|Home | About | Journals | Submit | Contact Us | Français|
Peroxisome proliferator-activated receptor-γ (PPARγ) agonists are commonly used for the treatment of diabetes although their PPARγ-dependent effects transcend their role as insulin sensitizers. Thiazolidinediones lower blood pressure (BP) in diabetic patients while results from conventional/tissue-specific PPARγ experimental models suggest an important pleiotropic role for PPARγ in BP control. Little evidence is available on the molecular mechanisms underlying the role of vascular smooth muscle cells (VSMC)-specific PPARγ in basal vascular tone.
We show that VSMC-selective deletion of PPARγ (SMPG KO) impairs vasoactivity with an overall reduction in BP. Aortic contraction in response to norepinephrine (NE) is reduced and vasorelaxation is enhanced in response to β-adrenergic receptor (β-AdR) agonists in vitro. Similarly, SMPG KO mice display a biphasic response to NE in BP, reversible upon administration of β-AdR blocker and enhanced BP reduction upon treatment with β-AdR agonists. Consistent with enhanced β2-AdR responsiveness, we found that absence of PPARγ in VSMC increased β2-AdR expression possibly leading to the hypotensive phenotype during the rest phase.
These data uncovered the β2-AdR as a novel target of PPARγ transcriptional repression in VSMC and indicate that PPARγ regulation of β2-adrenergic signaling is important in modulation of BP.
Management of type 2 diabetes involves controlling hyperglycemia in order to maintain blood glucose levels as similar as possible to those in normoglycemic individuals. Several therapeutic approaches are currently in use to achieve this goal1. At present, thiazolidinediones (TZDs) stand among the plethora of therapeutic interventions aimed at lowering blood glucose levels. TZDs are ligands for the peroxisome proliferator-associated receptor-γ (PPARγ), a key transcription factor involved in lipid metabolism with a central role in the regulation of insulin sensitivity2.
Compelling evidence directly relates PPARγ with additional effects on the cardiovascular system3. Growing data suggest that insulin resistance may be strongly linked to hypertension. For instance, hypertension is associated with substantial insulin resistance, even in patients without diabetes where TZDs have been found to decrease BP4. This effect has been observed in type 2 diabetic patients with hypertension5, and also in normotensive type 2 diabetic individuals6. In addition, dominant-negative mutations in the human PPARγ gene have been associated with severe hypertension at a very early age7. Recent clinical data have shown that treatment of these patients with pioglitazone significantly improved endothelium-dependent dilation in response to bradykinin without affecting the response to sodium nitroprusside (SNP)8. Rosiglitazone therapy also improved the in vivo reendothelization capacity of endothelial progenitor cells from diabetic patients9. However, although TZDs treatment lowered BP levels in diabetic patients and in murine models of diabetes and hypertension, a conventional PPARγ knockout experimental model, in which embryonic lethality was rescued to produce global PPARγ KO, surprisingly displayed a systemic hypotensive phenotype despite of severe insulin resistance10. In order to further understand the contribution of PPARγ to BP regulation in the vasculature, tissue specific knock-out animals have been generated. In that regard, it is noteworthy that endothelial-specific PPARγ KO mice did not show an apparent hypertensive phenotype unless otherwise induced by different challenges, such as high-salt water or feeding a high-fat diet11. Since vascular smooth muscle cells (VSMC) are essential to vascular tone modulation, we examined the contribution of VSMC-specific PPARγ function to regulation of vascular tone and BP. To this aim, we developed a highly VSMC-selective PPARγ knock-out mice using our SM22α-Cre knock-in mice12 crossed to PPARγflox/flox,13. The results presented here indicate that VSMC-selective PPARγ deficiency results in a hypotensive phenotype, apparently at odds with the TZD agonistic phenotypes. We provide evidence of PPARγ-dependent transcriptional repression of β2-AdR, uncovering a possible link between PPARγ deficiency with increased vascular relaxation and reduced BP.
BP was measured by radiotelemetry (PA-C20; Data Sciences International). The left common carotid artery was cannulated with the implant catheter, and the implant body was secured in the abdominal cavity. Mean systolic BP and diastolic BP were recorded after implantation of the devices. Mice were kept on a 12-hour light/12-hour dark cycle.
Mean ± SEM values were analyzed. Statistical comparisons between two groups were performed by Student’s t test, and among three groups were performed by one-way ANOVA. Aorta ring contraction/relaxation curves were adjusted with nonlinear regression first, and then were compared by two-way ANOVA and Bonferroni post-tests. Groups were considered significantly different if P values were <0.05.
Please see the detailed methods online
We generated vascular smooth muscle cell-selective PPARγ knockout (SMPG KO) mice by crossing PPARγflox/flox mice13 with the SM22α-Cre knock-in mice that we previously developed12. Currently available VSMC-selective PPARγ KO mice do not display a high efficient Cre-mediated excision of loxP sites and residual PPARγ expression could still be functional in the aortic tissue14,15. We used our knock-in mice in which Cre expression is driven by the endogenous SM22α promoter to circumvent that limitation. For the current studies, we used SMPG KO and littermate control (LC) mice (see methods online). RT-PCR analysis of the wild-type PPARγ allele (~700 bp) and the deleted allele (~300 bp) indicates that the functional PPARγ transcript in SMPG KO mice was fully abolished in the vascular smooth muscle cell layer of aortic tissue and only partially excised in the heart and stomach (Supplemental Figure 1A). Of significance, PPARγ protein expression in the smooth muscle cell layers from aorta was completely absent in SMPG KO mice as determined by Western-blot analysis using a PPARγ antibody (Supplemental Figure 1B).
To analyze the contribution of VSMC-specific PPARγ function to the systemic regulation of BP, we determined baseline systolic and diastolic BP of SMPG KO and LC mice by radiotelemetry (n=6). As depicted in figures 1A and 1B, LC mice showed the typical circadian regulation of BP in mice, higher during night time and lower during the day time. SMPG KO mice displayed a similar pattern of circadian rythmicity in BP although these mice showed a significantly lower baseline systolic (SBP) and diastolic BP (DBP) during the lights-on period, reaching minimal levels around 6pm (average BP from 2 pm to 6 pm, SBP: SMPG KO, 96±27 mmHg versus LC, 123±14 mmHg, p=0.03; DBP: SMPG KO, 84±15 mmHg versus LC, 102±14 mmHg, p=0.025) and recovering to the levels of LC mice during the active, lights-off period (average BP from 8pm to 12am, SBP: SMPG KO, 140±17 mmHg versus LC, 134±6 mmHg, p=0.19; DBP: SMPG KO 118±22 mmHg versus LC, 111±8 mmHg, p=0.21) (Figure 1A and 1B, doted lines). That SMPG KO mice present lower BP was confirmed by measuring systolic BP in mice ranging from 1 to 6 months old (Figure 1C). Measurements were taken between 2pm and 6pm using the tail-cuff method both in SMPG KO and LC mice (n=16). Consistent with the radiotelemetry data, the SMPG KO mice showed significantly lower BP compared to the LC controls during the less active period (103±11 mmHg, LC, versus 84±10 mmHg, SMPG KO on the first month of age; 107±10 mmHg LC versus 90±10 mmHg SMPG KO on the second month; 110±7 mmHg, LC, versus 87±9.6 mmHg on the third month; and 106±8 mmHg, LC versus 92±9 mmHg SMPG on 6 months-old mice) (Figure 1C). Also, SMPG KO female mice have similar hypotensive phenotype (data not shown).
We then investigated whether changes in BP in SMPG KO mice correlated with changes in metabolic parameters, insulin sensitivity and heart function. The results obtained for glucose tolerance and insulin tolerance tests (Supplemental Figures 2A and 2B) in SMPG KO did not differ from LC mice. We additionally determined whether reduction in BP was a result of changes in vasomotor activity, housing the experimental groups in metabolic cages to record VO2 consumption (Supplemental Figure 3A), VCO2 generation (Supplemental Figure 3B) and ambulatory activity (Supplemental Figure 3C). No significant differences among the experimental groups were observed either during the nocturnal or the diurnal period. We further demonstrated that heart function was not significantly different between SMPG KO mice and LC mice hearts as determined by echocardiography analysis (Supplemental Figure 4 and Supplemental Table 1), neither did we observe any differences in heart rate among the different genotypes during the telemetry measurements (data not shown). Histological analysis of aortic, heart and skeletal muscle tissue did not reveal any morphological changes between SMPG KO mice and LC mice (Supplemental Figure 5).
To determine whether reduced BP was caused by impaired vasoactivity, we tested the response of aortic rings isolated from SMPG KO and LC mice to various vasoconstrictors. Administration of phenylephrine (PE), an α-adrenergic receptor agonist, resulted in similar maximal contraction levels in aortic rings from SMPG KO and LC mice (Figure 2A). However, aortic rings from SMPG KO mice showed significantly lower maximum contraction in response to norepinephrine (NE), a dual α- and β-adrenergic receptor agonist, compared to rings from LC mice (Figure 2B). Of interest, NE showed a biphasic response in SMPG KO mice. Whereas the lower dose-dependent contraction in response to NE (10−9-10−7mol/L) was significantly reduced in SMPG KO mice, a higher dose (10−6mol/L) promoted a robust vasorelaxation in SMPG KO mice, an effect not observed in LC mice (Figure 2B, Figure 2C line b). The differential contractile response in the SMPG KO animals indicated a selective enhanced sensitivity to specific β-adrenergic receptors versus α-adrenergic receptor agonists (e.g. NE versus PE). This observation prompted us to further explore the β-AdR-dependent vasoactivity of SMPG KO mice. Pre-treatment of SMPG KO mice aortic rings with propanolol (Figure 2C, line a), a non-specific β2-AdR receptor antagonist, completely abolished SMPG KO mice aortic ring relaxation in response to NE (Figure 2C, line b). On the other hand, treatment of PE-preconstricted aortic rings with a selective β2-AdR agonist, terbutaline, induced greater relaxation in aortic rings from SMPG KO mice (Figure 2D, line b) when compared to LC mice (Figure 2D, line a). In contrast, both the endothelial-dependent aortic relaxation response to Ach (Figure 2E) and endothelial-independent vasorelaxation response to SNP (Figure 2F) were no significantly different in aorta rings of SMPG KO and LC mice, indicating that neither the endothelial response nor the endothelial independent VSMC relaxation was impaired in the SMPG KO mice.
We then examined whether there was a direct correlation in BP regulation upon treatment with β2-AdR receptor agonists or antagonists in vivo. Infusion of 10−6 mol/L of NE at 2 µl/min caused a sustained increase in BP in LC mice (Figure 3A). However, NE infusion in SMPG KO mice caused an initial increase in BP, followed by a gradual decrease (Figure 3B) recapitulating the response of the aortic rings in vitro. Co-infusion with 10−7mol/L propranolol (2 µl/min) and NE further increased BP in LC mice (Figure 3C) and completely abrogated the biphasic effect of NE in SMPG KO mice (Figure 3D). Additional experiments using 10−6mol/L terbutaline (2 µl/min), a specific β2-AdR subtype agonist, failed to alter BP in LC mice (Figure 3E), whereas the SMPG KO mice displayed higher susceptibility to the BP lowering effect of terbutaline (Figure 3F). These results confirmed that the β-AdR response was exacerbated in the SMPG KO in vivo.
The results above prompted us to hypothesize that PPARγ depletion may directly affect β2-AdR expression. Western blot analysis of protein extracts prepared from aortic vessels of LC and SMPG KO mice demonstrated that β2-AdR protein levels were significantly increased in SMPG KO mice compared with LC mice (Figure 4A), suggesting that β2-AdR could be a novel target of PPARγ-dependent transcriptional repression in vivo. These results were further confirmed by gain-and loss of-PPARγ in vitro. Adenoviral-mediated expression of PPARγ in VSMC strongly inhibits β2-AdR protein levels. Conversely, shRNA-mediated PPARγ knockdown leads to enhanced expression of the β2-AdR (Figure 4B and supplemental Figure 6), reinforcing the notion of a repressor effect of PPARγ on the β2-AdR. The effect of activation of PPARγ by TZDs was tested in SMPG KO mice compared to LC mice, with rosiglitazone added to the animal chow at the dosage of 10 mg/kg/day. After 2 weeks of rosiglitazone regime, β2-AdR mRNA expression was determined in isolated aortic SMCs from SMPG KO and LC mice compared to those of normal diet. Figure 4C shows again that β2-AdR mRNA is significantly upregulated in SMPG KO mice versus LC mice in normal diet, but increased after the rosiglitazone treatment in both SMPG KO and LC mice. This was surprising but not totally unexpected since TZDs may depict off-target effects independent of PPARγ16–18. No differences on α-AdR expression were observed in SMPG KO mice with or without rosiglitazone treatment (supplemental Figure S7). Therefore, we tested the effects of the highly selective non-TZD PPARγ ligand, GW7845, in VSMC in vitro. Ligand activation by GW7845 as well as overexpression of PPARγ in VSMC significantly reduces the β2-AdR mRNA expression (Figure 4D), suggesting an active role of PPARγ on β2-AdR repression. Furthermore, transient transfection with a mutated PPARγ construct (PPARγDBDmu), in which the DNA-binding domain is deleted, significantly abrogates the inhibitory effect of the GW7845 (1µmol/L) on β2-AdR mRNA expression. In basal conditions in vitro, up-regulation of PPARγ in VSMC causes inhibition of β2-AdR transcription while shRNA-mediated down-regulation of PPARγ leads to increased β2-AdR mRNA in the vehicle-treated cells (supplemental Figure 6). On the other hand, Rosiglitazone treatment (1µmol/L) results in up-regulation of β2-AdR independent of the levels of PPARγ, again indicating the possibility of PPARγ-independent effects (supplemental Figure 6). Taken together, these results strongly indicate that β2-AdR is a novel target of PPARγ transcriptional repression.
Consistent with these observations, we identified a putative PPARγ binding site (PPAR element) in the promoter of the β2-AdR located between nucleotides -1371 and -1382 upstream of the transcription start site. To determine the functionality of this putative PPAR element in the β2-AdR promoter, we first performed electrophoretic mobility shift (EMSA) analysis that confirmed the ability of PPARγ to effectively bind to this putative PPAR element newly identified in the β2-AdR promoter (-1371 and -1382) (see methods online) (supplemental figure 8). Next, chromatin domains in the β2-AdR promoter were scanned for PPARγ binding by chromatin immunoprecipitation (ChIP) analysis. As shown in Figure 5A, PPARγ detectably binds to the site located between -1371 to -1382 bp upstream of the transcription start site (lanes 7 and 8). An isotypic IgG antibody was included as a negative immunoprecipitation control (lanes 5 and 6). The negative control for the ChIP assay detected no binding of PPARγ to flanking regions located 15 kb upstream of β2-AdR promoter, where another putative PPAR binding site was predicted by bioinformatics analysis (lanes 1 to 4). These data strongly suggest that in the context of chromatin, the PPAR site (PPRE) between nucleotides -1371 to -1382 appears to be functional, showing binding of PPARγ to the β2-AdR promoter.
The transcriptional response of the β2-AdR promoter to PPARγ was further characterized using a luciferase reporter in vitro. The specificity of the identified PPRE within the β2-AdR promoter was additionally determined by specific site-directed mutagenesis of this PPAR binding site (see methods online). VSMC were transiently transfected with a 1.4 kb promoter (β2-WT) or the corresponding site-directed mutated promoter on the identified PPRE (β2-Mu). Cells were co-transfected with equivalent amounts of the Flag-PPARγ and then treated with 1µmol/L GW7845. As shown in Figure 5B, ligand activation of PPARγ as well as PPARγ transient overexpression reduced β2-AdR promoter activity, whereas luciferase activity from the mutated promoter did not show responsiveness to GW7845 treatment or PPARγ overexpression, indicating that PPARγ activation repressed β2-AdR expression in vitro.
Results from this study provide important insight on the role of PPARγ on vascular homeostasis by uncovering a potential direct antagonistic transcriptional control of the β2-adrenergic signals leading to a hypotensive phenotype. We used our previously described SM22α-Cre knock-in mice in which the expression of Cre recombinase is driven from the endogenous SM22α promoter to efficiently deplete PPARγ expression in smooth muscle cells12. The rationale for this strategy is two-fold: 1) to avoid the use of conventional transgenesis of SM22α-driven expression of Cre recombinase, which does not control the location of insertion into the mouse genome, and thus the recombinase expression is highly influenced by the surrounding sequences at the integration sites in the genome19; and 2) to ensure that the efficiency of Cre-mediated recombination is fully restricted to the endogenous transcriptional control of the SM22α promoter. In this way, we achieved full absence of PPARγ expression in vascular smooth muscle cells (VSMC), with only partial depletion in heart and stomach, concomitant with partial expression of Cre recombinase in these tissues.
In this experimental model, PPARγ deletion in VSMC has no effect on insulin sensitivity neither causes overt metabolic and cardiac dysfunction in the conditions used here, although our results clearly indicate that lack of PPARγ in VSMC reduces systemic BP during the rest-phase in these animals. Therefore, this BP lowering effect could be a consequence of a direct role of VSMC-dependent PPARγ function in the vasculature. Aortic relaxation was enhanced in response to β-adrenergic agents in SMPG KO mice and as a consequence BP was significantly reduced. Infusion with the β-adrenergic receptor antagonist propranolol increased BP mediated by NE in SMPG KO mice. Using terbutaline, a β2-AdR-specific agonist, SMPG KO mice responded reducing BP, but no significant effect was observed in LC mice, indicating an enhanced sensitivity to β2-AdR agonists in SMPG KO mice. The lack of appreciable different effects of α-adrenergic agents in terms of both aortic contraction and BP lowering indicate that the α-adrenergic response is not contributing to the observed hypotension in SMPG KO mice. This is consistent with the mRNA levels of α-adrenergic receptors not being affected in the SMPG KO animals.
These results provided insights on a novel PPARγ -dependent physiological response leading to the reduction in BP in these mice. The biphasic effect observed in the SMPG KO mice in terms of aortic relaxation as well as in BP indicated the possibility that these animals may have an enhanced β2-AdR response, which is known to mediate vasorelaxation without affecting that of the α-AdR response20. This prompted us to evaluate whether this was the result of a specific PPARγ-dependent transcriptional regulation of the components of the β-adrenergic signaling and uncovered β2-AdR as a novel target of PPARγ subjected to transcriptional repression by this nuclear receptor in the vessels in SMC in vivo. Increased levels of β2-AdR expression in SMPG KO mice renders them further susceptible to β2-adrenergic receptor agonist-mediated relaxation compared to control mice. These observations are consistent with a previously proposed model for the BP lowering effects of PPARγ KO mice. PPARγ depletion as well as activation with TZDs may lead to the release of yet unidentified co-repressors located on the promoter of genes involved in the control of the vascular tone10,16. In this report, we have identified the β2-AdR as such a target that could account, at least in part, for the enhanced hypotensive phenotype in PPARγ knock-out mice. Whether the PPARγ repressive effects result from recruitment of specific co-repressors and the dynamics of the transcriptional events modulating β2-AdR expression, remain to be determined. Furthermore, the observed increased expression of β2-AdR by rosiglitazone in a manner that is independent of PPARγ, uncovered an off-target effect of TZDs.
The occurrence of human mutations in the PPARγ gene with dominant-negative function was first described by Barroso et al7. Patients with such mutations develop insulin resistance, early-onset of diabetes and premature hypertension with the consequence of a severe cardiovascular dysfunction7. Endothelial dysfunction appears to be the primary cause of the development of an early hypertensive phenotype as experimentally determined through impaired Ach-mediated vasorelaxation in animal models expressing such PPARγ dominant-negative mutations in mice21–23. Moreover, TZDs treatment significantly decreases BP in experimental models of hypertension (e.g. Ang II infusion, spontaneous and DOCA-salt hypertensive rats, transgenic mice overexpression of the renin-angiotensin-aldosterone system)24–29. But specifically, lack of PPARγ in endothelial cells (EC) exacerbates hypertension, a phenotype that could not be reversed upon rosiglitazone treatment11, indicating that interference with PPARγ function may primarily cause endothelial damage, a key step for the occurrence of a hypertensive phenotype. In our SMPG KO mice, endothelial function is preserved, as indicated by the fact that Ach-mediated vasorelaxation is not altered in these mice compared to LC mice, which argues that the differential phenotype of hypotension in these animals depends primarily on VSMC-mediated vascular tone.
A recent report indicated that VSMC-specific PPARγ KO mice obtained by crossing PPARγflox/flox mice with SM22α-Cre transgenic mice, developed spontaneous pulmonary artery hypertension associated with right ventricular hypertrophy14. Deregulation of a bone morphogenetic protein/PPARγ/apoE signaling axis is proposed as the mechanism behind pulmonary hypertension. Consistently with this, PPARγ activation by rosiglitazone reduces pulmonary hypertension in ApoE KO mice in a high-fat diet and reduces pulmonary artery atherosclerosis in the insulin resistant animals. Yet, systemic BP in that experimental model was reported as unaffected30. Loss of PPARγ expression is also observed in the lungs of patients with severe pulmonary hypertension31. Recently, SM22α-Cre-driven PPARγ knock-out mice were reported to show loss of circadian rhythmicity in BP regulation with increased BP and heart rate during the light cycle15. This animal model shows notable differences from ours at several levels. First, the mechanism of transgenesis is essentially different: Cre-knock-in model versus transgenic SM22α-Cre expression. It is well-described that the expression of the endogenous SM22α gene is different when in its native genomic context than in transgenic models. Recently, Long et al. summarized some molecular clues on the regulation of several cis-regulatory elements for the spatial and temporal control of gene expression, giving emphasis on SMC-specific markers including SM22α32. These effects may translate in differential recombinase activity leading to significant PPARγ depletion in other tissues, including heart and kidney, with increased heart rate and evident sympathoadrenal dysfunction in those transgenic animals15, all of which may contribute to the differences with our model, in which no altered heart function was observed. Further differences include a loss of BMAL1 circadian expression in SM22α-Cre-driven PPARγ knock-out mice identifying this circadian gene as a target of PPARγ in both EC and VSMC15. In our experimental model, SMPG KO displayed shifted circadian BMAL1 expression compared to the littermate control mice (supplemental Figure S9). BMAL1 knockout mice displayed altered circadian rhythmicity of BP with an overall reduction of the mean arterial pressure33, which disagrees to the data that increased BP was observed in SM22α-Cre-driven PPARγ knock-out mice with impaired BMAL1 expression15. BP regulation is a multifactorial trait with combinatorial contributions of different factors in response to changing metabolic conditions. These conflicting observations suggest that it is likely that neither BMAL1 nor the β2-AdR described here alone will be the sole contributing factors for the changes in BP regulation observed in these intrinsically different PPARγ knockout models. Thus, in our SMPG KO mice, loss-of-PPARγ function in VSMC did not affect the circadian periodicity of BP regulation, whereas the enhanced β2-AdR expression contributed to the observed diurnal decrease of BP. Interestingly, a recent report did not show a loss of BP circadian rythmicity in the β1- and β2-AdR double KO mice in spite of clear effects on vascular tone34. Therefore, further studies aimed to dissect the molecular mechanisms and signal transduction pathways linking PPARγ with β-AdR, sympathoadrenal function and BP will be required.
The data presented here clearly indicates that lack of PPARγ in VSMC results in a significant reduction in systemic BP associated to increased expression of the β2-AdR concomitant with increased sensitization to β-adrenergic agonists.
Hypertension control is one of the major goals in the management of cardiovascular disease. The occurrence of hypertension in diabetic patients further aggravates their cardiovascular outcomes. In the clinical practice, thiazolidinediones (TZDs), agonists of the PPARγ stand out as one of the therapeutic drugs for diabetes control. They act as insulin sensitizers but also improve hypertension in diabetic patients. Experimental models have been developed to elucidate the molecular mechanisms by which PPARγ mediates these processes. Endothelial-, cardiac-, and smooth muscle-tissue specific gain-and loss-of PPARγ function animal models display a considerable range of effects on various aspects of the cardiovascular pathophysiology, such as pulmonary hypertension, cardiac remodeling or high-fat diet induced atherosclerosis. Here we describe that vascular smooth muscle cell-selective PPARγ deletion leads to systemic hypotension with a circadian component and we specifically identify the β2-adrenergic receptor (β2-AdR) as a novel gene subjected to PPARγ-dependent repression in the vasculature which will undoubtedly open new paradigms in the regulation of blood pressure and vascular tone.
This work was partially funded by National Institutes of Health (HL68878, HL75397 and HL89544 to Y.E.C.). L.C. and J.Z. were supported by American Heart Association Midwest Affiliate Fellowship (0625705Z) and National Career Development Grant (0835237N) respectively. Y.E.C. is an established investigator of American Heart Association (0840025N).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Authors have no conflict of interest to disclose.