Odor-Specific Deactivation Defects in a Drosophila Odorant-Binding Protein Mutant

Abstract

Insect odorant-binding proteins (OBPs) are a large, diverse group of low-molecular weight proteins secreted into the fluid bathing olfactory and gustatory neuron dendrites. The best-characterized OBP, LUSH (OBP76a) enhances pheromone sensitivity enabling detection of physiological levels of the male-specific pheromone, 11-cis vaccenyl acetate. The role of the other OBPs encoded in the Drosophila genome is largely unknown. Here, using clustered regularly interspaced short palindromic repeats/Cas9, we generated and characterized the loss-of-function phenotype for two genes encoding homologous OBPs, OS-E (OBP83b) and OS-F (OBP83a). Instead of activation defects, these extracellular proteins are required for normal deactivation of odorant responses to a subset of odorants. Remarkably, odorants detected by the same odorant receptor are differentially affected by the loss of the OBPs, revealing an odorant-specific role in deactivation kinetics. In stark contrast to lush mutants, the OS-E/F mutants have normal activation kinetics to the affected odorants, even at low stimulus concentrations, suggesting that these OBPs are not competing for these ligands with the odorant receptors. We also show that OS-E and OS-F are functionally redundant as either is sufficient to revert the mutant phenotype in transgenic rescue experiments. These findings expand our understanding of the roles of OBPs to include the deactivation of odorant responses.

Keywords: olfaction; olfactory; perireceptor.

Figure 1

Generation of OS-E/F null mutants. (A) Map of the OS-E/F genomic region on the right arm of the third chromosome. CRISPR-mediated replacement of the OS-E and OS-F genes with 3xP3>dsRed (black rectangle) is depicted. Solid triangles indicate the position of CRISPR/Cas9 cleavage sites. The dashed lines denote the regions of homology upstream and downstream of the OBP genes that were cloned into the donor vector (see Materials and Methods for details). Labeled arrows (a–d) indicate the positions and orientations of primers for PCR reactions used to identify correct integration of the DsRed gene (black rectangle) into the OS-E/F locus. Unlabeled arrows indicate gene-specific primers used to determine the presence of the OS-E and OS-F genes. (B) Agarose gel image of PCR fragments generated with the primers depicted in (A). PCR fragment sizes from controls and OS-E/F mutants confirm correct integration of the DsRed gene, and loss of the OS-E and OS-F genes in the mutant. Markers in left lane are a 1-kb ladder (Thermo Fisher Scientific). CRISPR, clustered regularly interspaced short palindromic repeats; OBP, odorant-binding protein.

Figure 2

Spontaneous and evoked activity from wild-type and OS-E/F mutants. (A) Cartoon of two trichoid and three intermediate sensilla classes normally expressing OS-E and OS-F, depicting the neurons expressing the characteristic odorant receptors defining each neuron class. (B) Average spontaneous activity of individual trichoid and intermediate neurons of the indicated genotypes in the absence of odorants. Or2a and Or43a, as well as Or88a and Or65abc, neuron responses were combined due to similarities in spike amplitudes. Or88a and Or65abc spontaneous rates were determined in the Or47b mutant (Wang et al. 2011). Or83c wild-type = 4.34 ± 0.8 spikes/s; Or83c OS-E/F mutant = 8.36 + 1.93 spikes/s (P = 1.4 × 10⁻⁷). Or88a/Or65abc wild-type = 10.46 ± 0.97 spikes/s; Or88a/Or65abc OS-E/F mutant = 20.31 ± 2.47 spikes/s (P = 0.0003). Or47b wild-type = 46.97 ± 3.00 spikes/s; Or47b OS-E/F mutant = 74.25 ± 7.42 spikes/s (P = 0.00029). n = 10–20 for each genotype. (C) Odor-induced responses of trichoid and intermediate neurons to the best-known activating ligands for each neuron type (n = 5–22) (Hallem et al. 2004; Galizia and Sachse 2010; Dweck et al. 2013; Ronderos et al. 2014; Pitts et al. 2016). Or83c wild-type Δ spikes/s = 35.9 ± 1.94; Or83c OS-E/F mutant Δ spikes/s = 64.65 ± 2.8 (P = 1.15 × 10⁻⁷). Or47b wild-type Δ spikes/s = 72.38 ± 4.48; Or47b OS-E/F mutant Δ spikes/s = 28.45 ± 5.73 (P = 3.21 × 10⁻⁶). Odorants and dilutions used to test each neuron were as follows: Or13a, 10% 1-octen-3-ol; Or83c, 1% farnesol; Or23a, 10% cyclohexanone; Or19a, 1% limonene, Or2a/Or43a, 10% benzaldehyde; Or67d, 1% cVA; and Or47b, 10% trans-2-hexenal. Error bars indicate SEM. All P-values determined by one-way ANOVA with post hoc Tukey’s test. (D) Responses of at4 neurons to known activating ligands (Dweck et al. 2015; Pitts et al. 2016). All odorants were used at a 10% dilution on the filter paper (see Materials and Methods for details). Or47b response to trans-2-hexenal: wild-type Δ spikes/s = 72.38 ± 4.48, OS-E/F mutant Δ spikes/s = 28.45 ± 5.73, and genomic rescue Δ spikes/s = 66.1 ± 9.91. ANOVA P-values: wild-type to OS-E/F mutant = 3.21 × 10⁻⁶, wild-type to genomic rescue = 0.51, and OS-E/F mutant to genomic rescue = 0.0033. n = 10–22. Error bars indicate SEM. All P-values determined by one-way ANOVA with post hoc Tukey’s test. cVA, 11-cis vaccenyl acetate.

Figure 3

Deactivation kinetics are abnormal in OS-E/F mutants to a subset of odorants. (A) Comparison mutant aI2a neuronal responses to farnesol. (B) Comparison of wild-type and OS-E/F mutant Or67d neurons to cVA. C) Comparison of wild-type and OS-E/F mutant at4 neuronal responses to trans-2-hexenal. In (A–C), representative traces are shown on the left, time courses of activation and deactivation as measured by binning spikes in 0.5-sec bins are shown on the right. Note delayed return to baseline activity in the mutants. (D) Time constant τ of deactivation of all tested olfactory neurons. Or83c wild-type = 99.2 ± 4.6 msec; Or83c OS-E/F mutant = 472.1 ± 58.2 msec (P = 5.14 × 10⁻⁶). Or67d wild-type = 153.7 ± 22.4 msec; Or67d OS-E/F mutant = 971.7 ± 112.5 msec (P = 1.69 × 10⁻⁶). Or47b wild-type = 725.69 ± 81 msec; Or47b OS-E/F mutant = 1830 ± 250.8 msec (P = 6.95 × 10⁻⁶). n = 10–20. Error bars indicate SEM. All P-values determined by two-tailed Student’s t-test. cVA, 11-cis vaccenyl acetate.

Figure 4

Odorant specificity of the Or83c deactivation defect. (A–C) representative traces are shown on the left, time courses of responses assayed by binning spikes in 500-msec bins on the right. (A) Responses of Or83c neurons from wild-type, and OS-E/F and Or83c mutant flies in response to 1% farnesol. Deactivation time constant (τ) = 92.3 ± 16.2 msec for wild-type and 404.0 ± 80 msec for OS-E/F mutants (P = 0.0015, n = 10). (B) Responses of wild-type, and OS-E/F and Or83c mutant flies to 0.1% farnesol. Despite lower peak activation, a prominent deactivation defect is still present in the mutants. τ = 109.4 ± 42.2 msec for wild-type and 870.5 ± 152 msec for OS-E/F mutants (P = 0.0014). n = 5. (C) Responses of wild-type, and OS-E/F and Or83c mutant flies to 10% 3-hexanol. No differences in deactivation are apparent between wild-type and OS-E/F mutants. Or83c mutants do not respond to 3-hexanol. τ = 115.2 ± 14.5 msec for wild-type, 109.6 ± 32.9 msec for OS-E/F mutants (P = 0.88). n = 5. Error bars represent SEM. All P-values determined by two-tailed Student’s t-test.

Figure 5

Activation kinetics for farnesol and cVA are not affected by loss of OS-E and OS-F. (A) Representative 500-msec traces from wild-type and OS-E/F mutant-aI2 sensilla responses to 0.01% farnesol. (B) Representative 3-sec traces from wild-type, OS-E/F mutant, and lush¹ Or67d neurons to 0.3% cVA. (C) Analysis of latency to activation for neurons and genotypes indicated. Or83c activation latency for wild-type is 228.3 ± 2.1 msec; for OS-E/F mutants, 225.8 ± 4.9 msec (P = 0.65). n = 10. Latency for Or67d activation is not significantly different in wild-type and OS-E/F mutants. Or67d activation latency in wild-type is 223 ± 34 msec and for OS-E/F mutants 299.8 ± 20 msec (P = 0.99). (D) Wild-type Or67d neurons exposed to 100% cVA have a latency for activation of 109.8 ± 2.2 msec. For lush¹ mutants, the latency for activation is 1941.7 ± 414.5 msec. lush¹ mutants are significantly different from wild-type (P = 4.25 × 10⁻⁵), n = 5–10 for each genotype. Error bars indicate SEM. All P-values are determined by two-tailed Student’s t-test. cVA, 11-cis vaccenyl acetate.

Figure 6

No genetic interactions between Snmp1 and OS-E/F mutants. (A) Representative 5-sec traces from wild-type, Snmp1 mutants, OS-E/F double mutants, or Snmp1/OS-E/F triple mutants in response to 100% farnesol. (B) Time constants of deactivation with the genotypes indicated: wild-type = 165.16 ± 27.1 msec, Snmp1 mutant = 396.7 ± 58.9 msec, OS-E/F mutant = 882.31 ± 121.92 msec, and Snmp1/OS-E/F mutant = 768.31 ± 88.7 msec. ANOVA P-values: wild-type to Snmp1 mutant = 0.0031; wild-type to OS-E/F mutant = 9.27 × 10⁻⁵; wild-type to Snmp1/OS-E/F triple mutant = 2.91 × 10⁻⁵; Snmp1 mutant to OS-E/F mutant = 0.006; Snmp1 mutant to Snmp1/OS-E/F mutant = 0.0064; and OS-E/F mutant to Snmp1/OS-E/F mutant = 0.46. n = 6–7 for each genotype. Error bars indicate SEM. All P-values were determined by one-way ANOVA with post hoc Tukey’s test.

Figure 7

OS-E and OS-F are functionally redundant for rapid neuronal deactivation. (A) Map of the OS-E/F gene region with black rectangles depicting coding regions. Arrows indicate transcription start sites. The lines beneath the map depict the DNA deletions that were used to evaluate the function of individual OBP genes. Breaks in the lines showing individual rescues show where the coding regions of either OS-E or OS-F were excised from the genomic rescue construct (see Materials and Methods for details). (B) Time constants of deactivation for Or83c, Or67d, and Or47b neurons with the genotypes indicated. Delayed deactivation present in the OS-E/F mutants is reversed by all three forms of the rescuing transgene. Deactivation time constants for Or83c neurons to farnesol (left side of graph): wild-type = 96.36 ± 19.24 msec, OS-E/F mutant = 375.98 ± 83.94 msec, genomic rescue = 75.34 ± 12.31 msec, OS-E rescue = 76.54 ± 4.15 msec, and OS-F rescue = 63.79 ± 9.48 msec. ANOVA P-values: wild-type to OS-E/F mutant = 0.0013, OS-E/F mutant to genomic rescue = 0.001, OS-E/F mutant to OS-E rescue = 0.0013, and OS-E/F mutant to OS-F rescue = 0.003, all nonsignificant P-values = 0.99. For Or67d responses to cVA (center part of graph): wild-type = 160.73 ± 17.23, OS-E/F mutant = 1249.58 ± 246.56, genomic rescue = 310.43 ± 80.34, OS-E rescue = 257.52 ± 50.11, and OS-F rescue = 322.24 ± 88.65. ANOVA P-values: wild-type to OS-E/F mutant = 9.8 × 10⁻⁵, OS-E/F mutant to genomic rescue = 0.0019, OS-E/F mutant to OS-E rescue = 0.0016; OS-E/F mutant to OS-F rescue = 0.0013, wild-type to genomic rescue = 0.90, wild-type to OS-E rescue = 0.98, wild-type to OS-F rescue = 0.94, genomic rescue to OS-E rescue = 0.99, genomic rescue to OS-F rescue = 0.99, and OS-E rescue to OS-F rescue = 0.99. For Or47b responses to trans-2-hexenal (right part of graph): wild-type = 676.42 ± 64.24 msec, OS-E/F mutant = 1882.5 ± 328.9 msec, genomic rescue = 629.12 ± 102.74 msec, OS-E rescue = 699.18 ± 184.7 msec, and OS-F rescue = 499.1 ± 99.4 msec. ANOVA P-values: wild-type to OS-E/F mutant = 2.5 × 10⁻⁴, OS-E/F mutant to genomic rescue = 3.8 × 10⁻⁴, OS-E/F mutant to OS-E rescue = 0.004, OS-E/F mutant to OS-F rescue = 0.001, wild-type to genomic rescue = 0.70, wild-type to OS-E rescue = 0.90, wild-type to OS-F rescue = 0.14, genomic rescue to OS-E rescue = 0.73, genomic rescue to OS-F rescue = 0.40, and OS-E rescue to OS-F rescue = 0.39. n = 5–10. Error bars indicate SEM. All P-values determined by one-way ANOVA with post hoc Tukey’s test. (C) Extra copies of OS-E and OS-F do not alter deactivation kinetics. Time constants of deactivation for Or83c, Or67d, and Or47b neurons with the genotypes indicated. Deactivation time constants for Or83c neurons to farnesol (left side of graph): wild-type = 87.13 ± 9.2 msec, 4X OS-E/F= 99.6 ± 9.9 msec, P = 0.43. For Or67d responses to cVA (center part of graph): wild-type = 130.1 ± 22.9 msec, 4X OS-E/F = 124.3 ± 14.2, P = 0.83. For Or47b responses to trans-2-hexenal (right part of graph): wild-type = 752.65 ± 64.6 msec, 4X OS-E/F = 658.7 ± 38.8 msec, P = 0.52. cVA, 11-cis vaccenyl acetate; lncRNA, long noncoding RNA; OBP, odorant-binding proteins.