ESF Darling Chestnut Science Summary
ESF Darling Chestnut Science Summary
December 15, 2023
Summary
This page includes summaries of scientific studies on Darling 54 chestnuts. Two separate chestnut blight inoculations (controlled, replicated intentional infections with the blight fungus) confirm that Darling 54 trees have smaller cankers than their non-transgenic siblings. Height measurements in well-replicated plots show that Darling 54 trees are shorter than their non-transgenic relatives, but that the magnitude of this effect varies by family. Updates available soon will include further details and statistical test outputs. Examples of homozygous Darling 54 trees (those which have inherited copies of OxO from both parents) are shown from two controlled pollinations. Finally, more detail is provided regarding the insertion site of OxO in Darling 54 compared to Darling 58.
Blight resistance trials: intentional blight fungus inoculations
Potted first-year seedlings (small stem assay)
This study was designed to test whether “stacking” OxO in hybrid backgrounds might result in increased blight tolerance. Canker length was measured two months after the inoculation (Figure 1). The most consistent result was that trees with OxO (regardless of hybrid background) had smaller cankers than their non-transgenic siblings. In addition, fewer OxO+ seedlings had visible, active cankers, suggesting some level of resistance to initial infection (Figure 2).
This inoculation was a blind study, meaning inoculations and observations were done without the tree identity or treatment group known to the data collectors; this ensures rigor and eliminates potential bias.
Figure 1: Canker length vertically along the stem after blight inoculations on potted seedlings. For all tree types, those with OxO (green boxpots) had much smaller cankers than those without (brown boxplots). Other control groups without OxO are shown in the right panel
Figure 2: Proportion of seedlings in each category with distinct, active blight infection (“takes”). OxO+ trees had fewer takes than OxO- and control trees.
Four-year-old trees in field plots
Four-year-old field trees were inoculated with blight fungus in June 2022. All Darling 54 trees (OxO+) were paired with non-transgenic full siblings (OxO-) from the same crosses. Cankers were measured throughout the summer to monitor their development (Figure 3). After three months, cankers on OxO- trees were more than three times larger on average than their Darling 54 siblings. There was also a substantial amount of variation in canker size between families (Figure 4).
As with the small stem assay above, only clearly visible cankers (“takes”) were included in these analyses, and there were substantially fewer apparent infections than on non-transgenic trees.
Figure 3: Canker growth over time. Cankers remained smaller and grew more slowly on OxO+ trees (green) than OxO- trees (brown).
Figure 4: Canker size 13 weeks after inoculation, for OxO+ (green) and OxO- (brown) siblings in each family. Cankers were consistently smaller in OxO+ than OxO- trees, though size varied more than in the potted small stem assays.
Tree height measurements
Common Garden Trials
A “Common Garden” is a plot design where the same tree types are planted in different locations to observe potential environmental effects on tree growth. This is also an ideal design to compare growth of several tree types at a single location. ESF’s Common Garden plot contains approximately 500 trees, including several diverse Darling 54 full-sibling families (OxO+ and OxO-), along with non-transgenic controls (Backcross, hybrid, and full Chinese). Additional plots with these same tree types were planted in Maine and Virginia.
These charts show heights of 3-year-old OxO+ (green) and OxO- (brown) full-sibling families growing at SUNY ESF, measured in fall 2023. The average effect of OxO presence is a ~20 cm (~ 8-inch) reduction in height (Figure 5). Differences in height vary considerably when OxO+ and OxO- trees are compared in separate families (Figure 6), and family background can affect height of both OxO+ and OxO- trees. These experiments are in early stages and will continue to be monitored for many years.
Figure 5: Height comparison between OxO- (brown) and OxO+ (green) trees in the ESF Common Garden plot. Darling 54 trees with OxO were on average shorter (~75cm mean height) than their non-transgenic siblings (~95cm).
Figure 6: Height comparison for OxO- (brown) and OxO+ (green) trees by family. Though on average OxO- trees are taller than OxO+ trees, there is considerable variation by family in both tree height, and the differences between transgenic and non-transgenic siblings. Tree heights for other non-transgenic control groups in the ESF Common Garden experiment are shown in the right panel.
Growth analysis in a replicated experiment across three states (NY, PA, VA)
The first experimental plots to include Darling 54 seedlings, non-transgenic full siblings, and other controls were planted in 2019, and replicated in NY, PA, and VA. Patterns of height growth varied across locations (states) and field settings (open field vs. wooded sites in NY) (Figure 7). When measurements were combined across all sites, average tree height as of 2021 was 67 cm for B3F3 trees, 65 cm for OXO- trees, and 43 cm for OXO+ trees. Thus, OXO+ were ~20-25cm shorter than other tree types. As with the Common Garden trials above, this work is ongoing, and this is a preliminary visualization of this dataset. More recent data (2022 and 2023) will be integrated soon.
Figure 7: Height growth trajectories for trees in open field plots in New York, Pennsylvania, and Virginia, and a shelterwood (partially wooded) plot in New York. Values reflect the mean and standard error (n = 24 observations per tree type per year).
Darling 54 and Homozygosity
Successful homozygous plant propagation
Classical genetic theory would predict that when two Darling chestnuts are crossed, 25% of the progeny would be expected to be homozygous (inherit two copies of the transgene). We measured how many transgene copies were present in 19 transgenic chestnuts obtained from six Darling x Darling crosses and identified three homozygous offspring. These homozygous shoots are being propagated in tissue culture and grown in pots. Our experience with these lines so far demonstrates that homozygous embryos develop into functional plants (Figure 8). This is fewer homozygous offspring than expected, but defining homozygosity rates would be premature at this stage due to small sample sizes from each cross to date. Darling trees are still too young to develop large numbers of female flowers. As they mature and become more prolific, ESF scientists and collaborators will be able to perform more crosses to better quantify homozygosity rates.
Figure 8: Production of homozygous plants via tissue culture at SUNY ESF. TGG004 shoots (A) two days and (B) one month after multiplication, respectively; (C) TGG002 & TGG004 plants 6 and 4 weeks after potting, respectively; TGA002 plants 7.5 months after potting in (D) the production chamber and (E) under high light for pollen production.
Transgene position in Darling 54
Darling 54’s transgene is located on Chromosome 4 within the sequence of a “Sal1 ortholog” (gene with high similarity to a gene known as Sal1 in other species). This gene has not been previously studied in chestnuts, so its function is hypothesized based on studies in other plants, where it has generally been linked to drought or salt tolerance. This insert location would not necessarily have prevented regulatory approval, but warranted further study of the resulting plants. Since development in 2016, numerous observations and environmental studies have been done on trees we now know are Darling 54 offspring, with no indication of compromised safety or risks that would affect pending regulatory approval. More in-depth genome sequencing and additional analyses on Darling 54 are underway and will be submitted to regulators soon.