Glucocorticoids (GCs) are potent therapeutic agents frequently used for treatment of number of conditions, including hematologic, inflammatory, and allergic diseases (1). The therapeutic effect of GCs administration has been repeatedly shown to vary interindividually due to numerous factors like gender, age, and particularly, genetic constitution of the treated individual. One extreme of the range of varied responses is a near- to complete failure to respond to GC therapy, described as GC insensitivity or resistance (2, 3). Similar variety is observed in unwanted, side effects accompanying GCs therapy—lower bone mineral density and osteoporosis, diabetes, dyslipidemia, obesity, impaired wound healing, or muscle wasting (4). Manifestation of the adverse effects is, to certain extent, dependent on the therapeutic dose and time of treatment; however, pharmacogenetic interactions have been implicated as substantial contributing factors (5, 6). Variation in genes coding for proteins directly involved in GC genomic actions were shown to alter GC sensitivity, namely, the GC receptor (NR3C1), recently summarized by Nicolaides and Charmandari (7). Clearly, other genes can also mediate the clinically relevant adverse effects or changes in response to GC. For instance, genome-wide association study (GWAS) of response to inhaled GCs in asthma patients identified a polymorphism in glucocorticoid-induced transcript 1 gene (GLCCI1) and further replicated the lower response of individuals carrying two copies of the variant allele in four independent cohorts (8). In another GWAS performed in acute lymphoblastic leukemia patients, dexamethasone (DEX)-induced pleiotropic side effects were significantly associated with variation in F2RL1 (F2R like trypsin receptor 1) gene (9). In an experimental, genetically designed model strain set, we showed previously that several genomic loci affect sensitivity toward DEX-induced unfavorable changes in carbohydrate and lipid metabolism (10). Later, we reported that a presence of defective Cd36/Fat-containing genomic segment substantially blunted DEX-induced glucose intolerance and dyslipidemia (11). Also, we have isolated a small region of rat chromosome 8, which sensitized its carrier toward DEX-induced, muscle-specific insulin resistance (IR). In the process of positional cloning of gene(s) connected to several aspects of metabolic syndrome based on results of linkage and association studies in segregating populations and recombinant inbred strains, we derived congenic strain SHR-Lx.PD5 (12). This strain harbors a limited rat chromosome 8 segment of polydactylous rat (PD/Cub) origin on the genomic background of the spontaneously hypertensive rat (SHR). While SHR represents one of the most studied models of essential hypertension (13), it shows also IR, dyslipidemia, and other features of metabolic syndrome (14). By contrast, the PD/Cub strain is a model of metabolic syndrome without elevated blood pressure (15), particularly sensitive to nutritional and pharmacologic challenges (10, 16–18). We have hypothesized that a PD/Cub-derived, GC-responsive gene present within the isolated region, Zbtb16 (zinc finger and BTB domain containing 16), is a prime candidate for the observed pharmacogenetic interaction (12). Our previous sequence analysis revealed a conservative amino acid substitution (T208S) and a 2,964-bp deletion in intron 2 of PD/Cub’s Zbtb16 gene (19). Zbtb16 is a pleiotropic transcription factor involved, among others, in processes underlying pathogenesis of practically all major constituents of metabolic syndrome, as reviewed recently (20). In this study, we created two new models, one of them carrying only the Zbtb16 gene from the original segment to retest its effect on DEX-induced metabolic syndrome features by comparing their metabolic profiles including global and tissue-specific insulin sensitivity.

In a follow-up to previous identification of limited genomic region responsible for DEX-induced dyslipidemia and muscle-specific IR, we present in this study further evidence that GC-inducible transcription factor Zbtb16 (25) plays a major role in such sensitization. Zbtb16, also known as Plzf (promyelocytic leukemia zinc finger), belongs to POK (POZ and Krüppel zinc finger) family of proteins and is involved in plethora of physiological processes including maintenance of spermatogenesis (26), stem cell self-renewal (27), hematopoiesis (28), limb development (29, 30) and regulation of metabolism (31, 32). The gene is quite conserved in mammals, the human and rat ZBTB16/Zbtb16 proteins show 96% identity over 673 amino acids constituting the protein. The new, single-gene congenic strain carrying the variant of Zbtb16 gene (3 kb intronic deletion + T208S substitution) showed increased plasma TG and diminished insulin sensitivity of skeletal muscle tissue. Both these observations correspond to the original findings in parental strain SHR-Lx.PD5 (with seven genes in its differential segment, including Zbtb16) when compared with SHR (12). Reduced visceral adiposity observed in the previous study was not confirmed despite the matching difference trend (P = 0.06). The concept of DEX-induced IR of muscle tissue is well known, and several mechanisms have been proposed to convey this effect including direct effects on glucose transporter GLUT4 trafficking (33), increase in glycogen synthase phosphorylation together with decrease in expression and insulin-stimulated phosphorylation of protein kinase B (34) or dysregulation of intramyocellular lipid metabolism (35). To the best of our knowledge, involvement of Zbtb16 in this process has not been reported so far. We have recently reviewed in detail the pleiotropic connections of Zbtb16 to the facets of metabolic syndrome, including IR and dyslipidemia (20). Zbtb16 is strongly induced by GCs (36) and at the same time it negatively regulates the insulin signaling pathway by decreasing the phosphorylation of IRS1, Akt, and FoxO1 in normal mice (32). This presents potential basis for a cross-talk involving DEX, Zbtb16, and insulin sensitivity of muscle, somewhat similar to the one reported for the role of Zbtb16 in regulation of hepatic gluconeogenesis previously. In the suggested model, Zbtb16 acts as a downstream effector of the PGC-1α/GR complex, activated by gluconeogenic signals. Activated PGC-1α/GR then induces Zbtb16 expression, triggering hepatic gluconeogenesis and, at the same time, negatively regulating the insulin signaling pathway (32). The insulin sensitivity of skeletal muscle determined repeatedly in DEX-treated SHR [unpublished data and Ref. (12)] is comparable to that of SHR-Lx.PD5^SHR-Zbtb16 strain (SHR basal 121 ± 13; insulin-stimulated 226 ± 25 nmol glucose/g/120 min vs. SHR-Lx.PD5^SHR-Zbtb16 basal 173 ± 23; insulin-stimulated 247 ± 38 nmol glucose/g/120 min) contrasting with decreased sensitivity in the single-gene congenic strain with the PD/Cub mutant allele of Zbtb16 gene as well as the original SHR-Lx.PD5 strain (SHR-Lx.PD5 basal 82 ± 6; insulin-stimulated 142 ± 10 nmol glucose/g/120 min vs. SHR-Lx.PD5^PD-Zbtb16 basal 55 ± 9; insulin-stimulated 144 ± 16 nmol glucose/g/120 min). In skeletal muscle, GCs were shown to cause IR by reducing transcription of IRS1, while increasing expression of proteins that interfere with insulin action, including protein tyrosine phosphatase type 1B, leukocyte common antigen-related protein (LAR) and p38MAPK (37). However, over 170 potential GC targets have been identified in muscle (38) affecting not only insulin sensitivity but also protein metabolism with effects extending to other organs, as documented in the muscle-specific GR knockout animals (39). While interactions between Zbtb16 and several of the implicated genes have been described, the identification of causal perturbed pathways in the SHR-Lx.PD5^PD-Zbtb16 remains to be established. The enhanced TG storage in liver due to GC administration (40) was more prominent in the SHR-Lx.PD5^SHR-Zbtb16 strain. In an experimental model with TALEN (transcription activator-like effector nuclease)-mediated Zbtb16 knockout, the heterozygous SHR-Plzf^+/− animals showed decreased concentrations of TG and cholesterol in liver and plasma (41). Hepatic Zbtb16 is upregulated in several murine models showing severe hepatic steatosis (32). However, given the higher fasting and satient plasma concentrations of TG in SHR-Lx.PD5^PD-Zbtb16, we can speculate that DEX-induced activation of liver Zbtb16 in the single congenic strain could cause a decrease of phosphorylation of FoxO1 (32), which in turn activated both gluconeogenesis (leading to increased glycemia) and increased expression of microsomal triglyceride transfer protein, augmenting very low-density lipoprotein, TG-rich particles assembly and secretion (42). The reduced total body weight and relative weight of liver in the original SHR-Lx.PD5 strain compared with SHR is consistent with the morphometric profile of the strain without the variant Zbtb16 allele, i.e., the SHR-Lx.PD5^SHR-Zbtb16 strain. Although at this point it is impossible to pinpoint any of the six genes present in the differential segment as a major contributor, the serotonin receptor Htr3a was recently shown to modulate energy expenditure and weight gain via adipose tissue (43) and can be thus considered as a potential candidate. The limitations of this study include the use of only male rats as sex-specific genetic architecture of the metabolic syndrome and its components was demonstrated (44). Also, as the experimental protocol was set up to mimic the original set of experiments comparing SHR rats to SHR-Lx.PD5 congenic, only one dose of DEX and a short-term administration in 5-month-old animals was used. It is possible that the sensitizing effect would be modified under different conditions, yet the observed robust changes, particularly regarding the blunted insulin sensitivity of the skeletal muscle tissue, would most likely not be affected. Further studies should address in detail the mechanisms and pathways, through which the Zbtb16 mediates the sensitization to DEX-induced IR of skeletal muscle. Also, the relevance of our findings for human condition needs to be validated. So far, there is not any pharmacogenetic study involving variation in human ZBTB16 gene available. In a study of 1517 non-diabetic subjects without hypolipidemic treatment, several single nucleotide polymorphisms in ZBTB16 gene were associated with adiposity measures, LDL, and total cholesterol (45), suggesting the importance of ZBTB16 for metabolic syndrome in humans. The metabolic disturbances including impaired glucose tolerance, dyslipidemia, and IR of skeletal muscle observed after DEX treatment in the congenic SHR-Lx.PD5^PD-Zbtb16 reveal the Zbtb16 locus as a possible sensitizing factor for side effects of GC therapy.

All experiments were performed in agreement with the Animal Protection Law of the Czech Republic (311/1997) which is in compliance with the European Community Council recommendations for the use of laboratory animals 86/609/ECC and were approved by the Ethical committee of the First Faculty of Medicine and the Ministry of Education, Youth and Sports of the Czech Republic (protocol ID MSMT-1461/2015-17).

VK and DK derived the new congenic substrains. OS, LS, MK, AK, and LK carried out the experimental components of the study and drafted the manuscript. FL participated in the design of the study and performed the statistical analysis. OS, VK, and LK conceived the study, participated in its design and coordination, and helped to draft the manuscript. All the authors participated in the manuscript preparation and read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.