Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Screens in fly and beetle reveal vastly divergent gene sets required for developmental processes

Screens in fly and beetle reveal vastly divergent gene sets required for developmental processes Background: Most of the known genes required for developmental processes have been identified by genetic screens in a few well-studied model organisms, which have been considered representative of related species, and informative—to some degree—for human biology. The fruit fly Drosophila melanogaster is a prime model for insect genetics, and while conservation of many gene functions has been observed among bilaterian animals, a plethora of data show evolutionary divergence of gene function among more closely-related groups, such as within the insects. A quantification of conservation versus divergence of gene functions has been missing, without which it is unclear how representative data from model systems actually are. Results: Here, we systematically compare the gene sets required for a number of homologous but divergent developmental processes between fly and beetle in order to quantify the difference of the gene sets. To that end, we expanded our RNAi screen in the red flour beetle Tribolium castaneum to cover more than half of the protein- coding genes. Then we compared the gene sets required for four different developmental processes between beetle and fly. We found that around 50% of the gene functions were identified in the screens of both species while for the rest, phenotypes were revealed only in fly (~ 10%) or beetle (~ 40%) reflecting both technical and biological differences. Accordingly, we were able to annotate novel developmental GO terms for 96 genes studied in this work. With this work, we publish the final dataset for the pupal injection screen of the iBeetle screen reaching a coverage of 87% (13,020 genes). Conclusions: We conclude that the gene sets required for a homologous process diverge more than widely believed. Hence, the insights gained in flies may be less representative for insects or protostomes than previously thought, and work in complementary model systems is required to gain a comprehensive picture. The RNAi screening resources developed in this project, the expanding transgenic toolkit, and our large-scale functional data make T. castaneum an excellent model system in that endeavor. Keywords: Gene function, Comparative genomics, RNAi screen, iBeetle, Tribolium castaneum, Drosophila melanogaster, Divergence of gene function, iBeetle-Base, FlyBase * Correspondence: Gbucher1@uni-goettingen.de Muhammad Salim Hakeemi, Salim Ansari, Matthias Teuscher and Matthias Weißkopf contributed equally to this work. Johann-Friedrich-Blumenbach-Institut, GZMB, University of Göttingen, Justus-von-Liebig-Weg 11, 37077 Göttingen, Germany Full list of author information is available at the end of the article © The Author(s). 2022 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Hakeemi et al. BMC Biology (2022) 20:38 Page 2 of 13 Background and head specification show the involvement of similar The function of genes is paramount for the biology of gene sets, although a few components appear to be re- any organism and, hence, the assignment of functions to quired for only some clades [12–15]. genes is a central question of biological research. How- Notably, the differences in gene functions documented ever, only in a very small number of genetic model spe- so far may be an underestimation of the real divergence, cies like the mouse Mus musculus, the zebrafish Danio because the prevailing candidate gene approach leads to rerio, the nematode Caenorhabditis elegans, and the vin- a systematic bias towards conservation. The genes to be egar fly Drosophila melanogaster have the functions of tested are usually chosen based on the knowledge of developmental genes been assayed in systematic screens. their ortholog’s involvement in other species. As a con- This restriction to a few model systems is a consequence sequence, unrelated genes are rarely tested and the in- of the necessity for an elaborate genetic and molecular volvement of unexpected genes in a given process is tool kit, which is extremely laborious to establish [1–4]. underestimated. Hence, approaches are needed to over- Unfortunately, it has remained unclear how representa- come this bias and to gain a realistic view on the degree tive findings in these model species actually are for their of gene function divergence. To that end, genes required clade or in other words, how quickly and profoundly for certain biological processes need to be identified in gene function diverges in evolution. Indeed, based on an an unbiased and genome-wide manner in non- astonishing degree of functional conservation with re- traditional organisms as well, even though this has spect to many genes, it has been assumed that the find- remained technically challenging and such large datasets ings gained in those few model species can be outside the classic model systems had not been available. transferred to a large degree. This assumption underlies The red flour beetle T. castaneum has recently been the use of the insect D. melanogaster as a model for a established as the only arthropod model organism apart number of human diseases, including developmental as- from D. melanogaster where genome-wide unbiased pects like heart formation. Conversely, even among in- RNAi screens are feasible. Based on the robust and sys- sects, dramatic evolutionary changes of gene functions temic RNAi response of this species, the iBeetle large- have been described with respect to homologous devel- scale screen was performed where random genes were opmental processes, calling into question the degree of knocked down and the resulting animals were scored for transferability. However, a systematic comparison of the a number of developmental phenotypes [16–18]. How- gene sets required for the same process in different spe- ever, the data gained in the iBeetle screen had covered cies has not been done and, hence, a quantification of di- only one third of the gene set, not allowing for robust vergence versus conservation remains badly missing. genome wide statements. Apart from its particularly Importantly, knowing the degrees of gene function di- strong and robust RNAi response, T. castaneum offers a vergence is relevant not only for understanding the evo- comparably large tool kit for analyzing gene function in- lution of biodiversity but also for applied research, e.g., cluding transgenic and genome editing approaches [19– for transferring knowledge from model systems to spe- 21]. cies relevant for medical applications or pest control. In this paper, we used an expanded dataset to assess Recently, the study of gene function in development the degree of divergence of the gene sets required for se- has been extended to non-traditional model organisms. lected developmental processes between fly and beetle Predominantly, candidate genes known for their function such as head, muscle and ovary development, and in the classical model systems have been tested in other dorso-ventral patterning. These processes are homolo- organisms. Subsequent comparisons revealed both, con- gous but show a different degree of evolutionary diver- servation and divergence of gene functions. For example, gence, which could be reflected in changes of gene axis formation in D. melanogaster has turned out to be a function. First, we determined genes that were essential rather diverged process partially based on different genes in the beetle for these processes but which had so far compared to other insects. The key anterior morphogen not been connected to them in D. melanogaster. These a of D. melanogaster, bicoid, is not present in most insects priori unexpected genes sum up to about 37% of the [5]. Instead, repression of Wnt signaling plays a central total genes identified to be required for either one or role in the red flour beetle Tribolium castaneum [6]asit both species. For 30% of these genes, no functional an- does in many animals including other insects, flatworms, notation had been available at FlyBase at all such that and vertebrates [7–10] - but not in D. melanogaster. The we provide the first functional Gene Ontology (GO) as- functions of genes of the Hox cluster, in contrast, appear signment for the respective ortholog group in insects. conserved over very large phylogenetic distances—al- Only two genes essential in T. castaneum did not have though some functional divergence has been linked to an ortholog in D. melanogaster, i.e., these processes the evolution of arthropod morphology [11]. Likewise, seem not much affected by gene gain or loss. We con- the gene regulatory network of dorso-ventral patterning clude that restricting genetic screens to one model Hakeemi et al. BMC Biology (2022) 20:38 Page 3 of 13 system only falls short of identifying a comprehensive set of essential genes. Further, our data reveals an unex- pected degree of divergence of gene function between two holometabolous insect species. Moreover, we present here an update of the dataset gained in the gen- ome wide iBeetle screen in T. castaneum. Our analysis is based on both, a dataset previously published comprising 5300 genes [17] and an additional 3200 genes screened as part of this project. In addition to those, we publish and make accessible (at iBeetle-Base) the phenotypes for an additional 4520 genes which were screened while the analysis presented here was ongoing. Hence, with this paper, the coverage of genes tested and annotated at iBeetle-Base sums to 13,020 Tribolium genes (78% of the predicted gene set), which will be the final number for the pupal injection screen of the iBeetle screen. Results Fig. 1 Quality controls of the primary screen. 178 positive controls Reaching the final dataset for pupal injections of the using 35 different genes were included. More than 88% of the large scale iBeetle screen positive controls were fully or partially recognized (left bar) while We added 3200 genes to the previously published 5300 7.3% were missed. 4% could not be analyzed due to technical genes of our large-scale iBeetle screen [17], reaching a lethality before the production of offspring. 7.1% of the negative coverage of 51% of the T. castaneum gene set of a total controls had annotations, i.e. they were false positive (right bar). These figures are similar to the first screening phase [17] of 16,593 currently annotated genes [22]. We followed the previously described procedure for the pupal injec- tion screen [17] with minor modifications (see the “Methods” section). In short, we injected 10 female analysis presented here because both analyses ran in pupae per gene with dsRNAs (concentration 1 μg/μl). parallel. We annotated the phenotypes of the injected animals and the first instar cuticle of their offspring using the Unexpected gene functions in developmental processes EQM system [23], the T. castaneum morphological We wanted to use our large-scale phenotypic dataset to ontology Tron [24], and a controlled vocabulary. The systematically compare the gene sets required for the data is available at the online database iBeetle-Base [25– same biological processes in T. castaneum and D. mela- 27]. As positive controls, an array of genes with known nogaster. We define a gene to be “required for” a phenotypes related to the processes under scrutiny was process, if its knock-down or mutation leads to a pheno- added to the screening and included both drastic and type in that process. To that end, we first identified in mild phenotypes (find details in the extensive descrip- an unbiased way all genes annotated with a phenotype tion of the first part of the screen in Schmitt-Engel et al. for a number of biological processes by searching iBee- [17]). Buffer injections were performed as negative con- tle-Base. Specifically, we scored for phenotypes indicative trols. These controls revealed a similar portion of false of functions in dorso-ventral patterning, head and negative and false positive data compared to the first muscle development, and in oogenesis. part of the screen (Fig. 1 and Additional file 1: Table The choice of the processes to be compared was based S1). The analysis presented in this work is based on all on two arguments. On one hand, we chose processes genes that had been screened in the first round of the that fell within our core expertise for sake of correct in- screen and the set of genes published with this publica- terpretation of phenotypes, for the opportunity for sub- tion. Taken together, the analyzed set of genes covered sequent detailed work, and to ensure comprehensive approximately 50% of the genome. About two thirds of knowledge of respective Drosophila data. On the other the Drosophila genes relevant for our analysis had been hand, we aimed at including developmental processes covered by the screen (Additional file 2: Fig. S1). While with different degrees of conservation. We assumed that this analysis was ongoing, we continued the screen as cellular processes such as muscle development might be well and have in the meanwhile reached a coverage of conserved. Indeed, for both, conservation of basic mech- 78% (13,020 genes). We publish these additional pheno- anisms between Drosophila and vertebrates was shown typic data (accessible online at iBeetle-Base) with this along with evolutionary differences [28–31], and the use article, but they were not included in the detailed of Drosophila as a model for vertebrate heart Hakeemi et al. BMC Biology (2022) 20:38 Page 4 of 13 development is widely adopted [32, 33]. Some aspects of searching T. castaneum, D. melanogaster, and M. muscu- dorso-ventral patterning are highly conserved among an- lus genomes for orthologs and paralogs. This analysis re- imals (e.g., the involvement of dpp/BMP versus sog) but vealed that only three genes with a novel function (appr. differences even between insects have been described as 3%) did not have a D. melanogaster ortholog (yellow in well [14, 34]. Likewise, a set of highly conserved genes is Fig. 2; see Additional file 2: Fig. S2 and S3 for phylogen- involved in anterior development in all animal clades etic trees). Evidently, lineage-specific gene loss or gain while clear differences were described even between flies explains only a minor part of the functional divergence and beetles [12, 35–39]. This divergence may to some of homologous developmental processes. degree be due to the fly-specific involution and reduc- Next, we asked whether the respective D. melanogaster tion of the larval head. Finally, telotrophic oogenesis of orthologs were known to be involved in other biological Tribolium is morphologically quite distinct from the processes or lacked any phenotype information. For this polytrophic oogenesis found in Drosophila [40–42] and analysis, we did not only use the genes identified in the both are quite different from the vertebrate process screen but included those that had previously been stud- probably representing a more divergent process. ied, as well. We looked up phenotype information of the For all these processes, we found a number of gene respective D. melanogaster orthologs on FlyBase (hom- functions that were expected based on D. melanogaster ology assignment done with OrthoDB v9). Among the knowledge (see below). This confirmed that the screen fly orthologs whose functional annotations did not design allowed the detection of these types of pheno- match with those from the iBeetle screen or published types. Importantly, we also found functions for genes so records of functional data, around two thirds (64.6%) far not connected to those processes (based on FlyBase had annotations that were related to other processes information [43, 44], PubMed searches, and scientist ex- than the ones studied in T. castaneum (Fig. 2). Import- pertise). The iBeetle screen is a first pass screen with a antly, one third of the genes (32.3%) did not have any focus on minimizing false negative results with the functional annotation in FlyBase. Hence, for those genes, trade-off of allowing for false positive annotations [17]. the iBeetle-screen had detected the first documented The likelihood for this type of error is further increased function of that ortholog group in insects. Importantly, by off-target effects and/or by strain-specific differences due to the lack of previous phenotypic information, in the phenotype, i.e. genes for which the described these genes likely would not have been included in a phenotype was not reproduced in another genetic back- classical candidate gene approach. ground, which we counted as “false positive” in order to be conservative [45]. Hence, we aimed at excluding false A quarter of Drosophila gene function annotations were positive annotations for the unexpected gene functions. not confirmed for T. castaneum First, we based our analyses only on genes for which In a complementary approach, we asked how many genes phenotypes had been annotated with a penetrance of > known to be required for a given process in D. melanoga- 50% in the primary screen. Further, we only used pheno- ster had been assigned related functions in the iBeetle types that were reproduced by RNAi experiments with screen. To that end, we first collected lists of genes re- non-overlapping dsRNA fragments targeting the same quired for those processes based on D. melanogaster gene. In order to exclude genetic background effects, we knowledge (expert knowledge, literature, and FlyBase) used another lab strain (our standard lab strain San Ber- (Additional file 4: Table S3). Then we mined iBeetle-Base nardino, SB) except for the muscle project where we to see how many of the beetle orthologs had an annota- needed to use the pBA19 strain, which has EGFP tion related to that process (Fig. 3A). About two-thirds of marked muscles [46]. This re-screening procedure re- those genes had actually been screened in T. castaneum sulted in a set of genes for which we can claim with high (Additional file 2: Fig. S1) and all following numbers are confidence that they are indeed required for these pro- based on the analysis of this subset. cesses in T. castaneum, but which previously were not A surprisingly large portion of genes (26.4%) known assigned to these in D. melanogaster (Additional file 3: to be required for these processes in D. melanogaster Table S2). did not show the expected phenotype in T. casta- neum (Fig. 3B). Assigning the first function to a gene versus extending previous annotations Enriching the GO information with data from Tribolium One reason for a lack of respective functional data in Gene ontology (GO) assignment is a powerful tool to es- FlyBase could be that the knocked-down beetle gene tablish hypotheses on the function of given gene sets does not have an ortholog in the fly. In order to test this, [47]. So far, there were no GO terms associated based we searched for the fly orthologs in orthoDB and by on T. castaneum data. The work presented here revealed manually generating phylogenetic trees based on that a surprisingly high portion of orthologous genes has Hakeemi et al. BMC Biology (2022) 20:38 Page 5 of 13 Fig. 2 Analysis of genes with unexpected gene functions found in Tribolium. A Numbers of genes, for which an unexpected function in the respective process was found in the iBeetle screen but had not been known from Drosophila. B Combined numbers for all four processes. Only three genes with novel gene functions in Tribolium had no ortholog in Drosophila (yellow). About two-thirds of genes with novel function had previous phenotypic annotations in FlyBase but relating to other biological processes (blue). Importantly, for one third of those genes, we had detected the first phenotype in any insect (green). We added this novel information to the GO database diverging functions in different organisms. To enrich the based on work in flies. However, there are several rea- GO database, we submitted GO terms with respect to sons to believe that the picture has remained incom- the biological process for all 96 re-screened genes with plete. On one hand, species-specific or technical functions in dorso-ventral patterning (GO:0010084), oo- limitations may have prohibited the identification of an genesis (GO:0048477), the development of embryonic involved gene in D. melanogaster. On the other hand, muscles (GO:0060538), and head (GO:0048568) [48]. evolution has led to functional changes such as the modification or loss of ancestral gene functions or the Discussion co-option of genes into a novel process. Unfortunately, Investigating one species falls short of a comprehensive it has remained unclear to what extent the gene sets de- view on gene function termined exclusively in flies would be representative of Large-scale screens in the leading insect model organism insects as a whole or if it is even appropriate to assume D. melanogaster have revealed gene sets required for the existence of representative gene sets. certain biological processes. As consequence, insect- Our systematic screening in a complementary model related GO term annotations are almost exclusively organism has revealed that the identified gene sets show Fig. 3 Beetle genes showing phenotypes expected from Drosophila. A Gene sets known to be required for a given process in Drosophila were compared to iBeetle data. Close to three quarters showed related phenotypes (blue) while others had no or different types of phenotypes (green). B Approximately one quarter of the genes known to be required for certain Drosophila processes were not required for that process in Tribolium. This analysis is based on the subset of genes which already had been screened in Tribolium (51%). Interestingly, we found orthologs for 66% of the respective Drosophila genes—this indicates that our screen was enriched for relevant genes Hakeemi et al. BMC Biology (2022) 20:38 Page 6 of 13 an unexpected degree of divergence (see Fig. 4 for num- one species (11% vs. 37%) may reflect both, biological bers, Fig. 5 for examples). Based on our calculations (see and technical differences (see detailed discussion below). details below) we estimate that only half of the gene Beyond the fly-beetle comparison, our findings provide functions are detected in both species (52%, column 4 of a compelling argument that focusing on single model Fig. 4A) while the remaining gene functions were found species falls short of comprehensively revealing the gen- either only in D. melanogaster (11%, column 4 of Fig. etic basis of biological processes in any clade separated 4A) or only in T. castaneum (37%, column 4 of Fig. 4A). by an evolutionary distance similar or larger than the We found no strong indication that the gene inventory one separating flies and beetles (i.e., around 370 million required for a process would be more conserved for years). Further, it shows that T. castaneum is an ex- those processes, which seem more conserved morpho- tremely useful screening system for insect biology, able logically. For instance, dorso-ventral patterning, which to reveal novel gene functions even in processes that we assumed representing an intermediate degree of con- have been studied intensely in D. melanogaster. servation showed the largest common gene set while the supposedly most conserved process, muscle formation, Estimating the portions of gene functions revealed in fly showed the lowest value. However, we note that we versus beetle found more Tribolium-specific gene functions for the In order to make a comprehensive and quantitative less conserved processes than for muscle development. comparison, we included in our comparisons all genes However, given the uncertainties with these numbers that are currently known to be involved in the respective (see the “Discussion” section below) and the fact that processes from both, beetle and fly. Our beetle data are morphological conservation of a process is hard to quan- based on both, our systematic screening of 51% of the T. tify, we hesitate to draw conclusions about the correl- castaneum gene set and on previous candidate gene ation of divergence of a biological process and the work. With respect to fly data, we rely on information involved gene inventory. available on FlyBase and our expert knowledge of the While these data were gained with respect to develop- processes under scrutiny. Given these different kinds of mental processes only, they strongly indicate that our sources and approaches, and the fact that we focus on current knowledge based on screening in one species ap- developmental processes, the data are – despite the pears to be much less comprehensive than previously comprehensive approach - prone to various types of un- thought. We believe that the different proportions of certainties. In the following, we first discuss the way we genes shown to be required for a specific process in only combined the numbers to calculate our estimation. Fig. 4 Many genes involved in a given process are detected only in one of the two species. We combined all genes found in the fly to be involved in our processes (column 1) and/or those genes that we identified in the iBeetle screen to be required in the same process (column 3) to assemble a set of genes comprising all genes currently known to be required in any insect for the processes analyzed here (column 4). Of the fly gene set (column 1) about two thirds had been tested in the iBeetle screen. Of those, three quarters showed a similar function in our beetle while one quarter appeared to be fly specific (column 2). The subdivisions of columns 1 and 2 are based on Figs. 2 and 3 and Additional file 2: Fig. S1. From the numbers in columns 1-3 we calculated the portions of genes of the combined insect gene set (column 4), which were detected only in Drosophila (11%), only in Tribolium (37%), or in both (52%). See text for details and discussion of potential systematic biases. B) Respective values for the single processes show that the Tribolium screening platform revealed 20–50% novel genes relevant for a process (i.e., which were not detected in Drosophila). See Additional file 5: Table S4 for calculations. Given these results neither model system can be used alone as a proxy for insects or protostomes in general and that Tribolium is a very useful complementary screening platform Hakeemi et al. BMC Biology (2022) 20:38 Page 7 of 13 Fig. 5 Examples for novel gene functions detected in Tribolium but not known from Drosophila. A Anterior part of a wildtype cuticle with head, thorax, and anterior abdomen. B In iB_03355 knock-down embryos, dorso-ventral patterning was disturbed such that the embryo has turned inside out, i.e. the legs and the head are located inside the trunk cuticle instead of outside. C In iB_04199 knock-down, anterior epidermal patterning was disrupted to different degrees. In mild phenotypes, just the most anterior part, the labrum, was affected (not shown), intermediate phenotypes lacked head and parts of the thorax (shown) and in strong phenotypes, only cuticle crumbs remained. D In the transgenic strain pBA19, the muscles are marked with EGFP. They are visible in vivo as elongated structures with a segmentally repeated pattern. E In iB_01159 (Tc-Unc-76 ) knock-down, the muscles were partially missing or detached such that some muscles adopted a rounded shape. F In wildtype ovaries, the nurse cells are located in the tropharium forming an elongated structure (marked by a white line). The first part of the vitellarium is marked by active cell division (marked here by phospho-histone 3 staining, PH3) and along the entire vitellarium, the oocytes increase in size (compare stars). G In iB_10431 knock-down ovaries, the tropharia were normal (white line) but no oocytes developed. The white structure is not part of the ovaries. Anterior is to the top in A–E, the pictures of the phenotypes are modified from iBeetle-Base. Scale bars are 100 μm Hakeemi et al. BMC Biology (2022) 20:38 Page 8 of 13 Subsequently, we will discuss some uncertainties and in specific genes without conserved functions likely is even how far they influence the estimation. higher than reflected in Fig. 4A. Of the genes known from D. melanogaster to be re- Second, we found quite different numbers for the four quired for the processes investigated here (n = 132; see processes under scrutiny (Fig. 4B). However, even in the Additional file 5: Table S4), we could compare 66% to process with the lowest portion of genes detected exclu- iBeetle data (column 1 in Fig. 4A; based on Additional sively in T. castaneum (muscle development), this por- file 2: Fig. S1; n = 87). Of those genes, 26% (n = 23) were tion was 21%, which still indicates a significant degree of not required for that process in T. castaneum (column 2 unexplored biology. in Fig. 4A; based on Fig. 3). With this statement, we Third, the D. melanogaster numbers could be influ- mean that the respective genes did not have any pheno- enced by false negative data. The data on FlyBase has type with respect to the biological process in question. not been gathered in one or few standardized screens They could have either no phenotype or a phenotype af- where all data were published—it is mainly based on fecting another process. Based on our positive controls, published results of single gene analyses. However, not the potential error affecting this statement is less than all genetic screens have reached saturation and not all 7.5% (see Additional file 1: Table S1). For our overall es- genes detected in large-scale screens may have been fur- timation, we extrapolated this share to the total number ther analyzed and published. Hence, the number of of genes required for the fly (dotted lines from column 2 genes in principle detectable in D. melanogaster might to column 4). A number of gene functions detected in actually be larger than the numbers extracted from Fly- the iBeetle screen had not been assigned such functions Base. In the iBeetle screen, in contrast, negative data was in D. melanogaster before (column 3 in Fig. 4A; based systematically documented, such that this type of uncer- on Fig. 2). When combining these numbers, we aimed at tainty is restricted to technical false negative data, which providing a minimum estimation for the divergence of we found to be around 15% in this first pass screen (Fig. detected gene functions (Column 4 in Fig. 4A). To be 1). This uncertainty could potentially increase the por- conservative, we assumed that all gene functions known tion of D. melanogaster-specific or conserved genes. from D. melanogaster but not yet tested in the iBeetle Fourth, theoretically, there may be false positive data al- screen would fall into the class of genes being required beit restricted to the set of genes detected in both spe- for both species (see numbers in green square in Add- cies. The reason is that iBeetle was a first pass screen, itional file 5: Table S4). Further, we scored each signal- where we aimed at reducing false negative data with the ing pathway as one case (finding mostly conservation) tradeoff that false positive data are enriched [17]. Al- even if single components of these pathways did not though finding similar phenotypes in two different spe- have divergent phenotypes. This conservative assump- cies will not in many cases be false positive, we tried to tion leads to the abovementioned minimum estimation minimize this error by manually checking the annota- of divergence in these gene sets (Column 4 in Fig. 4A; tions of the respective genes, excluding those that calculation in Additional file 5: Table S4). Of all genes showed a phenotype with low penetrance or in combin- currently known to be required for one of the processes ation with many other defects indicating a non-specific we studied, the portion of genes detected exclusively in effect. Of note, the issue of false positives is restricted to the fly (11%; n = 23) is much smaller than the one de- the genes detected in both species (column 2; based on tected only in the beetle (37%; n = 76) while the analo- Fig. 3). It does not apply to those genes detected only in gous function of half of the genes (52%; n =109)is the beetle but not the fly (column 3; based on Fig. 2) be- detected in both species. cause in this case, all phenotypes were confirmed by in- With this work, we present the first and a quite exten- dependent experiments with non-overlapping dsRNA sive dataset to estimate this kind of numbers. Still, some fragments in different genetic backgrounds such that confounding issues need to be considered. The first un- false positive results are excluded. In summary, while certainty stems from the fact that the beetle data is there are a number of uncertainties that we could not based on testing about 50% of the genes. In the second clarify with available data or methods, most of these un- part of the screen, we had prioritized genes that were certainties hint at underestimation rather than overesti- moderately to highly expressed, showed sequence con- mation of functional divergence between fly and beetle. servation, and had GO annotations. The prioritization Our work focused on developmental processes with apparently was successful as 66% of the gene functions different grades of assumed conservation and different known from D. melanogaster had been covered in the grades of previous knowledge. Morphologically, the iBeetle screen (Fig. 4A), which is much more than the muscle pattern and general development appear to be a 40% expected for an unbiased selection [17]. Hence, our quite conserved between these insects [31, 49] compared figures might be biased towards conserved gene func- to oogenesis where a number of morphological differ- tion. As a consequence, the overall portion of beetle- ences were described [40–42]. Given the background of Hakeemi et al. BMC Biology (2022) 20:38 Page 9 of 13 a strongly derived head morphology of first instar larvae were subsequently tested by RNAi lines in D. melanoga- but conserved adult heads and brains, both conservation ster where four of them indeed showed a related pheno- and divergence were found with respect to the genetic type. Likewise, some wing blister genes from D. control [6, 36, 50, 51]. Likewise, dorso-ventral patterning melanogaster were not annotated in the iBeetle screen. is relying on both, conserved and diverged gene regula- When we checked more specifically, this was often due tory networks [14, 52, 53]. Taken together, our selection to the lethality of the animal before the formation of appears to cover both conserved and diverged processes wings [17]. When we varied the timing of injection, two such that—at least for the genetic control of develop- of those knock-downs elicited wing blister phenotypes ment—our data can be generalized with some also in T. castaneum [17]. These data show that details confidence. of the screening procedure influence the subset of genes that are detected. Technical characteristics contribute to the detection of unequal gene sets Evolutionary divergence of gene function and derived Our numbers reveal that functionally comparable gene Drosophila biology may be larger than appreciated sets in two quite closely related model systems are far Most relevant for the field of functional genetics is our from identical. A question of obvious biological rele- conclusion that the degree of divergence of gene func- vance but not easily resolved is: to which degree do tions among holometabolous insects is larger than previ- these differences reflect the biologically meaningful di- ously assumed. Therefore, some genes are detected only vergence of gene functions, or alternatively, simply result in one species because the gene’s function is not re- from technical problems, i.e., reflect different strengths quired for that process in the other. This finding should and weaknesses of the respective screening methods and of course influence our thinking about using any insect model systems? as model for human development and diseases such as As discussed above, some degree of false negative data muscle fomation and congenital heart defect. may be expected in both model systems. In the case of Indeed, there is evidence supporting the notion of an the iBeetle screen, this will be restricted to technical unexpected degree of divergence with respect to muscle false negative data. In the D. melanogaster field, there development. Based on the iBeetle screen, a number of may be additional false negative data due to the lack of muscle genes identified in the iBeetle screen were more saturation of screens and/or lack of reporting of genes closely investigated in D. melanogaster [31, 49]. Despite that were not studied in detail. However, given the ex- quite some efforts, the negative data for fly orthologs ap- tent and comprehensiveness of work in the D. melano- peared to be true negative. For example, null mutations gaster field, we feel that this might not be of high of one of the genes found in our beetle, nostrin, did not relevance. As to the different strengths of screening pro- elicit a phenotype in D. melanogaster unless combined cedures, it is certainly true that the way screens are per- with a mutation of a related F-bar protein Cip4. Like- formed influences what sets of genes can be detected. wise, Rbm24 displays strong RNAi and mutant pheno- For instance, our parental RNAi approach knocked types in T. castaneum and vertebrates, respectively, but down both, maternal and zygotic contributions while D. melanogaster is lacking an Rbm24 ortholog, and func- some classic D. melanogaster screens affected only the tional compensation by paralogs was suggested to occur zygotic contribution. Hence, genes where maternal con- during D. melanogaster muscle development. Other tribution rescues the embryonic phenotype are easily genes including kahuli and unc-76 are expressed in the missed in the fly but not the beetle. For instance, paren- D. melanogaster mesoderm but only showed very subtle tal RNAi knocking down components of the aPKC com- somatic muscle phenotypes, if any, in Mef2-GAL4 driven plex leads to severe early disruption of embryogenesis in RNAi experiments or with CRISPR/Cas9 induced muta- T. castaneum while in respective D. melanogaster mu- tions, respectively (see Materials & Methods). By con- tants almost no defects are seen on the cuticle level (A. trast, their beetle counterparts had strong and penetrant Wodarz, unpublished observation). Conversely, our phenotypes in single knock-downs (e.g. see Tc-unc-76 in RNAi screen depended on the accuracy of gene annota- Fig. 5E) [31, 49, 54]. These data suggest that the function tions and our approach of screening for several pro- of genes or their relative contribution to this biological cesses in parallel may have reduced detection sensitivity. process has changed significantly. They also indicate that One striking example of the different strengths of the single gene view may be limited. Phenotypes depend screening designs is provided by wing blister phenotypes. on networks of interacting genes and this may allow for In the first part of the iBeetle screen, we detected 34 changes and replacements of individual components genes showing wing blister phenotypes where 14 did not while the overall network structure is maintained. There have related GO term annotation at FlyBase and 5 did are more striking examples of gene function changes. not have any GO annotation at all. Seven of these genes The gene germ cell-less was detected in the iBeetle Hakeemi et al. BMC Biology (2022) 20:38 Page 10 of 13 screen to govern anterior-posterior axis formation in the penetrance documented. For subsequent work, we only beetle while in D. melanogaster it is required for the for- considered phenotypes, which showed a penetrance of > mation of the posterior germ-cells [51]. Also, the D. mel- 50%. dsRNAs (1 μg/μl) were produced by Eupheria Bio- anogaster textbook example of a developmental tech Dresden, Germany. Different from the published morphogen bicoid does not even exist in T. castaneum procedure, the stink gland analysis was performed 21 [5] and yet other genes were found to act as anterior de- days after pupal injection (in the first screening phase, terminants in other flies [9, 10]. Along the same lines, this analysis had been performed after larval injection). the genes forkhead and buttonhead do not appear to be required for anterior patterning in T. castaneum but are Controls of the screen essential in flies [12, 39, 55, 56]. To assess the sensitivity and reliability of the screen, and These findings with respect to specific genes add to a to test the accuracy of each screener, we included ap- number of observations arguing for a comparatively high proximately 5% positive controls randomly chosen from degree of divergence due to the overall highly derived a set of 35 different genes. By and large, we used the nature of fly biology. The number of genes is much same positive controls as in the first screening phase smaller in D. melanogaster (appr. 14,000) compared to (see Table Table_S1_controls). However, Tc-zen-1 was T.castaneum (appr. 16,500). Further, a number of devel- excluded since the phenotypes were much weaker than opmental processes are represented in a more insect- in the previous screen, probably due to the degradation typical way in T. castaneum like for instance segmenta- of the dsRNA. We added new positive controls to score tion [57], head [50] and leg development, brain develop- for muscle and stink-gland phenotypes, which we took ment [58], extraembryonic tissue movements [59], and from novel genes detected in the first screening phase. mode of metamorphosis [60]. In most cases, the situ- New controls for muscle phenotypes: iB_06061, iB_ ation in the fly is simplified and appears to be stream- 05796, iB_03227, iB_01705; for stink gland and ovary lined for faster development. We think that these phenotypes: iB_02517; for head defects: iB_05442 (that biological differences might be the basis for divergence gene was not scored for its stink gland phenotype be- in gene function, which we just started to uncover. In cause it turned out to be too mild to be identified reli- the absence of similar large-scale comparisons in other ably in high throughput). In 143 cases (80.8%, n = 177), species, it remains open, whether an insect-typical gene the phenotypes of positive controls were fully recognized set even exists or whether one would rather have to (for comparison: in the first screening phase the respect- emphasize a constant change of gene function, such that ive numbers were: 90%, n = 201). In 14 cases (7.9%; any ancestral gene set simply “melts” away with evolu- phase 1: 4%) the phenotype was partially recognized. tionary time. This category includes complex phenotypes where half (one of two aspects: knirps, piwi, SCR, cta, cnc, iB_ Conclusion 01705, iB_05442) or two of three aspects (aristaless)of We found that the gene sets detected for the same pro- all phenotypic aspects were correctly identified. 13 phe- cesses in flies and beetles differ much more than ex- notypes were missed completely (7.3%, phase 1: 4% ). pected—only about 50% of the genes were detected in Tc-metoprene tolerant (Tc-met) was missed most fre- both species. Given the large divergence of gene sets quently, probably due to the fact that the embryonic leg found in different screening systems and the docu- phenotype was very subtle and in addition, the pene- mented cases of biological divergence of gene function, trance of the phenotype appeared to be lower than in we propose that a more systematic investigation on the the first screen (penetrance: less than 30%). Seven posi- divergence of gene function is needed. Further, the hy- tive controls (4%, phase 1: 1%) could not be analyzed pothesis independent screening now possible in T. casta- due to prior technical lethality, i.e. the premature death neum may be very helpful in that endeavor. of the injected pupae prevented the detection of the phenotype. In three cases wrong aspects were annotated Methods (false positive: 1.7%). Depending on the other annota- Screen tions these positive controls were valued as partially rec- We followed the tested and published procedures apart ognized (SCR) or missed (met, CTA). Find more details from some minor changes detailed below (please find an in Table Table_S1_controls. extensive description of the procedure in Schmitt-Engel Negative controls (buffer injections) were mainly et al. [17]). In particular, we used the same strains, injec- annotated correctly (no phenotype in 92.9%; phase 1: tion procedures, and incubation temperatures and incu- 96%) and just in 7 cases led to false positive annota- bation times. We injected 10 pupae per gene leading to tions (7.1%; phase 1: 2%) (Table Table_S1_controls; the collection of > 50 offspring L1 cuticles, which were sheet 2). analyzed. Phenotypes were annotated and their Hakeemi et al. BMC Biology (2022) 20:38 Page 11 of 13 Re-screen neoFRT}19A embryos. Single lines established from the Re-screening of selected iBeetle candidates was per- offspring were tested as heterozygotes over the balancer formed in order to probe for off-target and strain- FM7c. We used a T7 endonuclease assay for detecting specific effects. For that purpose, two independent sequence alterations near the target site as described in dsRNA fragments (original iB-fragments and one non- [62]. Our lethal Unc-76[CR007] allele carries a 16 nu- overlapping fragment, both at concentration 1 μg/μl) tar- cleotide deletion near the target site in the sequence geting the same gene were injected separately into a dif- ..TAT CCA CAC ACc aac ggt ttg gga tcc GGA TCC ferent genetic background (San Bernardino, SB strain), GGA TCC.. of the second exon (X: 2091152... 2091167, except for the muscle project where it is required to use r6.32; lower case letters represent the deleted DNA) that the pBA19 strain with EGFP marked muscles. The rest creates a frameshift in the ORF of all known isoforms of the injection procedures and analyses were performed (i.e., the frameshift occurs after T246 in Unc-76 RA to like in the screen. Note that with this approach, we can- -C and after T61 in Unc-76 RD). not exclude that phenotypes observed in one tissue are elicited by knock-down in another tissue (e.g., hormone- Supplementary Information induced morphogenesis may fail due to knock-down of The online version contains supplementary material available at https://doi. org/10.1186/s12915-022-01231-4. hormone production in a gland). Additional file 1: Table S1. Positive and negative controls of the Fly gene sets screen. Lists of genes involved in those processes were estab- Additional file 2: Figure S1. Diagram displaying the portion of genes lished by our experts of the respective processes. This known to be required for the processes in Drosophila, which were tested was supported by individual FlyBase searches for re- in Tribolium. Figures S2 and S3. Phylogenetic trees supporting our claim of absence of an ortholog in Drosophila. spective GO terms of the category “biological process”. Additional file 3: Table S2. Genes with phenotype in Tribolium but not For the analyses in Figs. 2 and 3, we only considered Drosophila. gene functions, which were based on experimental data Additional file 4: Table S3. Genes with phenotype in Drosophila as documented at FlyBase in the year 2017 (Dmel Re- compared to Tribolium data. lease 6.18) with updates in single case in 2020 (Dmel Re- Additional file 5: Table S4. Combined analysis leading to the numbers lease 6.32). given in Figure 4. Phylogenetic analysis Acknowledgements The Tribolium protein sequences from gene set OGS3_ We thank Mohamad Al Heshan, Elke Küster, and Claudia Hinners for help with injection and processing during the screen. proteins.fasta.gz (including changes from 2016/02/15, available from [27]) were used to retrieve the most simi- Authors’ contributions lar proteins of T. castaneum, D. melanogaster, and M. MSH, SA, MT, MW, DG, TK, and XW performed high throughput RNAi musculus using only one isoform. Multiple alignments screening and annotation of phenotypes. DG and TK led the screening were done with the ClustalOmega plugin as imple- teams. MK, MS, and GB led the screen. JD provided bioinformatics support and developed iBeetle-Base. DS did follow-up research in Drosophila includ- mented in the Geneious 10.1.3 software (Biomatters, ing the generation of mutants. JS, DG, JD, MSH, and GB gathered and inter- Auckland, New Zealand) using standard settings. Align- preted information on differences in gene function between beetle and fly. ments were trimmed to remove poorly aligned sequence GB and MSH wrote the first draft of the manuscript and prepared the figures. MF, SF, MS, and MK supervised the analyses, interpreted data, and edited the stretches. Phylogenetic trees were calculated using the manuscript. All authors have read and approved the final manuscript. FastTree 2.1.5 plugin implemented in Geneious. Funding Generation of Unc-76 mutations via CRISPR/Cas9 Deutsche Forschungsgemeinschaft (DFG) funded the research unit FOR1234 In order to generate the Unc-76 mutations, we essen- “iBeetle” including the following grants: BU 1443/6-2; BU 1443/7-2; Kl 656/7- 2; Bu 1443/8-2; Kl 656/8-2; Scho 1058/4-2; FR 696/4-2; RO 890/4-2; STA 1009/ tially followed the procedure described by Basset et al., 10-1; iBeetle-Base was supported by DFG LIS project 417202192 2013 [61]—please find an extensive description there. Bayer CropScience funded part of the continuation of the screen. For making the template for the guide RNAs, the Unc- The China Scholarship Council funded Xuebein Wan (201706760058). The funding bodies had no influence on the project’s design and 76 target sequence GGTTCAACGATCTGACCAGTG publication strategy. Open Access funding enabled and organized by Projekt was inserted between the T7 promoter and the gRNA DEAL. core sequence in the forward primer, gRNA_F. After an- nealing gRNA_F with SGRNAR, the template was PCR Availability of data and materials amplified with Q5 polymerase (NEB). Guide RNAs were The datasets generated and/or analyzed during the current study are available from the iBeetle-Base [25, 26] repository [27]. transcribed with Ampliscribe T7 Flash (epicenter), iso- The dsRNA fragments used to knock down the genes are commercially lated with the MEGAclear kit (Ambion), and injected to- available from Eupheria Biotech, their sequences are documented in the gether with Cas9 mRNA into w[1118] sn[3] P{ry+t7.2= iBeetle-Base. Hakeemi et al. BMC Biology (2022) 20:38 Page 12 of 13 Declarations 14. Lynch JA, Roth S. The evolution of dorsal–ventral patterning mechanisms in insects. Genes Dev. 2011;25(2):107–18. https://doi.org/10.1101/gad.2010711. Ethics approval and consent to participate 15. Stappert D, Frey N, von Levetzow C, Roth S. Genome-wide identification of Not applicable Tribolium dorsoventral patterning genes. Development. 2016;143(13):2443– 54. https://doi.org/10.1242/dev.130641. 16. Bucher G, Scholten J, Klingler M. Parental RNAi in Tribolium (Coleoptera). Consent for publication Curr Biol. 2002;12(3):R85–6. https://doi.org/10.1016/S0960-9822(02)00666-8. Not applicable 17. Schmitt-Engel C, Schultheis D, Schwirz J, Strohlein N, Troelenberg N, Majumdar U, et al. The iBeetle large-scale RNAi screen reveals gene Competing interests functions for insect development and physiology. Nat Commun. 2015;6(1): The authors declare that they have no competing interests. 7822. https://doi.org/10.1038/ncomms8822. 18. Tomoyasu Y, Denell RE. Larval RNAi in Tribolium (Coleoptera) for analyzing Author details 1 adult development. Dev Genes Evol. 2004;214(11):575–8. https://doi.org/10.1 Johann-Friedrich-Blumenbach-Institut, GZMB, University of Göttingen, 2 007/s00427-004-0434-0. Justus-von-Liebig-Weg 11, 37077 Göttingen, Germany. Current address: 19. Berghammer AJ, Klingler M, Wimmer EA. A universal marker for transgenic Institute of Clinical Pharmacology, University Medical Center Göttingen, insects. Nature. 1999;402(6760):370–1. https://doi.org/10.1038/46463. University of Göttingen, Robert-Koch-Str. 40, 37075 Göttingen, Germany. 20. Gilles AF, Schinko JB, Averof M. Efficient CRISPR-mediated gene targeting Department of Biology, Friedrich-Alexander-University Erlangen-Nürnberg, and transgene replacement in the beetle Tribolium castaneum. Staudtstr. 5, 91058 Erlangen, Germany. Current address: Department of Development. 2015;142(16):2832–9. https://doi.org/10.1242/dev.125054. Medical Bioinformatics, University Medical Center Göttingen, University of 21. Schinko JB, Weber M, Viktorinova I, Kiupakis A, Averof M, Klingler M, et al. Göttingen, Goldschmidtstr. 1, 37077 Göttingen, Germany. Current address: Functionality of the GAL4/UAS system in Tribolium requires the use of Department of Insect Biotechnology, Justus-Liebig-University Giessen, endogenous core promoters. BMC Dev Biol. 2010;10(1):53. https://doi.org/1 Heinrich-Buff-Ring 26-32, 35392 Gießen, Germany. Current address: Institute 0.1186/1471-213X-10-53. of Pathology, University Hospital Cologne, Kerpener Str. 62, 50924 Cologne, 22. Herndon N, Shelton J, Gerischer L, Ioannidis P, Ninova M, Dönitz J, et al. Germany. Current address: Institute of Neuropathology, University Hospital Enhanced genome assembly and a new official gene set for Tribolium Erlangen, Friedrich-Alexander-University Erlangen-Nürnberg, Erlangen, castaneum. BMC Genomics. 2020;21(1):47. https://doi.org/10.1186/s12864-01 Germany. Institute for Zoology/Developmental Biology, University of 9-6394-6. Cologne, Biocenter, Zülpicher Straße 47b, D-50674 Köln, Germany. 23. Mungall CJ, Gkoutos GV, Smith CL, Haendel MA, Lewis SE, Ashburner M. Integrating phenotype ontologies across multiple species. Genome Biol. Received: 15 July 2021 Accepted: 12 January 2022 2010;11(1):R2. https://doi.org/10.1186/gb-2010-11-1-r2. 24. Dönitz J, Grossmann D, Schild I, Schmitt-Engel C, Bradler S, Prpic N-M, et al. TrOn: An Anatomical Ontology for the Beetle Tribolium castaneum. PLoS References ONE. 2013;8(7):e70695. https://doi.org/10.1371/journal.pone.0070695. 1. Jorgensen EM, Mango SE. The art and design of genetic screens: 25. Dönitz J, Schmitt-Engel C, Grossmann D, Gerischer L, Tech M, Schoppmeier caenorhabditis elegans. Nat Rev Genet. 2002;3(5):356–69. https://doi.org/10.1 M, et al. iBeetle-Base: a database for RNAi phenotypes in the red flour 038/nrg794. beetle Tribolium castaneum. Nucl Acids Res. 2015;43(D1):D720–5. https:// 2. Kile BT, Hilton DJ. The art and design of genetic screens: mouse. Nat Rev doi.org/10.1093/nar/gku1054. Genet. 2005;6(7):557–67. https://doi.org/10.1038/nrg1636. 26. Dönitz J, Gerischer L, Hahnke S, Pfeiffer S, Bucher G. Expanded and updated 3. Patton EE, Zon LI. The art and design of genetic screens: zebrafish. Nat Rev data and a query pipeline for iBeetle-Base. Nucleic Acids Res. 2018;46(D1): Genet. 2001;2(12):956–66. https://doi.org/10.1038/35103567. D831–5. https://doi.org/10.1093/nar/gkx984. 4. St Johnston D. The art and design of genetic screens: Drosophila 27. iBeetle Consortium. iBeetle-Base: A Database of Tribolium Genes, Genomes melanogaster. Nat Rev Genet. 2002;3(3):176–88. https://doi.org/10.1038/ and RNAi phenotypes. http://ibeetle-base.uni-goettingen.de/. Accessed 19. nrg751. Nov 2018. 5. Brown S, Fellers J, Shippy T, Denell R, Stauber M, Schmidt-Ott U. A strategy 28. Bier E, Bodmer R. Drosophila, an emerging model for cardiac disease. Gene. for mapping bicoid on the phylogenetic tree. Curr Biol. 2001;11(2):R43–4. 2004;342(1):1–11. https://doi.org/10.1016/j.gene.2004.07.018. https://doi.org/10.1016/S0960-9822(01)00007-0. 29. Ahmad SM. Conserved signaling mechanisms in Drosophila heart 6. Fu J, Posnien N, Bolognesi R, Fischer TD, Rayl P, Oberhofer G, et al. development. Dev Dyn. 2017;246(9):641–56. https://doi.org/10.1002/dvdy.24530. Asymmetrically expressed axin required for anterior development in 30. Johnson MS, Lu N, Denessiouk K, Heino J, Gullberg D. Integrins during Tribolium. Proc Natl Acad Sci USA. 2012;109(20):7782–6. https://doi.org/10.1 evolution: evolutionary trees and model organisms. Biochim Biophys Acta. 073/pnas.1116641109. 1788;2009(4):779–89. https://doi.org/10.1016/j.bbamem.2008.12.013. 7. Glinka A, Wu W, Delius H, Monaghan AP, Blumenstock C, Niehrs C. Dickkopf- 31. Schultheis D, Schwirz J, Frasch M. RNAi screen in tribolium reveals 1 is a member of a new family of secreted proteins and functions in head involvement of F-BAR proteins in myoblast fusion and visceral muscle induction. Nature. 1998;391(6665):357–62. https://doi.org/10.1038/34848. morphogenesis in insects. G3 (Bethesda). 2019;9:1141–51. 8. Gurley KA, Rink JC, Sánchez AA. Beta-catenin defines head versus tail 32. Theis JL, Vogler G, Missinato MA, Li X, Nielsen T, Zeng X-XI, et al. Patient- identity during planarian regeneration and homeostasis. Science. 2008; specific genomics and cross-species functional analysis implicate LRP2 in 319(5861):323–7. https://doi.org/10.1126/science.1150029. hypoplastic left heart syndrome. eLife. 2020;9:e59554. https://doi.org/10. 9. Klomp J, Athy D, Kwan CW, Bloch NI, Sandmann T, Lemke S, et al. Embryo 7554/eLife.59554. development. A cysteine-clamp gene drives embryo polarity in the midge 33. Zaffran S, Frasch M. Early signals in cardiac development. Circ Res. 2002; Chironomus. Science (New York, NY). 2015;348:1040–2. 91(6):457–69. https://doi.org/10.1161/01.RES.0000034152.74523.A8. 10. Yoon Y, Klomp J, Martin-Martin I, Criscione F, Calvo E, Ribeiro J, et al. 34. Niehrs C. On growth and form: a Cartesian coordinate system of Wnt and Embryo polarity in moth flies and mosquitoes relies on distinct old genes BMP signaling specifies bilaterian body axes. Development. 2010;137(6):845– with localized transcript isoforms. Elife. 2019;8:e46711. https://doi.org/10. 57. https://doi.org/10.1242/dev.039651. 7554/eLife.46711. 35. Finkelstein R, Boncinelli E. From fly head to mammalian forebrain: the story 11. Averof M. Arthropod Hox genes: insights on the evolutionary forces that of otd and Otx. Trends Genet. 1994;10(9):310–5. https://doi.org/10.1016/01 shape gene functions. Curr Opin Genet Dev. 2002;12(4):386–92. https://doi. 68-9525(94)90033-7. org/10.1016/S0959-437X(02)00314-3. 12. Kittelmann S, Ulrich J, Posnien N, Bucher G. Changes in anterior head 36. Posnien N, Koniszewski NDB, Hein HJ, Bucher G. Candidate Gene Screen in patterning underlie the evolution of long germ embryogenesis. Dev Biol. the Red Flour Beetle Tribolium Reveals Six3 as Ancient Regulator of Anterior 2013;374(1):174–84. https://doi.org/10.1016/j.ydbio.2012.11.026. Median Head and Central Complex Development. PLoS Genet. 2011;7(12): 13. Kitzmann P, Weißkopf M, Schacht MI, Bucher G. A key role forfoxQ2in e1002418. https://doi.org/10.1371/journal.pgen.1002416. anterior head and central brain patterning in insects. Development. 2017; 37. Hunnekuhl VS, Akam M. An anterior medial cell population with an apical- 144(16):2969–81. https://doi.org/10.1242/dev.147637. organ-like transcriptional profile that pioneers the central nervous system in Hakeemi et al. BMC Biology (2022) 20:38 Page 13 of 13 the centipede Strigamia maritima. Dev Biol. 2014;396(1):136–49. https://doi. 59. Panfilio KA. Extraembryonic development in insects and the acrobatics of org/10.1016/j.ydbio.2014.09.020. blastokinesis. Developmental Biology. 2008;313(2):471–91. https://doi.org/1 38. Leclère L, Bause M, Sinigaglia C, Steger J, Rentzsch F. Development of the 0.1016/j.ydbio.2007.11.004. aboral domain in Nematostella requires β-catenin and the opposing 60. Snodgrass R. Insect Metamorphosis: Smithsonian Miscellaneous Collections, activities of Six3/6 and Frizzled5/8. Development. 2016;143(10):1766–77. V122, No. 9. Literary Licensing: Washington; 1954. https://doi.org/10.1242/dev.120931. 61. Bassett AR, Tibbit C, Ponting CP, Liu J-L. Highly efficient targeted mutagenesis of Drosophila with the CRISPR/Cas9 system. Cell Rep. 2013;4(1): 39. Schinko JB, Kreuzer N, Offen N, Posnien N, Wimmer EA, Bucher G. 220–8. https://doi.org/10.1016/j.celrep.2013.06.020. Divergent functions of orthodenticle, empty spiracles and buttonhead 62. Kondo S, Ueda R. Highly improved gene targeting by germline-specific Cas9 in early head patterning of the beetle Tribolium castaneum expression in Drosophila. Genetics. 2013;195(3):715–21. https://doi.org/10.1 (Coleoptera). Dev Biol. 2008;317(2):600–13. https://doi.org/10.1016/j. 534/genetics.113.156737. ydbio.2008.03.005. 40. Trauner J, Büning J. Germ-cell cluster formation in the telotrophic meroistic ovary of Tribolium castaneum (Coleoptera, Polyphaga, Tenebrionidae) and Publisher’sNote its implication on insect phylogeny. Dev Genes Evol. 2007;217(1):13–27. Springer Nature remains neutral with regard to jurisdictional claims in https://doi.org/10.1007/s00427-006-0114-3. published maps and institutional affiliations. 41. Teuscher M, Ströhlein N, Birkenbach M, Schultheis D, Schoppmeier M. TC003132 is essential for the follicle stem cell lineage in telotrophic Tribolium oogenesis. Front Zool. 2017;14(1):26. https://doi.org/10.1186/s12983-017-0212-2. 42. Bäumer D, Strohlein NM, Schoppmeier M. Opposing effects of Notch- signaling in maintaining the proliferative state of follicle cells in the telotrophic ovary of the beetle Tribolium. Front Zool. 2012;9(1):15. https:// doi.org/10.1186/1742-9994-9-15. 43. FlyBase Consortium. FlyBase - A Database of Drosophila Genes & Genomes. http://flybase.org/. Accessed 19. Nov 2018 44. Gelbart WM, Crosby MA, Matthews B, Rindone WP, Chillemi J, Twombly SR, et al. FlyBase: a Drosophila database. The FlyBase consortium. Nucleic Acids Res. 1997;25(1):63–6. 45. Kitzmann P, Schwirz J, Schmitt-Engel C, Bucher G. RNAi phenotypes are influenced by the genetic background of the injected strain. BMC Genomics. 2013;14(1):5. https://doi.org/10.1186/1471-2164-14-5. 46. Lorenzen MD, Kimzey T, Shippy TD, Brown SJ, Denell RE, Beeman RW. piggyBac-based insertional mutagenesis in Tribolium castaneum using donor/helper hybrids. Insect Mol Biol. 2007;16(3):265–75. https://doi.org/1 0.1111/j.1365-2583.2007.00727.x. 47. Carbon S, Ireland A, Mungall CJ, Shu S, Marshall B, Lewis S, et al. AmiGO: online access to ontology and annotation data. Bioinformatics. 2009;25(2): 288–9. https://doi.org/10.1093/bioinformatics/btn615. 48. GO Consortium. Gene Ontology. Database http://geneontology.org/. Accessed 19. Nov 2018. 49. Schultheis D, Weißkopf M, Schaub C, Ansari S, Dao VA, Grossmann D, et al. A Large Scale Systemic RNAi Screen in the Red Flour Beetle Tribolium castaneum Identifies Novel Genes Involved in Insect Muscle Development. G3 (Bethesda). 2019;9:1009–26. 50. Posnien N, Schinko JB, Kittelmann S, Bucher G. Genetics, development and composition of the insect head - A beetle’s view. Arthropod Struct Dev. 2010;39(6):399–410. https://doi.org/10.1016/j.asd.2010.08.002. 51. Ansari S, Troelenberg N, Dao VA, Richter T, Bucher G, Klingler M. Double abdomen in a short-germ insect: Zygotic control of axis formation revealed in the beetle Tribolium castaneum. PNAS. 2018;115:201716512. 52. Nunes da Fonseca R, von Levetzow C, Kalscheuer P, Basal A, van der Zee M, Roth S. Self-regulatory circuits in dorsoventral axis formation of the short- germ beetle Tribolium castaneum. Dev Cell. 2008;14:605–15. 53. Pechmann M, Kenny NJ, Pott L, Heger P, Chen Y-T, Buchta T, et al. Striking parallels between dorsoventral patterning in Drosophila and Gryllus reveal a complex evolutionary history behind a model gene regulatory network. Elife. 2021;10:e68287. https://doi.org/10.7554/eLife.68287. 54. Schultheis D. Identifizierung und Charakterisierung neuer regulatorischer Gene in der Muskelentwicklung durch einen genomweiten RNAi-Screen in Tribolium castaneum; 2016. 55. Weigel D, Jürgens G, Kuttner F, Seifert E, Jäckle H. The homeotic gene fork head encodes a nuclear protein and is expressed in the terminal regions of the Drosophila embryo. Cell. 1989;57(4):645–58. https://doi.org/10.1016/ 0092-8674(89)90133-5. 56. Wimmer EA, Cohen SM, Jackle H, Desplan C. Buttonhead does not contribute to a combinatorial code proposed for Drosophila head development. Development. 1997;124(8):1509–17. https://doi.org/10.1242/dev.124.8.1509. 57. Tautz D, Friedrich M, Schröder R. Insect embryogenesis – what is ancestral and what is derived? Development. 1994;1994(Supplement):193–9. 58. Farnworth MS, Eckermann KN, Bucher G. Sequence heterochrony led to a gain of functionality in an immature stage of the central complex: a fly- beetle insight. 2020;18(10). https://doi.org/10.1371/journal.pbio.3000881. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Biology Springer Journals

Loading next page...
 
/lp/springer-journals/screens-in-fly-and-beetle-reveal-vastly-divergent-gene-sets-required-f9UHunQimn

References (63)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
eISSN
1741-7007
DOI
10.1186/s12915-022-01231-4
Publisher site
See Article on Publisher Site

Abstract

Background: Most of the known genes required for developmental processes have been identified by genetic screens in a few well-studied model organisms, which have been considered representative of related species, and informative—to some degree—for human biology. The fruit fly Drosophila melanogaster is a prime model for insect genetics, and while conservation of many gene functions has been observed among bilaterian animals, a plethora of data show evolutionary divergence of gene function among more closely-related groups, such as within the insects. A quantification of conservation versus divergence of gene functions has been missing, without which it is unclear how representative data from model systems actually are. Results: Here, we systematically compare the gene sets required for a number of homologous but divergent developmental processes between fly and beetle in order to quantify the difference of the gene sets. To that end, we expanded our RNAi screen in the red flour beetle Tribolium castaneum to cover more than half of the protein- coding genes. Then we compared the gene sets required for four different developmental processes between beetle and fly. We found that around 50% of the gene functions were identified in the screens of both species while for the rest, phenotypes were revealed only in fly (~ 10%) or beetle (~ 40%) reflecting both technical and biological differences. Accordingly, we were able to annotate novel developmental GO terms for 96 genes studied in this work. With this work, we publish the final dataset for the pupal injection screen of the iBeetle screen reaching a coverage of 87% (13,020 genes). Conclusions: We conclude that the gene sets required for a homologous process diverge more than widely believed. Hence, the insights gained in flies may be less representative for insects or protostomes than previously thought, and work in complementary model systems is required to gain a comprehensive picture. The RNAi screening resources developed in this project, the expanding transgenic toolkit, and our large-scale functional data make T. castaneum an excellent model system in that endeavor. Keywords: Gene function, Comparative genomics, RNAi screen, iBeetle, Tribolium castaneum, Drosophila melanogaster, Divergence of gene function, iBeetle-Base, FlyBase * Correspondence: Gbucher1@uni-goettingen.de Muhammad Salim Hakeemi, Salim Ansari, Matthias Teuscher and Matthias Weißkopf contributed equally to this work. Johann-Friedrich-Blumenbach-Institut, GZMB, University of Göttingen, Justus-von-Liebig-Weg 11, 37077 Göttingen, Germany Full list of author information is available at the end of the article © The Author(s). 2022 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Hakeemi et al. BMC Biology (2022) 20:38 Page 2 of 13 Background and head specification show the involvement of similar The function of genes is paramount for the biology of gene sets, although a few components appear to be re- any organism and, hence, the assignment of functions to quired for only some clades [12–15]. genes is a central question of biological research. How- Notably, the differences in gene functions documented ever, only in a very small number of genetic model spe- so far may be an underestimation of the real divergence, cies like the mouse Mus musculus, the zebrafish Danio because the prevailing candidate gene approach leads to rerio, the nematode Caenorhabditis elegans, and the vin- a systematic bias towards conservation. The genes to be egar fly Drosophila melanogaster have the functions of tested are usually chosen based on the knowledge of developmental genes been assayed in systematic screens. their ortholog’s involvement in other species. As a con- This restriction to a few model systems is a consequence sequence, unrelated genes are rarely tested and the in- of the necessity for an elaborate genetic and molecular volvement of unexpected genes in a given process is tool kit, which is extremely laborious to establish [1–4]. underestimated. Hence, approaches are needed to over- Unfortunately, it has remained unclear how representa- come this bias and to gain a realistic view on the degree tive findings in these model species actually are for their of gene function divergence. To that end, genes required clade or in other words, how quickly and profoundly for certain biological processes need to be identified in gene function diverges in evolution. Indeed, based on an an unbiased and genome-wide manner in non- astonishing degree of functional conservation with re- traditional organisms as well, even though this has spect to many genes, it has been assumed that the find- remained technically challenging and such large datasets ings gained in those few model species can be outside the classic model systems had not been available. transferred to a large degree. This assumption underlies The red flour beetle T. castaneum has recently been the use of the insect D. melanogaster as a model for a established as the only arthropod model organism apart number of human diseases, including developmental as- from D. melanogaster where genome-wide unbiased pects like heart formation. Conversely, even among in- RNAi screens are feasible. Based on the robust and sys- sects, dramatic evolutionary changes of gene functions temic RNAi response of this species, the iBeetle large- have been described with respect to homologous devel- scale screen was performed where random genes were opmental processes, calling into question the degree of knocked down and the resulting animals were scored for transferability. However, a systematic comparison of the a number of developmental phenotypes [16–18]. How- gene sets required for the same process in different spe- ever, the data gained in the iBeetle screen had covered cies has not been done and, hence, a quantification of di- only one third of the gene set, not allowing for robust vergence versus conservation remains badly missing. genome wide statements. Apart from its particularly Importantly, knowing the degrees of gene function di- strong and robust RNAi response, T. castaneum offers a vergence is relevant not only for understanding the evo- comparably large tool kit for analyzing gene function in- lution of biodiversity but also for applied research, e.g., cluding transgenic and genome editing approaches [19– for transferring knowledge from model systems to spe- 21]. cies relevant for medical applications or pest control. In this paper, we used an expanded dataset to assess Recently, the study of gene function in development the degree of divergence of the gene sets required for se- has been extended to non-traditional model organisms. lected developmental processes between fly and beetle Predominantly, candidate genes known for their function such as head, muscle and ovary development, and in the classical model systems have been tested in other dorso-ventral patterning. These processes are homolo- organisms. Subsequent comparisons revealed both, con- gous but show a different degree of evolutionary diver- servation and divergence of gene functions. For example, gence, which could be reflected in changes of gene axis formation in D. melanogaster has turned out to be a function. First, we determined genes that were essential rather diverged process partially based on different genes in the beetle for these processes but which had so far compared to other insects. The key anterior morphogen not been connected to them in D. melanogaster. These a of D. melanogaster, bicoid, is not present in most insects priori unexpected genes sum up to about 37% of the [5]. Instead, repression of Wnt signaling plays a central total genes identified to be required for either one or role in the red flour beetle Tribolium castaneum [6]asit both species. For 30% of these genes, no functional an- does in many animals including other insects, flatworms, notation had been available at FlyBase at all such that and vertebrates [7–10] - but not in D. melanogaster. The we provide the first functional Gene Ontology (GO) as- functions of genes of the Hox cluster, in contrast, appear signment for the respective ortholog group in insects. conserved over very large phylogenetic distances—al- Only two genes essential in T. castaneum did not have though some functional divergence has been linked to an ortholog in D. melanogaster, i.e., these processes the evolution of arthropod morphology [11]. Likewise, seem not much affected by gene gain or loss. We con- the gene regulatory network of dorso-ventral patterning clude that restricting genetic screens to one model Hakeemi et al. BMC Biology (2022) 20:38 Page 3 of 13 system only falls short of identifying a comprehensive set of essential genes. Further, our data reveals an unex- pected degree of divergence of gene function between two holometabolous insect species. Moreover, we present here an update of the dataset gained in the gen- ome wide iBeetle screen in T. castaneum. Our analysis is based on both, a dataset previously published comprising 5300 genes [17] and an additional 3200 genes screened as part of this project. In addition to those, we publish and make accessible (at iBeetle-Base) the phenotypes for an additional 4520 genes which were screened while the analysis presented here was ongoing. Hence, with this paper, the coverage of genes tested and annotated at iBeetle-Base sums to 13,020 Tribolium genes (78% of the predicted gene set), which will be the final number for the pupal injection screen of the iBeetle screen. Results Fig. 1 Quality controls of the primary screen. 178 positive controls Reaching the final dataset for pupal injections of the using 35 different genes were included. More than 88% of the large scale iBeetle screen positive controls were fully or partially recognized (left bar) while We added 3200 genes to the previously published 5300 7.3% were missed. 4% could not be analyzed due to technical genes of our large-scale iBeetle screen [17], reaching a lethality before the production of offspring. 7.1% of the negative coverage of 51% of the T. castaneum gene set of a total controls had annotations, i.e. they were false positive (right bar). These figures are similar to the first screening phase [17] of 16,593 currently annotated genes [22]. We followed the previously described procedure for the pupal injec- tion screen [17] with minor modifications (see the “Methods” section). In short, we injected 10 female analysis presented here because both analyses ran in pupae per gene with dsRNAs (concentration 1 μg/μl). parallel. We annotated the phenotypes of the injected animals and the first instar cuticle of their offspring using the Unexpected gene functions in developmental processes EQM system [23], the T. castaneum morphological We wanted to use our large-scale phenotypic dataset to ontology Tron [24], and a controlled vocabulary. The systematically compare the gene sets required for the data is available at the online database iBeetle-Base [25– same biological processes in T. castaneum and D. mela- 27]. As positive controls, an array of genes with known nogaster. We define a gene to be “required for” a phenotypes related to the processes under scrutiny was process, if its knock-down or mutation leads to a pheno- added to the screening and included both drastic and type in that process. To that end, we first identified in mild phenotypes (find details in the extensive descrip- an unbiased way all genes annotated with a phenotype tion of the first part of the screen in Schmitt-Engel et al. for a number of biological processes by searching iBee- [17]). Buffer injections were performed as negative con- tle-Base. Specifically, we scored for phenotypes indicative trols. These controls revealed a similar portion of false of functions in dorso-ventral patterning, head and negative and false positive data compared to the first muscle development, and in oogenesis. part of the screen (Fig. 1 and Additional file 1: Table The choice of the processes to be compared was based S1). The analysis presented in this work is based on all on two arguments. On one hand, we chose processes genes that had been screened in the first round of the that fell within our core expertise for sake of correct in- screen and the set of genes published with this publica- terpretation of phenotypes, for the opportunity for sub- tion. Taken together, the analyzed set of genes covered sequent detailed work, and to ensure comprehensive approximately 50% of the genome. About two thirds of knowledge of respective Drosophila data. On the other the Drosophila genes relevant for our analysis had been hand, we aimed at including developmental processes covered by the screen (Additional file 2: Fig. S1). While with different degrees of conservation. We assumed that this analysis was ongoing, we continued the screen as cellular processes such as muscle development might be well and have in the meanwhile reached a coverage of conserved. Indeed, for both, conservation of basic mech- 78% (13,020 genes). We publish these additional pheno- anisms between Drosophila and vertebrates was shown typic data (accessible online at iBeetle-Base) with this along with evolutionary differences [28–31], and the use article, but they were not included in the detailed of Drosophila as a model for vertebrate heart Hakeemi et al. BMC Biology (2022) 20:38 Page 4 of 13 development is widely adopted [32, 33]. Some aspects of searching T. castaneum, D. melanogaster, and M. muscu- dorso-ventral patterning are highly conserved among an- lus genomes for orthologs and paralogs. This analysis re- imals (e.g., the involvement of dpp/BMP versus sog) but vealed that only three genes with a novel function (appr. differences even between insects have been described as 3%) did not have a D. melanogaster ortholog (yellow in well [14, 34]. Likewise, a set of highly conserved genes is Fig. 2; see Additional file 2: Fig. S2 and S3 for phylogen- involved in anterior development in all animal clades etic trees). Evidently, lineage-specific gene loss or gain while clear differences were described even between flies explains only a minor part of the functional divergence and beetles [12, 35–39]. This divergence may to some of homologous developmental processes. degree be due to the fly-specific involution and reduc- Next, we asked whether the respective D. melanogaster tion of the larval head. Finally, telotrophic oogenesis of orthologs were known to be involved in other biological Tribolium is morphologically quite distinct from the processes or lacked any phenotype information. For this polytrophic oogenesis found in Drosophila [40–42] and analysis, we did not only use the genes identified in the both are quite different from the vertebrate process screen but included those that had previously been stud- probably representing a more divergent process. ied, as well. We looked up phenotype information of the For all these processes, we found a number of gene respective D. melanogaster orthologs on FlyBase (hom- functions that were expected based on D. melanogaster ology assignment done with OrthoDB v9). Among the knowledge (see below). This confirmed that the screen fly orthologs whose functional annotations did not design allowed the detection of these types of pheno- match with those from the iBeetle screen or published types. Importantly, we also found functions for genes so records of functional data, around two thirds (64.6%) far not connected to those processes (based on FlyBase had annotations that were related to other processes information [43, 44], PubMed searches, and scientist ex- than the ones studied in T. castaneum (Fig. 2). Import- pertise). The iBeetle screen is a first pass screen with a antly, one third of the genes (32.3%) did not have any focus on minimizing false negative results with the functional annotation in FlyBase. Hence, for those genes, trade-off of allowing for false positive annotations [17]. the iBeetle-screen had detected the first documented The likelihood for this type of error is further increased function of that ortholog group in insects. Importantly, by off-target effects and/or by strain-specific differences due to the lack of previous phenotypic information, in the phenotype, i.e. genes for which the described these genes likely would not have been included in a phenotype was not reproduced in another genetic back- classical candidate gene approach. ground, which we counted as “false positive” in order to be conservative [45]. Hence, we aimed at excluding false A quarter of Drosophila gene function annotations were positive annotations for the unexpected gene functions. not confirmed for T. castaneum First, we based our analyses only on genes for which In a complementary approach, we asked how many genes phenotypes had been annotated with a penetrance of > known to be required for a given process in D. melanoga- 50% in the primary screen. Further, we only used pheno- ster had been assigned related functions in the iBeetle types that were reproduced by RNAi experiments with screen. To that end, we first collected lists of genes re- non-overlapping dsRNA fragments targeting the same quired for those processes based on D. melanogaster gene. In order to exclude genetic background effects, we knowledge (expert knowledge, literature, and FlyBase) used another lab strain (our standard lab strain San Ber- (Additional file 4: Table S3). Then we mined iBeetle-Base nardino, SB) except for the muscle project where we to see how many of the beetle orthologs had an annota- needed to use the pBA19 strain, which has EGFP tion related to that process (Fig. 3A). About two-thirds of marked muscles [46]. This re-screening procedure re- those genes had actually been screened in T. castaneum sulted in a set of genes for which we can claim with high (Additional file 2: Fig. S1) and all following numbers are confidence that they are indeed required for these pro- based on the analysis of this subset. cesses in T. castaneum, but which previously were not A surprisingly large portion of genes (26.4%) known assigned to these in D. melanogaster (Additional file 3: to be required for these processes in D. melanogaster Table S2). did not show the expected phenotype in T. casta- neum (Fig. 3B). Assigning the first function to a gene versus extending previous annotations Enriching the GO information with data from Tribolium One reason for a lack of respective functional data in Gene ontology (GO) assignment is a powerful tool to es- FlyBase could be that the knocked-down beetle gene tablish hypotheses on the function of given gene sets does not have an ortholog in the fly. In order to test this, [47]. So far, there were no GO terms associated based we searched for the fly orthologs in orthoDB and by on T. castaneum data. The work presented here revealed manually generating phylogenetic trees based on that a surprisingly high portion of orthologous genes has Hakeemi et al. BMC Biology (2022) 20:38 Page 5 of 13 Fig. 2 Analysis of genes with unexpected gene functions found in Tribolium. A Numbers of genes, for which an unexpected function in the respective process was found in the iBeetle screen but had not been known from Drosophila. B Combined numbers for all four processes. Only three genes with novel gene functions in Tribolium had no ortholog in Drosophila (yellow). About two-thirds of genes with novel function had previous phenotypic annotations in FlyBase but relating to other biological processes (blue). Importantly, for one third of those genes, we had detected the first phenotype in any insect (green). We added this novel information to the GO database diverging functions in different organisms. To enrich the based on work in flies. However, there are several rea- GO database, we submitted GO terms with respect to sons to believe that the picture has remained incom- the biological process for all 96 re-screened genes with plete. On one hand, species-specific or technical functions in dorso-ventral patterning (GO:0010084), oo- limitations may have prohibited the identification of an genesis (GO:0048477), the development of embryonic involved gene in D. melanogaster. On the other hand, muscles (GO:0060538), and head (GO:0048568) [48]. evolution has led to functional changes such as the modification or loss of ancestral gene functions or the Discussion co-option of genes into a novel process. Unfortunately, Investigating one species falls short of a comprehensive it has remained unclear to what extent the gene sets de- view on gene function termined exclusively in flies would be representative of Large-scale screens in the leading insect model organism insects as a whole or if it is even appropriate to assume D. melanogaster have revealed gene sets required for the existence of representative gene sets. certain biological processes. As consequence, insect- Our systematic screening in a complementary model related GO term annotations are almost exclusively organism has revealed that the identified gene sets show Fig. 3 Beetle genes showing phenotypes expected from Drosophila. A Gene sets known to be required for a given process in Drosophila were compared to iBeetle data. Close to three quarters showed related phenotypes (blue) while others had no or different types of phenotypes (green). B Approximately one quarter of the genes known to be required for certain Drosophila processes were not required for that process in Tribolium. This analysis is based on the subset of genes which already had been screened in Tribolium (51%). Interestingly, we found orthologs for 66% of the respective Drosophila genes—this indicates that our screen was enriched for relevant genes Hakeemi et al. BMC Biology (2022) 20:38 Page 6 of 13 an unexpected degree of divergence (see Fig. 4 for num- one species (11% vs. 37%) may reflect both, biological bers, Fig. 5 for examples). Based on our calculations (see and technical differences (see detailed discussion below). details below) we estimate that only half of the gene Beyond the fly-beetle comparison, our findings provide functions are detected in both species (52%, column 4 of a compelling argument that focusing on single model Fig. 4A) while the remaining gene functions were found species falls short of comprehensively revealing the gen- either only in D. melanogaster (11%, column 4 of Fig. etic basis of biological processes in any clade separated 4A) or only in T. castaneum (37%, column 4 of Fig. 4A). by an evolutionary distance similar or larger than the We found no strong indication that the gene inventory one separating flies and beetles (i.e., around 370 million required for a process would be more conserved for years). Further, it shows that T. castaneum is an ex- those processes, which seem more conserved morpho- tremely useful screening system for insect biology, able logically. For instance, dorso-ventral patterning, which to reveal novel gene functions even in processes that we assumed representing an intermediate degree of con- have been studied intensely in D. melanogaster. servation showed the largest common gene set while the supposedly most conserved process, muscle formation, Estimating the portions of gene functions revealed in fly showed the lowest value. However, we note that we versus beetle found more Tribolium-specific gene functions for the In order to make a comprehensive and quantitative less conserved processes than for muscle development. comparison, we included in our comparisons all genes However, given the uncertainties with these numbers that are currently known to be involved in the respective (see the “Discussion” section below) and the fact that processes from both, beetle and fly. Our beetle data are morphological conservation of a process is hard to quan- based on both, our systematic screening of 51% of the T. tify, we hesitate to draw conclusions about the correl- castaneum gene set and on previous candidate gene ation of divergence of a biological process and the work. With respect to fly data, we rely on information involved gene inventory. available on FlyBase and our expert knowledge of the While these data were gained with respect to develop- processes under scrutiny. Given these different kinds of mental processes only, they strongly indicate that our sources and approaches, and the fact that we focus on current knowledge based on screening in one species ap- developmental processes, the data are – despite the pears to be much less comprehensive than previously comprehensive approach - prone to various types of un- thought. We believe that the different proportions of certainties. In the following, we first discuss the way we genes shown to be required for a specific process in only combined the numbers to calculate our estimation. Fig. 4 Many genes involved in a given process are detected only in one of the two species. We combined all genes found in the fly to be involved in our processes (column 1) and/or those genes that we identified in the iBeetle screen to be required in the same process (column 3) to assemble a set of genes comprising all genes currently known to be required in any insect for the processes analyzed here (column 4). Of the fly gene set (column 1) about two thirds had been tested in the iBeetle screen. Of those, three quarters showed a similar function in our beetle while one quarter appeared to be fly specific (column 2). The subdivisions of columns 1 and 2 are based on Figs. 2 and 3 and Additional file 2: Fig. S1. From the numbers in columns 1-3 we calculated the portions of genes of the combined insect gene set (column 4), which were detected only in Drosophila (11%), only in Tribolium (37%), or in both (52%). See text for details and discussion of potential systematic biases. B) Respective values for the single processes show that the Tribolium screening platform revealed 20–50% novel genes relevant for a process (i.e., which were not detected in Drosophila). See Additional file 5: Table S4 for calculations. Given these results neither model system can be used alone as a proxy for insects or protostomes in general and that Tribolium is a very useful complementary screening platform Hakeemi et al. BMC Biology (2022) 20:38 Page 7 of 13 Fig. 5 Examples for novel gene functions detected in Tribolium but not known from Drosophila. A Anterior part of a wildtype cuticle with head, thorax, and anterior abdomen. B In iB_03355 knock-down embryos, dorso-ventral patterning was disturbed such that the embryo has turned inside out, i.e. the legs and the head are located inside the trunk cuticle instead of outside. C In iB_04199 knock-down, anterior epidermal patterning was disrupted to different degrees. In mild phenotypes, just the most anterior part, the labrum, was affected (not shown), intermediate phenotypes lacked head and parts of the thorax (shown) and in strong phenotypes, only cuticle crumbs remained. D In the transgenic strain pBA19, the muscles are marked with EGFP. They are visible in vivo as elongated structures with a segmentally repeated pattern. E In iB_01159 (Tc-Unc-76 ) knock-down, the muscles were partially missing or detached such that some muscles adopted a rounded shape. F In wildtype ovaries, the nurse cells are located in the tropharium forming an elongated structure (marked by a white line). The first part of the vitellarium is marked by active cell division (marked here by phospho-histone 3 staining, PH3) and along the entire vitellarium, the oocytes increase in size (compare stars). G In iB_10431 knock-down ovaries, the tropharia were normal (white line) but no oocytes developed. The white structure is not part of the ovaries. Anterior is to the top in A–E, the pictures of the phenotypes are modified from iBeetle-Base. Scale bars are 100 μm Hakeemi et al. BMC Biology (2022) 20:38 Page 8 of 13 Subsequently, we will discuss some uncertainties and in specific genes without conserved functions likely is even how far they influence the estimation. higher than reflected in Fig. 4A. Of the genes known from D. melanogaster to be re- Second, we found quite different numbers for the four quired for the processes investigated here (n = 132; see processes under scrutiny (Fig. 4B). However, even in the Additional file 5: Table S4), we could compare 66% to process with the lowest portion of genes detected exclu- iBeetle data (column 1 in Fig. 4A; based on Additional sively in T. castaneum (muscle development), this por- file 2: Fig. S1; n = 87). Of those genes, 26% (n = 23) were tion was 21%, which still indicates a significant degree of not required for that process in T. castaneum (column 2 unexplored biology. in Fig. 4A; based on Fig. 3). With this statement, we Third, the D. melanogaster numbers could be influ- mean that the respective genes did not have any pheno- enced by false negative data. The data on FlyBase has type with respect to the biological process in question. not been gathered in one or few standardized screens They could have either no phenotype or a phenotype af- where all data were published—it is mainly based on fecting another process. Based on our positive controls, published results of single gene analyses. However, not the potential error affecting this statement is less than all genetic screens have reached saturation and not all 7.5% (see Additional file 1: Table S1). For our overall es- genes detected in large-scale screens may have been fur- timation, we extrapolated this share to the total number ther analyzed and published. Hence, the number of of genes required for the fly (dotted lines from column 2 genes in principle detectable in D. melanogaster might to column 4). A number of gene functions detected in actually be larger than the numbers extracted from Fly- the iBeetle screen had not been assigned such functions Base. In the iBeetle screen, in contrast, negative data was in D. melanogaster before (column 3 in Fig. 4A; based systematically documented, such that this type of uncer- on Fig. 2). When combining these numbers, we aimed at tainty is restricted to technical false negative data, which providing a minimum estimation for the divergence of we found to be around 15% in this first pass screen (Fig. detected gene functions (Column 4 in Fig. 4A). To be 1). This uncertainty could potentially increase the por- conservative, we assumed that all gene functions known tion of D. melanogaster-specific or conserved genes. from D. melanogaster but not yet tested in the iBeetle Fourth, theoretically, there may be false positive data al- screen would fall into the class of genes being required beit restricted to the set of genes detected in both spe- for both species (see numbers in green square in Add- cies. The reason is that iBeetle was a first pass screen, itional file 5: Table S4). Further, we scored each signal- where we aimed at reducing false negative data with the ing pathway as one case (finding mostly conservation) tradeoff that false positive data are enriched [17]. Al- even if single components of these pathways did not though finding similar phenotypes in two different spe- have divergent phenotypes. This conservative assump- cies will not in many cases be false positive, we tried to tion leads to the abovementioned minimum estimation minimize this error by manually checking the annota- of divergence in these gene sets (Column 4 in Fig. 4A; tions of the respective genes, excluding those that calculation in Additional file 5: Table S4). Of all genes showed a phenotype with low penetrance or in combin- currently known to be required for one of the processes ation with many other defects indicating a non-specific we studied, the portion of genes detected exclusively in effect. Of note, the issue of false positives is restricted to the fly (11%; n = 23) is much smaller than the one de- the genes detected in both species (column 2; based on tected only in the beetle (37%; n = 76) while the analo- Fig. 3). It does not apply to those genes detected only in gous function of half of the genes (52%; n =109)is the beetle but not the fly (column 3; based on Fig. 2) be- detected in both species. cause in this case, all phenotypes were confirmed by in- With this work, we present the first and a quite exten- dependent experiments with non-overlapping dsRNA sive dataset to estimate this kind of numbers. Still, some fragments in different genetic backgrounds such that confounding issues need to be considered. The first un- false positive results are excluded. In summary, while certainty stems from the fact that the beetle data is there are a number of uncertainties that we could not based on testing about 50% of the genes. In the second clarify with available data or methods, most of these un- part of the screen, we had prioritized genes that were certainties hint at underestimation rather than overesti- moderately to highly expressed, showed sequence con- mation of functional divergence between fly and beetle. servation, and had GO annotations. The prioritization Our work focused on developmental processes with apparently was successful as 66% of the gene functions different grades of assumed conservation and different known from D. melanogaster had been covered in the grades of previous knowledge. Morphologically, the iBeetle screen (Fig. 4A), which is much more than the muscle pattern and general development appear to be a 40% expected for an unbiased selection [17]. Hence, our quite conserved between these insects [31, 49] compared figures might be biased towards conserved gene func- to oogenesis where a number of morphological differ- tion. As a consequence, the overall portion of beetle- ences were described [40–42]. Given the background of Hakeemi et al. BMC Biology (2022) 20:38 Page 9 of 13 a strongly derived head morphology of first instar larvae were subsequently tested by RNAi lines in D. melanoga- but conserved adult heads and brains, both conservation ster where four of them indeed showed a related pheno- and divergence were found with respect to the genetic type. Likewise, some wing blister genes from D. control [6, 36, 50, 51]. Likewise, dorso-ventral patterning melanogaster were not annotated in the iBeetle screen. is relying on both, conserved and diverged gene regula- When we checked more specifically, this was often due tory networks [14, 52, 53]. Taken together, our selection to the lethality of the animal before the formation of appears to cover both conserved and diverged processes wings [17]. When we varied the timing of injection, two such that—at least for the genetic control of develop- of those knock-downs elicited wing blister phenotypes ment—our data can be generalized with some also in T. castaneum [17]. These data show that details confidence. of the screening procedure influence the subset of genes that are detected. Technical characteristics contribute to the detection of unequal gene sets Evolutionary divergence of gene function and derived Our numbers reveal that functionally comparable gene Drosophila biology may be larger than appreciated sets in two quite closely related model systems are far Most relevant for the field of functional genetics is our from identical. A question of obvious biological rele- conclusion that the degree of divergence of gene func- vance but not easily resolved is: to which degree do tions among holometabolous insects is larger than previ- these differences reflect the biologically meaningful di- ously assumed. Therefore, some genes are detected only vergence of gene functions, or alternatively, simply result in one species because the gene’s function is not re- from technical problems, i.e., reflect different strengths quired for that process in the other. This finding should and weaknesses of the respective screening methods and of course influence our thinking about using any insect model systems? as model for human development and diseases such as As discussed above, some degree of false negative data muscle fomation and congenital heart defect. may be expected in both model systems. In the case of Indeed, there is evidence supporting the notion of an the iBeetle screen, this will be restricted to technical unexpected degree of divergence with respect to muscle false negative data. In the D. melanogaster field, there development. Based on the iBeetle screen, a number of may be additional false negative data due to the lack of muscle genes identified in the iBeetle screen were more saturation of screens and/or lack of reporting of genes closely investigated in D. melanogaster [31, 49]. Despite that were not studied in detail. However, given the ex- quite some efforts, the negative data for fly orthologs ap- tent and comprehensiveness of work in the D. melano- peared to be true negative. For example, null mutations gaster field, we feel that this might not be of high of one of the genes found in our beetle, nostrin, did not relevance. As to the different strengths of screening pro- elicit a phenotype in D. melanogaster unless combined cedures, it is certainly true that the way screens are per- with a mutation of a related F-bar protein Cip4. Like- formed influences what sets of genes can be detected. wise, Rbm24 displays strong RNAi and mutant pheno- For instance, our parental RNAi approach knocked types in T. castaneum and vertebrates, respectively, but down both, maternal and zygotic contributions while D. melanogaster is lacking an Rbm24 ortholog, and func- some classic D. melanogaster screens affected only the tional compensation by paralogs was suggested to occur zygotic contribution. Hence, genes where maternal con- during D. melanogaster muscle development. Other tribution rescues the embryonic phenotype are easily genes including kahuli and unc-76 are expressed in the missed in the fly but not the beetle. For instance, paren- D. melanogaster mesoderm but only showed very subtle tal RNAi knocking down components of the aPKC com- somatic muscle phenotypes, if any, in Mef2-GAL4 driven plex leads to severe early disruption of embryogenesis in RNAi experiments or with CRISPR/Cas9 induced muta- T. castaneum while in respective D. melanogaster mu- tions, respectively (see Materials & Methods). By con- tants almost no defects are seen on the cuticle level (A. trast, their beetle counterparts had strong and penetrant Wodarz, unpublished observation). Conversely, our phenotypes in single knock-downs (e.g. see Tc-unc-76 in RNAi screen depended on the accuracy of gene annota- Fig. 5E) [31, 49, 54]. These data suggest that the function tions and our approach of screening for several pro- of genes or their relative contribution to this biological cesses in parallel may have reduced detection sensitivity. process has changed significantly. They also indicate that One striking example of the different strengths of the single gene view may be limited. Phenotypes depend screening designs is provided by wing blister phenotypes. on networks of interacting genes and this may allow for In the first part of the iBeetle screen, we detected 34 changes and replacements of individual components genes showing wing blister phenotypes where 14 did not while the overall network structure is maintained. There have related GO term annotation at FlyBase and 5 did are more striking examples of gene function changes. not have any GO annotation at all. Seven of these genes The gene germ cell-less was detected in the iBeetle Hakeemi et al. BMC Biology (2022) 20:38 Page 10 of 13 screen to govern anterior-posterior axis formation in the penetrance documented. For subsequent work, we only beetle while in D. melanogaster it is required for the for- considered phenotypes, which showed a penetrance of > mation of the posterior germ-cells [51]. Also, the D. mel- 50%. dsRNAs (1 μg/μl) were produced by Eupheria Bio- anogaster textbook example of a developmental tech Dresden, Germany. Different from the published morphogen bicoid does not even exist in T. castaneum procedure, the stink gland analysis was performed 21 [5] and yet other genes were found to act as anterior de- days after pupal injection (in the first screening phase, terminants in other flies [9, 10]. Along the same lines, this analysis had been performed after larval injection). the genes forkhead and buttonhead do not appear to be required for anterior patterning in T. castaneum but are Controls of the screen essential in flies [12, 39, 55, 56]. To assess the sensitivity and reliability of the screen, and These findings with respect to specific genes add to a to test the accuracy of each screener, we included ap- number of observations arguing for a comparatively high proximately 5% positive controls randomly chosen from degree of divergence due to the overall highly derived a set of 35 different genes. By and large, we used the nature of fly biology. The number of genes is much same positive controls as in the first screening phase smaller in D. melanogaster (appr. 14,000) compared to (see Table Table_S1_controls). However, Tc-zen-1 was T.castaneum (appr. 16,500). Further, a number of devel- excluded since the phenotypes were much weaker than opmental processes are represented in a more insect- in the previous screen, probably due to the degradation typical way in T. castaneum like for instance segmenta- of the dsRNA. We added new positive controls to score tion [57], head [50] and leg development, brain develop- for muscle and stink-gland phenotypes, which we took ment [58], extraembryonic tissue movements [59], and from novel genes detected in the first screening phase. mode of metamorphosis [60]. In most cases, the situ- New controls for muscle phenotypes: iB_06061, iB_ ation in the fly is simplified and appears to be stream- 05796, iB_03227, iB_01705; for stink gland and ovary lined for faster development. We think that these phenotypes: iB_02517; for head defects: iB_05442 (that biological differences might be the basis for divergence gene was not scored for its stink gland phenotype be- in gene function, which we just started to uncover. In cause it turned out to be too mild to be identified reli- the absence of similar large-scale comparisons in other ably in high throughput). In 143 cases (80.8%, n = 177), species, it remains open, whether an insect-typical gene the phenotypes of positive controls were fully recognized set even exists or whether one would rather have to (for comparison: in the first screening phase the respect- emphasize a constant change of gene function, such that ive numbers were: 90%, n = 201). In 14 cases (7.9%; any ancestral gene set simply “melts” away with evolu- phase 1: 4%) the phenotype was partially recognized. tionary time. This category includes complex phenotypes where half (one of two aspects: knirps, piwi, SCR, cta, cnc, iB_ Conclusion 01705, iB_05442) or two of three aspects (aristaless)of We found that the gene sets detected for the same pro- all phenotypic aspects were correctly identified. 13 phe- cesses in flies and beetles differ much more than ex- notypes were missed completely (7.3%, phase 1: 4% ). pected—only about 50% of the genes were detected in Tc-metoprene tolerant (Tc-met) was missed most fre- both species. Given the large divergence of gene sets quently, probably due to the fact that the embryonic leg found in different screening systems and the docu- phenotype was very subtle and in addition, the pene- mented cases of biological divergence of gene function, trance of the phenotype appeared to be lower than in we propose that a more systematic investigation on the the first screen (penetrance: less than 30%). Seven posi- divergence of gene function is needed. Further, the hy- tive controls (4%, phase 1: 1%) could not be analyzed pothesis independent screening now possible in T. casta- due to prior technical lethality, i.e. the premature death neum may be very helpful in that endeavor. of the injected pupae prevented the detection of the phenotype. In three cases wrong aspects were annotated Methods (false positive: 1.7%). Depending on the other annota- Screen tions these positive controls were valued as partially rec- We followed the tested and published procedures apart ognized (SCR) or missed (met, CTA). Find more details from some minor changes detailed below (please find an in Table Table_S1_controls. extensive description of the procedure in Schmitt-Engel Negative controls (buffer injections) were mainly et al. [17]). In particular, we used the same strains, injec- annotated correctly (no phenotype in 92.9%; phase 1: tion procedures, and incubation temperatures and incu- 96%) and just in 7 cases led to false positive annota- bation times. We injected 10 pupae per gene leading to tions (7.1%; phase 1: 2%) (Table Table_S1_controls; the collection of > 50 offspring L1 cuticles, which were sheet 2). analyzed. Phenotypes were annotated and their Hakeemi et al. BMC Biology (2022) 20:38 Page 11 of 13 Re-screen neoFRT}19A embryos. Single lines established from the Re-screening of selected iBeetle candidates was per- offspring were tested as heterozygotes over the balancer formed in order to probe for off-target and strain- FM7c. We used a T7 endonuclease assay for detecting specific effects. For that purpose, two independent sequence alterations near the target site as described in dsRNA fragments (original iB-fragments and one non- [62]. Our lethal Unc-76[CR007] allele carries a 16 nu- overlapping fragment, both at concentration 1 μg/μl) tar- cleotide deletion near the target site in the sequence geting the same gene were injected separately into a dif- ..TAT CCA CAC ACc aac ggt ttg gga tcc GGA TCC ferent genetic background (San Bernardino, SB strain), GGA TCC.. of the second exon (X: 2091152... 2091167, except for the muscle project where it is required to use r6.32; lower case letters represent the deleted DNA) that the pBA19 strain with EGFP marked muscles. The rest creates a frameshift in the ORF of all known isoforms of the injection procedures and analyses were performed (i.e., the frameshift occurs after T246 in Unc-76 RA to like in the screen. Note that with this approach, we can- -C and after T61 in Unc-76 RD). not exclude that phenotypes observed in one tissue are elicited by knock-down in another tissue (e.g., hormone- Supplementary Information induced morphogenesis may fail due to knock-down of The online version contains supplementary material available at https://doi. org/10.1186/s12915-022-01231-4. hormone production in a gland). Additional file 1: Table S1. Positive and negative controls of the Fly gene sets screen. Lists of genes involved in those processes were estab- Additional file 2: Figure S1. Diagram displaying the portion of genes lished by our experts of the respective processes. This known to be required for the processes in Drosophila, which were tested was supported by individual FlyBase searches for re- in Tribolium. Figures S2 and S3. Phylogenetic trees supporting our claim of absence of an ortholog in Drosophila. spective GO terms of the category “biological process”. Additional file 3: Table S2. Genes with phenotype in Tribolium but not For the analyses in Figs. 2 and 3, we only considered Drosophila. gene functions, which were based on experimental data Additional file 4: Table S3. Genes with phenotype in Drosophila as documented at FlyBase in the year 2017 (Dmel Re- compared to Tribolium data. lease 6.18) with updates in single case in 2020 (Dmel Re- Additional file 5: Table S4. Combined analysis leading to the numbers lease 6.32). given in Figure 4. Phylogenetic analysis Acknowledgements The Tribolium protein sequences from gene set OGS3_ We thank Mohamad Al Heshan, Elke Küster, and Claudia Hinners for help with injection and processing during the screen. proteins.fasta.gz (including changes from 2016/02/15, available from [27]) were used to retrieve the most simi- Authors’ contributions lar proteins of T. castaneum, D. melanogaster, and M. MSH, SA, MT, MW, DG, TK, and XW performed high throughput RNAi musculus using only one isoform. Multiple alignments screening and annotation of phenotypes. DG and TK led the screening were done with the ClustalOmega plugin as imple- teams. MK, MS, and GB led the screen. JD provided bioinformatics support and developed iBeetle-Base. DS did follow-up research in Drosophila includ- mented in the Geneious 10.1.3 software (Biomatters, ing the generation of mutants. JS, DG, JD, MSH, and GB gathered and inter- Auckland, New Zealand) using standard settings. Align- preted information on differences in gene function between beetle and fly. ments were trimmed to remove poorly aligned sequence GB and MSH wrote the first draft of the manuscript and prepared the figures. MF, SF, MS, and MK supervised the analyses, interpreted data, and edited the stretches. Phylogenetic trees were calculated using the manuscript. All authors have read and approved the final manuscript. FastTree 2.1.5 plugin implemented in Geneious. Funding Generation of Unc-76 mutations via CRISPR/Cas9 Deutsche Forschungsgemeinschaft (DFG) funded the research unit FOR1234 In order to generate the Unc-76 mutations, we essen- “iBeetle” including the following grants: BU 1443/6-2; BU 1443/7-2; Kl 656/7- 2; Bu 1443/8-2; Kl 656/8-2; Scho 1058/4-2; FR 696/4-2; RO 890/4-2; STA 1009/ tially followed the procedure described by Basset et al., 10-1; iBeetle-Base was supported by DFG LIS project 417202192 2013 [61]—please find an extensive description there. Bayer CropScience funded part of the continuation of the screen. For making the template for the guide RNAs, the Unc- The China Scholarship Council funded Xuebein Wan (201706760058). The funding bodies had no influence on the project’s design and 76 target sequence GGTTCAACGATCTGACCAGTG publication strategy. Open Access funding enabled and organized by Projekt was inserted between the T7 promoter and the gRNA DEAL. core sequence in the forward primer, gRNA_F. After an- nealing gRNA_F with SGRNAR, the template was PCR Availability of data and materials amplified with Q5 polymerase (NEB). Guide RNAs were The datasets generated and/or analyzed during the current study are available from the iBeetle-Base [25, 26] repository [27]. transcribed with Ampliscribe T7 Flash (epicenter), iso- The dsRNA fragments used to knock down the genes are commercially lated with the MEGAclear kit (Ambion), and injected to- available from Eupheria Biotech, their sequences are documented in the gether with Cas9 mRNA into w[1118] sn[3] P{ry+t7.2= iBeetle-Base. Hakeemi et al. BMC Biology (2022) 20:38 Page 12 of 13 Declarations 14. Lynch JA, Roth S. The evolution of dorsal–ventral patterning mechanisms in insects. Genes Dev. 2011;25(2):107–18. https://doi.org/10.1101/gad.2010711. Ethics approval and consent to participate 15. Stappert D, Frey N, von Levetzow C, Roth S. Genome-wide identification of Not applicable Tribolium dorsoventral patterning genes. Development. 2016;143(13):2443– 54. https://doi.org/10.1242/dev.130641. 16. Bucher G, Scholten J, Klingler M. Parental RNAi in Tribolium (Coleoptera). Consent for publication Curr Biol. 2002;12(3):R85–6. https://doi.org/10.1016/S0960-9822(02)00666-8. Not applicable 17. Schmitt-Engel C, Schultheis D, Schwirz J, Strohlein N, Troelenberg N, Majumdar U, et al. The iBeetle large-scale RNAi screen reveals gene Competing interests functions for insect development and physiology. Nat Commun. 2015;6(1): The authors declare that they have no competing interests. 7822. https://doi.org/10.1038/ncomms8822. 18. Tomoyasu Y, Denell RE. Larval RNAi in Tribolium (Coleoptera) for analyzing Author details 1 adult development. Dev Genes Evol. 2004;214(11):575–8. https://doi.org/10.1 Johann-Friedrich-Blumenbach-Institut, GZMB, University of Göttingen, 2 007/s00427-004-0434-0. Justus-von-Liebig-Weg 11, 37077 Göttingen, Germany. Current address: 19. Berghammer AJ, Klingler M, Wimmer EA. A universal marker for transgenic Institute of Clinical Pharmacology, University Medical Center Göttingen, insects. Nature. 1999;402(6760):370–1. https://doi.org/10.1038/46463. University of Göttingen, Robert-Koch-Str. 40, 37075 Göttingen, Germany. 20. Gilles AF, Schinko JB, Averof M. Efficient CRISPR-mediated gene targeting Department of Biology, Friedrich-Alexander-University Erlangen-Nürnberg, and transgene replacement in the beetle Tribolium castaneum. Staudtstr. 5, 91058 Erlangen, Germany. Current address: Department of Development. 2015;142(16):2832–9. https://doi.org/10.1242/dev.125054. Medical Bioinformatics, University Medical Center Göttingen, University of 21. Schinko JB, Weber M, Viktorinova I, Kiupakis A, Averof M, Klingler M, et al. Göttingen, Goldschmidtstr. 1, 37077 Göttingen, Germany. Current address: Functionality of the GAL4/UAS system in Tribolium requires the use of Department of Insect Biotechnology, Justus-Liebig-University Giessen, endogenous core promoters. BMC Dev Biol. 2010;10(1):53. https://doi.org/1 Heinrich-Buff-Ring 26-32, 35392 Gießen, Germany. Current address: Institute 0.1186/1471-213X-10-53. of Pathology, University Hospital Cologne, Kerpener Str. 62, 50924 Cologne, 22. Herndon N, Shelton J, Gerischer L, Ioannidis P, Ninova M, Dönitz J, et al. Germany. Current address: Institute of Neuropathology, University Hospital Enhanced genome assembly and a new official gene set for Tribolium Erlangen, Friedrich-Alexander-University Erlangen-Nürnberg, Erlangen, castaneum. BMC Genomics. 2020;21(1):47. https://doi.org/10.1186/s12864-01 Germany. Institute for Zoology/Developmental Biology, University of 9-6394-6. Cologne, Biocenter, Zülpicher Straße 47b, D-50674 Köln, Germany. 23. Mungall CJ, Gkoutos GV, Smith CL, Haendel MA, Lewis SE, Ashburner M. Integrating phenotype ontologies across multiple species. Genome Biol. Received: 15 July 2021 Accepted: 12 January 2022 2010;11(1):R2. https://doi.org/10.1186/gb-2010-11-1-r2. 24. Dönitz J, Grossmann D, Schild I, Schmitt-Engel C, Bradler S, Prpic N-M, et al. TrOn: An Anatomical Ontology for the Beetle Tribolium castaneum. PLoS References ONE. 2013;8(7):e70695. https://doi.org/10.1371/journal.pone.0070695. 1. Jorgensen EM, Mango SE. The art and design of genetic screens: 25. Dönitz J, Schmitt-Engel C, Grossmann D, Gerischer L, Tech M, Schoppmeier caenorhabditis elegans. Nat Rev Genet. 2002;3(5):356–69. https://doi.org/10.1 M, et al. iBeetle-Base: a database for RNAi phenotypes in the red flour 038/nrg794. beetle Tribolium castaneum. Nucl Acids Res. 2015;43(D1):D720–5. https:// 2. Kile BT, Hilton DJ. The art and design of genetic screens: mouse. Nat Rev doi.org/10.1093/nar/gku1054. Genet. 2005;6(7):557–67. https://doi.org/10.1038/nrg1636. 26. Dönitz J, Gerischer L, Hahnke S, Pfeiffer S, Bucher G. Expanded and updated 3. Patton EE, Zon LI. The art and design of genetic screens: zebrafish. Nat Rev data and a query pipeline for iBeetle-Base. Nucleic Acids Res. 2018;46(D1): Genet. 2001;2(12):956–66. https://doi.org/10.1038/35103567. D831–5. https://doi.org/10.1093/nar/gkx984. 4. St Johnston D. The art and design of genetic screens: Drosophila 27. iBeetle Consortium. iBeetle-Base: A Database of Tribolium Genes, Genomes melanogaster. Nat Rev Genet. 2002;3(3):176–88. https://doi.org/10.1038/ and RNAi phenotypes. http://ibeetle-base.uni-goettingen.de/. Accessed 19. nrg751. Nov 2018. 5. Brown S, Fellers J, Shippy T, Denell R, Stauber M, Schmidt-Ott U. A strategy 28. Bier E, Bodmer R. Drosophila, an emerging model for cardiac disease. Gene. for mapping bicoid on the phylogenetic tree. Curr Biol. 2001;11(2):R43–4. 2004;342(1):1–11. https://doi.org/10.1016/j.gene.2004.07.018. https://doi.org/10.1016/S0960-9822(01)00007-0. 29. Ahmad SM. Conserved signaling mechanisms in Drosophila heart 6. Fu J, Posnien N, Bolognesi R, Fischer TD, Rayl P, Oberhofer G, et al. development. Dev Dyn. 2017;246(9):641–56. https://doi.org/10.1002/dvdy.24530. Asymmetrically expressed axin required for anterior development in 30. Johnson MS, Lu N, Denessiouk K, Heino J, Gullberg D. Integrins during Tribolium. Proc Natl Acad Sci USA. 2012;109(20):7782–6. https://doi.org/10.1 evolution: evolutionary trees and model organisms. Biochim Biophys Acta. 073/pnas.1116641109. 1788;2009(4):779–89. https://doi.org/10.1016/j.bbamem.2008.12.013. 7. Glinka A, Wu W, Delius H, Monaghan AP, Blumenstock C, Niehrs C. Dickkopf- 31. Schultheis D, Schwirz J, Frasch M. RNAi screen in tribolium reveals 1 is a member of a new family of secreted proteins and functions in head involvement of F-BAR proteins in myoblast fusion and visceral muscle induction. Nature. 1998;391(6665):357–62. https://doi.org/10.1038/34848. morphogenesis in insects. G3 (Bethesda). 2019;9:1141–51. 8. Gurley KA, Rink JC, Sánchez AA. Beta-catenin defines head versus tail 32. Theis JL, Vogler G, Missinato MA, Li X, Nielsen T, Zeng X-XI, et al. Patient- identity during planarian regeneration and homeostasis. Science. 2008; specific genomics and cross-species functional analysis implicate LRP2 in 319(5861):323–7. https://doi.org/10.1126/science.1150029. hypoplastic left heart syndrome. eLife. 2020;9:e59554. https://doi.org/10. 9. Klomp J, Athy D, Kwan CW, Bloch NI, Sandmann T, Lemke S, et al. Embryo 7554/eLife.59554. development. A cysteine-clamp gene drives embryo polarity in the midge 33. Zaffran S, Frasch M. Early signals in cardiac development. Circ Res. 2002; Chironomus. Science (New York, NY). 2015;348:1040–2. 91(6):457–69. https://doi.org/10.1161/01.RES.0000034152.74523.A8. 10. Yoon Y, Klomp J, Martin-Martin I, Criscione F, Calvo E, Ribeiro J, et al. 34. Niehrs C. On growth and form: a Cartesian coordinate system of Wnt and Embryo polarity in moth flies and mosquitoes relies on distinct old genes BMP signaling specifies bilaterian body axes. Development. 2010;137(6):845– with localized transcript isoforms. Elife. 2019;8:e46711. https://doi.org/10. 57. https://doi.org/10.1242/dev.039651. 7554/eLife.46711. 35. Finkelstein R, Boncinelli E. From fly head to mammalian forebrain: the story 11. Averof M. Arthropod Hox genes: insights on the evolutionary forces that of otd and Otx. Trends Genet. 1994;10(9):310–5. https://doi.org/10.1016/01 shape gene functions. Curr Opin Genet Dev. 2002;12(4):386–92. https://doi. 68-9525(94)90033-7. org/10.1016/S0959-437X(02)00314-3. 12. Kittelmann S, Ulrich J, Posnien N, Bucher G. Changes in anterior head 36. Posnien N, Koniszewski NDB, Hein HJ, Bucher G. Candidate Gene Screen in patterning underlie the evolution of long germ embryogenesis. Dev Biol. the Red Flour Beetle Tribolium Reveals Six3 as Ancient Regulator of Anterior 2013;374(1):174–84. https://doi.org/10.1016/j.ydbio.2012.11.026. Median Head and Central Complex Development. PLoS Genet. 2011;7(12): 13. Kitzmann P, Weißkopf M, Schacht MI, Bucher G. A key role forfoxQ2in e1002418. https://doi.org/10.1371/journal.pgen.1002416. anterior head and central brain patterning in insects. Development. 2017; 37. Hunnekuhl VS, Akam M. An anterior medial cell population with an apical- 144(16):2969–81. https://doi.org/10.1242/dev.147637. organ-like transcriptional profile that pioneers the central nervous system in Hakeemi et al. BMC Biology (2022) 20:38 Page 13 of 13 the centipede Strigamia maritima. Dev Biol. 2014;396(1):136–49. https://doi. 59. Panfilio KA. Extraembryonic development in insects and the acrobatics of org/10.1016/j.ydbio.2014.09.020. blastokinesis. Developmental Biology. 2008;313(2):471–91. https://doi.org/1 38. Leclère L, Bause M, Sinigaglia C, Steger J, Rentzsch F. Development of the 0.1016/j.ydbio.2007.11.004. aboral domain in Nematostella requires β-catenin and the opposing 60. Snodgrass R. Insect Metamorphosis: Smithsonian Miscellaneous Collections, activities of Six3/6 and Frizzled5/8. Development. 2016;143(10):1766–77. V122, No. 9. Literary Licensing: Washington; 1954. https://doi.org/10.1242/dev.120931. 61. Bassett AR, Tibbit C, Ponting CP, Liu J-L. Highly efficient targeted mutagenesis of Drosophila with the CRISPR/Cas9 system. Cell Rep. 2013;4(1): 39. Schinko JB, Kreuzer N, Offen N, Posnien N, Wimmer EA, Bucher G. 220–8. https://doi.org/10.1016/j.celrep.2013.06.020. Divergent functions of orthodenticle, empty spiracles and buttonhead 62. Kondo S, Ueda R. Highly improved gene targeting by germline-specific Cas9 in early head patterning of the beetle Tribolium castaneum expression in Drosophila. Genetics. 2013;195(3):715–21. https://doi.org/10.1 (Coleoptera). Dev Biol. 2008;317(2):600–13. https://doi.org/10.1016/j. 534/genetics.113.156737. ydbio.2008.03.005. 40. Trauner J, Büning J. Germ-cell cluster formation in the telotrophic meroistic ovary of Tribolium castaneum (Coleoptera, Polyphaga, Tenebrionidae) and Publisher’sNote its implication on insect phylogeny. Dev Genes Evol. 2007;217(1):13–27. Springer Nature remains neutral with regard to jurisdictional claims in https://doi.org/10.1007/s00427-006-0114-3. published maps and institutional affiliations. 41. Teuscher M, Ströhlein N, Birkenbach M, Schultheis D, Schoppmeier M. TC003132 is essential for the follicle stem cell lineage in telotrophic Tribolium oogenesis. Front Zool. 2017;14(1):26. https://doi.org/10.1186/s12983-017-0212-2. 42. Bäumer D, Strohlein NM, Schoppmeier M. Opposing effects of Notch- signaling in maintaining the proliferative state of follicle cells in the telotrophic ovary of the beetle Tribolium. Front Zool. 2012;9(1):15. https:// doi.org/10.1186/1742-9994-9-15. 43. FlyBase Consortium. FlyBase - A Database of Drosophila Genes & Genomes. http://flybase.org/. Accessed 19. Nov 2018 44. Gelbart WM, Crosby MA, Matthews B, Rindone WP, Chillemi J, Twombly SR, et al. FlyBase: a Drosophila database. The FlyBase consortium. Nucleic Acids Res. 1997;25(1):63–6. 45. Kitzmann P, Schwirz J, Schmitt-Engel C, Bucher G. RNAi phenotypes are influenced by the genetic background of the injected strain. BMC Genomics. 2013;14(1):5. https://doi.org/10.1186/1471-2164-14-5. 46. Lorenzen MD, Kimzey T, Shippy TD, Brown SJ, Denell RE, Beeman RW. piggyBac-based insertional mutagenesis in Tribolium castaneum using donor/helper hybrids. Insect Mol Biol. 2007;16(3):265–75. https://doi.org/1 0.1111/j.1365-2583.2007.00727.x. 47. Carbon S, Ireland A, Mungall CJ, Shu S, Marshall B, Lewis S, et al. AmiGO: online access to ontology and annotation data. Bioinformatics. 2009;25(2): 288–9. https://doi.org/10.1093/bioinformatics/btn615. 48. GO Consortium. Gene Ontology. Database http://geneontology.org/. Accessed 19. Nov 2018. 49. Schultheis D, Weißkopf M, Schaub C, Ansari S, Dao VA, Grossmann D, et al. A Large Scale Systemic RNAi Screen in the Red Flour Beetle Tribolium castaneum Identifies Novel Genes Involved in Insect Muscle Development. G3 (Bethesda). 2019;9:1009–26. 50. Posnien N, Schinko JB, Kittelmann S, Bucher G. Genetics, development and composition of the insect head - A beetle’s view. Arthropod Struct Dev. 2010;39(6):399–410. https://doi.org/10.1016/j.asd.2010.08.002. 51. Ansari S, Troelenberg N, Dao VA, Richter T, Bucher G, Klingler M. Double abdomen in a short-germ insect: Zygotic control of axis formation revealed in the beetle Tribolium castaneum. PNAS. 2018;115:201716512. 52. Nunes da Fonseca R, von Levetzow C, Kalscheuer P, Basal A, van der Zee M, Roth S. Self-regulatory circuits in dorsoventral axis formation of the short- germ beetle Tribolium castaneum. Dev Cell. 2008;14:605–15. 53. Pechmann M, Kenny NJ, Pott L, Heger P, Chen Y-T, Buchta T, et al. Striking parallels between dorsoventral patterning in Drosophila and Gryllus reveal a complex evolutionary history behind a model gene regulatory network. Elife. 2021;10:e68287. https://doi.org/10.7554/eLife.68287. 54. Schultheis D. Identifizierung und Charakterisierung neuer regulatorischer Gene in der Muskelentwicklung durch einen genomweiten RNAi-Screen in Tribolium castaneum; 2016. 55. Weigel D, Jürgens G, Kuttner F, Seifert E, Jäckle H. The homeotic gene fork head encodes a nuclear protein and is expressed in the terminal regions of the Drosophila embryo. Cell. 1989;57(4):645–58. https://doi.org/10.1016/ 0092-8674(89)90133-5. 56. Wimmer EA, Cohen SM, Jackle H, Desplan C. Buttonhead does not contribute to a combinatorial code proposed for Drosophila head development. Development. 1997;124(8):1509–17. https://doi.org/10.1242/dev.124.8.1509. 57. Tautz D, Friedrich M, Schröder R. Insect embryogenesis – what is ancestral and what is derived? Development. 1994;1994(Supplement):193–9. 58. Farnworth MS, Eckermann KN, Bucher G. Sequence heterochrony led to a gain of functionality in an immature stage of the central complex: a fly- beetle insight. 2020;18(10). https://doi.org/10.1371/journal.pbio.3000881.

Journal

BMC BiologySpringer Journals

Published: Feb 8, 2022

Keywords: Gene function; Comparative genomics; RNAi screen; iBeetle; Tribolium castaneum; Drosophila melanogaster; Divergence of gene function; iBeetle-Base; FlyBase

There are no references for this article.