The taste of ribonucleosides: Novel macronutrients essential for larval growth are sensed by Drosophila gustatory receptor proteins

Abstract

Animals employ various types of taste receptors to identify and discriminate between different nutritious food chemicals. These macronutrients are thought to fall into 3 major groups: carbohydrates/sugars, proteins/amino acids, and fats. Here, we report that Drosophila larvae exhibit a novel appetitive feeding behavior towards ribose, ribonucleosides, and RNA. We identified members of the gustatory receptor (Gr) subfamily 28 (Gr28), expressed in both external and internal chemosensory neurons as molecular receptors necessary for cellular and appetitive behavioral responses to ribonucleosides and RNA. Specifically, behavioral preference assays show that larvae are strongly attracted to ribose- or RNA-containing agarose in a Gr28-dependent manner. Moreover, Ca2+ imaging experiments reveal that Gr28a-expressing taste neurons are activated by ribose, RNA and some ribonucleosides and that these responses can be conveyed to Gr43aGAL4 fructose-sensing neurons by expressing single members of the Gr28 gene family. Lastly, we establish a critical role in behavioral fitness for the Gr28 genes by showing that Gr28 mutant larvae exhibit low survival rates when challenged to find ribonucleosides in food. Together, our work identifies a novel taste modality dedicated to the detection of RNA and ribonucleosides, nutrients that are essential for survival during the accelerated growth phase of Drosophila larvae.

Fig 1. Larval preference for ribose and RNA is not mediated by sugar Gr genes.

Two-choice preference assays for arabinose, ribose, deoxyribose, and RNA (panel a and c) and survival on these chemicals and nutritious sugars (panel b). (a) Preference for arabinose is independent on various sugar Gr genes (n = 12–28). The underlying data can be found in S1 Data. (b) Comparison of survival of w¹¹¹⁸ larvae when kept on different substrates (n = 3–8). After 72 hours, approximately 50% of the larvae survive on agarose-only substrate (median survival, dashed line). For simplicity, significant differences are only indicated for median survival time. Data are represented as mean ± SEM. “*” represents significant difference between the larval survival on different substrates and agarose (two-tailed Mann-Whitney U test, p < 0.05). Reduced survival rate of larvae kept on arabinose and deoxyribose might be due to interference of these chemicals with sugar metabolism. The underlying data can be found in S2 Data. (c) Larvae show strong preference for ribose (n = 12–36) and RNA (n = 6–36) when lacking Gr43a or the 8 sGr. Larvae are not attracted to deoxyribose (n = 6–24). As for fructose [4], ΔsGr larvae showed stronger preference for ribose than wild-type larvae. Concentration of all substrates was 100 mM in 1% agarose, except RNA (0.5 mg/mL in 1% agarose). Genotypes: w¹¹¹⁸ (control), w¹¹¹⁸; Gr43a^GAL4/Gr43a^{GAL 4}(ΔGr43a), and w¹¹¹⁸; ΔGr61a ΔGr64a-f/ΔGr61a ΔGr64a-f (ΔsGrs). The underlying data can be found in S1 Data. Gr, gustatory receptor; PREF, preference index; sGr, sugar Gr gene.

Fig 2. Expression of the 6 Gr28 genes in third-instar larvae.

(a) Graphic summary of Gr28 gene expression. Cells and neurons with their axons expressing the respective GAL4 driver are all shown in green. Brain is shown in grey, and the digestive system—including the pharynx, PV, and gut—are outlined. (b) Live GFP imaging of the larval head, showing expression of 3 genes (Gr28a, Gr28b.a, and Gr28b.e) in neurons of the TO. Gr28b.d is expressed in neurons of the DPS and VPS organs, while Gr28a is also expressed in the PPS organ. Neither Gr28a nor Gr28b.d are co-expressed with Gr43^GAL4 (see S1 Fig). Number of larvae with GFP positive taste neurons/total number of larvae analyzed were 7/7 for Gr28a, 4/4 for Gr28b.a, 0/5 for Gr28b.b, 0/7 for Gr28b.c, 5/5 for Gr28b.d, and 5/5 for Gr28b.e. (c) View of the brain and parts of the ventral nerve cord, showing different degrees of expression for each of the 6 Gr28 genes. The brains were stained with anti-GFP antibody (green) and counterstained with nc82 antibody (red). Number of larvae with GFP antibody–positive staining in the brain-VNC/number of brains analyzed were 3/3 for Gr28a, 5/5 for Gr28b.a, 3/3 for Gr28b.b, 5/5 for Gr28b.c, 3/3 for Gr28b.d, and 6/6 for Gr28b.e. (d) Live GFP imaging of the PV and midgut, showing expression of all Gr28 genes with the exception of Gr28b.d. Expression of Gr28b.a and Gr28b.e is broad and includes the PV and midgut, while expression of Gr28a and Gr28b.b is defined to a smaller area of the gut only. Number of larvae with GFP-positive cells/total number of larvae analyzed were 5/5 for Gr28a, 4/4 for Gr28b.a, 3/3 for Gr28b.b, 5/5 for Gr28b.c, 0/5 for Gr28b.d, and 4/4 for Gr28b.e. (e) Summary of tissues expressing each of the 6 Gr28 genes. Genotypes were w; UAS-mCD8GFP/Gr28x-GAL4, such that x refers to indicated Gr-Gal4 driver. Scale bar is 100 μm. For live imaging (panel b and d), at least 5 larvae for each genotype were analyzed, and GFP cells in taste sensilla and the gut were observed in each case for Gr28a, Gr28ba, Gr28b.e, Gr28b.d, and Gr28b.e; for staining (panel c), at least 3 brains for each genotype were analyzed, with GFP-positive neurons observed in each case. The images are good representatives of these experiments. DPS, dorsal pharyngeal sensory; GFP, green fluorescent protein; Gr28, gustatory receptor subfamily 28; PPS, posterior pharyngeal sensory; PV, proventriculus; TO, terminal organ; VPS, ventral pharyngeal sensory.

Fig 3. Genes of the Gr28 locus mediate larval taste preference for ribose, RNA, and arabinose.

Two-choice feeding assays of wild-type and Gr28 mutant larvae. (a) Larvae require the Gr28 genes for taste preference for arabinose (n = 12–51), ribose (n = 12–36), and RNA (n = 21–36); Genotypes: w¹¹¹⁸ (Control), w¹¹¹⁸; ΔGr28/ΔGr28 (ΔGr28) and ΔGr28/ΔGr28; Gr28 genomic rescue/+ (ΔGr28 g-rescue). (b) Single Gr28 genes rescue taste preference for ribose in ΔGr28 homozygous mutant larvae. Genotypes were w¹¹¹⁸ (lane 1), w¹¹¹⁸; ΔGr28/ΔGr28 (2), w¹¹¹⁸;ΔGr28/ΔGr28; Gr28a-GAL4/+ (3), w¹¹¹⁸;ΔGr28/ΔGr28; Gr28a-GAL4/UAS (4, 6, 8, 10, 12, 14), and ΔGr28/ΔGrGr28; +/UAS (5, 7, 9, 11, 13, 15) such that UAS represents indicated transgene (n = 12–36). (c) Gr28 mutant larvae expressing a single Gr28 gene in fructose-sensing (Gr43a^GAL -expressing) neurons show preference for ribose. Genotypes: w¹¹¹⁸ (lane 1), w¹¹¹⁸; ΔGr28 Gr43^GAL4/ΔGr28 Gr43^GAL4 (2), and w¹¹¹⁸; ΔGr28 Gr43^GAL4/ΔGr28 Gr43^GAL4; UAS/+, such that UAS represent indicated transgene (n = 12–30). Each bar represents the mean ± SEM of two-choice preference responses. Concentrations were 100 mM (arabinose and ribose) or 0.5 mg/mL (RNA) in 1% agarose. Red “*” represents significant difference between indicated genotype and w¹¹¹⁸ control (two-tailed Mann-Whitney U test, p < 0.05). Green “*” represents significant difference between indicated genotype and ΔGr28 Gr43a^GAL4 double mutant (w¹¹¹⁸; ΔGr28 Gr43^GAL4/ΔGr28 Gr43^GAL4). Two-tailed Mann-Whitney U test, p < 0.05). The underlying data can be found in S1 Data.

Fig 4. RNA and ribose are ligands for single Gr28 proteins.

(a) Terminal taste neurons expressing the Ca²⁺ sensor CaMPARI require Gr28a in order to respond to ribose and RNA (n = 4–9). Genotypes: w¹¹¹⁸; UAS-CaMPARI/+; Gr28a-GAL4/+ (control), w¹¹¹⁸; ΔGr28 UAS-CaMPARI/ΔGr28; Gr28a-Gal4/+ (ΔGr28), and w¹¹¹⁸; ΔGr28 UAS-CaMPARI /ΔGr28; Gr28a-GAL4/UAS-Gr28a (Gr28a rescue). (b) Expression of single Gr28 genes conveys ribose and RNA responses to fructose-sensing pharyngeal taste neurons (n = 4–13). Genotypes: w¹¹¹⁸; UAS-CaMPARI Gr43a^GAL4/+ (control), w¹¹¹⁸; Gr43^GAL4 UAS-CaMPARI/+; UAS-Gr28a/+ (Gr28a), w¹¹¹⁸; Gr43^GAL4 UAS-CaMPARI/+; UAS-Gr28b.a/+ (Gr28b.a), w¹¹¹⁸; Gr43^GAL4 UAS-CaMPARI/+; UAS-Gr28b.e/+ (Gr28b.e). Final concentration of all substrates was 100 mM in water except for RNA (0.5 mg/mL). Representative images of the indicated genotypes are shown above the graphs. Scale bar is 10 μm. Each bar represents the mean ± SEM of ratios of red and green fluorescence intensities. “*” represents significant differences between the preexposure (no PC light, no chemical) group and a substrate group (two-tailed Mann-Whitney U test, p < 0.05). The underlying data can be found in S3 Data. PC, photoconversion.

Fig 5. Ribonucleosides are essential nutrients for rapid larval growth and survival.

(a) Growth time in days from hatching of the first-instar larvae to eclosion (left) and survival rate (right) of larvae raised in different media shows that inosine and uridine are essential components. Larvae raised on HM grow slightly slower than, but have the same survival rate as, larvae raised on SCF. Replacing ribonucleosides with RNA (0.5 mg/mL) in HMΔ restores both growth time and survival rate, while replacing it with equimolar concentration of ribose fails to do so. Each bar represents the mean ± SEM (n = 4). Bars with different letters represent significant differences (two-tailed Mann-Whitney U test, p < 0.05). Genotype: w¹¹¹⁸. The underlying data can be found in S4 Data. (b) CaMPARI imaging of TO taste neurons shows that inosine and uridine, but none of the 3 other ribonucleosides, are potent ligands for Gr28 neurons. Uridine, cytidine (100 mM), and inosine (50 mM) were dissolved in water, while guanosine and adenosine were dissolved in DMSO and presented at concentration of 25 mM and 50 mM in water containing 25% and 10% f.c. DMSO, respectively. Each bar represents the mean ± SEM of ratios of red and green fluorescence intensities (n = 5–19). “*” represents significant differences between the preexposure (untreated) group to the groups with the indicated ligands applied (two-tailed Mann-Whitney U test, p < 0.05). Genotype: w¹¹¹⁸; UAS-CaMPARI/Gr28a-GAL4. The underlying data can be found in S3 Data. (c) Two-choice preference assay shows that larvae require the Gr28 genes to exhibit preference for uridine (50 mM, n = 12–24) and inosine (100 mM, n = 12–18). “*” represents significant difference between the genotypes (two-tailed Mann-Whitney U test, p < 0.05). All the genotypes are compared to control. Genotypes: w¹¹¹⁸, w¹¹¹⁸; ΔGr28/ΔGr28 and w¹¹¹⁸; ΔGr28/ΔGr28; genGr28/+. The underlying data can be found in S1 Data. f.c., final concentration; HM, holidic medium; SCF, standard cornmeal food; TO, terminal organ.

Fig 6. Larvae require Gr28 genes for efficient growth and survival when presented with HM and HMΔ food.

(a) About 40 eggs were deposited in 21-well microtiter plates containing 1 of 3 different foods: all wells containing HM (black; left), HMΔ (gray; middle), or a mixture of the two (12 HMΔ and 9 HM; right). Plates with either only HM or HMΔ medium were used to determine survival rate for complete (HM) or ribonucleoside-deficient (HMΔ) food. (b) Survival is displayed as percentage of flies hatched after eggs were deposited onto plate. For statistical analysis, survival in different foods was either compared across the same genotype (Control [black]: w¹¹¹⁸, ΔGr28 [red]: w¹¹¹⁸; ΔGr28/ΔGr28, and ΔGr28 g-rescue [green]: w¹¹¹⁸; ΔGr28/ΔGr28; genGr28/+), or different genotypes were compared against the same mixed food (light–dark checkered pattern). Each bar represents the mean ± SEM (n = 5–6). Bars with different letters represent significant difference (two-tailed Mann-Whitney U test, p < 0.05). The underlying data can be found in S4 Data. HM, holidic medium.