Mannosylglycerate and Di-myo-Inositol Phosphate Have Interchangeable Roles during Adaptation of Pyrococcus furiosus to Heat Stress

Abstract

Marine hyperthermophiles accumulate small organic compounds, known as compatible solutes, in response to supraoptimal temperatures or salinities. Pyrococcus furiosus is a hyperthermophilic archaeon that grows optimally at temperatures near 100°C. This organism accumulates mannosylglycerate (MG) and di-myo-inositol phosphate (DIP) in response to osmotic and heat stress, respectively. It has been assumed that MG and DIP are involved in cell protection; however, firm evidence for the roles of these solutes in stress adaptation is still missing, largely due to the lack of genetic tools to produce suitable mutants of hyperthermophiles. Recently, such tools were developed for P. furiosus, making this organism an ideal target for that purpose. In this work, genes coding for the synthases in the biosynthetic pathways of MG and DIP were deleted by double-crossover homologous recombination. The growth profiles and solute patterns of the two mutants and the parent strain were investigated under optimal growth conditions and also at supraoptimal temperatures and NaCl concentrations. DIP was a suitable replacement for MG during heat stress, but substitution of MG for DIP and aspartate led to less efficient growth under conditions of osmotic stress. The results suggest that the cascade of molecular events leading to MG synthesis is tuned for osmotic adjustment, while the machinery for induction of DIP synthesis responds to either stress agent. MG protects cells against heat as effectively as DIP, despite the finding that the amount of DIP consistently increases in response to heat stress in the nine (hyper)thermophiles examined thus far.

INTRODUCTION

Many thermophiles and hyperthermophiles have been isolated from marine geothermal areas and, accordingly, grow optimally in medium containing 2 to 4% (wt/vol) NaCl. Like the vast majority of halophiles, these organisms accumulate compatible solutes to balance the external osmotic pressure (1–3). However, the organisms from hot habitats accumulate unusual negatively charged solutes (sometimes designated thermolytes), which contrast with the neutral or zwitterionic nature of the solutes typical of mesophiles. Moreover, the intracellular content of solutes from (hyper)thermophiles increases not only with the NaCl concentration of the medium but also with the growth temperature, suggesting that the role of thermolytes goes beyond osmoprotection (4).

Screening for new solutes in (hyper)thermophiles showed that di-myo-inositol phosphate (DIP) and mannosylglycerate (MG) are the most widespread components of the solute pools in organisms adapted to grow at temperatures above 60°C (4). Typically, heat stress conditions lead to the accumulation of DIP and DIP derivatives, while MG increases in response to a supraoptimal salinity in the growth medium; curiously, the thermophilic bacterium Rhodothermus marinus and the hyperthermophilic archaeon Palaeococcus ferrophilus, which lack DIP biosynthesis, accumulate MG to cope with heat stress (2, 5). On the other hand, Archaeoglobus fulgidus strains unable to synthesize MG use diglycerol phosphate as an alternative solute for osmoadjustment (6).

DIP and MG are the two major solutes in Pyrococcus furiosus DSM 3638^T, an extensively studied hyperthermophilic archaeon able to grow at temperatures up to 103°C and regarded as a model organism for investigating the molecular basis of adaptation to high temperatures (7–9). Upon osmotic or heat stress, the total solute pool increases in concentration by approximately 10-fold. MG represented 70% of the total pool under salt stress, while DIP was the only solute accumulating at supraoptimal temperatures. Minor amounts of glutamate were used only for adjustment to low salinity (8). Therefore, the profile of the stress response in P. furiosus is representative of the general trend, insofar as MG and DIP are preferentially associated with osmoprotection and thermoprotection, respectively (2, 10). However, definite proof of their physiological roles is still lacking.

The present work was designed to address the following questions: do DIP and MG play a role in cell protection against stress? If so, do they have specialized activities in osmoadaptation and thermoadaptation? To answer these questions, mutants deficient in the synthesis of specific compatible solutes are required. Fortunately, the development of tools for the genetic manipulation of hyperthermophiles has advanced significantly in recent years (9, 11, 12). In particular, genetic tools to manipulate P. furiosus have been developed (13). The genetic system is based on a variant of P. furiosus DSM 3638^T designated COM1, which is highly competent for external DNA uptake and recombination. The genome sequence of strain COM1 showed extensive changes; however, the major metabolic features of the wild-type strain appear to be conserved (14).

P. furiosus is an ideal target organism for investigating the physiological role of MG and DIP because these are the major components of its solute pool, they are produced by de novo synthesis, and, importantly, the organism does not accumulate amino acids or other biomolecules from the external medium to cope with osmotic and heat stress (8). The preferential import of alternative solutes could complicate the interpretation of results, as reported for a DIP-deficient mutant of Thermococcus kodakarensis (15). The biosynthetic pathways of DIP and MG are depicted in Fig. 1.

FIG 1.

Biosynthetic pathways for di-myo-inositol phosphate (top) and mannosylglycerate (bottom). Abbreviations for enzymes: IPCT, CTP:l-myo-inositol-1-phosphate cytidylyltransferase; DIPPS, di-myo-inositol phosphate phosphate synthase (the gene encoding the phosphatase activity is unknown); MPGS, mannosyl-3-phosphoglycerate synthase; MPGP, mannosyl-3-phosphoglycerate phosphatase. Other abbreviations: DIPP, di-myo-inositol phosphate phosphate; 3-PGA, 3-phosphoglycerate; MPG, mannosyl-3-phosphoglycerate; GDP-man, GDP-mannose.

In this study, we deleted the genes encoding the key enzymes in the biosynthesis of MG and DIP in P. furiosus COM1 and studied the impact of these disruptions on growth as well as on the pattern of solute accumulation as a means to obtain insight into the physiological role of those ionic solutes.

MATERIALS AND METHODS

Strains and culture conditions.

P. furiosus strain COM1, a uracil auxotrophic strain, was used as the host strain for genetic manipulation (16). This organism was transformed by the method of Lipscomb et al. (13).

The physiologic studies of P. furiosus were performed with cells cultivated in the complex medium described previously (17), with the following modifications: 0.5% (wt/vol) maltose and 0.25% (wt/vol) yeast extract were used as carbon sources, and growth was performed without elemental sulfur. Cultures were grown in a 2-liter fermentation vessel with continuous gassing with a mixture of N₂ (80%) and CO₂ (20%) and stirring at 80 rpm. Cultures used as the preinoculum were successively passed in fresh medium at least 3 times before inoculation. Prior to inoculation (10% of an overnight culture), the medium was supplemented with cysteine (0.5 g/liter) and Na₂S (0.5 g/liter) and the pH was adjusted to 6.8. During growth, the pH was maintained at 6.8 by the addition of 10% (wt/vol) NaHCO₃. The optical density (OD) at 600 nm (OD₆₀₀) was used to assess cell growth. The temperature of the growth medium was monitored with a thermocouple probe (Fluke 80PK-22; Fluke Corporation, WA). The accuracy of the measurements over the temperature range of 85 to 100°C was better than ±0.5°C.

To study the effect of the NaCl concentration on the growth of P. furiosus COM1 (the parent strain), cells were grown at 90°C in medium with different NaCl concentrations (1.0%, 2.8%, 3.6%, 4.5%, and 5.5%, wt/vol, corresponding to 0.171, 0.479, 0.616, 0.770, and 0.941 M, respectively); the effect of temperature was investigated at different temperatures (88°C, 90°C, 93°C, 95°C, 98°C, and 100°C) in medium with the optimum NaCl concentration (2.8% NaCl). To determine the effect of osmotic and heat stress on the growth parameters and solute accumulation, cells were grown under optimal conditions (2.8% NaCl, 90°C), conditions of osmotic stress (4.5% NaCl, 90°C), and conditions of heat stress (2.8% NaCl, 98°C).

Construction of linear fragments for gene deletions.

The linear fragment aeΔMPGS was constructed to delete the gene encoding the mannosyl-3-phosphoglycerate synthase (MPGS; pfc_02085) (see Fig. S1 in the supplemental material). The upstream and downstream flanking regions (1.5 kbp) of the MPGS gene were amplified from the genomic DNA. The P_gdhpyrF marker cassette was also amplified from plasmid pGLW021. The 3′ primer of the upstream fragment (Δmpgs_prev_UFR) and the 5′ primer of the downstream fragment (Δmpgs_pfow_DFR) have 20 to 30 bases homologous to the 5′ and 3′ ends of the marker cassette, respectively. The three PCR fragments were joined by PCR using the 5′ primer of the upstream fragment (Δmpgs_pfow_UFR) and 3′ primer of the downstream fragment (Δmpgs_prev_DFR). This strategy was also used to construct a second linear fragment, the aeΔIPCT/DIPPS gene fragment, to obtain a mutant deficient in DIP synthesis by deleting the gene encoding the bifunctional enzyme CTP:l-myo-inositol-1-phosphate cytidylyltransferase (IPCT)/di-myo-inositol phosphate phosphate synthase (DIPPS) (pfc_04525). The primer sequences used in this work are presented in Table S1 in the supplemental material.

Transformation of strains.

P. furiosus COM1 was transformed with the aeΔMPGS or aeΔIPCT/DIPPS gene by natural competence to obtain the MG-deficient mutant or the DIP-deficient mutant, respectively. Selection was performed on solid defined medium without uracil. Genomic DNA isolation was performed as previously described (13). Briefly, cells from 1 ml of an overnight P. furiosus culture were harvested and suspended in 100 μl of buffer A (25% sucrose, 50 mM Tris-HCl, 40 mM EDTA, pH 7.4), followed by the addition of 250 μl of 6 M guanidinium HCl–20 mM Tris, pH 8.5. The mixture was incubated at 70°C for 5 min. Genomic DNA was extracted with 350 μl of phenol-chloroform-isoamyl alcohol (25:24:1; buffered at pH 8), followed by ethanol precipitation, and suspended in 50 μl of 10 mM Tris buffer, pH 8.0. PCR analyses confirmed that the MG- and DIP-deficient mutants lack the MPGS and IPCT/DIPPS genes, respectively. The positive mutants of both transformations were further purified by successive cultivation on solid medium without uracil. Additionally, each flank of the recombination region (1.7-kbp) was amplified and sequenced for each mutant. Sequence analysis revealed that the MPGS and IPCT/DIPPS genes were correctly deleted and replaced by the pyrF cassette.

Southern blot analysis.

Chromosomal DNA was extracted by the method previously described (18). Chromosomal DNAs of the MG- and DIP-deficient mutants were digested with XmnI/HindIII/BglII and HindIII/XmnI/SacI, respectively. The chromosomal DNA of P. furiosus COM1 was also digested with both sets of restriction enzymes. The resulting DNA fragments were separated onto a 1% agarose gel by electrophoresis and transferred to a nylon membrane. The probe (500 bp) against a region close to the 3′ end of the upstream region of the target gene was amplified by PCR and labeled following the protocol supplied by the manufacturer (ECL Direct nucleic labeling and detection system; GE Healthcare). Probe hybridization was carried out as previously described (19). The signals were detected with an enhanced chemiluminescence (ECL) system (GE Healthcare).

Extraction, identification, and determination of intracellular solutes.

Cells were harvested at the end of exponential phase by centrifugation (6,370 × g, 15 min) and washed twice with a solution containing all medium components except the carbon sources. Cell pellets were resuspended in water and disrupted by sonication. The total protein concentration was estimated using a Pierce bicinchoninic acid protein assay kit (Thermo). Intracellular solutes were extracted with boiling 80% ethanol, as described previously (20). Solvent was removed by rotary evaporation, and the residue was freeze-dried. The dry residue was extracted twice with a mixture of water-chloroform (2:1). After centrifugation, the aqueous phase was lyophilized and the residue was dissolved in ²H₂O for ¹H nuclear magnetic resonance (NMR) analysis. Formate was used as the concentration standard (15).

RT-PCR experiments.

The parent and DIP-deficient strains were grown under heat stress conditions (98°C, 2.8% NaCl) as described above. Cells were harvested at the end of exponential phase, and total RNA was extracted using an RNeasy minikit (Qiagen, Hilden, Germany). RNA samples were treated with Turbo DNase (Life Technologies, Grand Island, NY) to remove residual chromosomal DNA. To confirm the absence of chromosomal DNA in the RNA preparations, RNA samples were used as DNA templates for PCR using primers specific for the pfc_04085 gene, which encodes pyruvate ferredoxin oxidoreductase (POR) (see Table S1 in the supplemental material). Genomic DNA from P. furiosus COM1 was used as a positive control. Reverse transcription-PCR (RT-PCR) experiments were carried out using a OneStep RT-PCR kit (Qiagen, Hilden, Germany). Briefly, aliquots of total RNA (60 ng) were mixed with 400 μM each deoxynucleoside triphosphate, 0.6 μM specific primers, 1× buffer containing 2.5 mM Mg²⁺, and 1 μl of Qiagen OneStep RT-PCR enzyme mix for a reaction volume of 25 μl. Reverse transcription occurred at 50°C for 30 min, followed by a PCR activation step at 95°C for 15 min. PCR conditions consisted of 30 repetitive cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 1 min, followed by a final extension at 72°C for 10 min. The primers used to amplify the internal fragments of genes pfc_09170 (coding for Hsp20), pfc_08710 (coding for Hsp60), pfc_07215 (coding for HtpX), pfc_04085 (coding for POR), and pfc_09175 (coding for AAA⁺ ATPase) are shown in Table S1 in the supplemental material. The gene encoding POR was used as a control. RT-PCR products were separated by electrophoresis and visualized with ethidium bromide. Semiquantitative analysis was performed by using Quantity One software (Bio-Rad, Hercules, CA). Results are expressed as a ratio of the amount of the target PCR product to the amount of the POR product. These experiments were performed in duplicate.

RESULTS

Effect of temperature and NaCl concentration on growth and solute accumulation by the parent strain.

The growth profiles of P. furiosus COM1 at temperatures between 88 and 100°C were examined in medium containing 2.8% (wt/vol) NaCl. In accordance with the properties of the original strain of P. furiosus (7), the growth rate did not change appreciably over the temperature range of 88 to 98°C (Table 1 and results not shown). However, the maximum cell density was observed at about 90°C, and growth within 8 h was not observed at 100°C (Table 1; Fig. 2A). At the temperature for maximum growth, 90°C, strain COM1 grew in medium containing up to about 5.5% NaCl, with optimal growth occurring at approximately 2.8% NaCl (Table 1; Fig. 2B), a behavior typical of slightly halophilic organisms. Unlike for the temperature dependence, the cell density and the growth rate displayed parallel trends when the NaCl concentration was varied (Fig. 2B and results not shown).

TABLE 1.

Growth rate and maximal cell density of P. furiosus COM1 (parent strain) and mutants under optimal growth conditions, heat stress, and osmotic stress

P. furiosus strain	Growth rate (h−1)a
	90°C, 2.8% NaCl	98°C, 2.8% NaCl	90°C, 4.5% NaCl
Parent	0.65 ± 0.05 (0.88 ± 0.13)	0.70 ± 0.08 (0.32 ± 0.06)	0.39 ± 0.03 (0.49 ± 0.08)
MG deficient	0.56 ± 0.06 (0.71 ± 0.05)	0.62 ± 0.08 (0.29 ± 0.03)	0.25 ± 0.04 (0.39 ± 0.04)
DIP deficient	0.73 ± 0.06 (1.42 ± 0.06)	0.77 ± 0.03 (0.45 ± 0.07)	0.43 ± 0.03 (0.63 ± 0.15)

^aThe values are averages from three to six independent experiments. The optimal temperature for growth was 90°C; the optimal NaCl concentration was 2.8% (wt/vol). Values in parentheses are the maximal cell densities (OD₆₀₀).

FIG 2.

Maximal cell density (OD₆₀₀) for P. furiosus COM1 as a function of the growth temperature (A) and the NaCl concentration in the growth medium (B). The following NaCl concentrations were examined: 1.0, 2.8, 3.6, 4.5, and 5.5% (wt/vol), corresponding to 0.171, 0.479, 0.616, 0.770, and 0.941 M, respectively. The bars represent standard deviations obtained from three to six independent experiments. In the other cases, a single experiment was performed.

In view of the observed growth profiles, 90°C and 2.8% NaCl were selected as the optimal temperature and the optimal salt condition for growth of the COM1 strain, respectively. Heat stress was imposed by increasing the growth temperature to 98°C in medium containing 2.8% NaCl, while the effect of salt stress was studied by increasing the NaCl concentration to 4.5% in cultures grown at the optimal growth temperature. Typical growth curves obtained under optimal and stress conditions are shown in Fig. 3A. An increase of 8°C above the optimal temperature resulted in a clear impairment of the final cell density (the OD₆₀₀ varied from 0.88 to 0.32), while it did not significantly affect the growth rate (Table 1). The lack of an effect of temperature on the growth rate over the temperature range of 75 to 90°C was also reported for Thermococcus kodakarensis by Kanai et al. (21). On the other hand, in medium containing 4.5% NaCl, both the growth rate and the final OD were reduced by about 40% compared with those found under optimal growth conditions.

FIG 3.

Growth curves of P. furiosus COM1 (A), the MG-deficient strain (B), and the DIP-deficient strain (C) under optimal growth conditions (90°C, 2.8% NaCl) (diamonds), under heat stress (98°C, 2.8% NaCl) (circles), and under osmotic stress (90°C, 4.5% NaCl) (triangles). The growth rates were determined and are shown in Table 1. For each strain and condition, one representative curve of a total of three to six is shown.

In parallel, the accumulation of compatible solutes was investigated in cells collected at the end of the exponential phase (Table 2). MG was the predominant solute accumulated by strain COM1 under all growth conditions examined, corresponding to about 65%, 47%, and 78% of the total organic solute pool under optimal, heat stress, and salt stress conditions, respectively. Under optimal growth conditions, DIP was detected in vestigial amounts, but its concentration increased 7-fold between the optimal temperature and heat stress conditions, constituting 20% of the total solute pool at 98°C; in contrast, the concentration of MG decreased by 24% in response to the heat stress condition. Alanine, lactate, and aspartate were also detected in the ethanol extract, but in smaller amounts, and their concentrations did not change significantly with either of the stresses imposed (Table 2; Fig. 4A). Accumulation of lactate is intriguing, since the genome of P. furiosus lacks genes encoding a lactate dehydrogenase homolog (22). We postulate that lactate was taken up from the medium. NMR analysis confirmed the presence of lactate in yeast extract (17.9 mg of lactate per gram), leading to a final concentration of 0.5 mM lactate in the culture medium. The concentrations of intracellular lactate were highly variable among experimental replicates, and hence, the respective standard deviations are much greater than those estimated for the other solutes.

TABLE 2.

Quantification of organic solutes in P. furiosus COM1 (parent strain) and the MG-deficient and DIP-deficient mutants under optimal growth conditions, heat stress, and osmotic stress

Strain	Growth temp (°C)	NaCl concn (%, wt/vol)	Solute concn (μmol/mg of protein)a
			MG	DIP	Ala	Lac	Asp	InoP	Total
Parent	90	2.8	0.42 ± 0.03	0.02 ± 0.01	0.08 ± 0.01	0.06 ± 0.01	0.07 ± 0.01	ND	0.65 ± 0.04
	98	2.8	0.32 ± 0.06	0.14 ± 0.01	0.03 ± 0.02	0.15 ± 0.13	0.04 ± 0.01	ND	0.68 ± 0.14
	90	4.5	1.12 ± 0.00	0.07 ± 0.00	0.09 ± 0.04	0.05 ± 0.00	0.10 ± 0.04	ND	1.43 ± 0.06
MG deficient	90	2.8	ND	0.12 ± 0.03	0.17 ± 0.02	0.14 ± 0.02	0.10 ± 0.01	ND	0.53 ± 0.04
	98	2.8	ND	0.37 ± 0.05	0.05 ± 0.03	0.27 ± 0.10	ND	ND	0.68 ± 0.12
	90	4.5	ND	0.31 ± 0.05	0.21 ± 0.02	0.13 ± 0.04	0.42 ± 0.07	ND	1.08 ± 0.10
DIP deficient	90	2.8	0.79 ± 0.19	ND	0.19 ± 0.02	0.02 ± 0.01	0.05 ± 0.01	ND	1.06 ± 0.19
	98	2.8	0.47 ± 0.06	ND	0.06 ± 0.02	0.05 ± 0.03	0.03 ± 0.01	0.09 ± 0.03	0.70 ± 0.07
	90	4.5	1.19 ± 0.11	ND	0.19 ± 0.00	0.02 ± 0.00	0.12 ± 0.00	0.06 ± 0.01	1.58 ± 0.11

^aValues are the means of two to three independent experiments. MG, mannosylglycerate; DIP, di-myo-inositol phosphate; Ala, alanine; Lac, lactate; Asp, aspartate; InoP, l-myo-inositol-1-phosphate; ND, not detected.

FIG 4.

Composition of organic solutes in P. furiosus COM1 (parent strain) (A), the MG-deficient mutant (B), and the DIP-deficient mutant (C) under optimal growth conditions (90°C, 2.8% NaCl), heat stress (98°C, 2.8% NaCl), and osmotic stress (90°C, 4.5% NaCl). MG, mannosylglycerate; DIP, di-myo-inositol phosphate; Ala, alanine; Lac, lactate; Asp, aspartate; InoP, l-myo-inositol-1-phosphate. The values are the means of two to three independent experiments. Standard deviations are presented in Table 2.

The total solute pools at the optimal growth temperature and under heat stress conditions were identical (0.65 and 0.68 μmol/mg of protein, respectively), but in response to elevated salinity, the total pool of solutes increased approximately 2.2-fold. This increment is attributed largely to higher levels of MG, which increased 2.7-fold. The level of DIP also increased (3.5-fold), but the final concentration was much lower than that of MG, contributing only 5% to the total pool of solutes.

Construction of mutants deficient in MG or DIP synthesis.

The contribution of compatible solutes during osmo- and thermoadaptation was assessed by constructing mutants of P. furiosus COM1 unable to synthesize MG or DIP. To this end, the genes encoding the synthases involved in DIP and MG synthesis, IPCT/DIPPS and MPGS, respectively (Fig. 1), were deleted by double-crossover homologous recombination. The construction of the two mutants was confirmed by different techniques. PCR analysis corroborated the absence of the deleted genes. In addition, each flank of the recombination region was sequenced to show the correct replacement of the target gene by the pyrF cassette. Moreover, the genotype of each mutant was verified by Southern blotting using a 500-bp fragment upstream of the target gene and three restriction enzymes (see Fig. S2 in the supplemental material). The predicted hybridization pattern for single insertion of the marker cassette into the genome was obtained for both deletion mutants. Finally, analysis of the solute pool of the two mutant strains confirmed the absence of the targeted solute (MG or DIP) in cell extracts, as described below.

Effect of heat and salinity stress on growth and solute accumulation by mutants.

DIP-deficient and MG-deficient mutants were grown under optimal conditions and under stress conditions as described above for the parent strain. Under the optimal conditions, the MG-deficient mutant exhibited a growth rate and a final cell density similar to those of the parent strain. Curiously, the DIP-deficient mutant reached a higher final OD (1.42, in comparison with a final OD of 0.88 for the parent strain), and the growth rate was also slightly higher (Table 1; Fig. 3). Under heat stress conditions, the three strains exhibited similar growth parameters. Upon salinity stress, the growth of the MG-deficient strain was clearly impaired, but again, the DIP-deficient mutant grew slightly better than the parent strain. It seems as though MG is more important than DIP for the performance of this archaeon.

To interpret the results of the growth performance of the three strains, we compared the composition of the solute pools under optimal and stress conditions. Under optimal growth conditions, the absence of MG in the MG-deficient mutant was offset by an increase in the pools of DIP, alanine, and lactate (6-, 2.1-, and 2.3-fold, respectively) in comparison with the amounts in the parent strain (Table 2; Fig. 4A and B). While MG was by far the predominant solute in the parent strain, the solute pool in the MG-deficient strain comprised four solutes, DIP, alanine, aspartate, and lactate, in similar proportions. Like the parent strain, increasing the growth temperature led to a strong increase in the level of DIP (3-fold), which became the major solute, corresponding to about 55% of the total solute pool. Besides DIP, lactate and alanine were detected, but aspartate decreased to an undetectable concentration. The total concentrations of solutes were similar in the MG-deficient strain when heat stress and optimal conditions were compared, but the total concentrations clearly increased upon salt stress, primarily due to the increment in the pools of aspartate (4-fold) and DIP (3-fold). Even so, the total solute pool in the MG-deficient strain was significantly lower than that of the parent strain (1.08 and 1.43 μmol/mg of protein, respectively) when the strains were subjected to the same osmotic stress.

Except for the absence of DIP, the DIP-deficient mutant showed patterns of solute accumulation generally similar to those of the parent strain (Fig. 4C). MG was the major solute under all growth conditions examined, corresponding to about 75%, 67%, and 75% of the total organic solute pool under optimal, heat stress, and salt stress conditions, respectively. Notably, under optimal growth conditions, a significant increase in the total solute concentration of 1.6-fold was observed for the DIP-deficient mutant in comparison with that for strain COM1 (1.06 ± 0.19 versus 0.65 ± 0.04 μmol/mg of protein; Table 2). In the DIP-deficient mutant, in response to salinity stress, the level of MG increased by 50% and the total pool of solutes was about 50% and 10% higher than that in the MG-deficient and parent strains, respectively. l-myo-Inositol-1-phosphate, the substrate of the IPCT/DIPPS enzyme encoded by the deleted gene, was detected under conditions of heat and salt stress but not under optimal conditions. Raising the temperature from 90 to 98°C caused a decrease in the MG pool (1.7-fold). In the DIP-deficient mutant, the total solute pool was similar to that observed in the wild-type and MG-deficient strains (0.70, 0.68, and 0.68 μmol/mg of protein, respectively) when it was also challenged with heat stress.

The rationale behind this study relies on the assumption that the effects on the growth profiles of the mutants observed under stress conditions are caused by the deletion of the target solutes and not by the differential induction of other stress factors, such as heat shock proteins and chaperonins, which would undermine the interpretation of the results. The response of P. furiosus to heat stress was studied at the transcription level by Shockley et al. (23). Based on the results of these authors, we selected four genes that were clearly induced (4- to 7-fold) under heat stress, i.e., genes coding for a Hsp60-like chaperonin (PF1974), an Hsp20-like small heat shock protein (PF1883), one molecular chaperone belonging to the AAA⁺ family and homologous to VAT (PF1882), and an ATP-independent protease (PF1597). RT-PCR experiments were performed to confirm that the transcription levels of the corresponding genes in strain COM1 were identical in the mutant and control strains (see Fig. S3 in the supplemental material).

DISCUSSION

The variant strain P. furiosus COM1 showed a salinity profile similar to that of the original wild-type strain, P. furiosus DSM 3638^T, but the temperature profile was shifted downwards. Although both strains showed similar growth rates over the 88 to 98°C temperature range, the temperature for the maximum cell density of COM1 was 90°C and growth was not observed at 100°C (after 8 h), while the temperature for the maximum cell density of the wild-type strain was 95°C and some growth was observed at 100°C (7, 8). The explanation for the observed phenotypic differences could be related to the large number of changes encountered at the genome level (14). Exploring these differences by complementing the COM1 strain with genes that are functional in the DSM 3638^T strain provides an avenue to approach this problem.

As in P. furiosus DSM 3638^T, the COM1 variant accumulated MG in response to salt stress and DIP in response to supraoptimal temperatures. However, while DIP was by far the predominant solute in the wild-type strain at optimal and supraoptimal temperatures, MG was the major solute (over 65% of the total pool) in the variant strain under all conditions examined. It appears that the variant strain relies heavily on MG for both osmotic and thermal protection and DIP has a much smaller contribution. A comparison of the sequences of the regions (360 bp) immediately upstream of the genes encoding IPCT/DIPPS and MPGS in the genomes of the wild-type strain DSM 3638^T and the variant COM1 showed a perfect match. In a further attempt to find clues for the differentiated behavior of the strains, we looked for changes in putative transcription regulators: the sequences of 43 out of the 44 candidates annotated in the NCBI database were identical, and differences (58% identity) were found only in PF0054. For the moment, the origin of the phenotypic differences between the two strains remains obscure.

In this work, the genes involved in the synthesis of MG or DIP were disrupted in P. furiosus COM1 as a means to obtain insight into the role of these compatible solutes. If a specific solute were an important part of the machinery mounted by the cell to cope with stress, we would expect impaired growth under stressful conditions for a mutant unable to synthesize such a solute. Significantly, the growth of the MG-deficient mutant compared with that of the parent strain was negatively affected under salt stress but not under heat stress, an observation that supports a major role of MG in the osmoadaptation of this archaeon. In fact, the combined accumulation of DIP and aspartate in response to salt was not sufficient to offset the lack of MG. This observation, together with the consistent increase of MG upon osmotic stress in many (hyper)thermophiles (4), suggests that the regulatory network implied in MG accumulation is designed to respond effectively to osmotic imbalance. Surprisingly, the growth of the DIP-deficient mutant was as good as or even slightly better than that of the parent strain under either of the challenge conditions imposed. Comparison of the growth and solute profiles of the parent and the DIP-deficient strains under heat stress conditions suggests that MG can replace DIP for thermoprotection with similar performance. Indeed, the growth profiles were nearly superimposable, despite the substantial differences in the composition of the solute pools: while MG and DIP were the predominant solutes in the DIP-deficient and MG-deficient strains, respectively, the solute pool of the parent strain comprised a mix of MG and DIP in a proportion of approximately 2:1. Hence, it appears that DIP and MG can substitute for each other to protect against heat damage, but MG is more closely associated with osmoprotection.

The strategy of solute replacement for osmotic adjustment has been illustrated in a few mesophilic halophiles (24, 25). For example, in the bacterium Halobacillus halophilus, proline can be replaced by glutamate, glutamine, and ectoine to cope with osmotic stress (25). In Methanosarcina mazei, the lack of N^ε-acetyl-β-lysine is compensated for by the accumulation of glutamate and alanine, but at high salinity, the deficient mutant exhibited poorer growth than the parent strain. Thus, in this case, the replacement of N^ε-acetyl-β-lysine by alternative solutes was not fully adequate for osmotic adjustment (24). Replacement of compatible solutes for thermoprotection has also been reported in a DIP-deficient Thermococcus kodakarensis mutant, where the missing solute was replaced by aspartate, and this substitution had no effect on the ability of the strain to grow at supraoptimal temperatures (15).

The construction of P. furiosus mutants unable to synthesize either MG or DIP has provided more precise information on the triggers associated with the accumulation of specific solutes. In the mutant lacking MG synthesis, it is apparent that DIP accumulation is enhanced by heat stress and also by salt stress, revealing a certain flexibility of the sensing/regulatory mechanisms leading to DIP synthesis. In contrast, MG accumulation was clearly induced at elevated salinity but reduced under heat stress.

The molecular mechanisms involved in the regulation of DIP synthesis as part of the global heat stress response in hyperthermophiles have been investigated in Thermococcus spp. (21). The central role of the transcriptional regulator Phr is well established in members of the order Thermococcales (21, 26). At the optimal temperature, Phr binds to the promoter region of several genes, such as the hsp20 and AAA⁺ ATPase genes and the gene encoding IPCT/DIPPS, the enzyme implied in DIP synthesis, thereby blocking their transcription. Conversely, at higher temperatures, the affinity of Phr to the promoter region is decreased and transcription is activated. The activity of Phr in the regulation of DIP synthesis was validated in Thermococcus kodakarensis by disruption of the respective gene (21). Therefore, the molecular link between heat stress and DIP accumulation is fairly well established at the transcriptional level. On the other hand, the molecular mechanisms leading to enhancement of the DIP pool in response to elevated salinity remain obscure.

Herein we show that in P. furiosus the accumulation of MG increased in response to osmotic stress and was reduced at supraoptimal temperatures. This is the general trend for MG accumulation in (hyper)thermophiles; actually, upregulation of MG synthesis by heat was observed only in Rhodothermus marinus, which is unique, insofar as it possesses two pathways for the synthesis of this solute (27). In this thermophilic bacterium, MG synthesis is regulated at the translational level, as the amount of MPGS, the synthase involved in the 2-step pathway, increases in response to elevated salinity and decreases upon heat stress (5); on the other hand, the level of the synthase involved in the single-step pathway is selectively increased by heat. As demonstrated in this work, P. furiosus synthesizes MG exclusively via the 2-step pathway. It is therefore interesting to find that the outcome of MG regulation by stress in P. furiosus is identical to that observed for the 2-step pathway in R. marinus, although the specific regulatory mechanisms operating in bacteria and archaea are expected to be different.

In conclusion, this work shows that the lack of MG in P. furiosus was compensated for by DIP accumulation with a comparable efficacy in thermoprotection. MG and DIP played interchangeable roles in thermoprotection, while MG was primarily directed at osmoprotection. Future studies must determine the molecular mechanisms in the sequence of events from thermo- and osmosensing to the final accumulation of specific solutes.