Description
Introduction -
Within insects, the process of detoxifying xenobiotics and host secondary metabolites is a three-phase process that involves functionalization, conjugation, and excretion of these compounds. Expansions of known detoxification gene families (
e.g.
, cytochrome P450s) is associated with diet breadth and insecticide resistance (Ranson et al., 2002; Després et al., 2007; Rane et al., 2016). With the increasing availability of high-quality genomes for non-model organisms, including
Drosophila
species beyond
D. melanogaster
, it is now possible to perform large scale comparative studies (Robinson et al., 2011; Kim et al., 2021; Threfall and Baxter 2021). Careful manual annotation and curation of gene models can improve upon computational gene predictions in non-model species, which aids the accuracy of studies on gene and genome evolution (Mudge and Harrow 2016; Tello-Ruiz et al., 2019). To aid in these annotations, the Genomics Education Partnership (thegep.org) developed a curriculum involving web-based tools that allow undergraduates to engage in authentic course-based research focused on manually annotating genes in non-model species (Rele et al., 2023). The orthologous gene models, including the one presented here, then provide a reliable basis for further evolutionary genomic analyses when made available to the scientific community. The gene ortholog described here in
D. dunni
for
Glutathione S transferase O3
(
GstO3
), a member of the glutathione S-transferase gene family, was characterized as part of a developing dataset for a comparative study of detoxification gene families in the
immigrans
-
tripunctata
radiation of the genus
Drosophila
.
Within the subgenus
Drosophila
,
D. dunni
Townsend and Wheeler 1955 is placed in the
dunni
subgroup of the
cardini
species group in the
immigrans-tripunctata
radiation (Heed & Krishnamurthy 1959; Bächli 2005). Species in the
dunni
subgroup, including
D. dunni
, are distributed across the Caribbean (Heed & Krishnamurthy 1959). Members of the
cardini
group primarily feed and develop on fruit and flowers (Markow & O’Grady 2008). While some mushroom-feeding
cardini
subgroup members tolerate the mushroom toxin a-amanitin (Stump et al. 2011), the fruit/flower feeding
dunni
subgroup species do not (Erlenbach et al.
2023).
One class of phase II detoxification enzymes are the glutathione S-transferases (GSTs), which act by conjugating xenobiotics or products of phase I detoxification with a glutathione to make them more hydrophilic prior to excretion (Enayati et al., 2005). GSTs are classified into two families based on their cellular location (microsomal and cytosolic). Within insects, cytosolic or canonical GSTs are divided into six classes (delta, sigma, epsilon, zeta, theta, and omega; Enayati et al., 2005; Ketterman et al., 2011). In the omega class,
Glutathione S transferase O3
(
GstO3
) is a biochemically active GST (Kim et al. 2006). It is significantly upregulated following exposure to oxidative stressors, such as hydrogen peroxide and zinc (Yepiskoposyan et al., 2006; Li et al., 2008). Across 12
Drosophila
genomes,
GstO3
is highly conserved and expressed in all life stages of
D. melanogaster
(Walters et al., 2009).
We propose a gene model for the
D. dunni
ortholog of the
D. melanogaster
Glutathione S transferase O3
(
GstO3
) gene. The genomic region of the ortholog corresponds to the Augustus gene prediction JAECXC010000286.g1545.t1 in the ASM1815212v1 Genome Assembly of
D. dunni
(GCA_018152125.1 – Kim et al., 2021). This model is based on mixed sex adult RNA-Seq data from
D. dunni
(Erlenbach et al. 2023; https://doi.org/10.5061/dryad.hdr7sqvq2)
and
GstO3
in
D. melanogaster
using FlyBase release FB2024_02 (GCA_000001215.4; Gramates et al., 2022; Jenkins et al., 2022; Larkin et al.,
2021).
Synteny -
The reference gene,
GstO3,
occurs on
chromosome 3L in
D. melanogaster
and is flanked upstream by
CG6765
and
CG6683
and downstream by
sepia
(
se
) and
Glutathione S transferase O2
(
GstO2
). The
tblastn
search of
D. melanogaster
GstO3-PA (query) against the
D. dunni
(GenBank Accession: GCA_018152125.1 Genome Assembly (subject) placed the putative ortholog of
GstO3
in contig_291 (JAECXC010000286) which corresponds to Augustus gene prediction JAECXC010000286.g1545.t1 (E-value: 2e-139; percent identity: 74.27% as determined by
blastp
). The putative ortholog is flanked upstream by Augustus gene prediction JAECXC010000286.g1546.t1 and JAECXC010000286.g1547.t1, which correspond to
CG6765
and
CG6683
in
D. melanogaster
(E-value: 0.0 and 2e-52; identity: 63.37% and 49.54%, respectively, as determined by
blastp
;
Figure 1A
; Altschul et al., 1990). The putative ortholog of
GstO3
is flanked downstream by Augustus gene prediction JAECXC010000286.g1544.t1 and JAECXC010000286.g1543.t1, which correspond to
se
and
GstO2
in
D. melanogaster
(E-value: 5e-167 and 2e-127; identity: 90.12% and 68.80%, respectively, as determined by
blastp
). The putative ortholog assignment for
GstO3
in
D. dunni
is supported by the following evidence: the gene predictions surrounding the
GstO3
ortholog are orthologous to the genes at the same locus in
D. melanogaster
, gene expression data corresponds with each prediction, and local synteny is completely conserved, supported by E-values and percent identities, so we conclude that Augustus gene prediction JAECXC010000286.g1545.t1 is an ortholog of
GstO3
in
D. dunni
(
Figure 1A
).
Protein Model -
GstO3
in
D. dunni
has 2 coding sequences (CDS) within the genome sequence. The only unique protein sequence (GstO3-PA) is translated from 1 mRNA isoform (GstO3-RA;
Figure 1B
). Relative to the ortholog in
D. melanogaster
, the CDS number and protein isoform count are conserved (2 CDSs and 1 mRNA isoform
).
The sequence ofGstO3-PA
in
D. dunni
has 74.3% identity (88.0% similarity) with the
protein-coding isoform
GstO3-PA
in
D. melanogaster
,
as determined by
blastp
(
Figure 1C
). This level of divergence is not surprising given that
D. dunni
and
D. melanogaster
belong to two separate subgenera (
Drosophila
and
Sophophora
respectively) that diverged approximately 45-60MYA (Russo et al., 1995; Tamura et al. 2004; Obbard et al., 2012). Coordinates of this curated gene model are archived in the CaltechDATA repository (see “Extended Data” section below).
Methods
The annotation methods used in this project are adapted from those described in Rele et al. (2023), which includes algorithms, database versions, and citations for the complete annotation process developed for the Pathways Project. The methods for the current project are detailed in brief below with notes on significant differences between this protocol and the one described in Rele et al. (2023). The students use the GEP instance of the UCSC Genome Browser v.435 (https://gander.wustl.edu
;
Kent WJ et al., 2002; Raney et al., 2024) to examine the genomic neighborhood of their reference detoxification gene in the
D. melanogaster
genome assembly (Aug. 2014; BDGP Release 6 + ISO1 MT/dm6). Students obtain the protein sequence for the
D. melanogaster
target gene for a given isoform and use a
tblastn
search of the sequence against their target
Drosophila
species genome assembly (
D. dunni
(GCA_018152125.1 – Kim et al., 2021)) on the NCBI BLAST server (https://blast.ncbi.nlm.nih.gov/Blast.cgi, Altschul et al., 1990) to identify the putative ortholog location. Students compare the genomic neighborhood of the putative ortholog to that of the reference gene in
D. melanogaster
. This local synteny analysis includes a minimum of two upstream and downstream genes relative to the potential ortholog. As no RefSeq protein data is available for these species, comparisons are based on gene predictions that correlate with gene expression data in the putative ortholog neighborhood. Using the multiple alignment tracks feature in the Genome Browser, students examine other sets of genomic evidence, including Spaln alignment of
D. melanogaster
proteins, multiple gene prediction tracks (e.g., GeMoMa, Augustus, NSCAN PASA-EST), and RNA-Seq mixed sex adult expression data from the target species generated by Erlenbach et al. (2023; https://doi.org/10.5061/dryad.hdr7sqvq2). Information on the genomic structure information (e.g., CDSs, intron-exon number, number of isoforms) for the reference gene in
D. melanogaster
is retrieved using Gene Record Finder (https://gander.wustl.edu/~wilson/dmelgenerecord/index.html; Rele et al
.,
2023). To determine approximate splice sites within the target gene, a
tblastn
search using the CDSs from the
D. melanogaste
r reference gene against the putative ortholog location (10kb up- and downstream of the target gene prediction). Coordinates of the CDS(s) are refined by examining aligned RNA-Seq data, identifying canonical splice site sequences, and ensuring the maintenance of an open reading frame. Students confirm the biological validity of their target gene model using the FlySeq Gene Model Checker (https://gander2.wustl.edu/~wilson/genechecker-flyseq/), which compares the hypothesized target gene model’s structure and translated sequence against the
D. melanogaster
reference
gene. At least two independent models for this gene are generated. These models are reconciled by a third independent researcher to produce the final model presented here. Note: comparison of 5' and 3' UTR sequence information is not included in this GEP CURE protocol.