Whole genome expression profiling associates activation of unfolded protein response with impaired production and release of epinephrine after recurrent hypoglycemia

Abstract

Recurrent hypoglycemia can occur as a major complication of insulin replacement therapy, limiting the long-term health benefits of intense glycemic control in type 1 and advanced type 2 diabetic patients. It impairs the normal counter-regulatory hormonal and behavioral responses to glucose deprivation, a phenomenon known as hypoglycemia associated autonomic failure (HAAF). The molecular mechanisms leading to defective counter-regulation are not completely understood. We hypothesized that both neuronal (excessive cholinergic signaling between the splanchnic nerve fibers and the adrenal medulla) and humoral factors contribute to the impaired epinephrine production and release in HAAF. To gain further insight into the molecular mechanism(s) mediating the blunted epinephrine responses following recurrent hypoglycemia, we utilized a global gene expression profiling approach. We characterized the transcriptomes during recurrent (defective counter-regulation model) and acute hypoglycemia (normal counter-regulation group) in the adrenal medulla of normal Sprague-Dawley rats. Based on comparison analysis of differentially expressed genes, a set of unique genes that are activated only at specific time points after recurrent hypoglycemia were revealed. A complementary bioinformatics analysis of the functional category, pathway, and integrated network indicated activation of the unfolded protein response. Furthermore, at least three additional pathways/interaction networks altered in the adrenal medulla following recurrent hypoglycemia were identified, which may contribute to the impaired epinephrine secretion in HAAF: greatly increased neuropeptide signaling (proenkephalin, neuropeptide Y, galanin); altered ion homeostasis (Na+, K+, Ca2+) and downregulation of genes involved in Ca2+-dependent exocytosis of secretory vesicles. Given the pleiotropic effects of the unfolded protein response in different organs, involved in maintaining glucose homeostasis, these findings uncover broader general mechanisms that arise following recurrent hypoglycemia which may afford clinicians an opportunity to modulate the magnitude of HAAF syndrome.

Fig 1. Experimental design.

The schematic represents the basic protocol and the time line of the experiments. Before the antecedent treatments, a period of habituation was planned to minimize any external influences from travel and new housing. Body weight was monitored on a daily basis for each animal during the experiments.

Fig 2. Blood glucose levels during the antecedent treatments on days 1–3.

The values (mg/dL) are expressed as mean ± SE, n = 6.

Fig 3. Plasma glucose concentrations during the hyperinsulinemic-hypoglycemic glucose clamp.

The values (mg/dL) for twice-daily saline control (2RS) and twice daily RH groups (2RH) are shown as mean ± SE, n = 6 for each experimental group.

Fig 4. Plasma hormonal responses during the hypoglycemic clamp.

A) Epinephrine (ng/ml) and B) glucagon (pg/ml) responses in twice daily recurrently hypoglycemic (2RH) rats and in the corresponding saline group (2RS). Data are summarized from three independent experiments, n = 6 animals per group. Values are shown as mean ± SE, *p <0.05 or **p<0.002 vs. corresponding control at given time point.

Fig 5. Venn diagrams of differentially expressed genes.

A) Comparative analysis of DEGs in the 2RH0 and 2RH60 groups vs their respective saline controls (2RS0 and 2RS60); B) Comparative analysis of DEGs in 2RH groups—2RH30 and 2RH60 vs 2RHO respectively.

Fig 6. Enrichment analysis of DEGs in 2RH and 2RS experimental groups.

A) Shown is the distribution by process networks at 0 time point (2RH0 vs. 2RS0) and B) at 60 min time point (2RH60 vs. 2RS60). Sorting is done for unique genes and both signals (induced genes—shown in red and repressed genes—in blue) included. Top 10 process networks are listed based on their—log (p-value). For list of abbreviations see supplemental file S1 Text.

Fig 7. Functional analysis of DEGs at different time points following RH.

The distribution by process networks is shown, with top 10 significantly enriched GO items for differentially-expressed common genes in 2RH30 vs 2RH0 and 2RH60 vs 2RH0. For list of abbreviation see S3.

Fig 8. RH induces the unfolded protein response in rat adrenal medulla.

A schematic of the unfolded protein response in mammals is presented. In resting cells all three ER stress sensors are inactive due to association with Grp78. Accumulation of unfolded proteins leads to dissociation of Grp78 and activation of IRE1, ATF6 and PERK which reprogram transcription and translation in a concerted manner to restore homeostasis: increase the transcription of genes involved in protein folding (ATF6 and IRE1 signaling), attenuate global protein synthesis (PERK) by phosphorylating translation initiation factor 2 (eIF2a) while promoting the translation of ATF4. ATF4 controls the expression of CHOP, which in turn induces GADD34 –a negative feedback loop effector that terminates UPR signaling by recruiting protein phosphatase1 catalytic subunit resulting in dephosphorylation of eIF2a and recovery of protein synthesis ([37]). Selected genes affected only by RH in our experiments are indicated in bold.

Fig 9. Western blot analysis confirms activation of UPR following RH.

Total protein lysates from the right adrenal medulla of saline (2RS) and RH groups (2RH) were subjected to Western blot analyses as described in methods. Proteins were separated by SDS-PAGE, electroblotted and the membranes were sequentially probed with antibodies specific to Grp78, Derlin1, TH and GAPDH. Similar results were obtained in two separate replicate experiments.