Supporting Information

Files in this Data Supplement:

Fig. 6. Structural model of the ClpXP Complex. (A) Shown are the domain boundaries of ZBD and AAA⁺ in ClpX. (B) A model of the ClpXP complex. The structures shown are based on the solved structure of E. coli ClpP (1, 2), a model for the hexamer of the AAA⁺ domain of E. coli ClpX based on the solved monomeric AAA⁺ domain of H. pylori ClpX (3) and the hexameric E. coli HslU structure (4), and on our proposed trimer-of-dimers model for the ZBD of E. coli ClpX (5). Note that ClpX oligomers can bind to both ends of the ClpP cylinder. Structures were drawn using PyMOL (http://pymol.sourceforge.net).

Fig. 7. Binding of ZBD₂ and AAA⁺ to the arrays of lO and MuA. (A) Purified ZBD₂ or (B) AAA⁺ in the absence of nucleotides were incubated with peptide arrays containing a total of 3717 13-mer peptides spanning the sequences of 26 proteins, with a frame shift of two amino acids. The arrays for lO and MuA are shown as examples. Labels to the left give the residue number of the N-terminal amino acid of the first peptide of each row, while labels to the right give the number of the last peptide in each row.

Fig. 8. Effect of ATP on the binding of AAA⁺ to the peptide array. Purified AAA⁺ was incubated with peptide arrays in the absence (top panel) or presence (bottom panel) of 5 mM ATP. The array of H-2 class I histocompatibility antigen D-B alpha chain precursor [H-2D(B)] is shown as an example. Labels to the left give the residue number of the N-terminal amino acid of the first peptide of each row, while labels to the right give the number of the last peptide in each row.

Fig. 9. Peptide-ClpX interaction. (A) The ClpP-dependent degradation of lO (3.9 mM monomer) mediated by ClpX (1 mM monomer) and ClpP (1.2 mM monomer) was monitored by SDS/PAGE in the absence of peptide (top lane), in the presence of 50 m M lO^49-63 (second lane), 50 mM MuA^653-663 (third lane), or 50 mM SspB^154-165 (bottom lane) peptides. Aliquots were removed from the degradation mixture at the indicated time points. The chaperone was the last component added to the reaction mixture. (B) ELISA assays for the binding of 25 mM ZBD₂ or 1 mM AAA⁺ (monomer concentration), respectively, to 100 mg/ml of N-terminally tagged GFP or GFP-SsrA. Data points are the average of two experiments and normalized to GFP-SsrA absorbance. DVG-GFP and IYY-GFP correspond to His-tagged GFP with the peptides DVGVLVISARKGE (DVG) and IYYITGESLKAVE (IYY), respectively, introduced at the N terminus. Based on Table 1, the DVG peptide contains patterns ranked 3 and 7 preferentially bound by ZBD₂ and patterns ranked 4, 6, 9, 12, 16, and 18 preferentially bound by AAA⁺. The IYY peptide contains patterns ranked 1, 2, 4, 8, and 15 preferentially bound by ZBD₂ and the pattern ranked 1 preferentially bound by AAA⁺.

Fig. 10. Mutational analysis to determine the SspB₂ binding site on ZBD in ClpX. The ClpP-dependent degradation of GFP-SsrA (3.9 mM monomer) mediated by ClpX or different ClpX mutants (each at 1 mM monomer) and ClpP (1.2 mM monomer) are shown in the presence of 0 (a), 0.025 mM (b), and 0.165 mM (c) of SspB₂.

Fig. 11. Stability of different ZBD₂ mutants. (A) CD spectra at 10°C of ZBD₂ WT and different mutants (30 mM). (B) Thermal denaturation curves monitored by CD at 220 nm. (C) The oligomeric states of the ZBD₂ mutants were analyzed by size exclusion chromatography on a Superdex 200 HR 10/30 column. Molecular mass standards, in kDa, are shown along the top. 'VV' refers to void volume. Due to the column resolution and the structure of wild-type ZBD₂, the protein elutes near the 29 kDa marker even though the molecular mass of the dimer is about 14 kDa.

Fig. 12. Determination of the binding affinity of SspB^154-165 to ZBD₂ WT and ZBD₂(Y34W) using ITC. Shown are raw ITC (A and C) and integrated heat data (B and D) of addition of SspB^154-165 to ZBD₂. Titrations were done at 20°C. Solid lines in (B) and (D) represent the fit to a model of a single type of n identical and independent binding sites. The n and K_d resulting from the fits are given.

Fig. 13. Raw data and model of the binding affinity of SspB₂ and SspB^154-165 to ZBD₂ using DPI. Shown are raw sensor data of addition of free SspB^154-165 (A) or free SspB₂ (B) to immobilized ZBD₂. T, D, and M refer to thickness, density, and mass, respectively. Different additions of SspB^154-165 or SspB₂ are indicated with dashed lines. The concentrations of injected SspB^154-165 or SspB₂ are shown on the top x axis. (C) Shown are examples of the binding mode of SspB₂ to ZBD₂ drawn to scale. ZBD₂ on the sensor chip is drawn in light gray with the charged surface facing the chip. SspB₂ is shown as 2 triangles with the C-terminal tails binding to ZBD₂. Numbers on the left represent the thickness resulting from SspB₂ binding at saturation. Thickness measurements have an estimated error of 5 - 10%. Other modes of binding cannot be excluded. (D) The areas occupied by immobilized ZBD₂ and by bound SspB₂ in the low and high affinity interactions are given.

Supporting Text

Purified ZBD₂ (10 mg/ml) or AAA⁺ were incubated in buffer E (25 mM TrisHCl, pH 7.5, 150 mM NaCl, 0.1% Tween, 0.1% BSA, and 0.002% thimerosal) with peptide arrays previously blocked with 2% BSA in buffer E for 1 h. The bound protein was transferred to a nitrocellulose membrane using a Hoefer TE Series semidry blotter (Amersham Pharmacia Biosciences), at 0.75 mA/cm² for 30 min using buffer F (25 mM TrisHCl, pH 8.3, 192 mM glycine, 0.1% SDS, and 20% methanol). ZBD₂ and AAA⁺ were detected using rabbit polyclonal a-ClpX antibodies, and then visualized on film using Protein LA-Peroxidase (Sigma) and ECL substrate (Amersham Pharmacia Biosciences). Three independent peptide array incubation experiments were analyzed.

After scanning the film, spot volumes or intensities were measured using ImageQuant 5.0 (Molecular Dynamics) software. All of the spot volumes were then averaged and normalized. Peptides that had spot intensities greater than 75% were considered binders, otherwise they were classified as nonbinders. The number of each type of amino acid was counted in the binder and nonbinder sequences and the percent occurrence of each amino acid in each group (binders and nonbinders) and the normalized percent occurrences were calculated, as follows:

% occurrence in binders = (no. of aa 'x' in binders)/(total no. of aa in binders) x 100 (1)

Normalized % occurrence in binders = (%occurrence of aa 'x' in binders)/(%occurrence of aa 'x' on array) x 100 (2)

The z-test for comparing two proportions was used to determine whether the percent occurrences of the specific amino acids in the binder or nonbinder groups differed from their respective occurrences on the array at the 95% significance level.

The software Teiresias was used to search for recurring sequence patterns in the binder and nonbinder peptide groups. For each sequence pattern, the frequencies of the pattern in the binder (F_B) and nonbinder (F_NB) groups were determined by dividing the number of peptides in each group that contain the pattern by the total number of peptides in that group. The ratio R (R = F_B/F_NB) was calculated. Patterns that pass the z-test at 99.9% confidence level and that fulfill the criteria F_B ³0.045 and R ³4.0 were selected. The most general form of a sequence pattern is shown (Table 1). For example, a pattern will be reported as having [ILMV] for an amino acid instead of the version that specifically shows I, L, M, or V if both the general and specific forms of the pattern satisfy one of the three criteria listed above. The Teiresias and statistical analyses were implemented in a program developed in-house termed Sequence Array Analyzer.

1. Wang J, Hartling JA, Flanagan JM (1997) Cell 91:447-456.

2. Gribun A, Kimber MS, Ching R, Sprangers R, Fiebig KM, Houry WA (2005) J Biol Chem 280:16185-16196.

3. Kim DY, Kim KK (2003) J Biol Chem 278:50664-50670.

4. Bochtler M, Hartmann C, Song HK, Bourenkov G P, Bartunik HD, Huber R (2000) Nature 403:800-805.

5. Donaldson LW, Wojtyra U, Houry WA (2003) J Biol Chem 278:48991-48996.