Bacterized plant infusion is a popular culture medium for Paramecium, using Klebsiella pneumoniae for the bacterium and Wheat Grass Powder (WGP) for the plant. It has been thought that WGP plays a role in the growth of bacteria, which in turn serve as the direct food for paramecia. However, we found that bacteria suspended in saline solution were unable to support the growth of paramecia. WGP including no bacteria was able to support neither the growth nor the survival of paramecia; instead, it killed paramecia. The killing effect of the WGP-derived substance(s), estimated to be of molecular weight less than 1,000, was abolished when bacteria were once grown in the WGP and then eliminated, suggesting that bacteria might change the toxic substance into an inactive form. This inactivation of the toxic substance may be caused either by metabolization inside of the bacteria or by neutralization by means of bacteria-derived substance outside of the bacteria. The second alternative is likely, because paramecia were able to survive and grow in the WGP medium containing a sufficient amount of dead bacteria killed by formalin or kanamycin. Dead bacteria killed by autoclaving were ineffective, probably because bacterial contents were lost. These findings revealed an ectosymbiotic role of bacteria; they confer benefits upon paramecia not only as food but also as machinery to detoxicate a plant toxin.
Culture medium is fundamental to any kind of microbiological experiment. Experimental results may be different if biological materials prepared in different culture media are used. For example, the experimental outcomes in studies of Paramecium tetraurelia regarding the cellular and ciliary components of phospholipids, fatty acids and sterols (Andrews and Nelson, 1979; Hennessey et al., 1983; Kaneshiro et al., 1979; Rhoads and Kaneshiro, 1979), the growth rate, the final cell density, the accumulation of macronuclear DNA fragments (Holmes and Holmes, 1986), and the clonal life span (Fukushima et al., 1990) were different when chemically defined medium or axenic medium was used instead of bacterized plant infusions. The clonal life span of P. tetraurelia was reduced when the bacterized plant infusions prepared using a reduced concentration of the plant (Wheat Grass Powder, WGP) were used (Takagi et al., 1987). Human diploid fibroblasts, which normally have a limited division potential became able to divide indefinitely if serum in the standard culture medium was replaced with certain growth factors (Hayashi and Sato, 1976).
For the growth of Paramecium, two-component culture medium, composed of plants (WGP, lettuce) and bacteria (Klebsiella pneumoniae), has been most frequently used. It has been taken for granted that in the two-component culture medium, the plant infusion plays the role of allowing the growth of the bacteria, which in turn serve as the direct food for paramecia (Sonneborn, 1950, 1970). However, the role of each component in the growth of paramecia has not been examined critically, except for cases in which nutritional requirements were studied in axenic or chemically defined medium (Nerad and Daggett, 1992; Soldo, 1992; Van Wagtendonk, 1974).
In our laboratory, Paramecium has usually been cultured in WGP medium inoculated with K. pneumoniae 2 days before use. The density of K. pneumoniae increases from ca. 103 cells/ml to ca. 108 cells/ml during the 2 days in WGP medium. The present study addresses the simple question of whether Paramecium can be grown on bacteria alone. The negative answer that we found to this question led us to further studies giving fresh insight into the ectosymbiotic role of bacteria for Paramecium.
MATERIALS AND METHODS
Cells and maintenance of stock cultures
Wild-type stock 51 (mating type VII) of Paramecium tetraurelia was generally used. (When referred to as paramecia, we mean this stock of P. tetraurelia cells unless otherwise stated.) Other stocks used were G3 (mating type O) of P. caudatum, stock YM26 (mating type III) of P. multimicronucleatum collected from the wild by Mr. Y. Murakami of Ishinomaki Sensyu University, stock G312 (mating type undecided) of P. bursaria and stock 4B-1 (mating type unknown) of P. trichium provided by Dr. T. Takahashi of Hiroshima University.
The stock cultures were maintained in test tubes covered with stainless steel caps at 17°C, replacing half of the culture with an equal volume of fresh culture medium every 3 weeks. For starting experiments, cells from the stock culture were washed as described (Sonneborn, 1970) and grown as clones.
The standard culture medium was 0.5% phosphate-buffered WGP (Wheat Grass Powder; Pines International, Inc., USA) medium inoculated with Klebsiella pneumoniae 2 days before use (designated 2×WGP; Takagi et al., 1987). WGP is identical to Cerophyl, and K. pneumoniae is identical to Aerobacter aerogenes in former reports. We describe the method (basically after Sonneborn, 1970) in detail because the culture medium is critical in this study. WGP in double-distilled water (5 g/l) was boiled for 5 min. After cooling, WGP was removed by filtration through filter paper. The infusion contained in the remaining wet WGP-paste was squeezed out through gauze folded successively eight times. Na2HPO4·12H2O (0.5 g /l) was added, and the resulting infusion was bottled in flasks capped with silicone sponge closures (Sigma), and autoclaved for 20 min at a pressure of 1.2 atm. The WGP infusion was inoculated with bacteria at about 103 cells/ml and incubated at 26°C for 2 days to grow the bacteria to the saturation density of about 108 cells/ml. The bacterized WGP medium was adjusted to pH 6.8 with NaOH before inoculating with Paramecium. When various concentrations of WGP ranging from 6xWGP to 1/64xWGP were used, we first prepared 6×WGP (15 g/l) diluted in autoclaved 0.05% Na2HPO4·12H2O.
Lettuce juice medium was prepared as follows (basically after Miyake, 1968). Fresh lettuce leaves were washed and immersed in boiling distilled water for a few seconds to inactivate enzymes. The leaves were treated with the juicer repeatedly, squeezed out through gauze folded successively eight times and filtered through filter paper. The juice was stored in a freezer (−25°C). After thawing, the juice was diluted to 2% with double-distilled water, buffered by sodium phosphate (2 mM Na2HPO4·2H2O+Na2HPO4·12H2O, pH 6.8), and autoclaved. The bacterized lettuce medium, usually containing a saturation density of about 108 bacteria/ml, was adjusted to pH 6.8 with NaOH before inoculating with Paramecium.
K. pneumoniae was grown on agar slants (25 g/l bacto agar, 0.2 g/l glucose, 5 g/l yeast extract) in a test tube capped with a silicone sponge closure (Sigma) at 26°C. Other bacteria used were Escherichia coli (Tübingen-3, TF1002 and TF1003) provided by Dr. H. Tanabe of Kinki University, which were grown on LB plates containing 50 μg/ml ampicillin. Tübingen-3 (NeorI) contains ampicillin- and neomycin-resistance genes on a plasmid and a kanamycin-resistance gene in its chromosomal DNA. TF1002 (NeosII) contains the ampicillin-resistance gene but no neomycin-resistance gene on a plasmid and no kanamycin-resistance gene in its chromosomal DNA. TF1003 (NeorII) contains the ampicillin- and neomycin-resistance genes on a plasmid but no kanamycin-resistantce gene in its chromosomal DNA.
Bacteria from a slant were collected in 2 ml of distilled water that had been sterilized by autoclaving, and then were sedimented by centrifugation at 10,000 rpm (8,600 g) for 10 min. The harvested bacteria were resuspended in 0.5 ml of sterilized distilled water. The bacterial concentration of this suspension was 1010 cells/ml.
To eliminate bacteria from the bacterized 2×WGP, the culture was centrifuged for 15 min at 10,000 rpm and then filtered with DISMIC-25cs (Toyo Roshi Kaisha, Ltd.).
To prepare dead bacteria, a final concentration of 1 mg/ml kanamycin or 20% formalin was added to the suspension containing 1010 bacteria/ml. After 24 hr, the suspensions were centrifuged. Killed bacteria were washed three times in 2 ml of sterile distilled water and were resuspended in 0.5 ml of sterile distilled water. This suspension was checked for the absence of living bacteria by colony formation.
Bacteria are found growing in autoclaved WGP medium after Paramecium cells are introduced, because bacteria are brought in with the paramecia. Paramecia may live closely associated with bacteria so that they are not separable unless antibiotics are used. We therefore treated an early stationary phase culture of paramecia with 50 μg/ml kanamycin for 24 hr at 26°C. At this concentration of kanamycin, there were no detectable effects on paramecia, in terms of the rate of cell division in isolation cultures, the saturation cell density in mass cultures, and the rate of food vacuole formation.
Survival test of paramecia
Five bacteria-free paramecia were placed in each well of 3-well depression slides, each containing 0.5 ml of test medium. A set of 9 wells were prepared to obtain a single data point for experiments to examine the rate of death (or % survival) in each test medium. For daily check of survival, for example, 3 sets (27 wells of 9 depression slides) were prepared, and each set was monitored daily for 3 days. For hourly check of survival, the necessary numbers of sets were prepared depending on the length of time of the observations. All cultures were incubated in a moist chamber at 26°C. In the experiments in which bacteria-free WGP inoculated with kanamycin-treated paramecia was used, the absence of living bacteria was checked by colony formation.
SMBIII (Miyake, 1981) was used as physiological balanced solution (1.5 mM NaCl, 0.05 mM KCl, 0.4 mM CaCl2, 0.05 mM MgCl2, 0.05 mM MgSO4, 2 mM Na-phosphate buffer pH 6.8, 2×10−3 mM EDTA) for both paramecia and bacteria. Dryl's solution (2 mM sodium citrate, 1 mM NaH2PO4·2H2O, 1 mM Na2HPO4·12H2O, 1.5 mM CaCl2) (Dryl, 1959) was also used in some experiments. Bacteria can grow in SMBIII to ∼106 cells/ml.
2×WGP was subjected to ultrafiltration using a Diaflow membrane YM10 (Amicon) with a nominal cut-off molecular weight of 10,000. The resulting fraction with a molecular weight of less than 10,000 was subjected to ultrafiltration using a Diaflow membrane YM1 (Amicon) with a nominal cut-off molecular weight of 1,000. Thus, 2×WGP was separated into three fractions of molecular weight <1,000, 1,000∼10,000 and >10,000. Each fraction was supplemented with double-distilled water containing 0.05% Na2HPO4·12H2O in order to achieve the original volume.
Growth of paramecia in bacterized WGP medium with different buffers, different bacteria and different concentrations of WGP
The bacterized 2×WGP medium contains buffering chemicals, WGP-extracts and bacteria. We examined the effects of each of these components on the growth of paramecia.
WGP was dissolved in (a) 0.05% Na2HPO4·12H2O, or (b) SMBIII, or (c) Dryl's solution, or (d) 10 mM Tris-HCl pH 6.8, or (e) double-distilled water, and then inoculated with K. pneumoniae. The standard is (a). Paramecia grew well in all of the media with similar growth kinetics (data not shown).
WGP was dissolved in 0.05% Na2HPO4·12H2O, and then inoculated with (a) K. pneumoniae, or (b) Escherichia coli, Tübingen-3, or (c) E. coli, TF1002, or (d) E. coli, TF1003. The standard is (a). Paramecia grew similarly in all of the media, although there was a time-lag before starting exponential growth in Escherichia medium when transferred from Klebsiella medium (data not shown).
WGP of 10 different concentrations ranging from 6× to 1/64× (the standard is 2×) was dissolved in 0.05% Na2HPO4·12H2O and inoculated with K. pneumoniae. Two days later, bacterial density was counted in each medium and then 5 paramecia were introduced into each medium to study their growth kinetics. The growth was vigorous in 6×, 4×, 2× and 1× medium, weak in 1/8×, 1/16×, 1/32× and 1/64× medium, and intermediate in 1/2× and 1/4× medium (Fig. 1A). The saturation density of bacteria decreased correspondingly as the WGP concentration decreased (Fig. 1B); in vigorously growing cultures, bacterial density was on the order of 108 cells/ml.
Altogether, it was concluded that at least for short-term experiments, as in this case, different kinds of buffers and bacteria were similarly effective for the growth of paramecia. However, the concentration of WGP and density of bacteria were crucial for the growth of paramecia.
Bacteria alone can not support the growth of paramecia
The results in the previous section support the traditional idea that WGP is used for the growth of the bacteria and that a sufficient number of bacteria (on the order of 108 cells/ml) is necessary for the growth of paramecia. To test this idea, we concentrated bacteria, suspended them in saline solution (SMBIII) at various densities, and introduced 5 paramecia into each suspension. Paramecia could survive, but were unable to divide more than 2 times even in the medium containing bacteria at 109 cells/ml (Fig. 2); they began to grow if WGP was supplied in that medium (data not shown). We conclude, therefore, that the growth of paramecia can not be supported by bacteria alone, but requires a nutritional supply from WGP.
Paramecia die in WGP medium containing no bacteria
To answer the question if WGP alone can support the growth of paramecia, we placed paramecia at early stationary phase in WGP medium containing no bacteria. We have tried to wash out bacteria by repeatedly transferring paramecia into sterilized SMBIII before introducing them into WGP medium. It was, however, impossible to eliminate bacteria completely. Bacteria that escaped from washing grew finally to 106 cells/ml in SMBIII and to 108 cells/ml in WGP medium. We therefore used kanamycin. The treatment of paramecia with 50 μg/ml kanamycin for 24 hr was found sufficient both to eliminate bacteria and to keep paramecia uninjured: bacterial elimination was confirmed by colony counting on agar plates; the normality of paramecia was confirmed by obtaining identical results between treated and untreated cells in comparisons of the fission rate in isolation cultures, final cell density in mass cultures, and the rate of feeding activity (data not shown).
Five paramecia that were pretreated with kanamycin were placed in each well of depression slides containing 10 different concentrations of WGP ranging from 6× to 1/64× as in Fig. 1, and the number of surviving cells was counted for 3 days (Fig. 3A). Paramecia that were placed in 6×WGP–1/4×WGP died out within the period of observation, while those that were placed in 1/8×WGP–1/64×WGP remained alive, at least partly (Fig. 3A). These results strongly suggest that some substance(s) that is toxic to paramecia is contained in WGP. In 2×WGP containing no bacteria, paramecia began to die after about 7 hr of incubation, while in 2×WGP with bacteria, paramecia began to divide at this time (Fig. 3B).
To estimate the molecular weight of the toxic substance, 2×WGP was separated by ultrafiltration using two membranes into three fractions of molecular weights less than 1,000, from 1,000 to 10,000, and more than 10,000. Paramecia died out on the first day in the fraction of molecular weight less than 1,000, but survived for 3 days in the other two fractions, although the number of living cells decreased (Fig. 3C). It is supposed, therefore, that some substance(s) with molecular weight less than 1,000 is the main component of the toxic substance(s).
We then examined whether the above findings can be extended to species other than P. tetraurelia and to a plant other than WGP. First we studied how cells of P. caudatum, P. multimicronucleatum, P. bursaria and P. trichium grow or die in 2×WGP with and without bacteria. The results were similar to those with P. tetraurelia; every species grew in bacterized 2×WGP, but died out before long in 2×WGP in which bacteria were not inoculated (Fig. 4). We next studied how paramecia grow in lettuce juice with and without bacteria. The results were again similar to those with WGP; paramecia grew in bacterized lettuce juice at the concentrations of 0.5%, 1%, 2% and 4%, but died out at 8% (Fig. 5A), while in bacteria-free lettuce juice they died out faster as the concentration of the lettuce juice was increased (Fig. 5B).
Toxic substance(s) contained in WGP is eliminated by bacteria
Some toxic substance(s) is contained in 2×WGP, but paramecia can grow in 2×WGP in which bacteria are inoculated 2 days before use. It is therefore likely that the toxic substance is eliminated during the 2 days in which bacteria grew to the saturation density. To test this hypothesis, we conducted the following experiments.
The bacteria were eliminated from the 2-day-old bacterized 2×WGP by centrifugation and filtration, and paramecia were placed in this bacteria-free 2×WGP. Paramecia survived in this medium, but did not grow. Paramecia were able to grow if bacteria were inoculated again. Bacteria grew to 108 cells/ml in this medium (Fig. 6). Similar experiments were conducted by using 2xWGP in which bacteria had been incubated for 1, 2, 3, 4 and 5 days and then eliminated. The results were similar in all of the experiments: paramecia survived but did not grow (data not shown). These results show that the toxic substance in 2×WGP is eliminated by bacteria in a day or more, and support the conclusion in the previous section showing that paramecia require both bacteria and WGP for growth.
Do paramecia survive or die if they are placed in 2×WGP immediately after bacteria are inoculated? To answer this question, paramecia were placed in 2×WGP immediately after bacteria of various densities ranging from 103 to 108 cells/ml were inoculated. In 2×WGP in which bacteria were suspended at 103 and 104 cells/ml, paramecia died out within the period of observation; in medium containing 105 bacteria/ml, paramecia began to grow on the second day; in medium containing 106, 107 or 108 bacteria/ml, paramecia began to grow on the first day (Fig. 7A). This rapid detoxification by a sufficient amount of bacteria suggests the elimination of the WGP-toxin by adsorption on bacterial cell surface or by neutralization by means of bacteria-derived substances, rather than by metabolization inside of the bacteria.
To test further this possibility, we performed similar experiments using dead bacteria. Bacteria were killed by formalin, kanamycin or autoclaving. We confirmed, by colony formation, that bacteria were not alive. As shown in Fig. 7B, paramecia grew in 2×WGP in which bacteria killed by formalin or kanamycin were suspended at the density of 108 cells/ml, while they tended to be reduced in cell number and to die out early or late in 2×WGP, in which bacteria killed by auto-claving were suspended at the density of 108 cells/ml. Bacterial cell contents may remain almost intact when killed by formalin or kanamycin, but may have been lost when killed by autoclaving. Because bacteria killed by autoclaving were countable under microscope, they had not been broken to pieces but only partly destroyed. The elimination of the WGP-toxin may, therefore, be caused by neutralization by means of bacteria-derived substance rather than by adsorption on the bacterial cell surface, although the latter possibility can not be completely ruled out.
Paramecia, except for autotrophic species, are usually found in nature in water that is rich in organic substances and thus also rich in bacteria. Many kinds of combinations of plants (leaves and grains) and microorganisms (bacteria, yeast and algae) have been used for the cultivation of paramecia, although they differ greatly in suitability (Sonneborn, 1950, 1970). In our laboratory, phosphate-buffered WGP in which Klebsiella pneumoniae have been grown has long been used. Our stock of P. tetraurelia that has long been cultured in this medium was able to grow similarly in the medium prepared with different buffering solutions and with E. coli instead of K. pneumoniae. However, paramecia were unable to grow if WGP of low concentrations that resulted in a low density of bacteria was used (Fig. 1). This result supports the traditional interpretation of the role of plant infusion and bacteria as the food for paramecia: plant infusion is used for the growth of bacteria, and bacteria are used for the growth of paramecia (Sonneborn, 1950, 1970).
We found that this was not true, because neither as many as 109 bacteria/ml suspended in saline solution nor WGP without bacteria was able to support the growth of paramecia (Figs. 2, 3); both WGP and bacteria are necessary for the growth of paramecia. It was surprising that paramecia were not only unable to grow but were also killed in WGP containing no bacteria (Fig. 3). WGP apparently contains some toxic substance(s) as well as essential nutrients for paramecia. Similar substances are also contained in lettuce (Fig. 5). The key substance is of molecular weight less than 1,000 (Fig. 3C), although we can not exclude the possibility of involvement of additional components.
The most important finding in this study is that the toxic substance is changed to an inactive form by bacteria. Paramecia were killed in 2×WGP containing no bacteria, but survived in 2×WGP in which bacteria were once grown to 108 cells/ml in 2 days and then eliminated (Fig. 6). Interestingly, paramecia not only survived but also grew in 2×WGP in which 108 bacteria/ml killed by formalin or kanamycin were suspended (Fig. 7B). Since dead bacteria can not metabolize the WGP-toxin, the toxin may be eliminated either by adsorption on bacterial cell surface or by neutralization by some substance(s) released from killed bacteria. Bacteria killed by autoclaving looked intact in that they were countable under microscope and yet they were ineffective (Fig. 7B), suggesting that the detoxification by neutralization rather than by adsorption appears more likely. It may be that bacteria killed by autoclaving might have lost their cell contents more than those killed by formalin or kanamycin, or that the bacteria-derived substance could be a polypeptide and thus would have been inactivated by heat during autoclaving.
It also remains open whether the WGP-toxin is neutralized before or after it is incorporated into paramecia. This is related to the question whether the WGP-toxin acts externally on the cell surface of paramecia or acts internally after it is incorporated into food vacuoles. The result in Fig. 3B showing that paramecia began to die 7 hr after isolation into 2×WGP appears to support the internal action, although not decisive. Conjugating paramecia lacking the ability to form food vacuoles (Mikami, 1988) may be useful to examine this problem.
The results revealed in this study lead us to an understanding of the unexpected role of bacteria in the growth of paramecia. Bacteria play a critical role in saving the lives of paramecia not only by providing food resources but also by eliminating toxic components of food plants. Also in a physical sense, paramecia are living so closely with bacteria that bacteria can not be completely washed away from paramecia by passing repeatedly through sterilized saline-solution unless antibiotics are used. We thus understand that food bacteria living on and around paramecia may be regarded as ectosymbionts for host paramecia.
We are grateful to Dr. T. Harumoto for helpful discussions, to Dr. T. Takahashi for providing P. bursaria and P. trichium, to Dr. H. Tanabe for providing Escherichia coli, and to Mr. Y. Murakami for providing Paramecium multimicronucleatum. We thank to the referees for their useful comments.