Approaches To Studying Cellular Signaling: A Primer For ......boda has used several whole tissue...

TUTORIAL

Approaches To Studying Cellular Signaling:A Primer For MorphologistsKATHY KAY HARTFORD SVOBODA* AND WENDE R. REENSTRA

Many research projects will lead to understanding tissue and/or cell responses to extracellular influences either fromsoluble factors or the surrounding extracellular matrix. These types of investigations will require the understandingof signal transduction. This particular cell biological field has literally exploded with information and new technicalapproaches in the past 10 years. This article is directed toward investigators interested in using these newapproaches to study their systems. An overview of the general principles of signal transduction events including thetypes of receptors and intracellular signaling events is followed by an introduction to methods for visualizing signaltransduction. This is followed by an introduction to biochemical analysis and an example of combiningseveral approaches to understanding a tissue response to extracellular matrix stimulus. Anat Rec (New Anat) 269:123–139, 2002. © 2002 Wiley-Liss, Inc.

KEY WORDS: extracellular matrix; ECM; Rho; green fluorescent protein; GFP; FRET; FRAP; GST assays; methods; signaltransduction; morphology; imaging

INTRODUCTION

Cellular signaling is a complex balletof molecules interacting and stimulat-ing surrounding proteins, lipids, and

ions, resulting in cytoskeletal reorga-nization, modulation of differentia-tion, and induction of gene expres-sion. The choreography of events insignaling pathways has been a hottopic in the past few years. The inter-molecular reactions should be viewedas a fluid dynamic process in whichmultiple events may be occurring atthe same time. Many of the presentexperimental procedures documentsignal transduction snapshots of indi-vidual events, such as tyrosine phos-phorylation, without the benefit of visu-alizing the whole stage simultaneously.However, newer assays using fluores-cently tagged transfected proteins com-bined with live cell imaging analysiswill both confirm and modify currentperceptions of signal transductionevents. We have been focusing on inte-grin-mediated pathways, but many ofthe experimental approaches describedin this publication can be used for anysignal transduction cascade. Therefore,this article will naturally be directed to-ward how integrins and growth factorsinitiate and propagate their signals.

Before discussing specific ap-proaches to visualizing and biochem-ically analyzing signal transductionevents, we will provide suggestions on

how to understand the literature, ex-plain the general principles, and de-scribe simple cellular behavior assays.Both historical and current proce-dures will be explained, including thenecessary reagents and laboratoryequipment, with the goal of demon-strating that morphologically equippedlaboratories have the ability to conductthese experiments with minimal invest-ment in time and funds.

Many cell and developmental biolo-gists who are currently analyzing sig-nal transduction in their systems havea similar story. We really did not startour careers expecting to be talking inthree letter words and explaining cas-cades of reactions that led to our ob-servations. One of us started her sci-entific career studying how thecytoskeleton controlled cell shapethroughout the development of theoptic vesicle (Svoboda and O’Shea,1984, 1987), and then continued ex-amining the role of the actin cytoskel-eton in corneal epithelial responses toextracellular matrix (ECM) (Svobodaand Hay, 1987; Svoboda, 1992;Khoory et al., 1993; Yeh and Svoboda,1994; Hirsch et al., 1996; Svoboda etal., 1999a). After several years of de-scribing the morphologic changes in

Dr. Reenstra is the Associate Director ofResearch for the Department of Emer-gency Medicine at the Beth Israel Dea-coness Medical Center, Harvard Medi-cal School. She is also an instructorwhile completing her residency in emer-gency medicine. Dr. Reenstra has fo-cused her research on signal transduc-tion events in dermal fibroblasts andvascular endothelial cells in response tovarious stimuli, including growth factorsand oxygen tension. Dr. Svoboda is aProfessor of Biomedical Sciences andDirector of the Cell and Molecular Biol-ogy Core Facility at Texas A&M HealthScience Center at Baylor College ofDentistry. She also holds an AdjunctProfessorship in the Ophthalmology De-partment of the University of Texas,Southwestern Medical Center. Dr. Svo-boda has used several whole tissue de-velopmental models to study cell–ma-trix interactions, actin reorganization,and signal transduction events.*Correspondence to: K.K.H. Svoboda,Department of Biomedical Sciences,Texas A & M University System HealthScience Center, Baylor College ofDentistry, 3302 Gaston Avenue, Dallas,TX 75246. Fax: 214-828-8951; E-mail:[email protected]

DOI 10.1002/ar.10074Published online in Wiley InterScience(www.interscience.wiley.com).

THE ANATOMICAL RECORD (NEW ANAT.) 269:123–139, 2002

© 2002 Wiley-Liss, Inc.

cells, it was apparent that an under-standing was needed regarding howthe changes in cytoskeletal proteinswere controlled. This investigation leddown the path into signal transduc-tion (Svoboda et al., 1999b; Chu et al.,2000). In this publication, our objec-tive is to relate how a laboratoryequipped to examine subtle morpho-logic changes can use new tools todetermine which signal transductionmolecules may be responsible for thechanges.

GETTING STARTED

One of the biggest hurdles to learningsignal transduction is understandingthe language. This area of cell biologyis full of jargon and acronyms thatwill quickly trip up the newcomer(See Box 1 for a list of common acro-nyms). Therefore, the first objective isto learn the terminology. The firstword of caution is do not expect thenames of proteins to be logical, asmany were named for their first

known function, but now have beenshown to have a wider distributionand functional activity. In addition,there may be multiple names for thesame protein or the same names (orabbreviations) used for different pro-teins, adding to the confusion andfrustration to a new investigator.Many Web sites have been generatedrecently that cover the basic con-cepts (http://www.unige.ch/sciences/biochimie/Lafont/WA_CB.html orhttp://geo.nihs.go.jp/csndb/) and aregood places to start understanding thedifferent signal pathways. Many ven-dors have produced informative Websites and general literature thatshould not be overlooked. A recenthistorical review gives the perspectiveand nomenclature in the field of signaltransduction (Hunter, 2000) and maybe another possible starting point.Several good review articles on thepathways of interest (Mecham, 1991;Meredith et al., 1996; Cary and Guan,1999; Gonzalez-Amaro and Sanchez-Madrid, 1999; Duong et al., 2000; Rid-

ley, 2000a,b; Roovers and Assoian,2000) combined with a few colleaguesor students are the ingredients neededto start a journal club. The journalclub should start by reading the chap-ters in a modern cell biology text (Lo-dish et al., 2000), or a monograph onthe specific subject, followed by somereview articles, then tackle at least onecurrent article a week. I am suggest-ing this route because the learningcurve takes approximately 3–6months just to understand the lan-guage of selective signal transductioncascades. As an example, I am inter-ested in cell–matrix interactions, andthe understanding of these cascadeshave become increasingly compli-cated in the past 3 years; a recent lit-erature search (1998–April of 2001),by using integrin and signal transduc-tion as key words produced thousandsof citations and over a hundred re-views.

GENERAL PRINCIPLES OFSIGNAL TRANSDUCTION EVENTS

Receptors

Intracellular signaling is triggered bya cell surface event such as a receptor–ligand interaction, cell–cell contact,or cell–ECM contact (Figure 1) (Sas-try and Burridge, 2000). The contactwith other cells or the environmentcan stimulate many cellular reactions,including proliferation, motility, dif-ferentiation, or even programmed celldeath, termed apoptosis. The receptoror cell surface binding proteins areclassified by the protein structure andligand characteristics. The basic pro-tein domains for a receptor are an ex-tracellular ligand binding domain,membrane spanning domain, and acytoplasmic domain (Figure 1). Mostmodern cell biology textbooks list atleast four types of cell surface recep-tors, including the G-protein-coupled

Figure 1. Many different types of receptors are on cells that contribute to some of the samesignal pathways. In this illustration, G-protein–coupled receptor (GPCR); receptor tyrosinekinase, cell adhesion molecules (CAMs), selectins, integrins, syndecans, and cadherins arerepresented with some of their intracellular signal molecules identified. This image was usedto show that multiple signal pathways control focal adhesion assembly by coordinatingregulation of RhoA. Serum factors such as LPA signal by means of GPCRs to activateguanine-nucleotide exchange factors (GEFs) for RhoA. Growth factors (platelet-derivedgrowth factor [PDGF], epidermal growth factor [EGF]) act also, transmitting stimulatory andinhibitory signals to RhoA. Integrins transmit positive and negative signals to RhoA. Initialintegrin-substrate interactions decrease RhoGTP but activate p190RhoGAP through cSrc. Asthe cell spreads, integrins activate RhoA. Downstream of RhoA, p160ROCK stimulatesactin-myosin contraction and increased integrin clustering. Adapted with permission fromSastry and Burridge, 2000.

One of the biggesthurdles to learning

signal transduction isunderstanding the

language.

124 THE ANATOMICAL RECORD (NEW ANAT.) TUTORIAL

receptors, ion-channel receptors, ty-rosine kinase-linked receptors, and re-ceptors with intrinsic enzymatic activ-ity (Lodish et al., 2000). TheG-protein–coupled receptors (GPCR)are characterized by multiple trans-membrane domains (usually seven)that wind the protein in and out of themembrane in a serpentine conforma-tion (Figure 1, GPCR). The ion-chan-nel receptors are closely related andactually open a membrane channelwhen the ligand binds. Many cytokinereceptors are in the tyrosine kinase-linked class, as they lack intrinsic ac-tivity, but when the ligand binds, in-tracellular tyrosine kinases becomeactivated to generate cellular changes.The classic growth factor receptors dohave intrinsic kinase activity and,therefore, make up the fourth class of

receptors, the receptor tyrosine ki-nases, or receptor serine/threonine ki-nases (Figure 1). These receptors usu-ally have one transmembrane domainand at least two molecules mustdimerize to activate the signal.

In addition to these classic recep-tor classes, cells can respond to theirextracellular environment throughintegrin receptors or proteoglycanreceptors such as the syndecan fam-ily (Figure 1). Single transmembranedomains with large extracellular andmuch smaller cytoplasmic domainscharacterize these receptors. Thesyndecan molecules have long gly-cosaminoglycan chains that assist insequestering the fibroblast growthfactors close to the cell membrane(Couchman and Woods, 1999; Rich-ardson et al., 1999). The integrin re-

ceptors are heterodimers composedof � and � subunits. The family isvery large with at least 25 integrinheterodimers, including 19 � sub-units and 8 � subunits identified(Humphries, 2000). The integrins donot have kinase activity, but uponbinding to ECM molecules, some asso-ciated proteins become activated by au-tophosphorylation and then phosphor-ylate surrounding proteins to generatesignals for specific cellular activitiesthat usually change the actin cytoskel-eton. One of the first proteins to be-come phosphorylated is focal adhe-sion kinase (FAK).

Another concept in recent literatureis that different signaling moleculesmay be sequestered in membrane mi-crodomains, termed lipid rafts (Chap-man et al., 1999; Giancotti and Ruo-

BOX 1. Signal Transduction-Related Abbreviations Used in this Article

ACM, actin cortical matbax, bcl-associated � protein that accelerates cell death

(apoptosis)BL, basal laminaBMP, bone morphogenetic protein, a member of the TGF�

familyCAM, cell adhesion moleculescAMP, cyclic AMPcAPKs, cyclic AMP-dependent kinases, also called protein

kinase A (PKA)CD, cytochalasin DCGMP, cyclic GMPcLSM, confocal laser scanning microscopeCOL, type I collagenDAG, diacylglycerolDia, p140 DiaphanousECL, enhanced chemiluminescenceECM, extracellular matrixEGF, epidermal growth factorErk, extracellular signal regulated protein or externally reg-

ulated kinasesF-actin, filamentous actinFAK, focal adhesion kinaseFGF, fibroblast growth factorFN, fibronectinFRAP, fluorescence recovery/redistribution after photo-

bleachingFRET, fluorescence resonance energy transferGAP, GTPase activating proteinGDI, guanine-nucleotide dissociation inhibitorGEF, guanine exchange factorsGFP, green fluorescent protein; CFP (CyanFP); YFP (yel-

lowFP); and DsRed are variations that emit signal atdifferent wavelengths

GPCR, G-protein–coupled receptors

Grb2, growth factor receptor bound protein 2GST, glutathione-S-transferaseHRP, horseradish peroxidaseICE, interleukin-1 converting enzymeJNK/SAPK, Jun N-terminal kinase also known as SAPK,

stress-activated protein kinaseLPA, lysophosphatidic acidMAP kinase, mitogen-activated protein kinaseMAPKK, mitogen-activated protein kinase kinase, also

known as MEKMAPKKK, mitogen-activated protein kinase-kinase-kinase

or MEKKPDGF, platelet-derived growth factorP190RhoGAP, 190-kDa protein that functions as a GAP forRhoPAX, paxillinPI3kinase, phosphatidylinositol 3 kinasePIP2: phosphatidylinositol (4,5)-bisphosphatep-tyr, phosphotyrosinePY-20, an antibody that recognizes tyrosine phosphory-

lated proteinsROCK, Rho associated coiled-coil containing protein ki-

nase also known as p160 RhokinaseRho, a small GTPase family that includes Rac and cdc42SH2, Src homology 2 domainsSH3, Src homology 3 domainSmad, sma and mad homology signal proteins renamed

1996, act in the TGF� pathwaySOS1, son of sevenlessSrc, a kinase that was first described in the Rous sarcoma

virus, however, it is also found in normal cellsTEM, transmission electron microscopyTGF�, transforming growth factor betaTUNEL, Tdt-mediated dUTP-X nick-end labeling for apo-

ptosis

TUTORIAL THE ANATOMICAL RECORD (NEW ANAT.) 125

slahti, 1999) that may facilitate cross-talk between different receptors andsignaling pathways (Schwartz andBaron, 1999; O’Neill et al., 2000; Rid-ley, 2000a,b; Sastry and Burridge,2000).

Intracellular Signaling Proteins

One basic principle of signal trans-duction is that the proteins exist in atleast two states: activated and inacti-vated. At this time, we know of severalmethods to turn the signals on andoff, including phosphorylation, de-phosphorylation, intracellular loca-tion, nucleotides (ADP/ATP or GDP/GTP) cycles, and calcium/ion levels.In this section, an overview of differ-ent methods of producing intracellu-lar signaling will be briefly explained.

One of the first signal transductionmechanisms described was the GPCRcascade that generates the classic sec-ond messengers cyclic AMP (cAMP),cyclic GMP (cGMP), diacylglycerol(DAG), phosphoinositols, and calcium(Ca�2). When G-protein linked or hor-mone receptors become activated,they trigger a series of events at thecell surface that cause transient in-creases in these second messengermolecules (Lodish et al., 2000). Aswith other signaling events, there is atransient increase in the active formof the molecule followed by a rapiddecrease to produce a “signal.”Briefly, when the ligand binds to thereceptor, it causes a conformationalchange that allows the G-protein �subunit to bind to the cytoplasmic do-main of the receptor (Figure 1). Thisinteraction causes the exchange ofGDP for GTP on the � subunit and thedisassociation of the �� subunits fromthe � subunit. The activated (GTPbound) � subunit interacts with ad-enylyl cyclase, the membrane-boundenzyme responsible for producingcAMP. After activating adenylyl cy-clase, the � subunit reverts to the GDPstate and re-associates with the ��subunits.

Not only is the number of GPCRsvery large, but the number of � sub-units are also numerous, providing awide variety of signals to the cell.Once the G proteins are activated, thesignal can be amplified by interac-tions with other proteins as illustratedin Figure 1. Classically, adenylyl cy-

clase produces cAMP that activatesother kinases, termed the cyclic AMP-dependent kinases (cAPKs) also calledprotein kinase A (PKA). These kinasescan phosphorylate several substrates,depending on the specific stimulus(Lodish et al., 2000) to amplify thesignal from the cell surface. One effectof activation is the release of calciumfrom intracellular stores such as therough endoplasmic reticulum and mi-tochondria. Because free calcium lev-els in cells are maintained at very lowlevels, the rapid increase in calciumlevels from these intracellular storeshas been a way to visualize signalevents. Calcium can be labeled witheither quantitative ratiometric dyes orsingle-wavelength dyes to monitorthese rapid changes in cells after stim-ulation (Nuccitelli, 1994). Rapid se-questering of the free calcium ions by

molecules such as calmodulin thatbind several calcium ions turns off thecalcium signal.

Phosphorylation and dephosphory-lation provide another mechanism forsignal on and off switches. The en-zymes that phosphorylate other pro-teins are called kinases. The commonamino acids that become phosphory-lated are tryosine, threonine, andserine. As stated previously, the phos-phorylated proteins may be activatedby adding phosphate molecules, anddeactivated by removing the phos-phates, a function usually performedby another enzyme class, the phos-phatases. Some proteins can phos-phorylate themselves (“autophosphor-ylation”); once activated, they canphosphorylate surrounding sub-strates. An example of this type of pro-tein was discussed previously in de-

scribing the integrin molecules; focaladhesion kinase (FAK) becomes phos-phorylated when integrins bind totheir ligand, ECM. Once FAK be-comes phosphorylated, it will activateor phosphorylate paxillin, an actin-as-sociated protein, and another kinase,Src (Cary and Guan, 1999). Src canalso start activating surrounding pro-teins, creating an amplification of theoriginal signal. To complicate things,some proteins are inactive in thephosphorylated state and become ac-tive after dephosphorylation. There-fore, it is important to understand thepossible changes in the proteins be-fore starting an investigation. Cellsmay contain a constant amount of agiven protein in a pool, but the proteinhas to be in an active state to producea signal that will change the proteinsaround it (Svoboda et al., 1999a,b). Itis important to remember that, justbecause an mRNA for a given proteinis expressed, it does not indicate thatthe protein is produced or that it isactivated.

Another way that proteins get acti-vated involves binding to a nucleotidesuch as ATP or GTP. The protein isgenerally in an inactive state if theADP or GDP nucleotide is bound andbecomes activated when the ATP orGTP is bound. An example of this typeof activation is the small G-proteinfamilies, Ras and Rho (Bokoch, 2000;Ridley, 2000a; Sastry and Burridge,2000; Schwartz and Shattil, 2000;Settleman, 2000; Symons and Settle-man, 2000). These proteins alternatebetween the GTP-bound active form(on) and the GDP-bound inactiveform (off) to regulate other down-stream kinases. Many other proteinsregulate the “on” and “off” state of thesmall G-proteins. The guanine-nucle-otide exchange factors (GEFs) are the“on” signal, as they add GTP to theprotein. GTP-activating proteins(GAPs) are the “off” signal, as theyremove a phosphate to deactivate theprotein. Guanine-nucleotide dissocia-tion inhibitor proteins (GDIs) seques-ter the inactive protein in the cyto-plasmic pool. We have shown that oneof these regulatory proteins(p190RhoGAP; 190-kDA protein thatfunctions as a GAP for Rho) becomesphosphorylated very quickly in em-bryonic epithelia in response to cellsbinding ECM (Svoboda et al., 1999b).

For a cell to achieve adifferentiated phenotype

or respond tomicroenvironmentchanges, the ECM

molecules and theirreceptors must integrateboth form and function.


We have also shown that decreasingRho protein levels or activity de-creased other integrin signaling mole-cules.

Another way to determine whethera protein is activated is to determineits intracellular location. Some pro-teins move to specific intracellularstructures, such as focal adhesions(Sastry and Burridge, 2000; Turner,2000) when they become activated.Paxillin, �-actinin, and talin accumu-late at the focal adhesions in both mi-gratory and stationary cells. In addi-tion, integrin molecules becomeclustered at the focal adhesion of fi-broblast cells in culture.

Many proteins move to the plasmamembrane when they become acti-vated. The movement to the plasmamembrane may take several steps, in-cluding release from a cytoplasmicchaperone protein, acquiring a lipidtail, or both. The small G proteins(such as Rho) require both the releasefrom the GDI protein (Read et al.,2000; Michaelson et al., 2001) and alipid tail to move to the membrane.So, in addition to having a GTP boundto the protein, the protein itself movesto the site of action. Many of the smallGTPase proteins (Ras, Raf, Rac) fol-low similar intracellular translocationpatterns upon activation.

Other activated proteins move tothe nucleus and usually act as tran-scription factors. Examples of thistype of intracellular translocation are

some of the MAP kinase proteins(Kortenjann and Shaw, 1995; Treis-man, 1996; Roovers and Assoian,2000). MAP kinases can respond to avariety of extracellular signals, includ-ing osmotic stress, heat shock, cyto-kines, and mitogens (Garrington andJohnson, 1999). Two of the MAP ki-nases, the extracellular signal-regu-lated kinases (erk-1 and erk-2, also re-ferred to as erk-1/2), translocate to thenucleus after activation (Figure 2) toregulate the expression of varioustranscription factors (c-myc, c-fos,and c-jun, tcf, srf, elk-1, apl, atfn)(Kortenjann and Shaw, 1995; Gar-rington and Johnson, 1999). Activa-tion of the MAP kinase pathways hasbeen identified as a mechanism usedby integrins to regulate gene expres-sion leading to cell shape changes dur-ing cell spreading or migration (Rob-inson and Cobb, 1997; Schwartz andBaron, 1999; Schwartz and Shattil,2000), and as a cross-talk pathway be-tween integrins and growth factors(Schwartz and Baron, 1999; Sastryand Burridge, 2000).

Sometimes the signal protein needsto bind another protein, a chaperone,before translocating to the nucleus.Transforming growth factor beta(TGF-�) proteins activate a class ofproteins called Smads. These growthfactors bind to cell surface receptors(T�RI and T�RII) and activate specificSmad proteins (Smad 2 or 3) that thenbind to a chaperone protein, Smad 4

before translocating to the nucleus toact as transcription factors (Giorgioand Hemmati-Brivanlou, 1999; Mas-sague, 2000; Schiffer et al., 2000; Zim-merman and Padgett, 2000). Simi-larly, bone morphogenetic proteins(BMPs) bind to BMP receptors andactivate Smads 1 or 5, which bindSmad 4 to translocate to the nucleus.

In summary, intracellular re-sponses to cell surface receptors arecomplicated and poorly understood.Several studies have established recip-rocal linkages between ECM-inte-grins, growth factor signaling, cell–cell adhesion molecules, specializedmembrane domains (lipid rafts) andG-protein–linked receptors (Sastryand Burridge, 2000). In addition,cross-talk has been established be-tween the ECM and intracellular mi-togen stimulated pathways, the smallG-proteins, and the phosphoinositols(Porter and Hogg, 1998; Ossowski andAguirre-Ghiso, 2000). The cell’s mi-croenvironment and the resulting tis-sue profoundly influence each of theselinkages. Thus, for a cell to achieve adifferentiated phenotype or respondto microenvironment changes, theECM molecules and their receptorsmust integrate both form and func-tion. In contrast, mutated genes andaberrant interactions with the micro-environment may degrade this inte-gration, possibly resulting in malig-nant transformation or abnormaldevelopment (Boudreau and Bissell,1998). Recently, it has also becomeapparent that integrins regulate RhoGTPases and vice versa. Integrins andGTPases, therefore, might be orga-nized into complex signaling cascadesthat regulate cell behavior (Schwartzand Baron, 1999; Sastry and Burridge,2000; Schwartz and Shattil, 2000).

CELL BIOLOGIC ANALYSIS OFSIGNAL TRANSDUCTION

In the following sections, some spe-cific methods that have been used toestablish the signaling mechanismswill be explored.

Blocking Specific Kinases WithInhibitors

Signal transduction proteins havebeen the targets of many pharmaco-logic agents over the past 20 years.Many inhibitors are available for both

Figure 2. Fibroblast cells cultured on plastic substrate, serum starved for 24 h, and thenstimulated with fibronectin (FN). A: Anti-phosphotyrosine Western blot from cells (0) 30 s, 2,5, 10, 15, and 60 min after FN stimulation in the MAP kinase molecular weight region anddouble-labeled immunohistochemistry for erk 1 (green) and anti phosphotyrosine (red). B:In unstimulated cells, the erk distribution appeared diffuse throughout the cells. C: Whereasafter 15 min, many proteins are phosphorylated (red) and the erk appears to be movingmore centrally with some double-labeled proteins (yellow) moving toward the nucleus. D:At 45 min, the nuclei clearly contain both erk and phosphotyrosine (yellow nuclei).


generalized and specific interactionswith either receptors or specific ki-nases; we have provided a short inhib-itor list as an example of the varietyavailable (Table 1). The reader is cau-tioned to be extremely careful whenusing the inhibitors to include carriercontrols and a titration analysis foreach inhibitor on their system. Manyinhibitors may be specific for one ki-nase at low doses, but as the dose in-creases, they may affect a wider classof molecules. Many act as competitiveinhibitors for ATP binding sites andare reversible; therefore, a constantlevel of inhibitor is required through-out the experimental time both beforeand after cellular stimulation. The as-sessment of inhibitor effects shouldinclude morphologic, biochemical,and behavioral analysis as describedhere.

Cell Behavior Assays

To study the signal transductionevents in cells, cell behavior assaysneed to be developed that can bequick, easy, and preferably used onliving tissues and cells. These assaysmay determine the differences in cy-toskeletal elements, cell shape

changes, matrix binding, and migra-tion or differentiation characteristics.Attachment, spreading, and migrationassays were developed over 25 yearsago to determine whether cells re-acted differently to alternate sub-strates. These assays are now beingused to determine whether down- orup-regulating specific signal transduc-tion proteins change the cellular be-havior (Berrier et al., 2000).

Signal transduction proteins aremodulated by specific inhibitors to in-tracellular kinases or cell surface re-ceptors. A representative sample ofsome inhibitors and how they inter-fere with protein function is shown inTable 1; however, many more inhibi-tors have been developed, as theseproteins may be pharmacologic tar-gets. Another way to change the activ-ity of specific proteins is by transfect-ing or microinjecting cells withplasmids that express either inactiveor constitutively active proteins (Ber-rier et al., 2000). Alternately, blockingsynthesis with antisense oligonucleo-tides specific for the mRNA of a se-lected protein can down-regulate spe-cific proteins.

We routinely use two assays in our

corneal epithelial research project: ac-tin reorganization and ECM binding(Svoboda et al., 1999b). Corneal epi-thelia respond to ECM molecules byreorganizing the basal actin cytoskel-eton into long actin bundles (Figure4), termed the actin cortical mat(ACM). Once the cellular behaviorwas established, then we selectivelydecreased the activity of individual orgroups of proteins (by using specificinhibitors or antisense oligonucleo-tides) to establish whether those pro-teins are necessary for the actin toreorganize. Our results show thatmany signal transduction pathwayscontribute to this cytoskeletal rear-rangement (Figure 3).

The ECM-stimulated changes in ac-tin cytoskeleton organization havebeen well documented in this tissueby using transmission electron mi-croscopy (TEM) (Sugrue and Hay,1981, 1986; Svoboda and Hay, 1987)and confocal microscopy (Svoboda,1992; Svoboda et al., 1999b; Chu et al.,2000). In addition to the time saved byusing a fluorescent phalloidin stainingmethod compared with TEM, wefound that different ECM moleculesstimulated distinct actin bundle con-

TABLE 1. KINASE AND EGFR INHIBITORS

Kinase Inhibitor Action

MAP KinaseMEK/ERK PD98059 Competitive inhibition of ATP binding site

Olomoucine Competitive inhibition of ATP binding siteMEK1 PD184352 Directly affects MEK1ERK2 5-Iodotubercidin Competitive inhibition of ATP binding sitec-Raf ZM336372 Competitive inhibition of ATP binding siteMEK1/MEK2 U0126 Non-competitive with respect to MEK substrate, ERK,p38 MAPK SB202190 Potent, cell-permeable; inhibits phosphorylation of myelin basic protein no

effect ERK or JNKPI3 Kinase LY294002 Competitive inhibition of ATP binding site

Wortmannin Competitive inhibition of ATP binding site; irreversible inhibitor, inducesconformational change in catalytic domain

FAK echistatinSrc PP2 Potent and selective inhibitor for SRC family; inhibits activation of FAK and its

phosphorylation at Tyr577

Herbimycin A Potent and cell-permeable inhibitor of SRC familyRho C3 Exoenzyme ADP ribosylation of all Rho proteinp160 ROCK Y27632 Tyrosine phosphorylation inhibitionMLCK ML-7 Cell-permeable and highly selective inhibitor

ML-9 Cell-permeable and highly selective inhibitorA3 Hydrochloride Alkyl chain derivative

EGF-Receptor AG1478 Selective inhibitor abolishes ERK activationCompound 56 Most potent and specific inhibitor of tyrosine kinase activityPD174265 Reversible, potent, and cell-permeable inhibitor of EGF-R

FGF-Receptor SU5402 Inhibits FGF-R kinase activity, also inhibits aFGF induced tyrosinephosphorylation of ERK1 & 2-does NOT inhibit EGF-R kinase activity


figurations (Svoboda et al., 1999b).Also, we demonstrated that this tissuereorganizes the actin cytoskeleton inresponse to bombesin or lysophos-phatidic acid (LPA) (Svoboda et al.,1999b), which bind to the classic 7membrane spanning G-protein–typereceptors. The equipment needed tomorphologically assess the actin cy-toskeleton includes a microscope, ei-ther a TEM or a confocal microscope.As this tissue is between 20 and 30microns thick, we need to either phys-ically section the material for TEM oroptically section the tissue with con-focal microscopy. The actin wasstained with fluorescently taggedphalloidin for the confocal micro-scope, whereas traditional TEM fixa-tion and staining were used to deter-mine the organization of the actin byusing TEM analysis.

The advantages of confocal micros-copy include time and field size. Thephalloidin staining takes approxi-mately 30 min, allowing tissues to beviewed within an hour of experimen-tal treatment. The tissues are scanneden face with field sizes as large as 300square microns. In contrast, the fixa-tion, embedding, sectioning, andstaining preparation for TEM takes aminimum of 3 days and often may

take up to a week between the end ofthe experiment and morphologicanalysis. In addition, even at low-power observation, only very smallfields of cells can be observed. Manytimes while viewing TEM samples, wehad to search for actin bundles, as thespecific section may not have includedthe bundle. Furthermore, fine actinbundles were obscured from view bythe embedding medium, plastic.

The second morphologic assay weuse, ECM binding to the basal epithelialsurface, also depends on confocal mi-croscopy for analysis. In this assay, wefluorescently label ECM molecules (col-lagen or fibronectin), or more recently,we have purchased labeled type IV col-lagen or gelatin (Molecular Probes,Inc.) or collagen-coated beads. The la-beled ECM molecules are incubatedwith the tissues for designated times.The tissue is then rinsed, stained withphalloidin, and viewed with the confo-cal microscope (Figure 5). The bindingof the ECM over a temporal sequence isestablished, then used as an assay todetermine whether selective inhibitorsblock ECM binding in a specified time(Chu et al., 2000).

Other cell behavior changes thatcan be followed are cell attachment,shape change, migration, prolifera-

tion, or survival (Wayner et al., 1991;Hanks et al., 1992; Chen et al., 1994;Mooney et al., 1995; Meredith andSchwartz, 1997; Frisch, 1999). For ex-ample, numerous researchers use thein vitro wound-healing model inwhich the cells are grown to conflu-ence, and then a scrape or wound isplaced in the culture dish (Nobes andHall, 1999; Song et al., 2000). The cellsare observed moving over the woundsurface under various conditions todetermine what proteins are neces-sary for the cell migration to close thewound. This method is a very easyassay and has the further advantage ofliving cell observations.

Visualizing Signal TransductionEvents

Traditionally, signal transductionevents, such as calcium or pHchanges, were tracked with intracellu-lar fluorescent indicators. Transloca-tion of specific proteins was trackedwith immunohistochemistry (Figure6), caged proteins, or fluorescentmarkers for certain plasma mem-brane lipids. More recently, move-ment of specific proteins has beentracked by incorporating a fluorescentprotein gene, green fluorescent pro-tein (GFP), into genetic vectors encod-ing the protein to be studied.

In this section, the general princi-ples of how these different ap-proaches are used to visualize signaltransduction events will be described.However, many articles, reviews, andbooks that explain the details of eachapproach have been published. In ad-dition, many companies are develop-ing products designed to optimize vi-sualizing cell biological events withboth wide field and confocal micros-copy. All of these techniques requirehigh-resolution light microscopeswith light sources that can excite flu-orescent probes. Most also requiredigital recordings with additionalcomputer analysis.

Several short courses are availableto either introduce the novice or expe-rienced morphologist to these tech-niques. A few Web sites for furtherinformation include the following:The Marine Biological Laboratory hastwo courses (http://courses.mbl.edu/Courses/about.htm), the W.M. KeckCenter for Cellular Imaging also has

Figure 3. Schematic drawing of our proposed signaling hypothesis. The stages of epithelialresponse to extracellular matrix (ECM) may be divided into four parts: (1) first contact, (2)signal amplification, (3) actin polymerization and contraction, (4) integrin clustering leadingto increased ECM binding and further signal amplification. RhoGTPases increase actinorganization and contraction through several mechanisms as indicated in the diagram.Adapted with permission from (Chu et al., 2000).


workshops (http://www.cci.virginia.edu/workshop_fret.html), and the Im-age Facility at San Antonio has ayearly course (http://www.uthscsa.edu/csb/image/facility.html). In addi-tion to these courses that take from3–7 days, many national meetingshave short workshop courses. TheAmerican Association of Anatomistshas sponsored imaging workshops atthe national meeting for the last 5years (http://www.anatomy.org).

Calcium and pH

Changes in calcium or pH levels canbe accomplished by loading the cellswith ratiometric or single-wavelengthdyes. These fluorescent probes thatshow a spectral response upon bind-ing Ca2� that has enabled researchers

to investigate changes in intracellularfree Ca2� concentrations by using flu-orescence microscopy, flow cytom-etry, and fluorescence spectroscopy.Most of these fluorescent indicatorsare derivatives of the Ca2� chelatorsEGTA, APTRA, and BAPTA that weredeveloped by Tsien and his colleagues(Tsien, 1980) and, more recently,through scientists at Molecular Probes,Inc. (Haugland and Johnson, 1999)(http://www.probes.com/handbook).

The common ratiometric dyes forCa2� are fura 2 and indo-1. These dyeschange wavelength in the presence ofdifferent concentrations of ions. Sin-gle wavelength dyes include fluo-3,rhod-2, calcium green, calcium or-ange, calcium crimson, and fura red(Nuccitelli, 1994; Haugland and John-

son, 1999). These compounds requirea fluorescent microscope with theability to record relatively fastchanges in emission wavelengths byusing enhanced video, charged cou-pled device (CCD) cameras or veryfast confocal microscopes. Many arti-cles and even books have been pub-lished on this subject; readers are di-rected to these publications for moredetails (Nuccitelli, 1994).

Immunohistochemistry

Immunohistochemistry can also beused to locate signaling proteins. Thelocation itself may indicate whetherthe protein is activated, but two addi-tional approaches may confirm thestate of the target protein. Some ven-dors are producing antibodies to the

C ACMA Blebbing B DFigure 4. Single optical sections through the basal region of epithelial tissues that were isolated without the basal lamina, then cultured incontrol media (B) or type I collagen (COL, D). The epithelia were fixed and stained with fluorescein isothiocyanate–phalloidin then viewedon the Leica CLSM. Single optical sections were obtained from the optical plane indicated by the arrows in the drawings of the cornealepithelial tissue (A,C). Individual basal cells were approximately 8–10 �m in diameter. Phalloidin also labels the F-actin associated withlateral cell membranes but not nuclei. Actin staining through the cytoplasmic blebs appears punctate in the en face optical section (B,arrowheads). D: Epithelium cultured in the presence of COL for 2 h have extensive actin bundles that line up from cell to cell (arrowheads)across the 100-�m field. A,C: Schematic drawings. Adapted with permission from (Svoboda et al., 1999b). Scale bar � 10 �m.

Figure 5. Corneal epithelia were isolated without the basal lamina, incubated with media containing fluorescein isothiocyanate–fibronectin (FITC-FN) for a short time intervals (5 to 30 min). The epithelia were rinsed, fixed, and viewed on the confocal microscope in thexz optical plane. The intensity of bound FITC-FN was recorded at the basal cell surface (intensity wedge). All settings (pinhole, laser power,voltage, and offset) were the same for all epithelia. Average relative intensity measurements were obtained for each epithelial group (n �3). The FITC-FN increased from 5 to 30 min as analyzed with NIH Image (graph). Adapted with permission from (Svoboda et al., 1999b).


signal proteins in their activated state.These antibodies have recognitionepitopes that include the phosphate orother activating conformation. Anti-bodies against active epitopes (anti-active) are becoming available formany common signal transductionproteins. If the anti-active antibody isnot available for the protein, then analternate approach is to double labelthe cells with an antibody specific forthe target protein and another anti-body that recognizes all serine, threo-nine, or tyrosine phosphorylatedamino acids (Figure 6). In this ap-proach, the overlap of the two signalsmay indicate a site where the activeprotein resides, whereas the single-la-beled protein is not activated. Thismethod is even more informative if

combined with biochemical evidence(Western blots) that the proteins arebecoming phosphorylated in the sametime frame as the translocation of theprotein (Figure 6). However, the pro-tein colocalization is not conclusive,as the resolution of the confocal micro-scope is approximately 0.18 microns.

Constructed fusion proteins

Living cells can also be observed withGFP-tagged proteins. Vectors withGFP fused to the protein of interestare transfected into cells, tissues, andtransgenic animals to track the ex-pression patterns of the protein. Cur-rently, GFPs have been modified toproduce multiple emission wave-lengths, such as cyanFP (CFP), yel-

lowFP (YFP), and DsRed (Ayoob et al.,2000; Lansford et al., 2001) so thatthey can be used in combination tolabel multiple proteins in the samecell, tissue, or animal (Hakamata etal., 2001).

Transfection with GFP-tagged fu-sion proteins is a particularly power-ful method if the GFP-tagged proteinchanges cellular location after activa-tion as discussed previously (Zamir etal., 2000; Lansford et al., 2001; Li etal., 2001). A recent study used thismethod to determine the domain ofprotein–protein interactions betweenRho and RhoGDI proteins by mutat-ing specific amino acids in the GFP-tagged constructs (Michaelson et al.,2001). This approach requires exper-tise in expression vector plasmid con-struction, a cell type that can be trans-fected with the constructs, and afluorescent microscope to view thecells. Some GFP-constructs recentlyhave become commercially available,so in the future, a laboratory set up fortissue culture and morphology may beable to simply purchase the molecularbiology component.

A new twist is the recent report ofproducing a kinase substrate con-structed with a GFP reporter. A frag-ment of myelin basic protein that con-tains a single consensus ERK/MAPkinase phosphorylation motif (PRTP,where the threonine is phosphory-lated) was fused to GFP. The fusedprotein transfected into mammaliancells acted as kinase substrate. TheGFP-MBP fused protein became phos-phorylated after serum stimulation;furthermore, a specific MEK inhibitorblocked this change in phosphoryla-tion (Mandell and Gocan, 2001).

FRET and FRAP

Two imaging techniques that were de-veloped in the 1970s and 1980s re-cently have re-emerged as usefulmethods in combination with the GFPfusion protein constructs. Fluores-cence resonance energy transfer(FRET) (Matyus, 1992) and fluores-cence recovery/redistribution afterphotobleaching (FRAP) were first de-veloped for wide field epifluorescentmicroscopes and detected with CCDcameras (Matyus, 1992; Kenworthy,2001). FRET is the nonradioactivetransfer of energy from an excited

Figure 6. Fibroblast cells cultured on plastic substrate were serum starved for 24 h and thenstimulated with fibronectin (FN). The cells were either harvested for Western blot analysis atspecific time points, (0, 30 s, 2, 5, 10, 15, and 60 min) or prepared for double-label immu-nohistochemistry at three time points: 30 s (A,D), 10 min (B,E), or 45 min (C,F). The cells wereimmunolabeled for p190RhoGAP or FAK (green) and phosphotyrosine (red) and analyzedwith confocal microscopy. The images are single optical sections through the basal regionof the cells. The yellow areas show the overlap in signal from p190 RhoGAP or FAK and ptyr,indicating the intracellular location of phosphorylated target proteins. In unstimulated cells,the p190RhoGAP and FAK distribution appeared diffuse throughout the cells. Many proteinswere phosphorylated (red or yellow) in both samples after 10 min of FN stimulation (B,E).P190RhoGAP appeared to accumulate in the cytoplasm, whereas FAK was located in cellattachment areas. The phosphorylated proteins decreased in the cells after 45 min (C,F).


state donor to a nearby acceptor (Fig-ure 7). The donor in this case is afluorescent molecule that is excited bya specific wavelength of light; the do-nor emits a higher light wavelengththat excites the fluorescent acceptormolecule. Many fluorescent moleculescan be used as donor/acceptor pairs,and many are explained in the Molec-ular Probes Web site (http://www.probes.com/handbook). Donor/accep-tor pair examples are Cy3 and Cy5 orCFP and YFP. The energy transfer isdependent on the distance betweenthe fluorescent molecules. As the ac-ceptor absorbs the donor fluores-cence, the donor will quench and itslifetime will decrease. Therefore, sev-eral parameters may be measured todetermine that energy is actually be-ing transferred including (1) increasein acceptor fluorescence, (2) decreasein donor fluorescence, (3) increase indonor fluorescence after the acceptoris photobleached.

FRET is a very powerful light mi-croscopic technique, as it will allowthe investigator to determine whethertwo proteins are within 10–70 ang-stroms of each other rather than colo-calized with confocal microscopy (200nm apart). The probes for FRET can

be either directly labeled proteins orfusion proteins that contain GFP de-rivatives (see Box 2 for a specific ex-ample) (Sorkin et al., 2000; Mochizukiet al., 2001), directly labeled primaryantibodies or secondary antibodies(Fab fragments) (Kam et al., 1995;Dictenberg et al., 1998; Kenworthy,2001), or caged molecules. Again, sev-eral good resources are available toaid the investigator with detailed

methods and protocols (Sullivan andKay, 1999; Kenworthy, 2001).

FRAP is based on the common ob-servation that if a region of tissue orcells is exposed to excitation fluores-cent wavelength light for an extendedperiod of time, the area will lose itsfluorescence or “bleach.” The cells arepreloaded with the fluorescent mole-cule of interest, then a specific regionis photobleached and the movementof the fluorescent molecules back intothe bleached area (recovery) is mea-sured as described in Box 2 experi-ment on PC12 cells (Mochizuki et al.,2001). This method allows the inves-tigator to determine the mobility anddiffusion of small molecules in the cy-toplasm of living cells or record themovement of macromolecules (RNAor large proteins, drugs) into and outof cell organelles such as the nucleus.As these movements may be rapid, itis necessary to capture images in “realtime” with CCD cameras or use smallareas and special procedures withconfocal microscopy. It is importantto record the prebleach fluorescentlevels, photobleach a selected smallarea, record the recovery at low lightlevels to prevent further bleaching,

Figure 7. The schematic representation ofthe fluorescence resonance energy transfer(FRET) spectral overlap integral. Copied withpermission from The Handbook of FluorescentProbes and Research Products by MolecularProbes, Inc. (http://www.probes.com/hand-book/). [Color figure can be viewed in theonline issue, which is available at www.inter-science.wiley.com.]

BOX 2. Examples of FRET Technology in Practice

Two different groups have recently used this FRET tech-nology to directly study signal transduction in living cells.One group (Sorkin et al., 2000) analyzed the spatial andtemporal regulation of epidermal growth factor receptor(EGFR) interactions with the SH2 domain from the adaptorprotein Grb2 in living cells. To achieve this goal, theyproduced three fused proteins. They constructed thegrowth factor receptor (GFR) fused to CFP in the cytoplas-mic domain of the receptor. In addition, they produced twoconstructs of cytoplasmic proteins fused to YFP. The cellswere stably transfected with these fused proteins, thenstimulated with the growth factor. Stimulation by thegrowth factor resulted in the recruitment of a YFP-labeledcytoplasmic adapter protein to cellular compartments thatcontained GFR-CFP as well as a large increase in the FRETsignal. In particular, FRET measurements indicated thatactivated GFR-CFP interacted with YFP-labeled adapterprotein in membrane ruffles and endosomes (Sorkin et al.,2000). This is the first report of the cytoplasmic domain ofa receptor directly interacting with adaptor and amplifyingproteins.

The second group (Mochizuki et al., 2001) used thecombination of CFP and YFP to study the spatiotemporalimages of growth factor-induced activation of Ras and

Rap1. They made numerous fused protein constructs con-sisting of the Ras-binding domain of Raf (RafRBS) and apair of YFP and CFP constructs so that intramolecularbinding of GTP-Ras to the fused protein would bring theCFP close enough to the YFP to produce energy transfer.This construct was different as both the donor and accep-tor were on the same molecule, so that activation of thesignal depended on a conformational change in the proteinthat occurs when the protein binds to a specific activatedG protein. The experiments and controls showed thatgrowth factors activated Ras at the peripheral plasmamembrane and Rap1 at the intracellular perinuclear regionof COS-1 cells. In PC12 cells, nerve growth factor-inducedactivation of Ras was initiated at the plasma membraneand transmitted to the whole cell body. By using the FRAPtechnique, they demonstrated that Ras in neurites turnedover rapidly; therefore, the sustained Ras activity in neu-rites was due to high GTP/GDP exchange rate and/or lowGTPase activity, but not to the retention of the active Ras.This is a very powerful and innovative use of this technol-ogy to establish the relationship between growth factorstimulation and the intracellular pathways stimulated uponactivation of the receptors.


then quantify the time course of re-covery for complete analysis (Phairand Misteli, 2000).

Both FRET and FRAP require thatthe protein/organelle to be studied isfluorescently labeled in either fixedpreparations or in living cells. Load-ing the plasmids, caged molecules, orother markers presents a problem thathas been overcome by a variety ofmethods, including scrape loading,lipid carriers, osmotic shock, or elec-troporation. All methods require care-ful examination and analysis of thecells combined with controls for thespecific method. For example, it wasreported recently that the carriers forlipid transfer might also produce non-specific reflectance or autofluores-cence (Guo et al., 2001). In addition,the FRET experiments will requiremany controls to establish that thechange in fluorescence of the acceptormolecule is specific (Sorkin et al.,2000; Mochizuki et al., 2001; Kenwor-thy, 2001).

In summary, the equipment and re-agents needed to viewing signal trans-duction events are light microscopeswith recording devices and softwarethat can calculate and display ratioimaging wavelength emission levelsor changes in emission over time. Theindicators for some cellular changes,including calcium and pH, can be pur-chased as esters that easily cross cellmembranes. These indicators are sen-sitive to temperature and oxygenchanges and may need special tem-perature controlled environments foroptimum data collection. They mayalso require special microscope acces-sories. The GFP-labeled proteins re-quire the transfer of the DNA into thecell and not only successful expres-sion but also that the proteins are inthe correct conformation to interactwith other proteins and cellular ele-ments. Immunohistochemistry re-quires that the cells be fixed beforelabeling, resulting in the disadvantagethat the cells are dead. Thus even atemporal sequence may not tell thewhole story. New anti-active antibod-ies have made it possible to locateonly the activated forms of some sig-nal proteins. However, double-label-ing experiments will allow the visual-ization of additional proteins thatanti-active antibodies are not yetavailable. It is important to note that

not all antibodies work for immuno-histochemistry as some have been de-signed primarily for biochemicalanalysis by means of Western blots orimmunoprecipitation.

BIOCHEMICAL ANALYSIS OFSIGNAL TRANSDUCTION

As we were primarily trained in mor-phologic techniques, it was relativelyeasy to add biochemical and molecu-lar biological experiments to ourprojects. Basically, if you can cook,then you can do biochemistry, as itgenerally involves following previ-ously identified protocols and recipes.Many of the assays and proceduresare being packaged into kits with step-by-step instructions, opening up theprocedures to a wider audience. Thebasic procedures will be briefly de-scribed; however, it is recommendedthat a good laboratory protocol man-ual be consulted for further detailsand instructions (e.g., the Current Pro-tocols series; www.currentprotocols.com).

Western Blots

Traditionally, signal transductionevents have been described by usingbiochemical analysis. For example, alltyrosine-phosphorylated proteins canbe detected with an anti-phosphoty-rosine antibody on a Western blot ofcell lysates obtained after stimulationin a temporal sequence (Figure 8A).Alternatively, antibodies specific forother phosphorylated amino acids(serine or threonine) could also be an-alyzed. The cells were harvested atspecific times and placed into a lysisor SDS sample buffer that containsprotease and phosphatase inhibitors.The proteins were separated on a minigel electrophoresis apparatus (Bio-Rad) and then transferred to a mem-brane (Millipore). Once the proteinsare on the membrane, detecting thebound proteins is very similar to im-munohistochemistry. After appropri-ate washing and blocking steps, themembranes were incubated with pri-mary antibodies (in this case, an anti-body that recognizes all phosphory-lated tyrosine residues). Themembrane was washed and incubatedwith a secondary antibody conjugatedto horseradish peroxidase. The la-beled antibodies were detected with

chemiluminescent exposure to an x-ray film (Figure 8) or an imaging cam-era (Kodak Imaging Station). The ad-vantage of this type of blot is that itindicates all of the proteins that maybe tyrosine phosphorylated and candocument how individual proteinsmay increase or decrease in signal.This blot does not identify the specificproteins.

To identify the individual proteins,either this same blot can be strippedand reprobed with another primaryantibody (Figure 8B), or a sister blotcan be probed with the specific anti-body. Unfortunately, those sequentialWestern blots may not be conclusive;therefore, the current recommendedprocedure is to either obtain an anti-body that only recognizes the activatedprotein (Figure 9) or immunoprecipi-tate all tyrosine phosphorylated pro-teins, and then do a Western blot for thetarget protein (Figure 8C). Alterna-tively, the target protein can be immu-noprecipitated then all tyrosine-phos-phorylated proteins determined witha Western blot (Figure 8D). This ap-proach has the advantage that, if thetarget protein is bound to other pro-teins in a complex, multiple proteinsand their interactions may be identi-fied.

The equipment needed for Westernblots includes the electrophoresis ap-paratus and transfer equipment. Werecommend getting a mini gel, as it ismore economical, uses fewer re-agents, and requires less cellular pro-tein. We also purchase the prepouredgels in the appropriate concentrationof acrylamide. Although it is moreeconomical to pour the gels, the re-agents are toxic and necessary precau-tions need to be taken. In addition,technical time and consistency fromgel to gel are also factors for consid-eration.

Immunoprecipitation

Immunoprecipitation is relativelyeasy and requires the same equipmentas Western blots. Beads made from avariety of substances, including met-als, are purchased with secondary an-tibodies attached (available from a va-riety of vendors including Pierce,BioRad, Dynal). In the laboratory, wecross-link primary antibodies to a spe-cific protein or all tyrosine-phospho-


rylated proteins to the beads. The cellsare lysed in buffer containing pro-tease inhibitors and then incubatedwith the antibody-coated beads. Weprefer to use the beads that can becollected with a magnet, as they aremore efficient in our experience. Theproteins are separated by using SDSelectrophoresis, and then the proteinsare identified by using the proceduresdescribed for Western blots. However,the protein–protein interactions dem-onstrated by immunoprecipitationand Western blots from cell lysatemay not reveal the actual situation invivo and more experiments are neces-sary to demonstrate specific interac-tions.

GST Binding or “Pull-Down”Assays

The glutathione S-transferase (GST)binding or “pull-down” assay is simi-lar to immunoprecipitation, except itdetermines direct protein–protein in-teractions. Fragments of proteins orsmall whole proteins are produced ina bacterial expression system with aGST tag (Cytoskeleton, Upstate Bio-technology). In our experiments, wehave purchased GST-labeled proteinssuch as RhoGDP, RhoGTP, the Rhobinding domain of rhotectin (RBD-GST) that selectively bind RhoGTP, andappropriate GST controls. In experi-ments using RBD-GST, it was demon-strated that the corneal epithelial cellshave a biphasic response to collagen(Figure 10). A decrease in RhoGTP after15 min of collagen stimulation was fol-lowed by an increase at 30 min (Figure10) similar to the response reported byRen and coworkers in endothelial cells(Ren et al., 1999, 2000).

In another set of experiments, therelationship between Rho andp190Rho-GAP, GST-RhoGDP, andGST-RhoGTP was determined by us-ing affinity precipitations followed byWestern blots for p190Rho-GAP (GT-Pase Activating Protein) (Figure 11).Control GST beads precipitated asmall amount of p190Rho-GAP (Fig-ure 10). Cell lysates from unstimu-lated corneal epithelia bound equalamounts of p190Rho-GAP with inac-tive or active GST-Rho beads. In con-trast, in epithelia stimulated withLPA, the GST-RhoGTP beads boundthreefold more p190Rho-GAP than

the RhoGTP beads mixed with cell ly-sates from epithelia that were notstimulated with LPA (Figure 11). Inaddition, an unidentified lower molec-ular weight protein cross-reacted withthe p190Rho-GAP antibody alsobound to the GST-RhoGTP beads.

The GST fusion proteins can bemade with just those protein domainsthat may interact with other specificproteins. This property of the systemmakes it a very powerful method tostudy blots; immunoprecipitation andGST pull-down blots are a commonway to compare proteins in differentexperimental treatment groups. It isdifficult to compare blots that wereproduced on different days, becausethere are many variables that contrib-ute to the density of the protein bandsproduced by the procedure; however,it is fair to compare the intensity ofdifferent protein bands on the sameblot if the gel was loaded with equalamounts of total protein (Figure 11,

graphs). To overcome the variabilityfrom experiment to experiment, du-plicate or triplicate samples from eachtreatment group could be processed atthe same time, or a control group(A4331 cell lysate, Figure 9) can beadded as a standard. Many vendorsprovide a control cell lysate with anti-bodies as a courtesy or for a nominalcost.

Activity Assays

Another classic way to determinewhether a particular kinase is active isto expose the cell lysate to a knownsubstrate for the enzyme in the pres-ence of radioactive phosphate. Theproducts are separated by electro-phoresis (with or without immuno-precipitation), then the gel is exposedto x-ray film to determine whether theproteins incorporated the isotope.

Recently, some substrates for en-zymes that cleave specific sequences

Figure 8. Several proteins were phosphory-lated on tyrosine residues after epitheliawere incubated in collagen (COL) for 15min. A:Corneal epithelia (n � 20–30/group)were isolated without basal lamina and in-cubated with medium containing COL forvarious time intervals (0–60 min). The epi-thelia were removed from culture andplaced directly into Western blot samplebuffer. Tyrosine phosphorylated proteinswere detected with antibodies to phos-photyrosine (PY-20). Several proteins (mo-lecular mass 125, 60–70, and 40–45 and 30kDa) appeared to increase or decreasephosphorylation with time. The phospho-proteins were identified by several meth-ods: reprobing the same phosphotyrosineblot, separate parallel Western blots B, andimmunoprecipitation C. B: Samples of theindividual Westerns for each protein dem-onstrate that a �220-kDa protein was iden-tified as tensin, a 190-kDa protein was p190Rho-GAP, a 125- to 135-kDa protein wasfocal adhesion kinase (FAK), and a 60- to70-kDa protein was paxillin. C: All phospho-proteins were immunoprecipitated fromepithelia isolated without basal lamina withPY20 after COL stimulation for 1 h (�) com-pared with controls (-). The proteins wereseparated on 7.5% SDS-PAGE gels, and an-alyzed for tensin, p190RhoGAP, FAK andpaxillin (PAX). Tensin, FAK and paxillin co-immunoprecipitated with PY-20, however,p190RhoGAP did not, indicating that theepitope was masked or the protein was notphosphorylated. Adapted with permissionfrom (Svoboda et al., 1999b).


have also been developed to testwhether the enzyme is active in livingcells. Apoptosis, also known as pro-grammed cell death, causes suscepti-ble cells to undergo a series of enzy-matic and morphologic changes.During apoptosis, several signal trans-duction pathways and specific degrada-tive enzymes become activated. Thedegradative enzymes are in the caspasefamily of cysteinyl proteases, alsoknown as ICE-like, because they resem-ble the first member described, inter-leukin-1 converting enzyme (ICE).These enzymes may cleave essentialstructural components of the cell, in-cluding actin cytoskeletal elements, nu-clear lamins, and small nucleoproteins(Whyte, 1996; De Laurenzi and Melino,2000). These changes eventually lead tonuclear material condensing as the nu-clear DNA breaks down and cellularmaterial pinches off, forming apoptoticbodies.

Several companies have developeda class of fluorogenic protease sub-strates to determine whether the cellscontain active caspase-3. One of thecaspase-3 substrates (PhiPhiLux, On-coImmunin, Inc.) contains thecaspase-3 recognition sequence aspar-tic acid-glutamic acid-valine-aparticacid (DEVD) in a bifluorophore-deriv-itized peptide that mimics the struc-tural loop conformation present in the

native protease cleavage sites of glob-ular proteins. The compound is notfluorescent unless it is cleaved by en-dogenous caspase-3. Cells used in anexperiment are incubated for the lasthour in the presence of the caspase-3substrate PhiPhiLux then viewed witheither confocal or wide-field fluores-cent microscopes. Cells that do notcontain activated caspase-3 do nothave any fluorescence, whereas cellsthat do have activated caspase-3 willemit light in the wavelength of thesubstrate. These experiments can beconfirmed with a quantitative bindingassay for activated caspase-3(CaspACE Assay, Promega). To testwhether the reaction is specific forcaspase-3 activity, some tissues can bepretreated with the caspase-3 inhibi-tor Z-VAD-FMK.

Using a Combination ofWestern blots,Immunoprecipitation, and GSTPull-Down Assays

Sometimes proteins are particularlydifficult to analyze. A protein that wehave been investigating, p190RhoGAPcannot be immunoprecipitated withanti-phosphotyrosine (Figure 8C), butwe have demonstrated that it is ex-pressed in our model, and a 190-kDaprotein becomes activated by tyrosinephosphorylation very quickly afterECM or LPA stimulation (Figure 11).We have used the combination ofWestern blots demonstrating a molec-ular weight shift and GST pull-downassays to establish that this proteinchanges in response to ECM stimula-tion. The following section describesthe data from these experiments inmore detail to demonstrate how dif-ferent experiments provide evidencefor protein-protein interactions dur-ing signal transduction events.

To understand the timing of this ex-periment, it is important to know thatepithelial sheets require interactionswith soluble ECM molecules for aminimum of 15 min for ACM reorga-nization to occur in 2 h (Svoboda etal., 1999b). In addition, ECM mole-cules bind to the basal cell surface andimmediately initiate tyrosine phosphor-ylation signaling events as shown by ananti-phosphotyrosine Western blot(Figure 8A). One of the proteins thatbecame phosphorylated very early inresponse to ECM was p190RhoGAP(Svoboda et al., 1999b). However, at-tempts to immunoprecipitate the phos-phorylated form of this protein failed,because when it became phosphory-lated, it bound to other proteins in acomplex that masked the phosphory-lated tyrosines. To further investigatethe role of this protein in the signal cas-

Figure 9. Western blot analysis of epithelial extracts cultured with or without collagen (COL)(no treatment, NT, 2 h) to detect MAP kinase (erk-1 and erk-2) activity (A), total erk-2 (B),total MEK-2 (C), with or without pretreatment (24 h) with the MEK inhibitor, PD98059 (10 �M,50 �M, 100 �M). All inhibitor-treatment groups were COL stimulated for 2 h without inhibitor.A431 fibroblasts served as the positive tissue control. Activated erk-1 or erk-2 (denoted byasterisks) were detected with an antibody that only binds to phosphorylated erk-1 and 2(A). Phosphorylated erk-1 (44 kDa) and erk-2 (42 kDa) increased over time with COLstimulation in control and COL-stimulated epithelia (A). Pretreating the epithelia with theinhibitor did not appear to effect erk-2, but erk-1 appeared to decrease when comparedwith collagen-treated controls. Total erk-2 and MEK were determined by reprobing thesame membrane (B,C). The 42-kDa protein was erk-2, whereas the 44-kDa protein wasMEK-2 (C). The total protein levels of erk-2 and MEK-2 appeared constant in control andCOL-stimulated epithelia (C). Adapted with permission from Chu et al., 2000.

Figure 10. Glutathione S-transferase pre-cipitation using Rho binding domain ofrhotectin (Upstate Biotechnology). Twoconcentrations of collagen (COL) weretested (50 and 100 �g/ml) on epitheliaisolated without basal lamina and stim-ulated for 15 or 30 min compared withlysates from tissues incubated withoutcollagen (NT, no treatment). We foundthat there was a transient decrease inRhoGTP at 15 min, but by 30 min, thetissues treated with 50 �g/ml had a sig-nificant increase in RhoGTP.


cade, we needed to establish whetherp190RhoGAP changed protein levels ormolecular weight in response to ECM.Extracts from epithelial sheets cul-tured in the presence of collagen(COL) or fibronectin (FN) over vari-ous times (2–60 min) were analyzedfor p190RhoGAP to detect changes inprotein levels by using Western blottechniques (Figure 11A–C). The cellextracts were electrophoretically sep-arated on a 7.5% SDS gel overnight(25 V) to show molecular mass dif-ferences in p190RhoGAP. We found

that, with increasing exposure toCOL, there was a correspondingchange in molecular mass inp190Rho-GAP (Figure 11B,C). A highmolecular weight shift in p190Rho-GAP from COL-stimulated epitheliawas observed as early as 5 min thatbecame more prominent by 30 anddecreased by 60 min. The molecularweight shift peaked at 30 min for ep-ithelia treated with FN and COL andcould be illustrated by measuring therelative intensity of the higher molec-ular weight band (Figure 11, graph),

which increased in intensity at 30 minin the presence of FN and COL. Com-parison of the ratio of the high molec-ular weight band/lower molecularweight band could also be used to de-scribe the relationship between thetwo forms of the protein to show thatthe higher molecular weight form isprominent at 30 min.

To further analyze the relationshipbetween Rho and p190RhoGAP, weperformed GST-RhoGDP and GST-RhoGTP affinity precipitations fol-lowed by Western blots for

Figure 11. Western blot analysis of total protein of collagen (COL) -stimulated epithelia over various time points (NT [no treatment], 2, 5, 10,15, 60, min) to detect FAK (A) and p190RhoGAP (B,C). FAK was present as a single band in control epithelia (NT; A). However, a doubletwas detected by 5 min. The molecular weight shift was prominent in FAK by 10 min. After 15 min of COL stimulation, FAK appeared as adoublet of two protein isoforms. By 60 min, most of the detectable protein was present in the phosphorylated form (A). p190RhoGAP hada molecular weight shift in COL (B) and fibronectin (FN) (C) (graph) stimulated epithelia by 5 min and became more prominent by 10 minin the FN-stimulated cells and nearly completely in the higher molecular weight band by 30 min. The COL-stimulated epithelia maintaineda doublet through 60 min (B). D: Glutathione S-transferase precipitation of Rho binding proteins (D) by using GST pull-down assays. Celllysates were added in equal volumes to bovine serum albumin blocked GST beads (10 �l) conjugated to RhoGDP, RhoGTP, or GST controls.Specific proteins were identified with Western blots. Cell lysates from lysophosphatidic acid–stimulated epithelia precipitated more p190RhoGAP with RhoGTP than either RhoGDP or unstimulated cell lysates with RhoGTP (graph). ECM, extracellular matrix.


p190RhoGAP. These experiments re-quired several treatment groups, in-cluding control GST-coated beads, theinactive form of the protein (Rho-GDP), and an active form of the pro-tein (Rho-GTP). In addition, cell ly-sates from experimental (e.g.,lysophosphatidic acid [LPA] treated)and control conditions needed to becompared. The results demonstratedthat control GST beads precipitated asmall amount of p190Rho-GAP (Fig-ure 11D and densitometry graph). Celllysates from unstimulated corneal ep-ithelia bound equal amounts ofp190Rho-GAP with inactive or activeGST-Rho beads. In contrast, in epithe-lia stimulated with the direct Rhostimulator LPA, the activated GST-Rho beads bound more p190Rho-GAPthan the RhoGDP beads. In addition,an unidentified lower molecularweight protein also bound to the GST-RhoGTP beads. This blot was re-probed with an antibody specific forphosphotyrosine, and a small percent-age of the p190Rho-GAP proteinpulled down by Rho-GTP in the LPAstimulated tissue was tyrosine phos-phorylated.

These experiments gave us severalpieces of information about the be-havior of this protein in our cornealepithelial system. First, p190RhoGAPincreased in molecular weight in re-sponse to ECM. Second, RhoGTPbinds more p190RhoGAP in the pres-ence of LPA than in unstimulated ep-ithelia. Third, some of the phosphor-ylated p190RhoGAP may bind toRhoGTP. These data provide furtherevidence that the Rho signal pathwaymay have a significant role in ECM-and LPA-induced actin reorganiza-tion.

Clearly the biochemical assays arevery powerful in the analysis of signaltransduction events. With these meth-ods, different protein phosphorylationstates can be assessed either by usingspecific anti-active antibodies or com-bining immunoprecipitation withWestern blots. In addition, the GSTpull-down assays allow the analysis ofspecific protein–protein interactions.These approaches provide a means toquantify intracellular changes in re-sponse to a specific stimulus. With theaddition of specific inhibitors, investi-gators can determine the sequence ofevents. However, the weakness of

these approaches is that the cells haveto be destroyed by homogenization toproduce the cell lysates. This require-ment necessarily causes the loss of in-tracellular relationships and the tem-poral sequence of interactions, as thetissues have to be harvested at finitetimes. The biochemical analysis re-quires more tissue and protein thanmany morphologic approaches. In ad-dition, the laboratory has to be prop-erly equipped with the gel apparatusand transfer equipment. Often, the re-agents (antibodies, GST proteins, ac-tivity kits, and chemiluminescent ma-terials) are expensive.

CONCLUSION

Future experiments designed to dis-sect signal transduction pathways andtheir morphologic effects will requirea combination of all of these ap-proaches and surely new ones. Thereemergence of FRET and FRAP withcustom fusion proteins will revolu-tionize how signal transductionevents are understood, as living cellswill be observed. This type of experi-ment requires the cooperation of cellbiologists, molecular biologists, andmorphologists to design the experi-ments, provide the constructs, and vi-sualize the signal transduction events.In the near future, procedures will beperfected to transfect engineered pro-teins into embryonic tissues, possiblywith electroporation, to view signaltransduction events in whole tissues.

ACKNOWLEDGMENTS

Many students, technicians, and jour-nal club groups have contributed toour understanding of signal transduc-tion events. We thank Dan Orlow andChia L. Chu for providing Westernblots and immunoprecipitation dataand Jesus Acevedo, Petra Moessnerand Tamara Field for GST pull-downdata. We also thank my current signaltransduction journal club for criti-cally reading this manuscript. NIHEY08886 has supported this work for10 years.

LITERATURE CITED

Ayoob J, Shaner N, Sanger J, Sanger J.2000. Expression of green or red fluores-cent protein (GFP or DsRed) linked pro-

teins in nonmuscle and muscle cells. MolBiotechnol 17:65–71.

Berrier AL, Mastrangelo AM, Downward J,Ginsberg M, LaFlamme S. 2000. Acti-vated R-Ras, Rac1, PI 3-Kinase and PKC[small element of] can each restore cellspreading inhibited by isolated integrinb1 cytoplasmic domains. J Cell Biol 151:1549–1560.

Bokoch GM. 2000. Regulation of cell func-tion by Rho family GTPases. ImmunolRes 21:139–148.

Boudreau N, Bissell MJ. 1998. Extracellu-lar matrix signaling: Integration of formand function in normal and malignantcells. Curr Opin Cell Biol 10:640–646.

Cary LA, Guan JL. 1999. Focal adhesionkinase in integrin-mediated signaling.Front Biosci 4:D102–D113.

Chapman HA, Wei Y, Simon DI, Waltz DA.1999. Role of urokinase receptor andcaveolin in regulation of integrin signal-ing. Thromb Haemost 82:291–297.

Chen Q, Kinch MS, Lin TH, Burridge K,Juliano RL. 1994. Integrin-mediated celladhesion activates mitogen-activatedprotein kinases. J Biol Chem 269:26602–26605.

Chu CL, Reenstra RW, Orlow DL, SvobodaKKH. 2000. Erk and PI3 kinase are neces-sary for collagen binding and actin reor-ganization in avian corneal epithelia. In-vest Ophthalmol Vis Sci 41:3374–3382.

Couchman JR, Woods A. 1999. Syndecan-4and integrins: Combinatorial signaling incell adhesion. J Cell Sci 112:3415–3420.

De Laurenzi V, Melino G. 2000. Apoptosis.The little devil of death. Nature 406:135–136.

Dictenberg JB, Zimmerman W, Sparks CA,Young A, Vidair C, Zheng Y, CarringtonW, Fay FS, Doxsey SJ. 1998. Pericentrinand gamma-tubulin form a protein com-plex and are organized into a novel lat-tice at the centrosome. J Cell Biol 141:163–174.

Duong LT, Lakkakorpi P, Nakamura I, Ro-dan GA. 2000. Integrins and signaling inosteoclast function. Matrix Biol 19:97–105.

Frisch SM. 1999. Evidence for a functionof death-receptor-related, death-do-main-containing proteins in anoikis.Curr Biol 9:1047–1049.

Garrington TP, Johnson GL. 1999. Organi-zation and regulation of mitogen-acti-vated protein kinase signaling pathways.Curr Opin Cell Biol 11:211–218.

Giancotti FG, Ruoslahti E. 1999. Integrinsignaling. Science 285:1028–1032.

Giorgio L, Hemmati-Brivanlou A. 1999. Amolecular basis for Smad specificity.Dev Dyn 214:269–277.

Gonzalez-Amaro R, Sanchez-Madrid F.1999. Cell adhesion molecules: Selectinsand integrins. Crit Rev Immunol19:389–429.

Guo B, Pearce A, Traulsen K, Rintala A,Lee H. 2001. Fluorescence produced bytransfection reagents can be confusedwith green fluorescent proteins in mam-malian cells. Biotechniques 31:314–321.

Hakamata Y, Tahara K, Uchida H, SakumaY, Nakamura M, Kume A, Murakami T,


Takahashi M, Takahashi R, HirabayashiM, Ueda M, Miyoshi I, Kasai N, Koba-yashi E. 2001. Green fluorescent protein-transgenic rat: A tool for organ trans-plantation research. Biochem BiophysRes Commun 286:779–785.

Hanks SK, Calalb MB, Harper MC, PatelSK. 1992. Focal adhesion protein-ty-rosine kinase phosphorylated in re-sponse to cell attachment to fibronectin.Proc Natl Acad Sci U S A 89:8487–8491.

Haugland RP, Johnson ID. 1999. Intracel-lular ion indicators. In: Fluorescent andluminescent probes for biological activ-ity. Mason WT. Academic Press, Cam-bridge: 40–50.

Hirsch M, Chang K, Kao WW-Y, SvobodaKKH. 1996. Intracellular distribution oftype II collagen mRNA and prolyl 4-hy-droxylase in embryonic Avian cornealepithelia. Anat Rec 244:1–14.

Humphries MJ. 2000. Integrin structure.Biochem Soc Trans 28:311–339.

Hunter T. 2000. Signaling-2000 and be-yond. Cell 100:113–127.

Kam Z, Volberg T, Geiger B. 1995. Map-ping of adherens junction componentsusing microscopic resonance energytransfer imaging. J Cell Sci 108:1051–1062.

Kenworthy AK. 2001. Imaging protein-pro-tein interactions using fluorescence res-onance energy transfer microscopy.Methods 24:289–296.

Khoory W, Wu E, Svoboda KK. 1993. In-tracellular relationship between actinand alpha-actinin in a whole corneal ep-ithelial tissue. J Cell Sci 106:703–717.

Kortenjann M, Shaw PE. 1995. The growingfamily of MAP kinases: Regulation andspecificity. Crit Rev Oncog 6:99–115.

Lansford R, Bearman G, Fraser SE. 2001.Resolution of multiple green fluorescentprotein color variants and dyes using two-photon microscopy and imaging spectros-copy. J Biomed Optics 6:311–318.

Li HY, Kotaka M, Kostin S, Lee SM, KokLD, Chan KK, Tsui SK, Schaper J, Zim-mermann R, Lee CY, Fung KP, WayeMM. 2001. Translocation of a human fo-cal adhesion LIM-only protein, FHL2,during myofibrillogenesis and identifica-tion of LIM2 as the principal determi-nants of FHL2 focal adhesion localiza-tion. Cell Motil Cytoskel 48:11–23.

Lodish H, Berk A, Zipursky SL, MatsudairaP, Baltimore D, Darnell J. 2000. Molecu-lar cell biology. New York, NY: W.H.Freeman and Co.

Mandell JW, Gocan NC. 2001. A green flu-orescent protein kinase substrate allow-ing detection and localization of intra-cellular ERK/MAP kinase activity. AnalBiochem 293:264–268.

Massague J. 2000. How cells read TGF-beta signals. Nat Rev Mol Cell Biol1:169–178.

Matyus L. 1992. Fluorescence resonanceenergy transfer measurements on cellsurfaces. A spectroscopic tool for deter-mining protein interactions. J Photo-chem Photobiol B 12:323–337.

Mecham RP, editor. 1991. Laminin recep-tors. Annual Review of Cell Biology. PaloAlto, CA: Annual Reviews Inc.

Meredith JE, Schwartz MA. 1997. Inte-grins, adhesion and apoptosis. TrendsCell Biol 7:146–150.

Meredith JE Jr, Winitz S, Lewis JM, HessS, Ren XD, Renshaw MW, Schwartz MA.1996. The regulation of growth and in-tracellular signaling by integrins. En-docr Rev 17:207–220.

Michaelson D, Silletti J, Murphy G,D’Eustachio P, Rush M, Philips MR.2001. Differential localization of RhoGRPases in live cells: Regulation by hy-pervariable regions and RhoGDI bind-ing. J Cell Biol 152:111–126.

Mochizuki N, Yamashita S, Kurokawa K,Ohba Y, Nagai T, Miyawaki A, MatsudaM. 2001. Spatio-temporal images ofgrowth-factor-induced activation of Rasand Rap1. Nature 411:1065–1067.

Mooney DJ, Langer R, Ingber DE. 1995.Cytoskeletal filament assembly and thecontrol of cell spreading and function byextracellular matrix. J Cell Sci 108:2311–2320.

Nobes CD, Hall A. 1999. Rho GTPases con-trol polarity, protrusion, and adhesionduring cell movement. J Cell Biol 144:1235–1244.

Nuccitelli R, editor. 1994. A practical guideto the study of calcium in living cells.San Diego: Academic Press.

O’Neill GM, Fashena SJ, Golemis EA.2000. Integrin signaling: A new Cas(t) ofcharacters enters the stage. Trends CellBiol 10:111–119.

Ossowski L, Aguirre-Ghiso JA. 2000.Urokinase receptor and integrin partner-ship: Coordination of signaling for celladhesion, migration and growth. CurrOpin Cell Biol 12:613–620.

Phair RD, Misteli T. 2000. High mobility ofproteins in the mammalian cell nucleus.Nature 404:604–609.

Porter JC, Hogg N. 1998. Integrins takepartners: Cross-talk between integrinsand other membrane receptors. TrendsCell Biol 8:390–396.

Read PW, Liu X, Longenecker K, DipierroCG, Walker LA, Somlyo AV, Somlyo AP,Nakamoto RK. 2000. Human RhoA/RhoGDI complex expressed in yeast:GTP exchange is sufficient for transloca-tion of RhoA to liposomes. Protein Sci9:376–386.

Ren X-D, Kiosses WB, Schwartz MA. 1999.Regulation of the small GTP-bindingprotein Rho by cell adhesion and thecytoskeleton. EMBO J 18:578–585.

Ren X-D, Kiosses WB, Sieg DJ, Otey CA,Schlaepfer DD, Schwartz MA. 2000. Fo-cal adhesion kinase suppresses Rho ac-tivity to promote focal adhesion turn-over. J Cell Sci 113:3673–3678.

Richardson TP, Trinkaus-Randall V, Nu-gent MA. 1999. Regulation of basic fibro-blast growth factor binding and activityby cell density and heparan sulfate.J Biol Chem 274:13534–13540.

Ridley A. 2000a. Rho GTPases. Integratingintegrin signaling. J Cell Biol 150:F107–F109.

Ridley A. 2000b. What initiates actin poly-merization? Genome Biol 1:Reviews102.

Robinson MJ, Cobb MH. 1997. Mitogen-activated protein kinase pathways. CurrOpin Cell Biol 9:180–186.

Roovers K, Assoian RK. 2000. Integratingthe MAP kinase signal into the G1 phasecell cycle machinery. Bioessays 22:818–826.

Sastry SK, Burridge K. 2000. Focal adhe-sions: A nexus for intracellular signalingand cytoskeletal dynamics. Exp Cell Res261:25–36.

Schiffer M, von Gersdorff G, Bitzer M,Susztak K, Bottinger EP. 2000. Smadproteins and transforming growth fac-tor-beta signaling. Kidney Int Suppl 77:S45–S52.

Schwartz MA, Baron V. 1999. Interactionsbetween mitogenic stimuli, or a thou-sand and one connections. Curr OpinCell Biol 11:197–202.

Schwartz MA, Shattil SJ. 2000. Signalingnetworks linking integrins and rho fam-ily GTPases. Trends Biochem Sci25:388–391.

Settleman J. 2000. Getting in shape withRho. Nat Cell Biol 2:E7–E9.

Song QH, Singh RP, Richardson TP, Nu-gent MA, Trinkaus-Randall V. 2000.Transforming growth factor-beta1 ex-pression in cultured corneal fibroblastsin response to injury. J Cell Biochem77:186–199.

Sorkin A, McClure M, Huang F, Carter R.2000. Interaction of EGF receptor andgrb2 in living cells visualized by fluores-cence resonance energy transfer (FRET)microscopy. Curr Biol 10:1395–1398.

Sugrue SP, Hay ED. 1981. Response ofbasal epithelial cell surface and cytoskel-eton to solubilized extracellular matrixmolecules. J Cell Biol 91:45–54.

Sugrue SP, Hay ED. 1986. The identifica-tion of extracellular matrix (ECM) bind-ing sites on the basal surface of embry-onic corneal epithelium and the effect ofECM binding on epithelial collagen pro-duction. J Cell Biol 102:1907–1916.

Sullivan K, Kay S, editors. 1999. Methodsin cell biology: Green fluorescent pro-teins. Methods in cell biology. San Di-ego, CA: Academic Press.

Svoboda KKH. 1992. Embryonic cornealepithelial actin alters distribution in re-sponse to laminin. Invest OphthalmolVis Sci 33:324–333.

Svoboda KKH, Hay ED. 1987. Embry-onic corneal epithelial interaction withexogenous laminin and basal lamina isF-actin dependent. Dev Biol 123:455–469.

Svoboda KKH, O’Shea KS. 1984. Optic ves-icle defects induced by vincristine sul-fate: An in vivo and in vitro study in themouse embryo. Teratology 29:223–240.

Svoboda KKH, O’Shea KS. 1987. An anal-ysis of cell shape and the neuroepithelialbasal lamina during optic vesicle forma-tion in the mouse embryo. Development100:185–200.

Svoboda KKH, Orlow DL, Ashrafzadeh A,Jirawuthiworavong G. 1999a. Zyxin andvinculin distribution at the “cell-extra-


cellular matrix attachment complex”(CMAX) in corneal epithelial tissue areactin dependent. Anat Rec 254:336–347.

Svoboda KKH, Orlow DL, Chu CL, Reen-stra WR. 1999b. ECM stimulated actinbundle formation in embryonic cornealepithelia is tyrosine phosphorylation de-pendent. Anat Rec 254:348–359.

Symons M, Settleman J. 2000. Rho familyGTPases: More than simple switches.Trends Cell Biol 10:415–419.

Treisman R. 1996. Regulation of transcrip-tion by MAP kinase cascades. Curr OpinCell Biol 8:205–215.

Tsien R. 1980. New calcium indicators andbuffers with high selectivity againstmagnesium and protons: Design, synthe-sis, and properties of prototype struc-tures. Biochemistry 19:2396–2404.

Turner CE. 2000. Paxillin and focal adhe-sion signalling. Nat Cell Biol 2:E231–E236.

Wayner EA, Orlando RA, Cheresh DA.1991. Integrin avb3 and integrin avb5contribute to cell attachment to vitro-nectin but differentially distribute on thecell surface. J Cell Biol 113:919–929.

Whyte M. 1996. ICE/CED-3 proteases inapoptosis. Trends Cell Biol 6:245–248.

Yeh B, Svoboda KKH. 1994. Intracellulardistribution of beta-actin mRNA is po-larized in embryonic corneal epithelia.J Cell Sci 107:105–115.

Zamir E, Katz M, Posen Y, Erez N,Yamada KM, Katz BZ, Lin S, Lin DC,Bershadsky A, Kam Z, Geiger B. 2000.Dynamics and segregation of cell–matrixadhesions in cultured fibroblasts. NatCell Biol 2:191–196.

Zimmerman CM, Padgett RW. 2000.Transforming growth factor beta signal-ing mediators and modulators. Gene249:17–30.


Approaches To Studying Cellular Signaling: A Primer For ......boda has used several whole tissue...

Documents

Transcript of Approaches To Studying Cellular Signaling: A Primer For ......boda has used several whole tissue...