- IQ - +--u «hr 9 mm 1.
.. -_.. a! , ”J O»; z a.
. ‘- o—npnmm.f-.In
Eknfig‘l ”’9- 31"!
This is to certify that the

dissertation entitled ‘
Effcc+ of DfoCe ren‘l’ The rmno «2 m c. 5+9+es
O r\ Se 166- +0? Deter/m1 Act—Ar; of TM are t 09
Hormone. AUHo r\

presented by . . .
Frant< grail—Q7 Hilljuri’ne'?

has been accepted towards fulfillment
of the requirements for

PA i D . degree in EO£SClQNC€ QM

Rman Mutt-1h on

Maw

Major professor

 

Date Q/[fi/gé
7 L

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

MSU

 

 

 

RETURNING MATERIALS:
Place in book drop to

 

“BRANES remove this checkout from
All-ICIIIII. ‘your record. ‘FINES will
be charged if book is
returned after the date
stamped below.
‘J’U'L”Zii"d?§l

1+9 K205

 

 

 

EFFECT OF DIFFERENT THERMOGENIC STATES ON SELECTED
DETERMINANTS OF THYROID HORMONE ACTION

By
Frank Bradley Hillgartner

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Food Science and Human Nutrition

1986

ABSTRACT

EFFECT OF DIFFERENT THERMOGENIC STATES 0N SELECTED
DETERMINANTS OF THYROID HORMONE ACTION

By
Frank Bradley Hillgartner

Obese (ob/ob) mice and low protein-fed rats exhibit impaired and
enhanced adaptive thermogenesis, respectively, relative to corresponding
controls that may be mediated, in part, by changes in thyroid hormdne
action at the level of the nuclear 3,3',5-triiodothyronine (T3)
receptor or the peripheral conversion of thyroxine (T4) to metabolically
active T3. To investigate this possibility, microsomal iodothyronine
5'-deiodinase activity, nuclear T3 receptor concentration and affinity
and endogenous T3 content associated with the nuclear receptor were
measured in lean and obese mice. Only iodothyronine 5'-deiodination
was examined in low protein-fed rats.

Kinetic analysis of hepatic and renal iodothyronine 5'-deiodinase
revealed that maximal enzyme activity was lower in obese mice as early
as 1-2 wks of age relative to corresponding lean mice while the Km of
the enzyme was similar in both phenotypes. The above finding suggests
that T3 availability to thermogenic target tissues may be impaired in
obese mice. Scatchard analysis showed that the maximal binding capacity
(Bmax) and equilibrium dissociation constant (Kd) of solubilized
nuclear T3 receptors prepared from liver were similar in both 4 and
8-l0 wk-old lean and obese mice, indicating that reduced thyroid
hormone action in the latter phenotype was not caused by alterations

in nuclear T3 receptor concentration or affinity. In contrast, the

Frank Bradley Hillgartner
concentration of hepatic endogenous T3 associated with the specific
nuclear receptor was 13 and 26% lower in 4 and 8-10 wk-old obese
mice, respectively. These data correlate with reported changes in
hepatic thyroid hormone-sensitive enzymes in obese mice, consistent
with a diminished nuclear T3 signal initiating thyroid hormone action.
Decreased endogenous nuclear T3 concentration may be caused by the
reduction in iodothyronine 5'-deiodination. Alternately, lowered
endogenous nuclear T3 content may result from reduced transport of T3
from the plasma to the nuclear compartment.

Maximal (hepatic+renal) iodothyronine 5'-deiodination per 100 g
body weight was inversely related to enhanced adaptive heat production
in low protein (5% casein)-fed rats. However, a causal relationship
between the above parameters is unlikely since similar alterations in
maximal 5'-deiodinase activity were observed in 8% casein-fed rats
which failed to exhibit changes in adaptive thermogenesis. Hence,
control'points regulating thyroid hormone action other than peripheral
iodothyronine 5'-deiodination may mediate elevated adaptive heat

production in rats fed a low protein diet.

To my parents for their loving support

ii

ACKNOWLEDGMENTS

I wish to thank Dr. Dale R. Romsos for his advice and guidance

during the course of these studies.

iii

TABLE OF CONTENTS

LIST OF TABLES. . . . . . . . . .

LIST OF FIGIJRES O I O O O O O O O O O O 0 O O O O O O O O O O

REVEW 0F LITERATURE. O I O O I O O O O O C O O O O O O O O O

REGULATION OF IODOTHYRONINE 5'-DEIODINATION

OBESE(Ob/Ob)MICE.............

Introduction . . . . .

Materials and Methods. . . .
Results. 0 O C O l O O O O O
1315011881011 0 o o o o o o o o

IODOTHYRONINE 5'-DEIODINATION IN RATS FED

LACK OF CORRELATION WITH ENERGY

Introduction . . . . .

Materials and Methods. . . .
ReSfltS. I O O O O O O I O 0
Discussion . . . . . . . . .

NUCLEAR TRIIODOTHYRONINE RECEPTOR

OCCUPANCY IN LEAN AND OBESE (Ob/0b)

Introduction . . . . .

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

GENERAL CONCLUSIONS . . . . . . .

LIST OF REFERENCES. . . . . . . .

BALANCE

BINDING

iv

LOW

IN LEAN

PROTEIN

AND

DIETS:

CHARACTERISTICS AND

MICE. .

Page

vi

15
27
35

35

41
50

LIST OF TABLES

Table Page
1 Effect of age on iodothyronine metabolism in lean and
obese mice ....................... l7
2 Microsomal protein content of tissues from lean and
obese mice ....................... 20
3 Plasma T4 and T3 concentrations in lean and obese mice. 22
4 Composition of diets .................. 38

5 Effects of low protein diets on kinetic parameters of
iodothyronine 5'-deiodinase .............. 42

6 Effects of low protein diets on tissue weight, micro-
somal protein content and maximal iodothyronine 5'-
deiodinase activity expressed on a unit tissue weight

and body weight basis ................. 46
7 Energy intake and retention in rats fed low protein

diets ......................... 47
8 Effects of low protein diets on serum thyroid hormones. 49
9 Binding characteristics of the hepatic nuclear T3

receptor ........................ 53
lo T3 content of hepatic nuclear extracts ......... 57

LIST OF FIGURES

Figure Page

1 Representative Lineweaver-Burk plots of hepatic and
renal iodothyronine 5'-deiodinase in 4 week old lean
and obese mice ..................... l6

2 Effect of environmental temperature on iodothyronine
5'-deiodination in lean and obese mice ......... 23

3 Effect of thyroid hormone administration on iodothyro-
nine 5'-deiodination in lean and obese mice ...... 26

4 Lineweaver-Burk analysis of iodothyronine 5'-deiodinase
in rats fed low protein diets ............. 44

5 Scatchard plots of binding of T3 to solubilized hepatic

nuclear receptors in 4 and 8-l0 week old lean and obese
mice .......................... 65

vi

REVIEW OF LITERATURE

Cold- and diet-induced thermogenesis are defective in obese (ob/ob)
mice. Regarding the former process, several studies have shown that
the stimulation of heat production resulting from acute cold exposure
is severely impaired in obese mice (63). As a result, these animals
are unable to survive prolonged periods of cold exposure at 4°C.
Focusing on the latter process, it has been reported that the thermogenic
response to feeding a highly palatable "cafeteria" diet is reduced in
obese mice compared to their lean counterparts (65). It has been
postulated that alterations in the above thermogenic mechanisms may at
least be partly responsible for the increased energy efficiency in
obese mice. This hypothesis is consistent with the findings that
maintenance energy requirements are lower in obese mice than in lean
mice housed at 33°C (37). Furthermore, the maintenance energy require-
ment of obese mice remains unchanged as environmental temperature is
lowered from 33°C (thermoneutrality) to 25°C-30°C, while in lean mice
the maintenance energy requirement is significantly increased by 35
percent (69). The obese mouse has been employed as an experimental
model in the study of human obesity since both syndromes share
numerous physiological characteristics (7). This animal model may
also have utility in more basic studies examining the hormonal
regulation of heat production during conditions of impaired thermo-

genesis.

In contrast to the obese mouse, another experimental model has
been developed for the study of thermogenesis during conditions when
heat production is enhanced. It has been shown that feeding a low

protein diet to either weanling or adult Sprague-Dawley rats results

in increased energy expenditure and reduced energy efficiency (50,62).
In general, rats fed a low protein diet will maintain a normal or
slightly elevated energy intake (expressed per 100 g body weight) but
will gain significantly less body energy than animals fed an adequate
protein diet. Feeding animals a variety of palatable foods (cafeteria
feeding) has also been shown to stimulate thermogenesis (52).

However, the latter feeding procedure has a disadvantage in that
dietary composition cannot be strictly controlled. This problem is
circumvented by the low protein feeding regimen since semi-purified
diets of defined nutrient composition are employed.

Adaptive thermogenesis induced by either cold exposure or diet
appears to be regulated by a number of factors. The sympathetic
nervous system is thought to play a prominent role in initiating and
maintaining cold- and diet-induced thermogenesis (51). Thyroid
hormones also appear to be required for thermogenesis and probably act
in concert with the sympathetic nervous system. Evidence that thyroid
hormones are necessary for cold-induced thermogenesis is provided by
studies with thyroidectomized rats which reveal that these animals are
unable to increase metabolic rate in response to cold exposure and
hence exhibit poor cold tolerance (55). In addition, thermogenic
response to norepinephrine administration, commonly used as an index
of the maximal capacity of cold-induced thermogenesis, is greatly
diminished in thyroidectomized animals (61). Both cold- and
catecholamine-induced thermogenesis can be fully restored by thyroid
hormone replacement. Regarding the role of thyroid hormones in

regulating diet-induced thermogenesis, several studies with lean rats

have shown that increased heat production resulting from feeding either
a "cafeteria" or low protein diet is associated with a marked elevation
in serum 3,3',5-triiodothyronine (T3) concentration (67). In obese
rats, which have an impaired thermogenic response to diet, the rise in
serum T3 is absent (79). It has been suggested that the elevation in
serum T3 concentration may be important in the development of diet-
induced thermogenesis (67).

The above findings raise the possibility that altered thyroid
hormone action may be responsible, at least in part, for changes in
cold- and diet-induced thermogenesis. In support of this hypothesis,
several biochemical processes which have been proposed as mechanisms for
adaptive heat production have been shown to be sensitive to thyroid
status. They include Na+, K+-ATPase activity (36), protein turnover
(15,48), futile cycling in carbohydrate metabolism (23,25), mito—
chondrial a-glycerolphosphate shuttle activity (68), and proton
conductance pathway activity in brown adipose tissue (66). Obese mice
exhibit reductions in Na+,K+-ATPase (36), a-glycerolphosphate
dehydrogenase (26,47) and proton conductance pathway activity (22)
indicating impaired thyroid hormone expression. Conversely, low
protein-fed rats exhibit elevations in a-glycerolphosphate dehydro-
genase and proton conductance pathway activity (50) suggesting enhanced
thyroid hormone action.

The expression of thyroid hormone action at the cellular level
appears to be initiated by the interaction of T3 with specific low
capacity, high affinity receptors in the nucleus (43). These receptors

have been isolated in the non histone protein fraction and shown to be

present in a variety of thyroid hormone-sensitive tissues. T3 is
considered the main metabolically active substance since approximately
85 percent of the nuclear bound iodothyronine in liver and kidney has
been estimated to be T3 and the remaining 15 percent thyroxine (T4).

A positive relationship has been demonstrated between nuclear T3
receptor occupancy in vivo and induction of thyroid hormone responsive
enzymes, indicating that the latter process may play a role in regu-
lating thyroid hormone action (44,60). Currently, there are no data
available for nuclear T3 receptor occupancy in either lean or obese
mice or animals fed a low protein diet.

There is evidence that thyroid hormone expression may also be
influenced by nuclear T3 receptor concentrations. For example,
starvation has been shown to reduce maximal nuclear T3 binding capacity
in rat liver which is associated with an impaired induction of hepatic
malic enzyme by T3 administration (13). Guernsey and Morishige (19)
have reported that nuclear T3 receptor concentrations are also lower
in liver and lung of obese mice than in lean mice. However, the above
study did not account for leaching of T3 receptors from chromatin
during the incubation procedure. The extent of leaching may have
varied in lean and obese mice and therefore may have introduced
significant error in estimating phenotypic differences in nuclear T3
receptor concentration. In addition, the age of the animals was not
reported. It would be of interest to know how early in the obese
syndrome changes in nuclear T3 receptor concentration occur. More
detailed studies employing improved methodology are clearly needed to

confirm the above findings. Recent studies employing solubilized T3

receptors from liver revealed that neither nuclear T3 receptor concen-
tration nor receptor binding affinity was affected by low protein
feeding in weanling rats (59). Since extraction of T3 receptors from
chromatin may have varied in different dietary groups, the above
findings await confirmation by T3 binding studies in which efficiency
of extraction of the nuclear T3 receptor is taken into account.

Another control point which may regulate thyroid hormone action
is extrathyroidal 5'-monodeiodination of T4 to T3. Approximately
80 percent of the circulating T3 is derived from intracellular 5'-
monodeiodination of T4 in the peripheral tissues (32). This reaction,
which is catalyzed by the enzyme T4 5'-deiodinase, may be of regulatory
importance in whole animal energy metabolism since the activity of this
enzyme can influence intracellular and circulating T3 concentrations and
hence, the level of nuclear T3 receptor occupancy in target tissues
mediating adaptive thermogenesis. Although the quantitative contribu—
tion of various tissues to T3 production from T4 in vivo is unknown,
T4 5'-deiodinase specific activity assayed in vitro is greatest in
liver and kidney (10). This microsomal enzyme has an apparent
Michaelis constant (Km) of mS-lO uM for T4, is activated by thiols in
a two-step transfer (ping-pong) mechanism, is inhibited by propyl-
thiouracil and can use reverse T3 (3,3',5'-triiodothyronine, rT3) as
an alternate substrate (mnm0.5tfl0 (72). To date, very little is
known regarding T4 5'-deiodination during conditions of altered
adaptive thermogenesis. Information on iodothyronine metabolism in
lean and obese mice is completely lacking.

Recently, Smallridge et a1. (59) have reported that hepatic and

renal 5'-deiodinase activities were not altered by feeding a low

protein diet to rats even though plasma T3 concentrations were
markedly elevated in these animals. The above enzyme assays were
performed with tissue homogenates in which significant non-specific
substrate binding to non-enzyme protein occurs. Under such conditions,
no relationship between protein concentration in the homogenate enzyme
preparation and enzyme activity can be demonstrated. Therefore, the
specific activity of 5'-deiodinase will vary depending on the protein
content of the homogenate. Measuring enzyme activity under these
conditions is invalid and may lead to spurious results when making
comparisons between treatment groups. Studies incorporating improved
methodology are needed to determine the effects of low protein feeding
on 5'-deiodination.

The purpose of the following studies was to examine whether
alterations in thyroid hormone action during conditions of impaired or
enhanced thermogenesis are mediated by changes in either nuclear T3
receptor occupancy, nuclear T3 receptor concentration, peripheral
iodothyronine metabblism or combination of the above. To study the
regulation of thyroid hormone action during conditions of enhanced
thermogenesis, comparisons were made between rats fed low protein or
adequate protein diets. During conditions of impaired thermogenesis,

comparisons were made between lean and obese (ob/ob) mice.

REGULATION OF IODOTHYRONINE 5'-DEIODINATION

IN LEAN AND OBESE (Ob/Ob) MICE

Introduction

 

Obese (ob/ob) mice exhibit increased energy efficiency relative to
their lean counterparts which may be attributed to a defect in non-
shivering thermogenesis (63,64) and diet-induced thermogenesis (65).
Recently, work has focused on the regulatory role of the sympathetic
nervous system in mediating reduced adaptive thermogenesis in obese
mice. This has lead to the proposal that sympathetic activation of
the proton conductance pathway in brown adipose tissue is markedly
impaired in obese mice (31). There is also evidence that thyroid
hormones may play a role in mediating reduced energy expenditure in
obese mice. For example, thyroid hormones are necessary for the
expression of non-shivering thermogenesis since thyroidectomized rats
exhibit poor cold tolerance (55), a condition also observed in obese
mice (63). Thermogenic response to norepinephrine administration,
commonly used as an index of maximal capacity of non-shivering thermo-
genesis, is also greatly diminished in both thyroidectomized rats
(61) and obese mice (64). Furthermore, stimulation of thermogenesis
by feeding a low protein diet is associated with an elevation in
serum 3,3',5-triiodothyronine (T3) concentration in lean rats (77);
whereas, in genetically obese rats, which have an impaired thenmogenic
response to diet, the rise in serum T3 is absent (79). Finally,
several biochemical processes which have been proposed as mechanisms
for reduced heat production in obese mice are sensitive to thyroid
status. They include Na+,K+ATPase activity (36), protein turnover
(15,48) futile cycling in carbohydrate metabolism (23) and proton

conductance pathway activity in brown adipose tissue (66).

Extrathyroidal 5'-monodeiodination of thyroxine (T4) to the
metabolically active form T3 has been identified as one of several
control points regulating thyroid hormone action (32). Approximately
80% of the circulating T3 is derived from intracellular 5'-mono-
deiodination of T4 in peripheral tissues. This reaction, which is
catalyzed by T4 5'-deiodinase, may be of regulatory importance in
whole animal energy metabolism since it influences intracellular and
circulating T3 concentrations and hence, the level of T3 nuclear
occupancy in target tissues mediating thermogenesis. The quantita—
tive contribution of various tissues to T3 production from T4 in vivo
is unknown, but T4 5'-deiodinase activity assayed in vitro is greatest
in liver and kidney (10).

In both liver and kidney, T4 5'-deiodinase is isolated in the
microsomal fraction (14,33) and requires the presence of thiol reducing
agents such as dithiothreitol for maximal activity (34,72); reduced
glutathione may serve as a cofactor in vivo. Liver and kidney are
also active in monodeiodinating 3,3',5'-triiodothyronine (rT3), an
inactive intermediate product of T4, in the 5' position to yield
3,3'T2. Similarities in the subcellular distribution of the T4 and
rT3 5'-deiodinases and in the effects of inhibitors on deiodination
suggest that a single enzyme catalyzes 5'-deiodination of both T4 and
r13 in liver and kidney (27,35).

It is evident that altered thyroid hormone metabolism may be an
important factor contributing to the defective expression of adaptive
thermogenesis in the obese mouse. Since little is known about thyroid

hormone metabolism in lean and obese (ob/ob) mice, we have examined the

effects of age, environmental temperature and T3 administration on the
kinetic parameters of 5'-deiodinase in hepatic and renal microsomes

of each phenotype.

Materials and Methods

 

Materials. The 125I-labeled iodothyronines, (3' or 5') L—T4 (60-80
uCi/ug) and (3' or 5') L-rT3 (600-800 uCi/ug) were purchased from
Abbott Laboratories (N. Chicago, IL). 125I-labeled L-T3 (l50 uCi/ug)
was purchased from New England Nuclear Corp. (Boston, MA).
Dithiothreitol (DTT) and unlabeled L-T4 and L-T3 were obtained from
Sigma Chemical Co. (St. Louis, MO). Unlabeled L-rT3 was purchased
from Calbiochem (LaJolla, CA). All other chemicals were reagent grade.
Animals. Male C57 BL/6J lean (ob/+ or +/+) and obese (ob/ob) litter-
mates obtained from our breeding colony were weaned at 3 weeks of age
and housed individually in plastic solid-bottom cages at 25°C. All
animals were provided a stock diet (Wayne Rodent Blox from Continental
Grain Co., Chicago, IL) and water ad libitum. A 12 h;12 h light-dark
cycle was maintained throughout the studies.

Experiments. In experiment 1, iodothyronine metabolism was assessed

in 5 age groups (1, 2, 4, 6 and 8-10 weeks of age) of lean and obese
mice to compare the pattern of enzyme development between phenotypes
before and after the onset of gross obesity in ob/ob mice. Pheno-
types of pre-weanling mice were identified based on differences in
oxygen consumption (6). Body energy content of these mice was also
measured (69). Oxygen consumption and body energy content of 1 week
old lean and ob/ob mice were 3768:186 and 2472:83 ul OZ/h/g body weight
(p<0.05) and 10.8:0.3 and 12.2:O.9 kcal/animal (NS), respectively. Data

10

for 2 week old lean and ob/ob mice were 35792114 and 24511106 ul
OZ/h/g body weight (p<0.05) and 16.9:O.7 and 24.5:O.8 kcal/animal
(p<0.05), respectively. Thus, there was a detectable increase in body
energy of the ob/ob mice examined at 2 weeks of age, but not in mice
examined at 1 week of age. Phenotypes of post-weanling mice were
identified visually.

Because the adaptive response in heat production to changes in
environmental temperature is different in lean and obese mice (63) the
effect of environmental temperature (14°, 25°, and 33°C) on iodo-
thyronine metabolism was determined in lean and obese mice (experiment
Four week old mice were placed in temperature-controlled chambers
maintained at either 14°C or 33°C for 2 weeks. Animals maintained at
25°C were housed in the animal room.

In experiment 3, the effect of T3 administration on iodothyronine
metabolism was determined. This experiment was conducted so that a
comparison could be made with other hepatic enzymes which have been
shown to give a different maximal response to thyroid hormone adminis-
tration in lean and obese mice. Four week old mice were employed
since most of the physiological changes associated with the obese

syndrome have occurred by this age without the appearance of massive

obesity which may have secondary effects on T3-induced enzyme response.

Lean mice were injected intraperitoneally with 5 ug T3/lOO g body
weight twice daily for 7 days. Obese mice were injected with the same
amount of T3 as their lean littermates. Based on the data of Kaplan
and Utiger (28), this dosage was sufficient to produce a maximal

induction in 5'-deiodinase activity. Control animals received an

2).

11

equal volume of vehicle solution (155 mM NaCl, 2 mM NaOH). On day 7,

mice were killed 4 h after their last injection.

Tissue Preparation. All iodothyronine deiodinase assays were conducted

 

using a washed, crude microsomal fraction obtained from liver or kidney.
To prepare this fraction mice were killed by cervical dislocation and
the liver and kidneys quickly excised, washed with homogenization

buffer and placed on ice. Each tissue was homogenized at 0°C in 4 vol
of 0.25 M sucrose, 0.05 M potassium phosphate, 0.01 M EDTA buffer, pH
7.4 using a Potter-Elvehjem glass homogenizer. The homogenate was
centrifuged at 15,000 x g for 10 min at 4°C. The resulting supernatant
was passed through 3 layers of cheesecloth and centrifuged at 100,000 x g
for 65 min yielding the crude microsomal pellet. The microsomal pellet
was washed twice before being resuspended in the homogenization buffer
to produce a protein suspension of 5-6 mg/ml and 2-3 mg/ml for liver
and kidney, respectively. After taking a sample for protein determina-
tion (39) 2 mM DTT was added to the suspension before it was rapidly
frozen in acetone and dry ice. Thereafter, the microsomal preparation
was stored at -80°C until deiodinase assay.

Enzyme Assay.

 

Incubation Procedure. Microsomal T4 5'-deiodinase (T45'-D) and

 

rT35‘-deiodinase (rT35'-D) were measured using a radiolabeled tracer
procedure. The generation of radiolabeled T3 and iodide from T4 and
rT3, respectively, was monitored in each reaction. Both T45'-D and

rT35'-D were assayed separately in hepatic microsomes. Only rT 5'-D

3
was assayed in renal microsomes since the activity of T45'-D was very

low in this preparation.

12

Based on preliminary studies, optimal conditions were established
for T45'-D activity. The results indicated that T45'-deiodination was
maximal at pH 6.8, 37°C and 2 mM DTT using a 0.05 M potassium phos-
phate, 0.01 M EDTA incubation buffer. In addition, T45'-deiodination
was linearly related to microsomal protein concentration from 0.1 mg/ml
to 0.4 mg/ml during a l h incubation at a T4 concentration of 2 uM.
Hence, T45'-D activity was assayed using the above reaction conditions
and a microsomal protein concentration of 0.35 mg/ml. Enzyme reactions
were carried out in polystyrene culture tubes (75 x 12 mm) with
various amounts of unlabeled T4 added to 1251-14 (1.25 uCi/tube) to
give T4 concentrations ranging from 2 uM to 13 uM in a total incubation
volume of 285 ul. Incubations proceeded for 30-120 min at 37°C under
N2 gassing. Reaction times were adjusted so that less than 12% T4 was
converted to T3. Reactions were terminated by plunging the assay
tubes in cracked ice and immediately adding 0.5 m1 ice-cold ethanol.

rT3ES'-deiodination in hepatic microsomes was measured in 0.05 M
potassium phosphate, 0.01 M EDTA, 5 mM DTT, pH 8.0 at protein concen-
trations ranging from 0.01 mg/ml to 0.10 mg/ml. The above pH for the
incubation buffer was chosen based on the findings of Visser et al. (71)
showing that a pH 8.0 phosphate buffer gave the highest Vmax for
rT35'-D in hepatic microsomes compared to pH 6.5 and 7.2. Our own
observations revealed that rT35'-D required a slightly higher DTT
concentration (5 mM vs 2 mM) than T45'-D for maximal enzyme activity.
The reaction tubes contained 125I-rT3 (1.0 uCi/tube) and unlabeled rT3

to provide final rT3 concentrations ranging from 0.125 uM to 3.0 uM.

Enzyme reactions were carried out for 7-20 min at 37°C under N2

l3

gassing. The assay for rT3S'-deiodination in renal microsomes was
similar to that in hepatic microsomes except for two alterations in
the incubation buffer; 0.25 M sucrose was added to the 0.05 M
potassium phosphate, 0.01 M EDTA and 5 mM DTT buffer, and the pH was
adjusted to 7.4 rather than 8.0. These alterations of the incubation
buffer were made because the small amount of protein recovered from
kidneys of the mice required that a relatively large volume of original
suspension buffer be added to the incubation buffer to provide adequate
amounts of enzyme for assay. To minimize effects of adding relatively
large and varying amounts of suspension buffer to the incubation
buffer we employed an incubation buffer similar in composition to that
of the original suspension buffer. The range of microsomal protein
concentrations and incubation times for the renal rT3 5'-D assay were
0.015 mg/ml to 0.10 mg/ml and 7 min to 15 min, respectively. Reaction
times and/or protein concentrations for all rT35'-D assays were
adjusted so that less than 16% of the substrate was consumed during the
incubation. rT35'-D reactions were terminated by plunging the assay
tubes in cracked ice and adding either 0.5 m1 ice-cold ethanol or
0.1 ml 2% bovine serum albumin followed by 0.5 ml 10% trichloroacetic
acid depending on which method of enzyme product analysis was employed.
Since T3 is a reaction product of T45'-D, alterations in T3
degradation are a potential source of error when comparing hepatic
T45'-D among treatment groups. Therefore, T3 degradation was measured
in hepatic microsomes by monitoring the disappearance of 125I-T3.
Details of the procedure and reaction conditions for this assay are

the same as those for the T45'-D determination, except that T3

l4

degradation was measured at only one substrate concentration (0.35 uM
T3).

Analysis of Incubation Products. To determine the percentage
generation of T3 and iodide from T4 and rT3, respectively, and the
percentage degradation of 13, enzyme reactants and products were
separated by descending paper chromatography employing the solvent
system of hexane-tertiary amyl alcohol-2 N ammonia (1:10:11) (2).
Briefly, ethanolic extracts of reaction mixtures were spotted on
Nhatman no. 1 paper strips (Whatman Inc., Clifton, NJ) along with
appropriate carrier compounds. After development, T3 and T4 spots
were visualized with 0.5% ninhydrin while a 0.1% palladium chloride
solution was used to identify the iodide spot. Next, the paper
strips were cut into sections which were then measured for radioactivity
using a gamma counter. Values for the fractional generation of
reaction products or the fractional degradation of substrate obtained
from tissue-free control incubations were subtracted from the observed
experimental values of tissue-containing incubations to correct for
any non-enzymatic reactions and for 125I-labeled reaction contaminants.

In some experimental groups, iodide release from rT3 was measured
using the ion exchange method of Leonard and Rosenberg (35) rather than
paper chromatography. In this procedure, trichloroacetic acid extracts
of reaction mixtures were applied to Dowex SON-X8 columns (BioRad
Laboratories, Richmond, CA) equilibrated with 1.74 M acetic acid. Free
iodide was eluted from the column with three column volumes of 1.74 M
acetic acid divided into two aliquots. Results of the above method were

similar to those obtained using the paper chromatography procedure.

15

Treatment of Enzyme Data. Kinetic analysis was performed on data

 

derived from the T45'-D and rT35'-D assays. Enzyme activity (pmol
product formed/mg microsomal protein/min) was determined at each
substrate concentration and double reciprocal plots were constructed.
Slopes and intercepts of the linear plots were determined using
unweighted least squares analysis. In calculating 5'-deiodinase
activity, it was assumed that 125I-T4 was randomly distributed between
the 3' and 5' positions of 125I-T4 and 125I-rT3. Hence the observed
fractional generation of T3 and iodide from T4 and rT3, respectively,

was doubled to obtain the actual values.

Plasma T4 and T3 Determinations. Immediately prior to killing the

 

mice, blood samples were obtained from the orbital sinus for plasma T4
and T3 analysis. Plasma T4 was measured using a solid phase RIA
system supplied by Becton Dickinson Immunodiagnostics (Orangeburg, NY).
Plasma T3 was assayed based on the method of Nejad gt_gl, (41) which
uses charcoal-dextran to separate bound from free hormone.
Statistical Analysis. The data were subjected to analysis of variance.
Statistical comparisons were made with the Bonferroni t test or
Student's t test as noted in table footnotes and figure legends.
Results

Effect of Age on iodothyronine Metabolism. Effects of age and pheno-
type on the kinetic parameters of 5'-deiodinase in liver and kidney
derived from Lineweaver-Burke plots are summarized in Table 1. Sample
double reciprocal plots are presented in Figure l for 4 week old mice.
In examining the pattern of development of hepatic 5'-deiodinase
activity in mice, a large increase in T45'-D activity was observed

between 1 and 2 weeks of age. V x values could not be compared at

ma

l6

 

v- LIVER
I
z 3 2,0 , T4 5’-o
o "g --LEAN
g \ 240 . «roses: )1
c g
‘5‘ 3 zoo L
c: o
e a 160 P {fl
,2 E 120 » xv"
\ \
F - '
o no » f
E
5 40'3"

 

 

 

 

-.2 -.1 .1 .5 .3 :4 .1.
1/[T4] (uM'1)

 

 

LIVER
'- /
L‘ 14.TT35'D '
s x
o
8% 12» ,
( O
m fill
_l 2 a
I“ Q
a: a: e
a E
v- ‘5 4’
E 2
IIILL- --

 

 

 

 

-53-. 1 234SOTBO
1/[r1'3] (uM'1)

 

 

 

 

 

,. KIDNEY
I /_ ,x'
E 14 L I'T3 5 D id
3 S 12
2 E 10 .
3 o
‘- a
g a
a 5' °’
\
v- 2 43,.

 

4-.-; i2345CTBO
1/[r‘l’3] (uM’I)
Figure 1. Representative Lineweaver-Burk plots of hepatic and renal

iodothyronine 5'-deiodinase in 4 week old lean and obese
mice. Each line is derived from data of 5 animals.

I7

I

.umc_seoumu no: .oz .»_m>_uumammg mu_s mmmno new cmm— no; Amo.oqu :_s\:_muoca ms

\_osa co.on—.p use so.onmo._ mo: .Aepz: o.mv co_umcu:mu:ou muocumasm opm:_m a um umxmmmo .o1.m¢H+ .ummu1u
m.ucon:um an um=_Ecmume me man meow one we move camp we omega zoom Amo.ovav u:mcmaw.u one mcomza .mu:o_go>
mo mama—ace an um:_scmumu mm Amo.ovav ucwcmmuwu one mcmuum. ammo Luzop qugumgmaam ucmcmae_u zu_3 mc__ mama
on» :o meow: .=_E\:_muoca ma\mpo5w mm ummmocaxm m_ Amer: mm.ov =o_uocu:mucoo museumnam m_m:_m a an omxammo
cowuoumcmmu mp .»_m>_uumammg .z: ecu cwe\:_muoeq ms\m_o5a mm commonaxm one use m_mx—acc mxczm1cm>am3mcwn
an cm:_eLmumu mam: o-.mmwc use o-.mep zuon mo Ex can me> one .mu_e o1m mo towncoms ecu acmmmgamc mm:_o>

 

avo.vov.o nmo.onv.o a—o.0wvm.o mmo.o«mm.o oz wmmao
aao.oama.o amo.oapm.o amo.onm.o ammo.oH_a.o a: coon
-5x o-.mmho so=u_g-
uvpfiem— m—Hoow ouoom omwmmw oz mmmno
1 nun . «a 1 «no 1
uop+mm~ aPNHNQN mop+mmm mve+mmm oz :omo
-xaE> a-.mMPa so=e_x-
auam.o_nn.¢~ ao~.mnm.mc an~.mfim.om am.mnm.¢op m—.ene._m «mono
n.vpno.om— p.opwm.oe— m.m—Hu.pvp N.oflo.NNP m.oao.—m como
a n a m a a
1:o_umuocmmo » em>_41
avo.onmo.o mmo.oH—p.p moo.oHo—.P naoo.ofimo.— coo.oHe—.p mmmno
vo.onm.o o~.oHoo.o ammo.oHoo.o moo.oHN—.— oo.oHoF.— ammo
a on m a
15x o1.m PL Lw>w4u
.oamamme .ommaaem .ameomom .n_maaaa . .am_a~mw omoao
a_~a~mo_ noNNaoo_ noopama__ nogaaam aomamam coon
-xas> o-.mmho Lo>_4-
oz mo.owm.—p mm.pno.o— am.onm.m oz mmmao
oz mm.owc.pp ac.owm.p— mo.—Ho.o oz cum;
-sx o-.mae oo>an-
oz amm.—Hp.—N «m~.¢Hm.oN mm.~wo.vm + mmmoo
oz no.~um.¢e no.¢uo.o¢ mo.eno.mm + sumo

-xaE> o.meh Lo>_n-

 

op1w o c N _
1111 maxuocmza

 

x: .om<

 

wows omooo new comp cw sm__onmums m:_=ogxguouop co moo ao uummwm .P m_nmp

18

125 125
I-T4 to I-T3

the above ages since the fractional conversion of
at 1 week of age was too low to monitor at substrate concentrations
greater than 2.0 uM T4. However, if T45'-D activity at 2.0 uM T4 is
considered, 2 week old lean and obese mice had 237% and 222% higher
enzyme activity, respectively, than respective 1 week old animals.
After 2 weeks of age, maximal T45'-D activity in lean mice continued

to increase until a plateau was reached at 4 weeks of age, while in
obese mice maximal activity was already evident at 2 weeks of age.

In general, the developmental profile of hepatic rT35'-D was similar
to that of T45'-D in lean and obese mice. Hepatic rT35'-D in lean and
obese mice peaked at 2 weeks of age and remained high in the former
phenotype for the remainder of the experimental period but declined
significantly in the latter phenotype by 6 weeks of age. In all age
groups, hepatic 5'-deiodination was significantly lower in obese mice
than in respective lean mice. Furthermore, differences in both
T45'-D and rT35'-D activity between phenotypes were greater in post-
weanling mice than in pre-weanling mice.

Developmental alterations in the velocity of renal rT35'-D followed
a general trend of declining maximal activity with increasing age in
both lean and obese mice (Table 1). Renal rT35'-D activity was
significantly lower in obese mice at 2, 4 and 6 weeks of age relative
to their lean littermates. These differences in renal rT35'-D activity
between lean and obese mice were greatest at 2 weeks of age and became
smaller with increasing age. No difference in enzyme activity was

observed at 8-10 weeks of age.

19

Km values for hepatic and renal 5'-deiodinase were similar between
phenotypes in each age group (Table 1). However, in a few instances,
significant differences in the Km of hepatic and renal rT35'-D were
observed in different age groups of a given phenotype. Since the
magnitude of the above differences was small, it is probable that they
are of little regulatory importance.

Also presented in Table l are the effect of age and phenotype on
hepatic T3 degradation. In general, hepatic T3 degradation in lean
and obese mice followed a similar pattern of development as that
observed for hepatic 5'-deiodination. In addition, T3 degradation was
lower in obese mice than in lean mice in each age group, but only to a
significant extent in post-weanling mice.

Data for microsomal protein content (mg protein/organ) for liver
and kidney of each age group are presented in Table 2. In both lean
and obese mice, microsomal protein content in liver and kidney
progressively rose with increasing age until peaking at 6 weeks of
age. Between 6 and 8-10 weeks of age a slight decline in hepatic
protein content was observed in both phenotypes while renal protein
content remained constant. Hepatic protein levels were similar between
lean and obese mice at l, 2 and 4 weeks of age. However, at 6 and
8-10 weeks of age when gross hypertrophy of the obese liver was readily
apparent, microsomal protein content was greater in obese mice than in
lean counterparts. No significant differences in renal protein
content were observed between phenotypes at any age investigated.

Employing the data in Table 2, the calculation of maximal

5'-deiodinative capacity per organ (Vmax x microsomal protein/organ)

.P mpnmh cw umcwemc mam m_ons>m new mgmuump “nanomgqum new mcowumw>mgan<

.cmmeo\cwmaoaq as me ummmwgaxm mam :Pmpoca Fmsomoguws no» mmzpm>

 

 

 

um_.onop.m uo_.on_o.m amo.onmm.~ meo.onom.o oz «mono
omp.onom.m omm.onmm.m aao.onop.m woo.onom.o oz com;
.355.
amo.an.eN «uF.FHP.om um.onm.m_ n~.onm.e mF.oH_.~ mmmno
mm.onu.om um.onp.mm um.onm.e— np.onm.m up.onm.~ ammo
1Lm>wn1
op1m m c N _ maxuocmza
x3 .wm<

1F

wows mmmno ecu comp Ease mmammwu we ucmucou :Pmuoga pmsomocuwz .N epoch

21

could be performed. In general, maximal 5'-deiodination per liver was
significantly lower in obese mice than in lean mice at all age groups
studied with differences between phenotypes being greater in older
animals than in younger animals (results not shown). 5'-deiodinative
capacity per kidney was also significantly lower in obese mice at all
age groups except 8-10 weeks. The above differences in maximal 5'-
deiodinative capacity per liver or kidney between lean and obese mice
were mainly the result of phenotypic alterations in 5'-deiodinase
specific activity in these respective organs.

In Table 3, plasma T4 and T3 concentrations for 4, 6 and 8-10 week
old mice in experiment 1 are presented. Relative to lean mice, plasma
T4 concentrations in obese mice were significantly lower at 4 and 6
weeks of age, but were not different in 8-10 wk old mice. Plasma T3
concentrations were similar in lean and obese mice of each age group.

Effect of Environmental Temperature on Iodothyronine Metabolism.

 

Kinetic analysis of T45'-D and rT35'-D activity was performed in mice
housed at 14°C, 25°C and 33°C (Figure 2). Exposing lean and obese
mice to 14°C for 2 weeks stimulated maximal hepatic T45'-D activity by
89% and 80%, respectively, relative to mice housed at 25°C. Compared
to their lean littermates, hepatic T45'-D activity in obese mice was
similarly lower (55% and 53%) at 14°C and 25°C, respectively. Housing
lean mice at 33°C produced a 83% decline in hepatic T45'-D activity
compared to lean mice housed at 25°C. No data for obese mice housed
at 33°C are shown since hepatic T45'-D activity in obese mice at 33°C

was too low to accurately measure.

22!

 

 

 

Table 3. Plasma T4 and T3 concentrations in lean and obese mice
Concentration
T4(ug/dl) T3(ng/dl)

4 wk age §§2££imsn£_l

Lean 3.75:0.23a 105:Ba

Obese 2.67:0.18a* 85:6a
8 wk age

Lean 4.13:0.20a 131212a

Obese 3.57:0 27b 127:10b
14°C Experiment 2

Lean 3.52:0.15a 1281136

Obese 1.91:0.19a* 108ma
25°C

Lean 3.90:0.13a 103166

Obese 2.89:0.34b* 116:9a
33°C

Lean 3.58:0 18a .95:9a

Obese 1.72:0.lla* 68:6b*
Control Experiment 3

Lean 3.22:0.13 120:10a

Obese 2.81:0.23 115:9a
T3-treated

Lean + 871:34b

Obese + 1410 65b*

 

Plasma T4 and T3 concentrations are presented for Experiment 1 (effect
of age), Experiment 2 (effect of environmental temperature) and

Experiment 3 (effect of T3 administration).
5-6 animals.

Values are means:SEM of
Hormone concentrations for 6 week old mice in Experiment

1 are the same as those presented for mice housed at 25°C in Experiment
2. Means with different lower case superscript letters within the same
phenotype are different (p<0.05) as determined by analysis of variance.
*Means are different (p<0.05) from those of lean mice of the same treat-
ment group as determined by Student's t-test. +Honnone concentrations
were too low to be detected.

23

 

 

'°’ " LIVER 1 1 ' KIDNEY 1500
T4 5’-o rT35’-D

     

   

< 400

 

 

 

 

 

 

 

 

 

 

 

 

'0 LEAN
,_ 4;. Bones: 300
, u
C 301 . ‘2“ 3
E 2
1; 15 c 100 a
5 10. NA 3
3 14° 23° 33° 14° 23 33° -
° 9.
‘6 2.
r a
E 1730 LIVER «0.30 _,
a a '
0 "'°’ 0.20 .'.
30001
ons
730
500 040
C t
2301 c 1 0.05
1.57""?
33°

   

ENVIRONMENTAL TEMPERATURE

Figure 2. Effect of environmental temperature on iodothyronine 5'-
deiodination in lean and obese mice. Vertical bars
represent mean values for Vmax (pmol/mg protein/min)
whereas the numerals at the base of each bar denote Km
values (uM T4 or rT3). In the case of T3 degradation
vertical bars represent enzyme activity assayed at a
single substrate concentration of 0.35 uM T . Bars
within phenotype with different letters are different
(P< 0.05) as determined by analysis of variance. *Means
are different (P<’0.05) from those of lean mice housed
at same temperature as determined by Student's t test.

1 Km was significantly different (P‘=0.05) from values
observed in mice of same phenotype housed at 14 and 25'C.
There were no significant differences in Km between
phenotypes at a given environmental temperature. NA,
data were not available because enzyme activity was too
low to accurately measure.

24

The response of hepatic rT35'-D to environmental temperature
followed a similar pattern as T45'-D in both phenotypes housed at 14°
and 25°C (Figure 2). Housing lean and obese mice at 33°C produced a
marked decrease in hepatic rT3 5'-deiodination of 76% and 88%,
respectively, relative to mice housed at 25°C. Since the magnitude
of the above temperature-induced responses was different in lean and
obese mice, the percentage difference in hepatic rT35'-D activity of
obese mice housed at 33°C (-7l%) relative to their lean littermates
was much larger than the differences between lean and obese animals
housed at 14°C (-44%) or 25°C (-46%). The effect of environmental
temperature on hepatic T3 degradation followed a similar pattern of
regulation as hepatic 5'-deiodination.

Exposure to 14°C stimulated maximal renal rT35'-D activity in lean
mice (69%), but had no significant effect in obese mice when compared
to respective animals housed at 25°C (Figure 2). Housing lean and obese
mice at 33°C caused a decline in renal rT35'-D activity of 63% and 39%,
respectively, relative to animals housed at 25°C. Due to the above
quantitatively dissimilar temperature-induced responses in maximal
renal rT35'-D activity observed in lean and obese mice, the percentage
difference in rT35'-D in obese mice relative to their lean littermates
was greater at 14°C (-54%) than at 25°C (-25%). At 33°C, maximal renal
rT35'-D activity was actually higher (+26%) in obese mice than in lean
mice.

In experiment 2, no differences in the Km for hepatic and renal
5'-deiodinase were observed between phenotypes housed at a given

environmental temperature. However, comparisons within a given

25

phenotype across different environmental temperatures revealed that
the Km of hepatic rT35'-D was significantly higher in lean and obese
mice housed at 33°C than in respective mice housed at 14°C and 25°C.
Plasma T4 and T3 concentrations for mice housed at 14°C, 25°C and
33°C are presented in Table 3. In all temperature groups, plasma T4
concentrations were significantly lower in obese mice than in their
lean littermates. Environmental temperature had no effect on plasma
14 concentrations in lean mice, while in obese mice T4 concentrations
were significantly higher at 25°C relative to 14°C and 33°C. Plasma
T3 concentrations were significantly lower in obese mice housed at
33°C compared to their lean littermates. No significant differences in
plasma T3 concentrations were observed between phenotypes at 14°C and
25°C. In addition, there was no effect of environmental temperature on
plasma T3 levels in lean mice. However, in obese mice plasma T3
concentrations were significantly lower at 33°C relative to 14°C and
25°C.

Effects of T3 Administration on Iodothyronine Metabolism, Results from

 

the kinetic analysis of T45'-D and rT35'-D in mice injected with
either T3 or vehicle are presented in Figure 3. T3 administration
increased maximal hepatic T45'-D activity by 327% and 695% in lean and
obese mice respectively. In addition, the maximum velocity of hepatic
rT35'-D was stimulated 191% and 424% by T3 treatment in lean and obese
mice, respectively. In control animals, both T45'-D and rT35'-D were
significantly lower (50% and 44%, respectively) in obese mice than in
lean mice. However, no difference in either T45'-D or rT35'-D was

observed between T3-treated lean and obese mice, due to the greater

pmoI-mg protein" 1 min" 1

Figure 3.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
    

 

 

 

 

 

 

 

LIVER 1 KIDNEY .. mm
2” ‘ T4 5’-o b 1. rT3 sl-o ;
“0’ ’LEAN ’ I500
200’ Dosese * 12”
100» _ ‘ 000
120» '3
L ' , 600 O
30 I * I ' 300 3-
4o 7’ a ’ 7 ‘
11.5 0.4 10.3 -o.5o :
cowtnm 13 -TREATED CONTROL 1;, «warren 3
E
3
42°°ILIVER 1. 1, FLIVER 0..., 3
3500 T3 DEGRADfiTION 3
3000 11 10.00 5;
2400 "
«0.40
13001
1200 3 a ‘ 0.20
300» 77 H
. %- A ,
CONTROL 1;, -TREATED c0N1'n01. 13 -TREA1’ED
Effect of thyroid hormone administration on iodothyronine

5'-deiodination in lean and obese mice. Vertical bars
represent mean values for Vmax (pmol/mg protein/min),
whereas numerals at base of each bar denote Km values
(uM T4 or 133). In the case of T3 degradation vertical
bars represent enzyme activity at a single substrate
concentration of 0.35 uM T3. Bars within phenotype
with different letters are different P<'0.05) as
determined by analysis of variance. Means are
different (P< 0.05) from those of lean mice of same
treatment group as determined by Student's t test.

1’ Km was significantly different (P< 0.05) from values
observed in control mice of same phenotype. There were

no significant differences in Km between phenotypes of
a given treatment group.

27

T3-induced response in enzyme activity in the latter phenotype. T3
administration also produced a greater response in renal rT35'-D of
obese mice (+418%) than in lean mice (+285%). Lean and obese controls
had similar activities of rT35'-D, while T3-treated obese mice had
significantly higher renal rT35'-D activity than T3-treated lean mice.
No differences in the Km for hepatic and renal 5'-deiodinase were
observed between phenotypes of a given treatment group. However, Km
values for hepatic rT35'-D and renal rT35'-D were, in some instances,
significantly lower in T3-treated mice than in controls.

T3 degradation appeared to be regulated by T3 administration in a
similar manner as that observed for hepatic 5'-deiodination (Figure 3).
Compared to control animals, T3 administration stimulated T3 degrada-
tion in lean and obese mice by 296% and 544%, respectively. As
observed in the previous experiments, hepatic T3 degradation was
significantly lower in obese controls than in lean controls. However,
ho difference in T3 degradation was observed between T3-treated lean
and obese mice.

Effect of T3 administration on plasma T4 and T3 concentrations are
presented in Table 3. No significant difference in either plasma T4
or T3 concentrations was observed between lean and obese controls.
Compared to control mice, T3 levels 4 h after the last T3 injection
were 7-fold and 12-fold higher in lean and obese mice, respectively.
T4 levels in T3-treated mice were too low to be detected.

Discussion

 

In the present study, hepatic and renal 5'-deiodinase activity was

assayed in a washed, microsomal fraction under optimal conditions in a

28

range of protein concentrations that were linearly related to reaction
velocity. Using this assay system, kinetic analysis of hepatic and
renal 5'-deiodinase revealed that the substantial variations in Vmax
arising from different physiological states or phenotypes in the study
were not accompanied by alterations in Km, except in a few cases where
small differences were noted. Hence, it may be assumed that the above

alterations in V were probably the results of changes in enzyme

max
amount.

Both T45'-D and rT35'-D in livers of lean and obese (ob/ob) mice
responded similarly to the various physiological manipulations
examined in the current study. This finding is consistent with the
hypothesis that a single enzyme is responsible for 5'-monodeiodination
of both T4 and rT3 in liver (27). The above hypothesis has also been
proposed for kidney (35); therefore, changes in renal rT35'-deiodinase
which was assayed in this study may reflect changes in renal T45'-
deiodinase as well.

Based on the results of experiment 1, slight differences in the
developmental profile of hepatic and renal 5'-deiodinase were observed
between lean and obese mice (Table 1). These differences are
illustrated by the fact that the percentage reduction in hepatic or
renal 5'-D in obese mice relative to their lean counterparts varied as
a function of age; for example, differences in hepatic 5'-deiodinase
between lean and obese mice were greater in the post-weaning age
groups than in the pre-weaning age groups. An opposite trend was
observed for renal 5'-deiodinase where the greatest reduction in
enzyme activity in obese mice relative to lean littermates occurred in

younger animals rather than older animals. Hence, the developmental

29

pattern of 5'-deiodinase appeared to be a function of both phenotype
and tissue type. t

The reduction in hepatic 5'-deiodinase in ob/ob mice relative to
their lean counterparts in experiment 1 was not a secondary result of
overt obesity since significant differences between phenotypes were
observed as early as 1 week of age. There are no data confirming the
above findings although it has been reported that 5'-deiodination in 10%
liver homogenates is similar in 5 mo old lean and obese mice (16);
these data conflict with our results derived from microsomal prepara-
tions. The above discrepancy may be attributed to possible experi-
mental error associated with measuring 5'-deiodinase specific activity
in crude homogenates where a significant amount of non-specific
substrate binding occurs. For example, we have been unable to
demonstrate any relationship between homogenate protein concentrations
of the magnitude seen in 5-20% liver homogenates and 5'-deiodinase
activity (unpublished observations). Dilution of a 20% liver
homogenate by 1/2 or 1/4 resulted in no apparent change in enzyme
activity. Similar studies by others (33) employing 18% kidney
homogenates revealed that T45'-D activity actually increased with
homogenate dilution. It was concluded that homogenate dilution
greatly increased free substrate (T4) concentrations which apparently
more than offset the decrease in enzyme protein in the system.
Without a linear relationship existing between homogenate protein
concentration and enzyme velocity, the assay of 5'-dei0dinase specific
activity (product formed/min/mg protein) is not valid because the

above measurement will vary depending on the protein content of the

3O

enzyme preparation. This could lead to spurious results when making
physiological comparisons in 5'-deiodinase specific activity since
different treatment groups may yield homogenates of different protein
concentration as is the case in lean and obese mice.

Accompanying reduced T45'-D activity in obese mice relative to
respective lean mice are lowered rates of T3 degradation in each age
group of experiment 1. Since T3 is a product of T45'-D, the difference
in activity of T45'-D in lean and obese mice should be theoretically
underestimated. However, the rate of T3 degradation in hepatic
microsomes is much lower than the enzymatic production of T3 so that
the amount of error is diminished.

Results from experiment 2 revealed that the response of 5'-dei0-
dinase to environmental temperature was a function of both phenotype
and tissue type (Figure 2). Regarding the former determinant, it was
shown that in a given tissue enzyme response to environmental
temperature was quantitatively different in lean and obese mice. The
latter determinant was indicated by the fact that as environmental
temperature shifted from one extreme to another, changes in the
relative levels of 5'-dei0dinase in lean and obese mice followed
opposite trends in liver and kidney. The latter finding and the
earlier observations showing that developmental changes in 5'-
deiodinase were dependent on tissue type suggest that different
isoenzymes for 5'-deiodinase may exist in liver and kidney. To confinm
the above hypothesis, additional studies are needed which directly
compare the kinetic properties of the renal and hepatic enzyme in a

purified system.

31

In experiment 3, T3 administration was shown to produce a greater
maximal induction in both hepatic and renal 5'-deiodinase in obese mice
than in lean mice (Figure 3). Although T3 administration resulted in
higher plasma T3 levels in obese mice than in lean mice, it is unlikely
that this is responsible for the greater enzyme induction in the former
phenotype since the T3 dosage used probably produced a maximal enzyme
response. The latter supposition is supported by studies showing that
the thyroid hormone dosage required for maximal 5'-deiodinase induction
in lean rats (5 ug T4/100 g body weight/day for 5 days) (28) is lower
than the hormone dosage employed in the present study (10 ug T3/100 g
body weight/day for 7 days). Furthermore, we injected T3 which is a
more potent inducer of 5'-deiodinase than T4 (18). The differential
response of hepatic 5'-dei0dinase to T3 administration between lean
and obese mice was qualitatively similar to that reported for other
hepatic, T3-regulated enzymes such as mitochondrial glycerol-3-
phosphate dehydrogenase (26,47) and Na+,K+-ATPase (36). In addition,
the basal activity of all 3 enzymes is lower in untreated obese mice
housed at 25°C than in their lean counterparts (26,36,47). The lower
basal enzyme activity in obese mice does not appear to be the result
of lowered circulating T3 concentrations since the preponderance of
, the data from this study (Table 3) and other reports (40,76,78) show
that blood T3 concentrations are similar in lean and obese mice. An
alternative mechanism for the lowered basal enzyme activities in obese
mice may be reduced tissue sensitivity to T3. Indeed, nuclear T3
receptor concentration in liver has been reported to be lower in

obese mice than in lean mice (19). However, the above mechanism does

32

not account for the greater enzyme response to T3 administration in
obese mice. This discrepancy may be explained based on the proposal
that thyroid hormone action can be modulated by other hormonal and
metabolic factors at a level distal to the nuclear receptor (43). It
has also been postulated that T3 by itself may generate unidentified
regulatory factors which amplify the nuclear T3 signal at the post-
receptor level (38). One may therefore speculate that T3 administration
results in a greater production of such intracellular activation
factors in obese mice than in lean mice. Thus, alterations in the
post-receptor modulation of thyroid hormone action between lean and
obese mice may explain the augmented enzyme response to exogenous T3 in
the latter phenotype in the face of lowered nuclear T3 receptor concen-
tration.

In an attempt to extrapolate the in vitro results of the present
study to the in vivo situation, maximal 5'-deiodinase activity
determined in vitro and expressed on a per organ basis (Vmax x micro-
somal protein/organ) may be compared among lean and obese mice to
estimate phenotypic differences in whole animal T4 to T3 conversion.
The above exercise assumes that all other factors regulating enzyme
activity besides enzyme amount are equal in lean and obese mice and
that liver and kidney are responsible for a large fraction of the T3
production in vivo. In general, maximal 5'-deiodinative capacity per
liver or kidney was found to be lower in obese mice than in respective
lean mice in most of the age groups examined in experiment 1. Based
on these results, one would expect serum T3 concentrations to be lower

in obese mice than in lean mice, but as discussed earlier circulating

33

T3 concentrations appear to be similar in both phenotypes. Possibly
T3 metabolism in vivo is lowered to a similar extent as T4 to T3
conversion in obese mice. Another possibility is that other regula-
tory factors, such as cytosolic cofactor concentrations are operating
in vivo to maintain normal 5'-deiodination in obese mice in spite of
lowered enzyme capacity. In vivo kinetic studies with thyroid hormones
would be useful in resolving this issue.

In conclusion, the present study demonstrated that hepatic and renal
5'-deiodinase activity is adaptive to changes in age, environmental
temperature and thyroid hormone status in both lean and obese mice. The
magnitude of enzyme response to the above physiological alterations,
however, was shown to vary between phenotypes. Under basal conditions
at an environmental temperature of 25°C maximal 5'-deiodination per mg
microsomal protein or per organ was lower in obese mice than in lean
mice. Furthermore, significant differences between phenotypes were
observed as early as 1 week of age before external signs of obesity
were apparent. These findings suggest that altered 5'-deiodination may
play a role in the expression of reduced energy expenditure in obese
mice which is also observed as early as 1 week of age (6). Conceiv-
ably, the reduced conversion of T4 to T3 in obese mice would result in
lowered T3 availability and consequently, impaired thyroid hormone
action in tissues mediating adaptive thermogenesis. In conflict with
the above hypothesis is the observation that plasma T3 concentrations
are similar in lean and obese mice. However, in the case of liver,
intracellular T3 concentrations in obese mice may be reduced since local

T4 monodeiodination in euthyroid rats accounts for approximately 28% of

34

this pool (58). Also, plasma T3 concentrations may not accurately
reflect intracellular nuclear T3 receptor occupancy if there are
alterations in T3 transport across the cell membrane or changes in
nuclear T3 receptor concentrations (19). The above situation suggests
that coordinate alterations at several metabolic control points may
collectively be responsible for impaired thyroid hormone action in
obese mice. Further studies to examine T3 transport across the cell
membrane and nuclear T3 receptor occupancy are needed to explore this

possibility.

IODOTHYRONINE 5'-DEIODINATION IN RATS FED LOW PROTEIN DIETS:

LACK OF CORRELATION WITH ENERGY BALANCE

Introduction

 

Rats fed low-protein diets exhibit increased heat production and
reduced efficiency of energy retention (30,62,67). Recently, efforts
have been made to identify the mechanisms responsible for adaptive
thermogenesis induced by feeding low protein diets. It has been
proposed that increased heat production observed in rats consuming a
low protein diet results from sympathetic activation of brown adipose
tissue (50). This is supported by data showing that norepinephrine
turnover in brown adipose tissue is accelerated in rats fed a low
protein diet (30). There is also evidence that thyroid hormones may
play a role in mediating the thermogenic response to low protein diets.
Consumption of a low protein diet is associated with a marked rise in
serum 3,3',5-triiodothyronine (T3) concentration (50,67) as well as an
elevation in hepatic mitochondrial o-glycerol phosphate dehydrogenase
activity (50,68). In addition, several biochemical processes which
have been proposed as mechanisms of adaptive heat production are
sensitive to thyroid hormone status. They include Na+,K+-ATPase
activity (36), protein turnover (15,48), futile cycling in carbohydrate
metabolism (23) and proton conductance pathway activity in brown
adipose tissue (66).

Extrathyroidal 5'-monodeiodination of thyroxine (T4) to the
metabolically active form T3 has been identified as one of several
control points regulating thyroid hormone action (32). This enzymatic
reaction is most active in liver and kidney and accounts for approxi-
mately 80% of the circulating T3 in the body. Enhanced thyroid

hormone action in rats fed a low protein diet may be mediated by a

35

36

stimulation in iodothyronine 5'-dei0dination since the latter process
would augment the supply of T3 to thermogenic target tissues.

In the present study, we examined the kinetic parameters of
iodothyronine 5'-deiodinase assayed in hepatic and renal microsomes of
young rats fed low protein (5 or 8% casein) or normal protein (22%
casein) diets. In addition, the above enzyme data were related to
energy balance measurements performed in the respective dietary treatment
groups.

Materials and Methods

 

Materials. 1251-iaoe1ed (3' or 5') L-3,3',5'-triiodothyronine (r13)
(600-800 uCi/ug) was purchased from Abbott Laboratories (N. Chicago, IL).
Unlabeled L-rT3 was obtained from Calbiochem (LaJolla, CA).
Dithiothreitol (DTT) was purchased from Sigma Chemical Co. (St. Louis,
MO). All other chemicals were reagent grade.

Animals.

Experiment 1. Twenty-four 3 week old male Sprague-Dawley rats,

 

obtained from Harlan Industries, Inc. (Indianapolis, IN), were
individually housed in wire-bottomed cages in a room maintained at
23-25°C with a 12-hour light-dark cycle. After an initial adaptation
period of 5 days in which rats were provided a stock diet ad libitum
(Wayne Rodent Blox from Continental Grain Co., Chicago, IL), animals
were segregated into 3 groups of equal number and similar average body
weight. One group of animals was killed immediately for determination
of initial body energy content. The remaining two groups were fed

ad libitum either an 8% casein (low protein, LP-8) or 22% casein (normal

protein, NP) semi-purified diet for 17-20 days. Composition of the

37

diets is given in Table 4. Animals were provided water ad libitum.
Food intake and body weights were monitored every 3 days during
dietary treatment. At the end of the feeding trial, rats were decapi-
tated and the liver and kidneys were rapidly removed for subsequent
enzyme assay. Carcasses were saved for determination of final body
energy content.

Experiment 2. The experimental design and conditions of this

 

experiment were similar to that described for experiment 1 except that
4 groups of 8 rats were employed. One group was again killed at the
start of the experiment for determination of initial body energy. The
remaining 3 groups were fed ad libitum either a 5% (LP-5) or 6% (LP-6)
or 22% (NP) casein diet for 17-20 days.

Measurement of Iodothyronine 5Ldeiodination. Iodothyronine 5'-

 

deiodinase activity was determined in hepatic and renal microsomes
using the tracer technique of Leonard and Rosenberg (35) where genera-
tion of radiolabeled product, 1251' from 1251 labeled (3' or 5')
substrate, L-rT3, was monitored. rT35'-deiodinase was assayed rather
than T45'-deiodinase since iodide release by the former reaction is
easier to quantitate (ion exchange chromatography) than is 13
production by the latter reaction (paper chromatography). Differences
in rT35'-deiodination observed between treatment groups should reflect
differences in T45'-deiodination as well since it has been proposed
that a single enzyme is responsible for 5'-monodeiodination of both
T4 and rT3 in liver (27) and kidney (35). Previous studies have shown
that both enzymes respond similarly to various physiological manipula-

tions (20,27). Details of the assay procedure have been previously

38

Table “w Composition of Diets

 

 

Low Protein Normal Protein
Casein 5 6 8 22
Glucose 76 75 73 59
Basal mix _l9_ _l9_ __12_ __12_
100 100 100 100

 

Basal mix contained in 9/1009; corn oil, 10;
cellulose, 3.6; mineral mix, 4 (1+); vitamin mix, 1
( 4); choline chloride, 0.2; and methionine, 0.2.

39

reported (27). Briefly, hepatic and renal microsomes were prepared by
differential centrifugation in 0.25 M sucrose, 0.05 M potassium phos-
phate, 0.01 M EDTA, (pH 7.4). Microsomal suspensions (5-7 mg protein/ml)
were frozen in acetone and dry ice and stored at -80°C after adding 2 mM
DTT. Microsomal protein concentration was determined according to the
method of Markwell et a1. (39). Prior to enzyme assay, aliquots of
hepatic and renal microsomes were thawed and then diluted with incubation
buffer consisting of 0.05 M potassium phosphate, 0.01 M EDTA, 5 mM

DTT, pH 8.0 to yield final protein concentrations of 6-8 ug/ml and

12-14 ug/ml, respectively. The above protein concentrations were

within the range where rT35'-deiodinase activity was linearly

related to microsomal protein concentration. Diluted microsomes were
incubated with 125I-rT3 (1.0 uCi/tube) and unlabeled rT3 to provide

final rT3 concentrations ranging from 0.125 uM to 2.0 uM. Enzyme
reactions were carried out for 10 min at 37°C under N2 gassing.

Reaction times and microsomal protein concentrations for the rT35'-
deiodinase assay were adjusted so that less than 16% of the substrate
was consumed during the incubation. Enzyme reactions were terminated

by plunging the assay tubes in cracked ice and adding 0.1 m1 2%

bovine serum albumin followed by 0.5 ml 10% trichloroacetic acid.

To determine the fractional generation of iodide from rT3, enzyme
reactants and products were separated using the ion exchange method of
Leonard and Rosenberg (35). In this procedure, trichloroacetic acid
extracts of the reaction mixtures were applied to Dowex SON-X8 columns

(BioRad Laboratories, Richmond, CA) equilibrated with 1.74 M acetic

acid. Free iodide was eluted from the column with three column

40

volumes of 1.74 M acetic acid divided into two aliquots and counted in
a ganma counter. Values for the fractional generation of iodide from
rT3 obtained from tissue-free control incubations were subtracted from
the observed experimental values of tissue-containing incubations to
correct for any non-enzymatic reactions and for 125I-labeled contami-
nants.

Treatment of Enzyme Data. Kinetic analysis was performed on data

 

derived from the rT35'-deiodinase assay. Enzyme activity (pmol product
formed/mg microsomal protein/min) was determined at each substrate
concentration and double reciprocal plots were constructed. Slopes

and intercepts of the linear plots were determined using unweighted
least squares analysis. In calculating 5'-deiodinase activity, it was
assumed that 1251 was randomly distributed between the 3' and 5'
positions of 125I-rT3. Hence the observed fractional generation of

iodide from rT3 was doubled to obtain the actual values.

Determination of Efficiency of Energy Retention. Efficiency of energy

 

retention (%) was calculated as body energy gain (kilocalories)
divided by metabolizable energy intake (kilocalories) times 100. Body
energy gain during the dietary treatment period was determined as the
final energy content of each carcass minus that estimated at the
beginning of the experiment. Initial body energy was estimated using
a linear regression equation developed by relating body energy to body
weight of rats killed at the start of the experiment. Body energy
content was determined by homogenizing carcasses 1:1 (w/v) in water
and drying aliquots at 55°C. A sample of the dried aliquot was then

combusted in a bomb calorimeter (Parr Instrument Co., Moline, IL).

41

The metabolizable energy content of the diets was calculated based on
the assumption that the metabolizable energy values of casein, corn
oil and glucose were 4, 9 and 3.64 kcal/g, respectively. Kevonian

et al.(30) demonstrated in feeding trials that the low and normal
protein diets of this study have equivalent metabolizable energy
content.

Serum Total T4, Total T3 and Free T3 Determinations. Blood samples

 

obtained at the time of killing were analyzed for serum total T4, total
T3 and free T3 by radioimmunoassay. Serum total T4 and free T3 were
measured by using a solid phase system supplied by Becton Dickinson
Immunodiagnostics (Orangeburg, NY). Serum total T3 was measured based
on the method of Nejad et a1. (41) which uses charcoal-dextran to
separate bound from free hormone.

Statistical Analysis. All values are presented as means i SEM. The

 

data were subjected to analysis of variance and statistical comparisons
were made using the Bonferroni t test (17), or Student's t test as
noted in Table footnotes.

Results

Experiment 1. Effects of feeding an 8% casein (low protein) diet on the

 

kinetic parameters of hepatic and renal rT35'-deiodinase are presented
in Table 5. The Lineweaver-Burk plots from which the Vmax and Km of
5'-deiodinase were derived are shown in Figure 4. Consumption of an 8%
casein diet had no effect on maximal activity of hepatic 5'-deiodinase
expressed per mg microsomal protein, but stimulated Vmax of the renal
enzyme by 40%. In both tissues, the Km of the reaction was not affected

by dietary treatment.

142

.umou u m.ucouaam an uoc.seuume mo acmevemaxo co c.3u.: asooo on»

1cvomou uNN mcvucoamuccou an» ac omega saga Amc.ovqo acacmaupc one menoxa .maopa xc=m1eu>aozmcvg sec; uo>wcou mew:
omocveowwu1.m ocwcocxsuouoa pecan ecu u.uoaog no; mcwuoEmgma u.uo=.x .mpce.co a go zwm « memos an» acmmmeame mos—m>

 

can No.qu mmn mc.qw
see m~.o «man ee.o e.oaeu ENN
emu ma.qu. com” Nc.qu
ammo~ -.o mNH~ we.o e_eoeu um

N ucwsvamaxm

Hm“ ee.c ease mo.oe
ewe me.o can“ mm.o e.aaeu RNN
man. mo.qn. «man m¢.qn
.e~o~ c~.c aeo~ am.o e.aaeu em

d acme.emaxm

 

 

 

cawuoca as :_muoea as
\:'E\—oEn mpg z: \=.E\posa m»; z:
1xas> 15x 1xus> 1Ex
xocepx eu>.4

 

omuc.uo—ou1.m oc.coexguoeo. Co mcmuoEacaa uwuocvx co muoau :.ouoea :op we muumeuu on o—pah

Figure 4.

43

Lineweaver-Burk analysis of iodothyronine 5'-
deiodinase in rats fed low protein diets. Each
point represents the mean reciprocal enzyme velocity
of 8 animals at a given substrate concentration.
Lines were calculated by linear least squares
regression.

44

 

 

 

3.0

2.5

2.0.

 

 

1/1' RELEASED: (nmol/mg protein/min)“1

 

 

 

°'5/ ~22% CASEIN °'5 H2291. CASEIN
" ... 3% CASEIN ... 5% CASEIN
-2-'1 g 4 6 a -2-1 2 4 6 8
_ _1

93,. KIDNEY KIDNEY

 

 

 

 

 

 

45

Hepatic and renal tissue weight, microsomal protein content and
maximal iodothyronine 5'-deiodinase activity expressed on either a
unit tissue or body weight basis are presented in Table 6. Liver
weight per 100 g body weight was similar in LP-8'and NP animals while
kidney weight was reduced 18% in the former group. Hepatic microsomal
protein content expressed per 9 tissue and per 100 g body weight was
30 and 36% lower in LP-8 rats, respectively, relative to NP controls.
Renal microsomal protein content was reduced 21% when expressed per
100 g body weight, but was unchanged when expressed per 9 tissue.

In calculating maximal 5'-deiodinase activity on either a 9 tissue or
100 g body weight basis, Vmax (Table 5) was multiplied by the appro-
priate expression of microsomal protein content. Maximal hepatic
5'-deiodinase activity per 9 liver and per 100 g body weight was 22 and
27% lower in LP-8 rats than in NP animals. Maximal renal 5'-dei0dinase
activity per g kidney was slightly elevated (38%) in LP-8 rats while
enzyme activity per 100 g body weight was unchanged. The sum of
hepatic and renal maximal 5'-deiodination per 100 g body weight was
reduced by 26% in LP-8 rats.

Energy balance measurements for experiment 1 are summarized in
Table 7. Metabolizable energy intake expressed either on an absolute
basis or per unit body weight basis was higher, 9 and 27% respectively,
in LP-8 rats than in NP animals. Body energy gain in the LP-8 group
was also increased (18%); hence, the calculated efficiency of energy
retention was similar in both the LP-8 and NP groups.

Experiment 2. Because consumption of an 8% casein diet failed to

stimulate adaptive thermogenesis in experiment 1, one could not

116

.unou u a.ueoeaum an vocascuuoe no ucos.coaxo ea

c—gu.3 cacao uoE1E.-u~u u- oe.oeoauoccou oz» uo omega nae. Ama.ova. neocou._u ecu access .As—ospuuoaaoc .u:m.on mean a oo~xc¢ouoca as
we. cann.u oxcaouoca on. acouEOD evouocn _-omoco'3 an so._a.u_au an 0.9-“. co.uu-oc as» so nun, as» ea nuance: ogu a. noun—:u—au no:
era’s: mean a can an: ac. oaua.u a can go'swuo- oa-¢.uo.oe1.m page; ce- u.uqaog pecan-z .n—qnvca o .0 gum « «coon oz» neon-Lao; no:_a>

 

o~a ~.o« u.o« n.o« n.ou no.c« on ~.—« ~.n« m.o« ~.c«
~m~ c.o o.ofl ~.o o.- nooc nod noun oom~ o.o~ v.v cvomou umm
an ¢.c« m.o« n.c« ”.0“ n°.o« no K.~« ov~a ..o« n.o«
coo“ v.0 ~.—— ~.o ~.- oo.o ago cm.- awonc o~.o~ m.e cvumou «m
N unoE—Loaxm
can woou m.o« m.o« n.o« "coon o~« w.~« w.~« c.o« ~.c«
snn moo o.o ~.o o.o~ Nooo ~m~ m.o~ ~.o~ ".0— o.e cwouuu «mm
on m.o« coon ~.o« n.o« ~o.o« on o.~« o.~« m.o« ~.o«
«NoH v.~ 3~.~— c~.n o.o~ 3~m.o 3mm cm.mm oa.¢¢ 3~.—~ o.c c'omau flm
g acme—c396
wan—u: ago—o: usm.oz uzm.ox ago—o: uanoz ugm_ux

soon a co" seen a can 0533.3 o soon 0 cc” oeaa_o 0 seen a ecu seen a can «new.» u soon a co" oeaa.o a zoos a so”
\c—E\~0Ec .Nmeex—oac .\c_E~_oEc \c_ouoca as \c.ouocq me .xozmmbu a .xcrexpoEc 1xcaexposc \c.ouoco uEI\c_uuoca as .xosmmeIo

 

uu.>.uu< »u_>.uo< au.>_uu<
onscauowoo1.m panvxuz anneauo.oo1.m c.0uoca ”no.0: ouocvuo.uo1.m c.0uoca u:m.o3
—~coa a u.uoaoz paces poe_xox .aIOmocu.z mamn.~ u.»~ao: pan—us: panamocu.z oamm—h
90¢.aeou
xucvau cos—4

 

u.n~n ugm.ox soon we. ago—o3 onaa.u u.c: a co nomuocaxo au.»vuua ounc.vo.ou1.m oc.cocxzuoeo_
poE_on ecu acoucou c.0uoca pneumocu—E .uzo.oz mama.» co mauve c.uuoca ro— ac muooauu .o «poop

47

Table 7. Energy intake and retention in rats fed low protein diets

 

Percentage casein in the diet

 

 

 

Measurement Exp 1 Exp 2
811 22: 51 61 fizz}
1111:1111 body wt - g 59 :2“ 5812‘ 65 _+_1a 54 1 13 65 11‘
Final body wt - g 140 _+_ 2° 174 _+_ 3b 87 _+_ 1a 93 1 1b 173 _+_ 3c

HE intake - kcal/expt 1075 :21“ 9135 120b 723 £183 815 1 8” 990 :13"-

ME intake - kcal/lOOg b b
body wt/day 57 113 45 1 1 49 _+_ 1a 51 1 1a 43 39.2

Body energy gain - b b
kcal/expt 254 1128 215 i. 8 94 _+_ 6a 132 _+_ 7 214 _+_ 6‘

Efficiency of energy
retention - 3 23.71123 21.810.83 13.1_+_0.13a 16.5_+_0.8° 21.7_+__o.6c

 

Values are the means 2 SEM for 8 rats fed for 17-2D’daysIIIiE (metabolizable
energy) intake and body energy gain were determined as described in the
Materials and Methods. Efficiency of energy retention was calculated as body
energy gain (kilocalories) divided by ME intake (kilocalories) times 100.
Means with different lower case superscript letters within each experiment
were different (p<0.05) as determined by either Student's t test (expt 1) or
analysis of variance followed by the Bonferroni t test (expt 2).

48

determine whether iodothyronine 5'-deiodination is altered during
conditions when the former process is enhanced. Therefore, protein
content of the diet was further lowered to either 5% or 6% casein in an
effort to activate adaptive heat production. This modification
succeeded in stimulating adaptive thermogenesis since efficiency of
energy retention of LP-5 and LP-6 rats was reduced by 40% and 24%,
respectively, relative to NP animals (Table 7). Although absolute
energy intake was lower (27% and 18%, respectively) in the LP-5 and
LP-6 groups than in the NP group, energy intake expressed on a per unit
body weight basis was higher in LP-5 and LP-6 animals. Body energy
gain was markedly reduced in LP-S and LP-6 rats by 56% and 38%,
.respectively.

Iodothyronine 5'-deiodination was examined in rats fed LP-5
because this group exhibited a greater stimulation in adaptive heat
production than did LP-6 animals. In general, the alterations in
enzyme data observed in the LP-S group relative to NP animals were very
similar to those described for the LP-8 group in experimEht 1 (Table 5,
Table 6, Figure 4). Consumption of a 5% casein diet increased maximal
activity of renal 5'-deiodinase by 21%, but had no effect on Vmax of
the hepatic enzyme (Table 5). In both tissues, the Km of the reaction
was unaffected by dietary treatment. Hepatic microsomal protein
content and maximal hepatic 5'-deiodinase activity were reduced 39 and
33% when expressed per 9 liver and 40 and 36% when expressed per 100 g
body weight, respectively, in LP-5 rats compared to NP animals. The
above measurements for kidney were similar in both dietary groups.

Combined hepatic and renal maximal 5'-deiodination per 100 g body

49

Table 8. Effects of low-protein diets on serum thyroid honmones

 

 

Total T4- Total T3- Free T3-
ng/ml ng/ml pg/ml
Experiment 1
as Casein 46.311.73 2.34:0.123 6.70:0.283
2211 Casein 49.0;1.7a 1.5610.04° 5.95:0.23a
Experiment 2
5: Casein 44.8;1 .5a 2.41:0.10a 5.721029a
611 Casein 49.112.63 241510.093 13.31_+_0.31a
2211 Casein 415.5130a 1.42_+_o.05b 5.87_+_0.283

 

Values are the means i SEM for 8 rats. Serum was obtained from animaTs
at the end of each experiment and analyzed for thyroid hormones as
described in the Materials and Methods. Means with different lower case
Superscript letters within each experiment were different (p<0.05) as
determined by either Student‘s t test (expt 1) or analysis of variance
followed by the Bonferroni t test (expt 2).

50

weight was reduced by 33% in LP-5 animals. Liver and kidney weights

per 100 g body weight were unchanged in LP-5 rats.

 

Serum Thyroid Hormones. Effects of dietary treatments in
experiments 1 and 2 on serum thyroid hormone concentrations are presented
in Table 8. Total serum T3 was markedly increased in LP-5, LP-6 and
LP-8 rats by 70%, 51% and 50%, respectively, relative to corresponding
NP animals. However, serum free T3 concentrations were unaffected by
the dietary treatments. Total serum T4 concentrations were also not

altered by dietary manipulation.

Discussion

 

He previously reported (20) that iodothyronine 5'-deiodination is
reduced in obese (ob/ob) mice, a genetic model where adaptive thermo-
genesis is impaired. In the present study, rats were fed low protein
diets in an effort to examine iodothyronine metabolism during
conditions of enhanced adaptive heat production. However, in
experiment 1, we were unable to accomplish this objective because rats
fed an 8% casein diet failed to exhibit an elevation in energy expendi-
ture (Table 7). Subsequently, in experiment 2, the protein content of
the diet was further reduced to either 5% or 6% casein which finally
succeeded in stimulating adaptive thermogenesis. Energy intake (per
100 g body weight) was also elevated in the latter dietary groups. It
is possible that the reduced efficiency of energy retention in rats
fed the 5% or 6% casein diets was a secondary response to an elevation
in energy intake. But, this is unlikely because rats fed 8% casein
(experiment 1) exhibited an increase in energy intake as well, but

without a concommitant change in heat production.

51

Kinetic analysis of microsomal rT35'-deiodinase revealed that
consumption of a low protein (5% or 8% casein) diet had no effect on
the Vmax (expressed per mg microsomal protein) of the hepatic enzyme,
but increased the maximal activity of the renal enzyme (Table 5).
Alterations in Vmax, if any, probably represented changes in enzyme
amount since the Km of hepatic and renal 5'-deiodinase was not
affected by dietary treatment, and the enzymes were assayed under
optimal conditions. The above findings partially conflict with the data
of Smallridge et a1. (59) where consumption of a low-protein diet (9%
casein) had no effect on 5'-deiodination in either liver or kidney
homogenates. The discrepancy regarding effects of low protein feeding
on renal 5'-deiodinati0n may be attributed to the different assay
systems used to measure this enzyme. We have discussed earlier (20)
the problem caused by substrate binding to non-enzyme protein in
measuring 5'-dei0dinase in homogenate systems which may lead to
spurious results. In the present study, this problem was circumvented
by assaying 5'-deiodination in an isolated subcellular fraction under
conditions where enzyme activity was linearly related to protein
concentration.

In an attempt to extrapolate the in vitro results of the current
study to the in vivo situation, maximal 5'-deiodinase activity deter-
mined in vitro was expressed per 100 g body weight for each tissue
(Vmax x mg microsomal protein/100 g body weight) (Table 6) to account
for differences in microsomal protein content and body weight among
dietary treatment groups. When making comparisons of maximal 5'-

deiodinase activity/100 g body weight, it is assumed that all other

52

factors regulating enzyme activity besides enzyme amount are constant
among treatment groups. The possibility that availability of either
substrate or cofactor may have differentially influenced enzyme
activity in rats fed low and normal protein diets cannot be ruled out,
but there is currently no evidence indicating that this occurs in rats
fad low protein diets or in any other dietary manipulation.

Maximal hepatic 5'-deiodinase activity/100 g body weight was lower
in rats fed a 5% casein (low protein) diet than in animals fed a 22%
casein (normal protein) diet; while in kidney, maximal enzyme
activity/100 g body weight was similar in both treatment groups.
Combined hepatic and renal maximal 5'-deiodination/100 g body weight
was lower in 5% casein-fed rats than in 22% casein-fed controls
because the contribution of the hepatic enzyme to total 5'-deiodinase
activity/100 g body weight was much greater than the contribution of
the renal enzyme. Since increased energy expenditure caused by feeding
a 5% casein diet is associated with a reduction in iodothyronine
5'-deiodination/100 g body weight, an inverse correlation between the
above parameters may be indicated. However, a causal relationship is
unlikely because similar alterations in maximal hepatic and renal
5'-deiodinase activity/100 g body weight occurred in both the 5%
casein-fEd rats where reduced efficiency of energy retention was
observed and in the 8% casein-fed group where adaptive thermogenesis
was not stimulated. Since iodothyronine 5'-deiodinase activity was
measured only in the microsomal fraction, the validity of the above
comparisons of maximal enzyme activity/100 g body weight depends on the

degree of variation between low and normal protein-fed animals in the

53

fractional recovery of total tissue enzyme activity in this subcellular
fraction during differential centrifugation. Recovery of total tissue
5'-deiodinase in the microsomal fraction was not measured in the
present study; nevertheless, maximal possible differences in this
variable between low and normal protein-fed.animals can be determined
from dietary treatment differences in microsomal protein content
expressed per 9 tissue. If one assumes the worst case scenario that
the decrease in hepatic microsomal protein content per 9 liver

(Table 6) in 5 and 8% casein-fed rats is entirely the result of reduced
efficiency of recovery of the microsomal fraction during differential
centrifugation and considers the fact that hepatic and renal iodo-
thyronine 5'-deiodinase is localized in the microsomal fraction (14,33),
then maximal hepatic 5'-deiodinase activity/100 g body weight and
combined maximal hepatic and renal enzyme activity/100 g body weight
corrected to account for differences in the above parameter would be
similar in low and normal protein-fed animals, rather than decreased

in the former group if these measurements were left uncorrected. In
either case, the same conclusion would be made: that iodothyronine
5'-deiodination does not mediate changes in energy balance in low
protein-féd rats. A similar conclusion may be deduced separately

from the kinetic data for 5'-dei0dinase (Table 5), which are independent
of changes in efficiency of recovery of the microsomal fraction, since
Vmax and Km of hepatic and renal 5'-deiodinase varied in a similar
manner in both 5 and 8% casein-fEd rats even though these two groups
exhibited different responses in efficiency of energy retention

relative to normal protein-fed controls. The above conclusion

54

contrasts with our initial hypothesis suggesting that increased
adaptive thermogenesis resulting from consumption of a low protein
diet is at least partly mediated by an increase in peripheral iodo-
thyronine 5'-deiodination. Additional studies examining in vivo
thyroid hormone kinetics in animals fed low protein diets are needed
to confirm the above results.

In agreement with earlier reports (11,59,77), rats fed either a 5%,
6% or 8% casein diet had higher serum total T3 concentrations than
normal protein-fed controls without corresponding changes in serum free
T3 concentrations (Table 8). These changes in circulating T3 levels
do not correlate well with the observed alterations in energy expendi-
ture in these dietary groups. As an illustration, feeding rats a_5%
casein diet and an 8% casein diet produced different thermogenic
responses relative to normal protein-fed animals, yet similar changes
in serum total T3 and free T3 concentrations. This indicates that
changes in circulating thyroid hormone concentrations, whether in the
free or bound form, are not directly responsible for activating
adaptive thermogenesis in 5% and 6% casein-fed rats. But, one cannot
rule out the possibility that the elevation in serum total T3 in 5%
casein-fed rats has a permissive role in reducing efficiency of energy
retention.

In summary, it appears that control points regulating thyroid
hormone action other than peripheral iodothyronine 5'-deiodination and
circulating thyroid hormone concentrations are active in mediating
alterations in energy efficiency resulting from consumption of a low

protein diet. One such control point might be hepatic nuclear T3

55

receptor concentration; however, consumption of a low protein diet (9%
casein) was without effect on this parameter (59). It was not reported
whether this diet stimulated thermogenesis. Additional studies
focusing on other determinants of thyroid hormone action such as
thyroid hormone transport from the plasma to the nuclear compartment
and nuclear T3 receptor occupancy are needed to provide a more complete
understanding of the mechanisms involved in enhanced heat production

after consumption of low protein diets.

NUCLEAR TRIIODOTHYRONINE RECEPTOR BINDING CHARACTERISTICS

AND OCCUPANCY IN LEAN AND OBESE (Ob/Ob) MICE

Introduction

 

Obese (ob/ob) mice gain more body energy than their lean litter-
mates in part as a result of defective non-shivering (63,64) and
diet-induced thermogenesis (65). Diminished adaptive heat production
in obese mice may be associated with impaired thyroid hormone action
since thyroid hormones play a permissive role in conjunction with the
sympathetic nervous system in regulating this process (21). The above
hypothesis is supported by the findings that activities of the thyroid
hormone-sensitive enzymes, o-glycerolphosphate dehydrogenase (26,47)
and Na+,K+-ATPase (36) are lower in several tissues of obese mice than
in lean mice. Furthermore, obese mice exhibit characteristics of
hypothyroidism such as cold intolerance (64) that are partially
reversed by thyroid hormone administration (22,42).

Recently, efforts have been made to identify the specific mechanisms
responsible for reduced thyroid hormone action in obese mice. We
reported that one determinant of thyroid hormone action, thyroxine
(T4)5'-deiodination, is decreased in hepatic and renal microsomes of
obese mice as early as 1-2 weeks of age (20). Conceivably, reduced
conversion of T4 to metabolically active 3,5,3'-triiodothyronine (T3)
in obese mice would result in lowered T3 availability, and consequently,
impaired thyroid hormone expression in tissues mediating adaptive
thermogenesis. However, this hypothesis is not in accord with findings
that plasma T3 concentrations are similar in lean and obese mice
(20,40,76), suggesting that diminished T45'-deiodination in obese

mice may not directly mediate reduced thyroid hormone action.

56

57

Alternate mechanisms for impaired thyroid hormone expression in
obese mice include changes at the level of nuclear T3 receptors.
Binding of T3 to specific nuclear T3 receptor sites is generally
regarded as the point of initiation of thyroid hormone action (43).
Nuclear T3 receptors have been characterized in several thyroid hormone
responsive tissues (46). Possibly diminished thyroid hormone action in
tissues of obese mice is caused by a reduction in numbers of nuclear T3
receptor binding sites. Modulation of some indices of thyroid hormone
expression through changes in nuclear T3 receptor concentration has
already been proposed for other physiological states such as starvation
(8,12,13) and partial hepatectomy (13). Another possible mechanism for
impaired thyroid hormone action in obese mice is a reduction in
nuclear T3 receptor occupancy by endogenous T3. Previous studies
have shown that thyroidal stimulation of both Na+, K+-ATPase (60)
and o-glycerolphosphate dehydrogenase (44) is correlated with occupancy
of specific nuclear T3 binding sites.

In the present study, we examined the possibility that nuclear T3
receptor concentration and occupancy by endogenous T3 are altered in
obese (ob/ob) mice. This report focuses only on liver since current
concepts of the mechanism of thyroid hormone action are largely
derived from studies employing this tissue. Also, data are available
comparing the activities of thyroid hormone-sensitive enzymes in livers
of lean and obese mice during various physiological states. Analysis
of nuclear T3 binding and endogenous T3 concentration was conducted
in vitro in soluble (0.4 M NaCl) nuclear extracts. This preparation

was employed to obviate the problem of T3 receptor leaching from

58

chromatin that occurs during incubation of intact nuclei, and is a
potential source of error in measuring maximal binding capacity (3).

In addition, soluble nuclear extracts exhibit relatively low nonspecific
binding and are suitable for the measurement of endogenous T3 content
via direct radioimmunoassay.

Materials and Methods

 

Animals. Male C57 BL/6J lean (ob/+ 0r +/+) and obese (ob/ob) littermates
obtained from our breeding colony were weaned at 3 wk of age and housed
individually in plastic solid-bottom cages at 23-250C. All animals

were provided stock diet (Wayne Rodent Blox from Continental Grain,
Chicago, IL) and water ad libitum. A 12 h light-dark cycle was
maintained throughout the studies. Experiments were conducted with mice
at either 4 wk of age, a time point early in the development of the

obese syndrome, or at 8-10 wk of age when the expression of gross

obesity is well established.

Preparation of Nuclei and Solubilization of Nuclear Receptor Proteins.

 

The following procedures were performed at 0-4°C. Hepatic nuclei were
isolated according to the method of Nidnell and Tata (73). Briefly, 7
and 6 livers from 4 wk-old lean and obese mice and 4 and 3 livers from
8-10 wk-old lean and obese mice, respectively, were pooled and
homogenized (1:3) in 0.32 M sucrose, 3 mM MgC12 (25 strokes, 1200 RPM)
using a loose fitting teflon-glass homogenizer. The homogenate was
filtered through nylon mesh and then diluted with 0.6 volumes of
homogenizing medium and 0.25 volumes of water to lower the tissue and
sucrose concentrations. Next, 0.32 M sucrose, 3 mM MgCl2 was layered

underneath the homogenate followed by centrifugation at 700 x g for

59

10 min. The pellet of crude nuclei was resuspended in 2.4 M sucrose,
1 mM MgCl2 and centrifuged at 60,000 x g for 60 min. The resulting
pellet of nuclei was washed with 0.14 M NaCl, 3 mM M9012 and recovered
by centrifugation at 1000 x g for 10 min.

Nuclear T3 receptors were solubilized according to the technique
of Silva et a1. (57). The washed nuclear pellet was first suspended in
TEM buffer (30 mM Tris, 2 mM EDTA, 5 mM MgClz, 5 mM mercaptoethanol,
10% (v/v) glycerol (pH 8.0 at 25°C)) using a glass rod. An equal
volume of TEM buffer containing 0.8 M NaCl was then added to the
suspension while gently stirring. To equalize protein concentrations
in each extract the total combined volume of solubilization buffer
(TEM, 0.4 M NaCl) added to each nuclear pellet was approximately 1.15
and 1.00 mg/g liver homogenized for 4 wk-old lean and obese mice,
respectively, and 1.10 and 0.65 ml/g liver for 8-10 wk-old lean and
obese mice, respectively. T3 receptors were extracted while vortexing
the nuclear suspension on ice for 15 sec at 5 min intervals for 45 min.
The preparation was then centrifuged at 25,000 x g for 30 min; the
clear supernatant was decanted and used for subsequent measurement of
T3 binding and endogenous T3 concentration.

Efficiency of extraction of the nuclear T3 receptor was estimated
by injection of tracer levels of (12°I)T3 (400-600 mCi/mg; Abbott Labs,
N. Chicago, IL) in lean and obese mice of each age group 1 h before
removal of tissue. The fraction of radioactivity in the washed nuclear

pellet that was solubilized with the receptor was determined.

60

Nuclear T3 Receptor Binding Assay. Fresh nuclear extract (100 ul)

 

diluted 1:3 with TEM, 0.4 M NaCl was mixed with 100 ul of ('25I)13
diluted in TEM, 0.025% bovine serum albumin to give final (12°I)T3
concentrations of 0.2x10'10 to lelO'loM. The mixture was first
incubated at 30°C for 40 min to dissociate endogenous T3 from the
nuclear receptor followed by incubation at 0°C for 20 h (57). Bound
and free (1251)T3 were separated as described by Seelig et a1. (54)
using an-anion exchange resin. Briefly, 1.0 ml of 10% (w/v) ice cold
slurry of Dowex AG 1-X8 resin (200-400 mesh) in TEM, 0.2 M NaCl was
added to the T3 binding assay mixture and incubated on ice for 20 min
with mixing at 5 min intervals. The resin, which adsorbs free
(12°I)T3 and T3, was then pelleted by centrifugation at 2000 x g for
5 min. Radioactivity in the bound fraction (supernatant) and free
fraction (resin) was quantitated using a gamma counter. Nonspecific
binding was determined by addition of 10'6 M T3 to the nuclear extract
1251)T

containing ( 3.

Measurement of Endogenous T3 Concentration in Nuclear Extracts. Endo-

 

genous T3 concentration was dEtermined by a modification of the direct
radioimmunoassay procedure of Yagura and walfish (74). The assay
mixture consisted of 300 ul of undiluted nuclear extract or standard
containing 0-200 pg T3 in TEM, 0.4 M NaCl, 0.1% bovine serum albumin;
50 ul of TEM, 0.4 M NaCl containing 250 ug 8-anilinonaphltalene-l-
sulphonic acid (ANS), 25 ug bovine serum albumin and ~25,000 CPM
(1251)T3 (3300 mCi/mg; New England Nuclear, Boston, MA); and 100 ul
anti-T3 antibody (Monoform, Miles Scientific, Naperville, IL) diluted
1:300 in TEM, 0.4 M NaCl. 250 ug ANS/tube was used since this

6l

concentration was sufficient to minimize the binding of (1251”3

to the receptor, and at the same time did not significantly interfere
with binding of T3 to antibody. The mixture was incubated at 4°C for
24 h. Bound and free fractions were separated using the anion exchange
resin method described for the T3 binding assay. Nonspecific binding
was assessed by omitting antibody from the assay mixture. Values were
separately determined for standards and nuclear extract samples and
subtracted from corresponding binding data obtained from samples
containing antibody.

Treatment of Data. Data from the binding studies were analyzed by the
method of Scatchard (53). The endogenous T3 content of the nuclear
extract was taken into account in calculating total ligand binding at

a given (1251)T3 concentration. In expressing maximal binding capacity
and endogenous nuclear T3 concentration on a unit DNA basis, correction
for incomplete receptor solubilization was included. Fractional

occupancy of the T receptor was calculated as endogenous T3 content

3
multiplied by the fraction of hormone associated with the receptor as

determined by the application of anion exchange resin to the nuclear

extract labeled with (125

I)T3 in vivo; the product finally divided by
the maximal binding capacity.

Other Analyses. Protein was determined by the method of Hughes et al.

 

(24) which employs N-ethylmaleimide to neutralize interference in the
assay produced by mercaptoethanol. DNA was determined by the method
of Burton (9) as modified by Richards (49). Statistical comparisons

were performed using the Student's t test for paired data.

62

Results

Results of the nuclear T3 binding study are summarized in Table 9.
In 4 wk-old mice, extraction of the hepatic nuclear pellet with TEM,
0.4 M NaCl solubilized 16% more protein per mg DNA in obese preparations
than in lean preparations even though the efficiency of extraction of
the nuclear T3 receptor as measured in the in vivo labeling experiment
was similar in both phenotypes. Maximal binding capacity expressed on
either a mg protein or mg DNA basis was not different in 4 wk-old lean
and obese mice. In addition, maximal binding capacity expressed on a
whole organ basis was similar in lean and obese mice since DNA content
per liver was not different between phenotypes. The equilibrium

dissociation constant (Kd) for specific nuclear T binding was also

3
similar in 4 wk-old lean and obese mice (Table 9 and Figure 5).

In 8-l0 wk-old mice, significantly more (26%) hepatic nuclear
protein was extracted per mg DNA in obese preparations than in lean,
and a small decrease (5%) in efficiency of extraction of the nuclear
T3 receptor was observed in the former group (Table 9). In contrast to
the results for 4 wk-old mice, maximal binding capacity (expressed on
a mg protein basis) was reduced 30% in 8-10 wk-old obese mice relative
to their lean counterparts (Table 9). There was also a small, but
significant, reduction (8%) in maximal binding capacity in obese mice
when the results were expressed per mg DNA. Conversely, maximal binding
capacity expressed on a whole organ basis was elevated 33% in 8-10 wk
old obese mice since the DNA content per liver was 45% higher in this

phenotype. No difference in the Kd for specific nuclear T3 binding

was observed between phenotypes (Table 9 and Figure 5).

63

.Afio.oquww .mmu mam» use co mu_s cmm_ co «menu
scab Amo.ov¢v ucwcweuwu wee mcumz ...m:opuocmamca couqmumc emm.u=: m.n=.om m we mm « memos mg» mew mmzpo>
.m'm»_a:< ucmzuuoum soc» um:_muno wee: ax can xupumamu mc_ucwn _ms_xoe co mmuwewumm

 

 

 

 

 

 

 

No.o« eefiu m~« Hm“ c.o« ~o.ow . -.o«
N~.o «rmfimm rmem «samm «H.me ««NG.H ««N~.© mmmno
~o.oa Km « mHH “Na e.~« No.o« “H.oa
m~.o chw com mse m.mc mw.o m~.e can;
u_o-x3-oaum
No.9“ mm « mm« NH“ v.0“ mo.o« oo.o«
m~.o «Nod omo ome H.~o «w~m.o ~m.~ mmmao
No.o« mm « mew HN« m.o« eo.o« mo.o«
mN.o okmfi owe . com o.mm em.c m~.~ cam;
c_o-¥3-<
Azofiuofixv cameo <2: 35 :mmuoca as may, <zo as cameo
ex _OEC .oEC (has. .xocawuwccw umuuaauxm <zo we
ampuwmmwlmcwucwm Pme_xmz cowuumcuxm cwmuoaa
emmpusz as

 

couamume m» emu—us: ovumam; as» ho muwumpcmguoccgu acwucwm gm m_nwp

Figure 5.

64

Scatchard plots of binding of T3 to solubilized
hepatic nuclear receptors in 4 and 8-10 week old
lean and obese mice. Each point is the mean of 5
preparations. The average protein content per tube
was 38 and 37 ug for 4 week old and 32 and 30 ug
for 8-10 week old lean and obese mice, respectively.

65

 

1.6 n 4 WEEK

1.2 .
LEAN '-

OBESE wo-
O.8 .-

O.4 -

 

 

 

 

100

 

1,6 . 8-10 WEEK

BOUND/ FREE
[0
O

1.2 "

0.8 -

0.4 '

 

 

l l l L .1 k I l l
O

 

 

20 40 60 80 160
T3 BOUND (pM)

Figure 5.

66

Data for endogenous nuclear T3 concentration and fractional
nuclear T3 receptor occupancy are presented in Table 10. In 4 wk-old
animals, hepatic nuclear T3 concentration expressed on a mg protein
basis was 24% lower in obese mice than in lean mice. There was also a
small (13%), but significant, reduction in nuclear T3 concentration
expressed on a mg DNA basis in 4 wk-old obese mice. Nuclear T3 content
expressed on a whole organ basis was similar in both phenotypes. The
fraction of endogenous nuclear T3 specifically bound to the receptor,
as determined by an ion exchange resin test, was unaffected by pheno-
type (Table 10). As a result of the above finding and the previous
observation that maximal binding capacity per mg DNA was not different
in 4 wk-old lean and obese mice, a reduction in fractional nuclear T3
receptor occupancy (14%) of similar magnitude to that of endogenous
nuclear T3 concentration per mg DNA was calculated for the latter
phenotype.

Nuclear T3 concentrations were depressed even more in 8-10 wk-old
obese mice than in 4 wk-old obese mice (Table 10). Nuclear T3
concentration expressed on either a mg protein or mg DNA basis was 45%
and 28% lower, respectively, in 8-10 wk-old obese mice than in lean
counterparts. Nuclear T3 content per liver was unaffected by pheno-
type (Table 10). The fraction of nuclear T3 specifically bound to the
receptor was also not different between lean and obese mice. Fractional
nuclear T occupancy was 23% lower in nuclear extracts of 8-10 wk-old

3
obese mice than in corresponding lean controls.

67

.Aao.ovavww .mmm msmm as» ho wu.s cam. ac among soc» Amo.ovav ucmcmeawu mam mcam: .
.m:o.umcoamcq m mo mm a menus as» are mwapm> .moonumz new mpwaamucz any cw cmnweumwc
<Hm pumawo mg» mcwma uwcamoms mm: mucccuxa cm¢_u:c mpnapom Lo cowumepcmucou me maocmmoccm mg»

 

 

 

 

 

 

o.~w ~.H« saga man mfia
«.m.m~ m.m~ gem“ «wmom wwemfi mmmao
m.~« w.on mm a m a m a
m.mm m.om mafia mmm NNN cam;
upo-x3-ofium
m.~« a.o« me“ man m u
wo.mm m.om mow «mow swoom mmmno
m.~fl m.o« may ea“ c «
m.~e w.ow mmm omm mom cams
upo-x3-e
oon ooflx cmaeo <zo me :Pmuoea me
zucmaauuo couamumm m» venom appoupapumam mp ems» Foe» Poe»
mmm_u=zlemcoppumci maocmmoemw ho :o—pumad (cowuacucmucou mp campuaz

 

.muumauxm campus: upucam; ac ucwacou mp .oH m—nmh

68

Discussion

 

In determining whether changes in thyroid hormone action are caused
by alterations in either nuclear T3 receptor number or occupancy by
endogenous T3, it is probably most valid to express data for these
measurements on a DNA basis since this compound comprises the genome
that is transcribed in response to T3 binding to the specific nuclear
receptor. Data stated on a protein basis have little physiological
relevance since the amount of nuclear protein extracted per unit DNA
can vary between treatment groups as was the case with lean and obese
mice at both 4 and 8-10 wks of age (Table 9). Similarly, when
correlating the above measurements with indices of thyroid hormone
action such as Na+,K+-ATPase and a-glycerolphosphate dehydrogenase,
the latter should also be expressed in the same units (per unit DNA).

In the present study, hepatic nuclear T3 receptor concentration
(fmol/mg DNA) was similar in 4 wk-old lean and obese mice, and only
slightly reduced in obese mice at 8-10 wks of age (Table 9). Theo-
retically, maximal enzyme response to thyroid hormone administration
is limited by the concentration of nuclear T3 receptors in the tissue.
Therefore, the finding that maximal induction of hepatic Na+,K+-ATPase
(36) and a-glycerolphosphate dehydrogenase (56) by thyroid hormone
treatment results in equally high enzyme levels in lean and obese
mice is consistent with our observation that nuclear T3 receptor
concentration is not different in the above phenotypes (Table 9).

Our nuclear T3 binding results are in apparent conflict with the
study of Guernsey and Morishige (19) showing a marked reduction (56%)

in hepatic nuclear T3 receptor concentration in obese (ob/ob) mice.

69

There are several possible explanations for this discrepancy. First,
Guernsey and Morishige may have employed older mice (age of the animals
not reported) in their experiment than those used in the present study.
Possibly a substantial reduction in nuclear T3 receptor concentration
occurs in obese mice after 8-10 wks of age. Second, Guernsey and
Morishige conducted their binding studies with intact nuclei isolated
by repeated detergent treatment and centrifugation. Bernal and
DeGroot (3) have shown that a significant fraction (m50%) of the T3
receptors dissociate from detergent treated nuclei incubated at 20°C.
Receptor leaching would cause an underestimation of maximal T3

binding per unit DNA. The rate of receptor dissociation may have been
higher in nuclear preparations of obese mice than in lean mice, thus
producing an apparent reduction in maximal binding capacity in the
former phenotype. In the present study, we employed a solubilized
nuclear receptor preparation to overcome this problem as well as to
increase the "signal to noise" ratio of the binding assay. We also
determined the efficiency of extraction of the nuclear receptor to
account for the possibility that this variable might influence the
results. Finally, the different nuclear preparations employed by
Guernsey and Morishige (l9) and this study could produce divergent
results by another mechanism if intact isolated nuclei contain
inhibiting and/or activating intranuclear factors at sufficient
concentrations to affect nuclear T3 receptor binding in vitro. Indeed,
several nuclear components such as certain histone (l) and nonhistone

(5) proteins can modulate the activity of partially purified nuclear

T3 receptors derived from rat liver. Possibly intranuclear regulatory

70

factors differentially influence T3 binding in isolated nuclei of
lean and obese mice. Use of soluble nuclear extracts should reduce
the influence of these factors since the intranuclear environment
would be greatly diluted.

In contrast to the nuclear T3 kinetic binding data, fractional
nuclear T3 receptor occupancy was reduced 14 and 23% in 4 and 8-10
wk-old obese mice, respectively, relative to corresponding lean mice
(Table 10). This finding was mainly the result of changes in endo-

genous T concentration associated with the specific nuclear receptor

3
rather than alterations in nuclear T3 receptor number. Diminished
nuclear T3 receptor occupancy in obese mice is closely paralleled by
similar alterations in the concentration of the thyroid hormone-
sensitive enzyme, Na+, K+-ATPase. The number of hepatic Na+, K+-ATPase
enzyme units is reported to be decreased 24 and 46% (36) in 4 and

8 wk-old obese mice,respectively; a less pronounced depression in the
concentration of enzyme units in obese mice is evident in each age
group when the data are expressed on a DNA basis (D.R. Romsos,
unpublished results). In addition, Na+, K+-ATPase activity expressed
on a per organ basis is not different in both 4 and 8 wk-old lean and
obese mice (36) which is consistent with data showing a similar nuclear
T3 content per organ (Table 10) in each phenotype. The above correla-
tions between nuclear T3 receptor occupancy and enzyme expression
suggests that the former parameter plays a role in mediating impaired

thyroid hormone action in obese mice. This proposal is congruent

with data showing a greater percentage stimulation in Na+, K+-ATPase

activity in thyroid hormone-treated obese mice compared to similarly

71

treated lean mice (36) since a greater fractional increase in nuclear
T3 receptor occupancy would occur in the former phenotype.

There are several possible mechanisms that can cause a reduction
in nuclear T3 receptor occupancy in obese mice. First, a decrease in
plasma T3 concentration can produce corresponding changes in nuclear T3
receptor occupancy (45,70). This mechanism, however, may be disregarded
since circulating T3 concentrations are similar in lean and obese mice
(20,40,76). Second, a decrease in local T4 to T3 conversion in a
given tissue could result in a diminished supply of T3 to the nuclear
compartment and hence, occupancy of receptor sites. Indeed, microsomal
T45'-deiodinase activity is 32 and 40% lower in 4 and 8-10 wk-old obese
mice, respectively, than in corresponding lean mice (20). But, the
importance of this process in influencing hepatic nuclear T3 receptor
occupancy may be minimal since van Doorn et al. (70) have recently
shown using a continuous infusion technique that the hepatic nuclear
T3 pool is almost exclusively derived from the plasma T3 compartment
with little contribution coming from local conversion of T4 to T3.
Third, inhibition of binding of endogenous T3 to the specific nuclear
receptor by intranuclear regulatory factors could also cause a
reduction in nuclear T3 receptor occupancy. Presently, there is no
evidence either supporting or refuting this pr0posal in regards to
obese mice. A fourth and final mechanism for lowered nuclear T3
receptor occupancy in obese mice is a reduction in transport of plasma

T3 to the nuclear compartment. Altered T transport in obese mice is

3
consistent with the hypothesis of York (75) which proposes that changes

in membrane fluidity and phospholipid composition in obese mice may

72

affect the activity of membrane associated proteins, some of which may
be involved in transport processes. Further supporting this mechanism,
Kaplan et al. (29) have shown that tissue to plasma T3 ratios following
in vivo injection of tracer T3 are markedly reduced in livers of obese
mice relative to lean controls, indicating that transport across the
plasma membrane is diminished in the former phenotype. Additional
studies directly measuring transport of T3 from the plasma to the
nuclear compartment in livers of lean and obese mice are needed to
confirm the above proposal.

In summary, minimal differences in maximal binding capacity
expressed on a DNA basis were observed between lean and obese mice,
whereas nuclear T3 receptor occupancy was reduced in obese mice at 4
and 8-10 wk of age, the extent of the decrease being greater in 8-10
wk-old animals. Alterations in nuclear T3 receptor occupancy in obese
mice were correlated with reported changes in the thyroid hormone-
sensitive enzyme, Na+, K+-ATPase, thus suggesting a possible causal
relationship between these two parameters. Finally, it is proposed that
reduced hepatic nuclear receptor occupancy by endogenous T3 in obese
mice results primarily from an impairment in the transport of T3

from the plasma pool to the nuclear compartment.

GENERAL CON CLUS IONS

73

In the present studies, iodothyronine 5'-deiodination was examined
in lean and obese mice and rats fed low protein diets in an effort to
elucidate the mechanism(s) of altered thyroid hormone action in these
experimental animal models. It was hypothesized that alterations in T4
conversion to T3 in liver and kidney would influence the supply of
metabolically active T3 to thermogenic target tissues, resulting in
changes in nuclear T3 receptor occupancy. Reduced thyroid hormone
action in obese mice was shown to be paralleled by changes in microsomal
T45'-deiodination; however, the latter parameter was not correlated
with increased thyroid hormone expression in low protein-fed rats.
Nhether impaired thyroid hormone action in obese mice is caused by
reduced 5'-deiodination of T4 remains unclear since plasma T3 concen-
trations are not altered in these animals. Increased thyroid hormone
action in low protein-fed rats is probably mediated at another level
of the thyroid system.

To study an alternate mechanism of impaired thyroid hormone action
in obese mice, the binding characteristics and endogenous T3 occupancy
of solubilized hepatic nuclear T3 receptors were examined in these
animals. Maximal binding capacity and Kd were shown to be similar in
4 and 8—10 wk-old lean and obese mice, indicating that changes in
nuclear T3 receptor concentration and affinity do not mediate reduced
thyroid hormone expression in the latter phenotype. In contrast,
nuclear T3 receptor occupancy was decreased in obese mice, the extent
of the reduction being greater at 8-10 wks of age. Diminished
nuclear T3 receptor occupancy in obese mice was correlated with

reported changes in the activities of the thyroid hormone-sensitive

74

enzymes, Na+, K+-ATPase (36) and a-glycerolphosphate dehydrogenase
(47), indicating that a decreased nuclear T3 signal may mediate, in
part at least, impaired hepatic thyroid hormone action in this animal
model. Whether alterations in nuclear T3 receptor occupancy are
responsible for changes in thyroid hormone action in other thermogenic
tissues (i.e. brown adipose tissue) of obese mice remains to be
addressed.

The possibility that enhanced thyroid hormone action in low protein-
fed rats results from changes in the binding characteristics of the
nuclear T3 receptor has been previously examined by Smallridge et a1.
(59). They reported that the number and affinity of solubilized
hepatic nuclear T3 receptors were similar in low (9% casein) and normal
(18% casein) protein-fed rats. However, in calculating maximal binding
capacity, they did not account for incomplete extraction of the specific
T3 receptor from the nuclear pellet. This parameter could have varied
between the dietary treatment groups which would have produced correspond-
ing differences in maximal binding capacity expressed on a DNA basis.
Nevertheless, assuming that extraction efficiency is similar in low and
normal protein-fed rats, it may be concluded that enhanced thyroid
honmone action in the former group is not caused by alterations in
nuclear T3 receptor concentration and affinity, the same deduction
having been made during conditions of impaired thyroid hormone expression
in obese mice.

In summary, lowered nuclear T3 receptor occupancy may explain, in
part at least, hepatic resistance to normal circulating T3 concentrations

in obese mice. Studies are needed to determine whether this mechanism

75

is also applicable to the opposite situation of enhanced thyroid
hormone action in low protein-fed rats. In addition, the cause(s)
for reduced nuclear T3 occupancy in obese mice should be addressed.
Presently, an attractive hypothesis is impaired transport of T3 from
the plasma to the nuclear compartment which is consistent with the
theory of a general defect in membrane function proposed by York (75).
Direct measurements of hepatic T3 transport would be required to

investigate this possibility.

LIST OF REFERENCES

10.

LIST OF REFERENCES

Anselmet, A., J. Bismuth, M.M. Menezes-Ferreira, and J.
Torresani. Triiodothyronine nuclear receptor. Role of
histones and DNA in hormone binding. Biochim. Biophys. Acta
739:291-300, 1983.

Bellabarba, D., R.E. Peterson and K. Sterling. An improved
method for chromatography of iodothyronines. J. Clin. Endocrinol.
Metab. 28:305-307, 1968.

Bernal, J., and L.J. DeGroot. Thyroid hormone receptors: release
of receptor to the medium during in vitro incubation of isolated
rat liver nuclei. Endocrinology 100:648-655, 1977.

Bieri, J.G., G.S. Stoewsand, and G.M. Briggs. Report of the AIN
Ad Hoc Committee on standards for nutritional studies. J. Nutr.
107:1340-1348, 1977.

Bismuth, J., A. Anselmet, and J. Torresani. Triiodothyronine
nuclear receptor and the role of non-histone protein factors in
in vitro triiodothyronine binding. Biochim. Biophys. Acta 840:
271-279, 1985.

Boissonneault, G.A., M.J. Hornshuh, J.w. Simons, D.R. Romsos and
G.A. Leveille. Oxygen consumption and body fat content of young
lean and obese (ob/ob) mice. Proc. Soc. Exp. Biol. Med. 157:
402-406, 1978.

Bray, G.A. and D.A. York. Hypothalamic and genetic obesity in
experimental animals: an autonomic and endocrine hypothesis.
Physiol. Rev. 59:719-809, 1979.

Burman, K.D., Y. Lukes, F.D. Wright, and L. Nartofsky. Reduction
in hepatic triiodothyronine binding capacity induced by fasting.
Endocrinology 101:1331-1338, l977.

Burton, K. A study of the conditions and mechanism of the
diphenylamine reaction for the colorimetric estimation of
deoxyribonucleic acid. Biochem. J. 62:315-323, 1956.

Chopra, I.J. A study of extrathyroidal conversion of thyroxine

(T ) to 3,3',5-triiodothyronine (T3) in vitro. Endocronology 10l:
453-463, .

76

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

77

Cox, M.D., S.S. Dalal, C.R.C. Heard, and D.J. Millward. Metabolic
rate and thyroid status in rats fed diets of different protein-
energy value: the importance of free T3. J. Nutr. 114:1609-1616,
1984. ,

DeGroot, L.J., A.H. Coleoni, P.A. Rue, H. Seo, E. Martino, and
S. Refetoff. Reduced nuclear triiodothyronine receptors in

starvation-induced hypothyroidism. Biochem. Biophys. Res. Commun.
79:173-178, 1977.

Dillmann, N.H., H.L. Schwartz, and J.H. Oppenheimer. Selective
alterations in hepatic enzyme response after reduction of nuclear
triiodothyronine receptor sites by partial hepatectomy and
starvation. Biochem. Biophys. Res. Commun. 80:259-266, 1978.

Fekkes, D., E. van Overmeeren-Kaptein, R. Docter, G. Hennemann and
T.J. Visser. Location of rat liver iodothyronine deiodinating
enzymes in the endoplasmic reticulum. Biochim. Bi0phys. Acta
587:12-19, 1979.

Flaim, K.E., J.B. Li and L.S. Jefferson. Effects of thyroxine
on protein turnover in rat skeletal muscle. Am. J. Physiol.
235(2):E23l-E236, 1978.

Gambert, S.R. and T.L. Garthwaite. Serum thyroid hormone levels
and thyroid hormone metabolosm by 10% liver homogenate from adult
lean and obese (ob/ob) mice. Horm. Metab. Res. 13:588-589, 1981.

Gill, J.L. Design and Analysis of Experiments in the Animal and
Medical Sciences, vol. l. Iowa State University Press, Ames, IA,
1978.

Grussendorf, M. and M. Hufner. Induction of the thyroxine (T4)
to triiodothyronine (T3) converting enzyme in rat liver by
thyroid hormones and analogs. Clin. Chim. Acta 80:61-66, 1977.

Guernsey, D.L. and N.K. Morishige. Na+ pump activity and nuclear
T3 receptors in tissues of genetically obese (ob/ob) mice.
Metabolism 28:629-632, 1979.

Hillgartner, F.B., and D.R. Romsos. Regulation of iodothyronine
5'-deiodination in lean and obese (ob/ob) mice. Am. J. Physiol.
249 (Endocrinol. Metab. 12):E209-E218, 1985.

Hinms-Hagen, J. Thyroid hormones and thermogenesis. In:
Mammalian Thermogenesis, edited by L. Girardier and M.J. Stock.
London: Chapman and Hall, 1983, p. 141-177.

Hogan, S. and J. Himms-Hagen. Abnormal brown adipose tissue in
genetically obese mice (ob/ob): effect of thyroxine. Am. J.
Physiol. 241 (Endocrinol. Metab. 4):E436-E443, 1981.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

78

Huang, M.T. and H.A. Lardy. Effects of thyroid states on the
cori cycle, glucose-alanine cycle, and futile cycling of glucose
metabolism in rats. Arch. Biochem. Biophys. 209:41-51, 1981.

Hughes, J., S. Joshi, and D. Ascoli. Elimination of thiol reagent
interference during Lowry protein determination. Anal. Biochem.
117:1-5, 1981.

James, N.P.T. and P. Trayhurn. An integrated view of the metabolic
and genetic basis for obesity. Lancet ii:770-773, 1976.

Joosten, H.F.P. and P.H.N. van der Kroon. Role of the thyroid
in the development of the obese-hyperglycemic syndrome in mice
(ob/ob). Metabolism 23:425-436, 1974.

Kaplan, M.M., J.B. Tatro, R. Breitbart and P.R. Larson. Comparison
of thyroxine and 3,3',5'-triiodothyronine metabolism in rat kidney
and liver homogenates. Metabolism 28:1139-1146, 1979.

Kaplan, M.M. and R.D. Utiger. Iodothyronine metabolism in liver
and kidney homogenates from hyperthyroid and hypothyroid rats.
Endocrinology 103:156-161, 1978.

Kaplan, M.M., J.B. Young, E.A. Shaw. Abnormal thyroid hormone
binding to serum proteins in ob/ob and db/db genetically obese
mice. Endocrinology 117:1858-1863, 1985.

Kevonian, A.V., J.G. Vander Tuig, and D.R. Romsos. Consumption of
a low protein diet increases norepinephrine turnover in brown
adipose tissue of adult rats. J. Nutr. 114:543-549, 1984.

Knehans, A.H. and D.R: Romsos. Reduced norepinephrine turnover
in brown adipose tissue of ob/ob mice. Am. J. Physiol. 242
(Endocrinol. Metab. 5):E253-E26l, 1982.

Larsen, P.R., J.E. Silva and M.M. Kaplan. Relationships between
circulating and intracellular thyroid hormones: physiological and
clinical implications. Endocrine Rev. 2:87-102, 1981.

Leonard, J.L. and I.N. Rosenberg. Subcellular distribution of
thyroxine 5'-deiodinase in the rat kidney: a plasma membrane
location. Endocrinology 103:274-280, 1978.

Leonard, J.L. and I.N. Rosenberg. Thyroxine 5'-deiodinase activity
of rat kidney: observations on activation by thiols and inhibition
by propylthiouracil. Endocrinology 103:2137-2144, 1978.

Leonard, J.L. and I.N. Rosenberg. Iodothyronine 5'-deiodinase
from rat kidney: substrate specificity and the 5'-deiodination of
reverse triiodothyronine. Endocrinology 107:1376-1383, 1980.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

79

Lin, M.H., J.G. Vander Tuig, D.R. RomSos, T. Akera and G.A.
Leveille. Na+,K+-ATPase enzyme units in lean and obese (ob/ob)
thyroxine-injected mice. Am. J. Physiol. 237(3):E265-E272, 1979.

Lin, P.Y., D.R. Romsos, J.G. Vander Tuig and G.A. Leveille.
Maintenance energy requirements, energy retention and heat
production of young obese (ob/ob) and lean mice fed a high-fat
or high-carbohydrate diet. J. Nutr. 109:1143-1153, 1979.

Mariash, C.N. and J.H. Oppenheimer. Thyroid hormone-carbohydrate
interaction at the hepatic nuclear level. Federation Proc. 41:
2671-2676, 1982.

Markwell, M.A.K., S.M. Haas, L.L. Bieber and N.E. Tolbert. A
modification of the Lowry procedure to simplify protein determi-
nation in membrane and lipoprotein samples. Anal. Biochem. 87:
206-210, 1978.

Mobley, P.R. and P.U. Dubec. Thyroid hormone levels in the
developing obese-hyperglycemic syndrome. Horm. Metab. Res. 11:
37-39, 1979.

Nejad, 1., J. Bollinger, M.A. Mitnick, P. Sullivan and S.
Reichlin. Measurement of plasma and tissue triiodothyronine
concentrations in the rat by radioimmunoassay. Endocrinology
96:773-778, 1975.

0htake, M., G.A. Bray, and M. Azukizawa. Studies on hypothermia
and thyroid function in the obese (ob/ob) mouse. Am. J. Physiol.
233 (Regulatory Integrative Comp. Physiol. 3):R110-R115, l977.

Oppenheimer, J.H. Thyroid hormone action at the cellular level.
Science 203:971-979, 1979.

Oppenheimer, J.H., P. Coulombe, H.L. Schwartz and N.N. Gutfeld.
Non-linear (amplified) relationship between nuclear occupancy by
triiodothyronine and the appearance rate of hepatic a-glycerolphos-
phate dehydrogenase and malic enzyme in the rat. J. Clin. Invest.
61:987-997, 1978.

Oppenheimer, J.H., H.L. Schwartz, 0. Koerner, and M.I. Surks.
Limited binding capacity sites for L-triiodothyronine in rat liver
nuclei. Nuclear-cytoplasmic interrelationships, binding constants,
and cross reactivity with L-thyroxine. J. Clin. Invest. 53:
768-777, 1974.

Oppenheimer, J.H., H.L. Schwartz, and M.I. Surks. Tissue
differences in the concentration of triiodothyronine nuclear
binding sites in the rat: liver, kidney, pituitary, heart, brain,
spleen and testis. Endocrinology 95:897-903, 1974.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

80

Otto, w., T.G. Taylor and D.A. York. Glycerol release in vitro
from adipose tissue of obese (ob/ob) mice treated with thyroid
hormones. J. Endocrinol. 71:143-155, 1976.

Peavy, D.E., J.M. Taylor and L.S. Jefferson. Protein synthesis
in perfused rat liver following thyroidectomy and hormone treat-
ment. Am. J. Physiol. 240 (Endocrinol. Metab. 3):E18-E23, 1981.

Richards, G.M. Modifications of the diphenylamine reaction giving
increased sensitivity and simplicity in the estimation of DNA.
Anal. Biochem. 57:369-376. 1974.

Rothwell, N.J., M.J. Stock and R.S. Tyzbir. Mechanisms of
thermogenesis induced by low protein diets. Metabolism 32:257-
261, 1983. -

Rothwell, N.J. and M.J. Stock. Similarities between cold- and
diet-induced thermogenesis in the rat. Can. J. Physiol. Pharmacol.
58:842-848. 1980.

Rothwell, N.J. and M.J. Stock. A role for brown adipose tissue
in diet-induced thermogenesis. Nature 281:31-35, 1979.

Scatchard, G. The attraction of proteins to small molecules and
ions. Ann. NY Acad. Sci. 51:660-672. 1949.

Seelig, 5., H.L. Schwartz, and J.H. Oppenheimer. Limitations in
the conventional analysis of the interaction of triiodothyronine
with solubilized nuclear receptor sites. J. Biol. Chem. 258:
2154-2161, 1981.

Sellars, E.A. and 5.5. You. Role of the thyroid in metabolic
responses to a cold environment. Am. J. Physiol. 163:81-91, 1950.

Shimomura, Y., G.A. Bray, and D.A. York. Effects of thyroid
hormone and adrenalectomy on Na+, K+,-ATPase in the ob/ob mouse.
Horm. Metab. Res. 13:578-582, 1981.

Silva, E.J., H. Astier, M. Thakare, H.L. Schwartz, and J.H.
Oppenheimer. Partial purification of the triiodothyronine
receptor from rat liver nuclei. J. Biol. Chem. 252:6799-6805,
1977.

Silva, J.E., T.E. Dick and P.R. Larson. The contribution of local
tissue thyroxine monodeiodination to the nuclear 3,5,3'-triiodo-
thyronine in pituitary, liver and kidney of euthyroid rats.
Endocrinology 103:1196-1207, 1978.

Smallridge, R.C., A.R. Glass, L. Nartofsky, K.R. Latham and K.D.
Burman. Investigations into the etiology of elevated serum T
levels in protein-malnourished rats. Metabolism 31:537-542, 982.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

81

Somjen, D., F. Ismail-Beigi, and 1.5. Edelman. Nuclear binding
of T3 and effects on 002, Na-K-ATPase, and a-GPDH in liver and
kidney. Am. J. Physiol. (Endocrinol. Metab. 3) 250:E146-E154,
1981.

Swanson, H.E. Interrelations between thyroxine and adrenalin in
the regulation of oxygen consumption in the albino rat. Endo-
crinology 59:217-225, 1956.

Swick, R.N. and C.L. Gribskov. The effect of dietary protein
levels on diet-induced thermogenesis in the rat. J. Nutr. 113:
2289-2294, 1983.

Thurlby, P.L. and P. Trayhurn. The role of thenmoregulatory

thermo enesis in the development of obesity in genetically-obese
(ob/ob mice pair-fed with lean siblings. Br. J. Nutr. 42:377-385,
1979. '

Trayhurn, P. and N.P.T. James. Thermoregulation and non-shivering
thermogenesis in the genetically obese (ob/ob) mouse. Pfluegers
Arch. 373:189-193, 1978.

Trayhurn, P., P.M. Jones, M.M. McHuckin and A.E. Goodbody.
Effects of overfeeding on energy balance and brown fat thermo-
genesis in obese (ob/ob) mice. Nature 295:323-325, 1982.

Triandafillou, J., C. Gwilliam, and J. Himms-Hagen. Role of
thyroid hormone in cold-induced changes in rat brown adipose
tissue mitochondria. Can. J. Biochem. 60:530-536. 1982.

Tulp, D.L., P.P. Krupp. E. Danforth and E.S. Horton. Characteris-
tics of thyroid function in experimental protein malnutrition.
J. Nutr. 109:1321-1332, 1979.

Tyzbir. R.S., A.S. Kunin, M.M. Sims and E. Danforth. Influence of
diet composition on serum triiodothyronine (T3) concentration,
hepatic mitochondrial metabolism and shuttle system activity in
rats. J. Nutr. 111:252-259, 1981.

Vander Tuig, J.G., D.R. Romsos and G.A. Leveille. Maintenance
energy requirements and energy retention of young obese (ob/ob)
and lean mice housed at 33° and fed a high-carbohydrate or a high-
fat diet. J. Nutr. 110:35-41, 1980.

van Doorn, J., D. van der Heide, and F. Roelfsema. Sources and
quantity of 3,5,3'-triiodothyronine»in several tissues of the rat.
J. Clin. Invest. 72:1778-1792, 1983.

Visser, T.J., D. Fekkes, R. Docter and G. Hennemann. Kinetics of
enzymic reductive deiodination of iodothyronines: effect of pH.
Biochem. J. 179:489-495, 1979.

72.

73.

74.

75.

76.

77.

78.

79.

82

Visser, T.J., I.van der Does-Tobe, R. Docter and G. Hennemann.
Subcellular localization of a rat liver enzyme converting
thyroxine into tri-iodothyronine and possible involvement of
essential thiol groups. Biochem. J. 157:479-482, 1976.

Nidnell, C.C., and J.R. Tata. A procedure for the isolation of
enzymatically active rat-liver nuclei. Biochem. J. 92:313-317,
1964.

Yagura, T., and P.G. Nalfish. A new unextracted sample radio-
immunoassay method for hepatic endogenous nuclear L-triiodothyronine
content. Biochem. J. 208:641-649, 1982.

York, D.A. Alterations in membrane function, organization and
composition in the obese ob/ob mouse. Proc. Nutr. Soc. 44:189-
200, 1985.

York, D.A., H. Otto and T.G. Taylor. Thyroid status of obese
(ob/ob) mice and its relationship to adipose tissue metabolism.
Comp. Biochem. Physiol. 598:59-65, 1978.

Young, R.A., L.E. Braverman and R. Rajatanavin. Low protein-
high carbohydrate diet induces alterations in the serum thyronine-
binding proteins in the rat. Endocrinology 110:1607-1612, 1982.

Young, R.A., S-L Fang, J. Prosky and L.E. Braverman. Hepatic
conversion of thyroxine to triiodothyronine in obese and lean
Zucker rats. Life Sci. 34:1783-1790, 1984.

Young, R.A., O.L. Tulp and E.S. Horton. Thyroid and growth
responses of young Zucker obese and lean rats to a low protein-
high carbohydrate diet. J. Nutr. 110:1421-1431, 1980.

"11111111111'111111113