ABSTRACT

THE RELATION BETWEEN BOAR ODOR INTENSITY
AND SdLANDROST-16-ENe3-0NE CONTENT
IN PORK FAT

By
Raymond E. Thompson, Jr.

The present investigation was undertaken to determine
the relationship between boar odor scores for fat samples
derived from cryptorchid (males with retained testicles)
pigs and the levels of SIPandrost-16-en-3-one, which has
been claimed to be responsible for this undesirable odor.
In order to ascertain the levels of 51-androst-16-en-3-one,
a stable isotOpe dilution/carrier technique (Bieber, M.A.
g§H§;., 1972, Anal. Biochem. 47:264) was adapted for use
on fat samples. After addition of the deuterium labeled
Supandrost-16-en-3—one to the fat sample, the mixture of
deuterium and protium forms was purified. The resultant
fraction was analyzed by a combination of gas chromato-
graphy and mass spectrometry. The ratio of the deuterium
and protium forms was obtained by multiple ion detection
using the accelerating voltage alternator accessory of an
LKB 9000 gas chromatograph-mass spectrometer (Holland,‘g§
§;., 1973. Anal. Chem. 45:308).

Raymond H. Thompson, Jr.

The addition of the labeled form of the steroid enabled
small amounts of the compound to be partially purified and
separated by gas chromatography without total loss through
adsorption. Losses during isolation and analysis were auto-
matically accounted for by the method since the ratio of
the labeled and unlabeled species would remain constant.

The amount of unlabeled Bipandrost-16-en-3-one was deter-
mined from the ratio of the labeled to the unlabeled species.
This procedure was utilized in determining the level of 9&-
androst-16-en-3—one in 21 fat samples from cryptorchid pigs.

Deuterium labeled forms of the five most common 019-Afl6
steroids were synthesized in order to permit analysis by
stable isotope dilution. The labeled c19.¢§5 steroids
synthesized were as follows: d2-, d3- and d4-5dpandrost-
16-enp3-one; d1- and d4- SdPandrost-16-en-34-ol; d4-5d—
androst-16—en-3p—ol; d2-4,16-androstadien-3-one and
d2-5,16-androstadien-2p-ol.

Several fat samples from cryptorchid pigs were also
rated for intensity of the undesirable (boar) odor by either
a trained laboratory panel or by a meat industry panel and
the values were correlated with the levels of Sd-androst-
16-en—3—one determined by the isotope dilution procedure.

The correlation coefficient for the meat industry panel and the
level of Supandrost-16—en-3-one was only 0.40, which was not
statistically significant (P‘<.05). The corresponding corre-

lation coefficient for the trained laboratory panel was 0.51

Raymond H. Thompson, Jr.
which was significant at P‘<.05. Although the trained
laboratory panel apparently were able to detect the level
of Bapandrost-16-en-3-one, the relationship only accounted
for 25% of the variation. This indicated that Sarandrost-
16-en-3-one was not solely responsible for the undesirable
odor and that other related C19—£§6 steroids may account
for the low relationship. Further support for this view-
point is evident by the fact that several fat samples
having high odor intensity scores were found on analysis
to have only low levels of Sdpandrost-16-en-3-one.

The correlation coefficient between the odor scores
for the two panels was 0.52 (P<(.05). Although the relation—
ship was significant, the relationship accounted for only
27% of the variation in odor scores. Thus, the two panels
did not closely agree on the odor. Some reasons for the
possible differences are discussed.

Although the labeled 019-zl6 steroid standards were
synthesized and are now available in our laboratory, time
did not permit analysis. These labeled standard compounds
will be useful in subsequent studies on the significance of
019-cfl6 steroids in both pigs and human beings.

THE RELATION BETWEEN BOAR ODOR INTENSITY
AND 5dkANDROST-16-EN-3-ONE CONTENT
IN PORK FAT

by

3.9

as
Raymond R? Thompson, Jr.

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Food Science and Human Nutrition

1975

ACKNOWLEDGMENTS

The author expresses his appreciation to Dr. A. M.
Pearson for his direction and encouragement throughout
this study and during preparation of this thesis. The
author also wishes to thank Dr. L. L. Bieber, Dr. L. R.
Dugan, Dr. D. R. Dilley and Dr. J. W. Allen for serving
as members of the author's guidance committee.

Appreciation is specifically expressed to Dr. C. C.
Sweeley and Dr. J. F. Holland and their staff at the
lass Spectrometry Facility in the Department of Biochemistry,
Michigan State University, for allowing the author use of
their laboratory facilities and for guidance during the
course of the research reported herein.

Finally, the author is especially grateful to his wife,
Jackie, and children, Scott and Jon, for their support,

understanding and sacrifice.

ii

TABLE OF CONTENTS

Page
INTRODUCTION . . . . . . . o . . . . . 1
REVIEW OF LITERATURE . . . . . . . . . . 3
Incidence of Boar Odor . . . . . . . . 3
Disposition of Carcasses with Sex Odor . . 5
Chemical Identity of Boar Odor . . . . . 7
Odor Characteristics6 of C19-A16-Steroids . . 11
Occurrence of 019-4 6-Steroids . . . . . 14
Cg-A1:-Steroids in the Pig . . . . 15
0119-4 6-Steroids1gn the Human . . . . 17
Biosynthesis of C19-A -Steroids in the Boar 19
Analysis of 54-Androst-16-enp3-one in Adipose
Tissue . . . . . . . . . . . . 23
Distillation Procedure . . . . . . 23
Chromatographic Technique . . . . . 25
Boar Odor Intensity and SdrAndrost-16-en23-one
Content . . . . . . . . . . . . 25
EXPERIMENTAL . . . . . . . . . . . . 28
Materials . . . . . . . . . . . . 28
Chemicals . . . . . . . . . . 28
Solvents . . . . . . . . . . 29
Chromatography Supplies . . . . . . 29
Fat Samples from Cryptorchid Pigs . . . . 30
Purification and Quantitative Determination
of SdbAndrost-16-en-3-one . . . . . . 31
Preparation of Fat Samples . . . . . 31
Saponification of Fat Samples . . . . 32
Treatment of Fat with Lithium Aluminum
Hydride . . . . . . . . . . 33
Urea Inclusion Compounds . . . . . 33

iii

Vacuum Distillation . . . . .
Thin Layer Chromatography . . .
Liquid Column Chromatography . .

Silica Gel H Column . . .

AgNOB-Impregnated Silica Gel Column

Gas Chromatography . . . . .

Gas Chromatography-Mass Spectrometry .

Synthesis of C19-A16-Steroids . . .
6,6-d2-5q-Androst-16-en—3-one . .
5,6,6-d3-5d-Androst-16-en-3-one .
5,16-Androstadien-2F-ol . . . .
17-d1-5d-Androst-16-enp3d-ol . .
16,17-d2-5,16-Androstadien-2fbol. .
16,17-d2-4,16-Androstadien—3-one .
6,6,16,17-d4—5drAndrost-16-en-3-one
6,6,16,17-d4-5d-Androst-16-en-3d-ol
6,6,16,17-d4-5d-Androst-16-en-3p-ol

RESULTS AND DISCUSSION . . . . . . .

Comparisons of Levels of 5d-Androst-16-en93-one

and Boar Odor Scores 0 e e e e 0

Procedure for the Assay of Sd-Androst-16-en-

3-One e o o e e o e e e 0

Possible Errors in the Assay Procedure
Purification by LiAlH4 Reduction and Urea

Clathrate Formation . . . .
Deuterium Labeled C19-A1G-Steroids . .
6,6-d2-5d-Androst-16-enr3-one . .
5,6,6-d3-5ubAndrost-16—en-3-one .

Perdeuterio 5d-Androst-16-en-3fbol and

5d-Androst-16-en-3u-ol . . .
16,17-d2-5,16-Androstadien-3P-ol .
16,17-d2-4,16-Androstadienp3-one .
6,6,16,17-d4-5d-Androst-16-en93-one

WY , . . . . . . e o e e
BIBLIOGRAPHY . . . . . . . . . .

iv

Page

34
36
37
38
38
39
39
41
41
43
46
47
4s
49
so
52
53
54

54

62
71

73
76
77
81

83
91
103
107
111
113

TABLE

LIST OF TABLES

Comparison of the Concentration of SNPAndrost-
16-en-3-one and Boar Odor Intensity Scores by
Employees of a Meat Processing Plant . . . . 57

Comparison of the Concentration of SN-Androst-
16-en-3-one and Boar Odor Scores by
Laboratory Panel . . . . . . . . . . 60

Assay of SdPAndrost-16-en-3-one . . . . . 7O

FIGURE

10

11

LIST OF FIGURES

Pathways of Biosynthesis of 019-416-Steroids
in boar testis in vitro. . . . . . . . .

Graph of 51-Androst-16-en-3-one Concentration
vs. Boar Odor Intensity Scores . . . . . .

Gas Chromatogram of Sd-Androst-16-en—3-one
Isolated with Deuterium Labeled Carrier . . .

Gas Chromatogram of SQPAndrost-16-en-3-one
Isolated without Carrier . . . . . . .

Abbreviated Mass Spectra of Protium and
Deuterium Forms of Soc-Androst-16-en-3-one . .

Molecular Ion Intensity Recordings for Protium
and Deuterium Forms of Sx-Androst-16-en—3-one .

Mass Spectra of Protium and Deuterium Forms
Of 5«-AndrOSt-16-en-3-One e e e e e e 0

Mass Spectra of Protium and Deuterium Forms ‘
of Sq-Androst-16-en-3d-ol . . . . . . .

Mass Spectra of Protium.and Deuterium Forms
of Sq-Androst-16-en-3p-ol . . . . . . .

Mass Spectra of Protium and Deuterium Forms
Of 5 ,1 G-Androatadien-3P-Ol e o e o e e 0

Mass Spectra of Protium and Deuterium Forms
of zp-hydroxy-S-androsten-17-one . . . . .

_vi

Page

20

58

65

66

67

68

80

85

87

95

97

FIGURE
12

13

14

LIST OF FIGURES (CON'T)

Page

Mass Spectra of Monodeuterated and Protium
Forms of quAndrost-16-en-3d-ol . . . . . 101

Mass Spectra of Deuterium and Protium Forms
of 4,16-Androstadien-3-one . . . . . . . 106

Mass Spectra of Tetradeuterated and Protium
Forms of Sd—Androst-16-en-3-one . . . . .110

vii

INTRODUCTION

The undesirable odor frequently associated with the
cooking of meat from uncastrated sexually mature male pigs
(boars) has long been a problem to swine producers and the
meat packing industry. This "perspiration-like" or "urine-
like" odor is extremely offensive to many consumers. Approx-
imately 65% of boar pigs have been reported to be effected,
while only 1-5% of the females and castrated males have boar
odor (Williams _e_t_ a_l_., 1963). Although only 0.001596 of all
hogs are condemned for boar odor under federal meat inspection,
it has been estimated that strict enforcement of the regula-
tions could result in condemnation or restricted usage of
1,350,000 hogs or about 675,000,000 pounds of pork annually
(National Provisioner, 1967).

Sink (1967) Postulated that the odor was due to 019-416-
steroids and that they may function as sexual pheromones in
pigs. Patterson (1968) isolated the C1 9-A16-steroid, 5C-
androst-16-en-3-one, by vacuum distillation of heated boar
fat and identified it as the major contributor to boar odor.
Subsequent corraborative reports from two laboratories (Berry
gt_§l., 1971; Thompson 23,2l., 1972) have confirmed the
involvement of C19-A16-steroids in boar odor as suggested

earlier by Sink (1967).

2

The aim of this investigation was to develOp an accurate
method for the assay of Sd-androst-16-en-3-one in pork fat,
and to detemine the extent to which the intensity of boar
odor is related to the concentration of Sat-androst-16-en-3-one
in the pig. A major portion of this study was devoted to the
synthesis of deuterium labeled C19-A16-steroids to be used as
internal standards and carriers in quantitative analysis by
mass spectrometric reverse isotope dilution.

The technique of reverse stable-isotope dilution seemed
uniquely well suited for assay of C19-A16-steroids in bear
odor studies, and for studies of the metabolism of this class
of compounds in the pig and in humans. The addition of a
labeled form of a C19-A16-steroid would enable extremely
small amounts of the compound to be partially purified and
separated by gas chromatography without total loss by adsorpe.
tion. Losses during the isolation would be corrected for
since the ratio of the labeled and unlabeled species would
not be expected to change. The amount of unlabeled steroid
could be determined from the isotopic ratio resulting after
the addition of a known amount of the labeled species.

REVIEW OF LITERATURE

Boar Odor in Pork

Meat from sexually mature boar (uncastrated male)
pigs gives off a permeating undesirable odor upon heating,
which is extremely offensive to many consumers. This odor
has been characterized as "perspiration-like”, "onion-
like" or "urine-like" and occurs not only in boars, but
also in cryptorchids and to a lesser extent in sows,
barrows and gilts (Pearson gt gl., 1969). Lerche (1936)
indicated that the meat of boars gives off an objectionable
odor that becomes apparent as soon as the male pig reaches
sexual maturity. Castration and holding for 57-68 days
resulted in disappearance of the odor (Lerche, 1936).

Incidencg of Boar Odor

Self (1957) reported that sex odor occurred as frequently
in sows, gilts and barrows as in boars. Prior to this report
other workers largely assumed that the problem was confined
to sexually mature boars and cryptorchids (Pearson g§,§;,,
1969). Pearson 33 El. (1969) reviewed the incidence of sex
odor and showed that there was a definite sex-dependency,

4

with 64% of all intact males exhibiting boar odor compared
to only 1-5% for the females or the castrate males. A
similar incidence of boar odor in pork carcasses was
reported by the Meat Inspection Division of the United
States Department of Agriculture (USDA, 1968). Their
survey indicated that 57% of the boars and 33% of the
cryptorchids examined gave off the objectionable odor,
while approximately 20% more in each group were found to
have a slight odor. Only 11% of the sows were found to
give off a definite sexual odor, with 10% having a slight
odor. The limited amount of data available on cryptorchids
suggests that about 50% have sex odor, which would make
the incidence somewhat lower than for boars (64%). but
considerably higher than the 1-5% reported for gilts, sows
and barrows by Pearson gt_gl. (1969).

The discrepancies between the results of Self (1957)
and those of other workers (Williams 23 gl., 1963; USDA,
1968) may have been due to differences in the methods used
for evaluating sex odor (Pearson.gtԤl., 1969). The testing
procedure used by Self (1957) consisted of heating a portion
of the diaphragm muscle to 200°F for evaluation by a panel
selected without regard for their ability to recognize
sex odor. The reliability of such a procedure is question-
able on considering the variability observed in the ability
of people to detect boar odor (Pearson gt‘al., 1969). It

is now known that a unique aspect of the sex odor problem

5

is the variable and sex-related ability of the human to
detect the odor of the C19-196-steroids, which are primarily
responsible for the offensive odor (Griffiths and Patterson,
1970). Furthermore, the use of diaphragm muscle creates
confusion because of possible contamination from the contents
of the urinary and digestive tracts (Pearson.§t‘§l., 1969).
Finally, Self (1957) heated the sample to only ZOOOF,
whereas, Craig gt al. (1962) subsequently showed that a
temperature of 2120 F was required for Optimum detection.

On occasion, terms such as boar odor and sex odor
have been applied to more than one off-odor associated
with pig fat (Self, 1957). Sex odor is volatilized on
heating or cooking of pork or bacon containing the undesirable
aroma, but can not be readily detected in the uncooked meat
or cold precooked meat (Pearson gtflgl., 1971). The term
"sex odor" is meant to describe only this odor and not

other odors associated with the live animal or its carcass.

Disposition of Carcasses with Sex Odor

Federal Meat Inspection Regulations require the
condemnation of carcasses that give off a "sexual" or
"urine odor" and specify that disposal shall be determined
by heating tests after chilling (USDA, 1973). The meat
from all boars and cryptorchids not condemned may be used

in comminuted, cooked meat food products or for rendering.

The seriousness of boar odor from the economic stand-
point was emphasized in a recent article (National Provisioner,
1967). It was estimated that strict enforcement of the
regulations on sex odor could result in condemnation of
about one-half of the 300,000 boars and stags killed
annually under federal meat inspection. In addition, 550,000
sows would be condemned and 500,000 more would be restricted
for use in cooked sausage or for rendering. This would
affect 1,350,000 hogs or about 675,000,000 pounds annually,
without taking into account additional losses in fresh pork
sausage or in Canadian style bacon (Pearson, 1972).

Indications are that only 0.0015% of all hogs are
condemned for sex odor under federal inspection (National
Provisioner, 1967). Efforts to enforce the regulations are
frustrated by the varying levels of odor, the inability of
many inspectors to consistently recognize the odor and the
difficult task of heat testing every carcass (Pearson, 1972).
Condemnation of all carcasses having boar odor would result
in a serious economic loss, nevertheless, the pork industry
cannot afford to market any meat or lard which gives off
boar odor on heating.

Results of a study by Pearson 23 El. (1971) verified
that boar meat can be used in some meat products eaten
without further heating. Two large consumer taste panels

were utilized to ascertain the acceptability of 22 different

products differing only in that the pork utilized in them

7

did not contain boar pork. In general, the results indicated
that products containing boar meat are not readily distinguish-
able from similar control products providing the kitchen and
dining areas are kept distinctly separate, for it is during

the cooking process that the objectionable odor becomes

most apparent (Pearson gt gl., 1971).

Chemical Identity of Boar Odor

One of the earliest investigations into the identity
of the boar odor component(s) was a study by Craig|g§,§l.
(1962), which established that the undesirable odor was
localized in the fatty tissues of the carcass and was
concentrated in the nonsaponifiable fraction of the fat.
Subsequent efforts were made to identify the responsible
components by fractionation of the nonsaponifiable material
and gas chromatography, but these studies did not reveal
the identity of the responsible components (Williams and
Pearson, 1965)-

Sink (1967) published a theoretical paper in which
he proposed that sex odor was caused by C‘g-A1 6-s1:eroids
and that these compounds were serving as sexual pheromones
in the chemical communication between pigs. The theory was
based in part on comments made much earlier by Prelog gt‘gl.
(1944) about the musk-like odor of two C19-A16-steroids,
deandrost-16-en-3dpol and 5d-androst-16-en-3p-ol, both
of which were isolated for the first time from swine

8

testicular tissue (Prelog and Ruzicka, 1944). Prelog §t_§l.
(1945) reported that the corresponding Sp-isomers were
odorless and that the ketones, 4,16-androstadien-3-one and
SQ-androst-16-en-3-one, possessed pronounced urine-like or
perspiration-like odors. These workers also drew attention
to the superficial structural similarities between 5d-androst-
16-en-3-one and the structure of civetone and muscone. These
two macrocyclic compounds were isolated by Ruzicka (1926a,b,c)
and shown to have pheromonal significance in the sexual
behavior of the civet cat and the musk deer, respectively.

The similarity in the structures of civetone, muscone
and Supandrost-16-en-3-one as pointed out by Prelog and

Ruzicka (1944) are shown below:

0’ -
H
SqrAndrost-16-en-3-one

Civetone Muscone

9

Shortly after Sink (1967) published his theoretical
paper on the possible relationship of boar odor to the
C19-zg6-steroids, Patterson (1968) reported the isolation
and identification of 5a-androst-16-en-3-one in volatiles
stripped from boar fat under high vacuum and at elevated
temperatures (800 C). Gas chromatography of the complex
mixture revealed many components, with the "taint" compound
producing a very small response among the very last materials
to be eluted. Patterson (1968) was able to identify the
appropriate region of the chromatogram.by smelling the
column effluent after extinquishing the flame of the gas
chromatographic detector. Collection of the effluent
after repeated injections yielded enough material for mass
spectrometric analysis. The mass spectrum.suggested a
molecular weight of 272, and an empirical formula of C19H280.

Patterson (1968) then considered the ketosteroids of
the androstane series, which have been described in the
literature as possessing urine-like odors (Radt, 1959).

The Steroid Reference Collection of the Medical Research
Council of England was made available to Patterson (1968)

so that he could perform olfactory tests of many steroids,
most of which were not commercially available. Olfactory
examination of a number of monoketoandrostanes and mono-
ketoandrostenes revealed that several possessed a charact-
eristic odor, which was identical in quality to the odor

of the material isolated from boar fat. The strongest odors

10

were associated with steroids containing a keto group at
position 3 on the steroid nucleus and a hydrogen atom on
position 5, if present, in theci-orientation (Patterson,
1968). Chemical data and the "intense, urine-like odor"
of Surandrost-16-en—3-one as described by Radt (1959)
prompted Patterson (1968) to further examine this compound.
The compound was not available to him either commercially
or from the Steroid Reference Collection and no published
mass spectrum was available (Patterson, 1968). He made
the steroid by oxidation of the corresponding alcohol,
5d-androst-16-en-3p-ol, which was available from the
Medical Research Council Steroid Reference Collection.

He found that the odor, gas chromatographic retention time
and mass spectrum of 5X-androst-16-en—3-one agreed with
that of the "taint" compound isolated from boar fat.

The findings of Patterson (1968) plus corroborative
results by Beery and Sink (1971) and by Thompson gt‘gl.
(1972) confirmed the involvement of 019-8 6--steroids in
boar odor as suggested by Sink (1967). Isolation of’5d-
androst-16-en-3-one from the nonsaponifiable material of
boar fat (Beery and Sink, 1971; Thompson 23 gl., 1972)
supported the earlier studies by Craig and Pearson (1959),
Craig 23,9l. (1962) and Williams and Pearson (1965) showing
that sex odor was associated with the fatty tissues, and

more specifically with the nonsaponifiable material.

11

Odor Characteristics of C19-zg6-Steroids

54randrost-16-en-3-one has a pronounced perspirative,
urine-like odor, which is characteristic of a number of
androstane steroids of similar structure (Patterson, 1968).
The olfactory properties of the C19-lg6-steroids have been
referred to in a number of papers (Kloek, 1961: Comfort,
1971a; 1971b; Beets, 1971 ) and have been reviewed recently
by Katkov (1971).

The characteristic odors of this group of compounds
have been useful in detecting the presence of very small
quantities in tissues, plasma or urine extracts (Brooksbank
and Haslewood, 1950; 1952). For example, 5d-androst-16-en-
3-one was detected in boar fat extracts as it emerged from
a gas chromatographic column (Patterson, 1968). Brooksbank
and Haslewood (1950; 1952) were able to smell the musk-like
odor of a compound (later identified as Sapandrost-16-en-
3x-ol) in hydrolyzed urine extracts. Gower gtflgl. (1970)
described how the odor of’54-androst-16-enp3-one could be
detected on the hot syringe needle after injecting the
plasma extract from the spermatic vein onto a gas chromato-
graphic column.

The characteristic musk odor is not restricted to

16-steroids. The 3-hydroxy-51-androstanes also possess

019"
a musk-like odor (Prelog‘gt‘§;., 1945). although the 5p-
epiners are odorless. 0f thirty-three steroids investigated

by Beets (1962), the musk odor was detected in thirteen

12

derivatives of androstane.

Many ketosteroids of the androstane series are described
in the literature as possessing urine-like odors (Radt, 1959).
These compounds include 5u-androstan-3—one, Sarandrost-Z-en-
17-one, Supandrost-3-en-17-one, 5d-androst-16-en-3-one and
5p-androst-16-en-3-one. The last compound was reported as
having a distinct, but weaker urine-like odor than that of
the Sarepimer, which Radt (1959) has described as being
intense.

Patterson (1968) reported that the odor of Su-androstan-
3-one was strong while that of the Sp-epimer was weak. He
also stated that 4-androsten-3-one and 4,16-androstadien-
3-one possessed strong odors, while 3,5-androstadien—17-one
possessed a weak odor. He observed that some monoketoandro-
stanes possessed no odor. These included Susandrostan~17-one,
Surandrost-Z-en-17-one and Supandrostan-11-one.

A unique aspect of sex odor is the variation in and
possible sex-related ability to perceive this odor (Comfort,
1971a; Sink, 1973). To some consumers the sex odor is accept-
able or even pleasant, while to others the slightest hint of
this odor is enough to cause serious objection to or even
complete rejection of food products containing the undesir-
able aroma (Sink, 1973). Several researchers (Patterson, 1968;
Pearson.g§flgl., 1969; Sink, 1973) have noted that some people
are consistently able to perceive small amounts of the odor,

while others cannot detect it.

13

Griffiths and Patterson (1970) noted a sex-related
difference in the ability to detect 5drandrost-16-en-3-one.
Analysis of olfactory responses of fifty men and fifty women
to a pure sample of Sakandrost-16-en-3-one showed that 92%
of the women, but only 46% of the men could smell this steroid.
For the men and women able to detect the odor, there was no
significant difference between sexes in the individual
threshold values. However, women found the odor significantly
more unpleasant than did men, with most of the women subjects
rating the odor as extremely unpleasant. On a scale of 1 to
9 (extremely pleasant to extremely unpleasant) the average
score for women was 7.26 compared with 6.22 for men. This
has important practical implications since women, rather
than men, are most often involved in the preparation and
cooking of pork or bacon and make decisions as to whether
or not a product is acceptable.

A study by Klock (1961) of 100 men and 100 women revealed
that 29% of the men could not smell 3d-hydroxy-Sxpandrost-16-
ene, 38% described the odor as faint and 33%1as strong. The
corresponding percentages for women were 22% (no odor), 36%
(faint odor) and 42% (strong odor).

Griffiths and Patterson (1970) recorded threshold values
for 51kandrost-16-en—3-cne in 30 individuals selected for
their ability to smell the compound. Threshold values were
determined for the dry residue of the pure compound that had

2

been applied in ether solution to a 5 cm area on a watch-glass.

The values extended over a 2000-fold range of concentration

14

from 0.049 to 100.0 ng. These threshold values refer to the
concentration on the watch-glass and not the concentration

in the air, which obviously is very much lower. Griffiths
and Patterson (1970) found for both sexes that the most
sensitive subjects reported the odor to be more objectionable.
On smelling the same concentrations of Surandrost-16-en—3-one,
women found the odor more unpleasant than did men having

identical odor thresholds.

Occurrence of 019-[g6-Steroids

The existence of 16-dehydro-C19 steroids in the pig
(Prelog and Ruzicka, 1944) and in man (Brooksbank and
Haslewood, 1950) has been recognized for many years, but
very little is known of their biological significance
(Brooksbank gt,§l., 1972). The occurrence of Ctg-(g6-steroids
has been reviewed by Gower (1972) and more current information
can be found in papers by Brophy and Gower (1972), Lake and
Gower (1972), Saat 23,9l. (1972), Brooksbank gt 9;. (1972)
and Gustafsson (1973). Comparative aspects of steroid
metabolism in human and boar testes tissue was the subject of
recent papers by Ruokonen (1973) and Ruokonen and Vihko
(1974a,b).

15

16 .
019:1; -bteroids in the Pig

 

The first report of C19-A16-steroids in the pig was by
Prelog and Ruzicka (1944), who isolated relatively large
amounts of Sipandrost-16-en-3d—ol and 54-androst-16-en—3p-ol
'from boar testes. These same workers and subsequent investi—
gators (Gower, 1972) noted that large quantities of the 019-[Q6-
steroids occurred in pig testes, while the amount of androgenic
steroids was comparatively much lower. Recently, Booth (1972),
Claus 23,9l. (1971) and Ruokonen and Vihko (1974a) have shown
conclusively that the C19-zg6-steroids are quantitatively
more important than androgens-in the pig. In fact, Ahmed
and Gower (1968) demonstrated that the metabolism of pregnenolone
and progesterone is biased in favor of the formation of 019-zfi5-
steroids rather than androgens.

Claus gl‘al. (1971) have shown that 54+androst-16-en-3-one
and testosterone increase with age, both in testis tissue and
in peripheral blood plasma of boars. In testes, the concen-
tration of 51+androst-16-en-3-one was ten times higher than
that of testosterone. The level of 54kandrost-16-en-3-one
varied from 100 to 310 ng/g in comparison to values for
testosterone of 6 to 37 ng/g testes (Claus gtflal., 1971).
In peripheral blood plasma, the concentration of Sdpandrost-
16-en-3-one and testosterone was nearly the same, and increased
directly with age in the boar from 6 to 22 ng/ml plasma
(Claus ggmal., 1971). Plasma values in females (0.8 to 2.7
Jug/ml) and in castrated males (1.3 to 2.7 ng/ml) were 9.5-

16

and 8-fold lower, respectively, than corresponding values
of 6.0 to 22.3 ng/ml for boars (Claus gt_§l., 1971).

Claus 23 al. (1971) found relatively high concentrations
of 5d+androst~16~en-3-one in boar fatty tissue (1.03 to 7.49
Pg/g) and in boar parotid salivary gland (0.17 to 11.43 Pg/g).
Surandrost-16—en-3-one was not found in the fatty tissue or
the parotid gland of females or castrate males, in contrast
to the situation in peripheral blood where appreciable
amounts of this compound were detected (Claus gtial., 1971).

Ruokonen and Vihko (1974a,b) have also observed that
019-1g6-steroids occupy a quantitatively important position
in boar testis tissue. The principle compounds in the uncon-
jugated fraction were 5w—androst-16-enp3u-ol and 5xpandrost-
16-enp3p-ol (Ruokonen and Vihko, 1974a). In the monosulfate
fraction of boar testicular tissue, Ruokonen and Vihko (1974a)
found 5,16-androstadienp3p-ol was also one of the most
abundant steroids along with Sd-androst-16-en-3u-ol and
SdPandrost-16-en-3p-ol. No disulfate conjugates were
detected despite a careful search with a level of detection
of 5 ng/g testis tissue. Ruokonen and Vihko (1974a) did not
detect 5x—androst-16-en-3-one or testosterone in the testicular
tissue of the boar with a level of detection of 5 ng/g. In
contrast to this, Claus gt 5;. (1971) reported that the
concentration of’Sdhandrost-16-enp3-one in boar testis ranged
from 100 to 310 ng/g, while the concentration of testosterone
was 6 to 37 ng/g.

17

Analysis of boar urine (Gower and Katkov, 1969; Gower
23 al., 1970) has revealed the presence of 5d-androst-16-
en-3p-ol, conjugated as glucosiduronates, at a level of
approximately 2507P3/1° This contrasts with the situation
in human urine where the predominant 019-1g6-steroid is the
corresponding 31901 isomer present as glucosiduronates
(Brooksbank and Haslewood, 1950; 1952). The 34901 isomer
possesses an intense odor, which is characteristic of human
urine (Kloek, 1961) and explains the choice of the term
"urine-like" frequently used to describe boar odor in pork.
The term "perspiration-like" also appears to be an appro-
priate description for boar odor since it is now known
that human sweat also contains Sarandrost-16-en-3-one, the
steroid principally responsible for boar odor in pigs
(Gower, 1972)-

C19-A16-Steroids in the Human

 

Brooksbank (1962) found that relatively large quantities
of the odorous compound Susandrost-16-en-3K-ol are excreted
in human urine, with excretion values of approximately
1 mg/day for men and 0.4 mg/day for women. Urinary 5a-
androst-16-en-3u-ol excretion is very small in infants and
children but increases at puberty to a maximum in the young
adult. Thereafter, excretion decreases to lower values in

old age (Cleveland and Savard, 1964; Brooksbank, 1962).

18

Brooksbank gt gl. (1969) isolated 4,16-androstadien-
3-one from normal human blood plasma, and subsequently
showed that its concentration was higher in the male than
in the female (Brooksbank gt al., 1972). The only other
paper located which deals with C19-A6-steroids in human
plasma was by Gower and Stern (1969), who reported the
presence of 5d-androst-16-en-3d-ol in the peripheral plasma
of a woman with a virilizing adrenocortical carcinoma.
After adrenalectomy, the 5d-androst—16-en-3d-ol was no longer
detectable. Although clearly related to the endocrine
function of the testis, no specific function has yet been
attributed to C19-zg6-steroids in man (Brooksbank gt,al., 1972).

There is some evidence that the [gs-androstenes are
precursors for 3,16,17-trihydroxy-androstanes. In this
connection, Brooksbank.g£,gl. (1972) found four different
isomers of 3,16,17-trihydroxy-androstanes in human urine
after administration of radioactive 4,16-androstadien-3-one
to a normal male and a normal female subject. Gustafsson
(1973) showed that 4,16-androstadien—3-one, Sd-androst-16-en-
34-01, 5drandrost-16-en-3f—ol and 5,16-androstadien-3p-ol
were metabolized to 3,16f,17d-trihydroxylated steroids in

liver microsomal preparations.

19

Biosynthesis of 019-[16-Steroids in the Boar

Boar testis tissue has been shown to convert
pregnenolone and progesterone into 019-1g6-steroids
in 13339 (Gower and Ahmad, 1967; Ahmad and Gower, 1968;
Katkov and Gower, 1968; 1970). The conversion of
pregnenolone and progesterone to androgen is small by
comparison (Ahmad and Gower, 1968).

Katkov and Gower (1970) have shown that 5,16-androsta-
dien-3P—ol is produced from pregnenolone by boar testis
in zitgg, and is then converted into 4,16-androstadien-
3-one. 4,16-androstadien-3-one can be converted to the
A-ring saturated alcohols (SR-androst-16-en-3d-ol and
54-androst-16-en-zp-ol) by way of Sx-androst-16-en-3-one
as shown by Gower (1972) in Figure 1. 4,16-androstadien-
3-one can also be formed directly from progesterone ig,zit;g,
furnishing an alternative pathway for formation of 019-116-
steroids as shown in Figure 1. The pathway by way of 5,16-
androstadien-3F-ol has been reported by Saat 23.3}, (1972)
to be the predominant one in boar testes. Bicknell and Gower
(1971) have reported similar results from experiments with

human testes.

 

Hounwssonmausoosossitm guns—v noses sons mouse .oaees mm masses soon on
3833134130 no mesoneshmoaa. no 9335.8 .. 0.3m:

m on
— osoaoauomam

ono1n1soaesao 0.5513. c
osoInuso1m Tumoaegiom
m
.. so eoL\
T
.8 “NO

one Hososwohm

om
”surnusoum_1sooses<LVm Hounwusoaooseonessuo. m

AI:

m
.m tom Q .
. . Ono

21

A number of steroids have been tested as intermediates
in the conversion of pregnenolone and progesterone to 019-1(5-
steroids, but none of them have been shown to be effective
(Gower, 1972). The compounds tested included the following:
testosterone, 3P-hydroxy-5—androstene-17-one, 4,16-pregnadiene-
3,20-dione and 17x-hydroxy-4-pregnene-3,20-dione by Ahmad and
Gower (1968); epitestosterone and 3P,17w-dihydroxy-5—pregnene-
20-one by Gower and Ahmad (1967); 16q-hydroxy-4-pregnene-3,20—
dione by Matsui and Fukushima (1970); and 5-androstene-3B-ol
by Katkov and Gower (1970).

The 17qrhydroxyl-derivatives of pregnenolone and progesterone
are generally recognized as intermediates in the formation of
androgens in the testes (Eik-Nes, 1970). However, 17d~hydroxy~
pregnenolone and 1749hydroxy-progesterone are not converted
to 019-416-steroids in boar testes, suggesting that the C19—A16-
steroids are produced from pregnenolone and progesterone by
a unique pathway not requiring 17d-hydroxylation before side-
chain cleavage (Gower, 1972)

Loke and Gower (1972) considered the possibility that
other dihydroxylated C21-steroids might be intermediates in
C1g-lg6-steroid biosynthesis. They reported that radioactive
5-pregnene-3p,20P-diol was formed during the biosynthesis of
5,16-androstadiene-3f-ol in homogenates of boar testes.
However, they could not exclude the possibility of the con-
version of 5-pregnene-3f,20p-diol to pregnenolone before
the formation of 5,16-androstadiene-3P-ol. Although Ruokonen
and Vihko (1974a) did not detect 5-pregnene-3r,20P-diol in

22

boar testes, they did detect its monosulfated metabolite,
Sappregnane-3p,20p-diol.

The enzyme system involved in the conversion of preg-
nenolone to 5,16-androstadiene-3f-ol in boar testes has
been named "andien-p synthetase" by Lake and Gower (1971).
Katkov and Gower (1970) have shown that the enzyme system
requires NADPH and 02 for activity. Gower and Lake (1971)
studied the enzyme system in homogenates and subcellular
fractions of boar testes and showed the activity to be
mostly in the microsomal fraction.

Katkov .e_t_ gl. (1972) have shown that Sat-androst-16-en-
3K-ol and Sd-androst-16-en-3P-ol are formed by reduction of
Sq-androst-16-en-3-one in boar submaxillary glands. However,
they found no evidence for the conversion of 021-steroids
into C19-A16-steroids in the submaxillary glands. Therefore,
it seems likely that the C19-A16-steroids are formed in the
testes and transported to the salivary glands where reduction
of the ketone occurs. This could explain the occurrence of
5°handrost-16-en-3-one and 54-androst-16-en-3-t-ol in boar
saliva (Gower, 1972).

The adrenal cortex in the pig has a limited capacity for
019-416-steroid biosynthesis (Gower, 1972), which may explain
the occurrence of boar odor in sows, barrows and gilts. It
has been shown that 5w-androst-16-en—3ct-ol, Sq-androst-16-en-
3P-ol and 4,16-androstadiene-3-one are formed from pregnenolone
and from progesterone in the adrenal cortex 1! mm, but only
in very small quantities compared with the testes (Gower and
Ahmad, 1967; Ahmad and Gower, 1968).

23

Analysis of 5Q-Androst-16-en-3-one

in Adipose Tissue

Two techniques for determination of 54-androst-16-en-
3-one in adipose tissue have been published. The first
technique, which was developed by Patterson (1969) and
later used by Fuchs (1972) and Stinson (1972), utilizes
vacuum distillation followed by saponification, and sepa-
ration by thin layer chromatography and gas chromatography.
The second technique, which was developed by Claus £3,3l.
(1971), involves the addition of a radioactive internal
standard (54-3H-androst-16-en-3-one) to the sample followed
by two separate column chromatographic steps, followed by
thin layer and gas chromatography.

Digtillation Procedure

Patterson (1969) used kidney fat instead of subcutaneous
fat, since kidney fat contains less connective tissue and
can be obtained without damaging the carcass. In this
connection, Patterson (1969) reported that 54-androst-16-en-
3-one is present in similar concentrations in both kidney
fat and subcutaneous fat, but gave no values to substantiate
this claim.

Patterson (1969) used samples of kidney fat (150 g),

which were chilled, cut into small pieces, minced and then

24

melted at 55°C. The liquified fat was filtered through
butter muslin using a warm Bachner filter funnel connected
to an aspirator. The fat was then dried with anhydrous
sodium sulfate and filtered through Whatman No. 1 filter
paper under vacuum to give a clear oil.

For vacuum distillation, Patterson (1969) poured a 44 5
sample of the filtered fat into the lower section of a
"pot-still". The upper half of the "pot-still" consisted
of a cold finger containing liquid nitrogen. When the two
sections were joined together with an O—ring they formed
an efficient, short-path, molecular still in which the
distance from the fat surface to the condensing surface
was only a few centimeters. The fat was distilled with
stirring for 6 hours at 1 mm vacuum, while the base of

the still was immersed in water at 80°C.

Patterson (1969) removed the distillate from the surface
of the cold finger by rinsing with small volumes of ethyl
ether. These solutions were then combined and concentrated
lg m. The residue was saponified for 15 minutes with
0.1 N aqueous NaOH at room temperature. Then the solution
was poured into 90 ml of distilled water and the nonsapon-
ifiable material was extracted with two-30 ml volumes of
diethyl ether-hexane (3:17). The combined extracts were
dried with anhydrous sodium sulfate and concentrated to
250 ul. A 50 ul aliquot was purified by thin layer chroma-

tography and the appropriate zone was recovered for quantitative

25

analysis of 54+androst-16-en-3-one by gas chromatography.
Patterson (1969) performed gas chromatographic analysis by
comparing peak heights obtained with 10 ul injections of the
fraction containing 5drandrost-16-en-3-one with standards for
this compound. Internal standards were not used during
either the isolation procedure or the gas chromatographic
analysis. The efficiency of the overall procedure was
determined by checking recoveries of known quantities of
fix-androst-16-en-3-one added to fat, which had been previously

distilled to remove this compound.

Chromatoggaphic Technique

Claus 23 El- (1971) used a chromatographic method in
which fatty tissue (5 g) of an unspecified type was cut into
small pieces. After addition of 0.148 PC of tritiated 5.1.-
androst-16-en—3-one as an internal standard, the sample was
homogenized in 24 ml 0.1 N NaOH. The homogenate was
increased to a total volume of 50 ml by the addition of
another 25 ml portion of 0.1 N NaOH. It was then extracted
with 250 ml of dichloromethane-ethyl.acetate (1:1) in a
500 ml separatory funnel. The extract was washed with 50 ml
of distilled water, dried over anhydrous sodium sulfate and
evaporated to dryness lg Egggg,

Claus 33,gl. (1971) dissolved the residue in 2 ml
cyclohexane and applied it to the first silica gel column
(1.5 on x30 cm), after which it was eluted with cyclohexane-

26

ethyl acetate (98:2). The fractions containing radioactivity
were pooled, dried and applied to the second silica gel
column, which differed from the first column only by being
smaller in size (1 cmx28 cm). The radioactive fractions
were pooled, dried and further purified by chromatography

on silica gel thin layer plates developed in benzene-ethyl
acetate (99:1). The thin layer zone, which contained the
radioactivity, was removed and eluted with 5 ml of ethyl
acetate. An aliquot was withdrawn for calculation of the
recovery by counting its radioactivity. The remainder of

the sample was used for analysis by gas chromatography

using 4-androstene-3,17-dione as a chromatographic internal
standard. Recovery of the radioactive 5a-androst-16éen-3-one
was reported to be 50-60% when measured just prior to the

gas chromatographic analysis.

Boar Odor Intensity and
SdPAndrost-16-en-3-one Content

Fuchs (1972) and Newell g;,gl. (1973) reported positive
correlations between the concentration of 5d-androst-16-enp
3-one in pork fat and the odor scores determined by a panel
of experts, selected for their ability to smell Surandrost-
16-en-3-one and trained to distinguish its odor from other
odors associated with heated pork fat. Both Fuchs (1972)
and Newell g§_al. (1973) used the vacuum distillation-gas

27

chromatographic technique of Patterson (1969) to determine
the concentration of 54eandrost-16-en-3-one. Newell g; gl,
(1973) reported that the correlation coefficient between
odor intensity and the quantity of 5(-androst-16-en-3-one
was 0.53. Fuchs (1972) evaluated 19 samples and reported
a correlation coefficient of 0.75.

Fuchs (1972) excluded the three samples with the highest
concentration of’Sarandrost-16-en-3-one (5.3, 5.8 and 8.0 pg/g
fat) from his regression calculations. He reported that the
odor intensity reached a maximum value at about 2-4 Pg/g
fat, after which there was no incremental increase in odor
with increasing concentrations of Susandrost-16-en—3-one.

The concentration of 5i-androst-16-en-3-one ranged from
0.3 to 8.0 ’lg/g fat.

Newell _e_1; El. (1973) showed that odor scores from an
untrained panel did not correlate significantly with the
quantity of 5dpandrost-16-en-3-one in the fat. Members of
the untrained panel were able to detect the odor of 5(-
androst-16-enp3-one, but they were not trained to specif-
ically distinguish this odor from other odors that are

apparent when pork fat is heated.

EXPERIMENTAL

Materials

Chemicals

The common reagents were reagent grade. Special

reagents are listed below:

Aluminum isopropoxide,
electronic grade

Chloroiridic acid (H21r016-6H20) Ventron Corporation
Alfa Products
Lithium wire, packed P.0.Box 159
under mineral 011 Beverly, Mass.

Methyllithium, 2M in ether

Lithium aluminum hydride Chemicals Procurement
Laboratories, Inc.
p-toluenesulfonylhydrazine 18-17 130th Street

College Point, N.Y.

Ammonia-d3

Merck and Company, Inc.
Deuterium oxide, D20 Isotopes

111 Central Avenue
Lithium aluminum deuteride Teterboro, N.J.
Methyl alcohol-d
Trimet 1 has hite 97% Aldrich Chemical Co.

by p p ' 940 West St. Paul Ave.

3P-Hydroxy-5-androsten-17-one Milwaukee. Wisconsin

28

29

3w-Hydroxy-5x-androstan-17-one

3f-Hydroxy-5u-androstan-17-one

Sw-Androst-16-en-3-one 104 mg
51-Androst-16-en-3u-ol 22 mg
5d-Androst-16-en—3p-ol 50 mg
4,16-Androstadien-3-one 100 mg
5,16—Androstadien-3F-ol 50 mg

Solvents

General solvents

Benzene, acetone, cyclohexane,

ethyl acetate, ethyl ether, hexane
methanol and ethanol - "Distilled in
Glass"

Chr t Su lies

Silica Gel H

Thin layer plates, Type PLOP
"pre-absorption silica gel,
1 mm thickness, 20320 cm

Activated Alumina, F-20

Gas chromatography packing materials
3% and 5% OV-1 on Supelcoport,
100/120 mesh

5% SP-2401 on Supelcoport,
1oo/12o mesh

Sigma Chemical Co.
P.O.Box 14508
St. Louis, Mo.

Gifts from The Upjohn Co.
Kalamazoo, Michigan

All solvents were
redistilled before use

Burdick and Jackson
Laboratories, Inc.
Muskegon, Michigan

Brinkmann Instruments, Inc.
EM Reagents Division
Westbury, N.Y.

Quantum Industries

341 Kaplan Drive
Pairfield, N.J.

Alcoa Chemicals

Supelco Incorporated
Bellefonte, Pa.

30

Fat Samples from Cryptorchid Pigs

The objective of this project was to determine the
concentration of 5¢+androst-16-en-3-one in back fat from
cryptorchid pigs that had previously been evaluated for
boar odor by employees of two meat packing companies.

Each company was to supply 40 back fat samples (5 lbs each-
free of skin) that had been evaluated for boar odor by
procedures used under normal operating conditions. The
fat was removed from the carcasses of cryptorchid pigs in
the cooler and evaluated in the quality control laboratory
by smelling the odors produced upon touching the fat with
a hot iron. The samples were frozen, packaged in dry ice
and shipped to East Lansing by air freight. Upon arrival
in East Lansing the samples were coded with numbers taken
from a table of five-digit random numbers and stored in a
freezer at -20°C.

The samples were supposed to include a representive
number in each of the following categories: no boar odor,
slight boar odor, strong boar odor and pronounced boar odor.
Although packer A provided a total of 40 samples, only 23
were evaluated according to the above system. The 23 samples
were classified as: none-3, slight—8 and pronounced-12.
The other 17 samples were rated by other descriptive terms
such as very slight-3, slightly less than pronounced-1,
slightly pronounced-7, moderate-4, moderately intense-1 and

31

and very pronounced-1. Samples from packer B arrived
nine months late and included only samples described as
having no odor-17 and slight odor-25. No strong or pro-
nounced samples were sent because they reported that none

were found.

Purification and Quantitative Determination
of 5K-Androst-16-en-3-one in Fat Samples

The level of 51+androst-16-en—3-one in samples of

fat was determined by a reverse isotope dilution/carrier
technique (Sweeley 23 gl., 1966; Samuelsson.gp,pl., 1970;
Bieber ppflgl., 1972; Gordon and Frigerio, 1972). A known
amount of stable isotope labeled steroid (usually 21 P8 of
6,6-d2-5dbandrost-16-en—3-one) was added to the sample to
serve as a carrier and as an internal standard. The mixture
of protium.and deuterium form of 54-androst-16-en-3-one was
isolated and the ratio of protium form.and deuterium form
was found by combined gas chromatography-mass spectrometry.
The amount of 5K-androst-16-en-3-one in the original mixture
was calculated from the measured ratio and the known amount

of deuteriated material added.

Preparation of Fat Samples

Fat samples were prepared as described by Patterson

(1968,1969). A 150 g portion of fat was homogenized in a

32

Waring blender and the liquified fat was then warmed to

55°C. The warm fat was filtered through four layers of

cheese cloth using a warm Bdchner filter funnel connected

to an aspirator. The liquified fat was dried by mixing

with anhydrous sodium sulfate for 30 minutes and then
filtering through Whatman #1 filter paper on a warm Bdchner
filter funnel connected to an aspirator. The drying procedure

was repeated for each sample until a clear oil was obtained.

Saponification of Fat Samples

Fat was saponified according to a procedure recommended
for assay of vitamin E (Bunnel, 1967). A 2 g fat sample
was transferred to a 100 ml Erlenmeyer flask containing
15 ml ethanol, 100 mg pyrogallol and a few boiling chips.
The flask was connected to a reflux condenser and the
solvent refluxed for 5 minutes to displace air from the
flask. Approximately 1 g of potassium hydroxide pellets
was added through the reflux condensor and the refluxing
continued for 30 minutes.

The flask was cooled and the contents transferred to
a separatory funnel which contained 15 ml of hexane and
15 ml of water. The saponification mixture was shaken for
1 minute. The extraction was repeated twice with 8 ml of
hexane. The combined extracts were washed 5 times with
water followed by drying over anhydrous sodium sulfate.
The hexane extract was evaporated to dryness and the residue

was redissolved in 250 ul of ethyl acetate.

33
Treatment of Fat with Lithium Aluminum dri e

De Luca (1974) recommended treatment with lithium
aluminum hydride (LiAlH4) as an alternative to saponifi-
cation for fractionation of pork fat samples. The sample
(usually 2 g, prepared as described above) was dissolved
in 40 ml of tetrahydrofuran in a 50 ml Erlenmeyer flask
fitted with a ground glass stopper. The mixture was dried
over anhydrous sodium sulfate for 4 hours, and then the
sodium sulfate was removed by filtration. The solution was
transferred to a 50 ml conical shaped centrifuge tube.
Approximately 0.32 g of LiAlH4 was added slowly, and the
reaction mixture was held at room temperature for 72 hours
with occasional stirring.

The excess LiAlH4 was removed by centrifugation and
decanting. The supernatant was transferred to a 250 ml
separatory funnel and the excess LiAlH4 was destroyed by
dropwise addition of ethanol. The solution was washed 3
times with 60 ml of 0.1N’HCl and.3 times with 60 ml of
water. After drying over sodium sulfate, the solution

was evaporated to dryness and the residue was weighed.

W

The residue (usually 1.7 8) obtained after LiAlH4
treatment of fat samples was dissolved in methanol (30 ml/g)
previously saturated with urea at room temperature (Mareno and

Roncero, 1964). After 2 hours the crystals were separated

34

from the solution by filtration. The filtrate was transferred
to a 250 ml separatory funnel containing 100 ml of water and
the solution was extracted 3 times with 60 ml of ethyl ether.
The ether extracts were combined and washed 3 times with 60
ml of water. After sodium sulfate drying, the solution was
evaporated to dryness and the residue was weighed.

The included compounds were freed by treatment with
water. The crystals were dissolved in 150 ml of water and
the solution was extracted 3 times with 60 ml of ether. The
combined extracts were washed 3 times with 60 ml of water,

dried over sodium sulfate and evaporated to dryness.

Vgggg! Distillatiog

The analysis of Sukandrost-16-en-3-one involving
vacuum distillation was performed essentially as described
by Patterson (1969). For vacuum distillation, a 55 g
sample of filtered and dried fat was poured into the bottom
of a high capacity vacuum sublimator (Ke855700, Kontes
Glass Company). A Teflon-coated stirring bar was added
to the container. The flask and condenser were sealed
without grease using an "0”-ring, and the assembled unit
was connected to a vacuum manifold by a ball and socket
joint. The base of the sublimator was submerged in a
glycerol bath that was resting on a magnetic stirrer.

The temperature of the glycerol bath was controlled within
narrow limits by a circulating heater (HAAKE, Model FJ).

35

It was determined that a bath temperature of 87°C gave
the desired 80°C temperature in the fat during vacuum
distillation.

Liquid nitrogen was added to the condenser section
of the sublimator and also to another trap interposed
between the manifold and the mechanical vacuum pump
(Cenco, HYVAC 14). The stopcock on the sublimator was
cpened and the sublimator-manifold unit was evacuated very
slowly by gradually opening a small precision vacuum valve
(K—82505, Kontes Glass Company). The speed of stirring
was carefully controlled to avoid splattering of the fat
during the initial phase of evacuation, after which the
speed was adjusted so that the stirring bar revolved at
approximately one revolution per second. A large bore
vacuum valve was opened as soon as it was apparent that
no splattering would occur and the system was thoroughly
degassed.

The pumping system included an oil diffusion pump
(Cenco, #93423), which was connected to the system as soon-
as the pressure dropped to 1 mm Hg. The mechanical pump
was used to produce the required vacuum at the outlet of
the oil diffusion pump. The distillation was continued
for 6 hours after the pressure first drapped to 10-3mm Hg.
The glycerin bath was maintained at 87°C. The liquid nitrogen
traps were resupplied with refrigerant throughout the 6 hour
distillation period.

36

At the conclusion of the distillation period the
stapcock on the sublimator was closed, and the sublimator,
while still under high vacuum, was removed from the manifold
and glycerin bath. The sublimator remained at room temper-
ature and high vacuum until opened for analysis of the
distillate. Then air was slowly readmitted and the subli-
mator was disassembled. The distillate on the condenser
was removed by rinsing the glass surface with four 5 ml volumes
of ethanol and four 5 ml volumes ethyl ether. The combined
solvents were removed on a rotary vacuum evaporator. The
residue was dissolved in 10 ml of 0.1N K03 and saponified
for 15 minutes at room temperature. The solution was poured
into 90 ml of distilled water and the nonsaponifiable material
was extracted with three 30 ml volumes of ether-hexane (3:17).
The combined extracts were dried over sodium sulfate, trans-
ferred to a 5 ml Pierce Reactivial (Pierce Chemical Company)

and concentrated to 0.5 ml under a stream of nitrogen gas.

WW

Preparative thin layer chromatography was performed on
"pro-absorption” silica gel thin layer plates (Type PLQF,
1 mm thickness, 20 X20 cm, Quantum Industries). The sample
was applied to the center of the plate. A solution of
authentic 019-416-steroids was applied to the edges of the
plate to serve as a marker. The plate was developed in
benzene-ethyl acetate (95:5) in a tank under saturated

conditions.

37

Following development, the center of the plate was
covered with glass to protect the sample. Then the sides
of the plate were sprayed with a sulfuric-acetic acid
mixture (1 :1) to visualize the C19-A16-steroid standard.
The apprOpriate region of the plate was removed and placed
in a small fritted glass filter funnel (#36060, 2ml, medium
porosity, Corning Glass Works). The sample was recovered
from the silica gel by a 1 hour continuous elution with
four 1 ml volumes of chloroform-methanol (2:1) as recommended
by VandenHeuvel (1967). The solution was collected in a
5 ml Pierce Reactivial (Pierce Chemical Company) and
evaporated to dryness with a stream of nitrogen gas. The

residue was redissolved in 50 ul of ethyl acetate and stored

in a desiccator at -10°C.

Li d C Chromato a

Newly synthesized 019-416-steroids were purified by
chromatography on Silica Gel H. The glass column, 110 cm X
2 cm (i.d.), had a solvent reservoir at the top and a sintered
glass disk and Teflon stopcock at the bottom. Two kinds of
column material were used: (1) a 20 g column of Silica Gel H
for separations of steroids differing in the number and
position of hydroxyl and carbonyl groups; and (2) a 30 g
column of AgNOB-impregnated Silica Gel H for separation of
steroids differing in the number of carbon-carbon double bonds.
Silica Gel H was used without further preparation as received
from the supplier. The AgNOB-impregnated Silica Gel H was

38

prepared by the method of Katkov and Gower (1970) and stored

in the dark in a desiccator over phosphorous pentoxide.

Silica Gei H Column. This procedure was adapted from
that of Claus pp.§i. (1971). The column was filled to a
height of 14 cm with 20 g of Silica Gel H suspended in hexane.
The column was prewashed with 100 ml of benzene. The sample
was dissolved in a minimum amount of benzene (usually 2 ml)
and applied to the column. Elution was carried out with
benzene-ethyl acetate (9:1,v/v) and the eluate was collected
in 10 ml fractions. Nitrogen pressure was applied to achieve
a flow rate of 1 ml/minute. Each fraction was examined by
gas chromatography using a 5% SP-2401 column at 220°C.
5m-Androst-16-enp3dbol was eluted in fractions 9 and 10
and 51-androst-16-en—3p-ol was eluted in fractions 12,13,14
and 15.

wrinpremaied Silica El 001m. The procedure of
Katkov and Gower (1970) was used with only minor modifications.

The column was prepared as follows: the column was filled

with hexane and 4 g of alumina was poured into it followed

by 30 g of the impregnated gel. The alumina layer served as

an adsorbent for any AgNO3 displaced from the upper layer
(Katkov and Gower, 1970). The column was prewashed with

100 ml of benzene. The sample was applied in a benzene
solution (2 ml) and eluted with benzene-ethyl acetate (3:1,v/v).
Nitrogen pressure was applied to achieve a flow rate of

1 ml/minute. Gas chromatographic analysis of each 1C>nfl.

39

fraction was performed as described above. Fractions 12
through 19 contained 5drandrost-16-en-3-one and fractions

23 through 33 contained 4,16-androstadien-3-one.

Gas Chromatography

Gas chromatographic analysis was done with a Beckman
GC-4 instrument with a hydrogen flame detector using a
6 ft3¢2 mm (i.d.) glass column packed with 5% SP-2401, 3%
ov-1 or 5% ov-1 on 1oo/120 mesh Supelcoport. The column
had been previously cleaned, silanized and packed with suction
by the procedure recommended by Horning gp‘gi. (1967). Helium
carrier gas flow was adjusted to 22 ml/minute. Hydrogen and
oxygen flows were set at 60 ml/minute and 300 ml/minute,
respectively. Additional helium (carrier-makeup gas) was
introduced into the hydrogen ahead of the detector at 45 ml/minute
to improve the signal to noise ratio of the detector. The
samples, dissolved in ethyl acetate, were injected directly
onto the column. Authentic 019-‘QG-steroids were used as

standards in the analysis.

Gas Chromatogpaphy—Mass Spectrometpy (GO-MS)

 

Combined GC-MS was carried out on an LKB-9000, interfaced
to a dedicated minicomputer (PDP-8/I, Digital Equipment Company)
for data acquisition and reduction (Sweeley £3.2l-v 1970).

Data was displayed on a Tektronic Model 4002A storage scope
with keyboard terminal and a Tektronic Model 4601 hard copy

unit.

40

The coiled glass GC column was 6 ft in length by 2 mm
(i.d.) and was packed with 5% SP-2401 or 3% OV-1 on 100/120
mesh Supelcoport. Mass spectral measurements were recorded
at 70 eV ionizing energy with full accelerating voltage of
3.5 kV and 60 uA trap current.

The level of steroid in G0 effluents was determined by
mass spectrometric selected ion monitoring with the aid of
computer-control of fine focus, data acquisition, reduction
and display (Holland pp‘gi., 1973). The ratio of protium
and deuterium forms of Susandrost-16-en-3-one was determined
by continuous recording of the intensities of the molecular
weight ions. The lower mass of the pair was first focused
by manual adjustment of the magnetic field strength. The
higher mass was then focused by lowering the accelerating
voltage. The areas and heights of each ion peak was deter-
mined by the computer and the ratio of protium and deuterium
forms was determined. A blank ratio of these ions for the
pure reference deuterium form was obtained and this was sub-
tracted from the isotopic abundance ratios obtained for the

samples.

41

. 16
SyntheSls of 019‘A -Steroids

6,6-d2-5w-Androst-16-en-3-one

 

The activated hydrogens in 4,16-androstadiene-3-one
were exchanged for deuteriums by equilibration in alkaline
deuteriomethanol/deuterium oxide. The exchange procedure
was adapted from a procedure used by Djerassi and Takes
(1966) for deuteriation of 5drpregn-9-en-12-one. A solution
of 25 mg of 4,16-androstadiene-3-one in 5 ml of deuterio-
methanol was saturated with 20% sodium deuterioxide in
deuterium oxide and heated under reflux for 3 days. A few
drops of deuteriomethanol were added whenever the solution
became turbid due to supersaturation. The refluxing solution
was protected from air and moisture by maintaining the
system under an atmosphere of nitrogen. After cooling, the
solution was diluted with 10 ml of dry ethyl ether and washed
with two 5 ml volumes of deuterium oxide in a dry 60 ml
separatory funnel. The ether solution was dried over
anhydrous sodium sulfate. The solvent was evaporated on
a rotary vacuum evaporator and the dry residue was stored
in a dessicator over phosphorous pentoxide.

The product, 2,2,4,6,6-d -4,16-androstadiene-3-one, was

5
converted to 2,2,4,6,6-d5-5m-androst-16-en-3-one by lithium-
ammonia reduction (Dryden, 1972). Approximately 3 ml of

ammonia was transferred into a 10 ml flask fitted with a

42

dewar condenser (9253, Ace Glass Incorp.) and a side arm
pressure equalizing addition funnel (9494T, Ace Glass Incorp.).
The condenser was refrigerated with dry ice/isopropanol.

Approximately 2 mg of clean lithium wire was added to
the liquid ammonia and an intense blue color appeared in the
solution as the lithium dissolved. Color persistence for at
least 5 minutes indicated that the ammonia was suitable for
use. Rapid disappearance of the color indicated that the
ammonia was contaminated with colloidal iron and should not
be used (Campbell, 1972).

Approximately 25 mg of the deuteriated steroid in 2 ml
of tetrahydrofuran (freshly distilled from LiAlH4) was added
to the lithium/ammonia solution.in a slow stream. The mixture
was refluxed with stirring for 1/2 hour and then the reaction
was stopped by the addition of 37 mg of anhydrous ammonium
chloride. Approximately 6 ml of ethyl ether was added to the
reaction flask and the ammonia was allowed to evaporate. The
product was extracted from the reaction vessel with three 5 ml
volumes of ethyl ether. The combined ether extracts were
dried over anhydrous sodium sulfate and evaporated on a
rotary vacuum evaporator.

The product, 2,2,4,6,6-d5-5d-androst-16-en-3-one, was
dissolved in 10 ml of methanol-water (9:1) and refluxed for
12 hours to remove deuteriums in the «at-position to the ketone.
The solution was diluted with 15 ml of water and extracted
with three 15 ml volumes of ethyl ether. The combined extracts

43

were dried over anhydrous sodium sulfate and evaporated to
dryness on a rotary vacuum evaporator. The residue was
redissolved in methanol-water and the back-exchange was
repeated to insure removal of all the exchangeable deuteriums.
The steroid was reextracted and dried in the same way. The
purity of the compound was established by thin layer and gas
chromatography and mass Spectrometry. Attempts to crystallize
the product were not successful. The yield of 6,6-d2-5xr
androst-16-en-3-one was estimated by gas chromatography to

be 10 mg, resulting in an overall yield of 40%.

5,6,6-d3-5dpAndrost-16-en-3-one

 

This material was prepared in the same way as 6,6-d2
-5a-androst-16-en-3-one except that deuterioammonia was used
in the lithium-ammonia reduction.

The deuterium exchange of 4,16-androstadiene-3-one was
achieved by refluxing 87 mg of the steroid in 15 ml of
deuteriomethanol saturated with 20% sodium deuterioxide
in deuterium oxide. The refluxing solution was protected
from air and moisture by maintaining the system under an
atmosphere of nitrogen. Refluxing was begun before the
addition of sodium deuterioxide in deuterium oxide because
it was found that serious degradation occurred in the
previous exchange when the alkali was added prior to thorough
"degassing" of the steroid solution. After 1 hour of refluxing,
the 20% sodium deuterioxide in deuterium oxide was added drOp-

wise until turbidity indicated that the solution was saturated.

44

The reflux was continued for 3 days with periodical
additions of deuteriomethanol (9 ml total) as needed to
maintain the volume of solution.

After cooling, the solution was diluted with 15 ml of
deuterium oxide and the steroid was extracted with.two 15 ml
volumes of ethyl ether. The combined ether extracts were
dried over sodium sulfate and the solvent was removed under
vacuum. The residue of 2,2,4,6,6-d5-4,16-androstadiene-3-one
was stored in.a.dea1ccator over phosphorous pentoxide.

The deuteriated Ae-3-keto steroid was converted to the
corresponding 3-keto steroid by reduction in a solution of
lithium in deuterioammonia. The reaction was carried out
in a 25 ml triple-neck flask fitted with a pressure equalizing
addition funnel (9494T, Ace Glass Incorp.) and a small dewar
condenser (9253, Ace Glass Incorp.) joined to a drying tube.
All of the glassware was previously dried for 24 hours at
110°C and 30 inches of vacuum. The condenser was filled
with dry ice/isOprOpyl alcohol, and a small Teflon coated
magnetic stirring bar was added to the flask. The bottom
of the flask was immersed in a second dry ice/isopropanol
bath to provide more condensing surfaces during the addition
of the deuterioammonia to the reaction apparatus.

Approximately 4 ml of deuterioammonia and 7 mg of
clean lithium wire were added to the flask. An intense

blue color developed in the solution when the lithium

dissolved but the color did not persist for more than a few

45
minutes. The stubborn disappearance of the color indicated
that the lithium was being consumed by impurities in the
deuterioammonia (Campbell, 1972). This situation necessi-
tated the addition of several 7 mg portions of lithium.
wire before the blue color could be maintained in the
deuterioammonia solution.

A solution of approximately 87 mg of the 2,2,4,6,6-d5-
4,16-androstadiene-3-one in 6 ml of tetrahydrofuran (freshly
distilled from LiAlH4) was slowly added with stirring to the
lithium/deuterioammonia solution using the addition funnel.

After a 30 minute refluxing period the reaction was
stopped by the careful addition of 130 mg of ammonium chloride
(previously dried at 110°C and 50 inches of vacuum for 2
hours). The flask was removed from the refrigerant bath
and 10 ml of ethyl ether was added to the flask. The deuterio-
ammonia was allowed to evaporate at room temperature.

The flask was rinsed several times with ethyl ether and
the combined ether extracts were washed with water in a 60 ml
separatory funnel. The ether solution was dried over sodium
sulfate and the solvent was removed in vacuum.

The product, 2,2,4,5,6,6-d6-5mpendrost—16-enp3-one,
was back exchanged to remove tut-deuteriums by refluxing for
2 hours in 25 ml of methanol and 10 ml of water containing
0.2 mg of sodium hydroxide. The steroid material was recovb

ered by extraction with ethyl ether. The combined extracts

were washed with water to remove the alkali, dried over sodium

sulfate and evaporated to dryness under vacuum.

46

The product was a mixture of 36 mg of 5,6,6-d3-5d-
androst-16-en-3-one and 36 mg of 2,2,4,5,6,6-d6-5d;androst-
16-en-3p-ol. These two deuteriated steroids were purified
by preparative thin layer chromatography. The purity of the
5,6,6-d3-5upandrost-16-en-3-one was established by gas
chromatography and mass spectrometry.

ros iene- -

5,16-Androstadien-3p-ol was prepared according to a
procedure published by Matthews (1968) and Matthews and
Hassner (1972). The procedure involved methyllithium
reduction of the toluenesulfonylhydrazone derivative of
the corresponding 17-keto steroid.

3p-hydroxy-5-androstene-17-one (12.2 g) and p-toluene-
sulfonylhydrazine (10.2 g) were dissolved in 212 ml of
hot methanol and refluxed for 16 hours. The reaction
mixture was concentrated on a steam bath to a volume of
approximately 170 ml and cooled to room temperature. The
crystallized steroid hydrazone was recovered by filtration,
washed with 40 ml of methanol and dried in a vacuum oven
at 65°C for 12 hours. The reaction yielded 16.7 g of 3p-
hydroxy-S-androstene-17-one tosylhydrazone.

The hydrazone (16.7 g) was dissolved in 628 ml of 1,2-
dimethoxyethane (freshly redistilled from LiAle) in a 1 l
flask fitted with a 250 ml addition funnel, a drying tube
and a magnetic stirring bar. A 2.05 M ether solution of

47

methyllithium (30 ml) was transferred to the addition
funnel using appropriate precautions to avoid an explosion
or fire that can occur when methyllithium is exposed to a
moist atmosphere. The methyllithium/ether solution was
transferred to the addition funnel using a 30 cc syringe
fitted with a needle sufficiently long to withdraw the
liquid from the reagent bottle through a small hole in

the cap. The solution was withdrawn from a cold bottle
which contained a precipitate in order to avoid the mineral
oil impurity present in the reagent (Matthews, 1968).

The methyllithium solution was added to the hydrazone
solution over a 60 minute period. The highly colored reaction
mixture was stirred for 7 hours before it was stirred into
1250 ml of ice water. The precipitated material was digested
for 12 hours on a warm steam bath to aid its removal by
filtration. The filter cake was washed with water and dried
under vacuum at 50°C for several hours.

The product was recrystallized from.ethanol-water to
give a first crOp of 4.5 g and a second crop of 2.8 g of
5,16-androstadiene-3F-ol, a total yield of 61%. The purity
of the product was established by thin layer and gas chromato-
graphy and mass spectrometry.

17-d1-5i-Androst-16-en-3dyol

 

The procedure was adapted from one described by Matthews
(1968) and Matthews and Hassner (1972). 3dphydroxy-54-androstene-

48

17-one (14.4 g) and p—toluenesulfonylhydrazine (12.2 g)
were refluxed together in 250 ml of methanol for 12 hours.
The solution was concentrated on a steam bath to a volume
of approximately 180 ml and cooled to room temperature to
promote crystallization of the hydrazone. The hydrazone
was recovered by filtration, washed with 50 ml of methanol
and dried in a vacuum oven at 65°C for several days. The
yield of tosylhydrazone was 18.9 g.

The tosylhydrazone derivative was reduced with methyl-
lithium as previously described with the exception that
deuterium oxide (100 ml) was added slowly to the reaction
mixture just before it was stirred into the ice water.

The resulting material was digested on a warm steam bath

for 24 hours before the agglomerated precipitate was filtered,
washed with water and dried in a vacuum oven for 3 hours at
45°C. Recrystallization from ethanol-water gave 5.3 g of
17-d1-5upandrost-16-ene3dbol. The purity of the product

was established by thin layer and gas chromatography and

mass spectrometry.

16,11-d2-§.]6-Andr0§iadien-ip:pi

The hydrogen atoms at C-16 of 3P-hydroxy-5-androstene
17-one (12.2 g) were exchanged for deuterium atoms by refluxing
for 16 days in 212 ml of deuteriomethanol. p-Toluenesulfonyl-
hydrazine (10.2 g) was added to the solution and the tosyl-
hydrazone derivative of the steroid was formed by refluxing

49

for an additional 16 hours in the deuteriomethanol solution.
The volume of solution was reduced by approximately 20% by
directing a stream of nitrogen into the warm flask. Crystal-
lization of the tosylhydrazone occurred as the solution
cooled. The crystals were removed by filtration, washed with
40 ml of methanol and dried under vacuum at 65°C for 12 hours.
The reaction yielded 16.6 g of steroid hydrazone.

The hydrazone was dissolved in 730 ml of 1,2-dimethoxy-
ethane and treated with methyllithium as previously described.
The highly colored reaction mixture was stirred for 7 hours.
Then deuterium oxide (150 ml) was added slowly using an
addition funnel and the mixture was stirred for an additional
6 hours. The reaction mixture was stirred into 900 ml of
water and the resulting mixture was digested on a warm steam
bath for 12 hours to aid the recovery of the product by
filtration. The filter cake was washed with water and then
dried in a vacuum oven at 50°C for 5 hours. Recrystallization
from ethanol-water yielded 7.2 g of 16,17-d2-5,16-androsta-
dien-3’-ol. The purity of the product was established by
thin layer and gas chromatography and mass spectrometry.

16,11—d2-§,16-Androstadien-3-one

This compound was prepared by Oppenauer oxidation of
16,17-d2-5,16-androstadien-3p-ol using a procedure described
by Matthews (1968). A solution of 6 g of 16,17-d2-5,16-
androstadien-3f-ol in 300 ml of toluene and 47 ml

50

of cyclohexanone was distilled for 30 minutes to eliminate
traces of moisture. The distillate (approximately 10 ml)
was removed from the apparatus using a Stark-Dean moisture
receiver.

A solution of 3.1 g of aluminum isopropoxide in 30 ml
of toluene was added to the mixture and the combined solution
was refluxed for 2 hours. Water (80 ml) was added and the
volatile components were removed by 4 hours of steam
distillation. The residue was extracted with 500 ml of
chloroform and then with 300 ml of dichloromethane. The
combined extracts were washed with water, dried over anhydrous
sodium sulfate and concentrated to 25 ml under vacuum.
Hexane (100 ml) was added to the concentrate and the solution
was further concentrated until crystallization was imminent.
The crystals were recovered by filtration and rinsed with a
minimum amount of hexane before being dried for 6 hours in a
vacuum oven at 35°C. The yield of the reaction was 2.9 g
of 16,17-d2-4,16-androstadiene-3-one, the purity of which
was established by thin layer and gas chromatography and

mass spectrometry.

6,6,16,17-d4-5d-Androst-16-en-3-one

 

16,17-d2-4,16-Androstadiene-3-one (previously dried
for 24 hours at 50° C over P205) was converted to 2,2,4,
6,6,16,17-d7-4,16-androstadiene-3-one by refluxing 2 g

51

of the steroid for 4 days in 110 ml of deuteriomethanol
containing 10 drops of 20% sodium deuterioxide in deuterium
oxide. At the conclusion of this equilibration period the
perdeuterio—steroid was recovered from the deuteriomethanol
solution by crystallization brought about by the addition
of deuterium oxide. The crystallized steroid was recovered
by filtration and washed with a small amount of deuterium
oxide. This product was dried for 3 hours in a vacuum
oven at 40°C and then for 8 hours in a desiccator over P205
at 50°C. The exchange procedure yielded 1.70 gm of 2.2.4,
6,6,16,17—d7-4,16-andr0stadiene-3-one.

The 2,2,4,6,6,16,17-d7-4.16-andr0stadiene-3-0ne was
converted to the corresponding'Sxeh-3-one steroid by reduction
of the A4-double bond in lithium/ammonia solution. Lithium
wire (0.25 g) was dissolved in 250 ml of ammonia in a 500 ml
flask fitted with an addition funnel and a dewar condenser.
The ammonia was freshly distilled from sodium and passed
through a refrigerated trap (~21°C) to remove any water
contaminating the ammonia. A solution of 1.7 g of 2,2,4.
6,6,16,17-d7-4,16-andr0stadiene-3-0ne in 25 ml of tetra-
hydrofuran (freshly redistilled from LiAlHa) was added in
a slow stream to the lithium/ammonia solution. The reaction
mixture was stirred at reflux for 1 hour by a Teflon coated
stirring bar. The reaction was st0pped by the addition of
a saturated ammonium chloride in tetrahydrofuran until the
blue color disappeared. The ammonia was allowed to evaporate
and the product was extracted with ethyl ether. The ether

52

extract was washed 2 times with 1 volume of 0.1N HCl, 2 times
with 1 volume of 0.1M NaHCO3 and 3 times with 1 volume of
water. The solution was dried over sodium sulfate and the
solvent was removed in vacuum. The residue was purified

by chromatography on a column containing AgNO3-impregnated
silica gel (prepared as described earlier), and then recrys-
tallized from acetone-hexane. The yield was 67 mg of 2,2,4.
6,6,16,17-d7-5qrandrost-16-en-3-one.

Theq-deuteriums were removed by equilibration in
methanol-water and the final product, 6,6,16,17-d4-5¢+andr0st-
16-en-3-one, was recrystallized from the same solvents. Thin
layer chromatography, gas chromatography and mass spectrometry
were used to establish the purity of the product.

6 5 16 1 44Wt-16-en-3u-pi

The procedure was adapted from one published by Wheeler
and Wheeler (1972). A solution of 6,6,16,17-d4—5qpandrost—
16-en—3-one (0.63 g), chloroiridic acid (0.32 g), trimethyl
phosphite (3.3 ml) and water (6.5 ml) in 2-pr0panol (49 ml)
was boiled under reflux for 5 days. After cooling, the reaction
mixture was transferred to a 250 ml separatory funnel containing
50 ml of ether and 50 ml of water. The product was extracted
3 times with 50 ml of ether. The combined ether extracts were.
dried over anhydrous sodium sulfate and then evaporated to

dryness.

53

The residue was purified by chromatography on a column
packed with silica gel impregnated with AgNO3 as described
earlier. The product was further purified by chromatography
on a second silica gel column without AgN03. The appropriate
fractions were combined and evaporated to dryness. The product
was recrystallized from acetone-hexane to yield 0.11 g of 6,6,
16,17-d4-5xpandrost-16-en-3K-ol. The product purity was estab-

lished by thin layer and gas chromatography and mass spectrometry.

6,6,16,17-d4-515Androst-16-en-3P-ol

 

The procedure was adapted from one published by Takes and
Throop (1972). A mixture of 6,6,16,17-d4-5fl-androst-16-en-3-
one (145 mg) and lithium aluminum hydride (66 mg) in dry ether
(15 ml, freshly distilled from lithium aluminum hydride) was
heated under reflux for 3 hours. The excess hydride was then
decomposed by careful dropwise addition of water. The reaction
mixture was washed with 0.1N HCl to remove the residue from the
hydride. Then the solution was washed with water, dried with
sodium sulfate and evaporated to dryness.

Gas chromatographic analysis of the product revealed that
it consisted of 96% of the 3F-hydroxy compound and 4% of the
corresponding 3qshydroxy compound. The 3F-hydroxy steroid was
purified by chromatography on a silica gel column as outlined
earlier and crystallized from acetone-hexane. The reaction
yielded 73 m8 of 6 , 6 ,16 ,1 7-d4~5«-andr0st-16-en-3’-ol . The
purity of the product was established by gas chromatography

and mass spectrometry.

RESULTS AND DISCUSSION

Comparisons of Levels of 56-Androst-16-en-3-one

and Boar Odor Scores

Boar odor is not usually a serious problem in meat
packing plants processing only female and castrate male
hogs. Occasionally, however, cryptorchid pigs enter the
plant and go undetected until the discovery of testicles
in the abdominal cavity upon evisceration. The carcasses
must be tested for boar odor in the cooler so that any
possessing boar odor can be restricted for use in commi-
nuted cooked products, for rendering or condemned for use
as human food. The carcasses are tested by heating the
back fat with a hot iron and smelling the heated fat to
detect the odors produced. The test is simple in theory,
but in practice it is difficult because of the presence of
other odors in the plant and also because of variations in
the ability of different inspectors to detect the undesirable
odor.

The USDA proposed an amendment to the federal meat
inspection regulations that would exclude all cryptorchid
pork from the fresh pork market (National Provisioner, 1967).
Otherwise, they argued that it was impossible to insure
cryptorchid pork containing boar odor was excluded from the

54

55
fresh meat market because of the difficulties associated

with the testing procedure. The American Meat Institute,

a special interest group representing meat packers, argued
that the ruling was unnecessary since packing plant personnel
could test cryptorchid pork for boar odor. However, USDA
officials questioned whether or not the packers would be any
more effective than the USDA inspectors in insuring that
cryptorchid pork with the undesirable odor did not reach the
market place.

The objective of this study was to determine whether
packing plant personnel could estimate the levels of 5m-
androst-16-en-3-one in backfat of cryptorchid pigs by the
hot iron test for boar odor. This involved comparing the
results of the olfactory tests with actual levels of 5‘-
androst-16-en—3-one as determined by chemical analysis.
These results are shown in Table 1 where 21 cryptorchid
samples, previously tested for odor by personnel of a meat
packing firm, are listed in order of increasing concentration
of 519androst-16-enp3-one.

In general, the odor rankings of the plant personnel do
not accurately reflect the changing concentration of 51-
androst-16-ene3-one as can be seen in Table 1. This is
shown graphically in Figure 2 where the odor scores are
plotted against the concentration of 5a-androst-16-en-3-one.
For the purposes of this comparison the odor scores were
assigned numerical values as follows: no boar odor, 0; very

slight, 1; slight, 2; moderate. 31 slightly pronounced, 4;

56

pronounced, 5; and very pronounced, 6.

The odor scores did not correlate significantly (P<L05)
with the concentration of 5fl-androst-16-en-3-one (r-0.40),
which is in contrast to reports by Fuchs (1972) and Howell
pp pi. (1973) who found correlation coefficients of 0.75
and 0.53, respectively, between odor intensity and the level
of Sd-androst-16-enp3-one. The discrepancy between the results
reported herein and those of Fuchs (1972) and Newell pphgi.
(1973) is due, in part, to differences in the make-up of the
odor panels, and to differences in the instructions given to
them. The significant correlations reported by Fuchs (1972)
and Newell 23 pi. (1973) were obtained using panels composed
of individuals selected for their ability to smell 5(eandrost-
16-en-3-one and trained to distinguish its odor from other
odors associated with heated pork fat. Therefore, it is not
surprising that they were able to obtain positive correlations
between quandrost-16-en-3-one concentration and odor assign-
ments. In contrast, the plant personnel used in this study
were asked to rank the samples according to the degree of
undesirable odor, generally referred to as boar odor. Their
task was to score the samples according to the intensity of
the undesirable odor which was assumed to be unacceptable to
consumers of meat products.

The panel composed of packing plant personnel is more
like the untrained panel used by Newell ppflgi. (1973). Members
of this panel could detect the odor of 54-androst-16-enp3-one,
but they were not trained to specifically distinguish this

57

TABLE 1. Comparison of the Concentration of 5Q-Androst-
16-en-3-one and Boar Odor Intensity Scores
by Employees of a Meat Processing Plant

 

 

 

Sample Concentration of Bear a
Number SMPAndrost-16-en-3-one Odor
P876 fat Intensity
10174 40.10 0
14736 < 0.10 0
15025 40.10 2
74761 ’ 40.10 2
31751 (0010 1
15486 0.19 5
62342 0.22 4
63267 0.42 5
88651 0.42 4
75122 0.61 2
39528 0.67 2
17820 0.82 3
08899 1 .12 3
00556 1.23 3
43253 1 .24 5
94015 1.45 6
14924 1.56 4
70312 1.88 2
70707 1.88 5
02985 5.70 4
00911 6.97 5

 

a. Intensity: C. = none, 1 a very slight, 2 - slight, 3 :-
moderate, 4 = slightly pronounced, 5 a pronounced, 6 a
very pronounced.

Boar Odor Intensity

a)

 

.h

58

 

 

 

 

n- O
0 0 Q 0 0
1» o o o e
A 3 so 0
1+ 9
1D 6
II
V I If T T
0 1 1 2 5 6
(pg/g of Fat)

Concentration of 5¢3Androst-16-en-3-one

Figure 2. Graph of 51+Androst-16-enP3-one
‘vs. Boar Odor Intensity Scores

59

odor from other odors that are apparent when pork fat is
heated. Their odor scores did not correlate significantly
with the level of Sirandrost-16-enp3-one in fat (Newell gpflai.,
1973). which is in agreement with the results of the packing
house personnel as reported herein.

The same fat samples were odor tested in the laboratory
by a three member panel composed of individuals with demon-
strated ability to smell low levels of SdPandrost-16-enp3-one,
5d-androst-16-enp3w-ol, 4,16-androstadien-3-one and 5,16-
androstadien-3P-ol. The boar odor intensity, rated on a
scale of 0 (no boar odor) to 9 (pronounced boar odor), are
shown in Table 2 where the fat samples are listed in order
of increasing concentration of Sebandrost-16-en—3-one.

The average scores of this panel were positively corre-
lated (r=0.51, P<t05) with the level of 5dbandrost-16-enp3-
one. Even though the correlation coefficient for the level
of quandrost-16-en-3-one and the odor scores was significant,
it accounted for only 25% of the variation. This compares
closely with the correlation of 0.53 obtained in a similar
study by Newell‘ngpi. (1973) with a trained panel. The
correlation of 0.73 obtained by Fuchs (1972) is not directly
comparable with the correlation of 0.51 obtained in this
study because Fuchs (1972) decided to exclude from his calcup
lations those samples having the highest concentrations of
5d-androst-16-en-3-one (5 to 8 rig/g fat).

The ”urine-like" and ”perspiration-like" nature of boar

odor appears to result from a complex interplay of several

60

Table 2. Comparison of the Concentration of SihAndrost-
16-enp3-one and Boar Odor Intensity Scores by
Laboratory Panel

 

 

 

Sample Concentration of Boar a
Number 51-Androst-16-enp3-one Odor
Intensity
pg/g of fat
10174 0.10 0.0
14736 0.10 1.3
13025 0.10 1.8
74761 0.10 0.0
31751 0.10 0.3
15486 0.19 1.3
62342 0.22 3.3
63267 0.42 0.8
88651 0.42 3.3
75122 0.61 1.8
39528 0.67 1.5
17820 0.82 5.5
08899 1.12 0.5
00556 1.23 2.0
43253 1.24 1.3
94015 1.45 8.5
14924 1.56 4.8
70312 1.88 1.8
70707 1.88 3.0
02985 5.70 6.0
00911 6.97 4.8

 

a. Intensity:

0 (no boar odor) to 9 (pronounced boar odor).

61

undesirable odors, with 5¢randrost-16-en-3-one being an
important contributor. .As can be seen in Tables 1 and 2,

a high concentration of this steroid is a good indication
that the undesirable odor will be evident when the fat

sample is heated. However, a low level of quandrost-16-
en-3-one does not guarantee freedom from undesirable odors.
Therefore, it is unlikely that a chemical assay for Sue
androst-16-en-3-0ne alone would be helpful as an objective
test for boar odor, but an assay that included other 019-516-
steroids, especially the highly odoriferous 5w—androst-16-en-
306-01, might give a better indication of the extent of odor
to be expected in the sample.

During the evaluation of samples in this laboratory the
distinctive musky odor of 5N+androst-16-enp3«+ol was notice-
able in several of the samples, especially sample #88651 and
#14924. This observation prompted efforts to prepare a
deuterium labeled form of this compound to be used as an
internal standard in an assay using mass spectrometric reverse
isotopic dilution. The labeled compound was prepared as out-
lined earlier but unfortunately there was not enough time to
run the assay.

The failure of the meat plant personnel to discriminate
between samples with various levels of Sdhandrost-16-enp3-one
does not discredit the hot iron technique as a way of selecting
cryptorchid animals with no boar odor. All animals with levels
of 51-androst-16-en-3-one greater than 0.1 pg/g fat were rated
as having at least slight boar odor, which would disqualify

62

them for use in the fresh pork market. However, the selection
of carcasses eligible for use in cooked, comminuted products
(less than pronounced odor), as against carcasses to be
condemned or for rendering (pronounced odor), may be a.more

difficult task based on the observations presented here.

Procedure for the Assay

of SU-Androst-16-enp3-one

The original plan called for the assay of'5dbandrost-
16-en—3-one by the vacuum distillation method of Patterson
(1969). However, alternative procedures were considered very
early for several reasons. First, the distillation step
required constant supervision, and even with careful attention
to details it was difficult to reproduce the distillation
conditions. Another serious objection was the absence of a
suitable way to correct for manipulative losses during the
procedure. In addition, the saponification step was found
to result in some destruction of quandrost-16-enp3-one.
Coupled with these chemical and manipulative losses was the
uncertainty associated with gas chromatographic analysis
performed without an internal standard.

The technique of mass spectrometric stable isotope
dilution seemed uniquely well suited for this analysis. Losses
during the entire procedure could be automatically corrected

63

for through the use of a deuterium labeled form of 5<~androst-
16-enp3-one as an internal standard. Secondly, the sample
size could be substantially reduced because the deuterium
form.of 5d—androst-16-enp3-one could serve as carrier for
the smaller amount of the protium form being'measured.

The original intention was to check the accuracy of the
distillation technique by comparisons with the results from
a few samples analyzed by reverse isotope dilution. However,
this became impossible when the distillation apparatus was
destroyed by a fire in the laboratory. It was the loss of
this equipment that prompted the use of the reverse isot0pe
dilution/carrier procedure for analysis of all of the samples.
This was made possible by the generous cooperation of Drs.

0.0. Sweeley and J.F. Holland of the Mass Spectrometry Laboratory
in the Department of Biochemistry at Michigan State University.
The stable isotope dilution/carrier technique was used

to measure the levels of Supandrost-16-en-3-one which are

given for cryptorchid fat samples in Tables 1 and 2. Following
the addition of the deuterium labeled internal standard

(usually 21‘pg of 6,6-d2-5x-androst-16-en-3-one), the mixture

of deuterium and protium forms was purified by saponification
and TLC. The resultant fraction was analyzed by GC to check

its purity before analysis by combined GCqMS. The GC trace
shown in Figure 3 is typical of the results obtained for all

fat samples. It shows that the purification procedure involving
saponification and TLC resulted in a relatively pure sample

64

with 5s-androst-16-en-3-one eluting at about 7 minutes. The
peak at 7 minutes contained both the protium form of 5u~androst-
16—enp3-one, which was originally present in the sample, and

the dZ-form added as an internal standard and carrier.

Figure 4 shows a GC trace obtained for one sample to
which no deuterium labeled standard had been added. No
measureable amount of 5d-androst-16-en-3-one can be seen at
7 minutes. It is evident from the GC trace in Figure 4 that
ordinary GC techniques without carrier could not be used to
assay this sample. This sample was easily assayed by come
bined GCqMS when an excess of deuterium labeled Subandrost-
16-en-3—0ne was added as a carrier and internal standard.

The sample was found to contain Surandrost-16-enp3-one at
a level of 0.42 Pg/g fat (sample #88651, Table 3).

The mass spectrometer was used to determine the relative
amounts of protium and deuterium form present in the 5“?
androst-16-en-3-one peak for each of the samples. Acting as
a very specific and sensitive GC detector, the mass spectrometer
registered the intensity of the molecular weight ions of the
two forms as they emerged from the column. The molecular
weight ions of the protium form (m/e 272) and for the dz-form
(m/e 274) can be seen in the abbreviated mass spectra shown
in Figure 5. Figure 6 shows the result obtained by monitoring
the intensity of m/e 272 and 274 during the elution of a
typical sample. Peak heights were used in the isotopic
ratio determinations.

65

RESPONSE

DETECTOR

 

 

 

 

 

J UL-

‘ ‘

 

O 5 ‘IO 15 20
MINUTES

Figure 3. Gas Chromatogram of 5d—Androst-16-en-3-one
Isolated with Deuterium Labeled Carrier.

This material was isolated from 2 g of backfat to which had
been added 21 pg of 6,6-d2-5drandrost-16-en-3-one. The
asterisk identifies the peak containing the protium and
deuterium forms of 5K-androst-16-en-3-one. Conditions:

6 ftxz mm column packed with 3% sr-2401, operated at 183°C.

66

RESPONSE

DETECTOR

r-~

O 5 10 IS 20
MINUTES

Figure 4. Gas Chromatogram of SieAndrost-16-enp3-one

Isolated without Carrier.

This material was isolated from 2 g of backfat. The asterisk
denoted the region of the chromatogram corresponding to the
retention time of Surandrost-16-en-3-onc. Operating conditions
are the same as those described in the legend to Figure 3.

67

 

 

 

 

 

 

 

 

 

 

 

 

 

109 +
M
a 272
0' 504
9
g ..
. T Lr 'F' l rj T 147 T
b W
a 274
50¢
.4-
L—
..nannJ IL jlll [Ll min ‘1 a .1
0 r F I r r I r r f I r
250 270 290
m/e

Figure 5. Abbreviated Mass Spectra of Protium and
Deuterium Forms of SKPAndrost-16-cn-3-one.

a) SupAndrost-16-en-3-one; b) 6,6-d2-54-Androst-16-en-3-one
Spectra recorded at 70.0 eV with a GC inlet consisting of a
6 ft)(2 mm glass column packed with 3% SP-2401. The complete

mass spectra are shown in Figure 7.

68

I MIG

l
I
g :1- 274
‘2 1'
2 1'
33 I
as l I
11
[I
a: 1
E: I |
a 1.
3 1'
' I
1
1

 

IIIUTES

Figure 6. Molecular Ion Intensity Recordings for Protium

and Deuterium Forms of 5x-Androet-16—en-3-one.

Selected ion monitoring of the M+ ions for the p-form (m/e 272)
and dz-form (m/e 274) as they elute from the 00 column. The
sample was the same one shown in Figure 3. Operating conditions

are given in the legend to Figure 5.

69

The isotopic abundance ratios obtained for the exper-
iment are shown in Table 3 where the samples are listed in
the order of their analysis. The concentration of Sat-androst-
16-enp3-one (column D) was obtained by multiplication of the
difference in the protium/deuterium ratio (column B) by the
amount of deuterium carrier added to the original sample.

The difference in protium/deuterium ratio was obtained by
subtraction of the ratio at ion pair m/e 272/274 in the
pure deuteriated carrier from the ratio obtained for each
sample (column A). This difference was taken to be propor-
tional to the amount of protium form present in the sample.

In the analysis of some samples (#14736, 13025, 10175
and 74761) the protium/deuterium ratios (column A in Table
3) were smaller than the protium/deuterium ratio for the
pure deuteriated carrier. There are several possible expla-
nations for this incongruity.

The samples may have contained an impurity that co-chro-
matographed with Sqrandrost-16-enp3-one and produced an ion
at m/e 274, thereby reducing the ratio of m/e 272/274 below
that of the added deuteriated carrier. The SP-2401 liquid
phase of the GC column does not produce a measurable ion at
m/e 274. Therefore, column bleed was ruled out as a possible
source of an inflated intensity value for m/e 274.

If the deuteriated carrier was not sufficiently pure
to begin with, it could have become more purified during
its isolation from the fat sample. A lowering of the

70

Table 3. Assay of 5dpAndrost-16-enp3-one

 

 

 

 

 

sample protigzéiguterium cargfer 5m-Andrggt-16-ene
added to 3-0ne/g of fat
2 g fat
Ratio Difference sample
A .B C

(a-.024) (3-0/2)
carrier .024 -- -- -
carrier .024 -- -- -
31751 .032 .008 21 (0.10
14924 .173 .149 21 1.56
75122 .082 .058 21 0.61
14736 .020 H 21 < 0.10
62342 .045 .021 21 0.22
63267 .064 .040 21 0.42
13025 .016 (-) 21 <o.1o
88651 .064 .040 21 0.42
43253 .142 .118 21 1.24
02985 .566 .542 21 5.70
70312 .203 .179 21 1.88
08899 .131 .107 21 1.12
10174 .018 (-) 21 <o.1o
17820 .102 .078 21 0.82
74761 .016 (-) 21 <o.1o
39528 .088 .064 21 0.67
94015 .162 .138 21 1.45
15486 .042 .018 21 0.19
00556 .141 .117 21 1.23
00911 .688 .664 21 6.97
70707 .203 .179 21 1.88

 

a. Ratio (A) was smaller than ratio for carrier. Therefore,
values of¢<0.10 were assigned. b. The ratio (A) was only
slightly larger than ratio for carrier. Since the difference

(B) may not be significant, an arbitrary value of <0.10 was
assigned.

71

ratio of m/e 272/274 could then occur if the impurities
which were removed had been partly responsible for the

ion intensity at m/e 272 in the pure deuteriated carrier.
Errors of this type could have been.minimized if the back-
ground value (ratio of m/e 272/274 for the carrier) had
been determined on carrier material that had been treated
in the same way as the samples.

For the reasons discussed above, the concentration of
5N?androst-16-en—3-one in these four samples were assigned
values ofc<0.1 ug/g of fat. Sample #31751, with a calculated
level of 0.08 jig/g, was also reported as having <0.1 pg/g
of fat because it was considered doubtful that a ratio
difference (column B of Table 3) as small as 0.008 could
be significant in view of the possible errors in the assay,

which are discussed below.

Possible Errors in the Assay Procedure

 

The data presented in Tables 1, 2 and 3 is considered
to be sufficiently accurate to serve the purposes of this
study, and perhaps more reliable than other published data
for similar samples assayed without any corrections for
losses (Patterson, 1969; Fuchs, 1972; Stinson.ppflgi., 1972;
Newell pp,gi., 1973). Nevertheless, the levels of 5dr
androst-16-en-3-one reported herein are considered to be

approximations because of the possibility of errors caused

72

by "memory" effects in the GC-MS instrument, and the reliance
on single measurements for each sample.

Errors in isotopic abundance measurement could have
resulted from background due to a carry-over effect from
previous samples examined ("memory effect"). According to
Beynon (1960) the only solution in such cases is to remove
all memory of the previous sample which can be accomplished
by repeatedly flushing with the new sample until a constant
isotopic abundance ratio is achieved. Ratio measurements
made in this study were based on only single injections of
each sample so it is quite likely that some errors may have
resulted, especially in cases where a low level sample was
preceded by a much more concentrated one.

The first step in reverse isotope dilution analysis
should be the preparation of a working curve showing a
linear relationship between observed isot0pic abundance
ratios and actual weight ratios upon injection of a dilution
series comprised of mixtures containing protium and deuterium
forms of the compound being measured. The use of such a curve
would eliminate errors arising from differences in ion inten-
sities between the protium and deuterium forms which may
result from fragmentation isotope effects and mass discrim-
inations in the mass spectrometer.

No correction was made for the cross-contribution of
the M3+2 natural satellite isotope peak from the protium
form (MI, m/e 272) and the ion intensity measured for the

73
dZ-form which also appears at m/e 274. Since the intensity
of the MI+2 satellite ion will be only approximately 0.12%
of the intensity of the M+ ion, this correction can be
ignored for isotopic abundance measurements on mixtures
containing small amounts of the protium form.

Even though the reverse isotOpe dilution method did
not give absolute values for 5q+androst-16-en-3-one, exact
concentrations are not required to relate to odor scores.
The levels of Subandrost-16-en-3-one reported herein are
in agreement with the findings of other workers (Claus pp
gi., 1971; Rhodes and Patterson, 1971; Fuchs, 1972; Stinson
gp'gi., 1972; Newell 23 pi., 1973) who reported concentrations

of 0.1 to 7 ug/g of fat in studies using fat samples

from uncastrated male hogs.

Purification by LiAlH4 Reduction and Urea Clathrate Formation

 

Consideration was given to a purification scheme
involving LiAlH4 reduction and urea Clathrate formation
instead of saponification for the removal of glyceride
esters during the purification of Sqfandrost-16-enp3-one
from fat samples. This was prompted by the desire to reduce
the time and labor requirements of the assay and to reduce
destruction of Sq-androst-16-en-3-one. Results of preliminary
experiments showed that these combined steps yielded 0.6 g
of material from a 2 g fat sample. However, the samples
prepared in this way were not sufficiently pure for GC-MS
analysis without additional preparative chromatography steps.

74

Reduction of the glyceride esters with LiAlH would be

4
accompanied by reduction of the 3-ketone group of 5d-androst-
16-en-3-one. The resulting 3f-alcohol would be indistinguish-
able from Sueandrost-16-en-3P-ol originally present in the
sample. However, by the use of LiAlD4 instead of LiAlH4 it
would be possible to distinguish between the two since any
alcohol formed during the reduction would have deuterium
incorporated at the 3qrposition. Therefore, any 5d-androst-
16-en-3-one (MW 272) present in the original sample would be
converted to 3d-d1-5m-androst-16-en-3feol (MW 275). and a
carrier/internal standard such as 6,6-d2-5d-androst-16-en-
3-one (MW 274) would be converted to 3d;6,6-d3-5d+androst-

16-en-3f-ol (MW 277) as shown below.

 

 

A),
o: LiAlD
E 4 3°: ..
D 11
MW = 272 W = 275
a).
no \ o
\ .-
D H

D D

 

MW 3277

75

The ratio of the two isotopic forms could be determined
by measuring the intensities of their molecular weight ions
(d1-form, m/e 275; dB-form, m/e 277) in GC-MS analysis.

Then the concentration of 5u-androst-16-en-3-one originally
present in the sample could be calculated by multiplication
of the isotopic abundance ratio by the weight of carrier
added to the sample.

Although time did not permit an evaluation of the
entire procedure described above, the proposed method
appears to be superior in some respects to the methods

described earlier which involved saponification.

76

Deuterium Labeled C19-A16-Steroids

Deuterium labeled steroids were prepared to be used as
internal standards and carriers in the assay of 019-131 6-steroids
by mass spectrometric reverse isot0pe dilution. The general
features of the synthetic steps are discussed in this section,
while the details of the reactions are given under Experimental.
Deuterium labeled forms of the following five steroids were
prepared: 5q-androst-16-en-3-one (I), 5w-androst-16-en-3d-ol
(II), 54-androst-16-en-3P-ol (III), 5,16-androstadien-3P-ol
(IV), and 4,16-androstadien-3-one (V). The structural formulas

for each steroid are shown below:

 

77

6,6-d2-51-Androst-16-en-3-one

 

One convenient method for preparing a site-specifically
deuteriated form of 5l-androst-16-en-3-one would be to exchange
the four activated hydrogens in the1qrposition of the carbonyl
group. However, the resultant material would not have been
suitable as an internal standard in isotope dilution analysis
because the label would back exchange with protons in the medium.

An alternative approach was selected which led to the
introduction of two deuteriums at carbon-6, three carbon atoms
away from the carbonyl group (Shapiro gp‘gi., 1964). The

synthetic steps are shown below.

6 . .
D
0‘ reflux-3 days * 05.
CHBOD + DZO/Na

Li/NH3

reflux-3 hrs .
7' 0" illllll

CH3OH + HZO/Na .H

 

   
 

 

VII VIII

78

In the first step, the activated hydrogens in 4,16-
androstadien-3-one (V) were exchanged with deuteriums by
equilibration in methanol-OD/DZO. The product (VI) was
reacted with a solution of lithium metal in liquid ammonia
(Li/NH3). The Li/NH3 reduced only the A4-double bond and
yielded 2,2,4,6,6-d5-5weandrost-16-en-3-one (VII). The
exchangeable deuteriums at C-2 and C-4 were removed by back
exchange in methanol-OH/HZO. The two deuteriums attached
to C-6 are no longer exchangeable since the enone structure
that existed in the original material has been lost by
saturation of the double bond between C-4 and C-5. These
reaction steps resulted in a 40%»yield (10 mg) of 6,6-d2-
quandrost-16-en-3-one (VIII). _

The mass spectrum of 6,6—d2-515androst-16-en-3-one (VIII)
is shown in Figure'7a. By comparison with the mass spectrum

of the unlabeled form, shown in Figure'7b, it can be seen

that the molecular weight ion (M+) at m/e 272 has been shifted

to m/e 274 and the ion at m/e 257 (MI-CH3) has been shifted to

m/e 259 in the labeled steroid. Either of these ions could

be monitored for the measurement of 5‘-androst-16-en-3-one (I)

in gas chromatographic effluents by selected ion monitoring
mass spectrometry.

The 6,6-d2-5N-androst-16-en—3-one (VIII) prepared in
this experiment was used as an internal standard and carrier
in the assay of 5s-androst-16-en-3-one (I) in fat samples
evaluated earlier for boar odor intensity. The results

were discussed earlier herein.

.. .lrll '1 IIJIIII1IIIIIIII. I] lull Ill 1|.lu1lllllv ‘ (i

79

.oomfi pm mfiamsemzeomfi seememeo Aemms oma\oofiv
psoaooamazm co Hoqmlmm mm pea: Umxoma ABE m x pm ow Cezaoo mmmaw m mo Umpmfimcoo

pchH om one .>m on mo Hwfipcmpoq wcHNHCOH as us compooms mew: mspomqm mmmz

mconmucmlmalumOLGC¢|smnmonw.m.m Ao

monoumlcmlwanpmopUC¢nsm An mocclmlcmnwHipmopuc<nsmumcum.o Am

.mconmlcwlmanpmopcc<usm mo mEpom Esfipmpsmo @cm EZHpOLm mo mhpomam mmmz .w mpswfim

 

T

 

 

259

 

 

 

 

100

80~ a.

60“

4o
20‘

(z)

100 150 200 250

50

80

 

 

 

 

 

 

 

 

 

 

 

 

\T w \T w
L 1 l J L l 1 J
N :— p
. a _ . N
r
e *0 o
In LU) g
L
I 8 2
III N O
\E) a
0
L0
H
O
O
H
O
.9 m e
I I i r T T ‘
O O O O O O O O O
2 CD LO Q” N O (I) (0 N
q—a

ALISNBLNI BAIIVWHH

100 150 200 250

50

rVWx/E

81

5,6,6-d3-Su-Androst-16-en-3-one

 

This synthesis was undertaken because it was thought
that a dB-form of Sx-androst-16-en-3-one would be a better
internal standard than the dZ-form due to lower values for
background. In this case, the term "background" refers to
the relative amount of m/e 272 (M+ for the unlabeled form)
in the mass spectrum of the labeled form. Also, the use of
a d3-labe1ed standard would eliminate the cross contribution
that would exist between the natural isotope peak at MI+ 2
in the unlabeled form and the M+ peak in a dz-form.

dB-SK-androst-16-en-3-one was prepared as previously
described for the dz—form except that deuterioammonia was
used in the lithium/ammonia reduction instead of ordinary

ammonia as shown below.

D
D
reflux-3 days
0" _,, (y’lll!!!
CHaoD/DZO
D D D
v Li:::3//’/" . VI

_’reflux-3 hrs 3 ’ .
CH OH/H o 0" -
3 2 D

D D

  
   

 

 

X

82

The starting material, 4,16-androstadien-3-one (V),
was subjected to hydrogen-deuterium exchange prior to reduction
with Li/NDB, and then the deuteriums in thecx-position of (IX)
were removed by back exchange. The products of the above
reactions consisted of 34 mg (41% yield) of 5,6,6-d3—SQr
androst-16-en-3-one (X) and an equal amount of the corresponding
Qp-alcohol. The production of substantial amounts of alcohol
was predictable since over-reduction to an alcohol is a common
side reaction due to the presence of D20 in commercially
available ND3 (makes and Throop, 1972).

The mass spectrum of 5,6,6-d3-5x-androst-16-en-3-one (X),
shown in Figure 7c, shows that the M+ and Mf-CH3 ions now
appear at m/e 275 and 260 as expected. Unfortunately, the
spectrum also shows a background ion at m/e 272 which was not
expected. The intensity of m/e 272 for this d3-form was
expected to be much smaller than m/e 272 in the dZ-form shown
in Figure 7b. It was thought that the additional deuterium
introduced at 0-5 in the formation of the d3-form would reduce
the background at m/e 272 by an amount proportional to the
percentage of deuterium added at C—5.

The intensity of m/e 272 taken together with the intensity
of m/e 287 suggested that the d3-form was chemically impure,
and that m/e 272 was a chemical impurity and not an isotopic
impurity. Thin layer chromatography purification removed both
the ions at m/e 272 and 287, and resulted in a d3-product with
a background (intensity of m/e 272 / m/e 275) of 8 parts per

thousand.

83

Perdeuterio 5d~Androst-16-en-§p—ol and SdpAndrost-16-en-3upol

 

The mass spectra of 519androst-16-enp3uhol (II) and
51$androst-16-en—2P-ol (III), shown in Figure Ba and 9a,
are different only in the intensity of the ions produced.

Both compounds yield relatively intense M+ ions (m/e 274) and
M+-CH3 ions (m/e 259) that would be good candidates for mass
spectrometric selected ion monitoring of these compounds in
gas chromatographic effluents.

SQ-Androst-16-en-3P-ol and SK-androst-16-en-3fl-ol both
undergo the loss of two hydrogens from their MI ions and M+-CH3
ions upon electron ionization. Therefore, it would be desirable
if the deuterium labeled internal standards contained three or
more deuteriums. Otherwise, the measurement of the intensity
of m/e 274 and 259 (M+ and M+-CH3 for the unlabeled steroids)
would be complicated because of contributions to those masses
from the dZ-labeled internal standard.

A convenient way to prepare a d3dversion of subandrost-
16-en-zp-ol would have been to reduce the previously prepared
6,6-d2-5dpandrost—16-en-3-one (VIII) with LiAlD4 as shown below:

0” _ ‘—+ H 3 -
LiAlD4 D
D D n D

VIII XI

.s madman on ccmmma on» Ca cm>fiw mam m:o«pfivcoo mafiumamqo

Houem-ceumfiuenoaecauem-eeuaa.ma.o.m Ah ”Houemudeumauenoaecanem an

.HOIsmncmlmHIpmosvc<nsm go mEsom Esfinmusmo mam Esfipopm mo manomam mmmz .m madman

85

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

q vq ‘0 lN
l J J J l l J J
q f °
5 - + 5
z ————-j*_ Z
8 .
N O
H In
¢ eCV
N
O
O
N
L <3
__ g m t l
H
O
p O
; H i
m : O ,0
_ L0
I I r T r I I
O O O 3 O O O O O O
2 (D (D N 52 CD CD \T N

(z)

ALISNHINI HAIIVWHH

100 150 200 250

50-

IWW//€

 

 

86

.n mpswfim 0» @cmme on» Ca cm>fiw mam mCOfiancoo mcfiumsmao

Houmm-cmnmfi-ehoaecauemussuafl.ma.m.o Ah ”Hoummucmuwa-enoaec<uem

.Hoummncmumfinpmosvc¢usm mo mEsom Esfiamusmo vcm Esfipomm mo manomam mmmz

Am

.0 obzmfim

8?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

q \T w
:1
ZN 2“
O)
2 o
H to __
st N I-
N
O
O
(\J
0
;____—___LD
.0
:;e 2
:3 o
U’)

I I fl I I I ”'_
'O O O O O O O O O O
2 (I) L0 \1’ N 2 (I) CO \'I' (\I

(Z) ALISNEINI EMU/138

 

100 ISO 200 250

50

m/e

88

The product obtained from the above reaction, 3¢,6,6-d3-
Sd-androst-té-en-Sp-ol (XI), would be expected to be of high
isotopic purity since Tbkés and Throop (1972) reported that
the isotOpic purity of products from LiAth reduction are
usually equivalent to that of the LiAlD4 (99 atom f D).

While the above procedure would have been convenient
for preparing the zp-alcohol, the Su-epimer could not be
made by this procedure since little or none of the Sx-alcohol
can be obtained by reducing 3-keto—5fl-steroids with LiAlD4
(Wheeler and Wheeler, 1972). However, as much as 92%»of the
BQPalcohol can be obtained by reduction of 3-keto-desteroids
with iridium complex catalysts (Haddad g§H§;.. 1964). Although
there are no reports describing the introduction of deuterium
during this reaction, it seems possible that this could be
accomplished if the reaction were performed in isopropanol-OD
and D20 instead of protio-solvents. However, this was not
attempted.

An alternative approach was considered which involved the
use of 5,6,6-d3-5upandrost-16-enp3-one (X) as starting*material
instead of the dZ-form (VIII). The use of d3 starting material
would eliminate the need to incorporate deuterium.during the
reduction with iridium complex catalyst. For example, the 5,6,
6-d3-5dpandrost-16-en-3-one (X) prepared earlier could be
reduced with LiAlH4 to obtain 5,6,6—d3-51pandrost-16-enPQp-ol
(XII), while reduction with the iridium complex could be used
to obtain 5,6,6-d3-519androst-16-en934-ol (XIII) as shown on
the following page.

89

 

D
LiAlH4 D iridium
complex

HO XII

 

D D

Although the procedure just described seemed at first to
be a promising approach, the method was abandoned because of
the difficulties in making, handling and storing the large
quantities of ND3 required for the synthesis of 5,6,6-d3-5flr
androst-16-en—3-one (X). TBkés and Throop (1972) described
the preparation of ND3 by reaction of magnesium nitride with
D20, but this technique was not well suited for the production
of large amounts of ND3. Large scale preparation of ND3 by a
counter current exchange procedure has been described by
Chevallier and Jillian (1968), but this was not attempted.
Fétlzon and Gore (1966) used deuteriated propylamine instead
of ND3 and their results suggested that deuteriated prOpylamine
should be used in subsequent synthetic work of this type.

90

The technique finally selected for preparation of d-labeled
Sw-androst-16-en-ip-ol and Sarandrost-16-en-3e-ol involved the
reduction of d4-labeled Sdpandrost-16-en—3-one (XIV). The 2;-
alcohol (XV) was prepared from the corresponding 3-ketone (XIV)
by reduction with LiAlH4, while the 3dpalcohol (XVI) was prepared
from similar material by reduction with iridium complex catalyst
as shown below. The preparation of the d4-labeled starting
material (XIV) will be discussed later.

D
‘1,

I XIV

0" s

D
1

H
D
LiAlH4 ridium
complex
D
' D
d” ‘

XVI

.D

so“
HO xv
n n

II“!

D D

The mass spectra of the perdeuterio Bdrandrost-16-enp3’bol
(XV) and 3urol (XVI) are shown in Figure 9b and 8b. respectively.

An important feature in the mass spectrum of both compounds is

91

the absence of a background ion at m/e 274. The very low
intensity of this ion suggests that both compounds would be
good internal standards for mass spectrometric isotopic

dilution studies of thesetwo important C19-A16

-steroids.

While both d4-labeled standards contain isotopic
impurities (d1, m/e 275; d2, m/e 276; d3, m/e 277). these
should not interfere with reverse isotopic dilution analyses.
The dB-impurity probably originated from low deuterium incor-
poration at C-17, which is discussed later in connection with
the synthesis of 16,17-d2-5,16-androstadien-3p-ol. The appear-
ance of an ion at m/e 266 in the mass spectrum of the d4-labeled
Bp-alcohol (Figure 9b) is curious because it represents the
unlikely loss of 14 amu's from the M+ ion, which suggests that
it originated from a chemical impurity and that the steroid
should be purified further. The ion at m/e 280 was also

unexpected, and may originate from a chemical impurity.

16,17-d2-5,16-Androstadien-QF-ol

An effort was made to prepare this compound to be used as
an internal standard and carrier in the assay of 5.16-androsta-
dien-Bf-ol, an important intermediate in boar testes (Saat_g§
al., 1972) and in human testes (Bicknell and Gower, 1971), and
one of the most abundant steroids in boar testes (Ruokonen and
Vihko, 1974a). The procedure was analogous to the one used by
Matthews (1968) to make 5.16-androstadien-zf-ol except that a
d2-labeled 17-keto steroid (XVII) was used as starting material

 

BO

92
and D20 was added in the final step instead of 320. The

sequence of reactions are shown below:

 

 

.0
‘ D beH-Ts
D D
‘D
1. Reflux-CH3OD
HO 2 R fl CH 03-.
’°‘”‘"3 HO
+ tosylhydrazine
XVII XVIII
._-CH3Li
‘ D
G” d”
D20
——_*
HO
XIX 'J XI

The dZ-starting material (XVII) was prepared by enolic
exchange in.methanol-OD. This exchange was purposely done
without the aid of a base catalyst so that the hydrazone
derivative (XVIII) could be formed simply by addition of
tosylhydrazine to the methanol-OD solution of 17-keto steroid.

93

Decomposition of the hydrazone (XVIII) with CH3Li yielded

the stabilized anion complex (XIX), and addition of D O

2
resulted in the introduction of one deuterium at 0-17
(Shapiro and Heath, 1967).

A 53% yield of dideuterio-5,16-androstadien-2P-ol(XX)
was obtained, but the isotOpic purity of this material was
disappointingly low as can be seen by comparison of its
mass spectrum (Figure 10a)with that of the protium form
(Figure 10b). The deuterium labeled product (XX) contained
as much as 45% of a d1-impurity (m/e 275) and 4% of a do-
impurity (m/e 274).

The low isotopic purity was not caused by incomplete
deuterium exchange at C-16 of the 17-keto starting material
(XVII). This was ruled out after comparison of its mass
spectrum (Figure 11a) and the spectrum of the protium form
(Figure 11b). For example, the spectrum of the labeled form
shows that two deuteriums were added because the MI, M+-CH3,
Mfiazo and M+-(CH3+H20) ions appear 2 amu's higher in the
deuterium labeled form. The possibility that only one
deuterium was attached to C-16 and that the other one was
on the hydroxyl oxygen was excluded because the M+AHZO ion
appears at m/e 272 (M+-18) instead of m/e 271 (MI-19). Also,
the #40113 + 320) ion is found at m/e 257 (Mt-33) instead
of m/e 256 (M+-34). Furthermore, comparison of other ions
in the spectra of the deuterium and protium forms reveals
mass differences of either 2 amu's or no mass differences,

a result consistent with both deuteriums being located on

94

.m mazwfim 0p snowma one CH cm>fiw mam macapfiucoo mcfipwsooo

HOIQmuooeosonosoosues.m an mHousmacoeosenoaocauma.mImoueH.eH as

.Hoummncmflwmpmosn:«nma.m do mEsom Ezfiposm Ucm Esfiampsma mo mppomam mmmz

.oa seamed

95

 

 

250

 

 

U TA—f '—

 

 

 

 

’l
200

i
150

 

 

 

 

 

100

 

 

 

 

 

 

 

 

100

60—

 

100
80a

AiI SNELLNI 3A1 1V"! 38

I
0
L0

r
O
\T

207

250

200

150

100

50

m/e

96

cm pm chLOUmL

.>o m.mm so Hosesoeoa massacoe

new emacfi mnopa posses one mH> omcfimppo one: mapomam mmmz

sconeHucopmosvcmnmlzxopwmnnmm an

scoumaucmpwopccmfimwuoa.mfiHumnmxosczcumm Am
.mcouwHucopmos@CMImazxopvmsnmm

no meson Ezfiuosm cam Esflpmpsmo go mapomam mmmz .HH ossmfim

97

 

 

 

 

 

M
2%

262

Ill

 

 

 

 

257

 

 

150 200 250

100

 

 

50
b

 

 

I I I I I l
O O O O O O
\T N (I) LO \T (\l

100

ALI SNHINI 3A1 MN 38

100 150 200 250

50

we

so“

98

one end of the molecule at C-16.

The view that both deuteriums were on C-16 is supported
by the results of other workers who prepared 16,16-d2-17-
keto-steroids in high isotOpic purity (89%-d2; 11fi—d1)
after only 4 hours exchange in methanol-OD/deuterioxide
Tokes g; a;., 1967). The absence of base catalyst in the
procedure described herein is counterbalanced by the longer
exchange period which was used (7 days).

The results of a separate experiment indicate that the
low isotOpic purity was due to competition between the ether
solvent and D20 for protionation (deuteriation) of the anion
intermediate (XIX). In this experiment the tosylhydrazone
derivative of 3a-androst-Sa-androstanp17-one (XXI) was treated
with.methyllithium and then D20 was added as shown below:

H.
-HH-Te ll JXXII

a—‘E;;;;J"

D ' H
m &‘ h“
+
4&% 565

 

 

99

The isotopic purity of the product was low (Figure 12a;
44%—dj and 56%-do) which supported the suspicion that the
low isotopic purity observed earlier for 16,17—d2-5,16-androsta-

dien-QP-ol (XX) was the result of competition between D O and

2
ether, and not the result of incomplete enolic exchange at
C-16. Shapiro and Heath (1967) speculated about this compe-
tition in explaining the results of their experhmente in
which they obtained 17-d1-16-androstene with an isotopic
purity of 60%—d1 and «10%-do using diethyl ether as solvent.
By contrast, Fischer gt,a;. (1965) and Takes and Djerassi
(1969) obtained 89%--d1 and 931-d1, respectively, when a similar
reaction was performed with dioxane as solvent. These obserb
vations suggest that subsequent labeling experiments of this
type should be performed in dioxane solvent instead of dimethoxy-
ethane or diethyl ether in order to achieve the highest possible
isotopic purity.

Even a dz-form of 5,16-androstadienszpol of high isotopic
purity would not be a convenient internal standard because of
its tendency to give MI-Z ions upon.mass spectral analysis.
The MI-Z ion can best be seen in the spectrum of the pure
unlabeled form shown in Figure 101:. The H+-2 ion arising from
the d2-labeled internal standard would contribute to the variable
background which must be substracted from.measurements made
of the MI ion of the protium form.

The 16,17-d2-5,16-androstadien~3,—ol (XX) could be trans-
formed into a d3, d4, d5 or d7-form by procedures used by

100

.n madman on snowma one :H sm>fiw mam msoapavcoo wcfipmsmqo
Honsmucmnmanumoanc<usm An mHonsmIsolmaupmopcc¢nsmnaonwa Am
.HOIsmIchmHIpmopcc<Ism

mo mason Esfiuosm was cmpmsmusmnocoz mo snuomqm mmmz .ma maswfim

101

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100

 

W \T W
p e T
+ N + (r: ..
o r
L0
(V 3 ‘
N
O
O
N
O :
a —_e
O E:_:
- E3 if:
m C) .9
L0
I I I I I I I I
O O O O O O O O O
(I) LO \T N 2 CD (0 \T N
(Z) All SNHINI 3A1 1V7 38

100 150 200 250

50

m/e

102

Bjorkhem gtflal. (1973,1974) for preparation of otherlis-IP-ol
steroids with deuterium label on 0-2, C-3 and 0-4. The proposed

scheme of reactions is shown below:

D
‘1’

HO Oxidation

OI nauer I
PPG 3
0”

XX XXIV

 

 

103

The first step involves the conversion of the AP-zp-ol
steroid (XX) to the corresponding 44-3-one (XXIV) by Oppenauer
oxidation, described later herein. Deconjugation of the A4-
3—one by equilibration in t-butyl alcohol-OD and potassium
t-butoxide, followed by acetic acid-GOOD protonation
(deuteriation) of the anion would yield the perdeuterio
(AP-B-one (Ringold and Malhotra, 1962). This material.(XXV)
could be treated immediately with LiAlD4 which would result
in the AS-BP-ol (XXVI) with as many as 7 deuteriums. Alter-
natively, a dB-form (3«,16,17-d3), a d4-form (4.4.16.17-d4)
or a d5-form (3d,4,4,16,17-ds) could be prepared by variations
of this procedure as shown by ijrkhein 23,3l. (1974) in work
with 2p-hydroxy-5-androsten-17—one. Perdeuterio-5,16-androsta-
dien—if—ol made in this way would be a useful internal standard
and carrier in the assay of 5,16-androstadien-2fbol in the

tissues and fluids of pigs and humane.

16,17-dg-4,16-Androstadien-3-one

 

The dz-form of 4,16-androstadien-3-one (XXIV) was prepared
from 16,17-dZ-5,16-androstadien-3f-ol (XX) by Oppenauer
oxidation performed according to the procedure described by

Matthews (1968).
D
ID
‘D
‘D

enauer

Oxidation
XX

so 0: XXIV

104

The reaction gave an 88% yield (5.3 g) of d-labeled
4,16-androstadien-3-one (XXIV), but no plans were made to
use this material as an internal standard/carrier after
the extent of its isotopic impurity was revealed by the
results of mass spectral analysis shown in Figure 13a and b.

Unfortunately, this synthetic experiment was performed
before checking the isotopic purity of the starting material
(XX), and as a result, no useable material was obtained for
use as a carrier in the assay of 4,16-androstadienPB-one.
However, the product was valuable as starting material for
the preparation of other deuterium labeled standards,
including Sd—androst-16-en-3-one (XIV), Sdpandrost-16-enp
Ifbol (XV) and Se-androst-16-en-3d—ol (XVI).

The dO and d1 impurities seen in the mass spectrum
of 4,16-androstadien-3—one (Figure 13a) were a direct
result of the low isotOpic purity observable in the mass
spectrum of the starting material (XX) shown in Figure 10a.
It is now known that a more pure d-labeled 5,16-androstadienp
BfLol (XX) can be prepared with only minor modifications

of the methods, as discussed earlier herein.

 

105

.w mssmfim Op Ucmwma on» Ca cm>flm mam mcofipwvcoo weapmsmao

mcoumucmfivmpmoschuwfi.z An mmcoumncmfiumpmopoc<lwa.zumnlma.ma Am

.mCOImucmHGMpmoswcmumH.q mo mfieom Esfluoem new Esfismpsmo mo mepomam mmmz

.ma msswfim

106

tLL pt

com

 

 

 

 

 

m\E
o

 

m

H

are ; gees; i€_:
_

 

 

OOH

 

 

 

 

 

om

 

 

 

+2 rmN

.Z_ :37
r _:

 

 

 

 

OOH

ON
row
row
row

 

1&2

All SNHINI 3A1 .LV'IHH

(z)

107

6,6,16,17—d4-SdrAndrost-16-en-3-one

 

The steps used to prepare this material are shown below:

 

   

 

 

D n
G D f D
D
' fl 3 d s D
I re ux- ay +

O’ O:

‘ CH3OD + DZO/Na

D D D
XXIV XXVII
Li/NH3
D
I
D
G D
D lllll|
D
reflux-3 hrs

0’ D)

E 03303 + HZO/Na o: :

D D D H
D D XIV

XXVIII

The activated hydrogens in 16,17-d2-androstadienp3-one
(XXIV) were exchanged with deuterium by equilibration in
methanol-OD/DZO. The product (XXVII) was reacted with a
solution of lithium metal in liquid ammonia which reduced
the A4-double bond and yielded 2,2.4.6,6,16,17—d7-5Il-
Iandrost-16-enp3-one (XXVIII). The exchangeable deuteriums
at 0-2 and 0-4 were removed by back exchange in methanol-OH/
H20. The final product, 6,6 ,16,17-d4-5¢(-androst-1 6-en-3-one
CXIV), was chromatographed on.a column of silver nitrate

108

impregnated silica gel and crystallized from acetone-hexane.
The procedure gave 0.7 g or 357‘ yield of the deuterated
steroid. The mass spectra of the deuterium and protium
forms of Set-androst-16-en-3-one are shown in Figure 149.
and b. The absence of an ion at m/e 272 (M9 for protium
form) in the bar graph for the deuterium form is a good
indication that this material would be very useful as a
carrier and internal standard for the assay of Sat-Fandrost-
16-en-3-one by mass spectrometric reverse isotOpe dilution.
Time did not permit analysis of the fat samples using
all of the labeled steroids. However, the labeled compounds

are now available for future studies.

109

.n madman on ccowoa on» Ca cm>fiw msoaufivcoo wcfipmamao

mcoumucmnwfinumoauz<uom An mocclmIcmumauumoaoc<usmnzelwH.oaemem Am

ecoumncmumfiuumosvcausm mo masom Esfiuoem Ucm anaemusmo mo mauomqm mmmz

.za mssmfim

110

 

®\E

 

 

0mm com 02 OOH om

 

 

 

 

 

 

 

 

 

 

 

 

i_4
_ ,2
10¢
New .m.uo row
.2
co
emu
0mm OON
_m_ i 2.. _ a
_l_
mPN
+7— low
new low

 

 

 

 

OOH

All SNHINI 3A1 .LVIHH

(Z)

SUMMARY

A reverse isotope dilution/carrier technique was adapted
for determination of’quandrost-16-en-3-one. After the
addition of a deuterium labeled form of’Supandrost-16-en-3-
one to the fat sample, the mixture of deuterium and protium
forms was purified and the resultant fraction was analyzed
by combined gas chromatography-mass spectrometry. The
deuterium/protium ratios were obtained by multiple ion
detection using the accelerating voltage alternator acces-
sory of an LED 9000 gas chromatographsmass spectrometer.

The addition of a labeled form of the steroid enabled
extremely small amounts of the compound to be partially
purified and separated by gas chromatography without total
loss by adsorption. Losses during the isolation were auto-
matically accounted for since the ratio of the labeled and
unlabeled species would remain constant during isolation
and analysis. The amount of unlabeled 5dkandrost-16-en—
3-one was determined from the isotOpe ratio resulting after
the addition of a known amount of the labeled species.

This procedure was shown to be useful in analyzing for fir-
androst-16-enp3-one in fat samples.

Deuterium labeled forms of the five most common 019-
1:6 steroids were synthesized to provide standards for
quantitative analysis for these compounds in the pig. The

labeled C19-¢§6steroids that were synthesized include

111

112

the following compounds: d2-, d3-, and d4-5qkandrost-16-
en-3-one; d4-5qkandrost-16-en-3q-ol; d4-5x-androst-16-
ensawbol; d2-4,16-androstadienp3—one and d2-5,16-androsta-
dienrzp-ol.

Several fat samples from cryptorchid (males with retained
testicles) pigs were analyzed for sueandrost-ié-en-S-one by
stable isotope dilution and the levels were correlated with
odor intensity scores by both a trained laboratory panel
and a meat industry panel. The results for the meat industry
panel were not statistically significant. Although the
trained laboratory panel scores were significantly (P<LOS)
correlated with the levels of Surandrost-16-enp3-one, this
only accounted for 25% of the variation in scores. This
suggests that Subandrost-16-en-3-one is not the only compound
contributing to the undesirable odor in boar fat. Further
support for this viewpoint is evident by the fact that several
fat samples having high odor intensity scores were found on
analysis to have only low levels of’Su-andrcst-16-en-3-one.
It seems likely that the other 019-zgesteroids may account
for some of the discrepancies in the results.

Although the labeled 019-4fi5 steroid standards were
synthesized and are now available in our laboratory, time
did not permit analysis. These labeled standard compounds
will be useful in subsequent studies on the significance
of the 019-416 steroids in both pigs and human beings.

BIBLIOGRAPHY

BIBLIOGRAPHY

Ahmad, N. and D. B. Gower. 1968. The biosynthesis of some
androst-16-enes from C-21 and 0-19 steroids in boar
testicular and adrenal tissue. Biochem. J. 108:233.

Berry, K. E., J. D. Sink, S. Patton and J; H. Ziegler. 1971.
Characterization of swine sex odor components in boar
fat volatiles. J. Food Sci. 36:1086.

Beets, M. G. 1962. Some aspects of the odor problem. Perfum.
Cosmet. Savons 5:167.

Beets, M. G. 1971. Olfactory response and molecular structure.
In Handb o Senso o o p. 322. ed. L. H.
Beidler, Springer, Berlin.

Beynon, J. H. 1960. Mass Spectrometry and its Applications

tg Organic Chemistry, p. 67. Elsevier Publishing Co.
New, N. Y.

Bicknell, D. C. and D. B. Gower. 1971. Studies of 0,9 16-
unsaturated steroid excretion and formation in cases of
testicular feminization. J. Endocr. 49:264.

Bieber, M. A., C. C. Sweeley, D. J. Faulkner and M. R.
Petersen. 1972. Purification and quantitative determine
ation of Oecropia juvenile hormone. Anal. Biochem.

47:264.

113

114

Bjarkhem, I., R. Blomstrand and L. Svensson. 1974. Serum
cholesterol determination by mass fragmentography.
Clin. Chim. Acta 54:185.

ijrkhem, I., J. A. Gustafsson and J. Sjovall. 1973. A
novel fragmentation of trimethylsilyl ethers of 5p-
hydroxy-AP-steroids. Org. Mass Spec. 7:277.

Booth, W. D. 1972. The occurrence of testosterone and
Sx-dihydrotestosterone in the submaxillary salivary
gland of the boar. J. Bndocr. 55:119.

Brooksbank, B. W. L. 1962. Urinary excretion of androst-
16-en-3d-ol. Levels in normal subjects and effects of
treatment with trophic hormones. J. Endocr. 24:435.

Brooksbank, B. W. L., A. B. Cunningham and D. A. Wilson.
1969. The detection of androsta-4,16-dien-3-one in
peripheral plasma of adult men. Steroids 13:29.

Brooksbank, B. W. L. and G. A. D. Haslewood. 1950. The
nature of pregnanediol-like glucuronide. Biochem. J.
47:36.

Brooksbank, B. W. L. and G. A. D. Haslewood. 1952. The
isolation of androst-16-en-3xpol from women's urine.
Biochem. J. 51:286.

Brooksbank, B. W. L., D. A. A. Wilson and J.-A. Gustafsson.
1972. The metabolism in vivo of 4,16—androstadien-3-
one in man. Steroids Lipids Res. 3:263.

BrOphy, P. J. and D. B. Gower. 1972. 16-unsaturated
019-3-oxosteroids as metabolic intermediates in boar
testis. Biochem. J. 128:945.

115

Bunnel, R. H. 1967. Vitamin E assay by chemical methods.
In The Vitamins, vol. 6, eds. Gyorgy, P. and W. N.
Pearson, p. 97. Academic Press, New York, N. Y.

Campbell, J. A. 1972. Personal communication.

Chevallier, M. and J. Jullien. 1968. Countercurrent
exchange preparation of deuterated ammonia. Bull.
Soc. Chim. France. p. 4937.

Claus, K., B. Hoffmann and H. Karg. 1971. Determination
of Surandrost-16-en-3—one, a boar taint steroid in
pigs, with reference to relationships to testosterone.
J. Anim. Sci. 33:1293.

Cleveland, W. W. and X. Savard. 1964. Excretion of androst-
16-enp3u-ol. J. Clin. Endocr. Metab. 24:983.

Comfort, A. 1971a. Communication may be odorous. New
Scientist 49:412.

Comfort, A. 1971b. Likelihood of human pheromones. Nature
230:432.

Craig, H. B. and A; M. Pearson. 1959. Some preliminary
studies on sex odor in pork. J. Anim. Sci. 18:557.

Craig, H. B., A. M. Pearson and N. B. Webb. 1962. Fraction-
ation of the components responsible for sex odor/flavor

in pork. J. Food Sci. 27:29.

DeDuca, H. 1974. Personal communication.

116

Djerassi, C. and L. Tokes. 1966. Mass spectrometry in
structural and stereochemical problems. XCIII.
Further observations on the importance of interatomic
distance in the McLafferty rearrangement. Synthesis
and fragmentation behavior of deuteriumplabeled 12-
keto steroids. J. Am. Chem. Soc. 88:536.

Dryden, H. L., Jr. 1972. Reductions of steroids by metal-
ammonia solutions. In Organic Reactions in Stgygig
Chemistry, vol. 1, eds. Fried, J. and J. A. Edwards,
Van Nostrand Reinhold Company, New York, N. Y.

Eik-Nes, K. B. 1970. The Androgggs of yhe Tegyig.
M. Dekker Publishers, New York, N. Y.

Fetizon, M. and J. Gore. 1966. Specific deuteration ,to
a keto group: Reduction of 05/8 -ethylenic ketones with
lithium and deuterated propylamine. Tetrahedron
Letters 5:471.

Fischer, M., Z. Pelah, D. R. Williams und C. Djerassi. 1965.
Mechanismus der reduktion.von tosylhydrazonen.mit
komplexen metallhydriden. Chem. Ber. 98:3236.

Fuchs, G. 1972. The correlation between the 5x-androst-16-
ene-3-one content and sex odour intensity in bear fat.
Swedish Je Agr. R68. 1:233e

Gordon, A. E. and A. Frigerio. 1972. Mass fragmentography
as an application of gas-liquid chromatography—mass
spectrometry in biological research. J. Chromatogr.
73:401.

Gower, D. B. 1972. 16-unsaturated C19 steroids. A review
of their chemistry, biochemistry and possible physio-
logical role. J. Steroid Biochem. 3:45.

117

Gower, D. B. and N. Ahmad. 1967. Studies on the biosynthesis
of 16-dehydrosteroids. The metabolism of 4-D4C pregnen-
olone by boar adrenal and testis tissue igyyiyyg.
Biochem. J. 104:550.

Gower, D. B., F. A. Harrison and R. B. Heap. 1970. The
identification of 019-16-unsaturated steroids and
estimation of 17-oxosteroids in boar spermatic vein
plasma and urine. J. Endocr. 47:357.

Gower, D. B. and T. Katkov. 1969. Studies on the biosynthesis
and excretion of C19-zg6-steroids in the boar. In Heat

Production from Entire Male Animals, p. 261, ed. Rhodes,
D. N., J & A Churchill LTD. London.

Gower, D. B. and X. H. Loke. 1971. Studies on the subcellular
location and stability of the enzyme system.involved in
the biosynthesis of 5,16-androstadien-3f-ol from 3,-
hydroxy-S-pregnen-ZO-one (pregnenolone). Biochim.
Biophys. Acta 250:614.

Gower, D. B. and M. I. Stern. 1969. Steroid excretion and
biosynthesis with special reference to androst-16-enes
in a woman with a virilising adrenocortical carcinoma.
Acta Endocr. (th). 60:265.

Griffiths, N. M. and R. L. 3. Patterson. 1970. Human
olfactory responses to Srhandrost-16-enp3-one - Principal
component of boar taint. J. Sci. Fd. Agr. 21:4.

Gustafsson, JAA. 1973. The formation of 16-17 dihydroxylated
C19 steroids from 16-dehydro C19 steroids in liver micro—
somes from male and female rats. Biochim. Biophys. Acta
296:179.

118

Haddad, Y. M. Y., H. B. Henbest, J. Husbands and T. R. B.
Mitchell. 1964. Reduction of cyclohexanones to axial
alcohols via iridiumpcontaining catalysts. Proc. Chem.
Soc. p. 361.

Holland, J. F., C. C. Sweeley, R. E. Thrush, R. E. Tests,
and M. A. Bieber. 1973. On-line computer controlled
multiple ion detection in combined gas chromatography-
mass spectrometry. Anal. Chem. 45:308.

Horning, E. C., W. J. A. Vanden Heuvel and B. G. Creech.
1963. Separation and determination of steroids by

gas chromatography- Chapter 2. ‘In W
Anglygig, vol. 11. ed. Glick, D. Interscience, New York.

Xatkov, T. 1971. Studies on 019-116-steroids. Ph.D. Thesis,
University of London, London, England.

Katkov, T., W. D. Booth and D. B. Gower. 1972. The metab-
olism of 16-androstenes in boar salivary glands.
Biochim. Biophys. Acta 270:546.

Xatkov, T. and D. B. Gower. 1968. The biosynthesis of 5d-
androst-16-en-3-one from progesterone by boar testis
homogenate. Biochim. Biophys. Acta 164:134.

Katkov, T. and D. B. Gower. 1970. The biosynthesis of
androst-16-enes in bear testis tissue. Biochem. J.
117:533.

Xloek, J. 1961. The smell of some steroid sex-hormones
and their metabolites. Psychiat. Neurol. Neurochir.
64:309.

Lerche. 1936. Geschlechtsgeruch bei eberkastraten.
Z. fur Fleisch-und Milchhygiene 46:417.

119

Loke, K. H. and D. B. Gower. 1971. Further studies on the
biosynthesis of androsta-5,16-dien-3p-ol and the sub-
cellular location of the site of biosynthesis. Biochem.
J. 122:27p.

Loke, X. H. and D. B. Gower. 1972. The intermediary role
of 5-pregnene-3,,20II-diol in the biosynthesis of 16-
unsaturated C19 steroids in boar testis. Biochem. J.
127:545.

Mareno, J. M. M. and A. V. Roncero. 1964. The applications
of urea inclusion compounds in fat analysis. In.Aggly§ig

and Chgrgcterization of OilsI Fats and Fat Products. ed.
Boekenoogen, H. A. Interscience, New York, N. Y.

Matsui, M. and D. K. Fukushima. 1970. Studies on the
heterolytic fragmentation of pregnane-16,20—diol
derivatives to androst—16~enes. J. Org. Chem. 35:561.

Matthews, G. J. 1968. The reduction of )r-iodo azides. A
stereospecific synthesis of aziridines. Ph.D. Thesis,
University of Colorado.

Matthews, G. J. and A. Hassner. 1972. Synthesis of oxiranes,
aziridines and episulfides. In W
Steroid Chemisgy, vol II, eds. Fried, J. and J. A.
Edwards, p. 32, Van Nostrand Reinhold Co., New York.

National Provisioner. Editorial. Oct. 14, 1967. 157(16):7.

Newell, J. A., L. H. Tucker, G. C. Stinson and J. P. Bowland.
1973. Influence of late castration and diethylstil-
besterol implantation on performance of boars and on
incidence of boar taint. Can. J. Anim. Sci. 53:205.

120

Patterson, R. L. S. 1968. 5d-Androst-16-en-3-one: Compound
responsible for taint in boar fat. J. Sci. Food Agr.
19:37.

Patterson, R. L. S. 1969. Boar taint: Its chemical nature
and estimation. In Meat Production from Entire Male
Animals, p. 247, ed. Rhodes, D. N., JlA Churchill LTD.
London.

Pearson, A. M. 1972. Personal communication.

Pearson, A. M., S. Ngoody, J. F. Price and H. E. Larzelere.
1971. Panel acceptability of products containing boar
meat. J. Anim. Sci. 33:26.

Pearson, A. M., R. H. Thompson and J. F. Price. 1969.
Sex odor in pork. Proc. Meat Ind. Res. Conf. Am.
Meat Inst. Found., Chicago, p. 145.

Prelog, V., L. Ruzicka and P. Wieland. 1944. Steroids and
sexualhormone. Uber die herstellung der beid en
moschusartig riechenden.(gs-androstenole-(3). Helv.
Chim. Acta 27:66..

Prelog, V. and L. Ruzicka. 1944. Untersuchungen uber
organextrakte. Uber zwei moschusartig riechende
steroids aus schweinetestes-extrakten. Helv. Chim.
Acta 27:61.

Prelog, V., L. Ruzicka, P. Meister and P. Wieland. 1945.
Steroids und sexualhormone. Untersuchungen uber den
zuzammenhang zwischen konstitution und gerush bei
steroiden. Helv. Chim. Acta 28:618.

121

Radt, F. 1959. Editor of Elsevisr's Encyclopaedia of
Organic Chemistry, Series III, Vol 14-supplement,
2396s et seq. SpringeraVerlag, Berlin.

Rhodes, D. N. and R. L. S. Patterson. 1971. Effects of
partial castration on growth and the incidence of boar
taint in the pig. J. Sci. Fd. Agr. 22:320.

Ringold, H. J. and S. K. Malhotra. 1962. Deconjugation of
c(,f-unsaturated ketones. Tetrahedron Letters 15:669.

Ruokonen, A. 1973. Free and sulphate-conjugated 16duneat-
urated C‘9 steroids in human testis tissue. Biochim.
Biophys. Acta 316:251.

Ruokonen, A. and R. Vihko. 1974a. Steroid metabolism in
testis tissue: Concentrations of uncondugated and
sulfatsd neutral steroids in boar testis. J. Steroid
Biochem. 5:33.

Ruokonen, A. and R. Vihko. 1974b. Steroid metabolism in
human and bear testis tissue. Steroid concentrations
and the position of the sulfate group in steroid
sulfates. Steroids 23:1.

Ruzicka, L. 1926a. Constitution of civetone. Helv. Chim.
Acta 9:230.

Ruzicka, L. 1926b. Constitution of muscone. Helv. Chim.
Acta 9:715.

Ruzicka, L. 1926c. Further considerations on the constitution
of muscone. Helv. Chim. Acta 9:1008.

122

Seat, Y. A., D. B. Gower, F. A. Harrison and R. B. Heap.
1972. Studies on the biosynthesis in viyo and excretion
of 165unsaturated C19 steroids in the boar. Biochem. J.
1293657.

Samuelsson, B., M. Hamberg and C. C. Sweeley. 1970.
Quantitative gas chromatography of prostaglandin E
at the nanogram level. Anal. Biochom. 38:301.

1

Self, H. L. 1957. The problem of pork odor. Proc. of Am.
Meat Inst. Found. Res. Conf. 9:53.

Shapiro, R. H. and M. J. Heath. 1967. Tosylhydrazones. V.
Reaction of tosylhydrazonss with alkyllithium.rsagents.
A new olefin synthesis. J. Am. Chem. Soc. 89:5734.

Shapiro, R. H., D. H. Williams, H. Budzikiswicz and C.
Djerassi. 1964. Mass spectrometry in structural and
stereochsmical problems. LIII. Fragmentation and
hydrogen transfer reactions of a typical 3-keto steroid,
Supandrostan-3-one. J. Am. Chem. Soc. 86:2837.

Sink, J. D. 1967. Theoretical aspects of sex odor in
swine. J. Theorst. Biol. 17:174.

Sink, J. D. 1973. Lipid-soluble components of meat flavors/
odors and their biochemical origin. J. Am. 011 Chem.
Soc. 50:470.

Stinson, C. G., L. H. Tucker, G. M. Weiss and A. M. Martin.
1972. Bear meat: Tests for taint and consumer response.
Proc. 18th Meeting of Meat Res. Workers, Uhiv. of Guelph,
Guelph, Ontario Canada, p. 253.

123

Sweeley, C. C., W. H. Elliott, I. Fries and R. Ryhage. 1966.

Mass spectrometric determination of unresolved components
in gas chromatographic effluents. Anal. Chem. 38:1549.

Thompson, R. H., Jr., A. M. Pearson and X. A. Banks. 1972.
019-1g6 steroids contributing to sex odor in pork. J.

Agr. Food Chem. 20:185.

Tokes, L. and C. Djerassi. 1969. Mass spectrometry in
structural and stereochsmical problems. CLXXVI. The

course of the electron impact induced fragmentation of
androstane. J. Am. Chem. Soc. 91:5017.

Tokes, L., R. T. LaLonde and C. Djerassi. 1967. Mass
spectrometry in structural and stereochsmical problems.
CXXVI. Synthesis and fragmentation behavior of deuterium-

labeled 17-keto steroids. J. Org. Chem. 32:1012.

Tokes, L. and L. T. Throop. 1972. Introduction of deuterium
into the steroid system. In W
W, vol. 1, eds. Fried, J. and J. A. Edwards,

p. 164. Van Nostrand Reinhold Company, New York, N. Y.

U. S. D. A. 1968. Meat Inspection. Disposition of swine
carcasses with sexual odor. Fed. Reg. 33:10577.

Meat and Poultry Inspection Program.

U. S. D. A. 1973.
Sexual

Animal and Plant Health Inspection Service.
odor of swine. M-311.20.

Vandsnheuvel, F. A. 1967. Recovery of 0.5-1.0 pg amounts
of sterols and steroids from thin-layer chromatographic

plates. J. Lab. Clin. Med. 69:343.

124

Wheeler, D. M. S. and M. M. Wheeler. 1972. Reductions of
steroidal ketones. In W
m, vol. 1, eds. Fried, J. and.J. A. Edwards,
p. 91. Van Nostrand Reinhold Company, New York, N. Y.

Williams, L. D. and A. M. Pearson. 1965. Unsaponifiable

fraction of pork fat as related to boar odor. J. Agr.
Food Chem. 13:573.

Williams, L. D., A. M. Pearson and N. B. Webb. 1963.

Incidence of sex odor in boars, sows, barrows and
gilts. J. Anim. Sci. 22:166.

CH IGRN STRTE UNIV.

D1I|I3I| II III104