Antibiotic Susceptibility of Autochthonous Enterococcus Strain Biotypes Prevailing in Sheep Milk from Native Epirus Breeds Before and After Mild Thermization in View of Their Inclusion in a Complex Natural Cheese Starter Culture

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript focuses on a topic of significant practical value for traditional dairy processing: screening safe and functional lactic acid bacteria (LAB) strains from thermized antilisterial sheep milk to develop complex natural starter cultures (NSCs). By comparing the biotyping, antilisterial activity, and antibiotic resistance of Enterococcus and Streptococcus isolates before and after thermization, the study fills key gaps in understanding how thermization shapes the safety and functionality of autochthonous milk microbiota. The experimental design is generally rigorous, with comprehensive isolate characterization and clear alignment with industry needs for NSC development. However, several ambiguities and gaps in data interpretation and experimental scope exist, which need clarification to strengthen the manuscript’s conclusions.

 

 

1. Novel biotypes are identified but lack genomic validation. For example, E. faecalis 3D (isolate KFM48) is described as "strange" for fermenting L-arabinose (unusual for E. faecalis), but no whole-genome sequencing (WGS) or additional molecular tests (e.g., sodA gene sequencing) are used to confirm its taxonomic placement.
2. All functional assessments (e.g., antilisterial activity, enzymatic traits) are conducted in vitro. The manuscript does not test candidate NSC strains in actual cheese fermentation, so it cannot confirm whether strains retain their bioprotective/technological traits in a real-world context—critical for justifying their use in NSCs.
3. While phenotypic AR is thoroughly measured, the authors do not investigate the genetic basis of resistance.
4. For Streptococcus spp., you conclude they are "a priori unsafe" for NSC inclusion, but only tested enterococcal virulence genes. Will you supplement with PCR testing for Streptococcus-specific virulence markers (e.g., asa1, sua referenced in [38, 79]) to validate this safety claim, especially for S. lutetiensis with strong antilisterial activity?
5. You note E. faecium 1X isolates are fully antibiotic-susceptible and prevalent in TM. Have you tested their ability to compete with other LAB during cheese fermentation, a key trait for NSC viability?
6. E. hermanniensis KFM37 is MDR and vancomycin-resistant. Do you have data on its prevalence in Epirus sheep milk herds, and could it act as an AR gene reservoir for other LAB in NSCs?
7. For enzymatic traits (e.g., β-galactosidase in E. faecium), how do these in vitro activities correlate with actual milk acidification rates— a critical technological trait for NSCs?
8. Improve the quality of the tables and figures.

Author Response

Please refer to our responses to Reviewer 1 comments in the attached PDF file.

Sincerely

John Samelis 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript presents a detailed study of Enterococcus and Streptococcus isolates from raw and thermized sheep milk, with emphasis on biotyping, enzymatic activity, and antibiotic resistance. The work is relevant to artisanal cheese safety and the development of “natural starter” cultures. The dataset is extensive and well-documented, but the manuscript in its current form is overly long, densely written, and sometimes difficult to follow. The study would benefit from streamlining, stronger synthesis of key findings, and clearer connection to applied food safety frameworks. The practical cheese making relevance of the work , and the acidification or sensory properties of isolates are not discussed. I am concerned about the use of antilisteral in the title and discuss further below.

Major comments

The study is restricted to two “winter” raw sheep milk batches. Seasonal, lactation, and dietary effects are well known to influence milk microbiota, including the prevalence of enterococci and streptococci. Please acknowledge this limitation more explicitly in the Discussion. Ideally, broader sampling across seasons and lactation stages would be needed before generalizing which biotypes are truly dominant or safe candidates for starter cultures.

The study stops at phenotypic susceptibility (with limited prior vanA/vanB screening) yet argues for strain safety and potential inclusion in natural starters. Given EFSA’s strain-level requirements, this is problematic.

Overall the genetic analysis presented in this manuscript is limited. Species assignment relied on 16S rRNA/IGS sequencing from earlier work, while PCR was applied only for a restricted set of targets: structural enterocin genes (entA, entB, entP, bac31), vancomycin resistance genes (vanA/vanB), and selected virulence determinants in enterococci (agg, ace, espA, hyl, IS16, gelE). Importantly, no wider antibiotic resistance gene panel was tested (e.g., tet, erm, aac-aph, cat, pbp5, optrA/poxtA, cfr), nor were fluoroquinolone resistance mutations (gyrA/parC) or mobile genetic elements assessed. For streptococci, potential virulence loci were explicitly “not tested.” Given that the manuscript aims to identify safe, autochthonous strains suitable for natural starter development, this omission is significant. Current EFSA recommendations stress that strain-level safety assessment requires genotypic as well as phenotypic ARG data, preferably from whole-genome sequencing. I recommend that the authors (i) acknowledge this limitation clearly in the Discussion, and (ii) consider at least targeted ARG PCR screening, or ideally WGS analysis, for any strains proposed as safe starter candidates.

There are inconsistencies between 16S rRNA sequencing and API20 STREP results (e.g., S. equinus vs. S. pneumoniae; S. parauberis vs. S. agalactiae). These should be discussed more critically. Consider adding a recommendation for whole genome sequencing (ANI or phylogenomics) as the gold standard for resolving such discrepancies.

The discussion should emphasize practical implications: which species/biotypes show multidrug resistance and therefore cannot be considered safe.

Manuscript length and data presentation and clarity of writing

The paper is long with extensive tables. Many details (e.g., complete sugar fermentation patterns, extended enzymatic profiles, atypical biotypes) could be moved to supplementary materials. The main text should focus on the most relevant findings: which biotypes survived thermization, which carried virulence or resistance traits, and which could be considered safe candidates for natural starters.

The AR data are central but are currently buried in multiple long tables. A visual summary (heatmap, clustering, or bar charts by species/biotype) would greatly improve readability.

The sentences are often very long, making the text difficult to read. Consider simplifying sentence structure and reducing redundant parenthetical information. The paper is is heavy with acronyms (RM, TM, NSC, NSLAB, AR, Ent+) which while helpful to some do not help redability.

General comments

I am concerned about the use of the term antilisterial in the title. Its inclusion could be read as an endorsement that milk from specific sheep breeds is intrinsically inhibitory to L. monocytogenes, which the original data do not support. In the original study, only one fresh batch showed slight inhibitory activity. I therefore recommend removing antilisterial from the title and, if desired, referring to it with proper qualification in the text.

 

Please qualify “all Streptococcus spp. derived from sheep milk should be considered a priori unsafe” by mentioning Str. thermophilus.

Check spelling of thermized/thermised and adopt one or the other.

 

Comments for author File: Comments.pdf

Author Response

Please refer to our responses to Reviewer 2 comments in the attached PDF file

Thank you.

John Samelis

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The paper is significantly improved. I have major scientific and ethical concerns regarding the retention of antilisterial in the title as explained in my first review. If permitted by the EIC this risks major food safety issues. I cannot ethically approve this. I have no problem with antilisterial being qualified in the text and used. But this need to be done in a valid manner. 
you will be aware that the incidence of L monocytogenes in bulk sheep milk is typically 2 to 5%. Over the past 10 years as you know there have been major recalls of sheep cheese due to listeria. 

 

Author Response

Dear Editors

Please refer to the attached PDF file with our responses to the Reviewer 2 comments.

John Samelis

Author Response File: Author Response.pdf