Genome Mapping and Molecular Breeding in Cassava
Y .Lokko1, E. Okogbenin1, C.Mba2, A. Dixon1, A. Raji1 and M. Fregene3
1. International Institute of Tropical Agriculture (IITA), Oyo Road PMB 5320,
Ibadan, Nigeria. www.iita.org
2. Plant Breeding Unit, Joint FAO/IAEA Agriculture and Biotechnology Laboratory,
International Atomic Energy Agency Wagramer Strasse 5, P.O. Box 100, A-1400
Vienna AUSTRIA www.iaea.org
3. Centro Internacional de Agricultura Tropical, Apartado Aéreo 6713, Cali
Columbia, www.ciat.cgiar.org
Formatted: Spanish (Spain-Modern
Sort)
I. Introduction
Brief History of Crop
Cassava (Manihot esculenta Crantz) belongs to the family Euphorbiaceae, which are
characterized by lactiferous vessels composed of secretory cells. Its relatives in the
euphorbiaceae family include several commercially important plants, such as rubber
trees (Hevea brasiliensis), castor oil plants (Ricinus comunis) and ornamental plants
(Euphorbia spp.). Cassava it’s believed to have originated by hybridization between
two wild Manihot species, followed by vegetative reproduction of the hybrid.
The center of origin of cassava was first reported to be Central America including
Colombia, Venezuela, Guatemala and Southern Mexico due to the large number of
varieties present there (Sauer (1952; Roger 1965). This was later referred to as the
“minor” center of origin, and Brazil as the “major” center of origin (Ekanayake et al
1997). Furthermore, Manihot species found in the Central America region are only
distantly related to cassava (Fregene et. al. 1994;Schaal et. al. 1994; Roa et. al. 1997).
Schaal and Oslen, (1999), investigated the crop’s domestication based on haplotypes
of the single-copy nuclear gene glyceraldehyde 3-phosphate dehydrogenase (G3pdh),
in cassava and its relatives, which confirmed that cassava was likely domesticated
from wild M. esculenta, mainly from populations of subspecies flabellifolia along the
southern border of the Amazon basin.
Cassava grows in tropical and subtropics and was introduced into Africa and Asia by
Portuguese travelers in the 15th century (Jennings and Hershey, 1985; Allem, 1994).
Botanical Descriptors
Cassava, Manihot esculenta Crantz, is a perennial woody shrub with edible tuberous
root. It is also called yuca, (Spanish), manioc (French), and mandioca (Portugues).
The height of a mature cassava plant usually ranges from 1 to 2 m, although some
cultivars reach 4 m. Branching height can be as low as 20 cm, while some varieties
never branch and, as a result, never flower.
Cassava leaves are simple, with 3 and 9 leaf lobes (usually odd numbers) arranged
spirally around the stem. It is monoecious: male and female flowers are located on the
same plant. Some varieties flower frequently and regularly, while others flower rarely
or not at all. Flower production is important for breeding. Environmental factors, such
as photoperiod and temperature, influence flowering. Male flowers develop near the
tip, while female flowers develop closer to the base of the inflorescence (Ekanayake
et al 1997).
Cassava is generally propagated with stem cuttings, thereby maintaining a genotype.
Under natural conditions as well as in plant breeding, propagation by seed is common
and farmers in African are known to occasional use of spontaneous seedlings for
subsequent planting (Silvestre and Arraudeaus, 1988).
Cassava plant propagated from stem cuttings produces adventitious roots at the base
of the cuttings within a week, while Cassava plants propagated from seeds first
develop a tap root system. The adventitious roots develop into a fibrous root system
which within 30-60 days increase in diameter and become tuberous roots (also called
'storage roots’). These swell with time due to starch accumulation (Ekanayake et al
1997).
Cassava generally has a diploid genome (2n=36). However, some authors have
described it as a segmental allotetraploid with basic chromosome number x=9. Jos
and Nair, (1979) however, conducted studies on the meiotic behaviour of several
cassava genotypes and observed regular 18 bivalent formation of the chromosomes
typical of its diploid (2n=2x=36 chromosomes). Furthermore, the breeding behaviour
of allotetraploids is similar to that of diploid (Wricke and Weber, 1986). Although
most cassava genotypes studied are diploid, spontaneous polyploid such as triploids
(3n) and tetraploids (4n) of some genotypes have been reported (IITA 1980; Hahn et
al., 1980). Triploid and tetraploid plants differ from diploid plants in plant vigor and
leaf shape and size. Triploid plants usually grow and yield better than tetraploid and
diploid plants. The nucleic acid content of diploid cassava is 1.67 picogram per
nucleus that is 772 mega base pairs in the haploid genome (Awoleye et al., 1994).
The heterozygosity levels of different cassava accession have been further confirmed
from diversity studies using isozymes (Lefevere and Charrier, 1993), AFLP and SSRs
(Fregene et al 2000; Fregene et al 2003; Sanchez, et al 1999).
Economic Importance
Cassava accounts for approximately one-third of the total staple food production and
provides over 50% of the energy for more than 200 million people in sub-Saharan
Africa (IITA 1992). In Central Africa alone, cassava is estimated to provide over
1000 Kcal per day to 30 million people (Cock 1985). The storage roots form the basic
carbohydrate component of the diet and the leaves, which contain appreciable
amounts of vitamins, minerals and proteins are consumed as a preferred green
vegetable in many parts of Africa, providing protein, mineral and vitamins (Hahn
1989; Lancaster and Brooks, 1983).
The crop is grown almost exclusively as food in 39 African countries stretching
through a wide belt from the Island of Madagascar in the southeast to Senegal and the
Cape Verde Islands in the north-west (Hahn and Keyser 1985). Production is in areas
where the rainfall exceeds 600 mm over a period of at least 2 to 3 months and
altitudes range from sea level to 1800m. It is particularly important in those areas
where food supply is constantly threatened by environmental constraints such as
drought and pest outbreaks because of its ability to grow under conditions considered
as sub optimal for the majority of food crops. It can be harvested anytime from 6 to
24 months after planting, and can be left in the ground as a food reserve for
household food security in times of famine, drought and war (FAO, 2000; Best and
Henry 1992; Cock, 1985).
Global cassava production projected for 2004 is expected to be around 192 million
tonnes as was recorded in 2003, with Africa contributing 103 million (FAO, 2004).
Classical breeding Objectives
Breeding for increased yield, multiple pest and disease resistance, desirable
agronomic and consumer preference traits such as early vigor in plant growth (for
high foliage yield for leaf vegetable), appropriate plant architecture, and early bulking
of storage roots, combined with high dry matter content, low cyanide content, and
other favorable traits, e.g., easy peeling, have been the main breeding objectives.
Recently breeding for improved micronutrient content is being emphasized. Broadbased populations for different agroecologies are developed through recurrent
selection (intra and interspecific population improvement techniques) and
backcrossing as well as multiple crossing schemes to achieve. Superior individuals
are selected as parents, segregating families are generated by multiple crossing among
these elite clones complementing one another for various agronomic, consumer
quality and major pest resistance traits. The breeding values of the parents are
evaluated through progeny testing in seedling nurseries. Based on evaluations,
selected parental clones or half sib progenies are hybridized for further improvement
in a recurrent selection scheme. Backcrossing has also been a useful procedure for the
transfer of resistance into elite populations by providing resistant lines quickly to
prevent the severe infestation of relevant pests.
Classical breeding Achievements
Achievement in cassava breeding in Africa is in the development of a range of elite
genotypes , such as TMS 30572 and TMS 4(2)1425that combine high stable yields,
agronomy and consumer quality with acceptable levels of resistance to CMD and
CBB which are widely cultivated In Africa ( Hahn et al 1989). The adoption of some
of these multiple pest and disease resistance genotypes in Uganda for instance,
curtailed the devastating effect of the CMD on cassava production in that country
(Otim Nape et al., 1994).
The introduction of Latin American gemplasm into the breeding programs in Africa
has resulted in significant broadening of the genetic base of cassava in Africa.
Classical breeding has contributed to improved dry matter and starch content as well
as reducing the cyanogenic potential in cassava. ( could you fill in with some eg from
CIAT). An IFAD-financed research project executed by the National Center for
Research on Cassava and Tropical Fruits (CNMPF, the Brazilian acronym), Cruz das
Almas, CIAT and IITA beginning 1991 led to the collection and evaluation at
representative semi-arid sites of cassava germplasm in Northeast Brazil. Genotypes
with high efficiency for extraction of limited soil water, low levels of water loss
through transpiration, and resistance to severe mites attack were selected and
recombined by genetic crosses to form the basis of a breeding population for the semiarid agroecosystem (CIAT 1996). Subsequently, a farmer participatory plant breeding
(PPB) scheme was employed to evaluate recombinant progenies and select several
varieties adapted to the semi-arid region of Northeastern Brazil (Fukuda and Saad
2001). Improved cassava varieties developed in the project had 25-100% yield
increases over traditional cultivars in 41 communities of the states of Bahia and
Pernambuco of Northeast Brazil (CIAT and CNMPF 1999).
Recombinant seeds from the above breeding population for semi-arid ecosystems were
also introduced to Sub Saharan Africa (hereafter Africa) through IITA and evaluated in
the drier areas of Northern Nigeria. Evaluations revealed the potential of the
germplasm to increase productivity in the dry and hot conditions of Northern Nigeria
(IITA 1999). However, the presence of a major biotic constraint the cassava mosaic
disease (CMD) in Africa, a disease that is not found in Latin America, limited
immediate use of the germplasm but required introgression of CMD resistance into the
Latin American germplasm. Resistance to CMD was introduced by backcrosses and
thousands of recombinant seeds were produced and distributed to participating
countries in West and Central Africa. Evaluation under typical farmer conditions and
selection has lead to the selection of 25, 32, 10, 28 and 122 improved genotypes that
combine CMD resistance and adaptation to water stress for the semi-arid areas of
Burkina Faso, Chad, Ghana, Niger and Nigeria respectively (IITA 2000a). The
improved germplasm has extended considerably the range of cultivation of cassava
beyond its traditional area in the humid and sub-humid tropics into the semi-arid zone
of West and Central Africa by more than 100,000 hectares between 1989 and 1999
(IITA 2000b).
Limitations of classical breeding and utility of molecular mapping
Most traits studied in cassava are polygenic (Bryne, 1987; Rajendran, 1989; Amma and
Sheela 1995; Hahn et al. 1989). Variation in polygenic traits is attributed to quantitative
trait loci, QTL’s. Quantitave traits in plants and animals are studied using a variety of
genetic models and designs incluing the analysis of mating designs in segregating
populations, estimate effective factors using biometrical estimating techniques or
molecular markers or biometrical methods can be used to estimate effective factors for
quantitative traits (Lynch and Walsh, 1998; Zeng, et al., 1990). Most of the biometrical
estimating techniques do not estimate the exact number of genes but rather the number
of effective factors. Furthermore the detection of multiple genes for a trait using
segregation analysis alone is not efficient because of the differences due to genotype by
environment interaction. Molecular markers, which are not affected by environmental
conditions and are insensitive to gene interactions, are suitable for such studies.
The two main strategies used to identify molecular markers associated with traits of
interest are genetic linkage mapping and bulk segregant analysis (BSA) (Tanksley et al.,
1989; Michelmore et al., 1991; Giovannoni et al., 1991).
Molecular markers, which include biochemical (isozymes and storage proteins) and
DNA markers, exist in every genotype. They occur in large numbers and their
expression is independent of phenotypic value and as such are powerful tools of genetic
research (Soller and Beckmann, 1986; Beckmann and Soller, 1983). They allow
geneticist and plant breeders to locate and follow the numerous interacting genes that
determine a complex trait as well as tagging those controlled by single genes. Molecular
markers offer a spectacular improvement in the efficiency and sophistication of plant
breeding. They are i) developmentally stable, ii) detectable in all tissues, iii) unaffected
by environmental conditions, and iv) virtually insensitive to epistatic or pleiotropic
effects. They provide a choice of codominant markers which can be identified in
heterozygotes or dominant marker which are identified as present or absent subclasses
(Botstein et al., 1980; Williams et al., 1990). Molecular markers can provide. It is now
generally accepted that markers represent the most significant advance in breeding
technology that has occurred in the last few decades and are currently the most important
application of molecular biology to plant breeding.
Molecular markers have made an immense contribution to cassava breeding and
genetics, in the assessment of genetic diversity, taxonomical studies, understanding
the phylogenetic relationships in the genus, confirmation of ploidy and in the
development of genetic maps (Fregene et al., 1997; Fregene 2001; Lokko et al 2004).
II. Construction of genetic maps:
Brief history of mapping efforts
Genetic mapping uses genetic techniques such as crossbreeding and pedigree analysis to
construct genomic maps. Genetic linkage maps provide a direct method for selecting
genes via their easily detectable markers, a guide for sequencing experiments by
showing the positions of genes and other sequence features. Linkage maps are based on
recombination frequencies and the distance between points on a genetic map is a
reflection of the recombination frequencies between the points.
The first genetic maps constructed in an organism was in the fruit fly Drosophilia which
used genes which specified distinct phenotypes such as eye colour and wing shape as
markers (Sturtevant, 1931). In plants genetic maps of some crops species such as maize
and tomato were also among the earliest genetic linkage maps to be constructed
(Emerson et al 1935: MacAurther, 1934). These were based on morphorlogical
characters such as dwarfism, albinism and alter leaf morphology.
The first DNA marker based genetic linkage maps in plants were Brassica were
published in the late 1998’s to the early 1990 (Table1). These were mainly based on
RFLP markers, which led to the rapid construction of linkage maps in many plant
species and the placement of genes that control qualitative traits on these maps. In
cassava RFLP markers were also the first set of DNA markers to be used in mapping.
Table1 below
First generation maps constructed in cassava
Angel et al. (1993), initiated work on genome mapping in cassava with an investigation
on polymorphism in a range of cassava accessions and a wild relative with different
random genomic clones and restriction enzymes. They concluded that a combined use of
RFLP and RAPDs markers would lead to the construction of detailed map of cassava
and Gomez et al (1996) subjected 328 RAPD markers to linkage analysis in cassava.
Following this, the first cassava linkage map to be developed was published by Fregene
et al, (1997).
The map was based on an F1 population of two geographic divergent parents. The
female parent TMS I30572, with resistance to CMD, which was derived through
introgression from M. glazovii while, the South American male, parent CM 2177-2
(ICA-Cebucan) with no resistance to CMD. A hundred of 150 individuals were available
in the cross however a subset of 90 individuals were used to develop the first map.
In a heterozygous species, the segregating F1 is used (Lui, 1998; Pillay and Kenny, 1996;
Fregene et al., 1997). They scored single-dose restriction fragments (SDRFs), for the
linkage analysis and developed two separate linkage maps based on male and female
parents. SDRFs are DNA markers that are present in one parent and absent in the other
and segregate in a 1:1 ratio (absence: presence) in the progeny. They represent the
segregation equivalent of an allele at a heterozygous locus in a diploid or an
allopolyploid genome or a simplex allele in an autopolyploid and are suitable for linkage
analysis in an F1 population when there is the presence of a number of unique
segregating polymorphisms (heterozygosity) and normal meiosis in either or both
parents in mapping polyploid genomes (Wu et al., 1992; Williams, 1998).
One hundred and fifty-eight RFLP, 30 RAPD, 3 microsatellite, and 4 isozyme singledose markers, donated by the female parent of mapping population, were tested for
linkage using the MAPMAKER computer package (Lander et al. 1987). Map units (in
centiMorgans, cM) were derived using the Kosambi function (Kosambi 1944).
Maximum likelihood orders of markers were verified by the ‘‘ripple’’ function, and
markers with LOD value ≥2.0 were assigned to the framework map. Markers that
could not be placed with LOD≥2.0 were added to the map in the most likely interval
between framework markers. One hundred and thirty-two RFLP, the 30 RAPD and 3
microsatellite, and 3 isozyme loci defined 20 linkage groups spanning 931.6 cM, with
an average marker density of 1 marker every 8 cM (Fig. 1); 26 RFLP markers and 1
isozyme marker remained unlinked. The linkage groups were named alphabetically
until they can be correlated to earlier named chromosome karyotypes (Magoon et al.
1969). The most densely populated linkage group (D) spanned 51.2 cM, with 26
markers, while the least populated group (I), also the longest group, had 8 markers
spanning 80.6 cM. This wide range of marker density indicated differing degrees of
saturation of linkage groups with markers.
In Fig. 4, 139 (84%) attached to linkage groups by horizontal bars make up the LOD
2.0 framework map of cassava, the remaining 29 markers (in parenthesis) are placed
in the most likely intervals between framework markers (LOD≥2.0).
They estimated the length of the cassava genome based on the female map to be 1,
610 cM, covering about 60% of the cassava genome. This estimate was based on
probability that a randomly chosen pair of loci will lie within xcM of each other is
approximately 2x/G; where x is assumed to be small compared to the mean genetic
length of the chromosome (Hulbert et al. 1988).
The male derived map consisted of 107 RFLP, 50 RAPD, 1 microsatellite, and 1
isozyme single-dose markers, in 24 linkage groups with a total distance of 1, 220 cM.
The genome length was estimated 2, 010 cM. Intervals were observed to be larger in the
male-derived map than in the female-derived map and a paired t-test of showed
significantly (P(0.01) greater distances in the male-derived map, suggesting a reduced
recombination rate in gametes of the female parent. The mean interval length between
adjacent allelic bridges (markers common to both parents) in the female-derived map
was 38% less than in the male derived map.
Second generation maps constructed in cassava
In an attempt to increase the marker density on the map the next strategy was to
develop simple-sequence repeat (SSR) markers, markers that are polymerase chain
reaction (PCR)-based and highly polymorphic and map them on to the exiting map
and new maps to be developed. Mba et al., 2001, developed and characterised 172
SSR markers to saturate the existing linkage map. The markers were screening in 150
progenies of the mapping population TMS I30572 X CM 2177-2. Thirty six markers
were placed on the map and were evenly spread over the linkages groups. A higher
number of “allelic bridges” which are required for rigorous marker-assisted
quantitative genetic analysis in F1 progeny from non-inbred parents were obtained
using the SSR markers (30%, as against 10% obtained with RFLPs). Currently, 80
RAPD markers 239 RFLP markers and six EST’s have been placed on the cassava
genetic map (Fregene et al., 2001).
Map Coverage Now?????
The total number of RFLP, RAPD, and SSR markers on the cassava map is
830 markers, SSR markers make up roughly 25% of markers on the map.
Genetic mapping in a full-sib cross from non-inbred parents involves linkage
analysis of markers segregating independently in the female and male gametes
and leads to 2 independent maps (Fig???). Analogous groups can be identified
amongst the female- and male-derived linkage groups via markers that are
unique in both maps, also known as allelic bridges”. For the first time, 18
analogous linkage groups that may represent the 18 haploid chromosomes of
cassava have been identified. Previous mapping efforts with only RFLP
markers failed to resolve this problem due to insufficient allelic bridge
markers 27% of all markers compared to 30% with SSR markers. Efforts are
currently ongoing to develop a single map using the JOIN MAP (Stam 1993)
program. Nevertheless, the linkage map is not saturated as 22 and 40 markers
remain unlinked as well as the presence of 2 non-analogous linkage groups,
one from each parent.
With over 1000 cassava SSR primers, developed (Chavariagga et. al. 1998; Mba et.
al. 2001; CIAT 2003; Lopez and Verdier 2004, unpublished data), and the suitability
of AFLP marker technology to non-radioactive detection, efforts were made to
develop linkage maps of cassava from diverse genetic backgrounds based on simple,
PCR based techniques. Three genetic linkage mapping studies were initiated at IITA
specifically to map the genes conferring resistance to cassava mosaic disease (CMD),
develop molecular markers for CMD resistance in African cassava germplasm and to
Deleted: 300
assist the incorporation and enhancement of resistance to this disease in African,
Latin American and Asian cassava gene pools (IITA, XXX; Akano et al 1998). The
F1 populations were generated from crosses between the CMD resistant breeders line
TMS I30572 and a Nigerian landrace TME117 (local name Isunikankiyan); crosses
between the breeders line TMSI30555 with moderate susceptibility to CMD and
CMD resistant landraces TME3 (2nd Agric) and TME7 (Oko Iyawo) (Akano et al
1998; Lokko, et al., 2001;).
One hundred and thirty two genotypes each of populations TMS I30572 X TME117 and
TMS I30555 X TME3 were used for linkage map construction. A total of 125 markers
from 85 SSR primers were generated in for were subjected to linkage analysis in the
populations TMS I30572 X TME117. Sixty-two markers mapped in 19 linkage groups
(LGs), which covered 816.6 cM of the genome.
In the population TMS I30555 X TME3, 66 markers from 47 SSR primers and 85 AFLP
markers were subjected to linkage analysis also mapped in 19 LGs (Fig 3). Thirty-three
out of the 85 AFLP’s markers and 37 out of the 66 SSR’s, mapped in 19 LGs using a
LOD of 3.0 and recombination fraction of 0.17 to declare linkage. The map covered 788
cM. Generally in population C, the SSR and AFLP markers clustered in specific groups,
nine groups were made up of ALFP markers only, two groups had both AFLP and SSR
markers and the remaining eight were made up of SSR markers only.
Sixty-nine F1 s of TMS I30555 x TME7 were screened with 15 SSR primers and 11
AFLP markers and the data subjected to linkage analysis with Join Map (Van Ooijen et
al 2001). Thirty-six <aaxab> and 41 <abxaa> markers were scored from AFLP markers,
while five <aaxab>, three <abxaa> and five <abxac> markers were scored for the SSR
analysis giving a total of 90 markers. The chi square analysis revealed that 34 (37.78%)
<aaxab> markers , 37 (41.11%) markers and five (5.55%) <abxac> markers were
significant for the1:1 segregation ratio while 10 (11.11%) <abxaa> markers and 4
(4.44%) <aaxab> markers were significant for 3:1 segregation ratio. Forty-five out of the
90 markers mapped in 15 linkage groups with a total length of 812.32 cM of the
genome.
Despite the low saturation of the maps, the loci were randomly distributed
over linkage groups and the information from it and the map itself can be
utilised in cassava genetics. For instances, the markers could also be applied in
studies with similar cassava populations. AFLP primer pair E-ACT/M-CAG
which contributed markers in four linkage groups in the TMS I30555 X TME7
map, could be used in experiments such as in diversity studies, which require
coverage of most of the genome. The maps could also be used in linkage map
pooling with the TMS I30572 X CM 2177-2 map.
[Representative maps in figures??]
III Gene Mapping
The cassava genetic linkage map developed at CIAT currently has five known genes
including two CMD resistance genes CMD1 and CMD2 placed on it and a number of
quantitative trait loci (QTL) associated with some linkage groups (Fregene et al., 2001).
To date two CMD resistance genes CMD1 and CMD2 have been placed on the map
(Akano; et al., 2002; Fregene, 2001).
Bulk segregant analysis (BSA) was used to screen several SSR and AFLP markers in the
mapping populations TMS I30555 X TME3 and TMS I30555 x TME7 and to map
genes for resistance to CMD (Akano et al 2001; Lokko et al 2001). The populations
were characterised in multiple environments and the CMD status of each genotype used
to classify them into highly resistant, resistant susceptible and highly susceptible
individuals. In TMS I30555 X TME3, the CMD resistant and CMD susceptible bulks
were made up of 40 individuals. In TMS I30555 x TME7 10 individuals made up either
DNA bulk. In the two studies, the SSR marker SSRY28 was tightly linked to CMD
resistance (Fig XX).
Akano et al., (2002) identified a dominant cassava resistance gene CMD2 on linkage
group R of the male cassava framework map 8 cM from the marker SSRY28.
Marker-trait association detected by regression analysis of TMS I30555 x TME7 showed
that the marker, which accounted for 57.41% of total phenotypic variation for resistance.
The analysis further showed that another SSR marker, SSRY106 and AFLP marker EACC/M-CTC-225, accounted for 35.59% and 22.5% of the total phenotypic variation
for resistance respectively. Correlation coefficients for SSRY28-180 and E-ACC/MCTC-225 with CMD responses were significant and negative while correlation between
SSRY106-270 and the CMD responses were significant and positive. This suggests that
the CMD resistance genes associated with E-ACC/M-CTC-225 and SRY28-180 are
different from genes associated with SSRY106-270. Interval mapping analysis further
revealed the large QTL regions between SSRY28-180 and E-ACC/M-CTC-225
covering about 50% of the length of the linkage group and 3% of the length of genome
covered, which explained most of the phenotypic variation.
Later, the group at CIAT identified NS158 to be tightly linked to CMD resistance.
Representative gel pictures??
CMD Resistance F1 progeny CMD Susceptible F1 progeny
RP/SP/RB/SB
200bp
175bp
150bp
IV QTLs detected
QTLs associated with dry matter yield and dry matter percentage have been linked to
linkage group D of the cassava framework map (Fregene et al, 2001).
V Marker Asisted Breeding
Tools of genomics particularly molecular markers permit an immediate
increase of efficiency in cassava breeding pending when inbred lines will be
developed. Molecular markers are particularly useful for:
1. Using a large of progenitors, especially when no information on
combining ability is available, in breeding schemes and the subsequent
elimination of a large part of the resulting progeny very early in the
breeding cycle.
2. Increasing the number of replications and environments for testing
breeding lines (thus achieving a more accurate performance of the
genotypes), by working with reduced progeny sizes.
3. Increasing heritability by the elimination of the confounding effect of
the environment, this is particularly true in cassava where the first
stage of selection is conducted on a single plant per genotype or
low/absence of pathogen/pest pressure in initial sites of selection
(nurseries), or a trait that has a large environmental influence.
Formatted: Bullets and Numbering
4. The pyramiding of different sources of resistance against a single
pathogen or the combination of resistance to several pests and diseases
and other traits without resort to time consuming field evaluations, also
known as ‘express breeding’.
The success of conventional cassava breeding is inversely proportional to the
number of traits it attempts to improve. Combining many genes controlling
quantitative traits from diverse sources into a single variety is a long-term, highrisk venture that requires every available tool for success. MAS can be used to
achieve these goals more efficiently. For example, CMD-resistant donor parents
can be crossed to other parents with excellent resistance to CGM, and markers
used to select recombinants that combine resistance to CMD and CGM in a
single generation, without the need for field trials. Resulting selections can then
be crossed to other genotypes that carry , for example, high beta-carotene content
to produce multi-trait hybrids, again without need for field evaluations. The best
of these selections are then crossed to elite progenitors of the appropriate gene
pool, to capture genes for yield and adaptation, and the resulting hybrids are
selected with markers to eliminate those progenies that do not have resistance to
CMD, CGM and high beta-carotene content, leaving a smaller number of
progeny to be thoroughly evaluated in the regular breeding scheme.
Molecular marker-assisted selection for breeding resistance to CMD has been
successfully implemented for introducing resistance into elite gene pools at
CIAT (Fregene et al. 2004;CIAT, 2003,) and also to introgress resistance to
CGM and CMD in Local cassava varieties in Tanzania (Kullaya et al. 2004).
With the successful development of markers for resistance to CMD and CGM,
efforts have turned to development of markers for another major source of loss
to cassava production -- postharvest physiological deterioration (PPD).
Dramatically delayed PPD was found in Manihot walkerae, a wild relative of
cassava found in Mexico and USA (state of Texas) (Fregene and Mba 2004).
An accession of M. walkerae (MWal 001) was crossed extensively to elite
cassava varieties. A single successful genotype was found with delayed PPD.
The storage roots of the hybrid remained intact a month after harvest. Backcrosses
of this hybrid to elite progenitors of the CIAT cassava gene pools and selfed (S1)
populations were made for genetic mapping of the delayed PPD traits. Genetic
Formatted
mapping of the delayed PPD genes is progressing and following identification of
genes involved in the regulation of PPD, MAS will be used to combine these
genes with progenitors that already have combined CMD and CGM genes. CIAT
and a number of partners have initiated a project to genetically fortify cassava
with the inherent ability to produce higher levels of beta-carotene. This is one
way of combating the deficiency of this key micronutrient in areas where
cassava is a major staple. The experimental approach to increasing cassava
beta-carotene
content
includes
conventional
breeding
and
genetic
transformation. The discovery of a wide segregation pattern of root color in
two S1 families from the Colombian land race MCol 72 (cross code AM 273)
and the Thai variety MTAI 8 (AM 320), was the basis for molecular genetic
analysis of beta-carotene content in cassava (CIAT 2003). Three SSR markers
were found to be associated with beta-carotene content (Fregene and Mba
2004). One of these markers explained 30% of phenotypic variation for betacarotene content in the population used for this study. The homozygous state
of certain alleles of these markers translates into higher beta-carotene content,
suggesting that breeding for beta-carotene can benefit from molecular markers
to assist in combining these favorable alleles in breeding populations. Few
key traits in cassava hold greater potential for increasing cost-effectiveness via
molecular marker assisted selection, than root dry matter content (DMC). This
trait is usually measured at the end of the growth cycle. A number of genetic
and environmental effects influence DMC. It is usually highest before the
onset of the rains, but drops after the rains begin as the plant mobilizes starch
from the roots for re-growth of leaves (Byrne 1984). Defoliation from pest
and disease attacks can lower DMC. Breeding programs have been quite
successful in improving DMC, especially for industrial markets.
The entry point for developing markers associated with DMC was three diallel
experiments carried out from 2000 to 2002 (Jaramillo et al 2004). Diallels, in
this case made up of 90 families, are an ideal method to identify genes
controlling DMC that are useful in many genetic backgrounds. Estimates of
general and specific combining ability (SCA and GCA, respectively) for many
traits of agronomic interest were calculated, with an emphasis on DMC.
Based on GCA estimates, parents were selected to generate larger sized
progenies for DMC mapping.
Sizes of families in the original diallel
experiment were about 30 progenies, a rather small size for genetic mapping.
Parallel to the development of mapping populations was the search for
markers associated with DMC using two F1 families – GM 312 and GM 313
selected from the diallel experiment having parents with high GCA for DMC.
VI Map based cloning
High resolution mapping around the cassava genome region bearing a
dominant CMD resistance gene, CMD2, has lead to the identification of 2
markers, RME1 and NS158 that flank the gene at 1 and 2 cM respectively
(Moreno 2004, unpublished data). This year, positional cloning of CMD2
made progress with the construction of a bacterial artificial chromosome
(BAC) library and BAC contigs around CMD2. The BAC library construction
was carried out at the Clemson University Genome Institute with the
participation of CIAT via a graduate student, Ms Isabel Moreno. The BAC
library was brought back to CIAT and BAC plate, column and plate pools
were created. PCR amplification, using primers of the SSR marker NS158
and the SCAR marker RME1, of the ‘BAC Pools’ were employed to identify
BAC clones containing the markers closest to CMD2 for BAC contig
construction. The BAC contigs will be the start-off point for chromosome
walking to the gene of interest.
Plant material for BAC library construction was the African local variety
TME3 that has resistance to CMD mediated by the dominant gene CMD2.
Construction of the library was as described earlier by Tomkins et al. (1999a;
1999b; 2004). To estimate the distribution and average size of the clones, a
total of 370 clones from the TME3 library were picked at random and grown
overnight in 3ml of liquid LB medium + 12.5µg /µl Chlorampenicol. Plasmid
DNA was isolated, digested with Not I restriction enzymes and inserts
separated from the vector by pulsed-field electrophoresis.
Contig construction was by PCR amplification of ‘BAC pools’, namely ‘plate
pools’ (PP), ‘column pools’ (CP) and ‘row pools’ (RP). Briefly, all 192 384well plates were duplicated using a 384-pin replicator and allowed to grow in
LB/ Cloramfenicol (12.5 ug/ml) medium at 37o C over night. For BAC plate
pools, all the bacteria culture in a 384-well plate was combined into an
omnitray and 200ul of this transferred into a single well in a 96-well plate to
yield 2 ‘BAC pool’ plates. Simultaneously, every 10 plates of the library were
inoculated into a single 384-well plate using a 384-pin replicator to give 20
384-well plates. Each row of each of the 20 plates was inoculated, using a
sterile tooth pick, into a single well containing LB/ Cloramfenicol (12.5
ug/ml) medium to form ‘Row plates (RP)’ 5 RPs of 96-well plates in all
(Figure 12.24). The same was done for each column of the 20 384-well plates
combined into a single well to form 4 96-well ‘Column plates (CP)’ (Figure
12.25).
PLATE 1
PLATE 2
B
PLATE 3
PLATE 4
Figure 12.24. Schematic diagram of how the ‘Row plates’ were formed,
sixteen rows from each of the plates (on the right) were
combined into a single well (on the right).
PLATE 1
PLATE 2
2
PLATE 3
10
PLATE 4
17
18
11
12
13
14
15
16
19
20
21
22
23
24
PLATE 5
Figure 12.25. Schematic diagram of how the ‘column plates’ were formed, 24
column from each of the plates (on the right) were combined
PLATEinto
6 a single well (on the right).
The 4 column and 5 row plates were incubated at 37oC overnight. A total of
11 plates PP, RP and CP were obtained. For PCR amplification, 5ul of the
bacteria culture was transferred using a multi-pipette to a clean 96 well plate
and the bacteria pelleted at 4500rpm for 10 minutes in a Sorvall centrifuge.
The supernatant was discarded and the pellet re-suspended in 5ul of sterile
water and used as template for PCR amplification. PCR amplification
condition for the SSR marker was 2 mM of MgCl2, 0.2 mM of DNTPs, 0.2
uM of each primer, 1 U of taq-polimerasa, in a final volume of 50ul. Thermal
cycle profile was an initial denaturation step at 95ºC 2 min, 30 cycles 94º 30s,
55ºC 1 min, 72ºC 1 min, and a final extension step of 72ºC for 5 min. For the
SCAR marker, MgCl2 and DNTP concentrations were increased to 2.5 mM
and 0.4 mM respectively. PCR cycling conditions were: 95ºC 2 min, 30
cycles: 94ºC 30 seg, 52ºC 1 min, 72ºC 1 min; and a final extension of 72ºC for
5 min. PCR products were visualized in a 1.5% agarose gel stained with
ethidium bromide.
The TME3 BAC library is made up of 73 728 clones in 192 384-well plates.
Insert size ranged from 20kb to 130kb with an average insert size of 100kb
(Figure 12.26). The BAC library has a 10X coverage of the cassava genome.
kb
194.0
145.5
97.0
48.5
pCUGI-1
Figure 12.26. Random BAC clones from the TME3 library digested with Not 1
and run in a 1% agarose gel in 0.5X TBE at 14oC at 6V/cm
with a switch time of 5s-15s pulse time for 14 hours.
Results of screening the ‘BAC pools’ with NS158 yielded 2 positive clones
while screening with RME1 yielded 34 positive clones. NS158 is a single
copy SSR marker while RME1 was developed from a multiple copy RAPD
marker. The clones were digested with 20U of HindIII overnight and run
for 24 hours on 1.2% agarose gel to obtain a BAC clone fingerprint
(Figure ???.)
Figure ???. Fingerprinting of positive BAC clones.
The positive clones are listed below (Fig.8.28)
“PP” ( “Pool Plate”)
Plate”)
90
189
“RP”( “Row Plate”)
“PP” ( “Pool Plate”)
Plate”)
12
17
34
47
51
52
54
85
130
136
139
155
173
“RP”( “Row Plate”)
“CP”(
N
M
Column
18
19
“CP”(
Column
N
M
L
O
I
A
14
21
16
3
1
23
J
D
J
G
K
F
23
18
8
22
24
1
Figure ???. List of positive clones from the ‘BAC pool’ screening using
the SSR marker NS158 and the SCAR marker RME1.
VII Advanced works (Kole recently included this section.)
Third Generation maps; physical, RH
Yvonne we don’t have physical maps yet, we have some small BAC contigs
nothing close to physical maps
Tools developed: ESTs Microarrays
EST in cassava were first reported by Suárez, et al (2000) based on cDNA–AFLP
analysis of the mRNA from the parents of the genetic mapping population, and
obtained over 500 transcript-derived fragments (TDFs). ESTs were identified in about
half of the TDFs and six were assigned to the both the male and female derived maps.
Currently, further development of SSR markers involving untranslated regions of
cassava ESTs for SSR repeats are underway (Mba et al., 2001).
Formatted: Bullets and Numbering
A collaborative project between IITA and the USDA-Agricultural Research Service,
Biosciences Research Laboratory at Fargo, North Dakota USA was initiated to
generate ESTs from cassava, and other related Euphorbs such as leafy spurge, in
order to prepare gene catalogues that are requisite for development of DNA arrays.
A hybridisation test microarray containing clones from a leafy spurge EST-database
successfully identified gene expression changes associated with cytokinin imbalances
in poinsettia samples. Analysis of gene sequences from cassava and leafy spurge
databases indicated that many genes retain sufficient conservation (sequence
similarity) to cross-hybridise (Anderson et al., 2001). Using DNA macroarrays
developed using clones from a leafy spurge EST-database (Anderson and Horvath,
2001), we further demonstrated that at least 35% of the leafy spurge clones showed
greater than 2X above background hybridisation with dehydration-stressed cassava
target sample (Anderson et al., 2004). Based on this clones from the leafy spurge
EST-database are being used as probes to monitor gene expression profiles in cassava
tissues challenged with various environmental stresses as well as other Euphorbiaceae
family members. Genes involved in the G1/S and G2/M cell cycle progression in
cassava have also been identified (Anderson et al. unpublished).
To data 20,000 ESTs have been generated for clones isolated from the two cassava
libraries. Preliminary analysis of a subset of these revealed that 73.3% of the
sequences had coding potential, 3741 sequences were single unique genes and1451
were unigene clusters with an average cluster size of 3.53. A wide range of gene
functions could be assigned and microsatellite and SNP sites, which could be used in
marker development, were identified (Lokko et al unpublished).
Details on structural and function genomics
VII. Future Work
A comparative assessment of gaps vis a vis other crops/ models
The three CMD linkage maps require further saturation with markers for a more
extensive coverage of the genome. Addition SRR markers, high throughput markers
such as EST and SNPs, may be employed in these population SNP could also
facilitate mapping CMD resistance and other disease resistance genes of interest due
to the high frequency of SNPs in the genome, there is high possibility of them being
closely associated with resistance (Lohmann et al., 2000). SNP’s also offer the
opportunity to uncover allelic variation directly within expressed sequences of
candidate genes to develop halplotypes based on gamete phase disequilibrium for
analysis of quantitative traits.
With the recent development in EST in cassava, sequence tagged sites (STS), which
are a short DNA sequence generally, between 100-500 bp that is easily recognisable
and occurs only once in a chromosome or genome studied (Brown, 1999) can be
developed to further saturate the maps and facilitate fine mapping.
With the availability and enhance of the computer package JOIN MAP future work
on the cassava framework map TMS I30572 X CM 2177-2 would be to develop
consensus map, based on male and female-derived framework map. Following this, a
composite map of cassava, based on all exiting maps could then be used in planning
experiments, to construct a genome database, to compare QTL identities in different
genetic backgrounds, and for comparative mapping with other species. Composite
maps have been reported in Arabidopsis thaliana and in Maize (Beavis and Grant,
1991; Hauge et al., 1993). Comparative mapping is a strategy that uses information
obtained from the study of a species with a simple genome to make inferences about
the map position and function of genes in species with a complex genome (Brown,
1999). Comparative mapping has been used successful in human and mouse genome
mapping to increase the efficiency of mapping (Liu, 1998). It has been suggested that
comparative mapping has definite advantages in mapping plant genomes. For
instance, genes in wheat with a genome size of 17,000 Mb would be mapped from
information obtained from a small member of the Gramineae, rice, with a genome
size of 400 Mb (Gale and Devos, 1998; Liu 1998; Brown 1999).
For effective efficient map based gene cloning and associating candidate genes
with important biological or agronomic traits there is a need for a physical
map and an established correlation between the genetic maps and the physical
map. Efforts at fine mapping specific regions of the cassava framework map at
CIAT, followed by contig mapping with bacterial artificial chromosome (BAC)
libraries is underway (Fregene et al., 2001).
Table 1 First generation linkage maps of various plant species compared with
the first cassava linkage map.
Crop
Maize
Marker
RFLP
Population Type
F2
Reference
Helentjars et al 1987
Lettuce
RFLP, Isozymes
Intraspecific cross
Landry et al 1987
Potato
RFLP
Interspecific cross
Bonierbale et al 1988
Tomato
Pepper
Rice
Arabidopsis
Random cDNA
RFLP, Isozymes
RFLP
RFLP
Interspecific cross
F2
F2/F3
Tanskley et al 1988
Tanskley et al 1988
McCouch et al 1988
Chang et al 1988
B. oleracea
B. rapa
B. napus
Cassava
RFLP
RFLP
RFLP
RFLP, RAPD,
Isozymes, SSR
F2
F2
F2
Segregating F1
Slocum et al 1990
Song et al 1991
Landry et al 1991
Fregene et al., 1997
Martin’s references (please check for constituency with text)
CIAT (1996) Development of Cassava Germplasm for the Drier Tropics and
Subtroical Agroecosystems of Africa, Asia and Latin America. A Report by
CIAT and CNMPF/EMBRAPA to the International Fund for Agricultural
Development (IFAD). CIAT, Cali, Colombia. 30pp
CIAT (1999) Pogramme for the Enhancement of Income-Generation Potential
through the Development and Transfer of Improved Cassava Genetic
Material and Adoption of Post-Harvest Technologies. A Report by CIAT and
CNMPF/EMBRAPA to the International Fund for Agricultural Development
(IFAD). CIAT, Cali, Colombia. 40pp
CIAT, (2001). Annual Report Project SB2, Assessing and Utilizing Agrobiodiversity
through Biotechnology, CIAT, Cali, Colombia, pp 239-241.
CIAT 2003a. Annual Report IP3. Improved cassava for the developing world.
Pp8-1 to 8-3
DeVries, J., and Toenniessen, G. (2001). Securing the Harvest. Biotechnology,
Breeding and Seed Systems for African Crops. CABI Publishers, New York,
USA. 208pp.
FAO, (2000) www.fao.org. Internet Database.
(Fregene and Mba 2004). Marker-Assisted Selection (MAS) In: Hershey C.
(ed) Cassava Breeding. FAO, Rome Italy (in press)
Fukuda, W.M., and Saad, N. (2001). Participatory Research in Cassava Breeding
with Farmers in Northeastern Brazil. Working Document No. 99 CNPMF,
Cruz das Almas-Bahia, Brazil, 42pp.
IITA (1999) Annual report Project 14. International Institute of Tropical Agriculture,
Ibadan, Nigeria
IITA (2000b) Enhancement of Income-Generation Potential through the Development
and Transfer of Improved Cassava Genetic Material and Adoption of PostHarvest Technologies in Africa. A report by IITA to the International Fund
for Agricultural Development (IFAD). IITA, Ibadann, Nigeria. 200pp
IITA (2000b). Impact: The Contribution of IITA-imporved cassava to food security
in sub-Saharan Africa: an impact study. IITA, Ibadan, Nigeria. 10pp
Formatted
Kawano, K. (2001). The role of improved cassava cultivars in generating income for
better farm management. In: Howeler R.H., and Tan S.L. (eds). Cassava’s
Potential in the 21st Century: Present Situation and Future Research and
Development Needs. Proceedings of the Sixth Regional Workshop, held in
Ho Chi Minh city, Vietnam. Feb 21-25, 2000, CIAT, Bangkok, Thailand, pp
5-15.
Nweke, F., Spencer, D., Lynam, J. (2001) The Cassava Transformation: Africa’s
Best-Kept Secret. Michigan State University Press, East Lansing, USA. 273
pp.
Kullaya A.; K. Mtunda; H. Kulembeka; M. Ferguson; J. Marin; C. Ospina; E. Barrera;
A. Jarvis; N. Morante; H. Ceballos; Tohme J; M. Fregene (2004) Molecular
marker-assisted and farmer participatory improvement of cassava germplasm
for farmer/market preferred traits in Tanzania In: Alves and Tohme (eds).
Adding Value to a Small-Farmer Crop: Proceedings of the Sixth International
Scientific Meeting of the Cassava Biotechnology Network. 8-14 March
2004, CIAT, Cali Colombia. Book of Abstracts Pp70