In the present work, we analysed the microbiomes associated to the inner and external surfaces of UV cabins used in the Dermatology Service of the Hospital General Universitario de Valencia to treat skin pathologies. First, we wanted to shed light on whether the UV irradiation shaped the microbial communities of the cabins. Second, we wanted to explore the possible correlation between UV light-resistance and resistance to antibiotics commonly used to treat skin pathologies. For this, we used a double strategy based in culture-dependent (culturomics, colony identification and biological activity tests) and independent techniques (high-throughput 16S rRNA gene sequencing).
From the culturomics point of view, the microbial profiles we found are moderately diverse and include both environmental and human associated microbial taxa (Fig. 1A). From the most abundant taxa to the least, the high abundances of Staphylococcus, and to a lesser extent of Micrococcus and Bacillus, are not surprising since these genera are naturally present on the human skin [33–34]. Moreover, Kocuria and Pseudomonas have also been associated with skin disorders, such as psoriasis, and some Arthrobacter species are opportunistic human pathogens, such as A. creatinolyticus or A. woluwensis, or have been isolated from human clinical specimens [35–38].
However, both Kocuria and Arthrobacter species inhabit soils, being, thus, common environmental species [2], and the large genus Bacillus not only includes pathogenic species such as B. cereus or B. anthracis, but is also a typical environmental species in different natural habitats [39–40]. Some of these genera are also known by their tolerance to biotic and abiotic stressors. Specifically, Micrococcus spp. and Kocuria spp. have been isolated from polar environments and reported to be resistant to radiation [41–44], whereas Arthrobacter spp. are present in hot deserts [2, 45–46]. Moreover, Bacillus spp. are well known by their tolerance to stress given their ability to form resistance spores [47].
From the clinical point of view, we found relevant to analyze the presence of some health-threatening genera, such as Acinetobacter or Pseudomonas. Both genera have been described to have an innate adaptation ability, including the acquisition of antibiotic resistances [48–50]. Among them, A. baumanii and P. aeruginosa strains fall into the ESKAPE group of multi-drug resistant bacteria [51–52]. Moreover, other present genera such as Rhodococcus, Roseomonas or Stenotrophomonas host species that cause infection to immunocompromised patients, and Cryptococcus spp. have been reported to cause opportunistic infections [52–56].
Apart from the above-described ones, there is also a less abundant representation of some environmental-associated taxa, many of which have been isolated from varied environments such as Deinococcus, Domibacillus, Pantoea or Sphingomonas [57–60]. Others have been linked to isolated cases of fungaemia, bacteremia or sepsis, such as Aureobasidium, Kosakonia, Lysinibacillus or Massilia [61–64]. However, Massilia is also a common soil-inhabitant [65–66]. Regarding fungi, the ones we isolated in pure culture belonged to the genera Aureobasidium (five isolates), Ustilago, Cystobasidium, Rhodotorula and Cryptococcus (the last four represented by just one isolate). Despite the low number of fungal isolates selected, the fact that the yeast genera Aureobasidium, Rhodotorula and Cryptococcus had previously been reported to inhabit different hospital facilities is in accordance with our results [67].
The comparison of the isolation surface (inside or outside the cabins) revealed similar microbiomes. However, the existence of a cluster of exclusive genera in each location reveals some differences from the culturable point of view. The dominance of Staphylococcus in both isolation sources is in accordance with its widely known role in skin pathology [68]. In contrast, Pseudomonas, a sensitive genus, is only detected outside [69]. Curiously, all six isolates identified as Frigoribacterium spp. were isolated from the inner surfaces of the cabins (Fig. 1A). This genus was first described as a psychrophilic genus isolated from dust in a cattle barn in Finland [70].
As revealed by high-throughput 16S rRNA gene sequencing, the all the cabins and locations display similar taxonomic profiles in terms of α-diversity. However, in terms β-diversity some differences are observed at the genus level (Figs. 3A and 3B). The fact that cabin one (working in UVA) and cabin two (working in UVB) are the most similar ones suggests that UV does not have a significant effect in shaping the microbial biocenosis in our studies (Fig. 3B).
At the phylum level, the predominant phyla found (Pseudomonadota, Actinomycetota and Bacillota) are in accordance with the already described profiles found in hospitals by other authors. Moreover, the comparison of the microbial profiles found in the samples at the genus level with those of hospitals is also in accordance with previous studies, particularly due to the presence of Staphylococcus, Pseudomonas, Acinetobacter and Streptococcus (Figs. 4A and 4B) [27, 71]. Interestingly, the presence of the genera Rubellimicrobium, Deinococcus, Bacillus, Hymenobacter and Sphingomonas is in line with the already described microbial communities living on solar panels [3]. This may suggest that the studied microbiomes are a combination of both highly-irradiated surfaces, such as solar panels, and hospital environments. Although, at the genus level, there are differences between both cabins and locations (Fig. 4B), the most relevant genera according to their environmental or clinical interest do not show significant differences in their distribution, with the exception of Staphylococcus (Fig. 5C).
Finally, the genera Rubellimicrobium, Paracoccus and Corynebacterium are among the most abundant genera through NGS whereas they are completely absent in the strain collection (Figs. 4B and 1A, respectively). Biases in culturing techniques are well-known and our results support the importance of combining both culture-dependent and culture-independent techniques in microbial ecology.
We tested the hypothesis that the resistance to UV irradiation of the species isolated inside the cabins would be higher than that of the species isolated outside the cabins. However, our results did not support this statement (Fig. 6). All the strains isolated exclusively from the inner surfaces showed a significant decrease in the survival rate after 15 s or 30 s of UV irradiation, with the only exception being K. polaris at 15 s of UV irradiation treatment. Similarly, in the group of isolates obtained exclusively from the outer surfaces there were only two strains whose survival did not decrease significantly after 15 s of UV exposure: A. agilis and D. ficus. Moreover, in the comparison of species isolated from both the inner and outer surfaces of the cabins, no relevant differences were found out. There was only one outside M. luteus isolate displaying higher resistance that the inner strain (Fig. 6). This suggests that, in this case, UV exposure is no causing an adaptive response for bacteria (Fig. 6A).
According to this experiment, the most resistant strains were A. bussei, A. agilis, K. polaris, D. ficus and M. luteus (Figs. 6B and 6C). Some Kocuria strains have been reported as highly resistant to different types of radiation, as well as to synthesize carotenoids and encode genes related to oxidative stress [41, 44, 72–74]. Moreover, Deinococcus spp. have been extensively studied for its high resistance to radiation, which is a result of a combination of mechanisms such as robust DNA repair systems regulatory proteins, enzymatic and non-enzymatic antioxidant strategies [74–75]. On its part, the genus Arthrobacter hosts several multi-resistant species to different abiotic stressors [76–77]. Specifically, A. agilis has been reported to produce the C50 carotenoid bacterioruberin [78]. Finally, Micrococcus spp. have DNA repair mechanisms fundamental on their resistance to UV light [79–80], and M. luteus strains have been reported to resist high doses of gamma radiation [42]. These taxa are, thus, highly resistant to radiation and other stresses, and naturally inhabit soil and desert-like environments [2–3].
Even though for most of the strains the survival rate after UV treatment was significantly reduced compared to the non-irradiated control, many of them showed mid-viability after 15 s of treatment. This is the case of A. variabilis and M. esteraromaticum, isolated from the inner surfaces of the cabins, and A. bussei, C. flaccumfaciens and K. arsenatis, isolated both from the inside and the outside. Some Acinetobacter species have demonstrated to be able to cope with oxidative stress [81–82]. Moreover, the genus Microbacterium has extensive background on UV resistance and carotenoid synthesis as well [83–85]. In contrast, less has been described about Curtobacterium spp., but still there are reports on their tolerance to UV [86].
The reasons for the absence of an enrichment on resistant species inside the cabins may be varied (Figs. 6A and 6B). On the one hand, both surfaces are accessible to patients, which may be in contact constantly with both surfaces. This would explain also the high similarities found in terms of diversity. Moreover, the stress to which they are subjected (short pulses of UV light) may be less intense than the stress tested under laboratory conditions. From this perspective, the species tested may not represent a threaten.
As stated in the introduction, hospitals and sanitary environments increase the population of multidrug resistant pathogens. Considering that the surface of UV-cabins is constantly in contact with patients with skin pathologies, many of them with complementary treatments with antibiotics, and that the use of radiation may favor the selection of antibiotic resistant bacteria [30], we hypothesized that the isolates taken from the inside and the outside of these cabins may present an altered susceptibility to antibiotics. In this regard, we assessed the antibiotic resistance of the isolates from Table S2 to six antibiotics widely used in dermatology: AMC, DXT, GEN, MUP, AZM and CD (Fig. 7). The classification of the strains as sensitive or resistant has been done according to the European Committee on Antimicrobial Susceptibility (EUCAST) as expressed in the instructions of the MTS.
The presence of some resistant strains is confirmed, with special interest on the genus Staphylococcus. In this regard, there were found isolates resistant to DXT and GEN (S. hominis isolated from outside), MUP (S. haemolyticus), AZM (both S. cohnii isolates, S. haemolyticus and S. hominis from inside) and CD (both S. cohnii). The resistance to MUP is remarkable, as this antibiotic is specifically used in the treatment of topic dermal infections by Gram-positive cocci, which also explains the resistance values displayed by most of the tested strains (which are not Gram-positive cocci). Moreover, the resistance of S. haemolyticus to MUP has been already reported and is mediated by the gene mupA [87], and the rest of the mentioned staphylococci have reports on their multi-resistances [88–90]. There were some differences between strains isolated from different surfaces belonging to the same species, such as the case of S. hominis: the inside isolate is resistant to AZM whereas the outside one is resistant to DXT and GEN. As for S. rhizophila, our strain showed resistance to AMC, and there are reports on their multi-resistance to many antibiotics [91]. Regarding the strains for which a breakpoint has not been established, no susceptibilities can be assigned.
Taken together, the results obtained from both resistance assays (UV and antibiotics) reveal that most of microorganisms are not resistant to the antibiotics tested. Those that are, which are the staphylococci, do not show UV-resistance. The rest of the isolates we tested lack clinical interest as they are not common human pathogens. Therefore, the microbial community of the cabins is mainly composed of antibiotic-sensitive micro-organisms which display a diverse sensitivity to UV light, and a few potential pathogenic microorganisms that are sensitive to UV light (and that should thus be eliminated easily with UV-based sterilization devices). Moreover, the lack of reports of infections associated with the cabins supports a lack of substantial threat in their microbial content. However, the combination of the presence in the cabins of some microbial pathogens and the presence of antibiotic resistant genes poses an obvious potential problem linked to horizontal gene transfer (Fig. 8).