Molecular Analysis of Cryptococcus neoformans Mitochondrial Cytochrome b Gene Sequences

MATERIALS AND METHODS

RESULTS

The 386-bp fragment corresponding to nucleotides (nt) 445 to 830 of the Candida glabrata cyt b coding sequence (GenBank accession no. X53862 ) was analyzed in this study. The type strain of C. neoformans var. neoformans, IFM 48637, had no introns in this region of cyt b. However, all other strains investigated contained one intron that started after nt 46 of the sequence (Fig. 1); the size of this intron varied from 270 to 1,143 bp. Some isolates of C. neoformans var. gattii contained a second intron that began at nt 633 of the cyt b sequence (Table 1) or nt 363 of the coding sequence (Fig. 1). The size of the second intron varied from 515 to 529 bp. The sizes and locations of these introns were determined as described previously (2) through comparison of strains with and without introns to maximize amino acid identities. Isolates of C. neoformans var. neoformans contained the longest introns, which were 1,142 and 1,143 bp. Among isolates of C. neoformans var. gattii, IFM 47258, IFM 48818, and IFM 48822 contained introns similar in size (1,128 bp) to that of C. neoformans var. neoformans and with greater than 96% sequence identity. Although the intron sizes were variable, they were fixed for specific cyt b types (Table 1), with the exception of strains IFM 48636 and 48821 (cyt b type I) and IFM 5878 (cyt b type II).

• Open in new tab
• Download powerpoint

FIG. 1.

Comparison of coding sequences of the mt cyt b genes of various C. neoformans isolates. Dots, nucleotides that are identical to those of C. neoformans cyt b type I; arrows, positions of introns. Cyt b types are in parentheses.

Analysis of 386 bp of coding sequence of cyt b revealed 27 (6.99%) variable nucleotide sites (Fig. 1). To ensure that these variations were not due to polymerase errors, we used the TaKaRa Ex Taq polymerase, which has an approximately fourfold-lower error rate than standard Taq DNA polymerase. Moreover, we sequenced both strands of each PCR product and repeated each PCR and sequencing reaction. On the basis of these differences in cyt b, the C. neoformans strains we analyzed were divided into nine types: C. neoformans var. gattii comprising cyt b types I to V and C. neoformans var. neoformans comprising cyt b types VI to IX (Table 1 and Fig. 1). All strains with cyt b types I, IV, and V were serotype B; all strains with cyt b types II and III except IFM 5878 (serotype B) were serotype C. Serotype D strains had cyt b types VI and IX, and serotype A strains were cyt b type VIII. Of four serotype AD strains, one was cyt b type VII and the remaining three were cyt b type VIII. Although the cyt b sequences contained 27 variable nucleotides (Fig. 1), the deduced amino acid sequences revealed that only one of these substitutions was nonsynonymous. C. neoformans var. gattii isolates, with cyt b types I to V, had identical amino acid sequences (type AA-I). Similarly, isolates of C. neoformans var. neoformans, with cyt b types VI to IX, had identical amino acid sequences (type AA-II). Therefore, isolates of C. neoformans var. neoformans differed from those of C. neoformans var. gattii only at amino acid position 55 (Thr instead of Ser) (Fig. 2).

• Open in new tab
• Download powerpoint

FIG. 2.

Comparison of the deduced amino acid sequences encoded by the cyt b genes of various C. neoformans isolates. Dots, amino acids that are identical to those of C. neoformans cyt b type I; arrows, inserted positions of introns. cyt b types are in parentheses.

The phylogenetic trees (UPGMA) generated from the mt cyt b DNA and amino acid sequences are shown in Fig. 3. The strains of C. neoformans were distributed according to nine DNA types. The type strain of C. neoformans var. neoformans (IFM 48637) was the outgroup in the phylogenetic tree. IFM 48637 contained the most variable nucleotide sites in cyt b (Fig. 1) and showed only a distant relation with other strains, even those of C. neoformans var. neoformans. Strains with cyt b type V (C. neoformans var. gattii IFM 47258, IFM 48818, and IFM 48822) contained introns similar in both length and content to that of C. neoformans var. neoformans, and they were phylogenetically closer to C. neoformans var. neoformans. Analysis of mt cyt b sequences suggested the existence of nine DNA types of C. neoformans; however, the phylogenetic tree based on the deduced amino acid sequences divided the strains into only two varieties: C. neoformans var. neoformans and C. neoformans var. gattii.

• Open in new tab
• Download powerpoint

FIG. 3.

UPGMA-based trees showing the relations of various C. neoformans isolates generated from nucleotide sequences of the cyt b gene (exon) (a) and deduced amino acid sequences (b). Bar, number of nucleotide and amino acid substitutions per nucleotide site and amino acid site.

DISCUSSION

To our knowledge this is the first study where mt cyt b sequences were used to analyze the genetic relations of strains of C. neoformans. Isolates of C. neoformans var. gattii, the anamorphic state of F. neoformans var. bacillispora, had five types (I to V) for mt cyt b; however, the deduced amino acid sequences were identical. Similarly, the amino acid sequences encoded by mt cyt b genes of isolates of C. neoformans var. neoformans, which is the anamorphic state of F. neoformans var. neoformans, were identical although there were four cyt b types (VI to IX). This result is consistent with the varieties of C. neoformans var. neoformans and var. gattii distinguished by d-proline assimilation or reaction on l-canavanine-glycine-bromthymol blue agar (11, 14).

The present study also revealed that one serotype D strain (cyt b type IX) had no intron in the region sequenced; however, all other strains of C. neoformans contained one intron, and some isolates of C. neoformans var. gattii contained a second intron (Table 1 and Fig. 2). A recent study of the mt COX1 gene of C. neoformans indicated that the presence or absence of introns in COX1 is not serotype specific (22), which is similar to the outcome of our study. The same group also reported that serotype D strains contain more introns in COX1 than do serotype A strains. However, in our study of the cyt b gene, all serotype A and D strains except IFM 48637 contained a single intron and the introns from all strains were similar in size and nucleotide sequence. The first intron in cyt b of C. neoformans started at nt 47 (Fig. 2), which is the same position as intron 2 of Neurospora crassa cyt b gene. Rhodotorula acheniorum and Rhodotorula ferulica, two other basidiomycetous yeasts, have introns in the same location (2). Some isolates of C. neoformans var. gattii (cyt b types II, III, and IV) contained a second intron that began at nt 633 (Table 1), which is similar to the location of intron 5 of the Saccharomyces cerevisiae cyt b gene. These findings suggest two possible evolutionary events. The first possibility is that these introns appeared in these locations prior to the separation of these species and that some species lost introns over time. The second is that these introns appeared in these locations after separation of these species.

Analysis of the mt large rRNA gene and NADH dehydrogenase subunit 2 of C. neoformans revealed that serotype AD strains had either the serotype A or serotype D mtDNA genotype (29). In the present study, we obtained almost similar results for the distribution of cyt b gene types in serotype AD strains. Of four serotype AD strains, three had serotype A-specific cyt b (type VIII) and one had a unique specific cyt b (type VII) that was nearly identical to serotype D-specific cyt b (type VI), with only 1 nt difference.

In conclusion, we have shown that isolates of C. neoformans represent nine cyt b types; however, the deduced amino acid sequences indicate that there are only two varieties, C. neoformans var. neoformans and C. neoformans var. gattii.