Unveiling the Occupational Exposure to Microbial Contamination in Conservation–Restoration Settings

Assuring a proper environment for the fulfillment of professional activities is one of the Sustainable Development Goals and is contemplated in the One Health approach assumed by the World Health Organization. This particular study is applied to an often neglected sector of our society—the conservators/restorers—despite the many health issues reported by these professionals. Three different specialties (textiles, paintings and wood sculpture) and locations were selected for evaluation by placement of electrostatic dust cloths. After treatment of the samples, bacterial and fungal contamination were assessed, as well as mycotoxin determination, the presence of azole-resistant strains and cytotoxicity of the microorganisms encountered. Bacteria were only present in one of medias used and showed relatively low numbers. The highest level of contamination by fungi was identified in one of the textiles settings. The textile area also showed the highest variability for fungi. Aspergillus sp. are one indicator of possible environmental issues, and A. sections Fumigati and Circumdati were particularly relevant in two of the settings and identified in all of them. No mycotoxins were detected and the large majority of the fungi identified were non-cytotoxic. Overall, these can be considered low-contaminated environments but attention should be given to the Aspergillus sp. contamination. Additional studies are needed not only to make these results more robust, but also to test if the environmental sampling alone is the best approach in a setting where there is very little movement and dust displacement and where professionals are in very close proximity to the artefacts being treated, which may suggest the existence of a micro-atmosphere worth evaluating and comparing to the obtained results.


Introduction
Among all microorganisms, fungi cause the degradation of cultural heritage sites to the greatest extent [1]. In the case of archives, some of the fungi present in paper documents, surfaces and air from archives, libraries and museums are also a threat to human health [2]. Due to their enormous enzymatic activity and their ability to grow at low water activity levels values, fungi are able to inhabit and to decay paintings, textiles, paper, parchment,

Working Settings Assessed
This study sampling campaign was conducted between May and June 2021 in four locations, three in the city of Lisbon and another one in a smaller coastal town. This last one was a private home-based studio, where a single conservator handles historic and artistic objects that can be made from a variety of substrates, some organic, some inorganic and most a composition of both. The several objects in line to be treated are kept at this location and the work is seldom initiated and maintained in different objects at the same time to make up for the obligatory intervals some treatments demand. The other private studio is dedicated to the restoration of paintings. At the time of the evaluation it was being used by three conservators and a master's student. It is a relatively small studio, with two different stories and uses natural ventilation only. As happened in the previous case, there is also an accumulation of other paintings besides the ones that are being treated at the moment, either because the finalized work has not been collected, is between treatment phases or has not been initiated yet. The remaining two settings analyzed are both public run studios and both tend to the restoration of textiles. These are larger rooms and the only pieces that are on display are being treated by the conservators. At the time of the evaluation, one of the studios had 6 workers (this number can shift to 4 in case there is a low demand for conservation work) and the other had 4 workers in the premises ( Figure 1). of mycologists and restorers are needed to develop object specific methods for the conservation and treatment of contaminated objects.

Working Settings Assessed
This study sampling campaign was conducted between May and June 2021 in four locations, three in the city of Lisbon and another one in a smaller coastal town. This last one was a private home-based studio, where a single conservator handles historic and artistic objects that can be made from a variety of substrates, some organic, some inorganic and most a composition of both. The several objects in line to be treated are kept at this location and the work is seldom initiated and maintained in different objects at the same time to make up for the obligatory intervals some treatments demand. The other private studio is dedicated to the restoration of paintings. At the time of the evaluation it was being used by three conservators and a master's student. It is a relatively small studio, with two different stories and uses natural ventilation only. As happened in the previous case, there is also an accumulation of other paintings besides the ones that are being treated at the moment, either because the finalized work has not been collected, is between treatment phases or has not been initiated yet. The remaining two settings analyzed are both public run studios and both tend to the restoration of textiles. These are larger rooms and the only pieces that are on display are being treated by the conservators. At the time of the evaluation, one of the studios had 6 workers (this number can shift to 4 in case there is a low demand for conservation work) and the other had 4 workers in the premises (Figure 1). Figure 1. Sampling locations, OE1 and OE4 for textiles, OE2 for sculpture and OE3 for the painting's restoration studio.

Sampling Approach Characterization through Culture-Dependent Methods
The EDCs were placed in each sampling site ( Figure 2) for 30 days and transported under refrigeration (0-4 °C) to the laboratory for further analyses [26]. EDCs were Figure 1. Sampling locations, OE1 and OE4 for textiles, OE2 for sculpture and OE3 for the painting's restoration studio.

Azole Resistance Screenin
Sabouraud dextrose agar (SDA) (Frilabo, Maia, Portugal), either alone or supplemented with 4 µg/mL itraconazole (ITZ), 2 µg/mL voriconazole (VCZ), or 0.5 µg/mL posaconazole (PSZ), were used to screen fungal resistance to medical azoles (adapted from [30,31]). The A. fumigatus ATCC 204305 reference strain, and a pan-azole-resistant A. fumigatus were used as controls (both strains provided by National Health Institute Doutor Ricardo Jorge, IP). Briefly, SDA media plates inoculated with samples' extracts from all the EDC were incubated at 27 °C (to enable optimal conditions for fungal growth) for three to four days. After incubation, fungal colonies were counted and identified by microscopy, as previously described [32].

Molecular Detection of Aspergillus Sections
The extracts (8.8 mL) from the EDCs were used for molecular detection of Aspergillus sections [26]. Fungal DNA was extracted using the ZR Fungal/Bacterial DNA MiniPrep Kit (Zymo Research, Irvine, CA, USA) and molecular identification was performed by Real Time PCR (qPCR) using the CFX-Connect PCR System (Bio-Rad, Amadora, Portugal). Reactions included 1× iQ Supermix (Bio-Rad, Amadora, Portugal), 0.5 μM of each primer, and 0.375 μM of TaqMan probe in a total volume of 20 μL. Amplification followed a three-step PCR: 40 cycles with denaturation at 95 °C for 30 s, annealing at 52 °C for 30 s, and extension at 72 °C for 30 s.

Azole Resistance Screenin
Sabouraud dextrose agar (SDA) (Frilabo, Maia, Portugal), either alone or supplemented with 4 µg/mL itraconazole (ITZ), 2 µg/mL voriconazole (VCZ), or 0.5 µg/mL posaconazole (PSZ), were used to screen fungal resistance to medical azoles (adapted from [30,31]). The A. fumigatus ATCC 204305 reference strain, and a pan-azole-resistant A. fumigatus were used as controls (both strains provided by National Health Institute Doutor Ricardo Jorge, IP). Briefly, SDA media plates inoculated with samples' extracts from all the EDC were incubated at 27 • C (to enable optimal conditions for fungal growth) for three to four days. After incubation, fungal colonies were counted and identified by microscopy, as previously described [32].

Molecular Detection of Aspergillus Sections
The extracts (8.8 mL) from the EDCs were used for molecular detection of Aspergillus sections [26]. Fungal DNA was extracted using the ZR Fungal/Bacterial DNA MiniPrep Kit (Zymo Research, Irvine, CA, USA) and molecular identification was performed by Real Time PCR (qPCR) using the CFX-Connect PCR System (Bio-Rad, Amadora, Portugal). Reactions included 1× iQ Supermix (Bio-Rad, Amadora, Portugal), 0.5 µM of each primer, and 0.375 µM of TaqMan probe in a total volume of 20 µL. Amplification followed a threestep PCR: 40 cycles with denaturation at 95 • C for 30 s, annealing at 52 • C for 30 s, and extension at 72 • C for 30 s.
A non-template control and a positive control consisting of DNA obtained from a reference that belonged to the culture collection of the Reference Unit for Parasitic and Fungal Infections, Department of Infectious Diseases of the National Institute of Health, from Dr. Ricardo Jorge were used. These strains have been sequenced for ITS, B-tubulin, and Calmodulin.

Mycotoxins Analysis
Nineteen samples were screened for mycotoxins presence. In all samples, 38 mycotoxins were analyzed by HPL-MS (HPLC) Nexera (Shimadzu, Tokyo, Japan) with a mass spectrometry detector API 4000 (Sciex, Foster City, CA, USA) following the same laboratory procedures described in previous papers [27,28]. The mycotoxin concentration was calculated using external calibration. The limits of detection (LOD) obtained for each mycotoxin with the analytical method used are presented in Table 1.

Statistical Analysis
Data were analyzed using SPSS statistical software for windows, version 27.0. The results were considered significant at the 5% significance level. To test the normality of the data, the Shapiro-Wilk test was used. For the comparison of bacterial contamination, fungal contamination and fungal resistance, the Kruskal-Wallis test was used, since the assumption of normality was not verified and given the small size of the sample. To study the relationship between bacterial contamination, fungal contamination and fungal resistance, Spearman's correlation coefficient was used, since the assumption of normality was not verified. To assess species diversity, Simpson and Shannon indices, given by Shannon where p i is the proportion (n i /n) of individuals of one particular species found (n i ) divided by the total number of individuals found (n).

Viable Bacterial Contamination
Total bacteria contamination ranged from 0 to 21.23 CFU·m −2 in one of the textile's working area (OE1) and from 0 to 7.08 CFU·m −2 in the other one (OE4). It ranged from 3.54 to 10.62 CFU·m −2 in the paintings area (OE3) and in OE2 the counts were 7.08 CFU·m −2 . Among the sampled areas, no statistically significant differences were detected (χ 2 K−W (2) = 2.498, p = 0.287), with the sculpture area excluded, since it only had one observation. From the analysis of Figure 3 (boxplot), it can be seen that OE3 displays the highest contamination in TSA and the OE1 is the one presenting higher variability. No gram-negative bacteria (VRBA) were detected in any of the areas sampled.

Viable Fungal Contamination
Total fungal contamination in indoor sites was 233.5 CFU·m −2 ·day −1 on MEA and 28.3 CFU·m −2 ·day −1 on DG18 in the OE1 (textiles); 46 CFU·m −2 ·day −1 on MEA and 10.6 CFU·m −2 ·day −1 on DG18 in OE2 (sculpture); 173 CFU·m −2 ·day −1 in MEA and 31.8 CFU·m −2 ·day −1 in DG18 in the paintings area (OE3). The highest fungal counts were found in the OE1 textiles area (233.5 CFU·m −2 ·day −1 on MEA; 28.3 CFU·m −2 ·day −1 on DG18) (Fig-Figure 3. Bacterial contamination in the four sampled locations observed in TSA. On the top right is the boxplot for comparison of bacterial contamination in TSA medium between sampled areas (excluding OE2, since it has a single record). * severe outlier.  The highest number of fungal species was obtained on the OE1 (8 species MEA; 4 species DG18, more details on all identified genera are on Table 2), closely followed by OE4 (6 species MEA; 3 species DG18). Aspergillus section Fumigati was the most common In both MEA and DG18 medium, no statistically significant differences were detected between the sampled areas (χ 2 K−W (2) = 5.696, p = 0.058 and χ 2 K−W (2) = 0.177, p = 0.915, respectively). However, from the analysis of Figure 4 (boxplot), it can be seen that on MEA, OE3 was the one that presented the greatest fungal contamination and OE1 the one with the greatest variability. In DG18, one can see that OE1 and OE3 were the ones with the highest contamination, with the OE3 showing greater variability (sculpture area excluded as before).

Viable Fungal Contamination
Concerning fungal distribution per sampling location, Figure 4 presents the quantitative results and Figure 5 presents the qualitative results.
The highest number of fungal species was obtained on the OE1 (8 species MEA; 4 species DG18, more details on all identified genera are on Table 2), closely followed by OE4 (6 species MEA; 3 species DG18). Aspergillus section Fumigati was the most common species obtained in OE1 in MEA (43.94%) while Penicillium sp. was the most common species obtained on DG18 (50%) in this same location; in the sculpture area-OE2-the most common species were Aspergillus section Circumdati on DG18 (66.66%) and Penicillium sp. on MEA (53.85%), and in OE3 (paintings) the most common species observed was Cladosporium sp. both in MEA (87.76%) and in DG18 (44.44%). In OE4 Cladosporium sp. accounted for 44.84% in MEA, while Penicillium sp. was the prevalent genera on DG18 (60%).  The highest number of fungal species was obtained on the OE1 (8 species MEA; 4 species DG18, more details on all identified genera are on Table 2), closely followed by OE4 (6 species MEA; 3 species DG18). Aspergillus section Fumigati was the most common species obtained in OE1 in MEA (43.94%) while Penicillium sp. was the most common  Regarding Aspergillus sp., they were present in all the assessed environments. The highest value obtained in MEA (43.94%) was found in OE1 and in OE2 in DG18 (67%). On MEA, the areas with the highest values of the genera were the OE1 (43.94%), followed by the OE4 (8.16%). OE3 displays a lower percentage in MEA (2.04%). The genus was not identified in the MEA media in OE2 but accounts for 67% of the CFUs identified in this location when using DG18.
On DG18, two Aspergillus sections were identified, namely Circumdati (96.67%) and Fumigati (12.50%), also on MEA, two sections were reported, as follows: Fumigati (50.06%) and Nidulantes (4.08%). As for sections identification in OE1, one Aspergillus section was detected both on MEA (43.94% Fumigati) and DG18 (12.50% Fumigati). In OE2, no sections were detected on MEA and one section was detected on DG18 (66.66% Circumdati). In the painting area, OE3, two sections were identified, namely section Fumigati on MEA (2.04%) and section Circumdati on DG18 (33.33%). In the second textiles area, OE4, two sections were detected on MEA (4.08% Fumigati and 4.08% Nidulantes) and one section was detected on DG18 (30.00% Circumdati) ( Figure 6). Regarding Aspergillus sp., they were present in all the assessed environments. The highest value obtained in MEA (43.94%) was found in OE1 and in OE2 in DG18 (67%). On MEA, the areas with the highest values of the genera were the OE1 (43.94%), followed by the OE4 (8.16%). OE3 displays a lower percentage in MEA (2.04%). The genus was not identified in the MEA media in OE2 but accounts for 67% of the CFUs identified in this location when using DG18.

Fungal Growth in Azole-Supplemented Media
Regarding fungal contamination in azole-supplemented media, the results are presented in Figure 7. The most contaminated local was OE4 and the most frequent fungi in SDA (Saboraud Dextrose Agar) was Cladosporium sp. (5.0 × 10 2 CFU·m −2 ·day −1 ), followed by Penicillium sp. (1.0 × 10 2 CFU·m −2 ·day −1 ). Looking into each azole supplement individually, the most frequent fungi was Cladosporium sp. in voriconazole (VCZ) (7.8 × 10 1 CFU·m −2 ·day −1 ) and itraconazole (ICZ) (3.2 × 10 1 CFU·m −2 ·day −1 ), followed by Penicillium sp. in voriconazole (2.5 × 10 1 CFU·m −2 ·day −1 ). Aspergillus sections Flavi (3.5 CFU·m −2 ·day −1 ) and Fumigati (7.0 CFU·m −2 ·day −1 ) were observed in SDA but not in any of the azole added media. The media with PSZ recorded the lowest contamination rate in all settings. Among the sampled areas, no statistically significant differences were detected in any of the media (p's > 0.05) regarding fungal contamination. However, from the analysis of Figure 8A, it can be seen that, for the SDA, the textile area OE4 was the one with the highest values. The ICZ supplemented media (B) with the highest contamination was recorded in OE3. The textile working areas, both OE1 and OE4 presented similar results in in VCZ (C). Lastly, OE3 registered the highest contamination in PSZ (D). Among the sampled areas, no statistically significant differences were detected in any of the media (p's > 0.05) regarding fungal contamination. However, from the analysis of Figure 8A, it can be seen that, for the SDA, the textile area OE4 was the one with the highest values. The ICZ supplemented media (B) with the highest contamination was recorded in OE3. The textile working areas, both OE1 and OE4 presented similar results in in VCZ (C). Lastly, OE3 registered the highest contamination in PSZ (D).

Contamination of EDCs by Mycotoxins and Cytotoxicity Assessment
EDC sampling did not reveal the presence of any of the 38 mycotoxins evaluated at the sampled sites. Regarding the assessment of cell viability of the three distinct cell lines exposed to EDC, the results showed a majority of non-cytotoxic EDC, with only two samples exhibiting an IC50 value of 10 mm 2 /mL (one in A549 lung epithelial cells and another in SK cells).

Correlation Analysis
Only a significant positive correlation of moderate intensity was detected between bacterial contamination in TSA and fungal contamination in MEA (rS = 0.621, p = 0.013), revealing that greater bacterial contamination in TSA is related to greater fungal contamination in MEA (Table 3).

Contamination of EDCs by Mycotoxins and Cytotoxicity Assessment
EDC sampling did not reveal the presence of any of the 38 mycotoxins evaluated at the sampled sites. Regarding the assessment of cell viability of the three distinct cell lines exposed to EDC, the results showed a majority of non-cytotoxic EDC, with only two samples exhibiting an IC50 value of 10 mm 2 /mL (one in A549 lung epithelial cells and another in SK cells).

Correlation Analysis
Only a significant positive correlation of moderate intensity was detected between bacterial contamination in TSA and fungal contamination in MEA (r S = 0.621, p = 0.013), revealing that greater bacterial contamination in TSA is related to greater fungal contamination in MEA (Table 3).

Discussion
Conservators-restorers are a professional class that has yet to see some of its occupational hazards being correctly addressed [11,[34][35][36]. It is not difficult to imagine experiencing allergic respiratory or dermatological symptoms when handling old documentation or textiles that have been exposed to dust or pesticides in the past [12,34,37,38]. As far as health issues are concerned, allergic symptoms (eye and throat pruritus, nasal congestion) and traumatic disorders from the adoption of awkward and stressful body positions, are the top cause of absenteeism or even the abandonment of the profession (personal inquiry). Performing a comprehensive microbiological analysis is, therefore, an essential step to understand the environments where these workers perform their activities. The assessed locations, as mentioned earlier, are mostly dedicated to the treatment of organic-based art pieces (in this case textiles, canvas and wood-based sculptures).
In what concerns fungal contamination assessment, different results were obtained with the two different culture media applied (MEA and DG18), following the trend also found in other occupational environments already assessed [27,39]. In fact, a greater number of fungal counts was obtained in MEA; what is expected since DG18 favors the presence of xerophilic fungi and restricts some fungi with fast growing rates, such as the Mucorales order [27,40]. There appears to be no correlation to be made between the type of materials being handled and the results obtained. The differences in results, for both bacterial and fungal contamination, can be related to the accumulation of dust and the movement due to the performed activities indoors [41][42][43] that may cause its displacement and deposition on the EDC. In addition, outdoor air flow, besides human activities, was reported to be the leading factor responsible for the fungal contamination indoors [44]. The two locations where the fungal counts were higher had other artworks that were not being handled at the moment, which adds to the dust deposition, although this was more evident in OE3. The fungal counts follow the same tendency as bacteria results, which also deem OE1 as the more contaminated. Because conservators spend long hours in the same position, devoted to one task, future studies might find it useful to include the collection of dust from each particular art piece for analysis, as well as the surrounding dust, as performed in this study.
In terms of variability-for fungi only- Figures 5 and 6 confirm a higher variability in the two locations where textiles are handled. This variability is particularly noticeable in OE4, where three sections of Aspergillus were identified. Thus, the workstations OE1 and OE4 can be identified as hotspots for widespread Aspergillus and seen as a priority for risk management intervention. Previously, this genus was found to be the most prevalent on historical textiles and also reported that even with maintenance of recommended conditions, the growth of xerophilic fungi cannot be prevented [45]. In addition, several Aspergillus sections (Circumdati, Flavi, Fumigati and Nidulantes) considered as indicators of harmful fungal contamination were observed, indicating the need for the implementation of corrective measures [46,47]. For this particular case, and because it is organic materials that we are discussing, it is important also to address the impact these fungal and bacterial contaminations can have on the artwork itself. In fact, the textiles' microbial colonization can promote conservators-restorers occupational exposure to these microbiologic risks but also the biodeterioration of historical textiles. Fungi can promote the biodeterioration of cellulosic and proteinaceous archaeological textiles, whereas bacteria are the main players for silk biodegradation [45,48]. Bacteria, however, do need higher water availability, and these art objects are normally kept within safe intervals of relative humidity and temperature [49].
The screening of azole resistance revealed one textile handling environment (OE4) as the one with the highest fungal load in Sabouraud media. Although a limited fungal diversity was found, with predominant Cladosporium and Penicillium sp., two important Aspergillus sections with toxigenic potential-Flavi and Fumigati-were present in the assessed environments. The results come to add another Aspergillus genus to the ones already identified (Figure 6). These results are in accordance with previous studies assessing textile specimens contaminated by fungi in Slovene and Jordanian museums, in which the dominant contaminant fungal species also belonged to the genus Penicillium, Aspergillus and Cladosporium [45].
The fact that Cladosporium sp. and Penicillium sp. were able to grow at tested concentrations of voriconazole and itraconazole must be further investigated in order to determine fungal susceptibility to other commonly used medical azoles for the treatment of fungal infections in humans. Moreover, although not determined during this assessment, azoleresistant Fumigati isolates have been increasingly reported in different environments [50] and described as a potential health menace, especially for immunocompromised individuals [18,51]. A deeper knowledge of fungal susceptibilities to azoles or other biocides is also relevant to guide the adoption of better fungal control strategies in restoration environments and suitable policies on cultural heritage conservation, while ensuring the maintenance of the effectiveness of antifungals in the treatment of infections in humans and animals [52,53].
As reported, none of the mycotoxins analyzed were detected. This might be related with many factors, such as the occupational environment characteristics (e.g., humidity, temperature, availability of fungal nutrients), and the materials being used and handled [54,55]. However, this does not mean that exposure might not happen in this occupational environment since the environmental conditions are constantly changing. It might also depend of the previous contamination of the materials and pieces to be handled and their storage conditions. Further studies are warranted to confirm these scenarios.
No relevant cytotoxicity was observed in EDC samples and the reduced number of samples does not allow further conclusions. Nevertheless, the use of relevant cell lines to assess biological effects and estimate health risks is a valid strategy for risk assessment [56][57][58].

Conclusions
This work presents important findings concerning microbial contamination in an occupational setting not commonly studied. In this particular study, the workplace environment where textiles were handled revealed itself as more prone to a diverse fungal contamination and, more specifically, to Aspergillus sp. contamination. The results, obtained with the techniques identified above, show low contaminated environments overall, considering and comparing with other settings. Thus, after the results are analyzed and compared with different studies using the same methodology, the behavior and particularities of these particular professionals and settings may warrant a conjugation of different approaches. Conservation and restoration is a task that is developed slowly and with care, and this means the workers do not engage in activities that contribute to the aerosolization of particles and, therefore, of fungi and bacteria. This is good news because it reduces the exposure to air contaminants, and good cleaning practices may render the working place safe. However, future studies must accommodate not only the environmental approach, with the EDC placed strategically on the workplace, but also the analysis of the painting, textile, etc., being treated because the conservator works in close proximity to the artefact and shares a micro-atmosphere with the piece itself. Comparing the EDC results with the results obtained by vacuum cleaning, the artefact will possibly increase our knowledge on the particularities of these settings. As such, in future studies an innovative approach (One Health approach)-simultaneously targeting workplaces, workers (and users) and the cultural heritage-should be implemented to allow researchers to map the potential risk of microorganism's dissemination and then, if needed, define an appropriate remediation strategy to simultaneously protect the health of workers and users and prevent further biodeterioration on the cultural heritage artefacts.