Analysis of Pneumocystis carinii Introns

Abstract

Pneumocystis carinii is an ascomycete phylogenetically related to Schizosaccharomyces pombe. Little is known about gene regulation in P. carinii. The removal of introns from pre-mRNA requires spliceosomal recognition of the intron-exon boundary. In S. pombe and higher eukaryotes, this boundary and a branch site within the intron are conserved. We recently demonstrated that P. carinii cdc2 cDNA can complement S. pombe containing conditional mutations of cdc2, an essential gene involved in cell cycle regulation. We next tested whether P. carinii genomic cdc2 (with six introns) could also complement S. pombe cdc2 mutants and found genomic sequences incapable of this activity. Reverse transcriptase PCR confirmed the inability of the S. pombe cdc2 mutants to splice the P. carinii genomic cdc2. Analysis of 83 introns from 19 P. carinii protein-encoding genes demonstrated that the sequence GTWWDW functions as a donor consensus in P. carinii, whereas YAG serves as an acceptor consensus. These sequences are similar in S. pombe; however, a branch site sequence was not found in the P. carinii genes studied.

Pneumocystis carinii is a significant cause of morbidity and mortality in immunosuppressed patients, especially those with AIDS or malignancies or following organ transplantation (13, 20, 27, 30). P. carinii is a fungus which is phylogenetically classified as an ascomycete. Accordingly, P. carinii is related to the fission yeast Schizosaccharomyces pombe (6). The inability to continuously culture P. carinii is a major hindrance in understanding the biochemistry and cell biology of this important pulmonary pathogen (23). Consequently, little is known about gene regulation and expression in P. carinii.

The removal of intervening sequences (introns) from pre-mRNA is an important step in gene regulation, requiring a complex number of splicing molecules and an active spliceosomal structure which recognizes distinct features of introns in order to permit intron excision. In protein-encoding genes, the nucleotide sequences at the splice junctions between exon-intron, or donor sequences, and intron-exon, or acceptor sequences, are well defined and usually adhere to a consensus motif. Breathnach et al. (5) have shown that the nucleotide sequences between donor and acceptor splice junctions are not random. Instead, introns begin with GT and end with AG (5). This has since been validated for a large number of protein-encoding genes and currently has been expanded to MAG/GTRAGT for the donor site consensus and (Y)_nNYAG/G for the acceptor site consensus (15). While these rules apply to protein-encoding genes, they do not pertain to mitochondrial, tRNA, or rRNA splice junction sequences (15). In addition, a site upstream of the acceptor sequence, known as the branch site, is also conserved between S. pombe and higher eukaryotes. A detailed analysis of the nucleotide sequences comprising the splice junctions of P. carinii genes has not previously been performed (24).

Mutant strains of S. pombe have been useful in understanding the function of a number of genes from heterologous species, especially genes controlling the cell cycle which encode the cdc (cell division control) molecules (7–9, 12, 21). S. pombe strains containing temperature-sensitive mutations of a particular cell cycle control gene grow normally at the permissive temperature of 25°C but are arrested when grown at the restrictive temperature of 37°C. These S. pombe mutants, however, grow normally at the restrictive temperature if the defective gene is replaced by a functional gene from another organism, thereby complementing the defective gene (7–9, 21). A number of cDNA sequences from organisms as diverse as plants and mammals have successfully been used to complement S. pombe temperature-sensitive mutants in cell division control molecules. We have recently shown that the P. carinii cdc2 cDNA can complement temperature-sensitive S. pombe mutants in cdc2, allowing the yeast to proliferate at the restrictive 37°C temperature (26).

It has further been proposed that S. pombe may represent a good model for the study of eukaryotic gene expression and regulation, since the intron features in S. pombe are similar to those of higher eukaryotes (10, 29, 31, 32). For instance, Nurse and coworkers have spliced the viral simian virus 40 (SV40) small-t antigen in S. pombe, suggesting that some of the machinery required for splicing is conserved between S. pombe and other higher organisms (10). Since P. carinii is phylogenetically similar to S. pombe, we hypothesized that P. carinii introns might be spliced in S. pombe as well. Accordingly, we tested whether genomic P. carinii cdc2 sequences would complement growth of temperature-sensitive S. pombe cdc2 mutants, in a fashion similar to P. carinii cdc2 cDNA. In addition, we analyzed the structure of splice junctions from P. carinii protein-encoding genes to determine whether they conform to the general rules of eukaryotic splice junction consensus sequences.

Plasmid construction.

The plasmid pCDC2I was constructed as follows. The 1,200-bp region from the start codon to the stop codon (including six introns) of the P. carinii cdc2 gene was amplified by PCR with the 5′ NdeI primer, TTTTCATATGGAGCAATATCAGAGGTTAGAG, and the 3′ BamHI primer, TTTTGGATCCCTATAGCACCACATTAGATCTATT, with genomic P. carinii cdc2 previously cloned into the plasmid pGEM-7Z(f-) as a template (26). The PCR product was restriction digested with NdeI and BamHI and directionally cloned into pREP41. The plasmid pREP41 is an S. pombe nmt1 expression plasmid which is repressed by thiamine and contains the leu2 gene for auxotrophic selection of transformants on media lacking leucine (2, 14) pCDC2I was sequenced completely to confirm that no PCR errors were introduced into the construct. The plasmid pCDC2C is the pREP41 plasmid containing the P. carinii cdc2 cDNA, and pSPCDC2 is the pIRT2 shuttle vector containing the wild-type S. pombe cdc2 gene with four introns (26).

S. pombe transformation and complementation.

S. pombe temperature-sensitive cdc2 mutants were obtained as a gift from K. Gould, Vanderbilt University. The S. pombe mutants were grown overnight in yeast extract (plus supplements) medium at 30°C to an optical density at 595 nm of 1.0 and were electroporated with the plasmids pCDC2I, pCDC2C, and pSPCDC2 as previously described (18, 26). Following electroporation, the S. pombe cells were plated on minimal medium plates lacking leucine and thiamine and grown at 30°C for 5 days. Transformed colonies were tested for complementation by incubation at 37°C. Despite our prior success with complementing S. pombe cdc2 mutants with P. carinii cdc2 cDNA in the identical vector (26), we were unable to complement these temperature-sensitive S. pombe cdc2 mutants with the P. carinii cdc2 genomic sequence which contained six introns. We screened more than 10,000 colonies transformed with the genomic sequences at 37°C without identifying a single complementing colony. In comparison, the P. carinii cdc2 cDNA expressed in the S. pombe mutants yielded 10 complementing clones per 1,000 transformants.

Splicing analysis.

Reverse transcriptase PCR (RT-PCR) was used to confirm the inability of the S. pombe temperature-sensitive mutants to splice the P. carinii genomic cdc2 introns. Transformed S. pombe temperature-sensitive mutants harboring either pCDC2I, pCDC2C, or pSPCDC2 were grown in liquid minimal medium lacking leucine and thiamine to an optical density at 595 nm of 1.0 at 30°C, and the cell pellets were frozen at −70°C until needed. Total RNA was extracted by incubating the S. pombe pellets in buffer (10 mM Tris [pH 7.5], 10 mM EDTA, 0.5% sodium dodecyl sulfate) and an equal volume of phenol (pH 4.5) at 65°C for 60 min, followed by chloroform extraction and ethanol precipitation. The samples were then treated with DNase I at 37°C for 15 min. Five micrograms of total RNA was used from each preparation to make cDNA with an oligo(dT) primer and Moloney murine leukemia virus RT as previously described (26). PCR amplification of the cDNA was performed with 1 μM (each) primer set and 35 cycles of amplification. The primer set PC (5′ primer, TTTTCATATGGAGCAATATCAGAGGTTAGAG, and 3′ primer, TTTTGGATCCCTATAGCACCACATTAGATCTATT) flanks the P. carinii genomic cdc2 sequences of exons 1 and 7 of pCDC2I and the open reading frame of the P. carinii cdc2 cDNA of pCDC2C. The primer set SP (5′ primer, ATGGAGAATTATCAAAAA, and 3′ primer, AGATAATTTTGTTGCAAA) corresponds to the S. pombe genomic cdc2 sequences of exons 1 and 5 of pSPCDC2. The primer set SPI (5′ primer, ATGGAGAATTATCAAAAA, and 3′ primer, CTTGACAAAATGGTTAGT) corresponds to the S. pombe genomic cdc2 sequences of exon 1 and intron 4 of pSPCDC2. PCR amplicons were analyzed by agarose gel electrophoresis and visualized by ethidium bromide staining (Fig. 1). A single 1,200-bp amplicon was generated by RT-PCR from pCDC2I expressed in S. pombe, indicating that the pre-mRNA was transcribed but that intron splicing did not occur. This PCR amplicon was sequenced to verify that splicing did not occur. Correct intron splicing would generate a 900-bp product, which is evident in the RT-PCR of S. pombe expressing pCDC2C. Both the pre-mRNA and the spliced mRNA from S. pombe expressing pSPCDC2 were detected by RT-PCR.

FIG. 1.

Inability of S. pombe cdc2 temperature-sensitive mutants to splice P. carinii cdc2 introns. M, DNA ladder. Lanes 1 to 4, RT-PCR amplifications of total RNA from S. pombe cdc2 mutants transformed with the plasmids pCDC2C, pCDC2I, and pSPCDC2. Lane 1, pCDC2C (P. carinii cdc2 cDNA) amplified with primers flanking the open reading frame of the cDNA (corresponding to exon 1 and exon 7 of the genomic DNA) has a 900-bp product which is the correct size for the cDNA. Lane 2, pCDC2I (P. carinii cdc2 DNA with six introns) amplified with the identical primer pair has a 1,200-bp product which corresponds to the size of the pre-mRNA, demonstrating that the pre-mRNA is transcribed but that all six introns are retained in the mature mRNA. Correct splicing would have generated a 900-bp product as seen in lane 1. Lane 3, pSPCDC2 (S. pombe cdc2 genomic DNA) amplified with primers from exon 1 and exon 5 has an 891-bp product which is the correct size for the cDNA. As expected, the pre-mRNA is not visible. Lane 4, pSPCDC2 amplified with primers from exon 1 and intron 4 demonstrates that the pre-mRNA is transcribed (1,106-bp product).

Intron sequence analysis.

We next analyzed 83 introns from 19 P. carinii protein-encoding genes available through the GenBank database. These genes were chosen by searching the GenBank database for all P. carinii nucleic acid sequences and identifying genes which had introns present. We evaluated the three nucleotides preceding the exon-intron junction and the six nucleotides following the junction for the presence of a donor boundary consensus. We further evaluated the six nucleotides preceding the intron-exon junction and the three nucleotides following the junction for an acceptor boundary consensus. These sequences were aligned and evaluated for consensus by using GCG software. To further determine a branch site consensus, the acceptor sites were aligned and the 30 nucleotides upstream of the acceptor site were analyzed for consensus. Additionally, the four introns of S. pombe cdc2 were compared to the six introns of P. carinii cdc2 for the above features.

We observed that the splice junction sequences at the donor and acceptor sites in P. carinii conform to the consensus sequence for eukaryotic protein-encoding genes. Of the 83 introns from 19 P. carinii protein-encoding genes from the GenBank database, the sequence DWD/GTWWDW was found to be a consensus sequence for the exon-intron (donor) site and WWWYAG/DDW was found to be a consensus sequence for the intron-exon (acceptor) site (Table 1). Our analysis further revealed that in general P. carinii introns are small (Table 2), ranging from 38 to 100 bp in length. The P. carinii calmodulin gene has an anomalously large intron (424 bp), which is significantly larger than the other P. carinii introns.

TABLE 1.

P. carinii splicing donor and acceptor site consensus sequences from 83 introns^a

Nucleotide	Frequency for consensus
	Donor									Acceptor
	D	W	D	G	T	W	W	D	W	W	W	W	Y	A	G	D	H	W
G	16	6	53	83	1	3	9	32	8	5	6	7	0	0	83	27	11	13
T	22	26	14	0	78	13	37	23	39	35	54	38	60	1	0	14	28	38
A	31	42	14	0	0	61	32	22	26	39	23	36	11	82	0	30	32	25
C	14	9	2	0	4	6	5	6	10	4	0	2	12	0	0	12	12	7

^aThe consensus sequence represents 75% certainty for each position. The values are the frequencies at which each nucleotide is represented at each position of the sequence. The sequence appears in the International Union of Pure and Applied Chemistry nucleic acid code: D, not cytosine; H, not guanine; M, adenine or cytosine; N, adenine, cytosine, guanine, or thymine; R, adenine or guanine; W, adenine or thymine; Y, cytosine or thymine. 

TABLE 2.

P. carinii genes analyzed for spliceosome recognition components and intron content

GenBank accession no.	Protein encoded	No. of introns	Intron size range (bp)
AF001305	Protease 1	7	38–45
AF026546	Cdc2 cyclin-dependent kinase	6	46–57
AF061071	SUC1 Cdc2-binding protein	4	45–57
AF097334	CDC13	2	44–70
D31909	MSG99	1	48
D49831	Actin	7	43–49
L05466	Beta-tubulin	8	41–100
L05572	Calmodulin	3	60–424
L18918	Pentafunctional enzyme	1	45
M25415	Thymidylate synthase	4	45–55
M26495	Dihydrofolate reductase	1	43
M86602	Folic acid synthesis	3	43–47
M95294	Alpha-tubulin	8	41–69
M96931	DNA polymerase II	2	42–49
S77510	Thymidylate synthase	2	45–55
U14410	TATA binding protein	4	45–63
U17121	Ribosomal protein	2	41–54
U30790	GTP alpha subunit	9	40–50
U30792	GTP alpha subunit	9	41–48

As mentioned earlier, the branch site sequence is also an important determinant of pre-mRNA splicing. Studies of gene expression in the budding yeast Saccharomyces cerevisiae have shown that this yeast cannot correctly remove introns from the transcripts of other organisms (3, 11, 28). Although the donor and acceptor sites between S. cerevisiae and other eukaryotes are conserved, S. cerevisiae has a characteristic TACTAAC branch site sequence upstream of the acceptor site which is unique (10). In contrast, S. pombe and higher eukaryotes have a homologous CTRAY branch site consensus sequence (10, 29, 31). Interestingly, our analysis of P. carinii intron sequences could not identify a characteristic branch site sequence upstream of the 3′ acceptor site. In particular, P. carinii does not have homology to the branch sites in S. cerevisiae, S. pombe, or other higher eukaryotes.

A comparison of the nucleotide sequences comprising the splice junctions in S. pombe and P. carinii cdc2 genes revealed further important differences (Table 3). S. pombe cdc2 has four introns, while P. carinii cdc2 has six. All of the S. pombe cdc2 introns have the branch site sequence CTRAY upstream of the acceptor site, while this sequence was not present in the P. carinii introns. Further, although the donor and acceptor site sequences for P. carinii in general fit the consensus sequences denoted above, P. carinii cdc2 introns five and six deviated from this consensus at the donor site (Table 3). If the last two introns in P. carinii cdc2 had not been correctly spliced, then the catalytically active t loop of cdc2 (containing the threonine required for activation) would not be correctly expressed, which would render the Cdc2 protein nonfunctional (8, 17).

TABLE 3.

Comparison of P. carinii cdc2 introns with S. pombe cdc2 introns^a

Intron	Donor	Acceptor
P. carinii cdc2 intron 1	AAG/GcAtaa	TttaAG/GgA
S. pombe cdc2 intron 1	AAG/GtAggt	TgctAG/GaA
P. carinii cdc2 intron 2	aaG/GTAtca	aaTtAG/ACT
S. pombe cdc2 intron 2	tcG/GTAagt	ttTaAG/ACT
P. carinii cdc2 intron 3	GAa/GTAggT	taTTAG/tTT
S. pombe cdc2 intron 3	GAg/GTAtaT	gtTTAG/aTT
P. carinii cdc2 intron 4	Ttg/GTAAtt	tgttAG/GtT
S. pombe cdc2 intron 4	Tct/GTAAga	gtcaAG/GcT
P. carinii cdc2 intron 5	tag/gtcaag	tttcag/aat
P. carinii cdc2 intron 6	cag/gttttc	atgtag/aaa

^aLowercase indicates mismatches between the sequences. 

Removal of introns from pre-mRNA occurs in a catalytically active spliceosome and requires a number of small nuclear RNA protein (snRNP) molecules and accessory factors which recognize and bind conserved splice junction sequences in the pre-mRNA to allow precise excision of the introns (1, 19). A number of snRNP molecules have been identified and appear to have distinct roles in pre-mRNA splicing. For example, the U1 snRNP binds to the 5′ donor splice site, the U2 snRNP binds the branch site, and the tri-snRNP U4/U6.U5 binds the 3′ acceptor site through U5. Cleavage of the donor and acceptor sites with subsequent ligation of the exons occurs in a complex reaction (1, 19).

Alternative modes of splicing are also known to exist. Trypanosomes, nematodes, trematodes, and some mammalian cells utilize trans splicing as a means to regulate gene expression (4, 16, 22). trans splicing involves the joining together of exons from different pre-mRNA transcripts into a single mature mRNA. In trypanosomes, a unique spliced leader RNA contains a unique sequence (a spliced leader sequence) and the 5′ exon and donor site, which are required for splicing (4, 16). All trypanosomatid pre-mRNAs contain the spliced leader sequence (16). Although the splice donor and acceptor sites of trans splicing may conform to eukaryotic cis-splicing consensus sequences, the mechanism and the machinery involved in trans splicing are different. In addition to cis- and trans-splicing mechanisms of pre-mRNA processing, a special class of pre-mRNA introns from plants and vertebrates which contain noncanonical consensus sequences has been identified (25). These unique pre-mRNA introns have differences at both splice sites and the branch site, with AT replacing GT at the donor site and AC replacing AG at the acceptor site. This led to the discovery of a novel spliceosome utilizing four snRNPs that are unique but analogous to U1, U2, and U4-U6 snRNPs (25).

Nurse and coworkers discovered that, although S. pombe could correctly splice the intron from the SV40 small-t antigen, it was unable to do so for the SV40 large-T antigen (10). Both SV40 introns had the S. pombe branch site sequence, and yet the SV40 large-T antigen intron differed in the donor site sequence, replacing an adenine at nucleotide 5 of the donor sequence (conserved in S. pombe) with a thymine. This difference in the donor site was hypothesized to inhibit splicing (10). Indeed, P. carinii cdc2 introns four and six have a thymine at nucleotide position 5 in the donor site sequence (Table 3). These differences may have contributed to the lack of complementation of the P. carinii genomic cdc2 gene in the S. pombe cdc2 mutants if the introns preceding these had not been correctly spliced, but we determined that none of the six introns were spliced.

These data demonstrate that S. pombe could not correctly splice the six introns from the P. carinii cdc2 gene. Although we tested only a single P. carinii gene for splicing, the number of introns tested suggests that S. pombe may not be an ideal organism to study P. carinii gene regulation since no introns were spliced. P. carinii may use alternative modes of splicing introns from pre-mRNA, such as trans splicing, or may have evolved unique branch site recognition strategies for intron splicing. Additionally, P. carinii may require novel snRNPs and other spliceosomal components. Clearly, further investigations into the mechanisms of RNA processing in P. carinii, which may yield new insights into the genetic regulation of this important pulmonary pathogen, are needed.

Nucleotide sequence accession numbers.

Gene accession numbers used in this study are noted in Table 2.