Pneumocystis carinii: Has the Name Really Been Changed?

In a recent article in Clinical Infectious Diseases, Mofenson et al. [1] make what appears to be a statement of fact—that the name of Pneumocystis carinii infecting humans has been changed to Pneumocystis jiroveci and that the nomenclature P. carinii is now reserved for organisms infecting rats.

That differences exist among P. carinii organisms in human lungs and those from the lungs of lower animals has been known for several decades. Early studies showed that isolates from human lungs did not infect lower animal species. In 1972, the first evidence of antigenic differences between human and rat strains of P. carinii was reported [2], and a species difference was proposed. Extensive studies of antigenic and molecular structure, beginning in 1986, have provided convincing evidence of protein differences and genetic divergence between the organism obtained from human lungs and the organism obtained from lower animal species [3–7]. Thus, P. carinii from humans is well-established and generally accepted as a distinct form of the organism.

Less well-defined than the science is the taxonomy and nomenclature of P. carinii. In recent years, several attempts at nomenclature have evolved to deal with differences among organisms found in various animal species. A simple nomenclature, published in 1988, added the animal host of origin to P. carinii (e.g., P. carinii, humanus; P. carinii, ratti; etc.) [8]. A more detailed system was proposed by Stringer et al. [9] in 1994, also designating organisms by host of origin (e.g., P. carinii f. sp. hominis and P. carinii f. sp. carinii for human and rat species, respectively). This was realistic, rational, and functional and was gaining widespread acceptance. Of importance, either of the aforementioned systems allow for clearly identifying unique types or strains of P. carinii without the requirement of proving that there are actually distinct species of Pneumocystis. However, Stringer et al. in 2001 [10] proposed a new system based on the nucleotide sequences of the organism and subsequently proposed the controversial term of P. jiroveci [11]. This same nomenclature for human P. carinii was proposed >25 years ago and was not accepted by the community of P. carinii researchers [12]. The recent proposal drew prompt objections by some [13] and rebuttal by the authors [14]. Now, >3 years after the last proposal, the nomenclature of the organism remains confused. In reviewing the titles of articles involving Pneumocystis pneumonitis in humans published between January 2004 and July 2005, one finds that, of the 84 publications on the human disease, 24 (30%) used P. jiroveci, 33 (40%) used P. carinii, and 26 (31%) simply used the genus Pneumocystis without species designation. Identification of P. carinii that infects animals is also affected, with proposed nomenclature based on both host species and nucleotide sequence diversity [10, 11]. There appears to be confusion on the part of authors, reviewers, and editors as to an “official” nomenclature for P. carinii. Basically, there is no official or authoritative body to establish terminology for fungi. The most recent proposal of Stringer et al. [11] is based on the guidelines of the International Code of Botanical Nomenclature (ICBN). The purpose of this commentary is to address the need to reconsider the issue.

I believe there remains much confusion on the part of authors, reviewers, and journal editors as to the significance of the proposed change in Pneumocystis nomenclature, likely because the proposed change in nomenclature has not come from any authoritative body on taxonomy, nomenclature, or mycology, or from any scientific society or appointed study group. This raises several issues.

First, the impression is given that the name change was made after debate and consensus building [11, 14, 15]. This was not the case. At an International Workshop on Opportunistic Protists, one individual asked to make an unscheduled presentation and, during that presentation, announced that he had submitted the change from P. carinii to P. jiroveci for human-derived Pneumocystis according to ICBN guidelines. However, to this point in time, it is unclear whether there is any consensus, as evidenced by a simple search of publication databases. Furthermore, it is impossible to determine how often use of the new nomenclature is voluntary on the part of the authors or, as will be discussed below, directed by reviewers or editors.

Second, the proposed system of species designation includes provisions that differences in nucleotide sequences can also be used as the basis for species designation, despite the fact there are no distinct biological features associated with these changes (i.e., species are being suggested on the basis of genomic sequence differences, with little evidence of differences in biological behavior) [10]. Although there may be a legitimate argument for species designation on the basis of host range, using these genetic differences to define species raises many questions. For example, using the new systems, 5 or 6 distinct species of Pneumocystis that infect rats have been proposed, and 3 have been “formally” named [10, 15]. This is despite the fact that there are no unique biological features that would seem sufficient to distinguish them as distinct species. Furthermore, using this basis to define species, one must say that simultaneous infection with >1 species of Pneumocystis is common [16, 17]. Is this biologically plausible, especially with an organism acquired by the airborne route? For example, stable changes in the electrophoretic karyotype (both chromosome size and number) and/or gene sequences, often associated with a recognizable phenotype, are well described in Candida albicans. According to results obtained using molecular epidemiological probes, these genetic variants are designated as either different strains of C. albicans or as mutants derived from a parental strain, rather than as a new species. Should this not prompt serious discussion by the microbiologist, infectious diseases specialist, and/or immunologist to decide, in the context of what we know about Pneumocystis and other organisms, why these genetic differences, in the case of Pneumocystis, should be considered the basis for defining a unique species? Taxonomy takes into account all of the features of an organism for purposes of classification, not solely genetic sequences.

Third, the stated rationale for the change in nomenclature is to foster scientific understanding and communication. The success in achieving this goal is debatable. For example, a recent article in Infection and Immunity discussed mouse P. carinii as P. muris [18], which at the time was not yet named as a unique “species” and which has subsequently been named Pneumocystis murina [19]. Furthermore, the use of molecular changes in Pneumocystis that allow for multiple species to infect a common host (and which are not discernible on any biological basis) can hardly be considered to be an aide to fostering communication. Many reports have resorted to using only the genus name. What is the effect on scientific communication when experiments are described without defining what species of organism was under study? Speciation on the basis of genotype requires special technology available in only some of the research laboratories working with this organism. Use of this form of speciation for P. carinii that infects the rat gives an example of how cumbersome this system would be. Experiments done with the rat model of P. carinii would have to be verified as involving a single “species” of organism. If results obtained in a series of experiments performed over time revealed that the precise number and ratio of “species” of Pneumocystis infecting a given rat population were not identical, would this finding invalidate the biological results? To answer “no” to this question would negate the basis for the nomenclature. To answer “yes” to this question would result in invalidating reproducible biological and/or immunological findings on the basis of an arbitrary degree of genetic variability without any evidence that these genotypic changes in P. carinii involve any alterations in the fundamental biology of the organism. To date, there are no data to suggest that the various “species” of Pneumocystis that infect rats have biological differences that rise to the level of indicating unique species. The subtle differences that have been noted would be consistent with strain variation, at best. Strain variants of microbes can be quite stable and are thus consistent with what has been observed with P. carinii.

Fourth, a potential flaw in the proposed system is that it calls on individuals to submit new names for as-yet-unnamed Pneumocystis organisms. Because the decision to name species is left up to individuals, what is to prevent anyone from choosing whatever name they prefer, as Frenkel did with “jiroveci” [12]? Ferret P. carinii was first recognized and described in studies performed at St. Jude Children's Research Hospital, in Memphis, Tennessee. Thus, it would be consistent with microbiological convention to name this “species” of Pneumocystis after the city of Memphis, where the organism was first isolated, or even after the discoverer. Because the overwhelming majority of “species” are currently “undiscovered” at this point, anyone can submit a new species name for any of the Pneumocystis organisms that infect each mammalian host that has not yet been specifically named. If individuals choose such an approach, what effect will this have on the desire to have an organized system to name Pneumocystis derived from monkeys, chimps, rabbits, dogs, horses, cows, or goats, for example?

Finally, is the finding of multiple “species” that infect rats an inherent part of the biology of this organism, compared with P. carinii that infect other hosts, or is it merely a reflection of the detailed analysis that has been performed on rat P. carinii? Given what we know so far, it would not be surprising if, with more-careful analysis, the same molecular variation noted in different “strains” of rat Pneumocystis would be found in Pneumocystis infecting other mammals. Preliminary analysis of ferret-derived P. carinii suggests that this is in fact the case [20]. If we adopt this concept of “phylogenetic species” [19] for human P. carinii, what does this portend for the number of different “species” of Pneumocystis that infect humans? I am unaware of clinical data suggesting different virulence or pathogenicity among human P. carinii, but it is quite likely that there will be multiple “species” of Pneumocystis infecting humans if we resort to the concept of phylogenetic species for Pneumocystis. This leaves open the potential for more (rather than less) confusion when dealing with human P. carinii. This system can hardly be considered to be user friendly. At least a potential advantage of a system of naming limited to the host of origin is that the organisms could be easily identified by host species or by simple assays using antibody reagents, many of which already exist or could easily be produced.

The issue of organisms infecting humans needs to be specifically addressed. Hughes [13] makes the case that the name P. carinii should be used for organisms that infect humans and designated as originating from humans (P. carinii of human origin or P. carinii f. sp. hominis). P. carinii has little veterinary (i.e., economic) relevance, given the fact that it is an opportunistic pathogen. The reason one studies P. carinii is because of the disease it produces in human beings. Stringer et al. [14] make the case that, with well-known syndromes and disease agents, these name changes are difficult to adopt, because of the effect that they have on daily communications. However, they go on to state, the new information dictates that these name changes be made [14]. Although, as stated at the outset, the use of host-species specificity as a basis for speciation is a legitimate question for scientific debate, this concept clearly requires careful consideration. I propose that, if such a change is adopted, serious consideration be given to how to retain the name P. carinii for the organism that infects humans. Given the importance of P. carinii as a human pathogen, it makes little sense to disrupt the clinical literature that has been generated during the past half-century.

Many of the points raised above could be—and likely will be—debated. However, there is no debate about the significance of the comment in the literature that the new species name is “valid” according to ICBN guidelines [11, 14]. I believe that what use there is to date of the name P. jiroveci is because of a misunderstanding of the significance of the ICBN nomenclature by reviewers and/or editors who then insist on the use of the new name as scientifically accepted and correct. The ICBN does not make any decisions with regard to taxonomic placement of organisms or scientific correctness of organism nomenclature. Rather, this is simply a code of rules governing the process by which taxa (e.g., species and genera) are assigned names. Thus, when a name is considered to be “valid” by the ICBN, this simply refers to the process by which the name was determined to be available for use, and it should not be interpreted as an indication of scientific or taxonomic validity. Indication that a name is valid under the guidelines established by the ICBN carries no mandate whatsoever that the scientific community is obligated to use this name. Rather, it is up to the scientific community to decide whether a newly designated name is the correct name for a recognized taxon (e.g., species). If there is consensus in this regard, then the new name will be incorporated into general usage. If it is considered to be poor taxonomy, then the name will not be adopted. Therefore, there is no mandate to use the name P. jiroveci when describing Pneumocystis of human origin. I have been involved as author, reviewer, or collaborator in the recent reviews of many papers dealing with Pneumocystis in which reviewers or editors have insisted on the use of this new nomenclature, likely because they were unaware of the scientific meaning of comments in the literature that the new name is “valid” [11, 14]. This forced use of the new terminology represents editorial censorship and can hardly be considered evidence of scientific consensus. Although reviewers and editors have the right and obligation to enforce accepted nomenclature, they should not use their positions to set taxonomic policy. Options are available under the ICBN to preserve the use of the name P. carinii for the human pathogen if this course of action would lead to nomenclatural stability and avoid name changes.

In summary, we have made great advancements in our understanding of P. carinii. However, the newly proposed nomenclature is not an advancement. Clear communication can be fostered by 2 preexisting mechanisms that are based on the host of origin. We first suggested such an approach >15 years ago [8]. English is the universal scientific language, and there is little confusion about using the host species along with the designation P. carinii (for example, human-derived, rat-derived, or rabbit-derived P. carinii). As an alternative, when more formal or precise designation is needed to describe a specific molecular strain of Pneumocystis, the tripartite name (P. carinii f. sp. hominis) can be used [9], which again clearly identifies the strain of organism being studied. Either of these systems offers the clarity in communication that was the stated intent of the change in Pneumocystis nomenclature. Furthermore, the nature of the organism being reported in the current literature would still be clear if species designations are subsequently deemed to be necessary.

At this point in time, I think there is real debate as to whether there is consensus in the scientific and medical community working with P. carinii about the need for a name change and whether the proposed system of nomenclature represents an improvement over current methods. Any final decision should involve interested individuals who are not only expert in the study of P. carinii, but also in the area of mycological taxonomy and nomenclature, as well. Such a task force could be developed by an independent scientific body, such as the National Institutes of Health, the American Society for Microbiology, or another international scientific organization.

• Received March 8, 2005.
• Accepted July 29, 2005.

• © 2005 by the Infectious Diseases Society of America