C1 & RFP TITLE: BIOCHEMICAL MARKERS OR ENZYME CHANGES THAT MAY PRESAGE HE PRESENCE OF CANCER RFP No. NCI-CB-74-29 (PROJECT NO. CB-43902-S) SUBMITTED BY: THE BOARD OF TRUSTEES OF THE LELAND STANFORD, JR. UNIVERSITY PROPOSAL SUMMARY AND DATA SHEET NCI-CB-74-29 MAME OF OFFEROR The Board of Trustees of the Leland Stanford Jr. University PLACE OF PERFORMANCE (City, County, and Seate) Stanford, Santa Clara County, California ,TOTAL ESTIMATED TIME REQUIRED TO 1ST, VEAR (Detalied budget required) | 4TH YEAR ° COMPLETE THE PROJECT TOTAL $70, 808 ESTMATED 2D YEAR eT YEAR 76,130 3 years : CosTs, 3nO YKAA ~ 81,855 TYPE OF CONTRACT PROPOSED [2G cosv-nenssunseuent Oo cosT-mamne - CC cost-ecus-cineo-rae C] »xep-ence CJ otuee NAME OF INDIVIDUALISM AUTHORIZED TO NTtLE TELEPHONE HUMBER NEGOTIATE 415 321-2300 . : X2883 Jon A. Erickson Sponsored Projects Officer . 415 321-2300 Robert D. Simmons Contracts and Grants Manager X2883 . . Manager, 415 321-2300 Jason C. Mavis Sponsored Projects Office X2883 NAME OF INDIVIDUALIS) AUTHORIZED TO TITLE TELEPHONE NUMBER CAECUTE AND NEN CONTRACTS , Manager, 415 321-2300 Jason C. Mavis . Sponsored Projects Office X2883 Robert D. Simmons Contracts and Grants Manager 415 aaaa MUMBER OF DePLOVEES CURRENTLY M@ePLoveD . BOLLAR VOLUME OF GUBNESS FER Annus 7,000 . . . $60 Million Acceptance Period - The offeror agrees to allow_120_days from the date of this propoxal far acceptance thereof by the Government (120 days, if not otherwize apecified). MAME AND ADORESS OF COGMIZANT GOVERNMENT AVOIT AGENCY Defense Contract Audit Agency 400 ‘Cambridge Ave., Suite 404 . - Palo Alto, Calif. 94306 NAME OF PRINCIPAL INVESTIGATOR TEL EP HONE NUMBER EsSTimaTED SOCIAL SECURITY a ~ o~ _ NOURS WEEKLY weuneeen ° - | 415-321-1200 Prof. Joshua Lederberg X5801 two > 3 name or COANVESTIGATORS AND sENION TELEPHONE NUMER ESTIMATED / SOCIAL SECUMTY — 415-321-1200 | MOUntweaxty ae X5239 : twenty Dr. Alan Duffield 415-321-1200 \ Dr. Wilfred Pereira X5239 ‘Twenty > IT. III. IV. Vv. VI. VII. IX. INTRODUCTION.... TECHNICAL KEY Objectives a Background. . a, Experimental PERSONNEL... I. TABLE OF CONTENTS nd Scope, e®e¢eeeeens sevens eseeenve eee 1 i re 3 *eeeene eeeene @eeveve eevee @evens . 10 TOO Reet eee e cence ccc cenecece 15 FACILITIES AND EQUIPMENT... se eeeessecccceccse. 16 RELATED ACTIVITIES. eee eeecceceeceseseences, 16 PROGRAM MANAGEMENT. 6 eee eeeeeseessccceccecess, 17 REFERENCES...... TO eet Cee meee ee oe eee c ec cascnne 17 CURRICULA VITAE eee eee e ee eeececececscccseeces, 22 BUDGET.. APPENDICES (A) (B) - (C) Letter sig Letter sig Studies of Premature Mass Spect Te ett CCR Cero ree a seca nesceens 40 \ ned by W. R. Fair..........00., 46 ned by T.F.Long/J.R.Wilbur..... 47 the Urinary Metabolites of Infants Using Gas Chromatography FOMCETY ese ceeecesescccccccun,, 48 RECENT AND IN-PRESS REFERENCES (A) (B) (Cc) (D) (E) (F) Reference Reference Reference Reference Reference Reference Dee e ee ee eens cencceveesceneacs, 69 Gove eee eceeececsecenccesccecccs” 75 Te cececeecceeccassccuscessence, 99 AD ec ceseceescnecccnccensceccce, 107 NB eee cee ccccccucsceneseesececs. 115 NOs sees ee eeeeeeesesceneeeens 126 RFP No. NCI-CB-74~29 REP NO. NCL-UB-/4-29 II. INTRODUCTION The objectives of this work are to develop further the uses of gas-liquid chromatoqraphy (GC) and mass spectrometry (HS ) instrumentation, under computer management, for the study of derangements of human biochemical metabolisn due to neoplastic change. The efficacy of these analytical tools has been demonstrated when anplied to the determination of the metabolic content of limited populations of urine samples in the research laboratory environment. ile propose to enlarge the clinical investigative applications of GC/MS technology and to demonstrate its utility for the screening of urine samples from clinically diagnosed cancer patients with a view to identifying new metabolites, or altered levels of known metabolites, which might be diagnostic for cancer. Snecific qoals include the applications of GC/MS analysis capabilities to large and more diversified populations to establish better defined norms, deviations, and control parameters necessary to relate GC/iS analysis results to identifiable disease states. Control urine samples will be obtained from healthy volunteers and from opatients hosvitalized for disease other than cancer. These will be matched by age, sex, smoking history, drug and alcohol habits and clinical judaments of general state of nutrition length of illness (disturbance of work, exercise, and weight), and exposure to radiation and to surqery. It will not be possible to control dietary and medicinal ‘intake, but these factors will he monitored for their expected influence on urinary output. In addition sanples will he taken fron healthy relatives of oatients to serve as controls and as possible sources of evidence for aenetic-nmetahbolic predispositions to specific forms of cancer. As promising biochemical markers for cancer are recognized we will collaborate with clinicians in applying this methodolony for the practical early detection of the types of cancer studied. TI]. TECHNICAL A. Objectives and Scope. (1) Gas Chromatooraphy/‘ass Soectrometry Screening of Urine from Cancer Patients for the Identification of Metabolites Related to this Disease. We plan to use gas chromatography/mass spectrometry (GC/MS)to screen urine from control subjects and individuals suffering from the following forms of cancer: (a) prostatic cancer (b) bladder cancer (c) Hodakin’s Disease (d) Various lymphomas These samples will be obtained from the Clinical Research Center and the Uroloay Clinic of the Stanford University Medical -~2- RFP No. NCI-CB-74-29 School through the cnoveration of Dr. W. Fair (Associate Professor of Surgery, see attached letter of intent, Aopendix A.) As detailed in Section IIIC of this proposal these samples will be scrutinized to identify the metabolites present in the “acidic", "amino acid" and "carbohydrate" fractions usina an adaptation of the procedure described by Jellum and his croup (ref. 1). Comparison of the metabolic content of the urines from control and diseased subjects will then be made to detect any metabolites diagnostic for the various cancers studied. (2) The Quantitation of Urinary Beta-Aminoiso-butyric Acid (BAIB) Levels from Cancer Patients. During a pilot survey of the urinary metabolites excreted hy children with diagnosed leukemias we observed several who excreted massive amounts (excess of | aram/24 hours) of BAIS (I). Others in our studied pooulation excreted far smaller amounts of BAIB (milligqrans/24 hours) and frequently other connounds were seen to co-elute which made it imnossible to accurately quantitate for daily BAIB excretion. This experience led us to develon a method usinn mass fragmentography (see section IIIC (2)) for the quantitation of BAIB at the nanogqram level (ref. 23 copy attached). Subsequently other investiaators, usina an amino acid analyzer reported on the sianificance of urinary BAIB excretion in cancer patients (ref. 3). H.C 3° ~ CH-COOH HN Py We will assay for BAIB those clinical samples fron diagnosed cancers listed in section IIIA (1) together with samples received fron Drs. T. Lona and J. R. tilbur (Children’s Hosoital, Stanford University: See letter of intents Annendix 8). In this latter instance we will monitor urinary BAIB levels of children as they receive treatment (chemotherapy and radiation) with a view of followina BAIB excretion during the course of their disease. This approach might orovide an insiaht into the efficacy of dose levels of various chemotherapeutic agents and radiation levels used to check leukemia. In addition a relationship will he sought between urinary RAIB levels and the activity of the disease state with special emphasis on predicting the first indications of relapse during clinical therapy. This objective would continue the work initiated in our pilot stuay (ref. 2). (3) The Quantitation of Urinary Protein Amino Acid Levels in Cancer Patients The literature contains reports of abnormal patterns of urinary excretion of protein amino acids by cancer patients. Thus ref. 4 describes marked differences (using paner chromatography) in the excretion of methionine, threonine, valine, leucine, tyrosine, histidine and aspartic acid hetween normal ana leukemic patients. Another publication (ref 5) describes elevated urinary ae a eve Ate WETS eS -3- methionine (using paper chromatoaqraphy) in leukemic patients. Quantitation by GC of the urinary amino acid levels is difficult because of the number and complexity (overlanpins peaks) in the chromatogram. In addition gas chromatography is a non-specific method of detection (i.e. it detects a peak but it cannot identify what comnound(s) are present under the peak envelope). Apolying this laboratory’s recently developed technique of quadrupole mass fraqmentogranohy (ref 6% copy attached to the quantitation of urinary protein amino acids will enable us to measure these comoounds at the nanoqram level (refs 6,7). We intend to measure amino acid levels in a statistically Significant number of urines from normal and cancer patients. ile will then search for subtle changes in the quantitative levels of amino acids between the normals and the various tynes (section IIIA (€1)) of cancer investiqated. Experimentally our computer programs will have to be slightly modified so that a sincle analysis will suffice for the quantitation of BAI8 (section IIIA (2)) and the amino acids. IIIB. BACKGROUND The Instrumentation Research Laboratory was established in the Genetics Department under NASA auspices in 196]. Its task was to define and improve microanalytical methods for the detection of living processes that might be useful for the biolodical exploration of the planets. ‘any of the concepts that we explored have heen embodied in NASA’s planetary mission plans. However, we have not undertaken to desian and build hardware for such missions. Instead, we have served as experienced advisors to the experiment teams responsible for scientific studies on the Mariner and Viking Mars programs. Our work on GC/MS is one of several lines of instrumentation effort. Our original mandate from NASA included generous encouragement to seek health-related applications as a spin-off of the development work they were supoorting. However, they have not been able to support the full fledged extension of space-related technology to disease research per se. The oresent application also comes at a time when overall] funding for basic research by NASA is declining rapidly and may disappear within the next year. We have reduced our GC/MS laboratory staff in response to these cutbacks. It is therefore appropriate that we seek other supnort to help maintain this existing laboratory to apply its capabilities to problems of characaterizinga metabolites and their levels associated with carcinomatous disease. Qur focus on mass spectrometry (ref 8) originally stemmed from the exquisite sensitivity, speed, and specificity of this technique for the identification of oraanic molecules. ie have had some experience with instruments (like the Bendix Time-of-Flight ass Spectrometer) which can generate a complete low resolution mass’spectrum in 100 microseconds and whose sensitivity is limited by the statistics of the number of ionized fragments, and by the data handling problems of averaging -~4- repetitive spectra emerning at a rate of 10,N00 frames per second). tie have also been led to look at the computational challenges of AS from another standpoint - namely the mechanization of the scientific thinking that is entailed by the interpretation of 4 mass spectrum. This task has been the focus of the research in "artificial intelligence" of the DENORAL project. (See for instance ref. 9, 10.) The application of these instruments to human metabolic research requires another dimension, namely the separation of complex mixtures, 2.9., from hody fluids, into individual components. These are then availahle for identification by the mass spectrometer. Gas chromatogranhy has proven to be a useful companion to’ the mass spectrometer - the output gas stream can he fed directly to the inlet of the spectrometer and much of the carrier gas (helium) selectively daviated by a semi-permeable membrane. (Automated, continuous flow into a mass snectrometer of other chromatonraphic streams, ¢.9., hich pressure liquid chromatography, is a sneculation that may eventually materialize to great advantage but is not yet available for apnlications like ours). For some time then we have been developing the means to integrate GC with MS under computer management. The present project represents the systematic application of these skills to the recognition and identification of metabolic variations, viewed as clinical problems of cancer in man. The sample populations will comprise healthy (control) and patients with clinically diagnosed forms of cancer who are under treatment in either of the (1) Clinical Research Center$’ (2) Uroloay Clinic and (3) Stanford Children’s Hospital The Techniques of MS and GC The technique of mass spectrometry gives information about the structure of a molecular species by measuring the characteristic mass abundance pattern of fraaments resulting from ionizing the parent molecule. Ionization is- usually accomplished by electron bombardment. The compound under analysis must have a measurable vapor pressure at about 200 degrees C. (This temperature and a pressure of 0.0! microbars are the normal operating conditions of a GC=-coupled mass spectrometer ion source). The ionizina electron heam (70 eV energy) removes one electron from some of the molecules of the sample vapor to yield excited positive nolecular ions: M+ @ ~a=-->M(+) + 2e The molecular ion, M(+), is generally unstable (especially if at a high eneray of excitation) and may decompose within a few microseconds to yield a series of positively charged fraqment ions. Each fragment ion can in turn decompose to ions of lesser mass. These ions aré accelerated from the ion. source of the mass spectrometer into the analyzer region where they are separated according to their mass-to-charge ratio. In sector instruments ~5- Arr NO. NCI-CB-/4-29 this is done by. .a magnetic field, in quadrunole instruments by an electric field, and in time-of-flight instruments by an adjustable ion detection time delay. The mass spectrum of an oraanic compound thus consists of a table of positive ions of different masses and abundances. Molecular structure (for - instance number and location of heteroatoms) determines the frequency with which bonds will rupture subsequent to ionization, thereby producing, for the most part, a characteristic mass spectrum for each compound. Geometrical isomers may show subtle differences within their respective mass spectra owing to the influence of the geometry of neighboring grouos. Ontical enantiomers yield identical spectra. Although the technique of mass soectrometry was extensively used by petroleum chemists from the 1940%s, it was not widely utilized in organic chemistry until the late 1950%s. The first extensive monograoh on biochemical annvlications of mass spectrometry has just anpeared (ref 11). Our colleanues and close collaborators in the Stanford Chemistry Department, led by Professor Carl Djerassi, have been among the pioneers in the development of MS for natural product chemistry, esnecially as applied to steroids (4 books and in excess of 200 papers on various aspects of the theory and anplication of ‘5 have heen published by Prof. Djerassi’s qroun since 1961). During the 1960%s, mass soectronetry was annlied to many different tynes of oraanic comnootunds. The accumulation of these reference mass spectra was necessary to establish franmentation rules for the interoretation of unknown mass spectra. The exnerienced mass spectroscopist becones adept at recognizing the mass spectral Sianatures of those tynes of compounds with which he works hut he cannot encompass within his memory all the relevant information contained in the literature. In addition, many reference snectra determined by nass spectrometry laboratories have not been published. To overcone this orohlem, libraries of mass spectra are being compiled for computer storage and retrieval so that they will eventually be available for matching by computer against the mass soectra of unknown compounds. Proqress is now being made toward compiling libraries of mass spectra relevant to general metabolic studies. These will match the accumulation already available for soecial classes of organic molecules and for some druqs whose spectra are important for emergency toxicological analyses (ref 12). Instrumentation advances in mass spectrometry durina the past decade like improved sensitivity, direct counlina with GC and the use of computers for the routine recording and presentation of mass spectra, all facilitate the larae scale application of mass spectrometry to biomedical problems. Body fluids and other materials encountered in biomedical research are complex mixtures. For example, urine is Known to contain several hundred organic compounds at levels exceeding on the order of 1 nanogram per milliliter. The aas chromatoaqgraoh -is indispensable for the separation of such mixtures into discrete compontents. With medium resolution instruments, the mass ~é- RFP No. NCI-CB-74-29 soectrometer can be scanned repeatedly once every 2-4 seconds. The aas chromatograhic senaration of a urine mixture may require 40-50 minutes: the result is the accumulation of over 600 mass spectra per analysis. The simplest way to identify these mass spectra is to search a library of known compounds. Even if- the mass spectrum of a test compound does not reside in the library, the best match found may be a related compound. This can facilitate the manual interpretation hased on the chemist’s knowledge of and quesses about the rules of fraqmentation. The problem of comouterizing the identification of comoounds whose mass spectra are not in a library is addressed by the DENDHAL project (refs 9, 10). Computer programs have been developed to interpret mass spectra of unknown compounds from first principles (i.e., to emulate the reasoning processes of organic chemists). Frequently compounds of biochemical interest occur in small amounts in biological fluids. (Ry definition, many frontier problems concern compounds at the limit of easy detection hy existing techniques). Thus, the effectiveness of GC/#5 as a detector of biological materials is directly related to its sensitivity. Current systems routinely operate with sensitivities such that mixture components with as little as 50-3090 nanoaorams of material can be measured. This limitation is imoosed hy the following instrument-related factors. In order to record an interoretable mass spectrum, a low resolution mass spectrometer must have input to its source on the order of 5 nanocrans of ‘material per second. Since a aas chromatographic oeak lasts for approximately 5 to 30 seconds, in GC/MS operation some 25-159 nanograms oer GC peak are required for mass spectral analysis. Inherent in the gas chromatographic column and in the . semi~nermeable membrane separator (used to preferentially remove the helium carrier oas from the effluent stream) are losses of up to 50% which increase the input sample requirements in a oractical sense to about 50-300 nanograms of material oer GC peak analysis. Another limiting factor in the aoplication of the GC/#S technique to biological extracts concerns the volatility of the material to be assayed. Before the system can detect many non-volatile components (e.q., carbohydrates, amino acids, etc.), they must be converted into volatile derivatives which will pass through the gas chromatograoh at a maximum oven temperature of 300 degrees C. Above this temperature, column bleed fron the gas chromatographic phase will tend to enter the ion source and complicate the recorded spectra. Thus the GC/"“S techniague is restricted to those organic compounds which can be converted to volatile derivatives and is not, in ceneral, arplicable to inorganic compounds. A recent report (ref 13) describes the analysis by GC/MS of ketose diphosphates (as their trimethylsilyl (TMS) derivatives) while aldose diphosphates (also as their TMS derivatives) proved to he tno unstable to analyze. It is safe to assert, however, based on our own experience and the literature, that a broad soectrum of orqanic compounds of: biological significance will be amenable to analysis by GC/MS methods. -/|[- BAL NV) WL TUDE HTS? The routine screening of normal and abnormal hody metabolites, including drugs, in human body fluids (ref 14) is currently the object of several research programs. Various non-specific methods, including thin layer (ref 15,16), ion exchange (ref 17,18), liquid (ref 19), and gas chromatoqraohy (ref 20-23 and 24), are used primarily with the goal of separating a large number of unnamed constituent materials. Using these techniques, compound "identification" is made by a comparison of the nigration, under identical conditions, of the unknown spot with reference compounds. This anproach can lead to erroneous identifications, however. This point is illustrated by a recent article (ref 25) which describes the use of mass spectrometry in the identification of a case of isovaleric acidemia. Previously, the same patient was diagnosed as having butyric and hexanoic acidemia on the basis of chromatoarapnhic evidence alone. This type of error is especially important when analyzina a "new" (oreviously undescribed) metabolite where rigorous identification in various body fluids and tissues is essential. For positive identification using mass spectrometry, the separated components must he transferred from the chromatoqraphic medium to the mass spectrometer. The unknown spot can be leached from paper or thin layer chromatograms, or in the case of liquid chromatography, the solvent removed, and a mass spectrum recorded on the residual material. Trapping the various effluent components from a gas chromatoqraph, witn subsequent introduction “into the mass spectrometer, has been used. ThiS aporoach requires considerable time and is inefficient when anplied to complex separations. It has been superseded by the direct coupling of the gas chromatograph and mass spectrometer, Gas chromatograohy is unquestionably the most convenient separation technique to couple to the mass spectrometer because the carrier gas can be removed efficiently and easily as the analysis proceeds. For recent examples of the use of the GC/iK techniaue for the analysis of body fluids see refs 1,25-27. 8ased on the work cited, as well as our own on-qojing programs, the ability of the GC/MS technique for the analysis of body fluids is well established. ile have drawn upon the published literature in helping to design our experimental protocols (ref 1). III. B (2) MASS FRAGMENTOGRAPHY There is another mode of operation of a GC/MS system which enables greater sensitivities to be attained for the quantitation of KNOWN metabolites. If the mass spectrum and the aas chromatographic retention time of the compound to be quantitated are known, the mass spectrometer can he used as a specific detecting system for this compound. This technique is called mass fragmentography. Under these experinental conditions, the mass spectrometer is not scanned over the ENTIKE mass ranae but is directed to measure one or two SPECIFIC masses known to be characteristic of the compound(s) being quantitated. Consequently, there is an appreciable increase in sensitivity -3~ RFP No. NCI-CB-74-29 since the mass spectrometer samples only the significant data points and can intearate the signal longer. In this mode, existing GC/MS instrumentation matches the new fluorescent reagents for amines (reported to detect approximately 10 picomoles). It also embodies the snecificity of the mass spectrum at individual mass numbers. At greater cost and cumbersomeness, the WS can be extended to a quantum-counting range of sensitivity. These methods are therefore likely to he complementary to the special purpose methods, like fluorescimetry, which are often cheaper and more efficient for well defined classes of compounds. On the other hand, the history of pesticide analysis shows how the GC can also be made ultrasensitive at the cost of some loss of specificity. Usina deuterated analoas as standards for the test compound, quantitation can also be achieved at sub-nanogram levels. ie have recently exploited certain characteristics of the quadrupole mass spectrometer and its data system to develop a method for the quantitation of ten amino acids in soil extracts (ref 7, copy attached) and subsequently for the amino acid content of biological fluids (ref 6). This represents an advance in the technique of mass fraqmentocraphy since the sector mass spectrometers used up to this time, have been severely limited in the number of ions and the mass range they could monitor for any one experiment. Our technique of quadrupole mass fraqmentoography ‘was used for the quantitation of the amino acids in the urine of a patient with suspected branch chain amino aciduria (see ref. 6). As an example of the application of these methods to biomedical problems, we can use some recent studies we have undertaken on the urine of a patient suffering from acute lymphoblastic leukemia. The aas chromatoaranhic profile of the amino acid fraction of his urine showed the presence of an abnormal peak . The mass spectral analysis of this chromatographic peak identified this component as beta-amino isobutyric acid (BAIB) from a comparison with a literature spectrum of authentic m material (ref 28). The literature on human urinary BAIB excretion is extensive and it has heen reviewed in reference 29, Awapnara (ref 30) noted increased urinary BAIB excretion from patients receiving nitrogen mustard theraoy while Rubini (ref 31) reported hich levels of BAIB excretion in peoole exposed to excessive radiation levels. right and Fink (ref 32) found elevated urinary RAIB levels from mongoloid children and non-monaoloid mental defectives but these conclusions were subsequently disputed (ref 33). Harris in Enaland (ref 34) and Calchi-Novata in Italy (ref 35) found that 6-10% of a samole nonulation had elevated urine concentrations of BAIB while DeGrouchy and Sutton (ref 36) in a study of families of oriental extraction, and Yani, et al. (ref 37) with Japanese families, found about 40% of their subjects excreted elevated levels of BAIB. Gartler concluded that the variation underlyina BAIB excretion by New York families (ref 38), Apache Indians and Black Caribs of British Honduras (ref 39) was due to genetic ——— NS Eee SOU EE am ly difference at one locus. Later studies showed that thymine was probably a precursor of BAIB in man (ref 40). At least two authors (refs 29 amd ‘37) have questioned the accuracy of the methoas used for the detection and quantitation of BAIR in urine. The early work (refs 30 and 41), includina the first isolation of BAI8 from urine (ref 34a), was completed using paper chromatography and reaction with ninhydrin to develop the approoriate spot for detection and quantitation. As suggested (refs 29 and 37) other compounds in urine miaqrate with characteristics similar to BAIB thereby introducing an element of imprecision into the published quantitation results. In those instances where BAIB is present in only a small amount, or for that matter in larae amounts, it is always possible that another metabolite could be present within the confines of the "PAIB spot", Subsequent methods employed for BAI8 quantitation include thin layer chromatoaraphy (ref 42 ) and paper electrophoresis (ref 43). These methods must also he considered as non-specific for the estimation of BAIB in urine. Our own procedure, must be considered far sunerior to these other techniques for hoth the detection and quantitation of PAIB in urine because it combines the quantitative aspects of gas chromatoagraohy with the positive identification of a specific ion detector. se have observed several instances in which the gas chromatographic peak associated with BAIB was shown by mass spectrometry to contain other components. Using the technigue of mass fraqnentogranhy (refs 6 and 7) (conies attached), we can quantitate for BAIR under these conditions. In this techniaue, a characteristic fragment ion from the mass spectrum of BAIB is summed over the total ion chromatogram to give an area corresnonding to the amount of 8AIB oresent. Usina an internal standard and a characteristic ion from its mass spectrum and knowing the ionization efficiency of both FAIB and the internal standard, quantitation can be achieved. For a description of this work see ref 2 (copy attached). . In a pilot proaram we have examined urine samoles from over 30 patients suffering from leukemia and have observed various levels of BAIB in over 50% of these individuals. One oatient was excreting 1.2 grams of RAIE per day and immediately following druq therapy this compound could no longer be detected nor was it detected in a urine sample collected 3 months later. Awcpara (ref 30) claimed that all leukemic patients excrete varying amounts of BAIB which was not detected in normal urine. Killmann, et al. (ref 44) noted that two out of three patients with chronic granulocytic leukenia produced hich levels of BAIB but followina drug therapy, the level returned to normal. Lee, et al. (ref 45) examined the urine of 33 patients with leukemia receivina specific drug therapy and failed to observe PAIB excretion in any of these subjects. Our own observations, in agreement with the literature (refs 44-46), suagest that during the disease cycle, leukemic patients may excrete large amounts of BAI8 but this ceases following chemotherapy. Schreier (ref 41) studied the urinary excretion of BAIB in leukemic subjects and noted extremely variable results with "massive amounts" of BAIB present 10 APL WO. WOLTUDT/ 4-27 in the terminal stages of the disease. III. C. EXPERIMENTAL We use the following procedures to fractionate urine, blood, and amniotic fluid in preparation for their analysis by GC/iiS. In the case of plasma, the pvrotein is first precipitated (by the addition of ethanol) and the supernatant liquid dried IN VACUO and processed as for urine. Amniotic fluid is treated in the same way as urine. - URINE (pH = 1, internal standards added) ' § ' : + continuous ether extraction (18 hrs) v v v . organic phase aqueous phase v Vv (free acids) conn wrest enn +T and neutrals VY Y ‘ : A (carbohydrates) Camino acids) + HCl optional hydrolysis Cc B * and continuous ether i extraction Vv ee T Vv Vv organic phase aqueous phase vy , (hydrolyzed acids) Camino acids) D . E The experimental procedure used for workina with a fluid sample is as: follows. To an aliquot (2.5 ml.) of fluid is added 6N hydrochloric acid until the pH is I. Internal standards, n-tetracosane, 2-amino octanoic acid and trehalose are then added. Continunus ether extraction isolates the free acids (fraction A) which are then methylated and analyzed by GC/¥S. An aliquot of the aqueous phase (1.0 ml) is passed through an ion exchange resin (Dowex 50 H+) and the column washed with water (15 ml) to elute the carbohydrate fraction (C). This column is then washed with 3N aqueous ammonia to elute the amino acids and any basic compounds present (fraction 3). Fractions F and © are then concentrated to dryness and 8 derivatized by reaction with n-butanol/hydrochloric acid followed hy methylene chloride containing trifluoroacetic anhydride. This procedure derivatizes any amino acids (or water soluble amines) which are then sub jected to GC/MS analysis (fraction B). The carbohydrate fraction (C) is then derivatized by reaction at room temperature with Tri-Sil-Z. AS an additional option, one can hydrolyse portions of the urine to free those metabolites present in conjugated form, and one BO SN 8 SN RON rE -~|I- this is accomplished as follows: Concentrated hydrochloric acid (0.15 ml) is added to the urine (1.5 ml) after ether extraction and the mixture hydrolyzed for 4 hours under reflux. Ethyl acetate extraction then separates the hydrolyzed acid fraction (D) which is methylated and analyzed by GC/MS. A portion of the aqueous phase (0.5 ml), from hydrolysis of the urine, is concentrated to dryness and derivatized and analyzed for amino acids (Fraction E). tle do not propose to hydrolyze all the urine samples screened but rather to limit this step to those samples having an indication of the presence of unusual metabolites. In these instances it will be of interest to hydrolyze te urine to determine whether the metaholite of interest is also present in conjugated form. The above scheme represents a general method for preparing body fluids for analysis. The conditions required for yas chromatooraphic separation vary with different classes of - compounds. For instance, using our method for chromatogranhing derivatized free acids, amino acids, and carbohydrates, steroids will not be detected . Large molecular weiaht compounds (e.Ge, tripentides and mucopolysaccharide fragments) cannot be analyzed by this system without their degradation to smaller unit molecules. Also small neutral molecules (the so called "urinary volatiles") will be mostly lost during our isolation techniques. A suitable computer data system can significantly ease the tedious burden of analyzing the large amounts of data emanating from GC/MS instrumentation. The computer assists in instrument set-up and calibration, data collection and filina, and data reduction and analysis. je have developed considerasle exnerience with such data systems, both for low and high resolution instruments, since our first-desian was described in 1966 (ref 47, copy attached). Our most recent efforts have focussed on facilitating user interaction with the data system on automatina various aspects of data analysis (ref 8) and on establishing a library search system for the automatic identification of metabolites from their recorded mass spectra. At the end of July 1973 the NIH financial support ceased for the computer facility (ACME system of the Stanford University Medical School) to which our mass spectrometer was interfaced. At this time we commenced using a PDP-11/20 based mini-computer system (16K of core) and containing the following peripherals: disk storage (256K words), DEC tape drive for the recording of mass spectra, a line printer and a floating head disk (1.2 3s words) for the supoort of a library of mass spectra and the associated search routines. This system has the capacity to record on DEC tape those mass spectra recorded during the lifetime Canproximately 40 minutes in our procedure) of a GC/MS analysis of one fraction from a urine sample. This system also operates the mass spectrometer in the mass fragmentograohy analytical procedure. The automatic data reduction phase of our previous system (ref. 63 copy attached) is currently being coded to operate on the PDP -12- AGE NO. NUi~UCDb-/4~-29 11720 system. It is anticipated that the recoding of these programs will oe completed by Apvril 1, 1974. During the lifetine of this proposal, effort will be applied to improving the routines previously used. To this effect we have hudgeted 20% time for an experienced programmer to accomplish this task and to maintain existing software. Our object in screening uring samoles is to identify as broad a suite as nossible of those metabolites present. This results in a relatively low daily throuchout of samples but this is balanced by the detailed scrutiny each fraction receives. For instance our procedure fractionates urine into the following qeneral metabolic classes? (i) free acids and neutrals (isolated by solvent extraction) (ii) amino acids and bases (iii) sugars. A GC/MS analysis after aopropriate derivatization of one of these fractions occupies about 40 minutes hut with the tine required for preparation of the analytical conditions this relates to a sample through put of one per hour. In devotina 50% time to the screenina of urine an orcanic chemist would have a throuthput of one urine sample per day (time is required to nlot back data stored in the data system and for the interoretation of the results). Some additional efficiency would be venerated by limiting the GC/NS analysis to those urinary fractions disolaying anomalous peaks after a pre-screening analysis throuch an existing dual column qas chromtoaraph. The incresed efficiency might allow between 5 and 7 urine samples to he analyzed per week with each analysis consisting of the three metabolic classes shown at the bottom of the previous paracraph. This method for the screening of urine allows a broad range of significnt compounds to be analyzed but only quantitative estimates of their concentration (use of internal standards) to be presented. ore precise quantitation will be obtained from GC using the relative FID sensitivities of the unknown and internal standard (FID = flame ionization detector). This should be contrasted to the situation pertaining to mass fragmentoarachy where a more limited range of metabolites are measured very precisely. This data system will allow the necessary sophistication for the processing of body fluid samples to establish norms and to investigate specific urinary metaholic content for various forms of clinically diagnosed cancer. This involves extracting from the approximately 600 -snectra collected during each run, the 100 or so representing the components of the body fluid samole for identification. The raw spectra may be contaminated with background "column bleed" and some are composited with adjacent constituent spectra’unresolved by the GC. ~ {3- KEE NOU. WELT UDT ISLS We have begun to develop a solution to the problen of effectively increasing the resolution of the GC by comouter analysis of the data. These prosrams will allow us to automaticaly locate the body fluid constituent spectra and remove the distortions caused by hackground and peor resolution. These "“cleaned-uo" spectra can then be analyzed by library search techniques or first principles as necessary. By using a disk-oriented matrix transposition algorithm develoned for image processing applications, we can rotate the entire array of 700 spectra by 500 mass samples for each run. In this way we can gain convenient access to the "mass chromatocram" form of the data. This form of the data, displayed at a few selected masses, is used in mass fragmentography described elsewhere in this proposal. Mass chromatograms have the important property of displaying much higher resolution in localizina GC effluent constituents. The automatic analysis techniques currently being develoned for mass fragmentogqrans can be extended to this more general case. Thus hy transposing the raw data to the mass chromatoqram domain, we can systematically analyze these data for baselines, peak positions, and correct amplitudes thereahy deriving idealized mass spectra for the constituent naterials. These spectra are free from backaround contamination and influences of adjacent unresolved GC peaks. This aspect of our work is funded by NASA (see Section VI. Cytochemical Studies of Microorganisms.), but on implementation it will be available to this proposal The problem of spectrum idantification is addressed by using rapid library search techniques for the identification of previously encountered compounds. Those not in the library will be identified by the manual interoretation of the spectrum using other information as available. These results will be incorporated into the library to extend its domain of usefulness. This progressive compilation of a library of biomedically relevant compounds will sneed the throughout of the system. This library will be freely shared with other investicators. Eventually the extension of the library domain will be assisted by adapting the computer programs under development in the DENDRAL project (see for instance refs. 9 and 10). These seek to emulate the reasoning processes organic chemists employ in interpreting spectra using fraqnentation rules and other knowledge to infer the correct molecular structure from among those possible. The compilation of mass spectra is currently well underway and should be operational by June 1974. we are in contact with Dr. S. “Markey, University of Colorado iiedical Center, Denver, who is compiling a library of the mass spectra of biologically significant compounds. This laboratory has contributed mass soectra to this compilation and we will receive a copy of the final library. ITIC (2) and (3) Quantitation of Urinary Beta-Aminoiso-butyric Acid and Protein Amino Acids using Quadrupole Mass Fragmentoaraphy. IED NV we NNT eS -|4- In the preceding section we briefly touched upon the capability our laboratory has generated for the quantitation of BAIB and amino acids at the nanogram level using quadrupole mass fragmentoaraphy. In the case of the amino acids a deuterated amino acid mixture serves as an internal standard. %y exploiting characteristics of the quadrupole mass spectrometer~computer system we can quantitate in a single analysis many more components than previously was possible using mass fragmentography. The quantitation of urinary BAIB and amino acid levels are conveniently discussed togther since experimentally one analysis will suffice to quantitate both. The fractionation procedure for urine described in section IIIC (1 ) provides a derivatized fraction (B) consisting of a carboxybutyl, N-trifluoroacetate derivative which our procedures use for the quantitation of BAIB and protein aminoacids. This quantitation will as we have already demonstrated be at the nanoaram level of detection (refs. 2, 6, 7, 48, 49, copies attached). The PDP-!! computer system will calculate the concentrations of the following amino acids: BAIR, alanine, valine, alycine,isoleucine, leucine, proline, threonine, serine, phenylalanine, aspartic acid, glutamic acid, tyrosine and lysine present in the urine of control and cancer oatients. Statistical correlations will then be sought between the levels of these metabolites and the respective disease states of the patients. Qur publications (refs. 6,7) describing the mass | fragmentographic quantitation of amino acids did not measure tyrosine because we lacked deuterated tyrosine to serve as an internal standard. vie now have this deuterated amino acid and routinely analyze for tyrosine. One analysis using mass fragmentongraphy will be completed every hour. The data acquisition phase of the experiment occupies about 30 minutes, the computation and presentation of the results should occupy another 15 minutes and 15 minutes are allowed for resetting various experimental parameters (sample oreparation, resettina GC oven conditions, etc.). ie believe that three samples per day will be analyzed (50% time of one organic chemist for a weekly output of 15 samples 750 on an annual basis). GENERAL COMMENTS? An imoortant aspect of our experimental protocols will involve the clinical study of the efficacy of metabolites or their concentration levels we may identify for the early detection of cancer. This aspect of our program will be addressed by screening a large number of urines from healthy subjects and clinically following those who register positive in our tests. Qur clinical associates at Stanford see a large number of patients during their professional treatment of cancer and we expect to have ample control subjects and natients for the testing of any clinical applications resulting from this research project. At this tine we have not formalized the protocols which might be used in such a clinical study and prefer to wait until - the research develops results applicable to a detailed clinical RFP No. NCI-CB-/4-29 ~15- study. ‘ IV. KEY PERSONNEL J. Lederberg Prof. of Genetics & 5% Principal Investigator A Duffield Research Associate 50% WN. Pereira Research Associate 50% R. Tucker Computer Programmer 20% E. Steed Research Associate 10% N. Veizades Research Engineer 10% M. ilyche Research Assistant 30% M. Allan Secretary 10% J. Lederberg, Princioval Investicator Professor Lederbera is the Joseph D. Grant Professor of Genetics and assumes overall responsibility for the research program outlined in this proposal. His oriainal interest in the technique of GC/MS involved its use in exobiology for the detection of extra-terrestrial life. His present interest in GC/MS resides in its anplication to the analysis of the metabolic content of hody fluids. He is the orincipal investinator of a comprehensive grant to start a Genetics kesearch Center, one project of which is the use of GC/‘S to analyze the metabolic content of body fluids for the recoonition of genetic disease. Dr. Alan Ouffield will assume responsibility for the integrity of the urine analyses. His experience is in oraanic chemistry and the applications of mass spectrometry to nroolems in organic chemistry and biochemistry. Since 1970 ne has been active in the aoolications of GC/MS to biochemical and hiomedical oroblems. This has involved research in the analysis of various body fluids for metabolic content primarily with a view to recognizing genetic disease. In the oresent proposal he will conduct those experiments desianed to screen urine from cancer patients for snecific netabolites which may he idiosyncratic with the disease. Dr. Wilfred Pereira received his Ph.D. in pharmaceutical chemistry and has had exnerience in the analysis of body fluids by GC/MS. de developed the chemistry which led to the development of quadrupole mass fraqmentograohy as a sensitive, specific and quantitative single analysis for trace quantities of over le organic compounds. In the present proposal Or. Pereira will use quadrupole mass franmentooraphy for the quantitation of beta-amnino~isobutyric acid and orotein amino acids in the urines of control subjects and cancer patients. Mre Robert Tucker implemented and maintains the software used for the PDP 11/20 mini~comnuter system. He is responsible for the continuing develooment of the software requied for automatic reduction of mass fragqmentography data. Mr. Ernest Steed is responsible for the maintenance of the GC/MS system. The transfer lines (GC column, separator) in this -|6- RFP No. NCI~CB~/4-29 i system are all glass (the membrane separator is made by "r. Steed, who is an experienced glass blower). He also assumes responsibility for the maintenance of the mass spectrometer vacuum and electrical systems. Mr. Nicholas Veizades will be required to maintain the interface between the mass spectrometer and the PDP-11/720 computer system. He will also be responsible for some maintenance of the data system. Mrs. dyche will support the chemical portion of this program. She assists in routine laboratory functions, oreparation of standard solutions and data presentation from the mini~computer system. V. FACILITIES AND EQUIPMENT. ie will derive our clinical urine samples from Dr. @. R. Fair (Associate Professor of Surcoerys: see attached letter, Appendix A) and from Drs. J. FR. viilbur and T. Long, Stanford Children’s Hospital (see attached letter, Appendix 8). Analyses will be oerformed on existing GC and MS equipment in the Department of Genetics. iie have a two-column Varian Aerogradch 2100 aas chromtograph used for pre-screening and a Varian Aeronraph 1200 aas chromatograph interfaced to a Finnigan 1015 Quadrupole mass spectrometer by an all glass membrane separator of our own design. The mass spectrometer is interfaced-to a PDP=11/290 mini-conouter system (16 core) consisting of fixed head disk storage (255K) movable head disk storage (1.2 nillion) line printer, and for data storage, a DEC tane drive. jie have access to a Varian WAT 711 high resolution mass spectrometer (interfaced to a Hewlett-Packard 7610 gas chromatograph) in the Chemistry Department for the recording of HRMS data in those instances where unknown metabolites cannot be uniquely identified from GC/LRAS data alone. Also available to the project are the electronic facilities and the software exoerience of the Instrumentation Research Laboratory (Dr. &. Levinthal, UVirector) of the Department of Genetics. - VI. RELATED ACTIVITIES? Cytochemical Studies of Planetary Microorganisms. Explorations in Exobioloqy. (J. Lederberg, Principal Investigator). NASA Grant NGR-05-020-004. This grant sponsored the activities of the Instrumentation Research Laboratory of the Department of Genetics since 1962 and these are summarized in ref. &. The initial focus of this laboratory was to develop methodologies for the detection of extraterrestrial life and this led to the development of a sophisticated GC/MS-computer system (ref. 47). During the lifetime of this grant NASA has generously fostered anpolications of these methodologies for health related research. tiith recent budgetary restrictions which necessitated that NASA reduce its sponsorship of basic research we anproached other granting ADL WO. NULTUDT/4—-29 -|7- agencies in the field of public health for the financial support necessary to continue to apply advanced chemical methodologies to health related research. Analytical Methodology for Biochemical Monitoring. (J. Lederberg,Principal Investigator). NASA NGR~05-020-632, Under this grant several pilot projects were commenced aimed at devising non-invasive techniques for the diagnosis of human disease. An important aspect of this grant was its willingness to sponsor the software develooment required for the PDP-11/20 mini-computer system to be interfaced to our mass spectrometer. It was under this umbrella that our investigation of the metabolites present in urine from patients with leukemia and Hodgkin’s Disease were commenced. During the sponsorship of this grant we have aoplied mass fraqnentography to the analysis of body fluids. Another investination partly generated by this qrant was an investigation of the urinary metabolic content of prenature infants following controlled dietary intake. An early draft of a manuscript describing this research is attached to this application (see Appendix C). Genetics Research Center. (J. Lederbera, Principal Investigator) NIGMS Approved and awaiting funding. (Grant #PQ I-GM-—20832-01) The focus of this grant is the detection of clinical indicators of genetic disease. One section of the proposal deals with GC/MS methodology in the analysis of body fluids for the identification of metabolites which might be indicative of genetic disease. It is intended that this will not be a broad Screening program but rather that its emphasis resides in analyzing body fluids from those patients where there are clinical symtoms indicative of possible genetic abnormalities. Another application within this orant of GC/MS methodology is the analysis of amniotic fluid for prenatal diagnosis of hereditary inborn errors of metabolism. VIT. PROGRAM MANAGEMENT. The proposed program, because of its limited size and sharply focussed arena of anpolication does not require special management protocols to be drawn up. REFERENCES 1. Jellum, E. Stokke, 0., and Eldjarn, L., "Combined Use of Gas Chromatography, Mass Spectrometry?, and Computer in Diagnosis and Studies of Metabolic Disorders," Clin. Chem., 18, 0. soo, (1972). 2. Pereira, We. E., Summons, R. Es, Reynolds, ve Es, Rindfleisch, T. C.,and Duffield, A. Me, "The Quantitation of Beta~Amino-isooutyric Acid in Urine by Mass Fragmentography," Clin. Chim. Acta, in press (1974). BNA SL Ae AN I eS ~1[8= 3. Abe, Me, Takahashi, M., Nishidai, T., Suyama, S., Oshima, Se, and Teramatsu, T., "The Significance of Urinary Beta~Aminoisobutyric Acid in Cancer Patients," Int. J. Radiat. Biol., 24, 73 (1973). 4. Lee, M. B., Neill, D. We, and Bridges, J. “M., "Urinary Amino Acid Execretion in Subjects with Leukemia," J. Clin. Path., 20, 70 (1967). 5. Lee, Me. Be, Bolaoer, C. D., and Bridges, J. %., "Distribution of Amino Acids in Urine, Plasma, Leucocytes and Erythrocytes of Leukemic and Normal Subjects," Acta Haemat., 42, 86 (1969). 6. Summons, R. E., Pereira, fe E~, Reynolds, ji. E., Rindfleisch, T. C., and Duffield, A. “., "Analysis of Twelve Amino Acids in Biological Fluids by Mass Fraqgmentoygraphy Using a Quadrupole Mass Spectrometer~Computer System," Anal. Chem., 46, OOO (1974). 7. Pereira, il. E.~, Hoyano, Y., Reynolds, ti. E~, Summons, R. E., and Duffield, A. Me, "The Simultaneous Quantitation of Ten Amino Acids in Soil by Mass Fragmentography," Anal. Biochem., 55, pe236 (1973). 8. "Cytochemical Studies of Planetary Microoryanisms - Explorations in Exobiology," Annual reports from 1962 to the ‘present, NASA Grant NsG 81-60. 9. Smith, D.H., Buchanan, B.G., Englemore, R.S., Duffield, AwMe, Yeo, Ae, Feigenbaum, E.A., Lederhern, J., and Djerassi, C., "Applications of Artificial Intelliaence for Chenical Inference VII. An Approach to the Computer Interoretation of the High Resolution Mass Spectra of Comolex Molecules: Structure Elucidation of Estrogenic Steroids," J. Amer. Chem. Soc., 94, DP. 5962, (1972). 10. Smith, D. H., Buchanan, 8B. G., Englemore, R. S., Adlercreutz, H., Djerassi, C., “Applications of Artificial Intelligence for Chemical Inference. IX. Analysis of itixtures without Prior Separation as [Illustrated for Estrogens. J. Amer. Chem. Sace, 95, 6078 (1973). 11. "Biochemical Anplications of Mass Spectronetry," Edited by G. R. Waller, Wiley-Interscience, New York, 1972. 12. Biemann, K., Althaus, J.R., Biller, J, Donaghue, P.F., Evans, DeAc, Forster, He-Je, Hertz, H.S., Hianite, C.F., MUYORY, R.C., Preti, Ge, and Reinhold, V., Exnerientia, 26, Page 714, (1970). 13. Harvey, D. J., and Horning, M.G., "Characterization of the Trimethyl-Silyl-Derivatives of Sugar Phosphates and Related Compounds by Gas Chromatography-Mass Spectrometry,” J. Chromat., 76, Page 5! (1973). ~19- RFP No. NCI~CB-74-29 14. Schwartz, M.K., "Biochemical Analysis," Anal. Chem., 44, p. OR, (1972). 15. Heathcote, J.G., Davies, D.M., and Haworth, C., "The Effect of Desalting on the Determination of Amino Acids in Urine by Thin Layer Chromatograohy," Clin. Chim. Acta, 32, p. 457, C1971). 16. Davidow, 8., Petri, N.L., and Quame, B., "A Thin Layer Chromatographic Screening Procedure for Detecting Drug Abuse," Amer. J. Clin. Pathol., 50, op 714, (1968). 17. Efron, K. and wolf, P.L., "Accelerated Single-column Procedure for Automated iweasurement of Amino Acids in Physiological Fluids," Clin. Chem., 18, p 621, (1972). 18. Wwilson-Pitt, W., Scott, C.D., Johnson, i.F., and Jones, Ge, "A Bench-ton, Autometed, Hich-resolution Analyzer for Ultraviolet Absorhing Constituents of Rody Fluids," Clin. Chen., 16, pop. 657, (1970). 19, Burtis, C.A., "The Separation of the Ultraviolet-absorbinga Constituents of Urine by High Pressure Liquid Chromatography," J. Chromatog., 52, p 97, (1970). 20. Dalgliesh, C.E., Horning, &.C., Horning, {.G., Knose, -KeL., and Yaroer, Ke, "A Gas-liquid Chromatoaqraphic rrocedure for Separating a Wide Range of Metabolites Occurring in Urine or Tissue Extracts," Biochem. J., IDO), pe. 792, (1966). 21. Teranishi, Re, Mon, T.R., Robinson, AeRe, Cary, Pe, and Pauling, L., "Gas Chronatoaradnhy of Volatiles from Preath and Urine," Anal. Chem., 44, p. 18, (1972). 22. Pauling, L., Robinson, A.3., Teranishi, Re, and Cary, P., "Quantitative Analysis of Urine Vapor and Preath by Gas-liquid Partition Chromatography," Prec. Nat. Acad. Sci. USA, 68, p. 2374, (1971). - 23. Zlatkis, A. and Liebich, H.i., "Profile of Volatile Netabolites in Human Urine," Clin. Chem., 17, pe 592, CI9TI). 24. Jenden, D.J. and Cho, AwKe, “Applications of Intenrated Gas Chromatoocranhy/iiass Snectrometry in Pharmacology and Toxicology," Ann. Rev. of Pharmacology, 13, Pace 371, (1973). 25. Ando, T., Nyhan, ‘i.e Le, Bachmann, C., Rasmussen, Ke, Scott, R., and Smith, E. K., "Isovaleric Acidemia: Identification of Isovalerate, Isovalerylolycine and 3-Hydroxyisovalerate in Urine of a Patient Previously Reported as Having Butyric and Hexanoic Acidemia," J. Pediatrics, 82, 243 (1973). 25.6 Mrochek, Joke, Butts, W.C., Rainey, TheTes and Burtis, C.Ae, "Separation and Identification of Urinary Constituents by Use of Multiple~analytical Techniques," Clin. Chem., 17, 0. 72, _20- RFP No. NCI~CB-74-29 C1971). 27. Horning, E.C. and Hornina, h.Ge, “Human ‘Metabolic Profiles Obtained by GC and GC/MS," J. Chromatog. Scis, 9, Pp. 129, (1971). 28. Lawless, J.G. and Chadha, M.S., "Mass Spectral Analysis of C(3) and C(4) Aliphatic Amino Acid Derivatives," Anal. Biochem., 44, pe. 473, (1971). 29. Sutton, H. E., "BAIBuria in Metabolic Basis of Inherited Disease," New York, McGraw-Hill, 1960, pp.792-806. 30. Awapara, J., "The Leukemias," International Symposium (Detroit) New York, Academic Press, 1956, p. 353. 31. Rubini, J. Re, Cronkite, E. P., Bond, V., and rleednir, T. Me, "Urinary Excretion of BAI3 in Irradiated Hunan Reincs.”" Proc. Soc. Exper. Biol. and Wed., 100, Page 130 (1959). 32. wirignt, S. We and Fink, K., "The Excretion of PAIR in Normal, Monaoloid and Non-Mongoloid Mentally Defective Children," Am. J. Ment. Deficiency, 61, Page 530 (1957). 33. Berry, He. K., "Individual Metabolism Patterns: II. Excretion of BAIB." Metabolism, 9, Page 373 (1960). 34. Harris, He, “Family Studies on the Urinary Excretion of BAIB," Ann. Eugenics, 18, Page 43 (1953). 34a. Crumpler, H. R., Dent, C. E., Harris, H., and sestall, R. Ge, "BAIB, A New Amino Acid oftained from Human Urine," Nature, 167, page 307 (1951). 35. Calchni-Novati, C., Cepellini, R., Biancho, I., Silvestroni, E. and Harris, de, "BAIB Excretion in Urine. A Family Study in an Italian Population," Ann. Eugenics, 16, Page 335 (1954). : 36. DeGrouchy, J. and Sutton, H. E., "A Genetic Study of BAIB Excretion," Am. J. Human Genetics, 9, Page 76 (1957). 37. Yanai, Je, Kakimnoto, Y., Tsujio, T., and Sano, I., Am. J. Human Genetics, 21, Page 115 (1969). 38. Gartler, S. Me, "A Family Study of BAIB Excretion," An. J. Human Genetics 8, Page 120 (1956). 39. Gartler, S. M., Firschein, I. Le, and Kraus, B. S., "An Investigation into the Genetics and Racial Variation of BAIB Excretion," Am. J. Human Genetics, 9, Page 200 (1957). 40. Gartler, S. “%., "A Metabolic Investiaation of Urinary BAI8B Excretion in Man, Arch. Riochem. 80, 400 (1959). -2I- Rrr NO. NOL-CB—-/4-29 41. Schreier, "The Behavior of Amino Acids in Body Fluids During Develonoment and Growth: Physiology and Pathology in Amino Acid Pools," J.T. Holden, Editor, Elsevier, New York, Paae 276, (1962). 42. Goedde, H.ii. and Prunschede, H., "BAIB: A TLC Method for the Quantitative Estimation in Human Urine," Clin. Chim Acta, I[1, page 485, (1965). 43. Fessas, P.H., Koniavitis, Ae, and Zeis, P.eiie, Je Clin. Pathe, 22, Page 154, (1969). 44, Killmann, S.A., Rubini, J.R., Cronkite, E.P., and Bond, V.Pe, “Urinary Excretion of Beta-Amino Isobutyric Acid (BAIBA) in Leukemia and Polycythemia Vera," Acta Haemat., 25, Page &l, (1961). 45. Lee, MeBey Neill, DwaWe, and Bridges, Jales "Urinary Amino Acid Excretion in Subjects with Leukemia," J. Clin. Path., 20, Page 70, (1967). 46. Fink, K., Henderson, R. R., and Fink, R. A-, "BAIBt A Possible Factor in Pyrimidine Metabolism," Proc. Soc. Exptl. Biol. Med., 78, Page 135 (1951). 47. Reynolds, #W.E., Sacon, V.A., Bridges, J.C., Coburn, T.C., Halpern, B., Lederberg, J., Levinthal, E.C., Steed, &., and Tucker, R.Be, "A Computer Operated ifass Spectrometer System," Anal. Chem., 42, Page I122, (1970). 48. Pereira, ww. E., Summons, R. E., Rindfleisch, T. C., and Duffield, A. Me, "The Determination of Ethanol in Blood and Urine by Mass Fragmentography," Clin. Chim. Acta in press (1974). 49. Pereira, A. E., Summons, k. E., Rindfleisch, T. C., Duffield, A. Me, Zeitman, B., and Lawless, J. G., "Stable Isotope Mass Fragmentography: Quantitation and Hydrocen-Deuterium Exchange Studies of Eight Murchison Meteorite Amino Acids," Geochem. et Cosmochem. Acta, submitted for publication.