*3.1. Preparation of Wild-Type and Mutated Proteasome* α *Subunits*

Human proteasome α6 short isoform [*PSMA1* (P25786); residues 1–263] and α7- [*PSMA3* (P25788); residues 1–255] subunits were expressed and purified as described previously [23,24,26]. Genes encoding proteasome α1 [*PSMA6* (P60900); residues 1–246] and α4 [*PSMA7* (P60900); residues 1–248] were subcloned into the *Nde*I and *Sal*I sites of pET28b (Merck Millipore, Burlington, MA, USA), whereas the α2 gene [*PSMA2* (P25787); residues 1–234] was subcloned into the *Nde*I and *Xho*I sites of pRSFDuet-1 vector (Merck Millipore). In addition, genes encoding α3 [*PSMA4* (P25789); residues 1–261] and α5 [*PSMA5* (P28066); residues 1–241] were subcloned into the *Bam*HI and *Xho*I or *Sal*I sites of modified pCold-I and pCold-GST vectors (TaKaRa Bio Inc., Kusatsu, Japan), respectively, which contain the TEV protease cleavage site preceding the target genes. Monomeric mutants of α7 (designated as α7\* and α7F14A) were created via truncation of 22 N-terminal residues or introduction of the F14A mutation, respectively. The mutated α7 genes were subcloned into the *Nde*I and *Xho*I sites of the pET28b vector. In contrast, the single-ring mutant α7SR was generated by introducing the S96D, S100D, F102R, and Y104R mutations using the wild-type construct in pRSFDuet-1. All expression plasmids were introduced into *Escherichia coli* BL21-CodonPlus (DE3)-RIL (Agilent Technologies, Santa Clara, CA, USA).

For producing recombinant proteins, the *E. coli* cells harboring the expression plasmids were grown in Luria–Bertani medium containing 15 μg/mL kanamycin or 50 μg/mL ampicillin. The α7SR mutant was purified as employed for the wild-type α7. Briefly, except for α2, the recombinant proteins were purified from the soluble fraction obtained by sonication and centrifugation. The resultant cell lysates were subjected to affinity chromatography [Ni+-charged Chelating Sepharose or Glutathione Sepharose 4B (GE Healthcare, Chicago, IL, USA)], and further purified using anion-exchange (RESOURCE Q, GE Healthcare) and size-exclusion (HiLoad 26/60 Superdex 75 or 200 pg; GE Healthcare) columns. The α2 was purified from the inclusion bodies and refolded according to standard dilution methods using a buffer containing 20 mM Tris-HCl (pH 8.0), 400 mM L-arginine, 250 mM NaSCN, 1 mM oxidized glutathione, and 5 mM reduced glutathione. The refolded protein was further purified using a HiLoad 26/60 Superdex 75pg column (GE Healthcare).

#### *3.2. Determination of Molecular Mass*

The molecular masses of the human wild-type proteasome α subunits (α1–α7) and the mutated α7 proteins (α7\*, α7F14A, and α7SR) were estimated using SEC and native MS. In SEC, the samples (0.3–10 μM) were loaded onto a Superose 6 increase 10/300 GL column (GE Healthcare) with 20 mM Tris-HCl (pH 8.0) and 150 mM NaCl at a flow rate of 0.75 mL/min. For calibrating the column, ribonuclease A (13.7 kDa), chymotrypsinogen A (25 kDa), ovalbumin (43 kDa), bovine serum albumin (67 kDa), conalbumin (75 kDa), aldolase (158 kDa), ferritin (440 kDa) (GE Healthcare), and blue dextran 2000 were used. The elution profiles were recorded as absorbance values at 280 nm.

In native MS, buffer exchange of the purified α1–α7 proteins (30–50 μM monomers) was performed using 150 mM ammonium acetate (pH 6.8–8.0) with a Bio-Spin 6 column (Bio-Rad, Hercules, CA, USA). The buffer-exchanged samples (5–20 μM monomers) were immediately subjected to nanoflow electrospray ionization MS analysis with gold-coated glass capillaries made in house. Approximately 2–5 μL samples were loaded for each measurement. Buffer-exchanged α4 and α6 (4 μM monomers) were mixed with α7 (2 μM tetradecamer) at 20 ◦C for 1 h and subsequently analyzed using native MS. In contrast, α4 and α6 (4 μM monomers) were mixed with α7SR mutant (4 μM heptamer) and incubated as employed for wild-type α7. As for α2, the sample was incubated with α7SR beforehand, and the buffer exchange was carried out for the mixture, and then subjected to native MS measurements. Spectra were acquired on a SYNAPT G2-S*i* HDMS mass spectrometer (Waters, Manchester, UK) in the positive ionization mode, as previously described [25]. Spectrum calibration was performed using 1 mg/mL of cesium iodide and analysis was performed using the Mass Lynx software (Waters).

#### *3.3. AFM*

For AFM sample preparation, the α1, α4, or α6 (20 μM monomers) was mixed with an equal molar amount of α7\* at 20 ◦C for 1 h, and subsequently fractionated by SEC. The high-molecular-mass complex fractions were subjected to the AFM analysis. AFM was performed using a laboratory-constructed apparatus with cantilevers (7 μm long, 2 μm wide, and 90 nm thick) at room temperature [29]. Typical values of the spring constant, resonant frequency, and quality factor of the cantilever in an aqueous solution are approximately 0.2 N/m, 800 kHz, and 2, respectively. In AFM imaging, the free and set-point oscillation amplitudes were set to approximately 1.5 nm and 90% of the former, respectively. All samples were applied to either bare mica in 20 mM Tris-HCl (pH 8.0) with 150 mM NaCl, as previously described [26].

#### *3.4. EM*

The protein samples for EM measurement were prepared using the same protocol as in the AFM analysis. Negative-staining EM was performed using a conventional protocol, as previously described [30]. EM imaging of the α4/α7\* hetero-oligomeric complex was performed at room temperature using a JEOL JEM 2200FS electron microscope (JEOL Ltd., Tokyo, Japan) equipped with a field emission gun operating at an acceleration voltage of 200 kV. A total of 40 images were obtained using a DE20 direct detection camera (Direct Electron LP, San Diego, CA, USA) at a detector magnification of 50,000 with an energy slit width of 20 eV using the low-dose mode. The image size was set to 1.09 Å per pixel on the camera. After subjection of motion collection with the DE\_process\_frames.py script, the obtained images were processed with Relion 2.0 software [31]. Subsequently, 1,534-particle images were extracted from the 40 images and subjected to two-dimensional classification after sorting with cross-correlation coefficients.

**Supplementary Materials:** Supplementary Materials can be found at http://www.mdpi.com/1422-0067/20/9/2308/ s1.

**Author Contributions:** T.S. (Taichiro Sekiguchi), T.S. (Tadashi Satoh), E.K., and K.K. conceived and designed the study; T.S. (Taichiro Sekiguchi), T.S. (Tadashi Satoh), E.K., H.Y., and S.Y. performed protein designing and sample preparation; T.S. (Taichiro Sekiguchi), K.I., and S.U. performed native MS; T.K., H.W., and T.U. performed AFM; C.S. and K.M. performed EM; T.S. (Tadashi Satoh) and K.K. mainly drafted the manuscript.

*Int. J. Mol. Sci.* **2019**, *20*, 2308

**Funding:** This work was supported in part by the Grants-in-Aid for Scientific Research (Grant Numbers JP16H06280 to T.S., JP26460051 to E.K., JP17H05890 to H.Y., JP18H04512, JP18H01837 to T.U., and JP25102008, JP15H02491 to K.K.), by the Grants-in-Aid for Scientific Research on Innovative Areas-Platforms for Advanced Technologies and Research Resources, "Advanced Bioimaging Support" (JP16H06280), from the Ministry of Education, Culture, Sports, Science and Technology, Japan. This work was also supported by the Joint Studies Program in the Okazaki BIO-NEXT project of the Okazaki Institute for Integrative Bioscience (No. 303), by the Joint Research by Exploratory Research Center on Life and Living Systems (ExCELLS) (ExCELLS program No. 18-101 to T.U., and No. 18-402 to H.Y.), by Functional Genomics Facility, NIBB Core Research Facilities, and by SOKENDAI (The Graduate University for Advanced Studies).

**Acknowledgments:** The authors would like to thank Kumiko Hattori and Kiyomi Senda for their help in the preparation of recombinant proteins, Hiroki Kawamura for his contribution during the early stage of this study, and Michiko Nakano (Institute for Molecular Science) for constructing three-dimensional protein models useful for insightful discussion.

**Conflicts of Interest:** The authors declare no conflict of interest.
