March 2018 Issue of Wines & Vines

Grape Breeders No Longer Flying Blind

Low cost of DNA sequencing allows breeders to incorporate disease resistance

by Tim Martinson

Imagine you are a wheat breeder. Your overarching goal is to produce varieties that increase yield by incorporating superior genetic traits such as large heads with more seeds, short stems and disease resistance.
     You can make crosses of numerous elite parents that generate thousands of seedlings each year. Because each cross produces a family of hundreds of siblings, you can map the inheritance of yield components. It’s an annual crop, so you see results of your efforts within months. By taking the seeds to the Southern Hemisphere, you can cycle through two or more generations per year. Measuring yield is relatively simple—and cheap.
     This process has been remarkably successful. Winter wheat yields have increased by 0.4 bushels (bu) per acre each year since 1960, and wheat farmers now produce 45 bu/acre, up 80% from 25 bu/acre in 1960. To cite another example, corn yields have increased fivefold since 1930.
     Now imagine you are breeding wine grapes. Your goal is a new variety that combines powdery mildew resistance with excellent fruit quality and yield. You start by emasculating flowers (using tweezers to painstakingly remove pollen-producing stamens) and bagging each cluster to prevent stray pollen from fertilizing them. You then take pollen you have collected from a male parent and spread it over the emasculated flowers of the female parent. Each cluster might produce 50 to 150 seeds.
     You plant those seeds the next year and observe seedling plants, then discard those that are susceptible or have growth defects. You select the best 5% to 15% of seedlings and replant them to a vineyard. In another three or four years, you can finally evaluate fruit characteristics and discard another 90% of the vines. Soon you can sample wines made from the most promising single vine selections, and then throw out another 90%.
     Measuring yield and most other traits is costly, and a single year’s data is insufficient—three- to five-year averages are needed for this perennial crop. Other quality metrics (e.g., fruit chemistry) are important, as are sensory attributes of wines made from the fruit. Odds are you will need more generations to find a vine that combines all your desired traits.
     The bottom line is that, with grapes, it can take more than 15 years after the initial crosses to find out what you have, how it performs and to identify the most promising vines, which likely will be parents for the next generation. Crossing, selecting and evaluating new grapevines takes a lot more time and expense than evaluating a new hybrid strain of wheat or corn.

The numbers game and mapping
Progress in breeding is determined by numbers and time. Aside from the built-in time lag involved in growing grapes, the less obvious consequence is that each grapevine takes up more space and costs more to grow and evaluate than wheat or corn. Grapevines are planted at densities of less than 1,000 plants per acre, as opposed to wheat and corn at upwards of 100,000 per acre. And they require wire and trellis as they grow. That translates to much more expense for each cross and progeny evaluated.

     Wheat and corn breeders have historically been able to map the chromosomes and determine how traits are inherited much more easily—simply because they can crank out large families of progeny and evaluate them in one cropping cycle. Even before modern DNA sequencing, they were able to develop specialized genetic stocks (mutant collections, chromosome-deletion lines, near-isogenic lines) that provided tools for mapping genes.
     Grape geneticists, by contrast, faced a more complex task in determining how desirable traits were inherited and mapping these traits to specific locations on each chromosome. Finding even a few molecular markers for simple traits required years of effort. The first grape genetics maps were developed at Cornell University 25 years ago. At that time, it took five years to map a few traits, including genes for flower sex determination and “REN2” powdery mildew resistance.

Phenotypic selection
Because of this, grape breeding programs relied on observed phenotypes (vine characteristics), most often without much idea of their genetic basis. For example, some seedlings in Cornell’s “no-spray” block show remarkably few powdery mildew colonies on their leaves, but only now are we learning what gene or genes are associated with the observed resistance.
     The cost of maintaining experimental vineyards and the time it takes to evaluate phenotypes set an upper limit on how much genetic knowledge grape breeders could obtain. Compared to corn and wheat breeders, grape breeders have had fewer tools and a much more rudimentary genetic map in their toolbox. They have been flying blind.

The genetic revolution
and genetic markers

Sequencing technology has changed the equation. The cost of obtaining DNA sequencing information has fallen dramatically. For example, according to the National Institutes of Health (2017), the cost of obtaining a complete human genome sequence has dropped in the past 10 years from $10 million to just above $1,000, a 10,000-fold reduction in cost. The cost of sequencing each mega-base (1 million base pairs) now stands at a little more than a penny (1.2 cents), down from $500 in 2007. This dramatic drop in cost exceeds the historical doubling in computer chip circuitry every two years that “Moore’s law” famously predicted. DNA sequence information is now very inexpensive.
     For grape breeders and geneticists, the previous trickle of scarce genetic knowledge has turned into a flood of DNA sequence information. For the first time, there is enough sequence information to allow geneticists to make a detailed map of the 19 pairs of chromosomes and 500 million base pairs in the grape genome. This map enables them to locate genetic markers (short DNA sequences) associated with single gene loci, or what they call Quantitative Trait Loci (QTLs).
     Using inexpensive tests of the DNA from each seedling to detect these genetic markers, they can identify which seedlings resulting from their crosses are carrying the genes of interest. Instead of waiting years to identify which vines have the desired traits, they can find out almost overnight which seedlings carry the desired genes by snipping out a small leaf sample, extracting DNA and carrying out inexpensive lab-based analyses.
     To date, the USDA-funded VitisGen project has identified more than 70 marker-trait associations for disease resistance (powdery mildew, downy mildew, phomopsis, black rot), fruit quality traits (anthocyanins, skin color, sugar and acid content) and more. These markers are being put to use in selecting or discarding seedlings before planting them out in the field. Since 2014, more than 16,000 seedlings have been screened using marker-assisted selection.
     Moreover, DNA markers now give grape breeders the tools to stack multiple genes for disease resistance into the same variety. Field observations of phenotypes could only tell you whether a seedling was resistant or susceptible—but not which gene was responsible. Now DNA markers can tell you whether one, two, three or more resistance genes are present in one seedling. This should allow breeders to develop stable, long-lasting resistance with multiple modes of action.
     Before automated DNA sequencing, extracting genetic information was like turning on a faucet and watching it drip slowly. Now that the cost of DNA sequencing is trivial, the flow of information is more like a torrent from a firehose. The challenge for breeders and geneticists is to correctly interpret and use this flood of information. From now on, grape breeders won’t be flying blind.

Dr. Tim Martinson, a senior extension associate in the Cornell University School of Integrative Plant Science, works with regional extension educators and industry groups to provide growers and wineries with educational programs, workshops, newsletters and applied on-farm research that supports profitable production of grapes, grape products and wine.

This research was supported by the USDA-NIFA Specialty Crop Research Initiative (Award No. 2017- 51181-26829).

National Institutes of Health 2017. DNA Sequencing Costs. National Institutes of Health, National Human Genome Sequencing Project.  Accessed at:

Garris, A., L. Clark, C. Owens, S. McKay, J. Luby, K. Mathiason and A. Fennell. 2009.  Mapping of Photoperiod-induced Growth Cessation in the Wild Grape, Vitis riparia.  J. Amer. Soc. Hortic. Sci. 134:2, 261-272.

Shanshan Yang, Jonathan Fresnedo-Ramírez, Qi Sun, David Manns, Gavin L. Sacks, Anna Katharine Mansfield, James Luby, Jason Londo, Bruce Reisch, Lance Cadle-Davidson and Anne Fennell. 2016. Next generation mapping of enological traits in an F2 interspecific grapevine hybrid family. PlosONE 11(3): e0149560. doi:10.1371/journal.pone.0149560.

National Agricultural Statistics Service, USDA, 2017. Crop Production Historical Track Records. Accessed at:

Print this page   PRINTER-FRIENDLY VERSION   »
E-mail this article   E-MAIL THIS ARTICLE   »
Currently no comments posted for this article.