A novel fluorogenic substrate for detecting alkaline ... - SAGE Journals

2 downloads 0 Views 3MB Size Report
chloro-4-(3H)-quinazolinone (CPPCQ) (Figure 1) is a substrate of. Supported by NIH grant GM 38987 (RPH). * Correspondence to: Dr. Zhijian Huang, Molrmlar ...
0022-1554/93/$3.30 The Joumal of Histochemistry and CytochemipUy Copyright @ 1993 by The Histochemical Society, Inc.

Vol. 41, NO.2, pp. 313-317, 1993 Printed in US.A.

Technical Note

I A Novel Fluorogenic Substrate for Detecting Alkaline Phosphatase Activity in Situ' ZHIJIAN HUANG,' WEIMIN YOU, ROSARIA P. HAUGLAND, VIOLE'ITE B. PARAGAS, NELS A. OLSON, and RICHARD P. HAUGLAND Molecular Psobes, Inc., Eugene, Oregon. I

Received for publication May 11, 1992 and in revised form August 10, 1992; accepted September 18, 1992 (2T2678).

I We describe here the in situ detection of alkaline phosphatase (APase) activity with a new fluorogenic substrate, 2-(5'chloro-2-phosphoryl~henyl)-6-chlor~(3H)-q&azolinone (CPPCQ.CPPCQis very soluble and colorless. APase converts it into a rapidly precipitatingproduct, whose strong fluorescence marks the sites of APasc activity. The detected APase was either a probing enzyme anchored to epidenmal growth factor (EGF) receptors of fued human epidermoid carcinoma a l l line (A431) by biotinylated JiGF and streptavidin-AFase conjugates or an endogenous marker existing in a fixed canine kidney cell line (MDCK). With CPPCQ staining, the EGF receptors and the endogenous APase were both visualized by fluorescence miaoscopy as contrasting,

Introduction The traditional goal of enzyme histochemistry is to demonstrate in situ localizations of particular endogenous enzymes in tissues or cells (1). Recent advances in histochemical enzyme techniques have been expanded to detect and quantitate tissue or cell components such as antigens, antibodies, or gene sequences that are tagged with appropriate probing enzymes (2,3). In such enzyme-related histochemistry, alkaline phosphatase (APase; EC 3.1.3.1) is often targeted because of its biological and clinical significance and its suitability as a probe enzyme (2-5). Identification of a functionally active enzyme in situ, however, is normally achieved with a substrate system that directly or indirectly yields specific and visible precipitates at sites ofthe activity. Well-known substrate techniques for APase histochemistry include metal-phosphate precipitation, azo dye coupling, and indole dye precipitation (1,6,7).Unfortunately, none of these substrate systems can produce an insoluble and fluorescent product that rapidly precipitates at the enzyme sites with high resolution for light microscopic examination. Soluble and colorless 2 45'-chloro-2'-phosphoryloxyphenyl)-6chloro-4-(3H)-quinazolinone (CPPCQ) (Figure 1)is a substrate of

Supported by NIH grant GM 38987 (RPH).

* Correspondence to: Dr. Zhijian Huang, Molrmlar Probes, Inc., 4849

Pitchford Avenue, Eugene, OR 97402.

photostable, and well-resolved fluorescent stains. The EGF receptor staining was specific since it could be blocked by excesivc unlabeledEGP. In contrast,fluorescein-labeledEGF failed to specificallystain the EGF receptors under the same fluomcent miaosap. The endogenous APasestaining with CPPCQ was sensitive to heating, levamisole and L-homoarginine, showing an APase tissue specificity ofthe liveriboncl kidney type. Therefore, CPPCQ appears to be a now1 substrate dye for sensitiwfluor"ce Apase hktochemistry. ( J H i s t o c h ~CJT&U~ 41313-317, 1993) KEY WORDS: Fluomgenic substrate;Fluorescence enzyme histochemis-

try; Epidermal growth factor receptor; Endogenous alkaline phosphatase; A431; MDCK.

APase. On action of APase in solution, CPPCQ releases an insoluble (solid) and rapidly precipitating product. This precipitate product simultaneously fluoresces at 520 nm when excited at 320-420 nm. Therefore, CPPCQ should be us& for in situ staining offmed APase activities under fluorescent microscopes with visible illumination. In this work we succeeded in detecting both epidermal growth factor (EGF) receptors and endogenous APase activity with this new fluorogenic substrate under the light microscopic scales.

Materials and Methods EGF Binding Assay. This was to tat the chemically modified EGFs for their ability to bind to EGF receptors on the f k d cells (8). Human cpidermoid carcinoma A431 cells (American Type Culture Collection; Rockvillc, MD) grown in a 96-wcJl plate (Corning Glass Works;Corning, NY)with 10% fetal bovine serum DMEM medium (GibCO; Grand Island, NY)supplemented with 10 mM Hepes @H 7.4), 2 mM glutamine. and 50 pglml gentamicin were harvested at density of about 10' cells per well. The plate was washed (three times for 5 min) with PBS and fixed in 0.5 % formaldchyde-PBS solution for 15 min at room temperature. After washing (three times for 5 min) with PBS, a 200 p1 mLmue of 16 nM biotinylated EGF (B-EGF)(MolecularProbes; Eugene, OR) and various concentrationsoffluoresccin EGF (FEGF) or unlabeled EGF (Molecular Probes) in DMEM medium supplementedwith 10mM Hcpes (pH 7.4) and 1% bovine serum albumin (B-DMEM)was added to the plate. The plate was then incubated for 60 min at 37"C, followed by PBS washing (three times for 5 min). The plate was further incubated with 100 p1 of 0.5 pglml streptavidinb-ga-

3 13

314

HUANG, YOU, HAUGLAND, PARAGAS, OLSON, HAUGLAND

0

Figure 1. Structure of CPPCQ.

lactosidase conjugate in PBS for 15 min at room temperature. After washing (three times for 5 min) with PBS, 200 pl of 1 mM 3-carboxyumbelliferyl P-D-galactoside (Molecular Probes) in 0.1 M phosphate buffer (pH 7.0 containing 0.11 M mercaptoethanol and 1.0 mM MgC12) were added to the plate. After 30 min of reaction at room temperature, the fluorescence of the hydrolytic product, carbaxyumbelliferone,was measured in a CytoFhor 2300 fluorescence plate reader (Millipore; Bedford, MA) with an excitation filter of 360 nm, an emission filter of 460 nm, and a sensitivitysetting of4.Competitivebinding between B-EGFand other EGFs to the EGF recep tors of A431 cells was indicated as the residual activity of B-galactosidase on the plate which is, in turn,expressed by the fluorescence reading. Fhorescence of a control well in which B-EGF was omitted was subtracted from the experimental readings.

EGF Receptor Staining. In a staining dish, A431 cells cultured on coverslips in 10% fetal bovine serum-DMEM medium were washed (three times for 5 min) with PBS and fixed with 0.5% formaldehyde-PBS solution for 15 min at room temperature. After washing (three times for 5 min) with PBS, the coverslipswere incubated with 10 nM B-EGF or FEGF in B-DMEM (those with additional 1 pM unlabeled EGF served as controls) for 60 min at 37'C. After washing (three times for 5 min) with PBS, the slips stained with FEGF were examined under an Axioplan fluorescentmicroscope (Zciss; Thornwood, NJ) with a fluorescein filter set of480 nm excitation and 530 nm emission. The slips previously treated with B-EGF were washed (three times for 5 min) in 0.1 M Tris buffer (pH 7.8 containing 10 mM MgC12, 50 mM NaCI, and 0.1 mM ZnC12). and then incubated with 0.1 pg/ml streptavidin-APase conjugate (MolecularProbes) in Tris buffer at room temperature for 15 min. After washing (three times for > min) with Tris buffer, the slips were incubated with 30 pM CPPCQ (Molecular Probes) in Tris buffer that was previously filtered through a 0.2-pm filter (Millipore). After 15 min of incubation at room temperature, the slips were washed (three times for 10 min) and mounted with distilled water for examination and photographing under the fluorescent microscope with a filter set that permits 360 nm excitation and 520 nm emission. EndogenousA b Staining. In a staining dish. canine kidney MDCK cells (American TVpe Culture Collection)cultured on coverslips in DMEM medium supplemented with 10% calf serum, 10 mM Hepes (pH 7.4), 2 mM glutamine, and 50 pglml gentamicin were washed (three times for 5 min) with PBS and fixed in 4% formaldehyde-PBS solution for 15 min at room temperature. The coverslips were washed (three times for 5 min) with Tris buffer and then incubated with 10 pM pre-filtered CPPCQ solution in Tris buffer for 30 min at room temperature. After washing (three times for 10 min) and mounting with distilled water, the slips were then m i n e d under the fluorescent microscopewith a 360 nm excitation filter and a 520 nm emission filter. The fluorescent images were then processed with a Start I camera (Photomctrics; Tucson,AZ) and I m w - 1 software (UniversalImaging; Media, PA). The tissue specificity of the staining was examined for sensitivityto inhibitory factors such as heating (the coverslips

0

20

40

60

80

IO0

EGF concentration (nM) Figure2. Inhibition of B-EGF binding to fixed A431 cells by unlabeled EGF (solid circles) and F-EGF(open circles). Fluorescencereading represents the pgalactosidase activity associated with 8-EGF that remains on the A431 cells as a result of the competition.

in Tris buffer were heated at 70'C for 45 min before application of the CPPCQ substrate solution), levamisole (1 mM added to the CPPCQ solution), Lhomoarginine (10 mM added to the CPPCQsolution),and tphenylalanine (10 mM added to the CPPCQ solution) (Sigma; St Louis, MO).

Results Figure 2 shows the binding competition between B-EGF, F-EGF, and unlabeled EGF to the EGF receptors of A431 cells fixed on a 96-well plate. B-EGF associated with the cell plate at 10 nM could be virtually eliminated by unlabeled EGF at concentrations around 100 nM, as monitored by the fluorescence readings (solid circles in Figure 2), indicating specific binding of B-EGFto the EGF receptors of the A431 cells. The binding ability of F-EGF to the receptors can also be seen in its competition with B-EGF (open circles in Figure 2). Using the enzyme-mediated technique with 10 nM B-EGF and 30 WMCPPCQ, a contrasting and photostable fluorescentstaining of the EGF receptors on the fixed A431 cells under the fluorescent microscope was obtained, as shown in Figure 3b. The staining was associated solely with the A431 plasma membranes and could be resolved at the single-cell level even under the high magnification of 400. No nonspecific staining outside the cells was observed. The staining specificity was proven by: (a) the ability of B-EGF to specifically bind to the EGF receptots as shown in Figure 2; (b) the fact that addition of 1 pM unlabeled EGF simultaneouslywith 10 nM B-EGF blocked the stainingcompletely; and (c) the B-EGF staining concentration used, 10 nM, is compatible with the dissociation constant of the EGF receptors of high affinity (0.9 nM) (9). The control experiment with unlabeled EGF also indicated the lack of endogenous APase activity in A431 cells. Apparently the fluorescence spectrum of the CPPCQ hydrolytic product that is separated

SUBSTRATE FOR ALKALINE PHOSPHATASE

.-

315

.

-,

r-]

D

L

e-

o

Pm a *

''...

+?. e

-

"X

Figure 3. (a) Phasecontrastmicrograph of fixed A431 cells; (b) fluorescence micrograph of the EGF receptorson the cells. Original magnification x 400. Bar = 10 pm. Figure 4. (a) Phase-contrastmicrograph of fixed MDCK cells; (b)fluorescence micrograph of APase activity in the cells. Original magnification x 1000. Bar = 10 pm.

from the cell autofluorescence is responsible for the clean background seen in this control experiment. However, we failed to visualize the EGF receptors with 10 nM FEGF, although the binding experiment in Figure 2 shows that F-EGF does bind to the receptors. Increasing the F-EGF concentration (up to 100 nM) did not improve the signal and resulted in nonspecific staining that could not be blocked by 100-fold excess unlabeled EGF. Such faint staining of the EGF receptors with FEGF was also reported by Sorkin et al. (10). It might be explained by the significant photo-bleachingof fluorescein dye and the cell autofluorescence passing through the fluorescein filter set. We also used CPPCQ to stain the endogenous APase activity that exists in MDCK cells. On incubation with 10 FM CPPCQ, APase of MDCK cells was reproducibly visualized as green fluorescent deposits under the fluorescence microscope. These stains were associated solely with the individual cells and were distinguishable on a single-cell basis even at a high magnification of 1000 (Figure 4b). No nonspecific staining in the cell-free area was present.

Moreover, this APase staining characteristicallyappeared around the inner side of the MDCK plasma membranes, with no apparent cell-crossingstaining resulting from the product diffusion. Heating (70°Cfor 45 min) and levamisole (1.0 mM) completely inhibited the APase activity and eliminated the APase staining with CPPCQ. However, L-homoarginine (10 mM)and L-phenylalanine (10mM) exhibited only partial and slight inhibition of the APase, respectively. These inhibitory results agree with the properties of a liver/ bonelkidney-typeAPase (11-14), thus confiiing the biochemical specificity of the MDCK APase staining shown in Figure 4b. It is also interesting to note that the black spots appearing in the phase imaging of Figure 4a correspond to the fluorescent stains of Figure 4b. Apparently CPPCQs hydrolytic product also constitutes a visible contrast for the APase activity by light microscopy. There are two common stainingvariables for both the EGF receptors and the endogenous APase. The primary variable is the pH of the Xis buffer used to prepare the CPPCQ staining solution. The precipitation and fluorescenceof the CPPCQ hydrolysis prod-

3 16

uct, a weak phenol, can be reduced by phenolic ionization at high pH. For instance, at pHs higher than 8.0,the EGF receptor staining with CPPCQwas diminished and was characterized by random nonspecific stains outside the cells. The other variable is the CPPCQ concentration. In terms of sufficient signal and low nonspecific background, the optimal CPPCQ concentration for staining of the EGF receptor and the endogenous APase was found to be around 10-100 pM.

Discussion Most enzyme histochemical methods involve substrate systems that However, release color precipitates as enzymatic products (1,2,6,7). histochemical signals constructed with such color precipitates can be confused with cell stmctures that have the similar contrast. The fluorescenceapproach is considered advantageousbecause the distinctive signal can be analyzed by modem imaging techniques for high sensitivity and specificity(15).Burstone used several naphthyl phosphate derivatives to localize tissue APase activity by the direct formation of fluorescent precipitating products (naphthols) on the action of the enzyme (7,16).However, the naphthol products appeared to be too diffusible and formed coarse precipitates. Consequently, APase localization at magnifications greater than x 200 with these substrateswas impossible. Similarly, another fluorescence method based on azo-dyecoupling technique cannot demonstrate APase activity at high resolution owing to product diffusion (14). Obviously, a substrate that rapidly forms a highly insoluble and fluorescent product is desirable for sensitive and well-resolved enzyme histochemistry under fluorescent microscopes. It was previously proposed that sufficient affinity of the hydrolytic product for tissues is a prerequisite of a substrate useful for in situ detection (16).This requirement leads to special chemical designs by which the hydrolytic products can be trapped by tissues. With a given substrate and specimen, histochemical staining of APase also appears to be dependent on the methods of specimen processing. such as pre-treatment and fixation. For example, APase reactions with the metal-phosphate-basedsubstrate became detectable by electron microscopy when the tissue sections were pre-fixed Actually, with glutaraldehydeto prevent product diffusion (17~8). choosing a substrate that yields less solubleproduct for rapid precipitation could be a versatile altemativeto the tissue-affinityapproach. Despite the potential advantage ofversatility, the solubility approach is often precluded by the decrease in substrate solubility in parallel with the decrease in product solubility. Therefore, an ideal substrate for enzyme histochemistryshould be very soluble and should also be able to generate a much less soluble product. CPPCQ is such a substrate, having a large difference in solubility from its product. Its hydrolytic product, 2-(5’-chloro-2‘-hydroxyphenyl)-6-chloro-4-(3H)-quinazolinone (CHPCQ) is extremely insoluble. This could be attributed to the unique intramolecular hydrogen bonding between the 1-nitrogen and the 2’-hydroxylic hydrogen atom that is absent in CPPCQ (Figure 1). CHPCQ‘s solubility limit is around 0.1 mM, as determined by light scattering in solution, whereas CPPCQ is extremely soluble. This solubility difference causes the rapid precipitation of CHPCQ formed from CPPCQ hydrolysis. The rapid precipitation of CHPCQ then makes possible the localization of APase with CPPCQ at high resolution. Other advantagesof CPPCQ as an APase substrate include its rela-

HUANG, YOU, HAUGLAND, PARAGAS, OLSON, HAUGLAND

tive photostability and the unique spectrum of CHPCQ. The high photostabilityof CHPCQ permits a prolonged observation of APase activity. On the other hand, the cellular background fluorescence, which is typically around excitation 340 nm and emission 460 nm or excitation 450 nm and emission 515 nm (19,20), can be greatly reduced by use of a filter set that matches the unique fluorescence spectrum of CHCPQ (i.e., excitation 360 nm and emission 520 nm). The usefulness and advantagesof CPPCQ are illustrated by the EGF receptor and endogenous APase stainings in this work (Figures 3 and 4). These stainings are clean, bright, and durable, and are clearly associated with the cells in characteristic patterns at magnifications greater than x 400. The staining of endogenous APase in MDCK cells with CPPCQ has the most well-defined localization at the light microscopic level ( x 400) of all the previous methods (7,16).This APase staining pattern also seems to be in agreement with the observation that this enzyme is mostly localized to cell surfaces, as revealed by immunofluorescence and electron microscopic techniques (11,17,18).The in situ detection of the EGF receptor mediated by the APase amplification and CPPCQ is obviously more sensitive than that with FEGF. In conclusion, CPPCQ appears to be a novel substrate for sensitive and well-resolved APase histochemistryor APase-based histochemical detection under conventional fluorescence microscopes.

Acknowledgment We thank Ms Nan Minchow for the art work.

Literature Cited 1. BancroJD, Hand NM. Enzyme histochemistry.Oxford:Oxford Univer-

sity Press, 1987 2. PolakJM, Van Noorden S.An introduction to immunocytochemistry:

3.

4. 5.

6. 7. 8.

9.

10.

11.

12.

current techniquesand problems. Oxford Oxford University Press, 1984 Bloch B. Detection of messenger RNAs by in situ hybridization with nonradioactive probes. In Chesselet M-F, ed. In situ hybridization histochemistry. Boca Raton, E CRC Press, 1990:23 HarbJ, Vavasseur F,Chadcneau C. Expression of alkaline phosphatase in murinelymphomacells. Biochem Biophys Res Cm”m 199I;177:121 FarlqB, Kyeyune-Nyombi E, kbaux NM,Hall SL, Strong DD. Alkaline phosphatase activity from human osteosarcomacell line SaOS-2: an isoenzyme standard for quantifying skeletal alkaline phosphatase activity in serum. Clin Chem 1989;35:223 Lojda Z, Gossrau R. Schieblcr TH. Enzyme histochemistry:a laboratory manual. Berlin: SpringerVerlag, 1979 Burstone MS. Enzyme histochemistryand its application in the study of neoplasms. New York Academic Press, 1962 King IC, CatinoJJ.Nonradioactive ligand binding assay for epidermal growth factor receptor. Anal Biochcm 1990;188:97 Carraway KL, KolandJG, Grione RA.Visualizationd e p i d c d growth factor (EGF) receptor aggregation in plasma membranes by fluorcscence energy transfer. J Biol Chem 1989;264:8699 Sorkin AD, Tcslenko LV, Nikolsky NN. The endocytosis of epidermal growth factor in A431 cells: a pH of microenvironment and the dynamics of receptor complex dissociation. Exp Cell Res 1988;175:192 Bcnhan~FJ, h d r m PW, Knowles BB. Bronson DL, H ~ r i H. s Ahline phosphatase isozymes as possible markersof differentiation in human testicular teratocarcinoma cell lines. Dev Biol 1981;88:279 Borgcrs M.The cytochemical application of new potent inhibitors of alkaline phosphatases. J Histochcm Cytochem 1973;21812

SUBSTRATE FOR

ALKALINE P H O S P U S E

317

A, Livne E, Silbcrmann M, R e d & AZ.Activity ofalkaline phosphatase in rat skeletalm d e localized along the sarcolemma and endothelial cell membranes. J Histochem Cytochem 1991;39:199

13. Dolbeare F, Vanderlaan M, Phares M. Alkaline phosphatase and an acid arylamidax as marker enzymes for n o d and transformed WI-38 cells. J Histochem Cytochem 1980;28:419

17. S&

14. Ziomek CA, Lepire ML,Torres I. A highly fluorescent simultaneous

18. MulnardJ, Huygens R. Ultrasuucturallocalization of non-specific alka-

azo dye technique for demonstration of nonspecificalkaline phosphatase activity. J Histochem Cytochem 1990;38:437 15. Shotton DM. Video and opto-digital imaging microscopy. In Cherry RJ, ed. New techniques of optical microscopy and microspecuoscopy. Boca Raton, FL.: CRC Press, 1991:l

line phosphatase during cleavage and blastocyte formation in the mouse. J Embryo1 Fkp Morpho1 1978;44:121 19. Aubin JE. Autofluorescence of visible cultured mammalian cells. J

16. Buntone MS.Postcoupling, noncoupling, and fluorescence techniques

for the demonstration of alkaline phosphatase. J Natl Cancer Inst 1960;24:1199

Histochem Cytochem 1979;27:36 20. Bemson RC, Meyer RA, Zaruba ME. McKhann GM. Cellular autofluorescence-is it due to flavins?J Histochem Cytochem 1979;27:44