measurements of cell physiology: ionized calcium, ph and

AD-A264 28911111 Iiinliiiiiilll ll i ORT DOCUMENTATION PAGE

Ia. crvJao i v i arsmJ lb. RESTRICTIVE MARKINGSUNC.

2a. SECURITY CLASSIFICATION AU TIYELC,-TE 3. DISTRIBUTION/AVAILABILITY OF REPORT

Approved for public release;2b. OECLASSIFICATIONIDOWNGP SCItA 31993 distribution is unlimited

4. PERFORMING ORGANIZATION R NUMBEIR S MONITORING ORGANIZATION REPORT NUMBER(S)

NMRI 93-17

6a. NAME OF PERFORMING ORGANIZATION 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION

Naval Medical Research (If applicable) Naval Medical Command

Institute I

6c. ADDRESS (Cty, State, and ZIP Code) 7b. ADDRESS (City, State, and ZIP Code)8901 Wisconsin Avenufe Department of the Navy

Bethesda, MD 20889-5055 Washington, DC 20372-5120

8a. NAME OF FUNOING/SPONSORING 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER

ORGANIZATION Naval Medical (If applicable)

Research & Development Comman

6c. ADDRESS (City, State, and ZIP Code) 10. SOURCE OF FUNDING NUMBERS

8901 Wisconsin Avenue PROGRAM PROJECT TASK WORK UNITBethesda, MD 20889-5044 ELEMENT NO. NO. NO. ACCESSION NO.

62233N MM33C30.005 1051 DN249507

11. TITLE (Include Security Classification)

Measurements of cell physiology: ionized calcium, pH, and glutathione

12. PERSONAL AUTHOR(S)

Rabinovitch PS. Junel CH. Kavananh TI

13a. TYPE OF REPORT 13b. TIME COVERED IA. DATE OF REPORT (Year, Month, Day) I5. PAGE COUNT

Book chapter FROM TO 1993 " 30

16. SUPPLEMENTARY NOTATION In: Clinical Flow Cytometry: Principles and Application. Edited by Kenneth D. Bauer,

Ricardo E. Duque, T. Vincent Shankey. Baltimore: Williams & Wilkins. 1992 pp. 505-534

17. COSATI CODES I18. SUBJECT TERMS (Continue on reverse if necessary and identify by block number)

FIELD GROUP SUB-GROUP flow cytometry; calcium; pH; HIV infection; signal transduction

19. ABSTRACT (Continue on reverse if necessary and identify by block number)

0~ 93-10214

20. DISTRIBUTION /AVAILABILITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATION

UUNCLASSIFIEDIUNUMITED 0 SAME AS RPT. --OTIC USERS Unclassi fied

22a. NAME OF RESPONSIBLE INOIVIDUAL 22b. TELEPHONE (Include Area Code) 22c. OFFICE SYMBOL

Phyllis .Bluwn" Librarian. (301) 295-2188 MRL/NMRI

00 FORM 1473, e4 MAR 83 APR edition may be used until exhausted. SECURITY CLASSIFICATION OF THIS PAGEAll other editions are obsolete. UNCLASS I FI ED

egCAURITY CLASSIFICATION OF THIS PAG09 I

UNCLASSIFIED

SECURITY CLASSIFICATION OF THIS PAGEII

NTIS CRA&/

3U lfllo nced

Measurements of Cell Physiology: Ionized Calcium, o,.,______o__pH, and Glutathione

PETER S. RABINOVITCH, CARL H, JUNE, and TERRANCE 3. KAVANAGH r Seciation Iit

The flow cytometer can be used to measure a variety of pase C. In many cases, this process requires the intervening

functional parameters that are of increasing interest to cell activation of a guanine nucleotdde binding (G) protein. Phos-

biologists. The recent development of a number of new pholipase C causes the hydrolysis of a membrane phospho-fluorescent probes now permts the measurement of various lipid, phosphatidylinositol 4,5-bisphosphate (PIP2), which

intraceihilar free ion concentrations in single living cells, yields a water soluble product, inositol I ,4,5-trisphosphaxeAmong these ions are calcium," magnesium, sodium, ([P3), and a lipid, I ,2-diacylglycerol (DAG). [P3 then

pot~assiurn, arnd hydrogen (pH)D. In addition, intracellular causes the release of calcium from intracellular stores while

glutathione, crucial for maintaining the physiological redox DAG, in conjunction with calcium ions, activates proteinstape, can be easily and accurately measured by flow kinase C. Thus, a single agonist can result in the production

cy'tometry. Most previously available techniques to measure of at least two "second messengers," m aking this pathway acellular activation parameters determined the mean value for unique, bifurcating system. Calcium is ther, fore a third

a mixed population of cells. The flow cytometer has the messenger" that controls numerous cellular processes, acti-

unique capacity to permit the measurement of physiologic rating a broad variety of enzyme systems, both as a cofactorparameters in large numbers of single cells; it alows and in conjunction with the calcium-binding protein

correlation with other parameters, such as immunophenotype calmodulin. While it appears clear that the initial elevationand cell cycle; and, finally, it reveals heterogeneity within of ionized calcium is due to the release of intracellular cal-

the cell population, sometimes even in cells that were ciun stores, little is known about regulation of the influx of

previously thought to be homogeneous. In this chapter, flow calcium from extracellular sources that is necessary to sus-

cytometrMc pechniques to measure inoracellular calcium aran the response.

concentration, pH, and glutathione as well as theirapplications are described. Indicators of Intracellular Ionized Calcium

ConcentrationINTRACELLULAR IONIZED CALCIUM Until 1982, it was not possible to measure [Cya2bit in small

Introduction intact cells and attempts to measure cytosolic free calcium

were restricted mostly to large invertebrate cells where theEukanyotic cells in their resting state maintain an internal use of microelectrodes was possible. Bioluminescent indica-

calcium ion concentration that is far below that of the cx- tors, such as aequorin, a calcium-sensitive photoprotein, are

tracelilular environment. Ionized calcium has fr important well suited for certain applications (I). Their greatest limits-

role as a mediator of transmembrane signal transduction, and son is the necssity for loading into cells by microinjecthoni

elevations in intracellular ionized calcium concentradon or other forms of plasma membrane disruption. [Ca 2*] was

([Car ]) regulate diverse cellular processes. Measurement of first measured in diverse populations of cells with the devel-

[Ca 2"]i in living cells is thus of considerable interest to i U- opment of quiit2 (2). The indicator was easily loaded into

vestigators over a broad area of immunology and cell biol- small intact cells using a chemical technique developed by

ogy. Tsien (3). Cells are incubated in the presence of the acetox-

Calcium influx is thought to be initiated by membrane ymethyl ester of quin2. This uncharged form is cell permeant

depolarization, which opins voltage-gated channels, or by and diffuses freely into the cytoplasm, where it serves as a

the binding of ligands to receptor-operated channels. The nst- substrate for esterass. Hydrolysis releases the tetrasnionic

ter case is more commonly encountered by investigators us- form of the dye that is tapped inside the cell. Unfortunately,

ing flow cytomeuy, as these mechanisms prdooM inate in quiof 2 has several disadvantages that limit its application to

cells that a e not elecgtrically excitable; in ,his case, the bind- flow yt ofntry (4). Its relatvely low extinction coefficient

ing of agonis to its specific membrane reciepmor lg tivates e- and quantum yield have made detection of the dye at lowzymalc prnoc uses that r tsult in the activation of phosphomi- conceneations difficult; at highedr oncentrations, quir2 itself

d5 w ie v

ter cu i m co m monl en o n e e b,,, inve tig tor ,u,- for of the dy that is mtIamlmld insii th el n ot n atl y

506 PART II: CLINICAL APPLICATION

coo" "ooc coo Doc ( OC an individual cell. Rijkers et a]. (9) have suggested using theoro'oooo-o- ratio of fluo-3 to SNARF- I (SemiNaphthoRhodaFluor) fluo-NN N N N N o

0o,,o o 0 0 o0 rescence for this purpose (see subsequent section on pH).

SciC Preparation of Cells for Indo-1 or Fluo-3 Analysis

OH cUptake and retention of indo-1 and fluo-3 is facilitated by

EGTA FLUo3 INDO I the use of their penta-acetoxymethyl esters, using the

Figure 31.1. Structure of the [Ca'*1 probes indo-1 and fluo-3 as scheme described above. Approximately 20% of the total

wen as, for reference, the structure of the calcium chelator EGTA. dye is trapped in this manner during typical loadings. After

Note the identical calcium-binding domain in all three compounds. loading, the extracellular dye should be diluted 10- to 100l.fold before flow cytometric analysis (10). A typical protocol

buffers the [Ca2 ]i. Subsequently, Grynkiewicz et al. (5) de- for cell preparation might be:

scribed a new family of highly fluorescent calcium chelators 1. Incubate 0.5 - 20.0 x 10' cells,'ml in medium with 1 - 3 i.LM

that overcome most of the above limitations. One of these fluo-3 or indo-l (acetoxymethyl esters) at 37C for 30 min. For

days, indo-l (Il-[2amino-5-[6-carboxylindol-2-yl]-phe- cells that are difficult to load, or that result in heterogeneousyethane N,N,NN'-te- loading, 0.02% (final vol/vol) pluronic detergent F-127 (Molec-

noxy]a2ia2'(amino-g.3methlyphenoxy] ethane that make ular Probes) may be added to the above before incubation, withtraacetic acid)) (Fig. 31.1), has spectral properties that make subsequent incubation at 30'C (11, 12). An additional option forit especially useful for analysis with flow cytometry. In par- the use of fluo-3 is to retard transport of the deesterified dye out

ticular, indo- I exhibits large changes in fluorescent emission of the cell by adding 4 mM (final concentration) probenecid dur-

wavelength upon calcium binding (Fig. 31.2). As described ing loading. In initial experiments, we suggest using lympho-below, use of the ratio of intensities of fluorescence at two cytes, as they are easily and reliably loaded. Later, when thewavelengths allows calculation of [Ca2* ], independent of va- technique is validated on the flow cytometer, other cell t'pesriability in cellular size or intracellular dye concentration. can be utilized.The only significant drawback to the use of indo-l is the 2. For experiments with simultaneous immunofluorescence, incu-

requirement for ultraviolet (UV) excitation. bate aliquots of the above with saturating concentrations of R-

A practical alternative to indo- I became availab)e upon phycoerythrin (PE)-conjugated antibody (with fluo-3) or fluores-the description by Minta et a]. (6) of a fluorescein-based, cein isocyanate (FITC)- and'or PE-conjugated antibody (with

calcium-sensitive probe, fluo-3 (Fig. 31.1). This dye exhib- indo-l) at room temperature for 30 min. Antibodies should beazide-free in order not to poison cellular metabolism.

its an increase in fluorescence intensity with increasing 3. Centrifuge the above for 6 rain at lS0g at room temperature.

[Ca:*]i (Fig. 31.3). Fuo-3 is less sensitive than quin2 and Gently resuspend (do not vortex) in medium.at the desired cellindo-I at detecting small changes in [Ca-*]j, in part, because concentration (usually - 10,''m]) at room temtperature. The me-the Kd is higher (400 nM); conversely, this may be an ad- dium used for resuspension and analysis can be dictated primar-vantage if the experimental situation involves the distinciton ily by the metabolic requirements of the cells, subject only tobetween stimuli, all of which produce large (above 400 rim) reasonable pH buffering (preferably not solely bicarbonate, as it

[Ca 2*÷i responses. An additional advantage with fluo-3 is aill become alkaline) and the requirement that mM concentra-

that it can be used with other probes such as caged calcium tions of calcium be present. If analysis of Ca:- released from

chelators that may themselves require UV excitation (7). intracellular stores (independent of influx of extracellular Ca")

Furthermore, the use of fluo-3 permits, for the first time, the is desired, addition of 5 mM EGTA (ethylene glycol bis (.ami-

simultaneous use of other UV-excitable probes for flow noethyl ether)-N,N,N',N'-tetraacetic acid) to the cell suspensionsimulta suh use othoseu or cwill reduce Ca:* to -20 rnM, thus abolishing the usual cross-cytometry, such as those used for cell cycle analysis or mea- membrane gradient.surement of intracellular glutathione, allowing correlation of 4. Hold cells at room temperature (-22'C) until ready for analysis.these parameters with calcium responses. The primary disad- Leakage and compartmentalization of both dyes is accelerated atvantage of fluo-3 is that it does not have fluorescence higher temperatures. Warm an aliquot of cells to 37C for atproperties that allow ratiometric determinations. Therefore, least 5 mrin before analysis. Analyze at 37C. For a 10-min anal-calibration on a flow cytometer is more complicated because ysis, use at least 0.5 ml (usually containing -10' cells), so that

the signal is proportional to cell size and dye concentration cell transit is rapid from the warm sample chamber to the flow

as well as [Ca2*]J. Thus, the ability to measure responses in cell (this is usually through unheated small caliber tubing).

subsets of cells is more limited because the broad distribu- [Ca3 I• responses are highly temperature-dependent and reduced

don of fluorescence intensities of unstimulated cells often re- or even absent calcium responses will be seen at lower tempera-

suits in an overlapping distribution of the values from stimu- asea,.5. After obtaining a baseline mesurement for a sample, kinetic

lated and unstimulated cells. This problem can be minimized analyses are typically performed by quickly ceasing flow, re-by loading ,.tlls in the presence of pluronic F-127 (see be- moving the sample container, adding agonist, and rapidly re-low), which minimizes the cell-to-cell variation in loading starting flow (more sophisticated and rapid sample injectionwith fluo-3 (8), or by the simultaneous use of a second dye schemes are possible, but for the great majority of applications.that serves as an indicator of the magnitude of dye loading in the above -ill suffice).

CHAPTER 31: MEASUREMENTS OF CELL PHYSIOLOGY 507

100 Figure 31.2. Emission spectra for indo-1 as a functionof ionized calcium concentration. Fluorescence excitedat 356 nm was measured in a spectrofluorimeter. Atwavelengths below 400 nm. emission in the absence of

R calcium drops to a few percent of that seen in the pres

E ence of calcium. About 500 nm, there is more than a

L 75 fourfold converse difference. The lCal"I of EGTA buffer

A solutions used is shown at the right.TIVEI so 40 nM

NTEN onM

S 22550n1

TY .,"00 nm

0 1 I

400 450 500 550

WAVELENGTH (nM)

FLUO-3+ Co 2 t EMISSION tain the hydrophilic imperrneant dye and can be excluded bySC Mgating during subsequent analysis. The lower limit of useful

Ex=490nm intracellular loading concentrations of both indo- 1 and fluo-361.6 pM is determined by the sensitivity of fluorescence detection of

the flow cytometer, and the upper limit is determined by'C avoidance of buffering of [Ca 2 i] by the presencd of the cal-

cium chelating dye itself. In practice, one should use the5 least amoung of dye that is necessary to reliably quantitate

025 the fluorescence signal. In the case of indo-l, the dye has1 excellent fluorescence characteristics (30-fold greater quan-

turn yield than quin2; (5) and useful ranges of indo- I loading0.042 are much lower than the millimolar amounts required withquin2. In using the newer dye fluo-3, one should begii with

500 5 5 550 5t5 the same lower concentration that is found to be suitableWovelength (nm) with indo-1 and increase loading only if detection against

Figure 31.3. Emission spectra for fluo-3 as a function of Ionized cal- background autofluorescence becomes limiting (autofluores-cium concentration. The concentration of Ca* in the buffer solutions cence in the green spectrum is greater than in the blue orIs Indicated for each Individual spectrum. (Courtesy of R. Haugland, violet spectrum used by indo-1). For human peripheral bloodMolecular Probes, Eugene, OR). T-cells, we have found adequate detection of both dyes at or

above 1 1I.M (for indo-I intracellular concentrations achieved6. Cablbrate using one of the approaches discussed subsequently. If with this loading are -5 PM). Buffering of [Ca t* ]i in human

using iooomycif, it is n.ceay to be scrupulous in removing T-cels can be observed as a slight delay in the rise in [C&2*1iredu ionopbore before analyzing the next sample. Use and a retarded rate of return of [Ca2*] to baseline values. Inwashes with bleach and/or dimethyl sulfoxide (DMSO) to facili- the case of indo-1, this was seen when loading concentra-tate the removal of vaces of ionomnycin.tions above 3 I.M (22 ILM intracelular concentmtion) were

One hncidental benefit of the above loading strategy is used. A reduction in peak [Ca2*], occurred at even higherthat this mrcedure, like the more familiar use of fluorescein indol-l concentrations (10). Chused et Al. (13) have ob-diacews or carboxyfluorescein diacetate, allows one to dis- served slghtly greater sensitivity of murine B-cells to indo- Itingish between live and dead cells. The latter will not re- buffering, recommending a loading concentration of no


greater than I p.M. In side by side comparisons, we have (Fig. 31.2). The choice of filter combinations can lead tofound that calcium transients in B-cells are much more sensi- substantial differences in the magnitude of observed changestive to the effects of buffering by indo-1 than are T-cells. For in the violet-blue ratio, perhaps best summarized by thehuman platelets, a 2 j.M indo-I loading concentration has value R,./R, the range of change in indo-1 ratio observedbeen reported (14). Rates of loading of the dye esters can be from resting intracellular calcium (R) to saturated calciumexpected to vary between cell types, perhaps as a conse- (R.. 1 ). Filters centered at 530 nm (green) and 395 nm (vio-quence of variations in intracellular esterase activity. In pe- let) result in a R,.,/R ratio of 9.0 (individual instrumentsripheral human blood, more rapid rates of loading are seen in may vary somewhat). Replacing the 395-nm emission filterplatelets and monocytes than in lymphocytes. Even within with one centered on the peak "violet" emission of the cal-one cell ty e, donor- or treatment-specific factors may affect cium-bound indo-I dye (405 nm) reduces R,,AfR to 7.9. In-loading; for example, lower rates of indo-1 loading are seen stead, replacing the 530-nm filter with one centered on thein splenocytes from aged mice than from young rrice (15). peak "blue" emission of calcium-free indo-I dye (485 nM)

Indo- I has been found to be remarkably nontoxic to cells reduces R,,./R to 6.6. Thus, a greater dynamic range of thesubsequent to loading. Analysis of the proliferative capacity indo-I ratio is observed when light nearer to the isosbesticof human T-lymphocytes (10) loaded with indo-I has shown (450 nm) is avoided. Note that if a 525- to 530-nm bandpassno adverse effects on the ability of cells to enter and corn- filter is used for the longer indo-I emission wavelength, thenplete three rounds of the cell cycle. Similar results have been this same filter and photomulhiplier tube can be used forobtained with murine B-lymphocytes (13). This is especially (temporally displaced and gated) FITC emission.pertinent to the viable sorting of indo-I loaded cells based on[Ca2-)], as described subsequently. Calibration of Fluorescence to [Cal*],

Prior to the development of indo-1, [Ca2:]i determinationFlow Cytometric Analysis Technique with quin2 fluorescence was sensitive to cell size and intra-The instrumental setup for fluo-3 analysis is simply that used cellular dye concentration as well as [Ca2" ]i. This made nec-for routine FITC analysis, i.e., 488 nm excitation and detec- essary the calibration of each loaded sample (in fact, strictlytion of emission in a band centered near 525 nm. Considera- speaking of each individual assay) by determination of thetons for indo-I analysis are slightly more complex. The ab- fluorescence intensity of the dye at known Ca:" concentra-sorption maximum of indo-l is between 330 and 350 nm , tions, or at least at zero and saturating Ca-*. This is still thedepending upon the presence of calcium (5); this is well case for analyses with fluo-3. In contrast, with indo- I, use ofsuited to excitation at either 351-356 nrm from an argon-ion the Ca2"-dependent shift in dye emission wavelength allowslaser, or 337-356 nm from a krypton-ion laser, and (al- the ratio of fluorescence intensities of the dye at the twothough spectrally slightly less optimal) almost the same per- wavelengths to be used to calculate [Ca2*i],•

formance can be obtained by excitation at 325 nmn using a (R - R-&) Shelium-cadmium Laser. Laser power requirements depend [Ca:*]i = Kd - B• [Eq. 31.1]upon the choice of emission filters and the optical efficiency (R. - R) Sb

of the instrument; however, under optimal optical conditions where Kd is the effective dissociation constant (250 nM at(i.e., using quartz flow cells) as little as 5-10 mW excitation 37°C, pH 7.05), R, R.,, and R. are the fluorescence inten-appears satisfactory, within the range available from helium- sity ratios of violet-blue fluorescence at the experimental,cadmium lasers. In this regard, the stability of the intensity zero, and saturating [Ca 2"]i, respectively; and S.01&2 is theof the excitation source is less important in this application ratio of the blue fluorescence intensity of the calcium-freebecause of the use of the ratio of fluorescence emissions. and calcium-bound dye, respectively (5). Note that the Kd

An increase in [Cal*]i is detected with indo-I as an in- varies dramatically as a function of temperature, pH, andcrease in the ratio of fluorescence intensity from a lower to a ionic strength. The term SpISh2 is a constant that depends inhigher emission wavelength. The optimal strategy is to select largest part on the filters used for indo-l analysis. Determi-bandpass filters so that the collection of light near the isos- nation of R.,, R.-,- and SI/Sj,2 for calibration of indo ratiosbestic point is minimized and the collection of fluorescence allows the direct determination of [Cal*], using the abovethat exhibits the largest variation in calcium-sensitive emis- equation. Determintion of Ra.u is easily performed by addi-sion is maximized. The choice of filters used to select these tion of the calcium ionophore ionomycin to a cell sample (1-wavelengths is dictated by the spectral characteristics of the 3 tIg/ml final). Unfortunately, simply adding EGTA to cellsshift in indo-I emission upon binding to calcium (Fig. 31.2). treated with ionomycin and metabolic poisons does not re-The original spectral curves published for indo-1 (5) did not duce the fluorescence ratio of cells to R.P,. The details of adepict the large amounts of indo- I emission in the blue-green strategy that makes use of a spectrofluorimeter in addition toand green wavelengths; in practice, we find that there is the flow cytometer has been previously described (10). Be-more light available in the blue region than in the violet re- cause the ratiometric analysis is independent of cell size andgion, although the dynamic range of the calcium-sensitive total intracellular dye concentration as well as of instrumen-changes in the violet region exceeds that of the blue region tal variation in efficiency of excitation or emission detection,

CHAPTER 31 : MEASUREMENTS OF CELL PHYSIOLOGY 509

it is not necessary to measure the fluorescence of the dye in possible by the recent commercial availability of carefullythe calcium-free and calcium-saturated states for each indi- prepared calcium buffer solutions (Molecular Probes, Eu-\vidual assay. In principle, it is sufficient to calibrate the in- gene, OR).strument once, after which only R is measured for each sub-sequent analysis. In practice, the minimum quality control Display and Analysis of Resultsshould include determination of the value of R,,•/R (cellstreated with ionomycinlresting cells) for each experiment; Analyses with fluo-3 require the observation of changing flu-

day-to-day optical variations in the flow cytometer are usu- orescence intensities, while the analysis of indo-1 (and of

ally minimal (with the same filter set) and, thus, a narrow fluo-3, if normalized for cellular loading differences by us-

range of R-J/R values should be obtained. ing the ratio to fluorescence of a second dye) requires deter-

As an alternative to the combined use of a flow cytome- rmination of a ratio of two fluorescence intensities. Fortu-

ter with a spectrofluorimeter, Parks et al. (16) have proposed nately, commercial flow cytometers all have some provisionthat, with minor modification, the flow cytometer may be for a direct calculation of the value of the fluorescence ratio,used as a spectrofluorimeter. In essence, the fluorescence of either by analog circuitry or by digital computation. If cellu-a steady stream of dye is measured by the photomultiplier lar indo-1 loading is extremely heterogeneous, it may be de-and the photomultiplier voltage is analyzed as in a standard sirable to work with a logarithmic conversion of "'violet"spectrofluorimeter, or, even beter, the voltage is electroni- and "blue" emission intensities in order to observe acally converted to a pulse for processing by the flow cytome- broader range of cellular fluorescence. In this case, the in-ter. An alternative approach to create such a pulse is to rap- strument must permit the logarithm of the ratio to be calcu-idly strobe the laser, such that the beam is illuminated for lated by subtraction of the log "blue" from the log "vio-only I.sec intervals. Preliminary experience with this tech- let" signals (10).nique suggests that it is possible to determine all of the con- Plotted as a histogram of the ratio values, quiescent cell

stants necessary for calibration directly on the flow cytome- populations stained with indo-1 show narrow distributions ofter using EGTA-depleted and calcium-saturated indo-1 ration, even when cellular loading with indo-I is very heter-solutions. ogeneous, and coefficients of variation (CVs) of less than

An alternative approach to calibration can be based upon 10% are common (Figs. 31.4A and 31.4!). For simple fluo-3a regression curve that relates R to treated cells suspended in analyses, variation in dye loading will result in wider CVsa series of precisely prepared calcium buffers. Chused et al. (Figs. 31.4B and 31.41). In either case, the effects of pertur-(13) have suggested that ionomycin plus a cocktail of meta- bation of [Ca2* ]i by agonists can be noted by changes in thebolic poisons be used to collapse the calcium gradient to zero histogram profiles, observed by storing histograms sequen-([Ca 2"], = [Ca2"]o). Thus, this technique allows one to esti- tially over time, with subsequent analysis of data.mate [Ca 2 )i] without the need to determine R,, Sp or Sb:, A more informative and elegant display is obtained by aalthough it is subject to limitations in the precision with bivariate plot of indo- 1 ratio or fluo-3 fluorescence intensitywhich one can prepare a series of calcium buffers that yield vs. time. The bivariate data can be displayed as "dot plots"

known and reproducible free calcium concentrations. Accu- on which the indo-1 ratio or fluo-3 intensity of each cellracy of prediction of ionized Ca2 ÷ concentration in buffer (proportional to [Ca 2 *1i) is plotted on the Y-axis vs. time onsolutions depends on a variety of interacting factors, so thatcare must be exercised in formulating Ca2" standards. The the X-axis (Figures 4A and 48). Alternatively, the data can

caremus beexecisd i fomulaingCa2 stndads.The be presented as "isometric plots" in which the X-axis repre-ionized calcium concentration in an EGTA buffer system de- b rsne s"smti lt"i hc h -xsrpeinsonized c magnesium concentration; other metals such as sents time, the Y-axis is indo-I ratio or fluo-3 intensity, andpends on the ane l nntratum als such as the Z-axis is the number of cells (Figs. 31.4K and 31.4L). Inaluminum, iron, and lanthanum also bind avidly to EGTA these bivariate plots, kinetic changes in lCa2Ii, are seen with(17). In addition, the dissociation constant of Ca4÷-EGTA is mh greate plo n, lit ed on by the seer witha function of pH, temperature, and ionic strength (review 1, much greater resolution, limited only by the number of chan-19). For example, in an EGTA buffer (total EGTA 2 rM, nels on the time axis, the interval of time bc',een each chan-total Ca2* I mrM, ionic strength 0.1 at 37C), changing the nel, and the rate of cell analysis. Changes in the fraction of

pH from 7.4 to 7.0 can result in the ionized calcium increas- cells responding, the mean magnitude of the response, and

ing by more than 200 nM, a change that is approximately the heterogeneity of the responding populaiton are best ob-

twice the magnitude of that found in resting cells and that is served with these displays. For example, it can be seen in

easily measured on a flow cytometer. Finally, it is important Figures 31 .4A and 31.4K that for CD4* T-cells at I min af-

to prepare the buffers using the "pH metric technique" (19), ter the addition of antibody to CD3, the response is highly

in part because of the varying purity of commercially avail- heterogeneous, with some cells not yet responding and otherable EGTA (20). Detailed methods and a computer program cells achieving very high levels of [Ca 2*÷i. Figures 31.4Bthat will determine [Ca'j] as a function of pH, total calcium and 31.4L show that when the analysis of fluo-3 and indo-I

and magnesium concentrations, temperature, and ionic in the same cells are compared, the presence of the broaderstrength e available from one of this chapter's authors (C. CV of the measurement with fluo-3 results in a diminished

June). A practical alternative for obtaining buffers is made ability to resolve the heterogeneity of the response.

40.

as'

- * . D '. i

04

U~ 4"

U 4a 3 366dn~ Cmi Tie mm

C 60 0 Go4

40 - - 01 a 2

E -~ F

soAl

TIOG~ go15a im s

as ) & I a i

Fiue314 -


For many purposes, it is simpler and preferable to ex- cellular immunophenotype simultaneously with the [Ca2*],amine and compare a univariate parameter describing the assay, allowing alterations in [Ca 2*]j to be examined in, andcellular response vs. time. There are several such parameters correlated with, specific immunophenotypic subsets. FITC-that can be derived from the bivariate data for this purpose. and PE-conjugated antibodies can be used with indo-1, andCalculation of the mean Y-axis value for each X-axis time PE and allophycocyanine (APC) (or other red-excited dyes)interval allows presentation of the data as mean ratio or fluo- can be used with fluo-3. Only the fluo-3/PE combination canrescence intensity vs. time (10), as shown in Figures 31.4C be performed with a single-laser flow cytometer, the otherand 31.4D. Calibration of the ratio or intensity to [Ca'"]i combinations require a dual-laser flow cytometer.allows data presentation in the same manner as traditionally Numerous examples of the analysis of [Ca 2*)i in immu-

displayed by spectrofluorimetric analysis, i.e., mean [Ca2*]i nophenotypically defined subsets have been described (re.'s. time (Figs. 31.4E and 31.4F). While this presentation view, 22). Several illustrative examples will demonstrate theyields much of the information of interest in an easily dis- utility of this approach. Figure 31.5 shows the simultaneousplayed format, data relating to the heterogeneity of the analysis of CD4÷ and CD8÷ subsets of PBL, performed us-[Ca-*]i response are lost. Some of this information can be ing two-color inmmunofluorescence simultaneously with

displayed by a calculation of the "proportion of responding indo-I. This figure is also chosen in order to illustrate thecells." If a threshold value of the resting ratio distribution is importance of the interaction of receptors on the cell surface

chosen, i.e., one at which only 5% of control cells are above (in this case by cross-linking) and the relative contributionsthe threshold value, the proportion of cells responding by ra- of internal and external calcium stores.

1io elevations above this threshold vs. time yields a presenta- Addition of an anti-CD3 antibody alone (3 p.g/nm,tion that is informative of the heterogeneity of the response UCHT-l, Dako Corp) in this experiment had only a small(Figs. 31.4G and 31.4M). A more sophisticated measure of effect upon [Ca:÷]i in either CD4* or CD8" subsets (see 2-the proportion of responding cells can be derived by the ap- to 4-min interval of Fig. 31.5). This antibody differs from

plication of histogram subtraction techniques as illustrated in that of Figure 31.4, in that the laner produces a greaterFigures 31.41 and 31.4J: The measurement of unperturbed [Ca2 "] response in the absence of further cross-linking.

cells in the initial period of observation can be used as the When saturating amounts of goat antimouse antibody is"control" fluorescence or ratio distribution; at each subse- added (at time 4 rain), an intracellular calcium response is

quent time interval, the cumulative subtraction technique seen that is greater in CD4÷ cells than in CD8" cells. As the

(21) has been used to subtract the control distribution from extracellular medium used in this experiment contained little

the observed distribution. Comparison of Figure 31.4H with Cal*, the response from 4 to 7.5 rain represents mobilizationFigure 31.4. shows that the latter technique is less sensitive from intracellular stores. This response is characteristically

to the shape and width of the control distribution; this differ- of short duration (compare Figs. 31.5C and 31.SD to FiguresP ence is more dramatic in Figure 31.5, described below. 31.4E and 31.4F), which is the strongest evidence that the

extended phase of elevated [Ca:÷] is maintained by influx ofSimultaneous Analysis of [Ca2*], and Other extracellular Ca2÷. To illustrate that calcium channels have,Simlu neousAne almstes of ain fact, been opened by the cross-linked stimulus in FigureFluorescence Parameters

31.5, Ca:÷ was added to the extracellular medium at timeThe most common application for use of additional fluoro- 7.5 rain; this is accompanied by an immediate elevation ofchromes with [Ca2÷] analysis will be the determination of [Ca2÷]i, with a sustained response. Again, the magnitude of

Figure 31.4. Comparison of indo-1 and fluo-3 measurement of vs. time is shown in C and fluo-3 intensity vs. time is shown in D. ThetCa-I, and methods of displaying and analyzing calciurmi signaling in data were converted to calcium concentration vs. time by calibration,single cells as a function of time. Human PBL were loaded simultane- using measured constants in equation 1 for indo-1 (E) and by calibra-ously with 3 "Ig/ml indo-1 and 3 ipg'rnl fluo-3, stained with PE-CD4 tion with buffer solutions for fluo-3 (F), and the values were plotted vs.mAb, and then stimulated with anti-CD3 antibody (10 ipml G19-4, a time. Note that whereas the shapes of the curves for mean indo-1gift from Dr. Jeff Ledbetter, Oncogen, Seattle, WA) at approximately 1 ration (C) and fluo-3 intensity (D) are different (particularly in an ap-min after the start of analysis (note gap in data acquisition). Violet and parently slower return to baseline after 4 min in (D), once converted togreen indo-1 emission (UV-exclted) and green fluo-3 and orange PE [Cal-), the two measurements are essentially identical. The apparentemission (temporally delayed 488-nm exciled) were collected simulta- discrepancy in C vs. D is related to the different values of Km for indo-neously for each cell. Only results gated from PE-CD4* cells are dis- 1 and fluo-3. The percent cells responding with ICa2*] elevated be-played. In A and B, the results are displayed as a "dot plot" In which yond two standard deviations above the mean of the cells before anti-[Ca0'i Is plotted for each cell analyzed on a 100 x 100 pixel grid, body stimulation ("percent cells above threshold") are plotted forwhere the X-axis Is time and the Y-axis is indo-1 ratio or fluo-3 fluores- indo-1 (G) and fluo-3 (H). The percent responding cells calculated bycence Intensity. The number of cells per pixel is displayed by dark- cumulative curve subtraction (see text) are plotted in I and J for indo-1ness that ranges over 12 levels. At the bottom, K and L, isometric and fluo-3, respectively. Data analysis In this and subsequent figuresplots of the same experiment In A are shown; sequential histograms was pelormed with a sottware program written by one of the authorsare plotted In which the X-axis represents time, the Y-axis [Ca'** or (P.S. Rabinovhtch) ("MuliTime," Phoenix Flow Systems, San Diego,fluo-3 Intensity, and the Z-axIs number of ceft. The mean Indo-1 ratio CA).

Ar

04

cc.

.48

20 36

50 U 406

0j c

A 2 24

70 76

C go

- 0 - 0,

026

Is

c 561 3

Figure 31.& A-J

CHAPTER 31 : MEASUREMENTS OF CELL PI1YSIOLOGY 513

this final response is greater in CD4÷ cells than in CD8÷ population for their subsequent biochemical analysis orcells, growth. Results of artificial mixing experiments with Jurkat

Comparison of the percent of cells above a threshold (T-cell) and K562 (myeloid cell) leukemia lines indicate that(Figs 31.5E and 31.5F) with the proportion of responding subpopulations of cells with variant [Ca2*]i comprising <1%cells calculated by curve subtraction (Figs. 31.5G and of total cells could be accurately identified (10). Goldsmith31.5ff) shows a substantial difference between the two mea- and Weiss (25, 26) have reported the use of sorting on thesures: The latter calculation detects more responding cells basis of indo-I fluorescence to identify mutant Jurkat cellsthan the former. This difference is greater in Figure 31.5 that fail to mobilize [Ca2 j]i in response to CD3 stimulus, inthan in Figure 31.4, due in largest part to the presence in spite of the expression of structurally normal CD3/Ti comn-Figure 31.5 of a greater fraction of cells with elevated plexes. These experiments suggest that sorting on the basis[Carl] in the "resting" control population; in such a case, a of indo-I fluorescence can be an important tool for the selec-threshold set to exclude all but 5% of the "resting" population and identification of genetic variants in the biochemicaltion may be required to be at a relatively high [Cal*]j value, pathways leading to Ca2- mobilization and cell growth andwhereas the method of curve subtraction avoids this problem differentiation.(21).

As a second example, Figure 31.6 shows the analysis of Critical Aspects in the Analysis of [Cal ], by Flowmrtine splenocytes. While approximately 50% of murine Cytometrysplenocytes are B-cells, only approximately 10% are T-cells COMPLICATIONS IN LOADING AND INTRACELLULAR(CDM-positive), as seen frnm the bivariate analysis of FITC- ENVIRONMENTCD20 vs. PE-CD5 (Fig. 31 .6A). When a T-cell stimulus (an- Potential problems to be aware of in the use of [Ca-" dyetibody to the T-cell receptor, CD3) is applied to the spleno- indicators include compartmentalization, leakage or secre-cytes, only a very small response is visible in the total popu- tion 'of dye, quenching by heavy metals, and incompletelation. In contrast, when the indo-I analysis is gated upon deesterifcation of dye ester. The flow ctometric analysis ofthe CD5÷ CD20- cells, then the response of the minority [ ]usi iicatof dye es Thedicated anahis ofsubpopulation of T-cells can be seen to be vigorous, and, as fCa2 ]di using indicator dyes is predicated on achieving uni-

expeted theCD5CD20 cels how o rspone (ig. form distribution of the dye within the cytoplasm. In severalexpected, the CD w-CD20÷ cells show no response (Fig. cell types, the related dye fura-2 has been reported to be31.68). Conversely, when a B-cell stimulus (anti-IgM) is coprm taiewthnrgnes(728.nbonea-applied, the response of the immunophenotypically identi- compartmentalized within organelles (27, 28). In bovine aor-appied, the responstwie othe mgitudpheofthpicat identi- thetic endothelial cells, fura-2 has been reported to be localizedfled B-cells shows twice the magnitude of that seen in thetochondria; however, under those condiios, indo-I e-ungated total population of splenocytes, while the T-cellsubset, as expected, shows no response (Fig. 31.6C). mained diffusely cytoplasmic (29). Thus, it is ýossible that

there will be fewer problems with compartmentalization ofUsing other probes excited by visible light, it is possible indo-I than with fura-2. Some cell types, such as neutrophils

to analyze additional physiologic responses in cells simulta- andonocits and Some cell lines suoppos toprimrneously with [Ca2*],. "the simultaneous analysis of mere- and monocytes, and some cell lines (as opposed to primarybrane potential and [Ca2 -3]e has been accomplished by several cells) appear to be more susceptible to compartnentaliza-graotential, 4), and s~~imlhasnbeen accompished by sal* and tion. In addition, compartmentalization is enhanced by pro-groups (23, 24), and simultaneous analysis of [Ca2 )i, and longed incubation of cells at 37C. In general, it is advisablepHi will be illustrated in a subsequent section of this chapter. to examine the cellular distbution of tndo-1 or fluo-3 mi-

croscopically and, in each new application, to confirm theSorting on the Basis of [Ca-l• Responses expected behavior of the dye. This is done by determining

The ability of the flow cytometric analysis with indo-1 to the ratio of R.. to R as a control for each experiment, asobserve small proportions of cells with different [Ca 2'i] re- described below. In addition, one should store indo-I loadedsponses than the majority of cells suggests that the flow cy- cells at room temperature after loading and use the cellstorerter may be useful to identify and son variants in the promptly after loading. Since heavy metals may quench the

Figure 31.5. Analysis of lCa1'l in human PBL stained with indo-1 acleristic of mobilization of Cal* from intracellulu stores, without in-and goed to show PE-positive CD4 cells (A, C, E, 0, I) and PE- flux of extraoallutat Cal- (see text). Al 7 rain, 10 mM CaCI. was addedTexaw-red-tandem-oonjgae (ECD, Coulter Corp.) positive CD8 cells to the medium. A and B display "dot plot$" of indo-1 ratio vs. time; C(B, D, F, l, J). Cells were suspended In medium with 5 mM EGTA, and D show calibrated ICa'-) vs. time; E and F show the percent ofwhich reduced available extracellular Cal to approximately 20 nM cells above a threshold vs. time (the threshold ratio is set such that(10). At I min after the start of analysis, 3 ig/ml an'i-CD3 mAb only 5% of "resting" cells are above that ratio, see text); G an KUCHT-1 was added (Dako Cow.; tOi antibody by Itself produces only show the percent responding cens vs. time; and I and J show isomet-a small JIJa' responlse), and at 4 min, 20 p~riV goa antimouse anti- rio displays of fth Vxlnd ratio vs. time.body wee added. The retum to baseline after a M,-ck response Is char- -'

ii l i m il • • • •


e2 indicator dye fluorescence, the use of the membraneA ::permeant heavy metal chelator diethylenetriaminepentaacetic

D t"-. D 0acid (TPEN) may be advisable for cell lines that contain in-. creased amounts of heavy metals (30). In situations in which

loaded dye is not retained intracellularly for a sufficientlength of time, the dye is often being actively secreted and

0 6• the use of probenecid, a blocker of organic anion transport,Of may be beneficial (27)."M If, for a particular cell type loaded with indo-l, the mag-CL 4• nitude of change between R and R. is in good agreement

with the values predicted from spectral curves of indo-I in acell-free buffer, then it would be unlikely that the dye is ;n acompartment inaccessible to cytoplasmic Ca*, in a form un-responsive to [Ca"'], (i.e., still esterified, see below), or in acytoplasmic environment in which the spectral properties ofthe dye have been altered.

26 4T 6Y 16 It has been suggested that both fura, 2 and indo- I (and byFITO THY-i1O • extension, presumably fluo-3) may be incompletely deesteri-

fied within some cell types (31, 32). Since the fluorescence, B of the ester has little spectral dependence upon changes in

Ca'*, the presence of this dye form could lead to false esti-mates of [Ca2 ]i. Again, results of calibration experimentsare helpful in excluding this possibility. Further, it has been

2W• proposed that since indo-1 fluorescence, but not that of theindo-1 ester, is quenched in the presence of rmM concentra-

* dions of Mn 2÷, then, in the presence of ionomycin, MN"÷, .... .:,-,-. .. .. -2• can be used as a further test of complete hydrolysis of the

, . . .. indo-1 ester within cells (31).Because of its more recent introduction, there is rela-

tively little information available at this time regarding therelative importance of each of the above concerns when us-ing fluo-3. To illustrate, however, the possibility that each

a, 3_ . dye may be subject to different intracelular processing andS 4 s IS environments, Figure 31.7 shows results obtained with

TIME (MIN) human peripheral blood lymphocytes (PBL) loaded similta-4,A C"neously with indo-I and fluo-3. Using UV and 488-nm exci-

C Ctation, each dye was independently but simultaneously ana-lyzed in the same cells. After exposure to ionomycin, indo- Ishowed the expected response associated with calcium in-flux; however, after an initial increase in fluorescence,

29 ",fluo-3 intensity declined at the same time that the indo-I ra-dio remained stably maximal. The reasons for the different

""t "behaviors of the two dyes is not yet clear, although it almost"certainly is related to one or more of the points discussed

S. ...... .... above. This experiment demonstrates graphically the need

T shows the result of stimulation with anti-CD3 m.Ab 2C1 1 (25 "gWl).

i 4 sThe ungated total con response is small (doffed line), whereas theTIME (MIN) response galed upon the T-cell region of A is approximately seven-

Figure 31.& Analysis of splenic B- and T-cellS S6 mouse spleen fold larger (solid line), and B-cells (dashed line) show no response. Cculo were stained with PE-thy-1 (T-cells) and FITC B220 (B-cells). A shows stimulation with goat antimouse Ig (25 plmI): to tesponseshows the blvarlate pill of FITC vs. PE. Ffeen percent of spleno- gated to show only B-cells (dashed line) is twice as large as that of allcytes we T-cells (quadrant 4) and 44% are B-colls (quadrant 1). B cells (doffed line) and T-cells (solid line) show no response.

S- " , . - :!


S.t .... ..... . .loa

j 0 --- , so

I 40

2- LL 2'~t0 4 2 4

•e0 lee

, , 0

S 60.

20 2 4

Time (min) Time (min)Figure 31.7. lonomycin treatment of CD4" peripheral blood T-cells the partial return to baseline seen in fluo-3 intensity and not seen insimultaneously loaded with indo-1 and fluo-3. The indo-1 ratio is the indo-1 ratio. This observation was variable over a series of experi-shown in A and B. and the fluo-3 intensity is shown in C and D. Note ments (data not shown).

for care in assessing correspondence to [Ca2*]i by indicator DMSO and then scavenging residual ionophore by washingdyes. with a buffer containing 2% bovine serum albumin.

UNSTABLE OR BROAD BASELINE DISTRIBUTIONS POOR CELLULAR RESPONSE

Under typical conditions, the baseline [Ca2+i] distribution Failure of the cells to show proper response to various treat-should show little (<3%) variation from sample to sample ments may be due either to difficulty with the cells or withand should, with indo-l, be narrow and, most often, follow the instrument. To differentiate between these, the cellsa normal (Gaussian) distribution. Some cell lines may have should be stimulated with the calcium ionophore ionomycinasymmetric distributions and altered mean values of "rest- and the magnitude of R,. to R should be determined. If theing" [Ca2*)], which can often be ascribed to a subpopulation indo-I ratio of R,.,/R increases by the amount described pre-of cells with elevated [Ca2*]i. This may result from the im- viously, or if the fluo-3 fluorescence increases by approxi-paired viability of some cells, "spontaneous" activation of mately five-fold, then the instrument is functioning properly.some subset of cells, or presumably, biological beterogene- If the increase is less than expected, then one should obtainity within the cell population (for example, cells within cer- an independent preparation of cells, such as murine thymo-tain phases of the cell cycle). Sometimes, the baseline will cytes or human PBL. If these cells load properly and alsostart at a normal level and then rise with time. This may be respond poorly, then the instrument alignment should bedue to the failure to completely remove from the sample checked. With indo-l, not uncommonly, the violet or bluelines an agonist from a previous experiment. The most corn- signals may not be properly focused, or there might be inter-mon problem has been residual calcium ionophore; this can ference from a second laser. This problem can be pinpointedbe effidienly removed by first washing the sample lines with by analyzing separately the blue and violet signals after io-

516 PART II : CLINICAL APPLICATION

nophore treatment; the violet signal should increase approxi- single cells (44). In addition, there is evidence that calciummately threefold and the blue signal should decrease approx- elevations occurring after cellular stimulation may be oscilla-imatley twofold (Fig. 31.2). tory rather than sustained (45, 46), thus raising the possibil-

If the instrument is functioning properly, then the prob- ity that some cellular processes controlled by calcium may1em may be in the cells. The cells must be loaded with suffi- be frequency-modulated as well as amplitude-modulated.cient indo-I or fluo-3 to be detectable well above autofluo- Since flow cytometry cannot measure the calcium concentra-rescence. If there is any question, this should be checked tion inside a single cell as a function of time, it is not possi-independently with fluorescence microscopy, especially if ble to distinguish whether there is a subpopulation of cellsone has become familiar with the expected fluorescence of that is responding with a sustained response or, alternatively,loaded cells examined in this manner. If the cells are too dim whether there are two populations of cells, one that has ele-or excessively bright or if the dye is compartmentalized, the vated calcium concentration and one that has basal levels. Inability to detect calcium signals will be impaired. For un- spite of these limitations, determination of [Ca2*]i in largeknown reasons, the calcium signaling of B-cells, and not T-cells, is particularly sensitive to overloading with indo-1 (10 numbers of single cells using flow cytometry with indo-l of-13). The cells must be suspended in media that contains cal- fen great practical advantages and allows measurements of acium; occasionally, responses will appear blunted because of kind that is not possible to achieve by means of the alterna-the inadvertent resuspension of cells in a medium than con-tains no added calcium.

Applications of the Flow Cytometric Analysis of

EFFECT OF ANTIBODY LABELING [Cal*],

In the simultaneous analysis of [Ca-" ]i and immunofluores- The flow cytometric assay of cellular calcium concentrationcence, consider that the use of the antibody probe can itself has already been applied to a wide variety of cells, providingalter the cellular [Ca:"J]. It is becoming increasingly clear interesting and sometimes unexpected results. Examples ofthat the binding of monoclonal antibodies (mAbs) to cell sur- the initial applications of the technique are presented in re-face proteins can alter [Ca24 ]i, even when these proteins are cent reviews (47, 48). One of the first observations that wasnot previously recognized as part of a signal transducing made readily quantifiable by the flow cytometric analysispathway (10, 33-38). For example, antibody binding to CD4 was that there is great heterogeneity in the response of lym-will reduce CD3-mediated [Ca*i signals; if the anti-CD4 phocytes to mitogenic agonists. Using simultaneous immu-mAb is cross-linked to the CD3 complex, as with a goat- nofluorescence, some of this heterogeneity can be shown toantimouse mAb, the CD3 signals are augmented (39, 40). be related to immunophenotypic subsets; for instance, CD4*Antibody binding to CD8 has similar effects (unpublisheddata).cells show More vigorous responses to phy-t~hemagglutinin,data).

As a consequence of these concerns, a reciprocal staining concanavalin A, and mAb to CD3 (see Fig. 31.5) than do

strategy should be used whenever possible so that the cellu- CD8÷ cells (10, 49). Considerable use of this approach has

lar subpopulation of interest is unlabeled while undesired been made in the demonstration of differences between

cell subsets are identified by mAb staining. The CD4* sub- [Ca24 ]j activation requirements of different cell subsets and

set in PBL may be identified, for example, by staining with a subset specificities of activation pathways. The effects of an-combination of CD8, CD20, and CDI 1 mAbs (10), and the tibody binding to cell surface molecules, sometimes a com-CD5 subset can be identified by staining with CDl6, plication in labeling experiments (see above), has been ex-CD20, and HLA-DR mAbs (35). Finally, it is important to tensively employed to analyze signaling mechanisms, andreiterate that, when staining cells with mAbs for functional augmented, or even new relationships have been probed bystudies, antibodies must be azide-free for metabolic pro- cross-linking antibodies on the cell surface (39, 50). Flowcesses to be uninhibited. Commercial antibody preparations cytometric measurements with indo-1 have been performedmay thus require dialysis before use. to date with all nucleated blood cell types. Applications of

fluo-3 have been reported with most types, with reports ap-INTRINSIC LIMITATIONS OF FLOW CYTOMETRY pearing at a rapidly increasing pace. The combination of sen-

Flow cytometry is unable to detect heterogeneity of cellular sitivity, reliability, and ability to analyze large numbers ofcalcium concentrations within a single cell, and there are re- cells within cell subsets has made the flow cytometric assayports from assays using digital video microscopy that, in of [Ca2 ÷]i the preferred technique for a broad spectrum ofsome situations, calcium transients may be present only research applications.within compartmentalized sublocations (41, 42). It has been There are many potentially exciting clinical applicationsreported that the photoprotein aequorin may in some circum- of the flow cytometric assay of cellular calcium concentra-stances help to detect changes in cytosolic calcium not re- don (47). One topical example is the study of the effects ofported by indo-1 (43), although use of the two indicators is the human immunodeficiency virus (HIV) infection. In in vi-complementary because aequorin cannot measure Ca2* in tro studies, the 1IV-1 retrovirus was found to impair signal


100. of cells, in that subset may show an altered response. Dem-/-Control onstration of such heterogeneity in [Ca-'*]i signals would

¢60 o•have been impossible to discern in conventional assays car-Tied out in a fluorimeter where only the mean calcium re-

CO. 601 sponse is recorded.

C,

_40.. INTRACELLULAR pHZ t ,i PotVen t 3 Introduction

S20. I.:.. : "V'/\.. The pHi of mammalian cells is -7.2, and pHi appears to be

closely regulated (the coefficient of variation measured by0. flow cytometry is -5%). In mammalian cells, pHi is regu-

0 A0 80 120 160 200 240 280 320 360 400 lated by at least three mechanisms, including Na*o/H*i ex-Time (Seconds) change, sodium-dependent CI-.HCO3)-j, and HCO 3)-J

Figure 31.8. PBL from asymptomatic HIV-inlected patienis or an Cl- exchange. Acid extrusion is primarily accomplished byuninfected control donor were stained with FITC-CD8 and PE-CD5. the Na*/H÷, antiport and by sodium-dependent Cl'JThe cells were loaded with indo-1, stimulated with anti-CD3 antibody HCOC,)-i, while HCO(3)-./Ci- exchange has the major role38.1 (1 Wg/ml), and changes in calcium were monitored in the CD4*(CD5"CD8-) cells. The time course of the percent of responding cells for base extrusion. All of these mechanism., appear to beis shown. stimulated by a variety of growth factors -nd by phorbol es-

ters, presumably through the activation of protein kinase C.It was initially proposed that an alkaline shift in pH, itself

transduction in CD4 cells, the major target of HIV-I (31). might be a signal to regulate cell activation. A number ofRecent studies suggest that CD4 T-cells from HIV-infected reports purporting to show a role for pH, in cellular activa-patients also have impaired responses to T-cell receptor stim- tion are derived from measurements made in bicarbonate-ulation (G.P. Linene et al., unpublished). Due to the neces- free medium. Thus, nonphysiologic (HCOa)-free) bufferssity of analyzing live, retrovirus-infected cells on the flow were used so that only the Na* /H+ antiporner could operatecytometer, a closed fluidics system is used to prevent the and, thus, alkalinization was the dominant response. Recentgeneration of aerosols. Lymphocytes from patients with reports, performed in physiologic medium, show that cellu-early-stage HIV infection (Walter Reed stage 1f-ItT) have lar activation is, in fact, accompanied by acidification inbeen isolated, loaded with indo- I, and the CD4 subset hasbeen analyzed, byaded wathinon the the CD8 subset ofs many cases and that the net effect (acidification or alkalin-_= been analyzed by gating on the CD5bi÷ CD8- subset of ization) is a function of the medium in which the experimentcells. More than 80% of HIV-infected patients have anti- is conducted. Thus studies of pHi are important physiologicCD3-induced calcium responses that are below that of age- tools for the elucidation of various ion exchange mechanismsand sex-matched control donors. There is substantial patient- but cannot be taken as evidence per se of cellular activationto-patient variability, with some patients having minimally (52).diminished responses and others displaying severely reducedsignaling (Fig. 31.8). Thus, both normal CD4 T-cells after Indicators of Intracellular pHHIV infection in vitro, and the CD4 T-cells from patientswith early stage HIV infection have impaired signal trans- Until recently, the most commonly used probes were modifi-duction. It is likely that the mechanism of impaired signaling cations of fluorescein-fluorescein diacetate being the first-may differ in vitro vs. in vivo. During in vitro infections, generation pH probe-followed by carboxyfluoresceinmore than 50% of the CD4 cells can be shown to be in- diacetate (COFDA). Both of these dyes are limited by rela-fected. In contrast, <1% of CD4 cells from patients with tively poor retention inside loaded cells. An improved fluo-early-stage HV infection can be demonstrated to be in- rescein probe is 2',7'-bis-carboxyethyl-5(6)-carboxyfluores-fected, so that an indirect mechanism must exist for the virus cein (BCECF). As with indo-I and fluo-3, BCECF is loadedto impair signal transduction pathways in uninfected cells. In into cells using the acetoxymethyl ester, after hydrolysis, itany case, the existence of this abnormality may prove to be a has a negative charge of - 4 or - 5 and, therefore, leakssensitive test of the immune system function in MIV infec- more slowly than COFDA. The pKa of BCECF, 6.98, istions. Current studies in several laboratories are exploring near the pK- of resting cells, and there is a pH-dependentwhether or not this application may provide prognostic infor- shift in the excitation wavelength, making it possible to usermation and whether the assay may be useful for the manage. the ratio of fluorescence signals to correct for differences inmeat of this chronic illness. Study of MIV infction does il- loading and cell size. In addition, BCECF fluorescence ex-lustrate the utility of flow cytometry in the clinical setting in cited at 450 rui is pH-independent, while fluorescence atwhich only one immunological subset of cells, or a portion 500 nm is pH-dependent, allowing ratiomea-ic fluorescence


Table 31.1Fluorocbromes for Ratiometric Determination of pH Using Flow Cy-tomerry

Probe + pita Elduttia2a Emissiuo Refgnacts

Bis-carboxyetityl-carboxyfluoresein 6.98 ratio 439/490 535 ParadisO et al. (130)A eoymediy] ester (BCECF AM) 488 ratio 5201620 MusgrOvtet Wt. (131)Diamaoxy-dicyanobenzene (ADB); yields 8.0 -350 ratio 425/540 Cook and Fox (132)DicytAnydroquinone (DCII) after de-estenification Musgrove et Wl. (131)CArboxy SNARF-l acetoxymethyl acetate 7.50 514 or 530 ratio 575/670 AWhiker et a]. (55)

(SNARF-l)

emission analysis (Table 3 1. 1). In both cases, the magnitude SNARF-1of pH-dependent ratio shifts is relatively modest (for exam-Ex48nple, when the pH is raised from 6.5 to 7.5, the ratio of fluo-Ex48nrescence intensities after excitation of FDA at 436 rim and pH 6.0495 rim increases by only 1.45 to 1.55 times (53)); this ismuch less of a shift than is seen with SNARE- I (see below)and, thus, these dyes are largely supplanted by the latter.

Subsequently, several UV-excited pH probes were devel- 2oped. The most useful of these is I ,4-diacetoxy-2,3-dicy-a-noberizene (ADB). The cell-permeant ADB is hydrolyzed7.and trapped intracellularly to yield 2,3-dicyanohydroquinone(DCH) (54). DCH fluorescence is pH-dependent, while theesterified forms exhibit pH-independent fluorescence; there-fore, care must exercised to ensure complete deesterificationof the ADB. Ratiomnetric determinations of pH are possible p .using a single excitation source by measuring fluorescenceemiission at 429 nmi and 477 nrn (Table 3 1. 1). As with mea-surement of fCa2Ji, many laboratories will find the require- 550 600 650 700ment for UV excitation to be the primary limitation for use Wavelength (m)of this probe. Figure 31.9. Fluorescence emission spectra of SNARF-1 as a turic-

Recently, the most useful probe for pH1 measurement tion of pH. Excitation was at 488 nmn. (Courtesy of R. Haughland. Mo-has been introduced; a compound termed SNARE- I lectilar Probes, Eugene, OR).(SemiNaphthoRhodaFluor) (55). This is the first pH probe tohave fluorescence characteristics that provide truly useful Preparation of Cells: for SNARF-1 Analysisand sensitive ratiometric properties. SNARF- I appears to be As with indo- I, the optimal conditions for loading must bethe most promising reagent for use in flow cytomnetry, as it empirically determined for each cell type. Cells must behas convenient excitation spectra and exhibits large changes loaded only to the point where the fluorescence signal isin pH-dependent fluorescence (Fig. 31.9). The acid form of above autofluorescence and sufficient for detection by theSNARE- I has absorption maxima at 518 and 548 =i, while fo yoee.Atpcllaigpooo sthe base form excites maximally at 574 rum. The emission of fo yoee.Atpcllaigpooo sSNARE-i in acid is maximal at 587 =i and the basic form 1. Incubate cells (S2 x 10'/inl) in buffered (for example, HEMESemits maximally at 636 rim; there is an isosbestic point at (4-(2-hydroxyethyl)-1-PiperZine-cthanesufonic acid), see610 am. In practice, either the 530-nm line of a krypton-ion above for rationale for using bicabonate-free buffer or not) me-laser or the 514-nm line of an argon-ion laser are optimal for diuml with 2-6 jiM carboxy SNARM- (diacetate) at 30'C toexcitation, while emission should be collected at both 575 2. * fo 30'

mm ad 60 tm ad te rtioof hes sinal caculted Ex 2.Cenrifuge and resspend in fresh mediuim at fth desired cellrit n 4 i n h ation of theF1at48 seol sightlyes calulted.alx concentration (usually -l0'/ml). Store cells at 2?C and protect

citation ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~~fo ofSAEIa 48rmi ny lgtyls otml ih untrl analysis. The cel pellet wil appear faintly pink.because it has the advantage of permitting simultaneous 3. Warm aliquot of SNARE- I loaded cell to 37*C for 5 to 10 minanalysis of FITC probes (i.e., immnunofluorescence), this befo analysis. The medium and sheath fluid should be baf-will probably be the most comnmonly used excitation wave- fare saline. The rate of cell analysis can vary, commonly, Cellslength- awe analyzed at 200 to 300 cells/sec.

CHAPTER 31 MEASUREMENTS OF CELL PHYSIOLOGY 519

10 of high potassium buffers of different pH (with otherwiseidentical ionic composition). The cells should be loaded with

0 the pH probe and treated with the proton ionophore nigericinP- 7 (1 i.zg/rrm), added in order to equalize pHi and to buffer pH< (56). Nigericin produces the equilibrium [K*i]/[Ke] =c[H*i]/[H+o]. Thus, in the presence of nigericin, pHi can be7 calculated with the knowledge of [K*,], [K*.] and the buffer

cc _pH. Equilibrate tube at 37C for 15 min, and measurez 3 SNARF-I fluorescence ratio at 37TC. Note that nigericin isto toxic to human lymphocytes at >5 tgg/rrl, manifested by

loss of cellular fluorescence., ,A calibration curve can be constructed by plotting values

6.00 6.25 6.50 6.75 7.00 7.2' 7.50 7.75 6.00 of pHi vs. the SNARF-I fluorescence ratio (Fig. 31.10). ThepH ratio changes -3-fold between pH 6.0 and pH 8.0. Most of

the ratio shift occurs between pH 7 and pH 8, as the slope ofFigure 31.10. Calibration of the SNARF-1 ratio of fluorescence

(670/575 nm) to pH. Jurkat cells were loaded with SNARF-1 and the curve is much less between pH 6 and pH 7.resuspended at 37"C in buffers containing 135 nM K" with varying pHin the presence of nigericin. Excitation was at 514 nm. Simultaneous Analysis of pH and Other Fluorescence

ParametersFlow Cytometric Analysis of SNARF- 1-Loaded Cells As mentioned above, the fluorescence emission properties of

,emission should be collected SNARF-I allow simultaneous excitation of FITC probes us-As described above, SNARF-I msinsoldb olce ing the same 488-nm laser. Analysis of ph, in immunophe-

using orange and red bandpass filters centered at 570 rnm (or- nothpically -ne llsets usin o njugatebange an 64 to670am red. Te rtioof ed/rane fue- notypically-defined cell subsets using FITC-conjugated mAbange) and 640 to 670 nm (red). The ratio of red/orange fluo- is thus very straightforward. Similarly, single-laser [Ca 2 -]irescence intensity is proportional to increasing pH (Fig. measurements with fluo-3 simultaneously with pHi are easily31.10). See the corresponding section above on indo- I anal- performed. When a second UV laser is available, [Ca* ],y'ses for further discussion of ratiometric analysis. measurement with indo-I is preferred. Fig. 31.11 illustrates

SNARF-I is a vital dye and, therefore, dead cells are the analysis of [Ca* ]i and pH, in splenic T-cells. FITC-CD4efficiently excluded from analysis by electronically gating oncell tht hve luorscece.If ITCfluoescnceis ea- mAb was used to gate the analysis specifically upon this sub-cells that have fluorescence. If FI XC fluorescence is m ea- se of m i e p n cy s( ot h w ),a d he t o f U Vsured simultaneously, as for determination of cell subsets, set of grouine splenocytes (not shown), and the ftios of UV-suitable gating on this parameter would also be used. Using excited green and violet indo-l emission and 488-ncm-excitedsingle-beam 488-nm illumination for analysis of FITC fluo- red and orange SNARE-1 emission were each calculated.rescence requires compensation for overlap into the green The elevation in [Ca2 i reached its peak within 5 mi after

FITC bandpass from the leftward tail of the SNARE- I emis- antibody to the T-cell receptor (CD3) was added and, subse-

sion spectrum. This is analogous to the routine color com- quently, slowly declined (Fig. 31.11A). Intracellular pHi

pensation used when analyzing PE with FITC, except that reached its peak more slowly, 10 min after addition (a 0.2even greater crossover compensation may be required when unit alkalinization), and thereafter also slowly declined to-the SNARF-I fluorescence is bright. Alternatively, cells can ward baseline (Fig. 31.10B).be stained with amino-methycoumarin acetic acid (AMCA)-conjugated antibodies and fluorescence analyzed using UV INTRACELLULAR GLUTATHIONEillumination, thereby avoiding the need for compensation. Introduction

The effective Kd of SNARF-I for protons is 7.5, whichis near the pHi of resting cells (pHi = -7.2). The pHI of Glutathione (glutamylcysteinylglycine, GSH) is an importantstimulated cells may decrease or increase, depending on cell antioxidant tripeptide thiol that is involved in the scavengingtype, stimulus, and medium composition (52). As seen in of toxic oxygen products (57-59). In addition, GSH is in-Fig. 31.9, described below, SNARF-I has a greater sensitiv- volved in a number of other important reactions in the cell,ity in reporting alkaline shifts than acidification. including conjugation of xenobiotics, amino acid transport,

and deoxyribonucleotide synthesis (57). GSH is also impor-tant for the maintenance of cellular thiol redox status, and its

Calibration and Data Analysis redox state has been proposed to be a major determinant ofthe functioning of a number of enzymes (60-66) and of the

For each ixperiment, it is necessary to construct a calibration integrity of cytoskeletal proteins (67). The biochemical path-curve so that one may determine the fluorescence channel ways of GSH synthesis, renewal, and utilization ae outlinednumber as a funicon of pHK. Cells are suspended in a series in Fig. 31.12.

,"0


. ............... ............ ............... . . ...... .These roles for GSH have important consequences uponcell physiology and mitogenic activation. In general, aug-

• .menting cellular GSH with cysteine delivery agents such asN-acetycysteine, oxathiayolidine carboxylic acid, or gluta-

0 .thione esters has a positive effect on cell growth, whereasdepletion of GSH inhibits many cellular functions (57). GSHis known to influence cell growth and replication at various

il I F levels including GI/S-phase transition (68, 69) and very earlyO 40 events in mitogen-induced cell activation (70). We will dis-a cuss later in this chapter some evidence that GSH affectsz very early steps of signal transduction. Other aspects of cell

growth, such as protein synthesis, are also influenced by71 .GSH levels, since -y-glutamyl transferase-dependent amino

acid transport utilizes GSH (71). Interestingly, although acti-a4 8 2 s6 28 24 28 vation steps that are important for cell replication of lympho-

4 ": TIME(micytes are GSH-dependent, those activation steps that are nec-* -" essary for acquisition of differentiated functions are notU anecessarily GSH-dependent (72), although some workers

21 have reported that GSH content can influence these functions2" as well (73, 74).

.2GSH has also been shown to affect the responsiveness ofR 24 2 cells to various cytokines. For example, intracellular GSH

has been shown to enhance the action of IL-2 in lymphocytesC . . . (72, 73, 75, 76), whereas extracellular GSH has been shown

•, ,.,"; .: :. :" " " .'. " to decrease the effects of granulocyte.1macrophaee colonyso h.m''"stimulating factor (GMl-CSF) and platelet-derived growth

0 factor (PDGF) on granulopoiesis and fibroblast proliferation,,I- respectively (77, 78). For GM-CSF, this effect is thought to

be mediated through direct competition by GSH for the GM-CSF receptor. In fact, GSSG, the oxidized form of GSH wasfound to have the capacity to stimulate the GM-CSF receptord: directly (78). The mechanisms responsible fof-the inhibitory

Z effects of GSH on PDGF-stimulated fibroblast growth areit less well understood.

. I The activity of redox-responsive oncogene products andt .t DNA-binding factors responsible for gene regulation may"".... • . also be affected by the GSH status of the cell (79-85). For

TIME (min) instance, the interaction of c-fos and c-jun products are74s D known to be redox-sensitive (79). Another example is that of

oo 74, NFkB. Roederer and colleagues (83) have shown that this"74,4 nuclear binding factor can be inhibited from binding to regu-X M.. latory DNA binding domains by n-acetylcysteine, a drug thatX ri increases the GSH level of cells.

IL 72 Finally, GSH has recently received attention because of4 t t, its role in a number of pathological disease states, including

Figure 31.11. Simultaneous analysis of murine splenocytes loaded tumor cell resistance to chemotherapeutic agents (98, Chap-with ndo-1 and carboxy SNARF-1. Murine splenocytes were loaded ters 14 and 27, this volume), idiopathic pulmonary fibrosiswl klndo-1 and SNARF-1 and stained with FITC-conjugated anti-CD4 (86, 87), and acquired immune deficiency syndrome (83,mAb. Fluorescence was excited by the UV lines of a krypton-ion laser&Wnd fe 488-nm line from an argon-ion laser. The ratio of indo-1 fluo- 88-91).

rieosnce (395 nmt525 nm) and of SNARF-1 fluorescence (665 nm/575 rnm) is displayed vs. time. The analysis was gated on the FITC-CD4 Iluorescence at 525 nm. Anti-CD3 antibody 2C11 was addedding ft gap in the analysis. The calcium and pH values for individ- Fluorescent Indicators of Intracellular GSH

ual cl are plotted as dot plots in the upper panels, and the mean Several yeas ago Durand and Olive (92) published a reviewresponse vs. time is plotted in the lower panels. (Cae-I increased offurescent i nd Olincludish , a tfroM 0.13 ILM to 0.36 i.M and pH, increased 0.2 units. of fluorescent indicators for thiols, including GSH, that

might prove useful for flow cytomet€ic purposes. The prob-


OVERVIEW OF GLUTATHIONE METABOLISM

XENOBIOTIC \ \ I /CONJUGATION TION

PROTEIN TRANSPORT c oPRTENAMINO ACID '00THIOL REDOX GS-Bimane GLUTAMATE

MCB/GLLIrATHINE- ?GLUTAM). GtUrATHICE-S. GLYCINE CYSTEINE

PROTEINIXED T�RSPEPTDSE TPANSFERASES ) .GLUTAMYLCYSTEINEDISULFIDE OXOO 107 0-rHTAREDUCTASE ox7, ST) STHE TA SE

\ / LsYrETAST E GLUTATHIONE SYNTHESIS

GLUCOSE yNADP GSHH202

HEXOSE GLLUTATHIONE GLUTATHRONE SODMONOPHOSPWrE REDUCTASE PEROXIDASE

RIBUILOSE ... W' 1K DP _-' GSSGJ ý"K%*,. HO

5-PHOSPHATE 2

GLUTATHIONE REDOX CYCLE

Figure 31.12. Overview of glutathione (GSI' metabolism. GSH is hexose monophosphate shunt (also called the pentoss phosphatesynthesized in a two-step process (upper right hand portion of figure): shunt). GSH can also serve as a cosubstrate for glutathione transfer-cystelne is metabolized to g-glutamylcysleine (g-CC) by g-GC ase-mediated conjugation of xenobiotics (shown here is the conjuga-synthetase and then to GSH by GSH synthetase. In the glutathione tion of monochlorobimane (MCB). to the fluorescent glutathione-redox cycle (lower portion of figure) glutathlone peroxidase uses GSH bimane corjugate, top of figure). GSH also has important roles into reduce hydrogen peroxide and other peroxides and, in the process, amino acid transport and in the maintenance o( protein thiol redoxGSH becomes oxidized (GSSG). Reduction of GSSG to GSH by status (upper left of figure).glulathione reductase requires NADPH, supplied primarily by the

lem with all of the dyes reviewed was one of specificity for dye has the same disadvantage as MBB of potentially bind-GSH. Durand and Olive (92) concluded that a promising rea- ing to other nonprotein thiols in the cells such as cysteinegent at the time for flow cytometric determination of GSH and ,y-glutamylcysteine. Nonetheless, OFT still has the ad-was monobromobirnme (MBB). Poot et a). (93) used this vantage that, when bound to protein thiols, it has slightlyreagent in their investigation of human diploid fibroblast different spectral characteristics than when bound to GSHGSH metabolism, and employed n-ethylmneimide-treated (higher blue fluorescence and less green fluorescence), al-"'blinks" in order to control for nonspecific protein thiol lowing one to simultaneously measure the sulfhydryl redoxbinding, status of protein and of low molecular weight thiols (94).

Anodr reagent that has since been introduced for flow A major advance in the measurement of GSH by flowcytometric evaluation of GSH is o-phthaldialdehyde (OPT) cybometry was made by Rice et Al. (95), who used mono-(94). OPT i supposedly more specific for GSH than MBB chiorobimane (MCB) to detect changes in GSH status of in-because, in addition to reacting with the cysteine thiol group dividual normal and tumor cells. Many papers have sinceof GS]H, it also reacts with the amino terminus of GSH, been published applying this technique to a number of inter-forming a higly fluorescent cyclic smxte. However, this eWsng problem in biology and medicine (70, 96-402). This

J


dye owes its specificity for GSH to the fact that it is conju- TSO and BPO, then MCB/FCM would be a convenient as-gated to GSH by glutathione-S-transferases (GSTs) (see be- say for the polymorphic expression of this isozyme, the ex.low) and has relatively low nonenzymatic reactivity toward pression of which has been associated with a decreased riskGSH and other thiols (103). for certain cancers (108-111).

Another reagent, chloromethylfluorescein diacetate Figure 31.13A shows a comparison of human T-cefls(CMFDA) has recently been described for measuring intra- stained with MCB at two different concentrations. It is obvi-cellular GSH (104). Once inside the cell, this dye is deesteri- ous that the fluorescence of these cells is much higher whenfled to a hydrophilic cell impermeant product chloromethyl- one stains with 1 mM MCB. We have also found, in contrastfluorescein (CMF) by nonspecific esterases. CMF is then to Cook et a]. (96), that the rate of GSH conjugation inpresumably metabolized in a manner similar to MCB, yield- human cells is temperature-dependent. As seen in Figureing a GSH-MF fluorescent conjugate. The advantage pro- 31.13A, a significantly faster rate of fluorescence develop-vided by this dye is that excitation and emission characteris- ment is achieved when cells are incubated at 370C rathertics are similar to fluorescein once conjugated to GSH, than at room temperature, and the relative difference isallowing one to use this dye simultaneously with UV excita- greater with 60 pLM MCB than with 1000 I.LM MCB. Thisble dyes-for instance, with Hoechst 33342 for simultaneous result is consistent with the observations of Seidegird andGSH and DNA determinations (see below). The principal colleagues (108) who showed that the maximum activity to-disadvantage of this dye is that it is apparently less specific ward TSO occurs at 40C in peripheral blood mononuclearfor GSH than MCB (104). cells. In the studies published by Cook et al. (96) and

Ublacker et a]. (101), cells were reacted with MCB at roomtemperature (although it is mentioned in a footnote to the

Performing and Calibrating GSH Analysis Cook paper that they did not see a temperature dependence

FACTORS THAT AFFECT THE UTILITY OF MCB FOR GSH in the cell types they examined). Both of these groups haveMEASUREMENTS addressed the limitations of the MCBiFCM method whenWe and others have found that the conditions required to ad- measuring GSH in human cells (especially tumor cells). The

equately stain cells for GSH content with MCB depend upon specificity of the various GST isozymes toward MCB as a

several factors, including the concentration of dye used, function of temperature needs to be more fully explored.

staining time, and temperature. Cook et a]. (96) have re- Figure 31.13B shows that rodent cells (Rat-I fibroblasts)

cently reported that 50 ýaM MCB are sufficient to label most exhibit more rapid and higher MCB staining with lower con-

of the GSH in rodent cells; however, it appears that the Km centrations than human lymphocytes, which is consistent

toward MCB for human GSTs is much higher, suggesting with Cook et al. (96). Note, however, that appreciable dif-

that an appropriate concentration may be closer to I mM ferences between 60 IM and 1000 ILM MCB still remain,

MCB. Moreover, there are different isozymes of GST in dif- and that the temperature used for incubation has a smaller,

ferent cell types; classes described include at, a, and - but still reproducible, effect than is seen with human T-cells.

forms, grouped according to their mobility on isoelectric fo- Thus, for both rodent cells (Rat-i fibroblasts) and human

cusing gels. The GST isozymes are homo- and heterodimes cells (PBL), the concentration of dye used can substantiallyof peptides coded from multiple genes (105). Heterogeneity affect the amount of fluorescence seen on the cell sorter. The

in the expression of the different isozymes of GST in differ- choice of a dye concentration depends upon the cell type be-

ent cell types, coupled with the fact that different GST ing used, the Km for MCB for the GST isozymes of those

isozyme classes show different reactivity toward MCB cells, and the proportion of GSH one wishes to derivatize

(GST-~IL > GST-a > GST--7) (96, 101), means that one within the cell. Even with these higher concentrations of

must be very careful to substantiate flow cytometric data MCB, not all of the GSH has been derivatized. The basis for

with other conventional biochemically based assays of GSH this incomplete derivatization has been attributed to the inhi-

(see below). This caveat is especially relevant when examin- bition of GST by the GSH-bimane conjugate (96).

ing human tumor cells, many of which express primarily Nonetheless, we have found that lower concentrations of

GST-7r, the isoform with the lowest reactivity toward MCB. MCB (60 p.M) may be used with human lymphocytes to re-

Caution is also appropriate in the use of MCB for m - flect qualitative differences in GSH level. If one wishes only

ment of GSH in human lymphocytes, since there is differen- to determine if there is a correspondence of some parameter

tial expression of a GST-w form (GST-X), depending upon with relative GSH content, and not the exact amount of GSH

culture conditions and growth state (106), and since GST-I. present, these lower concentrations of MCB may suffice (see

expression has been shown to be polymorphic in PBL from calibration section below).

humans, as indicated by reactivity toward trans-stilbene ox-ide (T$O) and benzo-a-pymne-4,5-oxide (BPO) (106-110). PRACTICAL CONSIDERATIONS FOR STAINING WITH MCBIn fact, if GST-t. shows the same polymorphic distribution Stock solutions of the dye are prepared either in ethanol orIn the human population toward MCB as it does towards DMSO, such that at the final working concentration the total


400

30

20r

.+ 6O0iA 370

-0 16uM 21*

0 10 20 30 40 s0 60 70

.0

10 RAT- FIBRO LAST I240

0 10 20 3_40 0 604 70

TIME..... (MINUTES)3F~~gure~3 311..nluneo.tmprtread.ocetaio.n.h..- m ve.ro.h.m cin...a...ut ftes elsw s rnsee

neiso fursecedvlpen nMSstie uanTck t aohrtbecnanngMBta hdbe el2tteaprp0t

detuermnn the1 Inflence auoflurseteclswr ucl e temperature an ocnrto ntekmvdefrendencchie for thi ractiqono thes anl did 't-ifbbanstes.

nolume d lofrtescethnoedvlorpmen esnt enxcesied huma of te doantepend cntiing upo h ctivite hand specinicilieao the vapporiouscell suspenobsion. THusmraens Tcan beo storedat door inete is enzmperauehs ofb wTamon q plcedl tyaes Hnowv, thermcie mayd

ld aftefr up nefd mouth to. Calls are e icbthed wit ath deom alursoenbe pooelsofmentHmunaviabred fovr cjgtion. by shiws rea-siedcmpeeaturezor of Mainmed ium,30Cwae forh the plaedngth fof gstaine humact Tcells, 8suhow aswthin fibrbatsBohoeeldr pe. otieadeirkw aed, catl sCorte 37pC. After soutainng cerp.s Haleah kept o loesec tafserrt hngiedwt t10ont Ixcig andpoete fro m lnmsight uetecte atayi is048 per Afted on M thFan hngie t6 L ua -tsoe raetheflerir cthmetbaer.i -As1 wit MCD CMD rc.thcalweeqye-emrtuedpnnc on hs ec~nthatinsdi rare expmetedtob

vlmeD Is tnoethnolorest butO whesnconjugated tof the bydelltpenspciic. Wpnte hactvitesn speviiiiousl reofted vaprioenscellI Is cemon.7eaed trteafuoesents ca esorediatOine prdc.wthe Rat-ime ofiobasmousng thsraetcell s tyes soeetaied inyThis for dup to Ine month Cells, art sincuaed withdeeur the de-also wbe pol o0 G sounavionafnl) for 10mco onjuainbytiger

OST, ku bf fomm the fluorescent pOduct aredct waibe ifh RonspeIficr t estenge athvtyis low)aget. Cell mestocne inu

Ws reacdon own in.ost........u... inc.it.does.require.te.dark..ith.a..0...solu.io.........fr.10 min.(or .ori-

524 CHAPTER 31: MEASUREMENTS OF CELL PHYSIOLOGY

S50" Since many flow cytometric applications of the use of"MCB will be concerned with relative differences in GSH

C 45- content among different cell populations, a highly relevant-40 calibration procedure is to examine cells with different GSH"C) content by both FCM and HPLC or biochemical assay. MCB35- is easily used to discriminate between the degrees of GSH

depletion after treatments that oxidize or deplete intracellular

GSH. Figures 31.15A and 31.15B show a series of histo-2 25- grams of human CD4÷ and CD8÷ lymphocytes pretreated

LL with various doses of 1-chloro-2,4-dinitrobenzene (CDNB; a0 ¶0 20 30 GSH depleter). A dose-dependent reduction in MCB fluores-

cence (after correction by subtraction of autofluorescence) isGSH (nmol/mg Protein) by HPLC seen with increasing concentrations of CDNB in both T-cell

Figure 31.14. Correlation between GSH content measured by subsets. The measurement of the relative decline of MCBHPLC and MCB/flow cytometry. Chinese hamster V79 cells were fluorescence as a function of CDNB concentration used fortreated with 0, 1, 3, 6, 9, or 12 mM BSO for 12 hr. Cells were then MCB depletion reveals that the lower concentration of MCBprocessed for flow cytometry (45 mM MCB for 10 min at 37"C) or for (60 I.M) shows a greater relative depletion with increasingHPLC determination of total reduced GSH content. (Reprinled with doses of CDNB than does staining with 1 mM MCB (Fig.permission from Kavanagh TJ. Martin GM, Livesey TC, RabinovitchPS. Direct evidence of intercellular sharing of glulathione via meta- 31.15C), and a closer correspondence with GSH concentra-bolic cooperation. J Cell Physiol 1988;137:353-359.) don, as determined by HPLC (Fig. 31 .15E). This suggests

that a greater proportion of the MCB staining at I mM is

ions are made up with 5 mM1 in DM50. If the cells are to be attributable to nonspecific staining of protein thiols and that,stained in medium, care sadeupwiho 5 rtaken n toe esre or ex- in spite of the lower proportion of total GSH stained with 60clude serum in the medium, since serum contains variable pLM MCB (Fig. 31.13A), this staining concentration mayestera u activitynrt e mium sincthe reagent exracellularly yield a more linear relationship to alterations in GSH contentesteracell acivi hat wfonrt. tin this cell type. Figures 31.15D and 31.15F show 25% toto a cell impermeant form. 40% CDNB-resistant staining with MBB (higher with higher

MBB concentration), a result that is consistent with theCalibration and Data Analysis staining of other non-GSH thiols by MBB. For each staining

CALIBRATION BY HIGH-PERFORMANCE LIQUID protocol used in Figure 31.15, there was little difference be-CHROMATOGRAPHY (HPLC) AND TIETZE ASSAY tween CD4÷ PBL, CD8÷ PBL, and CD4-CP8- PBL (pri-

Since MCB fluorescence is dependent upon GST activity, marily B-cells). Data, such as that in Figure 31.15, suggestand since CMIFDA stains thiols and is not necessarily spe- that, while absolute and intercell-type comparisons of GSHcific for GSH (although under most circumstances GSH is using MCB may be more difficult, measurement of relativethe major low-molecular-weight-thiol in cells), it is impor- differences may be much more reliable in human lympho-tant to calibrate the level of fluorescence measured by flow cytes. This result is in agreement with the study of rodentcytometry with a more specific biochemical measure of cells by Ublacker et al. (101), but is in contrast with theGSH. We have used H.PLC analysis and various doses of the results they obtained with human tumor and monkey cells.GSH-depleting agent buthionine sulfoximine (BSO; an in- The latter showed disparity between flow cytometric GSHhibitor of -y-glutamylcysteine synthetase) to ascertain the quantitation and biochemical assay. As described previ-correlation between MCB fluorescence and GSH content in ously, this may be related to the predominance of the GSH-,'rmatched samples of Chinese hamster V79 cells (112). For isoform in these tumor cells, which has the highest Km forthis cell type, there is a very good correlation between mean MCB. Thus, while MCB fluorescence can, in some cellcellular fluorescence by flow cytometry and GSH content by types, yield excellent correlation with GSH, this may not beHPLC (Fig. 31.14). the case in other cell types, and reliance upon the flow cyto-

In order to determine MCB specificity for GSH, one can metric determinations should be made only after carefulsimply run stained cell lysates over an HPLC column with study of the relationship of MCB fluorescence to indepen-fluorescence detection and compare with known low-molec- dent assays of GSH content.ular-weight-thiol standards that have been derivatized withMBB (70). When this is done, we find that >99% of the GST ACTIVITYMCB fluorescence is associated with the GSH peak in pro- As discussed above, heterogeneity among cell types existswein-free cytosolic extracts of normal human PBL. with respect to the total GST activity, the isozymes ex-

Another method of calibration is the Tietze recycling as- pressed, and their Km toward MCB. It is therefore importantsay (113) or a modified version designed to handle multiple to address the question of GST activity toward MCB in asamples ina microtiter plate (114). particular cell type in order to insure that one is using the


10A. CD4 T CELLS B00 B.C T CELLS

so 4S IS 5 Control so8 451i5"'5 Control

z 60 z 60

*40 I 5 40

~20 CC 20

.5 1 2 5 10 .5 1 2 5 10MCB Fluorescence MCD Fluorescence

123,100 1009

oso0 80

0 GO0 so0a C

0 A.

*40 ~ 0 _ _ _ _

o2 5 O',w.m....... 101 02 03 04 00 S1 52 53 54 55

C;DNS lug/mi) C:ONB lug/mi)

E 19F .-

60~ -7- 60

aa

20 it.20

0 5 10 19 20 25 30 0 5 10 15 20 25 30

JOSH) (nM/mg Protein) JOSH) (nM/mg Protein)

Figure 31.15. Effect of treatment with 1-chloro-2.4-dmintrobenzene for CD40 cell (diamonds). CD6* cells (ckcbes) and CD4-.CDS-(CONS) on the GSH content of human CD4# and C08- PIBL as deter- cells (tdanglss). regardless of which concentration of MCB or MBBminted with MCM/Bow cylornetry and HPLC. A and B show MCS (60 was used. E compares the level of GSH-bimnane conjugate (as deter-ILM) fluorescence hi~storms ol CD4* and CD6* Cobl pretreated with mined by HPLC: In PBL stained with I mM MBS) with the mean relaCDNS at 0.,5, 15. and 45 mghmI for 15 mnin at 374C. C shows the ttvo fluorescence of CD4* cells, CD8 ceask and CD4-,CDS- celleffect of using two differedtcnetain of MCS (0MA closed stained with eWhe 00 ILM or I mM MC&. F shows a simrillar cornparl-symbols, end I mM, open symbol) on the measuemert of CONS- son for 60 pa.M ro I1mM MBB. Autofluoescence of unstained aesk hasInduced de-mo in fluorescenc as a percentage of the fluoresence been subtracted from all fluorescence Intensitis plotted in panels C-of fth nrrsa* ed cntols.a! D shows a simia analysi for MBD (60 I&M F.and I mid). There is a sknda proportionate decline in fluorescence


Z 48 .. c

S . ! : •" : ..' , .'to 08

WI

322

t6 4S6 3 S

00

U -J

TIME (Min.) "4 "

""ofMDcnuaini human CD ymhcte yf"wcyoery"

10 as

sa =',6 :,ii=

TIME (Min) 10

(3')tb otiig CSsc httefna ocnrto a 26 46 50 *m 1

m i su be was qui cp ced back in te l sor rn d the ainal- L280,

the meaST cnuactivnity froma sch dataobuteaso an inicton ofitrcetrieyy Clo yometry. Alhwoabseit ytga*cellularoheterog n denity integrTactivity. Trmpeiherman blCd weures blgeletoi ouev.lgfursec)o a- irbat tie

cencepende es idan that th concentration of

10H

in ths c:

istihty o ntold needn fcl ie n hwhsowarmedg 3rams. Afflurresesnenceldaing cor rectedne for) flm a c

pop rb con stagatin inaln concentrationtincbation wtime,. a1 rag

ind Temps ature. ed Wton nt tahents ( o115 an d Coeo an. (96)

ysshave describued frathetus of MC ton assesno GSTy cancdterinei in u*3.7 eainhpbtee elsz Hcnetachells.an exTampcfthvi sty p ofo scd ana talyssoa isdipresented ine- dtrie yM~t cytomety. We haefonshasiutnous a measuriaementg

Figulret31.6.nells in susensi ary Thel mat ro tepera- opr electronic cell volume ao f of of fra sta/nedreopewarmed to 37.tiab n a baseline autofluorescence yor ied data th a cn eas i b d tedf cell volue bian os-

rptdan d a predterine aliquot of thee cells is hn rnferdt addewred t o ohvlm n C loecne n a oeesl

(r meuropwm tube cntaining MCB. disinuihtbtwencotr concentration lo in ted cell smi. hstuensio is thenlly quickly oanatye forf a tn oe t C n a onf volu se b n her tofluoresence over tiitrme (sucally 10 m lso atnthe preedi otera- fluorescence to log v olmet, a l ows vue taria te ior-

pelarhturoe neitya rathes of T fluo visceTc e acmulatin aCBeflnoic- (fr eetrnic thtvolume.ld flothrwisenbe ompliated fby olume stief-

volumeo respectively. The logarithmic 3xes correspond to a three-dec-

appropriate MCB staining concentration, incubation time, adi range.and temperature. Watson et a]. (115) and Cook et a). (96)have described the use of MCB to assess GST activity incell. An example of this type of analysis is presented in cytometry. We have found thet simultaneous measurementFigu e 31.16. Cellts in suspension are held at room tempera- of GSH-imependent fluorescence and electronic cell volumeturee or prewarmnd to 37"C, and a baseline autofluorescence yields data that can easily be adjusted for cell volume bias inis established for 30 sew to I rain. Analysis is briefly inter- GSH content (Fig. 31.17,A). By acquiring data on log scalerupted tand a predetermined aliquot of these cells is added to for both volume and MCB fluorescence, one can more easilya room temperature or prewarmed tube containing MCB. distinguish between control and GSH-depleted cell popul•-

The suspension is then quickly analyzed for accumulation of dions. Compensation for volume by analyzing the ratio of logfluorescence over time (usually 10 rain) at the preferred tern- fluorescence to log volume, allows one to appreciate dif-

pertur. Initial ratee of fluorescence accumulation are indic- frences that would otherwise be complicated by volume ee-tire of G " activity. fects (112) (Figurs 31.17B and 31.17C). MCB analysis on

D/• NALSISinstruments that rely on sizing of the cells with forward-an*D= ANLYSISgle light scatter can be made either by gating on a specific

Stim GSH concentration but not absolute content, is regu- cell size to minimize volume effects, or by displaying bivari-lated in the cell, it is important to compensate for heteroge- ate histograms of the ratio of log fluoresce=c to log forwardneity in cell volume when assessing cellular GSH by fDow scatter.

•.. X


100 G$

'0 HIGH MCB SORTLu so GO io

7060

I--

O-0 102 0-0 808O0N N

LU 40

-J 30

0.

0 0-10 10-20 40-60 80-100DNRA

PERCENTILE OF MCB FLUORESCENCE

Figure 31.18. Proliferation of human CD4" lymphocytes sorled onMCB fluorescence intensity. Human lymphocytes isolated from pe-ripheral blood by density gradient centrifugation were stained withMCB (60 ýLM for 15 min at 37C) and PE anti-CD4 mAb. CD4" cellswere then sorted according to their MCB fluorescence for the lowest B LOW MCB SORT10%, 1Oth to 20th percentile, 40th to 6Dth percentile, and the highest

20% of the fluorescence histogram. Cells were then plated into 96-well plates precoated with anti-CD3 mAb in BrdU-containing medium.After three days, cells were harvested and assessed for proliferationby the BrdU/Hoechst method (129). There is a direct correlation be-tween the MCB fluorescence intensity of the unstimulated cells andtheir subsequent ability to proliferate. (Data kindly provided by Dr. A.Grossman.)

Role of GSH in Cell Activation

A large number of studies have shown that GSH is importantfor cell activation and replication. Most of these studies re-lied on the disruption of normal GSH homeostasis with drugs Fgur 31.19. Effect of GSH status on the ability of C54* cells to

that either deplete or enhance GSH content in order to show become activated by anti-CD3 mAb, as determined by RNA synthe-a GSH dependence for cell growth or function (69, 116- sis. Cells were stained with 60 ;.M MCB and sorted on GSH content

124). (upper and lower 30% of fluorescence histogram distribution), allowed

An alternative approach to the study of lymphocyte acti- to recover for 24 hr. and stimulated with solid phase anti-CD3 mAb.After 48 hr, cells were fixed in ethanol, stained with acridine orange.

vation is to sort on the basis of GSH content and then to and analyzed by flow cytometry for RNA and DNA content Cells with

assess growth in the sorted cells after they are allowed to high GSH (high MCB sort) show the ability to exit G, (A), as demon-

recover to their pretreatment levels of GSH (70). There is a strated by elevated RNA content in G,. S, and Go cells, whereas cells

direct correlation between initial GSH content in unstimu- sorted for low GSH content have nearly all remained in Go (B). (Re-printed with permission from Kavanagh TJ, Grossman A, Jaecks EP,

lated CD4÷ human lymphocytes and the capacity for subse- et al. Proliferative capacity of human peripheral blood lymphocytes

quent cell proliferation (Fig. 31.18). In an attempt to ascer- sorted on the basis of glutathione content. J Cell Physiol 1990;145:tain the GSH-sensitive steps in cell proliferation, we stained 472-480.)cells sorted on their GSH content with acridine orange inorder to assess their RNA and DNA content simultaneously; from the figure, this reagent, which effectively depletesthis allows one to distinguish activated G, cells from those GSH (Fig. 31.20B), as well as some protein thiols (126),that remain quiescent (Go) (122). Figure 31.19 shows that dramatically reduced calcium signaling in CD4÷ cells. Incells with low GSH remain in Go after mitogen stimulation, contrast to the results with T-cells, B-cells pretreated withsuggesting that there is an early step in cell activation that is NEM do not show similar reductions in calcium signalingblocked in low GSH cells. (data not shown), indicating that these two classes of lyrn-

The effect of GSH content on very early steps in trans- phocytes have very different thiol-redox requirements formnembrmne signal transduction can be directly evaluated by their activation. To direcdy assess the relationship of GSH tousing indo-l to assess mrogen-induced changes in intracel- transmembrance signaling, we have sorted CD4+ cells onlular calcium. Figure 31.20 shows the effects of prefetating the basis of their GSH content and subsequently evaluatedCD4* cells with n-ethylmaleimide (NEM) on their ability to mitogen-induced changes in indo-1 ratio. As shown in Fig-respond to stimulation with anti-CD3 mAb. As is apparent um 31.21, cells with high GSH show more vigorous [Ca&2 ,

i~r

528 PART II : CLINICAL APPLICATION

Effect of NEM on GSH and Ca

- 100O CD4+ MCB FLUORESCENCEsaoz (GLUTATHIONE CONTENT)0

O 603 gmlAT-D

o

0 40

J 20

A 0

n-2 0 oo-

0 CD4+ INTRACELLULAR CALCIUM

830 ug/mI ANTI-CD360

w 40

20

o0B 5 S 10 20 40

[NEMI (uM)

Figure 31.20. Effect of depletion of intracellular thiols upon intracet- cells were stimulated with 30 W.giml anliCO3 , Ab and jCa'-I waslular calcium signaling. Indo-l-loaded cells were exposed to various examined. There is a progressive inhibition of the rise in jCa'* as aconcentrations of NEM. which progressively deplete GSH (A). The function of increasing NEM concentration (B).

responses than cells with low GSH. These results suggest the neously will surely facilitate the identification of drug-resis-possibility that certain components of the transmembrane tant tumor cells from biopsy material, allowing one to ascer-

signaling apparatus in at least some cell types are sensitive to tain relationships between aneuploidy and changes in GSHthiol oxidation. It is known, for instance, that p561ck con- content and/or GST activity.tains a zinc finger motif and this should be sensitive to oxi-dation. Freed et al. (126) have assessed the influence of Sorting on the Basis of GSH: Isolation of GSHmembrane thiol groups in lymphocyte proliferation, IL-2 Variantsproduction, and receptor expression, and have reported simi-lar findings. Surface thiol groups are also important for the Since MCB is a viable stain, one can sort cells on the basisacfivation of neutrophils and monocytes (127). of their GSH content. Lee and Siemann (128) have usedMCB and cell sorting to isolate human ovarian tumor cellsGSK Variations During the Cell Cycle: Simultaneous with increased GSH content that were also resistant to ad-GSHEVariationsof Duing the Cs riamycin. However, the cells reverted to normal GSH after

growth in culture over time. Using repeated rounds of MCB

The ability of CMFDA to be used as a thiol stain excited staining and sorting, a number of Rat-1 fibroblast variantswith visible light has made it possible to simultaneously that have stable alterations in their basal GSH content haveevaluate GSH levels and DNA content in viable proliferating been isolated (Fig. 31.23). These cells have not yet beenpopuatin of cells (104). Figure 31.22 shows a bivariate charactized as to the basis of their altered GSH level.cyegram of Rat-I filroblasts staine for GSH with C1FDA However, likely possibilities include mutations in the genesand DNA with Howehst 33342. As one can appreciate from coding one of the two enzymes responsible for GSH synthe-the figur, thee is a twofold increase in fluorescence in G: sis ('v-glutamylcysteine synthetase; glutahione stlihele).cells; this incwrease appemar consistent with their increase in differential expression of GST isozymes, or deficiencies incell volume. The ability to evaluaft GSH ant DNA simuha- cysteine muspt. High GSH variants may have mutadons

• ,, • ,. .• , ,,, .4.• I . _

CHAPTER 31 MEASUREMENTS OF CELL PHYSIOLOGY 529

Sorted High MCB Sorted Low MCBlse ..... lob i .............

A .A

oe ... ... ....

6 *ii0t 2 4

44 6 0

-. i566 B .. r566 E480 480

U ......... U

326 320-

..... ~ ....

..... ....C.. . . . . . Fo ...........-... .

... ..-.w ....... ............. .. .......... .......

Tim (m. T ..me.(..)Figue 2121. ffec of ntraelluar GH cnten on nti-D3 mb- o hig andlow SH.ontet, .specivel mer.is. dimn.sh. rstimlate trssinmbrae sinaltranducton a CD humn T ym- spone inceil wih lo G51 cotentB an.E sow ..e.man.[a..phocles.Huma PBLwer staned ith 0 rM MC andPE ati clcuatedfromthe ataPresntedIn Aand., rspecivel.Th.e. i

CDa- inrb CD cell weasre steupradlwr3%o h ihrma n ekcIu epnei h ihGHsreMCBflorececedisriutinalowe t reovr or 4 r ad he clestaentelwGH otdclsCadF hwtepretg

lode wt ethrIno. M r etane it MB Cll esaie o rspnin cls aluatd rm hedtaprsntd nA ndDwith~~~~~~~~~~~~~~~.. M... reaie thei orignal. soFdM fursec ifr epciey hr sa prcal ihrpretg fclstaence (nt sown. Cllsloadd wth ndoi wre nalzed or ~a' ar abe t resondIn he ighGSHsored cllstha.inthelow.S.content~~~~~~~~~~~~~.. ...w.n sti.lu wih ..-V3r~ a ecne nFg oldcls31.4).A and show he kintics o I~a' transEnt nclssre

....... ...

530 PART 11: CLINICAL APPLICATIONT

is.. ... .... ~ -. .

4.1

Uin

040

U

HOECHSTr FLUORESCENCE US04

lse.

4.

28_

4Lee

Figure 31.22. Simultaneous measurement of DNA and thiol content measured with UV excitation and blue emission. The temporarilyIn rat-I fibroblasts. Live cells were stained with Hoechst 33342 (10 delayed signal from CMF fluorescence was measured with 488-nmmM) for 30 min at 37'C. During the last 10 min of Hoechst stairing, excitation and green emission. Cen aggregates were gated from thethe ceab were stained with CMFDA (10 mg/mI). Cells were then ana- analysis by displaying Hoechst-42 fluorescence vs. peak area (notlyzed with a dual-laser cell sorter. DNA-specific fluorescence was shown).

20,

6 ~~~~~~Figure21.23. MCSflorscence ad ee~cir ~onellvolumeofrat-I* ~fibroblast GSH variants isolated after seveal rounds of MCB staining

0 and viable sorting. Rat-i fibroblasts were stined with MCS (45 ILfor 10min at 3MC. placed Inaceml sorter, and sorted on the highest

10 and lowest 5% of the fluorescienice histogmra. Cells were the allowedWto recve and grow to confluwene at whid time the process was

9repeated unti separate population od cell appeared. The cell were.00 t1hen kdvlduail sorted Into web of a 96-we ftisue culture plaite anid

grown from single call Flones. The figure represents mean volumeand fluorescence of thes variants after analysis on a FACS Analyzer.The noinselected parental cell are shbown as an open symbiol. Note

S I ~th'at dlore lyin above the lin frcm the origin through the parental20 40 so so 100 120 cebthae" bwerimracelulairMCS conicarratabs than the parental

cab (atough many such calsk had larer volumes), while doris be-MCB Fluorescence low tialne whave anir Inese raosbiArMC1S cocnration.

CHAPTER 31 : MEASURAEMENTS OF CELL PHYSIOLOGY 531

in the regulation of these geneCs resulting in increased 17. Martell AE, Smith RM. Critical Stability Constants. Volume 1:amounts of product, such t~hat there is an abnormally high Amino Acids. New York: Plenumn Press, 1968;269-212.

leve of SH sntheis o conugaton.18. Harafuji H. Ogawa, Y. Re-extamination of the apparent binding 'con-leve of SH sntheis r cojugaion.stant of etltvlene Slycol bis(bea-aminoealtyl ether)NN.N,.N X

letuaactic acid with calcium around neutral pH. I Biochem 1980.87:ACKNOWLEDGMENTS 1305-1312.

We tars J.. Lebeter nd . Grssmnn or mny aluble 19. Moisescu DG, Pusch H. A pH-metric method for the dieterina~tion ofWe ~than D.A. LEabtonfr arvdnd A. L Gomanalysormany valabl Jn th eltv concentration of calcium to ECTA. Pflugers Archiv 1975;

nernan and R.C. Lee for their much-valued technical assistance. 20. Miller DJ. Smith GL. EGTA purity and the buffering of calcium ionsThis work ws supported in parn by the Naval Medical Research and in physiologgical solutions. Am J Physiol 1994;246:C]60-C]66.Development Command, Research Task No. M0095.003-1007, 21. Overton. %R. Modified histogram subtraction technique for analysisand by the National Institutes of Health Grant AGO 175 1. The opin- of flow cytometric data. Cytometuy 1989;99:639-626.ions and assertions expressed herein are those of the authors and are 22. Rabinovitch PS, June CH. Intracellular ionized calcium, membranenot to be construed as official or reflecting the views of the Navy potential, and pH. In: Olmerod MG, ed. Flow C)ytometrv: A PracticalDepartment or the naval service at large. Approach. Oxford: Oxford Press. 1990;161-185.

23. La~zzAi KG, Proto P). Simons ER. Simultaneous measurement of

REERNCES stimulus-induced changes in cytoplasmic Ca:. and in membrane po-tential of human necutrophils. J Biol Chem 1986;261:9710-9713.

1I. Blinks JR. Wier WVG, Hess P, Prendergast PG. Measurement of Ca:* 24. Ishida Y. Chused TM. Heterogeneity of lymphocytec calcium metabo-concentrations in living cells. Prot Biophys Mol Biol 1982;40:1-114. lism is caused by a T cell-specific calcium-sensitivt potassium than-

2. Tsien RY, Pozzans T, Rink TV. Calcium homeostasis in intact lympho- aed and sensitivity of the calcium AT7ase pump to membrane poten-cytes: cytoplasmic free calcium monitored w\ith a new. intracellularly tial. I Exp Med 1988;168:839-852.trapped fluorescent indicator. I Cell Dial 1982,94-.325-3.-4. 25. Goldsmith MA, Weiss A. Isolation and characterization of a -Vym-

3. Tsien RY. A non-disruptive technique for loading calcium buffers and phocyte somatic mutant with altered signal transduction by the antigenindicators into cells. Nature 1981l;290:527-528. receptor. Proc Nail Acad Sci USA 1987;S4;6879-6983.

4. ansm I, D~iuto L, amber C. inge cll nabsisof al- 26. Goldsmith MA. Weiss A. Early signal transduction by the antigen re-cium mobilization in ansti-imnunsoglobulin-stiniulated B lymnphocytes. ceptor without commitment to T cell activation. Science 1999;240:I Imsmunsol 19960136:_91-57. 1029-1031.

5. Grynkiewicz G, Poensie M, Tuien RY. A new generation of Cal + 27. De Virgilio F, Steinberg TH. Swanson JA. Silverstein SC. Fura-2 se-indicators with gre~atly improved fluorescence properties. I Biol Chem cretion and sequestration in macrophages. A blocker of organic anon1985;260:3440-345.UOtrnprreelthtteepoessocrvsammaetasot

6. Mints A, Kao .IPY. Tsien RY. Fluorescent indicators for cytosolic cal-trnptrealththserosesccrvaam baetasotciumbasd o rhdamne ad fuorscen cromohors. 1 Bol hemsystem for organic anions. I Immunsol 1988;140:915--920.

cium;2basedo roain n fureci crmphrs.IBolCe 28. Malgawli A, Milani D. Meldolesi J, Pozzans T. Fura.2 measurement ofconcenration.9Methds Cel cytosolic free Ca' in monolayecrs and suspensions o( various types of7. Teien RY. Fluorescent indicators of ion animalatins cells.s Cel CelB 97lS24-15 -

Biol 1989;30:127-156. -2clsJCelBo197051425.S. VndeberlicPA.Ceupen X.Flowc)-onitri mesurmen ofcy- 29. Steinberg SF, Bileziltian JP. A]-Awqati Q. Fura-2 fluorescence is lo-3.cl; taonbrh PAconri ienpn enoLela celow cAomti measurement of cy-

toplasmic free calcium in human peripheral blood T-lyznphocytes with 1)cal744t-mtChodj7nedohla els7. lsilI975Pfluo-3, a new fluorescent calcium indicator. I Immunsol Methods 3.Al):C7P. D4-C74F7.lrntM esaR, o7n .Ctoo1990-,1217:197-205. 3.Asa .D~rii ,Blrn ,Tin~ oznT yooi9.Rikes TJustement LB. Griffioen AW, Camsbier 3C. Improved Ca2* homeostasis in Ehrlich and Yoshida carcinomas. I Biol Chem

9. thodjfors mesrn nrcTlrC-wt lu-.Ctne 1985;260:2719-2725.1e9o9 fr0 esuin1itacllla23-wih92u-7.ytmer 31. Ltaclcoff A. Measuring cytosolic free calcium concentration in endo-

10. Rabinovitch PS, June CH, Grossmnann A. Ledbetter IA. Heterogeneity thelial cells with indo-1: The pitrall of using the ratio of two fluores-among Tells in inmvcelhular free calcium responses after mitogen stimu- cece intensities recorded at different wavelengths. Cell CalciumlaIsn with PHA or anti-CD3. Simultaneous use of indo-lI and irmmtoflu- 1986;7:233-248.otescome with flow cytotnevy. J Immnunol 1986;137.952-961. 32. Scanslon M, Williams DA, Fay PS. A Ca2 +-insensitive form of furs-

11. Coben LB. Salzberg BM, Davila HV, et aI. Changes in axon fluores- 2 associated with pol)%wophonuclear leukocytes. Assessment and ac-cence during activity: Molecular probes of membrane potential. I curat C&2 + measurement. I Biol Chem 1997;,262:6308463)2.Membir Biol 1974; 19:1-36. 33 Ansasetti C, Martin P3, June CH, et al. Induction of calcium fluxa and

12. Poenie M4, Alderton 3, Steinhtardt R, Tseen R. Calcium rises abruptly enhancement of cytolytic activity in natural killer cells by cross-link-and briefly throughout the cell at the onset of anaphase. Science 1986; inS of the sheep ezythrocyte binding protein (CD2) and the Pc-receptor233:386-889. (CD16). J Imrnunol 1987;139:1772-3779.

13. Classed TM. Wilson HA. Greeniblan D. et al. flow c)Iomenric analy- 34. Geppert TD, Wacholtz MC. Davis LS. Upsky PE. Activation ofsis of musine splenic B lymphocyte cytosolie free calcium response to humans T4 cells by cross-hinkzing class I M1I0C molecules. J Izmrotnolandi-ISM and antl-JgD. Cytomeny 1987;8:396-.404. 19U;.140:2155-2164.

14. Davies TA, Drons D, Weil G), Simons ERt. Flow cytoinetic measure- 35. June CH, Rabinovitchs PS, Ledbetter JA. CDS antibodies increase in.ments of cytoplasmic calcium change in human platelets. Cytornetry bbcelular ionized calcium concentration in T cells. J Iminunol 1987;19":9-138-142. 138-.2782-2792.

15. MhIlleRA, Jacobson B, Weil G. Simons ER. Diminished calcium in. 36. Pemotto A, D~rken B, Rabinovitch PS, Ledbetter ]A, MoldenhauerOniu n lectin-stimulated T cells frmw old mice. I Cell Physiol 1987. 0, Clark EA. CD19 monsoclonal antibody HD37 inhiubits anti-immu-132:337-42. nosiobulin-induced B cell activation and poif~eration. J Iminunol

16. Parks DR. Nozaki T. Dunne IF, Peweson LL. Flow cylometer adapta- 1987;138:2793-2799.tion for quantitation of immunofluorescen masents And for calbration 37. PC==u A. Rabinsovitch PS. Dorken B. Moldssabmer 0, Caik LA.a(d~y for soessurinS cellular Ca2- and pH. Cytometry 1987;.(SuMp Rok of the cD22 human B cel antigen in B cell risggeing by anti-

l)104. hnnaumoliobulia. J Imnusol 19S88;88:1791-179.

532 PART If: CLINICAL APPLICATION

38. Wilson HA, Greeniblan D. Taylor CW. et ad. The B lymphocyte Ca]- 60. Brodie AF, Reed DJ. Cellular recovery of slyceraldthydc.3.phos.cium response to anti-lg is diminished by membrane immunoglobulin phate dehydrogenase activity and Nhol status after exposure 10cros-linkage to the Pc gamma receptor. J Immunol 1987;138:1712- hydroperoxides. Arch Biochemn Biophys 1990;276:212-219.1718. 61. Josh] TG, Schirch V. The role of a critical sulfhydryl group in the

39. Ledbetter JA, June CHI. Grosmaire L.S, Rabinovitch PS. Crosslinkirig mechanism of serine hydroxymethyltransferase. Arm NY Acad SOiof ofsuaet antigens causes mobilization of intracellular ionized caJ 1990;585:339-345.chinm in T lymphocytes. Proc Nat] Acted Sci USA 1987;84:1384-1389. 62. Nalecz KA, Mil0ler M, Zambrowicz LB. Wojtczak L, Azzi A. Signif

40. Ledbetter JA, June CH. Rabinoviich PS. Grossmnann A. Tou TT, I icance and redox state of SH groups in pyruvate carrier isolated fromboden JB. Signal transduction through CD4 receptors: stiunulatory vs. bovine heant mitochondria. Biochim Biophvs Act& 1990;1016:272..inhibitory activity is regulated by CD4 proximity to the CD3/T cell 279.receptor. Eur J irninuniol 1988;38:525-532. 63. Niroomand F, R6ssle R, MtIlsch A, Bbhme E. Under anaerobic can-

41. Poeriti M. Tsien RY, Schntin.Vecrbulst AM. Sequentia~l activation and ditions, soluble giaanylate cyclase is specifially stimulated by gluta.lethal hit measured by [Ca'*), in individual evtolytic T cells and thione. Biochem Biophys Res Commnun 1969,161:75-40.targets. EMBO J 1987.6:2223-2232. 64. Stun Y. Oberley L2W. The inhibition of cataase by glutathione. Free

42. Williams DA, Becker PL. Fay FS. Regional changes in calcium un- Radic Biol Med 1989;7:595-602.derlying contraction of single smooth muscle cells. Science 1987;235: 65. Suzuki M, Capparelli AW, Jo OD, Yanagawa N. Thiol redox and1644-1648. phosphate transport in renal brush-border membrane. Biochim Bi-

43. Ware JA, Smith M, Salzman EW. Synergism of plateet-aggregating ophys Act~a 1990;1021:85--90.agents. Role of elevation of cytoplasmic calcium. J Clin Invest 1967; 66. Zeigler DM. Role of reversible oxidation-reduction of enzyme thiols-80:267-271. disulfides in metabolic regulation. Ann Rev Biochem 18;435-

44. Cobbold PH, Rink Ti. Fluorescence and biolumidnescence measure- 329.went of cytoplasmic free calcium. Biochem J 1987;248:.113-3:8. 67. Leung NIP, Chou IN. Relationship between 1-chloro-2.4-dinitroben.

45. Ambler SK, Poenie M. Tsien RY. Taylor P. Agonist-stimulated oscil- zetindued cytIoskceleaJ perturbations and cellular glutathione. Celllations and cycling of intracellular free calcium in individual culturedBilTiol185S6.muscle cells. .I Biol Chem 1988;263:1952-1959. Bo oio 9955-6

46. Wilson HA, Greenblatt D. Poenie M, Finkelman FD, Tsien RY. 68. Messina JP, Lawrence DA. Cell cy cle progression of glutathione.de-Crosslink-age of B lymphocyte surface imrnunogl)obulin by anti-Ig or pleted human peripheral blood mononuclear cells is inhibited at S

antgenindcesproongd ocilatin o inraclluar onied alcum.phase. J Iminunol 1989;143:1974-198).Saptge Mnued 1 rloge87:166:601-606. ~ la iniedcaci~n 69. Suthanthiran M, Anderson ME, Sharma V'K. Meister A. Glutathione

47. June CH. Rabinovitch PS. Flow cytometric measurement of cellular rgltsatvto-eedn N ytei nhgl uiidnrionized calcium concentration. Pathol Imimunopathol Res 1988;,7:409L- mal human T lymphocytes stimulated %-is the CD2 and CD3 antigens.432. Proc Natl Acad Sci USA 1990;87:3S42-3347.

48. Rabinovitch PS, June CHi. Measurement of intracellular free calcium 7.Kvng J rsmn ,Jek P ta.Poieaiecpctand membrane potent~ial. In: Melamred MR, Lindmo T, Mendelsohn of human peripheral blood lymphocytes sorted on the basis of gluts-MhL., eds. Flow cytometry and cell sorting. Second Edition. New thione content. J Cell Physiol 1990;145:472-'80.York: Wiley-Uss. 1990; 651-668. 71. Meister A. S-Oxoprolinuria (pyroglutamic aciduria) and other disor-

49. Grossmann A. Rabinovitch PS. Flow cytometeoy with indo-l reveals ders of the -y-glutamyl cycle. In: Wyngaarden JB. Stanbury lB. Fred-variation in intracellular free calcium waithin T-ccil subsets and be- rickson DS, Goldstein XL. Brown MS. eds. Metab~iic basis of inher-tween donors after mitogen stimulation. In: Burger G. Ploem JS , ited disease. New York: McGraw Hill, 1983;348-359.Goeritler K, eds. Clinical cytometry and histometry. London: Aca- 72. Smtyth MI. Glutathione modulates activation-dependent proliferationdeinic Press, 19970192-194. of human peripheral blood lymphocyte populations without regulating

50. Ledbiener ]A, Rabinovitch PS, Hellstr6m 1, Hellstr6m KE, Grosmaire their activated function. J Immunol 1991;146:1921-1927.LS5, June CHi. Role of CD2 crosslinking in cytoplasmic calcium re- 73. Liang SM, Liang CM, Hargrove ME, Ting CC. Regulation by gluta-sponises and T cell activation. Eur J Imrnunol 1988; 18: 1601-1609. tWone of the efect of lymphokines on differentiation of primary acti-

51. Liette GP, Hartzman RI, Ledhener JA, June CHi. {VV- I infected T vated lymrphocytes J Immunol 1991;146:1909-1913.cells exhibit a selective trantsembrane signalling defect through the 74. Drfige W, Ponmeyer GC, Schmidt H, Nick S. Glutathione augmentsCD31antigen receptor pathway. Science 1988;241:573-576. the activation of cytotoxic T lymphocytes in vivo. Immumobiology

52. Thomas, RC. Bicarbonate and pHi response Nature 1989;337:60). 1996;172:151-156.53. de Grooth BO, van Dam M. Swart NC, Willeinsen A, Greve J. Multi- 75. Gmnflnder H. Roth S, Eck HP, Gallas H. Mihin S, Dr6ge W. In-

plc wavelength illunination in Dlow cytometry using a single arc lamp teedeukin-2 mRNA expression, lyinphokinec production and DNA syn-and a dispersing element. Cytornetry 1987;8:445-452. thesis in glutathione-depleted T cells. Cell Immuniol 1990;130:520-

54. VaWe G, Raffacl A, Moroder L, Wllnsch E. Rubenst-oth-Bauer G. 58Fast intracellular pH determination in single cells by flow-cytoinety. 76. Liaing CM, Lee N, Cattell D. Liang SM. Glutathione regulates in-Naturwissenschaften 1961;68:265-266. terletikin-2 activity on cytotoxic T-cells. 3 Biol Chem 1989;264;

55. Whitaker IL, Haugland RP, Prendergast PG. Seminaphtho-fluores- 13519-13523.ceins wan -rhodafluors: dual fluorescenice pH indicators. Biophys J 77. Candon A. Lairivee P, Begini R. Extracellular glutathione suppresses1988-,53:197&. human lung fibroblast proliferation. Am J Respir Cell Mol Bidl 1990.

56. Thomas IA. Buchsbaum RN. Ziminzak A, Ricker E. Intracellular pH 2.7945.measurements in Ehrlich ascites tumor cells utilizing spectroscopic 78. Fetscb J, Maurer HR. Glutathione: An in vitro granulopoiesis inhibitorprobes generate in situ. Biochemistry 1979; 18:2210-2218. at nanomolar concentration, isolated from calf spleen. Lap Hematol

57. JMeisu A. Glutathione metabolism and its selective modification. I 1990;.18:322-325.Dial Chem I9M;263:17205--17208. 79. Abate C, Patel L. Rauscher F] M.I Curnu T. Redox regulation of fat

S8. Meister A, Anderson ME. Olutathione. Ann Rev Diochem 1983;52: and jun DNA binding activity in vitro. Science 1990-,249:1 157-1 161.711-760. 80. Hentze MW, Rousult TA, Harford ID, Klautiser RD. Oxidawtio-re

59. Read Dl. Glutaithitoce: Toxicological implications. Annu Rev duedoo and the molecular mechanism of a regulatory RNA-proteinPhumaco Taxiedl 1990*.3(603-631. ineraction. Science j989;244.357-359.


81. Keyse SM. Applegate LA, Tromvoukis Y, Tyrrell RM. Oxidant stress flow cytometry using monochlorobimane. Cancer Res 1991;51:1783-leads to traniscuiptional activation of the human beae oxygenase gene 1788.in cultured skin fibroblasts. MoD Cell Biol 1990;10:4967-4969. 102. Woronicz 3D, Rice GC. Simple modification of a commercial flow

82. Najita L, Sarnow P. Oxidation-reduction sensitive interaction of a cyftometer to trtple laser excitation. Simultaneous five-color fluores-cellular 50-kDa protein with an RNA hairpin in the 5' noncoding; re- c~ence detection, J 1mniunol Methods 19890120:2291-296.gion of the poliovirtis genome. Proc Nail Acad Sci LISA 1990,87: 103. Radkowsky AE, Kosower EM. Bimanes. 17. (Haloalk-vl)-1 .5-diazabi-5846.-5850. cycloI33.0] octadienediones (halo-9,1I0-dioxa-birnanes): Reactivity to-

83. Roederer M, St~aal F. Rajti PA. Ela SW, Henzenberg LA, Herzenberg ward the tnpeptide thiol, Clutathione. J Am Chem Soc 1996;108:LA. Cytokine- stimulated human immunodeficiency vmirs replication 4527-453).is inhibited by N-acetyl-L-cysieine. Proc Nail Acad Sci USA 1990;87: 104. Pont M. Kavanagh TI, Kang HC, Haugland RP, Rabinovwitch PS.489"-888. Flow cytorritric analysis of cell cycle-dependent changes in cell Whol

84. Rokutan K, Thomas JA, Sies H. Specific S-thiolation of a 30-kDa level by combining a new laser dye with Hoechst 33342. Cytomeurycylosolic protein from rat liver under oxidative stress. Etir J Biochem 1991;12:394-187.1999;179:233-239. 105. Picket CB, Lu AYH. Glutathione-S-transfirases: Gene structure regu-

85. Storz G. Tarusalia LA, Ames BN. Tanscriptional regulation of oxida- lation and biological function. Annu Rev Biochem 1999;58:743-764.tive suress-inducible genes: Direct activation by oxidation. Science 106. Jones SM, Brooks BA. Langley SC, ]dle JR, Hirom PC. Glutathione1990;248:189-194. tzansferase activities of cultured human lymphocytes. Carcinogenesis

86. Cantin A.M, Hubbard RC, Crystal RO. Glutathione deficiency in the 1988;9:395--398.epithelial lining fluid of the lower respiratory tract in idiopathic pul- 107. Jornes SM, Idle JR, Hirom PC. Differential expression of glutathionemonary fibrosis. Am Rev Respir Dis 1989;139:370-372. transferases by native and cultured human lymphocytes. Biochem

87. Cantin AM, North SL, Hubbard RC, Crystal RG. Normal alveolar Pharmacol 1988.37:4566-4590.epithelia] lining fluid contains high- levels of glutathione. J AppI 108. Scideglrd J, De Pierre 1W, Birberg W. Pilotti A, Pero RW. Charac-Physiol 1967;63-.152-157. terization of soluble glutathione transferase activity in resting mono-

88. Buhl R. Jaffe HA, Holroyd KY. et a). Systemic glutathione deficiency nuclear leuk-ocytes from human blood. Biochem Pharrmacol 1964;33:in symptom-free HIV-seropositive individuals. L.ancet 1989.2:1294- 3053-3058.1298. 109. Seideglrd J, Pero RW, Markowitz MM, Roush G. Miller DG, Beattie

89. Buhl R, Jaffe HA, Hoiroyd KI, et al. Glutathione deficiency and HIV El. Isoenrymne($) of glutathione transferase (class Mu) as a marker for[Letter). Lancet 1990,33i:5146. the susceptibility to lung cancer. A follow up study. Carcinogenesis

90. Eck HP, Gmtinder Ai. Hartmann M, Pet~zolds D, Daniel V. Dr~ge W. 1990;11:33-36.Low concentrations of acid-soluble thiol (cysseine) in the blood 110. Wienckec JX, Kelsey KT, Lamnela RA. Tosc~ano %%'. Geneticplasma of JIIV. I-infected patients. Biol Chem Hoppe Seiler 1989; polymorphisms in carcinogen metabolism predict substrate-induced370:.101-108. cylogenetic damage in humans. Prog Clin Biol Res 1990.340B: 137-

91. Kalebic T. Kinter A, Poli G, Anderson ME. Meister A, Fauci AS. 147.Suppression of human immnunodeficiency virus expression in chroni- 1ll. Seidegitd I, Pero RW, Stille B. Identification of the trans-stilbenecally infected monocysic cells by glutathione. gltathione ester, and oxide-active glutathione trasferase in human mononuclear leuk-oc.%tesN-acetylcysteine. Proc Natl Acad Sci USA 1991.88:966-99. and in liver as GSTI. Biochem Genet 1989;27:253-261.

92. Durand RE. Olive PE. Flow cytometry techniques for studying cellu- 112. Kavanagh T), Martin GM, EI-Fouly MIH, Trosko JE, Chang C-C,lar thiols. Radiat Res 1983;95:456-470. Rabinovitch PS. Flow cytometry and scrape-loading/dye trn~sfer as a

93. Poet M, Verlcerk A, Koster IF, .longkind MF De novo synthesis of rapid quantitative measure of intercellular communication- in vitro.glutathone in human fibroblasts during in vitro ageing and in some Cancer Res 1987-,47:6046-6051.metabolic diseases as measured by a flow cysometric method. Bi- 113. Tietze F. Enzymatic method for quantitative determination of na-ochim Biophys Actit 1986;883:580-594. Dogram amounts of total and oxidized glutathionie. Application to

94. Treumner J, Valet G. Flow-cytometric determination of glutaithione al- mammalian blood and other tissues. Anal Biochem 1969;27:502-522.teratiotis in vital cells by o-phthaldialdehyde (OPT) staining. Exp Cell 114. Baker MA, Cerniglia 01, Zaman A. Microtiser plate assay for theRes 1986;163:518-524. measurmenirt of glutuashione and glutathione disulfide in large numbers

95. Rice GC, Bump EA. Shrieve DC, Lee W, Kovacs M. Quanitistative of biological sample. Anal Biochem 1990;190:360-W6.analysis of cellular glutathione in Chinese hamster ovary cells by flow 115. Watson JV, Dive C, Workman P. Measurement of dynam-ic cellularcytometry utilizing monochlorobimane: Some applications to radiation events. In: Ornerod MG, ed. Flow cytometry: a practical approach.and drug resistance in vitro mad in vivo. Cancer Res 1986;46:1-6. Oxford: IRI. Press, 1990,241-264.

96. Cook ]A. Type SN, Mitchell lB. Differential specificity of 116. Fidelus RK. The generation of oxygen radicals: a positive signal fornaonochilorobimane for isozymes of human and rodent glutathione s- lymphocyte activation. Cell Immunol 1988;113:175--182.transferases. Cancer Res 1991 ;51:1606-1612. 117. Pidelus RK. Ginouves P, Lawrence D. Tsan MF. Modulation of inu-a-

97. Cook JA, Pass F9. Russo A, lype S, Mitchell II. Use of mono- cellular glutathione concentrations alters lymphocyte activation andchlorvkimane for Slutathione measurements in hamster and human tu- proliferation. Exp Cell Res 1987;170:269-275.mor cell lines. Int I Radiat Oncol Biol Pltys 1989;16:1321-1324. 118. Fidelus RK, Tsan MF. Enhancement of intracellular glutathione pro-

98. Hedley DW, Hallahan AR. Tripp EH. Flow cytomettic measurement motes lymphocyte activation by mitogen. Cell Immnunol 1996;97:155--of glutathione content of human cancer biopsies. Sr J Capcer 1990;61: 163.65-.68. 119. Fischman CM, Udey MC, Kurtz M. Wedneir Hi. Inhibition of lectin-.

9.Kavanagh TI, Martin GM. Uvesey IC, Rabinovitch PS. Direct evi. induced lymphocyte activation by 2-cyclohexcne- I-one: Decreased in-decem of intercellular sharing of glutathione via metabolic coopera. tracellular glutathione inhibits an early event in the activation se-ti... I Cell Playsiol 1988;137:353-359. quemc. 3Ilmnmunol 19981127:2257-2262.

100. Sbnleve DC, Bump EA, Rice GC. Heterogeneity of cellulair gluta- 120. Hamllos DL, Wedner HW. The role of glutathione in lymphocyte acti-thics. among cells derived from a murine fibrosarcoma or a human vaiorn. 1. Comparison of inhibitory effects of buthionine sulfoximinemeal cell carcinma detected by flow cytoasetrc analysis. J Biol and 2-cyclobexene-l-ane by nuclear size transformation. I lmmtmolOsemt 1988;263:14107-14114. 1985;135:2740--2747.

101. Wadmekr GA. Johns.. JA. Siegel FL, Mulcabey RT. Influence of 121. Hamilos DL, Zelarmey P, Mascali 31. Lymphocyte prolifenstios in10l1a11Iose S tasfeavaes as cellular glutathione determinatio by sbliua~ione-depleted lymphocytes: Direc relatonship between gluta.


thione availability and the proliferative response. Immunopharma- 127. Pettit CM. Hail ND. Surface thiol group involvement in neuuophi]eology 1989;18:223-235. and monocyu activation. Biochem Soc Trans J990;13:305--306.

122. Lacombe P. Kraus L, Fay M. Pocidalo J.J. Glutathione status during 128. Lee F, Siemann DW. Isolation by flow cytomneny of a human ovarianthe mitogenic response of rat splenocytes. Effects of oxygen cocenn- tamor cell subpopulation exhibiting a high glutathione content pheno.tration: F02 21% venus F02 7%. Biochimie 1986;68.555-563. type and increased resistance to adriamycin. fat J Radial Oncol Binl

123. Lacombe P, Kraus L, Fay M, Pocidalo JJ. Glutathione staus of rat Phys 1989;16:1315-1319.thymocytes and splenocytes during the early events of their ConA pro- 129. Rabinovitch PS, Kubbies M. Chen YC, Schindler D. Hoehn H. BrdU.liferative responses. Biochimie 1987;69,37-44. Hoechst flow cytometry: A unique tool for quantitative cell cycle anal.

124. Shaw JP, Chou I. Elevation of intracellular glutathione content associ. ysis. Exp Cell Res 1988;174:309--318.ated with mitogenic stimulation of quiescent fibroblasts. J Cell Physiol 130. Paradiso AM. Tsien RY, Macben TE. Digital image processing of in-1986;129:193-195. tracellular pH in gastric oxyntic and chief cells. Nature 1987;325:447-

125. Traganos F, Danynkie.cz 7, Sharpless T, Melamed MR. Simulta- 450.neous staining of ribonucleic and deoxynrbonucleic acids in unfixed 131. Musgrove E, Rugg C, Hedley D. Flow cytometric measurmen ofcells using acridine orange in a flow cytofluorometric system. I His- cytoplasmic pH: A critical evaluation of available fluorochromes.tochem Cytochem 1977;25:46-52. Cytomery 1986;7:347-355.

126. Freed BM. l.,empen N, Lawrence DA. The inhibitory effects of N- 132. Cook JA. Fox MH. Intracellular pH measurements using flowethylmbaimide, colchicine and cytochalasins on human T cell func- cytometry with I ,4 -diacetoxy-2,3-dicyanobenzene. Cytometry 1988;tions. J Immunopharmacol 1989;11:459-465. 9:441-447.

• ... 2..~L .=.. _ ,. , A .

measurements of cell physiology: ionized calcium, ph and

Documents