Among all insects, the Lepidoptera possess the most striking example of long-range communication by sex pheromones. Because of the central role of sex pheromones in mediating reproductive behaviors, moths are regarded as excellent models to address fundamental questions on the evolution of mate signaling. Hence, changes in the signal and/or response are argued to promote behavioral sexual isolation and speciation, and how phenotypic transitions are initiated are not fully understood, which has driven my curiosity to decipher the genetic causes of variation in signal production.

For a new mate recognition system to evolve, variations in signal and response need to be tightly coordinated. Current evidence shows that coordination does not result from genetic coupling of signal and response, but rather from the way selection acts on pheromone communication channels. Hence, selection operates more on seeking male moths than on female production, notably because of enhanced competition on males. The resulting asymmetric forces predict for changes to be initiated in pheromone production and subsequently tracked by males by ‘viability selection’. Males are tuned to the most common blend in a population but their preference distributions tend to be broader, as far as the new blend ratio or composition is in their response range, they are likely to track it without being affected by variation in signal characteristics.  The primary genetic changes underlying shifts in mate-signalling in moths thus primarily lie in the variable constraints on pheromone biosynthetic pathways (Figure 1) and the source of genetic variation in pheromone production genes. By investigating biosynthetic pathways and what mechanisms are source of pheromone variation, my research substantiates that biochemical novelties in the pheromone biosynthetic pathways can lead to the emergence of new pre-mating reproductive barriers, eventually leading to speciation.

Figure 1: Moth pheromones typically are mixtures of unsaturated fatty acid derivatives. Pheromone biosynthesis takes place by sequential modifications of ubiquitous fatty acid precursors coupled to acyl-coenzyme A (CoA), including beta-oxidation and desaturation reactions followed by enzymatic modification of the carbonyl end into an alcohol, aldehyde or acetate. From a unique fatty acid precursor (18:Acid), the successive combination of specific enzymatic steps (desaturation, b-oxidation, reduction, acetylation and oxidation) can lead to a variety of volatile pheromone components. Desaturation reactions are catalysed by several functional classes of desaturases, ∆9- and ∆11-desaturases being the most common types in Lepidoptera. E refers to trans and Z refers to cis geometrical isomers. The number before the hyphen (-) refers to the position of the double bond, the number after the hyphen refers to the number of carbon atoms in the acyl chain. MA. Liénard, PhD thesis.

 Using biochemical, molecular and functional approaches I investigate the genetic architecture of sex pheromone production. Especially, I characterized gene members of two key pheromone biosynthetic enzyme families (the fatty-acyl-CoA desaturases and reductases, and assessed the biological activity of multiple isolated transcripts in vitro after optimizing a yeast-based expression system, thereby determining the contribution of each functional gene to the synthesis of major pheromone precursor components.  This combinatorial approach allowed to elucidate how in vitro biochemical activities relate to in vivo biosynthetic pathways in adult female moths of Yponomeutidae, Lasiocampidae, Prodoxidae families  - and male Bicyclus butterflies, thus elucidating the functional evolution of biosynthetic genes across distantly related moth species.

Figure 2: Sex-pheromone components of small ermine moths (Lepidoptera, Ypnomeutidae) (upper left panel). Ventral view of Ypnomeuta rorellus (recently renamed Y. rorrella) extruding its terminal abdominal segments encompassing the sex pheromone-producing tissue (i.e. the intersegmental membrane between segments VIII and IX) (lower left panel). Phylogeny of lepidopteran FARs (Fatty-acyl-CoA reductases) including the Lepidoptera-specific subfamily involved in pheromone production (clade in red) (right panel). Liénard et al. 2010. PNAS 107:10955-60.

Studying the function of genes controlling pheromone production provides real insights into the molecular basis of variation of moth sex pheromones. Most duplication events leading to extant moth desaturase lineages occurred before the divergence of moths and flies around 300 million years ago,  however two more recent duplication events giving rise to the ∆9 C₁₈ > C₁₆ and ∆11-desaturase subfamilies occurred in the early Lepidopteran evolution. In a Lasiocampidae species, Dendrolimus punctatus, the ∆9 C₁₆ > C₁₈ desaturase acquired the ability to produce E9 monoenoic and dienoic pheromone precursors, demonstrating that the substrate preference of certain contemporary moth ∆9-desaturases is highly divergent from their ancestral activity and that orthologous gene members in flies and moths evolved new functions through divergent evolution. In the Yponomeuta genus (Figure 2), a multifunctional conserved FAR accounts for producing a broad range of alcohol pheromone precursors across sister species. The pgFAR (pheromone-gland specific) exhibits a chain-length preference for C14-acyls and plays a role in adjusting the ratios between all ∆11-alcohol precursors in proportions matching the final blend observed in females (Figure 3).

Figure 3: Pheromone biosynthesis and ratio specificity in Yponomeuta padella (Lepidoptera: Yponomeutidae). (A)  Ermine moth posing on hawthorn (Crataegus sp., Rosaceae). (B) Pheromone biosynthetic pathway towards the ∆11-unsaturated components in the Yponomeuta genus. (C) GC-MS analysis of ∆11-monounsaturated fatty-acyl intermediates produced by functional expression of the Ypa-∆11-desaturase in the YEpOLEX expression vector and the ole 1 elo 1 strain of the yeast Saccharomyces cerevisiae. The chromatogram trace represents dimethyl disulfide adducts (DMDS) from methanolyzed yeast extracts. In vitro, the ∆11-desaturase catalyses the introduction of a double bond at the eleventh carbon atom (characteristic ion at m/z 245) in several natural yeast fatty acids from C14 to C22. The Z11-14, E11-14 and Z11-16 acyls are produced in a 1.1:0.8:100 ratio, respectively. (D) The dot areas are proportional (%) to the ∆11-desaturase and pgFAR substrate preferences and to the final ratio between components. The reverse chain-length preference of the ∆11-desaturase and pgFAR for acyl substrates with 14 or 16 carbon atoms allows adjusting the final blend ratio. Liénard and Löfstedt (2010). Commun Integr Biol. 2010, 3(6):586-8. doi: 10.4161/cib.3.6.13177

In the Ostrinia genus, studying FAR biosynthetic enzymes between two closely related species Ostrinia furnacalis (Ofur) and O. nubilalis (Onub) pointed out the role of a single amino acid position causing a noticeable qualitative shift in pheromone composition between closely related species (Ofur F453C is unable to reduce the Z9-14:Acyl and Z12-14:Acyl precursors), while most additional non-synonymous changes between the two proteins contribute minor quantitative variation ensuring ratio specificity (Figure 4). Most pgFARs investigated until now appear to be an essential element in adjusting species-specific pheromone composition, alone or in combination with upstream biosynthetic desaturases. 

Figure 4: Cluster dendogram showing the distances between mutations at each of the eight amono acid selected (A). Pie charts show the ratio of components produced for WT (wild-type) and mutant pgFARs. One mutation is sufficient to determine the ability to catalyze the reduction of Z9-14 and Z12-14 Acyl precursors. The change in activity is remarkable, but the fine tuning of the proportions requires additional small effect mutations, which was demonstrated by producing double mutants carrying the F453 mutation plus one other of the seventh mutation. Biological activities were assessed by testing the function of each enzyme in a yeast-expression system followed by quantification of resulting enzyme activity using Gas Chromatography and Mass Spectrometry. Lassance, Liénard et al. PNAS (2013),110: 3967-72.

Altogether, the evolution of new ∆9, ∆6, ∆8,11, and substrate-specific ∆11-Acyl-CoA desaturases has promoted the structural diversification leading to variable pools of unsaturated fatty acid precursors for downstream pheromone biosynthetic enzymes. Fatty acyl reduction is the requisite downstream biosynthetic step at the interplay between precursors and final volatile molecules. One could question the chance for a new structure to become part of the released pheromone blend unless downstream enzymes are further capable of converting them in volatile components? As a first step to answer this question, we found out that pgFARs are capable of converting odd chemical structures in vitro (i.e. even rare ∆12 compounds ) suggesting that they are preadapted to accomodate upstream modifications in the biosynthetic machinery. In the light of multi-component blend usage in the Lepidoptera,  the pgFAR biochemical flexibility may well be a very useful feature in moth to  promote phenotypic transitions in the mating signals in Lepidoptera. Altogether, this research highlights a series of major genetic determinants and proximate mechanisms promoting variation among lepidopteran sexual signals, which forms the basis of reproductive isolation.