Slippage of degenerate primers can cause variation in amplicon length
Metabarcoding studies often employ degenerate primers to reduce amplification bias and increase the number of detected taxa. However, degeneracy has the disadvantage of lowering binding specificity although the exact mechanisms and potential biases introduced by such off-target amplification are not fully understood. We examined sequences recovered from the ten most abundant operational taxonomic units (OTUs) in two mock communities to investigate the specificity and binding behavior of five degenerate primer sets. Our results indicate that primers frequently bound 1-2 bp upstream in taxa where a homopolymer region was present in the amplification direction. As well, although less frequent, degeneracy occasionally led to primer binding 1 bp downstream. Some widely used primer sets were severely affected by this slippage effect, while others were not. Our study shows that primer slippage can produce taxon-specific length variation in amplicons and subsequent length variation in recovered sequences. While this variation will only have small impacts on OTU designation by clustering algorithms that ignore terminal gaps, primer sets employed in metabarcoding projects should be evaluated for their sensitivity to slippage. Moreover, steps should be taken to reduce slippage by improving protocols for primer design. For example, the flanking region adjacent to the 3' end of the primer is not considered by current primer development software although GC clamps in this position could mitigate slippage. While degeneracy is important to ensure the universality of a primer, binding in homopolymer regions should be avoided.