How do we create analytical controls for state?
The importance of appropriate analytical controls in eDNA research is
well established (e.g.,13,61).
These include field and laboratory controls that are designed to assess
contamination (negative
controls)62,
analytical precision (biological and technical replicates), and
sensitivity (positive controls). However, these controls do not account
for eDNA being present in different states nor do these controls allow
assessment of whether eDNA in each state(s) is accurately quantified.
Moreover, incomplete recovery of analytical controls typically leads to
the conclusion that PCR inhibition is involved. While this clearly is a
possibility, we propose that results could also be confounded because
current protocols may not completely extract DNA from all four states if
present in the sample. Therefore, additional analytical controls are
needed to disambiguate the cause of observed signal attenuation (e.g.,
PCR inhibition versus inefficient extraction across states).
There are various analytical controls employed in the eDNA literature,
but these are inconsistently applied. Some researchers (e.g.,63)
advocate multiplexing an assay for a given target species together with
an assay designed to detect a co-occurring species presumed to be
ubiquitous in the environment, such as algae (e.g., using a generalized
plant chloroplast DNA assay), to demonstrate that the PCR reaction was
not inhibited. Yet, because the state (Figure 1) and concentration of
any species’ eDNA is unknown, it cannot be used to assess relative rates
of PCR inhibition and/or inefficient eDNA recovery. To address this
issue, internal standards of known DNA concentration and state could be
applied at various stages in the workflow (Figure 4). Synthetic DNA has
been used as an internal positive control to quantitate the relative
degree of PCR inhibition, but this does not account for inefficient
extraction of different eDNA states. Applying a “spike in” control
prior to the extraction/precipitation step could result in some sorption
of the control DNA, but again the attenuation of the PCR signal could
not be used to discriminate between inhibition and inefficient recovery.
Developing analytical controls to assess whether eDNA is bound to
cellular debris, adsorbed to particles or dissolved in solution remains
a challenge. Size fractionation can be achieved by filtering a sample
through multiple filters of progressively smaller pore size and
subsequently extracting eDNA from each individual filter and the
filtrate. Assuming any DNA that passed through the filters into the
filtrate represents dissolved eDNA and potentially even particles, it is
possible to quantify this pool. However, eDNA recovered from the filters
cannot be separated into cellular bound vs. particle bound DNA without
utilizing extraction protocols optimized to recover only particle bound
or cellular debris bound eDNA. Protocols for separating soluble DNA
(i.e., extracellular and bound to particles) from insoluble DNA (i.e.,
still inside the cell) have been
developed60.
Their parallel application to known mixtures of cellular bound, particle
bound and dissolved eDNA could prove illuminating by separating out the
different states and analysing them separately for detection of a target
species or community. However, quantifying the eDNA in each category
before assembling the mixtures would be non-trivial and even then, the
approach could not easily assess the dynamic conversion of DNA between
states that may occur during the extraction. These issues
notwithstanding, the combined use of cellular material, plasmids (e.g.
as surrogates for organelles), synthetic DNA and varied adsorbent
materials, together with size fractionation and multiple extraction
techniques as applied across a gradient of environmental conditions
could yield novel insights concerning extraction efficiency among eDNA
states and the dynamic conversion processes between them when
selectively applied to each sample processing step (Figure 4).