Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Sep 12.
Published in final edited form as: Eur J Pharmacol. 2015 Jan 29;751:59–66. doi: 10.1016/j.ejphar.2015.01.032

Simple modifications to Methimazole that enhance its inhibitory effect on Tumor Necrosis Factor-α-induced Vascular Cell Adhesion Molecule-1 expression by human endothelial cells

Anuja Alapati a,b, Sudhir P Deosarkar c, Olivia L Lanier b, Chunyan Qi a,b, Grady E Carlson b, Monica M Burdick a,b, Frank L Schwartz d, Kelly D McCall a,d, Stephen C Bergmeier a,e, Douglas J Goetz a,b,f
PMCID: PMC5019189  NIHMSID: NIHMS659634  PMID: 25641748

Abstract

The expression of vascular cell adhesion molecule-1 (VCAM-1) on the vascular endothelium can be increased by pro-inflammatory cytokines [e.g. tumor necrosis factor – α (TNF-α)]. VCAM-1 contributes to leukocyte adhesion to, and emigration from, the vasculature which is a key aspect of pathological inflammation. As such, a promising therapeutic approach for pathological inflammation is to inhibit the expression of VCAM-1. Methimazole [3-methyl-1, 3 imidazole-2 thione (MMI)] is routinely used for the treatment of Graves’ disease and patients treated with MMI have decreased levels of circulating VCAM-1. In this study we used cultured human umbilical vein endothelial cells (HUVEC) to investigate the effect of MMI structural modifications on TNF-α induced VCAM-1 expression. We found that addition of a phenyl ring at the 4-nitrogen of MMI yields a compound that is significantly more potent than MMI at inhibiting 24 h TNF-α-induced VCAM-1 protein expression. Addition of a para methoxy to the appended phenyl group increases the inhibition while substitution of a thiazole ring for an imidazole ring in the phenyl derivatives yields no clear difference in inhibition. Addition of the phenyl ring to MMI appears to increase toxicity as does substitution of a thiazole ring for an imidazole ring in the phenyl MMI derivatives. Each of the compounds reduced TNF-α-induced VCAM-1 mRNA expression and had a functional inhibitory effect, i.e. each inhibited monocytic cell adhesion to 24 h TNF-α-activated HUVEC under fluid flow conditions. Combined, these studies provide important insights into the design of MMI-related anti-inflammatory compounds.

Keywords: Methimazole, Inflammation, Endothelial cell, Adhesion, Leukocyte

1. Introduction

Leukocyte adhesion to the vascular endothelium plays a crucial role in inflammation and is mediated, in part, by endothelial cell adhesion molecules [ECAMs; e.g., VCAM-1, ICAM-1, and E-selectin] (Luscinskas and Gimbrone, 1996). Pro-inflammatory cytokines such as tumor necrosis factor-α (TNF-α) up-regulate ECAM expression [e.g. TNF-α induces ECAM expression on cultured endothelial cells (Bevilacqua, 1993)]. While elevated ECAM expression is important for a normal inflammatory response, the ECAMs can be aberrantly expressed and contribute to pathological inflammation (Luscinskas and Gimbrone, 1996). For example, VCAM-1 has been implicated in many inflammatory diseases such as atherosclerosis, and arthritis (Carter et al., 2002; Cybulsky et al., 2001). Thus, compounds which inhibit cytokine-induced VCAM-1 expression could aid in treating pathological inflammation. Methimazole [3-methyl-1, 3 imidazole-2 thione (MMI)] is routinely used clinically for the treatment of Graves’ disease (Giuliani et al., 2010) and patients treated with MMI have decreased levels of circulating VCAM-1 (Wenisch et al., 1995).

Our group has extensively studied a phenyl derivative of MMI termed C10 (1a in this manuscript; 4-phenyl-3-methyl-1, 3 imidazole-2-thione). Adding the aromatic ring (phenyl group) to MMI potentially enhances cell membrane transit and thus increases activity (Seydel, 2002). Previous studies have revealed that 1a, compared to MMI, exhibits a significantly greater suppression of abnormal MHC gene expression in vitro (Giuliani et al., 2010). We have found that 1a is effective in various in vitro and in vivo models of disease or disease processes [e.g. (Benavides et al., 2010; McCall et al., 2010; Schwartz et al., 2009)]. Most germane to the present study, we previously reported that 1a can reduce TNF-α-induced VCAM-1 expression in human endothelial cells (Dagia et al., 2004). In the previous study, the effect of MMI on VCAM-1 expression was not determined. Thus, the first goal of the present study was to investigate the hypothesis that 1a is more active than MMI in inhibiting TNF-α-induced VCAM-1 expression by human endothelial cells.

Our second goal was to determine if small modifications to 1a could significantly enhance the ability to inhibit TNF-α-induced VCAM-1 expression. We focused on two modifications. First, we wished to see if a simple modification to the phenyl ring of 1a might enhance or decrease the activity of this class of compounds. We chose to use an electron donating group, methoxy, as an initial substitution. The strongly electron donating properties of this group should have some effect on the activity if the phenyl ring is involved in significant non-covalent interactions with the biological target. Second, the heterocyclic ring of 1a is an imidazole. Thiazoles (an imidazole with an N substituted for an S) are important antiinflammatory and immune-suppressive agents (Emami and Foroumadi, 2006; Nishikaku et al., 1994; Pattan et al., 2009; Siddiqui et al., 2009). In addition, previous studies have suggested that thiazoles may exhibit higher efficacy than imidazoles in rat models of inflammation with edema (Unangst et al., 1994). For these reasons, we probed the hypothesis that thiazole phenyl-MMI analogs are more active than imidazole phenyl-MMI analogs in reducing TNF-α-induced VCAM-1 expression.

2. Materials and Methods

2.1 Materials

The reagents for human umbilical vein endothelial cell culture and ELISA were described previously (Dagia and Goetz, 2003). 16% para-formaldehyde was obtained from Electron Microscopy Sciences (Hatfield, PA). The assay buffer was Hanks buffered saline solution with Ca2+ and Mg2+ (HBSS+) supplemented with 5% FBS. MMI was obtained from Sigma (St. Louis, MO). 1a and 1b were prepared according to reported procedures (Kohn et al., 2011; Theoclitou et al., 2002). Compounds 2a and 2b were prepared by the method reported by Gan et al. (Gan et al., 2010). They were prepared as 200 mM stock solution in 100% DMSO. Recombinant human TNF-α was obtained from R&D Systems (Minneapolis, MN). MTS was purchased from Promega (Madison, WI). The buffer for the adhesion assay was Dulbecco’s phosphate buffered solution (DPBS) with Ca2+ and Mg2+ (Life Technologies, Carlsbad, CA) supplemented with 0.1% BSA (Sigma).

2.2 Antibodies

Anti-CD106 (Anti-VCAM-1, clone: 1.G11B1) was purchased from Ancell (Bayport, MN). Human IgG was purchased from Sigma-Aldrich (St. Louis, MO). Anti-mouse IgG (heavy and light chain specific F (ab)’2 fragment, peroxidase conjugate) was purchased from Calbiochem (La Jolla, CA).

2.3 Cell Culture and treatment of HUVEC

HUVEC (Lonza, Walkersville, MD) and U937 monocytic cells (American Type Culture Collection, Manassas, VA) were cultured as previously described (Dagia et al., 2004; Dagia and Goetz, 2003). HUVEC were grown to 95–100% confluence prior to use in an assay. The cells were activated with TNF-α at a 10 ng/ml concentration, and treated with compound in the carrier (0.1% DMSO for ELISA, adhesion assays and real time PCR and 0.25% DMSO for MTS) for a 24 h period unless otherwise noted.

2.4 ELISA

The activated and treated HUVEC were washed, fixed in 1% paraformaldehyde for 20 min at 4°C, washed, and incubated with 5% FBS in HBSS+. Following that, the HUVEC were washed and incubated with primary antibody (e.g., anti-VCAM-1) at 10 μg/ml for 20 min at 4°C. The HUVEC were then washed and incubated with peroxidase conjugated anti-mouse IgG at 1:100 dilution for 20 min at 4°C. Subsequently, the HUVEC were subjected to 5–6 washes with the HBSS+. Thereupon, HUVEC were treated with OPD dissolved in phosphate citrate buffer containing sodium perborate. After 10 and 20 min incubations in the dark, the Optical Density (OD) values were read at 450 nm on a Multiskan MCC Spectrophotometer (Fisher Scientific, Dubuque, IA).

2.5 Cell viability (MTS) assay

20 μl of MTS/PMS solution was added to the wells containing the HUVEC as per the supplier’s protocol. Subsequently, the plate was incubated at 37°C for 1 h in a humidified 5% CO2 atmosphere. OD was read at 490 nm on the Multiscan MCC Spectrophotometer.

2.6 Parallel plate flow chamber assay

Flow adhesion assays were conducted using a parallel plate flow chamber (Glycotech, Rockville, MD) mounted on a Nikon TE 300 inverted microscope equipped with a CCD camera. HUVEC were plated in sterile 6.5 mm FlexiPerm gaskets (Sarstdet, Numbrecht, Germany) on 35 mm tissue culture dishes (Corning, Corning, NY). After reaching confluence, the HUVEC were treated for 24 hours with TNF-α (10 ng/ml) and compounds 1a, 1b, 2a, 2b (70 μM) or DMSO (carrier control) at 37°C and 5% CO2. The HUVEC were loaded into the flow chamber, the chamber placed on the microscope stage and a suspension of U937 cells (5 × 105 cells/ml) perfused over the HUVEC for 2 minutes at a wall shear stress of 1.8 dynes/cm2 using a syringe pump (Harvard Apparatus, Cambridge, MA). The interaction of the U937 cells with HUVEC was recorded for offline analysis as previously described (Dagia and Goetz, 2003). The number of adhering cells included all cells attaching from the free fluid stream.

2.7 Real-Time PCR

Total RNA was isolated from HUVEC (RNeasy Kit, Qiagen, Valencia, CA, USA), treated with DNase (RNase-Free DNAse Kit, Qiagen), and quantified with NanoDrop 2000c Spectrophotometer (Thermo Scientific, Waltham, MA). cDNA was synthesized using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA). Real-time PCR was performed using human VCAM-1 (Hs00365485_m1) and Actin, beta (VIC, Hs99999903_m1) Taqman Gene Expression Assays (Applied Biosystems) and amplified using the StepOnePlus Real-Time PCR system (Applied Biosystems). Fold changes in gene expression were calculated using the ΔΔCt method (Livak and Schmittgen, 2001).

2.8 Statistics

One-way analysis of variance (ANOVA) was performed to assess the statistical significance of the differences observed in the ELISA and the cell viability assays. If compounds exhibited significant differences, Tukey’s test was used for multiple pair wise comparisons.

3. Results

3.1 Modifications to MMI yield 1a and related compounds used in the present study

The parent structure for this work is the compound Methimazole [3-methyl-1, 3 imidazole-2 thione (MMI)] which is used clinically for the treatment of autoimmune disease (Cooper, 1984) (Fig. 1). As previously reported (Kohn et al., 2011) addition of a phenyl ring onto the 4-position of the imidazole ring of MMI yields phenyl methimazole [4-phenyl-3-methyl-1, 3 imidazole-2-thione] or C10 (compound 1a in this report; Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Compounds used in the present study. Methimazole (MMI) is used clinically for the treatment of autoimmune disease and is chemically defined as 3-methyl-1,3 imidazole-2 thione (The number 1 next to the nitrogen indicates the 1-nitrogen). Addition of a phenyl group to the 4th position of the imidazole ring of MMI yields phenyl methimazole (C10 – termed 1a in this paper). Addition of a methoxy group, at the para position of the phenyl ring of 1a, yields 1b. Substitution of the imidazole ring for a thiazole ring in 1a yields 2a. Addition of a methoxy group, at the para position of the phenyl ring of 2a, yields 2b. Note that 2b is the thiazole analog of 1b.

For this project we wished to examine two specific alterations to the core structure of 1a. The first was a simple substitution of the aromatic ring. The unsubstituted benzene ring of 1a was replaced with a 4-methoxyphenyl to provide 1b. The goal was simply to identify portions of the molecule that might be useful to modify in a larger structure activity relationship study. A second alteration of the core structure of 1a was the change of the unsubstituted N–H of the imidazole-2-thione ring to an S providing a thiazole-2-thione ring. Such a substitution might be expected to alter interactions with the pharmacological target if interactions of the somewhat acidic N–H are important.

3.2 Addition of a phenyl ring to methimazole significantly increases its ability to inhibit TNF-α-induced VCAM-1 expression by human endothelial cells

Dagia et al. have shown that 1a significantly reduces TNF-α-induced VCAM-1 expression by human endothelial cells while having limited effect on other ECAMs (Dagia et al., 2004). Interestingly, a comparison between 1a and MMI was not made in this previous work. Thus, as a first step in the present study, we compared, at a range of concentrations, the effect of 1a to that of MMI in terms of the ability to inhibit TNF-α-induced VCAM-1 expression. We have previously shown that TNF-α induces VCAM-1 mRNA and protein expression in HUVEC (Dagia and Goetz, 2003; Deosarkar et al., 2011). We chose to initially focus this study on TNF-α induced VCAM-1 protein expression (as opposed to VCAM-1 mRNA expression) since VCAM-1 protein expression is more functionally relevant than VCAM-1 mRNA. Previous work (Bevilacqua, 1993) has established that VCAM-1 expression reaches its peak at ~24 h post-TNF-α activation and our preliminary data demonstrated that treatment of HUVEC with 10 ng/ml TNF-α gave maximal levels of VCAM-1 protein expression (i.e. the level of VCAM-1 protein expression resulting from 10 ng/ml TNF-α was similar to the level of VCAM-1 protein expression that resulted from treatment with 5, 25 and 50 ng/ml of TNF-α). Thus, we chose to treat the HUVEC for 24 hours with 10 ng/ml for this study.

As shown in Figure 2A and 2B an ELISA reveals that 1a significantly reduced TNF-α-induced VCAM-1 protein expression on HUVEC at concentrations where MMI had little, if any, effect. For example at 50 μM, 1a inhibited over 40% of the VCAM-1 protein expression elicited by TNF-α while MMI had no inhibitory effect on VCAM-1 expression (Figures 2A and 2B). We determined the IC50, the concentration of compound needed to inhibit 50% of the VCAM-1 expression, for both 1a and MMI using linear regression. This was done for multiple experiments and the resulting IC50 values averaged to give the results presented in Figure 2C. As shown in Figure 2C, the IC50 for MMI (1135 μM) was over 17 times higher than that of 1a (66 μM). Thus, the addition of the phenyl ring onto MMI appears to significantly improve its potency in terms of inhibiting TNF-α-induced VCAM-1 protein expression.

Fig. 2.

Fig. 2

Open in a new tab

1a is significantly more potent than MMI in terms of the ability to inhibit 24 hr. TNF-α induction of VCAM-1 protein expression. (A and B) Confluent HUVEC monolayers were treated with TNF-α in the presence of varying concentrations of MMI or 1a. After 24 hours, the level of VCAM-1 protein present on the HUVEC was determined via ELISA as described in the methods. A ratio between the determined values and the value obtained when HUVEC were treated with carrier alone (DMSO) and TNF-α was multiplied by 100% to give the % activity. The % activity is plotted versus the concentration of MMI (A) or 1a (B). (C) The IC50 for MMI and 1a was determined from at least three separate experiments for each compound. The results for each compound were averaged to arrive at the IC50 for that compound in terms of the ability to inhibit 24 h TNF-α induction of VCAM-1 protein expression. Error bar is S.E.M.; * P < 0.05.

3.3 Addition of a para methoxy to the appended phenyl ring increases inhibition of 24 h TNF-α-induced VCAM-1 expression by human endothelial cells

We next sought to determine if some relatively simple modifications to 1a would further enhance the potency of the compound. Specifically we made a methoxy derivative of 1a (termed 1b), which is described in section 3.1 and depicted in Figure 1, and a thiazole derivative of 1a (2a; Fig. 1), and a thiazole derivative of 1b (2b; Fig. 1). For each compound, we used ELISA and performed dose response experiments to determine the effect of the compounds on 24 h TNF-α-induced VCAM-1 protein expression. The results are presented in Figure 3. This was done multiple times and the resulting IC50 values, determined from each replicate, averaged to give the IC50 value for each compound (Figure 4). The IC50 for 1a, 1b, 2a and 2b were 66 μM, 25 μM, 57 μM, and 36 μM, respectively (Figure 4). Each of the 1a derivatives has activity that is more similar to 1a than MMI, i.e. they are all at least 19 times more potent than MMI which had an IC50 of 1135 μM. Figure 4 reveals that substituting the imidazole ring for a thiazole ring does not have a significant effect on activity (compare 1a to 2a and 1b to 2b). However, addition of a para methoxy group to the appended phenyl does significantly increase activity (compare 1a to 1b and 2a to 2b).

Fig. 3.

Fig. 3

Open in a new tab

Modification of 1b affords additional compounds that inhibit 24 h TNF-α-induced VCAM-1 protein expression. Confluent HUVEC monolayers were treated with TNF-α in the presence of varying concentrations of (A) 1b, (B) 2a, or (C) 2b. After 24 hours, the level of VCAM-1 protein present on the HUVEC was determined via ELISA as described in the methods. A ratio between the determined values and the value obtained when HUVEC were treated with carrier alone (DMSO) and TNF-α was multiplied by 100% to give the % activity. The % activity is plotted versus the concentration of compound used in the treatment. Error bar is S.E.M.

Fig. 4.

Fig. 4

Open in a new tab

A graph comparing the average IC50 obtained for the various compounds. The IC50 for the compounds was determined from at least three separate experiments for each compound. The results for each compound were averaged to arrive at the IC50 for that compound in terms of the ability to inhibit 24 h TNF-α induction of VCAM-1 protein expression. There was no significant difference between the imidazole and thiazole paired compounds, specifically there was no difference between 1a and 2a nor between 1b to 2b. Addition of a para methoxy group to the appended phenyl does significantly increase activity; specifically 1b activity is significantly greater than 1a’s activity and 2b’s activity is significantly greater than 2a’s activity. The values presented are mean ± S.E.M. of at least three independent experiments. * P < 0.05.

During the final steps of the ELISA assay [after the addition of the secondary (detection) antibody] we observed that some HUVEC were washed away around the perimeters of the wells treated with 1b and 2b at 100 μM. Hence, for 1b and 2b, only data up to 50 μM were used to obtain the IC50. This observation led us to investigate the effect of the compounds on cell viability.

3.4 As measured by a MTS assay, the phenyl derivatives are more toxic than MMI and the thiazole derivatives appear to be more toxic than the imidazole derivatives

To gain insight into the effect of the compounds on cell viability, we performed MTS assays. The MTS assay measures metabolic activity which often correlates with the number of viable cells. For each compound, we performed an MTS assay for a range of concentrations (Figure 5). This allowed us to estimate the concentration of compound needed to reduce the MTS signal by 50% of the control value, i.e. the TC50. These experiments were done multiple times and the resulting TC50 values, determined from each replicate using linear regression, averaged to give the TC50 value for each compound (Figure 6). The TC50 for MMI, 1a, 1b, 2a and 2b was 2450 μM, 1240 μM, 1010 μM, 470 μM, and 520 μM respectively (Figure 6). Figure 6 reveals that MMI is significantly less toxic than any of the phenyl derivatives used in this study and that the thiazole derivatives are more toxic than their imidazole counterparts (2a vs 1a and 2b vs 1b).

Fig. 5.

Fig. 5

Open in a new tab

The effect of the compounds on MTS signal. Confluent HUVEC monolayers were treated with varying concentrations of (A) MMI, (B) 1a, (C) 1b, (D) 2a or (E) 2b. After 24 hours, an MTS assay was performed. A ratio between the determined values and the value obtained when HUVEC were treated with carrier alone (DMSO) was multiplied by 100% to give the % MTS signal. The % MTS signal is plotted versus the concentration of compound used in the assay. Error bar is S.E.M.

Fig. 6.

Fig. 6

Open in a new tab

TC50 for the compounds as assessed by an MTS assay. The concentration of compound that resulted in 50% decrease (TC50) in the MTS HUVEC signal is shown. MMI is significantly less toxic than any of the phenyl derivatives (* P < 0.05) and the thiazole derivatives are more toxic than the imidazole counterparts (2a vs. 1a, # P < 0.05; 2b vs. 1b, % P < 0.05). The values presented are mean ± S.E.M. of three independent experiments.

Several noteworthy observations were made during these assays. While performing the MTS assay for 2a, it was observed that at the higher concentrations (1000 μM and 500 μM) the “solution” contained “fuzzy structures” that were occupying a significant portion of the wells. At a concentration of 1000 μM of 2b or 2a small round bodies that appeared to be disintegrated cells/precipitates were seen throughout the wells. Therefore, only the MTS data for concentrations up to 250 μM for 2a and up to 500 μM for 2b were used to determine their TC50.

3.5 Toxic to inhibitory index of imidazole and thiazole derivatives of MMI

To gain insight into the effectiveness of the compounds relative to their potential toxic effects, we determined the ratio of TC50 to IC50. We term this ratio the “toxic to inhibitory index (TII)”. The results are presented in Figure 7 where it is shown that the TII for MMI, 1a, 1b, 2a, and 2b are 2, 19, 40, 8, and 14 respectively. Figure 7 reveals that all the phenyl derivatives of MMI appeared to exhibit a greater TII than MMI and that 1b appears to have the maximal TII.

Fig. 7.

Fig. 7

Open in a new tab

Comparison of the compounds’ toxic to inhibitory indices (TII). TII is calculated by taking the ratio of the TC50 to IC50 for each compound. The phenyl derivatives of MMI exhibited a greater TII than MMI. The values presented above are ratio ± S.E.M. as determined by propagation of error analysis.

3.6 MMI and each of the derivatives inhibit monocytic cell adhesion to 24 h TNF-α-activated human endothelial cells under fluid shear conditions

We sought to determine if each compound would have an inhibitory effect in a functional assay. During pathological inflammation, VCAM-1 helps mediate the adhesion of monocytes to the endothelium in the fluid dynamic environment of the circulation (Luscinskas and Gimbrone, 1996). Thus, we tested the effect of the compounds on monocytic (U937) cell adhesion to 24 h TNF-α-activated HUVEC in an in vitro flow adhesion assay that mimics conditions present in vivo (Gopalan et al., 1996). As shown in Figure 8, unactivated HUVEC supported very little U937 adhesion. In contrast, treatment of HUVEC with TNF-α for 24 hours dramatically increased U937 adhesion. Incubation of the HUVEC with a function blocking mAb to VCAM-1, prior to the adhesion assay but after the TNF-α 24 h treatment, revealed that the majority, but not all, of the adhesion appeared to be mediated by VCAM-1. Each of the compounds caused a significant reduction in adhesion when added to the HUVEC during the TNF-α treatment (Figure 8) and thus each demonstrates a significant inhibitory effect on function.

Fig. 8.

Fig. 8

Open in a new tab

The effect of the compounds on monocytic (U937) cell adhesion to HUVEC in an in vitro adhesion assay that mimics fluid dynamic conditions present in vivo. U937 monocytic cells were perfused over unactivated HUVEC or 24 h TNF-α-activated HUVEC in an in vitro flow chamber assay. To gain insight into the role of VCAM-1 in the adhesion, the HUVEC were pre-treated with a function blocking mAb to VCAM-1 immediately prior to the adhesion assay. In certain cases, the TNF-α activation was carried out in the presence of one of the compounds or DMSO (carrier control). Each compound caused a significant reduction in U937 adhesion compared to treatment with DMSO alone (* P < 0.05) – bracket indicates comparison of each bar under bracket to DMSO bar. Wall shear stress = 1.8 dynes/cm2. The values presented are mean ± S.E.M. of n ≥ 3. All compounds were used at 70 μM and the mAb concentration was 20 μg/ml.

3.7 MMI and each of the derivatives significantly reduce 24 h TNF-α-induced VCAM-1 mRNA expression by human endothelial cells

We have previously shown, using Northern blot analysis, that compound 1a inhibits 24 h TNF-α-induced VCAM-1 mRNA expression in human aortic endothelial cells (Dagia et al., 2004). To expand on this previous finding and to begin to explore the mechanism by which these compounds inhibit induced VCAM-1 protein expression, we used real time PCR to determine the compounds’ effects on induced VCAM-1 mRNA expression. HUVEC were treated for 24 h with TNF-α in the presence of one of the compounds or DMSO (carrier control). Subsequently, VCAM-1 transcripts were analyzed by PCR. As shown in Figure 9, each of the compounds significantly reduced 24 h TNF-α-induced VCAM-1 mRNA expression. This result, combined with our previous studies with 1a (Dagia et al., 2004), suggests that the compounds diminish TNF-α-induced VCAM-1 protein expression by inhibiting VCAM-1 mRNA expression. [Note that in these assays the MMI concentration (1200 μm) was significantly greater than the concentration of the other compounds (70 μm).]

Fig. 9.

Fig. 9

Open in a new tab

Each compound inhibits 24 h TNF-α-induced VCAM-1 mRNA expression by HUVEC. HUVEC were treated with TNF-α for 24 hours in the presence of the various compounds or carrier control (DMSO). Subsequently, RNA was collected, cDNA generated, and real-time PCR was performed. The % inhibition was determined relative to the carrier control using the ΔΔCt method. Each compound caused a significant reduction in VCAM-1 mRNA transcripts compared to treatment with DMSO alone (* P < 0.05) – bracket indicates comparison of each bar under bracket to DMSO bar. All compounds were used at 70 μM except MMI, which was used at 1,200 μM. The values presented are an average of n = 3 experiments.

4. Discussion

Our studies reveal that addition of a phenyl group to MMI dramatically increases the ability to inhibit TNF-α-induced VCAM-1 protein expression by human endothelial cells. Indeed, all four of the compounds tested, 1a, 1b, 2a and 2b had an IC50 value that was over an order of magnitude lower than the IC50 achieved for MMI (Figure 4 compared to Figure 2C). To our knowledge the ability of 1a and MMI to inhibit TNF-α-induced VCAM-1 expression has not previously been compared. However, others have observed the greater potency of 1a, relative to MMI. For example, 1a was reported to exhibit a 70–100 fold greater, compared to MMI, suppression of abnormal MHC gene expression in thyroid epithelial cells (Giuliani et al., 2010).

It is important to note that in parallel with the finding of increased activity, the toxic effects of the compounds on human endothelial cells appeared to also increase with the addition of the phenyl ring. Specifically, the TC50 value for MMI was two to five times higher than that observed for the phenyl appended compounds (Figure 6). Hence, it would appear that addition of the phenyl group augments the desired activity but also potentially increases unwanted toxic side effects. That said, it is true that all drugs have a desired, therapeutic, effect as well as a toxic effect(s) at some level. Insight into the potential usefulness of a compound can be gained by comparing the TC50 values to the IC50 values – a ratio we termed the toxic to inhibitory index (TII) (Figure 7). The phenyl appended compounds all have a higher TII (anywhere from 4 to 20 times higher) than MMI demonstrating that, while addition of the phenyl group appears to increase toxicity relative to MMI (Figure 6), the activity increases to a greater extent (Figure 4) suggesting that addition of the phenyl group does move the compound toward a structure that will have lower adverse side effects for the same therapeutic effect (Figure 7).

Interestingly, the data demonstrated that addition of a methoxy group to 1a or 2a significantly increased the activity by ~2.6 and ~1.5 fold (Figure 4) as determined by calculating a ratio of the IC50 for 1a vs 1b and 2a vs 2b, respectively. At the same time, there was not a significant difference in the toxicity with the addition of the methoxy group (Figure 6). Not surprisingly given these results, the TII data revealed that addition of the methoxy group appears to increase the TII by ~2 fold (compare TIIs of 1b and 1a as well as 2a and 2a in Figure 7). Hence, addition of the methoxy group to the appended phenyl would appear to be a step towards a more optimal compound and suggests that the phenyl ring may be involved in non-covalent interactions with the biological target. That said, it is important to note the observation we mentioned in the results section. Specifically, we observed that for both 1b and 2b at higher concentrations (e.g. 100 μM) some of the HUVEC were observed to detach from the tissue culture dishes during the washing steps. This was not observed with 1a or 2a. Loss of endothelial cells in vivo would obviously be of significant concern since a denuded endothelial cell layer can cause a plethora of problems some of which can be quite severe (Butcher and Nerem, 2007). Thus, while addition of the methoxy group did appear to be advantageous based on the TII results, the fact that 1b and 2b appeared to induce detachment of HUVEC, albeit at higher concentrations, is of concern when considering these compounds.

Many thiazole-based compounds are approved drugs (e.g. meloxicam) (Martin et al., 2012) or have been shown to be anti-inflammatory immune-suppressive agents (Nishikaku et al., 1994; Pattan et al., 2009). In addition, previous studies have demonstrated that thiazole-based compounds may exhibit higher efficacy than matched imidazole-based compounds in rat models of inflammation with edema (Unangst et al., 1994). These observations motivated us to make thiazole analogs of 1a and 1b. Interestingly, the thiaozle derivatives appeared to be no more potent than the imidazole derivatives (Figure 4) but did appear to be more toxic (Figure 6). It is unclear at this time why the thiazoles appear to be more toxic than the imidazoles and additional investigation is certainly warranted. Such future studies could include making additional imidazole/thiazole pairs and comparing them in the MTS assay.

To determine if the ability to reduce VCAM-1 protein expression translates into an inhibitory effect on function, we tested each compound in an in vitro dynamic cell adhesion assay that mimics flow conditions present in vivo. As shown in Figure 8, each of the compounds significantly inhibited U937 (a monocytic cell line) adhesion to 24 h TNF-α-activated HUVEC. Somewhat surprisingly, MMI had a significant inhibitory effect on adhesion (Figure 8) despite the fact that, at the concentration of MMI used in the adhesion assay (70 μm), there is little, if any, inhibition of VCAM-1 protein expression (Figure 2A). The finding with MMI underscores the fact that the U937 cells adhere to the HUVEC via VCAM-1-dependent as well as VCAM-1-independent mechanisms (Dagia et al., 2004). It is well established that TNF-α increases the expression of ICAM-1 and induces the expression of E-selectin both of which are functional endothelial cell adhesion molecules (Bevilacqua, 1993). Treatment of the TNF-α-activated HUVEC with a mAb to VCAM-1 caused a significant decrease in adhesion, but clearly did not eliminate all of the adhesion (Figure 8), strongly suggesting the presence of a VCAM-1-independent mechanism(s) for adhesion. Thus, the inhibition on adhesion observed with MMI (Figure 8) could be due to inhibition of a VCAM-1-independent adhesion mechanism.

To begin to explore the mechanism by which the compounds inhibit induced VCAM-1 protein expression, we probed VCAM-1 expression at the mRNA level using real time PCR. This analysis revealed that all of the compounds inhibit 24 h TNF-α-induced VCAM-1 mRNA expression (Figure 9). (Note that MMI was used at 1,200 μm and the other compounds were used at 70 μm in this assay.) This finding is in line with our previous result with 1a where we observed, via Northern analysis, that 1a inhibited 24 h TNF-α-induced VCAM-1 expression by human aortic endothelial cells (Dagia et al., 2004). In addition, we have previously reported that 1a acts to inhibit signal transduction (Courreges et al., 2012; Dagia et al., 2004; McCall et al., 2010; Schwartz et al., 2009) and our finding that the compounds inhibit TNF-α-induced VCAM-1 mRNA expression is consistent with the hypothesis that the compounds inhibit VCAM-1 via blockade of a signal transduction mechanism(s).

In conclusion, we have found that addition of a phenyl ring at the 4-position of MMI yields a compound that is significantly more potent than MMI at inhibiting 24 h TNF-α-induced VCAM-1 protein expression by HUVEC. Addition of a para methoxy to the appended phenyl group increases the inhibition, while substitution of a thiazole ring for an imidazole ring in the phenyl derivatives yields no clear difference in inhibition. Addition of the phenyl ring to MMI increases toxicity as does substitution of a thizole ring for an imidazole ring in the phenyl MMI derivatives. Each of the compounds reduced TNF-α-induced VCAM-1 mRNA expression and had a functional inhibitory effect, i.e., each inhibited monocytic cell adhesion to 24 h TNF-α-activated HUVEC under fluid flow conditions. Combined, these results provide important insights into the design of MMI-related anti-inflammatory compounds.

Acknowledgments

This work was supported by The National Institutes of Health [00550937 (DJG, KM)] and The National Science Foundation [1039869 (DJG, MMB)].

Footnotes

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.

References

  1. Benavides U, Gonzalez-Murguiondo M, Harii N, Lewis CJ, Sakhalkar HS, Deosarkar SP, Kurjiaka DT, Dagia NM, Goetz DJ, Kohn LD. Phenyl methimazole suppresses dextran sulfate sodium-induced murine colitis. Eur J Pharmacol. 2010;643:129–138. doi: 10.1016/j.ejphar.2010.06.003. [DOI] [PubMed] [Google Scholar]
  2. Bevilacqua MP. Endothelial-leukocyte adhesion molecules. Annu Rev Immunol. 1993;11:767–804. doi: 10.1146/annurev.iy.11.040193.004003. [DOI] [PubMed] [Google Scholar]
  3. Butcher JT, Nerem RM. Valvular endothelial cells and the mechanoregulation of valvular pathology. Philos Trans R Soc B-Biol Sci. 2007;362:1445–1457. doi: 10.1098/rstb.2007.2127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Carter RA, Campbell IK, O’Donnel KL, Wicks IP. Vascular cell adhesion molecule-1 (VCAM-1) blockade in collagen-induced arthritis reduces joint involvement and alters B cell trafficking. Clin Exp Immunol. 2002;128:44–51. doi: 10.1046/j.1365-2249.2002.01794.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cooper DS. Antithyroid drugs. N Engl J Med. 1984;311:1353–62. doi: 10.1056/NEJM198411223112106. [DOI] [PubMed] [Google Scholar]
  6. Courreges MC, Kantake N, Goetz DJ, Schwartz FL, McCall KD. Phenylmethimazole blocks dsRNA-induced IRF3 nuclear translocation and homodimerization. Mol Basel Switz. 2012;17:12365–12377. doi: 10.3390/molecules171012365. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cybulsky MI, Iiyama K, Li H, Zhu S, Chen M, Iiyama M, Davis V, Gutierrez-Ramos JC, Connelly PW, Milstone DS. A major role for VCAM-1, but not ICAM-1, in early atherosclerosis. J Clin Invest. 2001;107:1255–62. doi: 10.1172/JCI11871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dagia NM, Goetz DJ. A proteasome inhibitor reduces concurrent, sequential and long term IL-1b and TNF-a induced endothelial cell adhesion molecule expression and adhesion. Am J Phys. 2003;285:C813–C822. doi: 10.1152/ajpcell.00102.2003. [DOI] [PubMed] [Google Scholar]
  9. Dagia NM, Harii N, Meli AE, Sun X, Lewis CJ, Kohn LD, Goetz DJ. Phenyl methimazole inhibits TNF-a induced VCAM-1 expression in an IFN regulatory factor-1-dependent manner and reduces monocytic cell adhesion to endothelial cells. J Immunol. 2004;173:2041–2049. doi: 10.4049/jimmunol.173.3.2041. [DOI] [PubMed] [Google Scholar]
  10. Deosarkar SP, Bhatt P, Lewis CJ, Goetz DJ. A quantitative real-time PCR approach for the detection and characterization of endothelial cells in whole blood. Ann Biomed Eng. 2011;39:2627–2636. doi: 10.1007/s10439-011-0354-x. [DOI] [PubMed] [Google Scholar]
  11. Emami S, Foroumadi A. Synthesis of 4-(4-methylsulfonylphenyl)-3-phenyl-2(3H)-thiazole thione derivatives as new potential COX-2 inhibitors. Chin J Chem. 2006;24:791–794. [Google Scholar]
  12. Gan SF, Wan JP, Pan YJ, Sun CR. Water-Mediated Multicomponent Reaction: A Facile and Efficient Synthesis of Multisubstituted Thiazolidine-2-thiones. Synlett. 2010:973–975. [Google Scholar]
  13. Giuliani C, Bucci I, Montani V, Singer DS, Monaco F, Kohn LD, Napolitano G. Regulation of major histocompatibility complex gene expression in thyroid epithelial cells by methimazole and phenylmethimazole. J Endocrinol. 2010;204:57–66. doi: 10.1677/JOE-09-0172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gopalan PK, Jones DA, McIntire LV, Smith CW. Cell adhesion under hydrodynamic flow conditions. In: Coligan JE, Kruisbeek AM, Margulies DH, Shevach EM, Strober W, editors. Current Protocols in Immunology. J. Wiley; New York: 1996. pp. 7.29.1–7.29.23. [DOI] [PubMed] [Google Scholar]
  15. Kohn LD, Goetz DJ, Benavides-Peralta U, Gonzalez-Murguiondo M, Harii N, Lewis CJ, Napolitano G, Dagia ND. Methods for the amelioration of episodes of acute or chronic ulcerative colitis. US Patent. 2011;7:928,132. [Google Scholar]
  16. Livak KJ, Schmittgen TD. Analysis of Relative Gene Expression Data Using Real-Time Quantitative PCR and the 2–ΔΔCT Method. Methods. 2001;25:402–408. doi: 10.1006/meth.2001.1262. [DOI] [PubMed] [Google Scholar]
  17. Luscinskas FW, Gimbrone MA. Endothelial-dependent mechanisms in chronic inflammatory leukocyte recruitment. Annu Rev Med. 1996;47:413–421. doi: 10.1146/annurev.med.47.1.413. [DOI] [PubMed] [Google Scholar]
  18. Martin AI, Nieto-Bona MP, Castillero E, Fernandez-Galaz C, Lopez-Menduina M, Gomez-Sanmiguel AB, Gomez-Moreira C, Villanua MA, Lopez-Calderon A. Effect of cyclooxygenase-2 inhibition by meloxicam, on atrogin-1 and myogenic regulatory factors in skeletal muscle of rats injected with endotoxin. J Physiol Pharmacol Off J Pol Physiol Soc. 2012;63:649–659. [PubMed] [Google Scholar]
  19. McCall KD, Holliday D, Dickerson E, Wallace B, Schwartz AL, Schwartz C, Lewis CJ, Kohn LD, Schwartz FL. Phenylmethimazole blocks palmitate-mediated induction of inflammatory cytokine pathways in 3T3L1 adipocytes and RAW 264.7 macrophages. J Endocrinol. 2010;207:343–353. doi: 10.1677/JOE-09-0370. [DOI] [PubMed] [Google Scholar]
  20. Nishikaku F, Aono S, Koga Y. Protective Effects of D-Penicillamine and a Thiazole Derivative, Sm-8849, on Pristane-Induced Arthritis in Mice. Int J Immunopharmacol. 1994;16:91–100. doi: 10.1016/0192-0561(94)90064-7. [DOI] [PubMed] [Google Scholar]
  21. Pattan SR, Hullolikar RL, Dighe NS, Ingalagi BN, Hole MB, Gaware VM, Chavan PA. Synthesis and Evaluation of some New Phenyl Thiazole Derivatives for their Anti-Inflammatory Activites. J Pharm Sci Res. 2009;1:96–102. [Google Scholar]
  22. Schwartz AL, Malgor R, Dickerson E, Weeraratna AT, Slominski A, Wortsman J, Harii N, Kohn AD, Moon RT, Schwartz FL, Goetz DJ, Kohn LD, McCall KD. Phenylmethimazole decreases Toll-like receptor 3 and noncanonical Wnt5a expression in pancreatic cancer and melanoma together with tumor cell growth and migration. Clin Cancer Res Off J Am Assoc Cancer Res. 2009;15:4114–4122. doi: 10.1158/1078-0432.CCR-09-0005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Seydel JK. Drug-Membrane Interactions: Analysis, Drug Distribution, Modeling. Vol. 15. Wiley-VCH Verlag GmbH & Co KGaA; 2002. Drug-Membrane Interaction and Pharmacokinetics of Drugs; pp. 141–215. [Google Scholar]
  24. Siddiqui N, Arshad MF, Ahsan W, Alam MS. Thiazoles: A valuable insight into the recent advances and biological activities. Int J Pharm Sc Iences Drug Res. 2009;1:136–143. [Google Scholar]
  25. Theoclitou ME, Delaet NGJ, Robinson LA. Rapid parallel synthesis of combinatorial libraries of substituted 3-thio-1,2,4-triazoles and 2-thioimidazoles. J Comb Chem. 2002;4:315–319. doi: 10.1021/cc010094o. [DOI] [PubMed] [Google Scholar]
  26. Unangst PC, Connor DT, Cetenko WA, Sorenson RJ, Kostlan CR, Sircar JC, Wright CD, Schrier DJ, Dyer RD. Synthesis and Biological Evaluation of 5-[[3,5-Bis(1,1-Dimethylethyl)-4-Hydroxyphenyl]methylene]oxazoles, -Thiazoles, and -Imidazoles – Novel Dual 5-Lipoxygenase and Cyclooxygenase Inhibitors with Antiinflammatory Activity. J Med Chem. 1994;37:322–328. doi: 10.1021/jm00028a017. [DOI] [PubMed] [Google Scholar]
  27. Wenisch C, Myskiw D, Gessl A, Graninger W. Circulating selectins, intercellular adhesion molecule-1, and vascular cell adhesion molecule-1 in hyperthyroidism. J Clin Endocrinol Metab. 1995;80:2122–6. doi: 10.1210/jcem.80.7.7541802. [DOI] [PubMed] [Google Scholar]

RESOURCES