Deterioration-associated microbiome of a modern photographic artwork: the case of Skull and Crossbones by Robert Mapplethorpe

The preservation of cultural heritage, including ancient photographic materials, is of paramount importance in the field of conservation science . In this context, the microbial diversity of ‘Skull and Crossbones’, a 1983 photograph by Robert Mapplethorpe printed on silver gelatine, was assessed. We employed both culture-dependent and culture-independent methods to characterize microbial communities inhabiting this artwork. Vibrational Raman micro spectroscopy and FT-IR spectroscopy were utilized to assess the chemical degradation condition and characterize the chemical components of the silver gelatin print. The combination of molecular sequencing methods (Sanger and HTS approach) and non-invasive vibrational spectroscopy yielded valuable insights into the microbial communities thriving on photographic material and the chemical degradation of the print. Isolated fungal strains were added to the Fungal Collection at the University of Naples Federico II, and their deteriorative potential was investigated by adding substrates, commonly used in canvas photographs to the culture media. These results establish a link between microbial communities colonizing ancient photographic materials, paper decomposition, and the enzymatic patterns of the retrieved microorganisms. This information is invaluable for understanding and addressing biodeterioration progression on valuable works of art, such as historical photographs, which remain understudied.


Biodeterioration of modern artworks
Historical photographs represent valuable works of art with a finite life due to the process of biodeterioration.Photographic prints are susceptible to spontaneous deterioration, primarily due to the instability of silver images.Concerns related to their preservation date back to the early times of photography [1].In fact, internal factors may initiate the decay process, which is subsequently exacerbated by external factors, such as inappropriate selection of enclosure materials and unfavorable environmental conditions, including high relative humidity, and UV light exposure [2].
The rise of photographic materials as a cultural asset has been primarily observed over the last two centuries.Several techniques have been used to produce photographic materials, but in general they mainly consist of two components: a support structure and a photosensitive emulsion [3].The support often includes imageforming material, such as silver particles for black and white images and a binder [4], which is usually gelatin or albumin.Gelatin and albumen are two common binders in historical photographs, each offering distinct properties.Gelatin became popular at the end of the 19th century, and thanks to its versatility and stability, gelatin silver prints dominated black and white photography in the 20th century Albumen, predating gelatin, was extensively used in the mid-19th century for albumen prints, known for their rich tonal range and sharp detail.These binders played a crucial role in the construction of photographic prints, aiding in the adhesion of light-sensitive emulsions to various substrates such as paper, glass, or metal [5].
In the context of photographic films, various carriers have been used, such as cellulose nitrate, cellulose diacetate, cellulose triacetate, and polyethylene terephthalate.These components make photographic materials susceptible to microbial attack, particularly by fungi with cellulolytic and proteolytic activities.Fungi can indeed degrade photographs and films through hyphal penetration, attacking and breaking down the binder, leading to irreversible decay [6].Fungal colonies may also manifest as colored spots, leading to a loss of image sharpness, while oxidants and pigment deposition are the main culprits for staining and yellowing of photograph material [7].
However, high environmental relative humidity can lead to gelatin softening and subsequent increased solubility, resulting in changes in print brightness and detachment from the paper [8].Both gelatin and paper serve as the primary energy source sustaining biofilm metabolisms; biofilms actively dissolve gelatin and induce chemical alterations in the silver particles within the emulsion, thus causing gradual photograph dissolution.Both bacterial and fungal communities are responsible for attacking photographs, with the former being particularly destructive [9,10].Recent research has highlighted the compartmentalization of different species within biofilms, each occupying separate niches [11].
Until recently, the conservation and restoration of photographs received limited attention and were often treated as a distinct case of paper biodeterioration.The subject of this study is a silver gelatin printed photograph namely Skull and Crossbones, by Robert Mapplethorpe which is on permanent exhibition and visited by thousands of people annually at the Royal Palace of Caserta (CE, Italy).While it is generally recommended to store photographs in cool and dry environments with minimal light exposure to prevent biodeterioration, this is not feasible for Mapplethorpe's artwork, as the environmental conditions present a significant challenge to museum staff.
The photograph under study showed damages attributable to active biological degradation, and subsequently this study, it has undergone a restoration intervention.The objective of the study was to assess the extent of photo decay through a detailed diagnostic analysis.The correlation between microbial colonization and photographic material damage remains poorly understood [12], yet this information is essential for designing effective restoration approaches.
To achieve this, we first evaluated the extent of resident microbial community on the artwork, identifying the species composition of the biofilm using both culturedependent and -independent methods.While culture techniques are valuable for characterizing biofilm microorganisms, they are insufficient for assessing microbial diversity and the overall abundance as many of microorganisms in biofilms are uncultivable [13].Therefore, we complemented them with culture-independent techniques [14].
Culture-dependent methods were employed to isolate and identify microorganisms sampled from the artwork's substrates, followed by Sanger sequencing to determine their degradation potential.However, since not all microorganisms in biofilm samples are known to contribute to active biodeterioration, we conducted analyses to understand the degradation patterns of isolated microorganisms, focusing on their cellulolytic and proteolytic activity [15].
As a culture-independent method, High-Throughput Sequencing (HTS) was employed to obtain a comprehensive representation of bacteria and fungi inhabiting the Mapplethorpe's artwork as a whole biofilm at higher taxonomic levels.
Further over, we performed vibrational spectroscopic measurements on photographic samples (Raman spectroscopy for the front, and FTIR spectroscopy for the back) to assess the extent of chemical damage resulting from the presence of bacterial and fungal communities.This non-invasive microanalysis technique has been successfully employed to study alterations in photographic materials [16].
The present work represents the first characterization study of the microbial community and chemical degradation state of an artwork from the Terrae Motus modern art collection, thereby serving as a guide for subsequent research endeavors.To date, only a few studies have considered the features and mechanisms of biological degradation of photographic artworks, as further discussed below in this work.Additionally, it is worth noting that until now, the few published works on photographic artworks have focused either solely on microbial characterization or solely on their chemical degradation, without simultaneously considering both aspects [17][18][19][20][21][22].The combined approach used in this study paves the way for a far-sighted non-invasive methodology to study modern non-conventional artworks, such as photographs.Beside the identification of associated microbiomes and their degradative potential -intended as the capacity to break the cellulose and gelatin support, as well as the deposit of pigments -the spectroscopic methods here applied show the potential to map the severity of degradation degree, thus opening the route to the set-up of novel non-invasive diagnostic methods for the preservation of photographic material cultural heritage.

The terrae motus collection and robert mapplethorpe's photographs
The Terrae Motus Collection was established by Neapolitan art gallerist Lucio Amelio  in the early 1980s, following the massive earthquake in Campania and Basilicata dated November 23rd 1980.Amelio conceived the Terrae Motus project, centered on the theme of natural catastrophes, and invited 66 international artists, who collaborated with his gallery, the Modern Art Agency, to create artworks on this subject.Featuring renowned artists like Barceló, Cragg, Warhol, and Mapplethorpe, the collection debuted in 1984 at Villa Campolieto in Ercolano (Naples), and later traveled to Paris in 1987 before finding a permanent home at the Royal Palace of Caserta.One notable installation by Robert Mapplethorpe from the collection includes five black and white photographic prints, with the central photograph, "Skull and Crossbones" exhibiting significant signs of biodeterioration after nearly 40 years on display, prompting a comprehensive conservation assessment.Mapplethorpe's five sequence of photographs emulates the images of a polyptych and alludes to traditional religious paintings.The references to the symbols of Christ's passion are clear, such as: the crown of thorns in Jack with crown; and the profile of Magdalene in Jill Chapman.The shots were all taken in the artist's studio except for the central image with the skull, Skull and Crossbones, which was taken by Mapplethorpe in front of the church of Santa Maria delle Anime del Purgatorio ad Arco in via dei Tribunali, during one of his stays in Naples.The photos were printed, in all probability, by Tom Baril (1952), a historic printer, with whom Mapplethorpe collaborated since 1979 for the very high quality of his work, who continues to print the artist's works for the Mapplethorpe Foundation.The work was exhibited for the first time in Boston, at the Institute of Contemporary Art, in November 1983; later, together with all the other works of Terrae Motus at Villa Campolieto (Ercolano, Naples) in 1984.The state of conservation of the works was compromised by the diffused colonization by potentially biodeteriogenic microbial agents.The image layer of the photographs has been damaged by the metabolic action of the microbial community, in some cases irreversibly, compromising the aesthetics and correct perception of the work.In particular, the Skull and Crossbones print, for conservation reasons, can no longer be re-exhibited, paving the way for new solutions to reconstitute the aesthetic unity of the installation.

Sampling
Sampling was performed on the photo named Skull and Crossbones (Fig. 1).
To the naked eye, the photograph was extensively affected by microbial contamination on back of the photograph (BC) and, to a lesser extent, also on front of the photograph (FC).Two and four samples were respectively taken from the front and back canvas in two ways: (1) by gently pressing 2 × 2 cm adhesive tape samples on photographic material, which were immediately attached to autoclaved slides.From this second set of samples, approximately 1 cm of the tape was then cut using sterile blade, removed with sterile tweezers, and inserted into 2 mL tube tubes, which were used for HTS; (2) with sterilized cotton swabs that were deposited into sterile vials, transported at 4 °C and then stored at − 20 °C.These samples were used for DNA barcoding characterization using Sanger sequencing.

DNA barcoding analyses SANGER sequencing
Isolation of bacterial and fungal strains Each sterile swab used for sampling was added to 1 ml of sterile distilled water, then decimal dilutions were performed in triplicate.After dilutions, 100 µL were plated on dishes containing 30 mL of agarized Potato Dextrose medium (PDA, Merck, Germany) with chloramphenicol (0.1%) and 30 mL of agarized Tryptic Soy medium (TSA, Merck, Germany) with nystatin (0.2%) to isolate total fungi and bacteria, respectively.The PDA plates were incubated at 25 ± 2 °C for ten days, while the TSA plates were incubated at 28 ± 2 °C for 72 h.Once dishes were incubated, fungal and bacterial colonies were counted.Colony forming units expressed in CFU/cm 2 were determined.The result was calculated as the arithmetic mean of 3 independent repetitions.Then, individual colonies were pinched and repeatedly streaked onto agar plates to obtain pure cultures.The isolation of fungi was carried out with a sterile needle under a dissecting microscope.Fungal strains were cultivated using a culture medium PDA, while bacteria strains were cultivated on TSA.All the isolates were added to the novel fungal collection at University of Naples Federico II, namely FCUF [23].
Genomic DNA extraction, amplification, and identification of isolates DNA was extracted from each isolate with a modified DNA extraction protocol [24] and used for a Polymerase Chain Reaction with primers targeting the bacterial 16S rRNA gene (16S_forward: 5′-AGG ATG CAA GCG TTA TCC GG-3′; 16S_reverse: 5′-AAT CCC ATT CGC TCC CCT AG-3′) and the fungal Internal Transcribed Spacer region, including 5.8S gene (ITS1 5′-TCC GTA GGT GAA CCT GCG G-3′; ITS4: 5′-TCC TCC GCT TAT TGA TAT GC-3′).Additional regions, b-tubulin (Bt2a5′-GGT AAC CAA ATC GGT GCT TTC-3′;Bt2b 5′-ACC CTC AGT GTA GTG ACC CTT-3′) and translation elongation factor 1-a (EF1-728 forward 5′-CAT CGA GAA GTT CGA GAA GG and TEF1_reverse TAC TTG AAG GAA CCC TTA CC) were amplified to ensure a robust molecular identification of fungal isolates.Amplification reaction was carried out in a reaction volume of 25 µL containing a buffer reaction (1/10 volume of the supplied 10 × buffer), a deoxynucleoside triphosphate mixture (0.2 mM each), supplemented of 2.5 mM MgCl2, 0.5 mM of each primer and 1.25 U of Taq polymerase (EconoTaq, Lucigen).One microliter of DNA (ca.10 ng) was added to each reaction mixture, then the PCR reactions were run in a thermocycler (Applied Biosystems 2720, Thermo Fisher Scientific) according to conditions described for 16S gene [25], ITS 1-4 [26], BTUB [27], and TEF1 [28].Amplicons were purified using PEG8000 precipitation (PEG 15%, NaCl 2.5 M) and were sequenced using Bright Dye Terminator Cycle Sequencing Kit (ICloning).The amount of primer and template in a cycle sequencing reaction was according to the manufacturer's instruction.The reactions were purified using BigDye Xterminator Purification Kit (Applied Biosystems, Thermo Fisher Scientific) and 13 µL of template was run in an automated sequencer (3130 Genetic Analyzer, Life Technologies, Thermo Fisher Scientific).The characterization of the isolated strains was made employing BLASTN (Basic Local Alignment Search Tool-Nucleotide, https:// blast.ncbi.nlm.nih.gov/ Blast.cgi), identifying the microorganisms by choosing the highest percentage of identity with a 99% cut-off for both bacteria and fungi.A minimum E-value lower than E −4 was considered.All ITS and 16S sequences were submitted to GenBank.

Microbial genomic DNA extraction and PCR amplification
Four adhesive tapes from back of the photograph and two from front ca of the photograph were used for the DNA extraction (see point 1 in Sampling paragraph).Total DNA of the sampled microbial community was extracted from around 1 cm of adhesive tape using a modified protocol according to [29] (Additional file 1., Appendix S1).The DNA was resuspended in 20 μL of which 1 μL was analysed using a Qubit 3 Fluorometer (Invitrogen, Thermo Fisher Scientific) to determine the DNA concentration.To compare the Prokaryotes (archaea and bacteria) and fungi community composition and diversity in each adhesive tape sample, amplicon surveys of a portion of the partial 16S rRNA gene (targeting V3-V4 regions) and the intergenic transcribed spacer (targeting ITS2) were performed (respectively).The barcoded primer sets for 16S rRNA gene were Pro341F (5′-CCT ACG GGN BGC ASC AG-3′) and Pro805R primers (5′-GAC TAC NVG GGT ATC TAA TCC-3′) [30] and for ITS2 nuclear marker were ITS3 (5′-GCA TCG ATG AAG AAC GCA GC-3′) and ITS4 primers (5′-TCC TCC GCT TAT TGA TAT GC-3′) [31], respectively.Both forward and reverse primer contained Illumina overhang adapter sequences (5′-TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG-3′ + forward primer; and 5′-GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA G-3′ + reverse primer).After several optimization amplification tests, PCR was made in a 40 μL reaction mixture consisting of 2 µl of DNA template (ca.0.5-3 ng) and Kodaq 2X PCR MasterMix (Applied Biological Materials Inc., ABM) with 0.25 mM of each primer.The reaction conditions were as follows: an initial denaturation step of 94 °C for 3 min and then 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s and extension at 72 °C for 30 s.The final extension was performed at 72 °C for 5 min.A negative control without DNA was employed for each primer combination.A duplicate for each amplicon was carried out for reproducibility purposes.To remove possible microbial DNA contaminating from MasterMix and water, a modified pre-treatment before the PCR reaction was carried out using 2.5 U of Bsp143I (Sau3AI) (Thermo Fisher Scientific) for 37 °C for 30 min (plus 65 °C 20 min deactivation) [26]; a positive control of digestion reaction was also carried out using as DNA template 300 ng of pGem-3Zf + (Applied Biosystem, Thermo Fisher Scientific).The amplicons were quality checked and quantified using the UVIdoc HD5 gel documentation system (UVITEC) and a Qubit 3 Fluorometer, respectively.A quantitative of 25 μL (> 5 ng/ μL) of each amplicon was shipped to BMR Genomics s.r.l.(Padua, Italy) for sequencing using a MiSeq platform (2 × 300 paired-end sequencing; 2 × 50,000 reads/sample) (Illumina).
OTUs were picked using the script pick_open_ reference_otus.pyfrom QIIME with '-suppressalign-and-tree' flag employing the reference and taxonomy databases version 'sh_refs_qiime_ver8_ dynamic_all_02.02.2019' (https:// unite.ut.ee/ repos itory.php# uchime) for ITS2, '97_otu_taxonomy.txt'default from QIIME for 16S.Statistically tests to determine whether there is statistical significant difference between groups (in our case between the front and back canvas and between sample replicates) were performed inside QIIME using the non-parametric test ANOSIM with 999 permutations using a Bray-Curtis distance metric [37].This test returns a p-value which represents the proportion of permutations that yielded an ANOSIM test statistic (called R) that is equal or stronger than the R calculated from the actual data.

Cellulolytic activity of fungal isolates
The ability of fungal isolates to produce cellulases was determined following the method described by Teather and Wood (1982) using carboxymethyl cellulose (CMC 1%) agarized medium containing carboxymethyl cellulose 10 g; NaNO 3 1 g; K 2 HPO 4 1 g; MgSO 4 .7H2O0.5 g, KCl 1 g; FeSO 4 0.1 g, agar 20 g and distilled water l L was prepared.The medium was sterilized at 121 °C for 20 min and, subsequently, poured into plates square under a sterile flow.One agar disk (8 mm diam.) of 6 days old cultures of each fungal isolate was placed in the center of CMC plates and then incubated at 25 °C for 7 days.After incubation, plates are flooded with 1% Congo red dye and washed with 1 M NaCl after 15 min to stabilize the reaction [38].Because Congo red stain is held by only integral cellulose polymers, cellulase activity was indicated by the clear zones around fungal growth [39].A digital camera was used to capture photographs after staining the plates.Software ImageJ version 1.52 k [40] was used to measure fungal growth (as the diameter of the colony) and the diameter of the halo for subsequent calculation of the enzymatic index (EI), a semi-quantitative estimate of enzyme activity according to the following relation [38]: EI = Diameter of hydrolysis zone/Diameter of colony.

Proteolytic activity
Proteolytic activity was determined using gelatin hydrolysis assays [41].The fungal strains were inoculated into a culture medium with the same composition as that used for the cellulolytic activity test, replacing CMC with 0.5% gelatin as sole carbon source.The medium was sterilized at 121 °C for 20 min.
Fungi were inoculated into gelatin-containing plates, running three spots for each, and incubated in a thermostat at 25 °C for 7 days.When fully developed, the hydrolysis of the gelatin is was highlighted following the procedure described by Vermelho [42]: the plates were coated with a 0.1% solution of black starch in methanolacetic acid-water in the ratio 30:10:60 for 1 h at 30 °C and decoloring with a solution of methanol-acetic acid water, in the ratio 30:10:60 for 15 min; after removing the dye, the plates were flood with the bleach consisting of methanol + acetic acid + water (30:10:60) for 15 min to stabilize the reaction.Any areas of hydrolysis appear in the form of halos clarification.

Statistical analyses
All results were analyzed according to a completely randomized design with three replicates per sample.Analysis of variance (ANOVA) was used for the statistical analysis of the data, and significant differences were compared by Dunnett's test, using Prism software, from three independent replicate values.The value of p ≤ 0.05 was considered statistically significant.

Raman microspectroscopy
Raman microspectroscopy was carried out on the front page of the photograph.Indeed, during the analysis of the state of conservation of the photo, preliminary to the restoration intervention, sampling of microsections of the emulsion layer on the back of the photograph was carried out.The Raman spectra were collected at room temperature using a confocal micro-Raman spectrometer (Jasco, NRS-3100), working with the 514 nm excitation line from an Ar + laser (3 mW at the sample) in a backscattering configuration.The scattered light was analyzed with a spectrometer equipped with a Peltier-cooled 1024 × 128 pixel CCD photon detector (Andor DU401BVI).A holographic notch filter was used to reject the excitation laser line.Raman backscattering was collected using a diffraction lattice of 1200 grooves/mm and 0.01-0.20 mm slits, corresponding to an average spectral resolution up to 1 cm −1 [43].Wavelength calibration was performed by using cyclohexane Raman signal as a reference.The spectra acquisition times varied from 30 to 90 s, and they were triplicated for scope of reproducibility.Assignment of Raman spectra was carried out by comparison with literature data and web database (e.g.RRUFF library).

FT-IR spectroscopy
Fourier-transform infrared spectroscopy was carried out on the back of the photograph.This sampling was carried out on some fragments of the secondary photographic support, subject to microbial infection.The secondary support was then replaced during the restoration work.FT-IR measurements were performed using a Nicolet system (model 5700 FT-IR spectrometer) in ATR mode.About 1-2 mg of sample was put on the optic window (ZnSe crystal).Spectra were collected in the range of 4000-600 cm −1 , 128 scans, and a resolution of 4 cm −1 .

Molecular DNA analyses SANGER sequencing
Culture-dependent isolations The number of culturable fungi on the photographs ranged from 2.6 × 10 5 CFU/ cm 2 (Sample No. 1 back canvas) to 1.0 × 10 6 CFU/cm 2 (Sample No. 4 back canvas).A similar level of contamination was found on the front of the photograph, ranging from 1.1 × 10 5 CFU/cm 2 (Sample No.5 front canvas) to 2.1 × 10 5 CFU/cm 2 (Sample No.8 front canvas).Bacteria were the least numerous on the photographs, with their numbers ranging from 2.1 × 10 1 CFU/cm 2 (Sample No. 2 back canvas) to 1.6 × 10 2 CFU/cm 2 (Sample No. 6 front canvas).No algae nor cyanobacteria were retrieved from our samples.
DNA barcoding identification In total, nineteen fungal isolates and one bacterial isolate were recovered.Concerning the back side of the photograph, four distinct fungal species from three different genera were isolated.No bacterial strains were detected using the culture-dependent method from samples derived from the back of the photograph (Fig. 2).
As for the front front of the photograph, four distinct fungal from four different genera, plus one species of bacteria, were isolated (Fig. 3).
The recovered fungi belonged to Ascomycota (89%) and Basidiomycota (11%) phyla.The most abundant isolates belonged to the Cladosporium genus, followed by Penicillium.Purpureocilium, Sporobolomyces and Paraconiothirium genera were isolated in equal proportions.The genus Chaetomium was the least abundantly isolated.
The molecular identifications obtained with eukaryotic nuclear rRNA/ITS and prokaryotic 16S rRNA were submitted on Gene Bank.The processed accession numbers are OR816146-OR816164 for eukaryotic barcode (submission ID: SUB13982410), while the processed accession number is PP468626 (submission ID: SUB14304634) for prokaryotic barcode and are listed in Table 1.

Deteriorative potential
Given that not all microorganisms thriving on artworks pose a risk for their integrity and appearance, we investigated the enzymatic potential of the obtained isolates.Focusing solely on fungal isolates, given that they are generally more associated with deterioration compared to bacteria, all the fungal isolates showed cellulolytic activity, with enzymatic indices ranging between 0.8 ± 0.1 and 1.7 ± 0.1.However, not all isolates displayed a proteolytic activity (Table 2).

High-throughput sequencing
The percentage of reads that included adapters varied between 6.9% and 84.3% before their subsequent removal.Reads that did not pass the quality Phred score cutoff of 20 ranged from 2.5% and 24.4%.For the 16S analysis, the total number of OTUs spanned between 38,000 to approximately 94,000.After applying this filter, the median number of OTUs was 60,081.500.In the case of ITS2 analysis, the total number of OTUs ranged from 84,000 to around 173,000 reads, a median of 129,946.000.The ANalysis of SIMilarity (ANOSIM) test did not reveal statistical significance when samples were grouped based on the front and back localization of the canvas (16S, p-value 0.17; ITS2, p-value 0.28).However, significant values were observed when samples were analyzed individually, both for 16S and for ITS2 (p-values 0.001).Rarefaction plots to set sampling depth for alpha diversity are shown in Fig. 4.
Detailed relative abundances of Prokaryotes and fungi OTUs (from phylum to genus) are provided in Additional file 1: Tables S1 and S2.Subsequent bar plots for OTUs at the phylum and class levels are presented in Figs.5-6 for Prokaryotes and fungi, respectively.
A total of 147 genera across 28 different classes of bacteria were identified, with 17 and 6 accounting for ≥ 1% of relative total abundance (RA T ), respectively (Additional file 1: Table S1).One hundred and two OTUs were excluded because they could not be verified at the genus level.Overall, 43 classes were well taxonomically identified, belonging to 18 phyla.Considering the RA T ≥ 1%, this included 12 and 8, respectively (Fig. 5).The dominant phylum was Proteobacteria (37%), with Alphaproteobacteria and Gammaproteobacteria being the most representative classes (21.1% and 12.1%, respectively) in all sample from the photo (Fig. 5).Additionally, Betaproteobacteria was identified with a lower RA T (3.5%).The Proteobacteria genera identified for all samples (RA T ≥ 1%) were Sphingomonas (6.7%), Enhydrobacter (3.2%), Paracoccus (2.9%), Erwinia (2.7%), Acinetobacter (2.2%), Pseudomonas (1.8%), Haemophilus (0.8%), Phyllobacterium (0.8%) and Methylobacterium (0.6%).The phylum Cyanobacteria was also well represented by the class Oscillatoriophycideae with 18% (Fig. 5).The most represented family was Xenococcaceae (18%) with Chroococcidiopsis genus (13%).The genus Calothrix was present in only one sample (4 BC) with a lower relative abundance (RA) of 1%, while it was absent in two samples (6 FC, 2 BC) or in traces in the other samples.Overall, the observed bacterial diversity (RA T ≥ 1) did not show differences in composition but in the abundance of the sampled bacterial communities in Mapplethorpe's photograph.As previously mentioned, the composition of the fungal communities is shown in Fig. 6 and the relative abundances of fungal taxa are detailed in Additional file 1: Tables S2.
Fungal diversity at the phylum and genus levels was lower than that of the bacterial communities.Additionally, there were unassigned OTUs with a higher abundance compared to bacterial communities (Fig. 5 and Fig. 3 Proportions of retrieved fungi on front of the photograph under the phylum, order, class, family, genera, and species using culture dependent method (A); proportions of retrieved bacteria on front canvas under the phylum, order, class, family, genera, and species using culture dependent method (B) Fig. 6).This can be attributed to the marker's universality for certain eukaryotic taxa (i.e., Viridiplantae).A total of 135 genera across 14 different fungal classes were identified, with 16 and 8 accounting for ≥ 1% of the relative total abundance (RA T ), respectively.A significant portion of assigned fungi belonged to Ascomycota (49.9%), with Dothideomycetes as the dominant class with 33.1% of RA T .The classes Eurotiomycetes, Sordariomycetes, and Leotiomycetes were identified with lower RA T values (6.9, 4.9 and 2%, respectively).Saccharomycetes exhibited higher abundance in the front canvas (3.9% vs 1% back canvas).The dominant genera were Cladosporium (16.6%) and Alternaria (7%).Other genera with lower abundance but consistently observed in all sampling points included Aspergillus (2.4%) and Penicillium (1.8%).Considering an RA T for ≥ 1%, genera variably present or in trace amounts in sampling points were: Aureobasidium (1.6%), Stemphylium (1.4%), Gibberella (1.2%), Fusarium and Erysiphe (0.9%), as well as Hanseniaspora, Cadophora (0.7%), and Coniosporium (0.6%) (Additional file 1: Table S2).Hanseniaspora was more abundant on the front canvas (FC) sampling points [2.3 vs 0.1% back canvas, (BC)].
The phylum Basidiomycota (26.1%) was well represented by the class Malasseziomycetes (RA T = 18%), followed by Tremellomycetes and Agaricomycetes with lower RA T values (6 and 2.2%, respectively) (Fig. 6).The most represented genus was Malassezia (16%), while Dioszegia, Filobasidium and Vishniacozyma (Tremellomycetes) were present at all points with lower relative abundance values (2.1 -1.3%).Genera belonging to Agaricomycetes were well represented only in traces (Peniophora, Trametes and Pleurotos).Overall, the observed fungal diversity (RA T ≥ 1) did not show differences in composition but in the abundance of the sampled bacterial communities in Mapplethorpe photographs.
Additionally, among the other OTUs unidentified as fungi, Chlorophyta (Viridiplantae) was detected with the genus Trebouxia in all samples, with a variable abundance, more pronounced in the back of the photo (0.25 FC vs 0.8% BC) and Coccomyxa and Acutodesmus present only in trace amounts on the front canvas (RA ≤ 0.1%, data not shown).

Vibrational spectroscopies
Before delving into the results of the vibrational spectroscopic analysis of the gelatin-silver halide black and white prints, it is essential to note that the presence of various species, including both fungi and bacteria, on the photographic prints makes it challenging to conduct a detailed spectral analysis.This is because it becomes impossible to differentiate the contribution of each species.The examined prints were created on baryta papers, where silver halide salts were suspended in gelatin.
The Raman spectra of the microsamples taken from degraded and defaced areas on the front page of the photograph are presented in Fig. 7.The efflorescence in the spectra exhibit peaks at 1139, 990, 645, 620, 521, 458 cm −1 , which are associated with BaSO 4 [44], or baryte, the mineral situated between the support and the gelatin binder in gelatin prints.Additionally, signals in 1200-1720 cm −1 region may be attributed to gelatin and/or microorganisms: 1239 cm −1 (amide III) and  Paraconiothyrium fuckelii 0.9 ± 0.1 -1650 cm −1 (amide I).Notably, neither barium phosphates (978 cm −1 ) [45] nor anhydrite, CaSO 4 [46] were observed, in contrast to other cases of degradation.
Moreover, TiO 2 anatase features were detected (146, 384, 520, and 633 cm −1 corresponding to Eg, B1g, A1g,  S1 and Eg lattice vibration modes) without significant amount of titanates (which would be expected at 234, 660, and 1055 cm −1 ), as elsewhere reported [40].There was no detection of amorphous carbons (which typically appear around 1350 and 1660 cm −1 ).Overall, Raman spectroscopy of the front of the photograph reveals only chemical component expected in the original photograph, without highlighting any direct chemical degradation, except for the extensive gelatin digestion induced by the microorganisms.
For the support on the back of the photograph, fluorescence hindered Raman acquisition, so FT-IR spectroscopy was employed.FT-IR spectra of white, yellow, and brown regions of the back of the photograph were collected in ATR mode and are presented in Fig. 8.
The white and yellow regions exhibited very similar FT-IR spectra, indicating the acetylation of the cellulose support [48,49].Conversely, the brown regions revealed only cellulose [50] and no longer exhibited the initial acetate coating.This evidence clearly indicates a stepwise degradation of the multilayer structure on the retro support of the photograph.The yellow areas reasonably underwent to a reversible microbial attachment on the acetate coating, whereas in the brown area microbiological degradation reached the support paper.

Discussion
In this study, the microbial composition within "Skull and Crossbones," a significant photograph by Robert Mapplethorpe, was assessed.The identification of the associated microbiome found on both the front and back surfaces of the photograph was essential to evaluate its potential for causing degradation.Moreover, Raman and FT-IR measurements were respectively carried out on the front and the back of the photograph to assess the degree of chemical damage caused by the presence of the bacterial and fungal populations.
In detail, for the molecular analysis we selected the internal transcribed spacer (ITS 1-2) region, widely used as the official barcode for the molecular determination of fungal species and due to its ease of amplification and a well-defined barcode gap.Indeed, there is a vast amount of ITS data available on several International Nucleotide Sequence Databases (GenBank, EMBL and DDBJ).In particular, the entire ITS region or its subregions, namely ITS1 and ITS2, can be used as DNA barcoding markers for the Ascomycota and Basidiomycota phyla.However, the molecular identification success of the ITS region is higher in Basidiomycota than Ascomycota [51], primarily because some genera within the Ascomycota phylum, such as Cladosporium [52] and Penicillium [53], show narrow or no barcode gaps in the ITS sequence.Consequently, ITS may not offer sufficient variation to distinguish diverse fungal genera.Therefore, to achieve a deeper level of species identification, we used a combination of two other genes [54], β-tubulin and translation elongation factor 1-α (TEF1-α).The selection of these biomarkers was based on preliminary morphological data and the results obtained from ITS sequencing.ß-tubulin was chosen in accordance with established data recommending it as a secondary barcode for species identification in certain fungal genera, such as Penicillium and Purpureocillium, [55].On the other hand, TEF1-α marker was used to confirm the identification of some of plant and human-pathogenic fungal strains retrieved in the present study, including Basidiomycota species and genera Cladosporium.Indeed, literature data show the higher specificity and discriminatory power of TEF1-α with respect to ITS [56].The results obtained by combining ITS 1-2, ß-tubulin and TEF1-α markers are consistent with previous studies that have identified Cladosporium cladosporiales, Cladosporium sphaerospermum, Penicillium chrysogenum, Purpureacillium liliacinum and Chaetomium globosum as colonizers of photographic materials [57].These fungi are well known for producing a wide range of enzymes, including proteinases and cellulases, which can degrade the components of photographic materials, including gelatin and paper [58].For instance, the genus Penicillium can produce diffusive pigments and secrete malic and citric acid, along with cellulases, as reported for P. chrysogenum [59].
Chaetomium globosum produces various extracellular enzymes, including amylase, cellulase, laccase, lipase, pectinase, protease and chitinase [60].Furthermore, most of these fungal species are commonly present in indoor environments, consequently constituting a serious risk to cellulose-based materials, such as photographs and canvas paintings.Indeed, under suitable conditions these microbes can initiate biodeterioration processes.Microbes in indoor environments can initiate biodeterioration processes under suitable conditions.Notably, genera like Cladosporium (Dothideomycetes), Chaetomium (Sordariomycetes), Penicillium (Eurotiomycetes) and Purpureocilium (Sordariomycetes) have been identified as xerophilic fungal organisms in indoor settings [61].These fungi have the remarkable ability to produce a variety of pigments, including black, green, blue, purple, and violet, derived from different chemical classes like carotenoids, melanin, or quinones [62].It is interesting to note that Paracooniothyrium (Ascomycota) and Sporobolomyces (Basidiomycota) which we identified within the microbiota of the Skull and Crossbones photograph have never previously reported as biodeteriogens-or simply inhabitants-of photographs or indoor works of art.However, both isolates displayed significant cellulolytic activity, suggesting their potential to deteriorate materials with high cellulose content.For instance, Paraconiothyrium fuckelii is a well-known soil and plant pathogen fungus [63,64].Similarly, Sporobolomyces roseus belongs to Uredimiomycetes, typical plant pathogens (rusts and smuts).Although these fungi may be opportunistic, we cannot confirm their origin and whether their presence is due to a particular level of specialization in using paint components or to a circumstantial deposit of spores from the environment.To gain a better understanding of the microbial communities thriving on the Skull and Crossbones photograph, we also employed cultureindependent methods to identify the species that could not be cultivated (i.e., Basidiomycetes).In particular, we used high-throughput sequencing (HTS) techniques that are ultra-sensitive and are excellent tools for accurate metagenomic studies.Such methods have already been successfully employed to investigate the microbial presence on photographic-based materials.Prokaryotes and fungal communities identified on Mapplethorpe's photo using HTS showed significant complexity.In comparison to similar studies conducted on objects stored at the Tianjin Museum [65], the bacterial biodiversity found was higher, although some trends, such as the dominant occurrence of Proteobacteria, Firmicutes, and Actinobacteria, were confirmed.The bacterial composition profile exhibited a closer resemblance to that found on objects from Crete Museum, as reported by Saridaki et al. [66].At the genus level, the predominant presence of Paracoccus, one of the most abundant genera colonizing Mapplethorpe's photo, was observed, along with the occurrence of cyanobacteria belonging to Oscillatoriaceae.The unique characteristics of photographic materials may account for the distinctive composition of microbial communities, since surface characteristics influence their profile [12,67].Enhydrobacter is one of the major components of the community detected on silver gelatin photographs [68], and Sphingomonas, the most abundant genus found on Mapplethorpe's photograph, is a ubiquitous bacterium frequently isolated from various artifacts worldwide [69].Even at low temperatures, it is capable of degrading proteinaceous substrates, such as gelatin or albumen, commonly used as photographic binders [70].Overall, the most abundant bacteria identified on Mapplethorpe's artwork possess cellulolytic and proteolytic capabilities, potentially affecting the biodeterioration of photographic artworks.
The comparison between the results obtained from culture-dependent and culture-independent methods for bacteria detection is intriguing.The limited number of bacterial strains isolated from Mapplethorpe's photo may be attributed to a very low concentration of bacterial cells in the samples.The elevated levels of fungal contamination could potentially have contributed to a decline in bacterial concentration within the consortia.Additionally, it could be hypothesized that a significant portion of the bacterial components in the microbial population entered a resistance stage, limiting their growth potential in culture.An exception to this general pattern is the genus Bacillus, identified as the sole cultivable bacterium in the present study.Species within this genus typically produce spores capable of rapid growth in culture media [71].Furthermore, this discrepancy may be elucidated by the limitations of culture-dependent methods, which can only identify 0.1-1% of microorganisms present on artworks.This small fraction of microorganisms thrives under laboratory cultivation conditions, enabling their isolation, while the rest may consist of undiscovered species or species in a viable but non-cultivable state (VBNC) for which suitable laboratory conditions cannot be established [72].
However, the discovery of Bacillus as a cultivable bacterium aligns with previous studies indicating that this genus is frequently isolated from audiovisual materials [2].
The fungal community structure retrieved on The Skull and Crossbonnes photograph predominantly featured the phylum Ascomycota, followed by Basidiomycota, as reported for many art objects affected by fungi, regardless of substrate characteristics, such as drawings [73], stones [74,75], audio-visual materials [7].At the genus level, the fungi identified in this work were also previously described as colonizing photographic materials [57].The presence of a forested area around the building housing the Terrae Motus collection could account for the high prevalence of airborne fungi as Cladosporium and Alternaria.Furthermore, the significant occurrence of Malassezia, a human skin parasite, is likely associated with the high number of Museum visitors.Cladosporium spp.often form patinas or staining on gelatin positive print and negative films, as well as on the black and white albumen of paper photographs; a similar propension to colonize gelatin cellulose triacetate films was reported for Alternaria, particularly A. alternata [6] and Malassezia.Malassezia is one of the most frequently occurring genera in indoor samples of museum exhibition rooms [76].Consistent with Sanger sequencing results, HTS results also showed a higher presence of Ascomycota species than Basidiomycota.Other genera within the Ascomycota were also detected, such as Alternaria, Aspergillus, and Fusarium, which are commonly found in association with various works of art of historical and cultural heritage.Despite the high frequency of the Ascomycota phylum, HTS sequencing revealed a significant presence of the Basidiomycota phylum.Among Basidiomycota, the most abundant genus was Malassezia, which was not detected using culture-dependent methods.Malassezia, is a well-known dermatophyte opportunistic fungus, commonly associated with human skin as a causative agent of skin-related issues such as dandruff and seborrheic dermatitis.It is also found within other warm-blooded animals [77].Although Malassezia is an unusual contaminant in cultural heritage, recent studies have revealed its unexpected presence on cultural heritage artifacts, such as ancient manuscripts, paintings, sculptures, photograph, and textiles [78,79].The presence of this fungus on such objects poses new challenges for conservators and researchers due to its ability to degrade cellulose and produce lipolytic enzymes [80].Furthermore, its spores could easily spread into the surrounding closed environment where photograph is exposed, representing a serious health concern for visitors, restorers, and museum workers.The discovery of Malassezia on cultural heritage artifacts emphasizes the importance of continuous research and vigilance in the field of heritage preservation.
Lastly, the presence of taxa belonging to Anthophyta in trace amounts was also observed (data not shown), likely due to airborne or human contamination.
Overall, the presence of these described fungi in photographic materials and their ability to degrade cellulose and protein in vitro demonstrate that these microorganisms could be responsible for the degradation and appearance of foxing in historical photograph.All isolates in this study showed cellulolytic activity, while not all microorganisms showed proteolytic activity.
Finally, to evaluate the chemical degradation conditions, we used vibrational spectroscopies [81,82].In particular, the FT-IR analysis on the back of the photograph allowed to correlate the different colors of the cellulose support with different degradation stage, in terms of loss of its acetate coating.Conversely, the Raman analysis shows that the extensive microbial digestion of the gelatin at the front of the photograph is not accompanied by any direct chemical degradation.
Understanding the factors contributing to fungal infestation on cultural heritage artifacts and the consequent degradation is crucial for implementing effective mitigation strategies, considering the presence of organic residues, and environmental conditions promoting fungal proliferation.Additionally, environmental conditions, such as temperature and humidity fluctuations, may play a role in promoting fungal proliferation on vulnerable artifacts.The handling, exposure and storage conditions of artwork inhabited by human opportunistic pathogenic fungi should be evaluated with care, continuously monitored and, if necessary, improved.By understanding the implications of fungal infestations and implementing effective mitigation strategies, conservators can better safeguard these precious objects for future generations.

Conclusions
Even though museums of significant historical relevance, such as the Royal Palace of Caserta, maintain controlled indoor environment, fluctuations in relative humidity and temperature can still occur, potentially leading to the attachment of microorganisms, particularly fungi, on exhibited works of art.In addition, it is important to consider that the large number of visitors can facilitate the introduction of microorganism into the museum, which can contribute to chemical, physical, mechanical, or aesthetic degradation of the artworks.The combined application of molecular and spectroscopic methods has proven to be a valuable tool for understanding not only the composition and function of microbial communities colonizing ancient photographic material but also the extent of chemical alteration of the photograph under investigation.These findings can assist in identifying resilient microbes on the substrate, enabling the establishment of strategic interventions to control the museum's climatic conditions and visitor flow.Additionally, it is essential to continuously monitor the structure, composition, and activity of microbial communities to assess the impacts of control measures or conservation treatments.Therefore, the development of a standardized approach to investigate biodeterioration issues is desirable.
archives, photography, digitalization; GOG: care of the historical and artistic heritage; AV: funding acquisition, experimental design, reviewer; AP funding acquisition, experimental design, analysis of results, reviewer.All authors discussed the research outcome, as well as approved the final manuscript.

Fig. 1
Fig. 1 Sampling on Skull and Crossbones photograph on back of the photograph (BC) samples n.o.1-4 (A) and on front of photograph (FC) samples n.o.6-7 (B)

Fig. 2
Fig.2Proportions of retrieved fungi on back of the photograph under the phylum, order, class, family, genera, and species using culture dependent method

Fig. 6
Fig. 6 Proportions fungal operational taxonomic units (OTUs) under the phylum and class ranks using High-Throughput Sequencing data through metabarcoding analysis of ITS2 nrDNA marker for the seven tape samples [front of the photograph (FC): 6, 7; back of the photograph (BC): 1, 2, 3, 4].Further details for taxonomical level are provided in Additional file 1: TableS2

Fig. 7
Fig. 7 Raman spectra of the microsampled degraded and defaced areas on the front of the photograph.Ce (cellulose), I (amide I) and III (amide III)

Fig. 8
Fig. 8 ATR-FT-IR spectra of white, yellow, and brown regions of the back of the photograph.Ce (cellulose), CA (cellulose acetate).Star refers to carbon dioxide

Table 1
Identification of microorganisms sampled from Skull and Crossbones photograph using ITS barcode.BC = back of the photograph; FC = front of the photograph

Table 2
Deteriorative potential of fungal isolates derived from culture-dependent methods