Continuous Evolution of Proteases with Altered Specificity

Continuous Evolution of Proteases With Altered Specificity

The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters

Citation Packer, Michael Samuel. 2017. Continuous Evolution of Proteases With Altered Specificity. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:41141523

Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#LAA

Continuous Evolution of Proteases with Altered Specificity

A dissertation presented !

Michael S. Packer

The Committee on Higher Degrees in Biophysics

in partial fulfillment of the requirements

!for the degree of

Doctor of Philosophy

!in the subject of !

Biophysics

Harvard University Cambridge, Massachusetts

April 2017

Dissertation Advisor: Professor David R. Liu Michael S. Packer

Continuous Evolution of Proteases with Altered Specificity

Abstract

The following thesis work aims to establish a method for the generation of proteases with tailor-made substrate specificities. This programmability will enable the design of proteases that modulate the activity of a target protein of biotechnological or therapeutic relevance. In the first phase of this project, we developed and validated a system for the phage-assisted continuous evolution (PACE) of protease enzymes. In pilot feasibility studies, we used this system to evolve hepatitis C virus proteases that are resistant to small-molecule protease inhibitors. We then established that this PACE selection could also evolve specificity changes in two different model proteases: human rhinovirus (HRV) protease and tobacco etch virus (TEV) protease. After these proof-of-concept experiments, we began the phage-assisted continuous evolution (PACE) of TEV protease, which canonically cleaves

ENLYFQS, to cleave a very different target peptide sequence, HPLVGHM, that is present in IL-23, a cytokine implicated in inflammatory disease. A protease emerging from approximately 2,500 generations of PACE contains

20 non-silent mutations, cleaves human IL-23 at the target peptide bond, and inhibits IL-23 mediated signaling in cultured primary murine splenocytes. We characterized the substrate specificity of this evolved enzyme using a protease specificity profiling method, revealing a mixture of shifted and broadened specificity at the six positions in which the target amino acid sequence differed. On-going studies seek to expand the scope of protease PACE through: (1) the development of a negative selection scheme to ablate off-target proteolysis, (2) the adaptation of disulfide compatible E. coli strains for the PACE of human circulating proteases, and (3) the evolution Botulinum toxin proteases to enable the destruction of intracellular target proteins.

iii

Table of Contents

Table of Contents ...... iv! Table of Figures ...... v! List of Tables ...... vi! Acknowledgements ...... vii! Introduction ...... 8! Chapter 1: Design and Validation of a Positive Selection Scheme for the Continuous Evolution of Proteases 11! Section 1.1: Transducing protease activity into gene expression ...... 12! Section 1.2: Linking protease activity to phage propagation ...... 20! Section 1.3: Continuous evolution of resistance to HCV protease inhibitors ...... 22! Section 1.4: Continuous evolution of model proteases with altered specificity ...... 28! Section 1.5: Discussion ...... 32! Section 1.6: Methods ...... 34! Chapter 2: Continuous Evolution of a Model Protease to Cleave a Disease-Associated Human Protein ...... 39! Section 2.1: Choice of target substrate and protease ...... 40! Section 2.2: Continuous Evolution of TEV Variants that Cleave the IL-23 Target Peptide ...... 44! Section 2.3: Characterization of Evolved TEV Protease Variants ...... 58! Section 2.4: Substrate Specificity Profiling of an Evolved TEV Protease ...... 61! Section 2.5: Specificity Profiling Reveals Functionally Independent TEV Mutation Groups ...... 68! Section 2.6: Evolved TEV L2F Cleaves Human IL-23 ...... 70! Section 2.7: Evolved TEV Protease Deactivates IL-23 and Prevents IL-17 Secretion ...... 72! Section 2.8: Discussion ...... 76! Section 2.9: Methods ...... 78! Chapter 3: Extending the Utility of Protease PACE ...... 86! Section 3.1: Design of a negative selection scheme to ablate off-target proteolysis...... 87! Section 3.2: Validation of simultaneous positive and negative selection in protease PACE...... 90! Section 3.3: Attempts at PACE of FDA-approved therapeutic proteases containing disulfide linkages...... 93! Section 3.4: Adapting protease PACE to botulinum toxin light chain...... 95! Section 3.5: BoNT-LC can be evolved in protease PACE...... 99! Section 3.6: Evolutionary target substrates for BoNT-LC E...... 103! Section 3.7: Evolution of BoNT-LC E proteases that cleave PTEN...... 106! Section 3.8: Discussion...... 112! Section 3.9: Methods...... 114! References ...... 117!

Table of Figures

Figure 1: Development of a system to link protease activity to gene expression...... 14! Figure 2: Crystal structure of T7 lysozyme bound to T7 RNAP...... 15! Figure 3: Vector map of complementary plasmid (CP)...... 16! Figure 4: Vector map of expression plasmid (EP)...... 17! Figure 5: Vector map of accessory plasmid (AP)...... 18! Figure 6: Vector map of selection phage (SP)...... 19! Figure 7: Western blot showing PA-RNAP cleavage...... 20! Figure 8: PA-RNAPs link protease activity to phage propagation...... 22! Figure 9: HCV PA-RNAP response to protease inhibitors in E. coli cells...... 25! Figure 10: Vector map of mutagenesis plasmid (MP)...... 26! Figure 11: Continuous evolution of drug resistance in HCV protease...... 27! Figure 12: Vector map of accessory plasmid (AP) used in protease specificity reprogramming experiments...... 29! Figure 13: PACE of TEV proteases with altered P1 specificity...... 30! Figure 14: PACE of HRV proteases with altered P1’ specificity ...... 31! Figure 15: Evolutionary trajectories and representative evolved TEV protease genotypes...... 45! Figure 16: Luciferase activity assay of clones from the middle of PACE stage 1 of trajectories 1, 2, and 3...... 46! Figure 17: Luciferase activity assay after PACE stage 2 of trajectories 1 and 2...... 47! Figure 18: Luciferase activity assay after PACE stage 2 of trajectory 3...... 48! Figure 19: Luciferase activity assay of clones after PACE stage 4...... 49! Figure 20: Validation of protease PACE stringency modulation...... 50! Figure 21: Luciferase activity assay of clones after PACE stage 8...... 51! Figure 22: Epistatic interactions with TEV protease residue N177...... 58! Figure 23: Protein cleavage assay to identify the most active clone...... 59! Figure 24: HPLC assay of TEV protease kinetics...... 60! Figure 25: Evolved TEV protease cleaves wild-type, intermediate, and target substrates...... 61! Figure 26: Protease specificity profiling...... 63! Figure 27: Specificity profiles generated from libraries with three randomized substrate amino acids...... 66! Figure 28: TEV protease variants containing subsets of TEV L2F mutations are all active...... 68! Figure 29: Specificity profiles of TEV variants possessing wild-type-like specificity...... 69! Figure 30: Identification of IL-23 cleavage sites by Western blot and LC-MS...... 70! Figure 31: Identification of the cleavage site within IL-23 heterodimer by mass spectrometry...... 71! Figure 32: Identification of two cleavage sites within IL-23 monomer by mass spectrometry...... 71! Figure 33: Protease-mediated attenuation of IL-17 secretion in mouse splenocytes...... 73! Figure 34: Western blot of pre-mixed additives to splenocyte cell culture...... 73! Figure 35: Western Blot of pre-mixed additives to splenocyte cell culture...... 74! Figure 36: TEV L2F catalytically deactivates IL-23 and prevents IL-17 secretion in mouse splenocytes...... 74! Figure 37: TEV L2F must be pre-incubated with IL-23 to prevent IL-17 secretion in mouse splenocytes...... 75! Figure 38: TEV L2F is unaffected by the addition of FBS to in vitro cleavage assays...... 75! Figure 39: Schematic representation of negative selection components...... 87! Figure 40: Luciferase assay screen of polymerase variants for an orthogonal PA-RNAP...... 88! Figure 41: Vector map of negative selection AP...... 89! Figure 42: Simultaneous selection for cleavage of ENLYAQS and against cleavage of ENLYFQS...... 91! Figure 43: Luciferase assay of disulfide containing proteases in the presence of Erv1 and DsbC...... 94! Figure 44: SNARE-derived PA-RNAPs exhibit protease-dependent gene expression with cognate BoNT-LC...... 97! Figure 45: SNAP25 residues 187-206 disrupt PA-RNAP leading to high background transcription...... 98! Figure 46: PACE-evolved BoNT-LC B and F variants with high apparent activity on wild-type substrate...... 100! Figure 47: Luciferase assay of PACE-evolved BoNT-LC E clones exhibit apparent proteolysis of SNAP23...... 102! Figure 48: Luciferase assay of wild-type BoNT-LC E on a panel of point mutant substrates...... 107! Figure 49: Luciferase assay of evolved BoNT-LC E exhibit apparent proteolysis of stepping-stone two and three. 109!

List of Tables

Table 1: Target peptide scoring matrix based upon wild-type TEV specificity...... 40! Table 2: Protease target substrates identified by specificity scoring...... 41! Table 3: TEV protease mutations after PACE on HNLYFQS substrate...... 41! Table 4: Target peptide scoring matrix based upon known evolutionary potential...... 42! Table 5: Target substrates identified by TEV protease evolutionary potential...... 43! Table 6: PACE stage 1 mutations...... 46! Table 7: PACE stage 2 mutations...... 47! Table 8: PACE stage 2 mutations...... 48! Table 9: PACE stage 3 mutations...... 52! Table 10: PACE stage 4 mutations...... 53! Table 11: PACE stage 5 mutations...... 54! Table 12: PACE stage 6 mutations...... 55! Table 13: PACE stage 7 mutations...... 56! Table 14: PACE stage 8 mutations...... 57! Table 15: Kinetic parameters of wild-type and evolved TEV proteases...... 60! Table 16: Phage display enrichment values from selections on single site libraries...... 64! Table 17: Phage display enrichment values from selections on libraries with three randomized residues...... 67! Table 18: Plasmids used for PACE and protein expression...... 85! Table 19: Sanger and Illumina sequencing DNA primers...... 85! Table 20: Overnight propagation assay of simultaneous positive and negative selection...... 90! Table 21: Simultaneous positive and negative selection is compatible with protease PACE...... 90! Table 22: BoNT-LC B genotypes after 72h PACE on AP encoding VAMP1 60 amino acid substrate...... 99! Table 23: BoNT-LC F genotypes after 72h PACE on AP encoding VAMP1 60 amino acid substrate...... 99! Table 24: BoNT-LC E genotypes after 72h PACE on AP encoding SNAP25 substrates...... 100! Table 25: BoNT-LC E genotypes after PACE on AP encoding SNAP25 D179K or SNAP23...... 101! Table 26: Target peptide scoring matrix based upon wild-type BoNT-LC E specificity...... 104! Table 27: Candidate target substrates identified via wild-type BoNT-LC E specificity scoring...... 104! Table 28: Target peptide scoring matrix based upon BoNT-LC E specificity and evolutionary potential...... 105! Table 29: PTEN target substrate identified via revised scoring method...... 105! Table 30: Evolutionary stepping-stone substrates for lagoon 1...... 107! Table 31: Evolutionary stepping-stone substrates for lagoon 2...... 107! Table 32: Evolutionary stepping-stone substrates for lagoon 3...... 108! Table 33: Evolutionary stepping-stone substrates for lagoon 4...... 108! Table 34: BoNT-LC E genotypes after 72h PACE on AP encoding SNAP25 D179C or R180S substrates...... 108! Table 35: BoNT-LC E genotypes after 72h PACE on AP encoding SNAP25 D179C and R180S substrate...... 110! Table 36: BoNT-LC E genotypes after 72h PACE with mixing-transition from stepping-stone three to four...... 111!

Acknowledgements

Alon Rivel, my husband, for seeing me through the entire PhD and beyond.

My mom, dad, and sister for raising me to be curious, creative, and brave (A.K.A. a Scientist).

David R. Liu for mentoring me and serving as a role model. He has combined a group of fantastic researchers with ample scientific resources to create an environment where anything seems possible.

Aleks Markovic for keeping the Liu Lab running like a well-oiled machine.

Bryan Dickinson for taking me on as a rotation student, and then graciously allowing me to continue his work in setting up a protease PACE system. Without his pioneering efforts my thesis studies could not have transpired

Travis Blum; it has been my honor to have someone even interested in continuing where my work left off. It is an even greater pleasure to see the amazing directions only you could take it.

Holly Rees for assistance both technical and emotional in the execution of primary splenocyte cell culture assays.

Sunia Trauger for saving me from LC-MS purgatory and helping me see the light at the end of the tunnel.

Jacob Carlson, Kevin Esvelt, Aaron Leconte (PACE forefathers whom I have never met); for all the hardwork that went into this technology before my time in the Liu Lab.

9AM Millis for being an incredible PACE troubleshooting resource (Jeff Bessen, David Bryson, Liwei Chen, Kevin

Davis, Bryan Dickinson, Sherry Gao, Johnny Hu, Basil Hubbard, Tim Roth, Ning Sun, David Thompson, Travis

Blum, Nicole Gaudelli, Tina Wang, Ben Thuronyi, Mary Morrison, Ahmed Badran, Lena Afeyan).

Ahmed Badran for answering all of my molecular biology questions and being a guiding force for all PACErs.

Nicole Gaudelli, Alexis Komor, Bill Kim for being amazing friends and coworkers. I am forever grateful for being included in your genius Cas9 projects. It was a welcome distraction and a great learning experience.

Current and former Liusers that I have not yet mentioned all played a critical role in making the Liu Lab a fantastic work place: Chihui An, Brian Chaikind, Ryan Hili, Rick McDonald, John Zuris, Brent Dorr, John

Guilinger, Margie Li, Lynn McGregor, Jia Niu, Adrian Berliner, Alix Chan, Jonathan Chen, Luke Koblan, Jon Levy,

Phillip Lichtor, JP Maianti, Chris Podracky, Weixin Tang, Dmitry Usanov, Ariel Yeh, Christina Zeina, Manda

Arbab, James Nelson.

Jim Hogle and Michele Jakoulov for running Harvard Biophysics. You have created a PhD program with unparalleled scientific and academic independence.

Funding from NIH, DARPA, Broad Institute, NSF Graduate Research Fellowship and IPSEN Pharmaceuticals.

vii

Introduction

Parts adapted from: Dickinson, Packer, Badran, Liu, Nature Communications. 5 (2014).

Among the more than 600 naturally occurring proteases that have been described1 are enzymes that have proven to be important catalysts of industrial processes, essential tools for proteome analysis, and life-saving pharmaceuticals2-5. Recombinant human proteases including thrombin, factor VIIa, and tissue plasminogen activator are widely used drugs for the treatment of blood clotting diseases4. Researchers have engineered or evolved industrial proteases with enhanced thermostability and solvent tolerance6, 7. Similarly, a handful of therapeutic proteases have been engineered with improved kinetics and prolonged activity8-10. The potential of proteases to serve as a broadly useful platform for degrading proteins implicated in disease, however, is greatly limited by the native substrate scope of known proteases. In contrast to the highly successful generation of therapeutic monoclonal antibodies with tailor-made binding specificities11, the generation of proteases with novel protein cleavage specificities has proven to be a major challenge. For example, efforts to engineer trypsin variants that exhibit the substrate specificity of the closely related protease chymotrypsin were unsuccessful until researchers grafted the entire substrate-binding pocket, multiple surface loops, and additional residues from chymotrypsin12-15. This approach of replacing protease residues with amino acids from related proteases to impart specificity features from the latter16, 17 cannot provide proteases with specificities not already known among natural proteases, prompting researchers to instead turn to laboratory evolution to generate proteases with novel specificities18-22. Despite several decades of effort, no evolved proteases have yet been reported with more than one position of changed substrate specificity.

The evolution of a protease that can degrade a target protein of interest will almost always require changing substrate sequence specificity at more than one position, and thus may require many generations of evolution.

Continuous evolution strategies, which require little or no researcher intervention between generations23, therefore may be well-suited to evolve proteases capable of cleaving a target protein that differs substantially in sequence from the preferred substrate of a wild-type protease. In phage-assisted continuous evolution (PACE), a population of evolving selection phage (SP) is continuously diluted in a fixed-volume vessel by an incoming culture of host E. coli24. The SP is a modified M13 bacteriophage genome in which the evolving gene of interest has replaced gene III, a gene essential for phage infectivity. If the evolving gene of interest possesses the desired activity it will trigger expression of gene III from an accessory plasmid (AP) in the host cell, thus producing infectious progeny encoding active variants of the evolving gene. The mutation rate of the SP is controlled using an inducible mutagenesis plasmid (MP) such as MP6, which upon induction increases the mutation rate of the SP by > 300,000-fold25.

Because the rate of continuous dilution is slower than phage replication but faster than E. coli replication, mutations only accumulate in the SP.

Here we describe the development and application of a system for the continuous directed evolution of proteases. This system uses an engineered protease-activated RNA polymerase (PA-RNAP) to transduce polypeptide cleavage events into changes in gene expression that support phage propagation during PACE. We validate that this system successfully links the phage life cycle to protease activity for three distinct proteases. When performed in the presence of danoprevir or asunaprevir, two hepatitis C virus (HCV) protease inhibitor drug candidates currently in clinical trials, protease PACE rapidly evolved HCV protease variants that are resistant to each drug candidate. The PACE-evolved HCV protease variants are dominated by mutations previously observed in patients treated with these drug candidates. Encouraged by these results, we then used PACE to evolve both human rhinovirus (HRV) and tobacco etch virus (TEV) proteases that cleave substrates containing single amino acid substitutions at critical positions in their substrate recognition sequences. Together, these findings establish protease

PACE as a platform to reveal the vulnerability of protease inhibitors to the evolution of drug resistance and to rapidly generate proteases with novel specificities.

We then applied this platform to dramatically reprogram the specificity of TEV protease, which natively cleaves the consensus substrate sequence ENLYFQS, to cleave a target sequence, HPLVGHM, that differs at six of seven positions from the consensus substrate and is present in an exposed loop of the pro-inflammatory cytokine IL-

23. After constructing a pathway of evolutionary stepping-stones and performing ~2,500 generations of evolution using PACE, the resulting proteases contain up to 20 amino acid substitutions, cleave human IL-23 at the intended target peptide bond, and block the ability of IL-23 to stimulate IL-17 production in a murine splenocyte assay. These results establish a strategy for generating proteases with substrate specificities changed at several positions and the ability to cleave proteins implicated in human disease.

After such dramatic progress with a model enzyme (TEV protease), we sought to expand the purview of protease PACE to a number of FDA-approved protease therapeutics. Although, we were unable to adapt PACE to the evolution of enzymes containing disulfide linkages, which comprise the majority of therapeutic proteases, FDA approved C. botulinum toxins contain a metalloprotease domain absent of any disulfide linkages. These proteases have proven to be amenable to substrate specificity reprogramming in PACE, and because these toxins mediate intracellular delivery, they open up novel therapeutic avenues that we are only just beginning to explore.

Chapter 1: Design and Validation of a Positive Selection Scheme for the

Continuous Evolution of Proteases

Parts adapted from: Dickinson, Packer, Badran, Liu, Nature Communications. 5 (2014).

Section 1.1: Transducing protease activity into gene expression

PACE requires that a target activity be linked to changes in the expression of an essential phage gene such as gene III (gIII). To couple the cleavage of a polypeptide substrate to increases in gene expression, we engineered a

PA-RNAP that transduces proteolytic activity into changes in gene expression that are sufficiently strong and rapid to support PACE. T7 RNA polymerase (T7 RNAP) is naturally inhibited when bound to T7 lysozyme26. We envisioned that T7 lysozyme could be tethered to T7 RNAP through a flexible linker containing a target protease cleavage site. Ideally, the effective concentration of the tethered T7 lysozyme with respect to T7 RNAP would be sufficiently high that the T7 RNAP subunit would exist predominantly in the T7 lysozyme-bound, RNAP-inactive state. Proteolysis of the target sequence would disfavor the bound T7 RNAP:T7 lysozyme complex, resulting in the liberation of an active T7 RNAP and expression of gIII placed downstream of a T7 promoter (Figure 1A).

N-terminal fusions to T7 RNAP are known to be well tolerated, and in the crystal structure of T7 RNAP bound to T7 lysozyme, the C-terminus of T7 lysozyme is only 32 Å from the N-terminus of T7 RNAP, separated by a solvent-exposed channel27 (Figure 2). In light of this structural information, we linked the two proteins through these proximal termini. Since T7 lysozyme activity is toxic to host E. coli cells, we characterized catalytically inactive lysozyme variants and found that the inactive C131S lysozyme mutant retained its ability to inhibit T7

RNAP without impairing host cell viability.

To identify T7 RNAP–T7 lysozyme linkers that promote complex formation and result in an inactive polymerase subunit yet permit efficient proteolysis, we screened a small set of linkers consisting of Gly, Ser, and

Ala ranging in length from three to ten residues flanking each side of a target protease substrate. We designed PA-

RNAP constructs containing linker peptide sequences known to be cleaved by tobacco etch virus (TEV) protease,

HCV protease, or human rhinovirus-14 3C (HRV) protease. We assayed T7 RNAP activity using a luciferase reporter and observed that T7 lysozyme linked to T7 RNAP through at least 28 residues including the target protease substrate resulted in significant inhibition of RNAP activity (Figure 1B). To assay RNAP activation, we coexpressed each PA-RNAP variant from a plasmid (the complementary plasmid or CP, Figure 1C and Figure 3) together with each of the three proteases (expressed from the expression plasmid or EP, Figure 4) in E. coli cells that also harbored a plasmid encoding gIII and luciferase under control of the T7 promoter (the accessory plasmid or AP,

Figure 1C and Figure 5).

Expression of a protease that is not known to cleave the target amino acid sequence in a coexpressed PA-

RNAP did not result in enhanced gene expression as measured by luciferase activity (Figure 1D). In contrast, expression of a protease that is known to cleave the target sequence within the PA-RNAP resulted in 18- to 49-fold increase in gene expression for all three cognate combinations of protease and substrate. These data indicate that

PA-RNAPs are capable of transducing specific proteolytic cleavage activities into large changes in target gene expression.

Figure 1: Development of a system to link protease activity to gene expression. (a) Protease-activated RNA polymerase (PA-RNAP). T7 RNAP is fused to the natural inhibitor T7 lysozyme through a linker containing a protease target substrate sequence. While the linker is intact, the complex preferentially adopts the lysozyme-bound, RNAP-inactive state. Proteolysis of the target sequence favours dissociation of the complex, freeing active T7 RNAP to transcribe genes downstream of the T7 promoter. This study used an accessory plasmid (AP) in which the T7 promoter drives a tandem gIII-luciferase (lux) cassette. (b) Sequences of the protein linkers containing a target protease substrate used for each PA-RNAP, with T7 lysozyme residues in blue, protease substrates in red, T7 RNAP residues in green and linker regions in black. (c) Plasmids used for protease PACE. An AP that has gIII and luciferase (lux) under the control of the T7 promoter serves as the source of gIII in the cells. A complementary plasmid (CP) constitutively expresses a PA-RNAP variant with a protease target substrate sequence embedded in the linker. (d) PA-RNAP gene expression response in E. coli cells. Host cells were transformed with (i) an AP containing the T7 promoter driving gIII-lux; (ii) a CP that constitutively expresses a PA-RNAP including the TEV protease substrate, the HCV protease substrate, or the HRV protease substrate; and (iii) a plasmid that expresses TEV protease (orange bars), HCV protease (purple bars), or HRV protease (grey bars). Gene expression is activated only when the expressed protease cleaves the amino-acid sequence on the PA-RNAP sensor. The luminescence experiment was performed in triplicate with error bars indicating the standard deviation.

32 Å

T7 Lysozyme

T7 RNAP

Figure 2: Crystal structure of T7 lysozyme bound to T7 RNAP. Generated from PDB-1ARO27. The crystal structure of T7 RNAP in complex with the transcriptional inhibitor T7 lysozyme reveals that the carboxy-terminus of T7 lysozyme is approximate 32 angstroms away from the amino- terminus of T7 RNAP. This proximity makes possible the generation of an auto-inhibited fusion protein linked through a small flexible polypeptide linker.

Tuesday, May 20, 2014 1:39 PM Page 1 of 1 1303-1169

r te o 0 A 0 ad p m 0 a 15A ro 5 o p rig l in a n r r e r t n in b b t e l 1 a r 0 m p 0 i r 0 n o

a m

t o o

r t

Complementary Plasmid (CP) R

B S

s e d

0 5628 bp 8 0 m

y z

o s y l

r e k n T 0 i 7 0 L RN 0 AP 2 r e e c k n 3 in e 0 L u 00 q se on iti gn co re se tea Pro

Figure 3: Vector map of complementary plasmid (CP). This plasmid was used in HCV protease inhibitor resistance experiments. The complementary plasmid supplies constitutive expression of the protease-activated RNA polymerase (lysozyme-linker-T7 RNAP).

Tuesday, May 20, 2014 1:37 PM Page 1 of 1 pJC166d - SP R6K pBAD HCV

cat promoter

3500 r to a R6 in K gam ma rm or te igi l n a t n 0 a o 0 c 5 ti 0 0 c 3 0 e ir id f b 1

B n o t /

1 P Expression Plasmid (EP)

2 5 3542 bp 0

0 H

C 0 V 0 p 0 r o 1 t e a C s a e r a

2 00 R 0 0 P BS 50 B S 1 AD D p 8 ro mo ter Pc promoter

Figure 4: Vector map of expression plasmid (EP). This plasmid was used for arabinose-inducible protease expression during luminescence assays as well as for the purification of recombinant HCV protease.

Tuesday, May 20, 2014 1:40 PM Page 1 of 1 pJC173b - T1 term sd8 geneIII

bla promoter tor ina rm l te na tio ec 00 ir 60 R

id ) ep

b R A p r

L o

rf (A t a a

/o l

b n

tA i

e m

t r

1 e

0 t

0 l

0 a

o i

p t

S i

C r 1 c

0 0 s 1

0 n r

0 a

r e t

5 t o

Accessory Plasmid (AP) 1

m B

6533 bp 7

x ) l t u 0 x 8 n A 0 d e B 0 s d 2 n S e B p 4 R e 0 d 0 I 0 II e m y z o s ly 7 00 (T 30 or at in m ter T1 Figure 5: Vector map of accessory plasmid (AP). This plasmid was used in HCV protease inhibitor resistance experiments. The accessory plasmid supplied gene III and luxAB under control of the T7 promoter. The rrnB1 T1 terminator was placed downstream of the T7 promoter to lower the background transcription in the absence of protease that would activate the PA-RNAP.

Tuesday, May 20, 2014 1:38 PM Page 1 of 1 CE-SP-T7 P314T

i ) or l (+ tia complete (+) ori en gin ss tion e ca Gene II pro pli mot re er -) r ( al fo gn r si e g im in r g p a k A c G N a e R p II n 0 e 0 V 0 p 5 r o m 1 o 0 t 0 e 0 r

V X I

Selection Phage (SP) V

n I

I e 5892 bp V

I V r

4 e I t p 0 I X I I o r o 0 V m m 0 o o r t p e I I r I e g s a 0 I te 0 ro 0 r p 2 o V t HC a r in S to m B VI c r R fa te G er o 8 en ot c D e 300 m a S I p 0 ro 4 ro p NS mo 102 ter J23 Gene VI RBS

Figure 6: Vector map of selection phage (SP). This construct was used in HCV protease inhibitor resistance experiments. In this selection phage, gene III has been replaced by HCV protease followed by a J23102 promoter to replace the portion of the gene III coding sequence that also contains the gene VI promoter.

Section 1.2: Linking protease activity to phage propagation

Next we sought to use PA-RNAPs to link the life cycle of M13 bacteriophage to protease activity. We generated selection phage (SP) in which gIII was replaced by a gene encoding TEV protease, HCV protease, or

HRV protease (Figure 6). Without pIII, these phage are unable to propagate on wild-type E. coli cells. We engineered host E. coli cells containing two plasmids: (i) an AP that contains gIII and luciferase under the control of the T7 promoter, and (ii) a CP that constitutively expresses a PA-RNAP (Figure 1C). To be sure that the PA-RNAP selection scheme works as intended we analyzed the cleavage of the sensor by Western blot. We observed the loss of the Lysozyme-RNAP fusion and the formation of a new protein that corresponds to the size of T7 RNAP exclusively in the presence of protease phage that recognizes the host encoded PA-RNAP (Figure 7). To assay whether the host cells could support phage propagation in a protease-dependent manner, we performed activity- dependent plaque assays. We observed that plaque formation, a consequence of phage replication in solid media, only occurred with phage encoding a protease that can cleave the PA-RNAP within the host cells. Phage with mismatched protease/PA-RNAP combinations did not form plaques, indicating that phage encoding non-cognate proteases do not replicate, or replicate at a significantly reduced rate. These observations together establish that the

PA-RNAP system is capable of transducing protease activity of a phage-encoded protease into phage production.

PA-RNAP HCV HRV TEV

Phage None HCV HRV TEV None HCV HRV TEV None HCV HRV TEV

PA-RNAP 118kDa

Cleaved T7 RNAP 99kDa

Figure 7: Western blot showing PA-RNAP cleavage. The PA-RNAP is cleaved only by the protease that is known to recognize the target sequence. Band sizes of ~120 kDa and ~100 kDa correspond to the full PA-RNAP construct and the cleaved RNAP, respectively.

We next tested if the PA-RNAP-based selection supports the continuous propagation of phage encoding active proteases in the continuous liquid culture format required for PACE (Figure 8A). We maintained three host cell cultures, each harboring a CP expressing a PA-RNAP containing one of the three protease cleavage sites (TEV,

HCV, or HRV protease substrates), using chemostats diluted with fresh growth media at a fixed rate28. Each of these

host cell cultures continuously diluted lagoons seeded with various combinations of phage containing TEV, HCV, or

HRV protease. Lagoons seeded with phage encoding cognate proteases that can cleave the PA-RNAP within the host cells robustly propagated (108-1010 pfu mL-1 after 72 hours of continuous dilution at 1.0 lagoon volume per hour), while lagoons seeded with phage encoding proteases that do not match the PA-RNAP of incoming host cells washed out (< 104 pfu mL-1), demonstrating protease activity-dependent propagation in continuous liquid culture.

In order to determine if this system can selectively replicate phage carrying protease genes with a desired activity at the expense of phage encoding proteases that are unable to cleave the host-cell PA-RNAP, we performed protease phage enrichment experiments in a PACE format. We seeded a lagoon with a 1,000:1 ratio of TEV

SP:HCV SP, then allowed the phage to propagate in the lagoon while being continuously diluted with host cells containing a PA-RNAP with the HCV protease recognition site. We periodically sampled the waste line of the lagoon and amplified by PCR the region of the phage containing the protease genes. The TEV protease and HCV protease genes are readily distinguishable as PCR amplicons of distinct lengths. At the start of the experiment the

HCV protease phage were virtually undetectable by PCR amplification of the starting population and gel electrophoresis, while TEV protease dominated the lagoon (Figure 8B). After just 24 h of continuous propagation on host cells containing the HCV PA-RNAP, the TEV protease SPs were undetectable, while the HCV protease SPs were strongly enriched (≥ 100,000-fold enrichment over 24 hours).

We repeated this experiment with a 1,000-fold excess of HCV protease phage over TEV protease phage using host cells containing the TEV protease PA-RNAP (Figure 8C), and a third time using a 1,000-fold excess of

TEV protease phage over HRV phage and host cells containing the HRV protease PA-RNAP (Figure 2D). In all three of the enrichment experiments, continuous propagation rapidly and dramatically enriched phage encoding each cognate protease from a minute fraction of the starting phage mixture, while non-cognate proteases washed out of the lagoon (Figure 8). Collectively, these results indicate that this protease PACE system successfully links specific protease activity to the phage life cycle in a continuous flow format and can strongly and rapidly enrich phage that encode proteases with the ability to cleave a target polypeptide substrate.

Figure 8: PA-RNAPs link protease activity to phage propagation. (a) The protease PACE system. Fixed volume vessels (lagoons) contain phage in which gIII is replaced with a gene encoding an evolving protease. The lagoon is fed with host cells that contain an AP with the T7 promoter driving gIII and a CP that expresses a PA-RNAP. Phages infect incoming cells and inject their genome containing a protease variant. Only if the protease variant can activate the PA-RNAP by cleaving the linker encoding the target protease substrate, gIII is expressed and that SP can propagate. (b–d) Enrichment of active proteases from mixed populations using PACE. At time 0, a lagoon was seeded with a 1,000-fold excess of non-cognate protease-encoding phage over cognate protease-encoding phage. The lagoon was continuously diluted with host cells containing a PA-RNAP with either the HCV (b), TEV (c) or HRV (d) protease substrates. Lagoon samples were periodically analysed by PCR. In all three cases, phage encoding the cognate protease were rapidly enriched in the lagoon while phage encoding the non-cognate protease were depleted.

Section 1.3: Continuous evolution of resistance to HCV protease inhibitors

As an initial application of protease PACE, we continuously evolved protease enzymes to rapidly assess the drug resistance susceptibility of small-molecule protease inhibitors. Several HCV protease inhibitors are in late- stage clinical trials or are awaiting FDA approval29, 30. For some HCV protease inhibitor drug candidates, clinically isolated drug resistance mutations are known31. First we tested whether small-molecule HCV protease inhibitors can modulate protease activity in the protease PACE system. We observed that the incubation of host cells with either

32 33 danoprevir (IC50 = ~0.3 nM) or asunaprevir (IC50 = ~1.0 nM) , two second-generation HCV protease inhibitors, inhibited the cellular gene expression arising from the activity of HCV protease on the HCV PA-RNAP in a dose- dependent manner (Figure 9). These observations suggest that protease inhibitors can create selection pressure during PACE favoring the evolution of protease mutants that retain their ability to cleave a cognate substrate despite the presence of the drug candidates.

Based on the relationship between protease inhibitor concentration and gene expression in our system

(Figure 9) and initial trial PACE experiments, we selected 20 µM danoprevir as the final concentration to use in the culture media during attempts to continuously evolve drug-resistant HCV proteases. We inoculated two separate lagoons with HCV protease SP and propagated the phage on host cells containing the HCV protease PA-RNAP in the absence of any inhibitor for 6 h to allow the accumulation of mutations in HCV protease genes. Next, we added

20 µM danoprevir to the media that feeds into the host cell culture, and eventually into each of the two replicate lagoons. As a control, we propagated two replicate lagoons of HCV protease phage on HCV protease PA-RNAP host cells with no added protease inhibitor for the same time period. Throughout all of these experiments, we induced enhanced mutagenesis of the phage genome by activating a second generation mutagenesis plasmid (MP) in the host cells with 0.5% arabinose (Figure 10).

Phage populations at 6 and 28 h from replicate lagoons were analyzed by high-throughput DNA sequencing. No mutations were substantially enriched in the control lagoons propagated in the absence of any drug candidate (Figure 11C). In contrast, several mutations rapidly evolved in both replicate lagoons in the presence of danoprevir. Mutations at position D168 were predominant among these mutations. By 28 h, lagoon 1 with danoprevir contained 38.8% D168E, 8.3% D168Y, 2.1% D168A, and 1.1% D168V, while lagoon 2 with danoprevir contained 40.3% D168E and 10.7% D168Y (Figure 11C). Other genetic differences between the SPs of these two replicate populations, such as R130C (5.1% in lagoon 1, undetectable in lagoon 2) and T72I (10.8% in lagoon 2, undetectable in lagoon 1), suggest that cross-contamination did not lead to the observed protease variants in these experiments. These findings reveal that the presence of danoprevir caused the population of continuously evolving proteases to rapidly acquire mutations at D168.

To assay whether the PACE-evolved mutations confer danoprevir drug resistance in HCV protease, we purified recombinant HCV protease variants containing either of the two most highly enriched mutations, D168E

and D168Y. Each of these two mutations increase the IC50 of danoprevir by ~30-fold (wild-type HCV protease IC50

= 1.3 ± 0.1 nM; HCV protease D168E IC50 = 38.9 ± 2.4 nM; HCV protease D168Y IC50 = 34.4 ± 2.8 nM; IC50 ± standard deviation) (Figure 11D). Importantly, the D168E, D168A, and D168V mutations emerging from protease

PACE have been previously identified as common drug-resistance mutations in HCV isolated from patients treated with danoprevir31, 34.

To validate that protease PACE in the presence of a different HCV protease inhibitor can also result in the rapid evolution of drug-resistance mutations, we repeated PACE of HCV protease in the presence of asunaprevir, an

HCV protease inhibitor in phase III clinical trials, instead of danoprevir. We selected 75 µM asunaprevir as the final target concentration to use in the culture media based on dose-dependent gene expression assays (Figure 9). In order to allow diversity to emerge in the protease population, we first propagated HCV protease phage for 24 h without any inhibitor. Next, to ensure that the populations had sufficient time to evolve mutations that confer drug resistance, we propagated the populations for 24 h with 10 µM asunaprevir, Finally, we ramped up the asunaprevir concentration to 75 µM for 27 h in order to enrich those mutations that conferred robust drug resistance. HCV protease phage were also propagated for an identical amount of time without any added drug candidate for comparison. High-throughput DNA sequencing of phage populations at the end of the experiment revealed that mutations evolved at substantial levels in the asunaprevir-treated lagoons but not in the control samples (Figure

11C). In this experimental condition as well, mutations at position D168 were highly enriched. In the case of asunaprevir, however, the only substitution at this position to emerge at substantial levels from protease PACE was

D168Y, in contrast with the evolution of both D168E and D168Y during protease PACE with danoprevir.

In vitro assays of HCV proteases containing either mutation provides an explanation underlying the strong apparent preference of D168Y over D168E within asunaprevir-treated lagoons. D168Y increases the IC50 of asunaprevir by 30-fold, while D168E only increases the IC50 of asunaprevir by ~10-fold (wild-type HCV protease

IC50 = 6.9 ± 0.6 nM; HCV protease D168E IC50 = 53.5 ± 3.4 nM; HCV protease D168Y IC50 = 214.8 ± 31.9 nM;

IC50 ± standard deviation) (Figure 4e). Mutations at position D168 have been previously identified in replicon-based asunaprevir resistance experiments35 and the specific D168Y mutation has been observed to arise in hepatitis C patients treated with asunaprevir36. Collectively, these results establish that protease PACE in the presence of protease inhibitor drug candidates can very rapidly (1-3 days) reveal clinically relevant mutants that confer strong resistance to the drug candidates, without requiring extensive laboratory or clinical experiments.

Figure 9: HCV PA-RNAP response to protease inhibitors in E. coli cells. Host cells expressing the HCV PA-RNAP were incubated with HCV protease inhibitors danoprevir (a) or asunaprevir (b) for 90!min, followed by inoculation with HCV protease encoding phage. After 3!h, luminescence assays were used to quantify relative gene activation resulting from the PA-RNAP. Luminescence experiments were performed in triplicate with error bars depicting the standard deviation.

Tuesday, May 20, 2014 1:39 PM Page 1 of 1 pAB086k8 - MP CloDF13 pBAD dnaQ926 dam seqA cat (clone #5)

cat prom r oter inato erm al t on cti 00 re 50 di bi B n o clo /t at DF 4 c 1 1 3 P

r r n B 1

t s r

1 a e

0 n q

0 s A 0 0 c 0 A r

N i q 0 p e

a t 4 s i

o t i

n v Mutagenesis Plasmid (MP)

e l

t e

a r

r m S

5183 bp i n

o r

d a m

M o 3 d 0 0 i 0 dn Q926 0 fie 0 a 0 d 2 D m A u pB tS R 5 BS sd RBS

Figure 10: Vector map of mutagenesis plasmid (MP). This plasmid was used in HCV protease inhibitor resistance experiments. This plasmid provides arabinose-inducible expression of mutator genes dnaQ926, dam, and seqA.

Figure 11: Continuous evolution of drug resistance in HCV protease. (a,b) PACE condition timeline for evolution in the presence of danoprevir (a) or asunaprevir (b). The blue arrows indicate arabinose-induced enhanced mutagenesis, and the red arrow shows the timing and dosing of HCV protease inhibitors. (c) High-throughput sequencing data from phage populations in replicate lagoons (L1 and L2) subjected to danoprevir treatment at 28!h, asunaprevir treatment at 75!h, and no drug at 72!h. All mutations with frequencies >1% above the allele-specific error rate are shown. (d) In vitro analysis of danoprevir inhibition of mutant HCV proteases that evolved during PACE. (e) In vitro analysis of asunaprevir inhibition of mutant HCV proteases that evolved during PACE. For (d,e), evolved HCV protease variants were expressed and purified, then assayed using an internally quenched fluorescent-substrate (Anaspec). In vitro analyses were performed in triplicate with error bars calculated as the standard deviation.

Section 1.4: Continuous evolution of model proteases with altered specificity

Encouraged by these results, we next attempted to evolve proteases that cleave substrates containing single amino acid substitutions at critical positions in their substrate recognition sequences. In our first attempt at reprogramming substrate specificity we attempted to recapitulate the results of a previously published test case. A yeast-display screen was used to evolve TEV protease variants that cleave the single mutant substrate ENLYFE/S, a substrate known not to be accepted by wild-type TEV protease22. For these studies and all future studies involving specificity reprogramming, we used a single accessory plasmid (as opposed to the aforementioned CP+AP combination) to supply the protease-activated RNAP as well as gIII under control of the T7 promoter (Figure 12). In order to reprogram protease substrate specificity, we used a mixing strategy in which we performed 24 h of PACE on host cells expressing PA-RNAPs containing the wild-type substrate (ENLYFQ/S) to allow the protease population to diversify and increase in activity, followed by 24 h on a 1:1 mixture of two host strains each expressing either a PA-RNAP containing wild-type (ENLYFQ/S) or a PA-RNAP containing mutant (ENLYFE/S) substrate linkers, and finally 24-48 h of PACE on host cells expressing only the PA-RNAP containing the mutant substrate linker (Figure 13A). This protocol yielded TEV protease variants containing mutations at residues D148 and N177 both of which were reported as crucial determinants of TEV specificity at the modified P1 substrate position (Figure 13B). These TEV protease variants showed apparent activity on the ENLYFE/S substrate in luciferase assays and in vitro using purified proteins as shown in Figure 13C and Figure 13D.

In a second preliminary study, we sought to evolve HRV protease to cleave a target substrate (LEVLFQ/Y) that is known not to be cleaved by wild-type HRV protease37. Nevertheless, using the aforementioned mixing strategy PACE rapidly evolved HRV protease mutants that can cleave LEVLFQ/Y (wild-type substrate =

LEVLFQ/G) (Figure 14A). Virtually all evolved clones contained either mutation T143A or T143P (Figure 14B). In the NMR structure of HRV protease bound to substrate, residue T143 is position directly next to the substrate P1’ residue38. We hypothesize that mutations T143A or T143P alleviate the steric clash which the larger mutant substrate residue Tyr (as opposed to the wild-type Gly). The evolved HRV protease variants exhibit apparently proteolytic activity on both the wild-type and mutant target substrates in luciferase assays, whereas wild-type protease has no detectable activity on the mutant substrate (Figure 14C). These results collectively validate the ability of PACE to access protease mutants with the ability to cleave non-cognate substrates.

Tuesday, February 7, 2017 3:07 PM Page 1 of 1 MSP955

10000 or nat mi RepA er R CD t mp S nb A rr bla promo B ter p o S r 8 C P d rrn 1 s B 0 S 1 1 B e te R m rm zy in o a s t ly o e r r t e R i k B S in T S t L 7 T u 1 s p C d r r t e 8 e o V r k m m

E n i

o t

T i n

e t

e a

r A t

Accessory Plasmid I o I

/ r

( r T

L 7

T l

b y

7 i s d o

i z

N r y e m

A c

P t e

io d e

n p

a e n

l d

t e

e n t rm ) in ator B xA xlu

5000

Figure 12: Vector map of accessory plasmid (AP) used in protease specificity reprogramming experiments. A single accessory plasmid encodes constitutively expressed (proB promoter) PA-RNAP (Lysozyme-Linker- TEVcutsite-Linker-T7RNAP) as well as T7 promoter-controlled expression of gIII and the translationally coupled luciferase reporter (xluxAB). A lysozyme-dependent terminator (T1 terminator) is placed downstream of the T7 promoter to lower transcription of gIII-xluxAB in the absence of active protease. This plasmid encodes the low copy pSC101 origin of replication and is maintained with carbenicillin.

A B L1A E107D D148A N177K 0 24 48 72 96 h L1B N177K L1C D148A N177K L1D N177K AP L1E N177K ENLYFQS ENLYFES L1F Q73R S135F stop223 50% ENLYFQS + L1G S135F stop223 50% ENLYFES L1H S135F stop223 C D 60 GENLYFQSA ENLYF_/S Q E Q E Q E Protease NA NA WT WT N177K, N177K, GENLYFESA D148A, D148A, 50 E107D E107D

10 Fold Change in Luminescence in Fold Change 0 Wild Type N177K N177K, D148A TEV Protease Genotype

Figure 13: PACE of TEV proteases with altered P1 specificity. (A) PACE experimental condition timeline. Time in hours is depicted above the arrow; while protease substrate sequences used within the PA-RNAP are depicted beneath the arrow. The mixing period is indicated as a 50:50 mixture of two host strains expressing PA-RNAPs containing either the wild-type or the mutant substrate linker. (B) Genotypes after 96 h of PACE. Each row represents the genotype of TEV protease sequenced from a single clone of SP. Mutations at S135, D148 and N177 are strongly enriched. (C) Luminescence assays of PACE evolved SP clones. SP encoding TEV proteases bearing both the N177K and D148A mutations exhibit enhanced apparent activity on the single mutation ENLYFE/S substrate. (D) Protein cleavage assays of PACE-evolved TEV proteases. Approximately 1 µg of protease was incubated with 5 µg of a fusion protein construct in which MBP is linked to GST through a cleavable substrate linker containing the peptide ENLYFQ/S or ENLYFE/S. PACE-evolved TEV protease variant is capable of cleaving both the wild-type and mutant substrates.

A 0 24 48 72 96 h

AP LEVLFQG LEVLFQY 50% LEVLFQG + 50% LEVLFQY B C LEVLFQGP (WT) L1A T143P 81 71 LEVLFQYP L1B 61 L1C T143A 51 L1D T143P 41 L1E T143P 31 L1F T143A 21 L1G T143P 11 L1H T143P 1 Fold Change in Luminescence in Fold Change WT T143P T143A HRV Protease Genotype

Figure 14: PACE of HRV proteases with altered P1’ specificity (A) PACE experimental condition timeline. Time in hours is depicted above the arrow; while protease substrate sequences used within the PA-RNAP are depicted beneath the arrow. The mixing period is indicated as a 50:50 mixture of two host strains expressing PA-RNAPs containing either the wild-type or the mutant substrate linker. (B) Genotypes after 96 h of PACE. Each row represents the genotype of HRV protease sequenced from a single clone of SP. Mutations at T143 are strongly enriched. (C) Luminescence assays of PACE evolved SP clones. SP encoding HRV proteases bearing either the T143P and T143A mutations exhibit enhanced apparent activity on the single mutation LEVLFQ/Y substrate.

Section 1.5: Discussion

Previous efforts to use laboratory evolution to study HCV protease inhibitor resistance have relied on time- and labor-intensive approaches such as viral replication in mammalian cell culture or conventional protein evolution methods, which typically require months to complete39. By comparison, the continuous evolution of proteases can reveal key resistance mutations in as little as ~1 day of PACE. The speed of PACE and its ability to be multiplexed using many lagoons in parallel, each receiving a different drug candidate, and analyzed by high-throughput DNA sequencing of bar-coded lagoon samples, raises the possibility of screening future early-stage hit or lead compounds for their vulnerability to the evolution of drug resistance, before more resource-intensive optimization of in vivo properties or clinical trials take place. Rapid and cost-effective access to drug resistance susceptibility enabled by

PACE may enhance the more informed selection of more promising early-stage drug candidates for further development. This technique could also be applied to quickly screen a drug candidate across many distinct genotypic variants of a protease target (such as the six major HCV protease genotypes) to reveal each target variant’s potential to evolve mutations that abrogate the effectiveness of the drug candidate. As HCV patient isolates and replicon assays have already demonstrated differing drug resistance profiles among different HCV genotypes,40 this capability could also be used to rapidly identify patient-specific drug treatments that are more likely to offer long-term therapeutic effects on patients infected with specific HCV strains, even in the absence of previous data relating strain genotypes to drug effectiveness.

The development of protease PACE also expands the scope of PACE to evolve diverse biochemical activities. Prior to the publication of these results, PACE studies had only evolved RNA polymerases, which have activities that can be directly linked to changes in gIII expression. This study demonstrates how other types of enzymatic activities with no obvious direct connection to gene expression can nevertheless be evolved using PACE by establishing an indirect, but robust, linkage between the activity of interest and gIII expression. Since the publication of this work the Liu Lab has developed and/or published PACE selection schemes for evolving the activities of amino-acyl tRNA synthetases, high-affinity protein binding domains (Bt toxins, monobodies, etc), DNA binding domains (TALEs, Cas9, etc), CRISPR endoribonucleases, and enzymatic pathways that biosynthesize polyhydroxyalkanoates.

The protease PACE system provides a strong foundation for the continuous evolution of proteases with reprogrammed specificities. Previous work on reprogramming the DNA substrate selectivity of T7 RNAP

enzymes24, 28, 41, 42 demonstrated the ability of PACE to rapidly evolve enzymes that accept substrates very different from the native substrate. These reprogramming experiments relied on a “stepping-stone” strategy in which selection phage are transitioned between a series of intermediate substrates42, and are enhanced by the recent development of modulated selection stringency and negative selection during PACE28. In proof of principle experiments we have shown that these strategies coupled with protease PACE also enable the continuous evolution of model proteases with altered specificity for substrates containing a single amino acid substitution. These studies establish a foundation for subsequent research programs in which proteases are evolved with the tailor-made ability to selectively cleave proteins implicated in human diseases.

Section 1.6: Methods

PA-RNAP gene expression response in vivo. All plasmids were constructed by Gibson Assembly 2x Master Mix

(NEB); all PCR products were generated using Q5 Hot Start 2x Master Mix (NEB). E. coli strain S103028 were transformed by electroporation with three plasmids: (i) a complementary plasmid (CP) that constitutively expresses a PA-RNAP with one of the three protease cut sites (Figure 3), (ii) an accessory plasmid (AP, Figure 5) that encodes gIII-luciferase (translationally coupled) under control of the T7 promoter, and (iii) an arabinose-inducible expression plasmid for one of the three proteases (EP, Figure 4). The HRV protease gene was purchased as IDT gblocks and cloned into the expression vector. The MBP-TEV fusion protein was amplified by PCR from pRK79343. The MBP fusion was necessary for expression and solubility. We deployed a constitutively active HCV protease construct that includes the NS4a cofactor peptide44. Cells were grown in 2xYT media to saturation in the presence of antibiotics and 1 mM glucose, then inoculated into 1 mL fresh media containing 1 mM glucose and antibiotics in a 96 well culture plate. After 4.5 h, 150 µL of the cultures were transferred to a black-wall clear-bottom assay plate and

luciferase and OD600 measurements were taken using a Tecan Infinite Pro plate. The luminescence data was normalized to cell density by dividing by OD600.

Western blot of PA-RNAP sensor activation. E. coli cells transformed with an AP and a CP were grown to log phase, then infected with a 10-fold excess of protease-encoding phage. After 4.5 h, the cells were harvested by centrifugation at 5000 g for 10 min, and then resuspended in LDS Sample Buffer (Life Technologies). Samples were heated to 95 °C for 5 m and vortexed to shear genomic DNA. 4 µL of each sample was loaded onto a protein gel electrophoresis system (Bolt gel system, Life Technologies). The blot was performed using a PVDF membrane

(iBlot 2 system, Life Technologies). The membrane was blocked with 5% BSA TBST then incubated overnight with the primary antibody (5% BSA, TBST, 1:5000 anti-T7 RNAP mouse monoclonal, Novagen #70566). The membrane was washed three times, incubated with the secondary antibody (5% BSA, TBST, 1:5000 donkey anti- mouse, IR-dye conjugate, LI-COR #926-32212) for 60 min, washed three times, then visualized on a LI-COR

Odyssey at 800 nm. As seen in Figure 7, the PA-RNAP sensor is proteolyzed to a smaller band of anticipated molecular weight only in the presence of a cognate protease that can cleave the peptide sequence in each PA-RNAP linker.

Protease activity-dependent plaque assays. Protease phage were cloned using Gibson assembly and the aforementioned expression plasmids as templates. E. coli strain S1030 were transformed by electroporation with an

AP and a CP. After the transformed host cells were grown in 2xYT to OD600 ~ 1.0, 100 µL of cells were added to 50

µL of serial dilutions of protease-encoding phage. After one minute, 800 µL of top agar (7g / L agar in 2xYT) was added, mixed and transferred to quarter-plates containing bottom agar (15g / L agar in 2xYT). After overnight incubation at 37 °C, the plates were examined for plaques, which represent zones of slowed growth and diminished turbidity due to phage propagation.

PACE propagations and enrichment experiments. E. coli strain S1030 were transformed by electroporation with an AP, CP (one for each of the three PA-RNAPs), and a mutagenesis plasmid (MP, Figure 10) encoding arabinose- inducible expression of a dominant-negative mutator variant of dnaQ, wild-type dam, and wild-type seqA45-47.

Starter cultures were grown overnight in 2xYT supplemented with antibiotics and 1 mM glucose to prevent induction of mutagenesis prior to the PACE experiment. Host cell culture chemostats containing 80 mL of Davis rich media28 were inoculated with 2 mL of starter culture and grown at 37 °C with magnetic stir-bar agitation. At

-1 approximately OD600 1.0, fresh Davis rich media was pumped in at 80-100 mL h , with a chemostat waste needle set at 80 mL. This fixed dilution rate maintains the chemostat culture in late log phase growth, at which point it can be flowed into lagoons seeded with protease phage (initial titers were ~105 pfu mL-1). For these experiments, lagoon waste needles were set to maintain a lagoon volume of 15 mL, and host cell cultures were flowed in at 15-17 mL h-1.

Arabinose (10% w/v in water) was added directly to lagoons via syringe pump at 0.7 mL h-1 to induce mutagenesis.

Test propagations were conducted with cognate protease phage as well as non-cognate protease phage. Enrichment experiment lagoons were seeded with 1,000-fold excess of non-cognate protease phage. Lagoon samples were sterile-filtered at least every 24 h, and titers were assessed by plaque assay. Plaque assays were performed with

S1030 carrying pJC175e, a plasmid that supplies gIII under control of the phage-shock promoter28. Mock selections were monitored by PCR of the protease genes with forward primer (BCD582)

5'TGTTTTAGTGTATTCTTTCGCCTCTTTCGTT3' and reverse primer (BCD578)

5'CCCACAAGAATTGAGTTAAGCCCAATAATAAGAGC3' using filtered samples as templates. The distinct sizes of amplicons containing protease genes enabled evaluation of the relative abundance of cognate and non- cognate protease-encoding phage.

Inhibition of PA-RNAP response in host E. coli cells. Host cells were prepared by electroporation with an AP and the CP encoding the HCV-site PA-RNAP. We prepared 2xYT media with serial dilutions of inhibitors (danoprevir and asunaprevir, MedChemExpress) from stock solutions made in DMSO, and inoculated with a saturated starter culture of host cells. 150 µL cell cultures in a 96-well assay plate were incubated at 37 °C for 1.5 h to allow uptake of inhibitors, then infected with ~10 µL HCV protease phage (multiplicity of infection ~ 10). After 3 h of incubation at 37 °C, the luminescence of each culture was measured on a Tecan Infinite Pro plate reader and normalized to

OD600. In the absence of inhibitor, phage-encoded protease will activate the PA-RNAP leading to robust production of luciferase. Relative dose responses to inhibitors compared to control cells without drug were measured in triplicate.

Evolution of drug resistance in HCV protease using PACE. Host cells were the same as those in the HCV test propagation and enrichment experiments. Host cell culture chemostats containing 40 mL of Davis rich media were inoculated with 2 mL of starter culture and grown at 37 °C with magnetic stir-bar agitation. At approximately OD600

1.0, fresh Davis rich media was pumped in at 50 mL h-1, with a chemostat waste needle set at 40 mL. This adjustment was made to provide enough cell culture to feed two lagoons while also conserving media that contained small-molecule inhibitors. Lagoons were seeded with HCV protease phage and run in duplicate. Again, lagoon waste needles were set to maintain a lagoon volume of 15 mL, and host cell cultures were flowed in at 15-17 mL h-1.

Arabinose (10% w/v in water) was added directly to lagoons via syringe pump at 0.7 mL h-1 to induce mutagenesis.

After 6 h of propagation without any inhibitor, a filtered lagoon sample was taken, and danoprevir was added directly to the chemostat media at 20 µM with 2.5% DMSO to enhance solubility. A final time point was taken after 22 additional hours, and titers were measured by plaque assay on strain S1030 carrying pJC175e.

For the asunaprevir experiment, samples were taken every 12 h. After 24 h of propagation with no inhibitor, asunaprevir was added directly to the chemostat media at 10 µM with 2.5% DMSO. After an additional 24 h, asunaprevir dosage was increased to 75 µM and 5% DMSO. Titers were measured by plaque assay on strain

S1030 carrying pJC175e.

High-throughput sequencing of evolved populations. Strain S1030 carrying pJC175e were grown to saturation and used to inoculate fresh media. Host cells were infected with phage samples from the above PACE experiments and incubated for 5 h at 37 °C. DNA from infected cells was extracted using miniprep kits to yield concentrated template phage DNA (Epoch Life Science). PCR reactions were performed using Q5 Hot Start 2x Master Mix

(NEB) with a set of tiled primers. The PCR product from the first reaction was diluted ten-fold and 1 µL served as the template for the second PCR. The second PCR added Illumina adapters as well as barcodes; PCR products were purified from agarose gel (Qiagen) and quantified using the Quant-IT Picogreen assay (Invitrogen). Samples were normalized and pooled together to create a sequencing library at approximately 4 nM. The library was quantified by qPCR (KapaBiosystems) and processed by an Illumina MiSeq using the MiSeq Reagent Kit v3 and the 2x300 paired-end protocol. A single paired-end read of 600 bp is sufficient to cover the entire HCV protease gene.

Data was analyzed in MATLAB using the custom scripts. FASTQ files were automatically generated by the Illumina MiSeq. These files were already binned by sample barcodes and ready for transfer to a desktop computer. Each read was aligned to the wild-type HCV protease gene in the expected orientation using the Smith-

Waterman algorithm. Base calls with Q-scores below a threshold of 31 were converted to ambiguous bases, and the resulting ambiguous codons were turned into a series of three dashes for computationally efficient translation.

Ambiguous codons were translated into X’s, which were ignored when tabulating allele counts into a matrix. The script automatically cycled through each FASTQ file and saved the resulting allele count matrix in a separate subdirectory. At this stage, matrices for paired-end reads were added together and normalized to yield allele frequencies for each sample.

We relied on a wild-type control sample to assess PCR and sequencing bias. For this sample, we calculated the frequency of alleles that were not wild-type at each locus to yield the locus-specific error rate. We added

0.01(1%) to the locus-specific error rate to yield our variant call threshold. The allele frequency matrix for each sample was scanned for mutant alleles above the variant call threshold.

Purification and in vitro assays of evolved HCV variants. HCV protease variants were sub-cloned by Gibson assembly out of the phage genome and into the previously mentioned EP. EPs were transformed into NEB BL21

DE3 chemically competent cells. Starter cultures were grown to saturation, and 2 mL was used to inoculate 500 mL

LB. At OD600 = 0.6, cultures were transferred to 20 °C and induced with 0.5% arabinose for 6 h. Cells were harvested by centrifugation at 5,000 g for 10 m, and resuspended in lysis/bind buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, 5 mM imidazole). Cells were lysed by sonication for a total of 2 m, and then centrifuged for 20 m at 18,000 g to clarify the lysate. Supernatant was flowed through 0.2 mL His-pur nickel resin spin columns that were equilibrated with binding buffer (Pierce-Thermo). Resin was washed with 4 column volumes of wash buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, 20 mM imidazole). HCV protease was eluted in 4 column volumes of 50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, 200 mM imidazole. Samples were further purified by size exclusion chromatography on a SuperDex 75 10/300 GL column (GE Healthcare). Size exclusion was performed in 50 mM Tris-HCl pH 8.0, 100 mM NaCl, 10% glycerol, 1 mM DTT. Protein concentrations were determined by UV280 on a Nanodrop machine and calculated using an extinction coefficient of

19,000 cm-1 M-1 and a molecular weight of 23 kDa.

In vitro assays were performed using the commercial HCV RET Substrate 1 (Anaspec), an internally quenched probe that fluoresces upon proteolytic cleavage, according to the manufacturer’s instructions. Protease and inhibitors were incubated in assay buffer at room temperature for 5 m prior to addition of substrate. Fluorescence was measured every 30 s for 20 m by a Tecan Infinite Pro plate reader (excitation / emission = 355 nm / 495 nm).

Assays were performed at 30 °C with 40 nM protease, 7.5 µM substrate, and varying concentration of inhibitors in a final volume of 100 µL per well in black-wall clear-bottom assay plate. The assay buffer contained 50 mM Tris HCl pH 8.0, 100 mM NaCl, 20% glycerol, 5 mM DTT. Assays were performed in triplicate, and initial reaction velocities were calculated and normalized to controls without inhibitor. The data was fit to the Hill Equation using

Igor Pro with base and max parameters fixed at one and zero respectively. The resulting fits yielded IC50 values and standard deviations of the estimate.

Chapter 2: Continuous Evolution of a Model Protease to Cleave a Disease-

Associated Human Protein

Parts adapted from Packer, Rees, Liu, (2017).

Section 2.1: Choice of target substrate and protease

The biochemistry, substrate specificity48, and structure49 of TEV protease have been extensive studied, making it an ideal target for directed evolution. Our initial search for TEV protease target substrates was devised to find substrates similar in sequence to those that wild-type TEV protease could cleave. We populated a matrix with activity scores for each of the 20 possible amino acids at each of the seven positions within the TEV recognition motif. These scores roughly correspond to the published in vitro activity levels of TEV protease on mutant peptides48. This “specificity” matrix (Table 1) was then used to rank all possible heptapeptides within all human extracellular proteins. From this ranking, we manually curated a handful of disease-associated proteins in which the target peptide sequences were predicted to be solvent exposed based on their crystal structures (Table 2). Target peptides HFPYSQY from CCR5 (an HIV co-receptor whose degradation could provide immunity) and HSSYRQR from PDL1 (an immunosuppressive agent often overexpressed in cancers and whose degradation could prompt an anti-cancer immune response) both share a common His substitution at P6.

amino acid P6 P5 P4 P3 P2 P1 P1' A 0 0 0.1 0 0 0 0.5 R 0 0 0 0 0 0 0 N 0 0 0 0 0 0 1 D 0 0 0.1 0 0 0 0 C -1 -1 -1 -1 -1 -1 1 Q 0 0 0 0 0 1 0 E 1 0 0 0 0 0.5 0 G 0 0 0.1 0 0 0 1 H 0.5 0 0.5 0 0 0.5 0.25 I 0 0 1 0 0 0 1 L 0 0 1 0 0.1 0 0 K 0 0 0 0 0 0 0 M 0 0 0 0 0 0 0.25 F 0 0 0 0 1 0 0 P -0.5 0 -0.5 -0.5 -0.5 -0.5 -0.5 S 0 0 0.1 0 0.1 0 1 T 0 0 0.5 0 0.1 0 0 W 0 0 0 0 0 0 0 Y 0 0 0 1 0 0 0 V 0 0 0.5 0.25 0 0 0 Table 1: Target peptide scoring matrix based upon wild-type TEV specificity. We created a subjective rating matrix based upon our knowledge of wild-type TEV protease substrate specificity as assessed by published in vitro peptide cleavage assays. Key features include high ratings for consensus residues ENLYFQS as well as substitutions with high biochemical activity. We also introduced penalties for Cys residues due to disulfide formation in mammalian target proteins and for Pro due to its unique structural properties.

TEV site E X L Y F Q S CCR5 H F P Y S Q Y PDL1 H S S Y R Q R TNFa L G G V F Q L Black = Matches consensus Red = Requires evolution Table 2: Protease target substrates identified by specificity scoring. Target substrates were identified from the human extracellular proteome based upon ratings calculated using the above scoring matrix (Table 1). These three substrates were manually curated based upon the disease relevance of the target protein and the solvent-accessibility of target peptide.

For this reason we introduced the P6 His substitution first, which at the time had no precedent in the literature. Using the aforementioned mixing strategy we transitioned SP populations from the wild-type ENLYFQS substrate to the single mutant HNLYFQS substrate. In this experiment, we enriched for variants containing mutation N171D or N176T (Table 3). After our execution of these experiments, it was independently reported that mutations N171D and N176T promote TEV protease tolerance at the P6 position for uncharged residues such as threonine and proline18.

L1 A N176S B Q226Stop C S135F D E S135F F G H S135F L2 A D90G N185S B N176T C N171D N177Y D N171D N177Y F G D136E N176D Table 3: TEV protease mutations after PACE on HNLYFQS substrate. This table contains the clonal sequencing data from the end of PACE in which SP population was transitioned from hosts expressing ENLYFQS substrate to hosts expressing the HNLYFQS substrate. Each row corresponds to a single clonal sequence of TEV protease from the SP.

We then attempted mixing experiments to access double mutant substrates (HNPYFQS, HNSYFQS,

HNLYRQS, HNLYFQR) that would bring our protease populations closer to activity on the evolutionary target peptides HFPYSQY and HSSYRQR. These PACE attempts resulted in phage washout indicating that these evolutionary challenges were too difficult. Due to a concern that an unprecedented two substitutions within the substrate would have a synergistic effect on activity, we sought to decouple the mutations and access corresponding single mutant substrates (ENPYFQS, ENSYFQS, ENLYFQR). Mixing experiments attempting to evolve

recognition of these three single mutant substrates either ended in extremely low SP titers with no enriched mutations or complete washout. For the P4 Pro and P4 Ser substitutions we also attempted NNK site-saturation mutagenesis of TEV protease residues 139, 169, 178, 214, and 216. This library washed out of the lagoons and did not yield TEV variants with enhanced activity on either mutant substrate. A broad statement of evolutionary potential cannot be made, however it seems to be far more challenging to use protease PACE to alter TEV protease specificity at the P4 position than at the P1 and P6 positions (which we successfully achieved in Sections 1.4 and 2.1 respectively).

Using wild-type protease specificity to identify target substrates may fail to identify targets with high evolutionary potential. Although wild-type TEV protease is extremely specific for P6 Glu and P1 Gln, our studies and those of others show it is relatively easy to evolve TEV variants that recognize substrates mutated at these positions. Specificity scoring penalizes such substrates with directed evolution precedent18, 19, 21, 22, instead we created an “evolvability” scoring matrix (Table 4) that assigns an evolution difficulty score to each of the 20 possible amino acids at each of the seven positions recognized by TEV protease. We used this matrix to rank all possible heptapeptides within all human extracellular proteins and created a new short list of candidate target substrates (Table 5).

amino acid P6 P5 P4 P3 P2 P1 P1' A -1 0 1 0 2 -1 0 R 0 0 0 -1 0 0 0 N 0 0 0 0 0 0 0 D 0 0 1 0 0 0 0 C -1 -1 -1 -1 -1 -1 -1 Q 0 0 0 0 0 5 0 E 3 0 0 0 0 4 0 G -1 0 1 0 0 -1 0 H 3 0 1 0 0 4 0 I -1 0 1 0 1 -1 0 L -1 0 1 0 1 -1 0 K 0 0 0 1 0 0 0 M 0 0 0 0 0 0 0 F -1 0 0 0 3 -1 0 P -1 0 -1 -1 -1 -1 -1 S 0 0 1 0 1 0 0 T 0 0 1 0 1 0 0 W 0 0 0 0 0 -1 0 Y 0 0 0 4 0 0 0 V -1 0 1 1 1 -1 0 Table 4: Target peptide scoring matrix based upon known evolutionary potential. We created a subjective rating matrix based upon our knowledge of TEV protease substrate specificity and evolution of TEV proteases that accept single substrate changes. Key features include high ratings for consensus residues ENLYFQS as well as substitutions with known evolutionary solutions such as P6 His, P1 His, and P1 Glu. We also introduced penalties for Cys residues due to disulfide formation in mammalian target proteins and for Pro due to its unique structural properties.

TEV site E X L Y F Q S IL2RA H F V V G Q M IL23A H P L V G H M SELE H L V A I Q N SAA1/2 S D K Y F H A Black = Matches consensus Green = Previously evolved Yellow = Accepted Red = Requires evolution Table 5: Target substrates identified by TEV protease evolutionary potential. Target substrates were identified from the human extracellular proteome based upon ratings calculated using the above scoring matrix (Table 4). These four substrates were manually curated based upon the disease relevance of the target protein and the solvent-accessibility of target peptide.

The resulting candidate target sequences included HPLVGHM, a peptide found in human IL-23. IL-23 is a pro-inflammatory cytokine secreted by macrophages and dendritic cells in response to pathogens and tissue damage, ultimately promoting an innate immune response at the site of injury or infection. This immune response is mediated by IL-23-dependent stabilization of Th17 cells, a class of T helper cells that produce pro-inflammatory cytokines IL-17, IL-6, and TNFα50. Hyperactivity of this pathway can lead to a variety of autoimmune disorders including psoriasis and rheumatoid arthritis51. Monoclonal antibodies that neutralize IL-23 are FDA-approved drugs for the treatment of psoriasis and show promise in late-stage clinical trials for other autoimmune disorders52. These studies suggest that a protease that catalyzes IL-23 degradation may have anti-inflammatory activity.

The target peptide HPLVGHM differs from the TEV consensus substrate sequence, ENLYFQS, at six of seven positions. Two of these substitutions are predicted to not substantially impact TEV protease activity due to its low specificity at positions P5 and P1’, while the other four substitutions occur at positions that are known to be crucial specificity determinants of wild-type TEV protease (P6 Glu, P3 Tyr, P2 Phe, and P1 Gln). Indeed, substitution of TEV substrate P2 Phe or P1 Gln with the corresponding IL-23 substrate residue (P2 Gly or P1 His) has been shown to reduce TEV protease activity by more than an order of magnitude in each case48. Encouragingly, we had previously introduced the P6 His substitution, while other researchers have used site-saturation mutagenesis and an elegant yeast display screen to identify TEV mutants that accept P1 His instead of the P1 Gln22.

Section 2.2: Continuous Evolution of TEV Variants that Cleave the IL-23 Target Peptide

We verified that SP expressing wild-type TEV protease propagate robustly on host cells expressing a PA-

RNAP containing the TEV consensus substrate, ENLYFQS, in the linker connecting T7 RNAP and T7 lysozyme. In contrast, replacing the PA-RNAP linker with the target peptide HPLVGHM results in a failure of phage to propagate and rapid phage washout, consistent with the inability of wild-type TEV protease, or TEV variants containing a handful of immediately accessible mutations, to cleave the target IL-23 peptide.

Given the lack of previous success evolving proteases with specificity changes at more than one substrate residue, we anticipated that successful evolution of TEV protease variants that cleave the IL-23 target would require multiple evolutionary stepping-stones28, 42, 53, 54 to guide evolving gene populations through points in the fitness landscape that bring them successively closer to activity on the final target substrate. We designed three evolutionary trajectories such that substrate changes known to strongly disrupt the activity of wild-type TEV protease, including P6 His, P2 Gly, and P1 His,48 were introduced in the earliest stepping-stones (Figure 15A). We confronted these challenging substitutions first while the evolving protease populations had access to variants with wild-type-like levels of activity, reasoning that the likelihood of success was higher while proteases had sufficient activity to exchange for altered specificity without falling below a minimum activity threshold needed to survive selection. We introduced these challenging substrate changes one stepping-stone at a time to minimize the risk of washout and to illuminate at each stage how mutations within TEV protease altered substrate specificity.

We began all three evolutionary trajectories (Figure 15A) by introducing the P6 His substitution

(HNLYFQS) into the PA-RNAP and expressing a site-saturation mutagenesis library of TEV protease from the SP.

Using NNK codons, we randomized TEV protease residues N171, N176, and Y178, all of which are proximal to the

P6 substrate residue. This first PACE yielded variants with enhanced apparent activity on the HNLYFQS substrate

(Figure 16) and genotypes were highly enriched for D127A + S135F + N176I, or I138T + N171D + N176T (Table

6). Mutations N171D and N176T have been previously characterized as allowing P6 tolerance for uncharged residues such as threonine and proline18.

A Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Stage 6 Stage 7 Stage 8 1a Trajectory 1 NNK Residues: 146, 148, 167, 177 HNLVGHS IL-23 (38-66) 1b NNK Residues: NNK Residues: HNLYGHS proB proA 171, 176, 178 209, 211, 216, 218 50% proB+ HPLVGHM 50% proA HPLVGHM Q649S NNK Residues: 2a HPLVGHM HNLYFQS ENLYGQS 146, 148, 167, 177 HNLVGHS IL-23 (38-66) 2b HNLYGHS proB proA Trajectory 2 50% proB+ HPLVGHM 50% proA HPLVGHM Q649S 3a Trajectory 3 NNK Residues: HPLVGHM 209, 211, 216, 218 HNLVGHS IL-23 (38-66) 3b proB proA HNLYFQS HNLYFHS HNLYGHS 50% proB+ HPLVGHM 50% proA HPLVGHM Q649S IL-23 (38-66)IL-23 (38-66) 50% HNLYFQS + HPLVGHM Q649S Q649S 50% HNLYFHS proA

PACE stage Trajectory 1 1 D127A S135F N176I 2 D127A S135F N176I V209M W211I M218F 3 D127A S135F T146A D148P N176I N177R V209M W211I M218F 4 D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E 5 E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop 6 E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop 7 E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop 8 E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop

PACE stage Trajectory 2 1 I138T N171D N176T 2 D127A S135F N171D N176T V209M W211I M218F 3 D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E 4 D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E 5 D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E 6 R50G E107D D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E 7 T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F K229E 8 T17S H28L T30A N68D E107D D127A F132L S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F K229E

PACE stage Trajectory 3 1 D127A S135F N176I 2 D127A S135F D148A N176I 3 E107D D127A S135F T146A D148A N176I V209E W211L V216I M218W 4 E107D D127A S135F T146A D148A N176I V209F W211C M218L 5 H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop 6 H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop 7 H28Y T30A E107D D127A S135F T146A D148A S153N N176I V209F W211C K215E M218L Q226P P227A V228S K229stop 8 H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C K215E M218L Q226P P227A V228S K229stop Figure 15: Evolutionary trajectories and representative evolved TEV protease genotypes. (A) Across the eight stages of PACE along three diverging trajectories (shown in purple, blue, and orange), each arrow represents a PACE experiment with the corresponding substrate peptide and selection stringency parameters listed beneath the arrow. Increased selection stringency annotations are: Q649S (a T7 RNAP mutant with decreased transcriptional activity), proA (lower expression of substrate PA-RNAP), and IL-23 (38-66) (native IL-23 sequence in place of GGS linker). Numbers above the arrows denote TEV protease residues that were targeted in site- saturation mutagenesis libraries used to initiate that PACE experiment. In the first PACE experiment, wild-type TEV protease was mutagenized at the positions shown. All other libraries were generated using the protease genes emerging from the previous PACE stage as the PCR template. For PACE stages with no targeted mutagenesis, lagoons were inoculated with an aliquot of the phage population from the preceding experiment. (B) For each stage of each trajectory, a representative evolved protease clone emerging at the end of the stage (numbered in the first column) is shown. The mutations within each clone are listed as an individual row, illustrating the enrichment of new mutations during successive stages of PACE as well as genotypic differences between trajectories.

Trajectory A N171D N176T B R159I N171D N176T C S120N N176I E230A D D127A S135F N176I E D127A S135F N176I F D127A F132I S135F N176I G D127A S135F N176I L1 H D127A S135F N176I A S3R I138T N171D N176T B N12T I138T N171D N176T C I138T N171D N176T D I138T N171D N176T E I138T N171D N176T F I138T N171D N176T G I138T N171D N176T L2 1,2,3 H V36I I138T N171D N176T Table 6: PACE stage 1 mutations. Clonal sequencing data from the end of PACE stage 1 in Figure 15A (84 cumulative hours of evolution). Each row corresponds to a single clonal sequence of TEV protease from the SP. Lagoons 1 and 2 (L1 and L2) correspond to two separate biological replicate populations that underwent the same selection in PACE stage 1.

70000 ENLYFQS 60000 HNLYFQS 50000 40000 30000 20000

Luminescence/OD600 10000 0 WT N176I, R159T, R159T, I138T, Y178F N171D, N176T N171D, N176T N176T

Figure 16: Luciferase activity assay of clones from the middle of PACE stage 1 of trajectories 1, 2, and 3. TEV protease clones (corresponding genotypes are shown beneath the x-axis) after 36h of evolution on the first stepping-stone substrate show apparent proteolytic activity on both the wild-type substrate and the single mutant substrate HNLYFQS. Error bars represent the standard deviation of three technical replicates.

Next we pursued two parallel lines of PACE using either ENLYGQS (trajectories 1 and 2) or HNLYFHS (trajectory 3) as the second stepping-stone substrate. For trajectories 1 and 2, we diversified the population that emerged from PACE on the first stepping-stone (HNLYFQS) with NNK codons at TEV protease residues 209, 211, 216, and 218, which line the hydrophobic pocket that is occupied by the P2 Phe and performed PACE using host cells expressing ENLYGQS. The resulting population of TEV mutants is typified by the mutations N176I, V209M, W211I, M218F (Table 7), which confer apparent cleavage activity on both HNLYFQS and ENLYGQS substrates (Figure 17).

Trajectory A T70C T114P D127A S135F N176I V209A W211W V216I M218F B N176I V209M W211V V216V M218W C D127A S135F N176I V209M W211I V216V M218F E K65R D127A S135F N176I V209E W211L V216I M218W F Y11C D127A S135F N176I V209M W211I V216V M218W G D127V S135F N176I V209E W211L V216I M218W L6 1,2 H D127A S135F N176I V209M W211I V216V M218W Table 7: PACE stage 2 mutations. Clonal sequencing data from trajectories 1 and 2 after PACE stage 2 in Figure 15A (168 cumulative hours of evolution). 80000 ENLYFQS 70000 HNLYFQS 60000 ENLYGQS 50000 40000 30000 20000

Lumienscence/OD600 10000 0 uninfected WT L6A L6C L6G L6H

Figure 17: Luciferase activity assay after PACE stage 2 of trajectories 1 and 2. TEV protease clones from trajectories 1 and 2 (corresponding genotypes can be found in Table 7) after evolution on the second stepping-stone substrate, ENLYGQS, show activity on the wild-type substrate and both single mutant substrates (HNLYFQS/ENLYGQS). Error bars represent the standard deviation of three technical replicates.

In trajectory 3, we used a mixing strategy to access TEV proteases that could cleave the HNLYFHS stepping-stone double mutant substrate. Unlike PACE experiments initiated from a site saturation mutagenesis library, a mixing strategy relies on a transitional period of phage propagation on a mixture of two different host cell populations, one expressing an accepted substrate (HNLYFQS) and the other expressing the next stepping-stone substrate (HNLYFHS). Following this transitional period, the SP is propagated exclusively on hosts expressing the next stepping-stone substrate (HNLYFHS). The variants that emerged from this stage of trajectory 3 showed weak apparent activity on the double mutant substrate HNLYFHS (Figure 18), and only a single additional enriched mutation D148A (Table 8). Encouragingly, mutation at residue D148 has previously been reported to enable activity on ENLYFHS22.

Trajectory A D127A S135F N176I B D127A S135F N176I C S135F N176I D S120N S135F D148A N176I E D127A S135F N176I F D127A S135F N176I G D127A S135F D148A N176I R203Q L1 H S120N S135F D148A N176I A I138T N176T B I138T D148A N176T C I138T D148A N176T D I138T D148A N176T E I138T D148A N176T F I138T D148A N176T G I138T D148A N176T L2 3 H I138T D148A N176T Table 8: PACE stage 2 mutations. Clonal sequencing data from trajectory 3 after PACE stage 2 in Figure 15A (168 cumulative hours of evolution).

ENLYFQS 50000 HNLYFHS 40000 30000 20000 10000 0 Luminescence/OD600

Figure 18: Luciferase activity assay after PACE stage 2 of trajectory 3. TEV protease clones from trajectory 3 (corresponding genotypes can be found in Table 8) after evolution on the second stepping-stone substrate, HNLYFHS, show apparent activity on the wild-type substrate and the double mutant substrate, HNLYFHS. Error bars represent the standard deviation of three technical replicates.

Due to the low apparent activity of proteases emerging from this mixing experiment, we relied on the site- saturation mutagenesis strategy to evolve activity on the third stepping-stone, HNLYGHS. The TEV protease populations from trajectory 3 were randomized at sites implicated in P2 recognition (209, 211, 216, and 218), while for trajectories 1 and 2 the TEV protease population was randomized at sites 146, 148, 167, 177 as previously described for the reprogramming of TEV specificity at the P1 position22 (Figure 15A). The primers used to randomize TEV protease residues 167 and 177 must also encode the identity of intervening amino acids N171 and

N176. Although the population appeared to converge on N176I (Table 7), we reasoned that it was best to preserve genetic diversity at N176 by constructing one library with primers encoding N176I (trajectory 1) and another with

N171D + N176T (trajectory 2). Libraries constructed for all three trajectories were then subjected to PACE on host cells expressing the triple mutant substrate HNLYGHS. The variants emerging at this stage of trajectory 1 and 2 were enriched for mutations at residues 146, 148, and 177, consistent with acceptance of the newly introduced P1 substitution22. Similarly, clones from trajectory 3 exhibit mutations at residues 209, 211, and 218 that may promote acceptance of the newly added P2 Gly substitution. Regardless of trajectory, all clones emerging at this stage exhibit at least one mutation from each of three targeted mutagenesis libraries (Table 9), suggesting that they have evolved activity on the triple mutant substrate.

Given the known tolerance of TEV protease for amino acids at positions P5 and P1’, we speculated that proteases evolved to recognize the triple mutant substrate HNLYGHS might already exhibit activity on the final target substrate (HPLVGHM). Indeed, the populations arising from evolution on the triple mutant substrate successfully propagate in PACE on host cells producing the HPLVGHM substrate, and the resulting variants display weak apparent activity on the final target substrate (Figure 19). In order to evolve high levels of activity on the final target substrate from these weakly active mutants, we applied three strategies to increase selection stringency on all three trajectories: (1) express a lower concentration of the PA-RNAP substrate by using a weaker constitutive promoter (proA instead of proB)55; (2) substitute the flexible GGS linker that flanks our substrate with the native sequence from IL-23 (human IL-23 residues 38-66); and (3) introduce a mutation in the T7 RNA polymerase portion of the PA-RNAP that decreases transcriptional activity (Q649S)56. We confirmed that all three strategies indeed increased selection stringency (Figure 20).

200000 ENLYFQS 150000 HPLVGHM 100000 50000 0 Luminescence/OD600

Figure 19: Luciferase activity assay of clones after PACE stage 4. PACE evolved TEV SP clones (corresponding genotypes can be found in Table 10) from stage four of the evolutionary trajectories show proteolysis of HPLVGHM substrates within a protease-activated RNA polymerase as measured by downstream luciferase signal. These data indicate that the evolved enzymes were acquiring the desired phenotype, but higher selection stringency would be necessary in order to achieve catalytic activity similar to that of wild-type TEV protease. Error bars represent the standard deviation of three technical replicates.

35000 * p<0.005 uninfected ** p<0.0005 * unpaired t-test L2A 080215 30000 ** 25000 ** **

20000

15000

Luminescence/OD600 10000

5000

0 HPLVGHM IL23(38-66) HPLVGHM HPLVGHM IL23(38-66) proB proB proA proB proA Q649S Q649S

Figure 20: Validation of protease PACE stringency modulation. Using the highest activity TEV variant prior to stringency modulation in PACE, we performed protease-induced luminescence assays on APs that were expected to exert higher selection stringency. Prior to stringency modulation, the HPLVGHM proB AP exhibits robust protease-induced luminescence and 4.7-fold fold activation. When the flexible GGS-linkers in the PA-RNAP of the standard AP are replaced with the native sequence of IL-23 (amino acids 38-66) protease-induced luminescence is diminished (2.8-fold activation). When expression levels of the HPLVGHM PA-RNAP are lower due to a weaker constitutive promoter (proA instead of proB), we see much lower background and protease-induced luminescence as well as 3.3-fold activation. The introduction of deactivating mutation Q649S to the T7 RNAP portion of the PA-RNAP also causes a decrease in background and protease- induced luciferase signal (2.7-fold activation). When all three strategies are combined in a single AP, an even greater decrease in luciferase signal is observed (2.1-fold activation). Lower fold activation corresponds with higher selection stringency. Error bars represent the standard deviation of three technical replicates.

We first applied the lowered substrate concentration strategy using a mixing experiment to transition from proB to proA expression of the PA-RNAP; this experiment yielded modest changes in genotypes. Exploiting the ease of performing PACE on multiple lagoons in parallel, we implemented the other two strategies simultaneously on all three trajectories. The resulting six populations (trajectories 1a, 1b, 2a, 2b, 3a, and 3b; see Figure 15A) were carried forward into PACE on hosts expressing a PA-RNAP with both the IL-23 (38-66) linker and the attenuated

T7 RNAP mutant Q649S. In the final stage of PACE for all six populations, we used a proA promoter to generate less of the PA-RNAP containing the IL-23 (38-66) linker and the Q649S mutation. This series of stringency

modulation experiments produced variants with higher levels of apparent activity on the final HPLVGHM substrate

(Figure 21).

300000 ENLYFQS 250000 HPLVGHM

200000

150000

100000

Lumienscence/OD600 50000

Figure 21: Luciferase activity assay of clones after PACE stage 8. After multiple PACE stages with increasing positive selection stringency, many TEV protease variants (corresponding genotypes can be found in Table 14) exhibit markedly stronger apparent activity on the HPLVGHM substrate when compared with clones from previous PACE experiments such as those seen in Figure 19. Error bars represent the standard deviation of three technical replicates.

Trajectory A D127A S135F T146A D148P N176I N177M V209M W211I M218F B E106G D127A S135F T146A D148P N176I N177R V209M W211I M218F C D127A S135F T146A D148P N176I N177R V209M W211I M218F D D127A S135F T146R D148C H167P N176I N177G V209M W211I M218F E D127A S135F T146A D148P N176I N177W V209M W211I M218F F D127A S135F T146C D148P N176I N177M V209M W211I M218F G D127A S135F T146C D148P N177M S200G V209M W211I M218F L1 1 H D127A S135F T146C D148P N176I N177M V209M W211I M218F A D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E B D127A S135F T146C D148P N171D N176T N177M V209E W211L V216I M218W E223stop C D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E D D127A S135F T146C D148P N171D N176T N177M V209E W211L V216I M218W E V63I D127A F132S S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E F D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E G D127A S135F T146C D148P N171D N176T N177M V209E W211L V21 6I M218W E223stop L2 2 H D127A S135F T146C D148P N171D N176T N177M V209M W211I M218F K229E C E107D D127A S135F T146A D148A N176I V209S W211I M218W D E107D D127A S135F D148A N176I V209E W211L V216I M218W Q226stop E E107D D127A S135F D148A N176I V209S W211I M218W

52 F E107D D127A S135F T146A D148A N176I V209E W211L V216I M218W

G G32R E107D D127A S135F T146A D148A N176I V209E W211L V216I M218W Q226stop L3 3 H E107D D127A S135F D148A N176I V209F W211C M218L

Table 9: PACE stage 3 mutations. Clonal sequencing data from trajectory 1, 2, and 3 after PACE stage 3 in Figure 15A (264 cumulative hours of evolution).

Trajectory A D127A S135F T146A D148P N176I N177W V209M W211I M218F K229E Q233stop B D127A S135F T146A D148P N176I N177W V209M W211I M218F K229E C P8L D127A S135F T146A D148P N176I N177W V209M W211I M218F K229E D D127A S135F T146A D148P N176I N177W V209M W211I M218F K229E Q233stop E D127A S135F T146A D148P N176I N177W V209M W211I M218F K229E Q233stop F D127A S135F T146A D148P N176I N177W V209M W211I M218F K229E Q233stop G D127A S135F T146A D148P R159K N176I N177R V209M W211I M218F K229E L1 1a H D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E A D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E B D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E C D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E D D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E E D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E F D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E G D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F Q226stop K229E L2 2a H D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E A N12H E107D D127A S135F T146A D148A N176I V209F W211C M218L Q226stop B E107D D127A S135F T146A D148A N176I V209F W211C M218L P227S D E107D D127A S135F T146A D148A N176I V209F W211C M218L E K67N E107D D127A S135F T146A D148A S152N N176I V209F W211C M218L F E107D D127A S135F T146A D148A N176I V209F W211C M218L G E107D D127A S135F T146A D148A N176I V209F W211C M218L L3 3a H E107D D127A S135F T146A D148A N176I V209F W211C M218L A E106G D127A S135F T146S D148P N176I N177F V209M W211I M218F K229E B T118S D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E C D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E 53 D T118S D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E

E D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E F D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E G N12T D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E L4 1b H R80G D127A S135F T146A D148P N176I N177R V209M W211I M218F K229E A D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E B D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E D D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E E K6E D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E F T17A D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E L5 2b G N68D D127A S135F T146S D148P N171D N176T N177M V209M W211I M218F K229E A E106G E107D D127A S135F T146A D148A N176I V209F W211C M218L B E107D D127A S135F T146A D148A N176I M218F Q226stop C T30A R50K E107D D127A S135F T146A D148A N176I V209F W211C M218L D E107D D127A S135F T146A D148A N176I M218F Q226stop E E107D D127A S135F T146A D148A N176I M218F Q226stop F E107D D127A S135F T146A D148A N176I M218F Q226stop G R9C E107D D127A S135F T146A D148A N176I V209F W211C M218L L6 3b H E106G E107D D127A S135F T146A D148A N176I

Table 10: PACE stage 4 mutations. Clonal sequencing data from trajectory 1, 2, and 3 after PACE stage 4 in Figure 15A (336 cumulative hours of evolution).

Trajectory A E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop B E106G D127A S135F T146A D148P S153N S170A N176I N177R K184T V209M W211I M218F K229E C E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop D E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop E E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop F E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop G E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop L1 1 H D127A S135F T146A D148P S153N S170A N176I N177R V209M W211I M218F K229E A D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E B H28Y D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E C H28L D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E D K6E D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E E D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E F H28L D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E G K6E D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E L2 2 H D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E A E107D D127A S135F T146A D148A N176I V209F W211C M218L Q226S P227A V228S K229stop B H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop D H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop E H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop F E107D D127A S135F T146A D148A N176I V209F W211C M218L Q226S P227A V228S K229stop L3 3 H H28Y E107D D127A S135F T146A D148A N176I V209F W211C M218L Q226S P227A V228S K229stop

54 Table 11: PACE stage 5 mutations. Clonal sequencing data from trajectory 1, 2, and 3 after PACE stage 5 in Figure 15A (456 cumulative hours of evolution).

Trajectory A E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop B E2K P39S D127A S135F T146C D148P F162S S170A N176I N177S R203Q V209M W211I M218F Q226stop C E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop D E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop E E2K H28Y N68D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop F E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop G E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F E223G Q226stop L1 1a H E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop A E107D T114A D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E B H28L T30A E106D D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E

D K6E H28L D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F E223stop E K6E D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E F K6E D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E G K6E T30A D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E L2 2a H K6E D127A S135F T146S D148P F162S N171D N176T N177M V209M W211I M218F K229E

B H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop C H28Y K89E E107D M121T D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop D H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop E H28Y F40L E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop

G H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop L3 3a H H28Y E107D I109V D127A S135F T146A D148A S153N N171K N176I V209F W211C M218L Q226S P227A V228S K229stop A E2K D10N D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop B E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop

D E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop E K99Q E106G M124I D127A S135F T146A D148P S153N S170A N176I N177R V209M W211I M218F K229E F E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop G E2K H61R D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop L4 1b A H28L T114A D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F K229E B F5L T17A R50G E107D D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E

55 C R50G D127A C130R S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K220Q K229E D R50G I109V D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F E223stop K229E

E P13S R50G D127A F132V S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E L234R F R50G I109V D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E G D127A S135F T146S D148P F162S S170A N171D N176T N177M V209M W211I M218F K229E A231E L5 2b A H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop B H28Y N68D E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop C H28Y E107D M121T D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop D H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop E H28Y F40L E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop F T17A H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop G H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop L6 3b H H28Y E107D I109V D127A F132V S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop

Table 12: PACE stage 6 mutations. Clonal sequencing data from trajectory 1, 2, and 3 after PACE stage 6 in Figure 15A (528 cumulative hours of evolution).

Trajectory A E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop C E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop D T17A E24K H28L K97N D127P S135F T146A D148P S153N S170A N176I N177R V209M W211I M218F K229E E E2K D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop G T17A H28L K97N D127P S135F T146A D148P S153N S170A N176I N177R V209M W211I M218F K229E L1 1a H T17A H28L K97N D127P S135F T146A D148P S153N S170A N176I N177R V209M W211I M218F K229E A T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F K229E B T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F K229E C T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F F225L K229E D T30A T71P K89E D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F K229E E T30A T69P M124T D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211I M218F K229E F T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F K229E L2 2a H T30A D127A S135F T146P D148P Q150P F162A S170A N171D N176T N177M V209M W211V K215E M218F K229E A H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L B H28Y T30A E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226P P227A V228S K229stop C H28Y T30A E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226P P227A V228S K229stop D H28Y E107D S120N D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop E H28Y T30A E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L F H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop G H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop L3 3a H I18V H28Y E107D D127A S135F T146A D148A N176I V209F W211C M218L Q226P P227A V228S K229stop A E2K I109V D127A S135F T146C D148P S153N S170A N176I N177S R203Q V209M W211I M218F Q226stop B E2K E106D D127A S135F T146C D148P S153N S170A N176I N177S K184E R203Q V209M W211I M218F Q226stop C E2K T70P E107D D127A S135F T146C D148P S153N S170A N176I N177S R203Q V209M W211I M218F Q226stop D E2K T30A E106G D127A S135F T146A D148P S153N S170A N176I N177R V209M W211V M218F K229E T232A E E2K E107D L111F M124I D127A S135F T146C D148P S153N S170A N176I N177S R203Q V209M W211I M218F Q226stop F E2K E107D D127A S135F T146C D148P S153N S170A N176I N177S R203Q V209M W211I M218F Q226stop L4 1b G E2K E107D T114A D127A S135F T146C D148P S153N S170A N176I N177S R203Q V209M W211I M218F Q226stop A T17A H28L D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F F225S K229E B H28L H61Q D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F Q226S P227A V228S K229stop C T17A H28L D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F Q226S P227A V228S K229stop E T17A H28L D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F Q226S P227A V228S K229stop L234R F T17A H28L N68D D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F Q226S P227A V228S K229stop G E2K T17A H28L D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F Q226S P227A V228S K229stop A231E L5 2b H H28L N68D D127A S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F Q226S P227stop A H28Y V66G H75L E107D D127A S135F T146A D148A N176I V209F W211C M218L Q226S P227A V228S K229stop C H28Y E107D M121T D127A S135F T146A D148A N176I V209F W211C M218L Q226S P227A V228S K229stop D H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop

56 E H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop F H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop

L6 3b G H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop

Table 13: PACE stage 7 mutations. Clonal sequencing data from trajectory 1, 2, and 3 after PACE stage 7 in Figure 15A (600 cumulative hours of evolution).

Trajectory A E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop B E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop C E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop D E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop E E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop F E2K E107D D127A S135F T146C D148P S153N S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop G E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop L1 1a H E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop B H28L E107D D127A T128P S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F K229E C R9C H28L T30A N68D E107D D127A F132L S135F T146S D148P S153N F162S S170A N171D N176T N177M V182G V209M W211I M218F K229E F T17S H28L T30A N68D E107D D127A F132L S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F K229E L2 2a H H28L T30A N68D E107D D127A F132L S135F T146S D148P S153N F162S S170A N171D N176T N177M V209M W211I M218F K229E A H28Y Q104R E107D D127A F132S S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop B H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop C E24V H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop D H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop E E2G H28Y E107D D127A S135F T146A D148A N176I N185D V209F W211C M218L Q226S P227A V228S K229stop F H28Y N68D E107D D127A S135F T146A D148A N176I V209F W211C M218L Q226S P227A V228S K229stop G E24K H28Y N68D E107D D127A T128A S135F T146A D148A N176I V209F W211C M218L Q226S P227A V228S K229stop L3 3a H H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop A E2K D26Y E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218L Q226stop B E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop C E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop D E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop E E2K R80G D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I K215E M218F Q226stop F E2K T17A E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop G E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop L4 1b H E2K E107D D127A S135F T146C D148P S170A N176I N177S R203Q V209M W211I M218F Q226stop A F5L T30A Q73H T118A D127A C130R S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F P221T E222stop B T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F P221T E222stop C T30A V125G D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F P221T E222stop D T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F P221T E222stop E T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F K229E F T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F P221T E222stop G T30A D127A S135F T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F P221T E222stop L5 2b H T30A Q73H T118A D127A C130R S135F Q145P T146S D148P F162A S170A N171D N176T N177M V209M W211V K215E M218F P221T E222stop A N12T H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop B H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop C D10G H28Y K67E E107D D127A S135F T146A D148A S153N N176I Q193P V209F W211C M218L Q226S P227A V228S K229stop D H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop E H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop

57 F H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L Q226S P227A V228S K229stop L6 3b G H28Y E107D D127A S135F T146A D148A S153N N176I V209F W211C M218L

Table 14: PACE stage 8 mutations. Clonal sequencing data from trajectory 1, 2, and 3 after PACE stage 8 in Figure 15A (672 cumulative hours of evolution)

Section 2.3: Characterization of Evolved TEV Protease Variants

Mutations that arise early in long evolutionary trajectories can create a cascade of contingencies because subsequent mutations must be compatible with the preexisting genetic context, a phenomenon known as epistasis.

Genotypes suggest that epistasis strongly shaped the outcomes of trajectories 1 and 2, which were dominated by

N176I vs. N171D + N176T, respectively, prior to the third stage of PACE. During subsequent evolution, the amino acid identity at amino acid 176 appears to have dictated the optimal identity of residue 177, such that the combinations N176I + N177S, or N176T + N177M, predominate trajectories 1 and 2 respectively. Swapping the identity of N177 between clones from trajectories 1 and 2 results in a substantial loss of activity (Figure 22), further consistent with epistasis at this position. It is likely that these genetic differences between trajectory 1 and 2 also later led to the enrichment of distinct mutations outside of the substrate-binding site (Table 6-Table 14). For example, mutations at position 203 persisted only in trajectory 1, while mutations at positions 28, 30, 68, 132, and

162 were only abundant in trajectory 2.

160000 * * p<0.0001 140000 unpaired t-test * 120000 100000 80000 60000 40000

Luminescence/OD600 20000 0 L1F L1F N177M L2F L2F N177S Uninfected

Figure 22: Epistatic interactions with TEV protease residue N177. PACE evolved clones L1F and L2F, from PACE stage 8 of trajectories 1 and 2 respectively, exhibit robust apparent activity on the HPLVGHM substrate. When the identity of residue N177 is swapped between these clones, as in variants L1F N177M and L2F N177S, we see a significant decrease in apparent activity, suggesting that the optimal substitution for N177 depends upon the identities of other mutations within TEV protease. Error bars represent the standard deviation of three technical replicates.

Unsurprisingly, we observed a dramatically different outcome in the third trajectory, which not only experienced a different schedule of stepping-stone substrates but was also subjected to a mixing experiment instead of NNK mutagenesis at residues 146, 148, 167 and 177. Our data are consistent with a model in which a lack of diversification at these critical residues traps trajectory 3 in a local fitness maximum, evidenced by weak apparent activity on the final target substrate (Figure 19-Figure 21) and few genotypic changes after the fifth stage of

trajectory 3 (Figure 22B). Consequently, while all six populations yielded TEV variants with apparent activity on the final target, the TEV protease variants that exhibited the highest apparent activity on the final target were all derived from trajectories 1 and 2.

Three representative proteases from the end of trajectories 1 and 2 were purified and assayed in vitro for their ability to cleave a model protein substrate in which MBP and GST were fused through a linker containing the final HPLVGHM substrate sequence (Figure 23). All three evolved proteases cleaved the model substrate. We selected the most active clone (TEV L2F from trajectory 2, containing 20 non-silent mutations, Table 14) for detailed characterization. We assayed the kinetic parameters of this mutant enzyme on wild-type (ENLYFQS) and target (HPLVGHM) substrate peptides using a previously described HPLC method22. Unlike the wild-type enzyme, which exhibits no detectable activity on the HPLVGHM peptide, the L2F variant processes this substrate with approximately 15% of the catalytic efficiency (kcat/KM) with which TEV protease cleaves its native substrate (Table

15 and Figure 24). Compared to wild-type TEV, evolved TEV L2F appears to maintain nearly identical kinetics on the canonical ENLYFQS substrate, while experiencing only a modest 5-fold increased Km on the target substrate

HPLVGHM. These results collectively indicate that PACE generated a highly evolved mutant protease that cleaves a target substrate containing mutations at six positions with only modestly lower efficiency than wild-type TEV protease cleaves its consensus substrate.

L1F L2F L5B

MBP-GST

MBP

TEV GST

Figure 23: Protein cleavage assay to identify the most active clone. TEV protease variants from the final PACE experiment were overexpressed and purified. Approximately 1 µg of protease was incubated for 3 h at 30°C with 5 µg of a fusion protein construct in which MBP is linked to GST through a cleavable substrate linker containing the peptide HPLVGHM. Note that TEV protease variants L1F and L5B encode premature stop codons leading to products with approximately the same molecular weight as GST. Consequently, the intensity of the MBP product band best reflects reaction efficiency, leading to the conclusion that TEV L2F exhibits the highest catalytic activity among the clones tested.

-1 -1 Substrate Protease kcat (s ) Km (µM) kcat/Km (s*µM) ENLYFQS Wild-Type 0.23±0.011 170±22 1.3x10-3 HPLVGHM Wild-Type N.D. N.D. N.D. ENLYFQS TEV L2F 0.27±0.019 150±28 1.7x10-3 HPLVGHM TEV L2F 0.19±0.010 920±110 2.0x10-4 Table 15: Kinetic parameters of wild-type and evolved TEV proteases. N.D. = not detected (no product formation observed after 30 min of incubation of 1 µM protease and 2 mM substrate with a limit of detection of 1 nM product).

TEV Wild-Type A Product Substrate B 250 0.25

200 0.20

150 -1 0.15

ENLYFQS 0.10

100 Rate, s HPLVGHM kcat 0.2338 50 0.05 Km 173.1 0.00 Absorbance355nm (mAu) 0 15 17 19 21 23 25 27 0 200 400 600 800 1000 -50 [ENLYFQS Substrate], µM Time (minutes)

C TEV L2F D TEV L2F 0.15 0.25

0.20 0.10 -1 -1 0.15

0.10 Rate, s 0.05 Rate, s kcat 0.1865 kcat 0.266 0.05 Km 925.7 Km 152.4 0.00 0.00 0 500 1000 1500 2000 2500 0 200 400 600 [HPLVGHM Substrate], µM [ENLYFQS Substrate], µM

Figure 24: HPLC assay of TEV protease kinetics. (A) Synthetic peptide standards. TEV protease substrate peptides and the corresponding product peptides in a 1:1 mixture are separable by reverse-phase liquid chromatography. (B) Wild-type TEV protease (0.1 µM) was incubated for 10 min at 30 °C with ENLYFQS substrate concentration ranging from 50 to 800 µM. (C) TEV L2F protease (0.1 µM) was incubated for 10 min at 30 °C with HPLVGHM substrate concentration ranging from 50 to 2000 µM. (D) TEV L2F protease (0.05 µM) was incubated for 10 min at 30 °C with ENLYFQS substrate concentration ranging from 50 to 500 µM. Data was fit to a Michaelis-Menten enzyme kinetics model with error bars representing the standard deviation of three technical replicates.

Section 2.4: Substrate Specificity Profiling of an Evolved TEV Protease

Proteolysis assays on individual substrates reveal that evolved TEV protease L2F maintains the ability to detectably cleave starting and intermediate substrates while acquiring activity on the final IL-23 target (Figure 25).

A more comprehensive understanding of the substrate specificity of this evolved enzyme requires an unbiased protease specificity profile generated from a large number of substrate variants. To obtain such a profile, we applied a previously reported phage substrate display method (Figure 26A)57-59. M13 bacteriophage encoding pIII fused to a

FLAG-tag through a library of substrate linkers were immobilized on anti-FLAG magnetic beads. When incubated with a protease of interest, phage encoding cleaved substrates are liberated from the solid support, while phage encoding the intact substrates remain immobilized and are eluted with excess FLAG peptide. The abundance of each substrate in the cleaved versus eluted populations was measured by high-throughput DNA sequencing, yielding enrichment values (Table 16) and sequence logos (Figure 26B-E) that convey protease substrate specificity across all possible amino acids60.

ENLYFQSHNLYFQSENLYFHSENLYGQSHNLYFHSHNLYGHSHPLVGHMENLYFQSHNLYFQSENLYFHSENLYGQSHNLYFHSHNLYGHSHPLVGHM

MBP-GST

MBP

TEV GST

Wild Type TEV Protease TEV L2F Protease T17S, H28L, T30A, N68D, E107D, D127A, F132L, S135F, T146S, D148P, S153N, F162S, S170A, N171D, N176T, N177M, V209M, W211I, M218F, K229E Figure 25: Evolved TEV protease cleaves wild-type, intermediate, and target substrates. In a manner analogous to that described above in Figure 23, we assayed TEV proteases on a panel of substrate sequences. Approximately 1 µg of protease was incubated for 3 h at 30°C with 5 µg of a fusion protein construct in which MBP is linked to GST through a cleavable substrate linker containing the indicated amino acid sequence. WT TEV efficiently cleaves wild-type substrate, and to a much lesser degree processes single mutant substrates (HNLYFQS, ENLYFHS, ENLYGQS). Evolved TEV protease clone L2F yields a visible product band for the target substrate HPLVGHM. This evolved protease has also maintained activity on wild-type, single, double, and triple mutant substrates that were used as evolutionary stepping-stones in PACE.

We applied this specificity profiling technique to seven separate libraries, each containing a single randomized position within the canonical ENLYFQS substrate. These libraries are inherently biased by the identities of the residues that are held constant, but because of their small theoretical diversity, they are easy to construct and a single round of selection yielded robust enrichment values. We validated this method by enrichment of the consensus motif

EXLYFQS (where X = any amino acid) for wild-type TEV protease (Figure 26B). We also applied this substrate specificity profiling method to more complex libraries containing sets of three consecutive randomized amino acids within either the ENLYFQS or HPLVGHM substrate. The resulting specificity profiles from these larger libraries

(Figure 27) did not substantially differ from the results of the single-site libraries. In addition, the identity of the constant residues (ENLYFQS vs. HPLVGHM) had only modest impact on the resulting specificity profiles of TEV

L2F, although the P1 specificity of TEV L2F is more pronounced for P1 His, the target residue, and P6 Glu, the wild-type residue, in the context of HPLVGHM libraries (Figure 27 and corresponding enrichment values in Table

17).

When we compare the specificity profile of evolved TEV L2F (Figure 26C) to that of wild-type TEV, a number of differences are apparent: TEV L2F shows a broadening of specificity at P6, a shifting of P3 specificity towards aliphatic residues Ile and Val, a shifting of P1 specificity to include His, and a shifting of P1’ specificity towards aliphatic amino acids Ala, Ile, and Met. These changes are largely consistent with evolutionary pressure to cleave the target substrate HPLVGHM. A notable absence of altered specificity at the P2 Gly position suggests that affinity for this substitution may offer the largest remaining potential gains in target substrate cleavage efficiency.

Although the evolved L2F protease recognizes a shortened motif due to loss of P6 specificity, it retains the ability to reject the substantial majority of amino acids at each of the five others positions used by TEV. The overall specificity of the evolved L2F protease lies well within the range of specificities exhibited by natural proteases such as caspases, granzymes, clotting factors, and MMPs, which typically specify strongly only one or two positions and accept mixtures of several to many amino acids at other positions, yet retain sufficient overall specificity to mediate physiological signaling roles61-64.

Figure 26: Protease specificity profiling. (A) Overview of phage substrate display. M13 bacteriophage libraries contain pIII fused to a FLAG-tag through a randomized protease substrate linker. These substrate phage are bound to anti-FLAG magnetic beads and treated with a protease to release phage that encode substrates that can be cleaved by the protease. The remaining intact substrate phage are eluted with excess FLAG peptide. The abundance of all substrate sequences within the cleaved and eluted samples is measured by high-throughput sequencing. (B-E) For all assayed proteases, phage substrate display was separately performed on seven libraries, each with a different single randomized position within the ENLYFQS motif. The resulting enrichment values are displayed as sequence logos, with enrichment values above zero indicating protease acceptance, and values below zero indicating rejection. (B) wild-type TEV protease exhibits strong enrichment for the consensus motif EXLYFQS. (C) Evolved TEV L2F has broadened specificity at P6 and shifted specificity at P3, P1, and P1’ in accordance with the HPLVGHM target substrate. (D) Mutations I138T, N171D, N176T are sufficient to broaden P6 specificity. (E) Mutations T146S, D148P, S153N, S170A, N177M shift specificity at both P1 and P3.

Table 16: Phage display enrichment values from selections on single site libraries. Each sub-table within the larger table represents the amino acid enrichment values generated for the given genotype of TEV protease. Each row contains enrichment values from a selection performed on the library in which the corresponding position within the ENLYFQS motif was randomized. The enrichment value for each amino acid identity at a given position was calculated as frequencycleaved/frequencyelution-1. The cells are shaded on a linear scale from red to blue; this color scale is normalized for each sub-table with the lowest number in the sub-table being the darkest red and the highest number in the sub-table being darkest blue. Wild-type TEV A C D E F G H I K L M N P Q R S T V W Y P6 -0.19 0.08 1.03 5.69 0.99 -0.16 0.03 0.41 -0.20 0.42 0.31 -0.10 0.04 0.23 -0.14 -0.05 -0.01 0.26 0.37 -0.01 P5 0.30 -0.25 0.45 -0.07 0.25 -0.46 0.15 -0.19 -0.06 0.21 0.10 -0.51 -0.24 0.02 0.02 0.10 -0.04 0.20 0.14 0.24 P4 -0.34 -0.29 -0.31 -0.15 -0.33 -0.31 -0.40 5.16 -0.29 3.16 0.68 -0.32 -0.37 -0.39 -0.33 -0.35 -0.31 0.39 -0.21 -0.39 P3 -0.50 -0.52 -0.09 -0.54 0.84 0.30 -0.48 -0.11 -0.50 -0.37 -0.54 -0.45 -0.39 -0.52 -0.29 -0.56 -0.58 0.01 -0.26 4.08 P2 -0.19 -0.74 -0.48 -0.66 3.81 -0.39 -0.70 3.79 -0.69 0.48 0.85 -0.59 -0.03 -0.72 -0.62 -0.38 0.19 4.36 0.92 -0.72 P1 -0.22 -0.06 -0.14 -0.12 -0.16 -0.15 -0.08 -0.24 -0.24 -0.18 1.29 -0.02 -0.22 5.49 -0.27 -0.24 -0.17 -0.28 -0.16 -0.26 P1' 2.05 -0.50 -0.03 -0.48 1.60 1.96 1.22 -0.39 -0.58 -0.44 1.38 0.07 -0.68 -0.33 -0.62 2.77 -0.33 -0.60 2.24 2.60 TEV L2F T17S, H28L, T30A, N68D, E107D, D127A, F132L, S135F, T146S, D148P, S153N, F162S, S170A, N171D, N176T, N177M, V209M, W211I, M218F, K229E A C D E F G H I K L M N P Q R S T V W Y P6 0.50 -0.59 -0.04 0.02 -0.08 0.01 -0.01 0.11 0.02 0.02 0.07 -0.03 -0.03 -0.02 0.01 -0.07 -0.01 0.02 0.15 -0.72 P5 0.01 -0.57 -0.08 0.03 -0.05 0.00 0.09 -0.05 0.01 0.01 0.06 -0.23 0.07 0.03 0.09 -0.01 -0.03 0.00 -0.02 0.06 P4 0.03 -0.40 -0.21 -0.12 -0.11 -0.32 -0.41 3.80 -0.25 4.24 1.30 -0.16 -0.31 -0.38 -0.35 -0.32 -0.33 1.23 -0.20 -0.37 P3 -0.66 -0.86 -0.58 -0.79 1.42 -0.51 -0.13 1.82 -0.65 1.12 1.30 -0.78 -0.88 0.02 0.00 -0.85 -0.36 1.73 1.43 1.36 P2 0.50 -0.85 -0.83 -0.79 3.04 -0.64 -0.81 1.76 -0.82 0.53 -0.18 -0.70 -0.56 -0.80 -0.73 -0.28 0.02 1.94 2.67 2.22 P1 0.45 -0.66 0.11 1.74 1.97 -0.08 2.35 -0.70 -0.62 1.89 2.60 0.68 -0.54 3.04 -0.54 0.12 1.33 -0.76 -0.15 2.71 P1' 3.22 -0.51 0.98 1.36 2.07 2.13 1.99 3.16 -0.77 2.06 3.01 -0.80 -0.80 1.50 -0.74 2.45 2.26 2.69 2.52 2.87 TEV I138T, N171D, N176T A C D E F G H I K L M N P Q R S T V W Y P6 -0.81 -0.50 0.59 0.78 1.41 -0.51 -0.12 1.53 1.45 1.60 1.70 -0.42 1.50 1.06 -0.08 0.28 0.48 1.46 0.42 0.19 P5 0.63 -0.19 1.66 1.25 2.49 -0.76 0.85 2.32 -0.57 2.11 2.16 -0.06 2.39 0.93 -0.48 0.07 0.72 1.97 3.04 2.81 P4 -0.20 -0.22 -0.41 -0.25 -0.12 -0.16 -0.24 1.69 -0.13 4.15 -0.01 -0.13 -0.04 -0.14 -0.18 -0.21 -0.28 -0.25 -0.09 -0.27 P3 -0.26 -0.35 0.84 0.26 -0.08 1.62 -0.12 -0.33 -0.19 -0.25 -0.39 0.00 -0.19 -0.19 -0.12 -0.23 -0.30 -0.26 -0.26 3.35 P2 -0.06 -0.40 -0.10 -0.30 5.21 -0.18 -0.34 0.05 -0.29 -0.21 -0.38 -0.33 -0.77 -0.31 -0.18 -0.21 -0.26 1.74 -0.27 -0.48 P1 -0.10 -0.12 -0.47 -0.12 -0.13 -0.01 -0.20 -0.25 -0.06 -0.17 -0.11 -0.18 0.08 4.05 -0.08 -0.05 -0.07 -0.26 -0.11 -0.24 P1' 1.24 -0.31 -0.44 -0.29 0.20 0.74 0.05 -0.25 -0.33 -0.27 -0.04 -0.37 -0.36 -0.43 -0.40 2.35 -0.41 -0.41 1.24 -0.06 TEV T146S, D148P, S153N, S170A, N177M A C D E F G H I K L M N P Q R S T V W Y P6 0.48 0.35 1.31 2.68 0.52 -0.03 0.09 0.31 -0.16 0.26 0.36 0.18 0.15 0.34 -0.14 -0.01 0.10 0.21 0.18 0.19 P5 0.30 -0.48 0.20 0.30 0.28 -0.55 -0.06 0.42 -0.25 0.27 0.40 -0.23 0.30 0.30 -0.04 -0.08 0.17 0.32 0.30 0.17 P4 -0.34 -0.34 -0.39 -0.08 -0.35 -0.27 -0.37 4.37 -0.22 4.49 0.91 -0.20 -0.28 -0.24 -0.29 -0.33 -0.32 0.50 -0.31 -0.47 P3 -0.67 -0.73 -0.48 -0.59 0.83 0.02 -0.49 0.86 -0.72 -0.15 -0.38 -0.56 -0.74 -0.56 -0.45 -0.74 -0.69 0.81 0.66 2.27 P2 -0.36 -0.64 -0.58 -0.49 3.55 -0.47 -0.51 1.16 -0.65 0.14 -0.01 -0.42 -0.27 -0.56 -0.49 -0.37 -0.23 0.39 1.18 -0.43 P1 0.15 -0.43 0.14 1.21 1.55 -0.14 2.02 -0.78 -0.59 1.57 2.23 0.14 -0.61 2.16 -0.56 0.12 0.86 -0.78 0.08 1.99 P1' 1.00 -0.32 0.49 0.38 1.11 0.81 0.54 0.90 -0.60 0.33 0.73 -0.76 -0.81 0.13 -0.56 1.92 0.32 -0.08 1.00 2.71 TEV V209M, W211I, M218F A C D E F G H I K L M N P Q R S T V W Y P6 0.91 0.31 0.39 2.61 0.19 -0.01 0.01 0.22 0.31 0.29 0.28 0.09 0.11 0.17 -0.12 0.05 0.10 0.18 0.05 -0.46 P5 1.06 -0.37 1.42 1.05 1.65 -0.78 1.14 2.04 -0.55 1.47 1.45 0.36 1.88 0.60 -0.45 0.19 0.84 1.90 1.78 1.94 P4 -0.31 -0.22 -0.07 -0.13 -0.23 -0.19 -0.30 2.68 -0.13 4.45 0.19 -0.11 -0.17 -0.18 -0.23 -0.27 -0.27 0.26 -0.19 -0.26 P3 -0.39 -0.46 -0.03 -0.30 0.32 0.72 -0.34 -0.10 -0.34 -0.17 -0.44 -0.11 -0.23 -0.32 -0.03 -0.45 -0.44 -0.09 -0.16 3.38 P2 0.06 -0.50 -0.42 -0.49 3.77 -0.28 -0.50 -0.26 -0.51 -0.37 -0.27 -0.27 0.41 -0.56 -0.38 -0.29 -0.21 0.05 1.41 0.49 P1 -0.03 -0.07 0.06 0.20 -0.22 0.07 0.04 -0.30 -0.08 -0.25 0.51 -0.01 -0.07 3.92 -0.12 -0.07 -0.02 -0.29 -0.23 -0.29 P1' 0.79 -0.13 -0.07 -0.16 0.39 1.08 0.56 -0.33 -0.39 -0.36 -0.08 -0.50 -0.46 -0.30 -0.44 2.44 -0.24 -0.46 0.88 0.55 TEV H28L, T30A A C D E F G H I K L M N P Q R S T V W Y P6 0.56 0.18 0.24 0.33 0.58 0.04 -0.03 0.00 0.17 0.07 0.04 0.13 0.00 -0.05 -0.04 0.11 -0.01 -0.10 0.25 0.09 P5 0.11 -0.05 0.46 0.43 0.32 -0.39 -0.14 0.22 -0.45 0.32 0.66 -0.16 1.07 0.00 -0.38 -0.28 0.08 0.34 1.05 0.29 P4 -0.05 -0.03 -0.12 0.04 -0.11 0.03 -0.07 0.26 0.08 0.60 -0.12 0.04 0.06 -0.05 0.01 -0.01 -0.14 -0.10 -0.09 -0.20 P3 0.08 -0.07 -0.02 0.29 0.00 0.45 0.11 -0.16 0.23 -0.09 -0.20 0.22 0.11 0.03 0.04 0.03 -0.04 -0.20 -0.10 0.51 P2 0.07 -0.14 -0.13 -0.02 0.74 -0.08 -0.10 0.54 0.02 -0.09 -0.18 0.03 -0.35 -0.10 0.08 0.00 0.04 0.15 -0.18 -0.26 P1 -0.10 -0.09 0.16 0.01 -0.09 0.18 -0.07 -0.12 0.08 -0.16 -0.05 0.01 0.08 0.51 0.05 -0.02 -0.03 -0.15 -0.07 -0.14 P1' 0.02 -0.20 -0.21 -0.08 0.21 0.16 -0.04 -0.02 -0.18 -0.11 0.24 -0.21 -0.19 -0.22 -0.26 0.33 -0.07 -0.11 0.97 0.06 TEV T17S, N68D, E107D, D127A, F132L, S135F, F162S, K229E A C D E F G H I K L M N P Q R S T V W Y

Table 16 (Continued) P6 -0.85 -0.16 0.45 2.26 0.47 -0.06 -0.12 0.23 3.24 0.43 0.48 -0.09 -0.05 0.53 -0.08 -0.05 -0.02 0.16 -0.08 -0.05 P5 0.40 -0.42 0.10 0.90 0.92 -0.65 -0.31 1.30 -0.49 1.11 1.07 -0.44 1.43 0.59 -0.15 -0.30 0.53 1.13 1.47 0.92 P4 -0.21 -0.16 1.06 -0.06 -0.36 -0.06 -0.18 2.69 -0.12 2.92 0.00 -0.04 -0.10 -0.14 -0.12 -0.19 -0.21 0.00 -0.20 -0.36 P3 -0.27 -0.37 -0.34 -0.17 0.20 0.74 -0.23 -0.25 0.25 -0.26 -0.37 -0.08 0.06 -0.25 -0.09 -0.27 -0.35 -0.16 -0.15 2.34 P2 -0.25 -0.37 -0.24 -0.31 2.20 -0.33 -0.40 1.36 -0.39 -0.11 -0.17 -0.23 -0.31 -0.36 -0.21 -0.26 -0.16 0.80 0.19 -0.43 P1 -0.04 0.08 -0.45 0.08 -0.15 -0.12 0.00 -0.33 -0.12 -0.18 0.56 0.00 -0.02 2.36 -0.07 0.00 0.04 -0.36 -0.07 -0.27 P1' 0.91 -0.35 -0.10 -0.38 1.30 0.49 0.64 -0.05 -0.46 -0.26 0.58 -0.56 -0.41 -0.40 -0.50 1.69 -0.30 -0.35 2.37 1.85 TEV E107D, D127A, S135F, R203Q, K215E A C D E F G H I K L M N P Q R S T V W Y P6 -0.94 0.10 0.10 0.62 0.53 -0.01 -0.13 0.14 2.29 0.38 0.41 -0.11 -0.02 0.39 -0.04 -0.04 -0.08 0.14 0.00 0.44 P5 0.16 -0.39 -0.55 -0.44 0.12 -0.52 -0.44 0.51 0.36 0.46 0.68 -0.58 1.43 0.49 1.08 -0.37 0.19 0.33 0.24 -0.06 P4 -0.12 -0.07 0.56 -0.02 -0.20 -0.01 -0.18 1.05 0.02 0.94 0.03 0.06 -0.07 -0.06 -0.04 -0.13 -0.11 -0.06 0.00 -0.21 P3 0.00 -0.17 1.01 0.07 0.14 0.05 -0.10 -0.21 0.14 -0.02 -0.14 0.15 -0.12 -0.05 0.05 -0.09 -0.21 -0.18 0.14 0.35 P2 -0.12 -0.14 -0.10 -0.15 1.07 -0.24 -0.24 0.58 -0.11 -0.08 -0.18 -0.10 0.98 -0.17 0.01 -0.11 -0.13 0.27 0.07 -0.31 P1 -0.14 0.21 -0.49 -0.18 0.29 0.02 -0.18 -0.18 -0.07 0.12 0.13 -0.18 -0.03 0.76 -0.06 -0.11 -0.10 -0.24 0.60 -0.08 P1' 0.12 0.09 -0.27 -0.25 1.25 0.06 0.14 -0.21 -0.27 -0.15 0.14 -0.31 -0.27 -0.35 -0.30 0.72 -0.29 -0.35 1.50 1.24 TEV E107D, D127A, S135F A C D E F G H I K L M N P Q R S T V W Y P6 -0.81 0.25 0.57 2.22 1.16 -0.13 -0.08 0.90 2.59 1.14 0.94 -0.15 0.00 0.71 -0.14 -0.01 0.02 0.67 0.64 1.44 P5 0.15 -0.15 -0.08 0.20 0.42 -0.53 0.07 0.12 -0.18 0.27 0.47 -0.52 0.24 0.25 0.04 0.08 0.15 0.10 0.36 0.44 P4 -0.26 -0.14 0.80 -0.05 -0.20 -0.15 -0.27 3.37 -0.16 3.03 0.46 -0.02 -0.18 -0.29 -0.20 -0.25 -0.24 0.24 -0.03 -0.30 P3 -0.34 -0.40 -0.28 -0.27 0.73 0.75 -0.28 -0.15 -0.31 -0.25 -0.44 -0.16 -0.45 -0.32 -0.02 -0.41 -0.44 0.03 -0.12 2.57 P2 -0.37 -0.51 -0.54 -0.52 3.67 -0.55 -0.53 2.99 -0.56 0.09 0.03 -0.35 -0.18 -0.59 -0.41 -0.40 -0.22 2.52 1.22 -0.50 P1 -0.15 0.28 -0.47 -0.02 -0.11 -0.07 -0.02 -0.29 -0.09 -0.13 0.71 0.06 -0.07 2.88 -0.12 -0.03 -0.04 -0.36 0.05 -0.23 P1' 1.45 0.19 -0.04 -0.39 1.59 1.17 0.89 -0.27 -0.46 -0.35 1.25 -0.48 -0.56 -0.28 -0.46 2.27 -0.34 -0.49 1.89 2.40

WT TEV XXXYFQS TEV L2F XXXYFQS TEV L2F XXXVGHM

Enrichment Enrichment Enrichment