Report

Role of Inorganic Polyphosphate in Promoting Ribosomal Protein Degradation by the Lon Protease in E. coli

See allHide authors and affiliations

Science  27 Jul 2001:
Vol. 293, Issue 5530, pp. 705-708
DOI: 10.1126/science.1061315

Abstract

Inorganic polyphosphate (polyP), a polymer of hundreds of phosphate (Pi) residues, accumulates inEscherichia coli in response to stresses, including amino acid starvation. Here we show that the adenosine 5′-triphosphate–dependent protease Lon formed a complex with polyP and degraded most of the ribosomal proteins, including S2, L9, and L13. Purified S2 also bound to polyP and formed a complex with Lon in the presence of polyP. Thus, polyP may promote ribosomal protein degradation by the Lon protease, thereby supplying the amino acids needed to respond to starvation.

PolyP is found in all microbes, fungi, plants, and animals (1). In Escherichia coli, levels of polyP are low in the exponential phase of growth, but increase more than 100-fold in response to acute stresses such as amino acid starvation and the multiple stresses in stationary phase (2, 3). AnE. coli mutant deficient in polyphosphate kinase (PPK), the principal enzyme for the synthesis of polyP, shows an extended lag in growth when shifted from a rich to a minimal medium (downshift); addition of amino acids abolishes the growth lag (4). Degradation of intracellular proteins is important in providing amino acids for use in the synthesis of the enzymes required for adaptations to starvation (4,5). The mutant fails to increase protein turnover after the downshift and thus extends the lag (4). In yeast and animal cells, the bulk degradation of proteins in response to starvation and cellular differentiation occurs by a ubiquitin-style conjugation system (6). However, in bacteria, the mechanisms underlying the regulation of intracellular protein degradation during amino acid starvation remain unknown (5).

In E. coli, more than 90% of cytoplasmic protein degradation is energy dependent (7). The ATP-dependent protease Lon, conserved from bacteria to humans (8, 9), is responsible for the rapid turnover of both abnormal and naturally unstable proteins (7,10). Lon and two other proteases (ClpAP and ClpXP) are responsible for 70 to 80% of the energy-dependent protein degradation in E. coli (7). Here, mutants deficient in either the Lon or Clp proteases (but not the HslVU protease) showed an extended growth lag after the downshift, which was more severe in the double mutant (lon, clp) (Fig. 1A). The double mutant was also defective in protein turnover after the downshift (Fig. 1B), as is theppk mutant (4). Addition of amino acids (50 mg/liter) abolished the growth lag of the protease mutants; once adapted to minimal medium, the mutants grew well on it. Both the Lon and Clp proteases were therefore involved in the increase in protein turnover after the downshift. Because the concentrations of adenosine 5′-triphosphate (ATP) after the downshift were similar in the wild type and the ppk mutant, polyP is not likely to serve as an energy source. The rate of protein turnover of a triple mutant (lon, clp, Δppk-ppx) was virtually identical to that of the lon, clpdouble mutant (Fig. 1B), consistent with polyP operating in the same pathway for protein degradation as the Lon and Clp proteases.

Figure 1

Growth lag of E. coliprotease mutants after downshift and rates of protein turnover. (A) Escherichia coli MG1655 and its derivatives KY2966 (hslVU), KY2347 (lon,clpPX), and KY2350 (lon,clpPX, hslVU) (20) were grown to mid-log phase in an amino acid–rich medium and downshifted to a minimal medium (4). Growth was measured as the optical density at 600 nm (OD600) after the downshift. (B) Rates of intracellular protein degradation were measured (4); proteolysis was expressed as the percentage increase of the trichloroacetic acid–soluble radioactivity relative to the total incorporation (14C-leucine).

A linear chain of 32P-labeled polyP [∼700 inorganic phosphate (Pi) residues] was synthesized withE. coli PPK and [γ-32P]ATP (2, 3). Purified Lon protease (8) bound to the 32P-labeled polyP even in 10 mM Pi(Fig. 2A). Presumed complexes of Lon with polyP increased with the concentration of Lon; two bands appeared shifted from the position corresponding to polyP (Fig. 2A). When DNA (pUC119), which binds Lon (8), was added with equimolar amounts (equivalent to 26 times the mass of polyP), there was no effect on the binding of polyP to Lon, nor was there any detectable complex of bovine serum albumin (BSA) with polyP even at 100 times the molar amounts of Lon (Fig. 2A). Previous studies (7,8), as well as our assay by gel filtration chromatography, indicated that Lon formed a tetramer under these conditions. A filter-binding assay showed that Lon bound to polyP with a dissociation constant (K d) of about 0.5 nM irrespective of the presence of ATP, and that one tetramer of Lon bound to one molecule of polyP (Fig. 2B).

Figure 2

Binding of polyP by purified Lon. (A) 32P-labeled polyP (3) was incubated with purified Lon (8) at room temperature in the presence of 50 mM tris-HCl buffer (pH 7.4), 10 mM MgCl2, and 10 mM Pi and then subjected to 1% agarose gel electrophoresis (TAE buffer). PolyP used in the experiments was the long-chain polyP (700 Pi residues), unless indicated otherwise. (B) A polyP-binding assay of Lon (2.5 nM as a tetramer) was performed with a nitrocellulose filter (21). 32P-polyP was used at concentrations of 0.86 to 8.6 nM. A Scatchard plot yielded a straight line with a slope of −1/K d and intercept on the x axis corresponding to the maximum concentration of32P–polyP-Lon.

An E. coli lysate was fractionated by phospho-cellulose or DEAE-cellulose, and then each fraction was subjected to proteolysis by MBP-Lon (11) in the presence and absence of polyP. Among many proteins tested, a few were degraded only in the presence of polyP (Fig. 3A). NH2-terminal sequences of those proteins corresponded to those of ribosomal proteins S2, L9, and L13, respectively (12). PolyP at 1 μM (0.7 mM as Pi residues) was effective for degradation of the S2 and L13 proteins (Fig. 3B). In response to amino acid starvation, levels of polyP increase even at concentrations >14 μM (10 mM as Pi residues) (2, 3) and far exceed those required for the effective degradation of these proteins. The S2 protein remained stable even after 100 min in the absence of polyP, but was degraded by Lon in the presence of 1.4 μM polyP with a half-life of about 50 min (Fig. 3C); ATP was also required for the rapid S2 degradation by Lon. A shorter chain length of polyP (P65) was less active in stimulating Lon-dependent degradation of S2 as well as of other ribosomal proteins, whereas P15 was inactive (Fig. 3D). PolyP did not affect casein hydrolysis by the Lon protease. To monitor the in vivo degradation of the S2 protein after the downshift, we expressed a S2–V5-epitope fusion (13) in the wild type and the ppk mutant. The S2-V5 fusion was stable without the downshift but was degraded rapidly after the downshift, and only in the wild type (Fig. 3E); the S2-V5 fusion was stable in either the lon or ppk mutants after the downshift (Fig. 3E). Thus, the S2 protein was subject to Lon-dependent degradation in the presence of polyP in vivo as well as in vitro.

Figure 3

Degradation of ribosomal proteins by MBP-Lon or Lon in the presence of polyP in vitro and in vivo after the downshift. (A) A phospho-cellulose fraction (Fraction P9) obtained from the E. coli lysate (22) was incubated with 0.6 μg of MBP-Lon (11) in the presence of polyP (0.7 mM as Pi; 1 μM as polymer). After incubation at 37°C for 60 min, exopolyphosphatase (yeast PPX, 3 × 104 U) was added to degrade the polyP; after 5 min, the mixture was subjected to 12% SDS–polyacrylamide gel electrophoresis (PAGE) and then visualized by silver staining (Di-ichi). (B) Fraction P9 was incubated with MBP-Lon (0.6 μg) in the presence of polyP (0.01, 0.1, or 1 μM) for 60 min and analyzed as described in (A). (C) Fraction P9 was incubated with Lon (1 μg) in the presence of polyP (1.4 μM). Samples were removed from the reaction mixture at the indicated times and then analyzed as in (A). Yeast PPX was not used. (D) PolyP with chain lengths of 65 and 15 residues (Sigma) was used for degradation of ribosomal proteins (23). The ribosomal proteins (2 μg) were incubated with MBP-Lon (0.1 μg) in the presence of polyP (0.64 mM as Pi) for 60 min at 37°C and then analyzed as in (A). (E) S2–V5-epitope fusion (13) was expressed in the wild type and in theppk mutant on 2× YT medium for 2 hours in the presence of 0.2% l-arabinose. Cells were collected by centrifugation (10 min, 3000g) and resuspended in the 2× YT medium withoutl-arabinose (No downshift) or in the MOPS minimal medium without l-arabinose (Downshift). At the indicated times, total proteins (100 μl of the culture) were subjected to SDS-PAGE and Western analysis with an antibody to V5 epitope (Invitrogen).

The S2 protein binds late in the assembly process of the ribosome, is localized on the surface of the 30S subunit (14), and is essential for ribosomal function. Purified S2 bound to polyP, as determined by the filter-binding assay, with a K d of ∼12 nM; the binding to polyP was observed at a level of 55% even in the presence of a 10 times the mass of pUC119 DNA and at a level of 30% in the presence of a 10 times the mass of E. coli total RNA (Fig. 4A). The complex formation of Lon with substrate may be necessary for efficient degradation (15). Lon formed a complex with S2 in the presence of polyP (Fig. 4B), suggesting that binding of S2 to polyP helped in the formation of a complex with Lon. Addition of polyP (without substrates) did not stimulate the adenosine triphosphatase activity of Lon. Thus, the stimulation of S2 degradation can be ascribed mainly to the formation of a complex between Lon and S2 in the presence of polyP.

Figure 4

A presumed complex formation with Lon and S2 in the presence of polyP. (A) A polyP-binding assay of the purified S2 protein (24) was performed with a nitrocellulose filter (21). The purified S2 (0.45 μg) or BSA (2 μg) was subjected to binding with 0.04 μg of 32P-polyP (11 nM) in the presence of 0.45 μg of DNA (pUC119) or 0.48 μg of RNA (total E. coli RNA). (B) Purified MBP-Lon (12 μg) and S2 (50 μg) were incubated in 20 mM tris-HCl (pH 7.4) and 5 mM MgCl2 in the presence or absence of 0.2 μM polyP for 10 min at 37°C without ATP. The mixture (500 μl) was applied onto a 1-ml column embedded with amylose resin. After washing the column with 50 mM tris-HCl (pH 7.4) and 200 mM NaCl (fractions 1 to 4), MBP-Lon was eluted with the same buffer containing 10 mM maltose (fractions 5 and 6). Fractions were analyzed by SDS-PAGE.

Escherichia coli contains substantial amounts of ribosomal proteins, including S2, as disassembled (free) forms during exponential growth on a rich medium, but little on a minimal medium (16). About 17% of total S2 protein exists as the free form (16), and such ribosomal proteins may be subjected to Lon-dependent degradation (17). Thus, it is likely that the S2 protein found in fraction P9 was free from ribosomes, and that polyP was involved in the degradation of free ribosomal proteins after the downshift. Degradation of ribosomal proteins should release amino acids for synthesis of the key enzymes required for adaptations to starvation, as well as reduce translational activity during starvation. Most substrates for polyP-dependent degradation were basic ribosomal proteins that could bind to polyP [see supplementary material (18)]. In assays of polyP binding to proteins in E. coli lysates, most proteins were contained within the ribosome fractions, but also included Lon as well as ribosome-associated proteins. Thus, it is likely that the ribosomal proteins are the major substrates (in terms of mass) for this system. When we used intact ribosomes as substrates for Lon, Lon with polyP was ineffective in degrading intact ribosomes, but did act on ribosomes treated with ribonuclease (RNase) (18). Thus, polyP alone did not disassemble the ribosome, but polyP and Lon together degraded free ribosomal proteins, as well as those in the RNase-distorted ribosome. Our findings may provide insights into the regulation of protein degradation when cells are stressed or enter the stationary phase.

  • * To whom correspondence should be addressed. E-mail: akuroda{at}hiroshima-u.ac.jp

REFERENCES AND NOTES

View Abstract

Navigate This Article