Introduction

Penicillium, the phialidic, anamorphic, species-rich genus of the Trichocomaceae in the Eurotiomycetes, inhabits a variety of substrates and environments. Members of this genus are important in human life; they commonly appear as soil, storage, indoor and airborne fungi or as food contaminants (e.g., Pitt 1979; Domsch et al. 2007; CBS databases [http://www.cbs.knaw.nl/fungi/BioloMICS.aspx]). Fungal biodeterioration is well known to affect cultural properties such as catacombs (Saarela et al. 2004) and murals (Berner et al. 1997; Dhawan et al. 1993; Guglielminetti et al. 1994) in Europe. In addition, the roles of fungi in the deterioration of murals, as well as their decay mechanism, have been reviewed by Nugari et al. (1993) and Garg et al. (1995); those studies listed, at the species level and including Penicillium spp., fungi that have been isolated from deteriorated murals.

On the other hand, in Japan, microbiological surveys related to the conservation of stone chamber interiors and murals were conducted for the Ozuka and Chibusan Tumuli in Fukuoka Prefecture (Emoto and Emoto 1974), the Torazuka Tumulus in Ibaraki Prefecture (Emoto et al. 1983; Arai 1984, 1990a), and the Takamatsuzuka (Arai 1984, 1987, 1990b; Kigawa et al. 2004, 2006b) and Kitora (Kigawa et al. 2005, 2006a) Tumuli, both in Nara Prefecture. Anamorphic fungi such as species of Cladosporium, Doratomyces, Fusarium, Penicillium, and Trichoderma were detected from samples taken from these tumuli (cf. Tables 1 and 3 in Kiyuna et al. 2008). In most cases, however, identification of Penicillium isolates in these previous studies was limited to the generic level.

The Takamatsuzuka Tumulus (TT) and the Kitora Tumulus (KT), which are thought to have been built in late 7th century, are well known in Japan as invaluable cultural heritage sites because of their beautiful mural paintings, which were drawn directly onto thin plaster in the small stone chamber interior. The mural paintings of the TT were discovered in 1972, whereas the KT was discovered in 1983 and excavated in early 2004. The Japanese government designated the TT and KT as Special Historic Sites in 1973 and 2000, respectively.

The TT is located in the village of Asuka, Nara Prefecture, Japan. It is a circular burial mound measuring approximately 20 m in diameter. The stone chamber was constructed of cut slabs of volcanic tuff; the interior area is approximately 2.7 m deep, 1.0 m wide, and 1.1 m tall (for the schematic diagram, see Fig. 1 in Kiyuna et al. 2008). The colorful paintings adorn the inner walls and a ceiling of the stone chamber with various colors, including red, green, yellow, blue, and gold. The figures comprise star constellations (Seishuku) on the ceiling, and the sun and the moon (Nichi-Getsu), the four heavenly guardian gods (Shishin), and groups of men and women on the four walls. The group of women on the west wall is called the “Asuka beauties” (Asuka bijin) (see Fig. 2 in Kiyuna et al. 2008). All the paintings were designated as National Treasures by the government in 1974. In accordance with specialists’ advice, the Japanese Agency for Cultural Affairs decided in 1973 to preserve the murals on site and they were never opened to the public. It was thought that the beauty and stability of the plaster paintings could be preserved if conditions inside the stone chamber would be kept as before, naturally buried state; high humidity (100% RH) and cool temperature (14–20°C) could be established. In addition, the stone chamber was kept in darkness except for regular inspections by the staff involved in its conservation.

The KT is also located 1.2 km south of the TT in the same village of Asuka. Its circular tomb is approximately 14 m diameter; the stone chamber interior is approximately 2.4 m deep and 1 m wide and tall; a diagram of the stone chamber and related facilities is not illustrated here, but the KT adjacent small room corresponds to the adjacent space of the TT. Similar colorful murals are drawn on the thin plaster wall within the stone chamber, and a star chart on the ceiling is in the same style as in the TT. The interior environmental conditions are quite similar to those of the TT, and no conspicuous fungal colonies were seen for some time after the excavation. In 2004, in the course of the excavation, species of Acremonium, Fusarium, Penicillium, and Trichoderma appeared as the main contaminants (Kigawa et al. 2005, 2006a, 2007). In the case of the KT, the paintings on the plaster wall had become partly detached from their support. Considering such situation, it was decided to relocate these paintings to a controlled environment in July 2004. By the end of 2008, all the paintings except for possible ones hidden by thin wall layer, were relocated.

We have detected and reported the haplotypes of Fusarium and Trichoderma, the two predominant fungal colonizers, using the 28S rDNA D1/D2 region (28S) and EF-1α (=tef1: protein-coding gene translation-elongation factor 1-alpha) gene sequence-based analyses (Kiyuna et al. 2008). We have also proposed two new species of Candida assignable to the C. membranifaciens clade: C. tumulicola and C. takamatsuzukensis, isolated mainly from biofilm samples from the stone chamber interior of the Takamatsuzuka Tumulus (Nagatsuka et al. 2009). In the previous papers (Sugiyama et al. 2008, 2009; Kiyuna et al. 2008), we have mentioned that Penicillium isolates, which comprised one of the predominant fungal groups, were obtained from samples in the stone chamber interiors and adjacent spaces or rooms of both tumuli. Of 1,780 fungal isolates, 662 were assignable to Penicillium. And 181 of those isolates have been phenotypically assigned as Penicillium sp. 1.

In the present study, we reveal phenotypically and genotypically the identity of Penicillium sp. 1 isolates as the major Penicillium colonizers in the stone chamber environments in the Takamatsuzuka and Kitora Tumuli.

Materials and methods

Sample sources, isolation, and culturing

A total of 224 samples, including mold spots, viscous gels (biofilms), and mixtures of plaster fragments and soil, were collected from the stone chamber interior, spaces between the stone walls, and stone chamber exterior of the Takamatsuzuka Tumulus (TT) between May 2004 and May 2007. As well, a total of 149 samples were collected from the stone chamber interior and exterior of the Kitora Tumulus (KT) between June 2004 and September 2007. Isolation methods have been mentioned in Sugiyama et al. (2008, 2009) and Kiyuna et al. (2008). The isolates have been maintained on potato dextrose agar (PDA; Nihon Pharmaceutical, Tokyo, Japan). For strain data originally identified as Penicillium sp. 1 from the Takamatsuzuka and Kitora Tumuli, see Table 1 and supplementary Table S1. The 11 selected isolates are deposited with the Japan Collection of Microorganisms (JCM), RIKEN BioResource Center, Wako, Saitama Prefecture, Japan, as JCM 15985–15995 (Table 1). The remaining P. paneum isolates from both tumuli are maintained at the Biology Laboratory of the National Research Institute of Cultural Properties, Tokyo as lyophilised cultures (Table 1, supplementary Table S1).

Table 1 Strain data of TT and KT isolates, with major ubiquinone system data and the GenBank accession numbers, for 28S, ITS, β-tubulin, and lys2 gene sequences determined in this study
Full size table

Cultural and morphological observations

A total of 50 isolates, comprising 31 and 19 isolates from TT and KT, respectively, were used in the cultural and morphological observations. These isolates were selected from the standpoint of sampling sites and dates as well from the standpoint of kinds of substrates. The detailed data on isolates are shown in supplementary Table S1. All isolates were grown using the media and growth conditions adopted by Pitt (1979, 2000) and Frisvad and Samson (2004). The isolates were incubated on Czapek yeast autolysate (CYA) agar at 25°C, malt extract agar (MEA) at 25°C, 25% glycerol nitrate (G25N) agar at 25°C, yeast extract sucrose (YES) agar at 25°C, and creatine sucrose (CREA) agar at 25°C all for 7 days in darkness, respectively. Growth rates were measured on CYA inoculated with conidial suspension and incubated at 5, 10, 15, 20, 25, 30, 37 and 40°C all for 7 days in the dark. Furthermore, growth rates were determined in the moist chamber, about 100% RH. The colony colors of the isolates on all media were determined by using Kornerup and Wanscher’s color standard (1978). Microscopic slides were prepared from portions of the colonies grown on MEA plate cultures and were mounted in lactophenol mounting medium without dye (Bills and Foster 2004). Microscopic examinations were made using a fluorescence microscope BX51 (Olympus, Tokyo, Japan) with Normarski interphase contrast at up to ×1,000 magnification. All micrographs were taken with a Coolpix P5000 digital camera (Nikon, Tokyo, Japan).

Ubiquinone determination, a chemotaxonomic characteristic

Mycelia grown on Difco™ Potato Dextrose Broth (Becton–Dickinson, MD, USA) were used for the extraction of ubiquinones. The isolates and other authentic strains were cultured at 25°C for 3 days on a rotary shaker (Taitec, Saitama, Japan) at 125 revolutions per min−1. Extraction, purification, and determination of ubiquinones by high-performance liquid chromatography were carried out as described previously (Kuraishi et al. 1985, 1991). When one type of ubiquinone molecule constituted more than 90% of the types found in a particular strain, it was determined to be the major ubiquinone type (Kuraishi et al. 1985, 1991).

DNA extraction, PCR amplification, and sequencing

The isolates used for the DNA sequencing are listed in Table 1. Their genomic DNA was extracted using a DNeasy Plant Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The four gene regions sequenced were the 28S rDNA D1/D2 region (28S), ITS-5.8S rDNA (ITS), part of the β-tubulin, and part of the lys2 (aminoadipate reductase) gene. The primers used included NL1 and NL4 (O’Donnell 1993) for 28S, ITS5 and ITS4 (White et al. 1990) for ITS, Bt2a and Bt2b (Glass and Donaldson 1995) for β-tubulin, and lys2F and lys2R (An et al. 2002) for lys2. Polymerase chain reactions (PCR) were performed using puReTaq Ready-To-Go PCR beads (GE Healthcare, Buckinghamshire, UK). Thermal cycling was performed using a GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA, USA). Initial denaturation at 95°C for 5 min was followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s (28S, ITS)/58°C for 30 s (β-tubulin)/56°C for 1 min (lys2), extension at 72°C for 1 min, and then a final extension at 72°C for 10 min. The amplified DNA fragments were purified with a QIAquick PCR Purification Kit (Qiagen, Hilden, Germany). The PCR products of the lys2 gene were cloned using a PCR Cloning Kit (Qiagen, Hilden, Germany). Two to three white clones of each the PCR products were sampled. Sequencing of the PCR products and the cloned plasmids was performed using the BigDye Terminator Cycle Sequencing Kit v3.1 (Applied Biosystems, CA, USA). Reactions were incubated in a GeneAmp PCR System 9700 (Applied Biosystems, CA, USA) and the products cleaned using a DyeEx™ 2.0 Spin Kit (Qiagen, Hilden, Germany). Sequences were determined using electrophoresis in an ABI3130xl DNA sequencer (Applied Biosystems, CA, USA). The sequences determined in this study were deposited in GenBank/EMBL/DDBJ, and their accession numbers are given in Table 1. Other sequences downloaded from GenBank (Table 2) are shown in the respective molecular phylogenetic trees.

Table 2 Strain data used from molecular phylogenetic analyses with the major ubiquinone system and GenBank accession numbers for 28S, ITS, β-tubulin, and lys2 gene sequences
Full size table

28S and ITS molecular phylogenetic analyses

The sequences were assembled using ChromasPro 1.42 (Technelysium Pty Ltd., Tewantin QLD, Australia). Multiple alignments were performed using CLUSTAL W version 1.83 (Thompson et al. 1994), and then the final alignments were manually adjusted. Ambiguous positions and alignment gaps were excluded from the analysis. The alignments and trees were deposited in TreeBASE, accession number SN4551. The neighbor-joining (NJ) tree based on the respective gene sequences was constructed using the multiple alignments in MEGA ver3.1 (Kumar et al. 2004), with 1,000 bootstrap replicates (Felsenstein 1985).

Molecular phylogenetic analyses of β-tubulin and lys2 genes

The β-tubulin and lys2 gene analysis for Penicillium, the phylogenetic tree by NJ, and the maximum likelihood (ML) and Bayesian (Bayes) analyses were performed. The alignments and trees were deposited in TreeBASE, accession number SN4551. NJ trees based on the respective gene sequences were constructed using the multiple alignments in MEGA ver3.1 (Kumar et al. 2004), with 1,000 bootstrap replicates (Felsenstein 1985). ML trees based on the respective gene sequences were also constructed using the multiple alignments in PHYLIP version 3.6 (Felsenstein 2002), with 100 bootstrap replicates. The Bayesian approach to phylogenetic reconstruction (Rannala and Yang 1996) was implemented using MrBayes v3.1.1-p1 (Huelsenbeck and Ronquist 2001). The model of DNA substitution was calculated using Modeltest 3.0 (Posada and Crandall 1998). The results were obtained by the K80 model and gamma correction (G) rate variation among sites for β-tubulin, and the general time-reversible (GTR) model and G rate variation among sites for lys2. Bayesian Metropolis-coupled Markov chain Monte Carlo (MCMCMC) analyses (Mau et al. 1999) were performed with MrBayes for phylogenetic estimation inferred from the respective gene sequences. MrBayes was run for 1,500,000 generations for β-tubulin and 1,000,000 generations for lys2. Searches were conducted with four chains (three cold, one hot) with trees sampled every 100 generations. The average standard deviations of split frequencies were 0.008 and 0.006, respectively, at the end of the run. The confidence levels of nodes were measured by posterior probabilities obtained from the majority rule consensus after deletion of the trees during burn-in.

Results and discussion

Cultural and morphological characterization of Penicillium sp. 1 isolates

Cultural and morphological characterization of Penicillium sp. 1 was based on 50 representative isolates from the TT and KT stone chamber interiors and exteriors (for details, see Table 1 and supplementary Table S1).

Colonies on CYA at 25°C for 7 days grew rapidly and were 30–40 mm in diameter, radially sulcate, velutinous, predominantly greyish turquoise (25CD-6) or greyish green (28B2-3), and had a white (25A1-2) margin; the reverse was pale yellow. No growth was observed at 37°C for 7 days. Colonies on MEA at 25°C for 7 days grew rapidly, and were 43–50 mm in diameter, velutinous, and predominantly pale green (26A-3) to greenish white (26A-2), with a greenish white (25A-2) margin; the reverse was pale yellow. Colonies on YES at 25°C for 7 days grew rapidly, and were 40–43 mm in diameter, radially sulcate, velutinous, and predominantly greyish white (25A-2) to pale green (25A-3); the reverse was pale yellow. Colonies on G25N at 25°C for 7 days were 26–30 mm in diameter, velutinous, and predominantly greyish green (25B-5), with a white (25A-1) margin; the reverse was pale yellow. Colonies on CREA at 25°C for 7 days attained a diameter of 18–20 mm and were velutinous, predominantly pale green (25A-3). The optimum temperature for growth on CYA, in the four selected Penicillium sp. 1 isolates (T4519-5-5, T5916-6-1, T6517-1-2 and K5916-6-1), was 20–25°C; the colonies attained ca. 50–60 mm diameter after 7 days. The growth rates were similar to the moist chamber, about 100% RH. These results suggested that TT and KT isolates were mesophilic.

Conidiophores arose from the subsurface and submerged hypha, and stripes reached lengths of 100–200 (–250) μm, characteristically with very rough and tuberculate walls. The penicillus was typically terverticillate, occasionally quaterverticillate, with appressed elements; the rami were 20–25 (–30) μm long with tuberculate walls; the metulae were 12–15 μm long; and the phialides were ampulliform, 7.5–8.5 (–10) μm long, with short collula. Conidia were spherical, (3–) 3.5–4 (–4.5) μm in diameter, with smooth or finely rough walls, pale green to dark green, borne in long chains, and closely packed. No sclerotium-like structures were observed.

The above-mentioned cultural and morphological characterization of Penicillium sp. 1 isolates differed from those of P. roqueforti and P. carneum, both in Penicillium subg. Penicillium sect. Roqueforti by a pale yellow colony reverse on CYA and YES. The phenotypic characteristics of Penicillium sp. 1 isolates agreed well with those of the ex-holotype (CBS 101032) of P. paneum Frisvad, the third species in the same section and its previous descriptions (Boysen et al. 1996; Frisvad and Samson 2004), except for producing finely rough conidia. As a whole, the results of our morphological observations were in concordance with those of O’Brien et al. (2008). Further studies are needed to clarify whether or not the finely rough conidial formation is a common characteristic of P. paneum.

Chemotaxonomic characterization of Penicillium sp. 1 isolates

Three selected Penicillium sp. 1 isolates (T5916-6-1, T7214-8-1, and T7425-4-1) and the ex-holotype strain CBS 101032 and CBS 465.95 of P. paneum, which were determined in this study, had Q-9 as the major ubiquinone system, a chemotaxonomic characteristic (Table 1). It agreed with that of the remaining two species within the same section Roqueforti, P. roqueforti and P. carneum. As a result, three known species comprising P. roqueforti, P. carneum and P. paneum in the same section Roqueforti, have been characterized by the same ubiquinone system Q-9 (Table 2; cf. Kuraishi et al. 1991).

Molecular phylogenetic position of Penicillium sp. 1 isolates

Here we show evidence from our molecular phylogenetic analyses of four different gene sequences (28S, ITS, β-tubulin, and lys2; Figs 1, 2, supplementary Figs. S1 and S2) that the selected 15 isolates of Penicillium sp. 1, i.e., 13 from the TT and 2 from the KT (Table 1), are assignable to Penicillium paneum in sect. Roqueforti subg. Penicillium.

Fig. 1
figure 1

Cultural and morphological characteristics of Penicillium sp. 1 (isolate T5916-6-1). Colonies on CYA (a), MEA (b), G25N (c), CREA (d), and YES (e) at 25°C, 7 days. Conidiophores and penicilli (f, g), and conidia (h). Bars 10 μm

Full size image
Fig. 2
figure 2

Phylogenetic relationships among 15 TT and TK isolates and 28 known Penicillium isolates for which the sequences were obtained from GenBank, based on NJ analysis of ITS sequence data of 519 aligned nucleotide sites using MEGA ver3.1. Numbers on the branch nodes represent bootstrap support values (%) based on 1,000 replications; bootstrap values greater than 50% are indicated. T of the respective isolate numbers indicates isolates from TT, whereas K indicates those from KT. Right vertical bars indicate the section(s) and clade(s) defined by Samson et al. (2004). Strains with a superscript indicate the strains derived from; HT, Holotype; NT, Neotype; IT, Isotype; and ET, Epitype

Full size image

Using P. brevicompactum JCM 22849 (Penicillium subg. Penicillium sect. Coronata) as an outgroup taxon, the NJ tree (supplementary Fig. S1) was inferred from 594 bp of 28S sequences of 44 isolates of Penicillium and relatives. In the 28S phylogeny, all the TT and KT isolates were placed in a cluster comprised only of P. paneum, including the ex-holotype strain CBS 101032 (isolated from mouldy rye bread in Denmark) of P. paneum. However, the bootstrap value showed that the reliability of the phylogeny was not sufficient to endorse a taxonomic position at a species level because of a comparatively low confidence level (80%).

Using the same outgroup taxon as the 28S phylogeny, the NJ tree (Fig. 2) was inferred from 519 bp of ITS sequences of 43 isolates of Penicillium and relatives. In the ITS phylogeny, all the TT and KT isolates were placed in the cluster containing only P. paneum, including its ex-holotype strain CBS 101032. The bootstrap support showed a comparatively high value, 94%. Incidentally, ITS is now expected to be the primary target gene among possible barcodes for fungi (e.g., All Fungi Barcoding [http://www.allfungi.com/index.php]). An assessment of the DNA barcode for Penicillium subg. Penicillium was done by Seifert et al. (2007) using ITS, CO1 (mitochondrial gene cytochrome oxidase I; also known as Cox1), and β-tubulin gene sequences. CO1 provided barcodes for Penicillium subgen. Penicillium, superior to the resolution of ITS, but inferior to that of the β-tubulin gene. Our case study of the barcode for TT and KT fungal isolates is now in progress. The results will appear elsewhere in the near future.

Using Eupenicillium catenatum CBS 431.69 as an outgroup, the NJ, ML, and Bayesian trees (Fig. 3) were inferred from 254 bp of β-tubulin sequences of 31 isolates of Penicillium and relatives. In the β-tubulin phylogeny, all the TT and KT isolates were placed in a cluster containing only P. paneum, including its ex-holotype strain CBS 101032, with strong statistical supports. The bootstrap values were 99 and 100%, respectively, and the posterior probability was 1.00. The respective three species of sect. Roqueforti were well-supported by the bootstrapping, as was the topology of the species phylogeny demonstrated by Boysen et al. (1996). Very recently, O’Brien et al. (2008) adopted β-tubulin and acetyl CoA synthetase sequences to identify their isolates in this section. The β-tubulin sequence-based phylogeny of TT and KT Penicillium sp. 1 isolates were identical with that of P. paneum isolated from baled grass silage by O’Brien et al. (2008).

Fig. 3
figure 3

Phylogenetic relationships among 15 TT and TK isolates and 16 known Penicillium isolates for which the sequences were obtained from GenBank based on NJ, ML, and Bayesian analyses of β-tubulin gene sequence data of 254 aligned nucleotide sites, using MEGA ver3.1., PHYLIP version 3.6., and MrBayes. Numbers on branches indicate Bayesian posterior probability and bootstrap support values (%) based on 1,000 replications; bootstrap values greater than 50% are indicated. For other abbreviations, see Fig. 2

Full size image

In fungi, aminoadipate reductase converts 2-aminoadipate to 2-aminoadipate 6-semialdehyde. An et al. (2002) have already shown that the lys2 gene (encoding of aminoadipate reductase) fragment is useful for the phylogenetic analyses of ascomycetes. However, the prokaryotes that biosynthesize lysine through the aminoadipate pathway have no lys2 gene; they synthesize lysine from the aminoadipate using a different pathway (An et al. 2002, 2003; Nishida and Nishiyama 2000). The lys2 gene is fungal-specific (An et al. 2002).

Using P. brevicompactum JCM 22849 as an outgroup taxon, the NJ, ML, and Bayesian trees (not shown herein; see supplementary Fig. S2) were inferred from 384 bp of the lys2 sequences of 30 isolates of Penicillium and relatives. In the lys2 phylogeny, all the TT and KT isolates were placed in a cluster comprised only of P. paneum, including its ex-holotype strain CBS 101032. The bootstrap supports were 99 and 96%, respectively, whereas the posterior probability was 0.70. No sequence variation was observed in the 28S, ITS, and β-tubulin gene sequences from the Penicillium sp. 1 isolates. In the lys2 gene, however, T6517-1-2 differs isolates T7425-4-1, T7510-5-1, T7528-21-1, T4906-11-8 and K6203-2-1 by one or two nucleotides. Consequently, the lys2 gene may suggest a potential to distinguish populations of P. paneum in genealogical studies. Further analyses are needed to test the availability for fungi.

Our molecular phylogenies inferred from the 28S, ITS, β-tubulin, and lys2 gene sequences clearly indicate that the 15 selected Penicillium sp.1 isolates of TT and KT, as well as the ex-holotype and another strain of P. paneum, formed a monophyletic cluster with high bootstrap supports and posterior probability.

In our previous paper, we detected several haplotypes of Fusarium and Trichoderma isolates from the TT and KT stone chamber interiors and exteriors (Kiyuna et al. 2008). However, our molecular phylogenies suggest that there are no haplotypes within Penicillium sp. 1 (i.e., P. paneum). As shown in Table 1 and supplementary Table S1, 50 representative strains of Penicillium sp. 1, which were isolated from a variety of substrates such as mouldy spots and viscous gels (biofilms) on plaster walls, soil, air, and Isopoda’s body surface collected in different periods, are thought to be the same species, P. paneum.

On the other hand, the most abundant phylotype P. namyslowskii (=Geosmithia namyslowskii), a common soil colonizer, has been reported from the Chamber of Felines in the prehistoric painted cave of Lascaux in France (Bastian and Alabouvette 2009). P. paneum (this paper) and P. namyslowskii (Peterson 2000) shows a difference of 46 nucleotides in the ITS region. In addition to this, the former is characterized by the Q-9 system, whereas the latter has Q-8 (23%) + 9 (77%) as the major ubiquinone system (Ogawa et al. 1997). Therefore, there is a big difference between the two.

So far P. paneum has been isolated from mouldy rye breads, other foods (Frisvad and Samson 2004), and baled grass silage (O’Brien et al. 2006, 2008). The present report is thus the first account of P. paneum isolated from samples relating to the biodeterioration of cultural properties such as mural paintings. This novel finding will help elucidate the cause of biodeterioration of the TT and KT murals. Very recently an important role of arthropods in the dispersion of fungal spores and development has been expressed by Bastian et al. (2009) in the prehistoric painted cave of Lascaux. In the near future, we will discuss on roles of P. paneum in the biodeterioration of mural paintings and plaster walls, and also on the invasion route to the TT and KT stone chamber interiors.

In conclusion, the phenotypic and genotypic characteristics of Penicillium sp. 1 isolates from TT and KT samples, agreed well with those of the authentic strains (including the ex-holotype) and related references of P. paneum Frisvad (Boysen et al. 1996; Frisvad and Samson 2004; O’Brien et al. 2008).