Leukamenin E Induces K8/18 Phosphorylation and Blocks the Assembly of Keratin Filament Networks Through ERK Activation

Abstract

Leukamenin E is a natural ent-kaurane diterpenoid isolated from Isodon racemosa (Hemsl) Hara that has been found to be a novel and potential keratin filament inhibitor, but its underlying mechanisms remain largely unknown. Here, we show that leukamenin E induces keratin filaments (KFs) depolymerization, largely independently of microfilament (MFs) and microtubules (MTs) in well-spread cells and inhibition of KFs assembly in spreading cells. These effects are accompanied by keratin phosphorylation at K8-Ser73/Ser431 and K18-Ser52 via the by extracellular signal-regulated kinases (ERK) pathway in primary liver carcinoma cells (PLC) and human umbilical vein endothelial cells (HUVECs). Moreover, leukamenin E increases soluble pK8-Ser73/Ser431, pK18-Ser52, and pan-keratin in the cytoplasmic supernatant by immunofluorescence imaging and Western blotting assay. Accordingly, leukamenin E inhibits the spreading and migration of cells. We propose that leukamenin E-induced keratin phosphorylation may interfere with the initiation of KFs assembly and block the formation of a new KFs network, leading to the inhibition of cell spreading. Leukamenin E is a potential target drug for inhibition of KFs assembly.

Keywords: inhibitory effect of keratin filaments assembly 3; leukamenin E 1; phosphorylation of keratin 2.

Figure 1

Effects of leukamenin E on proliferation and apoptosis in human umbilical vein endothelial cells (HUVECs), primary liver carcinoma cells (PLC), and Panc-1 cells. (A) Chemical structure of leukamenin E. (B) Cell viability assay of HUVECs, PLC, and Panc-1 treated with leukamenin E (2 μM-4 μM) for 24 h based on MTT. (C) Evaluation of apoptosis of HUVECs (a,b), PLC (c,d), and Panc-1 (e,f) cells treated with leukamenin E by acridine orange/ethidium bromide (AO/EB) double staining. Four types of cells were observed by AO/EB staining: viable cells (V), early apoptotic cells (EA), necrotic cells (N), and late apoptotic cells (LA), as indicated by the arrows in the pictures. Bar, 40 µm. (D) Apoptotic rate in HUVECs, PLC, and Panc-1 cells treated with leukamenin E for 24 h. (E) Necrotic rate in HUVECs, PLC, and Panc-1 cells treated with leukamenin E for 24 h. Images represent typical cells from at least three independent experiments. The data shown represent the means ± standard deviation (S.D.) of three independent experiments. * p < 0.05, νersus the control group.

Figure 2

Cytochalasin B and colchicine cause the depolymerization of microfilament (MFs) and microtubules (MTs), respectively, in well-spread PLC and HUVECs but do not distinctly alter the keratin filament (KF) network. (A) Representative images of cell morphology and three cytoskeleton components in well-spread PLC and HUVECs treated with cytochalasin B (b,j,f,n) or colchicine (d,l,h,p) or DMSO (control, a,i,e,m,c,g,k,j,o). 1000×, Bar, 10 µm. (B) Representative images of cell morphology and three cytoskeleton components in well-spread PLC (a,b,c) and HUVECs (d,e,f) treated with cytochalasin B and colchicine. 1000×, Bar, 10 µm. The binucleate cells caused by cytochalasin B are indicated by arrows. Images represent typical cells from at least three independent experiments.

Figure 3

Effects of cytochalasin B and colchicine on KFs reassembly and cell spreading. (A) Representative images of KF network reassembly and cell morphology at 4–12 h after suspended PLC and HUVECs were seeded in media with cytochalasin B and colchicine (d,e,f,j,k,l) or DMSO (control, a,b,c,g,h,i), showing that treatment with cytochalasin B and colchicine does not inhibit KFs reassembly and cell spreading. 1000×, Bar, 10 µm. (B,C) Areas of cell spreading were measured at 4 h, 8 h, and 12 h after suspended HUVECs or PLC were seeded in medium-containing cytochalasin B and colchicine or in control medium (DMSO). The binucleate cells caused by cytochalasin B are indicated by arrows. Images represent typical cells from at least three independent experiments. The data shown represent the means ± standard deviation (S.D.) of three independent experiments.

Figure 4

Leukamenin E destroys the KF network in HUVECs and PLC. (A) Representative images of cell morphology and three cytoskeleton components in well-spread HUVECs and PLC treated with 2.0 μM leukamenin E (d,m,e,n,f,o) or 4.0 μM leukamenin E (g,p,h,q,i,r) or DMSO (control, a,j,b,k,c,l) for 24 h, showing that leukamenin E destroys the KF network (f,I,o,r) but slightly alters MFs (e,h,n,q) and MTs (d,g,m,p). 1000×, Bar, 10 µm. (B) Representative images of cell morphology and three cytoskeleton components in well-spread HUVECs and PLC treated with a combination of leukamenin E, cytochalasin B, and colchicine for 24 h, showing that leukamenin E destroys the KF network (c,f) independently of MFs (b,e) and MTs (a,d). 1000×, Bar, 10 µm. Images represent typical cells from at least three independent experiments.

Figure 5

Leukamenin E inhibits KFs reassembly and cell spreading in HUVECs and PLC. (A) Representative images of KF network reassembly and cell morphology at 4–12 h after the suspended PLC and HUVECs were seeded in media with leukamenin E (d,e,f,j,k,l) or DMSO (control, a,b,c,g,h,i), showing the inhibition effect of leukamenin E on KFs reassembly and cell spreading. 1000 ×, Bar, 10 µm. (B,C) Areas of cell spreading were measured at 4 h, 8 h, and 12 h after suspended HUVECs or PLC were seeded in medium-containing leukamenin E or DMSO (control). Images represent typical cells from at least three independent experiments. The data shown represent the means ± standard deviation (S.D.) of three independent experiments. * p < 0.05, ** p < 0.01, νersus the control group.

Figure 6

Effects of leukamenin E on keratin phosphorylation in PLC and HUVECs. (A) Leukamenin E stimulates phosphorylation of K8 at Ser73 and Ser431 and K18 at Ser52 but does not alter the expression of K8 and K18 in well-spread cells. (B) Leukamenin E stimulates phosphorylation of K8 at Ser73 and Ser431 and K18 at Ser52 in spreading cells. (C–E) Quantification of keratin or phosphorylated keratin by grey level analysis. The data represent the means ± standard deviation (S.D.) of three independent experiments. * p < 0.05, ** p < 0.01 versus control group.

Figure 7

Leukamenin E induces phosphorylation at K8-Ser431, K8-Ser73 and K18-Ser52 to increase keratin solubility in HUVECs. (A) Immunofluorescence imaging shows that leukamenin E leads to increased phosphorylated KFs fraction from keratin network. Keratin immunostaining was performed using the pK8-Ser431 antibody (b,e), pK8-Ser73 (h,k) antibody, pK18-Ser52 antibody (n,q), and K18 antibody (a,d,g,j,m,p). The mergers are shown in c, f, i, l, o and r, respectively. 1000×. (B) Graph depicts the quantification of K8 or phosphorylated K8 at Ser431/Ser73/Ser52 by fluorescence intensity analysis. (C) Protein content in supernatant and precipitation was determined by grey level analysis. (D) Cells were subsequently lysed, and the keratin content in the supernatant and pellet was analyzed by Western blotting using pan-keratin /pK8-Ser431/pK8-Ser73/pK18-Ser52antibody, which showed that leukamenin E increases soluble fraction of KFs or phosphorylated keratin in well-spread cells. Images represent typical cells from at least three independent experiments. The data shown represent the means ± standard deviation (S.D.) of three independent experiments. * p < 0.05, ** p < 0.01, νersus the control group.

Figure 8

Extracellular signal-regulated kinases (ERK) mediates leukamenin E-induced K8/18 phosphorylation and increased keratin solubility in HUVECs and PLC. (A–C) Leukamenin E-induced activation of ERK and P38 phosphorylation in HUVECs and PLC and their quantitative analysis. (D) ERK mediates leukamenin E-induced K8/18 phosphorylation. HUVECs and PLC were treated with 10 μM U0126 for 1 h followed by incubation with 2.0 μM leukamenin E for 24 h. Western blotting was performed using pK8-Ser431, pK8-Ser73, and pK18-Ser52 antibody. (E) Graph depicts the quantification of keratin by grey level analysis. (F) ERK mediates leukamenin E-induced increase of keratin solubility. Well-spread HUVECs were treated with 10 μM U0126 for 1 h followed by incubation with 2.0 μM leukamenin E for 24 h or the spreading HUVECs were treated with 10 μM U0126 for 1 h followed by incubation with leukamenin E at the indicated concentrations for 8 h. WB was performed using Pan-K antibody. (G) Graph depicts the quantification of keratin by grey level analysis. The data shown represent the means ± standard deviation (S.D.) of three independent experiments. * p < 0.05, ** p <0.01, νersus the control group.

Figure 9

Leukamenin E inhibits migration of HUVECs and Panc-1 cells. (A) Scratch wound healing assays were performed to detect the migration of the indicated cells. Images of the scratch wound were captured after 12 h of culture with or without leukamenin E. 100× multiplication. (B,C) Graph quantifying the rate of wound closure in HUVECs and Panc-1 cells. (D,E) HUVECs and Panc-1 cells were incubated with 2.0 μM and 4.0 μM. The number of cells migrating through size-limited pores was determined by inverted microscopy. The results are presented as means ± SEM. The data shown represent the means ± standard deviation (S.D.) of three independent experiments. * p < 0.05, ** p < 0.01, νersus the control group.

Figure 10

Working Model. Our data are consistent with a model in which leukamenin E-induced phosphorylation at K8-Ser73/431 and K18-Ser52 was involved in increased soluble fraction of KFs and altered polymerization. Leukamenin E may interfere with initiation of KFs assembly at the periphery of the cell and block step-by-step assembly of KFs and the formation of a new KFs network and then inhibit turnover of KFs in HUVECs.