Int. J . Cancer: 46, 604-607 (1990) 0 1990 Wiley-Liss, Inc.

Publication of the International Union Against Cancer Publication de I’Union Internationale Contre 1e Cancer

A TWO-SITE ENZYME-LINKED IMMUNOSORBENT ASSAY FOR CYTOKERATIN 8 B. SUNDSTR~M and T. STIGBRAND Deparment of Physiological Chemistry, University of UmeB, S-901 87 UmeB, Sweden. A monoclonal enzyme-linked immunosorbent assay ELISA) was developed for the determination in biological luids of cytokeratin 8, a potential marker for malignant diseases. Two monoclonal antibodies (MAbs), TS 3 and T S 4, with different epitope specificity, were selected from 4 cytokeratin-8 reactive antibodies. T S 3 was used for coating and TS 4 as HRP-conjugate, respectively. Antibodies were selected with the aim of optimizing the discriminatory capacity between cytokeratin 8 levels in sera from healthy persons and from patients with malignant diseases. In sera from healthy individuals the mean value was determined to be 3.1 2 2.3 ng/ml with an upper cut-off level of 7.8 ng/ml (+ 2 SD) using purified cytokeratin 8 as standard. Sera from patients with colon cancer and pancreatic cancer were found to have significantly elevated levels, showing a 4- to more than 10-fold increase compared with the normal level. In patients with ovarian cancer no significant elevation was seen. Cytokeratin 8 monitoring may be of value for patients with colon and pancreatic cancer.


Intermediate filaments (10 nm filaments) constitute a complex family of cellular structures, which can be subdivided into 5 major groups. One of these is the cytokeratin group (40-70 kDa) which is found in epithelial cells. These display marked heterogeneity and 19 different polypeptides have so far been identified. All cytokeratins appear to have a central, conserved a-helical rod domain, flanked by highly variable amino- and carboxy-terminal domains, which generate the diversity within the group. Sequence data and immunological characteristics have shown that cytokeratins can be further divided into 2 large subfamilies. Type-I (acidic) and Type-I1 (neutral-basic) cytokeratins (Lazarides, 1980, 1982; Weber and Geisler, 1984; Steinart et al., 1985). One member of each subfamily is required to form heterotypic complexes in vivo and in vitro (Quinlan et al., 1984; Franke et a l . , 1984, 1987; Hatzfeld and Franke, 1985; Hatzfeld et al., 1987). These complexes display heterogeneity with regard to the constituting subclasses of cytokeratins and these are differently expressed in all epithelial cells (Moll et al., 1982, 1983a; Quinlan et a l . , 1985). Cytokeratins constitute a significant part of most epithelial cells. Cytokeratin 8 is particularly abundant and can be identified in simple epithelial cells, both normal and transformed (Moll et al., 1983a; Osborn and Weber, 1983; Quinlan et al., 1985). Since many malignant tumours exhibit large areas of necrosis, causing release of cytokeratins into surrounding tissues, the appearance of these proteins in the circulation could be a possible marker for cell necrosis. The pore-solubility propeaies of intact cytokeratins favour retention at the site of cell necrosis under physiological conditions. Following partial enzymatic degradation, cytokeratins become more soluble (Franke et a l . , 1987; Hatzfeld et al., 1987) and can be detected in the circulation using specific immunoassays. Cytokeratins have thus the potential of being useful tumour markers for carcinomas associated with cell necrosis. The introduction of hybridoma technology (Kohler and Milstein, 1975) has made it possible to generate highly specific immunochemical reagents and a battery of MAbs against tumour-derived cytokeratins has been established (Sundstrom et al., 1989). We now report a 2-site monoclonal enzyme immunoassay using 2 of these antibodies. MATERIAL AND METHODS

Source of antigen Cytokeratin 8 (CK 8) was extracted from MCF 7 cells

(Sundstrom et a l . , 1989) and further purified by preparative SDS-PAGE (Laemmli, 1970). The 52.5-kDa band was extracted from the gel with 0.1% SIX in 0.1 M NaHCO, for 17 hr at room temperature. The protein content was determined by amino acid analysis. The preparation contained more than 90% CK 8 as judged by analytical SDS-PAGE followed by silver staining (Ansorge, 1982). This preparation was diluted to 5 kg/ml with 0.5% BSA, sodium phosphate buffer 50 m ~ PH, 7.5 was stored in aliquots at - 201°C until use. Monoclonal antibodies Four MAbs, TS 1, 3, 4 and 7 produced against tumourderived cytokeratins (Sundstrom et al., 1989) were used. The hybridoma lines were injected into Pristane-primed BALB/c mice for production of ascites. The immunoglobulin fractions were purified on Protein-A-Sepharose (Pharmacia, Uppsala, Sweden), and absorbed in 0.8 M glycine/NaOH, 3 M NaCl PH 8.9 and eluted with 0.1 M citrate buffer, pH 6.0. Horse-radish peroxidase conjugates (HRP) of the 4 MAbs were prepared according to Engvdl (1980) using glutaraldehyde for conjugation; 5 mg HRP (Sigma, St. Louis, MO) were used per mg protein A purified MAb and the conjugates were stored at 4°C in glyceroUwater 1:l (v/v). Monoclonal enzyme immunoassay fiw cytokeratin 8 Microtiter plates comprising 96-wells (Nunc, Roskilde, Denmark) were coated with 100 ~l of a solution containing 5 or 10 pg/ml protein-A-purified Mkb. The plate was incubated for 3-17 hr at 8°C and washed 3 times with 1% NaC1,0.05% Tween 20. Purified CK 8 (0-300 nglml) was used as standard, serially diluted in normal human serum or with 5% BSA in 50 m~ sodium phosphate-buffered saline, PH 7.5. Serum samples were diluted in 10 m~ phosphate-buffered saline, 0.05% Tween 20 PH 7.2 (PBSITween) supplemented with a non-related MAb, and 1OO-lp1 aliquots were assayed in duplicate. After washing, PBS/Twe:en-dilutedHRP-conjugated antibodies were allowed to react with antigen. After washing, enzyme activity was determined in citrate-phosphate buffer (0.1 M , PH 5.0) using 0.7 g o-phenylenediamine and 0.4 ml H,0,/1, and the absorbance at 492 nm was recorded after addition of 50 p1 H,SO,.


Clinical samples Venous blood samples were obtained from 70 healthy blood donors. Sera from a total of 89 patients exhibiting 50 ovarian epithelial malignant tumours, 11 colon cancers, 21 pancreatic cancers and 7 of other origin were obtained from Radiumhemmet, Karolinska Hospital and the Dlepartment of Gynaecology, University of UmeP. After clotting and centrifugation, sera were decanted and stored at - 20°C until use. RESULTS

Optimization of the assay In order to optimize the assay conditions, each step was tested to assure maximal efficiency. The coating conditions for all MAbs were investigated. Generally, 3 hr or more at + 8°C were required to obtain maximal coating, using 3 pg/ml MAb Received: April 19, 1990.




TS 1 with HRP-TS 4 n 50

% (A-AO)/AlOO


401 30 20 10

TS 3

TS 1

TS 4

TS 7 0

0HRP-18 4 HRP-18 7 FIGURE1 - Two-site sandwich ELISAs using different MAb combinations reactive with cytokeratin 8. Antibodies used for coating and as conjugate are indicated. HRP-18 1

HRP-18 3

in the coating solution. Coating at room temperature for up to 19 hr yielded only 70-90% of the values obtained at lower temperature. The PH dependence of the coating efficiency at PH 7.5; 8.5 and 9.6 was also tested and was found to vary for the different antibodies. For TS 1 and 7,O. 1 M NaHCO, PH 8.5 gave high and reproducible values. TS 3 gave similar coating at both PH 7.5 (phosphate-buffered saline) and PH 8.5 (0.1 M NaHCO,). TS 4 was found to coat optimally at 0.05 M NaHCO, PH 9.6. The incubation time with the samples was determined to be 3-6 hr in order to obtain the same reactivity with both standard and samples. The HRP-coupled MAb used as conjugate in the last step was used at a dilution of 1500 in PBS/Tween (approx. 100 ng conjugate/well) to give maximal sensitivity. Optimal incubation time with all MAb-conjugates was found to be 2-3 hr. Longer incubation resulted in loss of sensitivity. Of all combinations of antibodies tried, 16 different assays are presented (Fig. 1). As seen, when TS 1, TS 3 or TS 7 was used for coating, binding of HRP-conjugated TS 1, TS 3 and TS 7 was inhibited, confirming involvement of partially related epitopes. Of the 5 combinations tested, the 3 which gave the highest reactivity with CK 8 (Fig. 2) were further investigated by testing discriminatory potential using serum samples from cancer patients and normal individuals. The highest discriminatory capacity when normal and cancer sera were compared was found for the combination of TS 3 and HRP-TS 4 (Fig. 3). A standard curve with an imprecision profile is shown in Figure 4 covering a quantitation interval of 0-333 ng CK 8/ml. The least detectable dose was calculated to be below 3 ng/ml.

ovarian cancer

i Wnormal sera

TS 3 with HRP-TS 4 6o ?f(A-AO)/AlOO ~ - _ _

_ . _ . _ _ _ - - - ~







iI m ovarian cancer

X (A-AO)/AX)O 100,


Tk IlHRP-f8 4 18 WHRP-18 4 18 4/HRP-l8 3 18 4/HRP-T8 7 18 7/HRP-T8 4

Ipoaltlv 1


m poaltlv 2

0poaltlv 3 0normal 2 FIGURE2 - Reactivity pattern in 5 different assays for cytokeratin 8 using 5 serum samples, 3 of which contain different combinations of cytokeratins (positive 1-3), and 2 blood donor sera (normal 1-2).

LFiS normal sera FIGURE3 - Discriminatory capacity for 3 selected assays (TS 1/ HRP-TS 4, TS 3/HRP-TS 4, TS 7IHRP-TS 4) in tests with sera from normal and ovarian cancer patients.

The amount of non-related MAb added to the serum samples was titrated to ascertain that the response obtained was not related to non-specific interactions, i.e. cross-linking antibodies which are found in many sera (Boscato and Stuart, 1986, 1988; Zweig et al., 1988). As seen in Figure 5, as much as 5 p,g/ml was required for some of the sera tested, while the non-specific effect was eliminated with lower amounts for other sera.

El normal I

Clinical results Figure 6 summarizes the serum levels for normal sera, and sera from tumour patients. Healthy males and females exhibit






A 492 nm 1.21



0.8l b \


0' 0



-etandard wrw



- x- Impreciaion profile

FIGURE 4 - Standard curve with an imprecision profile for cytokeratin 8 assay using TS 3 for coating and HRP-TS 4 as conjugate.


abrun ?


marurn 4


norurn 2


barurn 3

oarurn 5



FIGURE 5 - Effects of different amounts of added non-related antibody for elimination of non-specific interaction in the 2-site ELISA for cytokeratin 8 . Five different serum samples were tested.

similar levels. The mean value was 3.1 f 2.3 ng/ml (mean 2 SD) which gives an upper cut-off level of 7.8 ng/ml. In sera from colon cancer patients cytokeratin 8 was highly significantly @ < 0.05) elevated with a mean value of 270 ? 535 ng/ml (mean t SD). All samples tested were above the upper cut-off level. An elevated serum concentration of CK 8 was also seen in sera from patients with pancreatic carcinomas (26.5 t 9.8 ng/ml, mean 2 SD). Despite the high incidence of samples with high CK 8 values (20/21 patients), the mean value in this group was only moderately above the cut-off level. Finally, in the group of ovarian cancers, only 20-25 % of the tested samples had elevated values of CK 8.


.... ..... ... *.

-11 1



One means of increasing organ specificity and discriminatory capacity of tumour markers is to select those which are selectively expressed in defined organs. Cytokeratins, which are selectively expressed in a very complex manner, in human tissues are a good example. Out of the 19 immunochemically different cytokeratins, CK 8 is highly expressed in certain human tissues and tumours derived from these tissues. Colon mucosa and adenocarcinomas derived from colon, for instance express CK 8, 18, 19, with CK 8 representing approximately 50% (Moll et al., 1982; Hatzfeld and Franke, 1985). Adenocarcinomas of pancreas express, in addition to the above, CK 7 and 17, resulting in a lower relative content of CK 8. Ovarian tumours contain cytokeratins 7, 8, 18 and 19 (Moll et al., 19836), giving patterns similar to pancreatic adenocarcinomas. Here we describe a 2-site ELISA for cytokeratin 8. The antigen used for generating antibodies was purified from solid human carcinomas (Sundstrom et a/., 1989) in order to ascertain immunoreactivity with epitopes relevant for monitoring of tumour patients. All antibodies investigated detect epitopes in the central rod portion of cytokeratin 8. The forms of cytokeratins with highest solubility usually contain the rod portion, which makes this assay favourable from a theoretical point of view, for use in clinical practice. Furthermore, the antibodies constituting the assay were selected with the aim of increasing their discriminatory capacity when control samples were compared to those from tumour patients. A total of 16 different combinations of antibodies was



ug/ml in test solution

'0 330

... lil..!. ......


....... .::.


....... :I:::::







J ovarlan

FIGURE 6 - Serum levels of cytokeralin 8 (nglml) in normal sera and sera from patients with colon, pancreas and ovarian cancer. The upper normal cut-off level is indicated ('7.8 ng/ml) by arrows. tested for that purpose. The MAbs that were selected, TS 3 and TS 4, have been shown by competitive experiments to bind to different determinants (Sundstrom ei' al., 1989) and to differ with respect to immunohistochemical staining patterns of normal human tissues (Sundstrom el al., 1989). The clinical samples that we have tested demonstrate the usefulness of the assay in monitoring certain groups of carcinoma patients. In patients with colon adenocarcinomas in particular, elevated values amounting to 4 to 100 times the upper cut-off level were observed. This is in good agreement with expression of CK 8 in such tumours. In sera from patients with pancreatic cancer, a high incidence of elevated values but lower amounts were determined. This observation is in agreement with the lower expression of CK 8. For patients with ovarian epithelial tumours, low incidence and low values were observed. Other human tissues which express CK 8 as major type-I1 cytokeratin are: hepatocytes, small intestinal mucosa, epithelial lung cells, oesophagus, stomach and breast cells. The significance of this CK 8 2-site assay has not yet been investigated in patients with tumours derived from these tissues.


ANSORGE,W., Fast visualisation of protein bands by impregnation in

potassium permanganate and silver nitrate. In: D. Stathakos (ed.) Efectro-

phoresis, Vol. 82, pp. 235-242, de Gruyter, Berlin (1982).

BOSCATO, L.M. and STUART, M.C., Incidence and specificity of interference in two-site immunoassays. Clin. Chem., 32, 1491-1495 (1986).

BOSCATO,L.M. and STUART,M.C., Heterophilic antibodies: a problem for all immunoassays. Cfin. Chem., 34, 27-33 (1988). ENGVALL, E., Enzyme immunoassay ELISA and EMIT. Meth. Enzymol., ' O 9 419439 (1980). FRANKE, W.W., SCHILLER, D.L., HARTZFELI),M., MAGIN,T.M., JORE-



CANO, J.L., MITTERNACHT, S . , COHLBERG, J.A. and QUINLAN,R.A., peptide patterns observed in certain human carcinomas. Differentiation, Cytokeratins: complex formation, biosynthesis, and interactions with des- 23, 256-269 (1983a). mosomes. In: A.J. Levine, G.F. Van de Woude, W.C. Topp and J.D. MOLL,R., LEVY,R., CZERNOBILSKY, B., HOH~WEG-MAJERT, P.,DALWatson (eds.), Cancer cell, the transformed phenotype, pp. 177-190, LENBACH-HELLWEG, G. and FRANKE, W.W., Cytokeratins of normal epCold Spring Harbor Laboratory, New York (1984). ithelia and some neoplasms of the female genital tract. Lab. Invest., 49, FRANKE, W.W., WINTER,S., SCHMID,E., S ~ L L N E P., R , HAMMERLING, 599-610 (19836). G. and ACHSTATTER, T., Monoclonal cytokeratin antibody recognizing a OSBORN,M. and WEBER,K.,Tumour diagnosis by intermediate filament heterotypic complex: immunological probing of conformational states of typing: a novel tool for surgical pathology. Lab. Invest., 48, 372-394 cytoskeletal proteins in filaments and in solution. Exp. Cell Res., 173, (1983). 17-37 (1987). QUINLAN, R.A., COHLBERG, J.A., SCHILLER,D.L., HARZFELD, M. and HATZFELD, M. and FRANKE, W.W., Pair formation of cytokeratins: For- FRANKE, W.W., Heterotypic tetramer (A2D2)complexes of non-epidermal mation in vitro of heterotypic complexes and heterologous recombinations keratins isolated from cytoskeletons of rat hepatocytes and hepatoma cells. of purified polypeptides. J . Cell Biol., 101, 18261841 (1985). J. mol. Biol., 178, 365-388 (1984). HATZFELD,M., MALER,G. and FRANKE, W.W., Cytokeratin domains QUINLAN,R.A., SCHILLER,D.L., HATZFELD,M., ACHTSTATTER, T., involved in heterotypic complex formation determined by in-vitro binding MOLL,R., JORECANO, J.L., MAGIN,T.M. and FRANKE, W.W., Patterns assay. J . mol. Biol., 197, 237-255 (1987). of expression and organization of cytokeratin intermediate filaments. In: K ~ H L E RG.T. , and MILSTEIN,C., Continuous cultures of fused cells se- E. Wang, D. Fishman, R.R.H. Liem and T.-T. Sun (eds.), Intermediate creting antibody of predefined specificity. Nature (Lond.), 256, 4 9 5 4 9 7 filaments, Ann. N.Y. Acad. Sci., 455, 282-306, (1985). (1975). STEINART, P.M., STEVEN,A.C. and ROOP,D.R., The molecular biology LAEMMLI, U.K., Cleavage of structural proteins during the assembly of the of intermediate filaments. Cell, 42, 411-419 (1985). SUNDSTR~M, B.E., NATHRATH, W.B.J. and STIGBRAND, T., Diversity in head of bacteriophage T4. Nature (Lond.), 227, 680-685 (1970). LAZARIDES, E., Intermediate filaments as mechanical integrators of cellu- immunoreactivity of tumour derived cytokeratin monoclonal antibodies. J. Histochem. Cytochem., 37, 1848-1854 (1989). lar space. Nature (Lond.), 282, 249-256 (1980). wool WEBER,K . and GEISLER,N., Intermediate filaments-from LAZARIDES, E., Intermediate filaments: a chemically heterogeneous, de- a-keratins to neurofilaments. A structural overview. In: A.J. Levine, G.F. velopmentally regulated class of proteins. Ann. Rev. Biochem., 51, 219Van de Woude, W.C. Topp and J.D. Watson (eds.), Cancer cell, the 250 (1982). transformed phenotype, pp. 153-159, Cold Spring Harbor Laboratory, MOLL,R., FRANKE, W.W. and SCHILLER, D.L., The catalog of human New York (1984). cytokeratins: patterns of expression in normal epithelia, tumours and cul- ZWEIG,M.H., CSAKO,G. and SPERO,M., Escape from blockade of intured cells. Cell, 31, 11-24 (1982). terfering heterophilic antibodies in a two-site immunoradiometric assay for MOLL,R., KREPLER,R. and FRANKE, W.W., Complex cytokeratin poly- thyrotropin. Clin. Chem., 34, 2589-2591 (1988).

A two-site enzyme-linked immunosorbent assay for cytokeratin 8.

A monoclonal enzyme-linked immunosorbent assay (ELISA) was developed for the determination in biological fluids of cytokeratin 8, a potential marker f...
395KB Sizes 0 Downloads 0 Views