Copyright ? 2007 by the Genetics Society of America
The in Silico Map-Based Cloning of Pi36, a Rice Coiled-Coil–Nucleotide-Binding
Site–Leucine-Rich Repeat Gene That Confers Race-Specific
Resistance to the Blast Fungus
Xinqiong Liu,*,†Fei Lin,* Ling Wang* and Qinghua Pan*,1
*Laboratory of Plant Resistance and Genetics, College of Resources and Environmental Sciences, South China Agricultural University,
Guangzhou, 510642, China and†Key Biotechnology Laboratory of State Ethnic Affairs Commission, College of Life Science,
South-Central University for Nationalities, Wuhan, 430074, China
Manuscript received May 4, 2007
Accepted for publication May 15, 2007
The indica rice variety Kasalath carries Pi36, a gene that determines resistance to Chinese isolates of rice
japonica variety Nipponbare was used for an in silico prediction of the resistance (R) gene content of the
intervalandhencefor theidentification ofcandidate gene(s)forPi36.Threesuchsequences,which allhad
amplified from the genomic DNA of a number of varieties by long-range PCR, and the resulting amplicons
were inserted into pCAMBIA1300 and/or pYLTAC27 vectors to determine sequence polymorphisms cor-
related to the resistance phenotype and to perform transgenic complementation tests. Constructs con-
taining each candidate gene were transformed into the blast-susceptible variety Q1063, which allowed the
identification of Pi36-3 as the functional gene, with the other two candidates being probable pseudogenes.
The Pi36-encoded protein is composed of 1056 amino acids, with a single substitution event (Asp to Ser) at
residue 590 associated with the resistant phenotype. Pi36 is a single-copy gene in rice and is more closely
related to the barley powdery mildew resistance genes Mla1 and Mla6 than to the rice blast R genes Pita, Pib,
Pi9, and Piz-t. An RT–PCR analysis showed that Pi36 is constitutively expressed in Kasalath.
blast, remains the most important pathogen in most rice-
trol. However, because of the instability of the pathogen
and a high level of variability in pathogenicity between
isolates (Ou 1979; Valent and Chumley 1994), host re-
sistance is typically only short-lived in disease-prone envi-
The establishment of durable resistance requires the iso-
lation of multiple R genes, as this simplifies the process of
R gene stacking into elite cultivars, via either marker-
aided breeding or transgenesis.
system to study plant–pathogen interactions (Valent
1990). Race-specific resistance closely follows the classi-
calgene-for-generelationship(Silue ´ etal.1992;Jiaetal.
HE filamentous ascomycete Magnaporthe grisea
(Hebert) Barr, which is the causal agent of rice
2000). The isolation and subsequent characterization of
R genes will help to unravel the molecular mechanisms
underlying the interaction between host and pathogen.
Although more than 50 rice R genes have been docu-
mented to date (Chen et al. 2005; Liu et al. 2005), only 6
(Pib, Pita, Pi9, Pid2, Pi2, and Pizt) have as yet been isolated
(Wang et al. 1999; Bryan et al. 2000; Qu et al. 2006; Chen
et al. 2006; Zhou et al. 2006). The sequences of 5 of these
(Pib, Pita, Pi9, Pi2, and Pizt) include both nucleotide-
while Pid2 encodes a receptor-like kinase.
Plants use R genes to detectthe presenceof pathogen,
and then to induce a spectrum of defense responses.
elicitors has been established by a variety of direct and
indirect experimental evidence (Jia et al. 2000; Gu et al.
2005; Dodds et al. 2006). The commonest class of R gene
encodes proteins containing an NBS–LRR domain (Bent
1996; Hammond-Kosack and Jones 1997; Hulbert et al.
2001). These have been classified into two types on the
basis of the presence/absence of an N-terminal TIR do-
main. Genes in the TIR group are only known among the
Bai et al. 2002). The non-TIR group typically includes
a coiled-coil (CC) domain at the N terminus. The NBS
region is thought to be involved in signal transduction
Sequence data from this article have been deposited with the EMBL/
GenBank Data Libraries under accession no. DQ900896.
1Corresponding author: Laboratory of Plant Resistance and Genetics,
College of Resources and Environmental Sciences, South China Agricul-
tural University, Guangzhou, 510642, China.
Genetics 176: 2541–2549 (August 2007)
cascades involving phosphorylation/dephosphorylation
events with either ATP or GTP (Traut 1994; Dangl and
Jones 2001), whereas the CC domain may facilitate ho-
modimerization of the proteins themselves or heterodi-
merization with other proteins, generating interactions
that lead to the repression of signaling (Moffett et al.
2002; Hwang and Williamson 2003). Several studies
of recognition specificity for the pathogen avirulence
factor(s) (Meyers et al. 1998). LRR-containing sequences
are prone to adaptive evolution (Parniske et al. 1997;
Mcdowell et al. 1998; Ellis et al. 2000; Sun et al. 2001),
and in particular, their insertions and deletions have
been shown to be responsible for both R gene loss of
function and recognition specificity (Anderson et al.
1997; Wulff et al. 2001). For example, particular loss-
of-function alleles of the Arabidopsis thaliana genes RPS2
andRPM1 differ from theeffectivewild typebyonly one
Grant et al. 1995).
The indica rice variety Kasalath (formerly coded as
Q61) confers a stable and high level of partial resistance
against Chinese isolates of rice blast. The resistance gene
Pi36 has recently been mapped to a location on chro-
mosome8 (Liuetal.2005).Inthispaper, wedescribethe
positional cloning of Pi36 gene based on a prior bio-
informatics analysis, long-range PCR (LR–PCR), and an
efficient transformation-competent artificial chromo-
some (TAC) vector-based transformation technique. We
believe that this approach should be widely applicable
within rice and also other plant species. The cloned Pi36
gene represents an important resource for the develop-
ment of durable resistance to rice blast, and along with
other R genes its sequence should inform the molecular
basis of disease resistance in plants.
MATERIALS AND METHODS
Candidate gene prediction: Candidates for Pi36 were iden-
FGENESH, RiceGAAS, and Gramene, using as a reference the
Nipponbare sequence for the 17-kb genomic region defined
by the flanking markers CRG2 and RM5647 (Figure 1A). To
verify which, if any, of these are true candidates for Pi36, the
sequence from the same genetic interval was derived from the
blast-resistant variety Kasalath and two blast-susceptible varie-
ties, Aichi Asahi and Lijiangxintuanheigu (LTH), using a PCR
walkingapproach. The intervalwasdividedinto 11overlapping
amplifiable fragments, and sequence comparisons were per-
formed between the alleles from the resistant and susceptible
generated by DNAStar software (http:/ /www.DNAStar.com).
Candidate gene cloning: Primer sets were designed to am-
regions, using the above-mentioned gene annotation. Restriction
sites to enable cloning were identified from the Kasalath genomic
sequence. Pi36 candidates were amplified from Kasalath genomic
400 mm dNTP, 100-ng template, and 0.2 mm of each primer. The
primer sequences, PCR conditions, and restriction enzymes used
are listed in Table 1. Purified LR–PCR products of candidates
Pi36-1 and Pi36-2 from three independent reactions were ligated
into the BamHI and PstI sites of the vector pCAMBIA1300 to
form R36L1CAM and R36L2CAM, respectively. For the longest
candidate, Pi36-3, the LR–PCR products were cloned into the AscI
site of TAC vector pYLTAC27 to form R36L3TAC. To improve
transformation efficiency, the Pi36-3 insert was later recloned
into the vector pCAMBIA1300AscI, in which an AscI site was
engineered into the multiple cloning sites. This construct was
named R36L3CAM. The constructs were validated by restriction
analysis and sequenced from both ends using the vector primers
CAM1300F and CAM1300R. Details of the vectors, constructs and
primers used are listed elsewhere (Table 1; supplemental Table S1
at http:/ /www.genetics.org/supplemental/).
Complementation analysis: Constructscontainingeachcan-
didate gene were transformed into Agrobacterium tumefaciens
Hercules, CA). Clone stability was tested as per Qu et al.
blast-susceptible rice variety Q1063, as described by Hiei et al.
(1994). Selfed T1and T2progeny were tested for reaction to
blast infection with pathogen isolates CHL39 and CHL273,
using the spray method described elsewhere (Pan et al. 2003;
Liu et al. 2005). For these phenotyping tests, Kasalath and
Q1063 represented, respectively, the positive and negative
controls. A number of resistant transgenic individuals were
randomly selected and subjected both to PCR verification for
the presence of the transgene, using the gene-specific primers
CRG4F and CRG4R (supplemental Table S1 at http:/ /www.
genetics.org/supplemental/), and to Southern hybridization
analysis to estimate the transgene copy number. For the latter
procedure, ?3 mg genomic DNA was digested to completion
with HindIII; the products were separated by 0.8% agarose gel
electrophoresis and were then transferred to a nylon mem-
brane (Hybond-N1, Amersham, Buckinghamshire, UK). A part
of the HptII sequence, amplified from the vector pCAM-
BIA1300 by primers HptF and HptR (see supplemental Table
S1), was labeled with a-½32P? by random primer labeling
(TaKaRa) for use as a hybridization probe. Southern hybrid-
ization was also used to infer the copy number of Pi36-like
genes in rice. Genomic DNA of Kasalath and the blast-
susceptible variety AS20-1 was digested to completion with
EcoRI, KpnI, or BamHI and probed with sequences amplified
from the 59-untranslated region (UTR), 39 UTR, and a part of
the largest intron (L-intron) of the Pi36 gene (see supple-
mental Table S1).
Gene expression analysis: Two-week-old seedlings of Kasalath
and the blast-susceptible variety LTH were inoculated with path-
ogen isolate CHL39 and maintained in a greenhouse. Leaf sam-
ples were collected at 0, 6, 12, 24, 48, and 72 hpi. Total RNA was
isolated using the TRIzol reagent (Invitrogen, Carlsbad, CA),
following the manufacturer’s instructions. The assessment of
gene expression levels was obtained in a two-step reverse
transcription PCR (RT–PCR) process. The initial RT reaction
used the SuperScript II reverse transcriptase kit (Invitrogen),
following the manufacturer’s instructions. For the second PCR
reaction, a 0.5–2 ml aliquot of the first reaction was used as
template. Each experiment was performed in replicate. To
enablediscrimination between the various RT–PCR amplicons,
the RT–PCRprimers(see supplementalTableS1athttp:/ /www.
genetics.org/supplemental/ and Figure 4C) were designed
from exonic sequence flanking the predicted Pi36 introns, and
genomic DNA was included as a negative control. Primers for
rice actin (supplemental Table S1) were used as a positive RT–
26, 29, 32, and 35 cycles.
2542 X. Liu et al.
Rapid amplification of cDNA ends: The 59 and 39 end
sequences of the cDNA were determined by rapid amplifica-
tion of cDNA ends (RACE) using the GeneRacer kit (Invi-
trogen), following the manufacturer’s instructions. We used
susceptible (LTH and AS20-1) plants, harvested 24 hpi. The
same RACE primers were used for both the resistant and
susceptible templates. The full-length cDNA was divided into
three amplifiable fragments. The 59 RACE was generated by
a nested PCR using the primary primer set of GeneRacer 59
primer and GSP1, and the second set of the GeneRacer 59
nested primer and GSP2; similarly, the 39 RACE was generated
by the set GSP3 and the GeneRacer 39 primer, and the inter-
GSP1 (see supplemental Table S1 at http:/ /www.genetics.org/
supplemental/). The relative locations of all the gene-specific
primers are shown in Figure 3C (note that the intermediate
RT–PCR fragment overlaps both the 59 and 39 RACE frag-
ments). The RACE products were ligated into the pGEM-T
vector (Promega, Madison, WI), following the manufacturer’s
instructions, and sequenced.
DNA and protein sequence analysis: DNA sequence simi-
larity analysis was performed using software BLASTN and
com/berry.html). Genomic sequence comparisons were per-
formed with pairwise BLAST (http:/ /www.ncbi.nlm.nih.gov/
BLAST/bl2seq/bl2.html), and protein sequence similarity
analysis was performed with BLASTP (Altschul et al. 1997).
Multiple sequence alignments were obtained with ProbCons
(Do et al. 2005), and based on these outputs, a phylogenetic
tree was generated using the molecular evolutionary genetic
analysis (MEGA) program (Nei and Kumar 2000). Boot-
strapping was used to provide a confidence estimate for each
branch point. The theoretical isoelectric point and protein
molecular weight were computed using DNAStar software.
The CC structure was predicted by COILS (http:/ /www.ch.
within a 17-kb interval (Figure 1A). To identify candi-
dates for the gene, the Nipponbare version of the 17-kb
interval was scanned by gene prediction software, re-
vealing the three NBS–LRR-type sequences Pi36-1, Pi36-
2, and Pi36-3. Both Pi36-1 and Pi36-2 were identified by
RiceGAAS (Figure 1B), while Pi36-3 was identified by
both Gramene and FGENESH. The Pi36-3 sequence
includes both Pi36-1 and Pi36-2. LR–PCR products, each
representing one of the alleles of the three candidate
sequences, were successfully generated; these had the
anticipated lengths of 5.9, 9.5, and 16.1 kb. To exclude
PCR artifact as a source of sequence variation, three
independent LR–PCR products were cloned for each of
the candidate genes from each of the templates, and
The Kasalath alleles differ from those in the susceptible
varieties by a number of base substitutions and small
Primer sequences used for the amplification of candidate genes for Pi36 by long-range PCR
aNucleotides corresponding to a restriction enzyme recognition site are underlined.
bEnzyme used for cloning.
cAll PCRs included an initial denaturation step (94?/2 min), followed by 30 amplification cycles under the conditions indicated. At the end of the cycling procedure, a final
incubation of 72?/10 min was given. A, 94?/30 sec, 65?/6.5 min; B, 94?/30 sec, 63.8?/10.5 min; C, 94?/30 sec, 62.8?, 17 min.
Rice Blast R Gene Pi36
similarity between the alleles was 97% (between Kasalath
vs. LTH). Gene prediction identified the same set of
three candidate genes in each of the three alternative
versions of the Nipponbare 17-kb segment.
Complementation analysis of the candidate genes:
The constructs containing each candidate gene were
Q1063. A total of 259 and 39 independent T0trans-
formants, respectively, were generated using R36L1CAM
and 63 (R36L3CAM) independent T0individuals were
obtained. The pattern of segregation for rice blast resis-
transgenic. All T1individuals derived from a T0parent
carryingR36L1CAM or R36L2CAM werehighly susceptible
to blast isolates CHL39 and CHL273 (which are both avir-
ulent on Kasalath and virulent on Q1063). Ten of 24 tested
T1families derived from a T0parent carrying R36L3TAC
segregated resistant vs. susceptible in a ratio between 1:3.5
derived from a T0parent carrying R36L3CAM varied from
To confirm the presence and stable integration of the
transgene Pi36-3, molecular assay was first conducted by
Southern blot analysis. The results showed that all the
resistant transgenic plants harbored the transgene of
copies of the transgene Pi36-3, although few transgenic
plants contained multiple copies (Figure 2). To further
verify steady inheritance of the transgene Pi36-3, two
T2lines, LX182 T2-2 and LX182 T2-6, whose progeny
segregated 3:1 for resistance, were chosen for a cosegre-
gationanalysis between blast resistance andthe presence
of the marker CRG4, which lies within Pi36 (Figure 1A).
of CRG4 (Figure 3), Pi36-3 must represent a functional
Structure of Pi36: Six RACE products from both the
59 and 39 end of the genomic sequence of Kasalath were
59 and 39 RACE products as well as an intermediate RT–
PCR fragment overlapped one another, thereby pro-
viding complete coverage of the transcribed region
(Figure 4C). The size and structure of Pi36 were deter-
the genomic DNA sequence. Pi36 contains a 3171-bp
coding region, interrupted by four introns (433, 5069,
124, and 259 bp) and flanked by a 65-bp 59 UTR and a
725-bp 39 UTR. A 123-bp intron is present within the 39
UTR (Figure 4C). The size and structure of Pi36 are
rather different from those predicted by annotation of
Figure 2.—Southern blot analysis of blast resis-
tant transgenic plants. Genomic DNA was isolated
from resistant transformed and susceptible non-
transformed plants. A fragment of hptII was used
as a probe. Lane 1, molecular weight marker
lHindIII; lane 2, Kasalath (resistant); lane 3,
Q1063 (susceptible); lanes 4–12, transgenic T1
Figure 1.—In silico map-based cloning of Pi36. (A) Physical
and genetic map surrounding the Pi36 locus. The numbers
below the map represent distances in kilobases, as estimated
from the Nipponbare genome sequence. The numbers in pa-
rentheses represent the number of recombinants/gametes in
the mapping population previously reported (Liu et al. 2005).
(B) Pi36 candidate genes. Pi36-1 and Pi36-2 were predicted by
RiceGAAS and Pi36-3 by both FGENESH and Gramene. The
shaded box represents the coding region, and the hatched
boxes represent predicted 59 promoter and 39 poly(A) re-
gions, respectively. The numbers above the map refer to loca-
tion on the reference Nipponbare genomic sequence. The
targets for the LR–PCR primers are indicated.
2544X. Liu et al.
the genomic sequence (Figure 4, B and C), which con-
sists of 13 introns and 14 exons, with the translation
stop codon at position 2,884,299. In the cDNA-derived
structure, the stop codon is shifted 59 by 2784 bp. Both
start codons, however, lie at position 2,872,609 (Figure
Sequence analysis of the Pi36-encoded protein: A
comparison of the deduced amino acid sequence of the
Pi36 alleles from two susceptible japonica varieties ½LTH
(pi36j1) and Nipponbare (pi36j2)? and the resistant
Kasalath identified nine substitutions and two deletions
among the alleles. In addition, six substitutions distin-
guish the alleles Pi36 and pi36i fromthe blast-susceptible
indica variety AS20-1. A global analysis suggests that just
one substitution at residue 590 defines the functional
Pi36 gene. The deduced 1056-amino acid sequence of
the Pi36-encoded protein has a molecular mass of 120
kDa and a calculated isoelectric point of 6.61, and con-
5). The GMGGLGKTT sequence (beginning at residue
IVIDDIWD (beginning at residue 286) and GSKILVTTRK
(beginning at residue 310) correspond, respectively,
to the kinase 2 and kinase 3a consensus motifs (Traut
1994; Grant et al. 1995). In addition, GVPLAIITIAS
(beginning at residue 372) and LKNCLLYL (beginning
at residue 427) correspond, respectively, to the con-
served R gene NBS domains 2 and 3 consensus motifs
(Traut 1994; Grant et al. 1995). The final conserved
NBS motif VHD (beginning at residue 501) corre-
sponds to the conserved MHD (methionine–histidine–
aspartate) motif. The C-terminal region of the protein
composed of ?15% leucine. The repeats, which are
based on an LxxLxxLxxLxL consensus, vary in length
between 22 and 44 amino acids. LRRs 14, 15, 16, and 17
show little or no similarity to the LRR consensus. The
trated in Figure 5. Finally, a COIL analysis (Lupas et al.
1991; http:/ /www.ch.embnet.org/software/COIL_form.
html) showed that a CC region is probably present
(P . 0.95) between amino acids 24 and 52. The CC
Figure 3.—Pi36 gene complementation test and molecular analysis of the transgenic lines. (A) Resistance phenotypes of the
Pi36 gene donor cv. Kasalath and its receptor cv. Q1063 as well as transgenic plants of two T2lines segregated into 3R:1S against
isolate CHL39. R, resistant; S, susceptible. (B) Cosegregation analysis of the resistance phenotype with the transgenes. The am-
plified fragment with the primer pair CRG4F and CRG4R was subsequently digested with HaeIII, and the digested product was
subjected to 1.5% agarose gel electrophoresis. M represents standard molecular weight marker DL2000.
Figure 4.—The structure of Pi36, as deduced
from its genomic and cDNA sequences. (A)
The target gene region. The numbers above
the map show genomic positions in the Nippon-
bare genomic sequence. (B) Gene structure as
deduced from the genomic DNA sequence. (C)
Gene structure as deduced from the cDNA se-
quence. The shaded box indicates an exon,
and the line an intron. The positions of 59 and
39 UTR (hatched boxes), the translation start co-
don (ATG), and the translation stop codon (TGA
or TAG) are also shown. The annealing targets of
the RACE and RT-PCR primers are indicated. (D)
Structure of the Pi36-encoded protein, in which
three tandem conserved domains are shown.
Rice Blast R Gene Pi36
region contains three perfect hxxhxxh and one hxxhxxx
motif (where h represents one of L, I, M, V, or F, and x is
Pi36 belongs to the CC–NBS–LRR family of R genes.
Phylogenetic analysis of Pi36: Southern hybridiza-
tion analysis was employed to estimate the copy number
of Pi36-related genes in rice. Only a single hybridizing
fragment was present in both resistant and susceptible
genotypes, indicating that Pi36 is a single-copy gene
(supplemental Figure S2 at http:/ /www.genetics.org/
supplemental/). Interestingly, only a single copy is pre-
and 93-11 by BLAST analysis, suggesting that Pi36 is a
single-copy gene in the rice genome. The evolutionary
relationshipbetweenPi36andtwelve R genesrelatedwas
assessed by a phylogenetic amino acid-based sequence
analysis using ProbCons and MEGA (Figure 6). The de-
gree of homology shared by these genes varies consider-
ably, and two heterogeneous groups can be recognized,
reflecting an early divergence in the evolution of the R
of these groups, whereas those derived from the di-
R genes were further classified into eight distinct sub-
groups: Mla1 (I), Pi36 (II), Mla6 (III), Pita (IV), Pib (V),
Piz-t/Pi9 (VI), Xa1 (VII), and rp3 (VIII) (Figure 6). Pi36
appears to be more closely related to the barley powdery
mildew R genes Mla1 and Mla6 than to the rice blast R
the LRR region (data not shown).
Expression pattern of Pi36: Only a faint band of the
expected size was detected when 23 cycles were used,
most likely due to the low expression level of the Pi36
gene. However, a stronger band was observed when 29,
32, and 35 cycles were applied (Figure 7). The results
revealed that no detectable differences in expression
could be observed either in a time course postinocula-
tion with blast pathogen, or between resistant and sus-
ceptible hosts. Thus, the expression of Pi36 is likely to
be constitutive, and is not induced by blast infection.
An efficient cloning strategy for Pi36: Positional
cloning is the accepted means to isolate genes where
only phenotype and genomic location are known. The
latter requires the prior generation of a high-resolution
genetic map in the region surrounding the target.
Generally, the map needs to be complemented with
Figure 5.—Deduced amino acid sequence of the Pi36 encoded protein. The seven conserved motifs forming the NBS region
are underlined. Residue 590, the single amino acid substitution distinguishing the blast-resistant from the blast-susceptible form of
the protein, is double underlined. The C-terminal LRR region is shown separately from the rest of the sequence.
2546X. Liu et al.
largeinsert (YAC orBAC)libraries.Where theserequire-
ments are met, the target can be narrowed down to a
single insert (or a contig) based on the presence of
closely linked hybridizing markers (Song et al. 1995;
in rice, because its almost complete genome sequence is
in silico. With respect to Pi36, the necessary high-
resolution genetic map had already been assembled
(Liu et al. 2005). Thus the identification of candidates
for Pi36 could be reduced to a bioinformatics-based
search of the relevant physical genomic segment. Valida-
tion of the candidates was then achieved using a trans-
by exploiting the capability of LR-PCR to amplify
sequences too large for conventional PCR (Feuillet
et al. 2003; Horvath et al. 2003; Song et al. 2003). In the
event, the three candidate genes were all recovered by
LR–PCR, avoiding the need to subclone from a large
Insert size in the pCAMBIA1300 vector is limited,
making it difficult to clone sequences as large as 10 kb.
However, the pLYTAC27 TAC vector tolerates a much
larger insert size (Liu et al. 2002) and was successfully
used to clone Pi36-3 (.16 kb). We were then able to
transfer the target fragment into a modified form of
pCAMBIA1300 for the complementation study. This
cloning large genes.
Pi36 belongs to the CC–NBS–LRR family of R genes:
Some 580 NBS-encoding genes have been identified in
the rice genome. Of these, ?490 belong to the CC–
NBS–LRR family and 101 are thought to be pseudo-
genes (Bai et al. 2002; Monosi et al. 2004). Of the three
candidates for Pi36, one is a functional copy, while the
other two are probably pseudogenes. Interestingly, the
structure of Pi36 deduced from the genomic sequence
was rather different from that deduced from the cDNA
sequence. There are 13 introns in the former, but only 5
in the latter (Figure 4). This resulted in removing stop
codons from genomic positions 2,884,229 to 2,881,515.
the right flanking marker RM5647 for 2784 bp (Figure
1). This is an additional evidence to suggest that the
rence of recombination in this region within a stretch
of 6.4 kb (Liu et al. 2005). Our results can be considered
in light of the hypothesis that automatic annotations
commonly inserted introns to remove stop codons or
rupting the NBS domain of R genes are more common
in cereals than in dicotyledonous species (Bai et al.
and three characterized rice blast-resistance genes (Pi-ta,
Bryan et al. 2000; Pib, Wang et al. 1999; Pi9, Qu et al.
2006) carry intron(s) in their NBS domain. One intron
is present in the NBS region of Pi36 (of length 5069 bp,
and beginning at amino acid residue 284; Figure 4C),
similar in size to that present in Pi9. Pi36 therefore has
a unique structure with respect to intron position and
size when compared with other rice genes. Conserved
splicing sites (gt and ag) were present at the intron/
exon junctions of introns 1, 2, 3, and 5, but at intron 4,
the splicing site was ag and ct (Figure 4C). Introns have
Figure 7.—Expression patterns of
Pi36 assayed by semiquantitative RT–
PCR. Two-week-old resistant Kasalath
and susceptible LTH seedlings were
inoculated with blast (isolate CHL39).
The expression of Pi36 was assayed at
various time points postinoculation. Ge-
nomic DNA (gDNA) serves as a control
to distinguish PCR products from cDNA
and gDNA. The rice Actin1 gene acted
as a positive control.
Figure 6.—Phylogenetic analysis of Pi36 with other 10 R
genes. Multiple amino acid alignments were conducted using
ProbCons and a neighbor-joining phylogenetic tree was gen-
erated using MEGA. Numbers on branches indicate the per-
centage of 1000 bootstrap replicates which support the
adjacent node. The unit branch length is 0.2 nucleotide sub-
stitutions per site, as indicated by the bar.
Rice Blast R Gene Pi36
been commonly detected in the 59 UTR region of R
et al. 2005) but seldom in the 39 UTR. Two 39 UTR
introns are present in Pi9 (Qu et al. 2006), and one in
Pi36. Whether these features of Pi36 have any biological
significance has yet to be determined.
The LRR regions of rice NBS–LRR genes vary con-
siderably in size and sequence, reflecting substantial
divergence in the R genes (Bai et al. 2002). The LRRs
occupy almost the entire C-terminal region of the CC–
NBS–LRR proteins (Meyers et al. 2003), but the repeats
are mostly imperfect, with only few conforming to any
the regions are leucine-rich but have no clearly distin-
guishable repetitive structure. A similar pattern pertains
to the Pi36 sequence, which encodes a 136-residue non-
LRR region at its C-terminus. Further research is needed
play any role in the determination of specificity with
respect to particular pathogen isolates.
Evolutionary relationships between Pi36 and other
NBS–LRR R genes: R genes are involved in the disease
resistance response in a wide variety of plant species.
They share a common structure and therefore probably
act via a common mechanism. In evolutionary terms,
it is widely assumed that the R genes have a common
origin (Caicedo et al. 1999). The functional and evolu-
tionary analysis of R genes is the focus of much current
research. Pi36 is a single-copy R gene, and hence could
represent a useful model for such functional and evolu-
tionary studies. At the protein level, the Pi36 product
most closely resembles the barley Mla1 and Mla6 pro-
teins, and is less closely related to the rice bast Pib and
Pita proteins. Multiple alignment of the amino acid se-
that nonconservative residue substitution was most fre-
quent in the LRR domain and least in the NBS domain,
supporting the widely held view that the LRR regions
are subject to diversifying selection, and that they are
et al. 1998; Sun et al. 2001).
A single amino acid mutation is responsible for the
resistance phenotype: Pathogen–plant coevolution op-
erates by simultaneous selection for avirulence genes in
the pathogen and resistance genes in the host (Stahl
and Bishop 2000). The direct interaction between an
uct was first shown for the rice Pita/Avr-Pita system (Jia
et al. 2000). The deduced Pita protein of a susceptible
host differs from that of a resistant one by a single
substitution of serine for alanine (Bryan et al. 2000).
Similarly, we have established that a single amino acid
difference distinguishes the resistant and susceptible
alleles of the Pi36 product. In this case, thereplacement
of asparagine by serine determines blast resistance. A
possible mechanistic explanation of the large biological
effect of this small sequence difference could relate to
the finding that whenNectriahaematococca mycelia invade
et al. 2000). Thus serine and asparagines residues may be
important for determining the resistance or resistance-
related response. A more likely reason is that sequence
variation at the active site affects molecular interactions
and therefore changes function (Hanzawa et al. 2005).
We are presently attempting the isolation of Avr-Pi36 to
and the pathogen.
We are grateful to Y Liu for the kind provision of the pYLTAC27
vector and to Robert Koebner for critical reading of the manuscript.
This research has been supported by grants from the National 973
project (2006CB1002006), the National 863 projects (2002AA2Z1002;
2006AA10A103; 2006AA100101), the Innovation Research Team Project
from the Ministry of Education of China (IRT0448), the Guangdong
Provincial Natural Science Foundation (039254), and the Special
of Education of Guangdong Province.
Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang
et al., 1997Gapped BLASTand PSLBLAST: a new generation of
protein database search programs. Nucleic Acids Res. 25: 3389–
Anderson, P. A., G. J. Lawrence, B. C. Morrish, M. A. Ayliffe, E. J.
Finnegan et al., 1997 Inactivation of the flax rust resistance
gene M associated with loss of a repeated unit within the leucine-
rich repeat coding region. Plant Cell 9: 641–651.
Bai, J., L. A. Pennill, J. Ning, S. W. Lee, J. Ramalingam et al.,
2002Diversity in nucleotide binding site-leucine-rich repeat
genes in cereals. Genome Res. 12: 1871–1884.
Bent, A. F., 1996 Plant disease resistance genes, function meets
structure. Plant Cell 8: 1757–1771.
Bent, A. F., B. N. Kunkel, D. Dahlbeck, K. L. Brown, R. Schmidt
et al., 1994
RPS2 of Arabidopsis thaliana: a leucine-rich repeat
class of plant disease resistance genes. Science 265: 1856–1860.
Bryan, G. T., K. S. Wu, L. Farrall, Y. L. Jia, H. P. Hershey et al.,
2000 A single amino acid difference distinguishes resistant
and susceptible alleles of the rice blast resistance gene Pi-ta. Plant
Cell 12: 2033–2045.
Caicedo, A. L., B. A. Schaal and B. N. Kunkel, 1999
molecular evolution of the RPS2 resistance gene in Arabidopsis
thaliana. Proc. Natl. Acad. Sci. USA 96: 302–306.
Chen, S., L.Wang, Z.Q.Que, R.Q.Panand Q.H.Pan, 2005
and physical mapping of Pi37(t), a new gene conferring resistance
to rice blast in the famous cultivar St. No. 1. Theor. Appl. Genet.
Chen, X. W., J. J. Shang, D. X. Chen, C. L. Lei, Y. Zou et al., 2006
B-lectin receptor kinase gene conferring rice blast resistance.
Plant J. 46: 794–804.
Dangl, J. L., and J. D. G. Jones, 2001
grated defence responses to infection. Nature 411: 826–833.
Do, C. B., M. S. Mahabhashyam, M. Brudno and S. Batzoglou,
2005ProbCons: probabilistic consistency-based multiple se-
quence alignment. Genome Res. 15: 330–340.
Dodds, P. N., G. J. Lawrence, A. M. Catanzariti, T. Teh, C. I. Wang
et al., 2006Direct protein interaction underlies gene-for-gene
specificity and coevolution of the flax resistance genes and flax
rust avirulence genes. Proc. Natl. Acad. Sci. USA 103: 8888–8893.
Ellis, J. G., P. Dodds and T. Pryor, 2000
disease resistance gene specificities. Trends Plant Sci. 5: 373–379.
Feuillet, C., S. Travella, N. Stein, L. Albar, A. Nublat et al.,
2003Map-based isolation of the leaf rust disease resistance
gene Lr10 from the hexaploid wheat (Triticum aestivum L.) ge-
nome. Proc. Natl. Acad. Sci. USA 100: 15253–15258.
Plant pathogens and inte-
The generation of plant
2548 X. Liu et al.
Grant, M. R., L. Godiard, E. Straube, T. Ashfield, J. Lewald et al.,
1995 Structure of the Arabidopsis RPM1 gene enabling dual
specificity disease resistance. Science 269: 843–846.
Gu, K., B. Yang, D. Tian, L. Wu, D. Wang et al., 2005
sion induced by a type-III effector triggers disease resistance in
rice. Nature 435: 1122–1125.
Hanzawa, Y., T. Money and D. Bradley, 2005
converts a repressor to an activator of flowering. Proc. Natl. Acad.
Sci. USA 102: 7748–7753.
Hiei, Y., S. Ohta, T. Komari and T. Kumashiro, 1994
transformation of rice (Oryza sativa L.) mediated by Agrobacterium
and sequence analysis of the boundaries of the T-DNA. Plant J. 6:
Horvath, H., N. Rostoks, R. Brueggeman, B. Steffensonm, D. Von
Wettstein et al., 2003 Genetically engineered stem rust resis-
tance in barley using the Rpg1 gene. Proc. Natl. Acad. Sci. USA
gene complexes, Evolution and utilization. Annu. Rev. Phytopa-
thol. 39: 285–312.
Hwang, C., and V. M. Williamson, 2003
ated intramolecular interactions in nematode recognition and
cell death signaling by the tomato resistance protein Mi. Plant
J. 34: 585–593.
Jia, Y. L., S. A. Mcadams, G. T. Bryan, H. P. Hershey and B. Valent,
2000 Direct interaction of resistance gene and avirulence gene
products confers rice blast resistance. EMBO J. 19: 4004–4014.
Kiyosawa, S., 1981Gene analysisfor blast resistance. Oryza 18: 196–
physical mapping of Pi36(t), a novel rice blast resistance gene lo-
of new transformation-competent artificial chromosome vectors
and rice genomic libraries for efficient gene cloning. Gene 282:
Lupas, A., M. Vandyke and J. Stock, 1991
from protein sequences. Science 252: 1162–1164.
Mackill, D. J., and J. M. Bonman, 1992
tance in near-isogenic lines of rice. Phytopathology 82: 746–749.
Mcdowell, J. M., M. Dhandaydham, T. A. Long, M. G. Aarts, S.
Goff et al., 1998 Intragenic recombination and diversifying se-
lection contribute to the evolution of downy mildew resistance at
the RPP8 locus of Arabidopsis. Plant Cell 10: 1861–1874.
are subject to divergent selection. Plant Cell 11: 1833–1846.
Meyers, B. C., A. W. Dickerman, R. W. Michelmore, R. M. Pecherer,
S. Sivaramakrishnan et al., 1999
encode members of an ancient and diverse protein family within
the nucleotide-binding superfamily. Plant J. 20: 317–332.
Meyers, B. C., A. Kozik, A. Griego, H. Kuang and R. W. Michelmore,
2003Genome-wide analysis of NBS-LRR-encoding genes in
Arabidopsis. Plant Cell 15: 809–834.
Moffett, M., G. Farnham, J. Peart, and D. C. Baulcombe,
2002Interaction between domains of a plant NBS-LRR protein
in disease resistance related cell death. EMBO J. 21: 4511–4519.
Monosi, B., R. J. Wisser, L. Pennill and S. H. Hulbert, 2004
genome analysis of resistance gene homologues in rice. Theor.
Appl. Genet. 109: 1434–1447.
Nei, M., and S. Kumar, 2000
Molecular Evolution and Phylogenetics.
Oxford University Press, Oxford.
Ou, S. H., 1979Breeding rice for resistance to blast, a critical view,
pp. 79–137 in Proceedings of the Rice Blast Workshop. International
Rice Research Institute, Mantla, Philippines.
Ou, S. H., 1985 Blast, pp. 109–201 in Rice Diseases, Ed. 2. Common-
wealth Mycological Institute, Kew Surrey, UK.
Pan, Q., Y. S. Liu, O. Budai-Hadrian, M. Sela, L. Carmel-Goren
et al., 2000 Comparative genetics of nucleotide binding site-
leucine rich repeat resistance gene homologues in the genomes
of two dicotyledons, tomato and Arabidopsis. Genetics 155:
R gene expres-
A single amino acid
Predicting coiled coils
Inheritance of blast resis-
Plant disease resistance genes
Pan, Q. H., Z. Hu, T. Tanisakaand L. Wang, 2003
the blast resistance gene Pi15, linked to Pii, on rice chromosome
9. Acta Bot. Sin. 45: 871–877.
Parniske, M., K. E. Hammond-Kosack, C. Golstein, C. M. Thomas,
D. A. Jones et al., 1997 Novel disease resistance specificities re-
sult from sequence exchange between tandemly repeated genes
at the Cf-4/9 locus of tomato. Cell 91: 821–832.
Qu, S., G. Coaker, D. Francis, B. Zhou and G. L. Wang, 2003
velopment of a new transformation-competent artificial chromo-
some (TAC) vector and construction of tomato and rice TAC
libraries. Mol. Breed. 12: 297–308.
Qu, S., G. Liu, B. Zhou, M. Bellizzi, L. Zeng et al., 2006
spectrum blast resistance gene Pi9 encodes a nucleotide-binding
site-leucine-rich repeat protein and is a member of a multigene
family in rice. Genetics 172: 1901–1914.
Rogers, L. M., Y. K. Kim, W. Guo, L. Gonza ´lez-Candelas, D. Li et al.,
2000Requirement for either a host- or pectin-induced pectate
lyase for infection of Pisum sativum by Nectria hematococca.
Proc. Natl. Acad. Sci. USA 97: 9813–9818.
Silue ´, D., J. L. Notteghem and D. Tharreau, 1992
gene-for-gene relationship in the Oryza sativa-Magnaporthe grisea
pathosystem. Phytopathology 82: 577–580.
Song, J. Q., J. M. Bradeen, S. K. Naess, J. A. Raasch, S. M. Wielgus
et al., 2003Gene RB cloned from Solanum bulbocastanum confers
broad spectrum resistance to potato late blight. Proc. Natl. Acad.
Sci. USA 100: 9128–9133.
Song, W. Y., G. L. Wang, L. L. Chen, H. S. Kim, L. Y. Pi et al., 1995
receptor kinase-like protein encoded by the rice disease resis-
tance gene, Xa21. Science 270: 1804–1806.
Stahl, E. A., and J. G. Bishop, 2000
the molecular level. Curr. Opin. Plant Biol. 3: 299–304.
Sun, Q., N. C. Collins, M. Ayliffe, S. M. Smith, J. Drake et al.,
2001Recombination between paralogues at the rp1 rust resis-
tance locus in maize. Genetics 158: 423–438.
Sun, X. L.,Y. L.Cao, Z. F. Yang, C. G. Xu, X. H. Li et al., 2004
gene conferring resistance to Xanthomonas oryzae pv. oryzae in rice,
encodes an LRR receptor kinase-like protein. Plant J. 37: 517–
Traut, T. W., 1994The functions and consensus motifs of nine
types of peptide segments that form different types of nucleotide
binding sites. Eur. J. Biochem. 229: 9–19.
Valent, B., 1990Rice blast as a model system for plant pathology.
Phytopathology 80: 33–36.
Valent, B., and F. G. Chumley, 1994
anism of genetic instability in the rice blast fungus, pp. 111–134
in Rice Blast Disease, edited by R. S. Zeigler, S. A. Leong and P. S.
Teng. CABI, Wallingford, UK.
Van Der Vossen, E. A., J. Gros, A. Sikkema, M. Muskens,
D. Wouters et al., 2005The Rpi-blb2 gene from Solanum bul-
bocastanum is an Mi-1 gene homolog conferring broad-spectrum
late blight resistance in potato. Plant J. 44: 208–222.
Vos, P., G. Simons, T. Jesse, J. Wijbrandi, L. Heinen et al., 1998
tomatoMi-1gene confersresistanceto bothroot-knot nematodes
and potato aphids. Nat. Biotechnol. 16: 1365–1369.
Wang, Z. X., M. Yano, U. Yamanouchi, M. Iwamoto, L. Monna et al.,
1999The Pib gene for rice blast resistance belongs to the nucle-
otide binding and leucine-rich repeat class of plant disease resis-
tance genes. Plant J. 19: 55–64.
Wulff, B. B. H., C. M. Thomas, M. Smoker, M. Grant and J. D. G.
Jones, 2001Domain swapping and gene shuffling identify se-
quences required for induction of an Avr-dependent hypersensi-
tive response by the tomato Cf-4 and Cf-9 proteins. Plant Cell 13:
Yoshimura, S., U. Yamanouchi, Y. Katayose, S. Toki, Z. X. Wang
et al., 1998 Expression of Xa1, a bacterial blight-resistance gene
in rice, is induced by bacterial inoculation. Proc. Natl. Acad. Sci.
USA 95: 1663–1668.
Zhou, B., S. Qu, G. Liu, M. Dolan, H. Sakai et al., 2006
amino-acid differences within three leucine-rice repeats between
Pi2 and Piz-t resistance proteins determine the resistance speci-
ficity to Magnaporthe grisea. Mol. Plant-Microbe Interact. 11: 1216–
Fine mapping of
Evidence for a
Plant-pathogen arms races at
Avirulence genes and mech-
Communicating editor: A. H. Paterson
Rice Blast R Gene Pi36