efficacy study (Figure 3e). Humane endpoint: tumor volume > 1,500 mm³, body weight loss > 20%, or signs of severe distress.

Statistical analysis: Two-sided Student’s t-test for group comparisons; Kaplan-Meier + log-rank (Mantel-Cox) for survival. All experiments performed ≥ 3 independent replicates minimum. GraphPad Prism v9.4.0.

Bacterial Isolation and Screening

Source animals: 45 distinct bacterial strains isolated from intestinal tracts of three lower-vertebrate species ¹:

Dryophytes japonicus (Japanese tree frog) — Ishikawa Prefecture, Japan
Cynops pyrrhogaster (Japanese fire-belly newt) — Okayama Prefecture, Japan
Takydromus tachydromoides (Japanese grass lizard) — Ishikawa Prefecture, Japan

Isolation method: Intestinal swabs plated on five selective agar media (desoxycholate, mannitol salt, standard method, LB, ATCC 543 agar); incubated anaerobically 3 days at 25 °C; individual colonies picked by sterile syringe needle under stereomicroscopic guidance; pure cultures established by serial streaking.

Species identification: 16S rRNA gene sequencing (BEX Co., Ltd., Tokyo) via Sanger sequencing + NCBI GenBank BLAST.

Preliminary biocompatibility screen (all 45 strains): Single IV tail-vein injection of 200 µL at 5 × 10⁹ CFU/mL into female BALB/c mice (7 weeks, n = 3/group). Criteria for advancement: (1) survival ≥ 7 days, (2) body weight loss < 20%, (3) no severe lethargy, respiratory distress, or neurological signs, (4) normal behavioral patterns and feeding (Supp. Table S1).

Nine strains advanced for antitumor evaluation (three from each host species) ¹:

Dryophytes japonicus: Priestia aryabhattai (No. 1), Rhodococcus qingshengii (No. 2), Ewingella americana (Nos. 3, 6, 7, 9, 10, 11 — multiple isolates)
Cynops pyrrhogaster: Citrobacter portucalensis (No. 16), Chryseobacterium gambrini (No. 22), Enterobacter ludwigii (No. 24)
Takydromus tachydromoides: Rathayibacter oskolensis (Nos. 28, 42), Microbacterium oxydans (Nos. 29, 32, 43), Arthrobacter humicola (No. 30)

In Vivo Antitumor Efficacy — Initial 9-Strain Screen

Results in the syngeneic Colon-26 model (n = 3 per group; 200 µL, 5 × 10⁹ CFU/mL IV; PBS control) ¹:

No antitumor activity:

Priestia aryabhattai — tumor growth indistinguishable from PBS controls

Significant tumor growth suppression (no complete regression):

Rhodococcus qingshengii — initial suppression to day 5, followed by tumor re-growth (transient effect)
Chryseobacterium gambrini — significant suppression vs PBS
Rathayibacter oskolensis — significant suppression vs PBS
Microbacterium oxydans — significant suppression vs PBS
Arthrobacter humicola — significant suppression vs PBS

Tumor regression achieved (complete or transient):

Citrobacter portucalensis — complete tumor regression by day 3; however, tumor recurrence observed after day 8, indicating incomplete long-term control
Enterobacter ludwigii — complete tumor regression by day 3; tumor recurrence after day 8 (same pattern as C. portucalensis)
Ewingella americana — complete tumor regression with no recurrence; 100% survival at 40-day endpoint (Kaplan-Meier log-rank p < 0.0001 vs PBS)

Critical observation: The three strains achieving tumor regression — E. americana, C. portucalensis, and E. ludwigii — are all facultative anaerobes. The authors propose this is mechanistically consistent with the selective colonization of hypoxic tumor microenvironments ¹. The other strains achieving suppression but not regression include obligate aerobes (e.g., R. qingshengii, M. oxydans, A. humicola), which may lack intrinsic tumor-targeting capability.

Body weight across all groups remained within ± 20% of baseline throughout the monitoring period ¹.

Primary Finding: E. americana vs Standard-of-Care Comparators

Head-to-head comparison in the Colon-26 model (n = 5 per group) ¹:

Treatment	Regimen	CR Rate (day 30)	Survival (day 60)
PBS	—	0/5 (0%)	0/5 (0%)
Liposomal DOX	4 × 2.5 mg/kg IV q2d	0/5 (0%)	low (Figure 3e)
Anti-PD-L1 antibody	4 × 2.5 mg/kg IV q2d	1/5 (20%)	low (Figure 3e)
E. americana	Single dose, 1 × 10⁹ CFU IV	5/5 (100%)	5/5 (100%) at ≥30 days beyond PBS

CR rate: p < 0.0001 (E. americana vs PBS, DOX, and anti-PD-L1 combined; Student’s two-sided t-test, Figure 3d).

Kaplan-Meier survival: log-rank (Mantel-Cox) p < 0.0001 (E. americana vs PBS; Figure 3e). Both anti-PD-L1 and DOX significantly suppressed tumor growth vs PBS but did not achieve consistent CR.

Dosing asymmetry: E. americana achieved 100% CR with a single administration. Anti-PD-L1 and DOX required four doses each on an every-other-day schedule (a total of ~8 days of treatment). This is a deliberate asymmetry: the bacterial treatment’s single-dose sufficiency reflects active tumor-homing and proliferation rather than bolus pharmacokinetics.

The authors note that liposomal DOX was chosen as the comparator (rather than free DOX) to represent state-of-the-art delivery technology; the result demonstrates that biological tumor-targeting outperforms passive EPR-based accumulation ¹.

Durable Immunological Memory

Tumor rechallenge experiment (Supp. Figure S5; main text p. 6 summary confirmed): On day 30 post-E. americana treatment, cured mice (n = 10) received subcutaneous injection of 1 × 10⁶ Colon-26 cells into the contralateral flank. Naïve age-matched BALB/c controls (n = 10) received the same challenge. Tumor growth monitored an additional 30 days (Methods p. 18) ¹. needs-supp-verification (full Kaplan-Meier curve and statistics are in Supp. Fig S5; main text confirms the 0/10 vs 10/10 outcome)

0/10 cured mice developed tumors
10/10 naïve controls developed tumors

Immunological memory persisted > 60 days (total observation from treatment initiation). The authors interpret this as establishment of adaptive immune memory against Colon-26 tumor antigens — a feature absent in standard chemotherapy and uncommon with single-agent checkpoint inhibition at the doses tested.

Mechanism: Intratumoral Colonization Dynamics

Colony assay from excised Colon-26 tumors following single IV E. americana injection (200 µL, 5 × 10⁹ CFU/mL; n = 3 independent experiments; Figure 4a) ¹:

Timepoint post-IV	Intratumoral bacterial load
0 h	~0 CFU/g (baseline)
3 h	detectable but low (Figure 4a baseline)
6 h	increased
24 h	~3,000-fold increase vs 3 h
48 h	further increase (highest measured titer)

Tumor-exclusive colonization: Colony assays on liver, spleen, kidney, heart, and lung at 24 h and 48 h post-injection recovered zero CFU in all five organs, confirming tumor-selective accumulation ¹. Physiological hypoxia in normal tissues (intestinal crypts, renal medulla) appears necessary but insufficient — the full tumor microenvironment constellation (CD47 “don’t eat me” overexpression, necrotic regions, aberrant vasculature, altered extracellular matrix) collectively creates the permissive niche.

Blood clearance: Bacterial colony assays in blood (collected from inferior vena cava at 5 min, 3 h, 24 h, 48 h post-injection) showed complete clearance by 24 h (Supp. Figure S8; main text p. 12 confirmed) ¹. Rapid blood clearance explains the acceptable safety profile despite a narrow therapeutic window. needs-supp-verification (colony count data and kinetic curve are in Supp. Fig S8)

Mechanism: Direct Cytotoxicity

Three-dimensional Colon-26 tumor spheroid co-culture experiments (Figure 4b-c; Supp. Data 1) ¹:

E. americana co-cultured with Colon-26 spheroids at 5 × 10⁴, 5 × 10⁵, 5 × 10⁶, 5 × 10⁷, and 5 × 10⁸ CFU
At the highest concentration (5 × 10⁸ CFU), spheroids were largely destroyed within 24 h
Dose-dependent spheroid disruption and cancer cell death observed across the concentration range
Time-lapse imaging documented progressive spheroid disintegration over 24–120 h

Cytolysins implicated: Virulence factor analysis of the type strain genome (ATCC 33852; NCBI RefSeq GCF_000735345.1; analyzed via Type Strain Genome Database) identified hemolysin and exotoxin as the likely direct-cytotoxicity mediators, confirmed by Supp. Figure S6 ¹. Important caveat: This virulence factor inference was derived from the publicly available type-strain genome — NOT from genome sequencing of the actual D. japonicus-derived isolate used in the experiments. genome-not-sequenced

CCK-8 (Cell Counting Kit-8) monolayer cytotoxicity assays against Colon-26 cells corroborated dose-dependent cancer cell elimination at all tested bacterial concentrations (Supp. Figure S7) ¹.

Mechanism: Immune Activation

Immunohistochemistry (IHC) — Day 1 post-treatment (Figure 4d-e)

Tumor sections stained for NKp46 (NK cells), F4/80 (macrophages), CD19 (B cells), CD3 (T cells), CXCR4 (neutrophils), TNF-α, Caspase-3, TUNEL + H&E. Quantification across 10 independent fields per tumor; mean ± SEM ¹:

Marker	Cell type / function	Result vs PBS
CD19 (B cell)	B lymphocytes	+3% (p < 0.001)
CD3 (T cell)	T lymphocytes	+5% (p < 0.0001)
CXCR4 (neutrophil)	Neutrophils	+30% (p < 0.0001)
TNF-α	Pro-inflammatory cytokine	significantly elevated (p < 0.0001)
Caspase-3	Apoptosis effector	significantly elevated (p < 0.0001)
TUNEL	Apoptotic DNA fragmentation	significantly elevated (p < 0.0001)
NKp46 (NK cell)	NK lymphocytes	NOT significantly elevated (ns)
F4/80 (macrophage)	Macrophages	NOT significantly elevated (ns)

The neutrophil response (+30% CXCR4+) is the largest single immune-cell expansion. Neutrophil effector mechanisms — NET formation, direct phagocytosis of tumor cells, and pro-inflammatory cytokine secretion — are proposed as major contributors to bacterial-mediated tumor destruction ¹.

Quantitative PCR — 6 h post-IV injection (Figure 4f)

qPCR on tumor tissue for immune markers (log₁₀ fold-change vs untreated control; ACTB as internal control; n = 3; mean ± SEM) ¹:

Gene	Cell / function	Direction	Significance
NKp46	NK cells	↓	p < 0.0001
F4/80	Macrophages	↓	p < 0.0001
CD19	B cells	↑	p < 0.01
CD3	T cells	↑	p < 0.001
Ly6G	Neutrophils	↑	p < 0.0001
IFN-γ	Th1 cytokine	↑	p < 0.0001
TNF-α	Pro-inflammatory cytokine	↑	p < 0.0001

NKp46 and F4/80 transcript levels are decreased at 6 h, consistent with rapid NK cell and macrophage depletion from the tumor tissue or redistribution. At 24 h (IHC timepoint), neither NK nor macrophage populations show significant changes — suggesting these lineages are neither recruited nor driven to expansion by E. americana within the 24-hour window studied. In contrast, T cells, B cells, and neutrophils show early qPCR upregulation (6 h) and sustained IHC elevation (24 h), indicating a coordinated adaptive-innate axis centered on these three populations.

IFN-γ elevation supports Th1 polarization and enhanced antigen presentation; TNF-α promotes T cell activation and proliferation. Both cytokines are also pro-apoptotic in the tumor microenvironment.

Proposed mechanism summary (Figure 5): E. americana selectively colonizes and proliferates within the hypoxic/immunosuppressive tumor microenvironment via its facultative anaerobic character. Direct oncolytic activity is mediated by secreted hemolysin + exotoxin. Host immune activation proceeds through PAMP (pathogen-associated molecular pattern) recognition via TLRs and NLRs, triggering innate inflammatory cascades that recruit neutrophils and T and B cells. The combined direct cytotoxicity + immune-mediated tumor destruction drives complete regression. After tumor elimination, persistent activated T and B cells provide immunological memory against tumor antigens.

E. americana Characterization

Taxonomy: Gram-negative, facultative anaerobic Enterobacteriaceae; isolated from Dryophytes japonicus (Japanese tree frog), Ishikawa Prefecture, Japan ¹
Pathogenicity: Low; associated with opportunistic infections in neonates and immunocompromised patients in published clinical literature (references 28, 29 in paper: Ioannou et al. 2024, Esposito et al. 2019) ²; generally not a pathogen in immunocompetent hosts. A 2025 case report (Ayik et al. Cureus 2025; PMID 40636648) documents E. americana sepsis in a cancer patient receiving chemotherapy, resolved with ceftriaxone — consistent with Iwata 2025’s characterization of antibiotic susceptibility as a key safety advantage ³
Antibiotic susceptibility: Susceptible to multiple clinically available antimicrobial agents — cited as a key safety advantage, as bacterial complications from treatment could be managed by targeted antibiotic intervention ¹
Growth optimum: Pearl Core E-MC64 medium at 25 °C; 5–10 day culture to optimal density
Virulence factor inference: Nutritional/metabolic factors, effector delivery systems, immune modulation, motility, adherence, regulation, biofilm, stress survival, invasion, and exotoxin identified from the ATCC 33852 reference genome (Figure 4c; Type Strain Genome Database) genome-not-sequenced
CD47 hypothesis (interpretive, not directly tested): The authors propose that E. americana’s tumor-exclusive localization is partly driven by tumor cell overexpression of CD47 (the “don’t eat me” signal), which suppresses phagocytic clearance of bacteria in the tumor but not in healthy organs where macrophage and NK cell activity rapidly eliminates circulating bacteria. This is consistent with the rapid blood clearance (undetectable by 24 h) and zero non-tumor colonization, but CD47 was not directly tested or blocked in this paper ¹. no-mechanism

Safety

CBC at day 30 post-treatment (Figure 6a; n = 5; mean ± SEM): All parameters within normal physiological ranges for BALB/c mice (manufacturer reference ranges); all comparisons E. americana vs PBS = not significant ¹:

Parameter	Comparison	Significance
WBC	vs PBS	ns
RBC	vs PBS	ns
HGB	vs PBS	ns
HCT	vs PBS	ns
MCV	vs PBS	ns
MCH	vs PBS	ns
MCHC	vs PBS	ns
PLT	vs PBS	ns

Organ histopathology (H&E; day 30; Figure 6b): Liver, spleen, heart, lung, and kidney examined by veterinary pathologists blinded to treatment. No evidence of tissue damage, inflammatory infiltration, necrosis, or other pathological alterations in any organ ¹.

Biochemical safety panel: Liver function (ALT, AST, ALP, total bilirubin), kidney function (BUN, creatinine), metabolic parameters (glucose, total protein, albumin), and electrolyte balance (Na, K, Cl) analyzed at days 1, 2, 3, 10, and 30. No significant differences vs PBS (Supp. Table S11) ¹.

Acute toxicity (Supp. Tables S14–S17): Hematological and biochemical analyses at 24 h, 48 h, 72 h, and 240 h post-injection. Acute responses are transient and self-limiting, resolving within 240 h (10 days) ¹. needs-supp-verification (detailed acute-phase CBC values are in Supp. Tables S14–S17; main text confirms resolution within 240 h)

Body weight: Body weight changes remained within ± 20% of baseline for all treatment groups throughout the monitoring periods. A slight body weight reduction was noted shortly after E. americana injection (consistent with mild acute inflammatory response), but no significant between-group differences were observed by day 15 post-treatment (Figure 3f) ¹.

Narrow therapeutic window: 1 × 10⁹ CFU was both the optimal efficacious dose and the maximum tolerated dose; doses exceeding this resulted in acute lethality (pilot dose-finding study, Supp. Figure S1). This narrow window is flagged by the authors as a critical consideration for clinical translation.

Comparison with Engineered Bacterial Therapeutics

The paper positions E. americana against the existing engineered-bacteria literature, particularly:

Salmonella VNP20009 (attenuated, genetically modified for tumor targeting) — safety concerns from spontaneous reversion needs-precedent-citation
Listeria monocytogenes attenuated strains — spontaneous mutation concerns needs-precedent-citation
E. coli Nissle-based engineered strains — spontaneous mutation safety risk needs-precedent-citation

The authors’ argument is that naturally occurring, antibiotic-susceptible strains with intrinsic tumor-targeting capability offer a superior safety profile: no risk of engineered-gene loss or spontaneous mutation restoring pathogenicity, and a clear antibiotic-intervention fallback. The claim that this is the “first reported demonstration that a naturally occurring gut bacterium isolated from a wild host organism achieved complete tumor regression following a single intravenous administration” is presented by the authors ¹, but this priority claim cannot be independently verified without a comprehensive literature sweep. needs-replication

Limitations and Gaps

Subcutaneous model, not orthotopic. The Colon-26 subcutaneous flank model recapitulates syngeneic immune interactions but does not reproduce the anatomical tumor microenvironment of colorectal cancer (mesenteric vasculature, liver proximity, GI peristalsis). Orthotopic colorectal models would provide stronger translational evidence. needs-human-replication
Virulence factor analysis from the type-strain reference genome, not the experimental isolate. The study’s virulence factor profiling (Figure 4c) used the publicly available ATCC 33852 genome (NCBI RefSeq GCF_000735345.1). The actual D. japonicus-isolated strain was NOT genome-sequenced. Differences in virulence gene content, IS-element insertions, or plasmid complement between the environmental isolate and the ATCC reference strain are unknown. genome-not-sequenced
Small per-group n in initial screen (n = 3). The 9-strain primary antitumor screen used n = 3 mice per group. While consistent with pilot-scale in-vivo oncology, this n is underpowered for detecting moderate-effect strains or for statistical modeling of dose-response. The head-to-head comparison used n = 5, which is slightly better but still small.
CD47 hypothesis not directly tested. Tumor-selective colonization is attributed partly to CD47 “don’t eat me” signaling in the tumor microenvironment, but no CD47 blocking, CD47-KO tumor lines, or anti-CD47 antibody co-treatment was performed. The exclusivity of tumor colonization (zero non-tumor CFU) is well-documented but its mechanistic basis remains correlational. no-mechanism
Murine immunology diverges from human in relevant ways. BALB/c mice have a Th2-biased immune background that may amplify or suppress specific arms of the E. americana response vs. typical human Th1/Th2 balance. Neutrophil biology, NK cell sensitivity, and TLR repertoire differ between mouse and human. The neutrophil-dominant response (+30% CXCR4+) in this model may not translate quantitatively to human tumors.
Narrow therapeutic window. 1 × 10⁹ CFU is both optimal and maximum tolerated; extrapolation to human-scale dosing (allometric scaling from 20 g mice to ~70 kg humans) is challenging and unpredictable. GMP-grade bacterial manufacturing at scale introduces additional consistency challenges (CFU batch variability, sterility, medium composition).
Single cancer type. All in-vivo efficacy data are from Colon-26 murine colorectal carcinoma. Applicability to other tumor types (pancreatic, breast, lung, melanoma — particularly “cold” tumors) is hypothesized based on the generality of tumor hypoxia but not demonstrated.
No human translational data of any kind. This is a first-in-class preclinical discovery paper. Phase 0 / Phase 1 dose-escalation in human solid tumors has not been initiated. needs-human-replication

Extrapolation to Humans

Dimension	Status	Notes
Pathway conserved in humans?	partial	Tumor hypoxia + facultative anaerobe colonization is a conserved principle; TLR/NLR PAMP recognition is conserved; neutrophil/T-cell/B-cell antitumor functions are shared, but quantitative contributions differ.
Phenotype conserved in humans?	unknown	No human tumor data; all in-vivo work is in syngeneic murine Colon-26. The subcutaneous model is further from human clinical reality than orthotopic or patient-derived xenograft models.
Replicated in humans?	no	No human study. needs-human-replication

Relation to Broader Bacterial Cancer Therapy Literature

This paper extends the JAIST Miyako lab’s prior work on tumor-resident bacteria as therapeutic agents [references 15–17 in paper: Goto et al. 2023 Adv Sci; Iwata et al. 2025 Nat Biomed Eng; Chintalapati et al. 2024 Biomed Pharmacother]. The novel contribution is: (1) systematic isolation from lower-vertebrate gut microbiota as a biodiscovery strategy, (2) demonstration that a naturally occurring (non-engineered) gut bacterium achieves complete tumor regression after a single IV dose, and (3) the immunological memory finding distinguishing this from conventional therapies. The lower-vertebrate microbiome biodiscovery concept is motivated by the observation that spontaneous tumor formation in wild amphibian populations occurs at remarkably low frequencies (reference 25: Torres-Dimas 2022 Cell Biology International), suggesting evolutionary selection for anti-carcinogenic host-microbiome interactions in amphibians.

Cross-References

ewingella-americana — atomic microbe page for Ewingella americana; hosts this paper’s characterization data
bacterial-cancer-therapy — class page for live bacterial cancer therapeutics
cancer — canonical cancer phenotype page; add E. americana to bacterial intervention evidence
dysbiosis — gut microbiome hallmark; lower-vertebrate microbiome biodiscovery framing
gut-microbiome-aging-shifts — aging shifts in microbiome composition; context for therapeutic potential of non-human microbiomes
immunosenescence — aging immune context; neutrophil and T-cell responsiveness to bacterial PAMPs may be impaired in aged hosts — a critical unaddressed variable
checkpoint-inhibitors — anti-PD-L1 direct comparator in this study; E. americana outperformed a representative checkpoint inhibitor
altered-intercellular-communication — SASP / cytokine network; TNF-α and IFN-γ elevation by E. americana operates via related cytokine circuits

Footnotes

iwata-2025-ewingella-americana-antitumor · local PDF confirmed via a local paper archive (path: (local PDF); download_status: completed) · in-vivo · model: female BALB/c syngeneic Colon-26; n=3–5 per group · Iwata S, Yamasita N, Asukabe K, Sakari M, Miyako E · Gut Microbes 2025; 17(1):2599562 · doi:10.1080/19490976.2025.2599562 · PMC:12710904 · OA gold (CC-BY-NC) ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³ ↩²⁴ ↩²⁵ ↩²⁶
doi:10.3390/antibiotics13060559 · Ioannou P, Baliou S, Kofteridis D · Antibiotics 2024;13:559 · review · Ewingella americana infections in humans — a narrative review; documents case series and clinical context for opportunistic pathogenicity in neonates and immunocompromised patients; cited in Iwata 2025 as evidence of low pathogenicity in immunocompetent hosts · needs-precedent-citation (not in a local paper archive) ↩
PMID 40636648 · Jafarova Ayik H, Eyupler C, Yassa G, Aksu C, Duman N · Cureus 2025 · case-report · E. americana sepsis in an immunocompromised cancer patient receiving chemotherapy; resolved with ceftriaxone; no prior E. americana case report in Turkish patients; not in a local paper archive · consistent with Iwata 2025 antibiotic-susceptibility safety framing; does not contradict efficacy findings ↩

🧬 Aging Wiki

Explorer

Discovery and characterization of antitumor gut microbiota from amphibians and reptiles: Ewingella americana as a novel therapeutic agent with dual cytotoxic and immunomodulatory properties