Making a Gene Map For Roses 

Sriyani Rajapakse and Robert E. Ballard (Clemson University)
David H. Byrne (Texas A&M University)

Clemson University and Texas A&M University have initiated a collaborative research project on gene mapping in roses with the long term objective of identifying genes involved in black spot resistance and fragrance. This article describes this project which is partially funded by the American Rose Society Research Trust. In the following, we will briefly examine what is a gene map, describe the two main traits that we are studying, explain the problems faced in studying these traits, and report our progress in mapping.

What is a Gene Map?

Similar to how a road map shows the location of cities and towns, a gene map marks the location of genes on chromosomes. Genes that control various traits are located on pairs of chromosomes in the nucleus of the cells. The chromosomes in the rose nucleus occur in sets of seven. Some roses have two sets of seven, or 14 chromosomes, while others have 21 (three sets of 7) or 28 (four sets of seven) chromosomes, respectively. The purpose of making a gene map is to assign traits to one of these seven basic chromosomes and then to pinpoint the location of the gene on that chromosome. These genes, called structural genes, code for enzymes that produce various biochemical compounds that lead to the expression of the traits. For chromosome mapping, DNA markers associated closely with the genes (traits) are identified and are used as tags to show the presence or absence of these genes in the rose being evaluated. Although gene maps have been or are now being constructed for many economically important crop plants, animals, and humans, one has not yet been constructed for roses.

Any inherited trait of practical significance can be mapped. These may include flower characters such as petal color, number, size or shape, fragrance, plant growth habit, and resistance to pests and diseases. For the Clemson rose project, we are interested in mapping genes that control two important traits, resistance to black spot and floral fragrance.

How Will a Gene Map be Used?

Finding exact location of a gene can lead to its isolation from other genes in the chromosome. Once a gene controlling the trait of interest is isolated, the gene can be transferred through genetic engineering to other plants lacking this gene. The genetic manipulation of this trait can produce new varieties with novel combinations of desirable characters. A good example of genetic manipulation of a trait in ornamental plants is the modification of flower color. The most important pigments in floral coloration are the flavonoids. Certain flavonoids, the anthocyanins, are responsible for the familiar reds and blues of many flowers. Anthocyanins have been studied extensively using molecular, biochemical, and genetic methods. Several genes that code for major enzymes that mediate the production of anthocyanins have been isolated (Davies and Schwinn, 1997). These genes have been used to genetically engineer flower color in a range of plants including petunia, gerbera, and chrysanthemum (Meyer et. al, 1987, Elomaa and Holton, 1994, Courtney-Gutterson, 1994).

Plant breeding using marker assisted selection is a second way to use a gene map. If you know where a gene is on a gene map, DNA markers that flank the gene of interest can be identified. Just like mileage markers on an interstate highway indicate where you are on the highway, the presence of the DNA markers would infer the presence of the desirable gene in the offspring. DNA markers can be detected soon after seed germination, as soon as a few leaves are available to use for DNA marker analysis. This process makes early detection of the presence of the gene possible even though the actual expression of the trait does not occur until much later in maturation. This is especially valuable, for example, when working with genes that control flowering in plants that take many years to bloom and fruit or traits such as disease resistance that take several years to evaluate. Therefore, DNA markers tightly linked to a gene can be used by breeders to select and cull breeding lines. Also if several genes exist for resistance to a disease, e.g., black spot, DNA markers can be used to "stack" these genes into one rose.

Additionally, once a gene is identified, its function and the biochemical steps leading to the trait being expressed can be studied and its genetic control understood.

How is a Map Constructed?

Gene mapping is a long-term project starting with identifying the genes to be mapped and creating appropriate study progeny. To be an informative progeny, individuals must show variation in the trait of interest, but otherwise should have a similar genetic background. For example, to find the genes responsible for resistance to black spot, some plants of the progeny have to be resistant to black spot while other plants have to be susceptible to the disease. The most common method to obtain this type of a progeny set is by crossing a resistant rose with a susceptible one. The offspring in the first generation (F1), or second generation (F2), or back cross progeny (BC) can be used in generating a map.

Evaluation of the trait in the progeny is necessary to ensure that the trait is variable and informative for mapping. Seedlings of the selected progeny are screened for resistance to the disease in the field. DNA markers are developed from the same progeny and those markers that are present only in the plants resistant to black spot and are absent in susceptible plants will be selected to tag the genes. The tightness of the association between the marker and the gene determines the effectiveness of the use that DNA marker as a tag in marker-assisted selection. Even the markers that are not associated with the trait are important in mapping, since they provide a framework for the chromosome maps on which genes are located. Mapping of all DNA markers and traits on chromosomes is carried out by a process known as linkage analysis.

Black Spot Resistance in Roses

Black spot is a major disease of garden roses characterized by irregularly shaped black spots along with leaf yellowing. This can lead to leaf defoliation, loss of plant vigor, less flower production and, in extreme cases, the death of the plant. It is currently controlled by the frequent application of fungicides. However, some roses have natural resistance to the disease.

Over two years of field testing at two sites in Texas: College Station (David Byrne) and Overton (Dr. H. Brent Pemberton) of Floribundas, Grandifloras, Hybrid Teas, Shrubs, Rugosa Hybrids, Species/hybrids, and Basye roses showed a wide difference in the reaction to black spot which ranges from death of the plant within two years (‘Peace’) to no spots at all (Rosa roxburghii). These replicated trials were rated 3 to 4 times a year for the percent of leaflets with black spot lesions and the amount of resultant defoliation on a 1 to 10 scale. The highest level of resistance was found within the Shrubs, Rugosa Hybrids, Species/hybrids, and Basye Roses classes.

Exploiting this genetic resistance and its transfer to susceptible Hybrid Tea and other roses is more effective and environmentally safe means for controlling the disease. In a previous American Rose article we have examined this aspect in detail (Rajapakse and Ballard, 1996). Up to now there has been little sustained effort in the development of black spot resistant roses. Although the transfer of this resistance is not a trivial matter and will take time, it is possible with a sustained and coordinated effort. Given this, it is important to identify and mark these resistance genes. The mapping work will help identify molecular markers for the specific resistance genes which can be used in combining multiple genes into one genotype to give it a broad based level of resistance. A rose progeny to study this trait has been produced by David Byrne and it is currently being used to develop a genetic map at Clemson University.

Fragrance in Roses

Fragrance is one of the most subtle, yet a familiar feature that characterizes many roses. Rose species and cultivars display substantial variation in scent as well as intensity of fragrance. Fragrance is a result of numerous volatile aromatic organic substances present in the flower. These substances include hydrocarbons, alcohols, aldehydes, ketones, esters, ethers, and other miscellaneous compounds (Flament et al., 1993). To be able to manipulate fragrance in roses through genetic engineering, the chemicals contributing to the fragrance of roses must be distinguished, their pathways of synthesis understood, and enzymes controlling these pathways identified. Then the genes controlling these enzymes have to be isolated and characterized. Gene mapping is a means to locate such fragrance genes and to identify the DNA markers associated with these genes.

Current biochemical understanding of fragrance in plants in general is poor, when compared to our knowledge of flower pigment formation. As mentioned before, flower color has been studied extensively and some genes that encode specific enzymes in the biosynthetic pathway of plant pigments have been isolated and have been used to manipulate flower color. However, no example is available now for genes that encode for enzymes related to fragrance. One major limitation to understanding the contribution of specific chemicals to fragrance is that unlike plant pigments, fragrance is not a result of just one or a few compounds, but is the result of the blending a large number of aromatic substances. More than 4000 volatile components from rose oils and petal extracts have been identified by gas chromatography and mass spectrometry (Flament et al., 1993)

Problems and Prospects in Studying Fragrance in Roses:

Two requirements are essential to study the inheritance of fragrance in roses and to map the trait. First and foremost, a progeny set of roses, parents and offspring, segregating for the trait is required. Without appropriate progenies, inheritance cannot be followed for any trait. Parental roses that differ in their level of fragrance will have to be crossed and offspring produced. In order to be an effective genetics and heritability study, either a F2 population obtained by self-fertilization of a selected F1, or a backcross population obtained by crossing one F1 with one of the parents will be required. Since it is very likely that many chemicals and therefore many genes may be involved, a large seedling population will be required to study through quantitative trait analysis of the many loci associated with fragrance. Once the progenies are evaluated for fragrance, genetic data such as variance components and heritability can be calculated.

Some of the problems encountered in producing such a rose progeny include insufficient knowledge about the fragrance in the parental lines, problems associated with self- and cross-incompatibility, differences in the ploidy level, low fertility of the F1 hybrids, ample seed production, seed germination, seedling establishment, and seedling growth and maintenance.

Such a progeny segregating for fragrance is not presently available for study. We welcome any populations produced by rose breeders that meet these requirements. There is a possibility that our current black spot mapping progeny might also segregate for floral fragrance, i.e., this population segregates for many other floral traits, and the parents have diverse genetic backgrounds. However, until fragrance traits are measured objectively within this population, we will not know.

This leads us to the second requirement in studying inheritance of fragrance, a standardized, objective method of evaluating fragrance. The major limitation in evaluating fragrance by chemical means is that more 4000 volatile, aromatic substances have been isolated from rose. However, all of these components may not contribute to fragrance as perceived by the human nose. Some components, which are present in minor quantities may be olfactively essential and may contribute more to the fragrance than some other components present in large quantities. We have not come across any study that attempted to relate these volatile compounds to human sensory evaluations. As the fragrance experienced by individuals varies from one person to another, a panel evaluation may be necessary for this purpose. The complexity of the nature of fragrance makes this a very difficult task. Short of these two essential requirements, it is difficult to make any progress in studying the genetics of fragrance methodically. Achieving each one of these will require greater effort and resources.

Our Current Research in Gene Mapping:

Early in the mapping research, we screened several roses from diverse genetic backgrounds as potential parents for producing our mapping progeny. Selection of parents is a critical step in the gene mapping process because there has to be sufficient variation between the parents that can be easily detected by using DNA markers. Without differences between the parents that are easy to score, progress in mapping is slow and tedious.

The parental plants screened exhibit a wide variability to resistance or susceptibility to black spot as well as in other floral characters. These plants are: (1) 86-7, a black spot resistant amphidiploid from two highly resistant diploid species, Rosa wichurariana X R. rugosa var. rubra; (2) 74-193, a black spot resistant rose from [R. carolina X Hugh Dickson] X [R. virginiana var. alba X Betty Morse]; (3) 82-1134 (Basye's Blueberry), considered susceptible (or sometimes moderately resistant) to black spot; (4) cv. Impatient, a tetraploid susceptible to black spot (5) cv. Angelface, a tetraploid moderately susceptible to black spot and, (6) Sunflare, a tetraploid susceptible to black spot.

DNA extracted from these roses was screened using a genetic marker analysis procedure known as RAPDs. Our results indicate that all of these parental plants exhibit a high degree of genetic polymorphism from each other, making any combination of parents suitable for producing mapping progeny. Therefore, selection of the parents for mapping was based on the number and fertility of F1 progeny produced, since the level of polymorphism does not appear to be a limiting factor in any cross between resistant amphidiploid and susceptible tetraploid parents. 

We are now analyzing the F2 rose progeny between 82-1134 and 86-7. This progeny segregates for black spot resistance as well as a number of other traits including flower color (pink vs. white), petal number (five vs. ten or more), presence or absence of thorns on the stems, and flowering time (once blooming vs. ever-blooming). DNA markers are now being developed in this progeny to create a framework of chromosome maps. A new DNA marker system was selected for this purpose as it is very effective in detecting differences between the parental plants and a large number of markers can be developed within a short period of time. Once the framework of the map is established, the markers will be analyzed for their association with the traits.

Chromosome                             Linkage Map


Courtney-Gutterson, N., 1994. The biologist’s palette: genetic engineering of anthocyanin biosynthesis and flower color. In: Ellis, B. E., Kuroki, G. W. and Stafford,H. (eds.), Genetic Engineering of plant Secondary Metabolism; Recent advances in Phytochemistry, Vol. 28. Plenum Press, New York, pp 93-124.

Davies, K. M. and Schwinn, K. E., 1997. Flower color. In: Geneve, R. L., Preece, J. E. and Merkle, S. A. (eds.) Biotechnology of ornamental plants. pp 259-294.

Elomaa P. and Holton, T., 1994. Modification of flower color using genetic engineering. Biotechnology and genetic engineering reviews 12, 63-88.

Flament, I., Debonneville, C. and Furrer, A., 1993. Volatile constituents of roses. In: Teranishi, R., Buttery, R. G. and Sugisawa, H. (eds.) Bioactive volatile compounds from plants. American Chemical Society Symposium Series No. 525. pp 269-281.

Meyer, P. , Heidmann, I., Forkmann, G. and Saedler, H., 1987. A new petunia flower color generated by transformation of a mutant with a maize gene. Nature 330, 677-678.

Rajapakse, S. and Ballard, R. E., 1996. Rose breeding and Biotechnology. American Rose Annual 82-86.