Genomic identification, annotation, and comparative analysis of Vacuolar-type ATP synthase subunits in Diaphorina citri

The hemipteran insect Diaphorina citri, or Asian citrus psyllid, is a vector for Candidatus Liberibacter asiaticus (CLas), the bacterium causing citrus greening disease, or Huanglongbing (HLB). Millions of citrus trees have been destroyed, and every grove in Florida, USA, has been directly affected by this disease. In eukaryotes, vacuolar-type ATP synthase (V-ATPase) is an abundant heterodimeric enzyme that serves the cell with essential compartment acidification through the active processes that transport protons across the membrane. Fifteen putative V-ATPase genes in the D. citri genome were manually curated. Comparative genomic analysis revealed that D. citri V-ATPase subunits share domains and motifs with other insects, including the V-ATPase-A superfamily domain. Phylogenetic analysis separates D. citri V-ATPase subunits into expected clades with orthologous sequences. Annotation of the D. citri genome is a critical step towards developing directed pest management strategies to reduce the spread of HLB throughout the citrus industry.


INTRODUCTION
Vacuolar (H + )-ATP synthase (V-ATPase) is a highly conserved eukaryotic enzyme [1]. Originally identified in the vacuole membrane, V-ATPase has a critical function in the plasma membrane and endomembrane system of almost every cell [2,3]. V-ATPase regulates the acidity of organelles, such as vacuoles, the Golgi apparatus, and coated vesicles, by translocating protons across their membranes and powering secondary transport processes. Structurally, V-ATPase has a noncatalytic transmembrane domain, the V 0 rotor, and a catalytic cytoplasmic domain, the V 1 stator. V-ATPase hydrolyzes adenosine triphosphate (ATP) into adenosine diphosphate (ADP), thus acting in the opposite manner to  [11]. https://www.protocols.io/widgets/doi?uri=dx.doi. org/10.17504/protocols.io.bniimcce  melanogaster (fruit fly) [20] were obtained by reciprocal BLASTp (RRID:SCR_004870) analysis of the nonredundant protein database at NCBI [8]. A neighbor-joining phylogenetic tree was constructed using MUSCLE (RRID:SCR_011812) multiple sequence alignment with the Poisson correction method and 1000 replicate bootstrap test using full-length protein sequences in MEGA version 7 (RRID:SCR_000667) for the transmembrane complex, the catalytic complex, and the accessory subunit Ac45, respectively (Figures 2-4) [21]. The sequence accession numbers used in these analyses can be found in Tables 2-4.
Comparative expression levels of D. citri V-ATPases throughout egg, nymph, and adult life stages in D. citri insects both exposed and not exposed to CLas were determined using RNA-seq data and the Citrus Greening Expression Network (CGEN) [9]. These gene expression levels were visualized using the pheatmap package in R (RRID:SCR_016418) [22,23]. Expression values for all samples discussed in this manuscript are visualized in  Table 5.  . Phylogenetic analysis of V-ATPase accessory subunit Ac45. The tree was constructed with MEGA7 software [21] using MUSCLE for alignment of amino acid sequences, followed by neighbor-joining analysis with 1000 bootstrap replications. Values greater than 50 are shown at nodes. Diaphorina citri is marked in bold. NCBI accession numbers are shown in Table 4.

DATA VALIDATION AND QUALITY CONTROL
Genes encoding all 13 subunits required to build a single V-ATP synthase enzyme, as well as an accessory subunit S1 (Ac45) gene, were annotated in D. citri. There were no additional subunits found in D. citri, as reported in other metazoans [2]. Although insect V-ATPases are known to contain 13 subunits, there is variation in the gene copy number for individual subunits among different species (Tables 6-8 duplication event occurred before the divergence of Hemimetabola and Holometabola ( Figure 2).
We identified complete genes in genome v3.0 for all the subunits except V-ATPase H.
Using genome-independent transcript sequences [24], we were able to determine that the 3′ portion of the V-ATPase H gene is located on chromosome 1, but the 5′ end of the gene is on  . Comparative expression levels of the Diaphorina citri V-ATPase genes encoding the V 1 , catalytic, and accessory subunits in D. citri insects reared on various infected and uninfected citrus varieties. V-ATPase H was annotated as two partial models and are both represented here separately as V-ATPase H (5′), denoting the 5-prime end of the gene, and V-ATPase H (3′), denoting the 3-prime end of the gene. Expression data were collected from the Citrus Greening Expression Network [9], with psyllid RNA-seq data from NCBI BioProjects PRJNA609978 and PRJNA448935, in addition to several published datasets [25][26][27][28][29]. Citrus hosts are abbreviated as Csin (Citrus  Accession number, bit score, query coverage (QC), and identity results from protein BLAST analysis of annotated D. citri V-ATPase transmembrane subunit genes to their putative orthologs.
one of the unplaced chromosomes that make up chromosome 0 (Table 1). Tables 2-4 show the results of protein BLAST analysis comparing the same insects as found in Tables 6-8.
BLAST results of annotated gene models had high query coverage to orthologs, supporting the completeness of the annotated gene models. In contrast, the sequence identities of Ac45, approximately 24-33%, show the highest divergence when comparing D. citri to other insects (Table 4). For the V 0 , transmembrane domain, subunits in Table 2, proteolipid subunit c (V-ATPase c) maintains some of the highest percentages of sequence identity, highlighting the importance of the protein function to form the c-ring that rotates and ultimately translocates protons across various membranes [2]. This is supported by the data shown in Table 6, in which a single gene copy for V-ATPase c is maintained across different orders of insects.
The Citrus Greening Expression Network (CGEN) was used to compare transcript expression levels in various regions of D. citri that have either been exposed to or not exposed to CLas infection [9,30,31].  Table 5).
These expression levels, coupled with the fundamental cellular nature and relatively even occurrence of V-ATPases, suggest that V-ATPase genes are good candidates for RNAi.  Accession number, bit score, query coverage (QC), and identity results from protein BLAST analysis of annotated Diaphorina citri V-ATPase catalytic subunit genes to their putative orthologs.

V-ATPase Ac45
Gene ID Dcitr07 g04330. 1 Comparative expression levels in transcripts per million (TPM) of the Diaphorina citri V-ATPase genes encoding the V-ATPase V 0 transmembrane, V 1 catalytic, and accessory subunits in D. citri insects reared on various infected and uninfected citrus varieties. V-ATPase H was annotated as two partial models and are both represented here separately as V-ATPase H (5′), denoting the 5-prime end of the gene, and V-ATPase H (3′), denoting the 3-prime end of the gene. Expression data were collected from the Citrus Greening Expression Network [9], with psyllid RNA-Seq data obtained from NCBI BioProjects PRJNA609978 and PRJNA448935, in addition to several published datasets [25][26][27][28][29]. Citrus hosts are abbreviated as Csin (Citrus sinensis), Cmed (Citrus medica), Cret (Citrus reticulata), and Cmac (Citrus macrophylla). Table 6. Gene copy comparison of V 0 transmembrane subunit V-ATPase genes in Diaphorina citri and orthologous insect genes.  Table 7. Gene copy comparison of V 1 catalytic subunit V-ATPase genes in Diaphorina citri and orthologous insect genes.  insect. However, it remains unknown at this time whether the elevated expression of V-ATPase c relative to other subunits in infected psyllids is associated with higher demand of these proteins in the cell. This should, therefore, be studied further in future research.

R. Grace et al.
Of the V 1 , catalytic domain, subunit genes, V-ATPase A and V-ATPase B maintain the highest percentages of sequence identity, consistent with the importance of their function in containing ATP binding sites at the V-ATPase subunits A/B protein interface (Table 3) [32].
Apart from D. melanogaster, V-ATPase A and B also maintain single copies of these two genes across different orders of insects, supporting their conserved nature compared with other genes of this enzyme (Table 7). V-ATPase A shows much higher expression than V-ATPase B across each measured variable, and V-ATPase G shows the highest expression in this group overall (Figure 6a). Unlike V-ATPase c, no significant differential expression was observed between the guts of insects reared on infected versus uninfected citrus trees  Table 5). A similar expression pattern can be seen throughout many of the V-ATPase catalytic genes, which may infer an interaction between these genes and pathogen infection. This warrants further investigations (Table 5).  (Tables 2, 6). In addition, the function of subunit e is still unknown for the transmembrane domain subunits [5]. Figure 4 shows the evolutionary relatedness of the D. citri Ac45 protein. It is a relatively new protein, critically associated with the assembly of a certain cell type V-ATPase, and is still being studied [6]. For this select group of insect species, Ac45 groups and forms a clade with the other hemipteran protein sequences (Figure 4). Ac45 is a variable gene when comparing V-ATPase across the domains of life, a paralog variability that is also seen among different orders of insects (Table 8) [6,33]. Ac45 has diverged the most of all the V-ATPase subunits in D. citri compared with other insects. This divergence is seen in phylogenetic analysis, denoted with longer branch lengths ( Figure 5), and is also supported in the values of the pairwise alignments, in which the protein shares very little sequence identity across the query lengths ( subunits (Figure 6a, Table 5). This likely reflects the limits in resolution with current whole RNA isolation and sequencing methods; nevertheless, it still indicates the relatively low total expression. The Ac45 protein has not been observed to exist in every cell type depending on the organism and so is not necessarily utilized by every V-ATPase in the psyllid [6]. Thus, the expression data agree with previously published research.

CONCLUSION
The V-ATPase is a fundamental enzyme that functions exclusively as an ATP-dependent proton pump in almost every eukaryotic cell. V-ATPase allows for the proper functioning of endosomes and the Golgi apparatus, and it generates a proton-motive force in organelles and across plasma membranes, which is utilized as a driving force for secondary transport processes [1]. Identification of these enzymes in the hemipteran, D. citri, provides a novel insect lineage for studies of insect evolution and biology, and may also provide potential targets for D. citri-specific molecular mechanisms for the management of HLB in citrus production systems [34][35][36]. D. citri shows no deviation in the expected copy numbers of each of the V-ATPase genes (Tables 6-8). The data collected from D. citri reveal consistency among the genes previously characterized as highly conserved, such as V-ATPase c, d, A, and B (Tables 2-4) [3,32]. While expression data were not available for V-ATPase d, V-ATPase c shows comparatively high expression levels overall, and differential expression -647.9 versus 1705.25 TPM -in the guts of adult psyllids fed on uninfected versus infected C. medica leaves ( Figure 5, Table 5). Conversely, the Ac45 gene shows low expression throughout life stages and tissues compared with other V-ATPase genes; however, the highly divergent nature of this gene may serve as a species-specific targeted approach to psyllid control (Table 4, Figure 6). In hemipterans, RNAi efficacy has been successfully demonstrated for psyllids, whitefly, and leafhoppers [34][35][36][37][38][39][40][41]; planthoppers [42,43]; bedbugs [44]; and others [45][46][47][48][49]. RNAi specifically targeting the V-ATPases in hemipteran insects has been reported for the corn planthopper, Peregrinus maidis (Ashmead) (Hemiptera: Delphacidae) [12]; the corn leafhopper, Dalbulus maidis (Hemiptera: Cicadellidae) [13]; the brown planthopper, Nilaparvata lugens (Stål) (Hemiptera: Delphacidae) [42]; and the bedbug, Cimex lectularius L. (Hemiptera: Cimicidae) [44], resulting in increased mortality and reduced fecundity. Thus, the highly divergent nature of these gene sequences provides unique targets that may serve as species-specific targeting for RNAi approaches in the management of psyllid vectors and other hemipteran pests [50,51].

REUSE POTENTIAL
The manually curated gene models generated through this D. citri community annotation project will be available as part of the official gene set version 3. Analysis of these data, including BLAST and expression profiling, can be conducted using the citrusgreening.org website [52] and Citrus Greening Expression Network (CGEN). The improved annotations presented in this study will facilitate experimental design to investigate the potential of V-ATPases as gene targets for therapies to control D. citri. Research considering differential expression patterns of V-ATPase transcripts in psyllids fed on CLas-infected plants should be conducted. Additional studies are also required to confirm the role of the Ac45 protein, as its divergent nature may provide novel and species-specific gene targets, potentially through the use of RNAi, to control psyllid populations and reduce the effects of pathogens such as CLas.

DATA AVAILABILITY
The datasets supporting this article are available in the GigaScience GigaDB repository [24].

EDITOR'S NOTE
This article is one of a series of Data Releases crediting the outputs of a student-focused and community-driven manual annotation project curating gene models and if required, correcting assembly anomalies, for the Diaphorina citri genome project [53].

ETHICAL APPROVAL
Not applicable.

CONSENT FOR PUBLICATION
Not applicable.