Download Gradzial-Variety Development - California Cling Peach Board

California Cling Peach Advisory Board
2006 Annual Report
Project Titles:
Development of New Cling Peach Varieties
Project Leaders:
Tom Gradziel & Carlos Crisostos
Cooperating Personnel:
M.A. Thorpe, C. Peace, R. Bostock , J. Adaskaveg
C. Peace and E. Ogundiwin
Location:
Dept. of Pomology, University of California at Davis
Breeding Synopsis.
Despite the challenging weather conditions during the spring bloom, the breeding
program has achieved it’s goal of over 6000 seed from controlled crosses, primarily
through the used of protected pollination structures and an emphasis on enforced selfpollinations of previous hybrids. (The hybridization of selected parents possessing
distinct but desirable traits allows their consolidation in the hybrids, while the subsequent
selfing of the best hybrid progeny allow the concentration of desirable genes from both
parents in a genetic background more adapted to California growing conditions). Over
18,000 seedling trees from breeding crosses were evaluated for fruit and tree quality.
Almost 500 individuals seedling trees were selected for further evaluation of processing
potential using the UCD pilot plant processing facility at Cruess Hall.
BioStar initiative to maximize breeding efficiency using recent biotechnologies. To
accelerate progress in developing improved fruit firmness disease/pest resistance, the
2006 processing peach Variety Development and Regional Testing programs underwent
dramatic expansion in research efforts though without increases in the level of industry
support. In previous years we have maximized our variety development and regional
testing efforts (also with very limited increases in industry funding) by combining day-today activities of the peach breeding program with a similarly industry-funded almond
breeding program. Because the more intensive peach and almond field management
operations (planting, pollination at flowering, harvest, etc.) were largely sequential with
little overlap between the two crops, we were able to maximize our labor and equipment
efficiency. The greater incentive for combining these seemingly disparate breeding
programs, however, involved our long-term efforts to enrich the previously very narrow
genetic variability within each crop through the exchange of genes between these and
related species. This more exotic germplasm also has unique value for basic research
on genetic control of plant development, allowing us to secure additional outside
funding to further supplement breeding efforts. Previous industry reports have
demonstrated the successful transfer to peach of a wide range of useful genes from
almond and its wild relatives, including resistance to brown rot, aphids, leaf curl, as well
1
as genes conferring desirable effects on tree structure and fruit quality. Because genes
are physically linked together on the DNA strand, the transfer of the small pieces of DNA
containing desirable genes inevitably result in some undesirable linked genes being
transferred as well. Over the last 6 years we have been involved in a rigorous, recurrent
(generation-by-generation) selection for desirable genes and against undesirable traits
as a way to purge out unwanted traits. This process, known as gene introgression, is
relatively inefficient in tree crops because of the long generation-to-generation time
periods involved, and because undesirable traits (such as high chilling requirement) may
be difficult to detect in any one season due to their ambiguous expression and/or
differences in expression in changing environments. The emerging biotechnologies
offer powerful tools to address these problems but are very expensive. In addition,
because this technology presently provides the 'cutting-edge' of our understanding of
genetic/developmental processes, the majority of current biotech funded research is
directed towards more basic questions such as identifying gene function (gene
discovery). Rather than a barrier, we saw this emphasis of biotech research on 'gene
discovery' as a potential advantage. A major (though frequently poorly appreciated)
problem for biotech-based gene discovery research is that to be 'discoverable' the target
needs to be a single gene whose expression is pronounced enough to be identifiable in
an otherwise chaotic genetic/environment background. Because of the inbred nature of
most of today's crops, the genetic variability is low and so presence of these novel
genes is less likely. For example, a biotech-based program to identify a commercially
useful brown-rot resistance gene in traditional California processing peach germplasm
would be unsuccessful simply because no such gene is present in the germplasm.
Molecular biologists are realizing that to increase the probability of success, they need to
examine genetic populations with the often conflicting attributes of being both genetically
diverse yet well characterized enough to demonstrate the presence of genes with
significant commercial value. During the last 15 years, the California processing peach
and almond breeding programs have successfully introgressed and partially
characterized a wealth of new genetic material available for crop improvement. By
pooling this resource with the demonstrated expertise in molecular and postharvest fruit
analysis of Carlos Crisosto's lab at the Kearney Agricultural Center in Parlier, California,
we have recently been successful in securing matching funds to current peach and
almond industry support to develop molecular tools to complement our more traditional
gene-discovery methods and so more efficiently exploit this rich germplasm. To
encourage collaboration with top-notch molecular researchers, (and to facilitate
paperwork), matching funds go entirely to the more expensive molecular research with
the applied breeding efforts utilizing the industry sponsored funding. A major goal of this
report is to document our progress in identifying and characterizing candidate genes for
more detailed molecular analysis. A second major objective is to illustrate how this
molecular technology can dramatically increase the efficiency and so effectiveness of
the applied variety development and regional testing programs. In addition, examples
of the frequently significant, though often poorly recognized, limitations of this new
biotechnology will be addressed.
2
Breeding objectives. Breeding objectives and their relative priorities are summarized in
Fig. 1. A short-term objective is
replacement of inferior varieties in the
Dixon-Andross and Halford-Corona
maturity periods. Longer-term objectives
Other
include development of commercially
New Traits
pre-Loadel
viable varieties ripening before Loadel
(to extend the processing period), the
Dixon-Andross
Impr. Quality
Other
development of resistance to fruit brown
rot and other commercially important
Halford-Starn
diseases, and the reduction in labor
inputs. In addition to the ongoing
Disease Resist.
improvement of current breeding lines
targeting these objectives, we continue
to evaluate new traits for their potential
value in solving current and future
Fig. 1. Breeding objectives: 2006.
production problems.
95
85
75
65
55
45
35
25
15
Number of Trees
Replacement of inferior varieties.
Progress in breeding efforts to replace inferior varieties is demonstrated by the recent
release and positive industry acceptance of the Andross-period variety Goodwin and the
Halford-period variety Lilliland, as well as the quantity and quality of advanced selections
now in grower tests (see Regional Testing report). Continued breeding progress can
also be gauged by the number of breeding progeny in each targeted category
demonstrating sufficient quality to merit further canning evaluation. Changes in the
numbers and distribution of UCD pilot plant processed selections is shown in Fig. 2.
Increased numbers of progeny have occurred in the before-Loadel and after-Corona
season-extension periods. (As
discussed in the regional
Ripe Date of Seedlings
testing report, many of the
1993 2007
post-Corona caning
evaluations are Halford70
Corona period selections
60
being tested for fruit structural
50
stability, both as post-ripening
40
holding ability on the tree, as
30
well as post-harvest storage).
20
Sizable increases in the
10
number of progenies can also
be seen in the targeted early0
harvest season (day 45 to 60)
and the extra-late harvest
Days from June 1
season (day 80 to 90). In
Figure 2. Distribution of processed breeding selections over the
1993, a distinct gap in the
2006 vs. 1993 seasons.
number of progeny maturing in
3
the Dixon-Andross period (day 45 to 55) is apparent despite intense efforts to select
parents targeting this season. Results from this early work suggested that the DixonAndross gap was the result of genetic deficiencies in the traditional California
germplasm and outside germplasm would have to be incorporated to overcome this
limitation. Accelerated efforts to incorporate new germplasm, particularly from eastern
Europe and Mexico, have been successful and significantly increased both the quantity
and quality of breeding progeny within this harvest season (Fig. 2). However, even in
advanced breeding lines derived from this new germplasm, a distinct decline in progeny
numbers is readily apparent which roughly corresponds with the maturity of the Dixon
variety (day 45). Our current working hypothesis is that this period represents a
threshold event for clingstone peach development (i.e. if the seedling tree fruit achieves
this stage of development by a certain time then development continues; if this
developmental threshold is not achieved by the critical time, then continued
development is delayed by 3-5 days). While the nature or even existence of this
threshold remains uncertain, we
speculate that it may involve the
transition from initial fruit cell
division to later cell specialization
(for example, cell lignification in the
development of the pit). We have
recently selected and placed in
regional grower testing a very high
quality processing peach,
designated Extra-Early#1, which
ripens at the targeted Dixon time
Figure 3. Extra-Early#1: Dixon-period but inconsistent
(Fig. 3). Evidence of a similar
maturity time?
bimodal distribution of harvest
time, even within this genotype or clone, has been collected in the last two years,
suggesting an inherent genetic/developmental instability at this ripening time. Other
evidence of developmental instabilities may be the preponderance of pit-breakdown (pit
fragments and red pit staining) common to this ripening period. Molecular fingerprinting
procedures are currently being used to examine the possibility that the two ripening
periods represent propagation mistakes. If it proves to be the same clone and it
continues to produce two distinctive phenotypes (ripening times) then a careful genetic
dissection using molecular techniques (as described in later sections) remains the only
option for characterizing and perhaps avoiding this problem. (See regional testing report
for more thorough evaluation of the Extra-Early#1, as well as the large number of
advanced selections in the Ultra-Early, Extra-Early, Early, Late, and Extra-Late maturity
periods).
Canning evaluations in 2005 and 2006 at both the UCD pilot plant as well as
cooperating commercial processors, have demonstrated an increasing number of
breeding selections possessing the fruit size, color and firmness required for commercial
acceptability (data and images available upon request). This larger pool of quality
selections has allowed the breeding program to put greater emphasis on attributes seen
as increasingly important in the maturing industry, such as disease resistance, fruit
structural stability with increased mechanical handling and storage, and uniform harvest.
4
Breeding for brown rot resistance. Progress in the breeding program for brown rot
resistance is demonstrated in the increasing number of advanced selections showing
improved field resistance (Ultra-Early#1, Extra-Late#4, Extra-Late#5, Extra-Late#6, and
Extra-Late#7) as well as the numbers and harvest season distribution of breeding
progeny screened in 2006 (Fig.
4). As pointed out by Dr.
Bostock in his accompanying
2006 report of laboratory
screening results for our
breeding progeny, uniformly
higher levels of brown rot
susceptibility were observed
during the later harvest season
relative to the early-season.
This has been attributed to
greater disease pressures at
these times (latent infections,
higher disease inoculum,
greater probability of rain
events, and greater stress on
late-season trees). The
skewed distribution also reflects
the distribution of our major
resistance parents (UltraEarly#1 ripens before Loadel,
Bolinha ripens midseason and
Extra-Late#5 ripens with or
Figure 4. Brown rot severity in 2005 & 2006 genotypes
after Corona). The unusually
arranged in order of sampling date.
cool early-season temperatures
followed by very hot midseason temperatures in 2006 also shifted the ripening date
distribution towards the late-season. The distribution of disease resistance over the
harvest season for 2005 is also shown in Fig. 4 where a more uniform resistance or
susceptible distribution is observed except in the third quarter of the season. This is
partly a consequence of our targeting resistance for the Extra-Early and Extra-Late
maturity seasons (as these are the times when rains and so disease explosions are
more likely to occur).
The importance of outside germplasm for developing resistance for California
adapted processing peaches is demonstrated in Fig. 5 where Dr. Bostock has sorted the
genotypes his lab evaluated in their order of resistance. Of the first 150 resistant
selections, all but three were derived from outside resistance sources. In addition,
examination of the inoculated fruit of these three lines indicates that they were picked
prior to the full-ripe stage when their resistance would be expected to be greater. To
guard against such false-resistances in immature fruit, we take color spectrum readings
of all tested fruit to quantify both their level of maturity, their remnant chlorophyll content,
and their potential for final fruit color. Other sources of resistance variability include
undetected damages to the fruit epidermis prior to inoculation, latent infections, poor
inoculation conditions in lab tests, and inconsistent levels of expression of the resistance
5
genes in different environments. The combination of these different and often
independent sources of variation can result in considerable year-to-year variability in
resistance scores for
individual selections.
To buffer against such
environmental and
genetic variability, we
are attempting to
combine resistances
from several
independent sources.
Fig. 6 shows the fruit
from several advanced
resistance lines as well
as documented
resistance mechanisms
identified in Drs.
Bostock’s and
Adaskaveg’s labs.
Figure 5. Composite of all genotypes evaluated during 2006 illustrating the
Mechanisms include
range in disease severity values. Values are presented from lowest (most
(from top to bottomresistant) to highest (most susceptible).
small picture sidebar in
Fig. 6) thicker layers of epicuticular waxes, denser epidermis cell walls, resistance to
epidermis and mesocarp cell wall breakdown with infection, and higher levels of antifungal phytochemicals such as phenolics. Currently, our resistance breeding strategy
involves combining resistance from multiple sources and selecting the resultant progeny
with the highest level and most stable field resistance. Since total resistance rather than
individual resistance mechanism is measured, the integration of multiple resistances is
not as efficient as we would like. While the histological and biochemical analysis of
individual mechanisms is overly tedious and beyond our level of expertise, the
identification of the underlying controlling genes would allow their individual identification
and selection using standard high throughput molecular screens. We now have
promising candidates for two of the resistance mechanisms/genes (anti-fungal
biochemicals with Drs. Bostock and Crisosto, and cell wall breakdown enzymes/genes
with Dr. Crisosto). Identification of the specific genes responsible for resistance would
allow their direct selection even at the seedling stage, thus allowing a much greater
efficiency and so breeding progress for brown rot resistance. This continuing integration
of molecular techniques for applied disease resistance breeding highlights 4 crucial,
though often unappreciated, requirements for success: 1) resistance has to be present,
2) resistance needs to be in advanced breeding lines of sufficiently good commercial
quality to assess field value, 3) the mechanism of resistance needs to be sufficiently
well-characterized
Figure 6. Sampling of different resistance lines as well as resistance components.
in identified
candidate genes,
and 4) the trait should be readily transmitted to progeny (i.e. controlled by 1 to 3 genes).
These requirements have also been key to our recent joint efforts to exploit molecular
markers to breed for greater resistant to fruit textural breakdown resulting from disease,
6
physical trauma, and postharvest storage.
Breeding for improved fruit flesh integrity using molecular markers. Fruit flesh
integrity defines the processing and fresh market peach industries. In order to achieve
the durable fruit textures needed for bulk handling, the processing peach industry almost
exclusively utilizes the nonmelting/clingstone peach rather then the melting/freestone
peach used for fresh market. The clingstone trait results in a poor separation of the
flesh from the stone, often leaving highly problematic lignified pit fragments in the
processed fruit. While the stone adhesion and flesh-melting traits appear independent,
clingstone is almost always inherited with non-melting flesh and freestone is always
inherited with melting flesh, indicating that they are controlled by the same gene or by
two genes which are so closely linked together as to behave as one unit in their
segregation to progeny. Even in the more durable clingstone/non-melting peaches,
textural integrity begins to collapse within days after full-ripe, requiring multiple harvests
to ensure ripe fruit durable enough for transport and processing. Recognizing that the
linkage between stone adhesion and fruit firmness had not been broken in efforts by a
large number of long-term breeding programs using traditional peach germplasm, our
processing peach breeding program began a program over 10 years ago to transfer
novel non-melting flesh types from almond and related species. By the early 2000's we
had developed a rather wide selection of novel freestone/non-melting to clingstone/longlasting fruit types, many of which were shown to be transferable to breeding progeny. In
the early 2000's we had also generated large number of progeny from the cross
between the Dr. Davis clingstone/non-melting peach and the Georgia Belle
freestone/melting peach as part of an informal collaboration with Dr. Crisosto's lab at
KAC to study the textural, biochemical and genetic nature of stone adhesion/flesh
melting traits. This genetic material, having already been initially characterized
morphologically and genetically (in terms of field inheritance) provided the foundation for
our successful BioStar proposal to characterized these fruit textural differences at the
molecular level and so allow a more precise understanding of the mechanisms involved
as well as their manipulation in applied breeding. In the following sections, the
chronology of this research is presented to demonstrate both the procedures and
potential of molecular tools to applied to plant breeding.
Collecting traditional and developing novel fruit firmness/pit adhesion types for
morphological, genetic and molecular characterization. The two major peach fruit
types in traditional varieties are the freestone/melting and clingstone/nonmelting types
(Fig. 7). A third type, clingstone/melting, can be recovered in certain breeding lines but
is almost always discarded as it appears to combine the worst commercial aspects of
both types. A prototype of the ideal type, freestone/nonmelting, as been recovered in
our breeding program from material introgressed from almond, as have ‘long-keeper’
processing peach selections showing a clingstone/nonmelting fruit type in which fruit
softening after the full-ripe stage is dramatically reduced when compared even to
clingstone-nonmelting types (Fig. 7). A comparison of fruit structure with softening
patterns further shows the distinctiveness of these 4 fruit types (Fig. 7). The lower thinsectioned images were allowed to dry slightly, as it allows a clearer perception of the
distinction between the inner fruit flesh (often visualized as straight strands radiating
outwardly from the pit) from the outer flesh region adjacent to the skin (where radiating
Figure 7. Different peach fruit stone adhesion and flesh melting types and graph showing
puncture/shearing force for ripe fruit from skin (left) to pit cavity (right).
7
strands of tissue often branch into a more complex tangled mass of strands growing
more perpendicular to the pit surface). The structural differences are also apparent in
graphs showing the amount of force needed to penetrate from the skin into the central
pit area (Fig. 7). In all fruit types the greatest force, and so most resistance to
penetration, is seen at the epidermis (though only slightly so for freestone/melting). At
full ripe, both the outer and inner tissue of freestone/melting is largely present as
disintegrating strands, and so presenting minimal resistance to the force gauge. By
comparison in our freestone/nonmelting type, the stone is clearly air-free but the
firmness of both the outer and inner flesh is comparable to clingstone/nonmelting (at this
full ripe stage). (As presented in previous reports, however, these types will soften to
almost freestone levels by day 4 and 5). The clingstone/nonmelting image shows both
the adhesion of a ' blanket' of flesh strands running across the pit surface as well as
attached to the outwardly radiating strands which make up the inner and outer flesh. A
high flesh firmness or integrity is apparent throughout the flesh at the full-ripe stage.
Finally, in another novel type derived from almond germplasm, the fruit behaves similar
to a clingstone/nonmelting at the full-ripe stage but, as has been documented in
previous reports (as well as the 2007 regional testing report), the fruit flesh maintains its
firmness or integrity for two weeks or more after the full ripe stage. The development of
these breeding lines demonstrated both that unique genetic controls were involved and,
further, that the inner and outer tissue showed different ripening patterns in different
genetic backgrounds.
EndoPG as a prime candidate gene controlling fruit softening. Because
endopolygalacturonase (endoPG) was previously shown to be associated with softening
in peach, the forms of this gene were studied in our novel breeding lines as well as
standard freestone/melting varieties, canning clingstone/nonmelting varieties and
progeny from the crosses between the two (Georgia Belle x Dr. Davis population).
Results are summarized in Fig. 8, where ESTs (, i.e. expressed sequence tags or
sequences of genes being actively expressed) of endoPG at fruit ripening are compared.
The two distinct dark bands associated with freestone melting flesh fruit types (FMF)
show that 2 separate but related genes of endoPG are being expressed (F and f) . In
the clingstone/melting flesh (CMF) types only the upper band was present though with
some individuals a closely paired, double-banding was apparent. The absence of the
lower FMF band in all clingstone (CMF & CNMF) types demonstrated the presence of
only one gene (‘f/f1’ in Fig. 8). Since peach is a diploid organism, (i.e. contains two
forms of each gene; one from the seed parent and one from the pollen parent), the tight
double banding in CMF fruit types indicates that different alleles or forms of the f-gene
were inherited from each parent (f/f1). Other CMF types inherited the same f-allele from
both parents (f/f or f1/f1). Significantly, in all processing-type clingstone/nonmelting flesh
(CNMF) varieties, the freestone or F-gene was either absent or nonfunctional leaving
only the f1 allele of the f-gene and in some cases no functional or expressed genes
(designated nn or null). The finding of this close correlation between fruit type and the
form of the endoPG gene appear to confirm our hypothesis that the endoPG gene is a
major player in determining fruit softening. The finding that a different form of this
endoPG gene appears to be responsible for the stone adhesion trait, was an added and
very exciting bonus. (Since endoPG is known to be responsible for the breakdown of
the intercellular matrix or glue that holds cells and cell strands together, its involvement
8
in fruit softening was intuitive. The possible control of the freestone/clingstone trait by
endoPG was also anticipated as a possibility, since, as is apparent in Fig. 7, the
freestone trait essentially involves a breakdown of the connections between the pitchannel strands and the pit-surface). As a final bonus, by knowing the sequence of the
endoPG-genes, we were able to develop PCR-based molecular fingerprinting
techniques which allowed the rapid and high-volume characterization of breeding lines
and varieties.
Rapid PCR-based molecular surveys of breeding lines and varieties. Using the
very rapid and efficient PCR approach for molecular (DNA) fingerprinting we were able
to evaluate over 100 varieties and breeding lines from the peach and almond breeding
programs. Since over 70 of these items were commercially important processing and
fresh market peach varieties, we believe the results (summarized in Fig. 9) represent an
accurate general picture of the genetic control of fruit softening in current peach
varieties. Results from the analysis of all commercial peach varieties were consistent
with our hypothesis that the endoPG gene had duplicated itself early in the evolution of
peach and almond in that one form had evolved to control stone adhesion while the
other form evolved to control fruit softening at ripening. Because this duplication
Figure 8. Gel banding patterns identifying different endoPG type genes (F & f) as well as different forms or
alleles of the f-genes, and their associated fruit types.
typically occurs as a copying error when the DNA is transcribed from mother to daughter
cells, they are very closely linked together. This close linkage explains why
freestone/melting and clingstone/nonmelting are always inherited together since the
probability of a random DNA break/rearrangement to separate freestone from melting,
for example, is highly unlikely within this very short distance. While both genes were
9
found to have multiple, often subtly different, forms or alleles, whenever any allele of the
F-gene was present, the fruit was freestone and melting flesh. Clingstone fruit were only
present when the F-gene was completely absent or nonfunctional allowing the f-gene,
whose alleles further determine flesh melting characteristics, to be expressed. Different
f-alleles determined different fruit types; if any f-alleles are present, then the clingstone
flesh is melting. If all alleles are f1-alleles or null-alleles (absence of functional gene)
than the flesh is clingstone/nonmelting. This advanced working hypothesis is also
consistent with previous genetic studies were freestone/melting (F/{any f-form}) is
dominant to (i.e. masks any expression of) clingstone/melting and
clingstone/nonmelting. Consequently, the freestone/melting trait is rapidly purged out of
any processing peach breeding program because 1) it is an undesirable commercial
type and, 2) it hides or masks the expression of the commercially important f-gene.
Similarly, because the f-allele, (which confers clingstone/melting and is commercially
undesirable), masks the expression of the commercially important f1-allele, it is rapidly
purged from processing peach breeding programs.
Figure 9. Sumary of fruit types, firmness & molecular data supporting the working
hypothesis of 2 separate multi-allelic genes controlling pit adhesion and fruit softening in
California processing and fresh market varieties.
10
EndoPG expression and activity. We are presently comparing the levels of endoPG
expression and activity in the different tissues of fruit types studied and the initial results
support our working hypothesis and, in addition, suggest a strategy for developing
improved processing peach types. Freestone/melting types consistently show very high
endoPG gene expression and biochemical activity in both the inner and outer flesh. In
nonmelting/clingstone types endoPG expression and activity are 50-fold lower and
located predominately in the inner flesh adjacent to the pit. Interestingly, in
nonmelting/clingstone genotypes completely lacking any functional alleles of the F-gene
and the f-gene (i.e. null or an nn alleles), endoPG activity was almost completely absent
and post-ripening fruit softening, while still occurring, was at the lowest level of any type
measured.
A high level of agreement between molecular fingerprinted genotypes, fruit flesh
endoPG expression and activity, and final fruit type, further supports our working
hypothesis. These molecular tools for fingerprinting (i.e. identifying specific forms of
specific genes) provides the breeding program with both precise knowledge of the
genetics involved as well as a more integrated knowledge of how these genes interact to
determine different fruit types. Thus, these molecular approaches offer powerful tools to
dissect or separate individual mechanisms conferring fruit firmness as well as allow a
more efficient recombination of the most desirable genes to achieve commercial goals.
While a powerful tool, however, this type of molecular dissection has certain limitations
in applied breeding programs because, on a whole-plant level, individual genes
frequently have multiple (and often unanticipated) consequences, and because different
developmental processes such as fruit ripening, often have multiple pathways to achieve
developmental goals. The final section of this report will document examples of such
situations to highlight their potential barriers to achieving breeding goals.
Dangers of over-dependence on molecular-based strategies for crop breeding.
Based solely on our molecular findings, the development of null endoPG fruit types (i.e.
complete absence of fruit flesh endoPG gene expression and biochemical activity) would
be a highly desirable goal for
processing
nonmelting/clingstone peach
breeding as it is associated with
the lowest rate of post-ripening
flesh softening. In fact, for the
past 10 years our breeding
program has been using the
parent Fla9,20-C (a very lowchill clingstone/nonmelting
fresh market peach developed
at the University of Florida from
Mexican germplasm), which
has been identified as a double
null. [It was the early analysis
Figure 10. Fruit of Fla9,20-C showing the intense red-stained
of this breeding selection which pit associated with the null condition for the endoPG gene.
first identified the occurrence of
null alleles for this gene]. Fig. 10 shows fruit samples of this selection collected 2004.
11
While conferring desirable fruit firmness, the trait is also associated with an intense red
pigmentation in the pit cavity in both the parent as well as breeding progeny. We are
presently using the molecular tools developed to target null genotypes in advanced
breeding lines in which these undesirable consequences might be minimized. For
example, we have recently identified another gene (designated high-liter) which appears
to shut down all red anthocyanin pigment production in the flesh and so could eliminate
the red-pit problem. (However, the presence of high-liter, itself, appears to be
associated with a poorer gloss or luminosity of the processed fruit, which is presently
perceived as one of the hallmarks of California processed peach quality). Another
breeding line identified as having very low post-ripening fruit softening rates were the
'stony-hard' selections derived from the Chinese Yumyeong peach. Suppressed
softening appears to be to due an independent mechanism here, possibly the shutting
down of the ethylene trigger for fruit ripening. The descriptor stony-hard refers to the
crisp, almost a pear-like texture of the ripe fruit. A number of stony-hard breeding lines
were tested in our program in the early 1990s before being discarded because of the
associated undesirable decline in processed fruit textural qualities (fruit were firm but
‘woody’ in texture).
While the molecular identification/dissection of endoPG-based flesh softening has
identified promising endoPG combinations and provided the tools to generate/monitor
their precise recombination, this molecular approach has also proved important in
identifying fruit firmness pathways independent and so possibly complementary to the
endoPG activity. (Though it was only with the recent application of the endoPG
molecular analysis that we became aware of the independence and so opportunities of
these alternative pathways). An example is seen in the advanced 'long-keeper' ExtraLate breeding selections (Extra-Late#4 to#7) presently being placed on regional grower
evaluations. Although possessing
the f1-allele, associated with
typical clingstone/nonmelting
processing peaches, these
selections appear to suppress
endoPG softening of ripe fruit to
levels comparable to the double
null and stony-hard genotypes,
yet without the undesirable
characteristics (Figs. 7 & 11).
(Because of the recent almondorigin of these lines, however,
opportunities for undesirable
linkages are relatively high,
requiring thorough testing at both
the field and molecular level
Figure 11. Extra-Late#5: a fruit type that maintains high
before consideration for varietal
fruit firmness independent of the endoPG gene.
release, (see discussion on
Regional Testing report).
Future research. Fruit softening via the endoPG and complementary pathways, while a
major objective of our collaborative efforts with Carlos Crisosto's molecular group
represents only one of the research avenues being evaluated within the larger fruit
Fig. 12
12
quality context. With Cameron Peace and Ebenezer Ogundiwin, the molecular
biologists directing the molecular testing, we are targeting a large number of traits of
commercial importance to processing peach and almond. In addition, we have
leveraged this research momentum to enlist the cooperation of a large number of
molecular biologists doing similar research at other national and international programs.
The combined efforts of these basic and applied researchers have resulted in the
development of an ambitious 'roadmap' or plan for a more aggressive application of
molecular tools to achieve applied breeding goals. Central to this progress has been the
development of a relatively complete genetic map of peach and almond (Fig. 12) by
combining the research results from these national and international programs. The
resultant genetic linkage map shows the location
Fig. 13.of a large number of different
molecular markers (i.e. the specific DNA sequence as determined by different molecular
approaches, PCR, SSR, etc.) as well as the location of known genes controlling
important traits. The map can be used to evaluate the presence of desirable and/or
undesirable linkages for candidate traits. For example in Fig. 12, the stick image labeled
'G 4' represents chromosome #4. Towards the bottom of the image, we can see the
location of endoPG gene determined from our research. The location or locus for the
'Bi' gene controlling internal flesh bleeding is very close to endoPG indicating that in
certain selections (as a hypothetical example, the red-stained double null selection
Fla9,20-C described above) specific desirable forms of the endoPG allele (the null form
in this case) may be physically linked and so almost always inherited with an undesirable
forms of the Bi gene (conferring red anthocyanin bleeding in the pit area). It this
association were confirmed, breeding programs could target the breakage of this linkage
using molecular markers to select even at the seedling stage (and so eliminating the
need to grow the very large numbers of plants required to field maturity). Alternatively,
the breeder could target the CS (color around stone) gene known to be located in the
central portion of chromosome 3 (G3 in Fig. 12) or the yet to be mapped high-liter gene,
to suppress all red anthocyanin production in plants whose other genetic components
would encourage it (i.e. shut down anthocyanin production upstream in the
developmental pathway).
13
Finally, the
high number of
molecular markers
on this peachalmond linkage
map provide
effective markers
to identify and
more thoroughly
characterized
additional traits of
commercial value.
For example, to
characterize the
previously
described high-liter
gene, we are
mapping both
molecular markers
as well as high-liter
trait expression in
progeny from
controlled crosses.
Due to the high
density of this map,
it is likely that
certain molecular
marker are located
very close to the
high-liter locus and so will almost always segregate with it in breeding progeny. Once
we have identified a molecular marker which co-segregates with the trait of interest, we
have identified the general location of that trait as well. Once identified, the entire
sequence of DNA within that region can be determined and individual embedded gene
sequences compared against known sequences of similar functioning genes. (Recent
advances in genomics and informatics have made accessible huge databases of known
gene sequences and function, as well as computer software allowing the rapid and
efficient comparison with hundreds of thousands of established sequences.) In addition,
genetic and molecular mapping at the very high resolutions now possible, can identify
the location (and through subsequent sequencing, the identity) of currently unknown
genes. An example is seen in Fig. 13 where the level of association between specific
molecular markers (here on a small segment of chromosome 5), and the trait of interest
(flesh browning) was compared over multiple years in our Georgia Belle x Dr. Davis
segregating population. The level of flesh-browning is normally difficult to evaluate
because it can very greatly depending on environmental factors such as temperature
and tissue stress. A multiyear, molecular mapping of this trait, however has identified a
gene locus which is estimated to account for approximately 40% of all flesh browning
observed in this population. In addition to identifying the presence of this gene, this
14
approach has also identified the specific location ( the qP-Brn5.1 region on chromosome
5) of the gene which, in turn, allows its sequence determination and ultimate
identification using procedures described above. The sequencing of the flesh-browning
locus will also allow the development of very precise molecular probes which can identify
other forms of this gene which may be inactive (no flesh browning) in the breeding
population (assuming such a gene exists). The sequence will also allow the
development of transgenes which might effectively turn off this gene in future genetically
engineered varieties. Equally important, using the huge genomic and informatics
databases developed in ongoing national and international molecular research, the
structure and putative function of the gene in question can often be predicted.
Increasingly, this 'forward-looking' approach to determining gene function from DNA
sequences can also be used in the reverse direction. For example, if the ongoing
research by Rick Bostock's lab identifies a specific phenolic compound associated with
resistance in processing peach, we might be able to predict a rough sequence for the
controlling gene by examining gene sequences for that or similar phenolics from other
plants (Arabidopsis, corn, tobacco, etc.) that are presently is listed in the genomic
database. We could then use that rough sequence to develop molecular probes to
specifically tag and efficiently identify the desired gene form even in seedling
populations or in environments not conducive to disease screening.
In summary, results from the first year of research have shown impressive
increases in our genetic knowledge and so breeding potential. However, for proper
application this knowledge needs to be tempered with field-based experience in the
traits of interest, particularly concerning potentially undesirable associated effects. A
second danger from the unbalanced use of molecular approaches to pursue breeding
goals is the risk of canalizing breeding efforts (for example, targeting the null form of
endoPG as the sole strategy to control fruit softening). Not only does such a narrow
perspective increase the risk of ‘dead-end’ research when unacceptable negative
consequences are recognized, but it eliminates the opportunities to identify independent
genetic solutions which may avoid undesirable associations. We believe this project has
achieved its goals of dramatically increasing molecular tools available to applied peach
breeding (at no additional cost to industry sponsors) while concurrently allowing
significant advances in our understanding of basic development pathways by exploiting
the inherent genetic variability developed by applied breeding programs in their
(typically) multifaceted approaches to targeted breeding goals.
15