The Plant Press

From The Plant Press, Vol. 24, No. 1, January 2021.

Molecular techniques in focus: Common misconceptions about tissue preservation for DNA 

By Gabe Johnson

Countless Ph.D. research botanists sincerely believe such misconceptions as: ethanol immediately destroys DNA in leaf tissue collections, leaves in both fresh and spent silica gel are equally dry, desiccated tissues rehydrate in absolute ethanol, and that EDTA inhibits all deoxyribonuclease (DNase) activity. Ethanol treatment of plant tissue for molecular studies is especially misunderstood by researchers in recent times, even though it can be a useful tool to improve DNA preservation for molecular studies. In this Information Age with the ability to sequence and process trillions of nucleotides of data in a single study, the capacity to isolate high quality DNAs must advance to utilize these new technologies to their fullest potential. Greater DNA quality and quantity are required for many of these next-generation sequencing methods than were necessary for traditional PCR-based Sanger methods. Obtaining sequenceable DNAs from a more complete taxonomic sampling will require an equally diverse set of DNA extraction tools, including ethanol preservation and pretreatment.

Biological tissues typically undergo a “treatment” such as freezing or drying prior to DNA extraction. The degree to which this treatment preserves the nucleic acid content of the leaf depends on the needs of the researcher and the logistical limitations of the field work. Such treatments can be as simple as allowing harvested leaves to dry in ambient conditions (usually with poor results) or the use of one of an array of desiccants, solvents, buffers, or cryogens (see Funk et al., Biodiversity Data J. 5: e11625; 2017). While a diversity of tissue treatment methods has been used to preserve DNA (i.e., Hamilton et al., Anal. Biochem. 49: 48-57; 1972), by the mid 1990’s silica gel desiccation was the default method for plant systematics studies (Chase et al., Taxon 40: 215-220; 1991). This preservation method has many virtues by being simple, inexpensive, non-toxic, and satisfactory for most major lineages of plants from algae to angiosperms, as well as fungi. However, the resulting DNA quality has limitations, and in particular is more fragmented than if extracted directly using fresh or frozen tissues. 

Just as DNA in herbarium specimens is better preserved for some lineages more than others, better DNA is extracted from silica dried tissues for certain taxa more than others. Tissue preservation media, dry or liquid, have a dual purpose: to protect the DNA within the tissue and to prepare the tissue for homogenization and efficient DNA isolation. Upon separation from the plant, a leaf undergoes a physiological stress response that initiates many cellular processes related to senescence: wound formation, polyphenol oxidation, and apoptosis. These cellular changes can rapidly degrade the DNA through programmed exonuclease activity and oxidative stress from reactive chemical species used in wound defense. 

It is a common misconception that dehydration can only occur in the absence of liquids. Immersing tissue in ≥96% ethanol also causes dehydration when water in the tissues rapidly diffuses into solution to reach equilibrium. This diffusion actually occurs faster than in air desiccated by solid media such as silica gel or calcium sulfate. While rarely used for plants, ethanol desiccation is the preferred method of preserving DNA in insects; they are then stored in ≥96% ethanol at sub-freezing temperatures. Ethanol does not inherently damage DNA, and is routinely used for nucleic acid precipitation. It cannot be stressed more emphatically that preserving tissues with ethanol for future DNA extractions is entirely different from using ethanol to prevent voucher specimens from rotting in the field. Herbarium voucher specimens collected in remote tropical field sites are routinely sprayed or soaked with 50-80% ethanol. This prevents microbial degradation of the specimen's morphological characters but poorly preserves its DNA, especially when coupled with subsequent high heat drying. 

To preserve leaf tissue with ethanol for DNA extraction, it must be torn into small fragments to increase surface area (and break the epidermis) for diffusion and immersed quickly into a vial of ≥95% ethanol at a general ratio of 1 mL per 1.0 cm² of leaf tissue (see image above). If the tissues contain considerable water, changing the alcohol will keep concentrations high and can enable better preservation, as is also employed with insects. Preserved samples should be kept in the dark; even the emissions of fluorescent office lights can degrade the DNA in leaf samples in as little as a week.

Ethanol desiccation confers irreversible changes to the elasticity of the plant cell wall (i.e., makes it brittle) and facilitates the removal of secondary chemicals from the tissue. In contrast, after silica gel drying, the leaf tissue contains all native substances and its cell walls regain elasticity upon rehydration. Just because a leaf was dried once in silica gel does not mean it stays at an optimal low humidity forever in a plastic bag. After the initial desiccation in silica gel, water vapor slowly seeps through the polyethylene container and into the desiccant and tissue. Eventually, the silica gel will become fully hydrated, and if not replaced, the leaf tissue will equilibrate with the humidity of the room. 

Buildings are generally kept at 40-60% relative humidity while a silica desiccator is about 30%. Although leaf tissue at ambient humidity contains significantly less water than a living leaf, the water available to the cell wall polysaccharides allow them to reform after mechanical deformation. This wall elasticity impedes homogenization in a bead beater and leads to reduced DNA yields. In contrast, leaves fully desiccated in ethanol do not regain the same elasticity after equilibrating to ambient humidity. 

Just as ethanol desiccation renders plant and fungal cell walls permanently brittle, it also irreversibly inactivates the plant cell's endogenous deoxyribonucleases, DNases (Linke et al., BioTechniques 49: 655-657; 2010). Contrary to popular belief, many families of nuclease enzymes are cofactor-independent and digest DNA in the absence of divalent cations such as Mg⁺² and Ca⁺². Since the concentrations and activity of cofactor independent nucleases vary among lineages of plants, these are not problematic for all plant systematists, which leads to common misunderstandings about them. Such DNases can quickly fragment in the DNA in the lysis buffer, even if it contains chelators like EDTA. While nucleases in silica desiccated leaves immediately regain their function upon rehydration, as little as briefly grinding tissue in ethanol can effectively inactivate all DNase activity (Adams et al., Mol. Ecol. 8: 681–684; 1999). 

Regardless of how leaf tissues were initially desiccated, subsequently treating samples in ethanol is a valuable method to improve DNA extractions by inactivating DNases, modifying polysaccharides to increase cell wall fragility, and removing many secondary metabolites. DNAs from silica dried material treated with ethanol before extraction are the same quality and quantity as those preserved in ethanol (Akindele et al., Conserv. Genet. Resour. 3: 409–411; 2011). Therefore, while ethanol desiccation is not necessarily a replacement for silica gel, it is a valuable accessory to be used with the various reagents and tools needed to best preserve the DNA of a particular organism in a certain location and quantity. 

Aside from its obvious limitations as a flammable liquid, ethanol is an invaluable resource for preserving plant nucleic acids in situ. It is unfortunate that misconceptions about its effect on plant collections have prevented its use more broadly in phylogenetic research. Using evidence-based research to demystify many long-held assumptions about collecting tissues for DNA extraction, molecular systematists can design better-informed collection strategies to obtain high quality DNAs from taxa long thought to be impossible to sequence.