ABSTRACT

THE INCORPORATION OF ASPARTATE AND MALATE INTO
THE PYRIDINE RING OF NICOTINE

by Theodore M. Jackanicz
Nicotinic acid is a precursor of the pyridine ring of
nicotine in tobacco metabolism. In animals and certain
molds such as Neurospora crassa, tryptophan is metabolized
to nicotinic acid. But in higher plants and microgrganisms

such as Escherichia 0011, this relationship does not exist.

 

In previous investigations glycerol and a four carbon
dicarboxylic acid, intimately related to the tricarboxylic
acid cycle, have been shown to be precursors. But the
exact nature of the dicarboxylic acid directly incorporated
into the pyridine ring, has not yet been elucidated.

In an attempt to study the incorporation of dicarboxylic
acids which might be used for pyridine ring biosynthesis,
aspartic and malic acids were employed as possible pre-
cursors. Although procedures have been developed previously
for the degradation of the pyridine ring, in the present
study a method for the isolation of each carbon of the ring
was devised which required fewer steps and resulted in a
greater over-all yield than was the previous case. lTherefore
much less nicotine was required in each degradative study.

1.

Aspartate-B-Cll+ and malate-B—C1 were hydrOponically

fed to Nicotiana rustica plants. The nicotine was isolated

Theodore M. Jackanicz

as the dipicrate. converted to the hydrochloride, and oxidized
to nicotinic acid. The nicotinic acid was methylated to tri-
gonelline acid sulfate which was oxidized to N-methyl-Z-
pyridone-S-carboxylic acid. Decarboxylation of the pyridone-
carboxylic acid and catalytic reduction of the ensuing product
produced N-methyl-Z-piperidone. Hydrolysis of the piperidone,
subsequent methylation to the betaine, followed by potassium
hydroxide fusion, yielded acetic acid from carbons 5 and 6 of
the original pyridine ring. and propionic acid from carbons

2, 3, and 4. The over-all yield of the fatty acids from
nicotine was about 15%. Therefore, about 2 g of nicotine was
ample starting material for each degradation.

u was studied, 57% of the label

When aSpartic acid-B-Cl
resided in carbon 3 of the pyridine ring, whereas 38% was in
carbon 2. When malic acid-3-Clu was employed, 61% of the
tracer appeared in carbon 3, whereas 38% was in carbon 2. The
remaining small percentage of radioactive carbon was
scattered among the other carbon atoms.

These results are discussed and compared to the experi-

ments of other workers in the field of nicotinic acid bio-

synthesis.

THE INCORPORATION OF ASPARTATE AND MALATE INTO THE
PYRIDINE RING OF NICOTINE
By :L’
\ (/111-

Theodore M! Jackanicz

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Biochemistry

1965

ACKNOWLEDGEMENT

I wish to express my sincere
gratitude to Dr. Richard U. Byerrum
for his guidance, patience, and
encouragement during this investiga-
tion.

In addition, I wish to thank
the National Institutes of Health
for providing funds to aid in this
study.

Finally. I wish to thank my wife,
Sandra, for uncomplainingly enduring
numerous lonely evenings over the past
three years.

TABLE OF CONTENTS

INTRODUCTION. . . . .
EXPERIMENTAL. . . .

Preparation of Plants.

Feeding of Plants.
Isolation of Nicotine.
Degradation of Nicotine.

Nicotinic acid

Trigonelline acid sulfate.

N-methyl- 2-pyridone-5-carboxyli

N-methyl- 2-pyridone.
N-methyl- 2-piperidone.

N-methyl— S-aminopentanoic acid
5- trimethylaminOpentanoic acid

PrOpionic and Acetic Acids .
Mechanism of Cleavage of 5- trimethylamino-

pentanoic acid

0
o

oooomcococco

coocpooocooo

O O O O O O O O O O C O

0 O O O O O O O O O O O 0

Degradation of Propionic and Acetic Acids.
Radioactivity .

Determination of
RESULTS . . . . . . .
DISCUSSION. . . . . .
SUMMARY . . . . . . .
BIBLIOGRAPHY. . . . .

11

I O O O O O O O O O O O I O O O O

LIST OF TABLES
Table
1. Isotope Distribution in the Pyridine Rigg of 26

Nicotine ngm Plants Fed Aspartate-B-C and
Malate-B-C ,

iii

Figure

LIST OF FIGURES

Degradation Sequence for the Pyridine Ring of
Nicotine.

Purification of Nicotinic Acid on Dowex-l-
Formate.

Separation of Propionic and Acetic Acids on
Chromosorb.

IsotOpe Distribution in the Pyridine Ring of
Nicotine Frfim Plants Fed ASpartate-B- C and
Malate-B- Cl

PrOposed Biosynthetic Pathway for Nicotinic
Acid in Higher Plants and Microorganisms.

iv

13

20

27

35

INTRODUCTION
The pyridine ring occurs in several alkaloids as well
as the vitamin niacin. and several coenzymes. Previous re-
search has shown that nicotinic acid is a precursor of the
pyridine rings of nicotine and anabasine as well as the
pyridone moiety of ricinine (1,2,3). Thus the problem of
pyridine ring synthesis in nicotine can be viewed as a study

of nicotinic acid biosynthesis.

OCH3
\ \ \ CN
N l N I
CH3 1% H
. C
Nicotine Anabasine Ricinine

In 19#9, Heidelberger etaal. found that rats could
convert tryptOphan to nicotinic acid (4). It was also
demonstrated that molds such as Neurospgra crassa could
synthesize nicotinic-acid from tryptophan (5). But three
separate groups found that tryptOphan was not converted to
nicotinic acid in higher plants (2,6,7). Yanofsky also
reported the same phenomenon in Escherichia 29;; (8).
Therefore a different pathway must exist in these latter
systems.

Ortega and Brown found that in resting g. ggl;_cells,

4 or succinate-2,3-Clu labeled mainly the

glycerol-1,3-C1
pyridine ring of nicotinic acid (9.10). But when succinate-

1,4—Cln was fed to this organism, most of the label was found

1

in the carboxyl group of nicotinic acid. From their limited
degradation data, they concluded that glycerol or a closely
related metabolite was condensed with a four carbon di—
carboxylic acid in such a way that one of the carboxyl groups
of the acid became the carboxyl group of nicotinic acid.

As previously mentioned, nicotinic acid was also found
to be a precursor of ricinine. Heller and Henderson fed
several C14 labeled precursors to castor bean seedlings.

They decided, as did Ortega and Brown, that glycerol or a re-
lated three-carbon compound condensed with a four-carbon
dicarboxylic acid so that one of the carboxyl groups became
the carboxyl group of nicotinic acid (ll).

Three groups have studied the Nicotiana alkaloids,
nicotine and anabasine. Leete found that acetate-Z-Clu was
incorporated into the pyridine ring of anabasine (12).
Griffith and Byerrum demonstrated that acetate-Z-Cln was
efficiently' incorporated into the pyridine ring of nicotine,
and that it was a vastly superior source compared to
acetate-l-Cln (13). Subsequently they demonstrated that both
glycerol-1,3-Cll+ and glycerol-Z-Clu were extensively in-
corporated into the pyridine ring. From their incorporation
data, they concluded that glycerol was incorporated as a
three carbon piece without fission of a carbon-carbon bond (lN).

In contrast, Dawson and Christman have studied nicotine

biosynthesis in Nicotiana tobaccum and believe that glycerol

is converted to a two carbon compound which then condenses
with citric acid to yield an intermediate which eventually
forms nicotinic acid (15). They based their conclusion on
the fact that in their studies, glycerol—Z-Clh was incor-
porated twice as efficiently as glycerol-1,3-C1b into the
pyridine ring. This indicated that one of the labeled
carbons in the 1,3-C1u labeled species-was cleaved.

Up until this point, the biosynthetic studies mainly
concerned incorporation studies into the pyridine moiety,
with no specific degradation to ascertain the actual loca-
tion of radiocarbon. Then in 1963, Griffith and Byerrum
developed a partial.degradation sequence whereby carbons 2,
3. and 6 were isolated,.and their specific activities
determined (16). After feeding.acetate-2-Clu and succinate-
2,3-cl". they found that about 75% of the label waseqnally
distributed between carbons 2.and 3 with only a few percent
located in carbon 6. The radioactivity of the nicotine
after feeding succinate was twice as great as nicotine from
the acetate.

More recently, Fleeker fed both glycerol-1,3-Clu and

~01” and determined the specific activity of each

glycerol-2
carbon atom in the pyridine ring of the biosynthesized'

nicotine (17). He found that 98% of the glycerol-Z-Cln was
incorporated into carbon 5. When glycerol-1,3-Cll+ was fed,
almost 70% of the label was equally divided between carbons

4 and 6, indicating that glycerol can provide carbons 4, 5,

and 6 of the ring. Almost 30% of the label resided in
carbons 2 and 3. The conversion of glycerol to acetate,
via glycolysis, and the entrance of the acetate into the
tricarboxylic acid cycle can explain this additional
incorporation.

As previously mentioned, Griffith and Byerrum found

that succinate-2,3-Clu

was incorporated into carbons 2 and

3 of the ring. Preliminary experiments in this laboratory
with aspartate-3—01“ had indicated that about half of the
incorporated label resided at carbon 3.of the pyridine ring,
and it was postulated that an equal percentage was located
at carbon 2. But since no other ring carbons could be
isolated at that time, it was not possible to draw any
definite conclusions concerning the biosynthesis of the
pyridine ring from aspartate. Since both succinate and
aSpartate were incorporated into the pyridine ring, it would
be of interest to see whether aSpartate, or another closely
related asymetric metabolite, were incorporated with
randomization of the label between carbons 2 and 3. or
whether the tracer might be specifically distributed, favoring
one carbon atom over the other. If the specific activity of
carbons 2 and 3 of the pyridine ring were appreciably
different after feeding.aSpartate-3-Clu, such results would
indicate that aSpartate was a more direct intermediate than
succinate or some other symmetrical metabolite of the tri-
carboxylic acid cycle. These results would also disclose

perhaps which of the two carboyxl groups becomes the carboxyl

group of nicotinic acid.

5

Several other degradations of the pyridine ring have
been published; all of them begin with nicotinic acid as the
product of the oxidation of nicotine or anabasine.

Fleeker converted nicotinic acid to S-trimethylamino-
pentanoic acid by eight separate reactions.(l7). The betaine
was then cleaved to acetate and propionate which were
degraded, via the Schmidt Procedure, to carbon dioxide, and
the specific activity of each carbon of the original pyridine
ring was determined. Leete has deve10ped a degradation
whereby nicotinic acid is converted to 1,3-dimethyl-2-phenyl-
piperidine (42,47). This product was divided into two
samples. One portion yielded carbons 2 and 3 of the original
pyridine ring after two additional reactions; the remaining-
sample was used to obtain carbon 6 after four reactions and
finally carbon 5 after three more reactions. Carbon 4 was
then obtained by difference.

Finally, Christman and Dawson have published a procedure
for degradation of the pyridine ring of nicotine (48).
Nicotinic acid was converted to nicotinamide which was
methylated to trigonelline amide and then oxidized with
potassium ferricyanide. The products of the oxidation were a
mixture of the corresponding 2- and 6-pyridones of trigonelline
amide, which were separated by column chromatography. The two
samples were then separately treated as follows. The amide

function was hydrolyzed to a carboxyl group which was then de-

carboxylated to N-methyl-Z—pyridone. One of the pyridone
samples would more accurately be termed N-methyl-6-pyridone,
since it represented the carbon atoms of the original
pyridine ring in a different manner than the other pyridone
sample. The pyridone, through six additional reactions,
yielded pentanoic acid which was degraded by the Schmidt
Procedure to obtain each separate carbon atom of the pyridine
ring. ‘

Therefore, the purpose of this study was to investigate
the incorporation of aspartate-3-Clu and malate-3-Cll+ into
nicotine and to develOp improved methods for degrading the
pyridine ring. Utilizing the degradative procedure developed
in this investigation revealed that about 60% of the tracer
resided in carbon 3 while about 40% appeared in carbon 2
after feeding both aSpartate-3-C1u and malate-3-C1u. Carbons
4, 5, and 6 contained essentially no radioactivity.

Dilution data revealed that aSpartate might be a more direct

precursor of the pyridine ring than malate.

EXPERIMENTAL AND RESULTS
Preparation of Plants

Due to their relatively high nicotine content,
Nicotiana rustica, var. humilus plants were used for the
present experiments. The plants were grown in a commercial
plant growth chamber at a day temperature of 27°C and a
night temperature of 25°C. exposed to 425 foot candles of
light for 16 hours per day. The seeds were planted in a
flat of vermiculite (commercial name for heat expanded mica)
and usually had germinated in ten to fourteen days. Between
one and two weeks after germination, each seedling was
transplanted into flats, atea distance of 5 cm from its
neighbor. The plants were watered as necessary (usually
every other day) and once a week were treated with a
nutrient solution of 0.1 g MgSO4o7H20, 0.6 g Ca(N03)2-4H20,
and 0.1 g KZHP04 per liter. I

At a height of 12-l5 cm the plants had reached the
Optimum state for feeding experiments. Since the rate of
gg’ggzg nicotine synthesis is prOportional to the rate of
new root growth (18), the roots were cut to a length of
about one cm, and the residual vermiculite rinsed with tap
water. A cotton plug was wrapped around the stem of each
plant for support and each was placed in a 125 ml Erlenmeyer
flask containing sufficient nutrient solution to cover the
roots (19). To promote new root formation and decrease the

growth of photosynthetic microBrganisms, the flasks were

7

wrapped with black construction paper (20). Every four or
five days the nutrient solution was discarded, the roots
washed with tap water, and new nutrient solution added.
Feeding of Plants

After two to three weeks of hydroponic growth, new
root systems had develOped and the plants were ready for
nicotine synthesis studies. The plants were removed from
the Erlenmeyer flasks, the roots rinsed with distilled
water, then blotted dry with tissue paper, and inserted into
15 cm x 5 cm beakers containing the desired precursor. Each
plant received 0.5 ml of aqueous solution containing 10 uc

of Cl“ present as 2.44 x 10'2 mmoles of the compound fed.

During the metabolic period, a few drops of water were added
to each beaker to keep the roots moist and also to rinse the
containers to insure maximum uptake of precursor.

ASpartate-3-C1u and malate-3-Clu, which were used as
precursors in the present study, were purchased from Volk
Radiochemical Company. Radioautography indicated that the
compounds were radiochemically pure. The Specific activity
of each sample was checked.

Isolation of Nicotine

After a four hour feeding period, the roots and beakers
were rinsed with distilled water, and the washings saved for
the later determination of unabsorbed precursor. The plants

were cut into small pieces.and ground in a Model CB-S Waring

Blendor for one minute at high speed using a ratio of one

9

liter of water per fifteen plants. The slurry was trans-
ferred to a three liter, ground glass, round bottom flask
and was made alkaline with calcium oxide. The mixture was
heated with a burner. the nicotine distilled azeotrOpically
with water, and the mixture collected in a two liter ground
glass round bottom flask, containing a few ml of 6N HCl.

To avoid charring the plant material, water was added at
intervals to the distillation flask so that the volume was
never less than one liter.

The distillation was continued until the distillate
failed to give a precipitate with silicotungstic acid, a
general reagent for detecting alkaloids (21). The distillate
was evaporated to dryness in 32339 leaving crude nicotine
dihydrochloride. The white precipitate was dissolved in a
few ml of water. transferred to a steam distillation flask,
made alkaline, and steam distilled. Again the distillate
was collected in 6N HCl, evaporated to dryness in 13339, and
the crude product dissolved in a small volume of methanol.
Addition of an equal volume of picric acid saturated methanol
to this solution precipitated nicotine dipicrate as a yellow
amorphous mass. The nicotine dipicrate was filtered on a
medium porosity fritted funnel, and was recrystallized from
water yielding yellow needles melting at 224-226°C*. The
reported value is 224—225°C (22).

“All melting points were corrected.

Degradation of Nicotine

After aspartate-»3«---Cll+ and malate-3-C1u were found to be
efficient nicotine precursors, the next step was to degrade
the pyridine ring to determine the Specific activity of each
carbon atom. The degradation sequence is shown in Figure l.
Nicotinic acid

The weighed nicotine dipicrate was transferred to a
steam distillation flask, dissolved in water, and made
alkaline with sodium hydroxide pellets. The nicotine was
steam distilled into 6N HCl until no precipitate with silico-
tungstic acid was observed in the distillate. The nicotine
dihydrochloride was transferred to a four liter Erlenmeyer
flask and about one liter of water was added. The pH was
adjusted to 9 with dilute potassium hydroxide. Then two
liters of potassium permanganate solution (eight moles per
mole of nicotine) was added dropwise over a one hour period
while the nicotine solution was stirred vigorously. The
flask was heated on a steam bath to complete the reaction and
to settle the manganese dioxide which had precipitated (23).

After 16 hours of heating the solution, ethanol was
added until the purple permanganate color disappeared. The
suspension was filtered first through a medium and then a
fine porosity fritted funnel. The manganese dioxide was
washed three times with 50 ml portions of hot water and the

washings were added to the filtrate. The water was removed

10

Figure l .

 

 

 

\ 1») coon ' cool 6)
_§Mn0u .‘(5) /’ l 3) (SE312§24, // l g H804
/ 7 '
633 (6)‘\N 2) 9//
CH
u 3
K3Fe(CN)6
OH'
V/

 

 

 

CH '
Ba(OH)2
t
A820 \ e 0
H000 CH I *7
31:1 3 °° i?
CH3 (533)3

11

 

(u)
HOOC

(5) CH3 0 2 (3)
O

350'C (6) coon CH3 (2)

HN3

KMnOu

 

C02

12

in 22239, and the white residue of impure nicotinic acid was
dissolved in about 50 ml water. Formic acid, 6N, was added
to give a solution of pH 1 to release the carbon dioxide
resulting from carbons 3, 4, and 5'of the pyrrolidine ring
plus the N-methyl carbon. The solution was then titrated to
a phenolphthalein endpoint with 2N potassium hydroxide. The
solution was concentrated to about 10 to 20 ml, placed on a
column of Dowex-l-formate, and eluted by the method of
Preiss and Handler (24). For a sample of 750 mg nicotine, a
column 60 cm x 3.5-cm sufficed.

Elution of the column with water and three different
concentrations of formic acid, released three ultraviolet
absorbing Species plus nicotinic acid. A typical elution
pattern is shown in Figure 2. The peaks were measured by
determining the absorption at 261 mu in a Beckman DB
SpectrOphotometer. The nicotinic acid fractions were combined
and the total quantity of compound present determined by
measuring the ultraviolet absorption, and comparing it to a
standard curve prepared from nicotinic acid samples of known
concentration. The identity of nicotinic acid was confirmed
by ultraviolet Spectrosc0py (31), paper chromatography, thin-
layer chromatography, and its melting point of 233-235°c.
The published melting point is 232°C (20).

 

0.05N 0.10 N
100 I‘ HQO COO HCOOH 0.2 N HCOOH

 

 

 

 

 

 

75-
Absorbance
at 261 mu
50F
25»
A /2'\0 [4U\ 60 'véuv
TUBE

Figure 2. Purification of Nicotinic Acid on Dowex—l-
Formate.

Triggnelline acidgsulfate

The nicotinic acid obtained in the previous reaction
was diluted with cold compound to give 4.3 g of nicotinic
acid (35 mmoles) which with 5 ml of dimethyl sulfate were
transferred to a 100 ml round bottom flask, equipped with a
Liebig condenser and a calcium chloride drying tube. The
sample was heated at 130°C for four hours and every 15 to
30 minutes the solution was stirred by means of a magnetic
bar (25). The flask was then cooled in ice, and the contents

were diluted with 50 ml of water plus 3.5 ml of 0.5 N H2804.

13

14

A little Norite was added, the flask heated on the steam
bath for a few minutes, and the mixture was filtered through
a medium porosity fritted funnel. The solution was
concentrated to about 10 ml by heating on a hot plate, and
was then diluted with absolute ethanol to provide a solution
90% in ethanol. Crystals started to form after a few
minutes. After about 30 minutes at room temperature, the
beaker was placed in the refrigerator for 30 minutes to
facilitate further crystallization. The crystals were
filtered on whatman No. 1 filter paper in a Bdchner funnel.
Concentration of the mother liquid, again followed by
addition of alcohol to 90% yielded a second crop of crystals
to give a total quantity of 6.4 g of trigonelline acid
sulfate which corresponded to 77% yield. Both samples
melted at 198-2oo°c. The recorded value is 199—20100 (25).
If one saves the final mother liquid and separately uses it
in the next reaction, an additional 300 mg of somewhat less
pure N-methyl-2-pyridone—5-carboxylic acid can be obtained.
N-methyl-2-pyridoneg5rcarngylic acid

Over a 90 minute period, 71 ml of a 32% solution of
potassium ferricyanide was added drapwise tosa stirred
solution of 6.4 g of trigonelline acid sulfate dissolved in
93 ml 2.5 N sodium hydroxide (26). After 45 minutes of
additional stirring at room temperature, the pH was adjusted
to 3.7 with 6 N H2804. Slowly needle-like crystals started

to form and after 1 hour at room temperature, the flask was

15

cooled in the refrigerator for an additional hour. The
crystals were filtered through Whatman No. 1 filter paper
in a Bflchner funnel and were washed with ice cold water
until they were white. The washings and mother liquid
were combined, the pH of the solution was lowered to 3.5.
and the above purification steps were repeated to yield a
second crOp of crystals. A total of 2a? g of product was
obtained which corresponded to 64% yield. An additional
150 mg of product can be obtained by continuous ether
extraction of the final mother liquid. The crystals melted
at 240-242°C; the published value for N-methyl-Z-pyridone-
5-carboxylic acid is 240-24l°C (26). The compound was
further characterized by its ultraviolet absorption peaks
at 254 mu and 298 mu (2?) and by an elemental analysis which
gave the following results:

C7H7NO3* Found: C, 54.90%; H, 4.62%; N, 9.13%

(153.18) Calculated: C, 54.90%; H, 4.6lfl; N, 9.21%

Since N-methyl-Z-pyridone—3-carboxylic acid has a
melting point of 183-184°C and an ultraviolet absorption
maximum at 320 mu (27), there is no question as to the

possibility of a mixture of the two acids being present.

*Microanalysis performed by Spang Microanalytical Laboratory,
Ann Arbor, Michigan.

N-methyl-2:pyridone

All of the pyridone carboxylic acid from the previous
reaction (2.70 g), 25 ml of redistilled quinoline, and 900
mg of freshly prepared c0pper chromite catalyst (28) were
placed in a two neck 50 ml round bottom flask. The flask
was fitted with a reflux condenser, immersed in a silicone
oil bath, and heated at 250-260°c until carbon dioxide was
no longer collected in a barium hydroxide trap which was
connected to the top of the condenser (29). The reaction
was usually completed in about five hours, whereupon the
mixture was cooled, and filtered through a medium porosity
fritted funnel into a long neck distillation flask. The .
quinoline was removed by steam distillation. When the dis-
tillate no longer gave a precipitate with silicotungstic
acid, all of the quinoline had been removed. The residue
was extracted five times with 70 ml portions of chloroform.
Removal of the solvent left a yellow oil with ultraviolet
absorption peaks at 225 and 298 mu. The published values
for N-methyl-Z-pyridone are 226 and 300 mu (30).
N-methyl-Z-piperidone

All of the N-methyl-2-pyridone from the previous
reaction, 30 m1 of 95% ethanol, and 200 mg of Englehard
brand ruthenium dioxide catalyst were transferred to a 100
ml flask designed especially for reductions. The flask was
placed in a Paar high pressure rocking bomb under 1000

pounds of pressure for two hours and was slowly heated to

16

17

100°C over this period (32). The mixture was filtered
through a fine porosity fritted funnel to remove the
catalyst. The lack of ultraviolet absorption at 298 mu
indicated essentially complete reduction. The solvent was
removed in 13233 leaving a very pale yellow oil.
N-methyle5:aminopentanoic acid

All of the product from the previous reaction was
refluxed in 75 ml of water for six hours with 6 g of the
octahydrate of barium hydroxide (33). The flask was
cooled, chips of dry ice were added to precipitate the
barium ions, and the barium carbonate precipitate was re-
moved by filtration through Whatman No. 42 filter paper
in a Bflchner funnel. Removal of the water under reduced
pressure left an oil which crystallized under vacuum
desiccation and melted at 125-126°C. The reported value
for N-methyl-S-aminOpentanoic acid is 126-12700 (3n).
5:trimethylaminopentanoic acid

All of the acid from the previous experiment was dis-
solved in 75 ml of methanol and transferred to a 250 ml
round bottom flask. Twenty g of freshly prepared silver
oxide and 5.0 ml of methyl iodide were added (35). A
stirring bar was placed in the flask and a Liebig condenser,
fitted with a calcium chloride drying tube, was inserted.
The flask was placed in a 70°C oil bath and stirred for 12
hours. During the reaction period, the flask was placed in

a dark room to lessen silver salt decomposition.

18

The silver salts were removed by filtering through a
fine porosity fritted funnel, and the solvent was collected
in a 50 ml long neck round bottom flask. Evaporation of
the methanol yielded an oil which crystallized under vacuum
desiccation. The betaine melted at 225-228°C (dec.) which
compares favorably with Christmanis value of 225°C (dec.).
(48).

Propionic and acetic acids

 

Ten g of potassium hydroxide was added to the betaine
and the long neck flask was immersed in a metal bath which
had been preheated to 300°C (35). The temperature was
quickly raised to 350°C and held there 1 5°C for 10 minutes.
The flask was cooled, the contents were dissolved in water,
and then transferred to a 125 ml steam distillation flask.
All residual volatile alkaline and neutral compounds were
removed by steam distillation of about 250 ml of water.
Addition of excess 6 N HZSOu to the remaining solution
generated prOpionic and acetic acids which were steam dis-
tilled and titrated to pH 8 with standard sodium hydroxide,
as determined with a pH meter.

Assuming that equal quantities of acetic and prOpionic
acids were present, the yield of acids from N-methyl-Z-
pyridone-S-carboxylic acid was about 40%. The water was re-
moved under reduced pressure leaving the sodium salts of the

acids, which were dissolved in about 0.2 ml water. Then one

19

drop of phenol red was added, followed by the addition of
solid potassium bisulfate until the indicator turned yellow.
At this point the distinctive odor of the fatty acids was
noticeable. The mixture was extracted with five 3 ml
portions of 5% n-butyl alcohol in chloroform and placed on a
Chromosorb* column 51 cm x 2.7 cm. The column had been
prepared by mixing 70 g of Chromosorb with 56 ml of 0.5 N
H280“ and then suSpending the mixture in chloroform pre-
viously equilibrated with 0.5 N H2804 (36).

Elution of the column with chloroform removed the
propionic acid; 10% n-butyl alcohol in chloroform eluted the
acetic acid. The contents of each tube from the column was
mixed with an equal volume of methanol and titrated to a
bromothymol blue endpoint with standard sodium hydroxide.
See Figure 3 for a typical elution pattern. The tubes con-
taining the two acids, as determined by titration, were
treated separately as follows.

The chloroform and methanol were removed in 13222
leaving the sodium salt in a few ml of water. The sclution
was transferred to a steam distillation flask, acidified
with 6 N 3230“, and the fatty acid was steam distilled. A
small aliquot was removed for paper chromatography to check

for cross contaimination between the two acids or the

*A coarse type of diatomaceous earth from Fisher Scientific.

20

presence of other acids (37). The remaining acid was
titrated with standard sodium hydroxide to pH 8 as measured
with a pH meter, and was concentrated to a known volume.
Aliquots were removed for counting and the rest used for

further degradations.

 

 

 

 

 

 

 

CHCI 10 n-BuOH in CHCl
3 3*; z 3
8 -
6,_
ml 0.100N
NaQH
4..
2»
£0 ' do ' 6b "
TUBE

Figure 3. Separation of PrOpionic and Acetic Acids on
Chromosorb.

Mechanism of Cleavage of 5-trimethylamin0pentanoicacid

In 1840, Varrentrap found that fusion of potassium
hydroxide pellets with oleic acid yielded acetic and palmitic
acids (38). Edmed observed a similar phenomenon by converting
octadec-Z-enoic acid to hexanoic and acetic acids (39). In
addition Cornforth and Hunter cleaved 2-hexenoic acid-l-Cll+
to butyric acid plus acetic acid-l-Clu (40). Also,

Cornforth et al. found that when 6-trimethylaminoheptanoic
acid was fused with potassium hydroxide, pentanoic and acetic
acids were the only products. All of these reactions bear.
striking similarities to the cleavage of the pentanoic acid
betaine.

As final proof of the mechanism of the cleavage, Fleeker
chemically synthesized nicotinic acid-4,6-C1u and via a
somewhat different degradative scheme than described in the
present study, converted it to 5-trimethylaminOpentanoic
acid (17). Potassium hydroxide cleavage, followed by
separation and subsequent radioassay of the acids, indicated
that cleavage occurred between carbons two and three of the
betaine.

Cornforth postulates a cleavage mechanism as follows:
H0-

' e ‘\‘H e e
(033)3EJ-CHZQH-CH2-CHZ-COO __. CH2=CH-CH2-CH2-COO ———.
4.
(CH3)3N

e e
CH3-CH-CH=CH-coo° ————: 033-032-000 +,CH3-COO

21

Degradation of Propionic and Acetic Acids

For the complete degradation of prOpionic acid, about
0.5 mmole of the sodium Salt was dried under vacuum in a 10
ml pear shaped ground glass flask (41). The flask was
cooled in ice, then 0.3 ml of 100% H2804 was added, and the
flask was gently shaken to dissolve the sodium propionate.
Again the flask was cooled and 50 mg (0.77 mmole) of sodium
azide was added. The mixture was warmed and gently shaken
to partially disSolve the azide. The flask was connected to
a train consisting of a tube of 5% potassium.permanganate
in 0.5 N H250“ and a tube of saturated barium hydroxide to
trap the carbon dioxide. Nitrogen was then bubbled through
the system and the reaction temperature was raised to 35°C
with a water bath. Over a 30 minute period the temperature
was raised to 60-70°C, and held there for an additional 30
minutes at which time the reaction was essentially completed.
The barium carbonate was filtered, dried and prepared for
radioactivity assay.

In order to recover the resulting ethylamine, the
permanganate tube was replaced with a tube containing 5 ml
of 0.2 N H280“; 3.5 ml of 5 N NaOH was added to the reaction
flask to liberate the ethylamine. The system was swept for
15 minutes with the bath temperature at 90-100°C.

The ethylamine sulfate solution was transferred to a 50
ml round bottom flask containing 5 ml of 5% potassium
permanganate, the flask was tightly stoppered, and the mixture

22

23

was heated in a 90-100°C water bath for 15 minutes. The
solution was transferred to a steam distillation apparatus,
acidified with 6N H280“, and the acetic acid steam dis-
tilled. The distillate was titrated, concentrated to a
small volume and transferred to a 10 ml pear shaped flask
where it was dried under vacuum.

The dry sodium acetate (about 0.3 mmoles) was treated
in the same manner as the sodium prOpionate, using pro-
portionally smaller quantities of sulfuric acid and sodium
azide. However, the initial bath temperature was 45°C and
was raised in 10 minutes to a final temperature of 70°C
which was maintained for 45 minutes.

The methylamine was recovered in the same manner as
the ethylamine. The methylamine sulfate was transferred to
a 25 m1 pear shaped flask containing 3 ml of 5% potassium
permanganate; 0.6 ml of 5N NaOH was added, the flask was
tightly stoppered, and the contents were heated for 15
minutes at 90-100°C. The flask was cooled, connected to the
usual train, and 6N H280“ was added to release the carbon
dioxide: The system was swept for 15 minutes with heating,
and the barium carbonate was treated in the usual manner.

«uDetermination of Radioactivity

Carbon dioxide was counted as barium carbonate in a

Nuclear-Chicago proportional gas flow counter, Model D-47,

with an automatic sample changer and Nuclear-Chicago Model

24

192A ultrascaler. The efficiency was determined and samples
corrected for self absorption. All other counting was per-
formed in a Packard Tricarb Scintillation Spectrometer, Model
3314. A one m1 aqueous aliquot of the radioactive sample was
mixed with ten ml of a solution consisting of 100 g of
reagent grade naphthalene, 10 g of 2,5-diphenyl-oxazole, and
250 mg of l,4-bis-2-(5-phenyloxazolyl)-benzene, dissolved in
1 liter of dioxane. All reagents were spectroquality. The
efficiency was determined and all compounds checked for
quenching. Since the absolute Counting efficiencies were
calculated, radioactive determinations in both systems could

be compared.

RESULTS

As mentioned above, the prOpionic and acetic acids
obtained from the potassium hydroxide cleavage were de-
graded so that each separate carbon atom of the original
pyridine ring could be isolated and its specific activity
determined. The results of this investigation are shown
in Table 1. when aspartate-3-Clu was used as a precursor,
57% of the label resided at carbon 3 of the pyridine ring
and 38% at carbon 2. Essentially the same results were
observed with malate-3-C14 experiments. Sixty-one percent
of the radioactivity appeared at carbon 3 of the pyridine
ring and 38% at carbon 2. In both experiments, an
insignificant percentage of the tracer appeared at carbons
4, 5, and 6. A visual presentation of the data appears in
Figure 4.

Dilutions* for aspartate ranged from 360 to 800 with
an average value of 600. For malate the values ranged from

1260 to 3270 with an average value of 2200.

“Specific activity of the precursor divided by the specific
activity of the nicotine isolated.

25

Table 1. Distribution of cl4 in the pyridine Rin of
am Plants Fed Aspartate-3-Cl

 

Nicotine Fr
Malate-3-C1
ASpartate-3-Clu
Compound S . Act. Percent*
opm/MM
N-methyl-S-
aminOpentanoic
acid 8480 100
Barium Carbonate
(carbon 2) 3220 38
Barium Carbonate
(carbon 3) 4840 57
Barium Carbonate
(carbon 4) 93.4 1.1
Barium Carbonate
(carbon 5) 288 3.4
Barium Carbonate
(carbon 6) 204 2.4

waists—3-01”

 

Sp. Act. Pgrggnt
2350 100

894 38
1430 61

19 0.6

33 1.4

24 1.0

*Percentage of radioactivity in a carbon atom as compared
to N-methyl-5-aminopentanoic acid. 3

26

 

1.1 '
.4
3 ,/’ 57 \\N I

2.4 \\ 38 |
N CH3

Nicotine from aspartate-3-Cll+

 

Nicotine from malate-3-C14

Figure 4. Percent incorporation of C1“ for the
pyridine ring of nicotine from Eeeding
experimentiuwith aspartate-3-C1 and
malate-3-C .

27

DISCUSSION

The results of the present experiments indicate that
aspartic and malic acids are efficient precursors of the
pyridine ring of nicotine. Furthermore the two acids seem
to be incorporated so that their 2 and 3 carbons respectively,
become the 2 and 3 carbons of the pyridine moiety. The
presence of label in carbon 2 of the pyridine ring after
feeding the 3-labeled precursors can be explained by assuming
that some of the malate is dehydrated to fumarate and that
some of the aspartate is deaminated to fumarate. Some of the
aspartate can also be deaminated to oxalacetate which can then
be metabolized to furmarate via malate. Since fumarate is a
symmetrical compound, the label is randomized between carbons
2 and 3. The fumarate is then converted back to malate and
aspartate. In effect, these malate and aspartate molecules
now have their isotope equally distributed between carbons 2
and 3. These randomized species can now enter the pool of
the carbon 3-labeled precursors and proceed into the pyridine
ring. or course one must assume that the inter-conversion of
these four carbon acids is such as to allow this randomiza-
tion process to occur to the extent that was observed.

One method of reducing randomization would be to feed
the precursor to the tobacco plants for a shorter time period,
perhaps for one hour instead of the usual four hours. Al-
though the actual incorporation of carbon-l4 into nicotine
would decrease during a shorter metabolic period, the 3-labeled
precursor would be less susceptible to randomization by inter-
conversion among other four carbon dicarboxylic acids.

28

29

Since the label in carbons 2 and 3 of the pyridine ring
is not equally distributed after feeding aspartate-3-Clu and

Ola, succinate, fumarate, and any other symmetrical

malate-3-
four carbon dicarboxylic acid may be eliminated as direct
precursors of the pyridine ring. This leaves malate,
aSpartate, and oxalacetate as possible choices. But perhaps
some other unknown four carbon acid could be the actual
precursor.

As mentioned earlier, aspartate has a lower dilution
factor than malate, and therefore seems to be a more direct
pyridine ring precursor. Dilution data, however, possess
limited value due.to effects such as metabolic pool sizes
and turnover rates of these pools. If aSpartate is a more
direct precursor of carbons 2 and 3 of the pyridine ring than
malate, it must have a dilution equal to or lower than
malate, since the malate would be diluted by non-radioactive
intermediates between malate.and aspartate in the pathway.
However, if oxalacetate were.in effect the.actual precursor,
conceivably the aspartate might be converted to oxalacetate
faster than malate is, and thereby seem to be a more
immediate precursor of the ring. On the other hand, if the
malate pool were considerably larger than the aspartate pool,
the malate-C14 would be diluted to a greater extent with non-
radioactive compound than aspartate was, and therefore appear

to be a more remote ring precursor than aspartate.

30

In order to determine the isotOpe distribution in the
pyridine ring, it was necessary to devise a degradative
procedure which permitted the isolation of each individual
carbon atom. The degradation developed in this study
proved to be shorter, faster to perform, and required less
starting material than previous procedures. Nicotine was
oxidized to nicotinic acid which was methylated to tri-
gonelline. Ferricyanide oxidation of this product yielded
only N-methyl-2-pyridone-5-carboxylic acid. The
correSponding isomer is not observed when trigonelline is
oxidized instead of trigonelline amide. After four addi-
tional reactions 5-trimethylamin0pentanoic acid is obtained.
This product, when fused with potassium hydroxide, yielded
acetic acids from carbons 5 and 6 of the original pyridine
ring, and prOpionate from carbons 2, 3,.and 4. The over-all
yield of fatty acids from nicotine was about 15%, and as.a
result about 2 g of nicotine sufficed for starting material.

As previously mentioned, nicotinic acid is a.precursor
of the pyridine rings of nicotine and anabasine, as well as
the pyridone ring of ricinine. Therefore, subsequent dis-
cussions will in effect concern nicotinic acid biosynthesis.
Several other research groups studying alkaloid biosynthesis
have also been able to isolate carbons 2 and 3 of the pyridine
ring after feeding carbon-l4 labeled tracers. Leete fed
acetate-Z-Clu to Nicotiana glauca plants and isolated radio-

31

active anabasine (42). Essentially all of the pyridine ring
activity was equally distributed between carbons 2 and 3.
Juby fed succinate-2,3-Clu to castor bean seedlings (43).
Isolation of ricinine synthesized by these plants and subse-
quent degradation of it, disclosed that carbons 2 and 3.each
contained about 38% of the label. The remaining activity
was associated with the cyanogen carbon which originally was
the carboxyl carbon of nicotinic acid.

In the two previous experiments, carbons 2.and 3 were
equally labeled, and no information could be obtained as to
the exact manner in which these precursors were incorporated
into the ring. But Gross and co-workers chose to use
aspartate-1,4-C14,N15 in a nicotinic acid biosynthesis study
in Mycobacterium tuberculosis (44). They found that 48% of
the radioactivity fed could be accounted for in the carboxyl
group of the isolated nicotinic acid.. Since during
biosynthesis of nicotinic acid one of the carboxyl groups of
asPartate was cleaved, the maximum expected incorporation of
carbon-l4 would be 50% of the original precursor. The 48%
value is in excellent agreement. Furthermore, their mass
Spectrometric le/Clu ratio was 0.91, indicating that the
nitrogen of the aspartate was incorporated with the carbon
skeleton. These observations support the role of aspartate
as the four carbon dicarboxylic acid precursor of nicotinic

acid.

32

Since Fleeker found that glycerol-1,3-Clu was incor-
porated into carbons 4 and 6 and that glycerol-Z-C14 was
incorporated into carbon 5 (17). a plausible view of nicotine

synthesis is:

A H OH 0 H —coos ° °
2 i 2 . / cos
' HOH + * H-COOH -——9 -——9 -——> I -—+«—+ Nicotine
A‘\ *
ACHZOH N32 N/

Two other research groups have attacked nicotinic acid
biosynthesis, but at a point subsequent to the condensation
of the two units which form the ring system. Since nicotinic
acid is a six carbon compound and the two suspected precursors
contain seven carbons, somewhere along the pathway, one carbon
atom must be cleaved from one of the intermediates. Of
course, there is the possibility that this carbon is lost
during the condensation reaction by some type of concerted
mechanism. But Andreoli et a1. investigated quinolinic acid
(2,3-pyridine-dicarboxylic acid) as a possible precursor of
nicotinic acid (45). They prepared a cell-free enzyme system
from E. 23;; which was incubated with quinolinic acid and 5-
phosPhoribosyl-l-perphosphate for three hours. They then
identified nicotinic acid, nicotinic acid ribonucleotide,
deamido-NAD, and NAD as reaction products in a ratio of about
42:15:1:4. These researchers believed that they had prepared
a crude enzyme system which catalyzed the condensation of

quinolinic acid and 5-phophoribosyl-l-perphOSphate to the

33

corresponding nucleotide, which was then decarboxylated to
nicotinic acid ribonucleotide. Furthermore they postulated
that the nicotinic acid obtained was a product of the
cleavage of the nucleotide, since when the 5-phosphoribosyl-
l-perphOSphate was omitted from the reaction medium, no
nicotinic acid was detected.

In separate experiments, Hadwiger et a1. studied
quinolinic acid metabolism in two different systems (46).

First, quinolinic acid-2.3.7.8-C11+

was fed to corn seedlings.
Radioactive nicotinic acid was isolated from the plants,

thus confirming the fact that quinolinic acid can be con-
verted to nicotinic acid. In a second experiment quinolinic
acid-2,3,7,8-Clu was fed to castor bean seedlings, and
radioactive ricinine was isolated. This finding again.
illustrated that quinolinic acid can be metabolized to
nicotinic acid. In addition, quinolinic acid was found to be
about 20 times more efficient than nicotinic acid as a
ricinine precursor. This phenomenon can be explained by
postulating that ”free" nicotinic acid is actually not a
ricinine precursor. Instead, as in the previously mentioned
situation with E. 221;, nicotinic acid ribonucleotide
appeared to be an intermediate, produced by the decarboxyla-
tion of quinolinic acid ribonucleotide. In this case, then,

nicotinic acid when fed, would first be converted to the

ribonucleotide before being metabolized to ricinine. Perhaps

34

no such enzyme exists which catalyzes the condensation of
nicotinic acid with 5-phoSphoribosyl-l-perphOSphate to the
ribonucleotide. Then the conversion would be non-Specifically
mediated by the quinolinic acid enzyme, and therefore proceed
at a much slower rate. Or if such an enzyme does exist, the
equilibrium might favor cleavage over condensation.

Hadwiger et al. also prepared a cell-free enzyme system
from castor bean seedlings which catalyzed the conversion of
quinolinic acid to nicotinic acid, nicotinic acid ribonucleo-
tide, and an unknown nicotinic acid derivative (46). These
compounds were the products of a four hour incubation period.
But when the reaction time was lowered to one hour, nicotinic
acid ribonucleotide was the only product observed. Again
this result points toward a ribosyl derivative of quinolinic
acid as the actual Species which is decarboxylated to
nicotinic acid ribonucleotide. This product can then be
hydrolyzed to nicotinic acid, or proceed via several bio-
synthetic reactions to various pyridine compounds.

Thus, assuming that aSpartate is the four carbon
dicarboxylic acid precursor for the 2 and 3 carbons and the
carboxyl group, and that quinolinic acid is an intermediate
in the pathway, Figure 5 illustrates a possible metabolic

sequence for the biosynthesis of nicotinic acid.

® 2'

H03 H
l/H
\clz/cmcoon HO coon I \coon I \ coon
H— cl— OH {CH-CODE coon 1K 003 /
I/ N
o/ /C\H

Figure 5.

Leete has recently pr0posed a biosynthetic scheme similar
to the one in Figure 5 (“9). He suggests that carbon 3.of
glyceraldehyde-3—phosphate is the source of carbon 6 of the
pyridine ring, and that carbon 1 of the aldehyde becomes carbon
4 in the ring. Additional information could be obtained con-
cerning the three carbon precursor for the 4, 5, and 6 carbons
of the pyridine ring of nicotine by feeding a compound such
as glyceraldehyde-3-phosphate, labeled in either carbon 1 or
carbon 3. With such an experiment, one should be able to
determine the orientation of this three carbon moiety in its
incorporation into the pyridine ring and therefore, determine
which carbon becomes carbon 6 of the ring and which becomes
carbon 4.

The actual determination of the intermediates before
quinolinic acid in the biosynthesis of the pyridine ring is
indeed a formidable task. One possible method of attack on
the problem would be to feed aSpartate-Clu to tobacco plants
for a very short metabolic period, perhaps 15 to 30 minutes.
Two dimensional paper chromatography of an extract of the

roots of these plants should separate the various carbon-l4

35

36

labeled compounds present. Using radioautography and
standard chromatographic solvent systems, one should then be
able to identify aspartate metabolites which are not related
to pyridine ring synthesis. HOpefully, one or more unique
spots would appear on the chromatograms, corresponding to

the possible intermediates shown in Figure 5. After
identifying these possible intermediates, they probably could
be synthesized and labeled with carbon-14. Feeding these
labeled compounds to tobacco plants and isolation of the
nicotine, would determine whether genuine precursors had been
employed. Of course, the final proof that these are actual
precursors, would be the.ability of a cell-free tobacco root
preparation to convert the suSpected intermediates to
quinolinic acid or some other pyridine compound. Unfortu-
nately the outlined approach is a very formidable one.and
requires much additional laboratory investigation. But it is
not an unrealistic suggestion and could well help to unravel

the pyridine ring enigma.

SUMMARY
A unique nicotinic acid degradation was devised which
permitted isolation of all of the carbons of the
pyridine ring individually.
ASpartate-3-C11+ and malate-‘3-C1)+ were fed to Nicotiana
rustica tobacco plants and found to be incorporated
almost exclusively into carbons 2 and 3 of the pyridine
ring of nicotine.
When aspartate-3-Clu was fed. 57% of the tracer resided
in carbon 3 and 38% in carbon 2. Malate-3-C1u labeled
carbon 3 with 61% of the isotope and carbon 2 with 38%.
These results are discussed and compared to data of
other researchers in the general area of nicotinic acid

biosynthesis.

37

10.

11.

12.

13.
14.

15.

l6.

170

18.

BIBLIOGRAPHY

R. F. Dawson, D. R. Christman, R. C. Anderson, M. L.
Solt, A. F. D'Adamo and U. Weiss, J. Am. Chem. Soc.,
z§.'2645 (1956).

E. Leete, Chem. Ind. London, 1252, 1270.

E. Leete and F. H. B. Leitz, Chem. Ind. London, 1252,
1572.

C. Heidelberger, E. P. Abraham, and S. Lepkovsky, J.
Biol. Chem., 122, 151 (1949).

C. W. H. Partridge, D. M. Bonner. and C. Yanofsky, J.
Biol. Chem., 124, 269 (1952). ’ ‘

J. Grimshaw and L. Marion, Nature, 181, 112 (1958).

L. M. Henderson, J. Someroski, D. R. Rao, P. H. L. WU.
T. Griffith, and R. U. Byerrum, J. Biol. Chem., 234,
93 (1959).

C. Yanofsky, J. Bacteriol., 68, 577 (1954).

M. V. Ortega and G. M. Brown, J. Am. Chem. Soc., §;,
4437 (1959).

M. V. Ortega and G. M. Brown, J. Biol. Chem., 255,
2939 (1960).

G. R. Waller and L. M. Henderson, Biochem. Biophys.
Research Communs., 5, 5 (1961).

E. Leete, Chem. Ind. London, 1258, 1477.
T. Griffith and R. U. Byerrum, Science, 122, 1485 (1959).

T. Griffith, K. P. Hellman, and R. U. Byerrum,
Biochemistry, l, 336 (1962).

D. R. Christman and R. F. Dawson, Chem. and Engr. News,
52, 40 (1961).

T. Griffith and R. U. Byerrum, Biochem. BiOphys.
Research Communs., $9, 293 (1963).

J. R. Fleeker and R. U. Byerrum, J. Biol. Chem., in
press.

M. L. Solt, R. F. Dawson, and D. R. Christman, Plant
Physiol., 35, 887 (1960).

38

19.

20.

21.
22.

23.

24.

25.

26.
27.

28.

29.

30.

31.

32.

33.

3"}.
35.

39

S. A. Brown and R. U. Byerrum, J. Am. Chem. Soc., 22.
1523 (1952).

B. L. Lamberts and R. U. Byerrum, J. Biol. Chem., 233.
939 (1953). ‘

J. R. Spies, Ind. Eng. Chem., Anal. Ed.. 2, 46 (1937).

T. A. Henry, The Plant Alkaloids, The Blakiston Co.,
Philadelphia, Pa., 1959. p. 39.

W. L. Alworth, R. C. DeSelms, and H. Rapaport, J. Am.
Chem. Soc., £6, 1611 (1964).

J. Preiss and P. Handler, J. Biol. Chem., 233, 488
(1958).

H. P. Sarett, W. A. Perlzweigg and E. D. Levy, J. Biol.
Chem., 135, 483 (1940).

J. W. Huff, J. Biol. Chem., 22;, 639 (1947).

W. I. M. Holman, and C. Wiegand, Biochem. J., 43, 423
(1948). '

W. H. Lazier and H. R. Arnold, ”Organic Synthesis,”
Coll. Vol. II, John Wiley & Sons, Inc., New York,
1943, p. 142.

M. Calvin, C. Heidelberger, J. C. Reid, B. M. Tolbert,
P. F. Yankwich, Isoto ic Carbon, John Wiley & Sons,
Inc., New York, I949, p. 260.

J. Maas, G. B. R. DeGraff, H. J. denHertog, Rec. Trav.

R. W. Green and H. K. Tong, J. Am. Chem. Soc., 28,
4893 (1955).

M. Freifelder and G. R. Stone, J. Org. Chem., 26, 3805
(1961).

H. McKennis, L. B. Turnbull, H. N. Wingfield, and
L. J. Dewey, J. Am. Chem. Soc., §Q, 1634 (1958).

L. Ruzicka, Helv. Chim. Acta, 4, 472 (1921).

J. W. Cornforth, G. D. Hunter, and G. PepJak, Biochem.
J., 2. 590 (1953).

40
36. J. W. Cornforth, G. D. Hunter, and G. PopJak, Biochem.
J., ii. 597 (1953).

37. E.(P. K§nnedy and H. A. Barker, Anal. Chem., 23, 1033
1951 .

38. F. Varrentrap, Ann. Chem., 35, 196 (1840).

39. F. Edmed, J. Chem. 3°C., 13, 362 (1898).

40. G. D. Hunter and G. POpJak, Biochem. J., 59, 163 (1951).
41. E. F. Phares, Arch. Biochem. BiOphys., 33, 173 (1951).

42. A. a. Friedman and E. Leete, J. Am. Chem. Soc., §§,
2141 (1963).

43. P. F. Juby and L. Marion, Can. J. Chem., 4;, 117 (1963).

44. D. Gross, R. R. Schfltte, G. Hubner, K. Mothes,
Tetrahedron Letters, 1263, 541.

45. A. J. Andreoli, M. Ikeda, Y. Nishizuka, 0. Hayaishi,
Biochem. Biophys. Res. Communs., $2, 92 (1963).

46. L. A. Hadwiger, s. E. Bodiei, o. a. Waller, a. K.
ghoésgn, Biochem. BiOphys. Res. Communs., i3, 466
19 3 .

47. E. Leete and A. a. Friedman, J. Am. Chem. Soc., §§.
1224 (1964).

48. D. R. Christman, R. F. Dawson, K. I. C. Karlstrom, J.
Org. Chem., 29, 2394 (1964).

49. E. Leete, Science, 242, 1000 (1965).

IIHlllUIHHIIUl||

3 1293