This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Saranak, J.
Right arrow Articles by Foster, K. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Saranak, J.
Right arrow Articles by Foster, K. W.

Next Article 

Eukaryotic Cell, October 2005, p. 1605-1612, Vol. 4, No. 10
1535-9778/05/$08.00+0     doi:10.1128/EC.4.10.1605-1612.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Photoreceptor for Curling Behavior in Peranema trichophorum and Evolution of Eukaryotic Rhodopsins

Jureepan Saranak and Kenneth W. Foster*

Physics Department, Syracuse University, 201 Physics Building, Syracuse, New York 13244-1130

Received 11 November 2004/ Accepted 18 July 2005


arrow
ABSTRACT
 
When it is gliding, the unicellular euglenoid Peranema trichophorum uses activation of the photoreceptor rhodopsin to control the probability of its curling behavior. From the curled state, the cell takes off in a new direction. In a similar manner, archaea such as Halobacterium use light activation of bacterio- and sensory rhodopsins to control the probability of reversal of the rotation direction of flagella. Each reversal causes the cell to change its direction. In neither case does the cell track light, as known for the rhodopsin-dependent eukaryotic phototaxis of fungi, green algae, cryptomonads, dinoflagellates, and animal larvae. Rhodopsin was identified in Peranema by its native action spectrum (peak at 2.43 eV or 510 nm) and by the shifted spectrum (peak at 3.73 eV or 332 nm) upon replacement of the native chromophore with the retinal analog n-hexenal. The in vivo physiological activity of n-hexenal incorporated to become a chromophore also demonstrates that charge redistribution of a short asymmetric chromophore is sufficient for receptor activation and that the following isomerization step is probably not required when the rest of the native chromophore is missing. This property seems universal among the Euglenozoa, Plant, and Fungus kingdom rhodopsins. The rhodopsins of animals have yet to be studied in this respect. The photoresponse appears to be mediated by Ca2+ influx.


arrow
INTRODUCTION
 
Since rhodopsins sharing sequence similarities are found in bacteria, archaea, eukaryotic microorganisms, and animals, it is possible to gain insight into the evolutionary progression and diversification leading to the receptor proteins of human visual photoreceptors as well as other photoreceptors. As part of this continuing project on the evolution and mechanisms of visual systems, the light-dependent responses of Peranema have been studied with respect to behavior, action spectra, the mechanism of activation, and signaling to the cell.

Peranema trichophorum is a colorless eukaryotic phagotroph of the Euglenophyta that lives in fresh water (8). It does not have a chloroplast but rapidly deforms in shape and is very effective at capturing and eating prey (41). Peranema voraciously takes any particles into its body by phagocytosis (1). This "feeding behavior" is clearly and commonly observed under the microscope. Its life cycle is not well studied, but from light microscopic observations Peranema appears to have several distinct stages of growth. After the cells are put into fresh medium, which is initially low on waste products, one sees small round cells of about 10 to 15 µm in diameter. These cells elongate into oval cells that swim freely in solution and fast in a helical euglenoid motion. Over 2 weeks, these cells elongate further to 25 to 70 µm in length, remaining 10 µm wide, and are seen to glide forward on a surface pulled by their leading anterior cilium, which is about the length of the cell body. (We use the word "cilium" rather than "eukaryotic flagellum" to describe the leading appendage that enables the cell to glide to emphasize that this appendage is not the more familiar rotary motor-driven flagellum of bacteria. The cilium controls linear motors that slide microtubules relative to each other in the axenome and is the same as the cilia of Ciliata. This choice of terminology follows the advice of Irene Manton [24] and Tom Cavalier-Smith [5], who says "the word flagellum should be dropped altogether from eukaryotic biology.") These fully grown or mature cells are pulled forward along a surface by the cell surface motility of the proximal five-sixths of the anterior cilium (32). The cilium is straight except for the waving anterior one-sixth, which presumably senses particles of food and obstacles. A second thinner cilium runs posteriorly and tightly appressed to the cell body, lying between the cell body and gliding surface, and is rarely visible under the light microscope (41). Apparently, the thin cilium does not participate in driving the cell forward. The cells are in sufficient contact with the surface to not rotate as they glide. In order to turn, the cell body curves and then straightens with its cilium pointing toward its new direction. The cells also curl their bodies into a ball, with fast beating and waving of their cilia, when they eat. Under our culture conditions, Peranema remains in this mature stage for about 3 to 4 weeks. The cells gradually become wider and denser at the last stage of their life cycle. They glide slowly, spend more time curled, and finally stay in the curled-up shape and die from starvation or senescence.

Mature Peranema cells glide, turn, and curl spontaneously in either light or dark. A movie of the curling can be seen at the Euglenoid Project website (http://www.plantbiology.msu.edu/triemer/Euglena/movies.htg/peranem2.mov). As originally noted by Shettles (35), the frequency of curling behavior is apparently enhanced by an increase in light, as shown by a shortening of the time before the next curl. Shettles mapped the location of photosensitivity to the primary cilium and the front half of the body. Both light and dark adaptations increase the sensitivity to light (35). At visible wavelengths, the sensitivity is well related to the light intensity and apparently is highest at 500 to 505 nm (35), which is typical for the spectral peak of a rhodopsin. While Shettles did not make any particular hypothesis about the nature of the receptor, the photoreceptor rhodopsin has been found in several evolutionary branches that diverged early from animals, including green algae (16) and fungi (34). Replotting of Shettles's data (35, 36) yielded an action spectrum which fit reasonably well with the rhodopsin standard curve (12) if one considers that the technology used did not provide sufficiently narrow-wavelength bands of light. These overly broad bands of light resulted in a flattened action spectrum. We are using the traditional (premolecular biology) operational definition of rhodopsin, namely, a protein that forms a chromophore with a retinal molecule and that has a light-sensing function. So far, all rhodopsins identified with this definition have subsequently proved to have similar (and probably homologous) amino acid sequences.

Since the available evidence suggests that rhodopsin might be the photoreceptor responsible for enhancing curling in Peranema, we used our established techniques for the identification of photoreceptors (12, 16, 34) to show here that rhodopsin is indeed the photoreceptor for the curling behavior of Peranema. We also confirm unusual features of the light/dark adaptation and link rhodopsin activation to calcium ions and channels during curling behavior.

The availability of rhodopsins in the exterior plasma membranes of microorganisms makes it relatively easy to incorporate and test for the physiological in vivo activity of rhodopsin analogs that can help to elucidate the mechanisms of receptor activation. Our report in this paper of a rhodopsin from the Euglenozoa kingdom (4) allows us to ascertain the universality of the activation properties of eukaryotic rhodopsins found previously in the Fungus and Plant kingdoms.

This finding also permits some speculation about the origin of rhodopsins in eukaryotes and their relationship to the rhodopsins of bacteria and archaea. Note that an analysis of the structures of bacteriorhodopsin and rhodopsin showed that they have the same overall topology of their polypeptide folds, with helices I to III and also helices VI and VII superimposing reasonably well (40). The amino acid sequences of helix VI are remarkably similar for all rhodopsins. We also note that a new consensus is emerging that bikonts such as Peranema, the green algae, ciliates, foramins, and cryptomonads diverged from unikonts such as the fungi and animals early in eukaryotic divergence (6, 7, 39). The finding of rhodopsin in Peranema supports the notion of a common eukaryotic ancestor having a rhodopsin photoreceptor.


arrow
MATERIALS AND METHODS
 
Culture. P. trichophorum was bought from Carolina Biological Supply Company, Burlington, NC. The cells came in a mixed culture in spring water with decaying material as food. They were kept in a constant room temperature of 22°C under fluorescent room light during the day and in the dark at night. The culture was used for experiments when there was a sufficient number of elongated cells to observe their curling behavior as they glided. These cells were good for experiments for about 3 to 4 weeks. Every 2 weeks, cells were transferred to about 30 ml of spring water (Spring Brook Springs II) in a culture jar with a grain of wheat and a small piece (1 to 2 mm) of hard-boiled egg yolk to produce the next batch of active cells. The gliding cells from the original or newly transferred jar were adapted to the dark for at least 30 min before the experiments. After a brief mixing and aeration step, an aliquot of 0.3 ml from the top half of the suspension was transferred to a glass microscope slide with a circular depression in the middle. The uncovered slide was placed on the microscope's X-Y translation stage. Moving the stage allowed us to track a cell during behavioral observation.

For characterization of dark/light adaptation, aliquots of the cell suspension were put through different protocols of adaptation as described below. For action spectrum experiments, the culture jar was kept in the dark 30 min before and during the experiments. Due to phagocytosis, Peranema takes up chemicals very rapidly. Therefore, for pharmacological study, the chemicals were incorporated with cell samples for only 5 min in the dark.

Experimental setup. The experimental setup is depicted in Fig. 1. All experiments were done in a constant-temperature (22°C) dark room. A phase-contrast microscope (Nikon Labophot-2) was equipped with a near-infrared (near-IR) charge-coupled-device video camera projecting the image of the cells onto a TV monitor. The cells were illuminated from below with dim near-IR light. A 10x dry objective was used both for magnification of the image of the cells and to concentrate the focused beam of stimulating light from above onto the cells. Except for the light/dark adaptation experiments, a 300-W tungsten lamp (EXR) in a Kodak carousel slide projector was used as a stimulating light source. Three-cavity interference filters with transmission peaks of 300, 334, 350, 380, 420, 460, 500, 546, 600, and 640 nm (10-nm bandwidth) from Microcoating (Westford, MA) were used to generate different wavelengths of light for stimulation. Optical density filters were used to vary the light intensity. The light beam was collected by three convex lenses onto a 45° dichroic mirror, which reflected wavelengths of <675 nm down to the objective lens to directly stimulate the cells in the field of observation while transmitting those above 675 nm from the near-IR observing light up through to the video camera.



View larger version (32K):
[in this window]
[in a new window]
  		FIG. 1. Schematic diagram of the experimental apparatus.

For light/dark adaptation studies, the stimulating light was the microscope illumination light without the near-IR filter. Light adaptation was done by exposing the cell suspension to a tungsten lamp at 5.0 mW/cm2 for 30, 60, or 120 min and then keeping it in the dark for 30 min before testing for the response time. One sample with 30-min light adaptation was tested without being kept in the dark. Dark-adapted samples were kept in the dark for the same 30 min and then an additional 30 min before being tested.

Test for response. The response time, defined as the time in seconds for a cell to curl (from a straight shape to a ball shape), was recorded manually with a stopwatch. The end point was defined as a tight curl if the cell body became a balled shape or the cell completed a 360° tight turn. A mature cell was identified and tracked to obtain the response time in the dark two times (geometric mean = D), followed by stimulation in the light (L) one time. The geometric mean is i = 1nai1n for the data sequence ai)i = 1n. Cells were only stimulated once because we found that prior exposure strongly influenced the response. Measuring the spontaneous response twice in the dark improved the estimate of the dark response time without extending the experiment too long. The light response, R, was usually defined as follows (except in Table 1, where R = D/L):

Cells that did not show spontaneous curling (in the dark) within 5 min were rejected. For action spectrum experiments, at least seven cells were observed at each intensity. The field of view was changed after one cell was observed to minimize the chance of using cells that were prematurely exposed to light. A new aliquot of cell suspension was used at each intensity. After testing to find approximately the right range of intensities to use, at least three intensities near the threshold were tested for each wavelength to plot the log-intensity linear-response curve.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Effect of light and dark adaptation on curling response

Plot of action spectra. The threshold action spectra for native rhodopsin and n-hexenal-opsin were plotted as previously described (12, 34). In brief, the threshold for each stimulating wavelength was obtained by linear extrapolation of its log-intensity linear-response curve (see Fig. 1 in reference 16 and Fig. 3 and 4 in reference 12). The log of sensitivity (reciprocal of threshold) was plotted against wavelength (nm)/energy (eV) using SigmaPlot. The action spectrum was fitted to the standard rhodopsin curve (33).



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 3. Action spectra of curling behavior. (A) Action spectrum of native untreated cells. (B) Action spectrum of cells whose chromophores have been bleached and then replaced by 5 pM n-hexenal. The spectra are shown on the same relative sensitivity scale, and the scales are in a standard proportion, where a 1-eV shift is the same length as 2 log sensitivity units (12).



View larger version (10K):
[in this window]
[in a new window]
  		FIG. 4. (A) Dose-response effect of EGTA on curling behavior. The response, R, was measured in seven cells for each data point after stimulation at 500 nm and 26 µW/cm2. (B) Effect of light and Ca2+ ionophore on curling rate. •, dark; {blacktriangledown}, light at 500 nm and 22 µW/cm2; {blacksquare}, light at 500 nm and 342 µW/cm2.

Chromophore removal and analog reconstitution. A nonreversible removal of the anticipated retinal of Peranema was attempted by adding hydroxylamine (to trap free retinal) to the cell suspension before exposure to high-intensity light. However, hydroxylamine as low as 5 nM was still toxic to the cells. Therefore, only strong light was used to bleach the cells and remove the chromophore. The cell suspension (0.3 ml) in the microscope chamber was laterally exposed to 500 nm (32-nm-bandwidth interference filter; Barr Associates, Westford, MA) of intensity up to 105 mW/cm2 for 2 min to bleach the rhodopsin molecules, removing the native chromophore. The bleached suspension was then allowed to take up 5 pM n-hexenal (Aldrich), which incorporated into rhodopsin, and the cells were tested immediately for their response time as described above. Testing of the response of each sample was completed within 30 min, the time period in which the native pigment remained suppressed by photobleaching and the n-hexenal activity was still effective.

It is essential to note that Peranema is enormously more sensitive to added chemicals than any other microorganism we have tested. The uptake of chemicals into the cells is evidently fast and efficient, probably due to the frequent phago/endocytosis used for their feeding. Hydroxylamine at >5 nM, n-hexenal at >5 pM, and retinal at >1 µM are toxic to the cells. The cells became tightly curled and eventually died upon exposure to these chemicals. We had to omit the hydroxylamine for irreversible bleaching and the vitamin E (0.25%) as an antioxidant for retinal and hexenal that we routinely used in our other studies. This was unfortunate since added antioxidant approximately doubles the period of activity for hexenal. We also observed that the regeneration of rhodopsin after bleaching without hydroxylamine took place after 30 min. The effects of added retinal and hexenal without added vitamin E last for 45 and 30 min, respectively, in Chlamydomonas (J. Saranak, unpublished). In the absence of added antioxidant after 3 h of incubation, no chromophore is detected in the binding site in Chlamydomonas and hence no activity is detected (19). Furthermore, the activity of n-hexenal analogues must be tested before they can be significantly oxidized into the carboxylic acid form that does not bind opsin and before they have evaporated. These constraints gave us a mere 30-min window for n-hexenal experiments. Fortunately, since uptake into the Peranema cell was almost instantaneous, we did not have to wait before testing. Since we were able to complete each test within 30 min, the regeneration of rhodopsin and the oxidation of retinal or hexenal did not affect our results. The above details are important if one desires to repeat these experiments.

Calcium experiments. To study the effect of calcium on curling behavior, we used the calcium ionophore A23187 [GenBank] (40 mM in dimethyl sulfoxide); a calcium chelator, the free acid form of EGTA; and a calcium channel blocker, nifedipine (50 mM in ethyl alcohol). Chemicals were purchased from Sigma and dissolved in spring water (Spring Brook Springs II) unless otherwise indicated. They were added to the cell suspension 5 min before testing. Subsequently, measurements of free Ca2+ in cell suspensions and spring water before and after the addition of EGTA were done using a calcium electrode (ISE25Ca-9; Radiometer Analytical, France) and a reference electrode (REF 201; Radiometer Analytical, France).

To detect the free calcium in the cells, we used Calcium Green 1 (5 mM in dimethyl sulfoxide; Molecular Probes, Eugene, Oregon) at 1 µl/1 ml of Peranema suspension incorporated on a shaker in the dark for 15 min and then observed with an Olympus BX51W1 microscope with 60x and 100x water immersion objectives at excitation/emission wavelengths of 500 and 535 nm.

Data analysis of delays in spontaneous (dark) and elicited (dark and light adaptation) responses. Because the distribution of response times followed a log normal distribution, the following calculations were carried out. For each cell, the geometric mean of the delays in spontaneous responses in the dark (D) was calculated. The ratio (R) of the rate of curling (reciprocal of delay) in response to a standard light stimulus (1/L) to the spontaneous (in the dark) rate of curling (1/D) was then calculated (R = D/L). The mean and the standard error of the mean (SEM) for each condition, which consisted of 2 to 12 cells, were also calculated using the PSIPlot computer program, and the data are presented in Table 1. Since the frequency distribution of reaction times was fit by the log normal distribution, the level of significance was calculated from the standard normal distribution using the natural logarithm values of D and L, as follows:

where s2ln D is the variance of ln D and nD is the number of measurements in the sample.


arrow
RESULTS
 
Spontaneous and light-stimulated curling. The frequency distribution of the response times (times to onset of curling) for spontaneous curling under constant conditions fits the log normal distribution curve (Fig. 2A). This fact guided all our analyses. The fit of the log normal distribution to the response time histogram for light stimulation (Fig. 2B) shows an about threefold decrease in geometric mean response time compared to that for spontaneous curling. (The geometric mean is the appropriate metric for log normal distributions, just as the mean is appropriate for normal distributions.) The geometric mean response time for light-stimulated curling was 11 +4/–3 s (value ± SEM; nL = 21) (Fig. 2B) versus 36 +6/–5 s (nD = 42) (Fig. 2A) for spontaneous curling of this set of cells. (Note that the SEM of a log normal distribution is symmetric on a logarithmic scale but skewed on a linear scale.) For these data, the probability that there is not a response to light is <0.0001 (level of significance). While possibly not significant, given the very limited amount of data, the light-stimulated log normal distribution shows an extra two cells responding earlier than expected. If significant, this would suggest that there could transiently be a higher probability of a response immediately following a step change in light than later. Such a transient increase in response is common in biological signal transduction networks and occurs whenever signal processing involves a differentiating function which leads to a faster response.



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 2. Time distribution of curling. Repeated measurements were made of the length of time before cell curling (time to curl). The figures are histograms of the natural Naperian logarithms of the time, in seconds. (A) Dark conditions; (B) immediately following light stimulation at 500 nm at 26 µW/cm2.

Light and dark adaptation was studied briefly here primarily to gain the experience of the light-dependent behavior of Peranema and to choose optimal parameters for action spectrum experiments. Our results on light/dark adaptation were similar to those of Shettles (35) in that the cells had the fastest response after 30 to 60 min of light followed by 30 min of dark (Table 1, compare first four rows). Longer light exposures (experiment d) reduced the response. Having a dark adaptation period after the light period also increased the response (compare experiments e and b). Thus, we have confirmed the finding that unlike the case in animals, both light and dark adaptation increased the response to light. This unique adaptation process is still not well understood.

Action spectra. To establish the native and n-hexenal chromophore action spectra, intensity-response curves for six wavelengths of light over the range of 300 to 640 nm were determined. Straight-line extrapolation of the intensity response to the log-intensity axis yielded the threshold for each wavelength, which varied >4 orders of magnitude. The reciprocal of the threshold, the sensitivity, is plotted as a function of photon energy in Fig. 3, resulting in the action spectrum of curling behavior.

The spectrum of the native chromophore fits well with the Chlamydomonas rhodopsin standard curve (33), with a peak of 2.43 ± 0.04 eV (510 nm) and a half-spectral width at a half-maximum of 0.21 eV. Bleaching of the chromophore with high-intensity (up to 105 mW/cm2) 500-nm light abolished the response to light. Light responses (R; see Materials and Methods) on a scale ranging from approximately zero to one were 0.33 and –0.13 for unbleached and bleached samples, respectively. As expected for rhodopsin, the light response could be restored by adding retinal (1 µM), which brought the response to 0.54. Bleaching causes rhodopsin to break down to free retinal and opsin. This set of results indicates that the loss of light response after bleaching is due to the loss of native chromophore. We have shown with Chlamydomonas (14, 15, 16) and Allomyces (34) that retinal and many retinal analogs that reach and incorporate into the retinal-binding site of opsin, including n-hexenal, can restore phototaxis. Because of n-hexenal's volatility and potential for relatively rapid oxidation, special care must be taken in its use (see Materials and Methods). For this study, we chose to use this small retinal analog, which has only two conjugated double bonds in the resulting chromophore, to substitute for the native chromophore. n-Hexenal is easy to incorporate with a rapid onset, and most importantly, has the bluest absorption, the largest spectral shift in action spectrum that we can measure; hence, it is the most obviously shifted among the available retinal analogs. As expected, n-hexenal at 5 pM restores the light-stimulated curling behavior in Peranema as soon as it is added and lasts for at least 30 min. Since we always measure threshold action spectra very near the threshold, as expected and intended, the level of significance for being different from behavior in the dark was low. For example, at 460 nm with n-hexenal incorporated, the point nearest the threshold had a level of significance of 0.35. Doubling the intensity, however, gave a response time (geometric mean, 19 +6/–5 s) different from that in the dark (34 +6/–5 s), with a level of significance or P value of 0.04. Thirty-five cells were used to obtain the 460-nm point, with 28 cells on average at each wavelength.

Since the action spectrum has the typical rhodopsin shape and the spectral peak anticipated for opsin with hexenal incorporated, clearly Peranema recovers its light response due to the presence of n-hexenal in the retinal-binding site. Since retinal was removed prior to the addition of the exogenous analog, activation of that analog could not transfer its signal to a native rhodopsin molecule via a retinal chromophore. The analog must activate the opsin directly by being incorporated into the available binding site. As a short chromophore with only two conjugated double bonds, the analog has an action spectrum (Fig. 3B) shifted to a higher energy (3.73 ± 0.04 eV [332 nm]) than that of the native pigment. This peak is similar to 3.66 ± 0.35 eV (339 nm), obtained from Allomyces zoospore phototaxis (34), and 3.50 eV (354 nm), obtained from Chlamydomonas phototaxis (15). n-Hexenal-opsin was as active as rhodopsin as a photoreceptor for curling behavior, as shown by its action spectrum (Fig. 3B). The red shift of the n-hexenal-opsin action spectrum from the free n-hexenal absorption spectrum (265-nm peak in methanol) indicates that hexenal forms a chromophore with opsin. The blue shift of its action spectrum from the native one indicates that hexenal-opsin has a physiological function as a photoreceptor for the curling behavior of Peranema. Together, the results form conclusive evidence that the euglenoid Peranema uses a rhodopsin as a photoreceptor. As anticipated for such a short chromophore, the low-energy cutoff slope of the hexenal-opsin action spectrum is less than that of the native chromophore.

Calcium effects. The motion observed when the cell curls could be described as a contraction of the cell body on one side. Shettles (37) found that the reaction time to light decreased with more calcium and was the shortest with calcium relative to that with magnesium, potassium, or sodium. Since this seemed similar to skeletal muscle contraction and since rhodopsins typically couple with calcium channels (2), we decided to test the hypothesis of calcium involvement in curling. The response is strongly dependent on the external Ca2+ concentration. Decreasing the external free Ca2+ concentration by EGTA (50 to 500 µM) inhibited the curling behavior in a dose-dependent manner (Fig. 4A). At these concentrations of EGTA, motility is unaffected. Note, however, that much higher concentrations of EGTA are toxic, causing the cell to permanently curl. Table 2 shows the free Ca2+ concentration as well as the pH of a Peranema cell suspension before and after the addition of EGTA. The values for spring water only are included for comparison. Note that the cells removed 85% of the free calcium in the suspension by binding or by sequestering it in concentrated vesicles. The vesicles were observed by labeling with Calcium Green 1, a dye that is fluorescent if it binds free calcium. Consequently, one cannot accurately predict the free calcium levels from consideration of the medium (spring water in this case) alone. Under our conditions, the calcium-chelating effect of EGTA very significantly lowered the free calcium level in the Peranema sample. Free Ca2+ was reduced from 2.7 x 10–4 M without EGTA to <10–8 M with 500 µM EGTA. As shown in Table 2, EGTA specifically reduced free Ca2+ up to 4 to 5 orders of magnitude but only reduced the pH from 7.7 to 6.4 at the highest dose. According to Shettles (37), this drop in pH corresponds to an increase in the average reaction time (n = 10), from 16.5 to 18.8 s. Shettles (37) also found that the response to light was quite normal from pH 5.5 to 8.5, and our conditions were well within that range. It is likely that the EGTA effect on curling behavior is predominantly due to a reduction of extracellular free Ca2+. To further elucidate the roles of calcium in curling behavior, the effects of a calcium ionophore (A23187 [GenBank] ) and a calcium channel blocker (nifedipine) were studied. A23187 [GenBank] at 20, 40, and 80 µM increased the spontaneous response rate, as shown in Fig. 4B. The effect is similar to that of light stimulation. Both the calcium ionophore and light stimulation result in dose-dependent increases in curling probability. In the presence of the calcium ionophore, light does not have much, if any, further effect on the response time. This implies that when the cell membrane is made permeable to Ca2+, light does not have a significant additional effect. Light probably causes a Ca2+ influx through Ca2+ channels in the cell membrane. Rhodopsin might, in addition, act as a channel. For the same level of light stimulation and batch of cells, the response, R, was 0.62, but with 100 µM nifedipine, which blocks L-type calcium channels, R was 0.21, and with 300 µM nifedipine, R was 0.15. This suggests a significant role of calcium channels apart from a direct role of rhodopsin. As anticipated if light induces calcium influx and intracellular calcium induces curling, reducing the entry of calcium will reduce the curling response. These results suggest that the curling behavior of Peranema is calcium dependent and that light induces a calcium influx.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Free Ca2+ concentration and pH of Peranema cell suspension and spring water with and without EGTA


arrow
DISCUSSION
 
Rhodopsin is the photoreceptor for curling in Peranema. As we have done previously with the green alga Chlamydomonas (16) and the fungus Allomyces (34), we have shown that retinal analogs incorporated into cells that have been depleted of their native pigment shift the action spectrum of a photoreceptor in a microorganism. Furthermore, we have found a native action spectrum consistent with a rhodopsin pigment. In both the case of Chlamydomonas and that of Allomyces, subsequent work by others found rhodopsin-like amino acid sequences in Chlamydomonas and other fungi such as Neurospora. Hence, Peranema has a rhodopsin photoreceptor which has yet to be proven to be of the same family as all other rhodopsins found so far in all domains of nature. This finding has many implications for the evolution of vision in eukaryotes.

Phylogenic significance of and sequence similarities in the rhodopsin family. Peranema rhodopsin joins the growing list of nonanimal eukaryotic rhodopsins, including those identified in dinoflagellates by the evaluation of protein sequences (30) and action spectral analysis (11, 12, 18), in cryptophytes by action spectra (9, 12) and protein sequence evaluation (22), in fungi by shifted action spectra (34) and protein sequence evaluation (3), and in green algae by shifted action spectra (16) and protein sequence evaluation (20, 38). Two other probable groups with rhodopsins suggested by action spectra are the ciliates (12, 25) and foramins (12). The rhodopsins of fungi and green algae show the greatest similarity to each other and then to archaeal sensory rhodopsin IIs compared to cyanobacterial rhodopsins. This association with archaeal rhodopsins and the ubiquity of these rhodopsins in different eukaryotic groups, both bikont and unikont, suggest that their last common ancestor probably had a rhodopsin photoreceptor. Mitchell (26) suggested for the origin of cilia that a phagocytic cell ancestor possibly developed a protocilium, which rather than bending, glided on a substrate and later evolved into a beating cilium. The phagocytic Peranema organism seems to have retained this capability, and its photosensitivity, which we show here depends on rhodopsin, is found in its gliding cilium (35). A possible single exception to this ancestral origin may be a dinoflagellate whose rhodopsin sequence shows high similarity to that found in cyanobacteria (28), implying perhaps that this rhodopsin may have come recently from a lateral or horizontal transfer from a cyanobacterium.

Spectral trends in evolution. Interestingly, as one proceeds along the evolutionary rhodopsin line from bacteria and archaea toward animals, there is a significant broadening of the spectral widths of absorption and action spectra. As previously noted (33), the archaeal pigment bacteriorhodopsin has a half-width at a half-maximum of 0.16 eV, bikont eukaryotic microorganisms such as dinoflagellates, cryptomonads, and green algae have wider spectra, with a half-maximum at 0.21 eV, and unikont eukaryotes such as fungi (Allomyces [34]) and animals have spectra that are wider still, at 0.31 eV. Peranema fits with these other bikont microeukaryotes. The consensus that unikonts and bikonts evolved from a common eukaryote is consistent with a single transition from 0.16 to 0.21 eV occurring between archaea/bacteria and eukaryotes. A second transition to 0.31 eV on the unikont line occurred after the bikont-unikont split and before the animal-fungal branching. The molecular basis for these spectral changes in evolution has not yet been reported.

Signal processing and behavioral trends in evolution. In bacteria and archaea, the rhodopsins frequently serve as proton or other ion pumps and simultaneously as sensory receptors, which control the probability of a reversal of the direction of rotation of the flagella changing the direction of the organism. These organisms do not have localized receptors, and the cell responds only to the transition of swimming into or out of a light beam. The ability to move in the direction of light is weak. In some nonmotile eukaryotes in which there is no visual function, such as the fungus Leptoshaeria, the proton pumping function is maintained (42). On the other hand, in animals rhodopsins are typically associated with localized eyes as visual receptors. Peranema's lack of a localized eye and its curling behavior, which contributes to weak photo avoidance, are more like characteristics of the prokaryotes than of the eukaryotes and suggest that the last ancestor of unikonts and bikonts may have had similar primitive behavior. If this is true, then the tracking type of phototaxis behavior evolved separately in the bikont and unikont lines. This would mean that all the physical principles of central location of the eye, looking orthogonal to the swimming axis, and using constructive interference in the eyes are the results of convergent evolution (18). All phototactic microeukaryotes with rhodopsin photoreceptors, apart from Peranema, have localized eyespots, which facilitate true tracking of light and true phototaxis.

Among bikonts, cryptomonads, green algae, and Peranema appear to use rhodopsin photoreceptors, while flavins are used for stramenopile and euglenoid phototaxis. An example of a flavin photoreceptor for phototaxis is an adenylyl cyclase activated by a flavin chromophore in Euglena (21). The last common ancestor of bikonts and unikonts probably had flavin photoreceptors, an alternative type of receptor available for visual systems. Why rhodopsins came to dominate for phototaxis and then vision in some kingdoms and flavin receptors became dominant in others is not clear. An additional ancestral function of rhodopsin may be its control of gene expression. This was observed for a rhodopsin in the green alga Chlamydomonas (17) and in rats (43). In the nonmotile fungus Neurospora, rhodopsin probably (3) controls gene expression.

Rhodopsin- and calcium-dependent curling. Our results suggest that Ca2+ plays a role in curling behavior in Peranema. Light stimulates the rhodopsin receptors, which through a presently unknown mechanism open Ca2+ channels, increasing Ca2+ influx to trigger curling. Calcium influx into the cell appears to mimic the effect of light in stimulating the probability of curling. In view of the demonstration that several of the rhodopsins of Chlamydomonas are light-gated channels (27), a reasonable hypothesis is that this is true for Peranema as well. This association probably extends to the common eukaryote ancestor. Certainly, an association between rhodopsin and calcium is common in many organisms (2) and may reflect a common origin.

Evolution of the rhodopsin activation mechanism and implications of n-hexenal activity in microeukaryotes for in vivo activation of rhodopsins. It is of interest to know the properties of the ancestral rhodopsins of unikonts and bikonts and how those properties have evolved since. Two types of rhodopsins probably existed side by side in these ancestors, including those which performed an energy conversion function like that of bacteriorhodopsin and those with a sensory function. For a strictly sensory function, no specific bond, such as the 11-12 bond in the human retinal chromophore, needs to be isomerized to activate a rhodopsin. Isomerization about any bond would appear to be sufficient. In vivo evidence was shown with the bikont Chlamydomonas rhodopsin (13, 14), for which active analogs were found with each bond individually blocked. More recently, this has also been found in unikont animal kingdom rhodopsins (10, 23). Hence, rhodopsins are not required during activation to sense a localized motion about a particular point of the chromophore, making it unlikely that such a motion is critical for its activation.

The observed activity of the truncated chromophore n-hexenal (Fig. 3) puts a limit on how much chromophore is needed to activate rhodopsin. n-Hexenal was also active with rhodopsin in the bikont Chlamydomonas (13, 14, 15) and in the zoospore of the unikont chytridiomycete fungus Allomyces (34). In Chlamydomonas, n-hexenal recovered the sensitivity of phototaxis at 354 nm to 2.9 orders of magnitude above the background, with no loss of sensitivity relative to the native chromophore and with an action spectrum consistent for n-hexenal-opsin amply demonstrating its activity (29). Whether this property has evolved to be different in animals has not been adequately tested. A chromophore strain may not be needed for activation, since n-hexenal lacks the ß-ionone ring and a significant part of the normal conjugated chain found in retinal. As evident from the n-hexenal structure, assuming isomerization about the only bond available for cis-trans isomerization, the only structural change that would occur is that the short saturated chain would flop the other way. Such a small change in such a weak-kneed molecule (only two conjugated double bonds in the chromophore) is very unlikely to produce significant strain on the site to result in activation of the photoreceptor. The opsin protein is specifically sensitive to what happens at the site near the lysine N to which the chromophore is attached, as this is the only area that neighbors the n-hexenal chromophore.

Without a need for the strain of the chromophore acting on the protein, something else must be driving the protein conformation change. Rousso et al. (31) used atomic force microscopy to show that bacteriorhodopsin's conformation changed with isomerization-locked chromophores (no strain on the receptor site) and that an electrically asymmetric chromophore was required for activation, presumably to establish charge separation in the excited state. This point was confirmed with analogs showing that this charge redistribution or displacement is necessary for initiating the bacteriorhodopsin photocycle (44). Hence, it would seem reasonable for the sensory role of rhodopsins in Peranema, Chlamydomonas, and Allomyces (13, 14, 15, 29), for which the n-hexenal is an active chromophore, that a similar charge displacement in the photoreceptor protein is sufficient to drive a conformation change and activation.

Peranema as a new model system. This work provides another unicellular model for the comparative study of rhodopsins. The model is convenient and inexpensive for studying rhodopsin activation and its transduction pathway. A particularly noteworthy advantage is that molecules of all sizes and charges can enter this organism simply by being placed in solution. We observed a whole Chlamydomonas cell being trapped inside Peranema by phagocytosis in seconds. While substances in the food vacuoles inside the cell are not guaranteed access to the intracellular compartment, chemicals that rapidly show toxicity must have gone to their sites of action. There is no sexual cycle or photosynthesis to interfere with behavioral studies. Saito et al. (32) have used Peranema as a model system for gliding motility. Furthermore, they have developed a monoxenic culture system so that it is possible to obtain sufficient bacterium-free cells for biochemical and molecular biological studies. The development of further biochemical and genomic analysis tools would seem appropriate.

Conclusions. Rhodopsin is the receptor that controls curling in Peranema. This response involves calcium signaling. The study of this curling photoresponse has yielded new insight into the evolution of eukaryotic vision and better predictions of the properties of the ancestral eukaryotic rhodopsin. This ancestral sensory rhodopsin does not appear to require isomerization of its chromophore for activation. Rather, its activation may be best described by an essential charge redistribution near the lysine nitrogen, followed by isomerization if necessary to move the ß-ionone ring of the native chromophore out of the way. Both unikont and bikont rhodopsin receptors are probably predominantly derived from an ancestral eukaryote with a characteristic spectral width of more than that of bacteriorhodopsins. The spectral width of Peranema's rhodopsin is consistent with that of all other bikonts measured and is distinct from the greater spectral width of all measured unikont rhodopsins. Peranema has possibly maintained several other traits postulated for the ancestral eukaryote, namely, phagocytosis, gliding behavior, and an association of rhodopsin with the cilium. This makes it an interesting organism for further study.


arrow
ACKNOWLEDGMENTS
 
Nisa Foster and Narie Foster started this work as their middle and high school science fair projects, and later Jeanne Hansen and Nicholas Peruzzi continued it as their high school summer projects. Their scientific curiosity and enthusiasm combined with industrious data collection contributed to this finding.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: 201 Physics Building, Syracuse University, Syracuse, NY 13244-1130. Phone: (315) 443-9220. Fax: (315) 443-9103. E-mail: foster{at}phy.syr.edu. Back


arrow
REFERENCES
 
    1
  1. Allen, J. R., J. J. Lee, S. H. Hutner, and J. Storm. 1966. Prolonged culture of the voracious flagellate Peranema in antioxidant-containing media. Protozoology 13:103-108.
    2. 2
  2. Baehr, W., and K. Palczewski. 2002. Photoreceptors and calcium, p. 632. Springer, Berlin, Germany.
    3. 3
  3. Bieszke, J. A., E. L. Braun, L. E. Bean, S. Kang, D. O. Natvig, and K. A. Borkovich. 1999. The nop-1 gene of Neurospora crassa encodes a seven transmembrane helix retinal-binding protein homologous to archaeal rhodopsins. Proc. Natl. Acad. Sci. USA 96:8034-8039.[Abstract/Free Full Text]
    4. 4
  4. Cavalier-Smith, T. 1981. Eukaryote kingdoms: seven or nine? BioSystems 14:461-481.[CrossRef][Medline]
    5. 5
  5. Cavalier-Smith, T. 1986. Cilia versus undulipodia. Bioscience 36:293.[CrossRef]
    6. 6
  6. Cavalier-Smith, T. 2004. Only six kingdoms of life. Proc. R. Soc. London Ser. B 271:1251-1262.[Medline]
    7. 7
  7. Douzery, E. J. P., E. A. Snell, E. Bapteste, F. Delsuc, and H. Philippe. 2004. The timing of eukaryotic evolution: does a relaxed molecular clock reconcile proteins and fossils? Proc. Natl. Acad. Sci. USA 101:15386-15391.[Abstract/Free Full Text]
    8. 8
  8. Ehrenberg, C. G. 1838. Die infusionsthierchen als vollkommene organismen. Verlag von Leopold Voss, Leipzig, Germany.
    9. 9
  9. Erata, M., M. Kubota, T. Takahashi, I. Inouye, and M. Watanabe. 1995. Ultrastructure and phototactic action spectra of two genera of cryptophyte flagellate algae, Cryptomonas and Chroomonas. Protoplasma 188:258-266.[CrossRef]
    10. 10
  10. Fan, G., F. Siebert, M. Sheves, and R. Vogel. 2002. Rhodopsin with 11-cis locked chromophore is capable of forming an active state photoproduct. J. Biol. Chem. 277:40229-40234.[Abstract/Free Full Text]
    11. 11
  11. Forward, R. B., Jr. 1974. Phototaxis by the dinoflagellate Gymnodinium splendens Lebour. J. Protozool. 21:312-315.[Medline]
    12. 12
  12. Foster, K. W. 2001. Action spectroscopy of photomovement, p. 51-115. In D.-P. Häder and M. Lebert (ed.), Photomovement. Comprehensive series in the photosciences, vol. 1. Elsevier, Amsterdam, The Netherlands.
    13. 13
  13. Foster, K. W., J. Saranak, F. Derguini, R. V. Jayathirtha, G. R. Zarrilli, M. Okabe, J.-M. Fang, N. Shimizu, and K. Nakanishi. 1988. Rhodopsin activation: a novel view suggested by in vivo Chlamydomonas experiments. J. Am. Chem. Soc. 110:6588-6589.[CrossRef]
    14. 14
  14. Foster, K. W., J. Saranak, F. Derguini, G. R. Zarrilli, R. Johnson, M. Okabe, and K. Nakanishi. 1989. Activation of Chlamydomonas rhodopsin in vivo does not require isomerization of retinal. Biochemistry 28:819-824.[CrossRef][Medline]
    15. 15
  15. Foster, K. W., J. Saranak, and P. A. Dowben. 1991. Spectral sensitivity, structure, and activation of eukaryotic rhodopsins: activation spectroscopy of rhodopsin analogs in Chlamydomonas. J. Photochem. Photobiol. B 8:385-408.[CrossRef][Medline]
    16. 16
  16. Foster, K. W., J. Saranak, N. Patel, G. Zarrilli, M. Okabe, T. Kline, and K. Nakanishi. 1984. A rhodopsin is the functioning photoreceptor for phototaxis in the unicellular eukaryote Chlamydomonas. Nature 311:756-759.[CrossRef][Medline]
    17. 17
  17. Foster, K. W., J. Saranak, and G. Zarrilli. 1988. Autoregulation of rhodopsin synthesis in Chlamydomonas reinhardtii. Proc. Natl. Acad. Sci. USA 85:6379-6383.[Abstract/Free Full Text]
    18. 18
  18. Foster, K. W., and R. D. Smyth. 1980. Light antennas in phototactic algae. Microbiol. Rev. 44:572-630.[Free Full Text]
    19. 19
  19. Hegemann, P., W. Gärtner, and R. Uhl. 1991. All-trans retinal constitutes the functional chromophore in Chlamydomonas rhodopsin. Biophys. J. 60:1477-1489.
    20. 20
  20. Hegemann, P., M. Fuhrmann, and S. Kateriya. 2001. Sensory algal photoreceptors. J. Phycol. 37:6668-6676.
    21. 21
  21. Iseki, M., S. Matsunaga, A. Murakami, K. Ohno, K. Shiga, K. Yoshida, M. Sugai, T. Takahashi, T. Hori, and M. Watanabe. 2002. A blue-light-activated adenylyl cyclase mediates photoavoidance in Euglena gracilis. Nature 415:1047-1051.[CrossRef][Medline]
    22. 22
  22. Jung, K.-H., and J. L. Spudich. 2004. Microbial rhodopsins: transport and sensory proteins throughout the three domains of life, p. 1-12. In W. M. Horspool et al. (ed.), CRC handbook of organic photochemistry and photobiology. CRC Press, Boca Raton, Fla.
    23. 23
  23. Kuksa, V., F. Bartl, T. Maeda, G.-F. Jang, E. Ritter, M. Heck, J. P. Van Hooser, Y. Liang, S. Filipek, M. H. Gelb, K. P. Hofmann, and K. Palczewski. 2002. Biochemical and physiological properties of rhodopsin regenerated with 11-cis-6-ring and -7-ring retinals. J. Biol. Chem. 277:42315-42324.[Abstract/Free Full Text]
    24. 24
  24. Manton, I. 1965. Some phyletic implications of flagellar structure in plants. Adv. Bot. Res. 2:1-34.
    25. 25
  25. Marangoni, R., S. Puntoni, L. Favati, and G. Colombetti. 1994. Phototaxis in Fabrea salina. I. Action spectrum determination. J. Photochem. Photobiol. B 23:149-154.[CrossRef]
    26. 26
  26. Mitchell, D. R. 2004. Speculations on the evolution of 9+2 organelles and the role of central pair microtubules. Biol. Cell 96:691-696.[CrossRef][Medline]
    27. 27
  27. Nagel, G., T. Szellas, W. Huhn, S. Kateriya, N. Adeishvili, P. Berthold, D. Ollig, P. Hegemann, and E. Bamberg. 2003. Channelrhodopsin-2, a directly light-gated cation-selective membrane channel. Proc. Natl. Acad. Sci. USA 100:13940-13945.[Abstract/Free Full Text]
    28. 28
  28. Nakamura, Y., T. Kaneko, S. Sato, M. Mimuro, H. Miyashita, T. Tsuchiya, S. Sasamoto, A. Watanabe, K. Kawashima, Y. Kishida, C. Kiyokawa, M. Kohara, M. Matsumoto, A. Matsuno, N. Nakazaki, S. Shimpo, C. Takeuchi, M. Yamada, and S. Tabata. 2003. Complete genome structure of Gloeobacter violaceus PCC 7421, a cyanobacterium that lacks thylakoids. DNA Res. 10(Suppl.):181-201.[CrossRef][Medline]
    29. 29
  29. Nakanishi, K., F. Derguini, V. J. Rao, G. Zarrilli, M. Okabe, T. Lien, R. Johnson, K. W. Foster, and J. Saranak. 1989. Theory of rhodopsin activation: probable charge redistribution of excited state chromophore. Pure Appl. Chem. 61:361-364.
    30. 30
  30. Okamoto, O. K., and J. W. Hastings. 2003. Novel dinoflagellate clock-related genes identified through microarray analysis. J. Phycol. 39:519-526.[CrossRef]
    31. 31
  31. Rousso, I., E. Khachatryan, Y. Gat, I. Brodsky, M. Ottolenghi, M. Sheves, and A. Lewis. 1997. Microsecond atomic force sensing of protein conformational dynamics: implications for the primary light-induced events in bacteriorhodopsin. Proc. Natl. Acad. Sci. USA 94:7937-7941.[Abstract/Free Full Text]
    32. 32
  32. Saito, A., Y. Suetomo, M. Arikawa, G. Omura, S. M. M. K. Khan, S. Kakuta, E. Suzaki, K. Kataoka, and T. Suzaki. 2003. Gliding movement in Peranema tricophorum is powered by flagellar surface motility. Cell Motil. Cytoskel. 55:244-253.[CrossRef][Medline]
    33. 33
  33. Saranak, J., and K. W. Foster. 1994. The in vivo cleavage of carotenoids into retinoids in Chlamydomonas reinhardtii. J. Exp. Bot. 45:505-511.[Abstract/Free Full Text]
    34. 34
  34. Saranak, J., and K. W. Foster. 1997. Rhodopsin guides fungal phototaxis. Nature 387:465-466.[CrossRef][Medline]
    35. 35
  35. Shettles, L. B. 1937. The response to light in Peranema trichophorum with special reference to dark-adaptation and light-adaptation. J. Exp. Zool. 77:215-249.[CrossRef]
    36. 36
  36. Shettles, L. B. 1938. Effect of ultraviolet light and X-rays on Peranema trichophorum. J. Cell Comp. Physiol. 12:263-272.[CrossRef]
    37. 37
  37. Shettles, L. B. 1939. The relation between salts, hydrogen-ion concentration and sensitivity to light in Peranema trichophorum. J. Exp. Zool. 81:473-477.[CrossRef]
    38. 38
  38. Sineshchekov, O. A., K. H. Jung, and J. L. Spudich. 2002. Two rhodopsins mediate phototaxis to low- and high-intensity light in Chlamydomonas reinhardtii. Proc. Natl. Acad. Sci. USA 99:8689-8694.[Abstract/Free Full Text]
    39. 39
  39. Stechmann, A., and T. Cavalier-Smith. 2002. Rooting the eukaryote tree by using a derived gene fusion. Science 297:89-91.[Abstract/Free Full Text]
    40. 40
  40. Teller, D. C., T. Okada, C. A. Behnke, K. Palczewski, and R. E. Stenkamp. 2001. Advances in determination of a high-resolution three-dimensional structure of rhodopsin, a model of G-protein-coupled receptors (GPCRs). Biochemistry 40:7761-7772.[CrossRef][Medline]
    41. 41
  41. Triemer, R. E. 1997. Feeding in Peranema trichophorum revisited. J. Phycol. 33:649-654.[CrossRef]
    42. 42
  42. Waschuk, S. A., A. G. Bezerra, Jr., L. Shi, and L. S. Brown. 2005. Leptosphaeria rhodopsin: bacteriorhodopsin-like proton pump from a eukaryote. Proc. Natl. Acad. Sci. USA 102:6879-6883.[Abstract/Free Full Text]
    43. 43
  43. Williams, T. P., A. Squitieri, R. P. Henderson, and J. P. P. Webbers. 1999. Reciprocity between light intensity and rhodopsin concentration across the rat retina. J. Physiol. (London) 516:869-874.[Abstract/Free Full Text]
    44. 44
  44. Zadok, U., A. Khatchatouriants, A. Lewis, M. Ottolenghi, and M. Sheves. 2002. Light-induced charge redistribution in the retinal chromophore is required for initiating the bacteriorhodopsin photocycle. J. Am. Chem. Soc. 124:11844-11845.[CrossRef][Medline]

Eukaryotic Cell, October 2005, p. 1605-1612, Vol. 4, No. 10
1535-9778/05/$08.00+0     doi:10.1128/EC.4.10.1605-1612.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Jekely, G. (2009). Evolution of phototaxis. Phil Trans R Soc B 364: 2795-2808 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Saranak, J.
Right arrow Articles by Foster, K. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Saranak, J.
Right arrow Articles by Foster, K. W.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»