Fat (fat) and Tubby (tub): Two Autosomal
Recessive Mutations Causing Obesity
Syndromes in the Mouse
D. L. Coleman and E. M. Eicher
This report describes the development of obesity syndromes in mice caused by two
autosomal recessive mutations, fat (fal), located on chromosome 8, and tubby (tub),
located on chromosome 7. Both mutations cause slowly developing but ultimately
severe obesity conditions. Although hyperinsulinemia, hyperactivity of the B cell of
the islets of Langerhans, and -cell degranulation are consistent features, these
obesity syndromes do not progress to severe diabetes. The many different singlegene mutations in the mouse that produce obesity-diabetes syndromes of varying
degrees of severity make the mutant mouse a powerful tool for analyzing the number
and nature of the primary defects than can cause obesity states.
From The Jackson Laboratory, Bar Harbor, Maine 04609.
This research was funded in part by NIH research
grants DK14461 (DLC), GM20919 and RR01 183 (EME).
We thank Christina Gott for originally noting the progenitor tubby male, Janice Southern for establishing
the mutant colony of tubby mice, and Elizabeth S. Russell for providing us with the C57BL/6J-HbbP congenic
strain. Dorothy B. Chapman and the late Richard H.
Copp provided colony management and technical assistance. The Jackson Laboratory is fully accredited
by the American Association of Laboratory Animal Care.
Address reprint requests to Dr. Coleman, Senior Stall
Scientist. The Jackson Laboratory, 600 Main St., Bar
Harbor, Maine 04609-0800.
Journal of Heredity 1990;81:424-427; 0022-1503/90/$2.00
Diabetes-obesity syndromes are present
in all human societies and are widespread
in the animal kingdom. Of special importance to research workers in this field is
the availability of small rodent models that
can be bred in the laboratory and maintained under controlled genetic and environmental conditions. Although the abnormalities associated with each obesity
type may have similarities with regard to
their overall development, the documentation that several different single genes
are involved in the different conditions
makes it unlikely that the various syndromes will be reduced to disturbances in
a single metabolic pathway. Single-gene
mutations maintained on standard inbred
backgrounds are particularly useful for
study since appropriate matings can be
made that produce predictable numbers
of unaffected and affected animals, all of
which are of the same genetic background,
differing only by the presence of a single
mutant gene or a small region of the chromosome containing the mutant gene. Several such single-gene mutations are known
in the mouse, and their characteristics and
interaction with inbred backgrounds have
been well described (Bray and York 1979;
Coleman 1978, 1982; Herberg and Coleman
1977). This report deals with two previously unreported obesity mutations that
occurred spontaneously in mouse colonies maintained at The Jackson Laboratory. With the rapid development of molecular techniques, these new murine
mutations causing obesity may play a fundamental role in elucidating the number
and nature of such defects and their relationships to each other.
Materials and Methods
All mice were housed three to five mice
per cage and fed natural ingredient chow
(Old Guilford 96, Emory Mouse Co., Guilford, Connecticut) ad libitum. We obtained
body weights and blood sugar concentrations weekly and made plasma insulin determinations at monthly intervals on selected normal and mutant mice. Blood (50
1l)for blood sugar determination was obtained from the retroorbital sinus, while
that for plasma insulin (0.5 ml) was obtained from the tail vein. Analytical and
histological procedures were as previously described (Coleman and Hummel 1967).
Results
The Fat (fat) Mutation
Origin and inheritanceof the fat mutation.
Obese mice were found in an HRS/J inbred
mouse colony in 1973 and progenitor mice
were provided to us for genetic and metabolic analyses. Breeding tests established that this obesity condition was
caused by an autosomal recessive mutation, which we named fat, gene symbol fat.
Forty (20.7%) of 176 offspring born to fat/+
parents developed the obese phenotype
establishing autosomal recessive inheritance. Allelism tests with the phenotypically similar diabetes (db, chromosome 4)
and obese (ob, chromosome 6) mutations
70
A
B
60
50
1-
40
a
30
U
o
20
10
l
0
10
20
30
AGE
10
20
30
(weeks)
and eight female -. O HRS/J-/at/fat mutants. The growth curve for normal HRS/J is similar to that of normal B6
Figure 1. (A) Growth curve of eight male H-and seven female
mice (Figure B), except that the attained body weight at 24 wk is greater, 32 g for males and 26 g for females; (B) Growth curve of six male --and five female A-A
normal littermates.
--O
6-tub/tub mutants compared with five male A-
produced no offspring having the obese
phenotype, which established that fat was
a new mutation, not a remutation at the db
or ob locus. Recent evidence suggests that
fat is located on chromosome 8 linked to
esterase-I (Es-i) (Paigen B, The Jackson
Laboratory, unpublished observations).
Physiologicalcharacteristicsof the fat mutation. Fat mutants develop obesity between 6-8 weeks of age (Figure IA). The
increase in body weight develops more
slowly than that seen with the more severe
obese (ob) and diabetes (db) mutations in
which the obese phenotype can be recognized at 18 days of age. Even so, the
obesity produced by the fat mutation can
be characterized as massive since mutants
attained body weights of 60-70 g by 24 wk
of age (Figure IA). No sex differences were
seen in the rate of weight gain in fat mutants in contrast to that seen with normal
littermate controls (data not shown) where
males after weaning are always heavier
than females. Most of the excess weight in
mutants represented increased weight of
body fat that extended to fat stores
throughout the body rather than being
confined to the axial and inguinal regions
as is typically observed by the pear-shaped
body conformation of the phenotypically
similar ob/ob and db/db mutants. Homozygous fat mutants are infertile, but not
sterile, and litters can be obtained from
homozygous male and female mutants if
they are mated before obesity develops at
5-8 wk of age.
Hyperglycemia was observed only transiently and just in males (Table 1) approaching 250 mg/ml at 7-8 wk and then
declining to within the normal range (120150 mg/dl). Female mutants remained
normoglycemic throughout the entire period of study. Hyperinsulinemia was a consistent feature in mutants of both sexes
and was severe at weaning (4 wk). The
hyperinsulinemia was associated with hypertrophy and hyperplasia of the islets of
Langerhans, features that became more
pronounced as the mutants aged. Varying
degrees of cell degranulation were consistent findings in males but were only obvious in females at 12 wk and older (Table
1). No signs of islet or a-cell atrophy were
observed in even the oldest mutants examined (11-13 mo).
To determine whether the fat/fatobesity
condition could be exacerbated to a fullblown diabetes condition, as seen in either
ob/ob or db/db mice maintained on the
C57BL/KsJ (BKs) inbred strain background, we transferred the fat mutation
from the HRS/J background onto the BKs
background via five cross-intercross cycles of breeding followed by inbreeding at
N5. On the BKs background, the syndrome
remained mild and similar in all parameters (rate of weight gain, blood sugar concentration, hyperinsulinemia, pancreatic
°
Table 1. Characteristics of HRS/J-fat/ut mice with age
Characteristic
Age (wk)
4
Blood glucose
(mg/dl)
Male
143 ±
141 ±
Female
Plasma immunoreactive insulin (AU/ml)
Male
612 ±
Female
417 ±
Islet hypertrophy
and hyperplasia
Male
Trace
Female
None
-cell degranulation
Male
Some
Female
None
8
12
175 ± 32
124 ± 10
16
199 ± 22
20
150 ± 22
147 ± 13
7.1
8.8
228 ± 40
121 ± 11
13
24
805 ± 190
988 ± 208
1,100 ± 96
1.752 ± 415
Some
Trace
Mild
Some
Moderate
Mild
Severe
Moderate
Some
None
Mild
Some
Moderate
Some
Moderate
Mild
133 ± 8.1
2,920 ± 479
2,240 ± 427
° Figures represent mean + SEM from six to eight mice of each group.
Coleman and Echer *New Obesity Mutants in the Mouse 425
Table 2. Mapping of the tub gene on chromosome 7
Cross: (Gpi-.l Hbbd +/Gpi-lb Hbb' tub)F, female x Gpi-1* Hbb' tub/Gpi-lb Hbb tub male
Loci inherited
Gpi-l
No.
of
bb
tub
offspring
Gpi-l"
Gpi-lb
Hbt
Gpb-bb
Hbb
+
rub
14
I
10
Gpi-l
Hbb
tub
13
Hbbd
+
8l
Gpi--IHbbd
Gpi-P
Hbb
rub
+
1
Gpi-I
Gpi-Pl
Hbb
Hbb
+
tub
CGl~pi-lb
Cpi
1
Total
Region of
recombination
None
Cpi- l-Hbb
Hbb-tub
0
0
47
Gpi-I-Hbb; Hbb-tub
Gene order with %recombination ± SE: Gpi-) - 44.7 + 7.2 - Hbb - 4.2 + 2.9 - tub
mative as to the relative order of Hbb and
Gpi-i, the actual proximal or distal location of tub relative to Hbb will need to be
confirmed.
Establishing the B6-HbbP +/Hbbs tub
strain. After we determined that tub was
located on chromosome 7 near the Hbb
locus, we established a breeding stock of
B6 mice segregating for both tub and Hbb
in order to use the Hbb locus to "tag" the
tubby mutant gene and thus select tub/+
and tub/tub offspring from informative
matings before their obese phenotype was
evident. This B6 stock was established as
follows: A female of the B6-Hbbcongenic
strain was mated to a B6-tub/tub male and
the resulting HbbP +/Hbb' tub offspring
morphology) to that seen in mutants maintained on the HRS/J inbred background.
Most significantly, plasma immunoreactive insulin (IRI) was high 886 + 95 uU/
ml at I yr and remained high up to 2 yr.
No signs of B-cell atrophy were seen in
BKs-fat/lfat mutants in contrast to the severe atrophy and drop in plasma IRI that
occurred at about 2-3 mo of age in either
diabetes (db) or obese (ob) mutants maintained on the BKs background.
The Tubby (tub) Mutation
Origin and inheritanceof the tub mutation.
The first tubby mouse noted was a C57BL/
6J (F125) male breeder located in the Animal Resources Colonies of The Jackson
Laboratory. This male, together with his
normal sister and their eight young, were
transferred to our research colony. Four
single-pair matings were established from
this litter. Both mice in one breeding pair
became obese and produced a litter of
young, all of which developed obesity after
weaning. None of the mice in the other
three pairs of mice developed obesity. Each
of these breeding pairs of mice, however,
did produce normal and obese young. We
concluded from this result that the newly
identified obesity condition was inherited,
probably as an autosomal recessive gene.
To determine the exact mode of inheritance of this new mutation, an obese male
was mated to two normal unrelated C57BL/
6J (B6) females. Because their nine offspring were normal, we concluded that the
mutation was not inherited as an autosomal dominant or X-linked gene. Three single-pair matings and one trio mating were
established from these nine young and
their offspring were classified at weaning
and at 8-10 wk of age for obesity. Approximately 25% of these offspring developed
an obese condition by 10 wk of age, suggesting that the obese condition was
caused by an autosomal mutation, hereafter named tubby, gene symbol tub.
To determine the genetic linkage of the
tubby mutation, a B6 homozygous tubby
female was mated to a CAST/Ei male. This
particular cross was chosen for mapping
tubby because the CAST/Ei inbred strain
is known to differ from the B6 strain at a
number of genetic loci. F, females were
then mated to B6-tub/tub males and 47
backcross offspring were typed for segregating loci, including the Hbb (hemoglobin beta-chain complex) and the Gpi-l
(glucose phosphate isomerase-1) loci, both
of which reside on chromosome 7, order
centromere - Gpi-l - Hbb. These data,
presented in Table 2, indicate that the tubby locus is located four map units from
Hbb on chromosome 7, probably distal to
Hbb. Because only two mice were infor-
were intercrossed. Offspring from this
cross were typed for their Hbb phenotype.
From this mating, mice that were heterozygous Hbb/ Hbb ', presumed to be HbbP +/
Hbb' tub, were mated to mice that were
homozygous Hbbs, presumed to be Hbb'
tub/Hbb tub. The next generation of mice
were again typed at weaning for their Hbb
genotype. This system of breeding is now
used to maintain the tubby mutation.
Physiological characteristics of the tub
mutant. Tubby mice are very similar to fat
mice except that the progression to obesity is slower (Figure B). The phenotype
cannot be recognized until 9-12 wk of age
and the average body weight at 24 wk is
less (46 g in tubby males vs 66 g in fat
males). Another difference between tubby
and fat mice is sexual dimorphisms with
respect to rate of weight gain, plasma in-
Table 3. Characteristics of C57BL/6J-tubtub mice with ageI
Age (wk)
Characteristic
4
8
12
16
20
147 ± 5.5
130 ±+3.3
139 ± 6.0
104 + 4.7
110 +±5.1
112 _+ 4.6
115 ± 3.9
120 ± 3.3
101 ± 3.0
120 + 4.6
83.4 ± 23
57.3 ± 16
96.7 ± 9.3
37.0 ± 7.6
258 ± 71
72 ±+ 14
418 ± 113
53.9 ± 7.3
None
None
None
None
Some
Moderate
Trace
Severe
Some
None
None
None
None
None
None
Some
Moderate
None
Blood glucose
(mg/dl)
Male
Female
Plasma immunoreactive insulin (U/ml)
Male
Female
69.9 ± 16
305 ± 64
Islet hypertrophy
and hyperplasia
Male
Female
None
d-cell degranulation
Male
Female
None
I Figures represent mean SEM obtained from six to eight mice of each group. Normal values are 130-150 mg/
dl for glucose and 28-37 eU/ml for insulin.
sulin concentration, and pancreatic morphology (Table 3). The attained average
weight at 24 wk of tub/tub females is only
38.6 ± 1.1 g compared to 46.4 + 0.9 g in
males. Mild hypoglycemia (110 mg/dl in
mutants vs 140 mg/dl in controls) associated with hyperinsulinemia develops in
mice of both sexes (Table 3). The hyperinsulinemia is mild at weaning and increases gradually with age. Histological
examination of the pancreas showed relatively normal islet morphology in mutants of both sexes compared to normal
littermates up to 12 wk of age, at which
time some enlargement of the islets occurred in males, a condition not seen in
females until much later (>24 wk). By 1
yr of age many large hypertrophied islets
with enlarged sinusoids were seen in tub/
tub males whereas these changes were not
nearly as evident in like-aged mutant females. Degranulation of the cells was
observed only in the late stages, suggesting that insulin synthesis and secretion in
tub/tub mice can be sustained at levels sufficient to maintain near normal 8-cell granulation in spite of hyperinsulinemia without degenerative changes in the islet or
E-cell being produced.
Tubby homozygotes, like fat homozygotes, are infertile, but not sterile, and can
produce litters when the animals are mated before severe obesity develops (up to
12 wk). The increased adipose tissue, as
in the fat mutation, is distributed over the
entire body.
The insulin II gene in the mouse is located on chromosome 7 (Lalley and Chirgwin 1984) suggesting that the tubby condition may result from some structural defect in insulin. HPLC analysis of the insulin
from tubby mutants showed no abnormalities, compared to normal littermates,
in the ratio of insulin I to insulin II or in
structure (based on retention time) showing that a structural defect in insulin is not
associated with this obesity syndrome
(Coleman DL, unpublished observations).
Discussion
The two obesity mutations tubby and fat
produce a more slowly developing obesity
syndrome than that in the phenotypically
similar obese and diabetes mutants. Additionally, problems in glucose homeostasis are mild and only transient. Hyperglycemia is only seen in homozygous fat males
between the ages of 7-11 wk whereas fat
females and tubby mutants of both sexes
remain normoglycemic or slightly hypoglycemic. In this regard, the mutations fat
and tubby produce an obesity syndrome
similar to that produced by the viable
yellow (Av) mutation. A mutants are
characterized by a severe but slowly developing obesity with modest hyperinsulinemia associated with normal glycemia
or even mild hypoglycemia (Coleman
1985). When the fat and viable yellow mutations were placed on the BKs inbred
background, the conditions remained mild
and diabetes was not seen. This differs
from the change from mild to severe diabetes seen when either the obese (ob) or
diabetes (db) mutation was placed on the
BKs inbred background. Tubby has not
been placed on the BKs background, but
it appears that in these milder obesity syndromes (viable yellow, fat and tubby) insulin synthesis and secretion can adjust
to meet the higher than normal demand
without producing islet atrophy and severe diabetes. Longevity studies have not
been done with either the fat or tubby mutants, but one would anticipate that the
lack of diabetes symptoms would allow a
lifespan similar to that seen with viable
yellow obese or diabetes maintained on
the diabetes protective B6 inbred background.
Histological examination of the pancreas showed that the syndromes produced
in tub/tuband fat/fatmice ranged from most
severe (hyperplasia, hypertrophy, and
marked degranulation of fS cells) in older
fat males to the nearly normal pathology
seen in middle-aged tubby females. One
striking feature of fatmutants on either the
HRS/J or BKs background is the pronounced hyperinsulinemia evident as early as 4 wk of age, at a time when obesity
is not detectable. Hyperinsulinemia of this
magnitude at this age without substantial
obesity is not typical of any of the other
known obesity mutants. The severe hy-
perinsulinemia sustained for most of the
lifespan of the fat mutants maintained on
the BKs background without islet atrophy
tends to dissociate chronic hyperinsulinemic stress from d-cell necrosis.
The five mutations obese (ob), diabetes
(db), viable yellow (A'), fat (fat), and tubby (tub) in the mouse offer the investigator
obesity-diabetes syndromes with a wide
range of severities, from mild obesity and
no diabetes to severe obesity with transient diabetes or severe obesity with severe and life-shortening diabetes. This
wide range of changes in the manifestation
of obesity coupled with varying degrees of
diabetes provide the investigator precision tools with which to advance our understanding of the causes and development of obesity-diabetes conditions. That
five genes in mice have been found that
cause obesities with varying degrees of severity suggests that similar genes will be
found in man and that human obesity may
have many possible genetic etiologies.
Mutant mice maintained on defined genetic backgrounds and used in conditions
of stringent environmental control should
provide the standardized material required to ultimately understand the causes
of obesity at the molecular level.
References
Bray GA and York DA, 1979. Hypothalamic and genetic
obesity in experimental animals: an autonomic and
endocrine hypothesis. Physiol Rev 59:719-809.
Coleman DL 1978. Obese and diabetes: two mutant
genes causing diabetes-obesity syndromes in mice.
Diabetologia 14:141-148.
Coleman DL, 1982. Diabetes-obesity syndromes-in mice.
Task force recommendations. Diabetes 31(Suppl 1):
1-6.
Coleman DL, 1985. Antiobesity effects of etiocholanotones in diabetes (db) viable yellow (A-) and normal
mice. Endocrinology 117:2279-2283.
Coleman DL and Hummel KP, 1967. Studies with the
mutation, diabetes, in the mouse. Diabetologla 3:238248.
Herberg L and Coleman DL, 1977. Laboratory animals
exhibiting obesity and diabetes syndromes. Metabolism 26:59-98.
Lalley PA and Chirgwin JM, 1984. Mapping of the mouse
insulin gene. Cytogenet Cell Genet 37:515 (abstr).
Coleman and Eicher - New Obesity Mutants in the Nfoue 427