The Drosophila kikkawai feature with NCBI Gene ID 108084518 was determined to be an ortholog of Drosophila melanogaster Sox102F, a member of the FlyBase High Mobility Group Box Transcription Factors gene group (FBgg0000748). Five isoforms were constructed using the GEP F element annotation protocol, the longest being novel isoform Sox102F-PNE (identified using the XM_017180752 RefSeq prediction and RNA-seq data). Among the isoforms found in both D. melanogaster and D. kikkawai, Sox102F-PB is the longest and exhibits a 1.18x coding span expansion due to transposable element insertion into an intron. All D. kikkawai protein isoforms contain the conserved domain HMG_box_dom (IPR009071).

Drosophila melanogaster Sox102F has been assigned to the High Mobility Group Box Transcription Factors gene family (Pfreundt et al., 2010; Phochanukul & Russell, 2010; Sessa & Bianchi, 2007). Proteins from this group regulate the Wnt signaling pathway and contain a characteristic 80 AA L-shaped DNA minor groove binding domain, which when bound to DNA induces DNA bending. According to FlyBase (release FB2024_02), the Sox102F gene is most likely orthologous to either the human SOX5 or SOX6 gene, having a DIOPT score of 9/14 when run against both SOX genes (Gramates et al., 2022; Hu et al., 2011). In humans, mutations in the SOX5 gene are related to Lamb Shaffer Syndrome, a neurodevelopmental disorder (Lamb et al., 2012). Due to its close association to the brain and development, Sox102F in Drosophila has been used to study Alzheimer's and heart disease in humans (Li et al., 2013, 2017). Drosophila kikkawai belongs to the melanogaster group of the Sophophora subgenus (NCBI taxonomy ID: 30033) (Schoch et al., 2020). This cosmopolitan species is tropical and subtropical, as it is not found above the latitude of 35° (Karan et al., 1998). D. kikkawai is one of four Drosophila species (along with Drosophila takahashii, Drosophila ananassae, Drosophila bipectinata) examined in the study of the Muller F element expansion and shows an approximate 1.7-fold increase in chromosome size when compared to the D. melanogaster F element (Leung et al., 2023).

D. kikkawai feature with NCBI Gene ID 108084518 is the putative ortholog of Sox102F. The ortholog assignment is supported by a tBLASTn (v2.15.0+; Camacho et al., 2009) alignment using the NCBI BLAST server of the D. melanogaster protein sequence for Sox102F-PA (FBpp0088312) against the entire D. kikkawai DkikHiC1 (GenBank Assembly Accession: GCA_030179895.1) assembly. The top hit maps to scaffold CM058227.1 (assigned to the F element) and reports an E-value of 7e-118, a percent identity of 74.02, and a percent coverage of 99. The coordinates for the top hit (i.e., the match with lowest E-value) correspond to the location of the D. kikkawai feature with Gene ID 108084518. The next best hit maps to scaffold CM058225.1 (assigned to the D element) and reports a higher E-value of 8e-25, a lower percent identity of 50.43, and a lower percent coverage of 70. Sox102F-PA is representative of the B, C, D and novel NE isoforms due to the significant CDS overlap among the isoforms. The results of three alignment tools within the genome browser (Spaln, BLAT, tBLASTn) map to the same region which corresponds to the location of the current gene model, providing strong evidence for the ortholog assignment, along with the E-value. Local synteny analysis provides further evidence for ortholog assignment. Sox102F is located on chromosome 4 (the F element) in D. melanogaster and surrounded by the genes bent (bt) (FBgn0005666), Mediator complex subunit 26 (MED26) (FBgn0039923), forkhead domain 102C (fd102C) (FBgn0039937), and Gigyf (Gyf) (FBgn0039936). In D. kikkawai, the orthologs of Eye-enriched kainate receptor (Ekar) (FBgn0039916) (Gene ID: 108079305) and Mediator complex subunit 26 (MED26) (Gene ID: 108079308) are located downstream of the Sox102F ortholog while the orthologs of forkhead domain 102C (fd102C) (Gene ID: 108084517), and CG31998 (FBgn0051998) (Gene ID: 108083269) are located upstream on the F element. As shown in Figure 1A, the two genes immediately flanking Sox102F in D. kikkawai are consistent with D. melanogaster while the next two genes in the genomic neighborhoods differ between the two species. The D. kikkawai feature with the Gene ID 108079305 was determined to be an ortholog of Ekar rather than an ortholog of bt based on the FlyBase BLASTp (v2.2.18; Altschul et al., 1990) search result of the protein product (XP_041632629) derived from the D. kikkawai RefSeq mRNA XM_041776695 against the D. melanogaster “Annotated proteins” database. The best BLASTp match is to D. melanogaster Ekar-PB with a normalized score of 1555.81 bits and an E-value of 0 (i.e., E-value < 1e-180). The next best hit to CG11155-PD also has an E-value of 0 but a lower score of 969.53 bits. Similarly, the D. kikkawai feature with the Gene ID 108083269 was determined to be an ortholog of CG31998 rather than Gyf based on the FlyBase BLASTp search result of the protein product (XP_017034489) derived from the D. kikkawai RefSeq mRNA XM_017179000 against the D. melanogaster “Annotated proteins” database. The best and only matches are to the A and B isoforms of the CG31998 gene where the top hit to D. melanogaster CG31998-PA reports a normalized score of 1338.94 bits and E-value of 0.

Characterizing the A, C and D isoforms for Sox102F. The Sox102F gene is located on the F element of D. kikkawai. Isoforms Sox102F-PA, Sox102F-PC, and Sox102F-PD in D. kikkawai are conserved relative to the orthologous isoforms in D. melanogaster and were annotated according to the protocol described in Rele et al., 2023. In both D. kikkawai and D. melanogaster, Sox102F-PA (BK067818), Sox102F-PC (BK067819), and Sox102F-PD (BK067820) are comprised of the same two sequences from the unspliced transcript while Sox102F-PB (BK067821), described in further detail below, is comprised of three coding sequences, two shared with the other isoforms and one unique initial CDS (Figure 1B). Further analysis of the Sox102F feature in D. kikkawai led to the discovery of a novel isoform named Sox102F-PNE (BK067822). Nucleotide sequence data reported are available in the Third-Party Annotation Section of the DDBJ/ENA/GenBank databases under the accession numbers TPA: BK067818-BK067822.

Characterizing Sox102F-PB and novel isoform Sox102F-PNE. The third CDS of the novel isoform overlaps with the open reading frame of the initial CDS of Sox102F-PB (inset of Figure 1B). The initial CDS of the Sox102F-PB isoform lacks splice junction support from the combined splice junction track in the GEP UCSC Genome Browser, and the best BLASTx (v2.15.0+) hit does not include the first 6 AA. There are no other nearby in-frame start codons. There were two options to retain the Sox102F-PB isoform, either to modify the gene structure by proposing a novel initial CDS or to truncate the CDS to the nearest start codon at 704,877-704,875. Based on the annotation strategy to construct the most parsimonious gene model compared to the D. melanogaster ortholog, the initial CDS for Sox102F-PB was truncated. Due to evidence of splice junctions and RefSeq predictions upstream of this start position, it was concluded that a novel isoform, Sox102F-PNE, whose CDS overlaps that of Sox102F-PB, exists in D. kikkawai (Figure 1B). Combined splice junctions JUNC00109258, JUNC00109265, and JUNC00109267 mapped to the DkikHiC1 assembly and small RNA-seq peaks from adult males and mixed embryo correspond to the splice boundaries predicted by BLAT (RefSeq mRNA XM_017180752), with the latter two junctions scoring reads greater than 10. A combined splice junction score of 10 indicates that the predicted intron is supported by 10 RNA-seq reads, which is the minimum support required for a novel isoform as per protocol. Sox102F-PNE becomes the longest Sox102F isoform in D. kikkawai. Sox102F is involved in the phenomenon known as the F element expansion. The expansion of the Sox102F gene was calculated using Sox102F-PB, the longest isoform whose ortholog can be found in D. melanogaster. The coding span (from start to stop codon and including introns) of the Sox102F-PB gene in D. kikkawai is 26,432 base pairs while its ortholog in D. melanogaster has a coding span of 22,317 base pairs. The ~4,000 base pair, or 1.18x, expansion is attributed to the insertion of LINE transposons (TEs) into the intron between Sox102F-PB CDS2 (3_9492_0) and CDS3 (4_9492_0) which are shared across all isoforms. The insertions of these TEs did not alter the gene structure or the predicted amino acid sequence. D. melanogaster has no identifiable TEs annotated in the corresponding intron.

Characterizing HMG_box_domain in Sox102F. As seen in the EMBOSS Needle (v6.6.0.0; Rice et al., 2000) alignment (Figure 1C), the HMG_box_domain (IPR009071; Paysan-Lafosse et al., 2023) has been identified in Sox102F-PA and is found to be shared in all isoforms, including the novel NE isoform. This confirms that the feature belongs to the High Mobility Group Box Transcription Factors gene family. Figure 1D depicts that the domain circled in red shows a much higher level of sequence conservation than the rest of the protein when compared to the orthologous D. melanogaster protein, alluding to its importance to protein function. Sequence outside of the red circle represent variable regions of lower sequence similarity that do not belong to the conserved domain and vary across species due to the accumulation of mutations over evolutionary time. Across 36 Drosophila species the HMG_box_dom is highly conserved in Sox102F which can be seen in a ROAST alignment of the terminal CDS (Figure 1E). Proteins belonging to the High Mobility Group Box Transcription Factors gene group at FlyBase (FBgg0000748) have been characterized as ubiquitous regulators of development by binding directly to the minor groove of DNA during transcription (Kamachi & Kondoh, 2013; Sessa & Bianchi, 2007). The Sox102F protein's role in development is consistent with the fact that the most abundant subset of supporting RNA-seq coverage is from mixed embryos.

The protocol used to annotate and reconcile the Sox102F gene model and neighboring gene models can be found in the Rele et al., 2023 paper. The annotations are based on the annotated gene models for FlyBase release FB2022_06 (D. melanogaster release 6.49) in the release 6 assembly (Hoskins et al., 2015). A mirror of the UCSC Genome Browser (v435) (Kent et al., 2002; Navarro Gonzalez et al., 2021) is maintained by the Genomics Education Partnership (GEP) at https://gander.wustl.edu. Within the D. kikkawai Hi-C genome browser, tracks displaying the results of experimental data (e.g., RNA-seq) and computational tools such as tBLASTn (v2.13.0+), Spaln (v2.3.3f), and BLAT (v37x1) were used support the assignment of the Sox102F ortholog. The D. kikkawai RNA-seq data was generated by the modENCODE project (Chen et al., 2014). The tBLASTn results report the region of the genome with the highest similarity to D. melanogaster protein coding sequences. The Spaln results report the region of the genome with the highest similarity to full-length D. melanogaster proteins. BLAT alignments report the region of the genome with the highest similarity to D. melanogaster transcripts.

We would like to thank Wilson Leung for developing and maintaining the technological infrastructure that was used to create this gene model. We would also like to thank Dr. Christopher Shaffer for supervising Mia Mo and Larissa LoBello in the reconciliation process and for his feedback on the manuscript.

This material is based upon work supported by the National Science Foundation (NSF) under Grant No. 2114661 to Dr. Cindy Arrigo. The Genomics Education Partnership (GEP) is supported by the NSF under Grant No. 1915544 and National Institute of General Medical Sciences of the National Institutes of Health under award number R25GM130517 to the Genomics Education Partnership (https://thegep.org/). The Genomics Education Partnership is fully financed by Federal moneys. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Supported by National Science Foundation (United States) 2114661 to Cindy Arrigo.

Supported by National Institutes of Health (United States) R25GM130517 to Laura Reed.

Supported by National Science Foundation (United States) 1915544 to Laura Reed.