Protein Engineering, Vol. 13, No. 10, 735-738,
October 2000
© 2000 Oxford University Press
Minor structural consequences of alternative CUG codon usage (Ser for Leu) in Candida albicans exoglucanase
Biochemistry Department, School of Medical Sciences, University of Otago, PO Box 56, Dunedin, New Zealand
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In some species of Candida the CUG codon is encoded as serine and not leucine. In the case of the exo-ß-1,3-glucanase from the pathogenic fungus C.albicans there are two such translational events, one in the prepro-leader sequence and the other at residue 64. Overexpression of active mature enzyme in a yeast host indicated that these two positions are tolerant to substitution. By comparing the crystal structure of the recombinant protein with that of the native (presented here), it is seen how either serine or leucine can be accommodated at position 64. Examination of the relatively few solved protein structures from C.albicans indicates that other CUG encoded serines are also found at non-essential surface sites. However such codon usage is rare in C.albicans, in contrast to C.rugosa, with direct implications for respective recombinant protein production.
Keywords: alternative codon usage/Candida albicans/crystal structure/exoglucanase
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The CUG codon normally codes for leucine but in the pathogenic fungus Candida albicans and also in some other species of Candida, it is recognized by an unorthodox seryl-tRNA with a 5' CAG 3' anticodon (Ohama et al., 1993
; Santos et al., 1993). Recent studies have suggested that this reassignment of the genetic code has been driven by selection and is not merely an aberration of nature (Santos et al., 1999). Usage of the CUG codon in C.albicans is, however, a comparatively infrequent event, some genes containing no such codon at all. CUG encoded serine incorporated into protein has been identified from in vitro translation experiments, amino acid sequencing and mass spectrometry (Kawaguchi et al., 1989; White et al., 1995; Zimmer and Schunck, 1995). Replacement of one or more strongly hydrophobic leucine residues by polar serine residues might be expected in some situations to perturb the three-dimensional structure of a protein or alter its surface character, with possible consequences for activity and/or stability. Therefore, in order to produce authentic recombinant protein from a C.albicans gene, it would appear desirable if not necessary to alter any CUG codons to a universal serine specific codon before expression in an appropriate heterologous host such as Saccharomyces cerevisiae or Pichia pastoris, both of which use the universal genetic code.
The EXG gene in C.albicans codes for the precursor of an exo-ß-1,3-glucanase (Chambers et al., 1993a) and contains two CTG triplets corresponding to position –17 in the prepro-sequence of 38 residues and position 64 in the mature protein sequence. The enzyme (Exg) is secreted to the cell wall where it is active in glucan metabolism and, it is thought, in helping to shape the cell wall during morphogenesis (Larriba et al., 1993). Exg, a member of the family 5 glycosyl hydrolases (Henrissat and Davies, 1997), can be isolated directly from the medium of C.albicans cultures. By expressing the gene in an EXG-deficient strain of S.cerevisiae the Leu64 recombinant version of the protein has also been produced (Chambers et al., 1993b). Having recently solved the crystal structure of the recombinant form (Cutfield et al., 1999), we were interested as to how position 64 could accommodate either a non-polar Leu or a polar Ser residue. We also hoped to shed light on the question of whether any selection advantage at the protein level might accrue as the result of alternative CUG codon usage. In this paper we report the three-dimensional structure of native Exg, confirming the presence of a serine at position 64 and allowing us to compare its molecular environment with that of Leu64 in the recombinant variant.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Native Exg was prepared from 10 l of C.albicans (ATCC 10261) culture medium as described previously (Chambers et al., 1993a
), then purified on a Pharmacia HR 5/5 phenyl Superose column under a reverse gradient of ammonium sulphate, 0.6–0.0 M over 30 min. Enzyme activity was assayed as described previously (Chambers et al., 1993b). Diffraction data for native Exg were collected at room temperature to 1.9 Å on a Rigaku R-AXIS II system (Massey University). The structure solution was straightforward as the crystals were isomorphous with the recombinant form (PDB accession code 1CZ1). In order to reduce phase bias, a 20-residue segment of the structure centred around position 64 was omitted from the structure factor calculations. Both 2Fo – Fc and Fo – Fc difference maps were examined and the omitted structure was built using O (Jones et al., 1991). Maximum likelihood restrained refinement was carried out using REFMAC (Murshudov et al., 1997). Family 5 exoglucanase sequences referred to, together with their Genbank accession codes, include the following: Candida albicans Exg (X56556); Saccharomyces cerevisiae Exg1 (M34341), Exg2 (Z46870) and Ssg1 (S52935); Debaryomyces occidentalis Exg1 (Z46871); Kluveromyces lactis Exg1 (Z46869); Yarrowia lipolytica Exg1 (Z46872); and Pichia angusta Exg1 (Z46868). |
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Purification of the enzyme to homogeneity, in both native and recombinant forms, was achieved through hydrophobic interaction chromatography. Under identical elution conditions the native form appeared ~3 min earlier than did the recombinant. The two forms were indistinguishable on denaturing gels and were shown to possess the same specific activity. The same crystallization conditions were successful for both forms, producing orthorhombic crystals (space group P212121); however the native protein crystals took significantly less time to grow than did the recombinant protein and were more reproducible. The crystal unit cell volume was 2% larger for the native Exg crystals. Crystallographic refinement of the native enzyme using data to 1.9 Å (Table I
) produced similar statistics to the recombinant form which had been solved at 1.85 Å resolution (Cutfield et al., 1999). Atomic coordinates for native Exg have been lodged in the protein database under accession code 1EQP. Analysis of the refined structure by PROCHECK (Laskowski et al., 1993) confirmed that the geometry and stereochemistry were of good quality. There was one outlier in the Ramachandran plot, Ser140, also seen in the recombinant Exg model. Backbone dihedral angles associated with Ser64 were within a few degrees of the equivalent angles in the recombinant (Leu64) molecule.
|
Careful examination of difference density around residue 64 in the native structure showed a serine side chain in alternative conformations that arise by means of a simple rotation around the C
–Cß bond (Figure 1a). The two positions for O
have approximately equal occupancy and atomic temperature factors. Both conformations are stabilized by a hydrogen bond involving the serine hydroxyl, in one case to the peptide carbonyl oxygen of Lys60 and in the other to a water molecule which in turn is hydrogen bonded to the peptide carbonyl oxygen of Gln110. In the recombinant protein the Cß–C bond of Leu64 coincides with the alternative conformation of Ser64 but the water molecule has been displaced by the side-chain methyl groups (Figure 1b). Gln68 has also moved slightly to compensate. This leaves Leu64 very exposed, stabilized by only two van der Waals interactions involving CB and CD1.
|
for the side chain of residue 64 in (a) native and (b) recombinant Exg. Note the alternative conformations for Ser64 in the native enzyme. The figure was drawn using BobScript (Esnouf, 1997). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The crystal structure of native Exg is essentially the same as that of the recombinant form produced in S.cerevisiae, except in the region of amino acid residue 64, which is a serine and not a leucine owing to the non-standard translation of the CUG codon in C.albicans. This residue is situated in a four-turn helix (residues 60–74) between the ß1 and ß2 strands of the (
/ß)8 barrel, its side chain directed out from the protein surface (Figure 2a) and well removed from the active site pocket. This would explain why both the native and recombinant forms display similar kinetic behaviour and also why the native form (Ser64) elutes earlier on a hydrophobic column. By forming a hydrogen bond in either of its two conformations the serine is well accommodated at position 64 (Figure 2b and c). From the database of preferred conformations in O (Jones et al., 1991), the arrangement where the serine hydroxyl is hydrogen bonded to the backbone is predicted to be the most probable. This reflects the tendency for the hydroxyl group of a serine in an -helical environment to hydrogen bond to the main chain carbonyl four residues back (Baker and Hubbard, 1984). The alternative conformation is also a predicted conformer and contacts a well-ordered water molecule. In the recombinant protein the side chain of Leu64 is accommodated with some minor solvent rearrangement, although it remains highly exposed (Figure 2d). In its favour, it could be argued, is its known high propensity for residing in an -helix (Koehl and Levitt, 1999). In general, solvent-accessible leucine residues are not uncommon in protein structures and their replacement by serine would not be likely to affect tertiary structure although it would affect the character of a hydrophobic surface patch. It is evident that the second CUG serine in Exg, which is present in the prepro-sequence, does not appear to be crucial for processing, secretion and activation of the enzyme as recombinant protein was successfully produced by S.cerevisiae using the full-length EXG gene.
|
Although it is well known that many surface sites on a protein can accept hydrophobic residues, it is rare for a surface position to remain hydrophobic over the course of evolution (Bowie et al., 1990
). It is relevant then to examine which amino acids appear at the equivalent of position 64 in homologous proteins, i.e. in other yeast/fungal family 5 exoglucanases. The closest known sequence to Exg is the product of the EXG1 gene from the yeast S.cerevisiae, which shares 58% overall identity (Chambers et al., 1993a). This protein possesses a lysine at position 64 while other relatives have Leu, Gln, Glu or Arg, hence there seems to be no specific requirement other than some preference for a polar residue. This suggests perhaps that CUG encoded serines, which are uncommon in C.albicans, may correspond to non-essential surface sites on the protein although more structures need to be examined to substantiate such a proposal. Of the seven other C.albicans protein structures currently lodged in the protein database, five do not involve the CUG codon and the other two, dihydrofolate reductase and old yellow enzyme, each contain just the one CUG in their mRNA. Both of these proteins were successfully expressed as bacterial recombinants and so have leucine incorporated at these sites. They are located near the ends of the polypeptide and, as with Exg, are well away from the active site.
It might be tentatively inferred then that, provided there are only a few CTG triplets in a particular C.albicans gene, it could be overexpressed in an heterologous host as is, without subsequent loss of protein function due to leucine replacing serine. This has implications for structural genomics initiatives involving C.albicans proteins given that there is an urgent need to identify new drug targets in this pathogen (Sternberg, 1994). Additional mutagenesis steps are time consuming, as illustrated in work on the industrially important lipase from Candida rugosa, a species of Candida in which alternative CUG translation is common. In the lip1 gene, 20 out of 47 serines are derived from CUG triplets, including the active site serine, so not surprisingly the heterologous expression of lip1 in S.cerevisiae produced an inactive lipase (Brocca et al., 1998). In order to produce fully functional recombinant protein from a yeast host, the modified gene containing universal codons for all of the 20 serines was completely synthesized. Although it was recognized that not all of these substitutions appeared necessary, it was clearly not worth the effort of identifying and selectively modifying by mutagenesis a subset of `vital' CUG codons.
As far as the C.albicans Exg is concerned, it would seem that there is no obvious selection advantage to the organism at the protein level in having the two serines instead of leucines. Possibly protein stability is marginally improved or perhaps a serine at position 64 is more conducive to favourable interactions within the cell wall environment. It seems more likely, however, that the codon ambiguity in C.albicans has provided a mechanism for adaptation through activation of the stress response as proposed by Santos et al. (1999).
|
Notes |
|---|
2 Present address: Molecular BioSciences Institute, Massey University, Private Bag, Palmerston North, New Zealand
1 To whom correspondence should be addressed. E-mail: john.cutfield{at}stonebow.otago.ac.nz
|
Acknowledgments |
|---|
This work was supported by the Health Research Council of New Zealand.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Baker,E.N. and Hubbard,R.E. (1984) Prog. Biophys. Mol. Biol., 44, 98–179.
Bowie,J.U., Reidhaar-Olson,J.F., Lim,W.A. and Sauer,R.T. (1990) Science, 247, 1306–1310.
Brocca,S., Schmidt-Dannert,C., Lotti,M., Alberghina,L. and Schmid,R.D. (1998) Protein Sci., 7, 1415–1422.[Web of Science][Medline]
Chambers,R.S., Broughton,M.J., Cannon,R.D., Carne,A., Emerson,G.W. and Sullivan,P.A. (1993a) J. Gen. Microbiol., 139, 325–334.
Chambers,R.S., Walden,A.R., Brooke,G.S., Cutfield,J.F. and Sullivan,P.A. (1993b) FEBS Lett., 327, 366–369.[Web of Science][Medline]
Cutfield,S.M., Davies,G.J., Murshudov,G., Anderson,B.F., Moody,P.C., Sullivan,P.A. and Cutfield,J.F. (1999) J. Mol. Biol., 294, 771–783.[Web of Science][Medline]
Esnouf,R.M. (1997) J. Mol. Graphics, 15, 132–134.[Web of Science][Medline]
Henrissat,B. and Davies,G. (1997) Curr. Opin. Struct. Biol., 7, 637–644.[Web of Science][Medline]
Jones,T.A., Zou,J.Y., Cowan,S.W. and Kjeldgaard,M. (1991) Acta Crystallogr., A47, 110–119.
Kawaguchi,Y., Honda,H.,Taniguchi,M.J. and Iwasaki,S. (1989) Nature, 341, 164–166.[Medline]
Koehl,P.and Levitt,M. (1999) Proc. Natl Acad. Sci. USA, 96, 12524–12529.
Kraulis,P.J. (1991) J. Appl. Crystallogr., 24, 946–950
Larriba,G., Basco,R.D., Andaluz,E. and Luna-Arias,J.P. (1993) Arch. Med. Res., 24, 293–299.[Medline]
Laskowski,R.A., MacArthur,M.W., Moss,D.S. and Thornton,J.M. (1993) J. Appl. Crystallogr., 26, 283–291.
Merritt,E.A. and Bacon,D.J. (1997) Methods Enzymol., 277, 505–524.[Web of Science][Medline]
Murshudov,G., Vagin,A. and Dodson,E.J. (1997) Acta Crystallogr. D53, 240–255.
Ohama,T., Suzuki,T., Mori,M., Osawa,S., Ueda,T., Watanabe,K. and Nakase,T. (1993) Nucleic Acids Res., 21, 4039–4045.
Santos,M.A.S., Keith,G. and Tuite,M.F. (1993) EMBO J., 12, 607–616.[Web of Science][Medline]
Santos,M.A.S., Cheesman,C., Costa,V., Moradas-Ferreira,P. and Tuite,M.F. (1999) Mol. Microbiol., 31, 937–947.[Web of Science][Medline]
Sternberg,S. (1994) Science 266, 1632–1634.
White,T.C., Andrews,L.E., Maltby,D. and Agabian,N. (1995) J. Bacteriol., 177, 2953–2955.
Zimmer,T. and Schunck, W.-H. (1995) Yeast, 11, 33–41.[Web of Science][Medline]
Received May 18, 2000; revised July 24, 2000; accepted August 10, 2000.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  W. L. Chaffin Candida albicans Cell Wall Proteins Microbiol. Mol. Biol. Rev., September 1, 2008; 72(3): 495 - 544. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Suzuki, T. Kashiwagi, and K. Hirayama Alternative CUG codon usage (Ser for Leu) in Pichia farinosa and the effect of a mutated killer gene in Saccharomyces cerevisiae Protein Eng. Des. Sel., March 1, 2002; 15(3): 251 - 255. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||