Botanical Studies (2008) 49: 93-99.
Corresponding author: E-mail:;
Tel: +86-22-23508241; Fax: +86-22-23497010.
A genetic linkage map based on AFLP and NBS markers
in cauliflower (Brassica oleracea var. botrytis)
, Qian-Chen ZHAO
, De-Ling SUN
, and Wen-Qin SONG
Tianjin Kernel Vegetable Research Institute, Tianjin 300382, P. R. China
Department of Genetics and Cell Biology, College of Life Sciences, Nankai University, Tianjin 300071, P.R. China
(Received March 26, 2007; Accepted October 4, 2007)
A genetic linkage map of cauliflower (Brassica oleracea var. botrytis) has been constructed
based on AFLP and nucleotide binding site (NBS) markers, in order to identify potential molecular markers
linked to important agronomic traits that could be useful for developing and improving the species. NBS
profiling was first used to map resistance gene analogues (RGAs) in cauliflower (Brassica oleracea var.
botrytis), which simultaneously allowed the amplification and mapping of genetic markers anchored in the
conserved NBS encoding domain of plant disease resistance genes. At the same time, the AFLP method was
also performed in this paper to construct an intervarietal genetic map of cauliflower. A total of 234 AFLP
markers and 21 NBS markers were mapped in an F
population derived by self-pollinating a single F
(a hybrid between "AD White Flower" and "C-8") based on seventeen AFLP primer combinations and two
degenerate primer/enzyme combinations. The markers were mapped into nine major linkage groups, spanning
668.4 cM, with an average distance of 2.9 cM between adjacent mapped markers. Each of the linkage groups
contained from 12 to 47 loci, and the distance between two consecutive loci ranged from 0 to 14.9 cM. The
AFLP markers were well distributed throughout the nine linkage groups, and eight linkage groups for the NBS
markers. Most NBS markers mapped in this study were organized in clusters, indicating that most of them
could be real RGAs. The maps we have generated provide a firm basis for mapping agriculturally relevant
traits, which will then open the way for application of a marker-assisted selection breeding strategy in this
Keywords: AFLP; Cauliflower (Brassica oleracea var. botrytis); Genetic linkage map; NBS profiling.
Abbreviations: NBS-LRR, nucleotide binding site-leucine-rich repeat; RGAs, resistance gene analogues;
PCs, primer combinations; LGs, linkage groups; cM, centiMorgans; R gene, resistance gene.
Over the past two decades, several genetic maps of
Brassica oleracea have been constructed. The first map
was based on the segregation of 258 restriction fragment
length polymorphism (RFLP) loci in a broccoli cabbage
population. The genetic markers defined nine linkage
groups, covering 820 recombination units (Slocum et
al., 1990). Subsequently, at least ten genetic linkage
maps were developed (Landry et al., 1992; Kianian et
al., 1992; Bohuon et al., 1996; Camargo et al., 1997;
Voorrips et al., 1997; Cheung et al., 1997; Hu et al., 1998;
Moriguchi et al., 1999; Sebastian et al., 2000; Chen et
al., 2002; Howell et al., 2002; Wang et al., 2005; Wang
et al., 2007). These maps were based on intraspecific
or intersubspecific populations. An intervarietal genetic
map of cauliflower (Brassica oleracea var. botrytis) was
missing. The high level of genetic variability within an
intraspecific population or the level of DNA polymorphism
in the parents may explain this. Nevertheless, having an
intervarietal linkage map is the most direct and efficient
approach for cauliflower breeding in the future since the
genetic information is derived from cauliflower.
Most studies have dissociated the isolation (cloning)
of resistance gene analogos (RGAs) from their genetic
mapping in segregating progenies. Most often, cloned
RGAs are mapped using RFLPs, which is often time-
consuming (Calenge et al., 2005). Modified amplified
fragment length polymorphisms (AFLP) and nucleotide
binding site (NBS) profiling were proposed as new
strategies by Hayes and Saghai-Maroof (2000) and by
Van der Linden et al. (2004) to generate, simultaneously,
polymorphism and specifically amplify highly conserved
motifs. Both methods are based on the simultaneous use
of an adapter primer matching a restriction enzyme site
Botanical Studies, Vol. 49, 2008
and of a degenerate primer targeting the NBS encoding
region. Genetic variation is sampled in the gene region
flanking the primer-binding site. Nevertheless, NBS
profiling has some advantages that improve specificity.
NBS profiling involves a two-step PCR procedure. The
first step is a linear (asymmetric) PCR with a limited
amount of the NBS specific primer. The asymmetric PCR
is then followed by an exponential PCR with an NBS
primer and an adapter primer. An amino linker effectively
blocks extension of the 3 end of the short arm in the
subsequent PCR reaction. The adapter primer can thus
only participate in the PCR reaction after extension of
the selective NBS primer. Amplification depends highly
on the selectivity and specificity of this NBS primer. To
further increase the specificity of the PCR procedure, the
actual amplification (exponential PCR) is preceded by
linear (asymmetric) PCR with only the NBS primer. In the
modified AFLP method, in the initial digestion and ligation
steps, approximately 87.5% of the soybean genome is
eliminated because Mse/Mse fragments are not detected
in the preceding steps. Nonetheless, NBS profiling can be
identified in all genome. NBS profiling was shown to be
highly effective in generating polymorphic markers with
high sequence homology for RGAs in several species (Van
der Linden et al., 2004).
The purposes of this paper were: 1) To construct an
intervarietal linkage map of cauliflower. This map could be
used to identify more markers, which would eventually be
linked to genes controlling important agronomic characters
and provide a powerful tool to be used in marker-assisted
breeding of cauliflower. 2) To demonstrate the reliability
and efficiency of the NBS profiling approach for mapping
homologous nucleotide binding site-leucine rich repeat
(NBS-LRR) genes in cauliflower, and 3) to determine
the distribution and characterization of NBS-LRR-like
sequences in the cauliflower genome.
Plant materials and DNA isolation
An F
population of 100 progenies was generated by
self-pollination of a single F
plant from an intervarietal
cross between "AD White Flower" as pollen parent
and "C-8" as seed parent. The mapping population was
supplied by the Tianjin Kernel Vegetable Research
Institute. "AD White Flower" is an inbred line, developed
from a general Chinese variety through several inbreeding
generations. "C-8" is a self-incompatible line derived from
an individual plant mutation of an exotic line.
Total genomic DNA of mapping progenies was isolated
following the CTAB method (Murray and Thompson,
1980) with some modifications. Fresh leaves (0.2 g) of
both lines were ground in liquid nitrogen. The frozen
powder was directly added to 2 ml lysis buffer (100
mM Tris-HCl, pH 8.0, 50 mM EDTA, 1.4 M NaCl,
0.2% ]-mercaptoethanol, 2% PVP and 1CTAB). After
incubation at 56XC for 30 min, 2 ml phenol: chloroform:
isoamyl alcohol (25:24:1) was added. The supernatant
was obtained by centrifugation at 10,000 g for 10 min
and extracted once with an equal volume of chloroform:
isoamyl alcohol (24:1). After centrifugation at 11,000 g
for 15 min, a two-thirds volume of cold isopropanol was
added to the supernatant. The mixture was incubated for
30 min at -20XC. DNA was pelleted by centrifugation at
11,000 g for 30 min at 4XC and dissolved in 100 gl TE (pH
8.0) buffer. RNA was removed from the DNA by treatment
with 200 gg gl-1 RNase A for 1 h at 37XC.
AFLP molecular marker analysis
AFLP analysis was performed according to Vos et
al. (1995) with minor modifications. Briefly, 100-150 ng
of DNA was digested with 1.5 U of both EcoR I and Mse
I (Shanghai Sangon, Shanghai China). Aft er ligation and
pre-amplification, selective amplification was conducted
by combining 30 ng of an EcoR I primer containing two
selective nucleotides and an Mse I primer containing
three selective nucleotides. Thermocycling (Mastercycler
Gradient 5331, Eppendorf Germany) was done with
35 cycles and included a 12-cycle touchdown (annealing
temperature was reduced from 65XC to 56XC a t 0.7XC pe r
step for 12 cycles, and subsequently maintained at 56XC
for 23 c yc les). Amplification products were separated in 6%
(w/v) denaturing polyacrylamide gels at 65 W for 2.5 h in 1
TBE buffer. After silver staining (Bassam et al., 1991), the
gels were dried at room temperature and photographed.
NBS profiling protocol
NBS profiling of the progeny was performed essentially
as described by Van der Linden et al. (2004). In brief,
DNA was digested for 4 h with Mse I, using 400 ng DNA
per individual. An adapter was ligated to the restriction
fragments. Adapter sequences and the adapter primer
Adapter short arm: 5-TGGGATCTATACTT-3 (with 3
amino group)
Two different degenerate primers, NBS2 (5-GTWG
THGTMITKGATGAYGTITGG-3), were then used for
a two-step PCR procedure. These primers were designed
from part of the conserved P-loop and Kinase-2 motif,
respectively, belonging to the NBS encoding region of
several plant disease resistance genes (Van der Linden
et al., 2004). The first PCR was linear, and only the
degenerate primer (either NBS2 or NBS5) was used. The
second PCR was exponential and was performed with both
the degenerate primer and the adapter primer. Both PCRs
were performed with an annealing temperature of 60XC
on a Mastercycler Gradient 5331 (Mastercycler Gradient
5331, Eppendorf Germany), using the following cycling
GU et al. X Cauliflower genetic linkage map
program: 30 cycles of 30 s at 95XC, 1 min 40 s at 60XC and
2 min at 72XC. The PCR products were separated in 6%
(w/v) polyacrylamide gels for 2.5 h at 65 W, then dried at
room temperature and photographed.
Genetic linkage map construction
The AFLP fragments were scored as dominant,
presence versus absence of bands, and therefore for the
parents, F
and F
generation markers were assigned to
either parental allele for map construction. AFLP markers
were designated by the PCs used, followed by a number
reflecting the fragment length on the gel (e.g., M-AAGe-
AC400 = Mse I +AAG/EcoR I +AC, band length is
400 bp). NBS markers were tested against the predicted
3:1 ratio, representing [homozygote + heterozygote]:
[homozygotes], by the presence: absence of the marker
band, respectively. NBS markers were named using the
same protocol/logic as AFLP (e.g., NBS2M60 = NBS2/
Mse I, band length is 60 bp).
Markers originating from each parent were scored
according to the standard coding system using A, B,
C, D and H of JoinMap, Version 3.0 (Van Ooijen and
Voorrips, 2001). Ambiguous genotypes were resolved
by assigning a blank score (-) to the individual locus for
map construction. Chi-square (q
) tests of goodness-of-fit
were performed on segregation data for all markers, with a
0.5% threshold level for significance. To determine marker
order within a linkage group, the following JoinMap
parameter settings were used: Rec = 0.40, LOD = 1.0,
Jump = 5. Markers were assigned to linkage groups (LGs)
by increasing the LOD score for grouping with steps of
one LOD unit. No order was forced during the linkage
analysis. Recombination frequencies were converted to
map distances in centimorgans (cM) using the Kosambi
mapping function (Kosambi, 1943).
Polymorphism analysis using AFLP markers
Initially, sixty-four AFLP PCs were screened for
polymorphism by using two parents. Seventeen most
informative EcoR I+ 2 / Mse I+3 PCs (Table 1) were
selected for mapping according to the number and
reliability of the polymorphic marker. Approximate 1,700
bands were produced by the seventeen PCs in the mapping
population with an average of 100 bands per PC. In
total, 339 polymorphic markers were scored as dominant
markers (Figure 1). The number of polymorphic AFLP
markers per PC ranged from 13 to 26, with a mean of 20
markers per PC (Table 1). Chi-square analysis revealed
234 of the 339 markers (69.9%) fitted a 3:1 ratio. The
remaining 105 (30.1%) showed segregation distortion
within the population.
NBS profiling marker development in
Two degenerate primer/enzyme combinations were
tested in the progeny: NBS2/Mse I and NBS5/ Mse I.
An average of 60-90 monomorphic bands per PC was
obtained. Thirty-one polymorphic bands were identified
with the two combinations and scored as dominant
markers (Figure 1). Ten markers (32.3%) had a distorted
segregation in the progeny as assessed by chi-square tests
and were discarded from data. 21 polymorphic markers
(68%) were reliably mapped on the linkage map (Figure 2).
NBS markers were mapped to 8 of the 9 linkage groups of
Table 1. Primer combinations for AFLP and NBS markers are shown with the number of polymorphic markers generated
for each of them.
Primer combination
Number of markers
Primer combination
Number of markers
Botanical Studies, Vol. 49, 2008
Figure 1. AFLP and NBS markers in a F
mapping population. Markers used for this study are indicated by arrow. Primer combinations
for AFLP and NBS profiling were M-CTT/E-AG and NBS5/MseI, respectively.
the genetic maps (Figure 2). Most markers were organized
in more or less wide clusters of 0-18 cM.
Linkage analysis and map construction
The 234 AFLP markers and 21 NBS markers were used
to construct a genetic map. Altogether, 255 of the 370
polymorphic markers were assigned to nine linkage groups
(LGs) (Figure 2) ranging in size from 12 to 60 markers
each. Map distances ranged from 42.6 to 113.0 cM per
LGs (Table 2). The 255 markers (229 loci) gave a total
map length of 668.4 cM and an average genetic distance
between adjacent mapped loci as 2.9 cM (Figure 2).
Linkage relationships of the 255 segregating markers were
established at a 3.0 . LOD . 10.0 and a recombination
fraction smaller than 0.5 Map distances were converted to
cM using the Kosambi mapping function (Kosambi, 1943).
According to Figure 2, we can see the names of markers
are shown at the right and their map position (cM) at the
left. All the markers appear to be distributed randomly
along all LGs, except LGs 3, 4 and 5, which show rather
few markers.
Several genetic linkage maps of B. oleracea have been
produced in the past using intraspecific or intersubspecific
mating systems. It is generally believed that the degree
of polymorphism is lower in an intervarietal population
than in an intraspecific or intersubspecific population. The
high level of genetic variability within an intraspecific
population or the level of heterozygosity of the parents
may explain this. However, Hu et al. (1998) compared
three previous independent B. oleracea linkage maps
and pointed out that the distances between markers
often varied from map to map. This demonstrated
that sequence rearrangement is a distinct feature of
this genome. Slocum et al. (1990) reached the same
conclusion. Their study suggested that a fairly high degree
of genetic rearrangement has occurred in the evolution
of B. oleracea. With respect to cauliflower breeding, we
believe that intervarietal crosses should be better than
intersubspecific. Results of this study suggest that an
intervarietal cross allows better map resolution and even
distribution in LGs.
Some published AFLP linkage maps show clustering
of these markers in centromeric regions, due to an excess
of repeats in this area and suppressed recombination
shrinking the genetic map relative to the DNA content
(Jeuken et al., 2001). In this study, do not find any mark on
centromeric regions of any LG (Figure 2). Chromosome
centromeric regions are usually conserved and may not be
polymorphic when self-pollinated. This is in contrast to the
clusters of markers mapped on an interspecific segregating
population (Pearl et al., 2004; Lan et al., 2008).
It was previously demonstrated that the NBS profiling
method allows the generation of polymorphic markers
among different cultivars of potato, tomato, barley,
and lettuce (Van der Linden et al., 2004). The present
investigation is the first to test this new method in
cauliflower and the first to demonstrate its utility for
GU et al. X Cauliflower genetic linkage map
Figure 2. Genetic linkage map of cauliflower (Brassica oleracea var. botrytis). Linkage groups are numbered LG1 through to LG9.
Markers are indicated at the right of each linkage group. Recombination distances are in Kosambis cM and are indicated at the left of
each linkage group.
Table 2. Marker number and map distance per linkage group.
Linkage group Marker Num. (loci) Marker Num AFLP/NBS (loci) Length (cM) Mean length (cM) LOD
Botanical Studies, Vol. 49, 2008
mapping studies using a segregating progeny. Thirty-one
polymorphic markers were observed in the intravarietal
progenies across the two-primer/enzyme combinations
tested, of which 21 markers could be mapped. Ten
unmapped markers had a distorted segregation in the
progeny, which could explain why they could not be
reliably mapped. Fewer NBS profiling markers were
obtained than AFLP markers. This is an inherent property
of such markers, as each in principle corresponds to a
member of the NBS-LRR RGA gene family that is present
in far lower numbers in the plant genome than restriction
sites (AFLP) (Syed et al., 2006). The usefulness of NBS
markers is their close linkage to potentially important
resistance genes. The NBS profiling molecular marker
methods have been validated for cauliflower.
In our map, most markers were assembled in clusters.
The best example is the cluster on linkage group LG6,
which includes four markers (Figure 2). Therefore these
markers correspond either to the same gene locus or to
very tightly linked loci. The clustering of RGAs of the
NBS-LRR type has previously been observed in numerous
studies. Most NBS-LRR and other R gene-like sequences
reside in large, extended arrays (Young, 2000). This
clustering of RGAs is not surprising, considering that plant
R genes often, though not always, belong to gene families
with evolutionarily related tandemly repeated genes or to
allelic series (Hulbert et al., 2001). The clustering of the
markers mapped in this study adds to the evidence that
most of them could be real RGAs.
Acknowledgements. The authors thank Ms Lili Liu from
the Tianjin Kernel Vegetable Research Institute of China
for planting and tending the mapping population. Financial
support for this study was provided by the National
Science Foundation of Tianjin (No.06YFJMJC10000).
Bassam, B.J., G. Caetano-Anolles, and P.M. Gresshoff. 1991.
Fast and sensitive silver staining of DNA in polyacrylamide
gels. Anal. Biochem. 196: 80-83.
Bohuon, E.J.R., D.J. Keith, I.A.P. Parkin, A.G. Sharpe, and D.J.
Lydiate. 1996. Alignment of the conserved C genomes of
Brassica oleracea and Brassica napus. Theor. Appl. Genet.
93: 833-839.
Calenge, F., C.G. Van der Linden, E. Van de Weg, H.J. Schou-
ten, G. Van Arkel, C. Denance, and C-E . Durel. 2005.
Resis tance gene analogues identified through the NBS -
profiling method map close to major genes and QTL
for disease res istance in apple. Theor. Appl. Genet. 110:
Camargo, L.E.A., L. Savides, G. Jung, J. Nienhuis, and T.C.
Osborn. 1997. Location of the self-incompatibility locus in
an RFLP and RAPD map of Brassica oleracea. J. Hered.
88: 57-60.
Chen, S.X., X.W. Wang, Z.Y. Fang, Z.H. Cheng, and P.T. Sun.
2002. Construction of Molecular Linkage Map of Brassica
oleracea var. alboglabra B. oleracea var. capitata by
RAPD analysis. Acta Horticulturae Sin. 29: 229-232.
Cheung, W.Y., G. Champagne, N. Hubert, and B.S. Landry.
1997. Comparison of the genetic maps of Brassica napus
and Brassica oleracea. Theor. Appl. Genet. 94: 569-582.
Howell, E.C., G.C. Barker, G.H. Jones, M.J. Kearsey, G.J. King,
E.P. Kop, C.D. Ryder, G.R. Teakleb, J.G. Vicenteb, and
S.J. Armstrong. 2002. Integration of the cytogenetic and
genetic linkage maps of Brassica oleracea. Genetics 161:
Hayes, A.J. and M.A. Saghai Maroof. 2000. Targeted resistance
gene m apping in s oybean us ing m odied AFLPs . Theor.
Appl. Genet. 100: 1279-1283.
Hulbert, S .H., C.A. We bb, S .M. S mith, and Q. S un. 2001.
Resistance gene complexes: evolution and utilization. Annu
Rev. Phytopathol. 39: 285-312.
Hu, J., J. Sadowski, T.C. Osborn, B.S. Landry, and C.F. Quiros.
1998. Linkage group alignm ent from four independent
Brassica oleracea RFLP maps. Genome 41: 226-235.
Jeuken, M., R. van Wijk, J. Peleman, and P. Lindhout. 2001.
An integrated interspecific AFLP map of lettuce (Lactuca)
based on two L. sativa L. saligna F
populations. Theor.
Appl. Genet. 103: 638-647.
Kianian, S.F. and C.F. Quiros. 1992. Generation of a Brassica
oleracea composit e RFL P map: li nka ge arra nge ments
among various populations and evolutionary implications.
Theor. Appl. Genet. 84: 544-554.
Kosambi, D.D. 1943. The estimation of map units from
recombination values. Ann. Eugen. 12: 172-175.
Lan, T.Y., R.Y. Chen, X.L. Li, F.P. Dong, Y.C. Qi, and W.Q.
Song. 2008. Microdissection and painting of the W
chromosome in Ginkgo biloba showed different labelling
patterns. Bot. Stud. 49: 33-37.
Landry, B.S., N. Hubert, R. Crete, M.S. Chang, S.E. Lincoln,
and T. Etoh. 1992. A genetic map for Brassica oler acea
based on RFLP markers de tected with expresse d DNA
se quenc es a nd mapp ing of resi st ance genes to rac e 2
of Plasmodiophora brassicae (Woronin). Genome 35:
Moriguchi, K., C. Kimizuka-Takagi, K. Ishii, and K. Nomura.
1999. A genetic map based on RAP D, RF LP, isozyme,
morphological ma rkers and QTL analysis for clubroot
resistance in Brassica oleracea. Breed. Sci. 49: 257-265.
Murray, M.G. and W.F. Thompson. 1980. Rapid isolation of high
weight plant DNA. Nucleic Acids Res. 8: 4231-4235.
P earl, H.M., C. Nagai, P.H. Moore, and D.L. Steiger. 2004.
Construction of a genetic map for arabica coffee. Theor.
Appl. Genet. 108: 829-835.
Sebastian, R.L., E.C. Howell, G.J. King, D.F. Marshal, and M.J.
Kears ey. 2000. An integrated AFLP and RF LP Brassica
oleracea linkage map from two morphologically distinct
doubled-haploid mapping populations. Theor. Appl. Genet.
100: 75-81.
GU et al. X Cauliflower genetic linkage map
Slocum, M.K., S.S. Figdore, W.C. Kennard, J.Y. Suzuki, and T.C.
Osborn. 1990. Linkage arrangement of restriction fragment
length polymorphism loci in Br as sica oler acea. Theor.
Appl. Genet. 80: 57-64.
Syed, N.H., A.P. Sorensen, R. Antonise, C. Van de Wiel, C.G.
Van der Linden, W. Vant Wes tende, D.A.P. Hooftman,
H.C.M. Den Nijs, and A.J. Flavell. 2006. A detailed linkage
map of lettuce based on S SAP, AFLP and NBS markers.
Theor. Appl. Genet. 112: 517-527.
Van der Linden, C.G., D.C.A.E. Wouters , V. Mihalka, E.Z.
Koc hie va , M.J .M. S m ulde rs , a nd B. Vos ma n. 20 04.
Efficient targeting of plant disease resistance loci us ing
NBS profiling. Theor. Appl. Genet. 109: 384-393.
Van Ooijen, J.W. and R.E. Voorrips. 2001. JoinMap Version 3.0,
Software for the Calculation of Genetic Linkage Maps.
Plant Res Int, Wageningen.
Voorrips, R.E., M.C. Jongerius, and H.J. Kanne. 1997. Mapping
of two genes for resistance to clubroot (Plasmodiophora
brassicae) in a populati on of doubled haploid lines of
Brassica oleracea by means of RFLP and AFLP markers.
Theor. Appl. Genet. 94: 75-82.
Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. Lee, M. Hornes,
A. Frijters, J. Pot, J. Peleman, M. Kupier, and M. Zabeau.
1995. AFLP : A new technique for DNA fingerprinting.
Nucleic Acids Res. 23: 4407-4414.
Wang, C.G., X.Q. Chen, H. Li, and W.Q. Song. 2007. RNA
e diting analysi s of mitochondrial nad3/rps12 genes in
c ytoplasmic m ale sterility and m ale-fertile cauliflower
(Brassica oleracea var. botrytis) by cDNA-SSCP. Bot. Stud.
48: 13-23.
Wang, X.W., P. Lou, H.J. He, B.J. Yang, Y.G. Zhang, and J.J.
Zhao. 2005. Construction of an AFLP-based genetic linkage
m ap us ing a doubled-haploid (DH) population derived
from a cross between chines e kale and broc coli. Acta
Horticulturae Sin. 32: 30-34.
Young, N.D. 2000. The genetic architecture of resistance. Curr.
Opin. Plant Biol. 3: 285-290.