NO860676L - PROCEDURE AND MEANS OF IN VITRO CANCER VIEWING. - Google Patents

Description

IN VITRO DETERMINATION OF STOMACH AND INTESTINAL CANCER

The present invention primarily relates to methods and systems for diagnosing mucinous cancers, and in particular cancer in the intestinal tract as well as cancer in the breast, lung, stomach, ovary and the like.

Mortality due to cancer of the intestinal tract ranks second after lung cancer in men and breast cancer in women. Although much attention has been directed to these conditions, the number of deaths has not been significantly reduced in recent years. Early detection of the cancer is important to achieve healing in those who are attacked. Cancer cells sometimes form substances that one cannot

detect or which are only found in very small concentrations in healthy individuals. These features can be used to develop innovative diagnostic techniques. Tests that aim to detect these cancer-related substances are of great help in differentiating between the individual with cancer compared to healthy individuals.

Carcinoembryonic antigen (CEA) is the most extensively investigated antigen in connection with bowel cancer and also the most studied of all the tumor-associated antigens. This antigen is present in the human fetal intestinal tract,

but are not present or present only in very small concentrations in the normal fully developed large intestine.

CEA was first detected in connection with colon cancer in humans by Gold and Freedman in 1965. 9 CEA was present in the blood of such cancer patients and was therefore first seen as an oncofetal antigen that gave great hope for an early diagnosis of cancer in the human digestive system. Later, it has been shown that measurement of CEA in blood is not specific enough to be valuable enough for an early diagnosis of bowel cancer, because it is often elevated in a number of non-neoplastic conditions that include bronchitis in smokers and liver damage. E.g. , 10% of the normal population show false positive results, in that these individuals have elevated concentrations of

CEA in blood, even if they do not have cancer or even benign tumors. Second, CEA is often undetectable

11-13

even in patients with rather large tumours, i.e. that the number of false negatives is significant (around 50% of patients with bowel cancer do not show increased blood concentrations of CEA). Despite these limitations, the CEA assay is still of significant importance in the preoperative assessment and postoperative treatment of cancer patients. CEA assays make up around 40% of current cancer tests.

Alpha-fetoprotein (AFP) is another oncofetal antigen that

have been found in high concentrations in fetuses but not in normal adults. Increased concentrations of AFP in blood are present in patients with primary liver cancer, secondary liver tumors, in most cases of some types of testicular cancer (teratoma); but also in case of toxic liver damage and liver cirrhosis. Despite this apparently limited utility, AFP represents about 20% of currently performed cancer tests.

More recently, monoclonal antibody-based radioimmunoassays

TM

for painting tumor-associated antigen CA 19-9 became 14-19 TM

commercially available. This CA 19-9 assay is claimed to have high sensitivity and specificity for all stages of pancreatic cancer, but with regard to the detection of colorectal cancer, this assay appears to have low sensitivity.

In 1980, a new mycin glycoprotein antigen was isolated from cancer of the colon.''" This compound is also present in normal developing small intestine and has therefore been named "small intestinal mycin antigen" (SIMA). SIMA was shown to be an oncofetal antigen , due to the fact that in an 8-12 week old fetus it is found in the entire gastrointestinal tract, but at birth it is no longer detected elsewhere than in the small intestine under normal conditions, but SIMA has been found in cancer in many various parts of the gastrointestinal tract and other organs. Another mycin compound, referred to as "colon mycin antigen" (LIMA) has also been isolated from normal colon. LIMA has also been shown to be an oncofetal antigen, and are formed to varying degrees in a variety of cancers. Extensive immuno-histological studies have shown that LIMA is antigenically distinct from SIMA and both are also distinct from CEA. These gut-associated mycin antigens have also been detected in some gastric tumors, gallbladder , lung, breast and ovary, as well as the pre malignant conditions in the above-mentioned organs. (Litt. ref. 1-8).

Results from immunohistochemical staining, using anti-SIMA and anti-LIMA monoclonal and polyclonal antibodies, of normal fetal and fully coiled tissue, and cancer from various organs are shown in table 1. It has now been discovered for the first time, that despite the high molecular weight of these mycinous glycoproteins, SIMA and LIMA can be detected in the blood serum of a large proportion of patients with colon cancer using diagnostic antibody assays. SIMA and LIMA are not found in measurable amounts or are found in very small amounts in the blood serum of healthy control individuals. Initial studies with a small group have already shown that a combination of SIMA and LIMA tests can be used to detect about 80% of patients with colorectal cancer, where the use of CEA tests alone detects only about 56% of such cancers. These SIMA and LIMA assays are thus more sensitive than CEA assays for detecting stage B and C (Duke's classification) cancers. In addition, LIMA can be detected in the blood of a large proportion of patients with ulcerative colitis in the colon. Such patients usually develop colon cancer. Likewise, SIMA and LIMA can be used to detect the presence of pre-cancerous conditions.

As used in this specification, the terms SIMA and LIMA are used not only to denote the mycin antigens which are described in more detail below, but also parts thereof which are antigenically related and which may occur in physiological fluids, especially blood serum or plasma.

In accordance with the present invention, an in vitro diagnostic method has been provided for detecting the presence of cancer cells or other cells in a patient that form mycin antigens, where the method comprises analyzing a sample of a physiological fluid from a patient in order to demonstrate the presence or absence of SIMA and/or LIMA in the above sample.

Preferably, the above-mentioned method comprises bringing a sample of a physiological fluid from the patient into contact with monoclonal or polyclonal antibodies against SIMA and/

or LIMA or fragments of the above antibodies, or the above antibodies or fragments thereof labeled with a

marker capable of providing a detectable signal, and detecting the presence of the binding between the above-mentioned antibodies or fragments thereof, or the above-mentioned labeled antibodies or fragments thereof, and their corresponding antigen.

Particularly preferred physiological fluids for use in methods of diagnosis are blood, blood serum or plasma.

On the other hand, one has provided an in vitro diagnostic kit or system for detecting the presence of cancer cells or other cells in a patient that form mycin antigens, where the kit or system comprises means for detecting the presence of SIMA and/or LIMA in a sample of a physiological fluid from the patient.

Particularly in accordance with this invention, there is provided an in vitro diagnostic kit or system for detecting the presence of SIMA and/or LIMA in a sample of a physiological fluid, such as blood or blood serum, from the patient, wherein the kit or system comprises monoclonal or polyclonal antibodies capable of an immunological reaction with SIMA and/or LIMA, or fragments thereof, together with indicator means for indicating the presence of an immunological reaction between the above antibodies and their corresponding antigens.

In a particular embodiment in accordance with the present invention, the diagnostic system further comprises a compound which is bound to a carrier of solid matter, wherein the above-mentioned compound comprises either

(i) SIMA and/or LIMA, or (ii) the above antibodies capable of immunologically reacting with SIMA and/or

LIMA.

This aspect of the present invention comprises a number of embodiments where different forms of indicator agents are used. In one embodiment of the present invention, the indicator means comprise a marker capable of providing a detectable signal, and wherein said marker is bound to said antibodies capable of reacting immunologically with SIMA and/or LIMA. In another embodiment, however, the indicator means comprise a marker capable of providing a detectable signal, wherein the above-mentioned marker is bound to other antibodies, wherein the above-mentioned second antibody is raised against the first-mentioned antibody and indicates the presence of the above-mentioned immunological reaction by binding to the the former antibody. Furthermore, the indicator agents may comprise protein A from staphylococci with a marker attached thereto. The markers used in any of these embodiments may include known enzymes, radioactive elements, fluorescent chemicals or compounds such as biotin.

The use of immunoassays for SIMA and LIMA has made it possible to detect these mycin antigens in all stages of colorectal cancer. These results therefore establish that this system is, if possible, more applicable and valuable

TM

than the CA 19-9 or CEA assays described above.

While the present invention is primarily directed at the detection of mycinous cancer, it will also be appreciated that immunoassays for SIMA and LIMA are valuable for the detection of inflammatory conditions such as ulcerative colitis or gastric ulcers, as well as the detection of pre-cancerous conditions.

As made clear in the preceding description, the mycin antigens SIMA and LIMA described herein can primarily be characterized by the source from which they are isolated

(using methods described in detail below). A further characterization and identification of these antigens can be achieved by using polyclonal and monoclonal anti-SIMA and anti-LIMA antibodies, and by using

immunohistochemical techniques, "competitive" immunoassay techniques to detect antigenic activity by inhibition of binding between anti-SIMA or anti-LIMA antisera and specific gastrointestinal tissue.

Further details of the present invention will be discussed below. Incidentally, it is noted that immunoassays for SIMA and/or LIMA in accordance with the present invention provide a very sensitive diagnostic tool for the detection of cancer cells or other mycin antigen-producing cells in a patient,

and in particular in the pre-operative phase in cancer patients to obtain an indication of the extent of the tumor,

and to obtain a baseline valuable for post-operative follow-up of patients who are operative, or by any other modality (e.g., radiotherapy, chemotherapy) to detect recurrence at the treatment stage, or to measure treatment efficacy efficiency. In such application, the diagnostic system in accordance with the present invention can primarily be combined with immunoassays for CEA or other known tumor-associated antigens, which are commonly used in such cases. In addition, the SIMA/LIMA (optionally in combination with CEA) diagnostic systems can be used to screen a population at risk of developing intestinal cancer, i.e. patients with ulcerative colitis or peptic ulcer, an area where the CEA diagnostic the systems alone have, after initial testing, been found to have no diagnostic value.

During work that has led to the present invention, a number of monoclonal antibodies have been produced by means of immunization of mice with mycin preparations extracted from colon cancer. Immunohistochemical labeling of tissue was used to classify 5 of these monoclonal antibodies as anti-SIMA (reactive with colon cancer and normal small intestine) and 5 as anti-LIMA (reactive with colon cancer and normal colon). Immunochemical analyzes showed that the affinity of the antibodies varied by as much as 100-fold for each class.

The following examples illustrate the production and characterization of mycin antigens, and commonly used methods for the production of polyclonal and monoclonal antibodies against SIMA and LIMA, and the use of these monoclonal antibodies in in vitro diagnostic methods for detecting the presence of cancer or other mycin antigen-producing cells in a patient in accordance with the present invention.

EXAMPLE 1

A. PURIFICATION AND CHARACTERIZATION OF MYCIN ANTIGENS.

Mycins were recovered from surgical specimens obtained by removal of colorectal cancer tissue. Areas of normal colon distant from the tumor were obtained from tissue removed along with the tumor and were examined by light microscopy to confirm that the tissue was normal before use. Samples from normal duodenum, jejunum, and colon were also obtained by autopsy from randomly selected individuals. Mycins were extracted from samples from colon cancer (LIMA and SIMA), normal colon (LIMA), or normal duodenum or jejunum (SIMA)

using the methods described below.

The mycin preparation used to immunize mice for monoclonal antibody production and screening was prepared from a cancer specimen diagnosed histologically as a "partially mycinous" adenocarcinoma of the sigmoid colon. The tissue sample (approximately 5 g) was cut into small pieces, then homogenized in 10 volumes of 4 M guanidine hydrochloride containing 24 mM EDTA, 10 mM N-ethylmaleimide, and 1 mM benzamidine HCl. After centrifugation at 20,000 xg for 20 min, the pellet was re-extracted using fresh guanidine hydrochloride. The mycin in the combined supernatants was then adjusted to a density of 1.35 g/ml with cesium chloride and centrifuged at 105,000 xg for 64 hours. The resulting gradient was separated into 6 equal fractions with densities ranging from 1.25 to 1.55 g/ml. Each fraction was dialyzed against distilled water and analyzed for protein (modified Bradford method) and hexose 21, as well as antigenic activity by determining its ability to inhibit specific immuno-fluorescent labeling of gastrointestinal tissue by by anti-SIMA and anti-LIMA antibody, and by means of immuno-assays using monoclonal antibodies (see more detail below). The fractions containing the greatest activity of antigen were re-fractionated on CsCl gradients,

and the fractions were analyzed as described above.

(i) LIMA

LIMA was found at the bottom of the gradient in the 2nd or 3rd fraction with a density of !.34 to 1.55 g/ml. The LIMA preparations were polydisperse and were usually found to have the highest antigenic activity at a density of around 1.4 - 0.08. Antigen activity, hexose and protein content in each fraction of a typical 3 CsCl gradient centrifugation of a LIMA preparation is shown in fig. let. Hexose/protein ratio (wt/wt)

for fractions 2 and 3 were respectively 0.5 and 0.25. The peaks denoting hexose, protein and antigen activity did not coincide; fraction 3 contained most of the antigenic activity. The total yield of LIMA from a typical experiment,

determined by enzyme-linked immunoassays (EIA) is about 35% of what is present in a tissue homogenate. The total purification of LIMA is in the order of 1000 times. SIMA is not usually detected in LIMA preparations from

normal colon. The successful use of LIMA preparations in "sandwich" immunoassays (where the same monoclonal antibody must bind more than once to the molecule) indicates that the antigenic determinant recognized by the monoclonal antibody must be repeated at least twice on the molecule.

The high molecular weight of LIMA was indicated by its inability to penetrate 7% polyacrylamide electrophoresis gels. Further studies using gel filtration chromatography showed that the LIMA antigen eluted in the exit volume on Sepharose 6B (which separates globular proteins in the approximate molecular weight range: 1 x 104 - 4 x 106, and dextrans with molecular weight: 104 - 100 <6>specifications from the Pharmacia catalogue) and Sepharose 2B (fractionates globular proteins in a molecular weight range of 7 x 10<4>4 x 10 7 and dextrans in the range: 10 5 -2x10<7>). Gel chromatography of LIMA on a Sephacryl 1000 column (1 cm x 65.5 cm) eluted in 10 mM phosphate

buffered salt solution (separates dextrans in a molecular weight range from: 5 x 10 r - 10 Q) is shown in fig. 2a. Most of the LIMA antigen activity was contained in this matrix and was eluted as a broad peak which may indicate polydispersity of the LIMA preparation. A similar degree of polydispersity is seen for many other proteoglycan molecules and the reason for this can e.g. be structural variations or an aggregation that affects the molecular weight, but not the antigenic effect of such molecules.

To better estimate the size of LIMA, ultracentrifugation was performed on glycerol gradients (10% to 20%) on a number of preparations to determine the sedimentation coefficients (Beckman SW 50 Ti roto, 22,000 rpm x 15 h or 42,000 rpm . x 5 h at 4°C). The gradients were fractionated and each fraction was analyzed for antigen activity using a "sandwich" assay. A typical LIMA gradient is shown in fig. 3a. From this and other experiments, average s-values with respect to weight were determined according to the following equation:

where ^<>> is the angular velocity, t is the centrifugation time (hours), Ro is the radius of rotation of the top of the gradient and Rt is the mean weight radius to which the peak of antigen activity has moved. The average s-value calculated from a number of different trials was 9.5 1.5.

LIMA was subjected to a number of different physical,

chemical and enzymatic treatments to obtain additional information regarding the nature of the antigen; this is summarized in table 2. The antigen activity for LIMA in a "sandwich" assay was stable when stored for at least 1 month at 40°C and up to 5 min. at 100°C. Boiling for 10 min. or longer resulted in a 40% loss in terms of activity. This stability of LIMA to boiling suggests that this antigen may be a highly glycosylated molecule and that the epitope contains or is protected by carbohydrate. Alkaline treatment of LIMA to break 0-glycosidic bonds between carbohydrates and protein was carried out at different concentrations of KOH: 0.01M, 0.05M, 0.10M, 0.25M, 0.50M (at 4°C for 2, 4 or 16 hours) after which it respectively

there were 85%, 70%, 70%, 55% and 45% antigen activity after treatment. Addition of a nucleophile such as 2-mercaptoethanol (0.1M) enabled the 3-elimination reaction to proceed under milder conditions (0.01M KOH,

at 4°C for 4 hours), with a resulting loss of 55% of antigenic activity. In addition to the partial loss of antigenic activity due to alkaline treatment of LIMA, analyzes on glycerol gradients showed a significant decrease in the average molecular weight value s (5.5) indicating a decrease in molecular weight (Fig. 3a). Experimental reduction of LIMA (in 6M guanidine HCL, 0.5M tris pH 8.1,

0.002M EDTA, 10 mM dithiothreitol, at 50°C for 2 hours, 4 hours or 16 hours followed by alkylation with 0.4M iodoacetic acid at 4°C overnight) had no significant effect on antigen activity.

Cleavage experiments with a number of very pure enzymes were

performed to determine the chemical nature of the LIMA antigen.

After treatment, the samples were boiled for 5 min. or neutralized (pepsin) to inactivate the enzymes for immunoassay. Neither pepsin nor clostripain (performed at a range of enzyme concentrations incubated with LIMA for 2, 4 or 16 hours) showed any significant effect on the antigenic activity of LIMA (table 2). Cleavage with papain (0.1 or 1.0 mg/ml) of LIMA resulted in a significant loss of antigenic activity with only 20% activity remaining after 2.4 or 16 h of cleavage, when activity was measured by of a "sandwich" assay. When the activity was measured in a competitive ELISA (where the measurement does not depend on the multivalence of the epitopes on the antigen) the activity remaining after 2 hours of cleavage with papain was 75%, after 4 hours - 65% and after 16 hours - 35%. This implies that at mild conditions of papain cleavage, LIMA can be degraded into fragments that mostly have only one epitope and therefore cannot be detected in a "sandwich" assay, but are detected in a competitive assay. These results show that the antigen is to a certain extent sensitive to papain, while it is essentially resistant to pepsin and clostripain.

(ii) SIMA

SIMA was found in 2 or 3 fractions of the gradient with a density between 1.29 and 1.45 g/ml. The SIMA preparations were polydisperse, but most of the antigenic activity was usually found at a density of 1.38 0.05. The antigen activity, hexose and protein content in each fraction of a typical third CsCl gradient centrifugation of a SIMA preparation is shown in fig. lb. The ratios between hexose/protein (wt/wt) for fractions 3 and 4, where the greatest antigenic activity is found, were 1.5 and 0.3 respectively. The yield of SIMA from a typical colon cancer sample is around 40%, determined by EIA. The purification achieved by these methods is about 1000 times, similar to the purification of LIMA. Most colorectal cancer tissue was found to contain significant amounts of LIMA as well as SIMA, while extracts from samples of normal duodenum and jejunum contained SIMA but no detectable amounts of LIMA. LIMA could be removed from "SIMA" preparations from colorectal cancer samples by passing the mixed mycin extracts through an anti-LIMA monoclonal antibody affinity column. The identification of SIMA in the preparation is based on the reaction with monoclonal antibodies and inhibition of immunofluorescence labeling of the small intestine by polyclonal and monoclonal antibodies. As described for LIMA, the use of SIMA in a "sandwich" immunoassay indicates the presence of a repeated epitope on the antigen molecule.

As was the case with LIMA, SIMA also does not penetrate 7% polyacrylamide electrophoresis gels. However, analysis by gel filtration on a Sepharose 2B column (1.5 cm x 7 cm), eluted in 25 mM tris pH 8.0 containing 4 M guanidine HCl, showed that SIMA was included in the gel and was polydisperse, where elution obtained a series of peaks (Fig. 4), and had a much lower molecular weight than LIMA. SIMA was also analyzed in a Sephakryl 1000 column, where again upon elution a broad peak with a much lower molecular weight than the entire LIMA preparation was obtained (see fig. 2a and b).

The s values were determined for a series of SIMA preparations on 10% to 20% glycerol gradients as described above for LIMA.

A typical SIMA gradient is shown in fig. 3b. The average s-value was calculated from a number of trials and was 4.8 - 1.4, which is much lower than the value for LIMA.

SIMA was found to be more stable to alkaline treatment and boiling than LIMA. No significant change in antigenic activity was observed, with alkaline treatment at conditions described above for LIMA. In addition, SIMA antigenic activity was not significantly affected by boiling for 5 min., and only 6% of the activity was lost after boiling for

10 minutes (see table 2).

SIMA was degraded using a series of purified enzymes as described for LIMA (Table 2). Clotripain had no effect on the antigenic activity of SIMA as measured by sandwich EIA. Degradation of SIMA using papain for 2 hours had no effect, while 50% of the antigenic activity was lost after 16 hours of degradation. Degradation of SIMA by pepsin resulted in a 70% loss of antigenic activity, in contrast to LIMA which was resistant to pepsin (10 pg/ml pepsin at room temperature overnight).

A number of refinements were introduced in the purification procedure of mycin as follows: 1. Perchloric acid (IM), treatment with sodium acetate and heating to 70°C, or boiling for 5 min. can also be used to improve the purity of the mycin preparations. 2. Repeated CsCl gradient ultracentrifugation has also been shown to give a markedly improved purity (up to 10

times) of mycin preparations.

3. Affinity monoclonal antibody chromatography can also be used in two ways: first eg, using anti-LIMA monoclonals to remove LIMA from a SIMA preparation (and vice versa); and second, by using adsorption and elution of mycin (SIMA) on anti-SIMA monoclonals to purify SIMA from other contaminating proteins. 4. Mycin antigens can thus be purified from normal and cancerous tissues using lectin columns, gel electrophoresis, ion exchange or gel filtration chromatography.

The various steps in the purification of mycin preparations are checked using competitive immunohistochemistry and EIA using polyclonal and monoclonal antibodies to determine SIMA and LIMA compounds in the preparations.

B. PREPARATION OF POLYCLONAL ANTISERA

The polyclonal anti-SIMA antiserum used in

these investigations were those made in rabbits by Ma et al (1980)''. Polyclonal antiserum recognizing LIMA was prepared by similar techniques involving the immunization of rabbits with mycin preparations extracted from normal colon in accordance with Ma et al (1980) "'".

These antisera, although relatively non-specific and cross-reactive, were initially used to control the steps in the purification of mycin. Polyclonal antisera prepared from very pure SIMA and LIMA can be used in "two-side sandwich" immunoassays as described below.

C. PREPARATION OF MONOCLONAL ANTIBODIES

Immunization:

Female Balb/C mice, about 8 weeks old, were immunized with about 50 jag per mice of partially purified cancer-mycin preparation in 4 subcutaneous and 1 intraperitoneal injection, first with complete auxiliary compound according to Freund, followed about 4 weeks later by the same type of injections in incomplete auxiliary compound. Approximately 12 days after the second immunization, the mice were bled through the tail vein, and serum antibody titers were determined by ELISA. Three weeks after the second immunization, mice with high serum titers were re-injected to reinforce the acquired immunity by intraperitoneal injections of 10 µg cancer-mycin preparations in PBS for four consecutive days, followed on the fifth day by killing the animals, removal of the spleen and conventional cell fusion with mouse myeloma lines.

Cell fusion, hybridoma growth, cloning:

A suspension of spleen cells was prepared from the spleen

an immunized mouse by careful treatment in phosphate-buffered physiological solution (GKN). 10 g spleen cells were co-cultured with 5 x 10 7 myeloma cells (X63 Ag 8.6.5.3) (from T. Stahelin, Hoffman La Roche, Basel, Switzerland) according to standard procedure. Spleen cells and myeloma cells were mixed in 50% PEG-4000 (GC grade), added slowly over 60 sec. After resting for 90 sec, the cells suspended in PEG were diluted slowly with 7 ml of GKN solution (added over 5 min.) then diluted to approximately 300 ml in RPMI 1640 medium containing 10% fetal calf serum, hypoxanthine, aminopterin and thymidine . The diluted cell suspension was then distributed into 288 wells of 24-well multiplates containing 10^ peritoneal macrophages per well. well. The medium was changed after about 7 days, and then as required in accordance with cell growth.

After screening by EIA, hybridomas secreting antibody to mycins were cloned twice by limiting dilution in microtiter plates containing peritoneal macrophages. For the production of large amounts of monoclonal antibodies, antibody-secreting hybridomas were cultured as fluid-filled tumors in 8-12-week-old mice, which 1-7 days before had been given 0.5 ml of Pristan injected intraperitoneally. Fluid from the above tumors was collected 2-4 weeks after injection of cells and stored at -20°C. Immunoglobulins were purified from the above fluids, as required, by protein A-sepharose affinity chromatography.

Hybridoma screening.

(i) EIA. Wells on polystyrene microtiter plates were filled with untreated cancermycin preparations (10 µg/ml in HCO~/CO~~ buffer pH 9.6) for 3 hours at 37°C. The wells were then washed four times in PBS/Tween. Hybridoma culture supernatants were incubated in the wells for one hour, then washed four times with PBS/Tween. The wells were then incubated with alkaline phosphatase-linked sheep anti-mouse immunoglobulin. After washing with PBS/Tween, followed by distilled water, p-nitrophenyl phosphate was added as substrate and positive wells (indicating supernatants containing antibody to mycins) were identified as those containing the yellow product, p-nitrophenol. (ii) Fluorescent immunohistochemistry. "Positive"

hybridomas identified by initial screening in EIA were then screened using immunofluorescence techniques for the possibility of monoclonal antibodies produced to label specific cells or tissues in the histological sections of normal or cancerous gastrointestinal organs. Supernatants from hybridoma cultures were incubated on tissue culture glass slides. After blocking non-specific binding sites using bovine serum albumin-

solution, the tissues were incubated with fluorescent

conjugated rabbit anti-mouse immunoglobulin. Excess labeled antibody was removed by washing, and the tissue was examined by fluorescent microscopy using U-V or near-band blue excitation.

Using the above techniques, 5 anti-SIMA monoclonal antibodies (which react with normal small intestine mycin and colorectal cancer mycin), and 5 anti-LIMA monoclonal antibodies (which react with normal colon mycin and colorectal cancer mycin) were identified.

Characterization of monoclonal antibodies.

The subset of antibodies was determined using enzyme-linked anti-IgG and anti-IgM antibodies (CSL). All of the monoclonal antibodies were found to be of IgG subgroups.

In order to fully characterize the monoclonal antibodies, their relative affinity was measured using two methods. First, purified IgG was obtained for each monoclonal antibody by protein A-sepharose affinity chromatography. The purified IgG was then titrated by EIA, and titer curves (turnover in EIA against protein concentration) compared as a measure of the relative affinity of the antibodies. Second, a competing EIA was developed (see below). The relative affinity constants were calculated using SIMA preparations as inhibitor together with anti-SIMA monoclonal antibody, and a LIMA preparation as inhibitor for anti-LIMA monoclonal antibody. The calculation of the relative affinity constants was

based on the following equation:

At the concentration of antigen (mycin) that produces a 50% inhibition in a competitive EIA, the concentration of free and bound antibody will be equal, ie (Ab) = (AbAg). Therefore, the relative af f ini tets constant Ka, = -7-r—r .

Ka,(Ag)The results from these calculations are shown in table 3.

As the real antigen concentration can be lower (if the preparations are not 100% pure or other proteins that can contaminate are present), the values for Ka are minimum values and can really be higher.

20

The method according to Friguet et al (1983) was used to determine whether each of the anti-SIMA monoclonals on the one hand and the anti-LIMA monoclonals on the other hand were bound to the same antigenic determinants on the SIMA and LIMA molecules, respectively . Using limiting amounts of antigen to cover the plate, each antibody was titrated individually and with all possible combinations. All the monoclonals produced in the experiments in question were found to bind to the same or overlapping antigenic determinants on the molecule.

EXAMPLE 2 IN VITRO DIAGNOSTIC IMMUNOASSAYS

Two immunoassays that have been developed to detect circulating mycin in blood samples are described below:

A. COMPETITIVE ELISA

(i) The steps involved in this experiment are shown in fig. 5. Initial experiments indicate that all mycin added to heparinized or EDTA-treated or untreated blood can be recovered and quantitated in diluted plasma or serum. Undiluted plasma or serum added to the immunoassay causes interference, while no effect on the experiments (e.g. aqueous buffers) was observed using plasma diluted -<J>q . A number of blood samples from patients with colorectal cancer, patients with ulcerative colitis and healthy controls were therefore analyzed for SIMA and LIMA using competitive ELISA. Plasma from healthy individuals containing added SIMA or LIMA was used as a standard. High levels of SIMA and/or LIMA were detected in 13/23 (about 60%) of patients with varying degrees of colorectal cancer, and high levels of LIMA (but not SIMA) were detected in 4/5 cases of ulcerative colitis. Neither SIMA nor LIMA was detected in blood samples from seven healthy controls in this experiment. (ii) To determine whether mycin could be detected at lower levels,

one planned a procedure for extraction and

concentration of mycin from plasma samples. An equal volume of 2M perchloric acid (PCA) was added to the serum or plasma samples for 15 min. at room temperature. The samples were centrifuged at 8,000 g for 2 min., the supernatants were removed and the pellets were washed in a volume of PBS twice the original sample volume, mixed and centrifuged as above. PBS was added to the combined supernatants to a final volume 10 times greater than the original sample volume. The samples were then concentrated/dialyzed 10 times in a Minicon B15 concentrator (molecular weight cut-off 15,000, Amicon Co.). The samples were then diluted to 10 times the original sample volume and concentrated 50 times. The PCA extract obtained was one fifth of the original sample volume. This sample was then analyzed by competitive ELISA, as described previously. Plasma samples from healthy volunteers with different mycin concentrations added and passed through the above procedure were used as standards. The results for the SIMA analyzes on samples from 39 patients with colorectal cancer, 2 patients with breast cancer, 5 patients with ulcerative colitis, and 7 controls showed that SIMA antigenic activity could be detected at low concentrations in some control patients and at higher concentrations in all with with the exception of 2 of the cancer patients. 60% of the samples from the patients with stage B colorectal cancer (defined as those cancers that have no obvious spread to other organs), 91% from stage C (defined as those cancers with detectable metastases in local lymph nodes) and 86% from stage D (cancer that has spread to distant organs, such as the liver) contained elevated mycin levels compared to controls. SIMA was also detected in a blood sample from a patient with stage A colorectal cancer, and 2 patients with breast cancer.

Analysis with respect to LIMA has only been performed with diluted serum samples, without the concentration/dialysis steps used in the SIMA experiments. Only a small proportion of patients with not far-advanced colorectal cancer had measurable LIMA-11%

in stage B, 29% in stage C; but a greater proportion (67%) of stage D cancer patients had elevated LIMA serum levels. A very large proportion (80%) of patients with ulcerative colitis showed high levels of circulating LIMA, even with this relatively insensitive assay. So this LIMA assay may be of value in the detection of inflammatory disease of the colon.

B. "SANDWICH" IMMUNOASSAYS

The above described monoclonal antibodies can also be used in a "sandwich" immunoassay.

This type of immunoassay has a number of advantages in improving the sensitivity of the analysis, and in addition there are fewer steps than in the competitive inhibitory analysis described above. Furthermore, "signal" in a "sandwich" immunoassay (ie using radioactivity or an enzyme-labeled antibody dye in the well) is directly proportional to the mycin antigen concentration, which greatly simplifies the calculation and treatment of the results compared to a competitive inhibitory assay.

The assay comprises a solid surface coated with monoclonal antibody to immobilize the mycin molecule, and the same monoclonal antibody labeled (with<125>I or an enzyme) to detect the presence of mycin bound to the coated antibody

(see fig. 6) .

a. Details of a SIMA or LIMA radioimmunoassay (RIA)

(Figure 6)

1. Wells of millititer HA plates (Millipore Corporation, USA) are coated with protein A-purified monoclonal antibody IgG (approximately 10 pg/ml) diluted in buffer (sodium carbonate/bicarbonate buffer pH 9, 6, 50 pl per well). The plates are incubated for 3-5 hours at 37°C. 2. To 100 µl of the serum samples or standards, add 200 µl of 0.2M sodium acetate pH 5.0, and heating at

70 - 1°C for 15 min., then centrifugation at

9000 g for 10 min. The standards contained 0, 1.5, 5.0, 15, 50, 150 and 500 µl/ml of SIMA or LIMA prepared in pooled serum obtained from healthy volunteers.

125 3. I-labeled monoclonal antibody diluted in PBS containing 1% BSA is then added to the acetate-extracted supernatant (approximately 0.1 pCi plus 100 µl of supernatant), and incubated at 37°C for 1 hour.

4. Coated wells of millititer plates are then washed

4 times with PBS/Tween and the plates are blocked for 30 min. with 1% BSA in PBS at 37°C. 5. After blocking, wash the plates three times with PBS/Tween. Samples (50 µl per well in two wells) preincubated with <125>I-labeled antibodies and then added to the wells, incubated overnight at 4°C. 6. After overnight incubation, wash the wells 5 times with PBS/Tween, wipe dry using filters,

and the filters at the bottom of each well are equipped and counted in a yarn feed counter.

The SIMA and LIMA levels are expressed as discretionary units/ml, determined relative to the protein content of the reference standard for SIMA and LIMA.

A number of experiments have been carried out to assess the reliability and reproducibility of the "sandwich" assay. Table 4 shows the variation obtained between determinations performed in triplicate in a typical analysis of control serum. For the serum assays of SIMA, the variability is greatest (cr = 10%) at levels of 1.5 p/ml where the assays are not considered reliable (therefore all values less than 5 u/ml SIMA are not given specific values). For serum levels of 5 to 500 U/ml SIMA, the intra-assay variability is 5.6% or less.

b. Details of a LIMA enzyme immunoassay (EIA)

(Figure 7)

1. Polystyrene microtiter plates (Nunc Immunoplates) are coated with anti-LIMA monoclonal antibody IgG (from hybridoma cell line 2C3) in bicarbonate buffer for 3 hours at 37°C. The plates are washed twice with wash buffer (PBS/Tween/sodium azide, pH 7.2-7.6),

and then dried.

2. 50 µl of a blocking solution (bovine serum albumin dissolved in PBS containing sodium azide (pH 7.4) is then added to each well of the plate and incubated at 37°C for 1 hour.

3. For 100 pl LIMA standard solution (standard LIMA

dissolved in normal human serum, containing sodium azide, to give an antigen concentration of

0, 1.5, 5, 15, 50, 150 and 500 units per ml) or serum samples, 200 µl acetate extraction buffer (0.2 M sodium acetate pH 5.0 containing sodium azide) is added, and heating at 70 - 1°C for 15 min. Cooling to room temperature and centrifugation at 10,000 g for 10 min. in an oscillating rotor.

4. Microtiter plates are washed three times with wash buffer and dried. 5. 50 µl supernatant from acetate/heat-treated standards or serum samples is added to each well (two wells per sample). Incubate plates at 37°C for 1 hour. 6. The microtiter plates are washed seven times with wash buffer and then dried. 7. 50 pl per well of alkaline phosphatase-conjugated anti-LIMA rrion ok 1 on a 11 antibody, diluted 1:300 in phosphate-buffered saline containing 1% bovine serum albumin is added. Incubate at 37°C for 1.5 hours. Wash 4 times in a wash buffer, 3 times in distilled water and then dry. Add 100 pl per well of substrate (p-nitrophenyl phosphate), dissolved in a concentration of 2 mg per ml in substrate dilution buffer (0.2 M sodium carbonate/bicarbonate buffer containing 0.1 M magnesium chloride). Incubate at 37°C for 20 min. 8. Add 100 pl per well of a solution that stops the enzyme reaction (0.5 M sodium hydroxide). 9. Read optical density at 405 nm in a suitable microtiter plate paper.

Evaluation of analysis results

(i) Construct a standard curve by plotting the average optical density obtained for each LIMA standard (on the Y-axis) against the corresponding LIMA concentration (X-axis). (ii) Using the average optical density of each serum sample, determine the corresponding concentration of LIMA in units per ml, from the standard curve.

If the sample requires further dilution (ie LIMA concentration > 150 U/ml) to give a value that can be read on the standard curve, the value obtained from the standard curve must be multiplied by the corresponding dilution factor to give LIMA -concentrate the ion in serum.

An example is shown in table 5 and fig. 8 is produced from data in table 5.

An example of the reproducibility of the above LIMA EIA is indicated in Table 6 which shows the variation obtained between determinations in triplicate in an analysis of serum samples containing different amounts of LIMA. The variability is highest for values less than 2.5 U/ml (coefficient of variation 10%) or values greater than 150 U/ml (coefficient of variation 12%). Serum levels outside these limits (ie <2 or >150 U/ml) are therefore considered unreliable. Samples containing > 150 U/ml LIMA should therefore be diluted and measured again.

c. Clinical results

Table 7 below shows an overview of results from samples from patients with various diseases and from healthy controls, and shows the results of SIMA and LIMA RIA (see above) and CEA serum analyzes for the purpose of comparing the number of samples that were only positive for SIMA, LIMA only, CEA only, or any combination of these three markers.

C. OTHER IMMUNOASSAY PROCEDURES

Another variation of the above "two-sided" immunoassay may be to use a polyclonal anti-mycin antiserum (ie, prepared in rabbits or sheep) to coat the solid phase. This would allow an unlabeled monoclonal antibody to be used as the "second" antibody in the assay, followed by detection of the mouse monoclonal with a labeled rabbit or sheep anti-mouse antibody. This system (or the opposite): monoclonal antibody for coating the solid phase and polyclonal antisera for detection) can provide better amplification of positive signals than when the monoclonal antibody is labeled directly. It should also be noted that the present invention encompasses the use of an enzyme substrate containing fluorine such as 4-methylumbelliferyl (4MU)-phosphate (for alkaline phosphatase) or 4MU-3-D-galacside (for 3-galactosidase ) in "sandwich" immunoassays. The same enzyme-linked monoclonal antibody can be used as that in the colorimetric method as described above. The expected increase in sensitivity will be around 10 times. Other known detection systems, such as protein A, avidin-biotin, etc., can also be used. Where labeled protein A is used as a detection system in these "sandwich" immunoassays, the antibody applied to the solid phase (either monoclonal or polyclonal) will be an F(ab) fragment, and the detectable antibody an intact molecule containing Fc segment. A number of different solid phases can be used with these systems in addition to the microtiter wells, such as different types of beads. In addition, membranes can be used as adsorbing surfaces to bind antigens.

LITERATURE REFERENCES

1. Ma, J., De Boer, W., Ward, H.A. 6 Nairn, R.C.

(1980). "Another Oncofoetal Antigen in Colonic Carcinoma" Br. J. Cancer 41 325-328. 2. Pihl, E. , Nairn, R.C, Hughes, E.S.R., Cuthbertson, A.J. & Rollo, A.J. (1981)-. "Mucinous Colorectal Carcinoma: Immunopathology and Prognosis" Pathology 12 439-447. 3. De Boer, W., Ma, J., Rees, J.W. & Nayman, J.

(1981). "Inappropriate mucin production in bile

bladder metaplasia and neoplasia - an immur.ohistological study" Histopa thol ogy 5_ 295-303 .

(1981). 4. De Boer, W., Ma, J. & Nayman, J. (1981). "Intestine-associated antigens in ovarian tumours: An immunohistologic study" Pathology 13 547-555. 5. Ma, J., De Boer, W. & Nayman, J. (19 82). "The presence of oncofoetal antigens in large bowel carcinoma". Aust. S N. Z. Journal of Surgery 5 2 30-34. 6. Ma, J., De Boer, W. & Nayman, J. (1982). "Intestinal mucinous substances in gastric intestinal metaplasia and carcinoma studied by immunofluorescence". Cancer 49 1664-1667. 7. Ma, J., Kandley, C. J. & De Boer, W. (1983). "An ovarian tumor specific mucin antigen - immunohistological and biochemical studies" Pathology 05 385-391. 8. Nayman, J., De Boer, W.6Ma, J. (1983). "Inappropriate nmcin production in Endodermal carcinoma - one point of view". Japanese Journal of Surgery 13 317-323. 9. Gold, P. and Freedman, S.O. (1965). Specific carcino-embryonic antigens of the human digestive system. J. Exped. Med., 122 , 467-481 . 10. Mackay, I.R. (1979). In. Immunodiagnosis of Cancer, Herberman and Mclntire (eds), Marcel Dekker, N.Y. 255. 11. Cuthbertson, A.M., Hughes, E.S.R., Nairn, R.C. Nind, A.P.P. and Pihl, E. (1981). CEA testing: Is it useful in colorectal cancer? With. J. Australia, 170 . 12. Cuthbertson, A.M., Hughes, E.S.R., Nairn, R.C. Nind, A.P.P. and Pihl, E. (1981). CEA testing in colorectal cancer. With. J. Australia, 201. 13. Pihl, E., McNaughtan, J., Ma, J., Ward, H.A. and Nairn, R.C. (1980). Immunohistological patterns of carcinoembryonic antigen in choriorectal carcinoma. Correlation with staging and blood levels.

Pathology,' 12, 7-13.

14. Herlyn, H. , Steplewskri, Z., Herlyn, D., and Koprowski, I. Choirectal carcinoma - specific antigen: Detection by means of monoclonal antibodies. Pro. Nati. Acad. Pollock. USA. 76:1438, 1979.

15. Koprowski, B., Herlyn, H., Steplewski, Z. & Sears, H.F., Specific antigen in serum of patients with colon carcinoma. Science 212:53, 1981. 16. Magnani, J., Brocklaus, M., Smith, D., Ginsburg, V., Blaszcyk, M., Mitchell, D., Steplewsky, Z. & Koprowski, H. A monoclonal antibody defined antigen of colon carcinoma. Science, 212:55, 1981. 17. Devillano, B. and coil. Radioimmunometric Assay for a Monoclonal Antibody defined Tumor Markers CA 19-9™. Clin. Chem. 29 :549-552. 1983. 18. Sears, H., Herlyn, J., Delvillano, B., Steplewski, Z. & Koprowski, H. A clinical evaluation of patients with colorectal cancer. Clinical immunology, 2, : 141, 1982. 19. Magnani, J., Steplewski, Z., Koprowski, H., & Ginsburg, V. Identification of the gastrointestinal and pancreatic cancer-associated antigen detected by monoclonal antibody 19-9 in the sera of patients as a mucin. Cancer Research : 43, 5489-5492, 1983. 20. Friguet, B., Djavadi-Ohaniance, L., Pages, J. , Bussaro, A., & Goldberg, M. A convenient enzyme-linked immunosorbent assay for testing whether monoclonal antibodies recognize the same antigenic site. Application to hybridomas specific for the 62-subunit of E. coli tryptophan synthase. J. Immunol. Methods. : 351-358, 1983. 21. Trerelyan, W.E. and Harrison, J.S.. Studies in yeast metabolism. I. Fractionation and microdetermination of cell carbohydrates. Biochem J. 50: 298-303, 1952.

Claims (14)

1. Procedure for in vitro diagnosis in favor of the presence of cancer cells or other cells that form mycin antigens in a patient, characterized by analyzing a sample of a physiological fluid from the patient to detect the presence of SIMA and/or LIMA in the above sample. 2. Method as stated in claim 1, characterized in that the above-mentioned sample is a sample of blood, blood serum or blood plasma from the patient. 3. Method as stated in claim 1, characterized in that the above-mentioned sample is brought into contact with antibody against SIMA and/or LIMA or fragments thereof, and detection of the presence of the bond between the above-mentioned antibody or fragments thereof and their associated antigens. 4. Method as stated in claim 1, characterized in that the above-mentioned sample is brought into contact with antibody against SIMA and/or LIMA or fragments thereof labeled with a marker capable of providing a detectable signal, and detection of the presence of the bond between above labeled antibody or fragments thereof and their associated antigens. 5. Method as stated in claim 3 or 4, characterized in that the above-mentioned antibody is a monoclonal antibody. 6. Method as stated in claim 3, 4 or 5, characterized in that the above-mentioned antibody or fragments thereof are labeled with an enzyme or a radioactive isotope. 7. Procedure as stated in claim 1, characterized in that: (i) antibody against SIMA and/or LIMA or fragments thereof is bound to a solid support, (ii) the above sample is brought into contact with the above solid carrier, and (iii) the presence of corresponding antigens in the above sample which are bound to the above antibody or fragments thereof bound to the above solid carrier is detected. 8. Method as stated in claim 7, characterized in that detection of the presence of associated bound antigens includes: (iv) contacting the above solid support with antibody to SIMA and/or LIMA, or fragments thereof, labeled with a marker capable of providing a detectable signal, and (v) detection of the binding between the above-mentioned labeled antibody or fragments thereof and the corresponding antigen bound to the above-mentioned solid support. 9. Diagnostic kit or system for in vitro detection of the presence of cancer cells or other cells that form mycin antigens in a patient, characterized in that the above-mentioned kit or system comprises means for detecting the presence of SIMA and/or LIMA in a sample of a physiological fluid from the patient. 10. Kit or system as stated in claim 9, characterized in that the above-mentioned kit or system comprises antibody capable of an immunological reaction with SIMA and/or LIMA, or fragments thereof, together rned indicator/means for detecting the presence of an immunological reaction between the above-mentioned antibody or fragments thereof and their associated antigens. 11. Kit or system as stated in claim 10, characterized in that the above-mentioned kit or system further comprises a carrier of solid substance to which the above-mentioned antibody or fragment is bound. 12. Kit or system as stated in claim 10, characterized in that the above-mentioned indicator means comprise a marker which is capable of providing a detectable signal and where the above-mentioned marker is bound to the above-mentioned antibody or fragments thereof. 13. Kit or system as stated in claim 10, characterized in that the above-mentioned indicator means comprise a marker which is capable of providing a detectable signal, and where the above-mentioned marker is bound to another antibody which is raised against the first-mentioned antibody and which indicates presence of the above-mentioned immunological reaction upon binding to the first-mentioned antibody. 14. Kit or system as stated in claim 12 or 13, characterized in that the above-mentioned marker is an enzyme or a radioactive isotope.