Migratory locusts may be guided toward plants by a molecular transporter best known for giving insect eyes their color, according to a study that links the protein Scarlet to olfactory behavior, brain chemistry and an unusual form of gene regulation. Researchers found that reducing activity of the locust gene scarlet(LOCMI17149) weakened the insects’ attraction to plant odors, including indole and β-ionone, while also altering the levels of two tryptophan-derived compounds in the brain. The work suggests that Scarlet does more than shuttle pigment precursors into cellular granules: in the nervous system, it may help tune the chemical signals that allow locusts to recognize food plants. The study also identifies a previously unreported regulatory mechanism involving an R-loop, a temporary RNA-DNA structure that forms within the gene itself. By influencing how tightly the local DNA is packaged and how efficiently RNA polymerase transcribes the gene, this structure appears to act as a molecular brake on olfactory sensitivity.
Scarlet belongs to the ABCG subfamily of ATP-binding cassette transporters, membrane proteins that use energy from ATP to move molecules across biological membranes. In insects, Scarlet typically works with another transporter called White to carry 3-hydroxykynurenine, or 3-HK, a metabolite of tryptophan, into pigment granules. There, 3-HK contributes to the production of ommochrome pigments responsible for characteristic red, brown or scarlet eye colors. Mutations in scarlet produce striking eye-color phenotypes in fruit flies, beetles, silkworms, moths and other insects. Yet earlier observations hinted that the protein also has neurological functions. In fruit flies, defects in scarlet have been associated with changes in brain biogenic amines, movement abnormalities and degeneration of dopamine-producing neurons. The new study extends that picture to the migratory locust, Locusta migratoria, an insect whose survival depends on interpreting odors released by host plants, neighboring locusts and potential threats. The researchers found four scarlet-like genes in the locust genome, but one, LOCMI17149, was especially abundant in the antennae and brain.
To test whether this gene was functionally involved in smell, the team used RNA interference, a technique that selectively reduces gene activity by introducing double-stranded RNA matching the target transcript. Fourth-instar locust nymphs received the treatment in their head cavities and were examined 72 hours later. Quantitative PCR showed that scarlet(LOCMI17149) messenger RNA had fallen significantly in both antennae and brains, while western blotting confirmed a corresponding decrease in the Scarlet protein. The behavioral consequences were substantial. In a Y-shaped olfactometer, where one arm carried plant-derived odors and the other carried clean mineral oil, the proportion of control locusts moving toward plant volatiles was about 53 percent. After Scarlet knockdown, that figure dropped to 22 percent. Responses to indole, a nitrogen-containing compound released by plants, fell from 58 percent in controls to 26 percent after knockdown. Attraction to β-ionone, a volatile associated with plant aromas, was also significantly reduced. The insects were not simply losing the ability to move; the change was specifically reflected in odor-guided choice.
Electrical recordings provided evidence that the behavioral effect began in the sensory system. Electroantennography measures voltage changes generated across an insect antenna when odorant receptors respond to a chemical. Antennae from locusts with reduced Scarlet activity produced weaker electrical responses to both indole and β-ionone than control antennae. The researchers then recorded local field potentials from the antennal lobe, the first major olfactory-processing center in the brain. These signals represent the coordinated electrical activity of populations of neurons. Exposure to either odor normally produced a rise in antennal-lobe activity, but Scarlet knockdown reduced both the overall signal and its power across the recorded frequency range. Power spectral density analysis, which tracks how neural energy is distributed over frequencies and time, showed diminished odor-evoked signatures rather than a general disruption caused by the mineral-oil carrier. Together, the results place Scarlet at multiple stages of olfactory processing, from the antenna’s initial response to the brain’s integration of odor information.
The researchers connected this sensory deficit to a metabolic shift involving 3-HK and xanthurenic acid, or XA. Scarlet is thought to transport 3-HK into intracellular vesicles or granules, where the compound can be stored or processed. When Scarlet was reduced, the amount of 3-HK detected in locust brains decreased, while XA levels rose sharply. One possible explanation is that impaired transport leaves more 3-HK in the cytosol, where it becomes available to enzymes that convert it into XA. The study tested whether this pathway, rather than Scarlet itself, could account for the behavioral effect. Knocking down kynurenine 3-monooxygenase, an enzyme involved in producing 3-HK, changed enzyme expression but did not significantly alter locust preferences for wheat-seedling extracts, indole or β-ionone. By contrast, injecting XA directly into the head cavities weakened attraction to plant odors and both test chemicals. XA-treated antennae also showed smaller electrical responses, and antennal-lobe recordings revealed reduced odor-evoked activity. The authors propose that excessive XA acts as a neuroactive signal that dampens olfactory transmission, potentially by interfering with glutamate signaling. XA has been reported to influence group II metabotropic glutamate receptors and vesicular glutamate transport, although its precise action in the locust olfactory circuit remains to be established.
The study’s most unexpected finding concerns how the locust controls Scarlet production. An R-loop forms when a newly synthesized RNA strand remains paired with one strand of DNA, displacing the other DNA strand and creating a three-stranded nucleic-acid structure. These structures are not merely accidental by-products of transcription. They can influence the initiation, pausing and termination of transcription, alter DNA repair and reshape nearby chromatin. Using computational prediction followed by DNA-RNA immunoprecipitation, the researchers identified a specific R-loop, which they named Scarlet-RL, within an intron of the scarlet(LOCMI17149) gene. The hybrid was sensitive to RNase H, an enzyme that selectively degrades the RNA component of RNA-DNA hybrids, confirming that it had the molecular properties of an R-loop. Additional experiments indicated that the RNA strand came from the sense transcript produced during scarlet transcription. When the researchers reduced RNase H1, the enzyme responsible for removing RNA from such hybrids, Scarlet-RL accumulated in locust brains.
More R-loop did not mean more gene activity. Instead, RNase H1 knockdown reduced both the precursor form of scarlet RNA and the mature messenger RNA, as well as the Scarlet protein. The precursor result is important because it indicates that the R-loop affects new transcription rather than merely destabilizing an already completed message. To manipulate the structure more directly, the team designed short locked nucleic-acid antisense oligonucleotides, or LNA-ASOs, that bind the RNA strand of Scarlet-RL. These chemically stabilized molecules are designed to recognize RNA with high affinity. Two of the three tested ASOs increased scarlet expression, with ASO1 producing the strongest effect. Injections of ASO1 raised nuclear precursor RNA, cytoplasmic mature RNA and protein levels in the brain. Blocking transcription with Actinomycin D showed that the altered RNA abundance could not be explained by a major change in messenger-RNA stability. Instead, the evidence points to Scarlet-RL controlling the birth of new transcripts.
Chromatin experiments provided a potential explanation. RNA polymerase II, the enzyme that copies protein-coding genes into RNA, was less abundant at the Scarlet-RL-containing intron when R-loop levels were increased by RNase H1 knockdown. Reducing Scarlet-RL with ASO1 had the opposite effect and increased polymerase occupancy. The R-loop was also associated with changes in histone modifications, chemical tags on DNA-packaging proteins that help determine whether a gene is accessible to the transcription machinery. Accumulation of Scarlet-RL reduced H3S10 phosphorylation, a mark generally associated with transcriptionally active chromatin, and increased H3K9 dimethylation, a repressive mark that can promote a less accessible chromatin state. ASO1 reversed these changes at the key intron. The findings suggest a chain of events in which an intronic RNA-DNA hybrid alters local histone chemistry, limits recruitment or progression of RNA polymerase II, reduces Scarlet production and ultimately shifts brain XA levels toward a state that suppresses odor responses. The investigators note that the enzyme responsible for the proposed H3K9 dimethylation has not yet been identified; G9a is a candidate, but direct recruitment remains to be demonstrated.
The work offers a new molecular connection between epigenetic regulation and insect behavior, while also highlighting important limits. RNase H1 knockdown can cause R-loops to accumulate throughout the genome, so its effects are not necessarily restricted to the scarlet locus. The more targeted ASO experiments strengthen the case for a direct role of Scarlet-RL, but the authors acknowledge that antisense oligonucleotides can have off-target effects and that mismatch controls or transcriptome-wide analyses will be needed to confirm specificity. The study also does not establish whether Scarlet always operates with White in locust olfactory tissues, nor does it resolve whether the transporter acts mainly in olfactory receptor neurons, supporting cells or glial cells. Even so, the results suggest that a protein historically associated with eye pigmentation can influence sensory decisions through a metabolite-sensitive neural pathway. Because Scarlet genes are widespread across insect groups, similar mechanisms may exist in other pests. A species-specific R-loop or its associated chromatin machinery could eventually become a target for RNA-based interventions designed to disrupt how locusts locate vegetation—although translating that possibility into field control will require substantial testing beyond the laboratory.
