Tải bản đầy đủ - 0 (trang)
Non-Transgenic TILLING, DEALING, and Deleteagene Approaches

Non-Transgenic TILLING, DEALING, and Deleteagene Approaches

Tải bản đầy đủ - 0trang


Mutagenesis and Functional Genomics in Cereals

Table 3 Important physical and chemical mutagenic agents and their mode of action

Mutagenic agents


Mode of action


(UV) rays




gamma rays



Fast neutrons



2-amino purine,







acridine orange



Ethyl methane

sulfonate (EMS)



Pyrimidine dimer

formation and

error during

DNA replication

Ionize water and

other organic


forming radicals

causing breaks in

DNA strands and

alterations in

purine and

pyrimidine bases



to DNA

Base pairing

mistake resulting

into A/T to G/C


Reduce fidelity of

DNA replication

by intercalating

between bases,



deletions or

additions that


induce frameshift


Modification of

guanine (G)

resulting into

G/C to A/T


Used in














Table 3


H. S. Balyan et al.


Mutagenic agents


Mode of action





Nitrous acid










Sodium azide






Modification of

thymidine (T)

resulting into

A/T to T/A


Modification of

cytosine (C)

resulting into

A/T to G/C


Modification of

cytosine (C)

resulting into

G/C to A/T


Produces inter and


DNA cross links;

cause large

deletions and

known as


Precise mechanism

of action


Produces inter and


DNA cross links

in conjunction

with UV; cause

large deletions

and known as


Used in










Not applied.

mutagens could be classified into those inducing point mutations (single

base pair change creating single nucleotide polymorphisms i.e., SNP), chromosomal aberrations, and sub-gene to gene level (kb) deletions; deletion

causing mutagens are popularly termed deletogens.

Mutagenesis and Functional Genomics in Cereals


Among chemical mutagens causing point mutations, ethyl methane sulfonate (EMS) has received a universal acceptance. EMS is considered the most

efficient mutagen for small genes (Greene et al., 2003). It induces high density of

G/C-to-A/T transitions with little or no chromosomal aneuploidy/aberrations, reduced fertility or dominant lethality (Ashburner, 1990). In a recent

reverse-genetics study involving 192 genes in an Arabidopsis population mutagenized by EMS, 99% induced changes were confirmed as G/C-to-A/T

transitions (Greene et al., 2003). In barley and rice also, nearly 70% transition

from G/C-to-A/T were noticed (Caldwell et al., 2004; Till et al., 2007). Until

now, most of the TILLING reverse-genetics resources involving detection of

SNP have been created using EMS (for details, see Section 3.1.3.). Only

recently, sodium azide (NaN3), an A/T-to-G/C transition inducing

chemical mutagen (Olsen et al., 1993) was used either alone for the creation

of barley TILLING populations (http://www.intl-pag.org/13/abstracts/

PAG13_P081.html) or in combination with methyl nitrosourea (MNU) in

rice (Till et al., 2007). However, a number of other point mutations inducing

chemical mutagens are available with the potential of producing different

constellations of mutations that lead to different projected codon usage and

phenotypes. These include the transversion inducing (A/T-to-T/A) N-ethylN-nitrosourea (ENU) (Stemple, 2004) and the transition inducing nitrous acid

(A/T-to-G/C), and hydroxylamine (C/G-to-T/A). Therefore, it would be

most appropriate to try and use other mutagenic agents with differing mechanisms besides EMS in the development of TILLING resources to maximize


With a view to induce deletion mutations in Caenorhabditis elegans, the

first organism used for DEALING studies, Jansen et al. (1997, 1999) and Liu

et al. (1999) used different chemical deletogens [ethylmethane sulfonate

(EMS), ethylnitrosourea (ENU), diepoxyoctane (DEO) and ultraviolet-activated trimethylpsoralen (UV-TMP)] as deletogens. The EMS and ENU

induced deletion mutations at comparable rate while DEO and UV-TMP

induced deletion mutations at a higher rate (see Liu et al., 1999). The deleted

segments varied from 700 to 2900 bp (average ~1400 bp) when screening was

performed in a 3 kb window (Liu et al., 1999). Most of these intragenic

deletions represented null or severe mutations resulting in phenotypes that

resembled wild-type or dramatic/lethal defects. Although, at present, extensive data on the efficiency and effectiveness of chemical deletogens is lacking in

crop plants, DEB and UV-TMP are being increasingly applied in cereal crop

plants. For instance, DEB has been used to induce mutations in the Xa21 gene

of rice changing disease resistance response against X. oryzae pv. oryzae (Wang

et al., 2004) and to create mutagenized population of indica rice cv. IR64 for

conducting reverse-genetics analysis at IRRI, Philippines (Wu et al., 2005).

DEB and UV-TMP are also being used for developing diploid wheat (Triticum

monococcum) and basmati rice reverse-genetics DEALING-resources through a

collaborative research project involving one University from the USA (led by


H. S. Balyan et al.

S. Kianian, North Dakota State University, Frago) and two institutions from

India (H. S. Balyan, Ch. Charan Singh University, Meerut and H. S. Dhaliwal,

IIT, Roorkee).

For a long time, physical mutagens such as fast neutrons have also been

shown to cause deletion mutations in plants. Fast-neutron treatments were

shown to induce mutations efficiently in A. thaliana (Koornneef et al., 1982).

Subsequent studies related with molecular characterization of Arabidopsis ga1-3

(Sun et al., 1992) and tomato Prf-3 (Salmeron et al., 1996) genes amply

demonstrated the ability of fast-neutron irradiation to induce deletion mutations. However, it is only as late as 2001 that the fast-neutron deletion

mutagenesis-based reverse-genetics methodology described as DeleteageneTM

was reported in model plants like Arabidopsis and rice for the first time (Li et al.,

2001, 2002). Fast-neutron and another important physical mutagen, gamma

rays, have been recently used to develop reverse-genetics resources in indica

rice cv. IR64 at IRRI, Philippines (Wu et al., 2005) and for forward-genetics

studies of wheat in Australia (Singh et al., 2006). However, data suggest that

these irradiation methods can produce large deletions (range of 100 kb) that

may be difficult to detect by simple PCR methods in non-sequenced genomes. We feel that a wider application of fast neutron based reverse-genetics

approach for discovering the biological functions of genes in a diverse set of

plant species including cereal crops will soon be a reality.

3.1.2. Mutagen treatment and population size

Most of the important cereal crop plants including wheat, barley, and rice

are self-fertilizing species while only a few species such as maize and rye are

cross-fertilizing. The mode of reproduction results in homozygosity and

homogeneity in the self-fertilizing species while heterozygosity and heterogeneity in cross-fertilizing species. Thus, the mode of reproduction and its

consequences have important bearing on the possible methods of mutagen

treatment and post-treatment handling of the population for detection of

induced mutations. In seeds, the embryo is already differentiated with one

to several genetically effective cells. Therefore, mutagen treatment of the

seed gives rise to chimeric M1 plants and segregation the mutations are

observed in M2 generation (progeny from self-fertilization of individual M1

plants) (Fig. 3A). In recent TILLING and DeleteageneTM studies involving

self-fertilizing species like Arabidopsis, wheat, rice, and barley, a similar

strategy was followed for the detection of base substitution and deletion

mutations (Caldwell et al., 2004; Li et al., 2001; McCallum et al., 2000a,b;

Slade et al., 2005; Till et al., 2003). In this scheme, DNA is isolated from

individual M2 plants from a family and pooled for subsequent mutation

detection. A part of the M3 seed harvested from M2 plants is stored for

future use, and the remaining seed may be used for rearing M3 progenies for

identification/confirmation of mutations and subsequent study. A similar



Seed treatment with chemical



Pollen treatment with mutagen

Raising of M1 population

(Plants will be +/+; +/−)

Pollination of mother variety plants with

mutagen treated pollens

Pollen culture

Raising of M2 population

Harvesting M1 seed and raising of M1

plants (Plants will be +/+ and +/−)

Raising of haploid (2n = 1x) plants

(Haploids will be + or −) and treatment

with colchicine

Isolation of DNA from individual plants

Raising doubled haploid (DH0) plants

(2n = 2x) (Plants will be +/+,−/−)

Individual M1

plant progenies

(Progenies will be

+/+, +/−, −/−)

M1 single seed

descent (SSD)


(Plants will be

+/+, +/−, −/−)

Preparation of 8-fold pools of DNA for

mutation detection by TILLING

Collection and storage of M3 seed;

use a portion of harvested seed for

raising M3 to identify and confirm

mutant progenies

Raising of M2 population; collection and

storage of M3 seed; use a portion of

harvested seed for raising M3 to identify

and confirm mutant progenies

Collection and storage of DH1 seed; use a

portion of harvested seed for raising DH1

to identify mutant progenies

Creation of database for data on

phenotypes and molecular mutations

In silico association of altered

phenotypes with mutations detected by


Figure 3

Flow chart showing steps involved in three different schemes of creation of mutagenized populations for TILLING.


H. S. Balyan et al.

strategy may be used with a cross-fertilizing species provided seed from

homozygous inbred lines instead of open pollinated source is used as the

starting material for mutagen treatment. Using seed from an open pollinated

source as the starting material for mutagenic treatment would complicate

matters related to the origin of mutations.

For generating TILLING resources in maize, a cross-fertilizing species,

Till et al. (2004a) recently used a modified strategy involving mutagen

treatment of the pollen rather than seeds (Fig. 3B). The treated pollens

were used to pollinate the donor or test-cross parent. The fusion of the

mutation-carrying male gamete with the wild-type female gamete would

produce M1 seed that would be heterozygous at the concerned locus. The

DNA from M1 plants is pooled for TILLING, and the selfed M2/M3 seed is

used for subsequent phenotyping of mutations. In the past, pollen mutagenesis coupled with pollen culture was recommended as an in vitro-mutagenesis

method for crop improvement (Castillo et al., 2001; Maheshwari et al., 1980;

Maluszynski et al., 1995, 2003; Taji et al., 2001; reviewed in Forster and

Thomas, 2005, and Szarejko and Forster, 2007). This approach so far has not

been utilized in experiments related to reverse-genetics analysis, though the

technique, once perfected, has tremendous potential as homozygous mutant

doubled haploid (DH) plants may be obtained in a single generation

(Fig. 3C), removing the complexity introduced by heterozygosity. As the

DHs would represent fixed inbred lines, molecular screening, phenotypic

analysis, identification of mutations as well as storage, and maintenance of

seed stocks would be a lot easier. However, the DH mutation population has

a disadvantage in that the lethal mutations will not be maintained and will be

lost. Such mutations hence will not be available to examine the basic

biological function of mutated genes conferring lethality.

While developing mutagenized populations for reverse-genetics analysis,

the determination of the optimum mutation frequency and thus appropriate

size of a suitable mutagenized population is crucial. The right balance

between the chances of detecting a mutation in the target gene and the

size of the population that can be handled with ease within the available

logistic limits, for a given mutagen treatment may be empirically determined. Mutagen treatment is usually applied in such a manner that it

produces sufficient lethality while allowing sufficient fertility, so that a

high frequency of induced mutations may be recovered in the mutagenized

population. To assess suitable mutagen dose, mutation frequencies in different plant species were indirectly estimated in the past on the basis of seedling

lethality, chlorophyll deficiency, seed set, etc. (Emery, 1960; Gilchrist and

Haughn, 2005; Kleinhofs et al., 1978; Li and Redei, 1969; Prina and Favert,

1983; Stemple, 2004). These are all phenotypic measures, which fail to

provide direct estimation of mutation frequency at molecular level, which

may be more relevant for the reverse-genetics experiments. Although precise estimates of induced global mutation frequency at the nucleotide level of


Mutagenesis and Functional Genomics in Cereals

a plant genome may be obtained directly through molecular analysis, routine

methods providing comprehensive estimates of mutation frequency are not

available at present. Though AFLP analysis (gain or loss of fragments), which

in a single 96-well experiment has the potential of screening 160 kb of

sequence has been tried to assess EMS, DEB, and gamma-irradiation induced

mutation frequency in barley, detailed results have not been reported (see

Waugh et al., 2006). Nucleotide level variation is a more desirable guide for

the quantitative measure of the mutation frequency to future reverse-genetics

studies in plants, but the available data should be treated with some degree of

caution as the estimates are based on limited experimental information.

Recent reports suggest that density of mutations range from 1/24 kb in

common wheat, a polyploid species, to 1/Mb in barley, a diploid species

(Table 4). In two other diploid species (maize and rice), density of mutations

was reported at 1/500 kb. In an Arabidopsis mutagenized population, mutation density of 1/235 kb or approximately 4 point mutations per eight-fold

pool (representing 768 plants = 96 wells  pooled DNA of 8 plants per

well) was reported (Greene et al., 2003). These data clearly support the long

held belief that polyploid species better tolerate mutational load than the

diploid species possibly due to their better genetic buffering capacity

afforded by multiple homoeoloci (Stadler, 1929). This would also mean

that to discover similar number of mutational events in diploids as in the

polyploids, one would probably need to analyze a much larger mutagenized

populations in diploid species.

Recent reports suggest working with relatively smaller M2 populations

of 8600 plants in barley (Caldwell et al., 2004). In case of maize, initially two

smaller TILLING populations were created and two more populations are

recently developed (see Weil and Monde, 2007). These two available

populations generated using EMS include one at University of Minnesota,

USA with 2370 mutant lines of B73, the inbred line used for maize genome

sequencing, and the other at Purdue University, USA that include 1276 line

of W22 inbred. The two other populations that have been created also

belong to B73 (at Iowa State University, USA) and W22 (at Purdue

University, USA) inbred lines. However, current efforts are on to develop

Table 4 Density of mutations determined in cereal crop species


Mutation density


Triticum aestivum

Triticum durum

Zea mays

Oryza sativa

Oryza sativa

Hordeum vulgare

1/24 kb

1/40 kb

1/500 kb

1/500 kb

1/497–1/500 kb


Slade et al. (2005)

Slade et al. (2005)

Till et al. (2004a,b)

Wu et al. (2005)

Till et al. (2007)

Caldwell et al. (2004)


H. S. Balyan et al.

much larger barley-TILLING resources following treatment of 60,000

seeds with NaN3 (S. Salvi, University of Bologna, Italy; Ist International

GABI-TILL Workshop, IPK, Gatersleben, Germany) and 80,000 seeds with

EMS (Nils Stein, IPK, Gatersleben, Germany; Ist International GABI-TILL

Workshop, IPK, Gatersleben, Germany). We believe that the remaining diploid cereal species also require large working M2 populations,

which would become available in due course. However, for genome-wide

Arabidopsis TILLING Project (ATP), McCallum et al. (2000a) suggested that

10,000 M2 plants would be sufficient for obtaining the desired mutation

using just a single primer pair per gene. Considering the low rate of mutation induction by deletogens and fast neutron, the required population size

has been estimated to be in excess of 100,000 plants for comprehensive

coverage (>95% probability of detecting a mutation in a target gene) if

the mutation rate is considered independent of the genome size (Li et al.,

2001, 2002). At IRRI, Philippines, 60,000 IR64 mutants have already been

generated using both physical and chemical mutagens and 38,000 independent

lines have been advanced to M4 generation for evaluation (Wu et al., 2005).

To allow availability of mutants, a mutant database has been created in International Rice Information system (IRIS; http://www.iris.irri.org).

3.1.3. Preparation of DNA pools

Creation of appropriate size DNA pools for high-throughput detection of

induced mutations is crucial for functional analysis of a gene through

reverse-genetics methods. Purity of the DNA preparation is also critical

for obtaining clean gels free from backgrounds. The average size of good

quality DNA should be at least 15 kb and it should be stable under standard

storage conditions (Comai and Henikoff, 2006). DNA is isolated from the

M2 generation of the seed treated populations and from M1 generation of

the pollen treated populations (Fig. 3A–C). DNA from individual plants is

pooled at the same concentration to achieve balanced representation. The

Seattle TILLING Project (STP) or formerly the ATP, which help establish

Maize TILLING Project at Purdue University (http://genome.purdue.edu/

maizetilling) ( Weil and Monde, 2007) routinely uses eight-fold pools that

are arranged in the bi-dimensional scheme so that the display of two

mutations in the same pools identifies the mutant individual. To improve

the high-throughput efficiency of TILLING for the identification of sugar

and starch mutants in Arabidopsis, a modified pooling strategy involving

64-fold super pools followed by 8 Â 8 sub-pools is being used (G. Strompen

and J. Lunn, Ist International GABI-TILL Workshop, IPK, Gatersleben,

Germany, 2006). Although desirable for high-throughput analysis, success

in detecting mutations in deeper pools of the large genome cereal crop

species is awaited. In C. elegans adduct lesion experiments 96-well plate

pools of DNA were used (Liu et al., 1999). In DEALING experiments on

diploid wheat (T. monococcum), several pooling strategies have demonstrated

Mutagenesis and Functional Genomics in Cereals


effectiveness for screening of 1000-fold pools (S. Kianian, unpublished data)

and a poison–primer approach, utilized in C. elegans for detecting small deletions, for screening at 5000-fold pools (Fig. 4, data from Riera-Lizara-zu). It is

apparent that the latter strategy of using large DNA pool size will allow

handling of large DEALING/ DeleteageneTM resource populations that will

compensate for the lower frequency of induced deletion mutations by fast

neutron or deletogens. Similarly, large DNA pools can also be applied to

DeleteageneTM but strategy for detection may have to be altered to account

for larger size deletions generated by agents.

3.2. Mutation detection technique in TILLING

For the first time, TILLING, as a reverse-genetics tool was used in Arabidopsis by McCallum et al. (2000a,b) and in Drosophila melanogaster by Bentley

et al. (2000). TILLING involves PCR amplification of DNA pools from

EMS mutagenized populations (Fig. 3) using gene-specific primers and

targeted identification of induced point mutations (base substitutions/

SNPs) in the gene of interest. The induced point mutations can lead to a

broader range of effects, including hypomorphic, hypermorphic, and neomorphic effects (Stemple, 2004). The nature of the induced lesion has also

implications for the experimental strategy to identify the type and genomic

location of the mutation. In the original TILLING experiments, McCallum

et al. (2000a,b) made use of denaturing high-performance liquid chromatography (DHPLC) to detect the base pair changes by heteroduplex analysis.

However, the need for newer mutation detection methods was felt as

DHPLC is unsuitable for high-throughput analysis and fails to provide the

location of the detected mutation without sequencing of the

whole amplified fragment. To overcome these drawbacks, new TILLING

protocol involving CELI enzymatic mismatch cleavage of heteroduplex

DNA (Oleykowski et al., 1998) on a LI-COR gel analyzer system (Lincoln,

NE; Middendorf et al., 1992) was reported by Colbert et al. (2001). This

protocol was first used by the STP for screening EMS mutagenized

Arabidopsis populations (Till et al., 2003). The CELI endonuclease, which

belong to the S1 nuclease family of single strand-specific (sss) nucleases, is

found in celery, Apium graveolens, (celery enzyme is known as CELI) and

many other plant species (Oleykowski et al., 1998). CELI cleaves DNA to

the 30 side of the mismatches and loops out in heteroduplexes between wildtype and mutant DNA, leaving duplexes intact. In this new method,

the PCR products amplified from pooled DNA samples using different

IRD700 and IRD800 labeled-primers are first incubated with CELI and

then the cleaved products are run and visualized on LI-COR gels. At

present, both home-made CELI enzyme preparations and commercial

mutation detection kits (SURVEYOR Mutation Detection Kit, Transgenomic) that uses CELI are being used (Qiu et al., 2004). In a few protocols,


H. S. Balyan et al.

800 bp




193 bp


Wild type



403 bp

1000 bp







Wild type DNA + + − −

Deletion DNA − − + +

Competing oligo + − + −

1000 bp

800 bp

600 bp









Mutant: wild type ratios




+ + + ++ + + + +


+ + +

+ + + ++ + + + +


+ + +


− +






Full length

primers and

poison primer


775 bp

300 bp



Figure 4 (A) Schematic representation of the wild-type and mutant barley (Hordeum

vulgare) GBSS I (waxy) structural gene. Exons are represented by filled boxes. The

primers used for PCR are represented by arrows. (B) Detection of the barley waxy

deletion using nested PCR with a poison primer. (A) First amplification step of wildtype, mutant (with the waxy deletion) and various dilutions of mutant to wild-type

DNA with the A and C external primers with and without the competing or poison

primer B. With no poison primer (only primers A and C), the wild-type DNA produces

a 1000-bp product while the deletion or mutant DNA produces an 800-bp fragment.

With the poison primer (primers A, B, and C), the wild-type DNA produces a 600-bp

product that results from the preferential amplification of DNA by the poison primer B

and the external primer C. With the poison primer mutant DNA still produces the

expected 800-bp product because there is no binding site for the poison primer B. With

the exception of the 1:1 dilution of mutant to wild-type DNA, poison primer B also

outcompetes the external A primer in the various dilutions. (B) Second amplification

Mutagenesis and Functional Genomics in Cereals


CELI has been replaced by mung bean nuclease. Recently, Endo-1, a

genetically engineered endonuclease, manufactured by Serial Genetics,

France (www.serialgenetics.com) has also been used as an alternative to

CELI for mutation detection following heteroduplex mismatch cleavage in

several plant species (A. Bendahmane, Ist International GABI-TILL Workshop, IPK, Gaterlseben, Germany). Endo-1 has the advantage of recognizing

all the different types of mismatches including base-substitutions and

INDELs with similar efficiency. The CELI/Endo-1 endonuclease-based

mutation detection assay has several advantages including simplicity of the

assay, home-made/commercial availability of enzymes, and insensitivity to

the location and types of different mismatches (see Yeung et al., 2005).

Protocols for high-throughput TILLING and detection of mutations have

been standardized. For example, a single TILLING run interrogates a total of

~750,000 base pairs (eightfold pooling  1000 bp  96 lanes = 768,000 bp)

and detects mutations equivalent to ~3000 sequencing lanes (assuming that a

1 kb fragment requires some four full sequencing lanes with four primers for

reliable detection of heterozygotes). The above TILLING protocol including the identification of the mutant in each pool and its sequencing is

estimated to be an order of magnitude more economical than full sequencing

(Henikoff and Comai, 2003). A summary of the currently available and

emerging technologies (modified following Comai and Henikoff, 2006) for

high-throughput discovery of mutations and polymorphism is given in

Table 5. It may be realized that all these technologies require very expensive

equipments where easy access may not be available to all research laboratories.

Following discovery of mutation(s) in a pool, DNA from the individual

plants of the pool is re-screened, and the mutant plant and the approximate

position of the mutation are identified. To complete the analysis, sequencing

of the amplified target from the mutant plant is carried out to identify the base

pair change (stop codon or mis-sense mutation) and consequent changed

phenotype. Freely accessible web-based software to aid TILLING has also

been developed. One of these software tools is CODDLE (for Codon Optimized to Detect Deleterious LEsions; http://www.proweb.org/cod/dle/).

CODDLE help in choosing the best region of the target gene and the optimal

primers to amplify PCR products of up to 1.5 kb that is suitable for CELIbased mismatch–cleavage method for TILLING (Till et al., 2003). Although,

CODDLE has served well with small genome model plant Arabidopsis, newer

procedures and modified software may need to be developed to carry out the

step of DNA from the first PCR step using the D and E internal primers. The wild-type

template produces a 775-bp product whether or not the poison primer was used in the

first PCR step. Similarly, the mutant template produces a 300-bp product with or

without the use of the poison primer. Without the poison primer, we begin to lose

the ability to detect the deletion at mutant: wild-type ratios of 1:100. With the poison

primer, the deletion can be detected at a ratio of 1:5000.

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Non-Transgenic TILLING, DEALING, and Deleteagene Approaches

Tải bản đầy đủ ngay(0 tr)