Cellular responses of the ciliate, Tetrahymena thermophila, to far infrared irradiation

Abstract

Infrared rays from sunlight permeate the earth's atmosphere, yet little is known about their interactions with living organisms. To learn whether they affect cell structure and function, we tested the ciliated protozoan, Tetrahymena thermophila. These unicellular eukaryotes aggregate in swarms near the surface of freshwater habitats, where direct and diffuse solar radiation impinge upon the water–air interface. We report that populations irradiated in laboratory cultures grew and mated normally, but major changes occurred in cell physiology during the stationary phase. Early on, there were significant reductions in chromatin body size and the antibody reactivity of methyl groups on lysine residues 4 and 9 in histone H3. Later, when cells began to starve, messenger RNAs for key proteins related to chromatin structure, intermediary metabolism and cellular motility increased from two- to nearly nine-fold. Metabolic activity, swimming speed and linearity of motion also increased, and spindle shaped cells with a caudal cilium appeared. Our findings suggest that infrared radiation enhances differentiation towards a dispersal cell-like phenotype in saturated populations of Tetrahymena thermophila.

---

Introduction

The infrared region of the solar spectrum is among the last to be studied biologically. Photons in this region have energies that are too low to excite electrons or disrupt chemical bonds in the compounds that make up cells. Nonetheless, they do interact with water and purified biomolecules in vitro. Low frequency modes of vibration in covalent bonds absorb them, and hydrogen bonds behave similarly.^1–3 Even so, the ways in which these interactions affect the structure and function of living cells are just beginning to be explored.

Reports that certain species of modern day reptiles and insects have thermoreceptor organs to detect infrared radiation were originally published in 1952 and 1964.^4,5 Since then, various types of such receptors have been found in organisms ranging from beetles and butterflies to snakes and bats.^6,7 For instance, jewel beetles have minute thoracic pits that capture the 2–4 µm rays emitted by forest fires. Each pit contains 50–100 dome shaped spherules, sensilla innervated by single dendrites, that expand as they absorb the radiation.⁷ These spherules are highly sensitive mechanoreceptors that enable the flying beetles to home in on fires as far as 60–100 miles away. Upon arrival, they mate and lay eggs in the bark of freshly burnt conifers free from the threat of predators.

Likewise, nocturnal feeding boa constrictors, pit vipers and vampire bats have facial pit organs that guide the strike toward the 8–12 µm infrared rays given off by warm-blooded prey.^6,7 In snakes, radiation of these wavelengths apparently induces structural changes in pit microanatomy that stimulate terminal nerve masses filled with mitochondria and calmodulin.^8,9 While the localization may indicate that ATP and calcium ions are needed for signal transduction, the molecular mechanisms are unknown. Thermal energy in the form of infrared radiation has been ever-present throughout evolutionary history. Cell membranes in tissues of widely divergent species have calcium-permeable ion channels that are sensitive to heat.¹⁰ Common ancestors of these species presumably embodied the genetic capacity to adapt to this environmental selection pressure. Thus, the descendants of ancient organisms alive today such as eukaryotic protists might still respond at the cellular level.

To investigate the effects of infrared radiation on single cells, we chose inbred strains of ciliated protozoan Tetrahymena thermophila for two reasons. First, they constitute genetically uniform populations of ancient eukaryotes with a distinguished history of research dating back to the 1950s. Second, like many eukaryotic microbes in aquatic environments, Tetrahymena swim upwards, a tendency that produces a teeming swarm near the water's surface.¹¹ There, atmospheric gases such as oxygen for aerobic metabolism dissolve, and edible organic debris and bacterial prey accumulate. Incident infrared radiation penetrates water up to about 100 µm, bombarding a superficial layer of cells that spans the upper reaches of the swarm at its interface with air.¹²Tetrahymena species have diversified and adapted to these conditions on the shallow fringes of ponds, lakes and streams worldwide. The consequences for growth and differentiation remain to be analyzed.

Typical of most ciliates, Tetrahymena display nuclear dimorphism and reproduce exclusively by binary fission. When starved, sexually mature cells of different mating types may form temporary pairs that exchange gamete nuclei in a complex process called conjugation.¹³ During these events, a germ line micronucleus gives rise to a genetically active macronucleus by programmed DNA sequence elimination, rearrangement and amplification. As a result, the polygenomic macronucleus contains more than 90% of the total nuclear DNA. During vegetative population growth, 80–90% of this macronuclear DNA condenses into heterochromatin-like chromatin bodies (CBs), and nearly half is expressed as messenger RNA (mRNA).¹⁴ Accordingly, CBs likely comprise specialized domains of potentially active and repressed genes.¹⁵ In cells irradiated as described here, CBs were noticeably smaller than controls during early stationary phase, when cells reached saturation density. The disparity led us to look for changes in chromatin composition, gene activity and physiological function. Here we report on the cellular effects of infrared radiation in T. thermophila.

Experimental

Full details of the materials and methods and tabulated data are available in the ESI.†

Results and discussion

Unaffected growth and fertility

To simulate radiation from the infrared region of the solar spectrum, we used a thermostatically regulated ceramic panel (Fig. 1). The overhead panel emitted a range of wavelengths of low intensity, noncoherent radiation with a broad peak between 8–13 µm (Fig. S1 of the ESI†). The peak corresponded to one of the two major atmospheric windows for infrared rays (3–5 µm and 8–13 µm) and was within the 3–1000 µm range designated as far infrared radiation (FIR) by the International Commission on Illumination (CIE).

Fig. 1 Diagram of the experimental apparatus. Wavelengths of noncoherent FIR emitted from the ceramic panel range from 4–25 µm with a maximum intensity near 10 µm (see Fig. S1 of the ESI†). At a distance of 14 cm from the panel, flasks rest on wooden rods to prevent heat conduction through the surface below. Control flasks also rest on rods in a conventional air convection incubator. Electronic sensors record water temperatures inside and air temperatures outside sentinel flasks (blue) every 30 min for 120 h after 5 ml cultures (amber) are inoculated. During this vegetative cycle, T. thermophila undergo logistic growth in which the rate depends on cell density and the availability of a limited supply of nutrients. Technical details are in Appendices 1.1–1.3.†

Shallow cultures of T. thermophila SB1969 cells, 5 ml in volume and 2 mm deep, were irradiated continuously in FIR translucent flasks undisturbed for five days in the dark (Fig. S2†). Control cultures were warmed in a conventional air convection incubator. Inside irradiated flasks, cells received cumulative doses of radiant energy with an irradiance of 2.80 ± 0.04 W m⁻² (mean ± SEM, n = 50). Inside control flasks, the value was sevenfold lower. Under these conditions, a near optimum growth temperature was maintained between 33 and 34 °C (Tables S1(a) and (b) of the ESI†).

Irradiated cells showed no signs of heat shock or increased cell death that might point to an extreme reaction to FIR. Asynchronous growth did not differ significantly between irradiated and control populations (Fig. S3A–C†). At 24 h in mid-log phase, the doubling time was 3.4 ± 0.1 h (mean ± SEM, n = 80) for both groups (Table S2a†). Genetic crosses showed no significant difference between irradiated and control cells in their capacity for sexual interaction (Tables S2(b) and (c)). Mating occurred normally, and the progeny had a high incidence of drug resistance consistent with undiminished fertility. In total, the conditions of irradiation did not affect vegetative reproduction or sexual conjugation.

Downsized chromatin bodies

During the log phase, irradiated cells were morphologically indistinguishable from variably shaped control cells. In the early stationary phase, 48 h after inoculation, transmission electron microscopy revealed that CBs in their macronuclei were discernibly smaller than controls (Fig. 2). Image analysis confirmed that the cross sectional area was 60% of control at that time (Table S3a†). In units of 100 nm², the mean values ± SEM were 185 ± 4 (n = 1070) versus 311 ± 6 (n = 1023). During the ensuing 24 h, irradiated cell CBs enlarged by nearly 30% and were 74% of controls at 72 h (Tables S3(b) and (c)†). Although the increase was not obvious to the unaided eye, the difference was highly significant (P < 0.0001, ANOVA). Control CBs, by contrast, were equivalent in size at 48 h and 72 h (Table S3(d)†). The disparity in ultrastructure raised the question of whether there might be changes in macronuclear chromatin composition in cells grown under FIR.

Fig. 2 Transmission electron micrographs of T. thermophila SB1969 cells. Irradiated and control cell nuclei are shown in early (48 h) and late (72 h) stationary phase of vegetative growth in laboratory culture. A few globular shaped nucleoli (NO) and hundreds of smaller ovoid patches of electron-dense, condensed chromatin called chromatin bodies (CBs) are scattered throughout the macronucleus (MAC). Fine fibrils, amorphous granular material and dispersed chromatin occupy the intervening space. The micronucleus (MIC) contains genetically inactive condensed chromosomes. At 48 h in FIR treated cells, CBs were discernibly smaller than those in controls (Table S3(a)†). By 72 h, when cells began to starve and die, CBs enlarged relative to 48 h but remained smaller than controls. Significant differences were confirmed by image analysis (Tables S3(b) and (c)†). In 72 h control cells, CB size was essentially unchanged from 48 h (Table S3(d)†). The images are representative examples. Scale bars equal 500 nm.

Unchanged DNA content

Because chromatin is made up of roughly equal amounts of DNA and histone proteins, we initially determined macronuclear DNA content. The amount varied with the phase of population growth in both groups, reaching a maximum between 30 and 35 pg at 48 h (Fig. S3D†). Values for irradiated cells did not differ significantly from controls (Tables S4(a)–(c)†). UV absorbance ratios indicated that the extracts of high molecular weight DNA were substantially free of contaminants. And assessment of DNA integrity by gel electrophoresis revealed no evidence of degradation (Fig. S4†). The results appeared to exclude decreased DNA content as a cause of the smaller CBs in irradiated cells. Partial dispersion of condensed chromatin seemed to be a more likely explanation.

Altered gene expression

How such a change in chromatin structure might relate to gene activity was unclear. Accordingly, we made a preliminary survey of gene expression using differential display reverse transcription polymerase chain reactions (RT-PCR). Differential display with random arbitrary primers revealed an intensely stained band of complementary DNA (cDNA) synthesized from the total RNA of 72 h irradiated cells (Fig. S5A†). The DNA sequences of five clones derived from the band matched that of mRNA for T. thermophila citrate synthase/14 nm filament protein (CS/14FP) in the GenBank database (Fig. S5B†).¹⁶ The observation was independently confirmed by comparing the abundance of CS/14FP mRNA in irradiated cells to controls by real time RT-PCR with specific primers. The estimated index of relative abundance, abbreviated F/C to stand for the ratio of mRNA in FIR treated cells to controls, was 8.9 (Tables 1 and S5c). Clones of four other differentially expressed cDNA bands contained sequences with no exact matches in the database.

Table 1 Summary of mRNAs with deviant F/C indices of relative abundance

mRNA	F/C Index
	24 h	48 h	72 h	Ref.
A preliminary survey of mRNA from key genes related to chromatin, metabolism and motility was made by real time RT-PCR of total RNAs (n = 5). Mean relative abundance of mRNA in FIR treated cells (F) versus controls (C) was estimated as an F/C index (Tables S5(a)–(d)). Indices that fluctuated twofold or more are in bold. Normalized levels of mRNA for these sixteen loci are shown in Fig. 3. Technical details are in Appendices 1.7–1.9 of the ESI.^1
Calmodulin	1.0	0.6	3.4	46
Ciliary dynein outer arm beta heavy chain	1.0	0.5	6.0	22
Citrate synthase/14 nm filament protein	1.0	1.0	8.9	16
Glyceraldehyde-3-phosphate dehydrogenase	1.0	1.0	3.1	21
Heat shock protein 70	2.2	0.6	1.0	18
Heat shock protein 82	1.4	0.5	1.8	18
Heterochromatin-associated protein 1-like protein	2.3	1.0	1.0	14
High mobility group B chromatin protein	1.0	1.0	5.3	20
High mobility group C chromatin protein	0.7	0.5	1.8	20
Linker histone H1	1.0	0.7	6.9	19
Core histones H2A.1 and H2A.2	1.0	0.6	2.5	19
Core histones H2B.1 and H2B.2	1.0	0.7	2.5	19
Core histone H3	1.0	0.4	4.7	19
Core histones H4–I and H4–II	1.0	0.7	2.3	19
Piwi related protein TWI1	0.5	0.5	0.4	31
Telomerase reverse transcriptase	3.6	1.0	0.6	17

---

Identification of CS/14FP came as a surprise, because it encodes the major regulatory enzyme of the citric acid cycle, which dominates aerobic metabolism in diverse taxonomic groups. In fact, citrate synthase is a constituent of nearly all cells in animals, plants, fungi, protists, bacteria and archaea. Tetrahymena CS/14FP is a dual function protein with widely differing activities.¹⁶ Three isoforms are translated from a single mRNA species. Two have citrate synthase catalytic activity, and all are regulated by post-translational phosphorylation. The enzymatic forms predominate throughout vegetative growth. During sexual conjugation, the third isoform polymerizes into 14 nm cytoskeletal filaments involved in pronuclear migration and oral morphogenesis. Much less is known about the regulation of gene expression. More work is needed to characterize CS/14FP transcription and enzyme activity in irradiated cells.

In view of these findings, we narrowed our focus to other highly conserved ‘housekeeping genes’ related to chromatin, metabolism and motility. In T. thermophila, such genes have well characterized genomic sequences, mRNAs, corresponding proteins and physiological functions. Based on this information, we designed primers for real time RT-PCR according to stringent hybridization criteria. Then, using extracts of total cell RNA, we synthesized, amplified and compared cDNAs under identical annealing conditions. Including CS/14FP, twenty-seven out of forty-three candidate mRNAs in the Tetrahymena Genome Database were examined (Tables S5(a)–(c)†). Sixteen had F/C indices that fluctuated twofold or more (Table 1).

In the log phase, twenty-one out of twenty-seven (78%) of the F/C indices equalled 1.0, suggesting that most of these mRNA levels were stable in irradiated cells (Table S5a†). Three were greater than 2.0: those for telomerase reverse transcriptase, heterochromatin-associated protein 1-like protein and heat shock protein 70.^14,17,18 The index for heat shock protein 82, whose synthesis is pre-eminent in the Tetrahymena heat shock response, was just 1.4.¹⁸ At 48 h, fifteen out of twenty-seven (56%) of the F/C ratios were marginally less than 1.0 (Table S5(b)†). By 72 h, eleven (41%) were greater than 2.0 (Table S5(c)†). Besides CS/14FP, these included the indices for all five nucleosomal histone proteins, a high mobility group chromatin protein, a key glycolytic enzyme, and the dynein molecular motor protein of cilia (Table 1).^19–22

Three mRNAs had F/C indices that equalled 1.0 at every point: starvation responsive cysteine protease CyP1, histone acetyl transferase and histone H2A variant H2A.Z.^23–25 The latter, formerly called hv1, is a highly conserved histone variant associated with genetically active chromatin and is essential for cell survival. Its real time RT-PCR threshold cycles (Cts), which are inversely proportional to the initial number of target sequences, were the lowest of the mRNAs surveyed. On those grounds, H2A.Z mRNA levels were predominant and temporally equivalent in irradiated and control cells. Consequently, we normalized the Cts for the twenty-six other mRNAs to the corresponding Ct of H2A.Z at each point. Wilcoxon's nonparametric signed ranks tests then demonstrated no significant difference between irradiated and control group mRNA levels at 24 h (Table S5(d)†). In contrast, irradiated cells had consistently lower levels at 48 h (P < 0.0001) and consistently higher levels at 72 h (P = 0.0003).

To visualize changes in mRNA levels over time, we normalized the Ct data in Tables S5(a)–(c) to that of the corresponding 24 h control for each of the sixteen loci shown in Table 1. Values were plotted in arbitrary units using side-by-side vertical bar graphs (Fig. 3). Several trends emerged. In log phase (24 h), mRNA levels of heat shock protein 70, heterochromatin associated protein 1-like protein and telomerase reverse transcriptase in irradiated cells were two to fourfold higher than the control. At population saturation density (48 h), control levels for eleven of the mRNAs were one to twofold higher than those in irradiated cells. When cells began to starve and die (72 h), control levels for all but one of the mRNAs were at their low point; the lone exception was piwi related protein TWI1. At that time, ten were significantly higher than the control in cells under FIR. Seven of these stood out: calmodulin, ciliary outer arm dynein beta heavy chain, citrate synthase/14 nm filament protein, glyceraldehyde-3-phosphate dehydrogenase, high mobility group B chromatin protein, linker histone H1 and H3 core histones.

Fig. 3 Normalized levels of selected mRNAs from T. thermophila SB1969 cells at 24, 48, and 72 h of culture. Levels of sixteen mRNAs with F/C indices that showed fluctuations of twofold or more (Table 1) are depicted in arbitrary units (A.U.) using side-by-side vertical bar graphs. For each mRNA, the threshold cycle (Ct) values from Tables S5(a)–(c) are normalized to that of the corresponding 24 h control. Vertical bars indicate the level in irradiated cells (black) versus control (white). Numbers are rounded off to significant digits, and error bars are omitted for clarity. Technical details are in Appendix 1.7 of the ESI.†

These major increases in mRNA abundance at 72 h were just as unexpected as the identification of CS/14FP. By then, the supply of nutrients had dwindled, waste products had accumulated, and population size had begun to decline. Control mRNA levels were generally lower than those at earlier times, consistent with down-regulation of gene expression (Fig. 3, Tables S5(a)–(c)). In most cases, levels in irradiated cells also decreased, but to a lesser extent, and CBs were smaller than controls (Fig. 2, Tables S3(a) and (b)). mRNA levels for starvation responsive cysteine protease CyP1 mRNA increased as predicted, and the values were nearly the same in both groups (Table S5(d)).²³ The equivalence supported the idea that there were no differences in starvation between the culture conditions that might preferentially affect gene expression.

Collectively, the results suggest that cells under FIR altered expression of genes encoding key proteins related to chromatin structure, intermediary metabolism and cellular motility. Tetrahymena gene activity adapts to ambient conditions including nutrition and temperature, and differential transcription predominantly determines mRNA abundance.²⁶ Nevertheless, some of the F/C indices we report may represent changes in mRNA stability. Temperature specific expression of at least one cell surface protein takes place this way.²⁷ Post-transcriptional mechanisms may also regulate expression of the genes for histones H1 and H4, whose mRNA stabilities increase in cells undergoing DNA replication.²⁶ The level of mRNA may vary during population growth as well. In rapidly shaken cultures, the concentration of mRNA for glyceraldehyde-3-phosphate dehydrogenase is highest in log phase and decreases gradually thereafter.²¹ Real time RT-PCR measures only steady state levels, so additional information is needed to distinguish these alternatives.

Discordant histone methylation

Covalent modifications of histone proteins associated with DNA in nucleosomes are epigenetic recognition marks that enable or inhibit the binding of regulatory factors to chromatin. In principle, they are thought of as primary determinants of chromatin structure and gene activity. For example, methylation of certain lysine residues may result in activation or repression of transcription.²⁸ To see if CB size and gene expression correlated with histone methylation, we looked at two such lysine residues in the amino-terminal tail domain of histone H3.

On immunoblots of histone enriched extracts, contrasts in chemiluminescent signals corresponding to antibodies against H3 lysine 4 (H3K4) and H3 lysine 9 (H3K9) were highest at 48 h (Fig. 4A–C). Then, CBs were smaller in irradiated cells, F/C indices were mainly less than 1.0, and the major shifts in mRNA abundance had yet to occur. Across the growth cycle, the predominance of signals from H3K4 was in accord with the fact that it is a major methylation site in genetically active macronuclei.²⁹ Signal intensities from dimethyl and trimethyl groups were much higher for H3K4 than for H3K9. The latter required film exposure times that were more than tenfold longer (Fig. 4A). Antibody staining patterns for dimethyl and trimethyl H3K4 were complete after 5 min. However, anti-dimethyl H3K9 showed only the upper band of the calf thymus histone H3 positive control at that time. Pattern completion took 60 min. Likewise, anti-trimethyl H3K9 showed weak signals at 5 min that were maximal after 60 min.

Fig. 4 Immunoblots of histone enriched proteins from irradiated and control T. thermophila SB1969 cells. (A) Chemiluminescent signals from blots probed with polyclonal antibodies to histone H3, dimethylated or trimethylated on lysine 4 (K4) and lysine 9 (K9). Film exposure times in minutes are in parentheses. All four primary antibodies recognized both isoforms of calf thymus histone H3. Decreased histone lysine methylation in FIR treated cells at 48 h appeared to correlate with the smaller CBs in Fig. 2. Protein standards greater than 95 kDa were the only bands visible in the gel stained with Coomassie brilliant blue after electroblotting. Absence of lower molecular weight bands confirmed that the transfer efficiency of histone enriched proteins exceeded 90%. (B) A Ponceau S stain showed that like amounts of protein were transferred to the membrane. (C) The H3 region of a gel processed in parallel and stained with Coomassie brilliant blue showed that equal quantities of H3 protein had been loaded in every lane. Abbreviations: C, control; CBB, Coomassie brilliant blue; CT, calf thymus; F, FIR treated; h, hours; K, lysine; 2me, dimethyl; 3me, trimethyl; PS, Ponceau S. The images are representative of three independent experiments. Technical details are in Appendices 1.10–1.12.†

Absence of detectable dimethyl H3K9 in the 24 h control also agreed with previous reports.^29,30 The presence of trimethyl H3K9 did not, because those studies focused exclusively on the lack of mono- and dimethyl forms. The reason for the absence of dimethyl H3K9 in the 24 h control is a matter of speculation. Mutation of the piwi related gene TWI1 abolishes methylation of H3K9 that is required for elimination of DNA in developing macronuclei during sexual conjugation.³¹ Beyond that, there is little direct evidence to go on. Mono- and dimethyl H3K9 are critical marks for gene silencing in diverse species, but whether they are universal silencing marks is controversial.²⁸ Transcriptionally active macronuclei may lack dimethyl H3K9, because there is scarcely any need for gene silencing in the highly amplified polygenome during vegetative growth. Assumptions about H3K9 methylation that are based on information from other organisms should be made with caution, however. Tetrahymena employs an alternative genetic code and may turn out to be an exception if its site usage for histone lysine methylation differs from that of other life forms.

Strong signals from antibodies to dimethyl and trimethyl H3K4 coincided with the maximum DNA content of macronuclei in 48 h controls. Then, the H3 mRNA level was more than double that of irradiated cells (Fig. 3 and S5b). This, too, may have contributed to the greater reactivity of trimethyl H3K9 relative to 24 and 72 h. By 48 h in irradiated cells, signals from anti-dimethyl H3K4 and anti-dimethyl H3K9 were sharply lower than those at 24 h. Their trimethyl counterparts looked more stable, but there was a clear imbalance in intensity in favour of control cells.

These widely discordant antibody reactivities invite further inquiry. Assuming the antigenic determinants were removed and not masked, mechanisms may include enzymatic demethylation, proteolysis of the H3 amino-terminal tail, and replacement by a nonmethylated histone variant. The first possibility is strengthened by the identification of a long-sought histone demethylase that cleaves the gene activating marks, mono- and dimethyl H3K4.³² The enzyme, referred to as LSD1 for lysine specific demethylase 1, is a nuclear amine oxidase homologue that is widely conserved among eukaryotes. Initially classified as a corepressor, it can also promote gene activation by removing the repressive marks, mono- and dimethyl H3K9.³³ The opposing activities reflect dynamic interactions with different protein partners in complexes recruited to particular loci on chromatin.³⁴ LSD1 has no effect on trimethyl H3K4 or trimethyl H3K9. Its potential role in reducing the antibody reactivity of dimethyl groups in histone H3 from irradiated cells remains to be clarified.

Dimethyl and trimethyl H3K9 are epigenetic marks found mainly in the genetically silenced domains of condensed heterochromatin in a wide range of organisms. The corresponding decreases in antibody reactivity at 48 h appear to correlate with the smaller CBs in irradiated cells. In addition, TWI1 gene expression is required for methylation of H3K9 in developing macronuclei during sexual conjugation.³¹ Whether TWI1 has a role in the assembly and maintenance of heterochromatin-like CBs in vegetative cells is unknown. The low levels of mRNA detected by RT-PCR suggest that the protein is present, however (Tables S5(a)–(c)†). If so, reduced expression of TWI1 also may have contributed to the decrease in CB size. A better understanding of H3K9 methylation in Tetrahymena may elucidate this point.

One possible interpretation of the images is that cells grown under FIR changed the configuration of methyl groups on H3K4 and H3K9 in early stationary phase. The consequences for chromatin structure and gene activity can only be surmised. In Tetrahymena, methylation of H3K4 correlates with active transcription, but the physiological significance is unclear.²⁹ Functional distinctions between mono-, di- and trimethylation have not been drawn. In yeast, dimethyl H3K4 is distributed across the coding region of active genes, while trimethyl H3K4 is restricted to the promoter region.²⁸ Should this pattern also occur in Tetrahymena, decreased antibody reactivity of these groups at 48 h may help explain the F/C indices that were less than unity at that time.

In contrast to H3K4, repressive proteins such as heterochromatin protein 1 (HP1) recognize methyl groups on H3K9. Drosophila HP1 controls elongation and capping of telomeres as well as silencing of nearby genes, and it is also associated with gene expression of heat shock protein 70 (hsp70).^35,36 Hsp70 is an essential part of the chaperone protein complex responsible for folding human telomerase into an active enzyme.³⁷ On immunoblots (Fig. 4A), dimethyl H3K9 was detected in irradiated cells during log phase (24 h). Then, RT-PCR showed increases in mRNA for telomerase reverse transcriptase, heterochromatin-associated protein 1-like protein and hsp70 (Fig. 3, Table S5(a)). The apparent correlation may indicate an interaction between FIR and the telomeric regions of CBs, a prospect for further investigation.

Variations in metabolism, motility and morphology

In light of these observations, we considered their implications for cell physiology. First, to assess metabolic activity, we used a colourimetric test called the MTT assay, which was statistically optimized for Tetrahymena.³⁸ It is based on chemical reduction of the yellow tetrazolium salt, MTT, to water insoluble crystals of the purple azo dye, formazan. Dimethyl sulfoxide dissolves the crystals in situ, and optical absorbance at 550 nm varies directly with the concentration of solubilized formazan. High absorbance implies high metabolic activity. The reduced pyridine nucleotide cofactor, NADH, is responsible for most of the reduction of MTT.³⁹ Current evidence suggests that the reaction is associated with enzymes of the endoplasmic reticulum, endosome/lysosome membranes and plasma membranes as well as mitochondria.

Irradiated cells were more active than controls at every point after zero time except 48 h (Fig. 5). The difference was greatest at 72 h when the absorbance exceeded control by more than 60%. Then, higher metabolic activity fit with the nearly threefold increase in mRNA abundance of glyceraldehyde-3-phosphate dehydrogenase, which catalyses reduction of NAD⁺ to NADH during glycolysis. Indeed, glyceraldehyde-3-phosphate dehydrogenase is a classical metabolic enzyme with newly recognized functions in the nucleus that include regulation of histone gene expression and maintenance of telomere structure.⁴⁰ At 96 h, increases in absorbance for both groups relative to 72 h were presumably due to the disproportionate death of cells with low metabolic activity.

Fig. 5 MTT assay of metabolic activity in T. thermophila SB1969 cells cultured at 33 °C. In this colourimetric test, the reduced pyridine nucleotide cofactor, NADH, is responsible for most of the reduction of MTT. Optical absorbance at 550 nm was calculated after subtraction of nonspecific background. High absorbance implies high metabolic activity. 0 h represents cells from a stationary phase stock culture from which inoculates were made. The data are means ± SEM of three independent experiments (n = 9). Technical details are in Appendix 1.13.†

Next, we analyzed cell velocity and trajectory to compare the metabolic changes with swimming behaviour. Image analysis after video microscopy showed that 72 h-irradiated cells swam 40% faster than controls, 392 ± 6 µm sec⁻¹versus 281 ± 10 µm sec⁻¹ (means ± SEM, n = 100). Statistical variance in irradiated cell velocity was roughly a third of the control, indicating that the magnitudes were more convergent (Table 2). Their trajectories were also much straighter. An index of track linearity was defined as the ratio of the start-to-finish linear distance to the actual swimming path length. Over equivalent path lengths, the index for irradiated cells was 0.94 against 0.77 for control (Tables S6(a) and (b)). The differences in these values were highly significant (P < 0.0001, ANOVA). Swimming behaviour and metabolism appeared to be enhanced in irradiated populations undergoing starvation.

Table 2 Swimming velocity at 72 h of culture in late stationary phase

Group	n	Sum	Mean	Variance
FIR	100	39239	392 µm sec^−1	3738
Control	100	28107	281	10785

---

ANOVA
Variation	SS	df	MS	F	F crit	P-value
In each experiment, cells from three independent cultures of FIR treated SB1969 cells or controls were pooled. Cells were immediately diluted to 10^5 ml^−1 with corresponding 72 h conditioned medium in 50 mm plastic petri dishes. Dishes were placed on a 33 °C ThermoPlate® (MATS-52RA, Tokai Hit) attached to a dark-field microscope (IX70, Olympus). After 20 min, cells were videotaped at 30 frames sec^−1 with a CCD camera (MC681SPD, Texas Instruments). VHS images were dubbed to a DVCAM cassette (PDV-184N, Sony), captured by a frame grabber board (LG-3, Scion Corp.) and stored on the hard disk of a personal computer. Images were calibrated at 1.84 µm pixel^−1 using a 10 µm grid etched on a glass slide. Captured movies were analysed by an image-processing programme for Windows software (Scion Image 4.03). We made a pixel counting macro programme to track cells in motion. Swimming velocity, path length and track linearity index of cells that swam straight were determined. Cells that swam backward or with helical trajectories were excluded from analysis. The data are representative of two independent experiments.
Between	619499	1	619499	85.3	3.89	<0.0001
Within	1437815	198	7262
Total	2057314	199

---

Greater velocity was in line with the relative increase in abundance of mRNA for outer arm dynein beta heavy chain, histone H1 and calmodulin, all of which participate in ciliary motion.^22,41,42 The swimming track linearity index of 0.77 for control cells agreed with a previous report, but the value of 0.94 for irradiated cells did not.⁴³ In that study, extremely low frequency electromagnetic fields (50 Hz) decreased path linearity to 0.68. The reduction was ascribed to direct interference with magnetic field sensitive calcium ion channels in the cell surface membrane. FIR may have affected cells by another signalling mechanism, resulting in fewer changes in direction per unit time that might tend to disperse the population.

Lastly, to find out if the differences in swimming behaviour and metabolism correlated with variations in cell morphology, we rapidly fixed cells for phase-contrast and scanning electron microscopy (Fig. 6 and S6). At 72 h, control cultures of inbred strain SB1969 contained a wide range of teardrop shaped cells that typify vegetative populations in the stationary phase (Fig. 6A, B). In irradiated cultures, spindle shaped cells with a caudal cilium of varying length were present in conspicuous numbers (Fig. 6C). Together with a caudal cilium, some cells appeared to have a barely discernible, slit-like aperture reminiscent of the cryptic oral apparatus that distinguishes the inducible dispersal cell phenotype of Tetrahymena (Fig. 6D).^44,45 Scanning electron microscopy also showed cells with a cleft that might appear to be a slit-like aperture under phase-contrast microscopy (Fig. 6E, F). Cells resembling the diamond shaped ‘compact fusiform’ cells thought to be an intermediate stage of phenotypic transformation were evident as well (Fig. 6G, H).⁴⁴ And what looked like cells undergoing transformation were observed in irradiated cultures of strain SB210, suggesting that the effect was not strain specific (Fig. 6I).

Fig. 6 Morphology of T. thermophila cells at 72 h of culture in late stationary phase. (A) Scanning electron microscopy (SEM) image of a typical vegetative cell. The cell body is teardrop shaped with its anterior end at top. Numerous cilia cover the external surface. The oral apparatus is prominently located left anterior. Scale bar equals 10 µm. (B-D, G and I) Phase-contrast micrographs of wet mounts of cells rapidly fixed in freshly mixed glutaraldehyde/osmium tetroxide. Scale bars equal 25 µm. (B) Inbred strain SB1969 cells exhibited a wide variety of teardrop shapes in a control culture. (C-I) FIR treated cultures. (C) Spindle shaped SB1969 cells, three bearing a caudal cilium that lies in a single focal plane (arrowheads). (D) A spindle shaped SB1969 cell with an inconspicuous slit-like aperture (large arrowhead) and a caudal cilium (small arrowhead). (E) SEM image of a similar cell with a cleft that might appear to be a slit-like aperture under phase-contrast microscopy. Scale bar equals 5 µm. (F) A close up view of the cleft. Scale bar equals 1 µm. (G) A roughly diamond shaped SB1969 cell with an emergent caudal cilium (arrowhead). (H) SEM image of another such diamond shaped cell. Scale bar equals 5 µm. (I) A slender spindle shaped SB210 cell with a caudal cilium (arrowhead). Images in panels B–D, G and I represent inspection of multiple fields from three independent experiments. Fig. S6 is an enlarged version of Fig. 6. Technical details are in Appendix 1.14 (see ESI†).

Further studies and perhaps new methods are needed to determine the rate of change in the number of rapidly swimming cells with a caudal cilium over time. The caudal cilium is a minute, fragile organelle that is difficult to resolve by light microscopy. In these cultures, which are beginning to starve and die at 72 h, cells vary in size, shape and orientation. The caudal cilium may lie in various positions in metachronal waves, some of which cut across multiple focal planes. When it is bent back behind the cell or lying across the cell body, it is indiscernible. Scanning electron microscopy also poses problems; the repeated centrifugation steps needed to process the cells invariably cause breakage and loss of cilia. Inherent difficulties such as these must be overcome before accurate and reliable measurements are possible.

Conclusion

Our findings suggest that Tetrahymena species have evolved heritable adaptations to incident infrared radiation that begin to disperse the population when competition for food reaches a maximum. The shifts in metabolism, motility, morphology and related gene expression in irradiated populations declining under starvation look as if they are compatible with this view. The decreases in CB size and histone methylation at cell saturation density further suggest that changes in chromatin structure precede those in physiological function.

Finally, there is currently no evidence to indicate T. thermophila improve their odds of survival by decreasing metabolism and motility to form resting cysts when food is depleted. On the other hand, fast swimming dispersal phenotypes of both T. thermophila and T. pyriformis can be induced by depriving cells of nutrients under laboratory conditions.^44,45 But extreme starvation in dilute salt solution at low cell density is required for cells to complete differentiation into these slender, elongated forms. They are not known to occur in cultures made with enriched medium. Thus, the measurements we report may represent average properties of cells in intermediate stages of differentiation. Future research will concentrate on the molecular basis of this phenomenon, the possibility that it involves calcium signalling, and its implications for other eukaryotes.

Acknowledgements

We thank E. Orias for the inbred strains of T. thermophila and H. Asai for constructive comments. We also thank M. Miyasita for expert assistance with scanning electron microscopy. Support for this investigation was provided in part by funds from Project B, Advanced Research Center for Science and Engineering, Waseda University.

References

1. M. Walther, P. Plochocka, B. Fischer, H. Helm and P. Uhd Jepsen, Collective vibrational modes in biological molecules investigated by terahertz time-domain spectroscopy, Biopolymers, 2002, 67, 310–313.
2. S. W. Smye, J. M. Chamberlain, A. J. Fitzgerald and E. Berry, The interaction between terahertz radiation and biological tissues, Phys. Med. Biol., 2001, 46, R101–R112.
3. Y. Maede, M. Ide and H. Kitano, Vibrational spectroscopic study on the structure of water in polymer systems, J. Mol. Liq., 1999, 80, 149–163.
4. T. H. Bullock and R. B. Cowles, Physiology of an infrared receptor—the facial pit of pit vipers, Science, 1952, 115, 541–543.
5. W. G. Evans, Infra-red receptors in Melanophila acuminata DeGeer, Nature, 1964, 202, 211.
6. A. L. Campbell, R. R. Naik, L. Sowards and M. O. Stone, Biological infrared imaging and sensing, Micron, 2002, 33, 211–225.
7. H. Bleckmann, H. Schmitz and G. von der Emde, Nature as a model for technical sensors, J. Comp. Physiol. A, 2004, 190, 971–981.
8. M. S. Grace, D. R. Church, C. T. Kelly, W. F. Lynn and T. M. Cooper, The Python pit organ: imaging and analysis of an extremely sensitive natural infrared detector, Biosens. Bioelectron., 1999, 14, 53–59.
9. V. Gorbunov, N. Fuchigami, M. Stone, M. Grace and V. V. Tsukruk, Biological thermal detection: Micromechanical and microthermal properties of biological infrared receptors, Biomacromolecules, 2002, 3106, 115.
10. I. S. Ramsey, M. Delling and D. E. Clapham, An introduction to trp channels, Annu. Rev. Physiol., 2006, 68, 619–647.
11. H. Winet and T. L. Jahn, Geotaxis in protozoa I. A propulsion-gravity model for Tetrahymena (Ciliata), J. Theor. Biol., 1974, 46, 449–46.
12. K. Hirota, A. Hashimoto, T. Honda, M. Shimizu and A. Watanabe, Constant drying of wet granular beds by far infrared radiation, Nippon Shokuhin Kogyo Gakkaishi, 1991, 38, 136–143.
13. K. M. Karrer, Tetrahymena genetics: Two nuclei are better than one, Methods Cell Biol., 2000, 62, 127–186.
14. H. Huang, E. A. Wiley, C. R. Lending and C. D. Allis, An HP1-like protein is missing from transcriptionally silent micronuclei of Tetrahymena, Proc. Natl. Acad. Sci. U. S. A., 1998, 95, 13624–13629.
15. H. Huang, J. F. Smothers, E. A. Wiley and C. D. Allis, A nonessential HP1-like protein affects starvation-induced assembly of condensed chromatin and gene expression in macronuclei of Tetrahymena thermophila, Mol. Cell Biol., 1999, 19, 3624–3634.
16. H. Kojima and O. Numata, Enzymatic form and cytoskeletal form of bifunctional Tetrahymena 49 kDa protein is regulated by phosphorylation, Zoolog. Sci., 2002, 19, 37–42.
17. T. M. Bryan, J. M. Sperger, K. B. Chapman and T. R. Cech, Telomerase reverse transcriptase genes identified in Tetrahymena thermophila and Oxytricha trifallax, Proc. Natl. Acad. Sci. U. S. A., 1998, 95, 8479–8484.
18. N. E. Williams and E. M. Nelsen, HSP70 and HSP90 homologs are associated with tubulin in hetero-oligomeric complexes, cilia and the cortex of Tetrahymena, J. Cell Sci., 1997, 110, 1665–1672.
19. X. Liu and M. A. Gorovsky, Cloning and characterization of the major histone H2A genes completes the cloning and sequencing of known histone genes of Tetrahymena thermophila, Nucleic Acids Res., 1996, 24, 3023–3030.
20. I. G. Schulman, T. Wang, M. Wu, J. Bowen, R. G. Cook, M. A. Gorovsky and C. D. Allis, Macronuclei and micronuclei in Tetrahymena thermophila contain high-mobility-group-like chromosomal proteins containing a highly conserved eleven-amino-acid putative DNA-binding sequence, Mol. Cell Biol., 1991, 11, 166–174.
21. Y. Zhao, S. Nakashima, M. Andoh and Y. Nozawa, Cloning and sequencing of a cDNA encoding glyceraldehyde-3-phosphate dehydrogenase from Tetrahymena thermophila: Growth-associated changes in its mRNA expression, J. Euk. Microbiol., 1997, 44, 434–437.
22. S. Toba, T. M. Gibson, K. Shiroguchi, Y. Y. Toyoshima and D. J. Asai, Properties of the full-length heavy chains of Tetrahymena ciliary outer arm dynein separated by urea treatment, Cell Motil. Cytoskeleton, 2004, 58, 30–38.
23. Y. Dou and M. A. Gorovsky, Regulation of transcription by H1 phosphorylation in Tetrahymena is position independent and requires clustered sites, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 6142–6146.
24. E. A. Wiley, C. A. Mizzen and C. D. Allis, Isolation and characterization of in vivo modified histones and an activity gel assay for identification of histone acetyltransferases, Methods Cell Biol., 2000, 62, 379–394.
25. Q. Ren and M. A. Gorovsky, The nonessential H2A N-terminal tail can function as an essential charge patch on the H2A.Z variant N-terminal tail, Mol. Cell Biol., 2003, 23, 2778–2789.
26. L. A. Stargell, K. M. Karrer and M. A. Gorovsky, Transcriptional regulation of gene expression in Tetrahymena thermophila, Nucleic Acids Res., 1990, 18, 6637–6639.
27. H. D. Love, Jr., A. Allen-Nash, Q. A. Zhao and G. A. Bannon, mRNA stability plays a major role in regulating the temperature-specific expression of a Tetrahymena thermophila surface protein, Mol. Cell Biol., 1988, 8, 427–432.
28. C. Martin and Y. Zhang, The diverse functions of histone methylation, Nat. Rev. Mol. Cell Biol., 2005, 6, 838–849.
29. B. D. Strahl, R. Ohba, R. G. Cook and C. D. Allis, Methylation of histone H3 at lysine 4 is highly conserved and correlates with transcriptionally active nuclei in Tetrahymena, Proc. Natl. Acad. Sci. U. S. A., 1999, 96, 14967–14972.
30. S. D. Briggs, M. Bryk, B. D. Strahl, W. L. Cheung, J. K. Davie, S. Y. R. Dent, F. Winston and C. D. Allis, Histone H3 lysine 4 methylation is mediated by Set1 and required for cell growth and rDNA silencing in Saccharomyces cerevisiae, Genes Dev., 2001, 15, 3286–3295.
31. Y. Liu, K. Mochizuki and M. A. Gorovsky, Histone H3 lysine 9 methylation is required for DNA elimination in developing macronuclei in Tetrahymena, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 1679–1684.
32. Y. Shi, F. Lan, C. Matson, P. Mulligan, J. R. Whetstine, P. A. Cole, R. A. Casero and Y. Shi, Histone demethylation mediated by the nuclear amine oxidase homolog LSD1, Cell, 2004, 119, 941–953.
33. E. Metzger, M. Wissmann, N. Yin, J. M. Muller, R. Schneider, A. H. F. M. Peters, T. Gunther, R. Buettner and R. Schule, LSD1 demethylates repressive histone marks to promote androgen-receptor-dependent transcription, Nature, 2005, 437, 436–439.
34. Y.-J. Shi, C. Matson, F. Lan, S. Iwase, T. Baba and Y. Shi, Regulation of LSD1 histone demethylase activity by its associated factors, Mol. Cell, 2005, 19, 857–864.
35. B. Perrini, L. Piacentini, L. Fanti, F. Altieri, S. Chichiarelli, M. Berloco, C. Turano, A. Ferraro and S. Pimpinelli, HP1 controls telomere capping, telomere elongation, and telomere silencing by two different mechanisms in Drosophila, Mol. Cell, 2004, 15, 467–476.
36. L. Piacentini, L. Fanti, M. Berloco, B. Perrini and S. Pimpinelli, Heterochromatin protein 1 (HP1) is associated with induced gene expression in Drosophila euchromatin, J. Cell Biol., 2003, 161, 707–714.
37. H. L. Forsythe, J. L. Jarvis, J. W. Turner, L. W. Elmore and S. E. Holt, Stable association of hsp90 and p23, but not hsp70 with active human telomerase, J. Biol. Chem., 2001, 276, 15571–15574.
38. N. Dias, A. Nicolau, G. S. Carvalho, M. Mota and N. Lima, Miniaturization and application of the MTT assay to evaluate metabolic activity of protozoa in the presence of toxicants, J. Basic Microbiol., 1999, 39, 103–108.
39. M. V. Berridge, P. M. Herst and A. S. Tan, Tetrazolium dyes as tools in cell biology: New insights into their cellular reduction, Biotechnol. Ann. Rev., 2005, 11, 127–152.
40. M. A. Sirover, New nuclear functions of the glycolytic protein, glyceraldehyde-3-phosphate dehydrogenase, in mammalian cells, J. Cell. Biochem., 2005, 95, 45–52.
41. L. Multigner, J. Gagnon, A. Van Dorsselaer and D. Job, Stabilization of sea urchin microtubules by histone H1, Nature, 1992, 360, 33–39.
42. J. J. Blum, A. Hayes, G. A. Jamieson, Jr. and T. C. Vanaman, Calmodulin confers calcium sensitivity on ciliary dynein ATPase, J. Cell Biol., 1980, 87, 386–397.
43. R. Hemmersbach, E. Becker and W. Stockem, Influence of low frequency electromagnetic fields on the swimming behavior of ciliates, Bioelectromagnetics, 1997, 18, 491–498.
44. E. M. Nelsen, Transformation in Tetrahymena thermophila: Development of an inducible phenotype, Dev. Biol., 1978, 66, 17–31.
45. E. M. Nelsen and L. E. Debault, Transformation in Tetrahymena pyriformis: Description of an inducible phenotype, J. Protozool., 1978, 25, 113–119.
46. T. Takemasa, T. Takagi, M. Edanatsu and Y. Watanabe, Calmodulin cDNAs from two species of Tetrahymena, Biochim. Biophys. Acta, 1992, 1132, 219–221.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed materials and methods; statistical analyses of temperature, population doubling time, sexual conjugation, chromatin body size, macronuclear DNA content, real time RT-PCR, path length and track linearity (Tables S1–S6); far infrared emission spectrum of the aluminium oxide/titanium oxide ceramic panel (Fig. S1); transmittance spectrophotometry of polystyrene composing the FIR translucent culture flask (Fig. S2); vegetative population growth (A–C) and macronuclear DNA content (D) of T. thermophila SB1969 cells left undisturbed (Fig. S3); assessment of SB1969 cell DNA integrity by agarose gel electrophoresis (Fig S4); differential display reverse transcription polymerase chain reactions (Fig. S5); and morphology of T. thermophila cells at 72 h of culture in late stationary phase (Fig. S6). See DOI: 10.1039/b601741j

---