Histochemical Journal 22, 426-432 (1990)

Biotinylated epidermal growth factor : a useful tool for the histochemical analysis of specific binding sites K. K A Y S E R 1., G. W E I S S E 2, H . J. G A B I U S 3 a n d T. H I N T Z E 2

1Langgewann 39, 19-6900 Heidelberg, Germany 2E. Merck, Frankfurter, Str. 250, Darmstadt 3Max Planck Institut fiir Experimentelle Medizin, Abt. Chemie, D34 GfSttingen, Germany (FRG) Received 23 October 1989 and in revised form 9 February 1990

Summary Epidermal growth factor (EGF)was labelled with biotin via modification of either the amino or carboxyl groups, using suitable reagents, namely biotinyl-N-hydroxysuccinimide ester or biotinamidocaproyl hydrazide. To assure that the specific binding capacity of EGF is retained despite its chemical modification, displacement of the EGF by biotinylated derivatives in a routine binding assay was performed. The inhibitory potency compared to unmodified EGF was only slightly reduced. This result is the prerequisite for testing the usefulness of biotinylated EGF in histochemistry. The biotinylated probes were applied to sections of human tumour tissue and of monkey organs (liver, kidney, uterus of Cynomolgus and Rhesus monkey) to localize the specific binding sites for EGF. Formalin-fixed, paraffin-embedded tissue sections were deparaffinized and incubated with the probes at a concentration of 10 ~tgml 1at room temperature for 60 rain. Specificbinding of the EGF was visualized by the avidin-biotin techniques (ABC). A positive reaction in conjunction with appropriate controls by competitive inhibition was seen for all monkey tissue sections and for the following number of cancer cases: breast carcinoma: 7/10; mesothelioma: 2/4; lung carcinoma: 1/3; colon carcinoma: 1/3. The staining properties were similar for both types of probes that differed in the functional group that is involved in modification by biotin attachment. However, the batches with modification of the amino groups stained more intensely and more distinctly than the carboxyl modified EGF. Overall, the data indicate that the ligand properties of the EGF are not impaired by biotinylation of the two types of functional groups. Thus, biotinylated EGF is a useful tool for histochemical detection and identification of EGF specific binding sites in mammalian tissue.

Introduction Receptor analysis is conveniently feasible by immunohistological tools. However, ligands themselves offer a suitable alternative. Their labelling to render them detectable should not impair their binding specificity. The application of such labelled ligands in receptor analysis thus warrants attention. The following study was performed in order to analyse the binding capacities of two synthetically different types of biotinylated epidermal growth factor (EGF) on routinely processed sections of h u m a n and monkey tissue using histochemical techniques. EGF is a protein consisting of 53 amino acids of known sequence with a molecular weight 6250 Da (Basu et al., 1987). Its presence has been demonstrated in normal h u m a n plasma and urine (Carpenter, 1987; Gill et al., 1987). This growth factor binds to specific receptors on the cell surface (Carlin & Knowles, 1986). The EGF receptor itself could be isolated from A431*To whom all correspondence should be addressed. 0018-2214/90 $03.00 +.12 9 1990 Chapman and Hall Ltd.

tumour cells and was shown to exhibit tyrosine protein kinase activity with ligand-stimulated autophosphorylation (Carpenter, 1987). These properties are similar to activities of other known peptide growth factor receptors such as those specific for insulin, platelet-derived growth factor, or virus-transforming gene products sic and abl (Gill et al., 1987; Heinemann & Jehn, 1985). The EGF appears to be involved in differentiation and in growth of mammalian cells (Damjanov, 1986), especially in tumour cells (Basu et aI., 1987; Hendler & Ozanne, 1984; Stoscheck & King, 1986). Tissue cultures of h u m a n small cell anaplastic carcinoma, for example, need media containing insulin, hydrocortisone, transferrin. Tissue cultures of h u m a n adenocarcinoma cells of lung apparently require addition of EGF (Damjanov et al., 1986). Presence of EGF-specific receptors could be demonstrated in a broad spectrum of h u m a n malignancies including neurogenous tumours, bladder cancer, or carcinoma of breast (Nicholson et al., 1988; Rios et al.,

KAYSER, WEISSE, GABIUS and HINTZE 1988) and lung (Veale et al., 1989), mainly by application of monoclonal antibodies. Studies on the detection of e n d o g e n o u s sugar receptors like lectins in lung cancer h a d clearly proven that a ligand, immobilized to a labelled carrier, offers a convenient alternative to monoclonal antibodies for detection of specific binding sites (Kayser et al., 1989 a,b). These findings p r o m p t e d the present s t u d y with a polypeptide without any previous conjugation to a carrier. It focuses on the question, w h e t h e r the labelled EGF can specifically bind to tissue sections despite the inevitable chemical modification of the polypeptide as well as the potentially harmful tissue processing.

Materials and methods Commercially available recombinant human EGF (Progen, Heidelberg, FRG) was biotinylated with biotinyl-Nhydroxysuccinimide ester at a ratio of 2 mg reagent for 200 ~tg EGF essentially as described for neoglycoproteins (Kohnke et al., 1989). To link covalently biotin to carboxyl groups a biotinhydrazide derivative was used in a reaction, similar to a protocol for biotinylation of nerve growth factor (Rosenberg et al., 1986) to EGF as follows. 300 btg EGF in 220 I*l 10 m u phosphate-buffered saline (PBS, pH 4.8) was diluted 1 : 10 with 10mM pyridine-HCl (pH 4.8). 16.5 mg biotinamidocaproyl hydrazide (Sigma, Munich, FRG), dissolved in 5 ml ofa I : I mixture of water and dimethylsulpf ~ "de, were added and the coupling reaction was started with further addition of 9 mg 1-ethyl-3-(3-dimethyl-aminopropyl)carbodiimide (Sigma, Munich, FRG). To avoid loss by gel filtration and to remove excess reagents, the mixture was extensively dialyzed in tubing, impermeable to molecules of molecular mass greater than 1000 Da, and was subsequently lyophilized. The presence of accessible biotin groups on the EGF was ascertained by probing with streptavidin-gold solution as described in detail elsewhere (Kohnke et al., 1989). Comparative displacement studies of iodinated EGF, prepared by the chloramide T method (Hunter & Greenwood, 1972) by EGF as well as by biotinylated EGF were performed in a routine receptor determination on three individual specimens of invasive ductal breast carcinoma, as described elsewhere (Macias et al., 1987). The biotinylated probes were applied to histological sections of formalinfixed (buffered formalin, 6.9 7.3), paraffin-embedded tissue, 4 - 5 btm thick. After deparaffination and rehydration through graded ethanols, the sections were incubated in 0.1% methanolic hydrogen peroxide for 30 rain. An equilibration with 0.1 M Tris buffer was followed by treatment with 1% bovine serum albumin (BSA) in Tris-buffered saline to minimize non-specific background staining by proteinprotein interactions. The incubation with the biotin-labelled

427 EGF at a concentration of 10 btg ml -I was performed at room temperature for 60 rain. The application of the ABC complex (Camon, Wiesbaden, FRG) and the development of the chromogenic product from the substrates diaminobenzidine-hydrogen peroxide followed after washing the sections twice in Tris buffer for 5 min. Finally, the sections were counterstained and mounted. Histochemical control experiments were performed by application of non-biotinylated EGF as competitive inhibitor of the labelled growth factor to exclude unspecific binding of the biotin groups, and to underscore the specificity of growth factor-receptor interaction. Positive and negative controls were performed as usual, i.e. by parallel performance of a positive staining slide (breast carcinoma) and by replacement of the biotinylated EGF with unlabelled EGF. Cases were considered to stain positively, only if a dark brown staining could be seen in all or in clusters of the corresponding ceils within the individual section.

Results A conventional binding assay with the iodinated ligand was employed for three t u m o u r specimens to determine w h e t h e r the biotinylated probes exhibit a similar inhibitory potency in comparison to the unmodified growth factor. Invariably, increasing concentrations of inhibitor progressively reduced the binding. In comparison, the biotinylated growth factors were similar or only slightly less effective as inhibitors in relation to native EGF, w h e n given at identical concentrations. Thus they are promising histochemical tools to localize binding sites for EGF in tissue sections. The synopsis of the material is s h o w n in Table 1. The selected sections originated from 20 cases of h u m a n cancer (ten carcinoma of breast, four mesothelioma, three colon carcinoma, and three carcinoma of lung), and of the following types of m o n k e y tissue: two kidneys (one rhesus, one cynomolgus), one liver (cynomolgus), and one uterus (cynomolgus). The staining was, in general, h o m o g e n e o u s a n d intensive, as s h o w n in Fig. 1, illustrating the positive staining of hepatocytes in the cynomolgus liver. The kidneys stained positively in the proximal tubuli (Fig. 2), the uterus for epithelial cells (not shown). Peripheral nerves s h o w e d positive binding capacities (Fig. 3). Notably, binding was not restricted to normal types of cells. A m o n g the breast carcinoma sections from 7/10 cases were considered positive with specific reaction to the proliferating cancer cells (Fig. 4). A strong positive staining of the cancer cell nuclei was seen in

Fig. 1. Cynomolgus liver showing a homogeneous positive binding to biotinylated EGF. x250. Fig. 2. Cynomolgus kidney showing a positive binding of the proximal tubuli and a negative binding of the glomeruli to biotinylated EGF. • Fig. 3. Peripheral human nerve showing a homogeneous binding to biotinylated EGF. x250.

KAYSER, WEISSE, GABIUS and HINTZE

429

Fig. 4. Human invasive ductal carcinoma of breast with focal microcalcinosis. Proliferating tumour cells bind positive to biotinylated EGF, stroma cells express no detectable binding capacities. • Fig. 5. An epithelial mesothelioma showing a positive binding of tumour cell nuclei to biotinylated EGF. x240.

some cases (Fig. 5). The staining b e h a v i o u r in the m e s o t h e l i o m a cases as well as in the colon cancer and lung cancer cases was similar. The results are s u m m a rized in Table 2. With respect to the type of modification the pattern of the staining of the two different labelled probes was nearly identical for each case

except the r h e s u s kidney. Here, positive staining for the amino g r o u p labelled EGF and no staining for the carboxyl g r o u p labelled EGF was observed. The control with the co-incubation of unlabelled EGF (60min) prior to the incubation with labelled EGF (60 min) to saturate EGF receptors revealed no staining

430

Biotinylated epidermal g r o w t h factor Table 1. Synopsis of material.

Monkey tissue:

Rhesus: Cyomolgus:

kidney kidney, liver, uterus

Human cancer:

Origin

N

males

females

age (years)

breast mesothelioma lung colon

12 4 3 3

4 3 1

12 2

43-57 53-67 51, 57, 63 62, 67, 69

Table 2. Histopathological evaluation of EGF receptors in monkey and in human cancer tissue by use of labelled EGF.

labelled via COOH

NH2

-

+

+ + +

+ + +

+ + + + (+) + (+)

+ + + + + + +

lobular lobular lobular invasive invasive invasive invasive invasive invasive invasive

Mesothelioma case 1 2 3 4

+ +

+ +

biphasic epithelial biphasic biphasic

Colon carcinoma case l 2 3

+

+

adeno adeno adeno

Lung carcinoma case 1 2 3

+

+

large cell small cell epidermoid

Tissue Rhesus kidney Cynomolgus kidney liver uterus Breast carcinoma case 1 2 3 4 5 6 7 8 9 10

Cell type

ductal ductal ductal ductal ductal ductal ductal

for both probes in the examined cases of m o n k e y tissue (liver) and h u m a n cancer (breast carcinoma). This result corroborated the conclusion, derived from the relative inhibitory p o t e n c y for native a n d biotinylated EGF, d e t e r m i n e d in the binding assays. Discussion

The EGF receptor is considered to be an integral m e m b r a n e protein, involved in two major cell func-

tions: signal transmission b y p h o s p h o r y l a t i o n of certain proteins o n their tyrosine residues via the kinase activity, and m o d u l a t i o n of the EGF-stimulated receptor activity (autoregulation) (Stoscheck & King, 1986). With respect to carcinogenesis, EGF has b e e n implicated in e n h a n c e m e n t of viral transformation of m a m m a l i a n cells (Fisher et al., 1979; Harrison & Auersperg, 1983), in the p r o m o t i o n of carcinogenesis by augmentation of the carcinogenic potential of m e t h y l c h o l a n t h r e n e in skin (Stoschek & King, 1986), and by i m m u n o s u p p r e s s i v e activity (for example suppression of T-cell function in delayed hypersensitivity reaction) (Roberts et aI., 1976). All these properties of EGF have led to an !ncreasing interest in the evaluation of expression of EGF-specific receptors, especially in h u m a n cancer. Consequently, monoclonal antibodies to the receptor have b e e n m a d e available (Waterfield et aI., 1982), and some of the p e r f o r m e d studies could d e m o n s t r a t e that EGF receptors are probably involved in the d e v e l o p m e n t and proliferation of certain h u m a n malignancies (Heinem a n n & Jehn, 1985; Basu et al., 1987; Veale et al., 1989,), Besides the immunological epitope-specific too! ionidated g r o w t h factors are suitable to detect and quantify u n o c c u p i e d ligand binding sites in h o m o g e m ates. The chemical modification o n tyrosine residues apparently did not impair the binding specificity. This observation e n c o u r a g e d the exploration of the possibility to r e n d e r biotinylated factors available for biochemical as well as cyto- and histochemical p u r p o ses (Hopkins et al., 1981; Rosenberg et al., 1986; Foxwell et al., 1988; N e w m a n et al., 1989; Wojchowski & Caslake, 1989; Yem et al., 1989). A m o n g these studies, one report already indicated that EGF can be biotinylated, while retaining its ligand properties (Hopkins et al., 1981). This result p r o m p t e d to extend the monitors of such a derivative and to assess the applicability of different types of modifying reagent. O u r s t u d y e m p h a s i z e d that this tool, accessible by c o n v e n i e n t labelling with commercially available reagents, can b e c o m e a reasonable alternative to receptor-specific monoclonal antibodies. The percentage of positive staining of biotinylated EGF in our s t u d y o n h u m a n breast carcinoma (7/10 cases) is within the same range as the reported extent of positive reaction to m o n o -

KAYSER, WEISSE, GABIUS and HINTZE clonal antibodies (Rios et al., 1988), and may thus similarly be of important prognostic value (Nicholson et al., 1988). However, it should be kept in mind in this context that immunological detection of a receptor is strictly confined to the recognition of the antigenic determinant or cross-reaction determinants. This technique does not infer functional activity. Moreover, detection of EGF-specific receptors in different animal species by application of only one specific antibody requires the conservation of similar antigenic epitopes. On the other hand, functional activity can be unequivocally determined by using a labelled ligand (EGF in our case) instead of a monoclonal antibody, as long as it can be demonstrated that the chemical labelling is not disturbing the physiological functions of the ligand. The results of our experiment clearly indicate that; a. the biotinylation of both the amino and carboxyl groups does not significantly affect the affinity of EGF to its receptors in histological specimens b. the labelled EGF specifically recognizes its receptive binding sites c. the practical use of the amino labelled probes is slightly superior to that of the carboxyMabelled probes d. biotinylated EGF can be used in h u m a n and in monkey tissue e. the EGF-receptor function is apparently not impaired by routine tissue processing. It can be concluded that the biotinylated EGF is useful for histochemical demonstration of EGF-specific binding sites that are present in h u m a n and in monkey tissue.

Acknowledgements The excellent assistance of T. Hellmann, A. Sch~ifer, and G. Wrist, the helpful suggestions of Dr L. Bill, and the financial support of Dr M. Scheel-Stiftung fi~r Krebsforschung and of the Verein zur F6rderung des biologisch-technologischen Fortschritts in der Medizin e. V. are gratefully acknowledged.

References B A S U , A . I M U R T H Y , U . / RODECK~ U.r H E R L Y N , M . / MATTES~

L. & DAS, M. (1987) Presence of tumour associated antigens in epidermal growth factor receptor from different human carcinomas. Cancer Res. 47, 2531-6. C A R L I N , C- R. & K N O W L E S , B. (1986) Biosynthesis and glycosylation of the epidermal growth factor receptor in human tumour-derived cell lines A431 and Hep3B. MoL Cell. Biol. 6, 257-64. CARPENTER, G. (1987) Receptors for epidermal growth factor and other polypeptide mitogens. Ann. Rev. Biochem. 56, 881-914. C A R P E N T E R , G. & C O H E N , S. (1979) Epidermal growth factor. Ann. Rev. Biochem. 48, 193-216.

431 I., MILDNER, B. & K N O W L E S , B. (1986) Immunohistochemical localization of the epidermal growth factor in normal human tissue. Lab. Invest. 55, 588-92. G I L L , G. S . , B E R T I C S , P. J. & S A N T O N , J. B. (1987) Epidermal growth factor and its receptor. Mol. Cel. Endo. 51, 16986. F I S H E R , P. B., B O Z Z O N E , H . J. & W E I N S T E I N , I. B. (1979) Tumour promotors and epidermal growth factor stimulate anchorage-independent growth of adenovirustransformed rat embryo cells. Cell 18, 695-705. DAMJANOV,

F O X W E L L , B., T A Y L O R , T., G R E I N E R , OL]VIERI,

V. & RYFEEL, B.

B.I M I H A T S C H ,

M.,

(1988) Biotinylated r e c o m -

interleukin-2. A tool for research on the interleukin-2 receptor./. ImmunoI. Meth. 113, 221-9. H A R R I S O N , J. & A U E R S P E R G , N . (1983) Epidermal growth factor enhances viral transformation of granulosa cells. Science 222, 771-8. HEINEMANN, V. & J E H N , U. (t985) Wachstumsfaktoren. Klin. Wochenschr. 63, 740-6. H E N D L E R , F. J. & O Z A N N E , B. W. (1984) Human squamous cell lung cancers express incFeased epidermal growth factor receptors. J. Clin. Invest. 74, 647-51. H O P K I N S , D. R . , B O O T H R O Y D , B. & G R E G O R Y , H . (1981) Early events following the binding of epidermal growth factor to surface receptors on ovarian granulosa cells. Eur. J. Cell Biol.24, 259-65. H U N T E R , W . H . & G R E E N W O O D , F. C. (1972) Preparation of iodine-131 labelled human growth hormone of high specific activity. Nature 194, 495-6. binant

K A Y S E R , K.r G A B I U S I H . J . , C I E S I O L K A I

T . , E B E R T / W. &

BACH, S. (1989a) Histopathological evaluation of application of labelled neoglycoproteins in primary bronchus carcinoma. Hum. Pathol. 20, 352-60. K A Y S E R , K . , H E I L , M. & G A B I U S , H . J. (1989b) Is the profile of binding of a panel of neoglycoproteins useful as a diagnostic marker in human lung cancer? Path. Res. Pract. 184, 621-9. K O H N K E - G O D T , B. & G A B I U S , H . J. (1989) Heparin-binding lectin from human placenta: purification and partial molecular characterization and its relationship to basic fibroblast growth factors. Biochem. 28, 6531-8. MACIAS, BERG,

A., C.,

AZAVEDO, PEREZ,

E.I R.

HAGERSTROM,

& SKOOG,

L.

T.,

KLINTEN-

(1987)Prognostic

significance of the receptor for epidermal growth factor in human mammary carcinoma. Anticancer Rev. 7, 45964. NEWMAN,

W . , B E A L L r I. D . , LEVINE~ 1VI. A . , C O N E , J. L . ,

RANDHAWA,

Z. I. & B E R T O L I N I ,

O. R.

(1989) Biotiny-

lated parathyroid hormone as a probe for the parathyreoid hormone receptor. J. Biol. Chem. 264, 16359-66. NICHOLSON,

S.,

HALCHROW~

A N G U S , B., C H A M B E R S t

P.,

SAINSBURYI

P., FARNDORN,

]'. R.

C.,

J. R. & H A R -

A. r. (1988) Epidermal growth factor receptor (EGFr) status associated with failure of primary endocrine therapy in elderly postmenopausal patients with breast cancer. Br. J. Cancer. 58, 810-14. RIS,

R I O S / M. A . , M A C I A S , A . , P E R E Z t R. L A G E / A . & S K O O G , L.

(1988) Receptors for epidermal growth factor and estrogen as predictors of relapse in patients with mammary carcinoma. Anticancer Res. 8, 173-6. R O B E R T S , ~l. L.j F R E S T O N , J. A . & R E A D E , P. C. (1976)

432

Biotinylated e p i d e r m a l g r o w t h factor Suppression of immune responsiveness by a submandibular salivary gland factor. ImmunoL 30, 811-14.

ROSENBERG,

M.

B.I H A W R O T ~

E. & B R E A K E F I E L D ~

X. O.

(1986) Receptor binding activities of biotinylated derivatives of B-nerve growth factor. J. Neurochem. 46, 641-8. SlrOSCH~CK, C. M. & KING, L. ~. (1986) Role of epidermal growth factor in carcinogenesis. Cancer Res. 46, 1030-7. V E A L E , D . , K E R R , N . , G I B S O N , G. J. & H A R R I S , A. L. (1989) Characterization of epidermal growth factor receptor in primary human non-small cell lung cancer. Cancer Res. 49, 1313--17. W A T E R F I E L D z M.

D.~ M A Y E S r E.,

STROOBANTt

P.,

BEN-

NETT~ P . , YOUNG~ S., G O O D F E L L O W /

P., BANTINGr

G.

s. & OZANN~, B. (i982) A monoclonaI antibody to the human epidermal growth factor receptor. J. Cell. Biochem. 20, 149-61. WOJCHOWSKI, J. & CASLAKE, L. (1989) Biotinylated recombinant human erythropoietins: bioactivity and utility as receptor ligand. Blood 74, 952-8. YEM,

A.

W.I

STAITE,

ZURCHER-NEELY, N.,

HEINRiKSON,

H.

A.~ R I C H A R D t

K. A.~

R.

& DEIBEL,

(1989)

M.

Biotinylation of reactive amino groups in active recombinant human interleukin-l~. J. Biol. Chem. 264, 176917.