The amino-terminal domain of chloroplast Hsp93 is important for its membrane association and functions in vivo

Abstract

Chloroplast 93-kD heat shock protein (Hsp93/ClpC), an Hsp100 family member, is suggested to have various functions in chloroplasts, including serving as the regulatory chaperone for the ClpP protease in the stroma and acting as a motor component of the protein translocon at the envelope. Indeed, although Hsp93 is a soluble stromal protein, a portion of it is associated with the inner envelope membrane. The mechanism and functional significance of this Hsp93 membrane association have not been determined. Here, we mapped the region important for Hsp93 membrane association by creating various deletion constructs and found that only the construct with the amino-terminal domain deleted, Hsp93-ΔN, had reduced membrane association. When transformed into Arabidopsis (Arabidopsis thaliana), most atHsp93V-ΔN proteins did not associate with membranes and atHsp93V-ΔΝ failed to complement the pale-green and protein import-defective phenotypes of an hsp93V knockout mutant. The residual atHsp93V-ΔN at the membranes had further reduced association with the central protein translocon component Tic110. However, the degradation of chloroplast glutamine synthetase, a potential substrate for the ClpP protease, was not affected in the hsp93V mutant or in the atHSP93V-ΔN transgenic plants. Hsp93-ΔN also had the same ATPase activity as that of full-length Hsp93. These data suggest that the association of Hsp93 with the inner envelope membrane through its amino-terminal domain is important for the functions of Hsp93 in vivo.

Figure 1.

Schematic representation of constructs used in this study. Numbers below each construct indicate the amino acid residue numbers, with the first amino acid of the precursor protein as 1. Proteins from Arabidopsis are specified with “at.” All others are from pea. D1 and D2, The ATPase I and ATPase II domains, respectively; N, the N domain; pr, precursor form; S, the spacer region between the D1 and D2 domains; TP, transit peptide.

Figure 2.

The Hsp93 N domain is important for Hsp93 membrane association. A, In vitro-translated [³⁵S]Met-prHsp93 and various deletion mutants (Tr; lane 1 of all panels) were treated with thermolysin directly (lane 2) or incubated with isolated pea chloroplasts (Cpt) under import conditions for 25 min. After import, a small portion of the chloroplasts were centrifuged through a 40% Percoll cushion to reisolate intact chloroplasts (lane 3). The rest of the reactions were pelleted and resuspended, and the chloroplasts were further digested with thermolysin. After digestion, intact chloroplasts were reisolated through a 40% Percoll cushion (lane 4). Some chloroplasts were further lysed hypotonically and separated into soluble (S; lane 5) and membrane (M; lane 6) fractions by centrifugation. The proteins in the soluble fraction were precipitated by TCA and dissolved with the protein extraction buffer. The membrane fractions were resuspended with the same volume of protein extraction buffer. Protein concentrations were then determined for all samples. Ten micrograms of proteins was loaded in all the chloroplast and soluble fraction lanes (lanes 3–5). For each membrane fraction, the same volume as its corresponding soluble fraction was loaded, so each pair of membrane and soluble fractions shown were derived from the same amount of chloroplasts. Samples were analyzed by SDS-PAGE and fluorography. All Tr lanes contain 20% of the in vitro-translated precursors used for the chloroplast (Cpt) lanes. B, The amount of imported mature proteins in the soluble and membrane fractions (lanes 5 and 6 of A) was quantified. The percentage in the membrane for each construct was calculated and then normalized to the wild type (WT). Data shown are means ± sd (n = 3). Data for prHsp93-ΔS were not included because no signal was obtained in the soluble fraction. [See online article for color version of this figure.]

Figure 3.

atHsp93V-ΔN cannot complement the visible phenotypes of an hsp93V mutant. A, Total proteins were extracted from 14-d-old seedlings of the wild type (WT), hsp93V, and two independent lines of hsp93V mutant transformed with atHSP93V or atHSP93V-ΔN grown on MS plates. Thirty micrograms of proteins from each plant was analyzed by immunoblotting using antibodies against Hsp93 and atTic40. The amount of atTic40 was analyzed as a loading control. B and C, Visible phenotypes of plants grown on an MS plate under a 16-h photoperiod for 14 d (B) or grown on soil under a 12-h photoperiod for 26 d (C). D, Chlorophyll contents of seedlings grown on MS plates under a 16-h photoperiod for 15 d. Data shown are means ± sd (n = 3).

Figure 4.

atHsp93V-ΔN cannot rescue the protein-import defect of an hsp93V mutant. In vitro-translated [³⁵S]Met-prRBCS (A and B) or [³⁵S]Met-prL11 (C and D) were imported into chloroplasts isolated from 14-d-old plate-grown plants of the indicated genotypes for 5 min. Line 2 of the atHSP93V transformants and line 11 of the atHSP93V-ΔN transformants shown in Figure 3 were used. Samples were analyzed by SDS-PAGE, stained with Coomassie blue, and dried for fluorography. Twenty micrograms of proteins was loaded in all lanes of the import samples. Tr, In vitro-translated precursors before import. In B and D, the amount of imported mature proteins was quantified and the amount imported into the wild type (WT) was set as 100%. Data shown are means ± sd of at least three independent experiments. [See online article for color version of this figure.]

Figure 5.

atHsp93V-ΔN does not affect the degradation of chloroplast GS2. Chloroplasts isolated from 27-d-old plate-grown plants of the indicated genotypes were incubated in the presence of ATP and an ATP regeneration system under constant light (A) or in the absence of ATP and the ATP regeneration system in the dark (B). Line 2 of the atHSP93V transformants and line 11 of the atHSP93V-ΔN transformants shown in Figure 3 were used. At 0, 1.5, and 3 h, an equal volume of chloroplast suspensions was taken and analyzed by immunoblotting using antibodies against Toc75 and chloroplast GS2. A total of 1.5 × 10⁶ chloroplasts were loaded in each lane. The amount of Toc75 was analyzed as a loading control. The arrow indicates the degraded GS2 fragment. WT, Wild type.

Figure 6.

atHsp93V-ΔN has reduced membrane association in vivo. Chloroplasts isolated from 14-d-old plate-grown plants of the indicated genotypes were lysed hypotonically and separated into soluble (S) and membrane (M) fractions by centrifugation. The proteins in the soluble fraction were precipitated by TCA and dissolved with the protein extraction buffer. The membrane fractions were resuspended with the same volume of protein extraction buffer. Protein concentrations were then determined for all samples. Eight micrograms of proteins was loaded in all the soluble-fraction lanes. For each membrane fraction, the same volume as its corresponding soluble fraction was loaded so that each pair of membrane and soluble fractions shown were derived from the same amount of chloroplasts. Samples were analyzed by SDS-PAGE. The top half of the gel was analyzed by immunoblotting with antibodies against Hsp93 and atTic40. The bottom half of the gel was stained with Coomassie blue to reveal the location of RBCS. WT, Wild type. [See online article for color version of this figure.]

Figure 7.

atHsp93V-ΔN has reduced association with Tic110. Chloroplasts isolated from 14-d-old plate-grown plants of the indicated genotypes were cross-linked with DSP. Eight hundred microliters of chloroplasts, in the concentration of 1 mg chlorophyll mL⁻¹, was used for the cross-linking. Chloroplasts were lysed, and the enriched envelope membrane fraction was collected and solubilized in 800 μL of immunoprecipitation buffer containing 1% n-decyl-β-d-maltopyranoside. The clarified supernatant of solubilized membranes was immunoprecipitated with the anti-Tic110 antibody. The immunoprecipitates (IP) were analyzed by SDS-PAGE followed by immunoblotting with antibodies as indicated on the left. The loading of the immunoprecipitate lanes was adjusted to make the amount of Tic110 precipitated in the two transgenic lines equal (equivalent to 207 and 120 μL of solubilized membranes from the atHSP93V and atHSP93V-ΔN transgenic lines, respectively, on the gel shown). 4% Env, Four percent of the solubilized membranes used in the corresponding immunoprecipitate lane. The atHSP93V-ΔN transgenic line is pale green and contains a higher number of chloroplasts per milligram of chlorophyll and thus a higher concentration of Tic110 in the solubilized membranes.

Figure 8.

N-domain deletion does not affect the ATPase activity of Hsp93. A, His₆-Hsp93 and His₆-Hsp93-ΔN were expressed and purified from the soluble fraction of E. coli and analyzed by SDS-PAGE and Coomassie blue staining. B, His₆-Hsp93 and His₆-Hsp93-ΔN have similar ATPase activities. Data shown are means ± sd of two independent experiments. [See online article for color version of this figure.]