Recognition between symbiotic Vibrio fischeri and the haemocytes of Euprymna scolopes

Abstract

The light organ crypts of the squid Euprymna scolopes permit colonization exclusively by the luminous bacterium Vibrio fischeri. Because the crypt interior remains in contact with seawater, the squid must not only foster the specific symbiosis, but also continue to exclude other bacteria. Investigation of the role of the innate immune system in these processes revealed that macrophage-like haemocytes isolated from E. scolopes recognized and phagocytosed V. fischeri less than other closely related bacterial species common to the host's environment. Interestingly, phagocytes isolated from hosts that had been cured of their symbionts bound five times more V. fischeri cells than those from uncured hosts. No such change in the ability to bind other species of bacteria was observed, suggesting that the host adapts specifically to V. fischeri. Deletion of the gene encoding OmpU, the major outer membrane protein of V. fischeri, increased binding by haemocytes from uncured animals to the level observed for haemocytes from cured animals. Co-incubation with wild-type V. fischeri reduced this binding, suggesting that they produce a factor that complements the mutant's defect. Analyses of the phagocytosis of bound cells by fluorescence-activated cell sorting indicated that once binding to haemocytes had occurred, V. fischeri cells are phagocytosed as effectively as other bacteria. Thus, discrimination by this component of the squid immune system occurs at the level of haemocyte binding, and this response: (i) is modified by previous exposure to the symbiont and (ii) relies on outer membrane and/or secreted components of the symbionts. These data suggest that regulation of host haemocyte binding by the symbiont may be one of many factors that contribute to specificity in this association.

Fig. 1

Vascularization of the E. scolopes light organ. A. Ventral dissection of an adult E. scolopes, revealing the bilobed light organ (circled in yellow). Eye (e) and tentacles (t) are also indicated. Scale bar = 1 cm. B. Higher magnification view of the light organ, dissected further to reveal the symbiont-containing central core (cc) tissue. Scale bar (for both B and C) = 2 mm. C. View of the same light organ, in which the circulatory system has been visualized by injecting the fluorochrome CellTracker Orange (Molecular Probes) through the cephalic artery (ca), located at the mantle/head interface. Fluorescent excitation reveals a high degree of vascularization of the central core (cc) tissue. Other highly vascularized organs in the vicinity include the brain (b) and the brachial hearts (bh).

Fig. 2

Differential binding of bacteria by squid hemocytes. Above: representative confocal microscopy images of isolated squid hemocytes stained with CellTracker Orange. Differential interference contrast (DIC) views of the cells are superimposed to reveal the extended pseudopodia. The extent to which different species of bacteria (green) bound to the hemocytes’ external surfaces was easily visualized and enumerated. Scale bar = 10 μm. Below: mean number of bacteria bound to each hemocyte was determined for at least ten hemocytes per microscopic field (n = 5) in three replicate experiments. Error bars indicate standard errors of the mean (SEM); asterisks denote levels of binding that were significantly different (p < 0.05) from that of V. fischeri ES114, as determined by ANOVA pair-wise analysis.

Fig. 3

Effect of curing squid of their symbiont population on hemocyte binding to different bacteria. Mean number of bacteria bound to hemocytes isolated from animals at different times after initiating antibiotic treatment to remove symbiotic V. fischeri. Adult squids were maintained in sterile seawater either in the normal symbiotic state or under antibiotic conditions that cured the light organ. Hemocytes were removed from either the symbiotic (normal) or cured (naïve) animals over 5 days for adherence assays (see Experimental procedures). The mean numbers of bacteria bound by symbiotic (white bars) or naïve (black bars) hemocytes were determined in three independent experiments. Error bars indicate SEM; asterisks denote levels of binding that were significantly different (p < 0.05) from those measured on day 1, as determined by ANOVA pair-wise analysis.

Fig. 4

Effect of an ompU mutation on V. fischeri evasion of hemocyte recognition and binding. The mean numbers of wild-type (ES114) or ompU (OM3) V. fischeri cells that were bound by isolated hemocytes was determined in three independent experiments. Restoration of an intact copy of ompU on pFA9 eliminated the elevated level of hemocyte binding of strain OM3. Addition of the V. fischeri cloning vector pV08 had no significant effect on OM3 binding. Error bars indicate SEM; asterisks denote levels of binding that were significantly different (p < 0.05) from those of V. fischeri ES114, as determined by ANOVA pair-wise analysis.

Fig. 5

Complementation of the V. fischeri ompU defect by viable, wild-type bacteria. A. Hemocyte binding of wild-type V. fischeri ES114 incubated either alone or in combination with cells of V. harveyi, V. parahaemolyticus, P. leiognathi, or V. fischeri OM3. B. Hemocyte binding of V. parahaemolyticus, P. leiognathi and V. harveyi in either the absence (white bars) or presence (black bars) of wild-type V. fischeri ES114. Error bars indicate SEM; asterisks denote levels of binding that were significantly different (p < 0.05) in the presence of V. fischeri ES114, as determined by ANOVA pair-wise analysis. C. Hemocyte binding of V. fischeri OM3 (ompU) either alone (white bar) or in the presence of wild-type V. fischeri ES114 cells (black bars). Values are the mean number of bacteria adhering to ten hemocytes in three independent experiments. Error bars indicate SEM; asterisk denotes levels of binding that were significantly different (p < 0.05) from those of V. fischeri ES114, as determined by ANOVA pair-wise analysis.

Fig. 6

Rates of phagocytosis of bacteria bound by hemocytes. A. Representative flow cytometry plot recorded 30 min after the start of the phagocytosis assay. The fluorescence of hemocytes exposed to TAMRA-labeled cells of V. fischeri ES114 was assessed by flow cytometry before (left panel) and after (right panel) the addition of trypan blue, which quenches the fluorescence of cells not phagocytosed. B. The fluorescence of hemocytes exposed to TAMRA-labeled cells of V. parahaemolyticus was assessed as described above. C. The extent of reduction in the mean fluorescence after the addition of the trypan blue (A and B) corresponded to the degree of internalization of the bound bacteria. This value was determined every 30 min after exposure of the bacteria to hemocytes, and was a measure of the relative rates of phagocytosis of V. fischeri (closed circle) and V. parahaemolyticus (closed triangle) cells.