18 Aug 2011

Nuclear and cytoplasmic LIMK1 enhances human breast cancer progression

Brice V McConell, Karen Koto and Arthur Guiterrez Hartmann University of Colorado

Presented by Rohit Shrivastava Research Fellow, ACTREC

Introduction to LIMK1
LIM domain containing serine/Theronine Kinase. Expressed in brain, spinal cord and sensory organs

LIMK1 substrates ADF, cofilins. Inhibitors phosphatases, Slingshots and chronophin

LIMK1

LIMK1- ch 7 and LIMK2, - ch 22 regulates actin reorganization

LIMK1 is downstream effector of Rho signalling pathway

B TG VE Structure of LIMK1 MP F GF Recept Proline/ orSerine

Rho GTP Rho Kinase RacGTP NE S Cdc4 2-GTP p21 Kinase 1&4

2 ZINC finge r

NL S

P
NH2

LIM1

LIM 2

N PDZ E S

NT508 KINASE - COOH L S

image adapted from Brice et al., (201

P
NH2

NH2

LIM 2 Hsp 90 LIM 2

NN PDZ E E SS NN PDZ E E SS

NT508 KINASE - COOH L S N KINASE L S P - COOH

Cofilin MicroTubule Tubulin

Cofilin

T508
P

14-33

NH2

LIM 1 BM P

LIM 2

NN PDZ E E SS

14-33

NT508 KINASE - COOH L S

Leads to high kinase activity, causes dendritogenesis

SS H LAT S1
Tumor suppressor regulates cell cycle and apoptosis. Inactivates LIMK1 in cytokinesis Causes downregulatio n of LIMK1 by dephosphoryla tion

p160ROCK LPA LPAR Rho mDia

LIMK
MLC Phosphatase

PDG F

PDGFR

Rac

p65PAK

LIMK

Bradyki nin

Bradyki nin- R

Cdc4 2

Filipodia

actin polymeriz ation

Lamellipodia

Stress Fibers

Significance of LIMK1

F-actin stabilizati on Actin :myosin crossilinki ng

Hypothesis
What we know
LIMK1 expression in human breast cancer cell lines (MCF-7, MDA-MB-231 and MDA-MB435) resulted in increased cellular invasion and xenograft tumor growth which were larger, more vascularized and more likely to metastasize and downregulation of LIMK1 abrogrates these changes. It is established that phosphorylation of cofilin and thus actin filaments are stabilized.

We wanted to know

Whether the sub-cellular localization, cytoplasmic versus nuclear, of LIMK1 affects its ability to promote the transformed phenotype.

Methods and Materials

Cell line

MDA MB 231 cells

Bgl II Sac II LIMK1 cDNA without start codon was taken from FPC-1-myc LIMK1 and inserted into Bgl II Sac II -cut pEGFP-C1-LIMK1 to produce pEGFP C1LIMK1. NLS-GFP-LIMK1 AND NES-GFP-LIMK1 were PCR amplified and along with GFP- LIMK1 ligated into Age I Pac I cut pQCXIN plasmid. NLS and NES tagged GFP LIMK1 were produced in pQCXIN retroviral vector. Phoenix cells were used to package pQCXIN based retroviruses. Packing cells were grown in appropriate medium and conditions according to manufacturers protocol. Virus containing supernatant was collected at different time intervals, syringe filtered and stored at -80 C. To infect MDA cells viral supernatant was diluted in growth medium and incubated overnight. Cells expressing stable fusions of GFP-LIMK1 were selected using G-418, while expression of EGFP tagged LIMK1 was detected by microscopy.

Plasmid Construct

Transductio n and generation of stable cell pool

Immunohistoche mical Analysis

Antigen retrival was performed by treating with sodium citrate and heating . Endogenous peroxidases were blocked by treating with hydrogen peroxide. Slides were then blocked with goat serum. Slides were then treated with goat polyclonal antibody overnight , which recognizes C-terminus of LIMK1 then treated with biotinylated anti goat IgG secondary antibody. All slides were then treated with avidin-biotinylated horse radish peroxidase . Antigen antibody complex was then visualized by treatment 231 cells expressing GFP or the various GFP -LIMK 1 with peroxidase substrate. Nuclei were visualized MDA- MBwith Mayers grown. Cells were then washed according to the fusions were hematoxylin.
protocol. Separate cytoplasmic and nuclear lysate were

Western blot

obatined. All lysis buffer were supplemented with Complete protease inhibitor cocktail. And PhosSTOP phosphatase inhibitor cocktail. Total protein (25-100 g) from each lysate was subjected to sodium dodecyl sulfate 10% polyacrylamide gel electrophoresis (SDS-PAGE) and electrophoresed proteins were transferred to Immobilon-P membranes. Membranes were blocked f or 1 - 2 hr in non-fat dry milk for phosphospecific primary antibodies. Primary antibodies were incubated overnight at4C. Polyclonal goat HRP-conjugated secondary antibodies against mouse, rabbit or rat were incubated on membranes for 1 hr at room temperature. After primary analysis, each blot was stripped using the Chemicon strong reblot reagent prior to re-probing with additional primary antibodies.

Immunofluor escence microscopy

MDA- MB- 231 cells expressing GFP or the various GFP -LIMK 1 fusions were fixed in 4 % PFA and then permeabilized with 0.5% Triton X - 100. Cells were then blocked for 1 h -2 h in PBST with 5% goat serum. Following the blocking incubation, sections were incubated overnight at 4C with phospho -FAK antibody. Sections were incubated for 1 h with a goat IgG secondary antibody counterstained with Alexa Fluor 647-conjugated phalloidin to visualize the filamentous actin.

Matrigel Invasion Assay

Matrigel-based trans-well invasion assays. 5 10 4 MDA-MB-231 cells in DMEM+0.1% BSA were plated in 24-well plates with DMEM+5% FBS as chemo-attractant. After 24 hours, the cells w ere fixed in 4% PFA and the invading cells on the underside of the filter were stained with Hoechst stain. Xenograft experiments were conducted in 7- 8 week old female nude mice. MDA - MB - 231 cells expressing each of the GFP- LIMK1 fusions as stable pools were harvested in PBS/EDTA and re-suspended in Matrigel. 2 106 MDA-MB-231 cells were injected bilaterally onto mammary fat pads #5 in a 50 l volume of Matrigel, with 6- 10 animals injected per cell line.

Nude mouse xenograft tumor Assay

Statistical Analysis

All data was analyzed using SAS software v 9.2, Pairwise comparisons of group means were conducted using the Tukey-Kramer

Results
Immunohistochemical Analysis

LIMK1 expression in both cytoplasm and nucleus of both normal mammary and human breast cancer tissue. Broken arrow depicts cytoplasmic LIMK1 staining, solid arrow depicts insignificant nuclear staining for LIMK1 and open arrow depicts nuclear LIMK1 staining.

Sub-celullar localization of GFP-LIMK1

GFP-alone and three GFP- tagged LIM K1 proteins are depicted graphically and color coded. GFP was fused to the N-terminus of the LIMK1 cDNA. Exogenous NLS and NES tags were fused to the N-terminus of GFP. us NLS and NES sequences target NLS-GFP-LIMK1 and NES-GFP-LIMK1 lear and cytoplasmic subcellular compartments. Fluorescence microscopy w Western blot analysis of cytoplasmic and nuclear fractions from MDAsualize subcellular localization of GFP fluorescence (green) in MDA-MB-2 MB-231 stable shed lines are drawn to outline the nucleus of individual PARP. transductants. Nuclear segregation is assayed by total cells. represents 20 microns. Cytoplasmic segregation is assayed by GAPDH. GFP-tagged LIMK1 is

Western Blot analysis of expression and phosphorylation status of GFP-LIMK1 fusion proteins and cofilin

n blot analysis of whole cell lysates from MDA-MB-231 stable transductants probed for (A) LIMK1 using onal antibody and for tubulin using anti-tubulin antibody; (B) Probed for phosphoryated LIMK1 at T508 (endogenous pLIMK1 is not detectable by Western blot with this antibody); (C) probed for total cofilin and, (D) probed for phosphoryated cofilin at Ser3 and tubulin.

Western Blot analysis of phosphorylation status of FAK signaling proteins

(A) Western blot analysis of whole cell lysates from MDA-MB-231 stable transductants probed with antibodies against pFAK, total FAK, pPaxillin, total Paxillin, pSrc, total Src, pAKT, total AKT, pErk1/2 and total Erk 1/2. (B) Western blot analysis for tubulin or GAPDH for each extract used

Immunofluorescence microscopic analysis of LIMK1 expression and its correlates with phosphorylation of FAK

-231 c ells expressing GFP-only or various GFP-LIMK1 fusions were fixed and stained with antibodies ag -FAK (red), and phalloidin against actin (gold). GFP fluorescence is green in this figure. Images were ob scence microscopy (non-confocal). Cellular regions with focal adhesion structures are marked by white ospho-FAK images. Scale bar represents 20 microns. Note: non-specific binding of secondary antibody s d.

Effect of cytoplasmic and nuclear LIMK1 on cellular invasiveness

Invasion assays of MDA-MB-231 cells expressing GFP-only (solid), NLS-GFP-LIMK1 (vertical), NES-GFP-LIMK1 (striped) or GFPLIMK1 (lines) after 24 hours of invasion through Matrigel-coated invasion chambers. Error bars represent SEM from 8 separate experiments performed in triplicate. Means of each treatment groups

Xenograft tumor growth studies in nude mouse

Expression of GFP-LIMK1 fusion enhanced tumor growth. The results represent two independent experiments (A/B and C/D) in which 2 10 6 cells MDA-MB-231 cells, in a 50 l volume of Matrigel, were injected bilaterally onto mammary fat pads #5 of female athymic-nude mice. Error bars represent SEM, and p-values for all three experimental groups are equal to or less than 0.05, compared to GFP-control. (A) Tumor volume was estimated by measurements from electronic calipers over the course of the assay. Statistical analysis was performed by mixed model linear regression analysis. Tumor weights were measured upon excision of the tumors in the study shown in (A). The p-value of GFPLIMK1 is 0.006, compared to GFP-alone (*). The p-value of NLS-GFP-LIMK1, compared to GFP-alone, is 0.37 (#). Statistical analysis of tumor weights was performed by pairwise comparison in one-way ANOVA.

Conclusion
Both nuclear and cytoplasmic targeted GFP-LIMK1 enhanced FAK/ paxillin /Src/AKT/Erk signaling by phosphorylation. Imaging study reveals GFP-LIMK1 is detected in both cytoplasm and nucleus. Expression of LIMK1 in nucleus resulted in the increased activation of FAK/paxillin/Src/AKT/Erk signaling by a mechanism still not clear. Possible role of cofilin. In vitro invasion data suggests that cofilin is the key factor regulating cell motility and invasion. But differential tumor promoting effects observed in vivo which could be de to some unknown LIMK1 pathways or because system reached threshold. Since LIM domain participates in transcription events, so it is possible that LIMK1 may mediate tumor progression via effects on p57kip2 or serum response factor both of these are regulated by cytoplasmic LIMK1.

Acknowledgement

Dr Vikram for inculcating a good habit of reading, for reference journal and for starting the Journal Club Areba & Charneet (Sarins Lab) for her help with plasmid construct and transduction Asha (Genetic Engineering) for her help with retroviral vector expression and plasmid packaging.

References Sayi, A (2008). Functional Analysis of the LIMK1 an its role in cell cycle progression. University of Basel. Sumi, T. et al., (2006). Differential activity regulation and subcellular localization of LIMK1 and LIMK2 during cell cycle. Experimental cell research. 1021-1030. Avizienyte, E and Frame, M.C. (2005). Src and FAK signalling controls adhesion fate and the epithelial-to-mesenchymal transition. Current opinion in cell biology. 17:542 547. Bernard, O. (2007). Lim Kinases, regulators of actin dynamics. International journal of biochemistry and cell biology. 39: 1071-1076. Nadella, K.S. et al., (2009). Regulation of actin function by protein kinase A-mediated phosphorylation of LIMK1. European molecular biology organization report. 10: 599-606