OUP user menu

The Trichoderma reesei hydrophobin genes hfb1 and hfb2 have diverse functions in fungal development

Sanna Askolin, Merja Penttilä, Han A.B. Wösten, Tiina Nakari-Setälä
DOI: http://dx.doi.org/10.1016/j.femsle.2005.09.047 281-288 First published online: 1 December 2005


Hydrophobins are fungal self-assembling proteins. Here, the hydrophobin genes hfb1 and hfb2 were deleted in Trichoderma reesei and their biological roles studied. Our results suggest that HFBI has a role in hyphal development and HFBII in sporulation. Sporulating colonies of the Δhfb2 strain were wettable and sporulation was only 50% of the parent strain. Colonies of Δhfb1 showed wettable and fluffy phenotype. In shaken liquid cultures, the hyphae of Δhfb1 were thinner and biomass formation was slower compared to the parent strain while in static liquid cultures no aerial hyphae were formed. Expressing the Schizophyllum commune hydrophobin SC3 in the Δhfb1 strain restored the formation of aerial hyphae.

  • Filamentous fungus
  • Gene deletion
  • Biological function
  • Morphology

1 Introduction

Filamentous fungi secrete moderately hydrophobic proteins, hydrophobins, which share eight conserved cysteine residues. Hydrophobins serve different functions and most of them are based on the ability to self-assemble into an amphipathic membrane at hydrophobic–hydrophilic interfaces. The hydrophobin membrane changes the hydrophobicity of the cell wall resulting in hyphal attachment, maintenance of open gas channels in fruiting bodies and spore dispersal by air [1,2]. The capability of hydrophobins to lower the water surface tension and to cover the emerging aerial hyphae with a hydrophobic membrane enables fungal hyphae to grow into the air from a liquid medium [3,4]. Hydrophobins have also roles in pathogenesis [5,6] and facilitate, for example, eruption of fruiting bodies through surface of host tree [7]. Two classes of hydrophobins have been distinguished based on different solubilities and hydropathy profiles [8]. The aggregates of class II hydrophobins are more soluble than that of class I hydrophobins and can be dissolved, for example, in 60% ethanol.

Three class II hydrophobin genes, hfb1, hfb2 and hfb3, have been isolated from Trichoderma reesei [911]. The hfb1 gene is expressed on glucose and sorbitol-containing media, but not on lactose or complex polysaccharides, such as cellulose [10]. The hfb2 [10] and hfb3 (E. Rintala and T. Nakari-Setälä, unpublished results) genes are expressed on sorbitol, cellulose and lactose-containing media, whereas their expression on glucose is relatively low if detected at all. HFBIII is mainly associated with the vegetative submerged mycelia (M. Linder and T. Nakari-Setälä, unpublished results). HFBI and HFBII have been isolated from the cell walls of vegetative submerged hyphae and aerial spores, respectively, as well as from the culture medium [9,10]. We report here the functional analysis of the hfb1 and hfb2 genes.

2 Materials and methods

2.1 DNA constructs

For deletion of hfb1, plasmid pTNS24 was constructed by cloning a 3.2 kb SphI–XbaI fragment from p3SR2 encompassing acetamidase gene amdS [12,13] into the MunI site of pEA10 [9] between 2.8 and 2.0 kb 5′ and 3′ flanking sequences of hfb1, respectively.

A 1.2 kb HindIII–BglI fragment of hfb2 5′ flanking region was cloned from pTNS8 [10] into the XhoI site of pARO21 [14] resulting in pTNS26. The 1.75 kb EcoRV fragment from pTNS8 encompassing the 3′ flanking region of hfb2 was cloned into the EcoRV site of pTNS26. The resulting plasmid pTNS27 contains the hygromycin resistance cassette (pki1 promoter, hygromycin B phosphotransferase gene hph and cbh2 terminator) in between the hfb2 flanking regions.

For expressing cDNA of Schizophyllum commune SC3 under the control of the 1.4 kb hfb1 and 1.2 kb hfb2 promoters, the plasmids pS3TR1.1 and pS3TR1.2 were used, respectively [15]. Both DNA constructs carry the 0.9 kb hfb1 terminator.

2.2 Fungal transformations and strains

Protoplasts of T. reesei QM9414 (VTT D-74075) [16] and Rut-C30 (VTT D-86271) [17] were transformed as described [18] with an SalI fragment of pTNS24 and an SphI–SpeI fragment of pTNS27. These fragments carry the deletion cassettes for hfb1 and hfb2, respectively. The Δhfb1 strain VTT D-99724 was transformed with the hfb2 deletion cassette to obtain the Δhfb1Δhfb2 strain. Strain VTT D-99724 was also co-transformed with pS3TR1.1 and pARO21 to express the SC3 cDNA. This cDNA was expressed in the Δhfb2 transformant VTT D-99726 by co-transforming this strain with pS3TR1.2 and p3SR2. Transformants were streaked three times on acetamide or hygromycin containing selective media, and transferred to potato dextrose agar for sporulation. Spore suspensions were plated on selective medium to obtain single spore colonies. Fungal transformants (Table 1) were screened by DNA and protein analysis. Another independent transformant of each strain were used to confirm the results reported below.

View this table:
Table 1

Fungal strains

T. reesei
QM9414VTT D-74075[16]
Δhfb1VTT D-99724This study
Δhfb2VTT D-99726This study
Δhfb1Δhfb2VTT D-99725This study
Δhfb1 expressing SC3Q139AThis study
Δhfb2 expressing SC3R26AThis study
Rut-C30aVTT D-86271[17]
Δhfb1aB8AThis study
Δhfb2aVTT D-99676This study
  • aRut-C30-based strains were only used on lactose-containing growth media due to the poor growth of QM9414 on that carbon source.

2.3 DNA analysis

Genomic DNA was isolated using Easy-DNA™ Kit (Invitrogen, The Netherlands). DNA of the hfb1 and hfb2 deletion strains were cleaved with PvuI and NcoI, respectively. These enzymes cut only within the acetamide and hygromycin cassettes, respectively, but not in the hfb1 or hfb2 genes. Genomic DNA of Δhfb1 and Δhfb2 transformants expressing SC3 was digested with HindIII and EcoRI. DNA fragments were separated by agarose gel electrophoresis and blotted onto nylon membranes (Magna Charge Nylon, Micron Separations Inc., USA) using standard procedures. The hfb1 and hfb2 gene fragments containing both coding and flanking sequences and the SC3 cDNA were labelled with digoxigenin (DIG High Prime, Boehringer Mannheim, Germany) and used as probes in Southern hybridisations at 68 °C according to the manufacturer's instructions. The hfb2 fragment was also labelled radioactively with [α-32P]dCTP and used as a probe in Southern hybridisation of the Rut-C30 transformants under stringent conditions [19]. Kodak XAR-5 X-ray film was exposed to the hybridised filters.

2.4 Protein isolation and analysis

To screen for SC3 production, strains were cultivated in shake flasks (50 ml of medium) or in microtiter plates at 28 °C for 5–6 days using Trichoderma minimal medium [18] buffered to pH 6 and supplemented with 3% glucose or lactose. Proteins were precipitated from culture filtrates with 10% trichloroacetic acid [20] and treated with trifluoroacetic acid [21]. After removal of the solvent by evaporation, proteins were dissolved in sodium dodecyl sulphate (SDS)-sample buffer (1% SDS, 5% glycerol, 2.5%β-mercaptoethanol, 0.005% bromophenol blue, 25 mM Tris/HCl buffer, pH 6.8). For Western analysis, 17.5% SDS polyacrylamide gels or slot blotting were used. The SC3 protein was detected with rabbit antiserum raised against SC3 [3]. The antiserum was purified to reduce non-specific binding by incubating in 1–2 w/v% acetone powder [22] of QM9414 for 60–90 min at 4 °C.

2.5 Solid cultivations

Spore suspensions (in 0.8% sodium chloride, containing 0.25% Tween 20 and 20% glycerol) were diluted and plated on Trichoderma minimal medium [18] containing 0.2% peptone and 2% w/v granulated agar (Difco, USA) such that 15–30 colonies per plate were obtained. Glucose or Solka floc cellulose (James River Inc., USA) (20 g l−1) was used as a carbon source and 0.1% Triton X-100 to restrict the growth of the colonies. Morphology of the colonies was examined with a stereomicroscope. Hydrophobicity was determined by placing 3 μl water droplets on colonies grown for 7 days.

2.6 Liquid cultivations

In shaken cultivation, 250 ml of Trichoderma minimal medium [18], buffered to pH 6 and supplemented with 3% glucose or lactose, was inoculated with 5 × 107 spores and cultivated at 200 rpm at 28 °C in a 2 l Erlenmeyer flask for 4–5 days. Samples were collected daily and assayed for dry weight, glucose (GOD-Perid kit, Roche, Germany) and lactose (lactose/β-d-galactose kit, Boehringer-Mannheim, Germany). For determination of hyphal thickness, 900 μl of culture was fixed by adding 100 μl 10 w/v% formaldehyde, diluted with 1% formaldehyde to approximately 1 g biomass l−1, stained for half an hour with lactophenol blue (Fluka, Switzerland) and analysed by light microscopy using analysis software imaging tool (Soft Imaging System).

In static cultivation, 0.4 ml Trichoderma minimal medium [18], buffered to pH 6 and supplemented with 3% glucose or lactose, was inoculated with 2 × 104 spores and incubated at 28 °C in 4 ml glass vials in light for three weeks. For in vitro complementation studies, 31 μl of aqueous solution of HFBI, HFBII or SC3 (0.7 g l−1) was added to the culture medium, water serving as a control. HFBI [23] and SC3 [24,25] were purified as described. HFBII was purified from the culture filtrate (M. Tenkanen, unpublished results) of T. reesei Rut-C30 (VTT D-86271) cultivated on lactose-containing medium in a Chemap LF.7 fermenter (working volume 10 l) essentially as described [26].

2.7 Sporulation assay

1.9 ± 0.2 × 106 spores were plated on potato dextrose agar (Difco, USA) plates. After growing for 7 days at 28 °C in the light, spores were collected with 6 ml 0.8% sodium chloride, containing 0.25% Tween 20 and 20% glycerol. Spore concentration was measured using a haemocytometer.

2.8 Electron microscopy

The mycelial samples were fixed in 0.1 M sodium acetate, pH 7.4, containing 2 w/v% paraformaldehyde and 2.5 v/v% glutaraldehyde (both electron microscopy grade) for 2 h at room temperature. The mycelium was separated by centrifugation at 20,000g for 10 min and washed five times with 0.1 M sodium phosphate, pH 7.4, for 5 min. The samples were post-fixed with 1% OsO4 in water, dehydrated with increasing amounts of ethanol (50%, 70%, 96%, 100%) and embedded in LX112 Epon. Ultrathin sections were stained with uranyl acetate and lead citrate (using Leica electron microscopy stain) and examined in a JEM 1200 EX or a JEM 1200 EX II (Jeol, Tokyo, Japan) electron microscope at 80 kV.

3 Results and discussion

3.1 HFBI has a role in hyphal development

In order to study the biological role of HFBI, the hfb1 gene was deleted in T. reesei strains QM9414 and Rut-C30. QM9414 Δhfb1 was cultivated on the hfb1 inducing (glucose) and repressing carbon source (cellulose) [10]. QM9414 was used as a control. On solid glucose medium, 3 days-old colonies of Δhfb1 were more fluffier than colonies of QM9414 (Fig. 1). The radial growth rate was not affected but formation of aerial hyphae was delayed. After 7 days of growth, morphological differences between Δhfb1 and QM9414 were no longer detected. However, the Δhfb1 strain had an easily wettable phenotype on glucose while QM9414 remained dry. The Δhfb1 and QM9414 strains were not sporulating on glucose medium. On cellulose, no significant difference in sporulation of the Δhfb1 and QM9414 strains was observed and both hyphae and spores of the strains were hydrophobic. Based on these observations, it is concluded that HFBI affects the hydrophobicity of T. reesei hyphae on glucose and affects formation of aerial hyphae in the early growth phase.

Figure 1

Colony morphology of the T. reesei parent strain and hydrophobin deletion strains. T. reesei QM9414 (a), the Δhfb1 (b), the Δhfb2 (c) and the Δhfb1Δhfb2 strain (d) were grown for 3 days on solid glucose-containing medium.

Sporulation efficiency was determined on potato dextrose agar medium, on which T. reesei tends to sporulate extensively. The Δhfb1 strain sporulated more poorly than QM9414 (Fig. 2) indicating that HFBI influences spore formation in T. reesei. HFBI probably affects sporulation through the formation of aerial hyphae. Reduced spore formation has also been reported for a Magnaporthe grisea strains, in which either the class I hydrophobin gene MPG1 or the class II hydrophobin gene MHP1, was inactivated [6,27,28]. In contrast to the mpg1 mutants, conidial viability of the mhp1 mutants was affected, however, the underlying mechanisms are not clear [6,27].

Figure 2

The effect of HFBI and HFBII hydrophobins on sporulation of T. reesei. The parent strain QM9414, the hydrophobin deletion strains Δhfb1, Δhfb2 and Δhfb1Δhfb2 and the Δhfb1 and Δhfb2 strains expressing SC3 were grown on potato dextrose agar plates for 7 days. The values indicate mean and standard deviation of five experiments.

In shaken liquid cultures on glucose, Δhfb1 produced less biomass than QM9414 especially in the exponential growth phase (Fig. 3(a)). One of the two Δhfb1 strains (VTT D-99723) reached the biomass level of QM9414 at the end of the cultivation in contrast to the other (VTT D-99724) probably be due to fluctuation of cell growth (Fig. 3(a)). Light microscope examination revealed that the hyphae of Δhfb1 were thinner than those of QM9414, the difference being most evident after 2 days of growth (Fig. 4(a)). No distinct differences in cell wall thicknesses were seen by transmission electron microscopy (Fig. 4(b)). On lactose, the growth (Fig. 3(b)) and the hyphal thickness of Rut-C30 Δhfb1 were comparable to its parent strain. Note that cellulose could not be used to follow biomass formation because of its insolubility. Moreover, the Rut-C30 background was used because QM9414 grows poorly on lactose. Since HFBI is not produced on lactose [10], another hydrophobin, such as HFBIII, is likely to replace its function. In fermenter cultivations on glucose, the growth properties and the hyphal thicknesses of Δhfb1 and QM9414 were similar [26]. Differences in hyphal thicknesses most likely occur earlier during the 3-day cultivation of inocula. Based on these observations, HFBI seems to have an impact on the overall hyphal thickness of T. reesei. In Aspergillus niger it has been shown that hyphal growth rate and diameter are correlated [29]. This suggests that the absence of HFBI somehow affects the growth rate, possibly by affecting the cell wall composition as has been shown for SC3 of S. commune [30].

Figure 3

Growth of hydrophobin deletion strains of T. reesei in shaken liquid cultivations. (a) The parent strain QM9414 (▪), the Δhfb1 strains (▴) VTT D-99724 (solid line) and VTT D-99723 (dashed line) and the Δhfb1 strain expressing SC3 (▵) were cultivated using glucose as a carbon source. (b) The T. reesei strain Rut-C30 (▪), Rut-C30 Δhfb1 (▴) and Rut-C30 Δhfb2 (•) were grown on lactose-containing media.

Figure 4

The effect of HFBI on hyphal diameter of T. reesei. (a) Hyphal diameter of QM9414, the Δhfb1 strain and the Δhfb1 strain expressing SC3 in shake flask cultures grown on glucose-containing medium for two days. Mycelial samples were fixed in 1% formaldehyde and stained with lactophenol blue dye. The results are represented as mean and standard deviation of 200 experiments. (b) Transmission electron micrographs of the strains QM9414 and Δhfb1 from the same cultivation. The bar represents 500 nm.

The effects of deletion of hfb1 on biomass formation and hyphal diameter in liquid cultures as well as colony morphology were most obvious in the early growth phase of the fungus (after 2–3 days of cultivation) when the production of HFBI had already initiated in the parent strain [10,26,31]. However, effects were less obvious after 7 days of growth. The T. reesei genome [32] contains several other hydrophobin genes apart from hfb1, hfb2 and hfb3. One or more of these genes may take over the function of hfb1 after a few days of growth.

For the escape of fungal hyphae from the medium into the air, surface tension needs to be reduced [4]. Both HFBI and HFBII were shown to be able to reduce the surface tension, at least when measured in water (S. Askolin and H.A.B. Wösten, unpublished results). The role of HFBI in aerial growth of T. reesei was studied in static liquid cultivation. In contrast to QM9414, Δhfb1 produced no aerial hyphae when grown on glucose (Fig. 5). The difference between the strains was observed during whole length of cultivation. Addition of purified HFBI to the medium of Δhfb1 restored the formation of aerial hyphae (Fig. 5). On lactose, both Rut-C30 Δhfb1 and Rut-C30 produced aerial hyphae. Emergence of aerial hyphae on this carbon source is probably aided by another hydrophobin due to absence of HFBI. These results suggest that HFBI enables T. reesei to grow into the air from liquid glucose-containing cultures.

Figure 5

Formation of aerial hyphae in static liquid cultivations. T. reesei QM9414, the hydrophobin deletion strains Δhfb1, Δhfb2 and Δhfb1Δhfb2 and the SC3 expressing Δhfb1 strain were grown on glucose-based medium. The Δhfb1, Δhfb2 and Δhfb1Δhfb2 strains were grown in the absence or presence of 50 mg l−1 purified HFBI or HFBII.

3.2 HFBII is involved with sporulation

The expression of hfb2 on different carbon sources differs from that of hfb1 [10]. The QM9414 Δhfb2, QM9414 Δhfb1Δhfb2 and parent strain were cultivated on cellulose, which induce expression of hfb2, and on glucose, on which hfb1 is expressed but the expression of hfb2 is low if detected at all [10]. On solid cellulose and glucose-containing medium, Δhfb2 and Δhfb1Δhfb2 colonies were hydrophilic, while the parent strain QM9414 was non-wettable. None of the strains sporulated on glucose. On cellulose and potato dextrose agar medium, spore production of the hfb2 deletion strains was reduced (Fig. 2). Simultaneous deletion of both the hfb1 and hfb2 genes reduced the spore formation on potato dextrose even further (Fig. 2). The colony morphology and formation of aerial hyphae of Δhfb2 did not differ from that of QM9414 on glucose and cellulose. Similar to the Δhfb1 colonies grown on glucose for three days, those of Δhfb1Δhfb2 were fluffy (Fig. 1). From these data, it is concluded that HFBII renders aerial hyphae and spores hydrophobic and is involved in aerial sporulation.

No differences were observed in biomass formation (Fig. 3(b)) and hyphal thickness in shaken liquid cultivations of the QM9414 Δhfb2 and QM9414 strain on glucose and the Rut-C30 Δhfb2 and Rut-C30 strain on lactose medium. Also in fermenter conditions on lactose-containing media, the growth properties of Rut-C30 and Δhfb2 resembled each other [26]. These results indicate that deletion of the hfb2 gene has no impact on the biomass formation.

In static liquid cultivation on glucose medium, both QM9414 and Δhfb2 formed white aerial hyphae (Fig. 5) due to the secretion of HFBI to the medium. In contrast, aerial hyphae formation was absent in Δhfb1Δhfb2. Purified HFBII only partly complemented HFBI when added to glucose-containing medium of Δhfb1 and Δhfb1Δhfb2 (Fig. 5). On lactose medium, the Rut-C30 Δhfb2 and parent strain both produced aerial hyphae. Since HFBI is not secreted under this condition, another hydrophobin, such as HFBIII, is expected to compensate the absence of HFBI. These observations show that HFBII has a minor role in mediating escape of hyphae from the substrate into the air.

3.3 SC3 complements the aerial hyphae-deficient phenotype of Δhfb1

To investigate whether the class I hydrophobin SC3 of S. commune can substitute for the class II hydrophobins HFBI and HFBII, Δhfb1 and Δhfb2 strains expressing SC3 under the hfb1 and hfb2 requlatory sequences, respectively, were constructed (Table 1). Expression of SC3 to the culture medium was verified using Western blotting (not shown). Expression level of SC3 in T. reesei under the hfb1 and hfb2 promoters has been estimated to be approximately 60 mg l−1 in similar cultures [15]. Expressing SC3 in the deletion strains did not restore colony morphology, biomass formation (Fig. 3(a)), hyphal thickness (Fig. 4(a)) and hydrophobicity of aerial hyphae and spores. Sporulation of the strains was even poorer than that of the parent strains (Fig. 2). However, the SC3 producing Δhfb1 strain formed aerial hyphae, in contrast to the parent strain Δhfb1, showing that SC3 can partially substitute for HFBI in vivo (Fig. 5). In agreement, when purified SC3 was added to the culture medium the hfb1 deletion strain also formed aerial hyphae (not shown). In contrast, HFBI or HFBII did not replace for SC3 when added in vitro into S. commune cultivations since these class II hydrophobins were proteolytically degraded in the culture medium (S. Askolin and H.A.B. Wösten, unpublished results).

To conclude, these results indicate that HFBI and HFBII play essential roles in fungal growth and development. HFBI facilitates the aerial growth of T. reesei on glucose medium whereas HFBII is suggested to aid spreading of aerial spores. The S. commune hydrophobin SC3 restored only the formation of aerial hyphae in Δhfb1. This shows that despite the common amphipathic nature of hydrophobin molecules specific hydrophobins are tailored to fulfil their functions in a particular fungus (see [2,33,34]).


We thank the technical assistance of Seija Nordberg and Tarja Haajanen in construction of the fungal strains, Karin Scholtmeijer for plasmids pS3TR1.1 and pS3TR1.2 and for help in construction of the SC3 producing strains, Michael Bailey for fermentation and the financial support of the Foundation of Fortum.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
View Abstract