Transcriptional patterns in both host and bacterium underlie a daily rhythm of anatomical and metabolic change in a beneficial symbiosis

Abstract

Mechanisms for controlling symbiont populations are critical for maintaining the associations that exist between a host and its microbial partners. We describe here the transcriptional, metabolic, and ultrastructural characteristics of a diel rhythm that occurs in the symbiosis between the squid Euprymna scolopes and the luminous bacterium Vibrio fischeri. The rhythm is driven by the host's expulsion from its light-emitting organ of most of the symbiont population each day at dawn. The transcriptomes of both the host epithelium that supports the symbionts and the symbiont population itself were characterized and compared at four times over this daily cycle. The greatest fluctuation in gene expression of both partners occurred as the day began. Most notable was an up-regulation in the host of >50 cytoskeleton-related genes just before dawn and their subsequent down-regulation within 6 h. Examination of the epithelium by TEM revealed a corresponding restructuring, characterized by effacement and blebbing of its apical surface. After the dawn expulsion, the epithelium reestablished its polarity, and the residual symbionts began growing, repopulating the light organ. Analysis of the symbiont transcriptome suggested that the bacteria respond to the effacement by up-regulating genes associated with anaerobic respiration of glycerol; supporting this finding, lipid analysis of the symbionts' membranes indicated a direct incorporation of host-derived fatty acids. After 12 h, the metabolic signature of the symbiont population shifted to one characteristic of chitin fermentation, which continued until the following dawn. Thus, the persistent maintenance of the squid-vibrio symbiosis is tied to a dynamic diel rhythm that involves both partners.

Fig. 1.

The diel cycle of the squid–vibrio symbiosis. (A) The mature host light organ. A ventral view of the host animal (Left) reveals the position of the bilobed light organ in the mantle cavity. After the dawn light cue, the animal expels the crypt contents into the environment (arrowhead; ventrally dissected animal, Center). The central epithelial core (arrow; Right), which harbors the bacterial symbionts, was removed from each lobe of the organ for array analyses (approximate position indicated by dashed circle). (B) The night-active period of the host (orange) corresponds to the largest symbiont population (blue lines), most of which is expelled at dawn. (C) Differential regulation of host (orange) and symbiont (blue) genes over the day as a percentage of the total host ESTs (nonredundant cDNA library of 13,962 sequences) or symbiont genes (3,802) probed. Time intervals #1–4, regulated at: #1,1’ = 2200 hours relative to 1600 hours; #2 = 0400 hours relative to 2200 hours; #3 = 1000 hours relative to 0400 hours; #4 = 1600 hours relative to 1000 hours.

Fig. 2.

The diel pattern of host cytoskeletal dynamics. (A) Differential regulation (based on per-spot per-chip (PSPC) values of P < 0.05) of cytoskeletal (i.e., microfilament- and microtubule-associated) genes over the time intervals examined: #1 = 2200 hours relative to 1600 hours; #2 = 0400 hours relative to 2200 hours; #3 = 1000 hours relative to 0400 hours; #4 = 1600 hours relative to 1000 hours. The number of regulated microarray features (genes) with a given annotation are in parentheses. *, genes regulated when the stringency was relaxed to PSPC values of P < 0.1; †, 11 genes were not significantly regulated, and one was significantly up-regulated; ‡, 11 genes were significantly down-regulated, and one was not significantly regulated. (B and C) TEMs of the apical surfaces of host crypt epithelia (upper right of each image) and the bacteria-containing crypt space (lower left of each image). (B) From the late morning through the evening time periods, a dense bacterial population associates closely with the lobate microvilli (mv) of the highly polarized host epithelia. (C) In the hours surrounding dawn, the host cell membranes are effaced of microvilli, and portions of the apical surfaces of these cells bleb into the bacteria-containing crypt spaces. This specific micrograph depicts host tissues right after the expulsion process, when population densities of the symbiont have not yet recovered.

Fig. 3.

Evidence of a diel pattern of symbiont metabolism. (A) Sequential expression of two groups of catabolic genes. Genes involved in either chitin utilization (light green) or glycerol utilization (magenta) are coordinately transcribed at peak levels during either the late night (interval #2; 2200–0400 hours) or morning (interval #3; 0400–1000 hours), respectively. Chitin genes include: circles, VF_0655; squares, VF_1598; triangles, VF_A0013; diamonds, VF_A0143; and hexagons, VF_A0715. Glycerol genes include: circles, VF_0072; squares, VF_A0235; triangles, VF_A0236; diamonds, VF_A0248; hexagons, VF_A0249; trapezoids, VF_A0250; and pentagons, VF_A0958. The relatively small, although statistically significant, fold-changes seen here have been noted in other studies of natural populations of genetically dissimilar strains (25). (B) Coordinate expression of distinct catabolic pathways. Proteins required for generation of ATP by the fermentation of chitin (left panel) or anaerobic respiration of glycerol or G3P (right panel) are indicated. Proteins that are apparently induced during Interval #2 or #3 [e.g., those in (A)] are indicated in light green or magenta, respectively. DHAP, dihydroxyacetone phosphate; GAP, glyceraldehyde phosphate; GlcNAc, N-acetyl glucosamine. (C) Metabolic capabilities of V. fischeri in culture. Growth yields per mole of substrate carbon [GlcNAc (light green bars) or glycerol (magenta bars)] in minimal-salts medium (52), and final pHs (indicated on the bars) were determined when cells were grown either fermentatively (F) with no electron acceptor or by anaerobic nitrate respiration (AR). (D) The relative energy-generation advantage of growth on GlcNAc relative to glycerol (light green bar) or glycerol relative to GlcNAc (magenta bar) were calculated by using a metabolic model based on the substrate-specific efficiency of ATP generation by fermentation or anaerobic respiration (18).

Fig. 4.

Fatty-acid composition of bacterial symbionts. The chain length, saturation, and relative amount of each of the fatty acids present in the lipids of symbiotic bacteria purified directly from the light organ (gray bars) were compared with those of the same bacteria subsequently grown in medium (black bars) or with those of the light-organ epithelial tissue (open bars).