responsive transcription factors. There were
also ABA transcription factor binding sites in 30
of the upregulated genes, while 119 had a bind
site for ethylene-responsive transcription factors.
This pathway matches the ABA or environment
triggered, calcium-dependent signal pathway
observed in maturing seeds, reinforcing the
similarity of turions and seeds on a molecular,
invisible level.
The same authors performed a second investigation of the rRNA depleted RNA-seq experiment in fronds and developing turions (Wang
et al. 2015). Since 26% of the total RNA
sequenced mapped to the chloroplast genome,
they had 1000-fold coverage of most genes after
stringent filtering. The PPR proteins are a massive
family, characterized by the 35 amino acid pentatricopeptide repeat motif that specifically binds
the 4th and 34th residues in the pfam model to an
RNA base, creating a pattern of these motifs that
bind to a specific RNA sequence (Barkan et al.
2012; Manna 2015). While these PPR proteins
are found in prokaryotes and eukaryotes acting in
splicing, processing, editing, stability, and translation of RNAs, this study focused on the
DYW-type PPRs that correct certain missense
mutations in the plastid genome by editing the
mRNA from a cytosine to uracil residue, thereby
creating a functional mRNA and protein product.
Mapping the RNA-seq reads and detecting C to U
SNPs with SAMtools revealed 66 sites of RNA
editing with an average efficiency of 76% and a
range of 6–100%. Comparison to developing
turions showed very similar gene expression with
no differentially expressed genes. There were,
however, six over and five under edited sites
(>two-fold difference, p value <0.05) in seven
genes during turion development compared to
fronds. So while expression was constant, 1/6th
of the sites were differentially edited, thereby
altering the functional protein abundance of seven
genes. These differences in editing efficiency
even varied as much as 8–100% at multiple sites
within the same gene due to the sequence-specific
nature of PPR protein editing. A phylogenetic
analysis with the Mega6 program revealed the 66
editing sites in Spirodela had an 81% overlap
with the 75 in coconut. There was a 42 and 38%
overlap with the 35 and 26 sites observed in the
more evolutionarily distant rice and maize. This
correlation confirms the hypothesis of a single
origin of RNA editing PPR proteins in the early
land plants like ferns, that have hundreds of
edited sites that were gradually reduced and differentiated over time to *80 in the basal monocots and 25–40 in the more recent angiosperm
species.
Another indirect study of Spirodela mRNA
expression was the degradome experiment found
in Fourounjian et al. (2019). The primary purpose of this experiment was to confirm the
cleavage activity of miRNAs on target mRNAs
and observe regulatory differences between the
biological triplicates of the control, 0 °C, 37 °C,
ABA, kinetin, copper, nitrate, and sucrose conditions by sequencing 28–63 million uncapped
mRNAs per library. While this degradome
sequencing is not a perfect correlate to mRNA
expression, the normalized read count of each
gene (not kilobase normalized), its expression
pattern can be viewed in this program hosted by
the Myers laboratory of the Danforth Center
https://mpss.danforthcenter.org/*private/dbs/
index.php?SITE=messing_SPIRODELA_PARE.
These patterns can even reveal unannotated
exons, since all reads were polyadenylated.
Finally, the miRNA cleavage study revealed that
15 genes, mainly well-conserved transcription
factors were expressed and cleaved in four or
more conditions, while 71% of the results were
condition-specific targets, many of which had
more structural and metabolic functions. Of these
conditional specific changes, sucrose created the
largest difference, followed by copper and heat
exposure. This large transcriptomic and metabolic change of sucrose addition suggests that
laboratory experiments modeling duckweeds in
outdoor applications should avoid this often
added media component.
The assembly of the Spirodela genome for
strains 7498 and 9509 provided not only a scaffold for easy and accurate mapping of RNA-seq
data, but a context for the gene expression. This
is both in a physical sense for the chromatin
13 Transcriptome Responses of Spirodela polyrhiza
135
also ABA transcription factor binding sites in 30
of the upregulated genes, while 119 had a bind
site for ethylene-responsive transcription factors.
This pathway matches the ABA or environment
triggered, calcium-dependent signal pathway
observed in maturing seeds, reinforcing the
similarity of turions and seeds on a molecular,
invisible level.
The same authors performed a second investigation of the rRNA depleted RNA-seq experiment in fronds and developing turions (Wang
et al. 2015). Since 26% of the total RNA
sequenced mapped to the chloroplast genome,
they had 1000-fold coverage of most genes after
stringent filtering. The PPR proteins are a massive
family, characterized by the 35 amino acid pentatricopeptide repeat motif that specifically binds
the 4th and 34th residues in the pfam model to an
RNA base, creating a pattern of these motifs that
bind to a specific RNA sequence (Barkan et al.
2012; Manna 2015). While these PPR proteins
are found in prokaryotes and eukaryotes acting in
splicing, processing, editing, stability, and translation of RNAs, this study focused on the
DYW-type PPRs that correct certain missense
mutations in the plastid genome by editing the
mRNA from a cytosine to uracil residue, thereby
creating a functional mRNA and protein product.
Mapping the RNA-seq reads and detecting C to U
SNPs with SAMtools revealed 66 sites of RNA
editing with an average efficiency of 76% and a
range of 6–100%. Comparison to developing
turions showed very similar gene expression with
no differentially expressed genes. There were,
however, six over and five under edited sites
(>two-fold difference, p value <0.05) in seven
genes during turion development compared to
fronds. So while expression was constant, 1/6th
of the sites were differentially edited, thereby
altering the functional protein abundance of seven
genes. These differences in editing efficiency
even varied as much as 8–100% at multiple sites
within the same gene due to the sequence-specific
nature of PPR protein editing. A phylogenetic
analysis with the Mega6 program revealed the 66
editing sites in Spirodela had an 81% overlap
with the 75 in coconut. There was a 42 and 38%
overlap with the 35 and 26 sites observed in the
more evolutionarily distant rice and maize. This
correlation confirms the hypothesis of a single
origin of RNA editing PPR proteins in the early
land plants like ferns, that have hundreds of
edited sites that were gradually reduced and differentiated over time to *80 in the basal monocots and 25–40 in the more recent angiosperm
species.
Another indirect study of Spirodela mRNA
expression was the degradome experiment found
in Fourounjian et al. (2019). The primary purpose of this experiment was to confirm the
cleavage activity of miRNAs on target mRNAs
and observe regulatory differences between the
biological triplicates of the control, 0 °C, 37 °C,
ABA, kinetin, copper, nitrate, and sucrose conditions by sequencing 28–63 million uncapped
mRNAs per library. While this degradome
sequencing is not a perfect correlate to mRNA
expression, the normalized read count of each
gene (not kilobase normalized), its expression
pattern can be viewed in this program hosted by
the Myers laboratory of the Danforth Center
https://mpss.danforthcenter.org/*private/dbs/
index.php?SITE=messing_SPIRODELA_PARE.
These patterns can even reveal unannotated
exons, since all reads were polyadenylated.
Finally, the miRNA cleavage study revealed that
15 genes, mainly well-conserved transcription
factors were expressed and cleaved in four or
more conditions, while 71% of the results were
condition-specific targets, many of which had
more structural and metabolic functions. Of these
conditional specific changes, sucrose created the
largest difference, followed by copper and heat
exposure. This large transcriptomic and metabolic change of sucrose addition suggests that
laboratory experiments modeling duckweeds in
outdoor applications should avoid this often
added media component.
The assembly of the Spirodela genome for
strains 7498 and 9509 provided not only a scaffold for easy and accurate mapping of RNA-seq
data, but a context for the gene expression. This
is both in a physical sense for the chromatin
13 Transcriptome Responses of Spirodela polyrhiza
135