Optimizing a MALDI-TOF MS database for detection of xerophilic fungi across environments

Review
General remarks

This is an excellent paper that provides a valuable contribution to fungal identification in heritage environments and can contribute significantly on the improvement of the MALDI-TOF MS database.
Materials and Methods Review:

Some minor comments for the improvement of the manuscript's readability would be the following:

1. The explanation of statistical analysis is good, but including specific statistical tests and parameters could perhaps enhance clarity for the readers.

2. The results are well-organized and provide great insights into the creation and validation of the supplementary database. However, some results parts seem to overlap with some of the discussion parts (ie. the part about how the media can affect fungal detection). Maybe merging these two parts could enhance readability!

3. In the conclusion part, it would be nice to have some more suggestions on practical applications and future research needed to further develop the MALDI-TOF MS platform.

Note:
This review refers to round 1 of peer review.

Review
To the authors: I am sorry that all formatting has gone. I hope you still can make heads and tails.

Optimizing a MALDI-TOF MS database for detection of xerophilic fungi across environments
by Christopher Campion, Victor Carp Kofoed and Anne Mette Madsen
National Research Centre for the Working Environment

The overall purpose of the study is to construct a MALDI-TOF database of spectra to make it possible to identify xerophilic fungi from environmental samples that are otherwise difficult or impossible to identify using standard methods. This is an important purpose since these fungal species are probably more widespread and harmful than previously thought.

General comments:
One major issue, though, is that all good scientific practice for identification uses three or more strains of the same species – preferrable unrelated – to conduct taxonomic, phylogenetic or chemical classification whenever possible (sometimes only one strain exists of a species). In this study one strain of each species was used, even though more strains of the same species exist. One strain can rarely represent a whole species in all its diversity. Furthermore, one cannot be sure of correct identity of a stain just because it comes from a culture collection (No guaranty, see the MTA for the IBT collection (https://dtu.bio-aware.com/page/Terms%20and%20conditions). Table 1 should state the origin/source (substrate, country) of the strains. Some strains may be the property of others (countries or researchers).

Also, I came across this new paper, which may be inspirational:
Rolland, N., Girard, V., Monnin, V., Arend, S., Perrin, G., Ballan, D., ... & Mounier, J. (2024). Identification of Food Spoilage Fungi Using MALDI-TOF MS: Spectral Database Development and Application to Species Complex. Journal of Fungi, 10(7), 456. DOI: 10.3390/jof10070456

The language of the manuscript is in places a bit incomprehensible, mainly due to too long sentences and incorrect punctuation.
It is a bit confusing that the supplementary database deals with xerophilic Aspergilli, which seems to be the main focus of the study, while the results are mostly about Cladosporium, Penicillium and other more common indoor fungi what are not known as xerophilic.

Throughout the manuscript, Aspergillus has been abbreviated As and Cladosporium as Cl. This is not the correct way to abbreviate genus names, which are always abbreviated with one letter: A. niger or C. herbarum.

The authors write that MALDI-TOF is a rapid “detection and identification” tool (line 12-13). However, fungi have to be sampled, isolated and grown in a standard broth before MALDI-TOF can be employed and then for identification only. If the sampling methods do not select for the xerophilic fungi in the first place (e.g. sampling on V8, MEA or even DG18) these fungi would never be detected.
If one doesn’t expect that xerophilic fungi is a problem, one would not use media that these fungi will grow on and never detect them – what is the advice to the building inspector?

Specific comments given at specific line numbers:

Introduction
32: use “controlled and/or prevented” instead of “limited”
63: it is “IBT” and not “ITB” – correct several placed
64: except that the strains have been described, what other selection criteria were used? are all selected species from museums or other relevant origin?

Will the supplementary database be made available to other researchers?

Materials and Methods
In general, the M&M needs more structure and perhaps a flowchart so the reader can follow the different experiments and how they are connected.

68: IBT
70-75: no need to repeat the strains, just refer to Table 1.
76-78: why was additional 100 mg/l chloramphenicol added to the DG18 medium? why were there used two different media? which strains were inoculated on which media?
86: insert reference for Sabouraud broth and what is the “either” for? Were the isolates also inoculated on other media? see 103.
86: how long did it take to get “sufficient” biomass, when 3 days were not enough?
97-100: I am not sure I understand the sentence or why.
103: insert references for the other three media (medium in singular). Why were the other media used? Did the isolates no produce biomass on Sabouraud?
124-126: what type of samples and from where? were the exposure samples inoculated particularly for this study or was the spectra from another (published?) study?
127: how can the authors validate the supplementary database with samples of unknown content?
128: When and where were these different samples collected and what was the occasion (visible fungal growth, water damage or health effects)?
133: what is a “Gesamtstaubprobenahme air samplers, GSP”?
135: what is an eSwab?
146: why were samples from home environment used? explain how these home samples relate to work environments.
150: what type of “material” was extracted?
156: replace “fungi” with “fungal” for fungal colonies, fungal concentration, fungal growth etc.
156-158: how was this done – please specify in each case.
152-158: as I understand this paragraph, the old samples were plated out and selected colonies from each plate where then isolated, extracted and analysed. Is this correctly understood?

Results
173: “As penicillium”?
175: replace “distinctions” with “discrepancies”
191: the drilling waste treatment plants are not mentioned in in M&M
204: the authors cannot validate the database using samples with unknown fungal content
208-211: how did the authors calculate these numbers? M&M does not state any method.
219-239: most results are on mesophilic fungi and not xerophilic fungi.
241: the media do not affect the presence of fungi, but different media detect/select for different fungi.
276-280: I have tried the IBT fungal collection and I cannot find any strain with the number IBT 34818 or any strain named A. destruens or A. salinarum.
296-300: the fungi identified from the museum (open air museum?) samples seem to be mostly outdoor fungi from plants and cereals.
305: sampling volume might be more important than sampling time.
308-314: how was the homes selected – randomly, from a database? and were they water damaged or did they not have any problems? Most of the fungal species listed are food-borne species.
326-333: this paragraph does not make sense.
275-276: could it be that the A. destruens is misidentified? What measures were taken to check/verify the identity of the strains used for the supplementary database?
286 and 292: the authors should not test environmental samples with unknown fungal compositions against a database with uncertain identifications. It might turn out to be correct, but as for now, the identity of the strains is not verified with more strains from e.g. different substrate or countries or different culture collections. Alternatively, strains could be grown according to Sklenar et al, (2017) and compared and verified according to their descriptions.

Discussion
266-269: it would be more relevant to evaluate the sampling protocols. If a high aW medium (e.g. V8 or MEA) is used during the initial sampling, the xerophilic fungi will never be detected.

Conclusion
The authors have included other fungal findings in their study but have not mentioned anything about the implications of these findings.
How will a study/method like this tackle dead/not viable fungal biomass, which can also have negative health implications on workers?
How does the MALDI-TOL method (a cultivation method) compared to non-cultivation methods, such like DNA using e.g. calmodulin primers, that can be used directly on an environmental sample? A DNA sequencing would be faster, detect both living and dead fungal biomass and be both qualitative and quantitative.

Figures
Increase the font size in all figures. The species names are very difficult to read.

Fig. 1: the quality of the figure is quite poor. Many of the lines are not visible and the colour scheme of the strain/isolate names are not readable.

Fig. 2: I suggest deleting the figure from the manuscript. The figure does only show the fungal concentrations from different sampling sites. However, the fungi have been collected in different ways, so A, B and C cannot be compared. Furthermore, the manuscript does not state that if the homes, the museum or the warehouse had any fungal problems and no other measurements (temperature, humidity etc) are given for the individual sampling sites.

Fig. 3: I suggest deleting the figure form manuscript. Same argument as for Fig. 2. Different sampling methods cannot be compared and there is no information on the sampling sites.

Fig. 4: this figure is interesting as it shows how important the sampling/isolation method is: xerophilic fungi can only be detected by using culture media that are non-standard and therefore would never be detected even if DG18 was used. Xerophilic fungi need media with very low aW in order to be detected and identified.
Some fungal species are labelled as xerophilic (in green) and others [not xerophilic or xerotolerant] in blue. There need to be references. I think that e.g. A. niger, A. montevidensis and A. flavus are considered xerotolerant species.

Table 1
Increase the table font (e.g. one below the main text) for easy reading.
Adapt the table layout from Sklenar et al. (2017) with species, strain no (also from other collections) and source (substrate and country). Correct species identity is not a guarantee just because they come from a culture collection. There is no need for the second column (section) since they are all from section restricti.
Either sort the strains according to species name or according to clade placement.
The web page is no longer accessible.
Are the authors sure that all the listed strains are available to the public? Sometimes strains are restricted and do not show up in the databases. Give also the identification numbers of the A. penicillioides that already is the Brukers BDAL Filamentous Fungi library.
I am not sure what the authors mean by the last # in the table. According to Index Fungorum A. destruens is not, as far as I know, been associated with P. salinarum or A. salinarum and it is neither mentioned in Greiner et al. (2014) or in Martinelli et al. (2017).
Martinelli, L., Zalar, P., Gunde-Cimerman, N., Azua-Bustos, A., Sterflinger, K., & Piñar, G. (2017). Aspergillus atacamensis and A. salisburgensis: two new halophilic species from hypersaline/arid habitats with a phialosimplex-like morphology. Extremophiles, 21, 755-773.

References
In the text use first author and et al. and year (e.g. Abdel-Maksoud et al. (2022)) instead of two authors and et al.
Use the APA style without the “quotation marks” throughout for uniformity:
Abdel-Maksoud, G., Abdel-Nasser, M., Sultan, M. H., Eid, A. M., Alotaibi, S. H., Hassan, S. E. D., & Fouda, A. (2022). Fungal biodeterioration of a historical manuscript dating back to the 14th century: an insight into various fungal strains and their enzymatic activities. Life, 12(11), 1821.

Some references are missing the journal name and others the page numbers:
414: Abdel-Maksoud
419: Bastholm
424: Borrego
426: Cappitelli
440: Honsig
446: Kujovic
454: Micheluz
458: Nastasi
467: Pitt
471: Pyzik
489: Sterflinger
500: Vacar
504: Visagie

Note:
This review refers to round 1 of peer review.