*3.2. Are Inositols Direct Insulin Mimetics Rather than Insulin Sensitizers?*

Apart from the proposed model of IPGs in insulin signaling, some early in vitro studies of inositol supplementation have shown the direct effect of inositol isomers in insulin signaling apart from IPG activity. It was shown that a 1 mM dose of DCI, DPIN, l-*chiro*-inositol (LCI; *cis*-1,2,4-*trans*-3,5,6-ciclohexanehexol), EI, and muco-I stimulated glucose uptake in rat L6 myotubes in vitro and also promoted that translocation of GLUT4 to the plasma membrane in L6 myotubes in vitro and the skeletal muscles of rats ex vivo to a similar extent as 100 nM of insulin [101]. It should be noted that cells were grown in a medium supplemented with fetal bovine serum (FBS) and starved 18 h prior to the glucose uptake assay with a medium containing 0.2% FBS, so the possibility of insulin traces remaining in the cell culture media should not be ruled out, as these could account for the inositol effect on GLUT4 translocation and glucose uptake at the given concentrations. MI was also supplemented at same concentrations and did not elicit any insulin-mediated response on glucose uptake [101].

When administered to endothelial cells in vitro, both 1 mM of MI and DCI promoted an increased phosphorylation of Akt, ERK1, and ERK2 in human vascular endothelial cells (HUVEC) to a greater extent than 100 nM of insulin [102]. Since cells were serum-starved before MI and DCI treatment, MI and DCI induced kinase phosphorylation in the absence of insulin. It is noteworthy that the Ras/MEK/ERK signaling pathway was also involved, as inositols are regarded to exert insulin-sensitizing effects on the PI3K/Akt/mTOR signaling pathway.

Our recent study has shown that the administration of DPIN to fasting rats promotes a significant reduction of circulating insulin without affecting plasma glucose levels [103]. These results may imply a direct action of DPIN on insulin signaling. An increase in ghrelin levels was also observed upon DPIN administration, which could account for the decreased secretion of insulin in pancreatic β cells [103].

These results show that inositols may act on upstream regulators of insulin signaling. Some experiments agree with this hypothesis. Sequoyitol pretreatment enhances insulin signaling with increased phosphorylation of IRS1 and Akt in HepG2 hepatocytes and 3T3-L1 adipocytes [104]. Interestingly, sequoyitol pretreatment also reverses decreased IR autophosphorylation in the presence of tumor necrosis factor (TNF-α), a well described inhibitor of IR activity [104,105]. TNF-α is known to also inhibit SMIT expression in cultured endothelial cells [106]. Inositol depletion might partially explain the insulin-sensitizing effect of sequoyitol on insulin-resistant cells but does not account for enhanced IR autophosphorylation.

Increased IR autophosphorylation has also been observed in primary hippocampal neurons from rats when administered 100 μM of DCI, DPIN, or INS-2 in a similar way to 1 μM insulin treatment [107]. Moreover, after media replacement with a serum-free HEPES buffer, DCI administration has been shown to promote IR internalization, a mechanism required for ERK activation, in a similar way to insulin [107]. Given that insulin was depleted from the media, again, this mechanism of IR trafficking from dendrites to soma elicits a direct effect of DCI as a free inositol in insulin signaling.

The way free inositols participate in the insulin signaling pathway remains unknown. Given that PLD levels are relatively high in all tissues, insulin stimulation of GPI-PLD might not be crucial for its activity. Somehow, a stimulus would be needed to increase IPG production, which would rely on the basal activity of GPI-PLD [108]. Based on the results obtained both in vitro and in vivo, the group of Ashida suggested an insulin-independent mechanism, implying the activation of PI3K and/or AMPK [101,109]. It has been described that INS-2 allosterically modulates PP2Cα [72], which is known to dephosphorylate and activate PI3K [74] and inactivate AMPK [75]. IPG production could account for GLUT4 translocation and enhanced glucose uptake in muscles. However, inositol supply and increased IPG production does not account for the direct actions of inositols regarding IR autophosphorylation, since their target starts insulin signaling downstream of PI3K and possibly IRS1.

Although unlikely, some hypotheses have arisen regarding the means of IR activity. One of them is the possible allosteric modulation of IR by inositols. A study has identified the molecular docking of MI with active sites of PPARγ, GLUT4, and IR [108]. However, contrasting docking analysis requires specific studies of protein–molecule inhibition or inhibition of the protein target to determine the correlation of effective interactions in a biological system and in silico predictions. Moreover, the fact that MI is predicted to interact with different molecules of the same signaling pathway undermines the reliability of this mechanism. Another possibility could involve the repression of negative regulators of IR. Protein tyrosine phosphatase 1B (PTP1B) is a well-described inhibitor of IR tyrosine kinase activity. PTP1B is activated under different stimuli, including the presence of TNF-α [110]. Other known inhibitors of IR are c-Jun N-terminal protein kinase (JNK) or suppressor of cytokine signaling 3 (SOCS3). It is still unknown if inositols may interact and downregulate one of the inhibitory pathways of IR autophosphorylation,

Mechanisms of the insulin signaling pathway may involve different downstream elements, but they all share a common activation of IRS1/2, PI3K, and Akt. In order to elucidate the target for free inositols, we propose an in vitro study of cultured hepatocytes or myocytes, as highly-responsive cells to insulin stimulus, and the blockade, one by one, of elements composing the cascade of the insulin signaling pathway (IR, IRS1, PI3K, PDK, Akt, mTOR, and AMPK) in a top-down manner. This could be easily achieved via the use of small interfering RNA (siRNA) and transient siRNA-mediated knockdown of IR and their downstream elements [111,112]. Measuring the GLUT4-translocation response of insulin-depleted cells to inositol addition could determine the exact point at which inositols enhance IR signaling in the absence of insulin. In vitro studies let us control medium conditions and eliminate external elements that could interfere with the insulin sensitivity. Alternatively, the use of specific inhibitors for each element could be considered. Identification of inositol targets would make way for further analysis of the exact mechanism of interaction, as well cyclitols, regarding their specificity.
