Evolutionary origin of cAMP-based chemoattraction in the social amoebae

Abstract

Phenotypic novelties can arise if integrated developmental pathways are expressed at new developmental stages and then recruited to serve new functions. We analyze the origin of a novel developmental trait of Dictyostelid amoebae: the evolution of cAMP as a developmental chemoattractant. We show that cAMP's role of attracting starving amoebae arose through recruitment of a pathway that originally evolved to coordinate fruiting body morphogenesis. Orthologues of the high-affinity cAMP receptor (cAR), cAR1, were identified in a selection of species that span the Dictyostelid phylogeny. The cAR1 orthologue from the basal species Dictyostelium minutum restored aggregation and development when expressed in an aggregation-defective mutant of the derived species Dictyostelium discoideum that lacks high-affinity cARs, thus demonstrating that the D. minutum cAR is a fully functional cAR. cAR1 orthologues from basal species are expressed during fruiting body formation, and only this process, and not aggregation, was disrupted by abrogation of cAR1 function. This is in contrast to derived species, where cAR1 is also expressed during aggregation and critically regulates this process. Our data show that coordination of fruiting body formation is the ancestral function of extracellular cAMP signaling, whereas its derived role in aggregation evolved by recruitment of a preexisting pathway to an earlier stage of development. This most likely occurred by addition of distal cis-regulatory regions to existing cAMP signaling genes.

Fig. 1.

Identifation of cAR-like sequences in four Dictyostelid species. (A) Alignment of cAR-like sequences from four test species with the D. discoideum cARs. DNA fragments of 543-627 bp were amplified from D. fasciculatum, D. minutum, P. pallidum, and D. rosarium genomic DNA by using degenerate oligonucleotides that match conserved sequences in the four D. discoideum cARs. After excision of a variable length intron at a conserved position (arrow), the derived amino acid sequences were determined and aligned by using clustal-x. Amino acid residues that are identical in the majority or at least four of the nine sequences are shaded gray. The conserved regions used for oligonucleotide design are shown for DdcAR1-4, for PpcAR, which is identical to TasA (24), and for DmcAR. The positions of the putative transmembrane (TM) domains 3-7 of DdcAR1 (35) are indicated. GenBank accession nos: A41238 (DdcAR1), A46390 (DdcAR2), A46391 (DdcAR3), A54813 (DdcAR4), and AB045712 (TasA). (B) Phylogenetic analysis of cAR-like sequences. The tree shown was derived by maximum likelihood analysis and Bayesian inference and is drawn to scale, as indicated by the scale bar (0.1 substitutions per site). Thick lines indicate nodes with 1.00 Bayesian inference posterior probabilities and 100% mlBP support. An alternative branching pattern among the two deepest cAR nodes favored by mlBP is indicated by a double-headed arrow. Four putative G protein-coupled receptor sequences were used to root the tree. N. crassa. Neurospora crassa. GenBank accession nos.: AAM20722 (AtGPCR), AAO62367 (DdcrlA), EAA35706 (NcGPCRα), and EAA28751 (NcGPCRβ). (C) Molecular phylogeny of Dictyostelids based on small subunit rRNA sequences. The tree shown was derived by using Bayesian inference and maximum likelihood analysis on 1,556 unambiguously aligned nucleotide positions. Sequences from solitary amoebae were used to root the tree.

Fig. 2.

Developmental regulation of cAR gene expression. Cells of the indicated five species were incubated on nonnutrient agar until fruiting bodies had formed. Total RNA was extracted at 2-h intervals, and the progression of development was photographed. Northern blots were probed at 65°C with [³²P]dATP-labeled DdcAR1 cDNA or with the [³²P]dATP-labeled DrcARI, DmcAR, PpcAR, and DfcAR PCR products, respectively, and washed at high stringency. (Bar, 200 μM.)

Fig. 3.

Cloning and cAMP-binding properties of DmcAR. (A) Cloning of DmcAR. Screening of a D. minutum genomic DNA library with the DmcAR PCR product yielded a 4.87-kb contig of three clones. This contig contains DmcAR and two flanking genes, which we denote DmSpkA and DmDtmA. These genes are most similar to the D. discoideum genes SpkA and DDB0217155, respectively, which occupy the same positions relative to DdcAR1 on chromosome 2 (36). The percentages of amino acid identity between the orthologous genes are indicated. (B) cAMP binding. car1car3 cells, transformed with either A15:DmcAR, A15:DdcAR1, or no construct, were incubated with 10 nM [³H]cAMP and assayed for cell-surface-associated [³H]cAMP-binding activity. (C) Competition curve for cAMP. A15:DmcAR- or A15:DdcAR1-transformed car1car3 cells were incubated with 1 nM [³H]cAMP and the indicated concentrations of cAMP and assayed for [³H]cAMP binding to the cell surface. The data are presented as percentage of ³HcAMP binding in the absence of cAMP and as a Scatchard plot (37) (Inset). B, bound; F, free cAMP; N, number of molecules. (D) Inhibition of [³H]cAMP binding by adenosine and 2′3′isopropylidene adenosine (IPA). The transformed cell lines were incubated with 10 nM [³H]cAMP and the indicated concentrations of adenosine and IPA and assayed for [³H]cAMP binding to the cell surface. The data are presented as percentage of [³H]cAMP binding in the absence of nucleosides. All data represent the means and SEM of two experiments performed in triplicate.

Fig. 4.

Complementation of Ddcar1car3 by DmcAR. (A) Restoration of development. The D. discoideum car1car3 mutant, its parent DH1, and car1car3 transformed with A15:DmcAR were incubated on nonnutrient agar at 22°C and photographed at 2-h intervals. (Bar, 100 μm.) (B) Oscillatory signaling. car1car3/A15:DmcAR cells were incubated for 5 h at 4 × 10⁵ cells per cm² on agar and subsequently tracked during 50 min at 10-s intervals by time-lapse videomicroscopy under phase-contrast illumination. Optical density waves were enhanced by image subtraction (30). The 256th video frame is shown. (Bar, 100 μm.) See Movie 1, which is published as supporting information on the PNAS web site.

Fig. 5.

Effects of SpcAMPS on Dictyostelid development. Cells from the indicated species were distributed at 2 × 10⁵ cells per cm² on nonnutrient agar (control) or agar with 10 or 300 μM SpcAMPS and incubated at 22°C. The progression of development was photographed at 2-h intervals. (Bar, 200 μM.)