Molecular Cell, Volume 46
Molecular Model of the Human 26S Proteasome
Paula C.A. da Fonseca, Jun He, and Edward P. Morris
Inventory of Supplemental Information
Figure S1 - Biochemical characterisation of the human 26S proteasome sample.
Related to Figure 1.
Figure S2 – Evaluation of cryo-electron microscope images of the human 26S
proteasome sample. Related to Figure 1.
Figure S3 - Estimation of resolution of the human 26S proteasome reconstruction.
Related to Figures 1 & 2.
Figure S4 - Rearrangement of 20S core subunits in the 26S proteasome. Related to
Figure S5 - The Rpt1-6 subcomplex. Related to Figure 3.
Figure S6 - Rpn1 and Rpn2. Related to Figure 4
Figure S7 – Structural coordinates for the PCI and MPN subunits of the 19S-RP.
Related to Figures 6 & 7.
Movie S1. Three-dimensional map of the human 26S proteasome in surface
representation rotating about its long axis. Related to Figure 1
Movie S2. Molecular model of the 26S proteasome. Related to Figure 7.
Supplemental Experimental Procedures
Figure S1 - Biochemical characterisation of the human 26S proteasome sample. (A)
SDS-PAGE gel showing the pattern of 26S proteasome subunits. (B) Native gel
electrophoresis. Left and centre panels show fluorescent images of gel stained for peptidase
activity with suc-LLVY-AMC, where the fluorescent signal is shown in dark against a light
background. Left panel shows activity for RP2-20S and RP-20S species (double capped and
single capped 26S proteasomes). In the centre panel activity is assessed in the presence of
0.02% SDS (w/v) and additionally shows activity for the small amount of uncapped 20S core
present in the 26S proteasome preparation. The fact that treatment with SDS only raises the
peptidase activity of uncapped 20S core present in the sample, but not that of the 26S
proteasome, indicates that the peptidase activity of the untreated 26S proteasome is intact.
Right panel is coomassie stained. The similar band intensities in the middle and right panels
indicate that the SDS treatment caused full 20S activation, with the faint bands for uncapped
20S core consistent with its low amount present in the analysed sample. (C) SDS-PAGE gel
immunoblotted to test for Rpn13. Lane 1: Cell lysate containing human Rpn13 expressed
with N-terminal double Step II tag. Lane 2: Hela cell extract (Abcam plc). Lane 3: Human
erythrocyte 26S proteasome preparation. No Rpn13 is detected in the Human erythrocyte
26S proteasome preparation. Lanes 1 and 2 serve as positive controls.
Figure S2 - Evaluation of cryo-electron microscope images of the human 26S
proteasome sample. (A) CCD recorded field of raw images in which side views of double-
capped 26S proteasome complexes are marked with red circles. (B,C) Rotational power
spectrum from A. (B) Greyscale representation where minima in fitted contrast transfer
function are marked as red arcs. (C) Graphical representation where minima in fitted
contrast transfer function are marked as vertical lines. Oscillations due to the contrast
transfer function extend beyond 7 Å.
Figure S3 – legend on next page
Figure S3 - Estimation of resolution of the human 26S proteasome reconstruction. (A)
Fourier shell correlation curve; ½ bit threshold curve (labelled) shown as solid line and 0.5
correlation coefficient threshold shown as dotted line giving rise to a resolution estimate of 7-
9 Å. (B-E) Comparison between the 26S proteasome structure (B,C – grey mesh with fitted
coordinates) and a model map calculated from the fitted 20S core low-pass filtered to 7 Å
(D,E – blue mesh with fitted coordinates). B & D are side view central sections. C & E are
top views sectioned at the interface between the α and β subunit rings.
Figure S4 – legend on next page
Figure S4 - Rearrangement of 20S core subunits in the 26S proteasome. Colour
coded cartoon representations of the α and β subunits independently docked into the 26S
proteasome map viewed down its long axis. (A) α subunit ring. (B) β subunit ring. For
comparison the α and β subunits as determined crystallographically in the mammalian 20S
core (Unno, Mizushima et al. 2002) are shown in grey. (C-E) Semitransparent side view
surfaces of the α and β rings with cartoons of fitted coordinates illustrating the effects of
subunit displacement. Narrow side windows at the interface between the α and β rings are
indicated by red arrows indicated in (C) and (D). (C) Model map calculated with the
independently fitted subunits. (D) Cryo-EM map. (E) Model map calculated from
mammalian 20S core crystal structure. (C) and (D) are low-pass filtered to 7 Å resolution.
Figure S5 – legend on next page
Figure S5 - The Rpt1-6 subcomplex. (A). Schematic diagram comparing the sequences of
Rpt1-6 and their archaeal homologue PAN. (B-F) Sections of the 26S proteasome map,
parallel to its long axis, represented as grey mesh with docked coordinates of the 20S core
Figure S6 - Rpn1 and Rpn2. (A) Sequence alignment of human and S. cerevisisae Rpn2.
(B) Identification of Rpn1 by antibody labelling. IGG molecules (red arrows) are observed
binding to the sides of the 19S-RP complexes. (C) Antibody labelling of Rpn2. IGG
molecules (red arrows) are observed binding to the ends of the 19S-RP complexes.
Figure S7 – legend on next page
Figure S7 – Structural coordinates for the PCI and MPN subunits of the 19S-RP.
Structural coordinates for the PCI (A) and MPN (B) subunits. Homology models for
Rpn3,5,7,9,12 were calculated using Phyre (Kelley and Sternberg 2009) and I-Tasser (Roy,
Kucukural et al. 2010). The x-ray structure coordinates of the Rpn8 MPN domain were used
as model for the MPN domain of Rpn11. The coordinates were fitted into the 19S-RP
densities of the 26S proteasome map with an overall correlation coefficient of 0.66,
indicating a good agreement between the densities and fitted coordinates, with the highest
correlation coefficient of 0.78 obtained for the fitting of the x-ray structure of Rpn6.
SUPPLEMENTAL EXPERIMENTAL PROCEDURES
Biochemical characterisation of the 26S proteasome
26S proteasome samples purified from human erythrocytes were obtained from
Biomol International. The samples were always kept in the presence of 2 mM ATP and 5 mM
MgCl2. The samples were analysed by SDS-PAGE in order to confirm their purity and their
typical subunit profile. Western blots showed that these preparations contain no Rpn13 (anti-
Rpn13 from human obtained from Abcam) in agreement with the previous analysis of human
erythrocyte 26S proteasome samples (Chen et al., 2010). Native gel electrophoresis was
performed in the presence of 2 mM ATP and 5 mM MgCl2 in all solutions. The peptidase
activity assay performed by incubating native gels in the presence 50 M Suc-LLVY-AMC
(Enzo Life Sciences), as previously described (Elsasser et al., 2005)
The 26S proteasome was diluted in 50mM Tris-HCl, pH 7.5, 5mM MgCl2, 2mM ATP
and 1mM dithiotreitol to a final concentration of ~0.15 mg/ml. This was applied to thin layers
of carbon freshly floated from mica and supported by 2 μm aperture Quantifoil grids and
flash frozen using a Vitrobot (FEI). Grids were mounted in a Gatan 626 cryoholder,
transferred into an FEI Tecnai TF20 electron microscope and data were collected at -178oC.
Low dose images (~25 e- Å-1) were recorded at 200kV accelerating voltage, 63,000x nominal
magnification and a defocus of 1-2.5 μm using a Tietz F415 CCD camera. The calibrated
sampling of the resulting images, binned to 2k x 2k pixels, was 2.82Å at the specimen level.
CCD images were screened and used for analysis if Thon rings in the power spectrum
extended isotropically beyond 7Å (Figure S2).
Single particle analysis and model building
Images were processed using Imagic (van Heel et al., 2000), Spider (Frank et al.,
1996) and locally developed programs. The CTF was measured and phase modulation
was corrected on whole CCD images by phase flipping in zones of contrast reversal.
Molecular images of double-capped 26S proteasomes were picked by hand using the
Boxer program from EMAN (Ludtke et al., 1999). Under the conditions used the majority of
molecular images were side views and could be clearly identified as double-capped
proteasomes with a 19S-RP at each end of the 20S core (Figure S2A, red rings) and these
were selected for analysis. A smaller proportion of side views of single-capped complexes
were observed, but these were excluded from the current analysis. Side views of uncapped
20S cores were seldom observed in the regions of the grid containing good 26S
proteasome images. Initial reference free class averages were calculated using the EMAN
refine2d procedure: an initial structure was calculated from these classes using Euler
angles assigned with reference to our earlier 26S proteasome structure (da Fonseca and
Morris, 2008). The structure was refined by multiple rounds of multireference alignment
using a brute-force cross-correlation approach and angular assignment by projection
matching using a combination of Imagic and Spider software. Three-dimensional
reconstructions with C2 symmetry imposed were calculated with a locally developed
program in which projection images are merged in Fourier space. The merging
incorporates weighting factors for the geometrical distribution of input data, contrast
transfer function and defocus envelope and applies a correction for the amplitude
modulation of the contrast transfer function to the merged Fourier components. To
minimise interpolation errors uninterpolated images are used as input and translations
required for alignment are achieved by phase shifting in Fourier space. The final map was
sharpened using a negative b-factor (Chen et al., 2009). Atomic coordinate data were fitted
to the 3D maps derived from cryo-electron microscope data with URO (Navaza et al.,
2002). Pymol (www.pymol.org) was used to visualise and create structure diagrams from
3D maps and atomic coordinate data.
Antibody labelling of Rpn1 and Rpn2
Antibodies against human Rpn1 and Rpn2 were obtained from Abcam. 26S proteasome
samples were incubated in the presence of antibody and loaded on Quantifoil grids covered
with a thin layer of carbon. The grids were stained using 2% (w/v) uranyl acetate and images
recorded as described above, expect that the imaging was performed at room temperature
and an electron dose
of ~100 e-Å-1 was used.
Cloning and expression of human full-length Rpn13
The Human Multiple Tissue cDNA Panel I (Clontech) contains the full-length cDNA of
human Rpn13. The full-length hRpn13 was amplified by PCR using primers
5’-GAATGCCCCCCGTTTTTAGTCCAGGCTCATGTCCTCCT-3’, generating a 15bp
overhang at both ends of the PCR product that precisely matches the ends of the linearized
vector. The PCR product was subcloned into PmeI linearized vector pOPNDS (Ziguo Zhang
and David Barford, unpublished) by the In-Fusion cloning system (Clontech). The pOPNDS
vector was derived from the pOPIN vector by inserting the DNA encoding a double Strep II
tag followed by a TEV cleavage site. The complete construct was transformed into E. coli
strain Rosetta 2 (Novagen). Cells were grown at 37°C to an OD600nm of ~0.5 and were
induced by addition of 0.5mM IPTG.
Chen, J. Z., Settembre, E. C., Aoki, S. T., Zhang, X., Bellamy, A. R., Dormitzer, P. R.,
Harrison, S. C., and Grigorieff, N. (2009). Molecular interactions in rotavirus assembly and
uncoating seen by high-resolution cryo-EM. Proc Natl Acad Sci U S A 106, 10644-10648.
Chen, X., Lee, B. H., Finley, D., and Walters, K. J. (2010). Structure of proteasome ubiquitin
receptor hRpn13 and its activation by the scaffolding protein hRpn2. Mol Cell 38, 404-415.
da Fonseca, P. C., and Morris, E. P. (2008). Structure of the human 26S proteasome:
subunit radial displacements open the gate into the proteolytic core. J Biol Chem 283,
Elsasser, S., Schmidt, M., and Finley, D. (2005). Characterization of the proteasome using
native gel electrophoresis. Methods Enzymol 398, 353-363.
Frank, J., Radermacher, M., Penczek, P., Zhu, J., Li, Y., Ladjadj, M., and Leith, A. (1996).
SPIDER and WEB: processing and visualization of images in 3D electron microscopy and
related fields. J Struct Biol 116, 190-199.
Ludtke, S. J., Baldwin, P. R., and Chiu, W. (1999). EMAN: semiautomated software for high-
resolution single-particle reconstructions. J Struct Biol 128, 82-97.
Navaza, J., Lepault, J., Rey, F. A., Alvarez-Rua, C., and Borge, J. (2002). On the fitting of
model electron densities into EM reconstructions: a reciprocal-space formulation. Acta
Crystallogr D Biol Crystallogr 58, 1820-1825.
van Heel, M., Gowen, B., Matadeen, R., Orlova, E. V., Finn, R., Pape, T., Cohen, D., Stark,
H., Schmidt, R., Schatz, M., and Patwardhan, A. (2000). Single-particle electron cryo-
microscopy: towards atomic resolution. Q Rev Biophys 33, 307-369.