Pathogenic diversity of Phytophthora sojae and breeding strategies to develop Phytophthora-resistant soybeans

Abstract

Phytophthora stem and root rot, caused by Phytophthora sojae, is one of the most destructive diseases of soybean [Glycine max (L.) Merr.], and the incidence of this disease has been increasing in several soybean-producing areas around the world. This presents serious limitations for soybean production, with yield losses from 4 to 100%. The most effective method to reduce damage would be to grow Phytophthora-resistant soybean cultivars, and two types of host resistance have been described. Race-specific resistance conditioned by single dominant Rps (“resistance to Phytophthora sojae”) genes and quantitatively inherited partial resistance conferred by multiple genes could both provide protection from the pathogen. Molecular markers linked to Rps genes or quantitative trait loci (QTLs) underlying partial resistance have been identified on several molecular linkage groups corresponding to chromosomes. These markers can be used to screen for Phytophthora-resistant plants rapidly and efficiently, and to combine multiple resistance genes in the same background. This paper reviews what is currently known about pathogenic races of P. sojae in the USA and Japan, selection of sources of Rps genes or minor genes providing partial resistance, and the current state and future scope of breeding Phytophthora-resistant soybean cultivars.

Introduction

Phytophthora stem and root rot (PSR), caused by the soil-borne Oomycete Phytophthora sojae (Kaufmann and Gerdemann 1958), is one of the most serious and widespread diseases of soybean [Glycine max (L.) Merr.] (Schmitthenner 1999). PSR is most often encountered when seeds are planted in poorly drained soils with a high clay content, and in fields subjected to temporary flooding and ponding. Disease can occur at any stage of soybean development from seedling to harvest, though it primarily affects seeds and seedlings (Schmitthenner 1985). When soybean plants are infected by P. sojae, the stem of the plant appears water-soaked and turns red-brown, and the infection results in wilting and the death of plants (Dorrance et al. 2003). P. sojae produces motile zoospores from infected tissues which can initiate further cycles of disease (Gijzen and Qutob 2009, Schmitthenner 1985). Large numbers of oospores can persist in the soil for many years without a host, and this may cause continuous crop losses (Kato 2010, Schmitthenner 1999). In addition, plants infected with P. sojae may become more vulnerable to infection by other soilborne pathogens.

In the USA, PSR was first observed in the state of Indiana in 1948 and in Ohio in 1951 (Kaufmann and Gerdemann 1958). PSR now occurs in most of the soybean growing regions of the USA, but is most common in the northern half of the country, where environmental conditions are generally more favorable for the pathogen (Dorrance and Schmitthenner 2000). P. sojae has been reported in soybean-producing areas in Asia, Africa, Australia, Europe, and North and South America (Schmitthenner 1999). In Japan, PSR was first observed in 1977 on Hokkaido, the northernmost island (Tsuchiya et al. 1978), and has subsequently been observed in Shizuoka, Yamagata, Akita, Saga, Niigata, Fukuoka, Hyogo, Toyama, Fukui, and Miyagi Prefectures (Sugimoto et al. 2006). In recent years, the yields of soybean and the income of soybean producers have decreased dramatically in the USA, where Wrather and Koenning (2006) estimated that annual crop damage from PSR between 2003 and 2005 averaged about $251.6 million. It is therefore essential to construct effective disease management strategies as quickly as possible.

Methods for reducing economic losses due to PSR include fungicide applications (Anderson and Buzzell 1982), planting resistant cultivars (Dorrance et al. 2003, Schmitthenner 1999), improving soil drainage (Schmitthenner 1985), modifying tillage practices (Workneh et al. 1998), and applying calcium-containing compounds (Sugimoto et al. 2005, 2007, 2008b, 2009, 2010a). Increasing public concern about environmental and health consequences of the widespread use of conventional fungicides has encouraged research on alternative disease control strategies (Sugimoto et al. 2010a).

Schmitthenner (1999) reported that the most effective way to reduce damage from PSR would be to plant resistant cultivars, and numerous sources of resistance have been identified. Two distinct types of host resistance have been described: (i) race-specific resistance conditioned by a single dominant Rps (“resistance to Phytophthora sojae”) gene and (ii) partial resistance conferred by multiple genes acting together. Partial resistance, sometimes referred to as tolerance or field resistance, is characterized by fewer rotted roots and disease progression at a much slower rate than what occurs in susceptible cultivars. Race-specific Rps genes have been widely used in commercial soybean cultivars (Dorrance et al. 2000, Slaminko et al. 2010). Agronomically competitive cultivars with resistance to PSR and other important diseases and pests would reduce both crop losses and production expenses, allowing producers to increase their incomes.

This review summarizes what is currently known about pathogenic races of P. sojae in Japan and the USA, methods used to identify and select soybean plants with Rps genes or partial resistance, and the current state and future scope of breeding soybean cultivars with race-specific or partial resistance to PSR.

Three characteristics of resistance to plant diseases

Van der Plank (1963) proposed the terms “vertical resistance” and “horizontal resistance” to describe two types of resistance found in plants. Vertical resistance confers absolute protection against some, but not all races of a pathogen, whereas horizontal resistance confers incomplete protection to all races of a pathogen. Fry (1982) proposed three characteristics of plant resistance to a pathogen: the magnitude of effect, the number of genes that govern the resistance, and the differential nature of the host’s reaction to pathogen races. According to Fry’s proposal, vertical resistance is conferred by a single gene with large and differential effects, and horizontal resistance is conditioned by multiple genes with small and non-differential effects. There are exceptions, however, such as the incomplete resistance that is conferred by a single gene in the rice/rice blast pathosystem (Zenbayashi et al. 2002). In this review, the terms race-specific resistance and partial resistance are used to refer to qualitatively and quantitatively inherited resistance, respectively. While this is convenient for the sake of simplifying discussion, the reader should bear in mind that some Rps genes may not confer complete immunity to certain races of the pathogen, resulting in a reaction that resembles quantitatively inherited partial resistance.

Race-specific resistance

Rps genes

Soybean cultivars and germplasm accessions differ in their reactions to different isolates of P. sojae (Kaufmann and Gerdemann 1958). Monogenic resistance conditioned by certain Rps genes has been providing reasonable protection against the majority of P. sojae populations in the USA for the last four decades (Bhattacharyya et al. 2005). Rps genes are thought to activate effector-triggered immune responses, similar to resistance (R) genes in other pathosystems (Dong et al. 2011). The first Phytophthora resistance gene was identified in the 1950s (Bernard et al. 1957). To date, 14 Rps genes at eight genomic loci have been reported (Sandhu et al. 2004) (Table 1). They are Rps1 (Bernard et al. 1957), Rps2 (Kilen et al. 1974), Rps3 (Mueller et al. 1978), Rps4 (Athow et al. 1980), Rps5 (Buzzell and Anderson 1981), Rps6 (Athow and Laviolette 1982), Rps7 (Anderson and Buzzell 1992) and Rps8 (Gordon et al. 2006, Sandhu et al. 2005). The Rps1 locus contains five functional alleles (Rps1a, 1b, 1c, 1d and 1k) (Bernard et al. 1957, Buzzell and Anderson 1992, Mueller et al. 1978), and the Rps3 locus contains three (Rps3a, 3b and 3c) (Mueller et al. 1978, Ploper et al. 1985). With the exception of Rps2, which confers incomplete resistance and is root-specific (Mideros et al. 2007), Rps genes usually provide absolute protection (i.e., immunity) against incompatible P. sojae races. Rps8 is the most recently identified Rps gene, and was discovered in PI 399073, a South Korean landrace (Sandhu et al. 2005). Near-isogenic lines (NILs), each carrying one of 14 Rps genes in the background of ‘Williams’ (rps; susceptible to Phytophthora sojae) were developed by Dr. Richard Bernard of the USDA-ARS in Urbana, IL, USA (Table 1; Bernard et al. 1991, Dorrance et al. 2004). This series of NILs included a line with the Rps1k gene from ‘Kingwa’ that was released as ‘Williams 82’ (Bernard and Cremeens 1988), the cultivar that was later used in the project to sequence the G. max genome (Schmutz et al. 2010).

Table 1.

Phytophthora resistance genes in soybean

Rps gene	Sourcea	Molecular linkage group	Citation
Rps1a	L88-8470	N	Bernard et al. (1957)
Rps1b	L77-1863	N	Mueller et al. (1978)
Rps1c	L75-3735	N	Mueller et al. (1978)
Rps1d	L93-3312, PI 103091	N	Buzzell and Anderson (1992)
Rps1k	L77-1794	N	Bernard and Cremeens (1981)
Rps2	L76-1988	J	Kilen et al. (1974)
Rps3a	L83-570	F	Mueller et al. (1978)
Rps3b	L91-8347	F	Ploper et al. (1985)
Rps3c	L92-7857	F	R. Nelson, personal communication
Rps4	L85-2352	G	Athow et al. (1980)
Rps5	L85-3059	G	Buzzell and Anderson (1981)
Rps6	L89-1581	G	Athow and Laviolette (1982)
Rps7	L93-3258	N	Anderson and Buzzel (1992)
Rps8	PI 399073	F	Gordon et al. (2006), Sandhu et al. (2005)

^a“L” lines are backcross lines developed by Bernard et al. (1991).

Racial diversity of Phytophthora sojae, and germplasm with effective Rps Phytophthora resistance genes

Breeding Phytophthora-resistant cultivars requires selection of effective Rps genes or parental lines with Phytophthora resistance. Thus, it is essential to be familiar with the racial diversity of P. sojae in soybean-producing regions. Grau et al. (2004) summarized the virulence pathotypes of North American P. sojae races on a set of eight differentials with either Rps1a, Rps1b, Rps1c, Rps1d, Rps1k, Rps3a, Rps6, or Rps7 (Table 2) using the hypocotyl inoculation method developed by Laviolette and Athow (1981). Single-zoospore isolates are generally used for the determination of race (Bhat et al. 1993). Since 1955, at least 55 physiologic races of P. sojae have been identified in the USA on the basis of the unique reaction patterns produced on the eight differential soybeans genotypes in Table 2 (Grau et al. 2004). The dominant races and racial diversity differ in each area (Grau et al. 2004). In the 1960s, Rps1a was the first resistance gene to be widely deployed in the USA, where it remained effective for approximately eight years (Grau et al. 2004, Schmitthenner 1985). Rps1a can still be found in approximately 5% of the commercial cultivars planted in the Midwestern USA (Slaminko et al. 2010). Rps1c, Rps1k, Rps3a and Rps6 were subsequently deployed in the North Central/Midwest region of the USA (Dorrance et al. 2003, Gordon et al. 2007) (Table 3). Rps1k confers strong resistance against a large number of North American P. sojae races, and has been the most stable and widely used Rps gene in the last two decades (Gao et al. 2005, 2008, Schmitthenner 1994) (Table 3). Slaminko et al. (2010) tested 3,533 USA commercial cultivars for their resistance to P. sojae, and found that 51% carried at least one Rps gene. Half of them had Rps1c, while another 40% of them had Rps1k-mediated resistance to PSR. Gordon et al. (2007) stated that while single gene-mediated resistance has been an effective means for managing this soilborne disease, it is necessary to continue identifying novel Rps genes.

Table 2.

Races of Phytophthora sojae from the USA classified using eight differentials with Rps1a, Rps1b, Rps1c, Rps1d, Rps1k, Rps3a, Rps6, or Rps7^a

Race	Virulence pathotype
0	—
1	7
2	1b, 7
3	1a, 7
4	1a, 1c, 7
5	1a, 1c, 6, 7
6	1a, 1d, 3a, 6, 7
7	1a, 3a, 6, 7
8	1a, 1d, 6, 7
9	1a, 6, 7
10	1a, 1b, 1c, 1d, 1k, 3a
11	6, 7
12	1a, 1b, 1c, 1d, 1k, 3a
13	6, 7
14	1c, 7
15	3a, 7
16	1b, 1c, 1k
17	1b, 1d, 3a, 6, 7
18	1c
19	1a, 1b, 1c, 1d, 1k, 3a
20	1a, 1b, 1c, 1k, 3a, 7
21	1a, 3a, 7
22	1a, 1c, 3a, 6, 7
23	1a, 1b, 6, 7
24	1a, 3a, 6, 7
25	1a, 1b, 1c, 1k, 7
26	1b, 1d, 3a, 6, 7
27	1b, 1c, 1k, 6, 7
28	1a, 1b, 1k, 7
29	1a, 1b, 1k, 6, 7
30	1a, 1b, 1k, 3a, 6, 7
31	1b,1c, 1d, 1k, 6, 7
32	1b, 1k, 6, 7
33	1a, 1b, 1c, 1d, 1k, 7
34	1a, 1k, 7
35	1a, 1b, 1c, 1d, 1k
36	3a, 6
37	1a, 1c, 3a, 6, 7
38	1a, 1b, 1c, 1d, 1k, 3a, 6, 7
39	1a, 1b, 1c, 1k, 3a, 6, 7
40	1a, 1c, 1d, 1k, 7
41	1a, 1b, 1d, 1k, 7
42	1a, 1d, 3a, 7
43	1a, 1c, 1d, 7
44	1a, 1d, 7
45	1a, 1b, 1c, 1k, 6, 7
46	1a, 1c, 3a, 5, 7
47	1a, 1b, 1c, 7
48 = 1b	(5), 7
49 = 5b	1a, 1c, (4), 6, 7
50 = 13b	(4), 6, 7
51	1c, 5, 6, 7
52 = 1b	(3b, 5), 7
53	1a, 1b,1c, 3a, 5, 7
54	1d, 7
55	1d, 3a, 3c, 4, 5, 6, 7

^aThis Table was adapted from Grau et al. (2004) with small changes.

^bAdditional differentials had been incorporated with Rps2, Rps3b, Rps3c, Rps4, or Rps5, but the original isolates were not reclassified.

Table 3.

Sources of Rps genes for developing new cultivars with resistance to P. sojae

Region	Cultivar or line (Rps gene)	Citation
USA	L77-1794 (Rps1k), PI 103091 (Rps1d)	Buzzell and Anderson (1992), Schmitthenner (1999)
USA (northern area)	L75-3735 (Rps1c), L77-1794 (Rps1k), L83-570 (Rps3a), L89-1581 (Rps6)	Dorrance et al. (2003), Gordon et al. (2007)
Hokkaido, Japan	Hayagin-1, KLS733-1	Tsuchiya et al. (1990)
Hyogo, Japan	PI 103091 (Rps1d), Gedenshirazu-1 (NDa), Ohojyu (NDa), Waseshiroge (NDa)	Sugimoto et al. (2006)
Hokkaido, Iwate, Miyagi, Yamagata, Fukushima, Ibaraki, Tochigi, Nagano, Shizuoka, Niigata, Toyama, Fukui, Hyogo, Tottori, Japan	PI 103091 (Rps1d), L77-1794 (Rps1k)	Moriwaki (2010)

^aND, not determined.

In Hokkaido, Japan, a total of 49 P. sojae isolates were recovered from soybean fields in 1985, and the reactions that each isolate induced on a set of eight differential cultivars (Table 1) from the USA were examined using the hypocotyl inoculation method (Tsuchiya et al. 1990). The reaction patterns that 45 of the 49 isolates produced on the eight differentials did not correspond to those of any of the 55 races previously identified in the USA. Six Japanese differential cultivars were subsequently selected to characterize the Hokkaido isolates (Tsuchiya et al. 1990) (Table 4). The 45 isolates represented 10 different virulence pathotypes (races A, B, C, D, E, F, G, H, I and J), and these were given letter designations to distinguish them from the numbers assigned to physiologic races from North America. Race D was the most prevalent, followed by races A, and J. Two cultivars, ‘Hayagin-1’ and ‘KLS733-1’, which were resistant to all 10 of the races from Hokkaido, were selected as resistance donors in soybean breeding programs (Tsuchiya et al. 1990) (Table 3).

Table 4.

Races of Phytophthora sojae reported in Japan using six Japanese cultivars as diferentials

Differentials	Racesa
	A	B	C	D	E	F	G	H	I	J	K	L	M	N	O
Isuzu	S	S	R	S	S	S	S	S	R	S	S	S	R	R	R
Chusei Hikarikuro	R	S	S	R	S	S	S	S	S	S	R	R	R	R	S
Kitamusume	S	S	S	S	S	S	S	S	S	S	R	R	S	R	S
Toyosuzu	R	R	S	R	S	R	S	S	S	S	R	R	R	R	R
Gedenshirazu-1	R	R	R	R	R	R	R	S	S	S	R	R	R	R	R
Ohojyu	R	R	R	S	R	S	S	R	S	S	S	R	R	R	R

^aS, susceptible; R, resistant.

Races A to J of P. sojae were reported by Tsuchiya et al. (1990).

Races K, L, M and N were reported in Hyogo from 2002 to 2004 (Sugimoto et al. 2006).

Race O was first noted in this study in the Sasayama region in 2006 (Sugimoto et al. 2010b).

In Hyogo Prefecture, which is famous in Japan for growing the black-seeded soybean cultivar ‘Tanbakuro’, PSR was first noted in 1987 (Sugimoto et al. 2006). P. sojae isolates were recovered from 164 fields between 2002 and 2008. The six Japanese differential soybean cultivars mentioned above were used for race determination of the 164 isolates with the agar medium inoculation method (Sugimoto et al. 2006). The results showed that race E was a major component of all the P. sojae populations, followed by races A, L, K, M, G, N and O (Sugimoto et al. 2010b). ‘Gedenshirazu-1’, PI 103091 (containing Rps1d) and ‘Ohojyu’ were resistant to most of the 164 isolates, and they were subsequently selected as sources of specific resistance from 10 Japanese cultivars and 14 differential cultivars with 14 Rps genes (Table 3). Rps1c, Rps3a and Rps6, which provided high levels of resistance to P. sojae races in the North Central USA, were ineffective in Hyogo (Sugimoto et al. 2011b).

Before 2009, Hokkaido and Hyogo were the only two regions of Japan for which there were published reports on the race distribution of P. sojae and the selection of cultivars that could be used as sources of resistance genes. Moriwaki (2010) collected 109 P. sojae isolates from 14 regions in Japan (Hokkaido, Iwate, Miyagi, Yamagata, Fukushima, Ibaraki, Tochigi, Nagano, Shizuoka, Niigata, Toyama, Fukui, Hyogo and Tottori), and examined the effectiveness of 14 Rps genes against them. Rps1d, Rps1k, Rps8, Rps1a, Rps1c, Rps7, Rps3b and Rps1b were found to provide resistance to 47–81% of the 109 isolates. Rps1d and Rps1k were the most effective resistance genes (Moriwaki 2010). These results corresponded with those of previous studies (Buzzell and Anderson 1992, Sugimoto et al. 2006), and indicated that these two Rps genes are potential sources of resistance that can be used to breed new resistant cultivars in the USA as well as in Japan (Table 3).

DNA markers linked to Phytophthora resistance genes

From the 1960s to 1980s, conventional breeding methods used phenotypic assays to identify seedlings carrying a Phytophthora resistance (Rps) gene, but this process is both time-consuming and costly (Young 1999). Great progress in DNA technology was made in the 1990s, resulting in the development and use of restriction fragment length polymorphism (RFLP), randomly amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP) and simple sequence repeat (SSR) markers and analyses. These facilitated the development of genetic maps for mapping genes and quantitative trait loci (QTLs) and the use of marker-assisted selection (MAS) of plants carrying one or more genes of high importance. During the last decade, the development of single nucleotide polymorphism (SNP) markers has greatly improved marker coverage of the soybean genome, while providing a class of marker that is suited to semi-automated analysis of DNA samples. Molecular markers make it possible not only to reduce the amount of time and labor expended, but also to allow efficient and accurate selection of Phytophthora-resistant plants. Since 1999, integrated genetic linkage maps of the soybean have been constructed using SSR, RFLP, RAPD, AFLP and SNP markers, classical traits and isozymes (Cregan et al. 1999, Cregan 2003, Diers et al. 1992, Song et al. 2004, http://soybase.org/). Currently, a genetic linkage map consisting of 20 linkage groups with approximately 1,500 SNP, 1,000 SSR markers, 700 RFLP and 73 RAPD markers, in addition to 46 classical trait loci, is available (Cregan 2003, Hyten et al. 2010, Song et al. 2004). This information allows researchers to identify molecular markers linked to important genes to use for MAS. With this information, the Rps1, Rps2, Rps3, Rps4, Rps5, Rps6, Rps7 and Rps8 loci, have been mapped to molecular linkage groups (MLGs) N, J, F, G, G, G, N and F, respectively (Cregan et al. 1999, Cregan 2003, Demirbas et al. 2001, Gordon et al. 2006, Lohnes and Schmitthenner 1997, Sandhu et al. 2005, Weng et al. 2001) (Table 1 and Fig. 1). The Rps4 and Rps8 loci mapped close to the Rps6 and Rps3 regions, respectively (Gordon et al. 2006, Sandhu et al. 2005). Diers et al. (1992) reported that the RFLP marker pT-5 was linked to the Rps5 locus (MLG G), but Demirbas et al. (2001) were unable to find SSR markers linked to it. Thus, SSR markers linked to each Rps gene except Rps5 have been reported (Demirbas et al. 2001).

Fig. 1.

SSR-based genetic linkage map of Rps genes on soybean molecular linkage groups (MLG), with corresponding Chromosomes (Chr) in parentheses. Genetic linkage map of the Rps3 and Rps8 region on MLG F (Chr 13) with the map distances reported by Cregan (2003) and Sandhu et al. (2004). Genetic linkage map of the Rps4 and Rps6 region on MLG G (Chr 18) with the map distances reported by Cregan (2003). Genetic linkage map of the Rps2 on MLG J (Chr 16) with the map distances reported by Cregan (2003). Genetic linkage map of the Rps1 and Rps7 region on MLG N (Chr 3) with the map distances reported by Cregan (2003). Genetic linkage map of the Rps1a and Rps7 region on MLG N (Chr 3) with the map distances reported by Weng et al. (2001). Genetic linkage map of the Rps1d region on MLG N (Chr 3) with the map distances reported by Sugimoto et al. (2008). ^a Molecular marker for single nucleotide polymorphism (SNP), because no SSR marker was found distal to Rps4.

Many researchers have studied the Rps1 locus, which carries five functional alleles, because of its effectiveness in controlling PSR. SSR markers linked to Rps1 (Cregan et al. 1999), Rps1a (Weng et al. 2001), Rps1b (Demirbas et al. 2001), Rps1c (Demirbas et al. 2001) and Rps1k (Bhattacharyya et al. 2005, Kasuga et al. 1997) were identified relatively early, but there were no published reports of molecular markers for the Rps1d gene until 2005 (Sugimoto et al. 2006), despite the fact that this gene was considered effective worldwide (Moriwaki 2010, Sugimoto et al. 2010b). Demirbas et al. (2001) were unable to find molecular markers linked to Rps1d because none of the markers closely linked to Rps1 on MLG N were polymorphic in a population derived from a cross between ‘Williams’ (rps) and L93-3312 (Rps1d). To identify markers for Rps1d, Sugimoto et al. (2008a) developed 123 F_2:3 lines from a cross between the traditional black soybean cultivar ‘Tanbakuro’ and PI 103091 (Rps1d). The results of virulence tests showed that the inheritance of Rps1d is controlled by a single dominant gene. Seven SSR markers (Sat_186, Satt631, Satt009, Satt675, Satt683, Satt624 and Satt080) on MLG N were linked to Rps1d, and a linkage map 44.0 cM in length was constructed. The closest markers, Sat_186 and Satt152, were mapped to positions 5.7 cM and 11.5 cM of this linkage map, respectively, on each side of the Rps1d gene. The accuracy of MAS was estimated to be 92.7% and 87.0% for MAS using Sat_186 and Satt152, respectively, by using progeny tests to confirm the presence of Rps1d. Selection efficiency was theoretically estimated to be 99.05% for MAS using both markers. We subsequently identified novel molecular markers even more closely linked to the Rps1d gene (Sugimoto et al. 2011b).

In Japan, Tazawa and Tezuka (2003) developed 143 F_2:3 lines from a cross between ‘Hayagin-1’ (resistant) and ‘Toyokomachi’ (susceptible) in order to identify markers for the Rps allele from Hayagin-1, which is resistant to all of the 10 P. sojae races in Hokkaido. Seven SSR markers on MLG N were associated with the Rps gene in Hayagin-1, and a linkage map was constructed. The closest marker, Satt152, was mapped 6.7 cM from the Rps gene, but the investigators were unable to identify a closer marker on the other side of the gene. These results suggest that the Rps gene in ‘Hayagin-1’ may be Rps1 or Rps7 according to information available at the SoyBase/Soybean Breeders Toolbox website (www.soybeanbreederstoolbox.org).

Isolation, characterization and evolution of Rps genes

Rps1k is the most frequently examined of the 14 Rps genes because Rps1k has provided stable and broad-spectrum Phytophthora resistance in the major soybean-producing regions of the USA for 40 years (Bhattacharyya et al. 2005, Schmitthenner et al. 1994). Kasuga et al. (1997) constructed a high-density linkage map of the Rps1k region, and the locus was mapped to a 0.13-cM interval between two AFLP markers. These markers were used to screen bacterial artificial chromosome (BAC) libraries to identify a BAC clone containing the Rps1k gene (Bhattacharyya et al. 2005). Rps1k was then isolated through positional cloning and transformation experiments (Gao et al. 2005, Gao and Bhattacharyya 2008). Sequence analysis of the cDNA clone showed that the sequence is a member of the coiled coil-nucleotide binding site-leucine rich repeat (CC-NBS-LRR) class of disease resistance genes. Bhattacharyya et al. (2005) reported that the soybean genome contains about 38 copies of a similar sequence, and that most of the copies are clustered in approximately 600 kb of contiguous DNA from the Rps1k region. Analysis of G. max sequence data released by the Soybean Genome Project, Department of Energy, Joint Genome Institute (http://www.phytozome.net/soybean; Schmutz et al. 2010) indicates the presence of 30 loci with genes on MLG N (Chr 3) that share sequence homology with Rps1k (Sugimoto et al. 2011b). The clustering of related genes at the Rps1k locus might have facilitated the expansion of Rps1 gene numbers and the generation of new recognition specificities. Graham et al. (2002) reported that resistance genes in soybean tend to be clustered in groups of genes that confer resistance to more than one type of pathogen. Plant disease resistance (R) genes often occur in clusters (Richly et al. 2002), which may facilitate the expansion of R gene loci and the generation of new R gene specificities through recombination and positive selection (Michelmore and Meyers 1998).

Although Rps1k was previously considered to be a single gene, two functional Rps genes (Rps1k-1 and Rps1k-2) were cloned from the Rps1k locus by Bhattacharyya et al. (2005), who sequenced the Rps1k region to try to gain a better understanding of the possible evolutionary steps that shaped the generation of Phytophthora resistance genes in soybean. As a result, Rps1k-3 from this Rps1k genomic region was evolved through intramolecular recombination between Rps1k-1 and Rps1k-2. They hypothesize that crossing over was one of the mechanisms involved in tandem duplication of CC-NBS–LRR sequences in the Rps1k region (Gao and Bhattacharyya 2008). Analyses of recombinants strongly indicated that at least one additional functional Rps gene maps next to the Rps1 locus (Bhattacharyya et al. 2005). Gao and Bhattacharyya (2008) also proposed from the sequencing analysis of the Rps1k region that the Rps1 locus is located in a gene-poor region where only a few full-length genes were predicted. These include two coiled coil-nucleotide binding-leucine rich repeat (CC-NB-LRR)-type Rps1k genes and retrotransposons (Gao and Bhattacharyya 2008). The abundance of repetitive sequences in the Rps1k region suggested that the location of this locus is in or near a heterochromatic region. Low recombination frequencies along with the presence of two functional Rps genes at this locus may explain why Rps1k has provided stable Phytophthora resistance in soybean for several decades.

The region containing Rps2 has been cloned and sequenced (Sandhu et al. 2005). It is comprised of three functional genes: (1) the powdery mildew resistance gene Rmd-c, (2) an ineffective nodulation gene named Rj2 and (3) Rps2 (Graham et al. 2000, 2002). This region contains at least nine resistance gene analogues (RGAs) similar to the Toll/Interleukin-1 receptor (TIR)-NBS-LRR class of resistance genes (Graham et al. 2002, Polzin et al. 1994). Rps2 was therefore proposed to be in the TIR-NBS-LRR class of resistance genes (Graham et al. 2002).

The Rps4 region also has been recently cloned and characterized, and genes similar to the CC-NBS-LRR resistance genes in the Rps1 locus were identified (Sandhu et al. 2004). Deletion of a disease resistance gene-like sequence leads to a loss of Rps4 function (Sandhu et al. 2004). Although Athow and Laviolette (1982) reported no linkages between Rps4 and Rps6, recent studies indicated that the two genes are either allelic or clustered (Demirbas et al. 2001, Sandhu et al. 2004) (Fig. 1).

The Rps3 locus has been mapped to a gene–rich region containing three additional disease resistance loci with one bacteria resistance gene (Rpg1) and two virus resistance genes (Rsv1 and Rpv1) (Sandhu et al. 2005). Three functional Phytophthora resistance genes were mapped to the Rps3 locus (Sandhu et al. 2005, http://soybeanbreederstoolbox.org/). Sandhu et al. (2005) reported that a novel Phytophthora resistance gene has been mapped to the Rps3 region.

The Rps8 gene mapped closely to the disease resistance gene-rich Rps3 region and this gene was located “below” Rps3 (Sandhu et al. 2005). Sandhu et al. (2005) reported that at least 11 disease resistance genes, including Rps8, have been mapped to this small genomic region. Other Rps genes could be cloned and characterized using information from SoyBase (http://soybeanbreederstoolbox.org/) and the sequence data from the soybean genome (http://www.phytozome.net/soybean). In addition, resequencing of regions where resistance genes reside could identify SNPs that may be directly responsible for phenotypic differences.

Partial resistance

Effectiveness of partial resistance in soybean

Continuous utilization of stable Rps genes in soybean cultivars grown in North America has resulted in selection pressures that promote the evolution of more pathogenic races of P. sojae (Grau et al. 2004). Numerous physiological races of P. sojae that can overcome the resistance conferred by the known Rps genes have been identified (Dorrance et al. 2003). Single Rps genes have been effective for 8 to 15 years, depending on inoculum density and environmental conditions (Schmitthenner 1985). The time cycle for loss of disease control with Rps1a and Rps1c was 8 to 10 years for each gene from the time of introduction of resistant cultivars with these genes (Schmitthenner et al. 1994). Cultivars having partial resistance are usually not severely damaged by P. sojae in the field, but are killed when hypocotyls or roots are inoculated with a compatible race (Schmitthenner 1985). It is reported that partial resistance is effective against all races of P. sojae (Dorrance et al. 2003, Schmitthenner 1985). Partial resistance can be evaluated by planting cultivars in the field and rating the incidence of diseased plants or assessing yield loss, or by using an assay in which agar cultures of a compatible race are placed 5 cm below the seeds of cultivars in long pots (Walker and Schmitthenner 1984a). The latter method gives more reliable results due to the better control over environmental conditions compared with the former method, which is influenced by the amount of rainfall and the population pathotypes. There is little P. sojae colonization of the roots of soybean cultivars with high levels of partial resistance (McBlain et al. 1991). Tooley and Grau (1984) proposed that this type of resistance would limit the lesion growth rate of the pathogen in host tissues and reduce the severity of root rot. Generally, partial resistance has been described as the relative ability of susceptible plants to survive infection without showing severe symptoms like death, stunting, or yield loss (Glover and Scott 1998). Dorrance et al. (2003) examined the effect of partial resistance on PSR incidence and seed yield of soybean in Ohio, and demonstrated that yields of soybean cultivars with partial resistance were not significantly different from those of cultivars with single Rps genes or Rps gene combinations in an environment with low disease pressure, indicating that genetic traits associated with high levels of partial resistance do not have a negative effect on yield. Walker and Schmitthenner (1983b) examined the heritability of tolerance to PSR in soybean. Heritability estimates were not affected by race-specific resistance, although soybean lines having incompatible Rps genes had a higher mean tolerance rating than lines which did not have Rps genes, indicating that race-specific resistance and tolerance were not completely independent.

‘Conrad’, a cultivar with high levels of partial resistance, developed more disease when inoculated 0–4 days after planting compared to infection 5 or more days afterwards, while ‘Resnik’, a cultivar with the Rps1k gene and moderate levels of partial resistance, showed no PSR symptoms (Dorrance and McClure 2001). Even if a cultivar with partial resistance is planted, additional control measures such as a combination of race-specific resistance with partial resistance, improved soil drainage, hilled row planting, or seed treatment with a fungicide might be necessary.

Evaluation of soybean cultivars with partial resistance to P. sojae

Dorrance and Schmitthenner (2000) examined 887 soybean plant introductions (from PI 273483 to PI 427107) from the USDA Soybean Germplasm Collection for partial resistance to P. sojae races 7, 17, 25, 30 and 31. A total of 438 (55.5%) accessions had high levels of partial resistance or tolerance to P. sojae (Table 5). Interestingly, 67.6% of the lines with high levels of partial resistance had originally been collected from South Korea, indicating that P. sojae may be endemic to East Asia. Dorrance et al. (2003) evaluated the partial resistance of 12 soybean cultivars in seven growing environments (i.e., site × year combinations). They found that the cultivar ‘Conrad’ (Fehr et al. 1989) possessed a high level of partial resistance or tolerance to PSR, and that partial resistance may provide protection when plants were subjected to diverse P. sojae populations (Table 5). Several other studies have also demonstrated that ‘Conrad’ was somewhat susceptible to all of the P. sojae races, but has strong partial resistance to the pathogen (Han et al. 2008, Jia and Kurle 2008, Weng et al. 2007). Mideros et al. (2007) reported that ‘General’ (with Rps1k) and ‘Jack’ had higher levels of partial resistance than other germplasm tested (Table 5) with the slant board assay. Jia and Kurle (2008) examined 113 early maturity group (MG) soybean PIs from the USDA germplasm collection for partial resistance to PSR, and ‘MN0902’ was found to have even higher partial resistance than ‘Conrad’. Two American cultivars (‘MN0302’ and ‘91B53’) and twelve accessions (PI 437161, PI 437700, PI 438148, PI 445831, PI 449459, PI 468377, PI 504484, PI 549051, PI 561308, PI 561389B, PI 592919 and PI 593975) had levels of partial resistance equivalent to ‘Conrad’ (Table 5).

Table 5.

Potential sources with partial resistance to Phytophthora sojae to breed new resistant cultivars

Cultivar or line with partial resistance trait	Citation
Conrad	Dorrance et al. (2003), Han et al. (2008), Jia and Kurle (2008), Sugimoto et al. (2010a), Weng et al. (2007)
General (with Rps1k)	Mideros et al. (2007)
Jack	Mideros et al. (2007)
438 accessions (PI 273483 to PI 427107) from the USDA Soybean Germplasm Collection	Dorrance et al. (2000)
MN0902, MN0302, 91B53, PI 437161, PI 437700, PI 438148, PI 445831, PI 449459, PI 468377, PI 504484, PI 549051, PI 561308, PI 561389B, PI 592919 and PI 593975	Jia and Kurle (2008)
Syoutou-1, Kitamijiro, Yuuhime, Horokanai-zairai, Wabash and Tim144	Yamashita (2008)

In Japan, 16 cultivars that are susceptible to ten P. sojae races from Hokkaido were examined for partial resistance in the field from 2005 to 2007 (Yamashita 2008). Six cultivars, including ‘Syoutou-1’, ‘Kitamijiro’, ‘Yuuhime’, ‘Horokanai-zairai’, ‘Wabash’, and ‘Tim144’, had higher partial resistance to PSR than the other ten cultivars (Table 5). The partial resistance of ‘Conrad’ was evaluated in field experiments conducted at five sites in Hyogo between 2006 and 2008. Disease incidence for ‘Conrad’ ranged from 0 to 3.4%, which was much lower than the 11.7 to 52.0% incidence observed on the susceptible cultivar ‘Tanbakuro’ (Sugimoto et al. 2010a). One Japanese cultivar with partial resistance significantly higher than ‘Conrad’ was discovered in the field experiments and the resistance was confirmed in laboratory examinations (Sugimoto et al. 2011a). These cultivars may be useful as sources of resistance for breeding new cultivars adapted to other parts of Japan or other countries where PSR is a problem.

DNA markers linked to partial resistance in soybean

The heritability of quantitatively inherited partial resistance is relatively high (Burnham et al. 2003, Han et al. 2008, Weng et al. 2007). To date, six QTLs associated with the partial resistance of ‘Conrad’ have been mapped to four different MLGs in soybean [two QTLs on MLG D1b+W (Chr 2), three QTLs on MLG F (Chr 13) and one on MLG J (Chr 16)] (Burnham et al. 2003, Han et al. 2008, Weng et al. 2007) (Table 6). Burnham et al. (2003) identified two QTLs in three F₄ populations (66 lines from ‘Conrad’ × ‘Sloan’, 79 lines from ‘Conrad’ × ‘Williams’ and 64 lines from ‘Conrad’ × ‘Harosoy’) using lesion length as an indicator of the level of partial resistance. The QTLs were mapped to MLG F (Chr 13) and D1b+W (Chr 2) using phenotypic data from growth chamber assays conducted at the seedling stage (Burnham et al. 2003). SSR markers Satt252 and Satt149 on MLG F and Satt579, Satt266 and Satt600 on MLG D1b+W were significantly associated with variation in stem lesion lengths. The QTLs on MLG F and D1b+W explained 21.4–35.0% and 10.6–20.7% of the genotypic variation for the three different populations, respectively. Han et al. (2008) detected three QTLs (named QGP1, QGP2 and QGP3) underlying tolerance to PSR in an F₇ population (‘Conrad’ × ‘OX 760-6-1’) consisting of 112 lines using growth chamber tests with three P. sojae isolates from China. Two markers, Satt509 (2.3–8.6 cM) and Satt343 (2.4–5.1 cM) on MLG F (Chr 13), were located near QGP1 and QGP2, respectively and marker OPL18₈₀₀ (2.35–10.63 cM from the QTL) on MLG D1b+W (Chr 2) was located near QGP3. Weng et al. (2007) recently identified a putative QTL on MLG J (Chr 16) in a ‘Conrad’ × ‘OX 760-6-1’ F₆ population consisting of 62 lines evaluated in the field at two locations. Satt414 and Satt596 were significantly (P < 0.005) associated with variation in reactions to PSR. The putative QTL was flanked by Satt414 (2.0 cM from the QTL) and Satt596 (4.5 cM from the QTL), which explained 13.7% and 21.5% of the total phenotypic variance, respectively.

Table 6.

Soybeans with partial resistance and the QTLs associated with Phytophthora stem and root rot tolerance

Cultivar or line with partial resistance	Population	Generation	P. sojae isolate	Number of QTL	Molecular linkage group (Chromosome)	Marker	Citation
Conrad	66 lines from Conrad × Sloan, 79 lines from Conrad × Williams, and 64 lines from Conrad × Harosoy	F4	OH25 (vir 1a, 1b, 1c, 1k, 7)	2	MLG F (Chr 13), MLG D1b+W(Chr 2)	Satt252 and Satt149 on MLG F, Satt579, Satt266, and Satt600 on MLG D1b+W	Burnham et al. (2003)
Conrad	62 lines from Conrad × OX 760-6-1	F6	Woodslee and Weaver	1	MLG J (Chr 16)	Satt414 and Satt596 on MLG J	Weng et al. (2007)
Conrad	112 lines from Conrad × OX 760-6-1	F7	JiXi, ShuangYaShan, JianSanJiang, Woodslee	3	MLG F (Chr 13), MLG D1b+W (Chr 2)	Satt509 and Satt343 on MLG F, OPL18800 on MLG D1b+W	Han et al. (2008)
PI 407162	298 lines from V71- 370 × PI 407162	F11	C2S1(vir 1a, 1b, 1c, 1k, 2, 3a, 3b, 3c, 4, 5, 6, 7)	3	MLG J (Chr 16), MLG I (Chr 20), MLG G (Chr 18)	Satt529 and Satt414 on MLG J, Sat_105 and Satt239 on MLG I, Satt235 and Satt163 on MLG G	Tucker et al. (2010)

Tucker et al. (2010) first identified three QTLs for partial resistance to a Midwestern USA isolate on MLG J (Chr 16), I (Chr 20) and G (Chr 18) using an interspecific recombinant inbred line (RIL) population consisting of 298 F₁₁ individuals derived from a cross between ‘V71-370’ (a G. max cultivar with moderate partial resistance) and PI 407162 (a G. soja accession with lower partial resistance) (Table 6). One PI 407126 gene at a major QTL on MLG J accounted for 22–42% of the phenotypic variation in three experiments. This QTL corresponds to the previously identified QTL on MLG J in ‘Conrad’ (Weng et al. 2007). It is unclear why the resistance gene at the QTL with the largest effect was inherited from the more susceptible parental line. A minor QTL gene from ‘V71-320’ that mapped to MLG I (Chr 18) explained 7–12% of the phenotypic variation, and the location of the QTL coincided with a previously mapped RGA designated as RGA018. Another QTL on MLG G, which explained 9–11% of the variation in three experiments, was derived from PI 407162 (susceptible), and the marker SLP142, a disease-related EST-SSR, was linked to this QTL. Tucker et al. (2010) did not have a good explanation for why resistance genes at the QTLs on MLG-J and G had been contributed by the more susceptible parent, PI 407162. The partial resistance QTLs reported here map to different genomic regions from those in which the known Rps genes have been mapped.

Mechanisms involved in partial resistance to Phytophthora stem rot

The molecular mechanisms and the defense responses associated with partial resistance to P. sojae in soybean have been examined in ‘Conrad’ and ‘OX 20-8’ (which carries Rps1a, but has a low level of partial resistance) (Vega-Sanchez et al. 2005). Correlations between the expression of defense-related genes and partial resistance between two cultivars were examined, and effective lesion-limiting mechanisms in ‘Conrad’ were found to occur primarily in the upper root section. At 72 hours after inoculation, transcript levels for PR1a, matrix metalloproteinase (MMP) and basic peroxidase (IPER) at the inoculation site and for IPER above the inoculation site were significantly higher in ‘Conrad’ compared with ‘OX 20-8’. They concluded that defense responses associated with the accumulation of PR1a, MMP, IPER and β-1,3-endoglucanase (EGL) mRNAs may contribute to the partial resistance response of soybean to P. sojae.

Thomas et al. (2007) indicated that the content of soybean root suberin (a complex biopolymer with a poly component associated with the cell wall and plasma membrane) was involved in partial resistance to P. sojae. ‘Conrad’ was found to contain significantly higher amounts of suberin than ‘OX 760-6’, a susceptible line. This correlation was supported by data from an analysis of nine cultivars and 32 recombinant inbred lines derived from a ‘Conrad’ × ‘OX 760-6’ cross.

Current status and future scope of developing Phytophthora-resistant soybean cultivars

Since P. sojae is now firmly established in many soybean production regions around the world, resistance to PSR will continue to be an important objective in many breeding programs. Slaminko et al. (2010) reported that a large percentage of soybean cultivars currently or recently marketed in the USA carry at least one Rps gene. Although race-specific resistance from certain Rps genes has been effective for many years, pathologists and breeders nevertheless recognize the value of partial resistance, which should be more durable and broader than resistance dependent on a single major gene. Similar benefits could also be achieved by “stacking” two or more Rps genes or an Rps gene and partial resistance genes in the same background. Molecular markers linked to Rps genes have been identified, and could be used for MAS to efficiently identify seedlings carrying the resistance gene. With this technique, backcrossing can be performed two or three times in a year to transfer the Rps gene into a different elite genetic background. In Hyogo Prefecture in Japan, the black-seeded Phytophthora-resistant line 262-1 (BC4) with Rps1d was developed as an alternative to the high-value, but PSR-susceptible cultivar ‘Tanbakuro’. This line was evaluated in the field from 2006 to 2008. The incidence of disease in this line was 0%, whereas control plants of ‘Tanbakuro’ had a disease incidence of 11.7–52.0% (Sugimoto et al. 2010a), showing that resistance conferred by some Rps genes provides adequate protection against PSR in Japan, where few Phytophthora-resistant cultivars have been developed.

Partial resistance controlled by QTLs would be more durable than resistance conditioned by single Rps genes, especially in the USA, where P. sojae races that can defeat even the most effective Rps genes have been detected (Dorrance et al. 2003). Several QTLs for partial resistance to P. sojae have been reported, and the molecular markers linked to the QTLs could be used for developing cultivars with partial resistance (Burnham et al. 2003, Han et al. 2008, Tucker et al. 2010, Weng et al. 2007). However, it is still difficult to integrate multiple QTLs at independent loci into a cultivar without some adverse effects. Dorrance et al. (2003) reported that plants with high partial resistance are still infected by P. sojae, which can impact yield under conditions of high inoculum and weather favorable to the pathogen. One of the most promising future strategies for controlling PSR would therefore be to combine partial resistance and race-specific resistance in Japan, as well as in the USA. This could maintain the effectiveness of resistance genes.

In response to the emergence of novel pathotypes, and to provide complete control of P. sojae, the identification of additional unique Rps genes is required. Sugimoto et al. (2011b) found a novel Rps gene from the Japanese soybean cultivar Waseshiroge. This Rps gene is either allelic to Rps1, or resides at a tightly linked locus in a gene cluster. Dorrance and Schmitthenner (2000) reported that the Republic of Korea is an area with many sources of resistance to P. sojae for both specific Rps genes and partial resistance.

Resistance to PSR could also be enhanced and prolonged if it is used in combination with other available tools and tactics in an integrated approach to disease management. While treatment of seeds with fungicides has proven effective for reducing seedling losses to P. sojae, more environmentally benign alternatives are needed. It was recently demonstrated that calcium compounds could reduce the severity of PSR in laboratory and field experiments (Sugimoto et al. 2008a, 2010a). Ca(HCOO)₂-A [Suicaru—a commercially available calcium formate formulation (Koei Chemical, Nagoya, Japan)] was the most effective in suppressing disease incidence among seven calcium compounds tested. By combining methods such as this with the planting of cultivars having both appropriate Rps genes and genes conditioning partial resistance, strategies for the integrated management of PSR of soybean can be both more sustainable and less threatening to the environment than some of the management practices currently used.