SEARCH

SEARCH BY CITATION

Keywords:

  • Chitin synthase;
  • Fungi;
  • Fungal evolution

Abstract

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References

Chitin, the structural component that provides rigidity to the cell wall of fungi is the product of chitin synthases (Chs). These enzymes are not restricted to fungi, but are amply distributed in four of the five eukaryotic ‘crown kingdoms’. Dendrograms obtained by multiple alignment of Chs revealed that fungal enzymes can be classified into two divisions that branch into at least five classes, independent of fungal divergence. In contrast, oomycetes and animals each possess a single family of Chs. These results suggest that Chs originated as a branch of β-glycosyl-transferases, once the kingdom Plantae split from the evolutionary line of eukaryotes. The existence of a single class of Chs in animals and Stramenopiles, against the multiple families in fungi, reveals that Chs diversification occurred after fungi departed from these kingdoms, but before separation of fungal groups. Accordingly, each fungal taxon contains members with enzymes belonging to different divisions and classes. Multiple alignment revealed the conservation of specific motifs characteristic of class, division and kingdom, but the strict conservation of only three motifs QXXEY, EDRXL and QXRRW, and seven isolated amino acids in the core region of all Chs. Determination of different structural features in this region of Chs brought to light a noticeable conservation of secondary structure in the proteins.


1Introduction

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References

Fungal cells are protected from the harsh conditions and the difference in osmotic pressure of the environment by an external layer made mostly of carbohydrates, the cell wall. This layer is considered to be a composite made by two types of components: (i) an amorphous matrix constituted by glycoproteins, polysaccharides, pigments, inorganic salts and lipids associated by non-covalent and in some cases covalent bonds; and (ii) microfibrils from a structural polysaccharide. The extraordinary mechanical properties of the cell wall are the result of this arrangement [1]. In most of true fungi the structural microfibrillar component responsible for cell wall rigidity is chitin, a polysaccharide made of N-acetylglucosamine (GlcNAc) units joined through β-1–4 glycosidic linkages. Although present in variable proportions, and sometimes a minor component as it happens in some yeasts, the compound is indispensable for the construction of the cell wall, and therefore for fungal survival under the normal hypotonic conditions found in nature and in artificial media [2]. Mutations that eliminate the capacity to synthesize chitin are lethal [3,4].

Chitin is the product of chitin synthases (Chs), enzymes from which we know more about the genes that encode them (CHS genes), than from the proteins themselves [5]. Chs utilize the nucleotide UDP-GlcNAc as sugar donor, and require a divalent metal ion for activity [2]. Studies with fungi have revealed that most (possibly all) of them contain more than one CHS gene. In other context we have discussed on the significance of CHS gene multiplicity, and suggested that it represents a valuable mechanism that provides fungi with the necessary plasticity required to survive during colonization of different ecological niches [6]. This property appears to have been preserved during evolution because it constitutes a trait favoring selection. In the present review we discuss the phylogenetic relationships of Chs, their probable evolution in the light of their distribution in several kingdoms, and some ideas regarding the conservation of their structure and catalytic domains.

2Classification of Chs

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References

Early studies on the relationships of fungal Chs, deduced from the sequence of fragments obtained by PCR, revealed the existence of three groups of these enzymes in fungi [7]. Further analyses performed with a larger group of whole or fragmentary Chs extended the suggested number of groups to five [8,9]. The results obtained clearly revealed the absence of a correlation between Chs similarities and the taxonomic relationships of fungi. Now, with a larger number of Chs sequences available from fungi, oomycetes and animals, we have proceeded to analyze the relationships among them. For this analysis we included only fungal Chs whose sequences are fully known. In the case of class V, animals and oomycetes where only a few genes have been completely sequenced, we included also the isolated fragments. Our results, obtained by the Clustal method, are shown in Fig. 1.

image

Figure 1. Maximal parsimony dendrogram displaying grouping of Chs from fungi, oomycetes and animals. Two divisions and five classes of fungal enzymes are recognized. Fungi: Ab, Agaricus bisporus; Arb, Arthroderma benhamiae; Af, A. fumigatus; Aq, Ampelomyces quisqualis; Bg, Blumeria graminis; Ca, C. albicans; Ci, Coccidioides immitis; Ed, Exophiala dermatitidis; En, Emericella nidulans; Fn, F. neoformans; Gv, Glomus versiforme; Nc, Neurospora crassa; Onu, Ophiostoma novo-ulmi; Pb, Paracoccidioides brasiliensis; Pc Penicillium chrysogenum; Pg, Pyricularia grisea; Phb, P. blakesleeanus; Pn, Phaeosphaeria nodorum; Rc, Rhizomucor circinelloides; Rm, Rhizopus microsporum; Sc, S. cerevisiae; Sp, Sch. Pombe; Tm, T. magnatum; Um, Ustilago maydis. Oomycetes: Aa, Achlya ambisexualis; Pca, Phytophthora capsici; Sm, Sa. monoica. Animals: Aea, Aedes aegypti; Bm, Boophilus microplus; Brg, Brugia malayi; Cb, Chrysomia bezziana; Cf, Ctenocephalides felis; Di, Dirofilaria immitis; Dm, Drosophila melanogaster, Hz, Helicoverpa zea; Lc, Lucilia cuprina.

Download figure to PowerPoint

Regarding only fungal enzymes, several aspects shown in this figure are worth to be discussed. First of all is the corroboration that Chs and fungal taxonomy are not related subjects. In the second place comes the observation that a clear division into two groups preceded the further separation of enzymes into the formerly described five classes. In order to conserve the original classification because of its widespread acceptance, we suggest to call these larger groups divisions. Division 1 includes classes I, II, and III, and division 2, classes IV and V. In the third place it may be noticed that, whereas classes I, II and III are clearly recognizable, separation of classes IV and V is more difficult to asses. This problem was clarified by alignment of fungal Chs sequences from classes IV and V only (Fig. 2). Lastly, it must be indicated that some classes appear to branch into defined subclasses. This occurs for class II enzymes that separate into two branches: one grouping enzymes from ascomycetes and basidiomycetes, and the other enzymes from zygomycetes only. The same happens in class I enzymes where Chs1 from Saccharomyces cerevisiae (that interestingly was the first Chs to be described [10] and constituted the class I prototype) and Chs2 from Candida albicans form a separate branch. In the case of these two enzymes this anomaly perhaps is due to their abnormal size, quite in excess of other enzymes belonging to division 1. Chs1 of Filobasidiella neoformans appears in the dendrogram in the limits between classes IV and V. Analysis of its conserved motifs (see below), and alignment of enzymes from classes IV and V only (Fig. 2), indicated that it belongs to class V.

image

Figure 2. Dendrogram showing grouping of fungal Chs belonging to classes IV and V. Same abbreviations as in Fig. 1.

Download figure to PowerPoint

The suggested existence of two Chs divisions is further supported by other differential characteristics of the corresponding enzymes. One of them is the relative molecular size of the proteins. Chs belonging to division 1 are smaller than division 2 enzymes. Statistical analyses of them gave the following results for division 1: min. 738 amino acid residues; max. 1131; mean, 910; S.D. 88.5. For division 2 the values were the following: min. 745; max. 1869; mean, 1187; S.D. 225. Another difference refers to the position of the pentapeptide QXRRW, the ‘signature sequence’ of Chs. This is located more towards the C-terminus in enzymes belonging to division 2. Statistical analyses for the position of the glutamine residue gave the following results for division 1 enzymes: min. amino acid residues 396; max. 756; mean, 544; S.D. 68. And for Chs from division 2: min. 372; max. 1548; mean, 952; S.D. 239. Conservation of different motifs in each division (see below) also supports the idea for separation of Chs into divisions.

By multiple alignment of the known sequences, we proceeded to classify fragmentary and complete Chs whose allocation to the different classes was unknown or uncertain. The results obtained indicated that zygomycetes, ascomycetes and basidiomycetes contain representatives from both divisions and the five classes of Chs. The sole exception was class III, without representatives in zygomycetes. It is possible that this may be due to the fact that not enough species have been thoroughly analyzed up to now. Nevertheless, when we proceeded to analyze the classification of the 10 Chs (mostly fragments) described in Phycomyces blakesleeanus[11], we found that four of them belonged to class II, one to class IV, and five to class V, but no representative of classes I and III were present.

Regarding Chs from animals and oomycetes, several aspects can be highlighted. In contrast to fungi, most (probably all) of which contain more than one Chs, in animals and oomycetes only one CHS gene has been detected per species. It is possible that this is because not enough analyses have been performed. But two pieces of evidence argue against this possibility: (i) all the enzymes from each group display high similarities among themselves; and (ii) all of them are clearly different to the fungal enzymes and constitute coherent groups in the dendrogram. This feature is more noticeable for Chs from animals.

3Conserved motifs and amino acids in Chs

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References

Multiple alignment of Chs established four kinds of hierarchical similarities: (i) sequences specific to each class; (ii) those common to each division; (iii) sequences conserved in each one of the different kingdoms represented; and finally (iv) those sequences conserved in all Chs. Class-specific and common sequences are presented in Table 1. Some of the conserved motifs had been noticed previously from analysis of a lower number of Chs [12]. It is noticeable that class I enzymes have no specific motif, and class II Chs have only a single short specific motif. Enzymes from these two classes share all their motifs among them and with class III enzymes. On the other hand class III Chs have several specific motifs, suggesting that this class may have evolved at a later period than the other ones.

Table 1.  Specific and common sequences of Chs
Specific sequencesCommon sequences
Class IDivision 1
NoneE F (T/ASK) X (M/L) (R/T) Y X A (A/VC) T (C/V/S)
 T (M/Y/S) Y N E (D/E/N)
Class IIW X K (I/V) X V X X (V/I) X D G
GXGPLF (C/V) (L/M) K (E/Q/A) X N X K K (I/L) N S H (R/L) W
 (L/I) (L/I/V) (D/E) (A/V/C) G T
 N P L V (A/Y) X Q N F E Y K (MI/L) S N I L D K (P/T) (L/TV) E S X (F/M) G (Y/F/H) (I/V) (S/T) V L P (G/A) A (F/L) (S/C) A Y R
Class IIIN M (Y/F) L A E D R I L C (F/W/Y) (E/D) (L/V) (V/A)
(D/E)YPVP(S/T)(A/P)I(Q/L)NAT D V P
RTLHGVM(Q/L)N(I/V)RDIE (F/L) (I/V) X Q R R R W (L/I) N (G/Q) X (F/L/M) (F/A) A
LNPE(V/I)CDivision 2
PLEQYFHG(P/A) G N R G K R D S Q
RMFFD A D T K (V/R)
WF(S/A)LAC G E T
LQF(I/V)LALGNRPKQ V (Y/F) E Y (Y/F) (I/V) S H (H/N) X X K (A/S) F E (S/A) X F G (G/S) (V/I) T C L P G (C/R) F
FRT(R/S)LVT L H X K N L L (H/S) L G E D R (Y/F/E) L (S/T)
 S Q (R/G) R R W I N S T (V/I) H N L
Class IVD D F (S/T) W G X T R
V(G/A)(F/Y)(L/I)TFGF 
TLDS(L/I)(A/S)XTDYP(N/S)SHKAll fungi
 I, Q X X E Y…L P (G/A)
KRHN(M/K)(A/C)(K/Q)(I/V)Y(A/T)G(F/Y)YII, L (A/G) E D R X L
QQR(V/I)P(M/I)III, Q (R/G) R R W (L/I) N
TFP(K/T)RK 
DLCGTFCFSM(Q/R)FAll Chs
Class VI, Q X X E Y
Y(S/D)D(R/K)RKLII, E D R X L
GL(Q/P)AIIII, Q X R R W

Motifs specific of class and separate divisions were assessed by alignment of full sequences. On the other hand, when analysis of all fungal Chs was attempted, loss of alignment of the most notorious conserved motifs, mainly the pentapeptide QXRRW was observed, probably because of the large difference in size (see above). In order to recognize conserved domains common to all Chs, we started reducing the length of the polypetides used for the analyses, watching for the conservation of alignment of the conserved motifs. Finally we ended with fragments of ca. 250 aa, where the QXRRW pentapeptide appeared located around aa residue number 150. A dendrogram obtained with these sequences gave results nearly identical to those obtained with the whole sequence. Alignment of these fragments revealed the motifs and conserved amino acid residues common to fungal and all Chs. Fig. 3 shows the results obtained with sequences from species representative of each fungal class and animals, and the only oomycete Chs whose sequence is fully known, Saprolegnia monoica. It must be recalled that the QXRRW sequence is absent in other Chs gene products from S. cerevisiae, known to play a role in regulation or structural organization of the catalytic polypeptide, but not carrying the active site themselves [5]. Its absence in Chs2 from Schizosaccharomyces pombe makes us suggest that also it does not code for a Chs catalytic polypeptide, but has other functions instead.

image

Figure 3. Alignment of the core region of Chs representative of each fungal class, oomycetes and animals. Amino acid residues conserved in at least four enzymes are shadowed. Same abbreviations as in Fig. 1.

Download figure to PowerPoint

As indicated above, the glutamine moiety of the ‘signature’ appears as the mean aa residue 544 in Chs belonging to division 1, and 952 in division 2 enzymes. In animals and oomycetes, position of the glutamine residue of the QXRRW pentapeptide relative to the length of the enzymes is in the order of 0.50–0.60, resembling more division 1 enzymes rather than enzymes belonging to division 2 (see above). It is interesting that only two exceptions to the canonical sequence of the pentapeptide (QRRRW) are known. Both are limited to the first arginine: QGRRW in a Chs from the ascomycete Tuber magnatum, and QARRW in the enzyme from the oomycete S. monoica. Incidentally, it may be noticed that the corresponding changes are not minor ones regarding structure, since they represent alterations of an extremely basic and structured domain, by a small neutral amino acid that permits more flexibility to the polypeptide.

The other conserved sequences and amino acids identified in all Chs are the following: QXXEY located at aa residues 650–654; LPG at aa residues 674–676, and EDRXL at aa residues 716–720. These and all further aa positions are referred, except when indicated otherwise, to aa residues from S. cerevisiae Chs1, taking into consideration that ScCHS1 was the first cloned CHS gene [10]. Absolutely conserved amino acids in all Chs isolated thus far are the following: G629, K662, E665, G669, R682, T744, and P747 (Fig. 3).

As noticed above, it appears that Chs belonging to division 1 form a more homogeneous group than division 2 enzymes. This assertion is supported by comparison of the central fragments of ca. 250 aa from different Chs. The following aa residues are conserved in division 1 Chs, whereas the corresponding aa residues in Chs belonging to division 2 are variable: NPLV644-647, NILDK649-653, YF694,695, F708, NM711,712, C721, E723, K727, W732, and L734. The opposite, conservation in division 2 enzymes only, occurs for a minor number of amino acids. These are the following: T672, which corresponds to T or S in other division 1 enzymes, but is always T in enzymes belonging to division 2; S723, which is occupied by A in division 2 enzymes; I754, corresponding to L in division 2 enzymes, A767, which is N in enzymes from division 2, and AF830,831, occupied by GL in all Chs belonging to division 2.

4Conserved structure and the active site of Chs

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References

It may be said that in general Chs have scant similarities with other proteins. Search for proteins with homology to any Chs in genomic banks gives only significant positive results with other Chs. Also use of programs that look for short sequences of homology or structural motifs, such as Pfam do not recognize any common homolog in the genomic banks.

Different approaches employed to determine the structure of the active site of Chs have provided results that agree only in part, or are even conflicting. Hydrophobic cluster analysis (HCA) of several β-glycosyl-transferases, including S. cerevisiae Chs1, revealed the existence of two conserved domains. One of them located at the N-terminal half (A), and the other one at the C-terminal half (B). Two aspartic residues located in the A domain and another one at B domain plus the motif QXXRW were suggested to be involved in the activity of the enzymes [13]. The conservation of these three aspartic residues, and the QXXRW motif, has been confirmed in other β-glycosyl-transferases, and their importance in catalysis has been pointed out [14]. Crystallographic analysis of a β-glycosyl-transferase from Bacillus subtilis provided evidence that the three aspartic residues were involved in binding of the UDP-sugar together with a divalent cation [15].

The QSLAVE program was employed in order to search for a possible fold in Chs. Comparison of the consensus obtained from several fungal Chs fragments with 94 homologous families of different folds [16] gave results suggestive that Chs belonged to the (β/α)8-barrel fold family. With the knowledge of the structure of some enzymes belonging to this family, a three-dimensional model for the catalytic domain of Chs was obtained using human aldolase as template. The model indicated that the catalytic center involved the following five amino acid residues, which were invariant in the enzymes analyzed: Thr88, Tyr90, Asn92, Lys219, and Asp256 (an error in the original paper of the authors, it must be Asp258). These residue numbers corresponded to a fragment from Rhizopus oligosporus Chs1 spanning aa residues 101–500 [16]. In turn these correspond to amino acids T453, Y455, N456, K572 and D602 from ScChs1.

It is important to indicate that the aa residues TXYN, and K are conserved only in division 1 and Saprolegnia Chs, but not in enzymes belonging to fungal division 2 or animals. Another cluster made by amino acid residues Glu400, Asp401, Leu404, Gln440, Arg443, and Try444 also appeared to be important. The first three aa residues of this cluster correspond to the EDRXL motif described above, and the last three aa residues correspond to the QXRRW sequence. Although agreeing with the same fold superfamily, different results were obtained by modelling of sequences of two Chs and other β-glycosyl-transferases, and comparing with the UDPGlc-binding domain of the phage T4 DNA-modifying β-glucosyl-transferase [17]. According to the authors the amino acids that made direct contact with the UDP-sugar were Asn215 (Asp in all others), Tyr 261, Arg269 (or Lys), and Glu272 (or Gln). These residues correspond to D441, Y494, K502, and E505 of ScChs2. Trp341, corresponding to the QRRRW motif also was considered to be important. As described above, multiple alignment of Chs fragments revealed that these amino acid residues are conserved in all Chs, giving credence to their suggested role in catalysis.

By use of directed mutagenesis of Chs2 from S. cerevisiae, it was concluded that two aspartic acid residues, Asp441, and Asp562, and Gln601, Arg604, and Trp605, constituted the active site of the enzyme [18]. Our results show that Asp441 (or Asn) and Asp 562 residues (this corresponding to the EDRXL motif), and the last three aa belonging to the QRRRW motif, are conserved in all Chs. Further experiments of mutagenesis by these authors [19] at the C-terminus of the enzyme led them to suggest that Asp800, Trp803, and Thr805 were conserved in Chs, and were involved in the active center, and that the three aspartic acid residues provided the polar residues necessary for catalysis. Nevertheless, alignment results provide evidence that these residues are not conserved in all Chs proteins. In the case of Chs1, and Chs3 of the yeast, mutation in any of the Arg residues of the QRRRW motif led to loss of function [20]. All these conflicting data indicate that not enough information has been collected yet to obtain the three-dimensional structure of Chs, and identify with absolute certainty the amino acid residues present in the active site of the enzyme.

Taking into consideration, (i) that Chs from all classes belonging to the different organisms perform a similar transglycosidase reaction, (ii) that Chs are ancient enzymes whose catalytic activity has been conserved during evolution, and (iii) that out of the indicated motifs and scant conserved aa residues, Chs possess reduced homology in their amino acid sequences, when all of them are compared; we utilized an alternative approach to understand the conservation of function in Chs. For this, we started from the location of the absolutely conserved motifs and amino acids in all Chs (see Fig. 3). Several interesting issues should be noticed from our analyses: (i) No conserved amino acid residues were observed beyond the QXRRW motif. This result is against the suggestion from Yabe et al. [19] (see above). (ii) Apparently the first Arg residue of the QRRRW motif is not required for activity, since it is altered in at least two Chs. (iii) Although only one Asp residue is absolutely conserved in all Chs, two other Asp residues that may be substituted by Asn in some enzymes could be identified: one at position 602 (Glu in UmChs3), and another one at position 745 in division 1 enzymes and S. monoica. This aa residue may correspond to an Asp present three aa residues beyond in Chs from division 2, and another one present four aa residues before in Chs from animals (see Fig. 3). The sole exception is ChsD from Aspergillus fumigatus that contains a Gln residue instead. Incidentally it is interesting to note that this enzyme departs from the rest of Chs in relation to conservation of important different aa residues, raising doubts as to whether it is indeed an active polypeptide.

It may be observed that the data obtained by other authors, as described above, partially agree with our results in regard to the possible role of different amino acid residues in catalysis, or at least in the conservation of the protein structure required for activity. Trying to reconcile conflicting data, we may summarize the amino acid residues suggested to be important for activity in all the described reports (unfortunately agreeing only in part), that coincide with their conservation in all the Chs that we have analyzed. These include at least some aa residues of the QXRRW motif. Others are the aa residues at the following positions: Asp/Asn602, Tyr654 (belonging to the QXXEY motif), Lys662, Glu665, Glu716, Asp717, Leu720 (these three from the EDRXL motif), and Asp/Asn745. The three Asp or Asn residues indicated above may play the role ascribed to the Asp triad in the function suggested for β-glycosyl-transferases [13,14].

With this information it might be possible to reconsider the three-dimensional structure of the core region of the proteins, where most of their similarities and the putative catalytic domains are found (see above). It is significant that search of analogues other than Chs to these polypeptide fragments in data banks gave again negative results. Comparison of the fragments by different parameters using the Protean software package, revealed that although the aa sequences of all the fragments might be difficult to conciliate, they conserved remarkable similar structural features. Results obtained with five Chs fragments, representative of the five fungal enzymatic classes are schematized in Fig. 4. It is known that Chs are proteins with a high number of transmembrane domains. By use of the Emini surface probability plot, it was observed that fungal Chs belonging to different classes had similar profiles, more noticeably among members of each division. In all of them, the QXRRW pentapeptide showed a high probability to be located at the surface (Fig. 4B). Also similar among Chs were the patterns of Kyte–Doolitle hydrophilicity plots (Fig. 4A). In all of them the pentapeptide appeared in a short hydrophilic stretch, catalogued as a flexible region by the Karplus–Schulz method, followed by hydrophobic domains. Analysis of several ca. 250 aa fragments, selected from the different Chs classes for the presence of transmembrane segments by TMpred, indicated the existence of three transmembrane helices in all cases. Two of them were located towards the N-termini, and one towards the C-termini of the Chs fragments in relation to the pentapeptide location.

image

Figure 4. Hydrophilicity (A) and surface probability (B) plots of core fragments from Chs representative of each fungal class. Plots were adjusted according to the position of their QXRRW motif. Amino acid positions are indicated in the abscissa at the upper part of (A). Same abbreviations as in Fig. 1.

Download figure to PowerPoint

Finally, analysis of secondary structure prediction was performed by means of the Psi-Pred method [21], utilizing the same five fragments indicated above. Analysis of similarities among the predicted structures was made by hand alignment, dividing the fragment in stretches of ca. 50 aa residues, each one aligned in relation to the conserved motifs: QXXEY, LP(G/A), LXEDRXL, and QRRRW, respectively (see above). The results were highly instructive (Fig. 5). It was observed that in all Chs classes the different conserved aa sequences maintained the same structural features and the same relative location among themselves. For example, the QXXEY domain was located at the second helix of the fragments, the LP(G/A) triad appeared in the fifth coil, the LXEDRXL motif was located at the start of the sixth helix, and the pentapeptide was located at the end of the seventh helix, surrounded by two coil regions. Not only the structure of the domains was conserved, but also the relative position of conserved amino acid residues located around each motif: G629, K662, E665, G669, L714, L720, T744 and P747. It may be stressed that all the motifs and single aa residues are also conserved in Chs from animals and oomycetes. The only exceptions are L674 substituted by Ser in Aedes, and L714 that is absent in Chs from animals.

image

Figure 5. Scheme of the predicted secondary structure of the core fragments from Chs representative of each fungal class. Plots were hand-aligned around each one of the conserved motifs (see Table 1). Conserved amino acids are represented in larger type. H, helix, C, coil, E, extended sheet. Same abbreviations as in Fig. 1.

Download figure to PowerPoint

Globally the results indicate a noticeable conservation of the protein structure, independently of the changes occurring at the aa level. This feature suggests that the relative location of certain key aa residues, as well as the peptide secondary structure, have been conserved during evolution and are responsible for the preservation of catalytic activity in Chs belonging to organisms very distant from an evolutionary point of view.

5Evolution of Chs

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References

Evolution of Chs must be analyzed considering three main points: (i) the conservation of only a limited number of domains and amino acids in the different enzymes; (ii) their presence not only in fungi, but also in invertebrates, oomycetes, algae, and some protists; (iii) the existence of several different classes of Chs in fungi, contrasting to a single group in the studied animals and oomycetes (see Fig. 1); and (iv) the lack of relationship between Chs similarities and the fungal phylogenetic relationships.

Comparison of the location of the motifs conserved in all Chs reveals that they are restricted to a short peptide stretch in the different classes of fungal Chs (Fig. 6). The region beyond this stretch towards the C-terminus of Chs belonging to division 2 is very short, contrasting with division 1 enzymes. The opposite occurs in the N-terminal region. Class-specific motifs extend through a short stretch, similar in length in all enzymes. With a single exception, division II enzymes do not posses class-specific domains along two thirds of the whole polypeptide length (Fig. 6). Accordingly, it may be speculated that the event that gave rise to separation of both divisions from the parental enzyme was the acquisition or loss of the N-terminus of the parental enzyme by a new family of Chs. One upsetting result is the apparent split of Chs into fungal divisions preceding separation of Chs from animals (see Fig. 1). Nevertheless, the fact that similarities among consensus sequences of fungal classes is significantly higher than the similarities of each one with animal enzymes indicates that the former may be due to some anomaly or else, less likely, that the animal precursor lost one of the enzyme divisions later on.

image

Figure 6. Schematic comparison of Chs and the location of consensus sequences. Horizontal boxes represent the relative average size of Chs belonging to the different fungal classes. Short vertical lines correspond to the relative position of the conserved motifs in each class. Chs were aligned according to the conserved motif QXRRW (III). Long vertical lines indicate the position of the three universal motifs QXXEY (I), EDRXL (II), and QXRRW (III) in class I Chs. Crosses correspond to the common motifs I and II from each Chs class.

Download figure to PowerPoint

Modern molecular techniques using different approaches, but mostly based on biochemical analyses, and comparisons of slowly evolving molecules, have drastically changed our concepts on phylogeny and evolution of living organisms. Instead of the previously accepted five kingdoms: plants, animals, fungi, protists and bacteria, biochemical characteristics and analysis of sequences of proteins and the small- and large-subunit ribosomal RNAs suggest a different phylogenetic tree. Accordingly, three superkingdoms or domains are recognized now: Eukaryotes, Eubacteria, and Archaea. In turn, Eukaryotes are divided into five kingdoms: Fungi, Plantae, Animalia, Alveolates (comprising ciliates, dinoflagellates and apicomplexans), and Stramenopiles (where diatoms, oomycetes, labyrinthulids, brown algae and chrysophytes are included) (see [22,23]). These are the so-called ‘crown kingdoms’. According to these concepts, oomycetes are no longer considered to belong to the fungal group. Regarding hyphochytriomycetes, it appears that they form a monophyletic group with oomycetes [23], and thus belong to the kingdom Stramenopiles. It is important to point out that chitin seems to be restricted to eukaryotes, and is present in members of all the ‘crown kingdoms’, with the exception of plants. Since it appears that separation of the eukaryotic groups occurred approximately 1 billion years ago, it may be concluded that Chs have their origin as a branch of antique β-glycosyl-transferases, once the plant kingdom had diverged about this time. This hypothesis is supported by the distribution of skeletal cell wall polysaccharides: cellulose in plants, and chitin in fungi. Further evidence is provided by results obtained by sequence analyses of srRNA [24] and actin [25] that suggest that animals and true fungi are closer to each other than to any other eukaryote kingdom, and share a more recent common ancestor than either does with the plant lineage. Phylogenetic analysis of a large number of enzyme sequences also indicates that animals and fungi evolved as separate groups after the separation of the plant kingdom, except that fungi have changed faster than the animal and plant lines [26]. The above hypothesis, and the cited suggestion that fungal and animal lineages have changed at a different pace, may explain why animals, in contrast to fungi, have a single family of highly similar Chs (see Fig. 1), and why the similarities in their Chs are higher than those from fungal groups. A different problem refers to Chs from Alveolates and Stramenopiles. In regard to oomycetes, as observed with animals, a single family of Chs with similarities to fungal division 1 Chs appears to exist (Fig. 1). We must recall that chitin is a non-essential and minor component in the cell wall of oomycetes and is accompanied by cellulose in both oomycetes and hyphochytriomycetes [27]. It is possible that Stramenopiles also separated from fungi before Chs diversification occurred, although the Chs phylogenetic tree, and the similarities of oomycetes Chs with division I enzymes, in contrast to division II and animal Chs, do not agree with this possibility. Alternative possibilities are a parallel evolution of oomycetes and division 1 enzymes, or late acquisition of a division 1-like enzyme by oomycetes by horizontal gene transfer. Examples of bona fide horizontal transfer of genetic information in fungi are not common, but they have been solidly substantiated [28]. It is evident that it will be necessary to await further studies of CHS genes from Alveolates and Stramenopiles in order to clarify this matter.

In regard to true fungi, the fact that members of both Chs divisions and five classes are present in all terrestrial fungi, suggests that diversification of the enzymes preceded that of fungal taxa. Analysis of 18S ribosomal RNA gene sequences, 5S sequences, and biochemical data support the idea of a former division of fungi into the Ascomycota–Basidiomycota and Zygomycota–Chytridiomycota lineages [29,30]. The origin of the ascomycete/basidiomycete clade has been set at about 600 Ma by use of the small subunit ribosomal sequences, assisted by a calibration with fossil records [31]. The relative dates of separation of the crown kingdoms and of the fungal groups sets the time at which the separation of Chs divisions and classes occurred.

It may be recalled that no Chs belonging to class III has been identified in zygomycetes. Whether this is due to deficient analysis, or is a true difference with ascomycetes and basidiomycetes will require analysis of a larger number of species and enzymes. In this regard it is unfortunate that no CHS gene from chitridiomycetes has been isolated yet. Homology analyses of the 10 Chs from P. blakesleeanus favor the second alternative. But then, how can this discrepancy be explained if we observe that appearance of Chs III class was preceded by separation of Chs divisions? The most appealing hypothesis is that the precursor of this group of enzymes was lost in the zygomycete ancestor during evolution. This hypothesis may be applied also to the observation that many fungal species do not contain representatives of all Chs divisions and classes. Nevertheless it is suggestive that class III Chs have the highest level of similarity among them, in comparison with enzymes from classes I and II, suggesting a more recent origin. This hypothesis agrees with the observation on the relative number of specific motifs present in each Chs class discussed above (see Table 1). The presence of more than one member of a specific Chs class in a single fungal species probably represents late gene duplication events. In the light of these results it is evident that analysis of Chs evolution may be extremely valuable by itself, and may contribute to our understanding of obscure problems regarding the phylogenetic relationships and evolution of fungi.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References

Original work of the authors was supported by CONACYT, México. J.R.H. and R.R.M. are National Investigators, México; J.M.G.P. is predoctoral fellow from CONACYT.

References

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Classification of Chs
  5. 3Conserved motifs and amino acids in Chs
  6. 4Conserved structure and the active site of Chs
  7. 5Evolution of Chs
  8. Acknowledgements
  9. References
  • [1]
    Ruiz-Herrera, J. (1992) Fungal Cell Wall: Structure, Synthesis, and Assembly, 248 pp. CRC Press. Boca Raton, FL.
  • [2]
    Ruiz-Herrera, J., Sentandreu-Ramon, R. and Martinez, J.P. (1992) Chitin biosynthesis in fungi. In: Handbook of Applied Mycology, Vol. 4. Fungal Biotechnology (Arora, D.K., Elander, R.P. and Mukerji, K.G., Eds), pp. 281–312. Marcel Dekker, New York.
  • [3]
    Shaw, J.A., Mol, P.C., Bowers, B., Silverman, S.J., Valdivieso, M.H., Duran, A., Cabib, E. (1991) The function of chitin synthases 2 and 3 in the Saccharomyces cerevisiae cell cycle. J. Cell Biol. 114, 111123.
  • [4]
    Beth-Din, A., Yarden, O. (2000) The Neurospora crassa chs3 gene encodes an essential class I chitin synthase. Mycologia 92, 6573.
  • [5]
    Bulawa, C.E. (1993) Genetics and molecular biology of chitin synthesis in fungi. Annu. Rev. Microbiol. 47, 503534.
  • [6]
    Ruiz-Herrera, J., Gonzalez-Prieto, J.M. and Ruiz-Medrano, R. (2001) Significance of the multigenic control of chitin synthase in fungi. In: Chitin Enzymology 2001 (Muzzarelli, R.A., Ed.), pp. 451–462. Atec Edizioni, Italy.
  • [7]
    Bowen, A.R., Chen-Wu, J.L., Momany, M., Young, R., Szaniszlo, P.J., Robbins, P.W. (1992) Classification of fungal chitin synthases. Proc. Natl. Acad. Sci. USA 89, 519523.
  • [8]
    Specht, C.A., Liu, Y., Robbins, P.W., Bulawa, C.E., Iartchouk, N., Winter, K.R., Riggle, P.J., Rhodes, J.C., Dodge, C.L., Culp, D.W., Borgia, P.T. (1996) The chsD and chsE genes of Aspergillus nidulans and their roles in chitin synthesis. Fungal Genet. Biol. 20, 153167.
  • [9]
    Xoconostle-Cazares, B., Specht, C.A., Robbins, P.W., Liu, Y., Leon, C., Ruiz-Herrera, J. (1997) Umchs5, a gene coding for a class IV chitin synthase in Ustilago maydis. Fungal Genet. Biol. 22, 199208.
  • [10]
    Bulawa, C.E., Slater, M., Cabib, E., Au-Young, J., Sburlati, A. (1986) The S. cerevisiae structural gene for chitin synthase is not required for chitin synthesis in vivo. Cell 46, 213225.
  • [11]
    Miyazaki, A., Ootaki, T. (1997) Multiple genes for chitin synthase in the zygomycete fungus Phycomyces blakesleeanus. J. Gen. Appl. Microbiol. 43, 333340.
  • [12]
    Ruiz-Herrera, J., Xoconostle-Cazares, B. (1995) Molecular and genetic control of chitin biosynthesis in fungi. Arch. Med. Res. 26, 315321.
  • [13]
    Saxena, I.M., M.R. Brown, Jr., Fevre, M., Geremia, R.A., Henrissat, B. (1995) Multidomain architecture of b-glycosyl transferases: implications for mechanisms of action. J. Bacteriol. 177, 14191424.
  • [14]
    Campbell, J.A., Davies, G.J., Bulone, V., Henrissat, B. (1997) A classification of nucleotide-diphospho-sugar glycosyl-transferases based on amino acid sequence similarities. Biochem. J. 326, 929942.
  • [15]
    Charnock, S.J., Davies, G.J. (1999) Structure of the nucleotide-diphospho-sugar transferase, SpsA from Bacillus subtilis, in native and nucleotide-complexed forms. Biochemistry 38, 63806385.
  • [16]
    Horsch, M. and Sowdhamini, R. (1996) A fold prediction for the catalytic domain of chitin synthases. In: Chitin Enzymology, Vol. 2 (Muzzarelli, R.A.A., Ed.), pp. 447–458. Atec Edizioni, Italy.
  • [17]
    Merz, R.A., Horsch, M., Nyhlén, L.E. and Rast, D. (1999) Biochemistry of chitin synthase. In: Chitin and Chitinases (Jollès, P. and Muzzarelli, R.A.A., Eds.), pp. 9–37. Birkhauser Verlag, Basel.
  • [18]
    Nagahashi, S., Sudoh, M., Ono, N., Sawada, R., Yamaguchi, E., Uchida, Y., Mio, T., Takagi, M., Arisawa, M., Yamada-Okabe, H. (1995) Characterization of chitin synthase 2 of Saccharomyces cerevisiae: implication of two highly conserved domains as possible catalytic sites. J. Biol. Chem. 270, 1396113967.
  • [19]
    Yabe, T., Yamada-Okabe, T., Nakajima, T., Sudoh, M., Arisawa, M., Yamada-Okabe, H. (1998) Mutational analysis of chitin synthase 2 of Saccharomyces cerevisiae: identification of additional amino acid residues involved in its catalytic activity. Eur. J. Biochem. 258, 941947.
  • [20]
    Cos, T., Ford, R.A., Trilla, J.A., Duran, A., Cabib, E., Roncero, C. (1998) Molecular analysis of Chs3p participation in chitin synthase III activity. Eur. J. Biochem. 256, 419426.
  • [21]
    Jones, T. (1999) Protein secondary prediction based on position-specific scoring matrices. J. Mol. Biol. 292, 195202.
  • [22]
    Sogin, M.L., Morrison, H.G., Hinkle, G., Silberman, J.D. (1996) Ancestral relationships of the major eukaryotic lineages. Microbiologia 12, 1728.
  • [23]
    Van de Peer, Y., De Wachter, R. (1997) Evolutionary relationships among the eukaryotic crown taxa taking into account site-to-site rate variation in 18S rRNA. J. Mol. Evol. 45, 619630.
  • [24]
    Kumar, S., Rzhetsky, A. (1996) Evolutionary relationships of eukaryotic kingdoms. J. Mol. Evol. 42, 183193.
  • [25]
    Drouin, G., Moniz de Sa, M., Zuker, M. (1995) The Giardia lamblia actin gene and the phylogeny of eukaryotes. J. Mol. Evol. 41, 841849.
  • [26]
    Doolittle, R.F., Feng, D.F., Tsang, S., Cho, G., Little, E. (1996) Determining divergence times of the major kingdoms of living organisms with a protein clock. Science 271, 470477.
  • [27]
    Bartnicki-Garcia, S. (1979) Fungal cell wall composition. In: Handbook of Microbiology, Vol. 4, 2nd Edn., pp 229–252. CRC Press, Boca Raton, FL.
  • [28]
    Rosewich, L.U., Kistlwer, H.C. (2000) Role of horizontal gene transfer in the evolution of fungi. Annu. Rev. Phytopathol. 38, 325363.
  • [29]
    Berbee, M.L., Taylor, J.W. (1992) Detecting morphological convergence in true fungi, using 18S rRNA gene sequence data. Biosystems 28, 117125.
  • [30]
    Walker, W.F. (1985) 5 S and 5.8 S ribosomal RNA sequences and protist phylogenetics. Biosystems 18, 269278.
  • [31]
    Redecker, D., Kodner, R., Graham, L.E. (2000) Glomelean fungi from the Ordovician. Science 289, 19201921.