TEXT

Top Abstract Text References

Pneumocystis carinii synthesizes a number of distinct ⁷ and ⁸ 24-alkylsterols but not ergosterol (the target of several antimycotics), and the organism scavenges cholesterol from its mammalian host (5, 6, 14-16, 18, 24). Most fungal sterols have an alkyl group consisting of one carbon at C-24 of the side chain. In contrast, the sterols of P. carinii and many plants have either one or two carbons at that site (C₂₈ and C₂₉ sterols). The P. carinii sterols are excellent chemotherapeutic targets because mammals cannot synthesize 24-alkylsterols.

The C₃₂ 24-alkylated lanosterol compound pneumocysterol (17) was detected in only trace amounts in organisms isolated from the corticosteroid-immunosuppressesd rat P. carinii pneumonia (PCP) model. The sterol profiles of organisms isolated from this animal model are consistent and reproducible from preparation to preparation (5, 14, 15, 18). In contrast, pneumocysterol was found in various percentages, from trace levels to 50% of the noncholesterol sterols, in organisms isolated from cryopreserved human lungs (16). Replicate analyses of these human-derived samples were consistent and reproducible, suggesting biochemical differences in organism populations. No correlation between human immunodeficiency virus infection (HIV) and high levels of pneumocysterol was found. However, in that earlier study, the possible effects of diet, nonprescribed drugs, or other factors could not be ruled out as the basis for broad variations in the accumulation of this sterol. In the present study, variations in pneumocysterol in samples from the same pair of human lungs were noted, and thus, more than six different sites in the lungs from the same individual were analyzed.

A formalin-fixed pair of lungs from an AIDS patient who died from PCP was provided by M. Pereira (Tufts Medical School, Boston, Mass.). Approximately 100-g pieces were excised from different sites and homogenized with distilled water in a Waring blender for 2 min, and total lipids were extracted (2) at room temperature for at least 2 h. The sterols were prepared and analyzed by gas-liquid chromatographic (GLC) methods as previously described (16-18).

Formalin does not have functions that would interact with sterols, and it was experimentally shown in a previous study that formalin fixation did not alter the sterols of rat lungs (19). Also, formalin-fixed lung tissue (17, 19) and Pneumocystis organisms isolated from cryopreserved human lungs (16) contained the same steroidal compounds. In the present study, the GLC components designated peaks 13, 16, 19, 20 and 24 were considered the organism's signature sterol profile [24-methylcholest-7-en-3-ol (fungisterol), 24-ethylcholestadiene-3-ol, 24-ethylcholest-7-en-3-ol, 24-ethylidenecholesta-7,24(28)- diene-3-ol, and 24-ethylidenelanost-8,24(28)-diene-3-ol (pneumocysterol), respectively]. In addition to cholesterol, GLC peaks 5, 8, 10, and 15 (desmosterol, campesterol, cholest-5-en-3-one, and -sitosterol, respectively) are believed to originate from the host.

In the samples taken from different sites of the Pneumocystis-infected human lung, the same major Pneumocystis sterols were detected, demonstrating qualitative similarities among the sampled areas (three examples are shown in Table 1). However, the relative proportions of these sterols at each site were different. The variations were most dramatically illustrated in the proportions of the distinct P. carinii sterols. In this individual with PCP, pneumocysterol composed only 29% of the noncholesterol sterols at one site, whereas it composed 49 and 54% of the noncholesterol sterols at the two other sites. Fungisterol composed 20% of the noncholesterol sterols at one site, whereas it accounted for only 2 and 6% of the noncholesterol sterols at the other sites. These findings were similar to those obtained for organisms isolated from human lungs from different subjects with PCP (16). Pneumocysterol was found in concentrations up to 50% of the noncholesterol sterols in some samples, whereas it was present in only trace amounts in other organism preparations (16); the sterol profiles of three independent organism preparations are shown in Table 1. In all analyses performed thus far on organisms from the experimental rat PCP model, pneumocysterol has been found in only trace amounts (14, 15, 18).

Pneumocystis is now recognized as a group of genetically diverse organisms (species and strains) which are specific for a given host species (7-9, 13, 22, 25). There is strong evidence for more than one distinct species or strain infecting the same mammalian host species and the same animal (3, 4, 13). Recent data obtained by nucleotide sequence analyses have demonstrated distinct genotypes infecting humans, and these organism populations can coexist in the same individual (10-12, 20, 21, 23).

It is unlikely that the results of the present study are due to varying proportions of different life cycle stages (e.g., trophic and cystic stages). The sterol compositions of trophic forms (24) and mixed life cycle stages (6, 15, 18) of P. carinii f. sp. carinii do not differ. A more likely explanation of the wide range of pneumocysterol and fungisterol in the samples from different regions of the same human lung is that it represents biochemical differences between distinct Pneumocystis genotypes coinfecting the same individual. These regional differences may result from infections that were initiated at distinct loci (clones) by individual invading organisms. This suggestion is supported by the identification within the same experimental rat lung of foci consisting of distinct populations (1). These foci represent either genetically distinct organism populations or populations expressing different major surface glycoprotein antigens.

In the previous report of Pneumocystis sterols in organisms isolated from cryopreserved human lungs, the samples were from HIV-positive and HIV-negative individuals (16). No correlation was observed between the accumulation of pneumocysterol and the HIV infection status of the patients. It was unlikely that diet or drug therapy (or possible unprescribed drugs taken) could cause these vast differences in sterol profiles. Nonetheless, these factors could not be ruled out in that study. In the present study, these factors can be ruled out because the samples were from the lungs of a single individual. Thus, the most parsimonious explanation for differences in sterol profiles at different PCP lung foci and in different organism preparations from different patients is the presence of genetically distinct organism populations that differ in their biochemical makeups. Thus, one population may accumulate large amounts of pneumocysterol and another phenotypic or genotypic population may normally produce or accumulate this sterol in only trace amounts. Intermediate pneumocysterol values can then be explained by the presence of mixed infections consisting of the two phenotypes or genotypes in various proportions.