Codon bias in the ß-lactam producer Acremonium chrysogenum Kerstin Jekosch and Ulrich Kück, Lehrstuhl für Allgemeine und Molekulare Botanik, D-44780 Bochum, Germany

Fungal Genetics Newsletter 46:11-13

*Table 2 in the print version of this article was mis-formatted and the correct version is available here*

In this paper we compile the G/C-content, the codon bias, and consensus sequences for the translation initiation and the intron splicing sites from 19 nuclear genes of the major ß-lactam antibiotic producer Acremonium chrysogenum. Our data are compared with those from other filamentous fungi, such as Aspergillus nidulans, Neurospora crassa, and Sordaria macrospora.

The filamentous fungus Acremonium chrysogenum is the industrially most important producer of the ß-lactam antibiotic cephalosporin C (reviewed by Brakhage 1998, Microbiol. Mol. Biol. Rev. 62:547-585). Sequence characteristics of this imperfect fungus are of great interest as prerequisite for the improvement of antibiotic biosynthesis. We present a comprehensive sequence analysis of all nuclear genes, published so far. We provide a compilation of codon usage, consensus sequences for translation initiation sites, and regulatory sequences relevant for intron splicing.

Genes used for the present analysis are 1) cystathionine beta-synthase (Acc # E08842), 2) homoserine 0-acetyltransferase (Acc # E08840), 3) cystathionine gamma-lyase (Acc # E08276), 4) CPC biosynthesis related gamma gene (Acc # E06692), 5) CPC biosynthesis related alpha gene (Acc # E06691), 6) alkaline protease, alp (Acc # D00923), 7) glyceraldehyde-3-phosphate dehydrogenase (Acc # E03375), 8) actin (Acc # E03374), 9) phosphoglycerate kinase E (Acc # 03373), 10) beta isopropylmaleate dehydrogenase, leu2 (Acc # E01906), 11) orotidine 5´-phosphate decarboxylase, pyr4 (Acc # X15937), 12) -(L--aminoadipyl)-L-cysteinyl-D-valine synthetase (Acc # M33522), 13) isopenicillin N synthase (Acc # S39881), 14) desacetoxycephalosporin C synthetase/hydroxylase, pcbAB (Gutiérrez et al. 1991, J. Bacteriol. 173:2354-2365), 15) desacetylcephalosporin C acetyltransferase, cefG (Acc # M91649), 16) esterase C, estC (Japan patent: Matsuda et al. 1992, Hei 4-144688), 17) beta tubuline, ßtub (Acc # X72789), 18) transcriptional repressor CREA, creA (unpublished), 19) transcription factor CPCR1, cpcR1 (Acc # AJ132014). No. 12 to 16 and 18 to 19 are involved in ß-lactam biosynthesis. The modes of action of genes 4 and 5 are not yet known.

The average G/C content is 61.5 ± 2.3%, which is higher than in A. nidulans (about 50 %, Lloyd and Sharp 1991, Mol. Gen. Genet. 230:288-294), N. crassa (54.1%, Edelmann and Staben 1994, Exp. Mycol. 18:70-81), and S. macrospora (56.7 %, Pöggeler 1997, Fungal Genet. Newsl. 44:41-44).

Similar to other fungi we found a bias for codons with a C in the third position (Table 1). While the termination codon TAA is preferred in N. crassa and S. macrospora, there is no such preference in A. nidulans and only little bias towards this codon in A. chrysogenum. The six least used codons in A. chrysogenum are TTA (Leu), TGT (Cys), AGA (Arg), ATA (Ile), GTA (Val) and AAA (Lys). Compared with the other three filamentous fungi, the same bias for rarely used and preferred codons is found in A. chrysogenum. An exception is found when the codons for serine are considered. Obviously, there is no significant preference for any one of the six possible tripletts.

The cephalosporin C-biosynthetic genes show different preferences for individual codons which correlates with their rate of expression as concluded from our own data (unpublished). The highly expressed pcbC and cefEF genes show only 15 and 17% low usage codons, while genes with a low transcriptional level, such as pcbAB, cefG and estC, contain 27 to 53% low usage codons. This correlation of the occurrence of rare codons and expression has already been reported for other organisms, e.g. Saccharomyces cerevisiae, Escherichia coli, Drosophila melanogaster (Zhang et al. 1991, Gene 105:61-67), S. macrospora (Pöggeler 1997) and A. nidulans (Lloyd and Sharp 1991, Mol. Gen. Genet. 230:288-294) and was supposed to affect translation rates.

The consensus sequence of the translation initiation context in A. chrysogenum does not significantly differ from S. macrospora and N. crassa. It should be noted that the open reading frame of the pcbAB gene starts with an GTG. The translation startpoint of the estC gene has not been identified yet. Following the ATG start codon, the three fungi show a preference for GCN, coding for alanine.

The average intron length in A. chrysogenum is 95 nt and this agrees well with the average intron length in S. macrospora (88 nt) but contrasts with that in N. crassa (63 nt). The average distance between the branch site and the 3´-splice site is 14 nt, and it is between 8 and 30 nt in length. In S. macrospora and N. crassa this distance varies from 12 to 22 nt and 14 to 30 nt, respectively. Similar the intron donor and acceptor consensus sequences show no significant deviation from those of the other fungi (Table 2). One exception is the cefG gene, the only ß-lactam biosynthesis gene containing introns with striking differences in the intron regulatory sequences (5´ intron donor G^TAGGTA and C^GGTGAG, 3´ intron acceptor AAAGT^, TGCTA^).

Acknowledgments: This work was supported by a grant of Hoechst Marion Roussel (Frankfurt a.M., Germany).

Table 1: Analysis of codon usage on the basis of 19 nuclear gene sequences. The codon usage values are given as whole numbers (N), as percent usage determined for each amino acid (%), and as relative synonymous codon usage (RSCU).
Codon AS N % RSCU Codon AS N % RSCU
GCT Ala 154 16 1.48 CTC Leu 394 41 3.78
GCC 466 52 4.47 CTA 58 6 0.56
GCA 101 11 0.97 CTG 313 32 3.00
GCG 178 20 1.71 AAA Lys 48 12 0.46
CGT Arg 71 12 0.68 AAG 362 88 3.47
CGC 248 41 2.38 ATG Met 187 100 1.79
CGA 58 10 0.56 TTT Phe 81 22 0.78
CGG 122 20 1.17 TTC 280 78 2.69
AGA 41 7 0.39 CCT Pro 103 17 0.99
AGG 109 18 1.05 CCC 252 41 2.42
AAT Asn 84 22 0.81 CCA 74 12 0.71
AAC 301 78 2.89 CCG 186 30 1.78
GAT Asp 163 28 1.56 TCT Ser 82 10 0.79
GAC 428 72 4.11 TCC 209 25 2.01
TGT Cys 21 18 0.20 TCA 86 10 0.83
TGC 99 82 0.95 TCG 189 22 1.81
CAA Gln 69 18 0.66 AGT 75 9 0.72
CAG 321 82 3.08 AGC 210 25 2.02
GAA Glu 99 17 0.95 ACT Thr 80 12 0.77
GAG 493 83 4.73 ACC 283 44 2.72
GGT Gly 163 21 1.56 ACA 94 15 0.90
GGC 414 53 3.97 ACG 190 29 1.82
GGA 94 12 0.90 TGG Trp 124 100 1.19
GGG 113 14 1.08 TAT Tyr 64 19 0.61
CAT His 89 28 0.85 TAC 259 81 2.49
CAC 234 72 2.25 GTT Val 101 14 0.97
ATT Ile 95 19 0.91 GTC 380 52 3.65
ATC 363 72 3.48 GTA 48 7 0.46
ATA 46 9 0.44 GTG 202 28 1.94
TTA Leu 9 1 0.09 TAA Term. 8 42 2.54
TTG 82 8 0.79 TAG 5 26 1.57
CTT 114 12 1.09 TGA 6 32 1.94

Table 2: A. chrysogenum (Ac) consensus sequences compared with those of Sordaria macrospora (Sm) and Neurospora crassa (Nc). Abbreviations: org. = organism, subscribed numbers indicate the percentage of occurence. 


org.     translation initiation context

Ac
    C44               A39  A39
C67     N  N  C61  A61          A94 T100  G95  G53  C53
    A33               C39  C39

Sm
        A38             C50                     G50
C75  C50     N  C63  A88       A63  A100  T100  G100       C50
        G38             A38                     T35

Nc
                  A
C  N  N  N  C  A     A  A  T  G  G  C
                  C

     consensus 5´intron donor

Ac     G41  ^  G95   T91   A64  A68  G91  T64
Sm     G67  ^  G100  T100  A72  A61  G83  T72
Nc     G    ^  G     T     A    A    G    T

     consensus branch site
Ac
                   A50
      G64  C100  T100        A95  C64  C41
                   G45

Sm
     A56           A56
         C100  T100       A94  C78  N
     G44           G33

Nc
     A        A     C
        C  T     A     C
     G        G     A



     consensus 3´intron acceptor

Ac
     A32  G36  C55 
                 A91 G91
     T32  A41  T27

Sm
     G33  A39  C56
                  A100  G100
     A27  T39  T44

Nc
     A  A  T
              A G
     T  T  C

Return to the FGN 46 Table of Contents

Return to the FGSC main page