The Search for Mutagens from Cooked Foods

Nonprofit O

rg.U

. S. P

ostage

PAID

Livermore, C

a.P

ermit N

o. 154

Science and Technology R

eviewLaw

rence Livermore N

ational LaboratoryP.O

.Box 808, L-664

Livermore, C

alifornia 94551Inside this issue:

The Search for Mutagens from Cooked Foods

July 1995

Lawrence

Livermore

National

Laboratory

Printed onrecycledpaper.

The Search for Mutagensfrom Cooked Foods

SCIENTIFIC EDITOR

William A. Bookless

PUBLICATION EDITOR

Dean Wheatcraft

WRITERS

Lauren de Vore, Robert D. Kirvel, andDale Sprouse

ART DIRECTOR

Ray Marazzi

DESIGNERS

George Kitrinos and Ray Marazzi

GRAPHIC ARTIST

Treva Carey

CONTRIBUTING ARTIST

John Maduell

COMPOSITOR

Louisa Cardoza

PROOFREADER

Catherine M. Williams

2 S&TR’s Premier Issue

3 Commentary on Science and Technology

4 The Laboratory in the News

FeaturesFood MutagensPotent mutagens are found in cooked foods, such as fried beef. This report, the first of two installments, describes our efforts to identify and measure the abundance of more than a dozen potential cancer-causing byproducts in a variety of foods characteristic of the Western diet.

6 The Role of Cooked Food in Genetic Changes11 The Challenge of Identification18 The Cooking Makes a Difference

Research Highlights26 This Hybrid Vehicle Uses Hydrogen28 Modeling for More Accurate Weather Forecasts

31 Patents and Awards

32 Abstract

S&TR Staff July 1995

LawrenceLivermoreNationalLaboratory

Printed in the United States of America

Available fromNational Technical Information ServiceU.S. Department of Commerce5285 Port Royal RoadSpringfield, Virginia 22161

UCRL-52000-95-7Distribution Category UC-700July 1995

This publication is a continuation of Energy andTechnology Review. The May and June issueswere not published.

Page 28

Page 6

Page 26

R E V I E W

About S&TR

About the Cover

This inaugural issue of Science andTechnology Review features the first of twoinstallments reporting the results of 17 years ofLaboratory research on the identification of toxicmutagenic compounds in cooked foods. For thecover, George Kitrinos of the S&TR design staffand photographer Jim Stoots created a collage ofthe foods common in the Western diet that werethe focus of this research. The story begins on p. 6.

The Lawrence Livermore National Laboratory, operated by the University of California for theUnited States Department of Energy, was established in 1952 to do research on nuclear weapons andmagnetic fusion energy. Science and Technology Review (formerly Energy and Technology Review) ispublished monthly to communicate, to a broad audience, the Laboratory’s scientific and technologicalaccomplishments, particularly in the Laboratory’s core mission areas—global security, energy and theenvironment, and bioscience and biotechnology. Rather than just inform people of theseaccomplishments, the publication’s goal is to help readers understand them and appreciate their value tothe individual citizen, the nation, and the world.

Please address any correspondence concerning S&TR (including name and address changes) to Mail Stop L-664, Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, CA 94551, ortelephone (510) 422-8961. S&TR is also available on the Internet at http://www.llnl.gov/str/str.html, andour electronic mail address is [email protected].

Prepared by LLNL under contractNo. W-7405-Eng-48

Electronic Access

M

O

CR

F

T N

PR

N F

GY

M EE O

ERAE

D

T

D

ITU

NE

A

EI

A

AST STE 1• 8 •68

LET

THE

E B

R

E

L IGH T

TH

EU

NIV

ER S I T Y O F C A

LIF

OR

NIA

Cov

er p

hoto

: Jam

es E

. Sto

ots

S&TR is available on the Internet athttp://www.llnl.gov/str/str.html. As referencesbecome available on the Internet, they will beinteractively linked to the footnote references atthe end of each article. If you desire moredetailed information about an article, click onany reference that is in color at the end of thearticle, and you will connect automatically withthe reference. Available after August 1.

ELCOME to the first issue of Science and TechnologyReview (S&TR), Lawrence Livermore National

Laboratory’s monthly magazine. Part of our responsibility as a national laboratory is to inform the country about ourwork and accomplishments as we address science andtechnology problems of national importance. As many of ourreaders are aware, we published S&TR’s predecessor Energyand Technology Review (E&TR) for two decades.

Over the years, E&TR evolved as we made changes informat, design, and style to enhance our ability to explain theLaboratory’s work to an audience with, frequently, a morethan basic scientific background. S&TR is a natural outgrowthof this evolutionary process. Upon the 20th anniversary ofE&TR, we thought it fitting to examine the publication onceagain to see what could be done to make it still more effective.We confirmed that the journal reflected the wide variety ofscience and technology being explored at the Laboratory.

As part of this review, we prepared a mission statement that reflects the essence of the new publication’s goal:

The purpose of S&TR is to communicate, to abroad audience, the Laboratory’s scientific andtechnological accomplishments, particularly inthe core mission areas of the Laboratory. Thepublication does more than just inform people ofthese accomplishments: it helps readersunderstand them and appreciate their value to theindividual citizen, the nation, and the world.

We also recognized that, like everyone else these days, ourreadership is extremely busy and often does not have the timeto read lengthy articles. Feature articles describing theLaboratory’s major accomplishments, projects, and programswill continue to be the mainstay of the journal. We willendeavor, however, to always provide these descriptions inlanguage understandable by educated, interested nonexperts.Feature articles will be supplemented with briefs and researchhighlights describing, in a few paragraphs or pages, progressin previously reported projects, recent achievements inongoing programs, and the start of exciting new researchefforts. Also included will be a list of recent patents awardedand licensing agreements signed, which will illustrate thescale and scope of the Laboratory’s work. In addition, a newcommentary column will provide a forum for members of the

Laboratory’s top technical management to give theirviews on important institutional and programmatic issues and on external events affecting the Laboratory’scurrent status and future direction.

Finally, we changed the name of our publication toScience and Technology Review to better reflect theLaboratory’s mission of applying a broad range ofscience and technology in the national interest.

In addition to the changes in the journal, we haveevaluated its availability and concluded that we couldsignificantly improve the dissemination of technicalinformation about Laboratory programs. During the next several months, we will be adding interested readersto the distribution list. We are also taking advantage ofmodern information technology by making S&TRavailable electronically on the Internet via theLaboratory’s home page at http://www.llnl.gov or directly at http://www.llnl.gov/str/str.html. Electronicaccess will lead to a wider distribution for the journal and to increased technical interactions with researcherselsewhere. As the Laboratory’s electronic library grows,we will include hypertext links to technical references inS&TR articles. In this way, we will be able to tailor ourarticles to a nonexpert audience while providing thosereaders who desire more specifics with easy access to thedetailed technical information.

To help us assess whether we are reaching the goalsset for S&TR, we encourage you to complete and returnthe survey printed on the inside back cover of themagazine. This survey will be a regular part of thejournal and our way of staying in close touch with theopinions and desires of our readers. You may also e-mailcomments to us by clicking on the [email protected] linkat the bottom of our S&TR home page.

Once again, welcome to S&TR. We hope that you willenjoy the publication and find its information interesting and useful.

William Bookless,Scientific EditorScience and Technology Review

HIS past year and a half has been one of the mostunsettled periods in the history of the Livermore

Laboratory. Our laboratory, like all federal laboratories, has felt the winds of change from Washington. Variouscommittees are questioning the way in which the federalgovernment supports scientific research and theappropriateness of certain programs. Indeed, such questioningis inevitable and necessary in light of trillion-dollar budgetdeficits, continuing economic uncertainties, and widespreadconcern about health care, social security, crime and violence,education, and other basic survival issues.

At the same time, many of the problems facing the nationand the world today involve science and technology. Forexample: • Ensuring national security, not only by maintaining the U.S.nuclear deterrent but also by stemming the proliferation ofweapons of mass destruction. • Understanding, remediating, and preventing damagingeffects of human activities on the environment. • Solving the mystery of the genetic code. • Developing advanced technologies, processes, and products(particularly those related to energy production,biotechnology, and electronics) that enhance the quality oflife while securing the nation’s preeminence in the globalmarketplace. • Improving the quality and reducing the cost of health care.

We believe that national laboratories like Livermore are as important now as they have ever been. Our overriding

mission of serving the nation through the application ofscience and technology remains unchanged. However, as thescience and technology required to solve important nationalproblems grows more complex, we must make sure that weexplain our work—and the value of that work—in ways thatare accessible and meaningful to a broad audience. And we must remember that the value of scientific research(particularly publically funded research) lies, to a large extent,in its ability to solve real-world problems.

No longer is it sufficient for scientists to communicateonly with other scientists through professional journals or attechnical conferences. Neither is it sufficient for laboratoriesto communicate primarily with each other and with theirfunding agencies. We must also reach the large numbers ofinterested, educated nonexperts—government representativesand congressional staffers, community leaders, and thegeneral public—all of whom through their taxes contribute tothe Laboratory’s funding and therefore have a vested interestin the Laboratory’s work.

Science and Technology Review is one of the principalmechanisms by which we inform and educate a broadreadership about our research programs and accomplishments.Much of the Laboratory’s research is at the cutting edge ofscience and technology, making it particularly challenging to describe state-of-the-art accomplishments and theirsignificance in widely understood terms. Our goal is that thearticles presented here represent the full range of projects atLivermore and convey the challenge and excitement ofworking at the frontiers of science and technology.

Commentary on Science and TechnologyS&TR’s Premier Issue

Communicating the Worth of Our WorkCommunicating the Worth of Our WorkWelcome to Science and Technology Review

Welcome to Science and Technology Review

W

T

2

Science & Technology Review July 1995 Science & Technology Review July 1995

Bill LokkeActing Deputy Director forScience and Technology

3

The Laboratory in the News

says physicist Tom Casper, who leads the Laboratory’seffort in the collaboration. “This technology is important not only to fusion, but also to other scientificapplications—such as medical systems or environmentalwork—where real-time access and control are needed fromvarious locations around the country.”Contact: Tom Casper (510) 422-0787 ([email protected]).

Researchers explore MACHOs in dark matterAn international team of scientists reports that Massive

Compact Halo Objects, or MACHOs, constitute less darkmatter in the Milky Way than was previously thoughtpossible. One conclusion is that some other unknown type ofobject makes up the dark matter. On the other hand, thefindings may mean that the current model for the halo ofdark matter in our galaxy is inaccurate. The findings arereported in the April 10 issue of Physical Review Letters.

In their search for MACHOs, the team observed threemicrolensing events instead of the 15 that were expected ifMACHOs made up all of the ubiquitous dark matter.Microlensing is the brightening of a star that indicates thepassage of a large object in front of it, thereby magnifying,as if through a lens, the light passing around the object. TheSeptember 20, 1993, issue of Nature called this group’s firstobservation of microlensing, which garnered worldwideattention, the “footprint of dark matter.”

The MACHO collaboration, which consists of 18researchers from eight institutions, is funded in part by theLaboratory. For the past two years, the collaboration hasmonitored some 8.6 million stars in the Large MagellanicCloud with sophisticated camera systems, each of whichcontains 16 million pictures, The search for dark matter willcontinue for at least three years at the great MelbourneTelescope at Mount Stromlo Observatory near Canberra,Australia.Contact: Charles Alcock (510) 423-0666 ([email protected]).

5The Laboratory in the News

Science & Technology Review July 1995Science & Technology Review July 1995

Goal is to spot small, covert nuclear testsLaboratory researchers are attempting to develop an

improved method for identifying covert nuclear tests. Currentidentification techniques are efficient at identifying detonationsof 150 kilotons or greater, the yield limit imposed by theThreshold Test Ban Treaty. They are not able, though, todifferentiate effectively between small evasive nuclear testsand other seismic events, such as earthquakes or miningactivity. Conceivably, proliferant nations might test withsmaller nuclear devices to take advantage of the fact that smallexplosions are more difficult to identify.

One promising identification method being explored by Labresearchers compares two kinds of seismic waves—called P andS waves—generated by both earthquakes and nuclear explosionsand looks for differences in the size of their spectral ratios atmany frequencies. A spectral ratio is the ratio of the spectralamplitudes of two signals such as a P and S wave.

A weakness of the approach is that it probably would not beable to discriminate between large-scale concentrated chemicalexplosions and small nuclear detonations, according to PeterGoldstein, the principal investigator working on the detectiontechnique for the Lab’s Nonproliferation, Arms Control, andInternational Security directorate. However, efforts are underway to develop techniques for identifying such explosions.Contact: Peter Goldstein (510) 423-1231 ([email protected]).

Lab and MIT operate tokamak via InternetLaboratory scientists teamed up with colleagues from

the Massachusetts Institute of Technology recently todemonstrate the first transcontinental operation of a fusionexperiment via the Internet. Using a Department of Energysubnet of the global Internet known as the Energy SciencesNetwork, managed from Livermore, MIT scientistssuccessfully operated the Alcator C-Mod reactor in Cambridgeover ESnet. On the first full day of remote experiments, 21 of35 shots were controlled from Livermore.

Although aimed at proving the technical feasibility ofrunning a tokamak remotely, the LLNL–MIT Internetcollaboration also allowed scientists to learn more aboutmanaging the efforts of research groups working at the sametime in different locations on joint equipment.

“This demonstration was the definitive test of controlling alarge, complex physics experiment from a remote location,”

4

Cutting system assists in Russian nuclear cleanupThe Laboratory is continuing its efforts to assist in the

dismantlement of nuclear weapons in the former SovietUnion. In early April, Lab engineers packed a 6.4-m (21-ft)van with a portable “water knife” cutting system designed topermit safe access to warheads and other “hot” materials.The system will be used for standby response in the event of accidents involving the transport of nuclear materials.

Although water knives are not new, adapting a “portable”system is a unique application. The pressure generator aloneweighs 771 kg (1700 lb) while the control panels needed forremote use weigh 136 kg (300 lb). To help move thepressure generator, the van has been fitted with a winch;many components have been adapted with lifts that allowassembly by just one person.

The van containing the cutting system was sent to a basenear Moscow. Another van containing a second cuttingsystem was shipped from Livermore in early June. Inaddition, a Laboratory team was dispatched to Russia toprovide training in the use of the emergency cutting systems.

Assembly of the vans and cutters has been the work of the Disablement Technology Group in the Laboratory’sNonproliferation, Arms Control, and International Securitydirectorate. The work is part of Department of Energy’sefforts in the Safe Secure Dismantlement Program, aninternational agreement designed to provide assistance andequipment to Russia, Ukraine, Belarus, and Kazakhstan.Contact: Norm Stewart (510) 243-7768 ([email protected]).

Study suggests aging, genetic damage connectionA study by Lawrence Livermore scientists has found that

older people have more DNA damage than youngerindividuals. DNA is the molecule that carries the body’sgenetic code. On March 14, James Tucker, senior biomedicalscientist, reported the results of the three-year study at theEnvironmental Mutagen Society’s annual meeting in St.Louis, Missouri.

Tucker’s research team studied chromosomes in bloodsamples from a group of individuals ranging in age from 20 to 80 as well as in blood taken from the umbilical cords ofnewborn infants. The scientists found that genetic damageincreased with age and shot up dramatically after age 50. “Weexpected to observe more genetic damage in older people, andthis seems to indicate that damage can accumulate throughnormal living,” said Tucker.

The most common type of damage observed wastranslocations, where chromosomes break and recombine withother chromosomes. Translocations were found more than 10times as often in people over the age of 50 than in newborns.

Tucker said the results will benefit researchers studying theeffects of radiation on people by providing a baseline againstwhich genetic damage in exposed individuals can bemeasured.Contact: James Tucker (510) 423-8154 ([email protected]).

Radar licensee, AlliedSignal, to develop auto systemAmerigon Inc. of Monrovia, California, has announced an

agreement under which AlliedSignal Inc., one of the autoindustry’s largest electronics suppliers, will join in thedevelopment of an auto radar system based on LLNLtechnology.

Amerigon Inc. holds the license from the Lab to use theradar advance, called Micropower Impulse Radar, forautomotive safety applications. Amerigon plans possibleapplications of the radar as a device to signal when vehiclesare in a driver’s blind spot, as a backup warning system, as aparking aid, and for triggering side-impact air bags.

The new technology was invented by Livermore engineerTom McEwan in connection with his work for the Lab’s Novalaser. The radar has been used as a part of a system to measurethe balance and power output of the laser’s 10 beams and asthe heart of the diagnostics that measure neutrons from thefusion reactions.

For $10 to $15 in off-the-shelf components, theMicropower Impulse Radar can do the same tasks asequipment costing $40,000. Since the new radar technologywas announced in March 1994, the Laboratory has receivedover 2000 calls from businesses and individuals in at least 15 nations. Contact: Tom McEwan (510) 422-1621 ([email protected]).

6

Science & Technology Review July 1995

CCORDING to a common, rathersimplistic notion, we are what we

eat. On a far more empirical level,epidemiological studies reveal aconnection between diet and adversehealth consequences. Many observeddifferences in cancer rates worldwide,including incidences of colon and breastcancer, are linked to variations inhuman diets.

Strong evidence suggests thatmutations are the initiating events in thecancer process. In other words, thecomplex sequence of cellular changesultimately leading to malignant tumorsis thought to begin with structuralchanges—mutations—within themolecular units that make up the genes.

For 17 years, LLNL researchers havebeen investigating certain biologicallyactive compounds in foods that cantrigger tumors in animals, at least afterexposure to high concentrations, byproducing cellular mutations.

At first glance, identifying themutagens that might put us at risk andunderstanding how they affect thebody appear to be simple matters. Infact, the opposite is true. Consider justa few of the questions that must beaddressed to understand the entirepicture of diet-induced mutations andpossible links to cancer. Exactly whatcompounds in foods are dangerous,how are the compounds formed duringcooking, in what amounts are they

present after cooking, and how toxicor cancer-causing are they? Whatchemical changes take placemetabolically at the molecular levelafter the mutagenic substances areconsumed? For example, what role do metabolic enzymes play, how isDNA affected, and how might tumorsbe triggered in the body’s somaticcells? What chemical, tissue, animal,and human models might be useful to estimate risk to the humanpopulation? Are all people affectedsimilarly, or are some resistant tocancer-causing effects? If people vary in cancer incidence, whataccounts for the differences insusceptibility?

7


The Role of Cooked Food in Genetic ChangesWhen food derived from muscle is heated, potent mutagens are produced. Fornearly two decades, LLNL researchers have studied the formation of toxicmutagenic compounds in red meats and other foods containing protein. Thisreport, the first of two installments, focuses on the identification of food mutagensand measurement of their abundance in cooked foods as a function of cookingtemperature and time.

A

FOOD MUTAGENS:The Role of Cooked Food in Genetic Changes

environmental sciences, and forensics(Figure 1). Our research requires toolssuch as accelerator mass spectrometryand nuclear magnetic resonancespectrometry, to name a few. TheLaboratory is one of the few places thatbrings together the broad expertise andstate-of-the-art analytic tools required tofully understand each important aspectof the problem of mutagens andcarcinogens in the human diet. The waywe became involved in this field ofresearch has much to do with our role as a national laboratory withinterdisciplinary research programs.

Mutagens are the damaging agentsthat can structurally change themolecular units that make up the genes(that is, the genetic material, DNA) orthe relation of one chromosome toanother. For many years, LLNLinvestigators have been studying someof the ways that x rays, ultraviolet light,and some chemicals in the environmentcan act as mutagens. Carcinogens areagents that incite the development of acancerous tumor or other malignancy.Some 80 to 90% of mutagenicsubstances are also carcinogenic. Morethan 50 years ago, scientists painted theskin of mice with extracts from heatedanimal muscle and found that theextracts were carcinogenic, but theresearch went no further.

By the early 1970s, Bruce Ames atthe University of California, Berkeley,had developed a biological test tomeasure the mutagenic potency(mutagenicity) of substances.1* In thelate 1970s, T. Sugimura, who directedresearch at the National Cancer Centerin Tokyo, applied the Ames method andpublished the fact that smokecondensate from cooking and thecharred surface of broiled fish and beefwere mutagenic.2 One year later, BarryCommoner, working at WashingtonUniversity, St. Louis, used the Amesmethod to show that cooking

temperature and time affect theformation of mutagens in food.3

The news that cooking amino acids(the building blocks of proteins) andmuscle-containing foods could bedangerous triggered considerablescientific interest around the world. In1978, biomedical researchers at LLNLwere working on the problem ofmutagenic chemicals produced by oilshale retorting and coal gasification.Because of our combined expertise inchemical analysis (including differenttypes of chromatography andspectrometry), biological analysis(including the Ames mutation assay),and our emerging program in geneticsand toxicology, we received a multiyearcontract from the National Institute ofEnvironmental Health Science (NIEHS)to take a detailed look into the problemof food mutagens. As it turns out, whathappens when oil shale and coal areheated is not so different from some of

9


Food Mutagens

Clearly then, isolating, identifying,and assessing the biological activity ofmutagenic compounds in food is adifficult problem requiring extensiveeffort. Table 1 is an overview of someof the research issues addressed andanalytic methods used in this field ofinvestigation. This series of articlesfocuses on the first five questions under“Issues” listed in Table 1. A secondinstallment in Science and TechnologyReview will address the remainingissues.

A simple analogy can help put a keyfeature of our work into perspective.The compounds we have beeninvestigating for nearly two decades—the aromatic heterocyclic amines—arepresent in cooked foods at very lowlevels, in the range of about 0.1 to 50 parts per billion. Isolating material atthe part-per-billion level is equivalent topouring a jigger of Scotch into the holdof a full supertanker and then trying to retrieve it again. Although thecompounds we study are present in verysmall amounts, they are also the mostmutagenic compounds ever found, andthey produce tumors in mice, rats, andmonkeys. Such knowledge, combinedwith the fact that these compounds arepresent in many foods characteristic ofthe Western diet and that certain dietsare known to influence tumors atseveral body sites, gives our research an extra sense of urgency.

LLNL’s Approach

The single aspect that bestcharacterizes our research on foodmutagens and carcinogens—and sets our work apart from almost all otherefforts around the world—is itsmultidisciplinary nature. Biomedicalscientists at LLNL routinely collaboratewith investigators working in analyticalchemistry, synthetic chemistry,quantum chemistry, physics, the

8


Food Mutagens

Table 1. Some of the required interdisciplinary research, analytic methods, and tools needed tounderstand the possible connection of mutagens in cooked food to cancer.

Issues Research required Analytic methods and tools

• What cooked foods contain mutagens? • Chemical extraction and • Gas chromatography (GC)• What are the mutagenic compounds? purification • Liquid chromatography (LC)• What amounts are produced? • Identification and • Mass spectrometry (MS)

quantification • High-resolution mass• Proof of structure spectrometry (HRMS)• Synthesis of isomers • Nuclear magnetic resonance

(NMR) spectrometry• Ames/Salmonella test• Monoclonal antibodies

• By what mechanism are mutagens • Study precursors and • Modeling mutagens from formed during cooking? reaction conditions in – creatine

chemical models – creatinine • Aqueous vs dry heating – amino acids• Vary cooking temperature – sugars

• Heavy isotope incorporation

• How potent (mutagenic) are • Mutagenicity research (e.g., • High-performancethe compounds? use chemical to induce liquid chromatography (HPLC)

mutations, and count • Ames/Salmonella testfrequency of mutant cells • Animal mutation studiesor chromosomal changes) – Mice

– Chinese hamster ovary (CHO)cell cultures

• How are mutagens activated • Study chemical intermediates • Cell modelsmetabolically? (bioactivation pathways) – Mammalian cell systems

• Modulate metabolism in – Bacterial cell culturescell models • Enzyme inhibitors

• Radioactive labeling

• How is DNA affected? • DNA damage and repair • Computational chemistry analysis• DNA binding analysis • 32P-postlabeling of DNA adducts• Data adduct analysis • Accelerator mass spectrometry (AMS)

• Models– Whole animals (in vivo)– Animal cells in culture (in vitro)– Bacterial assays

• How are tumors induced? • Carcinogenicity research • Animal models(e.g., assess tumor induction – Monkeysin various tissues – Ratsin laboratory animals) – Mice

• What are the health risks • Dose-response assessment • 32P-postlabeling of DNA adductsfrom exposure? in humans • Accelerator mass spectrometry

• What people are affected? • Adduct formation as an • Epidemiology• Who is most at risk? indicator of exposure

• Risk assessment• Extrapolation from animal

studies

*All references are at the conclusion of the third part ofthis installment on p. 25.

Figure 1. Cyndy Salmon,one of the researchers inthe LLNL food mutagenresearch group, pours acooked food sample into anextraction tube to prepare itfor subsequent analysis.

N the foods that make up theWestern diet, the most common

mutagens belong to a class collectivelycalled the amino-imidazoazaarenes(AIAs). Not all the known foodmutagens are AIAs, but the commonlyfound ones are. As shown in Figure 2,AIA compounds have one or twoaromatic ring structures fused to theimidazole ring. They also have anamino group (NH2) on the number-2position of the imidazole ring and canhave methyl groups (CH3) of varyingnumber and location.

Of the list of toxic substancesknown to be produced during cooking,the most important may well be theAIAs. Also referred to as heterocyclicamines, these compounds are potentmutagens produced at normal cookingtemperatures in beef, chicken, pork,and fish when fried, broiled, or grilledover an open flame. The pan residuesthat remain after frying also have high

mutagenic activity,indicating that meatgravies can be a sourceof exposure. Ourresearch suggests thatsmoke from cookingmuscle meats ismutagenic as well, butany such air exposure islikely to be far less thanthat from eating thecooked food. Otherfoods, such as cheese,tofu, and meats derived from organsother than animal muscle, have verylow or undetectable levels of AIAmutagens after they are cooked.

Extraction

Analyzing cooked foods formutagens requires many differentmethods (Figure 3). The toxiccompounds in food must first be

11


Food Mutagens

the chemical processes that occur whena hamburger is cooked.

Our work on food mutagens alsoparallels our interest in the mechanismsby which pesticides and many othertoxic chemicals can elicit adversebiological responses. For example,benzo[a]pyrene is a widely studiedpollutant found in combustion products,and it has been isolated from burned fatand cigarette smoke. However, thiscompound becomes carcinogenic onlyafter it interacts with DNA followingoxidation by metabolic enzymes. Theproduction of such enzymes and theirroles in changing the chemicalreactivity of compounds are part of thebody’s normal biological response tocertain foreign substances. We arelearning that similar “metabolicactivation” takes place before foodmutagens become harmful.

Today, our research is fundedprimarily by the National CancerInstitute, with additional support fromthe Laboratory Directed Research andDevelopment program and othersources. There are approximately 50 other prominent research teamsworldwide studying the heterocyclicamines. However, except for one otherprogram in Japan, ours is the only teamthat brings a truly multidisciplinaryapproach to the problem ofunderstanding mutagens andcarcinogens associated with cookedfood and their consequences at thecellular, genetic, and molecular levels.

A Problem of Strategy

Strictly speaking, it is inaccurate to say that cooked foods containmutagens. More precisely, certaincooked foods contain premutagenicsubstances (promutagens) that aremetabolized by enzymes naturallypresent in body tissues, leading to theformation of one or more reactivemutagenic substances. Conventionally,however, “promutagen” and “mutagen”are used synonomously, and we havefollowed that practice here unless thepoint being made about the researchdemands a precise distinction.

At the outset of our research, wewere faced with problems of strategy.Studying substances that are present atvery low concentrations imposes manyresearch constraints. If we focused ononly a few foods, as seemed wise, thenour results and their implications forpublic health might be misinterpreted.Instead, we decided to examine thefoods that are the principal sources ofcooked protein: meats (any muscle-containing food, including fish), eggs,beans, cheese, and tofu. Whereas weinitially focused on meats, especiallyfried beef, we have now expanded therange of foods to include cooked breadsand grain products, heated flour frommany different plant sources, and meatsubstitutes.

Over the years, our research has alsoevolved from relatively simple conceptsand approaches to more sophisticatedones. Initially, we had to identify themutagenic compounds in heated foodsbecause many were not known (that is,neither synthesized nor analyzed). Thus,we focused our efforts on identifyingthe chemical composition and structureof mutagens, assessing how different

cooking procedures affect the formationof mutagens, and determining theamount (abundance) of the mutagenicproducts. Even though chemicalidentification and quantification are stillimportant activities, our work hasexpanded to include many other aspectsof the problem.

For example, we developedtechniques to help us learn howmutagens are metabolized in the body.We use animals as models tounderstand complex metabolicpathways and are developing cell-culture methods that model humanmetabolic systems. One particularlyimportant issue is how metabolites (theintermediate products formed byenzymes) interact with the geneticmaterial. We need to know exactly whattakes place at the molecular level,including covalent binding with andstructural changes to specificcomponents of DNA. This work tapsthe skills and facilities in several relatedresearch programs, including theHuman Genome and DNA repairprojects. (See the April/May 1992 andApril 1993 issues of Energy andTechnology Review for morebackground on these two programs.)

In assessing the effects of low-levelexposure to food mutagens, we makeuse of Laboratory expertise inaccelerator mass spectrometry (AMS).Yet another part of the story is thedifferences among humans insusceptibility to cancer, which hasbecome our newest effort. In essence,our success in recent years is derivednot so much from simply applyingstandard analytical methods bythemselves as from combiningbiological analysis with state-of-the-artanalytical tools available at LLNL tostudy all aspects of the health risks,ranging from dietary exposure to effectsin model systems and humans.

10


Food Mutagens

FOOD MUTAGENS:The Challenge of Identification

I(c) The food mutagen IQ

CH3

NH2

(b) Imidazole ring with an amino and methyl group

CH3

NH2

3 2

154

CH

H

H(a) The imidazole ring

H

abbreviated

as

N N

NN

N

N

N

N

N

C

C

Imidazole ring

Figure 2. Structure of theamino-imidazoazaarenes(AIAs), also called heterocyclicamines. (a) The imidazole ring iscommon to all AIAs. Numbersshow the position of atoms onthis ring. (b) In the AIAs, anamino group and one or moremethyl groups are attached. (c) IQ, one of the potent AIAsfound in cooked meats, has twoaromatic rings attached to theimidazole ring. The mutagenicactivity of the differentheterocyclic amines varies byseveral orders of magnitude andcan be increased when one ormore additional methyl groupsare present.

Step 1. Extract mutagens from cooked food

Step 3. Detect Mutagenic Activity

Solid-phase extraction

Ames/Salmonella Test

Combine food extract,bacteria, and enzymes

Combine food extractand bacteria

Count revertant colonies

Count revertant colonies(baseline measurement)

Figure 3. Some of the steps required toextract, separate, purify, and confirm thepotency and chemical structure of mutagens incooked food. These steps show a typicalsequence of events during research on a givenmutagen. However, the sequence shown herecan vary depending on whether our objective isto study a known mutagen or to assess theproperties of a new candidate. Each of thesteps is described in more detail in the text.

Step 2. Separate and purify the many differentcompounds in the complex mixture

High-performance liquid chromatography (HPLC)

Pump

Separation column

Detector

Purified fraction

Step 4. Subsequent Characterization

Mass spectrometry

Determine molecularweight (MW) andchemical composition

Nuclear magnetic resonance (NMR) spectrometry

Determinedefinitivestructure

198 MW

IQ

224 MW

PhIP

CH3

NH2

N

N

N

N N

N

CH3

NH2

Spectrum review

Matching spectra

Confirm mutagenidentity

Ultravioletchromatogram

y axis = Absorbancex axis = Retention time

Detect peakabsorbance at aparticular retention time

Fluorescencechromatogram

y axis = Fluorescencex axis = Retention time

Increase sensitivityfor fluorescence only

1312

needed for the next step—testing formutagenic potency.

Detection of Mutagenicity

The most widely used detectionmethod for mutagenic potency is theAmes/Salmonella mutation test,1 whichis described in more detail in the box onp. 16. This test for mutagenic activity isexquisitely sensitive and relativelyinexpensive. It is also convenientbecause each analysis requires only 48hours, and many samples can beanalyzed in parallel (Figure 5).

The essential point to remember isthat the Ames test (step 3 in Figure 3)gives us a number by which we canexpress the mutagenic activity of agiven compound or food extract. Thisnumber by itself for a single mutagenwould have little meaning. However,we now have numbers for most of theknown mutagens in cooked foods andfor over a hundred additional mutagensfrom other sources, so we can comparethe mutagenic activity of many differentstructural types. When the Ames test is

used during initial screening for newmutagens and carcinogens, it serves as aguide to the chemical purification ofbiologically active molecules. It canalso be used to test and compare thepotency of newly synthesizedchemicals.

Characterization

Once a mutagen has been detected,we can characterize it further through avariety of analytical methods (step 4 inFigure 3). The type and sequence oftests depend on our objective for agiven mutagen (Figure 6). For example,we can routinely determine themolecular weight through massspectrometry and study the detailedchemical composition (the number ofhydrogen, carbon, and nitrogen atoms)by high-resolution mass spectrometry(HRMS). In mass spectrometry,complex compounds are broken up intoionized fragments, which areaccelerated through a magnetic fielduntil they strike a detector. Because thepath of an ionized fragment through the

field is determined by its inertia, we candetermine the mass of the various ionsby their spatial distribution on thedetector. Ultraviolet absorbancespectrometry and fluorescencespectrometry are other identificationmethods that are often combined withchromatography.

Substantially more effort is requiredif we want to identify a mutagen for thefirst time. For an unknown compound,we first need information on the atomiccomposition and the position of atomsin the molecule. This work requiresHRMS and nuclear magnetic resonance(NMR) spectra (step 4 in Figure 3)

15


Food Mutagens

chemically extracted beforepurification. Over the years, we andother researchers have dramaticallyimproved on the original extractiontechniques that required various acidsor mixed organic solvents in multistepschemes.

We now use solid-phase extraction,which is based on a method firstdescribed by G. A. Gross in 1990.4After homogenizing cooked food in ablender to obtain a uniform sample, we can extract a sample quickly andefficiently by passing it through a series of small tapered tubes containingchemically activated particles (see step 1 in Figure 3 and Figure 4). Thesmall amounts of organic solvents thatare needed during this solid-phaseextraction generate a minimum ofhazardous waste.

Separation and Purification

We use high-performance liquidchromatography (HPLC) for finalseparation and purification of theextracted compounds in a food sample

(see step 2 in Figure 3). Liquidchromatography is a standard techniquein chemistry labs. In HPLC, a liquidmixture is pumped under high pressurethrough a long, narrow tube filled withfine silica particles. This materialdifferentially retards the passage ofdifferent molecular components so thateach one exits after a characteristicdelay or retention time. Our recent solid-phase extraction method together withHPLC allows excellent quantificationfrom small samples (about a tenth of ahamburger patty, or one bite) and a 1- to2-day turnaround time for results.

For unknown mutagens, a separationis carried out in several stages. Weobtain about 100 fractions at the finalstage, where a “fraction” is one portionof the sample material that is capturedin a separate vial after exiting the HPLCdetector. One fraction at the final stageof separation contains as little as abillionth of the starting material.However, because the extracts frommeat and other food products cooked atelevated temperatures are tremendouslypotent, only a very small sample is

14


Food Mutagens

Figure 4. Researcher Cyndy Salmon usessolid-phase extraction to extract a sample bypassing it through a series of small cylinderscontaining small amounts of organicparticles.

Figure 5. Julie Avila, one of theresearchers in the LLNL foodmutagen research group, testsmutagens in cooked beef using theAmes/Salmonella test. (a) The foodsample is added to a mixturecontaining bacteria, nutrients, andenzymes needed for metabolism,and then (b) poured onto a petriplate. (c) Close up of growingbacterial colonies (calledrevertants) after 48 hours. Countingthe colonies gives us a number thatrepresents the sample’s mutagenicactivity.

(a)

(c)

Figure 6. Kathleen Dewhirstcombines methods, such asgas chromatography andmass spectrometry or liquidchromatography and massspectrometry, tocharacterize the foodmutagens in cooked meat.Mass spectrometry allowsus to determine themolecular weight of amutagen.

(b)

together with synthesis of all possibleisomers. Isomers are two moleculeswith the same number of atoms andmolecular weight but differentstructures. NMR spectrometry requiresthe highest quantity and sample purityof all the analytical methods, but it gives us the most definitiveinformation on chemical structure. Theexact chemical structure of a givenmutagen can be proven by comparing it with a known standard that issynthesized in the laboratory.

After the physical and chemicalproperties of a mutagen are known, wecan use the information to determinewhether that mutagen is present in othertypes of food. This approach gives us away to determine the dose of a givencompound in our diet and to assess thehuman risk associated with ingestingthat compound.

The Major Food Mutagens

Table 2 is a summary of the 14major mutagens that have beenidentified in at least one type of heatedfood to date.5 Notice that some of thecompounds have the same molecularweights. For example, 4-MeIQx and 8-MeIQx are isomer pairs and so are Trp-P-2 and Me-AåC. The ultravioletabsorbance spectra of two differentcompounds may be identical when theyare isomer pairs and differ only, forexample, in the position of a methylgroup on one of the rings. The similarproperties of isomers make themdifficult to separate usingchromatography. Likewise, other analytic tools do not alwaysdifferentiate between isomers.Additional mutagenic isomers havebeen synthesized for most of the foodmutagens in Table 2. The presence ofisomers means that we need to applyseveral different criteria foridentification purposes because nosingle property, such as an absorbance

spectrum, can uniquely identify all ofthe mutagens.

The compounds listed in Table 2 arenot the only mutagens or carcinogens infood. Researchers at LLNL andelsewhere have identified otherbiologically active compounds,including additional aromatic amines,nitrosamines, and hydrazines. However,

the heterocyclic amines we have beeninvestigating are among the mostabundant and potent substancesdetected to date. Because of theirpresence in cooked meats that arecommon in Western diets and theirassociation with certain types of cancerin laboratory animals, they warrantdetailed investigation.

17


Food Mutagens16


Food Mutagens

Table 2. Major mutagens that have been identified in at least one type ofheated food, such as fried beef or fish. The names of mutagens firstidentified at LLNL are in color.

Short name Chemical name Molecular weight

Phe-P-1 2-amino-5-phenylpyridine 170

TMIP 2-amino-n,n,n-trimethyl- 176imidazo[4,5-f]-pyridine

AåC 2-amino-9H-pyrido- 183[2,3-b]-indole

Glu-P-2 2-aminodipyrido- 184[1,2-a:3´,2´-d]-imidazole

Trp-P-2 3-amino-1-methyl- 1975H-pyrido[4,3-b]-indole

Me-AåC 2-amino-3-methyl- 1979H-pyrido[2,3-b]-indole

IQ 2-amino-3-methyl- 198imidazo[4,5-f]-quinoline

IQx 2-amino-3-methyl- 199imidazo[4,5-f]-quinoxaline

Trp-P-1 3-amino-1,4-dimethyl-5H- 211pyrido[4,3-b]-indole

4-MeIQ 2-amino-3,4-dimethyl- 212imidazo[4,5-f]-quinoline

8-MeIQx 2-amino-3,8-dimethyl- 213imidazo[4,5-f]-quinoxaline

4-MeIQx 2-amino-3,4-dimethyl- 213imidazo[4,5-f]-quinoxaline

PhIP 2-amino-1-methyl-6-phenyl- 224imidazo[4,5-b]-pyridine

4,8-DiMeIQx 2-amino-3,4,8-trimethyl- 227imidazo[4,5-f]-quinoxaline

The Ames/Salmonella Test: A Key to Our Research

Our success in detecting and identifying mutagens in cookedfoods is made possible by the interplay of many different types of chemical analyses, including chromatography and massspectrometry (Figure 3), and biological methods. The Ames testis an exquisitely sensitive biological method for measuring themutagenic potency of chemical substances. The Ames test byitself does not demonstrate cancer risk; however, mutagenicpotency in this test does correlate with the carcinogenic potencyfor many chemicals in rodents.

The test was developed in 1975 by Bruce Ames and hiscolleagues at The University of California at Berkeley. The Amesmethod is based on inducing growth in genetically altered strainsof the bacterium Salmonella typhimurium. To grow, the specialstrains need the amino acid histidine. However, when thechemical agent (mutagen) that is being studied is givento bacteria, some of the altered Salmonella undergomutations. Following a particular type of mutation, thebacteria can grow like the original “wild” (unaltered)strains without histidine. Because the mutant bacteriarevert to their original character with regard to thenutrient histidine, they are called “revertants.”

The Ames test yields a number—specifically, thenumber of growing bacterial colonies—which is ameasure of the mutagenic activity (potency) of atreatment chemical. This value is often expressed asthe number of revertants per microgram of a purechemical (mutagen) or per gram of food containingthat mutagen. Some pure mutagens result in hundredsof revertants per microgram, but many of thesubstances we have tested from cooked meat producehundreds of thousands of revertants per microgram.For example, in one strain of bacterium, the PhIPmutagen results in about 2000 revertants permicrogram, whereas another cooked food mutagen, IQ,results in 200,000. The illustration at the right showshow a food extruct is tested for its mutagenic activity.

In brief, a test begins by placing about 100 millionSalmonella bacteria in a petri dish containing a nutrientagar lacking histidine. A few bacteria willspontaneously revert in the absence of mutagens.Counting these revertant colonies gives us a baselineagainst which to check the validity of our complexlaboratory procedures. In a separate but essentiallyidentical histidine-deficient petri dish, another batch of

Salmonella bacteria are given a mutagen plus mammalianenzymes required for metabolism. (Adding such enzymes givesus a more realistic measure of the mutagenicity of a substancefor mammals. The enzymes are typically supplied from livercell extracts of rats given substances to increase levels ofmetabolizing enzymes.) Revertant bacteria grow into visiblecolonies. We simply count the colonies (equal to the number ofrevertants) after a standard time (48 hours) under standardgrowing conditions (37°C).

Different strains of altered Salmonella bacteria are availablefor the Ames test. The strains vary in sensitivity to specificmutagens. We used two strains, known as TA98 or TA100, formost of our recent work on fried beef and cooked grains. Thesestrains were generously supplied by Bruce Ames.

Control group

Mutagen group

Count spontaneousrevertants for baselinemeasurement

Incubate for 48 hours at 37˚C

Pour bacteria andfood extract onnutrient agar

Count colonies to obtainnumber of revertants pergram of food

Incubate for 48 hours at 37˚C

Pour bacteria, food extract, andenzymes on nutrient agar

of 244. However, the presence of thisnew mutagen in food has not beenverified.

Variations in Cooking

During the actual cooking of meatpatties, water and precursors move tothe hot, drying contact surfaces of themeat where reactions occur. Suchmigration, with water serving as thetransport vehicle, may account for theconcentration of precursors near themeat surface, which we have observedin several investigations. However,different cooking practices can lead to

very different results. For example,some mutagens are produced at allfrying temperatures, whereas othersmay require higher temperatures.Furthermore, when hamburger pattiesare grilled at high temperature over anopen flame, we can account for lessthan 30% of the mutagens in the meat.When cooking over an open flame,polycyclic aromatic hydrocarbons(different from AIA food mutagens)arise from fat that drips from themeat—this is an entirely differentmechanism than those that produceheterocyclic amines from heatedmuscle tissue itself. Thus, theformation of mutagens is complex andhighly dependent on the details ofcooking.

Preparation Principles

Given this complicated picture,what statements about food preparationcan we make with any certainty? Hereis a summary of some of the important

things we have learned about thecooking process:• Food mutagens can be producedboth with and without water present.Early reports suggested that water isessential to produce food mutagens.In later studies, dry heating actuallygives a greater percentage of certaintypes of mutagens compared withaqueous heating. We know, forexample, that the mutagen PhIP isformed relatively efficiently in dryheating reactions. We have alsofound that water tends to inhibit theformation of IQ-type mutagens.• Microwave pretreatment of meatcan reduce the formation ofheterocyclic amine mutagens. Whenmeat is microwaved for a fewminutes, a clear liquid is released,which contains many of theprecursors of mutagens. When theresulting liquid is drained off beforefrying, our studies show thatmutagenic activity, as measured bythe Ames test, and the amount of

19


Food Mutagens

OOKING practices can causelarge variations in the total

mutagenic activity and in the amountof specific mutagens present in muscle-containing foods. For example, theamount of mutagens in a cookedhamburger from a restaurant variesconsiderably from one vendor toanother and is often several-fold lowerthan that in a hamburger prepared inour laboratory (and presumably athome). The variation has much to dowith the details of food preparation,such as cooking temperature andcooking time. It is becomingincreasingly clear that there can bemany different routes and rates offormation for the different mutagenswe are investigating. Thus, a majorconcern is to identify the precursorsand specific reaction conditions thatlead to the formation of mutagensduring cooking. With this information,it may be possible to devise strategiesto reduce or prevent the formation ofmutagens.

Precursors

The reactions that produce mutagensin cooked food are not merely therandom coalescence of small fragments.We now know that the heterocyclicamines can be formed from singleamino acids (the building blocks ofproteins) or proteins when theseprecursors are heated alone. However,the temperatures required to producemutagens from amino acids or proteinsby themselves are higher than thosenormally used in cooking.

Muscle meats contain creatine andcreatinine. At more typical cookingtemperatures (greater than 150°C), oneor both of these two precursors reactwith the free amino acids and, in somecases, sugars to form a series ofheterocyclic amines more easily.

Modeling the Formation

We have modeled the formation of theimportant mutagen, PhIP (pronounced

“fip”), starting with the amino acidphenylalanine mixed with either creatineor creatinine, both of which are foundnaturally in animal muscle. Whenphenylalanine and creatine are mixed inthe proportion normally found in rawbeef and dry heated at 200°C, PhIP isproduced in amounts comparable tothose found after cooking beef. Figure 7shows the structures of phenylalanineand creatine and of the PhIP moleculethat is produced.

We have modeled the formation of several other food mutagens inadditional laboratory experiments. Forexample, the mutagen IQ can beformed with creatine, creatinine, andany of four different amino acids,again suggesting many differentpossible routes of formation.

Model reactions can help us identifynew mutagens as well. In one case, dryheating three precursors known to bepresent in meat led us to identify amutagen with two amino and twomethyl groups and a molecular weight

18


Food Mutagens

CN

N

N

N

CH

CH2

CH3

CH3

NH2

NH2

NH

NH2

HOOC COOH

L-Phenylalanine

PhIP

Creatine

Food Mutagens:The Cooking Makes aDifference Figure 7. Modeling the formation of the potent mutagen

PhIP. We combined two precursors, phenylalanine andcreatine, in amounts naturally found in raw beef. Aftersimple dry heating, PhIP was produced in yieldscomparable to those we obtain in beef after the cookingprocess. We have also labeled the two precursors withheavy isotopes to track the incorporation of specific atomswithin each precursor molecule into the PhIP molecule.Such work shows unequivocally the source of atoms thatmake up the mutagenic product.

important single source of heterocyclicamines in the typical American diet.However, several other popular cookedmeats, including fish, chicken, and pork,have been shown to produce a potentresponse in the Ames test.

Of the several different heterocyclicamine mutagens now identified, wewanted to know which ones are mostimportant (that is, most abundant bymass) in cooked muscle meats. To helpanswer this question, we compared theresults of many studies from LLNL andelsewhere. Specifically, we comparedthe mass percentages of differentmutagens in cooked muscle meats,including fried beef, broiled fish, andcommercially prepared beef extract. Wefound that the results were generallyconsistent among different laboratorieseven when different analytical methodswere used.

First, we did not detect significantlevels of three mutagens, Trp-P-1, Trp-P-2, and Glu-P-1, in any of our meatsamples. Second, we found that fourcompounds, IQ, 8-MeIQx, 4,8-DiMeIQx, and PhIP, contribute about80% of the mutagenic activity in thecooked muscle foods that were studied.Third, we found that PhIP alone canaccount for a startling 83 to 93% of themass of these four mutageniccompounds. Clearly, the analysis ofPhIP is important because it appears tobe, by far, the most abundantheterocyclic amine by mass incommonly eaten cooked meats. BecausePhIP is as carcinogenic as the othermutagens, its analysis becomes evenmore essential.

We examined the production of PhIPand other mutagens in beef at differentcooking temperatures and times. Thebox at the right gives the details on howwe prepare our fried beef. Figure 8shows the mutagenic activity, asmeasured by the Ames test, of all themutagens combined in a gram of beef

21


Food Mutagens

heterocyclic amine are 90 to 95% lowerthan they are in meat samples that arenot pretreated by microwave cooking.6The box below discusses this and othermethods that have been tested to reducethe formation of mutagens.• Different cooking methods producequite different results. In general,frying, broiling, and flame grillingmuscle meats produce more

heterocyclic amines and mutagenicactivity than other methods. Stewing,steaming, and poaching produce little orno mutagenic activity. Roasting andbaking have variable responses.• Heating temperature is extremelyimportant as is the time of cooking at agiven temperature. Our extensive findingson this important topic are best discussedaccording to the type of food product.

Cooked Muscle Meats

Fried beef patties appear to be themost commonly eaten cooked meatwith the highest mutagenic activity.Because of the high intake of fried beef(based on surveys from the U.S.Department of Agriculture and theDepartment of Health, Education, andWelfare), this food may be the most

20


Food Mutagens

Can the Mutagens in Cooked Beef Be Reduced?

Since mutagens were first observed in cooked meats, researchersin several different laboratories have explored various ways toreduce the amounts produced during food preparation. They havefound that mutagenic activity can be lowered by addingantioxidants, soy or cottonseed flour, tryptophan, and various otherfood additives or sugars either alone or with starch. However, noneof these additives is widely used commercially or at home.Consumer acceptance and possible changes in the taste, texture, andnutritional content of the cooked food need to be explored further.

Surveys have shown that more than 90% of American homeshave a microwave oven. As a practical way to reduce the mutagenand fat content of beef, we studied microwave pretreatment ofhamburger for various times before conventional frying either at200 or 250°C for 6 minutes per side. Our tests used a standardcommercial microwave oven set at 80% power for 0 to 3 minutes.The results were dramatic.

We found that the mutagen precursors in hamburger (creatine,creatinine, amino acids, and glucose), water, and fat were reducedup to 30% in the microwaved patties. The graph shows the amountof creatine remaining in the meat as a function of microwavepretreatment times. The fairly rapid loss of water-soluble mutagenprecursors and fat takes place in the clear liquid that is releasedafter microwaving. When this liquid is discarded before frying,mutagens in the cooked meat are reduced up to approximately 90%following frying, as shown in the table.

How is it possible that 90% of the mutagens disappear when the precursors are reduced by only 30%? The difference can beexplained by second-order reaction kinetics. For example, if tworeactants are needed, and each is reduced by 30%, then the productwould be reduced by about 50%. If three reactants are required andall are reduced by 30%, the product would be reduced by 70 to80%. It is also possible that some threshold level of precursor isnecessary to produce a mutagenic response or that some inhibitoris formed after microwave pretreatment. As with other techniquesto reduce mutagen content, the palatability of food may ultimatelygovern consumer acceptance of microwave pretreatment.

Microwave pretreatment time, min

500

43210

Pre

curs

or c

ompo

und

crea

tine,

m

g/10

0 g

of g

roun

d be

ef

400

300

200

Measured mutagenic activity, from the Ames test, in beef pattiespretreated in a microwave oven and then fried for 6 minutes at200 or 250°C.

Mutagenic activity from the Microwave time, Ames test, revertants per gram

min 200°C 250°C

0 450 14001.0 220 3691.5 130 2162.0 47 673.0 16 41

1200

121086Frying time, min

420

Mut

agen

ic a

ctiv

ity,

reve

rtan

ts p

er g

ram

of f

ried

beef 1000

800

230°C

190°C

150°C

600

400

200

0

Figure 8. A graph of themutagenic activity inbeef patties fried atthree differenttemperatures. Theessential point isthat mutagenicactivity increaseswith both fryingtemperature andtime.

One major difficulty in our dose-and exposure-assessment work is thatthe content of mutagens can varywidely even in the same kind of foodproduct when it is obtained fromdifferent suppliers or prepared bydifferent restaurants. Although therelative amounts of the heterocyclicamines are generally consistent amongdifferent studies and laboratories, theprecise amount of mutagen per gram ina given cooked food can span a tenfoldrange.

Hamburgers from fast-foodrestaurants generally have considerablylower levels of mutagens than thoseprepared at home. This result isprobably due to the fact that many fast-food restaurants cook their meat atmoderate temperatures on a grill or overopen flames for a short time. Becausethe meat patties are thin, the productsare not generally overcooked.

Because food-preparation practicesvary, over the years we have attempted toapproximate a range of cooking practicesthat are common in Americanhouseholds. In various experiments, foodswere pan fried, oven broiled, baked,boiled, stewed, grilled over coals, or leftraw. However, for the studies on red meatreported in this article (see Table 3), wepurchased ground beef, sold as containing15% fat, from a local market. We formedthe meat into patties weighing 100 grams(a little less than a quarter of a pound) andfried them on a commercial, electric,stainless-steel griddle for 2 to 10 minutesper side and at surface temperatures of150, 190, or 230°C. We monitored thegriddle surface with a digital probethermometer. After the meat was cooked,it was homogenized in a blender toproduce a uniform sample. Samples werefrozen at –4°C until extraction forsubsequent testing and analysis.

How We Fried the Burgers We Studied

Table 3. Content of four different mutagens in fried beef patties (expressed as nanogramsof mutagen per gram of beef) cooked at different times and temperatures.

Cooking time Cooking temperature of grill, °CMutagen per side, min 150 190 230

IQ 2 none 0.1 none4 none 0.1 0.156 0.1 0.45 0.6

10 0.1 0.85 0.7

8-MeIQx 2 none 0.1 0.74 none 0.25 0.46 0.2 1.3 5.6

10 0.6 1.3 7.3

4,8-DiMeIQx 2 none none 1.64 none 0.1 0.156 0.2 0.55 1.2

10 0.4 1.1 1.0

PhIP 2 none none 1.34 none 0.15 1.36 0.25 1.9 7.8

10 1.8 9.8 32.0

In fact, the mutagenic activity ofbreadsticks cooked for double theregular heating time is 20% that of ahamburger fried 6 minutes per side at210°C. In all cases, overcooking grainfoods led to much higher mutagenicactivity. Cooked garbanzo bean flourand the grain beverage powder, whichwe tested as purchased, had relativelyhigh mutagenic activity. Cooked riceand rye flour (containing no gluten), onthe other hand, showed no detectableactivity, and rice cereal showed verylittle. Fried tofu (soy bean curd) was notmutagenic, and the measured level ofactivity in meat-substitute patties(which are made from vegetableproteins) after frying was about 10% orless than that of a beef patty cookedunder the same conditions.

Table 4 summarizes the mutagenicactivity, as measured by theAmes/Salmonella test, for a variety ofcooked-grain food products. The resultsare expressed as mutagenic activityfrom the Ames test, so they cannot bedirectly compared with those in Table 3.(Recall that the numbers in Table 3represent a different measure, namelythe content by weight of individualmutagens expressed as nanograms ofmutagen per gram of beef.) Because wedo not yet know the identity of themutagens present in cooked grainproducts, we cannot provide theircontent by weight. However, to allowfor some comparison between cookedgrains and meat, we have included thevalues of mutagenic activity forhamburger cooked for three differenttimes at the end of Table 4.

Overall, the level of mutagenicactivity measured in heated nonmeatfoods is lower than that in cookedmeats. It is important to recognize thatthe cooked grains we studied lack thecreatine and creatinine levels thatexplain the formation of mutagens inmuscle meats during cooking. We arecurrently investigating the question of

why foods high in gluten are quitemutagenic in the absence of creatineand creatinine. We suspect that theamino acid, arginine, can substitute forthe creatine and creatinine precursorsfound in meat, but it may be a less

23


Food Mutagens

patty fried at 150°, 190°, and 230°C.We found no detectable heterocyclicamines after frying at 150°C for 2 to 4minutes. In general, increasing eitherthe temperature or time of cooking(specifically, frying on a solid metalrestaurant-type grill) causes a dramaticincrease in both the mutagenic activityand the total amount of mutagens

produced, especially PhIP and 8-MeIQx. For the most part, as shown in Table 3, the amount of individualmutagens in fried beef increasesproportionately with the cookingtemperature. A clear exception to this trend is the compound PhIP, which is produced at much greaterconcentrations at higher temperaturesrelative to the other mutagens we havestudied. When the cooking temperatureand time are increased, the PhIP contentof fried beef patties increases nearlyexponentially.

Mutagens from Grain?

We also recently used the Ames testto assess the mutagenic activity in manyheated foods derived from grainproducts. Our studies include cookedbreads (white, pumpernickel, crescentrolls, and pizza crust), breadsticks,heated flour from many different grainsources, breakfast cereals, grahamcrackers, and meat-substitute pattiesafter frying. These foods were eithertested as purchased without additionalcooking (for example, graham crackersand a grain beverage powder) or were cooked according to packageinstructions. In some studies, wedeliberately overcooked the grainproducts for twice the cooking time atthe specified temperature setting to seeif the mutagen content would increasewith continued cooking, as it does inmuscle meats.

Our studies generally demonstrateincreased mutagenic activity in grainfoods with cooking time, but the exactcomposition of the food is important.For example, when wheat gluten (theprotein in wheat seed) is heated alone at210°C in a beaker, it shows a potent,time-dependent mutagenic response(Figure 9). Because breadsticks are highin wheat gluten, they also show someactivity when heated normally andmuch higher activity when overcooked.

22


Food Mutagens

4000

1209060Heating time, min

300

Mut

agen

ic a

ctiv

ity, r

ever

tant

s pe

r gr

am o

f whe

at g

lute

n

3500

3000

2500

2000

1500

1000

500

0

Figure 9. Themutagenic activity ofwheat gluten increasesdramatically whenheated at 210°C for upto 2 hours. This potentresponse tells us thatone or more highlymutagenic chemicals,still unidentified, areformed with continuedcooking at hightemperature.

Table 4. Mutagenic activity of nonmeat food products (expressed as the number ofrevertants [mutants] per gram from the Ames/Salmonella test using the TA98 strain ofbacteria). Results for hamburger are given for comparison.

Mutagenic activity,Food sample revertants per gram

Flour from plant sources heated to 210°C for 60 minutesChemical-grade gluten 1330Food-grade wheat gluten 970Cornmeal 180Garbanzo flour 1890Teff flour 420Rice flour none detectedRye flour none detectedWheat flour for bread 320

Cooked food samples tested as purchased or cooked as directedWhite bread 2Pumpernickel 6Breadsticks 6Crescent rolls 1Pizza crust 3Graham crackers 4Grain beverage 320

Food samples cooked double the time directedWhite bread 5Pumpernickel 28Breadsticks 40Crescent rolls 4Pizza crust 8

Toasted breakfast cereals tested as purchasedRice-based 2.2Corn-based 4.4Wheat-based (various samples) 0 to 8.8

Commercial meat substitutes fried at 210°C for 6 minutes per sideGluten-based patties (various samples) 6 to 9.4Tofu none detectedFalafel 2.3Tempeh burger 23Tofu burger non detectedSoy-based patties 6.6Gluten, wheat, teff-based patties (230°C) 30

Hamburger fried at 210°C for 6 minutes per side 220Hamburger fried at 230°C for 6 minutes per side 800Hamburger fried at 250°C for 6 minutes per side 1400

amine mutagens, even if the meat iscooked well-done.• Most nonmeat foods, includingcooked grain products, contain lowerlevels of mutagens than cooked meats.

At least in rodents, we know thatfood mutagens trigger cancer in severaldifferent target tissues, such as the liver,colon, breast, and pancreas. In a follow-up installment in Science andTechnology Review, we will address thehealth risks to humans that may arisefrom exposure to heterocyclic amines.For this intriguing part of the story, wewill show how these highly toxiccompounds can react with the mostcritical macromolecule of all, DNA.With a connection established betweenfood mutagens, DNA damage, and thepotential for cancer, we will then try tomake sense of what all the numbers onmutagenic activity and mutagen contentin food mean for the average person.

Key Words: Ames/Salmonella assay;amino-imidazoazaarenes (AIAs);carcinogen; DNA adducts; heterocyclicamines; high-performance liquidchromatography (HPLC); mutagens—airborne, in cooked foods, in fried beef;mutagenicity; 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP); 2-amino-3-methyl-imidazo[4,5-f]quinoline(IQ); 2-amino-3,8-dimethyl-imidazo[4,5-f]quinoxaline (MeIQx).

References1. B. N. Ames, J. McCann, and

E. Yamasaki, “Method for DetectingCarcinogens and Mutagens with theSalmonella/Mammalian-MicrosomalMutagenicity Test,” Mutation Research31, 347–364 (1975).

2. T. Sugimura et al., “Mutagen-Carcinogens in Foods with SpecialReference to Highly MutagenicPyrolytic Products in Broiled Foods,” inOrigins of Human Cancer, H. H. Hiatt,J. D. Watson, and J. A. Winsten, Eds.,(Cold Spring Harbor Laboratory, Cold

Spring Harbor, New York, 1977), pp. 1561–1577.

3. B. Commoner et al., “Formation ofMutagens in Beef and Beef ExtractDuring Cooking,” Science 210, 913–916(1978).

4. G. A. Gross, “Simple Methods forQuantifying Mutagenic HeterocyclicAromatic Amines in Food Products,”Carcinogenesis 11, 1597 (1990).

5. For a general review of research on foodmutagens, see J. S. Felton and M. G.Knize, “Heterocyclic-Amine

Mutagens/Carcinogens in Foods,”Handbook of ExperimentalPharmacology, Vol. 94/I, C. S. Cooperand P. L. Grover, Eds.(Springer–Verlag, Berlin, Germany,1990), pp. 471–502.

6. J. S. Felton et al., “Effect of MicrowavePretreatment on Heterocyclic AromaticAmine Mutagens/Carcinogens in FriedBeef Patties,” Food ChemicalToxicology 32 (10), 897–903 (1994).(UCRL-JC-116450)

25


Food Mutagens

efficient mutagen precursor in cookedgrain products.

Before we can evaluate the riskassociated with cooked grains, we needto determine the mass of mutagens ineach food and to identify the specificmutagenic compounds that are present.Except for very low levels of PhIP inwheat gluten (accounting for only 4%of its mutagenic activity), our analysisdid not reveal any of the other mutagensfound in cooked meat or listed in Table 2. Because the mutagens incooked grain appear to be as potent asthe heterocyclic amines—and suchpotency is unusual, we suspect that themutagenic compounds may be newheterocyclic amines similar to those wehave found in cooked meats. However,more work needs to be done before weunderstand the entire picture.

What About Fumes?

Some studies have suggested thepossibility of an increased risk ofrespiratory tract cancer among cooksand bakers. When foods rich in proteinare heated, the fumes that are generatedsometimes contain many differentknown carcinogens, includingpolycyclic aromatic hydrocarbons andheterocyclic amines. Working withcolleagues at the University ofCalifornia at Davis, we recently studiedthe mutagenicity of fumes generatedwhen beef is fried at high temperatures.

We collected smoke from cookingby using a special sampling systemconsisting of a condenser, Teflon filters,and absorbent tubes containingpolyurethane foam and a resin. Wefound that the main volatile compoundsgenerated during frying were alcohols,alkanes, aldehydes, ketones, phenols,and acids. Their presence—wemeasured 34 different components—may account for much of the toxicity of

fume samples in bacterial tests. Twomutagens, PhIP and AåC, were themost abundant of the heterocyclicamines we measured in smoke, withAåC accounting for 57% of the totalweight of mutagens in the recoveredsamples. However, even though AåCseems to be the most volatile of ourquantified heterocyclic amines insmoke, its actual contribution to themutagenicity of fumes is negligiblebecause its mutagenic potency is lowerthan that of some other heterocyclicamines in smoke. We also detectedsignificant levels of MeIQx andDiMeIQx.

In a modified Ames test, one that ismuch more sensitive than our standardassay and uses two different strains ofbacteria, the fried meat extracts had30,700 revertants per gram (see box, p. 16 for a definition of “revertants”),whereas the fumes produced by fryinghad 10,400 revertants per gram of friedmeat. Thus, the fumes generated duringthe cooking of meat represent about one-third of the mutagenic activity measuredin the fried meat itself. It is important torecognize that the amount of mutagensinhaled is very low compared toconsuming solid, cooked meat.Nevertheless, the presence of toxiccompounds in meat fumes, even atrelatively low levels, could pose somerisk to food preparers who are exposed tothem for long periods over many years.

Cook to Manage Mutagens

Our research on food mutagens isnot specifically designed to generatepractical advice for diet- and health-conscious individuals. Many questionsremain unanswered in this highlycomplex field of investigation.Although food mutagens are extremelypotent, our preliminary estimates of riskare not alarming primarily because of

their low concentrations in food.Nevertheless, the amount of mutagensingested can be reduced by choice ofdiet and by modifying cooking practices.

Cooking Tips Summary• Fried beef has very high mutagenicactivity. Its popularity suggests that thisfood may be the most important sourceof heterocyclic amines in the typicalWestern diet.• Most, but not all, of the mutagenicactivity in fried beef can be accountedfor by known heterocyclic amines. Thesingle mutagen PhIP accounts for mostof the combined mass of mutagens infried beef cooked well-done.• The fumes generated during thecooking of beef have about one-thirdthe mutagenic activity measured in thefried meat itself. Occupational exposureover long periods could pose some risk,but probably much less than that fromconsuming the meat.• The fat content and thickness of meathave little effect on mutagen content,whereas the method and extent ofcooking have major effects. Frying,broiling, and barbecuing muscle meatsproduce more heterocyclic amines andmutagenic activity, whereas stewing,steaming, and poaching produce little orno mutagenic activity. Roasting andbaking show variable responses.• Both cooking temperature and timecan be manipulated to minimize theformation of mutagens. Increasing thefrying temperature of ground beef from200 to 250°C increases mutagenicactivity about six- to sevenfold.Reducing cooking temperature and timecan significantly lower the amounts ofmutagens generated and subsequentlyconsumed in the diet.• Microwave pretreatment of meat,followed by pouring off the clear liquidbefore further cooking, can substantiallyreduce the formation of heterocyclic

24


Food Mutagens

JAMES FELTON joined the Biomedical Sciences Division ofLawrence Livermore National Laboratory as a Senior BiomedicalScientist in 1976. He is currently the Group Leader of theMolecular Toxicology Group of the Biology and BiotechnologyProgram at the Laboratory. He received his A.B. in Zoology fromthe University of California, Berkeley, in 1967 and his Ph.D. inMolecular Biology from the State University of New York at

Buffalo in 1973. From 1969 until 1976, he was a Fellow of the National Institute ofHealth, initially in New York and later in Maryland.

In more than 147 professional publications, James Felton has explored the role ofdiet in carcinogenesis and mutagenesis. He has been a part of the Laboratory’sresearch on food mutagens since it began 17 years ago and has led it for the past 8 years.

About the Scientist

For further information contact James S. Felton (510) 422-5656 ([email protected]) or Mark G. Knize (510) 422-8260 ([email protected]).

To view this article with interactive links to these references, visit our Internet homepage athttp:llwww.llnl.gov/str/str.html. After August 1, click on references in color for immediate access toadditional specific information.

27


Research Highlight

The engine does not idle; rather, it shuts down each time theenergy-storage device is fully charged. To complete the powertrain, an electric motor is coupled to the wheels by a single-speed transmission. By turning the electric motor into agenerator during braking, our concept vehicle includes thefeature of regenerative braking. Thus, kinetic energy returns to the storage device when the brakes are applied.

If the engine/generator in a hydrogen-powered vehiclesupplies enough power for a fully loaded vehicle to climb hills at cruising speeds, then it performs much like today’s gasoline-powered automobiles. However, if theengine/generator supplies just enough power for averageenergy consumption, then it can serve as a range extender.The difference in power required for cruising versus hillclimbing is about a factor of four. We are designing a fullycapable concept car that can cruise and climb hills.

The Design Team’s ChallengesLLNL researchers are working on the technical details of a

new hydrogen piston engine with investigators at Los AlamosNational Laboratory and Sandia National Laboratories,California. Essentially, LLNL is responsible for the initialsystem studies, engine design, and combustion kinetics. Los Alamos investigators perform the computational fluid-dynamics modeling (combustion modeling) and integrate thisinformation into our vehicle simulation codes. Researchers atSandia’s Combustion Research Facility then do the engine-performance and emissions testing.

The need for a highly efficient vehicle and power train isdriven by the associated problem of onboard storage ofhydrogen fuel. Onboard fuel storage is perhaps the single mostdifficult task associated with our project. Table 1 shows twooptions we are considering for fuel storage: a cryogenic tankfor liquid hydrogen or a high-pressure tank for hydrogen gas.Without increased efficiency, the onboard fuel tank wouldneed to be about three times the volume listed in Table 1 and three times the size shown in the illustration; thatis, the tank would become so large as to be impractical. Weare applying the hybrid vehicle evaluation code (HVEC)developed at LLNL as a guide to select components thatmaximize efficiency and thus reduce fuel-tank volume andweight.

HVEC incorporates a wide range of details and complexity.The code calculates power-train dimensions, fuel economy,time to accelerate to 60 mph (96 km/h), hill-climbingperformance, and emissions. Our basic premise is that weneed to generate electrical energy at efficiencies of about42%, based on a generator that is 95% efficient and an engineefficiency of about 46%.

Our calculations show that an empty vehicle weighing 2508 lb (1140 kg) (see Table 1 for additional specifications)

would have a combined EPA urban/highway mileage of about80 mpg (expressed as gasoline-equivalent fuel efficiency).Such a vehicle would require only about 10.45 lb (4.75 kg) of hydrogen for a driving range of 380 miles (608 km). Forperspective, a kilogram of hydrogen has nearly the sameenergy content as a gallon of gasoline. Thus, our hydrogen-powered vehicle is extremely energy-efficient and hasemissions equivalent to those of electric vehicles when theemissions from power plants are included. And its gasoline-equivalent fuel efficiency of 80 mpg meets the goal set by PNGV.

With current technology, we believe that a general-purpose,low-emission, long-range vehicle that uses a hydrogen internalcombustion engine is now possible. Such a vehicle couldbecome competitive in the marketplace if hydrogen productionand distribution issues are addressed. These issues are beingstudied at the Laboratory and will be the subjects of Scienceand Technology Review articles in the future.

26


This Hybrid Vehicle Uses Hydrogen

ECENTLY, the Clinton Administration’sPartnership for a New Generation of

Vehicles (PNGV) set an automobile fuelefficiency goal of 80 mpg (34 km/L) toachieve responsible energy andenvironmental conservation. Even beforethe goal was announced, researchers atLLNL had joined investigators at Los AlamosNational Laboratory and Sandia NationalLaboratories, California, to design and test a hydrogenhybrid concept vehicle that will meet or exceed PNGVguidelines. The hydrogen piston engine they have designedgets mileage equivalent to the 80 mpg of a gasoline-poweredvehicle on the combined city–highway driving cycle.

Why Hydrogen Fuel?Hydrogen has several features that make it a serious

contender as an alternative fuel. It can be produced fromvarious domestic sources, including renewable sources; it can reduce emissions to near zero while maintainingperformance; and it can now be safely stored and transported.An immediate motive for moving to hydrogen is its potentialto improve urban air quality. In the longer term, such atransition would also benefit the balance of payments and theenergy security of the U.S. by reducing dependence onforeign oil.

Because hydrogen is a manufactured fuel, it is likely tocost more than fossil fuels for at least several decades. Thecost issue means that researchers need to exploit the use ofhydrogen fuel in those applications that have the highestleverage or payoff. One obvious application is intransportation. The energy efficiency of today’s automobilesis only about 18%.

Despite its many advantages, hydrogen has yet to becomea significant transportation fuel, even in advanced countries.Several factors hinder a transition from gasoline to hydrogen,including the absence of available vehicles with engines thatcan use this resource efficiently and the lack of an adequatedistribution infrastructure.

Hydrogen Fuel EfficiencyCurrent engine designs have low energy efficiency. Small

piston engines (in the range of about 40 kW or 54horsepower) have not been optimized specifically for

hydrogenfuel. The uniquecombustion properties ofhydrogen allow engines to runleaner and at a higher compressionratio than they do with hydrocarbon fuels. Energy efficiencyis a serious problem if consumers want a driving rangecomparable to that of today’s gasoline-powered vehicles.Thus, what we need are high-efficiency drive trains if we areto consider hydrogen seriously as an alternative fuel.Researchers at LLNL are showing that such drive trains arefeasible and that hydrogen has a genuine opportunity tocompete for the first time in the transportation sector.

Our studies demonstrate that considerable improvementover conventional automobile efficiency can be achievedthrough a hybrid-electric drive train. In this concept, all thechemical energy of the fuel is converted to electrical energyby means of a piston engine coupled to an electricalgenerator. The electrical energy can be stored in variousways, including an advanced battery, an ultracapacitor, or anelectromechanical battery (EMB), also known as a flywheelbattery. Of these three technologies, the EMB is closest tofull-scale demonstration. The flywheel battery, which will be the subject of a forthcoming article in Science andTechnology Review, has an energy recovery efficiency ofmore than 90% and a long lifetime. Compared to the EMB,today’s electrochemical batteries have an energy recoveryefficiency of about 70%.

How the Hybrid Hydrogen Vehicle WorksIn the hybrid concept vehicle we are developing (see the

illustration), stored electrical energy is extracted as needed bythe power demands for accelerating, cruising, and accessories.

This Hybrid Vehicle Uses Hydrogen

Research Highlights

R

Our conceptual design of a hydrogenhybrid vehicle features a large fuel tank forpressurized hydrogen. It has a gasoline-

equivalent fuel efficiency of 80 mpg anda driving range of 380 mi

(608 km).

Table 1. Some basic specifications and calculated performance

for the LLNL hydrogen hybrid vehicle.

General description

Five-passenger, engine-flywheel hybrid vehicleHydrogen internal combustion engineCryogenic or pressurized hydrogen-storage systemPrincipal accessory: air conditioning

Selected vehicle characteristics

Vehicle empty total weight 2508 lb (1140 kg)Power-train weight 578.6 lb (263 kg)Fuel-tank capacity 10.45 lb (4.75 kg)

of hydrogenLiquid-hydrogen tank volume @ 100 psi 28 gal (106 L)Liquid-hydrogen tank weight @ 100 psi 79 lb (36 kg) Pressurized-hydrogen tank volume @ 5000 psi 62 gal (235 L)Pressurized-hydrogen tank weight @ 5000 psi 141 lb (64 kg)

Aerodynamic-drag coefficient 0.24Rolling-friction coefficient 0.007Electric motor

Maximum continuous torque 100 N · mMaximum speed motor 11,000 rpm

Transmission efficiency 95%Hydrogen-engine efficiency 46%

Calculated performance

Combined 55% urban, 45% highwaygasoline-equivalent mileage ~80 mpg (34 km/L)

Driving range 380 mi (608 km)Time to reach 96 km/h (60 mph) 9.7 s

For further information contact J. Ray Smith (510) 422-7802 ([email protected]).

Fuel tank

Electric motor

Transmission

Flywheelbattery

Electrical generator

Two-cylinder motor

29


28


Research Highlights Weather Forecasting

make accurate short-term forecasts of precipitation andflooding. Between January 7 and January 11, three strongstorms hit California. Several areas experienced extensiveflooding as soils became saturated after the second and thirdstorms came ashore. The Russian River basin was among thehardest-hit areas, with an estimated $800 million in flood-related damage.

Large-scale forecast data (80-km resolution) from theNational Weather Service were used as input to the CARSsystem, and MAS simulations (20-km resolution) were runproducing precipitation fields for all of California for this timeperiod. MAS’s ability to calculate rainfall and snowfallseparately was essential for predictions of river flow, sincesnowfall does not immediately affect river flow.

California’s complex terrain can cause considerabledifferences in theprecipitation received byareas only a few milesapart. As a result, accurateestimates of localprecipitation are essentialfor accurate estimates ofriver flow in mountainousareas. To illustrate thisdependence, CARScomputations were madefor the area-averaged dailyrainfall for the entireRussian River basin(approximately 7000 km2)and compared withcalculations of the Hoplandwatershed (a smaller area,about 660 km2) within theRussian River basin, northof the Hopland gaugestation. The simulated dailyrainfall for the two areasdiffers by factors of two tothree (Figure 2a).

To evaluate CARS’sability to predict river flowand flooding, simulated precipitationvalues for the Hopland watershedwere compared with the observedprecipitation values for the first 12days of January (which were used bythe National Weather Service’sCalifornia–Nevada River ForecastCenter to model river flow). CARS

successfully simulated the amounts and timing of rainfall overthe Hopland watershed, except on January 10, where the modeloverestimated precipitation by a factor of two (Figure 2b). Uponfurther examination, this overestimation was found to haveresulted from excessive amounts of water vapor flux in theinput data for the CARS simulation, clearly demonstrating thedependence of regional predictions on accurate large-scale data.

Figure 2c plots the observed and simulated daily-meanriver flow volume of the Russian River at the Hopland gaugestation from January 1 through January 12. CARS simulatedthe river flow rate to within 10% accuracy during the floodstage. The overestimation of modeled river flow for January 11was due in part to the overpredicted rainfall for January 10, asnoted above. For the low flow periods before flooding,simulated river flow exceeded the observed river flow mainly

because of uncertainties inthe initial water content ofthe soils, a difficult variableto simulate.

These successfulpredictions of extremeprecipitation and river flowdemonstrate the applicabilityof the CARS system to short-term, local weatherforecasting. Such modelingwill not replace humanweather forecasters; rather,modeling can provideanother type of data to assistforecasters. As Jinwon Kim,one of CARS’s developers,remarks, “The value of front-line forecasting is that theforecasters have theexperience to interpret datafrom various sources. Ourgoal is to create a modelingsystem that can help improvethe accuracy of a forecastand the time span for whichit is valid.”

Improving short-term weatherforecasts is but one step toward thelong-range goal of understandingand predicting global climate changeand its regional impacts. Havingsuccessfully simulated the RussianRiver situation, CARS’s developersare moving ahead on several fronts.

Large-Scale Data

National Weather Service forecasts, global analysis data, general circulation model

Mesoscale Atmospheric Siomulation (MAS) Model

Regional atmospheric and land-surface information

TOPMODEL

Surface and subsurface hydrology at individual watersheds

Local Atmospheric and Hydrological Information

River flow, local weather, land-surface information

Temperature Humidity Rain Pressure

Rain and snow Radiation Temperature Humidity Wind Pressure

Runoff Soil water River flow

Hydrological characteristics of each watershed using digital

elevation data

Land Analysis System (LAS)

Figure 1. The CARS system. The MAS model takes

large-scale input data and telescopes it down to

simulate local precipitation and atmospheric variables,

which are then averaged over individual watershed

areas (obtained from LAS). LAS also computes

topographic characteristics for the watersheds.

TOPMODEL uses the precipitation and atmospheric

variables simulated by MAS together with the land

surface properties determined by LAS to compute river

flow and hydrology for the specific watersheds.

EATHER is fickle, especially in the varied terrains andmicroclimates of the western United States. California

and the other western states thrive or languish with their watersupply, as the pendulum swings between drought and deluge.All too often, “average” precipitation is merely an artifact ofarithmetic. Complicating the picture is the fact that the areareceives its year’s supply of water during the winter, and waterfor the dry summer must come almost entirely from reservoirstorage and mountain snowpacks. At the start of each winter,everyone—water district official, fire fighter, ski resortoperator, homeowner—wants to know if rainfall and snowfallwill be above or below average. Accurate assessments ofwintertime precipitation are particularly important for regionalwater management agencies as they attempt to managereservoir capacity and balance the water demands ofagricultural, industrial, urban, recreational, and environmentalinterests. In addition, since water supply is a limiting factor for urban and industrial development, regional planners areincreasingly concerned about the effect of global climatechange on local water resources.

Numerical simulation using general circulation models(GCMs) is one of the most important tools for understandingglobal climate and for projecting long-term climate change.Great strides have been made in recent years to couple modelsof atmospheric, terrestrial, and oceanic processes to providemore complete climate simulations. However, because of theircoarse spatial resolution (typically 100 km), it is difficult toapply these GCMs directly to regional forecasts. In California,for example, precipitation is closely related to topographicfeatures (e.g., the Coastal Range, the San Francisco Bay) withspatial scales of less than 100 km, too small to be resolved bya GCM. Increasing the resolution of the GCMs to provideregional simulations is beyond the capabilities of present andenvisioned computational resources.

Mesoscale models, nested within GCMs, are beingdeveloped to assess regional climate. As part of an effort toinvestigate regional-scale atmospheric flow, precipitation, andhydrology over various time scales and spatial resolutions,four LLNL researchers—Jinwon Kim, Norman Miller, DonaldErmak, and William Dannevik—have developed the CoupledAtmosphere-Riverflow Simulation (CARS) system. The

system consists ofthree unidirectionallycoupled models—MAS,LAS, and TOPMODEL(see Figure 1). CARS canbe nested either withinlarge-scale weather forecaststo predict regional weather andriver flow or within global climate analysis data to assessregional climate and long-term water resources.

The Mesoscale Atmospheric Simulation (MAS) model wasdeveloped jointly by LLNL and the University of Californiaat Davis.1 It models atmospheric processes, including thoseinvolved in storms, from which it computes localprecipitation, wind velocity, and other atmospheric variables.MAS computes rainfall and snowfall separately (using a bulkcloud microphysics scheme2), an important capability becausemountain snowpacks are major sources of summertime waterfor the western states.

The Land Analysis System (LAS) is a system of codestaken in part from software developed by the U.S. GeologicalSurvey3 and combined with numerous other codes and scriptsdeveloped at LLNL. It provides land surface characteristics(such as flow directions, topographic slopes, water channels,and hydrological characteristics) for individual watersheds,based on digital elevation data provided by LLNL’sAtmospheric Release Advisory Capability group. The areasand locations of the LAS watersheds are nested within thegrid points of the MAS model.

TOPMODEL is a hydrology model, developed originallyin 1979 at Lancaster University, England,4 and enhanced andexpanded over the years. LLNL’s version of TOPMODELtakes the watershed-averaged precipitation and atmosphericvariables from MAS together with the land surfacecharacteristics from LAS to simulate surface and subsurfacehydrology and river flow for individual watersheds.

The series of storms that struck Northern California inJanuary 1995 provided an effective test of CARS’s ability to

Modeling for More AccurateWeather ForecastsModeling for More AccurateWeather Forecasts

The terrain of theRussian River basin innorthern California, the

general area of ourrainfall modeling.

W

31


30


CARS’s hydrology simulation model is being extended toinclude other major river systems in California, specificallythe inflow to Lake Shasta, the Feather River, and theAmerican River. This expansion will make it possible to useCARS for simulating local weather and river flows overnorthern California’s major watersheds.

In collaboration with the National Weather Service, theCARS system is being used for experimental weatherprediction for the southwestern United States. Simulations arealso being run to test CARS’s ability to assess water resourcesover seasonal, multiyear, and decadal time scales, to model theeffect of such global phenomena as El Niño on regionalclimate, and to determine the effects of pollutants such ascarbon dioxide and aerosols on climate change.

References1. J. Kim and S.-T. Soong, “Simulation of a Precipitation Event in

the Western United States,” Proceedings of the 6th Conference on Climate Variations (January 1994), pp. 407–410. (UCRL-JC-114412)

2. H.-R. Cho et al., “A Model of the Effect of Cumulus Clouds onthe Redistribution and Transformation of Pollutants,” Journal ofGeographical Research 94 (ND10), 12,895–12,910 (1989).

3. S. K. Jenson and J. O. Domingue, “Extracting TopographicStructure from Digital Elevation Data for Geographic InformationSystem Analysis,” Photogrammetric Engineering and RemoteSensing 54 (11), 1593–1600 (1988).

4. K. J. Beven and M. J. Kirkby, “A Physically Based, VariableContributing Area Model of Basin Hydrology,” HydrologicalScience Bulletin 24, 43–69 (1979).

Weather Forecasting

Figure 2. (a) Simulated

precipitation for the entire

Russian River basin and

for the Hopland watershed.

(b) Observed and simulated

precipitation for the Hopland

watershed of the Russian

River basin. (c) Observed

and simulated river flow at

the Hopland gauge station

on the Russian River. The

CARS simulations are in

generally good agreement

with observed precipitation

and river flow, the

discrepancies

(overestimations) arising in

large part from inaccuracies

in the large-scale input data.

For further information contact Jinwon Kim (510) 422-1848 ([email protected]), Norman Miller (510) 423-1283 ([email protected]), Donald Ermak (510) 423-0146 ([email protected]), orWilliam Dannevik (510) 422-3132([email protected]).

01 2 3 4 5 6 7 8 9 10 11 12

250

500

750

1000

Rive

r flo

w vo

lum

e, m

3 /s

Hopland watershed Russian River basin

0

150

50

100

Prec

ipita

tion,

mm

/day

Simulated Observed

0

50

100

Prec

ipita

tion,

mm

/day

Hopland watershed Russian River basin

(b)

(a)

(c)

Day of January 1995To view this article with interactive links to these references, visit ourInternet homepage at http://www.llnl.gov/str/str.html. After August 1, clickon references in color for immediate access to additional specificinformation.

Each month in this space we report on the patents issued to and theawards received by Laboratory employees. Our goal is to showcasethe distinguished scientific and technical achievements of ouremployees as well as to indicate the scale and scope of the workdone at the Laboratory.

Patents and Awards

Patent issued to

Anthony M. McCarthy

Michael W. Droege, Paul R. Coronado, and Lucy M.Hair

Earl R. Ault and Terry W. Alger

Anthony M. McCarthy

Thomas C. Kuklo

Steven T. Mayer, James L. Kaschmitter, and Richard W. Pekala

Patent title, number, and date of issue

“Method for Forming Silicon on a Glass Substrate”

U.S. Patent 5,395,481Issued March 7, 1995

“Method for Making Monolithic Metal Oxide Aerogels”


“Metal Vapor Laser Including Hot Electrodes and Integral Wick”


“Method of Forming Crystalline Silicon Devices on Glass”


“Kinematic High Bandwidth Mirror Mount”


“Aquagel Electrode Separator for Use in Batteries andSupercapacitors”

U.S. Patent 5,402,306 Issued March 28, 1995.

Summary of disclosure

A method by which single-crystal silicon microelectronics may be fabricatedon glass substrates at low temperatures.

A method in which a metal alkoxide solution and a catalyst solution areprepared separately and reacted to produce transparent, monolithic metaloxide aerogels of varying densities.

A specifically designed electrode and wicking associated with the plasmatube of metal vapor lasers.

A method for fabricating single-crystal silicon microelectronic componentson a silicon substrate and transferring them to a glass substrate.

An adjustable high bandwidth mount for mirrors used in optical systems. The mount is adjustable along two perpendicular axes.

An electrode separator formed of aquagel with electrolyte in its pores forelectrochemical energy storage devices.

Patents

Awards

Dana Isherwood, the Laboratory’s legislative analyst, and Dick Post, aLaboratory Associate in the Energy Directorate, were elected fellows of theAmerican Association for the Advancement of Science (AAAS) inrecognition of their scientifically or socially distinguished efforts on behalfof the advancement of science or its applications.

Ralph Jacobs, director of New Technology Initiatives in the Laser Programat LLNL, was elected fellow of the American Physical Society. He washonored for “fundamental and applied contributions to the research anddevelopment for a wide variety of gaseous, solid, and liquid laser media.”

Tom McEwan and his “radar on a chip” were honored in April by theFederal Laboratory Consortium for excellence in transferring technologyfrom a laboratory to private business. The Consortium is an association ofDepartment of Energy research facilities that assists the U.S. public andprivate sectors in using technologies developed by federal researchlaboratories.

Secretary of Energy Hazel O’Leary presented Laboratory representativeswith the Management and Operation Contractor of the Year Award on

March 31 in recognition of its outstanding achievement in providingsubstantial contracting opportunities for small businesses. She cited oursocioeconomic program assisting small, women- and minority-ownedbusiness in securing procurement contracts with the Laboratory as the best ofits kind in the DOE complex.

The 1994 E. O. Lawrence Award has been awarded to Michael Campbell,head of the Laboratory’s lasers program, and John Lindl, scientific directorfor Inertial Confinement Fusion for distinguished leadership in helping topropel laser-driven inertial confinement fusion to the forefront of physicsresearch. The award was established in 1959 in memory of Ernest O.Lawrence to recognize outstanding contributions in the field of atomicenergy. Dr. Campbell was also the winner of the 1995 Edward TellerMedal. This award was established in 1989 to commemorate Teller’scontributions to fusion energy.

The Northern California Section of the American Institute of ChemicalEngineers has named a hazardous explosives cleanup process developed atthe Laboratory as Project of the Year. The award cited the project’s principalinvestigators Ravi Upadhye, Bruce Watkins, Cesar Pruneda, and BillBrummond. The process uses molten salt to safely dispose of wasteexplosives and explosive-like materials.


32


U.S. Government Printing Office: 1995/683-074-20015

Research on Food Mutagens

Of all the toxic substances produced during cooking, themost important are likely to be the heterocyclic amines. For17 years, LLNL researchers have been identifying these foodmutagens, measuring their abundance in cooked foodstypical of the Western diet, working to understand how theycan trigger malignant tumors in laboratory animals that havebeen exposed to high mutagen doses, and estimating theimportance of human exposures. Our success is largely afunction of the interdisciplinary approach we have taken toquantify food mutagens and to study their biological effects.LLNL investigators were the first to identify five of the mostimportant mutagens in heated food, including PhIP andDiMeIQx. We have shown that fried beef may be the mostimportant single source of heterocyclic amines in the humandiet and that PhIP accounts for most of the combined mass of

mutagens in fried beef cooked well-done. Most nonmeat foodscontain low or undetectable levels of these types ofcompounds, but some cooked protein-containing foods, suchas those high in wheat gluten, have significant levels ofunknown aromatic amine mutagens. Cooking time andtemperature significantly affect the amounts of mutagensgenerated. For example, reducing the frying temperature ofground beef from 250 to 200°C lowers the mutagenic activityby six- to sevenfold. Microwave pretreatment of meat anddiscarding the liquid that is formed also greatly reduces theformation of heterocyclic amines. Our related work on doseand risk assessment will be described in a forthcoming article.

Contacts: James S. Felton (510) 422-5656 ([email protected]) or Mark G. Knize (510) 422-8260 ([email protected]).

Abstract

4. How to receive more informationAdditional information, generally written for an expert scientificaudience, is available for the following articles. Please checkthe articles about which you would like more information.

❑ Food Mutagens: The Role of Cooked Food in Genetic Change

❑ This Hybrid Vehicle Uses Hydrogen❑ Modeling for More Accurate Weather Forecasts

5. Subscription information❑ Address correction ❑ New subscription(Please check appropriate box)❑ Please drop me from your mailing list

Name ____________________________________________

Title ______________________________________________

Organization ______________________________________

Address __________________________________________

City, State, Zip ______________________________________

Date you received this issue ________________________

Please respond by July 31, 1995, toFAX (510) 422-8803

Lawrence Livermore National LaboratoryP.O. Box 808, L-664Livermore, CA 94551

Science & Technology Review FAX Survey

1. Rate your interestHow interesting did you find the following articles?

very interesting interesting minimal none

Food Mutagens ❑ ❑ ❑ ❑Hydrogen Hybrid Vehicle ❑ ❑ ❑ ❑Weather Forecasts ❑ ❑ ❑ ❑

How would you rate the technical level of the content?acceptable too high too low didn’t read

Food Mutagens ❑ ❑ ❑ ❑Hydrogen Hybrid Vehicle ❑ ❑ ❑ ❑Weather Forecasts ❑ ❑ ❑ ❑

How interesting and valuable do you find our new sections?very interesting interesting minimal none

Laboratory News ❑ ❑ ❑ ❑Highlights ❑ ❑ ❑ ❑Patents/Awards ❑ ❑ ❑ ❑Commentary ❑ ❑ ❑ ❑

Comments ________________________________________

__________________________________________________

__________________________________________________

__________________________________________________

2. Please check the box that most closely describes yourorganization❑ DOE❑ Congressional leader or staff❑ LLNL employee❑ Other scientific or technical laboratory❑ Government agency or commission❑ Industry❑ Staff or student at educational institution❑ Local, county, or state official or staff❑ UC Regents or other official❑ Other______________________

3. Electronic accessDo you have access to the Internet?

❑ yes ❑ noIf so, did you try to use the electronic links to the references?

❑ yes ❑ noWere you able to access the references?

❑ yes ❑ no

What Do You Think?This issue of Science and Technology Review (formerly Energy and Technology Review)

represents a modified approach to communicating the work of the Laboratory. The intent ofour changes is to make this publication more interesting to a broader audience. Please give usyour reactions to our changes by answering the questions in section 1 and 2 below and faxingthem back to us at (510) 422-8803. You can also mail your response to the address below. Ifyou photocopy this form first, you won’t have to remove it from the publication.

Also note that this form can be used to ask for more information, to change address, to stopyour free subscription, or to add a colleague to the circulation. If you are requesting any ofthese, be sure to include your name, etc. in section 5.

R E V I E W

Documents

The Search for Mutagens from Cooked Foods