Human leukemic monocytes THP-1 (ATCC TIB 202) from American Type Culture Collection (Rockville, MD, United States) were grown in a suspension containing RPMI-1640 medium with 10% FBS, 100 μg/ml streptomycin, and 100 U/ml penicillin, in a humidified atmosphere of 5% CO₂ at 37°C (Pedersen et al., 2007; Lombardo et al., 2013). These cells show a large, round, single-cell morphology. The THP-1 monocytes were passaged twice a week by 1:4 dilutions and reseeded; only cells from passages no. 7–23 were used for the experiments.

Migration experiments were carried out by the use of Transwell (Corning) with an 8-μm polycarbonate membrane, 6.5-mm insert 24-well plate with serum-free RPMI-1640 medium containing 0.2% bovine serum albumin (BSA) in both chambers, for 4 h at 37°C (De Vito et al., 2012). THP-1 cells (about 200,000 cells/well) were placed in the upper chamber with RGD (10 μM), tetrac (10 μM), Ab-αvβ3 (8 μg/ml), wortmannin (100 nM), and PD98059 (10 μM), while T₃ (10^–7–10^–9 M), T₄ (10^–7–10^–11 M), MCP-1 (100 ng/ml), and IGF-1 (10 nM) were added to the bottom chamber. All αvβ3 inhibitors and L-NAME (1 mM) and GSNO (0.5 mM) were preincubated for 20 min at 37°C. After 4 h of incubation, at 37°C, cells migrated (from the top to the bottom part of the chamber) were counted with a modified Neubauer chamber. At the beginning, migration experiments were carried out with RPMI as medium and in the presence of serum (0.2% FBS) (24). Afterward, since data from literature (Tsaur et al., 2012) show that serum can act as a chemoattractant, we decided to perform experiments in a serum-free medium.

Cells were seeded in 60 mm × 15 mm Petri dishes with RPMI-1640 and stimulated with RGD, T₄, and IGF-1 the day after the seeding. Cells were counted after 72 h. The role of integrin αvβ3 on the proliferation of THP-1 cells was studied using RGD peptide (10 μM) as the integrin αvβ3 inhibitor. RGD peptide was preincubated 20 min before adding T₄ (100 nM), IGF-1 (10 nM), and MCP-1 (100 ng/ml). Moreover, we studied PI3K and MAPK pathways by the use of wortmannin (100 nM) and PD98059 (10 μM) preincubated 20 min before adding T₄, IGF-1, and MCP-1. Cells were counted with an optical microscope with a Neubauer chamber (Incerpi et al., 2014).

The production of nitric oxide (NO) in THP-1 cells was measured using DAF-2DA probe. DAF-2DA is a sensitive fluorescent indicator for the detection and bioimaging of NO. It is a cell-permeable derivative of DAF-2. Upon its entry into a cell, DAF-2DA is transformed into the less cell-permeable derivate, DAF-2, by cellular esterases, thus it prevents loss of the signal due to diffusion of the molecule from the cell. In the presence of oxygen, DAF-2 reacts with NO to yield the highly fluorescent triazolofluorescein (DAF-2T). At the time of the experiment, the cell suspension (8 × 10⁶ cells) was washed (1,200 rpm, 10 min) three times with 8 ml of PBS containing 5.0 mM glucose (90 mg/100 ml) to remove the serum, which may affect the action of the fluorescent probe. The incubation with the probe DAF-2DA at the final concentration of 5 μM was carried out for 30 min in the dark at 37°C (López-Figueroa et al., 2000). At the end of the incubation, cells were washed three times, centrifuged at 1,200 rpm for 5 min, and the final cell pellets were resuspended in 4 ml of PBS-glucose. Before the experiments, cells were recovered at 37°C for 30 min in the dark. Some samples of cells were preincubated with L-NAME (1 mM) for 30 min at 37°C. After the recovery, the cells were seeded in a 96-multiwell (200 μl/well) and stimulated with T₄ (100 nM), MCP-1 (100 ng/ml), ionomycin (2 μM), and LPS (1 mg/ml). Intracellular fluorescence was measured with spectrofluorometer at 37°C (Jasco, Analytical Instruments). Excitation and emission wavelengths were set at 495 and 515 nm, respectively, using 5- and 10-nm slits for the light paths. The measurements were performed at 0, 1, 2, 3, and 4 h.

The measurement of nitrite production was carried out by the Griess assay, a common method for the indirect determination of NO by the spectrophotometric measurement of nitrites. This method requires that nitrates are firstly reduced to nitrite and then determined by the Griess reaction (Ridnour et al., 2000; Colasanti et al., 2004). Herein, cells were seeded in a 24-multiwell and treated with LPS (1 μ g/ml), ionomycin (2 μM), MCP-1 (100 ng/ml), thyroxine (100 nM), and l-NAME (1 mM). Nitrite concentrations were tested on THP-1 cells at 4, 24, and 48 h, as described. At different time points, cell suspensions were removed and washed, while supernatants were frozen. In a 96-multiwell plate, a known volume of premixed Griess reagent (1% sulfanilamide, 0.1% naphthylethylenediamine dihydrochloride, and 2.5% H₃PO₄) was added to 70 μ of the supernatant of each sample. The reaction was carried out for 15 min at room temperature in the dark. At the end of the reaction, absorbance was measured at 550 nm using ELISA reader (Packard Fusion Microplate Reader). The concentration of nitrite in the supernatants was extrapolated using the calibration curve based on the known concentrations of sodium nitrite (NaNO₂ 0–50 μ M) reacted with the Griess reagent.

THP-1 cells were washed twice with phosphate-buffered saline (PBS), pH 7.4, and lysed in 50 mM Tris, pH 7.4, 150 mM NaCl, 1% Triton-X 100, 0.25% sodium deoxycholate, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM ethylene glycol tetraacetic acid (EGTA), 0.5% non-ionic detergent IGEPAL CA-630 (Sigma-Aldrich), 1 mM sodium orthovanadate, 20 mM sodium fluoride, 1 μg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml pepstatin A, for 30 min in ice. Whole-cell lysates were centrifuged at 6,000 g for 10 min at 4°C, and the supernatants were frozen at −80°C. The protein concentrations of cell extracts were determined by the Lowry protein assay (Lowry et al., 1951). The aliquots of cell extracts containing 30 μg of total proteins were resolved on 7% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred by electroblotting them on nitrocellulose membranes (Whatman GmbH, Dassel, Germany) overnight at 35 V with the Bio-Rad Trans-Blot apparatus. For the immunoassays, the membranes were blocked in 3% BSA fraction V (Sigma-Aldrich, St. Louis, MO, United States) in Tween 20-TBS (TTBS)/EDTA (10 mM Tris, pH 7.4, 100 mM NaCl, 1 mM EDTA, 0.1% Tween 20) for 1 h at room temperature. To analyze tyrosine phosphorylation of STAT1, immunoblotting was performed using specific antibodies anti-phosphotyrosine 701-STAT1 diluted in 1% BSA/TTBS-EDTA and then incubated overnight at 4°C. Expression levels of tyrosine phosphorylation of STAT1 were evaluated using corresponding specific antibodies anti-STAT1 or IgG anti-actin as the internal loading control. Antibodies used in the different immunoblottings diluted in 1% BSA/TTBS-EDTA were as follows: rabbit polyclonal antibodies anti-phospho-STAT1 (Tyr701), mouse monoclonal antibodies anti-STAT1, anti-actin rabbit polyclonal antibody. Immune complexes were detected with horseradish peroxidase-conjugated goat anti-rabbit or goat anti-mouse antiserum followed by enhanced chemiluminescence reaction (ECL LiteAblot PLUS, Euro Clone SpA) (Table 1). To reprobe membranes with antibodies having different specificities, nitrocellulose membranes were stripped for 5 min at room temperature with restoring Western blot stripping buffer (Pierce, Rockford, IL, United States) and then extensively washed with TTBS/EDTA.

In order to analyze the mechanism of interaction between T₄ and αvβ3 integrin, in silico molecular docking simulations have been carried out using the software AutoDock Vina (Trott and Olson, 2010; Di Muzio et al., 2017). This bioinformatics technique is used to predict the way by which two molecules, usually a macromolecule and a small molecule ligand, bind to each other in relation to their chemical structures. We have focused our attention on the interaction between T₄ and the αvβ3 integrin in its inactive form, both in the presence and in the absence of the peptide RGD, using the three-dimensional structure deposited in the Protein Data Bank with the code 1L5G (Xiong et al., 2002).

The results reported as means ± SD were analyzed with Student’s t-test and with one-way analysis of variance (ANOVA) followed by post hoc Bonferroni’s multiple comparison test. Differences were considered significant at p < 0.05.

First of all, we evaluated the effect of T₄ on THP-1 cell migration in the presence of both MCP-1 and IGF-1. At the beginning, migration experiments were carried out using RPMI as medium and in the presence of serum (0.2% FBS). Afterward, since serum can act as a chemoattractant, we decided to carry out experiments in a serum-free medium. T₄ alone did not affect cell migration stimulated by either MCP-1 or IGF-1; the stimulatory effect of MCP-1 was higher in comparison to that of IGF-1, according to the key role of MCP-1 as a chemokine (Figure 1). T₄ inhibited cell migration activated by either MCP-1 or IGF-1 by about 50%.

Experiments of dose–response of T₄ (10^–11–10^–7 M) show that the inhibitory effect of the hormone on the migration stimulated by MCP-1 was dose-dependent: T₄ (10^–7 M) prevented cell migration induced by either MCP-1 or IGF-1. The inhibitory effect of the thyroid hormone was significant starting at 10^–10 M (p < 0.01 with respect to MCP-1; Figure 2A). All experiments to be shown in the following were carried out with T₄ 10^–7 M, the physiological concentration of T₄. We also studied the effect of T₃ on cell migration induced by MCP-1 (Figure 2B). Our data show that T₃ alone, at different concentrations (10^–7–10^–9 M) did not affect cell migration, but when given together with MCP-1, T₃, as well as T₄, was able to prevent cell migration in a dose-dependent way starting at 10^–9 M (p < 0.001 with respect to MCP-1), but its inhibitory effect was lower with respect to that of T₄, although significant at all concentrations tested (Figure 2B); this may be due to the lower affinity of T₃ for the integrin with respect to T₄.

The αvβ3 integrin is involved in the invasion and migration of different cells (Cohen et al., 2014; Liu et al., 2016), besides being the plasma membrane receptor for thyroid hormone (Bergh et al., 2005) and a co-receptor for IGF-1 (Clemmons et al., 2007). Therefore, we also tested the effect of thyroid hormone on the migration induced by IGF-1 (10 nM). The possible role of integrin αvβ3 was studied using three well-known integrin αvβ3 inhibitors: Arg-Gly-Asp peptide (RGD), tetrac, a metabolite of thyroid hormone and a probe for the integrin αvβ3 and a human monoclonal antibody-αvβ3, LM-609. In the presence of IGF-1, the inhibitors of integrin αvβ3 were able to prevent the inhibitory effect of the hormone on the IGF-1-induced migration (Figure 3A). At variance with this, in the presence of MCP-1, T₄ and the inhibitors of the integrin, RGD, tetrac, or antibody-αvβ3, gave rise to a potentiation of the inhibitory effect of the migration stimulated by MCP-1 (Figure 3B).

We carried out experiments with T₃, to verify whether there were differences in the behavior of the two hormones, since T₃ binds integrin with about two orders of magnitude in the affinity lower with respect to T₄ (Bergh et al., 2005). The results obtained in the presence of T₃, in the cell migration using IGF-1 as a chemoattractant, are similar to those of T₄ in the presence of IGF-1, and all of the αvβ3 integrin inhibitors prevented the effect of T₃, analogously to T₄ (Figure 3C).

T₃ inhibited the migration induced by MCP-1, with a lower effect in comparison to T₄ (Figure 3D) and a different behavior with respect to T₄ (Figure 3B). In fact, all αvβ3 inhibitors were able to prevent the migration induced by MCP1, but T₃ + MCP-1 behaved in a different way in the presence of either RGD or tetrac or the antibody. In particular, two levels of inhibition of cell migration appear to be present in this type of experiments. In fact, the first level of inhibition, although significant with respect to MCP-1 alone, was found for MCP-1 in the presence of T₃, RGD, T₃ + RGD, or tetrac, although with a trend for the last to a higher effect. The second level of inhibition of migration was more effective and brought back the migration to basal level when MCP-1 was in the presence of T₃ and either tetrac or antibody or T₃ plus antibody (Figure 3D).

The signal transduction of both T₄ and MCP-1 shows the involvement of PI3K pathway for MCP-1 and MAPK pathway downstream the interaction of T₄ with integrin αvβ3 (Cohen et al., 2014; Lee et al., 2015; Shinderman-Maman et al., 2016). For this reason, we focused our study on these two pathways, by a pharmacological approach, using MAPK and PI3K inhibitors, PD98059 and wortmannin, respectively. IGF-1 stimulated cell migration through PI3K pathway. Interestingly, in the presence of the MAPK inhibitor, PD98059, T₄ did not inhibit the cell migration induced by IGF-1 (Figure 4A). These results indicate that T₄ prevents cell migration stimulated by IGF-1 through the MAPK pathway. MCP-1 induced cell migration through the PI3K pathway, since in the presence of wortmannin, the MCP-1-induced cell migration was inhibited up to the basal level (Figure 4B). As to the inhibitory effect of the thyroid hormone, T₃, the PI3K pathway is also involved. In fact, wortmannin prevented the inhibitory effect of T₃ on the cell migration stimulated by MCP-1 (Figure 5B). T₃ and T₄ behave in a similar way as to the role of the MAPK pathway in the inhibitory effect on migration induced by IGF-1 (Figures 4A, 5A). The results obtained from the study of the signal transduction indicate that both MCP-1 and IGF-1 induced cell migration through the PI3K pathway, but T₃ and T₄ acted in a different way: T₄ prevented cell migration induced by IGF-1 through MAPK, while this pathway is not involved in the inhibitory effect of T₃ in cell migration induced by MCP-1, and the presence of PD98059 did not significantly affect the inhibition of T₃ on MCP-1-stimulated cell migration. Wortmannin, instead, did prevent the inhibitory effect of T₃ on the migration induced by MCP-1, in agreement with the reported binding of T₃ to the S1 site of the integrin αvβ3 and downstream activation of the PI3K pathway (p < 0.05; Figure 5B; Lin et al., 2009).

IGF-1 enhances cell proliferation and survival (Laron, 2001; Hakuno and Takahashi, 2018), and IGF-1 significantly increased the proliferation of THP-1 monocytes with respect to control, and T₄ inhibited it; this effect was due to αvβ3 integrin (Figure 6A). In particular, we evaluated the role of integrin αvβ3 in cell proliferation; RGD alone did not affect cell proliferation but was able to remove the inhibitory effect of thyroid hormone (Figure 6A). These results are in agreement with data obtained from cell migration experiments, since both hormone responses, cell migration and proliferation, are mediated by integrin αvβ3. After evaluating the inhibitory effect of T₄ and the role of integrin αvβ3 in cell proliferation, we studied whether MAPK and PI3K pathways might be involved in this process by a pharmacological approach. We previously published that IGF-1 stimulated L6 myoblast proliferation through PI3K pathway, while the thyroid hormone prevented this process through the MAPK pathway (Incerpi et al., 2014). Moreover, as shown before (Figure 6B), these two pathways are involved in THP-1 proliferation induced by IGF-1. The results obtained are similar to those of migration experiments: IGF-1 stimulated cell proliferation through PI3K, as already reported in the literature (Baserga et al., 1997; Riedemann and Macaulay, 2006; Incerpi et al., 2014), whereas the inhibitory effect of T₄ in cell proliferation was mediated by MAPK, since PD98059, in the presence of IGF-1 and T₄, was able to revert the inhibitory effect of the thyroid hormone (Figure 6B).

MCP-1 also stimulated THP-1 cell proliferation, and the effect was inhibited by RGD, indicating that also, in this case, we have a genomic response mediated by the plasma membrane integrin αvβ3. Again, RGD was unable to bring back the inhibition of proliferation induced by T₄ but eventually increased it, a response parallel to the results on cell migration carried out with this chemokine (Figure 6C).

Migration of monocytes is required for routine immunological surveillance of tissues and their entry into inflamed sites, and NO plays a key role during inflammation (Sharma et al., 2007). Therefore, we studied the possible involvement of NO in the migration induced by MCP-1, as already reported (Biswas et al., 2001), and we tried to understand whether the effect of thyroid hormones might be related to a modulation of nitrite or reactive oxygen species (ROS) production. To this aim, we carried out experiments of cell migration using the NO inhibitor, L-NAME (1 mM), and an NO donor, GSNO (0.5 mM). The results of experiments carried out, with MCP-1 and T₄, show that L-NAME was able to potentiate the inhibitory effect of T₄ in the presence of MCP-1. As to the NO donor, GSNO alone stimulated cell migration with respect to control (p < 0.01), but GSNO in the presence of T₄ was able to bring the migration back to the basal level. GSNO did not affect the migration induced by MCP-1 (Figure 7A).

To assess the possible role of ROS, we used diphenylene iodonium (DPI), a nicotinamide adenine dinucleotide phosphate (NADPH) oxidase inhibitor, and a metabolite of thyroid hormones, tetrac, known to inhibit the interaction of the hormone with the integrin αvβ3 through binding to the RGD site or in its close proximity. DPI inhibits the migration induced by MCP-1, pointing to an involvement of the NADPH oxidase and ROS production in the cell migration stimulated by MCP-1. In the presence of DPI, T₄, and tetrac, we found a potentiation of the inhibitory effect on cell migration stimulated by MCP-1 (Figure 7B). These results indicate the involvement of ROS and NO in the cell migration activated by MCP-1 and perhaps a crosstalk between NADPH oxidase and integrin αvβ3, as previously hypothesized (Chuang et al., 2004; De Vito et al., 2012; Gnocchi et al., 2012). The high level of metabolic activity of the cells results in an increase of ROS that makes the antioxidant defense an important factor (Lombardo et al., 2013). We also studied the effect of baicalein, a flavonoid from the roots of Scutellaria baicalensis, on THP-1 cell migration. Baicalein is a strong antioxidant also at very low concentrations in THP-1 monocytes, reported to be able to modulate also NO production (Seo et al., 2011; Lombardo et al., 2013; Caioli et al., 2016). Baicalein was able to inhibit significantly the cell migration induced by MCP-1, as well as DPI (Figures 7B,C). In addition, the inhibitory effect of baicalein was significantly potentiated in the condition T₄ + MCP-1 + tetrac, similar to that of DPI (Figures 7B,C).

We also evaluated the production of NO in THP-1 human leukemic monocytes using the DAF-2DA fluorescent probe. Experiments were carried out, using ionomycin and LPS as positive controls, at 0, 1, 2, and 4 h. There was an increase of reactive nitrogen species (RNS) from 0 to 4 h, but we did not observe a significant difference between samples treated in the same time frame (Figure 8A). After that, the measurement of nitrite production, in THP-1 cells, was carried out by the Griess reaction (Figures 8B,C). This method was used to obtain a quantitative measurement of NO at 4–24 and 48 h after treatment. Our results show that the nitrite concentration was almost the same from 4 to 48 h and was very low (<5 μM). Although the results obtained suggest an involvement of NO in THP-1 cell migration, it was not easy to determine by the use of the fluorescent probe and the Griess assay. Probably, the THP-1 cells produce very little NO or not enough to be detected by the methods used in our laboratory.

STAT-1α is a signal transducer and activator of transcription that mediates cellular responses to IFNs, cytokines, and other growth factors. T₄ activates STAT (Lin et al., 1999a, b; Shih et al., 2004; Davis et al., 2005; Staab et al., 2012); therefore, we studied the possible modulation of STAT-1α in THP-1 monocytes after stimulation with T₄ in the presence and absence of either MCP-1 or IGF-1 through a Western blot analysis at different times: 30 min, 1 h, 2 h, and 4 h, the time of migration experiments. Our results were normalized with respect to the level of expression of STAT1, which was constant for all conditions and times. After 30 min of stimulation, there was a significant increase of STAT-1α tyrosine phosphorylation (P-Y701-STAT1) in all conditions, with respect to control, but after 1 h, there was a decrease that reached the basal value at 2 h to increase again at 4 h, in good agreement with similar values reported for different cells (Lin et al., 1999a; Figure 9). Interestingly, T₄ was able to prevent STAT-1α tyrosine phosphorylation induced by IGF-1 at 2 h that was still present after 4 h. At this time, we observed a second wave of STAT-1α tyrosine phosphorylation in the other conditions, with respect to 30 min, indicating a more stable activation. P-Y701-STAT-1α is an important transcription factor related to the activation of inducible nitric oxide synthase (iNOS) in macrophages as reported (Lin et al., 1999a, b; Arjcharoen et al., 2007).

In order to analyze the mechanism of interaction between T₄ and αvβ3 integrin, in silico molecular docking simulations have been carried out using the software AutoDock Vina (Trott and Olson, 2010; Di Muzio et al., 2017). We have focused on the interaction between T₄ and the αvβ3 integrin in its inactive form, both in the presence and in the absence of the peptide RGD, using the three-dimensional structure deposited in the Protein Data Bank with the code 1L5G (Xiong et al., 2002). Our results show that T₄ mainly binds at the interface between the two αvβ3 integrin subunits in the basal part of the macromolecule next to the cell membrane, a site that is different from the RGD binding site. Interestingly, T₄ was able to bind to this site both in the presence and in the absence of the RGD peptide. In particular, as mentioned above, the putative T₄ binding site is located in the extracellular space but very close to the plasma membrane, and interacting with both integrin subunits may stabilize the inactive conformation of the αvβ3 integrin (Figure 10). In fact, during integrin activation, the basal domains of the two integrin subunits must move away from each other (Xiong et al., 2002), a process that would be inhibited by T₄ binding in this region, at the interface between the two subunits. These results suggest a mechanistic interpretation of the data of migration induced by MCP-1 in the presence of T₄ and the αvβ3 integrin inhibitor RGD. In fact, experimental data of migration show that both RGD and T₄ inhibited the migration induced by MCP-1, and there was a significant potentiation in the inhibition of the migration when T₄ was in the presence of RGD or other inhibitors of the integrin. As mentioned above, this effect could be due to the presence on the integrin of a binding site for the hormone, different from the RGD site. At variance with results obtained with MCP-1, where both T₄ and RGD inhibit the migration induced by MCP-1 resulting in a potentiation of the inhibitory effect, data of migration in the presence of IGF-1 are different, but since no crystallographic structure is available for the active form of αvβ3 integrin, molecular docking simulations could not be performed on this integrin state. Our data on migration induced by IGF-1 in THP-1 monocytes are in agreement with previous data, published from different laboratories including our own, on the inhibitory effect of T₄ on the IGF-1-mediated actions in different cells (Incerpi et al., 2014) and papers from other groups, where it is reported that preincubation with RGD, but also tetrac and the Ab-αvβ3, prevented the hormone effect (Jones et al., 1996; Moeller et al., 2006; Moeller and Broecker-Preuss, 2011).