WO1992022817A1

WO1992022817A1 - Method to diagnose systemic mast cell disorders

Info

Publication number: WO1992022817A1
Application number: PCT/US1992/002368
Authority: WO
Inventors: L. Jackson Roberts; John A. Oates
Original assignee: Vanderbilt University
Current assignee: Vanderbilt University
Priority date: 1991-06-14
Filing date: 1992-03-25
Publication date: 1992-12-23
Anticipated expiration: 1993-12-14

Abstract

The present invention relates to an improved method to diagnose systemic mast cell disorders by detecting the presence or concentration of certain metabolites of prostaglandin D2. In particular, this invention provides a more accurate means to diagnose systemic mast cell disorders than the present standard test involving the detection of Me-histamine.

Description

METHOD TO DIAGNOSE SYSTEMIC MAST CELL DISORDERS

Background of the Invention

Field of the Invention The present invention relates to an improved method to diagnose systemic mast cell disorders by detecting the presence or concentration of certain metabolites of prostaglandin D₂. In particular, this invention relates to a more accurate method to diagnose systemic mast cell disorders than the conventional methods. This invention also relates to new compositions of matter comprising substantially pure metabolites of prostaglandin D₂.

Description of the Prior Art The biochemical diagnosis of systemic mast cell disorders involves demonstration of increased release of mast cell secretory products, which include hista ine, prostaglandin D₂, tryptase, and heparin.

In patients with systemic mastocytosis, it is frequently possible to demonstrate chronic overproduction of histamine and prostaglandin D₂ , which can probably be attributed to the increased mast cell burden in the body. During episodes in which mast cells are activated, mediator release in such patients increases further, usually be several fold. In contrast, in patients with no evidence of abnormal mast cell proliferation, mast cell mediators are usually normal between episodes of mast cell activation. Although a finding of normal levels of mast cell mediators at a time when a patient is asymptomatic, may suggest that the patient does not have mastocytosis, such a finding does not exclude the possibility that the patient has a disorder of systemic mast cell activation. Episodes of systemic mast cell activation experienced by patients both with or without abnormal mast cell proliferation are characteristically short¬ lived, usually lasting 15 to 60 minutes. Obviously, therefore, increased levels of mast cell mediators can be detected in biologic fluids only for similar brief periods of time. Therefore, it is important to obtain blood while a patient is actually experiencing and episode suspected to represent mastocyte activation and to collect fractional urines voided over short periods of time immediately after an episode.

The mediators responsible for the vasodilation that occurs during episodes of systemic mast cell activation are histamine and prostaglandin D₂. Roberts, L.J. II, Sweetman, B.J., Lewis, R.A. , Austen, K.F., Oates, J.A. : Increased Production of Prostaglandin D₂ in Patients with Systemic Mastocytosis. N. Engl. J. Med. 303:1400-4 (1980). It should be noted that although tryptase and heparin are not involved in mediating the vasodilation, documentation of their release can also be helpful as a means to assess mastocyte activation.

A number of methods are available to quantify histamine in plasma and in urine. Beaven, M.A. , Jacobsen, S., Horakova, Z.: Modification of the Enzymatic Isotopic Assay of Histamine and its

Applications to Measurement of Histamine in Tissues, Serum, and Urine. Clin. Chim. Acta. 37:91-103 (1972). Shaff, R.E., Beaven, M.A. : Increased Sensitivity of the Enzymatic Assay of Histamine: Measurement of Histamine in Plasma and Serum. Anal. Biochem. 94:425-30 (1979). Brown, M.J. , Ind, P.W. , Causon, R. , Lee, T.H.: A Novel Double Isotope Technique for the Enzymatic Assay of Plasma Histamine: Application to Estimation of Mast Cell Activation Assessed by Antigen Challenge in Asthmatics. J. Allergy Clin. Immunol. 69:20-4 (1982). Dyer, J., Warren, K. , Merlin, S., Metcalfe, D.D., Kaliner, M. : Measurement of Plasma Histamine: Description of an Improved Method and Normal Values. J. Allergy Clin. Immunol. 70:82-7 (1982). Tsuruta, Y. , Kohashi, K. , Ohkura, Y. : Simultaneous Determination of Histamine and N-tau-methylhistamine in Human Urine and Rat Brain by High Performance Liquid Chromatography with Fluorescence Detection. J. Chromatogr. 224:105-10 (1981). Skofitsch, G., Saria, A., Holzer, S.P.,

Lembeck, F. : Histamine in Tissue: Determination by High Performance Liquid Chromatography After Condensation with O-phehaldezlaehyde. J. Chromatogr. 226:53-9 (1982). Shore, P.A. , Burkhalter, A., Cohn, V.P. : A Method for the Fluorometric Assay of Histamine in Tissues. J. Pharmacol. Exp. Ther. 127:182-6 (1959). Oates, J.A., Marsh, E. Sjoerdema, A.: Studies on Histamine in Human Urine Using a Fluorometric Method of Assay. Clin. Chem. Acta. 7:488-97 (1962). Siraganian, R.P.: An Automated Continuous Flow System for the Extraction and Fluorometric Analysis of Histamine. Anal, biochem. 57:383-94 (1974). Meyers, G. , Donlon, M. , Kaliner, M. : Measurement of Urinary Histamine: Development of Methodology and Normal Values. J. Allergy Clin. Immunol. 67:305-11 (1981). Roberts, L.J. II, Oates, J.A. : Accurate and Efficient Method for Quantification of Urinary Histamine by Gas Chromatography Negative Ion Chemical Ionization Mass Spectrometry. Anal. Biochem. 136:258-63 (1984). Mita, H., Yasueda, H. , Shida, T. : Quantitative Analysis of Histamine in Biological Samples by Gas Chromatography-mass Spectrometry. J. Chromatogr. 181:153-9 (1980). Mita, H. , Yasueda, H. , Shida, T. : Simultaneous Determination of Histamine and N-tau-methylhistamine in Human Plasma and Urine by Gas Chromatography Mass Spectrometry. J. Chromatogr. 221:1-7 (1980). Keyzer, J. . , Bruekelman, H. , Elzinga, H. , Koopman, B.J. , Wolthers, B.J., Bruins, A.P.: Determination of Histamine by Chemical Ionization Mass Spectrometry: Application to Human Urine. Biomed. Mass. Spectrom. 10:480-4 (1983). Guesdon, J.L., Chevrier, D. , Mazie, J.C., David, B. , Avrameas, S.J.: Monoclonal Anti-histamine Antibody: Preparation, Characterization, and Application to Enzyme Immunoassay of Histamine. J. Immunol. Methods. 87:69-78 (1986). McBride, P.T., Bradley, D. , Kaliner, M.A. : Evaluation of a Radioimmunoassay for Histamine Measurement in Biological Fluids. J. Allergy Clin. Immunol. 82:638-46 (1988) . Unfortunately, many of them have problems related to specificity, sensitivity, and reliability. Gleich, G.J., Hull, W.M. Jr.: Measurement of Histamine: A Quality Control Study. J. Allergy Clin. Immunol. 66:295-8 (1980). Warren, K. , Dyer, J. , Merlin, S., Kaliner, M. : Measurement of Urinary Histamine Comparison of Fluorometric and

Radioisotopic-enzymatic Assay Procedures. J. Allergy Clin. Immunol. 71:206-11 (1983). Roberts, L.J. II, Aulsebrook, K.A. , Oates, J.A.: Comparative Evaluation of the Radioenzymatic Method for the Determination of Urinary Histamine with a Mass Spectrometric Assay. J. Chromatogr. 338:41-9 (1985). Stable isotope dilution mass spectrometric assay is probably the most accurate and reliable. Roberts, L.J. II, Oates, J.A. : Accurate and Efficient Method for Quantification of Urinary Histamine by Gas Chromatography Negative Ion Chemical Ionization Mass Spectrometry. Anal. Biochem. 136:258-63 (1984). Mita, H. , Yasueda, H. , Shida, T.: Quantitative Analysis of Histamine in Biological Samples by Gas Chromatography-mass Spectrometry. J. Chromatogr. 181:153-9 (1980). Mita, H. , Yasueda, H. , Shida, T. : Simultaneous Determination of Histamine and N-tau-methylhistamine in Human Plasma and Urine by Gas Chromatography Mass Spectrometry. J. Chromatogr. 221:1-7 (1980). Keyzer, J.J., Bruekelman, H. , Elzinga, H. , Koopman, B.J., Wolthers, B.J. , Bruins, A.P.:

Determination of Histamine by Chemical Ionization Mass Spectrometry: Application to Human Urine. Biomed. Mass. Spectrom. 10:480-4 (1983). Additionally, most of the methods are primarily research tools and are not available for routine diagnostic use.

Even when one uses an accurate analytical method, there are other significant problems associated with quantifying histamine in human biologic fluids as an index of endogenous histamine production. For example, because circulating basophils contain histamine, artifactual release of histamine from basophils during blood sampling may prevent plasma levels from accurately reflecting true circulating levels of histamine. Although this problem can be minimized by careful blood sampling, immediately placing collected blood on ice, and using a calcium chelator, such as EDTA, as an anticoagulant to suppress activation of basophils, it probably cannot be totally eliminated. Brown, M.J. , Ind, P.W. , Causon, R. , Lee, T.H. : A Novel Double Isotope Technique for the

Enzymatic Assay of Plasma Histamine: Application to Estimation of Mast Cell Activation Assessed by Antigen Challenge in Asthmatics. J. Allergy Clin. Immunol. 69:20-4 (1982) . Problems unrelated to accuracy of the method are also associated with quantification of urinary histamine. A variety of bacteria are capable of decarboxylating histidine to form histamine. Taylor, S.L.: Histamine Food Poisoning: Toxicology and Clinical Aspects. CRC Crit. Rev. Toxicol. 17:91-128 (1986) . Thus, patients with overproliferation of genitourinary tract bacteria can exhibit artifactually increased levels of histamine in urine. As might be expected, this problem is encountered more often in women than in men. Keyzer, J.J. , de Mouchy, J.G.R. , van Doormal, J.J. , van Voorst Vader, P.C. : Improved Diagnosis of Mastocytosis by Measurement of Urinary Histamine Metabolites. N. Engl. J. Med. 309:1603-5 (1983) . An example of the problem of apparent local formation of histamine in the genitourinary tract is illustrated in Fig. 1. In this figure levels of histamine (•) and N^τ-methylhistamine (■) were measured in consecutive 24-hour urine collections from a female patient. The doted horizontal line indicates upper limits of normal for both histamine and N^r-methylhistamine. In this female patient, levels of histamine ranged from normal to tenfold greater than normal, but N^τ-methylhistamine levels were normal in all urines. Both histamine and its metabolite, N^τ-methylhistamine, were quantified by highly accurate mass spectrometric assays. Roberts, L.J. II, Oates, J.A. : Accurate and Efficient Method for Quantification of Urinary Histamine by Gas Chromatography Negative Ion Chemical Ionization Mass Spectrometry. Anal. Biochem. 136:258-63 (1984). Morrow, J.D., Parsons, W.G. Ill, Roberts, L.J. : Release of Markedly Increased Quantities of Prostaglandin D₂ in vivo in Humans Following the Administration of Nicotinic Acid. Prostaglandins 38:263-74 (1989). Measurement of N^τ-methylhistamine circumvents the problem ofbacterial production of histamine, because further metabolic transformation of locally formed histamine does not occur. Keyzer, J.J., de Mouchy, J.G.R. , van Doormal, J.J. , van Voorst Vader, P.C. : Improved Diagnosis of Mastocytosis by Measurement of Urinary Histamine Metabolites. N. Engl. J. Med. 309:1603-5 (1983).

Although it may be difficult if not impossible to interpret whether an elevated urinary histamine in an isolated urine collection actually reflects increased systemic overproduction of histamine, an increase and subsequent decrease in urinary histamine excretion temporally associated with an episode suspected to involve mast cell activation can probably be interpreted with a reasonable degree of confidence. Such an example is shown in Fig. 2. This figure shows levels of urinary histamine excretion in a patient measured during two asymptomatic periods (time zero) and following an episode of intense flushing. The dotted line indicated upper limit of normal. Alternatively, simultaneous quantification of histamine metabolites virtually eliminates "false-positive" results. There is yet another problem with using histamine levels as a diagnostic indicator of systemic mast cell disease. Patients with systemic mastocytosis appear to have altered metabolism of histamine. Keyzer, J.J., de Mouchy, J.G.R., van Doormal, J.J., van Voorst Vader, P.C: Improved Diagnosis of Mastocytosis by Measurement of Urinary Histamine Metabolites. N. Engl. J. Med. 309:1603-5 (1983). Granerus, G. , Olafsson, J.H, Rouge, G. : Studies on Histamine Metabolism in Mastocytosis. J. Invest. Dermarat. 80:410-6 (1983) . Their urinary excretion of histamine metabolites, N^τ-methylhistamine and N³-methylimidazoleacetic acid, is elevated disproportionately to their excretion of histamine. Thus, there is diminished sensitivity with measuring histamine compared to metabolites of histamine as an index of overproduction of histamine in patients with systemic mastocytosis. This phenomenon is illustrated in Fig. 3. In this figure fold increase above normal is the urinary excretion of histamine (0) and N^r-methylhistamine (•) in eight 24-hour collections from a patient with histologically documented systemic mastocytosis. In all urines analyzed from that patient with systemic mastocytosis, levels of methylhistamine are elevated to a much greater degree than are levels of histamine, which are either normal or only minimally elevated in some samples.

Systemic mastocytosis is characterized by an abnormal proliferation of tissue mast cells. Symptoms of mastocytosis are primarily attributed to the release of mast cell mediators during episodes of systemic activation of the excessive numbers of mast cells. Thus, biochemical evidence for the release of increased quantities of mast cell secretory products can suggest or confirm, depending on the clinical situation, a diagnosis of systemic mastocytosis. A major advantage of the biochemical approach to the diagnosis of systemic mast cell disease is that it has allowed the recognition of a class of patients in whom episodes of systemic mastocyte activation can be unequivocally documented biochemically, but in whom clear-cut evidence of abnormal mast cell proliferation is lacking by current histologic criteria. Although the release of increased quantities of mast cell mediators can be demonstrated during episodes of mast cell activation in such patients, mediator levels are usually normal at quiescent times. By contrast, patients with proliferative mast cell disease (mastocytosis) usually exhibit chronic overproduction of mast cell mediators. Thus, to facilitate the accurate diagnosis of systemic mast cell activation disorders, a method to detect or determine chemical markers of this disease has been developed.

Summary of the Invention The present invention relates to an improved method to diagnose systemic mast cell disorders by detecting the presence or determining the concentration of metabolites of prostaglandin D₂ present in a sample of bodily fluid. The metabolites can be detected by mass spectroscopy. Additionally, an object of this invention is to provide a method to immunologically detect metabolites of prostaglandin D or antibodies to these metabolites. The metabolites of prostaglandin D₂ detected in biological fluid include: 9α-hydroxy-11.15-dioxo- 2,3,18,19,-tetranorprost-5-ene-l,20-dioic acid and 9α,llp-dihydroxy-15-oxo-2,3,18,19-tetranor prost-5-ene-l,20-dioic acid.

Brief Description of the Figures

Figure 1 shows levels of histamine and N- methylhistamine measured in consecutive 24-hour urine collections from a patient.

Figure 2 shows levels of urinary histamine excretion in a patient measured following an episode of flushing.

Figure 3 shows fold increase above normal in the urinary excretion of histamine and N^τ-methylhistamine in eight 24 hour urine collections from a patient with histologically documented systemic mastocytosis.

Figure 4 shows the metabolism of prostaglandin D₂ to a biologically active metabolite 9α,ll)3-prostaglandin F₂. Figure 5 shows the metabolism of prostaglandin D₂ to metabolite A and metabolite B.

Figure 6 shows the level of urinary excretion of Metabolite A of a prostaglandin D₂ in patients with suspected systematic mast cell disease. Figure 7 shows a comparison of urinary levels of Me-histamine with metabolite B of prostaglandin D₂ in patients with documented mastocytosis.

Figure 8 shows cyclization and derivatization scheme for GC NICI/MS analysis of the major urinary prostaglandin D₂ metabolite (metabolite B) .

Figure 9 shows analysis of urine from a normal human volunteer for prostaglandin D metabolite (metabolite B) . A) In the absence of indomethacin treatment B) During treatment with indomethacin (200 mg/day) . Cr-creatinine.

Figure 10 shows analysis of urine from a normal human volunteer for prostaglandin D metabolite (metabolite B) which illustrate skewing in the relative ratio of chromatographic peaks I and II compared to that of the internal standard. A) In the absence of indomethacin treatment; B) during treatment with indomethacin (200 mg/day) . Cr-creatinine.

Figure 11 shows standard curve for the analysis of prostaglandin D metabolite (metabolite B) by NICI/MS. Varying quantities of unlabelled prostaglandin D metabolite (metabolite B) were added to 4.14 ng of (¹⁸0₄-prostaglandin D metabolite (metabolite B) ) to give the ratios noted on the vertical axis. On the horizontal axis are plotted the actual ratios of m/z 514 (unlabelled prostaglandin D metabolite (metabolite B) to m/z 522 (labelled prostaglandin D metabolite (metabolite B) ) measured by selected ion monitoring analysis.

Figure 12 shows analysis of prostaglandin D metabolite, i.e. metabolite B in urine from a patient with systemic mastocytosis.

Detailed Description of the Invention One of the major mediators released from mast cells is prostaglandin D₂, which is a vasodilator. Roberts, L.J. II, Sweet an, B.J., Lewis, R.A. , Austen, K.F., Oates, J.A. : Increased Production of Prostaglandin D₂ in Patients with Systemic Mastocytosis. N. Engl. J. Med. 303:1400-4 (198_.0). However, prostaglandin D₂ is initially metabolized in vivo by 11-ketoreductase to the metabolite,

9α,ll)8-Prostaglandin F₂, which is a vasso pressor substance. See Figure 4. Liston, T.E., Roberts, L.J. II: Transformation of Prostaglandin D₂ to, 9α,11/3-(155)-trihydroxy-prosta-(5Z,13E)-dien-1-oic (9α,ll/3-Prostaglandin F₂ ) : A Unique Biologically

Active Prostaglandin Produced Enzymatically in vivo in Humans. Proc. Natl. Acad. Sci. USA 82:6030-4 (1985). To measure this metabolic event the inventor(s) , developed a mass spectrometric assay for the measurement of a metabolite of prostaglandin D₂,

9α-hydroxy-ll.15-dioxo-2,3,18,19,-tetranor-prost-5-ene- 1,20-dioic acid (Fig. 5, compound A) . Roberts, L.J. II: Quantification of the PGD₂ Urinary Metabolite 9cr-hydroxy-ll,15-dioxo-2,3,18,19-tetranorprost-5- ene-l,20-dioic Acid by Stable Isotope Dilution Mass Spectrometric Assay. Methods Enzymol. 86:559-70 (1982) .

Striking increases in the urinary excretion of this metabolite were found in patients with systemic mast cell disorders (Fig. 6) . Metabolite levels were determined in patients both with and without proliferative mast cell disease, primarily in urine collected surrounding episodes of suspected mastocytotic activation. Additionally, the inventors developed a mass spectrometric assay for the major urinary metabolite of prostaglandin D₂, which differs from the above metabolite in that the C-ll keto group has been reduced by ll-ketoreductase to a 11/3 hydroxyl (Fig. 5, compound B 9α,ll)3-dihydroxy-15-oxo-2,3,18,19-tetranorprost-5- ene-l,20-dioic acid) . The normal level of this metabolite in human urine was found to be 1.08 + 0.72 ng/mg creatinine (mean + S.D.) . This metabolite is also detectable in normal human plasma at levels of approxi atley 2-10 picogroams per ml. Levels of this metabolite of prostaglandin D₂ in urine correlate with levels of methylhistamine in urine from patients with systemic mastocytosis.

Surprisingly, recent data suggests that in some patients with mastocytosis, levels of this urinary prostaglandin metabolite (Fig. 5, Compound B) may be surprisingly a more sensitive biochemical indicator of disease than levels of urinary methylhistamine. In particular, this result can be seen in Fig. 7. Fig. 7 shows a comparison in the fold increase above the upper normal limit of urinary methylhistamine and the major urinary metabolite of prostaglandin D₂ in patients confirmed to have mastocytosis. Note that almost all of the dots lie to the right of the line of equality in the fold-increase in methylhistamine and prostaglandin D metabolite. The solid circled dots have normal methylhistamine levels, but substantially elevated prostaglandin metabolite. In such patients, a biochemical diagnosis of mastocytosis could not have been made by measuring methylhistamine but could be established by measuring the prostaglandin metabolite. Additionally the dots circled in a broken line show only minimally elevated methylhistamine levels in comparison to levels of the prostaglandin metabolite several-fold above the normal upper limit.

In addition to the above discussed assays, immunological assays for prostaglandin D₂ metabolites are contemplated by this invention. Substantially, pure prostaglandin D₂ metabolites can be used as an antigen in the process disclosed by Kohler and

Milstein, Nature 256:495 (1975) to make cell lines that secrete monoclonal antibodies. Additionally, polyclonal antibodies could be made to these antigens using conventional techniques. These antibodies could be used in a variety of standard immunological tests to detect or determine prostaglandin metabolites in a sample. In an alternative embodiment, the prostaglandin-like compound can be assayed using immuno or nucleic assays. Immunoassay is a suitable method for the detection of small amounts of specific fatty acid derivatives (5-500 picogram can be readily detected) . In particular, antibodies can be raised to these prostanoid-like compounds using conventional techniques. The antibodies can then be used in immunoassays to quantitate the amount of prostaglandin-like compounds in the biological fluid.

A small molecule (less than 5-10 kilodaltons) will usually not elicit the production of antibodies in experimental animals unless covalently linked to large immunogenic molecules prior to immunization.

Prostaglandin metabolites can be coupled via its carboxyl group to a carrier protein by the dicyclohexyl-carbodiimide method. Rich, D.H., etal.. The Peptides, 1:241-61 (1979); U.S. Patent No. 4,859,613. These compounds can be coupled to keyhole limpet hemocyanin (KLH) by the DCC method disclosed by Levine et al. Prostaglandines 20: 923-34 (1980) . For example, 9α,ll3-dihydroxy-15-oxo-2,3,18,19- tetranorprost-5-ene-l,20-dioic acid is dissolved in 100^1 of N,N-dimethyl formamide. 9α,110-dihydroxy- 15-OXO-2,3 ,18 ,19-tetranorprost-5-ene-l ,20-dioic acid is then activated by the addition of 3 mg of DCC in the presence of 3.5 mg of N-hydroxysuccinimide as a trapping agent. This reaction mixture is stirred for 30 minutes at room temperature. As the reaction proceeds the by-product dicyclourea will form a white precipitate. When the reaction is complete, this precipitate is removed by centrifugation. The surpernatant is then added to 6.25 gm of KLH in 0.5 ml of 0.1N NaHC0₃ and stirred for two hours at 4^βC. The conjugate is then extensively dialyzed against phosphate-buffered saline, pH 7.5 (0.15MNaCl, 0.005M NaHP0₄) . The conjugate is aliquoted and stored at -20^βC. It should be noted that many unsaturated fatty acids and their derivatives, are sensitive to oxidation, this does not hold for prostaglandins which have only a single double bond on each side of the chain.

These immunogens can then be used to raise antibodies. Rabbits can be immunized with the

9α,ll,3-dihydroxy-15-oxo-2,3,18,19-tetranorprost-5-ene-l, 20-dioic acid-KLH conjugate administered in Complete Freund's Adjuvant. One immunization dosing and scheduling is as follows: 100μ of conjugate in Complete Freund's Adjuvant will be administered at multiple subcutaneous sites for the primary immunization. At subsequent two week intervals 100μ of conjugate in incomplete freund's adjuvant will be administered at multiple subcutaneous sites to boost titers. Serum will be collected from animals one week following the initial boost and titers will be determined by ELISA. Similarly, myeloma that secrete monoclonal antibodies can be prepared following the techniques described by Kohler & Milstein (1975) . These antibodies can be used in immunoassays for prostaglandin metabolites. Typical assays for these types of compounds include enzyme-linked immunoassays, fluorescent immunoassays, and radioimmunoassays. The assays can be in the sandwich or competitive formats. See e.g. , David, U.S. Patent Nos. 4,736,110 and 4,486,530. In particular, a solid phase radioimmunoassay for the arachidonate derivatives 12(S)-HETE developed by Oxford Biomedical Research, Inc. could serve as a model for an immunoassay to prostaglandin metabolites.

Example 1: Quantification of major urinary metabolites in urine samples.

Preparation of r—0₄]-Labelled PGD-M: Unlabelled prostaglandin D metabolite (metabolite B) was chemically synthesized as described (Prakash, C. , Saleh, S., Roberts, L.J., Blair, I.A., Taber, D.F.: J. Chem. Soc. Perkin. Trans. 1:2821-16 (1988)) and subsequently converted to an [¹⁸0₄]-derivative for use as an internal standard by the method of Murphy et al. involving successive steps of methylation and alkaline hydrolysis with Li ¹⁸-OH. Murphy, R.C. , Clay, K.L.: In Method in Enzymology (Lands, W.E.M., and Smith, W.L. , Eds.) : Academic Press, New York. 86:547-51. This yielded an internal standard with an unlabelled blank of only 1 part per 10,000 when analyzed by GC/NICI-MS. The blank remained unaltered when the labelled standard was subjected to the assay procedure subsequently developed. Standardization of f-^O-^l-Labelled prostaglandin D metabolite (metabolite B) : Insufficient quantities of prostaglandin D metabolite (metabolite B) have been synthesized in pure form to weigh accurately for standardization. Therefore, the [¹⁸0₄]-prostaglandin ^D metabolite (metabolite B) was standardized against 6-keto-prostoglandin F_lα. An aliquot of the labelled prostaglandin D metabolite (metabolite B) was cyclized as described subsequently and the upper chain carboxyl was converted to a methyl ester. An aliquot of an accurately weighed quantity of 6-keto-prostoglandin F_lα was added and the mixture was converted to an 0-methyloxime, pentafluorobenzyl ester, trimethylsilyl ether derivative and subsequently analyzed by GC/NICI-MS. The negative ion mass spectra of this derivative of prostaglandin D metabolite (metabolite B) and the corresponding derivative of 6-keto-prostaglandin F^_α both generate essentially a single M-181 ion formed as the result of the loss of «CH C₆F₅. Blair, I.A., Banow, S.E., Waddell,

K.A. , Lewis, P.J., Dollery, C.T. : Prostaglandins. 23:579-89 (1982). Thus quantification was accomplished by determining the ratio of the M-181 ion for [¹⁸0 ]-labelled prostaglandin D metabolite (metabolite B) (m/z 522) to the M-181 ion for

6-keto-Prostaglandin F_lα (m/z 614) . In the absence of sufficient quantities of pure synthetic prostaglandin D metabolite (metabolite B) which can be weighed for standardization, standardization by GC/MS against 6-keto-prostoglandin F_lα was chosen because both it and prostaglandin D metabolite (metabolite B) are stable PGF-ring compounds of similar polarity which require methoximation for analysis. Since the two compounds were co-derivatized for anlysis in the same vial, any derivatization losses that occur would be expected to be similar for both compounds. This standardization was carried out six times with a precision of ± 12%. Important in regards to the accuracy of this method for quantification of prostaglandin D metabolite (metabolite B) based on the ratios of the M-181 ions for prostaglandin D metabolite (metabolite B) and 6-keto-Prostoglandin F_lα is whether the percent of the total ion current represented by the M-181 ions in the negative ion mass spectra of the two compounds are the same. These percentages were found to be essentially identical, differing by only 0.01%.

Purification and Analysis of Urinary PGD-M: To one ml of urine is added 2.07 ng of the [ 0 ]-prostaglandin D metabolite (metabolite B) internal standard. The sample is then acidified to pH 3 with 1M HCL and allowed to stand at room temperature for 30 minutes. This quantitatively results in cyclization of the molecule to Compound C in Figure 8. Confirmation that cyclization under these conditions is quantitative was established upon analysis of uncyclized standard subjected to the same treatment by thin layer chromatography which effectively separates the cyclized and uncyclized forms of the metabolite. Following this, the sample is applied to a C-18 SEP-PAK preconditioned with 10 ml methanol and 10 ml pH 3 water. The SEP-PAK is then washed with 10 ml pH three water followed by 10 ml heptane. The extracted compounds are then eluted with 10 ml of ethyl acetate: eptane (50:50) into a scintillation vial. The ethyl acetate:heptane is dried over 10 grams anhydrous Na₂S0₄ and then applied to a silica SEP-PAK prewashed with 10 ml methanol and 10 ml ethyl acetate. The SEP-PAK is subsequently washed with 5 ml ethyl acetate and compounds are then eluted into a reactivial with 5 ml ethyl acetate: ethanol (50:50) .

The ethyl acetate: ethanol is then dried under a stream of nitrogen and the residue reconstituted in 50 ul methanol. At this point, the upper side chain carboxyl of prostaglandin D metabolite (metabolite B) is a free acid (Figure 8, Compound C) . The upper side chain carboxyl is then converted to a methyl ester by adding excess ethereal diazomethane and allowing the mixture to stand at 20°C for five minutes. However, the lower side chain carboxyl is not methylated by this treatment because it is in the form of a lactone (Figure 8, Compound D) . The methylation solvents are evaporated under nitrogen and the residue is resuspended in 50 ul of acetonitrile. 200 ul of an aqueous solution of 3% methoxyamine-HCL are then added to the sample and allowed to react. This results in ring opening and formation of a 0-methyloxime derivative of the keto group at C-15 (Figure 8, Compound E) . One ml of ethyl acetate is then added and prostaglandin D metabolite (metabolite B) is extracted from the methoxyamine solution. The organic phase is removed and dried under nitrogen.

At this point, the upper side chain carboxyl of the prostaglandin D metabolite (metabolite B) molecule is methylated whereas the lower side chain carboxyl is a free acid. However, other acidic lipids which contain carboxyl groups that cannot undergo cyclization would have been methylated by the previous treatment with diazomethane. Therefore, considerable purification of prostaglandin D metabolite (metabolite B) from such potentially interfering lipids can be achieved at this point by incorporating an extraction step at neutral pH. At neutral pH, esterified impurities will be extracted whereas prostaglandin D metabolite (metabolite B) remains in the aqueous phase since the lower side chain carboxyl will predominantly be in the form of a carboxylate anion at pH 7. This extraction is accomplished by adding 1 ml of 50 mM borate buffer pH 7 to the dried sample followed by 1 ml of ethyl acetate. The mixture is vortexed and the organic phase is discarded. The aqueous buffer phase is then reduced to pH 3 with 1M HCL and prostaglandin D metabolite (metabolite B) is extracted into an equal volume of ethyl acetate. The aqueous phase is discarded and the organic phase evaporated under nitrogen.

Finally, the lower carboxyl group on the prostaglandin D metabolite (metabolite B) molecule is converted to a pentafluorobenzyl ester (Figure 8, Compound F) by adding 40 ul of a 10% solution of pentaflurobenzyl bromide in acetonitrile and 20 ul of a 10% solution of diisopropylethanolamine in acetonitrile. The mixture is incubated for 20 minutes at 37^βC and then dried under nitrogen. This procedure is then repeated a second time. Two cycles of esterification are performed since we have found that two cycles achieves more quantitative esterification of prostaglandins compared to one cycle, even if the length of the derivatization is allowed to proceed for a longer period of time with excess reagents.

After the sample has been dried, it is resuspended in 50 ul of methanol and subjected to further purification by thin layer chromatography using a solvent system of acetone:hexane (35:65) . The plates are developed to 13 cm. Prostoglandin F_2α methyl ester (5ug) is chromatographed on a separate lane and visualized by spraying with 10% phosphomolybdic acid in ethanol and heating. Compounds in the urine sample migrating in the region of Prostoglandin F_2α methyl ester (Rf-0.08) and up to 2 cm ahead are scraped from the silica gel and eluted with 1 ml of ethyl acetate.

PGF_2α methyl ester is used as a thin layer chromatography marker because of the unavailability of sufficient quantities of synthetic unlabelled prostaglandin D metabolite (metabolite B) for this purpose.

Prostaglandin D metabolite (metabolite B) is then converted to a trimethylsilyl ether derivative

(Figure 8, Compound G) by adding 20 ul N.O.-bis(trimethylsilyl)-trifluoroacetamide and 10 ul dimethylformamide and incubating at 40^βC for 10 minutes. The reagents are evaporated under nitrogen and the residue redissolved for GC/MS analysis in 10 ul of undecane which has been dried over a bed of calcium hydride.

As discussed above, the major ion generated in the NICI mass spectrum of the O-methyloxime, mono-methyl ester, mono-pentabluorbenzyl ester, bis-trimethylsilyl ether derivative of prostaglandin D metabolite (metabolite B) is m/z 514 which represents the M-181 (M-CH₂C_gF₅) carboxylate anion. The r L¹⁸n^υ4]-internal standard generates an anlogous ion at m/z 522. Quantification of endogenous prostaglandin D metabolite (metabolite B) is accomplished by selected ion monitoring analysis of the ratio of intensities of m/z 514 to m/z 522. The recovery of prostaglandin D metabolite (metabolite B) through the assay is in the range of 10-20%.

The above purification scheme was originally devised for assaying the metabolites. Recently, the inventors have found that with minor modifications sufficient purification of the metabolite can be accomplished without employing thin layer chromatography which makes the assay more efficient. In particular, these modifications include the elimination of the step of eluting the compound through silica SEP-PAK, See p. 17, 1. 27 and the elimination of thin layer chromatography process.

Assay Results: A representative selected ion monitoring chromatogram obtained from analysis of urine from a normal individual is depicted in Figure 9A and B. At the bottom is the m/z 522, selected ion monitoring chromatogram representing the r ^L18n^u4]-labelled internal standard. Two peaks are present representing the syn- and anti-methoxime isomers. The m/z 514 chromatogram at the top of the figure representing endogenous compounds reveals four peaks of approximately equal intensity (labelled I-IV) rather than two as is present for the internal standard. The two peaks in the m/z 514 chromatogram directly above the internal standard (I and II) in this urine are presumed to represent prostaglandin D metabolite (metabolite B) derived from endogenously produced prostoglandin D₂. This is supported by several lines of evidence. First, it was demonstrated in two volunteers treated for three days with either indomethacin (200 mg/day) , naproxen (4000 mg/day) , or ibuprofen (4000 mg/day) that levels of the first two peaks (I and II) were reduced by 67-99%. Figure 9B shows the analysis of urine obtained from the same individual as in Figure 9A during treatment with indomethacin. In the absence of indomethacin pretreatment, the level of prostaglandin D metabolite (metabolite B) was 0.96 ng/mg creatinine. Treatment with the cyclooxygenase inhibitor indomethacin reduced levels of both peaks I and II by approximately 90% to 0.10 ng/mg creatir .ne. Comparison of these chromatograms obtained with and without cyclooxygenase inhibition reveals a marked alteration in the ratio of the first two m/z 514 peaks (I and II) relative to the second two peaks (III and IV) . This indicates that whereas the first two peaks (I and II) represent metabolites of cyclooxygenase derived products (prostaglandin D₂) , the second two (III and IV) do not.

Peaks I and II have the same retention time on capillary gas chromatography as the ¹⁸0-labelled prostaglandin D metabolite (metabolite B) internal standard and that their intensity is markedly suppressed by cyclooxygenase inhibitors strongly suggests that they represent the endogenous metabolite of prostaglandin D₂. However, one cannot exclude the possibility that the material in peaks I and II may also be comprised, at least in part, by the analogous metabolite of cyclooxygenase derived prostoglandin F_2α. Although the analogous metabolites of prostaglandin D₂ and prostaglandin F_2α have the same basic structure and molecular weight and would thus generate the same M-181 ion (m/z 514) , they differ in regards to the stereochemistry of the cyclopentane ring hydroxyls. The cyclopentane ring hydroxyls in the prostaglandin D₂ metabolite are oriented trans due to the fact that 11-ketoreductase stereospecifically transforms prostaglandin D₂ to 9α,110-prostaglandin F₂. Liston, T.E., Roberts, L.J. II: Proc. Natl. Acad. Sci. USA 82:6030-4 (1985). However, the cyclopentane ringe hydroxyls of the analogous metabolite derived from prostaglandin F_2α are cis. Fortuitously, the prostaglandin F_2α metabolite is a by-product of the chemical synthesis of the prostaglandin D₂ metabolite. Prakash, C. , Saleh, S., Roberts, L.J., Blair, L.A. , Taber, D.F. J. Chem. Soc. Perkin Trans. 1:2821-6 (1988) . Therefore, we were able to examine the capillary gas chromatography characteristics of this compound. It was found to be separated from peaks I and II, eluting with a longer retention time in region of peaks III and IV.

Although the chromatogram obtained from analysis of prostaglandin D metabolite (metabolite B) in urine most often resembles that shown in Figure 9A, occasionally we have also obtained chromatograms from the analysis of urine from normal individuals which differ slightly as shown in Figure 10A. In this chromatogram, the relative ratio of the first two peaks (I and II) to each other in the m/z 514 chromatogram are skewed in comparison to the ratio of the two peaks representing the internal standard in the m/z 522 chromatogram. Specifically, peak II is higher relative to the intensity of peak I. Following treatment of this individual with indomethacin (200 mg/day for three days) , peak I was found to be suppressed by approximately 90%. In contrast, however, peak II was suppressed by only approximately 40% (Figure 10B) . This suggests that the second peak is comprised of more than one compound, one of which is derived from cyclooxygenase sources (prostaglandin D₂) and the other from noncyclooxygenase sources. This was further confirmed by analyzing small 0.5 cm sequential cuts of the area normally scraped on the thin layer chromatography plate which revealed that peak II is in fact comprised of at least two compounds with slightly different but overlapping R_f values. On the other hand, the analysis of seuqential thin layer chromatography cuts in this urine indicated that peak I is apparently comprised of only a single compound. Similar analyses have been performed on five urines from different individuals with similar results indicating that, whereas peak II in some urines represents a mixture of more than one compound, peak I, in each instance, appears to be comprised of a single compound. This was further confirmed by determining the stereochemical orientation of the cyclopentane ring hydroxyls. The cyclopentane ring hydroxyls of the prostaglandin D₂ metabolite are oriented trans. We have found that the compounds in peaks III and IV have hydroxyls which are exclusively oriented cis. This was determined by assessing their ability to form a cyclic boronate derivative. Liston, T.E., Roberts, L.J. : J. Biol. Che . 260:13172-80 (1985). Liston, T.E., Roberts, L.J. II: Proc. Natl. Acad. Sci. USA 82:6030-4 (1985) . This would be consistent with their being metabolites of noncyclooxygenase derived F-type prostaglandins since the cyclopentane-ring hydroxyls in the prostaglandin F₂-like compounds produced by this mechanism are predominantly oriented cis. Morrow, J.D., Hill, K.E., Burk, R.F., Nammour, T.M. , Badr, K.F., Roberts, L.J. II: Proc. Natl. Acad. Sci. USA, 87:938-37 (1990) . We found that peak II in urines such as shown in Figure 10 is comprised of a mixture of compounds with cis and trans hydroxyls. However, we have found that cyclopentane-ring hydroxyls in the material in peak I are exclusively oriented trans. This provides further evidence that peak I is not a mixture of compounds and that it is comprised entirely by the first methoxime isomer of the endogenous metabolite of prostaglandin D₂.

Assay Parameter and Validations: The lower limit of detection (signal to noise ratio of approximately 4:1) of prostaglandin D metabolite

(metabolite B) is in the range of 50 pg/mg creatinine. Several procedures were performed to establish the accuracy of this assay. Because, as discussed above, peak II in the m/z 514 chromatogram can be comprised of more than one compound, quantification is accomplished by comparing the ratio of peak I in the m/z 514 chromatrogram to the corresponding peak in the m/z 522 chromatogram. Initially a standard curve was construced by adding varying amounts of unlabelled prostaglandin D metabolite (metabolite B) to a fixed quantity of 4.14 ng of [ 0₄]-prostaglandin D metabolite (metabolite B) and the measured ratio of m/z 514 to m/z 522 to the expected ratio was compared. The standard curve was found to be linear over a concentration range of 16-fold (Figure 11) .

Experiments were then carried out to establish the precision and accuracy of this assay. Precision was measured by analyzing six 1 ml aliquots of urine obtained from a 24 hour collection from a normal volunteer. The mean of three replicate measurements of the ratio of m/z 514 to m/z 522 was determined for each sample. The precision was found to be ± 7%. Accuracy was assessed using the same urine. For this, 1.2 ng of unlabelled prostaglandin D metabolite (metabolite B) was added to another four 1 ml aliquots of the urine and reassayed. The amount of endogenous prostaglandin D metabolite (metabolite B) measured in the precision experiment was subtracted from the total measured and the accuracy of the assay to measure the added 1.2 ng of prostaglandin D metabolite (metabolite B) was calculated. The accuracy was found to be 96%.

The effect of storage of urine on levels of prostaglandin D metabolite (metabolite B) measured was also investigated. In two urine samples from each of three individuals, there was less than a 10% variation in the level of prostaglandin D metabolite (metabolite B) measured in urine which was analyzed immediately compared to the level measured following storage at -20'C x 1 year. Urinary Prostaglandin D Metabolite (metabolite B) Levels in Normal Humans: To establish the normal range of the urinary excretion of prostaglandin D metabolite (metabolite B) , aliquots of urine from 24 hour urine collections were obtained and analyzed from nine healthy men and nine healthy women. Normal levels were found to be 1.08 ± 0.72 ng/mg creatinine (mean ± 2 S.D.) . There were no significant differences in the levels of prostaglandin D metabolite (metabolite B) in males versus femals (men 1.19 ± 0.45 ng/mg creatinine vs. women 0.96 ± 0.24 ng/mg creatinine) .

Urinary Prostaglandin D Metabolite (metabolite B) Levels in Clinical Situations Associated With Increased Release of PGD₂: We then examined the ability of the assay to assess overproduction of PGD₂ in clinical situations in which increased quantities of prostaglandin D are known to be released in vivo. Prostaglandin D has been shown to be markedly overproduced in patients with systemic mastocytosis. Roberts, L.J. , Sweetman, B.J. , Lewis, R.A. , Austen, K.F., Oates, J.A. : N. Engl. J. Med. 303:1400-4 (1980) . Figure 12 shows the results of urine analyzed for prostaglandin D metabolite (metabolite B) in such a patient. Note in this chromatogram that the ratio of the first two peaks (I and II) relative to the second two (III and IV) is markedly altered opposite to what was found with treatment with indomethacin (Figure 9B) . The level of prostaglandin D metabolite (metabolite B) in this patient's urine was substantially elevated approximately 12-fold above normal at 12 ng/mg creatinine.

Ingestion of the hypolipidemic agent, niacin, evokes intense flushing. Recently we reported that niacin induces the release of large quantities of prostaglandin D₂ assessed by measuring levels of the prostaglandin D₂ metabolite, 9α,113-prostaglandin F₂, in plasma. Morrow, J.D., Parsons, W.G., Roberts, L.J.: Prostaglandins. 38:263-74 (1989). Analysis of urines from three individuals collected over a seven hour period following ingestion of 500 mg of niacin revealed elevated urinary excretion of prostaglandin D metabolite (metabolite B) ranging from 25-48 fold above normal. At present, the main use of the assay is in the diagnosis of patients with systemic mast cell activation disorders. However, mast cells are also activated to release prostaglandin D₂ during allergic reactions in "normal" allergic individuals. Thus, quantification of the metabolite should be useful in diagnosis for example anaphylactic reactions. Since anaphylaxis involves systemic activation of mast cells, it should be noted that systemic mast cell activation disorders would be inclusive of anaphylaxis. While the present invention has been described by reference to certain illustrative examples, various modifications and variants within the spirit and scope of the invention will be apparent to those skilled in the art.

Claims

I Claim:

1. A method to diagnose systemic mast cell disorders in a sample from a patient the improvement comprising: a) detecting the presence or determining concentration of metabolites of prostaglandin D₂ present in said sample; b) comparing 1(a) to a control to diagnose said disorders.

2. The method of claim 1 wherein said sample is human urine.

3. The method of claim 1 wherein said disorder is mastocytosis.

4. The method of claim 1 wherein said sample is human plasma.

5. The method of claim 1 wherein said metabolite is 9α-dihydroxy-ll,15-dioxo-2,3,18,19-tetranorprost- 5-ene-l,20-dioic acid.

6. The method of claim 1 wherein said metabolite is 9α,ll3-dihydroxy-15-oxo-2,3,18,19-tetranorprost- 5-ene-l,20-dioic acid.

7. A method to diagnose systemic mast cell disorders in a sample from a patient the improvement comprising: a) immunologically detecting or determining concentration of a metabolite of prostaglandin D₂ in said sample; b) comparing step (b) to a control; c) diagnosing said systemic mast disorders in said patient based on step (c) .

8. The method of claim 9 wherein said sample is human urine.

9. The method of claim 7 wherein said metabolite is 9α-dihydroxy-ll,15-dioxo-2,3,18,19-tetranorprost- 5-ene-l,20-dioic acid.

10. The method of claim 7 wherein said metabolite is 9α,110-dihydroxy-15-oxo-2,3,18,19-tetranorprost- 5-ene-l,20-dioic acid.

11. The method of claim 7 wherein said sample is human plasma.