Decoration of intramyocellular lipid droplets with PLIN5 modulates fasting-induced insulin resistance and lipotoxicity in humans

Abstract

Aims/hypothesis: In contrast to insulin-resistant individuals, insulin-sensitive athletes possess high intramyocellular lipid content (IMCL), good mitochondrial function and high perilipin 5 (PLIN5) levels, suggesting a role for PLIN5 in benign IMCL storage. We hypothesised a role for PLIN5 in modulating fasting-mediated insulin resistance.

Methods: Twelve men were fasted for 60 h, before and after which muscle biopsies were taken and stained for lipid droplets (LDs), PLIN5 and laminin. Confocal microscopy images were analysed for LD size, number, PLIN5 association and subcellular distribution.

Results: Fasting elevated IMCL content 2.8-fold and reduced insulin sensitivity (by 55%). Individuals with the most prominent increase in IMCL showed the least reduction in insulin sensitivity (r = 0.657; p = 0.028) and mitochondrial function (r = 0.896; p = 0.006). During fasting, PLIN5 gene expression or PLIN5 protein content in muscle homogenates was unaffected, microscopy analyses revealed that the fraction of PLIN5 associated with LDs (PLIN5+) increased significantly (+26%) upon fasting, suggesting PLIN5 redistribution. The significant increase in LD number (+23%) and size (+23%) upon fasting was entirely accounted for by PLIN5+ LDs, not by LDs devoid of PLIN5. Also the association between IMCL storage capacity and insulin resistance and mitochondrial dysfunction was only apparent for PLIN5+ LDs.

Conclusions/interpretation: Fasting results in subcellular redistribution of PLIN5 and promotes the capacity to store excess fat in larger and more numerous PLIN5-decorated LDs. This associates with blunting of fasting-induced insulin resistance and mitochondrial dysfunction, suggesting a role for PLIN5 in the modulation of fasting-mediated lipotoxicity.

Trial registration: trialregister.nl NTR 2042.

Keywords: Fasting; IMCL; Lipid droplet size; Lipotoxicity; Perilipin 5.

Fig. 1

Correlation of the change in IMCL content and S_I-index (μmol min⁻¹ kg⁻¹ pmol⁻¹ l) upon prolonged fasting (n = 11), suggesting that promoting the capacity to store lipids as IMCL ameliorates fasting-induced insulin resistance. This correlation appears to be driven by PLIN5-coated LDs (Fig. 6a, b)

Fig. 2

LD size and number increased upon prolonged fasting. (a) Representative images of muscle fibres from the fed and fasted state (magnification × 70; scale bar, 15 μm). LDs are stained in green and cell membranes in blue. (b) Quantification of LD size and number relative to cell area. (c) Frequency distribution of LD size. Data are based upon examination of 3,595 ± 522 and 4,766 ± 489 LDs per participant (n = 9) in the fed and fasted state and presented as mean ± SEM. *p < 0.05 and **p < 0.01 vs fed state. White bars, fed state; black bars, fasted state

Fig. 3

PLIN5 gene expression (n = 12) (a) and PLIN5 protein content (n = 11) (b) measured in whole-muscle lysates in the fed and the fasted state. Data are presented as mean ± SEM

Fig. 4

(a) Representative three-dimensional images of LD (green) and PLIN5 protein localisation (red) in the fed and fasted state (scale bar, 7 μm). Please visit ESM Videos 1 and 2 for animated three-dimensional reconstruction. (b) Average PLIN5 protein content in muscle fibres measured as intensity of PLIN5 staining. (c) Percentage of PLIN5 protein content associated with LDs. Data are based upon examination of 3,595 ± 522 and 4,766 ± 489 LDs per participant (n = 9) in the fed and fasted state and presented as mean ± SEM, *p < 0.05 vs fed state

Fig. 5

Changes in LD size (a) and number (b) are accounted for by PLIN5+ LDs. Data are based upon examination of 3,595 ± 522 and 4,766 ± 489 LDs per participant (n = 9) in the fed and fasted state and are presented as mean ± SEM, *p < 0.05 or **p < 0.01 vs fed state; ^† p < 0.001 PLIN5+ vs PLIN5− LDs. White bars, fed state; black bars, fasted state

Fig. 6

Correlations of the fasting-induced changes (fasted–fed) in the number of PLIN5+ LDs (per μm²) (a) and the number of PLIN5− LDs (per μm²) (b) with the reduction in S_I-index (μmol min⁻¹ kg⁻¹ pmol⁻¹ l) (n = 9)

Fig. 7

Correlations of the fasting-induced changes in LD size (μm²) of all LDs (a, d, g, j), PLIN5+ LDs (b, e, h, k) and PLIN5− LDs (c, f, i, l) with changes in mitochondrial ADP-driven state 3 (S3) respiration (pmol mg⁻¹ s⁻¹ [mtDNA copy no. × 10⁴]⁻¹) in permeabilised muscle fibres on a lipid-derived substrate octanoyl-CoA (MO) (a–c), octanoyl-CoA + glutamate (MOG) (d–f) and octanoyl-CoA + glutamate + succinate (MOGS) (g–i) and during maximal uncoupling (j–l) (n = 7)