Differential Regulation of the PanA and PanB Proteasome-Activating Nucleotidase and 20S Proteasomal Proteins of the Haloarchaeon Haloferax volcanii

Abstract

The halophilic archaeon Haloferax volcanii produces three different proteins (α1, α2, and β) that assemble into at least two 20S proteasome isoforms. This work reports the cloning and sequencing of two H. volcanii proteasome-activating nucleotidase (PAN) genes (panA and panB). The deduced PAN proteins were 60% identical with Walker A and B motifs and a second region of homology typical of AAA ATPases. The most significant region of divergence was the N terminus predicted to adopt a coiled-coil conformation involved in substrate recognition. Of the five proteasomal proteins, the α1, β, and PanA proteins were the most abundant. Differential regulation of all five genes was observed, with a four- to eightfold increase in mRNA levels as cells entered stationary phase. In parallel with this mRNA increase, the protein levels of PanB and α2 increased severalfold during the transition from exponential growth to stationary phase, suggesting that these protein levels are regulated at least in part by mechanisms that control transcript levels. In contrast, the β and PanA protein levels remained relatively constant, while the α1 protein levels exhibited only a modest increase. This lack of correlation between the mRNA and protein levels for α1, β, and PanA suggests posttranscriptional mechanisms are involved in regulating the levels of these major proteasomal proteins. Together these results support a model in which the cell regulates the ratio of the different 20S proteasome and PAN proteins to modulate the structure and ultimately the function of this central energy-dependent proteolytic system.

Proteasomes are large nanocompartments within the cell that catalyze the energy-dependent hydrolysis of proteins. The catalytic core particle (CP) responsible for peptide bond hydrolysis is a 20S cylinder of four stacked heptameric rings which is ubiquitous in archaea, eucaryotes, and actinomycetes (19, 22). The outer and inner rings of the CP are formed by proteins of related α and β superfamilies, respectively (6). A variety of components associate with the CP; most notable are the 19S cap regulatory particles (RP), which together with the CP form eucaryal 26S proteasomes. The RP of yeast can be separated into two multisubunit substructures: the lid and base domains. The base includes six Rpt (regulatory particle triple-A type I ATPase) subunits and not only exhibits chaperone activity (4) but also facilitates the energy-dependent degradation of globular proteins by 20S proteasomes (10). These Rpt subunits are closely related to archaeal proteasome-activating nucleotidase (PAN) proteins, of which only the Methanocaldococcus jannaschii Pan purified from recombinant Escherichia coli has been characterized (2, 38, 44). Substrate binding to this AAA ATPase activates the hydrolysis of nucleotides (e.g., ATP), which subsequently promotes the unfolding of substrates, opening of the 20S proteasome axial gate(s), and apparent translocation of substrate to the internal proteolytic chamber of 20S proteasomes.

Multicellular organisms synthesize CP and RP proteasomal subtypes by integrating distinct isoforms of one or more subunit. A classic example of this is the immunoproteasome of vertebrates, in which the gamma interferon-inducible isoforms of β1, β2, and β5 (designated as β1i, β2i, and β5i) are incorporated into CP subtypes (16, 29). Although not as well defined, expression of insect (20, 42) and plant (24) proteasomal subunit isoforms has also been detected at specific developmental stages and in response to external stimuli. Furthermore, recent advances in proteomics have demonstrated that at least 15 Arabidopsis CP and RP isoforms are incorporated into 26S proteasomes (41). Thus, in multicellular eucarya there are a variety of proteasome subtypes available for decoration by differential posttranslational modification (26) to provide an array of flexibility and diversity of function.

In contrast to the various proteasomal subtypes synthesized by multicellular organisms, most microorganisms encode a set of proteins that assemble into single CP and AAA ATPase RP (i.e., Rpt, PAN, and ARC) complexes (22). Although yeast synthesize seven α-type and seven β-type proteins, these different subunits assemble into a single 20S proteasome (12). Likewise, the six Rpt subunits of yeast are proposed to assemble into a single hexameric ring which forms the base of the RP (4). The two α-type and two β-type proteins of the actinomycetes Rhodococcus erythropolis also assemble into a single 20S proteasome (43). Recent studies, however, demonstrate that the haloarchaeon Haloferax volcanii encodes three proteins (α1, α2, and β) (37) that assemble into at least two 20S proteasome (CP) subtypes (14). One of these is a symmetric complex of α1 and β subunits (α1β CP), while the other (α1α2β CP) appears asymmetric with homogenous rings of α1 and α2 at each end. In addition, genome sequences reveal that other archaea encode one α and two β proteins which may form multiple 20S proteasome subtypes (22).

The demonstration that H. volcanii synthesizes more than one 20S proteasome subtype raises the possibility that select archaea may also synthesize more than one PAN regulatory protein. In this study, we demonstrate that H. volcanii synthesizes two PAN homologs (PanA and PanB), which are differentially regulated at the mRNA and protein levels in a manner coordinated with the CP isoforms. Although the transcripts of all five proteasomal components increased in parallel as cells entered stationary phase, at the protein level only PanB was coordinated with α2, while the levels of PanA were relatively constant and more consistent with α1 and β. These results suggest that differences in the levels of CP and RP proteins are involved in mediating the transition to stationary phase. In addition, these results suggest subunit isoforms form different proteasomal subtypes throughout growth to modulate energy-dependent proteolysis in H. volcanii.

MATERIALS AND METHODS

Materials.

Biochemicals and amplification-grade DNase I were from Sigma-Aldrich (St. Louis, Mo.). Other organic and inorganic analytical-grade chemicals were from Fisher Scientific (Atlanta, Ga.). Hybond-P membranes were from Amersham Pharmacia Biotech (Piscataway, N.J.). Desalted oligonucleotides were from Operon Technologies (Alameda, Calif.) and Integrated DNA Technologies (Coralville, Iowa). Digoxigenin-11-dUTP (2′-deoxyuridine-5′-triphosphate coupled by an 11-atom spacer to digoxigenin), alkaline phosphatase-conjugated antibody raised against digoxigenin, and nylon membranes for colony and plaque hybridizations were from Roche Molecular Biochemicals (Indianapolis, Ind.). Positively charged membranes for Southern hybridization were from Ambion (Austin, Tex.).

Strains, media, and plasmids.

Strains, oligonucleotide primers used for cloning, and plasmids are summarized in Table 1 or were as described previously (14). E. coli DH5α was used for routine recombinant DNA experiments. E. coli ER1647 was used to generate subgenomic libraries of H. volcanii chromosomal DNA. E. coli BL21(DE3) was used as a host for expression and purification of H. volcanii proteins. E. coli strains were grown at 37°C (200 rpm) in Luria-Bertani medium. H. volcanii strains were grown at 42°C (200 rpm) in complex medium ATCC 974. For analysis of mRNA and protein levels, H. volcanii WFD11 was selected to allow for future comparison to strains expressing genes from plasmids. WFD11 was grown to exponential phase (optical density at 600 nm [OD₆₀₀], <0.2) from an isolated colony, subcultured (1% [vol/vol]) to fresh medium, and grown to exponential phase (OD₆₀₀ of 0.080). This exponential-phase culture was used as a 1% (vol/vol) inoculum into fresh medium for the analysis.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Phenotype, genotype, or oligonucleotides for PCR amplification	Source
E. coli strains
DH5α	F−recA1 endA1 hsdR17(rk− mk+) supE44 thi-1 gyrA relA1	Life Technologies
BL21(DE3)	F−ompT [Ion] hsd SB (rB mB) (an E. coli B strain) with DE3, a λ prophage carrying the T7 RNA polymerase gene	Novagen
ER1647	F−fhuA2Δ(lacZ) r1 supE44 trp31 mcrA1272::Tn10(Tetr) his-1 rpsL104 (Strr)xyl-7 mtl-2 metB1 Δ(mcrC-mrr)102::Tn10(Tetr) recD1040	Novagen
H. volcanii strains
DS2	Isolate from the Dead Sea	23
DS70	DS2 cured of pHV2	36
WFD11	DS2 cured of pHV2	5
Plasmids
pUC19	Apr; cloning vector	New England BioLabs
pCR-BluntII-TOPO	Kmr; cloning vector	InVitrogen
pET24b	Kmr; expression vector	Novagen
pJAM806	Apr; 0.36-kb PCR fragment of H. volcanii genomic DNA blunt-end ligated into HincII site of pUC19; carries pan-specific probe used in Southern blotting	This study
pJAM624	Apr; 2.7-kb RsrII fragment of H. volcanii genomic DNA ligated into SmaI site of pUC19; carries panB′	This study
pJAM626	Apr; 6.2-kb SmaI fragment of H. volcanii genomic DNA ligated into HincII site of pUC19; carries panA	This study
pJAM631	Apr; 2.2-kb SacII genomic fragment from H. volcanii blunt-end ligated into HincII site of pUC19; carries panA	This study
pJAM638	Kmr; 1.2-kb fragment generated by PCR amplification from pJAM631 cloned into pCR-BluntII-TOPO; carries panA without a stop codon; 5′-CATATGATGACCG ATACTGTGGAC-3′ and 5′-AGGCTTAGCAAACGCGCGGGAGAC-3′ (NdeI and HindIII sites in bold)
pJAM639	Kmr; 1.1-kb fragment generated by PCR amplification from pJAM631 cloned into pCR-BluntII-TOPO; carries panAΔ120 without a stop codon; 5′-CATATGCGC GACAAGCTCCTCGAC-3′ and 5′-AGGCTTAGCAAACGCGCGGGAGAC-3′ (NdeI and HindIII sites in bold)	This study
pJAM642	Kmr; 1.2-kb NdeI-to-HindIII fragment of pJAM638 ligated into NdeI and HindIII sites of pET24b; PanA-His expressed in E. coli	This study
pJAM643	Kmr; 1.1-kb NdeI-to-HindIII fragment of pJAM639 ligated into NdeI and HindIII sites of pET24b; PanA(Δ1-40)-His expressed in E. coli	This study
pJAM1006	Kmr; 1.24-kb fragment PCR amplified from H. volcanii genomic DNA ligated into NdeI and XhoI sites of pET24b; PanB-His expressed in E. coli; 5′-CATATGTCA CGCAGTCCATCTCTCC-3′ and 5′-CTCGAGGTACTGGTAGTCCGTGAAG-3′ (NdeI and XhoI sites indicated in bold)	This study

DNA isolation and analysis.

Genomic DNA was isolated from H. volcanii DS2 as described previously (37). Plasmid DNA was isolated from E. coli strains using the QIAprep Spin Miniprep kit (QIAGEN, Valencia, Calif.). DNA fragments were isolated from 0.8% (wt/vol) SeaKem GTG agarose (FMC Bioproducts, Rockland, Maine) gels in 1× TAE electrophoresis buffer (40 mM Tris acetate, 2 mM EDTA; pH 8.5) using the QIAquick gel extraction kit (QIAGEN). Fidelity of PCR-amplified products was confirmed by DNA sequencing. Both strands of DNA were sequenced by the dideoxy chain termination method (27) with a LI-COR automated DNA sequencer (DNA Sequencing Facility, Department of Microbiology and Cell Science, University of Florida).

A DNA probe specific for H. volcanii genes encoding PAN homologs was generated using the degenerate oligonucleotides 5′-TAY GGN CCN CCN GGN CAN GGN AAR AC-3′ and 5′-AA NGK NCC NGG NGK NAR NAD NGC NGG-3′ (where R is A + G; Y is C + T; K is G + T; D is G, A + T; N is A + G + C + T). These oligonucleotides served as primers to amplify a 0.36-kb fragment from H. volcanii genomic DNA by PCR using Vent DNA polymerase. The PCR fragment was blunt-end ligated into the HincII site of pUC19 to generate plasmid pJAM806 (Table 1) and sequenced to confirm that the cloned DNA was pan-specific. The 0.36-kb XbaI-to-PstI fragment of pJAM806 was isolated and random prime labeled with digoxigenin-11-dUTP as previously described (37) to generate a pan-specific probe.

For Southern analysis, H. volcanii genomic DNA (2.5 μg per lane) was cleaved by a variety of restriction enzymes, including SmaI, SacII and RsrII. DNA fragments were separated by agarose gel electrophoresis, transferred to nylon membranes, and hybridized as previously described (37) with the following modifications. Membranes were equilibrated at 68°C for 4 h, after which the pan-specific probe was added to a final concentration of 25 ng per ml. After 14 h at 68°C, the membranes were washed twice with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and 0.1% (wt/vol) sodium dodecyl sulfate (SDS) (5 min per wash; 25°C) and twice with 0.5× SSC and 0.1% (wt/vol) SDS (15 min per wash; 68°C). Based on these Southern blotting data, RsrII (2 to 4 kb), SmaI (5 to 7 kb), and SacII (1.5 to 3 kb) subgenomic libraries were generated in plasmid pUC19. Plasmids carrying the panA gene (pJAM626 and pJAM631) and partial panB gene (pJAM624) (Table 1) were isolated from the subgenomic libraries by hybridization to the 0.36-kb pan-specific probe.

PCR was used to introduce restriction enzyme sites for directional cloning of panA, panAΔ120, and panB into plasmid pET24b for expression in E. coli (pJAM642, pJAM643, and pJAM1006) (Table 1). The start codons of these genes were positioned 8 bp downstream of the ribosome binding sequence of pET24b by using NdeI. For epitope tagging, the 3′ end of the gene was modified by PCR to remove the stop codon and provide an in-frame C-terminal addition of poly-histidine residues.

RNA isolation and analysis.

Total RNA was isolated from H. volcanii WFD11 with the RNeasy Mini kit for bacteria (QIAGEN) with the following modifications. Cells were centrifuged for 1 min at 18,000 × g and immediately resuspended in 100 μl of RNase-free deionized water. RNA was treated with DNase I according to the supplier (Sigma), and removal of DNA was confirmed by PCR. Quality of RNA was determined by 0.8% (wt/vol) agarose gel electrophoresis in 1× TAE buffer. One-step quantitative reverse transcription-PCR (RT-PCR) was performed using the Quantitect SYBR Green RT-PCR kit (QIAGEN) and iCycler MyiQ real-time PCR detection system (Bio-Rad). Quantitative RT-PCR was performed on a dilution series of total RNA using the oligonucleotide primers listed in Table 2. The total RNA concentration was determined by measuring absorbance at 260 nm. Reference standards used to determine the quantity of specific transcript included psmC-specific mRNA and plasmid DNA cleaved by HindIII which was specific for psmA (pJAM618), psmB (pJAM621), panA (pJAM642), and panB (pJAM1006) (14) (Table 1). For these latter genes, the PCR amplification efficiency for each dilution series of total RNA was equivalent to the amplification efficiency for the corresponding DNA standard. Thus, RT was not limiting and allowed the use of DNA standards for RT-PCR analysis. The psmC-specific mRNA standard was generated using the MAXIscript in vitro transcription kit according to the supplier's protocol (Ambion). Plasmid DNA carrying the psmC gene (pJAM619) (14) was linearized with HindIII, extracted with Tris-saturated phenol (pH 8.0) and chloroform (1:1), and precipitated with ethanol to remove residual RNases prior to in vitro transcription. RT-PCR included an RT step (50°C, 30 min) and a reverse transcriptase inactivation and DNA polymerase activation step (95°C, 15 min). This was followed by 40 cycles of denaturation at 94°C (15 s), annealing at 57.5°C (30 s), and extension at 72°C (30 s). Data were analyzed using the MyiQ single-color real-time PCR detection system software, version 1.0 (Bio-Rad). All RT-PCR and PCR amplification efficiencies were 90 to 100%, and melt curve analysis was performed to verify the uniformity of product.

TABLE 2.

Oligonucleotide primers used in quantitative RT-PCR

Primer	Direction	Sequence
α1	Forward	5′-CGGCGACCACCAAGAGCACC-3′
	Reverse	5′-GTCGTCCTCGGTCGCCAACAG-3′
α2	Forward	5′-CCCCGACGGACGCATCTATCA-3′
	Reverse	5′-GCGCGTCGTCGAGCTTGTGC-3′
β	Forward	5′-GCTCCGGCAGCCAGTACG-3′
	Reverse	5′-GTTCTGGTGGCGCTGGATG-3′
PanA	Forward	5′-GCAGGAGATCACCGCCCT-3′
	Reverse	5′-GGTGATTTCCTGAACCGTCGCAA-3′
PanB	Forward	5′-ACCGCCACGCCGAGTTAGTC-3′
	Reverse	5′-TGCCGTGCTGCTTGATGATGAC-3′

Protein purification and electrophoresis.

H. volcanii proteins were expressed from plasmids pJAM643 (PanAΔ_1-40-His), pJAM642 (PanA-His), and pJAM1006 (PanB-His) in recombinant E. coli BL21(DE3) as previously described (21). Cells were lysed in six volumes (wet weight/vol) of 20 mM Tris buffer, pH 7.2, containing 150 mM NaCl and 10 mM imidazole by passage through a French pressure cell at 20,000 lb/in². Cell lysate was filtered (0.22 μm) and applied to a Ni²⁺-Sepharose column (4.8 by 0.8 cm; Pharmacia) equilibrated in lysis buffer. The column was washed with 20 mM Tris buffer, pH 7.2, containing 2 M NaCl and 60 mM imidazole (10 ml). Protein was eluted in 20 mM Tris buffer, pH 7.2, containing 2 M NaCl and 500 mM imidazole.

Protein concentrations were determined by the Coomassie blue dye binding assay with bovine serum albumin as the standard (Bio-Rad) (3). Molecular masses of proteins were estimated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (17) using the standards phosphorylase b (97.4 kDa), bovine serum albumin (66.2 kDa), ovalbumin (45 kDa), carbonic anhydrase (31 kDa), soybean trypsin inhibitor (21.5 kDa), and lysozyme (14.4 kDa) (Bio-Rad).

Antibody preparation and immunoanalysis.

Prior to antibody preparation, the N-terminal sequences of PanA_Δ1-40-His and PanA-His were determined by Edman degradation, and the mass spectra of the tryptic fragments of PanB-His were determined by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (Applied Biosystems QSTAR XL hybrid LC/MS/MS; University of Florida ICBR Protein Chemistry Core). PanAΔ_1-40-His and PanB-His proteins purified from recombinant E. coli were separated by SDS-7.5% PAGE, excised from the gel, and used as antigen to raise polyclonal antibodies in rabbit (Cocalico Biologicals, Reamstown, Pa.). For immunoblotting, cells were harvested by centrifugation (5,000 × g; 10 min at 4°C) and stored at −70°C. Cell pellets were resuspended in 1 ml of deionized water and passed through a French pressure cell at 10,000 lb/in². Proteins were separated by SDS-7.5% and 12% PAGE and electroblotted onto Hybond-P hydrophobic polyvinylidene difluoride membranes (Amersham Biosciences) (14.5 h; 20 V at 4°C). Antigens were detected using primary polyclonal antibodies (α1, 1:20,000; α2, 1:2,000; β, 1:10,000; PanA, 1:3,000; PanB, 1:10,000) (14) and secondary horseradish peroxidase-conjugated anti-rabbit antibody raised in goats (Southern Biotechnology Associates, Inc., Birmingham, Ala.) and chemiluminescence with ECL Plus according to the supplier's protocol (Amersham Biosciences). Dilutions of the secondary antibody used with each of the 20S proteasome and PAN proteins were as follows: α1, 1:10,000; α2, 1:1,000; β, 1:10,000; PanA, 1:3,000; PanB, 1:10,000. Reference standards included βΔ-His, α1-His, and α2-His purified from recombinant E. coli as previously described (14) and PanA-His and PanB-His purified from recombinant E. coli as described above. For quantitative immunoblotting, the concentrations of antibodies and cell lysate were optimized to ensure the chemiluminescence signal was within the linear range of the standard curve. Chemiluminescent signal was detected by using a charge-coupled device camera (Versa Doc; Bio-Rad) and X-ray film (Hyperfilm; Amersham Biosciences). The levels of proteasomal proteins were determined using the Quantity One software package (Bio-Rad).

DNA and protein sequence analyses.

Clone Manager professional suite version 6.0 (Scientific & Educational Software, Durham, N.C.) and BioEdit sequence alignment editor version 5.0.9 (13) were used for DNA and protein sequence analyses. Deduced amino acid sequences were compared to sequences available in the GenBank, EMBL, and SWISS-PROT databases at the National Center for Biotechnology Information (Bethesda, Md.) by using the BLAST network server (1). For cluster analysis, protein sequences were aligned with Clustal W version 1.4 (30) using the neighbor-joining method set with a bootstrap value of 1,000. Amino acid extensions at N and C termini and insertions were removed. These core protein sequences were used to compute a distance matrix using the default settings of the Fitch-Margoliash and least-squares distance methods with PROTDIST version 3.5c. The TreeView program (25) was used to display these results as an unrooted dendrogram. The probability that a deduced protein sequence would adopt a coiled-coil conformation was predicted using COILS with weighted and unweighted MTK and MTIKL scoring matrices set to scanning windows of 21 and 28 residues (18).

Nucleotide sequence accession number.

The nucleotide sequences of the panA and panB genes of H. volcanii DS2 were assigned GenBank accession numbers AY627303 and AY627304.

RESULTS AND DISCUSSION

PAN genes (panA and panB) of H. volcanii.

To date, studies on archaeal PAN proteins have been limited to the biochemistry of a single enzyme (M. jannaschii Pan) purified from recombinant E. coli (38, 44). To further understand the function of haloarchaeal PAN proteins, the pan gene was targeted for isolation from H. volcanii, one of the few archaea with an established genetic exchange system. Two degenerate oligonucleotides were synthesized based on the conserved regions YGPPGTG and FRGPR(F/L)(F/L/I)AP of archaeal PAN homologs and the codon bias of H. volcanii (34). Using these oligonucleotides, a pan probe was generated by PCR and used for DNA hybridization to H. volcanii genomic DNA. This enabled the isolation of two genes (panA and panB) encoding PAN homologs from subgenomic libraries of H. volcanii. The panA gene encoded a 406-amino-acid protein of 45,664.31 Da and a calculated pI of 4.46 (Fig. 1). The deduced protein sequence was found to be identical to open reading frame RVO01394 and enabled closure of contigs 06262 and 06170 of the partial genome sequence of H. volcanii (http://zdna2.umbi.umd.edu/). In contrast to panA, the panB gene was not predicted by partial genome sequencing. The DNA sequence of this report enabled closure of contigs 06607 and 06375 of the partial genome sequence and completion of the entire panB gene. The panB gene was found to encode a protein of 412 amino acids (45,410.39 Da) and a calculated pI of 4.32. Based on these results combined with our group's previous work (37), all five proteasomal genes, including those encoding the α1, α2, and β subunits of the two 20S proteasome subtypes, are dispersed throughout the H. volcanii genome.

FIG. 1.

Amino acid sequence alignment of H. volcanii PanA and PanB. Identical residues are shaded in black. Functionally conserved and semiconserved amino acid residues are shaded in grey. Dashes indicate gaps introduced in the protein sequence alignment. Boxed residues were predicted with a >90% probability to form a coiled-coil conformation (see Materials and Methods). Consensus sequences of the Walker A and B boxes of the P-loop nucleotidase core are indicated below the alignment as GX₂GXGKT and DEXD, respectively (where X is any amino acid residue). The AAA ATPase second region of homology or SHR motif [(T/S)-(N/S)-X₅-DXA-X₂-R-X₂-RX-(D/E)] is also indicated. The N-terminal sequences of the PanA-His and PanA_Δ1-40-His antigens were identical to residues 2 to 14 and 41 to 51 of the deduced primary sequence, respectively. MALDI-TOF Q-STAR detected the mass spectra of 11 tryptic fragments of the PanB-His antigen, which encompassed 36% of the primary amino acid sequence. The masses (in daltons) and corresponding residue numbers of PanB-His were as follows: 716.3148, 30 to 34; 749.3711, 35 to 40; 1,004.5522, 330 to 337; 1,264.6408, 205 to 261; 1,307.7720, 317 to 328; 1,490.6807, 379 to 392; 1,548.7975, 301 to 313; 1,836.7691, 275 to 289; 2,180.0845, 118 to 137; 2,408.1564, 4 to 24; 2,739.3637, 138 to 162.

The deduced PanA and PanB protein sequences are closely related, with 60% identical and 78% similar amino acid residues (Fig. 1). The most significant region of identity was the central P-loop nucleotidase domain, which includes Walker A and B boxes, as well as the second region of homology typical of AAA ATPases (9). N-terminal residues of PanA and PanB (position 17 to 72 and 39 to 73, respectively) are predicted to adopt a coiled-coil conformation. In contrast to the central domain, however, the N-terminal coiled-coil domains were not highly related (17% identity) and differed in length by 21 residues (three helical turns). Likewise, C-terminal residues 324 to 406 of PanA were only 49% identical to residues 322 to 412 of PanB. This divergence in the primary sequence of the N and C termini, combined with the highly conserved central domain, suggests that PanA and PanB mediate similar yet distinct functions in H. volcanii. The N-terminal coiled-coil region of PAN/Rpt has been proposed to be involved in interactions with other RP subunits and/or substrate proteins (11). Furthermore, modification of the C terminus of Rpt1 (i.e., addition of a Flag-His₆ epitope tag in Saccharomyces cerevisiae) inhibits RP association with the CP in yeast (31). Thus, based on differences in the N and C termini, it is possible that the PanA and PanB proteins of H. volcanii recognize different substrate proteins and/or differ in their affinities for α1β and α1α2β 20S proteasome subtypes.

Cluster analysis was used to compare the H. volcanii PanA and PanB proteins to other archaeal PAN and eucaryal Rpt homologs (Fig. 2). The comparison revealed that PAN and Rpt proteins are related in primary sequence yet cluster into two major groups based on archaeal versus eucaryal origin. All of the eucaryal Rpt proteins were found to cluster into six subgroups (Rpt1 to -6), even when the genes were from multicellular organisms that encode more than six Rpt genes (i.e., humans, Caenorhabditis elegans, Arabidopsis thaliana, and Drosophila melanogaster). This subgrouping is consistent with biochemical evidence that suggests six different Rpt proteins assemble as a hexameric ring to form the RP interface which associates with the CP (35). The high level of sequence similarity among the PAN/Rpt proteins coupled with the species distribution of these genes has led to the suggestion that eucaryal Rpt genes originated by duplication and divergence from an ancestral proteasome-associated ATPase (32, 40). The few archaea that do not encode PAN proteins (i.e., Thermoplasma acidophilum, Thermoplasma volcanium, and Pyrobaculum aerophilum) likely encountered loss of this gene during evolution.

FIG. 2.

Dendrogram summarizing the relationships between archaeal PAN and eucaryal Rpt proteins. Cluster analysis was performed as described in Materials and Methods. The scale bar represents 0.1 nucleotide substitutions per site. Organism abbreviations: Hs, Homo sapiens; Ce, C. elegans; Dm, D. melanogaster; At, A. thaliana; Sc, S. cerevisiae; Gl, Giardia lamblia; Af, Archaeoglobus fulgidus; Hm, Haloarcula marismortui; Hb, Halobacterium sp. strain NRC-1; Hv, H. volcanii; Mj, M. jannaschii; Mk, Methanopyrus kandleri; Ma, Methanosarcina acetivorans; Mb, Methanosarcina barkeri; Mm, Methanosarcina mazei; Mt, Methanothermobacter thermoautotrophicus; Pa, Pyrococcus abysii; Pf, Pyrococcus furiosus; Ph, Pyrococcus horikoshii; Ap, Aeropyrum pernix; Ss, Sulfolobus solfataricus; Ne, Nanoarchaeum equitans. GenBank, EMBL, SwissProt, or locus tag accession numbers (in parentheses) were as follows: HsRpt1 (P35998), HsRpt2 (Q03527), HsRpt3 (P43686), HsRpt4 (Q92524), HsRpt5a (AAB24840), HsRpt5b (P17980), HsRpt6 (P47210), CeRpt1 (T20152), CeRpt2a (T31800), CeRpt2b (AAB42248), CeRpt3 (A88485), CeRpt4 (AAB70326), CeRpt5 (T33155), CeRpt6a (AAM48537), CeRpt6b (T27048), DmRpt1 (AAF59219), DmRpt2 (AAB34134), DmRpt3a (AAF48001), DmRpt3b (AAF54440), DmRpt4a (AAR46146), DmRpt4b (AAF08391), DmRpt4c (AAF49987), DmRpt5a (AAF56177), DmRpt5b (AAD24194), DmRpt6a (AAC63219), DmRpt6b (AAF57069), AtRpt1a (At1g53750), AtRpt1b (At1g53780), AtRpt2a (At4g29040), AtRpt2b (At2g20140), AtRpt3 (At5g58290), AtRpt4a (At1g45000), AtRpt4b (At5g43010), AtRpt5a (At3g05530), AtRpt5b (At1g09100), AtRpt6a (At5g19990), AtRpt6b (At5g20000), ScRpt1 (YKL145w), ScRpt2 (YDL007w), ScRpt3 (YDR394w), ScRpt4 (YOR259c), ScRpt5 (YOR117w), ScRpt6 (YGL048c), GlRpt1 (EAA40208), GlRpt2 (EAA39033), GlRpt3 (EAA42904), GlRpt4 (EAA41176), GlRpt5 (EAA42075), GlRpt6 (EAA42208), AfPan (AF1976), HmPanA (Contig93_11146_12363), HmPanB (Contig169_4249_3029), HbPanA (VNG0510g), HbPanB (VNG2000g), HvPanA (AY627303), HvPanB (AY627304), MjPan (MJ1176), MkPan (MK0878), MaPanA (MA4268), MaPanB (MA4123), MbPanA (Meth3002), MbPanB (Meth2182), MmPanA (MM1006), MmPanB (MM0798), MtPan (MTH728), PaPan (PAB2233), PfPan (PF0115), PhPan (PH0201), ApPan (APE2012), SsPan (SSO0271), NePan (NEQ186).

Of the archaea, only the haloarchaea and methanosarcina were found to encode more than one PAN homolog (indicated by group I [PanA] and group II [PanB]) (Fig. 2). Of these, the methanosarcina PanA proteins were more closely related to the single PAN homologs distributed among the Crenarchaeota, Nanoarchaeota, and Euryarchaeota. This close association suggests that the methanosarcina PanA are the most ancient of the PAN paralogs. In contrast, the haloarchaeal PanA and PanB proteins share a high level of identity, suggesting that they arose from a relatively recent gene duplication event which most likely occurred near the divergence of the haloarchaeal lineage. This is also based on the finding that the haloarchaeal PanA and PanB proteins are more closely related to their respective counterparts from other haloarchaea than to the homolog within the same species. Together these results suggest that the archaeal PAN paralogs, including those of H. volcanii, play nonredundant roles in cell physiology. In contrast, many of the eucaryal Rpt paralogs are nearly identical; for example, the Rpt2, Rpt4, Rpt5, and Rpt6 paralogs of A. thaliana share 94 to 99% identity.

PAN proteins (PanA and PanB) of H. volcanii.

The synthesis of multiple archaeal PAN proteins has not yet been reported. In fact, most archaea encode only a single PAN homolog, and several do not encode any PAN (e.g., Thermoplasma spp., Pyrobaculum). To address whether H. volcanii synthesizes both the PanA and PanB proteins, an immunoblotting approach was employed. Polyclonal antibodies were raised in rabbits against PanA_Δ1-40-His and PanB-His purified from recombinant E. coli. Antigen fidelity was confirmed by N-terminal sequencing and Q-STAR MALDI-TOF (Fig. 1). Western blot analysis of cell lysate prepared from stationary-phase H. volcanii was performed with these antibodies (Fig. 3). Both PanA and PanB proteins were detected, and based on migration by SDS-PAGE their molecular masses were estimated to be 54 and 65 kDa, respectively. Under similar electrophoretic conditions, the PanA-His and PanB-His purified from recombinant E. coli migrated as slightly larger proteins (55 and 66 kDa) due to the addition of six C-terminal histidine residues. The similar migration of the PanA and PanB proteins purified from recombinant E. coli to those of H. volcanii supports the conclusion that the gene sequences isolated from the H. volcanii chromosome code for the PanA and PanB polypeptides detected in cell lysate. It remains to be determined why the molecular masses calculated from the deduced PanA and PanB protein sequences were at least 8 kDa less than those estimated by SDS-PAGE gels. Although the separation of halophilic proteins by SDS gel electrophoresis usually overestimates molecular weight (33), this alone is unlikely to account for the observed differences. It is possible that the N-terminal region of these PAN proteins adopts a coiled-coil conformation under these conditions which retards their electrophoretic migration.

FIG. 3.

PanA and PanB are produced in H. volcanii. Cell lysate of stationary-phase H. volcanii DS70 (10 μg) (lanes 2 and 4) as well as PanA-His (45 ng) (lane 1) and PanB-His (20 ng) (lane 3) purified from recombinant E. coli were separated by SDS-7.5% PAGE. Protein was analyzed by Western blotting with anti-PanA (lanes 1 and 2) and anti-PanB (lanes 3 and 4) antibodies. Molecular mass standards are indicated to the right. Preimmune serum was included as a negative control (data not shown).

Levels of 20S proteasome and PAN proteins in H. volcanii.

Based on the results presented above coupled with our group's previous work (14), it is apparent that H. volcanii synthesizes five different proteasomal proteins, including three 20S (α1, α2, and β) and two PAN (PanA and PanB) proteins. This combination of proteasomal proteins is unique and suggests that the closely related α (α1 and α2) and PAN paralogs (PanA and PanB) may serve similar yet nonredundant functions in H. volcanii cell physiology. If the functions of these paralogs are distinct, it is quite possible that the levels of these proteins are differentially regulated based on growth and/or environmental conditions. To address this, the levels of the 20S proteasome and PAN proteins were estimated by quantitative immunoblotting throughout the course of growth of H. volcanii (Fig. 4). Based on this analysis, significant differences were detected in the levels of the proteasomal proteins. In particular, the expression patterns of the paralogous α1 and α2 as well as PanA and PanB proteins differed. The PanA and β proteins remained relatively constant throughout growth, each accounting for ∼0.4% of the total protein (Fig. 4). The α1 protein exhibited a modest but steady and linear (r² = 0.9679) increase over the same period, increasing by 0.0044% total protein per h from 0.16 to 0.39% of total protein as cells transitioned from exponential to stationary phase. In contrast, the PanB and α2 proteins increased rapidly as cells entered stationary phase. Though dramatic increases were seen in the levels of the α2 and PanB proteins between exponential and stationary phase, these proteins were present at notably lower levels than the three other proteasomal proteins (α1, β, and PanA). In stationary phase, the α2 and PanB proteins each accounted for only about ∼0.12% of the total soluble protein, one-third that of the α1, β, and PanA proteins.

FIG. 4.

Levels of 20S proteasome and PAN proteins of H. volcanii based on a quantitative immunoblot assay. Immunoblotting of α1 (•), PanA (○), β (▪), α2 (□), and PanB (▴) and growth (OD₆₀₀ [Δ]) are shown. Time (in hours) refers to the amount of time elapsed during growth. Protein levels are the percentage of total protein of cell lysate from three independent experiments. The overall trend in protein levels (including the reduction in the levels of α2, PanB, PanA, and β proteins in log phase) was similar for all three experiments.

Based on these results, 0.9 to 1.1% of the total protein of an exponential-phase culture of H. volcanii consists of 20S proteasomal and PAN proteins. During stationary phase, the overall levels of these proteins increase to ∼1.5% of total protein. These estimates of the percentage of 20S proteasomal protein to total protein are in good agreement with data based on specific activity of crude extract compared to 20S proteasomes purified from stationary-phase H. volcanii (data not shown) as well as that from other bacterial and eucaryotic cells (0.2 to 1.1% of the total soluble protein) (7, 15, 28). The estimates of H. volcanii PAN proteins are also consistent with the levels estimated for the Rhodococcus erythropolis ARC protein, a distantly related bacterial homolog. This ARC protein was determined to be 0.1% of total protein as analyzed by two-dimensional PAGE with polyclonal anti-ARC antibodies (39).

mRNA levels of 20S proteasome and PAN proteins in H. volcanii.

The differential expression of the five proteasomal proteins throughout growth of H. volcanii raises the possibility that the levels of these proteins are coordinated at the transcriptional level. The availability of the psmA, psmB, psmC, panA, and panB genes encoding α1, β, α2, PanA, and PanB proteins, respectively, allowed the mRNA levels of these genes to be investigated by quantitative RT-PCR. Surprisingly, the transcript levels of all five genes were found to increase from four- to eightfold as cells transitioned from exponential to stationary phase (Fig. 5). The psmC- and panB-specific mRNA levels displayed the most pronounced increase (seven- to eightfold), while the increases in panA-, psmB-, and psmA-specific mRNA levels were somewhat less (four- to sixfold). Interestingly, throughout growth, the overall levels of the psmC- and panA-specific mRNA were more than 10-fold higher than those specific for psmA, psmB, and panB. The psmC- and panA-specific mRNA accounted for ∼0.14 and 0.28%, respectively, of the total RNA from stationary-phase H. volcanii cells (Fig. 5). In contrast the psmA-, psmB-, and panB-specific mRNA accounted for only 0.01, 0.008, and 0.003%, respectively, of the total RNA in the same phase of growth (Fig. 5). The reason for these differences remains to be determined.

FIG. 5.

Levels of 20S proteasome and Pan-specific mRNA of H. volcanii. (A) mRNA levels were analyzed by quantitative RT-PCR from three independent experiments. Total RNA purified from exponential phase (gray) (OD₆₀₀ of 1.1) and stationary phase (black) (OD₆₀₀ of 3.2) is shown. (B) Total RNA concentration was determined by absorbance at 260 nm of a dilution series of the sample. rRNA separated by gel electrophoresis and estimated by ethidium bromide staining was used as an invariant control to ensure equal quality of sample. Lanes 1 and 2, total RNA prepared from exponential- and stationary-phase cells, respectively.

Since the genes coding for the 20S proteasome and PAN proteins are dispersed throughout the H. volcanii genome, a trans-activating factor(s) is most likely required to coordinate the regulated increase in transcript levels observed during the transition to stationary phase. The reason why all proteasomal gene transcript levels increased is not immediately evident, since only α2 and PanB proteins displayed a substantial increase at the protein level during this same period of growth, while α1, β, and PanA remained relatively constant. Previous analysis in eucaryotes, however, revealed that proteasome-specific mRNA levels do not necessarily correlate with protein (8). Thus, it appears that both transcriptional and posttranscriptional mechanisms are involved in regulating the levels of proteasomal proteins in H. volcanii.

Conclusion.

Our group's previous work (14) demonstrated that H. volcanii synthesizes at least two 20S proteasomes (the symmetrical α1β CP and asymmetrical α1α2β CP). These proteasomal subtypes differ substantially in residues located at the ends of the 20S cylinder as well as residues in a loop restricting the channel opening (based on protein modeling). Thus, changes in the ratio of α1 to α2 were proposed to modulate distinct structural domains in 20S proteasomes. This in turn may alter the gating and/or the type of ATPase (e.g., PAN) that associates with 20S proteasomes and, subsequently, regulate the type of protein substrate recognized for destruction. In this study, we identified two H. volcanii PAN homologs (PanA and PanB) which, based on primary sequences, are predicted to differ in their affinities for substrate proteins and/or 20S proteasome subtypes. The high levels and relatively constant expression of the PanA protein throughout growth paralleled that of the 20S proteasomal α1 and β proteins (Fig. 6). In contrast, the PanB and α2 proteins were expressed at relatively low levels but were highly regulated as cells transitioned from exponential to stationary phase. This suggests that PanA may associate with the α1β 20S proteasome and play a central role in the housekeeping functions of the cell, while the PanB and α1α2β 20S proteasomes function in auxiliary tasks. The increase in the protein levels of α2 and PanB from exponential to stationary phase raises the possibility that these subunits play a role during this transition period, perhaps acting to enhance proteolysis of specific proteins. Whether the PanA and PanB proteins associate as hetero- and/or homo-oligomers and whether other 20S proteasome subtypes exist in H. volcanii remain to be determined.

FIG. 6.

Working model for the modulation of 20S proteasome (CP) and Pan isoforms in H. volcanii. A transition is predicted to occur between a primarily α1β CP and PanA complex in log phase to a mixture of α1β and α1α2β CP isoforms and PanA- and PanB-containing complexes in stationary phase. Whether PanA and PanB associate as PanA, PanB, and PanA/PanB complexes to allow for association with the 20S proteasome isoforms and recognize a greater diversity of protein substrates in this latter stage of growth remains to be determined.