SUPPLEMENTARY INFORMATION Reengineering Protein Interfaces Yields Copper-Inducible Ferritin Cage Assembly Dustin J. E. Huard, Kathleen M. Kane and F. Akif Tezcan* Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093-0356 Supplementary Information contains: - Supplementary Tables 1-6 - Supplementary Figures 1-17 S1
Supplementary Results Supplementary Table 1. MALDI-TOF characterization of various ferritin mutants. For the actual mass spectra of unmodified and modified proteins, refer to Supplementary Figs. 2 and 3, respectively. Calculated Mass (Da) Observed Mass (Da) ΔC* 21,056 21,058 C53 ΔC* 21,031 21,034 C53 ΔC*-AEDANS 21,337 21,338 C53 ΔC*-Atto 590 21,758 21,764 (modified) 21,064 (unmodified) C116 ΔC* 21,030 21,055 C116 ΔC*-AEDANS 21,336 21,355 4His-ΔC* 21,069 21,086 C53 4His-ΔC* 21,044 21,047 C53 4His-ΔC*-AEDANS 21,350 21,333 (modified) 21,069 (unmodified) ΔH-MIC1 21,011 20,980 (H56L/H63R/H67E) MIC1 21,024 20,988 C53 MIC1 20,999 21,005 C53 MIC1-AEDANS 21,305 21,316 S2
Supplementary Table 2. X-ray data collection and refinement statistics. A single crystal was used during data collection for each variant. Cu-4His-ΔC* Cu-MIC1 Apo-MIC1 Cu- C53 MIC1- AEDANS Data collection Space group F432 F432 F432 F432 Cell dimensions a, b, c (Å) 179.3 179.6 181.1 180.4 α, β, γ ( ) 90 90 90 90 Resolution (Å)* 103.5-1.85 (1.95-1.85) 104.5-1.90 (2.00-1.90) 104.5-2.30 (2.42-2.30) 104.1-2.50 (2.64-2.50) R sym or R merge (%)* 6.0 (21.5) 10.2 (49.2) 9.7 (39.7) 8.5 ( 40.5) I / σi* 9.2 (3.2) 5.6 (1.4) 6.8 (1.9) 7.5 (1.9) Completeness (%)* 100 (100) 99.8 (100) 100 (100) 100 (100) Redundancy Refinement Resolution (Å) 51.8-1.85 89.8-1.90 90.5-2.30 40.4-2.50 No. unique reflections 21639 20051 11847 9197 R work / R free 16.9/19.7 21.8/24.1 22.0/27.2 22.4/30.0 No. atoms Protein 1416 1448 1446 1416 Ligand/ion 7 17 5 27 Water 215 165 93 51 B-factors Protein 15.3 27.3 25.6 38.1 Ligand/ion 29.7 37.5 34.1 61.7 Water 31.8 42.3 31.7 40.2 R.m.s. deviations Bond lengths (Å) 0.006 0.007 0.008 0.009 Bond angles ( ) 0.842 0.886 0.996 0.953 *Highest-resolution shell is shown in parentheses. S3
Supplementary Table 3. Parameters for sedimentation velocity measurements. MIC1 Monomer Cu-induced MIC1 cage apo-mic1 cage Cureconstituted MIC1 cage Concentration (µm) Buffer Density (g/ml) Buffer Viscosity (poise) Vbar (ml/g) Frictional Ratio Theoretical Sedimentation Coefficient (S) Observed Sedimentation Coefficient (S) 59.1 1.00724 0.010412 0.72 1.23867 2.091 2.16107 50.0 1.00706 0.010401 0.72 1.24652 18.22 19.9598 57.7 1.00724 0.010412 0.72 1.20000 18.14 19.3976 57.7 1.00706 0.010401 0.72 1.27949 18.24 20.3815 S4
Supplementary Table 4. Fe content of Fe-reconstituted ferritin variants. Data represent mean values ± s.d. of three independent measurements. Variant Fe Atoms/Cage ΔC* 317.6 ± 9.5 4His-ΔC* 227.0 ± 7.6 Cu-induced MIC1 cage 325.3 ± 16.2 S5
Supplementary Table 5. List of primers used in constructing site-directed mutants of ferritin. Variant Mutation Primer Sequence (5-3 ) (order of addition) ΔC* K86Q + C90E GCAGGACATTCAGAAGCCGGATGAGGACGATTGGG CCCAATCGTCCTCATCCGGCTTCTGAATGTCCTGC C102A GGCCTGAATGCGATGGAGGCGGCGCTGCATCTGG CCAGATGCAGCGCCGCCTCCATCGCATTCAGGCC C130A GAATGATCCGCACCTGGCGGATTTCATCGAAACGC GCGTTTCGATGAAATCCGCCAGGTGCGGATCATTC 4His-ΔC* R63H GCCATGAAGAACACGAGCACGCAGAG CTCTGCGTGCTCGTGTTCTTCATGGC L56H GAAAAACTTTGCGAAATACTTCCATCATCAGAGCCATGAAGAACACG CGTGTTCTTCATGGCTCTGATGATGGAAGTATTTCGCAAAGTTTTTC E67H TGAAGAACACGAGCACGCACATAAACTGATGAAACTGCAGA TCTGCAGTTTCATCAGTTTATGTGCGTGCTCGTGTTCTTCA MIC1 Y39E GTTTATCTGTCTATGAGCGAGTACTTCGACCGTGACGATG CATCGTCACGGTCGAAGTACTCGCTCATAGACAGATAAAC N74E CTGATGAAACTGCAGGAGCAGCGTGGTGGCCGC GCGGCCACCACGCTGCTCCTGCAGTTTCATCAG P88A GCAGGACATTCAGAAGGCGGATGAGGACGATTG CAATCGTCCTCATCCGCCTTCTGAATGTCCTGC H173A MIC1 H173A AGTACCTGTTTGACAAGGCCACCTTGGGTGACTCCG CGGAGTCACCCAAGGTGGCCTTGTCAAACAGGTACT D131A/E134A MIC1 D131A/E134A CCGCACCTGGCGGCTTTCATCGCAACGCATTACCTG CAGGTAATGCGTTGCGATGAAAGCCGCCAGGTGCGG C53 MIC1 K53C GTGGCCCTGAAAAACTTTGCGTGCTACTTCCATCATCAGAGCCAT ATGGCTCTGATGATGGAAGTAGCACGCAAAGTTTTTCAGGGCCAC C116 ΔC* E116C GTCAATCAGAGCCTGCTGTGCCTGCACAAGCTGGCAACG CGTTGCCAGCTTGTGCAGGCACAGCAGGCTCTGATTGAC C53 ΔC* K53C GTGGCCCTGAAAAACTTTGCGTGCTACTTCCTGCATCAGAGCCAT ATGGCTCTGATGCAGGAAGTAGCACGCAAAGTTTTTCAGGGCCAC S6
Supplementary Table 6. Crystallization conditions for ferritin variants. Protein Crystal Temp. Protein Precipitant Cu-4His-ΔC* 25 o C 712 µm apo protein in standard buffer 50 mm Tris (ph 8.0), 5 mm CaCl 2, 200 µm CuCl 2 Cu-MIC1 25 o C 642 µm in standard buffer 50 mm Tris (ph 8.0), 5 mm CaCl 2, 8% PEG 1900 MME, 700 µm CuCl 2 Apo-MIC1 25 o C 666 µm in standard buffer 50 mm Tris (ph 8.0), 10 mm CaCl 2, 4% PEG 400, 20 mm EDTA (in standard Cu-AEDANS- C53 MIC1 25 o C 666 µm in standard buffer buffer) 50 mm Tris (ph 8.0), 5 mm CaCl 2, 50 mm NaCl, 10% PEG 3350, 350 µm CuCl 2 S7
Supplementary Figure 1. SDS-PAGE characterization of ferritin mutants. All mutants were judged to be >85% pure. S8
Supplementary Figure 2. MALDI mass spectra of ferritin variants. For a list of corresponding masses, see Supplementary Table 1. S9
Supplementary Figure 3. MALDI mass spectra of ferritin variants subjected to chemical modification. For a list of corresponding masses, see Supplementary Table 1. For the SDS-PAGE characterization of the modified variants, see Supplementary Figure 4. S10
Supplementary Figure 4. Chemical modification of ferritin variants with IAEDANS and Atto 590 maleimide as characterized by SDS-PAGE. Prior to electrophoresis, all samples were treated with EDTA to remove any bound or unbound Cu that may quench IAEDANS fluorescence. The gel was first imaged using a UVP UV transilluminator equipped with a BioDoc-It imaging system to detect fluorescence by IAEDANS and Atto 590 (top). It was then stained with Coomassie for visualizing the protein bands (bottom). Lane 1: Molecular weight marker; Lane 2: ΔC* cage treated with IAEDANS (negative control); Lane 3: C53ΔC* cage treated with IAEDANS; Lane 4: C116ΔC* cage treated with IAEDANS (positive control); Lane 5: C534His-ΔC* cage treated with IAEDANS; Lane 6: C53MIC1 monomer treated with IAEDANS; Lane 7: C53ΔC* cage pretreated with Cu prior to labeling with IAEDANS; Lane 8: C116ΔC* cage pretreated with Cu prior to labeling with IAEDANS; Lane 9: C53 4His-ΔC* cage pretreated with Cu prior to labeling with IAEDANS; Lane 10: C53ΔC* cage treated with Atto 590 maleimide. S11
Supplementary Figure 5. Alternate views of the ferritin cage. a) View down the C3 symmetry axis. b) View down the C4 symmetry axis. One C2 dimer pair is highlighted in magenta as in Fig. 2. S12
Supplementary Figure 6. Identification of interfacial sites for grafting a stable Cu II coordination motif. a) The C 2 symmetric Cu 2 :MBCP1 2 structure directed by Cu II coordination to bis-histidine motifs on the MBPC1 surface. The resulting, stable Cu-coordination sites are defined by pairwise C α and C β distances (listed in the table below) among the four coordinating histidines. b) Pairwise C α and C β distances among residues 56, 60, 63, and 67 chosen in the HuHF C 2 interface for grafting a stable 4His Cu coordination site. S13
Supplementary Figure 7. Intersubunit interactions in the C 2 interface. The C 2 interface of ferritin viewed parallel to the C 2 symmetry axis. S14
Supplementary Figure 8. TEM imaging of ferritin variants. Images on the left and the right columns are obtained with and without uranyl acetate staining, respectively. The unstained samples clearly show the presence of Fe mineral within all variant cages. S15
Supplementary Figure 9. Backbone superposition of the C 2 dimers of native ferritin (green) and 4His-ΔC* (magenta). S16
Supplementary Figure 10. Sedimentation velocity profile of MIC1. The protein sample (569 µm in concentration) was prepared in the standard buffer supplemented with 10 mm EDTA. Despite its high concentration, MIC1 still remains monomeric (peak sedimentation coefficient of 2.2 S) in the absence of metals. S17
Supplementary Figure 11. Hydrodynamic properties of isolated MIC1 and its metal-mediated oligomers. a) (top) Size-exclusion chromatogram of a MIC1 solution obtained following isolation from inclusion bodies and metal-affinity chromatography; the majority of the protein is in a monomeric state with a minor fraction existing as large, but soluble aggregates that elute in the dead volume. (bottom) Size-exclusion chromatogram of MIC1 exchanged into solutions containing equimolar Cu II (blue trace), Ni II (green trace) and Zn II (black trace). b) Sedimentation velocity profiles of various fractions (indicated with numbers 1-4 in (a)) isolated via size-exclusion chromatography. S18
Supplementary Figure 12. Hydrodynamic properties of isolated D131A/E134A MIC1 and its Cu-mediated oligomers. a) Size-exclusion chromatogram of a D131A/E134A MIC1 solution obtained following isolation from inclusion bodies and metalaffinity chromatography (red trace), and upon exchange of the monomeric fraction into a Cu-containing solution (blue trace). b) Sedimentation velocity profiles of various fractions (indicated with numbers 1-2 in (a)) isolated via size-exclusion chromatography. In contrast to MIC1 (Supplementary Fig. 11), the majority of the as-isolated protein is found as large aggregates that elute in the dead volume of the SEC column; a small fraction is separated as monomeric species (elution time ~300 min). This monomeric fraction is nearly quantitatively converted into a 24meric cage upon Cu II binding (elution time ~155 min). S19
Supplementary Figure 13. Hydrodynamic properties of isolated H173A MIC1 and its Cu-mediated oligomers. a) Sizeexclusion chromatogram of a H173A MIC1 solution obtained following isolation from inclusion bodies and metal-affinity chromatography (red trace), and upon exchange of the monomeric fraction into a Cu-containing solution (blue trace). b) Sedimentation velocity profiles of various fractions (indicated with numbers 1-3 in (a)) isolated via size-exclusion chromatography. Judging from the SEC elution profiles, the monomeric fraction is converted upon reconstitution with Cu II into higher-order species (24mer or larger, elution time ~150 min), and what appears to be a dimeric form (elution time ~280 min, approximate concentration at elution is <50 µm). Only when this dimeric fraction is concentrated to above 200 µm for SV measurements (d), it is nearly fully converted to a 24meric cage. S20
Supplementary Figure 14. Backbone superpositions of Cu-4His-ΔC*, Cu-MIC1, and apo-mic1 structures. a) Cu-4His-ΔC* (green) vs. Cu-MIC1 (magenta). b) Cu-MIC1 (magenta) vs, apo-mic1 (cyan). S21
Supplementary Figure 15. Cu II coordination in the C 4 pore of the Cu-MIC1 cage. The 2F o -F c map is contoured at 1.5 σ. S22
Supplementary Figure 16. Chemical and thermal unfolding of various MIC1 species as monitored by CD spectroscopy at 222 nm. The coloring scheme matches that of Fig. 5 and is as follows: MIC1 monomer (magenta), Cu-induced MIC cage (blue), EDTA-treated MIC1 cage (green), Cu-reconstituted MIC1 cage (black). For both the chemical (a), and the thermal unfolding measurements, the data represent mean values ± s.d. of three independent measurements. S23
Supplementary Figure 17. Linear relationship between the unfolding free energy of various MIC1 cage species and GuHCl concentrations used to obtain cage stabilities at [GuHCl] = 0 M (as described in Online Methods). The coloring scheme is the same as that in Fig. 5. The data represent mean values ± s.d. of three independent measurements. S24