Supplementary Material Online (SOM)

Leo C. James, Pietro Roversi & Dan S. Tawfik

Antibody multi-specificity mediated by conformational diversity

The supplementary material includes Materials and Methods and additional dataregarding:

(1) Cloning sequencing and expression of SPE7 Fv.(2) Pre-steady-state kinetics(3) Crystal structure determination(4) Selection of the Trx-protein ligands

(1) Cloning sequencing and expression of SPE7 Fv.

SPE7 Fv was cloned from cDNA of the SPE7.49 hybridoma (1) into a pUC vector to

append pelB leaders on both chains and a 6xHis tag on the heavy chain (2). Its sequence

was found to be most similar to IgG anti-NP antibody B1-8 (3) — a germline antibody

with heavy chain V gene VH186.2 and light chain gene VL lambda 1 (Fig. S1 below).

The Fv fragment was expressed in the periplasm of E. coli and purified by Ni-NTA

chromatography essentially as described (2). Samples for crystallization were further

purified by gel filtration (S-200) where only a monomeric species was observed, and

dialyzed into 10 mM HEPES pH 7. Affinity measurements at equilibrium were

performed with the intact SPE7 antibody (Sigma) or its recombinant Fv fragment, by

either fluorescence quenching titration (for Kd <1 µM) or competitive ELISA for the

lower affinity ligands (4).

Figure S1. Heavy and light chain amino acid sequence of SPE7 Fv aligned with B1-8

(3). Residues with absolute identity are highlighted and in bold. CDRs are indicated as

boxed sections. B1-8 is a primary response antibody constructed from germline V(D)J

genes, including the Vh gene V186.2 and the Vl gene Vlambdal. SPE7 uses identical

germline genes and differs only at 1 position within CDR H2, 4 positions within CDR H3

and one position in CDR L3.

L1

L2

L3

H3

H2

H1

DFL16.1 JH2

j1d1

(2) Pre-steady-state kinetics

Pre-steady-state kinetics were performed and analyzed as described (4) using SPE7 Fv at

1 µM concentration in PBS (50 mM sodium phosphate, 100 mM sodium chloride, pH 7)

at 20°C. Data from at least 3 independent reactions were averaged for each ligand

concentration. Individual phases were generally traced for ����������� ����������������

within the dead time of the instrument (2 ms) was discarded before fitting with the

program Kaleidagraph. Fits to a single exponential were performed with equation (I): Ft=

F0. exp(-(1/τ). t) + F∞; where Ft is the fluorescence (in arbitrary units) at time t ; F0 is the

fluorescence of the free (uncomplexed) antibody; 1/τ (or kobserved) is the reciprocal

relaxation time; F∞ is the fluorescence at equilibrium (at t=∞).

(3) Crystal structure determination

Crystals were grown by the hanging drop method under the conditions described in Table

1. Hapten co-crystallization (Ab3) was performed by co-concentrating antibody with

hapten (due to low hapten solubility). The Trx-Shear3:Ab complex (Ab4) was co-

crystallized by adding equimolar amounts of the independently purified and expressed

proteins. For data collection, crystals were frozen at 100K in a cryo-stream without

addition of cryo-protectant. All datasets were collected on a rotating anode with a Mar-

345 image plate detector. A single crystal was used in each dataset. Data were indexed

using MOSFLM and scaled in SCALA (5). All structures were determined by AMORE

molecular replacement using the antibody B1-8 (1A6V) (5). Structures Ab1 and Ab3

were refined using REFMAC and ARP/warp (5) with rebuilding in O (6). Ab4 was

refined in BUSTER (7) and Ab2 in SHELXL97 (8) and CNS (9).

Ab1 isomer. The structure of free isomer Ab1 crystallizes in spacegroup P212121 with

one of each chain, H and L, per asymmetric unit (a.u.). The main packing interactions,

involving CDR loops, occur between the CDR 2 of H and L and are reproduced in all the

other structures. There is a single crystal packing interaction involving H3 (which is also

reproduced in the Ab3 structures) and none involving L3.

Ab2 isomer. The structure of free isomer Ab2 crystallizes in an apparent spacegroup of

I422 but the cell is not large enough to accommodate 2 copies per a.u. Twinning was

detected following examination of the acentric cumulative intensity distribution, which

was sigmoidal (Figure S2). Twinning was not unambiguously indicated from the

<I2>/<I>2 which has an average of 1.9 for all reflections (with a minimum of 1.59 for the

403 reflections in the resolution range 5.38-4.53 A) (Figure S3). The actual spacegroup is

I4 with one copy per a.u and twinning providing the apparent additional 2-fold diagonal

symmetry. Further indication that this 2-fold is due to twinning came from the molecular

replacement: no solution was found for I422 but replacement in I4 gives a strong

solution. Iterative rounds of refinement in SHELXL resulted in a model with an Rfactor

of 0.207 and a refined twin fraction of 0.498. The different Ab2 conformation is further

reinforced by the fact that when the CDR H3 conformations of either the Ab1 or the Ab3

model are brought into the I4 cell they clash heavily with symmetry related copies. Ab2

has the same packing interactions as Ab1 between H and L chain CDR2’s.

Ab3 isomer. All of the hapten-complexed structures crystallize in P212121 with 4

antibody fragments in the asymmetric unit; the molecules are labeled LH, NJ, MI and OK

(light chains labeled L,N,M and O; and heavy chains labeled H,J,I,K). The pairs of

molecules LH/NJ and MI/OK are related by a NCS-twofold axis along the <110>

direction that gives a pseudo-tetragonal packing. CDR 3 of heavy chains K and I contact

each other across this NCS twofold axis. The major CDR packing contact is the one

between CDR 2 of chains H and L (and the NCS-related one between CDR 2 of chains J

and N). The cell dimensions (a = 78.917Å; b = 78.88Å) and the symmetry of the

diffraction patterns are indeed compatible with a space group of P41212 but attempts to

build and refine in P41212 with 2 molecules per a.u. gave worse density maps and

refinement statistics. No indication of twinning was present in the intensity distribution

nor in plots of <|E**2-1|> in resolution shells. The LH and NJ molecules were found by

molecular replacement. The MI and OK molecules were built in the residual density, with

the heavy chains I and K nearly fully traced in the Alizarin Red complex, and nearly

completely disordered in the FUR, and DNPSer complexes. Refinement of the hapten

structures used two sets of hard NCS constraints: one set relating the non-CDR residues

of the HL/NJ pair and the other between non-CDR residues of the MI/OK pairs; CDR

regions were unconstrained.

Ab4 isomer. The SPE7:TRX-Shear3 crystal was refined with two sets of hard NCS

constraints: one relating the molecules in the HL/NJ pair and one between the two copies

of the TRX-Shear3 molecule (chains A and B). Packing resembles the Ab3 structures: the

HL and NJ molecules pack equivalently to the same molecules in the hapten structures;

the TRX-Shear3 A and B molecules are, approximately, in regions that in the hapten

structures were taken up by molecules MI and OK. The contacts between CDR 2 of

chains H and L (and the NCS-related one between CDR 2 of chains J and N) are

preserved in this crystal.

Effect of packing interactions on observed CDR H3 and L3 conformations. All

structures involve only a single crystal packing interaction involving H3 and none

involving L3. Thus, packing energies cannot explain the different conformations.

Furthermore, a similar packing arrangement is seen in all the orthorhombic structures

(free isomer Ab1 and complexed structures). If packing interactions were determining the

CDR conformations, these structures would be the same – but in fact, they are quite

different. The Ab2 crystals have a different spacegroup and unit cell and the lattice

cannot accommodate the Ab1 conformation. Therefore, to form a crystal, a different

solution conformation to Ab1 must have been used. Finally, as there is only a single

packing contact involving H3 and none involving L3 residues in Ab2 the energetic

contribution of packing to the conformation of this region must be small.

Model and data quality. All structures were validated using the program PROCHECK

(5). In all structures, residues for SPE7 are located in either energetically favored or

allowed regions on a Ramachandran plot with the exception of Thr53L (see Figure S4, an

example plot for Ab1). This residue forms part of a classic gamma-turn, which explains

the unfavorable Phi-Psi value. In the Fv-thioredoxin structure there are no thioredoxin

residues in disallowed regions. Data quality for all structures was good, as expected at

high resolution, with the exception of the disordered regions in chains I and K in Ab3

models. Views of the electron density for conformations Ab1-Ab4, describing regions

around the L3 and H3 loops, is presented in Figure S5 Part’s A and B.

Temperature factors. A comprehensive break-down of the temperature factors for all

structures is given in Table S1 below. The values are all within the expected range for the

resolution of each structure. However, there is variance in the conformational rigidity

between chains within the Ab3 models (Aliz, DNPS, Fur): the atomic vibration in chains

L,N and H,J is generally lower than in chains M,O and I,K. In DNPS and Fur this

conformational flexibility meant that much of the IK chains, in particular the CDR 1 and

2 loops, could not be built. This may imply what the other models confirm – namely, that

SPE7 does not adopt a single structure. The Ab3 data describe a low energy conformation

(as modeled in chains LN and HJ) and an energy landscape of different conformations in

equilibrium (as indicated by the atomic vibration in MO and IK). The temperature factors

provide a further indication of underlying conformational diversity in the form of

differences between the average B-factor of the H vs. L chain, in some structures. This is

in agreement with our observation that it is CDR H3 that undergoes the most radical

conformational change.

Sample homogeneity. All structures were crystallized from protein prepared from a

single clone of the recombinant Fv fragment of SPE7 expressed in E. coli. Thus, the

differences in conformation cannot be ascribed to heterogeneity of sequence.

Nevertheless, Figure 6S shows OMIT maps of the pivotal Trp93L from the two native

structures Ab1 and Ab2 demonstrating that the side chain is correct and the density is

unequivocal.

Figure S2. Plot of the cumulative intensity distribution for the twinned structure of the

free antibody isomer Ab2.

Figure S3. Plot of <|E**2-1|> for the twinned structure of the free antibody isomer Ab2.

Figure S4. A Ramachandran plot for the structure of the free antibody isomer Ab1.

A B

C D

Figure S5 – Part A. 2Fo-Fc electron density maps of the four different SPE7 binding siteconformations, contoured at the 1σ level. (A) Free Ab1, (B) Free Ab2, (C) Trx-Shear3complex Ab4 (Trx-Shear3 density is not shown) (D) DNP-Ser complex Ab3. Graphicsprepared in AESOP with maps output from REFMAC. Non-loop residues are notindicated for clarity.

Figure S5 – Part B. Zoom-in of 2Fo-Fc electron density maps for the four differentSPE7 conformations, contoured at 1σ level. (A) Free antibody isomer Ab1; (B) Freeantibody isomer Ab2; (C) Ab4, the antibody isomer complexed with Trx-Shear3 (densityfor the Trx-Shear3 antigen not shown); (D) The antibody-DNP complex, Ab3. A close-upview of the binding site is shown and key residues in each conformation are annotated.Whilst a similar orientation matrix has been used in each figure for clarity, rotation and z-slice adjustment have been used to maximize clarity. Maps were output from REFMACand are displayed in AESOP.

DNP

L93

H101

H105

H100

D

H101

H105

L93

A

L93H101

H105

H100

C

L93

H101H100

H107

B

Ab1

Figure S6. OMIT maps of Trp93L from Ab1 and Ab2 structures. Electron density (Fo-Fc) maps were output from REFMAC following refinement with a complete modelminus the Trp93L residue.

Table S1: Temperature factors for all structures by subset*

=================================================================================Structure Subset Natoms <B> rmsB MinB MaxB=================================================================================Ab1 L1 99 18.7810 2.9646 14.2900 31.9900Ab1 L2 43 23.9509 5.5178 16.9900 37.8500Ab1 L3 95 19.1346 4.5374 14.0900 35.3400Ab1 noCDR.L 571 21.8300 6.2545 13.9200 46.7100Ab1 L.All 808 21.2524 5.8874 13.9200 46.7100 ==================================================================Ab1 H1 91 18.5869 2.3581 14.3100 24.8200Ab1 H2 123 20.6033 5.6029 14.2600 39.3100Ab1 H3 69 24.9900 6.0582 16.7600 43.7500Ab1 noCDR.H 671 20.8875 5.7893 13.6800 44.4500Ab1 H.All 954 20.9281 5.7041 13.6800 44.4500 ===================================================================================================================================================Structure Subset Natoms <B> rmsB MinB MaxB=================================================================================Ab2 L1 99 20.4798 6.7098 10.4200 42.8300Ab2 L2 43 20.3553 6.3101 8.7000 42.6900Ab2 L3 95 19.3123 3.8185 12.8500 32.4200Ab2 noCDR.L 576 26.5814 10.4000 10.6700 62.2000Ab2 L.All 813 24.6597 9.7462 8.7000 62.2000 ==================================================================Ab2 H1 91 47.7282 5.8442 32.8000 59.3500Ab2 H2 99 48.8031 8.6994 19.8200 69.5600Ab2 H3 69 53.8038 6.5576 41.5000 66.1000Ab2 noCDR.H 661 40.0287 11.7324 12.1400 85.0500Ab2 H.All 920 42.7676 11.5984 12.1400 85.0500 ===================================================================================================================================================Structure Subset Natoms <B> rmsB MinB MaxB=================================================================================Ab3-Aliz L1 99 33.8768 5.3759 23.0700 49.5400Ab3-Aliz L2 43 35.1544 6.1169 28.6900 52.7800Ab3-Aliz L3 95 30.6056 4.9410 22.3200 52.7700Ab3-Aliz noCDR.L 585 33.2032 13.3867 20.5300 106.6900Ab3-Aliz L.All 822 33.0862 11.6968 20.5300 106.6900 ==================================================================Ab3-Aliz H1 91 33.8904 5.7901 19.4100 49.6100Ab3-Aliz H2 122 34.3515 7.9489 21.2800 66.3600Ab3-Aliz H3 69 37.6419 7.0984 27.0900 62.0600Ab3-Aliz noCDR.H 727 37.4653 15.1496 21.2500 100.0600Ab3-Aliz H.All 1009 36.7785 13.4659 19.4100 100.0600 ==================================================================Ab3-Aliz N1 99 33.9076 5.5210 23.0200 49.6600Ab3-Aliz N2 43 34.8714 6.1628 28.3200 52.1600Ab3-Aliz N3 95 30.4912 5.2588 23.1500 54.7600Ab3-Aliz noCDR.N 585 33.1541 13.4496 20.4200 100.8300Ab3-Aliz N.All 822 33.0269 11.7736 20.4200 100.8300 ==================================================================Ab3-Aliz J1 91 33.4878 5.5594 21.9700 49.0400Ab3-Aliz J2 122 34.5180 8.0893 21.5300 65.7200Ab3-Aliz J3 69 37.8117 6.6245 27.8700 63.3900Ab3-Aliz noCDR.J 727 37.5524 15.1612 21.7500 100.0100Ab3-Aliz J.All 1009 36.8367 13.4690 21.5300 100.0100 ==================================================================Ab3-Aliz M1 76 62.8564 8.6957 49.0000 96.0000Ab3-Aliz M2 43 62.3916 8.4667 53.8500 87.6300Ab3-Aliz M3 81 69.2383 5.5258 54.7200 77.8100Ab3-Aliz noCDR.M 558 51.7352 10.2852 34.8800 93.1400Ab3-Aliz M.All 758 55.3252 11.4663 34.8800 96.0000 ==================================================================Ab3-Aliz I1 91 68.1474 10.3006 47.2700 85.0800Ab3-Aliz I2 99 96.1499 6.6128 85.4800 100.8000Ab3-Aliz I3 69 77.4291 3.4375 68.5600 85.4700Ab3-Aliz noCDR.I 479 86.6790 10.6258 56.1200 100.8700Ab3-Aliz I.All 738 84.7996 12.3362 47.2700 100.8700 ==================================================================Ab3-Aliz O1 95 64.3385 7.9783 49.6300 95.1000Ab3-Aliz O2 43 61.5714 7.8019 52.7100 85.5000Ab3-Aliz O3 81 69.6437 5.5952 53.6800 78.3800Ab3-Aliz noCDR.O 557 51.9828 10.5101 32.2900 94.4500Ab3-Aliz O.All 776 55.8702 11.6168 32.2900 95.1000 ==================================================================Ab3-Aliz K1 57 78.3747 4.3733 66.9800 86.6800Ab3-Aliz K2 99 96.9452 6.3037 86.4300 100.4900

Ab3-Aliz K3 69 77.5803 3.5090 70.3200 85.4100Ab3-Aliz noCDR.K 467 87.9299 9.6453 65.9800 100.9700Ab3-Aliz K.All 692 87.4007 10.0409 65.9800 100.4900 ===================================================================================================================================================Structure Subset Natoms <B> rmsB MinB MaxB=================================================================================Ab3-DNPS L1 99 38.7832 4.6799 28.4700 49.9000Ab3-DNPS L2 43 42.0595 6.5631 34.3800 58.9900Ab3-DNPS L3 95 35.4488 4.1144 29.8000 48.6800Ab3-DNPS noCDR.L 566 36.0951 7.8720 26.5900 96.9700Ab3-DNPS L.All 803 36.6694 7.2941 26.5900 96.9700 ==================================================================Ab3-DNPS H1 91 40.4716 5.4555 31.3900 53.6700Ab3-DNPS H2 120 38.1588 7.1524 29.3600 65.8900Ab3-DNPS H3 69 44.9830 7.8310 32.6800 71.0800Ab3-DNPS noCDR.H 685 38.9468 10.0898 26.9000 92.5000Ab3-DNPS H.All 965 39.4242 9.4070 26.9000 92.5000 ==================================================================Ab3-DNPS N1 99 38.7589 4.7033 28.2600 50.0600Ab3-DNPS N2 43 42.0307 6.5449 34.1400 59.0100Ab3-DNPS N3 95 35.3872 4.1510 29.4600 48.3500Ab3-DNPS noCDR.N 566 36.0882 7.8937 26.7100 96.9800Ab3-DNPS N.All 803 36.6528 7.3136 26.7100 96.9800 ==================================================================Ab3-DNPS J1 91 40.4318 5.4337 31.3200 53.6900Ab3-DNPS J2 120 38.1034 7.1633 29.3400 65.8900Ab3-DNPS J3 69 45.0167 7.8083 32.8000 71.0600Ab3-DNPS noCDR.J 685 38.8751 10.0523 27.0500 92.5000Ab3-DNPS J.All 965 39.3650 9.3815 27.0500 92.5000 ==================================================================Ab3-DNPS M1 99 59.6133 16.5996 45.7500 100.3900Ab3-DNPS M2 43 99.4972 7.6719 92.5900 100.8300Ab3-DNPS M3 6 68.0900 1.3909 65.3500 69.6300Ab3-DNPS noCDR.M 532 64.1230 18.0307 34.0900 100.4900Ab3-DNPS M.All 680 65.7384 19.4376 34.0900 100.8300 ==================================================================Ab3-DNPS I1 10 82.8310 0.7650 81.4100 83.6300Ab3-DNPS I3 8 100.3000 0.2030 100.1000 100.9000Ab3-DNPS noCDR.I 203 98.8150 9.4547 85.5100 100.7600Ab3-DNPS I.All 221 99.1908 9.9789 81.4100 100.7600==================================================================Ab3-DNPS O1 99 59.6477 16.6268 45.7200 100.9400Ab3-DNPS O2 43 98.5388 7.6777 92.9300 100.1900Ab3-DNPS O3 6 68.2750 1.4360 65.4600 69.9200Ab3-DNPS noCDR.O 532 64.6452 18.3012 33.8800 100.8500Ab3-DNPS O.All 680 66.0929 19.5021 33.8800 100.1900 ==================================================================Ab3-DNPS K3 2 99.4350 0.3050 99.7300 100.0400Ab3-DNPS noCDR.K 191 94.0563 9.5262 87.5800 100.2200Ab3-DNPS K.All 193 94.1151 9.4942 87.5800 100.2200===================================================================================================================================================Structure Subset Natoms <B> rmsB MinB MaxB=================================================================================Ab3-Fur L1 99 39.1635 4.5709 30.0100 53.3900Ab3-Fur L2 43 41.7907 6.2487 34.1900 59.8900Ab3-Fur L3 95 34.5964 4.1993 27.7700 49.8700Ab3-Fur noCDR.L 560 34.9973 7.7605 21.6000 100.0900Ab3-Fur L.All 797 35.8336 7.2860 21.6000 100.0900 ==================================================================Ab3-Fur H1 91 40.2389 5.3429 30.6700 51.5300Ab3-Fur H2 120 37.9360 7.8445 23.1500 61.7500Ab3-Fur H3 69 47.5739 9.1114 34.6500 74.3500Ab3-Fur noCDR.H 688 39.8055 10.4015 26.2500 91.3600Ab3-Fur H.All 968 40.1682 9.8867 23.1500 91.3600 ==================================================================Ab3-Fur N1 99 39.1569 4.6281 29.7800 53.8200Ab3-Fur N2 43 41.9147 6.0930 34.1600 59.8300Ab3-Fur N3 95 34.6096 4.2215 28.0300 50.3300Ab3-Fur noCDR.N 560 34.9623 7.7693 22.6700 100.2800Ab3-Fur N.All 797 35.8164 7.2993 22.6700 100.2800 ==================================================================Ab3-Fur J1 91 40.2507 5.3702 30.8900 51.9200Ab3-Fur J2 120 37.9142 7.9036 23.2300 61.8500Ab3-Fur J3 69 47.6117 9.0813 34.5300 74.2400Ab3-Fur noCDR.J 688 39.7630 10.4318 26.1900 91.3000Ab3-Fur J.All 968 40.1391 9.9184 23.2300 91.3000 ==================================================================Ab3-Fur M1 62 78.6148 8.7290 67.8300 100.2800Ab3-Fur M2 27 83.2063 7.9807 75.4500 100.8000

Ab3-Fur M3 29 74.9514 6.4663 63.4500 82.5800Ab3-Fur noCDR.M 543 61.8104 11.8280 39.3300 94.8300Ab3-Fur M.All 737 64.8371 13.0427 39.3300 100.2800 ==================================================================Ab3-Fur noCDR.I 26 87.1273 2.4200 84.4400 93.8100Ab3-Fur I.All 26 87.1273 2.4200 84.4400 93.8100==================================================================Ab3-Fur O1 56 78.6264 8.3197 67.5800 100.2400Ab3-Fur O2 43 80.7719 7.0669 72.0300 100.4400Ab3-Fur O3 29 74.9383 6.4823 63.5800 82.6400Ab3-Fur noCDR.O 540 61.8319 11.9422 39.1000 94.6500Ab3-Fur O.All 668 65.0280 13.0426 39.1000 100.2400 ==================================================================Ab3-Fur noCDR.K 26 87.1927 2.4242 84.6100 94.0300Ab3-Fur K.All 26 87.1927 2.4242 84.6100 94.0300===================================================================================================================================================Structure Subset Natoms <B> rmsB MinB MaxB=================================================================================Ab4_Trx-Shear3 L1 99 54.9238 10.3113 27.7000 92.6500Ab4_Trx-Shear3 L2 43 52.0891 7.9937 39.7300 74.9100Ab4_Trx-Shear3 L3 95 50.2483 13.3794 29.6700 95.9600Ab4_Trx-Shear3 noCDR.L 558 41.8446 10.4616 20.7000 100.0000Ab4_Trx-Shear3 L.All 795 45.0316 11.8392 20.7000 100.0000 ==================================================================Ab4_Trx-Shear3 H1 91 51.8475 9.3499 31.1900 69.1400Ab4_Trx-Shear3 H2 123 40.6815 14.1237 21.9400 92.6000Ab4_Trx-Shear3 H3 69 45.2326 5.6889 37.1900 67.2700Ab4_Trx-Shear3 noCDR.H 680 43.9925 13.9233 22.0700 100.0000Ab4_Trx-Shear3 H.All 963 44.4007 13.4209 21.9400 100.0000 ==================================================================Ab4_Trx-Shear3 N1 99 57.5338 10.3113 30.3100 95.2600Ab4_Trx-Shear3 N2 43 54.6991 7.9937 42.3400 77.5200Ab4_Trx-Shear3 N3 95 52.8583 13.3794 32.2800 98.5700Ab4_Trx-Shear3 noCDR.N 558 44.4499 10.4362 23.3100 100.0000Ab4_Trx-Shear3 N.All 795 47.6384 11.8243 23.3100 100.0000 ==================================================================Ab4_Trx-Shear3 J1 91 55.2975 9.3499 34.6400 72.5900Ab4_Trx-Shear3 J2 123 44.1315 14.1237 25.3900 96.0500Ab4_Trx-Shear3 J3 69 48.6826 5.6889 40.6400 70.7200Ab4_Trx-Shear3 noCDR.J 680 47.4227 13.8464 25.5200 100.0000Ab4_Trx-Shear3 J.All 963 47.8367 13.3650 25.3900 100.0000

==================================================================

Table S1: Values are given by chain for; all atoms (chain.All), scaffold atoms only

(noCDR.chain), and each CDR. Ab1 and Ab2 have one copy per asymmetric unit

(chains L and H), Ab3 has four copies (chains L,N,M,O & H,I,J,K) and Ab4 has two

copies (chains L,N and H,J).

Table S2: Interactions between SPE7 Fv and its various ligands*

Antibody DNP-Ser Furazolidone Aq AzResidue Atom Atom Atom Atom Atom

O41L36 Nδ2O42

H50 Nε O62 O1O61 O5 O2 O3H50 Nη2O1 O6

H59 N2 O1 O4L93 Nε1 Oδ O1 O2

O62 O2 OH35 Nε2O3

Antibody ThioredoxinResidue Atom Residue AtomH101 O A107 NH105 OH A104 NζL34 OH A109

A108NN

H105 tyrosine ring A106 peptide backboneH102 tyrosine ring A35 Isoleucine side-chainH32 OH A41 NEH26 O A41 Nη1H28 Oδ1 A41 Nη2

* Residues and atoms involved in planar stacking or charged interactions are indicated initalics.

(4) Selection of the Trx-protein ligands

The Trx-protein ligands (Figure 4 of the article) were isolated using the FliTrx system

(Invitrogen) that displays on the surface of E. coli a cyclic peptide library in place of the

active site loop of thioredoxin (10). SPE7 was immobilized on plates and incubated with

the library. Unbound bacteria were washed away and bound bacteria removed either by

applying shear stress or by adding 0.5 mM DNP. Initial estimates of affinity were

determined by ELISA using an HRP labeled anti-thioredoxin antibody. Selected clones

were then sequenced and the Trx proteins re-cloned into the Invitrogen pCR-T7-TOPO

vector, expressed and purified by Ni-NTA and size exclusion chromatography. Affinities

were then measured by ELISA using purified protein. SPE7/Trx complexed crystals

were prepared by adding equimolar amounts of the separately prepared proteins.

References

1. Z. Eshhar, M. Ofarim, T. Waks, J. Immunol. 124, 775 (1980).2. J. McCafferty, Hoogenboom, H. R., Chiswell, D. J., Antibody Engineering. B. D.

Hames, Ed., PAS (IRL Press, ed. 1, 1996).3. T. Simon, K. Rajewsky, Protein Eng. 5, 229 (1992).4. A. B. Lindner, Z. Eshhar, D. S. Tawfik, J. Mol. Biol. 285, 421 (1999).5. CCP4., Acta Cryst. D50, 760-763 (1994).6. T. A. Jones, Zou, J. Y., Cowan, S. W., Kjeldgaard, Acta Crystallogr. A 47, 110

(1991).7. P. Roversi, E. Blanc, C. Vonrhein, G. Evans, G. Bricogne, Acta Crystallogr. D. 56,

1316 (2000).8. R. S. Herbst-Irmer, G. M., Acta Cryst. B54, 443 (1998).9. A. T. Brunger et al., Acta Crystallogr. D Biol. Crystallogr. 54 ( Pt 5), 905 (1998).10. Z. Lu et al., Biotechnology 13, 366 (1995).