Introduction
A new world of research opportunities has emerged with the advancement
of sequencing techniques. One of the fields that have benefited most is
the study of whole microbial communities, so-called microbiomes. This
novel method allows the study of bacterial mats and gut microbiome
linings without the need to cultivate each bacterium separately
(Caporaso et
al. 2012). Together with recently developed and improved bioinformatic
pipelines (Mothur, QIIME 2, etc.), we now have the means to classify and
assign taxonomy with a reasonable level of confidence
(Bolyenet al.2019).
As microbiome research moves away from model organisms and extends into
natural settings, new challenges of wildlife research and those arising
because of the variety of wildlife diets need to be tackled. One of the
drawbacks is that the separation of bacterial sequences from
non-bacterial signals, i.e. those from mitochondria (from the host) and
chloroplasts (from the diet) can sometimes be tricky
(Barottet al. 2011; Lundberg et al.2013). According to
the widely accepted endosymbiosis theory (Margulis [then known as
Sagan (1967), Gray
2017], mitochondria and chloroplasts were originally derived during
evolution from hijacked bacteria engulfed by other bacteria. Because of
this bacterial origin, some DNA sequences of organelles are strikingly
bacteria-like. This is also the case with reads obtained from high
throughput sequencing of 16S rRNA genes, the usual target of microbiome
studies. In the worst case, the resulting read coverage consists of many
reads assigned to mitochondria or chloroplasts.
Several ways are available to circumvent this problem; the most common
path is to increase the sequencing depth and then filter out the reads
assigned to the organelles. However, this technique results in an
expensive price tag for sequencing and may lead to highly skewed read
numbers depending on the provenance of the samples. Another option has
recently arisen: the use of DNA-PNA clamps as PCR blockers to prevent
the amplification of the specific mitochondrial or chloroplast sequences
(Fitzpatricket al. 2018). PNAs (peptide nucleic acids) are DNA-mimicking
molecules with outstanding hybridization properties
(Nielsen & Egholm
1999). The backbone of the molecules is constructed of N-(2-amino-ethyl)
glycyl (AEG) instead of the sugar-phosphate backbone of DNA
(Nielsen, Egholm &
Buchardt 1994). The nucleobases attached to this backbone are the same
as those in DNA, thereby allowing hybridization between the probe and
the bacterial DNA. PNAs are thus a powerful molecular tool in microbiome
research for dealing with samples with a high content of either host or
plant remnants in faecal pellets
(Lundberg et
al. 2013).
In this study, we tested the PNA-DNA clamps as a method for improving
microbiome discovery rates in bats (Tequila bat Leptonycteris
yerbabuenae ) and Galapagos mockingbirds (Mimus parvulus ). We
chose these two study organisms because they both rely on heavy
plant-based diets that sometimes can lead to masses of plant-based
faecal material producing high contamination signals from chloroplast
and mitochondria. Our study presents the first results obtained by using
PNA-DNA clamps to block the PCR amplification of chloroplast and
mitochondrial DNA from the diet in the gut microbiome of wildlife. The
method involves a cost-effective molecular technique, instead of the
filtering out of the unwanted sequencing reads.