Molecular Biology of Aspergillus niger: Results of Phase 1

SUMMARY

     The following data were produced during Phase 1 (1994-1996) of
our research and result from analyses of our A. niger cDNA library
as well as sequence information obtained from Genbank. After a
brief hiatus, the project has been revived.  
     Initial research focussed on the creation of a cDNA library
from A. niger and the isolation and characterization of cDNA clones
from this library which can be used to develop the molecular
biology of this species. Sixty-eight clones were partially
sequenced to assess the general quality of the library.
Introduction of library DNA into E. coli and S. cerevisiae
permitted isolation of functional cDNA clones which rescued
mutations in the his, trp and leu genes. Two of these genes, which
encode isoforms of á-isopropylmalate dehydrogenase, were
characterized in greater detail. This included complete sequence
analysis, developmental expression and genomic organization.
Analysis of codon usage for these and all
known A. niger genes was also performed. We have also attempted to disrupt
several A. niger genes.  

1. cDNA LIBRARY

a) Quality of library
     Our library consists of 2E6 independent transformants.
Restriction analysis of 121 clones indicated that 17 contained no
detectable insert. However, when 4 of these were subjected to
sequence analysis, 2 were found to contain inserts > 200 nt. Thus,
we estimate that > 90% of our transformants contain inserts with an
average size > 0.9 kbp. A poly(A) tail was
found at the 3' end of all 28 insert-containing clones examined indicating
that our directional subcloning was efficient with > 95% being in the
desired orientation.  

b) Sequence Analyses of Random Clones 
      I have acquired ca. 28 kb of sequence data representing 70
cDNA clones. Sixty-eight of these have been deposited with GenBank
in the database dbEST (accession numbers T82800-T82716).  Thirteen  of 
the 68 cDNA clones examined were found to share significant sequence
similarity to previously reported genes. These sequences include 5
ribosomal proteins, 2 histones, a putative protein kinase, xanthine
dehydrogenase, protein disulfide isomerase, and spermidine synthase. A
fourteenth clone exhibited significant similarity to a yeast expressed
sequence tag (EST) of unknown function. These genes share 33% - 92% amino
acid identity with their putative homolog. Similarities were obtained from
both lower eukaryotes like yeast and Aspergillus nidulans, as well as
higher organisms like Drosophila and mouse. None of the remaining A. niger
clones show significant similarity to previously reported sequences. Of
the 68 cDNAs sequenced only one was duplicated over the regions sequenced,
suggesting it is expressed at a relatively high level.

     The 3' ends of 28 of these clones were also examined and a
consensus sequence was determined. The context about the initiation
codon of 28 previously reported A. niger genes as well as for 7 of
our new genes was examined and a
consensus sequence determined.
 
     Considering the small number of A. niger genes reported to
date it seems likely that essentially all of the genes discovered
by the sequencing of random cDNAs clones will be new genes.
Therefore the assignment of gene function will depend primarily on
the availability of sequence data and functional information from
other organisms. 

c) Codon Usage in A. niger
     In many species highly expressed genes use almost exclusively
a preferred set of codons (Grantham et al, Bennetzen and Hall,
1982; Sharp and Shields) whereas weakly expressed genes do not.
Accordingly, the level of expression for a particular gene can
usually be predicted by the proportion of preferred codons, or the
codon usage bias (CBI), that it displays. Hence the CBI for an A.
niger gene could be used to predict its level of expression. Table II 
shows the CBI for 15 previously-reported A. niger genes and 12 of the
newly cloned and identified genes reported here. Initially, the CBI was
calculated using the codon bias of glucoamylase as the standard. This gene
was chosen because A. niger can express as much as 20 grams per liter
of glucoamylase. Using this CBI standard ribosomal protein genes
ranged between 0.26 and 0.85 and histones between 0.47 and 0.68.
The highest CBI (0.93)  belonged to ATP synthase subunit 9 and the
lowest CBIs belonged to genes encoding xanthine dehydrogenase
(0.11) and a protein kinase (0.23). Although the actual CBI values
change when we use other genes that are likely highly expressed
like acid proteinase or pectinase as the standard, the relative
values of the CBIs remain unchanged. This is due, in part, to a
pronounced pattern of codon usage in A. niger genes. An examination
of the sequences shown in Table II revealed that the frequency of
codons for each amino acid is dependent on the identity of the
nucleic acid at the third position with C > G = T > A. (Our study
found a ratio of 4073 : 2000 : 1881 : 696.) 

II. ISOLATION OF SPECIFIC GENES BY COMPLEMENTATION

1. Isolation of two genes for á-isopropylmalate dehydrogenase

a) Complementation of E. coli and S. cerevisiae mutants
     Five æg of cDNA library was transformed into Saccharomyces
cerevisiae strain BWGI-7a which is deficient in LEU2. Clones able
to grow in the absence of leucine arose at a frequency of 4E-4.
Plasmid DNA extracted from 4 of these transformants were amplified
in E. coli strain DH5àF' then introduced into E. coli strain
MC1066. All 4 plasmids complemented the leu2B- mutation in E. coli.
These experiments demonstrated that A. niger amino acid
biosynthetic genes function efficiently in E. coli and S.
cerevisiae and, by extension, these well-developed systems can be
used to screen our cDNA library for useful genes.  

b) Sequence analysis
      Restriction digest analysis suggested, and sequence data
confirmed, that the 4 isolated plasmids represented 2 cloned genes,
designated leu2A and leu2B. Sequence analysis revealed that leu2A
and leu2B contain open reading frames of 1092 nt and 1116 nt,
respectively, which share 55.5% identity at the nucleotide level
and 52.5% at the amino acid level.

     The predicted amino acid sequences of leu2A and leu2B were
found to share significant similarity with á-isopropylmalate
dehydrogenase from a wide range of eukaryotic and prokaryotic
species.

c) Codon Usage
     We also compared the codon biases of these ORFs with other A.
niger genes. Most A. niger genes exhibit a marked propensity
towards using codons having a C in the third position and avoid A
at this position. The high codon bias index (CBI) of leu2A (0.82)
indicates it shares the same codon bias of most A. niger genes
examined. Leu2B has the lowest CBI of all (-0.03), which indicates
there is little bias in its codon usage.
 
d) Southern Blot Analysis
     DNA blot analysis indicated there is probably only one copy of
each gene. 
 
e) Northern Blot Analysis
     RNA blot analysis revealed a pronounced difference in the
temporal accumulation of leu2A and leu2B transcripts. Leu2a RNA
accumulated to high levels during germination whereas leu2b RNA
reached maximum levels much later. Furthermore, leu2a but not leu2b
RNA accumulated to higher levels in minimal medium than in complete
medium.

f) Discussion
     The observed differences in the time and regulation of their
expression indicates these genes act in a complementary fashion.
Leu2A is expressed during early vegetative growth and is regulated
in part by leucine. During later growth, when the steady-state
concentration of leu2A transcripts is decreased, leu2B transcript
appears. We suspect the role of leu2B is to provide a basal level
of IPM dehydrogenase during conidiation when leu2A has become
inactive.Since leu2B exhibits little codon bias, its translation
draws less heavily on specific pools of tRNAs than does leu2A. This
may be an important consideration when cells are no longer
receiving nutrition from the environment, yet are being called
upon to produce conidiophores and conidia. Transfer-RNA pools may
change during this phase of growth. It is also possible that the
codon usage of leu2B dictates the level at which the protein is
synthesized, as has been found in yeast and E. coli (Bennetzen and
Hall, 1982). 

     IPM dehydrogenase catalyzes the third committed step in
leucine biosynthesis. This enzyme is thus in a key position for the
potential regulation of intracellular leucine concentration. At the
present time there are few selectable markers suitable for A.niger.
Amino acid biosynthetic genes such as leu2A will prove useful in
the engineering of industrially-important strains of A. niger.

     In E. coli and yeast, more highly expressed genes exhibit a
greater codon bias than less expressed genes. In these systems the
preferred codons are those corresponding to the most abundant tRNAs
(Bennetzen and Hall, 1982). The pronounced codon bias of highly
expressed genes in A. niger suggests that the amount of protein
produced by transgenic A. niger may depend on the codon bias of
introduced genes. High expression of bacterial ë-endotoxin in
plants (Adang et al. 1993) and whale myoglobin in E. coli (Springer
and Sligar, 1987) were obtained by changing the host sequences to
conform with the host codon bias. Thus, it may be possible to
manipulate the sequence of genes to maximize, or otherwise
modulate, the levels of the final product. 

2. Isolation of the A. niger genes for his and trp 
     A similar strategy has been used to isolate A. niger sequences
capable of complementing mutations in his and trp. Preliminary
sequence analysis indicated these clones encode the A. niger
homologs of the E. coli genes. Further analyses await funding.

III. GENE DISRUPTIONS

     We have attempted to disrupt, by homologous recombination, two
A. niger genes, one a putative protein kinase and the other, xxxx.
In one case, a 400 bp fragment was removed from the protein coding
region and replaced with the selectable, A. nidulans gene pyrG. In
the other case pyrG was simply inserted into coding region. A.
niger was transformed with these DNAs, either circular or
linearized. PyrG+ transformants were obtained, but PCR analysis
indicated the endogenous genes were not disrupted. 

IV. EXPRESSION CASSETTES

     Plasmids have been constructed with a view to expression
foreign genes in A. niger. Polymerase-chain reaction (PCR) was used
to amplify, from genomic DNA, the promoters of the A. niger genes
for pyrG and pyruvate kinase which are expected to provide for low-
and high- level constitutive expression (respectively). Each
promoter was placed just upstream of the coding region of the the
uida marker gene. The AMA sequence from pAB4-ARp1 was also added to
ensure maintenance of these plasmids in A. niger. Preliminary
results indicated that A. niger transformed with the construct with
the pyruvate kinase promoter (pPKGusAMA) expressed relatively high
levels of þ-glucuronidase (GUS). As expected, GUS was detected in
the cells but not in the medium. 

VII. ISOLATION OF NICB

VIII. CODON USAGE

      A computer program was written to extract codon usage data
from nucleic acid sequences, store and retrieve these data from a
data base, and to handle the computations required to analyze codon
usage.

Appendices: