Subscriber access provided by UNIV OF TEXAS DALLAS
Article
Systematic screening of synthetic gene-encoded enzymes for synthesis of modified glycosides Zachary Armstrong, Feng Liu, Hong-Ming Chen, Steven Hallam, and Stephen G. Withers ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b05179 • Publication Date (Web): 27 Feb 2019 Downloaded from http://pubs.acs.org on March 2, 2019
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
1 1
Title:
2
Systematic screening of synthetic gene-encoded enzymes for synthesis of modified glycosides
3 4
Authors:
5
Zachary Armstrong1, Feng Liu2, Hong-Ming Chen2, Steven J. Hallam1,3,4,7, Stephen G. Withers1,2,5,6, *
6 7
Author Affiliations:
8
1 Genome
9
Vancouver, BC V6T 1Z4, Canada
Science and Technology Program, University of British Columbia, 2329 West Mall,
10
2 Department
11
Columbia, Canada V6T 1Z1
12
3 Department
13
Canada V6T 1Z1
14
4 Graduate
15
V6T 1Z4
16
5 Centre
17
Vancouver, British Columbia, Canada V6T 1Z3
18
6 Department
19
Columbia, Canada V6T 1Z3
20
7 ECOSCOPE
21
V6T 1Z3
of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, British of Microbiology & Immunology, University of British Columbia, Vancouver, BC, Canada,
Program in Bioinformatics, University of British Columbia, Vancouver, BC, Canada, Canada
for High-Throughput Biology, Michael Smith Laboratory, University of British Columbia, of Biochemistry and Molecular Biology, University of British Columbia, Vancouver, British Training Program, University of British Columbia, Vancouver, British Columbia, Canada
22 23
*Corresponding Author Email:
[email protected] ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2 24
Abstract:
25
The construction of large, phylogenomically diverse libraries of synthetic genes from a sequence-
26
based family allows rapid evaluation of the substrate specificity encoded within the gene products and
27
identification of rare catalytic capabilities. For carbohydrate-active enzymes such as glycosidases this opens
28
up approaches to search for enzymes that are capable of catalyzing the hydrolysis and transfer of modified
29
sugars. Herein we report the screening of a library of E. coli clones each harbouring one of 175 synthetic
30
genes that represent the sub-families within the glycoside hydrolase family GH1 to identify a sub-set of
31
enzymes that cleave amino- and azido-glucosides modified at the 3-, 4-, and 6-positions. Eight of the most
32
active such enzymes were further characterized through detailed kinetic analysis and plate-based aglycone
33
specificity screening. Three glycosynthase mutants of each of these enzymes were generated in which their
34
catalytic nucleophile glutamate was substituted by alanine, glycine or serine and their synthetic capabilities
35
explored with -glucosyl fluoride and 3-, 4- and 6-amino and azido derivatives. The utility of seven of the
36
generated mutants was then demonstrated through the assembly of taggable disaccharides, trisaccharides,
37
glycolipid analogues and activity-based probes. Not only does this work present the creation of several
38
useful biocatalysts, it also provides a general framework that can be exploited to develop catalysts from
39
other glycoside hydrolase containing clone libraries.
40 41
Keywords:
42
Glycosynthase, biocatalysis, high-throughput functional screening, glycoside hydrolase,
43
CAZymes, phylogenomic library.
ACS Paragon Plus Environment
Page 2 of 23
Page 3 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
3 44
INTRODUCTION
45
The use of biocatalysts in chemical synthesis offers advantages of increased selectivity and
46
efficiency while reducing the amount of waste generated. Development of new or improved biocatalysts is
47
achieved either by modifying an existing enzyme or by mining the vastly unexplored repertoire present in
48
nature. The modification of biocatalysts is probably best achieved through directed evolution, which has
49
been used to create a cornucopia of improved biocatalysts1-4. However, the level of success attained in
50
directed evolution is often dictated by the choice of starting enzyme scaffold5-6.
51
Another route towards new biocatalysts, either in their own right, or as improved start points for
52
directed evolution, is to screen libraries of clones or isolates for the desired activity. This method, however,
53
is often limited by the ability to collect, curate and store large isolate libraries, all of which may have
54
differing growth requirements. Functional screening of metagenomic libraries created from environmental
55
DNA within a standard host is an excellent complement to traditional isolate library screening7-8, and has
56
been applied to a number of enzyme classes including glycoside hydrolases8-9 and esterases10-11. The
57
approach minimises bias due to growth requirements, though expression of target genes depends on host
58
transcription factors working heterologously thus typically results in low hit rates, resulting in a need for
59
large libraries12.
60
More recently, access to affordable large-scale DNA synthesis has enabled the systematic sampling
61
of natural diversity across an entire family of proteins. In a particularly relevant example, large scale gene
62
synthesis coupled with high-throughput screening enabled Heins et. al. to identify CAZy family 1 glycoside
63
hydrolase (GH1) enzymes that were both thermostable and tolerant to ionic liquids, and to define the
64
activity of a broad range of genes13. Although the cost of gene synthesis remains prohibitively high for most
65
researchers to perform such studies, it is our hope that the advancement of gene synthesis technologies will
66
make these types of studies more feasible. The benefits of this approach are several-fold. As with
67
metagenomic screening, protein expression is unbiased towards host growth conditions, but in contrast,
68
gene expression can be more tightly controlled by design. Additionally, genes from more than one
69
environment or species can be analysed concomitantly, providing a great diversity of catalysts. This method
70
also enables the use of ever-expanding gene databases, including genes derived directly from environmental
71
DNA.
72
In this light, we aimed to harness this diverse set of GH1 enzymes to help address a long-standing
73
problem in the biological sciences: ready access to synthetic oligosaccharides. The development of facile
74
methods for synthesis of peptides and oligonucleotides revolutionised the biological sciences. However,
75
despite excellent progress in automated glycan synthesis14-17 we still lack a universal and reliable method
76
for glycan assembly, largely because of the enormously greater regio- and stereochemical challenges posed.
77
Further, this approach will be limited for larger scale syntheses. The alternative to chemical synthesis
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
4 78
involves use of enzymes, most likely either the natural nucleotide phosphosugar-using glycosyltransferases
79
or synthetically useful mutant forms of glycoside hydrolases, glycosynthases18. These latter enzymes are
80
created by mutating the catalytic nucleophile residue of a retaining glycosidase (Glu or Asp in almost all
81
cases) to either Ala, Gly or Ser. Such mutants are hydrolytically incompetent with natural substrates, but
82
when presented with a glycosyl fluoride donor sugar of inverted anomeric configuration, will typically
83
transfer that sugar to an appropriate acceptor, often in stoichiometric yield, without subsequent hydrolysis.
84
The first glycosynthase generated was from the GH1 -glucosidase from Agrobacterium sp19 and many
85
have been developed since and used to create a variety of glycans including: glycolipids20, glycosidase
86
inhibitors21 and defined glycoproteins22-25.
87
This has been accomplished through the creation of glycosynthases from new families26, and
88
through directed evolution1. Our intent in this work was to explore the use of large synthetic gene libraries
89
as a means of identifying useful catalysts for the formation of specific glycosides that could not be
90
assembled with currently known enzymes. Of particular interest was the ability to assemble
91
oligosaccharides containing amino- or azido substituents at the 3, 4 or 6 positions. These would be useful
92
not only in the degradation/assembly of aminosugar-containing glycans, such as antibiotics or bacterial
93
surface polysaccharides, but also as a way of incorporating modifiable glycans into biomolecules under
94
mild conditions for subsequent “tagging”. More broadly this served as a test system for generating a pipeline
95
for the discovery and generation of custom biocatalysts for glycan assembly.
96
Herein we report the use of high-throughput screening of both glycone and aglycone specificity
97
within a synthetic gene-based library of expressing clones to identify promising candidate glycoside
98
hydrolases, which we then transform into glycosynthases and characterize (Figure 1). Screening of a library
99
of 175 GH1 enzymes13 against a library of six aryl glycosides bearing amino- and azido functionalities
100
identified 106 clones against at least one substrate. A set of eight competent hydrolases with the desired
101
glycone (-1 site) specificity to accommodate these modifications was selected for further investigation. The
102
aglycone (+1 site) specificities of these eight enzymes were then explored against a panel of 83 aglycone
103
“acceptors” in a second, plate-based screen, identifying optimal candidates to utilize as acceptors in
104
subsequent glycosynthase reactions. The alanine, serine and glycine mutants of the nucleophilic glutamate
105
residue were created for each of the eight GHs and the activities of the corresponding 24 mutants were
106
evaluated. Mutant forms of seven of the eight hydrolases selected acted as competent glycosynthases
107
yielding the desired glycans containing azido or amino substituents. The utility of these mutants in the
108
assembly of taggable activity-based probes was demonstrated.
ACS Paragon Plus Environment
Page 4 of 23
Page 5 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
5
109 110
Figure 1. Screening methodology. The -1 subsite specificity is evaluated through glycosidase
111
screening with fluorogenic substrates. Acceptor specificity is based on stimulation of reactivation of a
112
trapped 2-fluoroglycosyl enzyme through transglycosylation to a competent acceptor. The extent of
113
reactivation is assessed with a chromogenic substrate, giving an indication of +1 subsite specificity. The
114
corresponding glycosynthase can then be employed with cognate acceptor and donor substrates.
115 116
RESULTS AND DISCUSSION
117
Screening of Glycone (-1 Site) Specificity.
118
The library of 175 E. coli clones, each bearing one of the synthetic GH1 genes as characterized by
119
Heins et. al.13, was grown and lysed and the crude lysates were directly screened with a set of aryl
120
glycosides. Six substrates, each bearing a fluorogenic 4-methylumbelliferyl leaving group and either an
121
azido or amino group at the 3, 4, or 6 position of glucose were used to interrogate the activities of this set
122
of enzymes. Of the clones tested, 106 were active (z-score > 9) on at least one substrate, and 91 had activity
123
on more than one substrate (Figure 2). In general, specificity for the aminosugar substrates was broader
124
than for the azidosugars, with more clones capable of cleaving the amino-glucosides (4-NH2-Glc MU:
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
6 125
99/175, 6-NH2-Glc MU: 83/175, 3-NH2-Glc MU: 64/175) than the azido-glucosides (6-N3-Glc MU: 91/175,
126
3-N3-Glc MU: 2/175, 4-N3-Glc MU: 1/175). Further, the absolute activities (total fluorescence released) for
127
hits identified with the 3- and 4- substituted azido sugars were much lower than for the other hits,
128
corresponding to less than 1 % substrate turnover.
129
130 131
Figure 2. Screening Results. Phylogenomic tree of synthetic genes expressed showing the relative
132
fluorescent intensity released in each well. Each leaf corresponds to a protein and the bar height indicates
133
relative activity. Fluorescent values are given as the fraction of the maximum expected fluorescence.
ACS Paragon Plus Environment
Page 6 of 23
Page 7 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
7 134
Results for 3-N3-Glc MU and 4-N3-Glc MU are not shown. Parent tree with gene identifications is provided
135
in Supporting Information.
136 137
The finding of relatively few enzymes that are capable of cleaving the 3-N3-Glc MU and 4-N3-Glc
138
MU substrates is a reflection both of the considerable steric demand of an azide substituent compared to an
139
amine or the parent hydroxyl, as well as of the importance of the hydrogen bonds normally formed between
140
the enzyme and the substrate at those positions. Indeed the key residues involved in interactions with the
141
3- and 4-hydroxyls, His123, Gln20, Trp423 and Glu422 (numbering based on Phanerochaete
142
chrysosporium GH1) are highly conserved
143
tolerated, most likely because of the greater conformational flexibility possible at the 6-position, allowing
144
the substituent to adopt an orientation that minimises steric repulsion. The readier acceptance of an amine
145
substituent than an azide at C-3 or C-4 likely stems from its smaller size, as well as its hydrogen bonding
146
potential. Likewise, amine substitution at the 6 position also seems to be broadly tolerated. The hit rates
147
seen for the amino glucosides, in fact, nicely mirror the specificities determined for the GH1 -glucosidase
148
Abg by Namchuk and Withers29 through measurement of kinetic parameters for hydrolysis of a set of mono-
149
deoxyglycoside substrates in which each hydroxyl, individually, had been replaced by hydrogen. From
150
these parameters, the contributions of interactions with each hydroxyl to transition state stabilization were
151
determined, yielding ΔΔGo‡ values of 7.4, 2.5 and 2.9 kJ/mol for the 3-, 4- and 6-hydroxyls respectively,
152
very much in line with the lower tolerance seen here for substitution at C-3.
27-28.
The 6-azido substituent, however, is reasonably well
153
From the 106 hits so identified we chose 8 enzymes, which between them encompassed all the
154
activities sought, as candidates for transformation into glycosynthases (Table 1). As a first step, kinetic
155
parameters were measured for each modified substrate with each of these purified wild-type enzymes
156
(Supporting Information Table S1) with the exception of the Lac enzyme, for which activity on 3-NH2-Glc
157
MU, 3-N3-Glc MU, and 6-N3-Glc MU was below detectable levels. Since this enzyme is primarily
158
a 6-phospho-beta-glucosidase, it is not surprising that it has such sparing activity on the non-
159
phosphorylated modified glucosides. In many cases KM values were too high to be reliably
160
measured, so values of kcat/KM, the specificity constant, were determined instead through
161
measurements at low substrate concentrations. In general, for those enzyme/substrate pairs
162
where cleavage was detected, higher specific activities were seen with 6-modified substrates
163
(Supporting Information Table S1). Indeed, in two cases kcat/KM values measured with 6-modified
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
8 164
substrates were higher than for the parent glucoside. Also, interestingly, in some cases the 6-
165
azido glucoside was cleaved more rapidly than the 6-amino, and in others the reverse was seen.
166
Additionally, all enzymes capable of cleaving both 4-amino and 3-amino glucosides displayed
167
higher specific activities towards the 4-amino substrates. This again is very much in agreement
168
with the specificity studies on Abg noted above.
ACS Paragon Plus Environment
Page 8 of 23
Page 9 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
9 169
Table 1. Selected Genes and the Activities Identified Through Screening with Modified Glucosides. Genbank
AAZ81839.1
ACO44852.1 ACQ71106.1 CAQ67883.1 ABF87202.1 BAE87008.1 ABD82858.1 AAF37730.1
170
Species Alicyclobacillus acidocaldarius Deinococcus deserti VCD115 Exiguobacterium sp. AT1b Lactobacillus casei BL23 Myxococcus xanthus DK-1622 Phanerochaete chrysosporium Saccharophagus degradans 2-40 Thermobifida fusca
Abbr.
Ref.
Ali_GH1
Modified Glucosides Glc MU
3-NH2
4-NH2
6-NH2
3-N3
4-N3
6-N3
30
✔
✔
✔
✔
-
-
✔
Dei_GH1
-
✔
✔
✔
✔
-
-
✔
Exi_GH1
-
✔
✔
✔
✔
low
-
✔
Lac_GH1
-
✔
✔
✔
-
low
low
-
Myx_GH1
-
✔
✔
✔
✔
-
-
✔
Pha_GH1
31
✔
-
✔
✔
-
-
✔
Sac_GH1
-
✔
✔
✔
✔
-
-
✔
The_GH1
32
✔
-
-
-
-
-
✔
low: less than 1% of the total substrate was hydrolysed
171 172
Acceptor Specificity
173
To gain insight into the specificity of the +1 subsite, each of the 8 wild type enzymes was subjected
174
to aglycone specificity screening33. This method is based upon screening of relative rates of reactivation,
175
through transglycosylation, of a stabilised, but catalytically competent, glycosyl enzyme intermediate. The
176
slow substrate 2,4-dinitrophenyl 2-deoxy-2-fluoro--glucoside was first added to each of the eight enzymes
177
to form the stabilised intermediate. After removal of excess 2-fluorosugar the inactivated enzyme was
178
incubated for a fixed time period with a range of potential reactivating ligands and the amount of reactivated
179
enzyme formed in each well assessed by addition of a chromogenic substrate (pNP-Glc in this case).
180
Compounds that function as effective reactivators are likely to also be suitable glycosynthase acceptors,
181
allowing selection of the optimal candidate for each reaction.
182
All 8 enzymes were screened against a panel of 83 potential acceptors, including a variety of
183
glycosides, free sugars and alcohols (Supporting Information Table S3). As can be seen in Figure 3, each
184
enzyme displayed a different pattern of aglycone specificity. The extent of reactivation of each enzyme also
185
differed, with the maximum rates of reactivation for Ali_GH1, Pha_Gh1 and The_GH1 being fairly modest
186
(0.4 %, 2.1 % and 3.5 % of the uninhibited rate, respectively) when compared to those of Dei_GH1,
187
Exi_GH1, Lac_GH1, Myx_GH1 and Sac_GH1 (19.7 %, 83.8 %, 59.4 %, 47.2 % and 100 % of the
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
10 188
uninhibited rate, respectively). The majority of the best reactivators were aryl glycosides, with six of the
189
eight enzymes being reactivated fastest by an aryl glucoside (pNP-Glc, pNP-α-Glc or MU-α-Glc). The
190
exceptions to this were Pha_GH1, for which cellobiosides were best – suggesting a strong preference for
191
cello-oligosaccharide aglycones, and Sac_GH1, which reactivated fastest with pNP -D-fucopyranoside.
192 193
Figure 3. Acceptor Specificity. The top five reactivators and the relative initial rates observed are shown
194
for each of the wildtype enzymes. Enzymes screened are abbreviated as follows: (A) Ali_GH1, (D)
195
Dei_GH1, (E) Exi_GH1, (L) Lac_GH1, (M) Myx_GH1, (P) Pha_GH1, (S) Sac_GH1, and (T) The_GH1.
196
Rates are scaled to the best reactivator for each given enzyme. The control is the activity seen when no
197
reactivator was included.
198 199
Glycosynthases
200
We sought to make glycosynthases from each of the eight hits, with the hopes that they would be
201
competent at transferring modified glucosides onto a variety of molecules. Mutants were constructed in
202
which the conserved nucleophilic glutamate residue, for each enzyme, was replaced by one of three
203
different amino-acids (Serine, Alanine and Glycine) in the hopes that at least one of these would be an
204
active glycosynthase. All 24 mutant enzymes were expressed and purified from cultures grown on a 50 mL
ACS Paragon Plus Environment
Page 10 of 23
Page 11 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
11 205
scale. Initially enzymes were tested for glycosynthase activity using α-glucosyl fluoride (αF-Glc, 50 mM)
206
as a donor and pNP-Glc (10 mM) as an acceptor, this being the most widely accepted in the screen (Figure
207
3) and 20 µM of each enzyme. For six (Ali_GH1, Dei_GH1, Exi_GH1, Myx_GH1, Sac_GH1, The_GH1)
208
of the eight enzymes, at least one mutant acted as a glycosynthase with this donor/acceptor combination.
209
Mutants of the other two (Lac_GH1 and Pha_GH1) did not yield products. However, when Pha_GH1
210
nucleophile mutants were incubated with α–glucosyl fluoride and pNP--cellobioside (pNP-Cel), one of
211
the best reactivators for the wild-type enzyme, products were indeed seen. Since Lac_GH1 has a much
212
higher specificity constant for the hydrolysis of 6-phospho-glucosides than for glucosides (Supporting
213
Information Table S1), we suspected that the nucleophile mutants would be competent glycosynthases with
214
6-phosphosugar donors. Unfortunately, none of the Lac_GH1 mutants were able to effect glycosyl transfer
215
from 6-phospho-α-glucosyl fluoride to any of the top five reactivators from the screen that were tried.
216
To choose the best glycosynthase mutant from the three different variants we performed HPLC
217
analysis of small scale reactions. Reaction mixtures contained an equal amount of donor and acceptor (αF-
218
Glc and pNP-Glc or pNP-Cel at 5 mM). All glycine mutants displayed a low hydrolytic activity, which we
219
speculate is due to mis-incorporation of the wild-type glutamate, as has been reported previously for other
220
GH1 glycosynthases34-35. In those studies, mis-incorporation is seen for the gene containing the GGG codon
221
for glycine. In our case we used GGA, however, both of these codons differ at only one position from
222
codons for glutamate (GAG, GAA): future glycine-based glycosynthase creation should utilize codons
223
containing 2 substitutions (GGC, GGT). Of the serine and alanine mutants, the serine variant provided the
224
best yield of the major product for all enzymes except for The_GH1. However, the differences between
225
mutants were small (
