Incorporation of 13C glucose into nicotinamide in E. coli and in S. cerevisiae

Article abstract of DOI:10.1139/v01-056

The mode of incorporation into nicotinamide of label from ¹³C-labeled samples of D-glucose, in Escherichia coli and Saccharomyces cerevisiae, was determined by means of ¹³C NMR spectroscopy. The results, which confirm and extend early studies with radioactive tracers, permit a definitive choice to be made between alternative biogenetic proposals.

Full text of DOI:10.1139/v01-056

418
13
Incorporation of C glucose into nicotinamide in
E. coli and in S. cerevisiae
Ram Nath Gupta, Thomas Hemscheidt, Brian G. Sayer, and Ian D. Spenser
Abstract: The mode of incorporation into nicotinamide of label from 13C-labeled samples of D-glucose, in Escherichia
coli and Saccharomyces cerevisiae, was determined by means of 13C NMR spectroscopy. The results, which confirm
and extend early studies with radioactive tracers, permit a definitive choice to be made between alternative biogenetic
proposals.
Key words: nicotinic acid, nicotinamide, biosynthesis, glucose incorporation, 13C NMR.
Résumé : Faisant appel à la spectroscopie RMN du 13C, on a déterminé le mode d’incorporation du 13C d’échantillons
de D-glucose marqué dans le nicotinamide d’Escherichia coli et de Saccharomyces cerevisiae. Les résultats, qui confir-
ment et qui étendent la portée d’études antérieures à l’aide de marqueurs radioactifs, permettent de faire un choix défi-
nitif entre les diverses propositions biogénétiques.
Mots clés : acide nicotinique, nicotinamide, biosynthèse, incorporation du glucose, RMN du 13C.
[Traduit par la Rédaction] Gupta et al. 425
Introduction
these two precursors combine to yield the skeleton of
nicotinic acid has not been established.
Two major de novo pathways of nicotinic acid
biosynthesis are known (for a review see ref. 1). The key in-
termediate of both routes is quinolinic acid which in the two
pathways originates from different precursors.
The first of the pathways to the vitamin to be described is
found in mammalian tissues, in molds and fungi, in yeasts
under aerobic conditions, and in some higher plants. The
primary precursor of this pathway is tryptophan. The steps
and the enzymology of this route are well established (for a
review see ref. 1).
That another route to the vitamin must exist became prob-
able when it was shown that nicotinic acid in Escherichia
coli (2) and trigonelline, the methylbetaine of nicotinic acid,
in a higher plant (3), were not derived from tryptophan. This
second pathway, which would appear to be more widely dis-
tributed than the route from tryptophan,2 is found in
prokaryotes in general, e.g., E. coli (5), Mycobacterium tu-
berculosis (6) and other bacteria, and also in yeasts under
anaerobic conditions (7, 8). This route generates nicotinic
acid by condensation of a triose phosphate with a dicarbox-
ylic acid related to aspartic acid. The mechanism whereby
The first detailed incorporation studies related to this al-
ternative pathway of nicotinic acid biosynthesis were carried
out with higher plants.3 This early work with 14C-labeled
substrates established incorporation of label from glycerol,
and from dicarboxylic acids related to succinic acid (for
summary see ref. 13). It is certain from tracer work with
14C-labeled substrates that glycerol (and hence triose phos-
phate) supplies C-4, C-5, and C-6 of the ring. Label from
[1,3-14C]glycerol was found predominantly, and approxi-
mately equally, at C-4 and C-6, whereas label from [2-
14C]glycerol entered C-5 of the pyridine ring (13). Label
from D-[3-14C]glyceraldehyde entered mainly C-4 (55 ±
?4%), with the rest of the radioactivity spread throughout the
molecule (14).
The dicarboxylic acid supplies carbon atoms C-2, C-3 of
the ring system, and the two carboxylic acid groups of quinolinic
acid (and thence of nicotinic acid, after decarboxylation of
the carboxyl group at C-2). 14C from [2,3-14C]succinate as
well as from [2-14C]acetate labeled C-2 and C-3 of the pyri-
dine ring approximately equally, while label from [1-14C]ac-
etate, [1,4-14C]succinate, as well as from [4-14C]aspartic
Received December 22, 2000. Published on the NRC Research Press Web site on May 3, 2001.
R.N. Gupta, B.G. Sayer, and I.D. Spenser.1 Department of Chemistry, McMaster University, Hamilton, ON L8S 4M1, Canada.
T. Hemscheidt. Department of Chemistry, University of Hawaii at Manoa, 2545 The Mall, Honolulu, Hawaii 96822.
1Corresponding author (telephone: (905) 525-9140 ext. 23245; fax: (905) 522-2509; e-mail: spenser@mcmaster.ca).
2The view that nicotinic acid (niacin) and its congener, nicotinamide, components of NAD and related coenzymes, was universally
derived from tryptophan was firmly held until the mid-1950ies. But even today, current textbooks of biochemistry (4), if they refer
to niacin biosynthesis at all, persist in stating categorically that the sole route to nicotinic acid starts from tryptophan.
3Results, showing the precise distribution of label from 14C-labeled substrates within the pyridine ring of the alkaloids nicotine and
anabasine in intact tobacco plants and within ricinine in the castor bean plant, can be extrapolated to the origin of the nicotinic
acid ring system, since it has been demonstrated that the intact ring of nicotinic or of quinolinic acid serves as the precursor of the
pyridine ring of the tobacco alkaloids (9–11) and of ricinine (12).
4Experimental error not reported.
Can. J. Chem. 79: 418–425 (2001)
© 2001 NRC Canada
DOI: 10.1139/cjc-79-4-418

Gupta et al.
419
Fig. 1. 13C NMR spectrum of nicotinamide derived from D-
[13C6]glucose: (A) from E. coli (expt 1); and (B) from S.
cerevisiae (expt 2).
oxidase (EC 1.4.3.16), is present not only in E. coli (22) but
also in a higher plant (cotton) (23). It catalyzes the
dehydrogenation, in the presence of FAD, of L-aspartic acid
to yield 2-iminobutanedioic acid (2-iminosuccinic acid, re-
ferred to incorrectly as iminoaspartic acid) (24). However,
there appears to be some question whether or not in E. coli
the C-N fragment of aspartic acid is incorporated intact into
the vitamin (25). Certainly, 2-iminosuccinic acid is generated
also from oxalacetic acid plus ammonium sulfate (24).
The nadA gene product (“A protein”), quinolinate
synthase (26), catalyzes the condensation of 2-iminosuccinic
acid as the source of the fragment N-1,C-2,3,CO2H of nico-
tinic acid, with the triose phosphate as the source of the
fragment C-4,5,6. Both dihydroxyacetone phosphate and
glyceraldehyde phosphate react readily in this condensation
reaction. The identification of the former as the actual sub-
strate for the condensation reaction rests on a single experi-
ment with an inhibitor of triose phosphate isomerase (26).
The disposition of the triose on entry into the C3 unit C-
4,5,6 differs in alternative biogenetic proposals.
In a recently published hypothetical biogenetic sequence,
the carbon atom of the phosphotriose that bears the phos-
phate ester group is postulated to yield C-6 of nicotinic acid
(27). A similar disposition of the triose phosphate on entry
into nicotinic acid had been advanced earlier (21, 28, 29).
Alternatively, the phosphate ester carbon atom of the triose
was postulated to generate C-4 of the niacin ring system (1,
for a review see ref. 11, 14, 22, 30).
We now present results which establish that the latter
alternative is the correct one.
Results and discussion
acid entered solely into the carboxyl carbon of nicotinic
acid. It is noteworthy that label from [3-14C]aspartic acid, as
well as from [3-14C]malic acid, entered both C-2 (38 ± ?4%
of the total specific activity of the labeled product) and C-3
(57 ± ?4, 61 ± ?4%), rather than being confined to C-3 (15),
as would have been expected if aspartic acid had been incor-
porated intact.
We have employed NMR spectroscopy to examine the
mode of incorporation of 13C-labeled D-glucose into
nicotinamide in E. coli B WG2 (a pdxH– auxotroph) and in
S. cerevisiae strain ATTC 7752 (IFO 1234). These are the
micro-organisms that we had employed in our previous
investigations of pyridoxine (31) and thiamin biosynthesis (32).
The 13C NMR method had not hitherto been applied in the
study of niacin biosynthesis. To do so, a method of isolation
with minimal carrier dilution had to be developed. The method
that eventually proved successful, described in the experimental
section, was based on the isolation of nicotinamide, a weak
base that can be isolated from the cell hydrolysate by solvent
extraction and can then be further purified by liquid chroma-
tography. Nicotinamide is quite stable. There is no significant
hydrolysis of nicotinamide when it is heated in 0.1 M HCl so-
lution for 1 h. In earlier investigations of niacin biosynthesis
with 14C, significant carrier dilution was required when either
quinolinic acid or nicotinic acid, amphoteric compounds, were
isolated. To test the utility of the method, D-[13C6]glucose was
selected as the substrate whose incorporation was examined.
Incorporation into nicotinamide of multicarbon units derived
from this bond-labeled substrate5 was indeed observed both in
In E. coli, label from [1-14C]glycerol entered C-4 (57 ± ?4%)
and C-6 (40 ± ?4%) of quinolinic acid (16), while label from
[1-14C]dihydroxyacetone 1-phosphate entered C-4 exclusively
(17). Label from D-[1-14C]glucose (and probably also from
D-[6-14C]glucose) was spread over the whole molecule, with
preferential entry into C-4 (24 ± ?4%) and C-6 (25 ± ?4%)
whereas label from D-[2-14C]glucose entered C-5 preferen-
tially (78 ± ?4%) (16).
All available evidence thus points to the derivation of the
ring system of quinolinic acid, and hence of that of nicotinic
acid, from a triose plus a compound related to aspartic acid (18).
Two genes, nadA and nadB, that control the biosynthesis
of nicotinic acid in E. coli and S. typhimurium, have been
mapped and expressed (for a review see ref. 19, 20) and the
corresponding gene products have been characterized (20,
21). The nadB gene product (“B protein”), L-aspartic acid
5 These experiments cannot differentiate between the two pathways since the distribution of D-[13C6]glucose-derived label within niacin in
both pathways is predicted to be identical: In triose–aspartate-derived niacin the C3 unit C-4,5,6 is generated via a triose while the C2 unit
C-2,3 arises via a dicarboxylic acid that is ultimately also derived via a triose, but can randomize its label by tricarboxylic acid cycling. In
the aerobic route to niacin from glucose via tryptophan, a C3 unit of glucose is incorporated into C-4,5,6 of niacin via erythrose while the C2
unit C-2,3 arises via phosphoenolpyruvate.
© 2001 NRC Canada

420
Can. J. Chem. Vol. 79, 2001
Table 1. 13C NMR spectrum of nicotinamide: chemical shifts
Scheme 1. Observed incorporation of 13C into nicotinamide (see
and coupling constants.
Figs. 1–4).
Coupling constants (±0.5 Hz)
Chemical
Carbon
atom no.
shifta (ppm)
13C-13C
124.7
129.2
C-5
C-3
1JC-5 C-4
54.4
108.8
,
1JC-5,C-4 + JC-5,C-6
1
1JC-3 C-2
58.1
64.0
122.1
,
1JC-3 C=O
,
1JC-3,C-2 + JC-3,C=O
1
136.8
147.7
151.8
170.9
C-4
C-2
C-6
C=O
1JC-4, C-5
1JC-2,C-3
1JC-6,C-5
1JC=O,C-3
54.4
58.1
54.4
64.0
aChemical shift (ppm) relative to dioxane as internal standard (66.6 ppm
with respect to TMS at 0 ppm).
E. coli (expt 1, Fig. 1A) and in yeast (expt 2, Fig. 1B). 13C
NMR chemical shifts and coupling constants are summarized
in Table 1.
The appearance of multiplets straddling the signals due to
1
C-4 (136.8 ppm, JC-4, C-5 = 54.4 Hz), C-5 (124.7 ppm,
1
1JC-5, C-4 + JC-5, C-6 = 108.8 Hz), and C-6 (151.8 ppm,
1JC-6, C-5 = 54.4 Hz) and of doublets at the signals due to C-2
1
(147.7 ppm, JC-2, C-3 = 58.1 Hz) and C-3 (129.2 ppm,
1JC-3, C-2 = 58.1 Hz) that were observed in the 13C NMR
spectrum of nicotinamide from E. coli (expt 1, Fig. 1A) con-
firmed that a C3 unit and a C2 unit derived from D-[13C6]glu-
cose had been incorporated. Some scrambling of label had
taken place during the incubation period, as indicated by the
presence of a low intensity doublet around the carbonyl signal,
showing low level labeling of the CONH2 carbon atom and
coupling with C-3 of the ring. The labeling patterns observed
in these and subsequent experiments are summarized in
Scheme 1.
The fact that 13C incorporation into nicotinamide was de-
tected served to show that the experimental procedure can be
employed in the investigation of nicotinamide biosynthesis.
Using this method, an experiment was then carried out to
establish the disposition of the triose phosphate upon its en-
try into the C3 fragment C-4,5,6 of the pyridine ring of
nicotinamide in E. coli. The distribution pattern of label
from D-[1,2-13C2]glucose (expt 5, Fig. 2A) distinguished
among the two possible modes of entry of the triose into the
C3 unit C-4,5,6 of the ring. The spectrum of the nicotina-
mide sample derived from D-[1,2-13C]glucose (Fig. 2A) shows
Upon glycolysis, C-1 and C-2 of glucose give rise, respec-
tively, to the phosphohydroxymethyl carbon and the car-
bonyl carbon atom of dihydroxyacetone phosphate and then,
by triosephosphate isomerization, to the phosphohydroxy-
methyl carbon atom and the central carbon atom of D-
glyceraldehyde phosphate. The incorporation pattern of 13C
from D-[1,2-13C]glucose proves that incorporation of the
phosphotriose takes place in such a way that C-4, but not
C-6, of nicotinamide is derived from the phosphohydroxy
carbon atom, as had been proposed by Gholson and co-
workers (22) and others (1, 11, 14, 30).
1
that the C-5 signal (124.7 ppm, JC-5, C-4 = 54.4 Hz) is cou-
pled to the signal due to C-4 (136.8 ppm, 1JC-5, C-4 = 54.4 Hz).
As before, low intensity satellites are present at other carbon
signals, due to some scrambling of label. Since the signal
due to bond-labeling at C-5 (124.7 ppm) shows coupling to
only one other carbon, C-4, whereas in the spectrum of
D-[13C6]glucose-derived nicotinamide the signal for C-5
Incorporation of label from D-[13C6]glucose and from
D-[1,2-13C2]glucose into the C2 fragment C-2,3 of nicotina-
mide in E. coli must now be examined. This fragment,
together with the carbonyl carbon atom is derived from
2-iminosuccinic acid, the nadB gene product (see Introduction).
Since aspartic acid and related dicarboxylic acids origi-
nate from glucose-derived triose phosphates in several steps,
via [13C3]phosphoenolpyruvate and thence by carboxylation
to [1,2,3-13C3]oxaloacetate, an intact C2 unit derived from D-
[13C6]glucose was expected to enter C-2,3 of nicotinamide.
had shown coupling to two other carbon atoms (1JC-5, C-4
+
1JC-5, C-6 = 108.8 Hz), this coupling pattern shows that a C2
unit derived from the precursor was incorporated into C-4,5
of the nicotinamide.
© 2001 NRC Canada

Gupta et al.
421
Fig. 2. 13C NMR spectrum of nicotinamide derived from D-[1,2-
13C2]glucose: (A) from E. coli (expt 5); and (B) from S.
cerevisiae (expt 6).
Fig. 3. 13C NMR spectrum of nicotinamide derived from D-[4,5-
13C2]glucose (expt 7).
Fig. 4. 13C NMR spectrum of nicotinamide derived from sodium
[2,3-13C2]succinate (expt 12).
into the C2 unit C-2,3 of the ring was lowered, relative to the
13C incorporation into the C3 unit C-4,5,6. This was shown
by the observation that the ratios of the intensities of the
doublets C-6:C-2 and C-4:C-2 (5.3 and 5.1, respectively), in
the nicotinamide spectrum from the latter incubation, were
found to be more than twice as large as the corresponding
ratios in the spectrum of the sample obtained from the
incubation with glucose alone (expt 1) (C-6:C-2 = 2.1,
C-4:C-2 = 2.3). It should be mentioned, parenthetically, that
in a similar set of experiments with yeast (expts 4 and 2, re-
spectively), in which nicotinamide showed an incorporation
pattern similar to that of the nicotinamide derived from D-
[13C6]glucose in E. coli, these ratios were unaffected by the
presence of aspartic acid. This confirms that in yeast under
aerobic conditions the triose–aspartic acid route to
nicotinamide is at best of minor importance.
The appearance of doublets at the signals due to C-2
1
(147.7 ppm, JC-2, C-3 = 58.1 Hz) and C-3 (129.2 ppm,
1JC-3, C-2 = 58.1 Hz) (Fig. 1A) showed that this had indeed
occurred. Incorporation of label from glucose by this multistep
route had presumably taken place via quinolinic acid,
followed by decarboxylation, i.e., with loss of the labeled
oxaloacetate carboxyl group derived from C-3 of glucose.
The low-intensity multiplet around C-3 (129.2 ppm, 1JC-3, C-2
+
1JC-3, C=0 = 122.1 Hz), indicating incorporation of an intact
C3 unit into C-2,3,CO, is explicable on the basis of [1,2,3-
13C3]oxaloacetate recycling through the tricarboxylic acid
cycle to yield [2,3,4-13C3]oxaloacetate, which upon entry
into nicotinamide, loses the unlabeled carboxyl group.
Another indication for involvement of an intermediate de-
rived from aspartate was the observation that when incuba-
tion with D-[13C6]glucose was carried out in the presence of
excess unlabeled aspartic acid (expt 3), incorporation of 13C
This conclusion is borne out by the results of incorpora-
tion experiments with sodium [2,3-13C2]succinate. In E. coli
(expt 12, Fig. 4) label from this precursor was found in the
1
C2 unit C-2,3 (C-2: 147.7 ppm, JC-2, C-3 = 58.2 Hz; C-3:
© 2001 NRC Canada

422
Can. J. Chem. Vol. 79, 2001
Scheme 2. A possible mechanism of the triose phosphate–
aspartate pathway of nicotinic acid biosynthesis.
Scheme 3. Incorporation of carbon atoms of glucose into nico-
tinic acid. Solid lines represent incorporation of multicarbon
fragments. Numbers refer to the carbon atoms of glucose from
which indicated sites within nicotinic acid are derived.
the triose-derived C2 unit C-4,5 was much more highly en-
riched than the aspartate-derived C2 fragment C-2,3
(Fig. 2A). By contrast, in the tryptophan-derived aerobic S.
cerevisiae sample (expt 6) the C2 fragment C-2,3, derived
via phosphoenolpyruvate, was much more highly enriched
than the erythrose-derived C2 unit C-4,5 (Fig. 2B). In the
case of the samples derived from D-[U-13C6]glucose (Fig. 1)
the quantitative difference in the incorporation of
multicarbon units was minor. In nicotinamide derived from
D-[4,5-13C2]glucose in S. cerevisiae (expt 7, Fig. 3) coupling
was confined to C-5,6 (Scheme 3A).
1
129.2 ppm, JC-3, C-2 = 58.2 Hz) of nicotinamide.6 In the
spectrum of yeast nicotinamide, from the same precursor
(expt 13), incorporation of 13C was not detectable.
A mechanism for the biosynthesis of quinolinic acid (and
hence of nicotinic acid), based on the condensation of 2-
iminosuccinic acid with dihydroxyacetone phosphate, was ad-
vanced by Gholson and co-workers in 1982 (22). In order to
arrive at the desired target, quinolinic acid, from these precur-
sors, the proposed mechanism included steps that were equiva-
lent to an isomerization of dihydroxyacetone phosphate into D-
glyceraldehyde phosphate. Derivation of quinolinic acid by
condensation of 2-iminosuccinic acid with D-glyceraldehyde
phosphate rather than with dihydroxyacetone phosphate of-
fers a much simpler path to the product (Scheme 2). It would
be desirable to confirm by further experiment that the gene
product of nadA, quinolinate synthase, is indeed specific for
dihydroxyacetone phosphate, as seems to be generally as-
sumed on the basis of a single experimental observation
(26), or whether D-glyceraldehyde phosphate serves as the
substrate.
When singly 13C-labeled samples of glucose are used as
substrates, differentiation of the two pathways is possible.
The distribution of label observed in the nicotinamide ob-
tained in these experiments is shown in Scheme 1.
On the basis of well-established evidence (4) the trypto-
phan route is predicted to deliver label from D-[1-13C]glu-
cose into C-2 of niacin, while label from D-[2-13C]glucose is
predicted to enter C-3 as well as the carbonyl carbon. The
triose–aspartate route is expected to deliver C-1 of glucose
mainly into C-4 and C-3 of niacin, with some minor amount
entering C-2. By contrast, C-2 of glucose is delivered into
C-5 and C-2 of niacin, with a minor quantity entering C-3.
Niacin generated by the two different pathways should thus
be 13C-enriched at different sites (Scheme 3).
To determine which of the nicotinamide carbon atoms of
the samples derived from D-[1-13C]- and D-[2-13C]-glucose in
E. coli (expts 8 and 10) and in S. cerevisiae (expts 9 and 11)
had been significantly enriched, peak areas of each of the
signals in the 13C NMR spectra of these four samples were
compared with those of the corresponding signals in a spec-
trum of a standard nicotinamide sample. The signal areas in
the spectra of the 13C-enriched samples were normalized
using the signal of C-6 of nicotinamide as a baseline, since,
from the experiments with D-[1,2-13C2]glucose, the C-6 sig-
nal was known to be unenriched in each case. The signals
with maximum enrichment were thus identified. The differences
As already mentioned,5 the triose–aspartate route cannot
be distinguished from the tryptophan route on the basis of
the incorporation pattern of label derived from 13C bond-
labeled samples of D-glucose, since the sites of entry of
multicarbon units from glucose are similar in the two path-
ways (Fig. 1). But even though the distribution of the
multicarbon fragments is similar, the relative efficiency of
entry into the two sites of nicotinamide, C-4,5 and C-2,3,
was found to differ: Comparison of the spectra of nico-
tinamide derived from D-[1,2-13C2]glucose (Fig. 2) shows
that in the triose–aspartate-derived E. coli sample (expt 5)
6 The additional low level incorporation of C2 units into C=O,C-3; C-4,5 and C-5,6 shown in Fig. 4 is presumably due to tricarboxylic acid
cycle recycling of [2,3-13C2]succinate and due to gluconeogenesis, in the course of the 18 h incubation that was required to reach the desir-
able cell density in expt 12.
© 2001 NRC Canada

Gupta et al.
423
Table 2. Experimental details.
Expt
no.
No. of
incubations
Time
(h)
13C NMR
spectrum
Substrates
mg L–1
E. coli B WG2
1
D-[U-13C6]glucose
D-glucose
D-[U-13C6]glucose
D-glucose
100
900
100
900
300
200
800
400
600
300
700
200
1000
5 × 1 L
5 × 1 L
12
15
Fig. 1A
3
L-aspartic acid
D-[1,2-13C2]glucose
D-glucose
5
8
5 × 1 L
5 × 1 L
5 × 1 L
5 × 1 L
15
15
12
18
Fig. 2A
D-[1-13C]glucose
D-glucose
10
12
D-[2-13C]glucose
D-glucose
Na [2,3-13C2]succinate
D-glucose
Fig. 4
S. cerevisiae ATTC 7752 (IFO 1234)
2
D-[U-13C6]glucose
D-glucose
D-[U-13C6]glucose
D-glucose
1000
4000
1000
4000
1500
1000
4500
500
3500
5000
2000
1500
1000
8000
250
2 × 1 L
1 L
12
12
Fig. 1B
4
L-aspartic acid
D-[1,2-13C2]glucose
D-glucose
6
7
2 × 1 L
2 × 1 L
12
18
Fig. 2B
Fig. 3
D-[4,5-13C2]glucose
D-glucose
9
11
D-[1-13C]glucose
D-[2-13C]glucose
D-glucose
1 L
2 × 1 L
12
24
13
Na [2,3-13C2]succinate
D-glucose
2 × 1 L
20
L-threonine
L-proline
250
L-histidine
250
in the distribution of label from D-[1-13C]- and D-[2-
13C]glucose in the samples derived by the triose–aspartate
route (E. coli) and the tryptophan route (aerobically grown
S. cerevisiae) are manifest (Scheme 1).
5 × 1 L cultures. Yeast incubations were performed with 1 L
or with 2 × 1 L yeast cultures (Table 2).
Isolation of nicotinamide
The cells were harvested by centrifugation (10 min at
room temp) and the cells from the multiflask incubations
were pooled. The separated cells were suspended in dilute
HCl (0.1 M, 10 mL per 1 L batch of cells) and the mixture
was autoclaved for 1 h at 121oC. After cooling, acetate
buffer (1 M, pH 5.5) was added to adjust the pH of the sus-
pension to pH 4.7.
The suspension was centrifuged for 10 min at room temp.
The cell debris was sonicated twice in water (20 mL) and
the clear supernatant collected. The pH of the combined
supernatant was adjusted (if necessary) with acetate buffer
(0.2 M, pH 4.7).
The cell hydrolysate was spiked with 1 mL of an aqueous
solution containing 3 mg per mL nicotinamide. The solution
was centrifuged if necessary to remove suspended solids.
The supernatant was applied to a 80 mm × 10 mm column
of Amberlite CG50 (H+ form) and the drainings were
Experimental
13C-labeled compounds
All but one of the 13C-labeled compounds were purchased
from Cambridge Isotope Laboratories, Andover, MA. The
exception was D-[4,5-13C2]glucose, which was obtained from
Isotech Inc., Miamisburg, OH.
Micro-organisms
E. coli B WG2 and S. cerevisiae ATTC 7752 (IFO 1234)
were cultured and worked up as described earlier (32). E.
coli cultures grown to a cell density of 1.0 at 620 nm (12–
15 h) were found to contain ca. 60–80 µg L–1 nicotinamide.
Yeast cultures grown to a cell density of 1.6 or greater
(12–24 h) contained ca. 300–400 µg L–1. Incubations with
labeled compounds in E. coli cultures were carried out with
© 2001 NRC Canada

424
Can. J. Chem. Vol. 79, 2001
collected. The column was washed with water (120 mL).
The drainings and washings were combined and evaporated
at 50oC under vacuum.
Acknowledgments
Research grants from the National Institute of General
Medical Sciences, US Public Health Service, and from the
Natural Sciences and Engineering Research Council of Can-
ada (NSERC) are gratefully acknowledged. We thank Rich-
ard M. Pauloski for skilled technical assistance.
The residue was treated with 100 µL portions of a satu-
rated solution of sodium carbonate (total sodium carbonate
solution ca. 0.4–0.5 mL), until the pH was 9, followed by di-
chloromethane (25 mL). After mixing, anhydrous sodium
sulfate (5 g) was added and the mixture allowed to stand
overnight. The organic layer was separated and filtered
through a small column with a filter disc, containing ca. 2 g
anhydrous sodium sulfate.
The solid cake was stirred twice with dichloromethane
(15 mL), filtered and the combined filtrates evaporated. The
residue was dissolved in water (2 mL), treated with 1M HCl
(20 µL), and the solution applied to a 500 mg column of
SCX (a silica based strong cation exchange resin, Varian,
Harbor City, CA), which had been washed in turn with col-
umn volumes (ca. 10 mL) of 1M NH4OH, methanol, 1M
HCl, and methanol, followed by a water wash, until the
eluate was neutral. All the washings were carried out at a
medium flow rate by applying suction. However, the sample
was passed through the column at a slow rate by reducing
the suction. The column was drained completely by increas-
ing the suction to a maximum. The column was then washed
with water (5 mL) and methanol (2.5 mL) at a medium rate
and the product was then eluted with a mixture of dichloro-
methane (20 mL), methanol (2 mL), and concd. NH4OH
(0.2 mL). The elution was carried out by gravity or under
mild positive pressure of nitrogen.
References
1. (a) J.W. Foster and A.G. Moat. Microbiol. Rev. 44, 83 (1980);
(b) A.G. Moat and J.W. Foster. In Pyridine nucleotide co-
enzymes, chemical, biochemical and medical aspects. Edited
by D. Dolphin, R. Poulson, and O. Avramoviƒ. Wiley, New
York. 1987. Part B, pp. 1–24.
2. C. Yanofsky. J. Bacteriol. 68, 577 (1954).
3. E. Leete, L. Marion, and I.D. Spenser. Can. J. Chem. 33, 405
(1955).
4. (a) A.L. Lehninger. Principles of biochemistry. 2nd ed. Worth
Publishers, New York. 1993. p. 520; (b) D. Voet and J.G. Voet.
Biochemistry. 2nd ed. J. Wiley & Sons, New York. 1995. p.
741; (c) L. Stryer. Biochemistry. 4th ed. W.H. Freeman, New
York. 1995. p. 755; (d) R. Montgomery, T.W. Conway, A.A.
Spector, and D. Chappell. Biochemistry. 6th ed. Mosby, St.
Louis. 1996. p. 264; (e) T.M. Devlin. Textbook of biochemis-
try. 4th ed. Wiley-Liss, New York. 1997. p. 475; (f) G.L.
Zubay. Biochemistry. 4th ed. Wm. C. Brown, McGraw Hill,
Dubuque, IA. 1998. p. 661.
5. (a) M.V. Ortega and G.M. Brown. J. Am. Chem. Soc. 81, 4437
(1959); (b) M.V. Ortega and G.M. Brown. J. Biol. Chem. 235,
2939 (1960).
The eluate was evaporated to dryness to yield
nicotinamide, sufficiently pure for NMR. The recovery and
purity of nicotinamide was monitored by liquid chroma-
tography (HPLC) using a Beckman Ultrasphere (5 µm,
25 cm × 4.6 mm) ODS column. The column was eluted with
a mobile phase containing formic acid (400 µL), concd.
NH4OH (800 µL), and acetonitrile (15 mL) per L of water,
at a flow rate of 1 mL min–1. The peaks were detected at
254 nm. Under these conditions nicotinamide eluted at
13 min. The average recovery of spiked nicotinamide was
ca. 90%. When traces of impurities accompanied the nico-
tinamide sample, it was purified by HPLC, using a Beckman
Ultrasphere (5 µm, 25 cm × 10 mm) semipreparative col-
umn, using the same mobile phase that was used with the
analytical column, at a flow rate of 2.5 mL min–1.
6. D. Gross, A. Feige, R. Stecher, A. Zureck, and H.-R. Schütte.
Z. Naturforsch. 27 b, 1116 (1965).
7. F. Ahmad and A. G. Moat. J. Biol. Chem. 241, 775 (1966).
8. H.-D. Heilmann and F. Lingens. Hoppe Seyler’s Z. Physiol.
Chem. 349, 231 (1968).
9. E. Leete. Bioorg. Chem. 6, 273 (1977).
10. K.S. Yang, R.K. Gholson, and G.R. Waller. J. Am. Chem. Soc.
87, 4184 (1965).
11. E. Leete. In Alkaloids: Chem. Biol. Perspect. Vol.1. Edited by
S.W. Pelletier. J. Wiley & Sons, New York. 1983. pp. 85–152.
12. K.S. Yang and G.R. Waller. Phytochemistry, 4, 881 (1965).
13. I.D. Spenser. In Comprehensive biochemistry. Vol. 20. Edited
by M. Florkin and E.H. Stotz. Elsevier, Amsterdam. 1968.
pp. 275–281.
14. J. Fleeker and R.U. Byerrum. J. Biol. Chem. 242, 3042 (1967).
15. M. Jackanicz and R.U. Byerrum. J. Biol. Chem. 241, 1296
(1966).
Isolation of nicotinamide from the cell hydrolysate of S.
cerevisiae was carried out by the same procedure used for
the isolation from the E. coli cell hydrolysate.
16. J.L.R. Chandler, R.K. Gholson, and T.A. Scott. Biochim.
Biophys. Acta, 222, 523 (1970).
17. F.D. Wicks, S. Sakakibara, R.H. Gholson, and T.A. Scott.
Biochim. Biophys. Acta, 500, 213 (1977).
18. F.D. Wicks, S. Sakakibara, and R.K. Gholson. J. Bacteriol.
136, 136 (1978).
19. T. Penfold and J.W. Foster. In E. coli and S. typhimurium, Cel-
lular and molecular biology. Vol. 1, 2nd ed. ASM Press, Wash-
ington, D.C. 1996. pp. 721–730.
20. (a) R. Flachmann, N. Kunz, J. Seifert, M. Gütlich, F.-J.
Wientjes, A. Läufer, and H.G. Gassen. Eur. J. Biochem. 175,
221 (1988); (b) J. Seifert, N. Kunz, R. Flachmann, A. Läufer,
K.-D. Jany, and H.G. Gassen. Biol. Chem. Hoppe-Seyler, 371,
239 (1990).
NMR Spectroscopy
NMR spectra were acquired at McMaster and at the Uni-
versity of Hawaii. At McMaster: Bruker DRX 500 spectrome-
ter operating at 11.74 T, using a Bruker 2.5 mm microprobe;
pulse width 90o (8 µs); spectral width 28 985.5 Hz; recycle
delay 4.6 s; digital resolution 0.88 Hz per data point. At
University of Hawaii: GE Omega 500 spectrometer, using a
Nalorac 3 mm microtube; pulse width 60o; recycle delay 2 s;
32 k data points. Spectra were 125.776 Hz proton decoupled
13C NMR spectra of nicotinamide in D2O (100 µL). Approx-
imately 50 000 transients were required to generate the 13C
spectra showing satellites.
21. G.R. Griffith, J.L.R. Chandler, and R.K. Gholson. Eur. J.
Biochem. 54, 239 (1975).
© 2001 NRC Canada

Gupta et al.
425
22. S. Nasu, F.D. Wicks, and R.H. Gholson. J. Biol. Chem. 257,
626 (1982).
23. Y. Hosokawa, E. Mitchell, and R.K. Gholson. Biochem.
Biophys. Res. Commun. 111, 188 (1983).
24. S. Nasu and R.K. Gholson. Biochem. Biophys. Res. Commun.
101, 533 (1981).
25. (a) B.M. Steiner, J.T. Heard, Jr., and G.J. Tritz. J. Bacteriol.
141, 989 (1980); (b) J.T. Heard, Jr., and G. J. Tritz. Biochem.
Biophys. Res. Commun. 95, 1517 (1980); (c) J.T. Heard, Jr.,
B.M. Steiner, and G.J. Tritz. Current Microbiol. 5, 279 (1981).
26. N. Suzuki, J. Carlson, G. Griffith, and R.K. Gholson. Biochim.
Biophys. Acta, 304, 309 (1973).
27. T.P. Begley, C. Kinsland, S. Taylor, M. Tandon, R.
Nicewonger, M. Wu, N. Kelleher, N. Campobasso, and Y.
Zhang. Top. Curr. Chem. 195, 93 (1998).
28. J. Chen and G. J. Tritz. Microbios, 16, 207 (1976).
29. J.B. Foster and A.G. Moat. J. Bacteriol. 133, 775 (1978).
30. J. Arditti and J. B. Tarr. Am. J. Bot. 66, 1105 (1979).
31. R.E. Hill, K. Himmeldirk, I.A. Kennedy, R.M. Pauloski, B.G. Sayer,
E. Wolf, and I.D. Spenser. J. Biol. Chem. 271, 30 426 (1996).
32. K. Himmeldirk, B.G. Sayer, and I.D. Spenser. J. Am. Chem.
Soc. 120, 3581 (1998).
© 2001 NRC Canada