|Home | About | Journals | Submit | Contact Us | Français|
L-type Ca2+ channel (LTCC) blockers, represented by amlodipine and verapamil, are widely used anti-hypertensive drugs that also have anti-inflammatory activities. Plasminogen (Plg) is an important mediator of macrophage recruitment, and this role depends upon its interaction with Plg receptors (Plg-Rs). Plg-Rs include histone 2B (H2B), α-enolase, annexin 2 and p11, all proteins which lack signal sequences for cell-surface export. When human or murine monocytoid cells were induced to differentiate into macrophages, their Plg binding and Plg-Rs expression increased by 4-fold. These changes were suppressed by pretreatment with verapamil and amlodipine. Expression of the Cav1.2 LTCC pore subunit was induced in differentiated macrophages, and siRNA against this subunit suppressed the upregulation of Plg binding and Plg-Rs. In vivo, amlodipine and verapamil suppressed peritoneal macrophage recruitment in response to thioglycollate by >60% at doses that did not affect blood pressure. In drug-treated animals, macrophages migrated into but not through the peritoneal membrane tissue and showed reduced surface expression of Plg-Rs. These findings demonstrate that Plg-Rs expression on macrophages is dependent on Cav1.2 LTCC subunit expression. Suppression of Plg-Rs may contribute to the anti-inflammatory effects of the widely used LTCC blockers.
Involvement of plasminogen (Plg) and its active enzyme plasmin (Plm) in macrophage recruitment has been documented in a number of inflammatory models in Plg-/- mice 1-4. When bound to cell surfaces, Plg/Plm facilitates macrophage migration across adhesive substrates by direct degradation of ECM proteins and by activating MMPs. Plg-Rs are abundant on the surface of macrophages and are heterogeneous but are usually characterized by the presence of a C-terminal lysine, which interacts with the lysine binding sites (LBS) of Plg 5, 6. The major Plg-Rs expressed on macrophages are α-enolase, H2B, annexin 2 and p11 7-10. We and other have demonstrated that inflammation induces increased Plg binding to mouse macrophages and differentiated human macrophages 11-16. The mechanisms that promote upregulation of Plg-Rs during monocyte maturation are poorly understood because of none of the major Plg-Rs has a signal sequence for export through the conventional ER/Golgi pathway.
In excitable cells, L-type Ca2+ channels (LTCC) are voltage dependent Ca2+ channels composed of a pore forming α1 subunit, (Cav1.1, Cav1.2, Cav1.3 or Cav1.4) and several associated auxiliary subunits (α2-δ, β, γ). Cav1.2 pore subunits have also been detected in cells of leukocyte lineage 17-19. Besides lowering blood pressure, LTCC blockers have anti-inflammatory effects. Amlodipine, a long acting dihydropyridine, limits the progression of arteriosclerosis and decreases cardiovascular events 20, 21. Verapamil, a phenylalkylamine, also has anti-inflammatory properties 22. In search for a mechanism regulating exteriorization of Plg-Rs during monocyte to macrophage differentiation, we now report a profound effect of amlodipine and verapamil on exteriorization of H2B and other Plg-Rs in vitro. We further demonstrate that these drugs suppress macrophage recruitment and Plg-R expression on these cells in vivo.
Expanded material and methods sections are available as online data supplements.
Human monocytoid THP-1 cells were stimulated to differentiate with 250 U/ml IFNγ (eBioscience) + 100 nmol/L VD3 (Calbiochem) for 0-2 days in complete growth medium under conditions similar to those described by Kim et al12.
Rabbit polyclonal anti-α1 Cav1.2 (Alomone) was used to detect LTCC on various cell types. PE-conjugated anti-CD14 (eBioscience) was used to measure monocyte differentiation to macrophages.
Plg binding was assessed as described previously15.
Cell surface biotinylation on THP-1 cells followed by Western blot was performed as previously described15.
Cav1.2 subunit of LTCC was knocked down by siRNA designed to target Cav1.2 subunit (SiCav1.2, Dharmacon) using the nucleofection protocol described by Amaxa. SiControl RNA was a scrambled sequence and was provided by Ambion, TX
RNA isolation and RT-PCR were performed using RNeasy Mini Kits (Qiagen) and OneStep RT-PCR kit (Qiagen), respectively. The primers used to amplify Cav11.2 transcripts are previously described 18.
Fura-2 (Invitrogen) loaded THP-1 cells were stimulated with 1 μmol/L fMLP (Sigma), and Ca2+ mobilization was recorded using a fluorimeter (Photon Technology International) with dual excitation at 340 and 380 nm.
In vitro matrigel assay was performed as previously described15.
Amlodipine, verapamil or their vehicle controls were administered via subcutaneously implanted osmotic minipump, 0.25 μl/h (Alzet, Durect). Peritoneal inflammation was induced by i.p. administration of thioglycollate. Cells within the peritoneal lavage, collected three days after thioglycollate, were counted and peritoneal membrane (PM) that lines the pancreas was isolated for further study 23.
Anti-Mac3 (BD Bioscience) and biotinylated anti-rat antibody (mouse adsorbed, Vector) were used to stain sections of PM and developed with Vectastain ABC reagent (Vector) using DAB as a substrate and counterstained with hematoxylin.
Rabbit polyclonal anti-peptide antibodies against H2B, α-enolase, annexin 2, p11 or non-immune rabbit IgG, were reacted with 8 μm sections, followed by Alexa-488 labeled anti-rabbit IgG (Invitrogen).
Values are expressed as mean ± SD. P-values are based on paired student t tests. Results with p<0.05 were considered significantly different.
To investigate the molecular pathway(s) by which cell surface expression of Plg-Rs is modulated during monocyte activation, an in vitro model, THP-1 differentiation to macrophages, was implemented. When THP-1 cells were treated with IFNγ+VD3, expression of the macrophage differentiation marker, CD14, was substantially enhanced on day 1, and further increased on day 2 (Figure 1A). As measured by FACS, alexa-488 labeled Plg binding increased 4-fold at day 2 (Figure 1A). In parallel, cell surface expression of the major Plg-Rs on these cells, annexin 2, α-enolase, H2B and p11, increased by 3-4 fold (Figure 1B, first panel & C) as measured by Western blots with specific antibodies to each protein.15 CD14 expression was also increased at the cell surface and in whole cell lysates (Figure. 1B, & C). Consistent with a previous report 14, cellular annexin 2 expression increased with differentiation, both at the protein and mRNA levels (Figure 1B and 1C). However, increased cell surface expression of α-enolase, H2B and p11 occurred in the absence of detectable changes in total cellular expression at either the protein or mRNA levels, suggesting that changes in cell surface expression were independent of new protein synthesis (Figure 1B, second panel & C and Online Figure I). Similar results were demonstrated recently for α-enolase 16, using LPS to induce its cell-surface expression. Together, these observations suggest changes in protein trafficking.
THP-1 cells pretreated with Brefeldin A, which blocks transport of protein from the ER to the Golgi, reduced Plg binding by <25% (Figure 2A, p=0.02). Pathways involved in non-conventional protein secretion can include the ABC 1 transporter 24. However, glyburide, an ABC 1 transporter inhibitor, did not reduce Plg binding significantly (Figure 2A). Endolysosomal recycling (IL-1β) and the Na+/K+-ATPase (FGF-2) has also been implicated as a mediator of non-conventional protein surface expression 25, 26. However, methylamine, which blocks endosomal recycling 25 and oubain, a Na+/K+-ATPase inhibitor, had no effect on differentiation induced Plg binding to THP-1 cells (Figure. 2B). Increased [Ca2+]in is a primary cellular response associated with monocyte differentiation 27. Pretreatment of THP-1 cells with BAPTA-AM, a cell-permeant calcium chelator, blocked the IFNγ+VD3 induced increase in Plg binding by 98% (Figure 2B). Consistent with a role for [Ca2+]in in modulating Plg-R expression, ionomycin; a Ca2+ ionophore, enhanced Plg binding to the IFNγ+VD3 treated THP-1 cells by 34 ± 9.8 %. (p = 0.01).
To further investigate the role of [Ca2+]in in regulation of Plg binding, the effects of selective Ca2+ channel inhibitors were tested. Pretreatment of THP-1 cells with ω-agatoxin and ω-conotoxin; which block P-type and N-type Ca2+ channels, respectively, had no effect on Plg binding (Figure. 2B). However, two unrelated LTCC blockers, verapamil and amlodipine, effectively suppressed differentiation-induced Plg binding (Figure 2B). The concentration of these LTCC blockers (10 μmol/L) was selected based on a previous study of their effects on adhesion of THP-1 cells 17. LTCC blockers can also attenuate the synthesis of superoxide anions 28. However, pretreatment of cells with tempol (4-hydroxy-2,2,6,6-tetramethyl-piperidininoxyl), a stable membrane-permeable SOD mimetic, had only a modest effect (26%, p=0.06) on the increase in Plg binding induced by differentiation (Fig. 2B). Activation of LTCC is associated with cAMP generation 29 and IFN-α mediates α-enolase cell surface expression via MAPK/ERK1/2 activation30. Pretreatment of cells with the adenylate cyclase inhibitor, SQ22536, or ERK1/2 phosphorylation inhibitor, PD98059, had no effect on differentiation induced Plg binding (Online Figure II, C), even though PD98059 blocked differentiation induced cAMP generation and SQ22536 blocked ERK1/2 phosphorylation in the cells (Online Figure II, A&B).
Three independent approaches were taken to evaluate the presence and function of LTCC in the differentiated monocytoid cells. First, RT-PCR was performed on RNA isolated from THP-1 cells, human blood monocyte derived macrophages (HBMMΦ) and thioglycollate induced mouse peritoneal macrophages (TGMPMΦ). Using primers18 amplifying the IVS4-IVS6 domain of the Cav1.2 subunit of LTCC, an intense band of the predicted size (350 kb) was amplified in each cell type (Figure 3A). Human aortic smooth muscle cells (HASMC), used as a positive control31, also yielded a similar product. All of these cells also expressed LTCC β1 and β2 subunits (RT-PCR, not shown). Second, using FACS, an anti-Cav1.2 antibody against a Cav1.2 intracellular sequence epitope reacted well with permeabilized HASMC, THP-1, TGMPMΦ and HMDMΦ. This signal was substantially reduced by exposure to the immunizing peptide (Figure 3B). Additionally, Cav1.2 expression was enhanced upon treatment of THP-1 cells with IFNγ+VD3 for 0 to 2 days (Online Figure IIIA). Third, Ca2+ mobilization in response to fMLP increased with differentiation (Online Figure IIIB). This Ca2+ entry was blunted by amlodipine or verapamil, implicating LTCC in the [Ca2+]in response (Figure 3C). However, [Ca2+]in was unchanged when cells were treated with 50 mmol/L KCl to induce depolarization (not shown), suggesting that the LTCC present in THP-1 cells are not voltage-gated.
To eliminate the nonspecific actions of LTCC blockers, THP-1 cells were nucleofected with siRNA against Cav1.2 (SiCav1.2) or a control siRNA (Sicontrol) and then stimulated with IFNγ+VD3. The upregulation of Cav1.2 induced by differentiation was suppressed by SiCav1.2, but not by Sicontrol (Online Figure IV). Moreover, differentiation-induced Plg binding was reduced by ~90% (Figure 5A) by SiCav1.2 compared to Sicontrol. Interference with Cav1.2 expression also suppressed the [Ca2+]in response induced by fLMP in differentiated THP-1 cells (Figure 5C).
We sought to determine whether the dependence of Plg binding on LTCC and their function was due to suppression of one or more Plg-Rs. When THP-1 cells were pretreated with BAPTA-AM and then stimulated with IFNγ+VD3, BAPTA-AM prevented the cell surface upregulation of each of the four Plg-Rs (Figure 4A, first & second panels), without changing whole cell content of these proteins; i.e. trafficking of these Plg-Rs is Ca2+ dependent. In THP-1 cells pretreated with amlodipine or verapamil and then induced to differentiate, induction of each of the four Plg-Rs was reduced by the lower concentration (5 μmol/L) and was almost fully blocked by their higher concentration (10 μmol/L), with no effect on whole cell levels of these proteins (Figure 4B upper and lower panels). Similarly, SiCav1.2 but not SiControl suppressed surface expression without affecting whole cell levels of the Plg-Rs (Figure 5B).
LTCC inhibitors are known to have anti-inflammatory effects20-22, 28, and we sought to determine whether suppression of Plg-R expression might underlie this activity. To test this possibility, we turned to a widely used inflammatory model. Verapamil and amlodipine were administered to mice via subcutaneously implanted mini-pumps beginning 2 days prior to thioglycollate (TG) induced peritonitis, which involves blood monocyte differentiation into macrophages and their recruitment to the site of inflammation3, 32. Amlodipine inhibited macrophage recruitment by 54% (p=0.004) at 1 mg/kg/day, and by 87% (p = 9.22E-07) at 3 mg/kg/day compared to its vehicle control (Figure 6A). With verapamil, inhibition at the lower (1 mg/kg/day) and higher doses (3 mg/kg/day) was 82% (p = 1.24E-06) and 64% (p = 0.001), respectively. The differences in recruitment at the high and low dose of verapamil were not significant (p = 0.11). The vehicle control for each drug was not different from the recruitment observed in untreated mice. Even though these drugs are anti-hypertensive, hypotensive effects were only detected at the higher doses of both drugs (Table 1). At the lower doses at which both amlodipine and verapamil suppressed macrophage recruitment by 54 and 80%, respectively, systolic blood pressure was unaffected (Table 1). Similar effects of these doses of LTCC blockers on blood pressure were observed by Kataoka et al. 33 These results suggest that the drugs were employed in a biologically relevant range.
The reduction in macrophage recruitment in response to TG in mice deficient in Plg 3 arises from their accumulation in the membrane (PM) of the peritoneal cavity 23. If the decrease of macrophage recruitment by the anti-hypertensive drugs was associated with modulation of Plg binding and Plg-R expression, the anti-inflammatory effects of amlodipine and verapamil also might arise from the accumulation of macrophages in the PM. H&E staining demonstrated a large accumulation of cells in the PM of mice treated at the lower, non-hypertensive dose of amlodipine and verapamil (Figure 6B). The accumulated cells in the PM of the drug treated mice were identified as macrophages based on Mac3 staining (Figure 6C): the abundance of Mac3 positive cells was15% ± 2.5 with amlodipine and 27% ± 5.3 with verapamil vs. 3.3% ± 0.4, p = 0.001) 25 vs.3.4% ± 0.4 (p = 0.001) in PM of vehicle treated mice. Thus, the accumulation was ~7-fold enhanced by verapamil and ~5-fold with amlodipine (Figure 6D).
To determine whether the accumulation of macrophages in the PM of animals on the antihypertensive drugs was due to effects on Plg-R expression, we examined the expression of the four Plg-Rs of interest on the surface of macrophages arrested in the PM of the drug-treated mice compared to accumulated macrophages of Plg-/- mice. Confocal analyses of PM sections revealed H2B, α-enolase, annexin 2 and p11 on the surface (cells were not permeabilized) of arrested macrophages (Figure 7A) in Plg-/- mice. H2B was also present in the matrix of the PM in the Plg-/- mice. Upon verapamil treatment, α-enolase levels were unaffected on the macrophages in the PM of the Plg+/+ mice compared to Plg-/- mice (Figure 6A and B). However, cell surface distribution of annexin 2 and H2B were drastically reduced in the drug treated Plg+/+ versus the Plg-/- mice (Figure 7A and B). p11 cell surface distribution was also reduced in drug treated macrophages compared to Plg-/- macrophages (Figure 7A and B). These results suggest that LTCC blockers are anti-inflammatory, and this phenomenon occurs in concert with reductions of Plg-Rs on the surface of macrophages.
Receptors for Plg play an important role in regulating inflammatory cell recruitment15, 16, 23, 34. In the present study, we have sought to address the longstanding dilemma as to how Plg-Rs are regulated on inflammatory cells, using THP-1 differentiation to macrophages to drive surface expression of Plg-Rs. Stimulation of THP-1 cells with IFNγ+VD3 led to 4-fold increase in Plg binding, consistent with a previous report 13. Concomitant with this increase, stimulation of the cells for 24-48 h led to increased cell surface expression of the Plg-Rs. For H2B, α-enolase and p11, the increase in cell surface localization occurred without detectable increases in total cellular protein and mRNA levels. In contrast, cellular protein expression, mRNA levels and cell surface expression of annexin 2 were all increased with differentiation of the THP-1 cells, consistent with a previous report 14.
Several inhibitors of known non-conventional pathways of protein export were tested, using glyburide (to block the ABC 1 transporter), ouabain (Na+/K+ ATPase), and methylamine (endosomal pathway). All of these inhibitors had minimal effects on Plg binding and Plg-R expression. In contrast, the intracellular [Ca2+]in chelator, BAPTA-AM, suppressed the increment in Plg binding induced by IFNγ+VD3 by 98% and completely blocked the increased expression of the four Plg-Rs assessed. To further dissect the role of [Ca2+]in in regulation of Plg-Rs, we tested the effects of verapamil and amlodipine, two distinct LTCC inhibitors,. Both LTCC blockers completely suppressed the upregulation of Plg-Rs by IFNγ+VD3. In contrast, ionomycin, which increases [Ca2+]in, led to a significant increase in Plg binding; its modest effect may be due to repression of H2B gene expression by ionomycin 35.
In support of the pharmacological data implicating the LTCC and cytosolic calcium in regulation of Plg-R membrane localization, we found that THP-1 cells, as well as mouse and human macrophages, express the LTCC, Cav1.2. To our knowledge, this is the first study to document the presence of Cav1.2 transcript and protein in human and mouse macrophages. Differentiation of monocytes to macrophages was associated with increased Cav1.2 expression and a concomitant increase in fMLP-stimulated Ca2+ mobilization (measured with fura2). This response was blocked by amlodipine, verapamil and siRNA against Cav1.2. In macrophages, the LTCC may be non-voltage gated, since addition of KCl did not induce intracellular Ca2+ mobilization. This speculation is consistent with the nature of LTCC in other non-excitable cells like T and B lymphocytes 18, 19, and likely explains our inability to measure voltage-gated calcium channel activity using whole cell patch clamp (data not shown).
That knockdown of SiCav1.2 with siRNA showed the same inhibitory effects as verapamil and amlodipine on Plg-R cell surface expression suggests that the drug effects are due to their interactions with LTCCs. SiCav1.2 inhibited upregulation of Plg-R cell surface expression, Plg binding and Plg-mediated matrigel invasion (Online Figure V A&B). Anti-H2B Fab fragments, which block Plg binding to H2B on the cell surface 15, reduced Plm mediated THP-1 cell matrigel invasion by 45% compared to non-immune rabbit IgG Fab (Online Figure V B), but there was no additional effect of anti-H2B Fab on the SiCav1.2 treated cells (Online Figure V B) showing that the suppressive effect of SiCav1.2 on Plm mediated THP-1 matrigel invasion was indeed due to reduction of one or more candidate Plg-Rs.
Previous work by Brown et al 36 showed that verapamil reduced TNF-α mRNA expression in response to ultrafine particles in rat alveolar macrophages, but we did not observe changes in whole cell protein or mRNA expression of Plg-Rs in THP-1 cells by either verapamil or amlodipine. Studies of intracellular signaling have implicated adenylase cyclase and ERK/MAPK as pathways dependent on LTCC, including in IFNα mediated enolase cell surface expression. However in our study, inhibitors of these events failed to block upregulation of Plg-Rs. The details of the pathway(s) by which LTCC and intracellular Ca2+ modulate Plg-R cell surface localization remains to be elucidated in future studies.
Macrophage recruitment in a TG-induced mouse peritonitis model is reduced by >60% in Plg deficient mice3 and is associated with a large accumulation of these cells in the PM 23. In our study, we showed that both amlodipine and verapamil led to a marked suppression (82% to 87%) of macrophage emigration into the peritoneal cavity. The anti-inflammatory effects of amlodipine and verapamil have been demonstrated in previous studies in other models 20-22, 28, 33. As in Plg-/- mice, suppression of migration was associated with accumulation of macrophages in the PM of the verapamil and amlodipine treated mice. In comparing accumulated macrophages in mice treated with LTCC blockers and Plg deficiency, the drugs reduced cell surface expression of H2B and annexin 2 markedly. Cell surface localization of p11 was also reduced, but not as extensively, and α-enolase expression was unaltered. The availability of low levels of Plg-Rs may explain why the drugs were slightly less effective than absence of Plg in suppressing macrophage recruitment. In addition to cell surface localization, H2B protein was found in the ECM in the Plg-/- mice. Heparan sulfate proteoglycans are known to bind the N-termini of histones 37, and released histones have been shown to localize in the ECM 38. Matrix bound H2B may provide another mechanism for localizing Plg, enhancing its activation to Plm and, thereby, facilitating inflammatory cell migration.
In summary, our study identifies a pivotal role of increased [Ca2+]in in the upregulation of cell surface expression of Plg-Rs upon monocyte differentiation into macrophages. We further show that verapamil and amlodipine, two commonly used LTCC blockers, suppress surface expression of Plg-Rs. The influence of these drugs on cell surface expression of Plg-Rs has been demonstrated both in vitro and in vivo. This previously unrecognized activity of these widely used drugs may contribute to their beneficial anti-inflammatory effects.
The authors gratefully acknowledge Cleveland Clinic coworkers Christina Gaughan and Rui Chen for assistance in BP and [Ca2+]in measurement, respectively. We thank Dr. Jane Hoover-Plow and lab members, Dr. Maradumane Mohan for valuable discussion and technical help with in vitro experiments and Drs. Judy Drazba and John Peterson for help with microscopy studies.
Sources of Funding: This work was supported by NIH Grant HL17964 to EFP and AHA Fellowship 0825638D to RD.