Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XUS Government, 2003481143156Original ArticleH. K. Kagawa et al.Chaperonin alpha beta gamma subunits
Molecular Microbiology (2003) 48(1), 143–156
The composition, structure and stability of a group II chaperonin are temperature regulated in a hyperthermophilic archaeon
Hiromi K. Kagawa,1 Takuro Yaoi,2 Luciano Brocchieri,3 R. Andrew McMillan,4 Thomas Alton5 and Jonathan D. Trent4* 1 SETI Institute, 2035 Landings Dr., Mountain View, CA 94043, USA. 2 Department of Plant and Microbial Biology, University of California Berkeley, Berkeley, CA 94720, USA. 3 Department of Mathematics, Stanford University, Stanford, CA 94305, USA. 4 Astrobiology Technology Branch, NASA Ames Research Center, Moffett Field, CA 94035, USA. 5 Department of Biological Sciences, Western Illinois University, Macomb, IL 61055, USA. Summary The hyperthermoacidophilic archaeon Sulfolobus shibatae contains group II chaperonins, known as rosettasomes, which are two nine-membered rings composed of three different 60 kDa subunits (TF55 alpha, beta and gamma). We sequenced the gene for the gamma subunit and studied the temperaturedependent changes in alpha, beta and gamma expression, their association into rosettasomes and their phylogenetic relationships. Alpha and beta gene expression was increased by heat shock (30 min, 86∞∞C) and decreased by cold shock (30 min, 60∞∞C). Gamma expression was undetectable at heat shock temperatures and low at normal temperatures (75– 79∞∞C), but induced by cold shock. Polyacrylamide gel electrophoresis indicated that in vitro alpha and beta subunits form homo-oligomeric rosettasomes, and mixtures of alpha, beta and gamma form heterooligomeric rosettasomes. Transmission electron microscopy revealed that beta homo-oligomeric rosettasomes and all hetero-oligomeric rosettasomes associate into filaments. In vivo rosettasomes were hetero-oligomeric with an average subunit ratio of a:1b b:0.1gg in cultures grown at 75∞∞C, a ratio of 1a Accepted 5 December, 2002. *For correspondence. E-mail
[email protected]; Tel. (+1) 650 604 3686; Fax (+1) 650 604 1092.
© 2003 US Government
a:3b b:1gg in cultures grown at 60∞∞C and a ratio of 1a a:3b b:0gg after 86∞∞C heat shock. Using differential 2a scanning calorimetry, we determined denaturation temperatures (Tm) for alpha, beta and gamma subunits of 95.7∞C, 96.7∞C and 80.5∞C, respectively, and observed that rosettasomes containing gamma were relatively less stable than those with alpha and/or beta only. We propose that, in vivo, the rosettasome structure is determined by the relative abundance of subunits and not by a fixed geometry. Furthermore, phylogenetic analyses indicate that archaeal chaperonin subunits underwent multiple duplication events within species (paralogy). The independent evolution of these paralogues raises the possibility that chaperonins have functionally diversified between species. Introduction Chaperonins are double-ring structures found in nearly all organisms and composed of 60 kDa protein subunits referred to as heat shock proteins or HSP60s (Trent, 1996; Hartl and Hayer-Hartl, 2002). Heat and other stresses cause some organisms to increase the synthesis of their HSP60s, and it has been suggested that the chaperonins they form play essential roles in helping cells to recover from stress-related damage (Sanders, 1993; Peeples and Kelly, 1995; Trent et al., 1998). As proteins are damaged by HSP60-inducing stresses, a role for chaperonins in refolding damaged proteins was proposed (Edington et al., 1989; Hightower, 1991) and later generalized to include a role for chaperonins in de novo protein folding under non-stress conditions (for a review, see: Gething, 1997; Hartl and Hayer-Hartl, 2002). Although the chaperonin double ring isolated from cells is generally assumed to be the functional structure in vivo and protein folding is believed to be its primary function, both these assumptions have been questioned (Trent et al., 1998). Chaperonin double rings have been observed to assemble spontaneously into filaments in vitro (Harris et al., 1995; Furutani et al., 1998; Yaoi et al., 1998; Schoehn et al., 2000a), and evidence has been presented that such filaments exist in vivo (Trent et al.,
144 H. K. Kagawa et al. 1997). This suggested a function other than protein folding for at least some types of chaperonins. In addition to this possible structural role (Trent et al., 1997), other functions have been suggested, related to mRNA stabilization (Georgellis et al., 1995), gene translation by ribosomes (Ruggero et al., 1998) and membrane stabilization (Török et al., 1997; Trent, 2000). Observations of chaperonin filaments bring into question the assumption that chaperonin double rings are the structural unit in vivo, and the observed alternative activities raise the possibility that chaperonins have diversified functions in different organisms or cell types. Based on sequence and structural comparisons of chaperonins, two groups have been identified (Trent et al., 1991; Horwich and Willison, 1993). The so-called group I chaperonins are from Bacteria and from the chloroplasts and mitochondria of Eukarya, whereas the group II chaperonins are from Archaea and the eukaryotic cytoplasm. Group I chaperonins are composed of identical or two closely related subunits arranged in two stacked rings with seven subunits each. Group II chaperonins are composed of identical or diverse subunits arranged in rings of eight or nine subunits, depending on the organism. In the yeast Saccharomyces cerevisiae, for example, there is evidence of eight different subunits in each ring (Lin and Sherman, 1997). Among the Archaea, some thermophilic methanogens (e.g. Methanopyrus kandleri, Methanococcus jannaschii and Methanococcus thermolithotrophicus) have chaperonins with identical subunits (Furutani et al., 1998), whereas in the mesophilic methanogen Methanosarcina acetivorans, there are five different subunits (Galagan et al., 2002). Overall, within the 16 archaeal genomes sequenced to date, five have a single chaperonin gene, seven have two genes, three have three, and one has five. Of the 50 archaeal chaperonin sequences in the databases, most have >40% amino acid sequence identity. The majority of group II chaperonins in Archaea have eight subunits per ring and are referred to as ‘thermosomes’ (Klumpp and Baumeister, 1998), but the chaperonins in the thermoacidophilic Archaea in the family Sulfolobales have nine subunits per ring (Trent et al., 1991; Marco et al., 1994). To distinguish these Sulfolobus octadecameric chaperonins from thermosomes, we refer to them as ‘rosettasomes’ (Kagawa et al., 1995). We have reported previously that rosettasomes are composed of two types of HSP60s known as TF55 a and b, that TF55 a and b are among the most abundant proteins in Sulfolobus shibatae grown at optimal temperatures (75–83∞C) and that their synthesis increases at heat shock temperatures (85–88∞C) (Kagawa et al., 1995). It was also demonstrated that their abundance correlates with increased survival at lethal temperature or acquired thermotolerance (for a review, see Trent, 1996). Although these previous studies indicated that the rosettasome
contained two subunits, Archibald et al. (1999) identified a third related subunit (TF55 g) in Sulfolobus spp. by sequence analyses. Sequence information from Sulfolobus solfataricus (Charlebois et al., 1998) allows TF55 alpha, beta and gamma expression to be predicted based on codon usage (Karlin et al., 2001). This prediction suggests that the alpha and beta subunits in Sulfolobus spp. would be among the most highly expressed proteins, whereas gamma would be only moderately expressed or expressed only under specific conditions (J. Mrázek, personal communication). To test this prediction and to determine the relationship between TF55 alpha, beta and gamma, we cloned and sequenced the complete gamma gene from S. shibatae, compared the temperature-dependent expression of the three genes and determined their contribution to the structure of rosettasomes produced in vitro and in vivo. We developed a PAGE method that separates alpha, beta and gamma, and thereby determined the ratio of the three subunits in rosettasomes produced in vitro and in vivo. Using differential scanning calorimetry (DSC), we investigated the relative thermostability of the subunits and their impact on the thermostability of rosettasomes. Based on these results, we propose a model for the structure of rosettasomes in which the relative abundance of alpha and gamma functions to regulate the stability of rosettasomes at high and low temperatures respectively. Our phylogenetic analysis indicates that, although this model may apply to closely related species such as S. shibatae and S. solfataricus, it may not be meaningfully extrapolated to other organisms (even as close as Sulfolobus tokodaii) without experimental confirmation. More generally, phylogenetic analyses of all 50 currently known archaeal HSP60 sequences suggest that independent gene duplication events and sequence divergence reflect functional diversification of chaperonins between species.
Results The gamma gene Using the partial sequence information for the gamma gene reported by Archibald et al. (1999), we designed DNA primers and used standard and inverse polymerase chain reaction (PCR) techniques (Ochman et al., 1988) to clone the complete S. shibatae gamma gene and its flanking regions (Fig. 1). The sequence of the DNA fragment obtained by PCR contained a 1608 bp open reading frame (ORF) that we identified as the gamma gene by its homology to published gamma sequences (Archibald et al., 1999). In the region upstream from the assumed start codon, we found a hexanucleotide sequence (5¢-TTTATA3¢) that fits precisely the consensus for the archaeal ‘box © 2003 US Government, Molecular Microbiology, 48, 143–156
Chaperonin alpha beta gamma subunits 145 Fig. 1. The DNA and deduced amino acid sequences of Sulfolobus shibatae TF55-g. The 1608 bp gene codes for a 535-amino-acid protein. The gene is flanked by partial ORFs (ending at the arrow upstream and beginning at the arrow downstream). In the upstream sequence, there is a box A archaeal promoter element (position -43 bp ATG) followed by a putative transcription start (**) at the prescribed location from box A, and downstream by an oligo-T region typical of archaeal transcription terminators (dashed lines). The highly conserved ATP-binding motif of chaperonins (GDGTT) is underlined.
A’ promoter element ((C/T)TTA(T/A)A) (Reiter et al., 1990) and an 8 bp sequence in the vicinity of this ‘box A’, which fits the consensus (one mismatch) for the so-called BRE promoter element (Bell and Jackson, 1998). In the region located 25–28 bp from ‘box A’, there is a candidate transcription start site (purine following a pyrimidine) (Zillig et al., 1993). As in most archaeal genes, which lack Shine–Dalgarno sequences (Ma et al., 2002), gamma is also not flanked by a Shine–Dalgarno sequence. Beyond the stop codon for the gamma gene, we found a run of Ts, which is characteristic of transcription termination sites in Archaea (Zillig et al., 1993). The sequences 5¢ and 3¢ of the gamma gene contain ORFs that share significant sequence homology with the sequences flanking the gamma gene in S. solfataricus P2 (NCBI accession number AE006890). The 5¢ ORF is 73 bp and codes for part of a putative glycolsyltransferase (Fig. 1, arrow indicates the TGA stop codon). The 3¢ ORF is 183 bp and codes for part of a conserved hypothetical protein that is predicted to belong to the family of mandelate racemase © 2003 US Government, Molecular Microbiology, 48, 143–156
(Fig. 1, arrow indicates the ATG start codon). The DNA similarity indices for the 5¢ and 3¢ ORFs are 92.9% and 92.8% identity respectively. The gamma gene codes for a 535-amino-acid protein with a predicted molecular mass of 58.54 kDa and an isoelectric point of 5.5. The gamma protein contains 38% hydrophobic amino acids, 22% polar amino acids, 15% strongly acidic amino acids and 13% strongly basic amino acids. It does not contain either tryptophan or cysteine residues. A putative ATP-binding motif (GDGTT) is located in the region characteristic of nearly all chaperonins (Fenton et al., 1994). In S. shibatae, gamma is closely related to TF55 alpha and beta (Fig. 2) with 54% and 43% identity respectively. For comparison, the sequence identity between alpha and beta is 54%. Gamma is smaller than both alpha and beta with 25 fewer amino acids than alpha and 17 fewer than beta. Comparing all three proteins, alpha and gamma lack an 11amino-acid sequence present at the N-terminus of beta, whereas gamma and beta lack a six-amino-acid sequence
146 H. K. Kagawa et al. Fig. 2. Sequence alignment of alpha, beta and gamma proteins in which conserved positions are highlighted in black.
in the intermediate domain and a nine-amino-acid sequence at the C-terminus that are present in alpha. The overall similarity between alpha, beta and gamma suggests that they share a similar tertiary structure.
Temperature-dependent alpha, beta and gamma gene expression We compared the alpha, beta and gamma mRNA levels in cells exposed to different temperatures using Northern blot analyses (Fig. 3). Shifting mid-log phase cultures grown at 79∞C to 86∞C for 30 min (heat shock) significantly increased alpha and beta mRNA levels, and gamma mRNA levels decreased from low to undetectable (Fig. 3A). In contrast, shifting these cultures from 79∞C to 60∞C for 30 min decreased alpha and beta mRNA levels, but notably increased gamma mRNA levels. At 60∞C, S. shibatae’s doubling time increases from 5 h at 75∞C to ª10 h (Grogan, 1989). We cultivated S. shibatae at 60∞C for 20 h and determined the mRNA levels after
30 min incubations at temperatures ranging from 60∞C to 90∞C (Fig. 3B). At temperatures below 74∞C, beta mRNA levels were relatively low compared with levels observed in cells shifted to 74∞C and 86∞C. The highest beta mRNA levels were observed in cells shifted to 83∞C and 86∞C. In contrast, gamma mRNA levels were highest at 60∞C and declined with increasing temperature. To determine how rapidly S. shibatae adjusts gamma mRNA levels in response to a shift to low temperature, we transferred cells grown at 76∞C to 60∞C and monitored gamma mRNA levels after 0, 5, 15, 30 and 60 min (Fig. 3C). Gamma mRNA levels at 60∞C increased nearly twofold in 5 min and nearly 10-fold in 60 min. For comparison, at 86∞C, alpha and beta mRNA levels increase between five- and 10-fold in 30 min, but decline again to 75∞C (control) levels in 60 min (Kagawa et al., 1995). These results confirm previous observations that alpha and beta are heat shock proteins (Kagawa et al., 1995), and establish that gamma is not a heat shock protein. By the criterion of cold-induced synthesis, gamma may be classified as a ‘cold shock’ protein. © 2003 US Government, Molecular Microbiology, 48, 143–156
Chaperonin alpha beta gamma subunits 147 A
B
C
Temperature-dependent changes in alpha, beta and gamma protein ratios in vivo To determine how these changes in expression influenced the composition of rosettasomes, we developed a urea/ SDS-PAGE method (Alton PAGE) based on a method developed previously to separate different forms of actin (Storti and Rich, 1976). Using this method, we were able to separate the three subunits, despite the similarity of their molecular masses (a = 59.72 kDa, b = 59.68 kDa, g = 58.54 kDa), and found that, in cultures grown at 60∞C and 76∞C, all three subunits were present but, in cultures exposed to 86∞C for 30 min, only alpha and beta were present (Fig. 4). Although the intensity of the subunit bands shown in immunoblots (Fig. 4A) is not indicative of subunit ratios in cells because of the difference in immunoreactivity of the polyclonal antibodies used, these ratios were determined by purifying rosettasomes from cultures grown at different temperatures (Fig. 4B). In cultures grown at 60∞C, the subunit ratio in purified rosettasomes was ª1a:3b:1g. In cultures grown at 76∞C, this subunit ratio changed to ª1a:1b:0.1g. In cultures exposed to 86∞C for 30 min, the ratio was ª1a:1b with no detectable gamma. The subunit composition of rosettasomes from the different temperatures was quantified more accurately by densitometry (Fig. 4C). At 60∞C, rosettasomes contained 20% alpha, 61% beta and 19% gamma. At 76∞C, rosettasomes contained 41% alpha, 56% beta and 3% gamma. At 86∞C, rosettasomes contained 41% alpha and 59% beta, with no detectable gamma. Notably, on a percentage basis, beta remains approximately the same at all temperatures, whereas alpha and gamma change. These results are consistent with the Northern blot analyses.
The composition of rosettasomes in vitro
Fig. 3. Temperature-dependent gene expression of the three rosettasome subunits indicated by their mRNA levels using subunitspecific probes in Northern hybridizations. A. The subunit mRNA levels in S. shibatae cultures grown at 79∞C and exposed to 60∞C and 86∞C for 30 min. B. The beta and gamma subunit mRNA levels from 76∞C cultures incubated at 60∞C for 20 h and exposed to the temperatures indicated for 30 min (the intensity of each mRNA band was determined by densitometry and plotted against temperature; bottom). C. The gamma mRNA levels in a 76∞C culture (lane C) incubated at 60∞C for the indicated times (the equal intensity of ethidium bromidestained rRNA bands indicates that total RNA levels are the same in each lane; bottom).
© 2003 US Government, Molecular Microbiology, 48, 143–156
To characterize further the interaction between alpha, beta and gamma, we expressed the three subunits separately in Escherichia coli, purified them and studied the structures that they formed in vitro using non-denaturing and Alton PAGE (Fig. 5). Non-denaturing PAGE indicated that the subunits had distinct electrophoretic mobilities (Fig. 5A, lanes 1–3) and that, in the presence of ATP/Mg, alpha and beta formed distinct homo-oligomeric rosettasome bands, but gamma remained in subunits (Fig. 5A, lanes 4–6). Using mixtures of 1a:1b:1g, the hetero-oligomeric rosettasomes had electrophoretic mobility that was clearly different from the homo-oligomeric alpha band and slightly different from the beta band (Fig. 5A, lanes 4, 5 and 7). Changes in the intensities of the subunit bands indicated that all three subunits were present in the rosettasomes formed, but not in equal proportions (Fig. 5A, lane 7).
148 H. K. Kagawa et al. complex structures that are destroyed by electrophoresis (Yaoi et al., 1998). Rosettasomes formed in vitro were therefore analysed by transmission electron microscopy (TEM). In the presence of ATP/Mg, the structures formed by alpha, beta and gamma alone and in a 1:1:1 mixture are shown in Fig. 6. As indicated by non-denaturing PAGE, alpha formed discrete homo-oligomeric rosettasomes with the characteristic ninefold symmetry in ‘end views’ and four striations in ‘side views’ (Fig. 6A). For beta,
Fig. 4. The alpha, beta and gamma subunits in S. shibatae cultures from 60∞C, 76∞C and 86∞C. A. Total proteins were analysed using polyclonal antibodies against rosettasomes that qualitatively recognize alpha, beta and gamma. B. The subunits from purified rosettasomes from cultures grown at 60∞C, 76∞C and 86∞C were separated by Alton PAGE. C. The relative proportions of alpha, beta and gamma subunits (in B) were quantified using a UVP gel documentation system.
To confirm that rosettasomes contained all three subunits, the rosettasome band was excised from the non-denaturing polyacryamide gel, and the associated proteins were analysed by Alton PAGE (Fig. 5B). This procedure indicated that, indeed, the native rosettasome band contained three proteins with the same electrophoretic mobilities as alpha, beta and gamma, and that they were present in different ratios despite equal concentrations in solution. TEM images of alpha, beta and gamma structures We have reported previously that rosettasomes can form
Fig. 5. The effects of ATP/Mg on rosettasome formation by alpha, beta and gamma subunits indicated by non-denaturing and denaturing (Alton) PAGE. A. The mobilities of pure recombinant alpha, beta and gamma subunits (lanes 1–3). In the presence of 1 mM ATP and 25 mM MgCl2, alpha and beta formed rosettasome bands after 1 h at 76∞C (lanes 4 and 5), but gamma remained as subunits after 1 h at 60∞C (lane 6). In the presence of this concentration of ATP/Mg, equimolar mixtures of alpha, beta and gamma formed rosettasome bands after 1 h at 60∞C (or 75∞C) (lane 7). B. The rosettasome band formed from an equimolar mixture of alpha, beta and gamma subunits excised from non-denaturing polyacrylamide gels (left) and analysed by denaturing Alton PAGE. The rosettasome band contained approximately equal amounts of alpha and beta, but considerably less gamma, as indicated by the relative intensity of the bands (right). This is consistent with the amount of residual gamma subunit in the non-denaturing gel (bottom, left). © 2003 US Government, Molecular Microbiology, 48, 143–156
Chaperonin alpha beta gamma subunits 149 however, what were homo-oligomeric rosettasome bands in non-denaturing PAGE appeared as filaments of rosettasomes and bundles of filaments in TEM (Fig. 6B). Both non-denaturing PAGE and TEM indicated that gamma does not assemble into rosettasomes, although some amorphous aggregates and non-uniform round objects were seen in the TEM (Fig. 6C). With 1:1:1 mixtures of alpha, beta and gamma, the hetero-oligomeric rosettasome bands visible in non-denaturing PAGE were seen as a combination of rosettasomes and filaments in the TEM (Fig. 6D). These structures differed from beta filaments in their level of organization (Fig. 6B and D), but were indistinguishable from filaments formed by 1:1 mixtures of alpha and beta (not shown). All forms of rosettasome filaments are apparently unstable in non-denaturing PAGE. Thermostability of alpha, beta and gamma To determine how the composition of rosettasomes may affect their thermostability, we determined the denaturation temperatures of subunits and rosettasomes using differential scanning calorimetry (DSC) (Fig. 7). In Hepes buffer (pH 7.5), all three subunits denatured irreversibly; pure alpha at 95.7∞C, pure beta at 96.7∞C and pure gamma at 80.5∞C (Fig. 7A). We formed rosettasomes by adding ATP/Mg to mixtures of ab (subunit ratio 1:1) or abg (ratio 1:1:0.1), incubating for 1 h at 60∞C. The ab rosettasomes denatured at 93.8∞C, whereas the abg rosettasomes denatured at 93.5∞C (Fig. 7B). The impact of gamma on the stability of rosettasomes, however, is more clearly indicated by the calorimetric heat change (DH). The calculated DH for thermal denaturation of ab rosettasomes (1:1) was 3.01 ¥ 105 kcal mol-1 (calculated error 0.0249 ¥ 105). The DH for abg rosettasomes was 2.24 ¥ 105 kcal mol-1 (calculated error 0.0139 ¥ 105). This suggests that rosettasomes containing alpha and beta formed at 86∞C have a higher denaturation enthalpy than those containing alpha, beta and gamma formed at lower temperatures. Fig. 6. TEM images of the recombinant alpha, beta and gamma subunits in the presence of ATP/Mg. A. Pure alpha (1 mg ml-1) formed typical double-ring rosettasomes mostly visible as end views (arrowhead indicates side views). B. Pure beta (1 mg ml-1) formed rosettasomes (arrowheads) that stack spontaneously ring to ring to form ordered bundles of filaments. C. Pure gamma (1 mg ml-1) formed mostly amorphous aggregates with occasional round structures (arrowheads) that are smaller than rosettasomes and not of uniform size. D. An equimolar mixture of alpha, beta and gamma (0.5 mg ml-1 each) formed rosettasomes and less ordered and less bundled filaments; these filaments are indistinguishable from those formed by equimolar mixtures of alpha and beta (not shown). The 50 nm scale bar applies to all images.
© 2003 US Government, Molecular Microbiology, 48, 143–156
Phylogenetic analysis The phylogenetic relationships of the 50 currently known archaeal chaperonin genes are shown in Fig. 8. This updates the 30 sequence tree published by Archibald et al. (1999) and supports the division of cultivable Archaea into Euryarchaeota and Crenarchaeota and the positioning of a root between these two groups (for a review, see Forterre et al., 2002). Our analyses suggest that the gamma genes originated by duplication of an alpha gene in the progenitor of the Sulfolobales. The alpha gene is one of two paralogues (alpha and beta) resulting from a duplication of the primordial HSP60 gene that
150 H. K. Kagawa et al.
Fig. 7. The thermostability of rosettasome subunits and rosettasomes determined by differential scanning calorimetry (DSC). A. In 25 mM Hepes buffer (pH 7.5), the melting temperature or thermal mid-point of transition (Tm) of recombinant subunit alpha (dotted) was 95.7∞C, beta (dash) was 96.7∞C, and gamma (solid) was 80.5∞C. B. In the presence of ATP/Mg, the Tm of ab rosettasomes was 93.8∞C and of abg rosettasomes was 93.5∞C, but their calorimetric heat change (DH) was 3.01 ¥ 105 ± 2.49 ¥ 103 kcal mol-1 and 2.24 ¥ 105 ± 1.39 ¥ 103 kcal mol-1 respectively; the error was based on a two-state curve fit, using the Levenberg–Marquardt non-linear least squares method.
occurred before the radiation of the crenarchaeal species. The alpha–beta and alpha–gamma duplication events are indicated by a symbol ‘+’ in the crenarchaeal section of the tree (Fig. 8, upper right quadrant). From the length of the different branches, we can infer that, preceding the species radiation, the alpha paralogue differentiated much more than the beta paralogue. In contrast, the gamma paralogue differentiated only modestly after its duplication from alpha within the Sulfolobales lineage and before radiation of the Sulfolobus species. After speciation of S. tokodaii from the progenitor of S. shibatae and S. solfataricus, the gamma gene underwent independent differentiation within the two lineages. Long branches may indicate relaxation of functional constraints or positive
selection associated with the acquisition of new functions (Nei and Kumar, 2000). From our phylogenetic tree, we can also infer that the crenarchaeal paralogues originated from duplication events that are not correlated with those that gave rise to the paralogues observed in euryarchaeal species. These are the result of at least eight (possibly 10) other independent duplication events that occurred in the lineages of Methanobacterium, Thermococcus, Archaeoglobus and Thermoplasma (with one duplication each) or Haloferax (with two duplications) and Methanosarcina (with at least two, possibly four, duplications). We gain a new perspective on the relationship between archaeal HSP60 genes from the recently sequenced genomes of Methanosarcina mazei and acetivorans. Both species have three closely related archaeal-type HSP60s that arose from one, possibly two, duplication events before their split. M. acetivorans has two additional archaeal-type genes (hsp-4 and hsp-5) that are similar to each other (97% identity) but different from all other archaeal chaperonin subunits (ª 25% identity). The origin of the progenitor gene of hsp4 and hsp-5 is unresolved by our analysis. Surprisingly, in addition to the archaeal group II chaperonins, both M. mazei and M. acetivorans harbour the bacterial group I chaperonins (GroESL operon). The Methanosarcina GroELs have 55% identity with each other and 55–59% identity with bacterial GroELs. Bacterial-type HSPs have been described in Archaea and attributed to lateral transfer (Klump and Baumeister, 1998), but this is the first example of a bacterial-type HSP60 in Archaea, and the functional implications of this observation remain to be elucidated. Discussion We have studied the temperature-dependent expression of the three S. shibatae HSP60 genes, alpha, beta and gamma, that form the double-ring structures known as rosettasomes. We observed that the expression of alpha and beta genes is heat inducible, as described previously for S. shibatae (Kagawa et al., 1995) and observed for related genes in other Archaea (Emmerhoff et al., 1998; Izumi et al., 2001; Yoshida et al., 2002). We observed that the expression of the gamma gene, however, is cold rather than heat inducible. Using monoclonal antibodies, Izumi et al. (2001) reported that one of the two HSP60 genes (CpkA) in Thermococcus kodakaraensis (KOD1) is more abundant at low growth temperatures. We observed that the differential expression of alpha, beta and gamma in S. shibatae influenced the composition and stability of rosettasomes and that the composition of rosettasomes influenced filament formation. We predict that these changes in rosettasomes and filament formation have important structural and functional implications. © 2003 US Government, Molecular Microbiology, 48, 143–156
Chaperonin alpha beta gamma subunits 151
Fig. 8. Phylogenetic tree of archaeal chaperonins. The tree was derived by the neighbour-joining method (Saitou and Nei, 1987) from 50 sequenced archaeal HSP60s using the Ota and Nei (1994) distance transformation with the G-distribution parameter a = 1.42. Sequences from complete genomes are shown in bold. Paralogues from the same organism are distinguished by the descriptive name(s) reported in the databases or by gene names when the descriptive names are ambiguous or not available. Lineages relating alpha, beta and gamma proteins from Crenarchaeota are shown in the upper right quadrant. The root of the tree is indicated by a black circle and positioned in accordance with Archibald et al. (1999). Duplication events are indicated by the symbol ‘+’. Support values shown for each node include a value in bold (top row) that is the bootstrap proportion obtained by applying the neighbour-joining method and mutation rate G-distributed with parameter a = 1.42. Values in italics indicate the range of bootstrap proportions obtained by assigning to the a parameter the alternative values 0.5, 1.0, 1.5 and 2.0 or by inverting the SSPA similarity values (see Experimental procedures). Missing values indicate identity to the bold values. The second row shows the support to the node associated with the maximum likelihood procedure implemented in PUZZLE (Strimmer and von Haeseler, 1996). Blank circles hide unresolved topologies. A topology is considered not resolved if PUZZLE support values for the corresponding nodes are £50% or bootstrap proportions for a = 1.42 or for more than one of the alternative distance transformations are £50%. The scale bar indicates 50 mutational events over 397 aligned positions, corresponding to 12.6 mutations per 100 positions.
Structurally, the differential expression of rosettasome subunits creates in vivo conditions in which a diverse population of rosettasomes is produced. Our observation that alpha and beta can form both homo-oligomeric and hetero-oligomeric rosettasomes (cf. Fig. 6) indicates that © 2003 US Government, Molecular Microbiology, 48, 143–156
the diversity of structures in S. shibatae may be greater than suggested previously (Knapp et al., 1994; Kagawa et al., 1995; Archibald et al., 1999; Schoehn et al., 2000b). In other words, the changing ratio of subunits that we observed in S. shibatae can create a variety of differ-
152 H. K. Kagawa et al. ent double rings, which we propose will reflect the relative abundance of subunits in cells and the stability of their interactions, but not a single crystallographic symmetry. Under heat shock conditions, for example, the abundance and stability of subunits creates a mixed population of rosettasomes dominated by alpha and beta subunits, and excluding rosettasomes containing gamma. Observations of structural plasticity for related chaperonin subunits in other organisms (Waldmann et al., 1995; Yoshida et al., 1997; Schoehn et al., 2000a) suggest that heterooligomeric chaperonins in general may be more structurally diverse in vivo than indicated by crystallographic analyses. The structural diversity of chaperonins in other species, suggested by our observations of rosettasomes, remains to be elucidated. The temperature-dependent changes that we observed in subunit expression in S. shibatae suggest that the beta subunit is fundamental to the structure and presumably the function of rosettasomes. The observation that beta expression remains nearly constant at all temperatures, whereas alpha and gamma subunits vary, suggests that alpha and gamma may play a regulatory role modulating the function of rosettasomes in relation to temperature. Two current hypotheses that relate the structure of rosettasomes to their function in vivo are: (i) the double rings play a role in protein folding (Trent et al., 1991; Guagliardi et al., 1994); and (ii) the double rings are building blocks for filaments that play a role in the organization of the cytoplasm or in the stability of cell membranes (Trent et al., 1997; Trent, 2000). In the protein folding hypothesis, proteins fold within the central cavity of rosettasomes (Hartl and Hayer-Hartl, 2002), which implies that changes in the subunit composition that influence volume and reactivity of the central cavity would be of potential importance. The N- and C-termini of subunits are believed to project into and occlude the central cavity and, as these termini differ between subunits (cf. Fig. 2), changes in subunit composition of rosettasomes would indeed affect the central volume. Critical reactivity in the central cavity for protein folding has been associated with the apical domains of subunits (Ditzel et al., 1998), which are also regions that differ between subunits. In addition to the volume and binding properties of the central cavity, the composition of rosettasomes may influence how they may function in a protein folding process. It has been proposed that, for rosettasomes, which lack co-chaperonins (Saibil, 2000), this process depends on an ATP/Mg-dependent conformational change and dissociation of the subunits (Quaite-Randall et al., 1995). The differences in the thermostabilities of rosettasomes (cf. Figs 6 and 7) suggest that changes in subunit composition would affect their dissociation potential. It remains to be demonstrated how subunit composition would influence their ATPase activity and conformational flexibility. It also remains to be dem-
onstrated whether changes in rosettasome composition affect their ability to fold protein in vitro (Guagliardi et al., 1994) and, more critically, whether rosettasomes play a role in protein folding in vivo under any circumstances (Trent et al., 1998). An alternative hypothesis for the function of rosettasomes is that they are the building blocks for filaments that play a role in the cytoplasm or the membrane of cells (Trent et al., 1997; Trent, 2000). Evidence has been published that rosettasome filaments are present in vivo (Trent et al., 1997). It is suggested that such chaperonin filaments are a means of sequestering double rings that play a role in protein folding or that the filaments themselves play a structural role in cells (Trent et al., 1997; 1998). The observations that rosettasomes isolated from cells are hetero-oligomeric and that hetero-oligomeric rosettasomes assemble into filaments in vitro (cf. Fig. 6) support the hypothesis that filaments are functionally important in vivo. If rosettasome filaments, and not double rings, are the functional form in vivo, then the observed temperature-dependent changes in rosettasome composition must be interpreted in relation to their effect on filaments. The observed changes in rosettasome thermostability (cf. Fig. 7) would be expected to affect the stability of filaments, changing their length, flexibility or binding properties with components in the cytoplasm or on the cell membrane. Recent evidence indicates that, in S. shibatae, rosettasomes are associated with the cell membrane (J. D. Trent, H. K. Kagawa, C. D. Paavola, J. Howard and P. Chong, unpublished), which supports speculation that they are influencing the stability and permeability of the cytoplasmic membrane (Trent, 2000). We have demonstrated here that S. shibatae modifies the composition of its rosettasomes in response to temperature variations and that rosettasomes of different composition form filaments in vitro. The functional consequences of these changes in vivo require further investigation. Our phylogenetic analyses of archaeal chaperonins revealed that HSP60 genes in different archaeal species are the result of multiple duplication events (8–10 in Euryarchaeota and two in Crenarchaeota) that occurred independently early in the evolution of the different archaeal lineages, which is consistent with other phylogenetic analyses of chaperonins (Archibald et al., 1999; Archibald and Roger, 2002). The independent evolutionary histories of chaperonin genes suggest that, in different lineages, chaperonins may have evolved different functions. The gamma genes among the Sulfolobales, for example, diverged and evolved rapidly (cf. Fig. 8). Although the gamma genes (as well as the alpha and beta genes) are nearly identical in S. shibatae and S. solfataricus, they are much diverged from the gamma gene in S. tokodaii. Such rapid evolution suggests an accommodation to a different or more specialized function. In general, © 2003 US Government, Molecular Microbiology, 48, 143–156
Chaperonin alpha beta gamma subunits 153 the HSP60 genes in different lineages appear to have duplicated independently, and the different paralogues show different patterns of differentiation. Independent duplication events occurred early in the evolution of the Thermoplasma, Methanobacterium, Thermococcus– Pyrococcus and Archaeoglobus lineages, twice in the Halophile lineage, up to four times in the Methanosarcina lineage, and duplications and gene losses are apparent in Pyrococcus and Halobacterium. Different patterns of differentiation are seen in the alpha and beta paralogues in Archaeoglobus fulgidus that diverged only moderately and at similar rates, whereas in most other species, one of the paralogues has diverged significantly more than others (cf. Fig. 8; cct3 in Haloferax, gamma in Sulfolobus, and beta in Methanobacterium). These observations suggest that HSP60s in Archaea (and perhaps Eukarya and some groups of Bacteria) have been subject to intense evolutionary processing, possibly because of their high levels of expression (Karlin and Brocchieri, 2000). The duplications and modifications in HSP60s suggest that ‘chaperonins’ may be exploited for multiple functions in the cell, and it may therefore be inappropriate to ascribe a single function to them for all species. Our phylogenetic analyses suggest that, if HSP60s participate in a single, generic function, then this must result from parallel or convergent evolution. If, however, chaperonins have evolved different functions in different species, new functional hypotheses must be scrutinized. Conclusion We report the complete sequence of the gamma subunit of the S. shibatae chaperonins, known as rosettasomes. We observed that gamma is differentially expressed at low temperatures, whereas the related subunits, alpha and beta, are differentially expressed at high temperatures. Studies of the regulation of these genes may provide additional insights into archaeal gene regulation (Bell and Jackson, 1998). We demonstrate that the three subunits (alpha, beta and gamma) interact to form rosettasomes in vitro with a stoichiometry of 1a:1b:0.1g, whereas the ratio of subunits in rosettasomes isolated from cells changes with growth temperature. Using DSC, we observed that the thermostability of alpha, beta and gamma influences the thermostability of the rosettasomes. We propose that, in S. shibatae, the relative abundance of subunits in cells and not a single prescribed symmetry creates a population of rosettasomes with different properties. The diversification of subunits is expected to affect their function in vivo, which has yet to be determined. Our phylogenetic analyses reveal the independent evolution of archaeal chaperonin subunits and suggest that this class of molecules may have evolved different functions in different species that remain to be explored. © 2003 US Government, Molecular Microbiology, 48, 143–156
Experimental procedures Cloning and sequencing of the gamma gene The gamma gene was amplified by the PCR method from S. shibatae genomic DNA purified using Qiagen genomic tips. PCR primers (P1, 5¢-ATGAACTTAGAGCCTTCCTAT-3¢; and P2, 5¢-TTAACTCCATAAGAAACTTGT-3¢) were based on previously published partial gamma sequence information (Archibald et al., 1999). The inverse PCR method (Ochman et al., 1988) was used to obtain the complete gamma gene and its flanking regions. Briefly, AseI-digested genomic DNA was circularized by self-ligation, and a 1.2 kbp fragment was obtained by PCR amplification after 25 cycles (30 s at 94∞C, 1 min at 50∞C and 1 min at 72∞C), using Vent polymerase (New England BioLabs). The PCR fragment was ligated into pBluescript SK+ (Stratagene) to obtain a plasmid that was transformed into E. coli (strain DH5a). The gamma gene was sequenced on both strands by the dideoxy-chain termination method (Sanger et al., 1977), analysed using the program DNASTAR and submitted to GenBank (accession number AF313410).
Gamma gene expression in E. coli The complete gamma gene was PCR amplified from S. shibatae genomic DNA using two primers (primer 1, 5¢GAAAGAACATATGGCCTATTTATTAAGAGAAGGAACACAG3¢; and primer 2, 5¢-TAAAGTACTCGAGAAAACCTAAATAAA ATAATCATATCTTAAC-3¢). This fragment was cloned into the NdeI and XhoI sites of the plasmid vector pET22b (Novagen). Expression in E. coli strain BL21(DE3) ‘codon plus’ in LB media containing 50 mg ml-1 carbenicillin (Sigma) was under IPTG regulation. The alpha and beta genes, which were cloned previously (Kagawa et al., 1995), were similarly expressed.
Purification of recombinant alpha, beta and gamma Escherichia coli expressing alpha, beta or gamma subunits from S. shibatae rosettasomes was harvested by centrifugation at 7000 r.p.m. for 5 min at 4∞C (GSA rotor, Sorvall RC5). Cells were resuspended in a protease inhibitor cocktail (0.25 ml g-1 wet weight cells; Sigma), subjected three times to freeze/thawing at -80∞C and 21∞C and resuspended in two volumes of buffer A (25 mM Hepes, pH 7.5) containing lysozyme (46 400 U ml-1; Sigma) and benzonase (1 U ml-1; Sigma). The resuspension was incubated on ice for 30 min and sonicated for 2 min intervals every 10 min for 60 min at power setting 3, duty cycle 40% (Branson Sonifier 450). The lysate was centrifuged at 18 000 r.p.m. for 10 min at 4∞C (45Ti rotor; Beckman), and the decanted supernatant was heated for 30 min to 86∞C for alpha and beta and 73∞C for gamma. The heat-precipitated E. coli proteins were removed by centrifugation at 30 000 r.p.m. for 30 min at 4∞C (45Ti rotor; Beckman). The soluble alpha, beta or gamma fraction was purified by anion exchange chromatography (DEAESepharose fast flow, followed by MonoQ 10/10; Pharmacia) as described previously (Kagawa et al., 1995). Proteins were concentrated with a centrifugal concentrator (Centricon Plus30, 30K MWCO) as described by the manufacturer. Protein
154 H. K. Kagawa et al. concentrations were determined using the DC protein assay kit (Bio-Rad) with BSA standards. All proteins were stored in aliquots at -80∞C.
Purification of native rosettasomes Sulfolobus shibatae cultures incubated at 60∞C, 76∞C and 86∞C were harvested in stationary phase (ª 2 ¥ 109 cells ml-1). This required nearly 10 days of incubation at 60∞C and 4 days of incubation at 76∞C. S. shibatae did not grow at 86∞C, but late log cultures from 76∞C were incubated at 86∞C for up to 20 h before harvesting. Native rosettasomes from these cultures were purified as described above for the recombinant proteins with the following modifications. MgCl2 and KCl (10 mM each) were added to buffer A, heating steps were omitted and, between the DEAE and MonoQ chromatography, a 10–30% linear glycerol gradient was used to separate the 20S rosettasomes. This 30 ml glycerol gradient was centrifuged at 16 000 r.p.m. for 17 h in an SW28 rotor at 4∞C (Beckman).
Northern blotting hybridization Total RNA from S. shibatae grown at 76∞C to approximately mid-log phase and exposed to different temperatures in preheated water baths or a gradient thermocycler (MasterCycler gradient; Eppendorf) was purified using the RNeasy kit (Qiagen) as described by the manufacturer. Equal amounts of RNA, as determined by absorbance at 260 nm after DNase treatment, were separated by 6.6% formaldehyde denaturing 1% agarose gel electrophoresis buffered by HES (50 mM Hepes, 1 mM EDTA and 5 mM sodium acetate, pH 7.0). RNA was transferred to nylon membranes (Magnacharge; MSI) by capillary blotting with 20¥ SSC and probed with PCR products from the alpha, beta and gamma genes (regions 167–1003, 187–1527 and 359–1183 bp respectively). Probes were labelled with horseradish peroxidase using the ECL kit (Amersham) as directed by the manufacturer. The signal was recorded on X-ray film and quantified using an image acquisition system (BioChemi; UVP) with accompanying software (LABWORKS 4.0; UVP).
Subunit characterization by non-denaturing and Alton PAGE Rosettasomes were separated by 4–10% gradient non-denaturing PAGE (acrylamide–bis 37.5:1; Amresco) as described previously (Yaoi et al., 1998). Alton PAGE is based on a procedure used to separate different forms of actin (Storti and Rich, 1976) and consisted of a solution containing 6 M urea (5.4 g of urea) in 15 ml of 10% polyacrylamide separation gel solution containing 0.1% SDS, buffered with 375 mM Tris-HCl, pH 8.8. Polyacrylamide gels were dried, and proteins were quantified using an image acquisition system (BioChemi; UVP) and analysis software (LABWORKS 4.0; UVP).
Immunoblotting Proteins separated by Alton PAGE were electroblotted
(Trans-blot; Bio-Rad) to nitrocellulose membranes (Trans-blot transfer medium; Bio-Rad) and analysed as described previously (Kagawa et al., 1995) using polyclonal antibody raised in rabbits against the rosettasome proteins purified from S. shibatae cultivated at 76∞C. The signal was produced by peroxidase-labelled goat anti-rabbit IgG (affinity purified; Kirkegaard and Perry Laboratories) using the ECL kit (Amersham) and detected on X-ray film.
Electron microscopy Protein samples were attached to lacy carbon grids with ultrathin Formvar (Ladd Research Industries), stained with 0.22 mm filtered 2% uranyl acetate for 3 min and air dried at room temperature. The grids were viewed in a LEO 912 AB with tungsten filament at 100 kV. Images were recorded with a MegaView digital camera using ANALYSIS 3.5 software.
Differential scanning calorimetry (DSC) The alpha, beta and gamma proteins were analysed using an ultrasensitive differential scanning calorimeter (VP-DSC MicroCal; LLC). The pure proteins (1.5 mg ml-1 or ª17 mM) in 25 mM Hepes, pH 7.5, were scanned from 25∞C to 110∞C at 50∞ h-1. For studies of the effects of gamma on rosettasome stability, rings were assembled by the addition of ATP/ Mg (1 mM ATP, 25 mM MgCl2) to a 1 ml sample, incubated at 60∞C for 1 h and scanned against a reference containing ATP/Mg. Rosettasomes with a ratio 1a:1b:0.1g were made by adding ATP/Mg to a 1 ml solution containing 1.5 mg of alpha and beta and 0.15 mg of gamma. DSC data were processed using the program ORIGIN (MicroCal; LLC) to subtract the buffer–buffer reference scans from sample scans and to fit appropriate baselines. Tm and DH values were calculated by fitting the simplest two-state curve using the Levenberg–Marquardt non-linear least squares method.
Phylogenetic analyses Fifty archaeal chaperonin sequences from DNA and protein databases were aligned using the multiple sequence alignment program ITERALIGN (Brocchieri and Karlin, 1998). From this alignment, we identify 397 positions that were used to produce a high-confidence alignment of all sequences. This alignment was used to produce phylogenetic trees by the neighbour-joining (NJ) procedure (Saitou and Nei, 1987) and by the maximum-likelihood (ML) method implemented in the computer program PUZZLE (Strimmer and von Haeseler, 1996). In the neighbour-joining procedure, pairwise distances between sequences were estimated: (i) inverting SSPA (significant segment pairwise alignment) similarity values (Brocchieri and Karlin, 1998) as suggested by Feng and Doolittle (1997); and (ii) by the transformation of Ota and Nei (1994) based on the G-distribution of the position-dependent mutational rate. In the G-distribution for our sequence data, the parameter a was estimated as a = 1.42 by the procedure implemented in PUZZLE. Trees were also produced by setting a = 0.5, 1.0, 1.5 or 2.0 to evaluate the robustness of the results over different parameterizations. Neighbour-joining © 2003 US Government, Molecular Microbiology, 48, 143–156
Chaperonin alpha beta gamma subunits 155 trees were tested by perturbation analysis (‘bootstrap’) with 1000 independent resampling experiments.
Acknowledgements We thank A. J. Martin, C. Chen and A. Sanford for technical assistance, and S. Tornaletti for critical reading of the manuscript. This research was supported by DOE/BES funding.
References Archibald, J.M., and Roger, A.J. (2002) Gene duplication and gene conversion shape the evolution of archaeal chaperonins. J Mol Biol 316: 1041–1050. Archibald, J.M., Logsdon, J.M., and Doolittle, W.F. (1999) Recurrent paralogy in the evolution of archaeal chaperonins. Curr Biol 9: 1053–1056. Bell, S.D., and Jackson, S.P. (1998) Transcription and translation in archaea: a mosaic of eukaryal and bacterial features. Trends Microbiol 6: 222–228. Brocchieri, L., and Karlin, S. (1998) A Symmetric-iterated multiple alignment of protein sequences. J Mol Biol 276: 249–264. Charlebois, R.L., She, Q., Sprott, D.P., Sensen, C.W., and Garrett, R.A. (1998) Sulfolobus genome: from genomics to biology. Curr Opin Microbiol 1: 584–588. Ditzel, L., Lowe, J., Stock, D., Stetter, K.-O., Huber, H., Huber, R., and Steinbacher, S. (1998) Crystal structure of the thermosome, the archeal chaperonin and homolog of CCT. Cell 93: 125–138. Edington, B.V., Whelan, S.A., and Hightower, L.E. (1989) Inhibition of heat shock (stress) protein induction by deuterium oxide and glycerol: additional support for the abnormal protein hypothesis of induction. J Cell Physiol 139: 219–228. Emmerhoff, O.J., Klenk, H.-P., and Birkeland, N.-K. (1998) Characterization and sequence comparison of temperature-regulated chaperonins from the hyperthermophilic archaeon Archaeoglobus fulgidus. Gene 215: 431–438. Feng, D.-F., and Doolittle, R.F. (1997) Converting amino acid alignment scores into measures of evolutionary time: a simulation study of various relationships. J Mol Evol 44: 361–370. Fenton, W.A., Kashi, Y., Furtak, K., and Horwich, A.L. (1994) Residues in chaperonin GroEL required for polypeptide binding and release. Nature 371: 614–619. Forterre, P., Brochier, C., and Philippe, H. (2002) Evolution of the Archaea. Theor Popul Biol 61: 409–422. Furutani, M., Iida, T., Yoshida, T., and Maruyama, T. (1998) Group II chaperonin in a thermophilic methanogen, Methanococcus thermolithotrophicus. J Biol Chem 273: 28399– 28407. Galagan, J.E., Nusbaum, C., Roy, A., Endrizzi, M.G., Macdonald, P., FitzHugh, W., et al. (2002) The genome of M. acitivorans reveals extensive metabolic and physiological diversity. Genome Res 12: 532–542. Georgellis, D., Sohlberg, B., Hartl, F.-U., and Gabain, A.V. (1995) Identification of GroEL as a constituent of an mRNA-protection complex in Escherichia coli. Mol Microbiol 16: 1259–1268. © 2003 US Government, Molecular Microbiology, 48, 143–156
Gething, M.-J. (ed.) (1997) Guidebook to Molecular Chaperones and Protein-Folding Catalysts. Oxford: Oxford University Press. Grogan, D.W. (1989) Phenotypic characterization of the archaebacterial genus Sulfolobus: comparison of five wildtype strains. J Bacteriol 171: 6710–6719. Guagliardi, A., Cerchia, L., Bartolucci, S., and Rossi, M. (1994) The chaperonin from the archaeon Sulfolobus solfataricus promotes correct refolding and prevents thermal denaturation in vitro. Protein Sci 3: 1436–1443. Harris, J.R., Zahn, R., and Plückthun, A. (1995) Electron microscopy of the GroEL-GroES filament. J Structural Biol 115: 68–77. Hartl, F.U., and Hayer-Hartl, M. (2002) Molecular chaperones in the cytosol: from nascent chain to folded protein. Science 295: 1852–1858. Hightower, L.E. (1991) Heat shock, stress proteins, chaperones, and proteotoxity. Cell 69: 191–197. Horwich, A.L., and Willison, K.R. (1993) Protein folding in the cell: functions of two families of molecular chaperones, Hsp 60 and TF55-TCP1. Phil Trans R Soc London 339: 313– 326. Izumi, M., Fujiwara, S., Takagi, M., Fukui, K., and Imanaka, T. (2001) Two kinds of archaeal chaperonin with different temperature dependency from a hyperthermophile. Biochem Biophys Res Commun 280: 581–587. Kagawa, H.K., Osipiuk, J., Maltsev, N., Overbeek, R., QuaiteRandall, E., Joachimiak, A., and Trent, J.D. (1995) The 60 kDa heat shock proteins in the hyperthermophilic archaeon Sulfolobus shibatae. J Mol Biol 253: 712–725. Karlin, S., and Brocchieri, L. (2000) Heat shock protein 60 sequence comparisons: duplications, later transfer, and mitochondrial evolution. Proc Natl Acad Sci USA 97: 11348–11353. Karlin, S., Mrázek, J., Campbell, A., and Kaiser, D. (2001) Characterization of highly expressed genes of four fast growing bacteria. J Bacteriol 183: 5025–5040. Klumpp, M., and Baumeister, W. (1998) The thermosome: archetype of group II Chaperonins. FEBS Lett 430: 73–77. Knapp, S., Schmidt-Krey, I., Hebert, H., Bergman, T., Jörnwall, H., and Ladenstein, R. (1994) The molecular chaperonin TF55 from the thermophilic archaeon, Sulfolobus solfataricus. J Mol Biol 242: 397–407. Lin, P., and Sherman, F. (1997) The unique hetero-oligomeric nature of the subunits in the catalytic cooperativity of the yeast Cct chaperonin complex. Proc Natl Acad Sci USA 94: 10780–10785. Ma, J., Campbell, A., and Karlin, S. (2002) Correlations between Shine-Dalgarno sequences and predicted gene expression levels and operon structrues. J Bacteriol 184: 5733–5745. Marco, S., Ureña, D., Carrascosa, J.L., Waldmann, T., Peters, J., Hegerl, R., et al. (1994) The molecular chaperone TF55: assesment of symmetry. FEBS 341: 152–155. Nei, M., and Kumar, S. (2000) Molecular Evolution and Phylogenetics. Oxford: Oxford University Press. Ochman, H., Gerber, A.S., and Hartl, D.L. (1988) Genetic applications of an inverse polymerase chain reaction. Genetics 120: 621–623. Ota, T., and Nei, M. (1994) Estimation of the number of amino
156 H. K. Kagawa et al. acid substitutions per site when the substitution rate varies among sites. J Mol Evol 38: 642–643. Peeples, T.L., and Kelly, R.M. (1995) Bioenergetic response of the extreme thermoacidophile Metallosphaera sedula to thermal and nutritional stresses. Appl Environ Microbiol 61: 2314–2321. Quaite-Randall, E., Trent, J.D., Josephs, R., and Joachimiak, A. (1995) Conformational cycle of the archaeosome, a TCP1-like chaperonin from Sulfolobus shibatae. J Biol Chem 270: 28818–28823. Reiter, W.-D., Hüdepohl, U., and Zillig, W. (1990) Mutational analysis of an archaebacterial promoter: essential role of a TATA box for transcription efficiency and start-site selection in vitro. Proc Natl Acad Sci USA 87: 9509–9513. Ruggero, D., Ciammaruconi, A., and Londei, P. (1998) The chaperonin of the archaeon Sulfolobus solfataricus is an RNA binding protein that participates in ribosomal RNA processing. EMBO J 17: 3471–3477. Saibil, H. (2000) Molecular chaperones: containers and surfaces for folding, stabilising or unfolding proteins. Curr Opin Struct Biol 10: 251–258. Saitou, N., and Nei, M. (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4: 406–425. Sanders, B.M. (1993) Stress proteins in aquatic organisms: an enviromental perspective. Crit Rev Toxicol 23: 49–75. Sanger, F., Nicklen, S., and Coulson, A.R. (1977) DNA sequence analysis with chain terminating inhibitors. Proc Natl Acad Sci USA 74: 5463–5467. Schoehn, G., Hayes, M., Cliff, M., Clarke, A.R., and Saibil, H. (2000a) Domain Rotations between open, closed and bullet shaped forms of the thermosome, an archaeal chaperonin. J Mol Biol 301: 323–332. Schoehn, G., Quaite-Randall, E., Jimenez, J.L., Joachimiak, A., and Saibil, H. (2000b) Three conformations of an archaeal chaperonin, TF55 from Sulfolobus shibatae. J Mol Biol 296: 813–819. Storti, R.V., and Rich, A. (1976) Chick cytoplasmic actin and muscle actin have different structural genes. Proc Natl Acad Sci USA 73: 2346–2350. Strimmer, K., and von Haeseler, A. (1996) Quartet puzzling: a quartet maximum-likelihood method for reconstructing tree topologies. Mol Biol Evol 13: 964–969.
Török, Z., Horvath, I., Goloubinoff, P., Kovacs, E., Glatz, A., Balogh, G., and Vigh, L. (1997) Evidence for a lipochaperonin: association of active protein-folding GroESL oligomers with lopids can stabilize membranes under heat shock conditions. Proc Natl Acad Sci USA 94: 2192– 2197. Trent, J.D. (1996) A review of acquired thermotolerance, heat shock proteins, and molecular chaperones in archaea. FEMS Microbiol Rev 18: 249–258. Trent, J.D. (2000) Extremophiles in Astrobiology: Per Ardua ad Astra. Gravitational Space Biol Bull 13: 5–11. Trent, J.D., Nimmesgern, E., Wall, J.S., Hartl, F.-U., and Horwich, A.L. (1991) A molecular chaperone from a thermophilic archaebacterium is related to the eukaryotic protein t-complex polypeptide-1. Nature 354: 490–493. Trent, J.D., Kagawa, H.K., Yaoi, T., Olle, E., and Zaluzec, N.J. (1997) Chaperonin filaments: the archaeal cytoskeleton? Proc Natl Acad Sci USA 94: 5383–5388. Trent, J.D., Kagawa, H.K., and Yaoi, T. (1998) The role of chaperonins in vivo: the next frontier. Ann NY Acad Sci 841: 42–54. Waldmann, T., Nitsch, M., Klumpp, M., and Baumeister, W. (1995) Expression of an archael chaperonin in E. coli formation of homo - (alpha, beta) and hetero-oligomeric (alpha + beta) thermosome complexes. Fed Eur Biochem Sci 376: 67–73. Yaoi, T., Kagawa, H.K., and Trent, J.D. (1998) Chaperonin filaments: their formation and an evaluation of methods for studying them. Arch Biochem Biophys 365: 55–62. Yoshida, T., Yohda, M., Iida, T., Maruyama, T., Taguchi, H., Yazaki, K., et al. (1997) Structural and functional characterization of homo-oligomeric complexes of alpha and beta chaperonin subunits from the hyperthermophilic Archaeum Thermococcus strain KS-1. J Mol Biol 273: 635–645. Yoshida, T., Ideno, A., Suzuki, R., Yohda, M., and Maruyama, T. (2002) Two kinds of archaeal group II chaperonin subunits with different thermostability in Thermococcus strain KS-1. Mol Microbiol 44: 761–768. Zillig, W., Palm, P., Klenk, H.-P., Langer, D., Hüdepohl, U., Hain, J., et al. (1993) Transcription in Archaea. In The Biochemistry of Archaea (Archaebacteria), Vol. 26. Kates, M., Kushner, D.J., and Matheson, A.T. (eds). Amsterdam: Elsevier Science Publishers, pp. 367–392.
© 2003 US Government, Molecular Microbiology, 48, 143–156
