Large transgene arrays cause aberrant transcription and synthetic molting defects with myrf-1 mutation

Transgenes are a mainstay of modern biological research, permitting perturbation as well as read-out of gene expression or cellular states. In C. elegans, genome editing (Arribere et al. 2014; Dickinson et al. 2015; Paix et al. 2015; Dokshin et al. 2018; Ghanta and Mello 2020) or transposon-based single copy transgene insertion (Frøkjaer-Jensen et al. 2008; Frøkjær-Jensen et al. 2012) have become the methods .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 2 of choice for many of these applications, as they permit targeted manipulation and physiological expression levels. This contrasts with earlier approaches using transgene arrays consisting of multiple, concatenated copies of the transgene of interest (Mello et al. 1991). Following microinjection of DNA into the C. elegans germline, these arrays can be maintained extrachromosomally, but random segregation causes mosaic inheritance. This can be prevented by random integration of the array into of the genome, following X- or gamma ray (Mello and Fire 1995) or UV light/TMP treatment (Evans 2006) to induce DNA double-strand breaks. Certain problematic features of these integrated arrays are well understood: a high transgene copy number, which may also include truncated or otherwise abnormal copies, render them less suitable for mechanistic studies on gene expression and the repetitive nature makes them susceptible to heterochromatinization and small RNA-mediated silencing, especially in the germline (Kelly et al. 1997; Kim et al. 2005; Ashe et al. 2012). Yet, many multicopy integrated transgene arrays are still routinely used as gene expression and cell fate reporters because decades of research using them have established them as convenient tools and even de facto gold standards. For instance, in our field of research, C. elegans developmental timing, we and others have found multicopy integrated arrays immensely valuable to monitor molting, larval vs adult gene expression states, cell numbers or fates, and numerous publications have shown their value for this purpose. The underlying assumption for these experiments is that the transgenic array does not, or not substantially, alter the phenotype under investigation. Here, we show that this premise does not hold when we study the role of myrf-1 in molting. myrf-1 encodes a transcription factor that plays important roles in developmental timing control through the heterochronic pathway and in molting (Frand et al. 2005; Meng et al., 2017; Xia et al., 2021; Xu et al., 2024). Whereas strong hypomorph and null mutants fail to proceed through the normal four larval molts, previous work identified a single point mutation, myrf-1(mg412), that reactivated a larval molt reporter array, mgIs49[mlt-10p::gfp::pest; ttx-3::gfp], in adult animals, causing an aberrant molt and animal death (Frand et al. 2005). .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 3 Here, we show that the phenotype arises synthetically only in the presence of both myrf-1(mg412) and the mgIs49 array. Although mapping reveals that mgIs49 disrupts the prmt-9 gene on Chromosome IV, we exclude this as the cause for the phenotype. Instead, various arrays cause synthetic phenotypes with myrf-1(mg412), to variable extents. We find that integrated array sizes are consistently huge, ranging from 4 MB to 11 MB for each of eight examples that we mapped, accounting in the case of mgIs49 for > 50% of the 17.49 MB of the wild-type chromosome IV (and > 5% of the entire genome size). Moreover, for two arrays examined in greater detail, mgIs49 and the unrelated maIs105[col-19p::gfp] (Abbott et al., 2005; Ilbay et al. 2021), we observe substantial dysregulation of gene expression and accumulation of reads covering the promoter sequences utilized in the arrays even in the presence of wild-type myrf-1. Collectively, the myrf-1 synthetic phenotype appears best explained by generic features of the arrays such as large size or concatenation presumably at the core of the aberrant promoter transcription. Although we are reassured by the fact that the phenotype is not explained by the array alone but does additionally require the myrf-1 mutation, our findings strongly suggest that multicopy arrays have outlived their usefulness even as tissue or cell type markers and should be replaced whenever possible by single copy transgenes or endogenously tagged genes.

The mgIs49 transgene array is required for the myrf-1(mg412) mutant phenotype Recently, we identified myrf-1 as a candidate component of an oscillator, or clock, that times rhythmic development, especially molting, in C. elegans (Meeuse et al. 2023). Specifically, we found that RNAi-mediated depletion of this rhythmically transcribed gene caused animals to arrest development or die molting, consistent with earlier work showing that a myrf-1(tm2707) putative null mutation caused early larval lethality (Russel et al. 2011). (Russel et al. 2011) had additionally conducted a screen for molting regulators based on inappropriate activation of an integrated multicopy transgene, mgIs49[mlt-10p::gfp::pest; ttx-3::gfp]. (We will refer to this .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 4 transgene as mlt-10p::gfp::pest in the remainder of this article.) They identified animals carrying a mutation in myrf-1 (named pqn-47 at the time), mg412, that caused reactivation of this larval molt reporter in adults, initiation of a new molt, and ultimately death, when animals became trapped in their cuticle. In other words, myrf- 1(mg412) appeared to exhibit a retarded heterochronic phenotype, where adult animals failed to exit from the molting cycle. In seeking to study the molecular details of the function of MYRF-1, we outcrossed the myrf-1(mg412); mgIs49 animals to wild-type N2. Surprisingly, we found animals that were homozygous for the mutation but showed no evidence of cuticle entrapment and premature death when cultured on plates (Figure 1a). This suggested that an unlinked mutation was partly or fully responsible for this phenotype. Notably, these healthy animals lacked the mgIs49 array, suggesting that the array itself, or features linked to it, could contribute to the effect. To investigate whether myrf-1(mg412) was at all required for the phenotype seen in myrf-1(mg412); mgIs49 animals, we reverted the myrf-1 mutation in that strain back to the wild-type coding sequence. This suppressed lethality in adults grown at 25℃ (Table 1). Together these observations indicated that the myrf-1 mutation was required but not sufficient for the phenotype. We tested this hypothesis further by regenerating the mg412 mutation in an otherwise wild-type (non-transgenic) animal (Figure 1b) and subsequently crossing the new myrf-1(bch65) allele into mgIs49. The results were fully consistent with our hypothesis, since myrf-1(bch65) alone did not exhibit adult lethality, whereas myrf- 1(bch65); mgIs49[mlt-10p::gfp::pest] did, along with adult mlt-10p::gfp::pest reactivation (Table 1; Figure 1c). These results also support the phenotype arising from a combination of myrf-1 mutation and presence of the transgene array. We also note that the heterochronic extra molt phenotype reported for myrf- 1(mg412); mgIs49[mlt-10::gfp::pest] differs from that of a canonical retarded heterochronic mutation, lin-29 null (Azzi et al. 2020): When we assessed molting in a high-throughput luciferase assay (Olmedo et al. 2015; Meeuse et al. 2020), lin- 29(xe37) control animals showed the expected extra molts, whereas neither myrf- 1(bch65) nor myrf-1(mg412); mgIs49[mlt-10::gfp::pest] animals did (Figure 1d). .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 5 However, mgIs49 caused developmental delays irrespective of myrf-1 mutational status (Supplemental Figure 1). Figure 1. Adult lethality in myrf-1(mg412) animals is dependent on the presence of the mgIs49[mlt-10p::gfp::pest] array. a. Micrograph showing characteristic carcasses observed when growing myrf-1(mg412); mgIs49[mlt- 10p::gfp::pest] animals but not observed for myrf-1(mg412) animals. b. Chromatogram of relevant myrf-1(mg412) and myrf-1(bch65) sequences confirms CRISPR- mediated regeneration of the T364A missense change of mg412 in bch65 (red boxes). myrf-1(bc65) additionally contains three silent mutations (black boxes). c. Reactivation of the mgIs49[mlt-10p::gfp::pest] reporter in the presence of myrf-1 alleles at adulthood. Percentage of animals (n>87 for each genotype) showing GFP fluorescence from mgIs49[mlt-10p::gfp::pest] at indicated times. The animals were grown at 25℃ and scored for GFP expression from 47 to 72 hours using a stereomicroscope. Table 1b shows survival outcomes for these animals at 96 hours. d. Heatmap showing trend-corrected luminescence for the indicated strains (n>13), expressing luciferase from the eft-3 promoter. Each line represents one animal and traces are sorted by entry into first molt. Darker color represents low luminescence and is associated with molts. lin-29(xe37) animals dying at the juvenile-to-adult transition were censored. .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 6 Supplementary Figure 1. The mgIs49[mlt-10p::gfp::pest; ttx-3p::gfp] array causes developmental delay. Boxplots show larval stage length, intermolt length and molt length for animals of the indicated genotypes (n>13), expressing luciferase from the eft-3 promoter. lin-29(xe37) animals dying at the juvenile-to-adult transition were censored. A mlt-10p::gfp MosSCI reporter does not synergize with myrf-1(mg412) To investigate whether myrf-1 mutation could reactivate molting gene expression independently of the mgIs49 array, we used MosSCI to integrate a single-copy mlt- 10p::pest::gfp::h2b transgene in a targeted fashion into a landing site on chromosome III (Frøkjær-Jensen et al. 2012). We characterized two transgenic lines, .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 7 neither of which caused lethality, irrespective of whether this was assessed in a myrf-1 wild-type or myrf-1(mg412) or myrf-1(bch65) mutant background (Table 2). At the same time, there was no reactivation of the transgene in any of the adults (Figure 2). We conclude that not only the lethality phenotype but also reactivation of molting gene expression requires mgIs49 along with myrf-1(bch65) or myrf-1(mg412) mutations. Figure 2. A single-copy reporter of mlt-10 does not get reactivated in adulthood by myrf- 1(bch65). a., b. bchSi134[mlt-10p::pest::gfp::h2b] III reporter expression in WT and myrf-1(bch65) larvae, respectively, grown at 25℃ for 36 hours (a) or 70 h (b) after plating starved L1s. Images were acquired using the same exposure times for GFP. Scale bars: 10 µm. mgIs49 is a large integrated array disrupting prmt-9 To further characterize mgIs49, we used ONT long-read sequencing to map the insertion site(s) and estimate the total size of the integrated array (see Materials and Methods). This analysis revealed integration of a large array into a protein-coding gene (prmt-9) (Figure 3). The array’s size was estimated at ~8.8 Mb, or around 50% of the chromosome’s wild-type size (Supplementary Table 1). .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 8 To test whether prmt-9 inactivation caused the observed genetic interaction, we disrupted it by genome editing in wild-type animals. This failed to result in adult lethality both on its own and in combination with myrf-1(bch65) (Table 2), suggesting that prmt-9 mutation is not, or not alone, the cause of the observed phenotypes arising in presence of mgIs49. Figure 3. mgIs49[mlt-10p::gfp::pest] insertion disrupts the prmt-9 coding sequence on chromosome IV. a. IGV genome browser screenshot of the identified strain-specific integration in GR1395. Red and blue lines show ONT long reads in the + and – orientation respectively and their primary mapping to the genome; genes are shown on the bottom of the plot. The characteristic breakpoint alignment structures indicate soft-clipped reads flanking the insertion site of the transgene array. In many cases the clipped sequences can be mapped to components of the array and no other strain analyzed in this study shows a breakpoint pattern on this locus (data not shown). b. Schematic of the CRISPR/Cas9 editing of the prmt-9 coding sequence, showing its longer isoform. The alleles bch79, bch80 and bch81 contain exons 1-4, a partial exon 5, and a partial exon 10 of prmt-9. c. mRNA-seq reads mapping to the prmt-9 sequence in GR1395 mgIs49[mlt-10p::gfp::pest] and N2 (two replicates each). Reads corresponding to exons 9 and 10 of the longer prmt-9 isoform are absent in GR1395. .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 9 Other transgene arrays synergize with myrf-1(mg412) to variable extents It seemed possible that the synthetic interactions between myrf-1 mutation and mgIs49 were specific to this combination. Alternatively, transgenic arrays might more broadly cause phenotypes when combined with myrf-1. To distinguish between these possibilities, we crossed two other integrated arrays, expressing different transgenes, into myrf-1(bch65). Both arrays also caused synthetic adult lethality with myrf-1(bch65), although the effect was more moderate with maIs105[col-19::gfp] (Abbott et al. 2005) than with feIs5[sur-5p::luciferase::gfp + rol-6(su1006)] (Lagido et al. 2008) (Table 3). We wondered whether the synthetic phenotypes required array integration and therefore created extrachromosomal lines bearing mlt-10p::pest::gfp::h2b. Similar to mgIs49, each of the three extrachromosomal lines showed some degree of lethality at 96 hours even in a myrf-1 wild-type background. In one of the three lines, this phenotype was enhanced if myrf-1(bch65) was present. Taken together, we conclude that several but not all transgene arrays cause synthetic phenotypes with myrf-1 mutation. This implicates a more generic feature than the specific molecular composition or integration site of mgIs49 in the process, possibly including a large physical size and a repetitive nature. Integrated multicopy transgenes are megabases in size To get a broader view of the physical size range of transgene arrays, we applied the same ONT long-read sequencing and array characterization pipeline to six additional transgene strains used in the community. The results of this survey suggest that the size that we had determined for mgIs49 was in no way unusual: the estimated insertion sizes for each of the arrays range from ~4 to ~11 Mb, accounting for a significant fraction of the entire genome in those animals, and are on the same order of magnitude as individual chromosomes, which range from 14 to 21 Mb. The long-read sequencing data also allowed us to map candidate array integration sites (Supplementary Table 2). We selected maIs105[col-19p::gfp] for further validation, using SNP mapping (Davis et al. 2005) to test a putative insertion .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 10 on the right arm of chromosome V (Figure 4). We crossed maIs105 animals with CB4856 Hawaiian isolate males and selected, based on their GFP-expression status, F2 animals either homozygous for the array or devoid of it. The Hawaiian SNP snp_Y17D7B [3] (Davis et al. 2005) showed a strong segregation bias with GFP: While only 1/34 animals homozygous for the array segregated the Hawaiian SNP, 36/36 F2 animals lacking the array were homozygous for it. Hence, these data confirm that maIs105 is tightly linked to this SNP in the Y17D7B.3 gene (whose interpolated genetic position is V: 17.75). Of note, two additional candidate insertion sites were detected on chromosome II, along with multiple candidate integration sites in several other strains, consistent with the idea that array integration events can produce complex genome aberrations (Supplementary Table 2). Figure 4. maIs105[col-19p::gfp] insertion candidate site on the right arm of Chromosome V. IGV genome browser screenshot of the identified strain-specific integration in VT3855 (lin-46(ma467) maIs105, Ilbay et al., 2021). Features are as in Figure 3. Similar to GR1395, soft-clipped reads indicate the insertion site of the transgene array. The array insertion origin of these reads is further supported by mappings of the clipped sequences to array components. The discontinuity in coverage and the presence of a subset of reads aligning through the integration site suggest a complex genomic rearrangement associated with the integration. Two integrated multicopy transgene arrays cause gene expression changes and extensive promoter transcription To begin exploring whether the presence of such large transgene arrays affects gene expression, we performed mRNA sequencing in early L1 stage VT1367 maIs105[col-19p::gfp] and GR1395 mgIs49[mlt-10p::gfp::pest] animals. (We selected early L1 as they are devoid of the high-amplitude oscillatory gene expression that .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 11 occurs later in larval development and which can be a confounder in interpreting gene expression experiments involving single time point comparison (Aeschimann et al. 2017; Meeuse et al. 2020; Tsiairis and Großhans 2021; Bulteau and Francesconi 2022)). Each strain exhibited reproducible changes in expression relative to wild- type, when compared over two replicates each (Figure 5a). Using edgeR (Chen et al. 2025) with an FDR=0.05 and a minimum log2 fold change of 1, we found 59 down- and 76 upregulated genes in VT1367 maIs105[col-19p::gfp] and 273 down- and 199 upregulated genes in GR1395 mgIs49[mlt-10p::gfp::pest] (Figure 5a). Figure 5. Integrated multicopy arrays cause gene expression changes a. Scatter plots comparing the log2 expression changes between the two biological replicates for the strains VT1367 vs. N2 (left panel) and GR1395 vs. N2 (right panel). Differentially expressed genes (see Methods) are colored in red. b. Karyo-type plot depicting the log2 expression changes as a function of genomic position. Each dot represents a gene. Red lines represent the array integration sites. Blue lines represent sites on chr. II also showing array components found in Nanopore sequencing of VT3855, to which VT1367 is the parental strain (see Methods). We hypothesized that each array might preferentially perturb gene expression locally, i.e., close to its integration site, but saw no indication of such local expression perturbations (Figure 5b). However, we found evidence that there is extensive .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 12 transcription of noncoding regions contained in each array: mlt-10 (Frand et al. 2005) and ttx-3 (Hobert et al. 1997) promoter in mgIs49-containing animals and col- 19 (Abrahante et al. 1998; Abbott et al. 2005) promoter in maIs105-containing animals, respectively (Figure 6). Figure 6. Transcripts covering transgene promoters accumulate in the transgene array- containing lines Browser screenshots showing RNA-seq read alignments for the strains N2, VT1367 maIs105[col- 19p::gfp] and GR1395 mgIs49[mlt-10p::gfp::pest; ttx-3p::gfp] at the col-9, ttx-3 and mlt-10 loci, respectively. Promoter regions present in the arrays are shown as orange or cyan lines, depending on gene orientation. .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 13

First established through pioneering DNA transformation experiments more than 30 years ago (Mello et al. 1991), multicopy transgene arrays have remained widely used tools in C. elegans research. Known shortcomings, such as proneness to silencing, especially in the germline, and overexpression artifacts have all but disqualified their use for gene expression, but they have remained popular as markers of cell fates and states. Our work provides a cautionary tale, showing that transgene arrays are problematic even for these more limited applications: Several arrays, both integrated and extrachromosomal, caused synthetic phenotypes with a myrf-1 mutation. Since the phenotypes were co-dependent on mutant myrf-1, one could consider the arrays as a means to sensitize the system. However, without knowing what exactly causes this sensitization, it seems difficult to interpret any phenotypes in a mechanistically meaningful way. This becomes even more obvious when such composite phenotypes are tested for modification, suppression or enhancement, through additional mutations – is the contribution of the array, the separate mutation, or the interaction between the two modified? One hypothetical risk associated with transgene array integration is the disruption of host genes since integration occurs randomly. Our profiling data show that this is indeed what happened with the mgIs49 array, which is integrated in, and disrupts, the prmt-9 locus. However, although this may prove problematic in some contexts, it did not explain the myrf-1 synthetic lethality. In principle, the effects of such insertional mutagenesis would also seem more controllable – since insertion occurs randomly, testing of a separate integrant in the genetic context of choice would readily rule out arrays causing specific defects. (In reality, however, the logistics of strain generation and testing make reporting of results from independent integrants a rare scenario.) The fact that different transgene arrays, integrated as well as extrachromosomal, caused myrf-1 synthetic lethality appears more concerning, pointing to a more general issue. Although the extent of synthetic lethality varied, the .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 14 fact that it occurs independent of the component transgenes and integration status implicates more generic array features in the process, perhaps linked to their large size or repetitive nature. We were indeed surprised by the large, megabase sizes that we generally observed for all eight integrated arrays that we had assessed. However, this number is in good agreement with other reports, examining distinct arrays: ONT Minion sequencing allowed assembly of an 11 Mb extrachromosomal array (Lin et al. 2021) while (Mouridi et al. 2022) estimated a simple integrated array size at 5.5 Mb. We note that the smallest array we identified, at 4.1 Mb, resulted from biolistic bombardment of a recombineering construct (Sarov et al. 2006). Although generally believed to be smaller, such arrays may be prone to introducing genomic rearrangements (Praitis et al. 2001; Tyson et al. 2018). The risk of such rearrangements is likely a consequence of the DNA double strand repair rather than a specific consequence of biolistic transformation. Indeed, our preliminary analysis seeking to assign candidate integration sites for another seven sequenced strains, provided evidence for more than one possible integration site. For VT1367 maIs105[col-19p::gfp], which we characterized in greater detail, ONT sequencing data analysis revealed three candidate integration sites, two on chromosome II and one on chromosome V. SNP mapping confirmed that GFP expression was linked to the right arm of chromosome V, but the coverage is discontinuous and a subset of reads align through the integration site in the ncs-5 locus. We consider a complex genomic rearrangement as the most likely explanation of these observations. Our data show that multiple arrays interact genetically with myrf-1(mg412), but the ability to modulate mutant phenotypes does not appear limited to this allele. Thus, (Edelman et al. 2016) reported suppression of lin-42(0) mutant precocious alae formation by two out of three transgene arrays that they had tested. One of the suppressing arrays was mgIs49. Finally, our mRNA sequencing data revealed substantial aberrant transcription in mgIs49 and maIs105-carrying animals. While we do not know whether any of the mis-expressed genes contribute specifically to the myrf-1 lethality phenotype, the accumulation of transcripts extending across the transgene .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 15 promoters is striking. Whether these transcripts derive merely from the concatenation of truncated transgene copies remains to be determined, as does their functional relevance, but it seems quite possible that these aberrant transcripts could affect gene expression through anti-sense and other silencing mechanisms. Acknowledgments We thank Milou Meeuse for observing the loss of the extra molt phenotype upon outcrossing myrf-1(mg412); mgIs49 animals. We thank the FMI Functional Genomics Platform for ONT sequencing and RNA sequencing our samples, Anca Neagu for the gift of the mlt-10p reporter plasmid, Kathrin Braun for advice on preparation of samples for RNA sequencing and luciferase assays, and Lucas Morales Moya for help with the luciferase analyzer. Funding Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). IK heads the Friedrich Miescher Institute for Biomedical Research (FMI) C. elegans facility, which is supported by FMI core funding. M.W.M.M. received support from a Boehringer Ingelheim Fonds PhD fellowship. This work is part of a project that has received funding from the Swiss National Science Foundation (#310030_207470, to H.G.) and from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 741269, to H.G.). The FMI is core- funded by the Novartis Research Foundation

General C. elegans methods .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 16 Nematodes were grown under standard conditions at 20° unless indicated otherwise. Mutant alleles were crossed into appropriate genetic backgrounds and, whenever feasible, a wild-type segregant from the same cross was used as a control. Extrachromosomal arrays bchEx30, bchEx31 and bchEx32 are independent transgenic lines resulting from microinjecting the following mixture into adult gonads: mlt-10p 4 kb::pest::gfp::h2b, unc-119(+) at 10 ng/µl; unc-122p::gfp at 50 ng/µl; and 1 Kb Plus DNA Ladder (Invitrogen, 10787018) at 40 ng/µl. Scoring adult survival Gravid adults were allowed to lay eggs for 2 hours and were subsequently removed from the plates, which were cultured at 25℃. 24 hours later, larvae were counted and transferred to fresh plates. Surviving adults were counted at 96 hours. Genome editing by CRISPR/Cas9 Genome editing was performed as described by (Ghanta and Mello 2020). myrf-1(bch65) codes for a T364A missense mutation and was created in the N2 genetic background. It also includes silent mutations R370R, L371L, H372H. myrf-1(bch71), myrf-1(bch72) and myrf-1(bch73) are identical alleles that revert the myrf-1(mg412) mutation to the wild-type sequence. They were created in the

of the strain IFM363 myrf-1(mg412) II; mgIs49 [mlt-10p::gfp::pest + ttx- 3::gfp] IV; xeSi311 [eft-3p::luc::gfp::unc-54 3'UTR, unc-119(+)] V. prmt-9(bch79), prmt-9(bch80) and prmt-9(bch81) are identical prmt-9 truncation alleles, created in the background of the strain IFM332 myrf-1(bch65[T364A]). Single-copy transgene integration by MosSCI .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 17 Transgenic worms expressing the mlt-10p 4 kb::pest::gfp::h2b, unc-119(+) construct were obtained by single-copy integration into the site oxTi444 on chromosome III (Frøkjær-Jensen et al. 2012) to yield bchSi134 [mlt-10p::pest::gfp::h2b] and bchSi135 [mlt-10p::pest::gfp::h2b]. Luciferase assays Luciferase assays were performed as described (Meeuse et al. 2020). Briefly, gravid adults were treated with a bleaching solution to extract eggs. Single embryos were transferred into a well of a white, flat-bottom, 384-well plate (Berthold Technologies, 32505) by pipetting. Animals developed in 90 mL S-Basal medium containing E. coli OP50 (OD600 = 0.9) and 100 mM Firefly D-Luciferin (p.j.k., 102111). Plates were sealed with Breathe Easier sealing membrane (Diversified Biotech, BERM-2000). Luminescence was measured using a luminometer (Berthold Technologies, Centro XS3 LB 960) every 10 min for 0.5 s for 72 or 120 hours; the animals experience a temperature of 22℃. An automated algorithm was used to detect the hatch and the molts from the luminescence data. GFP imaging Fluorescent and Differential Interference Contrast (DIC) images were acquired using a Zeiss Axio Observer Z1 microscope with AxioVision software and Zen 2 (Blue Edition). Region selection and image processing was performed using Fiji. Oxford Nanopore DNA sequencing Mixed-stage animals were grown on a 10 cm NGM plate and washed three times with M9 buffer, pelleted and frozen in a dry ice-ethanol bath for 10 minutes. The DNA was extracted using the QIAGEN Puregene Cell Kit (158043) and Proteinase K (19131) as described in the Supplemental Protocol. DNA was eluted in 200 µl water per sample. .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 18 4µg of genomic DNA was fragmented to a target size of 15kb using G-Tubes (Covaris). Library preparation was performed using the Native barcoding kit v14 (SQK-NBD114.24) from Oxford Nanopore Technologies, following the manufacturer recommendations. Pooled libraries were sequenced on a PromethION flowcell (R10.4.1) using a P2 solo device. Processing of Oxford Nanopore DNA sequencing data Raw POD5 signal files were basecalled using Dorado (version 0.8.1) from Oxford Nanopore Technologies with the R10.4.1 high-accuracy model ([email protected]). Basecalling was performed on an NVIDIA A40 GPU. Basecalled reads were demultiplexed using dorado demux (kit SQK- NBD114-24) with barcode trimming enabled, yielding per-barcode read sets corresponding to the analyzed strains. Demultiplexed long reads were subsequently aligned to the C. elegans genome (ce11) using minimap2 (version 2.24) with the map-ont preset. Alignments were converted to BAM, sorted by coordinate, and low- confidence mappings were removed (minimum Q-score threshold of 8). These per- strain BAM files were used for downstream insertion-site discovery and array size estimation. Identification of candidate insertion sites from ONT reads To identify genomic insertion sites of large integrated transgenic arrays, aligned ONT reads from each strain were analyzed for breakpoint-like alignment signatures. We first extracted reads that showed at least 500 bp of terminal soft clipping, as expected for junction-spanning reads containing genomic sequence contiguous with non-reference array DNA. For each read, the genomic breakpoint coordinate was defined from the alignment boundary, taking strand orientation into account, and the number of reads supporting each chromosome-position pair was recorded and only candidate breakpoint sites with a minimum support of 5 reads were retained. To remove recurrent background signals, breakpoint coordinates detected in all other strains were pooled and subtracted from those of the target strain, retaining only .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 19 strain-specific sites. For each remaining candidate, we then determined how many reads overlapping the genomic breakpoint also aligned to reference sequences representing known array components. Candidate insertion sites were defined as strain-specific breakpoint hotspots supported by multiple soft-clipped reads, including reads with matches to array-derived sequence. Finally, insertion sites were confirmed by visual inspection in a genome browser. Estimation of total array sizes from ONT reads mapping to array components Total array sizes (over all possible insertions) were estimated from the relative number of long reads mapping to array components versus those mapping to the genome. For each sample, ONT reads were aligned to reference sequences representing known construct components using minimap2 (Li 2021) and the map- ont preset. Unmapped, secondary, and supplementary alignments were excluded from the construct-component BAM, and the corresponding genome-aligned BAM was used for normalization. Each read was counted only once per BAM file and its length was summed to obtain the total number of mapped base pairs for construct- component and genome alignments. Genome depth was then estimated as mapped genomic base pairs divided by the haploid C. elegans genome size. Finally, total array size per diploid genome was calculated by normalizing construct-mapped base pairs by this genome depth. RNA sequencing N2, GR1935 mgIs49 [mlt-10p::gfp::pest; ttx-3::gfp] and VT1367 maIs105 [col-19::gfp] L1 animals through starvation by hatching eggs into M9 buffer overnight. Following plating on OP50-seeded plates, animals were allowed to develop for 4 hours at 25˚C. The animals were washed three times and lysed in Trizol by repeated freeze- thawing cycles. RNA was extracted using the Direct-zol RNA Microprep kit (Zymo, .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 20 R2062). mRNA-seq libraries were generated using the Illumina Stranded mRNA Prep solution, according to the manufacturer’s protocol. Libraries were sequenced on a NovaSeq6000 flowcell, paired-end 2x56cycles. Demultiplexing was performed using bcl2fastq2. Processing of the RNA-seq data RNA-seq data was mapped to the C. elegans genome (ce11) using the align (qAlign) function from the QuasR (version 1.34.0) package in R with splicedAlignement = TRUE, using the aligner HISAT2 (version 1.10.0) including an exon-exon junction database. Gene expression levels were quantified using the qCount() function. For annotations, coding transcript info from WormBase WS270 was used. We further compensated for differences in library sizes by scaling each library to the average library size, and log2-transformed the data using a pseudocount of 8 (log2(x+8)). Given that fragments of the genes of mlt-10, ttx-3 and col-19 occur in the arrays, we removed the affected exons from those genes to only count reads that originated from the endogenous genes as opposed to the array. To determine those affected exons, we inspected the read alignments at those loci and removed the exons at the 5' end that overlapped sections with extensive read amplification. Differentially expressed genes between GR1395 mgIs49 [mlt-10p::gfp::pest; ttx- 3::gfp] and N2 as well as VT1367 maIs105 [col-19::gfp] and N2, respectively, were identified using the R package edgeR (version 4.8.2) (Chen et al. 2025). Genes with low expression were filtered out using filterByExpr. Quasi-likelihood F-tests were performed to assess differential expression, and genes with a false discovery rate (FDR) < 0.05 and minimum log2 fold change of 1 were considered significantly differentially expressed. Genetic mapping of maIs105 VT1367 maIs105 [col-19::gfp] hermaphrodites were mated with CB4856 Hawaiian males. Two kinds of F2 animals were analyzed: those homozygous for the array and .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 21 those homozygous for its absence. SNP analysis using the SNPs pkP5076 (V, -17), pkP5097 (V, 1), *** (V,6), uCE5-2609 (V, 13) and snp_Y17D7B[3] (V, 18) (Davis et al. 2005) of 72 animals placed maIs105 on the right arm of chromosome V (distal to snp_Y17D7B [3]). Data availability The sequencing data produced in this study are available at NCBI’s Gene Expression Omnibus (Edgar et al. 2002) under SuperSeries accession number GSE329497 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE329497).

Abbott AL et al. 2005. The let-7 MicroRNA Family Members mir-48, mir-84, and mir-241 Function Together to Regulate Developmental Timing in Caenorhabditis elegans. Dev Cell. 9(3):403–414. https://doi.org/10.1016/j.devcel.2005.07.009 Abrahante JE, Miller EA, Rougvie AE. 1998. Identification of Heterochronic Mutants in Caenorhabditis elegans: Temporal Misexpression of a Collagen::Green Fluorescent Protein Fusion Gene. Genetics. 149(3):1335–1351. https://doi.org/10.1093/genetics/149.3.1335 Aeschimann F et al. 2017. LIN41 Post-transcriptionally Silences mRNAs by Two Distinct and Position-Dependent Mechanisms. Molecular cell. 65(3):476-489.e4. https://doi.org/10.1016/j.molcel.2016.12.010 Arribere JA et al. 2014. Efficient marker-free recovery of custom genetic modifications with CRISPR/Cas9 in Caenorhabditis elegans. Genetics. 198(3):837–846. https://doi.org/10.1534/genetics.114.169730 Ashe A et al. 2012. piRNAs Can Trigger a Multigenerational Epigenetic Memory in the Germline of C. elegans. Cell. 150(1):88–99. https://doi.org/10.1016/j.cell.2012.06.018 Azzi C, Aeschimann F, Neagu A, Großhans H. 2020. A branched heterochronic pathway directs juvenile-to-adult transition through two LIN-29 isoforms. eLife. 9:e53387. https://doi.org/10.7554/elife.53387 Bulteau R, Francesconi M. 2022. Real age prediction from the transcriptome with RAPToR. Nat Methods. 19(8):969–975. https://doi.org/10.1038/s41592-022-01540-0 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 22 Chen Y et al. 2025. edgeR v4: powerful differential analysis of sequencing data with expanded functionality and improved support for small counts and larger datasets. Nucleic Acids Res. 53(2):gkaf018. https://doi.org/10.1093/nar/gkaf018 Davis MW et al. 2005. Rapid single nucleotide polymorphism mapping in C. elegans. BMC Genomics. 6(1):118. https://doi.org/10.1186/1471-2164-6-118 Dickinson DJ et al. 2015. Streamlined Genome Engineering with a Self-Excising Drug Selection Cassette. Genetics. 200(4):1035–1049 http://www.genetics.org/content/200/4/1035. https://doi.org/10.1534/genetics.115.178335 Dokshin GA, Ghanta KS, Piscopo KM, Mello CC. 2018. Robust Genome Editing with Short Single-Stranded and Long, Partially Single-Stranded DNA Donors in Caenorhabditis elegans Genetics. 210(3):781 Edelman TLB et al. 2016. Analysis of a lin-42/period Null Allele Implicates All Three Isoforms in Regulation of Caenorhabditis elegans Molting and Developmental Timing. G3: Genes, Genomes, Genet. 6(12):4077–4086. https://doi.org/10.1534/g3.116.034165 Edgar R, Domrachev M, Lash AE. 2002. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 30(1):207–210. https://doi.org/10.1093/nar/30.1.207 Evans T. 2006. Transformation and microinjection. WormBook. [published online ahead of print]. https://doi.org/10.1895/wormbook.1.108.1 Frand AR, Russel S, Ruvkun G. 2005. Functional Genomic Analysis of C. elegans Molting. Plos Biol. 3(10):e312. https://doi.org/10.1371/journal.pbio.0030312 Frøkjaer-Jensen C et al. 2008. Single-copy insertion of transgenes in Caenorhabditis elegans. Nature genetics. 40(11):1375–1383 http://www.nature.com/ng/journal/v40/n11/abs/ng.248.html. https://doi.org/10.1038/ng.248 Frøkjær-Jensen C, Davis MW, Ailion M, Jorgensen EM. 2012. Improved Mos1-mediated transgenesis in C. elegans. Nature Methods. 9(2):117–118. https://doi.org/10.1038/nmeth.1865 Ghanta KS, Mello CC. 2020. Melting dsDNA Donor Molecules Greatly Improves Precision Genome Editing in Caenorhabditis elegans. Genetics. 216(3):643–650. https://doi.org/10.1534/genetics.120.303564 Hobert O et al. 1997. Regulation of Interneuron Function in the C. elegans Thermoregulatory Pathway by the ttx-3 LIM Homeobox Gene. Neuron. 19(2):345–357. https://doi.org/10.1016/s0896-6273(00)80944-7 Ilbay O, Nelson C, Ambros V. 2021 C. elegans LIN-28 controls temporal cell fate progression by regulating LIN-46 expression via the 5' UTR of lin-46 mRNA. Cell Rep. 36(10):109670 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 23 Kelly WG, Xu S, Montgomery MK, Fire A. 1997. Distinct requirements for somatic and germline expression of a generally expressed Caernorhabditis elegans gene. Genetics. 146(1):227–238 Kim JK et al. 2005. Functional Genomic Analysis of RNA Interference in C. elegans. Science. 308(5725):1164–1167. https://doi.org/10.1126/science.1109267 Lagido C, Pettitt J, Flett A, Glover LA. 2008. Bridging the phenotypic gap: Real-time assessment of mitochondrial function and metabolism of the nematode Caenorhabditis elegans. BMC Physiol. 8(1):7. https://doi.org/10.1186/1472-6793-8-7 Li H. 2021. New strategies to improve minimap2 alignment accuracy. Bioinformatics. 37(23):4572–4574. https://doi.org/10.1093/bioinformatics/btab705 Lin Z et al. 2021. Formation of artificial chromosomes in Caenorhabditis elegans and analyses of their segregation in mitosis, DNA sequence composition and holocentromere organization. Nucleic Acids Res. 49(16):9174–9193. https://doi.org/10.1093/nar/gkab690 Meeuse MW et al. 2020. Developmental function and state transitions of a gene expression oscillator in Caenorhabditis elegans. Mol Syst Biol. 16(7):e9498. https://doi.org/10.15252/msb.20209498 Meeuse MWM et al. 2023. C. elegans molting requires rhythmic accumulation of the Grainyhead/LSF transcription factor GRH‐1. EMBO J. 42(4):EMBJ2022111895. https://doi.org/10.15252/embj.2022111895 Mello C, Fire A. 1995. Chapter 19 DNA Transformation. Methods Cell Biol. 48:451–482. https://doi.org/10.1016/s0091-679x(08)61399-0 Mello CC, Kramer JM, Stinchcomb D, Ambros V. 1991. Efficient gene transfer in C.elegans: extrachromosomal maintenance and integration of transforming sequences. The EMBO Journal. 10(12):3959–3970 Meng J, Ma X, Tao H, Jin X, Witvliet D, Mitchell J, Zhu M, Dong MQ, Zhen M, Jin Y, Qi YB. (2017) Myrf ER-Bound Transcription Factors Drive C. elegans Synaptic Plasticity via Cleavage-Dependent Nuclear Translocation. Dev Cell. 41(2):180-194.e7 Mouridi SE, Alkhaldi F, Frøkjær-Jensen C. 2022. Modular safe-harbor transgene insertion for targeted single-copy and extrachromosomal array integration in Caenorhabditis elegans. G3 Genes Genomes Genetics. [published online ahead of print]. https://doi.org/10.1093/g3journal/jkac184 Olmedo M, Geibel M, Artal-Sanz M, Merrow M. 2015. A High-Throughput Method for the Analysis of Larval Developmental Phenotypes in Caenorhabditis elegans. Genetics. 201(2):443–448. https://doi.org/10.1534/genetics.115.179242 Paix A, Folkmann A, Rasoloson D, Seydoux G. 2015. High Efficiency, Homology-Directed Genome Editing in Caenorhabditis elegans Using CRISPR-Cas9 Ribonucleoprotein Complexes. Genetics. 201(1):47–54. https://doi.org/10.1534/genetics.115.179382 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 24 Praitis V, Casey E, Collar D, Austin J. 2001. Creation of Low-Copy Integrated Transgenic Lines in Caenorhabditis elegans. Genetics. 157(3):1217–1226. https://doi.org/10.1093/genetics/157.3.1217 Russel S, Frand AR, Ruvkun G. 2011. Regulation of the C. elegans molt by pqn-47. Dev Biol. 360(2):297–309. https://doi.org/10.1016/j.ydbio.2011.09.025 Sarov M et al. 2006. A recombineering pipeline for functional genomics applied to Caenorhabditis elegans. Nat Methods. 3(10):839–844. https://doi.org/10.1038/nmeth933 Tsiairis C, Großhans H. 2021. Gene expression oscillations in C. elegans underlie a new developmental clock. Curr Top Dev Biol. 144:19–43 Tyson JR et al. 2018. MinION-based long-read sequencing and assembly extends the Caenorhabditis elegans reference genome. Genome Res. 28(2):266–274. https://doi.org/10.1101/gr.221184.117 Xia SL, Li M, Chen B, Wang C, Yan YH, Dong MQ, Qi YB. 2021. The LRR-TM protein PAN-1 interacts with MYRF to promote its nuclear translocation in synaptic remodeling. Elife 10:e67628. Xu Z, Wang Z, Wang L, Qi YB. 2024. Essential function of transmembrane transcription factor MYRF in promoting transcription of miRNA lin-4 during C. elegans development. Elife. 12:RP89903. .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 25 Tables Table 1: myrf-1(mg412) acts synergistically with mgIs49[mlt-10p::gfp::pest] to cause an adult lethality phenotype a b * All lines also contain xeSi311[eft-3p::luc::gfp::unc-54 3'UTR, unc-119(+)] V # Provided for comparison, data from Table 2. This line also contains xeSi311[eft-3p::luc::gfp::unc-54 3'UTR, unc- 119(+)] V Table 2: Neither a mlt-10p::gfp MosSCI reporter nor a prmt-9 deletion synergize with myrf-1(mg412) to cause an adult lethality phenotype a Relevant genotype* % death at 96 hours, 25°C (n) myrf-1(mg412); mgIs49[mlt-10p::gfp::pest] 98 (165) mgIs49[mlt-10p::gfp::pest] 14 (262) myrf-1(bch71[WT revertant of mg412]); mgIs49[mlt-10p::gfp::pest] 21 (279) myrf-1(bch72[WT revertant of mg412]); mgIs49[mlt-10p::gfp::pest] 32 (218) myrf-1(bch73[WT revertant of mg412]); mgIs49[mlt-10p::gfp::pest] 22 (219) Genotype % death at 96 hours, 25°C (n) mgIs49[mlt-10p::gfp::pest] 4 (106) myrf-1(mg412); mgIs49[mlt-10p::gfp::pest] 98 (90) myrf-1(bch65); mgIs49[mlt-10p::gfp::pest] 84 (87) myrf-1(bch65) # 0 (226) Genotype % death at 96 hours, 25°C (n) bchSi134[mlt-10p::pest::gfp::h2b] 0 (188) myrf-1(mg412); bchSi134[mlt-10p::pest::gfp::h2b] 1 (159) bchSi134[mlt-10p::pest::gfp::h2b] 0 (155) myrf-1(bch65); bchSi134[mlt-10p::pest::gfp::h2b] 0 (158) bchSi135[mlt-10p::pest::gfp::h2b] 0 (224) myrf-1(mg412); bchSi135[mlt-10p::pest::gfp::h2b] 1 (255) bchSi135[mlt-10p::pest::gfp::h2b] 0 (282) myrf-1(bch65); bchSi135[mlt-10p::pest::gfp::h2b] 0 (288) .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 26 b * All lines also contain xeSi311[eft-3p::luc::gfp::unc-54 3'UTR, unc-119(+)] V Table 3: Extrachromosomal and integrated arrays cause adult lethality phenotype in the presence of myrf-1(mg412) mutation a b Relevant genotype* % death at 96 hours, 25°C (n) myrf-1(bch65) 0 (226) myrf-1(bch65); prmt-9(bch79) 3 (144) myrf-1(bch65); prmt-9(bch80) 2 (149) myrf-1(bch65); prmt-9(bch81) 4 (143) Genotype % death at 96 hours, 25°C (n) bchEx30[mlt-10p::pest::gfp::h2b; unc-122p::gfp] 5 (314) myrf-1(bch65); bchEx30[mlt-10p::pest::gfp::h2b; unc-122p::gfp] 3 (151) bchEx31[mlt-10p::pest::gfp::h2b; unc-122p::gfp] 19 (170) myrf-1(bch65); bchEx31[mlt-10p::pest::gfp::h2b; unc-122p::gfp] 67 (105) bchEx32[mlt-10p::pest::gfp::h2b; unc-122p::gfp] 26 (78) myrf-1(bch65); bchEx32[mlt-10p::pest::gfp::h2b; unc-122p::gfp] 26 (147) Genotype % death at 96 hours, 25°C (n) feIs5[sur-5p::luciferase::gfp + rol-6(su1006)] 10 (291) myrf-1(bch65); feIs5[sur-5p::luciferase::gfp + rol-6(su1006)] 83 (125) maIs105[col-19::GFP] 12 (179) myrf-1(bch65); maIs105[col-19::GFP] 25 (200) .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 27 Supplemental Protocol DNA extraction from frozen worm pellets for Oxford Nanopore Sequencing, based on https://nanoporetech.com/document/extraction-method/c-elegans-dna QIAGEN Puregene Cell Kit (158043) and Proteinase K (19131) 1. Add 1.5 ml of Cell Lysis Solution to the 15 ml Falcon tube containing the frozen worm pellet, and allow the pellet to thaw in the lysis buffer. 2. Add 7.5 µl of Proteinase K and resuspend the pellet by pipetting with a 1 ml wide-bore tip. 3. Incubate the resuspended worms at 50°C for 1 hour; During this incubation, gently invert the tube 3 times every 30 minutes. 4. Repeat steps 1-3 once more. 5. Add 15 µl of RNase A and mix by inverting the tube. 6. Incubate the tube at 37°C for 30 minutes. 7. Place the tube on ice for 2 minutes. 8. Add 1 ml of Protein Precipitation Solution and pulse-vortex three times for 5 seconds. 9. Centrifuge at 2000 x g for 10 minutes. 10. Add 3 ml of isopropanol to a fresh 15 ml Falcon tube. 11. Pour the supernatant from step 9 into the Falcon tube with isopropanol. Discard the pellet. 12. Gently invert the tube 50 times. 13. Centrifuge at 2000 x g for 5 minutes. 14. Discard the supernatant and add 3 ml of 70% ice-cold ethanol to the pellet. Gently invert the tube several times to mix. 15. Centrifuge at 2000 x g for 2 minutes. 16. Discard the supernatant and remove as much ethanol as possible using sterile paper wipes. 17. Add 200 µl water, vortex for 5s. Incubate at 65 °C for an hour. Incubate overnight with gentle shaking. .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint 28 Supplementary Table 1: Array size estimation strain # reads mapping to construct Total bp construct reads # reads mapping to ce11 Total bp genome reads Genome depth Total array estimate (Mb) GR1395 268627 1479434959 4082261 16773265848 167.7 8.8202 HW143 22512 145554495 455140 2300983994 23 6.3257 SX346 16218 143146842 248341 1500816712 15 9.5379 VT3855 71469 414726268 974247 4559869817 45.6 9.0951 MRS387 64659 384296764 675902 3480718094 34.8 11.0407 PE254 57684 389794102 864314 4956895038 49.6 7.8637 PE255 49982 376376967 818510 5162372760 51.6 7.2908 OP768 12226 94694101 409690 2308839993 23.1 4.1014 Supplementary Table 2: Candidate integration sites Strain Integration Locus Chromosome Coordinate Genomic Support Array Support Breakpoint Support OP768 1/2 chrV 1986898 54 12 10 OP768 2/2 chrV 2344706 43 12 5 PE255 1/2 chrX 4257408 27 10 12 PE255 2/2 chrX 17200018 81 22 11 PE254 1/1 chrV 20165630 74 34 17 MRS387 1/1 chrX 14209166 19 14 14 MRS387 1/1 chrX 14209174 21 20 8 VT3855 1/3 chrV 18829757 91 28 10 VT3855 2/3 chrII 3483183 59 20 15 VT3855 3/3 chrII 8420157 30 29 12 SX346 1/2 chrII 14770046 10 9 6 SX346 2/2 chrX 5472798 53 31 23 SX346 2/2 chrX 5478852 75 14 5 SX346 2/2 chrX 5479078 76 14 5 SX346 2/2 chrX 5483955 48 25 21 HW143 1/2 chrV 14013649 48 31 7 HW143 2/2 chrV 7068894 47 14 7 GR1395 1/1 chrIV 12677932 81 79 28 GR1395 1/1 chrIV 12679597 101 16 48 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 5, 2026. ; https://doi.org/10.64898/2026.04.30.721998doi: bioRxiv preprint