I wrote myself and in collaboration with colleagues a substantial number of reviews on the taxonomy of Pseudomonas (Palleroni et al., 1973; Doudoroff and Palleroni, 1974; Palleroni, 1984; 2005; 2008) and, inevitably, the present one repeats details that were described before elsewhere. The main difference between this ‘Pseudomonas Story’ and previous reviews will be in the designation of its chapters, which carry titles that have been borrowed from episodes of the history of mankind, and soon I shall explain some arguments in support of this decision, which I adopt with the proviso that I am not trying to depreciate History or to make fun of a respectable group of bacteria.
Our story will fit comfortably within the span of the twentieth century plus the few years of the present one, and it will be divided into the following stages: (1) Beginning; (2) Middle Age; (3) Renaissance; (4) Modern Times and (5) Contemporary Outlook. These names became familiar to us as milestones in the history of Western civilization, but their meaning is wide enough to be applied to periods in the evolution of our knowledge on Pseudomonas species. Therefore, they will be used here as subtitles of a story that, thank God, has nothing to do with human political affairs.
In the absence of a reliable fossil record, a safe assumption is that the origin of the pseudomonads and other groups of bacteria may have occurred many millions of years ago, but their scientific study started when we became aware of their existence by means of microbiological techniques. The conventional ‘Beginning’ of the story is followed by the ‘Middle Age’, a name that sounds more compassionate than ‘Dark Age’, which has the meaning of a blackout. The name of the third stage could be ‘Restoration’, which is less presumptuous than ‘Renaissance’, and this is followed by ‘Modern Times’, which will have nothing in common with Constantinople or with the beautiful Chaplin's movie. The last or ‘Contemporary’ stage is not very rich in traditional taxonomic studies, because it is directed mostly towards nucleic acids and proteins and in exploring the relationships among the structures and functions of these macromolecules. In a way, this exciting stage, which shall be mentioned very succinctly, has its roots in previous taxonomic studies that gave us a frame of phylogenetic relationships among the species of Pseudomonas and of some of the related genera.
In any case, the following pages are not meant to be just an inventory of reminiscences, in spite of the predominance of episodes taken from my personal experience, and as far as it was possible, I did my best to restrain the subjective element to an acceptable proportion.
The name Pseudomonas occurred to professor Migula of the Karlsruhe Institute in Germany at the very end of the nineteenth century (Migula, 1894; 1900). His description of the new genus was short and inaccurate but, even so, it was accepted for publication. It reads: ‘Cells with polar organs of motility. Formation of spores occurs in some species, but it is rare (for instance: Pseudomonas violacea)’. That was all. We now know that Pseudomonas strains do not produce spores, and it seems possible that Migula may have been observing refractile granules of reserve materials, which often look like spores, but even with this stipulation, Migula's definition seems rather useless to be of help in the identification of members of this genus. However, the odds were on his side when shortly afterwards he decided to propose as type species Pseudomonas pyocyanea, later renamed P. aeruginosa. This was indeed a most fortunate decision, and therefore, this species will be described in some detail in a later section.
As far as I know, Migula never clarified the etymology of the word Pseudomonas in any of his writings, but it suggested to later taxonomists a direct derivation from the Greek monas or unit. Thus, we find in the 7th edition of Bergey's Manual of Determinative Bacteriology of 1957, the formal, somewhat telegraphic definition: ‘Gr. pseudes false; Gr. monas a unit, monad; M.L. fem. n. Pseudomonas false monad’. This etymology was transcribed without further comments by Michael Doudoroff and myself for the description of the genus in the 8th edition of the manual in 1974, and I inserted it unchanged in the first and second editions of Bergey's Manual of Systematic Bacteriology (Palleroni, 1984; 2005 respectively).
Migula's description does not mean much, and the above etymology elicited some bad taste humorous comments. Therefore, I thought that it may be of interest to revive now one other possibility for which I could never find any written confirmation from Migula himself, and that it may have only the value of hearsay. Some years before I met Professor Santos Soriano, my microbiology mentor at the University of Buenos Aires, Argentina, he had spent some time in Beijerinck's laboratory in Delft, Holland, and he also had visited the Karlsruhe Institute in Germany, and from both places, he brought the information that Migula had not traced directly the Greek ancestry of the name, but had simply created the name Pseudomonas for the resemblance of the cells to those of the nanoflagellate Monas in both size and active motility. Migula's primary intention may have been to call the bacteria ‘false Monas’, and not ‘false units’. In any case, the point is no longer of much importance.
2. Middle Age
Even though the idea of ‘Middle Age’ in History is usually perceived as a sort of hibernation period, some truly remarkable events took place at this time in our story. These facts did not have an immediate impact, and their significance was appreciated much later, but at the time of their conception, for one reason or another, members of the scientific community interested in Pseudomonas biology and taxonomy seemed to have been indifferent to them.
The most important event was the description of the remarkable nutritional versatility of Pseudomonas species, which was published in 1926 (that is, during the core of our Middle Age) in the thesis of L.E. den Dooren de Jong (1926). He was the last student of the famous microbiologist Beijerinck, and he had been assigned the project of examining the soil microflora with respect to the degradation of organic compounds as part of the process of carbon mineralization. The results were quite dramatic, in the sense that bacteria of the genus Pseudomonas appeared to be endowed with the ability of decomposing a large variety of organic molecules, including many that are often toxic to microorganisms of other groups and to higher organisms. As a consequence of these results, one of the important conclusions of the Delft thesis was a recommendation to use the high diversity in nutrition properties of the strains for their description and taxonomy, but both the findings and the recommendations only reached a small circle of scientists, in part because the thesis was written in Dutch, a language that was not familiar to many bacteriologists of other countries. Therefore, it is indeed a remarkable fact that some of the results presented in the thesis were transcribed extensively a few years later in the superb book on bacterial metabolism by Marjory Stephenson (1939), although she was more attracted by the nutritional diversity itself from the biochemical point of view than by the possible taxonomic applications of the screenings.
In spite of the fact that the American science was quite rich in outstanding names of pioneer bacteriologists, few of them seem to have manifested much interest in Pseudomonas. An exception was Erwin Frank Smith, who had a strong background in botany and microbiology and worked in the Laboratory of Plant Pathology of the U.S. Department of Agriculture. He recorded in several important monographs (Smith, 1904; 1911) his observations on bacterial diseases in plants, the starting point of our knowledge on the many variants of Pseudomonas syringae, which are specifically pathogenic towards plants of different species.
Finally, it is important to point out that during the early decades of the twentieth century took place the foundation of a society by the American bacteriologists that, under the heading of David H. Bergey, decided to start the publication of a manual of determinative bacteriology with periodic updated records of bacterial taxonomy. Unfortunately, some important contributions from foreign laboratories were excluded from the first editions of the manual, but the insularity of the manuals was corrected in later editions. However, the momentum imparted by the activity of the first two decades of the twentieth century was eventually lost during the ‘Middle Age’ of this story.
In the meantime, the number of species assigned to the genus was constantly growing to reach quite unmanageable numbers. In the particular case of Pseudomonas taxonomy, the corresponding chapters of later editions of the Manual were written by Haynes and Burkholder (1957), who performed a laborious and most remarkable job of creating determinative keys that included the great majority of nomenspecies, but the properties used in the keys, borrowed from the more familiar sanitary bacteriology and the enteric organisms, were of very limited usefulness for the description and taxonomic determination of new isolates in the laboratory.
The chaotic condition of Pseudomonas taxonomy was much aggravated by the enormous number of species assigned to the genus. A rough estimate showed that there were more than 800 names of species assigned to Pseudomonas towards the middle of the twentieth century.
At the time, I was involved in the study of the carbohydrate metabolism by a strain of Pseudomonas saccharophila (Palleroni, 1956), a species now transferred to a different genus. The single strain of this species had been isolated from a mud close to the Berkeley branch of the University of California by Dr Michael Doudoroff. Mike and I became close friends, and his enormous enthusiasm and biological vocation were for me a constant source of scientific inspiration. Therefore, in a relatively short time, I had been able to come out with very interesting findings in the biology and carbohydrate metabolism of this hydrogen bacterium, which was a handy supplement to the information already available on this subject. Consequently, it was simply inconceivable for me to switch to any other field.
Once in a while, however, both Mike and myself were involved in discussions that had their origin in Dr Roger Stanier's concern for the chaotic situation of Pseudomonas taxonomy and nomenclature (Stanier and van Niel, 1962). Of course, the main objective of these talks among the three of us was Roger's effort to convince Mike and myself of the convenience of doing something to correct the situation. Unfortunately, simple negative objections did not work against Roger's arguments, and the discussions took place periodically until, half convinced by the convenience of such a project, I decide to give it a try. None of us, unfortunately, realized at the time that the hook that attracted us had barbs that prevented any decision of moving to an altogether different field, or that taxonomy could give us a day the pleasure of a boisterous celebration to welcome some satisfactory conclusions.
However, Roger was very pleased with our response, and soon rescued from his large book collection a beat-up copy of den Dooren de Jong's thesis, and I remembered my nightmares to try to disentangle parts of the text, using my poor knowledge of German, of which Dutch is actually a dialect. Of course, I had to carry out most of the experiments, which included many biochemical reactions, and in particular it involved the job of testing many strains for utilization of a large number of organic compounds, using solid media dispensed into regular glass Petri plates. A pile of 10 or more plastic dishes of the modern brands of plates makes an almost perfectly vertical column, but in contrast, the landscape of several tables covered with unstable piles of the irregular glass Petri plates was an unwelcome addition to my nightmares. The operations are very well described in the enormous manuscript that resulted from our enterprise, and essentially differed from those of the original Dutch thesis in the use of the replica plating invented by Esther Lederberg (Lederberg and Lederberg, 1952), a welcome simplification of the original manipulations of poor den Dooren de Jong.
A convenient series of phenotypic properties of the Pseudomonas strains subjected to this screening was collected, and the results were used in the creation of differential tables based mostly on the nutritional characteristics, and occasionally on some details of interest displayed by the cell patches on the agar. Several other tests were included in the original studies to supplement the comprehensive phenotypic descriptions (Stanier et al., 1966). At present the time- and labour-consuming job of collecting nutritional properties has been simplified considerably by the use of commercial kits, which are supplemented by tests of various enzymatic activities, but the need for these additional determinations is not very obvious.
In what follows, I include the example of a condensed description of the type species P. aeruginosa strains as one of the many examples collected in the long manuscript.
Pseudomonas aeruginosa is the ideal type species for the genus. Because of the relative phenotypic homogeneity of the species, many strains may in fact serve the purpose of ‘type’. Someone has suggested that organisms of this species are represented by the largest number of individuals of any species on the planet, and one of the consequences of its abundance and of its activity as a peaceful inhabitant or as a vicious opportunistic human pathogen, is the enormous volume of scientific and technical literature that is available on its physiology and pathogenicity (Kockgether et al., 2008). In the section on ‘Contemporary Outlook’ some of the findings obtained by molecular biology techniques on this species will be summarized.
The specific name aeruginosa had occurred years before to Schroeter because of the colour of the colonies on certain media, a colour resembling that of copper rust or verdigris, hence green. Schroeter attached the name to the genus Bacterium, and all Migula did was to transfer the species to the new genus Pseudomonas of his invention. The cells are seen isolated or in small groups, and most of them have a single polarly inserted flagellum. Some cases of abnormal insertion are presented in the second edition of Bergey's Manual of Determinative Bacteriology. The cells are absolutely aerobic but capable of denitrification or nitrate respiration, and are commonly saprophytic or opportunistic animal and human pathogens, with some weak propensity to become plant pathogenic. A detailed study of the opportunistic pathogenicity has revealed the presence of about 10 pathogenicity factors, but in spite of the fact that P. aeruginosa can indeed become a very nasty customer in patients with low immunological defenses (particularly in cases of cystic fibrosis), it has been safely handled by workers with good health in some laboratories wholly dedicated to the study of this species without any serious health problems. I spent some time in the laboratory of Professor Bruce Holloway in Clayton, Australia, where the normal operation was buttressed for many years by hygienic measures and correct bacteriological manipulations, without any help of highly sophisticated working conditions.
Pseudomonas aeruginosa decomposes arginine in anaerobiosis through the arginine dihydrolase (not dehydrolase, as in some papers) mechanism, which, in the absence of oxygen, is able to generate the energy source required for motility. Therefore, the cells swim actively in anaerobiosis when arginine is present. Cells produce several typical pigments, among them pyocyanine and fluoresceine. The type strain is a reservoir of basic properties that are found in different combinations in the other species of the genus. The nutritional properties of P. aeruginosa are too numerous to be mentioned here and they can be found in other bibliographic sources, particularly in the monographic study by Stanier and colleagues (1966). The list of degradable compounds includes a respectable complement of aromatic molecules, and the benzene ring is ‘dearomatized’ following the ortho-cleavage mechanism.
The definitive paper describing a tentative taxonomic outline of our Pseudomonas species occupied many pages of one issue of the Journal of General Microbiology in the year 1966, clearly demonstrating the advantages of using adequate methodologies for the phenotypic characterization of the pseudomonads. The paper was well received, and in a few months it reached the category of a ‘citation classic’, an award that is not very common for papers of taxonomic subjects. However, at the beginning of the decade, Roger Stanier and C.B. van Niel (1962) had expressed the practical inadequacy of the current taxonomic procedures to support conclusions of phylogenetic value.
When the paper was in press, I had to leave the country to attend some problems in Argentina, and after a year, I returned to Berkeley, when the paper was out and I was pleased to see that it was a handy guide to some of the graduating collaborators for the performance of phenotypic studies of groups of recently isolated strains of Pseudomonas species that were added as a supplement of our original plans.
During my absence from the laboratory I had the opportunity of reading two recent papers with data on the conservative nature of the ribonucleic acid of ribosomes (Dubnau et al., 1965; Doi and Igarashi, 1965), and some ideas on the application of this approach to taxonomic studies started developing in my mind, although I did not see too clearly the need to add this approach to our current studies.
4. Modern Times
A number of strains of the various Pseudomonas phenotypic groups were chosen for in vitro DNA/DNA hybridization experiments, which I had learned from the late John L. Johnson at the University of Washington in Seattle (Johnson and Francis, 1975). The experiments gave results of two different classes. On the one hand, the results confirmed the similarities observed on phenotypic properties among strains of the same phenotypic group, but on the other hand, the hybridizations failed to detect any homology whatsoever between pairs of strains of different groups, suggesting deep genotypic differences among groups of species that had been assigned to the same genus.
Roger, Mike and I discussed these findings without seeing much of a way out, but at the moment came to my mind the results of the two papers on the high level of conservatism among ribosomal RNA molecules (Dubnau et al., 1965; Doi and Igarashi, 1965), and consequently I mildly suggested to attempt to switch to the study of these cell components. As expected, my suggestion was not received with much enthusiasm by Mike, because of our basic ignorance about RNA work.
However, in spite of this general negative response, I decided to do some exploratory experiments. The RNA was extracted from ribosome pellets, and I used the nucleic acid extracted from them in hybridization experiments versus total cell DNA. As the results were very interesting, they were repeated and improved by the able collaboration of Rebecca Contopoulou and Riyo Kunisawa. In summary, the very interesting outcome was that the species that until then had been collected under the generic name Pseudomonas (that is, Pseudomonas sensu lato) were shown to form five sharply defined rRNA groups, one of which included the type P. aeruginosa, all the fluorescent species (P. fluorescens, P. putida, P. syringae), and some non-fluorescent ones (P. stutzeri, P. alcaligenes, P. pseudoalcaligenes, P. mendocina). The name Pseudomonas was reserved for this cluster (named rRNA group I or Pseudomonas sensu stricto). In our paper, which came out in 1973, we suggested that the other RNA groups should be assigned to other genera of the same or different families, but we did not go any further. Eventually, as expected, this renaming process was done in other laboratories.
I liked very much our paper, and the only thing that I thought we should have corrected was the title, mentioning ‘RNA homologies’ instead of simply ‘nucleic acid homologies’, to emphasize the special nature of our novel research plans. In part this may have lowered the interest of the potential readers, but the references to this work were soon numerous to a point of converting this piece of work into another Pseudomonas‘citation classic’.
As I said, we did not continue by taking the logical step of applying new names to each of the rRNA groups, and this attitude may have shown once again that we were not behaving as regular taxonomists. Nevertheless, the truth was somewhat more complicated. Before the paper appeared, Roger, who had not participated in the RNA stage of the work, had left his post to accept a position at the Pasteur Institute in Paris, Mike was fighting a losing battle with cancer, and the Berkeley Pseudomonas team practically disintegrated as I decided to move to a different place or to return for good to Argentina.
Historically speaking, this was the first example of the use of RNA as a taxonomic tool in prokaryotes and, of course, it was soon followed by contributions from other laboratories, the first of which was Johnson's work on Clostridium (Johnson and Francis, 1975), and the results resembled those obtained on Pseudomonas. The work was described in a paper where the potential of rRNA homologies in taxonomic studies was welcome and appreciated.
After a few years, the Pseudomonas results were confirmed by the Belgian group headed by Josef DeLey and in his laboratory they were greatly amplified with the able help of Paul De Vos, by including strains of many other nomenspecies in the experiments (de Vos and de Ley, 1983). Technical improvements were developed in other laboratories, and they included the determination of sequences of oligonucleotides and eventually, total sequence comparisons among the 16S component of the rRNA groups.
The usefulness of rRNA homologies to define taxonomic hierarchies at the genus or higher level became clearly evident, and this approach, often spiced with a touch of aggressiveness, was universally adopted by taxonomists (Woese et al., 1984). Of course, the high conservative nature of the rRNA reduced the resolving power required for species characterization and differentiation, but the possibility of defining discrete groups corresponding to generic or supra-generic categories was a novel and very valuable contribution to bacterial taxonomy. In addition to the studies on rRNA, several additional genes of a less conservative nature were gradually included to search for more precise circumscriptions of species and, consequently, for an evaluation of their phylogenetic relationships.
5. Contemporary Outlook
The number of species assigned to the so-called Pseudomonas sensu stricto group of pseudomonads has grown already to a considerable number and it is still growing, which is a consequence of the remarkable adaptability of pseudomonads to the conditions of different environments. By ‘conditions’ I mean mostly the chemical composition, which has to satisfy the need of carbon, nitrogen and energy, and a rather limited range of temperatures and hydrogen ion concentrations; pseudomonads are less tolerant to wide variations in these last parameters, and this results in the absence of truly thermophilic or acidophilic strains.
In the old times, the names assigned to new species usually referred to some striking phenotypic property such as growth requirements, colour production, colony appearance, etc., but the possible variations in macroscopic details are limited, and nowadays the specific names derive from those of friends or colleagues, which makes them very happy but does not help very much the identification of strains upon re-isolation. Many years ago, names like aeruginosa, fluorescens, putida, aureofaciens, or the name of the host plants for the many variants of P. syringae, were useful identification elements.
The description of new species for publication has now to satisfy strict requirements demanded by the so-called polyphasic concept, where molecular data have to be supplemented by comprehensive phenotypic characterizations. Even so, the molecular data usually receive more attention than classical phenotypic properties such as gelatin liquefaction, maximal and minimal growth temperatures, denitrification, pigment production, nitrogen and carbon compounds required for growth, growth factors requirements, and others. The use of commercial kits has simplified considerably the time and labour-consuming job of the determination of the nutritional versatility and some related characteristics, among which is the free bonus of identification of some enzymatic activities. However, I have reviewed some manuscripts sent for publication in which the description of some new species did not include the average number and insertion of the flagella, or omitted a picture of the cells under the electron microscope or under phase contrast conditions in vivo or after special staining procedures for flagella, fimbriae or reserve materials.
To paraphrase a familiar expression taken from Hamlet, not ‘all is rotten in the state of Taxonomy’. The work that has been done on phenotypic characterization of species has been quite useful for identification purposes. However, of the fluorescent species, the ones that have received preferential attention are P. aeruginosa, P. putida, P. syringae.
A phenomenal volume of work dedicated to P. aeruginosa points to the remarkable versatility of this important species. From a description given in the 4th volume of Pseudomonas edited by J.-L. Ramos, we learn that the important strain of P. aeruginosa PAO (a designation that comes from PseudomonasaeruginosaOne) was described as the type strain of the species and assigned the number 131 in the Berkeley collection. The genomic analysis of this organism has revealed the presence in the chromosome of segments that are characteristic of a conserved core, in addition to a variable accessory part with segments of unrestricted gene flow within the populations. With 67% of G + C in the DNA and a genome of 6.3 mega-bases, PAO has 8.4% of the genome coding capacity acting as sensors interacting with the environment or with the host, permitting an adaptive response to aggression and to antibiotic toxicity. One-third of the genes have no homology to any previously reported sequences (Levesque, 2006).
A considerable amount of work was dedicated to P. putida KT2440 (Reva and Tümmler, 2008). This strain derives from a well-known toluene-degrading organism initially known as P. arvilla mt-2 and later renamed to P. putida mt-2. Work performed on its oligonucleotide usage resulted in a recommendation to use this property as a complement of 16S rDNA sequencing in the genomic analysis of pseudomonads. The book edited by Professor P. Cornelis on molecular approaches to Pseudomonas is an excellent source of information on the biology of P. aeruginosa PAO and P. putida KT2440.
The P. fluorescens strains have not received much attention and perhaps nobody can predict how long the eclipse will last. A large number of P. fluorescens strains of our collection were subdivided into seven biotypes, a decision inspired by a 1959 thesis work by Adriaan Fuchs at Delft under the guidance of Professor Kluyver. Many strains of the species were divided by Fuchs into varieties on the basis of two properties, denitrification and levan formation from sucrose, and our own work was in fact an adaptation of Fuchs' conclusions to our own collection (Fuchs, 1959). And that was that. None of the strains received preferential attention of the kind that was given to P. aeruginosa or P. putida strains. Two of the P. fluorescens biotypes (E and D) had individual species names, but they were later joined under one of them (P. chlororaphis) because of priority reasons. These two species have a content of GC in the DNA that is intermediate between P. aeruginosa and the bulk of the other fluorescent species and biotypes. Another biotype (F) included one strain that was previously named P. lemonnieri.
During my visit to Professor Holloway in Australia, I was asked my opinion on the best strains of P. fluorescens that could be explored from the genetic point of view using the methodology that was very effective with P. aeruginosa strains PAO and PAT (of course, P. aeruginosaTwo!) on the construction of genetic maps and in the demonstration of the chromosome circularity. To my chagrin, the P. fluorescens biotype names that I could mention had already been tested, but they seemed refractory to the current genetic approaches.
In contrast to the case of P. fluorescens, an interesting non-fluorescent species with very characteristic colony appearance and physiological properties, P. stutzeri, has been the subject of a great deal of very interesting work under the able direction of Jorge Lalucat of the University of the Balearic Islands in Mallorca (Lalucat et al., 2006). The studies on this species have supported the new concept of genomovar, and of the various genomovars already studied, some have been promoted to the species hierarchy.
The work of Jean-Marie Meyer and his collaborators at the University of Strasbourg (Meyer et al., 2002) has added a new and very interesting strategy for species identification, which is based in the determination of the type of siderophores produced by pseudomonads of the fluorescent and non-fluorescent groups, and it has been tested with very good results on a large number of well-characterized species.
In summary, things are going well in the taxonomic front and the information obtained in this field may serve as a useful phylogenetic background of the proteomic and genomic advances in the Pseudomonas Story.
Good sources of methodology are now available for the study of macromolecules to obtain information for taxonomic comparisons among Pseudomonas species. It goes without saying that this goal may be reached quite far in the future, but it is hoped that studies similar to those outlined above may be applied soon to other promising candidates, so that they may receive the attention that they deserve. In our forecasts, however, we have to keep in mind, as someone has said, that history makes us wise, but not necessarily clever.