THE PHENOTYPE OF the ob/ob mouse is characterized by

0013-7227/01/$03.00/0 Endocrinology 142(8):3421 3425 Printed in U.S.A. Copyright 2001 by The Endocrine Society Leptin-Deficient Mice Backcrossed to the BALB/cJ Genetic Background Have Reduced Adiposity, Enhanced Fertility, Normal Body Temperature, and Severe Diabetes J. QIU, S. OGUS, K. MOUNZIH, A. EWART-TOLAND, AND F. F. CHEHAB Department of Laboratory Medicine, University of California, San Francisco, California 94143-0134 A deficiency of leptin synthesis in mice results in a complex phenotype characterized by morbid obesity, diabetes, sterility, and defective thermogenesis. To determine whether the genetic background could alter the pleiotropic effects of leptin deficiency, we backcrossed the ob mutation for 10 generations from the C57BL/6J to the BALB/cJ genetic background. Compared with C57BL/6J ob/ob mice, BALB/cJ ob/ob mice showed at 27 wk of age a 35 40% reduction in body weight attributed to a 60% decrease in white adipose tissue mass. Food intake was not significantly different between the two obese strains, suggesting distinct utilization of energy intake. In the fed state, BALB/cJ ob/ob mice had elevated insulin and THE PHENOTYPE OF the ob/ob mouse is characterized by a morbid obesity, hyperphagia, transient hyperglycemia with late-onset diabetes, and life-long sterility (1). Despite intense investigations over half a decade aimed at uncovering its underlying physiological defect, it was only when positional cloning strategies came of age that a nonsense mutation (R105X) in a new gene was unveiled. The protein encoded by this gene was found to be a secreted protein named leptin (2). Administration of the recombinant hormone leptin to ob/ob mice rescues their metabolic and reproductive abnormalities (3 7) establishing an important role for this hormone in body weight homeostasis and reproductive biology. The search for leptin gene mutations among obese individuals led to the uncovering of rare homozygous cases that exhibited phenotypes similar to that of ob/ob mice (8) and responded well to recombinant leptin therapy, as evidenced by a decrease in food intake, reduction in body weight, and onset of normal reproductive function (9). Most of the experiments carried out on ob/ob mice were largely on mutant mice maintained on an inbred C57BL/6J genetic background. Hybrid vigor and genetic heterogeneity have long been known to confer a selective advantage to an organism, due mainly to the overall effects of modifier genes. Although the ob mutation has been mostly studied on the C57BL/6J background, a congenic line of ob/ob mice on the C57BL/Ks background was also generated. C57BL/Ks ob/ob mice exhibit initially an obese phenotype, but then start to lose weight progressively before dying prematurely of severe diabetes resulting from exhaustion and death of the islets of Langerhans (10). We recently described the phenotypes of F 2 ob/ob mice bred on a mixed C57BL/6J/BALB/cJ triglycerides levels, demonstrating a worsening effect on diabetes. At the reproductive level and in contrast to sterile C57BL/6J ob/ob mice, male and female BALB/cJ ob/ob mice were capable of reproducing after a mating period of 16 and 32 wk, respectively. At thermoneutrality, the body temperature of BALB/cJ ob/ob mice was 2.9 C higher than that of C57BL/6J ob/ob mice, whereas exposure of both groups to 4 C demonstrated a prolonged cold tolerance of BALB/cJ ob/ob mice. These studies show that the abnormalities caused by leptin deficiency can be genetically dissected and separated from each other, suggesting discrete pathways controlled by leptin modifier genes. (Endocrinology 142: 3421 3425, 2001) genetic background (11) and demonstrated that modifier genes on the BALB/c genome ameliorate the reproductive defect of ob/ob males and significantly contribute to an alleviation of their obese phenotype. In this communication we describe the characteristics of a new congenic line of ob/ob mice backcrossed for 10 generations on the BALB/cJ genetic background. BALB/cJ mice were previously found to be resistant to streptozotocin-induced diabetes (12) and have not been reported to exhibit a susceptibility to obesity. We found in this study that interaction of the BALB/cJ genome with homozygosity for the ob mutation results in a drastic decrease in fat accumulation, increased insulin resistance without a life-threatening condition, normal thermogenesis, and improved fecundity of the obese mice. Materials and Methods Derivation of the congenic line Inbred mice on the BALB/cJ background and ob/ob mice on the C57BL/6J background were purchased from The Jackson Laboratory (Bar Harbor, ME) and maintained at the University of California-San Francisco mouse facility under a standard regimen of alternating 12-h light and dark periods. All procedures were approved by the University of California-San Francisco committee on animal research. All mice were fed ad libitum a basic chow diet (FormuLab 5008, Ralston Purina Co., St. Louis, MO) with unrestricted access to water. Obligate heterozygous ob mice were generated by treating C57BL/6J ob/ob males with recombinant leptin as previously described (7), followed by mating with normal BALB/cJ females. F 1 ob/ males were then backcrossed to pure / BALB/cJ females to yield a 1:1 ratio of homozygous normal / and ob/ heterozygous mice in the F 2 generation. Heterozygous ob / mice were identified by PCR genotyping of the R105X mutation using a PCR assay (6). The process of backcrossing the ob / males to inbred BALB/cJ females was repeated for 10 generations to yield a congenic line of ob/ob mice on the BALB/cJ genetic background. During this process, 3421

3422 Endocrinology, August 2001, 142(8):3421 3425 Qiu et al. Genetic Separations of Abnormalities Caused by Leptin Deficiency 1 round of backcrossing involving a female ob / with a / BALB/cJ male was carried out to ensure transfer of the BALB/cJ Y chromosome to the new congenic line. Body weights, food intake, and adiposity Body weights of ob/ob males and females on the BALB/cJ and C57BL/6J backgrounds were determined regularly until 27 wk of age. Food intake was monitored daily for 18 d in C57BL/6J ob/ob (n 5) and BALB/cJ ob/ob (n 5) females housed in individual cages. Obese mice from each sex and group were killed at 27 31 wk of age, and the liver, white and brown adipose depots, and remaining carcasses devoid of any organs were weighed on an analytical balance. Metabolic assays Blood was collected by retroorbital sinus bleeding from five ob/ob males from each genetic background at 15 17 wk of age. Plasma glucose, cholesterol, triglycerides, and insulin levels in 24-h fasted and ad libitumfed mice were determined, respectively, on a Hitachi 747 Clinical Chemistry Analyzer (Hialeah, FL) and by a rat insulin RIA (Linco Research, Inc., St. Charles, MO). Cold challenge C57BL/6J ob/ob (n 3) and BALB/cJ ob/ob (n 7) females were housed individually and placed at 4 C. Rectal temperature recordings were determined before and hourly after exposure to cold with a precision thermometer (YSI, Inc., Yellow Springs, OH) equipped with a 0.2-mm probe (YSI, Inc., model 451). Fertility of BALB/cJ ob/ob mice To investigate the fertility of this new ob/ob line, 6- to 12-wk-old BALB/cJ ob/ob males (n 10) and BALB/cJ ob/ob females (n 4) were housed with proven wild-type breeding littermates over a period of 16 wk for the obese males and 32 wk for the obese females. As we regularly monitored the body weights of the obese mice and their mates, it was possible to retroactively determine their body weights at mating, considering a gestation period of 19 d. Statistics All values are reported as the mean sem. Significance levels were determined by unpaired t test for independent groups using the Statistica 4.1 software package (Statsoft, Tulsa, OK) for the Macintosh computer. Results Body weights, adiposity, and food intake The foremost striking difference between BALB/cJ ob/ob and C57BL/6J ob/ob mice was a reduction in body weight in the former group. At 5 wk of age, the body weights of C57BL/6J and BALB/cJ ob/ob mice were, respectively, 35.2 2.9 vs. 21.4 1.2 g for males (n 11 each group; P 6 10 7 ) and 30 2.9 vs. 20.7 1.3 g for females (n 10 each group; P 0.009). Differences in body weight between the two obese strains were highly significant throughout the monitoring period (Fig. 1), such that at 27 wk of age, the body weights of male and female BALB/cJ ob/ob mice were 27 and 31.5 g less than those of their C57BL/6J ob/ob counterparts. Figure 2 shows a photograph of two age-matched ob/ob females on the BALB/cJ and C57BL/6J backgrounds, depicting the impressive reduction of body weight in BALB/cJ ob/ob mice. To determine whether this reduction in body weight was due to a decrease in adiposity, we weighed the white and brown adipose tissue masses, liver, and carcasses of ob/ob mice from both strains at 27 31 wk of age (Fig. 3). Compared with their C57BL/6J ob/ob counterparts, BALB/cJ ob/ob mice had their white and brown adipose tissue masses reduced, respectively, by 2.5-fold (P 0.00002) and 1.5-fold (P 0.05) in males and by 2.2-fold (P 0.0000001) and 1.3-fold (P 0.06) in females. Furthermore, the liver was 1.6-fold (P 0.004) and 1.2-fold (P 0.05) smaller in male and female BALB/cJ ob/ob mice, respectively, vs. that in age- and sexmatched C57BL/6J ob/ob mice. Weights of the carcasses, which included muscle, bone, and brain, remained significant in both males (P 0.002) and females (P 0.0002), suggesting additional effects of the BALB/cJ genetic background. Food consumption by ob/ob mice on both genetic backgrounds was measured during an 8-d period to determine whether the decrease in food intake contributed to the noted differences in adiposity. We found that the cumulative food consumption of ob/ob mice from both strains was nearly identical and was not statistically significant (Fig. 4), demonstrating that food intake is not a contributing factor to the alleviation of the obese phenotype in BALB/cJ ob/ob mice. Thermogenesis At the initiation of this experiment, the body weights of the C57BL/6J ob/ob and BALB/cJ ob/ob females were 65.2 3.0 and 51.4 0.8 g, respectively (P 0.001). The ambient body temperatures of the C57BL/6J ob/ob and BALB/cJ ob/ob females were determined with a rectal probe and were 35.0 0.3 and 37.9 0.1 C, respectively (P 10 6 ). At 4 C, the temperature of C57BL/6J ob/ob and BALB/cJ ob/ob mice fell, respectively, to 29.0 1.9 and 35.8 1.1C(P 0.01) after 1 h and to 21.1 3.9 and 35.3 1.2C(P 0.001) after 2 h FIG. 1. Body weights of ob/ob males and females on the C57BL/6J or BALB/cJ genetic background from 4 27 wk of age.

Qiu et al. Genetic Separations of Abnormalities Caused by Leptin Deficiency Endocrinology, August 2001, 142(8):3421 3425 3423 FIG. 2. Typical photograph of two ob/ob females on the C57BL/6J or BALB/cJ genetic background depicting the reduced adiposity on the BALB/cJ background. FIG. 3. Adiposity and liver and carcass weights of BALB/cJ ob/ob ( ) and C57BL/6J ob/ob (f) males (M) and females (F) at 27 31 wk of age. *, P 0.05; **, P 0.002; ***, P 0.0002. Serum chemistries Plasma levels of glucose, insulin, triglycerides, and cholesterol were determined in the fasted and fed states in ob/ob males from both strains at 15 17 wk of age (Fig. 6). In the fasted state there was no significant difference in any measurement between groups except for glucose (P 0.04). However, in the fed state, triglycerides and insulin levels were, respectively, 3-fold (P 0.00003) and 2.5-fold (P 10 7 ) more elevated in obese BALB/cJ than in obese C57BL/6J mice. Glucose and cholesterol levels were not significantly different between the two groups. FIG. 4. Cumulative and daily food consumption of ob/ob males on the C57BL/6J (n 5; f) and BALB/cJ (n 5; ) genetic backgrounds. (Fig. 5) In compliance with an animal care regulation that prohibits animal death as the end point of an experiment, the C57BL/6J ob/ob mice had to be removed from the cold room after 1 h when their body temperature reached critically low levels, whereas the BALB/cJ ob/ob mice withstood the cold for up to 3 4 h but eventually succumbed to it. Fertility Six of 10 ob/ob males and all 4 females on the BALB/cJ genetic background were fertile, as determined by the 1 3 pregnancies induced by the obese fathers and the 1 2 pregnancies of the obese mothers (Table 1). The body weight range around the time of mating was 30.7 52.1 g for ob/ob males and 35.7 63.6 g for ob/ob females. There were no correlations between the number of pregnancies and the initial or at mating body weights of BALB/cJ ob/ob mice. Discussion The genetic makeup of an organism often accounts for the variation in penetrance associated with the phenotype of a particular mutation. To determine this effect in the context of an obese phenotype, we previously generated an F 2 inter-

3424 Endocrinology, August 2001, 142(8):3421 3425 Qiu et al. Genetic Separations of Abnormalities Caused by Leptin Deficiency FIG. 5. Body weights and core temperature of ob/ob females on the C57BL/6J and BALB/cJ genetic backgrounds duringa4ccold challenge. *, P 0.01; ***, P 0.0001. TABLE 1. Fertility of ob/ob males (n 10) and ob/ob females (n 4) on the BALB/cJ genetic background BW (g) Age (wk) Pregnancies Age at mating(s) (wk) BW at mating (g) Males 1 39.7 12 0 2 43.0 11 3 21, 22, 26 47.5, 48.7, 50.5 3 45.7 11 2 11, 24 45.7, 52.1 4 43.3 10 0 5 42.9 9 0 6 33.1 8 1 13 41.7 7 33.5 7 0 8 33.7 7 1 18 49.8 9 25.9 6 2 8, 19 34.3, 45.7 10 25.3 6 2 7, 13 30.7, 45.2 Females 1 39.0 10 1 11 41.7 2 33.6 8 2 13, 39 47.3, 63.6 3 29.3 8 2 10, 21 38.8, 52.6 4 23.4 6 1 9 35.7 The body weight (BW) and age of each mouse at the beginning of the mating period and near each mating time are shown. FIG. 6. Glucose, insulin, triglycerides, and cholesterol levels of C57BL/6J (f) and BALB/cJ ( ) ob/ob males at 15 17 wk of age. *, P 0.04; ***, P 0.0001. cross segregating the ob mutation on the mixed C57BL/6J and BALB/cJ genetic backgrounds (11). The resulting F 2 ob/ob mice showed variability in the expression of obesity, diabetes, and fertility, which was attributed to modifier genes from both genetic backgrounds. In this study we followed up on these experiments by backcrossing the ob mutation for 10 generations on the BALB/cJ genetic background. Intercrossing of the N10 ob heterozygotes resulted in a new congenic line of ob/ob mice on the BALB/cJ genetic background. Except for a differential C57BL/6J chromosomal segment of approximately 20 cm, which is retained from selection of the ob mutation at each round of backcrossing (13), the contribution of modifier genes from the BALB/cJ genetic background onto the ob/ob phenotype can now be assessed independently of the C57BL/6J genome. One of the major findings in BALB/cJ ob/ob mice was their decreased adiposity, which could not be attributed to decreased hyperphagia, because food intake was similar in both strains. Most likely, mechanisms that increase energy expenditure will be the first candidates to investigate the basis of their reduced adiposity. As C57BL/6J ob/ob mice have de-

Qiu et al. Genetic Separations of Abnormalities Caused by Leptin Deficiency Endocrinology, August 2001, 142(8):3421 3425 3425 fective thermogenesis and low body temperature (14), the finding that BALB/cJ ob/ob mice have normal ambient body temperature and are more tolerant to the cold challenge than C57BL/6J ob/ob mice suggests that cold-stimulated sympathetic outflow to brown fat in the absence of leptin can be corrected with the appropriate modifier genes. This observation further suggests that increased sympathetic activity of BALB/cJ ob/ob mice may be a contributing factor to their reduced obese phenotype. Although not yet determined, the general activity of BALB/cJ ob/ob mice appears to be increased compared with the well known passive nature of C57BL/6J ob/ob mice. The elevated serum insulin levels of fed BALB/cJ ob/ob mice demonstrate an insulin resistance state that is more severe than that of ob/ob mice on the C57BL/6J background and is quite distinct from that of the ob/ob strain bred on the C57BL/Ks background, which succumbs to the diabetic state as a result of pancreatic islet exhaustion (10). Thus, when coupled to an obese phenotype, the BALB/cJ background expresses modifier genes that are deleterious to glucose homeostasis. The elevated triglycerides levels of fed BALB/cJ ob/ob, but not C57BL/6J ob/ob, mice may be explained by decreased activity of endothelium-bound lipoprotein lipase, which normally is stimulated by insulin (15, 16). However, an insulin resistance environment such as that in obese mice, may explain at least in part the reduced synthesis and release of lipoprotein lipase leading to impaired triglycerides breakdown and accumulation in the circulation. Genetically, these hypotheses would imply that the modifier genes brought into the ob/ob phenotype by the BALB/cJ background greatly influence these pathways and that the uncovering of such factors, for example, by analysis of differentially expressed genes or microarray hybridization panels, will undoubtedly lead to the manipulation and understanding of these pathways. The sterility of ob/ob mice can be rescued with leptin treatment (6, 7) or through modifier genes brought into the ob phenotype by the mixed C57BL6J/BALB/cJ genetic background (11). In the present study the BALB/cJ ob/ob males exhibited a similar fertility as the F 2 C57BL6J/BALB/cJ ob/ob males. However, previously (11) we had found that the mixed C57BL/6J-BALB/cJ genetic background rescued the reproductive defect of only ob/ob males. In this study only 60% of BALB/cJ ob/ob males were fertile, suggesting that the mixed genetic background is more beneficial to male reproduction than the BALB/cJ inbred background alone. These differences may be attributed to the interaction of both genomes on reproductive function. Furthermore, the fertility of BALB/cJ ob/ob females, but not of C57BL/6J-BALB/cJ ob/ob females, demonstrates a stimulatory effect of BALB/cJ modifier genes on the female reproductive system of ob/ob mice. On the other hand, the idea that the obese state imposes a physical hindrance that might interfere with fecundity is not founded by our studies. For example, the body weights at mating of the BALB/cJ ob/ob mice ranged from 30.7 62.3 g. At these body weights, the C57BL/6J ob/ob mice remain sterile and fail to reproduce. Thus, adiposity is not a major physical factor for the failure of ob/ob mice to reproduce. Instead, BALB/cJ modifier genes, which remain to be unveiled, control the firing of the reproductive system and must play a critical role in reproductive pathways associated with obesity. Overall, the novel BALB/cJ ob/ob phenotype demonstrates the existence of modifier genes that can rescue a defective leptin pathway without leptin. The elucidation of the factors encoded by these modifier genes will not be an easy task and will require complex genetic and genomic approaches in central and peripheral tissues, but it will ultimately reveal whether these genes are part of the leptin pathway or of other dominant pathways. In either case, they will lead to an increased understanding of the mechanisms that govern body weight homeostasis, energy expenditure, and the reproductive system. Acknowledgments Received February 27, 2001. Accepted April 11, 2001. Address all correspondence and requests for reprints to: Dr. Farid F. Chehab, 505 Parnassus Avenue, Department of Laboratory Medicine, University of California, San Francisco, California 94143-0134. E-mail: chehabf@labmed2.ucsf.edu. This work was supported by NIH Grant HD-35142. References 1. Ingalls AM, Dickie MM, Snell GD 1950 Obese, a new mutation in the house mouse. J Hered 41:317 318 2. Zhang Y, Proenca R, Maffei. M, Barone M, Leopold L, Friedman JM 1994 Positional cloning of the mouse obese gene and its human homologue. Nature 372:425 432 3. Pelleymounter MA, Cullen MJ, Baker MB, et al. 1995 Effects of the obese gene product on body weight regulation in ob/ob mice. Science 269:540 543 4. Halaas JL, Gajiwala KS, Maffei M, et al. 1995 Weight-reducing effects of the plasma protein encoded by the obese gene. Science 269:543 546 5. Campfield LA, Smith FJ, Guisez Y, Devos R, Burn P 1995 Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks. Science 269:546 549 6. Chehab FF, Lim ME, Lu R 1996 Correction of the sterility defect in homozygous obese female mice by treatment with the human recombinant leptin. Nat Genet 12:318 320 7. Mounzih KM, Lu R, Chehab FF 1997 Leptin treatment rescues the sterility of genetically obese ob/ob males. Endocrinology 138:1190 1193 8. Montague CT, Farooqi IS, Whitehead JP, et al. 1997 Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature 387:903 908 9. Farooqi IS, Jebb SA, Langmack G, et al. 1999 Effects of recombinant leptin therapy in a child with congenital leptin deficiency. N Engl J Med 341:879 884 10. Coleman DL, Hummel KP 1973 The influence of the genetic background on the expression of the obese gene (ob) in the mouse. Diabetologia 9:287 293 11. Ewart-Toland A, Mounzih K, Qiu J, Chehab FF 1999 Effect of the genetic background on the reproduction of leptin-deficient obese mice. Endocrinology 140:732 738 12. Leiter EH, Le PH, Prochazka M, Worthen SM, Huppi K 1988 Genetic and environmental control of diabetes induction by multi-dose streptozotocin in two BALB/c substrains. Diabetes Res 9:5 10 13. Silver, L 1995 Mouse genetics. Oxford: Oxford University Press; 48 49 14. Trayhurn P, James WP 1978 Thermoregulation and non-shivering thermogenesis in the genetically obese (ob/ob) mouse. Pflugers Arch 73:189 193 15. Kern PA, Marshall S, Eckel RH 1985 Regulation of lipoprotein lipase in primary cultures of isolated human adipocytes. J Clin Invest 75:199 208 16. Fried SK, Russell CD, Grauso NL, Brolin RE 1993 Lipoprotein lipase regulation by insulin and glucocorticoid in subcutaneous and omental adipose tissues of obese women and men. J Clin Invest 92:2191 2198