History

Theophrastus was the first to describe powdery mildew in rose crops, around 300 B.C. In 1819 Wallroth designated pathogen as Alphitomorpha pannosa (^{Horst and Cloyd, 2007}), in 1829 he was transferred to the genus Erysiphe and classified as E. pannosa (Wallr.) Fr, the anamorph state as Oidium leucoconium Desm., described in 1851 and they described it in the genus Sphaerotheca relocated (^{Robert et al., 2005}). The fungus continued to identify as S. pannosa (Wallr.: Fr.) Lév., but some authors acknowledged the division of this species in ^{1914 by Woronichine}, in varieties, S. pannosa var. rosae, which infects the rosebush, and S. pannosa var. persicae, which infects pear and almond. In America there was controversy over the identity of the pathogen and taxonomically it was considered a different kind to that of Europe. Some studies showed that Sphaerotheca humuli (DC.) Burrill also infects Rosa spp., correspond to samples of America S. humuli, and Europe associated with S. pannosa (^{Salmon, 1900}; ^{Horst y Cloyd, 2007}). ^{Blumer in 1967} maintained the concept of these two species as causing powdery mildew of rose and identified S. pannosa as Sphaerotheca macularis (Wallr.: Fr.) Lind and was not considered different from S. humuli. In 1961 he compared Coyier fresh and herbal global materials; determined that S. pannosa and S. humuli are not clearly differentiated and powdery mildew of roses in the US S. pannosa var. rosae cause (^{Horst y Cloyd, 2007}). Currently, due to changes in the nomenclature of fungi, the pathogen is known as P. pannosa (Wallr Br.) De Bary (^{Braun and Takamatsu, 2000}; ^{Horst y Cloyd, 2007}).

Taxonomic classification

Podosphaera pannosa is a biotrophic pathogen whose teleomorph is located in the Ascomycetes; is heterotálico according Coyier and ^{Bender (1985)}. The fungus occurs in asexual state in crops and is identified as Oidium leucoconium Desem (^{Lediuk et al., 2010}). This pathogen belongs to the kingdom Fungi, Ascomycota phylum, subphylum Pezizomycotina, Leotiomycetes class, order Erysiphales, Erisiphaceae family, genus pannosa and Podosphaera species (^{NCBI, 2014}).

Disease cycle

The fungus in crops of roses in the hibernates field, primarily as mycelium on buds, occasionally casmotecios form on leaves, petals and stems (particularly around the thorns). In roses grown in a greenhouse the pathogen persists only as mycelium and conidia (^{Agrios, 2005}).

Ascospores or conidia of the fungus are spread by the wind into the tissues of plants, and temperature of 21 °C and relative humidity (RH) of 100 % spores germinate; germination decreases if there is a film of water on the leaves (^{Perera and Wheeler, 1975}; ^{Sivapalan, 1993}). In asexual development conidia germinate at 20 °C with 100 % RH about 2 to 4 h after being deposited on the host tissue; then a short germ tube occurs in one end of conidia and within 6 h an initial appressorium, which hyphae that directly penetrates the cuticle develops and reaches the epidermal cells which form haustoria (^{Agrios, 2005} is formed ). These have a core and are delimited by a membrane at 20 to 24 h is introduced into the cytoplasm of the epidermal cell. The cell is not damaged even though the cytoplasm is pushed and deformed by the haustorio the fungus. The haustorio absorbs soluble substances from the host cell, which are translocated to developing mycelium and conidial chains that form on the surface of leaves (^{Horst and Cloyd 2007}; ^{Whitaker and Hokanson, 2009}).

Hyphae, which formed the mycelium, will develop in order to produce erect and short conidiophores over the tissue; this process begins 48 h after germination of conidia (^{Agrios, 2005}). Initial conidiophores are formed as small swellings of hyphae above nuclei; these conidiophores elongate and nuclei are split, then the septa that separate the hyphae conidiophores are formed. In most cases conidiophores generate one conidia per day, but under optimal conditions of temperature (21 °C) and humidity (97 to 99 %) may form a chain of conidia 72 h after initial infection, but usually it takes 5 to 7 d. Mature conidia released 24 h later and form new colonies that produce conidiophores and conidia that cause new infections (^{Horst and Cloyd 2007}). In the sexual phase, the fungus can produce casmotecios with four to eight ascospores able to infect tissue rosebush and start new cycles of disease (^{Whitaker and Honkanson, 2009}).

Epidemiology

Among the varieties of rosebush there are differences in susceptibility to P. pannosa. Creeping roses, climbers and hybrid shrub with large flowers (Tea type) are likely. In contrast to the wichuraianas roses (climbing shrub) have high strength and Floribunda and Polyantha cultivars are more susceptible than hybrid Tea. The state of tissue growth at the time of infection is important because the fungus grows well in the young tissue, its development increases in new growth and decreases in mature (^{Horst and Cloyd 2007}). Also, shaded plantations, compact, with abundant growth of foliage or other factors that reduce airflow and increase moisture promotes the pathogen (^{Linde and Shishkoff, 2003}) is favored. In the development of P. pannosa, temperature and RH they are closely related. The complete cycle of infection occurs at temperatures between 15 and 25 °C and RH of 75 to 79 %, which favor infection structures of fungus, as mycelium, conidiophores, conidia appressoria and haustoria (^{Kashimoto et al., 2003}). ^{Longrée (1939)} reported that minimum, optimum and maximum temperature for the development ofP.pannosaare3to5,21and33°CandRH 97-99 %. Additionally, conidia support, without loss of viability, long periods at 0 °C and humid conditions (^{Price, 1970}). The optimal conidia germination occurs 2 to 4 h after depositing in the tissue if the temperature is 20 to 23 °C and RH 100 %, but temperatures near 30 °C inhibit (^{Xu, 1999}; ^{Horst y Cloyd, 2007}) and the maturation and release happens to 26.7 °C with RH of 40 to 70 % 15.5 °C day and night; HR of 90 to 99 % allows optimal training and conidia germination and infection, and several cycles under these conditions develop an epidemic (^{Horst y Cloyd, 2007}). The disease can also be caused by sexual development of the fungus, which is characteristic of a ascomiceto because it forms a fruiting body called casmotecio (^{Agrios, 2005}), which contains the asci, which in turn contain ascospores that when released and dispersed by the wind begin the process of primary infection disease.

Symptoms and signs

Symptoms of the disease develop rapidly in the aerial tissues, but leaves and buds are the most affected ( ^{and Comic Rankovic, 1997}; ^{Xu, 1999}; ^{Eken, 2005}). Early indications appear on young leaves as slightly elevated areas, often reddish, where signs of the disease will be formed with increased white powder on the underside and the upper leaf (^{Watkins, 1990}). Under favorable conditions, colonization extends throughout the sheet, causing it appear twisted or bent; mature leaves may not present the typical symptoms of the disease, but may have circular and irregular areas covered by the fungus and cause premature abscission (^{Horst and Cloyd, 2007}) or less distortion of mature leaves that eventually become necrotic (^{Agrios, 2005}; ^{Whitaker and Hokanson, 2009}). Mature leaves are resistant to powdery mildew and show no symptoms or only minor local lesions; when the damage is severe, the growth of leaves decreases photosynthetic processes are affected, flower buds and reduce growth (^{Watkins, 1990}).

The fungus can also infect flowers and grow abundantly on the pedicels, sepals and receptacles, especially when the flower bud not opened, so the infection produces flowers of poor quality (^{Horst and Cloyd, 2007}). In some cases they infected necrotic buttons and fall, and if the infection occurs in the flowers, the petals grow incomplete and irregular (^{Whitaker and Hokanson, 2009}). Damage can also occur in tissues succulent young stems, especially at the base of the thorns where powdery colonies are formed; this growth persists even in mature stems (^{Horst y Cloyd, 2007}). On young green stems and appear white spots formed by hyphae that coalesce and come to fully cover the growing apexes; due to infection, these apexes arch or crouch (^{Agrios, 2005}. Sometimes the fungus infects and colonizes reproductive buds before opening, so that the opening is inhibited or altered, the infection progresses to floral whorls , which discolor, atrophy and die.

Anamorph state

The anamorph state of Oidium leucoconium Desm. is characterized by white primary mycelium and secondary dense mycelium which form colonies of woolly white to gray appearance (^{Lediuk et al., 2010}). Their primary hyphae are hyaline with thin, smooth wall (3 to 9 μm wide) (^{Braun, 1987}; ^{Havrylenko, 1995}; ^{Braun and Cook, 2012}). Secondary hyphae are rough, sparsely branched and thick-walled (4.5 to 8 μm wide) (^{Braun and Cook, 2012}). In Mexico, the mycelium presented an average diameter of 4.7 to 6 μm (^{Felix-Gastélum et al., 2014}). Hyphal appressoria are almost indistinct, as bumps. Conidiophores emerge from the surface of the hyphal stem cells, may or not be found in the central part of this, they are erect and can measure up to 210 μm long with straight basal cell, sub-columnar 80x7.5 40 to 12 μm, followed by one or two short cells (^{Braun, 1987}; ^{Braun and Cook, 2012}). ^{Havrylenko (1995)} mentions that this conidiophores are of the Euoidium type, straight, with fibrosin bodies and basal cylindrical cell of between 50 to 80x10 to 12 μm, followed by 5 to 7 short cells. These features resemble those described by ^{Lediuk et al. (2010)}, but with a cylindrical basal cell size of 40 to 55x7 to 10 μm. ^{Felix-Gastélum et al. (2014)} also observed conidiophores of the Euoidium type, but recorded two to five ararey six subsequent cells. Conidia produce conidiophores chains, ellipsoidal-ovoid to doliform, of 20 to 33x10 to 19 μm (length/width) with germtubes more or less terminal to latteral, short or long, slim and simple of between 4 to 5 μm wide (^{Braun, 1987}; ^{Braun and Cook, 2012}). However, some authors disagree on the shape and dimensions of the conidia (^{Homma, 1937}; ^{Havrylenko, 1995}. ^{Comic Rankovic and 1997}; ^{Lediuk et al, 2010};. ^{Felix-Gastélum et al, 2014}.). This diversity may be due to factors such as humidity, host, age of the leaves and season (^{Homma, 1937}; ^{Salmon, 1900}) and the solution in which the samples are studied.

Teleomorph state

In winter, the fungus forms casmotecia (^{Agrios, 2005}) immersed in the mycelial layers, more or less gregarious of 70 to 115 μm in diameter, rarely larger, irregularly polygonal ascomatal rounded cells of 8 to 25 μm width (^{Braun, 1987}; ^{Braun and Cook, 2012}). The casmotecia diameter is 80 to 120 μm width and peridial cells width between 8 and 15 μm (^{Homma, 1937}; ^{Havrylenko, 1995}). In Korea, the casmotecia diameters are between 75 to 100 μm (^{Lee et al., 2011}) and ascomatal cells 10 to 20 μm wide (^{Shin, 1999}). In Yugoslavia, casmotecia showed diameters of 70 to 99 μm (^{Comic Rankovic and 1997}) with miceloid appendixes at the bottom, usually few in number, as mycelium, simple, often characteristically sinuous, contorted and intertwined with each other and with mycelium, sometimes short (shorter than the casmotecia diameter), sometimes rudimentary, with long appendages rarely exceeding 0.5 to 2 or 3 times the diameter of the casmotecia (^{Braun and Cook, 2012}). ^{Shin (1999)} indicated that the miceloid appendages measured between 4 to 6 μm wide, of variable length, usually 0.3 to 3 times the casmotecio diameter with 1 to 3 or 4 septa, basal septum of 10 to 20 μm distance from the base, smooth, moderate thick-walled, change to thin wall at the apex, light brown at the base that progress to becomes hyaline at the top.

Asci are hyaline, ellipsoidal to ovoid (subglobose), 70 to 100x50 to 80 μm (length/ width), sessile, containing four to eight ascospores, ovoid or ellipsoid to doliform of 16 to 28x9 to 20 μm (^{Braun, 1987}; ^{Braun and Cook, 2012}). Some authors disagree on the shape and dimensions of the asci and ascospores (^{Homma, 1937}; ^{Havrylenko, 1995}. ^{Comic Rankovic and 1997}; ^{Shin, 1999}; ^{Lee et al., 2011}).

Molecular studies on Podosphaera pannosa

^{Cook et al. (1997)} did not distinguish the anamorphic genera Sphaerotheca and Podosphaera the observation of the surface by conidia with scanning electron microscopy and claimed that their difference is based solely on the host plant. However, sequencing of the ITS region ribosomal DNA, based on the length of nucleotides, both genders and grouped coincided with the phylogenetic analysis (^{Takamatsu et al., 1998}). ^{Saenz and Taylor (1999)} reported that the combination of morphological and molecular analysis allowed to group genera Sphaerotheca and Podosphaera in one group. These studies support the theory that both genders are congeners and confirmed that all Sphaerotheca species belonging to the genus Podosphaera, so the fungus previously identified as S. pannosa (Wallr.: Ex Fr.) Lév. now called P. pannosa (Wallr.: Fr.) de Bary (fam Erisiphaceae, Cystotheceae tribe.) (^{Braun and Takamatsu., 2000}; ^{Braun et al., 2002}).

^{Alvarez et al. (2001)} found that amplification of the ITS region, the 5.8S rDNA gene with the primers ITS1 and ITS2, and confirmed the identity of 16 samples of P. pannosa with fragments of equal size (295 bp). They also corroborated identity by restriction enzymes (AluI and Hind I), which showed a pattern of bands equal for all samples with each enzyme.

^{Cunnington et al. (2003)} designed the PMITS1 and PMITS2 oligos, with which they reaffirmed the identity of P. pannosa on Rosa sp., Eucalyptus populnea F. Muell. and Eucalyptus sp. with 100 % similarity to both hosts. ^{Jankovics et al. (2011)} amplified the ITS region of rDNA using nested PCR with PMITS1 and PMITS2 oligos for the first reaction and ITS1F and ITS4 oligos for second; this allowed them to confirm the identity of P. pannosa on peach (Prunus persica (L.) Batsch) with identical or similar sequences by 99 % than those reported by ^{Saenz and Taylor (1999)} and ^{Mori et al. (2000)}.

^{Leus et al. (2006)} collected 26 samples of Podosphaera on Rosa sp. and Prunus spp. and by sequence analysis of the ITS region of rDNA, by nested PCR, with ITS1F and ITS4 oligos as the first reaction and oligos ITS5 and ITS4 as second reaction; they identified 24 sequences corresponding to P. pannosa of which one group 18 was identical to those reported by ^{Takamatsu et al. (2000)}. It was also identified P. pannosa in Mexico with a single reaction by the ITS1F and ITS4 initiators Rosa spp. (^{Felix-Gastélum et al., 2014}) and Vinca (Catharanthus roseus (L) G. Don) in United States (^{Romberg et al., 2014}). In France, the ITS1 and ITS4 a nucleotide sequence (Accession No. JN654341) which revealed 100 % identity with P. pannosa on Prunus cerasus (^{Hubert et al., 2012}) was obtained. With ITS5 and P3 oligos has confirmed the identity of P. pannosa (^{Takamatsu et al., 2000}; ²⁰¹⁰). Identity of P. pannosa was corroborated with ITS5 and P3 oligos, to obtain a 477 bp sequence, which was aligned in GenBank, with higher than 98 % similarity with AF011322, AB022348 and AB525937 of P. pannosa reported in Rosa spp. (^{Lee et al., 2011}).

Genetic diversity

Several pathogenic races of P. pannosa exist in both loci rose and both qualitative and quantitative resistance are present in the host (^{Linde and Debener., 2003}; Lei et al., 2006). ^{Mence and Hildebrandt (1966)} tested the resistance and susceptibility to P. pannosa of 17 varieties and six species of rose bush, confirmed the biological specialization of the fungus to note that the conidia on Rosa virginiana Mill. infected quickly Rosa rugosa Thunb. but no colonies developed in most varieties; they suggested the existence of two races, which differ in the host and production of conidia. In contrast, ^{Bender and Coyier (1984)} won nine samples of powdery mildew seven hybrid and two varieties of Rosa sp., Evaluated the virulence and identified five races of Podosphaera pannosa with different adaptation of its original host in Oregon, USA.

^{Linde and Debener (2003)} classified eight races after evaluate ten genotypes monoconidial rose eight samples of P. pannosa and showed that populations of P. pannosa exhibited high diversity of virulence genes. ^{Alvarez et al. (2001)} conducted studies with 16 samples pathogenicity of P. pannosa of rose six and observed varieties that no sporulation in variety Tineke^® with Sp6 sample but caused disease in other varieties, and the remaining samples sporulated in all cultivars. This suggests the presence of at least two pathotypes.

The races are differentiated only in some trials. To examine the possible existence of pathotypes ^{Leus et al. (2002)} obtained samples monosporic Sphaerotheca pannosa var. rosae, from pickings in several locations in Belgium, and evaluated in seven genotypes of rose bush with different levels of resistance. They did not obtain evidence of this phenomenon; but they revealed the existence of pathotypes of P. pannosa by the differential response in virulence when tested in vitro rose. In addition, some samples could also infect Prunus avium L. seedlings (^{Leus et al., 2003}). ^{Leus et al. (2006)} obtained samples of P. pannosa monoconidial Rosa sp. and Prunus sp. They collected in Belgium, Germany, France, Denmark, Israel and the Netherlands, characterized based on the differential reactions in vitro genotypes rosebush and Prunus avium and DNA sequence analysis of ITS rDNA region; They identified 24 samples of P. pannosa with varying degrees of host specificity; a first group of 18 samples was highly virulent on rosebush or no, or very little virulent P. avium, a second group of 4 samples was highly virulent on both hosts and a third group with 2 samples having identical sequences to the group 1, but they not infected the rosebush.

Disease management

The management of powdery mildew of the rose bush in the greenhouse is done mainly with synthetic fungicides (^{Passini et al., 2007}; ^{Scarito et al., 2007}; ^{Shetty et al., 2012}), every 7 to 14 d to protect the immature tissue susceptible (^{Xu, 1999}). Fungicides group demethylation inhibitors and inhibitors of ergosterol biosynthesis (^{Pasini et al., 1997}; ^{Pasini et al., 2007}) show biological effectiveness in controlling powdery mildew caused by P. pannosa; but repeated applications of these fungicides cause phytotoxicity can reduce the length of the stems and selection of resistant populations of the pathogen (^{Pasini et al., 1997}; ^{Daughtrey and Benson, 2005}). The group of strobilurins (inhibores of mitochondrial respiration), such as azoxystrobin, kresoxim-methyl, pyraclostrobin and trifloxystrobin exert efficient control of powdery mildew (Daughtrey and Benson, 2005). In this regard, ^{Tjosvold and Koike (2001)} observed that azoxystrobin (Heritage^®), kresoximmethyl (Cygnus^®) and trifloxystrobin (Compass^®) in 0.07 g dose L^-1, 0.12 g and 0.15 g L^-1 they controlled the disease without causing phytotoxicity or loss of plant vigor.

The development of powdery mildew is adversely affected by the presence of a sheet of water on the surface of the leaves (^{Perera and Wheeler, 1975}; ^{Sivapalan, 1993}). But this method is not recommended in commercial terms, because excess water may favor the development of other diseases such as downy mildew (Peronospora sparsa Berkeley) and black stain rosebush (diplocarpon rosae Wolf.) (^{Gullino and Garibaldi, 1996}; ^{Pasini et al., 1997}).

The film-forming polymers also are used for control; such products probably act as a chemical penetration pathogen or physical barrier and inhibit the development of the disease (^{Hagiladi and Ziv, 1986}). Other products such as baking soda and oils are also effective in controlling the disease (^{Horst et al., 1992}; ^{Pasini et al., 1997}).

Extracts of some plants, such as Azadirachta indica Adr. Juss., Reynoutria sachalinensis (F. Schmidt) Nakai, Macleaya cordata (Willd) R. Br., and Citrus paradisi Macf. significantly reduce infection by P. pannosa (^{Pasini et al., 1997}; ^{Newman et al., 1999}; ^{Wojdyla, 2001}; ^{Toppe et al., 2007}). In addition the anhydrous milkfat (0.7 %) and soybean oil (2 %) are used to control powdery mildew in potted rose plants; both products reduced the severity of disease in 2 to 5 % in relation to chemical control (greater than 40 % severity) and the control where the severity increased to 100 % (^{Ah Chee et al., 2011}). Besides, ^{Seddigh et al. (2014)} showed that the compost tea increased control of the disease as compared to chemical treatment with Penconazole (Topas^®).

Recent management alternatives for disease control, such as using resistance inductors can reduce the severity of damage by a pathogen. Dose of 0.1 to 0.2 mg mL^-1 of acibenzolar-S-methyl (Bion^®) provide effective control of the disease (Smith et al., 2012). Silicon has also been used to control some diseases; in hydroponics reduced the development of powdery mildew in rose and increased performance (^{Voogt and Sonneveld, 2001}). Four genotypes of roses, potted, with different levels of susceptibility to the disease were treated with a nutrient solution of 3.6 mM Si (100 ppm), supplied as metasilicate potassium, they decreased the severity of the disease by up to 48.9% (^{Shetty et al., 2012}). Aspirin in dose of 0.3 g L^-1 reduced the incidence and severity of disease in rose variety Cezane^® Classic (^{Torres et al., 2013}).

Some antagonistic fungi, bacteria and at least one insect (Thrips tabaci Lind.) were identified for the control of powdery mildew in rose bush (^{Bélanger et al., 1994}; ^{Eken, 2005}; ^{Elmhirst et al., 2011}) (^{Coyier, 1983}). Fungi as Ampelomyces quisqualis Ces., Cladosporium oxysporum (Berk. And Curt.), Tilletiopsis sp. and Verticillium lecanni (Zimm.) and other parasite or are antagonists of the rosebush powdery mildew (^{Belanger et al., 1994}; ^{Ng et al., 1997}; ^{Pasini et al., 1997}; ^{Agrios, 2005}). However, the results of large-scale trials with these antagonists have been unsuccessful. In addition, few of these trials have been carried out, because to achieve maximum control high levels of HR is required in the case of A. quisqualis and Tilletiopsis, which lose their effectiveness quickly lower to 90 % RH (^{Philipp et al., 1990}).

Mildew resistance

Some of the new rosebush varieties are resistant to the disease, but only some of them have high levels of resistance, presumably by the development of new strains of P. pannosa that break this resistance (^{Horst and Cloyd, 2007}). Among them, some are resistant in particular geographic areas, yet susceptible to another and even in the same locality, and resistant few years and susceptible others (^{Agrios, 2005}). ^{Linde and Debener (2003)} first showed the action of a single dominant resistance gene (Rpp1) against P. pannosa by repeated inoculations with monoconidial samples. This model based on a single dominant resistance gene was backed by isolation of genes closely linked to Rpp1 molecular markers (^{Linde et al., 2004}). In this regard, ^{Xu et al. (2005)} identified RGA22C, RGA7B RGA4A and markers linked to a resistance gene locus, called CRPM1 to pink brown powdery mildew. ^{Li et al. (2003)}introduced the AceAMP1 gene on the Rosa hybrid cv. Carefree Beauty^®, gene encoding a antimicrobial protein in transgenic plants; this showed increased resistance to P. pannosa using in vitro trials with detached leaves and plants established in greenhouse.

In addition, several QTLs of importance for powdery mildew resistance have been located in linkage maps for both populations of diploids and tetraploid roses (^{Crespel et al., 2002}; ^{Dugo et al., 2005}; ^{Linde et al., 2006}; ^{Hosseini Moghaddam et al., 2012}). Certain factors of monogenic resistance may lead to called boom-bust cycles (^{Thompson and Burdom, 1992}), which make resistance ineffective in a short time span. The quantitative resistance genes alternative to the monogenic resistance genes are hampered by the tetraploid nature of most cultivated roses and so an alternative to conventional resistance is to use enhancement of genes of a single dominant or QTLs or mildew resistance locus (MLO) (^{Kaufmann et al., 2012}).