JOURNAL OF BACTERIOLOGYVol. 88, No. 5, p. 1249-1256 November, 1964Copyright © 1964 American Society for Microbiology

Printed in U.S.A.

ACRIDINE ORANGE BINDING BYMICROCOCCUS LYSODEIKTICUS

ROLAND F. BEERS, JR.Department of Radiological Sciences, Johns Hopkins University, School of Hygiene and Public

Health, Baltimore, Maryland

Received for publication 25 May 1964

ABSTRACT

BEERS, ROLAND F., JR. (Johns Hopkins Univer-sity, Baltimore, Md). Acridine orange binding byMicrococcus lysodeikticus. J. Bacteriol. 88:1249-1256. 1964.-Micrococcus lysodeikticus cells bindacridine orange (AO) reversibly. The adsorptionisotherm is consistent with a highly cooperative-type binding similar to that observed with poly-adenylic acid. The cells exhibit a strong bufferingaction on the concentration of free AO whichremains constant (1 ug/ml) over a range from 5 to95% saturation of the cells by AO. The cells staineither fluorescent orange or green. The fractionstained orange is directly proportional to thequantity of dye adsorbed, indicating that thesecells bind a fixed amount of AO (10% of dryweight). The green-stained cells contain less than1% of the AO bound to orange-stained cells. Theresults suggest that the abrupt increase in amountof AO bound by the orange-stained cells occurswhen the concentration of free AO reaches athreshold concentration. Similar results were ob-tained with Bacillus cereus. Mg increases the freeAO concentration and the extent of binding capac-ity of the cells.

Acridine orange (AO), a cationic dye, has beenthe subject of extensive studies in recent yearsbecause of its metachromatic staining propertieswith cell constituents (Bertalanffy and Bickis,1956; Armstrong and Niven, 1957; Anderson,1957; Debruyn and Smith, 1958) and, more re-cently, because of its binding by polyelectrolytes,such as ribonucleic acid (RNA), deoxyribonu-cleic acid (DNA), and other negatively chargedpolymers (Peacocke and Skerrett, 1956; Beers,Hendley, and Steiner, 1958; Bradley and Wolfe,1959; Steiner and Beers, 1959; Ranadive andKargaonkar, 1960; Boyle et al., 1962). The bi-ological effects of AO, which include the photo-dynamic and mutagenic actions of the dye, con-stitute another large area of research interest(Hill, Bensch, and King, 1960; van Duijin, 1960;

Hirota, 1960; van Duijin, 1961; Orgel and Bren-ner, 1961).A shift of the absorption peak of the dye from

495 to approximately 465 m,u is attributed tostacking of the dye molecules in dimers or linearpolymers, either in solution or along the polymerto which the dye molecules are bound. RNA invitro and in cells produces such a spectral shiftof the dye (Bertalanffy and Bickis, 1956). Ashift of the absorption peak to approximately 502m,u and the marked increase in fluorescence of thedye are attributed to the formation of singlybound dye-polymer complexes at random pointsalong the polymer chain or possibly at specificsites. DNA in vitro and in cells gives this spec-tral shift to AO (Peacocke and Skerrett, 1956).However, studies with purified solutions of RNA,DNA, and synthetic polynucleotides showed thatthe color shift is also dependent upon the ratioof dye bound to dye-binding sites of the particularpolymer and, in the case of DNA, the physicalstate of the polymer (Beers et al., 1958; Bradleyand Wolfe, 1959).The majority of tissue and bacterial staining

studies published did not include correlation ofthe staining characteristics with adsorption iso-therms of the dye-binding process. Vinegar (1956)described some studies with tissue cultures inwhich the color shift of the dye was qualitativelycorrelated with the extent of dye binding. Venettaand Shure (1961) measured adsorption isothermswith Ehrlich ascites tumor cells. Hill et al. (1960)related uptake of acridine orange by fibroblasts toAO concentration, but recorded their results interms of the relative fluorescence of lysed sus-pensions of cells.

This paper describes the adsorption isothermsobtained with the bacterium, Micrococcus lyso-deikticus (Fleming, 1922), and AO, and corre-lates these findings with the staining character-istics of the dye as observed by fluorescencemicroscopy.

1249

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

J. BACTERIOL.

MATERIALS AND METHODS

Preparation of bacterial cells. M. lysodeikticuscells were grown in submerged culture accordingto methods described elsewhere (Beers, 1955).Cells were collected during the late log phase bycentrifugation, washed with distilled water, andused immediately or stored frozen at -15 C.Bacillus subtilis and B. cereus cells were alsogrown in shake flasks with the same medium butlacking the bicarbonate buffer. Because of thefragility of B. subtilis cells in distilled water, itwas necessary to wash these cells with a suitablebuffer [usually 0.01 M tris(hydroxymethyl)-aminomethane (tris) at pH 8.0] containing 1mM MgCl2. This procedure slows the leakage ofmaterial from the cells.

AO was obtained from Aniline Dye Co. as theZn salt. This was converted to the free base andfinally to the acid form (Beers, 1960). Concentra-tions were estimated from the optical density at495 m,u (Beers, 1960).

Adsorption isotherms. In a typical adsorptionisotherm experiment, 2-ml samples of cells rang-

ing in concentration from 20 ,ug to 4 mg/ml were

mixed with 2 ml of a buffer in a centrifuge tube;2 ml of AO in distilled water, ranging in concen-

tration from 2 to 15 Ag/m], were added and thecontents were mixed. The suspension of cells

was centrifuged in either a clinical centrifuge or

a Sorvall refrigerated centrifuge at 10 C. Be-cause temperature variations appeared to havelittle or no effect on the shape of the adsorptionisotherm, lower temperatures were finally se-

lected to minimize the transition of stained cellsto a yellow form (see Results). Approximately30 min were required for the completion of thisprocess. The pH, ionic strength, divalent cations,and other variables were altered as described inResults.The quantity of dye remaining in the super-

natant fluid was determined by the opticaldensity at 495 m,u at room temperature. A con-

trol curve of dye concentration against opticaldensity was used to estimate the concentrationof dye in micrograms per milliliter. Because ofthe hypochromic shifts in the absorption curve

of the dye at high ionic strengths, control opticaldensity curves were determined for deviationsfrom the Beer-Lambert law as a result of dyestacking in the solutions. The amount of dyeadsorbed to the cells was estimated by the differ-ence between total dye concentration and free

dye concentration. Data are recorded in mostinstances in the form, bound dye vs. free dye.All points represent the mean of duplicate ortriplicate samples.

Fluorescence microscopy. Suspensions of cellsat a concentration of 0.2% (dry weight pervolume) were used in varying concentrations ofAO. Samples of the cell suspensions used for theadsorption isotherm studies were employed.Each examination was carried out on a glassmicroscope slide protected with a cover slip bythe use of oil emersion optics. A Zeiss fluores-cence microscope with a high-pressure mercurylight source (HgHBO 200) was usedforthestudies.The ratio of green fluorescent-stained to

orange-stained cells was determined by countingup to 150 cells per field. For orientation, a gridwas used in the image field. Counts were made asrapidly as possible. Fields which contained alarge number of yellow-stained cells were notexamined.The duration of contact of the cells with the

dye over a period of 15 min to 2 hr did not affectsignificantly either the shape of the absorptionisotherm or the fluorescence microscopy quanti-tative studies. However, prolonged exposure ofthe stained cells to ultraviolet light on the micro-scope stage did cause "bleaching" of the orange-stained cells to a brilliant yellow fluorescence.This phenomenon was observed before (Keebleand Jay, 1962). To minimize extraneous effectsfrom either light or temperature, the proceduresfor measurement of both absorption isothermsand fluorescence studies were kept as brief aspossible.

RESULTS

Reversibility of staining. Dye binding by M.lysodeikticus cells is largely reversible (Table 1).In this experiment, a 30-ml suspension of cells ata concentration of 0.33 mg/ml was mixed withAO at a concentration of 21 ,g/ml in 0.01 Mtris (pH 8.0). After the suspended cells were re-moved by centrifugation, the concentration ofdye in the supernatant fluid was determined. Thecells were then suspended in 30 ml of freshbuffer. This procedure was repeated nine times.The quantity of dye removed by each washingwas the same within experimental error (stand-ard deviation i 18 %). The total dye removed bynine washings accounted for 60% of the dyeoriginally bound to the cells. Associated with

1250 BEERS

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

ACRIDINE ORANGE BINDING

the loss of dye was a progressive decrease in theratio of orange-stained to green-stained cells.

Adsorption isotherms. A representative seriesof adsorption isotherms for several concentra-tions of M. lysodeikticus cells is shown in Fig. 1.The total binding capacity of these cells variessomewhat from preparation to preparation, butaverages about 0.1 mg of dye per milligram(dry weight) of the cells. The buffering action ofthe bacterial cells is clearly shown at the higherconcentration of cells. Furthermore, the con-

centration of free dye is independent of the cellconcentrations above 0.67 mg/ml. These resultsare consistent with the data shown in Table 1.

B. cereus demonstrates a similar quantitativetype of adsorption isotherm, but the affinity forthe dye by B. subtilis is considerably less thanthat of the other species. Nevertheless, some

buffering action can be demonstrated.The physiological state of M. lysodeikticus

cells appears to have little or no effect on theirdye-binding properties. Shown in Fig. 2 are theadsorption isotherms for "live" and heat-killedcells [0.033 mg of cells (dry weight) per milli-liter in 0.01 M tris buffer (pH 8.0)]. Cells killedwith KCN also show no alterations in dye-bind-ing properties. Acetone-washed cells, however,lose about 50% of their capacity to bind the dye.A net loss of dry weight results from the acetonewashing procedure. No acetone-extractable andaqueous-insoluble material could be recoveredfrom the cells which appeared to combine with thedye.The hydrogen ion concentration has little or

no effect on the adsorption isotherm between pH

TABLE 1. Reversibility of dye binding byMicrococcus lysodeikticus

Washing no. Amt of dye in 30 ml ofwash*

Ag

0 38.61 55.52 41.53 47.64 44.85 34.76 39.67 32.68 34.49 39.8

* Mean, 41.0 4- 7.1 jig.

'J)

0

-00'

2 3 4uwg free AO per ml

FIG. 1. Adsorption isotherm of acridine orange

with Micrococcus lysodeikticus cells. Cells freshlyharvested from a 24-hr-old submerged culture were

washed with distilled water and suspended in 0.04 Mtris buffer (pH 8.0) with varying quantities of thedye.

6

'. S0)

ca.

0 4-o

c

m 30-2

2 3 4 5 6 7,ug free AO per ml

FIG. 2. Adsorption isotherms of acridine orangewith Micrococcus lysodeikticus cells freshly har-vested, heat-killed, or acetone-washed. Cells freshlyharvested from a 25.6-hr-old submerged culture werewashed with distilled water. Symbols: 0, controlcells suspended in 0.03$ M tris buffer (pH 8.0) withvarying quantities of the dye; X, cells suspended inwater and brought to 100 C for 6 min. After removalof the water, the cells were resuspended in 0.033 Mtris buffer (pH 8.0) with varying quantities of thedye; El, cells washed with anhydrous acetone, dried,and resuspended in 0.03$ M tris buffer (pH 8.0) withvarying quantities of the dye.

8 and 4.15. At pH 3, however, the extent of dyebinding is markedly reduced (Fig. 3). Qualita-tively, the adsorption isotherm at pH 3 is similar

o Controlne kHeat-killed

o Acetone-washedi X 0

II I

VOL. 88, 1964 1251

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

J. BACTERIOL.

to that at higher pH ranges, but the concentra-tion of free dye is higher under comparableconditions. Similar adsorption isotherms can beobtained with lower cell concentrations at thehigher pH ranges.

The divalent cation, Mg2+, increases the con-

centration of free dye at a given dye to cellratio, but the total dye-binding sites of the cellsare increased by as much as 50%. Figure 4 showsthe shift in the adsorption isotherm produced

12 x °pH 3.3

Ef 10 } pH 4. 15 ,'opH 6.0

2 3 4 5 6

,g free AO per ml

FIG. 3. Adsorption isotherms of acridine orange

with Micrococcus lysodeikticus cells at varying pH.

Cells freshly harvested from a 24-hr-old submerged

culture were washed with distilled water, and were

suspended (0.033%) in various buffers and the dye

(pH 3.3, 4.15, and 6.0, 0.033 M Na acetate buffer;

pH 7.22, 0.033 M tris buffer).

1 21

a)

0

c~0

.0.0I

0 2 4 6 8 10pug free AO per ml

FIG. 4. Adsorption isotherms of acridine orange

with Micrococcus lysodeikticus cells at differentMgCl2 concentrations. Cells freshly harvested froma 24-hr-old submerged culture were washed withdistilled water, and were suspended (0.033%) in0.033 M tris with varying concentrations of MgC12and the dye.

EI.-a)0 4

co 3

D

0, 2

2 3 4pg free AO per ml

5

FIG. 5. Adsorption isotherm of acridine orangewith formaldehyde-treated Micrococcus lysodeikticuscells. A 0.02% suspension of freshly harvested M.lysodeikticus cells was shaken with 5% formalde-hyde in 0.1 M tris buffer (pH 7.0) for 24 hr at 4 C.After removal of the formaldehyde solution, the cellswere washed with 0.1 M tris (pH 7.0), and were

suspended (0.033%) in the same buffer with varyingquantities of the dye.

by MgCl2 in 0.01 M tris (pH 8.0). This increaseddye-binding capacity in the presence of Mg isalso demonstrable in the correlation studieswith fluorescence microscopy (see below).Formaldehyde blocks dye binding by the M.

lysodeikticus cells. The adsorption isotherm offormaldehyde-treated cells at pH 8.0 in 0.01 M

tris is shown in Fig. 5.Fluorescence microscopy studies. M. lysodeikticus

cells freshly stained with AO show only two typesof staining. One is a brilliant fluorescent green-yellow and the other is a less brilliant fluorescentorange. There are no intermediate forms ofstained cells. The intensity of staining appearsto be uniform for a given cell stain type, regard-less of the cell to dye ratio of the preparation.

Cells which have been exposed to ultravioletlight for a period of time eventually lose eitherthe green- or orange-staining character, andbecome a bright yellow of varying intensity. Thisshift in color becomes very marked at pH valuesbelow 4.0.

B. cereus cells demonstrate similar stainingcharacteristics, despite the fact that they are rods.In contrast, B. subtilis cells stain heterogeneouslywith loci of green and orange regions on thesame bacterial cell. Moreover, much material islost from these cells in solutions containing no

Mg. This amorphous material stains green or

orange, or shows a mixture of the types of stain-

j i A Controlo H2CO-treated

i

BEERS1252

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

ACRIDINE ORANGE BINDING

1.0

=)

a) 0m

C-)

o ooa)

0 0

L-. -

0.5

.01 .02 .03 .04 .05mg A0/mg cells

FIG. 6. Fraction of Micrococcus lysodeikticuscells stained orange as a function of the quantity ofacridine orange bound to the cells. Cells freshlyharvested from a 24-hr-old submerged culture weresuspended (0.067%) in 0.1 M tris (pH 8.0) with orwithout 1 mM MgCI2 and varying quantities of thedye. Samples were removed and examined by fluores-cence microscopy to determine the fraction stainedorange. The quantity of dye bound to the cells wasdetermined as described in Materials and Methods.

1.0o Control

0,n x mM MgCI2

0~~~~m A0 /mgc-II

oc

cn

U,L.L

.01 .02 .03 .04 .05 .06mg AC/mg cells

FIG. 7. Fraction of Bacillus cereus cells stainedorange as a function of the quantity of acridineorange bound to the cells. Experimental conditionswere the same as in Fig. 6 but with a 0.033% sus-pension of cells.

ing. The cells which have lost most material havethe appearance of hollow orange-stained cylin-ders.

In Fig. 6 is plotted the fraction of M. lyso-deikticus cells stained orange as a function of theamount of dye bound per milligram of cells.The data fit a straight line with a slope equal tothe reciprocal of the quantity of dye bound permilligram of orange-stained cells, and with anintercept equal to the amount of dye bound permilligram of green-stained cells. In the absenceof Mg, the orange-stained cells contained ap-

proximately 0.07 mg of AO per milligram ofcells; in the presence of 1 mM MgCl2, this isincreased to approximately 0.11 mg of AO permilligram of cells, a 57% increase. The data arenot sufficiently accurate to estimate whetherthere was a change in the intercept (0.001 mgof AO per milligram of cells). Parallel studieswith lower concentrations of cells show that thelinear relationship holds with as much as 95%of the cells stained orange.

In Fig. 7 are the results with B. cereus cells. Thequantity of dye bound to the orange-stained cellsis approximately 0.05 mg of AO per milligram ofcells, and to the green-stained cells, 0.012 mg of-AO per milligram of cells. Mg has an equally strik-ing effect on B. cereus dye binding.

Similar results were obtained with spray-dried,heat-, and KCN-killed M. lysodeikticus cells.Lowering the pH to 4 results in the preferentialformation of the orange-stained cells. In contrast,treatment of the cells with HCHO prevents theformation of orange-stained cells; all remaingreen.Comparable studies with B. subtilis were not

possible because of the heterogeneous stainingproperties of the cells.Washing the stained cells with buffer results in

a reversal of the cell color from orange to green.The extent of color reversal is approximatelyproportional to the amount of dye removed.

DISCUSSIONThe essential findings of this study are as

follows: (i) the strong buffering action of the cellson the concentration of free dye, (ii) the reversi-bility of the dye-binding process(es), (iii) therelative insensitivity of the dye-binding processquantitatively and qualitatively to physiologicalchanges of the cells, and (iv) the linear relation-ship between the fraction of cells stained orangeand the quantity of dye bound to the cells.The reversibility of the dye binding gives us

considerable confidence that the data representtrue adsorption isotherms. Such a conclusioncould also be reached on the basis of the bufferingaction of the cells. The linear relationship betweenthe fraction of cells stained orange and the quan-tity of dye bound to the cells indicates that thequantity of dye bound to the orange-stained cellsis constant, regardless of the quantity of dyeadded or the fraction of cells stained.The buffering action of the cells can be ac-

counted for on the basis of a stacking mechanism

- o Control-x mM MgCI2

/~ x

.

0xx', . . . ,8

VOL. 88, 1964 1253

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

J. BACTERIOL.

300

200

H

c- Hc'Jf)loof

O0 )

FIG. 8. Plot of (AO)lKi

(AO)/K,

£ki (CI)* ki (SI)

to a pair of bound dye molecules, and (c._1) and(ca) are the respective concentrations of dyemolecules bound adjacent to pre-existing bounddye molecules stacked in arrays of n - 2 andn - 1 units.

Implicit in these definitions is the assumptionthat only the dye molecules at the end of thelinear array of stacked dye molecules may dissoci-ate from the bacterium. In other words, theprobability of a dye molecule dissociating whensandwiched between two bound dye molecules issmall enough to permit us to ignore this pos-sibility.The total concentration of dye molecules bound

to the bacteria as part of the linear series is.0

102 10 1(k, /KlI

FIG. 9. Plot of per cent saturation of cells byacridine orange as a function of free dye concentra-tion with varying values of k,/KI .

in the binding of the dye. A plausible model con-

sists of linear arrays of dye-binding sites com-parable to those found on linear polymers.The following set of dye-bacteria interactions

are presumbed to take place in the formation ofthe orange-stained cells:

AO + SI T C1;AO + C1 > C2;AO + c2 r± cs;

AO + c,.. ;=k cn;

(AO)(S,)/(c1) = k,(AO)(cj)/(c2) = k2(AO)(C2)/(CS) = kC(AO)(c.-,)I(c.) = kn

11.

12 .

13 .

in .

(SI) is the concentration of dye-binding sites;(cl) is the concentration of singly bound dye mole-cules; (c2) is the concentration of dye moleculesbound adjacent to a single dye molecule, (C3) is theconcentration of dye molecules bound adjacent

(CI) = (C1) + (C2) + -(cs) + 2(c4)....n

(cn)22

If we assume that k2 =k3 = .... kn = KI,then from equation 1 .... 1,, and 2 we obtain ageometric expression for (CI), the total concen-

tration of dye bound to the orange-stained cells.

(CI) = (C) [2 + 2r + 3r2 + 4r3

3.

+.** nr(n-1)] + (Ci) [1 + (1-r)-2]

where r = (AO)/K,. From equations 1 and 3 isobtained the following:

(AO)/KI (AO)2 2k,(CI)=

(AO) KI2 KI(SI) 4.1 - 2 K,

The buffering capacity of the cells can beshown with equation 4 from a plot of

(AO)IK, vs. 2K,(SC)

(Fig. 8). It is clear that as (AO)/KI approaches1 or 2K(C,)/K,(S1) increases to values greaterthan 100 the buffering capacity of the cells in-creases.

In Fig. 9 is a plot of the per cent saturation ofthe cells by the dye required to allow the free dyeconcentration to reach 95% of the maximal valueas a function of k, K, . It is clear that the cells areable to buffer the dye concentration when thecells have been only 5% saturated with the dye(Table 1, Fig. 1). Thus, the value of kl/Ki appearsto be of the order of 103.

2.

1254 BEERS

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

ACRIDINE ORANGE BINDING

Figure 10 shows the value of (AO)/KI as afunction of the quantity of stacked dye moleculesbound to the bacteria for different values ofkc/KI. The experimental points were obtainedfrom an adsorption isotherm with M. lysodeikticuscells, assuming a value for KI of 3.7 ,ug/ml (Table1, Fig. 1).

This argument is not valid, however, if thenumber of sites available for the formation of thesingly bound dye complexes is less than the totalnumber of sites available for the formation of thestacked dye complexes. Replacing (SI) by f(SI),where f is less than unity, reduces the magnitudeof KI required for a given per cent (CI).

Implicit in this model is the assumption thatr -+ 1 and that r cannot exceed 1, a necessary re-quirement if equation 2 converges. It follows,therefore, that, under the conditions in which(AO) --KI, (SI)/(ci) -k1/KI . As the concen-tration of available dye-binding sites decreases,the quantity of singly bound dye molecules de-creases proportionately. Moreover, if k1/KI is aslarge as 103, the actual concentration of singlybound dye molecules is very small, and can beneglected in any stoichiometric considerations. Itfollows, therefore, that the green fluorescent-stained cells are not representative of cl dye com-plexes. However, proof of this will have to waitfor adsorption spectral studies of individual cellsto determine whether the green fluorescence isabolished or masked by the orange staining. Inthe case of polyadenylic acid dye binding, thegreen fluorescence is masked rather than abol-ished.To account for the quantized nature of the

orange staining process, it is necessary to assumethat the individual bacterial cells behave assingle polymers. With the proviso that ki >> KIand that ci is vanishing be small, the formationof Cn is complete in each cell that binds the dye.Thus, only two types of stained cells are observed,one of which is completely saturated with the dye.An alternative explanation based on the quantita-tive relationship, k1 << k2 >> k3 = K1 , would holdonly if the number of sites available for single-bound dye molecules was considerably less thanthat available for the stacked dye molecules. Thisrequires a rather complicated mechanism involv-ing the release of dye-binding sites after theformation of monomer dye complexes. Againstthis hypothesis is the fact that the physiologicalstate of the bacteria has a negligible effect on theadsorption isotherm.

C 100Ic0.6 - d Il000 a

b

0.2- -

0 0.5 1.0AO/K1

FIG. 10. Plot of (AO)/KI as a function of thequantity of stacked acridine orange molecules boundto bacteria at different values of k,/KI . Open circlesare experimental points.

Divalent cations were shown to depress theformation of singly bound dye-polyadenylic acidcomplexes (Beers, 1959), to enhance the forma-tion of stacked-dye polymer complexes at theexpense of the single-bound complexes, and toincrease the concentration of free dye in equilib-rium with the dye-polymer complexes. There is,however, no apparent net increase in the quantityof dye bound by the polymer. It is not knownwhether Mg increases the extent of dye bindingby DNA. The mechanism by which Mg increasesthe binding capacity of the cells is not apparent.

Blockage of dye binding by formaldehydeappears to be related exclusively to the inter-ference of the dye stacking process. This is similarto the situation observed upon treating poly-adenylic acid or RNA with formaldehyde (Steinerand Beers, 1959), which results in a sterichindrance by the methene group when attachedto the 6-amino groups of adenine. Substitution ofthe amino group to give polyinosinic acid doesnot block the stacking process.The question remains as to the identity of the

sites of the bacterium responsible for the greenor orange staining. The cell walls of M. lysodeikti-cus contain a mucopeptide (Czerkauski, Perkins,and Rogers, 1963) which accounts for 85% of theweight of the cell wall. The polysaccharide com-ponent consists of N-acetylmuramic acid and N-acetyl glucosamine linked in glycosidic bonds.Other carboxyl groups are also present (glycineand glutamic acid). The cationic dye, safranine,readily binds to the mucopeptide to the extent of17 moles per 103 g of mucopeptide or 1.7 moleculesof dye per monomeric unit of the polysaccharide,i.e., approximately 50% saturation, assuming

VOL. 88, 1964 1255

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

J. BACTERIOL.

each carboxyl group is responsible for one dyemolecule. Thus, the chemical composition of theM. lysodeikticus cell wall is compatible with sur-face binding of AO for the green staining.The orange staining presumably involves pene-

tration of the dye into the interior of the cell. Itis quite likely that the total number of bindingsites exceeds those attributable to RNA andDNA and that other molecular species are

involved.The heterogeneous nature of the staining

process makes difficult any interpretation ofLD50 data with dyes, such as AO (Beers et a].,1958).

ACKNOWLEDGMENTS

This study was supported by an AmericanCancer Society grant and by Public HealthService grant RG 7988 from the National Insti-tutes of Health. The author gratefully acknowl-edges the technical assistance of Mary Sipes.

LITERATURE CITED

ANDERSON, E. S. 1957. Fluorescence microscopyin the study of nucleic acids. Visual observa-tion of deoxyribonucleic acid changes inbacteria during growth of bacteriophage.Nature 180:1336-1338.

ARMSTRONG, J. A., AND J. S. F. NIVEN. 1957.Fluorescence microscopy in the study ofnucleic acids. Histochemical observations on

cellular and virus nucleic acids. Nature 180:1335-1336.

BEERS, R. F., JR. 1955. Submerged culture ofMicrococcus lysodeikticus for large-scaleproduction of cells. Science 122:1016.

BEERS, R. F., JR. 1959. An examination of themechanism of action of polynucleotide phos-phorylase by the use of acridine orange. Ann.N.Y. Acad. Sci. 81:645-659.

BEERS, R. F., JR. 1960. Inhibition and activationof polynucleotide phosphoryhase by acridineorange. J. Biol. Chem. 235:726-732.

BEERS, R. F., JR., D. D. HENDLEY, AND R. FSTEINER. 1958. Inhibition and activation ofpolynucleotide phosphorylase through theformation of complexes between acridineorange and polynucleotides. Nature 182:242-244.

BERTALANFFY, L. VON, AND I. BICKIS. 1956. Identi-fication of cytoplasmic basophilia (ribonucleicacid) by fluorescence microscopy. J. Histo-chem. Cytochem. 4:481-493.

BOYLE, R. E., S. S. NELSON, F. R. DOLLISH, AND

M. J. OLSEN. 1962. The interaction of deoxy-

ribonucleic acid and acridine orange. Arch.Biochem. Biophys. 96:47-50.

BRADLEY, D. F., AND M. K. WOLFE. 1959. Aggrega-tion of dyes bound to polyanions. Proc. Natl.Acad. Sci. U.S. 45:944-952.

CZERKAUSRKI, J. W., H. R. PERKINS, AND A. J.ROGERS. 1963. A study of the composition andstructure of the cell-wall mucopeptide ofMicrococcus lysodeikticus. Biochem. J. 86:468-474.

DEBRUYN, P. P., AND N. H. SMITH. 1958. Compari-son between the in vivo and in vitro interactionof aminoacridines with nucleic acids and othercompounds. Exptl. Cell Res. 17:482-489.

DUIJN, C. VAN, JR. 1960. Effect of acridine orangeon living bull spermatozoa. Nature 187:1006-1008.

DUIJN, C. VAN, JR. 1961. Effects of light and opti-cal sensitization by acridine orange on livingbull spermatazoa. Nature 189: 76-78.

FLEMING, A. 1922. On a remarkable bacteriolyticelement found in tissues and secretions. Proc.Roy. Soc. (London) Ser. B 93:306-317.

HILL, R. B., JR., K. G. BENSCH, AND D. W. KING.1960. Photosensitization of nucleic acids andproteins. The photodynamic action of acridineorange on living cells in culture. Exptl. CellRes. 21:106-117.

HIROTA, Y. 1960. The effect of acridine dyes onmating type factors in Escherichia coli. Proc.Natl. Acad. Sci. U.S. 46:57-64.

KEEBLE, S. A., AND R. F. JAY. 1962. Fluorescentstaining for the differentiation of intracel-lular ribonucleic acid and deoxyribonucleicacid. Nature 193:695-696.

ORGEL, A., AND S. BRENNER. 1961. Mutagenesis ofbacteriophage T4 by acridines. J. Mol. Biol.3 :762-768.

PEACOCKE, A. R., AND J. N. H. SKERRETT. 1956.Interaction of amino acridines with nucleicacids. Trans. Faraday Soc. 52:261-279.

RANADIVE, N. S., AND K. S. KARGAONKAR. 1960.Spectrophotometric studies on the binding ofacridine orange to ribonucleic acid and deoxy-ribonucleic acid. Biochim. Biophys. Acta 39:547-550.

STEINER, R. F., AND R. F. BEERS, JR. 1959. Poly-nucleotides. V. Titration and spectrophoto-metric studies upon the interaction of syn-thetic polynucleotides with various dyes.Arch. Biochem. Biophys. 81:75-92.

VENETTA, B. D., AND A. Q. SHURE. 1961. Adsorp-tion isotherm studies of unfixed S-37-A andEhrlich ascites tumor cell using acridineorange. Exptl. Cell Res. 24:570-579.

VINEGAR, R. 1956. Metachromatic differentialfluorochroming of living and dead ascitestumor cells with acridine orange. Cancer Res.16:900-906.

1256 BEERS

on Novem

ber 18, 2018 by guesthttp://jb.asm

.org/D

ownloaded from