Thiazolidine Peracetates: Carbohydrate Derivatives that Readily Assign cis-, trans-2,3-Monosaccharides by Gas Chromatography-Mass Spectrometry Analysis

Article abstract of DOI:10.1021/acs.analchem.8b00976

A novel group of carbohydrate derivatives is described that uniquely assign cis/trans-2,3-aldose stereoisomers at low nanomolar concentrations. Aldopentoses, aldohexoses, or component aldoses from hydrolysis of polysaccharides or oligosaccharides react with cysteamine in pyridine to give quantitative formation of thiazolidines, which are subsequently peracetylated in a one-pot reaction. The nonpolar thiazolidines peracetate (TPA) derivatives are analyzed by gas chromatography and electron impact mass spectrometry (GC/EI-MS), each aldose giving rise to two TPA geometric isomers. The quantitative ratio of these diastereomers is dependent upon whether the parent monosaccharide is cis-2,3-(Rib, Lyx, Man, All, Gul, and Tal), or trans-2,3-aldose (Xyl, Ara, Glc, Gal, Ido, and Alt). TPAs generate observed EI-MS fragment ions characteristic of C1-C2 and C3-C4 bond cleavage of the parent sugars. This has been used to estimate the extent of metabolic labeling of microbial cell-wall carbohydrates, especially into the defining anomeric carbons and during aldolase / ketolase -catalyzed rearrangements.

Full text of DOI:10.1021/acs.analchem.8b00976

Subscriber access provided by Kaohsiung Medical University
Article
Thiazolidine Peracetates: Carbohydrate Derivatives that Readily
Assign cis-, trans-2,3- Monosaccharides by GC/MS analysis.
Neil P.J. Price, Trina M. Hartman, and Karl E. Vermillion
Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b00976 • Publication Date (Web): 08 Jun 2018
Downloaded from http://pubs.acs.org on June 11, 2018
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 9
Analytical Chemistry
1
2
3
4
5
6
7
Thiazolidine Peracetates: Carbohydrate Derivatives that Readily As-
8
9
sign cis-, trans-2,3- Monosaccharides by GC/MS Analysis.
Neil. P. J. Price,1* Trina M. Hartman,1 Karl E. Vermillion.2
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
U.S. Department of Agriculture, National Center for Agricultural Utilization Research, 1Renewable Product Tech-
nology and 2Functional Foods Research Units, 1815 N. University St., Peoria, IL 61604, USA.
ABSTRACT: A novel group of carbohydrate derivatives is described that uniquely assign cis/trans-2,3 aldose stereoiso-
mers at low nanomolar concentrations. Aldopentoses or aldohexoses, or component aldoses from hydrolysis of polysac-
charides or oligosaccharides, react with cysteamine in pyridine to give quantitative formation of thiazolidines, which are
subsequently peracetylated in a one-pot reaction. The non-polar thiazolidines peracetate (TPA) derivatives are analyzed
by GC/EI-MS, each aldose giving rise to two TPA geometric isomers. The quantitative ratio of these diastereomers is de-
pendent upon whether the parent monosaccharide is cis-2,3- (Rib, Lyx, Man, All, Gul, and Tal), or trans-2,3-aldose (Xyl,
Ara, Glc, Gal, Ido, and Alt). TPAs generate observed EI-MS fragments ions characteristic of C1 – C2 and C3 – C4 bond
cleavage of the parent sugars. This has been used to estimate the extent of metabolic labelling of microbial cell-wall car-
bohydrates, especially into the defining anomeric carbons and during aldolase/ketolase-catalyzed rearrangements.
sugars.8 We have previous described dialkyldithioacetal
acetate (DADTA) derivatives that have this desirable
INTRODUCTION
property, and which generate well-resolved base ion
The compositional determination of glycoconju-
peaks arising from C1-C2 bond cleavage of the parent
gates, polysaccharides, or oligosaccharides at low concen-
monosaccharide with charge retention at the C1 thiol
trations generally relies on acid hydrolysis to release the
groups.8 However, we also noted that the stability of the
component monosaccharides followed by derivativation
carbohydrate DADTA derivatives is poor and not suitable
and analysis by gas chromatography-mass spectrometry
for prolonged storage, and that the thioalkane reagents
(GC/MS).1-4 The chemical derivatization of the compo-
needed (for example, ethanethiol) are both volatile and
nent monosaccharides is to provide non-polar com-
highly odorous to work with in the laboratory.8
pounds with low volatility for improved chromatographic
properties. Three types of derivatives are currently in
In the present work we describe a new type of
general usage: alditol peracetates (AP),5 aldononitrile
carbohydrate derivatives that are suitable for analysis by
peracetates (PAAN),6 and trimethylsilyl methyl glycosides
GC/MS. Reaction of free aldoses (pentoses or hexoses), or
(TMS-MG).7 These each have various advantages, and are
the component aldoses arising from acid hydrolysis of
often used in combination, and are readily detected by
polysaccharides or oligosaccharides, with excess cysteam-
GC (for example, with flame ionization detection) or by
ine hydrochloride in pyridine resulted in the quantitative
formation of thiazolidines (Scheme 1).
GC/MS.
The electron impact mass spectral (EI-MS) data
obtained by GC/MS is useful to distinguish the monosac-
charide derivatives from potential non-carbohydrate
components or contaminants, but neither this nor colli-
sion-induced dissociation mass spectrometry (CID-MS)
provides much additional structural information. This is
because the monosaccharides are geometric isomers of
each other, and generally difficult to distinguish other
than by chromatographic retention times (requiring
known standards) or by resorting to less sensitive tech-
niques such as nuclear magnetic resonance (NMR) spec-
troscopy. In addition, the EI-MS fragmentation pathways
Scheme 1. One-pot preparation of thiazolidine peracetates
for AP, PAAN, and TMS-MG derivatives occurs with ioni-
zation at the non-reducing end of the parent monosac-
charides, and therefore provides little information about
the characteristic anomeric carbon (carbon C1) of the
(TPA) from aldose sugars.
These are readily peracetylated in a one-pot reaction with
acetic anhydride using the pyridine solvent from the prior
ACS Paragon Plus Environment

Analytical Chemistry
Page 2 of 9
Page 2 of 9
reaction as the base catalyst. Following water:ethyl ace-
tate partitioning the non-polar thiazolidines acetate
(TPA) derivatives are submitted to analysis by GC/EI-MS.
Each aldose gives rise to two TPA geometric isomers that
are readily separated by the GC chromatography. We re-
port that the quantitative ratio of these two isomers is
dependent upon whether the parent monosaccharide is a
cis-2,3- (Rib, Lyx, Man, All, Gul, and Tal), or a trans-2,3-
aldose (Xyl, Ara, Glc, Gal, Ido, and Alt). Hence, this
method is able to distinguish cis/trans-2/3 aldose isomers
at low concentration by GC/MS analysis alone. In addi-
tion, the high stability of the base ions generated by
GC/EI-MS (m/z 130 and 88) are characteristic for these
derivatives, and like the previously described DADTA
arise from C1-C2 bond cleavage with charge retention at
the C1 anomeric carbon of the parent sugar. We therefore
also demonstrate the potential for using this selective
ionization pathway of the new TPA derivatives for meta-
bolic flux analysis following incorporation of stable iso-
the estimates include the serial dilutions, TPA deriva-
tivation reactions, and solvent extraction steps, as well as
the GC response. Both TPA derivatives were linear over
0.4 – 4.0 nmole range with R-squared regression values of
0.9999 (Glc-TPA) and 0.9935 (Ara-TPA). At concentra-
tions less than 0.4 µmoles.mL-1 (0.2 nmoles injected) the
regression values dropped to 0.9989 and 0.9772, respec-
tively, indicating a minimal detection limit of 0.4 nmoles
injected onto the column. We also note that these quan-
titative lower limit values were similar to those obtained
for the GC analysis of Glc and Ara as either alditol acetate
or aldononitrile acetate derivatives.8
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
(ii)
Gas
Chromatography/Mass
Spectrometry
(GC/MS) Assignment of the 2,3-Stereochemistry of
Monosaccharides. All of the eight neutral hexose and
the four pentose aldose sugars were converted into TPAs
and analyzed by GC/MS (Table 1). As expected, each
monosaccharide gives rise to two GC peaks corresponding
to the 1S and 1R stereoisomers of the thiazolidine ring
(Fig. 1).
13
tope C-labeled metabolites into the anomeric carbon of
multiple aldose metabolic end-products. The new TPA
derivatization procedure is demonstrated for 32 neutral
aldoses and aminosugars, and is applied for the analysis of
four complex oligosaccharide and polysaccharide sugars.
RESULTS AND DISCUSSION
(i) Preparation and Analysis of Carbohydrate Thia-
zolidine Peracetates. The thiazolidine peracetate deriv-
atives (TPAs) were prepared from aldoses as described
previously.9 Carbohydrate samples were dissolved in a
small quantity (typically 1 mL) of saturated cysteamine
hydrochloride dissolved in pyridine. The thiazolidines
Fig. 1. Chromatographs of cis-2,3 (Tal, Lyx) and trans-
2,3 (Gal, Ara) TPAs. A/B peak height ratios are small
for cis-2,3 and large for trans-2,3 TPA derivatives,
where A and B is the order of elution from the GC.
o
are formed on heating (60 C, 15 min), and subsequently
peracetylated in situ by adding an equal volume of acetic
anhydride and reheating. The excess reagents are re-
moved by evaporation, and the TPAs are re-dissolved in
ethyl acetate. An important aspect of the thiazolidines is
that they are derived from the open chain form of the
original sugars, and therefore the analysis is not compli-
cated by pyranosyl or furanosyl cyclic forms. The
peracetylation also has several advantages. The thiazoli-
dine reaction is reversible via a Schiff’s base ring opening
of the thiazolidine group,9-11 but once N-acetylated on the
thiazolidine amino group the reverse reaction is preclud-
ed. The peracetylation step also allows a direct assess-
ment of the completeness of the thiazolidine reaction,
since any unreacted sugars should be immediately appar-
ent in the GC/MS analysis as peracetylated sugars. This
provides additional confidence that the conversion of the
sugars to TPA is indeed quantitative.
Unexpected, these two isomers were not produced in
equal amounts, but rather one peak was generally larger
than the other (Fig.1, Table 1). Examination of all of the
sugar TPAs showed a trend, where the earlier eluting GC
peak (Peak A) is greater for trans-2,3 monosaccharide
TPAs, and smaller for cis-2,3 TPAs (Table 1). Hence, it is
possible to distinguish the 2,3-stereochemistry of the
TPAs (Scheme 2) by GC analysis alone, irrespective of the
retention times.
To further assess the quantitative linearity of the
described GC method, and to determine the minimum
detection limit, a representative hexoaldose (D-glucose)
and pentoaldose (D-arabinose) were evaluated over the
concentration range 0.2 – 2.0 nmoles injected (supple-
mentary Fig. S.1). These assessments were initiated with
stock dilutions of the sugars (0.4 – 4.0 µmoles.mL-1) and
Scheme. 2. 1S, 1R stereoisomers of cis-2,3 and trans-2,3
carbohydrate TPAs.
ACS Paragon Plus Environment

Page 3 of 9
Analytical Chemistry
Page 3 of 9
To evaluate this more closely, we analyzed glyceralde-
hyde, erythrose, three 6-deoxy sugars (fucose, rhamnose,
and quinovose), two 2-deoxy sugars, two mono-methoxy
sugars, and six amino sugars (Table 1). Erythrose and the
6-
1
2
3
4
5
6
7
8
Table 1. Relative retention times and peak height data for cis- and trans-2,3 TPAs.
Peak Height
Peak Retention Time (min)
Monosaccharide TPA
2,3 - cis/trans
Peak A*
Peak B
Pattern
Peak A
Peak B
Difference
Xylose TPA
Arabinose TPA
Ribose TPA
Lyxose TPA
Mannose TPA
Glucose TPA
Galactose TPA
Allose TPA
Gulose TPA
Talose TPA
Altrose TPA
trans
trans
cis
cis
cis
trans
trans
cis
cis
cis
trans
trans
-
cis
cis
trans
trans
-
37915
92014
21126
20134
large/small
large/small
small/large
small/large
small/large
large/small
large/small
small/large
small/large
small/large
large/small
large/small
large/small
small/large
small/large
large/small
large/small
large/small
-
large/small
large/small
large/small
small/large
large/small
large/small
large/small
large/small
20.89
20.53
20.29
20.97
24.36
24.27
24.29
23.67
24.48
23.92
24.06
24.38
12.15
16.78
20.57
20.28
20.68
23.05
23.39
23.40
23.68
26.47
26.45
26.00
25.95
25.86
25.78
21.32
21.26
20.60
21.38
24.69
24.52
24.93
23.81
24.73
24.03
24.53
24.61
12.65
17.08
20.88
21.14
21.01
23.25
-
0.43
0.72
0.31
0.41
0.33
0.24
0.64
0.14
0.25
0.11
0.47
0.22
0.50
0.30
0.31
0.86
0.33
0.20
-
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
147925
419017
905383
2162551
6764241
1095563
1248147
36208
887460
38645
968181
93326
113170
603701
1429709
126790
80950
2606791
83635
1646740
3404070
1458540
5211330
19889
204874
841008
2090429
1955817
2927075
1860949
2945696
1232949
551104
31914
858085
111155
269718
161772
1115069
109307
-
Idose TPA
Glyceraldehyde TPA
Erythrose TPA
Rhamnose TPA
Fucose TPA
Quinovose TPA
2-deoxy glucose TPA
2 deoxy galactose TPA
2-O-Me glucose TPA
3-O-Meglucose TPA
Mannosamine TPA
ManNAc TPA
-
trans
trans
cis
2522815
79875
24.16
23.73
27.18
27.23
27.02
27.02
27.23
27.20
0.77
0.05
0.71
0.78
1.03
1.07
1.36
1.42
368569
3564379
424724
4424536
4342
cis
Glucosamine TPA
GlcNAc TPA
Galactosamine TPA
GalNAc TPA
trans
trans
trans
trans
17381560
275652
* peaks A and B are as labeled in Figure 1.
ManNAc (supplementary Fig. S.2.). This indicates an
involvement by the 2-N-acetyl in the specificity of the
thiazolidine ring closure step, possible due to steric hin-
drance, and prior to the subsequent per-O-acetylation.
deoxy sugars followed the same cis/trans-2,3 rules ob-
served for the neutral hexoses and pentoses (Table 1).
For glyceraldehyde the difference in peak A/peak B
height ratios is relatively small. However, the 2-deoxy
sugar TPAs did not conform to the cis/trans-2,3 peak
ratio rules. 2-DeoxyGal TPA gave a single peak, pre-
sumably due to unresolved diastereomers. However, 2-
O-MeGlc and 3-O-MeGlc both have large peak A/peak B
ratios characteristic of trans-2,3 TPA derivatives.
(iii) TPA Derivatives for the Analysis of Oligo- and
Polysaccharides. Having established the efficacy of
the TPA derivatives for GC analysis of monosaccharides
the procedure was subsequently evaluated for the analy-
sis of oligo- and polysaccharides (Fig. 2). Acid hydro-
lyzed samples were evaporated to dryness and converted
to TPA derivatives by the one-pot procedure. Raffinose
and stachyose were chosen because they both contain
glucose and galactose residue, but in a different ratio (1:1
and 1:2, respectively), plus a terminal ketose sugar (fruc-
tose) that cannot form as C1-TPA derivative. For these
examples the faster eluting TPA stereoisomers (Glc1 and
Gal1) gave overlapping peaks, but the slower isomeric
peaks (Glc2 and Gal2) are well resolved (Fig 2. A and B).
The Gal2 TPA is more intense for stachyose, consistent
with the greater quantity of galactosyl residues present
in that oligosaccharide (Fig. 2. B).
The amino sugars give the same TPA products
irrespective of whether they are 2-N-acetylated (Man-
NAc, GlcNAc, GalNAc) or not (ManN, GlcN, GalN), and
hence have the same GC retention times (Table 1). The
GlcN-TPA and GalN-TPA conform to the observed 2,3-
cis/trans rules, confirming the results seen for the neu-
tral sugars, but not the ManN-TPA, which was distinctly
different to the ManNAc-TPA (Table 1). To confirm
this, we prepared thiazolidines from ManN, ManNAc,
and ManNAc after de-N-acetylation by acid hydrolysis
(Supplementary Fig. S.2.).
After the subsequent
peracetylation step all three samples give the identical
ManNAc-TPA derivative, but only the ManNAc-TPA has
the intact 2-N-acetyl group during the Schiff
base/thiazolidine ring closure. The peak A/B ratio is
clearly smaller for the ManNAc-TPA derived from
Frost grape polysaccharide (FGP) is a recently
discovered, complex arabinogalactan isolated from na-
tive grapevine species, with emulsion properties and
high viscosity.12,13 Compositional analysis of FGP as al-
ACS Paragon Plus Environment

Analytical Chemistry
Page 4 of 9
Page 4 of 9
dononitrile acetates showed that it typically consists of
Ara, Gal, Xyl, and Man in a ratio 5: 3: 1: 0.3.7 The pre-
sent TPA analysis confirms these findings (Fig. 2. C).
Two GC peaks are evident for each of the pentose resi-
dues (Ara1 and 2, Xyl1 and 2) and two peaks for the hex-
oses (Gal1 and 2, Man1 and 2). Although Ara2 and Xyl2,
and Gal1 and Man1 TPAs are not fully resolved by the
chromatography, these components are readily identi-
fied by the baseline separation of the corresponding
second peaks.
tively. Molecular ions were not observed, but secondary
fragmentations occur by neutral eliminations of acetate
to give [M-120]+ (Fig. 3, supplementary Fig. S.4). Hence,
[M-120]+ ions m/z 377, 372, 372, and 371 are observed for
[U-13C]Glc-TPA, [1-13C]Glc-TPA, [2-13C]Glc-TPA, and un-
labeled Glc-TPA, respectively (Fig. 3).
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
Fig. 2. TPA analysis of oligo- and polysaccharides after acid
hydrolysis. A. raffinose (Glc, Gal; 1:1); B. stachyose (Glc,
Gal; 1:2); C. Frost grape polysaccharide (Ara, Xyl, Gal,
Man). TPA derivatives were analyzed by GC-MS.
(iv) EI-MS Ion Fragmentation Spectra of Thiazoli-
dine Peracetates. Electron impact (EI) ionization is a
‘hard’ ionization method that leads to extensive carbon-
carbon bond fragmentations that are often helpful for
structure determinations.14,15 The extent of these frag-
mentation pathways are readily traced by using stable
isotopes, leading to selective mass increase of specific
fragment ions.16,17 EI-MS spectra were obtained on the
hexosyl TPA derivatives of [U-13C]glucose, [1-13C]glucose,
[2-13C]glucose, and unlabeled glucose control (Fig. 3),
and compared with those for the pentosyl derivatives of
[U-13C]ribose, [1-13C]ribose, and unlabeled ribose (sup-
plementary Figs. S.3. and S.4.).
Fig. 3. EI-MS ion fragmentation spectra of thiazoli-
dine peracetates of selectively isotopically labeled
glucose. A. [U-13C]Glc; B. [1-13C]Glc; C. [2-13C]Glc; D.
unlabeled Glc control. E. Deduced ion fragmenta-
tions giving m/z 130 and 88 from C1-C2 bond cleav-
age, and m/z 214 and 172 from C3-C4 cleavage. The
[M-120] ion arises from successive neutral losses of
acetic acid.
All of the monosaccharides tested gave rise to
dominant m/z 130 ions due to C1 – C2 bond cleavage of
the parent sugars. This correspond to the stable thiazol-
idinium ion [C3H5SNAc]+ (calculated mass 130.0327 Da),
and that fragments further by a loss of ketene to give
[C3H6SN]+ (calculated mass 88.0221 Da). These ions
contain the carbon-1 of the parent monosaccharide, and
hence for the isotopically-labeled [U-13C]glucose, [1-
13C]glucose, [U-13C]ribose, and [1-13C]ribose TPA deriva-
tives these are observed at m/z 131 and m/z 89, respec-
Similarly, m/z 304, 300, and 299 [M-120]+ ions
are seen for [U-13C]Rib-TPA, [1-13C]Rib-TPA, and the un-
ACS Paragon Plus Environment

Page 5 of 9
Analytical Chemistry
Page 5 of 9
labeled Rib-TPA (supplementary Fig. S.4). The [M-120]+
ions fragment further by cleavage across the C3 – C4
bond of the parent sugar with retention of charge at the
derivatized end (Fig. 3, supplementary Fig. S.4). Hence,
the Glc-TPA series give rise to C3 – C4 cleavage ions at
m/z 217, 215, 215, and 214, respectively, and similarly for
the Rib-TPAs at m/z 217, 215, and 214. These ions subse-
quently also undergo a neutral ketene loss (-42 Da) to
generate significant ions at m/z 175, 173, and 172 (Fig. 3,
supplementary Fig. S.4).
89) and m/z 131/ (m/z 130 + 131) for each of the cellular compo-
nent monosaccharides (sugar TPAs) identified by the GC analysis.
1
2
3
4
5
6
7
8
The percentage 13C incorporations into the
anomeric carbon of these sugars were calculated from
m/z 89/(m/z 88 + 89) and m/z 131/(m/z 130 + 131) for
each of the cellular component monosaccharides (sugar
TPAs) identified by the GC analysis (Table 2). For the
Bacillus strain the fractional incorporations were typi-
cally 10 – 30% depending on the sugar residue and on
the metabolic precursor. The 13C enrichments of the
ribose C-1 are noticeably lower than those of the neutral
hexoses, presumably due to an active PPO pathway.19 In
general the incorporations from [1-13C]Glc and [1-13C]Fru
were similar, although the cellular Man shows a greater
enrichment from fructose than from glucose (Table 2).
This was also evident for the Streptomyces, for which
the Man m/z 88/89 and m/z 130/131 ratios are smaller
when the strain is grown on [1-13C]Fru (Fig. 4). A similar
effect is also seen for the GlcNAc sugar from the Strep-
tomyces strain (Fig. 4). The explanation for this greater
enrichment of Man and GlcNAc from 13C-Fru is that the-
se sugars are most often biosynthesized in a single en-
zymatic step from fructose-6-phosphate, involving
mannose-6-phosphate isomerase or glucosamine syn-
thetase, respectively.20,21
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
(v) Fractional 13C Incorporations at the Carbohy-
drate Anomeric Carbons. The characterized EI ioni-
zation fragmentation pathways can be valuable for iden-
tifying and quantifying the extent of incorporation of
isotopes into biological molecules during metabo-
lism.12,15 For the sugars this is particularly important for
the characteristic C-1 anomeric carbons,18 which are
readily lost as CO2 during oxidative metabolism of hex-
ose-6-phosphate to pentose sugars (the so-called phos-
phopentose oxidative (PPO) pathway).19 Moreover, C3-
C4 metabolism via the aldolase-catalyzed recombination
steps of the glycolytic-gluconeogenesis pathways also
leads to redistribution of the metabolic carbon flux.19
13
Metabolic incorporation of C into cellular car-
bohydrates was evaluated for two diverse bacterial spe-
cies, Bacillus and Streptomyces. The cells were grown in
culture on either [1-13C]glucose or [1-13C]fructose as a
carbon source, washed to remove excess soluble 13C-
labeled precursors, and then acid hydrolyzed. The hy-
drozylate was dried and the acid-labile monosaccharides
were converted to thiazolidine peracetates for analysis
by GC/MS. Six cellular sugars were detected in the hy-
drozylate (Rib, Glc, Gal, Man, GlcNAc, and GalNAc),
each giving rise to two diastereomer GC peaks (Table 2).
Table 2. Metabolic enrichment of 1-13C-labeled precur-
sors (Glu or Fru) into the cellular carbohydrate compo-
nents of Bacillus subtilis.*
Rt
1-13C-labeled Glucose
1-13C-labeled Fructose
Fig. 4. 13C/12C mass ratio analysis of cellular sugars
from Streptomyces cultured on [1-13C]Fru (A, B, C) or [1-
13C]Glc (D, E, F). The EI-MS ions (m/z 88, 89, and m/z
130, 131) from the C-1 carbons of Rib-TPA (A, D), Man-
TPA (B, E), and GlcNAc-TPA (C, F) are shown. Larger
[M+1]/M ratios are indicative of greater 13C enrichment.
Sugar TPA
min
% m/z 89
% m/z 131
% m/z 89
% m/z 131
Rib A
20.28
20.60
24.23
24.33
24.50
24.70
25.73
25.86
26.95
27.12
10.12
9.85
12.17
12.12
25.99
16.34
25.95
15.03
16.11
26.81
26.13
12.54
10.43
9.96
12.68
13.58
23.99
22.54
23.99
23.06
20.66
29.02
28.37
25.35
Rib B
Glc/Gal A
Man A
24.27
14.90
24.38
13.57
14.98
26.37
20.24
23.72
22.21
21.73
21.72
21.65
17.72
26.31
32.91
30.74
Glc B
Man B
MECHANISTIC STUDIES To evaluate a mechanistic
basis for the diastereoselective behavior of the TPA-
derivatives, NMR spectra were obtained of the D-Glc
and L-Ara thiazolidines prior to the peracetylation step
(Fig. 5). The thiazolidine H-1 protons were identified
(4.8 – 5.0 ppm) and the ratio of diastereomers was as-
sessed from the observed integrals. These data were
compared with the C1-epimer ratios obtained by GC
analysis of the corresponding Glc and Ara TPAs, which
includes the subsequent peracetylation procedure. The
GalNAc A
GlcNAc A
GlcNAc B
GalNAc B
*Bacillus cells were metabolically labeled in culture, washed to re-
move excess 13C-labelled precursors, then acid-hydrolyzed and
converted to thiazolidine peracetates for analysis by GC/MS. The 13
C
% incorporations at C-1 were calculated from m/z 89/ (m/z 88 +
ACS Paragon Plus Environment

Analytical Chemistry
Page 6 of 9
Page 6 of 9
diastereoselective A/B ratios were apparent from both
data sets (Fig. 5).
In this paper we have described aldose thiazol-
idine peracetates that are readily prepared in a one-pot
reaction and that are highly suitable for analysis by
GC/MS. We have characterized twenty seven different
monosaccharides using these new derivatives, including
all of the D-series, neutral aldopentoses and aldohexos-
es. As further examples, several oligosaccharides and a
novel arabinogalactan polysaccharide have also been
analyzed. The C1 – C2 and C3 – C4 IE-MS ion fragmen-
tation pathways has also been established for hexose
and pentose TPAs, and shown to be valuable for tracing
the enrichment of stable isotopes into carbohydrate
metabolism. The carbohydrate TPAs gave two diastere-
omeric peaks observed by GC, although unexpectedly
these were diastereoselective, dependent upon the C2 –
C3 stereochemistry of the parent sugar. Hence, this has
further value for establishing cis/trans-2,3- sugars based
on GC analysis alone, and is therefore generally applica-
ble on a sub-nanomole scale.
1
2
3
4
5
6
7
8
Thus, the H-1A and H-1B of Ara-thiazolidine are
observed at 4.88 ppm (J1,2 = 7.5 Hz) and 4.95 ppm (J1,2 =
3.9 Hz), in a percentage ratio of 78.0 : 22.0, in close
agreement with the GC peak A/B ratio of 78.1 : 21.9.
Similarly, the Glc-thiazolidine H-1A and H-1B NMR sig-
nals are at 4.92 ppm (J1,2 = 6.4 Hz, 59.9%) and 4.95 (J1,2
= 4.0 Hz, 40.1%), respectively, and are also in fair
agreement with the observed GC ratio of 51.4 : 48.6 (Fig.
5). These data indicate that the peak A/B ratio differ-
ences occur prior to the peracetylation step, and is most
likely determined during the thiazolidine ring for-
mation.
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
METHODS AND MATERIALS
Materials. Isotopically-labeled carbohydrates were
from Omicron Biochemicals, Inc., South Bend, IN,
and other reagents from Sigma-Aldrich, St. Louis,
MO. Frost grape polysaccharide was obtained from
Dr. Stephen Vaughn, Agricultural Research Service,
Peoria, IL. Monosaccharide stock solutions were
prepared at 100 mg.mL-1 in water and stored at -20 °C.
Cysteamine reagent was prepared by dissolved cys-
teamine hydrochloride (1 g) in dry pyridine (30 mL).
The reagent was stable for at least one week when
Fig. 5. Equivalence of Ara-thiazolidine (NMR signals,
left) and Ara-TPA after peracetylation (GC profile, right).
The C-1 epimers are labeled A or B.
1
stored in the dark at 4 °C. H and 13C nuclear magnet-
ic resonance (NMR) spectra were obtained on a
Bruker Avance instrument (Bruker BioSpin Corp.,
Billerica, MA). The spectra were recorded in D2O,
obtained from Cambridge Isotope Labs (Andover,
MA). The pulse sequences were supplied by Bruker,
and processing was done with Bruker TOPSPIN, ver-
sion 1.3. Matrix-assisted laser desorption/ionization
time-of-flight mass spectrometry (MALDI-TOF/MS)
was achieved on a Bruker Microflex reflecton instru-
ment (Bruker Daltonics, Billerica, MA) using 2, 5-
dihydrobenzoic acid as the matrix. Bacillus subtilis
MW10 was obtained for the Bacillus Genetic Stock
Center, Columbus, OH, and Streptomyces spp. NRRL
F-5065 from the ARS Microbial Culture Collection
housed in Peoria, IL.
Preparation of Carbohydrate Thiazolidine
Peracetates. Aqueous stock solutions (10 µL, 100
mg.mL-1) of the standard sugars were evaporated to
dryness in screw-capped reaction tubes. Dilution se-
ries were prepared were appropriate. Oligo- and pol-
ysaccharide samples were hydrolyzed with aqueous
trifluoroacetic acid (2 M; 2 mL, 120°C; 60 min), then
evaporated to dryness. Cysteamine reagent (1 mL)
was added to the sugar residues and were reacted on
The diastereoselective GC peak ratio probably
arise because of the more favorable formation of one C-1
epimer over the other, most likely during the formation
of the thiazolidine ring (Scheme 1). This occurs initially
by formation of a Schiff’s base between the aldose car-
bon 1 and the amino group of the cysteamine, and is
followed by a nucleophilic attack by the thiol group
leading to the thiazolidine ring closure. It is the facial
preference for this nucleophilic addition that deter-
mines the ratio of the two TPA geometric isomers for
each sugar, as seen in Table 1. This facial preference for
ring closure may be determined by hydrogen bonding in
the Schiff’s base intermediates (supplementary Fig. S.
5.). Thus, hydrogen bonding between the thiazolidine
imino nitrogen and the monosaccharide 2-OH group (2-
OH – N hydrogen bonding) may result in a transiently
stable 5-membered ring, or to the 3-OH group (3-OH –
N hydrogen bonding) resulting in a 6-membered ring
(supplementary Fig. S. 5.). Hence, these stabilized in-
termediates result in preferential thiazolidine ring clo-
sure, leading to a preferred diastereomer.
CONCLUSION
ACS Paragon Plus Environment

Page 7 of 9
Analytical Chemistry
Page 7 of 9
ACKNOWLEDGMENT
a hot block (60 °C; 20 min). The reactions were
cooled and without evaporating were treated with
acetic anhydride (1 mL, 60 °C; 20 min) to complete
the peracetylation. The completed TPA reactions
were cooled and evaporated to dryness. The residues
were dissolved in ethyl acetate (1 mL) and washed
with distilled water (2 mL). The upper, organic layer
was used directly for GC/MS analysis (1 µL injection,
equivalent to 1 µg of the underivatized parent sugar).
Gas Chromatography/Mass Spectrometry. GC/MS
analyses were performed on a Shimadzu Model GC
2010 Plus gas chromatograph equipped with an AOC
Model 20i autoinjector. The GC was interfaced with a
Shimadzu Model QP2010 Ultra mass-selective detec-
tor configured in electron impact (EI) mode. Chro-
matography was accomplished with a fused-silica
capillary Phenomenex ZB-5MSi column (30 m; 0.25
mm). Helium (18.6 mL/min) was used as the carrier
gas. The oven temperature was ramped over a linear
gradient from 150 °C to 300 °C at 5 °C/min, then held
for 5 min. The MS acquisition was started 6 min into
the run to occlude excess reagent peaks. Injector and
detector/interface temperatures were 275 and 300 °C,
respectively. Mass spectra were recorded in positive
ion mode over the range m/z 45 − 600.
Metabolic Experiments. Bacterial cells (Bacillus
subtilis and Streptomyces spp. NRRL F-5065) were
cultured in liquid TYD medium (tryptone (2 g.L-1),
yeast extract (2 g.L-1), glucose (6 g.L-1) and
MgCl2.6H2O (0.3g.L-1) as described previously.12. For
labeling experiments the glucose component was
substituted with a 13C isotopically labeled carbon
source, either [1-13C]glucose or [1-13C]fructose. The
cultures were grown aerobically on a shaker table
(200 rpm, 28 °C, 24 h). Cell pellets were harvested by
centrifugation (8000 rpm, 20min) and washed several
times with deionized water. The washed cells were
acid hydrolyzed with aqueous trifluoroacetic acid (2
M; 2 mL, 120°C; 180 min), then centrifuged to remove
cell debris. The supernatants were recovered, evapo-
rated to dryness on an airline, and the component
monosaccharides were converted to TPA derivatives
and analyzed by GC/MS as described above.
1
2
3
4
5
6
7
8
We thank Dr. Joseph O. Rich for pre-review of the man-
uscript. Mention of any trade names or commercial
products is solely for the purpose of providing specific
information and does not imply recommendation or
endorsement by the US Department of Agriculture.
USDA is an equal opportunity provider and employer.
ABBREVIATIONS
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
TPA, thiazolidine peracetate; GC, gas chromatography;
EI-MS, electron impact-mass spectrometry; CID, colli-
sion induced dissociation.
REFERENCES
(1) Medeiros, P. M.; Simoneit, B. R. T. J. Chroma-
togr. A 2007, 1141, 271.
(2) Mischnick, P. Adv. Polym. Sci. 2012, 248, 105.
(3) Chaplin, M. F. Monosaccharides, In Carbohy-
drate Analysis, A Practical Approach, Chaplin,
M. F.; Kennedy, J. F. ed. The Practical Approach
Series: Oxford, 1994; pp 27.
(4) Grice, I. D.; Wilson, J. C. In Microbial Glycobi-
ology: Structures, Relevance and Applications,
Holst, O.; Brennan, P. J.; von Itzstein, M. pp.
Elsevier: Amsterdam 2010; 233.
(5) Sawabdekeb, J. S.; Sloneker, J. H.; Jeanes, A.
Anal. Chem., 1965, 37, 1602.
(6) Guerrant, G. O.; Moss, C. W. Anal. Chem., 1984,
56. 633.
(7) Sweeley, C. C.; Bentley, R.; Makita, M.; Wells,
W. W. J. Am. Chem. Soc., 1963, 85, 2497.
(8) Price, N. P. J. Anal. Chem., 2004, 76, 6566.
(9) Ramazani, Y.; Levtchenko, E. N.; van den Heu-
vel, L.; van Schepdael, A.; Paul, P., Ivanova, E.
A.; Pastore, A.; Hartman, T. M.; Price, N. P. J.
Carbohydr. Res., 2017, 439, 9.
(10) Fife, T. H.; Natarajan, R.; Shen, C. C.; Bembi, R.
J. Am. Chem. Soc. 1991, 113, 3071.
(11) Luhowy, R.; Meneghini, F. J. Am. Chem. Soc.
1979, 101, 420.
(12) Price, N. P. J.; Vermillion, K. E.; Eller, F. J.;
Vaughn, S. F. J. Agric. Food Chem., 2015, 63,
7286.
(13) Hay, W. T.; Vaughn, S. F.; Byars, J. A.; Selling,
G. W.; Holthaus, D. M.; Price, N. P. J. J. Agric.
Food Chem. 2017, 65, 8754.
(14) Heinzle, E.; Yuan, Y.; Kumar, S.; Wittmann, C.;
Gehre, M.; Richnow, H.-H.; Wehrung, P.; Ad-
am, P.; Albrecht, P. Anal. Biochem., 2008, 380,
202.
(15) Wittmann, C. Microb. Cell Fact., 2007, 6, art.
no. 6.
ASSOCIATED CONTENT
Supporting Information. Five additional supplemen-
tary figures, and an additional reference.
AUTHOR INFORMATION
Corresponding Author
(16) Jennings II, M. E.; Matthews, D. E. Anal. Chem.,
2005, 77, 6435.
(17) Zamboni, N.; Fendt, S.-M.; Rühl, M.; Sauer, U.
Nat. Protoc., 2009, 4, 878.
*neil.price@ars.usda.gov
Author Contributions
The manuscript was written through contributions of all
authors. / All authors have given approval to the final
version of the manuscript.
ACS Paragon Plus Environment

Analytical Chemistry
Page 8 of 9
Page 8 of 9
(18) Jiang, W.; Wysocki, V. H.; Dodds, E. D.;
(20)Ichikawa, M., Scott, D. A., Losfeld, M.-E.,
Freeze, H. H. J. Biol. Chem. 2014, 289, 6751.
(21) Golinelli-Pimpaneau, B.; Le Goffic, F.; Badet, B.
J. Am. Chem. Soc., 1989, 111, 3029.
1
2
3
4
Miesfeld, R. L.; Scaraffia, P. Y. Anal. Biochem.
2010, 404, 40.
(19) Schwender, J.; Ohlrogge, J. B.; Shachar-Hill, Y.
J. Biol. Chem., 2003, 278, 29442.
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 Paragon Plus Environment

Page 9 of 9
Analytical Chemistry
Page 9 of 9
1
2
3
For Table of Contents Only
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
9
ACS Paragon Plus Environment