Cellular dissection of circadian peptide signals with genetically encoded membrane-tethered ligands

Abstract

Background: Neuropeptides regulate many biological processes. Elucidation of neuropeptide function requires identifying the cells that respond to neuropeptide signals and determining the molecular, cellular, physiological, and behavioral consequences of activation of their cognate G protein-coupled receptors (GPCRs) in those cells. As a novel tool for addressing such issues, we have developed genetically encoded neuropeptides covalently tethered to a glycosylphosphatidylinositol (GPI) glycolipid anchor on the plasma membrane ("t-peptides").

Results: t-peptides cell-autonomously induce activation of their cognate GPCRs in cells that express both the t-peptide and its receptor. In the neural circuit controlling circadian rest-activity rhythms in Drosophila melanogaster, rhythmic secretion of the neuropeptide pigment-dispersing factor (PDF) and activation of its GPCR (PDFR) are important for intercellular communication of phase information and coordination of clock neuron oscillation. Broad expression of t-PDF in the circadian control circuit overcomes arrhythmicity induced by pdf(01) null mutation, most likely as a result of activation of PDFR in PDFR-expressing clock neurons that do not themselves secrete PDF. More restricted expression of t-PDF suggests that activation of PDFR accelerates cellular timekeeping in some clock neurons while decelerating others.

Conclusions: The activation of PDFR in pdf(01) null mutant flies--which lack PDF-mediated intercellular transfer of phase information--induces strong rhythmicity in constant darkness, thus establishing a distinct role for PDF signaling in the circadian control circuit independent of the intercellular communication of temporal phase information. The t-peptide technology should provide a useful tool for cellular dissection of bioactive peptide signaling in a variety of organisms and physiological contexts.

Figure 1. Structure of GPI-tethered PDF isoforms

(A) The medium-linker isoform of GPI-tethered PDF (t-PDF-ML) contains trypsin signal sequence (blue), mature cleaved PDF peptide sequence (green), hydrophilic linker comprising the c-Myc epitope tag flanked by single glycine-asparagine (GN) repeats (yellow), and the GPI targeting signal from lynx1 protoxin (orange). After processing in the secretory pathway, the secretory signal and GPI targeting sequences are cleaved, and the C terminus is covalently linked to GPI whose aliphatic lipid chains are intercalated in the extracellular leaflet of the plasma membrane. (B) Schematics depicting PDFR and the isoforms of t-PDF (not to scale), which are identical to t-PDF-ML except as follows: The linker of t-PDF-LL contains c-Myc epitope flanked by four N terminal GN repeats and eleven C terminal GN repeats, t-PDF-SEC contains no GPI targeting sequence, and the PDF sequence of t-PDF-SCR has been replaced by a scrambled sequence comprising the same amino acids as PDF.

Figure 2. t-PDF activates cloned PDF receptor when co-expressed in vitro in mammalian tissue culture cells

Varying quantitites of cDNA encoding t-peptides are co-transfected into HEK 293 mammalian tissue culture cells with constant quantities of GPCR and cAMP-sensitive CRE-luciferase reporter cDNAs. Forty-eight hours after transfection, cells are either lysed for luciferase bioluminescence assay or kept intact and unpermeabilized for cell-surface anti-Myc ELISA assay. (A) t-PDF-ML and t-PDF-LL each dose-dependently increase steady-state intracellular cAMP, indicating activation of PDFR, with t-PDF-ML inducing greater increases than t-PDF-LL. t-PDF-SEC and t-PDF-SCR have no activity. (B) t-PDF-LL is expressed on the cell-surface at higher levels than t-PDF-ML and t-PDF-SCR, while t-PDF-SEC is undetectable. (C) t-PDF-ML only activates PDFR, and not the related receptors for the peptides DH31 or DH44. (D) t-DH31-ML, identical to t-PDF-ML except the PDF peptide sequence has been replaced with that for DH 31 (TVDFGLARGYSGTQEAKHRMGLAAANFAGGP), only activates DH31R, and not PDFR or DH44R. (Error bars are S.D; N = 3 repeats for all measurements).

Figure 3. t-PDF induces complex locomotor rhythms when expressed in vivo in all circadian clock neurons of transgenic flies

Male flies bearing UAS-t-peptide transgenes are mated to female flies bearing a tim(UAS)-GAL4 transgene to produce progeny expressing t-peptide in all circadian clock neurons. Free-running locomotor rhythms of individual male progeny entrained in 12h:12h LD conditions and then released into DD are categorized as rhythmic, complex rhythmic, or arrhythmic, and free-running periods are assigned, using Lomb-Scargle periodogram analysis. (A) t-PDF-ML is more active in vivo than t-PDF-LL, while t-PDF-SEC and t-PDF-SCR have no activity, thus recapitulating the relative in vitro activities shown in Figure 2. t-DH31-ML, while active against DH31R in vitro, does not influence free-running locomotor rhythms when expressed in vivo in circadian clock neurons. Bar graph depicts proportions of rhythmic (blue), complex rhythmic (yellow), and arrhythmic (red) flies of the indicated genotypes, with the notations in parentheses referring to specific chromosomal insertions, or combinations of two chromosomal insertions, of the UAS-t-peptide transgenes. n indicates the number of individual flies assayed, τ₀ indicates the average single free-running period of rhythmic flies, τ₁ the average shorter free-running period of complex rhythmic flies, τ₂ the average longer free-running period of complex rhythmic flies, and χ² the significance of χ² statistical comparison of the proportions for that genotype with that of tim > t-μO-MrVIA flies expressing a tethered conotoxin that has no activity in flies (***, p < 0.001). Average Lomb-Scargle periodogram powers are in parentheses following each free-running period component. (B) Representative free-running locomotor actograms of individual flies with the indicated phenotypes and genotypes. The gray bar indicates subjective day, and the black bar subjective night.

Figure 4. t-PDF expression in all clock neurons suppresses free-running arrhythmicity induced by pdf⁰¹ null mutation

Male pdf⁰¹ null-mutant flies bearing UAS-t-peptide transgenes are mated to female pdf⁰¹ null-mutant flies bearing tim(UAS)-GAL4 transgene to produce pdf⁰¹ null-mutant flies expressing t-peptide in all clock neurons. (A) Negative control pdf⁰¹ null-mutant flies expressing the inert t-μO-MrVIA conotoxin exhibit approximately 50% arrhythmicity, with the rhythmic flies exhibiting very weak rhythms, consistent with numerous published reports (see text). In contrast, very few pdf⁰¹ flies constitutively expressing t-PDF-ML in all clock neurons are arrhythmic, and instead predominately exhibit complex rhythms (***, p < 0.001, χ² test comparing each t-PDF-ML-expressing genotype to the t-μO-MrVIA-expressing control). (B) Averaged free-running DD actograms of all flies tested of the indicated genotypes. The induction of strong rhythmicity in pdf⁰¹ null-mutant flies by t-PDF-ML expression in all clock neurons is particularly apparent over the first week in DD, before individual flies have had the opportunity to drift out of phase with one another, thereby dispersing the population activity pattern depicted in the averaged actogram.

Figure 5. t-PDF expression in all clock neurons except for the PDF-secreting LN_V subset induces complex locomotor rhythms

Male flies bearing UAS-t-peptide transgenes are mated to female flies bearing both tim(UAS)-GAL4 and pdf-GAL80 (which suppresses GAL4 activation of UAS transgene expression in the PDF-secreting LN_Vs) transgenes to produce progeny expressing t-peptide in all circadian clock neurons except the PDF-secreting LN_Vs. (A) Proportions of locomotor phenotypes are different between each t-PDF-expressing genotype and the t-μO-MrVIA-expressing control (***, p < 0.001, χ² test). (B) Averaged actograms of flies of the indicated genotypes.

Figure 6. t-PDF expression in distinct partially overlapping subsets of clock neurons induces either single short-period rhythmicity or complex rhythmicity

Male flies bearing UAS-t-peptide transgenes are mated to cry24-GAL4 or cry16-GAL4 transgenes, which drive expression in distinct partially overlapping subsets of clock neurons. (A) t-PDF-ML expressed using cry24-GAL4 driver induces complex free-running locomotor rhythms similar to those induced using tim(UAS)-GAL4 driver (Figure 3). In contrast, t-PDF-ML expression using cry16-GAL4 driver induces only a modest degree of complex rhythmicity, and almost none when expressed at a higher dose simultaneously from two independent UAS-t-PDF-ML chromosomal insertions; rather, driving t-PDF-ML expression with cry16-GAL4 induces dramatic shortening of free-running period from ~25.5 hours in negative control flies expressing either t-PDF-SCR or t-μO-MrVIA (long-period phenotypes due to the cry16-GAL4 and cry24-GAL4 transgenes themselves have been previously reported; see text) to ~22 hours in flies expressing t-PDF-ML (overall ANOVA p < 0.001; p < 0.05 for paired comparisons to appropriate pooled controls using the Bonferroni Versus Control Test). ***, p < 0.001, **, p < 0.005, χ² test comparing each t-PDF-ML-expressing genotype to the appropriate pooled controls. (B) Averaged actograms of flies of the indicated genotypes demonstrate clearly the induction of complex rhythmicity by t-PDF-ML expressed using cry24-GAL4 and short-period rhythmicity using cry16-GAL4.