A genetic approach to the regulation of the constitutive forms of cytochrome P-450

MISSING IMAGE

Material Information

Title:
A genetic approach to the regulation of the constitutive forms of cytochrome P-450
Uncontrolled:
Cytochrome P-450
Physical Description:
viii, 156 leaves : ill., graphs ; 29 cm.
Language:
English
Creator:
Hawke, Roy Lee, 1951-
Publication Date:

Subjects

Subjects / Keywords:
Cytochrome P-450 Enzyme System   ( mesh )
Microsomes, Liver   ( mesh )
Testosterone -- metabolism   ( mesh )
Pharmacology and Therapeutics thesis Ph.D   ( mesh )
Dissertations, Academic -- Pharmacology and Therapeutics -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida.
Bibliography:
Bibliography: leaves 147-154.
Statement of Responsibility:
by Roy Lee Hawke.
General Note:
Photocopy of typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000373234
oclc - 10187564
notis - ACB2408
System ID:
AA00004897:00001

Full Text













A GENETIC APPROACH TO THE REGULATION OF THE
CONSTITUTIVE FORMS OF CYTOCHROME P-450





By



ROY LEE HAWKE


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1983

























TO LYNNE TWIEST


This dissertation is dedicated to my wife, Lynne Twiest, for her

support, patience, and sacrifice which were instrumental in the

attainment of this degree, and to my daughters Dawne and Heather who I

regret have had to make their own special sacrifices during the many

years of my education.














ACKNOWLEDGEMENTS

I would like to extend sincere thanks and appreciation to my

dissertation committee members, and to all of the faculty members and

support staff of the Department of Pharmacology who have helped me in

innumerable ways over the last four years.

I wish to especially acknowledge the heroic efforts of both Debbie

Baylor in the typing of these manuscripts as well as other "rush" jobs

which were often thrust upon her, and Jeff Jones who was forever asked

to make slides on a moment's notice or less.

I would like to acknowledge the expert technical assistance

provided by Lynn Raynor over the course of these studies. Her endless

effort in keeping the laboratory functioning smoothly has made things

easier on us all. I am truly grateful for the support and friendship

she has extended toward me over the years. The meticulous efforts of

my sister, Becky, who spent her summers working in our lab, are also

gratefully acknowledged.

Finally, I wish to express my gratitude to Allen Neims whose

greatest accomplishment was not in making this mere dissertation

possible, but rather instilling in me a genuine fascination for science

and an excitement for research which I hope to carry with me until the

end of my career. He allowed our student-teacher relationship to

continuously evolve such that I now have the confidence to think out

loud. I will sorely miss our collaboration and the many hours of

discussion spent on topics ranging from Gators to P-450.

















PREFACE


This dissertation is composed of five chapters representing

different aspects of my doctoral research. Each chapter has been

written in a standard manuscript style with an introduction, a

materials and methods section, results and discussion. Some

unavoidable redundancy will, therefore, be encountered in the

introduction and discussion of each chapter as a consequence of this

style of presentation.















TABLE OF CONTENTS


PAGE


ACKNOWLEDGEMENTS . . .

PREFACE. .. .. .. . .

ABSTRACT .. . .

INTRODUCTION . . .. .

CHAPTER ONE HETEROGENEITY OF CYTOCHROME P-450 IN THE
SITE-SPECIFIC HYDROXYLATIONS OF STEROIDS
BY AKR/J MOUSE LIVER MICROSOMES ..


iii

iv

vii

1



5


Introduction . .
Materials and Methods .
Results . .
Discussion . .


CHAPTER TWO


CHAPTER THREE


CHAPTER FOUR


THE INFLUENCE OF STRAIN AND SEX ON SITE-
SPECIFIC TESTOSTERONE HYDROXYLASE
ACTIVITIES OF MOUSE HEPATIC MICROSOMES.

Introduction. . ..
Materials and Methods ....
Results . .
Discussion. . .


DEVELOPMENT OF SITE-SPECIFIC TESTOSTERONE
HYDROXYLASES IN HEPATIC MICROSOMES FROM
AKR/J AND BALB/cJ MICE. . .

Introduction. .. .. .
Materials and Methods .
Results . . .
Discussion . .

GENETIC REGULATION OF HEPATIC MICROSOMAL
TESTOSTERONE 15a- and 16a-HYDROXYLASE
ACTIVITIES IN FEMALE AKR/J AND BALB/cJ
MICE . . .

Introduction ... .. .......
Materials and Methods ..
Results . .
Discussion. . .


85
88
92
109


. .










TABLE OF CONTENTS--Continued


PAGE
CHAPTER FIVE ENDOCRINE FACTORS IN THE REGULATION OF
HEPATIC MICROSOMAL TESTOSTERONE 15a- AND
16a-HYDROXYLASE ACTIVITIES IN AKR/J AND
BALB/cJ MICE. . . 113

Introduction. . . 113
Materials and Methods . 117
Results . . .. 122
Discussion. .... ..... 139

GENERAL CONCLUSIONS. .. . ... 144

REFERENCES . . . 147

BIOGRAPHICAL SKETCH. . . .. 155













Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


A GENETIC APPROACH TO THE REGULATION OF THE CONSTITUTIVE
FORMS OF CYTOCHROME P-450

By

Roy Lee Hawke

August, 1983

Chairman: Allen H. Neims
Major Department: Pharmacology and Therapeutics

The genetic, hormonal, and developmental regulation of various

forms of the cytochromes P-450 in untreated isogenic strains of mice

has been examined by virtue of the remarkable position specificity

demonstrated by these constitutive isozymes in the hydroxylation of

testosterone. Careful kinetic analysis has established optimal assay

conditions for the measurement of hydroxylase activity in mouse hepatic

microsomes, and a new radiochromatographic procedure for the separation

and quantitation of six monohydroxylated products of testosterone is

described.

Detailed statistical analysis of results from studies on the

effects of sex, strain, and postnatal development on the rate of

hydroxylation at the 6a, 6B, 7a, 15a, 158 and 16a positions of

testosterone provided evidence for both heterogeneity and polymorphism

in the cytochromes P-450 catalyzing these hydroxylations and/or in

their modes of regulation. The most interesting findings from these










studies related to the 15a and 16c hydroxylations. With respect to

testosterone 16a-hydroxylase activity, 1) rates were 2- to 3-fold

higher in microsomes from males than females of all eight strains

studied; 2) prominent strain differences were detected only between

female mice; and 3) postnatal development was more rapid than the

postnatal increase in cytochrome P-450 content and the development of

other hydroxylase activities. With respect to testosterone

15a-hydroxylase activity, 1) sexual dimorphism was strain-dependent in

that rates in females were 2- to 3-fold higher than rates in males in

only four of eight strains; 2) strain differences detected between

males were independent of those found between females; and 3) a 10-day

lag in postnatal development, which was observed for this activity

only, was independent of sex but not of strain.

Analysis of genetic crosses and endocrine studies revealed that

1) 16a-hydroxylase activity was reversibly inducible by androgens, and

a single genetic locus regulates the basal level of activity in female

mice; 2) the strain-specific sexual dimorphism of 15a-hydroxylase

activity is controlled by a single genetic locus which probably is

expressed in differences between strains in responsiveness to sex

hormones; and 3) the induction of 15a-hydroxylase activity by estradiol

benzoate was independent of strain while suppression of this activity

by androgen was strain-dependent.


viii














INTRODUCTION

The cytochrome P-450 monoxygenase system is involved in the

biosynthesis and/or catabolism of many endogenous compounds including

cholesterol, bile acids, corticosteroids, prostaglandins, and sex

steroids. The same enzyme system participates in the metabolic

oxidation of a variety of xenobiotics such as drugs, polycyclic

aromatic hydrocarbons, food additives, pesticides, and industrial

pollutants; it thus plays an important role in xenobiotic

detoxification and elimination. Where oxidative metabolism leads to

the generation of highly reactive electrophilic metabolites, such as

epoxides or N-hydroxides, the "detoxification" reactions can result in

carcinogenic, mutagenic or teratogenic metabolites. The broad

substrate and product specificity of the cytochrome P-450 oxidizing

system reflects the presence of several forms of the enzyme. Each of

these forms of cytochrome P-450 may be more or less species-specific.

Each form is likely to have more or less distinctive substrate and

product specificity and inducer and inhibitor sensitivity, and to the

under more or less distinctive control genetically, developmentally,

and distributionally. Therefore, the P-450 monoxygenase system

combines great breadth and specificity with regard to species, sex,

age, tissue, etc. Whether or not the various isozymes represent

distinct gene products and the extent to which post-translational

modification, allosteric modifiers, lipid environment (these enzymes

1










are membrane-bound), etc. modulate the subtrate and/or product

specificity are not resolved.

Although the existence of multiple isozymes of cytochrome P-450 in

mammalian liver is now well established, surprising uncertainty remains

with regard to the extent of heterogeneity and the physical, chemical,

and biological basis for multiplicity especially in the constitutive

forms characteristic of the untreated animal. These forms are of

particular interest because of the possibilitity that their regulation

and functions may differ from those of the inducible isozymes. The

constitutive isozymes represent the bulk of the cytochromes P-450

present in the untreated liver, are immunologically distinct from the

induced forms, and possess greater specificity towards endogenous

substrates. Although under most circumstances constitutive isozymes

account for a large proportion of hepatic xenobiotic metabolism,

surprisingly little is known about the degree of overlap in catalytic

properties between isozymes, the mechanisms by which these forms are

independently regulated by physiological and/or environmental factors,

and the extent of allelic polymorphism in the isozymes. Clearly such

issues are of great significance for understanding and predicting

xenobiotic interactions and human individuality in the response to

drugs.

This dissertation represents a study on the genetic, hormonal and

developmental regulation of various constitutive forms of the

cytochrome P-450 isoyzmes that contribute to the site-specific

monohydroxylations of testosterone in hepatic microsomes prepared from

untreated isogenic strains of mice. Several substrates, including









testosterone and other endogenous steroids, have been used earlier with

success to probe the heterogeneity of the cytochromes P-450 by virtue

of the remarkable regio- and positional selectivity exhibited by

certain isozymes towards these substrates. Indeed, changes in the

relative specificity for site-specific hydroxylations as a function of

species, age, sex and other factors have provided much of the initial

evidence for the existence of multiple forms of cytochrome P-450. Our

application of a common methodology in the comparative study of hepatic

microsomal steroid hydroxylations in male and female mice of several

isogenic strains provides us with valuable insight into the modes of

regulation and polymorphic expression of these constitutive isozymes of

cytochrome P-450. This study also serves as a useful probe of the

catalytic specificity and heterogeneity of those constitutive forms of

the enzyme active toward testosterone.

It is crucial to the rationale of this dissertation that the

investigation involve isogenic animals. Although heterogeneity in the

constitutive cytochromes P-450 probably reflects the presence of a

finite number of isozymic forms of this enzyme, the degree of

complexity could be exaggerated by the many allelic differences that

randombred or outbred animals can be expected to display. The use of

inbred strains prevents intrastrain allelic variations and provides a

homogeneous background for selected studies of the heterogeneity of

cytochrome P-450 by way of interstrain comparisons.

This thesis is composed of five chapters, the first of which

describes a convenient one-dimensional TLC-radiochromatographic

procedure for the separation and quantitation of six monohydroxylated










products of testosterone metabolism and the development of optimal

conditions for the measurement of hydroxylase activity in mouse hepatic

microsomes. Chapters two and three deal with the effects of sex,

strain, and postnatal development on the rates of hydroxylation at the

6a, 60, 7a, 15a, 158, and 16a positions of testosterone. Finally,

chapters four and five represent detailed studies on the genetic and

endocrine regulation of testosterone hydroxylase activities, primarily

at the 15a and 16a positions, in two selected strains of mice. A final

discussion serves to integrate conclusions reached in each chapter.

Each chapter has been prepared for submission as an independent

manuscript for publication.













CHAPTER ONE
HETEROGENEITY OF CYTOCHROME P-450 IN THE SITE-SPECIFIC HYDROXYLATIONS
OF STEROIDS BY AKR/J MOUSE LIVER MICROSOMES

Introduction

The concept of isozymic multiplicity has been used often to

explain the broad substrate and product specificities of the hepatic

microsomal monoxygenase system and its distinctive responses to various

inhibitors and inducers (1,2). Direct evidence for heterogeneity in

the cytochromes P-450 of mammalian liver is abundant. Several forms of

the enzyme have been resolved (3,4), and various isozymes have been

differentiated by subunit molecular weight, immunological properties,

catalytic properties, peptide maps, partial amino acid sequences, and

even nucleic acid hybridization (5,6).

Despite this progress, surprising uncertainty remains with regard

to the extent of heterogeneity, the degree of overlap in catalytic

properties between forms, and the mode of biological regulation of

individual forms and/or catalytic activities (7). In this situation,

where one does not know exactly how many forms there are or how they

relate to one another, unrecognized polymorphism would introduce

serious and unneeded complexity. Recent reports add credibility to

this concern (8,9) and underscore the importance of conducting selected

experiments with isogenic animals. Our lack of knowledge is especially

apparent in the case of the "constitutive" forms of hepatic cytochrome










P-450, here defined as those present in untreated animals. Notably,

only a small fraction of the cytochromes P-450 found in untreated rat

liver crossreacts with antibodies raised to the isozymes induced by

phenobarbital, polycyclic aromatic hydrocarbons, or pregnenalone 16a-

carbonitrile (10,11). These constitutive isozymes are of particular

interest because of the possibility that their substrate specificity,

function and regulation may differ from those of the inducible

isozymes.

Several substrates which yield multiple products, including

biphenyl (12), warfarin (13), and benzo[a]pyrene (14), have been used

successfully to probe the heterogeneity of the cytochromes P-450 by

virtue of the remarkable regio- and positional stereoselectivity

exhibited by certain isozymes toward these substrates. Testosterone,

an endogenous compound, has also served in this regard (15,16,17).

With rat liver microsomes, the relative specificity for hydroxylation

at the 7a, 16a and 68 positions changes with age (18,19), sex (20,21),

neonatal imprinting (20,22) and exposure to xenobiotics (23,24,25).

Some sex and strain differences in the pattern of hydroxylation of

testosterone have also been observed with mouse liver microsomes (26).

Our current work focuses on the site-specific monohydroxylation of

steroids, primarily testosterone, by hepatic microsomes from untreated,

inbred mice. We have found that the usual assay methods developed for

studies of testosterone metabolism in the rat are inappropriate in mice

because the array of metabolic products is different. This report

deals with 1) the development of a reasonably rapid and economical one-

dimensional TLC procedure suited to our purpose; 2) the development of






7


optimal assay conditions for the measurement of hydroxylase activity in

mouse microsomes; 3) the finding of substantial sexual dimorphism in

selective hydroxylations of testosterone by liver microsomes from AKR/J

mice; 4) the differential effects of potassium phosphate buffer on

selective hydroxylations of testosterone by preparations from these

animals; and 5) comparative experiments with progesterone as

substrate.










Materials and Methods
Chemicals. [4-14C]Testosterone (56.1 Ci/mol) and

[4-14C]-progesterone (57.2 Ci/mol) were purchased from New England

Nuclear, Waltham, Massachusetts. Radiochemical purities exceeded 98%

according to the TLC procedure described below. 68-, 7a-, 16a-, and

168-Hydroxytestosterone and 60-, 16a-, and 21-hydroxyprogesterone were

purchased from Steraloids, Inc., Wilton, New Hampshire. 28- and

15a-Hydroxytestosterone were gifts from G.D. Searle and Co., Chicago,

Illinois. 15a-Hydroxytestosterone was kindly provided by the Upjohn

Company, Kalamazoo, Michigan, and 15a-hydroxyprogesterone was a gift

from Dr. S. Solomon, McGill University, Montreal, Canada.

6a-Hydroxytestosterone and additional samples of several of the

compounds listed above were provided by Dr. D. N. Kirk from the Medical

Research Council Steroid Reference Collection.

Testosterone and progesterone, which were purified by

recrystallization twice from acetone/hexane and ethanol/water,

respectively, and NADPH were purchased from Sigma Chemicals, St. Louis,

Missouri. All other chemicals were purchased from the Fisher

Scientific Company, Pittsburg, Pennsylvania, and were of the highest

quality available.

Animals. Male and female AKR/J mice (8 weeks old, 20-25 g) were

purchased from Jackson Laboratories, Bar Harbor, Maine. Mice of the

same sex were housed in groups of up to six animals on corncob Sani-cel

bedding under a controlled environment for at least three weeks before

use. Smoking and the use of insecticides were prohibited, and 12-hour










light/12-hour dark cycles were maintained. Animals were provided

Purina Rodent Laboratory Chow and tap water ad libitum.

Preparation of microsomes. Mice were killed by decapitation

between 8 am and 10 am, and livers were excised and homogenized in ice-

cold 250 mM potassium phosphate buffer, pH 7.4, containing 150 mM

potassium chloride (5 ml/g of liver). Microsomes were prepared by

differential centrifugation (27) and were resuspended in 100 mM

potassium phosphate buffer, pH 7.4, to give a final suspension

containing approximately 10 mg microsomal protein per ml. The protein

content of microsomal suspensions was determined by the method of Lowry

et al. (28) with bovine serum albumin as standard.

Enzyme assays. Cytochrome P-450 content was measured by the

method of Omura and Sato (29) after microsomal suspensions were first

diluted ten-fold in 100 mM phosphate buffer, pH 7.4, containing 20%

glycerol (v/v) and 0.2% Emulgen 911 (v/v). Absorbance at 450 nm was

measured in relation to the steady baseline between 480-500 nm.

Virtually identical results were obtained without Emulgen as long as

careful baseline corrections were included.

The microsomal biotransformation of testosterone was assayed under

conditions similar to those described by Shiverick and Neims (30).

Reaction mixtures routinely contained 100 mNl potassium phosphate

buffer, pH 7.4, 10 mM MgCl2, 1.0 mM NADPH, 131 yM [14C]steroid,

testosterone or progesterone (0.5 pCi) added in 10 pl of ethanol, and

0.4-0.5 mg of microsomal protein (0.2-0.3 mg for progesterone

incubations) in a final volume of 1.0 ml. The enzymatic reaction was

initiated by the addition of NADPH and allowed to proceed for 10 min at










370 C before termination by the addition of 5 ml of methylene chloride.

Controls routinely lacked microsomes, but zero-time and

heat-inactivated samples gave similar results.

Separation and assay of testosterone metabolites. Each incubation

mixture was shaken with methylene chloride for 30 min before collection

of the organic phase, which was then evaporated to dryness at room

temperature under a stream of nitrogen. The residue, which contained

98% of the added radioisotope, was dissolved in 50 ul of ethyl acetate.

Aliquots (25 pl) of the extract and relevant standards were applied to

scored, methanol-washed, 20 x 20 cm, silica gel 60 F254-precoated,

plastic-backed thin layer chromatography plates, 0.2 mm thick (E.

Merck, Darmstadt, Germany). The plates were developed three times in

ether/benzene/methanol, 55/44/0.5 (v/v) in unlined tanks. Optimal

separation of the radiolabeled metabolites of testosterone was achieved

at 170 and an ambient humidity between 30 and 60%, conditions which

were associated with a development time of about 100 min per run. Even

at suboptimal conditions, the desired resolution could be achieved by

increasing the fraction of methanol in the developing solvent. Rarely,

a fourth development was needed to separate 15a-hydroxytestosterone

from a small zone of radioactivity just below it. Progesterone

metabolites were resolved by the same TLC system except that the plates

were developed only twice.

Radioactive zones on the TLC plates were located by autoradio-

graphy with SB-5 X-ray film (Eastman Kodak Co., Rochester, N.Y.).

After the plates were first covered by a single layer of Scotch 810

transparent tape, each zone was cut out and counted directly in 10 ml










of Liquiscint (National Diagnostics, Somerville, N.J.) in a Beckman LS

7000 liquid scintillation counter with external quench correction

(Beckman Instruments, Inc., Fullerton, CA); recovery of 14C applied to

the plate was greater than 95%. The Scotch tape, which was found not

to affect counting efficiency, was quite helpful in avoiding chipping

of the gel during cutting.

Metabolites were identified by comigration with unlabeled

reference standards in five different solvent systems. The standards

were visualized by fluorescence quenching. Identification and the

radiochemical purity of each putative hydroxytestosterone metabolite

was further assessed by acetylation. For these experiments, radio-

active zones were cut and eluted into ethyl acetate containing 40-50 Og

of the unlabeled standard. Sufficient numbers of each zone were

combined to yield about 15,000 cpm. Samples were evaporated to dryness

under nitrogen and incubated in 100 ~1 of pyridine and 100 pl of acetic

anhydride in foil-covered, tightly capped test tubes. After 24 hours,

the samples were dried and chromatographed as described above except

that the plates were developed only once. More than 85% of radiolabel

in each metabolite zone was recovered during TLC of the acetylated

derivatives. The Rf values of the radioactive areas were compared to

both internal and external acetylated reference standards.

Identification of progesterone metabolites was based only on

comigration with reference standards on the TLC system described

above.

Calculations and statistical analyses. The amount of each of the

various hydroxysteroid metabolites was computed from the fraction of






12


radiolabel in a particular zone relative to the total radioactivity

applied to that lane of the TLC plate. Specific activities of steroid

"hydroxylases" are expressed as nmoles of product formed per min per mg

microsomal protein. Molecular activities are defined here as units of

hydroxylase activity per nmole of cytochrome P-450. The data were

analyzed by the unpaired Student's t-test. A P-value of less than 0.05

was taken as significant. Unless otherwise indicated, each value is

expressed as a mean and standard error.










Results

Identification and TLC of testosterone metabolites. The

incubation of [14C]testosterone with NADPH, oxygen and hepatic

microsomes from AKR/J male mice resulted in the formation of a large

number of metabolites. This is seen in Figure 1 (left panel) which

depicts the reaction products resolved by a 2-dimensional TLC procedure

similar to that reported by Ford et al. (26). For routine analysis, we

developed a more rapid and economical one-dimensional TLC system. The

latter system was designed especially to separate and quantitate the

15a- and 16a-hydroxytestosterone metabolites because pilot experiments

had revealed these metabolites to be most significant for our purposes.

An autoradiogram of the one-dimensional chromatogram (Figure 1, right

panel) shows that twelve distinct zones of radioactivity were resolved.

Comparison of the one- and two-dimensional chromatograms revealed that

zones 1 and 2 (which comigrated with the 15a- and 16a-hydroxytesto-

sterone reference standards, respectively) as well as zones 4, 5, 6 and

7, were essentially devoid of obvious contaminants.

The relative mobilities of the metabolites contained in zones 1

through 8 and a set of hydroxytestosterone standards are presented in

Table 1. The metabolites were identified tentatively by comparison

with these external standards, as well as by comigration with internal

standards on one- and two-dimensional TLC. The presence of these

metabolites had been established by GC-mass spectroscopy (31).

Testosterone is in zone .11, and the metabolite in zone 5 is not yet

identified.














ZONE

12 kl mm,
12-
I1-I
10-
9-



8-- -"M


6- -
5-

4-
3-
I-


* ORIGIN


Autoradiograms of two-dimensional (left panel) and
one-dimensional (right panel) chromatograms of
[14C]testosterone metabolites. AKR/J male liver
microsomes were incubated with [14C]testosterone under
the conditions described in the text. Testosterone and its
metabolities were extracted and then spotted onto silica gel
F-254 plates. Two-dimensional chromatography was conducted
by first developing the plates horizontally two times with
chloroform/ethanol (92:8) and then developing the plates
vertically three times with ether/benzene/methanol
(55:44:0.5). The one-dimensional plates were developed
three times with the latter solvent system. The major
radiolabeled zones on the one-dimensional plate were
numbered arbitrarily with lowest numbers closest to the
origin.


Figure 1.


0


"a
































0CV)

JO





~LJ
LLJ C
0m4







O I--00
















OF--
S-- Ll LU











I0 U--
m- 01













O -

_J



0
-C)
-- LLJ <
)>- -4




UJ I --
MO




























04
-1C---



















C0l
<:tt
z > -iL
0 *-' U
00 Cl 3


4)
4-.
0








"U
cu
>)
*Pc_




=to


C-
4-4

4-I

o
*-3





0

0


o "
4o-
0 0


0








ac
C
0
0)
4-) -C



x V
0
o



I-


-- Co o t
10. o (0 0 LI

0 0 0 0


O CO C O










o M
* I











S0 0-







0 0 04





























S0 -
-4 -4


(c Rd-

00






NJI
1 I
*10
1 8









* -4
0










o0
.--


* *

1 0n








I c

0 0





I NC
I 0


cJ CMio) C0
-%- co LO LA










O 001


Ln Lf L

8 88
* *


0 00




co od
0 00


SLn to r- coC




















d I C C
.-* i *


0
4-3
o
































0


4-)
o


























r

a)





I-
























0)
UC
*r- 0
EN



3 0




EO
0 4-
EU
C- >


L -

w ei
e o
Eo





E C
a




U+)
0-








0-0
to-
E 1



uca
04-)









CJ
-r-
E -0

a)
40 4-







SD
4-) EU





r--

'U


n0
a 4-
$C X

U)
C







r-c






-4
r*r
o a






O
0- 0
1.) 0 -
U -o



00) *)




o-- c
I- a


m0



*w r ---
4- 4-



4C-

C3 *0 *. t
M) Cu c a









-r 4--, E-
'U 0 "C



0 LO
GE>
4- 0 4-)
r-) Tr -- *M-





S EU





00 "a
O C ) 4--





- C- *r- 3
i- e -




0E4.) C
U a 4 0














o i 0 $-
0) >0
a 4u -0 a) 0










4-' .- 0 0
0. to -0 41<










U -0 4--
4-' 0 5s- 5-
CO)4-> 3 4-

-- E '- /) ..
0 -00

c- *- (a 4-) 4-1
0 0 0 C
0 *r- X Q. 5-
0 4- 4-n
- *r-- U











0*a a)













<. &. 4 3 C L
S o' oU X
0 X0 c -
4- 0) .r- C 0
5- E *i- W

)0 C
U* a a c
S CU o)

o- 4 4rn0
4)

->U >) 14-
C C



0 *- w
*--.- 0 U)

S0) *- UL
dTn & EmdGM


E
x -o
a C >
S0"D 0






Q ',-*
4-) a) 4-





0 -4 -
>)
04->4-)









.) .0 .. a)
(U 0
4-> N J 5 -- .


U .-0 0 .
U 4-.) "0
U -C 0 -c


-u m + -




4 0 .. -C 0)O-



-t I 5 -)
ci- a) Ca > C 0
0 -O0 )- I 3-
4-' S- 4--U






C)- ( .9-- (A
-0- .- (0 -0,
5- U '-U 0. Q



4-) E X .




a)+J(a0 -) UU





-0 4 ) -P -r-
* 0r0E0

0 04 0 U-
CU int
S0)0-- Uo




(a -)o o .
m0 C O*>(
















0-)0 4-) .--
.0 -) 4to c0 E
- 0) c .4) a0 I-

0 0 4-)-





a- C : ( t- o
cU U 30 :E 0


(0 MC- 4-C --
a -1 4- 4 a) -C
*e R00 C






U? 0)v.E E
*- U4 CU-
J34> U m _
>a O CJ o i
>- e a *i
















"f 3 0.










Further identification, as well as assessment of the radio-

chemical purity, of the metabolites contained in zones 1-8 was

accomplished by acetylation of standards and the labeled material in

each zone. After acetylation, essentially all of the radioactivity in

zones 1 and 2 migrated with Rf's indistinguishable from the diacetyl

derivatives of 15a- and 16a-hydroxytestosterone, respectively (Table

1). As suggested by the results depicted in Figure 1, zones 3 and 8

contained more than one labeled compound. About 60% of the label in

zone 3 comigrated with the diacetyl 7a-hydroxytestosterone standard.

Zone 8 contained substantial quantities of both 28- and

16B-hydroxytestosterone, 55% and 28%, respectively, as well as an

unidentified product. Derivatization confirmed that the radiolabel in

zones 4, 6 and 7 was composed almost entirely of 150-, 6a-, and

6S-hydroxytestosterone, respectively, although a minor (approximately

20%) contaminant was found along with 15B-hydroxytestosterone in zone

4. The stabilities of individual diacetylated metabolites (or

standards) were not determined, and liability could have contributed to

some of the apparent contamination.

The [14C]testosterone metabolites produced by AKR/J male

hepatic microsomes were also identified by their comigration with

reference standards in 4 other TLC systems used in earlier studies of

the biotransformation of testosterone by rodent liver microsomes (Table

2). Careful examination of Table 2 reveals that none of these solvent

systems resolved all of the metabolites in question. Nonetheless, the

sum of these results supports our identification of metabolites.
































cOT
Z-O
O L
OC) -

C:) LLJZ
)C)

o C







_jO
0 V)













r- I--
w) C) M-


1- C- V0





>
LU0



JI -
CiLii-

0 3

L
= UL IL-





Li-
-)UJ -
UJ 0Z J0


-3



4-)






0

a)

-0


4-)
4-

ai)









4-)
00




I-4




S0







4-,
V)


ra




0 t,
41 4-0

F-


it0 0o
o o o
* *


o o C)







* *.




0 0 0








o o











C) c C

0 0 0























Ln fD co
0 0 0






-O -0 -O
-4 -4r


Mr Cw cOj r- C)
* *


I :- 0 CD

I C) C) C)


o co "o Un "n
C 0 0 0 0




co t no r co
C) I O O co
C) I C0 00






o 1.0 C o d 1o
qt- LO UD to CC)

o 0 C) 80 8)


I CO t- 4D
I CD0 C CO
I *
1 0 0 0


c ?- r-4 r- 0n

o o o o O


co o co
* I *
C) I CO 0 C



>1 8
0 0 0 L-
5- S- S- ()

) 0
I C I I +-
a 0 o c oa W
CO t 0v
-4 1-










19











L.

30 01
S- 4-)



-'CO~--.b .0 0

> 4..) 0 c
4J M -0 (U E

I-t 0---S-4)Vr -

o .9)4 rr- 000
14-- W-- UU -0
u. CO0 o 0
4. 0va 0 0 C

4-) '-, 3:L OL
4-3 --- r v .,--a
X E 0 C) 4-3 C
4C 0 00 ru
4-,\ .) LL -




a) 00

C S.- r D (A a
-- E U4-


ID o a CL-
4-3.0 00 c
:3 ~M n > 4-3
-0 C.4-) V4
a- ) 0 S- 00to
C)> *- 4-3
r V C ..C- C C)



j > 4t2 C a.CO 0

C- L) to 0 3:4-

4) CU 4-C
0 C3) c V) W C:C,


0 .---00 (U -a)
4-) m = ( =3 3:

'-00 too) to ->
OSCVICUVV 00
a = W~r 0 a L) 01

C 0 +J 0 4 Ca

0o ,- 4C a)C:= .p
U 0 a

a, 0 0 u a 0C v

C-) 0 4.a) 4-

0'C 4-3 V 0--
L J-- L. --




-o u CD :3 M4-3 3: (
-U S.. -

4-))V a 0

x ~ ~ 0 ~ 4 >-4.-0 4-)i
I4r- 4a ) 0to+.) C0


I ~ ~ ~ ~ S tC-.) -a m (
CL r 4- C:) 0 0)1 L-
C OJ 4-) 0 M (D a

LLJ E W to-- -- L) L
CO _j ,;r (u a (u o .,

:t F--c 0 L- u V a)
1- '01 _n o rJ a a










Characteristics of the enzymatic reaction. Typically, 20-25% of

the substrate, testosterone (initial concentration, 131 pM), was

consumed in 10 min upon incubation with 0.4-0.5 mg microsomal protein

The metabolites in zones 1 (15a-hydroxytestosterone) through 7

(60-hydroxytestosterone) accounted for more than 65% of the total

metabolism. 68-Hydroxytestosterone was the major metabolite formed in

the presence of both male and female liver microsomes. Each of these

metabolites was produced more rapidly in the presence of microsomes

than any other subcellular fraction.

Rates of formation of each of these hydroxytestosterone

metabolites were proportional to microsomal protein concentration

(Figure 2). Figure 3 depicts the relationship between time and the

production of metabolites. Rates of formation of 15a- and

16a-hydroxytestosterone were constant for 15 min. The rates of

formation of 7a-, 158-, 6a- and 68-hydroxytestosterones were constant

for at least 10 min. After 15 min of incubation, the rates of

formation of all products, especially 68- and 150-hydroxytestosterone,

decreased significantly.

The metabolism of testosterone by hepatic microsomes from male and

female mice. Figure 4 compares the relative rates of formation of six

hydroxytestosterone metabolites in the presence of microsomes prepared

from male and female AKR/J mice. A female predominance was observed in

hydroxylase activity at the 6a, 65, 158 and especially the 15a position

of testosterone. Hydroxylation at 7a was not affected by sex, and

hydroxylation at the 16a position of testosterone was uniquely

predominant in males. Table 3 compares the apparent Km and Vmax for


















*0 6a


0.6


MICROSOMAL PROTEIN (mg)


Effects of protein concentration on the rates of formation of
various hydroxytestosterone metabolites. The standard
reaction mixture, which was incubated for 10 minutes at
370C, contained 131M [14C]-testosterone, 100 mM phosphate
and NADPH as described in the text. The microsomes were
prepared from male AKR/J liver.


Figure 2.


_~ _I ____ _11_ ~_~_111~ ~_ ____ __




















oC
0



0-*
o
Oe




IL
O
S
E


z
0
o


Ir


w
L-




0



I
0




0:


16a


150


153


TIME (min)


Effects of time of incubation on the formation of various
hydroxytestosterone metabolites. The standard reaction
mixture, which was incubated at 370C, contained 0.4 mg
microsomal protein prepared from male AKR/J liver, 131 yM
[14C]-testosterone, 100 mM phosphate and NADPH as
described in the text.


Figure 3.


_ __UIUII_1I__UI__II___ ~_








23





















HYDROXYTESTOSTERONE METABOLITES


15a 16a


Ta 15.8


C-

I-;



4t
02
4%.
XE
0Q*

C,


dy d? dC d? cf d?


Rates of formation of six hydroxytestosterone metabolitl by
microsomes prepared from male and female AKR/J mice. [ C]-
testosterone was incubated with microsomes under standard
incubation conditions described in the text. Note different
scale for 6a-hydroxytestosterone. N=5 and indicates
P<0.01 for male-female comparison.


x
W M
X





z
0CW
W

0 E

W R


Figure 4.


__ ____























o 0 o 0 0 r-

* *c -* *

O C O 0 C -
m -i co CM o n c

I CN C4 C
Lfl


-a" rs in
* *
0 '-4 0


r- in)
T to*
41 L4


0

-4

r-4

Co








CO
















C)
0
C-t














on
U3


Z5 ? a c a ca
L 10 L O kn D0
1-4 -4 -1


0
CD



I-
C,
P---

Lui


1'-U
o-'-4

0_Q





-J




-J s:


IC- LLuJ



I--









.r-


C-)-
-4-
a-
< -
t^ i


E
0



EE

0
E
C
a|t-


C-


xr-
E
O



EE

0
E
-
C


C
0
c




o

art
0


x o
o -
(O
+J CD
V) +-
0 *'
*<- I




















U)
0

4-3
L4 SC C
C.iLO G)C .3-
-.3o E



U 4J,

4-) -) 0
0 4- C cI

4r 0 .
C( 4- C U
4- (Z 4-)
4 u- 4-)OM a)



kr~-o 4

> r -, (A
4-4..U 0

4-3-




'0 U 4-) U u
>) M QU C

C: 4->' 0n
u-odidi E fu



ej 4- 4-3 4- 0
$- .-.3U


a) =4iU 0)
> 'E




&-~ E
5-D E 0
4)4-7)0 0.



S4- S- 0
M- 0 4J


.A r. r


Qj ~ 4 4-.. t
C) E I- M)





E u- 0i S-- 0
cii ci).. E




I o 3 0.





S- ~ to C -
L-0 5-
4 (VQ) W.10)
cr U) 0o
a)Co Ci c)

_0EE -
1 cii 5- O- 5







tA (A C0 tn t
dii L-O4-)-- S-


43 U)C) U t) -#- o C.
.4) ) to 4j 0

a) E 4-. -di U)
I -I = -0 cu
I L UU )u ) 4-)
CY) *.'- C 4-1 3 4

CU 4- U -0 4
SEjOu(43 E
: 0 L Ca a >
fa -01r










production of each metabolite in the presence of male and female

microsomes. Lineweaver-Burk plots for each metabolite were essentially

linear with preparations from either sex. There were no significant

differences between male and female mice in the apparent Km values

for any of the six testosterone hydroxylations. For the 6a, 7a, 15a

and 16a hydroxylations, the apparent Km's for testosterone were low

(1.4 to 6.1 pM1) in comparison to those for 68 and 150 hydroxylations

(41 to 80 PM).

Maximal velocity for 15a hydroxylation was more than 3-fold

higher with female than male preparations, while Vmax for 16a

hydroxylation of testosterone was more than 3-fold higher with male

than female microsomes. Both Vmax's and relative rates for 6a,

68 and 158 hydroxylations were significantly higher in female

preparations, but differences in each case were less than 2-fold.

7a-Hydroxylation was the only activity studied that was not affected by

sex, although only 60% of the radiolabel in its zone is

7a-hydroxytestosterone. Admixture of liver microsomes from male and

female mice in 4 different proportions yielded additive activities for

production of the 15a-, 16a- and other hydroxytestosterones (data not

shown). Therefore, we find no evidence for the presence of a

reversible, non-stoichiometric activator or inhibitor for any of the

hydroxylations. The mixing experiments also decrease the probability

of product inhibition as an explanation for the sexual dimorphism.

The content of cytochrome P-450 in hepatic microsomes was not

dependent on sex (male and female, 0.816 and 0.760 nmoles P-450/mg

microsomal protein, respectively). Thus the sexually dimorphic










hydroxylase activities persist even when results are expressed in terms

of molecular activity. Indeed, we found less day-to-day variation in

molecular activities than in specific activities.

The effects of potassium phosphate concentration on testosterone

hydroxylation. All aforementioned experiments were conducted in 100 mM

phosphate buffer. Figure 5 presents the results of experiments in

which the relative rates of hydroxylation of testosterone by male and

female microsomes were studied at 30, 100 and 330 mM phosphate, all at

pH 7.4. The effects of changing potassium phosphate concentration were

strongly dependent on the position of hydroxylation, and in the case of

16a and 15a hydroxylations, on gender as well. An 11-fold increase in

phosphate concentration resulted in substantial inhibition (75 to 90%)

of 15a and 7a hydroxylation, little or no effect on 150 and

6a-hydroxylation, and a modest stimulation of 60 and 16a (male only)

hydroxylation. Male 15a hydroxylation was relatively insensitive to

changes in potassium phosphate concentration between 30 and 100 mF, but

female activity decreased 2-fold.

Progesterone. A limited number of experiments were conducted with

progesterone as substrate in order to gain some insight into the

substrate specificity of the hydroxylases under investigation. Results

with hepatic microsomes from male and female AKR/J mice are presented

in Table 4. Progesterone was actively metabolized, and 15a-, 16a-, 6B-

and 21-hydroxyprogesterone were identified as important products. Like

testosterone, 68 hydroxylation of progesterone was quantitatively most

important and exhibited a small but significant female predominance.

The female predominance seen in 15a hydroxylation with testosterone was







28


even more striking with progesterone in that a 5 to 6-fold female/male

ratio was observed. The male predominance in 16a hydroxylation with

testosterone (3-fold) was not seen with progesterone.
















HYDROXYTESTOSTERONE METABOLITES


S.-

Ue
4


it


30 30 100 330 30 100 330 30 100 330 30 100 330
POTASSIUM PHOSPHATE BUFFER (mM)


Figure 5.


Effects of potassium phosphate buffer concentration on the
rates of formation of different hydroxytestosterone
metabolites in the presence of hepatic microsomes from male
and female AKR/J mice. Livers from 2 mice were pooled for
preparation of microsomes. Incubation conditions are
described in the text; the initial [14C]-testosterone
concentration was 131 VM, but the concentration of potassium
phosphate, pH 7.4, was varied as indicated. All hydroxylase
activities are expressed in relation to activity at 100 mM
phosphate (defined as 100%) since that concentration is used
in the standard assay. n = 3. and + indicate P<0.05 for
30-100 mM and 100-330 mM comparisons, respectively.


__ __










30















c,
*- (D
0 *0>
mI to cn An
LO 0' Cj mD
SO- c)


S--r ea





<- E ""
tI- 2 o 4 xc




PA C x 0




. 0 L. ,


)- U- C- .,



< *- L L *,

Z- CU4 r -N )< 0 0. C

0 C w0 E =3 0C 4-



0 ,,-- "I 0 -- _
'-J a L I cV) 0 C
- 0 + -.t 0 0


m J< LO f % C C) 0 0

( OC) 4 E CU 4) a t a



:. 0r -
0- U, 4-. O 0 v v


OL J 0 0. 0
CoE mW-p E



>O E E
.~0Z E




Iw-- < t 1 O *n
< J (U U CD C94
M: Lr-I 't S- 0) 0)





J: IC O O 0 S- M 4-l
ODoU 0 ) -
- U- 0C E 0


C a1 0C a .0 NO o


.LL. ,- C .,- O C
CC OS. 4 WU4-a
_1 O tO 4- r
d ca 0 50 t S ) r- CA t

LLS 0 00 oL L- E.L & 0Wi
C) ) r 4 N Ca 0 4- 4- 4J
w or I W-a a to f
X. a) C>S.- 0 to
0L 0C E = E U a U
-4 0S W :- a
O LUM S *) *r
OL4 OE 0-
E U- <:Q- U U *- 0 C










Discussion

These results initiate a set of detailed studies with highly

inbred mice on the heterogeneity, substrate specificity and regulation

of the constitutive forms of cytochrome P-450. The overall project is

designed to control for possible allelic polymorphism in the P-450's

and later, to take advantage of potential strain differences. Our

initial goal was to establish a reasonably rapid and economical system

for assay of the major monohydroxylated metabolites of testosterone.

We have developed a one-dimensional radiochromatographic technique for

the assay of several hydroxytestosterone metabolites produced by mouse

liver microsomes. The method has better resolving power than other

reported TLC systems and is more rapid, less tedious and/or less

expensive than the two-dimensional TLC or the paper chromatographic

procedures that have recently been used in studies with mice (26,32).

We have focused on six hydroxytestosterone metabolites which

together comprise more than two-thirds of total products. Metabolites

were identified by comigration with reference standards in five

different solvent systems; identifications were confirmed by study of

diacetyl derivatives. While our method is good for assay of the 15a-,

16a-, 6a- and 63-hydroxytestosterone metabolites, 7a- and

15B-hydroxytestosterone account for only about 60% and 80% of the

labeled metabolites in zones 3 and 4, respectively. We are especially

cautious in the interpretation of 7a-hydroxylase activity because of

the susceptibility of this metabolite to spontaneous dehydration to the

4,6-diene both before and during TLC (31).










Large discrepancies exist in the reported rates of 16a, 7a and 68

hydroxylation of testosterone in the presence of hepatic microsomes

from various strains of mice (32-35). These discrepancies, which also

include differences in the rank order of the different hydroxylase

activities, could reflect allelic polymorphism between isogenic strains

of mice, but at least some of the discrepancy seems to be

methodological; it results from certain features peculiar to murine

hepatic steroid metabolism. Ford et al. (31) have reported the

two-dimensional TLC separation of several of the metabolites of

testosterone formed by mouse liver microsomes. At least two of these

metabolites, 15a- and 150-hydroxytestosterone, have not been reported

after incubation of testosterone with rat or rabbit microsomes. Mouse

liver microsomes also possess little if any 4-ene reductase activity

(31). Yet many earlier experiments with mice have been conducted

employing solvent systems (Table 2) and assay conditions optimized for

studies with rat or rabbit.

We have examined the rates of formation of each metabolite as a

function of various assay conditions. Subcellular fractionation

revealed that the hydroxylases in question were concentrated in the

microsomal fraction and that varying the NADPH concentration 2-fold did

not influence results. As discussed below, because the concentration

of potassium phosphate buffer had differen- tial effects on

hydroxylation at different positions, we elected to maintain phosphate

at a concentration nearly that found in liver cytosol (ca. 70 mM).

Under our standard incubation conditions, testosterone concentration

was at least 20 times greater than the Km's for the formation of all










products except 68- and 158- hydroxytestosterone. This assured

catalytic rates near Vmax for most products, and rates that were at

least 60% Vmax for all products. While the amount of each product

formed was proportional to both time of incubation and microsomal

protein concentration, our standard incubation conditions remain just

inside the limits of linearity for the formation of 68- and

158-hydroxytestosterone, perhaps reflecting the higher Km's for

these hydroxylations and/or differences between P-450 isozymes.

There was little variability between AKR/J mice of the same sex

with regard to rates of formation of six different hydroxytesto-

sterones; standard deviations approximated 15%. The rates of 15c, 158,

6a and 68 hydroxylation of testosterone were higher in female than male

AKR/J microsomes with 15a-hydroxylation exhibiting the greatest degree

of sexual dimorphism (3 to 4-fold female predominance) despite the lack

of sexual dimorphism in P-450 content. Hydroxylation at 7a was not

sexually dimorphic, and the rate of formation of 16a-hydroxytesto-

testosterone was 3 to 4-fold faster in male than female microsomes.

Sex-dependent differences in steroid and drug metabolism in rats have

been well documented (36,37), and recent evidence suggests that these

differences reflect the presence of sex-dependent forms of hepatic

microsomal cytochrome P-450 (38,39). Reports of sex differences in

drug metabolism in mice are inconsistent, perhaps reflecting

strain dependence and/or a smaller quantitative difference between the

sexes in mice (40,41). The sexual dimorphisms we find for 15a and 16a

hydroxylation of testosterone are larger than typically seen with other










substrates in mice, perhaps because it is an endogenous substrate.

Sex differences in the selective hydroxylations of testosterone were

also exhibited by A/J and 129/J (but not the C57BL/6J) mice (26).

Studies of testosterone metabolism in reconstituted P-450 systems

from animals induced with different xenobiotics have suggested that

hydroxylations at the 68, 7a and 16a positions are catalyzed primarily

by different forms of cytochrome P-450 (42,43,44). This concept has

been extended recently to untreated animals (45,46). By virtue of the

broad substrate and/or product specificity exhibited by certain forms

of cytochrome P-450, it is likely that hydroxylase activity at a given

position on the steroid nucleus reflects the activity of more than one

monoxygenase present in hepatic microsomes. Recently, Thomas et al.

(47) have established the existence of more than one testosterone 16a-

hydroxylase in rat liver microsomes. They found a large proportion of

testosterone 16a hydroxylation in microsomes from untreated rats to be

immunologically distinct from that due to cytochrome P-450b. Our

finding of a male predominance (3- to 4-fold) in 16a-hydroxylase

activity toward testosterone but not progesterone suggests that the

steroid 16a-hydroxylase activity of mouse hepatic microsomes may also

be heterogeneous. Based on analogous experiments, a similar

interpretation was proposed by Kremers et al. (48) for the rat. The

differential effects of potassium phosphate concentration also support

this conclusion. Male but not female testosterone 16a-hydroxylase

activity was stimulated by high potassium phosphate concentrations.

Our findings probably reflect a different distribution of P-450's

in male and female AKR/J liver, with each P-450 having more or less










discrete specificity with regard to the selective hydroxylation of

testosterone. Nonetheless, it seems quite probable that any given

"hydroxylase" (e.g. testosterone 16a-hydroxylase) is the result of

activities imparted by more than one enzyme. Certainly P-450's with

the functional propensity to direct most hydroxylation of testosterone

to the 15a or 16a positions can be distinguished from each other and

from other testosterone hydroxylases by comparison of sexes. To what

extent male and female forms of the P-450's differ qualitatively and/or

quantitatively cannot be surmised at this time. Kinetic experiments

reveal that the sex differences in AKR/J mice were not due to

differences in apparent Km's between sexes. Furthermore, mixing

experiments did not suggest a gender-dependent activator or inhibitor.

Our results could indicate differences in the concentrations of certain

forms of cytochrome P-450 between males and females.

It is not known whether the female predominance in 15a, and to a

lesser extent 158, 6a and 68 hydroxylation reflects quantitative or

qualitative differences in one or several P-450 isozymes.

Interestingly, the female predominance in 15a hydroxylation observed

with testosterone is even more striking with progesterone. The greater

sex difference results from higher rates of 15a hydroxylation toward

progesterone than testosterone with female microsomes; male

15a-hydroxylase activities are similar with both substrates. This

result suggests qualitative differences between the process of

15a hydroxylation in males and females and indicates that steroid

15a-hydroxylase is also a heterogeneous enzyme. This conclusion is

also supported by the effects of potassium phosphate. Rates for 15a






36


hydroxylation of testosterone in females were found to be 2-fold higher

at 30 mM potassium phosphate than at 100 mM while little difference was

observed in 15a-hydroxylase activity in males at these two

concentrations. This interpretation is more convincing when one

considers that the effects of phosphate concentration on other

hydroxylations (except 16a) are the same with male and female

preparations. More detailed experiments of the stimulatory and

inhibitory effects of potassium phosphate (cf. 49) are in progress.














CHAPTER TWO
THE INFLUENCE OF STRAIN AND SEX ON SITE-SPECIFIC TESTOSTERONE
HYDROXYLASE ACTIVITIES OF MOUSE HEPATIC MICROSOMES

Introduction
The isolation of a number of inducible cytochrome P-450 isozymes

from liver microsomes of rat (43,50) and rabbit (51,52) as distinct

proteins with different chemical and physical properties (3,4) has

established the multiplicity of mammalian hepatic microsomal cytochrome

P-450. Recently, attention has also been focused on the various forms

of cytochrome P-450 from untreated rats (4,6) and rabbits (45,53),

partly because immunological techniques have revealed a lack of

cross-reactivity between various inducible forms and the bulk of these

"constitutive" isozymes (10,11,47). The molecular basis for

heterogeneity within the constitutive forms is a particularly difficult

problem because each isozyme likely occurs in relatively small amounts.

Nonetheless, these forms are of special interest because of the

possibility that their physiological functions) and regulation may be

quite different from those of the inducible isozymes. These isozymes

likely have greater specificity than the inducible forms toward

endogenous substrates like steroids, and their regulation involves

factors such as neonatal and pubertal sexual differentiation (38,39).

The extent of heterogeneity in both inducible and constitutive

isozymes of P-450 could be exaggerated in studies of randombred or

outbred animals by the many allelic differences that such animals are










likely to display. Johnson et al. (8) have recently described

differences in progesterone 21-hydroxylase activity and form 1 P-450

between individual New Zealand White rabbits. Moreover, Vlasuk et

al.(9) have reported that Holtzman and Long-Evans rats, and even

different colonies within those strains, could be distinguished by the

characteristics of their phenobarbital-induced cytochromes P-450.

While it is not clear whether this degree of complexity reflects the

presence of genetically distinct P-450 apoproteins or is the result of

postranslational modifications) of one or more gene products, it is

clear that the heterogeneity of cytochrome P-450 and its regulation

represent a complex problem involving both genetic and environmental

influences.

Several substrates, including testosterone and other endogenous

steroids, have been used successfully to probe the heterogeneity of the

cytochromes P-450 by virtue of the remarkable regio- and positional

selectivity exhibited by certain isozymes. We reasoned that a detailed

comparative study of hepatic microsomal steroid hydroxylations in male

and female animals of several isogenic mouse strains could provide

valuable insight into the catalytic specificity, regulation and

polymorphic expression of several forms of cytochrome P-450 in

untreated animals. Strain differences in steroid hydroxylation have

been reported for many species (54,55,56) and in some instances the

patterns of inheritance have been worked out (57,58). In regard to the

hydroxylation of testosterone by mouse liver microsomes, large

discrepanices exist between reported rates for 16a, 7a, 60 and other

hydroxylations in studies by various investigators (26,32,33). These







39



differences could reflect a large degree of polymorphism among isogenic

mouse strains, but methodologic differences between laboratories

probably have contributed (59). In this study, the use of inbred

strains controls for intrastrain allelic variations and a common

methodology applies in all experiments. These factors allow more

meaningful interstrain and gender comparisons.










Materials and Methods

Materials. Testosterone [4-14C] (56.1 iCi/pmol) was

purchased from New England Nuclear, Waltham, Ma. The radiochemical

purity was greater than 98% as established by TLC on silica gel

F254 plates with the solvent system ether/benzene/methanol,
55:44:0.5 (v/v). Testosterone was obtained from Sigma Chemical Co.,

St. Louis, Mo., and was recrystallized twice from acetone/hexane.

Reference steroid standards were obtained through sources described

previously (59).

Precoated (0.2 mm) silica gel 60 F254 TLC plates (E. Merck,

Darmstadt, Germany) were purchased from Scientific Products, Brunswick,

N.J. SB-5 X-ray film for autoradiography was obtained from the Eastman

Kodak Co., Rochester, N.Y., and Liquiscint liquid scintillation

cocktail was obtained from National Diagnostics, Somerville, N.J.

NADPH was purchased from Sigma Chemicals, St. Louis, Mo. All other

materials were purchased from the Fisher Scientific Co., Pittsburgh,

Pa., and were of the highest quality commercially available.

Animals. Mice of 8 strains were obtained at the age of 6-8

weeks (20-25 g) from Jackson Laboratories (Bar Harbor, Me). Mice of

the same sex were housed 4-6 per cage on corncob Sani-cel bedding. The

animals were maintained on a 12 hr light/12 hr dark photoperiod and

were allowed both food (Purina Rodent Chow) and water ad libitum. All

mice were 11-14 weeks of age at the time of study.

Preparation of microsomes. Given the possibility of circannual

variation in testosterone hydroxylase activity (26), all animals were

studied between September and December. Male and female mice of a










given strain were killed by decapitation on the same day between 8 and

10 am. Livers were excised immediately, and gall bladders were

carefully removed. The livers were then rinsed, and each gram of liver

was homogenized with a motor-driven glass-Teflon homogenizer in 5 ml of

ice-cold 250 r~1 potassium phosphate buffer, pH 7.4, containing 150 mM

potassium chloride. Microsomes were prepared by differential

centrifugation. Briefly, the homogenate was centrifuged at 10,000 g

for 15 min, and the supernatant fraction was recentrifuged for 1 hr at

110,000 g. Microsomal pellets from 1 g of liver were resuspended in 2

ml of 100 mM potassium phosphate, pH 7.4, and assayed immediately.

Microsomal suspensions contained approximately 8-10 mg protein per ml

as determined by the method of Lowry et al. (28). Cytochrome P-450

content was determined by the method of Omura and Sato (29) after

ricrosomal suspensions were diluted 10-fold in 100 mM phosphate, pH

7.4, containing 20% glycerol (v/v) and 0.2% Emulgen 911.

Radiometric assay of testosterone hydroxylase activities. The

rates of formation of 6c-, 6B-, 7c-, 15a-, 15B-, and

16a-hydroxytestosterone were determined by the radiometric assay we

described earlier (59). Briefly, 0.4-0.5 mg microsomal protein were

incubated at 370 C in a shaking water bath with 14C-enriched

testosterone (131 nmol containing 0.5 iCi; final concentration, 131 pM)

added in 10 pl ethanol; 1.0 ymol NADPH (final concentration, 1.0 amM);

10.0 pmol MgC12 (final concentration 10.0 mM); and potassium

phosphate buffer, pH 7.4 (final concentration, 100 mM) in a final

volume of 1.0 ml. Reactions were started by the addition of NADPH and

stopped 10 min later by the addition of 5 ml of ice-cold methylene










chloride. Each sample was then transferred to a 10-ml Teflon-lined

screw cap test tube and shaken vigorously for 30 min. Phases were

separated by centrifugation, and the organic phase was collected and

evaporated to dryness at room temperature under a stream of nitrogen.

Residues were resuspended in 50 pl ethyl acetate and resolved by

TLC. Routinely, plastic-backed silica gel TLC plates were developed

3-times with ether/benzene/methanol (55:44:0.5, v/v). The testosterone

metabolites formed in the presence of hepatic microsomes from at least

one male and one female from each of the 8 strains were also resolved

by two-dimensional TLC; silica gel plates (20 x 20 cm) were developed

in the first dimension with chloroform/ethanol (92:8) and in the second

dimension 3-times with the solvent system described above. All

chromatograms were subjected to radioautography, and radiolabeled zones

were cut out and quantitated by scintillation spectrophotometry.

Statistics. Two-way analysis of variance was conducted for each

hydroxylase activity in order to test for the influence of sex, strain,

and the interaction between sex and strain on enzymatic activity (60).

Where the interaction between sex and strain was found to be

significant, Duncan's multiple comparison procedure was used to compare

the 8 strains at both levels of sex, and also to compare the two sexes

at each level of strain (61).










Results

The data collected in this study are reasonably straightforward;

the rates of formation of six different monohydroxytestosterone

metabolites in the presence of hepatic microsomes from males and

females of eight strains of mice were determined. Mean values for the

various testosterone hydroxylase activities for five animals of each

sex and strain are presented in Tables 5 (specific activity or

catalytic activity/mg microsomal protein) and 6 (apparent molecular

activity or catalytic activity/nmole P-450). Standard deviations for

the 16 groups of animals ranged from 10 to 30 percent of the respective

means.

Earlier experiments (59) with hepatic microsomes from male and

female AKR/J mice afford background for assessment of the results

presented in Tables 5 and 6. 1) The six metabolites under scrutiny

comprised more than 65 percent of the total biotransformation products

of testosterone. 2) The formation of each metabolite was proportional

to time of incubation and microsomal protein concentration under the

conditions of assay. 3) The initial testosterone concentration in

these experiments was 131 pM; the apparent Km's for production of

each metabolite were less than 7 vM except for 15p-hydroxytestosterone

(50 pM) and 6a-hydroxytestosterone (80 pM). In no case was more than

30 percent of the testosterone consumed during incubation.

Even a casual examination of the results depicted in Tables 5 and

6 revealed certain striking findings. 1) The rates of hydroxylation of

testosterone at six positions were quite similar in the various strains

of animals, especially among males. For example, 6i-hydroxylase


































0
'-4



UV)
LLV)
SU UVl

C)
0L 0:




LU
cL.- t/)

L3- aV)





V)L0
Ci-)


LUJ
V) LL.t



CD-
L- 0




c ~Lu (
C V)-' *_
Qn 1- 0uU







Li (-LU



m~ Li
S I- I


a- Ul
03 Z 3



Q: F- m
3=~ Ot:


UJ QL- C-







cr LLJ


C)W
LL- C)








LOPm
0J 1








L41
0 UJ



U-
0 O (/
<: UJ
t/ -
UJU 3
LI S S
1- /IS
=> >


*4-


0
.-


a-



oL




4-)















0






E
0-~--:
5-


UO
f- 4-.



4-)
0) J


Cv,
->-4 'd-(


Flo
m C0

U3
cS 0








rS
r- .-
3: a-4 r~~

cv)
U.-,


0 0 0 0 0 0


LO m r- i ID 1.
























%do O o t N0- CM
*
0 r-4 ) L-4 r- C- CM





.0 C,) cO cJ r- O0

*
c C) C4 -, C.



Co N-I CI CO
r- -c co CM C 0
*







n N N- o 1 0 >o,








n
CM CM CM 1O C4

CM C








4-
(a





*
CCO 0



















10
'. ? 9 C C'
s-












0 d CO m 0 Cc N
"2) 9-4 .-4 C'' C m
-1 .-4 -x



44_>





LIi C) C:) C) C) C) C)
I I I I I I

Li L0 t.0 N- 0 '0 '.0 C
cr> co o o co r~< <




a> .-< X3 l/l
<" J=
4^ U
X rt3

IjD 0Ib-C
1-^ 3 3 ^ 3:3: 3: t
U-l J O O O O~ O
-1 ~ ~ ~ ~ < i i i i i In
co s a o a a a a.
t- L| ,_ r ^- ,r










46











AL



II n m o -0
0-4 M CM4 '-4 C4 C4n










I- 0 0
LW Ln
<-O n LO to CA LO L



C3 U- 4.3 0


v- N NIDi )
04-e u- *



'--1 0 ) 'd I-







< w
c U I .-

. C) L- M*0 *r-




C-) I (0













C. 0
I-) L L 0 -. J O
U- 0 O +L
-4 c uj _i N- .-4







C: ( C) C) -
3 -UL CC- c' c0
- LL O- 2O4






















4-)
0-0 LU C C) C) N CM
CO LA c a -
e o cc












-00 4
CC :Z 12 2

C I-ll CO I







a) r:in t1. u 1c
0 CU






t- SS -*


































































































< 1 I I I I I
s a S m ai ao
L -l LO U( r- LD LO to
LLL -


to
4-)
a)


O
l-

*C 4-
(n X
>r ci1
to
C W

4 -)


--
40 c4I








4e-
*F O
uu L




i- O















E
C



















t o -
S0
X 0
Co S-
U

V) U)

ov
X X


a)
c E
(0



.- I



re0

C) E
0O

o .C
4- u
0










activity was about 10-fold higher than other activities in all strains.

2) The rate of 16a hydroxylation proceeded about 2 to 3-fold faster in

male than female preparations of all strains. 3) The rate of most

other hydroxylations, especially 15a hydroxylation, proceeded more

rapidly in female than male microsomes, but the effect was

strain-specific. 4) For most comparisons, mean specific activities and

apparent molecular activities lead to similar conclusions because mean

P-450 content per mg microsomal protein varied little between strains

or sexes. Of the eight strains studied, only C3H/HeJ and C57BL/6J

demonstrated a consistent sex difference in hepatic cytochrome P-450

content. In these two strains P-450 content was 20-30% greater in male

than female microsomes.

Results of a two-way analysis of variance for five hydroxylase

activities in 80 animals (5 males and 5 females for 8 strains) are

presented in Table 7 in terms of R2 values. 7a-Hydroxylation is

omitted from detailed analysis because only about 60% of radiolabel in

its TLC zone was 7a-hydroxytestosterone (59). "Total" R2 represents

the percent of the total variability for a given hydroxylase activity

that can be accounted for by the sum of the effects of sex, strain and

the sex-strain interaction. Unusually high values were obtained for

15a- and 16a-hydroxylase activities in that 85-90% of all individual

variability in these activities could be accounted for by these

variables. By comparison, only 50% of the variability in

6&-hydroxylase activity could be attributed to the same factors.

Except for 6B-hydroxylation, the effect of sex, strain and the

interaction between sex and strain were all significant determinants of










































LUJ

Lii
-r
CLLI






ZV)

0.: 0 j

lI---

CCO
Xi 2 tt.





- I-W I--


=c -






Lu X>-
P-4 f

cCC
U-LL












-----
>c
WZ j


Iu
U)

3


r-

c


0o

I-
















*e
0
I-


4--




t-



+

x

CD





cV.)
4-)



*-
Ual



x,


0
4-



*'- X


o
14- 0

U, -0
0 >)
0i- =:


a ca B m
uD LO UO t0
9-m~


LO


u

4-
a)





O--
U U
C



0 r- C
0e 0

cc t--



4-- 4
a0 a-



I .u
0 m



0e

>- 4-
u >
W -0 C-

to '"
C 4-)
4-C *r


4O -
*- U W




a-> *S 4-1
ur U










0 (a)
4-





CA >.
o, 0


Or- 4+



aU CD
0..L) C .








4-





S-
U
4-
.4-
c4
ct
C.-
*r


a)
L

<,









U3
*er



U













.
0)
r-














0
u









4c-


UC)
T-





a)

LU0
L4.-










the rates of position-specific monohydroxylation of testosterone (P <

0.05). For 16a hydroxylation, sex was the most important determinant.

For 15a hydroxylation, the interaction between sex and strain (i.e.,

the effect of one factor depends upon the level of the other factor)

was most important. In the other cases, strain was the most important

determinant.

Male and female comparisons of testosterone hydroxylase activities

for each strain are presented in Table 8. Sex ratios are given only

when the differences were statistically significant. Statistical

analysis of either specific activities or apparent molecular activities

results in similar interpretations except in a few cases where the

differences between the sexes was 40% or less. As indicated in Tables

5 and 6, the rate of 16a hydroxylation was 2-3 times higher in the male

preparations of all eight strains studied. In contrast, female

activity was greater for most other hydroxylase activities, but

differences were strain-specific. The most prominent strain-specific

sex difference was observed for 15a-hydroxylase activity. In the

AKR/J, A/J, 129/J and DBA/2J strains, 15a-hydroxylation proceeded 2-3

times faster in female than male microsomes while in the other four

strains no significant sex difference was observed. Other experiments

(see below) have revealed that 15a-hydroxylase activity is not

dependent on phase of the estrus cycle.

In order to assess the statistic significance of differences

between strains with regard to rate of site-specific testosterone

hydroxylation, the activities of male and female microsomes were

analyzed separately by Duncan's multiple comparison procedure

























+
+ -4

r-*







.1 44 4c(


ND N Ln

<1 C; 0


44 -


'-I
4
c'j


v)





C-)
>UJr


I-
OQC




I--
L)LzJ
2i 1

U UJ





c)b
S00


UJ >-<
LU c~>-
I- c_1



m o-






(/32


LOLU
-LLC


>-
C)

V) Uj
CCH


0 -I

U J


>-
La:
LLI
-jL


Er




La

a)



0
rt
Ec














































or
5-
L.
0











LL
-5

c














4- 0

C0






C-


'I-

-5-
-1 Cj

Ul)
C-)






r co
C). *

C-)
<_?N


LL- Z

E LL- U-


cc


S-3





Ln CQ












-S.

m cc






mc
N CO
CM 0


t- -- -
t0 0 L 0
O-- -4 1-1
m- : ,1


4-)

4-)
Cr-
u
to



U ..-



r.-
r 0

fu 4-


c' 0









Le L
E 0

*o
, -






C )

c S0



a

u f
So
1- ~








0c







o* C
4-
4-) S-

O *r-















4- (U
4C 4--

LL4-
U 0

















1c4 +


'5
I-





OJ













0
t-
C-






5-

0
-v




c0
0
1/1










(Figure 6). Comparisons were done on apparent molecular activities

instead of specific activities for reasons enumerated below. Many

differences between the strains (a level = 0.05) were observed in the

hydroxylase activity at each of four sites on testosterone.

Hydroxylation at 7a was omitted from consideration for the reasons

described above; 68 hydroxylation was not analyzed because it did not

reveal a significant dependence on strain (Table 7). Those strains

with hydroxylase activities that were statistically indistinguishable

are joined by lines in Figure 6. Such groupings between males were

distinct from female groups, and female mice generally exhibited a

greater degree of polymorphism for most hydroxylase activities. This

sex-strain interaction was most pronounced in the rate of 15a

hydroxylation (Figure 6). For example, while male activities of the

C3H/HeJ, AKR/J and SWR/J strains were not distinguishable, female

15a-hydroxylase activity of the AKR/J strain was 3- to 4-fold greater

than that of the C3H/HeJ and SWR/J strains. Male 15a-hydroxylase

activity varied only 1- to 2-fold among the eight strains of mice.

Although interstrain variation in the other hydroxylations was small, a

2 to 3-fold strain difference could be found depending on the sex,

strain and hydroxylase activity in question. No obvious correlations

indicative of groups of hydroxylase activities were detected.

Possible interstrain relationships in the relative pattern of

testosterone hydroxylation at all six positions were explored by

expressing each hydroxylase activity as a percentage of the total

hydroxylase activity (6a + 6S + 7a + 15a + 15B + 16a) observed in a

strain. The overall patterns of hydroxylase activity were nearly

























8.0r 4D
O OAJ AJ
S *C3
SASW M
SO6 8)
860

- Uo 1J
Q IJ
6.0 3.0


0 X
4-




l b n we n
S4.0- t -2 o













R 1 0

0acti s fr te 1 rox ions o








testosterone. Males and females were analyzed separately by
|o *















the Duncan's multiple comparison procedure. The strains

linked by a vertical solid line were indistinguishable
statistically at the a level of 0.05. Abbreviations: A/J,
AJ; AKR/J, AK; BALB/cJ, RL; C3H/HeJ, C3; C57RL/6J, B6;
DBA/2J, D2; SWR/J, SW; and 129/J, 1J.










identical among the different inbred male mice (Figure 7) although

small differences were found for each hydroxylation. In all strains

the rank order of hydroxylase activity was 6B > 16a > 6c > 7a > 15a >

15B. In contrast, female patterns of testosterone hydroxylation were

quite different in different strains (Figure 8). Only the C57BL/6J,

DBA/2J, C3H/HeJ and SWR/J strains had similar rank orders of

hydroxylase activity. Qualitative variations from this pattern were

seen for certain hydroxylase activities in the other four strains.

Changes in the pattern of hydroxylation at the 15a and 16a positions

were observed for the 129/J strain, and at the 15a, 16a and 7a

positions for the AKR/J and A/J strains. A change in the rank order of

only 7a-hydroxylase activity was noted for the BALR/cJ strain. This

difference did not reflect the contaminant that chromatographs with

7a-hydroxytestosterone because the unknown metabolite was found to

remain in equal proportion to the 7a-hydroxytestosterone metabolite in

both sexes of all eight strains as evidenced by two dimensional TLC.

Indeed, two-dimensional TLC did not reveal any previously unidentified

major strain-specific metabolites in either males or females which

might have been undetected by routine one-dimensional TLC or which

might have interfered with quantitation of the six hydroxylase

activities of interest.

Testosterone hydroxylase specific activities and cytochrome P-450

content of untreated animals were found to fluctuate from month to

month by as much as 50% during the course of these and subsequent

experiments. This variation in microsomal activities of untreated

animals has been observed by others (41,62). Ford et a]. (26) have













































SITES OF TESTOSTERONE HYDROXYLATION


Relative rates of formation of 6a-, 6B-, 7a-, 15a-, 15B- and
16a-hydroxytestosterone (expressed as a fraction of the rate
of formation of these 6 metabolites) in the presence of
hepatic microsomes from males of 8 strains of mice. Note
the 10-fold difference in the scale for 68 hydroxylation.


Figure 7.


_~__ __ ~~~_
















































SITES OF TESTOSTERONE HYDROXYLATION


Relative rates of formation of 6a-, 68-, 7a-, 15a-, 158- and
16a-hydroxytestosterone (expressed as a fraction of the rate
of formation of these 6 metabolites) in the presence of
hepatic microsomes from females of 8 strains of mice. Note
the 10-fold difference in the scale for 68 hydroxylation.


Figure 8.


L~IU-UI-I~III~-Y- I -- -I- ____~~. ~~~. ~--__ ____.1_------11











reported a circannual variation in total hepatic hydroxylase activity

of mice towards testosterone; total testosterone hydroxylase specific

activity was reported to be 3-4 times higher in December than in April.

We did not observe a circannual phenomenon, and we are able to

compensate for the variability noted above by expressing results as

apparent molecular activities, defined here as nmoles of product formed

per min per nmole cytochrome P-450. ANOVA indicated that in virtually

all cases unaccountable variability in data obtained with 80 animals is

decreased with use of apparent molecular activities in comparison to

specific activities.










Discussion

This study in inbred mice of the relationship between strain, sex

and the sex-strain interaction and site-specific monohydroxylations of

testosterone provides evidence for both heterogeneity and polymorphism

in the isozymes of hepatic microsomal cytochrome P-450 and/or their

modes of regulation. With the exception of 68 hydroxylation, a

surprisingly large portion (73-93%, Table 7) of individual variability

can be accounted for by strain, sex and/or the sex-strain interaction.

The significance of strain or the sex-strain interaction as

determinants of catalytic activity in individual animals indicates

genetic polymorphism in the 6a, 7a, 15a, 15B and 16a hydroxylations of

testosterone. Discrete phenotypic classes were not distinguished

because of the limited magnitude of the interstrain variations in

hydroxylase activities.

Because of the prominent interaction between sex and strain, more

detailed evaluation of polymorphism necessitated separate consideration

of males and females (Figure 6). Despite the striking conservation

among microsomal preparations from males of pattern and rank order of

hydroxylase activities (Figure 7), there are statistically significant

differences between strains in four activities (6a, 15a, 158, 16a).

Catalytic activities toward selected positions on the testosterone

molecule vary by about two-fold, and usually not in parallel with

changes in other hydroxylases in the same strain. These results bear

not only on polymorphism, but also suggest extensive isozymic

heterogeneity. We do not have sufficient information to indicate

whether the polymorphisms we have observed reflect differences in










isozyme-proteins or genetic alterations at any of several possible

levels of regulation. Both possibilities have precedent in that Dieter

and Johnson (56) have described differences between steroid

hydroxylases isolated from different strains of rabbits whereas the

regulatory Ah locus described by Nebert and colleagues is widely

appreciated (63).

Polymorphism among preparations from female mice is more striking,

especially with regard to 15a-hydroxylase activity. In only four of

the eight strains were female 15a-hydroxylase activities significantly

higher than that in males of the same strain. This is reflected by the

prominent sex-strain interaction as a determinant of activity. There

is no apparent correlation between this polymorphism and the Ah locus

since two strains which exhibit this dimorphism, A/J and AKR/J, are

responsive and non-responsive to polycyclic aromatic hydrocarbons,

respectively (64). Figures 6 and 8 also reveal several other

statistically significant differences in various hydroxylase activities

among preparations from the various female mice. The differences are

greater than those seen between males, yet once again parallelisms in

groups of strains or activities are not apparent. On the basis of the

data presented in Figures 7 and 8, one might conclude that testosterone

hydroxylase activity is highly conserved between inbred strains of mice

and female factors, presumably endocrine, are capable of perturbing

this homeostasis. Subsequent studies revealed the independence of

results from estrus (unpublished observations). Moreover, other

experiments in castrated males (unpublished) accentuated polymorphism.











It is now our opinion that endogenous testosterone actually blunts an

underlying polymorphism that is expressed more fully in the female.

The best evidence for involvement of multiple isozymes derives

from the influence of sex on the various hydroxylase activities. We

clearly can distinguish independence of 16c-hydroxylase and

15a-hydroxylase activities from each other and from the other

activities. Several other experiments also indicate substantial

heterogeneity in the constitutive P-450 isozymes in mouse liver that

contribute to testosterone hydroxylation. These include the

differential effects of phosphate concentration (59) and endocrine

manipulations (65). Indeed, based on the approach developed by Kremers

et al. (48) in rats, we have already presented evidence (59) for the

contribution of more than one isozyme to both 16a- and 15a-hydroxylase

activities in mice. Several other reports bear on the relationship

between our results and constitutive and inducible cytochrome P-450

isozymes of mice. Coumarin hydroxylation exhibits a 5 to 10-fold

polymorphism in both untreated and phenobarbital treated mice (66), yet

this polymorphism is not reflected in the metabolism of a large number

of other xenobiotics or in the 6p, 7a or 16a hydroxylation of

testosterone. To date five or six different forms of liver microsomal

P-450 from mice induced with different xenobiotics have been purified

(3,67). Two of these isozymes, A2 and C2 prepared from

phenobarbital-treated B6D2F1/J mice, catalyze the 7a and 6B

hydroxylation of testosterone, respectively (42). Additional physical

evidence for multiple forms of cytochrome P-450 in the mouse has been

provided by the high performance liquid chromatographic elution









profiles of solubilized hepatic P-450's in untreated phenobarbital and

B-naphthoflavone treated mice (68). To our knowledge, a constitutive

form of cytochrome P-450 has yet to be purified from mice.

Sex-dependent differences in steroid and drug metabolism by the

hepatic microsomal monoxygenase system have been well characterized in

the rat (36,37,69,70). Reports of such differences in mice are sparse

and often inconsistent. Until now, most of the P-450 associated sex

differences reported for mice have been strain-dependent and show a

small female predominance (40,71,72). Our results for the 6a, 68, 7a,

15a and 158 hydroxylations of testosterone support these observations.

Strain differences in the female predominance of 15c-hydroxylase

activity are substantial and offer the opportunity to study the genetic

constraints on the sexual expression of different forms of P-450.

Preliminary studies (73) suggest that the regulation of this activity

is similar to that described previously (74,75,76) for the

15-hydroxylase system of female rats active toward 5a-androstane-3a,

178-diol 3,17-disulfate. The only significant sex difference we found

to be strain-independent was 16a hydroxylation, which occurred 2 to 3

times faster in male preparations from all eight strains. This finding

of male predominance is distinctive for mouse hepatic P-450

monoxygenase activities and may be characteristic of mice in general.

Steroid hydroxylations have been studied most extensively in the rat,

and 16a hydroxylation is higher in males with several steroid

substrates (48,77). Hepatic 16a hydroxylation of testosterone in the

rat has been shown to be catalyzed by a sex-dependent form of

cytochrome P-450 that is neonatally imprinted (38,78). This activity










is distinct from rat phenobarbital-induced 16a-hydroxylase activity

(22, 47). We have previously shown (59) in the AKR/J strain that these

sex differences in 15a- and 16a-hydroxylase activities are not

associated with differences in apparent Km values for testosterone.

One observation seems to speak against the rather extensive

heterogeneity implied in the foregoing discussion, namely, the finding

that the rate of each hydroxylation parallels a two-fold time-to-time

variation in the P-450 content of hepatic microsomes. This, of course,

had prompted us to analyze most results in terms of catalytic activity

per nmole P-450 (apparent molecular activity). Only if these

hydroxylations are catalyzed either by one or two forms comprising the

bulk of the constitutive P-450, or by several forms under identical

regulation with the bulk of constitutive P-450, (or that we and others

introduce an unknown methodologic factor that influences recovery of

all cytochromes P-450) could we anticipate all hydroxylase activities

to be equally affected. All of these remain valid possibilities;

recent observations (Singh, Hawke and Neims, unpublished results) with

mice treated with polyIC to induce interferon yielded results that bear

on the question. In this circumstance, P-450 content decreased about

50% and all hydroxylase activities were affected nearly in parallel.

It is our opinion that the circannual variation in total testosterone

hydroxylase activity reported by Ford et al. (26) reflects a related

phenomenon although no reference to P-450 content was made.

In summary, we suppose that several isozymes of P-450 are involved

in the six hydroxylations of testosterone we studied. Each isozyme

probably has more or less discrete substrate (or position) specificity










and total hydroxylase activity at each position reflects the

contribution of more than one isozyme. Wle further presume that each

isozyme in a given strain of mice is under more or less discrete

regulation (e.g. by sex), yet capable of parallel changes with the bulk

of constitutive cytochrome P-450. Finally, we have also presented

evidence of polymorphism in several of the isozymes, with respect to

enzyme-protein and/or its regulation. It follows that the strain

differences detected in this study may be indicative of a much greater

degree of genetic polymorphism that would be apparent in studies of the

isolated forms of cytochrome P-450.














CHAPTER THREE
DEVELOPMENT OF SITE-SPECIFIC TESTOSTERONE HYDROXYLASES IN HEPATIC
MICROSOMES FROM AKR/J AND BALB/cJ MICE

Introduction

The cytochrome P-450 monoxygenase system is involved in the

biotransformation of a diverse array of compounds, from drugs and other

xenobiotics to endogenous substrates such as cholesterol, sex steroids,

fatty acids and prostaglandins (79,80). Originally, cytochrome P-450

was thought to be a single, catalytically nonspecific enzyme. There is

now substantial physical, catalytic and immunological evidence that

cytochrome P-450 consists of a family of hemoprotein isozymes each with

different and more or less overlapping catalytic specificities

(3,4,67). Most of this progress in establishing cytochrome P-450

heterogeneity has dealt with the inducible P-450 isozymes. Some recent

reports have focused on forms of cytochrome P-450 from untreated rats

(46) and rabbits (45,53), partly because immunological techniques have

revealed a lack of cross-reactivity between various inducible forms and

the bulk of these constitutive isozymes (10,11,47)

The extent of, and the molecular basis for, heterogeneity within

the constitutive forms of P-450 remain largely unknown because each

isozyme likely occurs in relatively small amounts in hepatic

microsomes. Nonetheless, these forms are of particular interest

because of the possibility that their physiological functions) and










regulations may be quite different from those of the inducible

isozymes. The isozymes probably have greater specificity than the

inducible forms toward endogenous substrates such as steroids.

We are studying the constitutive forms of cytochrome P-450 in

untreated animals by examining and comparing the position specific

monohydroxylations of testosterone in isogenic strains of mice. The

use of inbred strains has already provided valuable information on the

genetic regulation, structure and temporal expression of the inducible

isozymes associated with the Ah cluster (63). Identification of

genetic polymorphisms in the various testosterone hydroxylations have

enabled us to explore some of the genetic and hormonal controls

applicable to the constitutive isozymes (73,81). These results in turn

have been useful probes of the multiplicity and specificity of certain

constitutive forms of hepatic cytochromes P-450 active toward

testosterone.

Differences in the developmental patterns of monoxygenase

activities have also provided insight into the heterogeneity and

regulation of P-450. While many groups of different enzymes can be

shown to share similar developmental patterns (82,83), the maturation

of different hepatic monoxygenase activities has been observed to vary

with substrate (84), sex (85), strain (86) and xenobiotic induction

(87). While comparison of the developmental patterns for several

testosterone hydroxylase activities should provide insight into the

heterogeneity of the constitutive forms of cytochrome P-450, comparison

of patterns between two strains of mice may provide even more valuable

information on genetic control of temporal expression. Developmental










control of the induction of specific forms of hepatic microsomal

cytochrome P-450 by temporal genes has been proposed in a number of

species (88,89). The existence of temporal genes which modify the

expression of lysosomal and enzymes during development has been

suggested by Paigen and colleagues (82,90).

In this study we describe position specificity in the maturation

of testosterone hydroxylation and report a strain difference only in

the development of 15a-hydroxylase activity. The latter finding

suggests the existence of temporal genes in control of the expression

of certain constitutive cytochrome P-450 activities.










Materials and Methods

Testosterone [4-14C] (56.1 pCi/pmol) was purchased from New

England Nuclear, Waltham, Ma. The radiochemical purity was greater

than 98% as established by TLC on silica gel F254 plates with the

solvent system ether/benzene/methanol, 55:44:0.5 (v/v). Testosterone

was obtained from Sigma Chemical Co., St. Louis, Mo., and was

recrystallized twice from acetone/hexane. Reference steroids were

obtained from sources described previously (59). Precoated (0.2 mm)

silica gel 60 F254 TLC plates (E. Merck, Darmstadt, Germany) were

purchased from Scientific Products, Brunswick, N.J., and SB-5 X-ray
film for autoradiography was obtained from Eastman Kodak Co.,
Rochester, N.Y. Liquiscint liquid scintillation mixture was obtained

from National Diagnostics, Somerville, N.J.; Emulgen 911 was purchased
from the Kao-Atlas Co., Tokyo, Japan. NADPH, D-glucose-6-phosphate and

EDTA were purchased from Sigma Chemicals, St. Louis, Mo. All other
materials and reagents were purchased from the Fisher Scientific Co.,
Pittsburg, Pa., and were of the highest quality commercially
available.

Animals. Timed-pregnant AKR/J and BALB/cJ mice were received
from Jackson Laboratories, Bar Harbor, Me., between the 13 and 15th
days of gestation. Pregnancy had been determined by the observation of
a vaginal plug the day after mating and by gross observation or
palpation after the tenth day of gestation (91). After spontaneous
delivery, females and their litters were housed in separate cages on
corncob Sani-cel bedding. The animals were maintained on a 12 hr
light/12 hr dark photoperiod and allowed both food (Purina Rodent Chow)
and water ad libitum. Mice were weaned at 21 days of age and housed










4-6 per cage thereafter according to sex. At designated ages at least

3 male and female litter mate pairs were randomly selected from

separate litters for study. The liver of each animal was investigated

individually except where indicated.

Preparation of microsomes. Mice were killed by decapitation

between 8 and 10 a.m., and livers were excised and rinsed immediately.

All preparative steps were performed at 4 C. Liver homogenates from

28, 35, 49 and 77 day-old animals were prepared from minced tissue in

ice-cold 150 mM potassium chloride-250 mM potassium phosphate, pH 7.4,

as described previously. Microsomes were prepared by differential

centrifugation (27). Homogenates from younger mice were prepared by

identical procedures except that 150-170 mg of liver from 2 day-old

animals (3 pooled livers) or 300-350 mg of liver from 10 or 21 day-old

mice (2 pooled livers) were admixed with a fixed volume (5 ml) of

potassium chloride-phosphate buffer. Microsomal pellets were

resuspended in 100 mM phosphate buffer, pH 7.4, to yield a final

suspension containing approximately 15-20 mg microsomal protein per ml.

Protein content of microsomal suspensions was determined by the method

of Lowry et al. (28).

Enzyme assays. Cytochrome P-450 content was measured by the

method of Omura and Sato (29) with a Beckman 5260 Spectrophotometer

after microsomal suspensions or whole liver homogenates were diluted

five-fold in 100 mM phosphate buffer, pH 7.4, containing 20% glycerol

and 0.2% Emulgen 911. Peak heights were measured in relation to the

steady baseline observed between 480-500 nm. Glucose-6-phosphatase










activity was measured by the method described by Baginski et al. (92)

using 0.8-1.0 mg microsomal protein per incubation.

Radiometric assay for hydroxytestosterone metabolites. The

rates of hydroxylation of testosterone at the 6a, 60, 15a, 150 and 16a

positions were determined by the radioassay procedure we described

previously (59). Briefly, 0.4-0.5 mg of microsomal protein were

incubated at 370 C in a shaking water bath with NADPH (1.0 pmol),

MgC12 (10.0 nmol), potassium phosphate buffer (final concentration,

0.1 M, pH 7.4) and [14C] enriched testosterone (0.131 pmol

including at least 0.5 pCi in 10 pl ethanol) in a final volume of 1.00

ml. Measurement of activities in preparations from 2, 10 and 21 day

old animals required 1.0-1.5 mg of microsomal protein per incubation.

Reactions were initiated by the addition of NADPH and terminated after

10 min by the addition of 5 ml of ice-cold CH2C12. Each sample was

then transferred to a 10 ml screw-top Teflon-lined test tube, shaken

vigorously for 30 min, and centrifuged for 15 min to separate phases.

The organic phase was removed and evaporated to dryness at room

temperature under a stream of nitrogen. Residues were resuspended in

50 pl ethyl acetate, and the radiolabelled metabolites of testosterone

were separated by TLC on plastic-backed silica gel plates developed

three times with ether/benzene/methanol, 55:44:0.5 (v/v).

Statistical analysis. Statistical significance of differences

between sexes was calculated by Student's t-test; P < 0.05 was

considered significant. All values are expressed as mean and standard

error.










Results

Postnatal development of hepatic microsomal cytochrome P-450.

As depicted in Figure 9, both male and female AKR/J mice experienced a

substantial increase in cytochrome P-450 concentration in microsomes

between days 1 and 21 of age (0.16 to 0.91 nmol/mg of protein).

Microsomal cytochrome P-450 concentration increased only slightly more

between days 21 and 77. Adult values were 5-6 fold higher than those

at age 2 days regardless of sex.

The recovery of total cytochrome P-450 and another microsomal

enzyme, glucose-6-phosphatase, was measured in preparations from 10 and

77 day old mice of each sex to determine if recovery of microsomes

varied with age. At both ages, 55-62% of homogenate

glucose-6-phosphatase activity was recovered in the microsomal

preparation. The recovery of cytochrome P-450 in the microsomal pellet

was 55-60% and 70-75% of the homogenate content from 10 and 77 day old

animals, respectively. Since the yield of microsomal protein per gram

of liver changed little with postnatal maturation (range at all ages,

20.7-32.7 mg/gram liver), the pattern of P-450 maturation depicted in

Figure 9 would be the same had one expressed cytochrome P-450

concentration in terms of liver weight instead of microsomal protein.

Postnatal development of microsomal testosterone hydroxylase

activities. The developmental patterns of both testosterone 16a- and

15a-hydroxylase activities are illustrated in Figures 10 and 11. In

2-day-old female mice the specific activity of testosterone

16a-hydroxylase (Fig. 10) was about 12% of the adult value. This

activity increased 6 to 7-fold in both male and female mice during the

























.-0

- -0 --


/


/
/
0'


I I
28 35


4
49


AGE (days)


Cytochrome P-450 content in hepatic microsomes from AKR/J
male and female mice of varying age. Each value represents
the mean of 3 determinations.


1.2-




0.8-

O-

0.4-




0


I Ir r


Figure 9.


I-U~~~~-------~-~---I ----- .



































0 9


^-


----9--'"


F~ ~


I I I I I
21 28 35 49 77

AGE (days)


Figure 10.


Testosterone 16a-hydroxylase activity in hepatic microsomes
from male ( ) and female ( o ) AKR/J mice of varying age.
Each value represents the mean of 3 determinations.


0.5







0.3-


0-







a.a.
M
E




E:


U

U-
s,


I I
2 10


- ---- ~---- -










3 weeks after birth. After 21 days of age female 16a-hydroxylase

activity tended to plateau while activity in males continued to

increase such that there was a statistically significant difference

between sexes by day 28. By day 77, a 3-fold male predominance in

16a-testosterone hydroxylase had been attained.

In contrast, during the first 10 days after birth when 16a- and

all other hydroxylase activities were increasing (see below), the

specific activity of testosterone 15a-hydroxylase activity in AKR/J

mice remained unchanged (Figure 11). This 10-day lag in the postnatal

maturation of 15a-hydroxylase activity was independent of sex and in

sharp contrast to the 3-fold increase in cytochrome P-450 concentration

that occurred over the same interval. It follows, of course, that

15a-hydroxylase activity per nmole cytochrome P-450 (apparent molecular

activity) actually decreased substantially between days 2 and 10 of

age. Between 10 and 35 days of age, 15a-hydroxylase activity increased

ten-fold, an interval during which cytochrome P-450 concentration only

doubled. Male 15a-hydroxylase activity remained relatively constant

after 35 days of age but activity in females increased steadily such

that a 3-fold female predominance was manifest at 77 days of age.

We have already reported that testosterone 6a-, 68-, 158-, and

especially, 15a- and 16a-hydroxylase activities exhibited sexual

dimorphism in adult AKR/J mice (59). In the present study, no

significant gender differences were observed before day 28, and most

were not apparent until after the 35th day of age. The developmental

curves for 63-, 6a-, and 15p-hydroxylase activities in male and female

AKR/J liver are presented in Figures 12 and 13, respectively. In order
































* cr
0 9


I I I I I
2 10 21 28 35 49 77
AGE (days)


Figure 11.


Testosterone 15a-hydroxylase activity in hepatic microsomes
from male ( ) and female ( o ) AKR/J mice of varying age.
Each value represents the mean of 3 determinations.


So

0.

E



1-.


L-

a
(n


_i I~~~ __I__ ___ ~___ _I__C______~__I_____IIICI1I




































2 10 21 28 35 49 77


AGE (days)


Figure 12.


Testosterone 6a-( o ), 68-( o ), and 158-( A ) hydroxylase
activities in hepatic microsomes from male AKR/J mice of
varying ages. Cytochrome P-450 content (nmoles per mg
microsomal protein) is indicated (....) for comparison.
Specific activities were scaled to converge at age 35 days
for the purpose of prepubertal and postpubertal
comparisons. The relative specific activity of 1.0 at age
35 days is equivalent to 0.387, 3.654 and 0.184 nmoles
product per min per mg microsomal for 6a-, 68- and 15s-
hydroxylases, respectively. Each value represents the mean
of 3 determinations.


_ _








































AGE (days)


Figure 13.


Testosterone 6a-( a ), 60-( o ), and 158-( A ) hydroxylase
activities in hepatic microsomes from female AKR/J mice of
varying ages. Cytochrome P-450 content (nmoles per mg
microsomal protein) is indicated (....) for comparison.
Specific activities were scaled to converge at age 35 days
for the purpose of prepubertal and postpubertal
comparisons. The relative specific activity of 1.0 at age
35 days is equivalent to 0.445, 3.930 and 0.223 nmoles
product per min per mg microsomal for 6a-, 6a- and 15B-
hydroxylases, respectively. Each value represents the mean
of 3 determinations.


__










to distinguish relative rates of maturation before and after puberty,

the scales for each hydroxylase activity have been set such that all

curves converge to a common point at age 35 days. Results for total

cytochrome P-450 are included for clarity of comparison. Male and

female 6a-, 68- and 150-hydroxylase activities gradually increased

ten-fold between days 2 and 35 of age, in parallel to the five-fold

increase seen in cytochrome P-450 concentration. After puberty, 6a-

and 15B-hydroxylase activities in female preparations increased to

values greater than the corresponding levels in male microsomes.

Postnatal maturation of testosterone hydroxylase activities in

BALB/cJ mice. We have previously reported that, unlike the situation

in AKR/J mice, no female predominance in any testosterone hydroxylase

activity can be detected in the BALB/cJ strain (81). Therefore, it was

interesting to compare the development of hydroxylase activities in

female BALB/cJ mice with the AKR/J. Cytochrome P-450 concentration in

microsomal preparations from BALB/cJ mice was approximately 2-fold

higher than in AKR/J on days 2 and 10 of age (Table 9). To facilitate

comparison, hydroxylase activities are presented in Table 9 in terms of

apparent molecular activities or hydroxylase activity per nmole of

cytochrome P-450. The specific activity for any hydroxylation can be

computed by multiplying apparent molecular activity by age-specific

P-450 concentration which is also listed in Table 9 for convenience.

Developmental curves for 6a-, 68-, 168- and 16a-hydroxylase activity in

BALB/cJ were similar to those seen in the AKR/J strain. The striking

similarity in rates of maturation of these activities between the two

strains between age 2 and 10 days is revealed in Table 10, which










presents results as ratios of apparent molecular activities at the two

ages in each strain. These ratios of apparent molecular activities

between day 10 and day 2 should approximate 1 if cytochrome P-450 and

hydroxylase activity mature in parallel. 16a-Hydroxylase activities

increased 1.6 times faster than P-450 in both strains, whereas other

hydroxylases increase at rates within 30% of that of the increase in

P-450. The maturation of 15a-hydroxylase activity in the two strains

is quite different. As noted above, in AKR/J mice there was no

increase in 15a-hydroxylase activity between days 2 and 10; indeed,

specific activity decreased by 9% during this interval. This of course

is consistent with the low (0.33) ratio of apparent molecular

activities at days 10 and 2. During the same age interval, the

specific activity of 15a-hydroxylase in the BALB/cJ increased by 84%,

nearly in parallel with the increase in P-450.

In BALB/cJ females, which do not exhibit the adult female

predominance in most hydroxylase activities that is characteristic of

AKR/J mice, apparent molecular activities decreased between days 10 and

77 in all mice as P-450 content continued to increase to adult levels

(Table 9).















TABLE 9
DEVELOPMENT OF HYDROXYLASE ACTIVITIES IN TWO MOUSE STRAINS IN
RELATION TO THE DEVELOPMENT OF CYTOCHROME P-450


Position of Molecular activities or P-450 content
Hydroxylation

AKR/J BALB/cJ

DAY 2 DAY 10 DAY 77 DAY 2 DAY 10 DAY 77

15a 1.57 0.52 6.29 1.08 0.85 1.80

16a 1.76 2.85 2.08 2.78 4.53 3.32

158 1.27 1.10 2.71 1.69 1.29 1.11

6a 2.17 2.60 5.31 1.92 2.47 2.37

60 25.70 27.60 37.22 23.51 24.17 23.90

Cytochrome P-450 0.16 0.44 1.15 0.33 0.77 1.10

aHydroxylase activities are expressed as nmoles product formed per 10
mfin per nmole cytochrome P-450. Cytochrome P-450 contents are expressed
in terms of nmoles P-450 per mg microsomal protein. All values represent
the mean of 3 determinations. All standard deviations were 10 to 20
percent of respective means.















TABLE 10
MATURATION OF EACH HYDROXYLASE MOLECULAR ACTIVITY
BETWEEN DAYS 2 and 10 OF AGE IN AKR/J AND
BALB/cJ MICE


Ratio of Molecular Activitiesa
Position of Hydroxylation Day 10/Day 2
AKR/J BALB/cJ

15a 0.33 0.83

16a 1.62 1.63

150 0.87 0.76

6a 1.20 1.29

68 1.07 1.03


aThe ratios of hydroxylase activities
nmole P-450; see Table 9) at ages 10
been computed for 5 activities.


(expressed per
and 2 days have










Discussion
Independent patterns of maturation of the 7a, 68 and 16a

hydroxylation of testosterone by rat liver microsomes were first

observed by Jacobson and Kuntzman (93). Recently, differences in the

developmental pattern of several monoxygenase activities in untreated

C57BL/6N male mice have been submitted as evidence for the involvement

of distinct isozymes in the various activities (84). We have

previously studied the effects of sex and strain on the site-specific

hydroxylations of testosterone in inbred mice (81) and now extend the

experiments to include development to gain further insight into the

heterogeneity and regulation of the P-450 isozymes.

Our results provide developmental evidence for at least three

functional forms of cytochrome P-450 involved in the specific

hydroxylations of testosterone, namely, one involved in 6a, 68 and 158

hydroxylation, another in 16a hydroxylation, and a third in 15a

hydroxylation. The observation that 6a, 68 and 158 hydroxylase

activities cannot be distinguished developmentally suggests either

coordinate development of distinct P-450 isozymes or involvement of a

single form of cytochrome P-450 with broad specificity toward the 6a,

63 and 150 and perhaps other positions of testosterone. The fact that

all three activities show female predominance in adult AKR/J but not

BALB/cJ mice is consistent with both explanations. The concept of

coordinate development of three different enzymes has already been

established in the case of lysosomal acid hydroxylases in mice (82,90),

and could well apply to 6a, 68 and 158 hydroxylase activities since










recent biochemical and strain studies suggest some independence between

these activities (59,81).

The existence of a distinct P-450 isozyme with activity

especially for 16a-hydroxylation of testosterone is suggested by the

more rapid development of this activity relative to the postnatal

increase in cytochrome P-450 content and the other hydroxylase

activities, and by the fact that testosterone 16a hydroxylation is the

only activity to experience a three-fold postpubertal increase in male

mice only. Both characteristics apply to both AKR/J and BALB/cJ

animals. In contrast, 16a-hydroxylase activity exhibits little or no

increase until after the fourth week of age in rats (69). In these

animals 16a-hydroxylase activity is androgen-dependent in rats and

directed toward several steroid substrates (20,48,77). The distinctive

postnatal and postpubertal patterns of 16a hydroxylation may well

reflect regulation of two different isozymes bearing 16a-hydroxylase

activity. We have already presented evidence for multiplicity in mouse

16a-hydroxylase (59); moreover, in the rat phenobarbital-induced and

androgen-imprinted 16a-hydroxylases have been solubilized and resolved

(22,38,47).

The development of testosterone 15a-hydroxylase activity also

reveals postnatal and postpubertal patterns of maturation that are

easily distinguished from those of P-450 and other hydroxylases, but in

this case the distinctions are strain-specific. In AKR/J mice

15a-hydroxylase activity did not increase during the first ten days

after birth, and there was a three-fold increase in only female

activity after puberty. Neither situation applies to development of










15a-hydroxylase activity in BALB/cJ mice. As with 16a-hydroxylase

activity, we may be dealing with more than one 15a-hydroxylase and

strain-specificity in their regulation. The threefold rise in

testosterone 15a-hydroxylase activity in female AKR/J mice after

puberty may signify the emergence of a sex-dependent female form of

cytochrome P-450 which mainly catalyzes 15a hydroxylation with some

activity toward the 6a, 68 and 15g positions of testosterone. Indeed,

Gustafsson and Ingelman-Sundberg have described a 15-hydroxylase in

female rat hepatic microsomes active on steroids and steroid sulfate

conjugates (74,75,76), the activity of which has been associated with a

52,000 dalton protein found only in females (94). One of the most

intriguing questions raised by our results is whether or not this

postpubertal (or sexually dimorphic) difference between strains relates

to the differences among strains in the maturation of 15a-hydroxylase

activity during the first ten days after birth. There is no obvious

reason to link the capacity for expression of an adult female activity

or protein, with the delayed expression of 15a-hydroxylase activity in

young male and female pups; yet both are characteristics of AKR/J but

not BALB/cJ animals.

The difference in rates of maturation of 15a-hydroxylase

activity between days 2 and 10 of age in AKR/J (-9%) and BALB/cJ (+84%)

are perhaps best interpreted in terms of temporal genes. Paigen (95)

has suggested that genetic elements, called temporal genes, determine

the developmental program for individual proteins, and strain

differences have been described (90,96). The responsiveness of

drug-metabolizing enzymes to specific hormones has been shown to change







84


during development (97,98), and there is evidence in mouse, rat, and

rabbit (88,89) for temporal control of the induction of forms of

cytochrome P-450 associated with the Ah locus. It will be interesting

in future experiments to explore the potential relationship between

polymorphisms we have observed in the early neonatal developmental

patterns and sexual dimorphism in the adult.













CHAPTER FOUR
GENETIC REGULATION OF HEPATIC MICROSOMAL TESTOSTERONE 15a- AND
16a-HYDROXYLASE ACTIVITIES IN FEMALE AKR/J AND BALB/cJ MICE

Introduction

The cytochrome P-450 monoxygenase system is involved in the

biotransformation of a diverse array of compounds including drugs,

carcinogens, pesticides, steroids and fatty acids (79,80). The

existence of different isozymes of cytochrome P-450 in mammalian liver

is now well established with the isolation and purification of a number

of forms which vary in subunit molecular weight, immunological and

catalytic properties, and amino acid sequences (3,4) from rat

(43,46,50), rabbit (45,51-53) and mouse (42,67). But the extent of

cytochrome P-450 heterogeneity, the degree of overlap in substrate

specificity between isozymes and the mechanisms by which these isozymes

are independently regulated by physiological and environmental factors,

are not well understood.

Individual animals and people can be quite heterogeneous with

respect to both environmental exposures and genetic factors known to

influence drug metabolism. Sometimes it is the combination of

environmental and genetic factors that influence monoxygenase

activities. Genetic differences in sensitivity to induction of P-450

isozymes by polycyclic aromatic hydrocarbons has been attributed to

polymorphism in the Ah locus or cluster by Nebert and Jensen (63).










Indeed, the induction of specific forms of cytochrome P-450 by various

inducers has allowed some progress into the molecular biology of the

monoxygenase system as evidenced by recent nucleic acid hybridization

experiments (5,6). Less is understood about the constitutive forms of

cytochrome P-450, here defined as those isozymes present in untreated

animals. Immunological techniques have already revealed a lack of

structural homology between the major inducible forms and the bulk of

the constitutive P-450 isozymes in hepatic microsomes of untreated

rats (10,11,47). It seems likely that these constitutive isozymes will

differ from the inducible P-450's with regard to functions) and mode

of biological regulation. Studies into the molecular biology of the

constitutive isozymes would be greatly facilitated by identification of

genetic polymorphisms in the regulation and/or structure of these

forms.

While several examples of genetic polymorphism of drug oxidation

in man have been described (e.g. 99-102), the mechanism underlying

these polymorphisms remains unknown. Preliminary studies suggest that

the impairment in debrisoquine 4-hydroxylation resides at the level of

the microsomal monoxygenase system (103). The recent demonstration of

polymorphism of debrisoquine oxidation in various rat strains has

allowed more detailed biochemical studies (104,105). Vlasuk et al. (9)

have found inter- and intra-strain polymorphism among four molecular

forms of closely related phenobarbital-induced cytochromes P-450 from

rats. Dieter and Johnson (56) have attributed a strain difference in

the 68 hydroxylation of progesterone by highly purified preparations of

P-450 from untreated rabbits to the presence or absence of one of two











subforms. Progress has likely been slowed by the happenstance that

oxidative metabolism of any drug represents the cumulative action of

more than a single form of cytochrome P-450. Polymorphism in a single

isozyme might then be masked in the background of other contributing

isozymes except in instances where oxidation is predominantly catalyzed

by an isozyme with great substrate specificity. The coumarin

hydroxylase locus (106) remains the only monoxygenase polymorphic

allele to be mapped in the mouse.

Constitutive forms of P-450 may be more specific in the oxidation

of endogenous substrates (45,46,107) than inducible forms. It follows

that polymorphism associated with alteration in the structure or

regulation of constitutive cytochromes P-450 might be more easily

detected with endogenous substrates than with xenobiotics. Indeed,

strain differences in steroid hydroxylation have been reported for many

species (26,54,55,56). While some patterns of inheritance have been

worked out (57,58), in most instances the exact nature of these

polymorphisms is not yet known.

The study of the regiospecific hydroxylations of testosterone in

isogenic mouse strains promises to provide valuable information on the

genetic and hormonal factors regulating the predominant forms of

cytochrome P-450 in hepatic microsomes from untreated animals. We have

previously reported genetic and sexual polymorphisms in the

testosterone 15i- and 16a-hydroxylase activities of mice (59,81). Our

present studies were undertaken to elucidate the genetic and endocrine

basis of the sex and strain differences in testosterone hydroxylation

at these two positions in inbred mice.










Materials and Methods

Materials. Testosterone [4-14C] (56.1 pCi/pmol) was purchased

from New England Nuclear, Waltham, Ma. The radiochemical purity was

greater than 98% as established by TLC on silica gel F254 plates with

the solvent system ether/benzene/methanol, 55:44:0.5 (v/v).

Testosterone was obtained from Sigma Chemical Co., St. Louis, Mo. and

was recrystallized twice from acetone/hexane. Reference steroids were

obtained through sources previously described (59). Precoated (0.2 mm)

silica gel 60 F254 TLC plates (E. Merck, Darmstadt, Germany) were

purchased from Scientific Products, Brunswick, N.J., SB-5 X-ray film

for autoradiography was obtained from the Eastman Kodak Co., Rochester,

N.Y., and Liquiscint liquid scintillation mixture was obtained from

National Diagnostics, Somerville, N.J. Emulgen 911 was purchased from

the Kao-Atlas Co., Tokyo, Japan. NADPH and testosterone propionate

were purchased from Sigma Chemical Co. All other materials and

reagents were purchased from the Fisher Scientific Co., Pittsburgh,

Pa., and were of the highest quality commercially available.

Animals. Male and female AKR/J, BALB/cJ and BALB/cJ x AKR/J Fl

hybrid mice (CAKFI/J; 6-8 weeks old, 20-25 g) were purchased from

Jackson Laboratories, Bar Harbor, Maine. F2 and backcross animals were

bred at the University of Florida's Animal Resources Facility. Mice of

the same sex were housed in groups of six or less per cage on corncob

Sani-cel bedding in an animal room with a controlled environment.

Smoking and the use of insecticides were prohibited, and a 12-hour

light/12-hour dark cycle was maintained. Animals were fed Purina

Rodent Laboratory Chow and tap water ad libitum. Mice were maintained










under these conditions for at least three weeks before study at the age

of 77 days. Orchiectomies, ovariectomies and sham operations were

routinely performed under methoxyflurane anesthesia at the age of 56

days. Adequacy of the operative procedures was confirmed by inspection

at the time of death. Where indicated, intact, sham-operated or

gonadectomized mice received daily subcutaneous injections of 100 pg of

testosterone propionate in 0.2 ml of sesame seed oil for seven days.

Control animals received vehicle only. Mice were killed 24 hours after

the last injection.

Preparation of microsomes. Mice were killed by decapitation

between 8 am and 10 am, and livers were excised immediately and

carefully separated from gall bladders. The livers were then rinsed

and homogenized in 5 ml/gram liver of ice-cold 150 nM potassium

chloride 250 mM potassium phosphate, pH 7.4, with a motor-driven

glass-Teflon homogenizer. Microsomes were prepared by differential

centrifugation (27). Briefly, the homogenate was centrifuged at 10,000

xg for 15 min, and the resulting supernatant was centrifuged for 1 hr

at 110,000 xg. Microsomal pellets from 1 gram of liver were

resuspended in 2 ml of 100 rMl phosphate, pH 7.4, and were used

immediately. Microsomal suspensions contained approximately 8-10 mg

microsomal protein per ml as determined by the method of Lowry et al.

(28). Cytochrome P-450 content was determined by the method of Omura

and Sato (29) with a Beckman 5260 Spectrophotometer after microsomal

suspensions were diluted ten-fold in 100 mM phosphate, pH 7.4,

containing 20% glycerol and 0.2% Emulgen 911. Peak heights at 450 nm

were measured in relation to the steady baseline at 480 to 500 nm.










Radiometric assay for determination of testosterone hydroxylase

activities. Testosterone hydroxylase activities at the 6a, 68, 15a,

158 and 16a positions were determined by a radioassay procedure

described previously (59). Briefly, 0.4-0.5 mg of microsomal protein

were incubated at 370 C in a shaking water bath with NADPH (1.0 pmol),

MgCI2 (10.0 pnol), potassium phosphate (final concentration, 0.1 M, pH

7.4) and [14C]-enriched testosterone (0.131 nmol, 0.5 pCi in 10 pl

ethanol) all in a final volume of 1.00 ml. Reactions were initiated by

the addition of NADPH and terminated 10 min later by the addition of 5

ml of ice-cold methylene chloride. Each sample was then transferred to

a 10 ml Teflon-lined screw-capped test tube, shaken vigorously for 30

min and centrifuged to separate phases. The organic phase was removed

and evaporated to dryness at room temperature under nitrogen. Residues

were resuspended in 50 pl ethyl acetate and the radiolabelled

metabolites of testosterone were separated by TLC on plastic-backed

silica gel plates developed three times in ether/benzene/methanol,

55:44:0.5 (v/v).

Calculation and statistical evaluation of results. Testosterone

hydroxylase activities were expressed as apparent molecular activities

or nmoles of product formed per min per nmole of cytochrome P-450

assuming that 1 nmole [14C] hydroxytestosterone was derived from 1

nmole of [14C] testosterone. This was done to minimize the impact of

variation in cytochrome P-450 content (81). We do not know whether the

relatively small day-to-day variation in P-450 content/mg microsomal

protein is biological or methodologic. Expression of results in terms

of specific activity or product formed per mg microsomal protein










increased intrastrain variation a small amount but did not change basic

conclusions. The fluctuation in cytochrome P-450 content is not due to

induction by environmental exposure to polycyclic aromatic hydrocarbons

since the AKR/J strain is non-responsive (64). Statistical analysis of

the data was performed according to Student's t-test; P < 0.05 was

taken as significant. All values represent the mean of five

determinations and are expressed with standard errors.

The number of genes involved in determining quantitative

differences in testosterone hydroxylase activity between the BALB/cJ

and AKR/J mouse strains was estimated by the method of Falconer (108),

where N, the number of independent loci involved, is given by the

following relationship:

N = 1/8 (mean parent 1 mean parent 2)2
(F2 variance Fl variance)










Results
Site-specific hydroxylation of testosterone by hepatic microsomes

from AKR/J and BALB/cJ mice. Testosterone 6a-, 6B-, 7a-, 15a-, 15B-

and 16a-hydroxylase activities of hepatic microsomes prepared from

untreated male and female AKR/J and BALB/cJ mice are summarized in

Table 11. The various hydroxylase activities of preparations from

AKR/J males differ from those of the BALB/cJ males by less than 30%.

In contrast, only the rate of 7a hydroxylation of testosterone was

similar in microsomes from female mice of the two strains. The rate of

15a hydroxylation was 2-fold higher in microsomes prepared from female

AKR/J mice than in those from female BALB/cJ mice. The opposite held

for 16a hydroxylation, which was 2-fold higher in hepatic microsomes

from female BALB/cJ mice. Microsomal hydroxylase activities at the 6a,

68 and 150 positions of testosterone were 50-62% higher in female AKR/J

than in female BALB/cJ mice.

Comparisons of males and females within each strain revealed a 2-

and 3-fold predominance in the 16a-hydroxylase activity in males of the

BALB/cJ and AKR/J strains, respectively. In the AKR/J strain female

predominance was observed in hydroxylase activities at the 150, 6M, 68

and especially 15a (3-fold) positions of testosterone; no sexual

dimorphism was exhibited in the BALB/cJ strain with regard to any of

these site-specific hydroxylations. Microsomal cytochrome P-450

content did not differ significantly between strains or sexes (in

nmoles P-450 per mg microsomal protein: AKR/J male, 0.71; BALB/cJ

male, 0.75; AKR/J female, 0.63; BALB/cJ female, 0.72). We now consider