Nature Structural and Molecular Biology: doi:10.1038/nsmb · Nature Structural and Molecular...

Nature Structural and Molecular Biology: doi:10.1038/nsmb.2938


Supplementary Figure 1

Characterization of designed leucine-rich-repeat proteins.

(a) Water-mediate hydrogen-bond network is frequently visible in the convex region of LRR crystal structures. Examples are shown for the idealized L24 (DLRR_B) and L24→L28 fusion structure (DLRR_G3). Water molecules participating in the hydrogen bond (yellow dots) network are represented by spheres. (b) Super-helical shapes of the three idealized building block repeats. For clear visualization, dots tracing the global super-helix defined by the fitted parameters are overlaid with the LRR structures (rotation angle < 720°). The highly conserved leucine residues used for the parameter fitting are represented by spheres. See Supplementary Table 1 for the helical parameter estimation. (c) Structural alignments of the partial Ncap-L245 structure in DLRR_B (top) and L225 structure in DLRR_A (bottom) into the crystal structure of DLRR_E. Cα r.m.s. deviations for the alignments of DLRR_B and DLRR_A are 0.4 Å and 0.3 Å, respectively. (d) Structural defects in the initial fusion model of DLRR_G3. The crystal structure (magenta) of the junction module in DLRR_G3 is aligned with the initial model structure before design (gray) and the final model structure after design (green). The initial model contains large cavity and side chain clashes in the junction module, which are improved in the subsequent design procedure as shown in the final model structure (green). (e) SEC-MALS experiments for DLRR_D, DLRR_E, DLRR_I, DLRR_J, DLRR_K, and DLRR_L. Most of designs are monomeric even though some soluble aggregates/oligomers are observed in DLRR_I and DLRR_K.




Experimental characterization of six L22→L28 designs (DLRR_F).

In the top row, structure alignment (left) and sequence alignment (right) of the six junction module designs are represented. The building block sequences (L22 + L28) are shown in the first row of the sequence alignment for comparison. Far-UV CD spectra, thermal denaturation at 218 nm, and SEC-MALS are shown from left to right for each design.




Experimental characterization of six L24→L28 designs (DLRR_G).

In the top row, structure alignment (left) and sequence alignment (right) of the six junction module designs are represented. The building block sequences (L24 + L28) are shown in the first row of the sequence alignment for comparison. Far-UV CD spectra, thermal denaturation at 218 nm, and SEC-MALS are shown from left to right for each design. DLRR_G6 has one less {L28→L29} module than the others. The crystal structure of DLRR_G3 is shown in Figure 3d.




Experimental characterization of four L24→L32→L24 designs (DLRR_H).

In the top row, structural alignment of the four wedge module designs is represented with the structure. Sequence alignment of the four wedge module designs is shown with the building block and the native L32 module sequence (L24 + L32 + L24) in the first row of the alignment for comparison. Far-UV CD spectra, thermal denaturation at 218 nm, and SEC-MALS are shown from left to right for each design. Design DLRR_I has two identical L32 modules derived from DLRR_H1 (Supplementary Table 2). In SEC-MALS experiments, some soluble aggregates/oligomers are observed in addition to the monomeric status. The crystal structure of DLRR_H2 is shown in Figure 3e.




Characterization of designed junction modules.

(a) Sequence alignments between the designed junction modules and the top 3 naturally occurring sequences (square block) found in BLAST1 search for the non-redundant (NR) database. There are numerous sequence differences between the designed modules and the closest sequence in NR. Indeed, BLAST fails to find full length alignments for most of the junction sequences. (b) Comparison of structures of designed and naturally occurring junctions between LRR modules. Left: designed junction modules, Middle: the closeststructural matches found in the PDB using TMalign2, Right: structural alignment. The TMalign searches were carried out with the two-unit junction module structures (green) and one or two module structures next to the junction module are shown for both designed and natural structures (yellow) to make the ideality (lack of ideality) of the different structures clearer. Most junctions between different length LRR modules in the native structures occur near the caps where the structure becomes much less regular. This irregularity, evident in the right side of the images from native structures, makes it not possible to generate novel LRR’s with controlled curvature by combining multiple different types of modules simply using junctions already existing in the PDB. (c) Structural comparison between crystal structures and model structures generated by the iterative module assembly protocol described in Method. All model structures show high consistency to the crystal structures (r.m.s. deviationg in Table 2). (d) Native LRR proteins, internalin A (InlA, PDB ID: 1O6S, top left) and ribonuclease inhibitor (RI, PDB ID: 1A4Y, bottom left), achieve high affinity and specificity by having shapes closely conforming to the surfaces of the target proteins (human E-cadherin and ribonuclease A, respectively). Each protein has a curvature optimized to its target, resulting in well-packed complementary protein-protein interfaces with hot-spot clusters (shown by red sticks) at both the N and C termini. In contract, swapping the respective target for each of the LRR proteins (i.e. RI:E-cadherin, orange-cyan complex in the top right and InlA:ribonuclease, green-yellow complex in the bottom right) makes the clashes and large gaps in the binding interface.

Supplementary Table 1 Super-helical parameters of building block modules

LRR type Rise (Å) Radius (Å) Rotation angle (radian)

Number of repeat units used for fitting

Fitted RMSD (Å)

L22 2.34 18.67 0.24 8 0.09

L24 1.41 24.62 0.20 9 0.13

{L28→L29} 0.82 16.52 0.31 10 0.17

The L22, L24 and {L28→L29} repeats form unique solenoid shapes which can be described

by three super-helical parameters (radius: distance to the helical axis, rise: projected

distance along the helical axis between adjacent units, and rotation angle: rotation angle

about the helical axis between units). The global helical shapes and parameters are

estimated by fitting the three parameters to the repeat protein structures. For the parameter

fitting, one of the highly conserved positions, the second Leu in LxxLxLxxN/C motif, is used

as a representative for each repeat module. The Cα coordinates of the representative

positions are obtained from the crystal structures of DLRR_A (L22) and DLRR_B (L24), and

from the model structure of DLRR_C ({L28→L29}). Eight to ten Cα coordinates are used to

fit the same number of coordinates arbitrary generated from the three helical parameters.

RMSD between the two coordinate sets is minimized by using non-linear optimization

algorithm (constrOptim.nl) in alabama R package3,4. Initial helical parameters, the input of

the optimization procedure, are inferred from the transformation matrix between the first two

modules of the building block structures. After performing the optimization procedure, the

parameter of the lowest RMSD is used to represent the global helical shape of the idealized

building block structures (Supplementary Fig. 1b).


Supplementary Table 2 Module organization and module origins of the multiple fusion

designs in Figure 4c.

Design name Module organization Individual

modules Original design

DLRR_I Ncap–L242–JNL24→L32→L24–JNL24→L32→L24–L242

Ncap–L242

JNL24→L32→L24

JNL24→L32→L24

L242

DLRR_B

DLRR_H1

DLRR_H1

DLRR_B

DLRR_J Ncap–L224→L242–JNL24→L28→L29→[L28→L29]2

Ncap–L224

L242

JNL24→L28

L29

[L28→L29]2

DLRR_A

DLRR_B

DLRR_G3

DLRR_G3

DLRR_G3

DLRR_K Ncap–L242–JNL24→L32→L24–L243–

JNL24→L28→L29→[L28→L29]2

Ncap–L242

JNL24→L32→L24

L243

JNL24→L28

L29

[L28→L29]2

DLRR_B

DLRR_H2

DLRR_B

DLRR_G6

DLRR_G6

DLRR_G6

DLRR_L Ncap–L223→L243–JNL24→L32→L24–L243–

JNL24→L28→L29→[L28→L29]2

Ncap–L223

L243

JNL24→L32→L24

L243

JNL24→L28

L29

[L28→L29]2

DLRR_A

DLRR_B

DLRR_H2

DLRR_B

DLRR_G6

DLRR_G6

DLRR_G6


Supplementary Table 3 Number of possible fusion LRR structures with respect to the

number of repeat units.

Number of repeat units Number of possible LRR structures

Fold change (i) → (i+1)

5 64

6 145 2.266

7 327 2.255

8 736 2.251

9 1,655 2.249

10 3,720 1.976

11 8,360 2.247

12 18,786 2.247

13 42,213 2.247

14 94,853 2.247

15 213,134 2.247

16 478,909 2.247

17 1,076,100 2.247

18 2,417,996 2.247

19 5,433,237 2.247

LRR structures are generated by recursively following the edges of the network in Figure 4a.

The general module assembly starts from Ncap-L22 or Ncap-L24 in the network except

{L28→L29}n and each assembly (transition in the network) adds one repeat unit to the

structure. The number of repeat units in the table only considers the internal repeat units

excluding N-terminal capping domain.


Supplementary Table 4 Crystallization conditions

Design names Crystallization conditions

DLRR_A 22% PEG3350 w/v, 0.1 M MES pH 6.0, 0.2 M NaCl

DLRR_E 20% PEG 1000 v/v, 0.1 M Na/K phosphate pH 6.2

DLRR_G3 2 M ammonium sulfate, 0.1 M Bis-Tris Ph 5.5

DLRR_H2 22% PEG 3350 w/v, 300 mM Ammonium sulfate, unbuffered

DLRR_I 24% PEG 3350 w/v, 0.2 M ammonium sulfate, 0.1 M HEPES pH 7.5, 0.1 M proline

DLRR_K 20% PEG-3000, 0.1 M Tris pH 7.0, 0.2 M Ca(OAc)2


Supplementary Table 5 Designed sequences

> DLRR_A

ETITVSTPIKQIFPDDAFAETIKANLKKKSVTDAVTQNELNSIDQIIANNSDIKSVQGIQYLPNLKTLKLSNNKITDISAL

KQLNNLGWLDLSNNGITDISALKNLASLHTLDLSNNGITDISALKNLDNLHTLDLSNNGITDISALKNLDNLHTLDLS

NNGITDISALKNLTSLHTLDLSNNGITDISALKNLDNLETLDLRNNGITDKSALKNLNNLKgslehhhhhh

>DLRR_B

TITVSTPIKQIFPDDAFAETIKANLKKKSVTDAVTQNELNSIDQIIANNSDIKSVQGIQYLPNVRYLALGGNKLHDISAL

KELTNLGWLNLSNNQLETLPQGVFEKLTNLTTLNLSNNQLTSLPQGVFERLASLTTLNLSNNQLTSLPQGVFERLT

NLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNQLTSLPQGVFERLTNLKTL

NLSNNQLQSLPTGVDEKLTQLTgshhhhhh

>DLRR_C

LDLSNQNKTKEDCREIARELKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARAL

KQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEG

AAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDL

SNCNLTKEACREIARALKQATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgshhhhhh

>DLRR_D



NNGITDISALKNLTSLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLP

QGVFERLTSLTTLNLSNNQLTSLPQGVFERLTNLKTLNLSNNQLQSLPTGVDEKLTQLTgshhhhhh

>DLRR_E



NLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLHTLDLSNNGITDISALKNLDNLHTLDL

SNNGITDISALKNLDNLHTLDLSNNGITDISALKNLTSLHTLDLSNNGITDISALKNLDNLETLDLRNNGITDKSALKN

LNNLKgslehhhhhh

>DLRR_F1




NNGITDISALKNLTSLHTLDLSNNGIENFSAMSNLENLKTLNLSNNRVTKEACKAIAKALKRATSLHELHLSNNNIGE

EGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETL

DLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALK

QATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgshhhhhh

>DLRR_F2



NNGITDISALKNLTSLHTLDLSNNGIENFNALRNLENLKTLNLSNNRVTKDACEAIAEALKRATSLHELHLSNNNIGE

EGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETL

DLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALK

QATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgshhhhhh

>DLRR_F3



NNGITDISALKNLTSLHTLDLSNNGIENFEAMRNLENLKTLNLSNNRLTKEACKAVAEALKRATSLHELHLSNNNIG

EEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLET

LDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARAL

KQATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgshhhhhh

>DLRR_F4



NNGITDISALKNLTSLHTLDLSNNGITNVSALKNLENLKTLNLSNNNITKEACKAIAEALKRATSLHELHLSNNNIGEE

GAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLD

LSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQ

ATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgshhhhhh

>DLRR_F5




NNGITDISALKNLTSLHTLDLSNNGIRNLEAMRNLENLKTLNLSNNNVTKEACSALAEALKRATSLHELHLSNNNIG




>DLRR_F6



NNGITDISALKNLTSLHTLDLSNNGIRNFEAMRNLENLKTLNLSNNNFTKEACSALAEALKRATSLHELHLSNNNIG




>DLRR_G1



NLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNQLTSLPDGVLERLTNLKTL

NLSNNQITKEVCRHVAKILKQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALK

QATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEGA

AELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgshh

hhhh

>DLRR_G2



NLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNQLTSLPDGVFERLTNLKTL

NLSNNQLTKEACRIVAKMLKQLASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALK



hhhh

>DLRR_G3





NLSNNQLTKEACRAVANALKQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARAL

KQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEG

AAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgsh

hhhhh

>DLRR_G4



NLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNQLTSLPDGVLERLTNLKTL

NLSNNQITKEVCRLVAKFLKQLASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALK



hhhh

>DLRR_G5




NLSNNQITKEVCRMVAKVLKQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARAL

KQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLHELHLSNNNIGEEG

AAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATSLHELHLSNNNIGEEGKAWLEEARRHPGSTLETgsh

hhhhh

>DLRR_G6



NLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNQLTSLPKGVLERLTNLKTL

NLSNNQITKEVCRHVAELLKQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALK

QATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATSLHELHLSNNNIGEEGK

AWLEEARRHPGSTLETgshhhhhh


>DLRR_H1


KELTNLGWLNLSNNQLETLPQGVFEKLTNLTTLNLSNNQLTSLPQGVFERLASLTTLNLSNNNIANINDQMLEGLT

NLTTLNLSHNNLARLWKHANPGGPIYFLKGLTNLTTLNLSSNGFDEIPREVFKDLTSLTTLNLSNNQLTSLPQGVFE

RLTNLKTLNLSNNQLQSLPTGVDEKLTQLTgshhhhhh

>DLRR_H2


KELTNLGWLNLSNNQLETLPQGVFEKLTNLTTLNLSNNQLTSLPQGVFERLASLTTLNLSNNNLANLNDKVFEGLT

NLTTLNLSNNNLARLWKHANPGGPIYFLKGLTNLTTLNLSNNGFDEFPKEVFKDLTSLTTLNLSNNQLTSLPQGVF

ERLTNLKTLNLSNNQLQSLPTGVDEKLTQLTgshhhhhh

>DLRR_H3


KELTNLGWLNLSNNQLETLPQGVFEKLTNLTTLNLSNNQLTSLPQGVFERLASLTTLNLSNNNLANLNDRLLEGLT

NLTTLNLSNNNLARLWKHANPGGPIYFLKGLTNLTTLNLSNNGFDEFPREVFKDLTSLTTLNLSNNQLTSLPQGVF


>DLRR_H4


KELTNLGWLNLSNNQLETLPQGVFEKLTNLTTLNLSNNQLTSLPQGVFERLASLTTLNLSNNNLANLNDRVFEGLT

NLTTLNLSNNNLARLWKHANPGGPIYFLKGLTNLTTLNLSNNGFDELPKEVFKDLTSLTTLNLSNNQLTSLPQGVF


>DLRR_I


KELTNLGWLNLSNNQLETLPQGVFEKLTNLTTLNLSNNQLTSLPQGVFERLASLTTLNLSNNNIANINDQMLEGLT

NLTTLNLSHNNLARLWKHANPGGPIYFLKGLTNLTTLNLSSNGFDEIPREVFKDLTSLTTLNLSNNNIANINDQMLE

GLTNLTTLNLSHNNLARLWKHANPGGPIYFLKGLTNLTTLNLSSNGFDEIPREVFKDLTSLTTLNLSNNQLTSLPQG

VFERLTNLKTLNLSNNQLQSLPTGVDEKLTQLTgshhhhhh

>DLRR_J




NNGITDISALKNLTSLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLP

QGVFERLTNLKTLNLSNNQLTKEACRAVANALKQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCN

LTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATTLH

ELHLSNNNIGEEGKAWLEEARRHPGSTLETgshhhhhh

>DLRR_K


KELTNLGWLNLSNNQLETLPQGVFEKLTNLTTLNLSNNQLTSLPQGVFERLASLTTLNLSNNNLANLNDRVFEGLT

NLTTLNLSNNNLARLWKHANPGGPIYFLKGLTNLTTLNLSNNGFDELPKEVFKDLTSLTTLNLSNNQLTSLPQGVF

ERLTNLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNQLTSLPKGVLERLTN

LKTLNLSNNQITKEVCRHVAELLKQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIA

RALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATSLHELHLSNNNIG

EEGKAWLEEARRHPGSTLETgshhhhhh

>DLRR_L


KQLNNLGWLDLSNNGITDISALKNLASLHTLDLSNNGITDISALKNLDNLHTLDLSNNGITDISALKNLTSLTTLNLSN

NQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNNLA

NLNDRVFEGLTNLTTLNLSNNNLARLWKHANPGGPIYFLKGLTNLTTLNLSNNGFDELPKEVFKDLTSLTTLNLSN

NQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTNLTTLNLSNNQLTSLPQGVFERLTSLTTLNLSNNQLT

SLPKGVLERLTNLKTLNLSNNQITKEVCRHVAELLKQAASLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSN

CNLTKEACREIARALKQATTLHELHLSNNNIGEEGAAELVEALLHPGSTLETLDLSNCNLTKEACREIARALKQATS

LHELHLSNNNIGEEGKAWLEEARRHPGSTLETgshhhhhh

*C-terminal linkers and 6x His tags are shown in lower case.

*AS or TS in the regular repeat sequences are for inserting the restriction sites (NheI and SpeI).


Supplementary References

1. Altschul, S.F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search

programs. Nucleic Acids Res 25, 3389-402 (1997).

2. Y. Zhang, J. Skolnick, TM-align: A protein structure alignment algorithm based on TM-score,

Nucleic Acids Research 33, 2302-09 (2005)

3. R Core Team R: A language and environment for statistical computing R Foundation for Statistical

Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/ (2012)

4. Ravi Varadhan, alabama: Constrained nonlinear optimization. R package version 2011.9-1.

http://CRAN.R-project.org/package=alabama (2012)


Date post:	12-Mar-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

Nature Structural and Molecular Biology: doi:10.1038/nsmb · Nature Structural and Molecular...

Documents