5. A SURVEY OF THE BACTERIA
A. The Eubacteriales and related assemblages
The group which naturally presents itself first is that of the Eubacteriales. These organisms always possess rigid cell walls, demonstrable in the larger species by plasmolysis. Motility is by no means universal, but where it does occur, it is always flagellar. Cell division is always by transverse fission. While true nuclei are absent, there have been a number of reports in recent years, based on the use of reliable cytological techniques, in which claims have been made for the presence of discrete masses of chromatin material which could be interpreted as primitive chromosome-like structures (Badian, 1933, Stille, 1937, Piekarski, 1937. The work of Schaede (1940) has, however, cast some doubt on these previous results). The endospore, found in certain groups in the Eubacteriales, is highly characteristic, occurring nowhere else among the bacteria. A few members (Azotobacter spp.) form cysts. Endospores and cysts are the only two types of resting stages found in the Eubacteriales. On a purely morphological basis one can arrange the representatives of the Eubacteriales into a hypothetical tree with four main ascending branches, as was done by Kluyver and van Niel (1936). Certain facts which have come to light since 1936 have tended to weaken parts of this scheme, but in spite of that it is still useful for illustrating broad trends within the group. Also, some of these trends clarify the relationships between the Eubacteriales and other assemblages. A slightly modified form of the tree is given in figure 1.
Starting from the hypothetical primitive coccus type, the first line leads through the micrococci to the sarcinae, culminating in the spore-forming sarcinae. These forms may be either gram positive or gram negative. Motility occurs infrequently in members of this group, and spore formation has so far been reported only in Sporosarcina ureae.
The second line consists of the polarly flagellated rods, starting with the pseudomonas type and leading through the vibrios to the spirilla. The representatives of this line are all gram negative with one possible exception. This is the species Listeria (Listerella) monocytogenes (Pirie 1940) which was originally described as a gram-positive polarly flagellated rod (Pirie 1927) but which has' been more recently reported as peritrichously flagellated (Paterson 1939), and hence perhaps does not belong in this line. Spore formation is rare, being well established for only one species, Sporovibrio dessulfuricans (Starkey 1938). There have also been reports, so far unsubstantiated, of spore formation in Spirillum species (see Starkey 1938, and Lewis 1940). The third line is morphologically highly diverse. It takes its origin in the streptococci, and passes through the lactic acid rods, the propionic acid and other corynebacteria, and the mycobacteria to the actinomycetes. All members of this line are gram positive and none form endospores. Until very recently motility was thought to be absent in the higher forms although motile streptococci (see Koblmiller, 1935) had been known for some time. However, the work of Topping (1937) has shown that there are organisms of the Mycobacterium-Proactinomyces type which are flagellated. This fact may ultimately necessitate a radical revision of this line since, if motility is conceded to occur all along it, there is no valid means of differentiating between gram-positive peritrichously flagellated rods of the Kurthia type (at present placed in the fourth line) and organisms belonging to the genera Mycobacterium and Corynebacterium. The fourth line of the Kluyver-van Niel scheme comprises all peritrichously flagellated rods; both gram-positive and gramnegative forms occur. It is here that the great majority of sporeformers have up till now been placed. This line is also an unsatisfactory one because, particularly in the spore-forming representatives, a number of morphologically rather distinct types can be recognized. It is probable that future work will show the necessity of drastic revisions, but on the basis of present knowledge it is difficult to make satisfactory modifications of this and the preceding line.
The above-mentioned four lines were postulated primarily on the basis of cell shape and mode of insertion of flagella. It is of importance to note that the two most satisfactory groups share an additional character, viz., the homogeneous behavior of the members with respect to the gram stain. In the remaining two lines no such uniformity exists at the present time. This makes it necessary to determine the relative values of the type of flagellation and of the gram reaction for systematic purposes. The various recent reports on the difficulties encountered in definitely ascertaining the mode of insertion of flagella (Pijper, 1930, 1931, 1938, 1940; Pietschmann, 1939; Conn, Wolfe and Ford, 1940) lead one to suspect that the gram stain may ultimately appear to be the more valuable.
There are two large groups which show obvious morphological relationships to the Eubacteriales; namely, the Actinomycetales and the photosynthetic bacteria.
The Actinomycetales have already been mentioned in connection with the third line in the Eubacteriales from which they are clearly derived. The delimitation of this order is a difficult and necessarily arbitrary matter, since the morphological series which takes its origin in the lactic acid bacteria runs practically unbroken to the most complex of actinomycetes. From the purely determinative standpoint it is probably most satisfactory to start the Actinomycetales with the Proactinomycetaceae, the group in which a primitive mycelium formation occurs. In this case a clear and simple delimitation of the order is possible. As was mentioned in the critique of Bergey's system, the inclusion of the mycobacteria and corynebacteria leads to confusion, since these forms can so readily be taken for representatives of the Eubacteriales. On the other hand, the dividing line between the genera Mycobacterium and Proactinomyces is a tenuous one. Like the Eubacteriales, the Actinomycetales always possess rigid cell walls. Schaede (1940) has shown that the distribution of chromatin material in them is identical with that in members of the Eubacteriales. The question of motility in this order must remain open for the present, although it seems likely that the motile organisms described by Topping (1937) should be placed here. All representatives of the Actinomycetales are grampositive; for this reason the inclusion here of the polarly flagellated gram-negative genus Mycoplana, tentatively suggested by Waksman (1940), is undesirable. Endospores are never formed in the Actinomycetales; the characteristic reproductive structures in this order are conidia, formed by fragmentation of the aerial hyphae.
The photosynthetic bacteria were first rationally treated in the system of Pringsheim (1923), who recognized their differences from the colorless sulphur bacteria with which they had so long been associated, and created for them the order Rhodobacteriale8. Kluyver and van Niel placed them in the Eubacteriales, an action which was entirely justified from the strictly morphological standpoint. These organisms are morphologically indistinguishable from the true bacteria, falling into the Pseudomonas-Vibrio- SpiriUum and Micrococcus-Sarcina lines. All species are gramnegative and non-spore-forming. Nothing is known about the distribution of chromatin material. Chromatophores are absent, the photosynthetic pigments being evenly distributed throughout the cell, as in the blue-green algae. Physiologically, these organisms differ from green plants in a number of respects. Bacteriochlorophyll is chemically slightly different from chlorophylls a and b. Photosynthesis is accomplished only in the presence of reducing substances, and never accompanied by oxygen production.
The photosynthetic bacteria form a homogeneous group whose photosynthetic metabolism sets them off from the Eubacteriales. For this reason, it seems wise to keep them as a separate order, recognizing nevertheless their close relationship to groups in the true bacteria.
The three assemblages discussed so far-Eubacteriales, Actinomycetales and Rhodobacteriales-are a well-knit, closely related natural group whose relationships to other bacteria and to nonbacterial microorganisms are not very clear.
A number of workers (Drechsler, 1919; Vuillemin, 1912, 1925) have postulated a relationship between the Actinomycetales and the Eumycetae, but beyond the superficial resemblances in the mycelial nature of growth and the formation of conidia (oidia) by hyphal fragmentation there is little support for this hypothesis. True nuclei do not occur in the actinomycetes, which, as mentioned previously, show a typical eubacterial arrangement of the chromatin. Furthermore, the width of the individual hyphae is always of bacterial dimensions, never approaching that of the true fungi. Negative evidence is the complete absence of sexual reproduction in the actinomycetes. The superficial similarities between molds and actinomycetes are probably to be regarded as an example of convergence. The only relationship which has been seriously suggested for the Eubacteriales is one with the Myxophyta. Although the close relationship existing between bacteria and blue-green algae was stressed by 19th century microbiologists such as Cohn, van Tieghem, and Hansgirg (the two former workers treated them as one group), the importance of this concept has been appreciated less in recent times.
The common features of true bacteria and blue-green algae may be summarized as follows:
1. Absence of true nuclei.
2. Absence of sexual reproduction.
3. Absence of plastids.
One major difference, however, is the complete absence of flagellar motility in the Myxophyta, whose representatives are either immotile or exhibit creeping motility.
It is among the Chroococcales, the most primitive assemblage of the Myxophyta, that we find forms closely resembling the Eubacteriales. A Chroococcus sp., for example, would be indistinguishable from a Micrococcus sp. if it were to lose its photosynthetic pigments. The genus Eucapsis would be similarly indistinguishable from the genus Sarcina. Thus it seems at least possible that the primitive blue-green algae of the Chroococcus type have developed from the Eubacteriales as a second photosynthetic line, at first paralleling morphologically the purple bacteria, but undergoing in the course of time a far more complex morphological evolution which resulted in the development of the two higher orders, the Hormogonales and the Chamaesi- -phonales. If this were the case we must assume that the most primitive blue-green algae were non-motile, being derived from a non-motile group in the true bacteria somewhere close to the primitive coccus type, and that the very characteristic creeping motility of the Myxophyta developed at some later time during their evolution, probably in one of the branches of the Chroococcales. It should be realized that this is a speculative digression, which does not affect the systematic proposals we shall put forward. It is at least certain that morphologically the Myxophyta resemble the true bacteria far more closely than they do any of the other algal groups.
B. The Myxobacteria
The second major natural group of the Schizomycetes which we can recognize is that of the myxobacteria. These are forms which have been but little studied by bacteriologists, and our knowledge of them is due almost entirely to the work of such cryptogamic botanists as Thaxter and Jahn.
The vegetative myxobacterial cell differs fundamentally from that of the true bacterium. In the first place, it does not have a rigid cell wall (Thaxter, 1892; Jahn, 1924; Krzemieniewski, 1928); instead there appears to be only a delicate plasma membrane which gives the cell a power of flexibility never found in the true bacteria. Secondly, flagellar motility is completely absent in the myxobacteria, which exhibit instead a creeping motion very similar to that found in blue-green algae. All species studied so far are motile. Division is always by transverse fission. In their staining reactions the myxobacteria differ markedly from the true bacteria; with ordinary aniline dyes the cells are for the most part rather faintly stained, but certain granules within the cells become intensely colored. These granules have been interpreted (Badian 1930) as consisting of chromatin material which acts in a manner similar to that of chromosomes; more work is needed, however, to substantiate these claims. The vegetative cells are always rod-shaped, two types being found: cells of even width with blunt ends, and pointed spindle-shaped cells.
The life cycle of the typical myxobacterium is divided into a swarm stage in which the vegetative cells creep across the substrate in loose masses held together by a slimy material, and a fruiting stage in which the vegetative cells aggregate and pile up at certain points on the substrate to form fruiting bodies which often are of considerable structural complexity. On or in the fruiting bodies resting cells are borne. In the family Myxococcaceae these are spherical, highly refractive cells, each of which is formed by the rounding up and shortening of a single vegetative cell. In the remaining families the resting cells are merely shortened vegetative cells.'
Under favorable conditions the resting cells germinate and give rise anew to vegetative cells.
The supposed difficulties of obtaining pure cultures have caused the study of the myxobacteria to be confined largely to the fruiting stage. Hence myxobacteria in which the fruiting stage is absent have not been recognized until recent years. These forms, the cytophagas, possess typically myxobacterial vegetative cells (Stanier 1940); some exist only in the vegetative stage, whereas others form microcysts in a manner similar to that of the Myxococcaceae, with the sole difference that organized fruiting bodies are not produced (Krzemieniewska 1930, 1933; Imsenecki and Solntzeva, 1936).
In the myxobacteria we once again find a well-defined, closeknit group whose relationships are doubtful. It is at least certain that they are only remotely, if at all, related to the Eubacterialem- Actinomycetales-Rhodobacteriales assemblage. Some workers (Thaxter 1904; Vahle 1909) have suggested a relationship with the Myxomycetae, particularly with the Acrasieae, in which group there also occurs a swarm stage (pseudoplasmodium) consisting of individual vegetative cells. However, as in the case of the putative Actinomycetales-Eumycetae relationship, this is based largely on superficial resemblances in the manner of growth. All the Acrasiae have definitely amoeboid vegetative cells with true nuclei.
It seems more likely (Jahn 1924, Stanier 1940) that the myxobacteria have developed through forms such as the cytophagas from rod-shaped, motile types among the lower blue-green algae (Chroococcales) by loss of chlorophyll and adaptation to a saprophytic existence. If this were the case, creeping motility would have developed at a comparatively early stage in the evolution of the Myxophyta rather than among the Hormogonales. Jahn's (1924) claim to have observed creeping motility in unicellular blue-green algae would, if substantiated, provide strong support for this hypothesis.
C. The Spirochaetes
Our knowledge of the spirochaetes as a general group is still regrettably scanty. They are unicellular, spiral organisms, often considerably elongated, which, like the myxobacteria, possess a degree of flexibility never found in the Eubacteriales. They are motile either by means of an elastic axial filament or a modified fibrillar membrane. Division is always by means of transverse fission.
Their relationships are obscure. It has at times been suggested that they are related to the protozoa; however, the absence of true nuclei argues against this. On the other hand, their peculiar and specialized method of locomotion separates them sharply both from the true bacteria and from the myxobacteria. At present it seems impossible to arrive even at a tentative hypothesis as to their phylogeny. This is a group on which both morphological and physiological studies are greatly needed.
D. Other organisms commonly placed among the bacteria
Although the spirochaetes are the last of the clearly recognizable major groups of bacteria, there are still a number of organisms not falling into any of the three groups so far discussed which have in the past been included in the Schizomycetes. These are chiefly the forms placed by Bergey et al. in the orders Chlamydobacteriales and Caulobacteriales, with the addition of the nonphotosynthetic members of the Thiobacteriales.
The colorless representatives of the Thiobacteriales have been divided into two morphologically quite distinct families, the Beggiatoaceae, and the Achromatiaceae. The genera Beggiatoa, Thiothrix, and Thioploca, comprising the former of these families, possess a number of morphological characters (shape, size, method of locomotion, sheath formation) which make it possible to consider them as Hormogonales which have lost their pigment system. They represent the colorless counterparts of the bluegreen genera Oscillatoria, Phormidium, and Schizothrix. This has long been recognized by botanists and microbiologists, such as Winogradsky (1888) and Pringsheim (1923).
The family Achromatiaceae contains organisms which so farhave been studied exclusively from field collections. The genus Thiospira, whose representatives are morphologically typical spirilla, is so different from the other genera that there is every reason for its removal from the family and inclusion in the Eubacteriales in the family Pseudomonadaceae. The remaining genera are all very similar to one another morphologically, and it would seem wise to consider them as members of a single genus (Achromatium) pending more careful studies under controlled conditions. The relationships of these organisms are at present obscure. Their size and internal structure makes it unlikelythat they are related to any of the three major groups of bacteria. Judging from the most reliable reports, (e.g., Nadson 1914) their method of locomotion appears to resemble that of the bluegreen algae.2 In this connection it is tempting to compare them with such members of the Myxophyta as Synechococcus aeruginosus, to which they might well be related in the same manner as Beggiatoa is to Oscillatoria.
The order Caulobacteriales, created by Henrici (1935) and accepted by Bergey et al. in the most recent edition of the Manual (1939), comprises a heterogeneous collection of organisms whose sole unifying character is the possession of a stalk. The provisional nature of the order is illustrated by the following quotation from Henrici (1935):
It may be questioned whether the grouping of all stalked bacteria into a separate order is a "natural" arrangement, i.e., whether these organisms are actually phylogenetically related. Concerning this, we do not have enough information to venture an opinion, but the same is equally true of the other orders of bacteria as now defined (our italics). The genera Nevskia, Gallionella, and Caulobacter appear to comprise organisms which are morphologically members of the Eubacteriales on the basis of our definition of that order. The species Nevskia pediculata is probably an unstalked lactic acid bacterium identical with Betabacterium vermiforme (Ward) Mayer.3
Nevskia ramosa, on the other hand, as shown by the illustrations of both Famintzin and Henrici (1935), possesses true stalks which are characteristically dichotomously branched in a manner similar to those of Gallionella. Obviously pure culture studies on these organisms are needed, and the relationships are as yet problematical. Nevertheless, as far as one can judge from the present descriptions, Nevskia ramosa and the Gallionella species seem to have morphological characters in common which are not encountered in the other genera of Henrici's order Caulobacteriales. The occurrence of longitudinal fission and budding in the members of the family Pasteuriaceae shows that these organisms at least differ markedly from the true bacteria. Even after the extensive morphological studies of Kingma Boltjes (1936) on pure cultures of one member of the family, Hyphomicrobium vulgare, it is impossible to arrive at any well-founded conclusion concerning the relationships of the group.
Of the organisms included in the order Chlamydobacteriales, Clonothrix fusca is probably a colorless blue-green alga whose counterpart might be found in the myxophytal genus Rivularia. The relationships of Crenothrix polyspora and of the genera Leptothrix and Sphaerotilus are not clear. The morphological characteristics of Crenothrix are so outstandingly different from those of the other two genera that a single family containing all of them seems indefensible.