STUDIES ON THE BIOGENESIS OF THE PYRROLIDINE RING
OF NICOTINE IN THE TOBACCO PLANT

By

Burton L. Lamberts

A THESIS

Submitted to the School for Advanced Graduate Studies
of Michigan State University of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1958

ProQ uest Number: 10008598

All rights reserved
INFORM ATION TO ALL USERS
The quality o f this reproduction is dependent upon the quality of the copy subm itted.
In the unlikely event that the author did not send a com plete m anuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

uest
ProQuest 10008598
Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 4 8 1 0 6 -1 3 4 6

ACKNOWLEDGMENT

The author wishes to express his sincere appreciation
to Dr. Richard TJ0 Byerrum, whose interest, encouragement,
and guidance have greatly facilitated the completion of
this problem, and to other members of the Department of
Chemistry for helpful suggestions from time to time. The
author is also indebted to Dr. L„ J. Dewey of the Medical
College of Virginia for his assistance at various times and
for supplying the labelled dibromocotinine •
Finally the author wishes to thank the Department of
Chemistry, the Atomic Energy Commission, and the National
Institutes of Health for providing the funds necessary to
support this work.

STUDIES ON THE BIOGENESIS OF THE PYRROLIDINE RING
OF NICOTINE IN THE TOBACCO FTA NT

By

Burton L. Lamberts

AN ABSTRACT

Submitted to the School for Advanced Graduate Studies
of Michigan State University of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

Year 195$

Approved

5

u

a .

ABSTRACT

Glutamic a c i d ^ - C 14 was administered hydroponically to a high
nicotine strain of tobacco, Nicotiana rustica var. humilis.

The

nicotine which was isolated after a two-week feeding period was radio­
active .
A permanganate oxidation of the nicotine yielded nicotinic acid
and potassium bicarbonate as products of pyrrolidine-ring cleavage.
The potassium bicarbonate, representing positions 3* hy and 5 of the
pyrrolidine ring, was isolated as barium carbonate.

Decarboxylation

of the nicotinic acid permitted recovery of the carbon from position 2
of this ring as barium carbonate.

The pyridine resulting from the

decarboxylation was counted as the picrate.

About 10 per cent of the

carbon-ll; was found with the pyridine ring.

Half of the radioactivity

of the pyrrolidine ring was located at position 2; the remainder was
associated with positions 3 *

and f>.

From the similar pattern of labelling obtained in a comparable
study of nicotine with ornithine-2-C3-4 , it had been postulated that the
radioactivity of the pyrrolidine ring was distributed equally between
positions 2 and. 5 A procedure was developed to degrade nicotine, by which the carbon
representing position 5 of the pyrrolidine ring can be recovered as
barium carbonate.

Nicotine was treated with bromine in acetic acid

iv

and the resulting perbromide was reduced to cotinine.

Hydrolysis of

the cotinine with barium hydroxide yielded a product identified as
U- (3 1-pyridyl)-U-methylaminobutyric acid.

Since the sidechain of this

compound tends to cyclize in neutral or acidic aqueous solutions
compound was benzoylated in dilute sodium hydroxide.

the

The product was

isolated and identified as h-( 3 1-pyridyl)-l4.-N-methylbenzoylaminobutyric
acid.

Esterification of this compound with diazomethane in ether ,

followed by an a m m o n o l y s i s y i e l d e d U-(3 *-pyridyl)-U-N-methylbenzoylaminobutyramide. A Hofmann degradation of this compound permitted
recovery of the carbon from position
barium carbonate.

of the pyrrolidine ring as

The other product of this reaction c o n t a i n i n g the

remaining carbon skeleton of the nicotine molecule 3 was isolated and
identified as a thiourea derivative of 3-(3 1-pyridyl) -3-N-methylbenzoylaminO’
propylamine.
Nicotine^ labelled by glutamic acid~2-Cx4j was degraded by this
procedure.

The intermediate produced at each step of the degradation

was isolated and counted.

About half of the carbon-ll; was found at

position 5> of the pyrrolidine ring$ the amine derivative contained the
remaining radioactivity.

A similar degradation was carried out on

dibromocotinine (3^-di b r o m o - ^ - O * -pyridyl) -N~methyl-2-pyrrolidone),
from nicotine labelled by ornithine-2-C14.

The distribution of radio­

activity in the pyrrolidine ring was essentially the same as was found
in the glutamic acid-2-C14 study.
The similarity in the pattern of nicotine-labelling by glutamic
acid-2-C14 and ornithine-2-C14 supports the postulation of a symmetrical

v

intermediate along the pathway of synthesis of the pyrrolidine ring.
Several possible intermediates^ which may conceivably be formed from
the intact amino acid carbon chain ^ have been discussed.

vi

TABLE OF CONTENTS
Page
INTRODUCTION.............................................. ........
EXPERIMENTAL AND RESULTS..................................

6

Preparation of Plants..................................... ...
Uptake of Glutamic A c i d
..............
Isolation and Purification of Nicotine..............
11
Determination of Radioactivity................
Preliminary Biosynthetic Studies with Glutamic Acid-2-C14...
Administration of Radioactive Glutamic A c i d . ..«•••.•••.
Degradation of Radioactive Nicotine
.........
1} Oxidation of Nicotine with Potassium
Permanganate...............
2) Decarboxylation of Radioactive Nicotinic Acid..
Results..................................... ...........
Studies of Nicotine Degradation..........
Isolation of Cotinine........................ '.........
Isolation and Identification of )+-(3 1 -pyridyl) -5-methyl aminobutyric A c i d .................................
Some Attempted Degradations of J4— (31-pyridyl)-It-methylaminobutyric A c i d .............
Isolation and Identification of U - O 1-pyridyl )-]_{.-Nmethylbenzoylaminobutyric Acid.
..........
Some Attempted Reactions with U-( 3 1-pyridyl)-I4.-Nmethylbenzoylaminobutyric A c i d ............
hO
Isolation and Identification of !;-(3 1-pyridyl )-U-Nmethylb enz oylaminobutyramide..............
5.2
.........................
Decarboxylation
Isolation and Identification of the Amine Derivative...
Degradation of Nicotine Labelled by Glutamic Acid-2-C14
Results
.....................
Degradation of Nicotine Labelled by Ornithine-2-C14.........
Results
....................................
DISCUSSION

.................................................

s u m m a ....................................................

68

REFERENCES.................................................

69

APPENDIX..................................................... ...

vli

I

6
8
12
13
13
19
19
23
214
28
29
31
3U
36

hh
56
52
5U
55
57

60

73

LIST OF TABLES
TABLE

I

II

III

IV

V

Page

Incorporation of Radioactivity into Nicotine by
Ornitbine-2-C14 and Glutamic Acid-2-C2-4 ....................

15

Effects of Light Shields and the Quantity of Glutamic Acid
Administered on the Incorporation of Radioactivity into
Nicotine by Glutamic Acid-2-C14.....

20

Location of Radioactivity in the Nicotine Molecule After
the Administration of Glutamic Acid-2-C14^ with Isolation
of Carbon 2 of the Pyrrolidine Ring.......................

25

Isolation of Carbon 5 of the Pyrrolidine Ring of Nicotine
After the Administration of Glutamic Acid-2-C14...........

55

Isolation of Carbon 5 of the Pyrrolidine Ring of Nicotine
After the Administration of Qrnithine-2-C14. •. * ..........

57

viii

INTRODUCTION

1

INTRODUCTION

The significance of glutamic acid in the metabolic processes of
living organisms has long been recognized.

To a large extent, the

metabolic activity of this amino acid is related to protein synthesis
or to the reactions of the tricarboxylic acid cycle.

However, animal

and microorganism studies have shown that glutamic acid may also be
converted to ornithine or to proline, without cleavage of the carbon
chain.
Dakin (1), in 19135 observed the glucogenic effect of glutamic
acid, ornithine, and proline on diabetic animals and suggested a meta­
bolic relationship among these amino acids.

Stetten and Schoenheimer

(2 ) demonstrated, by isotopic labelling, the conversion of proline into
glutamic acid and ornithine in the rat.

In another study with mice,

these workers noted the conversion of ornithine, labelled with deuterium,,
into glutamic acid and proline (3)*

Sallach, Koeppe, and Rose (I4)

found relatively high concentrations of carbon-lU in proline and
arginine isolated from rats fed glutamic acid-C14.
Glutamic acid has also been shown by Tatum, Abelson, Vogel, and
others (5 ) to be a precursor of proline and ornithine in the bacterium
E. Coli.

Studies of mutant strains of *T. Utilis and N. Crassa have

offered evidence of similar interconversions among these amino acids
in fungi (6 ).

Reviews by Stetten (7) and Vogel (7) of these

2

interrelations in animals and microorganisms provide many details
that will not be covered in this study.
The information on this subject with regard to higher plants is,
however, quite meager.

Morgan and. Marion (8) obtained radioactive

glutamic acid from extracts of alfalfa seedlings fed carbon-lU-labelled
ornithine.

"When pyroxidine was added to the nutrient medium with the

ornithine, radioactive proline was isolated.

Naylor and Tolbert (9)

administered glutamic acid, uniformly labelled with carbon-ll; to
barley seedlings.

Although radioactive glutamine,

-aminobutyric acid,

aspartic acid, and various constituents of the tricarboxylic acid
cycle were detected, labelled proline and ornithine were not reportedj
hence, were evidently not found in significant amounts.
Pathways of biogenesis of the alkaloid, nicotine, would very probably
be influenced by such interconversions, were they to take place in the
tobacco plant.

Robinson (10) first proposed that the pyrrolidine ring

of nicotine arises from ornithine.

Trier (11), on the other hand,

postulated the formation of both rings of the nicotine molecule from
proline.

Prior to the general availability of carbon-ll; labelled

compounds, there was no decisive experimental confirmation of either
of these suggestions.

Early studies of alkaloid biogenesis involved

feeding supposed precursors to plants and comparing the alkaloid con­
tent of treated plants to that of controls.

Klein and Linser (12),

for example, observed an increase in the nicotine content of tobacco
after solutions of proline, ornithine, and glutamic acid had been

3

injected into the plant steins.

Fetrosini (13) noted that proline,

glycine, 3-indole acetic acid, tryptophan, alanine, aspartic acid, and
glutamic acid promoted, in decreasing order of activity, the synthesis
of nicotine in germinating seeds of tobacco cultured in the dark.
Dawson (ll;) has reported an increased nicotine content in leaves of
tobacco shoots cultured in aqueous solutions of 1-proline, nicotinic
acid, l-pyrrolidone-carboxylic acid, and d-glutamic acid.

However, in

subsequent experiments, Dawson (l£) did not observe a synthesis of
nicotine when proline and nicotinic acid were administered to initially
nicotine-free tobacco leaves.
Similar studies were carried out with other alkaloids containing
pyrrolidine rings.

Klein and Linser (16) noted an apparent increase

in stachydrine, an alfalfa alkaloid, when solutions of proline,
ornithine dihydrochloride, or sodium glutamate were injected into the
stems of Stachys palustris L ., Stachys recta L ., and Galeopsis
ochroleuca Lan.

James (17) showed significant increases in the

hyoscyamine content of Atropa belladonna, following the feeding of
ornithine or arginine to experimental plants.
However, variations of metabolism among the plants under study,
due to individual differences and, possibly, to seasonal effects, could
defeat attempts to reproduce experimental data by this method.

Hence,

it would be quite difficult to distinguish the real from the simply
apparent precursor.
With the advent of carbon-lit-labelled compounds, however, studies
of alkaloid biogenesis have been placed on a firmer basis.

Leete,

b

Spenser, and Marion (18), employing Datura stramonium, observed the
incorporation of ornithine-2-C14 into the pyrrolidine ring of hyoscyamine.
The activity appeared either to be distributed between carbons 1 and
of hyoscyamine (i.e., carbons 2 and 5 of the pyrrolidine ring) or to be
located at one of these positions.

Morgan and Marion (8) found no

incorporation of ornithine-2-C14 into the pyrrolidine skeleton of
stachydrine, but this was attributed to a lack of synthesis of the
alkaloid under the particular conditions employed in their experiments.
Dewey and Byerrum (19) administered ornithine-2-014 to threemonth-old tobacco plants (N. rastica), and obtained radioactive nicotine.
Essentially all of the radioactivity was found to be present in the
pyrrolidine ring— half of this radioactivity was located at carbon 2,
with the remainder associated with carbons 3* ^

and 5*

An independent

study by Leete (20), feeding ornithine, similarly labelled, to sixmonth old plants (Nicotiana tabacum L.) confirmed these observations.
Considering the ornithine-proline-glutamate relationship in the
metabolic functions of mammals and microorganisms, the early studies
of alkaloid precursors, and the labelled ornithine experiments, it was
decided to administer glutamic acid-2-C14 to tobacco plants.

The in­

vestigation was conducted with the following objectives in mind:
1) to test the capacity of glutamic acid to serve as a precursor of
the pyrrolidine ring of nicotine and, 2) in the event glutamic acid
functioned as a precursor, to establish the sites of radioactivity in
the pyrrolidine ring through degradative studiesj 3) to compare the
labelling effects of ornithine-2-C14 and glutamic acid-2-C14 by

5

carrying out similar degradative studies of nicotine from the labelled
ornithine-feeding experiments.

A similar pattern of labelling by

each of these amino acid precursors would suggest the formation of one
or more common intermediates in the pathway of nicotine biogenesis.

EXPERIMENTAL AND RESULTS

6

EXPERIMENTAL AND RESULTS

Preparation of Plants

The tobacco plants used in this study were of a high nicotine
strain, Nicotiana rustica, var. humilis .

The seeds were planted in a
i
greenhouse, in flats containing Vermiculite.
After two or three weeks

the plants were transplanted, being spaced at intervals of two to three
inches.

Every day the plants were watered and twice a week they

received a nutrient solution of the following composition:

1 mg.

MgS04 ;7H20, 1 g. K 2HP04 , 5-8 g. Ca(N03 )2 «I|N20 in four liters of tap
water.

After two to three months, depending somewhat on seasonal

conditions, the plants had attained the desired height of five to seven
inches.

The plants were then removed from the flats, rinsed briefly

with tap water to loosen the Vermiculite clinging to the roots, and
the roots were trimmed to within about two to four mm. from the plant
stems.

This technique, described by Dewey (21), was designed to force

the plants to develop new root systems.

Dawson (22) has shown that

nicotine is synthesized in the roots of tobacco plants; hence, improving
the health and vigor of root systems should promote the synthesis of
nicotine and, accordingly, the incorporation of carbon-lU into the
nicotine molecule.

Following a rinse with distilled water, each plant

was placed in a 125-ml. Erlenmeyer flask in a nutrient medium.

This

solution was prepared by diluting, with two parts of water, one part
XA commercially available heat expanded mica-

7

of a stock solution of the following composition:

Ca(N03 ) *l4H 20 , 2.61

g.; MgS04 .7H20, 756 mg; KC1, 500 mg.; K H 2P04 , 500 mg.; (NH4 )2S04 ,
500 m g .5 FeCl3 .6H20, 5«6 mg.; water, added to a volume of two liters.
A thin cotton plug was inserted into the mouth of each flask to retard
evaporation of the solution and to lend support to the plant.
After a two-week period., u nder,conditions of natural lighting, the
plants had usually developed extensive root systems.

In early stages

of this development, plants would occasionally wilt and die, particularly
during experiments conducted in the hot summer months.

Normally, growth

of new roots was not attended by appreciable growth of other parts of
the plants.
To prepare the plants for hydroponic administration of the glutamic
acid-2-Cx4, the plants were transferred to clean 125-ml. Erlenmeyer
flasks.

To each flask was added 50 ml. of a nutrient medium which was

prepared as follows:

one part of the stock solution was diluted with

two parts of a fungicide solution (to be described later), and the
resulting solution was aerated for 10 or 15 minutes.

Five-tenths ml.

of an aureomycin solution (50 mg. per 50 ml. of solution) was also
added to the flasks.
In some of the initial experiments, a comparison was being made
with one of Dewey1s ornithine-feeding experiments (21), hence the roots
of the plants were not shielded from light.

Later, however, the flasks

were covered with shields of black construction paper to promote root
growth; this practice was followed from the beginning of the rootgrowing period until the plants were eventually harvested.

8

The plants were kept in a hood daring the period of administration
of the glutamic acid-2-C14. A nutrient medium (stock solution diluted
with two parts of water) was supplied as required by the plants.
Artificial lighting was provided twelve hours daily during this period
by two 36-inch, 30-watt fluorescent tubes and by a 100-watt incandescent
bulb, placed approximately fourteen inches above the plants.

The light

intensity thus supplied the plants was about 200 foot-candles.

Uptake of Glutamic Acid

Since Bewey (21) has shown that ornithine may be administered to
tobacco plants hydroponically, it was decided to follow the same
procedure with glutamic acid.

Before feeding the glutamic acid-2-C14,

however, an uptake study was undertaken with non-radioactive glutamic
acid.
In the first uptake experiment, 52.5 nil. of solution was added to
each of twelve 125-ml. Erlenmeyer flasks.

Each flask contained 50 ml.

of nutrient solution (16.6 ml. stock solution diluted with two parts of
water), 0.5 mg. of aureomycin, and four mg. of .glutamic acid, adjusted
to a final volume of 52.5 ml. with water.
Four tobacco plants were removed from a flat and the roots soaked
briefly with tap water.

Following careful removal of the Vermiculite,

the roots were rinsed with distilled water and the plants transferred
to four of the flasks.

Six root fragments were introduced into each of

four other flasks to determine the extent of glutamic acid loss to
plant root microorganisms.

The remaining flasks served as controls.

9

The flasks were placed in the hood and artificial light was
supplied for a twelve-hour period each day.

Oxygen was administered

at the beginning and end of each lighting-period.

After forty-eight

hours , the plants were removed from the flasks , the roots were rinsed
with the nutrient solution and the washings were added to the flask
contents.

Due to loss of solution by plant uptake and evaporation^

most of the solution volumes had dropped below 52*5 ml.

The flask

volumes were restored with nutrient solution and the contents of each
flask filtered separately through 'Whatman 2 paper.

A ten-ml. aliquot

was then removed from each flask and diluted to 100-ml., in preparation
for the glutamic acid analysis.
Glutamic acid was determined by the method of Gordon (23).
A standard curve of glutamic acid v s . optical density was prepared from
known quantities of glutamic acid5 and covered the range of concentra­
tions encountered in this study.

Zero to 10 ml. aliquots of a 10 y/ml.

solution of glutamic acid were added to seven 100-ml. volumetric flasks.
The nutrient solution of inorganic salts and aureomycin,, of the concen­
tration described above but minus the glutamic acid., was diluted 1:20
with water and 10-ml. aliquots of the diluted solution were added to
each volumetric flask.

Then 2.3 ml. of a freshly prepared solution of

sodium hypochlorite (prepared by bubbling chlorine through 0.5 M sodium
hydroxide and diluting 1:100 with 0.1 M sodium hydroxide) was added to
each flask to oxidize the glutamic acid.

Exactly ten minutes after

each sodium hypochlorite addition, 0.8 ml. portions of an oxidizable

10

2
green dye

solution were added to the flasks to consume the excess

sodium hypochlorite.
oxidation.

The dye faded to a yellow-orange color upon

Exactly ten minutes after each addition of the dye solution,
3

each flask received a ten-ml. aliquot of a sulfate buffer

to stabilize

the colors, and the flask volumes were adjusted to 100 ml. with water.
The optical density of each flask was determined with a Beckman Model B
Spectrophotometer, at 630 mu, at 10-minute intervals after each buffer
addition.
To analyze the solutions from the uptake experiments, a 10-ml.
aliquot of each diluted sample (as described above) was added to a
100-ml. volumetric flask.

The subsequent steps were those followed with

the known quantities of glutamic acid; the concentration of each sample
was determined from the standard curve of concentration vs. optical
density.
The experimental results indicated that, though essentially all of
the glutamic acid had been absorbed by the plants in the li.8-hour period,
there was also an average loss of 38 peu cent of the glutamic acid in
the flasks containing root fragments.

The cloudy appearance of these

solutions suggested that the loss was due to fungi; hence, a solution
4

of fungicide was prepared.

The fungicide

was only slightly soluble in

water— less than one part fungicide per 100,000 parts of water.

2Fast Green FCF (The Matheson Co., E. Rutherford, N. J.).
3150 ml. of 10 M H 2S04 and 2ii5 nfL- of NaOH, diluted to one liter.
4A commercial brand, "Ubavit NeuM (Schering-Kahlbaum A . G. Berlin).

11

In the second uptake experiment, twelve 125-ml. Erlenmeyer flasks
were set up as before except that the nutrient solution was composed
of one volume of stock solution, diluted with two volumes of fungicide
solution.

In other respects, the experiment was conducted in the same

manner as the first experiment.

A new standard curve was prepared to

compensate for possible discrepancies arising from the use of the
fungicide.
The results of this experiment showed that 95 per cent of the
glutamic acid had been absorbed by the plants, with an average loss of
h per cent of the glutamic acid in the root-fragment flasks.

The solu­

tions of these flasks were clear and appeared to be free of fungal
growth.

Hence, with the subsequent administration of glutamic acid-2-C14,

in the later experiments, both fungicide and aureomycin were added to
the nutrient solutions.

Isolation and Purification of Nicotine

At the end of each administration period, the plants were removed
from the flasks, the roots were rinsed with distilled water and blotted
on cheesecloth.

The plants were cut into small pieces and dried under

infrared heat lamps.

Temperature was regulated to facilitate rapid

drying without heating the plant material above 70 to 80° C.

The dried

material was thoroughly ground in a mortar, mixed with 20 per cent by
weight of calcium hydroxide, and steam distilled in a Kjeldahl flask.
Distillation was continued until the distillate no longer gave a white
precipitate with silicotungstic acid, a reagent used to detect the

12

presence of alkaloids*

The distillate was collected in a small quantity

(usually about £ ml.) of 6N HC1, and concentrated to 30-lj.0 ml. with a
flash evaporator.

Nicotine was purified by two successive azeotropic

distillations with water, using a Widmer column as described by
Smith (2U).

The distillation was effected from an alkaline medium and

the distillate collected in 6N HC1.

The acidic solution was concen­

trated to dryness under reduced pressure, and the resulting hydrochloride
dissolved in a small amount of water and methanol.

Nicotine dipicrate

was precipitated upon addition of an excess of methanol saturated with
picric acid.

After a short period of cooling, the dipicrate was

separated by filtration with a sintered glass filter, washed with
methanol, and recrystallized from hot water.

(M.p., 219-2215 Recorded

value 218°).

Determination of Radioactivity

All measurements of radioactivity were made with a Nuclear-Chicago
Model 192X scaler (Nuclear Instruments and Chemical Corporation,
Chicago 10, Illinois) and a Tracerlab M°del SC-16 proportional flow
counter (Tracerlab, Inc., Boston 10, Massachusetts).

The counting gas,

purchased from Tracerlab, Inc., consisted of 90 per cent argon with
10 per cent methane.

The counting system was I4O per cent efficient, as

determined by a National Bureau of Standards N a 2C 140 3 sample.

5A11 melting-point values recorded, in this thesis were obtained by
means of capillary melting-point tubes. The melting points are
uncorrected.

13

The materials to be counted were plated on aluminum planchets^
2,83 sq, cm. in area.

Large crystals were pulverized in a small mortar

before transfer to planchets.

The surface of each plate was made as

smooth and uniform as possible before counting.

Barium carbonate

samples were plated from a slurry in 95 per cent ethanol,, as described
by Galvin (25).

Activities were corrected to 11infinite thinness1* by

reference to appropriate self-absorption curves.

(See Appendix for

calculations.)

Preliminary Biosynthetic Studies with Glutamic A c i d - 2 14

Administration of Radioactive Glutamic Acid
In the first experiment with glutamic acid-2-C 14, the radioactive
compound was diluted with non-radioactive glutamic acid.

As in the

uptake study,, each plant received four mg. of glutamic acid.

However 9

since Dewey (21) reported an increased synthesis of carbon-llj. labelled
nicotine when ornithine-2-C3-4 was supplied at two different times dur­
ing the administration period^ it seemed desirable to follow this
procedure with the glutamic acid.
Forty plants were prepared^ as described previously.

Each plant

then received two mg. of the glutamic acid5 having a calculated radio­
activity of 2.5 x 105 cpm.

Five days later another two mg. of the

glutamic acid was added to each flask.

The plants were harvested nine

days later., and the nicotine was isolated as the dipicrate.

However,

the nicotine dipicrate contained no significant radioactivity.

Since

this result could mean that the reservoir of radioactive glutamic acid,

iu

presumably present in each plant and available for nicotine synthesis,
had been inordinately diluted by addition of non-radioactive glutamic
acid, it was decided to omit the latter in the following experiment.
In the second experiment, each of forty plants received 0*35 m gof glutamic acid-2-C14, having a calculated radioactivity of 8 x 10s cpm.
Five days later, another 0.35 mg. was given to each plant ; again, after
nine days, the plants were harvested.

The nicotine dipicrate from this

harvest showed a specific activity of U«36 x 103cpm/mM.

Since this

experiment was modelled after an experiment of Dewey (21), which involved
ornithine-2-C14 of equivalent radioactivity and molar quantity to the
glutamic acid-2-C14, it was hoped that a valid comparison could be made
between glutamic acid and ornithine as precursors of nicotine.

When

this experiment was repeated with twenty plants the specific activity
of the isolated nicotine dipicrate was 2.7U x 103cpm/mM.

The discrepancy

between these results is believed to have been due to seasonal conditions.
Later in the course of this study it was discovered that the count­
ing instrument employed in the present studies was twice as efficient
as the instrument Dewey had used in his experimental work.

Hence, the

necessary correction was made when the ornithine-2-C14 and the glutamic
acid-2-C14 experiments were compared.

This comparison is shown in

Table I.
From the foregoing experiments, it is evident that glutamic acid
is incorporated into nicotine, but, when ornithine and glutamate were
fed in equivalent quantities and radioactivities, the nicotine from
the o m i thine-fed plants was considerably more radioactive.

However,

15

TABLE I
INCORPORATION OF RADIOACTIVITY INTO NICOTINE BY
ORNITHINE-2-C14 AND GLUTAMIC ACID-2-C14

Amino Acid
Administered

Experiment

Ornithine-2-014

Ratio of M a x .
Spec. Activity
of Ornithine:
Glutamic acid

Max. Spec.
Activity of
Nicotine d.p.
(reported)

Max. Spec.
Activity of
Nicotine d.p.
(corrected )-X-

--

1.21 x 105

6.05 x 105

--

Glutamic Acid-2-C14

2

h .36 x 103

1.09 x 104

56:1

Glutamic Acid-2-014

3

2.7U x 103

6.85 x 103

88:1
72:1 (average)

Assuming each Instrument to be 100 per cent efficient.

before degradative studies of nicotine could be undertaken to establish
the focal points of radioactivity in the molecule, it was necessary to
increase the supply of the carbon-lLi. labelled nicotine.
In a. fourth experiment, each of fifteen plants received a total of
5.16 mg. of the glutamic acid-2-C14 (calculated radioactivity, 1.17 x 107
cpm) over the two-week experimental period.

The nicotine dipicrate

(360 mg.) had a maximum specific activity of I.J4.0 x 104 cpm/mM.
Sufficient nicotine was thus available for a preliminary degradation,
following a dilution of the radioactive with non-radioactive nicotine.
The preliminary degradation of this radioactive nicotine did demon­
strate a similar pattern of labelling of the pyrrolidine ring by glutamic
acid and ornithine, as will be described later in this study.

Hence,

a further series of degradations of the nicotine seemed desirable,
although this would require about seven to eight times the amount of

16

nicotine used in the preliminary study.

However, the foregoing method

of acquiring the radioactive nicotine for the degradative studies
seemed impractical because of the expense of the radioactive glutamic
acid and its apparently low degree of incorporation into nicotine.
Therefore, the possibility of increasing still further the synthesis of
carbon-ll|.-labelled nicotine by the plants was studied.
Shields of black construction paper were fitted to the Erlenmeyer
flasks to protect the roots from light.

Following this modification,

the radioactivity of the nicotine was increased several fold over that
of the previous experiments; hence, shields were employed in all of the
subsequent plant-feeding experiments.

The author believes this increased,

incorporation of carbon-ll; into nicotine to be due partly to the over­
all improvement in health of plants upon shielding the roots from light;
particularly, however, it is believed due to the development of
healthier and more extensive root systems than were observed in the
previous experiments.
As mentioned before, however, the radioactivity of the nicotine
isolated from plants fed glutamic acid-2-C14:, in this manner may vary
considerably with any given experiment.

This is believed to be a

seasonal effect; at any rate, in conjunction with the "safety in numbers"
maxim, it seemed practical to treat a large group of plants with a
comparatively small amount of the radioactive compound per plant to
obtain the necessary labelled nicotine for the degradations.
Hence, as the growing plants reached the desired size, they were
utilized for the synthesis of radioactive nicotine.

Approximately one

17

hundred forty plants were harvested, in six groups, besides the plants
of the previously described experiments.

The method of handling these

plants was essentially the same as outlined in the previous experiments
except that, with each group, a total of one mg. of the glutamic acid
was administered per plant.
Although the direct purpose of these plant-growing operations was
the accumulation of a supply of radioactive nicotine, nevertheless,
certain comparative experiments were thereby rendered feasible.

For

example, it was noted that the radioactivity of the nicotine, isolated
in experiments 3 and 1;, was roughly proportionate to the amount of
glutamic acid-2-C14 administered to the plants.

These experiments,

however, were conducted during the summer and autumn months respectively;
hence, an attempt was made to clarify this effect by feeding two groups
of plants simultaneously, to eliminate the influence of seasonal change.
Six plants were fed a total of one mg. of glutamic acid-2-C14 per plant.
Another group of six plants, similar to the first group with respect
to the size and age of the plants, received a total of three mg. of the
glutamic acid-2-C14 per plant.

After the usual two-week administration

period, the plants of both groups were harvested, and the nicotine
isolated as the dipicrate.

In this comparison, however, the difference

in the amounts of nicotine isolated from each group of plants should
be considered.

The second group seemed to contain slightly more nicotine

than was found in the first group.

Assuming that each group synthesizes

nicotine at approximately the same rate after the glutamic acid-2-C14
administration, and that the isolated nicotine dipicrate is a measure

18

of the total nicotine, the excess nicotine contained in the total
reservoir of the second group of plants may be considered to function
as a diluent.

Upon application of this dilution factor to equate the

total nicotine content of each groups the radioactivity of the nicotine
was found to be approximately proportionate to the amounts of radio­
active glutamic acid administered— at least up to three mg. of glutamic
acid per plant.

This result would indicate that the glutamic acid

pool in the plant is much larger than the quantity of glutamic acid
administered.
As mentioned previously, Dewey (21) reported an increased synthesis
of nicotine in tobacco plants that had been forced to grow new roots,
following removal of the original roots.

Another comparative experiment

was set up to assess this effect under the conditions employed through­
out the present study.

The roots of the first group of ten plants

(Experiment 10, Table II) were removed, thus forcing the plants to grow
new roots.

The second group (Experiment 11, Table II) contained the

same number of plants, of approximately the same average height ;
however, these plants were simply removed from the flats with the
original roots left intact.

Each plant received a total of one mg. of

glutamic acid-2-C14 (calculated activity, 2,81* x 106cpm), administered,
as before, in two portions.
two-week period.

The plants were harvested after the usual

The nicotine from the first group showed about twice

the radioactivity found in the nicotine from the second group of plants
when a conversion factor was applied in an attempt to equate the
nicotine content of each group, as described in the previous experiment.

19

Even "though relatively few plants were involved in this experiment, the
results support Dewey*s observations.
Comparative data of the various harvests is provided in Table II.

Degradation of Radioactive Nicotine
1)

Oxidation of Dicotine with Potassium Permanganate.

Dewey and

Byerrum (19) have shown that ornithine-2-C14 is incorporated into the
pyrrolidine ring of nicotine.

Half of the original radioactivity was

located at carbon 2 , with the remaining activity associated with carbons
3, U, and 5> of the pyrrolidine ring.

Since glutamic acid and ornithine

may follow a similar route of incorporation into nicotine, it was
decided to repeat this degradation with some of the radioactive nicotine
isolated in the previous studies after the glutamic acid administrations.
This method of degradation, originally described by Laiblin (27),
consists in the oxidation of nicotine, by neutral potassium permanganate,
to nicotinic acid and potassium bicarbonate.

The potassium bicarbonate,

resulting from oxidation of carbons 3 , U, and 5 of the pyrrolidine ring,
may be decomposed with acid and the liberated carbon dioxide collected
as barium carbonate by bubbling it through barium hydroxide.
Initially, twd grams of radioactive nicotine dipicrate, prepared
by diluting the radioactive with non-radioactive material, were dis­
solved in a minimal quantity of hot water, recrystallized, and the
resulting homogeneous sample of nicotine dipicrate was counted.

The

nicotine (0.39 g.) from 1.5 g. of the recrystallized sample was
recovered by azeotropic distillation with water, as previously described,

20

5H
*
H
• O 73
o
0 >7, 0
Php
q
CO •H •H
p
>
• •H o
P o
CD O •H
ISJ
0
S3
•H
-|1
O
• o
tiD •H
s
CO

73
I
—I
CD

0
P
ctJ
Ph
O
■H
£3
•H

^

•71

N
t
<

O O
O
i
—1 i
—I 3 i—I
X
3
NO JS £
o
OA C-— -=$ OJ X
-Tf CM iH CM CM

^

H1

^

0

0

o
i
—1
3
X
^ On
OO \A
CM rH

o
3 3
3
© <S ox
nO O n OO
CA On CM

OO 0— CXI o NO 1A 7a o
C-— NO T-A vO NO oo ca ca
oAONPf-OAXACArArH

CNJCO rA
c— CM o
i
—ICM _0-

Q

Q

p

o

s

73

M

P
rw
SH

M
^
PS M
rS *
rv
S

X

X

X

X

X

X

X

£

0 P
O
CM

o MD

T-A

c-—

CM

CM

■A

ro

"L A

O

CM ON O
OO

[>-

--7J

& « H

O

•

•

P

0—

CM

i
—|

P

CN

CM

•

•

*

p
i—
s>

73
CD

P

0
O

t—I
X
VT\

3

^
NO

•

3

^
NO

i—I

t3 ~s3

q O a}
f P r—1
CJ
Ph
73
•
*H
pH
M O ©

rH

S

Q

CD

P Pa rH
O
Ph

On
i— 1

i—1

C3 fH O

P=-t P
£ ^0 ^0

Pa

C—
CM 0
1-0
73 P
q

CO

&

0
P

rH ctJ
SH r—i

CD

s
S ^
S K. ^

Ph
0

73

0 S3

CO

"S

Ph

0

M

I—I

Ph

0
Ph

[-1 n-J
CD O
> H
•H
P 0
O
0 bjO

P
0

p>

W

73
O
•H

0
rP rX
Fh 0
0
• ts1

Ph

S

w

Ph

0

0

^-D

1
o
•H
73
0

Ph q
CO •H

73
•H

3

0

0
O

o
o3-

O•

O

CD

Q

Q

0
O

0

0

Q

Q

0
Q

x

rd

rd

rd

rd

H

CO

CO

CO

OO

T-A

.-S
CO

CO

H

CM

CM

CM

CO

CM

CM

rH

* O» O*

o
•

rd

0
O
rH

x

nO

C—

CD

P

H
0

T O
1P

*% 73
NO 0
H Ph
0
<H P
O 0
•H
w

q

-p ■H
0 S
0 73
£ 0
0

•o •o • o

£

g

P\ i—I i—t i
—I i
—I co i
—|

Ph
q 73
O •H
O

Ph

0
O

O
*r~j

q

O
_0"

c—
CO

O
CM

©A
i—1

-H
-Cj"

CO
|—I

vO
IN—

nO

nO

O
[—I

O
(—]

O 0
•H P

P

a

3

hO

rQ 0
£ S>
O -H
O P
O
0
H

CM

C A ^ f l A M D

A-

CO

On

O

rH

21

except that no hydrochloric acid was added to the receiving vessel.
The UOO ml. of distillate was transferred to a liter Erlenmeyer flask
and was treated with 80 ml. of an aqueous solution of potassium perman­
ganate (2.3U g*., i.e., calculated as six times the weight of nicotine).
The potassium permanganate solution was added in five-ml. quantities
at three-to-five-minute intervals ; with each addition, the flask contents
were thoroughly mixed.

Initially, the oxidation was quite rapid, as

evidenced by the disappearance of the purple permanganate color and the
appearance of brown manganese dioxide.

However, the rate of oxidation

decreased with each addition of the reagent; after the final addition of
the permanganate solution, the mixture was placed on a steam bath and
was heated for about ten hours -

to complete the oxidation and to promote

coagulation of the manganese dioxide.

At the end of this time, the flask

was permitted to cool, the manganese dioxide was removed by filtration
and was washed thoroughly with hot water.

The washings and filtrate

were combined to give a nearly colorless solution which was concentrated
to dryness with a flash evaporator.
The residue, consisting of the potassium salt of nicotinic acid
and of potassium bicarbonate, was dissolved in about 50 ml. of water.
Acidification of this solution with dilute nitric acid released carbon
dioxide which was swept into a saturated solution of barium hydroxide
by a slow stream of nitrogen.

The resulting barium carbonate was isolated,

washed with hot water, and dried for one hour at 110° C.

The barium

carbonate was plated, dried in a vacuum desiccator, and counted.

22

Self-absorption corrections were made from a self-absorption curve,
prepared for barium carbonate (36).

After the solution of nicotinic

acid was treated with dilute ammonium hydroxide until just alkaline to
litmus, it was concentrated to dryness.

Water was added to dissolve

the residue, and the solution was again concentrated to dryness.

The

residue was dissolved in 10-15 ml. of water and 0.1 M silver nitrate was
added in excess to form the insoluble silver salt of nicotinic acid.
The salt was separated on a sintered glass filter and washed with
distilled water.

After slurrying the solid in about 50 ml. of hot water,

hydrogen sulfide was introduced to decompose the silver salt.

The silver

sulfide was removed by filtration, the filtrate decolorized with a small
amount of charcoal, and the solution concentrated to dryness under
reduced pressure.

The solid residue was transferred to a sublimation

apparatus which was placed in a mineral oil bath and evacuated with a
vacuum pump.

The bath was slowly heated to l50° and was maintained at

this temperature until no more sublimate appeared to form; this procedure
was repeated four times to purify the nicotinic acid.

With each sub­

limation, the amount of brownish residue remaining in the tube was
diminished; with the final sublimation, only a trace of impurity was
observed.

About 120 mg. of the nicotinic acid was collected,

230°; recorded value, 232°).

A comparison of the infrared

(m.p.,

spectrum of

the experimental preparation with that of commercial nicotinic acid
confirmed the identification and purity of the product.

The nicotinic

acid was plated, dried in a vacuum desiccator, and counted.

23

2)

Decarboxylation of Radioactive Nicotinic A c i d .

The nicotinic

acid obtained from the oxidation of nicotine was divided into two
portions, which were decarboxylated in separate operations.

Following

the procedure described by Dewey (21), the nicotinic acid was mixed with
an excess of calcium oxide and transferred to a small flask, which was
part of the apparatus employed in the demethylation of nicotine (26).
A delivery tube was connected to the sidearm of the flask, and a stream
of nitrogen was passed through the flask and delivery tube into a small
centrifuge tube containing a solution of methanol saturated with picric
acid.

The flask, immersed in a copper oxide bath, was slowly heated

with a Bunsen burner.

As droplets of pyridine began to condense on the

walls of the delivery tube, the temperature was raised to 300-350°., and
maintained at this level for 15 to 20 minutes.

The apparatus was

permitted to cool and the delivery tube was rinsed several times, with
methanol saturated with picric acid., into the centrifuge tube.

After

the solution had been cooled in the refrigerator for about an hour, the
precipitate was separated with a sintered glass filter and washed with
methanol.

The pyridine pic rate samples from the two decarboxylations

were combined and recrystallized from hot water.

Since further purifi­

cation of this sample was believed necessary, 10 mg. of recrystallized
inactive pyridine picrate was mixed with 10 mg. of the above recrystallized
sample and the mixture was recrystallized from hot water.

The pyridine

picrate was plated from a slurry in 95 V eT cent ethanol, dried in a
6

vacuum desiccator, and counted.

A n a l . Per cent calculated for

6The microanalyses reported in this thesis were performed by the
Spang Microanalytical Laboratory, Ann Arbor, Michigan.

2h

C2.2.HeN 40 7 :

C, b2,Q6j H, 2.62; N, 18.18; per cent found:

H 5 2.75; N, l8 .ll;.

C , U2.91i

M.p., I63“l6l4°5 m.p. recorded, l65°*

Upon decarboxylating the nicotinic acid in an excess of calcium
oxide, calcium carbonate was formed.

Before recovering the carbon

dioxide, a delivery tube was directed from the flask into a saturated
solution of barium hydroxide.
of the flask by

A glass Y-tube, attached to the side arm

tygon tubing, was also connected to a dropping funnel

and to a nitrogen tank.

Dilute nitric acid, introduced into the Y-tube

through the dropping funnel, was forced into the flask by a stream of
nitrogen.

The liberated carbon dioxide was swept into the barium

hydroxide by the nitrogen.

The barium carbonate was isolated and plated,

as described previously.

Results
The data obtained upon counting the samples of nicotine and its
degradation products, described above, are presented in Table III.

The

results of this degradation reveal a labelling-pattern for the nicotine
which is very similar to that reported by Dewey and Byerrum (19) for
nicotine from tobacco fed ornithine-2-C:L4.
The oxidation of nicotine by the neutral potassium permanganate
yields nicotinic acid, potassium bicarbonate, and methylamine.

Through

cleavage of the pyrrolidine ring, carbon 2 is converted to the carboxyl
carbon of nicotinic acid; carbons 3 , b > and 5 are oxidized to potassium
bicarbonate and the N-methyl grouping is converted to methylamine.
It is not definitely known whether or not any of the methyl carbon is

25

TABLE III
LOCATION OF RADIOACTIVITY IN THE NICOTINE MOLECULE AFTER THE
ADMINISTRATION OF GLUTAMIC ACID-2-C14, WITH ISOLATION OF
CARBON 2 OF THE PYRROLIDINE RING

Maximum Specific Activity
(cpm/mMole x 10’"3 )

Compound
(1) Nicotine dipicrate

2.7

(2) Barium carbonate^' (from nicotine oxi­
dation)

1.1

(3) Nicotinic acid

1.3

(U) Barium carbonate (from decarboxylation)

1.2

(5) Pyridine picrate

0.2

_X.

Calculations based on the assumption that three moles of barium
carbonate are formed per mole of nicotine.

converted to bicarbonate.

Laiblin (27), however, has isolated methyl­

amine from the oxidation mixture, and reports a strong odor of methyl­
amine throughout the oxidation procedure.

In the present study, as in

the previous studies of Dewey (21), it has been assumed that the bi­
carbonate is essentially the product of carbons 3 , U, and 5Hence, after the bicarbonate was converted to barium carbonate,
(Line 2, Table III) the original count of the barium carbonate was
multiplied by three in computing its specific activity, to permit a
comparison of the barium carbonate and nicotine.

The specific activity

of the barium carbonate is seen to be about half that of the nicotine,
and in the same order of activity as was found for the nicotinic acid

26

(Line 3) •

This closely resembles the pattern of incorporation of

ornithine-2-C14 into the pyrrolidine ring, reported by Dewey (21).
"When the nicotinic acid was decarboxylated, the carboxyl carbon
(originally carbon 2 of the pyrrolidine ring) , was isolated as barium
carbonate and counted.

The specific activity of this barium carbonate

(Line Lj., Table 111) was about 90 per cent of that of the nicotinic acid.
The remaining activity of the nicotinic acid was associated with the
pyridine ring, which was counted as pyridine picrate (Line

Table III).

In previous studies with ornithine-2-C14 by Dewey and Leete, Dewey
(21) found no activity in the pyridihe (isolated as the picrate),
following decarboxylation of nicotinic acid.
nicotinic acid

by refluxing it

Leete (20) decarboxylated

in quinoline with a copper chromite

catalyst.

The activity of the

carboxyl carbon was essentially thesame

as that of

the nicotinic acid;

hence, no activity was assumed to be

present in

the pyridine, which, however, was not isolated.

The biogenesis of the pyridine ring of nicotine has never been fully
elucidated.

Recently certain evidence (28) has indicated that this ring

may arise from “active acetate” or, perhaps* from some intermediary
product of acetate metabolism.

Studies of nicotine biogenesis, involv­

ing ribose-l-C14 (29) and pyruvate-3-C14 (30), suggest that the pyridine
ring precursor is related to the intermediates of glycolysis or of the
tricarboxylic cycle.

Since glutamic acid is a very active metabolite in

living systems, it may quite conceivably be linked with this precursor
of the pyridine ring.

27

In the previous studies of nicotine biogenesis by Dewey and Leete,
working with ornithine-2-C14, the nicotine was demethylated to isolate
the N-methyl grouping of the pyrrolidine ring.

Dewey (21) found less

than two per cent of the total radioactivity of the nicotine with this
carbon atom* Leete (20) reported no significant activity at this position.
Hence, no attempt was made in the present study to demethylate nicotine.
From these studies of nicotine biogenesis, it has been postulated
(20,21) that the pyrrolidine ring is formed from some type of symmetrical
intermediate.

The studies with omithine-2-C14, mentioned above, have

shown that half of the radioactivity of the nicotine molecule is located
at carbon 2 of the pyrrolidine ring, with essentially all of the remain­
ing radioactivity associated with carbons

I4,, and 5> of this ring.

If the pyrrolidine ring is formed from some type of symmetrical inter­
mediate, without a previous cleavage of the carbon chain of the glutamic
acid or ornithine precursor, the remaining radioactivity would probably
be located at carbon 5 of this ring, rather than with carbons 3 or I4.
Hence, a further series of degradations were undertaken to investi­
gate this possibility.

The primary purpose of the intended study was to

isolate carbon 5 of the pyrrolidine ring, using more of the radioactive
nicotine, labelled by glutamic acid-2-C14.

In the event that carbon 5

would be found radioactive, it was hoped that a comparison of its radio­
activity might be made with the residual carbon skeleton of the nicotine
molecule.

28

Studies of Nicotine Degradation

Huber, Etard, Pinner, and others, discovered several methods for
degrading the pyrrolidine ring of nicotine, when the structure of this
compound was under study.

The oxidation with permanganate, described

by Laiblin (27) has already been discussed.

However, the nature of the

^s-^-xo^ residue, attached to the pyridine ring of nicotine, was actually
established by a study of the action of bromine on nicotine.

This study,

pursued at times by several investigators (31*32 ) was more thoroughly
investigated by Pinner (33)*

Pinner succeeded in isolating several

brominated products, which were in turn subjected to further degradative
treatments.

Although many of these products were obtained only in small

quantities, this was sufficient for Pinner1s purpose of elucidating the
structure of nicotine.
When nicotine is treated, with bromine in acetic acid, an oily
perbromide is formed.

Reduction of the perbromide with zinc and hydro­

chloric acid forms compound I, which Pinner named “cotinine

Compound I

N

Cotinine is a stable compound which, according to Pinner (33)* is not
affected by heating with barium hydroxide or by the action of hydrochloric acid at l80°C
Prior to the present study, an attempt was made at this laboratory
(3U) to isolate carbon 5 of the pyrrolidine ring as the carboxyl carbon

29

of benzoic acid.

Cotinine was treated with phenyl magnesium bromide

and subsequently oxidized with potassium permanganate.

These reactions,

successful in degradative studies of hyoscyamine (18), were -unsuccessful
when applied to cotinine degradation.

In other studies, through which

the lactam structure of cotinine was confirmed, Frank eriburg and Vaitenkunas
(35) degraded the pyridine ring of cotinine to obtain N-methyl-5pyrrolidone-2-carboxylic acid.
to the problem under study.

These reactions, however, are not relevant

Recently McKennis, Turnbull, Wingfield, and

Dewey (37) have reported the hydrolysis of cotinine.

The product of

hydrolysis has been identified as iq—(3 1-py^idyl) -U-methylaminobutyric
acid which is represented thus:
H
•C-CH2C H 2€OOH
NH
gh3
(mPB&cid)

Compound II

U” (31-pyridyl)“U-methylaminobutyric acid

Hence, the 5 carbon of the pyrrolidine ring of nicotine may be converted
to the carboxyl carbon of this substituted butyric acid.

Since

decarboxylation of this compound should yield the desired carbon atom
as carbon dioxide, it was decided to prepare this material to permit
further decarboxylation studies.

Isolation of Cotinine
The procedure employed for the isolation of cotinine was essentially
that of Pinner (33).

Occasionally small modifications were introduced

as seemed of advantage in the present studies.

30

Thirty-two ml. of bromine in 80 ml. of 80 per cent acetic acid was
added slowly with cooling to ll| ml. of 95 per cent nicotine dissolved
in 65 ml. of 80 per cent acetic acid..
separated as a dark reddish oil.

Almost immediately the perbromide

After a few minutes of heating^ this

mixture was added slowly to 375 ml. of hot water to effect solution of
the perbromide.
yellow crystals.

Upon cooling^ the perbromide crystallized as orangeThe perbromide was separated by filtration (Whatman

No. 2 paper on Buchner funnel)^ transferred to a beaker and covered with
50 ml. of 6N hydrochloric acid.

The transfer was completed by rinsing

the remaining perbromide crystals into the beaker from the paper and
filtering with about an equal volume of water.

Zinc dust was added in

small amounts to the mixture to reduce the perbromide.

The customary

practice has been to place the perbromide-acid mixture in an ice bath,,
which was permitted to return to room temperature as the reaction pro­
ceeded.

The beaker was covered to exclude light because,, on the basis

of several experimentsy it was believed that side reactions were thereby
reduced^, yielding a less colored product.

Then the zinc dust was added

periodically until the perbromide had dissolved and a colorless or pale
yellow solution resulted.

The excess zinc was removed by filtration and

the solution was neutralized with sodium hydroxide.

The resulting zinc
7

hydroxide was removed by filtration^ washed with water several times
and the washings were added to the filtrate.

This solution was transferred

7Usually the operation was continued until no positive test was
obtained with silicotungstic acid.

31

to a separatory funnel, made strongly alkaline with sodium hydroxide,
and extracted repeatedly with chloroform.

Upon removing the chloroform,

(which was usually recovered by distillation), crude cotinine remained
as a brownish oil.

The crude material was transferred to a 50-ml. Glaisen

flask (a few ml. of water were added to facilitate this transfer), which
was connected to a distillation flask (39)* which served as a receiver.
The distillation apparatus was evacuated^ heat was supplied with a small
heating mantle, and the cotinine distilled.

The cotinine usually

separated as a colorless or pale yellow opalescent oil.

The yield of

cotinine by this procedure ranges from about hS to 65 per cent.

Isolation and Identification of h-(3' -pyridyl)-U-methylaminobutyric Acid
The method (38) commonly employed in the isolation and purification
of the i_j.—(3* -pyridyl}-l^-methylaminobutyric acid is outlined below.
The

cotinine, prepared as described above, was transferred to a

100-ml. round-bottomed flask.

Ten ml. of water and 10 gms. of Ba(0H)2 .8H20

were added to the flask and the mixture refluxed for about twelve hours.
The flask was then permitted to cool, and the flask contents were concen­
trated to dryness under reduced pressure at J4.0 to 50°C.

The solid

obtained was extracted several times with acetone to remove non-hydrolyzed
cotinine.

A small quantity of water was added to the flask and the

resulting solution was saturated with carbon dioxide to separate the
barium as barium carbonate.

This precipitate was removed by filtration

with a sintered glass filter, the precipitate was washed with water,

8 Ibid

32

and the washings and filtrate were combined.

This solution was placed

in an ice bath and the carbon dioxide saturation was repeated once or
twice, until no further barium carbonate precipitation was observed.
After removal of the remaining barium carbonate, the filtrate was concen­
trated to dryness under reduced pressure at 25-30° C., using a flash
evaporator.

The product, usually a yellow oil, was dissolved in a minimal

amount of absolute alcohol at 60°C., and an equal volume of acetone was
added.

It was found convenient to filter the alcoholic solution through

a sintered glass filter into a 50-ml. suction flask.

Excess solvent

was removed by drawing a stream of air through the flask and gently
warming the solution.

Acetone was added after the temperature had been

brought to 60° and the solution reduced to a minimal volume (usually a
trace of crystalline product appeared at this stage).

The product

crystallized from the solution as white neediest after several hours of
refrigeration, it was separated by filtration, washed with a little
acetone and ethyl ether, and dried in a vacuum desiccator.
When crude cotinine was hydrolyzed, the final product isolated was
frequently gummy and impure; hence, it was found to be more practical
to distil the cotinine prior to the hydrolysis.

No consistent melting-point values could be obtained, but this
difficulty may have been related to the hygroscopic nature of the
HAPE&cid and its tendency to crystallize as a hydrate. Melting-point
figures ordinarily ranged from 120 to 125°^ although a value of 132-133°
has been reported (37)* Anal. Per cent calculated for C10H:L4N202.l/2

33

H 20:

0, 59.08; H, 7 .UUj N, 13.78; per cent found:

C, 59-21, 59*30;

H, 7.51, 7 •Uh; N, 13.82, 13.77.
Evidence supporting formula II was obtained by a neutralization
equivalent.

Since MAPBA.cid may be considered a substituted N-methyl

amino acid, the determination was carried out in 85 per cent ethanol.
After a quantity of the crystalline material had been dried for 10 hours
in a vacuum desiccator over Drierite, a 0.1051 g. sample was transferred
to a small beaker and dissolved in 50 ml. of 85 per cent ethanol.

The

solution was stirred by a magnetic stirring-bar and bubbles of nitrogen
were introduced by a capillary to reduce interference of carbon dioxide.
The electrodes of a Beckman glass-electrode pH meter, which had been
calibrated against a standard buffer solution, were placed in contact
with the solution and standardized sodium hydroxide was added dropwise.
A pH reading was taken for each 0.05 ml. addition of the standard base.
A plot was made of pH vs. volume of sodium hydroxide added.

From the

point of inflection of the curve, the volume of standard sodium hydroxide
was determined, which permitted calculation of the neutralization equi­
valent.

.Theoretical N.El: 1 9 Ehsperimentally determined N.E.: 195.

Although the elemental analysis would suggest a theoretical molecular
weight of 203 (considering a half molecule of water of hydration), the
drying conditions employed prior to the elemental analyses may have
been less rigorous than those described for the neutralization equivalent
determination, which could account for this discrepancy.

McKennis et al.

(37) have obtained the MAPBAcid as a monohydrate after air drying.

3k

Upon drying in a vacuum desiccator (1 mm.) over potassium hydroxide,
the water of* hydration was readily removed.

Hence, it seems feasible

that the water of hydration was lost by drying the MAPBAcid as described
above.
McKennis and co-workers (IjjO) have reported that compound XI develops
a red color when exposed to vapors of cyanogen bromide.

A red color

developed when some of the crystalline material, prepared as described
above, was exposed to cyanogen bromide vapors.

Some Attempted Degradations of It-(31-pyridyl)-U-methylaminobutyric Acid
After a survey was made of some possible methods of decarboxylating
the MAPBAcid, several preliminary experiments were carried out in an
effort to select the most practical method.

An effort to convert the

carboxyl grouping to an amide by means of thionyl chloride and concen­
trated ammonium hydroxide yielded a brownish solid; however, when some
of the characteristic amide tests were applied to this material, the
results were negative.

A n attempt to form the silver salt by adding

silver nitrate or silver acetate to an aqueous solution of the MAPBAcid
was not successful.

(In a later study, the silver salt was prepared by

the method described by Anker (Ul) .

However, as will be discussed,

considering the difficulties encountered with the silver salt of a
derivative of MAPBAcid, the author believes the silver salt of MAPBAcid
to be of doubtful value as a degradative intermediate.)
The preparation of the barium salt was then attempted, since
aliphatic acids are frequently decarboxylated by heating barium or

35

calcium salts.

Seventy mg. of the MAPBAcid was heated with a slight

excess of Ba(OH)2 .8H20 , under an atmosphere of nitrogen.
decomposed at 78 , dissolving the MAPBAcid.
at about 80

The hydrate

The temperature was held

until the water had evaporated, then raised slowly to

about 250 . After acidifying the residue and sweeping nitrogen through
the flask into barium hydroxide, only a trace of barium carbonate was
observed.

A blank sample of the barium hydroxide hydrate was treated

in the same manner; about an equal amount of carbonate was formed.
Abderhalden and Kantzsch (I4.2) have reported that glutamic acid is not
decarboxylated by heating its calcium salt; rather, a lactam is formed
at 180-185°.

Possibly cyclization also occurs upon heating the barium

salt of MAPBAcid.

McKennis ejb al. (I4O) have observed the cyclization

of the MAPBAcid side chain in acidic or neutral aqueous solutions,
whereby cotinine is reformed.

Since most of the conventional methods

of decarboxylation involve acidic or neutral conditions, it seemed very
desirable to consider methods of blocking the methylamino grouping of
the MAPBAcid before attempting a decarboxylation.

Some of the common

reagents employed to this end— acetyl chloride, benzoyl chloride, and
3 ,5 -dinitrobenzoyl chloride— were tested briefly, but no derivatives were
isolated.

However, since the compound MAPBAcid possesses a carboxyl and

a pyridyl grouping, its derivative would remain in ionized form under
acidic, basic, or neutral conditions; hence an acetylated or benzoylated
derivative is not readily separated by an adjustment of pH.
Nevertheless, since the Schotten—Baumann reaction has been a successful
means of benzoylating primary and secondary amino groups, it was

36

considered quite probable that a derivative could be formed with
MAPBAcid by this method.

Furthermore, during the preliminary tests

with benzoyl chloride, a striking difference in the colors of the
solutions was noted when MAPBAcid was compared with a cotinine or
water blank.

Hence it was decided to study this reaction further and

attempt to isolate the benzoylated derivative.

Isolation and Identification of U - O 1^pyridyl)~U-U~methylbenzoylamino-buiyric Acicf
" 1
The Schotten-Baumann reaction, commonly carried out in 10 per cent
sodium hydroxide, frequently involves the reaction of benzoyl chloride
with a primary or secondary amino grouping.

If the MAPBAcid would undergo

this reaction, it might be depicted as shown:

C-CH2CH 2C00H

+

+ 2Na0H

C-CH2CH 2C00Na
N-CH3
+ NaCl + 2H20
C«0

An extensive series of experiments were carried out in an effort
to isolate the benzoylated derivative.

Since benzoic acid is a product

of the hydrolysis of excess benzoyl chloride, the benzoic acid was
usually removed by filtration, after adjusting the pH to about 3*5.
In one experiment, after removal of benzoic acid, a blue precipitate
was formed when a solution of copper acetate was added.

Assuming the

material to be a copper salt, it was slurried in water, treated with
hydrogen sulfide, and thereby decomposed.

Upon removal of the water,

a white solid remained which appeared to be organic in nature, melted
at about 170°, and gave an acidic reaction when it was dissolved in

37

water and tested with litmus.

This product was later identified as

3 1-pyridyl)-U-N-methylbenzoylaminobutyric acid which has been
abbreviated to N-benzoyl MkPB&cid and is indicated as structure III.
H
C-CH2C H 2C00H

c h 3°

(III)

U-(3' -pyridyl)-U-N -methylb enzoylamino butyric acid

(N-benzoyl MAPBA,cid.)

However the yield of the above experiment was very low.

In sub­

sequent experiments } the proportions of reagents used and strength of
the sodium hydroxide were varied in attempts to improve yields.

Later

it was discovered that^ probably due to the presence of the benzoyl
grouping^ the white product can be crystallized from water.

The method

which was eventually adopted to obtain the N-benzoylated. compound is
outlined below 5 the procedure of Steiger (ii3) was followed in the initial
steps of the benzoylation.
Six hundred mg. (about 3 in moles) of MAPBAnid was dissolved in 3 ml.
of IN sodium hydroxide in an 8-inch test tube.

A magnetic stirring-bar

was placed in the tube5 which was placed in an ice bath.

Four ml. of IN

sodium hydroxide and O.U ml. of benzoyl chloride were introduced dropwise into the solution over a period of l£ to 20 minutes5 during which
time the solution was stirred vigorously.
developed a bright red color.

The solution at this stage

The stirring-bar was then removed., the

test tube was stoppered^ and placed on a shaker.

The tube was shaken

in an ice bath which was permitted to return to room temperature.

38

After the benzoyl chloride has dissolved and its odor disappeared
(usually 2 to 3 hours, occasionally longer), the sample was removed
from the shaker.
brown.

The red color had faded at this stage to a yellow-

The solution was transferred to a small beaker, placed in an ice

bath, and acidified with dilute hydrochloric acid to a pH of 3 *5
(a Beckman pH meter was used to measure the pH in this operation).
The precipitated benzoic acid was removed by filtration through a sintered
glass filter, washed with a little cold water, and the washings added to
the filtrate.

The filtrate was then made alkaline with dilute ammonium

hydroxide and concentrated to dryness under reduced pressure, at 1+0-500 C.
Extraction of the residue with cold n-butanol separated the benzoylated
product from the bulk of inorganic salts present.

The salts were removed

by filtration and the butanol extract was concentrated to dryness with
a flash evaporator at 5>0-60°C. A few ml. of water were added to this
residue to facilitate removal of the remaining butanol as an azeotrope.
The yellowish residue was dissolved in methanol and filtered with a
sintered glass filter into a 50-ml. suction flask containing 5 to 10 ml.
of water.

A stream of air was drawn through the filter into the flask

and the flask was warmed to about 50 to 60° C. to facilitate removal of
the methanol.

When the flask was permitted to cool to room temperature,

the solution cloudedjwith refrigeration, a white crystalline product
separated.

Since a trace of benzoic acid may be present in the product,

it was purified by recrystallization from water.
The N-benzoyl MAPBAcid crystallized as white needles in about 20-30
per cent yieldj the melting point was 170°.

A n a l . Per cent calculated

39

for C 17H-lqN 20 3 :

C, 6 8 .

H, 6 .08; N., 9.U0; per cent found:

6 8 .l4.5 s H* 6.09 , 6.1U; N, 9.38^ 9.31.

C 3 68.U2^

Unlike the MAPBAcid, the benzoylated

compound was not hygroscopic , and did not appear to decompose when a
sample was heated on a steam bath at a pH of 2.

No change in color was

observed when this compound was exposed to cyanogen bromide vapor.

From

an inspection of the structures of various compounds which give the
positive red color with cyanogen bromide

(HU)

including MAPBAcid, it

appears that a primary amino or methylamino grouping is commonly present.
The negative test with the N-benzoyl MAPBAcid may be explained by the
fact that the N-methyl grouping of MAPBAcid has been altered through
benzoylation.

Further evidence that the crystalline product is repre­

sented by structure III was obtained by determination of a neutralization
equivalent.

After a quantity of the product was dried in a vacuum

desiccator, 0.1211 g. was dissolved in 50 ml. of 85 per cent ethanol.
The remaining steps of the determination were carried out in essentially
the same manner as has been described for the determination of the
neutralization equivalent of MAPB Acid.

The neutralization equivalent

was calculated on the basis of the data obtained.
298; Experimentally determined N.E., 300.

Theoretical N.E,

The plot of pH vs. ml. sodium

hydroxide indicates that the N-benzoylMAPBAcid behaves as a typical
weak acid,, with a pK value of approximately 6 .6 .

The compound appears

to be quite soluble In methanol, ethanol, and chloroform; difficultly
soluble in water; almost insoluble in ether and carbon tetrachloride.

ko

Buly jy c ^ A ^ d.6^ ^ eac^-ons

ti-(3Arpyfidyl^-U-U-methylb enzoylamino-

Decarboxylation of the N-benzoylated M&PBA.cid was attempted
initially by means of the Hunsdiecker reaction.

This method involves

treatment of the silver salt of a carboxylic acid with bromine in an
organic reagent^ such as carbon tetrachloride.

The general reaction is

shown:
R-CH2-C00Ag + B r 2

CC14
----

R-CH2-Br + C 0 2 + AgBr

Since the decarboxylation apparatus and reagents must be moisture-free }
bromine was shaken with concentrated sulfuric acid and carbon tetra­
chloride was dried over phosphorus pentachloride before the experiments
were begun.

All glassware was oven-dried at 110° C.

The silver salt of the N-benzoyl HAPBA.cid was prepared from silver
nitrate and ammonium hydroxide 5 as described by Anker (1*1).

When the

silver salt was dried in a vacuum oven at 110°, as has been recommended
with certain preparations by the Hunsdiecker reaction (Hi)* the salt
decomposed after a few hours.

Several attempts were made to effect the

drying by lyophilizing the saltj or by combining a lyophilizing with a
vacuum-oven drying at 70-80° for 10-to 12 hours.
In carrying out this reaction^ the evolved carbon dioxide was swept
into a saturated solution of barium hydroxide by a stream of nitrogen^
which had been passed through calcium chloride.

Usually 150 to 300 mg.

quantities of the silver salt were slurried in 10 ml. of the carbon
tetrachloride with a calculated equivalent quantity of bromine.

The

apparatus used for these attempted decarboxylations will be described
later, when the Hofmann reaction is discussed.

U1

When the salt was dried by lyophilizing (up to 15 hours), no
significant decarboxylation could be effected through the Hunsdiecker
reaction.

With heating, the silver salt frequently tended to decompose.

Although this method of decarboxylation may be quite effective for this
compound under the correct experimental conditions, it was abandoned,
in view of the unpromising results, as another method was concurrently
being investigated.
The Hofmann method of degrading an aliphatic amide to an amine may
be represented by the following equation:
R-CH2CO-NH2 + Br2 + JiHaOH = R-CH2-NH2 + Na2C03 + 2NaBr + 2H20
The first attempts to convert N-benzoyl MAPBA.cid to an amide were by
conversion of the carboxylic acid to an acid chloride by means of thionyl
chloride.

The amide was obtained from the acid chloride by treatment of

the acid chloride with concentrated ammonium hydroxide or with benzene
saturated with ammonia.

The general procedure was to dissolve the

N-benzoyl HAFBAcid in a few ml. of dry chloroform, add an equivalent of
thionyl chloride, and reflux the mixture for periods varying from 30
minutes to two hours.

The chloroform was removed under reduced pressure

before the ammonium hydroxide (or ammonia in benzene) was added.
Although positive hydroxamate tests were obtained with hydroxylamine.HC1
and ferric chloride, to indicate some amide synthesis, the yields were
of very doubtful significance.

No crystalline amide could be isolated,

nor did preliminary attempts to decarboxylate the amide preparations
yield more than traces of carbon dioxide.
synthesizing the amide was investigated.

Hence, another method of

U2

Isolation and Identification of l j - ( 3 * -pyridyl) -li-N-methylbenzoylamino--- --- ------------ ------- ------butyric Acid
Methyl esters of carboxylic acids are frequently synthesized in
quantitative yields by the action of diazomethane on carboxylic acids.
In the present studies, ethereal solutions of diazomethane (U5) were
employed to synthesize the methyl ester of the N-Benzoyl MAPBAcid.
No attempts were made to purify or crystallize the ester.

Following

the method of Yang and Rising (U6) for the conversion of alpha-amino
acid esters to amides, N-Benzoyl MAPBAmide (Compound IF) was prepared by
the action of ammonia on the methyl ester of N-Benzoyl MAPBAcid.

CH3
(IF)

k~(3 1-pyr idyl)-li-N-methylb enz oylaminobutyramide

(N-Benzoyl MAPBAmide)
After several experiments, involving various proportions of reagents
and the N-Benzoyl M&PBAcid, and comparing ether and chloroform as
esterification media, the following general procedure was adopted to
synthesize Compound IF:
One-hundred-fifty mg. of dry N-Benzoyl MAPBAcid (0.5 mmole) was
transferred to a 125-ml. suction flask, placed in a hood.

About 30 to

U0 ml. of a solution of diazomethane in ether was then added to the
flask. The diazomethane was prepared by adding b00 mg. of nitrosomethylQ
urea to 20 ml. of hO per cent potassium hydroxide, covered with about

8The author is indebted to Mr. Arlelgh Dodson for a preparation of
this compound.

h3

25 ml. of ether, in an Erlenmeyer flask.

This mixture was placed in an

ice bath until the solid nitrosomethylurea had reacted.

The yellow

solution of diazomethane in ether was partially decanted into a beaker
containing a few pellets of potassium hydroxide.

Ether was added to

the Erlenmeyer flask and decanted into the beaker until the ether layer
in the flask was colorless $ essentially all of the diazomethane had
then been transferred to the beaker.

After a drying period of l£ to 20

minutes, the ethereal solution of diazomethane was added to the sample
of N-Benzoyl MAPBAcid, as has been mentioned.
A glass-covered magnetic stirring-bar (3 / 8 x 3/81') was then placed
in the solution to slurry the N-Benzoyl MAPBAcid.

Since the ester is

soluble in ether, whereas the acid is not, the course of esterification
could be conveniently followed by visual observation.
nitrogen were evolved as esterification proceeded.

Bubbles of

Occasionally it was

necessary to prepare and add more of the diazomethane solution^ however,
the N-Benzoyl MAPBAcid was usually dissolved within six to ten hours.
At the end of this period, the ether and diazomethane were removed by
drawing a stream of air through the flask by means of a water aspirator.
The residue in the flask, which was usually a pale yellowish glassy
material, was dissolved in about 25 ml. of methanol, which had been
saturated with ammonia at 0°C.

The stoppered flask was permitted to

stand at room temperature, with occasional shaking, for about 100 hours.
The methanol was then removed under reduced pressure, at 30-U0°G., and
the residue was dissolved in a few ml. of chloroform.
usually added to the solution at this stage.

Charcoal was

Although it was difficult

w*

to remove all of the color, a partial decolorizing appeared to give rise
to a purer product without seriously affecting yields.

With no addition

of charcoal, a brownish impurity was frequently precipitated with the
desired product.
The charcoal was removed by filtration and the volume of chloroform
was reduced by heating the solution.
until a white cloud had formed.

Ether was added to the hot solution

With refrigeration, the N-Benzoyl

MfiLPB&mide crystallized as white plates, which melted at II4I-IU20 C.
The average yield of the amide was about 85 per cent.
Evidence supporting structure IV was obtained by elemental analysis.
Anal.

Per cent calculated for C-^H-Lg^O^:

per cent found:

C, 68.65$ H, 6.1*5$ N, llj.-135

C, 68.72, 68.80; H, 6.1*7* 6.53$ N, ll*.23* 11*.09#

Further confirmation of amide synthesis was obtained by a ferric hydroxamate test.

The characteristic wine-red color developed after a few

minutes.

Decarboxylation
The Hofmann degradation of the N-Benzoyl MAPBAmide was carried out
in a three-necked, 125-ml. pear-shaped flask.

A 1-ml. pipette had been

sealed to a 2U/U0 joint and cut to such length that, when fitted into
the flask, it extended to about 1* cm. from the base of the flask.

The

joint was also connected by ground glass fittings to a small tube of
ascarite, which was, in turn, joined to a tank of nitrogen by tygon
tubing.

A glass piece, containing a stopcock, was fitted by a ground

glass joint to a second neck of the flask5 the piece was also connected

1+5

by

"tygon "tubing to a small funnel.

Glass fittings led from the third

neck of the flask to a saturated solution of barium hydroxide.

Hence^

a slow stream of carbon dioxide-free nitrogen could be swept through
the flask into the barium hydroxide to trap evolved carbon dioxide.
A carbonate-free solution of sodium hydroxide (approximately 0.25 N)
was prepared by adding 20 ml. of 0.5 N barium

chloride

to500 ml. of0.5

N sodium hydroxide and diluting the resulting

solution

toone liter.

After the barium carbonate had settled^ the supernatant was decanted
into a liter bottle 3 which was stoppered with

a siphon

assembly connected

to a tube of carbon dioxide absorbent.
The general procedure employed in the decarboxylation (1+7) is out­
lined below.
One hundred fifty mg. (0.5 m mole) of N-Benzoyl MAPBAmide was trans­
ferred to the pear-shaped flask.

The-flask., connected to the decarboxyl­

ation assembly^ was clamped into position in a small beaker of ice-water^
which had been placed on a magnetic stirring base.

A glass-covered

magnetic stirring-bar was placed into the flaskj another magnetic stirringbar and a thermometer were placed into the beaker.

The system was then

flushed with nitrogen (which is not passed through barium hydroxide at
this stage).

A sodium hypobromite solution was prepared by dissolving

0.025-0.030 ml. of bromine (approximately 0.6 m mole) in 12 ml. of the
carbonate-free sodium hydroxide^ in an ice bath.

The clear yellow

solution of sodium hypobromite was introduced into the flask by pipette^
as rapidly as possible^ to minimize the transfer of atmospheric carbon

U6

dioxide into the system.

The stirring was then begun., to dissolve the

solid. N-Benzoyl MAPBAmide in the sodium hypobromite.

After another brief

period of flushing^ the nitrogen was directed through the solution of
barium hydroxide.

When no barium carbonate formation was evident in the

barium hydroxide^ the beaker of ice-water was warmed with a microburner.
When the temperature rose to 50-60° C . y the clear yellow solution began
to cloud due to the formation of barium carbonate (from the reaction of
the residual barium ions in the hypobromite with the carbon dioxide of
decarboxylation).

The temperature was maintained at 70-80° C. for about

twenty minutes to complete the reaction.

The flask was then permitted

to cool to room temperature and 3 ml. of IN hydrochloric acid was intro­
duced through the funnel to make the solution slightly acidic.

A little

water was used to rinse the acid into the flaskj this column of water
also served to prevent escape of carbon dioxide when the stopcock of
the connecting glass piece was opened.
The system was flushed with nitrogen for about two hours.

At the

end of this period^ the barium carbonate was collected on a 2.1 cm. filter
paper (Whatman 5h0)y using a stainless steel filtering apparatus.

The

barium carbonate was washed with a little water and dried to a vacuum
desiccator (or in an oven at 110°, if desirable).

Yields of barium

carbonate ranged from U0 to 60 per cent.

Isolation and Identification of the Amine Derivative
The acidified medium^ remaining in the flask after the decarboxylation, may contain some unreacted N-Benzoyl MAPBAmide with the presumed

hi

product of the reaction, 3— (3T-pyridyl) -3-N -methylbenzoylamlnopropyfamine
designated by the abbreviation N-benzoyl MAPPAmine.

Some of the N-Benzoyl

MAPBAcid may also be present, due to hydrolysis of the amide.

It seemed

desirable to isolate the amine from the reaction mixture, since identifi­
cation of the amine would serve as evidence that the desired carbon atom
had been separated from the amide.

Furthermore, in later degradations

of radioactive materials, the radioactivity of the N-benzoyl MAPPAmine
plus that of the isolated carbon atom should total the radioactivity of
the N-Benzoyl HAPBlmide.
Separation of the amine, (indicated by structure V) from the other
constituents of the reaction mixture was accomplished by preparing a
thiourea derivative of the amine.
H
6 -CH2CH2N H 2

3 - (3 *-pyridyl)-3 -N-methylbenzoylaminopropylamine
(V)

(N-benzoyl MAPPAmine)

The general procedure that has been followed is described below:
The acidified solution remaining in the pear-shaped flask was made basic
to litmus with IN sodium hydroxide, the solution was transferred to a
separatory funnel, and extracted with chloroform until the extracts gave
no positive tests with the silicotungstic acid solution.

(In the experi­

ments performed, this chloroform extract has given basic tests with
litmus).

The chloroform was removed under reduced pressure, and the

residue was dissolved in 2 or 3 ml- of methanol.

Then 0.06 ml. of

58

phenyl isothiocyanabe was added and the mixture refluxed gently for
about

10

minutes. Water was added dropwise to the hot solution until

a white cloud formed.
derivative appeared.

Upon cooling, shiny white plates of the. amine
These were recrystallized from water and methanol,

washed with water, and dried in a desiccator.

M.p., 197-198°.

The constitution of the derivative, structure VI, was supported by
elemental analysis.

A n a l . Per cent calculated for C 23H 24N 40S:

H, £.98; N, 13.85; per cent found:

C, 68.52; H, 6 .08 ; N, 13.97.

C, 68.29;
The

derivative was also tested qualitatively for the presence of sulfur. .
A sodium fusion was carried out on a few mg. of the material, dissolved,
in methanol.

When the fusion mixture was filtered and the filtrate

tested with 0.1 per cent sodium nitroprusside and a solution of lead
acetate, the positive results gave evidence of the presence of sulfur
in the derivative.

C - € H 2G H 2-N-C-n(/

CH3

n

Thiourea derivative of
3 - (3 1-pyridyl)-3 -N -methylb enz oylpropylamine
(VI)

It should be noted that, although the N-benzoyl MAPPAmine has not
been identified as such and crystallized, this compound could serve as
a useful intermediate if further degradative studies should be desirable.
The presence of the amino grouping should facilitate attempts to shorten
the sidechain by one carbon atom.

Hence it is conceivable that each

o
o

O 4>
W co
w o

W
o

«

a

o
o
w

H
t-yi
*
g
O

/N

H

w

u
E

S
CQ

o

W

w
«o
otu
W o •»

o jzj o (

DSGHADATION OF NICOTINE WITH ISOLATION OF CARBON FROM POSITION 5 OF PYRROLIDINE RING

w
3
u?

o

o

o
■H eo

tJ

rH

co

to

o

o

52

carbon atom In the pyrrolidine ring of the nicotine molecule may be
recovered separately.

Degradation of Nicotine Labelled by Glutamic Acid-2-C14

The radioactive nicotine, isolated from the tobacco plants "which
had received glutamic acid-2-C14, yielded a total of 3.8 g, of nicotine
dipicrate.

For the purpose of degradation, this sample was mixed with

10.2 g. of non-radioactive nicotine dipicrate and the material was ground
in a mortar.

The nicotine dipicrate (equivalent to 3.66 g. of nicotine)

was mixed thoroughly in a 250-ml. round-bottomed flask; three samples
were then removed, plated, and counted.

Since the sample counts agreed

within experimental error3 sodium hydroxide was added to the flask and
the nicotine was recovered by an azeotropic distillation with water
through a Widmer column.

After about 250 ml. of distillate had been

collected in a beaker containing 15 ml. of 6 N hydrochloric acid, and
the distillate showed no further traces of nicotine., the solution was
concentrated to dryness on a flash evaporator.

The resulting nicotine

hydrochloride was dissolved in 10 ml. of water and 38 ml. of glacial
acetic acid was added.

The perbromide was formed by the addition of

8 ml. of bromine in 20 ml. of 80 per cent acetic acid to the nicotine
solution.

Seventy-five ml. of hot water dissolved the oily perbromide;

upon cooling, the characteristic orange-red crystals were formed.
The perbromide was covered with U5 ml. of 2N hydrochloric acid and
reduced with zinc dust.

To preserve the radioactive material as much as

possible, the zinc was not removed as the hydroxide; rather, the sample

53

was made strongly alkaline, so that the hydroxide, initially precipitated,
was redissolved.
chloroform.

The crude cotinine was separated by extraction with

Because of the small quantity of radioactive material, a

special distillation apparatus was constructed by joining two 10-ml.
distilling flasks (39)*

After the crude cotinine was transferred to

this apparatus, and the system had been evacuated by a vacuum pump, the
9
flask containing the crude material was lowered into a Wood*s metal
bath, which was heated with a microburner until the cotinine was distilled.
(Under these conditions, cotinine distilled at a bath temperature of
170-175° U.)

The cotinine (2.0 g.) was transferred to a 100-ml. round-

bottomed flask, to which was added 15 ml. of water and 10 g. of Ba(0H)2*
8H20, and the mixture was refluxed for twelve hours.

The MAPBAcid

(O.U65 g.J m . p *, 120-121° C.), obtained in the manner previously described,
was plated and counted.
The MAPBAcid (0.B50 g.) was benzoylated, as has been described.
The product was re crystallized from water and samples of this N-Benzoyl
MAPBAcid (0.163 g.j m.p., 169-170° C.) were plated and counted.
The N-Benzoyl MAPBAcid was then esterified with diazomethane, and
the ester was permitted to stand for 130 hours under 20 ml. of methanol
saturated at 0° with ammonia.

The N-Benzoyl MAPBAmide (O.llj.2 g.j m.p.,

[.2° C.) was crystallized from chloroform and ether.

After several

samples were plated and counted, the amide was decarboxylated and 0.056
g. of barium carbonate was isolated, in the manner previously described.
Samples of the barium carbonate were plated and counted.
9A n alloy of bismuth, lead, cadmium, and tin.

5k

The amine derivative was isolated and purified (O.O78 g.$ m.p._,
197-198° C.).

Samples of the derivative were plated and counted.

Results
The data obtained upon counting the various samples are presented
in Table IV.
s,

TABLE IV
ISOLATION OF CARBON 5 OF THE PYRROLIDINE RING OF NICOTINE AFTER
THE ADMINISTRATION OF GLUTAMIC ACID-2-C14

Compound

Maximum Specific Activity
(cpm/mM x 10“3)

1) Nicotine dipicrate

6.7

2) MAPBAcid

6 .I4

3 ) N-Benzoyl MAPBAcid

6.3

U) N-Benzoyl MAPBAmide

6 .U

5) Barium carbonate

3.1

6) N-Benzoyl MAPPAmine derivative

2.9

The specific activities of the intermediates in the degradation procedure 9
of lines 2 y 3 y and L* are approximately equal to the specific activity
of the nicotine dipicrate.

Hence., essentially all of the original

radioactivity was present in the amide (line I4) • Upon degradation of
the amide., one carbon atom was separated from the molecule^ and was
isolated as barium carbonate (line 50-

This carbon atom— originally

carbon 5 of the pyrrolidine ring of nicotine— contained approximately
half of the radioactivity of the amide.

The amine derivative (line 6 )^

containing the residual carbon skeleton of the amide^ possessed a

55

specific activity equal to that found with the barium carbonate •

From

tne previous study of nicotine degradation by potassium permanganate s
it is evident that the greater part of this radioactivity of the amine
is associated with the carbon atom joining the pyridine ring and the
sidechain (i.e.^ the 2 carbon of the pyrrolidine ring)*

degradation of Nicotine Labelled by 0rnithine-2~C14

The series of degradations to be discussed was initiated on 2.06

go of dibromo cotinine (3J5-dibromo~5-(3t“Pyridyl)-N-methyl-2-pyrrolidone) .
10

The radioactive nicotine from which the dibromocotine-

was synthesized

was isolated from tobacco plants (var. N. rustica), to which ornithine-2-C14
had been administered hydroponically.
Samples of the original nicotine dipicrate were plated and counted.
The dibromocotinine (33) (which- may be obtained from the previously
discussed perbromide by a sulfurous acid reduction) was covered with 10
ml. of &N hydrochloric acid and reduced with zinc dust.

The subsequent

stepSj resulting in the isolation of crude cotinine , were similar to
those carried out with the radioactive glutamic and have been described
previously.

The crude cotinine was distilled with the apparatus discussed

in the glutamic acid—2-£J14 studies.

The distillate (0.6 g.) was trans­

ferred to a 50-ml. round-bottomed flask to which 5 ml* of water and 5 g.

10This sample was kindly furnished by Dr. L. J. Dewey.

56

of Ba(0H)2 «8H20 were added.

After this mixture was refluxed for 12

hours, the flask was permitted to cool and the MAPBAcid (0.350 g.;
m.p., 121

C.) was isolated.

Samples of the MAPBAcid were plated and

counted.
The benzoylation was carried out on 0.330 g. of this preparation,
dissolved in 1.7 ml. of IK sodium hydroxide.

Another 2 ml. of the IN

sodium hydroxide and 0.2 ml. of benzoyl chloride were added in the
manner previously discussed; the recrystallized N-Benzoyl MAPBAcid
(0.100 g.; m.p., 169-170° C.) was plated and counted.

The 0.100 g.

sample was esterified with diazomethane and the product was dissolved
in 20 ml. of methanol, saturated with ammonia at 0°.

After 100 hours,

during which period the solution was shaken occasionally, the N-Benzoyl
MAPBAmide (O.O8I1 g.; m.p., lljl-llj.20 C.) was isolated; samples of this
preparation were plated, and counted.
Unfortunately the 8I4. mg. of the amide preparation seemed to be
virtually the lower limit of quantity necessary for conducting a
decarboxylation— at least, under the selected experimental conditions.
However, it was decided to carry out the decarboxylation without dilut­
ing the amide preparation with non-radioactive carrier.
Using calculated quantities of reagents, the decarboxylation was
effected and the carbon dioxide was separated as barium carbonate.
However, only 15 mg. of barium carbonate were collected.
was plated and counted.

This material

Because of the difficulties of counting small

quantities, an 8.8 mg. sample of the barium carbonate was diluted with
two parts of non-radioactive barium carbonate and the mixture was plated

57

and counted.

The calculated specific activity of the original material^

based on the count of the diluted sample 9 was not significantly dif­
ferent from the values obtained by counting the undiluted barium carbonate.
The amino derivative (0.027 g.> m.p._> 19U-1950 C.) was isolated and
recrystallized; samples were plated and counted.

Results
The data obtained upon counting the samples is shown in Table V.

TABLE V
ISOLATION OF CARBON 5 OF THE PYRROLIDINE RING OF NICOTINE AFTER THE
ADMINISTRATION OF 0RNITHINE-2-C14

Compound

Maximum Specific Activity
(cpm/mM x 10~3)

1) Nicotine dipicrate

8.1

2} MAPBAcid

8.2

3) N-Benzoyl MAPBAcid

8.3

It) N-Benzoyl MAPBAmide

8.1

5) Barium carbonate

3.0

6 ) N-Benzoyl MAPPAmine

2.9

A comparison of Tables IV and V reveals a striking similarity in
the patterns of nicotine labelling by glutamic acid-2-C14 and omithine2-C:l4.

The intermediates of degradation of lines 2 y 3* and lt5 had the

same radioactivity as the nicotine dipicrate of line 1.

The barium

carbonate of line 5— derived from carbon 5 of the pyrrolidine ring of
nicotine— was radioactive5 although this radioactivity was somewhat

58

less than half of the radioactivity found with the N-BenzoylMAPRAjnide.
However } the ainine derivative (line 6) possessed approximately the same
amount of radioactivity as the barium carbonate •
The reason for the rather low specific activities of the barium
carbonate and the amine derivative has not been immediately apparent s
although it is probably associated with the difficulty of handling and
counting these small samples.

It is possible that the radioactive

barium carbonate was diluted by a small quantity of non-radioactive
carbonatej originating during the isolation of the radioactive carbonate*
from the contact of atmospheric carbon dioxide with the excess barium
hydroxide being removed.
However , the studies of Dewey and Byerrum (19) and of Leete (20).,
with ornithine-2 -C14, (discussed in the Introductory portion of this
thesis) have established that half of the radioactivity of nicotine
labelled by ornithine-2-C14 Is located at carbon 2 of the pyrrolidine
ring,

in another study with ornithine-2-C14 ^ Leete (1|8) oxidized the

radioactive nicotine with nitric acid.

A nitropyrazole^ obtained as a

by-product of nitric acid oxidation., possessed about half of the radio­
activity of the nicotine. Since carbon 5 of the pyrrolidine ring is
presumably expelled during formation of this nitropyrazole^ this finding

lends support to the data of Table V (i.e., that carbon 5 is radio­
active., and possesses a specific activity equal to carbon 2 of the
pyrrolidine ring) .
Hence it Is concluded^ from these considerations and from the data
presented in Tables IV and V* that the pyrrolidine ring of nicotine is

59

labelled in essentially the same manner by ornithine-2-C14 and glutamic
acid-2-€14.

DISCU3SI0I

60

DISCUSSION

From the preceding study, it is evident that the pyrrolidine ring
of nicotine is labelled in essentially the same manner by glutamic acid2-C14 and ornithine-2-C14.

The radioactivity of the pyrrolidine ring

was found to be distributed equally between positions 2 and 5 ^ when
either glutamic acid-2-C14 or ornithine—2-€14 was employed as a precursor.
The studies on nicotine biogenesis of Dewey (21) and Leete (20)
with ornithine-2-C14 , previously discussed, led to the postulation of a
symmetrical ihtermediate in the pathway of biosynthesis of the pyrrolidine
ring.

This intermediate, which is evidently derived from the intact

carbon chain of ornithine, has given rise to a certain amount of specu­
lation, regarding its structure and manner of formation.
One consideration suggested that the intermediate might be putrescine,
a four-carbon symmetrical diamine, which could arise directly from the
decarboxylation of ornithine 9

Another possibility is the conversion of

ornithine to glutamic - X -semialdehyde by transamination or oxidative
deamination.

The semialdehyde could be oxidized to glutamic acid, which,

by loss of the amino grouping, would give rise to alpha-keto glutaric
acid.

Decarboxylation of the latter compound, in the tricarboxylic acid

cycle, would yield another symmetrical compound, succinic acid.

The

succinic acid could also conceivably result from the oxidation of
succindialdehyde, which could arise from putrescine, through oxidative
deamination.

61

Another pathway to a possible symmetrical intermediate involved
cyclization of the semialdehyde to give A* -pyrroline-5-carboxylic acid.
Becarboxylation of this compound would produce pyrroline, which could
become symmetrical b y a shifting of the double bond.

Also, the symmetri­

cal compound pyrrolidine results from the decarboxylation of proline,
which could arise from the reduction of

A* ~pyrroline-5-carboxylic acid.

Although putrescine is generally associated with the metabolism of
microorganisms, certain evidence suggests that it may also play a role
in the metabolism of higher plants.

Cromwell (Jj.9) reported the prepara­

tion of an aizyme system from A_. belladonna, capable of oxidizing
putrescine to ammonia and an aldehyde.

Furthermore, he isolated small

quantities of putrescine from the leaves and upper stems of A. belladonna
and LL stramonium. When putrescine was injected into D. stramonium L .,
Cromwell (50) found an increased synthesis of hyoscyamine. On the other
hand, Diaper, Kirkwood, and Marion (5l) have reported that the hyoscyamine
isolated from D. stramonium L., following administration of 1,U~C14
putrescine, was not radioactive.

However, as has been mentioned earlier,

Leete, Spenser, and Marion (18) demonstrated that ornithine-2-C14 is
incorporated into the pyrrolidine ring of hyoscyamine.

On the basis of >

these radioactive studies, the immediate precursor of the pyrrolidine
ring of hyoscyamine would appear to be more closely related to ornithine
than to putrescine.
Leete (52) has recently found putrescine to be a precursor of the
pyrrolidine ring of nicotinej however, the efficiency of putrescine in
this respect, as compared to ornithine, has not been reported.

62

From t-he studies of Cromwell (50) with hyoscyamine, it is conceiv­
able that, in the biosynthesis of nicotine, putrescine may be oxidatively
deaminated to ^ -aminobutyraldehyde.

This compound might then either

cyclize to pyrroline or possibly condense with the precursor of the
pyridine molecule before cyclizing.

If, however, putrescine were con­

verted to succindialdehyde through oxidative deamination at both ends
of the molecule, condensation of the succindialdehyde with ammonia would
yield the symmetrical compound 2,5-dihydroxy pyrrolidine.

The pyrrolidine

ring might then be formed by loss of water from this intermediate,
followed by a reduction.

Hence, it is conceivable that some synthesis

of the pyrrolidine ring of nicotine may be effected’ by way of putrescine,
although this pathway is probably of minor significance.
Succinic acid has also been mentioned as a possible symmetrical
intermediate in the formation of the pyrrolidine ring of nicotine.
Reduction of the succinic acid would yield succindialdehyde, which has
been mentioned above.

If a condensation of the succinic acid and ammonia

were to occur, 2,5-diketopyrrolidone would be formed, which might possibly
be considered a precursor of the pyrrolidine ring.

Presumably glutamic

acid could be incorporated into the pyrrolidine ring either by way of
succinate or by way of ornithine and putrescine.
However, it has recently been reported (38) that carbon-lU-labelled
succinate was administered to tobacco plants.

"When the nicotine was

isolated, no significant radioactivity was found to be associated with
the pyrrolidine ring.

Hence, it would appear that little, if any,

pyrrolidine ring synthesis occurs by this pathway.

The present study

63

would seem to confirm this finding.

From Table I, it is seen that

ornithine is incorporated into the pyrrolidine ring to a far greater
extent than is glutamxc acid.

Xf succinate were an important inter­

mediate in the biosynthesis of nicotine, the opposite result should
have been found, since ornithine may be converted to succinic acid by
way of glutamic acid.
Another possible pathway of pyrrolidine ring biosynthesis involves
the intermediates glutamic--semialdehyde and A 1-pyrroline-5-carboxylic
acid,, from which the symmetrical intermediates pyrroline or pyrrolidine
might arise.

It should be mentioned, in this connection, that pyrroline

and pyrrolidine may also conceivably be formed from ornithine by way of
A 5-pyrroline-5-carboxylic acid.

Loss of the alpha-amino grouping of

ornithine through oxidative deamination or transamination would yield
alpha-keto delta-amino valeric acid, which, by cyclizing, would give the
A 5-pyrroline-5-carboxylic acid.

Glutamic acid could presumably form

from the G-keto glutaric semialdehyde and a-keto glutaric acid.

Some

evidence for this sequence has been obtained from studies in vitro with
animal tissue preparations (5>3 )*
However, the bulk of evidence shows that the former pathway is more
important in animal systems and microorganisms— that is, by way of
glutamic-semialdehyde and A*-pyrroline-5-carboxylic acid.

Stetten (f>U),

for example, conducted comparable experiments in which two samples of
ornithine, labelled in either the alpha or delta position with N 15, were
fed to mice.

Following the isolation of glutamic acid and proline from

the animal proteins, it was apparent that the major loss of N 15 had

6k

been from, the delta position of ornithine, which indicated that glutamic
semialdehyde is the likely intermediate in this conversion of ornithine
to glutamic acid or proline.
Other studies of IS. coli, hh crassa, and Th utilis, have shown
that the principal pathway in the conversion of glutamate to proline by
these organisms is by glutamic- Jf- semialdehyde and A 1-pyrroline-5>~
carboxylate (5) •

This finding has been established by isotopic competi­

tion experiments and by the administration of glutamic semialdehyde,
prepared synthetically, to mutant strains of these microorganisms.
The present study suggests that one or more symmetrical inter­
mediates are involved with the incorporation of glutamic acid or ornithine
into the pyrrolidine ring of nicotine.

This does not necessarily imply

the function of a common intermediate for glutamic acid and ornithine,
although, on the basis of the previous discussion, such a conclusion
would seem reasonable.

There is little information extant, regarding

the glutamate-omithine-proline relationship in higher plants, as was
noted earlier.

If, however, ornithine and glutamic acid give rise to

glutamic semialdehyde in the tobacco plant, as would be plausible on
the basis of animal and microorganism studies, two symmetrical inter­
mediates could result— namely, pyrroline and pyrrolidine.

At present,

there is no conclusive evidence favoring one or more of these or other
intermediates in the synthesis of the pyrrolidine ring of nicotine.
The biosynthesis of the pyridine ring of nicotine has never been eluci­
dated, although certain evidence (28,29,30) appears to link the immediate
precursor of this ring to intermediates of glycolysis or the

65

Furthermore, Dawson (55) y in bis studies of

tricarboxylic acid cycle.

nicotine, has reported some incorporation of radioactive nicotinic acid
into the pyridine ring.
Hence, in view of the uncertainty regarding the biosynthesis of
this ring., the manner of condensation of the two rings of nicotine has
never been explained.

The precursor of the pyrrolidine ring may or may

not cyclize before a condensation with the pyridine ring precursor.
It is possible that the precursors of these rings may be linked before
either ring has formed.

However, such considerations are merely specu­

lative at present and must necessarily await the -evidence of further
studies.
Present knowledge indicates that the carbon skeleton of the pyrroli­
dine ring probably originates with the carbon chain of glutamic acid.
Calvin (56) has observed, that C l40 2 is rapidly incorporated into glutamic
acid, during the process of photosynthesis.

Following the fixation of

a molecule of ,
C 1402 into the initial acceptor, ribulose diphosphate,
the atom of carbon-ll* may be traced through the intermediates of glycoly­
sis into the tricarboxylic acid cycle, to alpha-keto glutaric acid.
Glutamic acid is formed from alpha-keto glutaric acid by reductive
amination, in quantities that are contingent on the immediate require­
ments of the plant.

This reaction is catalyzed reversibly by the enzyme,

glutamic dehydrogenase (57

which is found in a wide variety of plant

tissues (58}.
Naylor and Tolbert (9) administered glutamic acid, uniformly
labelled b y carbon-lU* to barley seedlings; then, after periods of

66

one-half to three hours, methanol-water extracts of the plant material
were chromatographed.

The greater part of the radioactivity of these

extracts was found with glutamine,

-amino butyric acid, and aspartic

acidj no significant evidence of radioactive arginine, ornithine, or
proline was reported.
However, current knowledge of living systems limits the source of
the carbon chain of ornithine or proline to glutamic acid.

In a study

!L* co^-:i-? ^’or example, Abelson (5 ) has found, by isotopic competition
experiments, that glutamic acid is the precursor of ornithine and proline
in this microorganism.

It seems very likely that a similar relation

obtains among these amino acids in higher plants, although little experi­
mental evidence for this is available.
The experimental studies, mentioned above, suggest that the
reservoirs of free ornithine or proline in the higher plants are prob­
ably small relative to the reservoir of glutamic acid.
Varda (59)5 and Roberts and Wood
in tobacco leaves.

(60)

Commoner and

have studied the free amino acids

Glutamic acid was found in significant quantities.

Proline appeared to be present to a lesser extent, but no ornithine was
reported, although traces of arginine were noted.

If ornithine is

synthesized from glutamic acid by way of g l u t a m i c - if-semialdehyde, the
reduction of the glutamic acid to the semialdehyde may be the ratedetermining step.

The experiments of Abelson (5) with E. coli would

lend some support to this possibility in the bacterial system.
The mechanism of the reduction of glutamic acid to the semi­
aldehyde is not known.

However, Black and Wright (6l) have reported

67

that, aspartic acid forms B-semialdehyde by way of B-aspartyl phosphate
in cell-free extracts of baker* s yeast.

By analogy^ it is possible

that 2f -glutamyl phosphate is an intermediate in. this conversion.
From the present study., it is evident that^ when ornithine-2-C14
and glutamic acid-2 -C14 are administered to tobacco plants in comparable
experiments } the nicotine from the ornithine-fed plants is considerably
more radioactive.

These results may be explained if it is assumed that

the conversion to the symmetrical precursor occurs more readily with
ornithine than with glutamic acid_, and that the reservoir of free
ornithine in the plant is relatively small compared to glutamic acid^
as seems very likely.

The relation of proline to the immediate precursor

of the pyrrolidine ring is uncertain.

In his early work^ Dawson (lit)

noted an increased synthesis of nicotine in tobacco shoots cultured in
aqueous solutions of 1-proline.

However, a study of nicotine biogenesis

with prdline^ labelled by carbon-ll^ is desirable.

The capacity of

pro line to function as a precursor of the pyrrolidine ring might then
be assessed in relation to the foregoing studies with ornithine and
glutamic acid.

SUMMARY

68

SUMMARY
1* Tobacco plants that had been fed glutamic acid-2-C14 synthesized
radioactive nicotine.

The nicotine was considerably less radioactive

than had been reported for a comparable study with ornithine-2-C14.
2.

A degradation of the radioactive nicotine demonstrated that about

ten per cent of the carbon-llj. resided with the pyridine ring.

Half of

the remaining radioactivity was located at position 2 of the pyrrolidine
ring.
3 • A procedure was developed for the degradation of the nicotine
molecule which permits isolation of the carbon atom from position 5 of
the pyrrolidine ring.

The residual carbon skeleton of the molecule may

be recovered intact as a derivative of a substituted aliphatic primary
amine.

The intermediate produced at each step of this degradation has

been isolated and identified.
h* Hicotine labelled by glutamic acid-2-^G14 was degraded by this

procedure.

Half of the carbon-lU was associated with position 5 of the

pyrrolidine ring.
derivative.

The remaining radioactivity was found with the amine

Another degradation demonstrated that ornithine-2-G14

effects the same pattern of C 14-incorporation into the pyrrolidine ring
as was found in the glutamic acid study.
5* The similarity in this pattern of labelling by glutamic acid-2-C14
and ornithine-2 -C14 supports the postulation of a symmetrical intermediate
along the pathway of incorporation of the amino acid carbon chain into
the pyrrolidine ring.

HEFEBEHCSS

69

REFERENCES

1. Dakin, H. D., J. Biol. Chem. 13, 513 (1913).
2. Stetten, M. R., and Schoenheimer, R., J. Biol. Chem. 153, 113 (1955).
3. Roloff, M., Ratner, S., and Schoenheimer, R., J. Biol. Chem., 1 3 6 ,
561 (191*0).
--5. Sallach, H. J., Koeppe, R. E., and Rose, ¥. C., J. Am. Chem. Soc.
£3, U500 (1951).
5. Vogel, H. J., and David, B. D., J. Am. Chem. Soc., 7^ 109 (1952).
Vogel, H. J., Proc. Natl. Acad. Sci. U. S. 39, 578 ^1953).
Tatum, E. L., Proc. Natl. Acad. Sci. U. S. 31, 215 (1955).
Abelson, P. H., J. Biol. Chem. 206, 335 (195Sj.
6. Bonner, D., Am. J. Botany 33_, 788 (1956).
Strassman, M. and ¥einhonse, S., J. Am. Chem. Soc., 75? 1726 (1952).
7. Stetten, M. R., !tA Symposium on Amino Acid Metabolism ,n edited by
McElroy, ¥. E., and Glass, H. B., The John Hopkins Press, Baltimore,
Md. (1955)j, PP. 277-290.
Vogel, H. J., ibid., pp. 335-356.
8. Morgan, A- and Marion, L., Can. J. of Chem., 35? 1705 (1956).
9. Naylor, A. ¥., and Tolbert, N. E., Physiologia Plantarum 9j 220 (1956).
10. Robinson, R., J. Chem. Soc. Ill, 876-899 (1917).
11. ¥5nterstein, E., and Trier, G., 1,Die Alkaloids/' Bomtraeger, Berlin,
Second Edition, 1931.? PP« 356-357* 871.
12. Klein, G. and Linser, H., Planta 20, 570-575 (1933).
13. Fetrosini,

G.,

II Tabacco 58? 39~55 (1955)*

15. Dawson,
Dawson,

R. F.,
Science, 87? 257 (1938).
R. F., Plant Physiol., 15, 579-591 (1939).

15. Dawson,

R. F.,

C. A.,7127e (1955)•

Advances in Enzymology, 2? 203-251 (1958).

16. Klein, G. and Linser, H., Z. physiol. Chem. 209.? 75 (1932).

70

17. James^ W. 0., Nature 158, 655 (1956).
18. Leete, E., Spenser, I. D., and Marion. L., Canad. J. Chem. 32,
1116 (195U)^
19. Dewey, L. J., Byerrum, R. U .
Acta 18, lUl (1955).

and Ball, C. D., Biochem. Biophys.

20. Leete, E., Chemistry and Industry, 537 (1955).
21. Dewey, L. J., "Studies on the Biosynthesis of Nicotine and Lignin,"
Ph. D. Thesis, Michigan State University, 1955*
22. Dawson, R. F., Am. J. Botany, 29, 66-71 (1952).
23. Gordon, H., Anal. Chem. 23, 1853 (1951).
25. Smith, C. R., Ind. Eng. Chem., 35, 251-252 (1952).
25. Calvin, M. C., Heidelberger, C., Reid, J. C., Tolbert, B. M., and
Yankwich, P. F., "Isotopic Carbon," John Wiley and Sons, New York,
1959, p. 119.
26. Brown, S. A., "Studies on Methylation Reactions in Plants: The
Origin of the Methyl Carbon of Nicotine Formed by Nicotiana rustica,"
Ph. B. Thesis, Michigan State College, 1951*
27. Laiblin, R., Ann. 196, 129-182 (1879).
28. Leete, E., Bell, V. B., and Leitz, F. H. B., Abstracts, San
Francisco Meeting, American Chemical Society, April, 1958.
29. Griffith, T., "A Study of Ribose Metabolism in the Tobacco Plant,"
Ph. D. Thesis, Michigan State University, (1958).
30. Fronters-Aymat, M. E., "A Study of Pyruvate Metabolism in the
Tobacco Plant," M. S. Thesis, Michigan State University, 1956.
31. Huber, C., Ann. 1313 257 (1865).
32. Cahours, A.^ and Etard, A.? Bull. soc. chim. (2), 35? 559 (I880) .
33. Pinner, A.? Ber., 25? 2807 (1892).
35* Byerrum, R. U., personnal communication.
35. Frankenburg, W. G., and Vaitekunas, A. A.? J. Am. Chem. Soc. 79?
159 (1957).

71

36. Yankwieh, P. E., Norris, T. H., and Huston, J., Anal. Chem. 19.

H39 (1957).

— ^

37 * McKennis, H., Turnbull, L. B., Wingfield, H. N., and Dewey, L. J.,
J. Am. Chem. Soc. 80, 1635 (1958).
3 8 . Dewey, L. J*, personal communication.
39. Fisser, L., "Experiments In Organic Chemistry," D. C. Heath and Co.,
Boston, 1955* pp. 213-220.
HO. McKennis, H., Turnbull, L. B., and Bowman, E. R., J. Am. Chem. Soc.
7£; 63H2 (1957).

Hi.

Anker, H.

S., J. Biol. Chem. 195;

H 2 . Abderhalden,

E., and Kantzsch, K.,

177 (1952).
Z. physiol. Chem. 65,

H3.

Steiger, R., J. Org. Chem. 9j 396

HH*

Larsen, P. S., Haag, H. B., and Finnegan,
and Exp. Therapeutics, 86, 239 (19H6).

HH 8 (1910).

(19HH) .
J. K., J. of Pharmacol,

H5- Weinstock, H. H.; and May, E. L., J. Am. Chem. Soc. 62, 3266

H6.

Yang, P. S., and Rising,

M.M., J.

(19H0).

Am. Chem. Soc. 53, 3183 (1931).

H7. Wallis, E. S., and Lane,
J.F., "The Hofmann Reaction," Organic
Reactions, Vol. Ill, Chapter 7; J- Wiley and Sons, I9H6 .

H 8 . Leete,

E. and Siegfried,

K.J., J.

H9.

Cromwell,

B. T., Biochem. J., 3£,

50.

Cromwell,

B. T., Biochem. J.,

Am. Chem. Soc.
722-726

79, H529 (1957).

(19H3) .

37, 717-722 (19H3).

51. Diaper, D. G. M., Kirkwood, S., and Marion, L., Can. J. of Chem.

29,

96H (1951).

52. Leete, E., J. Am. Chem. Soc. (in press).
53. Krebs, H. A.; Enzymologia

5H.

53 (1939).

Stetten, M. R., J. Biol. Chem. 189,

H99 (195l) *

55. Dawson, R. F., Christman, D. R., Anderson, R. C., Solt, M. L.,
DtAdamo, A. F., and Weiss, U., J. Am. Chem. Soc. 7 8 , 2655 (1956).
Dawson, R. F., Christman, D. R., D*Adamo, A. F., Solt, M. L.; and
Wolf; A. P.; Chemistiy and Industry, 100 (1958).

72

56.

Calvin, M., J. Chem. Ed.

26, 639 (19^9).

57.

Ton Euler, H. and Adler, E . , Enzymologia

58.

Damodaran, M. and Nair, K. R., Biochem. J. 32^, 106U (1938).
Bonner, J. and Wildman, S. G., Arch. Biochem. 10, U97 (19U6).

59.

Commoner, B. and Tarda,

60.

Roberts, E. A. H., and Wood, D. F., Arch. Biochem. and Biophysics,

N.,

21 (1939).

J. Gen. Physiol.

36, 791 (1953).

33, 299 (1951).
61.

Black, S., and Wright, Iff. G., Fed. Proc.

13 > 18U (195U) •

73

APPENDIX

The formula used in correcting the observed count to zero sample
thickness was:

A

« C o* M
¥ *b

m

where A m « maximum specific activity (counts/minute/millimole)
C0 = observed count (counts/minute)
M

- molecular weight of compound

¥

- weight of sample counted

b -= fraction of maximum activity at the sample thickness used
(T)— obtained from self-absorption curve
Sample calculation:
Nicotine dipicrate —

C 0 - 178 c.p.m.
M

, ¥ = 5U.9

- 620 ^ T = 1 9 -k mg/cm2 9 b = 0.31

A m *=

~ 6.U9 x 103 c.p.m./mM