Evolution of a fatty acyl–CoA elongase underlies desert adaptation in Drosophila

Traits that allow species to survive in extreme environments such as hot-arid deserts have independently evolved in multiple taxa. However, the genetic and evolutionary mechanisms underlying these traits have thus far not been elucidated. Here, we show that Drosophila mojavensis, a desert-adapted fruit fly species, has evolved high desiccation resistance by producing long-chain methyl-branched cuticular hydrocarbons (mbCHCs) that contribute to a cuticular lipid layer reducing water loss. We show that the ability to synthesize these longer mbCHCs is due to evolutionary changes in a fatty acyl–CoA elongase (mElo). mElo knockout in D. mojavensis led to loss of longer mbCHCs and reduction of desiccation resistance at high temperatures but did not affect mortality at either high temperatures or desiccating conditions individually. Phylogenetic analysis showed that mElo is a Drosophila-specific gene, suggesting that while the physiological mechanisms underlying desert adaptation may be similar between species, the genes involved in these mechanisms may be species or lineage specific.

The PDF file includes: Figs. S1 to S9 Tables S1 to S4 Legend for dataset S1 Other Supplementary Material for this manuscript includes the following:

Figure S6
. Phylogenetic tree of all elongase genes identified in Drosophila melanogaster, Aedes aegypti, Apis mellifera, Bombyx mori, and Tribolium castaneum.The elongase genes of each species are denoted with a different color.The phylogenetic tree was inferred by the Maximum Likelihood method using amino acid sequences with 1000 bootstrap tests.The numbers next to nodes represent bootstrap values.The scale bar indicates the number of changes per site.The tree showed that three elongase genes, bond, CG31523, and sit, have one-to-one orthologs across all five species.mElo is clustered with a few Drosophila genes suggesting that this gene is likely to be Drosophila specific.Trace (n.s.) 6.9 ± 0.2 (***) Table S3.CHC profiles for D. mojavensis with mElo knocked out in females (top) and males (bottom).ISO1, ISO2, and ISO3 are isofemale lines established from the parental population.M3.5, M3.9, and M3.11 are independent mElo-knockout lines with 5 bp insertion, 90 bp deletion, and 10 bp deletion on the third exon of mElo.The abundance is in ng / fly.N = 6.Linear mixed effects models were applied to determine the difference of each CHC between the wild type and knockout lines of D. mojavensis.The three isofemale wild type and independent knockout strains were included as random effects.15.8 ± 1.9 14.8 ± 1.9 12 ± 1.3 18.9 ± 4.2 7.5 ± 1.9 8.3 ± 2.3 Table S4.Primers or Oligos used for this study Name Primers (5' to 3')

Figure S1 .
Figure S1.Phylogeny of elongases in the mElo loci of D. melanogaster and D. mojavensis.The coding sequences of these genes were used to generate the phylogeny using the Maximum Likelihood method with the GTR model and 1000 bootstraps.The phylogenetic analysis showed that the D. mojavensis orthologue of mElo is GI20347.

Figure S2 .
Figure S2.RNA in situ hybridization of fatty acyl-CoA elongase genes in the D. melanogaster and the D. mojavensis mElo loci on adults.In D. melanogaster, CG18609 RNA transcript was detected in the adult oenocytes.In D. mojavensis, GI20343, GI20345, and GI20347 RNA transcripts were detected in the adult oenocytes.The expressions of all four genes are sexually monomorphic.Arrowheads point to oenocytes.Filled arrowheads indicate visible expression detected and open arrowheads indicate no visible expression.

Figure S4 .
Figure S4.GC-MS chromatograms of mbCHCs in three homozygous Dmoj/mElo knockout strains of D. mojavensis, M3.5, M3.9, and M3.11.In all three knockout strains, levels of 2MeC30 and 2MeC32were reduced and levels of 2MeC28 were increased compared to the wild-type control.

Figure S5 .
Figure S5.Knockout of the mElo orthologue GI20347 in D. mojavensis did not lead to significant differences in survival at 37°C in a non-desiccating environment.Differences in survival between the wild type and Dmoj/mElo knockout strains of D. mojavensis were determined using the linear mixed effects model with the variation within each group (iso-female or independent knockout strains) being random effects.No significant differences were observed (Female: P = 0.4; Male: P = 0.2).

Figure S7 .
Figure S7.Monthly average maximum temperatures in the Sonoran Desert.The plot of monthly average maximum temperatures in a climatic station (GILA BEND 2 SE, AZ US) in the Sonoran Desert from December 1892 to August 2022.The climatic station is located at the coordinate 32.93803, -112.68109.The red dotted line indicates 37°C.The data were obtained from NCEI-NOAA (https://www.ncei.noaa.gov/).

Figure S8 .
Figure S8.Model showing the production of mbCHCs in D. melanogaster and D. mojavensis.In D. melanogaster, the elongase mElo elongates 2MeC24 to 2MeC26 and 2MeC28.In D. mojavensis, the elongase mElo elongates 2MeC28 to 2MeC30 and 2MeC32, while elongation to 2MeC28 is due to another elongase, possibly GI20345, which is expressed in D. mojavensis oenocytes and can elongate shorter mbCHCs to 2MeC28 when overexpressed in D. melanogaster oenocytes.As the knockout of mElo did not fully reduce the production of 2MeC30 in D. mojavensis, we hypothesize that another elongase may also be involved in the synthesis of mbCHCs up to 2MeC30 in D. mojavensis.

Table S1 . CHC profiles of female (top) and male (bottom) D. melanogaster mEloKO and mElo rescue strains. The
abundance is in ng / fly.Student's t-test was used to determine the difference of each CHC between each transgenic line and the control.P-values were adjusted using Benjamini-Hochberg correction at alpha = 0.05.N =4-5.n.s.: not significant; *: P < 0.05; **: P < 0.01; ***: P < 0.001.