Previous Article | Next Article ![]()
Journal of Bacteriology, January 2003, p. 165-174, Vol. 185, No. 1
0021-9193/03/$08.00+0 DOI: 10.1128/JB.185.1.165-174.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Department of Microbiology and Cell Science, University of Florida, Gainesville, Florida 32611-0700
Received 3 July 2002/ Accepted 2 October 2002
|
|
|---|
1,
2, and ß) which are classified in the 20S proteasome superfamily. The
1 and ß proteins alone form active 20S proteasomes; the role of
2, however, is not clear. To address this,
2 was synthesized with an epitope tag and purified by affinity chromatography from recombinant H. volcanii. The
2 protein copurified with
1 and ß in a complex with an overall structure and peptide-hydrolyzing activity comparable to those of the previously described
1-ß proteasome. Supplementing buffers with 10 mM CaCl2 stabilized the halophilic proteasomes in the absence of salt and enabled them to be separated by native gel electrophoresis. This facilitated the discovery that wild-type H. volcanii synthesizes more than one type of 20S proteasome. Two 20S proteasomes, the
1-ß and
1-
2-ß proteasomes, were identified during stationary phase. Cross-linking of these enzymes, coupled with available structural information, suggested that the
1-ß proteasome was a symmetrical cylinder with
1 rings on each end. In contrast, the
1-
2-ß proteasome appeared to be asymmetrical with homo-oligomeric
1 and
2 rings positioned on separate ends. Inter-
-subunit contacts were only detected when the ratio of
1 to
2 was perturbed in the cell using recombinant technology. These results support a model that the ratio of
proteins may modulate the composition and subunit topology of 20S proteasomes in the cell. |
|
|---|
The subunits that form 20S proteasomes have been classified into two related superfamilies (
and ß) (9). The
proteins form the outer rings (18) and are required for the ß proteins to be processed during formation of inner rings which harbor the active-site N-terminal threonine (13, 25, 26, 31, 38). The number of subunits forming 20S proteasomes varies. Many lower eucaryotes such as yeast produce a single symmetrical 20S proteasome of 14 different subunits (i.e., two copies each of
1 to
7 and ß1 to ß7) (20). Higher eucaryotes express additional subunits (e.g., ß1i, ß2i, ß5i) that form auxiliary 20S proteasomes (e.g., the immunoproteasome) (19). Bacterial 20S proteasomes are typically much simpler, being composed of a single
-type and a single ß-type subunit (12). The bacterium Rhodococcus erythropolis, however, is an exception and synthesizes a 20S proteasome composed of two
and two ß subunits (36). Interestingly, genome analysis suggests that increased 20S proteasome complexity may actually be widespread among the Archaea, with several Creanarchaeotes and Euryarchaeotes encoding three proteasomal proteins (typically a single
-type and two ß-type open reading frames [ORFs]) (27).
To date, Haloferax volcanii is the only archaeon that has been shown to synthesize three different proteins (
1,
2, and ß) that are classified in the 20S proteasome superfamily (33). The
1 and ß proteins form active 20S proteasomes; however, prior to this study, it was not clear whether
2 formed 20S proteasomes, and if so, with which proteins it associated. To further examine the unusual nature and topology of H. volcanii 20S proteasomes,
1,
2, and ß were separately expressed with an epitope tag and purified from H. volcanii. Recombinant Escherichia coli was used to analyze the apparent flexibility of
subunit associations. In addition, 20S proteasomes purified from wild-type and mutant cells were analyzed by native gel electrophoresis and cross-linking. This series of approaches provided a model for how the composition and topology of 20S proteasomes may be modulated by the ratio of
subunits in the cell.
|
|
|---|
Strains, media and plasmids. Bacterial strains, oligonucleotide primers, template DNA, and plasmids are summarized in Table 1. E. coli strains were grown in Luria-Bertani medium (37°C, 200 rpm). H. volcanii strains were grown in complex medium (ATCC 974) (42°C, 200 rpm). Media were supplemented with 100 mg of ampicillin, 50 mg of kanamycin, or 0.1 mg of novobiocin per liter as needed.
|
View this table: [in a new window] |
TABLE 1. Strains and plasmids used for this study
|
1 and
2 in E. coli, the psmA-H6 and psmC-H6 genes were positioned downstream of psmC and psmA, respectively, using DNA fragments isolated from the pET24b-based plasmids.
For expression of epitope-tagged proteins in H. volcanii, the modified proteasomal genes (psmA-H6, psmB-H6, and psmC-H6) were isolated from the pET24b-based plasmids by restriction enzyme digestion (Table 1) and were blunt end ligated into the BamHI and KpnI sites of plasmid pBAP5010. The DNA fragments used for ligation included not only the modified gene but also the ribosome binding sequence and T7 terminator of the original pET24b. The start codon of each modified gene was positioned 75 bp downstream of the Halobacterium cutirubrum rRNA P2 promoter. This resulted in the generation of shuttle expression plasmids pJAM202, pJAM204, and pJAM205 for the synthesis of ß-,
1-, and
2-His proteins, respectively. For the generation of H. volcanii psmC-1, strain WFD11 was transformed with a suicide plasmid (pJAM201) and selected for growth in the presence of novobiocin. Colonies were screened for the absence of
2 by immunoblotting using anti-
2 antibodies (described below). Strains which did not produce detectable levels of
2 were further screened for double recombination by PCR using the oligonucleotides 5'-TTCATATGAACCGAAACGACAAGCAGG-3' and 5'-TGTTACCGTCGGTCGGGCTCCCGTCTG-3', with genomic DNA as a template.
DNA purification and transformation. Genomic DNA was isolated from H. volcanii strains as previously described (33). Plasmid DNA was isolated using a Quantum Prep plasmid miniprep kit (Bio-Rad, Hercules, Calif.). DNA fragments were eluted from 0.8% SeaKem GTG agarose (FMC Bioproducts, Rockland, Maine) gels with 1x TAE buffer (40 mM Tris-acetate, 2 mM EDTA, pH 8.5) using the QIAquick gel extraction kit (Qiagen, Valencia, Calif.). H. volcanii WFD11 cells were transformed according to the method of Cline et al. (8) using plasmid DNA isolated from E. coli GM2163 strains.
Protein synthesis and purification. Expression of heterologous genes was induced from plasmids in E. coli BL21(DE3) as previously described (26). Epitope-tagged proteasome proteins were synthesized from plasmids in recombinant H. volcanii strains grown to late stationary phase. Protein purification steps were done at room temperature unless otherwise indicated. Buffers typically included high salt (2 M) to mimic the ionic strength of the cytosol of this halophilic archaeon. For non-histidine-tagged proteins, buffers were supplemented with 1 mM dithiothreitol. For all purifications, centrifugations were done at 16,000 x g (20 to 30 min, 4°C). Cells were harvested by centrifugation and stored at -70°C. Cells were resuspended in 2.5 to 6 volumes (wt/vol) of lysis buffer and lysed by passage through a French pressure cell at 20,000 lb/in2 followed by centrifugation. Dialysis was at 4°C for 16 h followed by centrifugation. Samples were filtered (0.45-µm-pore-size filter) prior to column application. Fractions were monitored for peptidase activity using N-Suc-LLVY-Amc (N-succinyl-Leu-Leu-Val-Tyr-7-amido-4-methylcoumarin) or by staining with Coomassie blue R-250 after separation by reducing 12% polyacrylamide gel electrophoresis (PAGE) with sodium dodecyl sulfate (SDS) (33). Molecular mass standards for SDS-PAGE included phosphorylase b (97.4 kDa), serum albumin (66.2 kDa), ovalbumin (45 kDa), carbonic anhydrase (31 kDa), trypsin inhibitor (21.5 kDa), and lysozyme (14.4 kDa) (Bio-Rad). Samples were stored at 4°C. Native molecular masses were determined by applying samples to a calibrated Superose 6 HR 10/30 column (Pharmacia) as previously described (33). Molecular mass standards included serum albumin (66 kDa), alcohol dehydrogenase (150 kDa), ß-amylase (200 kDa), apoferritin (443 kDa), and thyroglobulin (669 kDa) (Sigma).
(i) Purification of
1 and
2 from recombinant E. coli.
The
1 and
2 proteins were purified from E. coli strains carrying plasmids pJAM618 and pJAM619, respectively. Cells were lysed in 50 mM Tris-Cl buffer at pH 7.2 containing 150 mM NaCl (buffer A). Lysate was dialyzed into 50 mM Tris-Cl buffer at pH 7.2 containing 2 M NaCl (buffer B). The supernatant was dialyzed into 50 mM Tris-Cl buffer at pH 7.2 containing 1.5 M NaCl-2.2 M (NH4)2SO4 (buffer C). The supernatant (6 mg of protein per ml) was applied to a DEAE-cellulose (Sigma) column (2.6 cm by 10 cm) equilibrated with buffer C. The column was washed with buffer C, and protein was eluted with 50 mM Tris-Cl buffer, pH 7.2, containing 2.1 M NaCl-1.6 M (NH4)2SO4 (buffer D). Sample was dialyzed into buffer B, concentrated (5 to 10 mg ml-1) by dialysis against PEG 8000, and applied to the Superose 6 column equilibrated in buffer B.
(ii) Purification of histidine-tagged proteasomal proteins.
The histidine-tagged
1,
2, ß, and ß
proteins were purified from E. coli strains carrying plasmids pJAM622, pJAM623, pJAM621, and pJAM620, respectively. Cells were lysed in buffer B containing 10 mM imidazole. Sample was applied to a Ni2+-Sepharose column (4.8 by 0.8 cm) (Pharmacia) equilibrated in lysis buffer and washed with buffer B (10 ml) containing 60 mM imidazole. Protein was eluted in buffer B containing 500 mM imidazole. Sample was applied to a Superose 6 HR 10/30 column equilibrated in buffer B containing 10% glycerol. Similar procedures were used for coexpression of the
proteins in E. coli, with the following modifications. E. coli strains carrying plasmids pJAM600 and pJAM601 were lysed in buffer A. Supernatant was dialyzed into buffer B with 5 mM imidazole. Aliquots were heated (15 min, 37°C), chilled (15 min, 0°C), and centrifuged. Heat-treated and unheated samples were applied to the Ni2+-Sepharose column. Similar procedures were used for purification of the His-tagged
1,
2, and ß proteins from H. volcanii carrying plasmids pJAM204, pJAM205, and pJAM202, with the following modifications. Buffers were supplemented with 10% glycerol. Cell lysate was directly subjected to Ni2+-Sepharose chromatography for analysis of proteasomal subunit associations and ratios. For specific activity measurements (Table 2), 20S proteasomes were purified to homogeneity by treating cell lysate with PEG8000 and heat, as previously described (33), prior to affinity and gel filtration chromatography.
|
View this table: [in a new window] |
TABLE 2. Subunit ratios of 20S proteasomes purified from H. volcanii strains
|
Antibody preparation and immunoanalysis.
Proteins (ß
-His,
1, and
2) purified from recombinant E. coli were separated by SDS-PAGE, excised from the gel, and used as antigens to raise polyclonal antibodies in rabbits (Cocalico Biologicals, Reamstown, Pa.). For chromogenic Western blots, antigens were detected using primary antibodies (1:10,000) and alkaline-phosphatase-conjugated anti-rabbit antibody raised in goat (1:20,000) (Southern Biotechnology Associates, Inc., Birmingham, Ala.). For quantitative Western blots, antigens were detected using primary antibodies (
1, 1:10,000;
2, 1:1,000; ß, 1:4,000) and horseradish peroxidase-conjugated anti-rabbit antibody raised in goat (
1, 1:10,000;
2, 1:2,000; ß, 1:4,000) by chemiluminescence with ECL Plus according to the supplier's recommendations (Amersham Pharmacia Biotech). The ß-His,
1-His, and
2-His proteins purified from recombinant E. coli were included on these blots as quantitative standards (2 to 80 ng per lane).
|
|
|---|
1,
2, and ß in H. volcanii.
To determine whether
2 is involved in the formation of active 20S proteasomes, the
2 protein was expressed with an epitope tag (
2-His) from plasmid pJAM205 in H. volcanii. This plasmid was designed for constitutive transcription of the modified proteasome gene (psmC-H6) from the H. cutirubrum rRNA P2 promoter. The T7 terminator and ribosome binding sequence from pET24b were included to facilitate protein synthesis. Although leaderless transcripts have been described for the haloarchaea (3, 4, 7, 11, 28, 29), the 3'-end 16S rRNA sequences of E. coli and H. volcanii are highly related (HO-AUUCCUCCACUAGGUUGG for E. coli [5] and HO-UCCUCCACUAGGUCGG for H. volcanii [accession no. AB074566]). Thus, it was predicted that the pET24b-based ribosome binding site used for protein synthesis in E. coli would function in H. volcanii.
For comparison to
2-His, the genes encoding
1 and ß proteins were similarly modified (psmA-H6 and psmB-H6 encoding
1-His and ß-His) and separately synthesized in vivo using plasmids pJAM204 and pJAM202. The epitope tag (His) consisted of a seven-amino-acid linker followed by six histidine residues at the C terminus of the protein to enable Ni2+-affinity purification. Based on three-dimensional modeling (33), this type of modification was expected to result in minimal, if any, perturbation to 20S proteasome structure. This modeling approach also suggested that the C termini of all three subunits would be located on the surface of the 20S proteasome cylinder and, thus, would be accessible for binding during affinity chromatography. In contrast, the N termini of
-type proteasomal subunits are essential for auto-assembly into the heptameric rings of 20S proteasomes (38). Furthermore, the ß subunit of 20S proteasomes purified from H. volcanii is processed at its N terminus to expose the presumed active-site threonine (33).
Quantitative immunoblotting was used to analyze the levels of 20S proteasome proteins produced in the recombinant and parent strains of H. volcanii. Cell lysate was prepared from stationary-phase cultures and probed with polyclonal antibodies specific for each of the proteasomal proteins (see Materials and Methods). This approach revealed that cells transformed with the expression plasmid pJAM204 synthesized
2 (with or without His) at four times the level of the parent strain. Similarly,
1 (with or without His) was elevated five- to sevenfold when homologously expressed with an epitope tag. When ß-His was synthesized with the 49-residue propeptide, the overall levels of ß were enhanced only twofold. Accumulation of unprocessed forms of ß-His was not observed (detection limit of 1.2 x 10-13 mol per cell). Modification of operon structure and/or amplified copy number of the proteasomal genes is the likely reason for the observed increases in the levels of the proteasomal proteins in the recombinant strains. The basis for the differences in the overexpression levels of ß compared to
1 or
2 remains to be determined. It is possible that posttranscriptional differences influenced synthesis, since similar genetic elements were used during plasmid construction (i.e., P2 promoter, T7 terminator, and ribosome binding sequence) and the positioning of the ORFs with respect to these elements was identical. However, differences in native genetic elements located within the ORFs that may drive transcription or translation cannot be ruled out. Interestingly, the observed perturbations in the levels of the individual proteasomal proteins had no apparent influence on the growth rate or cell yield based on comparison of recombinant and parent strains grown on rich medium under microaerobic conditions (200 rpm, 42°C) (data not shown).
2 protein is a 20S proteasome subunit.
Ni2+-affinity and gel filtration chromatography were used to purify and analyze the epitope-tagged protein complexes from the recombinant H. volcanii strains. Initial analysis of
2-His by Coomassie blue staining revealed that the 1,519-Da epitope tag resulted in its comigration with
1 on SDS-PAGE gels. Therefore, Western blots using antibodies specific for each proteasomal protein were included for comparison.
An analysis of proteins purified from H. volcanii(pJAM205) expressing the
2-His protein (Fig. 1) revealed two
2-specific polypeptides with molecular masses consistent with
2 and
2-His. An additional, unexpected
2-specific polypeptide was observed (fractions 35 and 44) which had a molecular mass of 800 Da less than that of
2-His but greater than that of the wild-type
2 protein. The reason for this third
2-specific antigen has not been determined. It is possible that
2-His is susceptible to cleavage, either in the cell or during purification. It is also possible that an internal start codon is recognized which results in the synthesis of an
2-His polypeptide with a shortened N terminus. This second explanation, however, is less likely based on the sequence of psmC (33).
![]() View larger version (38K): [in a new window] |
FIG. 1. Association of the 1, 2, and ß proteasomal proteins as demonstrated by purification of 2-His-containing complexes. (A) Proteins were purified from recombinant H. volcanii(pJAM205) using Ni2+-Sepharose and Superose 6 gel filtration. Gel filtration fraction numbers are indicated on the x axis. Total protein measured by A280 ( ) and N-Suc-LLVY-Amc-hydrolyzing activity () are indicated. (B) Proteins were separated by reducing 12% PAGE with SDS and analyzed by Coomassie blue (CB) staining or Western blotting using antibodies raised against 1, 2, and ß as indicated on the right. H. volcanii(pBAP5010) cell lysate (lane 1) and Ni2+-purified fractions (lane 2) were included as controls. H. volcanii(pJAM205) cell lysate (lane 3), Ni2+-purified fractions (lane 4), and Superose 6 gel filtration fractions 25 to 47 are shown. Arrows to the right indicate 1 (band 1)-, 2 (bands 2 and 3)-, and ß (band 4)-specific proteins.
|
2-His was associated with high-molecular-mass complexes of 600 kDa (fractions 29 to 33) (Fig. 1). The complexes were comparable in structure to 20S proteasome particles when analyzed by transmission electron microscopy (TEM) (data not shown). Likewise, their chymotrypsin-like activity was similar to that of 20S proteasomes purified from wild-type H. volcanii (Table 2) (33). The recombinant 20S proteasomes were composed of
1,
2 (with or without His), and ß in a molar ratio of approximately 1:1:3 (Table 2). The reason for the less-than-1:1 ratio of
to ß subunits is unclear and contrasts with the 1:1 ratio observed for the other H. volcanii 20S proteasomes described in this study (Table 2). It may be a reflection of the methods employed for analysis, since the
2-His-containing 20S proteasomes were the only complexes requiring quantitative immunoblotting for estimating the molar ratio of subunits (due to the comigration of
2-His and
1 on SDS-PAGE gels).
Complexes containing
2-His with molecular masses of less than 600 kDa were also purified; however, these had no apparent peptidase activity (Fig. 1). These inactive fractions were composed of
1 and
2-His in varying ratios, with the predominant protein peak estimated to be dimers to tetramers of the
proteins. Unmodified
2 was not detected in these low-molecular-mass fractions.
Together, these results reveal that
2 can associate with
1 and ß to form active 20S proteasomes in H. volcanii. The presence of a second ß subunit (i.e., ß2) which comigrates with the ß subunit on SDS-PAGE cannot be ruled out at this time. However, this speculative ß2 protein is not detected when sequencing the N terminus of the ß subunit. Furthermore, Southern blotting using a degenerate probe based on the N-terminal sequence of ß (33) and the partial genome sequence of H. volcanii (V. G. DelVecchio, personal communication) do not predict a second ß protein. Based on the results presented above, altering the C terminus of
2 by the addition of a poly-His tag did not inhibit 20S proteasome formation or peptidase activity. However, the preferential accumulation of unmodified
2 in the 20S proteasome fractions suggests that this modification influenced the affinity of
2 for 20S proteasome complex formation.
For comparison,
1-His-containing proteins were purified from H. volcanii(pJAM204) and analyzed using similar methods (Fig. 2). Western blotting with
1-specific antibody revealed that a significant portion (10 to 25%) of the
1-His polypeptide was cleaved to generate a protein slightly (500 Da) smaller than
1-His but larger than the wild-type
1 protein. This
1-specific antigen was present in the majority of
1-His complexes (fractions 31 to 41). Internal start codons are not predicted to generate this shortened form of the
1-His protein. Whether a related protease is involved in cleavage of both
1- and
2-His remains to be determined.
![]() View larger version (43K): [in a new window] |
FIG. 2. Association of the 1, 2, and ß proteasomal proteins as demonstrated by purification of 1-His-containing complexes. The figure is similar to Fig. 1, except plasmid pJAM204 was substituted for pJAM205. (A) Proteins purified by gel filtration. (B) Proteins separated by SDS-PAGE. Arrows to the right indicate 1 (bands 1 to 3)-, 2 (band 4)-, and ß (band 5)-specific proteins.
|
2-His, the
1-His protein was purified in high-molecular-mass (600 kDa) complexes. These complexes contained all three proteasomal proteins (
1,
2, and ß) in a molar ratio of approximately 4:1:5 (Table 2). This high-molecular-mass fraction had biochemical properties similar to the H. volcanii wild-type 20S proteasomes (33), including chymotrypsin-like peptidase activity (Table 2) and architecture (as visualized by TEM). Interestingly, a large portion of
1-specific antigen purified as heptamers in association with
2 (fractions 36 to 41). In fact, the ratio of heptamers to 20S proteasomes was almost 4:1 as estimated by A280. Low levels of
1-His and
2 were also detected in complexes smaller than heptamers. Both heptamers and fractions of lower molecular mass had no measurable N-Suc-LLVY-Amc hydrolyzing activity. Thus, when the ratio of
1 to
2 was modified by overproduction of
1-His, the rate-limiting step in assembly of
1 into proteasomes appeared to be after the formation of heptamers. In contrast, high-level production of
2-His resulted in the accumulation of low-molecular-mass dimers to tetramers, suggesting that heptamerization limits incorporation of
2 into proteasomes. These results reveal that
1 and
2 are capable of interacting, since these two proteins were found as dimeric to heptameric complexes that did not contain ß-specific proteins.
Overproduction of unprocessed ß-His in H. volcanii(pJAM202) facilitated purification of 20S proteasomes that were composed of
1-,
2-, and ß-specific proteins in a molar ratio of approximately 3:1.5:5 (Table 2). The complexes had typical 20S proteasome structures and were fully active in hydrolyzing N-Suc-LLVY-Amc (Table 2). In contrast to the
proteins, when unprocessed ß protein was overproduced in the cell the majority of ß protein was processed and purified as 20S proteasomes (data not shown). However, about a third of the ß protein purified as dimers to monomers. Based on N-terminal sequence analysis of these low-molecular-mass fractions, the ß protein was processed to expose the same N-terminal threonine described previously for H. volcanii 20S proteasomes (33). These low-molecular-mass fractions were ß specific, with no other proteins detected by Coomassie blue staining of SDS-PAGE gels. Although further investigation is needed, these results reveal that not all processed ß protein is associated with 20S proteasomes in these recombinant cells.
Overall, these findings are consistent with the current model for archaeal 20S proteasome assembly (2, 38) and provide evidence that
2 associates in 20S proteasomes with active ß subunits. The results also substantiate the finding that
1 is the predominant
-type subunit incorporated into 20S proteasomes. The
2 protein was only about one-third of the total
protein incorporated into the ß-His-containing 20S proteasomes (Table 2). Together, these results demonstrate that 20S proteasomes composed of all three subunits (
1,
2, and ß) can assemble in H. volcanii.
Association of
1 and
2.
To further address the interactions of the
-type subunits, E. coli strains were modified to independently synthesize
1 and
2 with and without C-terminal histidine tags. Salt was included in the purification buffers at high concentrations to mimic the unusually high ionic strength of the H. volcanii cytosol. Based on TEM, the
1 and
2 proteins produced separately in recombinant E. coli formed rings after dialysis into buffer supplemented with 2 M salt (Fig. 3). E. coli strains were also constructed that (i) coexpressed
1-His and
2 and (ii) coexpressed
1 and
2-His. Affinity and gel filtration chromatography of these proteins after dialysis into high salt (2 M) revealed that
1 and
2 were able to associate as heterogeneous dimers to heptamers in recombinant E. coli (data not shown). These results suggested that there was flexibility in the types of
1 and
2 interactions that were possible. Modifying the ratios of
1 to
2 during expression in E. coli directly influenced the composition of the
-type heptamers that were formed (i.e., homo- to hetero-oligomers).
![]() View larger version (140K): [in a new window] |
FIG. 3. Recombinant 1 and 2 proteins form ring structures. Transmission electron micrographs of negatively stained proteins are shown. An end-on view of a 20S proteasome (A) purified from H. volcanii(pJAM204) is included for comparison. (B and C) Typical end-on view of 1 (B) and 2 (C) purified from recombinant E. coli. (D) Typical end-on view of ring observed for the 200-kDa fraction purified from H. volcanii(pJAM204) which contains 1-His, 1, and 2 proteins. Bar, 12 nm. Samples were prepared and viewed on a Zeiss EM-10CA transmission electron microscope as previously described (33).
|
-type proteasomal subunits. Often when eucaryal
-type proteins are independently expressed in recombinant E. coli, the proteins self-associate, even though this type of interaction is not evident in purified 20S proteasomes. For example, the human
7 (HsC8) and Trypanosoma brucei
5 self-assemble into single, double, and even four-stacked protein rings when synthesized in recombinant E. coli (16, 35). In addition, the human
7 protein forms hetero-oligomeric rings when coexpressed with
1 (PROS27) and
6 (PROS30), both of which are adjacent to
7 in the outer rings of wild-type 20S proteasomes (15). Likewise, four different 20S proteasomes can be synthesized by expressing different combinations of R. erythropolis
1,
2, ß1, and ß2 in recombinant E. coli (36). However, only a single 20S proteasome containing all four subunits has been purified from R. erythropolis.
Association of
1 and
2 in 20S proteasomes from H. volcanii.
Previous work revealed that the 20S proteasomes of H. volcanii dissociate into monomers when exposed to low-salt buffers (33). Similar results have recently been observed for the 20S proteasome purified from Haloarcula marismortui (14). During this study it was discovered that low levels of CaCl2 (10 mM) stabilized the H. volcanii 20S proteasomes in the absence of salt without influencing peptide (N-Suc-LLVY-Amc)-hydrolyzing activity. This enabled 20S proteasomes purified from H. volcanii to be separated by native gels pre-equilibrated in the presence of CaCl2.
To directly address the type of
1-
2 subunit interactions present in H. volcanii, a protein fraction (A) was purified from stationary-phase wild-type cells that contained all three subunits in an
1-
2-ß molar ratio of 1:1:2 (Table 2). For comparison, 20S proteasomes composed of
1 and ß in an equimolar ratio were purified from a mutant strain (psmC-1) that did not produce
2. Both the
1-
2-ß proteasome (fraction A) and
1-ß proteasome migrated as distinct bands on native gels, suggesting that each formed a separate complex (Fig. 4). These two complexes were responsible for the majority of chymotrypsin-like peptidase activity (N-Suc-LLVY-Amc-hydrolyzing activity) detected in the cell lysate (data not shown), and both 20S proteasomes hydrolyzed N-Suc-LLVY-Amc at comparable rates (Table 2). This suggests that the
1-
2-ß proteasome identified in this study is the prevalent ancillary 20S proteasome in stationary-phase H. volcanii.
![]() View larger version (62K): [in a new window] |
FIG. 4. Native gel of H. volcanii 20S proteasomes reveals two distinct complexes. Lanes 2 and 3, 1- 2-ß and 1-ß proteasomes purified from H. volcanii WFD11 and psmC-1, respectively; lanes 1 and 4, equimolar mixture of the two proteasomes. Native gels containing 5% polyacrylamide were pre-equilibrated and run with 25 mM Tris-192 mM glycine buffer (pH 8.3) supplemented with 10 mM CaCl2. Proteins were stained with Coomassie blue R-250.
|
subunit interactions. The
1 and
2 proteins purified separately from recombinant E. coli were included as controls. The cross-linked products were separated by SDS-PAGE and analyzed by immunoblotting using antibodies specific for each proteasomal protein (Fig. 5). Probing with the anti-
2 antibody revealed a distinct band (63 kDa) which was present for the
1-
2-ß and ß-His proteasomes but was not detected for the
1-ß proteasomes. This antigen appeared to be an
2-specific dimer, since it was readily separated from all
1- and ß-specific products and migrated analogously to the
2 dimer from recombinant E. coli. A separate
1-specific band (65 kDa) was identified for all three of the 20S proteasomes examined. This common
1-specific band migrated similarly to the
1 dimer from recombinant E. coli. In addition,
1-ß dimers were observed for all three proteasomes and
2-ß dimers were detected for all but the
1-ß proteasomes (data not shown).
![]() View larger version (54K): [in a new window] |
FIG. 5. Cross-linking of H. volcanii 20S proteasomes reveals 1- 1 and 2- 2 contacts. The 20S proteasomes (300 nM) and proteins (10 µM) were incubated (10 to 60 min, 37°C) in buffer B containing 10% glycerol and 0.22% glutaraldehyde. The cross-linking reaction was quenched with 1x 2 M Tris-Cl at pH 8.5 followed by precipitation with 10% trichloroacetic acid. Products and molecular mass standards (see Materials and Methods) were separated by reducing 7.5% PAGE with SDS. Specific antigens were detected by chromogenic Western blotting using antibodies raised against 1 (lanes 1 to 3 and 7 to 10), 2 (lanes 4 to 6 and 11 and 12), and ß -His (data not shown). Samples included the 1-ß proteasome from H. volcanii psmC-1 (lanes 1, 6, and 8), the 1- 2-ß proteasome from H. volcanii WFD11 (lanes 2, 5, 10, and 11), 1 (lanes 3 and 7) and 2 (lane 4) from recombinant E. coli(pJAM618) and E. coli(pJAM619), and ß-His-containing proteasomes from H. volcanii(pJAM202) (lanes 9 and 12). Arrowheads indicate putative dimers.
|
1-
2 interactions cannot be ruled out, these results strongly support the existence of an
1-
2-ß proteasome in stationary-phase cells which is composed primarily of
1-
1,
2-
2,
1-ß, and
2-ß contacts. Thus, based on our current understanding of 20S proteasome structure, it is likely that this proteasome is formed from four homoheptameric rings with two inner ß-rings and two outer
-rings of different subunit composition (
1 and
2). This would be the first example of a 20S proteasome with this type of asymmetry.
Although heterogeneous heptamers of
1 and
2 were observed in recombinant E. coli as well as recombinant H. volcanii, it is possible that these intermediate complexes were formed as a result of perturbations in the ratio of
1 to
2. Alternatively, it may be that
1-
2 associations do exist in 20S proteasomes but were not detected by our cross-linking methods. It is also possible that the 20S proteasome complexes purified from stationary-phase cells are a snapshot of those synthesized in the cell. The interactions between
1 and
2 may actually be dynamic; the cell may generate assembly intermediates composed of
1-
2 contacts that enable it to rapidly transition from
1-ß to
1-
2-ß proteasomes during growth.
|
|
|---|
subunit composition and topology. Why H. volcanii would synthesize more than one type of 20S proteasome remains to be determined. Since the
1 and
2 proteins share only 55.5% identity, there are significant structural differences in the homoheptameric rings formed by these two proteins. These differences are predicted to include residues preceding the N-terminal
-helix HO, located at the ends of the cylinder, as well as the loop restricting the 13-Å channel opening (based on comparison to the Thermoplasma acidophilum 20S proteasome crystal structure [25]). Thus, it is proposed that by modulating the levels of
1 and
2 proteins H. volcanii may control distinct structural domains at the ends of the 20S proteasome cylinder. This in turn may influence the gating and/or type of ATP-dependent unfoldases that interact with the 20S proteasome and, potentially, the type of substrate recognized for destruction.
This work was supported in part by the NIH (GM57498-03) and the Florida Agricultural Experiment Station (Journal Series R-09097).
|
|
|---|
-type subunits. J. Mol. Biol. 275:113-121.[CrossRef][Medline]
-type proteasomal subunit HsC8 forms a double ring-like structure, but does not assemble into proteasome-like particles with the ß-type subunits HsDelta or HsBPROS26. J. Biol. Chem. 272:10080-10086.
inducible exchanges of 20S proteasome active site subunits: why? Biochimie 83:367-372.[Medline]
5 subunit in Trypanosoma brucei proteasome can self-assemble to form a cylinder of four stacked heptamer rings. Biochem. J. 344:349-358.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»