Supplementary Information. for. Defining the geometry of the two-component proteasome degron

Transcription

1 Supplementary Information for Defining the geometry of the two-component proteasome degron Tomonao Inobe 1,2, Susan Fishbain 1, Sumit Prakash 1, and Andreas Matouschek 1 1 Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, 2205 Tech Drive, Evanston, IL and Robert H. Lurie Comprehensive Cancer Center, Northwestern University, 303 East Superior Street L3-125, Chicago, IL 60611, USA 2 Laboratory for Structural Neuropathology, RIKEN Brain Science Institute, Wako, Saitama , Japan Correspondence should be addressed to A. M. (matouschek@northwestern.edu) 1

2 Supplementary Results Supplementary Figure 1. Proteasome substrates are folded prior to degradation. The DHFR domains in the proteasome substrates are folded and addition of the DHFR ligand methotrexate stabilizes these proteins against degradation as expected 1. (a) Substrates that consisted of a Ub 4 -tagged DHFR domain followed by a 44 amino acid initiation region were protected from degradation by the addition of methotrexate. Degradation reactions were performed at 30 C with 50 nm proteasome in the presence (black circles) or absence (red diamonds) of 100µM methotrexate. (b) In longer substrates in which a titin domain is inserted between the initiation region and the DHFR domain, degradation should proceeded from the initiation region until the proteasome encounters the methotrexate-stabilized DHFR domain. To test this prediction, a substrate consisting of an N-terminal UbL domain followed by DHFR, an I27 domain, and a C-terminal 102 amino acid-long initiation region was degraded at 30 C by 50 nm proteasome in the presence (black filled circle) or absence (red diamonds) of 100 µm methotrexate. In the presence of methotrexate, the full-length substrate was converted into a smaller fragment (black open circles), as expected. Methotrexate did not affect the degradation rate. Data points represent mean values and error bars show standard errors calculated from three experiments. 2

3 a 100 b 100 Partial degradation product % Remaining % Remaining Time (min) Time (min) Supplementary Figure 1 Inobe et al.

4 Supplementary Figure 2. The dependence of degradation rates on initiation region length is not due to details of the amino acids sequence of the initiation region. To test whether the initiation region length dependence of degradation rates is dominated by amino acid sequence preferences of the proteasome, we compared behavior of initiation regions derived from two unrelated genes. To do so we constructed two sets of substrates consisting of a DHFR domain targeted to the proteasome by an N-terminal Ub 4 tag (red diamonds) or an UbL tag (green circles) and unstructured C-terminal tails of different lengths to serve as initiation regions. The tails were derived either from Saccharomyces cerevisiae cytochrome b 2 (solid symbols, solid line) or subunit 9 of the Neurospora crassa F o ATPase (open symbols, dashed line). Initial degradation rates by 50 nm proteasome at 30 C are plotted against the length of the initiation region. The degradation rates show similar dependences on initiation region length for both amino acid sequences. Data points represent mean values and error bars show standard errors calculated from three repeat experiments. 4

5 Initial degradation rate (%/min) DHFR DHFR Length (aa) Supplementary Figure 2 Inobe et al.

6 Supplementary Figure 3. Spacing requirement for Ub 4 degrons is not determined by the number of ubiquitin moieties in the tag. It is possible that the proteasome recognized the Ub 4 tag by its most N-terminal ubiquitin moiety, perhaps through a UbL receptor. In this case, the remaining ubiquitin moieties in the Ub 4 tag could function as spacers in the same way as the titin I27 insertions. To test this hypothesis, we constructed a series of substrate proteins in which the ubiquitin tag consisted of only one (a) or two (b) ubiquitin moieties. In these proteins, the ubiquitin tags and an initiation region of 44 amino acids in length were attached to a DHFR domain and separated by zero, one, two, or three titin I27 domains. (a,b) Degradation kinetics by 50 nm proteasome at 30 C for substrates with no (red), one (orange), two (black), or three (green) I27 domain insertions targeted to the proteasome by either one (a) or two ubiquitin moieties (b). Substrates with single ubiquitin tags were not degraded, presumably because the single ubiquitin moiety is not recognized effectively by the proteasome, as expected 2. Thus, it is unlikely that any ubiquitin tags, including the Ub 4 tag are recognized by the UbL receptors. Substrates with di-ubiquitin tags were degraded efficiently and displayed the same degron spacing requirements as substrates with Ub 4 tags. Therefore, the ubiquitin moieties in the Ub 4 degron do not serve as spacers and the differences in degron geometries required for efficient degradation with Ub 4 and UbL tags are probably due to the locations of their receptors on the proteasome. Data points represent mean values and error bars show standard errors calculated from three repeat experiments. 6

7 a % Remaining Ub 1 44 aa b Time (min) % Remaining Time (min) Ub 2 44 aa Supplementary Figure 3 Inobe et al.

8 Supplementary Figure 4. Dependence of degradation rates on degron spacing is not affected by UbL-UBA adaptor proteins remaining in the proteasome preparations. Proteasome preparations are expected to contain substoichiometric amounts of UbL-UBA proteins (e.g., [3]) but those present in the proteasome preparation used here do not affect the observed dependence of degradation rates on degron component spacings. Control experiments with proteasome prepared from a Δrad23,Δdsk2,Δddi1 disruption strain, which lacks UbL-UBA proteins, show the same dependence of degradation rates on spacing for ubiquitin (a) and UbL degrons (b) as observed for experiments performed with proteasome from a wildtype strain. Plots of degradation rates of Ub 4 -tagged (a) and UbL-tagged (b) substrates with 44 amino acidlong initiation regions against the number of inserted I27 domains show equivalent results for proteasome purified from yeast strain YYS40 (red diamonds) and from yeast strain SY488 in which the genes for the UbL-UBA proteins Rad23, Dsk2 and Ddi1 were disrupted (black circles). Degradation reactions were performed at 30 C. Data points represent mean value and error bars show standard errors calculated from three repeat experiments. The strain SY488 (RPN11-TEV-ProA::HIS rad23::ura dsk2::trp ddi1::kan) was kindly provided by Suzanne Elsasser and Dan Finley, Department of Cell Biology, Harvard University Medical School. The triple deletion proteasome was purified by following a protocol described previously 4, with modifications. Cell pellets were resuspended in 2 volumes of lysis buffer (50 mm Tris-HCl [ph 7.5], 5 mm MgCl 2, 10% glycerol, 2 mm DTT, 2 mm ATP, 20 mm creatine phosphate, 0.2 mg/ml creatine phosphokinase) and lysed by pressure homogenizer. The lysate was cleared by centrifugation followed by filtration, incubated with IgG resin (Sigma) for 1 hr at 8

9 4 C, and the resin washed with 50 bed volumes of wash buffer (50 mm Tris-HCl [ph 7.5], 5 mm MgCl 2, 10% glycerol, 50mM NaCl, 1mM DTT, 1 mm ATP). The proteasome was eluted from the beads by cleavage with AcTEV protease (Invitrogen). The IgG resin was washed with 3 volumes of TEV elution buffer (50 mm Tris-HCl [ph 7.5], 5 mm MgCl 2, 10% glycerol, 1 mm ATP, 0.5 mm TCEP) and then incubated with elution buffer containing 150 units of AcTEV protease at room temperature for 1 hr. The proteasome was then washed out of the column with elution buffer and the AcTEV protease was removed by incubation with talon affinity resin (Qiagen) at 4 C for 15 min. Proteasomes were stored at -80 C in 15% glycerol. 9

10 a Ub 4-tagged substrate Initial degradation rate (%/min) ( ) n 44 aa Number of I27 domains wt rad23 dsk2 ddi1 b UbL-tagged substrate ( ) n 44 aa Initial degradation rate (%/min) wt Number of I27 domains rad23 dsk2 ddi1 Supplementary Figure 4 Inobe et al.

11 Supplementary Figure 5. The stability of the titin domain does not affect degradation reactions significantly. Proteasome substrates were assembled from I27 domains in which cysteine residues were replaced with alanine to avoid complications from oxidation and nonspecific labeling in the FRET experiments. Constructs containing I27 domains with the wildtype cysteine residues intact (red diamonds) are degraded by the proteasome as efficiently as those with the cysteine to alanine replacement (black circles) whether they are targeted for degradation through a Ub 4 tag (a) a UbL tag (b). Substrates consisted of an N-terminal proteasome binding tag followed by DHFR, an I27 domain, and a C-terminal 102 amino acid-long initiation region; degradation reactions were performed at 30 C. Data points represent mean values and error bars show standard errors calculated from three repeat experiments. 11

12 a % Remaining Time (min) b % Remaining Time (min) Supplementary Figure 5 Inobe et al.

13 Supplementary Methods Substrate proteins Proteins were derived from two different domains: the 27th Ig (I27) domain of the giant muscle protein titin 5 and Escherichia coli DHFR 6. I27 contained the mutation Cys47/63 Ala to remove potentially reactive sulfhydryl groups. The proteasome targeting signal was either a Ub 4 tag, which consisted of four ubiquitin moieties containing the mutation Gly76 Val connected by their N and C-termini through a six-residue linker (Gly-Ser-Gly-Gly-Gly-Gly) 7, or a UbL tag, which consisted of residues 1-77 of Saccharomyces cerevisiae Rad23 3,8. Either tag was connected to the N-terminus of DHFR through a linker; the linker sequence was Gly-Ser-Leu- Ser-Ala-Met-Gly for the Ub 4 tag and Thr-Ser-Leu-Ser-Ala-Met-Gly for the UbL tag. The initiation regions were derived from Saccharomyces cerevisiae cytochrome b 2 or Neurospora crassa subunit 9 of the F o ATP synthase. Specifically, the cytochrome b 2 constructs consisted of a short linker (Pro-Arg), followed by a stretch of cytochrome b 2, and ended with a hexahistidine tag. The cytochrome b 2 sequences consisted of the first 6, 15, 21, 26, 36, 55, or 94 amino acids of cytochrome b 2. All lysine residues in the cytochrome b 2 sequence were mutated to arginine. The resulting sequences were referred to by their total length as 15, 24, 29, 34, 44, 64, and 102 amino acid initiation regions. The F o ATPase subunit 9 constructs consisted of a short linker and the first 13, 23, 33, 49, or 69 amino acids of subunit 9 of the F o ATPase or of amino acids 1 to 69 of subunit 9 of the F o ATPase followed by the same sequence again. In contrast to the cytochrome b 2 constructs, the Lys residues were not mutated and no hexahistidine tag was attached. The resulting sequences 13

14 were referred to by their total length as 15aa, 25aa, 35aa, 51aa, 71aa, and 142aa initiation regions. I27 domains were inserted between DHFR and the initiation region and were connected to each other in frame through a two-residue linker (Gly-Ser). For the FRET experiments, the sequence coding for egfp 9 was fused to the N-terminus of DHFR through an eight-residue linker (His-Met-Leu-Cys-Thr-Pro-Ser-Arg), and a sequence coding for Cys-Cys-Gly-Pro-Cys- Cys was inserted into the 44 amino acid long cytochrome b 2 initiation region immediately before the histidine tag. All genes were constructed using standard molecular biology techniques and verified by DNA sequencing. 14

15 References 1. Johnston, J.A., Johnson, E.S., Waller, P.R. & Varshavsky, A. Methotrexate inhibits proteolysis of dihydrofolate reductase by the N-end rule pathway. J. Biol. Chem. 270, (1995). 2. Thrower, J.S., Hoffman, L., Rechsteiner, M. & Pickart, C.M. Recognition of the polyubiquitin proteolytic signal. EMBO J. 19, (2000). 3. Schauber, C. et al. Rad23 links DNA repair to the ubiquitin/proteasome pathway. Nature 391, (1998). 4. Leggett, D.S. et al. Multiple associated proteins regulate proteasome structure and function. Mol. Cell 10, (2002). 5. Politou, A.S., Gautel, M., Pfuhl, M., Labeit, S. & Pastore, A. Immunoglobulin-type domains of titin: same fold, different stability? Biochemistry 33, (1994). 6. Rood, J.I., Laird, A.J. & Williams, J.W. Cloning of the Escherichia coli K-12 dihydrofolate reductase gene following mu-mediated transposition. Gene 8, (1980). 7. Stack, J.H., Whitney, M., Rodems, S.M. & Pollok, B.A. A ubiquitin-based tagging system for controlled modulation of protein stability. Nat. Biotechnol. 18, (2000). 8. Watkins, J.F., Sung, P., Prakash, L. & Prakash, S. The Saccharomyces cerevisiae DNA repair gene RAD23 encodes a nuclear protein containing a ubiquitin-like domain required for biological function. Mol. Cell. Biol. 13, (1993). 9. Yang, T.T., Cheng, L. & Kain, S.R. Optimized codon usage and chromophore mutations provide enhanced sensitivity with the green fluorescent protein. Nucleic Acids Res. 24, (1996). 15