Insulin-stimulated glucose uptake partly relies on p21-activated kinase 2 (PAK2), but not PAK1, in mouse skeletal muscle

Abstract

The group I p21-activated kinase (PAK) isoforms PAK1 and PAK2 are activated in response to insulin in skeletal muscle and PAK1/2 signalling is impaired in insulin-resistant mouse and human skeletal muscle. Interestingly, PAK1 has been suggested to be required for insulin-stimulated GLUT4 translocation in mouse skeletal muscle. Therefore, the present investigation aimed to further examine the role of PAK1 in insulin-stimulated muscle glucose uptake. Pharmacological inhibition of group I PAKs, IPA-3 partially reduced (−20%) insulin-stimulated glucose uptake in isolated mouse soleus muscle (p<0.001). However, since there was no phenotype with genetic ablation of PAK1 alone, consequently, the relative requirement for PAK1 and PAK2 in whole-body glucose homeostasis and insulin-stimulated muscle glucose uptake was investigated. Whole-body respiratory exchange ratio was largely unaffected in whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO and double whole-body PAK1 and muscle-specific PAK2 knockout mice. In contrast, glucose tolerance was mildly impaired in mice lacking PAK2 specifically in muscle, but not PAK1 KO mice. Moreover, while PAK1 KO muscles displayed normal insulin-stimulated glucose uptake in vivo and in isolated muscle, insulin-stimulated glucose uptake was slightly reduced in isolated glycolytic extensor digitorum longus muscle lacking PAK2 alone (−18%) or in combination with PAK1 KO (−12%) (p<0.05).In conclusion, glucose tolerance and insulin-stimulated glucose uptake partly rely on PAK2 in glycolytic mouse muscle, while PAK1 is dispensable for whole-body glucose homeostasis and insulin-stimulated muscle glucose uptake.

1. Introduction

Skeletal muscles account for the majority of insulin-mediated whole-body glucose disposal (DeFronzo et al., 1985; Baron et al., 1988) and muscle insulin resistance is an early defect in the pathophysiology of peripheral insulin resistance and type 2 diabetes mellitus (DeFronzo et al., 1985; Kahn, 2003). As diabetes globally is approaching epidemic proportions, it is important to understand the mechanisms regulating glucose uptake by skeletal muscle.

Insulin stimulates glucose uptake in skeletal muscle by activation of a signalling cascade that leads to the translocation of glucose transporter (GLUT)4-containing vesicles to the sarcolemma and transverse tubuli as well as increasing microvascular blood flow (Marette et al., 1992; Ploug et al., 1998; Lauritzen, 2013; Sjoberg et al., 2017). This signalling cascade has been proposed to include activation of p21-activated kinase 1 (PAK1) downstream of PI3K (Tsakiridis et al., 1996; Wang et al., 2011; Tunduguru et al., 2014). PAKs are serine/threonine kinases and involved in numerous signalling networks regulating essential cellular activities, including cell proliferation, differentiation, apoptosis, and cytoskeleton dynamics (Bagrodia & Cerione, 1999; Bokoch, 2003; Hofmann et al., 2004; Chiang & Jin, 2014). Group I PAKs (PAK1-3) are downstream targets of the Rho GTPases Cdc42 and Rac1 (Manser et al., 1994). Previous studies suggest that only PAK1 and PAK2 are expressed in skeletal muscle, whereas PAK3 mRNA and protein expression is below the detection limit (Arias-Romero & Chernoff, 2008; Tunduguru et al., 2014; Joseph et al., 2017). In muscle cells and mouse skeletal muscle, PAK1 is proposed to be required for GLUT4 translocation in response to insulin stimulation (Wang et al., 2011; Tunduguru et al., 2014), downstream of Rac1 (Khayat et al., 2000; JeBailey et al., 2007; Ueda et al., 2010; Sylow et al., 2013a, 2014; Raun et al., 2018). Thus, together with Akt which is proposed to regulate GLUT4 translocation via phosphorylation of the Rab GTPase-activating protein TBC1D4 (Sano et al., 2003; Thong et al., 2005; Kramer et al., 2006), Rac1-PAK1 activation is suggested to be necessary for insulin-stimulated GLUT4 translocation.

Upon activation of PAK1 and PAK2, conformational changes allow autophosphorylation of T423 and T402, respectively, thereby relieving the autoinhibition of PAK1 and PAK2 (Manser et al., 1994; Benner et al., 1995; Gatti et al., 1999). In vastus lateralis muscle from subjects with obesity and type 2 diabetes, phosphorylation of PAK1/2 at T423/402 in response to insulin was 50% reduced compared to healthy controls (Sylow et al., 2013a). Likewise, insulin-stimulated pPAK1/2 T423/402 was diminished in palmitate-treated insulin-resistant L6 myotubes, even though upstream of PAK1/2, insulin-stimulated Rac1-GTP binding (i.e. activation) was not impaired (Stierwalt et al., 2018). Together, these studies (Sylow et al., 2013a; Stierwalt et al., 2018) associate dysregulated activity of PAK1 and PAK2 with insulin resistance. In addition, a pharmacological inhibitor of group I PAKs, IPA-3 abolished insulin-stimulated GLUT4 translocation and glucose uptake in L6 myoblasts and myotubes overexpressing myc-tagged GLUT4 (L6-GLUT4myc), respectively (Tunduguru et al., 2014). This indicates that group I PAKs are required for insulin-stimulated glucose uptake. The effect of IPA-3 has largely been ascribed to inhibition of PAK1, as whole-body genetic ablation of PAK1 in mice impaired glucose tolerance (Wang et al., 2011; Chiang et al., 2013) and blocked insulin-stimulated GLUT4 translocation in skeletal muscle (Wang et al., 2011). Further supporting PAK1 being the major group I PAK isoform regulating GLUT4 translocation, insulin-stimulated GLUT4 translocation was unaffected by a 75% knockdown of PAK2 in L6-GLUT4myc myoblasts (Tunduguru et al., 2014). The suggested downstream mechanisms whereby PAK1 regulates GLUT4 translocation include simultaneous cofilin-mediated actin depolymerization and N-WASP-cortactin-mediated actin polymerization (Chiu et al., 2010; Tunduguru et al., 2014, 2017).

The present investigation aimed to further examine the role of PAK1 in insulin-stimulated glucose uptake in mature skeletal muscle. Therefore, we performed a systematic series of pharmacologic and genetic experiments to analyze the involvement of PAK1 in the regulation of insulin-stimulated glucose uptake in mouse skeletal muscle. We hypothesized that PAK1 would be necessary for glucose uptake in response to insulin. Contradicting our hypothesis, our results revealed that insulin-stimulated glucose uptake was unaffected by genetic ablation of PAK1 alone. Consequently, using transgenic double PAK1/2 knockout mice, the relative requirement for PAK1 and PAK2 in whole-body glucose homeostasis and insulin-stimulated glucose uptake in skeletal muscle was investigated. Our results revealed that insulin-stimulated glucose uptake partly relies on PAK2 in glycolytic mouse muscle, while PAK1 is dispensable for whole-body glucose homeostasis and insulin-stimulated muscle glucose uptake.

2. Methods

2.1. Ethical Approval

All animal experiments complied with the European Convention for the protection of vertebrate animals used for experimental and other scientific purposes (No. 123, Strasbourg, France, 1985; EU Directive 2010/63/EU for animal experiments) and were approved by the Danish Animal Experimental Inspectorate (License number: 2015-15-0201-00477). The experiments conformed to the principles and regulations as described in the Editorial by Grundy (2015).

2.2. Animals

All mice were maintained on a 12:12-hour light-dark cycle and housed at 22°C (with allowed fluctuation of ±2°C) with nesting material. The mice were group-housed. Female C57BL/6J mice (Taconic, Denmark) were used for the inhibitor incubation study. The mice received a standard rodent chow diet (Altromin no. 1324; Brogaarden, Denmark) and water ad libitum.

2.2.1. Whole-body PAK1^−/− mice

For a complete overview of the different cohorts of genetically modified mice including the number of mice in each group and the age at the experimental interventions, see Table 1. Whole-body PAK1^−/− mice on a C57BL/6J background were generated as previously described (Allen et al., 2009). The mice were obtained by heterozygous crossing. PAK1^−/− mice (referred to as PAK1 KO) and age-matched paired littermate PAK1^+/+ mice (referred to as controls) were used for experiments. One cohort of female and male mice was used for measurements of body composition, glucose tolerance and insulin-stimulated glucose uptake in isolated muscle. The mice received a standard rodent chow diet and water ad libitum. For measurement of in vivo insulin-stimulated glucose uptake in chow- and 60E% high-fat diet (HFD; no. D12492; Brogaarden, Denmark)-fed PAK1 KO mice, a distinct cohort of mice were assigned to a chow or HFD group. Chow-fed mice were used for measurement of in vivo insulin-stimulated glucose uptake. HFD-fed mice received the diet for 21 weeks and were used for body composition, glucose tolerance and in vivo insulin-stimulated glucose uptake. Energy intake was measured over a period of 10 weeks in a distinct cohort of mice. Number of mice in each group: Chow, n = 8/8 (Control/PAK1 KO); HFD, n = 8/8. Mice had access to their respective diet and water ad libitum.

Table 1:

Overview of the different cohorts of genetically modified mice.

			Age (weeks)
Cohort	Diet	Number of replicates	Start diet intervention	ITT	GTT	Body composition	Home cage calorimetry	Tissue dissection	Terminal experiment
Whole-body PAK1−/− mice	Chow	Control, n = 6/7 (female/male) PAK1 KO, n = 4/8			8–20	7–19		12–24	Insulin-stimulated glucose transport in isolated muscle
Whole-body PAK1−/− mice	Chow	Control, n = 14/8 PAK1 KO, n = 6/4						10–24	In vivo glucose uptake
Whole-body PAK1−/− mice	60E% HFD	Control, n = 7/7 PAK1 KO, n = 11/5	6–16		20–30	24–34		27–37	In vivo glucose uptake
Double PAK1−/−;PAK2fl/fl;MyoDiCre/+ mice	Chow	Control, n = 6/4 PAK1 KO, n = 5/4 PAK2 mKO, n = 6/4 1/m2 dKO, n = 6/3						10–16	Insulin-stimulated glucose transport in isolated muscle
Double PAK1−/−;PAK2fl/fl;MyoDiCre/+ mice	Chow	Control, n = 9/11 PAK1 KO, n = 8/10 PAK2 mKO, n = 12/9 1/m2 dKO, n = 9/14.		11–24	13–26	16–29	23–33

2.2.2. Double PAK1^−/−;PAK2^fl/fl;MyoD^iCre/+ mice

Double knockout mice with whole-body knockout of PAK1 and conditional, muscle-specific knockout of PAK2 (PAK1^−/−;PAK2^fl/fl;MyoD^iCre/+) were generated as previously described (Joseph et al., 2017). The mice were on a mixed C57BL/6/FVB background. PAK1^−/−;PAK2^fl/fl;MyoD^iCre/+ were crossed with PAK1^+/−;PAK2^fl/fl;MyoD^+/+ to generate littermate PAK1^−/−;PAK2^fl/fl;MyoD^iCre/+ (referred to as 1/m2 dKO), PAK1^−/−;PAK2^fl/fl;MyoD^+/+ (referred to as PAK1 KO), PAK1^+/−;PAK2^fl/fl;MyoD^iCre/+ (referred to as PAK2 mKO), and PAK1^+/−;PAK2^fl/fl;MyoD^+/+ (referred to as controls) female and male mice used for experiments. Age-matched littermate mice were used for experiments. One cohort of female and male mice was used for measurement of insulin-stimulated glucose uptake in isolated muscle. A distinct cohort of mice was used for whole-body metabolic measurements. In the latter cohort, characterization of contraction-stimulated muscle glucose uptake has previously been described (Møller et al., 2020). For some of the metabolic measurements, only a subgroup of mice was used as indicated in the relevant figure legends. Mice received standard rodent chow diet and water ad libitum.

2.2.3. Mouse genotyping by PCR

Mouse genotyping was performed as previously described (Møller et al., 2020). In short, an ear punch was digested overnight in 100 μL Viagen lysis buffer plus Proteinase K at 55°C followed by 45 min at 85°C. After spin at 1000 × g for 5 min, the supernatant was diluted 10 times in TE (pH 8.0) with yellow colour (50 pg mL⁻¹ Quinoline Yellow). 5 μL of this was used in a 25 μL real-time quantitative PCR reaction containing Quantitect SYBR Green Master Mix (Qiagen), 200 nM of each primer (Table 2) and blue dye (5 pg mL⁻¹ Xylene Cyanol). The reactions were furthermore spiked (100 times less than the samples) with a heterozygote sample as a positive PCR control. The samples, including no sample controls (TE), were amplified in an MX3005P real-time PCR machine (95°C, 10 min $\to$ {95°C, 15 sec → 58°C, 30 sec → 63°C, 90 sec} × 50 → melting curve 55°C → 95°C). The Ct values were used to access allele presence by comparison to the no DNA controls (spike values) such that the Ct value should be at least 2 Ct below the no sample controls to indicate the presence of the allele. Amplification efficiency in the individual reactions was estimated by the sigmoid method of Liu and Saint (Liu & Saint, 2002) to ensure that the Ct’s could be compared within primer sets. The genotype was later verified by immunoblotting on samples from muscle tissue.

Table 2:

Primers sequences used for mouse genotyping.

Gene target	Forward primer (5’−3’)	Reverse primer (5’−3’)
Pak1 wildtype	CCCCCGCAGCAAATAAAAAGA	CCCTGTGACAGCATCAAAACCA
Pak1 floxed	CCCCCGCAGCAAATAAAAAGA	GGAAAAGCGCCTCCCCTACC
Pak2 wildtype	GAATGAAGCCCGAGTTCAAGTCCC	CTGCATCAATCTATTCTGACTATGACAGGT
Pak2 floxed	TGCAGGTGCAGTGTGACAGAGA	TGAGCGGATCCACCTAATAACTTCGT
MyoD wildtype	GCTCAGGAGGATGAGCAATGGA	ATAAGGGACACCCCCACCCCAAG
MyoD iCre	GGATCCGAATTCGAAGTTCCTATTCTCT	CCAAGGGCCTCGGAAACCTG

2.3. Body composition

Body composition was analyzed using magnetic resonance imaging (EchoMRI-4in1TM, Echo Medical System LLC, Texas, USA).

2.4. Glucose tolerance test (GTT)

Prior to the glucose tolerance test (GTT), chow- and HFD-fed PAK1 KO mice and control littermates fasted for 12 hours from 10 p.m. In a distinct GTT, chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice and control littermates fasted for 6 hours from 6 a.m. D-mono-glucose (2 g kg⁻¹ body weight) was administered intraperitoneal (i.p) and blood was collected from the tail vein and blood glucose concentration determined at the indicated time points using a glucometer (Bayer Contour, Bayer, Switzerland). Incremental Area Under the Curve (AUC) from the basal blood glucose concentration was determined using the trapezoid rule. For measurement of plasma insulin, glucose was administered i.p. on a separate experimental day (1–2 weeks after the GTT) and blood was sampled at time points 0 and 20 minutes, centrifuged and plasma was quickly frozen in liquid nitrogen and stored at −20°C until processing. Plasma insulin was analyzed in duplicate (Mouse Ultrasensitive Insulin ELISA, #80-INSTRU-E10, ALPCO Diagnostics, USA). Homeostatic model assessment of insulin resistance (HOMA-IR) was calculated according to the formula: Fasting plasma insulin (mU L⁻¹) X Fasting blood glucose (mM)/22.5.

2.5. Insulin tolerance test (ITT)

Prior to the insulin tolerance test, chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice and control littermates fasted for 4 hours from 6 a.m. Insulin (0.5 U kg⁻¹ body weight) was administered i.p. and blood was collected from the tail vein and blood glucose concentration determined using a glucometer (Bayer Contour, Bayer, Switzerland) at time point 0, 15, 30, 60, 90 and 120 minutes. For two female control mice and four female PAK2 mKO mice, the ITT had to be stopped before the 120’-time point due to hypoglycemia (blood glucose <1.2 mM). Thus, blood glucose was not measured in these mice for the last couple of time points.

2.6. Home cage indirect calorimetry

One week prior to the calorimetric measurements, chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice and control littermates were single-housed in specialized cages for indirect gas calorimetry but uncoupled from the PhenoMaster indirect calorimetry system (TSE PhenoMaster metabolic cage systems, TSE Systems, Germany). After a 2-day acclimation period coupled to the PhenoMaster indirect calorimetry system, oxygen consumption, CO₂ production, habitual activity (beam breaks) and food intake were measured for 72 hours (TSE LabMaster V5.5.3, TSE Systems, Germany). On day 2, mice fasted during the dark period followed by refeeding on day 3. Respiratory exchange ratio (RER) was calculated as the ratio between CO₂ production and oxygen consumption.

2.7. Incubation of isolated muscles

Soleus and extensor digitorum longus (EDL) muscles were dissected from anaesthetized mice (6 mg pentobarbital sodium 100 g⁻¹ body weight i.p.) and suspended at resting tension (4–5 mN) in incubations chambers (Multi Myograph System, Danish Myo Technology, Denmark) in Krebs-Ringer-Henseleit buffer with 2 mM pyruvate and 8 mM mannitol at 30°C, as described previously (Jørgensen et al., 2004). Additionally, the Krebs-Ringer-Henseleit buffer was supplemented with 0.1% BSA (v/v). The mice were euthanized by cervical dislocation after muscle isolation. Isolated muscles from female C57BL/6J mice were pre-incubated with 40 μM IPA-3 (Sigma-Aldrich) or as a control DMSO (0.25%) for 25 minutes followed by 30 minutes of insulin stimulation (60 nM; Actrapid, Novo Nordisk, Denmark). Isolated muscles from chow-fed PAK1 KO were pre-incubated for 30 minutes followed by 30 minutes of insulin stimulation (0.6 nm or 60 nM). Isolated muscles from chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice or control littermates were pre-incubated for 20 minutes followed by 20 minutes of insulin stimulation (60 nM). 2-deoxyglucose (2DG) uptake was measured together with 1 mM 2DG during the last 10 min of the insulin stimulation period using 0.60–0.75 μCi mL⁻¹ [³H]-2DG and 0.180–0.225 μCi mL⁻¹ [¹⁴C]-mannitol radioactive tracers essentially as described in (Jørgensen et al., 2004). Muscle-specific [³H]-2DG accumulation was measured in the lysate with [¹⁴C]-mannitol as an extracellular marker and related to the specific activity of the incubation buffer.

2.8. In vivo insulin-stimulated 2-deoxyglucose uptake in PAK1 KO mice during a r.o. ITT

To determine 2DG uptake in skeletal muscle of PAK1 KO mice and littermate controls, [³H]-2DG (Perkin Elmer) was administered retro-orbitally (r.o.) in a bolus of saline containing 66.7 μCi mL⁻¹ [³H]-2DG ($~$ 32.4 Ci/mmol) corresponding to ∼10 μCi/mouse in chow-fed mice or ∼15 μCi/mouse in HFD-fed mice (6 μL g⁻¹ body weight), as previously described (Raun et al., 2018). The [³H]-2DG saline bolus was with or without insulin (Actrapid, Novo Nordisk, Denmark). Decreased insulin clearance has been observed by us (Raun et al., 2018) and others in obese rodent (Strömblad & Björntorp, 1986; Brandimarti et al., 2013) and human (Jung et al., 2018) models. Therefore, to correct for changes in insulin clearance, 0.5 U kg⁻¹ body weight of insulin was administered in chow-fed mice whereas only 60% of this dosage was administered to HFD-fed mice. Prior to stimulation, mice fasted for 4 hours from 07:00 and were anaesthetized (7.5/9 mg [Chow/HFD] pentobarbital sodium 100 g⁻¹ body weight i.p.) 15 minutes before the r.o. injection. Blood samples were collected from the tail vein after the r.o. injection and analyzed for glucose concentration using a glucometer (Bayer Contour, Bayer, Switzerland) at the time points 0, 5 and 10 minutes. After 10 minutes, skeletal muscle (gastrocnemius, quadriceps, triceps brachii and soleus) were excised after cervical dislocation of the mouse. The muscles were rinsed in saline, and quickly frozen in liquid nitrogen and stored at −80°C until processing. Blood was collected by punctuation of the heart, centrifuged and plasma was quickly frozen in liquid nitrogen and stored at −80°C until processing. Plasma samples were analyzed for insulin concentration and specific [³H]-2DG activity. Plasma insulin was analyzed in duplicate (Mouse Ultrasensitive Insulin ELISA, #80-INSTRU-E10, ALPCO Diagnostics, USA). Tissue-specific 2DG-6-phosphate accumulation was measured as described (Ferre et al., 1985; Fueger et al., 2004). To determine 2DG clearance from the plasma into the specific tissue, tissue-specific [³H]-2DG-6-P was divided by AUC of the plasma-specific [³H]-2DG activity at the time points 0 and 10 minutes. To estimate tissue-specific glucose uptake (glucose uptake index), clearance was multiplied by the average blood glucose levels for the time points 0, 5, and 10 minutes. Tissue-specific 2DG clearance and glucose uptake were related to the weight of the analyzed muscle tissue and time.

2.9. Muscle processing

Prior to homogenization, gastrocnemius, quadriceps, and triceps brachii muscles were pulverized in liquid nitrogen. All muscle were homogenized 2 × 30 sec at 30 Hz using a Tissuelyser II (Qiagen, USA) in ice-cold homogenization buffer (10% (v/v) Glycerol, 1% (v/v) NP-40, 20 mM Na-pyrophosphate, 150 mM NaCl, 50 mM HEPES (pH 7.5), 20 mM β-glycerophosphate, 10 mM NaF, 2mM PMSF, 1 mM EDTA (pH 8.0), 1 mM EGTA (pH 8.0), 2 mM Na3VO4, 10 μg mL⁻¹ Leupeptin, 10 μg mL⁻¹ Aprotinin, 3 mM Benzamidine). Homogenates were rotated end-over-end for 30 min at 4°C and lysate (supernatant) generated by centrifugation (10,854–15,630 × g) for 15–20 min at 4°C.

2.10. Immunoblotting

Lysate protein concentration was determined using the bicinchoninic acid method using bovine serum albumin (BSA) standards and bicinchoninic acid assay reagents (Pierce). Immunoblotting samples were prepared in 6X sample buffer (340 mM Tris (pH 6.8), 225 mM DTT, 11% (w/v) SDS, 20% (v/v) Glycerol, 0.05% (w/v) Bromphenol blue). Protein phosphorylation (p) and total protein expression were determined by standard immunoblotting technique loading equal amounts of protein. The polyvinylidene difluoride membrane (Immobilon Transfer Membrane; Millipore) was blocked in Tris-Buffered Saline with added Tween20 (TBST) and 2% (w/v) skim milk or 5% (w/v) BSA protein for 15 minutes at room temperature, followed by incubation overnight at 4°C with a primary antibody (Table 3). Next, the membrane was incubated with horseradish peroxidase-conjugated secondary antibody (Jackson Immuno Research) for 1 hour at room temperature. Bands were visualized using Bio-Rad ChemiDocTM MP Imaging System and enhanced chemiluminescence (ECL+; Amersham Biosciences). Densitometric analysis was performed using Image LabTM Software, version 4.0 (Bio-Rad, USA; RRID:SCR_014210). Coomassie Brilliant Blue staining was utilized as a control to assess total protein loading (Welinder & Ekblad, 2011) and for each sample set, a representative membrane from the immunoblotting is shown. For total protein expression, each data point presented the average of the protein expression in the left and right muscle from the same mouse.

Table 3:

Antibody Table

Antibody name	Antibody ID (RRID)	Manufacturer; Catalog Number;	Species Raised in; Monoclonal or Polyclonal	Antibody dilution	Blocking buffer
Akt2	AB_2225186	Cell Signaling Technology; 3063	Rabbit, Monoclonal antibody	1:1000	2% milk
pAkt S474*	AB_329825	Cell Signaling Technology; 9271	Rabbit; Polyclonal antibody	1:1000	2% milk
pAkt T309*	AB_329828	Cell Signaling Technology; 9275	Rabbit; Polyclonal antibody	1:1000	2% milk
GLUT4	AB_2191454	Thermo Fisher Scientific; PA1–1065	Rabbit; Polyclonal antibody	1:1000	2% milk
PAK1	AB_330222	Cell Signaling Technology; 2602	Rabbit; Polyclonal antibody	1:1000	2% milk
PAK2	AB_2283388	Cell Signaling Technology; 2608	Rabbit; Polyclonal antibody	1:1000	2% milk
pPAK1/2 T423/402	AB_330220	Cell Signaling Technology; 2601	Rabbit; Polyclonal antibody	1:1000	5% BSA
TBC1D4	AB_492639	Millipore; 07–741	Rabbit; Polyclonal antibody	1:1000	2% milk
pTBC1D4 T649**	AB_2651042	Cell Signaling Technology; 8881	Rabbit; Monoclonal antibody	1:1000	2% milk

^*Nomenclature for Akt2 was used for pAkt S474 and T309 (equivalent to pAkt1 S473 and T308).

^**Mouse nomenclature was used for pTBC1D4 T649 (equivalent to human T642).

2.11. Statistical analyses

Data are presented as mean ± S.D. or when applicable mean ± S.D. with individual data points shown. Statistical tests varied according to the dataset being analyzed and the specific tests used are indicated in the figure legends. Datasets were normalized by square root, log10, inverse or inverse square root transformation if not normally distributed or failed equal variance test. If the null hypothesis was rejected, Tukey’s post hoc test was used to evaluate significant main effects of genotype and significant interactions in ANOVAs. P < 0.05 was considered statistically significant. P<0.1 was considered a tendency. Except for mixed-effects model analyses performed in GraphPad Prism, version 8.2.1. (GraphPad Software, La Jolla, CA, USA; RRID:SCR_002798), all statistical analyses were performed using Sigma Plot, version 13 (Systat Software Inc., Chicago, IL, USA; RRID:SCR_003210). Due to missing values ascribed to hypoglycemia, differences between genotypes and the effect of insulin administration were assessed with a mixed-effects model analysis in Fig. 9G–H.

Figure 9: Whole-body insulin tolerance unaffected by genetic ablation of PAK1 and PAK2.

(A-C) Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) in both sexes combined (A) and in female (B) and male (C) chow-fed whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates. The number of mice in each group: Control, n = 9/11 (female/male); PAK1 KO, n = 8/10; PAK2 mKO, n = 10/9; 1/m2 dKO, n = 9/13. Data were evaluated with a two-way ANOVA to test the factors ‘PAK1’ and ‘PAK2’ thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. (D-F) Basal blood glucose concentration (fasted state) immediately before an insulin tolerance test (ITT) in both sexes combined (D) and in female (E) and male (F) PAK1 KO, PAK2 mKO, 1/m2 dKO mice and control littermates. The number of mice in each group: Control, n = 9/10 (female/male); PAK1 KO, n = 6/9; PAK2 mKO, n = 10/8; 1/m2 dKO, n = 8/13. Data were evaluated with a two-way ANOVA to test the factors ‘PAK1’ and ‘PAK2’ thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. (G-I) Blood glucose levels related to the basal concentration during an insulin tolerance test (ITT) in both sexes combined (G) and in female (H) and male (I) chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice or control littermates. The number of mice in each group: Control, n = 9/10 (female/male); PAK1 KO, n = 6/9; PAK2 mKO, n = 10/8; 1/m2 dKO, n = 8/13. For two female control mice and four female PAK2 mKO mice, the ITT had to be stopped before the 120’-time point due to hypoglycemia (blood glucose <1.2 mM), and thus blood glucose was not determined for these mice for the last couple of time points. Data were evaluated with five two-way ANOVAs to test the factors ‘PAK1’ and ‘PAK2’ at time point 15, 30, 60, 90 and 120, respectively, thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of insulin administration were assessed with a mixed-effects model analysis to test the factors ‘Genotype’ and ‘Time’. Main effects (ME) are indicated in the panels. Significant one-way ANOVAs and interactions in two-way (RM when applicable) ANOVAs were evaluated by Tukey’s post hoc test: Effect of insulin administration vs. time point 0’ **/*** (p<0.01/0.001); Control vs. PAK1 KO ¤/¤¤ (p<0.05/0.01); Control vs. PAK2 mKO (£)££ (p<0.1/0.01); Control vs. 1/m2 dKO ††† (p<0.001); PAK1 KO vs. 1/m2 dKO ‡ (p<0.05); PAK2 mKO vs. 1/m2 dKO $ (p<0.05). Data are presented as mean $\pm$ S.D. or when applicable mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line.

3. Results

3.1. Pharmacological inhibition of group I PAKs partially reduces insulin-stimulated glucose uptake in mouse soleus muscle

To investigate the involvement of group I PAKs in insulin-stimulated glucose uptake in mouse skeletal muscle, initially, we analyzed 2DG uptake in isolated soleus and EDL muscle in the presence of a pharmacological inhibitor, IPA-3. While glucose uptake in vivo is influenced by glucose delivery, GLUT4 translocation and muscle metabolism (Wasserman et al., 2011), glucose delivery is constant in isolated skeletal muscle and surface membrane GLUT4 is the limiting factor (Hansen et al., 1995, 2000; Zisman et al., 2000). Therefore, 2DG uptake in isolated muscles likely reflects GLUT4 translocation. IPA-3 is a well-characterized inhibitor of group I PAKs (PAK1-3) (Deacon et al., 2008; Tunduguru et al., 2014) and shown to completely abolish insulin-stimulated GLUT4 translocation and glucose uptake in L6-GLUT4myc myoblasts and myotubes, respectively (Tunduguru et al., 2014). In soleus, 2DG uptake increased 4.5-fold upon maximal insulin-stimulation, an increase that was partly reduced (−20%) in IPA-3-treated muscles (Fig. 1A). IPA-3 did not significantly (p=0.080) impair insulin-stimulated (+2.4-fold) 2DG uptake in EDL muscle (Fig. 1B). We confirmed that IPA-3 treatment abolished insulin-stimulated phosphorylation of (p)PAK1 T423 in both soleus and EDL muscle (Fig. 1C+D), indicating decreased insulin-stimulated activation of PAK1. In contrast, insulin-stimulated pAkt T309 (Fig. 1E+F) and pAkt S474 (Fig. 1G+H) were unaffected by IPA-3 treatment, suggesting that IPA-3 did not interfere with insulin signalling to Akt. Altogether this initial screening using a pharmacological inhibitor suggests that group I PAKs are partially involved in insulin-stimulated glucose uptake in isolated mouse muscle.

Figure 1: Pharmacological inhibition of group I PAKs partially reduces insulin-stimulated glucose uptake in mouse soleus muscle.

(A-B) Insulin-stimulated (60 nM) 2-deoxyglucose (2DG) uptake in isolated soleus (A) and extensor digitorum longus (EDL, B) muscle $\pm$ 40 μM IPA-3 or as a control DMSO (0.25%). Isolated muscles were pre-incubated for 25 minutes followed by 30 minutes of insulin stimulation with 2DG uptake analyzed for the final 10 minutes of stimulation. (C-H) Quantification of phosphorylated (p)PAK1 T423, pAkt T309, and pAkt S474 in insulin-stimulated soleus (C, E, and G) and EDL (D, F, and H) muscle $\pm$ 40 μM IPA-3 or as a control DMSO (0.25%). Some of the data points were excluded due to the quality of the immunoblot, and the number of determinations was n = 8/7 (DMSO/IPA-3) for pAkt T309and S474 in soleus muscle. (I-J) Representative blots showing pPAK1 T423, pAkt T309, and pAkt S474 in soleus (I) and EDL (J) muscle. The representative blots for pPAK1 T423 in soleus and EDL, respectively, are cropped from the same membrane. Data were evaluated with a two-way repeated measures (RM) ANOVA. Main effects (ME) are indicated in the panels. Interactions in two-way RM ANOVAs were evaluated by Tukey’s post hoc test: Insulin stimulation vs. basal **/*** (p<0.01/0.001); IPA-3 vs. DMSO (#)/#/##/### (p<0.1/0.05/0.01/0.001). Unless otherwise stated previously in the figure legend, the number of determinations in each group: Soleus, n = 9/8 (DMSO/IPA-3); EDL, n = 9/8. Data are presented as mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line. A.U., arbitrary units.

3.2. PAK1 knockout does not affect whole-body glucose tolerance or insulin-stimulated glucose uptake in isolated skeletal muscle

Hypothesizing that the effects of IPA-3 were attributable to the inhibition of PAK1, we next sought to confirm our findings in mice with whole-body genetic ablation of the PAK1 isoform (PAK1 KO) and therefore a complete knockout of PAK1 in skeletal muscle (Fig. 2A). In chow-fed mice, fat mass, lean body mass, body weight and energy intake (Fig. 2B–C) were similar between whole-body PAK1 KO and littermate controls, as also previously reported (Ahn et al., 2016). During a GTT, the lack of PAK1 had no effect on the blood glucose response (Fig. 2D–E) or plasma insulin concentration (Fig. 2F). Likewise, HOMA-IR, a measure of basal glucose homeostasis (Fig. 2G), and both submaximal and maximal insulin-stimulated 2DG uptake in isolated soleus and EDL muscle was unaffected by PAK1 KO (Fig. 2H–I). Thus, unexpectedly, genetic ablation of PAK1 alone did not impair whole-body glucose tolerance, or skeletal muscle insulin sensitivity (submaximal insulin-stimulated glucose uptake) or insulin responsiveness (maximal insulin-stimulated glucose uptake) in divergence to previous reports (Wang et al., 2011; Chiang & Jin, 2014; Tunduguru et al., 2014).

Figure 2: PAK1 knockout does not affect whole-body glucose tolerance or insulin-stimulated glucose uptake in isolated skeletal muscle.

(A) Representative blots showing PAK1 protein expression in gastrocnemius, quadriceps, and triceps brachii muscle from PAK1 knockout (KO) mice or control littermates. Coomassie Brilliant Blue staining was utilized to assess total protein loading. Coomassie Brilliant Blue staining of a representative membrane from the immunoblotting of the sample set is shown. (B) Body composition (FM: Fat mass; LBM: Lean body mass; BW: Body weight) in gram in chow-fed PAK1 KO mice (n = 12) or control littermates (n = 13). Data were evaluated with a Student’s t-test. (C) Energy intake in chow-fed PAK1 KO mice (n = 8) or control littermates (n = 8). Energy intake was monitored in a separate cohort of mice. Data were evaluated with a Student’s t-test. (D) Blood glucose levels during a glucose tolerance test (GTT) in chow-fed PAK1 KO mice (n = 9) or control littermates (n = 10). Data were evaluated with a two-way repeated measures (RM) ANOVA. (E) Incremental Area Under the Curve (AUC) for blood glucose levels during the GTT in panel D. Data were evaluated with a Student’s t-test. (F) Plasma insulin values during a GTT in chow-fed PAK1 KO mice (n = 10) or control littermates (n = 13). Data were evaluated with a two-way RM measures ANOVA. (G) Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) in chow-fed PAK1 KO mice (n = 10) or control littermates (n = 13). Data were evaluated with a Student’s t-test. (H-I) Submaximal (0.6 nM) and maximal (60 nM) insulin-stimulated 2-deoxyglucose (2DG) uptake in isolated soleus (H) and extensor digitorum longus (EDL; I) muscle from PAK1 KO mice or littermate controls. Isolated muscles were pre-incubated for 30 minutes followed by 30 minutes of insulin stimulation with 2DG uptake analyzed for the final 10 minutes of stimulation. The number of determinations in each group: Soleus-Control, n = 6/7 (Submax/max); Soleus-PAK1 KO, n = 7/7; EDL-Control, n = 6/6; EDL-PAK1 KO, n = 7/7. Data were evaluated with two two-way RM measures ANOVA. Main effects (ME) are indicated in the panels. Data are presented as mean $\pm$ S.D. or when applicable mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line.

3.3. PAK1 is dispensable for insulin-stimulated glucose uptake in lean or diet-induced insulin-resistant mice in vivo

Our findings on insulin-stimulated glucose uptake in isolated muscle from chow-fed PAK1 KO mice conflicted with a previous study reporting impaired glucose tolerance in PAK1 KO mice and defects in insulin-stimulated GLUT4 translocation in skeletal muscle in vivo (Wang et al., 2011). Therefore, we further explored the effect of PAK1 KO on insulin-stimulated glucose uptake in skeletal muscle in vivo. Additionally, we fed a subgroup of PAK1 KO and control littermate mice a 60E% HFD for 21 weeks to investigate the role of PAK1 in insulin-resistant muscle. Insulin administration lowered blood glucose by 5.4 $\pm$ 0.5 mM (−47%) in chow-fed control mice (Fig. 3A). In HFD-fed control mice, blood glucose dropped 3.0 $\pm$ 0.9 mM (−26%) upon insulin administration (Fig. 3B). Lack of PAK1 had no impact on either basal blood glucose or whole-body insulin tolerance on either of the diets (Fig. 3A–B). Insulin increased glucose uptake in muscles from chow-fed (Gastrocnemius: 8.1-fold; Quadriceps: 8.5-fold, Triceps brachii: 12.3-fold; Soleus: 8.9-fold) and HFD-fed (Gastrocnemius: 3.5-fold; Quadriceps: 1.8-fold, Triceps brachii: 4.3-fold; Soleus: 11.6-fold) control mice (Fig. 3C–D). Consistent with our findings in isolated muscle, we observed no effect of PAK1 KO on basal or insulin-stimulated glucose uptake in vivo in muscles of either chow-fed mice or in insulin-resistant muscles from HFD-fed mice (Fig. 3C–D). Like glucose uptake, 2DG clearance from the plasma was unaffected by PAK1 KO in all of the investigated muscles (Fig. 3E–F). Importantly, circulating [³H]-2DG availability was unaffected by genotype on both of the diets (Fig. 3G). As in chow-fed mice, lack of PAK1 in HFD-fed mice had no effect on fat mass, lean body mass, body weight or energy intake (Fig. 4A–B). Similarly, whole-body glucose tolerance (Fig. 4C–D), plasma insulin concentration during the GTT (Fig. 4E) and HOMA-IR (Fig. 4F) were unaffected by PAK1 KO in HFD-fed mice. Thus, PAK1 is dispensable for in vivo insulin-stimulated muscle glucose uptake in both the healthy lean and the diet-induced insulin-resistant state.

Figure 3: PAK1 is dispensable for insulin-stimulated glucose uptake in lean or diet-induced insulin-resistant mice in vivo.

(A-B) Blood glucose levels during a retro-orbital insulin tolerance test (r.o. ITT) in chow- (A) and 60E% high-fat diet (HFD)-fed (B) PAK1 knockout (KO) mice or control littermates. The number of mice in each group: Chow, n = 12/6 (Control/PAK1 KO); HFD, n = 8/11. Data were evaluated with a two-way repeated measures ANOVA. (C-F) Insulin-stimulated (Chow: 0.5 U kg⁻¹ body weight; HFD: 60% of insulin administered to chow-fed mice) glucose uptake index and 2-deoxyglucose (2DG) clearance in gastrocnemius (Gast), quadriceps (Quad), triceps brachii (Triceps) and soleus muscle from chow- (C+E) and HFD-fed (D+F) PAK1 KO mice or control littermates. The number of mice in each group: Chow-Saline, n = 10/4 (Control/PAK1 KO); Chow-Insulin, n = 12/6; HFD-Saline, n = 6/5; HFD-Insulin, n = 8/11. Data were evaluated with a two-way ANOVA for each of the muscles. (G) Plasma [³H] counts 10 minutes after retro-orbital (r.o.) administration of a bolus of saline containing [³H]-labelled 2DG ([³H]-2DG) with or without insulin. The number of mice in each group: Chow-Saline, n = 10/4 (Control/PAK1 KO); Chow-Insulin, n = 12/6; HFD-Saline, n = 6/5; HFD-Insulin, n = 8/11. Data were evaluated with two two-way ANOVAs to test the factors ‘stimuli’ (basal vs. insulin) and ‘genotype’ (control vs. PAK1 KO) in chow-fed and HFD-fed mice, respectively. Main effects (ME) are indicated in the panels. Data are presented as mean $\pm$ S.D. or when applicable mean $\pm$ S.D. with individual data points shown.

Figure 4: PAK1 knockout does not affect whole-body glucose tolerance in diet-induced obese mice.

(A) Body composition (FM: Fat mass; LBM: Lean body mass; BW: Body weight) in gram in 60E% high-fat diet (HFD)-fed PAK1 knockout (KO) mice (n = 17) or control littermates (n = 14). Data were evaluated with a Student’s t-test. (B) Energy intake in HFD-fed PAK1 KO mice (n = 8) or control littermates (n = 8). Energy intake was monitored in a separate cohort of mice. Data were evaluated with a Student’s t-test. (C) Blood glucose levels during a glucose tolerance test (GTT) in HFD-fed PAK1 KO mice (n = 17) or control littermates (n = 13). Data were evaluated with a two-way repeated measures (RM) ANOVA. (D) Incremental Area Under the Curve (AUC) for blood glucose levels during the GTT in panel C. Data were evaluated with a Student’s t-test. (E) Plasma insulin values during a GTT in HFD-fed PAK1 KO mice (n = 16) or control littermates (n = 11). Data were evaluated with a two-way RM ANOVA. (F) Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) in HFD-fed PAK1 KO mice (n = 16) or control littermates (n = 11). Data were evaluated with a Student’s t-test. Main effects (ME) are indicated in the panels. Data are presented as mean $\pm$ S.D. or when applicable mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line.

3.4. Whole-body substrate utilization is unaffected by genetic ablation of PAK1 and PAK2

Since there was no phenotype with genetic ablation of PAK1 alone, we next sought to determine the relative contribution and involvement of PAK1 and PAK2 in insulin signalling and glucose uptake in skeletal muscle. Double knockout mice with whole-body knockout of PAK1 and muscle-specific knockout of PAK2 (1/m2 dKO) were generated as previously described (Joseph et al., 2017). By crossing 1/m2 dKO mice with littermate controls, a cohort was generated consisting of whole-body PAK1 KO, muscle-specific PAK2 (m)KO, 1/m2 dKO and littermate control mice. Control littermates were PAK1^+/− heterozygous and PAK2^fl/fl mice without muscle-specific Cre expression. While no band for PAK1 could be detected in muscles lacking PAK1, muscles lacking PAK2 displayed only a partial reduction in band intensity in the immunoblots for PAK2 (Fig. 5A). This was likely because PAK1 KO mice are whole-body knockouts, while PAK2 mKO mice are muscle-specific and other cell types within skeletal muscle tissue could thus contribute to the signal obtained in the PAK2 immunoblots.

Figure 5: Combined whole-body genetic ablation of PAK1 and muscle-specific PAK2 knockout reduces lean body mass.

(A) Representative blots showing PAK1 and PAK2 protein expression in soleus and extensor digitorum longus (EDL) muscle from chow-fed whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates. Coomassie Brilliant Blue staining was utilized to assess total protein loading. Coomassie Brilliant Blue staining of a representative membrane from the immunoblotting of the sample set is shown. (B-D) Body weight of both sexes combined (B) and in female (C) and male (D) chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice or control littermates. (E-G) Body composition (FM: Fat mass; LBM: Lean body mass) in gram in both sexes combined (E) and in female (F) and male (G) chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice or control littermates. (H-J) Body composition in percentage in both sexes combined (H) and in female (I) and male mice (J).The number of mice in each group: Control, n = 9/11 (female/male); PAK1 KO, n = 8/10; PAK2 mKO, n = 12/9; 1/m2 dKO, n = 9/14. Data were evaluated with a two-way ANOVA to test the factors ‘PAK1’ (PAK1^+/− vs. PAK^−/−) and ‘PAK2’ (PAK2^fl/fl;MyoD^+/+ vs. PAK2^fl/fl;MyoD^iCre/+) thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. Main effects (ME) are indicated in the panels. Significant one-way ANOVAs were evaluated by Tukey’s post hoc test: Control vs. 1/m2 dKO †/††/††† (p<0.05/0.01/0.001); PAK1 KO vs. PAK2 mKO § (p<0.05); PAK1 KO vs. 1/m2 dKO ‡/‡‡/‡‡‡ (p<0.05/0.01/0.001); PAK2 mKO vs. 1/m2 dKO ($)/$/$ $ (p<0.1/0.05/0.01). Data are presented as mean $\pm$ S.D. with individual data points shown.

As previously shown (Joseph et al., 2017, 2019), 1/m2 dKO mice weighed less (−12%) than control littermates (Fig. 5B–D) due to reduced (−12%) lean body mass (Fig. 5E–G). Body weight and lean body mass decreased to the same extent in 1/m2 dKO mice, leaving lean body mass percentage largely unaffected (Fig. 5H–J). Using calorimetric chambers, we monitored whole-body metabolism for 72 hours during the light and dark period. On day 2, the mice fasted during the dark period followed by refeeding on day 3. Oxygen uptake (VO₂; Fig. 6A–C) and RER indicative of substrate utilization (Fig. 6D–F) were largely unaffected by genotype with only a slightly higher RER in mice lacking PAK2, either alone or in combination with PAK1 KO, upon fasting. Similar substrate utilization was obtained despite reduced habitual activity in the dark period in mice lacking PAK2, an effect largely driven by a decreased activity in 1/m2 dKO mice and increased activity in male PAK1 KO mice (Fig. 7A–C). Supporting the lower activity levels, energy intake was decreased (−11%) in 1/m2 dKO mice compared to PAK1 KO mice on day 1 (Fig. 7D) due to lower energy intake in the dark period (Fig. 7E). Upon refeeding, energy intake was reduced in mice lacking PAK2 (Fig. 7D–E) driven by a lower energy intake in the dark period in female mice lacking PAK2 (Fig. 7F–I). Altogether, these data suggest that lack of PAK1 and/or PAK2 does not compromise metabolic regulation during the light/dark period or in response to fasting/refeeding.

Figure 6: Whole-body substrate utilization is unaffected by genetic ablation of PAK1 and PAK2.

(A-F) Oxygen uptake (VO₂; A-C) and respiratory exchange ratio (RER; D-F) in both sexes combined (A+D) and in female (B+E) and male (C+F) chow-fed whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates during the light and dark period recorded over a period of 72 hours. On day 2, the mice were fasted during the dark period and then refed on day 3. The number of mice in each group: Control, n = 5/3 (female/male); PAK1 KO, n = 6/5; PAK2 mKO, n = 6/5; 1/m2 dKO, n = 6/7. Data were evaluated with two two-way ANOVAs to test the factors ‘PAK1’ and ‘PAK2’ in the light and dark period of day 1, respectively. Data for day 2 and 3 were evaluated similarly thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of the light and dark period were assessed with three two-way repeated measures (RM) ANOVA to test the factors ‘Genotype’ (Control vs. PAK1 KO vs. PAK2 mKO vs. 1/m2 dKO) and ‘Period’ (Light vs. Dark) on day 1, 2 and 3, respectively. Main effects (ME) are indicated in the panels. Interactions in two-way (RM when applicable) ANOVAs were evaluated by Tukey’s post hoc test: Light vs. dark period */**/*** (p<0.05/0.01/0.001). Data are presented as mean $\pm$ S.D. or when applicable mean $\pm$ S.D. with individual data points shown.

Figure 7: Habitual activity is reduced in the dark period in mice lacking PAK2.

(A-I) Activity (beam breaks; A-C) and energy intake (D-I) in both sexes combined (A+D+F) and in female (B+F+H) and male (C+G+H) chow-fed whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates during the light and dark period recorded over a period of 72 hours. On day 2, the mice were fasted during the dark period and then refed on day 3. The number of mice in each group: Control, n = 5/3 (female/male; for energy intake, n = 4/3); PAK1 KO, n = 6/5; PAK2 mKO, n = 6/5; 1/m2 dKO, n = 6/7. Data (A-C+E+H-I) were evaluated with two two-way ANOVAs to test the factors ‘PAK1’ and ‘PAK2’ in the light and dark period of day 1, respectively. Data for day 2 and 3 were evaluated similarly thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of the light and dark period were assessed with three two-way repeated measures (RM) ANOVA to test the factors ‘Genotype’ (Control vs. PAK1 KO vs. PAK2 mKO vs. 1/m2 dKO) and ‘Period’ (Light vs. Dark) on day 1, 2 and 3, respectively (for energy intake on day 2, differences between genotypes were evaluated with a one-way ANOVA). Total energy intake on day 1 and day 3 (D+F-G) were evaluated with two two-way ANOVAs to test the factors ‘PAK1’ and ‘PAK2’ on day 1 and day 3, respectively. Differences between genotypes and the day were assessed with one two-way RM ANOVA to test the factors ‘Genotype’ and ‘Day’ (Day 1 vs. Day 3). Main effects (ME) are indicated in the panels. Interactions in two-way (RM when applicable) ANOVAs were evaluated by Tukey’s post hoc test: Light vs. dark period */**/*** (p<0.05/0.01/0.001); Day 1 vs. Day 3 €€/€€€ (p<0.01/0.001); Control vs. PAK1 KO ¤¤ (p<0.01); Control vs. 1/m2 dKO † (p<0.05); PAK1 KO vs. PAK2 mKO §/§§ (p<0.05/0.01); PAK1 KO vs. 1/m2 dKO ‡/‡‡/‡‡‡ (p<0.05/0.01/0.001); PAK2 mKO vs. 1/m2 dKO $ (p<0.05). Data are presented as mean $\pm$ S.D. or when applicable mean $\pm$ S.D. with individual data points shown.

3.5. Glucose tolerance is reduced in mice lacking PAK2 in skeletal muscle

To test dependency on PAK1 and/or PAK2 in glucose handling and insulin sensitivity, we next investigated glucose and insulin tolerance in the 1/m2 dKO mouse strain. Blood glucose concentration in the fed state was similar between the genotypes (Fig. 8A–C). The blood glucose response to a glucose load was slightly increased in mice lacking PAK2 in skeletal muscle as evident by the increased area under the blood glucose curve (Fig. 8D–I). This was primarily driven by impaired glucose tolerance in female PAK2 mKO mice (Fig. 8E–F+H–I). Plasma insulin concentration during the GTT was unaffected by lack of PAK1 (Fig. 8J). In contrast, plasma insulin in male, but not female, PAK2 mKO mice was slightly elevated compared to 1/m2 dKO mice and tended (p=0.055) to be higher than control littermates (Fig. 8K–L), indicating impaired insulin sensitivity in male PAK2 mKO mice. HOMA-IR was unaffected by lack of PAK1 and/or PAK2 (Fig. 9A–C). In addition, even though fasted blood glucose immediately before an ITT was modestly reduced in PAK2 mKO mice (Fig. 9D–F), the blood glucose response to an ITT was largely unaffected by lack of either PAK1 and/or PAK2 (Fig. 9G–I). Thus, despite slightly impaired glucose tolerance in mice lacking PAK2 either alone or in combination with PAK1 in skeletal muscle, neither adverse effects on plasma insulin nor defects in insulin sensitivity could be detected.

Figure 8: Glucose tolerance is reduced in mice lacking PAK2 in skeletal muscle.

(A-C) Blood glucose concentration in the fed state (8 a.m.) in both sexes combined (A) and in female (B) and male (C) chow-fed whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates. Control, n = 9/11 (female/male); PAK1 KO, n = 8/10; PAK2 mKO, n = 12/9; 1/m2 dKO, n = 9/14. Data were evaluated with a two-way ANOVA to test the factors ‘PAK1’ (PAK1^+/− vs. PAK^−/−) and ‘PAK2’ (PAK2^fl/fl;MyoD^+/+ vs. PAK2^fl/fl;MyoD^iCre/+) thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. (D-F) Blood glucose levels during a glucose tolerance test (GTT) in both sexes combined (D) and in female (E) and male (F) chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice or control littermates. The number of mice in each group: Control, n = 8/11 (female/male); PAK1 KO, n = 8/10; PAK2 mKO, n = 10/9; 1/m2 dKO, n = 8/13. Data were evaluated with six two-way ANOVAs to test the factors ‘PAK1’ and ‘PAK2’ at time point 0, 15, 30, 60, 90 and 120, respectively, thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of glucose administration were assessed with a two-way repeated measures (RM) ANOVA to test the factors ‘Genotype’ (Control vs. PAK1 KO vs. PAK2 mKO vs. 1/m2 dKO) and ‘Time’ (0 vs. 15 vs. 30 vs. 60 vs. 90 vs. 120). (G-I) Incremental Area Under the Curve (AUC) for blood glucose levels during the GTT in panel D-F in both sexes combined (G) and in female (H) and male mice (I). Data were evaluated with a two-way ANOVA to test the factors ‘PAK1’ and ‘PAK2’ thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. (J-L) Plasma insulin values during a GTT in both sexes combined (J) and in female (K) and male (L) chow-fed PAK1 KO, PAK2 mKO, 1/m2 dKO mice or control littermates. The number of mice in each group: Control, n = 9/10 (female/male); PAK1 KO, n = 8/9; PAK2 mKO, n = 10/9; 1/m2 dKO, n = 9/13. Data were evaluated with two two-way ANOVAs to test the factors ‘PAK1’ and ‘PAK2’ at time point 0 and 20, respectively, thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of glucose administration were assessed with a two-way RM ANOVA to test the factors ‘Genotype’ and ‘Time’. Main effects (ME) are indicated in the panels. Significant one-way ANOVAs and interactions in two-way (RM when applicable) ANOVAs were evaluated by Tukey’s post hoc test: Effect of glucose administration vs. time point 0’ */*** (p<0.05/0.001); Control vs. PAK2 mKO £ (p<0.05); PAK1 KO vs. PAK2 mKO §§ (p<0.01); PAK2 mKO vs. 1/m2 dKO $ (p<0.05). Data are presented as mean $\pm$ S.D. or when applicable mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line.

3.6. Insulin-stimulated glucose uptake relies partially on PAK2 in EDL, but not soleus muscle, while PAK1 is not involved

To determine the relative contribution and involvement of PAK1 and PAK2 in glucose uptake in skeletal muscle, we next investigated insulin-stimulated 2DG uptake in isolated soleus and EDL. Soleus muscle lacking PAK1 and PAK2 displayed normal insulin-stimulated 2DG uptake compared to control littermates (Fig. 10A,C). In contrast, lack of PAK2 in EDL muscle caused a modest reduction either alone (PAK2 mKO: −18%) or in combination with PAK1 KO (1/m2 dKO: −12%) in insulin-stimulated 2DG uptake (Fig. 10B,D). Thus, in oxidative soleus muscle, group I PAKs are surprisingly dispensable for normal insulin-stimulated glucose uptake, whereas in glycolytic EDL muscle PAK2 is partly required.

Figure 10: Insulin-stimulated glucose uptake relies partially on PAK2 in EDL, but not soleus muscle, while PAK1 is not involved.

(A-B) Insulin-stimulated (60 nM) 2-deoxyglucose (2DG) uptake in isolated soleus (A) and extensor digitorum longus (EDL; B) muscle from whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates. Isolated muscles were pre-incubated for 20 minutes followed by 20 minutes of insulin stimulation with 2DG uptake analyzed for the final 10 minutes of stimulation. Data were evaluated with two two-way ANOVAs to test the factors ‘PAK1’ (PAK1^+/− vs. PAK^−/−) and ‘PAK2’ (PAK2^fl/fl;MyoD^+/+ vs. PAK2^fl/fl;MyoD^iCre/+) in basal and insulin-stimulated samples, respectively, thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of insulin stimulation were assessed with a two-way repeated measures (RM) ANOVA to test the factors ‘Genotype’ (Control vs. PAK1 KO vs. PAK2 mKO vs. 1/m2 dKO) and ‘Stimuli’ (Basal vs. Insulin). (C-D) Δ2DG uptake in soleus (C) and EDL (D) muscle from panel C-D. Data were evaluated with a two-way ANOVA to test the factors ‘PAK1’ and ‘PAK2’ to assess the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. The number of determinations in each group: Control, n = 9/10 (soleus/EDL); PAK1 KO, n = 8/9; PAK2 KO, n = 10/10; 1/m2 dKO, n = 9/9. Main effects (ME) are indicated in the panels. Significant one-way ANOVAs and interactions in two-way (RM when applicable) ANOVAs were evaluated by Tukey’s post hoc test: Insulin-stimulation vs. basal control *** (p<0.001); Control vs. PAK2 mKO £ (p<0.05); PAK1 KO vs. PAK2 mKO § (p<0.05). Data are presented as mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line.

3.7. PAK2 regulates TBC1D4 protein expression and signalling

Lastly, we looked into the effects of PAK1 and PAK2 on insulin-stimulated signalling. All groups displayed normal basal and insulin-stimulated pAkt T309 (Fig. 11A–D) and pAkt S474 (Fig. 11E–H) and Akt2 protein expression (Fig. 11I–J) compared to control littermates in both soleus and EDL muscle. In contrast, lack of PAK2 increased protein expression of Akt’s downstream target, TBC1D4 in soleus muscle (PAK2 mKO: +47%; 1/m2 dKO: +20%) (Fig. 12A), while reducing TBC1D4 expression in EDL (PAK2 mKO: −33%; 1/m2 dKO: −9%) (Fig. 12B). In soleus, basal and insulin-stimulated pTBC1D4 T649 was similar in all groups (Fig. 12C–D), suggesting that even with increased TBC1D4 expression, signalling through this pathway was normal. Concomitant with the decreased TBC1D4 expression in EDL muscle, lack of PAK2 reduced insulin-stimulated pTBC1D4 T649 (Fig. 12E–F) driven by attenuated (−46%) insulin-stimulated pTBC1D4 T649 in PAK2 mKO mice compared to control littermates (Fig. 12F). Knockout of TBC1D4 has been associated with lower GLUT4 protein abundance in some muscles (Lansey et al., 2012; Wang et al., 2013). Whereas GLUT4 protein expression was unaffected by lack of either PAK1 and/or PAK2 in soleus (Fig. 12G), GLUT4 protein expression was mildly reduced in EDL in PAK2 mKO mice compared to littermate controls (Fig. 12H). However, in a slightly older cohort (10–16 vs. 26–35 weeks of age), GLUT4 expression was reported to be unaffected by lack of PAK1 and/or PAK2 in EDL muscle but reduced in soleus muscle from 1/m2 dKO mice compared to control littermates (Møller et al., 2020). Yet, in the present investigation, the slightly reduced insulin-stimulated glucose uptake in EDL muscle lacking PAK2 was concomitant with downregulated TBC1D4 signalling and GLUT4 expression potentially explaining the requirement of PAK2, but not PAK1, in insulin-stimulated glucose uptake.

Figure 11: Insulin-stimulated Akt phosphorylation is unaffected by genetic ablation of PAK1 and PAK2.

(A-J) Quantification of phosphorylated (p)Akt T309, ΔpAkt T309, pAkt S474, ΔpAkt S474 and total Akt2 protein expression in insulin-stimulated (60 nM) soleus (A-B, E-F, and I) and extensor digitorum longus (EDL; C-D, G-H, and J) muscle from whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates. Total protein expression is an average of the paired basal and insulin-stimulated sample. (K) Representative blots showing pAkt T309, pAkt S474, and total PAK1, PAK2, and Akt2 protein expression in soleus and EDL muscle. Coomassie Brilliant Blue staining was utilized to assess total protein loading. Coomassie Brilliant Blue staining of a representative membrane from the immunoblotting of the sample set is shown. Protein phosphorylation was evaluated with two two-way ANOVAs to test the factors ‘PAK1’ (PAK1^+/− vs. PAK^−/−) and ‘PAK2’ (PAK2^fl/fl;MyoD^+/+ vs. PAK2^fl/fl;MyoD^iCre/+) in basal and insulin-stimulated samples, respectively, thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of insulin stimulation were assessed with a two-way repeated measures (RM) ANOVA to test the factors ‘Genotype’ (Control vs. PAK1 KO vs. PAK2 mKO vs. 1/m2 dKO) and ‘Stimuli’ (Basal vs. Insulin). Total protein expression and Δ-phosphorylation were evaluated with a two-way ANOVA to test the factors ‘PAK1’ and ‘PAK2’ thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. Main effects (ME) are indicated in the panels. The number of determinations in each group: Control, n = 9/10 (soleus/EDL); PAK1 KO, n = 8/9; PAK2 KO, n = 10/9; 1/m2 dKO, n = 9/9. Data are presented as mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line. A.U., arbitrary units.

Figure 12: PAK2 regulates TBC1D4 protein expression and signalling.

(A-H) Quantification of phosphorylated (p)TBC1D4 T649, ΔpTBC1D4 T649 and total TBC1D4 and GLUT4 protein expression in insulin-stimulated (60 nM) soleus (A, C, D, and G) and extensor digitorum longus (EDL; B, E, F, and H) muscle from whole-body PAK1 knockout (KO), muscle-specific PAK2 (m)KO, PAK1/2 double KO (1/m2 dKO) mice or control littermates. Total protein expression is an average of the paired basal and insulin-stimulated sample. Some of the data points were excluded due to the quality of the immunoblot, so the number of determinations for GLUT4 in soleus muscle: Control, n = 6; PAK1 KO, n = 5; PAK2 KO, n = 6; 1/m2 dKO, n = 6. (I) Representative blots showing pTBC1D4 T649 and total TBC1D4 protein expression in soleus and EDL muscle. Coomassie Brilliant Blue staining was utilized to assess total protein loading. Coomassie Brilliant Blue staining of a representative membrane from the immunoblotting of the sample set is shown. Protein phosphorylation was evaluated with two two-way ANOVAs to test the factors ‘PAK1’ (PAK1^+/− vs. PAK^−/−) and ‘PAK2’ (PAK2^fl/fl;MyoD^+/+ vs. PAK2^fl/fl;MyoD^iCre/+) in basal and insulin-stimulated samples, respectively, thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes and the effect of insulin stimulation were assessed with a two-way repeated measures (RM) ANOVA to test the factors ‘Genotype’ (Control vs. PAK1 KO vs. PAK2 mKO vs. 1/m2 dKO) and ‘Stimuli’ (Basal vs. Insulin). Total protein expression and Δ-phosphorylation was evaluated with a two-way ANOVA to test the factors ‘PAK1’ and ‘PAK2’ thereby assessing the relative contribution of PAK1 and PAK2. Differences between genotypes were evaluated with a one-way ANOVA. Main effects (ME) are indicated in the panels. Significant one-way ANOVA and interactions in two-way (RM when applicable) ANOVA were evaluated by Tukey’s post hoc test: Control vs. PAK2 mKO £ (p<0.05). Unless otherwise stated previously in the figure legend, the number of determinations in each group: Control, n = 9/10 (soleus/EDL); PAK1 KO, n = 8/9; PAK2 KO, n = 10/9; 1/m2 dKO, n = 9/9. Data are presented as mean $\pm$ S.D. with individual data points shown. Paired data points are connected with a straight line. A.U., arbitrary units.

4. Discussion

Here, we undertook a systematic investigation into the requirement of the group I PAK isoforms in muscle glucose uptake and whole-body metabolic regulation. In contrast to previous literature, our results firmly show that PAK1, surprisingly, is dispensable for insulin-stimulated glucose uptake in skeletal muscle, while PAK2 may play a minor role. Using a cohort of whole-body PAK1 KO mice and a distinct cohort of transgenic mice lacking either PAK1 (whole-body), PAK2 (muscle-specific), or jointly lacking both PAK1 and muscle PAK2, we show that PAK1 is not required in insulin-stimulated muscle glucose uptake in vivo or in isolated muscles. In accordance, we found no effect of whole-body PAK1 KO on glucose tolerance in either mice fed a standard chow diet (insulin sensitive mice) or in mice fed a HFD (insulin-resistant mice). In contrast, PAK2 seemed partially required for insulin-stimulated glucose uptake in glycolytic EDL muscle. This could potentially explain the slightly impaired glucose tolerance with the muscle-specific knockout of PAK2 in mice. Nevertheless, supporting only a minor role for skeletal muscle PAK2 in the whole-body substrate utilization, RER was largely unaffected by lack of PAK1 and/or PAK2 and only slightly elevated in mice lacking PAK2 when challenged by fasting.

In a previous study, the increase in GLUT4 abundance at the plasma membrane in response to insulin was completely abrogated in PAK1 KO muscle as measured by subcellular fractionation of homogenates of hindlimb skeletal muscles (Wang et al., 2011), suggesting that PAK1 is indispensable for insulin-stimulated GLUT4 translocation. Although glucose uptake was not assessed in that study (Wang et al., 2011), this indicated a key role for PAK1 in regulating glucose uptake in mouse skeletal muscle. Surprisingly, PAK1 was not required for insulin-stimulated glucose uptake in our hands. Furthermore, in our study, both chow- and HFD-fed PAK1 KO mice displayed blood glucose concentrations similar to control littermates during a GTT. This was in contrast to previous studies reporting impaired glucose tolerance in chow-fed PAK1 KO mice (Wang et al., 2011; Chiang et al., 2013) and elevated fasting blood glucose in PAK1 KO mice fed a western diet (45E% fat) (Ahn et al., 2016). Instead, despite the previous finding that insulin-stimulated GLUT4 translocation was unaffected by a 75% knockdown of PAK2 in L6-GLUT4myc myoblasts (Tunduguru et al., 2014), we found that muscle-specific PAK2 KO slightly impaired whole-body glucose tolerance and insulin-stimulated glucose uptake in mouse skeletal muscle.

These discrepancies between our and previous findings are difficult to delineate but might be due to methodological differences. Wang et al. (2011) used a crude fractionation method to measure GLUT4 translocation, whereas we analyzed the direct outcome hereof: glucose uptake. However, although discrepancies between the presence of GLUT4 at the plasma membrane and glucose uptake have occasionally been reported in cell culture studies (Somwar et al., 2001a, 2001b; Funaki et al., 2004), the insulin-induced increase in 3-O-methylglucose uptake correlates with the increase in cell surface GLUT4 protein content in human skeletal muscle strips (Lund et al., 1997). Given that fasting time affects metabolism (Andrikopoulos et al., 2008; Carper et al., 2020), this discrepancy is more likely explained by the longer fasting time (16 hours) used by Wang et al. (2011) prior to measurement of insulin-stimulated GLUT4 translocation and additionally the use of a supraphysiological insulin dose (21 U kg⁻¹ body weight i.p.). Another potential explanatory factor could be that our studies were conducted in both female and male mice, whereas past studies in PAK1 KO mice have been conducted in 4–6 months old male mice (Wang et al., 2011; Ahn et al., 2016). However, our data suggest no major differences between female and male mice in the response to lack of PAK1 and/or PAK2 on the whole-body metabolic parameters measured. Instead, the discrepancies between our and previous findings could be due to an effect of age, as our studies were conducted in mice at different ranges of age (10–37 of weeks age at the terminal experiment). In fact, age-dependent myopathy and development of megaconial mitochondria have been reported in the 1/m2 dKO mice (Joseph et al., 2019). Regardless, even though a role for group I PAKs in age-related insulin resistance should be further investigated, our investigation suggests that group I PAKs are largely dispensable in regulating whole-body glucose homeostasis or insulin-stimulated glucose uptake in skeletal muscle.

In the current study, pharmacological inhibition of group I PAKs inhibited muscle glucose uptake in response to insulin. Similarly, IPA-3 previously inhibited insulin-stimulated GLUT4 translocation and glucose uptake in L6-GLUT4myc myoblasts and myotubes, respectively (Tunduguru et al., 2014). IPA-3 is a non-ATP-competitive allosteric inhibitor of all group I PAKs (PAK1, 2, and 3). IPA-3 binds covalently to the regulatory CRIB domain of group I PAKs, thereby preventing binding to PAK activators, such as Rac1 (Deacon et al., 2008). Although IPA-3 is reported to be a highly selective and well-described inhibitor of group I PAKs that does not affect other groups of PAKs or similar kinases tested (Deacon et al., 2008), pharmacological inhibitors often have off-target effects (Davies et al., 2000). Thus, in a kinase screen IPA-3 has been reported to significantly inhibit Akt2 and Glycogen Synthase Kinase (GSK)-3α/β (Deacon et al., 2008), while the inhibitor is also suggested to alter the redox potential of cells (Rudolph et al., 2013). However, it is also possible that acute IPA-3-induced inhibition of group I PAKs elicits more potent effects compared with jointly knockout of PAK1 and PAK2 because the transgenic manipulations have been present from birth and may thus have resulted in compensatory changes. The development of inducible muscle-specific group I PAK deficient models could help clarify this. Importantly, any possible compensatory mechanisms cannot be via redundancy with group I PAKs, as PAK1 and PAK2 are removed genetically, and even in 1/m2 dKO mice, PAK3 cannot be detected at the protein level (Joseph et al., 2017). This emphasizes that group I PAKs are largely dispensable for insulin-stimulated glucose uptake in skeletal muscle with only PAK2 playing a minor role.

We hypothesized that group I PAKs, in particular PAK1, would be significantly involved in insulin-stimulated glucose uptake because of the established necessity of the upstream activator, Rac1 (Khayat et al., 2000; JeBailey et al., 2007; Ueda et al., 2010; Sylow et al., 2013a, 2014; Raun et al., 2018). Our findings suggest that Rac1 does not exclusively mediate insulin-stimulated glucose uptake through group I PAKs. Another downstream target of Rac1 is RalA. GLUT4 translocation induced by a constitutively activated Rac1 mutant was abrogated in L6-GLUT4myc myoblasts upon RalA knockdown (Nozaki et al., 2012) and, importantly, overexpression of a dominant-negative mutant of RalA reduced GLUT4 translocation in response to insulin in mouse gastrocnemius muscle fibres (Takenaka et al., 2015). Accordingly, Rac1-mediated regulation of insulin-stimulated GLUT4 translocation could be via RalA. Additionally, Rac1 is an essential component in the activation of the NADPH oxidase (NOX) complex (Abo et al., 1991; Bedard & Krause, 2007). In L6-GLUT4myc myotubes, reactive oxygen species have been reported to induce NOX2-dependent GLUT4 translocation in response to insulin (Contreras-Ferrat et al., 2014). A recent study suggested a role for Rac1 in the regulation of muscle glucose uptake through activation of the NOX2 in response to exercise (Henríquez-Olguin et al., 2019). Since Rac1 is required for both contraction- and insulin-stimulated glucose uptake in isolated mouse muscle (Sylow et al., 2013b, 2013a), Rac1 could also be involved in insulin-stimulated glucose uptake via NOX2 activation. Yet, insulin-stimulated NOX2 regulation in mature muscle remains to be investigated. Consequently, future studies should aim to investigate other players downstream of Rac1 since group I PAKs seem to be largely dispensable for glucose uptake in mature skeletal muscle (Møller et al., 2020).

5. Conclusion

Based on our present findings, we conclude that PAK2, but not PAK1, is partially required for insulin-stimulated glucose uptake in EDL muscle. Thus, the present study challenges that group I PAKs, and especially PAK1, are major regulators of whole-body glucose homeostasis and insulin-stimulated glucose uptake in skeletal muscle.

Key points summary.

• Muscle-specific genetic ablation of PAK2, but not whole-body PAK1 knockout, impairs glucose tolerance in mice

• Insulin-stimulated glucose uptake partly relies on PAK2 in glycolytic EDL muscle

• Contrasting previous reports, PAK1 is dispensable for insulin-stimulated glucose uptake in mouse muscle