Bridging the Gulf: Phytophthora and Downy Mildews Are Connected by Rare Grass Parasites

Downy mildews and root and foliar rots caused by Phytophthora are among the most destructive plant pathogens and therefore have attracted considerable attention during the past two decades. Although it has been realized that a close phylogenetic relationship exists, so far sharp distinction has been made between the obligate biotrophic downy mildews and the hemibiotrophic Phytophthora. In the study presented here, it is shown that a continuum of character states from hemibiotrophic Phytophthora species to obligate biotrophic downy mildews is present. Intermediate character states between downy mildews and Phytophthora species exist in several rare parasites of grasses, which are not embedded within the major clades of the downy mildews but are placed sister to these, with unresolved affinities to both these clades and to Phytophthora. They still have retained traits hitherto thought to be exclusive for Phytophthora. A careful review of previous research is presented and it is highlighted that uniquely for downy mildews, Poakatesthia may form an intracellular mycelium, growing through several host cells. In addition, scanning electron microscopy reveals that sporangiophore growth is not determinate in Viennotia and that outgrowth from sporangiophores is very similar to Phytophthora infestans. It is concluded that the sharp morphological distinction between downy mildews and Phytophthora species (that are often placed in separate families and even different orders), is rather artificial, since all features thought to be exclusive to Phytophthora or the downy mildews are united in the rare grass-parasitizing down mildew genera Viennotia and Poakatesthia and the enigmatic genus Sclerophthora. Therefore, several paradigms regarding the distinction between Phytophthora and the downy mildews need to be reconsidered.


Introduction
Phytophthora species are among the most destructive rot-causing pathogens of plants, responsible for several catastrophic events, like the sudden oak death in North America, caused by Ph. ramorum [1] and the Irish Potato Famine caused by Ph. infestans [2,3]. These pathogens can be grown on synthetic media and are hemibiotrophic in nature. The downy mildews can cause severe damage in several important crops, such grape, cucurbits, sunflower, spinach, lettuce, sorghum, millet, tobacco and spinach. Downy mildews are generally believed to be obligate biotrophic and fully dependent on living host cells -although the reasons why are still obscure -so they can not be grown on media. Downy mildews are mainly host genus or even host species specific, while Phytophthora species often have a much wider host range. In addition to these characteristics, there are three main morphological characteristics that are thought to distinguish Phytophthora species from downy mildews. First, downy mildews do not form intracellular mycelium, but invade host cells only by haustoria [4], while Phytophthora is able to grow through both living and dead cells [5]. Second and equally important, sporangiophore growth is terminated in downy mildews, i.e. once sporangia are formed no further growth takes place, while in Phytophthora the sporangiophore may grow further after sporangia have formed. Third, it is generally believed that in downy mildews all sporangia ripen simultaneously, while in Phytophthora sequential maturation takes place. These differences were thought to be of major importance and have been used as an argument to postulate a deep divide between downy mildews and Phytophthora, placing the downy mildews in a family of its own, the Peronosporaceae, while Phytophthora -on the basis of its thallus growth and general morphological characteristics -has usually been placed in the family Pythiaceae of the order Pythiales [6,7]. Recent molecular phylogenetic investigations [8][9][10][11][12], however reveal that downy mildews are very closely related to a paraphyletic Phytophthora. Only in one study [10], was there significant support for a monophyly of the downy mildews, but this result could be attributed to an artefact due to a too narrow search radius, especially in Maximum Likelihood [13] phylogenetic reconstruction [14]. However, in all molecular phylogenies computed so far, the clade containing both downy mildews and Phytophthora was found to be monophyletic, often with maximum support [15,10]. This appears at odds with the developmental and morphological differences between these two groups and the fact that Phytophthora infected plants usually rot and die, whilst those infected with downy mildews may remain almost or completely asymptomatic for long periods and may even recover, particularly in secondary infections of natural populations. However, the effectors secreted by these pathogens are similar [16][17][18][19]. The complex interactions involved in pathogenesis [20][21][22] differ mostly in respect of the final necrotic phases of infection in Phytophthora which is initiated by necrosis inducing proteins [23]. Although necrosis can also be observed in later stages of infection in some downy mildews, such as Bremia, only seldom do plants die from infections with this pathogen. Bremia is not culturable on artificial media and therefore it is unlikely that necrotic parts of infected plants can serve as a source of nutrition for the hyphae. In general, the downy mildews have evolved into well adapted obligate biotrophic pathogens, which may be transmitted with the seeds of their host plants and many hardly cause observable symptoms. This is especially the case in Basidiophora and some species of Hyaloperonospora, which systemically infect their hosts and in effect develop endosymbiotically [24].
Because of the close phylogenetic relationship revealed by molecular studies and the similarity in the effector genes between the hemibiotrophic Phytophthora and the obligate biotrophic downy mildews, the question then whether some missing links between these two groups still exist. Recent molecular phylogenetic studies [10][11][12]25], revealed that the graminicolous downy mildews appear to be sister clades to the monophyletic major clades of the downy mildews, which contain around 95 % of all species so far described (i.e. brassicolous downy mildews -Hyaloperonospora, Perofascia; downy mildews with coloured conidia -Peronospora, Pseudoperonospora; downy mildews with pyriform haustoria -Basidiophora, Benua, Bremia, Novotelnova, Paraperonospora, Plasmopara, Plasmoverna, Protobremia). However, in none of the analyses to date did these apparent sister relationships receive high bootstrap support and therefore this molecular evidence must be treated cautiously. If the before mentioned artefact in phylogenetic reconstruction [10,14] is taken into account, then the affinities of the graminicolous species to the other clades of the downy mildews or to Phytophthora remains unclear. Göker et al. [10] concluded that the evolution of the brassicolous downy mildew genera Hyaloperonospora and Perofascia had started from hosts in the Poaceae and Thines et al. [11] also noted that the possibility that downy mildew evolution in general might have started in the Poaceae cannot excluded. The basic question arising from this molecular work, addressed in this study, is whether there are any characteristics in the graminicolous downy mildews that are reflecting the close relationship of downy mildews and Phytophthora and the rather basal position of the graminicolous downy mildews that was revealed by recent molecular phylogenetic investigations.

Results and Discussion
In light microscopy (LM), the apophyses, which are typical for Phytophthora, can also be observed in Viennotia (Fig. 1). In scanning electron microscopy (SEM), the similarity of the apophyses in Phytophthora and Viennotia becomes even more apparent (Fig. 1). In both species, not only one, but up to three apophyses on an individual ultimate branchlet can be observed (Fig. 2). These apophyses are formed, when, after a sporangium is dispersed, the ultimate branchlet of the sporangiophore grows out again. The process of producing a sporangium, dispersal and outgrowth may repeat several times. In both Phytophthora and Viennotia, sporangiophores with three apophyses have been observed, which means that four times, a sporangium was produced on a single ultimate branchlet. The process of outgrowth and apophyse formation in Viennotia is illustrated in Figure 2. It should be noted that in contrast to Phytophthora, outgrowth takes place almost simultaneously in Viennotia. In both species, however, the growth of the sporangiophore is not terminated by the formation of sporangia.
The possibility for Phytophthora-like outgrowth in the genus Viennotia is unique for downy mildews. Outgrowth in Viennotia and Phytophthora takes place trough the middle of the scar left after sporangium dispersal and not as previously depicted by de Bary [26], and subsequently copied by many authors, from a swelling below the pedicel, on which the first formed sporangium may still sit [27,28]. Scars on the apophyses that should be present if development was the result of the inflation of branchlets below the original tip could not be observed, although 500 terminal branchlets were scrutinized for each genus. Also in Plasmopara halstedii, abnormal sporulation with outgrowth from a sporangiophore bearing sporangia has been observed on sunflower roots by Novotelnova [29]. Determinate growth of sporangiophores in downy mildews versus indeterminate sporangiophore growth in Phytophthora has been considered the most important characteristic in delimitating downy mildews. Therefore the finding that in Viennotia sporangiophore growth is indeterminate provides important evidence that sharp distinction made between Phytophthora and downy mildews based on morphogenetic criteria is rather artificial.
It is generally assumed that in downy mildews hyphal growth only takes place extracellularly and only the determinate haustoria intrude into host cells, whilst in Phytophthora mycelium may grow through both living and dying cells. Intracellular mycelium is not uncommon in Phytophthora species, (e.g. Phytophthora cinnamomi; 5), although many species of this genus also produces digit-like haustoria. In Poakatesthia penniseti, a rare downy mildew parasite of Pennisetum glaucum, it has been observed that intracellular, callosecovered, hyphae form from haustoria, which may grow through several host cells before they enter the apoplast again [11]. Interestingly, Tippett et al. [5] have noted constrictions at sites where the intracellular mycelium passes from one cell into another which is similar to the observations made for Poakatesthia [11]. Intracellular mycelium apart from haustoria has so far not been reported from any other downy mildew. However, Sclerophthora, some basal Pseudoperonospora and Peronospora species, as well Peronosclerospora, which general mycelium morphology is much alike Poakatesthia [30], have not been investigated thoroughly in this respect. The development of an intracellular mycelium from haustoria in Poakatesthia penniseti raises questions regarding the obligate biotrophy of this downy mildew. Unfortunately, as only the type collection is available as a herbarium specimen, this conclusion can not be supported by cultivation experiments.
The finding of Tokura [31] that Sclerophthora is able to grow on artificial media is not unexpected, as there is a long-standing and ongoing debate [32][33][34][35][36][37][38], as to whether this pathogen should actually be considered a member of the genus Phytophthora. In a recent phylogenetic reconstructions, the placement of this genus could not be unambiguously resolved [12]; though it could be shown unequivocally that it belongs to the Peronosporales and not to the Saprolegniomycetidae, as placed by Dick [7]. That the placement of Sclerophthora is controversial becomes already apparent from its name: ''Sclero'' is referring to the thick-walled oospores, which are similar to those of the graminicolous downy mildew Sclerospora [28,32,34], while ''phthora'' refers to the Phytophthora-like vegetative mycelium including sporangiophores. The main reasons for the inclusion of Sclerophthora in the downy mildews have been the morphology of the oospores and the determinate sporangiophore growth. However, like in Phytophthora, sporangia do not form synchronously in this genus [39]. The main argument for inclusion of Sclerophthora in Phytophthora has been the similarity in the shape of sporangia und the supporting sporangiophores [35]. However, as the shape of sporangiophores and sporangia can be highly misleading characters, as has become apparent from recent molecular phylogenetic investigations in downy mildews [11,15,25], the exclusion of the genus from Phytophthora on grounds of oospore morphology might still be justified. Also the apparently fast and rather long independent evolutionary history, which can be deduced from the long branch for Sclerophthora in Thines et al. [12], contributes some evidence for this conclusion. Certainly, the rather isolated position of this genus on the phylogenetic trees, suggests it should not be considered to be a member of the genus Phytophthora.
The second downy mildew genus which has been reported to be culturable on artificial media is Sclerospora [40], which has always been accepted as a distinct genus. Recent phylogenetic reconstructions clearly place Sclerospora amongst the downy mildews [14,15] and not with the Saprolegniomycetidae as proposed by Dick [7]. The report of cultivation by Tiwari & Arya [40] was supported by a photograph showing sporangiophores from which new sporangiophores were repeatedly arising. However, the successful cultivation of neither Sclerophthora nor Sclerospora has been reported again and so it is possible that the observations made by Tiwari & Arya [40] and Tokura [31] were fortuitous and not readily repeatable. Nevertheless, these results support the obvious assumption that obligate biotrophy has gradually evolved in the basal downy mildews until complete dependence on living host cells was established. Likewise, the synchronous development of sporangia seems to have gradually evolved, as in Sclerophthora the ripening of sporangia is asynchronous [39], whereas in Viennotia, sporangiogenesis is synchronous.
Therefore, in addition to the LM and SEM observations reported here, a careful examination of the literature also reveals that several characters states are in common for some genera of the graminicolous downy mildews with lasting sporangiophores (Graminivora, Poakatesthia, Viennotia), Sclerophthora and Phytophthora. Taking into account the comparison given in Table 1, it becomes apparent that not a single trait that was thought to separate the downy mildews from Phytophthora is always present in downy mildews and that most characteristics thought to be exclusive to Phytophthora can also be observed in lineages of the graminicolous downy mildews.

Conclusion
The three main paradigms about downy mildew distinction (obligate biotrophy, absence of intracellular mycelium, determinate sporangiophore growth) need to be rethought, as there are downy mildews species which show traits that are at variance with these basic assumptions.
The arguments listed above necessitate a critical revision of the paradigm that downy mildews are much different from Phytophthora. Phytophthora and downy mildews are both phylogenetically and morphologically connected by ''bridging taxa'' and share an intimate relationship. The downy mildews are therefore to be considered a more specialised sister-taxon to some advanced Phytophthora lineages, as they have evolved from a Phytophthora-like ancestor at a time, when some other basal lineages of Phytophthora s.l. had already parted from the downy mildew-Phytophthora s.str. lineage (which also includes Phytophthora infestans). Whether the ''bridging taxa'' of the graminicolous downy mildews take an intermediate phylogenetic position between Phytophthora and all other downy mildew clades or whether the plesiomorphic, Phytophthora-like, character states that can be observed in some downy mildews have been lost several times independently can so far not be resolved.
It seems likely that in the enigmatic genus Sclerophthora and in the rare downy mildew species Poakatesthia penniseti and Viennotia oplismeni, several pathogenicity effector genes and developmental genes intermediate to the model organisms in Phytophthora and the downy mildews could be found. Therefore, the investigation of these organisms might shed light on the shift from hemibiotrophy to obligate biotrophy and thereby help to decipher the most basal chapter in downy mildew evolution.

Oomycete material
Of the extremely rare parasites Poakatesthia penniseti and Viennotia oplismeni only the type collections are available (IMI 137328c and IMI 103944, respectively), which were investigated in cause of this study. The Phytophthora specimen depicted in this study is deposited in the herbarium of the University of Hohenheim (HOH) as HUH 992.

Microscopy
For light microscopy (LM) of sporangia and sporangiophores, small pieces of the samples were transferred to 5% chloral hydrate to restore turgidity. Preparation for scanning electron microscopy was done as previously described [41].