Minireview:

Are Hydrophobins and/or Non-Specific Lipid Transfer Proteins Responsiblefor Gushing in Beer? New Hypotheses on the Chemical Nature ofGushing Inducing FactorsSusanne Hippeli* and Erich F. ElstnerLehrstuhl für Phytopathologie, Labor für Biochemische Toxikologie, Technische UniversitätMünchen, Wissenschaftszentrum Weihenstephan, Am Hochanger 2, D-85350 Freising-Weihenstephan, Germany. Fax: (49) (8161) 714538. E-mail: S. Hippeli@lrz.tum.de

* Author for correspondence and reprint requests

Z. Naturforsch. 57c, 1Ð8 (2002); received July 12/August 2, 2001

Gushing, Non Specific Lipid Transfer Proteins, Hydrophobins

Gushing of beer is characterised by the fact that immediately after opening a bottle a greatnumber of fine bubbles are created throughout the volume of beer and ascend quickly underfoam formation, which flows out of the bottle. This infuriating gushing phenomenon hasbeen, and still is, a problem of world-wide importance to the brewing industry.

It is generally assumed that the causes of malt-derived gushing are due to the use of“weathered” barley or wheat and the growth of moulds in the field, during storage andmalting. We now develop a hypothesis connecting several lines of evidence from differentlaboratories. These results indicate that the fungal hydrophobins, hydrophobic componentsof conidiospores or aerial mycelia, are gushing-inducing factors. Furthermore, increased for-mation of ns-LTPs (non-specific lipid transfer proteins), synthesised in grains as response tofungal infection, and their modification during the brewing process may be responsible formalt-derived gushing.

Introduction

Regarding the causes of gushing, the researchgroup of Carlsberg (Gjertsen et al., 1963; Gjertsen,1967) proposed to divide them into two types:“primary gushing” which occurs periodically andappears to be related to the quality of malt, and“secondary gushing” which is due to faults duringbeer production or to the incorrect treatments ofpackaged beer.

As a result of the continued efforts to clarify thecause of gushing in beer for the past 50 years ithas been shown that the primary gushing is mostprobably caused by mould growth, either beforethreshing (field fungi like Fusarium, Alternaria,Stemphylium, Cladosporium), during storage(storage fungi like Aspergillus, Penicillium or Rhi-zopus) or during malting (Gjertsen et al., 1965;Amaha et al., 1973; Gyllang and Martinson, 1976;Haikara, 1980; Weideneder, 1992; Niessen, 1993;Zepf, 1998).

Prentice and Sloey (1960) and Sloey and Prentice(1962) clearly showed, that especially the Fusaria

0939Ð5075/2002/0100Ð0001 $ 06.00 ” 2002 Verlag der Zeitschrift für Naturforschung, Tübingen · www.znaturforsch.com · D

cause changes in the analytical values of malt sam-ples, such as increase in α-amylase activity, solublenitrogen, extract and gushing. Gjertsen et al. (1965)confirmed in experimental malting tests that inocu-lation of several Fusarium cultures onto barleygrains during stepping could yield gushing malts.The addition to the mash of mycelial extracts or cul-ture filtrate of Fusarium failed to show gushing inbeer. The authors concluded that gushing is notcaused by the Fusarium itself but as result of an in-teraction between Fusarium and the germinatinggrain. Several research groups tried to isolate andto characterise substances from culture filtrates ofdifferent moulds, but also from malt, mash, wortand gushing beer, which could induce malt-derivedgushing (see below), but up to now, the exact chem-ical nature and the physical and chemical mecha-nisms of these substances are unknown.

In this communication new hypotheses on themolecular identity of factors responsible for theprimary gushing based on published data will bepresented.

2 S. Hippeli and E. F. Elstner · Minireview: Hydrophobins, ns-LTPs and Beer Gushing

The “hydrophobin story”

In the 70th Amaha, Kitabatake and co-workerssucceeded in the isolation and characterisation ofa gushing-inducing substance, produced by Nigro-spora sp. no 207 in a liquid culture medium(Amaha et al., 1973; Kitabatake and Amaha, 1974,1977). The authors showed that the “Nigrosporagushing factor” (NGF) must be a hydrophobicpolypeptide with a molecular weight of 16.5 kDaand an isoelectric point of 4.0. NGF is water solu-ble, heat stable, surface active, and resistant to di-gestion by proteases. Amino acid analyses re-vealed that NGF is lacking the four residuesmethionine, histidine, tyrosine and tryptophan, butcontaining 16 cysteine residues forming 8 intramo-lecular disulfide bonds.

According to its molecular and chemical proper-ties it is most likely that NGF is identical to a hy-drophobin dimer (Table 1). Hydrophobins, whichwere discovered in the 90th, are small secreted fun-gal proteins with molecular weights between 7 and20 kDa (Wessels et al., 1991). They are among themost surface-active molecules and self-assemble toan amphiphilic film at any hydrophobic-hydrophi-lic interface. These films are insoluble and exhibita typical ultrastructure of a mosaic of interwovenrodlets (Kershaw and Talbot, 1998; Wösten and deVocht, 2000).

Hydrophobin encoated rodlet layers are com-mon features of aerial structures including mush-room caps, spore surfaces and aerial hyphae. Theyhave been shown to be important in many mor-phogenetic processes, including sporulation, fruitbody development and infection structure forma-

Table 1. Structural and chemical properties of the Nigro-spora-derived gushing factor and a hydrophobin dimerin comparison.

Nigrospora-derived gushing factor or hydrophobin dimer

� Fungal polypeptide� Water soluble� Heat stable� Resistant against proteases� Surface-active� MW: 16.5 kDa or 14Ð40 kDa� 8 intramolecular disulfide bonds� Lack of tryptophan and tyrosine� Lack of methionine and histidine� Secretion into cultur media� No infection process necessary for production

tion. Hydrophobins are therefore likely to beubiquitous in filamentous fungi (Wessels, 1997).

The water soluble monomers contain 8 cysteineresidues forming 4 intramolecular disulfide bondsin a particular manner within their amino acid se-quence. The aromatic amino acids tyrosine andtryptophan are always lacking, methionine andhistidine could likewise be absent. Besides thisstructural similarities hydrophobins and the Japa-nese gushing factor are secreted in submergedliquid cultures without an infection of barley orwheat grains.

In view of the above mentioned data we makethe following assertion: NGF is most likely iden-tical to a hydrophobin dimer. According to theirstrong surface activity hydrophobins are able toinduce gushing in beer. Furthermore, the rodletlayers with their typical rough surfaces may serveas condensation nuclei for the release of CO2-bub-bles. As components of conidiospores or aerialmycelia from field- as well as storage-fungi the en-try of hydrophobins via contaminated grains intobeer is easy to imagine.

The role of “ns-LTPs”

The Japanese succeeded in the isolation of a sec-ond gushing-inducing factor, which they extractedfrom wort (Kitabatake, 1978). The malt sampleused to prepare the wort was a commercial lotproduct in the northern Europe, which was foundto be gushing active. Unfortunately there are nodata, whether the malt was contaminated or notand if, what kind of fungi were present.

The isolated gushing factor was found to bemainly composed of polypeptides (57%) accompa-nied with some portions of glucose (17%). Thesubstance is acidic in nature and the molecularweight was estimated to be approximate 10 kDa.The amino acid composition was different fromthat of NGF and is characterised by extremely lowlevels of basic amino acid residues and compara-tively higher amounts of acidic residues. The gush-ing factor is water soluble, heat stable, surfaceactive and resistant to various enzymes such asproteases and carbohydrases; only Pronase wasable to inactivate the wort factor.

Weideneder (1992) extracted gushing-inducingsubstances (GIS) from wheat malt which wasstrongly contaminated with Fusarium culmorum.

S. Hippeli and E. F. Elstner · Minireview: Hydrophobins, ns-LTPs and Beer Gushing 3

These GIS were shown to be water soluble, heatstable and surface active. They were polypeptideswith molecular weights of about 10 kDa and ex-treme acidic isoelectric points between 1.8 and 3.5.In his comprehensive work Weideneder investi-gated the molecular weight distribution of proteinsextracted from control (uninfected) wheat malt(Fig. 1, lane a) and from wheat grains infected byFusarium (Fig. 1, lane b). The SDS-polyacrylamidegel electrophoretic pattern of the infected malt-extract gave a broad band in the region of 10 kDain comparison to the control, meaning a largequantity of protein. By preparative isoelectric fo-cusing Weideneder (1992) split up the protein ex-tract from contaminated wheat grains in 30 sub-fractions depending on pH (Fig. 2). After dialysisagainst water every subfraction was tested to begushing active by addition of two ml of the dialy-sate to 500 ml bottled non-gushing beer. The gush-ing activity of the bottled beer was then measuredby the method described in detail by Weideneder(1992) and estimated with the volumes (ml) over-flowed from the opened bottle. As shown in Fig. 2only the subfractions with pH values below 4.0exhibited a gushing-inducing potential. Further gelelectrophoretic analyses of the protein-compo-sition of these acidic subfractions revealed that nota single protein but quite a number of proteinswith different isoelectric points were contained inevery subfraction (Fig. 3). Weideneder did notcomment or discuss these findings and he did not

Fig. 1. SDS-polyacrylamide gel electrophoretic patternof protein extracts derived from uncontaminated wheatgrains (lane a) and with Fusarium contaminated wheatkernels (lane b). Experimental conditions and prepara-tion of protein extracts were as described byWeideneder (1992).

Fig. 2. Protein content, pH values and gushing activitiesof subfractions from a protein extract derived fromwheat kernels, infected by Fusarium. The gushing activi-ties of the subfractions were tested by addition of 2 mlof the corresponding dialysates to 500 ml bottled nongushing beer and were estimated with the volumes (ml)overflowed from the opened bottles. Experimental con-ditions and preparation of the protein-extract and thesubfractions respectively were as described by Weiden-eder (1992).

determine the molecular masses of these proteins.Remembering the molecular weight distribution ofthe unfractionated protein extract (Fig. 1, lane b),we conclude that they could possibly be about10 kDa.

Small polypeptides exhibiting properties like theJapanese wort component or the gushing-inducingsubstances from Weideneder are to be found in

Fig. 3. SDS-polyacrylamide gel electrophoretic patternof proteins contained in the acidic subfractions shown inFig. 2. The proteins of every subfraction were seperatedwith the aid of isoelectric focusing (IEF). Experimentalconditions were as described by Weideneder (1992). M:IEF-Marker. The original illustration was not anymoreavailable. This Fig. was scanned what explains the poorquality. The same holds for Fig. 1.

4 S. Hippeli and E. F. Elstner · Minireview: Hydrophobins, ns-LTPs and Beer Gushing

the multigenic family of non-specific lipid transferproteins (ns-LTPs). ns-LTPs are ubiquitous lipidbinding proteins in plants. They have been com-prehensively reviewed by Yamada (1992) or Kader(1996). Two main groups of ns-LTPs, ns-LTP1 andns-LTP2 have been identified with molecularmasses of about 10 and 7 kDa, respectively.

ns-LTP1 are synthesised with a N-terminal sig-nal peptide, which indicates that they follow thesecretory pathway, in agreement with their locali-sation in extracellular layers, i.e. cell walls or cutin.Furthermore, ns-LTP1 are abundant soluble pro-teins of the aleurone layers from grain endo-sperms!

Determination of the three-dimensional struc-tures of ns-LTP1 has shown that these proteinsshare a common folding, stabilised by four disul-fide bonds, and are characterised by a four helixbundle covered by a long C-terminal arm with noregular secondary structure. The loose folding ofthe hydrophobic side chains in the protein corecan produce a cavity, which is able to bind dif-ferent types of monoacylated and diacylated lipidmolecules. Such cavities have been seen in barley,wheat, maize and rice ns-LTP1 (Heinemann et al.,1996; Gincel et al., 1994; Shin et al., 1995; Chavolinet al., 1999; Gomar et al., 1996; Lee et al., 1998).

ns-LTP1 is water soluble, heat stable and sur-face-active adsorbing at air-water interfaces (Subi-rade et al., 1995).

The isoelectric point of ns-LTP1-proteins rangebetween 8.8 and 10, depending on the plantsource. The basic nature of ns-LTP1 is in contrastto the acidic nature of the Japanese wort compo-nent and the gushing-inducing substances fromWeideneder, but this inconsistence will be ad-dressed later.

As already mentioned ns-LTP1 can facilitate thetransfer of a broad range of lipids from donor toacceptor membranes Ð in vitro! The true in vivofunction remains to be established. It was sug-gested that they are involved in the formation ofcutin layers (Sterk et al., 1991). The observation ofantimicrobial activities of ns-LTP1 in vitro led tospeculation of a possible role in the defence ofplants against pathogens (Garcıa-Olmedo et al.,1995). Furthermore, several LTP genes in barleyare upregulated in response to both the infectionby various strains of fungal pathogens as well asabiotic environmental factors such as extreme

temperatures and salt or drought stresses (Garcıa-Olmedo et al., 1995).

However, in brewing technology the formationand stability of foam are important criteria con-cerning beer quality. Sørensen and co-workers(1993) reported that barley LTP1 is involved inbeer foam formation. LTP1 purified from beercreated a good beer foam with high potential in afoam assay. In contrast LTP1 purified from barleydisplayed only poor foaming properties. The au-thors concluded that LTP1 is modified during themalting and brewing processes, whereas the modi-fied LTP1 species is the most important one forthe beer foam potential.

Jegou and co-workers (2000) investigated thechemical modifications of barley LTP1 that occurduring the malting and brewing process, by com-paring the proteins isolated from seeds and beer.In one protein fraction containing barley LTP1 theauthors found a major compound with a molecularmass of 9689 Da and a minor peak with a molecu-lar mass of 9525 Da with the aid of electrospraymass spectrometry technology (Fig. 4a). The for-mer is identical with the theoretical mass of barleyLTP1, the latter corresponds to LTP1 having lostthe C-terminal tyrosine residue via carboxypepti-dase activity in seeds.

The mass spectrum of the LTP1-fraction frombeer showed multiple peaks, whereas the first twopeaks were identical with the barley LTP1 peaks(Fig. 4b). The masses of the other peaks corres-ponded to adducts of one or several units with amass of about 162 Da. Jegou et al. (2000) showedthat these adducts correspond to glucose units thatwere covalently bound upon the protein via Mail-lard-reaction.

The Maillard-reaction is a well known non-enzy-matic reaction which can form covalent linkagesbetween reducing sugars and protein-free aminogroups and is preferred by high-temperature treat-ments during the brewing processes.

Besides glycosylation the authors observedcleavage of disulfide bonds. In case of reductionof all four disulfide bonds the protein becomes to-tally unfolded. Glycosylation increases the solubil-ity, whereas unfolding leads to an exposition of hy-drophobic amino acid residues which are normallyinvolved in the formation of the hydrophobic cav-ity. Both processes increase the amphiphilicity ofmodified LTP1 coming along with a better effi-

S. Hippeli and E. F. Elstner · Minireview: Hydrophobins, ns-LTPs and Beer Gushing 5

Fig. 4. Deconvoluted and recon-structed electrospray mass spectrafrom multicharged ions spectra ofLTP1 from barley seeds (a) and frombeer (b). Experimental conditionswere as described by Jegou et al.(2000).

ciency to form foam in comparison with nativebarley LTP1.

Regarding to the Maillard-reaction reducingsugars react with free amino acid groups of pro-teins as mentioned above. Available groups arethe ε-NH2 of the side chain of lysine residues andthe guanidino amino group of arginine residues.This is a very important point, because lysine aswell as arginine are basic amino acids, which are

Table 2. Amino acid composition of barley lipid transfer protein 1 and kind of modifi-cation.

Amino acid Amount of Modificationbasic/acidic residues

Lysine 4 4 5 neutralisation via Maillard-reactionArginine 5 5 5 neutralisation via Maillard-reactionHistidine 2 2 5 oxidative conversion to aspartic acidGlutamic acid 2 2Aspartic acid 5 5Proline 6 65 oxidative conversion to glutamic acidThreonine 3Serine 8Glycine 9Alanine 4Valine 6Leucine 6Isoleucine 6Methionine 1Cysteine 8Tyrosine 3Asparagine 9Glutamine 4

Total 11 basic/7 acidic residues no basic residues/15 acidic residues

together with the histidine residues responsible forthe basic nature of ns-LTP1. It was shown, thatunder oxidative conditions histidine could convertto aspartic acid and proline to glutamic acid(Brnjas-Kraljevic et al., 1991).

With regard to the amino acid composition ofnon modified barley LTP1 (Table 2) the proteincontains eleven basic and seven acidic amino acidresidues. Modifications via glycosylation and oxi-

6 S. Hippeli and E. F. Elstner · Minireview: Hydrophobins, ns-LTPs and Beer Gushing

dation reveal a partial or total loss of basic resi-dues and an accumulation of negatively chargedresidues. These modifications must be accompa-nied with a shift of the isoelectric point from basicto acidic. One can imagine, that depending onmode and degree of modification quite a numberof LTP1-species could be generated with molecu-lar masses among 10 kDa and multiple acidic iso-lectric points. Remembering the gushing-inducingsubstances of Weideneder (Fig. 3) we assert, thatthe proteins he found in the acidic, gushing induc-ing subfractions must be modified LTP1-specieswith acidic isoeloctric points.

Remembering the chemical properties of theJapanese wort component (MG: 10 kDa; acidic IP;composition: amino acids (57%) and glucose(17%); water soluble; heat stable and surface-active) it seems that these gushing factor is like-wise identical with modified LTP1.

Concluding we make a second assertion: TheJapanese wort component as well as the gushing-inducing substances of Weideneder are most likelymodified ns-LTP1-species.

Modified ns-LTP1-species are normal andwanted ingredients of beer, because they mainlycontribute to foam formation as mentioned above.We assert, that gushing occurs whenever the con-tent of modified ns-LTP1-species exceeds a thresh-old value. Up to now two hypothetical reasons foran oversupply of modified ns-LTP1 would be con-ceivable: During growth and ripening and also dur-ing harvesting and storage barley or wheat grainsare more or less contaminated with a considerablevariety of microorganisms. In response to the infec-tion the LTP1-genes in embryos are upregulated re-sulting in an increased level of defensive LTP1-pro-tein compared to uninfected grains. The synthesisof LTP1 in response to pathogen infestation couldoccur in the field, where especially Fusarium couldexhibit strong LTP1-inducing potentials due togrowing inside the kernels. But also during storageand germination induction of LTP1 could continueor just set on in response to storage fungi or othermicroorganisms. An interesting point is that pri-mary gushing was observed sometimes although no(visible) contamination of the harvest (especiallybarley) could be determined (private information).May be in this special cases the activity of storagemicroflora was sufficient to exceed the tolerableLTP1-levels in the finished beers.

The second possibility in causing elevated LTP1-levels in beer rests on metabolic activities ofmicroorganisms populate the grains resulting inthe release of cell wall bound ns-LTP1, which nor-mally remains in the draff.

Indirect evidence for this second possibility hasbeen offered by Zepf (1998). He infected steriliseddraff as relevant but on nutrient poor mediumwith multiple Fusarium strains to investigate thepotential of these fungi in producing a gushingactive factor.

With the knowledge about the acidic nature ofgushing-inducing substances he analysed the pHvalues of worts prepared from the infected draffsand surprisingly he found not a decrease but anincrease in pH in most of the samples (Fig. 5).Zepf had no idea or explanation for these findingsand there was no further working on this topic.

We assert, that the increase of pH was due tothe release of basic compounds either originatingfrom the fungi themselves or from the nutrientmedium as a consequence of metabolic activityduring fungal growth. In view of the results ofGjertsen et al. (1965), mentioned above andWeideneder (Fig. 1) and regarding the fact, thatnative LTP1 is basic in nature, we assert, that theincrease in pH is due to a release of non-modifiedLTP1 Ð a side effect of destructive metabolic pro-cesses caused by Fusarium.

Fig. 5. pH-values of wort samples which were preparedfrom draff, infected with different Fusarium strains. Un-contaminated draff served as control. Experimental con-ditions were as described by Zepf (1998).

S. Hippeli and E. F. Elstner · Minireview: Hydrophobins, ns-LTPs and Beer Gushing 7

Conclusions

The chemical nature of already described malt-derived gushing inducing factors has been eluci-dated. By studying early publications from the 60th

and 70th , which demonstrated the close connec-tion between mould growth on grains and primarygushing, and very recently published data concern-ing biochemical and molecular properties of sur-face-active proteins, we conclude that the two

Amaha M., Kitabatake K., Nakagava A., Yoshida J. and Kershaw M. J. and Talbot N. J. (1998), HydrophobinsHarada T. (1973), Gushing inducers produced by and repellents: proteins with fundamental roles in fun-some mould strains. Proc. Eur. Brew. Conr., Congr., gal morphogenesis. Fungal Genet. Biol. 23, 18Ð33.Salzburg, 381Ð39. Kitabatake K. (1978), A wort component responsible for

Brnjas-Kraljevic J., Pifat G., Herak J. N. and Jürgens G. gushing in beer. Bull. Brew. Sci. 24, 21Ð32.(1991), EPR evidence for the oxidation-induced for- Kitabatake K. and Amaha M. (1974), Production ofmation of negatively charged species on the low-den- gushing factor by a Nigrospora sp. in liquid culturesity lipoprotein surface. Free Rad. Res. Commun. 14 media. Bull. Brew. Sci. 20, 1Ð8.(5Ð6), 307Ð313. Kitabatake K. and Amaha M. (1977), Effect of chemical

Chavolin D., Douliez J. P., Marion D., Cohen-Addad C. modifications on the gushing-inducing activity of a hy-and Pebay-Peyroula E. (1999), The crystal structure drophobic protein produced by a Nigrospora sp.of a wheat non specific lipid transfer protein (ns-LTP) Agric. Biol. Chem. 41 (6), 1011Ð1019.complexed with two molecules of phospholipids at Lee J. Y., Min K., Cha H., Shin D. H., Hwang K. Y. and2.1 A resolution. Eur. J. Biochem. 264, 562Ð568. Suh S. W. (1998), Rice non specific lipid transfer pro-

Garcıa-Olmedo F., Molina A., Segura A. and Moreno M. tein: the 1.6 A crystal structure in the unliganded state(1995), The defensive role of nonspecific lipid-transfer reveals a small hydophobic cavity. J. Mol. Biol. 276,proteins in plants. Trends Microbiol. 3 (2), 72Ð74. 437Ð448.

Gincel E., Simorre J.-P., Caille A., Marion D., Ptak M. Niessen M. L. (1993), Entwicklung und Anwendung im-and Vovelle F. (1994), Three-dimensional structure in munchemischer Verfahren zum Nachweis wichtigersolution of a wheat lipid transfer protein from multi- Fusarium-Toxine bei der Bierbereitung sowie myko-dimensional 1H-NMR data. Eur. J. Biochem. 226, logische Untersuchungen im Zusammenhang mit dem413Ð422. Wildwerden (Gushing) von Bieren. Dissertation an

Gjertsen P. (1967), Gushing in beer; its nature, cause and der Technischen Universität München.prevention. Brew. Digest 42 (5), 80Ð84. Prentice N. and Sloey W. (1960), Studies on barley

Gjertsen P., Trolle B. and Andersen K. (1963), Weath- microflora of possible importance to malting andered barley as a contributory of gushing in beer. Proc. brewing quality. I. The treatment of barley duringEur. Brew. Conv. Congr., Brussels, 320Ð341. malting with selected microorganisms. Proc. Amer.

Gjertsen P., Trolle B. and Andersen K. (1965), Studies Soc. Brew. Chem., 28Ð32.on gushing caused by microorganisms, specially Fusa- Shin D. H., Lee J. Y., Hwang K. Y., Kim K. K. and Suhrium spec. Proc. Eur. Brew. Conv. Congr., Stockholm, S. W. (1995), High resolution crystal structure of the428Ð438. non specific lipid transfer protein from maize seed-

Gomar J., Petit M.-C., Sodano P., Sy D., Marion D., lings. Structure 3, 189Ð199.Kader J.-C., Vovelle F. and Ptak M (1996), Solution Sloey W. and Prentice N. (1962), Effects of Fusariumstructure and lipid binding of a non specific lipid isolates applied during malting on properties of malt.transfer protein extracted from maize seeds. Protein Proc. Amer. Soc. Brew. Chem., 25Ð29.Sci. 5, 565Ð577. Sørensen S. B., Bech L. M., Muldbjerg M., Beenfeldt T.

Gyllang H. and Martinson E. (1976), Studies on the and Breddam K. (1993), Barley lipid transfer pro-microflora of malt. J. Inst. Brew. 82, 350Ð352. tein 1 is involved in beer foam formation. Tech. Quar-

Haikara A. (1980), Gushing induced by fungi. In: Mono- terly Master Brew. Assoc. Am. 30, 136Ð145.graph VI Ð Relationships between Malt and Beer, Sterk P. Booij H., Schellekens G. A., van Kammen A.European Brewery Convention, pp. 251Ð258. and de Vries S. C. (1991), Cell specific expression of

Heinemann B., Andersen K. V., Nielsen P. R., Bech the carrot EP2 lipid transfer protein gene. Plant CellL. M. and Poulsen F. (1996), Structure in solution 3 (9), 907Ð923.of a four-helix lipid binding protein. Protein Sci. 5, Subirade M., Salesse C., Marion D. and Pezolet M.13Ð23. (1995), Interaction of non specific wheat lipid transfer

Jegou S., Douliez J.-P., Molle D., Boivin P. and Marion D. protein with phospholipid monolayers imaged by fluo-(2000), Purification and structural characterisation of rescence microscopy and studies by infrared spectro-LTP1 polypeptides from beer. J. Agric. Food Chem. scopy. Biophys. J. 69, 974Ð988.48, 5023Ð5029. Weideneder A. (1992), Untersuchungen zum malzverur-

Kader J.-C. (1996), Lipid transfer proteins in plants. sachten Wildwerden (Gushing) des Bieres. Disserta-Annu. Rev. Plant Physiol. Mol. Biol. 47, 627Ð654. tion an der Technischen Universität München.

groups of small, cysteine-rich, water soluble, heatstable and surface-active polypeptides, the fungalhydrophobins and the plant-derived ns-LTPs, areresponsible for malt-derived gushing in beer.

Acknowledgement

This work was financially supported by the“Wissenschaftliche Station für Brauerei in Mün-chen e. V.”.

8 S. Hippeli and E. F. Elstner · Minireview: Hydrophobins, ns-LTPs and Beer Gushing

Wessels J. G. H. (1997), Hydrophobins: Proteins that Wösten H.A:B: and de Vocht M. L. (2000), Hydropho-change the nature of the fungal surface. Adv. Micro- bins, the fungal coat unravelled. Biochim. Biophys.bial Physiol. 38, 1Ð45. Acta 1469, 79Ð86.

Wessels J. G. H., de Vries O. M. H., Asgeirsdottir S. A. Yamada M. (1992), Lipid transfer proteins in plants andand Schuren F. H. J. (1991), Hydrophobin genes in- microorganisms. Plant Cell Physiol. 33, 1Ð6.volved in formation of aerial hyphae and fruit bodies Zepf M. (1998), Gushing-Ursachenfindung anhand vonin Schizophyllum. Plant Cell 3 (8), 793Ð799. Modellversuchen. Dissertation an der Technischen

Universität München.