Step 1: Induction of haploidy
Introduction
The production of doubled haploid (DH) lines by in vivo haploid induction in
maize is initiated by pollinating source germplasm, from which DH lines are to
be developed, with a specific genotype called “inducer”. For maternal haploids,
the inducer is used as male parent to induce the production of seeds with
haploid embryo on the ears of the female parent.
Characteristics of the haploid inducers
Inducer genotypes should have high haploid induction rate (HIR). It is defined as
the percentage of haploid seeds found in the total number of seeds produced.
The HIR of currently available inducers in the public sector varies from
approximately 2 to 8% (Table 1).
F1
Table 1: Inducers and their haploid induction rate (HIR).
plant
Figure 1: Schematic description of doubled haploid (DH) line development with the in
vivo haploid induction approach. 1. Haploidy is induced by pollinating the source
germplasm with pollen from a haploid inducer genotype. 2. After shelling, putative
haploid seeds are identified based on the expression of seed coloration. 3. These seeds
are treated with mitotic inhibitors to artificially double their chromosomes and produce
DH plants. 4. DH plants are self-pollinated to produce seeds for maintenance and
multiplication of the DH line.
Biological mechanism – still unknown!
To date, neither histological nor cytological or genetic studies have
unambiguously explained the mechanism of in vivo haploid induction.
Principally, two mechanisms have been conceived:
1) One of the two sperm cells coming from the pollen of the inducer is
defective, yet is able to fuse with the egg cell. During subsequent cell
divisions, the chromosomes of this sperm cell degenerate and are eliminated
stepwise from the primordial cells. The second sperm cell, which is
functional, fuses with the central cells and leads to the development of a
regular triploid endosperm.
2) One sperm cell is not able to fuse with the egg cell, but instead triggers
haploid embryogenesis. The second sperm cell, which is functional, fuses
with the central cells as under the first hypothesis. However, in the latter
case, seed abortion is expected if the functional sperm cell fuses with the egg
cell, but the defective one fuses with the central cells or if the central cells
remain unfertilized for other reasons.
Current research topics
H
plant
Furthermore, inducers should
 be well adapted to the environmental conditions present when pollinating the
source germplasm for haploid production,
 have excellent pollen shedding ability,
 carry a system for the identification of haploid seeds or seedlings, and
 allow for acceptable multiplication rates.
Obviously, the number of plants of source germplasm to be pollinated by pollen
of the inducer for obtaining the desired number of haploids needs to be adjusted
according to (a) the HIR of the employed inducer and (b) the genetic
background of the source germplasm owing to its influence on the expression of
marker systems. Stock6 was the first inducer line developed in maize (Coe
1959). In the past decade, the University of Hohenheim, Germany, has
developed two inducer inbreds: RWS and UH400. These inbred lines and their
single cross RWS×UH400 are widely employed for in vivo haploid induction in
temperate maize breeding programs around the world.
Figure 2: Pollen collection from inducer plants (left) and pollination of the
silks of the source germplasm (right).
Continuous research efforts are required to develop new inducer genotypes that possess enhanced HIR, carry improved or additional haploid identification
system(s), and exhibit excellent agronomic characteristics. Further, all inducers reported in the literature have been developed from temperate maize germplasm and
were mainly evaluated for HIR and agronomic performance under temperate climatic conditions. Hence, these are not adapted to the agro-climate prevalent at
(sub)tropical research stations. The most serious constraints are susceptibility to (sub)tropical diseases and limited pollen production under hot conditions. Therefore,
we are collaborating with the International Maize and Wheat Improvement Center (CIMMYT) to develop inducers that are well adapted to (sub)tropical climates by
crossing (sub)tropical CIMMYT maize lines with the temperate inducers. Selection criteria in the selfing and backcrossing generations include agronomic
characteristics, such as vigour, disease resistance, tassel size (pollen production), and seed production, along with specific inducer characteristics, such as high HIR
and good expression of specific markers for identification of haploid seeds or seedlings. Development of such inducers is expected to greatly facilitate the
implementation of the DH technique in (sub)tropical maize breeding programs around the world.
References:
Coe EH (1959). Am Nat 93:381-382
Barret P, Brinkmann M, Beckert M (2008). Theor Appl Genet 117:581–594
Lashermes P, Beckert M (1988). Theor Appl Genet 76:404-410
Röber FK, Gordillo GA, Geiger HH (2005). Maydica 50:275-283
Shatskaya OA, Zabirova ER, Shcherbak VS, et al. (1994). Maize Genet Newsl 68:51
Step 2: Identification of haploid seeds
Introduction
The second key to success for large-scale application of the DH technology in
maize breeding programs is the ability to distinguish the seeds with haploid
embryo from normal diploid hybrid seed obtained when pollinating plants of the
source germplasm with inducer pollen. This is commonly done with phenotypic
markers. These markers can be expressed either as seed traits (e.g., seed
colour, Nanda and Chase 1966) or as biochemical (e.g., herbicide resistance,
Geiger et al. 1994) and morphological traits of the seedlings (e.g., glossy leaf
surface, Coe 1959; liguleless, Chase 1947). The advantage of the seed
coloration marker system is that haploid seeds can be identified directly after the
harvest of the ear. This saves time and resources as compared with the
systems that are only applicable on seedlings, such as biochemical or
morphological markers.
Requirements
 Marker
expression should occur as early as possible, preferably on the
seeds produced by pollination with pollen of the inducer.
of haploids should be simple, quick, low-cost, reliable, and
accurate.
Haploid inducer:
purple embryo & endosperm
×
+
F1
plant
H
plant
Diploids
Haploids
colorless
embryo
(haploid)
 The embryo marker must be dominantly inherited.
 Identification
Source germplasm:
colorless
purple
endosperm
(triploid)
purple
endosperm
(triploid)
+
purple
embryo
(diploid)
Figure 2: What happens during the induction cross? When the unpigmented source
germplasm is pollinated with inducers which carry dominant seed coloration markers,
successful fertilization with inducer pollen is visible as a purple colored aleurone
(cellular layer of the endosperm) which indicates a fully functional, triploid endosperm.
The haploids differ from diploids by their colorless scutellum (part of the embryo)
indicating a haploid embryo of solely maternal origin.
 Marker expression should be independent of the genetic background of the
source germplasm and stable across different environments.
Inducers RWS and UH400 carry the dominant marker gene R1-nj (Nanda and
Chase 1966, Neuffer et al. 1997). This gene belongs to a family of genes that
regulates the expression of the pigment anthocyanine. The alleles of the locus
R are unique among plant genes because specific alleles control the
pigmentation in very specific tissues (Li et al. 2001). The allele R1-nj produces
a phenotype with purple scutellum and aleurone which can be used as embryo
and endosperm marker, respectively.
Endosperm coloration
Figure 3: Typical ears harvested from an induction cross: the seed color
marker works in white and yellow maize alike!
C
B
Embryo coloration
Figure 1: Marker traits carried by the inducer: purple colored stalk and purple
colored seed (scutellum and aleurone, acting as embryo and endosperm
marker, respectively).
A
Figure 4: Classifying seeds from
an induction cross.
Current research topics
Figure 5: Seeds commonly found on
induced ears: (A) haploids, (B) normal
F1 seeds, and (C) completely unpigmented seeds resulting from color
inhibition or (unintended) outcrossing.
The R1-nj marker system cannot be used for those source germplasm exhibiting anthocyanine seed coloration (such as e.g. “blue maize”, a delicacy in Mexico) or
those carrying specific inhibitor genes (Coe and Sarkar 1964). To overcome these limitations, we are working on the development of alternative haploid
identification systems such as root coloration (as an addition to seed coloration) and oil content in the embryo. Measuring the oil content of individual seeds with
near-infra-red spectroscopy equipment would be extremely fast and unbiased by human scoring.
References:
Coe EH, Sarkar KR (1964). J Hered 55:231-233
Li Y, Bernot JP, Illingworth C, et al. (2001). Genetics 159:1727-1740
Nanda DK, Chase SS (1966). Crop Sci 6:213-215
Neuffer MG, Coe EH, Wessler SR (1997). Cold Spring Harbor Laboratory Press, New York
Step 3: Artificial chromosome doubling
Introduction
How does colchicine work?
Planting seeds with haploid embryo would result in
plants which are mostly sterile because they carry
only one set of chromosomes. Consequently, such
plants can neither be maintained nor multiplied by
self-fertilization. Although spontaneous chromosome
doubling does occur in maize and some haploid plants
are also partially fertile, their frequency is too low to
exploit these in commercial maize breeding.
Depending on the genotype, spontaneous occurrence
of chromosome doubling in haploids has been
reported to vary greatly (see, e.g., Chase 1969,
Deimling et al. 1997, Geiger et al. 2006). In order to
employ the DH technology in large-scale maize
breeding programs, a cheap, easy and reliable
artificial chromosome duplication system is needed.
Gayen et al. (1994) developed a protocol using
colchicine, the alcaloid produced by Colchicum
autumnale.
Colchicine works as mitotic inhibitor. Mitosis is the process of nucleus division in somatic cells. During
mitosis, DNA replicates and the duplicated chromatides are usually pulled towards the two poles by
microtubules and the cell divides into two daughter cells (see scheme to the right). Colchicine binds to
tubulin, a protein of the microtubules, and prevents the microtubules from pulling the chromatides to the
poles. As a result, the haploid genome is duplicated but the cell does not divide, thus, a doubled-haploid
(i.e., diploid genome in a single cell) is formed.
F1
H
plant
Normal mitosis
plant
DNA replication
Two
haploid
daughter cells
Mitosis
Inhibited mitosis
DNA replication
One
doubled haploid
(=diploid) cell
Routines for colchicine treatment
Figure 1: Haploid seeds are
germinated under controlled
conditions (28°C, dark, ~ 3 days).
Anti-fungal treatment of the seeds
is recommended.
Figure 2: Once the
coleoptiles of seedlings have
reached a length of 2 cm,
their tip is cut with a razor
blade to facilitate penetration
of the colchicine solution.
Figure 3: The seedlings are submerged in a
0.06% colchicine solution for 8 hours. Properly
labelled mesh bags are used to separate
seedlings of different source germplasm.
A
B
Figure 7: At 3- to 4-leaf stage,
seedlings are ready for transport
to the field station.
Figure 5: After treatment, seedlings are
rinsed with tap water and planted into
biodegradable or plastic pots filled with
soil.
A
Figure 6: The seedlings are kept in the
greenhouse for 7-10 days to recover from
the stress imposed by the colchicine
treatment. They need proper watering but
excess moisture must be avoided.
B
Figure 8: After several days of adaptation to the local conditions, the seedlings
are transplanted to the field, either manually (A) or with a planting machine (B).
Figure 4: Common laboratory glass
containers can be used for small-scale
treatments (A), and a custom-made
stainless steel tank (B) equipped with an
electric pump and a timer is suitable for
large-scale application. Using a pump
and a timer, the staff doesn`t come in
contact with the chemical solution.
Filling of the tank is started and stopped
automatically.
Current research topics
Colchicine is a toxic and costly chemical and requires elaborate procedures for
residual waste removal. Furthermore, comprehensive safety guidelines must be
followed. For small national maize breeding programs, particularly in
developing countries, this may be an insurmountable problem, thus preventing
the adoption of the DH technology despite its advantages. Therefore, we are
currently investigating alternatives, such as components of herbicides (e.g.,
amipropos methyl, oryzalin, and pronamide) (Häntzschel and Weber 2010).
Herbicides are less toxic and have the advantage that they may also be applied
by spray-application with haploid seedlings after planting in the field.
Furthermore, we are trying to exploit the genetic ability of haploids to
spontaneously double their chromosomes or exhibit male and female fertility to
some extent.
References:
Chase SS (1969). Bot Rev 35:117-167
Deimling S, Röber FK, Geiger HH (1997). Vortr Pflanzenzüchtung 38:203-224
Gayen P, Mandan JK, Kumar R, et al. (1994). Maize Genet Newsl 68:65
Geiger HH, Braun MD, Gordillo GA et al. (2006) Maize Genet Newsl 80:28–29
Häntzschel KR, Weber G. (2010) Protoplasma 241:99-104
Doubled haploids
in maize breeding
& research
Institute of Plant Breeding, Seed Science, and Population Genetics (350a)
Step 4: Seed multiplication of DH plants
Introduction
The generation of plants obtained from the colchicine treated seedlings is
designated D0 generation. Each of these D0 plants is a unique genotype. Provided
that their chromosomes have doubled and the plants survive under field conditions,
D0 plants produce D1 seed. This seed represents the newly developed, completely
homozygous DH line.
One can expect a survival rate of 5 % which means that 5 % of the haploid seeds of
a source germplasm will result in DH lines. Several factors affect this success rate,
such as:
“False” F1 plants derived from hybrid seeds, which were wrongly classified as
haploids during the selection of seeds, can easily be identified. They are
vigorous, often tiller and have a purple coloured stalk (as inherited from the
inducer parent). They also have a highly branched tassel shedding lots of
pollen. In contrast, “real” D0 plants are short and weak, have fertility problems
(e.g., only partial anther emergence due to incomplete chromosome
F1
H
duplication), and produce little pollen.
plant
plant
 Type of source germplasm (e.g. landraces are expected to have a higher
genetic load that may negatively affect germination and/or growth);
 Accuracy of haploid seed identification system (misclassification leads to a high
rate of normal F1 progeny in the D0 nursery (Figure 1). These must be eliminated
as soon as possible to prevent any confusion);
 Efficiency of chromosome doubling system (typically, with colchicine treatment,
chromosomes are not doubled uniformly in all cells of a seedling; this may
result in partial infertility (Figure 2) and hinder maintenance and seed increase
through selfing);
false „positives“
Figure 1: „False“ F1 plants (arrows) amongst the DH plants in the D0 nursery
can be easily identified by their vigour and the purple-colored stalk.
 Field (climatic/agronomic) conditions at the experimental station (Figure 3).
The best lot of land on a station with light soil as well as drainage and irrigation
facilities should be allocated to the D0 nursery. Further, application of optimal
agricultural practices, such as timely application of the required dose of
fertilizers and plant protection measures, is important to ensure survival of D0
plants. In addition, well trained staff is essential to ensure optimal use of the
small quantities of pollen produced and for the reliable identification and
elimination of “false” plants from the D0 nursery.
Figure 2: Different types of tassels of colchicine-treated maize plants. (A) Tassel with many
emerged anthers on main and side branches. Pollen is released upon touching anthers. (B)
Only one side branch of the tassel has anthers emerged that shed pollen. (C) Several
anthers have emerged on the main and side branches but no pollen is released upon
touching the anthers. (D) No anthers emerged. In general, if the small quantity of pollen is
handled carefully, it is usually sufficient for effective self-pollination and production of some
seeds, so that the genotype can be maintained.
A
Figure 4: Few seeds are often found on ears
harvested from D0 plants. Seed production on
such plants is expected to improve in
subsequent cycles of DH production because
(i) selection occurs in source germplasm for
genes imparting favorable response to
haploid induction and artificial chromosome
doubling, and (ii) the persons involved gain
experience in handling the system. Hence,
the more frequently source germplasm
passes through the haploid induction and
artificial chromosome doubling processes, the
higher will be the success rate in future
breeding cycles.
B
Figure 3: Non-coated, transparent glassine bags (size approx. 6×20
cm) are most suitable to collect pollen from putative doubled haploid
plants for self-pollination (A). As pollen production is often limited in
these plants, the transparent bags allow visual assessment of the
quantity of pollen collected for self-pollination. If necessary, the
pollination can be repeated the next day. Shading nets may help
reduce stress imposed on seedlings by radiation in the D0 nursery (B).
Figure 5: DH lines display complete uniformity within lines and great
diversity between lines (different DH lines planted in each row).
We gratefully acknowledge the financial contributions of the Eiselen-Foundation Ulm, Germany, and the Tiberius Group, Stuttgart,
Germany, for our research to improve the DH technology for breeding programs of temperate and tropical maize.
Vanessa Prigge
The scientific and editorial contributions of B.S. Dhillon,
Albrecht E. Melchinger
W. Schipprack, and B. Devezi-Savula are greatfully acknowledged.
melchinger@uni-hohenheim.de vaprigge@uni-hohenheim.de