Open Access

Genome sequence of the filamentous soil fungus Chaetomium cochliodes reveals abundance of genes for heme enzymes from all peroxidase and catalase superfamilies

  • Marcel Zámocký1, 2Email author,
  • Hakim Tafer3,
  • Katarína Chovanová2,
  • Ksenija Lopandic3,
  • Anna Kamlárová2 and
  • Christian Obinger1
BMC Genomics201617:763

https://doi.org/10.1186/s12864-016-3111-6

Received: 16 August 2016

Accepted: 22 September 2016

Published: 29 September 2016

Abstract

Background

The ascomycetous family Chaetomiaceae (class Sordariomycetes) includes numerous soilborn, saprophytic, endophytic and pathogenic fungi which can adapt to various growth conditions and living niches by providing a broad armory of oxidative and antioxidant enzymes.

Results

We release the 34.7 Mbp draft genome of Chaetomium cochliodes CCM F-232 consisting of 6036 contigs with an average size of 5756 bp and reconstructed its phylogeny. We show that this filamentous fungus is closely related but not identical to Chaetomium globosum and Chaetomium elatum. We screened and critically analysed this genome for open reading frames coding for essential antioxidant enzymes. It is demonstrated that the genome of C. cochliodes contains genes encoding putative enzymes from all four known heme peroxidase superfamilies including bifunctional catalase-peroxidase (KatG), cytochrome c peroxidase (CcP), manganese peroxidase, two paralogs of hybrid B peroxidases (HyBpox), cyclooxygenase, linoleate diol synthase, dye-decolorizing peroxidase (DyP) of type B and three paralogs of heme thiolate peroxidases. Both KatG and DyP-type B are shown to be introduced into ascomycetes genomes by horizontal gene transfer from various bacteria. In addition, two putative large subunit secretory and two small-subunit typical catalases are found in C. cochliodes. We support our genomic findings with quantitative transcription analysis of nine peroxidase & catalase genes.

Conclusions

We delineate molecular phylogeny of five distinct gene superfamilies coding for essential heme oxidoreductases in Chaetomia and from the transcription analysis the role of this antioxidant enzymatic armory for the survival of a peculiar soil ascomycete in various harsh environments.

Keywords

Chaetomium cochliodes Peroxidase-catalase superfamily Peroxidase-cyclooxygenase superfamily Peroxidase-chlorite dismutase superfamily Peroxidase-peroxygenase superfamily Heme-catalase super family

Background

The ascomycetous family of Chaetomiaceae (class Sordariomycetes) includes numerous soilborn, saprotrophic, endophytic and pathogenic fungi that apparently exhibit a large flexibility in their adaptation to various growth conditions and living niches. In Mycobank (www.mycobank.org) currently up to 451 members of this abundant fungal family are registered but only from two representatives (i.e. Chaetomium thermophilum and Chaetomium globosum) the completely sequenced genomes are available. Analysis of the genome of C. thermophilum [1] mainly focused on the presence of genes coding for nucleoporins of high thermal stability, whereas the draft genome of Chaetomium globosum [2] was mainly asked for diverse genes coding cellulolytic pathways.

The filamentous fungus Chaetomium cochliodes was long considered to be a variant of Chaetomium globosum (cf. the NCBI taxonomy database at www.ncbi.nlm.nih.gov/taxonomy) but already in very early studies e.g. [3] it was shown that C. cochliodes produces the antibiotic chaetomin which was shown to be highly active mainly against Gram-positive bacteria. Additionally, studies from our laboratory revealed differences between C. globosum and C. cochliodes in the primary sequence and expression profile of peroxisomal catalase-peroxidases [4]. These findings – together with the fact that peroxidases participate in diverse fungal secondary metabolism pathways [59] – prompted us to sequence the entire genome of Chaetomium cochliodes strain CCM-F232 for detailed comparative studies.

Here we release the draft genome of C. cochliodes, reconstruct its phylogeny and analyse the occurrence of abundant genes coding for heme containing peroxidases and catalases with respect to the recently described four distinct heme peroxidase superfamilies [10] and the heme catalase super family [11]. Interestingly, representatives from all five (super)families were found including putative bifunctional catalase-peroxidase, cytochrome c peroxidase, hybrid B peroxidases, cyclooxygenase-like enzymes, dye-decolorizing peroxidases, heme thiolate peroxidases as well as large- and small-subunit monofunctional catalases. The occurrence of this large number and variability of genes encoding heme hydroperoxidases in C. cochliodes is discussed in comparison with related fungal genomes. We support our genomic findings with a first round of a quantitative expression analysis of selected genes from all mentioned superfamilies involved in the catabolism of H2O2.

Methods

Source and cultivation of Chaetomium cochliodes and isolation of genomic DNA

Chaetomium cochliodes CCM F-232 was obtained from Czech Collection of Microorganisms at the Masaryk University, Faculty of Natural Sciences in Brno, Czech Republic. The composition of the incubation medium and the growth conditions were the same as described previously [4].

Genomic DNA from 100 mg of frozen fungal mycelium was isolated with the method of Carlson [12] by using 2 % CTAB in a modification suitable for genome sequencing described in [13]. Finally, extracted DNA was completely dissolved in TE buffer (10 mM Tris–HCl 1 mM EDTA, pH 8.0) to a final volume of 100 μL. The concentration of obtained sample was measured in Nanodrop 2000 (Thermo Scientific, Waltham, MA, USA).

Library preparation for DNA sequencing

Approximately 1 μg of high quality genomic DNA was fragmented in 50 μl Low TE buffer (10 mM Tris pH 8.0, 0.1 mM EDTA) by BioRuptor UCD-200 sonication system (Life Technologies, Carlsbad, CA, USA) to obtain a population of ~190 bp long fragments. The length and the quantity of generated fragments were assessed by Bioanalyzer chip technology (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instructions. The protocol of the Library Builder™ System (Life Technologies, Carlsbad, CA, USA) was used for adaptor ligation, nick repair and fragment purification. The selection of 270 bp long fragments was conducted by the Pippin Prep instrument (Sage Science, Beverly, MA, USA) according to the manufacturer’s instructions. Library quantification was carried out using the TaqMan qPCR protocol of Life Technologies.

Genomic DNA sequencing and ORF prediction

Whole genome sequencing was carried out using the Ion Proton Technology (including the Ion AmpliSeq library preparation kit, Template OT2 200 kit, Ion PI sequencing 200 kit, and the Ion PI chip kit version 2; Life Technologies, Carlsbad, CA), according to the instructions of the manufacturer. A total of 34.746 Mbp, with a median read length of 180 bp, were assembled into a draft genome containing 6036 contigs (N 50, 14,381). The genome assembly was performed with Newbler 2.9. Genome coverage of this sequencing was 316 x. The entire genome shotgun project was deposited at GenBank under accession LSBY00000000, BioProject PRJNA309375, BioSample SAMN04432217. For comparative genomic analyses of Chaetomium cochliodes genes Ensembl Fungi (http://fungi.ensembl.org/index.html) was used.

For gene prediction in sequenced C. cochliodes contigs, HMM-based methods FGENESH and FGENESH+ located at www.softberry.com [14] trained for closely related C. globosum & C. thermophilum were used. For all peroxidase and catalase genes they were also curated manually.

Reconstruction of fungal phylogeny

Selected DNA sequence spanning the region from the 3′ end of the 18S rDNA, the complete ITS1, 5.8S rDNA, ITS2 and the 5′ end of the 28S rDNA from corresponding C. cochliodes contigs was aligned with 33 related sequences from Ascomycetes in exactly the same region obtained from GenBank (Table 1). This DNA alignment was performed with the Muscle program [15] implemented in MEGA 6 package with its default parameters and 100 iterations. For subsequent phylogeny reconstruction MEGA 6 program suite [16] was applied on this 2474 bp long DNA alignment containing the typical fungal barcode motif [17]. Maximum likelihood method with 1000 bootstrap replications and general time reversed substitution model were applied. Further, uniform rates of substitutions with invariant sites and involvement of all aligned sites with nearest-neighbour interchange and very strong branch swap filter were selected as optimised parameters. The resulting tree was rendered with Tree Explorer of the same MEGA package. For additional verification, the same 2474 bp long DNA alignment was subjected to phylogeny reconstruction using MrBayes 3.2 [18]. Majority consensus tree was obtained from all credible topologies sampled by MrBayes over 200,000 generations (with a standard deviation of split frequencies below 0.01) by using the same GTR substitution model with gamma distributed rate variation across sites and a proportion of invariable sites.
Table 1

List of all DNA sequences with their GenBank accession numbers used for phylogeny reconstruction in the region 18S, ITS1, 5.8S, ITS2, 28S-rDNA

Abbrev.

Fungus

Taxonom. family

GB accession #

[bp] used for phyl.

Ccoch

Chaetomium cochliodes

Chaetomiaceae

KT895345

2217

Celat

Chaetomium elatum

Chaetomiaceae

M83257

2211

Cg1

Chaetomium globosum CBS148.15

Chaetomiaceae

NT_166001

2245

Cg2

Chaetomium globosum (endophyt)

Chaetomiaceae

DQ234257

2219

Cg3

Chaetomium globosum isol. W7

Chaetomiaceae

JQ686920

2219

Cthe

Chaetomium thermophilum

Chaetomiaceae

GCA_000221225

2237

Coacu1

Colletotrichum acutatum 1

Glomerellaceae

AJ301905

2227

Coacu2

Colletotrichum acutatum 2

Glomerellaceae

AJ301906

2225

Cocir

Colletotrichum circinans

Glomerellaceae

AJ301955

2216

Cococ

Colletotrichum coccodes

Glomerellaceae

AJ301957

2218

Codem

Colletotrichum dematium

Glomerellaceae

AJ301954

2220

Colup

Colletotrichum lupini

Glomerellaceae

AJ301959

2200

Cotri

Colletotrichum trifolii

Glomerellaceae

AJ301942

2231

Cotru

Colletotricum truncatum

Glomerellaceae

AJ301937

2213

Fgram

Fusarium graminearum

Nectriaceae

NC_026477

2188

Gcin

Glomerella cingulata

Glomerellaceae

AJ301952

2198

Hgri

Humicola grisea

Chaetomiaceae

AY706334

2202

Lsak

Lecanicillium saksenae

Cordycipitaceae

AB360363

2236

Masp

Madurella sp. TMMU3956

Sordariaceae

EU815932

2271

Mhin

Myceliophthora hinnulea

Chaetomiaceae

JQ067909

2099

Mthe

Myceliophthora thermophila

Chaetomiaceae

NC_016478

2217

Mgram

Mycosphaerella graminicola

Mycosphaerellaceae

NC_018212

2195

Matr

Myrothecium atroviride

Stachybotriaceae

AJ302002

2223

Mcin1

Myrothecium cinctum 1

Stachybotriaceae

AJ301996

2204

Mcin2

Myrothecium cinctum 2

Stachybotriaceae

AJ302004

2202

Mver

Myrothecium verrucaria

Stachybotriaceae

AJ301999

2222

Ncr

Neurospora crassa

Sordariaceae

FJ360521

2230

Pan

Podospora anserina

Lasiophaeriaceae

FO904938

2196

Sfim

Sordaria fimicola

Sordariaceae

X69851

2256

Taus

Thielavia australiensis

Chaetomiaceae

JQ067908

2160

Tter

Thielavia terrestris

Chaetomiaceae

NC_016459

2233

Tasp

Trichocladium asperum

Chaetomiaceae

AY706336

2202

Tatr

Trichoderma atroviride

Hypocreacea

NW_014013638

2251

Vcil

Volutella ciliata

Nectriaceae

AJ301967

2214

Reconstruction of molecular phylogeny of protein superfamilies

Selected protein sequences translated from C. cochliodes contigs (Table 2B) and similar protein sequences coding for various peroxidases and catalases (i.e. hydroperoxidases deposited at PeroxiBase http://peroxibase.toulouse.inra.fr with direct links to GenBank & UniProt) were aligned with the Muscle program [15] using default parameters and 100 iterations. Obtained alignments were inspected and ambiguously aligned regions were excluded from further analysis. Resulting alignments were subjected to protein phylogeny reconstruction using MEGA 6 [16] with optimized parameters according to lowest Bayesian information criterion scores (Additional file 1: Table S1). Maximum likelihood method with 100 bootstraps was chosen using the best substitution model for each alignment (WAG in three cases and LG in two cases cf. Additional file 1: Table S1 for details), gamma distribution of rates (four categories) and the presence of invariant sites. Nearest-neighbour interchange was used as heuristic method and very strong branch swap filter was applied. The same protein alignments were subjected to phylogeny reconstruction using MrBayes 3.2 [18]. Majority consensus tree was obtained from all credible topologies sampled by MrBayes over 500,000 generations (with a standard deviation of split frequencies below 0.10) by using the same substitution model as in MEGA. Resulting trees were rendered with FigTree graphic suite available at http://tree.bio.ed.ac.uk/software/figtree as cladograms with transformed branches.
Table 2

List of potentially all genes coding for enzymes involved in H2O2 metabolism in contigs of C. cochliodes genome

Gene name

In contig #

Seq. identity*

Closest neighbour**

# Introns

Gene-superfamily relations

A) genes coding for enzymes producing H2O2

CcochCuZnSOD

0613

98 %

CgCuZnSOD

4

Copper/zinc superoxide dismutase superfamily (SODC)

CcochDAAO

0702

85 %

Mth_G2QLH3

3

Flavin D-amino acid oxidase (peroxisomal)

CcochFeMnSOD1

0353

91 %

TthFeMnSOD

2

Iron/manganese superoxide dismutase superfamily

CcochFeMnSOD2

1984 & 0805

93 %

CgFeMnSOD1

3

Iron/manganese superoxide dismutase superfamily

CcochFeMnSOD3

0879

94 %

CgFeMnSOD2

1

Iron/manganese superoxide dismutase superfamily

CcochFlOx1

0556

55 %

Colgra_E3Q5F0

1

GMC superfamily (flavin oxidases)

CcochGlOx1

0600

53 %

Scap_A0A084G823

7

GMC superfamily (flavin oxidases): glucose oxidase

CcochNOx1

0029

93 %

CgNox2

2

NADPH oxidase

B) genes coding for enzymes degrading H2O2

CcochkatG1

0012

93 %

CgkatG1

none

peroxidase-catalase superfamily: bifunctional catalase-peroxidase

Ccochccp

0676

95 %

Cgccp

2

peroxidase-catalase superfamily: cytochrome c peroxidase

Ccochpox2a

3115 & 3438

68 %

Cthepox2a

1

peroxidase-catalase superfamily: Family II prob. manganese-dependent

CcochhyBpox1

3712 & 3350

100 %

CghyBpox1

none

peroxidase-catalase superfamily: hybrid B peroxidase

CcochhyBpox2

0794

93 %

CghyBpox2

1

peroxidase-catalase superfamily: hybrid B peroxidase

Ccochcyox1

2133 & 0418

83 %

CgCyOx1

3

peroxidase-cyclooxygenase superfamily: cyclooxygenase

Ccochlds

1074 & 4463

91 %

Cglds1

5

peroxidase-cyclooxygenase superfamily: linoleate diol synthase

Ccochdyprx

0391

89 %

Cgdyprx

none

peroxidase-dismutase superfamily: Dyp_B peroxidase (fusion w. PFL)

Ccochhtp1

1650

92 %

Cghtp1

3

peroxidase-peroxygenase superfamily: heme-thiolate peroxidase

Ccochhtp2

2302

95 %

Cghtp3

3

peroxidase-peroxygenase superfamily: heme-thiolate peroxidase

Ccochhtp3

1018

85 %

Cghtp4

2

peroxidase-peroxygenase superfamily: heme-thiolate peroxidase

Ccochvcpo

0469 & 1446

93 %

Cgvcpo

3

non heme peroxidases: vanadium haloperoxidase

Ccochgpx

0466

84 %

Mthgpx

1

non-metal peroxidases: glutathione peroxidase

Ccoch1cysprx

1586

96 %

Cg1cysprx

1

non metal peroxidases: 1-cysteine peroxiredoxin

Ccoch2cysprx

1595

99 %

Cg2cysprx

2

non-metal peroxidases: typical 2-cysteine peroxiredoxin

CcochprxII

0388 & 1977

95 %

CgprxII

1

non-metal peroxidases: atypical 2-cysteine peroxiredoxin

CcochkatA1

0438 & 2821

94 %

Cgkat1

2

heme catalase superfamily: large subunit heme catalase

CcochkatA2

1883 & 2899

87 %

Cgkat2

3

heme catalase superfamily: large subunit heme catalase

CcochkatB1

0511

86 %

Cgkat3

2

heme catalase superfamily: small subunit heme catalase

CcochkatB2

0351

67 %

SschkatE

2

heme catalase superfamily: small subunit heme catalase

* - With closest known phylogenetic neighbour

** - Abbreviations of peroxidase & catalase gene names are explained in Additional file 3: Table S2, Additional file 5: Table S4, Additional file 6: Table S3, Additional file 7: Table S5 and Additional file 8: Table S6

Transcriptional analysis of genes involved in peroxide catabolism with RT-qPCR

To study the level of expression of genes involved in peroxide catabolism either non-induced C. cochliodes samples or samples induced in the early exponential phase of growth with 5 mM H2O2 or 5 mM PAA (final concentration, added only for last 30 min.) were used for total RNA isolation with RNeasy Plus Mini kit (Qiagen, Netherlands). Obtained RNA samples were directly subjected to RT-qPCR assays in AriaMx6 device (Agilent Technologies, Sancta Clara CA, USA) using the Brilliant III Ultra Fast SYBR Green Master Mix (also from Agilent Technologies) with specific primers for selected genes listed in Table 3.
Table 3

List of primers for C.cochliodes peroxidase & catalase genes

B

Primer description

Sequence in 5′ → 3′ direction

Tm [°C]

PCR prodct size [bp]

hyBpox1

CcochhyBpox1Fwd

CGAGAAAACAGATATTCTAGAAGCCA

60.1

116

CcochhyBpox1Crev

TTCTACCGGCACCTAAATTGTT

56.5

 

hyBpox2

CcochhyBpox2aFWD

GTTCATTTAGCAGGAGGTCAGG

60.3

119

CcochhyBpox2aREV

TGTCACTGCTCGAGTTAGCATT

58.4

cyox1

CcochCyox1bFWD

GCCTTCAAACTCCTCAACAAAG

58.4

117

CcochCyox1bREV

GTAGCCGTCATGGAGGTTGTAT

60.3

lds

CcochLDS3FWD

AACTTACACCATCTCCCGTGTC

58.4

127

CcochLDS3REV

GTCGTACTGAGCGTCGTTGTAA

60.3

dyPrx

CcochDyprxBfwd2

AAAGGAATGTCGAACCAAAAGA

54.7

135

CcochDyprxBrev1

GCCGAGAGTAAAATCTGGAATG

58.4

htp1

CcochHtp1fwd2

ATCTTCAACCAGACCATCTTCG

58.4

114

CcochHtp1rev2

GAACGACTTGGACTCGATCTG

59.8

katA2

CcochkatA2_IFWD

GAATCAACAAGACGCTTTGTGG

63.5

202

CcochkatA2_IREV

TAGGTGGGTTAGCAAGTGAGAG

63.3

katB1

CcochkatB2_2REV

TAAACACAAAGTCCTGGTTCCC

58.4

207

CcochkatB1_2REV

TGGAAAAGGCGCCGTAGTCG

61.4

katB2

CcochkatB2_1FWD

GGGGCGAGTTTGAGGTGACC

63.5

198

CcochkatB2_2REV

TAAACACAAAGTCCTGGTTCCC

58.4

Results and discussion

Overview of the sequenced genome of Chaetomium cochliodes CCM F-232

In total 6036 contigs were obtained from the genomic DNA of C. cochliodes strain CCM F-232 deposited at GenBank under accession LSBY00000000, BioProject PRJNA309375, BioSample SAMN04432217. 4141 of these contigs were larger than 500 bp. The genome size of the complete assembly was determined to be 34,745,808 bp. This value is very near to previously determined size of closely related C. globosum (updated to 34.9 Mb) [2]. The GC content of the entire genome of C. cochliodes was estimated to 55.95 % which is a slight difference to the corresponding value for C. globosum (55.6 %). The average size of C. cochliodes large genomic contigs (>500 bp) in this experiment was determined as 8256 bp, the N50 contig size was 14,381 bp and the largest assembled contig comprised 109,425 bp. As a quality control Phred quality scores were determined according to Illumina device: the portion of Q40+ bases was 34,112,976 (99.83 % of the whole genome sequence draft) whereas Q39 bases portion was only 59,430 (0.17 %). Prediction of all possible ORFs of C. cochliodes with Chaetomia-optimised FGENESH suite [14] led in both DNA strands to a total value of 10,103. This count is lower than the estimation for mesophilic C. globosum [2] but much higher than the estimation for C. thermophilum [1] or related thermophilic fungi. A brief comparison of three related fungal genomes is presented in Table 4. The average count of exons per predicted C. cochliodes gene was calculated as 3 with FGENESH.
Table 4

comparison of three related Chaetomia genomes

Organism

Reference

Genome size [bp]

Comparison with C.coch.

Predicted ORFs

C. cochliodes

this work

34,745,808

 

10.103

C. globosum

[2]

34,886,900

100.41 %

11.048

C. thermophilum

[1]

28,322,800

81.51 %

7.165

Phylogeny reconstruction in the 18S r DNA – ITS1 – 5.8S r DNA – ITS2 – 28S r DNA region

First, we were interested in the exact phylogenetic position of Chaetomium cochliodes. For this purpose we have reconstructed the DNA phylogeny of its 2217 bp region spanning the region from the 3′ end of the 18S rDNA, the complete ITS1, 5.8S rDNA, ITS2 and the 5′ end of the 28S rDNA containing the highly conserved locus described as universal fungal barcode [17]. Besides all corresponding DNA sequences for species of the Chaetomiaceae family currently available in GenBank, also sequences from related ascomycetous families were included in this reconstruction (Table 1). The DNA alignment used for the phylogeny reconstruction (Additional file 2: Figure S1) reveals clear differences (i.e. substitutions, insertions and deletions) in the sequence of C. cochliodes if compared with corresponding sequences of C. globosum in the entire region. The phylogenetic output presented in Fig. 1 (obtained by two independent methods) clearly segregates Chaetomium cochliodes from closely related C. elatum which is a root-colonizing fungus whose genome is not yet sequenced [19]. Both these fungi are separated from a sister clade represented by three different DNA sequences within this region coding for various C. globosum strains with a high statistical support. This figure clearly demonstrates that the thermophilic representatives (mainly C. thermophilum but also e.g. T. terrestris and M.thermophila) of the Chaetomiaceae family can be considered as basal lineages of the Chaetomia clade thus suggesting that mesophily has evolved only secondarily in this lineage. Our results correlate with the previous work on thermophilic fungi [20] and particularly on the thermostability of Chaetomiaceae [21] where C. cochliodes was not included at that time.
Fig. 1

Phylogenetic relationship among 34 Ascomycetes reconstructed from the conserved region spanning 18S-ITS1-5.8S-ITS2-28S rDNA genes. Maximum likelihood method from MEGA6 with 1000 bootstraps and MrBayes method over 200,000 generations were applied on the same DNA sequence alignment 2,474 bp long (Additional file 2: Figure S1). Bootstrap values above 50 & posterior probabilities are shown, respectively. Scale bare represents the frequency of ML substitutions

Putative heme peroxidases & catalases in Chaetomium cochliodes

Intracellular hydrogen peroxide is a by-product of various physiological pathways but, unique among all reactive oxygen species, it serves also as an important signalling molecule in apoptosis and ageing [22]. In filamentous fungi hydrogen peroxide was shown to be implicated in essential proliferation and differentiation processes [23]. Thus we have performed this genomic screening for all possible ORFs coding for a) enzymes supposed to release H2O2 during their reaction and b) two main types of enzymes involved in the catabolism of hydrogen peroxide in a novel genome of a soil Ascomycete. With TBLASTX method we could identify 8 genes for various oxidoreductases producing H2O2 (Table 2A) and up to 20 distinct genes belonging to various heme and non-heme peroxidase superfamilies as well as to the heme catalase superfamily. Overview on all these genes together with their introns composition is presented in Table 2B. All presented sequences are from contigs of the genome project deposited at GenBank under accession LSBY00000000, BioProject PRJNA309375, BioSample SAMN04432217. From Table 2 it is obvious that genes coding H2O2 degradation exhibit a higher diversity than genes coding H2O2-releasing enzymes. Detected genes for non-heme peroxidases include vanadium-containing haloperoxidase, glutathione peroxidase as well as 1-cysteine and 2-cystein peroxiredoxins. This work focuses further on genes coding for heme peroxidases.

As was presented recently, there are at least four heme peroxidase superfamilies and one heme catalase superfamily that arose independently during a convergent evolution. They differ in overall fold, active site architecture and enzymatic activities [10]. The following sections aim to discover all genes for representatives of all five superfamilies within the genome of C. cochliodes and to determine their exact phylogenetic positions. Heme peroxidases are found in all kingdoms of life and typically catalyse the one- and two-electron oxidation of a myriad of organic and inorganic substrates. In addition to the basal peroxidatic activity distinct families show pronounced catalase, cyclooxygenase, chlorite dismutase or peroxygenase activities.

Peroxidase-catalase superfamily

The peroxidase-catalase superfamily is currently the most abundant peroxidase superfamily in various gene and protein databases. It is comprised of three distinct families (Families I, II and III formerly known as classes) and hybrid peroxidases that represent transition forms (clades) between these families. Here we present an updated reconstruction of the phylogeny of this largest known heme peroxidase superfamily analysed previously [24, 25]. Our updated input included already 632 complete sequences and is presented in Fig. 2. We focus here on the phylogenetic positions of all representatives (ORFs) found in Chaetomia.
Fig. 2

Reconstructed phylogeny of the peroxidase-catalase superfamily with focus on newly sequenced Chaetomia ORFs. The complete tree from 632 full length sequences with 536 sites aligned is presented with collapsed branches that do not contain any Chaetomia sequences. Distinct subfamilies are labelled in different colours. C. cochliodes sequences are labelled red. Values in nodes represent bootstrap values above 50 (from maximum likelihood analysis) and posterior probabilities (from Mr. Bayes), respectively. Abbreviations of peroxidase names are listed in Additional file 3: Table S2. Abbreviations of taxa: Pb, Proteobacteria; As, Ascomycota; Ba, Basidiomycota; Chy, Chytridiomycota; St, Stramenopiles; Chl, Chlorophyta; Vi, Viridiplantae

Family I of the peroxidase-catalase superfamily typically contains catalase-peroxidases (KatG), ascorbate peroxidases and cytochrome c peroxidases (CcP) [24]. A HGT-event from Bacteroidetes to filamentous Ascomycetes was previously reported as a peculiarity of katG gene family evolution [26]. Circular tree of the whole superfamily clearly demonstrates that all katG1 genes from the Chaetomiaceae family (cf. Additional file 3: Table S2 for abbreviations) apparently are late descendants of this HGT event (Fig. 2 left upper part). Within the upper clades we observe a basal position of the thermophilic variants from which their mesophilic counterparts descended. However, a question remains whether only the coding region of katGs was transferred from bacteria to fungi or whether some neighbouring regions were also included in such a transfer? We demonstrate for the gene encoding KatG1 in C. cochliodes (i.e. CcochkatG1) that the regulatory elements located on 5′ and 3′ regions embedding the ORF are clearly of eukaryotic origin (Fig. 3). In the promoter region there is (besides the GC box) a typical regulatory sequence – the “CCAAT” box involved in eukaryotic oxidative stress response [27]. In the 3′ untranslated region the poly-A site for corresponding mRNA formation can be predicted with a high probability. Thus, we can conclude that a prokaryotic katG was inserted in the fungal genome but received a typical eukaryotic transcription regulation during later evolution. The main physiological role of KatG in C. cochliodes is most propable dismutation of metabolically-generated hydrogen peroxide to molecular oxygen and water, similar to typical (monofunctional) catalases (see below) [24, 26]. In addition to KatG Chaetomia contain genes (ccp) encoding cytochrome c peroxidases (CcP, Fig. 2 – middle of the left part). The relationships among the fungi presented in the CcP phylogenetic analysis suggest that this protein has evolved vertically throughout Ascomycetes. For ccp genes from both C. globosum and C. cochliodes a basal lineage represented by C. thermophilum and M. thermophila is apparent in the reconstructed tree. The physiological role of CcP is still under discussion.
Fig. 3

Presentation of the promoter region for CcochkatG gene showing typical eukaryotic regulatory elements for a HGT-related bacterial gene. Sequence analysis was performed in Contig 0012 between positions 43,000 - 47,000 with FGENESH software [14], drawn to scale

Further phylogenetic reconstruction of the peroxidase-catalase superfamily reveals that in C. cochliodes but not in C. globosum a Family II representative is present (Fig. 2 – lower part). This is very surprising for such closely related fungal species. However, the Family II representative from C. cochliodes has its closest neighbour in C. thermophilum. Family II ascomycetous genes code for hypothetical heme peroxidases with yet unknown reaction specificity but are closely related with well investigated basidiomycetous manganese and lignin peroxidases (Fig. 2, labelled violet). The latter are involved in oxidative degradation of lignin-containing soil debris and typically use Mn2+ or small organic molecules as electron donors.

Additional representatives from the peroxidase-catalase superfamily in C. cochliodes include two paralogs of hybrid B heme peroxidases discovered as a new gene family only recently [25]. Hybrid-type B peroxidases are present solely in fungi but are related to Family III (comprised of numerous plant secretory peroxidases, labelled green in Fig. 2) and also to Family II (fungal secretory peroxidases mentioned above). The basal lineage for the first paralog (CcochHyBpox1) together with its closely related C. globosum counterpart appears among mesophilic Sordariomycetes (Fig. 2 upper part). The second variant (CcochHyBpox2) containing besides the peroxidase domain also an additional C-terminal WSC (sugar binding) domain is not closely related with C. globosum ortholog (Fig. 2 right). Thus, both these HyBpox paralogs are not the result of a recent gene duplication but segregated rather early in the evolution of fungal genomes. Transcription analysis (Table 5 & Additional file 4: Figure S2) reveals a slight induction of both hyBpox genes selectively with peroxyacetic acid in the cultivation medium. In contrast, previous results [4] reveal a constitutive mode of expression for distantly related katG1 gene with hydrogen peroxide and peroxyacetic acid.
Table 5

Transcription analysis of 9 selected genes for peroxide catabolism in C. cochliodes recorded with RT-qPCR. Quantitative values representing relative changes of the transcription level were obtained by comparison of the expression of a particular gene in 30 min. induced vs. non induced samples. The constitutively expressed ITS1 region was used as internal standard for normalization

Changes in expression levels against non-induced control*

Analysed gene

Sample with 5 mM H2O2

Sample with 5 mM PAA

CcochhyBpox1

1.5 x

3.0 x

CcochhyBpox2

0.3 x

1.7 x

Ccochcyox1

0.3 x

2.3 x

Ccochlds

0.4 x

1.8 x

Ccochdyprx

3.3 x

18.5 x

Ccochhtp1

2.7 x

2.9 x

CcochkatA2

1.1 x

0.5 x

CcochkatB1

0.4 x

1.1 x

CcochkatB2

0.6 x

1.9 x

* Changes in the expression levels compared to the control sample (with the reference value of 1.0) were calculated as relative quantities due to the formula RQ = 2 – ΔΔCq where Cq is the quantification cycle of each RT-qPCR reaction. Presented are average values of triplicates for each listed gene and each inducer. Typical amplification plots and melting curves are presented in Additional file 4: Figure S2

Peroxidase-cyclooxygenase superfamily

Members of the peroxidase-cyclooxygenase superfamily (comprised of Families I - VII) are widely distributed among all domains of life. In many cases they are multidomain proteins with one heme peroxidase domain [10, 28]. Family IV is comprised of bifunctional cyclooxygenases possessing both peroxidase and cyclooxygenase activities. They are involved in various physiological and pathophysiological processes [29]. In mammals they are located in the luminal membrane of the endoplasmatic reticulum and mediate the conversion of free essential fatty acids to prostanoids by a two-step process [30]. The structure and function of the two distinct human paralogs (constitutive COX-1 and inducible COX-2) were intensively investigated but a comprehensive analysis of their diverse paralogs among eukaryotic microbes or even among prokaryots was only recently reported [31]. Evolutionary relationships among fungal cyclooxygenase genes were not analysed in sufficient detail yet.

Our current reconstruction based on the phylogeny of selected members from the whole superfamily (comprising 204 unique genes) is presented in Fig. 4. Genome analysis suggests the occurrence of two representatives of this superfamily in Chaetomia, a cyclooxygenase-like enzyme and a linoleate diol synthase. Cyclooxygenase genes from C. cochliodes and C. globosum share their closest phylogenetic neighbour (Fig. 4 upper part left) in the genome of M. mycetomatis, a human pathogenic fungus that grows optimally at room temperature [32]. No cyclooxygenase genes were found in thermophilic fungi so far. In contrast, the evolutionary reconstruction of another important subfamily of Family IV, linoleate diol synthases, reveals a very similar pattern for Chaetomiaceae as already described for the previous superfamily. Corresponding part of the tree (Fig. 4 – upper part right) demonstrates that genes encoding linoleate diol synthases (lds) from thermophilic fungi (M. thermophila and C. thermophilum) represent basal lineages for corresponding genes in mesophilic Chaetomia. Only recently it was shown that fatty acid diol synthases are unique fusion proteins containing a N-terminal heme peroxidase domain joined with a C-terminal P450-heme thiolate domain for conversion of unsaturated fatty acids to dihydroxy-fatty acids [33]. These enzymes are an essential part of the psi factor sexual inducer cascade in various fungi [34]. Their exact physiological role within the life cycle of Chaetomiaceae needs to be elucidated in the future. Our first round of transcription analysis revealed around 2-fold induction of expression of both cyox1 and lds genes in a medium with peroxyacetic acid (Table 5 and Additional file 4: Figure S2).
Fig. 4

Reconstructed phylogeny of the peroxidase-cyclooxygenase superfamily with focus on Chaetomia ORFs. The complete tree from 204 full length sequences with 1,053 aligned sites is presented. C. cochliodes sequences are labelled red. Distinct subfamilies are labelled in different colours. Values in nodes represent bootstrap values above 50 (from maximum likelihood analysis) and posterior probabilities (from Mr. Bayes), respectively. Abbreviations of peroxidase names are listed in Additional file 5: Table S3. Abbreviations of taxa: Ac, Actinobacteria; Acb, Acidobacteria; Cy, Cyanobacteria; Prb, Proteobacteria; Plb, Planctomycetes (bacteria); As, Ascomycota; Ba, Basidiomycota; Mu, Mucoromycota; St, Stramenopiles; Cn, Cnidaria; De, Deuterostomia

Peroxidase-chlorite dismutase superfamily

Our next screening within the C. cochliodes genome focused on the presence of genes encoding dye-decolorizing peroxidases (DyPs). These heme enzymes were first isolated from soil basidiomycetes but were further shown to be present in a wide variety of fungi and bacteria [35]. DyPs catalyse the H2O2-mediated oxidation of a very broad substrate range. Originally, fungal representatives were found to degrade bulky dyes. A detailed structure- and sequence-based comparison demonstrated that DyPs together with chlorite dismutases and chlorite-dismutase like proteins (EfeB, HemQ) constitute the CDE superfamily [36], also designated as peroxidase-chlorite dismutase superfamily [10]. The reconstructed evolution of DyPs within this superfamily is shown in Fig. 5. In fungal genomes mainly representatives of the subfamilies DyP-type D and DyP-type B can be found as paralogs. Interestingly, in the genome of C. cochliodes only a fused version of DyP-PFL is present, i.e. an N-terminal DyP peroxidase domain connected with a C-terminal pyruvate formate-lyase (PFL) domain known as a glycyl radical containing region [37]. This unique gene fusion was detected also in other distantly related prokaryotic & eukaryotic genomes [38]. The PFL domain can be activated by PFL activase, a radical SAM superfamily member [39], but the significance of a PFL fusion with a peroxidase domain remains elusive. We could detect a putative PFL activase in C. cochliodes contig 00230 revealing 81 % identity with CHGG_03160 from C. globosum and other putative PFL activases from filamentous fungi. Thus, C. cochliodes possesses both components necessary for the glycyl radical formation with yet unknown physiological function. A HGT event with a high bootstrap support in the clade of fused DyPs B can be observed between proteobacteria and ascomycetous fungi (Fig. 5 and Additional file 6: Table S4 for abbreviations). As the fused DyP B-PFL proteins are yet hypothetical, their physiological relevance has to be determined among Chaetomiaceae. Our first round of transcription analysis of dyprx gene exhibited the highest induction observed among all 5 superfamilies followed in this study with hydrogen peroxide (3-fold) and mainly with peroxyacetic acid (18.5-fold) in the cultivation medium (Table 5).
Fig. 5

Reconstructed phylogeny of the peroxidase-dismutase superfamily with focus on newly discovered Chaetomia sequences forming a separate clade of DyP-Bs together with fused bacterial representatives from which they were derived by a HGT event. The complete tree from 282 full length sequences is presented with 655 sites aligned. C. cochliodes sequence is labelled red. Distinct subfamilies are labelled in different colours. Values in nodes represent bootstrap values above 50 (from maximum likelihood analysis) and posterior probabilities (from Mr. Bayes), respectively. Abbreviations of peroxidase names are listed in Additional file 6: Table S4. Abbreviations of taxa: vir, DNA viruses; Ac, Actinobacteria; Aci, Acidobacteria; Bi, Bacteroidetes; Chl, Chloroflexi (bacteria); Cy, Cyanobacteria; Dei, Deinococci; Fi, Firmicutes; Pb, Proteobacteria; Pmc, Planctomycetes; As, Ascomycota; Ba, Basidiomycota; Alv, Alveolata; Amb, Ameboflagellates; De, Deuterostomia; Mol, Mollusca

Peroxidase-peroxygenase superfamily

Heme-thiolate peroxidases from Fungi and Stramenopiles constitute the peroxidase-peroxygenase superfamily [10]. Enzymes encoded by htp genes represent probably the most versatile catalysts among peroxidase superfamilies thus catalysing on one side classical heme peroxidase reactions and on the other side monooxygenase (monohydroxylation) reactions like cytochrome P450s [40]. The reconstructed phylogenetic tree for the peroxidase-peroxygenase superfamily (Fig. 6) reveals the distribution of three gene paralogs of this superfamily within the Chaetomium cochliodes genome. The presence of multiple gene paralogs in genomes of ascomycetous fungi is frequent and occurred by repeated gene duplications of this rather short gene but the phylogenetic distribution of C. cochliodes paralogs is variable (Fig. 6). Whereas there is a thermophilic basal lineage for CcochHTP2 and CcochHTP3 and their corresponding counterparts in C. globosum, the situation for paralog CcochHTP1 is different. Corresponding genes from pathogenic fungi represent a basal lineage for closely related CcochHTP1 and CgHTP1. It is unknown so far whether these three putative heme-thiolate peroxidases exhibit different enzymatic properties but they were segregated early during the evolution of fungal genomes and thus they all may be interesting for biotechnological applications. We have also performed transcription analysis of htp1 gene paralog resulting in almost 3-fold induction both with hydrogen peroxide and peroxyacetic acid present in the cultivation medium (Table 5).
Fig. 6

Phylogeny of the peroxidase-peroxygenase superfamily representing numerous gene paralogs of this superfamily among Chaetomiaceae. The complete tree from 172 full length sequences is presented with 287 sites aligned. C. cochliodes paralogs are labelled red. Distinct subfamilies are labelled in different colours. Values in nodes represent bootstrap values above 50 (from maximum likelihood analysis) and posterior probabilities (from Mr. Bayes), respectively. Abbreviations of peroxidase names are listed in Additional file 7: Table S5. Abbreviations of taxa: As, Ascomycota; Ba, Basidiomycota; Mu, Mucoromycota; St, Stramenopiles

Putative heme catalases in Chaetomia

Typical (monofunctional) heme catalases are enzymes that very efficiently dismutase hydrogen peroxide to oxygen and water. In contrast with heme peroxidases they can both reduce and oxidize hydrogen peroxide and have negligible peroxidatic activity [41]. Heme catalases represent a monophyletic group that evolved as a distinct gene family from prokaryotes to almost all lineages of eukaryotes [11]. In Fig. 7 the phylogeny focused on fungal heme catalases is presented. There are 3 distinct clades of genes for typical catalases defined by Klotz et al. [42]. In fungi only representatives of Clade 2 (large subunit, secretory catalases) and Clade 3 (small subunit, mostly peroxisomal catalases) can be found. There are up to four gene paralogs of a catalase gene within C. cochliodes genome that underlines the importance of mostly monofunctional catalases for the removal of H2O2. There are thermophilic basal lineages for the large subunit secretory catalases CcochKatA1, CcochKatA2 and their C. globosum counterparts, a situation very similar to the peroxidase superfamilies. In contrast, there are mesophilic basal lineages for the small subunit peroxisomal catalases CcochKatB1 and CcochKatB2 (Fig. 7 – on the right). In particular, CcochKatB1 and CgKatB1 have a basal lineage among catalases from various soil and phytopathogenic fungi. Surprisingly, CcochKatB2 has no counterpart in the closely related genome of C. globosum. Putative catalase from a widely distributed soil fungus S. schenckii shares a common ancestor with this unique small subunit peroxisomal catalase of C. cochliodes (Fig. 7). Possible involvement of C. cochliodes four catalase isozymes in the defence against oxidative stress was analysed by RT-PCR. Obtained results in the early exponential phase of fungal growth show only a slight induction of the paralog katB2 in the medium containing peroxyacetic acid (Table 5).
Fig. 7

Reconstructed phylogeny of the heme catalase super family with focus on Clade 2 and 3 representing the distribution of Ascomycetous large subunit as well as small subunit catalases (labelled in different colors). The complete tree from 222 full length sequences is presented with 546 sites aligned. C. cochliodes paralogs are labelled red. Distinct clades are labelled in different colours. Values in nodes represent bootstrap values above 50 (from maximum likelihood analysis) and posterior probabilities (from Mr. Bayes), respectively. Abbreviations of peroxidase names are listed in Additional file 8: Table S6. Abbreviations of taxa: Ar, Archaea; Ac, Actinobacteria; Aci, Acidobacteria; Bi, Bacteroidetes; Chl, Chloroflexi (bacteria); Cy, Cyanobacteria; Dei, Deinococci; Fi, Firmicutes; Pb, Proteobacteria; Pmc, Planctomycetes; As, Ascomycota; Ba, Basidiomycota; Chy, Chytridiomycota; Zy, Zygomycota; Cn, Cnidaria; Ich, Ichthyosporea; Chlph, Chlorophyta; BMagno, basal Magnoliophyta; My, Mycetozoa; Cryp, Cryptogams, Eudi, Eudicotyledons, Mctd, Monocotyledons; De, Deuterostomia; Ec, Ecdysozoa

Conclusions

In conlusion genomic sequence analysis revealed that Chaetomium cochliodes is closely related to C. globosum & C. elatum. These three filamentous fungi are mesophilic but probably have thermophilic ancestors as revealed from their basal lineage. C. cochliodes contains heme peroxidases and catalases from all so far described superfamilies. Ascomycetous genes encoding catalase-peroxidase and dye decolorizing peroxidase were obtained during the evolution by horizontal gene transfer from various bacteria. Several heme peroxidases of Chaetomia like hybrid heme B peroxidase, linoleate diol synthase or DyP-type B form fusions with additional functional domains that might enable a broader catalytic variability. Furthermore cytochrome c peroxidase, manganese and three paralogs of heme-thiolate peroxidases are found in addition to typical (monofunctional) catalases of large and small subunit architecture. Our transcription analysis reveals the highest induction of a fused dyprx gene with hydrogen peroxide and mainly with peroxyacetic acid in the cultivation medium followed by moderate inductions of htp1 and hyBpox1 genes.

Abbreviations

CcP: 

Cytochrome c peroxidase

CldL: 

Chlorite dismutase-like protein

CTAB: 

Hexadecyltrimethylammonium bromide

HGT: 

Horizontal gene transfer

HMM: 

Hidden Markov model

KatG: 

Bifunctional catalase-peroxidase

LDS: 

Linoleate diol synthase

LiPOX: 

Lignin peroxidase

ML: 

Maximum likelihood phylogeny

MnPOX: 

Manganese peroxidase

ORF: 

Open reading frame

PAA: 

Peroxyacetic acid

PEG: 

Polyethylene glycol

PFL: 

Pyruvate formate-lyase

RT-qPCR: 

Quantitative real-time PCR

SOD: 

Superoxide dismutase

WSC: 

Cell-wall integrity & stress response component

Declarations

Acknowledgements

Our research was supported by the Austrian Science Fund (FWF, project P27474-B22), by the Slovak Grant Agency VEGA (grant 2/0021/14) and by the Slovak Research and Development Agency (grant APVV-14-0375). We thank the company Hermes LabSystems for providing us with AriaMx6 device.

Availability of data and materials

All used DNA sequences are deposited in GenBank (Table 1). All protein sequences that were used for reconstruction of phylogenies are listed in Additional file 3: Table S2, Additional file 6: Table S4, Additional file 5: Table S3, Additional file 7: Table S5 and Additional file 8: Table S6. If possible their PeroxiBase accession number is given to find them at (http://peroxibase.toulouse.inra.fr) if no PeroxiBase accession numbers exist yet their UniProt (http://www.uniprot.org) accession numbers are given.

Authors’ contributions

MZ selected the fungus, designed all experiments, performed all molecular phylogeny analyses and wrote the manuscript; AK cultivated the fungus and performed genomic & transcription analyses; KC optimised the isolation of fungal DNA and performed genomic & transcription analysis; KL prepared the genomic DNA for sequencing and contributed to the discussion; HT performed the sequencing and assembled the contigs; CO evaluated the classification and phylogeny of peroxidases & catalases and finalized the manuscript. All authors read and approved the final manuscript.

Competing interests

The authors’ declare that they have no competing interests.

Consent to publish

Not applicable (this manuscript does not contain any individual persons data).

Ethics approval and consent to participate

Not applicable for this fungal genomic study. None of here analysed genes of Chaetomia was used in experimental cloning research (yet).

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

(1)
Department of Chemistry, Division of Biochemistry, University of Natural Resources and Life Sciences
(2)
Institute of Molecular Biology, Slovak Academy of Sciences
(3)
Department of Biotechnology, University of Natural Resources and Life Sciences

References

  1. Amlacher S, Sarges P, Flemming D, van Noort V, Kunze R, Devos DP, Arumugam M, Bork P, Hurt E. Insight into structure and assembly of the nuclear pore complex by utilizing the genome of a eukaryotic thermophile. Cell. 2011;146:277–89.View ArticlePubMedGoogle Scholar
  2. Cuomo CA, Untereiner WA, Ma LJ, Grabherr M, Birren BW. Draft genome sequence of the cellulolytic fungus Chaetomium globosum. Genome Announc. 2015;3:e00021–00015.View ArticlePubMedPubMed CentralGoogle Scholar
  3. Geiger WB, Conn JE, Waksman SA. Chaetomin, a new antibiotic substance produced by Chaetomium cochliodes. J Bacteriol. 1944;48:527–36.Google Scholar
  4. Zámocký M, Sekot G, Bučková M, Godočíková J, Schäffer C, Farkašovský M, Obinger C, Polek B. Intracellular targeting of ascomycetous catalase-peroxidases (KatG1s). Arch Microbiology. 2013;195(6):393–402.View ArticleGoogle Scholar
  5. Li C, Shi L, Chen D, Ren A, Gao T, Zhao M. Functional analysis of the role of glutathione peroxidase (GPx) in the ROS signaling pathway, hyphal branching and the regulation of ganoderic acid biosynthesis in Ganoderma lucidum. Fungal Genet Biol. 2015;82:168–80.View ArticlePubMedGoogle Scholar
  6. Narasaiah KV, Sashidhar RB, Subramanyam C. Biochemical analysis of oxidative stress in the production of aflatoxin and its precursor intermediates. Mycopathologia. 2006;162:179–89.View ArticlePubMedGoogle Scholar
  7. Nielsen KF, Gravesen S, Nielsen PA, Andersen B, Thrane U, Frisvad JC. Production of mycotoxins on artificially and naturally infested building materials. Mycopathologia. 1999;145:43–56.View ArticlePubMedGoogle Scholar
  8. Ye Y, Xiao Y, Ma L, Li H, Xie Z, Wang M, Ma H, Tang H, Liu J. Flavipin in Chaetomium globosum CDW7, an endophytic fungus from Ginkgo biloba, contributes to antioxidant activity. Appl Microbiol Biotechnol. 2013;97:7131–9.View ArticlePubMedGoogle Scholar
  9. Sivakumar V, Thanislass J, Niranjali S, Devaraj H. Lipid peroxidation as a possible secondary mechanism of sterigmatocystin toxicity. Hum Exp Toxicol. 2001;20:398–403.View ArticlePubMedGoogle Scholar
  10. Zámocký M, Hofbauer S, Schaffner I, Gasselhuber B, Nicolussi A, Soudi M, Pirker KF, Furtmüller PG, Obinger C. Independent evolution of four heme peroxidase superfamilies. Arch Biochem Biophys. 2015;574:108–19.View ArticlePubMedPubMed CentralGoogle Scholar
  11. Zámocký M, Furtmüller PG, Obinger C. Evolution of Catalases from Bacteria to Humans. Arch Microbiology. 2008;10:1527–47.Google Scholar
  12. Carlson JE, Tulsieram LK, Glaubitz JC, Luk VMK, Kauffeldt C, Rutledge R. Segregation of random amplified DNA markers in F1 progeny of conifers. Theor Appl Genet. 1991;83:194–200.View ArticlePubMedGoogle Scholar
  13. Healey A, Furtado A, Cooper T, Henry RJ. Protocol: a simple method for extracting next-generation sequencing quality genomic DNA from recalcitrant plant species. Plant Methods. 2014;10:21.View ArticlePubMedPubMed CentralGoogle Scholar
  14. Solovyev V, Kosarev P, Seledsov I, Vorobyev D. Automatic annotation of eukaryotic genes, pseudogenes and promoters. Genome Biol. 2006;7 Suppl 1:10.11–2.View ArticleGoogle Scholar
  15. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucl Acids Res. 2004;32:1792–7.View ArticlePubMedPubMed CentralGoogle Scholar
  16. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: molecular evolutionary genetics analysis version 6.0. Mol Biol Evol. 2013;30:2725–9.View ArticlePubMedPubMed CentralGoogle Scholar
  17. Schoch CL, Seifert KA, Huhndorf S, Robert V, Spouge JL CAL, Chen W, Fungal Barcoding Consortium. Nuclear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi. Proc Natl Acad Sci U S A. 2012;109:6241–6.View ArticlePubMedPubMed CentralGoogle Scholar
  18. Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Höhna S, Larget B, Liu L, Suchard MA, Huelsenbeck JP. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst Biol. 2012;61:539–42.View ArticlePubMedPubMed CentralGoogle Scholar
  19. Violi HA, Menge JA, Beaver RJ. Chaetomium elatum (Kunze: Chaetomiaceae) as a root-colonizing, commensalistic associate, or pathogen? Am J Botany. 2007;94:690–700.View ArticleGoogle Scholar
  20. Morgenstern I, Powlowski J, Ishmael N, Darmond C, Marqueteau S, Moisan MC, Quenneville G, Tsang A. A molecular phylogeny of thermophilic fungi. Fungal Biol. 2012;116:489–502.View ArticlePubMedGoogle Scholar
  21. Van den Brink J, Facun K, De Vries M, Stielow JB. Thermophilic growth and enzymatic thermostability are polyphyletic traits within Chaetomiaceae. Fungal Biol. 2015;119:1255–66.View ArticlePubMedGoogle Scholar
  22. Giorgio M, Trinei M, Migliacco E, Pelicci PG. Hydrogen peroxide: a metabolic by-product or a common mediator of ageing signals? Nat Rev Mol Cell Biol. 2007;8:722–8.View ArticlePubMedGoogle Scholar
  23. Papapostolou I, Sideri M, Georgiou CD. Cell proliferating and differentiating role of H2O2 in Sclerotium rolfsii and Sclerotinia sclerotiorum. Microbiol Res. 2014;169:527–32.View ArticlePubMedGoogle Scholar
  24. Zámocký M, Furtmüller PG, Obinger C. Evolution of structure and function of class I peroxidases. Arch Biochem Biophys. 2010;500:45–57.View ArticlePubMedGoogle Scholar
  25. Zámocký M, Gasselhuber B, Furtmüller PG, Obinger C. Turning points in the evolution of peroxidase-catalase superfamily - molecular phylogeny of hybrid heme peroxidases. Cell Mol Life Sci. 2014;71:4681–96.View ArticlePubMedPubMed CentralGoogle Scholar
  26. Zámocký M, Droghetti E, Bellei M, Gasselhuber B, Pabst M, Furtmüller PG, Battistuzzi G, Smulevich G, Obinger C. Eukaryotic extracellular catalase-peroxidase from Magnaporthe grisea - Biophysical/chemical characterization of the first representative from a novel phytopathogenic KatG group. Biochimie. 2012;94:673–83.View ArticlePubMedPubMed CentralGoogle Scholar
  27. Thön M, Al Abdallah Q, Hortschansky P, Scharf DH, Eisendle M, Haas H, Brakhage AA. The CCAAT-binding complex coordinates the oxidative stress response in eukaryotes. Nucl Acids Res. 2010;38:1098–113.View ArticlePubMedGoogle Scholar
  28. Zámocký M, Jakopitsch C, Furtmüller PG, Dunand C, Obinger C. The peroxidase-cyclooxygenase superfamily. Reconstructed evolution of critical enzymes of the innate immune system. Proteins. 2008;71:589–605.View ArticleGoogle Scholar
  29. Kudalkar SN, Rouzer CA, Marnett LJ. The Peroxidase and Cyclooxygenase Activity of Prostaglandin H Synthase. In: Raven E, Dunford B, editors. Heme Peroxidases, vol. RSC Metallobiology. Cambridge: Royal Society of Chemistry; 2015. p. 247–71.Google Scholar
  30. Chandrasekharan NV, Simmons DL. The cyclooxygenases. Genome Biol. 2004;5:241.View ArticlePubMedPubMed CentralGoogle Scholar
  31. Gupta K, Selinsky BS. Bacterial and algal orthologs of prostaglandin H2 synthase: novel insights into the evolution of an integral membrane protein. Biochim Biophys Acta. 2015;1848:83–94.View ArticlePubMedGoogle Scholar
  32. Mir F, Shakoor S, Khan MJ, Minhas K, Zafar A, Zaidi AKM. Madurella mycetomatis as an agent of brain abscess: case report and review of literature. Mycopathologia. 2013;176:429–34.View ArticlePubMedGoogle Scholar
  33. Shin K-C, Seo M-J, Oh D-K. Characterization of a novel 8 R,11 S -linoleate diol synthase from Penicillium chrysogenum by identification of its enzymatic products. J Lipid Res. 2016;57:207–18.PubMedPubMed CentralGoogle Scholar
  34. Brodhun F, Gobel C, Hornung E, Feussner I. Identification of PpoA from Aspergillus nidulans as a fusion protein of a fatty acid heme dioxygenase/peroxidase and a cytochrome P450. J Biol Chem. 2009;284:11792–805.View ArticlePubMedPubMed CentralGoogle Scholar
  35. Rahmanpour R, Bugg TDH. Structure and Reactivity of the Dye-decolorizing Peroxidase (DyP) Family. In: Raven E, Dunford B, editors. Heme Peroxidases, vol. RSC Metallobiology. Cambridge: Royal Society of Chemistry; 2015. p. 334–57.View ArticleGoogle Scholar
  36. Goblirsch B, Kurker RC, Streit BR, Wilmot CM, DuBois JL. Chlorite dismutases, DyPs, and EfeB: 3 microbial heme enzyme families comprise the CDE structural superfamily. J Mol Biology. 2011;408:379–98.View ArticleGoogle Scholar
  37. Becker A, Fritz-Wolf K, Kabsch W, Knappe J, Schultz S, Volker Wagner AF. Structure and mechanism of the glycyl radical enzyme pyruvate formate-lyase. Nat Struct Biol. 1999;6:969–75.View ArticlePubMedGoogle Scholar
  38. Strittmatter E, Plattner DA, Piontek K: Dye-Decolorizing Peroxidase (DyP). Encyclopedia of Inorganic and Bioinorganic Chemistry, Online 2014.Google Scholar
  39. Frey PA, Hegeman AD, Ruzicka FJ. The Radical SAM Superfamily. Crit Rev Biochem Mol Biol. 2008;43:63–88.View ArticlePubMedGoogle Scholar
  40. Hofrichter M, Ulrich R. Oxidations catalyzed by fungal peroxygenases. Curr Opin Chem Biol. 2014;19:116–25.View ArticlePubMedGoogle Scholar
  41. Zámocký M, Gasselhuber B, Furtmüller PG, Obinger C. Molecular evolution of hydrogen peroxide degrading enzymes. Arch Biochem Biophys. 2012;525:131–44.View ArticlePubMedPubMed CentralGoogle Scholar
  42. Klotz MG, Loewen PC. The molecular evolution of catalatic hydroperoxidases: Evidence for multiple lateral transfer of genes between prokaryota and from bacteria into eukaryota. Mol Biol Evol. 2003;20:1098–112.View ArticlePubMedGoogle Scholar

Copyright

© The Author(s). 2016

Advertisement