Homologs of ToxB, a host-selective toxin gene from Pyrenophora tritici-repentis, are present in the genome of sister-species Pyrenophora bromi and other members of the Ascomycota

Abstract

Pyrenophora tritici-repentis requires the production of host-selective toxins (HSTs) to cause the disease tan spot of wheat, including Ptr ToxA, Ptr ToxB, and Ptr ToxC. Pyrenophora bromi, the species most closely related to P. tritici-repentis, is the causal agent of brown leaf spot of bromegrass. Because of the relatedness of P. bromi and P. tritici-repentis, we investigated the possibility that P. bromi contains sequences homologous to ToxA and/or ToxB, the products of which may be involved in its interaction with bromegrass. Multiplex polymerase chain reaction (PCR) revealed the presence of ToxB-like sequences in P. bromi and high-fidelity PCR was used to clone several of these loci, which were subsequently confirmed to be homologous to ToxB. Additionally, Southern analysis revealed ToxB from P. bromi to have a multicopy nature similar to ToxB from P. tritici-repentis. A combination of phylogenetic and Southern analyses revealed that the distribution of ToxB extends further into the Pleosporaceae, and a search of available fungal genomes identified a distant putative homolog in Magnaporthe grisea, causal agent of rice blast. Thus, unlike most described HSTs, ToxB homologs are present across a broad range of plant pathogenic ascomycetes, suggesting that it may have arose in an early ancestor of the Ascomycota.

Introduction

Host-selective toxins (HSTs), a diverse set of molecules known only to be produced by plant pathogenic fungi, function as essential disease determinants in a number of plant host–pathogen interactions (reviewed in Walton, 2000, Markham and Hille, 2001, Wolpert et al., 2002). They are implicated in pathogenesis/virulence because they are toxic only to hosts susceptible to the fungus, and in turn, toxin production by the fungal pathogen is strictly correlated with disease elicitation on a susceptible host. For a number of HSTs, causality in disease has been confirmed molecularly (Panaccione et al., 1992, Ciuffetti et al., 1997, Yoder et al., 1997, Tanaka et al., 1999, Churchill et al., 2001, Johnson et al., 2001). HSTs range from low-molecular weight products of multifunctional enzymes or complex enzymatic pathways encoded for by complex gene clusters to protein products of single genes. In general, the genetic determinants of their production are discontinuously distributed in that they are present only in HST-producing, pathogenic isolates and are completely absent in nonpathogenic counterparts (Ahn and Walton, 1996, Ciuffetti et al., 1997, Yoder et al., 1997, Tanaka et al., 1999). Horizontal gene transfer is often invoked as an explanation for such irregular distribution (reviewed in Rosewich and Kistler, 2000, Walton, 2000).

Pyrenophora tritici-repentis, causal agent of tan spot of wheat, produces both proteinaceous and nonproteinaceous HSTs. The proteinaceous HSTs Ptr ToxA (Synonyms: Ptr necrosis toxin, Ptr toxin, ToxA [Ciuffetti et al., 1998]) (Ballance et al., 1989, Tomas et al., 1990, Tuori et al., 1995), encoded for by the single copy ToxA gene (Ballance et al., 1996, Ciuffetti et al., 1997), and Ptr ToxB (Synonym: Ptr chlorosis toxin [Ciuffetti et al., 1998]) (Strelkov et al., 1999), encoded for by the multicopy ToxB gene (Martinez et al., 2001, Martinez et al., 2004, Strelkov et al., 2006), cause necrotic and chlorotic lesions, respectively, on correspondingly sensitive wheat. Ptr ToxC, which appears to be a nonionic, polar, low-molecular weight HST (Effertz et al., 2002), causes chlorosis on a wheat line distinct from that sensitive to Ptr ToxB (Lamari et al., 1995) and awaits full characterization. Eight races, each with a unique complement of Ptr ToxA, Ptr ToxB, and Ptr ToxC, and a correspondingly unique host range, have been formally described (reviewed in Strelkov and Lamari, 2003). Races 1, 2, 7, and 8 produce Ptr ToxA; races 5–8 produce Ptr ToxB; and races 1, 3, 6, and 8 produce Ptr ToxC. Nonpathogenic race 4 isolates do not produce any of these three HSTs.

In contrast to other HSTs, which are present only in toxin-producing pathogenic isolates, the ToxB gene is unique in that related genes are found in races of P. tritici-repentis that do not produce Ptr ToxB (Martinez et al., 2004, Strelkov et al., 2006, Andrie et al., 2007), nor are pathogenic on Ptr ToxB-sensitive wheat (Lamari et al., 1995). In particular, single copies of distinct ToxB homologs are present in the genomes of P. tritici-repentis races 3 and 4. The ToxB homolog in race 3 shares all but the first six nucleotides with the wild-type ToxB open reading frame (ORF); however, upstream sequences are distinct between these two loci (Strelkov et al., 2006). Matching ToxB homologs, which share 86% identity with ToxB, have been identified in two different nonpathogenic P. tritici-repentis race 4 isolates, isolate SD20 (Martinez et al., 2004), within which the homolog is named toxb, and isolate 90-2 (Strelkov et al., 2006). Although the sequences flanking wild-type ToxB share some homology with the flanking sequence of the race 4 homolog, the majority of the flanking sequences are unique to each locus (Martinez et al., 2004, Strelkov et al., 2006). The presence of distinct sequences upstream of ToxB in race 5 and of the ToxB homologs in races 3 and 4 could account for the observed differences in their transcriptional regulation. Wild-type ToxB in race 5 is transcribed in both mycelia and conidia, whereas the ToxB homologs in races 3 and 4 are only transcribed in conidia. Other non-Ptr ToxB-producing races of P. tritici-repentis do not possess ToxB homologs.

The closest relative to P. tritici-repentis is its sister-species Pyrenophora bromi (Zhang and Berbee, 2001), the causal agent of brown leaf spot of smooth bromegrass (Bromus inermis) (Chamberlain and Allison, 1945). Smooth bromegrass (hereafter referred to as bromegrass) is considered one of the best cool-season pasture grasses and is one of two bromegrass species cultivated for permanent pastures to any extent in North America (Vogel et al., 1996). Throughout the bromegrass-growing regions of the US and Canada, brown leaf spot is one of the most widespread and destructive diseases of this forage crop (Kaufmann et al., 1961, Elliott, 1962, Smith and Knowles, 1973, Berkenkamp, 1974, Vogel et al., 1996). Susceptibility to brown leaf spot is prevalent among commonly grown cultivars of bromegrass (Zeiders and Sherwood, 1986) and, like P. tritici-repentis in its interaction with wheat (Faris et al., 1996, Stock et al., 1996, Gamba et al., 1998, Anderson et al., 1999, Effertz et al., 2002), susceptibility to P. bromi is dominant to resistance (Berg et al., 1983). Additionally, the presence of conspicuous chlorotic halos beyond the boundaries of fungal hyphae suggests the involvement of a phytotoxin in disease (Sherwood, 1996).

Because of the relatedness of P. bromi and P. tritici-repentis, and their grass hosts (Barker et al., 2001, Hsiao et al., 1994), it is possible that P. bromi possesses similar disease mechanisms as P. tritici-repentis. Thus, we initiated this study to determine if sequences homologous to the HST genes of P. tritici-repentis are present in P. bromi. Not only did we reveal the presence of ToxB homologs in P. bromi, Southern analysis mapped on a phylogeny of P. tritici-repentis and its close relatives suggests that the distribution of ToxB-like sequences extends throughout the genus Pyrenophora into other genera of the Pleosporaceae. Additionally, a search of available fungal genomes revealed a putative homolog in a member of another class of the Ascomycota.