Options
Research article

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

Authors
  • Christopher Campion orcid logo (National Research Centre for the Working Environment, Copenhagen Ø, Denmark)
  • Victor Carp Kofoed orcid logo (National Research Centre for the Working Environment, Copenhagen Ø, Denmark)
  • Jeppe Lund Nielsen orcid logo (Department of Chemistry and Bioscience, Aalborg University, Denmark)
  • Anne Mette Madsen orcid logo (National Research Centre for the Working Environment, Copenhagen Ø, Denmark)

Abstract

Xerophilic fungi can proliferate in dry conditions and have been detected in cultural heritage sites and libraries. To protect the staff from fungal exposure and ensure the preservation of heritage collections, research must be focused on improving detection protocols for xerophilic fungi. Matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry provides a strong method for the identification of fungi; however, this is dependent on the reference database. The aim of this study was to investigate whether it is possible to develop a supplementary matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry database of xerophilic/xerotolerant Aspergillus species. The database is intended to complement the current Bruker library; for this 19 Aspergillus species grown in four different broth media were included. The database was evaluated on samples from domestic homes, a museum and a warehouse. It was possible to create a database with mass spectra from the 19 species. For some species, it was possible to generate mass spectra from the four tested broth media, while other species required specific media and extended incubation time. Isolates from environmental samples identified by the Bruker fungi library were not misidentified by the supplementary database while some previously unidentified isolates (Aspergillus conicus, Aspergillus domesticus, Aspergillus glabribes and Aspergillus pseudogracilis) could be identified. Using low water-activity agar media had a profound effect on detection of these xerophilic/xerotolerant Aspergillus species. This work shows that it is possible to develop a supplementary matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry database for the identification of xerophilic/xerotolerant Aspergillus species, and that low-water activity broth media are recommended for the construction of a database and the following application.

Keywords: detection methods, cultural heritage, home exposure, indoor air, MALDI-TOF MS, museum, occupational exposure, Thermowood, xerophilic Aspergillus spp.

How to Cite:

Campion, C., Kofoed, V. C., Nielsen, J. L. & Madsen, A. M., (2025) “Optimising a MALDI-TOF MS database for the detection of xerophilic fungi across environments”, UCL Open Environment 7(1). doi: https://doi.org/10.14324/111.444/ucloe.3244

Downloads:
Download PDF
View PDF
Download XML

680 Views

Published on
29 Jul 2025
Peer Reviewed
License
Creative Commons Attribution 4.0

Introduction

Uncontrolled fungal growth can compromise the preservation of heritage collections [1–4], leading to the deterioration of cultural heritage artefacts [5–12]. Furthermore, exposure to airborne fungi may pose a health risk for museum staff and visitors [13–15].

Common strategies to prevent fungal growth in museum artefacts, exhibition rooms and storage repositories include controlling indoor climate parameters such as temperature, humidity and sunlight, complemented by various methods to eliminate fungal growth [5,6]. However, xerophilic/xerotolerant species of the genus Aspergillus [16], Cladosporium [17] and Penicillium [18] have the ability to proliferate in extreme dry and water controlled [low water-activity (aw) or low relative humidity (RH)] conditions and many reports have identified xerophilic/xerotolerant fungi in museums, libraries, cultural heritage sites and their respective repositories [19–25] – some also being reported in places despite stringent climate-controlled environments [26]. These observations suggests that xerophilic/xerotolerant fungi growing at low RH/aw, might be overlooked when attempting to limit fungal growth in such environments.

Matrix-assisted laser desorption/ionisation-time-of-flight mass spectrometry (MALDI-TOF MS) provides an efficient method for the identification of cultivable fungal isolates. MALDI-TOF MS relies on measurements of the mass–charge ratio of extracted proteins as mass spectra. For microbial identification species-specific mass spectra of unknown isolates are compared to a reference spectra database. Depending on the degree of spectral similarity, this will result in identification to the genus or species level. The accuracy and specificity of MALDI-TOF MS identification therefore depend on the extent of the reference spectra database. The current Bruker database for MALDI-TOF MS contains several xerophilic/xerotolerant Penicillium and Cladosporium species, but only a few xerophilic/xerotolerant Aspergillus species. The Bruker database was initially made to support identification of hospital-related species. However, it is possible to add spectra of other species to a local library, thus a recent study has constructed a library on food-related species [27]. To better ensure a safe indoor air and work environment for the staff and ensure the preservation of our heritage collection, further research is needed to broaden a rapid detection and identification protocols of xerophilic/xerotolerant Aspergillus fungi.

This study aimed to evaluate whether it is possible to construct a supplementary MALDI-TOF MS database for the identification of xerophilic/xerotolerant Aspergillus spp. For this, we selected 19 xerophilic/xerotolerant Aspergillus spp., originating from the Institut for Bioteknologi (IBT) fungal culture collection (Table 1) and tested different broth media for cultivation. The supplementary database was evaluated on environmental samples, and different agar media were compared for the detection of xerophilic/xerotolerant species.

Table 1.

Overview of xerophilic/xerotolerant fungi included in this study

Species* Clade IBT no.** Additional reference no. Country/source MSPs Spectra comprising MSP
Total SAB*** CY20S MY50 MY70
A. destruensa A. conicus 34818 NRRL 145T = CBS 593.91 = DTO 079-A8 = IMI 358691 = CCF 5462 USA, Maryland/maize seed 4 44 9 11 2
A. villosus A. conicus 34822 NRRL 25813T = CCF 5531 = DTO 356-C9 UK, Kirkhill/unknown 4 40 6 9 4
A. gracilis A. conicus 34817 NRRL 4962T = CBS 539. = DTO 351-H7 = CCF 5478 = ATCC 16906 = IMI 211393 USA, South Pacific/gun firing mechanism 4 21 5 4 1
A. domesticus A. conicus 34814 DTO 079-F2T = CCF 5464 = NRRL 66616 Netherlands, Tiel/wallpaper 4 28 7 4 4
A. conicus A. conicus 34288 EXF 7667 Slovenia, Ljubljana/Oil painting canvas 4 30 5 3 2
A. pseudogracilis A. conicus 34813 CCF 5505T = EMSL No. 2765 = DTO 356-F3 = NRRL 66620 USA, California/child carrier 4 23 11 9 2
A. reticulatus A. penicillioides 34819 CCF 5518 = EMSL No. 1272 = NRRL 58903 USA, Idaho/air 4 22 4 7 2
A. canadensis A. penicillioides 34642 CCF 5548 = KAS 6194 = DTO 356-H9 = IBT 34520 = NRRL 66614 Canada, Wolfville/house dust 4 30 6 13 4
A. clavatophorus A. penicillioides 34632 NRRL 25873 = CCF 5453 USA, Georgia/mouldy paper 1 32 - - -
A. hordei A. penicillioides 34631 NRRL 25830 = CCF 5485 USA, Minnesota/insulation board 4 56 7 11 4
A. infrequens A. penicillioides 34524 NRRL 25868T = CCF 5486 = DTO 356-D6 USA, Illinois/wheat 4 34 4 4 2
A. penicillioidesb A. penicillioides 34815 CCF 5498 = EMSL No. 2440 = DTO 356-E7 USA, New Jersey/green fabric binders 3 21 - 5 4
A. magnivesiculatus A. penicillioides 34816 NRRL 25866T = CCF 5488 Japan, Tokyo/fermented fish 4 52 12 14 12
A. halophilicus A. halophilicus 34881 DTO 271-F4 = CCF 5825 Netherlands/textile 2 - - 8 12
A. vitricola A. vitricola 34272 EXF-10383 = CCF 5655 Slovenia, Ljubljana/oil painting canvas 4 11 5 4 2
A. glabripes A. vitricola 34820 CCF 5474 = DTO 356-E8 = EMSL No. 2462 = NRRL 66618 Trinidad/office folder 4 18 5 5 6
A. restrictus A. restrictus 33630 CBS 118.33 UK/cotton fabric 1 6 - - -
A. caesiellus A. restrictus 34621 CCF 5448 = EMSL No. 1383 USA, Pennsylvania/air 4 14 4 1 2
A. pachycaulis A. restrictus 34812 NRRL 25824 = CCF 5492 = DTO 356-D2 = IBT 34521 = IBT 34812 USA, Washington DC/unknown 4 17 4 5 3

  • The number of MSPs created for each fungal isolate and the number of spectra used to create the MSP for each of four broth media are presented.

  • *All species belong to the Restrici section.

  • **https://dtu.bio-aware.com.

  • ***SAB broth.

  • aUpdated name – Aspergillus salinarum (ex type of Aspergillus destruens, synonymy – Phialosimplex salinarum [28]).

  • bA. penicillioides is present in Bruker BDAL Filamentous Fungi Library version 4 (MaldiBiotyperDBUpdate_V4_Fungi-856(RUO)).

Materials and methods

The study is divided into three parts. In the first part, a supplementary MALDI-TOF MS database was constructed using spectra generated from 19 xerophilic/xerotolerant Aspergillus spp. In the second part, previously stored spectra were analysed using the supplementary database to see whether previously identified or unidentified species were identified as some of the newly added species. In the third part, samples from different environments were cultivated and fungi were identified. The aim was to ensure that the supplementary database did not misidentify species which were already identified by the Bruker database, and to investigate the presence of xerophilic/xerotolerant species in these samples. Additionally, we aimed to evaluate the ability of different agar media to support the growth of xerophilic and xerotolerant fungi, from collected samples.

Xerophilic/xerotolerant Aspergillus species for the supplementary database

A total of 19 xerophilic/xerotolerant identified Aspergillus species were obtained from the IBT fungal culture collection. These isolates were originally isolated in different countries and from different environments and materials such as air, house dust, mouldy paper, wallpaper and paintings (Table 1). These species were all the Aspergillus sect. Restricti and contained representatives of the Aspergillus restrictus/Aspergillus conicus clade, the Aspergillus vitricola clade, the Aspergillus penicillioides clade and the Aspergillus halophilicus clade [16].

The fungi were cultivated on Dichloran Glycerol Agar (DG18, Thermo Fisher Scientific Oxoid, Basingstoke, UK) [29] supplemented with 100 mg/L chloramphenicol. The species A. halophilicus was grown on Malt Yeast 50% Glucose Agar (MY50G) [30] as it does not grow well on DG18. All fungi were incubated at 25 °C for 14–21 days. Of the 19 Aspergillus spp. included in this study, A. penicillioides was already present in the Bruker BDAL Filamentous Fungi Library version 4 [MaldiBiotyperDBUpdate_V4_Fungi-856(RUO)] (Bruker, Bremen, Gemany), and a strain of A. penicillioides was included in the supplementary library as a positive control.

Samples for MALDI-TOF MS

Protein extraction was performed using a modified version of the manufacture’s ethanol-formic acid extraction protocol. In brief, the fungal isolates were inoculated in 1.5 mL Eppendorf tubes containing a broth and incubated at room temperature for 1–14 days until visible growth was observed. Samples were inactivated by centrifugation, discarding the supernatant and resuspending the pellet in 70% ethanol. The ethanol was removed, and the sample was dried at room temperature, after which 5–20 μL (depending on pellet size) 70% formic acid was added. After 2 min, the same volume (5–20 μL, depending on pellet size) of acetonitrile was added. The suspension were then centrifuged and 1 μL of the supernatant was spotted onto a MALDI 96 main spectral profile (MSP) polished steel target plate (Bruker, Bremen, Germany) along with 1 μL (α-cyano-4-hydroxycinnamic acid (HCCA) matrix solution (#8255344, Merck, Darmstadt, Germany).

Spectra were acquired on a Microflex LT mass spectrometer (Bruker Daltonics Inc.), using the Bruker Biotyper software (v. 3.1) with the Filamentous Fungi Library version 4, MaldiBiotyperDBUpdate_V4_Fungi-856(RUO).

MSP creation for the supplementary MALDI-TOF MS database

Each species in the supplementary database was represented by a series of MSPs, constructed by combining raw mass spectra obtained from the species isolates inoculated in four different liquid broth media [Sabouraud (SAB) agar [31], filtered MY50 [32], Czapek yeast 20% sucrose broth (CY20) [31] or filtered Malt Yeast 70% Glucose Broth (MY70)], at room temperature until growth was visible (which was for 3–5 days for all species except for A. halophilicus, which had to be incubated for 14 days, and A. halophilicus should have a double layer of matrix in contrast to other species).

For fungal identification, the standard procedure in this laboratory was to inoculate isolates in SAB broth, therefore, the majority of the MSPs were created based on raw mass spectra obtained from isolates inoculated in SAB broth. Our previous studies and other studies show that SAB supports growth of many different Aspergillus species [31,33]. MSPs were later created based on mass spectra obtained from isolates inoculated in MY50, MY70 or CY20 broth. Protein extraction was as described above.

Prior to MSP creation, all the raw spectra were visually examined using the flexAnalysis software (Bruker Daltonics Inc.), where spectra with outlier peaks and low mass to charge ratio (m/z) were excluded. Then, using the MALDI Biotyper Compass Explorer module, raw mass spectra were ‘preprocessed’ by default settings, which includes Mass Adjustment, Smoothing, Baseline Subtraction, Normalisation and Peak picking. This was followed by creating a principal component analysis plot of the spectra, enabling further examination and removal of outlying spectra. Finally, the MSPs obtained for all species were presented as an MSP dendrogram using the MALDI Biotyper Compass Explorer module (Bruker, Bremen, Germany). This analysis enables visualisation of how close the MSPs representing the different species are related to one another, reflected by an arbitrary distance level, normalised to a maximum of 1000. In summary, for each species, up to four reference MSPs were created based on the broth medium used for inoculation (SAB, MY50, CY20 and MY70).

Search for xerophilic/xerotolerant Aspergillus species retrospectively

Previously obtained and stored mass spectra, originating from an unrelated study on drilling waste treatment plants, were used to perform a test of the supplementary database, using the Bruker Biotyper 3 software and the BDAL Filamentous Fungi Library version 4 [MaldiBiotyperDBUpdate_V4_Fungi-856(RUO)]. Exposure samples had previously been plated on DG18 and SAB agar. The number of mass spectra obtained from isolates grown on DG18 agar (nDG18) and SAB agar (nSAB) was 197 and 116, respectively. Further information on the study has been published elsewhere [34].

Evaluation of the supplementary MALDI-TOF MS database on various environmental samples

The supplementary database was evaluated on environmental samples from three different environments, utilising different sampling techniques. The environments included in this study were domestic homes (nsamples = 27), a museum (nsamples = 15) and a warehouse (nsamples = 9). Environmental samples can be collected using different methods, depending on the environment and the aim of the research. Therefore, for each environment, a different sampling method was utilised. Air samples were taken in the warehouse using Gesamtstaubprobenahme (GSP, CIS by BGI Inc.) sampling actively at 3.5 L/min. Electrostatic dust collectors (EDC) were used in homes as passive samplers. In an open-air museum environmental surface swabs (eSwab collection and transport system designed for microbiological sampling) were taken from the surfaces of artefacts.

The warehouse was used for storage of gardening and roadwork equipment, as well as tools and machinery used by workers for the maintenance of public spaces. The warehouse was selected for this study, because fungi seemed to grow on the walls, and it is important to protect the workers from exposure. The warehouse is made of heat-treated pine wood (Thermowood), and fungal growth was suggested to be caused by fluctuations in the humidity. Personal GSPs (nsamples = 6) were carried by workers when they were present in the warehouse in the morning and sampled for 11–21 min and the stationary GSPs (nsamples = 3) sampled for 240 min at specified locations in the warehouse. The GSP samplers were mounted with 37 mm polycarbonate filters (pore size of 0.8/1 μm; Merck, Damstadt Germany).

The museum is an open-air facility consisting of historical houses, buildings and indoor exhibitions representing various eras throughout history. Here, surface samples were collected using eSwab (Copan’s Liquid Amies Elution Swab, eSwab; Copan, Brescia, Italy) from artefacts in various locations. The museum was selected because the staff expressed concerns that the furniture and artefacts were colonised by fungi. Samples were collected from surfaces that appeared to be colonised by fungi, surfaces that appeared clean and surfaces covered in dust.

The homes were randomly selected and with no described water damage. The home environments utilised long-term passive sampling (11–19 days) of dust using an EDC cloth (ZEEMAN, Alphen, The Netherlands, surface exposure area: 209 m2).

Quantification and identification of fungi from environmental samples

Sampled material was extracted using an extraction solution [MilliQ water, 0.85% sodium chloride (NaCl) and 0.05% Tween80] while shaking (GSP and eSwab – 15 min at 500 rpm; EDC – 60 min at 300 rpm). All samples were stored at −80 °C in 30% glycerol until plated.

The GSP samples were plated on DG18 and MY50 agar. The eSwab samples were plated on SAB, DG18 and MY50 agar. The EDC samples were plated on DG18, MY50 and malt extract agar (MEA). In the case of agar plates being overgrown, making colony counting and isolation impossible, the samples were plated in 10 × dilutions (MilliQ water, 0.85% NaCl, 0.001% Bac. peptone, Merck KGaA, Darmstadt, Germany). All agar plates were incubated at 25°C; DG18, SAB, MEA for 1 week and MY50 for 3 weeks.

Following incubation, fungal colonies were counted. Fungal concentrations from EDC and GSP samples were calculated, taking into account the number of colonies, how much samples were diluted, sampling time and volume extracted, and for GSP samples also the flow rate.

The fungal colonies were prepared for MALDI-TOF MS as described above and identified using the Bruker database as well as the extended database. Identification of isolates were analysed as technical duplicates, with the cut-offs: identification (ID) scores <1.75 were unidentified, ID scores 1.75–1.85 were identified at genus level and ID scores >1.58 identified to species level.

Data analysis and visualisation

Data analysis was conducted using R v.4.2.3 [35] using the packages ‘tidyverse’ [36], ampvis2 [37] and ggplot2 [38] for data handling and visualisation. Data on fungal species (xerophilic/xerotolerant out of total) on different agar media were compared using Fisher’s Exact Test.

Results

Construction of a supplementary MALDI-TOF MS database

MSPs were successfully created from all four broth media for 15 out of the 19 xerophilic/xerotolerant Aspergillus species. However, for Aspergillus clavatophorus, A. restrictus, A. penicillioides and A. halophilicus, it was not possible to obtain spectra from all media (Table 1). It was only possible to obtain spectra of A. halophilicus if the fungus was inoculated in broth media with low water activity (MY50 and MY70) for 10 to 14 days. In contrast, spectra of A. restrictus, A. clavatophorus and A. penicillioides were of insufficient quality when inoculated in MY50, MY70 or CY20 (Table 1).

Discrimination between the 19 xerophilic/xerotolerant Aspergillus species of the supplementary database was evaluated by an MSP dendrogram (Fig. 1). The majority of the MSPs formed species-specific groups, with the exception of Aspergillus destruens and Aspergillus penicilloides, which were indistinguishable from each other. Generally, the MSP species groups formed clades, with similar patterns to previously published data based on genomic data, albeit with some discrepancies [16]. The Aspergillus restrictus clade, consisting of Aspergillus pachycaulis, Aspergillus caesiellus and A. restrictus were grouped together, forming a distinct cluster. The A. conicus clade (A. villosus, Aspergillus gracilis, Aspergillus domesticus, Aspergillus pseudogracilis, A. conicus, A. destruens), instead of forming one clade, formed two clades of A. villosus, A. gracilis and A. domesticus, A. conicus, A. pseudogracilis, with A. destruens being indistinguishable from A. penicillioides from the A. penicillioides clade. The two species included from the A. vitricola clade (Aspergillus glabripes, A. vitricola) were grouped with species of the A. penicillioides clade and A. conicus clade. The species included from the A. penicillioides clade (Aspergillus magnivesiculatus, Aspergillus hordei, A. penicillioides, A. clavatophorus, Aspergillus infrequens, A. canadensis), were split into two distinct clades.

Figure 1
Figure 1

AN MSP dendrogram based on MSPs of the 19 species of xerophilic/xerotolerant Aspergillus, using the MALDI Biotyper Compass Explorer module. The distance level [arbitrary unit (a.u.), normalised to a max of 1000], reflects the differences between species and how related the MSPs are to one another. Clade designation based on genomic data is indicated by clade-specific colours and marked on the right-hand side of the image [16].

When queried against the Bruker fungi database, only A. penicillioides, which already was present in library, was correctly identified. The remaining MSPs were not identified, as expected. The MSPs of A. destruens were indistinguishable from A. penicillioides and would be identified as A. penicillioides by the Bruker fungi library. The MSPs were kept in the supplementary database regardless.

Search for xerophilic/xerotolerant Aspergillus species retrospectively in stored mass spectra

The supplementary database was used to identify mass spectra previously obtained from a study on drilling waste treatment plants [34]. Xerophilic/xerotolerant fungi had been reported in this environment, therefore, a total of 313 mass spectra originating from isolates grown on DG18 and/or SAB agar plates (nDG18 = 197; nSAB = 116), respectively, were attempted to be re-identified, using the supplementary database. Of these, two previously unidentified isolates (originating from the same DG18 agar plate) were now identified as A. caesiellus, with ID scores >2.00 (Fig. A1).

The two agar media used in this study appeared to affect the total number of xerophilic/xerotolerant species detected (Fig. A1). We observed that the samples inoculated on DG18 agar plates, contained a higher fraction of xerophilic/xerotolerant species (n = 7/21) of the genera Penicillium, Cladosporium and Aspergillus when compared to the samples detected on SAB agar plates (n = 3/14) – although the differences were not significant (p = 0.70).

Xerophilic Aspergillus species in environmental samples

The supplementary database was further evaluated using environmental samples collected from domestic homes (n = 27, via EDC), a museum (n = 15, via eSwab) and a warehouse (n = 9, via GSP) (Fig. 2). Across all environments, isolates identified to species level by the Bruker fungi library were not misidentified by the new supplementary database.

Figure 2
Figure 2

Heatmap of fungal species identified (y-axis) across three environments (x-axis). Numbers represent the number of isolates found. The colour gradient represents the log10-transformed geometric mean (GM) concentration (CFU/m2/day, CFU/location and CFU/m3). Penicillium CCC is an abbreviation for P. camemberti, P. commune or P. cyclopium, which cannot be distinguished by MALDI-TOF MS. Species marked with * are identified by the supplementary database.

The fungal concentration of the air exposure samplers (GSP) was between 268 and 1.15 × 104 colony forming units (CFU)/m3. The fungal concentration of the sedimented dust (EDC) was between 205 and 4.45 × 104 CFU/m2/day. The fungal surface concentrations (eSwab) were between 50 and 9.2 × 106 CFU/location range (Fig. A2).

Xerophilic Aspergillus species, not present in the Bruker library, were detected in the home and the warehouse samples when using the supplementary database. The xerophilic Aspergillus species identified using the supplementary database were A. pseudogracilis (nine isolates), A. glabripes (one isolate), A. domesticus (three isolates) and A. conicus (one isolate). Aspergillus domesticus and A. pseudogracilis were identified in two of the environments tested (Figs 3–5). Other xerophilic/xerotolerant fungi were identified from these environments using the Bruker library, including A. glaucus (six isolates), Penicillium brevicompactum/olsonii (64 isolates), Penicillium chrysogenum (15 isolates), Penicillium corylophilium (14 isolates) and Cladosporium herbarum (27 isolates) identified in several samples (Figs 3–5).

Figure 3
Figure 3

Fungal species in homes (EDC samples) on DG18, MEA and MY50 agar. Numbers represent the number of isolates in samples. The concentration is represented as GM of samples. Xerophilic/xerotolerant fungi from the supplementary database are highlighted in red, xerophilic/xerotolerant species present in the Bruker fungi library in green and other fungal species in blue. Penicillium CCC is an abbreviation for P. camemberti, P. commune or P. cyclopium, which cannot be distinguished by MALDI-TOF MS.

Figure 4
Figure 4

Fungal species in surface eSwab from a museum on DG18, MY50 and SAB agar. Numbers represent the no. of isolates in samples. The concentration is represented as GM of samples. Xerophilic/xerotolerant species present in the Bruker fungi library are highlighted in green, and other fungal species in blue. Penicillium CCC is an abbreviation for P. camemberti, P. commune or P. cyclopium, which cannot be distinguished by MALDI-TOF MS.

Figure 5
Figure 5

Fungal species in personal GSP samples from a warehouse on DG18 and MY50 agar media. Numbers represent the no. of isolates in samples. The concentration is represented as GM of samples. Xerophilic/xerotolerant fungi from the supplementary database are highlighted in red, xerophilic/xerotolerant species present in the Bruker fungi library in green, and other fungal species in blue. Penicillium CCC is an abbreviation for P. camemberti, P. commune or P. cyclopium, which cannot be distinguished by MALDI-TOF MS.

The home environments were dominated by the genus Cladosporium, with Cladosporium cladosporioides and C. herbarum being present in high concentrations of 7.0 × 103 CFU/m2/day and 4.4 × 103 CFU/m2/day, respectively. Species of Penicillium were also present in high concentrations. The three xerophilic Aspergillus species identified from these environments, A. domesticus, A. glabripes and A. pseudogracilis, using the supplementary database were reported in concentrations of 520 CFU/day, 410 CFU/day and 210 CFU/day, respectively (Fig. 3).

The most abundant fungi detected in the museum were Penicillium aurantiogriseum (6.5 × 104 CFU/location), P. chrysogenum (1.1 × 104 CFU/location) and isolates belonging to the Cladosporium genus (2.2 × 104 CFU/location). No additional xerophilic Aspergillus species were identified from the museum samples using the supplementary database (Fig. 4).

The warehouse samples were dominated by Penicillium camemberti commune cyclopium (Penicillium CCC) (5.5 × 103 CFU/m3), Cladosporium langeronii (2.2 × 103 CFU/m3) and the yeast Filibasidium magnum (2.8 × 103 CFU/m3). In this environment, A. pseudogracilis, A. domesticus and A. conicus were found in concentrations of 270 CFU/m3, 39 CFU/m3 and 100 CFU/m3, respectively (Fig. 5).

Effect of media on detection of xerophilic Aspergillus species in environmental samples

The various agar media used for fungal growth tended to affect the detection of xerophilic fungi (Figs 3–5). In the home environment, a larger fraction of the detected fungi tended to be xerophilic species when using MY50 (10/21 = 48% of species) and DG18 (8/29 = 27% of species), contra MEA (5/26 = 19% of species); (for MY50 vs MEA, p = 0.059). None of the xerophilic Aspergillus species included in the supplementary database were detected on DG18 or MEA. A similar trend was observed in the warehouse environment with 7/10 (70%) of species on MY50 being xerophilic species compared to 1/5 (20%) of species on DG18 (p = 0.12). This trend, however, was not observed in the museum environment (MY50, 6/22 = 27% of species; DG18, 5/25 = 20% of species; SAB, 7/25 = 28% of species) (p = 0.77).

Discussion

In this study, a MALDI-TOF MS supplementary database consisting of 19 xerophilic Aspergillus species was constructed for the purpose of complementing the current Bruker fungi library for identification of these species in environmental samples.

Construction of a supplementary MALDI-TOF MS database of selected xerophilic Aspergillus species

Culture condition and sample preparation methods can influence the mass spectrum obtained [39,40]. Therefore, the database was constructed using MSPs based on raw mass spectra originating from isolates inoculated in four different broth media. Raw mass spectra of the highly xerophilic fungus, A. halophilicus, were only obtained when the fungus was inoculated in broth media of low water activity for at least 14 days. The inability to obtain raw mass spectra from A. halophilicus might be partly explained by it being obligately xerophilic and not readily cultured in high water-activity media [41–44]. This suggests that, when working with environments where extreme/obligate xerophilic fungi thrive, it is relevant to re-evaluate the protocols (i.e., inoculation media and incubation time) used to inoculate unknown fungal isolates, to obtain mass spectra and hence ensure species identification.

Importantly, none of the MSPs of the supplementary database matched entries in the Bruker fungi library except for the species A. destruens identified as A. penicillioides. The MSP dendrogram showed that the MSPs formed species-specific clusters of the various xerophilic Aspergillus species. An exception of the species-specific clusters was A. penicillioides and A. destruens, for which MSPs were indistinguishable. By the time of submission, we observed that the species designation was updated for A. destruens (strain IBT no. 34818), which is now designated Aspergillus salinarum. This species is a halophilic fungus isolated from a hypersaline environment (Table 1). However, this does not explain why MSPs of these two species are indistinguishable.

Regarding the overall clade structure of the MSP dendrogram, some disagreements in the phylogeny between proteomic-based data and published phylogenetic data based on genomic data were observed [16]. It should be noted that the supplementary database was based on a single strain per fungal species, which is insufficient for taxonomic characterisation of fungal species. Future studies, improving the supplementary database, by including biological replicates, will be relevant to provide better characterisation and clarify discrepancies.

Testing the supplementary MALDI-TOF MS database

The supplementary database was first tested on previously acquired spectra from a study on drilling waste treatment facilities. Two unidentified isolates were now identified as A. caesiellus, demonstrating how already performed studies can easily be revised with supplementary fungal databases. Isolates identified to species level by the Bruker fungi library were not identified by the supplementary database, and thus it did not lead to misidentification.

Testing the supplementary MALDI-TOF MS database on environmental samples

The supplementary database was used to complement the Bruker database from three different environments. Isolates of xerophilic Aspergillus species were identified in two of these environments (homes and warehouse).

The museum samples did not reveal any xerophilic Aspergillus species using the supplementary database, despite other xerophilic/xerotolerant species being identified by the Bruker database from this environment. It should be mentioned that the museum environment from which these samples originate was not a closed, confined or climate-controlled site, and may not have favoured the growth of the specific xerophilic Aspergillus species included in this study.

The warehouse was selected for this study because the walls were described as being dry and yet had fungal growth on the walls, and therefore the presence of airborne xerophilic fungi could be expected to be present. In line with this, the supplementary database detected two species, A. glabripes and A. pseudogracilis. Other xerophilic/xerotolerant were also identified from this environment by the Bruker database, including A. glaucus, Aspergillus pseudoglaucus, P. brevicompactum and Penicillium corylophilum. The low number of fungal species detected in this environment could be explained by the relatively short sampling time for the GSP samplers [11–21 min for the personal samplers (n = 6) and 240 min for the stationary samplers (n = 3)], resulting in a small sampled volume. The sampling for personal samplers covered the period the workers had tasks in the warehouse in the morning.

We analysed samples from 27 randomly selected homes. The xerophilic Aspergillus species identified by the supplementary database (A. domesticus, A. glabripes and A. pseudogracilis) constituted only a small part of the potential exposure to fungi in the homes. This is in accordance with what could be expected, as fungi in home environments may have several sources and not only sources supporting growth of xerophilic fungi. The relatively low concentration of xerophilic Aspergillus species, on the one hand, suggests that these species do not pose a problem in these homes. On the other hand, their presence, combined with the fact that methods selecting for these xerophilic fungi have not been widely used in previous studies in homes, highlights the need for further studies on their prevalence in homes. Previous studies in homes also using EDCs for sampling have typically used DG18 agar [45] and SAB agar [46]. The most common species observed are in accordance with what has previously been found in homes, such as certain Cladosporium (Cladosporium sphaerospermum, C. herbarum, Cladosporium cladosporides), Penicillium (P. brevicompactum, P. camemberti, P. chrysogenum, Penicillium citrinum, Penicillium commume, Penicillium glabrum, P. olsonii) and Aspergillus (A. glaucus) species [17,47,48].

Effect of media on isolation and identification of xerophilic fungi

The type of agar media tended to have an effect on detection of xerophilic/xerotolerant species. With the exception of A. caesiellus (Fig. A1), all xerophilic Aspergillus species from the supplementary database were only detected on MY50. Some xerophilic/tolerant species, such as P. brevicompactum/olsonii, P. chrysogenum, C. cladosporioides, C. herbarum, A. pseudoglaucus and A. glaucus, were detected regardless of the media used. It should be noted that the MY50 plates were incubated longer than DG18, MEA and SAB agar plates as it takes more time for fungi to grow on this water-restricted medium. One explanation as to why the xerophilic Aspergillus species in the supplementary database were not detected on DG18 (with exception of A. caesiellus, Fig. A1), and that fewer xerophilic/xerotolerant fungi were observed on other media, might be that they were outcompeted by faster growing species on less restrictive media.

As described previously in this discussion, the broth medium also has an impact on the creation of MSPs for the new database, and raw mass spectra of A. halophilicus could only be obtained if it was inoculated in MY50 or MY70 while A. clavatophorus only formed raw mass spectra on SAB agar. These results, along with other studies [49], underline the importance of appropriate media (i.e., low aw media) and incubation time, both for constructing and for using the database.

Implications and future directions

Xerophilic/xerotolerant Aspergillus species were detected across multiple environments and sample types though they constituted only a small fraction of all fungi. For future studies, we recommend incorporating low activity media (e.g., MY50) into both initial plating and broth cultivation protocols to obtain more knowledge on the presence of xerophilic/xerotolerant fungi in indoor environments. While this may not be necessary for routine surveys, it could be warranted in cases where environmental conditions, such as dry, climate-controlled buildings, suggest a potential niche for xerophilic species. The data from the warehouse, along with prior reports from museum settings [26,50], indicates that xerophilic fungi may persist even under stringent environmental controls. As such, targeted methods may be crucial for accurate assessment of fungal exposure in vulnerable settings.

Conclusion

It was possible to construct a supplementary MALDI-TOF MS database successfully which complemented the Bruker database, with the combined databases identifying additional xerophilic/xerotolerant Aspergillus species in different environmental samples, previously going undetected, and therefore provided a more in-depth characterisation of various environments. The supplementary database has the strength that spectra are obtained from fungi grown in different broth media.

For future constructions of MSPs of xerophilic/xerotolerant Aspergillus species and the subsequent use of the MSPs for species identification we recommend using broth and agar media agar with low water activity and using an extended incubation time of 3 to 5 days in broth media. For A. halophilicus an even longer incubation time and double the matrix are needed. Future studies expanding the MALDI-TOF MS database by including more strains of each species may improve the quality of the database and clear observed discrepancies.

Acknowledgement

Jakob Blæsbjerg Hoof DTU BIOENGINEERING is acknowledged for his contribution to the study.

Authorship contribution

CC: conceptualisation, writing – draft, review and editing, analysis, investigation; AMM: writing – draft, conceptualisation, project administration, resources, supervision, writing – review and editing; VCK: writing – draft, review and editing; JLN: review and editing.

Open data and materials availability statement

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations and conflicts of interest

Research ethics statement

Not applicable to this article.

Consent for publication statement

The authors declare that research participants’ informed consent to publication of findings – including photos, videos and any personal or identifiable information – was secured prior to publication.

Conflicts of interest statement

The authors declare no conflicts of interest with this work.

References

[1]  Sterflinger, K; Pinar, G. (2013).  Microbial deterioration of cultural heritage and works of art – tilting at windmills?.  Appl Microbiol Biotechnol [online] 97 (22) : 9637–9646, Available from:. DOI: http://dx.doi.org/10.1007/s00253-013-5283-1

[2]  Sterflinger, K. (2010).  Fungi: their role in deterioration of cultural heritage.  Fungal Biol Rev [online] 24 (1–2) : 47–55, Available from:. DOI: http://dx.doi.org/10.1016/j.fbr.2010.03.003

[3]  Romero, SM; Giudicessi, SL; Vitale, RG. (2021).  Is the fungus Aspergillus a threat to cultural heritage?.  J Cult Herit [online] 51 : 107–124, Available from:. DOI: http://dx.doi.org/10.1016/j.culher.2021.08.002

[4]  Stratigaki, M; Armirotti, A; Ottonello, G; Manente, S; Traviglia, A. (2024).  Fungal and bacterial species richness in biodeteriorated seventeenth century Venetian manuscripts.  Sci Rep [online] 14 (1) : 7003. Available from:. DOI: http://dx.doi.org/10.1038/s41598-024-57228-2

[5]  Pyzik, A; Ciuchcinski, K; Dziurzynski, M; Dziewit, L. (2021).  The bad and the good-microorganisms in cultural heritage environments-an update on biodeterioration and biotreatment approaches.  Materials (Basel) [online] 14 (1) 177 Available from:. DOI: http://dx.doi.org/10.3390/ma14010177

[6]  Cappitelli, F; Cattò, C; Villa, F. (2020).  The control of cultural heritage microbial deterioration.  Microorganisms [online] 8 (10) 1542 Available from:. DOI: http://dx.doi.org/10.3390/microorganisms8101542

[7]  Vacar, CL; Mircea, C; Parvu, M; Podar, D. (2022).  Diversity and metabolic activity of fungi causing biodeterioration of canvas paintings.  J Fungi (Basel) [online] 8 (6) 589 Available from:. DOI: http://dx.doi.org/10.3390/jof8060589

[8]  Abdel-Maksoud, G; Abdel-Nasser, M; Sultan, MH; Eid, AM; Alotaibi, SH; Hassan, SE. (2022).  Fungal biodeterioration of a historical manuscript dating back to the 14th century: an insight into various fungal strains and their enzymatic activities.  Life (Basel) [online] 12 (11) 1821 Available from:. DOI: http://dx.doi.org/10.3390/life12111821

[9]  Savković, Ž; Stupar, M; Unković, N; Ivanović, Ž; Blagojević, J; Vukojević, J. (2019).  In vitro biodegradation potential of airborne Aspergilli and Penicillia.  Naturwissenschaften [online] 106 (3–4) : 8. Available from:. DOI: http://dx.doi.org/10.1007/s00114-019-1603-3

[10]  Sterflinger, K; Voitl, C; Lopandic, K; Pinar, G; Tafer, H. (2018).  Big sound and extreme fungi–xerophilic, halotolerant aspergilli and penicillia with low optimal temperature as invaders of historic pipe organs.  Life (Basel) [online] 8 (2) 22 Available from:. DOI: http://dx.doi.org/10.3390/life8020022

[11]  Borrego, S; Guiamet, PS; Vivar, I; Battistoni, PA. (2018).  Fungi involved in biodeterioration of documents in paper and effect on substrate.  Acta Microscopica 27 (1) : 37–44.

[12]  Kujović, A; Gostinčar, C; Kavkler, K; Govedić, N; Gunde-Cimerman, N; Zalar, P. (2024).  Degradation potential of xerophilic and xerotolerant fungi contaminating historic canvas paintings.  J Fungi (Basel) [online] 10 (1) 76 Available from:. DOI: http://dx.doi.org/10.3390/jof10010076

[13]  Wiszniewska, M; Walusiak-Skorupa, J; Pannenko, I; Draniak, M; Palczynski, C. (2009).  Occupational exposure and sensitization to fungi among museum workers.  Occup Med (Lond) [online] 59 (4) : 237–242, Available from:. DOI: http://dx.doi.org/10.1093/occmed/kqp043

[14]  Górny, RL; Harkawy, AS; Ławniczek-Wałczyk, A; Karbowska-Berent, J; Wlazło, A; Niesler, A. (2016).  Exposure to culturable and total microbiota in cultural heritage conservation laboratories.  Int J Occup Med Environ Health [online] 29 (2) : 255–275, Available from:. DOI: http://dx.doi.org/10.13075/ijomeh.1896.00630

[15]  Wiszniewska, M; Swierczynska-Machura, D; Cezary, P; Walusiak-Skorupa, J. (2010).  [Fungal allergy among art conservators: prevalence, risk factors and clinical symptoms].  Med Pr 61 (2) : 133–141.

[16]  Sklenář, F; Jurjević, Ž; Zalar, P; Frisvad, JC; Visagie, CM; Kolařík, M. (2017).  Phylogeny of xerophilic aspergilli (subgenus Aspergillus) and taxonomic revision of section Restricti.  Stud Mycol [online] 88 : 161–236, Available from:. DOI: http://dx.doi.org/10.1016/j.simyco.2017.09.002

[17]  Segers, FJ; Meijer, M; Houbraken, J; Samson, RA; Wösten, HA; Dijksterhuis, J. (2015).  Xerotolerant Cladosporium sphaerospermum are predominant on indoor surfaces compared to other Cladosporium species.  PLoS One [online] 10 (12) e0145415 Available from:. DOI: http://dx.doi.org/10.1371/journal.pone.0145415

[18]  Lutsenko, V; Polishchuk, L; Hong, J; Fomina, M. (2019).  Xerotolerant strain of Penicillium chrysogenum MF18_10 isolated from the damaged walls of Saint Sophia’s Cathedral, Kyiv.  Faktori eksperimental’noi evolucii organizmiv [online] 25 : 14–19, Available from:. DOI: http://dx.doi.org/10.7124/FEEO.v25.1132

[19]  Polo, A; Cappitelli, F; Villa, F; Pinzari, F. (2017).  Biological invasion in the indoor environment: the spread of Eurotium halophilicum on library materials.  Int Biodeterior Biodegrad [online] 118 : 34–44, Available from:. DOI: http://dx.doi.org/10.1016/j.ibiod.2016.12.010

[20]  Bastholm, C; Richter, J; Aalling, A; Bjerre, A; Madsen, A. (2023). Xerophilic fungi have a taste for Danish art and cultural heritage In:  2nd International Conference on Moisture in Buildings (ICMB23). 3 July 2023, p. 2 Available from:. DOI: http://dx.doi.org/10.14293/ICMB230048

[21]  Montanari, M; Melloni, V; Pinzari, F; Innocenti, G. (2012).  Fungal biodeterioration of historical library materials stored in Compactus movable shelves.  Int Biodeterior Biodegrad [online] 75 : 83–88, Available from:. DOI: http://dx.doi.org/10.1016/j.ibiod.2012.03.011

[22]  Micheluz, A; Manente, S; Tigini, V; Prigione, V; Pinzari, F; Ravagnan, G. (2015).  The extreme environment of a library: xerophilic fungi inhabiting indoor niches.  Int Biodeterior Biodegrad [online] 99 : 1–7, Available from:. DOI: http://dx.doi.org/10.1016/j.ibiod.2014.12.012

[23]  Piñar, G; Dalnodar, D; Voitl, C; Reschreiter, H; Sterflinger, K. (2016).  Biodeterioration risk threatens the 3100 year old staircase of hallstatt (Austria): possible involvement of halophilic microorganisms.  PLoS One [online] 11 (2) e0148279 Available from:. DOI: http://dx.doi.org/10.1371/journal.pone.0148279

[24]  Piñar, G; Poyntner, C; Lopandic, K; Tafer, H; Sterflinger, K. (2020).  Rapid diagnosis of biological colonization in cultural artefacts using the MinION nanopore sequencing technology.  Int Biodeterior Biodegrad [online] 148 104908 Available from:. DOI: http://dx.doi.org/10.1016/j.ibiod.2020.104908

[25]  Liu, Z; Zhang, Y; Zhang, F; Hu, C; Liu, G; Pan, J. (2018).  Microbial community analyses of the deteriorated storeroom objects in the Tianjin museum using culture-independent and culture-dependent approaches.  Front Microbiol [online] 9 802 Available from:. DOI: http://dx.doi.org/10.3389/fmicb.2018.00802

[26]  Bastholm, CJ; Madsen, AM; Andersen, B; Frisvad, JC; Richter, J. (2022).  The mysterious mould outbreak – a comprehensive fungal colonisation in a climate-controlled museum repository challenges the environmental guidelines for heritage collections.  J Cult Herit [online] 55 : 78–87, Available from:. DOI: http://dx.doi.org/10.1016/j.culher.2022.02.009

[27]  Rolland, N; Girard, V; Monnin, V; Arend, S; Perrin, G; Ballan, D. (2024).  Identification of food spoilage fungi using MALDI-TOF MS: spectral database development and application to species complex.  J Fungi [online] 10 (7) 456 Available from:. DOI: http://dx.doi.org/10.3390/jof10070456

[28]  Greiner, K; Peršoh, D; Weig, A; Rambold, G. (2014).  Phialosimplex salinarum, a new species of Eurotiomycetes from a hypersaline habitat.  IMA Fungus [online] 5 (2) : 161–172, Available from:. DOI: http://dx.doi.org/10.5598/imafungus.2014.05.02.01

[29]  Hocking, AD; Pitt, JI. (1980).  Dichloran–glycerol medium for enumeration of xerophilic fungi from low-moisture foods.  Appl Environ Microbiol [online] 39 (3) : 488–492, Available from:. DOI: http://dx.doi.org/10.1128/aem.39.3.488-492.1980

[30]  Pitt, JI; Hocking, AD. (2009). Introduction In:  Pitt, JI, Hocking, AD AD (eds.),   Fungi and food spoilage. Boston, MA: Springer US, pp. 1–2.

[31]  Guinea, J; Peláez, T; Alcalá, L; Bouza, E. (2005).  Evaluation of Czapeck agar and Sabouraud dextrose agar for the culture of airborne Aspergillus conidia.  Diagn Microbiol Infect Dis [online] 53 (4) : 333–334, Available from:. DOI: http://dx.doi.org/10.1016/j.diagmicrobio.2005.07.002

[32]  Skaar, I; Stenwig, H. (1996).  Malt-yeast extract-sucrose agar, a suitable medium for enumeration and isolation of fungi from silage.  Appl Environ Microbiol [online] 62 (10) : 3614–3619, Available from:. DOI: http://dx.doi.org/10.1128/aem.62.10.3614-3619.1996

[33]  Madsen, AM; Rasmussen, PU; Frederiksen, MW. (2024).  Fungal and bacterial species on biowaste workers’ hands and inhalation zone, and potential airway deposition.  Waste Manag [online] 183 : 290–301, Available from:. DOI: http://dx.doi.org/10.1016/j.wasman.2024.05.018

[34]  Daae, HL; Graff, P; Foss, OAH; Kofoed, VC; Afanou, AK; Frederiksen, MW. (2024).  A cross-sectional study on occupational exposure to microorganisms, endotoxin, hydrogen sulfide, and dust during work at drilling waste treatment plants.  Ann Work Expo Health [online] 68 (1) : 58–77, Available from:. DOI: http://dx.doi.org/10.1093/annweh/wxad069

[35]  R Core Team R. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing.

[36]  Wickham, H; Averick, M; Bryan, J; Chang, W; McGowan, LDA; François, R. (2019).  Welcome to the Tidyverse.  J Open Source Softw [online] 4 (43) 1686 Available from:. DOI: http://dx.doi.org/10.21105/joss.01686

[37]  Andersen, KS; Kirkegaard, RH; Karst, SM; Albertsen, M. (2018).  ampvis2: an R package to analyse and visualise 16S rRNA amplicon data.  BioRxiv [online], 299537 Available from:. DOI: http://dx.doi.org/10.1101/299537

[38]  Wickham, H. (2016). Programming with ggplot2 In:  ggplot2. Use R! [online]. Cham: Springer, Available from:. DOI: http://dx.doi.org/10.1007/978-3-319-24277-4_12

[39]  Honsig, C; Selitsch, B; Hollenstein, M; Vossen, MG; Spettel, K; Willinger, B. (2022).  Identification of filamentous fungi by MALDI-TOF mass spectrometry: evaluation of three different sample preparation methods and validation of an in-house species cutoff.  J Fungi (Basel) [online] 8 (4) 383 Available from:. DOI: http://dx.doi.org/10.3390/jof8040383

[40]  Topić Popović, N; Kazazić, SP; Bojanić, K; Strunjak-Perović, I; Čož-Rakovac, R. (2023).  Sample preparation and culture condition effects on MALDI-TOF MS identification of bacteria: a review.  Mass Spectrom Rev [online] 42 (5) : 1589–1603, Available from:. DOI: http://dx.doi.org/10.1002/mas.21739

[41]  Micheluz, A; Pinzari, F; Rivera-Valentín, EG; Manente, S; Hallsworth, JE. (2022).  Biophysical manipulation of the extracellular environment by Eurotium halophilicum.  Pathogens [online] 11 (12) 1462 Available from:. DOI: http://dx.doi.org/10.3390/pathogens11121462

[42]  Hocking, AD; Pitt, JI. (1988).  Two new species of xerophilic fungi and a further record of Eurotium halophilicum.  Mycologia [online] 80 (1) : 82–88, Available from:. DOI: http://dx.doi.org/10.1080/00275514.1988.12025502

[43]  Andrews, S; Pitt, JI. (1987).  Further studies on the water relations of xerophilic fungi, including some halophiles.  J Gen Microbiol [online] 133 (2) : 233–238, Available from:. DOI: http://dx.doi.org/10.1099/00221287-133-2-233

[44]  Stevenson, A; Hamill, PG; Medina, Á; Kminek, G; Rummel, JD; Dijksterhuis, J. (2017).  Glycerol enhances fungal germination at the water-activity limit for life.  Environ Microbiol [online] 19 (3) : 947–967, Available from:. DOI: http://dx.doi.org/10.1111/1462-2920.13530

[45]  Frankel, M; Timm, M; Hansen, EW; Madsen, AM. (2012).  Comparison of sampling methods for the assessment of indoor microbial exposure.  Indoor Air [online] 22 : 405–414, Available from:. DOI: http://dx.doi.org/10.1111/j.1600-0668.2012.00770.x

[46]  Normand, AC; Ranque, S; Cassagne, C; Gaudart, J; Sallah, K; Charpin, DA. (2016).  Comparison of air impaction and electrostatic dust collector sampling methods to assess airborne fungal contamination in public buildings.  Ann Occup Hyg [online] 60 (2) : 161–175, Available from:. DOI: http://dx.doi.org/10.1093/annhyg/mev075

[47]  Nastasi, N; Haines, SR; Xu, L; da Silva, H; Divjan, A; Barnes, MA. (2020).  Morphology and quantification of fungal growth in residential dust and carpets.  Build Environ [online] 174 (10) 106774 Available from:. DOI: http://dx.doi.org/10.1016/j.buildenv.2020.106774

[48]  Knudsen, SM; Gunnarsen, L; Madsen, AM. (2017).  Airborne fungal species associated with mouldy and non-mouldy buildings – effects of air change rates, humidity, and air velocity.  Build Environ [online] 122 : 161–170, Available from:. DOI: http://dx.doi.org/10.1016/j.buildenv.2017.06.017

[49]  Visagie, CM; Boekhout, T; Theelen, B; Dijksterhuis, J; Yilmaz, N; Seifert, KA. (2023).  Da Vinci’s yeast: Blastobotrys davincii f.a., sp. nov.  Yeast [online] 40 (1) : 7–31, Available from:. DOI: http://dx.doi.org/10.1002/yea.3816

[50]  Paiva de Carvalho, H; Sequeira, SO; Pinho, D; Trovão, J; da Costa, RMF; Egas, C. (2019).  Combining an innovative non-invasive sampling method and high-throughput sequencing to characterize fungal communities on a canvas painting.  Int Biodeterior Biodegrad [online] 145 104816 Available from:. DOI: http://dx.doi.org/10.1016/j.ibiod.2019.104816

Appendix

Figure A1
Figure A1

Retrospective search in previous obtained mass spectra for xerophilic/xerotolerant fungi. The mass spectra are from fungi in drilling waste plants. Fungi from the supplementary database are highlighted in red, xerophilic/xerotolerant species present in the Bruker library in green, non-xerophilic/xerotolerant fungi in blue. (a) Total species identified. (b) Species identified based on agar media (nDG18 = 197; nSAB = 116). Penicillium CCC is an abbreviation for P. camemberti, P. commune or P. cyclopium, which cannot be distinguished by the MALDI-TOF MS library.

Figure A2
Figure A2

Concentrations of fungi across three environments, each using different sampling techniques: (a) EDC, homes; (b) surface eSwab, museum; C GSP air samplers, warehouse. The x-axis shows log10-transformed concentrations (CFU/m2/day, CFU/location and CFU/m3). The black dots represent GM values of the sample, while the smaller dots represent outliers.

 Open peer review from Spyros Efthymiopoulos

Review
This study makes a valuable contribution to the field of mycology and environmental health. The enhancement of the detection capabilities of MALDI-TOF MS for xerophilic fungi produces insight to surpass current limitations of the identification methods. Future work focusing on expanding the database, improving data accessibility, and providing detailed performance metrics will further solidify its role as a vital tool in fungal detection.

Note:
This review refers to round 2 of peer review.

 Open peer review from Birgitte Andersen

Review
Second review

Dear authors, including Jeppe Lund Nielsen as new author.

The manuscript has improved a lot, however, there are a couple of things that I think need to be addressed/corrected/improved.

General comments

I still think that the authors should make their data available for other researchers as suggested in the Data Availability Statement of the journal. The point of sharing data is that other researchers do not have to start from scratch in every project but can stand on the shoulders of others.

Furthermore, I also still think that the authors – in the conclusion – should write some think like “MALDI-TOF requires living/viable fungal cultures for identification, however, it should be noted that both living and dead fungal fragments may cause negative health effects in the work environment.”

--------

Specific comments

Table 1. It would help the reader if the species were in alphabetic order instead of alphabetical according to clade. Also, the is one * for species and one * for SAB. Find another symbol for SAB.

Figure 1 is still too small and fuzzy with missing lines. The labels are very difficult to read, though the colour scheme helps.

Line 200: A. penicillioides instead of A. penicillium

Authorship contributions: Jeppe Lund Nielsen’s contributions need to be stated.

Note:
This review refers to round 2 of peer review.

 Open peer review from Spyros Efthymiopoulos

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 and may pertain to an earlier version of the document.

 Open peer review from Birgitte Andersen

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 and may pertain to an earlier version of the document.