Methods in cell biology, volume 128

In vitro systems for thestudy of microtubule-based cell polarity in fission yeast 1 Nu´ria Taberner * , Andries Lofx, Sophie Roth * , Dimitry Lamersx, Hans Zeijlemakerx, Marileen Dogter

Methods in Cell

Biology Building a Cell from its

Component Parts

Volume 128

Philadelphia, USA &

Institut Curie, Paris, France

Methods in Cell

Biology Building a Cell from its

Department of Biochemistry & Biophysics, University of

California San Fransisco at Mission Bay, USA

AMSTERDAM • BOSTON • HEIDELBERG • LONDON

NEW YORK • OXFORD • PARIS • SAN DIEGO

SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO

Academic Press is an imprint of Elsevier

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

First edition 2015

Copyright© 2015 Elsevier Inc All Rights Reserved

No part of this publication may be reproduced or transmitted in any form or by any means,electronic or mechanical, including photocopying, recording, or any information storageand retrieval system, without permission in writing from the publisher Details on how toseek permission, further information about the Publisher’s permissions policies and ourarrangements with organizations such as the Copyright Clearance Center and theCopyright Licensing Agency, can be found at our website:www.elsevier.com/permissions.This book and the individual contributions contained in it are protected under copyright bythe Publisher (other than as may be noted herein)

Notices

Knowledge and best practice in this field are constantly changing As new research andexperience broaden our understanding, changes in research methods, professionalpractices, or medical treatment may become necessary

Practitioners and researchers must always rely on their own experience and knowledge inevaluating and using any information, methods, compounds, or experiments describedherein In using such information or methods they should be mindful of their own safetyand the safety of others, including parties for whom they have a professional responsibility

To the fullest extent of the law, neither the Publisher nor the authors, contributors, oreditors, assume any liability for any injury and/or damage to persons or property as amatter of products liability, negligence or otherwise, or from any use or operation of anymethods, products, instructions, or ideas contained in the material herein

ISBN: 978-0-12-802450-8

ISSN: 0091-679X

For information on all Academic Press publications

visit our website athttp://store.elsevier.com

Jose´ Alvarado

FOM Institute AMOLF, Amsterdam, The Netherlands; Massachusetts Institute of

Technology, Cambridge, MA, USA

R Ayadi

Natuur- en Sterrenkunde and LaserLab, Vrije Universiteit, De Boelelaan,

Amsterdam, The Netherlands

Swathi Ayloo

Department of Physiology and the Pennsylvania Muscle Institute, Perelman

School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA;

Department of Biology Graduate Group, School of Arts and Sciences at the

University of Pennsylvania, Philadelphia, PA, USA

Hella Baumann

London Research Institute, Cancer Research UK, London, UK

Kevin Carvalho

Institut Curie, Centre de Recherche, Paris, France; CNRS, UMR168, Paris,

France; UPMC Univ Paris 06, UMR 168, Paris, France; Univ Paris Diderot,

Sorbonne Paris Cite´, Paris, France

Guilherme Pereira Correia

Wellcome Trust/Cancer Research UK Gurdon Institute and Department of

Biochemistry, University of Cambridge, Cambridge, UK

Michael Diehl

Department of Chemistry, Rice University, Houston, TX, USA; Department of

Bioengineering, Rice University, Houston, TX, USA

Marileen Dogterom

Department of Bionanoscience, Kavli Institute of Nanoscience, Delft University of

Technology, Delft, The Netherlands

Christine M Field

Department of Systems Biology, Harvard Medical School, Boston, MA, USA;

Marine Biological Laboratory, Woods Hole, MA, USA

Daniel A Fletcher

Department of Bioengineering & Biophysics Program, University of California,

Berkeley, CA, USA; Physical Biosciences, Lawrence Berkeley National

Laboratory, Berkeley, CA, USA

xi

Adam Frost

Department of Biochemistry, University of Utah, School of Medicine, Salt LakeCity, UT, USA; Department of Biochemistry and Biophysics, University ofCalifornia, San Francisco, San Francisco, CA, USA

Kinneret Keren

Department of Physics, Technion-Israel Institute of Technology, Haifa, Israel;Russell Berrie Nanotechnology Institute, Technion-Israel Institute of Technology,Haifa, Israel; Network Biology Research Laboratories, Technion-Israel Institute ofTechnology, Haifa, Israel

Joe¨l Lemie`re

Institut Curie, Centre de Recherche, Paris, France; CNRS, UMR168, Paris,

France; UPMC Univ Paris 06, UMR 168, Paris, France; Univ Paris Diderot,

Sorbonne Paris Cite´, Paris, France; Current address: Department of Molecular

Biophysics and Biochemistry, Nanobiology Institute, Yale University, New Haven,

CT, USA

Binyong Liang

Center for Membrane Biology and Department of Molecular Physiology and

Biological Physics, University of Virginia, Charlottesville, VA, USA

Allen P Liu

Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI,

USA; Department of Biomedical Engineering, University of Michigan, Ann Arbor,

MI, USA; Cellular and Molecular Biology Program, University of Michigan,

Ann Arbor, MI, USA; Biophysics Program, University of Michigan, Ann Arbor,

MI, USA

Andries Lof

FOM Institute AMOLF, Amsterdam, The Netherlands

Martin Loose

Department of Systems Biology, Harvard Medical School, Boston, MA, USA;

Institute of Science and Technology Austria, Klosterneuburg, Austria

Maya Malik-Garbi

Department of Physics, Technion-Israel Institute of Technology, Haifa, Israel

Timothy J Mitchison

Department of Systems Biology, Harvard Medical School, Boston, MA, USA;

Marine Biological Laboratory, Woods Hole, MA, USA

Kiyoshi Mizuuchi

Laboratory of Molecular Biology, National Institute of Diabetes, and Digestive and

Kidney Diseases, National Institutes of Health, Bethesda, MD, USA

Victoria L Murray

Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI,

USA

Phuong A Nguyen

Department of Systems Biology, Harvard Medical School, Boston, MA, USA;

Marine Biological Laboratory, Woods Hole, MA, USA

Anand Radhakrishnan

Department of Chemistry, Rice University, Houston, TX, USA; Department of

Bioengineering, Rice University, Houston, TX, USA

Enas Abu Shah

Department of Physics, Technion-Israel Institute of Technology, Haifa, Israel;Russell Berrie Nanotechnology Institute, Technion-Israel Institute of Technology,Haifa, Israel

Nu´ria Taberner

Department of Bionanoscience, Kavli Institute of Nanoscience, Delft University ofTechnology, Delft, The Netherlands

Nathaniel Talledge

Department of Biochemistry, University of Utah, School of Medicine, Salt Lake

City, UT, USA; Department of Biochemistry and Biophysics, University of

California, San Francisco, San Francisco, CA, USA

Lukas K Tamm

Center for Membrane Biology and Department of Molecular Physiology and

Biological Physics, University of Virginia, Charlottesville, VA, USA

James A Taylor

Laboratory of Molecular Biology, National Institute of Diabetes, and Digestive and

Kidney Diseases, National Institutes of Health, Bethesda, MD, USA

David Tsao

Department of Chemistry, Rice University, Houston, TX, USA; Department of

Bioengineering, Rice University, Houston, TX, USA

Anthony G Vecchiarelli

Laboratory of Molecular Biology, National Institute of Diabetes, and Digestive and

Kidney Diseases, National Institutes of Health, Bethesda, MD, USA

Astrid Walrant

Wellcome Trust/Cancer Research UK Gurdon Institute and Department of

Biochemistry, University of Cambridge, Cambridge, UK

Hans Zeijlemaker

FOM Institute AMOLF, Amsterdam, The Netherlands

Katja Zieske

Department of Cellular and Molecular Biophysics, Max Planck Institute for

Biochemistry, Martinsried, Germany

The cell interior is another world that we are only beginning to explore Although

there are a number of approaches for examining the inner workings of the cell,

the reductionist approach of building up complexity appeals to many with physical

science and engineering backgrounds Such experiments are appealing because the

complexity of the cellular organism can be built up systematically adding in one new

element at a time and dialing up the concentrations of each component in a

system-atic way The hope is that such bottom-up approaches will garner new quantitative

insights at many levels At the lowest level, we seek to elucidate single molecule

activities to be able to have quantitative numbers for analytical models and

simula-tions As new components are added to one another, their activities can combine

synergistically to create novel emergent behaviors Since biology is inherently out

of equilibrium, there is energy being injected that can enhance such synergies

mak-ing even simple-seemmak-ing systems appear complex and inherently interestmak-ing Such

emergence of complex behavior from interactions of simple components is likely

to have been critical for the origins of life Further, reductionist scientists are seeking

to determine the minimal diversity of components to create cell-like structures and

activities Especially in eukaryotic systems, the number of components controlling

cell function appears unnecessarily large and the methods for cellular control appear

baroque By determining the minimal variety of proteins of each function to perform

a specific task, scientists are beginning to create a molecular toolkit of activities

This avenue of research also has implications for a new area of cellular

bioengi-neering based on cell-like nanodevices

Another interesting issue of the cell is the variety and diversity of systems one

can choose to work on Of course, reductionists are seeking to give fundamental

and universal insight to these processes, so they typically focus on aspects of the

cell that appear fundamental As such, many of the authors contributing to this

volume have chosen to focus on the cytoskeleton and membrane systems to

ulti-mately combine the two into cell-like patterns and organizations

We have grouped together contributed chapters that have similar focuses, starting

with wholly in vitro reconstitutions of cytoskeleton These contributions include

methods for combining different motors and cytoskeletal components in defined

ways to produce more complex behaviors, as well as methods to combine

cytoskel-etal assemblies with fabricated devices such as chambers or pillar arrays

Next we move to the membrane systems reconstituted in vitro Contributions in

this area include chapters on reconstituting membrane fission and fusion, as well as

reconstituting important biological processes that normally take place on membrane

surfaces, such as cell division and polarity

Finally, we finish with systems that attempt to recreate cells as encapsulated

systems of chemicals that can act together as a machine to perform functions These

chapters describe methods for encapsulating protein machines within vesicles or

xvii

droplets The chapters of this book thus span a range of spatial scales from singleprotein molecules to vesicle- and cell-sized structures capable of complex behaviors.These represented systems certainly are not complete, and many more scientistsare applying quantitative, systematic strategies to understand DNA, the nucleus,organelles, and countless other systems It is our hope that this volume will beinstructive to ultimately enable researchers to bridge the gaps between differentexperimental systems Until all the reduced systems are combined together, wewill not truly be able to reconstruct a working cell.

Jennifer RossWallace F Marshall

xviii Preface

In vitro systems for the

study of

microtubule-based cell polarity in

fission yeast

1

Nu´ria Taberner * , Andries Lofx, Sophie Roth * , Dimitry Lamersx,

Hans Zeijlemakerx, Marileen Dogterom * , 1

*Department of Bionanoscience, Kavli Institute of Nanoscience, Delft University of Technology,

Delft, The Netherlands

xFOM Institute AMOLF, Amsterdam, The Netherlands

1 Corresponding author: E-mail: M.Dogterom@tudelft.nl

CHAPTER OUTLINE

Introduction 2

1 Rationale 3

2 Materials 5

2.1 Materials for Method 1: Elongated Glass Wells with TiO2Overhang 5

2.1.1 Microfabrication 5

2.1.2 Surface functionalization 6

2.1.3 Assay 6

2.2 Materials for Method 2: Elongated Water in Oil Emulsion Droplets 8

2.2.1 Microfluidic chip 8

3 Methods 9

3.1 Method 1: Elongated Glass Wells with TiO2Overhang 9

3.1.1 Design 9

3.1.2 Deposition of chromium and titanium oxide by electron-beam evaporation 9

3.1.3 Postannealing of titanium oxide 10

3.1.4 UV-lithography with S1813 11

3.1.5 Reactive ion etching 11

3.1.6 Wet etching with KOH 11

3.1.7 Surface functionalization 11

3.1.8 Assays 12

3.2 Method 2: Elongated Water in Oil Emulsion Droplets 14

3.2.1 Microfluidic design 14

3.2.2 Masks for microfluidic chip fabrication 16

Methods in Cell Biology, Volume 128, ISSN 0091-679X, http://dx.doi.org/10.1016/bs.mcb.2015.02.008

© 2015 Elsevier Inc All rights reserved. 1

3.2.3 SU-8 negative mould 16

3.2.4 PDMS chip 16

3.2.5 Microfluidic setup and droplet formation 17

3.2.6 Lipids composition 18

3.2.7 Assays 18

Discussion and Perspectives 19

Acknowledgments 20

References 20

Abstract

Establishment of cell polarity is essential for processes such as growth and division In fission yeast, as well as other species, polarity factors travel at the ends of microtubules to cortical sites where they associate with the membrane and subsequently maintain a polarized activity pattern despite their ability to diffuse in the membrane In this chapter we present methods to establish an in vitro system that captures the essential features of this process This bottom-up approach allows us to identify the minimal molecular requirements for microtubule-based cell polarity We employ microfabrication techniques combined with surface functionalization to create rigid chambers with affinity for proteins, as well as microfluidic techniques to create and shape emulsion droplets with functionalized lipid boundaries Preliminary results are shown demonstrating that a properly organized micro-tubule cytoskeleton can be confined to these confined spaces, and proteins traveling at the ends of growing microtubules can be delivered to their boundaries

INTRODUCTION

Fission yeast (Schizosaccharomyces pombe) is a unicellular eukaryote that is frequently used as a model system to study the cell cycle and cell polarity It has

a rod-like shape of about 3 mm diameter that grows unidirectionally from 7 to

14 mm by extension from the cell poles and divides by medial fission (Mitchison

& Nurse, 1985) Spatio-temporal recognition of the poles and the center of the cell are achieved by specific proteins associated with the cell poles, such as the kelch repeat protein tea1p and the kinase pom1p (reviewed inHuisman & Brunner, 2011; Martin, 2009) Tea1p localizes in large clusters at the poles of the cell and recruits the machinery for cell growth (reviewed inChang & Martin, 2009) Pom1p forms a concentration gradient from the poles to the middle of the cell acting as a ruler for cell length (Martin, 2009; Moseley et al., 2009; Moseley & Nurse, 2010; Padte, Martin, Howard, & Chang, 2006) As the cell grows, pom1p concentration at the medial site decreases, which eventually triggers mitosis (Ba¨hler & Pringle, 1998; Padte et al., 2006)

Microtubules have been shown to play a major role in the establishment and maintenance of these polarity patterns During interphase, they nucleate from the spindle pole body at the nucleus and form bipolar bundles that extend towards the poles of the cell (Janson et al., 2007; Sawin & Tran, 2006) Complexes of tea1p

2 CHAPTER 1 Microtubule-based cell polarity in fission yeast

and the SH3 domain-containing protein tea4p are delivered to the poles by

microtu-bule plus ends (Behrens & Nurse, 2002; Feierbach, Verde, & Chang, 2004) and

teth-ered at the cell cortex via the membrane-associated prenylated anchor protein

mod5p (Snaith, Samejima, & Sawin 2005) and the ERM (ezrin-radixin-moesin)

family protein tea3p (Dogson et al., 2013) With the help of the phosphatase

dis2p, tea1petea4p induce membrane association of pom1p at the cell poles by

recruiting it in its unphosphorylated state (Hachet et al., 2011) Then, clusters of

pom1p diffuse at the membrane independently from tea1petea4p (Dogson et al.,

2013), auto-phosphorylate, dissociate, and return to the cytoplasmic pull in its

phos-phorylated state (Saunders et al., 2012) Microtubules can also deliver complexes

away from the poles (followed by T-shaped cell growth) in physically bent cells

(Minc, Bratman, Basu, & Chang, 2009) or mutants with shorter microtubules

(Verde, Mata, & Nurse, 1995) Therefore, microtubule organization is essential

for a proper formation of polarity patterns

According to current thinking, in addition to a properly organized microtubule

cytoskeleton, a large number of molecular interactions (providing e.g., affinity for

the membrane, affinity for microtubules, regulatory control through

phosphoryla-tion, etc.) are required for proper localization of polarity factors Several questions

remain however unanswered, and it is difficult to establish which molecular

interac-tions are sufficient for proper cell-end localization in in vivo experiments To

com-plement the large body of in vivo work, we therefore aim to engineer a simplified in

vitro system and use it to find the minimal components needed to establish and

main-tain a polarized protein pattern at the membrane

We hypothesize that a minimal in vitro system capable of mimicking the yeast

polarity machinery should have the following ingredients:

1 Dynamic microtubule organization in elongated confinement with rigid

In this chapter, we describe two different experimental setups that fulfill some or all

of the requirements mentioned above Results concerning the ability to establish

po-larity patterns in these minimal systems will be published elsewhere

Dynamic microtubules grown from slow hydrolysable GTP analogue GMPCPP

(Guanosine-5’-[(a,b)-methyleno]triphosphate) “seeds” are confined to two types of

elongated geometries that mimic the fission yeast cell shape (Figure 1, top): glass wells

(Method 1,Figure 1, middle) and water-in-oil emulsion droplets (Method 2,Figure 1,

bottom) The microtubules self-organize in both types of confinement along the

longest direction Microtubuleþend tracking proteins (þtips) with a fused His-tag

(six histidines) are inserted in the system They can both tip track microtubules andreversibly bind to Ni(II)-NTA moieties bound to the confinement boundaries.

In method 1, the glass wells are rectangular with different aspect ratios Theirwalls are functionalized with tris-Ni(II)-NTA moieties (Kdz 10 nM) (Figure 1,middle) (Lata & Piehler, 2005; Zhen et al., 2006) with the same method as described

inTaberner et al (2014) Tris-Ni(II)-NTA binds His-tagged proteins cally with avidity (multivalent binding of three Ni(II) with six histidines) muchhigher than mono-Ni(II)-NTA affinity, whose complete binding of a His-tagged pro-tein depends strongly on the density of mono Ni(II)-NTA at the surface Selectivefunctionalization of the walls and passivation of the bottom surface is achieved byfirst coating all surfaces with PLL-PEG-Tris-Ni(II)-NTA (Bhagawati et al., 2013),photo-cleaving it from the bottom surface with deep UV irradiation (Azioune,Storch, Bornens, The´ry, & Piel, 2009), and subsequent passivation of the bottom sur-face with PLL-PEG In the previously reported method, the walls are sheltered fromdeep UV irradiation by and embed chromium mask with an overhang on top of thewells Since chromium is highly reflective for visible light, fluorophores under the

overhang appear up to three times brighter (data not shown) due to reflections of the

excitation and emission light Therefore, it is not possible with this method to

compare the fluorescence intensity between aþtip comet away from the wall and

close to it Here, we replace the chromium overhang by a 85 nm titanium oxide

(TiO2) layer, which absorbs most of the deep UV light, while still being transmissive

in the 450e600 nm wavelength range used in fluorescent experiments As shown in

Taberner et al (2014), dynamic microtubules can deliver and tether

His-tagged-tip1p (fission yeast’s CLIP170p) protein clusters to the coated walls This system

can be used to study the emergence of a polarized cortical pattern via delivery by

self-organized microtubules in the absence of diffusion at the boundaries

In method 2, His-tagged fused þtips interact with Ni(II)-NTA lipids freely

diffusing at the boundary The confinement in droplets provides a closed system

compared to method 1, where proteins could diffuse in and out of the glass wells

This method combines recent achievements in our lab to polymerize microtubules

inside droplets with functionalized lipids at the watereoil interface (Laan, Roth,

& Dogterom, 2012; Roth, Laan & Dogterom, 2014) with a method fromBoukellal,

Selimovic, Jia, Cristobal & Fraden (2008)to produce and store droplets with a

pre-determined elongated shape With this method, visualization of single dynamic

mi-crotubules is difficult, but it is ideal to test the global emergence and maintenance of

polarity patterns

For tip tracking proteins, we use either the yeast analogue of EB1p, mal3p with a

6His-tag, or the kinesin-cargo system tea2petip1p with a 6His-tag on tip1p (Busch,

Hayles, Nurse & Brunner, 2004; Bieling et al., 2007)

We first explain in detail the fabrication of the glass wells and the yeast-shaped

water-in-oil emulsion droplets, followed by two example assays: (1) protein binding

to the glass walls or functionalized lipids, and (2) dynamic microtubule growth and

self-organization in both types of confinement Here, only proof-of-principle results

are shown

2.1.1 Microfabrication

Special equipment:

All the microfabrication steps except evaporation were performed inside a clean

room (class ISO 6)

Homemade electron-beam evaporator with the Pfeiffer RVC 300 gas controllerOxford Plasmalab 80 plus (Oxford instruments, UK)

Snijstaal Keramiekoven (Snijstaal B.V, The Netherlands)

Surface profiler, Alpha step 500 (KLA-Tencor Corporation, USA)

2510 Ultrasonic Cleaner (Branson, USA)

MJB3 mask aligner for UV exposure (Su¨ss MicroTec)

Binary chromium/soda lime mask (Delta Mask, The Netherlands)

FEI Helios Nanolab 600 DualBeam (FEI Company, USA)

Leica EM ACE600 sputter coater (Leica Microsystem, Germany)

Jasco V-530 UV/VIS Spectrophotometer (Jasco Analytical Instrument, Japan)Reagents:

Shipley MicropositÒS1813 G2 positive UV-resist (Microresist, Germany)MicropositÒMFÒ-319 developer (Microresist)

Fresh made base piranha (ddH2O:NH4OH:H2O2in 5:1:1)

2.1.2 Surface functionalization

Special equipment:

Compact UV-ozone cleaner (Uvotech, USA)

Binary chromium/quartz mask (Delta Mask)

Reagents:

1 mM ethylene glycol tetraacetic acid (EGTA), pH 6.8) at the stated stock tration and stored at80C unless stated otherwise.

concen-k-casein from bovine milk (C0406 SigmaeAldrich); 20 mg/mL solution(k-casein)

two-step coupling process, as described inBhagawati et al., 2013

Nickel(II) sulfate (656895-10G, SigmaeAldrich), 10 mM NISO4solution inMRB80 pH 7.5 stored at room temperature

2.1.3 Assay

Special equipment:

AirfugeÒAir-driven ultracentrifuge (Beckman Coulter, USA)

IX81F-ZDC2 microscope (Olympus, Japan) with a spinning disk confocal headCSU-X1 (Yokogawa, Japan) 60X and 100X oil immersion objectives and EmCCDcamera iXon3 (Andor,UK) Excitation lasers 488 and 561 nm (Andor, UK)Glass slides (Menzel Gla¨sser, Germany)

TesaÒdouble-sided tape (15 mm wide)

6 CHAPTER 1 Microtubule-based cell polarity in fission yeast

Diamond glass cutter

Valap (vaseline, lanolin, paraffin wax melted at equal concentrations)

Reagents:

All reagents were dissolved in MRB80 buffer at the stated stock concentration

and stored at80C unless stated otherwise.

Microtubule polymerization:

Tubulin from bovine brain (TL238 Cytoskeleton, Inc., USA); 100 mM

Fluorescent HiLyte 488 tubulin from porcine brain (TL488M Cytoskeleton, Inc.);

50 mM rhodamine tubulin from porcine brain (TL590M Cytoskeleton, Inc.)

Guanosine 50-triphosphate sodium salt hydrate (G8877 SigmaeAldrich);

50 mM (GTP) Guanosine-50-[(a,b)-methyleno]triphosphate, sodium salt

(NU-405 Jena BioScience); 10 mM (GMPCPP)

Glucose oxidase from Aspergillus niger (G2133 SigmaeAldrich);

20 mg/ml dissolved in 200 mM DL-Dithiothreitol (646563 SigmaeAldrich)

with 10 mg/mL catalase from bovine liver (C9322 SigmaeAldrich) (glucose

oxidase)

GMPCPP-stabilized microtubule seeds:

polymerization cycle with GMPCPP in MRB80 (this cycle is done to remove

residual GTP) For the first polymerization step, a tubulin mix of fluorescently

labeled (HiLyte 488) and nonlabeled tubulin at a ratio of 12:88 (total of 20 mM

The mix is then centrifuged 5 min at 30 psi The microtubule pellet is resuspended

in MRB80 buffer, at 80% of the initial volume and left on ice for 20 min At this step,

the microtubules are depolymerized Then, 1 mM GMPCPP is added to the mix and

placed at 37C for 30 min The microtubules polymerize again The seeds are

centrifuged 5 min at 30 psi, resuspended at room temperature in 400% of the initial

volume of MRB80 supplemented with 10% glycerol Finally, the seeds are snap

frozen in liquid nitrogen and stored at80C Thawed seeds can be kept at room

temperature for a week

þtips:

Kinesin Tea2p was expressed in Escherichia coli and purified as inBieling et al

(2007)

The following proteins were purified as inMaurer, Bieling, Cope, Hoenger, and

Surrey (2011)and the His-tags were removed if necessary with AcTEVÔ protease

(12575-015 Invitrogen) at 4C according to the provider’s protocol.

6His-TEV-Mal3-mCherry was obtained fromTaberner et al (2014)

Unlabeled 6His-TEV-mal3 was obtained as inBieling et al (2007)

6His-TEV-eGFP-Tip1 was obtained as inTaberner et al (2014)

2.2 MATERIALS FOR METHOD 2: ELONGATED WATER IN OIL

Surface profiler, Alpha step 500 (KLA-Tencor Corporation, USA)

Binary chromium/soda lime mask (Delta Mask, The Netherlands) for the dropletchannels

Photomask on film substrate (Selba S.A, Switzerland) for the buffer reservoirReagents:

Different SU-8 Permanent Epoxy Negative Photoresist (MicroChem, USA)depending on the desired height: 2005 for 5 mm, 2010 for 10 and 15 mm, and

3025 for 40 mm Even though the 2000 series were used, it is advisable to usethe 3000 ones since they produce better adhesion to the silicon wafer

Developer mr-Dev 600 (Micro Resist Technology GmbH, Germany)

PDMS chip fabrication:

Special equipment:

Spin coater SPIN150i Table-Top (SPS-Europe, The Netherlands)

Eppendorf Centrifuge 5702 (VWR International B.V., The Netherlands)Vacuum desiccator

Corona treater, model BD-20AC (Electro-Technic Products INC, USA) Oven.Razor blades

Harris Uni-CoreTMcutting tip of 2 Ø mm (Ted Pella, Inc., USA)

Reagents:

Ben-elux, The Netherlands)

Microfluidic setup and droplet formation:

Special equipment:

Fluiwell-4C device (Fluigent)

Micrewlock tubes: T341-2TTP of 2 and 0.5 mL (Simport, Canada)

Hamilton syringe (Hamilton-Bonaduz, Schweiz) 1 mL Syringe

23G Tygon tubing (Elvesys, France)

8 CHAPTER 1 Microtubule-based cell polarity in fission yeast

Oil phase:

Mineral oil (M5904, SigmaeAldrich)

SpanÒ80 (SigmaeAldrich)

1,2-dioleoyl-sn-glycero-3-phospho-L-serine (sodium salt) (Avanti Polar Lipids

Alabaster, AL, USA) (DOPS)

1,2-dioleoyl-sn-glycero-3-[(N-(5-amino-1-carboxypentyl)iminodiacetic

acid)suc-cinyl] (nickel salt) (Avanti Polar Lipids Alabaster, AL, USA) (Ni(II)-NTA-DGS)

Water phase (protein mix):

The used protein mix contains the same reagents as the ones for the glass wells

with the addition of the following components:

Dextran, Alexa FluorÒ647, 10,000 MW, Anionic, Fixable (Life technologies,

As explained above, selective functionalization of the walls requires an overhang on

the wells that absorbs deep UV light, but does not reflect in the range of wavelengths

used for fluorescent experiments Moreover, to make microfabrication possible, the

material used for the overhang should etch at a lower rate than the glass during the

wet etch step We found that a good balance can be obtained with a 85 nm layer of

TiO2(see transmission spectra inFigure 2(B)), which is conventionally used as

anti-reflection coating (Richards, 2003) To avoid enhanced interface-etching between

the titanium oxide and the glass, a thin chromium (Cr) barrier layer of 1.5 nm is

added in between

Figure 2(A)shows an overview of the methodology used to produce TiO2

shel-tered microwells First, a 1.5 nm layer of chromium followed by 85 nm layer of

amorphous titanium oxide are deposited on coverslips by electron-beam

evapora-tion The titanium oxide is then postannealed to form TiO2crystals, able to sustain

later etching of the glass The well patterns are applied by UV-photolithography

Subsequent etching of the TiO2 and Cr is achieved by anisotropic reactive ion

etching followed by isotropic wet etching of the glass

3.1.2 Deposition of chromium and titanium oxide by electron-beam

evaporation

To remove dust particles and organic residues from the production process and

ensure a good cohesion between the deposited layers, the following steps need to

be taken: clean the glass slides with base piranha (15 min at 75C), flush them

with demineralized water, rinse them in isopropanol, and finally blow-dry them withdry nitrogen Without further treatment, transfer the slides to the electron-beamevaporator To ensure an adequate vacuum (<5$107mbar) before evaporation,

the system is baked-out for 14 h at 100C.

Deposit a 1.5 nm chromium layer at a deposition rate of 0.05 nm/s Furtherdeposit a 85 nm layer of titanium oxide by evaporating Ti3O5at 0.05 nm/s in anelevated O2partial pressure (2$104mbar) Slides with titanium oxide layers that

were deposited at lower O2partial pressures showed lower transmission by visualinspection However, they all were, after annealing, resistant to KOH during theglass etching

3.1.3 Postannealing of titanium oxide

After deposition (within 6 h), anneal the slides at 400C for 2 h in the oven inambient air (with a ramp up of 1 h) The titanium oxide layer crystallizes and formsanatase crystals (examined with EBSD, data not shown) that are visible under a mi-croscope (Figure 2(C)) The grain size (z5 mm) is influenced by the evaporationconditions (lower O2partial pressure results in a smaller grain size, data not shown).Also, additional time between evaporation and annealing reduces the grain size (datanot shown) Layers deposited at O2 partial pressure lower than 2$104mbar,

increased in thickness up to 20% during annealing We think that the shortage of

FIGURE 2

(A) Scheme of the microfabrication steps of TiO2sheltered glass wells (B) Transmissionspectra of 85 nm TiO2on a glass substrate: Calculated (dashed line, with the programThinfilm,www.thinfilm.hansteen.net, which solves the Maxwell’s equations for multiple thinfilms using the transfer-matrix method) and measured with a spectrophotometer (continuousline) (C) Microscope image of glass slides where both the wells (appearing black) and TiO2crystals are visible (D) SEM image of a cross section of a well Note that the structurescontain an extra platinum layer deposited on top to ensure a good cross section cut Thislayer is much thinner under the overhang, since this area is less accessible

10 CHAPTER 1 Microtubule-based cell polarity in fission yeast

O2is supplemented during annealing, since higher partial pressure samples did not

show a thickness increase The transmission difference between slides caused by

different O2 conditions disappeared after annealing, also suggesting that O2 is

supplemented

3.1.4 UV-lithography with S1813

To remove dust particles and ensure a clean surface for the S1813 resist, clean the

cover slips with base piranha (15 min at 75C), flush with demineralized water,

rinse in isopropanol, and finally blow-dry with dry nitrogen Spin coat S1813 at

2000 rpm for 45 s and then soft-bake 60 s at 115C to form a layer of around 1 mm.

Expose it to a dose of 125 mJ/cm2365 nm UV light, through a chromium/soda lime

mask, with a Su¨ss MJB3 mask aligner Develop the exposed resist by submerging the

cover slips in MF319 for 40 s and rinse in demineralized water The pattern of the

mask now becomes visible in the S1813 resist layer (Figure 2(C))

3.1.5 Reactive ion etching

The resist pattern is transferred into the TiO2 by reactive ion etching with SF6

is etched by CHF3(20 sccm) and O2(20 sccm) gas (at 30 mTorr, 20C, forward

power 200 W, ICP 100 W) for 30 s If this final step is performed too long, the

exposed glass gets roughened due to inhomogeneous etch rates caused by impurities

in the cover slips

The resist layer is removed by sonication for 5 min in acetone and iso-propanol

Complete etching of the chromium and titanium oxide can be checked with a surface

profiler

3.1.6 Wet etching with KOH

Submerge the cover slips into a 40% w/w KOH solution at 75C This step will

ho-mogeneously and isotropically etch the glass with a rate ofz300 nm/h.Figure 2(D)

shows a cross section made with the FEI Helios of the glass well and the titanium

oxide overhang The image was made by first sputtering a 20 nm gold layer (with

the Leica sputter coater) for electric conduction Az300 nm platinum layer was

then deposited by electron beam-induced deposition (with the FEI Helios) to protect

the top layer Finally with a focused gallium ion beam a cross section was milled

The titanium oxide overhang has become thinner during the ion reactive etching

pro-cess This will lead to undesired higher reflection of visible light (up to 40% for a

42 nm layer as calculated from Thinfilm) The ion reactive etching process was

not optimized to create an anisotropic TiO2etch, a higher forward power and lower

pressure could improve this

3.1.7 Surface functionalization

After microfabrication, each glass slide can be cut (with a diamond glass cutter) in

three long pieces of 8 24 mm, which will serve for three different experiments

Prepare the surface for PLL-PEG coating by irradiating with deep UV (185 and

254 nm) for 30 min with the Compact UV-Ozone cleaner Pipet 10 mL of 0.2 mg/mLPLL- PEG-tris-NTA onto the microfabricated slide and place it inside a plastic seal-able bag so that the droplet wets all the surface of the slide while preventing evap-oration Sonicate it a few seconds to remove possible air bubbles at the wells and let

it at room temperature for 30 min Blow-dry it with dry nitrogen and irradiate 2e

3 min with deep UV in the Compact UV-Ozone cleaner Note that a thin layer ofTiO2is not completely opaque to deep UV, therefore too long exposure will removecompletely the coating at the walls

Form a flow cell of around 15 mL by sticking two parallel stripes of TesaÒdoublesticky tape (separated around 5 mm) on a glass slide previously cleaned with 1%Hellmanex, ethanol 70%, and demineralized water Put the prepared glass slide ontop, forming a channel with the microstructures on the ceiling Different liquid so-lutions can be flown by capillarity by placing absorbent paper on the other side of the

Assay on His-tagged protein binding to the tris-Ni(II)-NTA coated walls:Prepare on ice 20 mL of the following mix in MRB80: 500 nM 6His-mal3-mCherry, 50 mM KCl, 0.3 mg/ml k-casein, 0.4 mg/mL glucose oxidase, and

50 mM glucose Centrifuge 8 min at 30 psi and insert the supernatant in the flowcell.Figure 3(B)shows a confocal image of the sample, during protein mix incuba-tion, and after flushing it away During incubation, 6His-mal3-mCherry appears both

in the bulk and at the walls (Figure 3(B), top) After flushing away the protein mix,6His-mal3-mCherry remains bound to the walls, at least for some minutes(Figure 3(B), bottom)

Assay on microtubule organization inside elongated wells:

Take 2 mL of thawed microtubule seeds and add 8 mL of MRB80 at room perature Centrifuge it for 5 min at 30 psi and reserve the pelleted seeds

tem-Prepare in ice 20 mL of the following mix in MRB80: 16 mM unlabeled tubulin,0.875 mM rhodamine tubulin, 150 nM 6His-eGFP-tip1, 10 nM tea2, 100 nM mal3,

1 mM GTP, 2 mM ATP, 0.3 mg/mL k-casein, 50 mM KCl, 0.4 mg/mL glucoseoxidase, 50 mM glucose, and 0.1% methyl cellulose Centrifuge 8 min at 30 psi tosediment possible protein aggregates and use it to resuspend the pelleted seeds Incu-bate it in the flow cell and seal the sides with melted valap Observe the sample at

25e26C to trigger microtubule polymerization.

12 CHAPTER 1 Microtubule-based cell polarity in fission yeast

Figure 3(C) shows an example of a confocal fluorescent image Microtubules

grown from seeds diffuse close to the surfaces, both inside the wells and on top

of the TiO2 layer (Figure 3(C), bottom) However, they tend to align along the

longest axis only inside the wells 6His-eGFP-tip1 appears both in the reservoir

and partially coats the walls (Figure 3(C), top) Due to fast microtubule fluctuations,

it is difficult to observe tip tracking (images from both channels are taken

consecu-tively) The amount of protein bound to the walls is now much lower than in the

pre-vious test because His-tag binding to tris-Ni(II)-NTA is concentration-dependent

This avidity can be further reduced with addition of imidazole in the mix Imidazole

competes with the His-tag for Ni-binding sites This system allows to test the

FIGURE 3

(A) Schematic representation of surface functionalization of the glass wells (B) Assay to

test 6His-mal3-mCherry binding to tris-Ni(II)-NTA functionalized walls Top: During

incubation, 6His-mal3-mCherry appears enriched at the walls Bottom: After flushing out,

6His-mal3-mCherry remains at the walls, at least for a few minutes Scale bar: 10mm (C)

Dynamic microtubules grown in the tris-Ni(II)-NTA functionalized glass wells in the presence

of 150 nM of 6His-eGFP-tip1p Top: Fluorescent confocal image of 6His-eGFP-tip1p

Bottom: Fluorescent confocal image of microtubules (D) Delivery events of 6His-eGFP-tip1p

to the tris-Ni(II)-NTA functionalized walls by microtubules Confocal images at different

times: before, during and after contact of the microtubule end with the wall 6His-eGFP-tip1p

remains at the wall even after microtubule catastrophe Top: fluorescent tubulin Bottom:

fluorescent 6His-eGFP-tip1p

emergence of polarity driven by self-organized microtubules delivering tip trackersthat reversibly bind to the walls (results and detailed discussion will be publishedelsewhere) Examples of 6His-eGFP-tip1 delivery to tris-Ni(II)-NTA coated walls

Therefore, the microfluidic chip we use consists of two PDMS layers(Figure 4(A)): A thin layer (z30 mm high) on top of a PDMS coated glass slidewhere droplets are formed and stored, and an additional thick layer (z0.5 cm)with a buffer channel reservoir on top Buffer can be inserted in the upper channel

FIGURE 4

(A) Schematic representation of the double-layered microfluidic chip for creation ofelongated water-in-oil emulsion droplets (B) Schematic representation of the reservoir anddroplet channel design Zoom shows one module for creation and storage of an elongateddroplet

14 CHAPTER 1 Microtubule-based cell polarity in fission yeast

reservoir with a syringe (no flow needed), while the flow in the lower channels is

controlled with a MFCSTM-EZ system

The droplet formation and storage design is an adaptation of the one from

Boukellal et al (2008) It consists of several modules that both create and store

drop-lets placed in serial with an inlet and outlet reservoirs (Figure 4(B)) The inlet

reser-voir contains a filter to retain possible PDMS broken pieces produced when

connecting the tubes The resistivity of the filter should preferably not be higher

than the sum of the droplet modules one

Each droplet module has a storage, a bypass, and a restriction channels of widths

“a,” “b,” and “c,” respectively (Figure 4(B), zoom) As detailed inBoukellal et al

(2008), the device is first filled with oil A plug of water phase is then introduced,

followed again by the oil phase The widths are such that the water phase is first

pushed to the storage, blocked at the oil filled restriction, and once the storage is

filled, proceeds through the bypass (a> b > c) This is because, when the water

plug arrives at the intersection with the bypass, it should deform to enter to the

nar-rower bypass Therefore, it proceeds towards the storage In this case the bypass acts

as a capillary valve (Eijkel & van den Berg, 2006) However, once the water phase

reaches the even narrower restriction, the water flows though the bypass where less

deformation is needed (b> c) (Figure 6(B), left) Once oil is flown again, it pushes

the water phase, both towards the restriction and the bypass, leaving a droplet in the

storage (Figure 6(B), middle and right andFigure 6(A))

We want to keep close to a fission yeast size:z3 mm diameter Limited by the

1.5 mm lithography resolution We decided to use height¼ a ¼ 5 mm, b ¼ 3 mm, and

c¼ 1.5 mm Wider designs were also made for testing purposes

FIGURE 5

Microfluidic chip fabrication steps

3.2.2 Masks for microfluidic chip fabrication

Since the reservoir design does not need micrometer precision, it can be made with alow cost flexible photomask by Selba S.A However, the formation-storage chipshould be made in a soda-lime chromium mask with micrometer precision (Deltamask) These masks are then used to transfer the pattern to an SU-8-coated siliconwafer by photolithography, which will serve as negative mould for PDMS

3.2.3 SU-8 negative mould

Photolithography of SU-8 is a standard process The steps are: spin coat the SU-8,bake at 95C, expose to UV light through the photomask, bake at 95C, develop

different SU-8 suitable for each thickness We produced three masters for the tion-storage with 5 (with SU-8 2005, spin coated at 3800 rpms for 30 s), 10, and

forma-15 mm (with SU-8 2010, spin coated at 4000 and 1700 rpms, respectively for

32 s); and one master for the reservoir of 40 mm (SU-8 3025, spin coated at

1800 rpms for 45 s) However, since SU-8 ages, the spinning speeds given are ative and might have to be adjusted The samples were exposed to 195, 185, 195, and

indic-250 mJ/cm2365 nm UV light for the respective 5, 10, 15, and 40 mm thicknesses

3.2.4 PDMS chip ( Figure 5 )

All parts of the PDMS chip are made with 10 weights PDMS prepolymer RTV615and 1 weight of the curing agent The two components are mixed in a plastic weight-ing boat, placed in a 50 mL falcon tube (BD Falcon), and centrifuged at 300 rcf for

5 min to remove air bubbles

FIGURE 6

(A) Artist impression of elongated water-in-oil emulsion droplets in the microfluidic chip (B)Bright field view of droplet formation in 10 15 mm storage with 10 mm high channels Thewater phase appears darker than the oil (C) Fluorescent microscopy images on the equator

of droplets formed in 10 15  10 mm storage with mal3p-mCherry (left) and mCherry (right) (D) Fluorescent microscopy images of tubulin on the equator of a dropletformed in 5 11  5 mm storage Microtubules can be seen aligned along the longestdirection

6h-mal3-16 CHAPTER 1 Microtubule-based cell polarity in fission yeast

For the reservoir channel, make a pot-like structure with aluminum foil around

the silicon master and pour on it around 40 mL of PDMS Put it in a vacuum

dessi-cator until all the air bubbles are gone (around 1.5 h) Then bake it at 100C for 1 h

to cure it Remove the aluminum foil with the help of a razor blade and peel the

PDMS off by gently pulling it from one side Cut with a razor the nonflat edges

of the piece and punch 0.5 mm holes at the extremes of each reservoir channel

with a Harris Unicore cutting tip To avoid dust in the channels, one can put some

possible dust from the surfaces by putting some tape on it and removing it

For the droplet channel, a thin layer of PDMS (around 30 mm) is spin coated at

1000 rpm for 20 s with an initial step of 800 rpm for 5 s After spinning, bake it at

100C for at least 15 min.

Bind the reservoir channel on top of the droplet one by applying for a few

sec-onds O3plasma on both pieces with the corona treater and quickly placing the

reser-voir channel on top of the other one, properly aligned The PDMS layers should stick

without need of extra pressure Bake it at 100C for at least 1 h Cut with a razor

blade the extra PDMS part of the droplet channel that has no reservoir piece on

top and peel it off Peel off the rest of the chip with both layers by carefully pulling

it from one side Protect the side with the channels with tape and cut each modular

chip with the razor Punch the inlet and outlet holes as before

For the cover-slip support, spin PDMS at 4000 rpm for 30 s with a previous step

at 200 rpms for 100 s Bake it for at least 15 min at 100C.

Finish the chip by binding the two-layer formed before with the PDMS coated

cover-slip This is done by applying for a few seconds O3plasma on both pieces

with the corona treater and quickly placing the two-layer chip on top of the

cover-slip Bake it at 100C overnight to obtain very hydrophobic PDMS channels.

3.2.5 Microfluidic setup and droplet formation

Buffer with the same osmotic pressure as the droplets is inserted inside the reservoir

channels with a syringe and 23G Tygon tubes connected to the channels via a cut

piece of a 1.2 Ø mm needle

The flow for the droplet channels is controlled by a pressure controller MFCSTM

-EZ system connected to a Fluiwell accessory with Micrewlock tubes PEEK 125

Ø mm tubes directly connect the Micrewlock tubes in the Fluiwell to the inlets of

the microfluidic chip

Both the droplet formation and observation of the fluorescent microtubule assay

are done with a spinning disk confocal microscope

Droplets are formed by flowing first a mix of oil with lipids, followed by the

water phase containing the protein mix and, again, the oil phase Since there is

only one inlet, each phase is added by exchanging the PEEK tubes For 5 mm

high channels, the initial oil phase flows easily with 1e2 bars The protein mix needs

up to 4 bars to reach the filter of the inlet We found important to push the water

phase till the filter of the inlet before exchanging the tube back to the oil one Indeed,

we experienced the water phase to be pushed back to the PEEK tube when releasingthe pressure of the water phase prior to tube exchange.

A low pressure of a few hundred millibars can then be applied to the second oilphase Pressures have to be tested for each design and chip since the resistance ofthe channels might vary from one chip to another Once the droplets are formed,lower the flow of oil to keep the droplets close to the restriction without deformingthem

3.2.6 Lipids composition

For the proper functioning of the microfluidic chip, it is essential for the PDMS to bevery hydrophobic The water phase must not wet the surfaces Protein sticking toPDMS will make water wet the surface The use of surfactants such as SpanÒ80helps preventing protein interaction with the PDMS and the oil because it isabsorbed at the watereoil interface much faster than lipids Indeed surfactant ab-sorption is diffusion limited, while lipid absorption appears to be more complexand can take up to several minutes for full coverage (Pautot, Frisken, & Weitz,

2003) The amount of SpanÒ80 needed depends on the lipid composition and theuse of tweenÒ20 in the water phase An excessive amount of surfactant mightlead to very stable droplets and the undesired accumulation of several small dropletsnot coalescing in the storage channel In our hands, for 0.03% TweenÒ20 in the wa-ter phase, a molar ratio of SpanÒ80, DOPS and Ni(II)-NTA-DGS of 90:4:1 with0.5 mg/mL of lipids in oil (without counting SpanÒ80) was used for His-tag proteinbinding experiments and a 96:3.6:0.4 for the microtubule assay Prepare the oilphase by first mixing DOPS and Ni(II)-NTA-DGS in chloroform in a glass vial(use glass syringes) Dry it in nitrogen flow Add the SpanÒ80 and mineral oil.Mix by pipetting several times Sonicate for 30 min to dissolve the lipids in oil

3.2.7 Assays

Here we first present the control experiment testing the specific binding of tagged proteins to Ni(II)-NTA-DGS lipids We then describe a protocol to growdynamic microtubules from seeds

His-Assay on His-tagged protein specific binding to Ni(II)-NTA lipids:

The binding assay mix contains 500 nM 6His-mCherry or

mL glucose oxidase, 50 mM glucose, in MRB80 with 50 mM KCl and 0.7 mMdextran 467 Centrifuge 8 min at 30 psi to sediment possible protein aggregatesand transfer the supernatant to a 0.5 mL Micrewlock tube to prepare the droplets

as explained above

Figure 6(C)shows an example of a droplet with mal3p-mCherry (left) No cific binding of mal3p-mCherry to the water-oil interface is observed (Figure 6(C),left) We next show the results obtained when introducing 6His-mal3-mCherry in themix (middle) Proteins are now visible both inside the droplet and at the lipid-covered interface, confirming the specific interaction of 6h-mal3-mCherry withNi(II)-NTA lipids

unspe-18 CHAPTER 1 Microtubule-based cell polarity in fission yeast

Assay on microtubule organization in elongated droplets:

Mix on ice the following 10 mL MRB80 solution reserving 1 mL (for seeds that

will be added later): 26 mM unlabeled tubulin, 1.75 mM HiLyte 488 tubulin, 5 mM

GTP, 0.03% tween-100x, 5 mg/mL BSA, 5 mg/mL kappa-casein, 50mM KCl,

0.4 mg/mL glucose oxidase, 50 mM glucose, and 0.6 mM Dextran 647 Centrifuge

8 min at 30 psi to sediment possible protein aggregates Transfer the supernatant

to a 0.5 mL Micrewlock tube and add 1 mL of seeds Prepare the droplets as

explained above Once the droplets are formed, set the temperature controlled

cham-ber of the microscope to 26e30C to trigger polymerization.Figure 6(D)shows an

example of a droplet with dynamic microtubules inside They align in the longest

direction

DISCUSSION AND PERSPECTIVES

We described two in vitro setups with necessary features to reconstitute

microtubule-based establishment of cell polarity: (1) Microtubule organization in elongated

confinement, (2) tip tracking proteins with reversible tunable affinity for the

confine-ment boundaries Moreover, protein diffusion at the confineconfine-ment boundary is

allowed in the water-in-oil emulsion droplet method

The semi 2D glass chambers are an alternative to the previously described

method in Taberner et al (2014) Here, illumination is homogeneous along the

whole microwell Therefore, observations of microtubuleewall interactions are

free of illumination artifacts Moreover, since TiO2is not completely absorbent to

deep UV, the exposure time can serve to decrease the percentage of

tris-Ni(II)-NTA at the walls This provides a new way to tune the affinity of His-tagged

pro-teins, alternative to imidazole

This setup can serve to study microtubule-based delivery and tethering of

pro-teins to boundaries, microtubule organization in elongated cavities, and their

com-bination to assess the establishment of polarity Polarity maintenance, here,

depends on frequency of microtubule contacts to the poles, efficiency of protein

transfer from the microtubule tip to the wall, and rate of protein unbinding from

the walls This system will help understand microtubule organization in fission

yeast, as well as the polarity pattern of tea1

The elongated droplet system provides a new method to study microtubule

organization as well as polarity Here, polarity can be assessed under

condi-tions that allow for protein diffusion at the membrane This setup offers a

simplified frame to study both the tea1 and the pom1p polarity patterns

with a reliable control on droplet size and shape Moreover, the continuous

flow of oil in principle opens the possibility to introduce (oil-soluble) drugs

to the system to, for example, trigger microtubule polymerization Note that

the microfluidic chip can also be designed to produce elongated droplet in

a vertical position, such that the poles of the droplets can be imaged at higher

resolution

We thank Lutz Langguth for advice on thin layer optics Magdalena Preciado-Lopez, VandaSunderlı´kova´, Cristina Manatschal, and Michel Steinmetz for help with the purification ofproteins and for discussions We thank Pierre Recouvreux and Roland Dries for help withcloning techniques and for discussions This work is part of the research program of the

“Stichting voor Fundamenteel Onderzoek de Materie (FOM)” which is financially supported

by the “Nederlandse organisatie voor Wetenschappelijk Onderzoek (NWO).”

Bhagawati, M., You, C., & Piehler, J (2013) Quantitative real-time imaging of protein interactions by LSPR detection with micropatterned gold nanoparticles Analyt-ical Chemistry, 85, 9564e9571

protein-Bieling, P., Laan, L., Schek, H., Munteanu, E L., Sandblad, L., Dogterom, M., et al (2007).Reconstitution of a microtubule plus-end tracking system in vitro Nature, 450, 1100e1105.Boukellal, H., Selimovic, S., Jia, Y., Cristobal, G., & Fraden, S (2008) Simple, robust storage

of drops and fluids in a microfluidic device Lab on a Chip, 9(2), 331e338

Busch, K E., Hayles, J., Nurse, P., & Brunner, D (2004) Tea2p kinesin is involved inspatial microtubule organization by transporting tip1p on microtubules DevelopmentalCell, 6, 831e843

Chang, F., & Martin, S G (2009) Shaping fission yeast with microtubules Cold SpringHarbor Perspectives in Biology, 1(1), a001347 http://dx.doi.org/10.1101/cshperspect.a001347

Dodgson, J., Chessel, A., Yamamoto, M., Vaggi, F., Cox, S., Rosten, E., et al (2013) Spatialsegregation of polarity factors into distinct cortical clusters is required for cell polaritycontrol Nature Communications, 4, 1834

Eijkel, J C T., & van den Berg, A (2006) Young 4ever - the use of capillarity for passive flowhandling in lab on a chip devices Lab on a Chip, 6, 1405e1408

Feierbach, B., Verde, F., & Chang, F (2004) Regulation of a formin complex by the tubule plus end protein tea1p Journal of Cell Biology, 165, 697e707

micro-Hachet, O., Berthelot-Grosjean, M., Kokkoris, K., Vincenzetti, V., Moosbrugger, J., &Martin, S G (2011) A phosphorylation cycle shapes gradients of the DYRK familykinase Pom1 at the plasma membrane Cell, 145, 1116e1128

Huisman, S M., & Brunner, D (2011) Cell polarity in fission yeast: a matter of confining,positioning, and switching growth zones Seminars in Cell & Developmental Biology,

22, 799e805

20 CHAPTER 1 Microtubule-based cell polarity in fission yeast

Janson, M E., Loughlin, R., Loı¨odice, I., Fu, C., Brunner, D., Ne´de´lec, F J., et al (2007).

Crosslinkers and motors organize dynamic microtubules to form stable bipolar arrays

in fission yeast Cell, 128, 357e368

Laan, L., Roth, S., & Dogterom, M (2012) End-on microtubule-dynein interactions and

pulling-based positioning of microtubule organizing centers Cell Cycle, 11(20),

3750e3757

Lata, S., & Piehler, J (2005) Stable and functional immobilization of histidine-tagged

proteins via multivalent chelator headgroups on a molecular poly(ethylene glycol)

brush Analytical Chemistry, 77, 1096e1105

Martin, S (2009) Microtubule-dependent cell morphogenesis in the fission yeast Trends in

Cell Biology, 19(9), 447e454

Maurer, S P., Bieling, P., Cope, P., Hoenger, A., & Surrey, T (2010) GTPgS microtubules

mimic the growing microtubule end structure recognized by end-binding proteins

(EBs) PNAS, 108(10), 3988e3993

Minc, N., Bratman, S V., Basu, R., & Chang, F (2009) Establishing new sites of polarization

by microtubules Current Biology, 19(2), 83e94

Mitchison, J M., & Nurse, P (1985) Growth in cell length in the fission yeast

Schizosacchar-omyces pombe Journal of Cell Science, 75, 357e376

Moseley, J B., Mayeux, A., Paoletti, A., & Nurse, P (2009) A spatial gradient coordinates

cell size and mitoticentry in fission yeast Nature, 459(7248), 857e860

Moseley, J B., & Nurse, P (2010) Cell Division Intersects with Cell Geometry Cell., 142(2),

189e193

Padte, N N., Martin, S G., Howard, M., & Chang, F (2006) The cell-end factor pom1p

inhibits mid1p in specification of the cell division plane in fission yeast Current Biology,

16, 2480e2487

Pautot, S., Frisken, B J., & Weitz, D A (2003) Production of unilamellar vesicles using an

inverted emulsion Langmuir, 19, 2870e2879

Richards, B S (2003) Single-material TiO2 double-layer antireflection coatings Solar

Energy Materials and Solar Cells, 79(3), 369e390

Roth, S., Laan, L., & Dogterom, M (2014) Reconstitution of cortical dynein function

Methods in Enzymology, 540, 205e230

Saunders, T E., Pan, K Z., Angel, A., Guan, Y., Shah, J V., Howard, M., et al (2012) Noise

reduction in the intracellular Pom1p gradient by a dynamic clustering mechanism

Devel-opmental Cell, 22, 558e572

Sawin, K E., & Tran, P T (2006) Cytoplasmic microtubule organization in fission yeast

Yeast, 23, 1001e1014

Shim, J., Cristobal, G., Link, D R., Thorsen, T., Jia, Y., Piattelli, K., et al (2007) Control and

measurement of the phase behavior of aqueous solutions using microfluidics Journal of

the American Chemical Society, 129(28), 8825e8835

Snaith, H A., Samejima, I., & Sawin, K A (2005) Multistep and multimode cortical

anchoring of tea1p at cell tips in fission yeast EMBO Journal, 24, 3690e3699

Taberner, N., Weber, G., You, C., Dries, R., Piehler, J., & Dogterom, M (2014)

Reconstituting functional microtubule-barrier interactions Methods in Cell Biology,

120, 69e90

Verde, F., Mata, J., & Nurse, P (1995) Fission yeast cell morphogenesis: identification of new

genes and analysis of their role during the cell cycle Journal of Cell Biology, 131(6),

1529e1538

Watson, J M., & Baron, M G (1995) Precise static and dynamic permeation measurementsusing a continuous-flow vacuum cell Journal of Membrane Science, 106, 259e268.Zhen, G., Falconnet, D., Kuennemann, E., Vo¨ro¨s, J., Spencer, N D., Textor, M., et al (2006).Nitrilotriacetic acid functionalized graft copolymers : a polymeric interface for selectiveand reversible binding of histidine-tagged proteins Advanced Functional Materials, 16,

243e251

22 CHAPTER 1 Microtubule-based cell polarity in fission yeast

Microtubules, MAPs, and

Kasimira T Stanhope, Jennifer L Ross 1

Molecular and Cellular Biology Graduate Program, Department of Physics,

University of Massachusetts Amherst, Amherst, MA, USA

1 Corresponding author: E-mail: rossj@physics.umass.edu

3.4 Analysis and Notes 34

4 Cell-like Patterns from Gliding Prebundled Microtubule Filaments 35

4.1 Reagents and Buffers 35

Methods in Cell Biology, Volume 128, ISSN 0091-679X, http://dx.doi.org/10.1016/bs.mcb.2015.02.003

© 2015 Elsevier Inc All rights reserved. 23

The cell is an inherently nonequilibrium environment where countless nanoscalemachines, called enzymes, use energy to perform work opposing the entropicmixing force of diffusion Such machines work in concert to push and pull biolog-ical molecules, macromolecules, and networks to where they need to be in timeand space No place is this energetic dance more obvious than in the cytoskeleton.Perhaps this is because the cytoskeleton organizes the cell’s interior, is used asthe highway system to transport goods and services, and obviously and dramati-cally rearranges during cell division, motility, and development Whateverthe reason, groups have been studying cytoskeletal organization in cells for de-cades (Kirschner & Schulze, 1986; Ne´de´lec, Surrey, & Karsenti, 2003; Dogterom

& Surrey, 2013)

In vitro reconstitution experiments of cytoskeleton organization using purifiedcomponents have been worked on for many years (Dogterom & Surrey, 2013).Several groups have worked on this problem, which is rich and deep due to theextensive number of cytoskeleton binding, cross-linking, and translocating proteins.Many of the published procedures involve difficult steps including nanofabrication(Ne´de´lec, Surrey, Maggs, & Leibler, 1997; Laan & Dogterom, 2010), use of extractswith many unknown factors (Cahu & Surrey, 2009; Pinot et al., 2009; Brugue´s,Nuzzo, Mazur, & Needleman, 2012), and high-end microscopy methods (Surrey,Nedelec, Leibler, & Karsenti, 2001; Brugue´s et al., 2012) The difficulty to perform-ing these experiments has limited the number of researchers working on the problemand made reproducing prior results difficult

In an effort to simplify the system to something reproducible and easy toperform, we decided to modify a simple filament-gliding assay In this chapter,

we describe our simple filament-gliding assay and then add modifications to theassay including adding cross-linking proteins and more filaments Finally, wedescribe a system that recapitulates cell-like microtubule organizations similar tothose found in mitosis using only three protein components

24 CHAPTER 2 Microtubules, MAPs, and motor patterns

1 METHODS

Here, we outline the experimental methods to create cell-like patterns in vitro based

on a simple microtubule-gliding assay powered by kinesin-1 motor proteins (Liu,

Tu¨zel, & Ross, 2011; Pringle et al., 2013) We systematically add more microtubules

or other types of associated proteins to probe how the patterns change in increased

complexity The microtubule-associated protein (MAP) we have used are the plant

antiparallel cross-linker, MAP65-1 This is the plant homolog of PRC-1

(mamma-lian) or Ase1 (yeast)

MICROTUBULES

The gliding assay with microtubules and kinesin-1 is a simple assay that

undergrad-uates can do in any lab No special chambers or glass treatment is needed Here, we

describe the flow chambers, reagents and buffers, and how the experiment is

performed

2.1.1 Materials

Coverslip (22
22
1.5 mm, No 1.5, Thermo Fisher Scientific)

Coverslip (22
30
1.5 mm, No 1.5, Thermo Fisher Scientific)

Glass slide (25
75
1 mm, No 1.5, Thermo Fisher Scientific)

Permanent double-sided clear plastic tape (3 M)

5-min Zpoxy epoxy (Pacer)

2.1.2 Chamber construction

1 Clean cover glass and coverslip with double distilled water and ethanol Dry with

a kimwipe and leave under a petri dish to protect from dust

2 Place double-sided tape 4e5 mm apart on the cover glass creating a horizontal

flow chamber (Figure 1(A))

3 Place 22
22 mm coverslip on top of the tape and press down to seal Press only

on the tape and not in the middle of the path because you can crack the cover

glass When pressing on the tape, it should become more transparent indicating

that the chamber is sealed (Figure 1(A) and (C))

2.1.3 Notes

The chamber is 0.1 mm deep because the double stick tape is 100mm thick The

cover glass is 22 mm long, setting the length of the flow path These two lengths

are fixed, so the volume of the chamber is totally determined by the width of the

flow path between the two pieces of double stick tape To make a 10mL chamber

volume, place the pieces of tape 4.5 mm apart

FIGURE 1

Simple filament-gliding assay chamber construction and use (A) Typical flow chamber withflow path parallel to long axis of slide The top shows a schematic where the tape is gray andthe flow path is denoted in blue (dark gray in print versions) The bottom shows the actualchamber with a sealed tape that is transparent (B) Cross-flow chamber uses a longer coverglass with the flow direction perpendicular to the slide long axis enabling the flow path to beaccessible when the cover glass is down on an inverted microscope The top shows aschematic with the tape in gray and the flow path in blue (dark gray in print versions) Thebottom shows the actual sealed chamber (C) Example of a chamber where the double-sidedtape making the flow path is not sealed Comparing the look of the tape to that in (A), it is clearthat the tape is less transparent and thus not making contact with both the cover glass andthe slide (D) Schematic of the side view of the flow chamber with the cover glass down on aninverted microscope (E) Schematic of the surface treatment of the flow chamber in order toperform a gliding assay Kinesin coats the cover glass first The microtubules are flowed inlater to bind to the kinesin The kinesin motors walk to the plus-end pushing the microtubulewith the minus-end forward (F) Time series of a gliding assay Two clear filaments gliding aredenoted by arrowheads The time between frames is 55 s The scale bar on the last frame

is 5mm

26 CHAPTER 2 Microtubules, MAPs, and motor patterns

The chambers are used coverslip side down on a modern inverted microscope for

epifluorescence This makes it difficult to flow in more reagents while imaging To

have a chamber with an accessible flow path, you can build a cross-flow chamber

with a 22
30 mm coverslip perpendicular to the slide (Figure 1(B)) When the

cross-flow chamber is placed on the inverted microscope, you can still pipette

into the chamber

We often work with the chamber open at the ends of the flow path because we do

not image very long (<1 h) If longer time imaging is needed, we seal the chamber

with 5 min Zpoxy

Piperazine-N,N0-bis(2-ethanesulfonic acid) (PIPES), 1 M stock solution, pH 6.8

using KOH (K-PIPES) or NaOH (Na-PIPES) (Sigma)

Tubulin, unlabeled and fluorescently labeled (cytoskeleton or homemade, using

the method ofPeloquin, Komarova, and Borisy (2005))

in 4C.

aggregates at 360,000
g in microultracentrifuge at 4C Use supernatant and

polymerize microtubules at 37C for 30 min Add Taxol to a final concentration

of 20mM and incubate for 20 min at 37C to equilibrate the Taxol Centrifuge

for 15 min at 10,000
g to pellet microtubules Resuspend in PEM-100 with

20mM Taxol Store on the bench for up to 2 weeks

Kinesin-1 truncated form, K560, with green fluorescent protein (GFP) and 6xHis

tag purified with Nickel-NTA agarose beads, as previously described (Pierce &

Vale, 1998) Store aliquots in80C.

Deoxy: 10 mg/mL glucose oxidase, 4.7 mg/mL catalase, in double distilled

water Spin down solids and remove the saturated solution at the top Store in

4C for up to a week.

PEM-100 Make the day of the assay and do not store for reuse

PEM-100 Add the starred items just before use so that do not become

exhausted before the imaging begins Make the day of the assay and do notstore for reuse.

Microtubule dilution: Typically dilute stock of microtubules 1:100 (final

1 Make a standard flow chamber along with needed reagents Keep reagents and

buffers on ice except for microtubule stock, Taxol stock in dimethyl sulfoxide(DMSO), and microtubule dilution Combine all reagents except ATP, deoxy,and glucose, which should be added just before the experiment so they do notbecome exhausted

2 Thaw kinesin aliquot(s) on ice.

3 Flow in 10mL of kinesin into the flow chamber To allow kinesin to bind to glass,incubate for 5 min Two aliquot flow throughs may be needed depending onactivity of kinesin prep

4 Flow in 10mL of BSA wash buffer into the flow chamber Wick out originalkinesin-1 wash using a filter paper or kimwipe as you pipette in the next 10mL

5 Flow in 10mL of microtubule dilution into the flow chamber Incubate for 2 min

6 During the 2 min incubation add glucose and deoxy to the Motility Mix buffer.

7 Flow in 10mL of Motility Mix buffer to the flow chamber The Motility Mixactivates the gliding assay by giving kinesin the ATP source to move themicrotubules

8 Image using epifluorescence Take movies with 3e5 s between frames,

shut-tering in between to avoid photobleaching or photodamage Movies are saved asstacks of tiffs or ND2 files using Nikon Elements software

We use ImageJ to open the movie data and analyze the motion of gliding filaments.Filament gliding can be analyzed in two different ways: (1) Kymographs or (2) Fila-ment end tracking

2.4.1 Kymographs

One-dimensional motion can easily be observed using a kymograph or spaceetimeplot Kymographs are often used in single molecule assays where the track ofthe motor protein is the microtubule, in which case it is obvious where the one-dimensional path lies For a gliding microtubule, it still moves in one dimension,but the direction can change over time and the path may not be obvious evenwhen watching the recorded data movie playback (Figure 2(A))

1 In order to determine the path, we use ImageJ to collapse the stack into a single

frame by going to Image/Stacks/Z Project (Figure 2(B))

2 In this menu, there are a number of different methods to collapse a stack The

best method for detecting the motion of the filaments is the standard deviation

28 CHAPTER 2 Microtubules, MAPs, and motor patterns

FIGURE 2

Data analysis of gliding microtubules (A) Time series of gliding microtubules same as

Figure 1, except the frames are vertical The time between frames is 55 s The scale bar is

5mm (B) Standard deviation z-projection of the movie portrayed in (A) The location of the

left side of the microtubule marked with a wide arrowhead in (A) is marked with an asterisk

This frame is scaled up 200% compared to the frames in part (A) (C) The same standard

deviation z-projection shown in (B) with the segmented line drawn along the trace of the path

of a moving microtubule with the thick arrowhead in (A) The beginning of the segmented line

that was drawn is denoted with a “1.” The label as serves as a label for denoting the

kymograph from the movie later (e.g., Kymograph1 corresponds to the track 1) (D) The

segmented line selection is restored on the first frame of the movie (E) A kymograph is

created from the movie showing the direction of the moving microtubule The asterisk

denotes the same end of the microtubule denoted by the asterisk in (B) The top, left corner is

also the origin of the coordinate axis of the kymograph where t¼ 0 and x ¼ 0 (F) The average

speed can be measured from the kymograph by measuring the number of horizontal

distance pixels (Dx) over a set time displacement in the vertical direction (Dt) The velocity is

computed as v¼ Dx/Dt