Phylogeny of Leptographium Qinlingensis Cytochrome P450 Genes and Their Expression When Grown on Different Media or Treated With Terpenoids


 Leptographium qinlingensis is a fungal associate of the Chinese white pine beetle (Dendroctonus armandi) and a pathogen of the Chinese white pine (Pinus armandi) that must overcome the terpenoid oleoresin defences of host trees. We identified and phylogenetically analysed the cytochrome P450 (CYP) genes in the transcriptome of L. qinlingensis. Through analyses of the growth rates on different nutritional media and inhibition by terpenoids, the expression profiles of six CYPs in the mycelium of L. qinlingensis grown on different media or treated with terpenoids were determined. The CYP evolution predicted that most of the CYPs occurred in a putative common ancestor shared between L. qinlingensis and G. clavigera. This fungus is symbiotic with D. armandi and has more similarity with G. clavigera, which can retrieve nutrition from pine wood and utilize monoterpenes as the sole carbon source. Some CYP genes might be involved in the metabolism of fatty acids and detoxification of terpenes and phenolics, as observed in other blue-stained fungi, which also indicates the pathogenic properties of L. qinlingensis in Chinese white pine.


Introduction
Pathogens overcome the effects of terpenoids produced by conifers by active detoxi cation mechanisms. The ascomycete Leptographium qinlingensis is an active participant in the large-scale death of Pinus armandi, and it is associated with the Chinese white pine beetle (Dendroctonus armandi) ( The constitutive and induced defensive system of P. armandi consists of a multitude of monoterpenes, sesquiterpenes and diterpenes (Chen et al. 2006) and can stop or delay beetles from entering trees (Erbilgin et al. 2003). Moreover, monoterpenes present in the resin, including pinene, limonene and carene, can injure or kill beetles and inhibit fungal growth (Reid and Purcell 2011;Dai et al. 2015aDai et al. , 2015b. Previous studies have shown that symbiotic fungi destroy bleeding cells, block resin canals in host trees and kill epithelial cells, thereby resulting in disorders of the nutrient and water metabolism of the host (Chen and Tang 2002). The nitrogen concentrations in phloem infected with beetle-associated fungi were increased compared to those in uninfested phloem (Ayres et al. 2000). Associated fungi have been shown to provide nutritional support to bark beetles ( CYPs are also involved in the de novo synthesis of secondary metabolites. In Grosmannia clavigera, which is a pathogen of pines associated with Dendroctonus ponderosae, CYP65BJ1 is located in a secondary metabolite biosynthetic gene cluster and highly upregulated after treatment with monoterpenes that may produce aromatic polyketides (Lah et al. 2013). In pathogenic fungi, these compounds (e.g., a atoxin, fumonsin, trichothecene, gliotoxin) are often toxic to the host species and represent important virulence factors (Proctor et al. 2003;Yu et al. 2004; Yu and Keller 2005;Balibar and Walsh 2006;Kimura et al. 2007). Moreover, three phytotoxins (6methoxymethyleugenin, maculosin and cerevisterol) of P. armandi seedlings are synthesized by L. qinlingensis (Li et al. 2012).
Ophiostoma piceae is a wood-staining fungus that grows on a mixture of monoterpenes and diterpenes, although compared with G. clavigera, it cannot utilize monoterpenes as a carbon source (DiGuistini et al. 2011 ;Haridas et al. 2013). The wood of trees, logs and lumber has a high carbon/nitrogen ratio (Zabel and Morrell 1992). Compared with O. piceae, which grows more e ciently in drier pine wood, G. clavigera colonizes healthy or stressed living pine trees and can manage the high concentrations of defence chemicals produced by its pine host. Thus, O. piceae has slower growth rates than G. clavigera on rich media and wood Haridas et al. 2013). O. piceae and G. clavigera can grow on a variety of sugars (mannose, maltose and starch, a stored tree nutrient) and can acquire additional sugars by degrading wood hemicelluloses (Zabel and Morrell 1992;Fischer and Holl 1992;Fleet et al. 2001;Schirp et al. 2003). However, triglycerides and fatty acids can occasionally be used as carbon sources, which are ultimately processed through ß-oxidation and glycolysis pathways ).
Treatment with a terpenoid blend or pine phloem extract for associated fungi of bark beetles (D. ponderosae and D. armandi) always induces speci c CYPs (Lah et al. 2013;Dai et al. 2015b). Thus, cytochrome P450 enzymes that are highly induced by terpenes and metabolize or utilize monoterpenes are considered the major mechanisms that enable fungal resistance to monoterpenes (Lah et al. 2013;Wang et al. 2014). The nutrition of culture media could have an effect on the tolerance of G. clavigera to terpenes (Kligun et al. 2017).
In the work reported here, we identi ed and phylogenetically analysed CYPs in the transcriptome of L. qinlingensis. Analyses of the growth rates on different nutrition media, inhibition of growth by terpenoids and expression pro les of six CYPs in the mycelium of L. qinlingensis grown on different media or treated with terpenoids indicated that CYPs may detoxify pine defence compounds and could be in uenced by different nitrogen/carbon sources.

Strain and Growth Conditions
Leptographium qinlingensis (NCBI Taxonomy ID: 717526) was deposited at the College of Forestry, Northwest A&F University (Yangling, China).
Leptographium qinlingensis was grown on an MEA medium containing 1% Oxoid Malt Extract Agar and 1.5% Agar Technical (Oxoid Ltd., Basingstoke, Hampshire, UK) and topped with cellophane, and the pH was adjusted to 5 ~ 6.

Fungal growth under different nutrition
We characterized the effect of different nutrients on the growth rate of L. qinlingensis. The fungal strain was acclimatized at room temperature for 1 week on 25 mL MEA media following long-term storage at 4°C. According to the treatment for the mountain pine beetle-fungal symbiont Grosmannia clavigera (DiGuistini et al. 2011), mycelial plugs were transferred to a new Petri dish containing 25 mL of six different media [wood (W): 10 g/plate Chinese white pine sawdust; 1.5% granulated agar; starch (S); organic nitrogen (ON); inorganic nitrogen (IN); olive oil (OO); Chinese white pine methanol extract (CWPE): complete medium (0.17% YNB, 1.5% granulated agar, 1% maltose, 0.1% PHP, 0.3% asparagine) with 200 µl of the crude Chinese white pine methanol extract (Dai et al., 2015).
All plates were incubated at 28°C in the dark, and growth (in cm) was measured every 4 days in four directions and averaged until the strain brought the fungus to the edge of the plate. For the six different nutrition media, the growth rates were obtained by calculating the area of the colony. To assess whether different parameters affect the growth rate, we performed curve tting with a logistic equation [Y = A/(1 + B·e − kt ), where Y is the size of the colony (cm 2 ) and t is the culture time] using SPSS software (IBM SPSS Statistics, Chicago, IL, USA).

Identi cation of Leptographium qinlingensis P450s
Total RNA was isolated from mycelia grown on MEA medium for 7 days according to the protocol supplied with the E.Z.N.A.™ Fungal RNA Kit (Omega Bio-Tek, Norcross, GA, USA), and its integrity was assessed on 1% agarose gels and quanti ed by spectrophotometry with a NanoDrop 2000 (Thermo Scienti c, Pittsburgh, PA, USA). The purity was estimated by the A260/A280 equation (µg/mL = A260 × dilution factor × 40).
Samples were shipped on dry ice to Annoroad Gene Technology Co., Ltd. (Beijing, China) for paired-end sequencing. During the QC steps, an Agilent 2100 Bioanalyser and ABI StepOnePlus Real-Time PCR System were used for quanti cation and quali cation of the sample library. Finally, the library was sequenced using an Illumina HiSeq™ 2000 system. Raw data were processed with Perl scripts to ensure the quality of the data used in further analyses. For paired-end sequencing data, both reads were ltered out if any reads of the paired-end reads were adaptor-polluted.
The reads were assembled using Trinity (Grabherr et al. 2011), and unigene sequences were identi ed as candidate coding regions with TransDecoder to nd an open reading frame (ORF).
Trinotate was used to perform the functional annotation of unigenes and ORFs. The functional annotation included homology searches of known sequence data (BLAST), protein domain identi cation (PFAM), protein signal peptide and transmembrane domain prediction (SignalP), and comparison to current annotation databases, namely, the UniProt (Universal Protein), eggNOG (evolutionary genealogy of genes: Non-supervised Orthologous Groups) and GO (Gene Ontology) pathway databases. Protein function information could be predicted from the annotation of the most similar proteins in those databases.
To identify all of the unique P450 transcripts in the hybrid assembly, we assessed these unigenes and translated ORFs against the BLASTx, BLASTp, PFAM, and eggNOG (evolutionary genealogy of genes: Non-supervised Orthologous Groups) databases (e-value < 0.00001) to identify potential P450 sequences. The remaining unigenes were identi ed as potential P450 genes in L. qinlingensis (Table S1).
We downloaded the P450 protein sequences from Grosmannia clavigera kw1407 (53) A pair of primers for 6 annotated P450 sequences was designed to screen the putative P450 genes (Table S2). PCR ampli cations were performed in a C1000 thermocycler (Bio-Rad, Hercules, CA, USA). P450 genes were ampli ed under the indicated conditions in 20 µL reactions containing 1 µl cDNA, 0.25 µM of each primer and 1× EcoTaq PCR SuperMix (TransGen Biotech, Beijing, China). An initial 5 min step at 94°C was followed by 30 cycles of 30 s at 94°C, 30 s at Tm and 30 s at 72°C, with a nal extension for 10 min at 72°C.
The PCR products were visualized on 1% agarose gels stained with 1× DuRed and compared with a 2K plus DNA marker (TransGen Biotech, Beijing, China). Amplicons were puri ed, and the reaction product was cloned using the pMD™ 18-T Vector (TaKaRa, Dalian, China). Cloning reactions were transformed into DH5α chemically competent Escherichia coli cells, and a total of 5 clones with inserts were sequenced directly by GenScript USA Inc. The sequences were manually edited with DNAMAN to obtain the insert sequences. Blastx searches of partial-length sequences were performed against the NCBI database.
Information on the L. qinlingensis CYP65 genes was determined based on corresponding genes from Magnaporthe oryzae, N. crassa, Sordaria macrospora, Penicillium marneffei and Talaromyces stipitatus from the NCBI, and information on the CYP56BJ gene was determined based on corresponding genes from the genus Grosmannia (G. clavigera, G. aureum, G. penicillata) as well as Leptographium longiclavatum and L. terebrantis (Lah et al. 2013), and these data were used in the phylogenetic analyses.

Real-Time Fluorescent Quantitative PCR
We generated and analysed transcript-level data from two sets of growth conditions. For the rst set of conditions, mycelia were generated from a suspension of 5 × 105 spores spread on cellophane on the surface of six different nutrition media as above.
Total RNA isolation of the fungi was performed as described above. cDNA synthesis was performed using the protocol described in the FastQuant RT Kit (with gDNase) (Tiangen Biotech Co., Beijing, China) using 2 µg total RNA in a 20 µl nal reaction volume. The cDNA synthesis program was as follows: 42°C for 15 min and 95°C for 3 min. The cDNA was stored at -20°C.
For six P450 genes and the reference gene EF (Dai et al. 2015b), speci c primers were designed using Primer Premier 5.0 (Table S2). To estimate the qPCR e ciency and validate the primers for each gene, a linear regression analysis was performed between the mean values of the quanti cation cycles (Cq) of different dilutions (1.0, 10 − 1 , 10 − 2 , 10 − 3 , and 10 − 4 ) of cDNAs and the initial concentration. These dilutions were made from a cDNA pool, and 2 µl of each dilution was used as a qPCR template. PCR was performed three times for each gene, and its e ciency was estimated with the Eq. (10 − 1/slope − 1) × 100, where the E value and R 2 are shown in Table S2. Moreover, a melting curve reaction was performed to evaluate their speci city.

Fungal growth in different culture media
We compared the growth of L. qinlingensis on six culture media with different carbon and nitrogen sources at 28°C (Fig. 1). The growth of L. qinlingensis on media with organic nitrogen, inorganic nitrogen and wood was fast, and the growth on media with starch and Chinese white pine methanol extract was slow. However, L. qinlingensis showed the lowest growth rate on the medium with olive oil.
The logistic curve t the growth curve of L. qinlingensis on the six culture media (R 2 > 0.97). According to the logistic curve tting of the growth curve, the growth in exion day of L. qinlingensis growth on organic nitrogen, inorganic nitrogen and wood media occurred after approximately 8 d and the growth in exion day on starch medium or complete medium with Chinese white pine methanol extract occurred after approximately 12 d (Table 1). However, the growth in exion day occurred after over 20 d for the medium with olive oil as the only carbon source (Table 1). Logistic equation: Y = A/(1 + B·e − kt ), Y means size of the colony (cm 2 ), t means culture time, A is maximum size of the colony, B is parameter, and k is maximum of relative growth rate.

Transcriptome assembly and annotation
Among the predicted ORFs, 17,040 corresponded to our acceptance criteria (see Methods), and 10,735 of these ORFs were at least 200 amino acids long. Within the annotated transcriptome of L. qinlingensis, we identi ed genes and gene families for secondary metabolite processing and cytochrome P450. We also identi ed homologous O. piceae, G. clavigera and N. crassa proteins based on reciprocal best BLAST hits. Some of the major gene families for secondary metabolite processing in L. qinlingensis are shown in Table 3.  Table S1.

CYPome of L. qinlingensis
We identi ed 56 cytochrome P450 (CYP) genes in the 17,040 ORFs of the L. qinlingensis transcriptome (Table S1). Thirty-nine CYP genes with ORFs at least 200 amino acids long were used for phylogenetic analyses with the CYPome from G. clavigera, O. piceae, S. schenckii and N. crassa. We found more examples of recognizable orthologues of P450s for G. clavigera in our comparison than in the other fungal species (Fig. 2). According to the nomenclature of 54 CYP genes of G. clavigera, the CYP genes of L. qinlingensis represent 18 different CYP families.
Six CYP genes were ampli ed and sequenced for accurate sequence information using primers designed according to transcriptome annotation. The sequences were submitted to the Cytochrome P450 Nomenclature Committee (Nelson 2009) as CYP61A1, CYP582C, CYP537D6, CYP65BJ4, CYP578E and CYP52Z4. The speci c sequences of CYPs were submitted to GenBank under accession numbers MT178256-MT178261. The amino acid sequence had the highest identity with G. clavigera except for CYP52Z4, which was between partial-length sequences with respect to the matched GenBank sequences (Table 4). CYP52Z4 had high identity with the n-alkane-inducible cytochrome p450 protein of Pochonia chlamydosporia (Table 4).  (Altschul et al. 1990).
One gene was classi ed into the CYP65B family, whose members in other fungi were shown to be involved in terpene bioconversions (Kimura et al. 2007). The phylogenetic analysis with a maximum likelihood tree (model: T92 + G + I, -lnL = 5724.207, G = 1.69, I = 0.19) suggested that CYP65BJ4 of L. qinlingensis was conserved within the CYP65BJ1 subclade of the Grosmannia genus (Fig. 3).

RT-qPCR
To determine whether the P450 genes were involved in the utilization of different nutrition (carbon and nitrogen) sources, we analysed six CYP gene expression pro les of L. qinlingensis grown on the following culture media: W (wood), S (starch), ON (organic nitrogen), IN (inorganic nitrogen), OO (olive oil) and CWPE (Chinese white pine methanol extract). Statistically signi cant differences were found among these culture media for six CYPs ( Table 5). In mycelia grown on complete medium with Chinese white pine methanol extract (CWPE), six CYPs were signi cantly overexpressed (Fig. 4). However, the expression of CYPs was signi cantly downregulated in mycelia grown on inorganic nitrogen medium (Fig. 4). Signi cant overexpression of CYP582C and CYP52Z4 was found in mycelia grown on the other three kinds of media (wood, starch and olive oil), and the medium with olive oil as the only carbon source signi cantly downregulated CYP61A1 (Fig. 4).  Fig. 4 with different letters.
To discover these six L. qinlingensis CYPs with a possible role in the detoxi cation of pine defence chemicals, we analysed the expression pro les of CYPs from mycelia grown on MEA medium treated with monoterpenes and turpentine at three different concentrations for 24 h. The transcription levels of most CYPs were signi cantly changed after exposure to the terpenoids (Table 6). CYP61A1 was only signi cantly downregulated after treatment with limonene at a 10% concentration (Fig. 5). The transcription level of CYP582C was signi cantly overexpressed after treatment with 3-carene and β-pinene at a 5% concentration but downregulated after treatment with limonene and turpentine at 20% (Fig. 5). For CYP537D6, signi cant overexpression was found only after treatment with 10% α-pinene and 5% β-pinene (Fig. 5). The transcription level of CYP65BJ4 was signi cantly downregulated after treatment with 10% and 20% β-pinene and turpentine. Treatment with 3-carene caused overexpression at 5% and 20% but downregulation at 10% for CYP65BJ4, although the opposite changes in expression were observed after treatment with limonene (Fig. 5). For CYP578E, signi cant overexpression was found after treatment with all terpenoids at almost all concentrations. Similar to CYP578E, the transcription level of CYP52Z4 was overexpressed after treatment with α-pinene, 3-carene and β-pinene but downregulated after treatment with limonene and turpentine at 20% (Fig. 5). However, L. qinlingensis cannot utilize olive oil as a carbon source well, which is possibly because this medium consists of fatty acids. Fatty acids can be used as a carbon source in G. clavigera, although their utilization might require processing via ß-oxidation and glycolysis pathways  We studied the nutrition utilization of L. qinlingensis grown on host tree wood, multiple sugars and fatty acids. The host chemical compound tolerance of L. qinlingensis was determined with the MIC test, and the induction of CYP genes by monoterpenes and pine extract was identi ed. This fungus is symbiotic with D. armandi and has considerable similarity with G. clavigera, which can retrieve nutrition from pine wood and utilize monoterpenes as a  A maximum likelihood phylogeny tree of cytochrome P450s. Circular phylogram of the 39 predicted P450 protein sequences from L. qinlingensis, along with the P450s identi ed from the genome sequences of the Grosmannia clavigera kw1407, Ophiostoma piceae UAMH 11346, Sporothrix schenckii 1099-18 and Neurospora crassa OR74A. Some branch lengths are longer than might be expected due to their partial sequence length.  Quantitative expression of the six P450 genes (mean ± SE) in L. qinlingensis following treatment with different terpenoids. CYPs expressions were normalized with respect to EF1.