Red. In deciding which factors to measure, we elected to follow several that had already been investigated in human myotubes (IL6, TNFa, MCP-1) and others that had the potential to influence metabolism and/or inflammation. During the first 24 hours levels of secreted myokines ranged over four orders of magnitude, with MCP-1, IL8, GROa and follistatin amongst the most abundant factors (Fig 1A). T2D myotubes released significantly higher amounts of IL6, IL8, MCP-1, TNFa and GROa compared to ND cells, with a strong tendency for IL15 (p = 0.056) release to also be elevated. After 48 hours in culture, secretion of IL15 was significantly higher from T2D myotubes (Fig 1C). Even though all cultures were fully confluent, myotube protein content can vary considerably between individuals. To take this into account, myokine secretion was order PD98059 normalized against cell protein. When this adjustment was made, secretion of IL6, IL8, IL15, GROa, TNFa, MCP-1 and follistatin were all significantly higher from T2D myotubes, depending on the time Roc-A molecular weight evaluated (Fig 1B 1D). ND and T2D myotubes secreted similar amounts ofTable 2. Circulating cyto- and chemokine levels. [factor] (pg/mL) IL-1b IL-6 IL-8 IL-10 IL-15 IFNg TNFa MCP1 MIP1a GROa VEGF Follistatin *p<0.05 vs ND, not adjusted for age or BMI. doi:10.1371/journal.pone.0158209.t002 ND 1.96 ?0.56 4.56 ?2.54 7.88?1.02 5.23 ?1.01 3.80 ?0.50 6.79 ?3.64 3.99 ?0.36 303 ?26 2.61 ?1.00 472 ?88 97.5 ?16.4 1366 ?153 T2D 1.68 ?0.65 3.07 ?1.17 10.26 ?1.31 7.89 ?2.22 5.72 ?1.89 5.92 ?1.87 6.16 ?0.52* 277 ?31 8.09 ?5.18 885 ?193* 82.6 ?26.5 2159 ?327*PLOS ONE | DOI:10.1371/journal.pone.0158209 July 25,5 /Myokine Secretion in Type 2 DiabetesFig 1. Myokine secretion from differentiated hSMC obtained from non-diabetic (ND, open bars) and Type 2 Diabetic (T2D, solid bars) subjects. (A) Media concentrations of selected factors after 24 hr of culture. (B) 24 hr secretion normalized against total cell protein for each individual set of cells. N = 8?3 for ND, 8?8 for T2D. (C) Media concentrations after 48 hr in culture. (D) 48 hr secretion normalized against cell protein. n = 8?8 for ND, 5?4 for T2D. Results are average + SEM. *p<0.05 vs ND doi:10.1371/journal.pone.0158209.gIL1b and VEGF. Of other factors of interest, IFNg in the media was consistently below the limit of detection, regardless of the time of culture, and IL10 levels often did not exceed the lower limit of detection; this was seen with both ND and T2D hSMC.Regulation of myokine secretionMany of the myokines we found to be secreted to a greater extent by T2D myotubes are known pro-inflammatory cytokines and chemokines. Next, we investigated the impact of exposure to an inflammatory stimulus, lipopolysaccharide (LPS, 1 mg/mL) [26] on the secretion of selected myokines. While statistical analysis of treatment effects on myokine secretion was performed on absolute values, due to the wide range in absolute values, results are presented relative to the value in untreated (control) cells from each individual subject. Secretion of a number of myokines was increased over 24 hr of treatment with LPS (Fig 2A). In myotubes from ND subjects, LPS-induced changes in secretion attained statistical significance for GROa, IL15 and TNFa. In myotubes from T2D subjects, in addition to GROa, IL15, and TNFa, secretion of IL6 and IL8 was also stimulated to a significant degree by LPS treatment. LPS had no effect on secretion of follistatin (data not shown) or VEGF from either ND or T2D myotube.Red. In deciding which factors to measure, we elected to follow several that had already been investigated in human myotubes (IL6, TNFa, MCP-1) and others that had the potential to influence metabolism and/or inflammation. During the first 24 hours levels of secreted myokines ranged over four orders of magnitude, with MCP-1, IL8, GROa and follistatin amongst the most abundant factors (Fig 1A). T2D myotubes released significantly higher amounts of IL6, IL8, MCP-1, TNFa and GROa compared to ND cells, with a strong tendency for IL15 (p = 0.056) release to also be elevated. After 48 hours in culture, secretion of IL15 was significantly higher from T2D myotubes (Fig 1C). Even though all cultures were fully confluent, myotube protein content can vary considerably between individuals. To take this into account, myokine secretion was normalized against cell protein. When this adjustment was made, secretion of IL6, IL8, IL15, GROa, TNFa, MCP-1 and follistatin were all significantly higher from T2D myotubes, depending on the time evaluated (Fig 1B 1D). ND and T2D myotubes secreted similar amounts ofTable 2. Circulating cyto- and chemokine levels. [factor] (pg/mL) IL-1b IL-6 IL-8 IL-10 IL-15 IFNg TNFa MCP1 MIP1a GROa VEGF Follistatin *p<0.05 vs ND, not adjusted for age or BMI. doi:10.1371/journal.pone.0158209.t002 ND 1.96 ?0.56 4.56 ?2.54 7.88?1.02 5.23 ?1.01 3.80 ?0.50 6.79 ?3.64 3.99 ?0.36 303 ?26 2.61 ?1.00 472 ?88 97.5 ?16.4 1366 ?153 T2D 1.68 ?0.65 3.07 ?1.17 10.26 ?1.31 7.89 ?2.22 5.72 ?1.89 5.92 ?1.87 6.16 ?0.52* 277 ?31 8.09 ?5.18 885 ?193* 82.6 ?26.5 2159 ?327*PLOS ONE | DOI:10.1371/journal.pone.0158209 July 25,5 /Myokine Secretion in Type 2 DiabetesFig 1. Myokine secretion from differentiated hSMC obtained from non-diabetic (ND, open bars) and Type 2 Diabetic (T2D, solid bars) subjects. (A) Media concentrations of selected factors after 24 hr of culture. (B) 24 hr secretion normalized against total cell protein for each individual set of cells. N = 8?3 for ND, 8?8 for T2D. (C) Media concentrations after 48 hr in culture. (D) 48 hr secretion normalized against cell protein. n = 8?8 for ND, 5?4 for T2D. Results are average + SEM. *p<0.05 vs ND doi:10.1371/journal.pone.0158209.gIL1b and VEGF. Of other factors of interest, IFNg in the media was consistently below the limit of detection, regardless of the time of culture, and IL10 levels often did not exceed the lower limit of detection; this was seen with both ND and T2D hSMC.Regulation of myokine secretionMany of the myokines we found to be secreted to a greater extent by T2D myotubes are known pro-inflammatory cytokines and chemokines. Next, we investigated the impact of exposure to an inflammatory stimulus, lipopolysaccharide (LPS, 1 mg/mL) [26] on the secretion of selected myokines. While statistical analysis of treatment effects on myokine secretion was performed on absolute values, due to the wide range in absolute values, results are presented relative to the value in untreated (control) cells from each individual subject. Secretion of a number of myokines was increased over 24 hr of treatment with LPS (Fig 2A). In myotubes from ND subjects, LPS-induced changes in secretion attained statistical significance for GROa, IL15 and TNFa. In myotubes from T2D subjects, in addition to GROa, IL15, and TNFa, secretion of IL6 and IL8 was also stimulated to a significant degree by LPS treatment. LPS had no effect on secretion of follistatin (data not shown) or VEGF from either ND or T2D myotube.