Cytochrome oxidase subunit 2 gene in Neurospora crassa mitochondria.

The nucleotide sequence of the cytochrome oxidase subunit 2 (COX2) gene has been obtained from cloned mitochondrial DNA segments of Neurospora crassa. The coding sequences have been identified on the basis of protein sequence homology with the subunit 2 of cytochrome oxidase from yeast and man. The postulated precursor of the N. crassa subunit 2 protein is 250 amino acids long, with a molecular weight of 28,700. As in the tRNA and rRNA genes, the subunit 2 gene is flanked by G + C-rich palindromic sequences, which are highly conserved in N. crassa mitochondria. Three major transcripts have been detected by Northern blot hybridization. A transcript of 1100 bases is tentatively considered the fully processed mRNA. Furthermore, S1 nuclease protection experiments have revealed that the putative subunit 2 mRNA has a 330 nucleotide long 5' leader sequence.

The nucleotide sequence of the cytochrome oxidase subunit 2 (COX2) gene has been obtained from cloned mitochondrial DNA segments of Neurospora crassa. The coding sequences have been identified on the basis of protein sequence homology with the subunit 2 of cytochrome oxidase from yeast and man. The postulated precursor of the N . crassa subunit 2 protein is 250 amino acids long, with a molecular weight of 28,700. As in the tRNA and rRNA genes, the subunit 2 gene is flanked by G + C-rich palindromic sequences, which are highly conserved in N. crassa mitochondria.
Three major transcripts have been detected by Northern blot hybridization. A transcript of 1100 bases is tentatively considered the fully processed mRNA. Furthermore, S1 nuclease protection experiments have revealed that the putative subunit 2 mRNA has a 330 nucleotide long 5' leader sequence.
Known mitochondrial genomes from different sources exhibit very important differences in their organization. Mammalian mitochondrial genomes differ from those of lower eukaryotes in gene organization, in number of genes, and in the absence of split genes and intergenic regions (1-5). The mitochmdrial genomes of Saccharomyces cereuisiae and Aspergillus nidulans show a very low G + C content, mainly because of the high A + T content of the intergenic regions, and also because of the strong bias for A + T-containing codons. The size of the mitochondrial genome of N. crmsa is 60 kilobase pairs, only slightly smaller than that of S. cereuisiae, while its G + C base composition (40%) is more similar to that of mammalian mitochondrial DNA. The higher G + C content of the mitochondrial genome of N. crassa might reflect a different codon usage, a different structure of intergenic regions, or a higher number of genes.
The overall organization of the mitochondrial genome of N. crmsa has been investigated by using yeast mitochondrial gene-specific probes (6, 7). The present work reports the sequence of the gene for the cytochrome oxidase subunit 2 (COX2) and its flanking regions. The gene for this subunit has been sequenced in yeast (ll), mammals (1-3), and maize (12). The amino acid sequence is conserved, but an intron has been found in the maize gene. In N. crassa, this gene has no introns. Work by RajBhandary and co-workers has demonstrated the existence of peculiar structures in the intergenic regions * Part of this research was supported by a grant from the Istituto Pasteur/Fondazione Cenci Bolognetti and from Progetti Finalizzati, Ingegneria Genetica del Consiglio Nazionale delle Ricerche. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. around the rRNA (S), several tRNA (9), and COX3 genes (10) of N. crassa mitochondrial DNA. Similar structures have been observed in the flanking regions of the COX2 gene. To investigate the possible function of these structures, a transcript analysis has also been performed.

MATERIALS AND METHODS
Plasmid Construction-Plasmid pE2/35 contains the EcoRI fragment 4 of wild type strain SL74A N. crassa mtDNA, cloned into the EcoRI site of pBR322. This fragment was shown to contain the genes for subunit 2 of cytochrome oxidase and for subunits 6 and 9 of the mitochondrial ATPase (6, 7). Plasmids pP16/34 and pP2/36 were obtained by the subcloning of pE2/35 (digested with PstI) into the PstI site of pBR322. They contain, respectively, the 930-and 750-bp' long PstI fragments of Fig. 1.
DNA Hybridization-Restriction digests of DNA were separated by electrophoresis on 1.5% agarose and then transferred to nitrocellulose (13). Mitochondrial DNA from the yeast petite DS 200A1 (Ref. 11  DNA Sequencing-Restriction fragments were labeled at their 5' ends with [Y-~'P]ATP (specific activity 2000-3000 Ci/mmol, from New England Nuclear), in the presence of T4 polynucleotide kinase, or a t their 3' ends with [a-'*P]cordycepin (specific activity 2000-3000 Ci/mmol, from New England Nuclear) in the presence of terminal transferase (17). The labeled fragments were separated into single strands on 6 or 8% polyacrylamide gels and sequenced by the method of Maxam and Gilbert (17).
SI Nuclease Mapping-The 930-bp long DNA fragment of Fig. 2 was cleaved with HinfI restriction endonuclease. The resulting fragments were labeled a t their 5' ends with T4 polynucleotide kinase and [Y-~*P]ATP and then separated into single strands on 6% polyacrylamide gels. About 5-10 ng of HinfI-PstI fragment, representing the coding strand (see Fig. 6) was hybridized for 2 h at 45 "C with 10 gg of N. crassa mtRNA and then treated with 600 units of S1 nuclease (Boehringer Mannheim) in 100 pl a t 37 "C for 60 min. The SIresistant fragments were fractionated on 6% polyacrylamide, 8 M urea gels, alongside a guanine-specific reaction (17) of 5' end-labeled DNA of known sequence.

PVUII
Pyu I1 Hinf I Hinf I Taq  the two portions of the COX2 gene. All restriction sites were confirmed by sequencing. The arrows represent the sites, the direction, and the extent to which the sequences were derived. The sequences starting from the PstI sites were also confirmed by labeling these sites at their 3' ends.
probe was obtained from the petite strain 200A1 (11) and contained only part of the gene sequence (probe 1, Fig. 1). The EcoRI fragment 4 was cloned in pBR322 (pE2/35, see "Materials and Methods"). To more accurately localize the gene within this fragment, the clone pE2/35 was digested with PstI and analyzed by Southern blotting and hybridization with the previously described probe. A PstI fragment 930 bases long that showed positive hybridization with the probe was cloned (pP16/34) and sequenced. The nucleotide sequence showed that this fragment contained only the region corresponding to 180 amino acids of the NH2-terminal end of the gene.
To identify the adjacent PstI fragment that contained the remaining part of the gene, a probe carrying the COOHterminal portion of the yeast gene was prepared from the same petite (probe 2, Fig. 1). Hybridization of the probe to a 750-bp fragment permitted the identification of the second part of the gene (Fig. 1). This latter fragment was cloned (pP2/36) and sequenced.
DNA Sequencing-The two fragments were sequenced using the restriction sites shown on the map in Fig. 2. In most cases, both complementary strands were sequenced and restriction sites were crossed. The sequence shown indicates a single PstI site within the gene. This site was generated by joining the sequence of the two fragments at the PstI site, assuming that no extra sequences occur between the two fragments. The comparison of the amino acid sequence obtained this way shows a high homology with the amino acid sequences of known subunit 2 genes (Fig. 4), implying that this assumption is correct.
The DNA sequence of the COX2 gene and its flanking regions is shown in Fig. 3. The regions discussed below can be distinguished in this sequence.

(a)
A G + C-rich region starts from the PstI site at the 5' end of the sequence and extends for 200 nucleotides. The sequence adjacent to the PstI site is complementary and inverted compared to that determined by Yin et al. (9) in a different region of this genome (see Fig. 5). The entire G + C-rich region shows several peculiarities: i.e. the sequence between nucleotides 82 and 140 contains a 10 nucleotide long sequence repeated six times with only minor modifications, and the region between nucleotides 150 and 190 contains several stretches of purines alternating with stretches of pyrimidines. The nucleotide sequence is that of the nontranscribed strand. The reading frame is written in the three letter amino acid code. A sequence repeated six times in tandem is underlined. The palindromic sequences around PstI or "PstIlike" sites are bored. Dashed lines represent "PstI-like" sites.  shows a high degree of homology with the corresponding yeast and mammalian proteins (Fig. 4).

I~G G G G G G C A T A A T T R A A T G T G A T G G G G A C C T T G A G C A A A~A G G G G~r~G A G T G G A G G A~G G~~C C C C~G A G G G A A G C T G A G C G A A
The N. crassa COX2 protein (deduced from the nucleotide sequence) is 250 amino acids long and has M , = 28,700. The amino acid sequence of the NH2 terminus of the mature subunit 2 of cytochrome oxidase in N. crassa has been determined by Machleidt and Werner (18). The reading frame, derived from the nucleotide sequence, exactly matches the known amino acid sequence. However, there are two AUG initiator codons located 3 and 12 triplets upstream of the aspartic acid codon that represents the first residue of the mature protein, thus suggesting the existence of a precursor polypeptide.
( d ) The sequence following the terminator codon is G + Crich and contains several palindromic regions, including two sequences that are highly homologous to the stem structures which contain PstI sites (Fig. 5).
Codon Usage-- Table I shows the codon usage in the COX2 gene of N. crassa mitochondrial DNA. As already shown with the COX1 (19) and COX3 (10) genes of N. crassa, there is a strong bias for codons terminating in uridine or adenosine. A similar bias has been observed in yeast (5) and Aspergillus nidulans (20) mitochondrial DNA. The codon AUA, which in mammalian (1-3) and yeast (21) mitochondria probably codes for methionine, is here considered an isoleucine codon. Comparison of the COX2 coding regions of yeast and N. crassa shows that, 7 times out of 12, the codon AUA in N. crassa corresponds to AUU or AUC in yeast. It is interesting to note, moreover, that the CGN series of arginine codons is not used at all, as in the case with both yeast ( 5 ) and A. nidulans (20).
Transcript Analysis-Mitochondria1 RNA from N. crassa was denatured with glyoxal, fractionated on agarose gels, transferred to diazobenzyloxymethyl paper, and analyzed by hybridization with a nick-translated probe from within the structural gene. The results, reported in Fig. 6a, show hybridization signals corresponding to RNA species of about 3200, 1850, and 1100 bases. Similar results were obtained by van den Boogaart et al. (22). To localize the 5' end of the 1100 nucleotide long transcript that is thought to be the mature mRNA, a nuclease S1 protection experiment was performed by using the Weaver and Weissman modification of the Berk and Sharp procedure (23).
A single-stranded DNA fragment 555 nucleotides long, carrying 160 nucleotides of the coding region and 395 nucleotides of the upstream flanking region, was used. The results which are reported in Fig. 6b, show that a fragment of 485 k 10 bases was protected from S1 digestion. This indicates that the 5' end of the putative messenger is localized 330 5 10 nucleotides upstream of the initiator codon. A schematic drawing is shown in Fig. 6c.

DISCUSSION
The 1520-nucleotide sequence containing the gene coding for the cytochrome oxidase subunit 2 (COX21 of N . crassa mitochondrial DNA has a single reading frame, beginning with an AUG initiator codon and ending with a UAA terminator codon. Starting from codon 13 of this sequence, this  I LIG LIVISIWIMILIY  T I I I V I T~Y ; -I S I -I K~  I G L~V I E L H D N I MIYIY ~I v l v I L F I V I V I G I W I I I L~L   SIIII R N I Y I I I S T K I   E~L~I T FIH DIH A L M I I F LIIIcIFI-----ILIv L Y A L F L T LITIT I__""""""_ " " " " " " " " " " " " " " "  . 6. Analysis of the transcripts of the COX2 gene. a, agarose electrophoresis pattern of mitochondrial RNA denatured by glyoxal (track 1 ) and hybridization of nick-translated fragment pP16/34 to the same RNA blotted onto diazobenzyloxymethyl paper (track 2). The arrow indicates a transcript of 1100 bases that may represent the mature mRNA. b, localization of the 5' end of the postulated mature mRNA by S1 nuclease analysis.
Track I , specific cleavage at guanine of a 5' end-labeled single-stranded DNA fragment of known sequence (850 nucleotides long). Track 2, probe hybridized to mtRNA and then subjected to S1 nuclease treatment (see "Materials and Methods"). c, interpretation of the S1 experiment. The numbers above the lines indicate the lengths of the fragments.