An optimized fractionation method reveals insulin-induced membrane surface localization of GLUT1 to increase glycolysis in LβT2 cells

Abstract

Insulin is an important regulator of whole-body glucose homeostasis. In insulin sensitive tissues such as muscle and adipose, insulin induces the translocation of glucose transporter 4 (GLUT4) to the cell membrane, thereby increasing glucose uptake. However, insulin also signals in tissues that are not generally associated with glucose homeostasis. In the human reproductive endocrine axis, hyperinsulinemia suppresses the secretion of gonadotropins from gonadotrope cells of the anterior pituitary, thereby linking insulin dysregulation to suboptimal reproductive health. In the mouse, gonadotropes express the insulin receptor which has the canonical signaling response of IRS, AKT, and mTOR activation. However, the functional outcomes of insulin action on gonadotropes are unclear. Here, we demonstrate through use of an optimized cell fractionation protocol that insulin stimulation of the LβT2 gonadotropic cell line results in the unexpected translocation of GLUT1 to the plasma membrane. Using our high purity fractionation protocol, we further demonstrate that though Akt signaling in response to insulin is intact, insulin-induced translocation of GLUT1 occurs independently of Akt activation in LβT2 cells.

1. Introduction

The Hypothalamic-Pituitary-Gonadal (HPG) axis regulates fertility and gonadal tissue function in both females and males through the concerted action of the gonadotropic glycoprotein hormones luteinizing hormone (LH) and follicle-stimulating hormone (FSH), both secreted by the gonadotropes of the anterior pituitary. Secretion of gonadotropins in turn is stimulated by hypothalamic neurons which secrete gonadotropin-releasing hormone (GnRH). Activation of the cognate GnRH receptor (GnRHR) expressed by gonadotropes results in secretion of gonadotropins and activation of transcriptional, translational, and adaptive endoplasmic reticulum responses including the unfolded protein response. In the gonad, gonadotropins regulate reproduction and fertility through stimulation of ovarian follicular growth, steroidogenesis, and ovulation in the female and steroidogenesis and spermatogenesis in the male. Thus, dysregulation of gonadotropin secretion can have a profound effect on reproductive health. Notably, the response to GnRH is also associated with increased glycolysis and mobilization of glucose through GLUT1, which is translocated to the plasma membrane by GnRHR activation¹. In the anterior pituitary, GLUT1 is highly expressed in gonadotropes, and the secretory response is correlated with GLUT1 expression levels. In the female mouse, GLUT1 expression is also correlated with the estrous cycle stage, peaking at estrus, the time of the surge in LH secretion that precipitates ovulation. Knockdown of Slc2a1 mRNA, which encodes the GLUT1 protein, results in reduced GnRH-stimulated LH secretion, indicating the central role of GLUT1 in supporting the signaling response to GnRH^2,3.

Dysregulation of the HPG axis can lead to various reproductive disorders, such as lack of reproductive development, precocious or delayed puberty, polycystic ovary syndrome (PCOS), hyperandrogenism, oligo- or amenorrhea, or hypogonadism. Many reproductive disorders are linked to derangements in energy balance and inflammation, and changes in systemic metabolism are a common disruptor of the HPG axis. Insulin is a prime example. Complex relationships between insulin and serum gonadotropins have been observed in humans^4–9. In lean healthy women, infusion of insulin suppressed baseline and GnRH-stimulated LH⁹, while in both normal cycling women and women with PCOS, it is apparent that insulin is negatively correlated with LH levels¹⁰. Given that hyperinsulinemia is commonly associated with conditions that impact reproductive health including obesity and PCOS, it is important to understand the molecular mechanisms that may be responsible for the observed regulation of LH by insulin^4,7,11.

The primary canonical role of insulin is to regulate glucose homeostasis by facilitating the uptake of glucose from the circulation into target tissues. Insulin signals through the insulin receptor (IR), a transmembrane receptor tyrosine kinase. Upon binding, insulin induces a conformational change that activates receptor autophosphorylation. This event begins a signaling cascade that results in the phosphorylation and activation of protein kinase B (Akt). Akt then phosphorylates a Rab GTPase-activating protein called Akt substrate of 160kDA (AS160). This phosphorylation event deactivates AS160 which negatively regulates the trafficking of glucose transporter 4 (GLUT4)-containing vesicles. This deactivation leads to the fusion of GLUT4-containing vesicles to the cell membrane, where GLUT4 then facilitates glucose uptake. Recently, we and others have demonstrated that glucose metabolism and homeostasis is integral to the regulation of gonadotropin secretion from the pituitary^1–3. Further, we have also demonstrated that IR and GnRHR can potentiate or interfere with their signaling responses depending on the context^6,12. Given that GnRH-induces translocation of GLUT1 to the membrane of gonadotropes, and that insulin canonically regulates glucose uptake, we surmised that regulation of GLUT1 could be occurring in gonadotropes and be a functional outcome of insulin stimulation.

Using the gonadotropic cell line LβT2, we show insulin-stimulated GLUT1 and phosphorylated Akt subcellular redistribution with the use of an optimized subcellular fractionation protocol. With this protocol, we show that insulin induces GLUT1 membrane localization in an Akt independent manner despite phosphorylated Akt being found in cytosolic fractions¹³. Regulation of GLUT1 by insulin may provide the mechanistic basis of the context-dependent signaling interaction between insulin and GnRH for the physiological observation of insulin-GnRH co-regulation of gonadotropin section.

2. Materials and Methods

Tissue culture

The female mouse-derived LβT2 gonadotrope cell line^14–16 was maintained in high-glucose (4.5 g/L) HEPES-buffered DMEM supplemented with penicillin/streptomycin and 10% fetal bovine serum (FBS: FB-11, Omega Scientific, CA) at 37°C in a humidified atmosphere of 5% CO₂. EA.hy926 cells (ATCC, #CRL-2922) were grown at 37°C, 8% CO₂ in 10% FBS-DMEM (Gibco, #10–013-CV and #10437–028) supplemented with fresh 20% pre-conditioned media every two days and cultured to passage 8. HeLa cells were grown in DMEM containing 10% (v/v) FBS, 100 units/ml penicillin, and 0.1 mg/ml streptomycin in a humidified incubator at 5% (v/v) CO₂. PC12 cells were cultured in Dulbecco modified Eagle medium (DMEM) cell culture medium supplemented with 10% fetal bovine serum (FBS) and 5% donor horse serum (DHS) at 37°C with 5% CO₂.

Insulin treatment and Akt inhibition Treatment

To test the effects of insulin, LβT2 cells were seeded at 2 × 10⁵ cells per cm², cultured for 24 h, and pretreated with serum-free DMEM for 12–16 h prior to incubation with 1nM insulin (1pmol/L) (I9278 Sigma) or 1μM MK-2206 (Selleck, #S1078) treatment. MK-2206 was dissolved in DMSO and DMSO was used as the vehicle control.

Subcellular Fractionation Established protocol

The Abcam fractionation protocol published online at https://www.abcam.com/protocols/subcellular-fractionation-protocol is widely used for separating cellular components. We have optimized this protocol that we refer to as the Established protocol (ED) from herein. The Established Protocol is utilized to separate mitochondrial, nuclear, membrane, and cytoplasmic fractions and is summarized in Figure 1A. Cells were grown to confluence in 10 cm² plates. Cellular monolayers were washed twice with ice cold PBS. Next, 500μL of ice-cold subcellular fractionation buffer (20mM HEPES pH 7.4, 10mM KCl, 2mM MgCl₂, 1mM EDTA, 1mM EGTA, 1mM dithiothreitol, and Protease Inhibitor Cocktail (III)) was added, and cells were harvested by scraping. Cells were lysed by incubating on ice for 15 min and sheared by passing through a 27G needle 10 times. Lysates were incubated on ice for an additional 20 min. The nuclei were pelleted by centrifugation at 720 × g for 5 min at 4°C and the supernatant, which includes the mitochondria, membrane, and cytosol, was saved for further processing. The nuclear pellet was washed and resuspended in 500μL of fresh subcellular fractionation buffer, passed through a 25G syringe 10 times, and clarified of debris by centrifugation at 720 × g for 10 min at 4°C. In the Established protocol, the supernatant from this step (Step 6) is discarded. However, we analyzed this fraction (Waste from 6) for subcellular protein fractionation markers by immunoblot (Supplemental Figure 1). The cleared nuclear pellet was resuspended in 500μL TBS in 0.1% SDS and the clear nuclear suspension was sonicated for 3 sec on ice at a power setting of 2-continuous to shear genomic DNA and homogenize the suspension. The supernatant containing the cytoplasm, mitochondria, and membrane, was centrifuged for 5 min at 10,000g to pellet the mitochondria. The supernatant was transferred to a new Eppendorf tube and the mitochondrial pellet was processed exactly like the nuclear pellet. The fraction containing the membrane and cytoplasm was ultracentrifuged at 100,000 × g for 1 h at 4°C. The supernatant was saved as cytoplasm fraction 1. The membrane pellet was washed, resuspended in 500μL of subcellular fractionation buffer, passed through a 25G needle, and re-centrifuged for 45 min at the same settings while the cytoplasmic fraction was washed. The supernatant fraction was saved as the cytoplasmic fraction 2 and the membrane fraction was processed the same way as the mitochondria and nuclear fractions. Samples were mixed with 2x Laemmli Sample Buffer (LSB) containing 200mM dithiothreitol, boiled, resolved by SDS-PAGE and immunoblotted.

Figure 1.

Schematic of the Established and OMI subcellular fractionation protocols. (A) In the Established protocol, samples are lysed followed by centrifugation to produce a pellet containing the nuclear fraction and supernatant containing membrane, cytosolic, and mitochondrial proteins. The nuclear pellet is washed with fractionation buffer and centrifuged again. Supernatant from this centrifugation step is discarded (Waste from 6) and the nuclear pellet resuspended. The supernatant from the initial centrifugation of this protocol undergoes two centrifugations at increasing g force to pellet the mitochondrial fraction, followed by the membrane fraction. The remaining supernatant is the cytosolic fraction. (B) In the OMI protocol, samples are lysed and then centrifuged to produce supernatant containing the membrane and cytosolic proteins and a pellet with nuclear proteins. The supernatant is further processed via ultracentrifugation to obtain the cytosolic fraction and a pellet containing membrane proteins. The membrane and nuclear fraction are recombined, followed by centrifugation to obtain the nuclear and membrane fractions. M=membrane, C=cytosolic, N=nuclear, m=mitochondria

Subcellular Fractionation OMI protocol

Cells were washed twice with ice cold 1x PBS and then harvested by scraping in 1x PBS. Subsequently, the cells were centrifuged for 5 min at 2655 × g at 4°C, resuspended in 250 μL of 1x hypotonic lysis buffer (20 mM HEPES, pH 7.4, 10 mM NaCl, 3 mM MgCl2, Roche protease cocktail inhibitors and 1 mM PMSF), and incubated on ice for 30 min to allow for cell lysis. Cells were passed through a 27 ½ G needle 10 times to maximize lysis and spun down twice at 2655 × g to pellet the nuclear membrane. After BCA quantification of protein concentration, protein concentrations were normalized to the sample with the lowest protein concentration (2–2.5 mg/mL). Pellets were washed twice and combined in 250 μL of hypotonic buffer. The supernatant was ultracentrifuged (Rotor TLA120.2 Beckman Coulter Cat. #343778) at 96000 × g for 1 h at 4°C. The supernatant was the cytoplasmic fraction. The membrane pellet was resuspended in 250 μL of hypotonic buffer, and the sample was passed through a 27 ½ G syringe 10 times, and centrifuged for 45 min at 96000 g at 4°C. The supernatant was combined with the cytoplasmic fraction for a total of 500 μL. The membrane pellet was resuspended in 250 μL of hypotonic buffer, combined with the nuclear fractions and solubilized by adding NP-40 at a final concentration of 0.5%. The combined membrane and nuclear fractions were centrifuged twice at 2655 × g to pellet the nuclear proteins. The supernatant is the cell membrane fraction. The nuclear pellets were combined in 500 μL of hypotonic buffer and sonicated to solubilize. This protocol is summarized in Figure 1B and can be found in detail in Appendix A.

Flow Cytometry

LβT2 cells were grown in complete media (4.5 g/l glucose with sodium pyruvate) supplemented with antibiotics and 10% fetal bovine serum. After serum starvation and insulin and MK-2206 treatment, cells were washed with warm PBS, then harvested with 0.25% trypsin. Cells were washed again with PBS and then resuspended in PBS containing Zombie NIR Live/Dead fixable stain (Biolegend, San Diego, CA, cat# 423105) diluted 1:400. Samples were incubated on ice for 20 min at 4°C protected from light, and washed with FACS buffer (0.1%BSA, 2mM EDTA in 1x PBS). Following the wash with FACS buffer, cells were blocked with TruStain FcX (anti-mouse CD16/32) antibody (Biolegend, RRID: AB_1574975) in FACS buffer and stained for 20 min at 4°C protected from light with rabbit anti-GLUT1 Alexa647 (Abcam, ab195020, 1:50, RRID:AB_2783877) in 50 μL per 1 × 10⁶ cells. Cells were fixed according to protocol (Fixation Buffer, Biolegend, 421801). Single stained Ultracomp eBeads and ArC Amine Reactive Compensation Beads (Thermofisher) were used for compensation samples. Samples were acquired in technical duplicates on a FACSCanto II flow cytometer (BD Biosciences, NJ, USA) and analyzed with FlowJo Software (Treestar, Ashland, OR).

Imaging Flow Cytometry

1 × 10⁶ LβT2 cells in 2 mL of media were seeded for each treatment in a 6 well plate: The following day, the cells were serum starved by replacing the media with serum free media. After 12 to 16 h serum starvation, GnRH groups were treated with 10 nM GnRH, and the remaining groups received PBS as the control. The cells were then trypsinized, and live-dead staining was performed using 1:1600 Zombie green (Biolegend Cat # 423111) in PBS. Cells were then blocked to ensure specific binding of the GLUT1 antibodies with 1:50 TruStain FcX^™ PLUS (anti-mouse CD16/32) Antibody (Biolegend, cat # 156603, RRID: AB_2783137) in FACS buffer (PBS + 0.1% BSA + 2mM EDTA) followed by surface staining for GLUT1 with 1:400 rabbit anti-GLUT1 Alexa647 (Abcam, ab195020, 1:50, RRID:AB_2783877) in FACS buffer. The cells were then fixed with 4% paraformaldehyde and permeabilized with 1x perm/wash buffer according to the manufacturer protocol (Biolegend, cat # 421002). For intracellular staining, anti-GLUT1 was used at a 1:800 dilution. The cells were then washed with 1x perm/wash buffer and resuspended in PBS for imaging flow using the Amnis ImageStreamX Mk II. Imaging flow cytometry was repeated three independent times. Data was analyzed with the IDEAS version 6.0 software.

Signaling Assays

LβT2 cells were seeded in a 12-well plate and allowed to adhere overnight before serum starvation. Serum-starved cells were treated with 1nM insulin (I9278 Sigma) for 30 min at 37°C. For Akt inhibition prior to insulin stimulation, serum-starved cells were preincubated at 37°C with 1μM MK-2206 (Selleck, #S1078) or vehicle for 1 hr and subsequently treated for 30 min with 1nM insulin. Cells were washed with ice cold PBS, then lysed in 2x LSB containing 200 mM dithiothreitol and boiling. Equivalent protein from cell lysates were resolved by SDS-PAGE and immunoblotted.

Immunoblotting

Samples were prepared with LSB containing dithiothreitol and resolved by SDS-PAGE on 9% gels. Proteins were transferred to PVDF membranes. Membranes were blocked for 1 h in 4% BSA and probed for rabbit anti-Na+/K+ ATPase (Cell Signaling Inc, 1:1,000, Cat #3010S, RRID: AB_2060983), mouse anti-P84 (GeneTex, 1:2000, Cat # GTX70220, RRID: AB_372637), mouse anti-GAPDH (GeneTex, 1:50,000, Cat # GTX627408, RRID: AB_11174761), and rabbit anti-GLUT1 (Abcam, 1:200, Cat # ab15309, RRID: AB_301844) overnight at 4°C. Rabbit anti-GLUT1 (Abcam, 1:200, Cat # ab15309, RRID: AB_301844) was previously validated using LβT2 cells and stably transduced LβT2 cells expressing an shRNA targeted against GLUT1 mRNA (scl2a1)¹. HRP-conjugated secondary antibody (Santa Cruz Biotechnology, goat anti-rabbit IgG-HRP: sc-2004, RRID: AB_631746 and goat anti-mouse IgG-HRP: sc-2005. RRID: AB_631736, both at 1:15,000) were used at room temperature for 1 h. Blots were visualized by chemiluminescence and imaged using a Pxi system (Syngene) or film. After imaging, blots were stripped in Restore Western Blot Stripping buffer (Thermo Scientific, Cat# 21059) for 15 min and reblotted with a subsequent primary antibody at 4°C overnight. Quantitative densitometry was performed using Image J software v1.52i (https://imagej.nih.gov/ij/download.html). To ensure accurate results, multiple exposures for all blots developed on film were taken. Quantification was performed on blots with maximal signal without saturation. To determine the relative expression of protein by immunoblot, the raw signal for each marker on a single exposure of a blot was transformed into a percentage of total signal for that protein. These values are then divided by the sample with the highest expression to derive a relative expression value. The relative expression values of experimental markers are normalized to the relative expression values of organelle-specific markers and then presented as fold change over control. For example, membrane GLUT1 is normalized to the internal control of Na+/K+ ATPase. For enrichment calculations, the percentage of total signal for each sample was calculated for each target and expressed as the percentage of signal per target in each subcellular fraction.

Early Endosomal Antigen-1 (EEA1) Immunoprecipitation

Early endosomes from cytoplasmic fractions derived from the OMI fractionation protocol were enriched by adapting the approach described in Gosney et al. 2018¹⁷. Protein concentrations were determined by BCA and normalized. Approximately 0.44μg of EEA1 antibody was added to 400μL of cytoplasmic fraction and allowed to incubate overnight at 4°C. Protein G beads were washed 3 times in PBS and resuspended in fractionation buffer. Next, 10μL of beads were added to each sample and samples were rotated at 4°C for 1 h. The endosomes associated with the anti-EEA1 antibody/protein G complex were pelleted by centrifugation and the supernatant was set aside for analysis by immunoblot. Samples were then washed 3 times by adding fractionation buffer followed by centrifugation. Finally, affinity-purified endosomes (pellet) were eluted by resuspending the beads in 30μL of 2x LSB containing dithiothreitol. For immunoblotting, 15μL of eluted endosome-associated protein was used for analysis.

Extracellular flux analysis

For mitochondrial stress tests, LβT2 cells were seeded on poly-l-lysine coated XF96 plates at 6 × 10⁴ cells per well in HEPES-buffered DMEM supplemented with 10% FBS and antibiotics as described above. Cells were cultured for 24 h then serum starved overnight. Prior to XF analysis LβT2 cells were washed once with extracellular flux (XF) assay media (5 mM HEPES buffered-DMEM containing 10 mM glucose, 4 mM l-glutamine, and 1 mM sodium pyruvate). The media was changed to XF assay media with or without insulin and equilibrated for 1 h before XF analysis. Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured using a modified mitochondrial stress test procedure as follows: basal OCR followed by sequential addition of media or 10 nM GnRH (Injection 1), followed by 2.5 μM oligomycin (Calbiochem) (Injection 2), 500 μM 2,4 Dinotrophenol (DNP) (Sigma-Aldrich) (Injection 3) and 1 μM rotenone + 1 μM antimycin A (Enzo) (Injection 4) with the XF96 Extracellular Flux Analyzer (Agilent). Data were extracted using the Wave software v2.6.1 (Seahorse Bioscience, https://www.agilent.com/en/products/cell-analysis/cell-analysis-software/data-analysis/wave-desktop-2-6) and manually analyzed using JMP software v14.0 and Graphpad Prism 10.

Immunofluorescence

Mouse experiments were performed following an approved protocol in accordance with the UCI Institutional Animal Care and Use Committee regulations. Whole pituitaries were dissected from wild-type C57BL/6 male and female mice at 9–10 weeks of age. Individual pituitaries were dissected into ice-cold PBS and then dispersed individually incubation with 0.25% collagenase Type IV and 0.25% trypsin–EDTA (1x) (Life Technologies) as previously described¹. Primary pituitary cells were seeded at a density of 1.5 × 10⁶ cells per cm² on poly-l-lysine (Sigma-Aldrich Inc.) coated 8 chambered glass slides (Lab-Tek). The cells were cultured overnight at 37°C and 5% CO₂ in high-glucose HEPES-buffered DMEM with 10% FBS, Next cells were serum starved for 12 hr before treatment with or without GnRH for 30 min. Cells were washed in ice cold PBS, then fixed for 15 min with 4% formaldehyde diluted in PBS, followed by a PBS wash. Cells were stained with LH anti-sera raised in guinea pig from the National Hormone Peptide Program (NHPP)¹⁸ at a 1:2000 dilution and anti-GLUT1 described above. Secondary staining for LH was performed with goat anti-guinea pig Alexa Fluor 568 (1:800 dilution, Molecular Probes Cat# A-11075, AB_141954). Slides were mounted with Vectashield mounting medium with 4′, 6-diamidino-2-phenylindole (DAPI). Wide-field fluorescence images were captured at 60x with a Nikon TE2000-U microscope (Nikon America, Melville, NY) equipped with an X-Cite 120PC collimated light source (Lumen Dynamics Group, Mississauga, ON, Canada) and DAPI-1160A, FITC-5050A-000, and mCherry-c filter sets (Semrock,  Rochester, NY) using a CoolSnap EZ monochrome camera (Photometrics, Tucson, AZ) using NIS-Elements software (Nikon America).

Statistics

All experiments were repeated at least three times independently, and reported values are presented as the means ± standard error of the mean. JMP software v14.0 (SAS Institute, https://www.jmp.com/en_us/home.html) was used to perform all statistical analysis. Analysis was performed on raw, normalized values or Box Cox transformed data to correct for heteroscedasticity. Data were evaluated by t-test or ANOVA and appropriate post hoc testing unless otherwise indicated. Specifically, data was analyzed using a mixed linear model with GLUT1 expression as the dependent variable. Both the treatment (with insulin or without insulin) and cell passage were included as nested independent variables. This mixed model allowed for us to account for the response of the cells to insulin by independent experiment. This approach adjusts for the effects of different cell passages and dates of experiments. With this adjusment, a t test was performed considering the nested independent variables. A value of p < 0.05 as indicated was considered significant. Asterisks denote significance from the control. Data sharing a letter are not significantly different from each other, while data marked by letters exclusive to another data point are significantly different from each other. Multivariate modeling demonstrated that individual experiments introduced significant variability. Therefore statisical tests including t-tests and ANOVAs were adjusted to account for that source of variability.

3. Results

3.1. An optimized subcellular fractionation method

To determine redistribution of GLUT and Akt in the gonadotropes, we chose subcellular fractionation coupled with western blotting. Subcellular fraction is a biochemical technique in which cells are disrupted and homogenized by detergents and mechanical homogenization (high-pressure, sonication and/or osmotic shock), followed by separation of cellular components based on their sedimentation coefficients using centrifugation. While subcellular fractionation enables quantitative analysis, caveats remain. Most fractionation methods for purifying specific subcellular compartments produce crude impure fractions, and we found that standard methods of fractionation failed to generate high purity for LβT2 cells^15,16, a mouse gonadotropic cell line. We optimized a subcellular fractionation protocol and demonstrated an improvement in the enrichment of fraction marker proteins using LβT2 cells. As a benchmark, we compared a publicly available Abcam subcellular fractionation protocol (Established protocol, Figure 1A) to our optimized OMI protocol (Figure 1B). The OMI protocol eliminates steps that discard cell material and instead uses a series of recombining fractions and centrifugations to produce improved subcellular fractions (Supplemental Figure 1). Figure 2A shows immunoblot analysis of organelle-specific marker proteins for the nuclear (N), cytosolic (C) and membrane (M) fractions. In immunoblots of nuclear, cytosolic, and membrane fractions from the Established protocol, the nuclear matrix protein marker p84¹⁹, though mostly in the nuclear fraction, also appears as faint bands in the cytosolic and membrane fractions. GAPDH, the cytosolic marker, is present primarily in the cytosolic fraction and most notably, Na⁺/K⁺ ATPase plasma membrane marker appears in the nuclear fraction instead of the intended membrane fraction. In contrast to the Established protocol, the OMI protocol revealed distinct and dense bands of p84 at ~75 kDa in the nuclear fraction, GAPDH at ~ 37 kDa in the cytosolic fraction, and a band at ~ 100 kDa for Na⁺/K⁺ ATPase in the membrane fraction (Figure 2A). Although minimal GAPDH is present in the membrane fraction with the OMI protocol, the high-density bands of appropriate organelle-specific markers in each fraction and the lack of equivalent molecular weight bands in the other fractions indicate little cross contamination overall.

Figure 2.

Optimized fractionation method (OMI Protocol) yields higher enrichment of subcellular markers than the Established Abcam subcellular fractionation protocol. (A) Representative immunoblots of LβT2 cellular fractions using the Established protocol (n=4) and the OMI Protocol (n=3). Na+/K+ ATPase is a marker for the membrane fraction (M), p84 for nuclear (N), and GAPDH for the cytosol (C). (B) The percent of subcellular fraction protein markers in each fraction, quantified from immunoblots. Blue bars are quantified from the Established Protocol (ED) and red bars from the OMI Protocol. The percent of signal from the designated marker in each fraction is presented as mean +/− S.E.M. Data were analyzed by student’s t test after running a mixed model with the independent variables treatment and cell passage nested. An asterisk indicates significance between the ED protocol compared to OMI protocol at P < 0.05.

Quantification of the organelle-specific markers in each fraction was performed to assess enrichment. This analysis indicated that the recovery of proteins expected in the nuclear, cytosolic, and membrane fractions from the OMI protocol were significantly improved over the Established Protocol and was very high, >97%, 92%, and 95%, especially for Na⁺K⁺ ATPase (Figure 2B). Relative to the Established protocol, the OMI protocol minimizes cross-contamination of selected markers for enriched nuclear, cytosolic, and membrane fractions. Further, we found that this protocol was applicable to rat adrenal gland PC12 cells, human epithelial HeLa cells, and human umbilical-vein endothelial derived EA.hy926 cells in addition to LβT2 cells. The Established and OMI protocols produced results similar to the LβT2 cells for PC12 cells (Supplemental Figure 2A). Using the Established Protocol, GAPDH was observed in PC12 cytosolic fractions, whereas Na⁺/K⁺ ATPase appeared in the nuclear fraction instead of the membrane fraction. In HeLa cells, the Established protocol resulted in the presence of GAPDH in both the cytosol and membrane fractions with Na+/K+ ATPase being completely mis-localized to the cytosolic fraction (Supplemental Figure 2B). This contamination produced by the Established protocol was eliminated using the OMI protocol (Supplemental Figure 2B), where Na+/K+ ATPase, p84 and GAPDH were detected predominantly in the proper cellular fractions. With the Established protocol, fractions of the endothelial EA.hy926 cells showed similar patterns of contamination to the LβT2 and PC12 and HeLa cells (Supplemental Figure 2C). In contrast, the OMI protocol of fractionation resulted in high marker enrichment for all EA.hy926 fractions. From this data, we conclude that compared to the Established protocol, our optimized OMI protocol fractionation method reduces cross-contamination of nuclear, cytosolic, and membrane fractions with minimal waste and consistent outcomes across a variety of mammalian cell lines. With these improvements in the Established protocol, the OMI protocol will allow us to accurately study and quantify the trafficking of GLUT1 among cellular compartments in LβT2 cells.

3.2. Insulin regulates the subcellular translocation of GLUT1 in LβT2 cells

Using our optimized OMI Protocol, we examined distribution of GLUT proteins and Akt in LβT2 cells. GnRH from the hypothalamus stimulates gonadotropes in the pituitary to synthesize and secrete LH, a key regulator of reproduction. We have demonstrated that in response to GnRH, gonadotropes mobilize glucose transporter 1 (GLUT1) to the cell surface and increase glycolysis¹. This translocation of GLUT1 to the membrane is required for maximum secretion of LH. Furthermore, insulin, which regulates the shuttling of GLUT4 to the cell surface in insulin sensitive tissues, has an additive effect on GnRH signaling and LH production in gonadotropes depending on context^6,12. In addition, in some cell types, insulin regulates GLUT1²⁰, which is typically insulin independent. Therefore, we hypothesized that insulin may increase the translocation of GLUT1 to the surface of gonadotropes as a mechanism to increase glycolysis, which in turn would support LH production.

To partially test our hypothesis, we used cultured LβT2 cells stimulated with or without insulin and used the OMI protocol to assess the distribution of GLUT1 because GLUT1 supports LH secretion and insulin augments LH secretion from gonadotropes²¹ (Figure 3). Immunoblot detection of GLUT1 in fractions isolated from insulin-treated LβT2 cells revealed that GLUT1 was localized to the cytosol and the membrane as expected under basal conditions (Figure 3A). We found that insulin causes a modest but statistically significant increase in membrane GLUT1 (Figure 3B) as quantified by densitometry. The magnitude of this response is similar to published work quantifying GnRH-induced GLUT1 translocation¹. Standard GnRH-induced redistribution of GLUT1 was visually confirmed both in LβT2 cells and primary gonadotropes (Supplemental Figure 3).

Figure 3.

In response to insulin, GLUT1 is trafficked to the membrane in LβT2 cells. (A) Representative immunoblots of cellular fractions of LβT2 cells probed for GLUT1 in response to 30 min treatment with insulin. (B) Protein expression of GLUT1 derived from immunoblots and normalized to organelle-specific markers (e.g. normalized to p84 for nuclear) and plotted as fold expression relative to no insulin treatment. Data is presented as mean +/− S.E.M.. Data were analyzed by student’s t test after running a mixed model with the independent variables treatment and cell passage nested. An asterisk indicates significance between control and insulin treated for each fraction at P < 0.05. N=3–4.

If insulin does indeed result in increased translocation of glucose transporters to the membrane, we would expect that this would result in the functional outcome of increased glycolysis. To test whether insulin increased glycolysis in LβT2 cells, we performed extracellular flux analysis. As expected and previously published¹, GnRH results in an immediate increase in extracellular acidification rate (ECAR, a proxy for glycolysis) upon addition to the cells and minimal impact on the oxygen consumption rate (OCR, a proxy for oxidative phosphorylation by the mitochondria) (Figure 4A,B). The switch to increased glycolysis is evident by a significant decrease in the OCR:ECAR ratio (Figure 4C). We find that pretreatment with insulin increases basal glycolysis and results in a similar OCR:ECAR ratio as GnRH with no observed additive or inhibitory effect on GnRH-induced glycolysis (Figure 4A–C).

Figure 4.

Insulin induces glycolysis in gonadotropes and GLUT1 translocation is not dependent on Akt. (A,B) OCR (A) and ECAR (B) from mitochondrial stress test extracellular flux profiles for LβT2 cells ± 10nM GnRH, 10 nM insulin or GnRH + insulin. (C) OCR:ECAR ratio derived from the LβT2 mitochondrial stress test profile in (A and B) after GnRH stimulation, analyzed by two-way ANOVA with multiple comparisons. (D) Surface expression of GLUT1 quantified by flow cytometry. LβT2 cells were treated with or without insulin +/− MK-2206 for 30 min then analyzed by flow cytometry. n=3. Data is presented as mean +/− S.E.M.. Data was analyzed by two-way ANOVA with multiple comparisons. The bars labeled with different letters are significantly different from each other, P < 0.05.

Using flow cytometry and surface staining, we confirmed the increase of surface GLUT1 in response to insulin that was observed with the OMI fractionation method (Figure 4D). Insulin is known to induce phosphorylation of Akt downstream of the insulin receptor in adipose and muscle. This is also true for gonadotropes⁶. In insulin-sensitive tissues, phosphorylated Akt induces the translocation of GLUT4 to the cell membrane. We therefore tested whether insulin-induced increases of GLUT1 at the membrane of LβT2 cells were dependent on Akt. We used the Akt allosteric inhibitor MK-2206 that we have previously demonstrated inhibits phosphorylation and activation of Akt in vitro²². Surprisingly, we found no evidence that the insulin-induced increases of membrane GLUT1 were dependent on Akt based on treatment with MK-2206 (Figure 4D). In fact, MK-2206 caused an increase in GLUT1 surface expression when used alone on LβT2 cells. We conclude that GLUT1 is significantly trafficked to the cell surface resulting in increased glycolysis in gonadotropes through unconventional insulin signaling.

3.3. Optimized subcellular fractionation method reveals localization of pAkt in the cytoplasm of LβT2 cells

Although Akt is known to regulate the docking and fusion of GLUT4 vesicles to the plasma membrane, we found that insulin-induced GLUT1 translocation was not impacted in the presence of the Akt inhibitor MK-2206, which targets the closed inactive Akt conformation preventing T308 and S473 phosphorylation²³. Given these findings, we examined insulin-induced phosphorylation of Akt in LβT2 cells using anti-phospho-Akt S473 antibodies (Figure 5A,B). We found that Akt is indeed phosphorylated in response to insulin in LβT2 cells as early as 2 min post stimulation with Akt phosphorylation persisting up to 1 h. We next used the OMI protocol to determine whether we could assess the subcellular localization of insulin-stimulated Akt phosphorylation in LβT2 cells. LβT2 cells were treated with insulin prior to fractionation using our OMI protocol and then fractions were immunoblotted for phosphorylated and total Akt in the various fractions. We found that total Akt was localized to both the cytosol and the plasma membrane. Remarkedly, phosphorylated Akt was primarily associated with the cytosolic fraction (Figure 5C,D) and insulin caused a marked and significant ~1.5 fold increase in phosphorylated Akt.

Figure 5.

Phosphorylated Akt localizes in the cytoplasm, not at plasma membrane or in endosomal fractions in LβT2 cells. (A) Representative immunoblot of insulin-stimulated pAkt time course in LβT2 cells. (b) Quantification of pAkt immunoblots represented in (B). Data is presented as mean +/− S.E.M. (n=4). Significance was determined by ANOVA and post hoc analysis with Dunnett’s comparison to control test on Box-Cox transformed values. Asterisks indicate a significant difference (P< 0.05) from time 0 min. (C) Representative immunoblot of cellular fractions of LβT2 cells probed for Akt and pAkt after 15 min stimulation with insulin. (D) Quantification of pAkt immunoblots from the cytoplasmic fraction represented in (C). Data represents the proportion of total pAkt found in the C fraction after normalization to total Akt and is presented as mean +/− S.E.M. and was analyzed by Student’s t test (n=6), P<0.05. (E) Representative immunoblot demonstrating endosomal markers EEA1 and Rab5 are associated with the cytosol by subcellular fractionation (n=3). (F) Representative immunoblot of subcellular association of pAKT. LβT2 cells were stimulated with insulin for 15 min. The LβT2 cells were then fractionated as in (C) and the cytosolic fractions from LβT2 cells were immunoprecipitated with anti-EEA1 antibodies. The Supernatant (S) left behind after immunoprecipitation and the Pellet (P), the endosomal fraction, were probed by immunoblot, (n=3).

Several groups have shown that phosphorylation of Akt localized in the cytosol occurs in association with endosomal membranes^24–26. To determine whether the phosphorylated Akt we observed in the cytosol is associated with endosomes, we probed fractions from LβT2 cells for endosomal specific markers. Rab5, a member of the Ras-like small GTPase superfamily important for endosome sorting, and early endosome antigen 1 (EEA1), a membrane bound Rab5 effector protein specific to the early endosome, were present only in the cytosolic fraction as expected (Figure 5E). We then immunoprecipitated endosomes from the cytosolic fraction using anti-EEA1 antibodies. The endosomes were pelleted (P), leaving supernatant (S) that was free of early endosomes and depleted of most endosomes as determined by the absence of EEA1 and low expression of Rab5 detected by immunoblotting (Figure 5F). We then probed the separated endosome pellets (P) and endosome “free” supernatants (S) for phosphorylated Akt and found that total Akt was present in both fractions with a higher amount in the endosomal pellets. However, phosphorylated Akt was only associated with the supernatant under both basal and insulin-stimulated conditions (Figure 5F). This indicates that phosphorylated Akt may redistribute from endosomes and preferentially localizes to the cytosol or that phosphorylated Akt in LβT2 is phosphorylated on organelle membranes not immunoprecipitated by EEA1. We conclude that in LβT2 cells, insulin phosphorylates Akt and the resultant phosphorylated Akt is in the cytosolic fraction but is not associated with identifiable endosomes via the OMI protocol. This finding may indicate why insulin-induced trafficking in LβT2 is Akt independent.

Using our improved fractionation protocol, we can clearly distinguish unique translocation behavior of intracellular proteins that would otherwise not be detected. Together, these data show that in gonadotropes, insulin indues GLUT1 translocation in an AKT independent manner despite having intact and functional Akt signaling.

4. Discussion

We have utilized the OMI protocol to assess glucose transporter translocation in the gonadotropic LβT2 cell line. We demonstrated that insulin induces an increase of GLUT1 at the plasma membrane in LβT2 cells. Unexpectedly, we report that blockade of Akt phosphorylation with the Akt inhibitor MK-2206 did not disrupt insulin-induced redistribution of GLUT1 to the plasma membrane, but on its own increased GLUT1 localization to the membrane. This unusual result highlights the interesting nature of glucose transport and insulin signaling in the gonadotrope^1,5,6,12,27, and indicates the need to explore Akt-independent insulin signaling in gonadotropes. Further experiments to validate the inhibition of Akt activity and determine the impact on GLUT1 translocation, especially in primary cells, is merited.

GLUT4 translocation is known to be mediated via Akt-dependent signaling initiated through the insulin receptor, a receptor tyrosine kinase, in insulin sensitive tissues such as muscle and adipose. GLUT1 on the other hand is responsive to glucose concentration to maintain constitutive glucose uptake, especially in low glucose concentrations. However, GLUT1 in the gonadotrope is regulated through activation of the GnRH receptor, a G coupled-protein receptor and a potential Akt-independent mechanism. Specifically, GnRH induces translocation of GLUT1 to the cell surface of gonadotropes but does not induce phosphorylation of Akt^1,6. Although the role of insulin signaling in the gonadotrope is not clear, insulin is known to augment GnRH signaling and induction of luteinizing hormone transcription and translation^6,12. Given the central role of glucose metabolism in gonadotrope function, the insulin receptor and GnRH receptor crosstalk that results in increased GLUT1 expression and/or membrane translocation may be an important factor in the metabolic regulation of reproduction. Independent of Akt, insulin could mediate the accumulation of GLUT1 at the plasma membrane by augmentation of GnRH receptor signaling or through regulation of GLUT1 internalization. In gonadotropes, insulin increases GnRH-induced expression of early growth response-1 (EGR-1), an example of the crosstalk between insulin and GnRH²⁸.

A clear potential mechanism of insulin impact on LβT2 cell GLUT1 translocation could be through TXNIP, a transcription factor that is glucose sensitive. In response to high glucose, TXNIP dose dependently increases²⁹. It blocks production of GLUT1 mRNA while simultaneous promoting GLUT1 endocytosis³⁰. Knockdown and overexpression experiments elegantly demonstrated that lack of TXNIP increases glucose uptake³¹. Given that insulin is a known suppressor of TXNIP, inhibition of TXNIP by insulin in LβT2 cells could be the Akt-independent pathway promoting GLUT1 translocation.

Using the OMI protocol, we assessed phosphorylation of Akt in LβT2 cells. Capitalizing on the ability to use proteins for additional assays downstream of the OMI protocol, we found minimal association of phosphorylated Akt with endosomes by separating endosomes from the cytosolic fractions via immunoprecipitation. Canonically, inactive Akt is recruited to the plasma membrane and phosphorylated. Once phosphorylated, Akt leaves the plasma membrane and acts on cytosolic and potentially nuclear substrates. Recent studies challenge this paradigm, suggesting that Akt needs to maintain interaction with inositol phospholipids at the plasma, lysosomal, endosomal, or other endomembrane^24–26,32. Indeed, studies show that Akt is rapidly dephosphorylated after leaving cellular membranes. In contrast, our results support a different paradigm. We show that in LβT2 cells treated with insulin, phosphorylated Akt appears in the cytoplasmic fraction (Figure 5C), and the Akt that is associated with the endomembrane fraction of the cell lysate is not phosphorylated (Figure 5F). The observation that phosphorylated Akt is found in the cytoplasmic fraction 15 min after insulin treatment, combined with our observation that Akt remains phosphorylated for at least 1 h after insulin treatment of LβT2 cells, suggests that in this system, active Akt may not significantly remain associated with endomembranes. While this appears to contradict recent reports, these experiments did not measure Akt activity nor whether Akt is continually being phosphorylated over time. Additionally, it is also possible that Akt is phosphorylated on endosomes and then released into the cytosol¹³. While our observations suggest that Akt could still be phosphorylated and active while not associated with cellular membranes, further investigation is required to verify these initial observations. Alternatively, these observations may explain why the canonical insulin signaling pathway in gonadotropes is not intact.

In conclusion, we found two new observations in LβT2 cells that were made possible by an optimized subcellular OMI fractionation protocol. These findings include that in this cell line, insulin regulates GLUT1 protein localization and that phosphorylated Akt may not be associated with endosomal membranes. Confirmation of these mechanisms in primary gonadotropes will provide a basis for understanding the observed clinical impacts of hyperinsulinemia on gonadotropin production.

Supplementary Material

Figure S2

Figure S2. Optimized fractionation method is generalizable to other cell lines.

Figure S3

Figure S3. GnRH induces membrane translocation of GLUT1 in LβT2 and primary female gonadotropes.

Figure S1

Figure S1. Established protocol results in waste of cellular material.

Appendix A. OMI Subcellular Fractionation Protocol

A. Reagents

Name	Company	Catalog Number
Phosphate Buffered Saline (PBS)	Gibco	14190–144
Pierce Protein Kit	Thermo Scientific	23225
Bovine Serum Albumen	Thermo Scientific	23210
Nonidet™ P 40 Substitute solution	Sigma Millipore	98379
Protease Inhibitor Roche	Roche	11697498001
Phenylmethylsulfonyl Fluoride (PMSF)	Calbiochem	52332

B. Recipes

Stock Hypotonic Buffer*

*Buffer can be stored at 4°C. For each experiment, make fresh working hypotonic buffer.

Stock Hypotonic Buffer
Reagent	Stock Concentration	Volume	Final Concentration
Tris-HCl, pH 7.4	1M	1 mL	20 mM
NaCl	5M	0.5 mL	10 mM
MgCl2	1M	150 μL	3 mM
ddH2O	N/A	48.35 mL
Final Volume = 50 mL

Working Hypotonic Buffer*

*Only the working hypotonic buffer will be used during experimentation, not stock hypotonic buffer.

Working Hypotonic Buffer
Reagent		Volume	Final Concentration
Hypotonic Solution		9.9 mL	N/A
Protease inhibitors Roche		1 tablet	N/A
PMSF		100 μL	100 mM
Final Volume = 10 mL

C. Preparation

• Cool down centrifuges to 4°C.

• Make 10 mL working hypotonic buffer.

• Keep working hypotonic buffer cool on ice.

• Make 10% NP-40 solution.

D. Protocol

General Notes

• 5 to 10 million adherent cells (10 cm2 dish) are required for sufficient protein yield.

• Keep cells on ice throughout the entirety of the protocol.

• Read through the entire protocol before beginning.

Cell Harvest and Lysing

1. Gently wash cells with 5mL of cold PBS 2 times, aspirating PBS after each wash.

2. After second PBS wash, add 200 μL of PBS.

3. Scrape cells with cell scraper and collect in pre-chilled and pre-labeled Eppendorf tube.

4. Add another 200 μL of PBS to the plate, scrape cells, collect, and combine into the same Eppendorf tube.

5. Spin down cells at 4°C at 105 × g for 3 min.

6. Aspirate PBS and resuspend the pellet in 250 μL of working hypotonic buffer.

7. Vortex for 10 sec and incubate on ice for 30 min.

8. Homogenize using a 27G syringe at least 10 times to ensure the cells are lysed.

BCA and Normalization

1. Perform a BCA assay using your preferred method of protein quantification. We utilize the Pierce BCA Protein Assay.

2. Utilize a plate reader to detect the colorimetric change of the samples. Calculate concentration from the BCA assay by calculating the protein concentration through a BSA standard curve manually or via the protein concentration calculation software.

3. Normalize all samples in the range of 2.0 mg/mL to 2.5 mg/mL by adding equal amounts of protein to new Eppendorf tubes in a final volume of 250 μL of hypotonic buffer.

4. Optional. This is a possible stopping point. Samples can be stored at 4°C until next day.

Separation of Cytosolic Fraction from Membrane and Nuclear Components

1. Centrifuge the normalized homogenate for 10 min at 2655 × g at 4°C.

2. Save the pellet and supernatant in different tubes. The pellets contain the nuclear fraction and the supernatant contains mitochondria, membrane and cytosolic proteins.

3. Centrifuge the supernatant from step 2 for 10 min at 2655 × g at 4°C.

4. Save the pellet and save supernatant in different tubes. As in step 2, the pellets contain the nuclear fraction and the supernatant contains mitochondria, membrane and cytosolic proteins.

5. Transfer the supernatant from step 4 into high speed-centrifugation tubes.

6. Weigh the high-speed ultracentrifugation tubes and their contents. Use the weight to prepare balance tubes.

7. Centrifuge the supernatants for 1 h at 96000 × g at 4°C in an ultracentrifuge.

8. Without disrupting the pellet, gently wash the pellets from previous steps 2 and 4 twice with 250 μL of hypotonic working buffer. After the second wash, add 250μL to one pellet, resuspend and then transfer the resuspended pellet to the second tube containing a pellet. Resuspend to combine the pellets from steps 2 and 4 in a total volume of 250μL.*These steps can be completed during the 1 h of ultracentrifugation.*These resuspended pellets contain nuclear proteins so save on ice for further processing.

9. Transfer supernatants from the ultracentrifuge to a new Eppendorf tube. This supernatant is your cleared cytosolic fraction.

10. Resuspend the remaining membrane pellet in 250 μL of hypotonic buffer and pass through 27G syringe 10 times to homogenize. *This step maximizes the yield of membrane proteins and increases recovery of cytosolic proteins.

11. With the homogenized pellet, repeat Steps 6 and 7.

12. Add the supernatant from the repeated step 7 to the Eppendorf tube with the cytosolic fraction from step 9 for a total volume of 500μL. This is the final cytosolic fraction. Use the remaining membrane pellet in the next section.

Separation of Membrane Fraction from Nuclear Fraction

1. Resuspend the membrane pellet in 225 μL of hypotonic buffer and combine with the resuspended nuclear pellet for a total of 475μL.

2. Add 25 μL of 10% NP40.

3. Vortex for 10 sec.

4. Incubate samples for 30 min on ice.

5. Centrifuge for 10 min at 2655 × g at 4°C.

6. Transfer the supernatant to a new Eppendorf tube and save the pellet. The pellet is your nuclear fraction.

7. Centrifuge the supernatant a second time for 10 min at 2655 × g at 4°C.

8. Collect the supernatant and save as final membrane fraction. Save the pellet. This is your 2^nd nuclear pellet. *In this tube, the pellet will be very small, so handle the sample carefully.

9. Gently wash the remaining pellets 3 times with hypotonic buffer, discarding the buffer each time.

10. Combine the pellets in a total of 500 μL of hypotonic buffer and save as the final nuclear fraction.

11. Sonicate the membrane and nuclear fractions at 10% amplitude for 10 sec.