Green Fluorescent Protein: A Novel Viability Assay for Cryobiological Applications

ity

lehinsedors’tedere

idewithedectedotaythat

Cryobiology40, 360–369 (2000)doi:10.1006/cryo.2000.2258, available online at http://www.idealibrary.com on

Green Fluorescent Protein: A Novel Viability Assayfor Cryobiological Applications

Gloria Elliott,* John McGrath,* and Elahe Crockett-Torabi†*Department of Mechanical Engineering and†Department of Surgery, Michigan State University,

East Lansing, Michigan 48824 U.S.A.

Assessment of tissue viability following the application of a freezing protocol is challenging due to the paucof viability assays that can be used dynamically,in situ.Cells transfected with a green fluorescent protein (GFP)vector actively produce GFP, which is retained intracellularly. Because of its constitutive and heritabexpression, GFP fluorescence of transfected cells may have significant utility as a viability assay for cells wittissues. As a first step toward application to tissues, this work seeks to establish the validity of this GFP-baassay in cell suspensions by comparing the results to other accepted measures of viability. To the authknowledge, this is the first use of GFP in cryobiology applications. Intracellular GFP fluorescence was evaluafollowing slow freezing. Nontransfected and GFP-transfected rat 3230 adenocarcinoma (R3230AC) cells wfrozen at 1°C/min to minimum temperatures between25 and230°C and then immediately thawed in a 37°Cwater bath. Samples were assayed using the common viability indicators trypan blue and ethidium brom(EtBr). A regression analysis of recovery measured with the GFP assay as a function of recovery measureda trypan blue assay gave a correlation coefficient of 0.97. A similar correlation coefficient, 0.95, was determinfor recovery assessed by the GFP assay as a function of recovery measured by an EtBr assay. Nontransfand GFP-transfected cells responded similarly to slow freezing, indicating that GFP transfection did nsignificantly alter the response of cells to typical freezing conditions. The excellent correlation of GFP assresults with those of two common viability assays suggests that the GFP-based assay is valid for cells andfurther development of a tissue viability assay based on transfection is appropriate.© 2000 Academic Press

Key Words:green fluorescent protein; viability.

Evaluation of the viability of individual cells,fte

yo-Alceeryrelto

e iia-onetiobeto

x-tioen

or an identical fresh untreated sample.” Thereions.g.,at

umerefor

ea-

ia-d tod inthesed.singnif-

uc-

0.ES

as well as cells within tissues and organs, afreezing is of fundamental importance for crpreservation and cryosurgery research.though cryopreservation seeks to sustainand tissue viability and conversely, cryosurgseeks to devitalize neoplastic tissue, bothon some measurable function of viabilitydetermine protocol success. Although therno universally agreed upon definition for vbility that suits all contexts, and perhaps npossible, the authors embrace the convenadopted by Pegg (24). That is, viability canconsidered “the ability of a treated sampleexhibit a specific function or functions, epressed as a proportion of the same funcexhibited by the same sample before treatm

Received February 15, 2000; accepted May 19, 200This research was supported by NSF Grant B

9816734.

3600011-2240/00 $35.00Copyright © 2000 by Academic PressAll rights of reproduction in any form reserved.

r

-ll

y

s

n

nt,

exist a number of assays that evaluate functof the cell, such as metabolic activity (eMTT (32), Alamar blue (35)), and others thindicate cell membrane integrity (e.g., ethidibromide (32), trypan blue (30)). However, thare a limited number of assays availabledetermining tissue viability, especiallyin situ,and none that provide continuous dynamic msurements.

Although some of these traditional cell vbility assays have been successfully applietissues (22, 32), these methods are limitethat they can only provide information abouttissue biopsyin vitro, and the information iconstrained to the time the biopsy was obtainIn addition, studies onin vitro tissue sectioncannot account for the effects of circulatblood in the vasculature, which can be a sigicant factor in a number of applications.

In the context of evaluating cryosurgical s-

cess, Rabinet al. (26) have used perfusion fix-a in2 hef .A tica cest itr inf

i-c ot 7,2 iona cab oft Thr toe n as o

of

nodi-nd

flu-nsun

ire-ionthereell

iofwsluasasastro

as-ed

sesd

heef-lso

permit chronic and long-term studies on the

jel--rst

tlyotic-

useein3).uc-

NAofthebe

s iningthe

esisain

ions,dy-

drizewtheta-(7).g a

man

s ofantP).

FPres-asd iserfor

(9).cal-ellentanye if

361NOVEL GFP-BASED VIABILITY ASSAY

tion coupled with delivery of the vital sta,3,5-triphenyltetrazolium chloride, after t

reezing protocol, butprior to tissue biopsylthough this obviously permits a more realisssessment of cryosurgery protocol suc

han analogousin vitro tissue slice studies,emains an end-point assay, with the injuryormation localized in time.

Intravital dyes used in conjunction with mroscopy, without fixation, are now the statehe art for dynamic tissue viability imaging (29). However, the techniques for dye perfusnd observation are invasive and the dyese toxic, consequently requiring termination

he host before recovery from anesthesia.estricts the duration of dynamic evaluationxperiments that can be completed withiingle anesthesia period. Thesein situassays d

not permit dynamic and long-term evaluationviability.

The assay proposed in this work doesrequire staining or fixation of the tissue to incate viability, but relies on the expression arentention of an intracellular protein, greenorescent protein (GFP), produced by a trafected cell. Because the fluorescent compois produced intracellularly, there is no requment for electroporation techniques or diffus(passive or active) to insert the probe intocell. Moreover, no exogenous substrate isquired to produce the fluorescence. The ccan be assayed without altering the systemany way, other than illuminating with lightthe proper excitation wavelength. This allocontinuous, real-time, nondestructive evation of GFP fluorescence when optical accespossible. These features may provide the bfor a powerful tissue viability assay as wellan assay for isolated cells. Because the induction of dyes is not required, limitationssociated with diffusion of dyes are supercedMoreover, biopsy is not necessary to asviability. GFP fluorescence may be assessereal time, in a living host, without sacrificing tanimal, allowing the study of host-mediatedfects on the protocol of interest. This would a

s

-

f

n

is

t

-d

-sn

-isis

-

.s

in

same animal or system.Green fluorescent protein is native to the

lyfish speciesAequorea victoria.The gene responsible for the production of GFP was ficloned by Prasheret al. (23) and subsequenexpressed in both prokaryotic and eukarycells by Chalfieet al. (3). The unique fluorescent properties of GFP underlie its currentin diverse applications, ranging from protlocalization to ion channel localization (1Transfected cell lines are produced by introding a GFP-encoding segment of foreign D(Clontech, Palo Alto, CA) into the cell lineinterest. In this study we have focused onR3230 AC tumor cell line, because it canutilized to grow mammary adenocarcinomathe rat, a system of interest for our ongostudies. The transfected cell manufacturesGFP protein using its inherent protein synthmechanism. Furthermore, the vector can remstably expressed in successive generatwhich makes it especially attractive for stuing tumor metastasis. Chishimaet al. (4–6, 8)and Yanget al. (33) have effectively utilizetransduction of a GFP vector to charactesuccessive stages of metastatic colony groof human lung adenocarcinoma and the mstatic potential of hamster ovarian tumorsMore recently this group has been developinGFP-based mouse model of metastatic huprostate cancer (34).

There are numerous variations or mutantthe wild-type GFP (e.g., 15). The GFP mutused in this work is the enhanced GFP (EGFEGFP is a red-shifted variant of wild-type Gthat has been engineered for brighter fluocence (353 brighter than wild type). EGFP ha single excitation peak at 488 nm (12) anwell-suited for use with standard FITC filtcubes. This variant has also been optimizedimproved expression in mammalian cellsOther mutants have been optimized for loization within particular organelles and ccompartments (14), as well as for differspectral ranges (12, 15), which indicates mpossibilities for specifically targeting damag

a GFP-based tissue viability assay can be estab-

benu12

oit

emntstycerein

deo

thnd

catorit

touor oatell

ion

yonThveas

arsu

licassdi-ecttioninau-s

t f cev uti-l e

same concept but implemented differently. The-adi-ia-ingandsaysilityy atnteg-in--

i-entob-e-y,

adus-9).a-

ithto-G

d,

sure2ov-entoellsnted

n-to a

aTE-stheent

362 ELLIOTT, MCGRATH, AND CROCKETT-TORABI

lished.Wild-type GFP and EGFP are known to

distributed throughout the cytoplasm andcleus when expressed in mammalian cells (Because GFP accumulates in the cytoplasmnormally metabolizing transfected cells,seems likely that damage to the plasma mbrane could result in leakage of cell conteincluding GFP. In this work a GFP viabiliassay, based on a loss of GFP fluorescendamaged cells, is compared to the viabilitysults determined by the membrane integritydicators trypan blue and ethidium bromiSlow freezing is used to produce a rangedamage levels. This method of challengingcells has implications for both cryosurgery acryopreservation.

If the disappearance of GFP fluorescencebe shown to correlate with accepted indicaof viability (trypan blue, ethidium bromide),may have potential as anin vivo indicator oftissue viability. The goal of this work wasevaluate the potential of using the native flrescence in transfected cells as an indicatocell and tissue viability, by first establishing ththe GFP-based viability assay is valid for csuspensions.

A cell viability assay based on a transfectprocess has also been explored by Coomeet al.(11) and Baumstark-Khanet al. (1). The assadeveloped by Coome is based on the expressiluciferase that has been transfected into cells.assay requires addition of a substrate, howeand as currently implemented, has little utilityan in situ viability assay for tissues. A similGFP-based assay, requiring no exogenousstrate, was developed for biotechnology apptions by Huntet al. (16) to quantitatively assecolony growth in bioreactors. Their studies incated that GFP fluorescence as measured spphotometrically increased with cell concentrato a maximum that was dependent on the origplating density. The high correlation of GFP florescence to cell number (R2 5 0.997) supporthe use of GFP fluorescence as an indicator oiability as proposed in this work. The assayized by Baumstark-Khanet al.was based on th

-).f

-,

in--.fe

ns

-f

ofisr,

b--

ro-

l

ll

colony forming ability of wild-type and GFPtransfected CHO cells was measured after irration. They found that colonies surviving irradtion expressed GFP and that the survivfractions were comparable for transfectedwild-type cells. None of these GFP-based aswere explored as the basis for a tissue viabassay. The current study looks more directlGFP fluorescence retention as a membrane irity indicator, a generally accepted functionaldicator of cell viability in cryobiological applications.

MATERIALS AND METHODS

Cell culture. A rat mammary adenocarcnoma (R3230AC) cell line and an equivalenhanced GFP-transfected cell line weretained from Drs. Chuan-Yuan Li and Mark Dwhirst (Department of Radiation OncologDuke University Medical Center). Cells hbeen transfected with a pEGFP-NI plasmiding DMRIE liposomes and G418 selection (1Cells were cultured in Dulbecco’s modified Egle’s medium (D-MEM) supplemented w10% fetal bovine serum, 4% penicillin strepmycin solution (10,000 units/ml penicillinsodium and 10,000mg/ml streptomycin in0.85% saline), and 1% fungizone (250mg/ml)(all products from Gibco BRL, Grand IslanNY), with incubation at 37°C and 5% CO2atmosphere. Cells were harvested by expoto a 0.25% trypsin solution (Gibco BRL) formin at room temperature, subsequently reming the trypsin solution by pipette, and thincubating the dry flask for 3 min at 37°Crelease the cells from the flask surface. Cwere then resuspended in supplemeD-MEM.

Freezing injury. GFP-transfected and notransfected cell suspensions were preparedconcentration of 13 105 cells/ml in D-MEM.Aliquots (220ml) were added to 53 50 mmplastic vials (Corning, NY) and placed inprogrammable refrigeration bath (Neslab R140) at 25°C for 2 min. The solution wanucleated, and a 1-min period allowed forlatent heat of fusion to dissipate. A conveni

range of freezing damage was produced in both

m2 urer amp plew nda g,2

ndetereragere-s-

1ndn

ingi-

hoithreyol-th.omachlorc.0

amGBougeat00agngenth

gatBredlls3%

EtBr-positive cells and this percentage in-ure.iche-the

uo-all

fro-ents

ingthelly,thisslym-on-tonwl-aFPns,use

imeme

is

surell asmorn-

ed

ofon-m-reg-of

laterum-eg-ntsto

col-


of these cell lines by freezing from25°C toinimum temperatures between210°C and30°C at 1°C/min. When the bath temperat

eached the desired minimum temperature, sles were held in the bath for 10 s. The samere then rapidly thawed by immersion agitation in a 37°C water bath. After thawin0 ml of the cell suspension was added to 20ml

of trypan blue solution (0.1% in PBS), aviability was assessed using a hemacytomData points represent the average of threelicates, each replicate representing the aveof sixteen 13 1024 ml volume hemacytometwell counts (;80 cells per replicate). The rmaining 200ml of cell suspension (for tranfected R3230 AC cells only) was added towell of a 96-well tissue-culture microplate aeither 20ml of 1.25 mM ethidium bromide iD-MEM media (EtBr assay) or 20mL of D-MEM media (GFP assay) was added. Followa 10- to 15-min period of incubation, the mcroplates were transferred to a Nikon Diapinverted fluorescent microscope equipped wa 100-W mercury arc lamp. The wells weilluminated with blue light (450–490 nm) binserting a FITC filter cube (Chroma Technogy Corp., Brattleboro, VT) into the light paSeven color digital images were captured frapproximately the same regions within ewell using a SPOT thermoelectric-cooled coCCD camera (Diagnostic Instruments, InSterling Heights, MI). It took approximately 1min to locate each assay region, focus the frof interest, and obtain a three-channel Rimage of each test area. The average cell cper control well image was 133 (absolute ran86–253 cells); therefore, each control replicrepresented the average of approximately 1cells. Images were later analyzed using ImPro software (Media Cybernetics, Silver SpriMD). The total number of fluorescent grecells was counted in the GFP assay, andtotal number of fluorescent green, EtBr-netive (nonred) cells was counted in the Eassay. Although the number of EtBr staincells was not explicitly counted, control wewere generally observed to have less than

-s

r.p-e

t

,

e

nt,

e0e,

e-

creased with decreasing minimum temperatA color cube-based counting method, whidentifies all cells within the color range slected by the user, was utilized. In this casecells with the brightest and weakest GFP flrescence were used to set the color limits. Inassays, cell counts were normalized to nonzen control counts. Each assay point represthe average of three replicates.

Photobleaching and photodamage.AlthoughGFP is known to be resistant to photobleach(13) this effect needs to be evaluated incontext of each system of study. Additionabecause one of the goals of developingassay is to provide a means of continuoumonitoring cells and tissue, the potentially daaging effects of photoabsorption warrant csideration. For example, Miller and Silvers(20) were able to kill neurons within a feminutes by preloading them with Lucifer Yelow CH and irradiating with blue light fromfluorescent microscope. Although bright Gfluorescence is desirable for most applicatiothe capacity to absorb exciting energy can cadamage to system of study if the exposure tis excessive. The limiting total exposure tibefore irreversible photodamage occurstherefore a parameter of interest.

To assess the effects of continuous expoon the intensity of GFP fluorescence, as weon the viability of the cells, 24-well flat-bottotissue-culture-treated polystyrene plates (Cing, NY) were plated at a density of 13 104

cells/ml. After 24 h the medium was removand a solution of 2.5mM EtBr in fresh mediawas added to the wells. Following 10 minincubation a population of plated cells was ctinuously exposed to blue light, and digital iages were taken of this same population atular intervals over 40 min. Four replicatesthis assay were performed. Images wereanalyzed using Image Pro Software. The nber of green fluorescent, ethidium bromide-native (nonred) cells was counted. Cell couwere normalized to the starting cell countsdetermine the fraction recovered in platedonies.

ellsot

cemged tbleriein

d inPro

toflutereac

na as 1.f cove dec er

reet’s

sested.asd

1 f0 nal-y nbt nsityo ndt acht FPi thec ingo

ACc (1 ellsf rer ns-f oft the

FPryGFP


Because GFP is thought to leak out of cwith a damaged membrane, to assess phbleaching it is necessary to evaluate fluoresintensity in cells that are not membrane-copromised. The ethidium bromide assay imathat were captured at 40 min were examineisolate specific cells that remained viathroughout the test period. In each test sethree to five fluorescing cells were selectedthe 40-min images and identified and markeall of the preceding images using Imagesoftware. All images were then convertedgray scale and the gray scale intensities oforescence of these specific cells were demined as a function of time. Replicates repsent the average fluorescent intensity of esubpopulation of cells.

RESULTS

Freezing injury.The recovery of cells frozet 1°C/min to various end temperatures, asessed by the three assays, is shown in Fig.reeze damage curve was obtained, with rery decreasing as the minimum temperaturereased. For all minimum temperatures th

FIG. 1. (a) Recovery of cells frozen at 1°C/mfluorescence, GFP fluorescence with EtBr costaassessed by GFP fluorescence with recovery asfluorescence with recovery assessed by trypan b

o-nt-so

s

-r--h

s-A--

e

was no statistical difference between the thmethods used to assess recovery (Studenttest,a 5 0.05). It was assumed that in all cathe sample parameter was normally distribuA linear regression analysis of GFP viabilitya function of ethidium bromide viability yieldea correlation coefficient of 0.95 (y 5

.1271x). A similar correlation coefficient o

.97 was obtained for a linear regression asis of GFP viability as a function of trypalue viability (y 5 1.0628x). Figure 2 illus-

rates the change in GFP fluorescence intef cells after freezing to different minimum e

emperatures. The total number of cells in eemplate was approximately equal. The Gntensity in the damaged cells is so low thatells are undetectable without counterstainr use of phase-contrast microscopy.The response of nontransfected R3230

ells, as assessed by trypan blue exclusionn 5), was comparable to that of transfected c

or the25 to220°C minimum end temperatuange (data not shown). The single nontraected data points fell within the error barshe transfected cell data for three of four of

o various minimum temperatures, as assessed by Gand trypan blue exclusion. (b) Correlation of recovesed by EtBr. (c) Correlation of recovery assessed byData represent averages of three replicates61 SEM.

in tin,seslue.

clupreze

luegee).i-

t troc in-c ane mip

re

imeofre-gease

ob-entpa-uo-

th a

nme

roltage


temperatures investigated by trypan blue exsion. This suggests that, as expected, theence of GFP does not alter the basic freethaw response.

Photobleaching and photodamage.At full il-lumination (shutter completely open) with blight, the critical limit before detectable damaaccrued was 15 min (3% of cells EtBr positivSignificant damage (.10% of cells EtBr posive) was observed beyond 30 min. Conells, which were not illuminated, but wereubated at room temperature, did not showvidence of damage at the end of the 40-eriod.Normalized gray scale intensity measu

FIG. 2. Fluorescent images of GFP-transfectemperatures and thawing in a 37°C water bath:

-s-–

l

yn

-

ments are shown in Table 1 as a function of tfor cells that remained viable after 40 minexposure. Data beyond 30 min were notported due to the likelihood of photodamaartifacts in the intensity data. A steady decrein fluorescent intensity over time can beserved for these cells. At 20 min the fluorescintensity has decreased by 15%. This is comrable to the approximate 20% decrease in flrescent emission at 510 nm measured wimonochromator by Cubittet al. (13) when theS65T mutant of GFP was irradiated at 280with light from a xenon lamp for the samduration. The fluorescent intensity of contcells that had been left on the microscope s

R3230 AC cells after freezing to various minimumunfrozen control; (b)210°C; (c)220°C; (d)230°C.

ted(a)

id

tivilitole

endeddtru

is

hyso

nte toris-rteof

ho23tio

Good agreement between GFP and conven-ob-

epe

1 int is

asehenmns

ypeol

hanororet of

-erywithghery attered.ayP-

f 10ery

ableuch

ofre-

ry.pe-ry

ilityd inw-ays

f 1ned

o-oto-

asres-

TABLE 1sur

T cteR


for 30 min, without constant illumination, dnot decrease.

DISCUSSION

Some aspects of GFP are of interest relato the presumed basis of a GFP-based viabassay. These include the size of the GFP mcule and its stability. Yanget al. (31) describthe tertiary structure of GFP as cylindrical acomposed of a tightly woven, 11-strandb-sheet, with ana-helix inside the cylinder anshort helical segments on each end. This sture is frequently referred to as ab-barrel orb-can. The fluorophore portion of the proteinencapsulated within the barrel, utilizing theaend segments as scaffolding (31).

As the tertiary structure suggests, GFP is picochemically very stable. Purified samplesGFP have been shown to be highly fluorescepHs ranging from 5.5 to 12 and thermally stabl65°C (2). To our knowledge, stability charactetics at low temperatures have not been repoDenaturation of the GFP will result in a lossfluorescence, primarily because the chromopis no longer encapsulated and protected (10,Fluorescence is recovered upon renaturahowever (2).

GFP Fluorescence Intensity as a Function of Expoime for a Select Population of Plated GFP-Transfe3230 AC Cells

Time(min)

Normalized fluorescentintensity SEM

0 1.000 0.0001 0.980 0.0192 0.987 0.0073 0.988 0.0094 0.986 0.0155 0.983 0.0176 0.970 0.0177 0.968 0.0158 0.964 0.0249 0.970 0.026

10 0.964 0.01515 0.928 0.02020 0.848 0.01830 0.783 0.012

Note.Data represent averages of four replicates.

ey-

c-

-fat

d.

re).

n,

tional membrane integrity indicators wastained, as is evident from the correlations (R 50.95 EtBr; R 5 0.97, trypan blue) for freezinjury at slow freezing rates. Although the sloof the correlation curves was greater thanboth cases, the deviation from the ideal fithought to be a dynamic effect, resulting fromslight variation in the time for GFP to diffuinto and ethidium bromide out of the cells. Twater-soluble cylindrical GFP monomer is 3in diameter and 4 nm long (31). Calculatiobased on the proposed structure of wild-tGFP yield a molecular weight of 26,862 g/m(17). GFP is thus a much larger molecule tethidum bromide (mol wt, 394.3 g/mol)trypan blue (mol wt, 960.8 g/mol) and therefcan be expected to diffuse more slowly outhe cell. This combination of diffusion dynamics can give rise to small differences in recovat a given assay time. Recovery measuredthe GFP assay would be expected to be hithan recovery with EtBr or trypan blue assaa specific point in time, giving a slope greathan 1 in the regression analysis, as observ

Image Pro analysis was performed in a wsuch that the threshold for distinction of GFpositive cells was chosen as 1 color index (opossible) above background. This was a vconservative measure and included as vicells with fluorescence diminished by as mas 90%. A less stringent threshold (e.g., 50%average starting fluorescence) would havesulted in uniformly lower values of recoveAlternatively, the threshold value could be scifically determined by fitting the GFP recovedata to the recovery data of another viabindicator, such as EtBr or trypan blue as usethis study. From a statistical standpoint hoever, for the threshold level utilized, the assgave equivalent results (Student’st test, a 50.05) and therefore the threshold value ocolor index above background was maintaithroughout all experiments.

Although it can be difficult to separate phtobleaching effects from leakage due to phdamage, some degree of distinction wachieved by measuring the decrease in fluo

ed

cence intensity for cells that remained viablea-enth

uoext

ghtagni-thece

e-ncle-mse

n.erthdi-a

gegvi-thepidor

nalhe

, the asisskrothe

thedursa

lsodexiamiaamosFPrac

terization before general applicability of the as-, theeennottoof

ed inn-ion:rme-

le-theinglls,

ng.yetilarm-ors

rans-ofef-

es-therusthisble

s ael-

y at-

thator ofe-

issuefule

xam-fors are

Li,ed-


after 40 min of exposure to blue light illumintion. A steady but slight decrease in fluorescintensity was observed for these cells overtime interval assayed, indicating that the flrescence is quite stable for reasonably longposure times (;10 min). It is not certain thathis decrease in intensity isn’t due to slileakage of cytoplasmic GFP caused by damto the membrane, which has not fully mafested itself yet. Given the relative size ofviability markers, however, red fluorescenfrom ethidium bromide would likely be dtected before significant loss of fluorescedue to diffusion of the much larger GFP mocule. To avoid inadvertant photo-induced daage, cells assayed in this work were not expoto blue light illumination for longer than 5 mi

Some semiquantitative observations wmade over the course of these experimentsalso support the use of GFP as a viability incator. No intracellular GFP fluorescence wobserved following rapid freezing by plunfreezing (;500°C/min) and slow thawin(;10°C/min). EtBr assay also revealed 0%ability for these conditions, suggesting thatGFP assay will be a valid viability assay at rafreezing rates as well. Overgrowth of cellslack of fresh media can give rise to nutritiostresses that eventually lead to cell death. Wthese circumstances were allowed to occurGFP fluorescence was found to decay in timlong as these conditions were allowed to perAt various stages EtBr was added to the flaand the relative fraction of membrane- compmised cells was qualitatively comparable todisappearance of GFP fluorescence.

Although the method used to challengecells corresponds to the injury experienceding cryopreservation and cryosurgery, the asis not limited to these applications. It may aprove to be a valid viability assay for a wirange of phenomena, including tissue hypodrug-induced tissue death, and hypertherThis assay has the added advantage of dynin vivo assessment, without sacrifice of the hanimal. There are several aspects of a Gbased viability assay that need further cha

te--

e

e

-d

eat

s

nest.s-

-y

,.ict--

say can be claimed, however. For exampleturnover rate of cytoplasmic GFP has not bestablished for cells that are damaged, yetmembrane-compromised. This could leadfalse positive viability assessment if this typedamage is present and the sample is assaya narrow time window after insult. The coverse situation also needs further evaluatevents that lead to transient changes in peability may cause GFP leakage suggestingthal damage, when in actuality this is notcase. Assaying at a second time point, allowGFP to regenerate in non-lethally affected cewould circumvent this possible shortcomiHowever, such appropriate time points haveto be established. These limitations are simto those experienced with conventional mebrane integrity assays. Although the authhave had general success cryopreserving tfected cells without noticable compromiseGFP expression, quantitative studies of thefects of low temperature on constitutive exprsion are warranted. The same is true of omodes of cellular insult. Although ubiquitoapplication has not been demonstrated inwork, the authors feel that there is considerapotential for utilization of GFP transfection ageneral viability assay. In particular the devopment for tissues appears to be especialltractive.

SUMMARY

The data presented in this study suggestGFP fluorescence can be used as an indicatcell viability for studying freezing/thawing phnomena in isolated cells. GFP transfectionalso expected to have great utility as a tisviability assay. It should be particularly usefor dynamic in situ viability assessment. Thpresent results suggest that further studies eining the use of a GFP-based viability assaycells in tissues should be pursued. Such studiecurrently under way in our laboratory.

ACKNOWLEDGMENTS

The authors thank Drs. Mark Dewhirst and ChuanDepartment of Radiation Oncology, Duke University M

ical Center, for the generous donation of cell lines andian

this

, M.ro-V

of-

ndarke..,ff-

res-ac-).an,an,zedtein

da,a-cen

.,sion

by.

.,an

sta-A

o.

1 dere of

1 M.,lucit of

1 S-tein

13. Cubitt, A. B., Heim, R., Adams, S. R., Boyd, A. E.,im-ins.

1 .,P.,Or-s-s.),999.

1 ngthreen

1 . M.tive,etic.

1 ntence

1 in,cent.

1 , J.,ofvia

2 ofin-

2 andbi-

2 ndc-

2 ren-turein.

2 ues,

2 andFlu-.

2 ndry

,3,5-

2 G.,pi-


expertise. We are also grateful for the assistance of DDoeing, who performed some of the cell counts inwork.

REFERENCES

1. Baumstark-Khan, C., Palm, M., Wehner, J., OkabeIkawa, M., and Horneck, G. Green fluorescent ptein (GFP) as a marker for cell viability after Uirradiation.J. Fluorescence9(1), 37–43 (1999).

2. Bokman, S. H., and Ward, W. W. RenaturationAequoreagreen fluorescent protein.Biochem Biophys. Res Commun.101,1372–1380 (1981).

3. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W., aPrasher, D. C. Green fluorescent protein as a mfor gene expression.Science263,802–805 (1994)

4. Chishima, T., Miyagi, Y., Li, L., Tan, Y., Baranov, EYang, M., Shimada, H., Moossa, A. R., and Homan, R. M. Use of histoculture and green fluocence protein to visualize tumor cell host intertion. In Vitro Cell. Dev. Biol.33, 745–747 (1997a

5. Chishima, T., Miyagi, Y., Wang, X., Baranov, E., TY., Shimada, H., Moossa, A. R., and HoffmR. M. Metastatic patterns of lung cancer visualilive and in process by green fluorescence proexpression. Clin. Exp. Metastasis15, 547–552(1997b).

6. Chishima, T., Miyagi, Y., Wang, X., Tan, Y., ShimaH., Moossa, A. R., and Hoffman, R. M. Visualiztion of the metastatic process by green fluoresprotein expression.Anticancer Res.17, 2377–2384(1997c).

7. Chishima, T., Miyagi, Y., Wang, X., Yamaoka, HShimada, H., and Moossa, A. R. Cancer invaand micrometastasis visualized in live tissuegreen fluorescent protein expression.Cancer Res57, 2042–2047 (1997d).

8. Chishima, T., Yang, M., Miyagi, Y., Li, L., Tan, YBaranov, E., Shimada, H., Moossa, A. R., PenmS., and Hoffman, R. M. Governing step of metasis visualized in vitro.Proc. Natl. Acad. Sci. US94, 11573–11576 (1997e).

9. Clontech Protocol No. PT2040-1, Version NPR67195, p. 6.

0. Cody, C. W., Prasher, D. C., Westler, W. M., Prengast, F. G., and Ward, W. W. Chemical structurthe hexapeptide chromophore of theAequoreagreenfluorescent protein.Biochemistry 32, 1212–1218(1993).

1. Coombe, D. R., Nakhoul, A. M., Stevenson, S.Peroni, S. E., and Sanderson, C. J. Expressederase viability assay (ELVA) for the measuremencell growth and viability.J. Immunol. Methods215,145–150 (1998).

2. Cormack, B. P., Valdivia, R. H., and Falkow, S. FACoptimized mutants of the Green Fluorescent Pro(GFP).Gene173,33–38 (1996).

e

,

r

t

,

-

f-

Gross, L. A., and Tsien, R. Y. Understanding,proving and using Green Fluorescent ProteTrends Biochem. Sci.20, 448–455 (1995).

4. De Giorgi, F., Ahmed, Z., Bastianutto, C., Brini, MJouaville, L. S., Marsault, R., Murgia, M., Pinto,Pozzan, T., and Rizzuto, R. Targetting GFP toganelles.In “Methods in Cell Biology: Green Fluorecent Proteins” (K. F. Sullivan, and S. A. Kay, EdVol. 58, pp. 75–85. Academic Press, San Diego, 1

5. Heim, R., Prasher, D. C., and Tsien, R. Y. Wavelemutations and post-translational auto-oxidation of GFluorescent Protein.Proc.91,12501–12504 (1994).

6. Hunt, L., Jordan, M., De Jesus, M., and Wurm, FGFP-expressing mammalian cells for fast, sensinon-invasive cell growth assessment in a kinmode.Biotechnol. Bioeng.65(2), 201–205 (1999)

7. Inouye, S., and Tsuji, F. I.AequoreaGreen FluoresceProtein: Expression of the gene and fluoresccharacteristics of the recombinant protein.FEBSLett. 341,277–280 (1994).

8. Kolb, V. A., Makeyev, E. V., Ward. W. W., and SpirA. S. Synthesis and maturation of green fluoresprotein in a cell-free translation system.BiotechnolLett. 18, 1447–1452 (1996).

9. Li, C-Y., Shan, S., Huang, Q., Braun, R. D., LanzenHu, K., Lin, P., and Dewhirst, M. W. Initial stagestumor cell-induced angiogenesis: Evaluationskin window chambers in rodent models.J. Natl.Cancer Inst.92, 143–147 (2000).

0. Miller, J. P., and Selverston, A. I. Rapid killingsingle neurons by irradiation of intracellularlyjected dye.Science206,702–704 (1979).

1. Morise, H., Shimomura, O., Johnson, F. H.,Winant, J. Intermolecular energy transfer in theoluminescent system.Biochemistry13(12), 2656–2662 (1974).

2. Muldrew, K., Hurtig, M., Novak, K., Schachar, N., aMcGann, L. Localization of freezing injury in artiular cartilage.Cryobiology31, 31–38 (1994).

3. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Pdergast, F. G., and Cormier, M. J. Primary strucof theAequorea victoriaGreen-Fluorescent ProteGene111,229–233 (1992).

4. Pegg, D. E. Viability assays for preserved cells, tissand organs.Cryobiology26, 212–231 (1989).

5. Prendergast, F. G., and Mann, K. G. Chemicalphysical properties of Aequorin and the Greenorescent Protein isolated fromAequorea forskaleaBiochemistry17, 3448–3453 (1978).

6. Rabin, Y., Julian, T. B., Olson, P., Taylor, M. J., aWolmark, N. Evaluation of post cryosurgery injuin a sheep breast model using the vital stain 2triphenytetrazolium chloride.Cryo-Letters19, 255–262 (1998).

7. Schuder, G., Vollmar, B., Richter, S., Pistorus,Fehringer M., Feifel, G., and Menger, M. D. E

Illumination fluorescent light microscopy for theine to

lb,, E.di-re

.,an

pec

hetein

32. Yang, H., Acker, J., Chen, A., and McGann, L. In situ

, Y.,andpo-een

v, E.,ande-

, K.,.,

eirect


vivo study of rat hepatic microvascular responscryothermia.Hepatology29(3), 801–808 (1999).

28. Shimura, O. Structure of the chromophore ofAequoreaGreen Fluorescent Protein.FEBS Lett.104(2), 220–222 (1979).

29. Wanner, G. A., Mica, L. L., Wanner-Schmid, E., KoStefan A., Hentze, H., Trentz, O., and WolfgangInhibition of caspase activity prevents CD95-meated hepatic microvascular perfusion failure andstores Kupffer cell clearance capacity.FASEB J.13,1239–1248 (1999).

30. Weybright, P., Millis, K., Campbell, N., Cory, D. Gand Singer, S. Gradient, high resolution, magicgle spinning 1H Nuclear Magnetic Resonance stroscopy of intact cells.Magn. Reson. Med.39(3),337–45 (1998).

31. Yang, F., Moss, L. G., and Phillips, G. N., Jr. Tmolecular structure of Green Fluorescent ProNature Biotechnol.14, 1246–1251 (1996).

-

--

.

assessment of cell viability.Cell Transplant.7(5),443–451 (1998).

33. Yang, M., Hasegawa, S., Jiang, P., Wang, X., TanChishima, T., Shimada, H., Moossa, A. R.,Hoffman, R. M. Widespread skeletal metastatictential of human lung cancer revealed by GrFluorescent Protein expression.Cancer Res.58,4217–4221 (1998).

34. Yang, M., Jiang, P., Sun, F., Hasegawa, S., BaranoChishima, T., Shimada, H., Moossa, A. R.,Hoffman, R. M. A fluorescent orthotopic bone mtastasis model of human prostate cancer.CancerRes.59(4), 781–6 (1999).

35. Zhang, Z., Rydel, R. E., Drzewieke, G. J., FusonWright, S., Wogulis, M., Audia, J. E., May, P. Cand Hyslop, P. A. Amyloidb-mediated oxidativand metabolic stress in rat cortrial neurons: no devidence for a role of H2O2 generation.J. Neuro-chem.67, 1595–1606 (1996).

Documents

Green Fluorescent Protein: A Novel Viability Assay for Cryobiological Applications