In vivo glucoregulation and tissue-specific glucose uptake in female Akt substrate 160 kDa knockout rats

The Rab GTPase activating protein known as Akt substrate of 160 kDa (AS160 or TBC1D4) regulates insulin-stimulated glucose uptake in skeletal muscle, the heart, and white adipose tissue (WAT). A novel rat AS160-knockout (AS160-KO) was created with CRISPR/Cas9 technology. Because female AS160-KO versus wild type (WT) rats had not been previously evaluated, the primary objective of this study was to compare female AS160-KO rats with WT controls for multiple, important metabolism-related endpoints. Body mass and composition, physical activity, and energy expenditure were not different between genotypes. AS160-KO versus WT rats were glucose intolerant based on an oral glucose tolerance test (P<0.001) and insulin resistant based on a hyperinsulinemic-euglycemic clamp (HEC; P<0.001). Tissue glucose uptake during the HEC of female AS160-KO versus WT rats was: 1) significantly lower in epitrochlearis (P<0.05) and extensor digitorum longus (EDL; P<0.01) muscles of AS160-KO compared to WT rats; 2) not different in soleus, gastrocnemius or WAT; and 3) ~3-fold greater in the heart (P<0.05). GLUT4 protein content was reduced in AS160-KO versus WT rats in the epitrochlearis (P<0.05), EDL (P<0.05), gastrocnemius (P<0.05), soleus (P<0.05), WAT (P<0.05), and the heart (P<0.005). Insulin-stimulated glucose uptake by isolated epitrochlearis and soleus muscles was lower (P<0.001) in AS160-KO versus WT rats. Akt phosphorylation of insulin-stimulated tissues was not different between the genotypes. A secondary objective was to probe processes that might account for the genotype-related increase in myocardial glucose uptake, including glucose transporter protein abundance (GLUT1, GLUT4, GLUT8, SGLT1), hexokinase II protein abundance, and stimulation of the AMP-activated protein kinase (AMPK) pathway. None of these parameters differed between genotypes. Metabolic phenotyping in the current study revealed AS160 deficiency produced a profound glucoregulatory phenotype in female AS160-KO rats that was strikingly similar to the results previously reported in male AS160-KO rats.


Abstract
The Rab GTPase activating protein known as Akt substrate of 160 kDa (AS160 or TBC1D4) 24 regulates insulin-stimulated glucose uptake in skeletal muscle, the heart, and white adipose 25 tissue (WAT). A novel rat AS160-knockout (AS160-KO) was created with CRISPR/Cas9 26 technology. Because female AS160-KO versus wild type (WT) rats had not been previously 27 evaluated, the primary objective of this study was to compare female AS160-KO rats with WT 28 controls for multiple, important metabolism-related endpoints. Body mass and composition, 29 physical activity, and energy expenditure were not different between genotypes. AS160-KO 30 versus WT rats were glucose intolerant based on an oral glucose tolerance test (P<0.001) and 31 insulin resistant based on a hyperinsulinemic-euglycemic clamp (HEC; P<0.001). Tissue 32 glucose uptake during the HEC of female AS160-KO versus WT rats was: 1) significantly lower 33 in epitrochlearis (P<0.05) and extensor digitorum longus (EDL; P<0.01) muscles of AS160-KO 34 compared to WT rats; 2) not different in soleus, gastrocnemius or WAT; and 3) ~3-fold greater 35 in the heart (P<0.05). GLUT4 protein content was reduced in AS160-KO versus WT rats in the 36 epitrochlearis (P<0.05), EDL (P<0.05), gastrocnemius (P<0.05), soleus (P<0.05), WAT 37 (P<0.05), and the heart (P<0.005). Insulin-stimulated glucose uptake by isolated epitrochlearis 38 and soleus muscles was lower (P<0.001) in AS160-KO versus WT rats. Akt phosphorylation of 39 insulin-stimulated tissues was not different between the genotypes. A secondary objective was 40 to probe processes that might account for the genotype-related increase in myocardial glucose 41 uptake, including glucose transporter protein abundance (GLUT1, GLUT4, GLUT8, SGLT1), 42 hexokinase II protein abundance, and stimulation of the AMP-activated protein kinase (AMPK) 43 pathway. None of these parameters differed between genotypes. Metabolic phenotyping in the 44 current study revealed AS160 deficiency produced a profound glucoregulatory phenotype in 45 female AS160-KO rats that was strikingly similar to the results previously reported in male 46 AS160-KO rats. Key Words: insulin resistance; glucose transport; myocardial glucose uptake; sex differences Introduction 49 The Rab GTPase activating protein known as Akt substrate of 160 kDa (also known as 50 AS160 or TBC1D4) is highly expressed by multiple tissues, including skeletal muscle, the heart, 51 and white adipose tissue (WAT) [1][2][3][4][5]. These tissues are important sites for insulin-mediated 52 glucose disposal, and phosphosite-specific phosphorylation of AS160 by Akt is crucial for 53 insulin-stimulated GLUT4 glucose transporter exocytosis and enhanced glucose transport. 54 Accordingly, understanding the relationship between AS160 and glucose uptake in these 55 tissues has implications for whole body glucoregulation and insulin sensitivity.

56
AS160 deficiency in humans [6], mice [3,4] and rats [5] results in whole body insulin 57 resistance, but there is limited knowledge about the effects of AS160 deficiency in females, 58 regardless of species. Published research in humans has not addressed the possibility that 59 AS160 deficiency might not have identical consequences on males versus females [6]. Several 60 studies using AS160-KO mice reported data only [4,7] or mostly [8] in males. Other studies 61 reported data for both male and female AS160-KO mice for some, but not all outcomes [3,9]. 62 Hyperinsulinemic-euglycemic clamps (HEC) have been performed only in male AS160-KO mice 63 [4], and in vivo tissue-specific glucose uptake has been reported in male, but not female, mice 64 [3,9] Optics), and body mass was assessed in rats at 7-8 weeks-old (n = 6 for each genotype).

187
Another cohort of rats was anesthetized (intraperitoneal injection of ketamine/xylazine cocktail, 188 50 mg/kg ketamine and 5 mg/kg xylazine) at 11 weeks-old, and body mass was measured in 189 these rats. Skeletal muscles (extensor digitorum longus, EDL,epitrochlearis,gastrocnemius,190 and soleus), WAT, and the heart from these animals were sampled and weighed.

191
Indirect calorimetry, physical activity and food intake 192 Indirect calorimetry and physical activity were analyzed in rats (n = 6 for each genotype)

Plasma insulin and non-esterified fatty acids (NEFA)
233 Plasma was sampled at -10 and 120 minutes during the HEC and used to measure plasma 234 insulin via ELISA. Plasma samples from -10, 80, 90 and 120 minutes were used to measure NEFA 235 levels with a colorimetric assay.

236
Isolated skeletal muscle procedures 237 At 1600 to 1700 on the night preceding the isolated skeletal muscle procedures, food 238 was removed from the cages of the rats (aged 9 weeks). Rats (n = 8-9) used for insulin 239 treatment experiments were anesthetized with intraperitoneal injections of SP (50 mg/kg), and 240 rats used for the 5-aminoimidazole-4-carboxamide-1-β-D-ribofuranoside (AICAR) experiment 241 were anesthetized with a ketamine (50 mg/kg)/xylazine (5 mg/kg) cocktail (K/X). In preliminary 242 experiments, we found no significant difference for glucose uptake by isolated muscles isolated 243 from animals anesthetized with SP compared to K/X (not shown Data were statistically analyzed with SigmaPlot 13.0 (San Jose, CA). Comparisons 296 between the WT and AS16O-KO rats were performed with an unpaired two-tailed t-test. 297 Comparisons between contralateral muscles that were incubated ex vivo (either ±insulin or 298 ±AICAR) were analyzed using a paired two-tailed t-test. Two-way ANOVA was performed to 299 compare more than two groups (ex vivo muscles incubated ±insulin or ±AICAR), and Holm-300 Sidak post hoc tests were used to identify the source of significant variance. Data were 301 expressed as means ± SEM. P-values ≤0.05 were considered statistically significant.

303
Genotype confirmation 304 Genotype was determined using qPCR with DNA from tail tip samples. Lack of AS160 in 305 tissues from AS160-KO rats was also substantiated via immunoblotting for AS160 protein after 306 the animals were euthanized (Figure 1). Oral glucose tolerance test (OGTT) 421 Glucose levels at baseline were slightly (12.6%), but significantly (P<0.05) greater for 422 WT compared to AS160-KO females (Fig 2A) Hyperinsulinemic-euglycemic clamp (HEC) 436 Although 6 WT and 6 AS160-KO rats underwent an HEC, 2 rats from each genotype 437 were not successfully clamped because of technical problems. A catheter placed in one AS160-438 KO rat failed during the HEC, and the coefficient of variation (CV) for glycemia exceeded 20% 439 for three rats (WT, n=2; AS160-KO n=1). Accordingly, the data from these 4 rats were not 440 included in the statistical analysis of parameters dependent on blood glucose or insulin (Table   441 4), tissue glucose uptake and insulin signaling. The CV for the remaining 4 WT rats (3-14%) 442 was comparable to the CV for the remaining 4 AS160-KO rats (5-12%), and the reported values 443 for blood or plasma parameters, tissue glucose uptake, and phosphorylated signaling proteins 444 from the HEC experiment represent the data collected from these 8 animals. No significant 445 differences between AS160-KO compared to WT rats was found for blood glucose values, plasma insulin values, glucose turnover rate (GTR) or hepatic glucose production (HGP;  were not detectable in tissues analyzed from AS160-KO rats (Fig 1). Among all of the tissues 498 evaluated, there were no significant genotype-related differences for either pAkt Ser 473 or pAkt 499 Thr 308 (Fig 4). In addition, genotype-related differences were not found for TBC1D1, an AS160 500 paralog (Fig 5). GLUT4 glucose transporter protein abundance was determined in each of the tissues 514 after HEC. GLUT4 protein levels were much less in AS160-KO versus WT rats in the 515 gastrocnemius, EDL, epitrochlearis, soleus, heart and WAT ( Figure 6). GLUT1 glucose 516 transporter protein content was not different between genotypes in any of these tissues 517 (Supplemental Fig 1). In addition, hexokinase II abundance was not different between 518 genotypes for any of these tissues (Supplemental Fig 2). Because Akt2 is required for insulin-stimulated glucose uptake by the heart [33], we 525 determined Ser 474 phosphorylation (pAkt2 Ser474 ) in the heart from the HEC rats and found no 526 difference between the AS160-KO and WT rats (Fig 7). Greater heart sodium-dependent 527 glucose cotransporter 1 (SGLT1) expression has been found in diabetic animals concomitant 528 with attenuated GLUT1 and GLUT4 levels [34], but we observed SGLT1 protein content did not 529 differ between groups (Fig 7). Calcium ATPase (SERCA) has been linked to greater glucose 530 uptake by the heart [35]. The abundance of myocardial SERCA2 was not different between 531 genotypes (Fig 7). It was possible that the increased glucose uptake in AS160-KO heart was 532 related to the reduced metabolism of another energy source, e.g., fatty acids. Therefore, we 533 analyzed the abundance of the CD36 fatty acid translocase protein that is important for myocardial fatty acid uptake [36]. No significant genotype-related difference was observed for 535 CD36 levels in the heart (Fig 7). Activation of AMPK can lead to increased glucose uptake by the heart [37]. We 542 determined the phosphorylation of AMPK at Thr172 (pAMPK Thr 172 ), which is a major 543 mechanism for increasing the enzyme's activity [38]. We also assessed the phosphorylation of 544 its substrate acetyl CoA-carboxylase (pACC Ser 79 ) that is often used as an indicator for AMPK 545 activity. However, no difference was detected between WT and AS160-KO for either pAMPK 546 Thr 172 or pACC Ser 79 in the heart (Fig 8). Furthermore, phosphorylation of TBC1D1 on an 547 AMPK-phosphosite, Ser237 (pTBC1D1 Ser 237 ), was also undistinguishable in the heart of WT 548 compared to AS160-KO rats (Fig 8). We also assessed the abundance of GLUT8 glucose transporter protein, which is known 555 to be expressed by the heart [39]. However, no genotype-related difference was found for 556 myocardial GLUT8 abundance (Fig 9). Because major metabolic fates of glucose include 557 conversion to lactate or mitochondrial oxidation, we also determined the abundance of LDH and 558 multiple components of the electron transport chain and oxidative phosphorylation (NDUFB8, 560 detected for LDH (Fig 9) or any of the mitochondrial proteins that were studied (Supplemental 561 Fig 3). Soleus and epitrochlearis muscles were studied ex vivo with and without an insulin dose 569 (500 µU/ml) that corresponded to the plasma insulin values that the animals were exposed to 570 during the HEC. Glucose uptake by insulin-stimulated soleus (P<0.001) and epitrochlearis 571 (P<0.001) muscles isolated from WT rats was much greater than values in AS160-KO rats (Fig   572  8). Glucose uptake by the epitrochlearis from AS160-KO animals was greater for insulin-573 stimulated compared to paired muscles without insulin (P<0.001). Glucose uptake with insulin 574 was increased compared to without insulin in the soleus (P<0.05) and the epitrochlearis 575 (P=0.05) from AS160-KO rats. Glucose uptake in the epitrochlearis without insulin was 576 significantly greater (P<0.05) for WT versus AS160-KO rats (Fig 10). 577 578 Figure 10. greater with insulin versus without insulin (Fig 11). pAS160 Thr 642 was undetectable in either 590 epitrochlearis or soleus muscles from AS160-KO rats (Figure 11 A-B). With insulin treatment, a 591 substantial increase was found for the phosphorylation of Akt Ser 473 and Thr 308 (Fig 11 C-F  The greatest genotype-difference in the female rats was the markedly greater 662 myocardial glucose uptake rate in the AS160-KO compared to WT rats. We previously reported 663 an almost identical relative increase in heart glucose uptake in male AS160-KO versus WT rats 664 [5]. In both sexes, the difference in myocardial glucose uptake occurred in spite of much lower 665 GLUT4 protein abundance in the AS160-KO compared to WT rats. In neither sex was the 666 greater glucose uptake attributable to altered myocardial abundance of GLUT1, HKII or SGLT1.

667
Furthermore, in the females, we also assessed heart GLUT8, and found that the abundance of 668 this glucose transporter was also unaltered by AS160 deficiency. Taking together all of these 669 results, there is no evidence that the genotype effect on myocardial glucose uptake is linked to 670 greater expression of glucose transporter proteins. Our working hypothesis is that the result is 671 attributable to altered subcellular localization of glucose transporter proteins, with greater cell 672 surface glucose transporter content in the AS160-KO compared to WT rats of either sex.

673
GLUT4 is the most abundant glucose transporter protein expressed by the heart, and GLUT4 674 translocation is normally the dominant mechanism for increasing myocardial glucose uptake 675 [40]. Therefore, even though AS160 deficiency was characterized by a substantial decline in 676 total GLUT4 abundance in the heart, increased GLUT4 translocation may be hypothesized to 677 play a role in the observed 3-fold increase in myocardial glucose uptake.

678
AMP-activation has been reported to increase myocardial GLUT4 translocation and 679 glucose uptake [41]. Accordingly, we evaluated the activation of AMPK phosphorylation on 680 Thr172 because increased AMPK phosphorylation on this site results in a substantial increase 681 in AMPK activity [38]. We also assessed the phosphorylation of AMPK's substrate acetyl CoA-682 carboxylase (ACC) which is often used as a surrogate indicator of AMPK activation [42,43]. In 683 addition, we determined phosphorylation of the Rab-GTPase activating protein TBC1D1 on 684 Ser231, an AMPK-phosphomotif that has been implicated in TBC1D1's regulation of glucose 685 uptake [44]. However, there were no differences between genotypes for phosphorylation of