Purifying selection and birth-and-death evolution in the class II hydrophobin gene families of the ascomycete Trichoderma/Hypocrea

Abstract

Background: Hydrophobins are proteins containing eight conserved cysteine residues that occur uniquely in mycelial fungi. Their main function is to confer hydrophobicity to fungal surfaces in contact with air or during attachment of hyphae to hydrophobic surfaces of hosts, symbiotic partners or themselves resulting in morphogenetic signals. Based on their hydropathy patterns and solubility characteristics, hydrophobins are divided into two classes (I and II), the latter being found only in ascomycetes.

Results: We have investigated the mechanisms driving the evolution of the class II hydrophobins in nine species of the mycoparasitic ascomycetous genus Trichoderma/Hypocrea, using three draft sequenced genomes (H. jecorina = T. reesei, H. atroviridis = T. atroviride; H. virens = T. virens) an additional 14,000 ESTs from six other Trichoderma spp. (T. asperellum, H. lixii = T. harzianum, T. aggressivum var. europeae, T. longibrachiatum, T. cf. viride). The former three contained six, ten and nine members, respectively. Ten is the highest number found in any ascomycete so far. All the hydrophobins we examined had the conserved four beta-strands/one helix structure, which is stabilized by four disulfide bonds. In addition, a small number of these hydrophobins (HFBs)contained an extended N-terminus rich in either proline and aspartate, or glycine-asparagine. Phylogenetic analysis reveals a mosaic of terminal clades containing duplicated genes and shows only three reasonably supported clades. Calculation of the ratio of differences in synonymous vs. non-synonymous nucleotide substitutions provides evidence for strong purifying selection (KS/Ka >> 1). A genome database search for class II HFBs from other ascomycetes retrieved a much smaller number of hydrophobins (2-4) from each species, and most were from Sordariomycetes. A combined phylogeny of these sequences with those of Trichoderma showed that the Trichoderma HFBs mostly formed their own clades, whereas those of other Sordariomycetes occurred in shared clades.

Conclusion: Our study shows that the genus Trichoderma/Hypocrea has a proliferated arsenal of class II hydrophobins which arose by birth-and-death evolution followed by purifying selection.

Figure 1

Amino acid alignment of a portion of the extended N-terminus of QID3, Ha_2c, Hv_21a and the Passalora fulva hydrophobin PF1 (for accession number see Table 4). The gap (indicated by a "-") is of different length in the four sequences and therefore not shown. Black background indicates absolutely conserved amino acids. Grey background indicates conservation in at least three of the four proteins.

Figure 2

Amino acid alignment of the class II hydrophobins of Trichoderma/Hypocrea used in this study. The aa sequences were trimmed to show only the area from the first to the eight cysteine. Absolutely conserved aa's are within a black background, and functionally conserved aa's highlighted in grey. The symbols and letters over the alignment show the position of the four beta-strands (S1–S4) and the single helix (indicated by a horizontal cylinder). The sequence below the alignment proposes an updated consensus sequence for the Trichoderma/Hypocrea class II HFBs, as derived from this study: therein, the cysteines are in red and numbered in order of their appearance in the sequence; X denotes any amino acid, and the subscript the number of them; "al" denotes any aliphatic, hydrophobic amino acid (A, V, L, I,)

Figure 3

Phylogenetic analysis of Trichoderma class II hydrophobins. The already published proteins from H. jecorina (HFB1-6) is marked in red. Branchess receiving significant support (> 50% bootstrap values) are indicated with a fat line. Significantly supported clades, which contain hydrophobins from at least 3 different species, are underplayed in grey. The vertical bars mark the clades termed HFB1/2, HFB3 and HFB4.

Figure 4

Plot of one-by-one comparisons of K_a vs. K_s for individual hydrophobin genes within clade "HFB1/2" (●), "HFB3" (■) and "HFB4" (▲) (for explanation of clades see Fig. 6). The dotted line indicates the position of K_a/K_s = 1.

Figure 5

Phylogenetic tree of the nucleotide sequences of Trichoderma/Hypocrea class II hydrophobin genes by the split decomposition method. The "HFB4" clade, whose branch shows the least reticulate network, is highlightened in grey.

Figure 6

NJ analysis of amino acid sequences of class II hydrophobins from Trichoderma and other ascomycetes. Conditions and design of figure are similar as for Fig. 3. Accession numbers and/or genome database entries for the non-Trichoderma sequences are provided in Table 2. The inset on the right bottom shows the topology of an unrooted Bayesian tree (the three Trichoderma clades being highlightened in grey).

Figure 7

Gene expression of the Trichoderma/Hypocrea hydrophobins. (A) Number of ESTs found for the respective hfb genes during screening of the TrichoEST database. Species are abbreviated as follows: TL, T. longibrachiatum; TA, T. asperellum, HL, H. lixii; TAE, T. aggressivum var. europeae; TS, T. stromaticum; TCK, T. cf. koningiopsis. Strain numbers are given in the legend to Table 2. Individual hydrophobin genes are indicated by their respective numbers (cf. Table 1), and no number is given for species in which only a single hydrophobin gene has been detected. (B) Expression of the H. atroviridis gene members of the "HFB4" clade. SM, submerged cultivation; SF, surface cultivation. "5×" indicates that the 5-fold amount of PCR product had been loaded onto the gel.