Pancreatic Cancer Cell-Conditioned, Human-Derived Primary Myotubes Display Increased Leucine Turnover, Increased Lipid Accumulation, and Reduced Glucose Uptake

Metabolic alterations occurring in cancer cells have been seen to also occur in other tissues than cancerous tissue. For instance, cachexia, peripheral insulin resistance, or both are commonly seen in patients with cancer. We explored differences in substrate use in myotubes conditioned with the medium from a pancreatic cancer cell line, PANC-1, or primary human pancreatic cells, hPECs. Protein turnover was assessed using scintillation proximity assay, glucose and oleic acid handling were analyzed by substrate oxidation assay. We performed qPCR to study gene expression and immunoblotting and proteomic analyses to study protein expression. PANC-1-conditioned myotubes had an imbalance in protein turnover with decreased accumulation, increased decay, and decreased MYH2 gene expression. Glucose uptake decreased despite increased insulin-stimulated Akt phosphorylation. Fatty acid uptake increased, whereas fatty acid oxidation was unchanged, leading to accumulation of intracellular lipids (TAG) in PANC-1-conditioned myotubes. Secretome analyses revealed increased release of growth factors and growth factor receptor from PANC-1 cells, potentially affecting muscle cell metabolism. Myotubes exposed to pancreatic cancer cell medium displayed altered energy metabolism with increased protein/leucine turnover and lipid accumulation, while glucose uptake and oxidation reduced. This indicates production and release of substances from pancreatic cancer cells affecting skeletal muscle.


Introduction
Cancer causes great morbidity and mortality and is a major public health concern worldwide. Pancreatic cancer, with pancreatic ductal adenocarcinoma (PDAC) being the most common type, is one of the cancers with the poorest prognosis with a 5-year survival rate of 5% [1].
Metabolic alterations are well-known occurrences with cancer. Cancer cells have higher uptake of glucose and produce more lactic acid, even in the presence of oxygen, compared to normal cells (the Warburg effect) [2]. Alterations in fatty acid metabolism in cancer cells are also increasingly more recognized [2]. Metabolic changes have, however, also been seen to occur in other tissues than the cancer tissue itself.

Ethical Approval
Skeletal muscle biopsies were obtained after written informed consent and approval from the Regional Committees for Medical and Health Research Ethics (REK) North, Tromsø, Norway (ref. no. 2011/882). The study adhered to the Code of Ethics of the World Medical Association (Declaration of Helsinki).

PANC-1 Cells
PANC-1 cells were cultured in DMEM high glucose (25 mmol/L) supplemented with 10% FBS (PANC-1 medium) according to the manufacturer's instructions. Cells were kept in 75 cm 2 Nunc™ flasks before being cultured for experiments and then kept in a humidified in a 5% CO 2 atmosphere at 37 • C, and the medium was changed every 2-3 days.

hPECs
hPECs were cultured in epithelial cell medium with supplements according to the manufacturer's instructions (hPEC medium). Cells were kept in 75 cm 2 Nunc™ flasks before being cultured for experiments and then kept in a humidified 5% CO 2 atmosphere at 37 • C, and the medium was changed every 2-3 days. Experiments were performed on cells in passages 6-8.

Skeletal Muscle Cells
Biopsies were obtained from the musculus vastus lateralis of healthy male donors after obtaining written informed consent and ethical approval. To establish multinucleated human myotubes, satellite cells were first proliferated as myoblasts, as previously described [10], in DMEM-Glutamax TM (5.5 mmol/L glucose) medium supplemented with 10% FBS, 10 ng/mL of hEGF, 0.39 µg/mL of dexamethasone, and 0.05% BSA. At approximately 80% confluence, the medium was changed to DMEM-Glutamax™ (5.5 mmol/L glucose) supplemented with 2% FBS and 25 pmol/L of insulin to initiate differentiation into multinucleated myotubes for 7 days, in a humidified 5% CO 2 atmosphere at 37 • C, and the medium was changed every 2-3 days. Experiments were performed on cells in passages 3-4.

PANC-1-and hPEC-Conditioned Myotubes
Due to different proliferation rates, an unequal number of PANC-1 cells and hPECs were seeded to achieve an approximately equal cell number at the time of harvesting. PANC-1 cells (~1 × 10 6 cells/flask) and hPECs (~2 × 10 6 cells/flask) were cultured in respective media in 75 cm 2 flasks for 24 h. The cells were washed, and both cell types were cultured in DMEM high glucose (PANC-1 medium without FBS). After 48 h, the media were collected and spun down at 1000 rpm/4 • C/15 min, and the supernatants were frozen at −80 • C. Glucose and lactate levels were measured in the collected media, as described [10], and the glucose concentration was adjusted to be the same in both media (approximately 22 mmol/L). Human skeletal muscle cells were cultured and differentiated in cell culture plates. In the last 4 days of differentiation, the cells were conditioned with the medium collected from hPECs (chPEC) or PANC-1 cells (cPANC-1). The conditioned medium was added in a ratio of 50:50 with DMEM-Glutamax™ (5.5 mmol/L glucose) supplemented with 2% FBS and 25 pmol/L of insulin.

Scintillation Proximity Assay (SPA)
Myotubes were cultured in a 96-well ScintiPlate. After differentiation and conditioning, the myotubes were exposed to [ 14 C] leucine (37 kBq/mL, 0.8 mmol/L), followed by measurements of protein (leucine) accumulated in the cells by scintillation proximity assay (SPA) [11]. [ 14 C] leucine was monitored over the next 24 h (measurements at 0, 2, 4, 6, 8, and 24 h) using a 2450 MicroBeta 2 scintillation counter (PerkinElmer). Thereafter, the medium was changed to DMEM with 0.8 mmol/L of leucine, 10 mmol/L of HEPES, 0.5% BSA, and 0.1 mmol/L of glucose, and the decay of [ 14 C] leucine was monitored over 6 h (measurements at 0, 2, 4, and 6 h). The amount of radioactivity in the cells was related to the total cell protein content measured according to Bradford [12].

Substrate Oxidation Assay
Skeletal muscle cells were cultured in 96-well CellBIND ® microplates. After differentiation and conditioning, D-[ 14 C(U)] glucose (18.5 kBq/mL, 200 µmol/L) or [1-14 C] oleic acid (18.5 kBq/mL, 100 µmol/L) was added during 4 h of CO 2 trapping [11]. CO 2 production was measured in the presence or absence of 1 µmol/L of FCCP or 0.5 µmol/L of rotenone in combination with 2.5 µmol/L of antimycin A or 1 µmol/L of oligomycin. CO 2 production and cell-associated radioactivity (CA) were assessed using a 2450 MicroBeta 2 scintillation counter (PerkinElmer). The amount of radioactivity in the cells was related to the total cell protein content measured according to Bradford [12]. The sum of 14 CO 2 and CA was considered as the total substrate uptake.

Thin-Layer Chromatography and Measurement of Acid-Soluble Metabolites (ASM)
Skeletal muscle cells were cultured in 12-well CellBIND ® microplates. After differentiation and conditioning, the myotubes were incubated with [1-14 C] oleic acid (18.5 kBq/mL, 100 µmol/L) for 4 h. Thereafter, the cells were washed twice with PBS and harvested with 250 µL of 0.1% SDS per well. Cellular lipids were analyzed by extraction of the homogenized cell fraction, separation of lipids by thin-layer chromatography, and quantification by liquid scintillation [13]. A non-polar solvent mixture of hexane:ether:acetic acid (65:35:1) was used to separate the lipids. The amount of neutral lipids was related to the total cell protein concentration determined by Pierce protein assay.
Measurement of acid-soluble metabolites (ASM), reflecting incomplete fatty acid oxidation (β-oxidation) and mainly consisting of tricarboxylic acid cycle metabolites, was performed, as described and modified by Bakke et al. [14]. In short, 100 µL of the radiolabeled medium was transferred to an Eppendorf tube and precipitated with 300 µL of cold HClO 4 (1 mol/L) and 30 µL of BSA (6%). Thereafter, the tube was centrifuged at 10,000 rpm/10 min/4 • C before 200 µL of the supernatant was measured by liquid scintillation of a Packard Tri-Carb 1900 TR (PerkinElmer, Boston, MA, USA).

Glycogen Synthesis
Differentiated and conditioned cells, cultured in 6-well CellBIND ® plates, were starved for 1.5 h in glucose-free DMEM before being exposed to serum-free DMEM (5.5 mmol/L glucose) supplemented with D-[ 14 C(U)] glucose (18.5 kBq/mL, 0.67 mmol/L), in the absence or presence of 20 nmol/L or 100 nmol/L of insulin for 3 h. The cells were washed twice with PBS and harvested in 1 mol/L of KOH. Protein content was determined by Pierce protein assay before 20 mg/mL of glycogen and more KOH (final concentration 4 mol/L) were added to the samples. The incorporation of D-[ 14 C(U)] glucose into glycogen was measured [15].

Immunoblotting
Myotubes were cultured in 25 cm 2 flasks and harvested in Laemmli buffer (0.5 mol/L of Tris-HCl, 10% SDS, 20% glycerol, 10% β-mercaptoethanol, and 5% bromophenol blue). Proteins were electrophoretically separated on 4-20% Mini-Protean TGX™ gels with Tris/glycine buffer (pH 8.3), followed by blotting to nitrocellulose membranes and incubation with antibodies. For total Akt and Akt phosphorylation, immunoreactive bands were visualized with enhanced chemiluminescence (Chemidoc XRS, Bio-Rad, Copenhagen, Denmark). and quantified using Image Lab software (version 6.0.1). The level of Akt phosphorylated at serine 473 was normalized for an internal control (HEK cell-line sample) and the total Akt level on the same gel. Immunoreactive bands of total AMPK and AMPK phosphorylation were visualized using Image Studio Lite (version 5.2). The level of AMPK phosphorylated at threonine 172 was normalized for an internal control (HEK cell-line sample) and the total AMPK level on the same gel.

RNA Isolation and Analysis of Gene Expression by qPCR
Skeletal muscle cells were cultured in 6-well CellBIND ® plates. After differentiation and conditioning, total RNA was isolated using the QIAGEN RNeasy Mini Kit and reversetranscribed on a PerkinElmer 2720 Thermal Cycler (25 • C/10 min, 37 • C/80 min, and 85 • C/5 min) using a High-Capacity cDNA Reverse Transcription Kit and TaqMan reverse transcription kit reagents. Primers (Supplementary Table S1) were designed using Primer Express ® (Thermo Fisher Scientific, Waltham, WA, USA). qPCR was performed using the StepOnePlus Real-Time PCR system (Thermo Fisher Scientific, Waltham, WA, USA). Target genes were quantified in duplicate, carried out in a 25 µL reaction volume. All assays were run for 44 cycles (95 • C/15 s followed by 60 • C/60 s). Two housekeeping genes were examined, glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and ribosomal protein lateral stalk subunit P0 (RPLP0). Gene expression levels were comparable regardless of the choice of the housekeeping gene; data shown are presented relative to RPLP0. Forward and reverse primers were used at a concentration of 30 µmol/L.

Proteomics
Re-analysis of proteomic data from a previous analysis [10] was performed. Mass spectrometry proteomics data were deposited to the ProteomeXchange Consortium via the PRIDE partner repository (dataset identifier PXD017613). STRING software (version 11.5) was applied to identify functional protein association networks using annotated key words: "growth factors", "growth factor binding proteins", and "cytokines".

Presentation of Data and Statistics
All values are reported as means ± SEM. The value n represents the number of individual experiments, each with at least duplicate observations. Statistical analyses were performed using GraphPad Prism (version 8.3.0) for Windows or SPSS Statistics (version 28). The unpaired t-test was used to evaluate the effects on myotubes conditioned with either hPEC or PANC-1 medium. The paired t-test was used to evaluate the effects of treatment on myotubes conditioned with the medium from either of the two cell types. A p-value of <0.05 was considered statistically significant.

Measurement of Leucine Incorporation in Myotubes Indicated Decreased Protein Synthesis, Increased Protein Decay, and Lower mRNA Expression of MYH2 after Being Conditioned with the Medium from Pancreatic Cancer Cells
As cancer-induced cachexia leads to increased degradation of protein in skeletal muscle [4], we wanted to examine the protein content in myotubes after conditioning with the medium from hPECs (chPEC exposed) or PANC-1 cells (cPANC-1 exposed). We found that the overall protein content significantly decreased after 4 days of conditioning with cPANC-1 and chPEC media ( Figure 1A). To further explore the difference seen in the protein content, we traced the cellular accumulation of [ 14 C] leucine as a measure of protein synthesis in myotubes and thereafter evaluated the decay of radioactivity from the cells. The accumulation of leucine over 24 h was significantly less for cPANC-1exposed myotubes compared to chPEC-exposed cells ( Figure 1B,C). Further, the decay of radioactivity measured over 6 h increased for cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 1D,E).
As cancer-induced cachexia leads to increased degradation of protein in skeletal muscle [4], we wanted to examine the protein content in myotubes after conditioning with the medium from hPECs (chPEC exposed) or PANC-1 cells (cPANC-1 exposed). We found that the overall protein content significantly decreased after 4 days of conditioning with cPANC-1 and chPEC media ( Figure 1A). To further explore the difference seen in the protein content, we traced the cellular accumulation of [ 14 C] leucine as a measure of protein synthesis in myotubes and thereafter evaluated the decay of radioactivity from the cells. The accumulation of leucine over 24 h was significantly less for cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 1B,C). Further, the decay of radioactivity measured over 6 h increased for cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 1D,E).

Figure 1.
Incorporation and decay of leucine in myotubes after conditioning with medium collected from hPECs or PANC-1 cells. Myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed) during the last 4 days of differentiation. (A) Protein content was measured using Bradford protein assay. Muscle cells were incubated with [ 14 C] leucine (37 kBq, 0.8 mmol/l), and (B) cellular accumulation of leucine was measured over 24 h by scintillation proximity assay (SPA). (C) Overall accumulation of [ 14 C] leucine was calculated as the area under the curve. (D) The medium was then changed to label-free medium, and decay of radioactivity was measured over 6 h by SPA. (E) Overall decay over 6 h was calculated as the area under the curve. Data are presented as the mean ± SEM from four individual experiments, each with three replicates (n = 12). * Statistically significant for cPANC-1-exposed myotubes versus chPEC-exposed myotubes (p ≤ 0.05, SPSS linear mixed-model analysis).
The mRNA expression of selected genes important for protein synthesis and cachexia was also measured. The mRNA expression of ATP hydrolyzing motor protein myosin heavy chain 2 (MYH2) significantly reduced in cPANC-1-exposed myotubes compared to chPEC-exposed cells, while no changes were found for the serine-threonine kinase mech- Figure 1. Incorporation and decay of leucine in myotubes after conditioning with medium collected from hPECs or PANC-1 cells. Myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed) during the last 4 days of differentiation. (A) Protein content was measured using Bradford protein assay. Muscle cells were incubated with [ 14 C] leucine (37 kBq, 0.8 mmol/L), and (B) cellular accumulation of leucine was measured over 24 h by scintillation proximity assay (SPA). (C) Overall accumulation of [ 14 C] leucine was calculated as the area under the curve. (D) The medium was then changed to label-free medium, and decay of radioactivity was measured over 6 h by SPA. (E) Overall decay over 6 h was calculated as the area under the curve. Data are presented as the mean ± SEM from four individual experiments, each with three replicates (n = 12). * Statistically significant for cPANC-1-exposed myotubes versus chPEC-exposed myotubes (p ≤ 0.05, SPSS linear mixed-model analysis).
The mRNA expression of selected genes important for protein synthesis and cachexia was also measured. The mRNA expression of ATP hydrolyzing motor protein myosin heavy chain 2 (MYH2) significantly reduced in cPANC-1-exposed myotubes compared to chPEC-exposed cells, while no changes were found for the serine-threonine kinase mechanistic target of rapamycin (MTOR), mitogen-activated protein kinase 8 (MAPK8), the myogenic regulatory factor myoblast determination protein 1 (MYOD1), or the immune response regulator nuclear factor kappa B (NFκB); see Figure 2. anistic target of rapamycin (MTOR), mitogen-activated protein kinase 8 (MAPK8), the myogenic regulatory factor myoblast determination protein 1 (MYOD1), or the immune response regulator nuclear factor kappa B (NFκB); see Figure 2. Expression of genes related to protein turnover in myotubes after conditioning with medium collected from hPECs or PANC-1 cells. Myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed) during the last 4 days of differentiation. mRNA was isolated, and the expression of myosin heavy chain 2 (MYH2), mechanistic target of rapamycin (MTOR), myoblast determination protein 1 (MYOD1), mitogen-activated protein kinase 8 (MAPK8), and nuclear factor kappa B (NFκB) was assessed. The expression of each gene was normalized for the expression of the housekeeping gene ribosomal protein lateral stalk subunit P0 (RPLP0). Data are presented as the mean ± SEM from two individual experiments, with three replicates each (n = 6) relative to expression in chPEC-exposed cells. * Statistically significant in cPANC-1-exposed myotubes versus chPEC-exposed myotubes (p ≤ 0.05, unpaired t-test).

Myotubes Conditioned with PANC-1 Culture Medium Had Lower Glucose Uptake and Oxidation Compared to hPEC-Conditioned Myotubes
The link between pancreatic cancer and peripheral insulin resistance is not well understood. We wanted to explore whether conditioning skeletal muscle cells with media from hPECs and PANC-1 cells influences glucose metabolism. We found that cPANC-1exposed myotubes had lower uptake and oxidation of [ 14 C] glucose compared to cells exposed to chPEC medium ( Figure 3A,B).  Expression of genes related to protein turnover in myotubes after conditioning with medium collected from hPECs or PANC-1 cells. Myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed) during the last 4 days of differentiation. mRNA was isolated, and the expression of myosin heavy chain 2 (MYH2), mechanistic target of rapamycin (MTOR), myoblast determination protein 1 (MYOD1), mitogen-activated protein kinase 8 (MAPK8), and nuclear factor kappa B (NFκB) was assessed. The expression of each gene was normalized for the expression of the housekeeping gene ribosomal protein lateral stalk subunit P0 (RPLP0). Data are presented as the mean ± SEM from two individual experiments, with three replicates each (n = 6) relative to expression in chPEC-exposed cells. * Statistically significant in cPANC-1-exposed myotubes versus chPEC-exposed myotubes (p ≤ 0.05, unpaired t-test).

Myotubes Conditioned with PANC-1 Culture Medium Had Lower Glucose Uptake and Oxidation Compared to hPEC-Conditioned Myotubes
The link between pancreatic cancer and peripheral insulin resistance is not well understood. We wanted to explore whether conditioning skeletal muscle cells with media from hPECs and PANC-1 cells influences glucose metabolism. We found that cPANC-1exposed myotubes had lower uptake and oxidation of [ 14 C] glucose compared to cells exposed to chPEC medium ( Figure 3A,B).  Expression of genes related to protein turnover in myotubes after conditioning with medium collected from hPECs or PANC-1 cells. Myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed) during the last 4 days of differentiation. mRNA was isolated, and the expression of myosin heavy chain 2 (MYH2), mechanistic target of rapamycin (MTOR), myoblast determination protein 1 (MYOD1), mitogen-activated protein kinase 8 (MAPK8), and nuclear factor kappa B (NFκB) was assessed. The expression of each gene was normalized for the expression of the housekeeping gene ribosomal protein lateral stalk subunit P0 (RPLP0). Data are presented as the mean ± SEM from two individual experiments, with three replicates each (n = 6) relative to expression in chPEC-exposed cells. * Statistically significant in cPANC-1-exposed myotubes versus chPEC-exposed myotubes (p ≤ 0.05, unpaired t-test).

Myotubes Conditioned with PANC-1 Culture Medium Had Lower Glucose Uptake and Oxidation Compared to hPEC-Conditioned Myotubes
The link between pancreatic cancer and peripheral insulin resistance is not well understood. We wanted to explore whether conditioning skeletal muscle cells with media from hPECs and PANC-1 cells influences glucose metabolism. We found that cPANC-1exposed myotubes had lower uptake and oxidation of [ 14 C] glucose compared to cells exposed to chPEC medium ( Figure 3A,B).  . Impact of conditioning myotubes with medium from hPECs or PANC-1 cells on glucose metabolism. In the last 4 days of differentiation, myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed). Thereafter, myotubes were incubated with 200 µmol/L of D-[ 14 C(U)]glucose (18.5 kBq/mL) for 4 h before glucose uptake and oxidation were measured. (A) Uptake was calculated as captured CO 2 + cell-associated (CA) [ 14 C] glucose. (B) Complete oxidation was measured as trapped CO 2 . Data are presented as the mean ± SEM from three individual experiments, each with eight replicates (n = 24). * Statistically significant in cPANC-1-exposed myotubes versus chPECexposed cells (p ≤ 0.05, unpaired t-test).

Akt Phosphorylation Was Higher under Basal Conditions and When Treated with Insulin in Myotubes Conditioned with PANC-1 and hPEC Culture Media
Insulin-induced glycogen synthesis was not significantly different between chPECexposed myotubes and cPANC-1-exposed cells ( Figure 4A). There was an insignificant reduction in the mRNA expression of the insulin-dependent glucose transporter SLC2A4 (p = 0.079) in cPANC-1-exposed cells compared to chPEC-exposed myotubes ( Figure 4B). To further examine the response of insulin in these cells, we assessed Akt phosphorylation at serine 473 in cells treated with or without insulin for 15 min. Akt phosphorylation was higher in cPANC-1-exposed myotubes compared to chPEC-exposed myotubes, both under basal conditions and after insulin treatment ( Figure 4C,D).

Akt Phosphorylation Was Higher under Basal Conditions and When Treated with Insulin in Myotubes Conditioned with PANC-1 and hPEC Culture Media
Insulin-induced glycogen synthesis was not significantly different between chPECexposed myotubes and cPANC-1-exposed cells ( Figure 4A). There was an insignificant reduction in the mRNA expression of the insulin-dependent glucose transporter SLC2A4 (p = 0.079) in cPANC-1-exposed cells compared to chPEC-exposed myotubes ( Figure 4B). To further examine the response of insulin in these cells, we assessed Akt phosphorylation at serine 473 in cells treated with or without insulin for 15 min. Akt phosphorylation was higher in cPANC-1-exposed myotubes compared to chPEC-exposed myotubes, both under basal conditions and after insulin treatment ( Figure 4C,D). . Impact of conditioning myotubes with medium from hPECs or PANC−1 cells on insulin responses. Myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed) during the last 4 days of the differentiation period. (A) After starving myotubes for 90 min in a medium containing no glucose, insulin-stimulated glycogen synthesis was measured as the incorporation of D-[ 14 C(U)]glucose (18.5 kBq/mL, 0.67 mmol/L) into glycogen in the presence or absence of 100 nmol/L of insulin for 3 h. (B) mRNA expression of solute carrier family 2 member 4 (SLC2A4) was normalized for the housekeeping gene ribosomal protein lateral stalk subunit P0 (RPLP0), and the expression in cPANC-1-exposed myotubes is presented relative to chPEC-exposed cells (n = 3-6). (C) Immunoblot showing Akt phosphorylated at serine 473 (pAkt) and total Akt and in cells treated with or without 100 nmol/L of insulin for 15 min. (D) Quantified immunoblot showing the pAkt/total Akt ratio, adjusted to the internal control (HEK cells) and presented as a percentage of basal chPEC (n = 4). Data in (A,B,D) are presented as the mean ± SEM. * Statistically significant in cPANC-1-exposed myotubes versus chPEC-exposed myotubes (p ≤ 0.05, unpaired t-test). . Impact of conditioning myotubes with medium from hPECs or PANC−1 cells on insulin responses. Myotubes were conditioned with the medium from human pancreatic epithelial cells (chPEC exposed) or a human pancreatic cancer cell line (cPANC-1 exposed) during the last 4 days of the differentiation period. (A) After starving myotubes for 90 min in a medium containing no glucose, insulin-stimulated glycogen synthesis was measured as the incorporation of D-[ 14 C(U)]glucose (18.5 kBq/mL, 0.67 mmol/L) into glycogen in the presence or absence of 100 nmol/L of insulin for 3 h. (B) mRNA expression of solute carrier family 2 member 4 (SLC2A4) was normalized for the housekeeping gene ribosomal protein lateral stalk subunit P0 (RPLP0), and the expression in cPANC-1-exposed myotubes is presented relative to chPEC-exposed cells (n = 3-6). (C) Immunoblot showing Akt phosphorylated at serine 473 (pAkt) and total Akt and in cells treated with or without 100 nmol/L of insulin for 15 min. (D) Quantified immunoblot showing the pAkt/total Akt ratio, adjusted to the internal control (HEK cells) and presented as a percentage of basal chPEC (n = 4). Data in (A,B,D) are presented as the mean ± SEM. * Statistically significant in cPANC-1-exposed myotubes versus chPEC-exposed myotubes (p ≤ 0.05, unpaired t-test).

Myotubes Exposed to PANC-1-Conditioned Medium Showed Increased Lipid Accumulation and Responded Less to Electron Chain Complex Inhibitors Compared to hPEC-Conditioned Myotubes
We previously showed that PANC-1-conditioned myotubes had reduced ability to oxidize energy substrates compared to non-conditioned myotubes [10]. In this study, we found no difference in incomplete fatty acid oxidation measured as ASM ( Figure 5A) or basal complete fatty acid oxidation to CO 2 ( Figure 5B) in cPANC-1-exposed compared to chPEC-exposed myotubes. Mitochondrial dysfunction is a common feature of cancer cachexia, which occurs when the mitochondria work less efficiently than they should due to another disease or condition, such as cachexia [4] or aging [16]. We explored the mitochondrial responsiveness by using electron transport chain (ETC) inhibitors. Treatment with oligomycin or a combination of rotenone and antimycin A revealed that cPANC-1exposed myotubes were less sensitive to the inhibitors; thus, the oxidation of oleic acid after inhibitor exposure was higher in cPANC-1-exposed cells compared to chPEC-exposed myotubes ( Figure 5B). Further, total lipid accumulation from oleic acid and the level of free oleic acid (FFA) were higher in cPANC-1-exposed cells compared to chPEC-exposed cells ( Figure 5C,D). Incorporation of oleic acid into free fatty acids (FFA), triacylglycerol (TAG), and cholesterol ester (CE) was also higher in cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 5D).

Myotubes Exposed to PANC-1-Conditioned Medium Showed Increased Lipid Accumulation and Responded Less to Electron Chain Complex Inhibitors Compared to hPEC-Conditioned Myotubes
We previously showed that PANC-1-conditioned myotubes had reduced ability to oxidize energy substrates compared to non-conditioned myotubes [10]. In this study, we found no difference in incomplete fatty acid oxidation measured as ASM ( Figure 5A) or basal complete fatty acid oxidation to CO2 ( Figure 5B) in cPANC-1-exposed compared to chPEC-exposed myotubes. Mitochondrial dysfunction is a common feature of cancer cachexia, which occurs when the mitochondria work less efficiently than they should due to another disease or condition, such as cachexia [4] or aging [16]. We explored the mitochondrial responsiveness by using electron transport chain (ETC) inhibitors. Treatment with oligomycin or a combination of rotenone and antimycin A revealed that cPANC-1exposed myotubes were less sensitive to the inhibitors; thus, the oxidation of oleic acid after inhibitor exposure was higher in cPANC-1-exposed cells compared to chPEC-exposed myotubes ( Figure 5B). Further, total lipid accumulation from oleic acid and the level of free oleic acid (FFA) were higher in cPANC-1-exposed cells compared to chPEC-exposed cells ( Figure 5C,D). Incorporation of oleic acid into free fatty acids (FFA), triacylglycerol (TAG), and cholesterol ester (CE) was also higher in cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 5D).  . * Statistically significant in cPANC-1-exposed versus chPEC-exposed myotubes (p ≤ 0.05, paired t-test). CE, cholesteryl ester; DAG, diacylglycerol; FFA, free fatty acids; PL, phospholipid; TAG, triacylglycerol.
The phosphorylation of AMPK at threonine 172, relative to total AMPK, was also markedly lower in cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 6A,B). Lastly, mRNA expression of the fatty acid transporter CD36 (fatty acid translocase, FAT) was higher, while mRNA expression of the ETC complex III (UQCRB) and the nuclear receptor peroxisome proliferator-activated receptor alpha (PPARA) was lower in cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 6C).
The phosphorylation of AMPK at threonine 172, relative to total AMPK, was also markedly lower in cPANC-1-exposed myotubes compared to chPEC-exposed cells (Figure 6A,B). Lastly, mRNA expression of the fatty acid transporter CD36 (fatty acid translocase, FAT) was higher, while mRNA expression of the ETC complex III (UQCRB) and the nuclear receptor peroxisome proliferator-activated receptor alpha (PPARA) was lower in cPANC-1-exposed myotubes compared to chPEC-exposed cells ( Figure 6C). Figure 6. AMPK phosphorylation and gene expression in PANC-1-and hPEC-conditioned myotubes. In the last 4 days of differentiation, myotubes were conditioned with the medium from human primary pancreatic epithelial cells (chPEC exposed) or a pancreatic cancer cell line (cPANC-1 exposed). (A) Immunoblot showing AMPK phosphorylated on threonine 172 (pAMPK) and total AMPK. (B) Quantified immunoblots showing the pAMPK/total AMPK ratio adjusted to the internal control (HEK cells) and presented as a percentage of basal chPEC. (C) mRNA expression of ETC complex III (UQCRB), fatty acid transporter (CD36), and peroxisome proliferator-activated receptor alpha (PPARA), delta (PPARD) and gamma (PPARG). Data in (B,C) are from three individual experiments, each with two biological replicates, presented as the mean ± SEM (n = 6). * Statistically significant in cPANC-1-exposed versus chPEC-exposed myotubes (p ≤ 0.05, paired t-test). In the last 4 days of differentiation, myotubes were conditioned with the medium from human primary pancreatic epithelial cells (chPEC exposed) or a pancreatic cancer cell line (cPANC-1 exposed). (A) Immunoblot showing AMPK phosphorylated on threonine 172 (pAMPK) and total AMPK. (B) Quantified immunoblots showing the pAMPK/total AMPK ratio adjusted to the internal control (HEK cells) and presented as a percentage of basal chPEC. (C) mRNA expression of ETC complex III (UQCRB), fatty acid transporter (CD36), and peroxisome proliferator-activated receptor alpha (PPARA), delta (PPARD) and gamma (PPARG). Data in (B,C) are from three individual experiments, each with two biological replicates, presented as the mean ± SEM (n = 6). * Statistically significant in cPANC-1-exposed versus chPEC-exposed myotubes (p ≤ 0.05, paired t-test).

Secretomes from hPECs and PANC-1 Cells Indicate Increased Secretion of Certain Cytokines, Growth Factors, and Growth-Factor-Binding Proteins Potentially Affecting Metabolism in Skeletal Muscle Cells
Secretomes from hPECs and PANC-1 cells were examined using quantitative label-free proteomics (data previously published [10]). Cytokines, growth factors, and growth-factorbinding proteins in the secretomes were classified using STRING software ( Table 1). Most of these proteins were more abundant in PANC-1 medium compared to hPEC medium, including stem cell factor KIT ligand (KITLG), fibroblast growth factor 19 (FGF19), metallopeptidase inhibitor 1 (TIMP1), insulin-like growth factor 2 (IGF2), and insulin-like growth-factor-binding protein 2 (IGFBP2), while the amount of transforming growth factor beta-2 (TGFB2) reduced (Table 1). Transforming growth factor beta-2 −1.110 Color coding: growth factors (annotated keywords in STRING), growth-factor-binding proteins (annotated keywords in STRING), and cytokines.

Discussion
In this study, we found that myotubes exposed to a conditioned medium from PANC-1 cells had decreased protein content compared to myotubes exposed to hPEC medium. The PANC-1-medium-exposed myotubes also showed decreased overall leucine accumulation, increased overall leucine decay, and decreased gene expression of MYH2. Further, glucose uptake and oxidation decreased in cPANC-1-exposed cells. Moreover, we observed higher total cellular accumulation of oleic acid and a higher amount of FFA, TAG, and CE, supported by increased CD36 gene expression, in cPANC-1-exposed myotubes compared to chPEC-exposed cells. Basal fatty acid oxidation was equal in chPEC-exposed and cPANC-1exposed myotubes; however, ETC complex inhibitors had less impact on cPANC-1-exposed cells, and mRNA expression of the ETC-complex-III-related gene UQCRB reduced in cPANC-1-exposed myotubes.
Imbalance in protein metabolism due to hyperactive catabolism of muscle protein and decreased muscle protein mass is a common hallmark of cancer cachexia and is often already present at the time of diagnosis for many patients with pancreatic cancer [17]. When exploring the protein content in PANC-1-exposed myotubes, we found that the protein content decreased. Therefore, we further explored leucine accumulation and decay in cPANC-1-exposed myotubes. We found decreased accumulation and increased decay of radioactive leucine, suggesting an overall decrease in protein in these cells. This was also evident when comparing the protein amount from the two different cells, as the protein content was lower in cPANC-1-exposed compared to chPEC-exposed myotubes. In addition, the gene expression of MYH2 was significantly lower in cPANC-1-exposed myotubes. MYH2 is one of the three isoforms of myosin heavy chain, encoding a myosin heavy chain expressed in fast-twitch-type 2A fibers [18], the others being MYH1, expressed in fast-twitch-type 2X fibers, and MYH7, expressed in slow-twitch-type 1 fibers [19]. Grown in culture, human satellite cells mature into myotubes that express all three isoforms [20]. Mice studies have uncovered selective atrophy in type II fiber types during cachexia [21]. Muscle wasting in cancer cachexia is linked to increased NFκB activation in muscle [22]; however, we observed no difference in NFκB gene expression between our conditioned myotubes. The serine-threonine kinase mTOR is ubiquitously expressed and part of two different protein complexes named mTORC1 and mTORC2 [23]. The role of mTOR in cancer cachexia seems rather two-faced. On the one hand, cachexic muscle has reduced gene expression of mTOR complex 1 (mTORC1), which results in decreased protein and lipid synthesis; on the other hand, inhibiting mTORC1 protects against cachexia by upregulating autophagy and inhibiting pro-cachectic factors [23]. We observed no significant difference in MTOR mRNA expression between cPANC-1-exposed and chPEC-exposed myotubes.
Further, we found that cPANC-1-exposed myotubes had lower uptake and oxidation of glucose compared to chPEC-exposed myotubes, whereas insulin-induced glycogen synthesis was not different between the two conditions. Liu et al. showed that the diabetes typically found in patients with pancreatic cancer is characterized by profound peripheral insulin resistance, which impairs skeletal muscle glycogen synthesis and storage, and is also associated with a post-insulin resistance defect and no difference in SLC2A4 gene expression [24]. The insulin receptor is not thought to be affected in pancreatic-cancerrelated diabetes, which suggests that the defect might be further down the cascade toward glucose uptake and glycogen synthesis.
In our experiments, total lipid accumulation was higher in cPANC-1-exposed compared to chPEC-exposed myotubes and significantly more oleic acid was found as FFA or incorporated into TAG and CE, accompanied by increased gene expression of CD36. Tumor-bearing animals have shown to have a 60% increase in FAT in skeletal muscle [25]. Disruption of normal mitochondrial activity, leading to mitochondrial dysfunction, is associated with lipid metabolic disorders. However, increased lipid accumulation (i.e., TAG accumulation) can also lead to mitochondrial dysfunction [26]. The accumulation of cytosolic fatty acids will, in turn, increase the production of reactive oxygen species, which then leads to decreased mitochondrial biogenesis, increased mutation rates, and aging-like impaired mitochondrial function [27]. When comparing fatty acid oxidation in the conditioned myotubes, we observed decreased effects of mitochondrial inhibitors in cPANC-1-exposed cells. These are similar findings as seen in the mitochondria of aging cells [16]. Previous work comparing satellite cells from old versus young mice found that the muscle cells from young mice showed a decrease in basal respiration after addition of oligomycin, while the cells from old mice were less responsive, indicating that these cells already were respiring at minimal capacity [28]. However, the cells from old mice remained responsive to FCCP, though to a lesser extent than the cells from young mice [28]. This is similar to our results and indicates aging-like impaired mitochondrial function in myotubes after cPANC-1 exposure. Additionally, mitochondrial dysfunction in subjects with lipid storage diseases was associated with significant loss of activity in ETC complex I-IV compared to the control [29]. Studies on rat muscle cells found that only small changes in complex III activity (approximately 5% inhibition) was sufficient to change respiration [30]. Similar to these findings, we found a lower mRNA level of ETC complex III, UQCRB, in our cPANC-1-exposed myotubes compared to chPEC-exposed cells.
AMPK phosphorylation related to total AMPK was significantly lower in cPANC-1-exposed myotubes. As a major cellular energy sensor, AMPK acts as a moderator of skeletal muscle lipid metabolism [31]. Activation of AMPK through exercise is found to be beneficial as it increases glucose uptake independent of insulin [31]. AMPK also increases oxidative phosphorylation in skeletal muscle through transcriptional regulation [31]. The phosphorylation of AMPKα is inversely correlated with lipid accumulation in murine skeletal muscle C2C12 cells [31]. This is in line with the increased accumulation of lipids found in our cell model after cPANC-1 exposure, also linked to the typically increased lipid accumulation seen in cells with mitochondrial dysfunction [27].
Investigating relevant genes, we found a significant decrease in PPARA gene expression in cPANC-1-exposed compared to chPEC-exposed myotubes. PPARs are important metabolic regulators in the body, with different expression patterns across metabolic tissues. They are also involved in tumor development, though expression levels vary between different tumor types and stages of cancer [32]. Although PPARα agonists have been re-ported to show antitumor effects in colon carcinogenesis, it is still disputed whether PPARα represses or promotes cancer [32]. In patients with cachexia due to chronic obstructive pulmonary disease (COPD), PPARA mRNA expression decreased compared to patients with non-cachectic COPD [33]. PPARα may play a role in glucose usage in aged muscle, and decreased muscle glycogen concentrations have been detected in aged PPARα-deficient mice [34]. This supports our findings on mitochondria and glucose handling, implying an interference of the cancer cell secretome on myotube energy metabolism, causing decreased mitochondrial function and thereby increased lipid accumulation.
Previously, we showed the different metabolic profiles of hPECs and PANC-1 cells, where PANC-1 cells oxidized more glucose and less oleic acid compared to hPECs [10]. In addition, PANC-1 cells secreted a higher amount of proteins compared to hPECs [10]. Secreted substances from pancreatic cancer cells have been proposed to alter metabolism in skeletal muscle and induce peripheral insulin resistance and cachexia in skeletal muscle [7]. When comparing the proteins secreted from PANC-1 cells and hPECs, we found a higher abundance of some growth factors and growth-factor-binding proteins in the conditioned medium from PANC-1 cells compared to that from hPECs. These included the stem cell factor KITLG, often found to be upregulated in several types of cancers [35]; fibroblast growth factor 19 (FGF19), whose upregulation is linked to hepatocellular carcinoma and unfavorable outcomes [36]; and tissue inhibitor of metalloproteinases-1 (TIMP1), which is typically upregulated in patients with PDAC [37]. TIMP1 is associated with a poor clinical outcome for cancers [38] and is suggested as a biomarker for cachexia [39]. In addition, IGF2 and IGFBP2 were more abundant in the secretome of PANC-1 cells compared to hPECs. These are tumor growth promotors and, therefore, often increased in different tumors [40]. Increased IGF2 has also been found to lead to activation of Akt [41]. This could potentially explain the higher basal level of Akt phosphorylation seen in our cPANC-1exposed compared to chPEC-exposed myotubes. Previously, both increased and decreased Akt activity was observed during cancer cachexia [23].
Several of the typical circulating factors during cachexia, such as IL-1, IL-6, and TNFα [4], were not found in the secretome from PANC-1 cells and hPECs. However, in vivo, these factors or cytokines might come via the tumor cells' interaction with immune cells or other cells, which was missing in our in vitro experiments only using pancreatic cells and skeletal muscle cells. We also need to consider that there might be other factors in conditioned media influencing the myotubes. To mention one, we previously showed that glucose levels were significantly lower in PANC-1 medium compared to hPEC medium, while lactate levels significantly increased [10]. Although glucose levels were adjusted to be equal in the conditioned media in this study, lactate levels were not corrected.
Our data support the suggestion that cancer cells release substances that impact metabolism in skeletal muscle cells. In addition, an overall reduction in protein content, decreased gene expression of MYH2, decreased leucine accumulation, and increased leucine decay found in myotubes conditioned with PANC-1 medium indicate that the cancer cell secretome may affect the muscle protein content with similarities to the cachexia often seen in cancer. Our results also might indicate mitochondrial dysfunction in PANC-1conditioned myotubes, an important component in the wasting disorder often preceding atrophy development. A limitation of our study is that only a few selected markers associated with cachexia were examined and that the activity of several signaling pathways involved in muscle atrophy (Li et al., 2021) was not examined [42]. Another limitation is that we are solemnly studying the effects of conditioning primary human skeletal muscle cells with media from pancreatic cells using an in vitro system, as described before. Cachexia is a complicated process involving different cells and organs, and in vivo, there is a much more extensive cross-talk network [4]. In addition, our findings are results from one pancreatic cancer cell line, and the implications of conditioning primary human myotubes with media from other pancreatic cell lines might give different results.

Conclusions
Our data showed that substances released from a pancreatic cancer cell line can induce changes in energy metabolism and protein turnover in primary human myotubes. Changes in glucose and oleic acid metabolism, as well as leucine turnover, are affected in a manner resembling cachexia. Our research contributes to pancreatic cancer research; however, further studies are needed to elucidate the mechanistic pathways leading to cancer-induced changes in skeletal muscle metabolism.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/metabo12111095/s1, Table S1: Description of primers. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
Mass-spectrometry-based proteomics have previously been used to study global protein secretion from the two cell types. The full data set has been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with dataset identifier PXD017613.

Conflicts of Interest:
The authors declare no conflict of interest.