Virulence of isolates of Ustilago tritici collected in Manitoba and Saskatchewan, Canada, from 1999 to 2007

Abstract

Breeding for resistance to the loose smut pathogen, Ustilago tritici, in wheat is a traditional disease management approach in western Canada. The development of resistant wheat lines relies on accurate methods for identification of effective host resistance. Knowledge of the virulence frequencies of races in the pathogen population is essential to identify effective host resistance. The objective of this work was to characterize the virulence phenotypes of U. tritici isolates collected from 1999 to 2007 in Manitoba and Saskatchewan. Sixty-four isolates of U. tritici from durum wheat and 59 from hexaploid wheat were assessed for virulence. Teliospore suspensions of each isolate were injected into florets of the wheat differential host series by needle inoculation at mid-anthesis. Mature seed from inoculated florets were grown to determine the per cent infected plants. Eighty-five per cent of durum wheat isolates were race T32, possessing virulence to differential lines ‘Mindum’ (TD-1), and ‘Wakooma’ (TD-19). Virulence to ‘Pentad’ (TD-11) was found in only 5% of the durum wheat isolates. Races T7, T5, T64 and T10 were the most commonly identified races from the hexaploid wheat isolates of U. tritici. Six previously unreported races were identified: races T61, T64, T65, T66, T67 and T68. None of the isolates possessed virulence to ‘Sonop’ (TD-14), while >50% of the isolates possessed virulence to ‘Kota’ (TD-4), ‘Little Club’/‘Reward’ (TD-5A), ‘Reward’ (TD-7) and ‘Thatcher’/‘Regent’/‘Reward’ (TD-12A). Virulence to TD-6, which was previously uncommon from Manitoba and Saskatchewan, was possessed by 20% of the hexaploid wheat isolates. The natural population of U. tritici in Manitoba and Saskatchewan continues to evolve in terms of virulence, but effective sources of resistance are still currently available.

Résumé

La sélection pour la résistance à l’agent pathogène du charbon nu, Ustilago tritici, chez le blé est une approche de gestion traditionnelle des maladies dans l’Ouest canadien. Le développement de lignées de blé résistantes est fondé sur des méthodes précises permettant la détection d’une résistance efficace de l’hôte. La connaissance des fréquences de virulence des races chez les populations d’agents pathogènes est essentielle à l’identification d’une résistance de l’hôte efficace. Le but de ces travaux était de caractériser les phénotypes de virulence d’isolats d’U. tritici collectés de 1999 à 2007 au Manitoba et en Saskatchewan. En tout, 64 isolats d’U. tritici provenant de blé dur et 59, de blé hexaploïde ont été évalués pour leur virulence. Au stade mi-anthèse, des suspensions de téliospores de chaque isolat ont été injectées dans les florules des séries hôtes différentielles de blé par inoculation. Les graines mûres provenant des florules inoculées ont été mises en terre pour déterminer le taux de plants infectés. Quatre-vingt-cinq pour cent des isolats de blé dur étaient de la race T32, affichant de la virulence à l’égard des lignées différentielles ‘Mindum’ (TD-1) et ‘Wakooma’ (TD-19). Seulement 5 % des isolats de blé dur étaient virulents à l’égard de ‘Pentad’ (TD-11). Les races T7, T5, T64 et T10 étaient celles les plus couramment identifiées chez les isolats de blé hexaploïde d’U. tritici. Nous avons identifié six races qui n’avaient jamais été signalées auparavant: les races T61, T64, T65, T66, T67 et T68. Aucun de ces isolats n’était virulent à l’égard de ‘Sonop’ (TD-14), tandis que plus de 50% de ces derniers l’étaient à l’égard de ‘Kota’ (TD-4), ‘Little Club’/‘Reward’ (TD-5A), ‘Reward’ (TD-7) et ‘Thatcher’/‘Regent’/‘Reward’ (TD-12A). Vingt pour cent des isolats de blé hexaploïde ont affiché de la virulence à l’égard de TD-6, ce qui était rare auparavant rare au Manitoba et en Saskatchewan. La population naturelle d’U. tritici au Manitoba et en Saskatchewan continue d’évoluer en ce qui a trait à la virulence, mais des sources efficaces de résistance sont actuellement disponibles.

Keywords:

• loose smut
• races
• Triticum aestivum
• Triticum turgidum var. durum
• Ustilago tritici

Mots clés:

• charbon nu
• races
• Triticum aestivum
• Triticum turgidum var. durum
• Ustilago tritici

Introduction

The commercial use of lines of wheat (Triticum aestivum L., T. turgidum L. var. durum) with resistance to the loose smut pathogen, Ustilago tritici (Pers.) Rostr., is an effective and environmentally friendly disease management approach. Many wheat breeding programmes in western Canada aim to incorporate loose smut resistance genes into new wheat lines, despite the fact that breeding for loose smut resistance is expensive in terms of time and resources. These breeding efforts have led to numerous commercial wheat cultivars with effective resistance to loose smut, such as ‘Glenlea’, which possesses at least four genes for resistance (Knox et al. 2008) and is completely resistant to a broad spectrum of loose smut races collected from Canada and around the world (Nielsen 1983).

The development of disease resistant wheat lines requires identification and utilization of sources of effective resistance. Effective resistance is defined as resistance to most if not all of the virulence phenotypes, or races, in the pathogen population. Identification of the race structure of the pathogen population in the wheat production area for new cultivars is essential for identifying effective resistance. Resistance and virulence in the wheat–U. tritici pathosystem follow the gene-for-gene model (Oort 1963). This host–pathogen relationship allows for the study of the diversity of avirulence or virulence genes within populations of U. tritici by inoculation of different isolates of the pathogen onto wheat lines differing in their resistance, called a differential host set. The most common differential host set used in studies with U. tritici contains some lines originally used by Oort (1944), with modifications over time by Cherewick (1953), Nielsen (1987), Nielsen and Dyck (1988), and Nielsen and Thomas (1996). This differential host series is not a set of isogenic lines which differ in only one resistance gene, but is a collection of lines that have shown differing resistance to different isolates of U. tritici. Menzies et al. (2003) used Person’s geometric rule to show that while the durum wheat lines in this differential host series appear to have single genes for resistance, the hexaploid wheat lines are mostly multigenic. This differential host set has been used to identify races in populations of U. tritici from more than 30 countries (Nielsen 1987).

The natural population of U. tritici in western Canada has been extensively studied in terms of identification of races. Hanna and Popp (1932) were the first to identify the existence of physiological races of U. tritici in North America, and Hanna (1937) identified four physiological races from Manitoba. This work was continued by Cherewick (1953), who identified 10 races, and Nielsen (1987) who identified 18 races from Canadian isolates of U. tritici, as well as numerous races from worldwide collections. Menzies et al. (2003) conducted the largest study of Canadian isolates by conducting virulence assessments on 609 isolates of U. tritici collected between 1964 and 1998. All but four of the isolates were from Canada, with 574 of the isolates collected in Manitoba and Saskatchewan. The results showed a changing population structure, with some races being common at the beginning but not the end of the collection period, while other races, such as race T10, were less frequent in the early years but very common in the later years. Fourteen races previously not found in the Canadian population were identified during their study (Menzies et al. 2003). The current study represents a continuation of the characterization of the virulence dynamics of U. tritici in Canada, particularly Manitoba and Saskatchewan.

The objective of this study was to determine the virulence characteristics of isolates of U. tritici collected in Manitoba and Saskatchewan, from 1999 to 2007. The data were used to determine the presence and frequency of virulence genes and identify the races in the pathogen population.

Materials and methods

Nielsen (1987), Nielsen and Thomas (1996) and Menzies et al. (2003, 2009) provide a thorough review of the races of U. tritici and techniques for their study. The methods used here were those of Nielsen and Thomas (1996), unless otherwise noted.

Ustilago tritici isolate collection and maintenance

Isolates of U. tritici were collected during annual surveys in Manitoba and central and eastern Saskatchewan from 1999 to 2007, as reported in the Canadian Plant Disease Survey. Ten collections from western Saskatchewan were kindly provided by Dr Ron Knox (AAFC, Swift Current, SK) in 1999. In addition, four isolates were obtained in 1999 from smutted plants of breeding lines growing in the field at Glenlea, MB, which had been increased in New Zealand during the winter of 1998/1999. The origin of U. tritici infecting these lines may be New Zealand or Canada, but were considered to be of New Zealand source. One smutted spike was collected per field (or breeding line in the case of the New Zealand isolates) to avoid the possibility of a mixture of genotypes (Nielsen 1987). Each collection was assigned a collection number, allowed to dry at room temperature in a paper envelope for a few days and stored at 5°C until used.

Inoculum preparation

Teliospores from smutted spikes were suspended in tap water by breaking off a small portion of the smutted spike and rubbing it between two fingers under water. The inoculum was visually adjusted to approach a concentration of 1 g of teliospores per L of water by comparison with previously prepared standard suspensions which resembled weak black tea. Normally, 50 mL of suspension was prepared at a time and stored for up to 5 days at 5°C.

Inoculation of the wheat differential set and assessment of infection

Inoculation of wheat spikes was conducted by injection of the teliospore suspension into the florets using a hypodermic needle and syringe. Florets of the wheat hosts were inoculated at mid-anthesis, which is the optimum time for infection (Menzies et al. 1999). Two or three spikes were inoculated per isolate on each wheat differential line (Nielsen 1987; Nielsen & Dyck 1988; Nielsen & Thomas 1996). The differential host set contains lines from two different wheat species; durum wheat (T. turgidum L. var. durum) and hexaploid wheat (Triticum aestivum L.). In general, isolates of U. tritici from durum wheat do not infect hexaploid wheat and isolates from hexaploid wheat do not infect durum wheat (Menzies et al. 2009), with the exception of TD-13, which is the universal suscept for all races of U. tritici on Triticum spp. (Nielsen 1987). Isolates collected from durum wheat were inoculated onto the durum differentials ‘Mindum’ (TD-1), ‘Pentad’ (TD-11), PI 298554/CI 795 (TD-13) and ‘Wakooma’ (TD-19). Isolates collected from hexaploid wheats were inoculated onto the hexaploid differentials ‘Renfrew’ (TD-2), ‘Florence’/‘Aurore’ (TD-3), ‘Kota’ (TD-4), ‘Little Club’/‘Reward’ (TD-5A), PI69282 (TD-6), ‘Reward’ (TD-7), ‘Carma’/‘Reward’ (TD-8A), ‘Kearney’ (TD-9), ‘Red Bobs’ (TD-10), ‘Thatcher’/‘Regent’/'Reward’ (TD-12A), PI 298554/CI 795 (TD-13), ‘Sonop’ (TD-14), H44/‘Marquis’ (TD-15), ‘Marroqui 588' (TD-16), ‘Marquillo’/‘Waratah’ (TD-17) and ‘Manitou*2ʹ/‘Giza 144ʹ (TD-18). The inoculated heads were threshed after reaching maturity. Seed from each spike were planted either in the field or in soil beds in the greenhouse, and the mean percentage of smutted plants was determined at flowering.

Determination of virulence

An isolate of U. tritici was rated as virulent to a differential line if more than 10% of the inoculated seeds produced smutted plants, and avirulent if 10% or fewer plants were smutted. If an infection level of 0–20% was obtained on a differential line, the line was re-inoculated to confirm the per cent infection. The final per cent infection in these cases was the mean of the two (or in some cases three) tests. Normally, the differential lines showed a high level of infection (>50%) or 0% infection. Race nomenclature was according to Nielsen and Thomas (1996) and Menzies et al. (2003).

Results and discussion

One hundred and twenty-three isolates of U. tritici collected between 1999 and 2007 were assessed for their virulence, with 64 isolates from durum wheat (collected between 1999 and 2003) and 59 from hexaploid wheat (collected between 1999 and 2007). Five of the durum wheat isolates were collected in Manitoba and the rest in Saskatchewan. Twenty-five hexaploid wheat isolates were collected from Manitoba, 30 from Saskatchewan, and four from New Zealand. The races identified in this study are presented in Table 1.

Table 1. Races of Ustilago tritici identified from 123 isolates collected in Manitoba and Saskatchewan (including four from a New Zealand source) between 1999 and 2007, and the reactions of the races on the wheat differential host series.

	Wheat differential line (TD-)
Race	1	2	3	4	5A	6	7	8A	9	10	11	12A	13	14	15	16	17	18	19
T1				C	C		C						C
T3	C^1												C
T5							C					C	C
T7				C	C		C						C		C
T10				C	C		C	C	C			C	C					C
T11							C						C
T18				C	C		C	C	C				C
T26													C						C
T28					C		C						C		C
T32	C												C						C
T33	C										C		C						C
T42^2													C
T44				C	C		C					C	C		C			C
T61^3							C					C	C
T62^3				C	C								C		C
T63^3			C	C	C		C	C	C			C	C		C		C	C
T64^3				C	C	C	C	C	C			C	C					C
T65^3			C		C	C	C			C		C	C					C
T66^3			C	C	C	C	C		C	C		C	C			C	C	C
T67^3					C		C	C	C			C	C		C
T68^3		C	C				C			C		C	C

¹The differential designations are based on Nielsen (1987) and Nielsen and Thomas (1996). Races T1–T44 were reported by Nielsen (1987) and Nielsen & Thomas (1996), races T45–T 58 were reported by Menzies et al. (2003) and races T59 and T60 were reported by Kaur et al. (2014). C indicates a compatible reaction.

²Universal avirulent race collected from durum wheat (T. turgidum L. var. durum) in 2001. Previous reports of a similar race profile have been of isolates collected from hexaploid wheat (T. aestivum L.). However, normally races of U. tritici that infect durum wheat do not infect hexaploid wheat, and vice versa.

³New races (previously unreported) of U. tritici detected from isolates collected in Manitoba and Saskatchewan, and from New Zealand sources (T63 and T64).

Five races were identified from the 64 durum wheat isolates (Table 2). The most common race was race T32, which was the phenotype of 55 (85%) of the isolates. Race T32 was first identified in the U. tritici population from durum wheat in Manitoba and Saskatchewan in 1979 and has been the predominant race (>50% frequency) since 1988 (Menzies et al. 2003). Races T26 and T33 were the next most common races identified during 1999–2003, with each race represented by three isolates (5%). Race T26 was first identified in pathogen populations from Manitoba and Saskatchewan in 1978, but has not been a common race, whereas race T33 was first identified in 1980, and was a common race until 1998 (Menzies et al. 2003). Race T3 was represented by one isolate in 2001 (Table 2). Race T3 was commonly found on durum wheat between 1965 and 1977, but has not been common since then. Race T42, which is virulent only on TD-13 (the universal suscept in the differential host set), has been previously identified from U. tritici isolates collected in 1995 on hexaploid wheat (Menzies et al. 2003) and was identified from two durum wheat isolates collected in 2001 (Table 2). Virulence to the durum wheat differential lines TD-1 and TD-19 was observed in 92 and 95%, respectively, of the U. tritici isolates tested from 1999 to 2003. Most of the registered durum wheat cultivars grown at this time would have been susceptible to isolates with virulence to TD-1 and TD-19 (Menzies, unpublished data) and Menzies et al. (2003) reported that the virulences to these differential lines are randomly associated. It is possible that the prevalence of these virulence genes in U. tritici races on durum wheat is the result of selection pressure by ineffective resistance genes in the commercial durum lines. Virulence to TD-11 was found in only three isolates (5%), all race T33. Virulence to TD-11 has been reported in Canada since the 1930s (Hanna 1937; Cherewick 1953), in race T4 before 1977, and in race T33 from 1982 to 1998 (Menzies et al. 2003).

Table 2. Incidence (number of isolates) of races of Ustilago tritici collected from durum wheat (Triticum turgidum L. durum) in Manitoba and Saskatchewan during 1999–2003.

Race^1	Year
	1999	2000	2001	2002	2003
T3			1
T26			3
T32	15	12	14	11	3
T33	1		2
T42			2

¹See Table 1 for race differential reactions.

Thirteen races were identified from the 55 hexaploid wheat isolates of U. tritici collected in Manitoba and Saskatchewan from 1999 to 2007, excluding the four races identified from the New Zealand sourced seed (Table 3). Six of the races have not been previously reported before: races T61, T64, T65, T66, T67 and T68. The most common races identified were races T7 (14 or 25% of isolates), T5 (11 or 20% of isolates), T64 (9 or 16% of isolates) and T10 (8 or 15% of isolates). Races T10 and T5 were the most commonly identified races in Manitoba and Saskatchewan in previous years (Menzies et al. 2003) but were surpassed by race T7 in this study. Race T7 was found at incidences of 10% or lower from 1990 to 1998. Races T1 and T18 were commonly found at low percentages in 1990–1998, whereas races T11 and T44 were sporadically identified by Menzies et al. (2003) and these trends continued in 1999–2007 (Table 3).

Table 3. Incidence (number of isolates) of races of Ustilago tritici collected from hexaploid wheats (Triticum aestivum L.) in Manitoba and Saskatchewan during 1999–2007.

Race^1	Year
	1999	2000	2001	2002	2003	2004	2005	2006	2007
T1	1								1
T5	5			2	1		2		1
T7	5^2		2	5	1		1
T10	4				4
T11			1
T18	1	1		1					1
T28	1^2
T44						1
T61	1
T62	1^2
T63	1^2
T64		2		3				4
T65		1
T66		1
T67			2
T68									1

¹See Table 1 for race differential reactions.

²One of the isolates of T7, and each isolate of T28, T62 and T63 from the 1999 collections were from infected plants which were grown from seed sourced in New Zealand.

The identification of six new races of U. tritici is an indication of the genetic variability in the pathogen population. For the most part, these new races represent a re-assortment of virulence found in common races of the pathogen, which can be explained by the sexual reproduction the pathogen undergoes each time it infects a host (Nielsen & Thomas 1996). Race T64 could be of concern because it represents the addition of TD-6 virulence to race T10, a race possessing virulence to eight differentials (Table 1). Virulence to TD-6 was not common in previous years (Menzies et al. 2003), so the identification of 11 (20%) isolates (nine of T64 and one of each of T65 and T66) indicates its presence in the pathogen population has become more common. Races T63, T64 and T66 represent new races with increased virulence to the differential lines compared with the previously identified races. Races T63, T64 and T66 possess virulence to 11, 9 and 12 of the differential lines, respectively, which is more than the most virulent race (T10) found previously (Table 1). Menzies et al. (2003) observed that the Canadian population of U. tritici was dominated by races with wide virulence compared with populations from other countries, such as India (Kaur et al. 2014) or eastern Siberia (Nielsen & Tikhomirov 1993). The presence of U. tritici races with a broader virulence spectrum in Canadian populations was considered to arise from a greater emphasis on breeding for resistance in commercial wheat lines, and possibly the use of different or more diverse sources of resistance in Canada.

The frequency of virulence to differentials TD-7, TD-5A, TD-4 and TD-12a was quite high in the U. tritici isolates from Manitoba and Saskatchewan at 96%, 78%, 67% and 56%, respectively. None of the isolates possessed virulence to TD-14, while only 2–5% of the isolates possessed virulence to TD-2, TD-3, TD-10 and TD-16. This indicates that some of these sources of resistance are still effective against the pathogen population in Manitoba and Saskatchewan.

Four races were identified from the four U. tritici isolates from plants grown from New Zealand sourced seed (Table 3). The origin of these isolates cannot be conclusively determined because the seed used for growing the plants came from a Canadian wheat breeding winter nursery in New Zealand. The seed may have been infected with isolates originating in New Zealand, or they may have become infected by teliospores originating from other plants in the nursery that were originally infected by teliospores in Canada where the seed originated. Race T7 was identified from a New Zealand isolate and is a common race in Canada (Table 3, Menzies et al. 2003), so it is possible that the isolate originated from Canada. However, race T28 has not been identified in Canada before (Nielsen 1987; Menzies et al. 2003) and races T62 and T63 have not been identified from Canada or elsewhere prior to this study. The virulence possessed by races T28 and T62 is similar to that of race T7, except that race T28 does not possess virulence to TD-4 and race T62 does not possess virulence to TD-7 (Table 1). This does not, however, indicate that Canada is the origin of these isolates of U. tritici. Race T63 possesses virulence to 11 of the differential hosts, which could indicate a Canadian origin because isolates from Canada tend to possess more virulence than isolates from other parts of the world (Menzies et al. 2003), but this is not definitive proof.

The results of this study show that the natural populations of U. tritici in Manitoba and Saskatchewan continue to evolve in terms of virulence. The pathogen population on durum wheat was dominated by race T32, which is similar to previous studies (Menzies et al. 2003), with a decline in the incidence of race T33, reflecting the decline in the incidence of virulence to differential TD-11. The pathogen population on hexaploid wheat was dominated by an increase in the incidence of race T7, and six new races were identified. These new races are significant in that three of them (T64, T65 and T66) possess virulence to TD-6, and 16% of the isolates tested were race T64. Virulence to TD-6 has been reported previously in Canada, but not at such high frequencies. New hexaploid wheat lines for commercial production are tested for resistance to U. tritici prior to registration as part of the requirements for western Canada through the Prairie Grain Development Committee. These lines are inoculated with a mixture of races T2, T9, T10 and T39 (Menzies et al. 2009). Races T2 and T39 possess virulence to the wheat differential line TD-6 (Nielsen 1987). The commercialization of new wheat lines with TD-6 resistance may be selecting for races of U. tritici with virulence to TD-6, which may explain the increase in this virulence in the pathogen population. Nevertheless, there are still effective sources of resistance to this pathogen, and the development of resistant wheat lines, along with other control practices, have been effective in maintaining the incidence and severity of this disease at low levels as observed during recent annual surveys (Menzies et al. 2014, 2015, 2016).

Acknowledgements

The technical support of Cheri Saramaga and Zlatko Popovic is gratefully acknowledged.