When you think of model organisms soybeans are probably low on the list. Large, long life cycles, resistant to genetic transformation and regeneration. I used to think that way, too. My career in plant biology started with the quintessential model organism, Arabidopsis thaliana. That species has a lot of properties that make it easy to work with and I do think Arabidopsis will be at the forefront of basic plant biology for the foreseeable future. But as I have come to learn, soybeans also can be good models in many ways. Partially due to work Mozza and our competitors have done but also some natural qualities make them good subjects.
Transformation
Historically soybeans have been a pain to transform. From the day of transformation to the time when you are able to work with homozygous T1 plant tissue is 6 to 9 months. That method which we call “cotyledon node transformation”, described succinctly here1 has been the go-to method. There are many slight variations on this method but they are all roughly the same. Recently there has been developments in getting that timeline down.
The embryonic axis method described in this paper2 is a significant improvement that reduces workload per event and shaves a few months off the timeline. The paper is lacking details so look to some of their patents that happen to provide better recipes. Also, knowing someone who has done it before helps significantly as the dissection technique is not obvious from the literature. This3 method from Syngenta gets the timeline to first T1 seeds down to 6-7 weeks. You can also find methods out there for easy somatic embryo transformation and regeneration that do work and get you mature seed tissue to work with in under 12 weeks.
If your research requires stably transformed seed tissue, soybean can actually beat Arabidopsis if you use one of the newer methods. Arabidopsis seeds are so small that you will probably need to grow them out to at least the T1 generations, possibly the T2 generation to have a stable insert and enough tissue to do biochemistry or sequencing on. Soybeans embryos are so big that you can genotype beans from your T0 plant and have enough tissue left over to do your work on. One bean is enough to do many many western blots, generate Illumina libraries and still have some left over. So you don’t need to take the plants out so many generations to do useful work. You do have to be clever with the genotyping (assessing copy number and zygosity) but it really does save a lot of time.
If you don’t need stably transformed tissue there is hairy root transformations that are relatively simple, for those of you doing rhizome research. And for ultra fast transformations, soybean cotyledons are at least somewhat agreeable to protoplasting and PEG transformations4.
Genetics
The genome of soybean is larger than Arabidopsis. But unlike corn, I have not found it to be much more complex. The intergenic spaces are larger but that doesn’t effect much. If anything, the probability of your TDNA landing in something important is lower, so you get more good events. This is also aided by the fact that most genes have either 2, 4 or 8 duplicates/homologs. so even if you land in the middle of a coding sequence, there is a high probability the plant has a backup gene so the event is not lethal or oddly-phenotypic. But what this also means is making mutant lines is near impossible in soybean. Single mutations rarely have any phenotype. Instead we have learned to use RNAi to simultaneously target all the homologous genes simultaneously. siRNAs don’t care that 2 different mRNAs came from different chromosomal loci. So overall, the genome is bigger and has many duplicate genes but it doesn’t really matter.
The ancestors of modern soybean varieties may have been polyploids but the modern varieties act like diploids, the quantity of mobile elements is much much lower than in corn and it does not readily out-cross. So genes segregate like you would expect with simple Punnett squares. The self pollination (and lack of hybrid vigor) makes it easy to maintain many stable lines in close quarters. But the difficulty in crossing does become a hurdle sometimes. We had to fly soybean breeders in to our lab in Los Angeles to teach us how to do it. So that’s one strike against soybeans as a model organism. We do make the crossing hard on ourselves by intentionally growing smaller plants, which means the flowers are smaller, which makes the process more difficult.
Greenhouse
The space required to grow soybeans is definitely greater than Arabidopsis. There is no getting around that. If you grow them indoors under LEDs then electricity, water, fertilizer costs all are significantly higher for soybean than for Arabidopsis. We also chose to grow in rockwool (popular in the cannabis industry) instead of soil which also adds expense. Some people choose to do glass houses with augmented light which works too.
To give you a sense of the scale we are looking at here, the rockwool Gro-Blocks are 15cm on each side. We clip the apical meristems which prevents them from getting tall and viney. With our lighting conditions and the clipping, the plants get about 30-40cm tall. The racks in our greenhouse have 4 levels and slide around like library stacks. We can fit around 3000 plants in a 800 square foot room.
If you try to transfer all of your Arabidopsis protocols to soybean, you will find that you need a greenhouse the size of a city block. The smarter strategy is to find ways to make do with fewer plants.
We do all of our genotyping on plants before they get transferred to these 15cm Gro-Blocks. As an example, if we collect the seeds from a T0 individual and want to replant and find only the homozygous T1’s, we stick germinated seedlings in really small rockwool cubes packed close together and grow them until the first trifoliates expand. Then collect tissue samples and find the homozygous T1s and only transfer 1 or 2 of those to the large Gro-Blocks. This practice keeps the required space for a given line to a minimum and keeps cost down. We try to only have plants in the Gro-Blocks if they will be taken all the way to seed or if embryonic tissue is going to be removed for analysis.
The amount of embryonic tissue you can get from a single plant is also much higher than Arabidopsis. So the need to grow many plants is reduced. When I worked with Arabidopsis many years ago I would always have at least one 1020 tray in the greenhouse per line I was working with. Overall the space required isn’t much different. Not sure about the opex per equivalent grow space, though.
Transient
Luth, Diane, Katey Warnberg, and Kan Wang. “Soybean [Glycine Max (L.) Merr.].” In Agrobacterium Protocols: Volume 1, edited by Kan Wang, 275–84. Methods in Molecular Biology. New York, NY: Springer, 2015 ↩︎
Cho HJ, Moy Y, Rudnick NA, Klein TM, Yin J, Bolar J, Hendrick C, Beatty M, Castañeda L, Kinney AJ, Jones TJ, Chilcoat ND. Development of an efficient marker-free soybean transformation method using the novel bacterium Ochrobactrum haywardense H1. Plant Biotechnol J. 2022 May;20(5):977-990. doi: 10.1111/pbi.13777. Epub 2022 Feb 24. PMID: 35015927; PMCID: PMC9055811. ↩︎
Zhong H, Li C, Yu W, Zhou HP, Lieber T, Su X, Wang W, Bumann E, Lunny Castro RM, Jiang Y, Gu W, Liu Q, Barco B, Zhang C, Shi L, Que Q. A fast and genotype-independent in planta Agrobacterium-mediated transformation method for soybean. Plant Commun. 2024 Dec 9;5(12):101063. doi: 10.1016/j.xplc.2024.101063. Epub 2024 Aug 13. PMID: 39138866; PMCID: PMC11671754. ↩︎
https://www.jove.com/t/57258/a-simple-method-for-isolation-soybean-protoplasts-application-to ↩︎