Talaromyces atroroseus, a New Species Efficiently Producing Industrially Relevant Red Pigments

Some species of Talaromyces secrete large amounts of red pigments. Literature has linked this character to species such as Talaromyces purpurogenus, T. albobiverticillius, T. marneffei, and T. minioluteus often under earlier Penicillium names. Isolates identified as T. purpurogenus have been reported to be interesting industrially and they can produce extracellular enzymes and red pigments, but they can also produce mycotoxins such as rubratoxin A and B and luteoskyrin. Production of mycotoxins limits the use of isolates of a particular species in biotechnology. Talaromyces atroroseus sp. nov., described in this study, produces the azaphilone biosynthetic families mitorubrins and Monascus pigments without any production of mycotoxins. Within the red pigment producing clade, T. atroroseus resolved in a distinct clade separate from all the other species in multigene phylogenies (ITS, β-tubulin and RPB1), which confirm its unique nature. Talaromyces atroroseus resembles T. purpurogenus and T. albobiverticillius in producing red diffusible pigments, but differs from the latter two species by the production of glauconic acid, purpuride and ZG–1494α and by the dull to dark green, thick walled ellipsoidal conidia produced. The type strain of Talaromyces atroroseus is CBS 133442


Introduction
Monascus species are known to produce six major azaphilone pigments being the yellow monascin and ankaflavin; the orange monascorubrin and rubropunctatin and the red monascorubramine and rubropunctamine, in addition to more than 20 related pigments [1,2]. Another azaphilone series of yellow pigments is even more widespread in Talaromyces, i.e. the mitorubrins [3][4][5]. The red pigment producer Monascus purpureus has been used primarily in Southern China, Japan and Southeast Asia for making red rice wine, red soybean cheese and Anka (red rice) [6]. A problem is that some samples of Monascus-fermented rice have been found to contain the mycotoxin citrinin [7], but also that Monascus isolates also often produce mevinolin, a drug that is also unwanted in foods [2]. The production of such mycotoxins and drugs limits the use of Monascus for industrial purposes, but since citrinin has not been found in any Talaromyces species, the latter may be a good alternative for red pigment production.
Studies have shown that polyketide azaphilone Monascus red pigments and/or their amino acid derivatives are naturally produced by Talaromyces aculeatus, T. pinophilus, T. purpurogenus and T. funiculosus [8,9]. Talaromyces amestolkiae, T. ruber and T. stollii also produce azaphilone polyketides, as recently described by Yilmaz et al. [10], but in those three species the pigment are not diffusing into the growth medium. Talaromyces amestolkiae and T. stollii were isolated from immuno-compromised patients and are potential human pathogens, while T. purpurogenus produces mycotoxins such as rubratoxins A and B, rugulovasins, and luteoskyrin [10]. These factors limit the use of these species for biotechnological production of azaphilone pigments.
In the current study we describe a new Talaromyces species, T. atroroseus, which secretes large amounts of Monascus red pigments, without the production of any known mycotoxins.

Strains
Cultures were obtained from the CBS-KNAW Fungal Biodiversity Centre culture collection, Utrecht, the Netherlands. Fresh isolates deposited in the working collection of the Department of Applied and Industrial Mycology (DTO) housed at CBS, and strains from the IBT collection at DTU Systems Biology in Kongens Lyngby, Denmark were also included in this study. Strains are listed in Table 1. KAS strain numbers are from the fungal collection of Keith A. Seifert, Ottawa, Canada.

Morphological analysis
Macroscopic characters were studied on agar media Czapek-Dox yeast autolysate agar (CYA), CYA supplemented with 5 % NaCl (CYAS), yeast extract sucrose agar (YES), creatine sucrose agar (CREA), dichloran 18 % glycerol agar (DG18), oatmeal agar (OA) and malt extract agar (Oxoid) (MEA). The isolates were also tested on CYA at 37 °C and on Blakeslee malt extract agar (MEA2). All media were prepared as described by Samson et al. [11]. The strains were inoculated in three points onto media in 90-mm Petri dishes and incubated for 7 d at 25 °C in darkness. After incubation, the colony diameters on the various agar media were measured. Colonies were photographed with a Canon EOS 400D. Species were characterized microscopically by preparing slides from MEA. Lactic acid was used as mounting fluid. Specimens were examined using a Zeiss AxioSkop2 plus microscope.

DNA extraction, PCR amplification and sequencing
Strains were grown for 7 to 14 d on MEA prior to DNA extraction. DNA was extracted using the Ultraclean TM Microbial DNA isolation Kit (MoBio, Solana Beach, U.S.A.). The extracted DNA was stored at -20 °C. The ITS regions, regions of the β-tubulin and RPB1 genes were amplified and sequenced according to methods previously described [12][13][14][15].

Data analysis
Sequence contigs were assembled using Seqman from DNAStar Inc. Newly generated ITS, β-tubulin and RPB1 sequences were included in a data set obtained from the Samson et al. [15] study. Data sets were aligned using Muscle software within MEGA5 [16]. Neighbour-joining analysis on the individual data sets was performed in MEGA5 and confidence in nodes determined using bootstrap analysis with 1000 replicates. Talaromyces galapagensis (CBS 751.74 T ) was selected as a suitable out-group in all the phylogenies. The newly generated sequences were deposited in GenBank (accession numbers, see Table 1 and Figures 1-3).

Extrolites
Cultures grown on CYA and YES for 7 d at 25 °C were used for extrolite extractions. Extracts were analysed by HPLC using alkylphenone retention indices and diode array UV-VIS detection as described by 17-19, using three 6 mm agar plugs. Standards of extrolites from the collection at DTU Systems  [18]. The extrolite extractions from T. atroroseus CBS 133450, CBS 113154 and CBS 123796 were also analysed by ultra high performance liquid chromatography high-resolution mass spectrometry (UHPLC-HRMS). Liquid chromatography was performed on an Agilent 1290 Infinity LC system with a DADdetector coupled to an Agilent 6550 iFunnel Q-TOF with an electrospray ionization source. The separation was performed on a 2.1 x 250 mm, 2.7 μm Poroshell 120 Phenyl-Hexyl column (Agilent) at 60 °C with a water-acetonitrile gradient (both with 20 mM formic acid) going from 10 % (vol/vol) to 100 % acetonitrile in 15 min followed by 2.5 min with 100 % acetonitrile and then returning to the start conditions for 2.5 min for equilibration before next sample. All time the flow rate was kept at 0.35 mL/min. HRMS was performed in ESI + and extrolites were identified with targeted search on accurate mass of [M+H] + and [M+Na] + using Agilent MassHunter Qualitative Analysis B.06.00 software and a database of potential extrolites in T. atroroseus with support from UV-VIS spectra. The list of compounds searched for including the extrolite standards can be found in Table S1.

The electronic version of this article in Portable Document
Format (PDF) in a work with an ISSN or ISBN will represent a published work according to the International Code of Nomenclature of algae, fungi, and plants, and hence the new names contained in the electronic publication of a PLOS ONE article are effectively published under that Code from the electronic edition alone, so there is no longer any need to provide printed copies. In addition, new names contained in this work have been submitted to MycoBank from where they will be made available to the Global Name Index. The unique MycoBank number can be resolved and the associated information viewed through any standard web browser by appending the MycoBank number contained in this publication to the prefix http://www.mycobank.org/MB. The online version of this work is archived and available from the following digital repositories. PubMed Central, LOCKSS.

Results and Discussion
The relationship between the Talaromyces atroroseus sp. nov. and its close relatives were studied using multigene phylogenies, bason on ITS, RPB1 and β-tubulin sequences. The aligned datasets were 482, 888 and 374 bp long, respectively. The new species resolved in a clade together with other red pigment producing species such as T. albobiverticillius, and T. minioluteus. Talaromyces purpurogenus resolved in a distantly related clade (Figures 1-3). Within the red pigment producing clade, T. atroroseus resolved in a distinct clade separate from all the other species in all three phylogenies, confirming its unique nature.
Historically red pigment production caused a lot of confusion and resulted in numerous misidentifications in literature. This is especially true for Talaromyces purpurogenus, T. ruber, Penicillium sanguineum and P. crateriforme. Penicillium purpurogenum and P. rubrum were described by Stoll [20]. In their monograph Raper and Thom [21] also described P. purpurogenum and P. rubrum. No type material was available for P. rubrum therefore Raper and Thom [21] used two strains to describe P. rubrum, NRRL 1062 (= CBS 370.48) and NRRL 2120 (= CBS 133452). Pitt [22] synonymized P. rubrum, P. crateriforme and P. sanguineum with P. purpurogenum. The issues in the T. purpurogenus complex were clarified by Yilmaz et al. [10] who synonymized Penicillium crateriforme and P. sanguineum with T. purpurogenus and they described T. ruber as a distinct species. NRRL 1062 remained as T. ruber but NRRL 2120 (= CBS 133452) is a different species than T. ruber. Our results showed that NRRL 2120 is T. albobiverticillius. Raper and Thom [21] based the Penicillium purpurogenum description on NRRL 1061 (= CBS 364.48). However our results show that NRRL 1061 is a typical T. atroroseus strain.
Both Talaromyces purpurogenus and T. atroroseus are common in soil, indoor environments, and fruits. Talaromyces atroroseus resembles T. purpurogenus and T. albobiverticillius in producing red diffusible pigments, but differs from the latter two species by the production of glauconic acid, purpuride and ZG-1494α (Table 2 and Figure 4) and by the dull to dark green thick walled ellipsoidal conidia produced. Barton et al. [26,27] and Barton and Sutherland [28] reported glauconic acid from P. purpurogenum IMI 090178, which in the present study has been re-identified as T. atroroseus, while ZG-1494α was reported from P. rubrum CBS 238.95 [36], which is also a typical T. atroroseus. Talaromyces atroroseus, T. purpurogenus and T. albobiverticillius differ from T. ruber, T. amestolkiae and T. stollii by their production of red diffusible pigment. In Table 3 many red pigment producers identified as Penicillium species are listed, that may either be T. purpurogenus, T. ruber, T. albobiverticillius or T. atroroseus. The strains listed in Table 3 were not available for us, so their exact identity cannot be verified.   Many Talaromyces species produce striking diffusing red pigments, especially T. purpurogenus, T. atroroseus, T. albobiverticillius, T. minioluteus, and T. marneffei. These red pigments are typically composed of the azaphilone pigments ( Figure 5) monascorubrin, rubropunctatin, threonine derivative of rubropunctatin, monascorubramine, PP-R (= 7-(2hydroxyethyl)-monascorubramine), rubropunctamine, Nglutarylrubropunctamine, and PP-V [8,9,43,44,61]. The same family of azaphilones are also known from red rice, where different species of Monascus have grown [1,2]. These red pigments are of interest for the industry as they are stable and non-toxic and can be used as food colorants [62]. The azaphilone pigments can react with amino acids, hence their name, and give intense dark red colours. In addition some of these species produce yellow azaphilone pigments, such as monascin, ankaflavin, monascusone A and B, xanthomonascin A, and another series of yellow mitrorubrin azaphilones: mitorubrin, mitorubrinol, mitorubrinol acetate, mitorubrinic acid, and many other related compounds [5]. Many of these pigments have been reported from or found in T. atroroseus in this study (Table 2 and Table 4). The potential for pigment production has in this study only been investigated in small scale on solid media; however, T. atroroseus also produce pigments in liquid cultures under the right conditions [8,46]. The  potential for up scaling the production of red pigments needs to be investigated thoroughly. Even though sequence variations were observed for Talaromyces albobiverticillius strains, morphologically they were similar. Two strains used for the original description of T. albobiverticillius were received from Dr. Sung-Yuan Hsieh [63]. These included the type strain CBS 133440 T and CBS 133441. These strains were isolated from soil in Taiwan and produce white conidial masses and intense soluble red pigment on various media ( Figure 6). However, other freshly isolated T. albobiverticillius strains produce densely sporulating colonies and do not show any stability for red pigment production. Some of the Talaromyces albobiverticillius strains did not produce any soluble pigment such as CBS 133444 and CBS 133448. Strains that did produce red pigments include CBS 113168, and CBS 133452. On MEA only the degraded or mutated  Figure 5). Two strains of T. albobiverticillius (CBS 133440 T and CBS 133441) have globose to subglobose, smooth conidia; however, the remaining strains produce ellipsoid to fusiform smooth conidia ( Figure 5). Even though two clades were observed in the phylogenies there are no concordance between observed clades and morphological characters as discussed above. As such, they are considered here as representing one species. Raper and Thom [21] mentioned a number of colour mutations they observed in strains of P. citrinum and P. chrysogenum. They stated that colour mutations are encountered as the most common and conspicuous types of mutations, especially considering mature conidia. Mutations can often be observed when a strain loses its green pigment in its conidia, resulting in a white or tanned colour. Colour mutants are regularly encountered among the strains which were exposed to artificial stimulations such as ultra-violet, X-ray radiations and neutron bombardment [21].
Talaromyces atroroseus is considered as the optimal producer of industrially important yellow and red soluble pigments. Another option as a suitable producer of red soluble azaphilone pigments is T. albobiverticillius. However T. albobiverticillius produces soluble red pigment only in some strains. We speculate that the mitorubrins produced by Talaromyces atroroseus are of the (-)-form, as they have been shown to be that for the closely related Talaromyces purpurogenus (at that time identified as Penicillium rubrum) [64,65]. However, Natsume et al. [66] and Suzuki et al. [67] found both (+) and (-)-forms in the genus Talaromyces, while mitorubrins in Hypoxylon and other related genera are of the (+)-form [68][69][70]. Although T. purpurogenus is another good producer of diffusible red azaphilone pigments, this species also produce a series of mycotoxins, such as rubratoxin A and B and luteoskyrin in addition to extrolites that may be toxic if injected intraperitoneally (spiculisporic acid) [71] or in the veins of cats (rugulovasine A and B) [72,73]. Talaromyces purpurogenus can thus not be recommended for industrial production for red pigments.

Conclusion
Talaromyces atroroseus is a new species that produce large amounts of red pigments that can be potentially used for colouring foods, as it does not produce any known mycotoxins. Certain strains of T. albobiverticillius may also be used for these purposes. Table S1. Table S1 contains the extrolites searched for by ultra high performance-liquid chromatography-diode array detection-high resolution mass spectrometric detection (UHPLC-DAD-HRMS) the fungal extracts analysed. The table also includes data on the available standards used in the study.