Are asymmetric inheritance systems an evolutionary trap? Transitions in the mechanism of paternal genome loss in the scale insect family Eriococcidae

Abstract Haplodiploidy and paternal genome elimination (PGE) are examples of asymmetric inheritance, where males transmit only maternally inherited chromosomes to their offspring. Under haplodiploidy, this results from males being haploid, whereas under PGE, males inherit but subsequently exclude paternally inherited chromosomes from sperm. Their evolution involves changes in the mechanisms of meiosis and sex determination and sometimes also dosage compensation. As a result, these systems are thought to be an evolutionary trap, meaning that once asymmetric chromosome transmission evolves, it is difficult to transition back to typical Mendelian transmission. We assess whether there is evidence for this idea in the scale insect family Eriococcidae, a lineage with PGE and the only clade with a suggestion that asymmetric inheritance has transitioned back to Mendelian inheritance. We conduct a cytological survey of 13 eriococcid species, and a cytological, genetic, and gene expression analysis of species in the genus Cystococcus, to investigate whether there is evidence for species in this family evolving Mendelian chromosome transmission. Although we find that all species we examined exhibit PGE, the mechanism is extremely variable within Eriococcidae. Within Cystococcus, in fact, we uncover a previously undiscovered type of PGE in scale insects that acts exclusively in meiosis, where paternally inherited chromosomes in males are present, uncondensed, and expressed in somatic cells but eliminated prior to meiosis. Broadly, we fail to find evidence for a reversion from PGE to Mendelian inheritance in Eriococcidae, supporting the idea that asymmetric inheritance systems such as PGE may be an evolutionary trap.

vortexed the mixture and transferred it into a spin column.Then we followed the manufacturer's instructions from the Isolate II genomic DNA kit (Bioline) from the first ethanol wash step, finally adding 50μl of EB buffer to elute the DNA.
For C. campanidorsalis, TruSeq DNA library preparation and Illumina 150 base-pair paired-end sequencing (HiSeq2500) was completed at Macrogen Inc. (Republic of Korea) using one sixth of a lane.Potential SSR loci were identified using the QDD2 pipeline software package (Meglécz et al., 2009) and BLAST to ensure that none were within known coding regions.Using Primer3 (Untergasser et al., 2012) within QDD2, we designed primer pairs to amplify fragments between 90 and 400 base pairs in length, with a melting temperature between 57 °C and 63 °C and defaults for other parameters.
For C. echiniformis, we sequenced the sample at Edinburgh Genomics with MiSeq (250bp paired end seq with 350bp inserts) to generate low-coverage whole genome sequence data.We trimmed reads with fastp with parameters --cut_by_quality5 --cut_by_quality3 --cut_window_size 4 --cut_mean_quality 20 (v 0.12.3)(Chen et al., 2018), and assembled the reads using default settings with CLC assembly cell (v5.0.0,Qiagen).We used this assembly to generate primers with QDD (Meglecz et al., 2014).For primer generation, we used the default setting with the exception that we set the minimum PCR product size to 120bp.We also did not do the optional contamination check step (step4 in Meglecz et al. 2014).Instead, we chose simple trinucleotide microsatellites as the target regions, generated primer sets that produced amplicon sizes ranging from 120-300bp, and blasted the contigs these primers were found on to the nr nucleotide database on NCBI, excluding any primers from contigs that blasted to non-Metazoan species.For both species, we chose 24 primer pairs and used the nine that produced the most consistent signals when multiplexed in groups of three primer pairs per PCR reaction.

Microsatellite PCR and thermocycling conditions
For PCR reactions, we used the Type-it microsatellite PCR kit (Qiagen), following the manufacturers guideline with a few variations due to the use of M13 fluorescent primers (Schuelke, 2000).We conducted PCR reactions in a total of 15μl, with 7.5 μl Type-it Mastermix, 0.375μl of the forward primer mix (with each primer at a concentration of 2μM), 1.5μl of the reverse primer mix (also at 2μM), 0.5μl of the M13 fluorescent primer (6FAM or VIC, at 5μM), 3.625μl, and 1.5μl of DNA.The thermocycling conditions for the microsatellite PCR was as follows (following guidelines from Schuelke, 2000): 94°C x 5 min, 30 x (94°C x 30sec, 56°C x 45sec, 72°C x 45sec), 8 x (94°C x 30sec, 53°C x 45sec, 72°C x 45sec), 72°C x 10min.
We aimed to genotype 20 males and the mother of each family for each microsatellite loci.However, due to differences in the number of sons collected for each family, and some DNA extractions or primer sets not working on some males, we analysed fewer than 20 males for some families (or for some primers for a family).For family TLS_091, we analysed 13 males, for family TLS_087 we analysed 15 males, for families LGC_01363 and LGC_02525 we analysed 17 males, for family TLS_095 we analysed 18 males, and for family TLS_100 we analysed 19 males.

RNA extractions for gene expression analysis
For RNA extractions from females, we extracted RNA from a small amount of body tissue from the females as we did not want contamination from germ tissue in the sample.
We first took the female tissue out of RNAlater and rinsed it briefly in sterile 1X PBS before adding 50μl of Trizol to the sample and crushing the sample with a micropestle.We then added 950μl of Trizol and 200μl of chloroform, shook the sample by hand, transferred the supernatant to a new tube, added another 200μl of chloroform, and transferred the supernatant into a new tube for a second time.We then added 500ul of isopropanol and 1μl of linear acrylamide and stored the sample overnight at -20°C.The next day, we centrifuged the sample at 4°C for 15 min, removed the isopropanol from the samples, and added 1ml of freshly prepared 70% EtOH.We inverted the sample several times, centrifuged at 4°C for 15 min (13,000 rpm), and removed the EtOH with a pipette.We did the EtOH cleaning step twice for each sample.We dried the pellet at room temperature, then resuspended the pellet in 40ul distilled H 2 O.We then performed a gDNA digestion, adding 1μl 10X reaction buffer with MgCl2 (ThermoScientific), 1μl DNase I, 8μl water, and 0.25μl of RNase inhibitor for every 1μl of RNA in the sample.We heated the sample at 37°C for 30 min, then added 1μl 50uM EDTA and heated at 65°C for 10 minutes.Finally, we used the RNA Clean & Concentrator kit (Zymo Research) following the manufacturer instructions and eluting the samples in 60ul of H20.
For the RNA extractions from male samples, we used the PureLink RNA Mini Kit (ThermoFisher Scientific), using a slightly modified protocol.We first briefly rinsed the samples in 1X PBS, then added 50μl Trizol into each sample and crushed the tissue with a micropestle.We then added 350μl of Trizol, briefly microcentrifuged the tubes, and transferred the supernatant to a clean tube.We added 80μl of BCP (1-Bromo-3chloropropane), shook the sample by hand for 15 sec and incubated on ice for 3 min.We then centrifuged the samples at 4°C for 15 min (13,000rpm for all centrifugation steps).We transferred the supernatant to a new tube and added an equal volume of freshly prepared EtOH, mixing the tube by vortexing.We transferred the supernatant to a spin column and centrifuged the sample for 30 sec.We discarded the flow through and added 350μl of Wash Buffer I to the sample.We centrifuged for 30 sec, then added 80ul of the DNase mixture (made of 8μl 10X DNase I reaction buffer, 10μl resuspend DNase I, and 62μl of RNase-free water) onto the membrane of the spin column.We incubated this mixture for 15 min at room temperature, then added 350μl of Wash Buffer I to the sample and centrifuged for 30sec.
We then placed the spin cartridge into a new collection tube, added 500μl of Wash Buffer II, and centrifuged the sample for 30sec.We repeated this step once, then centrifuged the sample for an extra minute to dry the membrane.We placed the spin cartridge into a clean eppendorf tube, added 30μl RNase-free water to the membrane, and incubated the sample for 1min.We centrifuge the sample for 2 min and stored the sample at -80°C.We performed a cDNA amplification of the male samples as the yield of RNA from these samples was small due to their small size.In order to do this, we used the Ovation RNAseq System V2 (Tecan), following the manufacturer protocol.

Supplementary Tables/ Figures
Supplementary Figure 1 nuclei (blue circles), which are degraded following meiosis.In the Comstockiella system, male somatic cells look similar to Lecanoid species, but some paternal chromosomes are eliminated from cells just prior to meiosis.The number eliminated is variable between/within species, and pycnotic nuclei can be either present or absent depending on the number eliminated.In the Diaspidid system, paternally inherited chromosomes are eliminated from all cells of males in early development, so heterochromatic bodies are absent from the male soma and pycnotic nuclei are not present after meiosis, which has only one division (as there are no paternal chromosomes).Finally, the last row represents a hypothetical scenario of what somatic cells and meiosis might look like in a species which lacks PGE transmission but evolved from an ancestor with this type of reproduction.In this case, the paternally inherited chromosomes form into viable sperm (i.e.no pycnotic nuclei) and there are two divisions in meiosis and no heterochromatic bodies in somatic cells.

.
Differences in the mechanism of PGE in scale insect species.All scale insects exhibit inverted meiosis (i.e.sister chromatids separate in the first division of meiosis while homologous chromosomes separate in the second division) and chromosomes segregate according to their parent of origin.There are three types of PGE in scale insects; Lecanoid, Comstockiella, and Diaspidid.In the Lecanoid system, paternally inherited chromosomes (blue, maternal chromosomes in red) are retained in males throughout development, and are condensed as heterochromatic bodies in somatic cells (blue circle representing the condensed paternal chromosomes in a ball).During meiosis, paternal chromosomes are present throughout meiosis, but form into non-viable pycnotic

Table 1 .
Sample collection Information for samples used in this study.

Table 2 .
Summary of samples stained and the results of cell staining for somatic cells as well as male tissue undergoing meiosis.For Tanyscelis sp. and Capulinia jaboticabae, we used sequence from a related species in the same genus in the phylogeny.

Table 3 .
Accession numbers for sequences used in figure 2.