Sequence diversity of transcripts from Pneumocystis carinii gene families MSR and PRT1.

Abstract

P. carinii has three gene families, MSG, PRT1, and MSR, the members of which are interspersed at the ends of chromosomes. Only one of these families, MSG, has been characterized with respect to transcription. It appears that only one MSG gene is transcribed per organism at any given time because MSG transcripts come from a unique locus called the UCS. This locus encodes the Upstream Conserved Sequence (UCS) present at the beginning of messenger RNA molecules that encode different MSG proteins [4–6]. The UCS in messenger RNA molecules encodes the signal peptide motif needed to direct nascent MSG proteins into the endoplasmic reticulum, the first step in their journey to the cell surface. MSG on the surface lacks the UCS-encoded peptide, suggesting that it is removed by proteolytic cleavage. A possible site for such processing is marked by a conserved lysine-arginine motif, which would presumably be a target for socalled dibasic processing proteases (DPP), which are also known as subtilisin-like proteases. Interestingly, the PRT1 family encodes DPPs. The number of PRT1 genes has not been determined, but most if not all of the members of this gene family are linked to MSG genes, and 80% of characterized telomeric gene clusters contain at least one PRT1 gene. These clusters average 1.5 PRT1 genes each. There are approximately 34 telomeres in the P. carinii genome. Therefore, there would appear to be between 34 and 51 PRT1 genes. Based on antibody studies, at least one PRT1 protein appears to be located on the surface of the microbe [2]. P. carinii telomeric gene clusters also contain members of a third gene family, MSR, the members of which encode peptides that are related to MSG. MSR and MSG genes are distinguished by dependence, or lack thereof, on the UCS locus for expression [3]. MSG genes that are not linked to the UCS locus appear to be silent, but MSR genes have not been found at the UCS, and MSR messenger RNA molecules lack the UCS. Instead of the UCS, MSR messenger RNA molecules begin with sequences that map to MSR genes. These regions of MSR messenger RNA molecules encode a motif that would be expected to direct MSR proteins into the secretory pathway. Therefore, it seems probable that MSR proteins either are secreted or are on the surface of P. carinii. There are probably between 34 and 51 copies of the MSR gene family. Previous studies of RNA made from P. carinii organisms from the lungs of immunosuppressed rats have established that transcripts from multiple MSG, PRT1, and MSR genes are often present in a single animal. While these observations show that multiple genes can be expressed in a population of P. carinii, they do not show that multiple genes are expressed in an individual organism. In the case of MSG, the involvement of the UCS provided a molecular tag that could be used to discern transcriptional regulation even though all of the organism populations analyzed were expressing more than one MSG. In these studies, populations of P. carinii were found to differ with respect to the MSG genes that were present at the UCS locus. Comparisons of UCS-linked genes and MSG messenger RNA molecules in different populations indicated that genes at the UCS are represented by mRNA, while genes not at the UCS are not. This correlation can be explained with a simple model whereby linkage to the UCS locus is necessary and sufficient for MSG gene transcription. Recently, we discovered that the number of MSG genes at the UCS locus could be drastically reduced by restricting the number of organisms that contribute to the original infection [1]. We found that immunosuppression of a P. carinii-colonized rat that has been kept in isolation since weaning and remains in isolation during immunosuppression produced P. carinii populations expressing one MSG in at least 90% of organisms. The discovery of these populations provided an opportunity to detect co-regulation, if it occurs. In the present study, we used a P. carinii population dominated by organisms with one UCS-MSG sequence to address two questions. (1) Do transcripts encoding MSG have the same sequence as the UCS locus? (2) Is transcription of the other two gene families, MSR and PRT, coregulated with MSG?