BGPT: Paper Review: Leaf litter associated fungi from South Africa to control fungal plant pathogens

Explore by Goal

Quick Explanation Copied

Key takeaway

The paper combines a large South African leaf-litter fungal collection with an OSMAC/SAHA approach and a HTS antifungal assay panel, yielding 340 active extracts and libertamide as a novel metabolite with low-µg/mL potency vs Zymoseptoria tritici under the tested conditions.

Long Explanation

Paper Review (Science Reports 2025)

Leaf litter associated fungi from South Africa to control fungal plant pathogens

Received 31 Mar 2025; accepted 30 Jun 2025.

What they claim (in-scope)

232 leaf-litter-associated fungal strains (South Africa) were phylogenetically classified (ITS + 28S) and extracted using an OSMAC library approach; 2,362 extracts were screened by HTS against four phytopathogens.
SAHA and fermentation medium altered metabolomic profiles and antifungal activity distributions, with the most consistent SAHA-driven improvement reported for B. cinerea and M. grisea.
A novel molecule, libertamide, was isolated from Libertasomyces aloeticus CF-168990; it showed IC50 ≈0.91–1.30 µg/mL against Zymoseptoria tritici with reported HepG2 ED50 ≈25.14 µg/mL, and activity was tied to production yield modulation by SAHA.

1) Visual overview: library → hits → dereplication → new compound

Notes on epistemic status: the paper reports counts, but the treemap does not imply causality; it simply visualizes reported pipeline metrics.

2) HTS hits across the 4 phytopathogens

The paper states: 113 hits for B. cinerea (4.78%), 29 for C. acutatum (1.23%), 36 for F. proliferatum (1.52%), and 41 for M. grisea (1.74%).

3) SAHA modulation (explicit example: B. cinerea hit counts)

The manuscript explicitly reports B. cinerea hits rising from 11 (non-supplemented) to 76 (with SAHA).

4) Libertamide: fermentation yield boost under SAHA

The paper reports libertamide yields (mg/L) under the tested media/SAHA conditions: 188 (MXCK), 229 (MCKX+SAHA), 303 (BRFT), 395 (BRFT+SAHA).

5) Libertamide potency vs pathogens (IC50)

The paper reports libertamide IC50 values against multiple pathogens; the strongest activity reported is vs Zymoseptoria tritici (0.91–1.30 µg/mL depending on readout), with weaker potency vs several other fungi (e.g., >60 µg/mL for some panel members).

6) Libertamide vs cytotoxicity (HepG2 ED50)

The study reports HepG2 ED50 ≈25.14 ± 2.45 µg/mL, and Z. tritici IC50 ≈0.91–1.30 µg/mL.

7) Methods and statistical logic (what matters for scientific scrutiny)

Hit selection uses readouts + stringent thresholds

The paper uses absorbance and fluorescence readouts to estimate % conidial germination inhibition, and selects hits via statistical criteria (z′ score and outlier logic) and explicit dual-readout thresholds rather than a fixed 50% inhibition rule.

Dereplication is crucial—but can hide novelty

They perform LC-HRMS chemical dereplication using internal purified/NMR-confirmed databases plus manual searches against natural products references, reporting 92 known molecules in 187 active extracts. This protects against false novelty, but it can also miss truly novel compounds if they share fragments/isotopic patterns that map onto known spectra imperfectly or if thresholds/annotation confidence are not exhaustively validated in the text.

8) Skeptical critique: strengths vs gaps/unknowns

Strengths (evidence-backed)

Scale + breadth: 232 strains; 2,362 extracts; multi-target HTS panel. This reduces the risk that activity is idiosyncratic to one pathogen assay condition.
Two readouts + explicit hit thresholds: dual-criterion design (absorbance and fluorescence) improves confidence vs single readout artefacts, assuming thresholds and z′ usage are robust.
Mechanistic direction via culture epigenetics: SAHA is used as a structured perturbation with observed, pathogen-dependent directionality (increase for B. cinerea/M. grisea; decrease for C. acutatum/F. proliferatum).
Proof-of-concept includes deconvolution: libertamide is isolated by bioassay-guided fractionation, structurally characterized by NMR/LC-HRMS, and tested in dose-response against additional strains with cytotoxicity measured in HepG2.

Key limitations / uncertainties (what could change conclusions)

Dereplication confidence vs novelty: “tentatively dereplicated” and “unknown molecules” are discussed, but the text excerpt does not show confidence thresholds, false match rates, or how many “unknown” features persist after stricter validation. This matters because SAHA can alter metabolite profiles in ways that could create spectral shifts leading to misannotation.
Activity attribution in crude extracts: for many hits, activity is inferred from co-occurrence of candidate known compounds; without purified testing for each compound-feature pairing, causality remains uncertain for those extracts.
Assay-to-field translation gap: HTS microdilution on fungal conidia is a controlled in vitro system; the paper excerpt provided does not include greenhouse/field efficacy for libertamide specifically (and it does not demonstrate formulation, persistence, non-target impacts, or pathogen life-stage relevance).
Statistical thresholding and outliers: the paper uses z′ and outlier approaches that can be sensitive to plate/layout, normalization, and how z′ is computed per assay run; without the full statistical methods (not fully shown in excerpt), there is a risk that hit lists are partially driven by distributional assumptions.

9) Evidence-based “what would disprove or change this”

If libertamide’s IC50 potency does not reproduce in independent labs/assay formats (especially when assay readouts differ), then the claimed potency/selectivity would weaken.
If SAHA does not consistently shift metabolite profiles in the same direction under replicate fermentations with the same strain and media batches, then SAHA’s role as a reliable perturbation for hit-rate improvement would be less certain.
If dereplication/database matching reanalysis with independent metabolomics pipelines identifies different candidate structures for the “tentatively dereplicated” hits, then the inferred links between specific metabolites and antifungal activity would need revision.

Disclaimer: This critique is based on the full-text excerpt you provided for the paper and only on explicitly stated details from that text; absence of detail ≠ absence of the underlying method.

Feedback:

Updated: April 24, 2026