Analysis of Expressed Sequence Tags From Two Starvation, Time-of-Day-Specific Libraries of Neurospora crassa Reveals Novel Clock-Controlled Genes

Corresponding author: Jay C. Dunlap, Department of Genetics, Dartmouth Medical School, Hanover, NH 03755., jay.c.dunlap{at}dartmouth.edu (E-mail)

Communicating editor: J. ARNOLD

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In an effort to determine genes that are expressed in mycelial cultures of Neurospora crassa over the course of the circadian day, we have sequenced 13,000 cDNA clones from two time-of-day-specific libraries (morning and evening library) generating 20,000 sequences. Contig analysis allowed the identification of 445 unique expressed sequence tags (ESTs) and 986 ESTs present in multiple cDNA clones. For 50% of the sequences (710 of 1431), significant matches to sequences in the National Center for Biotechnology Information database (of known or unknown function) were detected. About 50% of the ESTs (721 of 1431) showed no similarity to previously identified genes. We hybridized Northern blots with probes derived from 26 clones chosen from contigs identified by multiple cDNA clones and EST sequences. Using these sequences, the representation of genes among the morning and evening sequences, respectively, in most cases does not reflect their expression patterns over the course of the day. Nevertheless, we were able to identify four new clock-controlled genes. On the basis of these data we predict that a significant proportion of the expressed Neurospora genes may be regulated by the circadian clock. The mRNA levels of all four genes peak in the subjective morning as is the case with previously identified ccgs.

THE ascomycete Neurospora crassa has a long history as a model organism for both classical and molecular genetics in general (DAVIS 2000 ) as well as for investigation of circadian rhythmicity (DUNLAP 1999 ). Several efforts have been made recently to map and sequence the Neurospora genome (Fungal Genome Resources at http://gene.genetics.uga.edu/ and Munich Information Centre for Protein Sequences Neurospora crassa Database at http://www.mips.biochem.mpg.de/proj/neurospora/) as well as to address the expression of genes at different stages of the Neurospora developmental cycle (Neurospora Genome Project at http://www.unm.edu/~ngp/; NELSON et al. 1997 ; DOLAN et al. 2000 ).

One of the most notable aspects of N. crassa is the fact that a great part of its life is tightly controlled by the circadian clock. Not only the formation of macroconidia but also the expression of genes important for general metabolism occur in a circadian fashion (LOROS 1998 ). Using differential screening procedures of cDNA libraries, several genes have been shown to be under clock control (LOROS et al. 1989 ; BELL-PEDERSEN et al. 1996 ). These studies have confirmed the impression that a substantial proportion of Neurospora genes might be regulated by the circadian clock, but the extent of this control is yet unclear. Therefore, it was decided to use an expressed sequence tag (EST) sequencing approach to address the question of clock regulation at a larger scale as well as to identify novel ESTs and to examine global gene expression under conditions (starvation in the dark) not previously examined.

EST sequences are highly valuable in genomic approaches. EST data can be used to identify novel genes, search for homologous genes in different organisms, analyze alternative splicing, and identify chromosomal locations of genes and other applications (PANDEY and LEWITTER 1999 ; OHLROGGE and BENNING 2000 ). With large numbers of ESTs available, so-called "digital Northerns" or "electronic transcriptional profiling" can be performed by counting the number of ESTs for a given gene within the EST population (PANDEY and LEWITTER 1999 ; PRADE et al. 2001 ). Also, cDNA clones with known sequences are used to establish cDNA microarrays with which genome-wide expression patterns can be investigated (DUGGAN et al. 1999 ).

In the project described here, cDNA clones from two mycelial libraries of N. crassa were partially sequenced. The libraries were made from time-of-day-specific tissues and had been used to characterize clock-controlled genes (ccgs) earlier (BELL-PEDERSEN et al. 1996 ). The ESTs were organized into contiguous sequences (contigs) and compared to the sequences within the National Center for Biotechnology Information (NCBI) database. Besides genes that have homologs in the NCBI database, the sequencing yielded a large number of yet unknown ESTs (50% of 1431 different genes identified within the libraries). Northern blots were hybridized with probes derived from 26 cDNA clones, and 4 of these genes were found to be rhythmically expressed. Again, this confirms the prediction that a significant proportion of N. crassa genes may be regulated by the circadian clock.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and growth conditions:
The following N. crassa strains were used: frq⁺ strain 87-3 (bd; a) and long period mutant 585-7 (bd; a; frq⁷). Neurospora media (Vogel's) were as described (DAVIS and DESERRES 1970 ). Culture conditions for rhythmic RNA analysis and light induction experiments were performed according to published methods (LOROS et al. 1989 ).

Preparation and analysis of RNA:
RNA was prepared as described previously (YARDEN et al. 1992 ). Northern blots, slot blots, and probing with DNA probes or riboprobes were performed according to standard techniques (MANIATIS et al. 1982 ).

Template preparation for DNA sequencing:
The libraries from which the sequenced clones are derived have been described previously (BELL-PEDERSEN et al. 1996 ); they were not further amplified or modified except for those modifications that normally arise during the process of phagemid excision (see below). Template preparation, sequencing, and analysis were performed at the Advanced Center for Genome Technology, University of Oklahoma. For each library, infection of Escherichia coli strain SOLR with aliquots from the mass library excision converted the single-stranded phagemids to the double-stranded form. The appropriate volume of F₁ lysate containing the single-stranded phagemids was added to the SOLR cells to yield 200–300 colonies per 2 x 10⁷ of infected SOLR cells. Single white colonies were picked to individual wells of a 96-well block containing 1.5 ml TB with ampicillin per well. After growth, triplicate 100-µl aliquots of each cell culture were mixed with glycerol (final concentration 17%) and stored at -70°. Individual clones are available publicly from the Fungal Genetics Stock Center, University of Kansas Medical Center (Kansas City, KS; http://www.fgsc.net). Double-stranded DNA template was robotically isolated from the remaining cell culture using a modified plasmid preparation procedure (http://genome.ou.edu/proto.html; BIRNBOIM and DOLY 1979 ) on a Biomek 2000 (Beckman, Fullerton, CA).

DNA sequencing:
Partial nucleotide sequences of the cDNA inserts were determined by single-pass sequencing using the Big Dye sequencing system (Perkin Elmer, Norwalk, CT; ROSENBLUM et al. 1997 ) with either the universal forward (5' GACGTTGTAAAACGACGGCC) or the universal reverse primer (5' CACAGGAAACAGCTATGACC). Generally, 0.2 µg of each template was cycle sequenced with 6.5 pmol primer and 2 µl of a 1:4 diluted AmpliTaq DNA polymerase Big Dye termination mix in a final volume of 5–7 µl. Cycling conditions were 60 cycles of 95° for 10 sec, 50° for 5 sec, and 60° for 4 min in Perkin Elmer thermocyclers 9600 or 9700. Reactions were purified by centrifugation through Sephadex G-50 microtiter plate spin columns (BODENTEICH et al. 1994 ) and electrophoresed on an ABI 377 sequencer (Applied Biosystems, Foster City, CA) for 7 hr at 2.4 kV. The collected raw data was manually retracked and analyzed using the ABI sequencing analysis software. See http://www.genome.ou.edu/proto.html for more detailed descriptions of protocols.

Informatics:
DNA sequences were automatically surveyed for quality using the phred software (EWING et al. 1998 ). Sequences were screened for the presence of vector sequences, low quality sequences (Phred-16), and small inserts (<100 bp) and were trimmed to remove poly(A) tail and vector sequences. All sequences were screened for homology to ribosomal RNA, mitochondrial DNA, and E. coli genomic sequences using BLASTN (ALTSCHUL et al. 1997 ) and matches were removed. To reduce the redundancy of the libraries and determine the relative abundance of the members, overlapping cDNA sequences were identified and organized into contigs using the phrap software (P. Green, copyright 1994–1996, http://bozeman.mbt.washington.edu/phrap.docs/phrap.html) with a window size of 14 and a minimal score of 80. A contig here, then, arises from the alignment of clustered ESTs on the basis of sequence similarity. The derived peptide sequences corresponding to the cDNA sequences were compared with the protein database available through the NCBI (Bethesda, MD) using the BLASTX algorithm (ALTSCHUL et al. 1997 ). EST sequences and results of comparisons can be obtained through the website http://www.genome.ou.edu/fungal.html. The site also provides the opportunity to search the EST sequences using BLAST. The EST sequences were also deposited in the dbEST database of GenBank (BOGUSKI et al. 1993 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Sequence analyses:
Plasmid DNA from >13,000 clones derived from two N. crassa cDNA libraries was sequenced using the universal reverse or universal forward primer (Fig 1). The libraries have been used previously to identify clock-controlled genes (ccgs) and correspond to "morning" (circadian time CT1) and "evening" (CT13; BELL-PEDERSEN et al. 1996 ). For >53% of all clones, high-quality sequence data (see MATERIALS AND METHODS) were obtained with both primers, 28% of all clones were sequenced only with the forward primer, and 19% only with the reverse primer. Typically 300–500 bp of high-quality sequence was obtained. Sequences obtained with the reverse primer should correspond to the 3' end and are designated in the database with an "r1" at the end of the sequence identification number; sequences obtained with the forward primer should correspond to the 5' end and have "f1" at the end of their identification number. Exceptions occurred for those few clones in which the cDNA inserts were inserted in reverse orientation. The sequences are available in a publicly searchable format as described in MATERIALS AND METHODS.

View larger version (8K): In this window In a new window Download PPT slide   Figure 1. Derivation of ESTs. Sequences generated using the forward and reverse primers correspond to the 5' and 3' ends of the cDNA clones and are denoted with f1 and r1, respectively, at the end of their sequence identification number. cDNAs were directionally cloned using asymmetric primers and linkers as described in BELL-PEDERSEN et al. 1996 .

Contig analysis and cDNA identities:
We obtained 9148 sequences from the evening and 10,871 sequences from the morning library, representing a total of 1431 different genes. Of these genes, 445 are represented only once within the libraries, whereas 986 are represented by more than one EST. The EST sequences were organized into contigs using the phrap software (P. Green, copyright 1994–1996, http://bozeman.mbt.washington.edu/phrap.docs/phrap.html). To identify potential homologs to the Neurospora genes, the BLASTX algorithm (ALTSCHUL et al. 1997 ) was used to translate each contig nucleotide sequence into the six possible reading frames and to compare the predicted protein sequences with the NCBI protein sequence database. In previously published sequence analyses of Neurospora ESTs (NELSON et al. 1997 ), matches (putative homologs) were divided into highly significant (P/E values 10^-20), moderately significant (10^-5–10^-19), and weakly significant (10^-2–10^-4). Here, we present only those matches that are at least moderately significant. The data upon which this analysis is based can be viewed at the Genetics website at http://www.genetics.org/supplemental/, http://www.genome.ou.edu.fungal.html, and http://www.dartmouth.edu/~jdunlap/. The clones are shown according to the classification scheme utilized by the Expressed Gene Anatomy Database (EGAD, available at http://www.tigr.org/docs/tigr-scripts/egad_scripts/role_report.spl; WHITE and KERLAVAGE 1996 ). A total of 710 of 1431 sequences (50%) correspond to characterized genes, including genes of known function as well as putative open reading frames. The remaining 721 genes (50%) do not have any significant matches within the NCBI database. Of the total set of 1431 genes, 35% were represented in the New Mexico Neurospora EST database (NELSON et al. 1997 ), and of these about half had sequence homologs in the NCBI database and half did not.

Transcript abundance in the morning and evening libraries:
With a large number of ESTs available, so-called "digital Northerns" or "electronic transcriptional profiling" can be performed by counting the number of ESTs for a given gene within an EST population. Formulas have been developed to estimate the statistical significance of this electronic transcript profiling (AUDIC and CLAVERIE 1997 ), but few efforts have been made to evaluate the correlation between digital Northerns and laboratory methods for the analysis of gene expression. Several reports describe approaches to correlate these data for human or plant EST databases (HUANG et al. 1999 ; BORTOLUZZI et al. 2000 ; HWANG et al. 2000 ; MEKHEDOV et al. 2000 ), but no data have been published previously to achieve this for fungal gene expression profiling (see, however, PRADE et al. 2001 ). Therefore, efforts were made to evaluate whether the relative abundance of cDNAs in the morning vs. the evening library reflects the situation in vivo. Insert DNAs of 26 cDNA clones representing contigs that contain multiple cDNA clones were used to hybridize Northern or slot blots with RNA from time courses. Growth and harvest of tissues was performed as described previously (BELL-PEDERSEN et al. 1996 ); the clones used as probes are given in Table 1. Among the 26 genes represented by the cDNAs, 19 should be differentially expressed with P > 0.999 according to published formulas (http://igs-server.cnrs-mre.fr; AUDIC and CLAVERIE 1997 ). Two genes have probabilities of being differentially expressed between 0.95 and 0.999, and 5 have probabilities of <0.95; cDNAs representing these latter ones were chosen as "negative controls" (Table 1). These results indicate that preferential occurrence in either the morning or evening library did not generally correspond to rhythmic expression in vivo or to enhanced expression at the time of day indicated by the occurrence of transcripts in the libraries (Table 1). The majority of the genes investigated showed more or less even expression over the course of the circadian day or displayed small peaks at random times (data not shown). Also, previously characterized ccgs (BELL-PEDERSEN et al. 1996 ) were in most but not all cases represented in the libraries according to their expression patterns in vivo (Table 2). As has been shown previously, ccgs peak preferentially in the late night or early morning (BELL-PEDERSEN et al. 1996 ), but this pattern could be found within the EST sequences only in five out of seven ccgs (Table 2). Possible explanations for these findings are discussed below (see DISCUSSION).

Identification of clock-controlled genes:
Of the 26 cDNA clones used to probe Northern or slot blots, 4 displayed rhythmic expression patterns (Table 3, Fig 2). These were named ccg-13, ccg-14, ccg-15, and lyz and corresponded to contigs 1432, 1421, 1411, and 1442 (Table 3). Probes derived from representative cDNA clones were hybridized to RNA from both frq⁺ and frq⁷ cultures to verify that the period length of the rhythm is dependent on the frq allele (21.5 hr for frq⁺ and 29 hr for frq⁷) and, therefore, that the expression is clearly under clock control. As shown in Fig 2, expression of ccg-13 and ccg-14 was cyclic, displaying high amplitudes, and the troughs of their rhythms were very consistent (Fig 2B). The other two genes, ccg-15 and lyz, displayed much lower amplitudes, and the variation between different assays appears to be much greater (Fig 2B). Nevertheless, a consistent pattern of peaks and troughs could be observed in both frq⁺ and frq⁷ strains for ccg-15 as well as lyz, and although clock control is not as strong as with other ccgs, the expression of these genes appears to be clock regulated. Expression of the new ccgs peaked between CT0 and CT6; therefore, they were morning specific as are previously identified ccgs (BELL-PEDERSEN et al. 1996 ). The newly identified ccgs were tested for light induction by hybridization to RNAs that were extracted from mycelia exposed to continuous light for intervals of 15 min to 2 hr, but none of them was light induced under these conditions (data not shown). Riboprobes were generated for both strands of the new ccgs and, as expected, all of them were found to encode mRNAs within their f1 strands (Fig 1).

View larger version (39K): In this window In a new window Download PPT slide   Figure 2. Rhythmic transcript accumulation of newly identified ccgs in frq+ and frq7 (29-hr period) strains. (A) Northern blot hybridization of total RNA (40 µg) extracted from frq+ and frq7 strains after growth in darkness for the indicated constant darkness (DD) times. The corresponding CT (circadian time) is given below the DD values. Probes used for hybridization are indicated on the left. (B) Densitometry of ccg mRNAs. Normalization was to ethidium-bromide-stained rRNAs on the gels. The lowest value was set to 1. For each graph, at least three independent experiments were used for the calculations of averages and error bars, two of which were conducted with a frq+ strain, and one with a frq7 strain. (In the case of lyz, the last three data points consist of only two independent experiments, one with a frq+ strain and one with a frq7 strain.)

Sequence similarities and chromosomal locations of the new clock-controlled genes:
Comparison of the newly identified ccgs to the databases revealed the following (Table 3): ccg-13 shows strong sequence similarity to a phase-specific protein from the ascomycete Ajellomyces dermatitidis (GenBank accession no. AF277086, P/E value 3e^-10). The putative 5' nontranslated region of contig 1432 (representing ccg-13) shows identity to the end of cosmid H37F12, which was sequenced by the University of Georgia mapping project (available at http://gene.genetics.uga.edu/). This cosmid was mapped to linkage group I by the same group (linkage data at http://www.fgsc.net/mappingdata/mapping1.htm); therefore, ccg-13 should be located on linkage group I. ccg-14 is very similar to snodprot1 from the ascomycete Phaeosphaeria nodorum (GenBank accession no. AF074941, P/E value 7e^-39), a protein that displays strong similarity to cerato-platanin, a phytotoxin from the ascomycete Ceratocystis fimbriata with N-terminal homology to the hydrophobin family (PAZZAGLI et al. 1999 ). It is contained in contig 9A11 from the MIPS N. crassa database (http://www.mips.biochem.mpg.de/proj/neurospora/), which maps to linkage group V. ccg-15 displays strong sequence similarity to an SPS2 homolog from Saccharomyces cerevisiae (Saccharomyces Genome Database via RefSeq, accession no. NP_009634, P/E value 9e^-45) and is contained in clone 3e7 from the MIPS N. crassa database, which also maps to linkage group V. Thus, although there are clear sequence homologs to ccg-13, ccg-14, and ccg-15, there is at present nothing about the biology of Neurospora to suggest biologically (or phenotypically) meaningful names for these genes, and so they have provisionally been identified as ccg-13, 14, and 15, pending functional studies. Contig 1442 shows the strongest homology to lysozyme from the ascomycete Chalaropsis sp. (Swissprot accession no. P00721, P/E value 4e^-85; FELCH et al. 1975 ); therefore, the gene was named lyz. This sequence was not present in either of the Neurospora mapping projects (MIPS N. crassa database or Fungal Genome Resource at http://gene.genetics.uga.edu/).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Many aspects of the developmental cycle of N. crassa are under the control of the circadian clock (LOROS 1998 ). As was shown previously, clock control of gene expression can be observed in submerged shaken cultures, although development is suppressed under these conditions (LOROS et al. 1989 ). cDNAs were therefore sequenced from two mycelial libraries that correspond to two different times within the circadian day (BELL-PEDERSEN et al. 1996 ) to identify genes that are expressed under these conditions and to find new clock-controlled genes.

A large number of previously unknown genes were identified by sequencing cDNAs from two mycelial libraries of N. crassa:
The RNAs used as templates for the construction of the cDNA libraries sequenced in these studies were derived from cultures held under starvation conditions in the dark for 43 hr (BELL-PEDERSEN et al. 1996 ). cDNA libraries sequenced in previous studies in Neurospora correspond to genes expressed in vegetative mycelial, conidiating, or sexually developing cultures (NELSON et al. 1997 ; DOLAN et al. 2000 ). Thus, because the growth conditions were distinct, we expected to identify a large number of novel genes, and this appears to be the case. Among the 1431 genes that were identified within the 20,000 sequences generated, 721 (50%) did not show any homology to previously identified genes within the NCBI database, whereas 710 genes have matches among the sequences within the NCBI nonredundant databases, as of November 2000. As the total number of Neurospora genes is estimated to be between 10,000 and 14,000 (KELKAR et al. 2001 ), it is surprising that the number of expressed genes identified here represents only 10% of the potentially expressed sequences of N. crassa. However, the ESTs generated within this project provide a valuable tool for fungal researchers and can be used for further studies of gene expression, including as probes for microarrays in transcriptional profiling experiments.

The relative abundance of cDNAs in the two libraries does not necessarily reflect gene expression patterns in vivo:
To our knowledge, there have been few reports of "digital Northerns" that were confirmed by conventional methods at a larger scale (see, however, PRADE et al. 2001 ). Hwang and co-workers identified genes involved in cardiac hypertrophy (HWANG et al. 2000 ), and HUANG et al. 1999 found several genes differentially expressed in prostate cancer by EST profiling. Both groups confirmed their results using Dot Blot and reverse transcription PCR techniques to survey the expression of 22 and 7 candidate genes, respectively. In an approach to construct the transcriptional profile for skeletal muscle, BORTOLUZZI et al. 2000 found that electronic expression profiles correlated with serial analysis of gene expression results for overall expression levels of the highly expressed genes in muscle. Efforts have been made to investigate the genes for plant lipid biosynthesis and to compare electronic expression profiles with enzyme activities in vivo (MEKHEDOV et al. 2000 ). The authors find similarities as well as differences between in silico and in vivo, but as enzyme activities do not necessarily reflect mRNA expression patterns, the reasons for these differences remain to be clarified.

In this investigation, hybridization of RNA time courses with 26 probes derived from cDNA clones representing different contigs did not reveal a strong correlation between the relative abundance of a cDNA among the EST sequences and the expression pattern of the gene in vivo. Of the 19 genes with a high probability for differential expression (AUDIC and CLAVERIE 1997 ), three turned out to be under clock control and showed peak mRNA levels in the subjective morning, whereas the other 16 did not reveal consistent patterns of differential expression (Table 1). Among the two genes with probabilities for differential expression between 0.95 and 0.999, none is under clock control (Table 1). Screening of 5 genes that displayed probabilities of <0.95 for differential expression did reveal one clock-controlled gene (Table 1), which might indicate that these significance levels lead to many false-positive but few false-negative results. Overall, the number of false positives is relatively high, being 82%, which does not indicate a strong correlation between EST abundance and in vivo expression.

There are several possible explanations for this finding. One might be that the number of sequences obtained is still too low to result in a representative picture of gene expression. The number of genes encoded by the Neurospora genome has been estimated between 10,000 and 14,000 (NELSON et al. 1997 ; KELKAR et al. 2001 ); therefore, sequencing 13,000 cDNA clones might not be enough to make reliable predictions about the situation in vivo. This possibility is strengthened by the finding that the five out of seven previously identified ccgs, which are predicted correctly by the digital Northern (Table 2), are among the N. crassa genes with the highest expression levels in vivo as well as the highest count numbers in the libraries. This might also indicate that our EST data are valuable to estimate overall expression levels, similar to investigations done by other groups (BORTOLUZZI et al. 2000 ; MEKHEDOV et al. 2000 ), but that the more subtle changes regulated by the circadian clock are not yet visible among the number of ESTs generated to date. Also to be taken into account is the fact that genes that are neither under clock control nor under any stringent control mechanism under the chosen growth conditions may display random fluctuations of expression. This could mimic the peak and trough patterns of clock-controlled genes when viewed only at two time points.

Another issue to be considered is the evaluation of the statistical significance of the data available. Most of the genes used as probes displayed differences of counts in the morning vs. the evening libraries that indicate differential expression with probability levels of >0.999 according to formulas developed to evaluate the outcome of electronic transcriptional profiling (AUDIC and CLAVERIE 1997 ). However, this significance level might still be too high given that HWANG et al. (1999) used significance levels <0.0002 to identify putative candidate genes for differential expression in cardiac hypertrophy and still found the ratio of false positives to be 10–30%. Also, the formula does not require a minimal number of EST counts, which might be required to make valuable predictions (see above). Plus, the ratios were calculated with the assumption that the overall number of mRNAs present in the mycelium is the same in the morning and the evening. This assumption may not be correct, as MARTENS and SARGENT 1974 found that overall RNA contents in Neurospora mycelium grown on solid medium cycle rhythmically over the circadian day, although with low amplitude. The authors found that the amplitude of the RNA rhythm was even lower in liquid medium, but as they did not use starvation conditions, this might not be true for the mycelia that were used for the libraries in the present study. Also, their study dealt with overall RNA content, not with mRNA.

If the number of mRNAs is greatly different in the morning vs. the evening, then sequencing about the same number of cDNA clones for both libraries will not give an accurate picture. Also, the presence of multiple cDNAs for highly expressed genes (such as ccg-1, ccg-2, and ccg-7; see Table 2) may lead to an underrepresentation of genes with lower expression, especially at the time when the highly expressed genes reach their peaks; the 20 genes with the highest count numbers constitute 32% of all ESTs in the morning library and only 11% of all ESTs in the evening library (data not shown). Another caveat might be that some of the clones in the libraries were sequenced from both ends and others only from one end, meaning that some clones would be counted twice. But as this occurs randomly within the whole data set, it may be assumed that this would not influence the overall outcome of the analysis. Still another reason might be that the two libraries were made from different strains (the morning library from a frq⁷ strain and the evening library from a frq⁺ strain; BELL-PEDERSEN et al. 1996 ) to harvest tissues at the same developmental stage but at different times within the circadian cycle. Although the strains were believed to be isogenic, except for the frq allele, the formal possibility exists that strain differences might contribute to the divergence of sequencing data from in vivo expression, although until now, no difference in overall expression levels between frq⁺ and frq⁷ strains was observed (BELL-PEDERSEN et al. 1996 ).

Four new clock-controlled genes were identified:
Hybridization of rhythmic RNA time courses with probes representing different contigs resulted in the identification of four new clock-controlled genes, ccg-13, ccg-14, ccg-15, and lyz. The new ccgs vary greatly in their amplitudes and robustness of rhythm (Fig 2). Since the signals and the robustness of the rhythm seen with ccg-15 and lyz were significantly reduced, there might be some doubts about the extent of rhythmicity. Nevertheless, these data are reported as they may indicate different levels of clock control. The Neurospora clock is thought to exert its function over a variety of different pathways and is also interconnected with other levels of regulation such as nutrition and development (LOROS 1998 ; DUNLAP 1999 ). Therefore, it is likely that a variety of ccgs are expressed with different degrees of amplitude and robustness of rhythm. The new ccgs peak in the early subjective morning, as do previously identified clock-controlled genes (BELL-PEDERSEN et al. 1996 ). ccg-13 shows similarity to a phase-specific protein from A. dermatitidis (GenBank accession no. AF277086), ccg-14 is strongly homologous to snodprot1 from the ascomycete P. nodorum (GenBank accession no. AF074941), and ccg-15 displays strong sequence similarity to a sporulation-specific protein (SPS2) homolog from S. cerevisiae (Saccharomyces Genome Database via RefSeq, accession no. NP_009634), but the actual function of all three genes in Neurospora remains to be elucidated. lyz is quite similar to lysozyme from the ascomycete Chalaropsis sp. (Swissprot accession no. P00721) and therefore almost certainly encodes the corresponding enzyme in Neurospora; it might be involved in defense mechanisms. The fact that 4 out of 26 genes tested displayed rhythmic expression confirms the prediction that a significant proportion of N. crassa genes is under clock control. Future studies may reveal additional clock-influenced gene expression patterns in Neurospora.

	FOOTNOTES

¹ These authors contributed equally to this work.

	ACKNOWLEDGMENTS

The authors thank Dr. Christian Heintzen for sharing materials and stimulating discussions. This research was supported by grants from the National Institute of Health (R37-GM 34985 to J.C.D. and MH44651 to J.C.D. and J.J.L.), the National Science Foundation (MCB-0084509 to J.J.L.), the Norris Cotton Cancer Center core grant at Dartmouth Medical School, and by a grant from the National Science Foundation EPSCoR program to B.A.R. M.N. received an Emmy-Noether-Fellowship from the German Science Foundation (Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg, Germany).

Manuscript received October 11, 2000; Accepted for publication December 19, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHAFFER, J. ZHANG, and Z. ZHANG et al., 1997  Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402[Abstract/Free Full Text].

AUDIC, S. and J.-M. CLAVERIE, 1997  The significance of digital gene expression profiles. Genome Res. 7:986-995[Abstract/Free Full Text].

BELL-PEDERSEN, D., M. L. SHINOHARA, J. J. LOROS, and J. C. DUNLAP, 1996  Circadian clock-controlled genes isolated from Neurospora crassa are late night to early morning specific. Proc. Natl. Acad. Sci. USA 93:13096-13101[Abstract/Free Full Text].

BIRNBOIM, H. C. and J. DOLY, 1979  A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1522[Abstract/Free Full Text].

BODENTEICH, A., S. CHISSOE, Y. F. WANG and B. A. ROE, 1994 Shotgun cloning as the strategy of choice to generate templates for high-throughput dideoxynycleotide sequencing, pp. 42–50 in Automated DNA Sequencing and Analysis Techniques, edited by M. D. ADAMS, C. FIELDS and J. C. VENTER. Academic Press, London.

BOGUSKI, M. S., T. M. LOWE, and C. M. TOLSTOSHEV, 1993  dbEST—database for "expressed sequence tags.". Nat. Genet. 4:332-333[Medline].

BORTOLUZZI, S., F. D'ALESSI, C. ROMUALDI, and G. A. DANIELI, 2000  The human adult skeletal muscle transcriptional profile reconstructed by a novel computational approach. Genome Res. 10:144-349.

DAVIS, R. L. and D. DESERRES, 1970  Genetic and microbial research techniques for Neurospora crassa.. Methods Enzymol. 27A:79-143.

DAVIS, R. W., 2000 Neurospora: Contributions of a Model Organism. Oxford University Press, New York.

DOLAN, P. L., D. O. NATVIG, and M. A. NELSON, 2000  Neurospora proteome 2000. Fungal Genet. Newsl. 47:7-24.

DUGGAN, D., M. BITTNER, Y. CHEN, P. MELTZER, and J. TRENT, 1999  Expression profiling using cDNA microarrays. Nat. Genet. 21:10-14[Medline].

DUNLAP, J. C., 1999  Molecular bases for circadian clocks. Cell 96:271-290[Medline].

EWING, B., L. HILLIER, M. C. WENDL, and P. GREEN, 1998  Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res. 8:175-185[Abstract/Free Full Text].

FELCH, J., T. INAGAMI, and J. HASH, 1975  The N, O-diacetylmuramidase of Chalaropsis species. V. The complete amino acid sequence. J. Biol. Chem. 250:3713-3720[Abstract/Free Full Text].

HWANG, D. M., A. A. DEMPSEY, C.-Y. LEE, and C.-C. LIEW, 2000  Identification of differentially expressed genes in cardiac hypertrophy by analysis of expressed sequence tags. Genomics 66:1-14[Medline].

HUANG, G. M., W. NG, J. FARKAS, L. HE, and H. A. LIANG et al., 1999  Prostate cancer expression profiling by cDNA sequencing analysis. Genomics 59:178-186[Medline].

KELKAR, H. S., J. GRIFFITH, M. E. CASE, S. F. COVERT, and R. D. HALL et al., 2001  The Neurospora crassa genome: cosmid libraries sorted by chromosome. Genetics 157:979-990[Abstract/Free Full Text].

LOROS, J. J., 1998  Time at the end of the millennium: the Neurospora clock. Curr. Opin. Microbiol. 1:698-706[Medline].

LOROS, J. J., S. A. DENOME, and J. C. DUNLAP, 1989  Molecular cloning of genes under the control of the circadian clock in Neurospora. Science 243:385-388[Abstract/Free Full Text].

MANIATIS, T., E. F. FRITSCH and J. SAMBROOKE, 1982 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

MARTENS, C. L. and M. L. SARGENT, 1974  Circadian rhythms of nucleic acid metabolism in Neurospora crassa.. J. Bacteriol. 117:1210-1215[Abstract/Free Full Text].

MEKHEDOV, S., O. MARTINEZ DE ILARDUYA, and J. OHLROGGE, 2000  Toward a functional catalog of the plant genome. A survey of genes for lipid biosynthesis. Plant Physiol. 122:389-401[Abstract/Free Full Text].

NELSON, M. A., S. KANG, E. BRAUN, M. CRAWFORD, and P. DOLAN et al., 1997  Expressed sequences form conidial, mycelial, and sexual stages of Neurospora. Fungal Genet. Biol. 21:348-363[Medline].

OHLROGGE, J. and C. BENNING, 2000  Unraveling plant metabolism by EST analysis. Curr. Opin. Plant Biol. 3:224-228[Medline].

PANDEY, A. and F. LEWITTER, 1999  Nucleotide sequence databases: a gold mine for biologists. Trends Biochem. Sci. 24:276-280[Medline].

PAZZAGLI, L., G. CAPPUGI, M. GIAMPAOLO, G. CAMICI, and A. SANTINI et al., 1999  Purification, characterization and amino acid sequence of cerato-platanin, a new phytotoxic protein from Ceratocystis fimbriata f. sp. platani. J. Biol. Chem. 274:24959-24964[Abstract/Free Full Text].

PRADE, R. A., P. AYOUBI, S. KRISHNAN, S. MACWANA, and H. RUSSELL, 2001  Accumulation of stress and inducer-dependent plant-cell-wall-degrading enzymes during asexual development in Aspergillus nidulans.. Genetics 157:957-967[Abstract/Free Full Text].

ROSENBLUM, B. B., L. G. LEE, S. L. SPURGEON, S. H. KHAN, and S. M. MENCHEN et al., 1997  New dye-labeled terminators for improved DNA sequencing patterns. Nucleic Acids Res. 25:4500-4504[Abstract/Free Full Text].

WHITE, O. and A. R. KERLAVAGE, 1996  TDB: new databases for biological discovery. Methods Enzymol. 266:27-40[Medline].

YARDEN, O., M. PLAMANN, D. EBBOLE, and C. YANOFSKY, 1992  cot-1, a gene required for hyphal elongation in Neurospora crassa encodes a protein kinase. EMBO J. 11:2159-2166[Medline].

This article has been cited by other articles:

				M. W. Vitalini, R. M. de Paula, C. S. Goldsmith, C. A. Jones, K. A. Borkovich, and D. Bell-Pedersen Circadian rhythmicity mediated by temporal regulation of the activity of p38 MAPK PNAS, November 13, 2007; 104(46): 18223 - 18228. [Abstract] [Full Text] [PDF]

				A. Mendoza-Mendoza, T. Rosales-Saavedra, C. Cortes, V. Castellanos-Juarez, P. Martinez, and A. Herrera-Estrella The MAP kinase TVK1 regulates conidiation, hydrophobicity and the expression of genes encoding cell wall proteins in the fungus Trichoderma virens Microbiology, July 1, 2007; 153(7): 2137 - 2147. [Abstract] [Full Text] [PDF]

				T. Akao, M. Sano, O. Yamada, T. Akeno, K. Fujii, K. Goto, S. Ohashi-Kunihiro, K. Takase, M. Yasukawa-Watanabe, K. Yamaguchi, et al. Analysis of Expressed Sequence Tags from the Fungus Aspergillus oryzae Cultured Under Different Conditions DNA Res, May 31, 2007; (2007) dsm008v1. [Abstract] [Full Text] [PDF]

				J. J. Loros, J. C. Dunlap, L. F. Larrondo, M. Shi, W. J. Belden, V. D. Gooch, C.-H. Chen, C. L. Baker, A. Mehra, H. V. Colot, et al. Circadian Output, Input, and Intracellular Oscillators: Insights into the Circadian Systems of Single Cells Cold Spring Harb Symp Quant Biol, January 1, 2007; 72(0): 201 - 214. [Abstract] [PDF]

				M. W. Vitalini, R. M. de Paula, W. D. Park, and D. Bell-Pedersen The Rhythms of Life: Circadian Output Pathways in Neurospora. J Biol Rhythms, December 1, 2006; 21(6): 432 - 444. [Abstract] [PDF]

				Y. Liu and D. Bell-Pedersen Circadian Rhythms in Neurospora crassa and Other Filamentous Fungi. Eukaryot. Cell, August 1, 2006; 5(8): 1184 - 1193. [Full Text] [PDF]

				H. Kim and M. A. Nelson Molecular and Functional Analyses of poi-2, a Novel Gene Highly Expressed in Sexual and Perithecial Tissues of Neurospora crassa Eukaryot. Cell, May 1, 2005; 4(5): 900 - 910. [Abstract] [Full Text] [PDF]

				D. M. Kupfer, S. D. Drabenstot, K. L. Buchanan, H. Lai, H. Zhu, D. W. Dyer, B. A. Roe, and J. W. Murphy Introns and Splicing Elements of Five Diverse Fungi Eukaryot. Cell, October 1, 2004; 3(5): 1088 - 1100. [Abstract] [Full Text] [PDF]

				M. W. Vitalini, L. W. Morgan, I. J. March, and D. Bell-Pedersen A Genetic Selection for Circadian Output Pathway Mutations in Neurospora crassa Genetics, May 1, 2004; 167(1): 119 - 129. [Abstract] [Full Text] [PDF]

				K. A. Borkovich, L. A. Alex, O. Yarden, M. Freitag, G. E. Turner, N. D. Read, S. Seiler, D. Bell-Pedersen, J. Paietta, N. Plesofsky, et al. Lessons from the Genome Sequence of Neurospora crassa: Tracing the Path from Genomic Blueprint to Multicellular Organism Microbiol. Mol. Biol. Rev., March 1, 2004; 68(1): 1 - 108. [Abstract] [Full Text] [PDF]

				M. Nowrousian, G. E. Duffield, J. J. Loros, and J. C. Dunlap The frequency Gene Is Required for Temperature-Dependent Regulation of Many Clock-Controlled Genes in Neurospora crassa Genetics, July 1, 2003; 164(3): 923 - 933. [Abstract] [Full Text] [PDF]

				H. Kim, R. L. Metzenberg, and M. A. Nelson Multiple Functions of mfa-1, a Putative Pheromone Precursor Gene of Neurospora crassa Eukaryot. Cell, December 1, 2002; 1(6): 987 - 999. [Abstract] [Full Text] [PDF]

				M. A. R. Milla, E. Butler, A. R. Huete, C. F. Wilson, O. Anderson, and J. P. Gustafson Expressed Sequence Tag-Based Gene Expression Analysis under Aluminum Stress in Rye Plant Physiology, December 1, 2002; 130(4): 1706 - 1716. [Abstract] [Full Text] [PDF]

				U. Nair, J. L. Ditty, H. Min, and S. S. Golden Roles for Sigma Factors in Global Circadian Regulation of the Cyanobacterial Genome J. Bacteriol., July 1, 2002; 184(13): 3530 - 3538. [Abstract] [Full Text] [PDF]

				M. L. Shinohara, A. Correa, D. Bell-Pedersen, J. C. Dunlap, and J. J. Loros Neurospora Clock-Controlled Gene 9 (ccg-9) Encodes Trehalose Synthase: Circadian Regulation of Stress Responses and Development Eukaryot. Cell, February 1, 2002; 1(1): 33 - 43. [Abstract] [Full Text] [PDF]

				H. S. Kelkar, J. Griffith, M. E. Case, S. F. Covert, R. D. Hall, C. H. Keith, J. S. Oliver, M. J. Orbach, M. S. Sachs, J. R. Wagner, et al. The Neurospora crassa Genome: Cosmid Libraries Sorted by Chromosome Genetics, March 1, 2001; 157(3): 979 - 990. [Abstract] [Full Text]