WO1989002599A1

WO1989002599A1 - Detection of survived, undegraded carbohydrate epitopes as diagnostic of intestinal diseases

Info

Publication number: WO1989002599A1
Application number: PCT/US1988/003051
Authority: WO
Inventors: Sen-Itiroh Hakomori; Edward D. Nudelman; Anil K. Singhal
Original assignee: The Biomembrane Institute
Priority date: 1987-09-08
Filing date: 1988-09-06
Publication date: 1989-03-23
Also published as: AU2489888A

Abstract

Methods for detecting intestinal diseases in a patient by the use of a feces sample or a blood-derived sample are disclosed. Monoclonal antibodies with a high degree of affinity for survived, undegraded carbohydrate epitopes in feces or blood-derived samples are disclosed. The detection of the carbohydrate antigens is diagnostic of intestinal diseases.

Description

DETECTION OF SURVIVED, UNDEGRADED CARBOHYDRATE EPITOPES AS DIAGNOSTIC OF INTESTINAL DISEASES

Technical Field

The present invention relates generally to the diagnosis of abnormalities, diseases or disorders of the intestinal tract by the detection of carbohydrate antigens in human feces or blood-derived samples. In particular, monoclonal antibodies with a high degree of affinity for the carbohydrate antigens permit detection.

Background of The Invention Colorectal Cancer

Cancers of the intestinal tract represent a significant percentage of the total cancers in the United States and other nations. Early diagnosis is essential. to prevent radical surgery or loss of life. Diagnosis has focused on the analysis of human blood and feces. As is described in detail below, the use of blood appears to have limitations that the past methods were unable to overcome and the past methods applied to feces yield information of limited value.

In order to detect or treat any form of cancer, it is necessary to be able to distinguish the transformed cells from the normal cells. Comparisons of normal and colonic cancer cells have revealed both that the latter cells possess unique glycolipids and glycoproteins and that they accumulate certain glycolipids and glycoproteins common with normal cells.

Glycolipids of both the neutral class and the ganglioside (i.e., sialic acid-containing) class have been found in cancer cells. Some of these glycolipids are identical in structure to certain of the Lewis (Le) blood group antigens, e.g., Le^a and Le^b, and their chemi cal isomers Le^x and LeY. All these fucosylated structures are greatly accumulated in colonic cancer. Others of these glycolipids are unique to the cancer cells, e.g., sialosyl-Le^x.

Polyps

Colonic adenomas ("polyps") were found in 58% of men and 47% of women in a survey in the United States. The majority of colonic adenocarcinomas arise from adenomas or neoplastic polyps. Several reports suggest that colonic adenomas have neoplastic potential (Muto et al., "The Evolution of Cancer of the Colon and Rectum," Cancer 36: 2251-60, 1975; B. C. Morson and D. W. Day, "Pathology of Adenomas and Cancer of the Large Bowel: Large Bowel Cancer," Churchill Livingstone, Inc., New York, 1981), and especially dysplasia of the tissue can be correlated with premalignant potential (Jass, "Relation Between Metaplastic Polyp and Carcinoma of the Colorectum," Lancet 1: 28-30, 1983).

Previous Attempts to Detect Colorectal Cancer

Monoclonal antibodies (MAbs) have been produced by the hydridoma technique of Kohler and Milstein against a number of the tumor-associated glycolipids and glycoproteins, often before the structure of the antigen has been elucidated. With the discovery of glycolipids and glycoproteins unique to or at elevated levels in the cancer cells, and with the production of MAbs to them, came the hope that cancer would be detectable at an early stage. In the past, blood or serum detection of tumorassociated antigens has proven to be a less than satisfactory method from which to detect colorectal cancer. First, although Le^x antigen was found to accumulate in lung, gastric, and colonic cancer (Yang and Hakomori, J. Biol. Chem. 246: 1192-1200, 1971), the blood level of Le^x antigen in patients with various cancers was found not to be elevated (Kannagi et al., Cancer Res. 46: 2619-26, 1986). Second, screening blood samples with MAbs specific for sialosyl-Le^a antigen (anti-N-19-9 and CSLEA1) and for sialosyl-Le^x antigen (CSLEX1) showed that the sera of patients with colonic cancer tested positive with these MAbs in less than 45% of the cases (Chia et al., Cancer Res. 45: 435-37, 1985). These high levels of false negatives were seen despite the fact that most of the sera were from patients with cancers of Stages III or IV. Accordingly, in the past, blood detection of tumor-associated antigens has not been a reliable method of diagnosis.

A number of reasons may be advanced for the lack of success, prior to the present invention, in detecting colorectal cancer from a blood sample. Blood travels from the gastrointestinal tract via the portal vein to the liver. The blood leaving the liver has a reduced concentration of glycolipids due to trapping. Furthermore, the liver biotransforms other glycolipids not removed from circulation. Alternatively, blood may contain constituents that interfere with some MAb-antigen reactions. While additional explanations may be possible, the fact remains that the detection of markers for colorectal cancer using blood as the sample source has encountered difficulties. Feces have proved to be useful in the clinical diagnosis of various disorders, e.g., fat analysis for the detection of digestive disorders or malabsorption and microscopic examination for parasites and their eggs and larvae. However, analysis of human feces for the clinical diagnosis of gastrointentinal diseases has been very limited because of the lack of systematic biochemical and immunochemical information regarding the precise composition of feces.

In the past, the use of feces to detect an intestinal disorder has been limited to the benzidine reaction for the detection of heme derivatives and metabolites as indicators of intestinal bleeding. Sakamoto et al., in U.S. Patent No. 4,579,827, describe the production of MAbs which are purported to be specific to human gastrointestinal cancers and useful in serological analyses and tissue immunopathology. Sakamoto et al. suggest the use of human "wastes" as a possible sample source.

Thus, there is a need in the art for methods of detecting colonic cancer that avoid the use of bloodderived samples, or overcome the difficulties of using such samples, and that function reliably with specimens from patients at earlier stages of the cancer. The present invention, through the use of feces, fulfills this need and further provides other related advantages. Glycolipids and Glycoproteins in Feces Feces consist of residues of undigested food, epithelial components shed from gastrointestinal mucosa, and secretions therefrom. The composition of the undigested residue from food varies extensively depending on the type of food and state of digestion. In contrast, the chemical composition of the material shed from intestinal mucosa and secretions is relatively constant. However, the majority of shed and secreted material from intestinal mucosa is enzymatically degraded, and only a small number of resistant structures are found in feces. Specific types of carbohydrates bound to either lipids (called "glycosphingolipids") or to proteins (called "glycoproteins") present in feces are those resistant species. Therefore, the composition of glycoproteins and glycolipids in feces may greatly reflect the physiological a/id pathological state of the intestinal epithelia. In fact, glycosphingolipids, which are the major components of intestinal epithelia, have been shown to change dramatically after birth when the intestinal flora becomes established. The glycosphingolipid composition of infant feces, called "meconium," reflects essentially that of the intestinal mucosa (Karlsson and Larson, J. Biol. Chem. 254: 9311, 1979). The glycosphingolipid composi tion of rat feces has been shown to depend greatly on the presence of bacterial flora in the intestine (Gustafsson et al., J. Biol. Chem. 261: 15294, 1986). A number of tumor-associated antigens have been established as glycosphingolipids (for a review see Hakomori, Annu. Rev. Immunol. 2 : 103, 1984). The present invention shows that some of these tumor antigens are resistant to enzymatic degradation by intestinal secretions, or bacterial flora, and are detectable in feces from cancer patients. This may be further extended to analysis of glycoproteins, since the same epitopes found in glycosphingolipids are also carried by glycoproteins. Furthermore, the changes in carbohydrate antigens are not only present in cancer but may also be found in various other intestinal disorders, including polyps (Abe et al., Cancer Res. 46 : 2639, 1986). Therefore, feces analysis focused on carbohydrate epitopes as provided in the present invention may also be useful for detection of various types of intestinal diseases.

Disclosure of the Invention

Briefly stated, the present invention in one aspect provides a method for detecting the presence or absence of intestinal diseases from a human feces sample. Intestinal diseases include any intestinal disorders or abnormalities. The method generally comprises contacting a sample of feces from a patient with an antibody specific for a disease-associated carbohydrate antigen and then detecting the presence or absence of the disease by the presence or absence of an immunocomplex formed between the antibody and the antigen.

In a preferred embodiment of this aspect of the invention, the disease to be detected is colorectal cancer. Particularly preferred MAbs in this regard include SH-1, which is directed against the Le^x antigen and is of the IgG₃ isotype, and SH-2. A cell line, ATCC #HB9499, that produces the SH-1 MAb is disclosed. The invention is also applicable to other intestinal disorders, such as colonic polyps. In this regard, MAbs directed against the LeY antigen, such as AH-6, are preferred. In another embodiment of this aspect of the invention, a method is provided for detecting the presence or absence of intestinal disease from a sample of human feces that has been extracted to concentrate the antigenic glycolipids, glycoproteins, and carbohydrates. This method is necessary when the disease-associated carbohydrate antigen cannot be detected in unextracted feces. This will be the case when the antigen is present in feces at a lower concentration and/or when the MAb has a lower affinity. The extraction step is applicable to natural state feces or in a lyophilized form.

In another embodiment of this aspect of the invention, a method is provided for detecting the presence or absence of intestinal disorders when the diseaseassociated carbohydrate antigen cannot be detected in extracted feces. The concentration of the antigen in the feces and the affinity of the MAb for the antigen will be determinative of whether this method is necessary. The method comprises fractionating the components of the extracted feces. The fractionation step may be enhanced by purification of the glycolipid, glycoprotein, and carbohydrate fractions by chromatography. Examples of the types of chromatography used include ion-exchange and hydrophobic. Either in conjunction with or in place of the purification step, the glycolipid, glycoprotein, and carbohydrate fractions may be separated by thin-layer chromatography, such as high-performance, thin-layer chromatography (HPTLC).

The present invention in another aspect provides a method for detecting the presence or absence of intestinal diseases from a human blood-derived sample. The method generally comprises contacting a blood-derived sample from a patient with an antibody specific for a disease-associated carbohydrate antigen and then detecting the presence or absence of an immunocomplex formed between the antibody and the antigen. A preferred blood-derived sample is serum. In a preferred embodiment of this aspect of the invention, a method is provided which generally comprises contacting a blood-derived sample with an antibody to capture an intestinal disease-associated antigen. The sample is then contacted with a conjugate of a reporter group and an antibody to the intestinal disease-associated antigen, and the presence or absence of an intestinal disease is detected by the presence or absence of an immunocomplex formed between the conjugate and the antigen. Other aspects of the invention will become evident upon reference to the following detailed description and attached drawings.

Brief Description of the Drawings Figure 1 (A, B, C and D). TLC immunostaining of fecal glycolipids with monoclonal antibody SH-1 defining Le^x determinant.

(A) Lane 1, Le^x ceramide pentasaccharide; lane 2, difucosyl Y2 (dimeric Le^x, III³V³Fuc₂nLc₆); lane 3, case 25; lane 4, case 26; lane 5, case 27; lane 6, case 39; lane 7, case 40; lane 8, case 24; lane 9, case 28; lane 10, case 29; lane 11, case 30; lane 12, case 33; lane 13, case 34. Lanes 3-7 and lanes 9 and 13 are glycolipids of feces from Le(a-b-) individuals; lanes 8, 10-12 are glycolipids of feces from Le(a-b+) individuals.

(B) Lanes 1-4, cases 35, 36, 37 and 38, respectively; lanes 5-9, cases 41, 42, 44, 45 and 46, respectively; lane 10, upper neutral glycolipid fraction from tumors; lane 11, Le^x glycolipids from blood group 0 blood cells; lanes 12 and 13, glycolipids of feces from healthy individuals; lanes 14, Le^x ceramide pentasaccharide; lane 15, difucosyl y2 ; lane 16, case 47; lane 17, case 48; lane 18, case 50. All the cases in this panel are glycolipids of feces of Le(a-b+) individuals.

(C) Lane 1, standard mixture of Le^x ceramide pentasaccharide and difucosyl Y2 ; lane 2, case 44; lane 3, case 50; lane 4, case 31; lane 5, case 32; lane 6, case 43; lane 7, case 49; lane 8, case 51; lane 9, case 19; lane 10, case 20; lane 11, case 21; lane 12, upper neutral glycolipids of tumors; lane 13, upper neutral glycolipids of 0 blood cellsr lane 14, case 31; lane 15, case 32; lane 16, case 43; lane 17, case 49; lane 18, case 51. Only cases 44 and 50 are Le(a-b+); all other cases in this panel are Le(a+b-) individuals.

(D) Lanes 1-5, cases 19, 20, 21, 22, and 23, respectively; lanes 6-12, cases 52, 53, 54, 55, 56, 57 and 58, respectively. These are all glycolipids from feces of patients with non-malignant diseases or healthy individuals. Lane 13, upper neutral glycolipids from tumors; lane 14, upper neutral glycolipids from blood group 0 blood cells. Information on all cases is listed in Table I.

Figure 2 (A, B and C) . TLC immunostaining of fecal glycolipids with monoclonal antibody SH-2, defining di- and trifucosylated Le^x determinant.

(A) Lane 1, difucosyl Le^x; lane 2, O upper neutral standard; lanes 3-23, glycolipids from cancer patients with Lewis status of a-b⁺; lane 24, difucosyl Le^x; lane 25, 0 upper neutral standard; lanes 26-33, cancer patients with Lewis status of a-b⁺.

(B) Lanes 1-5 and 7-8, glcolipids from cancer patients with Lewis status of Le^a-b- _; lanes 6 and 9, tumor upper neutrals isolated from cancer tissue.

(C) Lanes 1-6, cancer patients with Le^a+b- status; lane 7, tumor upper neutral standard.

Figure 3 (A, B and C) . TLC immunostaining of fecal glycolipids with monoclonal antibody SH-2.

(A) Lane 1, difucosylated Le^x; lane 2, 0 upper neutral standard; lanes 3-27, control feces glycolipids from individuals with Le^a-b+; lane 28, tumor upper neutral standard.

(B) Lane 1, difucosylated Le^x; lane 2, 0 upper neutral; lanes 3-10, normal individual with Le^a-b- status.

(C) Lanes 1-4, normal individuals with Le^a+b- status.

Figure 4 (A, B and C). TLC immunostaining of fecal glycolipids with monoclonal antibody 1B2. (A) Lanes 1 and 23, 0 upper neutral; lanes

2-22 and 24-31, patients with colon cancer with Le^a-b+ status.

(B) Lanes 6 and 9, 0 upper neutral; lanes 1-5 and 7-8, colon cancer patients with Le^a-b- status. (C) Lanes 1-6, colon cancer patients with

Le^a+b- status; lane 7, O upper neutral.

Figure 5 (A, B and C). TLC immunostaining of fecal glycolipids with monoclonal antibody 1B2.

(A) Lanes 1 and 27, 0 upper neutral; lanes 2-26, normal individuals with Le^a-b+ status.

(B) Lane 1, 0 upper neutral; lanes 2-9, normal individuals with Le^a-b- status.

(C) Lanes 1-4, normal individuals with Le^{a +b-} status. Figure 6 (A, B and C). TLC immunostaining of fecal glycolipids with monoclonal antibody AH6.

(A) Lane 1, trifucosyl LeY; lanes 2-31, colon cancer patients with Le^a-b+ status.

(B) Lanes 5 and 6, tumor upper neutral standard; lanes 1-5 and 8-9, colon cancer patients with

Le^a-b- status.

(C) Lanes 1-6, colon cancer patients with Le^a+b- status; lane 7, tumor upper neutral.

Figure 7 (A, B and C). TLC immunostaining of fecal glycolipids with monoclonal antibody AH6. (A) Lane 1, trifucosyl LeY; lanes 2-26, control individuals with Le^a-b- status; lane 27, tumor upper neutral standard.

(B) Lane 1, trifucosyl LeY,' lanes 2-29, control individuals with Le^a-b- status.

(C) Lanes 1-4, control individuals with Le^a+b- status.

Figure 8. Solid phase assay with SH-1 and fecal glycolipids from (a) normal patients and (b) cancer patients.

Figure 9. Solid phase assay with SH-2 and fecal glycolipids from (a) normal patients and (b) cancer patients.

Figure 10. Direct detection of Le^x antigen in feces of normal and cancer patients by a combination of monoclonal antibodies SH-1 and SH-2.

Figure 11. Comparative levels of Le^x antigen in sera of normal and cancer patients.

Figure 12. Levels of Le^x antigen in sera at different stages of colon cancer.

TABLE I Analysis o f feces glycolipids from cases of colorectal cancer and controls.

Expressiont in feces

Blood Group detected by

Case

No. Init. Age Sex Site Lewis ABO SH1 SH2 1B2 AH6

1 RM 48 M D a-b+ O +++

2 YS 74 F R a+b- O ++

3 YA 55 F T nd B ++

4 RF 77 F S nd O -

5 OK 64 M S nd B -

6 KE 57 M R nd A -

7 TS 76 F R nd B -

8 KK 69 M R nd AB -

9 SK 70 F S nd B -

10 JS 56 M T nd O +

11 TS 80 F S nd A -

12 KA 85 F R nd AB -

13 TK 56 M R nd AB -

14 IM 73 M R nd B -

15 ET 74 M T nd B +++

16 HK 78 M R a-b- A ++

17 KY 75 M R nd A -

18 HS 53 M S a-b- A +++

24 YA 86 M C a-b+ O - +/- + +

25 RM 55 M R a-b- O +/- + +/- +

26 TN 48 F R a-b- A + + +/- +

27 IT 62 M R a-b- AB - - + +++

28 TE 67 F R a-b+ A + ++ + +

29 YO 59 F S a-b+ O +++ ++ + ++ ++

30 YN 76 F A a-b+ A - +/- - -

31 KG 55 M R a+b- A + + + -

32 JO 77 M S a+b- A +/- - - -

33 JT 44 M R a-b+ 0 +++ +++ +++ +++

34 TK 75 F A a-b+ A +/- + +/- +

35 KS 60 M R a-b+ A + ++ +/- ++

36 FT 71 F R a-b+ B - - - -

37 IY 76 F D a-b+ A + - - -

38 IC 58 M A a-b+ B - +/- - +/-

39 AT 68 M R a-b- B + ++ + -

40 KK 77 F R a-b- A +++ +++ +++ +++

41 KS 77 M R a-b+ O - + +/- - Expression in feces

Blood Group detected by

Case

No. Init. Age Sex Site Lewis ABO SHI SH2 1B2 AH6

42 SY 78 M S a-b+ O +++ +++ +/- +++

43 SS 70 M S a+b- AB ++ + + -

44 RO 67 F R a-b+ B ++ ++ - -

45 CS 65 F A a-b+ O - + - +++

46 IF 70 F A a-b+ B - - + -

47 SM 80 M R a-b+ O - + - ++

48 RI 85 M S a-b+ A +++ +++ + +++

49 KS 75 M S a+b- B + - - -

50 SS 68 M T a-b+ AB + + ++ +

51 KS 33 M T a+b- O +/- - +/- -

89 CF 56 F A a-b+ O ++ - +

90 MH 45 F S a-b+ B - + -

91 SE 58 M S a-b+ A ++ + +

92 MS 43 M R a-b- AB + - -

93 KT 68 F R a-b+ A + ++ -

94 TS 67 M S a-b+ A - + +

95 MI 76 F S a-b+ A + ++ +

96 KC 84 M R a-b+ O + - +

97 IS 40 F R a-b- B +/- + -

98 KS 73 M R nd A +++ + +

99 KG 72 M R a-b+ A - - -

100 SH 72 M T a-b+ B - - -

101 SG 44 F A a+b- A

102 HK 66 M S a-b+ O - - - ft_jBSTITUT <___• I Control Cases

Expression in feces

Blood Group detected by

Case

No. Init. Age Sex Site Lewis ABO SH1 SH2 1B2 AH6

19 MM 67 F gallbladder nd - stone 20 RI 34 M healthy nd A -

21 KU 65 F gallbladder nd A - stone

22 TA 71 M gastric nd AB - ulcer

23 TK 62 M diabetes nd B - mellitus 52 ES 45 M pancreaa-b+ B - - - - titus

53 SI 78 F gallbladder a-b+ O - - + ++ stone

54 NT 26 M healthy a-b+ AB - - - -

55 HM 29 M healthy a-b+ O - - - ++

56 YN 24 M healthy a-b+ O - - + ++

57 SI 28 M healthy a-b+ O - - - - 58 MH 31 F total a-b+ O - - ++ ++ colectomy

59 SK 43 F diabetes a-b+ B + - - mellitus

60 KT 35 M gallbladder a-b- AB - - - stone

61 YH 19 M Crohn's a-b+ B - +/- +/- disease

6 -2- TS 6-5- _M gallbladder a-b+ A - - - stone

63 HA 71 M liver a-b- O +/- +/- + cirrhosis

64 RK 65 M myasthenia a-b+ B + - - gravis

65 YT 19 M chronic a-b+ B - - + hepatitis

66 FY 34 F myasthenia a-b+ 0 - - + gravis Expression in feces Blood Group detected by

Case

No. Init. Age Sex Site Lewis ABO SH1 SH2 1B2 AH6

67 RF 61 F diabetes a-b+ O - - + mellitus

68 RS 60 F acute a+b+ A +/- ++ ++ pancreatitus

69 FH 48 M bile duct a+b- A + + - ssttoonnee

70 YK 43 M ulcerative a-b+ B + +++ +++ colitis

71 FM 60 F diabetes a-b+ A + - ++ mellitus

72 AU 47 F acute a-b- A - - ++ hepatitis

73 KK 61 F diabetes a-b- O - - + mellitus

74 TM 23 M healthy a-b+ B + - +

75 JS 24 M healthy a+b- O - - -

76 TA 24 M healthy a-b- A - +/- +/-

77 MK 28 M healthy a-b+ O - - -

78 KS 26 M healthy a-b+ O +/- - ++

79 TT 27 M healthy a-b+ B - + +

80 YS 24 M healthy a+b- A - + -

81 TT 24 M healthy a+b- AB - - -

82 YT 23 M healthy a-b- O - +/- --

83 SH 24 M healthy a-b+ B +/- - ++

84 HS 25 M healthy a-b+ O - - ++

85 TS 25 M healthy a-b+ AB - - +

86 YN 27 M healthy a-b+ B - - -

87 HN 24 M healthy a-b+ O - - +

88 HA 31 M healthy a-b- A - - -

A, ascending colon; D, descending colon; R, rectum;

T, transverse colon; S, sigmoid colon; nd, not determined.

ITUTE SH££T Best Mode for Carrying Out the Invention

Although the following discussion pertains primarily to two intestinal diseases, viz., colorectal cancer and colonic polyps, it will be appreciated that there are numerous other intestinal diseases which possess disease-associated carbohydrate antigens capable of being detected in human feces, and blood-derived samples, by the methods disclosed in the present invention.

Detection in Feces

Glycolipids and glycoproteins are continuously released into feces. In newborn infants, the glycolipid and glycoprotein composition of feces reflects almost exactly the composition of these structures in gastrointestinal epithelia. After birth, when bacterial intestinal flora are introduced, the glycolipid and glycoprotein compositions begin to change. In adults, the composition of feces is greatly modified, e.g., glycolipids and glycoproteins are degraded, by bacterial enzymes. The present invention involves the determination that tumor-associated carbohydrate antigens survive undegraded in human feces, and blood-derived samples, and the development of methods linking their detection to the presence of intestinal disorders. Examples of tumorassociated carbohydrate antigens include Le^x-type, di-Le^x-type, sialylated Le^x-type, LeY-type, T, Tn, and sialyl-Tn-type antigens. The presence of. these antigens was detected using MAbs. Because the MAbs are against the carbohydrate portion of tumor-associated glycolipids, the carbohydrate antigens detected in the feces, and blood-derived samples, may be derived from any tumor-associated glycolipid or glycoprotein containing the appropriate carbohydrate determinant. Furthermore, the cleavage of the carbohydrate portion from a glycolipid or glycoprotein, such as a mucin-type, and subsequent degradation of the lipid or protein portion appears not to affect the recognition by the MAbs of the carbohydrate determinant. The survival of undegraded carbohydrate antigens in feces may result from an absence of the necessary glycosidases.

In one aspect of the present invention, the detection of intestinal diseases can be performed by one or more of the following methods involving a sample of feces. All three methods have in common the use of an antibody against an intestinal disease-associated carbohydrate antigen to detect the presence or absence of the intestinal disease. The methods differ by the amount of manipulation that a sample of feces undergoes prior to reaction with an antibody.

In the first method, a sample of feces is reacted without extraction. For example, a sample of feces may be reacted with a microtiter plate, or another type of solid phase media, coated with an antibody against an intestinal disease-associated carbohydrate antigen. In the second method, a sample of feces is extracted prior to reaction with an antibody. An example of an extraction solution is isopropanol-hexane-water. The feces may be extracted in its natural state or after lyophilization.

In the third method, a sample of feces is both extracted and fractionated. Fractionation may be accomplished, for example, by dissolving an extracted sample of feces in chloroform-methanol and partitioning by the addition of water. Detection of intestinal disease-associated carbohydrate antigens in the fractionated, extracted sample of feces may be performed by a variety of ways. For example, a microtiter plate may be coated with the sample, an antibody to an intestinal disease- associated carbohydrate antigen added, and the presence or absence of an immunocomplex detected. If necessary, additional steps, such as purification of glycolipid and glycoprotein fractions by chromatography or separation of the fractions, may be used. Ion-exchange and hydrophobic are typical types of chromatography procedures. High- performance, thin-layer chromatography is a typical separ ation procedure and carbohydrate antigens may be detected immunologically without removing from the TLC plate.

In general selection of a particular method disclosed will be dependent upon the concentration of the carbohydrate antigen in the feces and the affinity of the MAb for the antigen. As the concentration of the carbohydrate antigen and/or the affinity of the MAb increases, the less manipulation of the feces sample will be required. Detection of an immunocomplex formed between a

MAb and a disease-associated carbohydrate antigen may be accomplished by a variety of techniques. For example, a reporter group may be attached to a primary MAb or to an antibody against the primary antibody (secondary Ab, e.g., rabbit anti-mouse), or to molecules such as protein A that detect immunocomplexes. Reporter groups which may be used include radioisotopes, fluorophores, enzymes, luminescers, and dye particles.

The present invention provides a MAb, SH-1, of the isotype IgG3 that is unique in its ability to bind to a smaller portion of the carbohydrate chain of Le^x. SH-1 recognizes the terminal five carbohydrate residues of Le^x, penta-Le^x. Penta-Le^x has the following formula:

Gal βl - - > 4GlcNac βl - - > 3Gal βl - - > 4Glc - ceramide

3

I Fucal

It is possible that less than five residues is the minimum necessary for recognition. In contrast, the MAbs to Le^x devloped by others require at least six carbohydrate residues. SH-1 is produced by immunization of mice with purified glycoplipid antigen coated on Salmonella minnesota, the method used previously by Young et al. (J. Exp. Med. 150: 1008-19, 1979). By immunostaining with SH-1, a positive correlation was found between the presence of Le^x antigen in feces of patients with colorectal cancer and its absence in feces of normal subjects and those with non-malignant diseases (Figure 1). Since Le^x is known to accumulate in colonic cancer cells, presumably its presence in feces represents a shedding of the antigen, or cellular debris containing the antigen, into the intestinal lumen.

Other MAbs used in the present invention include: SH-2 for dimeric Le^x, FH-6 for sialylated Le^x (Fukushi et al., J. Biol. Chem. 259: 10511-17, 1983), and AH-6 for LeY (Abe et al., J. Biol. Chem. 258: 11793-97, 1983). The MAb 1B2 reacts with the terminal Galβl - -> 4GlcNac of polylactosamine structures. Although immunostaining of fractionated, extracted feces samples with SH-1 yielded the clearest results (Figure 1), a difference between feces from colorectal cancer patients and normal subjects was observed as well when the immunostaining was performed with SH-2 (Figures 2 and 3 ), 1B2 (Figures 4 and 5), AH-6 (Figures 6 and 7), FH-6, NBH-2, HH-8, TKH-1, TKH-2 and Cu-1. The antigen structures defined by these antibodies are shown in Table II.

The differences in immunostaining patterns between feces samples from normal and colorectal cancer cases by immunoblotting with any of the above-mentioned MAbs were observed with colorectal cancer patients of stages II-IV. Further, the incidence of positive staining in feces of colorectal cancer was found to occur irrespective of the location of the tumor (ascending, transverse, descending colon, sigmoidal color or rectum) or the blood group ABO status of the host. Positive cases were more frequent, however, in Le^a-b- individuals as compared with Le^a-b+ or Le^a+b- individuals.

Detection in Blood-Derived Sample

In another aspect of the present invention, a method is provided for the detection of intestinal diseases from a blood-derived sample. Such a sample is contacted with an antibody against an intestinal disease-associated carbohydrate antigen and the presence of an immunocomplex formed between the antibody and the antigen. The term "blood-derived sample" includes whole blood, plasma and the preferred form, serum. Examples of intestinal disease-associated carbohydrate antigens include Le^x-type, di-Le^x-type, sialylated Le^x-type, LeY-type, T, Tn and sialyl-Tn-type antigens. Suitable antibodies include SH-1, SH-2, 1B2, AH-6, FH-6, NBH-2, Cu-1, HH-8, TKH-1 and TKH-2.

In a preferred embodiment of this aspect of the invention, a blood-derived sample is first contacted with an antibody to capture an intestinal disease-associated antigen. The MAbs SH-1 and SH-2 are preferred. The presence or absence of the intestinal disease is detected by the presence or absence of the "captured" antigen. One way to detect the captured antigen is, following the first step, to contact the sample with a conjugate of a reporter group and an antibody to the intestinal disease-associated antigen. The MAbs SH-1 and SH-2 are preferred. To summarize the examples which follow, Example I describes the extraction and preparation of glycolipids from feces. Example II describes glycolipid analysis. Example III provides the preparation of monoclonal antibody SH-1. Example IV describes immunostaining of glycolipids on TLC plates. Example V provides a solid phase assay of fecal glycolipids. Example VI describes the direct detection of Le^x antigen in feces. Example VII provides a serum assay for the detection of Le^x antigen.

The following examples are offered by way of illustration and not by way of limitation.

EXAMPLE I Extraction and Preparation of Glycolipids from Feces

Feces were lyophilized and then powdered by applying pressure. A 10 to 15 ml volume of feces powder was placed in a 125 ml bottle, mixed with 100 ml of isopropanol-hexane-water (55:25:20), and sonicated for 30 min with frequent vigorous shaking. Each sample was filtered over a Buchner funnel with Whatmann No. 1 filter paper, and the residues were saved. The extracts in the filtrates were evaporated to dryness in a rotary evaporator, dissolved in chloroform-methanol (2:1 v/v), and transferred to a screw-cap tube (1.5 x 12 cm) and supplemented with chloroform-methanol (2:1) up to a volume of 1.2 ml. Folch partition (Folch et al., J. Biol. Chem. 191: 819-31, 1951) was performed by adding 2.0 ml of water. The upper phase was separated by centrifugation, and the lower phase was repeatedly partitioned by addition of chloroform-methanol-water (1:10:10).

The three upper phases were pooled, evaporated to dryness, and the residue was dissolved in 5 ml of water and dialyzed for three days in a Spectropor dialysis tube 3000 (Spectrum Medical Industries, Los Angeles, Calif). The dialysate was evaporated with ethanol to dryness. The dialyzed upper phases were placed over a small DEAE-Sephadex A-25 column with 2 ml column volumes, and the neutral glycolipids were eluted with chloroform-methanol-water (30:60:8) followed by elution with the same solvent containing 0.05 M ammonium acetate to elute monosialogangliosides. The column was washed with the same solvent containing 0.45 M ammonium acetate to elute di- and trisialogangliosides. The procedure is essentially the same as described by Yu and Ledeen (J. Lipid Res. 13: 680-86, 1972). The lower phases were evaporated with ethanol, and each sample was placed on a 2 ml column volume of Florisil (Floridin Chemical Co., Tallahassee, Fla.) in dichloroethane. Glycolipids were eluted with dichloroethane-acetone (1:1), which were deacetylated in sodium methoxide in chloroform-methanol and neutralized with Dowex 50 (H+) (Saito and Hakomori, J. Lipid Res. 12: 257-59, 1971).

The upper neutral glycolipids eluted from DEAE-Sephadex were purified on a C18 Bond-Elut column (Analytichem International, Harbor City, Calif.), as described by Kundu and Suzuki (J. Chromatogr. 224: 249-56, 1981). The ganglioside fraction was dialyzed and purified on a Bond-Elut column as described above.

EXAMPLE II

Glycolipid Analysis The three purified fractions (the neutral glycolipids from the lower phase, the neutral glycolipids from the upper phase, and the gangliosides) prepared from feces according to Example I were subjected to the sphingosine analysis, as described by Naoi et al. (Anal. Biochem. 58: 571-577, 1974).

The samples conta ining the same chemical quantity of glycolipids as expressed by sphingosine content were analyzed by high-performance, thin-layer chromatography (HPTLC) (The Baker Chemical Co., Phillipsburg, N.J.). The neutral glycolipids from the lower phase were separated in the solvent chloroform-methanol-water (100:90:6), and the upper neutral glycolipids were separated in chloroform-methanol-water (50:40:10). The ganglioside fraction was separated in chloroform-methanol-water (50:40:10) containing 0.1% CaCl₂.

The orcinol-sulfuric acid staining patterns of the glycolipids separated by HPTLC was performed by spraying the plates with a fine mist from a 0.5% orcinol/ 10% H₂SO₄. solution. The plates were then heated at 120°C for 1 to 3 min.

EXAMPLE III

Preparation of Monoclonal Antibody SH-1

Directed to Le^x Antigen

1. Purification of Antigen:

Lewis^x antigen (III³ Fuc nLc⁴) was purified from a human liver tumor metastasized from colonic cancer. The tumor was homogenized three times in isopropanol: hexane:H₂O and filtered. The organic extract was evaporated to dryness and partitioned three times by the method of Folch. The combined upper phase was evaporated and dialyzed against distilled water in a spectrapore 3 (3500 mol. wt. cut-off) dialysis tubing. The dialysate was evaporated using excess ethanol and subjected to ion-exchange chromatography on DEAE-Sephadex A-25 column (4 x 50 cm). The sample was applied in chloroform:methanol:water (30:60:8). The non-binding pass-through fraction was collected which contains Lewis^x antigen. The sample was evaporated and subjected to highpressure liquid chromatography (HPLC) on a iotrobead system consisting of isopropanol:hexane:water (55:40:5) to 55:25:20.

The fractions were pooled on the basis of staining with orcinol-sulfuric acid. The glycolipids migrating between standard H₁ and H₂ glycolipids were pooled and acetylated using pyridine-acetic anhydride. The acetylated glycolipids were applied to a preparative TLC plate and developed in dichloroethanol: acetone: water (40:60:0.03). Each isolated band was analyzed by NMR and permethylation analysis, and Le^x band was collected.

2. Immunization Protocol:

Balb/c mice (female, 8-weeks-old) were immunized with the purified Lewis^x antigen (III³ Fuc nLc4). The antigen (40 μg/100 μl ethanol) was injected into 800 μl phosphate-buffered saline (PBS) at 37°C. The solution was further mixed with 250 μg of acid-treated S. minnesota (1 mg/ml PBS). The mixture was incubated at 37°C for 30 min and lyophilized. The lyophilized powder was resuspended in 1 ml PBS.

Mice were immunized every 10 to 14 days apart by tail vein injection of 250 μl of antigen suspension of PBS.

3. Fusion and Screening Protocol:

Three days after the last injection, animals were killed by cervical dislocation and spleens were aseptically excised. Lymphocytes were fused with mouse mye^¬ loma SP² cells (5:1) using polyethylene glycol. Clones were screened after 11 days of fusion. The clones at this stage were small, and it was hoped that only high-affinity antibodies would test positive at that stage. Clones were screened using a Pandex machine, in which antigen is coated on submicron polystyrene particles. Antigen-coated beads are mixed with the antibody supernatents, followed by the addition of FITC-goat anti-mouse IgG and IgM. This assay is 2- to 4-fold more sensitive than traditional radioimmunoassay or ELISA assay and requires only 10 ng of antigen/well as opposed to 50 to 100 ng/well in other assays. This screening procedure further facilitated the selection of a high-affinity antibody to Lewis^x. The clones that tested positive were cloned by single-cell dilution and were also tested by TLC immunostaining. The clone 8G3 producing MAb SH-1, showing high reactivity in the Pandex assay and in TLC immunostaining, was selected and frozen.

EXAMPLE IV Immunostaining of Glycolipids on TLC Plate

The three purified glycolipid fractions were each separated by HPTLC according to Example II. The TLC plates were immmersed for 1 min in a solution of 0.5% polyisobutyl methyl acrylate in ether. The TLC plate was dried and soaked for 2 hours in 5% bovine serum albumin (BSA) in P^/saline to avoid nonspecific adsorption of antibody. After washing with P_i/saline, the plate was incubated overnight with 1 μg/ml of one of the antibodies described below in P_i/saline containing 1% BSA. The incubation was followed by sequential incubations with 1 μg/ml rabbit anti-murine IgM antibody solution and with [¹²⁵I]-protein A solution. After extensive washings with Pi/saline, the TLC plates were subjected to autoradiography using Kodak x-ray films. The neutral glycolipid fraction from the upper and lower phases, separated on TLC plates, was blotted by the following monoclonal antibodies: SH-1 for Le^x (SH-1 is an IgG₃), SH-2 for dimeric Le^x, FH-1 for Le^x (Itzkowitz et al., Cancer Res. 46: 2627-32, 1986), FH-4 for dimeric Le^x (Fukushi et al.), AH-6 for LeY (Abe et al. J. Biol. Chem. 258: 11793-97, 1983), and 1B2 for unsubstituted type 2 chain (Young et al., J. Biol. Chem. 256: 10967-72, 1981).

The ganglioside fraction, separated on TLC, was blotted with the following monoclonal antibodies: CSLEA for sialyl Le^a (Chia et al., Cancer Res. 45: 435-37, 1985), FH-7 for disialyl Le^a (Nudelman et al., J. Biol. Chem. 261: 5487-95, 1986), and IB9 for sialyl 2->6Gal (Hakomori et al., J. Biol. Chem. 259: 4672-80, 1984). EXAMPLE V Solid Phase Assay of Fecal Glycolipids Glycolipids isolated from feces of either normal or colon cancer patients were subjected to a solid phase assay. In this assay, the microtiter plates were coated with 100 ug/well of glycolipid in ethanol and the ethanol was evaporated to dryness. Plates were then blocked with 3% bovine serum albumin (BSA) for 1 hour. This was followed by incubation with culture supernatant of monoclonal antibody SH-1 or SH-2 for 2 hours. The plates were then washed and incubated with peroxidase conjugated goat anti-mouse IgG for 1 hour. After washing, plates were developed using the peroxidase substrate O-phenylenediamin (OPD). The color was read at O.D. 490 in an automated ELISA reader.

Glycolipids from feces of both normal and cancer patients were compared (Figures 8 and 9) - - 8 out of 10 cancer patients contained significantly higher levels of long chain Lewis^x glycolipids (recognized by SH-2), whereas only 1 out of 10 normal samples contained high levels of these glycolipids. Eight out of ten cancer patients also showed significantly high reactivity to monoclonal antibody SH-2, whereas only 2 out of 10 normal samples contained high levels.

EXAMPLE VI

Direct Detection of Le^x Antigen in Feces

Microtiter plates were coated with MAb SH-2 overnight and blocked with 3% BSA. Lyophilized feces (5 mg) were suspended in phosphate buffered saline (PBS). Samples were vortexed, sonicated, and centrifuged. The PBS-washed feces were added to the microtiter plate coated with SH-2. After 2 hours incubation, the plates were washed and incubated with peroxidase conjugated SH-1 for 1½ hours. The plates were washed and developed using the peroxidase substrate O-phenylenediamine (OPD). The color was read at O.D. 490 using an automated ELISA reader.

By this direct binding assay, 40% of feces from cancer patients contained higher levels of Le^x antigen, whereas 13% of the normals were positive (Figure 10).

The cut-off for the normal range is an O.D. 490 of 0.2 which represents normal ± standard deviations.

EXAMPLE VII Detection of Le^x Antigen in Human Sera

Sera from normal and cancer patients were assayed by a combination of monoclonal antibody SH-1 and SH-2 (Figures 11 and 12). MAb SH-2 was coated at the bottom of a microtiter plate overnight. The plate was blocked with 3% BSA. Human sera were diluted in 1% BSA containing 0.2% mouse sera. The sera was then added to the SH-2 coated plate and incubated for 2 hours. After washing the plate with PBS, peroxidase conjugated SH-1 was then added and incubated for li hours. The plates were then washed and the color developed with OPD. The color was read at O.D. 490 using an ELISA reader.

From the foregoing, it will be appreciated that, although specific embodiments of the invention have been described herein for purposes of illustration, various modifications may be made without deviating from the spirit and scope of the invention.

Claims

1. A method for the detection of intestinal diseases in a patient, comprising the steps of: contacting a sample of feces from said patient with an antibody against an intestinal disease-associated carbohydrate antigen; and detecting the presence or absence of said intestinal disease by the presence or absence of an immunocomplex formed between said antibody and said antigen.

2. The method of claim 1, additionally comprising, before the step of contacting, the step of extracting a sample of feces of said patient to yield an extracted sample of feces.

3. The method of claim 2 wherein the step of extracting further comprises extracting said feces with iso-propanol-hexane-water.

4. The method of claim 2, additionally comprising, after the step of extracting, the step of fractionating said extracted sample of feces of said patient to yield a fractionated, extracted sample of feces.

5. The method of claim 4, additionally comprising, after the step of fractionating, the step of purifying glycolipids and glycoproteins in the fractionated, extracted sample of feces by chromatography.

6. The method of claim 5, additionally comprising, after the step of purifying, the step of separating said purified glycolipids by high-performance, thin-layer chromatography.

7. The method of claim 1 wherein the carbohydrate antigen is selected from the group consisting of an Le^x-tyρe, di-Le^x-type, sialylated Le^x-type, LeY-type, T, Tn and sialyl-Tn-type.

8. The method of claim 1 wherein the antibody is selected from the group consisting of SH-1, SH-2, 1B2, AH-6, FH-6, NBH-2, Cu-1, HH-8, TKH-1 and TKH-2.

9. The method of claim 1 wherein the step of detecting further comprises determining the presence or absence of a reporter group bound to said antibody, to an antibody against said antibody, or to a molecule that reacts with said immunocomplex.

10. A method for the detection of intestinal diseases in a patient, comprising the steps of: contacting a blood-derived sample from said patient with an antibody against an intestinal disease-associated carbohydrate antigen wherein the carbohydrate antigen is selected from the group consisting of an Le^x-type, di-Le^x-type, sialylated Le^x-type, LeY-type, T, Tn and sialyl-Tn-type; and detecting the presence or absence of said intestinal disease by the presence or absence of an immunocomplex formed between said antibody and said antigen.

11. The method of claim 10 wherein the antibody is selected from the group consisting of SH-1, SH-2, 1B2, AH-6, FH-6, NBH-2, Cu-1, HH-8, TKH-1 and TKH-2.

12. A method for the detection of intestinal diseases in a patient, comprising the steps of: contacting a blood-derived sample from said patient with an antibody SH-2 against an intestinal diseaseassociated carbohydrate antigen; contacting said sample with a conjugate of antibody SH-1 and a reporter group; and detecting the presence or absence of said intestinal disease by the presence or absence of an immunocomplex formed between said conjugate and said antigen.

13. The method of claim 12 wherein the carbohydrate antigen is selected from the group consisting of an Le^x-type, di-Le^x-type, sialylated Le^x-type, LeY-type, T, Tn and sialyl- Tn-type.

14. The method of claim 12 wherein the blood-derived sample is serum.

15. The cell line HB 9499.

16. A monoclonal antibody produced by the cell line of claim 15.

17. A monoclonal antibody that competitively inhibits the formation of an immunocomplex between the antibody of claim 21 and penta-Le^x.

18. A monoclonal antibody of the isotype IgG₃ produced by a hybridoma formed by the fusion of cells from a myeloma line and spleen cells from a mouse previously immunized with an Le^x.

19. A monoclonal antibody that specifically binds to penta-Le^x.

20. The monoclonal antibody of any of claims 15-19, for use as an active therapeutic substance.