**4. Conclusions**

The immense phylogenetic diversity of photosynthetic prokaryotes was demonstrated by the wide systematic range of these bacteria. Bacteria considered in this communication were cultured representatives from six phyla (the cyanobacteria were not considered here) with 15 orders, 27 families, and 90 genera. The most ancient representatives of the phototrophic bacteria, the first that made bacteriochlorophyll (chlorophyllide reductase, BchXYZ) and performed photosynthesis were the phototrophic green bacteria, in particular, those with a type-I photosystem (*Chlorobi*, *Heliobacterium*, *Chloracidobacterium*) (Figures 1 and 2). Among those with a type-II photosystem, the *Chloroflexi* have by far the most ancient roots [3], and *Proteobacteria*, together with their photosystem, diversified to the present-day forms much later (Figures 1 and 3). There was an apparent large gap in the evolution of photosynthesis in the phototrophic green bacteria and in the *Proteobacteria*. This is an indication for the loss of early stages of the photosystem present in the Proteobacteria (Figures 1 and 2).

Today phototrophic *Proteobacteria* are by far the most diverse and the most abundant in the environment and have to be considered the most successful to adapt to the largely oxic environment. If we consider that the basic divergences within the *Rhodobacterales* (e.g., the separation of *Rhodobacter* and *Roseobacter*) have occurred approx. 1 billion years ago [40] and that the first photosynthetic prokaryotes have evolved approx. 3.2–3.5 billion years ago, it is reasonable to conclude from the phylogeny of photosynthesis that phototrophic *Proteobacteria* appeared around 2–2.5 billion years ago. If we use these rough estimates as a guide for the interpretation of the phylogenetic relations of photosynthesis, we can conclude that the ancestors of the green bacteria dominated the field over approx. a billion years and quite likely ancestors of the strictly anaerobic *Chlorobi* played a prominent role in the sulfur oxidation during this time. The *Chlorobi* maintained their strict phototrophic and also an anaerobic way of life up to today and consequently are pushed back to the few anoxic/sulfidic ecological niches that receive light. In the early ages also, the photosystem type-II originated and presumably soon separated into a system represented by our present-day *Chloroflexi* and a system that developed later within the phototrophic *Proteobacteria*. If we assume a common origin of the photosystem type-II in *Chloroflexi* and *Proteobacteria*, the system, as we know it from the *Proteobacteria*, is an advanced stage of a parallel development that diversified together with these bacteria much later. Ancient forms that could represent a link between the two type-II photosystems, apparently, were extinct or survivors have not ye<sup>t</sup> been detected.

The most ancient roots of photosynthesis among *Proteobacteria* are found in the *Alphaproteobacteria* (excluding *Rhodobacterales*) and *Betaproteobacteria* with often unsupported deep divergences and long lines to the present-day representatives, the species/strains studied. Photosynthesis in *Gammaproteobacteria* diversified significantly later with the origin of the *Ectothiorhodospira* group, predating that of the others (*Halorhodospira*, *Chromatiaceae*, *Cellvibrionales*). As the photosynthesis phylogeny in general terms was congruen<sup>t</sup> with the RNA phylogeny (Figures 1–4), it was concluded that this type-II photosystem diversified together with the *Proteobacteria*.

Compared to the phylogenetic depth and the systematic width found in the radiation of the *Alphaproteobacteria*, relatively few genera are known of these bacteria, with the exception of *Rhodobacterales*. The *Rhodobacterales,* on the other hand, represent the most recently diverged group. These bacteria apparently are the most successful to live in our mostly oxic world today, are most versatile in their metabolism, are well adapted to live in the oxic environment, and represent one of the largest orders of phototrophic bacteria living today. A second large gap in the evolution of photosynthesis is, in fact, seen between the *Rhodobacterales* and all other *Proteobacteria* (Figure 3). Another large group, which is clearly separated from the others, but diversified earlier than the *Rhodobacterales* is represented by the *Chromatiales*. These bacteria characteristically are adapted to the borderline between the anoxic/sulfidic and the oxic environment, have found ecological niches over billions of years and survived successfully until today.

The phototrophic bacteria included in this investigation were representatives of most of those that are known and in laboratory culture today. Therefore, the presented data gave a comprehensive basis of the phylogeny of anoxygenic photosynthesis, although the view was limited due to the fact that all those that have escaped cultivation attempts or for other reasons have not been cultured could not be considered. As genetic studies with communities of phototrophic purple bacteria from marine coastal sediments based on the PufLM metagenomic diversity demonstrated that many of those present (or close relatives thereof) already had been cultivated [7], it was concluded that a grea<sup>t</sup> part of those out in nature have already been identified, at least from coastal marine sediments. Nevertheless, we will certainly continue finding new species and phylogenetic lines of phototrophic bacteria, in particular when unstudied or poorly studied locations and environments are investigated. Comprehensive metagenome studies on a grea<sup>t</sup> number of environmental communities might even detect missing links of photosynthesis evolution.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2076-2607/7/11/576/s1.

**Author Contributions:** Cultivation of bacterial cultures, DNA extraction, and purification was performed by T.R.; genomic sequencing and quality assurance by S.K.; sequence assembly, retrieval from databases, phylogenetic calculations, tree construction, and visualization by S.C.N. Sequence annotation, retrieval of sequences from annotated genomes and databases, design of the study, as well as the writing of the manuscript was made by J.F.I. All authors contributed to and approved the work for publication.

**Conflicts of Interest:** The authors declare no conflict of interest.
