The authors have declared that no competing interests exist.
Conceived and designed the experiments: WC FC. Performed the experiments: WC TX. Analyzed the data: WC TX YS. Contributed reagents/materials/analysis tools: WC TX YS FC. Wrote the paper: WC FC.
Fungal amylolytic enzymes, including α-amylase, gluocoamylase and α-glucosidase, have been extensively exploited in diverse industrial applications such as high fructose syrup production, paper making, food processing and ethanol production. In this paper, amylolytic genes of 85 strains of fungi from the phyla Ascomycota, Basidiomycota, Chytridiomycota and Zygomycota were annotated on the genomic scale according to the classification of glycoside hydrolase (GH) from the Carbohydrate-Active enZymes (CAZy) Database. Comparisons of gene abundance in the fungi suggested that the repertoire of amylolytic genes adapted to their respective lifestyles. Amylolytic enzymes in family GH13 were divided into four distinct clades identified as heterologous α- amylases, eukaryotic α-amylases, bacterial and fungal α-amylases and GH13 α-glucosidases. Family GH15 had two branches, one for gluocoamylases, and the other with currently unknown function. GH31 α-glucosidases showed diverse branches consisting of neutral α-glucosidases, lysosomal acid α-glucosidases and a new clade phylogenetically related to the bacterial counterparts. Distribution of starch-binding domains in above fungal amylolytic enzymes was related to the enzyme source and phylogeny. Finally, likely scenarios for the evolution of amylolytic enzymes in fungi based on phylogenetic analyses were proposed. Our results provide new insights into evolutionary relationships among subgroups of fungal amylolytic enzymes and fungal evolutionary adaptation to ecological conditions.
Starch is the major carbohydrate storage product of green plants as a result of photosynthesis and makes up an important part of carbon and energy sources widely consumed among animals, plants and microorganisms
As heterotrophic microorganisms, fungi utilize polysaccharide substrates through a complement of hydrolytic enzymes secreted into the environmental niches to digest large organic molecules into smaller molecules that may then be absorbed as nutrients. Some fungi, for example members of the genus
α-amylases act on α-1,4-glycosidic bonds with the endo-hydrolysis of the long polysaccharide chains into shorter maltooligosaccharides and α-limit dextrins
Glucoamylases, also known as γ-amylases, catalyse hydrolysis of α-1,4 and α-1,6 glucosidic linkages to release β-D-glucose from the non-reducing ends of starch and related poly- and oligosaccharides
α-glucosidases hydrolyze α-1,4 and/or α-1,6-linkages of saccharides to liberate α-D-glucose from the non-reducing end
Amylolytic enzymes of microorganisms, in particular filamentous fungi, from the families GH13 and GH15 often possess starch-binding domains facilitating attachment and degradation of raw starch
Fungal amylolytic enzymes as the major industrial source play an important role in starch processing. There have been extensive studies focused on the identification and regulation of fungal amylolytic genes
Putative amylolytic genes from 85 strains of fungi were identified by HMMER searches and numbers of the annotated amylolytic genes were compared among these fungi (
Phylum | Taxonomic group | Species | Strains | Abbreviation | α-amylases(GH13) | GH15 | α-glucosidases(GH13) | α-glucosidases(GH31) |
Ascomycota | Dothideomycetes |
|
JN3 | Lm | 2 | 3 | 3 | 5 |
|
SN15 | Pn | 2 | 3 | 3 | 7 | ||
|
IPO323 | Zt | 5 | 1 | 4 | 7 | ||
Eurotiales |
|
NRRL 1 | Acl | 7 | 6 | 4 | 3 | |
|
NRRL 3357 | Afl | 5 | 3 | 5 | 6 | ||
|
Af293 | Afu | 6 | 5 | 4 | 4 | ||
|
IFO 4308 | Ak | 8 | 2 | 1 | 5 | ||
|
CBS 513.88 | An | 6 | 2 | 2 | 4 | ||
|
RIB40 | Aor | 6 | 2 | 5 | 5 | ||
|
NIH2624 | At | 7 | 2 | 2 | 6 | ||
|
FGSC A4 | En | 7 | 2 | 2 | 5 | ||
|
M7 | Mr | 3 | 3 | 4 | 2 | ||
|
NRRL 181 | Nf | 7 | 5 | 4 | 6 | ||
|
Wisconsin54-1255 | Pc | 7 | 3 | 4 | 8 | ||
|
ATCC 18224 | Pm | 5 | 4 | 1 | 5 | ||
|
ATCC 10500 | Ts | 5 | 4 | 3 | 5 | ||
Onygenales |
|
G186AR | Aca | 3 | 1 | 0 | 3 | |
|
CBS 113480 | Aot | 1 | 1 | 0 | 2 | ||
|
RS | Ci | 2 | 1 | 0 | 2 | ||
|
str. Silveira | Cp | 2 | 1 | 0 | 2 | ||
|
Pb01 | Pb | 3 | 1 | 0 | 2 | ||
|
CBS 127.97 | Te | 1 | 1 | 0 | 2 | ||
|
CBS 118892 | Tr | 1 | 1 | 0 | 2 | ||
|
CBS 112818 | Tto | 1 | 1 | 0 | 2 | ||
|
HKI 0517 | Tv | 1 | 1 | 0 | 2 | ||
Orbiliomycetes |
|
ATCC 24927 | Aol | 2 | 4 | 2 | 2 | |
Pezizomycetes |
|
Mel28 | Tm | 4 | 1 | 1 | 3 | |
Saccharomycotina |
|
WO-1 | Ca | 0 | 3 | 2 | 1 | |
|
CD36 | Cd | 0 | 3 | 2 | 1 | ||
|
CBS 138 | Cga | 0 | 1 | 0 | 1 | ||
|
MYA-3404 | Ct | 0 | 2 | 5 | 1 | ||
|
ATCC 42720 | Cl | 0 | 1 | 2 | 2 | ||
|
CBS767 | Dh | 0 | 2 | 2 | 1 | ||
|
DBVPG#7215 | Ec | 0 | 1 | 0 | 1 | ||
|
ATCC 10895 | Eg | 0 | 1 | 0 | 1 | ||
|
NRRL Y-1140 | Kl | 0 | 1 | 2 | 1 | ||
|
CBS 7435 | Kp | 0 | 1 | 0 | 1 | ||
|
CBS 6340 | Lt | 0 | 1 | 4 | 1 | ||
|
NRRL YB-4239 | Le | 0 | 1 | 1 | 1 | ||
|
ATCC 6260 | Mg | 0 | 1 | 3 | 1 | ||
|
CBS 4309 | Nca | 0 | 1 | 0 | 1 | ||
|
CBS 421 | Nd | 0 | 1 | 0 | 1 | ||
|
DL-1 | Op | 0 | 1 | 1 | 1 | ||
|
YJM789 | Sce | 0 | 1 | 3 | 1 | ||
|
CBS 6054 | Sst | 0 | 2 | 5 | 2 | ||
|
CBS 4417 | Tp | 0 | 1 | 0 | 1 | ||
|
CBS 1146 | Td | 0 | 1 | 3 | 1 | ||
|
CLIB122 | Yl | 0 | 1 | 0 | 1 | ||
|
CBS 732 | Zr | 0 | 1 | 0 | 2 | ||
Sordariomyceta |
|
B05.10 | Bf | 6 | 4 | 0 | 3 | |
|
CBS 148.51 | Cgo | 4 | 2 | 2 | 3 | ||
|
CM01 | Cm | 1 | 2 | 1 | 3 | ||
|
PH-1 | Gz | 1 | 3 | 5 | 3 | ||
|
74030 | Gl | 2 | 2 | 2 | 2 | ||
|
M1.001 | Gg | 5 | 3 | 3 | 4 | ||
|
kw1407 | Gc | 0 | 2 | 1 | 1 | ||
|
QM6a | Hj | 1 | 2 | 2 | 3 | ||
|
70–15 | Mo | 4 | 2 | 1 | 3 | ||
|
CQMa 102 | Mac | 1 | 2 | 3 | 4 | ||
|
ARSEF 23 | Man | 1 | 2 | 3 | 4 | ||
|
ATCC 42464 | Mt | 3 | 2 | 2 | 3 | ||
|
mpVI 77-13-4 | Nh | 1 | 2 | 5 | 4 | ||
|
OR74A | Ncr | 4 | 2 | 2 | 4 | ||
|
FGSC 2508 | Nt | 4 | 2 | 2 | 4 | ||
|
1980 UF-70 | Ssc | 6 | 4 | 1 | 3 | ||
|
k-hell | Sm | 5 | 3 | 2 | 4 | ||
|
NRRL 8126 | Tte | 1 | 3 | 2 | 4 | ||
|
VaMs.102 | Va | 2 | 4 | 3 | 3 | ||
|
VdLs.17 | Vd | 2 | 4 | 3 | 4 | ||
Taphrinomycotina |
|
yFS275 | Sj | 6 | 2 | 0 | 3 | |
|
972h- | Sp | 7 | 2 | 1 | 4 | ||
Basidiomycota | Agaricomycotina |
|
okayama7#130 | Cc | 4 | 4 | 2 | 3 |
|
WM276 | Cgt | 5 | 2 | 2 | 3 | ||
|
var. neoformans B-3501A | Cn | 5 | 2 | 2 | 3 | ||
|
S238N-H82 | Lb | 4 | 2 | 1 | 3 | ||
|
FA553 | Mp | 2 | 3 | 0 | 4 | ||
|
Mad-698-R | Pp | 2 | 2 | 0 | 6 | ||
|
H4-8 | Sco | 8 | 3 | 1 | 3 | ||
|
var. lacrymans S7.3 | Sl | 3 | 2 | 1 | 4 | ||
Pucciniomycotina |
|
98AG31 | Ml | 3 | 4 | 0 | 3 | |
|
CRL 75-36-700-3 | Pg | 1 | 3 | 0 | 2 | ||
Ustilaginomycotina |
|
SRZ2 | Sr | 1 | 1 | 2 | 3 | |
|
1 | Um | 1 | 1 | 1 | 3 | ||
Chytridiomycota | Chytridiomycetes |
|
JAM81 | Bd | 0 | 1 | 0 | 2 |
Zygomycota | Mucoromycotina |
|
RA 99-880 | Ro | 1 | 6 | 0 | 2 |
Taxonomy information of above fungi is extracted from Taxonomy Browser in NCBI (
The distribution of amylolytic genes from the tested fungi also suggested a strong relationship between the repertoire of amylolytic enzymes in fungal genomes and their saprophytic lifestyle. Members of the genus
For the phylum of Basidiomycota, fungi from Agaricomycotina had more abundance than those from Pucciniomycotina and Ustilaginomycotina in amylolytic gene distribution.
The phylogeny of GH13 including α-amylases and α-glucosidases was analysed among the tested fungi and members of the GH13 family were divided into four clades for studying their protein features (
A. The inner circle was the phylogenetic tree of the GH13 amylolytic enzymes from 85 fungal genomes and the root was put at the mid-point of the longest span across the tree. The tree was inferred by FastTree from the alignments of GH13 amylolytic enzymes constructed by HMMER packages against the profile hidden Markov model of PF00128 and edited on iTOL. The bootstrap values at the inner nodes are displayed by the color that the related edges are marked in red with the values less than 800 in 1000 replicates and otherwise maintain in dark. The outer is the taxon represented as species abbreviation (shown in
Previous studies revealed that the α-amylase family shared a common catalytic domain in the form of a (β/α)8-barrel, a domain of eight parallel β-strands surrounded by eight α-helices
Clade I with two main branches contained the fewest amount of α-amylases among the four clades. The first branch with a cluster of five putative α-amylases from the taxonomic group Agaricomycotina (2), Orbiliomycetes (1), Pezizomycetes (1) and Sordariomyceta (1) showed motif loss, containing only the first three conserved regions up to the conserved position 201. Homology searches using Blastp revealed that these putative α-amylases showed a large functional homogeneity with their animal counterparts. This was surprising, since fungal α-amylases were generally considered to be more related to each other than to the α-amylases from animals
The putative α-amylases in the second branch were from Agaricomycotina (4), Pucciniomycotina (3) and Sordariomyceta (1). Homology searches showed that the α-amylases exhibited high sequence similarity with their counterparts from Actinomycetes. Previous studies indicated that some of the bacterial α-amylases originated from repeated horizontal gene transfer from Eukarya
Most of the α-amylases in the tested fungi were branched into two clades (Clade II and Clade III) based on their phylogenetic relationships. The α-amylases in each clade were from a wide range of taxonomic groups and their phylogeny was generally in agreement with their taxonomic groups such as the α-amylases in close relatives were more likely to be clustered together. Conserved domain searches of consensus sequences using Blastp against NCBI’s Conserved Domain Database showed that the catalytic domains of Clade II were recognized as similar to eukaryotic α-amylases (cd11319, E-value: 0e+00) while the catalytic domains of Clade III were recognized as similar to bacterial and fungal α-amylases (cd11318, E-value: 4.48e-163)
The α-amylases were also shown to occur as multiple genes in a number of the tested fungi especially in the taxonomic group Eurotiales. Close phylogenetic relationships of some α-amylases from the same species suggested an occurrence of gene duplication. Previous studies revealed gene duplications of α-amylases in many living organisms from animals, plants, fungi and bacteria
All annotated α-glucosidases were clustered into Clade IV. The conserved structure and catalytic mechanism within GH13 enzymes are believed to represent a common evolutionary origin
Members of the GH15 family from the tested fungi were divided into two clades based on their phylogenetic relationships (
A. Phylogenetic tree of the GH15 family and B. Primary and secondary structure features of the two clades. For details see legend of
Despite the shared catalytic residues, Clade Ishowed many differences when compared to Clade II especially as some deletions in genes belonging to Clade I resulted in loss of one conserved helix as mentioned above. Moreover, homology searches using Blastp revealed that Clade I reflected an unambiguous assignment to the GH15 family without clear function. The proteins in Clade I were from a wide range of taxonomic groups involving the phyla Ascomycota, Basidiomycota and Zygomycota especially from the fungi with redundancy of glucoamylase genes. The widespread presence of these GH15 proteins suggested a specific function, currently unknown, but probably non-essential. It seems that Clade I was evolved from one of the GH15 forms existing in ancestral fungi and this form was later eliminated in many fungi with selection pressure against the other GH15 form evolved as Clade II in evolution.
The proteins in Clade II annotated as glucoamylases were found in all tested fungi. Generally, the phylogeny of fungal glucoamylases was divided into several main branches, probably due to the multiplicity of glucoamylase forms existing in ancestral fungi. However, fungal glucoamylases showed a conservative pattern in evolution. Glucoamylases from related species were clustered in the tree. It is worth mentioning that glucoamylases in the Saccharomycotina grouped together in the phylogenetic tree, suggesting a common evolutionary origin. This also supports the view mentioned above, namely that the fungi in the taxonomic group of Saccharomycotina were probably evolved from the common ancestral fungi. Another conserved feature of glucoamylases was reflected in their gene number. Glucoamylase genes were presented in each of the tested fungi but are maintained at relatively low number. The conserved evolution in glucoamylases reflected their important roles in fungi, and suggests that they may be essential.
These enzymes were divided into four major clades on the basis of sequence comparisons (
A. Phylogenetic tree of GH31 α-glucosidases and B. Primary and secondary structure features of the four clades. For details see legend of
Primary structural analyses of GH31 α-glucosidases in the tested fungi displayed some characteristic residues. Among them, the invariant Asp182 (nucleophile) and Asp257 (acid/base) (numbering of GH31 consensus in
Conserved domain searches of both consensus sequences revealed specific matches to lysosomal acid α-glucosidases (cd06602, E-value: 0e+00). It is worth mentioning that the enzymes in these two clades were all from a wide range of taxonomic groups. This widespread presence suggests multiple forms of lysosomal acid α-glucosidases in ancestral fungi.
As mentioned above, Clade III (with two main branches) suggested a different evolutionary process in view of the new signature surrounding the catalytic nucleophile. In the upper branch, the putative α-glucosidases reflected a close phylogenetic relationship with their bacterial counterparts based on homology searches, some of which, such as from the taxonomic group Eurotiales, were with specific hits to the bacterial α-glucosidases (cd06594). As these enzymes are present in a few species, they may have been horizontally transferred from bacteria.
The putative α-glucosidases in the other branch of Clade III came from a wide range of fungi including the Ascomycota, Basidiomycota and Chytridiomycota. Homology searches revealed that these enzymes were phylogenetically related to their bacterial counterparts. But their catalytic domains showed non-specific hits to current identified groups in NCBI’s Conserved Domain Database. Probably, these enzymes belonged to a new clade with the signature of DNNE adjacent to the catalytic nucleophile.
The conserved domain of Clade IV showed matches to neutral α-glucosidases (cd06603, E-value: 0e+00). The putative α-glucosidases belonging to this large branch were positively identified in all the tested taxonomic groups. Moreover, the phylogeny of α-glucosidases in this branch was highly in agreement with their taxonomic relationships. This suggests that this α-glucosidase clade is evolutionarily conserved and may be essential in fungi.
About 10% of microbial amylolytic enzymes contain starch-binding domains appended to catalytic modules to mediate the binding of raw starch
The family CBM20 is known as a classical C-terminal starch-binding domain of microbial amylases
Multiple alignments of putative proteins were performed by aligning them to the profile hidden Markov model of PF00686 with HMMER package. Residues assigned to match states were reserved for the profile analysis and their consensus logo and numbering were generated by Jalview. Protein sequence ID is represented as species abbreviation followed by serial number and domain position.
The family CBM21 is known as the N-terminally positioned starch-binding domain of
A, B and C correspond to the alignments of CBM21, 25 and 48 adjusted against the profile hidden Markov models of PF03370, PF03423 and PF02922 respectively.
The CBM25 family was established based on revealing a novel type of starch-binding domain with two copies in a bacterial α-amylase
The CBM48 family was established containing the putative starch-binding domains from the pullulanase subfamily
In our analysis, amylolytic enzymes with starch-binding domains were merely from filamentous fungi. No hits of four domains were showed in amylolytic enzymes from the tested yeasts and mushrooms. Interestingly, except the glucoamylase from
Starch-binding domains have been revealed an independent evolution to the catalytic domains
In this study, the genomic distribution, architecture and phylogeny of amylolytic enzymes including α-amylase, gluocoamylase and α-glucosidase in the available genomes of 85 fungal strains were investigated. Genomic distribution of amylolytic genes suggests their adaptation to the lifestyles of the fungi, at least with respect to starch degradation. Evolutionary significance of the adaptation may lie in their mode of survival, especially in saprobism for obtaining nutrients. Putative starch-binding domains of CBM20, CBM21, CBM25 and CBM48 are concentrated in phylogenetically related amylolytic enzymes from filamentous fungi, especially in Ascomycota. It supports the separate evolution of starch-binding domains to the individual enzymes and suggests their acquisition occurring in certain phylogenetic groups of amylolytic enzymes.
Phylogenetic analyses showed evidence for likely evolutionary events, such as horizontal gene transfer, gene duplication, and gene loss for amylolytic enzymes. We raised a hypothetical scheme for the evolution of genes encoding amylolytic enzymes in fungi (
A. Evolutionary scenarios for the GH13 enzymes. A few α-amylases identified as heterologous α-amylases might be transferred from animals and Actinomycetes. Eukaryotic, bacterial and fungal α-amylases correspond to subfamilies GH13_1 and GH13_5, respectively. GH13 α-glucosidases seem evolved from ancestral α-amylase. B. Evolutionary scenarios for the GH15 enzymes. The function of novel GH15 branch is currently unknown. C. Evolutionary scenarios for the GH31 enzymes. The enzymes in the group of temporarily named bacterial α-glucosidase are phylogenetically close to their bacterial counterparts. They may constitute a new clade of GH31 α-glucosidases in fungi.
Overall protein sequences of 85 strains of fungi from the phyla Ascomycota, Basidiomycota, Chytridiomycota and Zygomycota were used in this study (
The annotation pipeline of amylolytic genes in selected fungi was in a two-step procedure of identification and annotation. The identification step of the families GH13, GH15 and GH31 was performed by using HMMER 3.0 (
Distribution of four carbohydrate-binding module families CBM20, CBM21, CBM25 and CBM48 involving in starch binding was surveyed in the annotated amylolytic enzymes. Profile hidden Markov models of PF00686 (CBM20 family), PF03370 (CBM21 family), PF03423 (CBM25 family) and PF02922 (CBM48 family) from Pfam database were used for HMMER searching against all annotated enzymes. The hits passed MSV, Bias, Vit and Fwd filters were selected as the putative domains.
Alignment of amino acid sequences in the GH13, GH15 and GH31 families were carried out by HMMER package against the corresponding profile hidden Markov models. Phylogenetic trees from alignments of protein sequences were constructed by FastTree version 2.1.4 by maximum likelihood methods (
In this study, structural features were explored in groups of homologous proteins based on their phylogenetic relationships to reveal subfamily-specific conservation patterns, essentially conserved within each subfamily but differing across subfamily. Multiple protein sequence alignments built by HMMER package were edited by Jalview version 2.7
Consensus logos automatically generated by Jalview were used for visualization of the conservation of primary structure by plotting a stack of amino acids for each position. Secondary structures of consensus sequences extracted from the alignments were predicted by Jpred Server version 3.0.1 embedded in Jalview to exploit evolutionary information from multiple sequences