This Article Abstract Full Text (PDF) All Versions of this Article: 145/7/3307    most recent Author Manuscript (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Goren, H. J. Articles by Kahn, C. R. Search for Related Content PubMed PubMed Citation Articles by Goren, H. J. Articles by Kahn, C. R.

Glucose Homeostasis and Tissue Transcript Content of Insulin Signaling Intermediates in Four Inbred Strains of Mice: C57BL/6, C57BLKS/6, DBA/2, and 129X1

Research Division, Joslin Diabetes Center and Department of Medicine, Harvard Medical School, Boston, Massachusetts 02215

Address all correspondence and requests for reprints to: H. J. Goren, Department of Biochemistry and Molecular Biology, University of Calgary, 3330 Hospital Drive Northwest Calgary, Alberta, Canada T2N 4N1. E-mail: goren{at}ucalgary.ca

	Abstract

Top Abstract Introduction Materials and Methods Results and Discussion Conclusion References

 
Transgenic mice phenotypes generally depend on the background strains used in their creation. To examine the effects of genetic background on insulin signaling, we analyzed glucose homeostasis in four inbred strains of mice [C57BL/6 (B6), C57BLKS/6 (KLS), DBA/2 (DBA), and 129X1] and quantitated mRNA content of insulin receptor (IR) and its substrates in insulin-responsive tissues. At 2 months, the male B6 mouse is the least glucose-tolerant despite exhibiting similar insulin sensitivity and first-phase insulin secretion as the other strains. The 129X1 male mouse islet contains less insulin and exhibits a higher threshold for glucose-stimulated first-phase insulin secretion than the other strains. Female mice generally manifest better glucose tolerance than males, which is likely due to greater insulin sensitivity in liver and adipose tissue, a robust first-phase insulin secretion in B6 and KLS females, and improved insulin sensitivity in muscle in DBA and 129X1 females. At 6 months, although males exhibit improved first-phase insulin secretion, their physiology was relatively unchanged, whereas female B6 and KLS mice became less insulin sensitive. Gene expression of insulin signaling intermediates in insulin-responsive tissues was generally not strain dependent with the cell content of IR mRNA being highest. IR substrate (IRS)-1 and IRS-2 mRNA are ubiquitously expressed and IRS-3 and IRS-4 mRNA were detected in significant amounts in fat and brain tissues, respectively. These data indicate strain-, gender-, and age-dependent tissue sensitivity to insulin that is generally not associated with transcript content of IR or its substrates and should be taken into consideration during phenotypic characterization of transgenic mice.

	Introduction

Top Abstract Introduction Materials and Methods Results and Discussion Conclusion References

 
WITH THE MANY advances made in molecular biology, investigators have been able to create transgenic mice where genes are mutated, knocked in, knocked out, ubiquitously, in a tissue-specific site, or in a time- dependent manner (1, 2). Historically, mouse strains selected for creation of transgenics were chosen for their ease of use. For example, the FVB/N strain, because of its large zygote, relative resistance to microinjection of pronuclei, and good fertility and fecundity, has found wide use (3). Cultured embryonic stem (ES) cells used in creation of transgenic mice are derived from the 129 mouse (4), and transformed ES cells are implanted into blastocysts, typically of the C57BL/6 (B6)³ strain. The chimeric mouse so produced carries the genetic background of both the recipient strain and the 129 strain. The 129 strains, however, are a complex family of sub-strains (5), and therefore the genome of the chimeric mouse is more heterogeneous than 2-fold. With the mating of a male chimeric mouse with a female B6 or other strain, the F2 generation will carry the mutated gene as well as genetic material from the 129 mouse, the B6 mouse, and the breeding mouse. The genetic background of mice used in creation of transgenics have proven countless times to contribute to their phenotypes (2, 5, 6, 7, 8). Accordingly, to correctly interpret the physiological consequence of a mutated gene, two approaches have evolved: 1) back-crossing transgenic mice into recipient strains, and 2) careful physiological analyses of recipient strains before ES cell implantation.

Creation of congenic strains such as back-crossing 10 generations to the recipient strain reduces 129 background genes to 0.1% of all genes (5). In such mice, the phenotype should reflect the role of the mutated gene. One exception is if flanking 129 genes to the mutated gene contributes to the phenotype (9). Another exception is when the recipient mouse has modifying genes that alters the response of a mutated gene (10). Detailed physiological analyses of recipient strains should forewarn the potential for the latter to occur.

The interaction of genetic and environmental factors can produce changes in insulin action, insulin secretion, hyperglycemia, and type 2 diabetes (11). To aid in our understanding the role of different genes in type 2 diabetes, a large number of transgenic mice have been created (12, 13, 14, 15). Not unexpectedly, the phenotypes of these mice are sometimes different when they are created in different laboratories (16). Described in this report are the measurements of physiological parameters associated with insulin action in mouse strains frequently used in creation of transgenic mice.

Insulin regulates plasma glucose by inhibiting gluconeogenesis in the liver and by stimulating glucose uptake and metabolism in muscle and fat (13, 17). These responses are measured with glucose tolerance tests, insulin tolerance tests, and fasting and fed glucose and insulin levels. These measurements were made in four inbred strains of mice: B6, DBA/2 (DBA), and 129X1, strains widely used in transgenic studies, and C57BLKS/6 (KLS), a B6 mouse contaminated with DBA (18, 19). Because first-phase glucose-stimulated insulin secretion is an important regulator of glucose homeostasis (20, 21), we also measured this phase of insulin secretion in response to glucose and arginine.

Insulin signaling is initiated subsequent to binding to the insulin receptor. The insulin receptor (IR) complex autophosphorylates and subsequently phosphorylates a number of cytoplasmic proteins, among which are IR substrate (IRS)-1, IRS-2, IRS-3, and IRS-4. To determine whether different inbred strains of mice express different levels of insulin signaling intermediates, we measured mRNA for the IR and for each of the IRSs in liver, muscle, fat, islet, brain, kidney, and spleen. In addition, mRNA for the IGF-I receptor was measured; IGF-I receptor binds insulin weakly and is able to elicit some insulin responses (22).

We have found that B6, KLS, DBA, and 129X1 mice do not conform in how they maintain glucose homeostasis, in what they consider euglycaemia, and in how much insulin they need to achieve this. Furthermore, at the mRNA level, gene expression of the measured insulin signaling intermediates is with few exceptions neither tissue nor strain dependent. These findings will be of value in interpretation of phenotypic data of transgenic mice previously reported, and in selection of strains for creation of additional transgenic mice in functional analyses of insulin signaling intermediates.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results and Discussion Conclusion References

 
B6 and KLS mice were purchased from The Jackson Laboratory (Bar Harbor, ME), and DBA and 129X1 mice were purchased from Taconic Laboratory (Germantown, NY). Mice were fed ad libitum and kept on a 12-h light, 12-h dark cycle. All protocols for animal use and euthanasia were reviewed and approved by the Institutional Animal Care Committee of the Joslin Diabetes Center and were in accordance with the guidelines of the National Institutes of Health.

Physiological parameters
For insulin tolerance tests (21), fed mice were injected ip with insulin [1 U/kg body weight (b.wt.); Humulin, Lilly (Indianapolis, IN)] and tail vein glucose was measured at 0, 15, 30, and 60 min after injection using an automatic glucometer (One Touch, Lifescan, Milpitas, CA). Insulin tolerance tests were performed at 1300–1400 h.

For glucose tolerance tests, mice were fasted overnight (14 h) then injected ip with glucose (2 g/kg b.wt.). Tail vein blood glucose was measured as described above at 0, 15, 30, 60, and 120 min after injection.

For insulin secretion assays, mice were fasted overnight and then injected ip with glucose (3 g/kg b.wt.) or with arginine (0.3 g/kg b.wt.) to determine glucose-stimulated secretion or arginine-stimulated secretion, respectively. An automatic glucometer measured tail vein blood glucose at 0, 2, and 5 min after stimulus injection. In addition, 30 µl of whole blood were collected in 3 µl heparin (5000 U/ml) at these time points.

Fasting glucose levels represent the mean of glucose plasma concentrations at zero time point in glucose tolerance tests and in insulin secretion assays. Fed glucose levels are mean glucose concentrations at 0 time in insulin tolerance tests and when plasma was collected for fed insulin measurements. Fasted insulin or leptin levels are means of measured insulin or leptin, respectively, at zero time points from glucose- and arginine-stimulated secretion assays. To determine fed insulin or leptin levels 30-µl blood samples were collected at 1300–1400 h on days when mice were not used for physiological assays. Whole blood samples were centrifuged (10 min at 10,000 rpm), supernatants were collected, and stored at –20 C. Using mouse insulin and mouse leptin as standards, we measured supernatant insulin and leptin levels, respectively, with ELISA kits (Crystal Chem, Downer’s Grove, IL).

To isolate mouse islets, bile ducts of Nembutal-anesthetized mice were injected, anteretrogradely, with collagenase. After exsanguination, the pancreas was excised from the mouse and incubated for 16 min at 37 C. Isolated islets and acinar cells were enriched for islets after their centrifugation over Histopaque 1077 (Sigma, St. Louis, MO) (23). The latter was washed free of Histopaque with Media 1099 plus 10% fetal calf serum, and individual islets were picked under a dissecting microscope (GZ7 Stereozoom; Leica, Deerfield, IL). Size-matched islets from 2-month-old male B6 and 129X1 mice, were incubated at 37 C for 48 h in RPMI-10% fetal calf serum medium with glutamine. Islets (25 islets/well) were washed free of medium and incubated 37 C, 30 min with 200 µl RPMI medium containing 5, 8, or 11.1 mM glucose. Medium was removed, and islets were acid-ethanol extracted. Insulin content in media and extracts were measured immunologically, as described above.

Tissue mRNA content
After removal of the pancreas, as described above, liver, muscle, fat, kidney, spleen, and brain were excised rapidly, immediately frozen in liquid nitrogen, and stored at –80 C. Isolated islets, not used for insulin secretion assays, were also stored at –80 C. Tissues (10–30 mg) were ground under liquid nitrogen. Ground tissues and islets were lysed in RLT buffer containing 10 µl/ml 2-mercaptoethanol (RNeasy Mini Kit; QIAGEN, Valencia, CA). DNA was sheared by centrifugation of lysates through Qiashredder columns (QIAGEN). RNA was isolated from filtrate as per instructions in the Rneasy Mini Kit technical bulletin. All RNA preparations were deoxyribonuclease I (QIAGEN) digested, 15 min 20 C, in 50 mM Tris, 10 mM MgCl₂, 0.1 mM dithiothreitol (pH 7.5), and then purified with Rneasy Mini Kit Columns. RNA was stored at –20 C.

Transcript copy numbers of IR, IGF-I receptor, IRS-1, IRS-2, IRS-3, and IRS-4 were measured using the Applied Biosystems (ABI, Foster City, CA) Taqman procedure in an ABI 7700 instrument. Primers (Joslin Diabetes Center DNA Core Facility) and probes (ABI) specific for each IRS protein, IR, and IGF-I receptor were synthesized (Table 1). The ABI 7700 uses a fluorescence detection system to follow PCRs, and a plot of fluorescence vs. cycle number generates a sigmoid curve. For this study, cycle threshold (C_t) was the cycle number at a fixed point just above the lower inflection point of a sigmoid curve.

(1A)

(1B)

Reaction conditions to determine GAPDH mRNA included the absence and presence of reverse transcriptase. Because in the absence of reverse transcriptase no cDNA is formed, then under this reaction condition the analysis assured the absence of genomic DNA in the RNA preparation.

Plasmids containing cDNA for each of the IRSs (24) were used to correlate C_t with moles of an IRS in a reaction well. Multiplication of the moles of an IRS with Avogadro’s number converted the moles to copy numbers. Real-time PCR analyses of several concentrations of each of the IRS plasmids were used to prepare calibration curves that compared mRNA^IRS copy number ({mRNA^IRS}) to C_t (equations 2–5). These equations are applicable when reaction mixtures contain 90 nM primers and 100 nM probe. The plasmids also confirmed that each primer-probe combination were uniquely specific for the IRS for which they were designed.

(2)

(3)

(4)

(5)

The sequence of the probes for IR and IGF-I receptor differs at nucleotide 9 (thymidine/cytosine, IR/IGF-I receptor). Consequently, they can serve as probes for either receptor; specificity is achieved through choice of primer pair. We observed that the fluorescent tags at the 5' ends of probes predicted the choice of calibration curves (equations 2–5). Accordingly, to calculate the relative mRNA expression of the IR (mRNA^IR) and IGF-I receptor (mRNA^IGF-IR), we used equation 2 (5'-Fam-probe) and equation 3 (5'-Tet-probe), respectively. However, because these probes with different primer pair could be used to quantitate the other receptor, data so generated supported the use of these equations as described above and provide an estimate of the expression levels.

The total content of RNA in a mammalian cell is reported to be 26 pg (25). Because total RNA content (nanograms) in each reaction well was determined from GAPDH analysis, the number of cells this represented could be determined (1 ng RNA = 38.5 cells). This allowed transcript copy number to be expressed per cell. However, in light of the uncertainty of constancy of GAPDH mRNA content in cells between tissues and between strains, we express our data as relative mRNA expression.

We confirmed the reproducibility of the experimental and analysis methods. That is, IRS-1 mRNA was determined in six tissues, taken from five different KLS mice, and each tissue analyzed two to four times. The SE of measurement within an extract was less than 10%, and with two or three exceptions the variability between tissue extracts from different mice was also less than 10% (data not illustrated).

Statistics
Paired and unpaired Student’s t tests (Sigmaplot or Sigmastat) were used to test for significance of difference between samples and between mice. When two data sets had a P value above 0.05, they were considered not significantly different.

	Results and Discussion

Top Abstract Introduction Materials and Methods Results and Discussion Conclusion References

 
Physiology of inbred strains of mice
Roles for a wide variety of molecules in physiological processes have been examined using mice in which a gene of interest has been mutated, overexpressed, or deleted in the whole animal or in specific tissues (1). In the case of insulin, several molecules in the insulin signaling cascade, including IRS-1, -2, -3, and -4 and a variety of post-receptor molecules have been knocked out by several groups (12, 13, 14, 16, 26, 27, 28, 29, 30, 31, 32, 33, 34). The phenotype of knockout mice between laboratories, however, is not always identical. For example, the phenotype of the IR knockout mouse created by the Acilli laboratory (genetic background 129Sv and B6) (35) is more severe than the one created by Joshi et al. (genetic background 129Sv, B6 and B6D2F1, a B6-DBA cross) (36), and the phenotype of the IRS-2 knockout mouse of Kadowaki and colleagues (37) (genetic background CBA, ICR, and B6) is less severe or even normal in comparison to that created by Withers et al. (genetic background 129Sv and B6) (38). Although in creation of three of four of the latter transgenic mice ES cells from the 129 strain were used, each ES cell had a different origin. Because significant genetic variation among the 129 substrains exists (5), then, depending on the source of the ES cell, each transgenic mouse could elicit a different phenotype. In addition to ES cell, the choice of mouse strain used to breed to chimeric males will influence the phenotype. Thus, mice heterozygous for the absence of IR and IRS-1 are insulin resistant, but the degree of resistance is dependent on whether the breeding mouse is B6, DBA, or 129Sv (39). Because both the source of an ES cell and mice used in breeding influence the phenotype of transgenic mice (2), we examined the ability of the B6, KLS, DBA, and 129X1 inbred strains to maintain glucose homeostasis under physiological and pharmacological conditions, information that should be helpful in interpretation of transgenic studies.

Male mice.
In the fasted state, glucose levels ranged from 68–102 mg/dl (Fig. 1B) and insulin levels from 276–377 pg/ml (Fig. 1D) in 2-month-old B6, KLS, DBA, and 129X1 mice. Fasting glucose was lower in 129X1 mice (P = 0.001) but higher in B6 mice (P = 0.004) when compared with the KLS or DBA strains. In DBA mice, fasting insulin was higher than in the KLS (P = 0.03) and 129X1 strains (P = 0.001). Fed insulin levels (Fig. 1D) for the four inbred strains of mice were not statistically different (P > 0.05), but fed glucose levels (Fig. 1B) showed considerable variations with B6> KLS> DBA>129X1 (P < 0.05). With feeding, the increase in insulin and glucose was statistically significant for each inbred strain of mouse (P < 0.001). Because fed insulin levels were similar in the four inbred strains of mice, the presence of differences in fed glucose levels suggested that the determinant of glucose homeostasis in these mice was sensitivity of their fat, muscle, and liver tissues to insulin.

View larger version (32K): [in this window] [in a new window]   FIG. 1. Glucose homeostasis in inbred strains of male mice at 2 months of age. Male mice, 8–10 wk of age, were subjected to glucose tolerance tests (A) and insulin tolerance tests (C) as described in Materials and Methods. Fasting glucose (B) are the plasma glucose levels at t = 0 for glucose tolerance tests and for acute-phase insulin secretion assays. Fed glucose (B) are the plasma glucose levels before the collection of plasma for fed insulin measurements. D, Tail-vein blood after an overnight fast (Fasting insulin) or at 1300 h. (Fed insulin), was collected and plasma insulin (d) was measured (ELISA). The inbred strains were B6 (n = 15), KLS (n = 15), DBA (n = 12), and 129X1 (n = 12). Values are mean ± SEM (unseen error bars are within symbols). In panel A, * denotes significant difference (P < 0.05) from data of DBA and KLS mice. In panel B, fasting glucose was significantly higher in B6 and lower in 129X1 mice than in DBA or KLS mice, and fed glucose levels were significantly different (*) between each strain and significantly higher than fast levels. In panel C, the relative decrease in glucose in DBA and 129X1 strains were significantly different (*) at 15 and 30 min from B6 and KLS mice, but at 60 min only the 129X1 mouse was significantly different (+) from KLS and DBA mice. In panel D, DBA has significantly elevated (*) fasting insulin, whereas other fasting and all fed insulin levels in the four strains of mice were not statistically different. All fed insulin levels were statistically higher than fasted insulin levels (P < 0.001).

View larger version (22K): [in this window] [in a new window]   FIG. 2. Acute-phase glucose- and arginine-stimulated insulin secretion. Male and female mice, 2 months of age, were fasted overnight. Glucose (A and C; 3 g/kg b.wt.) or arginine (B and D, 0.3 g/kg b.wt.) were injected ip and 30 µl tail-vein blood was collected at 0, 2, and 5 min in 3 µl heparin (5000 U/ml). Plasma insulin was measured (ELISA). Results are mean ± SEM for B6 (n = 15), KLS (n = 15), DBA (n = 12), 129X1 (n = 12). At 2 min post stimulus injection, insulin levels were statistically (*, P < 0.05) elevated as shown. Between strains at 2 min post ip glucose injection insulin levels in B6, KLS, and DBA male mice were similar (P > 0.1) and greater than in 129X1 mice (A, P < 0.05), and in female B6 and KLS were similar (P > 0.1) and greater than in DBA (P < 0.05) and the latter was greater than in 129X1 mice (C, P < 0.05). Insulin levels 2 min post ip arginine injection in male and female mice were greater than before stimulus injection (B and D, P < 0.05), were highest in male and female B6 mice (P < 0.05), and in male mice insulin levels were higher in KLS and DBA mice than in 129X1 mice (P < 0.05).

B6, KLS, and DBA strains, in insulin tolerance tests, at 60 min had cleared approximately 40% of their initial fed glucose levels; 129X1 clearance was slightly less (Fig. 1C). These results suggested that a pharmacological dose of insulin is similarly effective in clearing plasma glucose in these mice (3+ relative insulin sensitivity of muscle for B6, KLS, and DBA, and 2+ for 129X1 mice; Table 2). DBA and 129X1 mice, however, demonstrated different kinetics of insulin-stimulated glucose clearance. Because insulin stimulates the translocation of glucose transporter-4 (Glut-4) from endosomal membranes to plasma membranes in fat and muscle (41), differences in rates of glucose clearance, as measured in insulin tolerance tests, likely reflect differences of rates of Glut-4 translocation and/or rates of insulin signal transduction.

Notably, the first phase of glucose-stimulated insulin secretion was absent in 129X1 mice (Fig. 2A; zero relative glucose sensitivity; Table 2). Furthermore, the amount of insulin secreted in response to arginine in 129X1 mice was less than in all the other strains (P = 0.03), and was much lower than in B6 mice (P = 0.01, Fig. 2B; relative arginine sensitivity of islets B6 3+, KLS and DBA 2+, and 129X1 +; Table 2). Although 3 g glucose/kg b.wt. was unable to stimulate insulin secretion in 129X1 mice, glucose did potentiate arginine-stimulated insulin secretion (Fig. 3), a finding that indicates that glucose can mediate its signaling events in 129X1 ß-cells (42, 43).

View larger version (15K): [in this window] [in a new window]   FIG. 3. Acute-phase insulin secretion in 129X1 male mice. Male 129X1 mice, 2 months old, were fasted overnight. Glucose 3 g/kg b.wt., arginine 0.3 g/kg b.wt., or glucose and arginine at the latter doses were injected ip. Tail-vein blood was collected and assayed for insulin content. Results are mean ± SEM (n = 12). At 2 and 5 min post stimulus injection, insulin levels were statistically (*, P < 0.05) elevated.

View larger version (20K): [in this window] [in a new window]   FIG. 4. Acute-phase glucose-stimulated insulin secretion in B6 and 129X1 male mice. Two-month-old male B6 mice and 129X1 mice were fasted overnight. Tail-vein blood was collected at 0 and 2 min post ip glucose injection. Glucose (A) and insulin (B) levels were measured as described in Materials and Methods. B6 mice (n = 9) were injected with 3 g/kg b.wt. glucose. 129X1 mice were injected with 3 g/kg b.wt. (n = 3), with 6 g/kg b.wt. (n = 2), with 9 g/kg b.wt. (n = 6), or with 12 g/kg b.wt. (n = 7). Fasting glucose/insulin levels (t = 0) are mean ± SEM (n = 18) for 129X1 mice. *, Signifies significant difference (P < 0.05) from fasting levels.

View larger version (24K): [in this window] [in a new window]   FIG. 5. Insulin secretion from islets isolated from B6 and 129X1 male mice. Isolated islets from 2-month-old male B6 and 129X1 mice were incubated, 37 C, 48 h in RPMI medium with fetal calf serum. A, Islets, 25 per well (n = 4), were incubated with serum-free medium containing 5, 8, or 11.1 mM glucose, 37 C, 30 min. Insulin content of supernatant and of islets was measured, as described in Materials and Methods. B, In a separate set of wells (n = 4), insulin and DNA content of islets, were measured. Asterisk denotes statistically significant difference (P < 0.05) between B6 and 129X1.

Female mice.
The degree of susceptibility a mouse has to diabetes is strain and gender dependent (45). Not unexpectedly, the gender of the four inbred strains had a marked effect on their ability to respond to a glucose challenge and to insulin. First, fasting glucose levels were lower in female mice (56–69 mg/dl) than in male mice (68–102 mg/dl; P < 0.001 for B6, KLS, and DBA mice and P = 0.014 for 129X1 mice; Fig. 1B cf. Fig. 6B). In B6 and DBA strains fasting insulin levels were gender independent (P > 0.05) but in KLS and 129X1 female mice fasting insulin levels were higher (481.4 ± 77.4 pg/ml cf. 299.1 ± 19.9 pg/ml, P = 0.02; and 397.3 ± 36.4 pg/ml cf. 276.2 ± 15.0 pg/ml, P < 0.001, respectively). Fed glucose levels were uniformly lower in female than in male mice (116–146 mg/dl cf. 121–180 mg/dl; P < 0.05; Fig. 6B cf. Fig. 1B). Fed insulin levels were lower in female B6 and KLS mice (P < 0.05) but were similar in DBA and 129X1 mice (P > 0.1, Fig. 6D cf. Fig. 1D). These findings with fasting and fed glucose and insulin levels indicated that in B6, KLS, DBA and 129X1 strains, female mice are generally more sensitive to insulin than their respective males.

View larger version (29K): [in this window] [in a new window]   FIG. 6. Glucose homeostasis in female inbred strains of mice at 2 months of age. Female mice 8–10 wk old were subjected to glucose tolerance tests (A), and insulin tolerance tests (C), and fasting and fed glucose (B) and insulin levels (D) were measured as described in legend to Fig. 1. The inbred strains of mice were B6 (n = 15), KLS (n = 15), DBA (n = 12), and 129X1 (n = 12). In panel A, asterisks denote which strain differ significantly (P < 0.05) from other strains at time points between 15 and 120 min. In panel B, fasting glucose levels in KLS (n = 57) and 129X1 (n = 18) mice were lower (*, P < 0.05) than in B6 (n = 60) or DBA (n = 36) mice, fed glucose were higher (*, P < 0.05) in B6 (n = 12) mice and lower (*, P < 0.05) in 129X1 mice (n = 12) than levels in KLS (n = 14) or DBA (n = 12) mice, and fed glucose levels were significantly higher than fasting glucose levels in each mouse strain (*, P < 0.05). In panel C, at 60 min the % decrease of initial glucose was greater in DBA and 129X1 mice (*, P < 0.05). In panel D, fed insulin levels were lower (*, P < 0.05) in KLS mice than in other strains, and fed insulin levels were higher than fasting insulin levels in B6, KLS, and DBA mice (*, P < 0.05).

For 129X1 mice, muscle appears to be the tissue responsible for the improvement because, in insulin tolerance tests, insulin lowered glucose levels 60% in 60 min in female 129X1 mice (relative insulin sensitivity for muscle 3+; Table 2), whereas in male mice it lowered glucose approximately 35% (Fig. 6C cf. Fig. 1C; P < 0.05; relative insulin sensitivity for 129X1 male muscle 2+; Table 2). Insulin sensitivity of DBA muscles is gender independent as insulin tolerance tests in male and female mice were similar (Fig. 1C cf. Fig. 6C; P > 0.05, relative insulin sensitivity 3+; Table 2). In contrast, female B6 and KLS muscles were less insulin sensitive than B6 and KLS male muscles; i.e. insulin cleared glucose 20–30% in female mice compared with 40–50% in male mice (Fig. 6C cf. Fig. 1C; P < 0.05; relative insulin sensitivity of muscle 2+ female cf. 3+ male; Table 2).

After an ip injection of glucose, the glucose excursion was higher and persisted longer in males than in females (Fig. 6A cf. Fig. 1A; P < 0.001). That is, in the four inbred strains studied, female mice were more glucose tolerant than male mice. The improved glucose tolerance, however, was not due to improved acute phase insulin secretion because peak (2 min post ip glucose injection) insulin levels were statistically equivalent (P > 0.1) in males and females (Fig. 2, A cf. C; relative glucose sensitivity 3+ male and female; Table 2). Because B6 and KLS female mice have less insulin-sensitive muscles than muscles in male B6 and KLS mice (Table 2), it follows then that hepatic and adipose tissues in female B6 and KLS mice are more insulin sensitive. Although improved, the relative insulin sensitivity of female B6 mouse liver and fat tissues is not as great as these tissues in female KLS mice: B6 mice have higher fed glucose levels (Fig. 6B) and poorer glucose tolerance (Fig. 6A). Accordingly, the relative insulin sensitivities of these tissues in B6 and KLS female mice are ‘2+’ and ‘3+’, respectively (Table 2), improvements over the insulin sensitivities in male mice (+ and 2+; Table 2). Finally, because insulin sensitivity of DBA muscles is gender independent, then the improved glucose tolerance in female DBA mice may be attributed to improved insulin sensitivity of their liver and fat tissues (relative insulin sensitivity 2+ cf. 3+; Table 2).

The improved insulin sensitivity of tissues in female mice lowered their fasting glucose levels compared with their male counterpart (Fig. 6B cf. Fig. 1B; P 0.01), but fasting insulin in female compared with male mice was only lower in the 129X1 mouse (P = 0.02). Thus, because fasting insulin concentrations were gender independent but fasting glucose levels were gender dependent, then gluconeogenesis in female B6, KLS, and DBA livers is lower. The latter findings augment the improvement in insulin sensitivity in livers of female mice over livers in male mice, as discussed above (Table 2).

Peak insulins in first phase arginine-stimulated insulin secretions were statistically higher in KLS, DBA, and 129X1 females than in males (Fig. 2, B cf. D; P < 0.05; relative arginine sensitivity 3+ for female mice cf. 2+ or + for male mice; Table 2). B6 insulin levels 2 min post ip arginine injection were gender independent (Fig. 2, B cf. D, P > 0.1), and in females were statistically equivalent to those in the three other inbred strains studied (Fig. 2D; P > 0.05; relative arginine sensitivity 3+; Table 2).

Effect of aging.
Insulin signaling regulates the aging process; i.e. disruption of signaling slows the process (46, 47, 48, 49). Because we have found that insulin sensitivity is strain and gender dependent, we examined whether with time insulin sensitivity itself is changed: i.e. does aging regulate insulin sensitivity?

Male.
As the four inbred mouse strains aged from 2–6 months, they demonstrated minimal changes in physiology (Fig. 7). Tissue sensitivity to insulin, however, did change. Fed glucose levels in the B6 mouse strain did drop about 35 mg/dl between 2 and 6 months of age (P < 0.001; Fig. 7A). Because glucose levels during the 30- and 60-min time points were unaffected during the insulin tolerance tests for B6 mice (Fig. 7C), then the lower fed glucose levels likely reflect improved insulin sensitivity in liver or fat tissue. The latter improvement did not manifest itself in altered fed insulin levels because these were equivalent in B6 mice at 2 and 6 months (P > 0.1; Fig. 7D). Fed insulin in KLS, DBA, and 129X1 strains were also unaffected with aging (P > 0.1; Fig. 7, H, L, and P). In contrast to fed insulin, fasting insulin rose with age in all strains except in the DBA mouse (Fig. 7, D, H, L, and P; P < 0.05). Fasting glucose (zero time values in glucose tolerance tests, Fig. 7, B, F, J, and N), however, was not affected by age. That is, to prevent excess gluconeogensis under fasting conditions, an increase in insulin is required, suggesting that insulin sensitivity of livers in B6, KLS, and 129X1 mice decreased with age (Table 2). The improvement with age of fed glucose in B6 mice, therefore, is likely a result of improved insulin sensitivity of fat tissue (Table 2).

View larger version (38K): [in this window] [in a new window]   FIG. 7. Effect of aging on glucose homeostasis in inbred strains of male mice. Mice at 6 months of age [B6 (n = 10), KLS (n = 9), DBA (n = 6), and 129X1 (n = 6)] were subjected to glucose tolerance tests (B, F, J, and N) and insulin tolerance tests (C, G, K, and O). Fed glucose (A, E, I, and M) and fed and fasting insulin levels (D, H, L, and P) were also measured. To facilitate comparison, data from Fig. 1 are plotted along with 6-month data. Asterisks denote significantly different values (P < 0.05) as a result of mice aging from 2–6 months. At 6 months, fed insulin is greater than fasting insulin (*, P < 0.05) in the four strains of mice.

Insulin secreted during the acute phase after an ip glucose injection improved in KLS and DBA mice (3- to 4-fold), P < 0.001 (Fig. 8A), suggesting that the events regulating the secretory cascade in ß-cells improves with age in these mice (Table 2). Although at 2 min post ip glucose injection the level of insulin was higher in 6-month-old 129X1 mice than in 2-month-old 129X1 mice (Fig. 8A; P < 0.05), 6-month-old 129X1 mice continued to exhibit a minimal first-phase response; i.e. fasting insulin levels were also higher (Fig. 7P). Arginine-stimulated insulin secretion increased only in aging KLS mice (P = 0.03, Fig. 8B). Because the arginine initiated signaling path in ß-cells is different from the glucose initiated path (50, 51), then in male KLS mice both signaling paths improve with age. At 6 months, 129X1 mice do not exhibit a first-phase response to arginine because 2-min insulin levels post ip arginine injection were not significantly different (P < 0.05) from fasting insulin levels. That is, islet sensitivity to arginine decreased with age (Table 2).

View larger version (24K): [in this window] [in a new window]   FIG. 8. Acute-phase glucose- and arginine-stimulated insulin secretion in male and female mice at 2 and 6 months of age. Mice were fasted overnight and injected ip with glucose (A and C) or arginine (B and D) as described in Fig. 2. At 0, 2, and 5 min post stimulus injection, tail vein blood was collected. Insulin content of plasma was determined (ELISA). Insulin secretions followed time-courses similar to Fig. 2. Illustrated are peak, 2 min, insulin values (mean ± SEM). At 2 months of age, numbers of male and female mice used were as in Fig. 2. At 6 months of age, male mice tested were B6 (n = 9), KLS (n = 9), DBA (n = 6), and 129X1 (n = 6), and female mice tested were B6 (n = 9), KLS (n = 9), DBA (n = 6) and 129X1 (n = 5). Asterisks denote significant difference (P < 0.05) between 6- and 2-month values.

With aging, insulin tolerance in 129X1 mice is unaffected (Fig. 7O), suggesting unaffected muscle sensitivity to insulin (Table 2). Liver sensitivity to insulin, as previously discussed, is depressed in 6-month-old 129X1 mice (Table 2). Fed glucose (Fig. 7M) and fed insulin (Fig. 7P) were unaffected upon aging. To maintain constancy of fed glucose and fed insulin with aging, an increase in fat sensitivity to insulin is suggested in 129X1 mice (Table 2).

Glucose tolerance assays in 6-month-old 129X1 mice showed a significant increase in peak glucose concentrations (P = 0.03) 30 min post ip injection of 2 g/kg b.wt. glucose (Fig. 7N). It is likely that the depressed liver sensitivity to insulin in 6-month-old 129X1 mice (Table 2) would lead to higher glucose levels in glucose tolerance tests.

Female.
Insulin tolerance improved in KLS mice upon aging from 2–6 months; i.e. at 2 months steady-state levels decreased to approximately 70% of initial glucose levels, whereas at 6 months the decrease was down to approximately 30% (Fig. 9G; P < 0.001). These findings suggested that insulin sensitivity in muscles in KLS mice improved with age (Table 2). Although B6 mice appear to have improved insulin tolerance (Fig. 9C), the data were not significantly different (P = 0.11 and 0.09 at 30 and 60 min, respectively). DBA insulin tolerance was also age independent (Fig. 9K). Thus, age does not affect insulin sensitivity in muscles of B6 and DBA mice. The 129X1 mice, in contrast, are highly age dependent because their muscles appear to be insulin intolerant (Fig. 9O; Table 2).

View larger version (39K): [in this window] [in a new window]   FIG. 9. Effect of aging on glucose homeostasis in inbred strains of female mice. B6 (n = 10), KLS (n = 8), DBA (n = 6), and 129X1 (n = 5) female mice at 6 months of age underwent glucose tolerance tests (B, F, J, and N) and insulin tolerance tests (C, G, K, and O). Fed glucose (A, E, I, and M) and fed and fasting insulin levels (D, H, L, and P) were measured. To facilitate comparison, 2-month data (Fig. 6) are included. Significant differences (*, P < 0.05) between 2- and 6-month values are noted. At 6 months, fed insulin was significantly elevated (*, P < 0.05) over fasting insulin in B6 (D), DBA (L), and 129X1 (P) mice.

Fed glucose (Fig. 9, A, E, I, and M) decreased with age in each female mouse strain. The level of fed insulin rose in B6 mice upon aging (Fig. 9D), which may explain the drop in fed glucose. In KLS mice, lower fed glucose may be associated with improved muscle sensitivity to insulin. Interestingly, at both 2 and 6 months plasma insulin levels in KLS mice were independent of feeding state (Fig. 9H, P > 0.1). In DBA and 129X1 mice, there was a drop in fed glucose indicating an improvement in insulin sensitivity in fat tissues for DBA mice (Table 2) and in fat or liver tissues for 129X1 mice.

During 2 h post ip 2 mg/kg b.wt. glucose injection, 6-month-old female B6 mice were exposed to more glucose than were 2-month-old B6 mice (Fig. 9B, P < 0.05); i.e. B6 mice with aging are less glucose tolerant. Because acute-phase glucose-stimulated insulin secretions were similar in 2- and 6-month-old B6 female mice (Fig. 8A; P > 0.1), and because muscle sensitivity was not affected by age (Table 2), then impairment of glucose tolerance in aging B6 female mice reflects a loss of insulin sensitivity in liver and fat tissues (Table 2).

In glucose tolerance tests, tail vein blood glucose fell more slowly from its maximum in older KLS mice (Fig. 9F; P < 0.001). Because acute-phase glucose-stimulated insulin secretions were similar at 2 and 6 months (Fig. 8C; P > 0.05) and because muscle sensitivity improved with age (Table 2), the poorer glucose tolerance would suggest a loss of insulin sensitivity in fat and liver tissues of KLS mice (Table 2). In DBA mice, where insulin secreted during the first phase was higher (Fig. 9C; P < 0.05), glucose tolerance (Fig. 9J) was unchanged suggesting that the elevated insulin was necessary to compensate for lower liver sensitivity to insulin (Table 2). In 6-month-old 129X1 mice, where muscle appeared to be insensitive to insulin, an improvement of insulin sensitivity in fat tissue (Table 2) would explain the lack of an effect of age on glucose tolerance (Fig. 9O).

Peak levels of insulin in arginine-stimulated first-phase insulin secretions from 6-month-old B6 mice and DBA mice doubled approximately to those from 2-month-old mice (Fig. 8D; P < 0.05; Table 2). In KLS and 129X1 mice, however, age did not affect first phase arginine-stimulated insulin secretion (Fig. 8D; P > 0.1; Table 2).

Leptin.
Leptin and insulin interact physiologically at several levels. For example, leptin inhibits insulin secretion from islets (23), mice lacking leptin are hyperglycemic and insulin resistant, and mice lacking the long form of the leptin receptor are diabetic (2). We measured, therefore, leptin levels in the four inbred mouse strains at 6 months of age.

Leptin levels in the fasting state in male mice were not statistically different between strains (Fig. 10A; P > 0.05). With feeding, leptin levels rose in B6, KLS, and 129X1 mice (P < 0.05) but not in DBA mice (P = 0.32). In the fed state, B6 and 129X1 mice had similar circulating leptin levels (P > 0.1): 11250 ± 1250 and 11750 ± 1000 pg/ml, respectively, that were higher than those in KLS and DBA mice (P < 0.05). The highest fold increase of leptin upon feeding occurred in the 129X1 mouse.

View larger version (16K): [in this window] [in a new window]   FIG. 10. Plasma leptin in inbred strains of mice at 6 months of age. Tail vein blood samples were collected from fasted and fed inbred strains of male (a) and female (b) mice at 6 months of age. Plasma leptin content (ELISA) was determined. Male mice tested were B6 (n = 10), KLS (n = 9), DBA (n = 6), and 129X1 (n = 6). Female mice tested were B6 (n = 10), KLS (n = 8), DBA (n = 6), and 129X1 (n = 5). In male mice, fed leptin levels were lower in DBA mice than in KLS mice and both of these mice exhibited lower fed leptin levels than in B6 or 129X1 mice (*, P < 0.05). In female mice both fasting and fed leptin levels were significantly different (*, P < 0.05) between strains, except for fasting levels between KLS and 129X1 mice and fed levels between DBA and 129X1 mice. Except for male DBA mice, fed leptin levels were significantly elevated over fasting leptin levels (*, P < 0.05).

Summary of physiological assays
We have found that mouse strains have unique characteristics in how they maintain euglycemia under physiological and pharmacological conditions. Table 2 summarizes relative insulin sensitivity of muscle, fat, and liver tissues in the four inbred strains of mice in both male and female mice. We also suggest the relative ability of islets in the four strains to respond to a glucose and arginine stimulus. In male mice, glucose intolerance was associated with impaired insulin sensitivity in liver and fat, and the most intolerant strain was B6. Female mice are more glucose tolerant, but they use different physiological approaches to achieve improvement. For example, liver and fat tissues in B6, KLS, and DBA females respond better to insulin than they do in B6 males, whereas in 129X1 mice, it is their muscle sensitivity to insulin that improves. We have found that the effect of aging on glucose homeostasis is dependent on mouse strain and gender. In KLS male mice, first-phase insulin secretion in response to a glucose or arginine stimulus is improved with age but their liver sensitivity to insulin decreases. In DBA male mice, glucose-stimulated acute phase insulin secretion improved but fat sensitivity to insulin fell. Aging female B6 and KLS mice become less glucose tolerant, because insulin sensitivity in their liver and fat tissues decreased.

Gene expression
Sensitivity to a stimulus is dependent on the level of expression of signaling molecules in a tissue that responds to the stimulus (52). Because differences in insulin sensitivity between tissues and between strains were observed, we measured relative transcript content for the early insulin signaling molecules in insulin-responsive and nonresponsive tissues from each inbred strain.

Relative transcript content in responsive and nonresponsive tissues
Using real-time quantitative PCR (RT-PCR) in the Applied Biosystems 7700 instrument, RNA extracted from mouse tissues was analyzed for mRNA of insulin signaling components. Figure 11 illustrates relative mRNA expression for the IR, for the IGF-I receptor, and for each of the four IRS’s in tissues from 2-month-old male KLS mice. Every tissue, including tissues considered nonresponsive to insulin (kidney and spleen), expressed similar amounts of IR mRNA; i.e. relative expression values that ranged from 1000–2000 (Fig. 11A). These values are similar to the estimate of 1500 IR mRNA per rat liver cell, or 91.8 attomol IR mRNA per µg total RNA (53). The presence of IR mRNA in most tissues is consistent with autoradiography data where IR is detectable in both responsive and nonresponsive tissues (54).

View larger version (27K): [in this window] [in a new window]   FIG. 11. Relative mRNA expression for early insulin signaling molecules in tissues of male C57BLKS/6 mice. Taqman real-time PCR (ABI 7700) was used to measure mRNA for IR, IGF-I receptor, and each IRS in RNA extracts from muscle, fat, liver, brain, kidney, and spleen. Tissues were excised from 2-month-old male KLS mice. Details of RNA preparation, analyses, and quantitation are described in Materials and Methods. One to five tissue extracts (each from a different mouse) were analyzed. Mean ± SEM are illustrated. To detect IRS-4 mRNA expression, additional RNA was added to reaction wells.

Consistent with an earlier report (55), IRS-1 was expressed ubiquitously, and with two exceptions (muscle and spleen), levels of IRS-1 and IRS-2 mRNA within a tissue were similar (Fig. 11A; P > 0.1). Muscle contained about 3-fold more IRS-1 mRNA than IRS-2 mRNA (P < 0.001), and spleen contained 10 times more IRS-2 mRNA than IRS1 mRNA (P < 0.001). Generally, the relative expression of IRS-1 or IRS-2 mRNA was 5–10% of a tissue’s IR mRNA expression (Fig. 11A). Relative IRS-3 mRNA expression was equivalent to that of IRS-1 and IRS-2 mRNA expression in fat tissue (Fig. 11, B cf. A; P > 0.05), but in all other tissues IRS-3 mRNA expression levels were at 10% or less of the level expressed in fat (Fig. 11B). The relative amount of IRS-4 mRNA expression in brain was greater than its expression of IRS-3 mRNA (Fig. 11B; P < 0.05). In all other tissues of the KLS male mouse, IRS-4 mRNA expression was substantially less than their IRS-3 mRNA expression (Fig. 11B; P < 0.001). In addition to brain, fat tissue contained measurable amounts of IRS-4 mRNA (Fig. 11B). Previously, IRS-4 mRNA was detected in mouse liver (56) and kidney (55, 56).

Relative transcript content of insulin signaling intermediates in muscle, fat, and liver in male inbred strains of mice
In an attempt to correlate insulin sensitivity of tissues (Table 2) to their expression of insulin signaling genes, we measured mRNA levels for IR, IGF-I receptor, IRS-1, IRS-2, IRS-3, and IRS-4 in muscle, fat, and liver, from the four inbred male mice (Fig. 12). Surprisingly, the relative expression of any single gene was generally not strain dependent. In liver, the relative amount of mRNA for any of the insulin signaling molecules was not statistically different between strains (P > 0.05). Except in the comparison of IRS-2 mRNA expression between the 129X1 mouse and the DBA mouse (Fig. 12D; P < 0.05), the relative expression of mRNA for the insulin signaling molecules in muscle was not strain dependent (Fig. 12, A–D). In fat tissue, a statistical difference in IRS-1 mRNA expression between the KLS mouse and the DBA mouse (Fig. 12C; P < 0.05), in IRS-2 mRNA between the KLS mouse and the B6 mouse (Fig. 12D; P < 0.05), and in IRS-3 between the DBA and the 129X1 mouse (Fig. 12E; P < 0.05) was observed. In all other fat tissue mRNA measurements, there was no statistical difference in relative expression between strains (Fig. 12, A–E; P > 0.05).

View larger version (30K): [in this window] [in a new window]   FIG. 12. Relative mRNA expression of genes for insulin signaling molecules in insulin-responsive tissues from inbred strains of male mice. RNA extracts from muscle, fat, and liver from 2-month-old inbred strains of male mice were measured for mRNA content of: A, IR; B, IGF-I receptor; C, IRS-1; and D, IRS-2. Relative expression of IRS-3 mRNA in fat tissue is also illustrated (E). Details for these analyses are given in Materials and Methods and in legend to Fig. 11. *, Statistical differences between an expressed gene in a tissue in different strains are noted.

Fat tissue was the only tissue to express significant levels of IRS-3 mRNA (Figs. 11 and 12E). Previously, IRS-3 mRNA was reported in fat and kidney tissues of BALB/c or mice on a B6/129 mixed background (55, 56).

Transcript content of insulin signaling intermediates in islets of male inbred strains of mice
ß-Cells have proven to be insulin-responsive tissues because insulin regulate its own synthesis (57), and its presence is necessary for glucose to stimulate first-phase insulin secretion (58). Furthermore, because stimulated acute phase responses were strain dependent, we measured islet content of transcripts for the IR, the IGF-I receptor, and the four IRSs. Similar to fat, liver, and muscle, islet IR mRNA content in KLS, DBA, and 129X1 was higher than its content of the other transcripts (Fig. 13; P < 0.05). Among the inbred strains, the transcript content of IR in KLS and 129X1 mouse islets was greater than in B6 or DBA mouse islets (P < 0.01). KLS mouse islets contained twice as much IGF-I receptor mRNA as did islets from the other three strains (P < 0.01). The transcript content of IRS-1 or IRS-2 in islets, unlike IR and IGF-I receptor, were strain independent (P > 0.1). Islet content of mRNAs for IRS-3 and IRS-4 were similar to the levels seen in the KLS liver (Fig. 11) and were strain independent. Notably, relative mRNA expression of genes for insulin signaling intermediates in islets was 50% or less of those present in muscle, liver, or fat (Figs. 13 cf. 12; P < 0.001).

View larger version (26K): [in this window] [in a new window]   FIG. 13. Relative mRNA expression of genes for insulin signaling molecules in islets from inbred strains of male mice. RNA was extracted from islets isolated from 2-month-old male mice. mRNA for IR, IGF-I receptor, IRS-1, and IRS-2 were measured and quantitated as described in Materials and Methods. Illustrated are mean ± SEM of analyses of three to five islet RNA preparations. *, Significant differences (P < 0.05) in gene expression within a strain are noted.