Skip to main content
NCBI home page
Integrative and Comparative Biology logoLink to Integrative and Comparative Biology
. 2024 Apr 2;64(2):556–564. doi: 10.1093/icb/icae009

Problems with Paralogs: The Promise and Challenges of Gene Duplicates in Evo-Devo Research

Kevin D Deem 1,✉, Jennifer A Brisson 2
PMCID: PMC11406157  PMID: 38565319

Synopsis

Gene duplicates, or paralogs, serve as a major source of new genetic material and comprise seeds for evolutionary innovation. While originally thought to be quickly lost or nonfunctionalized following duplication, now a vast number of paralogs are known to be retained in a functional state. Daughter paralogs can provide robustness through redundancy, specialize via sub-functionalization, or neo-functionalize to play new roles. Indeed, the duplication and divergence of developmental genes have played a monumental role in the evolution of animal forms (e.g., Hox genes). Still, despite their prevalence and evolutionary importance, the precise detection of gene duplicates in newly sequenced genomes remains technically challenging and often overlooked. This presents an especially pertinent problem for evolutionary developmental biology, where hypothesis testing requires accurate detection of changes in gene expression and function, often in nontraditional model species. Frequently, these analyses rely on molecular reagents designed within coding sequences that may be highly similar in recently duplicated paralogs, leading to cross-reactivity and spurious results. Thus, care is needed to avoid erroneously assigning diverged functions of paralogs to a single gene, and potentially misinterpreting evolutionary history. This perspective aims to overview the prevalence and importance of paralogs and to shed light on the difficulty of their detection and analysis while offering potential solutions.

Motivation

A core focus of evolutionary developmental biology (evo-devo) is to discover the conserved and derived aspects of developmental gene regulation and function that underlie organismal diversity. For example, determining how novel structures such as the insect wing originated requires investigating the role of a common set of genes in homologous tissues across clades or in serially homologous tissues across segments (for reviews, see Clark-Hachtel and Tomoyasu 2016; Tomoyasu 2021). Likewise, identifying shared gene regulation between nonhomologous structures may indicate evolution via co-option of gene regulatory networks, as is the case with some butterfly eyespots (Murugesan et al. 2022). These analyses rely heavily on the reliable assignment of homology to genes in species separated by huge evolutionary distances, a task that is greatly complicated by gene duplication. Gene duplication is an incredibly important evolutionary process, providing a tabula rasa for the evolution of new gene functions (Ohno 1970). Duplicated genes (paralogs) are especially difficult to detect and functionally characterize in nonmodel organisms, and often go overlooked. Because they are so understudied, they are an exciting, underexplored frontier for evolutionary biology. Below, we discuss the promise and challenges of studying paralogs in evo-devo research.

Origins and evolutionary fates of paralogs

Gene duplication has been implicated as the primary source of new genes (Ohno 1970), and as such, a major driving force of evolution. Genes duplicate by a variety of mechanisms, including small-scale duplication events like tandem duplication from unequal crossing over or large-scale events such as chromosomal segmental duplications or even whole-genome duplication. Likewise, clusters of tandemly duplicated genes can be copied via whole-genome duplication, creating duplicates of duplicates (Kuzmin et al. 2022). While all of these genes are considered homologous, there is specific terminology used to delineate different versions of genes derived from different types of duplication events, or even speciation events (reviewed in Koonin 2005). For this paper, we will largely focus on the terms paralog and ortholog. Genes that arise from a duplication event are paralogs, while genes that arise through speciation are orthologs.

While classical models of paralog evolution predicted that the vast majority of duplicated genes would quickly acquire and fix nonfunctionalizing mutations (Ohno 1970; see Force et al. 1999 for quotes from others echoing that sentiment), the observation that far more gene duplicates exist than predicted has led to a revised model of duplicate gene retention (Force et al. 1999). It is now widely accepted that there are a variety of evolutionarily important functional paths for new paralogs to take following a duplication event (reviewed in Kuzmin et al. 2022). Paralogs may continue to perform the same function (redundancy), divvy up ancestral functions (sub-functionalization), acquire new functions (neo-functionalization), or perform some combination of these three (Force et al. 1999; Kuzmin et al. 2022). For instance, imagine a gene that is expressed in tissues A, B, and C. Following duplication and divergence, the resultant paralog 1 may be expressed in tissues A and B and paralog 2 expressed in tissues B and C, as well as in another tissue D. In this case, paralogs 1 and 2 function redundantly in tissue B and have sub-functionalized in tissues A and C, and paralog 2 has neo-functionalized in tissue D. This is a simplified case where sub- and neo-functionalization occurred only via changes in gene expression domains, due to changes in paralog regulation. In reality, sub- or neo-functionalization can occur via regulatory or protein-coding changes, or both. Protein sequence changes can additionally result in alterations in protein–protein, protein–DNA, or protein–RNA interactions, or enzymatic activity (Kuzmin et al. 2022).

Paralogs in insect evolution

Paralogs important for the evolution of segmented body plans

Hox genes present an excellent example of how gene duplications followed by neo-functionalization have massively contributed to animal development and morphological evolution. In bilaterian animals, clustered paralogs termed the homeobox or Hox genes coordinate anteroposterior patterning of body segments (see Gehring et al. 2009; Iimura et al. 2009 for reviews). The ancestral Hox cluster was the product of repeated small-scale, tandem duplications of a primordial urhox gene (reviewed in Gehring et al. 2009; Singh and Krumlauf 2022). Insects have a single complement of 10 Hox genes usually contained within a single cluster, although this cluster has been split up in some insects (Pace et al. 2016; Mulhair and Holland 2024). Each insect Hox gene, except for zen and ftz (Falciani et al. 1996; Damen 2002), specifies the identity of a segment or group of segments of the body, the order of which along the anteroposterior axis curiously corresponds to the order of the gene within the cluster (so-called colinearity) (Lewis 1978; Kaufman et al. 1980; Gehring et al. 2009). However, early in the vertebrate lineage, two whole-genome duplications produced three additional Hox clusters, resulting in four in total. Functional redundancy among the duplicated Hox clusters relaxed constraint on their members (i.e., only one copy of each Hox gene was needed to perform anterior–posterior patterning). As a result, some vertebrate Hox genes have gained new functions, such as proximo-distal patterning of limbs, while others have become nonfunctional (Singh and Krumlauf 2022).

Paralogs important for insect flight

Other examples of paralogs with important roles in animal morphology are the nubbin/pdm2 and tiptop/teashirt paralogs in Drosophila melanogaster, which are transcription factors that play integral roles in the development of the flight machinery. Nubbin and teashirt are important for the development and delineation of the blade and hinge/body wall attachment of the wing, respectively (Zirin and Mann 2007).

The paralogs pdm1 (nubbin) and pdm2 in D. melanogaster derive from a single ancestral nubbin ortholog, which duplicated early in the Brachycera lineage ∼200 mya (Loker and Mann 2022). Studies in other insects with a single nubbin ortholog suggest an ancestral function of this gene in both central nervous system and wing development (Li and Popadić 2004; Tomoyasu et al. 2009; Biffar and Stollewerk 2014; Fisher et al. 2021; Tidswell et al. 2021; Fernandez-Nicolas et al. 2022). In Drosophila, while both nubbin and pdm-2 are expressed in the developing nervous system, pdm-2 is not expressed in the wing, a result of sub-functionalization (Loker and Mann 2022). Intriguingly, upon loss of nubbin function, pdm-2 gains expression in the wing and rescues its development, exhibiting functional redundancy (Loker and Mann 2022).

The Drosophila paralogs tiptop (tio) and teashirt (tsh) derive from an ancestral tiotsh gene important for wing and body wall patterning (Shippy et al. 2008; Zhang et al. 2020; Fisher et al. 2021; Medina-Jiménez et al. 2021). The duplication event of the ancestral tiotsh gene occurred within the Diptera, ∼260–125 mya (Wiegmann et al. 2011). During the development of adult structures in flies such as the wing hinge, tio and tsh are expressed in the same domain and function redundantly (Bessa et al. 2009). However, tio and tsh have sub-functionalized in Drosophila embryonic development and are expressed in different domains (Laugier et al. 2005). Much like in the case of nubbin and pdm-2, loss of tio function in the early embryo causes gain of tsh expression in the tio domain and results in developmental rescue, revealing further functional redundancy (Laugier et al. 2005). In both of these cases, sub-functionalization is achieved via repression of one paralog by another, which is alleviated upon loss of the repressing paralog (Laugier et al. 2005; Bessa et al. 2009; Loker and Mann 2022). Paralog repression may be an important way for genes to overcome detrimental dosage effects resulting from duplication, while simultaneously specializing in distinct ancestral functions (Force et al. 1999; Gout and Lynch 2015; Kuzmin et al. 2022).

Paralogs important for the loss of insect flight

Pea aphid (Acyrthosiphon pisum) males present an interesting example of a recently duplicated paralog that functions to suppress wing development. Our group recently showed that the presence of a follistatin paralog (fs-3) is polymorphic in pea aphid populations (Li et al. 2020). Wingless males have the fs-3 paralog present on their X chromosomes, while winged males do not. In addition to fs-3, there are two other fs paralogs in the pea aphid genome, fs-1 and fs-2, present on autosome 1 (Fig. 1A). The three paralogs share highly similar coding (Fig. 1B) and amino acid sequences (Fig. 1C), but the gene structure of fs-1 is quite different from that of fs-2 and fs-3 (Fig. 1A). These paralogs are far younger than any yet discussed, with both duplication events estimated to have occurred <40 mya. The similarities among these three paralogs have caused numerous problems in our attempts to study the evolutionary events leading up to fs-3′s role in male winglessness. These include issues with the proper assembly of gene loci (Li et al. 2020), and the design and function of in situ hybridization probes, cloning primers, and quantitative reverse transcription polymerase chain reaction (qRT-PCR) primers. In the following sections, we will detail how these issues can arise with highly similar paralogs.

Fig. 1.

Fig. 1

Open in a new tab

Similarities and differences among pea aphid fs paralogs. (A) Diagram of the gene structures of fs paralogs. Transcriptional start sites are marked with arrows, exonic sequences are represented by vertical bars, and introns are depicted as horizontal lines. Scale bar: 10 kb. Changes that have occurred in intronic sequences are evident in differential spacing of exons between paralogs. (B) Alignment of fs-2 and fs-3 mRNA sequences (tan arrows) to that of fs-1. Red shading of fs-2 and fs-3 indicates sequence conservation with fs-1. The coding sequence of fs-1 is shown as a maroon arrow. While mismatches are present within the coding region, sequence is dramatically less conserved in the noncoding 5′- and 3′-untranslated regions (UTRs). Scale bar: 200 nt. (C) Amino acid alignment of fs paralogs. Conserved residues are highlighted in yellow. A conservation track is shown above the aligned residues in which darker shades of gray or black indicate increasing or perfect sequence identity, respectively. Aside from the N- and C-termini, all three protein sequences are highly similar. Alignments were performed and visualized using SnapGene® software (from Dotmatics; available at snapgene.com).

Strategies for successful paralog detection and analysis

The identification and analysis of paralogs relies heavily on the detection of unique sequence within and surrounding duplicated loci. Changes in the amino acid sequence between paralogs accrue relatively slowly, as there is a high probability that changes are deleterious. Coding sequences typically diverge more rapidly than amino acid sequences (via accumulation of synonymous changes) and noncoding sequences diverge faster still (Li et al. 2007). Because of this, even paralogs that diverged tens of millions of years ago can have highly similar amino acid sequences, but divergent noncoding sequences (Li et al. 2007). In noncoding regions, regulatory sequences may be somewhat conserved, but they usually are dispersed among more divergent intergenic and intronic sequences (Li et al. 2007). In the following sections, we will describe the challenges this presents for genome assembly, gene expression analysis, and functional analyses of paralogs, as well as strategies for addressing these issues.

Paralog identification

Genome assembly and annotation

Paralogs are often misassembled or overlooked during de novo genome assembly and annotation (Denton et al. 2014). Short-read assemblers often collapse reads from the highly similar exonic sequences and nearby intronic sequences of recent paralogs into single misassembled genes that are a mishmash of paralogous sequences (Mastretta-Yanes et al. 2014; Li et al. 2020). In other cases, individual paralogous exons and their surrounding unique intronic sequence may become isolated within their own unplaced scaffolds. If these lone exons do not code for conserved protein domains, they can be difficult or impossible to homologize to other genes (Denton et al. 2014).

The use of long-read sequencing technologies greatly assists in the detection of paralogs. Long-read assemblies can resolve entire paralogous gene structures regardless of exon similarity and even large chromosomal duplications or rearrangements, thanks to read lengths exceeding the size of many genes (Chaisson et al. 2019; Vollger et al. 2019; Logsdon et al. 2020; Fadaie et al. 2021; Zhang et al. 2022). Indeed, our group used Oxford Nanopore Technologies (ONT, Oxford, UK) long reads to assemble the genomes of winged and wingless males separately, to resolve the X-chromosome insertion containing the fs-3 locus in the pea aphid (Li et al. 2020). Still, genome assemblies exclusively using long reads can suffer from read inaccuracy and contiguity issues, and many of the best genome assemblies use a hybrid approach (Chaisson et al. 2019; Logsdon et al. 2020). However, for cases in which hybrid assembly is not ideal, such as highly complex or repetitive genomes, the recent increase in accuracy for certain long-read technologies, such as HiFi reads from Pacific Biosciences (PacBio, Menlo Park, CA), may allow for long-read-only assembly (Hon et al. 2020). Unfortunately, the majority of genome assemblies to date have been produced exclusively from short reads, making it challenging to know with confidence how many paralogs exist in these species. The development of specific pipelines for the high-quality de novo hybrid genome assembly of nontraditional model organisms (Jaworski et al. 2020) and the recent releases of such assemblies (Nowak et al. 2019; Tang et al. 2021; Shu et al. 2023; Wang et al. 2023) provide hope for the reliable detection of paralogs in published genomes in the near future.

Analyzing paralog expression

Examining paralog expression patterns and levels can help determine whether the gene duplicates have likely retained similar functions or whether one or both have taken on new or different roles. For instance, one might expect no expression of a paralog following pseudogenization, mutually exclusive expression nested within the ancestral domain for subfunctionalized paralogs, expression in a new domain if a paralog has neo-functionalized, or expression in the same domain if paralogs are acting redundantly. This section will cover the challenges and potential paths forward for the analysis of paralog expression patterns using common molecular techniques.

Immunohistochemistry

One way to discern paralog expression patterns is via immunohistochemistry, a technique that visualizes where a protein is present across tissues by using an antibody that specifically binds to that protein (Schwartzbach et al. 2016). However, the similarity of paralogous proteins can cause an antibody to recognize both paralogs, making it impossible to resolve individual paralog protein expression patterns (Fig. 1C; Li et al. 2007). This issue can potentially be addressed by using only the specific regions in which paralogous protein sequences differ as an antigen when making the antibody (Lee et al. 2016). However, antibodies raised against such peptide fragments may be limited in their ability to recognize the native protein, a requirement for staining tissues to analyze expression (Lee et al. 2016; GenScript n.d.). Alternatively, it is possible to use genome editing to tag paralogous proteins with epitopes recognizable by commercially available antibodies, but such experiments can be tricky and labor intensive (see “Expression analyses using CRISPR-Cas9”). A more practical solution may be to leverage the decreased similarity of paralogous mRNA sequences, as discussed next.

In situ hybridization or hybridization chain reaction

In situ hybridization (ISH) and hybridization chain reaction (HCR) visualize the mRNA expression patterns of genes across tissues (see Higo et al. 2023 for a review of the techniques), and are another way to explore the similarities and differences among paralogs. HCR has recently gained popularity for its use in nontraditional model insects and other arthropods, and a universal protocol has even been developed (Bruce et al. 2021). One advantage mRNA detection has over immunohistochemistry is that paralogous nucleotide sequences are more divergent than protein sequences, so it is often possible to design paralog-specific ISH or HCR probes. Still, paralogs can share binding affinity for the same in situ probes because of their nucleotide similarity, causing problems of cross-reactivity and unclear individual expression patterns. To increase probe-binding specificity, adjustments can be made to annealing temperatures and times, the number of stringency washes, or the types of chemical reagents. Alternatively, the more divergent noncoding 5′- and 3′-untranslated regions (Fig. 1 B) are attractive regions for ISH or HCR probe design, potentially solving issues of probe cross-reactivity. However, these regions often have low guanine-cytosine (GC) content and low-sequence complexity, which can make it difficult to design appropriate probes.

qRT-PCR

A method for detecting mRNA levels across tissues is quantitative Reverse Transcription-Polymerase Chain Reaction (qRT-PCR). A critical step in qRT-PCR experiments is the design of suitable primer pairs, with more stringent requirements than traditional PCR experiments, for target and control genes. The primer pairs require similar amplicon lengths (>200 bp), annealing temperatures (dependent on GC content), and amplification efficiencies, and ideally span an exon–exon boundary to avoid amplification of contaminating genomic DNA (Bio-Rad n.d.). Usually only a handful of candidate primer pairs meet these requirements, making the design of paralog-specific qRT-PCR primers especially challenging because they must also be designed within limited regions of unique sequence. qRT-PCR should be able to distinguish expressions between more similar paralogs than ISH or HCR because only small dispersed unique sequences are required. When this is not the case, RNA sequencing (RNA-seq) provides a more sensitive solution for analysis of highly similar paralog mRNA expression.

RNA-seq

Problems with paralogs sharing similar coding sequences also affect RNA-sequencing (RNA-seq) pipelines such as de novo transcriptome assembly and differential gene expression level analysis. Reads from paralogs sharing an identical or highly similar coding sequence can be misassembled into a single transcript. Likewise, if the genome contains misassembled paralogous gene loci, mapped read counts will not correctly reflect the expression level of either paralog. Furthermore, even if the assembly is correct for one or more paralog, identical or similar reads may map multiple times, complicating differential gene expression analyses. Multiple factors must be accounted for to correctly determine differential paralog expression: correct assembly, sufficient read length, proper read trimming, stringent read-quality mapping parameters, high read coverage, and potential exclusion of all multiply mapped reads from the analysis. One attractive alternative is the use of long-read RNA-seq, which may produce a high-quality de novo transcriptome with properly assembled paralogs (Oikonomopoulos et al. 2020).

Enhancer–reporter assays

When analyses of paralog expression domains prove too difficult due to highly similar gene products, an alternative is the enhancer–reporter assay, in which a putative enhancer is cloned upstream of a core promoter and reporter gene and reintegrated into the genome. The resultant reporter gene expression can be used to infer the domains of gene expression driven by the enhancer in its native context (reviewed in Kvon 2015). As previously mentioned, the noncoding sequences of paralogs diverge more quickly than their coding counterparts, making it easier to design primers targeting the enhancer sequences of a specific paralog. While the enhancers themselves might exhibit significant conservation, the noncoding sequence surrounding them is typically sufficiently divergent for specific primer design. However, finding enhancers in the genome, and then successfully analyzing their expression in a nontraditional model organism, can be a challenging task (see Tomoyasu and Halfon 2020 for a review on the topic).

Still, the use of enhancer–reporter assays to study gene regulation in nontraditional model insects has recently accelerated. Tools have now been developed for in silico enhancer prediction (Asma and Halfon 2019) and enhancer functional evaluation (Lai et al. 2018; Deem et al., submitted for publication) that function in a variety of insects, including beetles (Lai et al. 2018; Asma et al. 2024; Deem et al., submitted for publication), butterflies (Murugesan et al. 2022), bees (Asma et al. 2024), and mosquitoes (Schember et al. 2023; Asma et al. 2024). One major advantage of these tools is cost reduction, as enhancer prediction is virtually free to run (Asma and Halfon 2019) and the cross-species compatibility of the reporter vectors allows for fast and affordable prescreening of candidate enhancers in D. melanogaster (Lai et al. 2018; Murugesan et al. 2022; Schember et al. 2023; Asma et al. 2024; Deem et al., submitted for publication).

Expression analyses using CRISPR–Cas9

The CRISPR–Cas9 (Clustered Regularly Interspaced Short Palindomic Repeats - CRISPR-associated-protein 9) genome editing system, which employs a user-designed guide RNA to target a double-strand break in the genomic DNA via the Cas9 nuclease, can be used to knock in various constructs for visualizing gene expression (Bukhari and Müller 2019; Rezazade Bazaz and Dehghani 2022). In the fruit fly D. melanogaster, many CRISPR–Cas9-based strategies analyze the expression of genes by targeting noncoding regions with guide RNAs. This is especially useful for studying recently duplicated paralogs, where the noncoding regions are most likely to be unique and thus suitable for targeting (Fig. 1A and B) (Li et al. 2007; Loker and Mann 2022).

CRISPR-mediated insertion cassettes, introduced into introns via CRISPR–Cas9, use splice acceptor and donor sites to create an artificial exon and add exogenous coding sequence into a gene of interest (Li-Kroeger et al. 2018). In principle, this can be used to insert a protein tag into a specific paralog for fluorescent visualization by a variety of means, including immunohistochemistry as previously mentioned (Li-Kroeger et al. 2018).

Another use of the CRISPR–Cas9 system is to add a sequence containing a core promoter and reporter gene to an intron or nearby intergenic sequence of a gene of interest, which endogenous enhancers can activate (so-called enhancer trap) (O'Kane and Gehring 1987), allowing visualization of paralog-specific expression patterns. While many enhancer traps use transposon sequences to insert randomly as part of large-scale screens (Bellen et al. 1989; Trauner et al. 2009), the use of CRISPR–Cas9 allows targeting of the construct to a particular paralog of interest (Rezazade Bazaz and Dehghani 2022). Unfortunately, CRISPR–Cas9 also comes with its own shortcomings. Namely, double-stranded DNA breaks are far more likely to be repaired with only small insertions or deletions, without integrating large exogenous DNA cassettes, making CRISPR knock-in generally much less efficient than knock-out (Rezazade Bazaz and Dehghani 2022). Additionally, while there is a growing list of insects in which CRISPR–Cas9 has been implemented (Singh et al. 2022), this type of genome editing is technically challenging in many species because of their biology, and therefore working protocols are missing for most nonmodel systems.

Functional analysis of paralogs

Although expression levels and patterns can be revealing as to the putative function of paralogs, the ultimate test is to examine paralogs functionally. While the general challenges of sequence similarity also apply here, solutions for the functional analysis of paralogs are offered below.

RNA interference

RNA interference (RNAi) is a popular method for gene loss-of-function analysis in insect evo-devo and has been established in a number of species (see Zhu and Palli 2020 for a review of RNAi in insects). For this, a double-stranded RNA of ∼100–1000 bp must be synthesized with sequence corresponding to the mRNA sequence of a particular gene of interest and then injected into the organism. The cell processes this double-stranded RNA into ∼21–25 bp single-stranded fragments that degrade endogenous mRNA with perfect sequence complementarity (Zhu and Palli 2020). For highly similar paralogs, proper RNAi design presents the same potential problems as the design of in situ hybridization probes or qRT-PCR primers, where even 5′- or 3′-untranslated regions of paralogous mRNAs may not be sufficiently different to avoid cross-reactivity. Additionally, it should be noted that there are still many species in which RNAi has not been established or is challenging to perform. For these cases, an alternative method for loss-of-function analysis is again CRISPR–Cas9-mediated genome editing (when it is tractable in a given system).

Functional analysis using CRISPR–Cas9

As previously mentioned, CRISPR–Cas9 genome editing induces double-strand DNA breaks at the binding site of designed guide RNAs and has been established in a variety of insects (Rezazade Bazaz and Dehghani 2022; Singh et al. 2022). When using two or more guide RNAs spaced apart, large deletions of the intervening sequence can be achieved (Rezazade Bazaz and Dehghani 2022). This allows loss-of-function analysis of even recently duplicated paralogs via the deletion of whole exons or promoters with guide RNAs designed in highly divergent introns and intergenic sequences such as in pea aphid fs paralogs (Fig. 1A).

Conclusions

Gene duplication followed by divergence is a major driving force in morphological evolution. However, the detection of gene duplicates in newly sequenced genomes is a particularly difficult and often overlooked task, requiring advanced and expensive sequencing techniques and specialized software. Given the prevalence of short-read-based genome assembly strategies, along with their present inability to reliably assemble paralogous loci (Denton et al. 2014; Mastretta-Yanes et al. 2014), gene duplication and divergence is an understudied frontier in evolutionary biology. We have reviewed a few individual cases where paralogs have proven to be crucial to changes to animal form and function (Laugier et al. 2005; Bessa et al. 2009; Loker and Mann 2022; Singh and Krumlauf 2022), but undoubtedly many other important paralog stories remain undiscovered.

There is hope moving forward, not only for more reliable detection of paralogs in newly sequenced species, but also for species whose genomes have been sequenced using short-read technologies. Hybrid assembly with long reads corrected by short reads or high-accuracy long-read-only assembly is becoming a viable option for even small genome assembly projects in nonmodel species (Nowak et al. 2019; Hon et al. 2020; Jaworski et al. 2020; Tang et al. 2021; Shu et al. 2023). Additionally, short-read sequencing has already been performed for many species, so refinement by hybrid assembly would only require the cost of the long reads. However, this prospect relies on access to raw reads from previous sequencing projects, highlighting the importance of submitting raw reads (in addition to assembled contigs and scripts used for analysis) to publicly available databases for any genome project. For identification of paralogs, the ideal scenario is to obtain a highly contiguous hybrid short-read-corrected long-read genome assembly (at least of the paralogous loci), along with validation of the transcribed regions via long-read RNA sequencing and/or cloning of the transcript, and PCR validation of important noncoding regions.

Finally, techniques for analyzing the function of paralogs continue to improve. Of particular interest is the development of ever more sophisticated molecular tools in a variety of insect models, especially enhancer detection and reporter assays (Lai et al. 2018; Murugesan et al. 2022; Schember et al. 2023; Asma et al. 2024; Deem et al., submitted for publication). These tools help reveal more than the divergent expression patterns of paralogs. They allow dissection of the causal changes in gene regulation that arise between divergent paralogs facilitating their sub- and neo-functionalization. Studying these changes is critical, as they are the first steps in the creation and modification of gene regulatory networks underlying much of morphological evolution. Thus, with these tools and the continued release of accurate chromosome-scale hybrid assemblies, the future holds great promise for the study of paralogs in evo-devo research.

Notes

From the symposium “Evolution, Physiology, and Biomechanics of Insect Flight” presented at the annual meeting of the Society for Integrative and Comparative Biology, January 2-6, 2024, Seattle, WA, USA.

Contributor Information

Kevin D Deem, Department of Biology, University of Rochester, Rochester, NY, 14620.

Jennifer A Brisson, Department of Biology, University of Rochester, Rochester, NY, 14620.

Funding

This work was supported by the National Institute of General Medical Sciences of the National Institutes of Health [R35GM144001 to J.A.B.]; and the NSF Postdoctoral Research Fellowships in Biology Program [DBI-2305817 to K.D.D.].

References

  1. Asma  H, Halfon  MS. 2019. Computational enhancer prediction: evaluation and improvements. BMC Bioinformatics. 20:174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Asma  H, Tieke  E, Deem  KD, Rahmat  J, Dong  T, Huang  X, Tomoyasu  Y, Halfon  MS. 2024. Regulatory genome annotation of 33 insect species. bioRxiv. 10.1101/2024.01.23.576926. [DOI] [Google Scholar]
  3. Bellen  HJ, O'Kane  CJ, Wilson  C, Grossniklaus  U, Pearson  RK, Gehring  WJ. 1989. P-element-mediated enhancer detection: a versatile method to study development in Drosophila. Genes Dev. 3:1288–300. [DOI] [PubMed] [Google Scholar]
  4. Bessa  J, Carmona  L, Casares  F. 2009. Zinc-finger paralogues tsh and tio are functionally equivalent during imaginal development in Drosophila and maintain their expression levels through auto- and cross-negative feedback loops. Dev Dyn. 238:19–28. [DOI] [PubMed] [Google Scholar]
  5. Biffar  L, Stollewerk  A. 2014. Conservation and evolutionary modifications of neuroblast expression patterns in insects. Dev Biol. 388:103–16. [DOI] [PubMed] [Google Scholar]
  6. Bio-Rad . n.d.; qPCR assay design and optimization. www.bio-rad.com/en-us/applications-technologies/qpcr-assay-design-optimization?ID=LUSO7RIVK, accessed 4-12-2024 [Google Scholar]
  7. Bruce  HS, Jerz  G, Kelly  SR, McCarthy  J, Pomerantz  A, Senevirathne  G, Sherrard  A, Sun  DA, Wolff  C, Patel  NH. 2021. Hybridization chain reaction (HCR) in situ protocol v1. protocols.io. 10.17504/protocols.io.bunznvf6. [DOI]
  8. Bukhari  H, Müller  T. 2019. Endogenous fluorescence tagging by CRISPR. Trends Cell Biol. 29:912–28. [DOI] [PubMed] [Google Scholar]
  9. Chaisson  MJP, Sanders  AD, Zhao  X, Malhotra  A, Porubsky  D, Rausch  T, Gardner  EJ, Rodriguez  OL, Guo  L, Collins  RL  et al.  2019. Multi-platform discovery of haplotype-resolved structural variation in human genomes. Nat Commun. 10:1784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Clark-Hachtel  C, Tomoyasu  Y. 2016. Exploring the origin of insect wings from an evo-devo perspective. Curr Opin Insect Sci. 13:77–85. [DOI] [PubMed] [Google Scholar]
  11. Damen  WGM. 2002. Fushi tarazu : a Hox gene changes its role. Bioessays. 24:992–5. [DOI] [PubMed] [Google Scholar]
  12. Denton  JF, Lugo-Martinez  J, Tucker  AE, Schrider  DR, Warren  WC, Hahn  MW. 2014. Extensive error in the number of genes inferred from draft genome assemblies. PLoS Comput Biol. 10:e1003998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fadaie  Z, Neveling  K, Mantere  T, Derks  R, Haer-Wigman  L, den Ouden  A, Kwint  M, O'Gorman  L, Valkenburg  D, Hoyng  CB  et al.  2021. Long-read technologies identify a hidden inverted duplication in a family with choroideremia. HGG Adv. 2:100046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Falciani  F, Hausdorf  B, Schröder  R, Akam  M, Tautz  D, Denell  R, Brown  S. 1996. Class 3 Hox genes in insects and the origin of zen. Proc Natl Acad Sci USA. 93:8479–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fernandez-Nicolas  A, Ventos-Alfonso  A, Kamsoi  O, Clark-Hachtel  C, Tomoyasu  Y, Belles  X. 2022. Broad complex and wing development in cockroaches. Insect Biochem Mol Biol. 147:103798. [DOI] [PubMed] [Google Scholar]
  16. Fisher  CR, Kratovil  JD, Angelini  DR, Jockusch  EL. 2021. Out from under the wing: reconceptualizing the insect wing gene regulatory network as a versatile, general module for body-wall lobes in arthropods. Proc Biol Sci. 288:20211808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Force  A, Lynch  M, Pickett  FB, Amores  A, Yan  Y, Postlethwait  J. 1999. Preservation of duplicate genes by complementary, degenerative mutations. Genetics. 151:1531–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gehring  WJ, Kloter  U, Suga  H. 2009. Evolution of the Hox gene complex from an evolutionary ground state. Curr Top Dev Biol. 88:35–57. [DOI] [PubMed] [Google Scholar]
  19. GenScript . n.d.; Antigen strategy for successful antibody production. www.genscript.com/protein-antigen-vs-peptide-antigen.html, accessed on December 4, 2024 [Google Scholar]
  20. Gout  J-F, Lynch  M. 2015. Maintenance and loss of duplicated genes by dosage subfunctionalization. Mol Biol Evol. 32:2141–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Higo  S, Ishii  H, Ozawa  H. 2023. Recent advances in high-sensitivity in situ hybridization and costs and benefits to consider when employing these methods. ACTA Histochem Cytochem. 56:49–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hon  T, Mars  K, Young  G, Tsai  Y-C, Karalius  JW, Landolin  JM, Maurer  N, Kudrna  D, Hardigan  MA, Steiner  CC  et al.  2020. Highly accurate long-read HiFi sequencing data for five complex genomes. Sci Data. 7:399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Iimura  T, Denans  N, Pourquié  O. 2009. Establishment of Hox vertebral identities in the embryonic spine precursors. Curr Top Dev Biol. 88:201–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jaworski  CC, Allan  CW, Matzkin  LM. 2020. Chromosome-level hybrid de novo genome assemblies as an attainable option for nonmodel insects. Mol Ecol Resour. 20:1277–93. [DOI] [PubMed] [Google Scholar]
  25. Kaufman  TC, Lewis  R, Wakimoto  B. 1980. Cytogenetic analysis of chromosome 3 in Drosophila melanogaster: the homeotic gene complex in polytene chromosome interval 84A-B. Genetics. 94:115–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Koonin  EV. 2005. Orthologs, paralogs, and evolutionary genomics. Annu Rev Genet. 39:309–38. [DOI] [PubMed] [Google Scholar]
  27. Kuzmin  E, Taylor  JS, Boone  C. 2022. Retention of duplicated genes in evolution. Trends Genet. 38:59–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kvon  EZ. 2015. Using transgenic reporter assays to functionally characterize enhancers in animals. Genomics. 106:185–92. [DOI] [PubMed] [Google Scholar]
  29. Lai  Y-T, Deem  KD, Borràs-Castells  F, Sambrani  N, Rudolf  H, Suryamohan  K, El-Sherif  E, Halfon  MS, McKay  DJ, Tomoyasu  Y. 2018. Enhancer identification and activity evaluation in the red flour beetle, Tribolium castaneum. Development. 145:dev160663. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Laugier  E, Yang  Z, Fasano  L, Kerridge  S, Vola  C. 2005. A critical role of teashirt for patterning the ventral epidermis is masked by ectopic expression of tiptop, a paralog of teashirt in Drosophila. Dev Biol. 283:446–58. [DOI] [PubMed] [Google Scholar]
  31. Lee  B-S, Huang  J-S, Jayathilaka  LP, Lee  J, Gupta  S. 2016. Antibody production with synthetic peptides. In: S  Schwartzbach, O  Skalli, T  Schikorski, editors. High-resolution imaging of cellular proteins: methods and protocols. methods in molecular biology. New York (NY): Springer Science+Business Media. p.25–48. [DOI] [PubMed] [Google Scholar]
  32. Lewis  EB. 1978. A gene complex controlling segmentation in Drosophila. Nature. 276:565–70. [DOI] [PubMed] [Google Scholar]
  33. Li  B, Bickel  RD, Parker  BJ, Saleh Ziabari  O, Liu  F, Vellichirammal  NN, Simon  J-C, Stern  DL, Brisson  JA. 2020. A large genomic insertion containing a duplicated follistatin gene is linked to the pea aphid male wing dimorphism. eLife. 9:e50608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Li  H, Popadić  A. 2004. Analysis of nubbin expression patterns in insects. Evol Dev. 6:310–24. [DOI] [PubMed] [Google Scholar]
  35. Li  L, Zhu  Q, He  X, Sinha  S, Halfon  MS. 2007. Large-scale analysis of transcriptional cis-regulatory modules reveals both common features and distinct subclasses. Genome Biol. 8:R101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Li-Kroeger  D, Kanca  O, Lee  P-T, Cowan  S, Lee  MT, Jaiswal  M, Salazar  JL, He  Y, Zuo  Z, Bellen  HJ. 2018. An expanded toolkit for gene tagging based on MiMIC and scarless CRISPR tagging in Drosophila. eLife. 7:e38709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Logsdon  GA, Vollger  MR, Eichler  EE. 2020. Long-read human genome sequencing and its applications. Nat Rev Genet. 21:597–614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Loker  R, Mann  RS. 2022. Divergent expression of paralogous genes by modification of shared enhancer activity through a promoter–proximal silencer. Curr Biol. 32:3545–55. e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mastretta-Yanes  A, Zamudio  S, Jorgensen  TH, Arrigo  N, Alvarez  N, Piñero  D, Emerson  BC. 2014. Gene duplication, population genomics, and species-level differentiation within a tropical mountain shrub. Genome Biol Evol. 6:2611–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Medina-Jiménez  BI, Budd  GE, Janssen  R. 2021. Panarthropod tiptop/teashirt and spalt orthologs and their potential role as “trunk”-selector genes. EvoDevo. 12:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Mulhair  PO, Holland  PWH. 2024. Evolution of the insect Hox gene cluster: comparative analysis across 243 species. Semin Cell Dev Biol. 152–153:4–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Murugesan  SN, Connahs  H, Matsuoka  Y, Gupta  MD, Tiong  GJL, Huq  M, Gowri  V, Monroe  S, Deem  KD, Werner  T  et al.  2022. Butterfly eyespots evolved via cooption of an ancestral gene-regulatory network that also patterns antennae, legs, and wings. Proc Natl Acad Sci USA. 119:e2108661119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Nowak  RM, Jastrzębski  JP, Kuśmirek  W, Sałamatin  R, Rydzanicz  M, Sobczyk-Kopcioł  A, Sulima-Celińska  A, Paukszto  Ł, Makowczenko  KG, Płoski  R  et al.  2019. Hybrid de novo whole-genome assembly and annotation of the model tapeworm Hymenolepis diminuta. Sci Data. 6:302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Ohno  S. 1970. Evolution by gene duplication. New York (NY): Springer-Verlag. [Google Scholar]
  45. Oikonomopoulos  S, Bayega  A, Fahiminiya  S, Djambazian  H, Berube  P, Ragoussis  J. 2020. Methodologies for transcript profiling using long-read technologies. Front Genet. 11:606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. O’Kane  CJ, Gehring  WJ. 1987. Detection in situ of genomic regulatory elements in Drosophila. Proc Natl Acad Sci USA. 84:9123–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Pace  RM, Grbić  M, Nagy  LM. 2016. Composition and genomic organization of arthropod Hox clusters. EvoDevo. 7:11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Rezazade Bazaz  M, Dehghani  H. 2022. From DNA break repair pathways to CRISPR/Cas-mediated gene knock-in methods. Life Sci. 295:120409. [DOI] [PubMed] [Google Scholar]
  49. Schember  I, Reid  W, Halfon  MS. 2023. Conserved and novel enhancers regulate expression of Aedes aegypti single-minded in the embryonic ventral midline. bioRxiv. 10.1101/2023.08.01.551414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Schwartzbach  SD, Skalli  O, Schikorski  T, editors. 2016. High-resolution imaging of cellular proteins: methods and protocols: methods and protocols. New York (NY): Springer New York. [Google Scholar]
  51. Shippy  TD, Tomoyasu  Y, Nie  W, Brown  SJ, Denell  RE. 2008. Do teashirt family genes specify trunk identity? Insights from the single tiptop/teashirt homolog of Tribolium castaneum. Dev Genes Evol. 218:141–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Shu  X, Yuan  R, Zheng  B, Wang  Z, Ye  X, Tang  P, Chen  X. 2023. Chromosome-level genome assembly of Microplitis manilae Ashmead, 1904 (Hymenoptera: Braconidae). Sci Data. 10:266. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Singh  NP, Krumlauf  R. 2022. Diversification and functional evolution of HOX proteins. Front Cell Dev Biol. 10:798812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Singh  S, Rahangdale  S, Pandita  S, Saxena  G, Upadhyay  SK, Mishra  G, Verma  PC. 2022. CRISPR/Cas9 for insect pests management: a comprehensive review of advances and applications. Agriculture. 12:1896. [Google Scholar]
  55. Tang  M, He  S, Gong  X, Lü  P, Taha  RH, Chen  K. 2021. High-quality de novo chromosome-level genome assembly of a single Bombyx mori with BmNPV resistance by a combination of PacBio long-read sequencing, Illumina short-read sequencing, and Hi-C sequencing. Front Genet. 12:718266. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Tidswell  ORA, Benton  MA, Akam  M. 2021. The neuroblast timer gene nubbin exhibits functional redundancy with gap genes to regulate segment identity in Tribolium. Development. 148:dev199719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Tomoyasu  Y. 2021. What crustaceans can tell us about the evolution of insect wings and other morphologically novel structures. Curr Opin Genet Dev. 69:48–55. [DOI] [PubMed] [Google Scholar]
  58. Tomoyasu  Y, Arakane  Y, Kramer  KJ, Denell  RE. 2009. Repeated co-options of exoskeleton fomation during wing-to-elytron evolution in beetles. Curr Biol. 19:2057–65. [DOI] [PubMed] [Google Scholar]
  59. Tomoyasu  Y, Halfon  MS. 2020. How to study enhancers in non-traditional insect models. J Exp Biol. 223:jeb212241. [DOI] [PubMed] [Google Scholar]
  60. Trauner  J, Schinko  J, Lorenzen  MD, Shippy  TD, Wimmer  EA, Beeman  RW, Klingler  M, Bucher  G, Brown  SJ. 2009. Large-scale insertional mutagenesis of a coleopteran stored grain pest, the red flour beetle Tribolium castaneum, identifies embryonic lethal mutations and enhancer traps. BMC Biol. 7:73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Vollger  MR, Dishuck  PC, Sorensen  M, Welch  AE, Dang  V, Dougherty  ML, Graves-Lindsay  TA, Wilson  RK, Chaisson  MJP, Eichler  EE. 2019. Long-read sequence and assembly of segmental duplications. Nat Methods. 16:88–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Wang  Z, Ma  X, Zhu  J, Zheng  B, Yuan  R, Lu  Z, Shu  X, Fang  Y, Tian  S, Qu  Q  et al.  2023. Chromosome-level genome assembly of Chouioia cunea Yang, the parasitic wasp of the fall webworm. Sci Data. 10:485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Wiegmann  BM, Trautwein  MD, Winkler  IS, Barr  NB, Kim  J-W, Lambkin  C, Bertone  MA, Cassel  BK, Bayless  KM, Heimberg  AM  et al.  2011. Episodic radiations in the fly tree of life. Proc Natl Acad Sci USA. 108:5690–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Zhang  R, Zhang  Z, Huang  Y, Qian  A, Tan  A. 2020. A single ortholog of teashirt and tiptop regulates larval pigmentation and adult appendage patterning in Bombyx mori. Insect Biochem Mol Biol. 121:103369. [DOI] [PubMed] [Google Scholar]
  65. Zhang  Z, An  HH, Vege  S, Hu  T, Zhang  S, Mosbruger  T, Jayaraman  P, Monos  D, Westhoff  CM, Chou  ST. 2022. Accurate long-read sequencing allows assembly of the duplicated RHD and RHCE genes harboring variants relevant to blood transfusion. Am J Hum Genet. 109:180–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Zhu  KY, Palli  SR. 2020. Mechanisms, applications, and challenges of insect RNA interference. Annu Rev Entomol. 65:293–311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Zirin  JD, Mann  RS. 2007. Nubbin and teashirt mark barriers to clonal growth along the proximal–distal axis of the Drosophila wing. Dev Biol. 304:745–58. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Integrative and Comparative Biology are provided here courtesy of Oxford University Press

RESOURCES