High resolution "ultra performance" liquid chromatography coupled to oa-TOF mass spectrometry as a tool for differential metabolic pathway profiling in functional genomic studies

The combination of a new 1.7 mum reversed-phase packing material, and a chromatographic system, operating at ca. 12,000 psi, (so-called ultra performance liquid chromatography, UPLC) has enabled dramatic increases in chromatographic performance to be obtained for complex mixture separation. This increase in performance is manifested in improved peak resolution, together with increased speed and sensitivity. Here, we show that UPLC offers significant advantages over conventional reversed-phase HPLC amounting to a more than doubling of peak capacity, an almost 10-fold increase in speed and a 3- to 5-fold increase in sensitivity compared to that generated with a conventional 3.5 microm stationary phase. The first functional genomic application of UPLC-MS technology is illustrated here with respect to multivariate metabolic profiling of urines from males and females of two groups of phenotypically normal mouse strains (C57BL19J and Alpk:ApfCD) and a "nude mouse" strain. We have also compared this technology to conventional HPLC-MS under similar analytical conditions and show improved phenotypic classification capability of UPLC-MS analysis together with increased ability to probe differential pathway activities between strains as a result of improved analytical sensitivity and resolution.     

Physiological variation in metabolic phenotyping and functional genomic studies: use of orthogonal signal correction and PLS-DA

Metabolic phenotyping, or metabotyping, is increasingly being used as a probe in functional genomics studies. However, such profiling is subject to intrinsic physiological variation found in all animal populations. Using a nuclear magnetic resonance-based metabonomic approach, we show that diurnal variations in metabolism can obscure the interpretation of strain-related metabolic differences in two phenotypically normal mouse strains (C57BL10J and Alpk:ApfCD). To overcome this problem, diurnal-related metabolic variation was removed from these spectral data by application of orthogonal signal correction (OSC), a data filtering method. Interpretation of the removed orthogonal variation indicated that diurnal-related variation had been removed and that the AM samples contained higher levels of creatine, hippurate, trimethylamine, succinate, citrate and 2-oxo-glutarate and lower levels of taurine, trimethylamine-N-oxide, spermine and 3-hydroxy-iso-valerate relative to the PM samples. We propose OSC will have great potential removing confounding variation obscuring subtle changes in metabolism in functional genomic studies and will be of benefit to optimising interpretation of proteomic and genomic datasets.     

Decline in male mouse pheromone with age

An age-related decline in urinary-borne pheromone was found in male C57BL/6J mice aged from 2 to 30 months. Pheromone activity, estimated by bioassay, declined sharply after about 10 months of age. Two other strains of mice tested (DBA/2J and CBA/HT6J) also appeared to show an age-related decline in pheromone activity. Within each strain, however, pheromone activity was consistently similar to or higher than that of the C57BL/6J male mice. The DBA/2J and BALB/cWt strains appeared to be high pheromone producers, and the C57BL/6J and CBA/HT6J strains, low producers. This report is the first demonstration of a decline with age in male mouse pheromone activity. This decline appears to be synchronized with the well-defined loss of reproductive function in female mice.     

Differential DNA-repair activity in prespermiogenic cells of various mouse strains

The present report demonstrates differential DNA-repair activity among 14 strains of immature (20 ± 2 days old) male mice (inbred strains: C57BL/6J, RF/J, Nude homo/nu, RIII/2J, Pl/J, AKR/J, Nude hetro/nude, C3H/HeJ, SWR/J, SM/J, ST/J, LP/J, BALB/cJ and random-bred strain: CD-1). The prespermiogenic cells were isolated and enriched by collagenase-trypsin digestion of seminiferous tubules and subsequent 3% albumin-gradient centrifugation. Enriched prespermiogenic cells demonstrated a viabiilty greater than 95% by trypan blue exclusion criteria. For in vitro unscheduled DNA synthesis (UDS) determination, prespermiogenic cells (10⁶ cells/ml) were incubated with methyl methanesulfonate (0.4 mM) in the presence of 20 mM hydroxyurea (HU). At 20 mM HU concentration, 90% of S-phase DNA activity in prespermiogenic cells was inhibited and thus, the net UDS activity following MMS exposure was readily determined. MMS-induced UDS activity in the CD-1 mouse strain was both linear up to 4 h of incubation and dose-dependent at 4 h incubation. The apparent K_m for MMS-induced UDS activity in prespermiogenic cells was approx. 1.8 × 10^?4 M. Of the 14 mice strains tested, C57BL/6J and RF/J exhibited the highest DNA-repair activity, while BALB/cJ, LP/J, and ST/J showed the lowest. A maximal difference in UDS activity fo 3.5-fold was observed between C57BL/6J and BALB/cJ. Furthermore, a 2.5-fold difference was also noted between RF/J and LP/J mouse strains. Thus, wide variations in DNA-repair activity among 14 mouse strans were clearly demonstrated. Whether genetically select mouse strains with the lowest DNA-repair activity should have greater sensitivity toward environmental mutagens needs to be tested.     

Metabolic responses to leptin in obese db/db mice are strain dependent

Obese, diabetic C57BL/Ks db/db mice that lack the long-form leptin receptor exhibit no decrease in body weight or food intake when treated with leptin. Here we compared responses to leptin in two strains of db/db mice: C57BL/6J mice that are hyperglycemic and hyperinsulinemic and C57BL/Ks that are hyperglycemic and normo- or hypoinsulinemic. Chronic intraperitoneal infusion of 10 microgram leptin/day partially reversed hyperglycemia in C57BL/6J male mice but exaggerated the diabetic state of female mice. Bolus intraperitoneal injections of 40 microgram leptin/day did not effect glucose in either strain of male db/db mice, whereas chronic intraperitoneal infusion of 20 microgram leptin/day significantly reduced fasting blood glucose in male mice from both strains, especially C57BL/6J mice. Food intake, body weight, rectal temperature, and body fat did not change. Chronic intraperitoneal infusion of 10 microgram leptin/day significantly reduced body fat in lean db/+ C57BL/6J but not in C57BL/Ks mice. Thus peripherally administered leptin is active in mice that have only short-form leptin receptors, and the response is dependent on the method of leptin administration and the background strain.     

Tests of genetic allelism between four inbred mouse strains with absent corpus callosum.

Inbred mouse strains that lack the corpus callosum connecting the cerebral hemispheres in the adult differ from the C57BL/6J strain at several relevant but unknown loci. To identify at least one major locus that influences axon guidance, different strains showing phenotypically similar defects were crossed to test for allelism. If the F1 hybrid between two strains with the same brain defect is phenotypically normal, it is much more likely that the two strains will differ at fewer loci than will an acallosal strain and C57BL/6J. This approach proved to be very informative. Five reasonable models of inheritance involving two or three loci were assessed, and the data justified rejection of all but one hypothesis. A total of 479 mice were obtained from four inbred strains prone to absence of the corpus callosum (BALB/cWah1, BALB/cWah2, I/LnJ, and 129/ReJ), one normal strain (C57BL/6J), and 11 F1 hybrids among them. Because the size of forebrain axon bundles is generally greater in mice with larger brains, and because whole brain size is certainly polygenic, the phenotypically normal groups were used to derive a standard index of the degree of corpus callosum deficiency relative to brain size. Results demonstrated clearly that the hybrid between BALB/cWah1 and 129/ReJ is normal, whereas the crosses among the BALB/c substrains and I/LnJ yielded many mice with deficient corpus callosum. I/LnJ crossed with 129/ReJ also produced some animals with callosal defects. The data were consistent with a model in which the difference between BALB/c and 129/ReJ involves two loci, whereas the defect in I/LnJ involves homozygosity at three loci, which impairs development more severely.(ABSTRACT TRUNCATED AT 250 WORDS)     

Strain-dependent differences in responses to exercise training in inbred and hybrid mice

The aim of this study was to characterize the response to exercise training in several mouse strains and estimate the genetic contribution to phenotypic variation in the responses to exercise training. Male mice from three inbred strains [C57Bl/6J (BL6), FVB/NJ (FVB), and Balb/cJ (Balb/c)] and three hybrid F(1) strains [CB6F1/J (CB6 = female Balb/c x male BL6), B6F F(1) (female BL6 x male FVB), and FB6 F(1) (female FVB x male BL6)] completed an exercise performance test before and after a 4-wk treadmill running program. Distance was used as the primary estimate of endurance exercise performance. FVB mice showed the greatest response to training, with five- to sevenfold greater increases in distance run compared with BL6 and Balb/c strains. Specifically, BL6, FVB, and Balb/c strains increased distance by 33, 172, and 23%, respectively. A similar pattern of changes across strains was observed for run time (17, 87, and 11%) and work (99, 287, and 57%). As a group, F(1) hybrid mice derived from BL6 and FVB strains showed an intermediate response to training (61%). However, further analysis indicated that training responses in FB6 F(1) mice (80%) were approximately 2.5-fold greater than responses in B6F F(1) mice (33%, P = 0.08). A similar pattern of changes between FB6 and B6F F(1) mice was observed for run time (44.5 and 17%) and work (141 and 59%). These data demonstrate that there are large strain-dependent differences in training responses among inbred mouse strains, suggesting that genetic background contributes significantly to adaptation to exercise. Furthermore, the contrasting responses in B6F and FB6 F(1) strains show that a maternal component strongly influences strain-dependent differences in training responses.     

Strain differences of ether susceptibility in mice

The susceptibility to ether in the following six strains of mice was tested: C57BL/6, DBA/2, BALB/c, C3H/He, ICR and ddY. Mice of 4 weeks old were exposed to a flow of air containing various concentrations of ether for 90 min and the mortalities were assessed. The C57BL/6 strain was the most resistant and the C3H/He strain was the most sensitive to the lethal effect of ether. The susceptibilities of the closed colony mice, ICR and ddY, were intermediate between those of C57BL/6 and C3H/He mice. The DBA/2 and BALB/c strains were more sensitive than these closed colony mice and made up a sensitive group with the C3H/He strain. The LD50 values for ether in male mice of C57BL/6 and C3H/He were 6.0 and 3.1% atm, and in female mice of these strains were 6.6 and 3.2% atm, respectively. The ED50 value of ether which was accompanied by loss of righting reflex after exposure for 10 min was also higher in male C57BL/6 mice than in male C3H/He mice.     

H-Y antigen: No evidence for alleles in wild strains of mice

H-Y antigen(s) coded or controlled by the Y chromosome in a variety of wild mouse strains have been compared with those of the inbred laboratory strains C57BL/6 (B6) and C57BL/10 (B10). H-Y antigen(s) were detected by H-2-restricted cytotoxic T cells from B6 and B10 female mice primed in vivo and boosted in vitro with syngeneic male spleen cells: There was no difference in the degree of H-Y specific lysis of male cells from the C57BL strains and of F₁ hybrids or B6 congenic mice carrying the Y chromosome from the wild mouse strains examined. This result indicated that at the level of target cell specificity the H-Y antigen(s) from wild and laboratory strains were indistinguishable. H-Y antigen(s) were also found to be indistinguishable at the level of the in vitro induction of the anti H-Y cytotoxic response: F₁ female mice, primed in vivo and boosted in vitro with homologous F₁ male cells, all made H-Y-specific responses and where it could be examined, the target cell specificity of the anti-H-Y cytotoxic cells showed that B10 male cells as well as the homologous F₁ male cells (where the Y chromosome was derived from the wild strain) were good targets. Finally, possible differences in H-Y transplantation antigens between the wild strains and the B10 laboratory strain were examined by grafting F₁ male mice, the progeny of B10 females, and wild strain males with B10 male skin. These grafts were not rejected during an observation period of more than 9 months. Taken together, neither the cytotoxic data nor the skin graft data provide any evidence for allelism of H-Y even though the mouse strains examined were collected from widely disparate geographical locations.     

Anesthesia in the mouse using a combination of ketamine and promazine

An anesthetic combination of ketamine and promazine was tested in 25 male and 25 female mice of each of two strains. The mean induction time was 5 minutes for the C57BL/6We mouse and 8 minutes for the DBA/2We mouse. No sex differences were observed. The mean duration time was 45 minutes for the C57BL/6We male, 53 minutes for the C57BL/TWe female, 29 minutes for the DBA/2We male, and 35 minutes for the DBA/2We female. The mean recovery time was 39 minutes for the C57BL/6We male, 38 minutes for the C57BL/6We female, 21 minutes for the DBA/2We male, and 24 minutes for the DBA/2We female. The drug combination provided effective anesthesia for these strains of the laboratory mouse for a period of 30--50 minutes after a single intramuscular injection.     

Factors affecting the glycosphingolipid composition of murine tissues

The effects of genetic strain, sex, age, and pathological state on the distribution and concentration of glycosphingolipids in mouse kidneys and livers were studied. The concentrations of glycosphingolipids and phospholipids in the kidneys and livers of different strains were compared. The major glycosphingolipid in the kidneys of male and female BALB/c, C3H/He, C57/BL, A, and C57xA (F(1) hybrid) mice was a sulfatide, monoglycosyl-(3-sulfate) ceramide; monoglycosyl ceramide was the major component in livers. The kidneys of males of all strains contained significant amounts of diglycosyl ceramide, but those of females contained, at most, only traces. Glycosphingolipid concentration in the male kidneys of C57/BL and C57xA (F(1) hybrid) was much higher than in the female and was also much higher than in the male kidneys of C3H/He, BALB/c, and A strains. The kidneys of "old" (36 wk) male and female C3H/He mice contained much higher proportions of monoglycosyl ceramide than 10-12-wk-old adults. The distributions of glycosphingolipids in kidneys of female C3H/He mice with BP8 ascites tumors and male C57xA (F(1) hybrid) mice with EL4 ascites tumors differed from those in the normal mice. An unknown lipid, present in all glycosphingolipid extracts from kidney and liver, was tentatively identified as cholesterol sulfate.     

Genetic variation in androgen regulation of ornithine decarboxylase gene expression in inbred strains of mice

Previous studies have indicated that androgen regulation of certain gene products in murine kidney is genetically controlled. In the present work, the expression of renal ornithine decarboxylase (ODC) gene(s) was used as a biological marker to study androgen responsiveness of eight inbred strains of mice (A/J, C57BR/cdJ, 129/J, C57L/J, BALB/cJ, SM/J, RF/J, and C57BL/6J). Kidneys of untreated females from these strains did not have significantly different basal ODC activities or ODC mRNA concentrations. However, renal enzyme concentrations in intact male mice exhibited marked strain-dependent variation; three strains (RF/J, SM/J, and C57BR/cdJ) had 5- to 20-fold higher activities than the other five strains. Renal ODC mRNA content showed similar genetic variability in the male mice; animals with highest enzyme activity had higher mRNA levels than those with low activity. These results could not be explained by differences in either serum testosterone levels or renal nuclear androgen receptor content, suggesting that the animals were differentially sensitive to endogenous androgens. To evaluate further the androgen regulation of ODC gene expression, female mice were treated with testosterone-releasing implants for 5-7 days. The two strains (A/J and C57BL/6J) that had low enzyme activity in response to endogenous testosterone in male mice also showed blunted responses to exogenous androgen administration, as measured by the induction of ODC and its mRNA. The relative distribution of the two mRNA species coding for ODC (2.2 and 2.7 kb in size) exhibited strain-dependent variation that did not, however, correlate with the androgen responsiveness. Studies of the mRNA levels in reciprocal F1 hybrids of C57BR/cdJ and C57BL/6J mice suggested that androgen sensitivity of ODC gene expression, at least in these crosses, was inherited in an autosomal dominant manner.     

Genetic polymorphisms among C57BL/6 mouse inbred strains

Mice from the inbred C57BL/6 strain have been commonly used for the generation and analysis of transgenic and knockout animal models. However, several C57BL/6 substrains exist, and these are genetically and phenotypically different. In addition, each of these substrains can be purchased from different animal providers and, in some cases, they have maintained their breeding stocks separated for a long time, allowing genetic differences to accumulate due to individual variability and genetic drift. With the aim of describing the differences in the genotype of several C57BL/6 substrains, we applied the Illumina^® Mouse Medium Density Linkage Mapping panel, with 1,449 single nucleotide polymorphisms (SNPs), to individuals from ten C57BL/6-related strains: C57BL/6JArc, C57BL/6J from The Jackson Lab, C57BL/6J from Crl, C57BL6/JRccHsd, C57BL/6JOlaHsd, C57BL/6JBomTac, B6(Cg)-Tyr ^c?2j /J, C57BL/6NCrl, C57BL/6NHsd and C57BL/6NTac. Twelve SNPs were found informative to discriminate among the mouse strains considered. Mice derived from the original C57BL/6J: C57BL/6JArc, C57BL/6J from The Jackson Lab and C57BL/6J from Crl, were indistinguishable. Similarly, all C57BL/6N substrains displayed the same genotype, whereas the additional substrains showed intermediate cases with substrain-specific polymorphisms. These results will be instrumental for the correct genetic monitoring and appropriate mouse colony handling of different transgenic and knockout mice produced in distinct C57BL/6 inbred substrains.     

Differential effects of pentobarbital and ethanol in mice

The duration of the loss of righting reflex (RR) after ethanol, 4 g/kg, intraperitoneally (i.p.), was significantly longer in “long-sleep” (LS) than in “short-sleep” (SS) mice. This effect was shown to be correlated with differences in brain sensitivities to ethanol. In contrast, pentobarbital sodium (PB), 50 mg/kg, i.p., produced a significantly longer loss of RR in SS than in LS mice. The PB concentrations in the brain were the same in both mouse strains at the time of RR recovery suggesting equal sensitivities of the central nervous systems to PB. The rates of disappearance of PB from the blood were the same in both strains, but the apparent volume of distribution of PB in the LS strain was greater than in SS mice.In addition, C57BL/6J mice were found to be more sensitive than DBA/2J mice to PB, 50 mg/kg. In contrast, C57BL mice are known to be less sensitive than the DBA strain to ethanol. The PB concentration in the brain of DBA mice at the recovery of the RR was significantly greater than in C57BL mice. The apparent volumes of distribution of PB were not different in the two strains, but the rate of disappearance of PB from the blood of C57BL mice was significantly greater than for the DBA strain. In conclusion, factors which govern the brain sensitivities of selected mouse strains to ethanol and pentobarbital may not be equivalent.     

Electroencephalographic changes with age in male mice.

Electroencephalographic (EEG) changes, as measured by the awake state, slow-wave sleep (SWS), rapid-eye movement (REM) patterns and ratio of REM/total sleep, were recorded in aging male mice of DBA/2J and C57BL/6J strains. Results indicate that there is a significant increase in the awake state accompanied by significant decrease in SWS with advancing age for both strains, although these changes appear more pronounced in DBA/2J mice than C57BL/6J mice. Of considerable significance is the finding that REM sleep is absent in mice of DBA/2J strain at 23.5 months of age. Based on these findings, the conclusion was reached that strain DBA/2J ages significantly faster than C57BL/6J. The difference in aging between the two strains emphasizes the need for additional studies dealing with genetic aspects of aging.     

BSB: A New Mouse Model of Multigenic Obesity

We report here a new mouse model of multigenic obesity. Backcross progeny ((C57BL/6J x Mus spretus)F1 x C57BL/6J), designated as BSB mice, range from 1% to 50% body fat. Since both parental strains are relatively lean, the wide range of the phenotype in the BSB mice indicates the involvement of multiple genes to produce obesity. Obesity in BSB mice results from increases in both intraabdominal and subcutaneous fat and is associated with hyperinsulinemia, hyperglycemia, and hyperlipidemia. Female and male BSB mice do not differ in the degree of obesity obtained. Stimulated plasma corticosterone levels are reduced in obese male and female mice. The development of appropriate genetic markers and statistical methods have made it feasible to analyze quantitative polygenic traits in animal models by employing F2 or backcross progeny. Thus, this BSB model is uniquely suited to the genetic analysis of the multifactorial quantitative trait of obesity and its associated phenotypes. (OBESITY RESEARCH 1993;1:271–280)     

Dimethylnitrosamine metabolism: II. In vitro activation of dimethylnitrosamine by hepatic and renal tissues from crosses among BALB/c, DBA and C57BL mice

Crosses among BALB/c, C57BL and DBA mice were performed to investigate the genetic mechanisms involved in metabolism of DMN by renal and hepatic tissues. Liver S-9 fractions from parental strain DBA had the greatest potential to activate DMN and liver fractions from parental strain BALB/c had the lowest. No age or sex-related differences were observed within strain. Crossing of either C57BL or DBA to BALB/c mice resulted in F1 hybrids with liver microsomal enzymes that gave results similar to the BALB/c parental strain. There were no sex or age differences within crossbred strains in the potential of liver to activate DMN. In contrast male DBA and C57BL parental mice renal S-9 fractions did not differ significantly from each other but did differ significantly from male BALB/c renal fractions and from female and immature animals of all strains. Crossing of either DBA or C57BL mice with BALB/c mice resulted in male F1 hybrids whose renal S-9 fractions did not differ significantly from males of the parental BALB/c strain. In all instances, male renal S-9 fractions had a significantly greater potential to activate DMN than female or immature animals. F1 DBA X C57BL hybrids had renal S-9 fractions that did not differ significantly from the parental strains. These data suggest that the gene(s) for low DMN metabolism of BALB/c mice are apparently dominant over the genes from both DBA and C57BL. The exact genetic or physiological mechanism needs further elucidation.     

Diet-induced obesity in genetically diverse collaborative cross mouse founder strains reveals diverse phenotype response and amelioration by quercetin treatment in 129S1/SvImJ,PWK/EiJ,CAST/PhJ,and WSB/EiJ mice

Significant evidence suggests protective effects of flavonoids against obesity in animal models, but these often do not translate to humans. One explanation for this disconnect is use of a few mouse strains (notably C57BL/6 J) in obesity studies. Obesity is a multifactorial disease. The underlying causes are not fully replicated by the high-fat C57BL/6 J model, despite phenotypic similarities. Furthermore, the impact of genetic factors on the activities of flavonoids is unknown. This study was designed to explore how diverse mouse strains respond to diet-induced obesity when fed a representative flavonoid. A subset of Collaborative Cross founder strains (males and females) were placed on dietary treatments (low-fat, high-fat, high-fat with quercetin, high-fat with quercetin and antibiotics) longitudinally. Diverse responses were observed across strains and sexes. Quercetin appeared to moderately blunt weight gain in male C57 and both sexes of 129S1/SvImJ mice, and slightly increased weight gain in female C57 mice. Surprisingly, quercetin dramatically blunted weight gain in male, but not female, PWK/PhJ mice. For female mice, quercetin blunted weight gain (relative to the high-fat phase) in CAST/PhJ, PWK/EiJ and WSB/EiJ mice compared to C57. Antibiotics did not generally result in loss of protective effects of quercetin. This highlights complex interactions between genetic factors, sex, obesity stimuli, and flavonoid intake, and the need to move away from single inbred mouse models to enhance translatability to diverse humans. These data justify use of genetically diverse Collaborative Cross and Diversity Outbred models which are emerging as invaluable tools in the field of personalized nutrition.     

不同品系小鼠对代谢性高尿酸血症造模的影响

目的：分别以昆明种小鼠及ICR、C57BL/6J小鼠为研究对象,比较在复制高尿酸血症模型时可能的小鼠品系差异,并通过降尿酸药物别嘌呤醇与非布索坦验证选择降尿酸药物筛选时选用不同品系动物造模的影响。方法：采用不同剂量次黄嘌呤腹腔注射联用尿酸酶抑制剂氧嗪酸钾皮下注射给药,测定不同造模时段各品系小鼠血清尿酸值。结果：ICR、C57BL/6J小鼠对高尿酸血症造模耐受显著高于昆明种小鼠,在腹腔注射次黄嘌呤500mg/kg,皮下注射氧嗪酸钾300mg/kg时,才可获得稳定的可用于药物筛选的高尿酸血症模型。结论：选择高尿酸血症在体模型时,昆明种小鼠灵敏度高于ICR小鼠以及近交系的C57BL/6J小鼠。     

cis-acting determinants of basal and lipid-regulated apolipoprotein A-IV expression in mice

The levels of apolipoprotein A-IV (apoA-IV) mRNA are regulated by dietary lipid in the liver of both the mouse and rat. Thirteen different inbred mouse strains were fed a high lipid diet, and the effect on apoA-IV liver mRNA levels was examined. It was found that each strain responded in one of two ways. Mice of four strains had higher liver apoA-IV mRNA levels as compared with syngeneic mice fed a normal chow diet. Mice of the other nine strains had decreased liver apoA-IV mRNA levels as compared with syngeneic mice fed a normal chow diet. Using F1 hybrids between mice from BALB/c, C3H, and C57BL/6 and between 129 and C57BL/6, as well as recombinant inbred strains derived from a cross between BALB/c and C57BL/6, we have shown that both the normal level of liver apoA-IV mRNA in the chow-fed mice and the lipid-dependent regulation of apoA-IV mRNA levels are controlled by cis-acting genetic elements. The apoA-IV mRNA levels in mice fed a normal diet varied dramatically among strains, with the largest difference (90-fold) being between the 129/J inbred strain and the C57BL/6J strain. In addition, we have examined the expression of apoA-IV during mouse development. ApoA-IV mRNA is expressed early in mouse liver (16 days postcoitum), whereas others have shown previously that rat liver apoA-IV mRNA is undetectable until 14 days after birth. ApoA-IV mRNA levels in the intestine and apoA-I mRNA levels in the liver and intestine, by contrast, mirror the pattern seen in the rat.