Comparative mapping of the 22q11.2 deletion region and the potential of simple model organisms

22q11.2 deletion syndrome (22q11.2DS) is the most common micro-deletion syndrome. The associated 22q11.2 deletion conveys the strongest known molecular risk for schizophrenia. Neurodevelopmental phenotypes, including intellectual disability, are also prominent though variable in severity. Other developmental features include congenital cardiac and craniofacial anomalies. Whereas existing mouse models have been helpful in determining the role of some genes overlapped by the hemizygous 22q11.2 deletion in phenotypic expression, much remains unknown. Simple model organisms remain largely unexploited in exploring these genotype-phenotype relationships. We first developed a comprehensive map of the human 22q11.2 deletion region, delineating gene content, and brain expression. To identify putative orthologs, standard methods were used to interrogate the proteomes of the zebrafish (D. rerio), fruit fly (D. melanogaster), and worm (C. elegans), in addition to the mouse. Spatial locations of conserved homologues were mapped to examine syntenic relationships. We systematically cataloged available knockout and knockdown models of all conserved genes across these organisms, including a comprehensive review of associated phenotypes. There are 90 genes overlapped by the typical 2.5 Mb deletion 22q11.2 region. Of the 46 protein-coding genes, 41 (89.1 %) have documented expression in the human brain. Identified homologues in the zebrafish (n = 37, 80.4 %) were comparable to those in the mouse (n = 40, 86.9 %) and included some conserved gene cluster structures. There were 22 (47.8 %) putative homologues in the fruit fly and 17 (37.0 %) in the worm involving multiple chromosomes. Individual gene knockdown mutants were available for the simple model organisms, but not for mouse. Although phenotypic data were relatively limited for knockout and knockdown models of the 17 genes conserved across all species, there was some evidence for roles in neurodevelopmental phenotypes, including four of the six mitochondrial genes in the 22q11.2 deletion region. Simple model organisms represent a powerful but underutilized means of investigating the molecular mechanisms underlying the elevated risk for neurodevelopmental disorders in 22q11.2DS. This comparative multi-species study provides novel resources and support for the potential utility of non-mouse models in expression studies and high-throughput drug screening. The approach has implications for other recurrent copy number variations associated with neurodevelopmental phenotypes.

Background 22q11.2 deletion syndrome (22q11.2DS, MIM #188400/ #192430) is the most common micro-deletion syndrome in humans with an estimated prevalence of at least 1 in 4000 live births [1,2]. Formerly known as velocardiofacial or DiGeorge syndrome, this multi-system condition is associated with a broad range of developmental features including congenital cardiac and palatal anomalies, intellectual disabilities, hypoparathyroidism, and subtle facial dysmorphism [2][3][4]. Developmental delay and later onset disorders affecting the nervous system are particularly common [5]. These include attention deficit hyperactivity disorder [6,7], anxiety disorders [8,9], autism [10,11], epilepsy, schizophrenia [2,12], and early-onset Parkinson's disease [13,14]. The phenotypic manifestations of the syndrome are thought to be related at least in part to reduced gene dosage in the 22q11.2 deletion region that in turn interferes with normal protein functioning [15].
The typical associated~2.5 Mb 22q11.2 deletion is present in >85 % of individuals with 22q11.2DS [16,17], while a smaller proximal nested~1.5 Mb deletion occurs in~10 % of cases [18,19]. The associated 22q11.2 deletions are mediated by segmental duplications, or low-copy repeats (LCRs) that confer susceptibility of the region to copy number variation through non-allelic homologous recombination [20,21]. The penetrance and variable expressivity of major associated phenotypes appear to be largely independent of deletion size [22,23]. The few mRNA sequencing and protein expression studies of individuals with 22q11.2DS published to date [24][25][26][27][28][29][30][31][32][33] illustrate the complexity of linking specific genes to the phenotypes associated with this disorder. Much remains to be known about the individual and collective roles of 22q11.2 deletion region genes in modulating associated phenotypes. Model animals will undoubtedly play an essential role in this discovery process.
Mouse models have already been proven useful for characterizing the molecular function of 22q11.2 genes and establishing a link between certain genes and 22q11.2DS associated phenotypes [34]. The syntenic region on mouse chromosome 16 has a high degree of gene conservation to the human 22q11.2 region. Current engineered mouse models include deletions of large portions of the syntenic region and mutations of individual genes [34,35]. However, simple model organisms could also prove to be powerful tools for investigating genomic disorders such as 22q11.2DS. Their ease of genetic manipulation, amenability to high-throughput behavioural screening, and short generation times make simple organisms attractive potential resources. The potential for simple model organisms to reveal the genetic mechanisms underlying 22q11.2DS phenotypes remains essentially unexamined however.
As an initial step in determining the utility of simple model organisms in the study of 22q11.2DS, we generated an updated, comprehensive 22q11.2DS human gene map and investigated the evolutionary conservation status of genes within the 22q11.2 region in three common model organisms: the zebrafish, Danio rerio (D. rerio), the fruit fly, Drosophila melanogaster (D. melanogaster), and the worm, Caenorhabditis elegans (C. elegans). We included the otherwise well-reviewed mouse models [34,35] for comparison. We then conducted a comprehensive review of gene function and phenotypic alterations related to 22q11.2 gene homologue disruptions and developed a novel comprehensive resource of available knockout and knockdown models. The results may help to accelerate the identification of novel genotype-phenotype correlations in 22q11.2DS and inform pathogenesis of, and drug development for this disorder and its commonly associated features.

Human 22q11.2 region characterization
The human 22q11.2DS deletion region, genetic content, and order were mapped from NCBI Gene Homo sapiens Annotation Release 105 using Affymetrix CytoScan HD (Santa Clara, CA, USA) array mean breakpoints (chr22:18,820, 303-21, 489,474) ascertained from 16 patients with confirmed 22q11.2 deletions (Fig. 1). Fourteen of the 16 patients had deletions covering most of the 22q11.2 region (~2.5 Mb) while two had smaller, nested proximal deletions. The same region was obtained with a larger, previously described patient population (n = 99) using Affymetrix Human SNP 6.0 breakpoints [3,36]. All patients provided consent and the study was approved by local research ethics boards [3]. We accessed the Database of Genomic Variants to establish the corresponding locations of major UCSC segmental duplications across the deletion region (http:// dgv.tcag.ca/dgv/app/home; accessed 1 December 2014). Build GRCh37 gene coordinates were used to ensure congruency across all databases used in this study. We omitted the few genes that move outside the 22q11.2 deletion region in build GRCh38 (i.e., the segment flanked by proteincoding genes TMEM191B…RIMBP3).

Gene conservation and function in model species
To identify putative orthologs of human 22q11.2 region protein-coding genes in the zebrafish (D. rerio), fruit fly (D. melanogaster), worm (C. elegans), and mouse (M. musculus), we employed the reciprocal best hits method, i.e., the protein products of genes in two different genomes represent the best hit in the opposite genome, using protein Basic Local Alignment Search Tool (blastp) analysis with the UniProtKB database (http://www.uniprot.org/uni prot/) including both Swiss-Prot and TrEMBL entries (accessed 1 December 2014). We ran blastp using each of the 46 22q11.2 deletion region protein-coding genes as a query against all proteins annotated in each genome of interest, using default settings and a maximum E-value threshold of 1 × 10 −6 [38,39]. We also required coverage of at least 50 % of any of the protein sequences in the alignments. In instances of multiple protein isoforms due to alternative splicing, the "canonical" sequence, as identified by UniProtKB, was selected for blastp analysis. To find orthologs as reciprocal best hits, we sorted blastp hits from the highest to the lowest bit score. Using this sorting method, the first hit was therefore the best hit. If the next hit had the very same score, there would be more than one hit (the method can therefore produce multiple orthologs). The same procedure was performed in the opposite direction. In the zebrafish, an organism that has undergone genomewide duplication, we included multiple hits if the scores of putative homologues were very similar, and both hits were consistently identified across multiple databases (e.g., RefSeq). NCBI Entrez Gene was then used to individually search all putative orthologs to establish organism-specific gene location (Additional file 1). The conservation status of the seven 22q11.2 region miRNAs identified was examined using miRBase21 (accessed in December 2014) [40]. Human non-coding genes (n = 10) including one readthrough transcript, and pseudogenes (n = 27) in the 22q11.2 deletion region were not investigated further.
To identify available knockout and knockdown models of the identified 22q11.2 region homologues and collate their phenotypic manifestations, we conducted a systematic search (accessed 1 December 2014) of species-specific databases including: WormBase (http://www.wormbase.org/), FlyBase (http://flybase.org/), ZFin (http://zfin.org/), and MGI (http://www.informatics.jax.org/) databases for C. elegans, D. melanogaster, D. rerio, and M. musculus, respectively. A secondary PubMed literature review confirmed that all studies examining orthologs in our model organisms of interest were included. Knockouts (homozygote) were defined as mutant models that did not produce a functional protein product due to a premature stop codon, a disruptive insertion, or full excision of a gene. For all model organisms discussed, we note only the availability of homozygous knockouts; these are more difficult to generate and are required for heterozygous knockout animals (the result of a cross of a homozygous knockout with a wild-type strain). We have however provided the known phenotypes of heterozygous knockout models for genes conserved across all examined organisms. Knockdown models were defined as those with reduced gene expression induced by any technology that interfered with the translation of a gene after it had been transcribed. The 17 genes conserved across model organisms were examined to document the availability of phenotypic information of mutant models. Single gene mutations have been reported in humans for 22q11.2DS genes, however these were outside the scope of this study.
In the typical 22q11.2 region, there were also 27 pseudogenes, one read-through transcript (classified as a noncoding RNA) (SEPT5…GP1BB), nine non-coding RNA genes, and seven microRNAs (miRNAs; Fig. 1). Recent studies propose that the miRNA processing protein Pasha, encoded by DGCR8, which lies within the 22q11.2 deletion (See figure on previous page.) Fig. 1 Genetic landscape of the human 22q11.2 region. The typical~2.5-Mb 22q11.2DS deletion spans 90 RefSeq genes (see text for details). Region breakpoints are mediated by four chromosome specific low-copy repeats (LCRA-D; approximate locations shown). Gene expression, indicated by a green circled check mark, was established using The Human Brain Transcriptome. Data for decreased expression with hemizygosity were collated from experimentally demonstrated [24,[28][29][30][31][32][33] reductions in gene expression in blood cells from patients with 22q11.2DS. Gene names within a rectangle denote the 17 genes conserved across the mouse, zebrafish, fruit fly, and worm region, may play a role in modifying genome-wide expression of target genes that contribute to the neuropsychiatric phenotypes associated with 22q11.2DS, together with the region's high density of miRNAs [36,[41][42][43]. Three of the seven miRNAs (MIR185, MIR1306, MIR1286) have been found to be expressed in the brain, while two were not (MIR3618, MIR649), [44] and the other two (MIR4761, MIR6816) have yet to be investigated.

22q11.2 region gene conservation in model organisms
The well-studied mouse syntenic region of the human proximal (1. The zebrafish also exhibited a high degree of gene conservation to the human 22q11.2DS region. In total, 37 (80.4 %) of the 46 protein-coding human homologues had putative homologues in the zebrafish (Table 1, Fig. 2). Compared to the genes contained in the human 22q11.2 region, the fruit fly had available homologues for 22 (47.8 %) protein-coding genes, and the worm had 17 (37.0 %) available homologues.
No miRNAs were conserved in the zebrafish, fruit fly, or worm. The two miRNA conserved in the mouse were MIR185 and MIR1306. Human non-coding genes (n = 9), a read-through transcript, and pseudogenes (n = 27) in the 22q11.2 deletion region were not investigated in the model organisms studied.

Availability of knockout and knockdown models of 22q11.2 region homologues in model organisms
The high proportion of conserved protein-coding genes and their arrangement in the mouse has permitted the construction of contiguous multi-gene deletion models. These include several short and long deletion models, as previously reviewed [34,35]. Of the 40 homologous genes in the mouse, 31 (77.5 %) had available homozygous knockouts (Table 1). Notably, no mouse knockdown models were identified.

Availability of phenotypic information for conserved 22q11.2 region genes in animal models
Examination of the 17 genes conserved across species showed substantial variability in the availability and comprehensiveness of phenotypic information for homozygous and heterozygous knockouts, and knockdown models ( Table 2). In the mouse, there were 13 genes with mutants available, of which nine had some form of phenotypic characterization. Zebrafish also appeared to be under-investigated; of 12 conserved genes with mutants available, phenotypic information was available for only six. Notably, a DGCR8 knockout has not been phenotypically assessed in zebrafish. More phenotypic information was available for mutants in the fruit fly (13 of 17 genes) and worm (15 genes), possibly due to the use of forward genetic screens in these organisms [47,48]. Only six genes were phenotypically characterized across all three of the simple model species (SLC25A1, UFD1L, TBX1, MED15, PI4KA, and SNAP29). Findings from these phenotypic studies are discussed below in the context of clinical manifestations of 22q11.2DS. We note that, as for all genetic studies, it is important to determine whether phenotypic effects are related to a specific background strain. This is an important consideration for all model animals including the mouse [49], zebrafish [50], fruit fly [51], and worm [52]. For example, Prodh homozygous in mice mutants were shown to be defective in prepulse inhibition [53], but this effect was dependent on genetic background [54].

Discussion
Here, we defined a comprehensive gene map of the most common human micro-deletion syndrome, 22q11.2DS, and conducted the first systematic examination of 22q11.2 deletion region gene conservation in simple model organisms. We developed a comprehensive resource of available knockout and knockdown models of conserved 22q11.2    Genes ordered by proximal to distal 22q11.2 locus position; "-" indicates no knockout (KO) or knockdown (KD) model available. Lethality for KD models is not included as this may vary based on when the gene is suppressed region homologues. Our results demonstrate that of the human 22q11.2 region protein-coding genes, a substantial proportion is conserved in simple model organisms. These mutant model organisms are amenable to extensive phenotypic characterization. This novel comparative multi-species resource can be used to provide initial insights into how simple animal models could be used to investigate the multi-system congenital and neurodevelopmental conditions associated with hemizygous 22q11.2 deletions.
Advantages of using a non-murine animal model in the study of 22q11.2DS The amenability of the zebrafish, fruit fly, and worm to genetic manipulation compared with the mouse could facilitate rapid and cost-effective generation of individual targeted gene and multi-gene mutations (Table 1). Such manipulations will be essential to functional studies and gene variant interpretation of 22q11.2 region genes. Moreover, the ease of genetic manipulation could rapidly improve our understanding of how the 22q11.2 deletion may interact with the rest of the genome to mediate the variable expressivity of 22q11.2DS associated phenotypes through mechanisms such as translational modifications due to the loss of one copy of DGCR8 and 22q11.2 region miRNA genes [55]. The relatively high proportion of 22q11.2 deletion region gene knockouts available in the mouse compared with the zebrafish and the worm is notable given that the process of developing knockout mouse models is expensive and time-consuming, calculated to take on average of about 1 year at a cost of more than US $12,000 [56]. With the advent of the CRISPR/ Cas9 system, the cost and speed of developing knockouts for all organisms will substantially decrease [57].
Unlike the mouse, knockdown models for 22q11.2DS homologues are available for the zebrafish, fruit fly, and worm (Table 1). Gene knockdown technologies are advantageous because they can be used to reduce gene expression in a dose-dependent manner with a high degree of specificity [58,59], essential for examining 22q11.2DS, where dose-dependency is thought to underlie phenotypic changes [15]. All knockdowns found for 22q11.2 homologues in the zebrafish were generated using morpholinos [60]. Knockdowns in the worm and fruit fly commonly employed RNA interference (RNAi) methods. Notably, it is possible to knockdown two genes simultaneously using combinatorial RNAi in C. elegans and D. melanogaster [61,62]. There are, as yet, no examples of this for 22q11.2 region genes, although such experiments could yield critical insights into the possibility of epistatic effects between 22q11.2 region genes that may mediate the complex expression of 22q11.2DS associated phenotypes [54]. Although RNAi technology is available in some vertebrates including the mouse, it is rarely used due to the length of RNAi generation times and the lack of simplicity compared with invertebrates [58].

Suitability of non-murine model organisms to study 22q11.2DS
Useful and valid model organisms should demonstrate examples of convergence with the human 22q11.2DS phenotype, thus providing proof-of-principle of the utility of these organisms in characterizing gene function and disease modeling (e.g., as we observed here for TBX1). Cellular and phenotypic observations already made in lower animals could identify new avenues of investigation regarding the roles of particular genes in different 22q11.2DS phenotypes. The limited data available make it difficult at present to make phenotypic comparisons between species that could help to indicate conserved molecular functions of 22q11.2 genes, especially as such data were rarely collected in the context of 22q11.2DS (Table 2). However the data reveal opportunities for novel functional studies of these genes.
One example of particular relevance to neurodevelopmental processes comes from PRODH, which encodes a mitochondrial enzyme that metabolizes L-proline [63], an amino acid involved in modulating glutamatergic and GABA-ergic transmission [64]. A report of severe psychomotor delay in a male with a homozygous deletion [65] indicates PRODH may also be an important candidate gene for motor functioning. Movement abnormalities are commonly observed in individuals with 22q11.2DS, including hypotonia in infancy, delayed gross-motor milestones in childhood [66], susceptibility to antipsychotic-induced movement disorders [67], and early-onset Parkinson's disease [13], but it remains unclear which of the 22q11.2 genes are involved in these processes. Observations in the fruit fly, however, suggest a novel role for PRODH in motor pathways (Table 2). Its PRODH homologue, slgA, is prominently expressed in the nervous system during embryonic development and shows proline dehydrogenase activity [63]. Fruit flies with a homozygous PRODH mutation demonstrate severe locomotor defects and indecisive movement patterns compared with wild-type flies in an activity chamber assay [63]. Notably, this observation in the fruit fly spurred the study of Prodh in locomotion in mouse models, where the effects of Prodh have been less clear [53,54]. Further study on the role of PRODH in mediating 22q11.2DS associated motor deficits is warranted.
Gene knockout or knockdown technologies have only recently been used in non-mouse model organisms, specifically in the context of 22q11.2DS phenotypes. One of the few examples is SLC25A1, a mitochondrial citrate transporter important for proper mitochondrial functioning. In zebrafish, knockdown of the SLC25A1 homologue, slc25a1a, during embryonic development causes mitochondrial depletion and gross morphological defects that recapitulate some features of 22q11.2DS (Table 2). Zebrafish treated with a slc25a1a morpholino showed a dosedependent phenotype, with higher doses leading to more severe developmental dysmorphic abnormalities such as a flattened head and a marked reduction in the size of the entire cranial region, including the brain [68]. Additionally, fish with marked slc25a1a depletion had small hearts surrounded by pericardial edema. These results indicate that slc25a1a plays a role in cardiac as well as craniofacial and brain development, all cardinal features of 22q11.2DS. Notably, these phenotypes were rescued in treated animals when autophagy was blocked [68]. SLC25A1 exemplifies the potential for studies using knockdown models of simpler organisms to investigate the molecular underpinning of 22q11.2DS phenotypes that may yield novel therapeutic targets.
A novel avenue of investigation based on the high degree of gene conservation of mitochondrial genes in the 22q11.2DS region is also now indicated. Mitochondrial dysfunction is implicated in the etiology of brain-based disorders associated with 22q11.2DS, including developmental delay, schizophrenia, and Parkinson's disease [69][70][71], and perturbation of mitochondria and related pathways affects key cellular processes such as cell migration, apoptosis, and synapse formation [72]. Aberrant expression of mitochondrial genes, already documented in 22q11.2DS patients for five of the six (PRODH remains unexamined) 22q11.2 mitochondrial genes [73], all found in the human brain (Fig. 1), could mediate susceptibility to neurological conditions in individuals with 22q11.2DS. For example, MRPL40 knockouts in the fruit fly show defects in neurogenesis that compromise neurodevelopment (Table 2), and knockdowns of TXNRD2 in worms overexpressing human beta-amyloid peptide as a model for Alzheimer's disease were more susceptible to muscular dysfunction and paralysis [74]. More comprehensive studies of neurodevelopment and neurodegeneration in lower organisms could shed light on the molecular function of these critical proteins.
Individual organism suitability for studying specific 22q11.2DS phenotypes Like the mouse [34,35], zebrafish, fruit fly, and worm models are unable to singularly recapitulate all of the 22q11.2DS associated phenotypes. A limitation for all animal models is the challenge presented by complex disease phenotypes, such as the range of psychiatric disorders associated with 22q11.2DS. Particularly for simple organisms, limited behaviors, and the paucity of reproducible tests make it difficult to study such phenotypes. Although a considerable number of genes are conserved in the zebrafish, fruit fly, and worm, incomplete conservation limits the ability to fully investigate the roles and possible interactions between all 22q11.2 region genes.
However, these issues should not discourage the further study of specific 22q11.2DS orthologs. As discussed below for each proposed model organism, orthologs in lower animals could be very useful for clarifying the roles of 22q11.2 deletion region genes in basic neurodevelopmental trajectories (e.g. development of neuronal components), as well as organ development. For example, despite the absence of a heart in the worm, mutants of mls-1 (the ortholog of TBX1) indicate that mls-1 is involved in the specification of non-striated muscle during development [75], suggesting the potential utility for these mutants in studying how TBX1 mediates heart development in higher organisms. Discretion is essential when deciding which model organism to use to study particular 22q11.2DS phenotypes. However, organismspecific characteristics and developmental trajectories do suggest that certain 22q11.2DS phenotypes are particularly well suited for study in each of these non-mouse models.
Congenital cardiac defects, involving cardiovascular molecular development, are major manifestations of 22q11.2DS [3] where the molecular underpinnings are difficult to analyze in mammalian models, since there is rapid death without an intact cardiovascular system during embryonic development [76]. Improvements in generating conditional gene deletion models using technologies such as the Cre-loxP and Tet-On/Tet-Off system in mice offer the potential to circumvent these issues, but have not been fully developed in the context of investigating congenital heart defects [77,78]. Recently, the zebrafish has emerged as a highly advantageous vertebrate model for studying early cardiovascular development, largely due to the ability of the zebrafish embryos to obtain oxygen in the absence of blood circulation through passive diffusion. 1This permits survival of the initial stages of embryonic development and allows investigation of even severe cardiovascular defects [79]. The additional optical transparency of zebrafish embryos, used in combination with tissue-specific expression of fluorescent proteins, permits visualization of early molecular processes [80]. In humans, TBX1 is associated with heart defects, palatal anomalies, facial dysmorphism, and low calcium levels [81,82] albeit each with reduced penetrance [3]. In the zebrafish, inhibiting the TBX1 homologue leads to developmental defects of the pharyngeal arches, aortic arches, and thymus ( Table 2). Abnormal cardiac morphology was visible in roughly 20 % of knockdowns, with compromised cardiac performance in nearly all injected embryos [83]. The zebrafish provides a unique opportunity to study early development and how 22q11.2 gene dosage affects these processes through the use of knockdown technologies.
The fruit fly is particularly well-suited for the study of the myriad brain-related disorders associated with 22q11.2DS [3,6,9,13,84], including intellectual disability, and other neurodevelopmental [85] and neurodegenerative disorders [86]. The benefits of using D. melanogaster include its short generation time along with a high reproductive rate and the availability of powerful genetic and molecular tools. Genes of interest can be readily manipulated in a time and cell-or tissue-specific manner using well-established tools such as the GAL4/UAS-system [86]. Together with well-characterized developmental stages, a simple and defined nervous system, and the ability to conduct large-scale behavioral and neurophysiological assays, the fruit fly has proven a valuable tool in the study of genomic and neurological disorders [87]. Similar opportunities for study are possible for 22q11.2DS, although there have been few studies to date targeted to 22q11.2DS (Table 2). Nevertheless, studies in the fruit fly have already provided some novel insights into the molecular function of 22q11.2 deletion region genes pertinent to the associated neurological conditions. For example, in the context of identifying novel intellectual disability candidate genes in a large-scale screening study, flies with a knockdown of SNAP29 were found to have profound synaptic defects characterized by abnormal basal neurotransmission [88] (Table 2).
To better assess the role of 22q11.2DS genes in developmental processes, additional molecular information is needed, such as where and when a gene is expressed, and elucidation of its protein-protein interactions. An ideal model for these experiments is the worm. In addition to having conserved fundamental biological processes and homology with mammals, the worm is noteworthy for being highly amenable to forward and reverse genetic screens [89,90]. One 22q11.2DS gene provides an example of the utility of using the worm system to study 22q11.2DS is DGCR8, a component of the "microprocessor" complex essential for genome-wide miRNA production [91] that may mediate the expression of multiple 22q11.2DS associated phenotypes in patients, including schizophrenia [41][42][43]. In mice, Dgcr8 has been implicated in altering the biogenesis of genome-wide brain miRNA [92]. Indeed, miRNAs were first discovered in 1993 in genetic screens performed in the worm and were initially thought to be phenomena unique to nematode biology [93,94]. It was later found that miRNA are widely conserved among eukaryotes as functional non-coding RNAs [95] and their role has been further studied in the mouse [96], zebrafish [97], and fruit fly [98].
DGCR8 was first identified as a candidate gene for miRNA processing based on a genome-wide two-hybrid analysis of D. melanogaster where the protein product, Pasha, was shown to interact with Drosha [99,100]. Inactivation of a temperature sensitive allele of DGCR8 in C. elegans lead to the accumulation of protein products of other genes elsewhere in the genome, particularly let-7, and a reduction in life span [101]. Similar mechanisms could be associated with the unexplained premature mortality reported in individuals with 22q11.2 DS [102]. In another study investigating gene-gene interaction mechanisms, RNAi targeting the worm DGCR8 homologue resulted in exacerbation of the uncoordinated motor phenotype caused either by a mutation in the human tau-FTDP-17 homologue or by unc (uncoordinated) mutations [103]. This is of interest in light of motor defects seen in patients with 22q11.2DS [104,105]. Comparable interactive mechanisms, involving effects of hemizygosity of DGCR8 on expression of mRNA across the genome, may also be operating in 22q11.2DS [41][42][43]. Differences in mRNA expression may also be mediated by another 22q11.2 region gene, DGCR14, whose ortholog in C. elegans has been implicated in promoting proper mRNA splicing when splice sites are compromised [106]. Notably, conservation is a key means of determining relevance of a non-proteincoding sequence in humans [107].
Another as yet unexplored area in the context of model organisms and 22q11.2DS relates to the development of pharmaceutical agents to treat 22q11.2DS related phenotypes. Collectively, D. rerio, D. melanogaster, and C. elegans are particularly suited to screening chemical libraries for potential drug development, circumventing the high financial and time investment for M. musculus [108].

Study limitations
We used a reciprocal best hits method to identify putative orthologs of human 22q11.2 genes in simple model organisms, a common and well-established method to probe orthology based on sequence similarity [39]. In this exploratory study, we did not restrict to a minimum sequence identity in order to identify all putative orthologs of the human 22q11.2 genes. Similarity in protein sequences does not necessarily translate to conserved function or patterns of gene expression across species. Further experiments are needed to assay possible conserved functional roles [109,110] of the homologues identified here. Importantly, we found that our results were consistent with previous homology relationships described for 22q11.2DS genes (e.g., PRODH [111], UFD1L [112], DGCR6 [113], MED15 [114], TSSK2 [115], and TXNRD2 [116]). Using these methods, we identified multiple putative homologues for individual 22q11.2 deletion region genes in the zebrafish, possibly related to gene duplication [45]. In these cases, functional studies are required to examine which homologue may have a conserved function. Additionally, using the stringent criteria for the reciprocal best hits method, homology may be difficult to establish for genetically distant species. For example, query coverage was too low to identify a reciprocal best hit for the mitochondrial gene ZDHHC8 in the fruit fly and worm. Our analyses were restricted to protein-coding genes and miR-NAs. The examination of other non-coding genes will be