Lipophilic, Fluorogenic /3.-GALACTOSIDASE Substrates

lacZ gene expression

Detecting

lipophilic, YU-ZHONG RICHARD Molecular

fluorogenic ZHANG, JOHN P. HAUGLAND’ Probes,

in living

/3.-galactosidase

J. NALEWAY,

KAREN

Inc., P.O. Box 22010, Eugene,

Oregon

ABSTRACT Current methods for detecting IacZ expression in transformed cells are limited because they require such harsh conditions that viability of the cells after detection is drastically reduced. To overcome this problem, we developed a series of new substrates for detection of IacZ expression in living cells under standard culture or physiological conditions. After incubation with these fluorogenic substrates, cultured lacZ-positive mammalian cells appear morphologically normal, continue to divide, and retain the fluorescent product. Because the product is so well retained, fluorescence intensity can be quantitatively related to the level of gene expression. We have demonstrated this correlation using transformed yeast cells bearing various plasmids, each containing the lacZ gene and a unique promoter sequence with known capabilities for promoting gene expression in yeast. Zhang, Y.-z.; Naleway, J. J.; Larison, K. D.; Huang, Z.; Haugland, R. P. Detecting lacZ gene expression in living cells with new lipophilic, fluorogenic /3-galactosidase substrates. FASEBJ. 5: 3108-3113; 1991. Key

Words:

substrates

1HE

fluorescein

di-/3-D-galaclopyranoside

lacZ gene expression

LACZ

GENE,

WHICH

quanhitation

ENCODES

.

/3-galactosidase

of promoter

3108

coli,

/3-galactosidase

97402,

efficiency

in

new

substrates

D. LARISON,

ZHIJIAN

HUANG,

AND

USA

METHODS Substrate

synthesis

The lipophilic fluorescein digalactoside substrates, CFDG, were prepared by the Koenigs-Knorr method used by Rotman to synthesize FDG (6, 7). The starting materials for these reactions, however, were the lipophilic fluorescein amides (Molecular Probes, Inc., Eugene, Oreg.). The products were chromatographically isolated as colorless powders. Characterization of C,2FDG: C, 60.13%, H, 6.18% (calculated for C44H57N017: C, 60.60%, H, 6.60%); m.p. 198-200#{176}C (decomposed); Rf, 0.44 (7:1:1:1 ethyl acetate:methanol:acetic acid:water); HPLC retention (k) = 4.09 on a C-2 HPLC column eluted with a 20-80% gradient of CH3CN in water containing 1% acetic acid; HPLC purity > 99%; background fluorescence < 20 ppm vs. an equivalent concentration of fluorescein; 1H-nmr (d6-DMSO) #{244}: lO.4(s,1H,N-H), 8.4(s,1H), 7.8(d,1H), 7.2(d,IH), 7.0(s,2H), 6.7(m,4H), 5.2(s,2H), 4.9(s,2H), 4.9(m,2H,2 x H-i), 4.7(s,2H), 4.5(s,2H), 3.7-3.2(m,12H),

Other

is one of the most commonly used reporter genes in molecular biology (1-3). It has previously been demonstrated that the substrate fluorescein di-3-Dgalactopyranoside (FDG)2 can be used to distinguish transformed cells expressing the lacZ gene (4, 5). However, the conditions for using the FDG substrate are harsh. Introduction of the substrate into the cells requires that the cellular membranes be permeabilized through a process such as hypotonic shock. Once the substrate is introduced, the cells must be maintained in isotonic conditions at 4#{176}C to retard leakage of the fluorescent product (fluor-escein) from the individually stained cells (4, 5). These conditions necessarily reduce cell viability after lacZ marker gene detection and severely limit FDGs use for detecting lacZ expression in whole, multicellular organisms. To overcome these limitations, we have developed a series of novel substrates lipophilic analogs of FDG - that pass freely through the cellular membrane under normal physiological or culture conditions. Like FDG, the substrates are nonfluorescent until specifically hydrolyzed by the glycosidic enzyme. Unlike FDG, the fluorescent products are very well retained in the cell at physiological temperatures with minimal leakage or transfer between marker-positive and marker-negative cells, even through several cell divisions. Escherichia

cells with

Cell

2.2(t,2H),

compounds lines,

yeast

had

1.6(m,2H),

i.2(s,16H),

similar

analyses.

and

plasmids

strains,

0.8(t,3H).

Both NIH/3T3 and CRE BAG 2 cells were obtained from American Type Culture Collection (Rockville, Md.). Yeast strains EG123 (MATa leu2 ura3 trpl his4-519 can!) and HR125-5d (MATa leu2-3 leu2 -112 ura3-52 his3 his4 trpl), along with the plasmids bearing the lacZ gene, were generously provided by David C. Hagen of The University of Oregon, Eugene, Oreg. Cell

culture

and

substrate

loading

Both 3T3/NIH and CRE BAG 2 cells were grown on coverslips in a humidified atmosphere of 5% CO2 in pH 7.4 culture medium, consisting of Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% calf serum, 50 tg/ml gentamycin, 300 sg/ml L-glutamine, and 10 mM HEPES (N[2-hydroxyethyl]piperazine-N’-[2-ethanesulfonic acid], Gibco BRL, Gaithersburg, Md.). Stock solutions (10 mM) of either 5-octanoylaminofluorescein di-/3-D-galactopyranoside (C8FDG)

‘To whom correspondence should be addressed, Probes, Inc., P.O. Box 22010, Eugene, OR 97402, 2Abbreviations:

C8FDG, C ,2FDG, DMSO,

FDG,

fluorescein

at: Molecular USA.

di-fl-D-galactopyranoside;

5-octanoylaminofluorescein 5-dodecanoylaminofluorescein dimethyl sulfoxide; m.p.,

di-13-D-galactopyranoside; di-/3-D-galactopyranoside; melting point; ONPG, o-nitrophenyl-13.D-galactopyranoside;OD600, absorbance (i.e., optical density) at 600 nm; R1, retardation factor; DMEM, Dulbecco’s modified Eagle’s medium.

0892-6638/91/0005-3108/$01.50.

© FASEB

or 5-dodecanoylaminofluorescein di-fl-D-galactopyranoside (C12FDG) in 20% dimethyl sulfoxide (DMSO) were prepared. The stock solutions were kept sealed in a brown bottle and stored at -20#{176}C.Stock solutions were diluted with fresh culture medium to the concentration desired for staining and then filter-sterilized through an ACRODISC filter (0.20-/Lm pore size) (Gelman Sciences, Ann Arbor, Mich.). The medium in which the cells were cultured was removed and replaced with the medium containing the substrate. Before microscopic examination, the coverslips with attached cells were rinsed with fresh culture medium. Proliferation

assay

Cultured CRE BAG 2 cells were trypsinized and suspended in culture medium. To ensure homogeneous cell distribution, we mixed the cells thoroughly with gentle pipetting. Equal volumes of this cell suspension were then transferred to eight Petri dishes. The cells were allowed to recover for 5 h in the incubator. Duplicate sets of these dishes were then loaded with either 1 or 0.1 mM C12FDG (as described previously), allowed to incubate for I h, and then returned to normal culture medium. The medium from a third pair of dishes was removed and the cells were treated for approximately 2 s with 100 tl of 50% methanol in culture medium, a treatment that effectively kills the cells. The killed cells were then rinsed and returned to normal culture medium. The cells in a fourth pair of culture dishes remained untreated. All dishes were returned to the incubator. Twenty-two hours after treatment, the relative cell numbers in the eight dishes were assessed using an ethidium bromide assay. Thorough pipetting ensured that the cells were completely suspended after trypsination. The cell suspensions were centrifuged and then resuspended in identical volumes of 0.05% saponin (Sigma Chemical Company, St. Louis, Mo.). Ethidium bromide was then added to a final 5 tM concentration. A 200-bd volume of each of the resuspended cells was transferred to a well in a 96-well plate. Relative cell numbers were determined using Millipore’s CytoFluor plate reader (excitation at 485 nm; emission at 620 nm). Retention

assay

CRE BAG 2 cells that had been grown to near confluency were trypsinized and removed from their Petri dish with gentle pipetting. The cells were then spun down and resuspended to a density of approximately 106 cells per milliliter. This cell suspension was then divided into three equal parts, each of which was incubated with 50 jLM FDG, C8FDG, or CI2FDG. FDG was introduced by osmotic loading as described by Nolan et al. (4). We then incubated the cells under standard culture conditions, using gentle agitation to keep the cells suspended. At 60-mm intervals, I ml of cell suspension was taken from each container. Cells were collected by centrifugation. The fluorescence of both cells and medium was determined using Millipore’s Cytofluor 2300 fluorescence multiwell plate scanner (excitation at 485 nm; emission at 530 nm) (Millipore Corp., Bedford, Mass.). Yeast

cell

!acZ GENE EXPRESSION

/3-galactosidase

activity

ford,

CELLS

yeast

Mass.).

The

ONPG

assays

were

performed

as

previ-

ously described (10) and the activities are reported in modified Miller units (9). We used linear regression to calculate the slopes of the lines shown in Fig. 5, and were then able to calculate the enzyme activities using the following equation: Enzyme where is the

activity

d is the dilution absorbance

a value

that

(i.e.,

correlates

=

(slope

factor optical

x d)

#{247} OD600

for the cell samples density)

of the

to the cell density

cells

of each

and 0D600 at 600

nm,

suspension.

RESULTS Structure

of the

FDG

analogs

We hoped that by making FDG more lipophilic, we would create a fluorogenic 13-galactosidase substrate that could be loaded under normal culture or physiological conditions without requiring harsh permeabilization of the plasma membrane. This was accomplished by attaching fatty acyl chains of varying lengths to the 5-position of FDG’s fluorescein moiety (Fig. 1). We hypothesized that once these

0--D-GaI

HNC(CH2)CH3

IN LIVING

of IacZ-positive

The transformed cells were grown in a special minimal medium for plasmid maintenance. After reaching a density of about 10 cells/ml, cells were collected by centrifugation and resuspended in Z buffer (9) at one-tenth the original volume. For the assays using the o-nitrophenyl-13-Dgalactopyranoside (ONPG) substrate, the cells were permeabilized by adding toluene (1%) and sarcosyl (0.5%) to the suspension. For the C12FDG substrate permeabilization was not required, but the cells were first preincubated with 300 iM chloroquine (Sigma Chemical Co., St. Louis, Mo.) for 20 mm at room temperature. A 10 mM stock solution of CI2FDG in 20% DMSO was diluted in Z buffer to 50 zM and filter-sterilized through an ACRODISC filter (0.20-tm pore size) (Gelman Sciences, Gaithersburg, Md.). We then mixed equal volumes of substrate and yeast cell suspension for a final substrate concentration of 25 tiM. The fluorescence intensity of each plasmid-bearing yeast was then measured using Millipore’s CytoFluor 2300 fluorescence multiweli plate scanner (excitation at 485 nm; emission at 530 nm) (Millipore Corp., Bed-

transformation

Plasmids pLG312S, pSL555, pSLII96, pSL974, and pASS were obtained from David C. Hagen of The University of Oregon, Eugene, Oreg. Yeast cells were transformed by the lithium acetate method (8). Yeast strain EG123 was used for all plasmids except pLG312S, which was transformed into strain HR125-5d.

DETECTING

Assaying cells

(rt

=

0-20)

Figure

1. Structure of the lipophilic analogs of the 13-galactosidase substrate, fluoresceindi-3-D-galactopyranoside (FDG). Membrane permeability of both the FDG analog and itsfluorescent product can be modified simply by varying n, thus changing the length of the fatty acyl chain.

3109

lipophilic substrates were internalized, they would be cleaved by /3-galactosidase, and the resulting fluorescent products would be retained within the cell. We used two of the synthesized substrates-FDG modified with 12- and 8-carton chains (C,2FDG and C8FDG)-to study the usefulness of these substrates for detecting lacZ expression in transformed cells. Detecting lacZ gene cells using CI2FDG

expression

in cultured

mammalian

Early lower

on, we discovered that it was possible to use a much concentration of the lipophilic substrates to stain lacZpositive cells than is required for FDG (10-50 zM vs. 2 mM). Presumably, the lipophilic fluorescent product accumulates within individual cells, whereas the product from the enzymatic cleavage of FDG passes freely through the membrane and into the medium. In Fig. 2, CRE BAG 2 cells [NIH/3T3

cells that have been transfected with the Moloney leukemia virus bearing the E. coli lacZ gene (Ii)] were stained with either C,2FDG or FDG. FDG was introduced by osmotic loading as described by Nolan et al. (4), whereas C,2FDG was simply added to the culture medium. Hydrolysis of both substrates at 37#{176}C was rapid. As can be seen, those lacZ-positive cells incubated with C,2FDG were characterized by a bright perinuclear staining with minimal background fluorescence, whereas those incubated with FDG were diffusely stained with high fluorescence in the surrounding medium. Control 3T3 cells, which lack the lacZ gene, showed little or no fluorescent staining except for traces of lysosomal staining, probably from acidic hydrolysis of the substrate. The fluorescence in the lysosomal compartment increased if cells were grown to confluency. In general, however, background fluorescence can be avoided by preincubating the cells with chloroquine or any other reagent that increases the pH in lysosomal compartments. To demonstrate that these new substrates are nontoxic, we incubated both 3T3 and CRE BAG 2 cells with 1 mM C,2FDG ( 20-fold the concentration normally used) under standard culture conditions for 24 h. This treatment had no noticeable influence on either morphology or viability. Cells cultured for several days in medium containing 60 ItM substrate also showed no detectable cytotoxicity. To confirm that C12FDG has no toxic effects, we did a proliferation assay using ethidium bromide to determine relative cell numbers. Ethidium bromide is a nucleotide-binding dye that is commonly used in viability assays (12). The spectrum of nucleotide-bound ethidium bromide is sufficiently different from that of fluorescein to allow us to use this dye to determine relative cell numbers without interference from the C,2FDG enzymatic product. We first incubated lacZ-positive cells with either 1.0 or 0.1 mM CI2FDG for 1 h, rinsed them thoroughly with culture medium, and then returned them to the incubator for 22 h. These cells were then permeabilized and treated with ethidium bromide. The ethidium bromide signal for the C,2FDG-treated cells was indistinguishable from that for untreated cells that were incubated for the same length of time, indicating that the rate of cellular proliferation is unaffected by CI2FDG treatment even at concentrations exceeding those required for effective C ,2FDG staining. The ethidium bromide signal from those cells that had been treated with methanol 22 h before the assay was diminished considerably compared with that of the untreated control, indicating that methanol was effective in stopping cellular proliferation. The relative fluorescence of the four populations was 1.00 for the untreated cells, 1.01 for cells treated with 0.1 mM C,2FDG, 0.98 for cells treated murine

3110

Vol.5

December

1991

Figure

2. Staining

of CRE

BAG

2 cells

with

the

fluorogenic

strates C,2FDG and FDG. A) Cells have been incubated with 50 jsM C,2FDG. B) Cells have been loaded with using culture

ments tures.

osmotic conditions.

shock

and As can

then

incubated

be seen,

that require the cells At these temperatures,

leaks from the cells, causing

FDG

for cannot

60

sub-

for 60 mm 50 brM FDG

mm under be used

normal for

experi-

to be kept at physiological temperathe cleavage product-fluorescein-

the poor signal-to-background

ratio

observed in this photograph. The cells were observed using a Zeiss Axioplan microscope equipped with a fluorescein filter set. FDG and its lipophilic analogs can also be excited by an argon laser at its principal wavelengths; it is therefore feasible to use confocal microscopy to further analyze cells stained with these substrates (data not shown).

with 1.0 mM C2FDG, and 0.30 for methanol-treated cells. The standard deviations for duplicate measurements ranged from 0.02 to 0.13. It is unclear, however, whether irradiation of lacZ-positive cells stained with CI2FDG affects cell viability It has been shown that cells stained with FDG can be cloned after fluorescence-activated sorting (4). As the behavior of the irradiated FDG molecule should be similar to that of irradiated CI2FDG, this suggests that C,2FDG-stained cells should also survive irradiation, at least at the levels required for flow cytometric methods. Retention of the fluorescent product in the lacZ-positive cells was excellent. Cells grown in medium containing 60 ftM C,2FDG were still fluorescent, albeit with decreased intensity, for up to three cell divisions after replacement with substrate-free medium. Prolonged treatment of cells with these substrates clearly does not affect their viability

The FASEB Journal

ZHANG

El AL.

Figure

3. Retention

characteristics

of lacZ-positive

cells

that

have

been

incubated with 50 jeM FDG (A), C8FDG (B), and C,2FDG (C). The fluorescence of both cellsand medium was determined using a fluorescence multiwell plate scanner (excitation at 485 nm; emission at 530 nm). In all three graphs, the fluorescence intensity of the cells is shown by the unbroken line whereas the fluorescence intensity

of the

culture

medium

is indicated

by the

dashed

line.

Be-

the substrates are nonfluorescent until specifically cleaved by intracellular 13-galactosidase, the fluorescence intensity of the medium correlates with the amount of leakage of fluorescent product from the cells. Retention of fluorescent product seems to

cause

correlate with the chain length of the covalently attached fatty acyl chain; the leakage of FDG’s fluorescent product into the medium is rather drastic (A), whereas that of CSFDG is slow but steady (B) and that of C,2FDG (C) is negligible.Unstained lacZ-positivecells that underwent the same treatment showed no fluorescence in

either

of lacZ gene expression (13). After reaching the appropriate level of growth, the transformed cells were assayed for lacZ gene expression as measured by /3-galactosidase activity We measured the /3-galactosidase activity of these transformed yeast cells using both ONPG (10) and our fluorogenic substrate, C,2FDG. For the C,2FDG substrate, the cells were first preincubated with chloroquine, a reagent that increases the pH of both phagocytic and lysosomal vacuoles. This treatment eliminates the fluorescence background in these cellular compartments, probably by curtailing acidic hydrolysis

of CI2FDG.

After treated

incubation yeast cells

with 25 ILM C,2FDG, chloroquinebearing a plasmid suitable for

guish the yeast strains. As shown in Fig. 4, lacZ-positive cells bearing the plasmid with the unmodified promoter sequence were brightly fluorescent, whereas those bearing a piasmid with a greatly modified upstream activation sequence showed only dim fluorescence. The lacZ-negative cells remained unstained for at least 2 h. We

then

undertook

the quantitation

of /3-galactosidase

activity of these yeast cells by measuring intensity over time. As shown in Fig. 5,

the cells or the medium.

f3-

galactosidase expression became fluorescent within 15 mm. Using microscopic techniques alone, we could easily distin-

their fluorescence approximately 30

mm Retention

characteristics

of FDG,

C8FDG,

and

C12FDG

Our preliminary investigations suggested that the fluorescent products from the enzymatic cleavage of our lipophilic FDG analogs were not equally retained, and this difference in retention seemed to be related to the chain length of the attached fatty acyl chain. To investigate these differences, we assayed the fluorescence intensity of cells stained with FDG, C8FDG, and CI2FDG, along with the culture medium in which they were incubated, using a multiwell fluorescence plate reader. Cells were collected at timed intervals after addition of substrate and centrifuged. We measured the fluorescence intensity of the resuspended cells, along with the supernatant collected after centrifugation, to determine both the fluorescence accumulation within the cells and leakage of the fluorescent product into the medium. Although the initial intensity of FDG staining of these cells appears to mimic that of C8FDG and C12FDG (Fig. 3), fluorescence of FDGstained cells quickly diminishes, and the level of fluorescence in the surrounding media increases proportionally In contrast, staining with the C8FDG substrate appears to result in significant accumulation of intracellular fluorescence. This is followed, however, by significant leakage (>30% in 3 h) of the fluorescent products into the extracellular medium. Cells stained under identical conditions with the even more lipophilic C12FDG show a steady, time-dependent increase in cell-associated fluorescence with minimal leakage into the medium. Quantitation of lacZ expression in yeast each with a unique promoter sequence

transformants,

Because the C12FDG substrate generates a product that is so well retained by cells, we predicted that it could be used to quantify lacZ expression. To test this hypothesis, we used a series of plasmids containing the lacZ gene, each bearing a yeast promoter with a uniquely modified upstream activation sequence. When transfected into the appropriate strains of Sacc/zammyces cerevisiae, each plasmid confers a different level

DETECTING

!acZ GENE EXPRESSION IN LIVING

CELLS

elapsed before the lacZ-positive yeast began to accumulate measurable amounts of fluorescent product, regardless of the plasmid they were bearing. We attribute this lag to the time required for the cells to internalize the substrate. After this lag, each lacZ-positive yeast strain showed a linear increase in fluorescence with time. In addition, each yeast strain accumulated fluorescent product at a characteristic rate, as reflected in the differences in the slopes of the various lines. We used the slopes of these lines to calculate the 13galactosidase that were

activity of each

yeast strain, obtaining

consistent with our assays using

the ONPG

values

sub-

Figure

4. C,2FDG staining of Saccharomyces cerevisiae cells expressing the lacZ gene. Cell suspensions of yeast strain EG123 bearing plasmids pLG-312S (A), pSL974 (B), and YEp24 (C) were incubated with

25

brM

C,2FDG

in Z buffer for 20 mm at room temperature.

Cells bearing plasmid pLG-312S, which showed 550 units activity in an ONPG assay, became brightly fluorescent (A), whereas cells bearing plasmid pSL974, which showed only 5 units activityin an ONPG assay, produced very dim fluorescence (B). Yeast cellsbearing the YEp24 plasmid showed no activity in an ONPG assay and also remained nonfluorescent (C). Photos A,, B,, and C, are the Nomarski differential interference contrast (DIC) image of these

cells. Photos A2, B2, and C2 show the fluorescent

image

of the same

cells.

3111

>‘

us C

a) C a) C-) C

a) C-) U)

a) 0

a) >

0 a)

80

Incubation

time (minute)

Figure 5. Measurement of the time-dependent increase in fluorescence of lacZ-positive yeast strains incubated in 25 iM C,2FDG. Yeast cells were transformed with various plasmids, each capable of conferring characteristic levels of lacZ expression. The fluorescence intensity of each plasmid-bearing yeast was then measured using a fluorescence multiwell plate reader (excitation at 485 nm; emission at 530 nm). We used five time points (60-140 mm) to calculate the enzyme activities by linear regression (see Table 1). The correlation coefficient was greater than 0.98 for the three transformed yeast lines with the highest activity, and 0.96 and 0.93 for pSL974 and pASS, respectively. with those obtained

The results (shown in Table using the ONPG assay.

1) are

consistent

strate (Table 1), as well as those reported in the literature (13). These observations suggest that the fluorescence intensity as measured over time correlates well with the level of lacZ expression in these yeast cells.

DISCUSSION To load cells with the fluorogenic /3-galactosidase substrate FDG, one must first permeabilize the plasma membrane, which requires rather harsh conditions (4, 5). Our hope was that by making FDG more lipophilic, we could produce a substrate that could be loaded under physiological conditions, thus broadening the applications of this otherwise useful substrate. We accomplished this by synthesizing a series

of FDG analogs, each with a fatty acyl chain attached to the 5-position of FDG’s fluorescein moiety. In this paper, we describe experiments using the 12- and 8-carbon fatty acyl analogs for detecting lacZ expression in both yeast and NIH/3T3 transformants. It is believed that these new substrates enter the cell by embedding in the outer layer of the cell membrane as glycolipids and then transferring to the inner layer of the membrane by a flip-flop mechanism (14). Once inside the cell, the substrate is cleaved by /3-galactosidase, producing a fluorescent fluorescein derivative, which because it contains a lipophiic carbon chain, is apparently well retained in the membrane. The lipophilic FDG analogs reported here should facilitate many of the methods used for analyzing lacZ-positive cells. For instance, semiquantitative techniques developed for in situ analysis of promoter efficiency using the FDG substrate have been limited because they typically require a confocal laser scanning system (15). Using the CI2FDG substrate, these techniques should be more precise and should require simpler and less expensive equipment. In addition, these new FDG analogs may extend methods that have been developed to study embryonic cells marked with the lacZ gene methods that previously required killing the animal (11, 16, 17). Using the CI2FDG substrate, for instance, researchers may be able to undertake the in vivo analysis of the lineage of cells infected with lacZ-bearing retrovis-uses (11, 16, 18). Similarly, these substrates may be useful for the in vivo detection of genomic regulatory elements, using enhancer-trap methods (17, 19-21). In this case, however, the FDG analog of choice may be one that is less well retained by the cells for example, the lipophilic FDG analogs containing shorter alkyl chains such as CBFDG. The time course of expression of the gene in a particular cell could then be characterized; the increase in intracellular fluorescence that marks gene expression within a given cell would diminish once the gene is turned off. Our investigations of C8FDG and CI2FDG provide convincing evidence that these lipophilic FDG analogs are useful reagents for detecting and quantifying lacZ expression in living organisms. For other lacZ transformants, however, the levels of lacZ expression or intrinsic /3-galactosidase may differ from those of the transformants described here, as may the cell type-specific permeability of the plasma membrane. Thus, the optimal substrate or loading conditions may differ from those used in the present experiments. In general, however,

find

wide

1. Comparison

cells using

ONPG

Plasmid

of t/ze/3-galactosidase and C,2FDG assays

-Ga1actosidase activity ONPG assay

pSL555 pSL1 196 pLGi312S pSL974 piSS

3112

Vol. 5

activity of lacZ-positive

35 22 18 5 0.6

December

1991

yeast

C,2FDG

145 90 64 20 2

assay

expect

application

these new

/3-galactosidase

in molecular

substrates

to

biology.

The authors wish to thank Nan Minchow for preparing the figures and Dr. David C. Hagen for his technical help and encouragement throughout. This research was supported by a National

TABLE

we

Institutes

of Health

grant

GM38987

to R.P.H.

REFERENCES 1. Bassford, P., Beckwith, J, Berman, M., Brickman, E., Casadaban, M., Guarente, L., Saint-Girons, I., Sarthy, A., Schwartz, M., Shuman, H., and Silhavy, T. (1978) Genetic fusions of the lac operon: a new approach to the study of biological processes. In The Operon (Miller, J. H., and Reznikoff, W. S., eds) pp. 245-262, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York 2. Guarente, L., and Ptashne, M. (1981) Fusion of Escherichia coli lacZ to the cytochrome c gene of Saccharomyces cerevisiae. Proc. Natl. Acad. Sd. USA 78, 2199-2203

The FASEB Journal

ZHANG

ET AL.

3. Rose, M., Casadaban,

M. J., and Botstein,

D. (1981) Yeast genes

to /3-galactosidase in Escherichia coIl can be expressed normally in yeast. Proc. Nail. Acad. Sci. USA 78, 2460-2464 4. Nolan, 0. P., Fiering, S., Nicolas, F., and Herzenberg, L. A. fused

J.

(1988) Fluorescence-activated mammalian transduction

cells based of Escherichia

2603-2607 5. Emilie, D., Peuchmaur,

cells analysis

and sorting of viable

on /3-D-galactosidase activity after coil lacZ Proc. NaiL Acad. Sd. USA 85,

M.,

Barad,

M., Jouin,

H., Maillot,

M.-C., Couez, D., Nicolas, J.-F, and Malissen, B. (1989) Visualizing interleukin 2 gene expression at the single cell level. Eur. J. Immunol. 19, 1619-1624 6. Koenigs, W., and Knorr, E. (1901) Ueber einige derivate des traubenzuckers und der galactose. C/rem. Ber. 34, 957-981

7. Rotman, B., Zderic,J. A., and Edelstein, M. (1963) Fluorogenic substrates for /3-D-galactosidase and phosphatase derived from fluorescein (3,6-dihydroxyfluoran) and its monomethyl ether. Proc. Nail. Acad. Sci. USA 50, 1-6 8. Itoh, H., Fukuda, Y., Murata, K., and Kimura, A. (1983) Transformation of intactyeastcellstreatedwith alkali cations. j Bacteriol. 153, 163-168 9. Miller, J. (1972) Experiments in Molecular Genetics, Cold Spring Harbor

Laboratory,

Cold

Spring

Harbor,

New

York

10. Jarvis, E. E., Hagen, D. C., and Sprague, G. F., Jr. (1988) Identification of a DNA segment that is necessary and sufficient for a-specific gene control in Saccharomyces ceievisiae: implications for regulation of a-specific and a-specific genes. Molec. Cell. Biol. 8, 309-320 11. Price, J., Turner, D., and Cepko, C. (1987) Lineage analysis in the vertebrate nervous system by retrovirus-mediated gene transfer. Proc. Nail. Acad. Sci. USA 84, 156-160 12. Taylor, I. W., and Milthorpe, B. K. (1980) An evaluation of DNA fluorochromes, staining techniques, and analysis for flow cytometry. J. Hisiochem. Cytochem. 28, 1224-1232 13. Hagen, D. C., McCaffrey, G., and Sprague, G. F., Jr. (1991)

DETECTING

!acZ GENE EXPRESSION

IN LIVING

CELLS

Pheromone responseelements are necessaryand sufficient for transcription of the FUSJ gene Cell. Biol. 11, 2952-2961 14. Homan, R., and Pownall, H. J. (1988) Transbilayer diffusion of phospholipids: dependence on headgroup structure and acyl chain length. Biochim. Biophys. Ada 938, 155-166 15. Ikenaka, K., Fujino, I., Morita, N., Iwasaki, Y., Miura, M., Kagawa, T., Nakahira, K., and Mikoshiba, K. (1990) Reliable transient promoter assay using fluorescein-di--Dgalactopyranoside substrate. DNA Cell Biol. 9, 279-286 16. Sanes, J. R., Rubenstein, J. L. R., and Nicolas, J. F. (1986) Use of a recombinant retrovirus to study post-implantation cell lineage in mouse embryos. EMBOJ. 5, 3133-3142 17. O’Kane, C. J., and Gehring, W. J. (1987) Detection in situ of genomic regulatory elements in Drosophila. Proc. Nail. Acad. Sd. USA 84, 9123-9127 18. Turner, D. L., and Cepko, C. L. (1987) A common progenitor for neurons and glia persists in rat retina late in development. Nature (London) 328, 131-136 basal and pheromone-induced Sacchromyces cerevisiae. Molec.

19. Bier,

E., V#{228}ssin, H., Shephard,

S., Lee,

K., McCall,

K., Barbel,

S., Ackerman, L., Carretto, R., Uemara, T., Grell, E. H., Jan, L. Y., and Jan, Y. N. (1989) Searching for pattern and mutations in the Drosophila genome with a P-lacZ vector. Genes Dee.

3, 1273-1287 20. Bellen, H. J.,

O’Kane, C. J., Wilson, C., Grossniklaus, U., Kurth-Pearson, R., and Gehring, W. J. (1989) P-elementmediated enhancer detection: a versatile method to study development in Drosophila. Gene Dee. 3, 1288-1300

21. Kl#{228}mbt,C.,Jacobs,J. R., and Goodman, C. S. (1991) The midline of the Drosophila central nervous system: a model for the genetic analysis of cell fate, cell migration, and growth cone guidance. Cell 64, 801-815 Received for publication July 15, 1991. Accepted for publication September 20, 1991.

3113