spiral.imperial.ac.uk · web viewappl microbiol biotechnol 97:4429–4442. doi:...
TRANSCRIPT
Amplification of Ehrlich pathway enhances production of higher alcohols in engineered Yarrowia
lipolytica
Ewelina Celińska1*, Monika Borkowska1, Wojciech Białas1, Monika Kubiak1, Paulina Korpys1, Marta
Archacka1, Rodrigo Ledesma-Amaro2,3, Jean-Marc Nicaud2
¹Department of Biotechnology and Food Microbiology, Poznan University of Life Sciences, ul. Wojska
Polskiego 48, 60-627 Poznań, Poland
²Micalis Institute, INRA-AgroParisTech, UMR1319, Team BIMLip: Integrative Metabolism of Microbial
Lipids, Domaine de Vilvert, 78352 Jouy-en-Josas, France
3 Imperial College London, Centre for Synthetic Biology, South Kensington Campus, London SW7 2AZ,
United Kingdom
*Corresponding author: Ewelina Celińska, phone: +48618466007, e-mail:
Abstract
Microbial cells can produce a vast spectrum of chemical compounds, including those most desired by
the global chemical market, as for example higher alcohols, which are promising alternative fuels and
chemical feedstock. In the current research, we investigated the effects of the Ehrlich pathway
amplification on higher alcohols production in Y. lipolytica, which directly follows our previous
findings concerning elucidation of putative molecular identities involved in this pathway. To this end
we constructed two alternative expression cassettes composed of previously identified genes and
cloned them under the control of constitutive pTEF promoter, and by this - released them from
extensive native regulation. The effects of the pathway amplification were investigated upon
provision of different, Ehrlich pathway-inducing amino acids (L-Phe, L-Leu, L-Ile and L-Val). Depending
on the expression system (cassette-genetic background of the host) we could observe different
effects on the targeted compound production. In general, amplification of the Ehrlich pathway in
many cases led to increased formation of a respective higher alcohol from its precursor by on
average 1.5 to >> 2-fold. We observed interesting effects of bat2Δ on 2-PE and 2-PEAc formation,
significant relationship between L-Val and L-Phe catabolic pathways, and extensive “cross-induction”
of the derivative compounds synthesis by a non-direct precursors.
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
12
Keywords:
higher alcohols, amino acid metabolism, Ehrlich pathway, bioconversion, Yarrowia lipolytica,
metabolic engineering
Abbreviations
AA – Amino Acid
BCAA – Branched Chain Amino Acid (L-Leu, L-Val, L-Ile)
aroma AA – Amino Acid with aromatic residue (L-Phe, L-Trp, L-Tyr)
GGFs– Golden Gate Fragments
GGVs – Golden Gate Vectors
GGVA – Golden Gate Vector bearing Assembly
GGA - Golden Gate Assembly
2-MEB – 2-Methyl-1-Butanol
2-MEBAc - 2-Methyl-1-Butyl Acetate
2-MEP – 2-Methyl-1-Propanol
2-MEPAc - 2-Methyl-1-Propyl Acetate
2-MEPacid - 2-Methyl Propanoic acid/Isobutyric acid
3-MEB – 3-Methyl-1-Butanol
3-MEBAc - 3-Methyl-1-Butyl Acetate
3-MEBacid - 3-Methyl Butanoic acid/ Isovaleric acid
2-PE – 2-Phenylethanol
2-PEAc – 2-Phenethylacetate
ARO8 [YALI0E20977p] - aminotransferase
BAT2 [YALI0F19910p] – aminotransferase
ARO10 [YALI0D06930p] - decarboxylase
Adh [YALI0F24937p]- alcohol dehydrogenase
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
34
Introduction
Growing awareness of the environmental burden imposed by chemical and fuel industry, as
well as uncertain supply of the fossil fuels are the major driving forces for development in the green
chemistry area. Apart from lower environmental impact, biotechnological processes employing
indigenous or engineered potential of microbial metabolism, may turn out to be superior in terms of
efficacy and economic feasibility compared to the chemical synthesis, as already demonstrated by
several examples (Nakamura and Whited 2003; Schrader et al. 2004). Amongst top most desired
chemicals on the global market, the authorities and stakeholders name several diols (2,3-butanediol,
1,3-propanediol), organic acids (succinic acid, lactic acid), lipid / alkene-derived metabolites
(ethylene, fatty acids, fatty alcohols) and higher alcohols (butanol, isobutanol) (Sauer et al. 2008;
Alper and Stephanopoulos 2009; Borodina and Nielsen 2014; Jansen and van Gulik 2014; Dai and
Nielsen 2015; Generoso et al. 2015; Nielsen and Keasling 2016). Higher alcohols are aliphatic,
medium-chain compounds of more than three carbon atoms with branched or aromatic side-
residues. Due to their useful chemical properties, higher alcohols have found applications in
numerous industrial processes, starting from production of alternative fuels with superior
physicochemical properties, through serving as chemical feedstock, to aromatization of various
goods (Nozzi et al. 2014). Microbial cells can produce a vast spectrum of chemical compounds,
including those most desired by the global chemical market. Higher alcohols can be produced in
microbial cells de novo by sequestering carbon skeletons from the central carbon metabolism, or in a
course of bioconversion from structurally-related precursors. While de novo synthesis seems more
reasonable in terms of the culture media costs (no need for the precursors), it is by far less efficient
than bioconversion (Schrader et al. 2004). Bioconversion pathway per se is decoupled from the core
carbon metabolism and thus the precursor is not dissipated in the competing pathways. As such,
carefully optimized bioconversion processes may approach perfect atom economy, expressed in high
yield of the product per consumed precursor. Higher alcohols, either synthesized de novo or in a
course of bioconversion, are inherently coupled with AA metabolism. Catabolism of L-leucine (L-Leu),
L-isoleucine (L-Ile), L-valine (L-Val) or L-phenylalanine (L-Phe) leads to the formation of higher
alcohols, namely 3-MEB (Isoamyl alcohol; CAS 123-51-3; 3-MBE), 2-MEB (Active Amyl Alcohol; CAS
137-32-6; 2-MBE), 2-MEP (Isobutanol; CAS 78-83-1; 2-MPE) and 2-Phenylethyl Alcohol (2-PE; CAS 60-
12-8; 2-PE), respectively. The metabolic pathway of branched chain AAs (L-Val, L-Leu, L-Ileu),
aromatic AAs (L-Phe, L-Tyr, L-Trp) and L-Met has been initially described by Ehrlich (Ehrlich 1907),
and scrutinized in the following studies employing Saccharomyces cerevisiae (Dickinson et al. 2003;
Hazelwood et al. 2008) Figure 1. In terms of the yeast cell physiology, the Ehrlich pathway is used to
seize nitrogen from AA. In terms of the chemical industry – it leads to formation of valuable chemical
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
56
compounds of significant interest. These chemical compounds derive from the carbon skeletons of
the AA which basically remain intact. The Ehrlich pathway is initiated by the action of
aminotransferase, which uses 2-oxoglutarate or other 2-oxo acids as amino acceptors to the
corresponding keto acids (phenylpyruvic acid from L-Phe, α-ketoisocaproic acid from L-Leu, α-
ketoisovaleric acid from L-Val, and α-ketomethylvaleric acid from L-Ile). For some AAs (e.g., L-Ala),
the resulting 2-oxo acid (pyruvate), can be readily metabolized in the central carbon metabolism by
the producer cell. However, in the other cases, such as aroma AA and BCAA, the 2-oxo acids resulting
from transamination are not intermediates of the central metabolism, implying that they cannot be
used as auxiliary carbon sources. The 2-oxo acids are secreted out of the cells as they are, or may be
also transformed into higher alcohols or derivatives by the yeast cells before they are excreted into
the growth medium (Vuralhan et al. 2005). Synthesis of each of the higher alcohols may be
accompanied by concomitant formation of its ester (3-MEBAc (CAS 123-92-2), 2-MEBAc (CAS 624-41-
9), 2-MEPAc (CAS 110-19-0)) or fusel acid (3-MEBacid (CAS 503-74-2) or 2-MEPacid (CAS 79-31-2)),
which also find numerous applications in industrial processes. Noteworthy, higher alcohols and their
esters constitute an important fraction of aroma compounds used in aromatization of various
commodities, including foodstuff and cosmetics. In contrast to fuel, chemical or polymer industry,
aroma production is a specific branch which is strongly influenced by the consumer acceptance. Thus,
highly-valued natural origin of the compound is an additional parameter driving its final market price.
According to European regulation on flavors (EEC No 1334/2008), aroma chemical compound can be
labelled as “natural” when “the flavoring substance is obtained by appropriate physical, enzymatic or
microbiological processes from material of vegetable, animal or microbiological origin either in the
raw state or after processing”. This in turn, makes biotechnology a powerful alternative compared to
chemical synthesis, contributing to substantial burden imposed on the natural environment and
providing “synthetic” compounds, or to exploitation of natural resources, which are highly
susceptible to seasonal variations, and/or hard-to-foresee climate-geographical factors.
Amongst microbial species potentially assisting production of higher alcohols, yeasts occupy
an important place, with the leading role assigned to a model species S. cerevisiae. A number of
studies investigating molecular background of the compounds synthesis (Dickinson et al. 2003;
Hazelwood et al. 2008; Styger et al. 2013), inventive strategies of the pathway engineering (Brat et al.
2012; Avalos et al. 2013; Nozzi et al. 2014) or intensification of the production processes employing
this species, have been published to date. Apart from S. cerevisiae, the other yeast species are
becoming attractive alternatives in microbial production of higher alcohols (Sibirny and Scheffers
2002; Forti et al. 2015; Grondin et al. 2015; Celińska et al. 2018). Although their superior potential
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
78
towards high-level production of higher alcohols has been demonstrated, the number of studies
within this area stay far behind the model species.
A list of valuable biomolecules synthesized by either native or engineered metabolism of
Yarrowia lipolytica is continuously increasing (Alper and Stephanopoulos 2009; Groenewald et al.
2014; Madzak 2015; Ledesma-Amaro and Nicaud 2016). Apart from a number of bioprocesses
employing Y. lipolytica’s metabolism shown to be highly promising in the research context, several
technologies are operating in the industry, like the production of citric acid (Pfizer Inc. and ADM,
USA), erythritol (Baolingbao Biology Co., China), eicosapentaenoic acid-rich oils (DuPont, USA),
nutrient-rich fodder yeast (Skotan SA, Poland; (Białas et al. 2015), or enzymes (Oxyrane, Belgium
and UK; Artechno, Belgium; Mayoly Spindler, France) (Darvishi Harzevili 2014; Groenewald et al.
2014), demonstrating high propensity of this yeast to serve as an industrial workhorse. For a long
time, Y. lipolytica has been employed in industrial production of creamy-peach aroma compound γ-
decalactone from ricinoleic acid, commercialized by BASF. Recently, we have shown that Y. lipolytica
is an efficient producer of another highly-valued higher alcohol – 2-PE in a course of bioconversion
from L-Phe (Celińska et al. 2013). In the following study we performed whole-proteome analysis in
order to fish-out molecular identities involved in the bioconversion of the AA to the higher alcohol
(Celinska et al. 2015). As mentioned above, the Ehrlich pathway is executed via three subsequent
reactions: deamination, decarboxylation and oxidation, catalyzed by transaminase (like ARO8/9 or
BAT1/2 [EC:2.6.-.-.]), 2-oxo acid decarboxylase (like ARO10 or PDC1/3/5 [EC:4.1.1.-]), and
dehydrogenase (like ADH1–5 in S. cerevisiae; [EC:1.1.1.90]), respectively. It is known, that in S.
cerevisiae each step can be conducted by several possible enzymes implying that the Ehrlich pathway
is highly redundant (Lilly et al. 2000; Dickinson et al. 2003; Vuralhan et al. 2003; Vuralhan et al. 2005;
Lilly et al. 2006; Hazelwood et al. 2008). Although the general overview of the Ehrlich pathway was
known, involvement of particular molecular identities in this pathway had not been elucidated
before in Y. lipolytica. Based on the proteomic data, we observed presence or over-representation of
specific proteins which, based on their sequence, function and similarity to S. cerevisiae’s “template”
pathway, could be directly linked to the general pathway of higher alcohols formation. The first step
of the Ehrlich pathway in Y. lipolytica could be catalyzed by ARO8 [YALI0E20977p; 2.6.1.5] and BAT2
[YALI0F19910p; 2.6.1.57] homologues whose presence was identified in the analyzed proteome.
YALI0D06930p [EC:4.1.1.43] was a single homolog of ARO10/PDCs which has been identified with
good matching results only in samples where the Ehrlich pathway was induced. This protein has been
also identified with good matching results, when blasting all the homologous sequences
(ARO10/PDC1/PDC5/PDC6) from S. cerevisiae against Y. lipolytica protein database. The last activity
that is directly involved in AAs catabolism via Ehrlich pathway is dehydrogenase. Determining the
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
910
molecular identities of the dehydrogenase involved in the Ehrlich pathway presents a challenge
(Hazelwood et al. 2008), Dickinson, Salgado and Hewlins (2003) have shown that any of the alcohol
dehydrogenases (Adh1–5 and Sfa1) encoded within S. cerevisiae genome can catalyze this final
reaction, demonstrating high redundancy of this step. In the previous study we have observed high
over-representation of a single dehydrogenase having broad-range substrate specificity -
YALI0F24937p. Significant induction of its expression, implied its involvement in the catabolism of
AAs via the Ehrlich pathway. Consequently, based on comparative whole proteome analysis we
identified candidate genes for genetic manipulations oriented towards enhanced formation of higher
alcohols in Y. lipolytica.
In the current research, we investigated the effects of the Ehrlich pathway amplification on
higher alcohols production in Y. lipolytica. To this end we constructed two alternative expression
cassettes composed of previously identified aminotransferases encoded by YALI0E20977g or
YALI0F19910g (either ARO8 or BAT2, depending on the cassette variant), decarboxylase encoded by
YALI0D06930g, and YALI0F24937g coding for broad substrate specificity dehydrogenase. The genes
were cloned under the control of constitutive promoter pTEF and Lip2 terminator, thus released
from native regulation mechanisms. The three-gene assemblies were transformed in two Y. lipolytica
strains (JMY2101 and JMY3150) differing in the genetic background related to the Ehrlich pathway
(JMY3150). The effects of the pathway amplification were investigated upon provision of different,
Ehrlich pathway-inducing AAs in comparison to the reference strain. It is known that at greater
concentrations, higher alcohols exert toxic effect toward microbial cells. Thus, to avoid toxic effect of
higher alcohols overproduction, we applied suboptimal conditions of the precursor feeding, to
enable observation of the modifications effects.
Materials and Methods
Strains and routine culturing conditions
E. coli and Y. lipolytica strains used in this study are listed in Table S1. All cultivations required
for molecular biology protocols complied with the standards described in (Barth and Gaillardin 1996;
Sambrook and Russell 2001). Briefly, E. coli strains were routinely maintained on LB medium (g/L: 10,
bacteriological peptone, 10, NaCl, 5, yeast extract; liquid or solidified with agar, 15 g/L)
supplemented with appropriate antibiotic when necessary (ampicillin at 100 µg/L; kanamycin 40
µg/L), at 37°C, 250 rpm in rotary shaking incubator (Biosan). Y. lipolytica strains were cultured in YNB
(g/L: 5, ammonium sulfate, 1.7, YNB without AA and ammonium sulfate, 20, glucose) or YPD (g/L: 10,
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
1112
yeast extract, 20, bacteriological peptone, 20, D-glucose) medium (liquid or solidified with agar, 15
g/L), at 30°C, 250 rpm.
Molecular biology protocols
If not stated otherwise, all the molecular biology protocols followed the methodologies
described in (Sambrook and Russell 2001). All oligonucleotides and plasmids used in this study are
listed in Table S2.
E. coli and Y. lipolytica transformations were conducted according to the heat-shock or LiAc
methodologies described in (Sambrook and Russell 2001) and (Barth and Gaillardin 1996),
respectively. Genomic DNA isolation from Y. lipolytica, plasmid isolation from E. coli, DNA fragments
extraction from agarose gel or purification of DNA fragments were all conducted using appropriate
kit from A&A Biotechnology (Poland) – Genomic Mini AX Yeast, Plasmid Mini, Gel-Out, Clean-Up.
Restriction digestion of DNA fragments was done using either NotI enzyme (Thermo Scientific) or
BsaI (New England Biolabs). Routine colony PCR with E. coli biomass was run using Taq DNA
polymerase (A&A Biotechnology; Poland) while colony PCR with Y. lipolytica biomass was conducted
using Phire Hot Start II DNA Polymerase (Thermo Scientific). Phire Hot Start II DNA Polymerase was
also used for amplification of any GGFs from Yarrowia’s genomic DNA, and for assembly PCR
procedure enabling elimination of internal BsaI recognition sites found within ARO8 (YALI0E20977p),
BAT2 (YALI0F19910p), ARO10 (YALI0D6930g) nucleotide sequences. In silico restriction analysis was
done using BioEdit’s Restriction Map tool (Hall 1999) and confirmed with control restriction
digestion. Elimination of undesired recognition sites from the targeted sequences was confirmed
through sequencing (GATC Biotech, Germany; Genomed, Poland), and BsaI restriction digestion. The
reactions were conducted according to protocols provided by the manufacturers.
Modular cloning – Golden Gate Assembly and positive clones selection
The modular cloning protocol followed a previously set standard (Celinska et al. 2017). All the
plasmids and primers used in this study are listed in Table S2. Briefly, a set of thirteen 4 nt overhangs
was developed together with the corresponding destination vector pSB1A3-RFP,
(http://parts.igem.org/Collections). Genomic DNA sequence of Y. lipolytica CLIB122 used in this study
as GGFs amplification template can be acquired from GRYC database (http://gryc.inra.fr/). All the
GGVs (GGF-bearing donor vectors) were constructed on a pCR Blunt II TOPO vector’s backbone
(Thermo Scientific) and a corresponding GGF amplicon, according to the instruction provided by the
manufacturer. Ligation mixture was transformed into one of E. coli host strains (JM109, TOP10 or
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
1314
DH5α), and correct GGVA organization was confirmed through restriction digestion and
electrophoresis. DNA concentration was assessed using NanoDrop device (Thermo Fisher Scientific).
The Golden Gate reaction mixture contained precalculated equimolar amounts of each GGF and the
destination vector (50 pmoles of ends), 2 µL of T4 DNA ligase buffer (New England Biolabs), 5 U of
BsaI, 200 U of T4 and ddH2O up to 20 µL. The following thermalcycling profile was applied: [37°C - 5
min, 16°C - 2 min] x 60, 80°C - 5 min, 15°C ∞. Subsequently, the reaction mixtures were used for E.
coli JM109 transformation. White colonies were screened for identification of complete GGA through
colony PCR covering adjacent elements (e.g. P1_F and G1_R), followed by plasmid isolation,
restriction digestion and multiplex PCR. Complete GGAs were linearized with NotI and transformed
into JMY2101 and JMY3150 strains. Clones appearing after 48 h incubation at 30°C on YNB-selection
plates were replica-plated on fresh YNB plates. All the clones were screened for the GGA presence
through colony PCR covering adjacent GGFs. All the strains confirmed to bear a complete GGA were
deposited as glycerol stocks at -80°C.
Batch bioconversion cultures
Batch bioconversion cultures to be analyzed through GC were conducted with at least two
subclones of each type selected from deposited library: JMY2101_GGA077, JMY3150_GGA077,
JMY2101_GGA097 and JMY3150_GGA097; each cultured in triplicate. JMY2900 control strain was
cultivated in parallel with each culture run as a non-transformed reference. The cultures were
conducted in shake flasks of total volume 50 mL and working volume 5 mL, at 30°C, 250 rpm. The
basal production medium contained in g/L: 50, glycerol, 1 mL, mineral salt solution II (Barth and
Gaillardin 1996), 15, KH2PO4, 0.5, MgSO4x7H2O, 0.2, yeast nitrogen base. The medium was
supplemented with appropriate AA, either: L-Phe, BCAA (Sigma Aldrich) up to 2 g/L. The supplements
were prepared as 10-fold concentrated stocks, sterilized through filtering through 0.22 µm (Millex,
Millipore). The cultures were continued for 72 h and the samples were collected at 24 h intervals and
centrifuged (15 min, 15 min, 22,639 x g; Biofuge R, Hareus). Cellular pellets and supernatants were
frozen separately at -20°C until analysis. The samples were used for analysis of biomass production
and targeted aroma compounds concentration. Biomass accumulation was assessed through OD600
measurements using UV/Vis Spectrophotometer SPECORD (Analytik Jena, Germany). Concentration
of targeted metabolites retained in the culture medium was determined through gas
chromatography analysis (GC; procedure described below).
Batch bioconversion cultures to be analyzed through GC-MS were conducted in 300 mL
Erlemeyer flasks in a working volume of 100 mL, in the same medium and culturing conditions as
described above. Three independent cultivations were conducted for each recombinant strain type
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
1516
(JMY2101_GGA097, JMY3150_GGA097, JMY2900). Samples were collected at 48 h of culturing and
subjected to GC-MS analysis.
Extraction of volatile compounds in the culture media
After thawing, the supernatant samples were vortexed, centrifuged (15 min, 22,639 x g;
Biofuge R, Hareus) and passed through 0.45 µm filter (Millex, Millipore). Extraction of the aroma
compounds was conducted through mixing the filtrate with equal quantities (GC analysis) or 9:1 ratio
(GC-MS analysis) of dichloromethane, vigorous shaking, and phases separation via centrifugation (15
min, 22,639 x g; Biofuge R, Hareus). The organic phase was collected and subjected to either GC or
GC-MS analysis. Comparison of the extraction efficiency was assessed by conducting parallel
cultivations in different culture volumes (5 mL vs 100 mL) followed by extraction with either 0.5 or
0.1 volume of dichloromethane.
GC analysis
The GC analysis of the organic phase (1 µL) was conducted using Agilent Technologies 7890A
System gas chromatograph equipped with a Zebron™ ZB-WAXplus™ GC capillary column and FID
detector. Mobile phase: hydrogen (30 mL/min), air (300 mL/min), helium (20 mL/min). The oven
programme was as follows: 80°C for 1 min, 80°C to 120°C, at 20°C/min, followed by 120°C to 220°C,
at 120°C/min, and final hold at 220°C for 3 min. Temperature of detector - 250°C. Data were analyzed
using Agilent Chemstation software. Compounds were identified and quantified through comparison
with analytical standards of known concentrations: 1) L-Leu derivatives: 3-MEB (CAS 123-51-3), 3-
MEBAc (CAS 123-92-2), 3-MEBacid (CAS 503-74-2); 2) L-Val derivatives: 2-MEP (CAS 78-83-1), 2-
MEPAc (CAS 110-19-0), 2-MEPacid (CAS 79-31-2); 3) L-Ile derivatives: 2-MEB (CAS 137-32-6), 2-
MEBAc (CAS 624-41-9); 4) L-Phe derivatives: 2-Phenylethyl Alcohol (CAS 60-12-8), 2-Phenylethyl
Acetate (CAS 103-45-7). All analytical standards were purchased from Sigma-Aldrich. Presence and
quantity of all the compounds was determined in all analyzed samples, however, only the “non-zero
concentration” compounds are shown in the respective figures in the Results section.
GC-MS analysis
GC-MS analysis was conducted at Wielkopolska Center of Advanced Technologies (Poznan,
Poland). GC-MS SCION TQ (Bruker) equipped with BP5MS capillary column (5% Phenyl
Polysilphenylene-siloxane) 30 m x 0,25 mm (length x inner diameter), 0,25 µm mobile phase
thickness was used. Mobile phase: helium at 1 mL/min. Oven temperature profile: initial column
temperature: 40°C - 3.5 min, followed by 10°C/min up to 250°C, 250°C - 5 min hold. Total analysis
time: 29.5 min. Temperature of injector: 185°C, ion source: 250°C, transfer line: 200°C. Ioninzation 70
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
1718
eV, scanning range: 25-300 m/z. Each sample was analyzed in technical triplicate. The obtained raw
data were analyzed using Bruker MS Workstation 8.0 Software. Compounds contained in the sample
were identified through comparison of spectra against NIST database.
Results
Design and construction of recombinant strains
DNA constructions were designed and assembled according to previously described
methodology (Figure 2AB). DNA sequence analysis of the ARO-genes revealed that ARO8, BAT2 and
ARO10 all bear two internal BsaI sites, which were eliminated through assembly PCR with
corresponding “BsaI del” primers (Table S2) and verified through sequencing and in silico translation.
Assemblies were first constructed and amplified in E. coli cells. Subsequently, correct constructions
were transformed into Y. lipolytica JMY2101 and JMY3150 strains. Clones bearing complete
assemblies were selected by colony PCR with primers covering adjacent biobricks (gene and
regulatory element) (Figure 2C). Positive clones of each of the four types (JMY2101 bearing either
GGA077 and GGA097 assemblies, or JMY3150 bearing GGA077 and GGA097 assemblies) were
deposited as glycerol stocks and subjected to further studies.
Phenotype evolution and analyses
Representative clones of the recombinant strains: JMY2101_GGA0077, JMY2101_GGA0097,
JMY3150_GGA0077 and JMY3150_GGA0097, were selected and subjected to bioconversion cultures
with L-Phe, L-Leu, L-Ile and L-Val as the sole nitrogen sources. The results are presented in Figures 3
and 4, as well as in supplementary material S1 and S2.
In general, presuming that overall biomass accumulation reflects feasibility of a given AA
assimilation, L-Val was the most easily utilized AA amongst all strains (13.82±0.5 OD600), followed by
L-Leu (13.64±0.35 OD600 average in JMY2900 and JMY2101), which assimilation was impaired in
bat2Δ JMY3150 strains (9.64±0.63 OD600; Figure S1.C.G, S2.C.G). In GGY0077 strains, the weakest
growth was observed upon feeding with L-Ile and L-Phe (9.18±0.06 and 8.61±0.39 OD600), while in
GGY0097 strains – the lowest OD600 readouts were observed in cultures with L-Phe as the sole
nitrogen source (7.1±0.15 OD600; Figure S1.A.E, S2.A).
With respect to the strains transformed with GGA0077 cassette cultured in the bioconversion
medium supplemented with L-Phe (Figure 3.A), more than 50% increase in 2-PE formation was
observed when compared to the reference strain (average for the recombinants 0.65±0.064 vs
0.424±0.078 g/L for the reference strain), while the growth was not significantly compromised in the
engineered strains (8.6±0.52 vs 10.04±1.04 OD600; Figure S1.A). The final specific titer of 2-PE was
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
1920
exactly 2-fold higher for the strains bearing recombinant ARO genes under the control of constitutive
pTEF promoter (average 0.084±0.001 vs 0.042 g/OD600, for the recombinants and the reference
strain, respectively; Figure S1.B). Considering the accompanying metabolites, all the recombinants
synthesized ~50% higher quantities of 2-MEPacid (co-product of L-Val bioconversion) while cultured
in L-Phe-supplemented medium, which was demonstrated by both the final concentration and
specific titer (average 0.2±0.0014 vs 0.13±0.02 g/L±SD (1.54-fold increase) and average 0.023 vs
0.012 g/OD600 (1.92-fold increase), for the recombinants and the reference strain, respectively;
Figure 3.B and S1.B). The levels of the other analyzed compounds (in g/L; 3-MEBacid, 2-PEAc) were
more deviating than actually different between the recombinants and the reference strain, however,
the final concentration and specific titer of 2-MEB (L-Ile derivative) were uniformly more than 2-fold
higher for all the recombinants than for the reference strain (0.011±0.001 vs 0.0038±0.0003 g/L±SD
and 0.0012 vs 0.00038 g/OD600 for the recombinants and the reference strain; Figure 3.B, S1.B).
Upon cultivation with L-Phe as the sole nitrogen source, the bat2Δ genotype in GGY0077s slightly
improved 2-PE formation, suggesting a kind of competition for the NH4+ residues between “aromatic
AA” and “BCAA” bioconversions, initiated by ARO8 and BAT2 aminotransferases, respectively, which
was relieved upon deletion of BAT2 gene in JMY3150. This statement is corroborated by the results
obtained with the JMY2101 and JMY3150 strains transformed with GGA0097 cassette (Figure 4.A and
S2.B). Additional copy of BAT2 aminotransferase in JMY2101 (altogether 2 copies – one under native
regulation, and the other – under pTEF), ultimately led to slightly lower 2-PE formation in a course of
L-Phe bioconversion (g/L±SD: JMY2101_GGA097: 0.37±0.05 / 1.37-fold decrease, JMY3150_GGA097:
0.48±0.042 / 1.06-fold decrease, reference strain: 0.508±0.018), which was also illustrated by specific
titers of 2-PE per OD600 (0.053 2-PE g/OD600 giving 1.36–fold decrease for JMY2101-derivatives,
0.066 2-PE g/OD600 giving 1.1–fold decrease for JMY3150-derivatives and 0.072 2-PE g/OD600 for
JMY2900; Figure S2.B). Consequently, it can be stated that pTEF-driven expression of ARO10 and Adh
did not result in higher bioconversion of L-Phe to 2-PE when BAT2 was concomitantly expressed,
either in JMY2101 or JMY3150. On the other hand, amplification of ARO8 and the following genes of
the Ehrlich pathway, enhanced 2-PE synthesis by more than 50% in the recombinants in general.
Further enhancement by more than 60% was observed upon deletion bat2Δ.
It is commonly now known that AA metabolism is more an extensive net of biochemical
reactions than a linear pathway. Hence, we also studied performance of here developed Y. lipolytica
recombinants upon provision of the other AA, known to be metabolized through Ehrlich pathway,
namely BCAA (L-Leu, L-Val, L-Ile). Considering recombinants bearing constitutive expression of the
Ehrlich pathway in GGA077 variant cultured in BCAA-containing media, expectedly production of 2-
PE was rather low, reaching on average 0.01 – 0.02 g/L (Figure 3.D,F,H). Nevertheless, the highest 2-
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
2122
PE concentrations were observed in the case of JMY3150_GGA077 strains (doubled ARO8, ARO10
and Adh, in bat2Δ) cultured in L-Val containing media (0.02±0.004 g/L / 4.35-fold increase for the
JMY3150-derivatives, 0.007±0.001 / 1.52-fold increase for the JMY2101-derivatives, 0.0046±0.001
g/L for JMY2900; g/L±SD), again suggesting that bat2Δ genotype is advantageous for 2-PE formation
(Figure 3.H). Cultivation of the GGA077-bearing strains in L-Ile-containing media showed that the
GGA077 modification was the enabling factor for detectable 2-PE formation, as the reference strain
did not formed this metabolite at all, as shown by 2-PE and its corresponding ester levels (2-PEAc)
(Figure 3.F). Also, an increase in 2-PE and its ester formation was observed for the GGA077-bearing
strains cultured in L-Leu-supplemented medium (average for the recombinants 0.0083±0.0014 vs
0.0043±0.0011 g of 2-PE/L±SD for the reference; 1.93-fold increase; Figure 3.D).
Interestingly, amplification of the Ehrlich pathway (GGA077) uniformly led to increased
formation of a respective higher alcohol, being derivative of the provided BCAA: 3-MEB from L-Leu
(0.76±0.038 / 2.38-fold increase for JMY2101-derivatives, 0.56±0.042 / 1.75-fold increase for
JMY3150-derivatives, and 0.32±0.04 for JMY2900; g/L±SD) (Figure 3.C), 2-MEP from L-Val
(0.077±0.012 / 2-fold increase for the JMY2101-derivatives, 0.086±0.049 / 2.26-fold increase for the
JMY3150-derivatives, and 0.038±0.004 for JMY2900; g/L±SD) (Figure 3.G), and 2-MEB from L-Ile
(0.025±0.002 / 1.6-fold increase for the JMY2101-derivatives, 0.023±0.002 / 1.5-fold increase for the
JMY3150-derivatives, and 0.0157±0.002 for JMY2900; g/L±SD) (Figure 3.E), which was also reflected
by specific titers of the derivative compounds (Figure S1.D, F, H). It seems that amplification and
constitutive overexpression of the broad range decarboxylase and oxidoreductase and relatively
broad range aminotransferase ARO8, improved formation of the higher alcohols. Even more
interestingly, the accompanying fusel acid production was concomitantly decreased for the
recombinants when compared to the reference strain: 3-MEBacid from L-Leu (0.021±0.0005 / 1.43-
fold decrease for the JMY2101-derivatives, 0.01±0.005 / 3-fold decrease for the JMY3150-derivatives,
and 0.03±0.003 for JMY2900; g/L±SD) and 2-MEPacid from L-Val (0.026±0.0044 / 3.88-fold decrease
for the JMY2101-derivatives, 0.043±0.0041 / 2.35-fold decrease for the JMY3150-derivatives,
0.101±0.0042 for JMY2900; g/L±SD). Which was also demonstrated by corresponding specific titer
values (Figure S1.D, H). Furthermore, we noted that for GGY0077 strains, the titers of 2-MEB (L-Ile
derivative) were increased upon supplementation with its “not direct” precursor BCAA - L-Leu
(0.85±0.043 / 2.36-fold increase for the JMY2101-derivatives, 0.54±0.07 / 1.5-fold increase for the
JMY3150-derivatives, and 0.36±0.076 for JMY2900; g/L±SD) (Figure 3.D). Both concentrations and
specific titers of the two isoforms: 2-MEB and 3-MEB in a course of L-Leu bioconversion were highly
corresponding (0.058 / 2.5-fold increase and 0.0637 / 2.5-fold increase in g of 3-MEB and
2-MEB/OD600 for the JMY2101-derivatives, 0.0657 / 2.6-fold increase and 0.0585 / 2.3-fold increase
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
2324
in g of 3-MEB and 2-MEB/OD600 for the JMY3150-derivatives, and 0.023 and 0.0259 g of 3-MEB and
2-MEB/OD600 for the reference; Figure S1.D). Moreover, it seems that ARO8 aminotransferase
presence / constitutive expression (GGY0077) does not impose such an unfavorable effect on the
BCAA-derived higher alcohols production, as it was observed for the BAT2 and 2-PE/2-PEAc synthesis.
It was rather observed, that the BCAA-derived higher alcohol formation could benefit from higher
abundance of the broad substrate specificity decarboxylase and the dehydrogenase. Upon
overexpression of the Ehrlich pathway with ARO8 aminotransferase in the G1 position (GGY0077), we
also observed an interplay between L-Phe and L-Val catabolic pathways. In general, provision of L-Val
promoted production of L-Phe derivatives - 2-PE and its derivative ester (which was more
pronounced in bat2Δ genotype background), when compared to L-Leu and L-Ile, while L-Phe
enhanced production of isobutyrate and 2-MEP (derivatives of L-Val) in these recombinant strains.
The second expression cassette GGA0097 differed from the GGA0077 by bearing BAT2
aminotransferase gene in the G1 position. Upon cultivation of GGA097-bearing strains in L-Leu-
supplemented medium, it was observed that BAT2 gene under native regulation is required for
maintaining growth phenotype corresponding to the reference strain, while lack of this gene resulted
in impaired growth (1.36-fold decrease in OD600 value; Figure S2.C). Surprisingly, such phenomenon
was not observed for the cultures supplemented with L-Ile and L-Val, where the recombinants’
growth remained at the native level by pTEF-BAT2 (Figure S2.E, G). The lower growth rate in the
JMY3150-derivatives was accompanied by slightly higher level of 3-MEB synthesis when expressed
per volume unit (peak concentration at 48 h 0.341±0.0343 vs 0.26±0.034 g/L±SD for JMY3150-
derivatives and the reference strain; 1.31-fold increase), which altogether resulted in importantly
increased specific titer (JMY3150-derivatives vs JMY2900: 0.103±0.026 and 0.028 of 3-MEB
g/OD600±SD at 48h; 3.67-fold increase; Figure 4.B, S2.D). On the other hand, for the JMY2101-
derivatives cultured in L-Leu-based medium the growth characteristic was maintained, still the peak
concentration of the derivative higher alcohol at 48 h was only slightly improved, when expressed in
g/L (0.32±0.025 vs 0.26±0.034 g/L±SD for JMY2101-derivatives and the reference strain; 1.23-fold
increase), but anyway improved specific titer (0.0377±0.0016 3-MEB g/OD600±SD at 48h; 1.34-fold
increase; Figure 4.B, S2.D). With respect to the accompanying metabolites, only 2-PE and 2-PEAc
were synthesized at detectable levels by GGY0097 strains in L-Leu-based cultures. While for the
JMY2101-derivatives, the increase in 2-PE and 2-PEAc formation was relatively low (2-PE:
0.079±0.014 vs 0.049 g/L±SD / 1.6-fold increase, 2-PEAc: 0.33±0.048 vs 0.154 g/L±SD / 2.14-fold
increase, for the recombinants vs the reference), in bat2Δ strains 2-PE and 2-PEAc formation was
strongly enhanced (more than 4-fold in both cases) up to 0.21±0.023 and 0.745±0.047 g/L±SD at 72 h
of culturing, even upon supplementation with L-Leu and not L-Phe (Figure 4.C); which was also
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
2526
reflected by significantly enhanced specific titers: 0.074±0.011 / 6.7-fold increase in 2-PEAc
g/OD600±SD at 72h, and 0.021±0.004 / 6-fold increase in 2-PE g/OD600±SD at 72h for
JMY3150_GGA0097 (Figure S2.D).
As mentioned above, constitutive expression of the Ehrlich pathway in GGA0097 variant,
reconstituted growth impairment observed for bat2Δ strains cultured in L-Ile and L-Val-containing
media. Upon cultivation with L-Ile, the GGA0097 modification increased production of 2-MEB, its
major bioconversion product, only in the JMY2101 strains (at 24 h: 0.35±0.02 vs 0.27±0.015 g/L±SD /
1.3-fold increase for JMY2101 derivatives vs the reference), while in the JMY3150-derivatives, its
production remained the reference level (Figure 4.D). In all strains, 2-MEB was converted to its
corresponding acetate ester, starting from approx. 48 h. In GGY0097 strains cultures with L-Ile, we
observed slightly increased titers of 3-MEBacid (derivative of L-Leu; 0.0132±0.002 / 1.4-fold increase
for the JMY2101 derivatives, 0.011±0.0008 / 1.16-fold increase for the JMY3150 derivatives, and
0.0095±0.0017 g/L±SD for JMY2900; Figure 4.E); while production of 2-PE was observed only in bat2Δ
strains, which was an unique quality. Surprisingly, none of GGY0097s differed significantly from
JMY2900 in production of major L-Val products (2-MEP and its ester) (Figure 4.F). However,
JMY3150_GGA0097 strains cultured with L-Val produced more 2-PE and 2-PEAc, when compared to
the JMY2900 reference strain (2-PEAc at 24 h: 1.163±0.003 / 1.43-fold increase for the JMY2101
derivatives, 0.96±0.08 / 1.18-fold increase for the JMY3150 derivatives, and 0.813±0.04 g/L±SD for
JMY2900; 2-PE at 48 h: 0.42±0.05 / 2-fold increase for the JMY2101 derivatives, 0.22±0.03 / 1.05-fold
for the JMY3150 derivatives, 0.21±0.05 g/L±SD for JMY2900; Figure 4.G). At the following time
points, the 2-PEAc concentration dropped from over 1 g/L to zero in the JMY3150 cultures, which
was not the case for the remaining strains. Moreover, the production of 2-PE was continued only in
the JMY3150-derivatives which again indicates important effect of bat2Δ on 2-PE formation and also
important relationship between L-Val and L-Phe pathways.
Discussion
The current study directly follows our previous findings concerning elucidation of putative
molecular identities involved in bioconversion of precursory AA to higher alcohols via Ehrlich
bioconversion pathway in Y. lipolytica (Celinska et al. 2015). Consequently, here we modified
previously identified elements of the Ehrlich pathway in Y. lipolytica cells by either its overexpression
under strong, constitutive promoter or knock-out, and investigated effects of such manipulations on
the yeast cell metabolism. To the best of our knowledge, this is the first study on engineering the
Ehrlich pathway in Y. lipolytica, as previous reports concerning characterization and manipulation of
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
2728
the Ehrlich pathway were mostly conducted in S. cerevisiae, with only few examples conducted with
other yeast (Hua and Xu 2011; Bolat et al. 2013; Kim et al. 2014b; Romagnoli et al. 2015; Stribny et al.
2016). Previous attempts of the Ehrlich pathway manipulation in S. cerevisiae mainly relied on single
gene deletion or overexpression which led to corresponding reduction or increase in the derivative
bioconversion products (2-MEB, 3-MEB, 3-MEBacid, 2-MEPacid) (Vuralhan et al. 2005; Lilly et al.
2006; Styger et al. 2013). More complex strategy relied on deletion of Ald3 gene, responsible for
phenylacetaldehyde oxidation, competing with 2-PE production, in a strain overexpressing ARO9 and
ARO10 and ARO80 transcription factor, resulted in increased formation of the derivative higher
alcohol (Kim et al. 2014a). Another interesting approach aimed at improvement of de novo
isobutanol production (2-MEP) by re-location of part of Ehrlich pathway to mitochondria
(decarboxylation and reduction) (Brat et al. 2012). Natively, the upstream de novo 2-MEP pathway is
located in mitochondria, while the downstream steps, executed by α-ketoacid decarboxylase and
alcohol dehydrogenase of Ehrlich pathway, are localized in cytoplasm. In order to secure availability
of intermediates without the need for transportation between the compartments, the researchers
attempted mitochondrial engineering. Targeting the entire pathway to mitochondria significantly
improved titer, yield and productivity of the higher alcohol, compared with a partly cytoplasmic
equivalent (Brat et al. 2012).
Due to possible differences in the experimental setups, direct comparison of own and
literature data concerning any bioprocess is not always fully justified. Nevertheless, such
comparisons allow to locate own data in relation to state-of-art. The aim of the present study was
manipulation with putative genes of Ehrlich pathway in Y. lipolytica, as identified in our previous
study, and examination of its effect on higher alcohols production. Thus, the bioconversion cultures
were conducted under suboptimal conditions in terms of amount of the precursory AA. The reason
behind such approach was to avoid a toxicity effect exerted by higher alcohols towards the producer
cells prior to observation of differences caused by the introduced modifications. Considering overall
titers of higher alcohols that were be obtained from the cultures, the highest levels of 2-PE were
obtained in L-Phe supplemented cultures of JMY3150_GGAA0077 and JMY2101_GGA0077 (> 0.7
g/L), followed by approx. 0.5 g/L produced by JMY3150_GGA0097. The highest levels of 2-PE’s
acetate ester (2-PEAc), which is highly valued aroma compound, were reached surprisingly in L-Val or
L-Leu supplemented cultures (and not L-Phe) of JMY3150_GGA0097 strain (> 1.0 and > 0.74 g/L).
Isobutanol (2-MEP) was accumulated at the highest concentration (0.347 g/L) in L-Val supplemented
cultures of JMY3150_GGA0097, while 2-MEB and its isoform (3-MEB) – in L-Leu supplemented
cultures of JMY2101_GGA0077 (0.85 and 0.75 g/L, respectively). As reported previously 2-MEP could
be accumulated up to 0.16 g/L by wild type S. cerevisiae, and up to 0.45 g/L by mutant strain, out of 2
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
2930
g/L of L-Leu supplemented into the culture medium (Abe and Horikoshi 2005). Its isomer – n-butanol,
could be accumulated from provided glycine via artificial pathway up to 0.092 g/L in engineered S.
cerevisiae (Branduardi et al. 2013). Industrial S. cerevisiae JP1 strain cultured in L-Leu supplemented
(at 9.8 g/L) bioreactor cultivations, produced up to 1.1 g/L of 3-MEB, together with 0.08 g/L of 2-MEP
and 0.05 g/L of 2-PE (Espinosa Vidal et al. 2015). Up to date, the highest levels of 2-MEP that have
been synthesized de novo (up to 0.63 g/L) were reported for recombinant S. cerevisiae strain having
part of Ehrlich pathway re-located to mitochondria (Brat et al. 2012).
In here designed expression cassettes, aminotransferase-encoding genes (ARO8 or BAT2)
were located in the first transcription unit. Depending on the assembly variant, either YALI0E20977g
or YALI0F19910g gene, encoding ARO8 or BAT2 aminotransferase homologs from S. cerevisiae,
respectively, was cloned in this position. ARO8 aminotransferase is mainly associated with
deamination of L-Phe, however, it has been experimentally shown that it possess relatively broad
substrate specificity, as it can act on L-Phe as the amino group donor with α-ketoisocaproate and α-
keto-γ(methylthio)butyrate, providing the connection node between L-Phe and L-Leu and L-Met
metabolic pathways (Urrestarazu et al. 1998; Iraqui et al. 1999). On the other hand, BAT2 is known to
be involved in both the first step of BCAA decomposition, as well as the last step of their synthesis (
BAT1 isozyme in mitochondria), while no earlier reports on its activity with aroma AA have been
published to date (according to www.uniprot.org; UniProtKB - P47176). In the current study we have
observed that while overexpression of the Ehrlich pathway in GGA0077 variant had significant
modulatory effect on several higher alcohols formation upon cultivation in L-Phe but also BCAA-
supplemented media (higher production of a respective higher alcohol and increase / decrease in the
fusel acid formation), the opposite setup (GGA0097 and L-Phe supplemented medium) had the
adverse effect on targeted higher alcohol production. These results well correspond with the
abovementioned findings on substrate specificity of the two aminotransferases – ARO8 and BAT2 in
S. cerevisiae. It seems unlikely that YALI0E20977p (ARO8) could directly act on BCAA, however, our
data suggest it is highly probable that it can interact with intermediates of BCAA catabolism, as it was
shown for S. cerevisiae ARO8. Interestingly, our current results indicate that overrepresentation of
BAT2 activity decreases production of aroma AA-derived metabolites (2-PE and 2-PEAc), while its
reduction (by generation of bat2Δ genotype) – leads to their higher formation. This observation
uniformly applied to all the analyzed experimental variants (different GGAs, BAT2+ or Δbat2
genotype, different AA fed to the culture media). Altogether it can be concluded that aroma AA and
BCAA pathways initiated by ARO8 and BAT2 aminotransferases, respectively, compete / act
synergistically in terms of sequestering amino groups from AA. This phenomenon can be seen as
competition, since it decreases titers of targeted higher alcohol and its ester, but considering the cell
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
3132
physiology, it dissipates the carbon skeletons and amino groups into different pathways, making the
process more efficient. Such dissipation of amino groups into different products has been earlier
observed in many studies on higher alcohols production (e.g. Espinosa Vidal et al. 2015; Celińska et
al. 2018). Moreover, it has been shown in S. cerevisiae that generation of Aro8Δ genotype underlies
the transcriptional upregulation of ARO10, which leads to increased de novo production of 2-PE
during growth on glucose and causes substantial changes in the absolute and relative intracellular
concentrations of AAs (Romagnoli et al. 2015). Our data suggest existence of similar phenomenon
operating between BAT2 aminotransferase and ARO10 decarboxylase (as seen in bat2Δ), however,
further research is need to support such conclusion.
As demonstrated previously, overexpression or deletion of BAT2 gene in S. cerevisiae
triggered respectively enhanced or decreased formation of derivative higher alcohols and acids from
their corresponding precursory BCAA (Lilly et al. 2006). In the present study, overexpression of BAT2
under control of strong pTEF promoter enhanced formation of 3-MEB from L-Leu in terms of specific
titer, as well as 3-MEBacid from L-Ile by both JMY2101- and JMY3150-derivatives. Interestingly, for
unbiased induction of 2-MEB formation from L-Ile, an additional BAT2 copy under native regulation
was necessary, since in JMY3150-derivatives bearing GGA0097 cassette the metabolite production
remained at the native level, while in JMY2101-derivatives a significant increase in 2-MEB formation
from L-Ile was observed. Importantly, the concentrations of the keto-acid (isovalerate) and the
downstream higher alcohol (2-MEB) were highly correlated in the GGA0097-bearing strains, cultured
under L-Ile provision. L-Ile promoted production of 2-PE solely in JMY3150-derived strains, which was
not observed for the reference and JMY2101-derived strains. Such “cross-induction” of the derivative
compounds synthesis by a non-direct precursors has been shown earlier for S. cerevisiae (Lilly et al.
2006; Espinosa Vidal et al. 2015). In a study by (Lilly et al. 2006), it was observed L-Val promotes
formation of 2-MEP and 2-MEPacid (straightforward bioconversion products), L-Leu and L-Ile are
both converted to 3-MEB and 3-MEBacid, and also, that L-Leu indirectly contributes to 2-PE
formation. It was indicated that supplementation of the growth media with individual BCAAs results
in increased levels of the corresponding higher alcohols and acids but also – the other, non-direct
bioconversion products (Lilly et al. 2006). Our current results are in agreement with those presented
by (Lilly et al. 2006), as we have observed that for example L-Val improved production of 2-PE and its
derivative ester in JMY3150_GGA077/GGA097 strains; L-Leu enhances production of 2-MEB (L-Ile
metabolism), especially in the case of JMY2101_GGA077 strains; L-Phe stimulates production of
isobutyrate (derivative of L-Val) in JMY2101-/JMY3150_GGA077 strains; ARO8 overexpression in
presence of native BAT2 (JMY2101) – strongly increases titers of 2-MEB (L-Ile derivative) in a course
of L-Leu bioconversion. As observed earlier in S. cerevisiae (Vuralhan et al. 2005), induction of an
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
3334
Ehrlich pathway for catabolism of a given AA leads to the formation of significant amounts of fusel
alcohols and acids from the other precursor, which implies that conversion of BCAA, aroma AA and
sulfur-containing AAs to the corresponding fusel alcohols and acids, via an Ehrlich pathway, involves
common broad-substrate-specificity enzyme activities (Vuralhan et al. 2003; Vuralhan et al. 2005;
Hazelwood et al. 2008). Such construction of the pathway enables flexible interchanging between the
individual sub-pathways. In our previous study (Celinska et al. 2015), we have observed high over-
representation of 2-oxoisovalerate dehydrogenase E1 (YALI0D08690g and YALI0F05038g; primarily
associated with L-Val), isovaleryl-CoA dehydrogenase (YALI0E12573g; primarily associated with L-Leu)
and 3-oxoacid CoA transferase (YALI0F26587g; primarily associated with L-Leu) as well as 3-
hydroxyisobutyryl-CoA hydrolase (YALI0D06215g; primarily associated with L-Val) upon provision of
L-Phe as the sole nitrogen source. This suggests that apart from tight connection between BCAAs
catabolic sub-pathways, such tight linkage functions also for the other, more structurally distinct
substrates metabolized via the Ehrlich pathway in Y. lipolytica, as well.
Although the crucial role in the Ehrlich bioconversion pathway is attributed to the initializing
aminotransferase (Romagnoli et al. 2015), the important contribution of the following keto acid
decarboxylase and dehydrogenase cannot be neglected (Romagnoli et al. 2012). Indeed, previous
studies with Kluyveromyces marxianus have shown that upon overexpression of ARO10 and the
following alcohol dehydrogenase (ADH2) genes from S. cerevisiae significantly improved de novo
synthesis of 2-PE (Kim et al. 2014b) In the present study we focused on a single gene, encoding
decarboxylase activity - YALI0D06930p (a homolog of ARO10 from S. cerevisiae), which was the only
protein, putatively catalyzing the keto acid decarboxylation reaction, identified in the previous
proteomic study (Celinska et al. 2015). As earlier proved in S. cerevisiae (Vuralhan et al. 2005), ARO10
is responsible for a broad- substrate-specificity decarboxylase activity involved in production of fusel
alcohols and acids, showing superior kinetic properties when compared with the other isozymes –
PDCs (Romagnoli et al. 2012). In S. cerevisiae, several different decarboxylases can operate in the
Ehrlich pathway. For L-Leu catabolism, KID1 decarboxylase acts with keto isocaproate , for L-Val – any
of the three isozymes of pyruvate decarboxylase encoded by PDC1, PDC5, and PDC6 can
decarboxylate α–ketoisovalerate, while the broadest range of isozymes can conduct this reaction
during L-Ile degradation, including PDC1, PDC5, PDC6, YDL080c, or YDR380w (Dickinson et al. 2003;
Hazelwood et al. 2008). Nevertheless, only ARO10-encoding gene’s expression was strongly up-
regulated in the presence of L-Phe, L-Leu and L-Met (the transcript level was at least 15-fold higher
than in cultures grown with ammonium sulfate), which coincided with induction of the Ehrlich
pathway. Such strong up-regulation was not observed for PDC5, PDC6, PDC1 and THI3. Since only
ARO10 was differentially and highly transcribed upon cultivation in different nitrogen sources, it was
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
3536
concluded that it operates as the most potent and flexible decarboxylase in the Ehrlich pathway
(Vuralhan et al. 2005). Moreover, overexpression of ARO10 in S. cerevisiae strain, in which the five
chromosomal decarboxylase genes had been deleted, was sufficient to restore a broad-substrate-
specificity decarboxylase activity. More research is needed to elucidate molecular bases of this step
of Ehrlich pathway in Y. lipolytica. It might be useful to identify specific decarboxylases in order to
more precisely manipulate with the Ehrlich pathway, to ultimately obtain narrowed spectrum of
metabolites. On the other hand, it might be challenging to reach more “focused” Ehrlich pathway, as
it is known that the individual steps are highly redundant, the enzymatic activities have broad-
substrate specificity, and the intermediates are shared between different branches, as for example α-
oxoisovalerate (L-Val metabolism) which is known, shared intermediate of BCAAs turnover.
It was previously shown that 2-MEP, 3-MEB, 2-MEB can all be re-metabolized by S. cerevisiae
cells in a specific culturing conditions, namely stationary phase, after exhaustion of glucose, ethanol
or acetate carbon sources in long-term batch cultures, using minimal medium with glucose as the
initial carbon source and L-Leu as the sole nitrogen source (Vuralhan et al. 2005). As neither growth
nor ester formation were observed concomitantly with depletion of the higher alcohols, the authors
concluded that this re-metabolisation might play a role in the maintenance of the redox balance
through reduction of NAD+, while oxidizing the higher alcohols into their corresponding aldehydes in
the opposite direction to the final reductive step of the Ehrlich pathway, and could potentially be
used as carbon sources after further oxidation to the corresponding acid and assimilation by β-
oxidation. Our cultures were conducted until the major carbon source was exhausted from the media
(72 h), and we have not observed such a phenomenon. Probably, prolonged cultivations could
answer whether such a phenomenon occurs also in Y. lipolytica cells. However, we observed that 2-
MEB (GGA097 L-Ile), 2-MEP (GGA097 L-Val) are converted to its acetate ester which production is
correlated with decrease in the alcohol concentration. Such correlation was not observed for 2-MEB
(GGA077, cultured with L-Ile), 2-MEP (GGA077, cultures with L-Val), and also 2-MEP (both GGAs, L-
Leu) where we determined just presence of the corresponding acetate ester, but not concerted with
decrease in the alcohol concentration.
Our current data also demonstrate that, similarly to S. cerevisiae, in Y. lipolytica the AA
precursors can be partly transformed into the corresponding fusel acid. It was earlier suggested that
the ratio between the final concentrations of synthesized acid or alcohol depends on the redox state
of the cell (Dickinson and Dawes 1992). Interestingly, we observed that overexpression of the Ehrlich
pathway in GGA077 variant and cultivation of the resulting strains in L-Leu, L-Ile and L-Val-containing
media led to significantly decreased synthesis of the fusel acid, when compared to the reference
strain. In the view of former findings in S. cerevisiae it might indicate favorable change in the redox
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
3738
state of the cell followed by the modification or, overexpression of dehydrogenase YALI0F24937p
with high preference for reduction reaction and low reversibility.
Conclusions and Perspectives
In summary, the present study reports effects of amplification of putative Ehrlich pathway in
Y. lipolytica on production of valuable higher alcohols. Here adopted modification strategy
contributed to improved production of targeted higher alcohol in many cases, but also enabled
observation of several interesting phenomena. First of all, we observed that bat2Δ genotype
promotes formation of aromatic higher alcohol and its derivative acetate ester, suggesting either
competition between the aroma AA- and BCAA-initiated pathways in sequestering ammonium
residues, or existence of regulation mechanism between aminotransferase and decarboxylase genes,
as it was observed for S. cerevisiae. Correspondingly, doubling BAT2 gene (JMY2101_GGA0097)
decreases formation of 2-PE from L-Phe. Moreover, we observed an interplay between L-Val and L-
Phe pathways, especially in GGA0077-bearing strains (cross-induction of derivative higher alcohols).
Interestingly, upon overexpression of GGA0077 cassette genes, we observed enhanced formation
higher alcohols deriving from L-Leu and L-Val, and concomitantly decreases fusel acid production.
This study deals with primary engineering of aroma AA and BCAA three-step bioconversion
pathway, ultimately leading to many valuable higher alcohols. Obviously, production of higher
alcohols from pure AA is not economically feasible. Still, as demonstrated by several studies
(discussed above) application of metabolic engineering strategies may successfully redirect necessary
carbon skeletons from central metabolism to be further converted by the Ehrlich pathway activities.
On the other hand, to add to sustainability of higher alcohols production from AA, a protein-rich raw
materials like e.g. whey, which is not completely absorbed by the market, could serve as the cost-
effective feedstock. Such solution would definitely constitute an added value of the higher alcohols
production process, but also would complement the biorefinery concept.
Acknowledgements
This study was co-funded by the European Union via the European Regional Development Fund
within the framework of the Innovative Economy Operational Program 2007–2013 (No. 01.01.02-00-
074/09), and was also financially supported by the Ministry of Science and Higher Education Project
entitled “Exploitation of biotechnological processes in production of safe food commodities and
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
3940
chemical compounds” (No. 508.864.00). We thank to Eng. Martyna Przybylak for her technical
assistance.
References
Abe F, Horikoshi K (2005) Enhanced production of isoamyl alcohol and isoamyl acetate by
ubiquitination-deficient Saccharomyces cerevisiae mutants. Cell Mol Biol Lett 10:383–388.
Alper H, Stephanopoulos G (2009) Engineering for biofuels: exploiting innate microbial capacity or
importing biosynthetic potential? Nat Rev Microbiol 7:715–723. doi: 10.1038/nrmicro2186
Avalos JL, Fink GR, Stephanopoulos G (2013) Compartmentalization of metabolic pathways in yeast
mitochondria improves the production of branched-chain alcohols. Nat Biotechnol 31:335–341.
doi: 10.1038/nbt.2509
Barth G, Gaillardin C (1996) Yarrowia lipolytica. In: Wolf K (ed) Nonconventional Yeasts in
Biotechnology A Handbook. Springer Berlin Heidelberg, pp 313–388
Białas W, Czerniak A, Dobrowolska A, Wojciechowska J, Grajek W (2015) Improvement of microbial
stability of innovative dietary supplement based on the biomass of Yarrowia lipolytica. Prog Res
Technol Agro-Food Ind 70:5–20.
Bolat I, Romagnoli G, Zhu F, Pronk JT, Daran JM (2013) Functional analysis and transcriptional
regulation of two orthologs of ARO10, encoding broad-substrate-specificity 2-oxo-acid
decarboxylases, in the brewing yeast Saccharomyces pastorianus CBS1483. FEMS Yeast Res
13:505–517. doi: 10.1111/1567-1364.12051
Borodina I, Nielsen J (2014) Advances in metabolic engineering of yeast Saccharomyces cerevisiae for
production of chemicals. Biotechnol J 9:609–620. doi: 10.1002/biot.201300445
Branduardi P, Longo V, Berterame NM, Rossi G, Porro D (2013) A novel pathway to produce butanol
and isobutanol in S. cerevisiae. Biotechnol Biofuels 6:68.
Brat D, Weber C, Lorenzen W, Bode HB, Boles E (2012) Cytosolic re-localization and optimization of
valine synthesis and catabolism enables inseased isobutanol production with the yeast
Saccharomyces cerevisiae. Biotechnol Biofuels 5:65. doi: 10.1186/1754-6834-5-65
Celińska E, Bonikowski R, Białas W, Dobrowolska A, Słoma B, Borkowska M, Kubiak M, Korpys P,
Grajek W (2018) Pichia cactophila and Kluyveromyces lactis are Highly Efficient Microbial Cell
Factories of Natural Amino Acid-Derived Aroma Compounds. Molecules. doi:
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
4142
10.3390/molecules23010097
Celińska E, Kubiak P, Białas W, Dziadas M, Grajek W (2013) Yarrowia lipolytica: The novel and
promising 2-phenylethanol producer. J Ind Microbiol Biotechnol 40:389–392. doi:
10.1007/s10295-013-1240-3
Celinska E, Ledesma-Amaro R, Larroude M, Rossignol T, Pauthenier C, Nicaud JM (2017) Golden Gate
Assembly system dedicated to complex pathway manipulation in Yarrowia lipolytica. Microb
Biotechnol 10:450–455. doi: 10.1111/1751-7915.12605
Celinska E, Olkowicz M, Grajek W (2015) L-Phenylalanine catabolism and 2-phenylethanol synthesis
in Yarrowia lipolytica-mapping molecular identities through whole-proteome quantitative mass
spectrometry analysis. FEMS Yeast Res. doi: 10.1093/femsyr/fov041
Dai Z, Nielsen J (2015) Advancing metabolic engineering through systems biology of industrial
microorganisms. Curr Opin Biotechnol 36:8–15. doi: 10.1016/j.copbio.2015.08.006
Darvishi Harzevili F (2014) Biotechnological Applications of the Yeast Yarrowia lipolytica. Springer,
Cham
Dickinson JR, Dawes IW (1992) The catabolism of branched-chain amino acids occurs via 2-oxoacid
dehydrogenase in Saccharomyces cerevisiae. J Gen Microbiol 138:2029–2033. doi:
10.1099/00221287-138-10-2029
Dickinson JR, Salgado LEJ, Hewlins MJE (2003) The catabolism of amino acids to long chain and
complex alcohols in Saccharomyces cerevisiae. J Biol Chem 278:8028–8034. doi:
10.1074/jbc.M211914200
Ehrlich F (1907) Über die Bedingungen der Fuselölbildung und über ihren Zusammenhang mit dem
Eiweißaufbau der Hefe. Berichte der Dtsch Chem Gesellschaft 40:1027–1047. doi:
10.1002/cber.190704001156
Espinosa Vidal E, de Morais MA, François JM, de Billerbeck GM (2015) Biosynthesis of higher alcohol
flavour compounds by the yeast Saccharomyces cerevisiae: Impact of oxygen availability and
responses to glucose pulse in minimal growth medium with leucine as sole nitrogen source.
Yeast 32:47–56. doi: 10.1002/yea.3045
Forti L, Di Mauro S, Cramarossa MR, Filippucci S, Turchetti B, Buzzini P (2015) Non-Conventional
Yeasts Whole Cells as Efficient Biocatalysts for the Production of Flavors and Fragrances.
Molecules 20:10377–10398. doi: 10.3390/molecules200610377
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
4344
Generoso WC, Schadeweg V, Oreb M, Boles E (2015) Metabolic engineering of Saccharomyces
cerevisiae for production of butanol isomers. Curr. Opin. Biotechnol. 33:1–7.
Groenewald M, Boekhout T, Neuvéglise C, Gaillardin C, Van Dijck PWM, Wyss M (2014) Yarrowia
lipolytica: safety assessment of an oleaginous yeast with a great industrial potential. Crit Rev
Microbiol 40:187–206. doi: 10.3109/1040841X.2013.770386
Grondin E, Shum A, Sing C, Caro Y, Raherimandimby M, Lalaniaina A, James S, Nueno-palop C, Marie
J, Petit T (2015) A comparative study on the potential of epiphytic yeasts isolated from tropical
fruits to produce flavoring compounds. Int J Food Microbiol 203:101–108. doi:
10.1016/j.ijfoodmicro.2015.02.032
Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for
Windows 95/98/NT. Nucleic Acids Symp. Ser. 41:95–98.
Hazelwood LA, Daran J-M, van Maris AJA, Pronk JT, Dickinson JR (2008) The Ehrlich pathway for fusel
alcohol production: a century of research on Saccharomyces cerevisiae metabolism. Appl
Environ Microbiol 74:2259–2266. doi: 10.1128/AEM.02625-07
Hua D, Xu P (2011) Recent advances in biotechnological production of 2-phenylethanol. Biotechnol
Adv 29:654–660. doi: 10.1016/j.biotechadv.2011.05.001
Iraqui I, Vissers S, André B, Urrestarazu A (1999) Transcriptional induction by aromatic amino acids in
Saccharomyces cerevisiae. Mol Cell Biol 19:3360–3371.
Jansen MLA, van Gulik WM (2014) Towards large scale fermentative production of succinic acid. Curr
Opin Biotechnol 30:190–197. doi: 10.1016/j.copbio.2014.07.003
Kim B, Cho BR, Hahn JS (2014a) Metabolic engineering of Saccharomyces cerevisiae for the
production of 2-phenylethanol via Ehrlich pathway. Biotechnol Bioeng 111:115–124. doi:
10.1002/bit.24993
Kim TY, Lee SW, Oh MK (2014b) Biosynthesis of 2-phenylethanol from glucose with genetically
engineered Kluyveromyces marxianus. Enzyme Microb Technol 61–62:44–47. doi:
10.1016/j.enzmictec.2014.04.011
Ledesma-Amaro R, Nicaud JM (2016) Yarrowia lipolytica as a biotechnological chassis to produce
usual and unusual fatty acids. Prog Lipid Res 61:40–50. doi: 10.1016/j.plipres.2015.12.001
Lilly M, Bauer FF, Styger G, Lambrechts MG, Pretorius IS (2006) The effect of increased branched-
chain amino acid transaminase activity in yeast on the production of higher alcohols and on the
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
4546
flavour profiles of wine and distillates. FEMS Yeast Res 6:726–743. doi: 10.1111/j.1567-
1364.2006.00057.x
Lilly M, Lambrechts MG, Pretorius IS (2000) Effect of increased yeast alcohol acetyltransferase
activity on flavor profiles of wine and distillates. Appl Environ Microbiol 66:744–753. doi:
10.1128/AEM.66.2.744-753.2000
Madzak C (2015) Yarrowia lipolytica: recent achievements in heterologous protein expression and
pathway engineering. Appl Microbiol Biotechnol 99:4559–4577. doi: 10.1007/s00253-015-6624-
z
Nakamura CE, Whited GM (2003) Metabolic engineering for the microbial production of 1,3-
propanediol. Curr. Opin. Biotechnol. 14:454–459.
Nielsen J, Keasling JD (2016) Engineering Cellular Metabolism. Cell 164:1185–1197. doi:
10.1016/j.cell.2016.02.004
Nozzi NE, Desai SH, Case AE, Atsumi S (2014) Metabolic engineering for higher alcohol production.
Metab Eng 25:174–182. doi: https://doi.org/10.1016/j.ymben.2014.07.007
Romagnoli G, Knijnenburg TA, Liti G, Louis EJ, Pronk JT, Daran JM (2015) Deletion of the
Saccharomyces cerevisiae ARO8 gene, encoding an aromatic amino acid transaminase,
enhances phenylethanol production from glucose. Yeast 32:29–45. doi: 10.1002/yea.3015
Romagnoli G, Luttik MAH, Kötter P, Pronk JT, Daran JM (2012) Substrate specificity of thiamine
pyrophosphate-dependent 2-oxo-acid decarboxylases in Saccharomyces cerevisiae. Appl
Environ Microbiol 78:7538–7548. doi: 10.1128/AEM.01675-12
Sambrook J, Russell D (2001) Molecular Cloning: A Laboratory Manual, 3rd Editio. Cold Spring Harbor
Laboratory Press
Sauer M, Marx H, Mattanovich D (2008) Microbial Production of 1,3-Propanediol. Recent Pat
Biotechnol 2:191–197.
Schrader J, Etschmann MM, Sell D, Hilmer JM, Rabenhorst J (2004) Applied biocatalysis for the
synthesis of natural flavour compounds--current industrial processes and future prospects.
Biotechnol Lett 26:463–472. doi: 10.1023/B:BILE.0000019576.80594.0e
Sibirny A, Scheffers L (2002) Thematic section: Biochemistry, Genetics, Biotechnology and Ecology of
Non-conventional Yeasts. FEMS Yeast Res. doi:
http://onlinelibrary.wiley.com/doi/10.1111/j.1567-1364.2002.tb00097.x/pdf
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
4748
Stribny J, Romagnoli G, Pérez-Torrado R, Daran J-M, Querol A (2016) Characterisation of the broad
substrate specificity 2-keto acid decarboxylase Aro10p of Saccharomyces kudriavzevii and its
implication in aroma development. Microb Cell Fact 15:51. doi: 10.1186/s12934-016-0449-z
Styger G, Jacobson D, Prior BA, Bauer FF (2013) Genetic analysis of the metabolic pathways
responsible for aroma metabolite production by Saccharomyces cerevisiae. Appl Microbiol
Biotechnol 97:4429–4442. doi: 10.1007/s00253-012-4522-1
Urrestarazu A, Vissers S, Iraqui I, Grenson M (1998) Phenylalanine- and tyrosine-auxotrophic mutants
of Saccharomyces cerevisiae impaired in transamination. Mol Gen Genet 257:230–237.
Vuralhan Z, Luttik MA, Tai SL, Boer VM, Morais MA, Schipper D, Almering MJ, Kotter P, Dickinson JR,
Daran JM, Pronk JT (2005) Physiological characterization of the ARO10-dependent, broad-
substrate-specificity 2-oxo acid decarboxylase activity of Saccharomyces cerevisiae. Appl
Environ Microbiol 71:3276–3284. doi: 10.1128/AEM.71.6.3276-3284.2005
Vuralhan Z, Morais MA, Tai SL, Piper MDW, Pronk JT (2003) Identification and characterization of
phenylpyruvate decarboxylase genes in Saccharomyces cerevisiae. Appl Environ Microbiol
69:4534–4541. doi: 10.1128/AEM.69.8.4534-4541.2003
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
4950
Figure Captions:
Figure 1. The Ehrlich pathway. Metabolic pathway of branched chain AAs (L-Val, L-Leu, L-Ileu),
aromatic AAs (L-Phe, L-Tyr, L-Trp) and L-Met catabolism initialized by transamination by
aminotransferase to alpha-keto acid, followed by decarboxylation by decarboxylase to fusel
aldehyde, and either oxidation or reduction by dehydrogenase to fusel alcohol or fusel acid,
respectively. CAS numbers are provided for those compounds that were analyzed in this study.
Figure 2. Y. lipolytica recombinant strains construction system. A.B. Golden Gate Assembly platform
dedicated to Y. lipolytica with three transcription units, bearing homologous genes encoding
activities involved in higher alcohols synthesis pathway: ARO8 [YALI0E20977p] /BAT2 [YALI0F19910p]
– aminotransferases; ARO10 [YALI0D06930p] – decarboxylase; Adh [YALI0F24937p]- dehydrogenase.
A. Complete set of donor vectors constructed on a backbone of pCR Blunt II-TOPO vector (Thermo‐
Scientific) bearing kanamycin resistance gene (KanR). Each building block is flanked with respective
BsaI recognition sites and predesigned 4 nt overhangs with signatures A–M. The overhangs allow to
introduce all the elements of the expression cassette in the correct position and orientation.
Abbreviations: InsUP/InsDOWN – sequences targeting insertion in the 5'/3' region (zeta); M –
selection marker (URA3); P1/2/3– constitutive promoters (pTEF) governing expression of the gene of
interest in transcription unit 1, 2 and 3, respectively; G1/2/3 – gene of interest in transcription unit 1,
2 and 3 respectively; T1/2/3– constitutive terminators (tLip2) located in the transcription unit 1, 2
and 3 respectively. B. Complete GGVAs (Golden Gate Assembles in destination vector pSB1A3-RFP
from iGEM collection) in 3-transcription unit (TU)-bearing format. Each TU is composed of a
promoter – pTEF, a gene of interest: BAT2/ARO8/ARO10/Adh, and a terminator – tLip2); Complete
GGVA bears a selection marker (URA3) and is flanked with integration targeting sequences (InsUP
and InsDOWN); signatures A to M and the following nucleotide sequences mark position and
sequence of the 4 nt overhangs, mentioned in part A; NotI: NotI recognition sites for release of the
expression cassettes prior to Y. lipolytica transformation. C. Electrophoretic separation of amplicones
obtained in colony PCR with biomass Y. lipolytica recombinant or parental strains, corresponding to
regions: ARO8-tLip2T1 (lanes: 1, 5, 9, 13, 17), ARO10-tLip2T2 (lanes: 2, 6, 10, 14, 18), Adh-tLip2T3
(lanes: 3, 7, 11, 15, 19) and ura3 (lanes: 4, 8, 12, 16, 20); lanes 1-4 (clone 1), 5-8 (clone 2), 9-12 (clone
3), 13-16 (clone 4), 17-20 (negative control – not transformed strain). Positive clones: 1 and 3. M:
Gene Ruler Express.
Figure 3. Metabolite profiles generated by GGY0077-bering Y. lipolytica strains in bioconversion
cultures supplemented with L-Phe (A.B), L-Leu (C.D.), L-Ile (E.F), and L-Val (G.H). The graphs represent
either kinetic (A.C.E.G) or end-point (B.D.F.H.) data. Legends indicate the type of detected metabolite
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
5152
and type of strain. X axis: time of culturing in h (A.C.E.G) or type of strain (B.D.F.H.). Y axis:
concentration of a given metabolite in g/L. Mind main and auxiliary Y axes. Error bars indicate mean
values from biological duplicate, each analyzed in triplicate.
Figure 4. Metabolite profiles generated by GGY0097-bering Y. lipolytica strains in bioconversion
cultures supplemented with L-Phe (A.), L-Leu (B.C.), L-Ile (D.E.), and L-Val (F.G.). The graphs represent
either kinetic (A.B.D.F.G) or end-point (C.E.) data. Legends indicate the type of detected metabolite
and type of strain. X axis: time of culturing in h (A.B.D.F.G) or type of strain (C.E.). Y axis:
concentration of a given metabolite in g/L. Mind main and auxiliary Y axes. Error bars indicate mean
values from biological duplicate, each analyzed in triplicate.
Supplementary Material
Figure S1. Biomass growth and specific titers of metabolites generated by GGY0077-bering Y.
lipolytica strains in bioconversion cultures supplemented with L-Phe (A.B), L-Leu (C.D.), L-Ile (E.F), and
L-Val (G.H). The graphs represent either kinetic (A.C.E.G) or single-point (B.D.F.H.) data. Legends
indicate the culturing time or type of detected metabolite. X axis: type of strain. Y axis: absorbance at
600 nm in OD units (A.C.E.G) or specific titer in g/OD600 (B.D.F.H.). Mind main and auxiliary Y axes.
Error bars indicate mean values from biological duplicate, each analyzed in triplicate.
Figure S2. Biomass growth and specific titers of metabolites generated by GGY0097-bering Y.
lipolytica strains in bioconversion cultures supplemented with L-Phe (A.B), L-Leu (C.D.), L-Ile (E.F), and
L-Val (G.H). The graphs represent either kinetic (A.C.E.G) or single-point (B.D.F.H.) data. Legends
indicate the culturing time or type of detected metabolite. X axis: type of strain. Y axis: absorbance at
600 nm in OD units (A.C.E.G) or specific titer in g/OD600 (B.D.F.H.). Mind main and auxiliary Y axes.
Error bars indicate mean values from biological duplicate, each analyzed in triplicate.
Table S1. List of E. coli and Y. lipolytica strains used in this study
Table S2. List of plasmids and primers used in this study
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
5354