Engineering Enzyme Substrate Scope Complementarity for Promiscuous Cascade Synthesis of 1,2‐Amino Alcohols

Abstract Biocatalytic cascades are uniquely powerful for the efficient, asymmetric synthesis of bioactive compounds. However, high substrate specificity can hinder the scope of biocatalytic cascades because the constituent enzymes may have non‐complementary activity. In this study, we implemented a substrate multiplexed screening (SUMS) based directed evolution approach to improve the substrate scope overlap between a transaldolase (ObiH) and a decarboxylase for the production of chiral 1,2‐amino alcohols. To generate a promiscuous cascade, we engineered a tryptophan decarboxylase to act efficiently on β‐OH amino acids while avoiding activity on l‐threonine, which is needed for ObiH activity. We leveraged this exquisite selectivity with matched substrate scope to produce a variety of enantiopure 1,2‐amino alcohols in a one‐pot cascade from aldehydes or styrene oxides. This demonstration shows how SUMS can be used to guide the development of promiscuous, C−C bond forming cascades.

the lysis supernatant ran over the bead bed for purification by Ni-affinity chromatography. The column was washed with 4 column volumes of 20 mM imidazole, 50 mM KPi buffer (pH = 8.0).
Washing with higher concentrations of imidazole resulted in slow protein elution. ObiH was eluted with 250 mM imidazole, 50 mM KPi buffer (pH = 8.0). Elution of the desired protein product was monitored by the disappearance of its bright pink color (resulting from the release of ObiH) from the column. The protein product was dialyzed to < 50 μM imidazole in 50 mM Tris-HCl buffer (pH = 8.01) or 50 mM KPi buffer (pH = 8.08). Purified enzyme was flash frozen in pellet form by pipetting enzyme dropwise into a crystallization dish filled with liquid nitrogen.
The enzyme was transferred to a plastic conical and stored at -80 °C until further use. Frozen pellets were thawed at room temperature and centrifuged before use. The concentration of protein was determined by Bradford assay after freeze-thawing using bovine serum albumin for a standard concentration curve. Generally, this procedure yielded > 300 mg per L culture.

Cloning, expression, purification, and storage of RgnTDC
A codon-optimized copy of the Ruminococcus gnavus tryptophan decarboxylase (RgnTDC) gene was purchased as a gBlock from Integrated DNA Technologies. This DNA fragment was inserted into a pET-22b(+) vector by the Gibson Assembly method. 3 BL21 (DE3) E. coli cells were subsequently transformed with the resulting cyclized DNA product via electroporation.
After 30 min of recovery in LB media at 37 °C, cells were plated onto LB plates with 100 µg/mL ampicillin (AMP) and incubated overnight. Single colonies were used to inoculate 5 mL TB + 100 µg/mL AMP (TB-AMP), which were grown overnight at 37 °C, 200 rpm. Expression cultures, typically 1 L of TB-AMP were inoculated from these starter cultures and shaken (180 rpm) at 37 °C. After 3.5 to 4 hours (OD600 > 1.5), the expression cultures were chilled on ice.
After 45 min on ice, expression was induced with 1 mM IPTG, and the cultures were S13 supplemented with 0.5 mM indole. Cultures were expressed overnight at 23 °C with shaking at 180 rpm. Cells were then harvested by centrifugation at 4300xg at 4 °C for 15 min. Cell pellets were frozen and stored at -20 °C until purification.
To purify TDC, cell pellets were thawed on ice and then resuspended in lysis buffer (50 mM KPi buffer (pH = 8.0), 1 mg/mL Hen Egg White Lysozyme (GoldBio), 0.2 mg/mL DNaseI (GoldBio), 1 mM MgCl2, and 200 μM pyridoxal 5′-phosphate (PLP)). A volume of 4 mL of lysis buffer per gram of wet cell pellet was used. After 45 min of shaking at 37 °C, cells were sonicated with a ½ in. tip for 10 min (1 s on; 1 s off). The resulting lysate was then spun down at 75,000×g to pellet cell debris. Ni/NTA beads (GoldBio) were added to a gravity column and the lysis supernatant ran over the bead bed for purification by Ni-affinity chromatography. The column was washed with 4 column volumes of 20 mM imidazole, 50 mM KPi buffer (pH = 8.0). Washing with higher concentrations of imidazole resulted in slow protein elution. TDC was eluted with 250 mM imidazole, 50 mM KPi buffer (pH = 8.0). Elution of the desired protein product was monitored by the disappearance of its bright yellow color (resulting from the release of TDC) from the column.
The protein product was dialyzed to < 50 μM imidazole in 50 mM Tris-HCl buffer (pH = 8.01) or 50 mM KPi buffer (pH = 8.08). Purified enzyme was flash frozen in pellet form by pipetting enzyme dropwise into a crystallization dish filled with liquid nitrogen. The enzyme was transferred to a plastic conical and stored at -80 °C until further use. Frozen pellets were thawed at room temperature and centrifuged before use. The concentration of protein was determined by Bradford assay after freeze-thawing using bovine serum albumin for a standard concentration curve. Generally, this procedure yielded > 100 mg per L culture. Protein purity was analyzed by sodium dodecyl sulfate-polyacrylamide (SDS-PAGE) gel electrophoresis using 12% polyacrylamide gels.

Cloning, expression, purification, and storage of VlmD
A codon-optimized copy of the Kitasatospora setae valine decarboxylase (VlmD) gene was purchased as a gBlock from Integrated DNA Technologies. This DNA fragment was inserted into a pET-22b(+) vector by the Gibson Assembly method. 3 BL21 (DE3) E. coli cells were subsequently transformed with the resulting cyclized DNA product via electroporation. After 30 min of recovery in LB media at 37 °C, cells were plated onto LB plates with 100 µg/mL ampicillin (AMP) and incubated overnight. Single colonies were used to inoculate 5 mL TB + 100 µg/mL AMP (TB-AMP), which were grown overnight at 37 °C, 200 rpm. 0.5 L of TB-AMP was inoculated from these starter cultures and shaken (180 rpm) at 37 °C. After 3 hours (OD600 ~ 1.0), the expression culture was chilled on ice. After 1 h on ice, expression was induced with 1 mM IPTG, and the cultures were expressed for 16 hours at 23 °C with shaking at 180 rpm. Cells were then harvested by centrifugation at 4300xg at 4 °C for 15 min. Cell pellets were frozen and stored at -20 °C until purification.
To purify VlmD, cell pellets were thawed on ice and then resuspended in lysis buffer (50 mM KPi buffer (pH = 7.0), 1 mg/mL Hen Egg White Lysozyme (GoldBio), 0.2 mg/mL DNaseI (GoldBio), 1 mM MgCl2, and 200 μM pyridoxal 5′-phosphate (PLP)). A volume of 4 mL of lysis buffer per gram of wet cell pellet was used along with 2.5% BugBuster. After 1 h of shaking at 37 °C, the lysate was spun down at 75,000×g to pellet cell debris. Ni/NTA beads (GoldBio) were added to a gravity column and the lysis supernatant ran over the bead bed for purification by Niaffinity chromatography. The column was washed with 4 column volumes of 20 mM imidazole, 50 mM KPi buffer (pH = 7.0). VlmD was eluted with 250 mM imidazole, 50 mM KPi buffer (pH = 7.0). Elution of the desired protein product was monitored by the disappearance of its bright yellow color (resulting from the release of VlmD) from the column. The protein product was dialyzed to < 50 μM imidazole in 50 mM KPi buffer (pH = 7.0). Purified enzyme was flash frozen in pellet form by pipetting enzyme dropwise into a crystallization dish filled with liquid nitrogen.

S15
The enzyme was transferred to a plastic conical and stored at -80 °C until further use. Frozen pellets were thawed at room temperature and centrifuged before use. The concentration of protein was determined by Bradford assay after freeze-thawing using bovine serum albumin for a standard concentration curve. This procedure yielded 70 mg per L culture.

Cloning, expression, purification, and storage of SOI
A codon-optimized copy of the Pseudomonas sp. VLB120 StyC (SOI) gene was purchased as a gBlock from Integrated DNA Technologies. This DNA fragment was inserted into a pET-22b(+) vector by the Gibson Assembly method 3 . BL21 (DE3) E. coli cells were subsequently transformed with the cyclized DNA product via electroporation. After 45 min of recovery in TB media at 37 °C, cells were plated onto LB plates with 100 μg/mL AMP and incubated overnight. Following initial cloning into the pET-22b(+) vector, the gene encoding SOI was transferred to a pBAD vector (p15A origin, KAN resistant).
Single colonies were used to inoculate 10 mL TB-KAN, which was grown overnight at 37 °C, 200 rpm. Expression cultures, typically 1 L of TB-KAN, were inoculated with starter cultures (1% inoculum) and shaken (200 rpm) at 37 °C. After ~3 hours (OD600 = ~0.6), the expression cultures were chilled on ice. After 30 min on ice, expression of the protein was induced with 0.2% w/v arabinose. The cultures were expressed for 16-24 hours at 20 °C with shaking at 200 rpm. Cells were then harvested by centrifugation at 4,300×g at 4 °C for 10 min. Cell pellets were extruded from a syringe and flash frozen in liquid nitrogen as small pellets to give flash frozen wet cells.
Reactions were quenched via addition of 300 μL acetonitrile (ACN) followed by 300 μL H2O, which were then centrifuged to aggregate enzyme, and injected onto UPLC-MS for product detection. Product m/z areas were used to determine relative amounts of product formed ( Fig   S1).

Mutagenesis of wt-RgnTDC
Based on a previously reported structure of an RgnTDC-inhibitor complex (PDB ID: 4OBV), we modelled 3-hydroxyhomophenylalanine (3b) bound substrate into the active site (Fig 3a). H120 was chosen for site-saturation mutagenesis (SSM). Primers were purchased from Integrated DNA Technologies (IDT). For this site, three primers encoding the degenerate codons NDT, VHG, and TGG at the codon of interest were mixed in a 12:9:1 ratio, respectively. 4 The gene library was amplified first as two separate fragments and then combined via polymerase chain assembly (PCA) to form full-length RgnTDC gene mutagenized at the site of interest. 3 The corresponding gene was then inserted into a pET-22b(+) vector as described above and then transformed into BL21 (DE3) E. coli cells and plated on LB + 100 µg/mL AMP agar plates.
To investigate triple mutants, primers encoding H120N and L126M were used to add these mutations onto W349Y-and W349F-containing RgnTDC gene sequences via PCA. These two gene sequences were transformed and expressed as above. RgnTDC NMY = H120N, L126M, W349Y; RgnTDC NMF = H120N, L126M, W349F. Individual colonies were used to inoculate 5 mL TB + AMP in biological triplicate for each mutant and grown for 16 h at 37 °C. Variant strain cultures were used to inoculate 5 mL TB + AMP, which was expressed, induced, and collected as described above. Cell pellets were frozen at -20 °C prior to use. Cell pellets were thawed at room temperature and then resuspended in 1 mL lysis buffer (50 mM KPi buffer (pH = 8.0), 1 mg/mL Hen Egg White Lysozyme (GoldBio), 0.2 mg/mL DNaseI (GoldBio), 1 mM MgCl2, and 300 μM PLP). Lysis was conducted at 37 °C for 1 h. Lysates were centrifuged at 20,000 xg for 15 min, and the supernatants were used for the following reactions: 90 μL lysate supernatant + 10 μL 100 mM β-OH amino acid (3b, 4b, or 5b). Reactions progressed at 37 °C for 4 h and quenched via addition of 180 μL ACN to 20 μL reaction solution. Quenched reactions were then diluted with 200 μL H2O, centrifuged at 20,000 xg for 10 min to pellet enzyme, and then products analyzed via LC-MS using single ion retention channels for the expected products ( Fig S8).

RgnTDC des-hydroxy amino acid activity comparison
Freeze-thawed RgnTDC variants were spin-filtered, and the supernatants used for the following reaction: 10 mM amino acid, 400 μM PLP, 10 μM RgnTDC (for Leu/homoPhe) or 0.05 μM RgnTDC (for Trp/Phe), and 50 mM KPi buffer pH = 8.0 (final volume = 100 μL). Reactions were allowed to proceed for 16 h at 37 °C. Reactions were quenched via addition of 300 μL ACN followed by 300 μL H2O, which were then centrifuged to aggregate enzyme, and injected onto UPLC-MS for product detection. Product m/z areas were used to determine relative amounts of product formed (Leu/homoPhe) or UV-vis peak areas at 254 nm (Phe) or 280 nM (Trp) were used to determine relative conversions ( Fig S4).

RgnTDC diastereoselectivity for β-OH moiety
Freeze-thawed RgnTDC variants were spin-filtered, and the supernatants used for the following reaction: 10 mM p-Br, β-OH Phe (6b) 80:20 threo:erythro d.r., 400 μM PLP, 10 μM RgnTDC variant, and 50 mM KPi buffer pH = 8.0 (final volume = 100 μL). Reactions were allowed to proceed for 15 min at 37 °C. Reactions were quenched via addition of 300 μL ACN followed by 300 μL H2O, which were then centrifuged to aggregate enzyme, and injected onto UPLC-MS for product detection. UV-vis peak areas at 254 were used to determine relative concentrations of remaining amino acid starting material to observe changes in d.r. over the course of the reaction ( Fig S11).
For standard amino acid specificity determination, freeze-thawed RgnTDC variants were spinfiltered, and the supernatants used for the following reactions: 2 mM each substrate (His, Leu, Tyr, Phe, Trp), 400 μM PLP, 1 μM RgnTDC variant, and 50 mM KPi buffer pH = 8.0 (final volume = 100 μL). Reactions were conducted in duplicate for 10 min at 37 °C. Reactions were quenched using 500 μL ACN, centrifuged to pellet aggregated enzyme, and injected onto UPLC-MS for product detection. Product m/z's were used to quantify relative product abundances ( Fig S4).
Reactions were quenched using 500 μL ACN, centrifuged to pellet aggregated enzyme, and injected onto UPLC-MS for product detection. Product m/z areas were used to quantify relative product abundances. Each reaction condition was conducted in duplicate.
Additionally, RgnTDC activity was assayed independently under different buffer conditions. Freeze-thawed RgnTDC NMY was spin-filtered, and the supernatant used for the following reaction: 10 mM β-OH homoPhe (3b), 400 μM PLP, 10 μM RgnTDC NMY , and 50 mM Tris/KPi buffer (final volume = 100 μL). Buffer pH's ranged from pH = 7.0 to pH = 8.5. Reactions were conducted for 16 h at 37 °C at 180 rpm. Reactions were quenched using 300 μL ACN, followed by 300 μL H2O, centrifuged to pellet aggregated enzyme, and injected onto UPLC-MS for product detection. Product m/z areas were used to quantify relative product abundances. Each reaction condition was conducted in duplicate ( Fig S10)

Co-solvent conditions:
Freeze-thawed ObiH and RgnTDC H120N was spin-filtered, and the supernatants used for the

Determination of turnover numbers for TDC variants via Marfey's derivation:
Freeze-thawed RgnTDC variants were spin-filtered, and the supernatants were used in the following reaction: 10 mM amino acid (3b, 5b, or 6b) 500 μM PLP, 50 μM RgnTDC variant (for 3b reaction), 20 μM RgnTDC variant (for 5b reaction), or 0.3 μM RgnTDC variant (for 6b reaction), and 50 mM KPi buffer pH = 8 (final volume = 100 μL). Reactions were allowed to proceed in a 37 C incubator for 16 h prior to quenching with 100 μL ACN. The samples were S22 centrifuged to pellet aggregated enzyme and diluted an additional 1:10 with 50 mM KPi buffer pH = 8.0 to ensure the total amine concentration for the Marfey's derivatization was low.
In a microcentrifuge tube, 30 μL of diluted reaction mix (1 equiv., 0.05 mM final total amines from unreacted β-OH amino acid substrate, and formed 1,2-amino alcohol product) was added to a solution of 120 μL of 18.75 mM NaHCO3 (150 equiv., 7.5 mM final concentration

Determination of analytical substrate scope yields via Marfey's derivation:
For analytical cascade yields, freeze-thawed ObiH and RgnTDC NMY were spin-filtered, and the supernatants used for the following reactions: 10 mM aldehyde/styrene oxide (2a-22a), were quenched via addition of 50 μL 1 M HCl and diluted to 1 mL with 1:1 ACN:H2O. Reaction solutions were analyzed on UPLC-MS by looking at absorbance at 340 nm and comparing relative product peak areas to the derivatized Arg peak area (Fig 4; Fig S12)

Calculation of product enantiomeric ratios via Marfey's derivation:
A Marfey's derivatization reaction was performed to assess the enantiomeric ratios of formed 1,2-amino alcohols following the above procedure with purified material. However, the Marfey's reaction was conducted both in the presence of L-FDVA or D-FDVA to ensure enantiomer separation via our UPLC-MS method. Enantiomeric ratio (e.r.) was determined by Marfey-derivatized product peaks integrated at 340 nm.

Stereoisomer determination of 22c:
Freeze-thawed RgnTDC NMY was spin-filtered, and the supernatant used for the following reactions: 10 mM 22b (65:35 anti:syn for the γ-Me and β-OH groups; prepared as previously described, 2 400 μM PLP, 0 or 10 μM RgnTDC NMY , 50 mM KPi buffer pH = 8.0 (final volume = 100 μL). Reactions were conducted in duplicate for 16 h at 37 °C. Reactions were quenched using 300 μL ACN, followed by 300 μL H2O, centrifuged to pellet aggregated enzyme, and injected onto UPLC-MS for product detection. Product m/z's were used to identify substrate and product peaks. Selective amino acid stereoisomer conversion was observed only for the minor syn amino acid diastereomer along with appearance of a new product peak, which we attribute to be the syn 1,2-amino alcohol (Fig S13). After the reaction, the insoluble product was isolated by vacuum filtration, washed several times with H2O, and dried under vacuum. The product was isolated as a yellow crystalline powder (170.8 mg) in a 74% yield. Product purity analyzed by 1 H NMR was determined to be 96%. Acetone (2.08 ppm) and H2O (3.33 ppm) were determined to be the main impurities. After the reaction, the insoluble product was isolated by vacuum filtration, washed several times with H2O, and dried under vacuum. The product was isolated as a yellow crystalline powder (153.3 mg) in a 65% yield. Product purity analyzed by 1 H NMR was determined to be 84%. Acetone (2.08 ppm) and H2O (3.33 ppm) were determined to be the main impurities.

Synthesis of Marfey's Reagent
Biocatalytic cascade synthesis of 1,2-amino alcohols *We found that purification of the protonated amines was generally easier than purification of the deprotonated amines on C18. We additionally observed high Tris (buffer) retention under basic purification conditions, whereas Tris eluted early as a single peak under acidic conditions. For these reasons, product isolations from C18 were generally performed at pH < 1 via acidification using HCl. We still occasionally observed remaining Tris in isolated product (singlet at 3.6 ppm) and Thr (doublet at 1.4 ppm).
Although acid/base extraction was used for the purification of several of the below compounds, considering the surprising simplicity of C18 purification and high resulting compound purity, we opted for column chromatography purification of many products.

S26
Additionally, since RgnTDC H120N was equally active on phenylserine amino acids, we used this enzyme for many syntheses instead of RgnTDC NMY , simply due to catalyst availability.

S28
The reaction solution was then heat-treated at 75 °C for 20 minutes to aggregate protein, diluted with 50 mL ACN, and centrifuged at 4000 xg for 10 min. The pellet was washed with 10 mL 1:1 ACN:H2O. The combined supernatants were concentrated via rotary evaporation down to ~5 mL, and 6 M HCl was added to bring the pH < 1. The resulting suspension was injected onto a 30 g C18 flash chromatography column and the product eluted early at 1% MeOH. Productcontaining fractions were pooled and concentrated via rotary evaporation, frozen @ -80 °C, and lyophilized until dry. The product was isolated as the HCl salt as a white powder (105 mg) in 96% yield. Enantiopurity was determined by subsequent Marfey's analysis to be 97.5:2.5 e.r.
(R:S).     The reaction solution was then heat-treated at 75 °C for 20 minutes to aggregate protein, diluted with 40 mL ACN, and centrifuged at 4000 xg for 10 min. The supernatant was concentrated via rotary evaporation down to ~3 mL, and 6 M HCl was added to bring the pH < 1. The resulting suspension was injected onto a 30 g C18 flash chromatography column and the product eluted at 1% MeOH. Product-containing fractions were pooled and concentrated via rotary evaporation, frozen @ -80 °C, and lyophilized until dry.

S32
Later fractions were observed to contain significant product. These fractions were pooled and concentrated via rotary evaporation. This aqueous solution was then basified with 6 M NaOH and extracted thrice with 50 mL EtOAc. The desired product was then extracted from the EtOAc layer thrice with 50 mL dilute HCl, which were subsequently combined and concentrated via rotary evaporation. This solution was then frozen @ -80 °C and lyophilized until dry.   The reaction solution was then heat-treated at 75 °C for 20 minutes to aggregate protein, diluted with 25 mL ACN, and centrifuged at 4000 xg for 10 min. The supernatant was concentrated via rotary evaporation down to ~15 mL. The pH of the solution was adjusted via addition of 6 M NaOH to > 11 followed by 5x extraction with 50 mL EtOAc. The organic extractions were combined and evaporated to ~50 mL. The product was then extracted from the organic layer via addition of 5x 50 mL dilute HCl (pH ~3). The resulting aqueous layer was concentrated via rotary evaporation and was transferred to a pre-tared flask, frozen @ -80 °C, and lyophilized until dry. The product was isolated as the HCl salt as a white crystalline powder (25 mg) in 29% yield. Enantiopurity was determined by subsequent Marfey's analysis to be > 99:1 e.r. (R). The reaction solution was then heat-treated at 75 °C for 20 minutes to aggregate protein, diluted with 50 mL ACN, and centrifuged at 4000 xg for 10 min. The pellet was washed with 10 mL 1:1 ACN:H2O. The combined supernatants were concentrated via rotary evaporation down to ~40 mL. The pH of the solution was adjusted via addition of 6 M NaOH to > 11 followed by 5x S40 extraction with 75 mL EtOAc. The organic extractions were combined and evaporated to ~50 mL. The product was then extracted from the organic layer via addition of 3x 50 mL dilute HCl (pH ~3). The resulting aqueous layer was concentrated via rotary evaporation and was injected onto a 30 g C18 flash chromatography column. The product eluted at 1% MeOH. Productcontaining fractions were pooled and concentrated via rotary evaporation, frozen @ -80 °C, and lyophilized until dry. The product was isolated as the HCl salt as a white crystalline solid (25mg) in 33% yield. Enantiopurity was determined by subsequent Marfey's analysis to be > 99:1 e.r. (R).

Structure Solution and Refinement
The systematic absences in the diffraction data were consistent for the space groups P21 and P21/m. The E-statistics strongly suggested the non-centrosymmetric space group P21 that yielded chemically reasonable and computationally stable results of refinement [5][6][7][8]. A successful solution by intrinsic phasing provided most non-hydrogen atoms from the E-map.
The remaining non-hydrogen atoms were located with an alternating series of least-squares cycles and difference Fourier maps. All nonhydrogen atoms were refined with anisotropic displacement coefficients. All hydrogen atoms (except those bound to non-C atoms) were included in the structure factor calculation at idealized positions and were allowed to ride on the neighboring atoms with relative isotropic displacement coefficients. The absolute configuration of the C2 chiral atom is R. The final least-squares refinement of 121 parameters against 2083 data resulted in residuals R (based on F 2 for I≥2σ) and wR (based on F 2 for all data) of 0.0350 and 0.0960, respectively. The final difference Fourier map was featureless.
CCDC 2207847 contain the supplementary crystallographic data for this paper. These data are   Colored bars represent relative amounts of each product formed, and diamonds represent mM total product produced, as determined by single-ion retention areas. The wild-type sequence is denoted by a grey diamond. Relative product amounts and mM total product were averaged from all wells with the given sequence. The greyed-out section of the graph indicates measurements that were indistinguishable from noise. Figure S4. Activity of RgnTDC variants with des-β-OH amino acids. a) Single-substrate activity comparisons. Reaction conditions: 10 mM amino acid, 400 μM PLP, 0.05 μM RgnTDC (for Trp/Phe) or 10 μM RgnTDC (for homoPhe/Leu), and 50 mM KPi buffer pH = 8.0 (final volume = 100 μL). Reactions were allowed to proceed for 16 h at 37 °C. Reactions for Trp were conducted in triplicate and all other reactions were conducted in duplicate. b) Analysis of RgnTDC variant specificities for standard amino acids. Reaction conditions: 2 mM each substrate (His, Leu, Tyr, Phe, Trp), 400 μM PLP, 1 μM RgnTDC variant, and 50 mM KPi buffer pH = 8.0 (final volume = 100 μL). Reactions were conducted in duplicate for 10 min at 37 °C. Average product single ion retention (SIR) areas were integrated to quantitate relative amounts of product formed. Figure S5. Screening results from the H120N + L126X site-saturation mutagenesis library. Colored bars represent relative amounts of each product formed, and diamonds represent mM total product produced, as determined by single-ion retention areas. The wildtype sequence is denoted by a grey diamond, and sequenced wells are labeled. Relative product amounts and mM total product were averaged from all wells with the given sequence. The greyed-out section of the graph indicates measurements that were indistinguishable from noise. Figure S6. Screening results from the H120N + Y312X site-saturation mutagenesis library. Colored bars represent relative amounts of each product formed, and diamonds represent mM total product produced, as determined by single-ion retention areas. The wildtype sequence is denoted by a grey diamond, and sequenced wells are labeled. Relative product amounts and mM total product were averaged from all wells with the given sequence. The greyed-out section of the graph indicates measurements that were indistinguishable from noise. Figure S7. Screening results from the H120N + W349X site-saturation mutagenesis library. Colored bars represent relative amounts of each product formed, and diamonds represent mM total product produced, as determined by single-ion retention areas. The wildtype sequence is denoted by a grey diamond, and sequenced wells are labeled. Relative product amounts and mM total product were averaged from all wells with the given sequence. The greyed-out section of the graph indicates measurements that were indistinguishable from noise. Figure S8. RgnTDC variants lysate activity verifications on single substrates. Reaction conditions: 90 μL lysate supernatant + 10 μL 100 mM β-OH amino acid (3b, 4b, or 5b) at 37 °C for 4 h. Single ion retention (SIR) peak areas were used to determine product abundance. No product was detected for reactions with 4b. Figure S9. Active-site comparisons between various amino acid decarboxylases. PDB ID's are denoted for each structure (RgnTDC = 4OBV, PasYDC = 6EEM, SspMetDC = 7CII). The residues in the black box are analogous to the H120 residue of RgnTDC. The structure of VlmD was generated using SwissModel with 7CII serving as the parent structure.