martin karplus

Boardpage ARI 12 April 2006 0:2

Martin Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

Spinach on the Ceiling:A Theoretical Chemist’sReturn to BiologyMartin KarplusDepartment of Chemistry and Chemical Biology, Harvard University, Cambridge,Massachusetts 02138; and Laboratoire de Chimie Biophysique, ISIS, Universite LouisPasteur, F67083 Strasbourg, France; email: [email protected]

Annu. Rev. Biophys. Biomol. Struct.2006. 35:1–47

The Annual Review ofBiophysics and BiomolecularStructure is online atbiophys.annualreviews.org

doi: 10.1146/annurev.biophys.33.110502.133350

Copyright c© 2006 byAnnual Reviews. All rightsreserved

1056-8700/06/0609-0001$20.00

Key Words

molecular dynamics, NMR coupling constants, trajectorycalculations, reaction kinetics

AbstractI was born in Vienna and came to the United States as a refugeein October 1938. This experience played an important role in myview of the world and my approach to science: It contributed tomy realization that it was safe to stop working in fields that I feltI understood and to focus on different areas of research by askingquestions that would teach me and others something new. I describemy experiences that led me from chemistry and physics back to myfirst love, biology, and outline some of the contributions I have madeas part of my ongoing learning experience.

1

ANRV275-BB35-01 ARI 10 April 2006 15:34

Contents

EARLY YEARS IN EUROPE . . . . . . . 2A NEW LIFE IN AMERICA. . . . . . . . 5BEGINNING OF SCIENTIFIC

INTERESTS . . . . . . . . . . . . . . . . . . . . 7COLLEGE YEARS . . . . . . . . . . . . . . . . . 11POSTDOCTORAL SOJOURN IN

OXFORD AND EUROPE . . . . . . . 17FIVE YEARS AT THE

UNIVERSITY OF ILLINOIS:NMR AND COUPLINGCONSTANTS. . . . . . . . . . . . . . . . . . . 19

MOVE TO COLUMBIA ANDFOCUS ON REACTIONKINETICS . . . . . . . . . . . . . . . . . . . . . . 21

RETURN TO HARVARDUNIVERSITY AND BIOLOGY . 27

HEMOGLOBIN: A REALBIOLOGICAL PROBLEM . . . . . . 31

PROTEIN FOLDING . . . . . . . . . . . . . 33ORIGINS OF THE CHARMM

PROGRAM . . . . . . . . . . . . . . . . . . . . . 34THE FIRST MOLECULAR

DYNAMICS SIMULATION OFA BIOMOLECULE . . . . . . . . . . . . . 36

APPLICATIONS OFMOLECULAR DYNAMICS. . . . . 38

FUTURE OF MOLECULARDYNAMICS. . . . . . . . . . . . . . . . . . . . . 39

EPILOGUE . . . . . . . . . . . . . . . . . . . . . . . . 40

EARLY YEARS IN EUROPE

I was born in Vienna, Austria, in 1930 andseemed destined to become a physician. Forseveral generations, there had been one ormore physicians in the family, partly becausemedicine was a profession in which Jewsin Austria could work with relatively littlehindrance from discrimination. Neither mybrother nor any of my numerous cousins dis-played any interest in becoming a doctor, un-like me who at the age of five went aroundbandaging chairlegs and other substitutes forbroken bones. I was fascinated by the sto-

ries of various relatives, including my Un-cle Paul, who was a superb clinician. So, myfamily concluded that I was to become “the”doctor.

My paternal grandfather, Johann PaulKarplus, was a professor at the medical schoolof the University of Vienna, where his re-search led to the discovery of the functionof the hypothalamus. My maternal grandfa-ther, Samuel Goldstern, ran a private clinicthat specialized in the treatment of rheuma-tological diseases by use of a mildly radioac-tive mud. The clinic, which I visited often be-cause my grandparents had their apartmenton an upper floor of the building, was calledthe Fango Heilanstalt. I had always assumedthat Fango was a made-up name and learneddecades later that fango is nothing more thanthe Italian word for mud. Many of the patientsat the Fango were wealthy Arabs from theMiddle East, an example of how the history ofArabs and Jews has long been intertwined andnot always in the present destructive manner.

My childhood home was located in theViennese suburb named Grinzing, wellknown as a wine growing area. There weresmall informal inns (Heurige) where wesometimes went in family groups for relaxedevenings eating and drinking (mainly theadults) the fruity young white wine of theregion. The Heurige were also great placesfor children because between eating sausages,cheese, and bread, they could play in thegarden.

In the early 1930s, when owning an auto-mobile was still relatively rare, we already hada small car, a “Steyer Baby.” One day whenit was parked in front of our house, I scootedinto the driver’s seat and pretended to drive. Iinadvertently released the brake, and the carstarted to roll downhill. I was terribly fright-ened as the car approached a pit at the endof the street. Miraculously, I steered the carso it turned just before reaching the pit andstopped. I recall the slope of the hill as be-ing steep and the pit as very deep. Forty-fiveyears later on a visit to Vienna with my fam-ily, we went to see my childhood home, which

2 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

had been appropriated by the Nazis duringthe Second World War. I discovered that the“steep” hill was a very gentle slope and the“pit” a shallow ditch, so that the danger wasmore in my young mind than in reality.

Stories from my early childhood, most ofwhich I know from their retelling by my par-ents, aunts, and uncles, indicate that I was astrong-willed independent child (in a positivesense) and a brat (in a negative sense). Onesuch story concerns my “escape” at the age ofthree from a summer daycare center, whereI apparently did not like the way I was be-ing treated. One morning I simply walked outand somehow, despite the center being severalmiles from our house, I made my way home.Both the daycare people and my parents, whohad been notified, were extremely worriedabout my disappearance and were searchingfor me. Although I was scolded, my parentswere so happy to see me that my punishmentconsisted of a mild reprimand. The venturewas justified in my mind because after thatI was allowed to stay home. Then there wasthe infamous “spinach incident.” My belovednanny, Mitzi, told me that I must eat myspinach. (Popeye did not exist in Austria, butunfortunately spinach did.) With all the ve-hemence I could muster, I took a spoonful ofthe spinach and threw it at the ceiling. Thespinach stain remained visible on the ceilingfor a long time and was pointed out at appro-priate moments when my parents wanted toindicate what a naughty child I was.

Traditionally we summered at an Austrianlake or on the Adriatic coast of Italy withseveral families that were either relatives orfriends with children of similar ages. Such ex-tended family activities were an integral partof my growing up and gave me confidence anda sense of belonging. One day at the beach, afriend of my parents picked me up and cud-dled me, much to my dismay. I yelled out, “Ichbin ein Nazi” (“I am a Nazi”), which so shockedher that she dropped me. Clearly, I had some-how realized, presumably from listening to myparents and others, that being a Nazi was theworst possible thing to be.

Already before the Nazis entered Austriain 1938, our life had changed significantly,even from the viewpoint of an eight year old.Among our neighbors were two boys of agescomparable to my brother, Robert, and me.They were our “best friends,” and we playedregularly with them. In the spring of 1937,they suddenly refused to have anything to dowith us and began taunting us by calling us“dirty Jew boys” when we foolishly continuedto try to interact with them. Similar problemsoccurred at school with our non-Jewish class-mates. Prior to this, my school experience infirst grade and the beginning of the secondhad been wonderful, in part because I had agreat teacher, Herr Schraik, not the least ofwhose outstanding attributes was that his wiferan a candy store. When my class was readyto advance to second grade, the parents peti-tioned that Herr Schraik be “promoted” withus and, because of his outstanding record as ateacher, this request was granted. Neverthe-less, in the middle of that school year (1937–1938) he was no longer allowed to teach. Hewas Jewish and the authorities had decidedthat any contact with him would contaminatethe minds of the children. The new teacherwas incompetent and blatantly anti-Semitic—he constantly criticized the Jewish studentslike myself, independent of how well or poorlywe were doing. The situation became so badthat my parents took me out of school.

One small, but not insignificant, part ofmy schooling in Austria had to do with mybeing left-handed. When I started first grade,I was obliged to learn to write with my righthand. Whatever the supposed psychologicalconsequences of that may be, I have alwaysbeen grateful that right-handedness in writingwas imposed on me. This was particularly truewhen I first went to the United States andsaw the contortions children went through towrite with their left hand. Clearly everything,at least in Western European languages (notHebrew or Arabic), is set up for right-handedpeople.

On March 13, 1938, the German Nazitroops crossed the border into Austria and

www.annualreviews.org • From Chemistry Back to Biology 3

ANRV275-BB35-01 ARI 10 April 2006 15:34

completed the Anschluss, the “joining” of Aus-tria with Nazi Germany, which had beenspecifically forbidden by the Versailles Treatyafter the First World War. The Germans hadbeen “invited” into Austria by the puppetSeyss-Inquart, who took over the Austriangovernment after Chancellor Schuschniggwas forced to resign. The night before, myfamily and some friends were listening to theradio in the living room, which was darkenedto conform to the curfew and black-out re-quirements in case of air attacks. I was furious,not because of what was happening politically,but rather because my parents had planned anearly birthday party (my actual birthday beingMarch 15), and I was far from being the centerof attention. Hitler entered Austria in triumphand his troops were welcomed by enthusiasticcrowds. (To this day, more than 65 years later,I have mixed feelings about visiting Austria,which I rarely do, because anti-Semitismseems nearly as prevalent now as it was then.However, I recently learned that I am still of-ficially an Austrian citizen, so I have dual na-tionality.) A few days after the Anschluss, mymother, brother, and I left Austria by train forSwitzerland. My parents had been concernedabout Hitler’s takeover of Austria for sometime. For the previous three years, my AuntClaire, who had studied in England, had beenteaching English to me and my brother, Bob.Well before March 13, train tickets had beenpurchased and a bed-and-breakfast “pension”had been reserved in Zurich.

The most traumatic aspect of our depar-ture was that my father was not allowed tocome with us and had to give himself up to beincarcerated in the Viennese city jail. In part,he was kept as a hostage so that any money wehad would not be spirited out of the country.My mother reassured my brother and me, say-ing that nothing would happen to him, thoughof course she herself had no assurance that thiswas true. At that time, the Nazi governmentstill allowed Jews to leave Austria, as long asthey left their money behind. One way of spir-iting money out of the country was to buy di-amonds and hide them in clothing and food.

During our train trip to Zurich, the guards atthe German/Austrian border meticulously re-duced a beautiful large sausage to thin slices,presumably searching for hidden diamonds.Fortunately, the sausage could still be eaten,and Bob and I were not too bothered by thisevent.

My parents hoped to go to America, butvisas to the United States were granted onlyto applicants who had an “affidavit,” a docu-ment from an American citizen guaranteeingtheir financial support. Many would-be im-migrants would have been allowed to leaveAustria (or Germany) but were not permit-ted to enter the United States. Some of themended up in other countries (South America,Australia, and New Zealand), but, as historyhas recorded, many were not able to leave anddied in concentration camps. My father’s olderbrother Edu had immigrated to the UnitedStates some years earlier and had become chiefengineer at the General Radio Corporation inBoston, where he invented the Variac, then awidely used device for continuously varyingthe electric voltage. The president of GeneralRadio, Mr. Eastman, provided the requiredaffidavit, enabling us to obtain visas for theUnited States.

Arranging for the journey to the UnitedStates took several months. After leavingAustria, my mother, brother and I lived inZurich. Bob and I enrolled in a neighbor-hood public school where we rapidly learnedSchwyzertutsch (the Swiss German dialect).Speaking Schwyzertutsch was not only a wayof belonging, but it also provided us with asecret language which my mother did not un-derstand. When summer came, we left Zurichand went to La Baule, a beach resort in Brit-tany, France, on the Atlantic coast, whereour Uncle Ernst Papanek had established asummer colony for refugee children. Thechildren were mainly from Jewish families,though there also were some whose parentswere political refugees. During the late 1930sand early 1940s, Ernst organized a number ofthese children’s homes, which saved the livesof many children. He described these efforts

4 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

in his autobiography, Out of the Fire. Ernst’sphilosophy was that one could rely on thecommon sense and intelligence of children;one of the hallmarks of the homes was thatthey were run as cooperatives, with the chil-dren’s input playing a significant role. Bob andI, and our cousins Gus and George (Ernst’schildren), as well as newfound friends, spenta blissful summer in La Baule swimming andbuilding sand castles. Although food was inshort supply, we were well nourished. I did notrealize it then, but my mother and the othergrownups were extremely worried about thefuture. Somehow they kept this from us andgave us a happy summer. Both Bob and Ilooked back on this period as a wonderful ex-perience. Roberto Benigni’s film Life Is Beau-tiful brought back memories of those days.

The visas finally arrived, passage wasbooked, and the three of us were ready toleave for the United States. Although therehad been no news from my father, he miracu-lously turned up at Le Havre a few days beforeour ship, the Ile de France, was scheduled todepart for New York. From my point of view,it was exactly what my mother had told mewould happen: We would all go to America to-gether. When my father joined us in Le Havre,Bob and I asked him what jail had been like.He told us that he had been treated well in jailand cheerfully described how he had passedthe time teaching the guards to play chess.One aspect of my father’s personality, whichstrongly influenced both my brother and me,was to make something positive out of anyexperience.

We never learned the full details of my fa-ther’s release from jail. My Uncle Edu hadposted a $5000 bond for his release and af-ter the war, several Viennese (e.g., a jailer,someone concerned with running the pris-ons) wrote us claiming credit for his release.There was no evidence as to who had donewhat, if anything. Nevertheless, my parentssent CARE packages to all the claimants andwished them well. (CARE, Cooperative forAssistance and Relief Everywhere, was a hu-manitarian organization that distributed non-

perishable food packages after the SecondWorld War to help overcome food shortagesin the war-torn countries.) My parents di-rected food packages to people who mighthave helped free my father, as well as to manyother people we knew (e.g., my nanny, Mitzi)who had survived the war.

A NEW LIFE IN AMERICA

We arrived in New York Harbor early in themorning on October 8, 1938, and I stood onthe deck watching the Statue of Liberty ap-pear out of the mist. The symbolism asso-ciated with the Statue of Liberty may seemtrite today (and somewhat deceptive given ourpresent immigration policies), but in 1938 itwas special for me. Most of the immigrationformalities had been taken care of by UncleEdu, so that a few hours after our arrival weboarded a train to Boston. (My uncle had ahouse in an exclusive part of Belmont, a sub-urb of Boston, which at that time did not al-low Jews to live there. We rarely were invitedto visit him when we first came, and whenwe did we had to hide our “Jewishness”; inparticular, we were not to say anything beforeentering the house to avoid the neighbors’ be-ing aware of our foreign accent.) During ourinitial weeks in the United States, we werelodged in a welcoming center in Brighton,where a large mansion had been transformedinto an interim home for refugee families. Wewere taught about America (what it was likefor foreigners to live in Boston), given lessonsto improve our English, and aided in the stepsrequired to be allowed to remain in the UnitedStates as refugees.

Soon we were ready to start a new life. Myparents rented a small apartment in Brighton(part of Greater Boston), and Bob and I im-mediately entered the local public schools, aswe had in Zurich. I was in the third gradeand had the good fortune to have a teacher (Ihad a crush on her) who gave me special En-glish lessons after school. At the age of eight,my English advanced so rapidly by being inschool and by playing with the neighborhood


ANRV275-BB35-01 ARI 10 April 2006 15:34

children that these special lessons, alas, lastedonly a few months.

I very much wanted to be accepted by mynew country, and while living in Allston I wasa street kid in every sense, hanging aroundwith my friends, playing stick-ball and othersuch games, occasionally stealing candy justfor the fun of it. For a while, I refused to speakGerman at home, despite my parents’ limitedEnglish. One afternoon when I was playingin the street, I tripped in my haste to get outof a car’s way and ended up with my foot un-der the car’s rear tire. The driver had stoppedand gotten out to see what was wrong, butwith me screaming and crying it took him awhile to understand that he should move thecar off my foot. Once he did, thankfully myfoot was only mildly sore. He wanted to driveme home and report to my parents what hadhappened, but I insisted that I was fine andwould get home by myself. It was true that Iwas not hurt, but my primary concern was tokeep my parents from knowing what had hap-pened, how I spent my time, and specificallythat I played in the streets.

After we came to the United States, ourcomfortable life in Vienna was a thing of thepast. We were now relatively poor, althoughmy father had brought some money in theform of valuable stamps, which he had pur-chased while we were still in Austria. Despiteour economic straits, my parents did every-thing to ensure that our lives were as un-changed as possible. The first summer wewere in the States, both of my parents workedas domestics—my father as a handyman andmy mother as cook and cleaning woman—fora wealthy family who had a summer house inNew Hampshire. We lived in a small house onthe grounds, where Bob and I had an idyllicsummer. The next summer my parents weresimilarly employed at a boys’ camp, again toenable Bob and me to spend the summer ascampers. My parents both took courses to helpthem find better employment. My father stud-ied machining at the Wentworth Institute andmy mother studied home economics at Sim-mons College. With the United States enter-

ing the War, employment was high, and myfather joined an airplane pump factory after hegraduated from a one-year course. He rapidlyadvanced in the company to become a qualityinspector and worked at this company untilhe retired 20 years later. My father frequentlytold stories to Bob and me about problems hehad solved or how he had suggested improve-ments in the pumps. The way he described hisideas (he had been educated in physics at theUniversity of Vienna) helped arouse our sci-entific curiosity. After finishing an undergrad-uate program, my mother found a position asa hospital dietician, similar to the type of workshe had done in Vienna at the Fango, her fa-ther’s hospital. She continued her educationand received a Master’s degree at the age of 65.

Motivated by their concern for our ed-ucation, my parents moved to Newton (asuburb of Boston), where the schools wererecognized as superior to the Boston publicschools. My parents bought a small house in apleasant neighborhood in West Newton, andI attended the Levi F. Warren Junior HighSchool. To my knowledge, we were the firstJewish family to live there. One Saturday, af-ter we had lived in West Newton for aboutsix months, an FBI agent knocked on thedoor and politely requested to see my father.The agent explained that he was investigat-ing a complaint from our next-door neigh-bor, who had telephoned the FBI to reportthat every morning as he was leaving for work,my father would step out on the front porch,turn around, and make the Nazi salute whileshouting “Heil Hitler.” The FBI agent ap-peared rather embarrassed and said that he re-alized that such an accusation against a Jewishrefugee from Nazi Austria was ridiculous. Af-ter some questioning (about our family historyand present status), he got up to leave and toldus that nothing further would happen.

My junior high teachers soon realized thatI was bored with the regular curriculum, sothey let me sit in the back of the classroom andstudy on my own. What made this experienceparticularly nice was that another student, avery pretty girl, was given the same privilege,

6 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

and we worked together. The arrangementwas that we could learn at our own pace with-out being responsible for the day-to-day ma-terial but had to take the important exams.Several dedicated teachers at Warren JuniorHigh helped us when questions arose, particu-larly with science and mathematics. With thisfreedom, we explored whatever interested usand, of course, did much more work than wewould have done if we were only concernedwith passing the required subjects.

In nonacademic activities, I participatedlike everyone else (I was not particularly goodin sports, though I enjoyed soccer a lot) andmade a number of friends that formed a close-knit group throughout high school. As part ofour education, we had to choose several tech-nical courses. The two I chose were printingand home economics, the latter not becauseit was essentially all girls, but because the stu-dents did real cooking. I had become inter-ested in cooking early on and used to spendtime in the kitchen with my mother andgrandmother, who both cooked simply butwell with the freshest ingredients. The finalexam had us prepare a dinner for the class,with each group responsible for one course.

BEGINNING OF SCIENTIFICINTERESTS

Soon after we moved to Newton, Bob wasgiven a chemistry set, which he augmentedwith materials from the school laboratory anddrug stores. He spent many hours in the base-ment generating the usual bad smells andmaking explosives. I was fascinated by his ex-periments and wanted to participate, but heinformed me that I was too young for suchdangerous scientific research. My plea for achemistry set of my own was vetoed by myparents because they felt that this might notbe a good combination—two teenage boysgenerating explosives could be explosive! In-stead, my father had the idea of giving mea Bausch and Lomb microscope. Initially Iwas disappointed—no noise, no bad smells,although I soon produced the latter with the

infusions I cultured from marshes, sidewalkdrains, and other sources of microscopic life.I came to treasure this microscope, and morethan 60 years later it is still in my possession.One especially rewarding aspect of my work-ing with the microscope was that my father,who was a thoughtful observer of nature (heliked to fish with a simple line, not to catchfish, but because it gave him an excuse to sit bya stream and watch the fish and observe theirbehavior), spent a lot of time with me and wasalways ready to come and look when I haddiscovered something new. I had found an ex-citing new world and looked through my mi-croscope whenever I was free. The first timeI saw a group of rotifers I was so excited bythe discovery that I refused to leave them, noteven taking time out for meals. They were themost amazing creatures as they swam acrossthe microscope field with their miniaturerotary motors. (The rotifers come to mind to-day in relation to my research on the small-est biological rotatory motor, F1-ATPase.) Myenthusiasm was sufficiently contagious that Ieven interested some of my friends. It was aspecial occasion when they came to my houseand looked at the rotifers through the micro-scope.

This was the beginning of my interest innature study, which was nurtured by my fatherand encouraged by my mother, even thoughit was still assumed that I would go to medicalschool and become a doctor. One day my clos-est friend, Alan MacAdam, saw an announce-ment of the Lowell Lecture Series (a Bostoninstitution, originally supported by a Brah-min family—the Lowells), which organizedevening courses on a wide range of subjectsat the Boston Public Library that were freeand open to the public. The organizers invitedexcellent lecturers from the many universitiesin the Boston area, as well as from nonaca-demic disciplines. The series that had caughtAlan’s eye was entitled “Birds and Their Iden-tification in the Field,” to be given by LudlowGriscom, the curator of ornithology at theMuseum of Comparative Zoology at HarvardUniversity. Alan and I occasionally walked in


ANRV275-BB35-01 ARI 10 April 2006 15:34

the green areas in Newton, particularly theNewton Cemetery, and looked for birds withmy father’s old pair of binoculars. Togetherwe attended the first lecture, which had agood-sized audience, although it was not clearwhether most of the people came simply tohave a nice warm place in winter rather thanbecause of their interest in birds. I was en-thralled by the lecture, which provided in-sights into bird behavior and described thelarge number of different species one couldobserve within a 50-mile radius of Boston. Iwas amazed that it was possible to identify agiven species from “field marks” evident evenfrom a glimpse of a bird, if one knew how andwhere to look. Alan did not attend the sub-sequent lectures, but I continued through theentire course. At the end of the fourth or fifthlecture, Griscom came up to me and askedme about myself. He then invited me to joinhis field trips, and a new passion was born.From that time on, my treasured microscopewas relegated to a closet, and I devoted myfree time to observing birds on my own, aswell as with Griscom and his colleagues, withthe Audubon Society, and other groups thatorganized field trips.

The culmination of these trips was the an-nual “census,” usually held at the height ofthe bird migration in May. This was an ac-tivity sponsored by the Audubon Society, andthe objective was to observe (see or hear) thelargest number of different species within agiven 24 hours. Each year, Griscom organizedone such field trip, inviting only a select groupof “birders” to participate. The census lasteda full 24 hours, starting just after midnightto find owls in the woods and rails and otheraquatic species in the swamps. There was acarefully planned route, based on known habi-tats and recent sightings of rare species. Thespecific itinerary was worked out in a meet-ing, at which everyone contributed what in-teresting birds they had sighted recently, butGriscom made the final decision on how wewould proceed. As the youngest (by far) inthe group, I was assigned special tasks. One ofthese tasks—perhaps not the most pleasant—

was to wade into the swamp at night (fortu-nately there was a moon) and scare up birds sothat they would fly off and could be identifiedby their calls. On this census trip we found 160or so different species, a record at the time forthe area.

I became intrigued by alcids, of which thenow extinct, flightless Great Auk was the mostspectacular member of the family. I persuadedmy family to go for summer vacation to theGaspe Peninsula in Canada because there wasa famous rocky island just offshore that hadnesting colonies of two kinds of alcids, puffinsand guillemots, as well as gannets, spectacularlarge black and white sea birds. Although onecould not visit the island sanctuary (it was for-bidden in order to protect the nesting birds),I borrowed a telescope from the Audubon So-ciety to view the birds. We drove through theGaspe Peninsula by car and spent the nights inbed-and-breakfast places. One strong impres-sion of the trip through New Brunswick andthe Gaspe area was that the houses in mostof the villages were in much poorer condi-tion than those on the coast, which were sup-ported by the tourist trade. Moreover, eachsuch small village was dominated by an out-sized church, indicating the power of religionin these communities.

Many of the alcids go south to NewEngland for the winter. Usually, they are farout to sea, but when there is a storm they arelikely to be blown close to shore, so that suchstorms (particularly Nor’Easters) present thebest opportunity to see rare species, such asthe tiny Little Auk, which is only eight inchesin length but can survive in the roughest seas.In the winter of 1944, school was closed be-cause of a heavy snowstorm, and I took theearly morning train up to Gloucester, a townon Cape Ann north of Boston, and hiked outto the shore. I sat on the steps of one of thelarge mansions, shuttered for the winter, andlooked out to sea through my binoculars. Theday started well as I quickly spotted several al-cids in the cove below. As I was getting chilledafter a couple of hours and ready to walk backto the train station, a car pulled up. A couple of

8 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

men got out and walked toward me. At first,I naively presumed they were other birders,interested in what I had seen. I soon realizedthat they did not look like birders—no binoc-ulars for one thing and not really dressed fora day in the snow. Moreover, they approachedme in a rather aggressive manner, asking mewhat I was doing and why. They did not be-lieve that I was sitting in such a storm lookingfor birds. Shortly after, they showed me theirpolice badges and bundled me into the car,not for a ride to the train station, but insteadto the Gloucester police station. I was only14 years old and very frightened and becameeven more so when I realized from their ques-tions that they thought I might be a Germanspy. That I was an immigrant from Austria,spoke German, and had German-made Zeissbinoculars did not help. They suspected meof signaling to submarines off the Americancoast, preparing to land saboteurs or what-ever. It took hours of interrogation and sev-eral phone calls to the Audubon Society beforethe officers finally decided that I was not do-ing anything wrong and drove me to the trainstation. That was the last time I ventured onsuch a trip by myself.

On one of the field trips to Newbury-port with Griscom, I spotted an unusual gull.When I pointed it out to him, he concluded,after looking through his telescope, that hehad never seen a bird like that and that weshould try to “collect” it, a euphemism forshooting the bird. He had a license to carrya “collecting gun” with a pistol grip and along barrel that made it easier to aim. Becausethe bird was far away and separated from usby mud flats, only partly exposed at low tide,I was given the task of collecting the bird. Iwaded out fairly close to the bird and success-fully shot it, even though I had never fired agun other than at fairs. After a careful com-parison with birds in the Museum of Com-parative Zoology collection, Griscom wasconvinced that we had found something new,a hybrid between a Bonaparte’s gull (commonin America) and a European black-headedgull (common in Europe but rare in North

Figure 1A photograph of me with the hybrid gull showing its wing feathers, whichplayed the essential role in its identification.

America), which had somehow crossed theocean. The bird was in the museum collec-tion with its wing feathers, which were essen-tial for its identification, beautifully spread out(Figure 1). (I say “was” because when I wentrecently to find it, the gull had disappeared.My reason for going to look for it was todiscuss with Scott V. Edwards, the newly ap-pointed Professor of Ornithology at Harvard,the possibility of sequencing the DNA tofind out whether the result would confirmGriscom’s identification.)

I entered Newton High School in the fallof 1944 but soon found that I did not have thesame supportive environment as in elemen-tary and junior high school. My brother, Bob,


ANRV275-BB35-01 ARI 10 April 2006 15:34

had graduated from Newton High School twoyears before and had done exceedingly well.My teachers presumed that I could not mea-sure up to the standards set by my brother.Since I had always been striving to keep upwith Bob and his friends, this just reinforcedmy feelings of inferiority. Particularly un-pleasant were my interactions with the chem-istry teacher. When my brother suggested Icompete in the Westinghouse Science Tal-ent Search, the chemistry teacher, who was incharge of organizing such applications, toldme that it was a waste of time for me to enterand that it was really too bad that Bob had nottried instead. However, I talked to the highschool principal and he gave me permissionto go ahead with the application. I managedto obtain all the necessary papers without en-couragement from anyone in the school. Atest was given as part of the selection pro-cess, and I found a teacher who was willingto act as proctor. I did well enough to be in-vited as one of the 40 finalists to Washington,DC. Each finalist had a science project for ex-hibition in the Statler Hotel, where we werestaying. My project was on the lives of al-cids, based in part on the trip to the GaspePeninsula and some of the field studies I hadmade during New England winters. The var-ious judges spent considerable time talkingwith us, and the astronomer Harlow Shap-ley, who was the chief judge, charmed mewith his apparent interest in my project. Iwas chosen as one of two co-winners. (At thattime, there was one male and one female win-ner; Rada Demereck and I were co-winners.)The visit to Washington, DC, was a greatexperience, especially because we met Pres-ident Truman, who welcomed us as the fu-ture leaders of America. Moreover, winningthe Westinghouse Talent Search made up forthe discouraging interactions with some ofmy high school teachers. Their attitude con-trasted with that of my fellow classmates, whovoted me “most likely to succeed.”

My final forays into ornithology took placeduring several summers at the end of highschool and after I entered college. In 1947,

I had a summer internship at the MarylandPatuxent Research Refuge of the Fish andWild Life Service, the only National WildlifeRefuge established to conduct research. Pub-licity about the harmful effect of DDT onbird life had prompted studies at the refuge.We collected eggshells as part of a field sur-vey of two several-acre plots, one sprayed withDDT at the normal level and the other with-out DDT. My task was to analyze eggshellsfor their DDT contents and to determine thedifferences between shells from the two plots(their thickness and other features). I also con-ducted a census twice a day of the birds in thetwo areas, determined the number of nestingpairs, and collected any dead birds I found.The summer was an exciting one for me anda fine introduction to field research and labo-ratory work as part of a team. It was a very hotsummer, and all my (older) colleagues drankbeer to relax in the late afternoon, so I joinedthem. I did not like the bitter taste initially butrapidly learned to enjoy beer.

Thanks to a meeting with ProfessorRobert Galambos, who did research on theecholocation of bats in the basement ofMemorial Hall at Harvard, I was invited byhis collaborator, Professor Donald Griffin, tojoin his group in a study of bird orientationthat was to take place in Alaska during thesummer of 1948. (Griffin and Galombos haddemonstrated in 1940 that bats used echolo-cation to orient themselves and find their prey,though Lazzaro Spallanzani had already sug-gested this in 1794, but apparently no one be-lieved him.) Our team was based at the Arc-tic Research Laboratory in the town of PointBarrow, Alaska, which is located at latitude71◦ north, the northern-most point of landon the North American continent. The lab-oratory was run by the Office of Naval Re-search (ONR), primarily to study how hu-mans (presumably soldiers and sailors) canadapt to life in the arctic winter. Its director,Lawrence Irving from Swarthmore College,had a broad view of the laboratory mission andhad invited our group to use the facilities. (Itis worth mentioning that before there were

10 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

organizations to support civilian research,like the National Science Foundation, ONRwas the leading government agency in thisarea.) Our primary interest was in goldenplovers, which nested on the tundra in ar-eas not far from the laboratory. Atlantic andPacific golden plovers nested together, andthe two species separated in the fall to mi-grate over a thousand miles or more south-ward to their respective winter homes. Wetrapped some plovers of both species and at-tached radio transmitters to all of them andmagnets to half of them. We then releasedthe plovers 20 to 50 miles from their nests,both in the Atlantic and Pacific direction, andfollowed them at a distance with a small air-plane. The idea was to ascertain if the birdswith magnets would have a more difficulttime returning to their nesting area. Therewere suggestive results (the birds with mag-nets seemed to get more disoriented thanthose without, though they all found their wayhome), but Griffin felt that what we had foundwas not conclusive proof that the plovers hada magnetic sense, and the work was neverpublished.

In Umiat, an observation camp at a dis-tance from the main laboratory, I organizedmy first experiment with the aid of the otherscientists, who must have been amused by myyouthful enthusiasm. (I was 17 years old andby far the youngest member of the researchteam.) The experiment involved several nest-ing pairs, including robins. Three people par-ticipated in the observation of the nests; eachperson stood a 4-h shift twice a day becausethere were 24 hours of daylight. We foundthat the parents fed the robins over the en-tire 24-h period and, interestingly, that theyoung robins left the nest earlier than didtheir cousins, who nested in Massachusetts. Iwrote a paper (50) describing the results, withthe conclusion that the survival value of theshorter time in the nests, which were highlyexposed to local predators, made up for thedangers of the longer flights required to reachthe summer nesting area in the Arctic. It isnot clear whether my youthful conclusion was

correct, although it did stimulate a number ofpapers, both pro and con, and the paper con-tinues to be cited (106). Also, I noticed that atUmiat, where we were on our own, the normal24-h day stretched to about 30 h; we stayedawake for 22 or so hours and then slept forabout 8 hours.

The following summer Griffin invited meto Cornell University, where he was on thefaculty before joining the Biology Depart-ment at Harvard. In addition to conductingexperiments in the Griffin lab, I enjoyed“hanging out” with college students andschool teachers who were taking summercourses at Cornell. I initially worked on batecholocation and was very impressed by theway the bats were kept in the refrigeratorbetween experiments; they went to sleep,hanging from a rod all in a row. At Griffin’ssuggestion, I then focused on trying to condi-tion pigeons to respond to a magnetic field totest the results of an article that had concludedthat pigeons use the earth’s magnetic field tonavigate. I was doubtful about the paper be-cause I thought the analysis was flawed. Myattempts at conditioning the pigeons provedelusive, but we never published our negative(to me, positive) results. Subsequently, otherexperiments have shown that pigeons, as wellas wild birds, do use the earth’s magnetic fieldas an aid in navigation. This experience taughtme that being skeptical is essential in sci-ence, but that it is also important to be re-ceptive to new ideas—even if you do not likethem.

COLLEGE YEARS

I entered Harvard in the fall of 1947. Therewas never any question about my wanting toattend Harvard and I did not apply to anyother school. In addition to the Westinghousescholarship, I received a National Scholarshipfrom Harvard to cover the cost of living oncampus. Otherwise I would have had to live athome to save money. I would not have mindedthis, since I was not a rebellious teenager ea-ger for independence and distance from my


ANRV275-BB35-01 ARI 10 April 2006 15:34

parents. However, as I soon discovered, muchof the Harvard experience took place outsideof classes at dinner and in evening discussionswith friends.

At first I still intended to go to medicalschool but changed my mind during my fresh-man year. My teenage ornithological studies,fostered by Griscom and Griffin, had alreadyintroduced me to the fascinating world of re-search, where one is trying to discover some-thing new (something that no one has everknown); I began to think about doing re-search in biology. I had concluded that toapproach biology at a fundamental level (tounderstand life), a solid background in chem-istry, physics, and mathematics was imper-ative, and so I enrolled in the Program inChemistry and Physics. This program, uniqueto Harvard at the time, exposed undergradu-ates to courses in both areas at a depth thatthey would not have had from either onealone. It had the additional advantage, frommy point of view, that it was less structuredthan chemistry. For example, it did not requireAnalytical Chemistry, certainly a good courseas taught by Professor James J. Lingane, butone that did not appeal to me. Although Ishopped around for advanced science coursesto meet the rather loose requirements, I alsoenrolled in Freshman Chemistry because itwas taught by Leonard Nash. A relatively newmember of the Harvard faculty, Nash had thedeserved reputation of being a superb teacher.Elementary chemistry in Nash’s lectures wasan exciting subject. A group of us (includingDeWitt Goodman, Gary Felsenfeld, and JohnKaplan—my “crazy” roommate, who becamea law professor at Stanford) had the specialprivilege that Nash spent extra time discussingwith us a wide range of chemical questions, farbeyond those addressed in the course. Theinteractions in our group, though we werehighly competitive at exam times, were alsosupportive. This freshman experience con-firmed my interest in research and the deci-sion not to go to medical school.

Harvard provided me with a highly stimu-lating environment as an undergraduate. One

aspect was the laissez-faire policy, which al-lowed one to take any courses with the in-structor’s permission, even without having theformal prerequisites. The undergraduate deansaid it was up to me to decide and, if a courseturned out to be too much for me, that wouldbe “my” problem. I enrolled in a wide rangeof courses, chosen partly because of the sub-ject matter and partly because of the outstand-ing reputation of the lecturers; these coursesincluded one in Democracy and Governmentand another in Abnormal Psychology. More re-lated to my long-term interests were some ad-vanced biology courses, which I registered inwithout having to suffer through elementarybiology and biochemistry. Two memorablecourses were George Wald’s Molecular Basisof Life and Kenneth Thimann’s class on plantphysiology with its emphasis on the chemistryand physiology of growth hormones (auxins)in plants. Both professors were inspiring lec-turers and imbued me with the excitement ofthe subject. These courses emphasized thatbiological phenomena (life itself ) could beunderstood at a molecular level, which hasbeen a leitmotif of my subsequent research ca-reer. Wald’s course also introduced me to themechanism of vision, which led to my first pa-per on a theoretical approach to a biologicalproblem (42).

Although I remember my undergraduatecareer at Harvard as a formative experiencethat furthered my interest in the world of sci-ence, it was reminiscent of my high schooldays in that my brother had preceded me,had been a stellar student in the Program inChemistry and Physics, and was in the pro-cess of completing a PhD at Harvard withE. Bright Wilson, Jr., and Julian Schwinger.I spent considerable time with Bob and hisfellow graduate students in Wilson’s group. Iwas tolerated, I suppose, as Bob’s little brother,though one day I made my mark when I solveda problem (they were always “challenging”each other) before any of them did. Giventhe importance of this “success” in my life,I am fond of the problem and restate it here:“It is agreed that to divide a pie between two

12 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

people so that both are satisfied, one is allowedto cut the pie in two and the other chooses.The problem is how to extend this concept(dividing or choosing) to three or more peo-ple, so that everyone is satisfied.” There is aspecial solution for three people and a gen-eral solution for any number. After solvingthis problem, I was accepted as part of thegroup by my brother and his friends, and dur-ing the afternoons when I could escape fromthe many labs I had to do, I would often jointhem. Their discussions of science exposed meto new ideas that I would not have come acrossotherwise.

The legendary Elementary Organic coursetaught by Louis Fieser was a standard part ofthe Program in Chemistry and Physics, butI thought it would be a waste of time: Thecourse had the reputation of requiring a verytedious laboratory and endless memorization.An early version of the well-known textbookby Louis Fieser and Mary Fieser was avail-able in lecture-note form and, rather than en-rolling in the course, I tried to learn organicchemistry by reading it on my own, not withcomplete success. After studying the lecturenotes, I enrolled in Paul Bartlett’s AdvancedOrganic because that course taught the phys-ical basis of organic reactions. It was an excel-lent course, though difficult for me becauseone was supposed to know many organic re-actions, which I had to learn as we went along.At one point, Bartlett suggested that we readLinus Pauling’s Nature of the Chemical Bond,which had been published in 1939 based onhis Baker Lectures at Cornell. The Nature ofthe Chemical Bond presented chemistry for thefirst time as an integrated subject that could beunderstood, albeit not quite derived, from itsquantum chemical basis. The many insights inthis book were a critical element in orientingmy subsequent research in chemistry.

At the end of three years at Harvard Ineeded only one more course to completethe requirements for a bachelor degree. Dur-ing the previous year I had done researchwith Ruth Hubbard and her husband, GeorgeWald. (Although Hubbard was scientifically

on par with Wald, she remained a SeniorResearch Associate, a nonprofessorial ap-pointment, until very late in her career whenshe was finally “promoted” to Professor. Thiswas not an uncommon fate for women in sci-ence.) I mostly worked with Hubbard on thechemistry of retinal, the visual chromophore,because she had a deeper knowledge of chem-istry than Wald. When I brought up my needto find a course for graduation, Wald sug-gested that I enroll in the physiology courseat the Marine Biological Laboratory in WoodsHole, Massachusetts. This course was one ofthe few non-Harvard courses that were ac-cepted for an undergraduate degree by theFaculty of Arts and Sciences. The physiol-ogy course was widely known as a stimulatingcourse designed for postdoctoral fellows andjunior faculty. The lectures in the physiologycourse by scientists who were summering atWoods Hole, while doing some research andenjoying boating and swimming, offered stu-dents a state-of-the-art view of biology andbiological chemistry. (The course still exists,although its subject matter has shifted towardcell biology, of greater current interest.) Forme, the only undergraduate in the course, itwas a wonderful experience. I not only learneda great deal of biology but I also met severalpeople, including Jack Strominger and AlexRich, who became lifelong friends.

Woods Hole was an exciting place. Amongthe famous scientists that I met there wereOtto Loewi, who had received a Nobel Prizein Physiology and Medicine (1936) for thediscovery of the chemical basis of the trans-mission of nerve impulses, and Albert SzentGyorgi, who had won a Nobel Prize, also inPhysiology and Medicine (1937) for discov-ering vitamin C and showing that it existedin high concentration in paprika, a staple ofthe Hungarian diet. His little book, Natureof Life, A Study of Muscle (116), helped in-spire my interest in doing research in biology.Like the Nature of the Chemical Bond, its stresson the logic of the subject helped to arousemy interest. These scientists held court inthe afternoons at the Woods Hole beach and


ANRV275-BB35-01 ARI 10 April 2006 15:34

fascinated us young people with their discus-sions of new experiments and scientific gossip.In addition, there was an active student life,since many of the senior researchers broughtalong students from their labs. One of the jus-tifications for having the laboratory in WoodsHole was that there were a wide range of ma-rine animals which were caught for use inexperiments. A prime example is the squid,whose giant axon was the ideal system forstudies of the mechanism of nerve conduc-tion. Because most of the experiments usedonly a small fraction of the animal, each weekor so I collected some of the leftover labora-tory squids and lobsters and prepared a feastfor myself and friends, who provided bread,wine, and salad.

In considering graduate school during mylast year at Harvard, I had decided to go to theWest Coast and had applied to chemistry atthe University of California at Berkeley and tobiology at the California Institute of Technol-ogy (Caltech). Accepted at both, I found it dif-ficult to choose between them. Providentially,I visited my brother, Bob, who was workingwith J. R. Oppenheimer at the Institute of Ad-vanced Studies in Princeton, New Jersey. Bobintroduced me to Oppenheimer, and brieflyto Einstein. When Oppenheimer asked mewhat I was doing, I told him of my dilemma inchoosing between U.C. Berkeley and Caltechfor graduate school in chemistry or biology.He had held simultaneous appointments atboth institutions and strongly recommendedCaltech, describing it as “a shining light ina sea of darkness.” His comment influencedme to choose Caltech, and I discovered thatOppenheimer’s characterization of the localenvironment was all too true. Pasadena itselfheld little attraction for a student at that time.However, camping trips in the nearby desertand mountains and the vicinity of Hollywoodmade up for what Pasadena lacked.

I had become very interested in films andorganized a classic film series at Caltech dur-ing my time there. This enabled me to previewmany films that I had always wanted to see.The series showed mainly silent films accom-

panied by live piano music played by fellowstudents. (One of them was Walter Hamilton,a crystallographer, who died very young.) Insearching for films, I gained access to sev-eral production studios, where I would askthe librarians to lend me films for our non-profit Caltech series. The high point was myvisit to the Chaplin studio. The reception-ist was not particularly forthcoming, but thenCharlie Chaplin himself walked in and askedwhat I wanted. He seemed intrigued by theidea that a science student was interested infilms, and I asked him about the possibil-ity of showing Monsieur Verdoux. He saidit had been withdrawn (for political reasons),but told the librarian to let me have someof his early short films which at the timewere not available to the public. I had al-ways been a Chaplin fan and this meeting wasone of the very special events of my graduatecareer.

At Caltech, I first joined the group of MaxDelbruck in biology. He had started out asa physicist but, following the advice of NielsBohr, had switched to biology. With SalvadorLuria and others, he had been instrumentalin transforming phage genetics into a quanti-tative discipline. His research fascinated me,and I thought that working with such a personwould be a perfect entree for me to do gradu-ate work in biology. There were many brightand lively people working with Delbruck. Iparticularly remember Seymour Benzer, likeDelbruck, a former physicist. We had discus-sions of phage genetics, biology, and a va-riety of subjects of mutual interest. Amongother things, it was Seymour who introducedme to horsemeat, which became a staple ofour household in Altadena. The law in LosAngeles was such that at the supermarkethorsemeat was sold only as meat for dogs.Not deterred by that, horse filet, which costa fraction of the corresponding cut of beef,turned up in my cooking as horse stroganoff,horse Cordon Bleu, and so on. Each of themembers of our household, which includedSidney Bernhard, Gary Felsenfeld, and Wal-ter Hamilton, had specific tasks to do. Mine,

14 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

with Sidney, was cooking, and the horsemeat(as well as the local seafood) was regularlyserved.

After I had been in the Delbruck groupfor a couple of months, Delbruck proposedthat I present a seminar on a possible areaof research. I intended to discuss my ideasfor a theory of vision (how the excitation ofretinal by light could lead to a nerve im-pulse), which I had started to develop whiledoing undergraduate research with Hubbardand Wald. Among those who came to my talkwas Richard Feynman; I had invited him to theseminar because I was taking his quantum me-chanics course and knew he was interested inbiology, as well as everything else. I began theseminar confidently by describing what wasknown about vision but was interrupted after afew minutes by Delbruck’s comment from theback of the room, “I do not understand this.”The implication of his remark, of course, wasthat I was not being clear, and this left me withno choice but to go over the material again.As this pattern repeated itself (Delbruck say-ing “I do not understand” and my trying toexplain), after 30 minutes I had not even fin-ished the 10-minute introduction and was get-ting nervous. When he intervened yet again,Feynman turned to him and whispered loudenough so that everyone could hear, “I canunderstand, Max; it is perfectly clear to me.”With that, Delbruck got red in the face andrushed out of the room, bringing the semi-nar to an abrupt end. Later that afternoon,Delbruck called me into his office to tell methat I had given the worst seminar he had everheard. I was devastated by this and agreed thatI could not continue to work with him. It wasonly years later that I learned from reading abook dedicated to him that what I had gonethrough was a standard rite of passage for hisstudents—everyone gave the “worst seminarhe had ever heard.”

After the devastating exchange withDelbruck, I spoke with George Beadle, thechairman of the Biology Department. He sug-gested that I find someone else in the de-partment with whom to do graduate research.

However, I felt that I wanted to go “home” tochemistry and asked him to help me make thetransfer. Once in the Chemistry Department,I joined the group of John Kirkwood, who wasdoing research on charge fluctuations in pro-teins, as well as on his primary concern withthe fundamental aspects of statistical mechan-ics and its applications. I undertook work onproteins and research started out well. It wascomplemented by a project involving IrwinOppenheim and Alex Rich. Kirkwood’s coursein Advanced Thermodynamics was famous forits rigor, and the three of us, with Kirkwood’sencouragement, worked together to preparea set of lecture notes for the course. Each ofus was responsible for writing up some of thelectures and the other two read them over.This was very useful for our learning thermo-dynamics and the set of notes was circulatedwidely. Some years later, Irwin Oppenheimprepared an improved version of the notesand it was published as a text entitled ChemicalThermodynamics (75).

In the spring of 1951, as I was getting im-mersed in my research project, Kirkwood re-ceived an offer from Yale. Linus Pauling, whowas no longer taking graduate students, askedeach student who was working with Kirkwoodwhether he would like to stay at Caltech andwork with him. I was the only one to acceptand, in retrospect, I think it was a very goodchoice. Initially, I was rather overwhelmed byPauling. Each day upon arriving at the lab, Ifound a handwritten note on a yellow piece ofpaper in my mailbox which always began withsomething like “It would be interesting to lookat. . . .” As a new student I took this as an or-der and tried to read all about the problem andwork on it, only to receive another note thenext day beginning in the same way. When Iraised this concern with Alex Rich and otherpostdocs, they laughed, pointing out that ev-eryone received such notes and that the bestthing to do was to file them or throw themaway. Pauling had so many ideas that he couldnot work on all of them. He would commu-nicate them to one or another of his students,but he did not expect a response. After I got


ANRV275-BB35-01 ARI 10 April 2006 15:34

over that, my relation with Pauling developedinto a constructive collaboration.

Given Pauling’s interest in hydrogenbonding in peptides and proteins, he proposedthat I study the different contributions to hy-drogen bonding interactions for a biologicallyrelevant system, but I felt this would be toodifficult to do in a rigorous way. Because quan-tum mechanical calculations still had to bedone with calculating machines and tables ofintegrals (something difficult to imagine wheneven log tables have followed dinosaurs intooblivion), we had to find a simple enoughsystem to be treated by quantum mechani-cal theory. I chose the bifluoride ion (FHF−)because the hydrogen bond is the strongestknown, the system is symmetric, and onlytwo heavy atoms are involved. (Today, such“strong” hydrogen bonds have become popu-lar in analyses of enzyme catalysis, althoughthere is no convincing evidence for theirrole.)

I sometimes felt intimidated when I wentin to talk with Pauling, but it was wonderfulto work with him and be exposed to (al-though not necessarily understand) his intu-itive approach to chemical problems. One dayI asked Pauling about the structure of a cer-tain hydrogen bonded system (i.e., whetherthe hydrogen bond would be symmetric, asin FHF−). He paused, thought for a while,and gave a prediction for the structure. WhenI asked him why, he thought again and of-fered an explanation. I left his office and soonrealized that his explanation made no sense.So I went in to ask Pauling again, thinkingthat perhaps he would come up with a dif-ferent conclusion. Instead, Pauling said thathe believed that the predicted structure wascorrect, but he proposed an entirely differentexplanation. After going over his analysis, Iwas again dissatisfied with his rationale andcaught up with him as he was leaving his of-fice. He said that he still believed the con-clusion was valid and produced yet anotherrationale. This one made sense to me, so Idid some crude calculations which indicatedthat Pauling was indeed correct. What amazed

me then, and still does today, is that Paul-ing came to the correct conclusion, apparentlybased on intuition, without having workedthrough the analysis. He “knew” the right an-swer, even if it took more thought to figure outwhy.

The research was very rewarding, all themore so because of the intellectual and so-cial atmosphere of the Chemistry Departmentat Caltech. The professors—like Pauling,Verner Schomaker, and Norman Davidson—treated the graduate students and postdoc-toral fellows as equals. We participated inmany joint activities that included trips intothe desert, as well as frequent parties held atour Altadena house, where Feynman wouldoccasionally come and play the drums. At onesuch party, Pauling disappeared for a whileand I discovered him out in the backyard onhis knees collecting snails, which had infestedour yard, for his wife Ava Helen to cook fordinner. (It was only later in France, when I col-lected my own snails, that I learned how com-plicated it was to prepare them—the snails hadto fast for a week—so that now, looking back,I am not sure what the Paulings did with theirsnails.)

Pauling’s presence attracted many post-doctoral fellows to Caltech. When I wasthere the group included Alex Rich, JackDunitz, Massimo Simonetta, Leslie Orgel,Edgar Heilbrunner, and Paul Schatz. Inter-acting with them (as a graduate student, Iwas the “baby” of the group) was a wonder-ful part of my Caltech education and manyof them became my friends. Sadly, MassimoSimonetta, who remained a dear friend andcolleague after his return to Italy and whom Ivisited regularly in Milan, as well as on ski tripsto Courmayeur in the winter and Portofinoin the summer, died suddenly from virulentleukemia in 1986.

My parents had given me their old car as agraduation present, and several times duringmy Caltech career I drove across the countryto our home in Newton, Massachusetts, forpart of the summer. Each time I took a differ-ent route, once through Canada with visits to

16 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

the Banff and Jasper National Parks, and an-other time through the Deep South. On onesuch trip while driving through Texas on avery hot summer day, my friends and I de-cided to take a swim and cool off. We passedone swimming area, but it was full of people.Because we were unshaven and dirty from thelong drive, we looked along the river for a qui-eter place to swim. About a mile downstreamwe came upon another swimming area withbroken-down steps leading to the water. Asit was deserted, we decided it was the perfectplace. After we had been in the water for about10 minutes, a couple of pickup trucks drove upand several men jumped out with guns at theready. What turned out to be the local law en-forcement officers ordered us out of the waterand demanded to know what we thought wewere doing. . . “white folks swimming in anarea reserved for niggers.” They had noticedthe Massachusetts license plate on the car andhad concluded we were Northern “trouble-makers.” After some effort we succeeded inexplaining that, given our scruffy state, we hadjust not wanted to bother other (white) peopleand the officers let us go, with the admoni-tion that we had better drive straight throughTexas without stopping anywhere, which wedid.

My initial attempt at a purely ab initioapproach to the bifluoride ion failed and Isoon realized that it was necessary to in-troduce experimental information concerningthe atomic states to obtain a meaningful esti-mate of the relative contribution of covalentand ionic structures. I developed a methodfor doing this and completed my work onlyto discover that William Moffitt (who wasto be my predecessor at Harvard) had justpublished a similar approach called themethod of Atoms in Molecules. He had pre-sented the method in a more general andelegant formulation (91) compared with mytreatment, which focused specifically on thebifluoride ion. Although there were signifi-cant differences in the details of the method-ology, I felt so discouraged by the similaritiesthat I never published “My Great Idea,” as

Verner Schomaker, one of the members of mythesis committee, called it. Not too surpris-ingly, I was having a difficult time writing mythesis. (I retained my interest in this type ofapproach, and some years later Gabriel Balint-Kurti joined my group as a graduate student,and we proposed an improved version of thetheory, which we called the OrthogonalizedMoffitt method, and applied it to the poten-tial energy surfaces for simple reactions (4).)Such calculations are mostly of historical in-terest today, when fast computers and ab initioprograms are widely used without empiricalcorrections, except for complex systems likebiomolecules.

POSTDOCTORAL SOJOURN INOXFORD AND EUROPE

One day in October 1953, Pauling came intothe office I shared with several postdocs andannounced that he was leaving in three weeksfor a six-month trip and that “it would be nice”if I finished my thesis and had my exam beforehe left. This was eminently reasonable, sinceI had finished the calculations some monthsbefore and I had received a National ScienceFoundation (NSF) postdoctoral fellowship togo to England that fall. Pauling’s “request”provided just the push I needed, even thoughthe introduction was all I had written thus far.With so much to get done, I literally wrotenight and day, with my friends typing and cor-recting what I wrote. In this way, the thesis wasfinished within three weeks, and I had my fi-nal PhD exam and celebratory party. After abrief visit with my parents in Newton, I left forEngland and arrived shortly before Christmas1953.

In my NSF postdoctoral application I hadproposed to work with John Lennard-Jonesin Cambridge, England. He, however, hadleft Cambridge to become Principal of KeeleUniversity in 1953, and so I had to alter myplans. Instead I joined Charles Coulson atOxford University, where he had an activegroup in theoretical chemistry at the Math-ematical Institute. One member was Simon


ANRV275-BB35-01 ARI 10 April 2006 15:34

Altmann, who greatly improved my limitedknowledge of group theory and who, with hiswife, Bochia, “adopted” me while I was in Ox-ford. In addition, there were visitors such asDon Hornig and Bill Lipscomb, who were onsabbaticals.

I was 23 years old when I arrived inEngland. Having worked continuously all theway through graduate school, I was eager tohave the sojourn in Europe provide expe-riences beyond science. The NSF postdoc-toral fellowship provided a generous (at thetime) salary of $3000 per year, which was suf-ficient to do considerable traveling. I tookthe NSF guidelines about following the cus-toms of the institution quite literally, per-haps more so than was intended, and traveledthroughout Europe outside of the three six-week terms when I was in residence in Oxford.In fact, upon my arrival in Oxford shortly be-fore Christmas, I went in to see Coulson, gothim to sign my NSF form, and immediatelyleft on the first of many trips to Paris. Al-though there were few scientific interactionsduring these extended trips, I learned muchabout the peoples and their cultures, art, ar-chitecture, and cuisine, all of which continueto play an important role in my life. One tripinvolved driving a Volkswagen across Europethrough Yugoslavia to Athens; another an ex-tended visit to France, Switzerland, and Italy.On the latter, I first saw the Lake Annecy andthe Haute Savoie region. I concluded that Iwanted to own a chalet in one of the uppervalleys for summer and winter vacations, butI could not afford it then and for many yearsto come.

I became interested in photography at thattime. Upon completing my PhD, my familyhad given me a Leica IIIC, a superb cam-era, which my Uncle Alex had brought tothe United States from Vienna. Throughoutmy travels, I took photographs, particularly ofpeople, using a trick I had developed. The Le-ica had a long focus lens with a reflex viewer,which enabled me to face away from my sub-ject; I could take photographs of crowds andindividuals without their being aware of it. Al-

though the resulting Kodachrome slides arenow more than 50 years old, they remainin excellent condition. My wife, Marci, hadthe idea that the slides, which had hardlybeen looked at over the years, should beprinted and enlarged. This endeavor has, ina sense, come full circle. During the academicyear 1999–2000 I was Eastman Professor atOxford. While there, we were introduced to amarvelous photographic craftsman, Paul Sims(Colourbox Techunique). He made beauti-ful prints from the collection of slides, and aselection was exhibited at my 75th birthdaycelebration at NIH in Bethesda, Maryland(101).

During the two years in Oxford as a post-doctoral fellow, I spent more time thinkingabout chemical problems than actually solv-ing them. My aim was to find areas wheretheory could make a contribution of generalutility in chemistry. I did not want to do re-search whose results were of interest just totheoretical chemists. Reading the literature,listening to lectures, and talking to scientistslike Don Hornig and the Oxford physicistH.M.C. Pryce, I realized that magnetic res-onance was a vital new area. Chemical appli-cations of magnetic resonance were in theirinfancy and it seemed to me that nuclear mag-netic resonance (NMR), in particular, was afield where theory could make a contribution.Chemical shifts, for example, could provide ameans of testing theoretical calculations, but,of even greater import, quantum mechani-cal theory could aid in interpreting the avail-able experimental results and propose newmeasurements.

My first paper in chemistry on thequadrupole moment of the hydrogenmolecule, obtained from different approx-imate wavefunctions, was in this vein (51).Although this was a short paper, I spent anenormous amount of time rewriting andpolishing it before I finally was ready tosubmit it. When students seem to face similarproblems in finishing their first paper, I oftentell them about what I went through and thatpublishing (or being ready to publish) one’s

18 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

work becomes easier with each succeedingpaper.

FIVE YEARS AT THEUNIVERSITY OF ILLINOIS: NMRAND COUPLING CONSTANTS

As my postdoctoral fellowship in Oxford(1953–1955) neared its end, I was looking fora position to begin my academic career in theUnited States. With my growing interest inmagnetic resonance, I focused on finding aninstitution that had active experimental pro-grams in the area. One of the best schools fromthis point of view was the University of Illi-nois, where Charles Slichter in Physics andHerbert Gutowsky in Chemistry were doingpioneering work in applying NMR to chem-ical problems. The University of Illinois hada number of openings in Chemistry at thattime because the department was undergo-ing a radical renovation; several professors,including the chairman Roger Adams, had re-tired. Pauling recommended me to the Uni-versity of Illinois and the department offeredme a job without an interview and withoutwaiting for a recommendation from Coulson.The latter was fortunate, because Coulsonhad written that, although he had no doubtabout my intellectual abilities, I had done verylittle work on problems he had suggested. Iaccepted the offer from Illinois without visit-ing the department, something unimaginabletoday with the extended courtships that havebecome an inherent part of the academic hir-ing process. The University of Illinois offeredme an Instructorship at a salary of $5000 peryear; the department offered nothing like thepresent-day start-up funds, and I did not thinkof asking for research support.

Although the University of Illinois was avery good institution with excellent chemistryand physics departments, it was located in asmall town in the flat rural Midwest, whereI could not imagine living for more than fiveyears. Having had such a good time as a post-doctoral fellow traveling in Europe, I wasready to get to work, and Urbana-Champaign

seemed like a place where I could concentrateon science with few distractions. The pres-ence of four new instructors—Rolf Herber,Aron Kupperman, Robert Ruben, and me—plus other young scientists on the faculty, suchas Doug Applequist, Lynn Belford, and E.J.Corey, led to a very interactive and congenialatmosphere.

I focused a major part of my researchon theoretical methods for relating nuclearand electron spin magnetic resonance param-eters to the electronic structure of molecules.The first major problem I examined wasconcerned with proton-proton coupling con-stants, which were known to be dominatedby the Fermi contact interaction. What madecoupling constants of particular interest wasthat for protons, which were not bonded, theexistence of a nonzero value indicated thatthere was an interaction beyond that expectedfrom localized bonds. In the valence bondframework, which I used in part because ofmy training with Pauling, nonzero couplingconstants provide a direct measure of the de-viation from the perfect-pairing approxima-tion. To translate this qualitative idea into aquantitative model, I chose to treat the vici-nal coupling constant in a molecule like ethylalcohol, one of the first molecules to haveits NMR spectrum analyzed experimentally.Specifically, I chose to study the HCC‘H’fragment as a function of the HCC‘H’ dihe-dral angle, a relatively simple system consist-ing of six electrons (with neglect of the innershells). I believed that it could be describedwith sufficient accuracy for the problem athand by including only five covalent valence-bond structures. To calculate the contribu-tions of the various structures, I introducedsemiempirical values of the required molecu-lar integrals. Although the HCC‘H’ fragmentis relatively simple, the calculations for a se-ries of dihedral angles were time consumingand it seemed worthwhile to develop a com-puter program. This was not as obvious in1958 as it is now. Fortunately, the ILLIAC, a“large” digital computer at that time, had re-cently been built at the University of Illinois.


ANRV275-BB35-01 ARI 10 April 2006 15:34

If I remember correctly, it had 1000 wordsof memory, which was enough to store myprogram. The actual program was written bypunching holes in a paper tape. If you made amistake, you filled in the incorrect holes withnail polish so that you could continue the pro-gram, the output appearing on spools of paper.Probably the most valuable aspect of havinga program for this type of simple calculation,which could have been done on a desk calcu-lator, was that once the program was knownto be correct, a large number of calculationscould be performed without having to worryabout arithmetic mistakes.

Just as I finished the analysis of the vic-inal coupling constants (52), I heard a lec-ture by R.V. Lemieux on the conformationsof acetylated sugars. I do not remember whyI went to the talk, because it was an organicchemistry lecture, and the chemistry depart-ment at Illinois was rigidly separated into divi-sions, which had a semiautonomous existence.Lemieux reported measurements of vicinalcoupling constants and noted that there ap-peared to be a dihedral angle dependence, al-though the details of the behavior were notclear. The results were exciting to me becausethe experiments confirmed the theory, at leastqualitatively, before it was even published.

E.J. Corey, who was an assistant profes-sor at Illinois and later became a colleague atHarvard, was one of the people with whomI had dinner on a fairly regular basis at theTea Garden, a passable Chinese restaurant inUrbana-Champaign. We often discussed re-cent work of mutual interest and one day Idescribed the studies that I had made of vic-inal coupling constants. Corey immediatelyrecognized the possibility of using the re-sults for structure determination and pub-lished what is probably the first applicationof my results in organic chemistry (7). Notlong after, the theory appeared in a compre-hensive review of the use of NMR in organicstructure determinations (19), and someoneintroduced the name Karplus equation for therelationship I had developed. This proved amixed blessing. Many people attempted to ap-

ply the equation to determine dihedral anglesof organic compounds. They found some de-viations of the measured coupling constantsfrom the predicted values for known struc-tures and published their results, commentingon the inaccuracy of the theory.

As happens too often with the applicationof theoretical results in chemistry, most peo-ple who used the so-called Karplus equation didnot read the original paper (52) and thus donot know the limitations of the theory. Theyassumed that because the equation had beenused to estimate vicinal dihedral angles, thetheory said that the coupling constant de-pends only on the dihedral angle. By 1963,having realized organic chemists tend to writeand read Communications to the Journal of theAmerican Chemical Society, I published such aCommunication (56). In it, I described variousfactors, other than the dihedral angle, that areexpected to affect the value of the vicinal cou-pling constant; they include the electronega-tivity of substituents, the valence angles of theprotons (HCC’ and CC’H), and bond lengths.The main point of the paper was not to pro-vide a more accurate equation but rather tomake clear that caution had to be used in ap-plying the equation to structural problems.My closing sentence, which has often beenquoted, was the following: “Certainly with ourpresent knowledge, the person who attemptsto estimate dihedral angles to an accuracy ofone or two degrees does so at his own peril.”

In spite of my concerns about the limita-tions of the model, the use of the equation hascontinued, and the original paper (52) is oneof the Current Contents “most-cited papers inchemistry”; correspondingly, the 1963 paperwas recently listed as one of the most-citedpapers in the Journal of the American Chemi-cal Society (20a). In addition, there have beenmany empirical “extensions” of the equation;perhaps the most complex published form(46) uses a 12-term expression. Equationshave been developed for vicinal coupling con-stants involving a variety of nuclei (73) (e.g.,13C-CC’-H, 15N-CC’-H), and they have beenapplied in areas ranging from inorganic to

20 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

organic to biochemistry. An important morerecent application is the use of these relation-ships as part of the data employed in structuredetermination of proteins by NMR (76, 89).The vicinal coupling constant model, whichwas developed primarily to understand devi-ations from perfect pairing, has been muchmore useful than I would have guessed. “Inmany ways my feeling about the uses and re-finements of the Karplus equation is that of aproud father. I am very pleased to see all thenice things that the equation can do, but it isclear that it has grown up and now is living itsown life” (59).

I continued to work on problems in NMRand ESR (electron spin resonance) becausenew areas of chemistry were being studied bythese spectroscopic methods and it seemedworthwhile to try to provide insights fromtheoretical analyses of some of these appli-cations. Examples are a study of the hyper-fine interactions in the ESR spectrum of themethyl radical (53) and the contributions ofπ-electron delocalization to the NMR cou-pling constants in conjugated molecules (54).My general approach to the magnetic proper-ties of molecules was summarized in an arti-cle entitled “Weak Interactions in MolecularQuantum Mechanics” (55). The choice of ti-tle was apt because the energies involved incoupling constants and hyperfine interactionsare indeed weak relative to the electron voltsof bond energies, excitation energies, and ion-ization potentials that are the bread and but-ter of quantum chemistry. However, the titlealso had a facetious aspect in that my brotherhad been working on what the physicists call“weak interactions” (72).

At Illinois, my officemate was AronKuppermann. Our instructorship at Illinoiswas the first academic position for both ofus, and we discussed science, as well as pol-itics and culture, for hours on end. We soonbecame fast friends. He and his wife, Roza,lived in an apartment next to mine and ofteninvited me for dinner. Our friendship has con-tinued for more than 50 years, even though Ileft Illinois to go to Columbia University and

Aron moved to Caltech. We see each otheronly once in several years, but having Aronand Roza as friends provides a special conti-nuity in my life.

Aron and I decided that, although we wereon the faculty, we wanted to continue to learnand would teach each other. I taught Aronabout molecular electronic structure theory[we published two joint papers on molecu-lar integrals (64, 77)] and Aron taught meabout chemical kinetics, his primary area ofresearch. Aron is officially an experimental-ist, but he is also an excellent theoretician, aswas demonstrated by his landmark quantummechanical study of the H + H2 exchangereaction with George Schatz. This work wassome years in the future (it was published in1975) (78), but in the late 1950s we both feltthat it was time to go beyond descriptions ofreactions in terms of the Arrhenius formula-tion based on the activation energy and pre-exponential factor. My research in this areahad to wait until I moved to Columbia Uni-versity, where I would have access to the re-quired computer facilities.

MOVE TO COLUMBIA ANDFOCUS ON REACTIONKINETICS

During the summer of 1960 I participated inan NSF program at Tufts University with thepurpose of exposing high school and small col-lege science teachers to faculty actively en-gaged in research. Our task was to presentsome modern chemical concepts in a way thatwould help the teachers in the classroom. BenDailey, one of the organizers of the program,asked me one day as we were standing nextto each other in the washroom whether Iwould consider joining the chemistry facultyat Columbia University, where he was a pro-fessor. Because I had already been at Illinoisfor four of the five years I had planned tostay there, I responded positively. I heard fromColumbia shortly thereafter and received anoffer to join the IBM Watson Scientific


ANRV275-BB35-01 ARI 10 April 2006 15:34

Laboratory with an adjunct associate profes-sorship at Columbia.

The Watson Scientific Laboratory wasan unusual institution to be financed by acompany like IBM. Although the laboratoryplayed a role in the development of IBM com-puters, many of the scientists there were do-ing fundamental research. The Lab had beenfounded in 1945 near the end of World WarII to provide computing facilities needed bythe Allies. Its director, Wallace Eckart, is per-haps best known for his highly accurate per-turbation calculation of the three-body prob-lem posed by the motion of the earth aroundthe sun in the presence of the moon; the H+ H2 reaction, which I studied while at theLab, is also a three-body problem in the Born-Oppenheimer approximation. When Eckartdescribed the position at the Watson Lab tome, he made clear that staff members werejudged by their peers for what they did intheir research and not for their contribution toIBM. The presence of outstanding scientistson the staff, such as Erwin Hahn, SeymourKoenig, Alfred Redfield, and L.H. Thomas(of Thomas-Fermi fame), supported this de-scription and made the place very attractive.Moreover, the Watson Lab had a special ad-vantage for me in that it had an IBM 650, anearly digital computer, which was much moreuseful than the ILLIAC because of its greaterspeed, larger memory, and simpler (card) in-put. (No more nail polish!) I was to have ac-cess to considerable amounts of time on theIBM 650 and to receive support for post-docs, as well as other advantages over a regu-lar Columbia faculty appointment. This was aseductive offer, but I hesitated about accept-ing a position that, in any way, depended on acompany, even a large and stable one like IBM.This was based, in part, on my political out-look, but even more so on the fact that indus-try has as its primary objective making a profit,and all the rest is secondary. By contrast, myprimary focus was on research and teaching,which are the essential aspects of a university,but not of industry. Consequently, I repliedto Columbia and the Watson Lab that the of-

fer was very appealing, but that I would con-sider it only if it included a tenured positionin the chemistry department, even though Iagreed initially to be at the Watson Lab aswell. Columbia acceded to my request, andafter some further negotiation I accepted theposition for the fall of 1960.

The environment at the Watson Lab wasindeed fruitful, both in terms of discussionswith other staff members and of the avail-able facilities. I was able to do research therethat would have been much more difficult atColumbia. However, not unexpectedly, the at-mosphere gradually changed over the years,with increasing pressure from IBM to dosomething useful (i.e., profitable) for the com-pany, such as visiting people at the much largerand more applied IBM laboratory in York-town Heights, essentially doing internal con-sulting. I decided in 1963 that the time hadcome to leave the Watson Lab, and I movedto the full-time professorial position that waswaiting for me in Chemistry at Columbia.(IBM closed the Watson Lab in 1970.) Giventhat experience, I always warn my studentsand postdocs about accepting jobs in industry.They may well have an exciting environmentwhen they first join the staff, receive a signif-icantly higher salary than they could at a uni-versity, and not have to worry about obtaininggrants to support their research. What I urgethem to remember is that a new managementteam can take over at any time, particularly ifthe company is not doing well, and decide tocut down on the research budget. (Researchin the short run only spends money, even ifit can finally produce a profit.) This attitudehas led to layoffs of individual scientists or theclosing of beautifully equipped research labo-ratories that were built only a few years before.It is of primary importance that your objec-tives (in my case, teaching and research) bethe same as those of the institution where youwork. This requirement is ideally satisfied ina good university but cannot be guaranteed inindustry.

I continued research in the area of mag-netic resonance after moving to New York.

22 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

One reward of being at Columbia was thestimulation provided by interactions withnew colleagues, such as George Fraenkel,Ben Dailey, Rich Bersohn, and Ron Bres-low. Frequent discussions with them helped tobroaden my view of chemistry. In particular,my interest in ESR was rekindled by GeorgeFraenkel and we published several papers to-gether (62, 66, 80), including a pioneeringcalculation of 13C hyperfine splittings (62).Although the techniques we used were rathercrude, the results provide insights concern-ing the electronic structure of the moleculesconsidered and aided in understanding themeasurements. Many of the weak interac-tions, which could be used to provide in-formation about the electronic structure ofmolecules, were a real challenge to estimatein the mid-1960s. Now they can be calcu-lated essentially by pushing a button with pro-grams like the widely used Gaussian package.The high-level ab initio treatments that areused routinely today have the drawback that,even though the results can be accurate, theinsights obtained by the earlier, simplified ap-proaches are often lost in the complexity of thecalculation.

My interest in chemical reaction dynamicshad deepened at Illinois through many dis-cussions with Aron Kuppermann, as alreadymentioned, but I began to do research in thearea only after moving to Columbia. Therewere several reasons for this. There is no pointin undertaking a problem if the methodologyand means for solving it are not available: It isimportant to feel that a problem is ripe for so-lution. (This has been a guiding rule for muchof my research—there are many exciting andimportant problems, but only when one feelsthat they are ready to be solved should one in-vest the time to work on them. This rule hasturned out to be even more important in theapplication of theory to biology, as we shallsee later.) Given the availability of the IBM650 at the Watson Lab, the very simple re-action, H + H2 → H2 + H, which involvesan exchange of hydrogen atom with a hydro-

gen molecule, could now be studied by theoryat a relatively fundamental level. Moreover,early measurements made by Farkas & Farkasin 1935 (26) of the rate of reaction over awide temperature range provided importantdata for comparison with calculations. A sec-ond reason for focusing on chemical kineticswas that crossed molecular beam studies werebeginning to provide much more detailed in-formation about these reactions than had beenavailable from gas phase or solution measure-ments. The pioneering experiments of Taylor& Datz opened up this new field in 1955 (117),although it was not until 2000 that Datz re-ceived the prestigious Enrico Fermi Awardin recognition of this work. Many groups ex-tended their original crossed molecular beamexperiments and showed that it was possibleto study individual collisions and determinewhether they were reactive. Thus, calculatedreaction cross sections, rather than overall rateconstants, could be compared directly with ex-perimental data.

To do a theoretical treatment of this, or anyother reaction (including the protein foldingreaction), a knowledge of the potential energyof the system as a function of the atomic co-ordinates is required; it is necessary to knowthe potential energy surface or energy land-scape, as it is now called. Isaiah Shavitt, whowas working with me as a postdoctoral fel-low at the Watson Lab on quantum mechan-ical calculations, had developed new methodsfor evaluating multicenter two-electron inte-grals (109), and he used the H + H2 poten-tial surface as his first application (110). Eventhough this reaction involved only three elec-trons and three nuclei, the theoretical surfacewas expected to be useful only for determin-ing the general features, and a more accuratesurface was needed for calculating reactionattributes for comparison with experiment.(Five years later Liu (86) was able to calculatean accurate surface for the H + H2 exchangereaction.) Thus, it was necessary to resortto so-called semiempirical surfaces, whoseform was given by a quantum mechanical


ANRV275-BB35-01 ARI 10 April 2006 15:34

model with parameters determined fromexperiment.

Already in 1936, J. Hirshfelder and B.Topley, two students of H. Eyring, had at-tempted a trajectory calculation of the H +H2 reaction with the three atoms restricted tomove on a line, for simplicity (40). In a tra-jectory calculation one determines the forceson the atoms, here the three hydrogen atoms,from the potential energy surface and inte-grates Newton’s equation, F = ma, to ob-tain the positions of the atoms as a function oftime. Hirschfelder and Topley used a three-body potential for the reaction based on theHeitler-London method. They calculated afew steps along the trajectory but were notable to finish the calculation, so we do notknow (and never will since we do not havethe initial velocities) whether the trajectorywas reactive. The potential had a well in theregion where all three atoms were close toeach other (“Lake Eyring” as it was called),which was expected to give a three-body com-plex under the collision conditions appro-priate for the reaction. Ab initio quantummechanical calculations, such as that ofShavitt, indicated that this was incorrect, i.e.,there was a simple activation barrier. Thus, toobtain a meaningful description of the H +H2 reaction, it was necessary to introduce amore realistic potential function. Moreover,what was needed was the reaction surface infull three-dimensional space, rather than re-stricting the hydrogen atoms to positions on aline.

Richard Porter, a graduate student withF.T. Wall, had used a surface without LakeEyring to improve the collinear collision cal-culations (119). Much impressed by Porter, Iinvited him to join my group at Columbia as apostdoctoral fellow. At Columbia, we rapidlydeveloped a semiempirical extension of theoriginal Heitler-London surface for the H+ H2 reaction, based on the method of di-atomics in molecules, and calibrated the sur-face with ab initio quantum calculations andexperimental data for the reaction (100). Thissurface, which is known as the Porter-Karplussurface, has an accuracy and simplicity thatled to its continued use in many reaction ratecalculations by a variety of methods over theyears.

Once the surface was developed, we wereready to undertake the first full three-dimensional trajectory calculation for theexchange reaction. Dick Porter was a funda-mental contributor to the H + H2 research,which also involved R.D. Sharma, anotherpostdoctoral fellow. With the availability oflarge amounts (by the standards of the day) ofcomputer time on the IBM 650 at the WatsonLab, we were able to calculate enough trajec-tories to obtain statistically meaningful results(69). Determination of the reaction cross sec-tion required the calculation of a series of tra-jectories with an appropriately chosen rangeof initial conditions, such as relative velocitiesand impact parameters. The trajectories startwith the reactants far apart (so far that theinteraction between the hydrogen atom and

−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→Figure 2The exchange reaction between a hydrogen atom and a hydrogen molecule. (a) Schematic representationof the reaction with definitions for the distances RAB, RAC, and RBC. (b) Contour plot of the potentialenergy surface for a linear collision as a function of the distances RAB and RBC with RAC = RAB + RBC;the minimum-energy path is shown in red. (c) Same as panel (b), but in a three-dimensionalrepresentation. (d ) Energy along the reaction coordinate corresponding to the minimum-energy path inpanels (b) and (c); the transition state is indicated in yellow. (e) A typical trajectory for the reactive,three-dimensional collision; the three distances RAB, RAC, and RBC are represented as a function of time.( f ) A typical trajectory for the nonreactive collision; as shown in panel (e). In both panels (e) and ( f ) theinteractions between the three atoms are limited to a very short time period (yellow background); thismirrors the narrow potential energy barrier in panel (d ). The figure has been reproduced, withpermission, from Reference 24.

24 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

hydrogen molecule are negligible), let themcollide in the presence of the interaction po-tential, and then follow the atoms until theproducts are again far apart. By looking atwhich atoms are close together, one can de-

termine whether a reaction has taken place(Figure 2). As can be seen from the figure,the reaction (i.e., the time during which thethree atoms are interacting) takes only a fewfemtoseconds. This illustrates a fundamental


ANRV275-BB35-01 ARI 10 April 2006 15:34

point, namely that many simple reactions havea small rate constant, not because of the el-ementary reaction rate, which is fast whenit occurs, but because of the large activationenergy, which makes most thermal collisionsnonreactive. Within the approximation thatclassical mechanics is accurate for describingthe atomic motions involved in the H + H2

reaction and that the semiempirical Porter-Karplus surface is valid, a set of trajectoriesmakes it possible to determine any and all re-action attributes, e.g., the reaction cross sec-tion as a function of the collision energy. Theultimate level of detail that can be achievedis an inherent attribute of this type of ap-proach, which I was to exploit 15 years laterin studies of the dynamics of macromolecules.Although there are significant quantumcorrections for H + H2, the results for thereaction cross section as a function of inter-nal energy and the rate constant as a functionof temperature provided insights concerningthe fundamental nature of chemical reactionsthat are as valid today as they were 40 yearsago when the calculations were performed.

As in many of my papers in which a newmethod was developed (69), I tried to presentthe detail necessary for the reader to repro-duce and use what had been done. (We hadto work through it all, so why not save oth-ers the effort?) The requirement that thework be reproducible is often cited as a stan-dard for publishing papers, but, in practice,few papers are written in this way. Recently,I was pleased to learn that our paper wascited by George Schatz (105) as one of thekey twentieth-century papers in theoreticalchemistry. Schatz pointed out that what wehad called the quasiclassical trajectory method(classical trajectories with quantized initialconditions, which are very important for theH2 molecule because of its large zero-pointvibrational energy, some of which is avail-able in the transition state) was still widelyused, even though now full quantum dynam-ics calculations for the H + H2 reactions anda small number of other reactions are avail-

able. Moreover, Schatz states, “The KPS pa-per stimulated research in several new direc-tions and ultimately spawned new fields.”

After the H + H2 reaction calculation,Dick Porter and I collaborated on the text-book Atoms and Molecules (68), which devel-oped quantum chemistry at the introductorylevel for students of physical chemistry. Itwas based on a lecture course I had givenat Columbia and then at Harvard Univer-sity. (Teaching is a very good discipline, whichforces the instructor to understand a subjectwell enough so it can presented in a fash-ion clear enough for students to understand.)We thought that writing a book based on thelecture notes would be simple to do, but infact it was an enormous task. The book wasfinished only because Dick and I were to-gether for a summer on Martha’s Vineyardat a “writing camp” for scientists, sponsoredby an imaginative young publisher, Bill Ben-jamin. Later, while spending a semester at theWeizmann Institute in Rehovot, Israel, I usedpart of the time to correct the proofs, whichin my case inevitably led to significant rewrit-ing. The book has been a success, particularlyas a source of material for teachers of physicalchemistry.

Unlike quantum dynamics calculations,the quasiclassical trajectory method was eas-ily extended to more complex systems. Onestudy that I remember as particularly interest-ing was done by Martin Godfrey, that of theK + CH3I → KI + CH3 reaction, in whichthe collisions involved an orientated CH3Imolecule (63). We were stimulated to do thiscalculation by an ingenious experiment per-formed by Richard Bernstein, who was oneof the outstanding contributors to the field ofcrossed molecular chemistry but died young.He, indeed, oriented the CH3I reactant rela-tive to the incoming K+ ion so that one couldstudy the effect of the orientation on the re-action and obtain additional information forcomparison with the calculation; the two pa-pers were published together in the Journal ofthe American Chemical Society (6).

26 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

RETURN TO HARVARDUNIVERSITY AND BIOLOGY

In 1965, it was time to move again. Columbiaand New York City were stimulating places tolive and work, but I felt that new colleagues ina different environment would help to keepmy research productive. I had incorporatedthis idea into a “plan”: I would change schoolsevery five years and when I changed schoolsI would also change my primary area of re-search. It was more exciting for me to workon something new, where I had much to learnso as to stay mentally young and have newideas.

When it became known that I was plan-ning to leave Columbia University, numerousschools invited me to join their faculty. Withconsiderable difficulty, I narrowed the choicedown to U.C. Berkeley and Harvard and de-cided to visit each place for a semester dur-ing a sabbatical year (1965–1966). I enjoyedmy stay in both places; discussions with col-leagues were stimulating and it was very hardto decide between the two schools. I particu-larly enjoyed my interactions at U.C. Berke-ley with Bob Harris, a gifted theoretician who10 years before had been my very first researchstudent while he was an undergraduate at theUniversity of Illinois. We spent many hourstogether talking science and politics. This wasthe era of the Vietnam antiwar movement,and I was introduced to police brutality dur-ing some of the marches in Berkeley, a centerof the movement, and particularly in neigh-boring Oakland, which supported the war, incontrast to the Berkeley citizens.

This was not my first experience with po-litical protest, however. I had participated inthe 1950s as a Caltech graduate student inmeetings organized against the death penaltysentences of Julian and Ethel Rosenberg. Inthe mid-1970s, my brother, Bob, was initiallyrefused a security clearance to do research atthe Livermore Laboratory. Because his par-ticipation was deemed important to the labo-ratory, the administrators appealed and foundout that the reason for the refusal was my pres-

ence at a Unitarian Church in Los Angelesat a Rosenberg protest meeting two decadesprevious. Once this item in his FBI file wasexposed, Bob rapidly received the necessaryclearance.

By contrast, the one time I needed a clear-ance, it was not granted. I was invited to par-ticipate in a disarmament study sponsored inWoods Hole by the National Academy ofSciences in the late 1970s. Every participanthad to have a security clearance so that whenthe report was issued, no one could say thatthe people involved did not have access tothe required information. I had not requesteda clearance previously, and it was arrangedthat I apply for one. I arrived in Woods Holeand attended the opening meeting, which waspublic, but was not allowed into the sessionsafter that. While awaiting my clearance to par-ticipate in the month-long meeting, I relaxedwith my family and worked on my own re-search in a pleasant nearby hotel with a swim-ming pool, courtesy of the sponsors. Thisalso permitted me to renew my ties with theMarine Biological Laboratory. The disturb-ing element in my FBI record was the one thathad raised problems for my brother, namelythat I had attended the Rosenberg protestmeeting. It is still amazing to me that theU.S. government wastes so much time andmoney recording the attendance at meet-ings that one should have the right to at-tend without fear of future harassment ordiscrimination.

Another such experience occurred while Iwas on the faculty of the University of Illi-nois in 1955. Shortly after I moved there, Ibegan to receive a number of visits from FBIagents. There were always two of them; onewas from the local (Chicago) office and theother was a different person each time. Theyquestioned me about my political views andfocused on the time I had spent in Yugoslaviawhile I was a postdoctoral fellow at Oxford. Ihad been there as a tourist and photographer,as already mentioned, but the FBI apparentlyhad other ideas. On one such visit, the “chang-ing” agent suddenly began speaking in a


ANRV275-BB35-01 ARI 10 April 2006 15:34

language I did not understand. It turned outto be Serbian and was apparently meant totrick me. What exactly the FBI was lookingfor I never discovered. The visits stopped af-ter about six months.

A variety of factors led me to chooseHarvard, one of which, in retrospect, wasthe fact that I had been a Harvard under-graduate. Also, I felt that the Berkeley envi-ronment and its weather were just too niceand that the distractions from work were toogreat, as evident from the activities of a con-siderable portion of the Berkeley chemistryfaculty at that time. I came to Harvard asProfessor in the fall of 1966 and received thetitle of Theodore William Richards Professorin Chemistry in 1979. Although such chairsmean little at Harvard, other than that thefunds for one’s salary come out of a specialendowment, I was pleased to receive this ti-tle for two reasons. First, the previous holderof the chair had been E. Bright Wilson, ahighly respected member of the departmentfor his science, his humanity in dealing withstudents, and his high standards of intel-lectual honesty. Second, this chair was theonly one in chemistry named after a scientist(the first American to win the Nobel Prizein Chemistry) instead of the donor of thefunds.

At Harvard I continued to do research inthe some of the areas that I had developedat Illinois and Columbia, including the studyof hyperfine interactions in ESR (102) and theuse of quasiclassical trajectory methods for thestudy of reactions (37). With the results fromthe trajectory calculations, Keiji Morokuma,B.C. Eu, and I undertook a study of the rela-tion between reaction cross sections and tran-sition state theory as a test of this widely usedmodel in chemistry (93, 94). The applicationof many-body theory to the electronic struc-ture of atoms and molecules (17, 31), as anextension of methods developed in physics,was also of interest to me, in part to under-stand better what was involved in this devel-opment. I have always found it very helpfulwhen I see a new idea or method to apply it to

some problem, even if the specific research isnot that significant.

After I had been at Harvard for only ashort time, I realized that if I was ever to re-turn to my long-standing interest in biologyI had to make a break with what had beenthus far a successful and very busy researchprogram in theoretical chemistry. I felt thatI had grasped what was going on in elemen-tary chemical reactions and the excitement oflearning something new was no longer there.The initial qualitative insights obtained fromrelatively simple approaches to a new prob-lem are often the most rewarding. This is notto imply that the field of gas-phase chemi-cal reactions has not continued to flourish.It is still active, with ever-finer details con-cerning reactive collisions being elucidated.For example, I very much enjoyed attend-ing a meeting on reaction dynamics at theFritz Haber Institute in Berlin in 1982, whereI learned about the exciting research goingon; I, however, gave a lecture on protein dy-namics (58). The meeting also brought hometo me that other people with skills differentfrom mine are better able to contribute to theadvanced technologies now required in thisarea.

I planned to take a six-month leave in thefall of 1969 and chose the Weizmann Institute,in part because it had an excellent library. I wasaware of Shneior Lifson’s work on polymertheory and his reputation of being an open-minded scientist, as well as a marvelous story-teller. I wrote Shneior asking whether I couldcome for a semester, and he kindly invited meto join his group. The sabbatical gave me theleisure to read and explore a number of areasin which I hoped to do constructive researchby applying my expertise in theoretical chem-istry to biology. Discussions with Shneiorand visitors to his group helped me in theseexplorations.

A key, although accidental, element in mychoice of a problem for study in biologywas the publication of Structural Chemistryand Molecular Biology, a compendium of pa-pers in a volume dedicated to Linus Pauling

28 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

for his 65th birthday. I had contributed an ar-ticle entitled, “Structural Implications of Re-action Kinetics,” (57) which reviewed someof the work I have already described inthe context of Pauling’s view that a knowl-edge of structure was the basis for under-standing reactions. However, it is not myarticle that leads me to mention this vol-ume, but rather an article by Ruth Hub-bard and George Wald entitled, “Paulingand Carotenoid Stereochemistry.” They re-viewed Pauling’s contribution to the under-standing of polyenes with emphasis on thevisual chromophore, retinal. The article con-tained a paragraph, which I reproduce herebecause it describes an element of Pauling’sapproach to science that greatly influenced myresearch:

“One of the admirable things about LinusPauling’s thinking is that he pursues it alwaysto the level of numbers. As a result, there isusually no doubt of exactly what he means.Sometimes his initial thought is tentative be-cause the data are not yet adequate, and thenit may require some later elaboration or re-vision. But it is frequently he who refines thefirst formulation.”

On looking through the article, it was clearto me that the theory of the electronic absorp-tion of retinal and its geometric changes onexcitation, which play an essential role in vi-sion, had not advanced significantly since mydiscussions with Hubbard and Wald duringmy undergraduate days at Harvard. I realized,in part from my time in Oxford with Coul-son, that polyenes, such as retinal, were idealsystems for study by the available semiem-pirical approaches; that is, if any biologicallyinteresting system in which quantum effectsare important could be treated adequately atthat time, retinal was it. Barry Honig, whohad received his PhD in theoretical chemistryworking with Joshua Jortner, joined my re-search group at that time. He was the perfectcandidate to work on the retinal problem. Itwas known that the retinal (see Figure 3) isnot planar. It is twisted about the C12–C13 sin-gle bond, and this was thought to play a role in

Figure 3The different form of retinal considered in the calculations. The figure hasbeen reproduced, with permission, from Reference 42.

the photoisomerization reaction (the C11–C12

double bond charges, from cis to trans) thatgives rise to the visual signal. No structureof retinal was available, so it was not knownwhether the twist led to a 12-s-cis or 12-s-transconfiguration (Figure 3). Honig did a calcula-tion with a Huckel one-electron Hamiltonianfor the π-electron system and a pairwise non-bonded energy function for the sigma bondframework of the molecule (42). The theorypredicted that the structure was 12-s-cis.

We felt that this result with its implicationfor visual excitation was appropriate for pub-lication in Nature. We submitted the paper,which received excellent reviews, but cameback with a rejection letter stating that be-cause there was no experimental evidence tosupport our results, it was not certain that theconclusions were correct. This was my first ex-perience with Nature and with the difficulty ofpublishing theoretical results related to biol-ogy, particularly in high impact journals. Theproblem is almost as prevalent today as it wasthen, i.e., if theory agrees with experiment itis not interesting because the result is alreadyknown, whereas if one is making a predic-tion, then it is not publishable because thereis no evidence that the prediction is correct. Iwas sufficiently upset by this reaction to ourwork that I called John Maddox, the insightful


ANRV275-BB35-01 ARI 10 April 2006 15:34

Editor of Nature, and explained the situationto him. Apparently, I was successful, as the pa-per was accepted. A subsequent crystal struc-ture verified our prediction (36). In a reviewof studies of the visual chromophore (43), Inoted that “Theoretical chemists tend to usethe word ‘prediction’ rather loosely to refer toany calculation that agrees with experiment,even when the latter was done before theformer.”

The study of the retinal chromophore gaverise to a sustained effort in my group con-cerned with the properties of retinal and otherpolyenes. The continuing effort was fostered,in part, by Bryan Kohler and Bryan Sykes,two assistant professors who had joined theChemistry Department and were doing ex-periments that provided challenges for theory.They were part of a group of young faculty(it also included William Reinhardt and RoyGordon, plus William Miller, a Junior Fel-low), which made the department particularlystimulating at that time. Their offices were lo-cated along a narrow corridor on the groundfloor of Converse Laboratory, with my officeat one end. My daily strolls down this corridorprovided many occasions for scientific discus-sion. A collaboration with Sykes led to one ofthe earliest uses of vicinal spin-spin couplingconstants and the nuclear Overhauser effectin NMR to determine the conformations ofa biomolecule (retinal in this case) (41). Thetechnique, in a much more elaborate imple-mentation, is now the basis of most proteinNMR structure determinations. Kohler andhis student, Bruce Hudson, were doing highresolution spectral studies of simple polyenes(e.g., hexatriene) and had observed a very weakabsorption below the strongly absorbing tran-sition, which is the analog of the one involvedin retinal isomerization. They suggested thatthere exists a forbidden transition, which wasnot predicted by simple (single-excitation)models of polyene spectra, such as the oneused by Honig in his study of retinal. KlausSchulten, then a graduate student, workingjointly with Roy Gordon and me, introduceddouble excitations into the Parisar-Parr-Pople

(PPP) approximation for π-electron systemsand found the low-lying (forbidden) state inhexatriene and octatetraene (108). A numberof related studies followed. Arieh Warshel hadjoined my group at Harvard after we met atthe Weizmann Institute, where he had beena graduate student with Lifson. He extendedthe polyene model by introducing a quantummechanical Hamiltonian that refined the PPPmethod for the π-electrons and by treatingthe sigma-bonded framework by a molecularmechanics approach fitted to a large set of ex-perimental data (120); the method, like thesimpler model used by Honig, was an earlyversion of the quantum mechanical/molecularmechanical (QM/MM) approach that is nowwidely employed for studying enzymatic re-actions (28). We used the method to calcu-late the vibronic spectra of retinal and relatedmolecules (120). Subsequently, a collabora-tion was initiated with Veronica Vaida, amember of the chemistry faculty, and hergraduate student Russ Hemley. He extendedthe approach we had developed for excitedstates to molecules such as styrene (39),which Vaida and her students were studyingexperimentally.

In the 1970s I moved to Mallincrodt fromthe Converse offices, where the large am-phitheatre lecture hall had been renovatedinto a three-story integrated space to housethe physical chemistry faculty and the theo-retical students. The renovated area, known asthe “New Prince House” (Prince House wasan old Cambridge house near the ChemistryDepartment where the theoretical studentshad offices for a number of years) promotedinteraction among all occupants—senior andjunior faculty and the theoretical postdocs andgraduate students who had offices in the lowerdepths of the tri-level complex. Its loungearea equipped with an espresso coffee machinewas ideal for generating discussions. Amongmy many interactions over the 20-year pe-riod this complex existed, none proved morefruitful than those with Chris Dobson, whowas a junior faculty member in the depart-ment from 1978 to 1980 before returning to

30 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

Oxford. Our collaborations continue to thisday, as described below.

HEMOGLOBIN: A REALBIOLOGICAL PROBLEM

Another scientific question that appearedready for a more fundamental investigationwas the origin of hemoglobin cooperativity,the model system for allosteric control in bi-ology. Although the phenomenological modelof Monod, Wyman, and Changeux (92) hadprovided many insights, it did not attempt tomake contact with the detailed structure ofthe molecule. I had already begun workingon hemoglobin with Robert Shulman, then atBell Labs, who had measured the paramag-netic NMR shifts of the heme protons, andwe had developed an interpretation of the re-sults on the basis of the electronic structureof the heme (111). In 1971 Max Perutz hadjust determined the X-ray structure of deoxyhemoglobin, which complemented his earlierresults for oxy hemoglobin (98). By compar-ing the two structures, he was able to pro-pose a qualitative molecular mechanism forthe cooperativity. Alex Rich, now a professorat the Massachusetts Institute of Technology,had invited Perutz to present two lecturesdescribing the X-ray data and his mecha-nism. After the second lecture, Alex suggestedthat I come to his office to have a discussionwith Perutz. Perutz was sitting on a couch inAlex’s office and eating his customary banana.I asked him whether he had tried to formulatea quantitative thermodynamic mechanismbased on his structural analysis. He said noand seemed very enthusiastic, although I wasnot sure whether he had understood what Imeant. Having been taught by Pauling thatuntil one expressed an idea in quantitativeterms, it was not possible to test one’s re-sults, I went away from our meeting thinkingabout the best way to proceed. Attila Szabohad recently joined my group as a gradu-ate student, and the hemoglobin mechanismseemed like an ideal problem for his theo-retical skills. The basic idea proposed by Pe-

rutz was that the hemoglobin molecule hastwo quaternary structures, R and T, in agree-ment with the ideas of Monod, Wyman, andChangeux; that there are two tertiary struc-tures, liganded and unliganded for each of thesubunits; and that the coupling between thetwo is introduced by certain salt bridges whoseexistence depended on both the tertiary andquaternary structures of the molecule. More-over, some of the salt bridges depended onpH, which introduced the Bohr effect on theoxygen affinity of the subunits. These ideaswere incorporated into the statistical mechan-ical model Szabo and I developed (115). Itwas a direct consequence of the formulationthat the cooperativity parameter n (i.e., theHill coefficient) varied with pH. This was indisagreement with the hemoglobin dogma atthe time and led a number of the experimen-talists in the field to initially disregard ourmodel, which was subsequently confirmed byexperiments.

When we began working on the model,I discussed our approach with John Edsalland Guido Guidotti, both biology professorsat Harvard. Edsall was well known for hisdeep understanding of protein thermodynam-ics and Guidotti was an expert on hemoglobin.There were a number of parameters in themodel and we had chosen their values byuse of physical arguments. Because the val-ues of the parameters were estimated, the re-sults from the model gave only approximateagreement with experiment. Guidotti warnedme that such results would not be accepted bythe hemoglobin community, in particular, andbiologists, in general. Consequently, we in-verted the description of the model. We usedexperimental data to determine the parame-ters so that the agreement with experimentwas excellent and then justified the values ofthe parameters with the physical argumentswe had developed. During the formulation ofour ideas, we often asked Guidotti which ofcertain experiments were to be trusted, sincethe nearly overwhelming hemoglobin litera-ture contained sets of data that disagreed witheach other, without any comment from the


ANRV275-BB35-01 ARI 10 April 2006 15:34

authors indicating which measurements werecorrect and why.

The paper describing the hemoglobinwork was written in Paris, much of it at AuxDeux Magots, a left-bank cafe famous as ameeting place for writers and philosophersfrom the time of Jean-Paul Sartre. I was onsabbatical leave during 1972–1973 and offi-cially at the Universite de Paris XI in Orsay,a suburb of Paris, with the group of JeannineYon-Kahn, a pioneer in experimental studiesof protein dynamics. However, I spent muchof my time in Paris at the Institut de BiologyPhysico-Chimique on rue Paul et Marie Curiein the 5th Arrondissement. Having often vis-ited Paris since my postdoctoral days and livedthere on a sabbatical, I had begun to considerthe possibility of moving to Paris on a per-manent basis in 1970. I had been at Harvardfor the canonical five years, and the idea ofreturning to Europe was tempting. Given theanti-Semitism and Nazi-leaning parties thatstill were prevalent in Austria, I had no desireto return to the country of my birth. Franceoffered many attractive aspects of Europeanlife and culture, and I believed that I could dohigh-level research there in theoretical chem-istry and its biological applications. After the1968 revolution, the immense Universite ofParis, with more than 300,000 students, hadbeen divided into a dozen campuses. Withtrue French rigor, they were named Paris I,Paris II, etc., although they now have names,instead of merely numbers. Orsay (Paris XI)was one of three science campuses, and it wascertainly the best. However, it had the draw-back that it was about 40 minutes by the RER(commuter rail) from Paris. If I was going tomove to a Paris university, I wanted to livein Paris itself. Consequently, I focused on thetwo other scientific universities (Paris VI andParis VII) that were intertwined on the Jussieucampus, a block of ugly modern buildings.Their saving grace was a central location inthe area where the Halles aux Vins had beenlocated before World War II. The neighbor-ing streets were still dotted with good inex-pensive restaurants dating back to the area’s

previous existence and now thriving on thefaculty and student clientele.

In discussions with colleagues who hadurged me to live in France, a serious obsta-cle became clear. I was a tenured professor atHarvard and not surprisingly was willing tomove only if I was offered a permanent po-sition in Paris. However, French universityprofessors were civil servants and only Frenchcitizens could be civil servants. Because ob-taining French citizenship without losing myAmerican one was out of the question at thetime (it is now possible and our son, Mischa,has dual citizenship), I was ready to give upthe idea of moving to Paris. Many things inFrance were achieved then (and still are) bypolitical influence. Jacques DuBois, a chem-istry professor at Paris VII with connections tothe Pompidou government, said he would tryto “arrange the situation.” I did not know ex-actly what he meant but hoped that the tenureproblem could be solved. On that basis I tooka leave of absence from Harvard.

With only a verbal commitment of a per-manent position, I moved the major part of myresearch group (including David Case, BruceGelin, and Iwao Ohmine, among others) fromHarvard in the fall of 1974. At Paris VII emptylaboratory spaces awaited us. We bought of-fice furniture and computing equipment andwent to work. One thing that made the transi-tion much easier was that Marci Hazard, whohad joined the lab as secretary in May, camealong. Many of the logistical problems (e.g.,finding where to purchase what we needed)were solved by her, and she played a key rolein the cohesion of the group. As the yearwent on, DuBois reported on his progress inregularizing my status. I had come as a Pro-fesseur Associe, which is an annual appoint-ment open to non-French citizens. Finally, inJanuary a decree was published in the offi-cial government register (much of the Frenchgovernment functions by decrees that do notrequire votes of the National Assembly). Thisdecree exempted university professors fromthe citizenship requirement, which made pos-sible my appointment as a tenured professor.

32 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

Not everything was resolved, however, andthe complexity of dealing with the French ad-ministration led me to renounce my dreamand return to Harvard. The decree remainedvalid, however, and subsequently I receiveda number of thank-you letters from non-French scientists who had for many yearsbeen appointed annually as Professeur Associeand suddenly received a permanent position.( Jean-Pierre Hanson, who was born in Lux-embourg, told me recently that he believes heis one of the first people to have profited from“my” decree.)

During this period I started spending sum-mer vacations with my family in the foothillsof the Alps above Annecy and its stunninglake, an area which I had first seen on mypostdoctoral trips in the early 1950s. My col-leagues at Harvard viewed such absences fromCambridge as improper. However, I foundthat being away gave me a chance to think,undistracted by everyday pressures. Contem-plative hikes in the mountains provided thebackdrop for my reading and thinking andplayed an essential role in developing new ar-eas of research. In fact, I had asked my NSFprogram director whether these trips werejustified under the conditions of my grant. Hisconclusion was that, given their importanceto my research, they constituted an appropri-ate, even if somewhat unorthodox, summerprogram. In 1974 I finally found a plot ofland with a magnificent view in Chalmont, asmall hamlet in the Manigod valley above Lacd’Annecy. A chalet was built, which has beenour summer home for 30 years.

PROTEIN FOLDING

In 1969 I was challenged by the mechanismof protein folding during a visit by Chris An-finsen to the Lifson group at the WeizmannInstitute. We had many discussions of hisexperiments on protein folding, which hadled to the realization that proteins can refoldin solution, independent of the ribosome andother aspects of the cellular environment (2).[Of course, it is now known that some proteins

have more complex folding mechanisms andrequire chaperones, such as the supramolec-ular complex GroEL, to fold. This molecu-lar machine is one for which we have usedmolecular dynamics simulations to elucidatethe mechanism (87, 118).] What most im-pressed me was Anfinsen’s film showing thefolding of a protein with “flickering helicesforming and dissolving and coming togetherto form stable substructures.” The film was acartoon, but it led to my asking him, in thesame vein as I had asked Perutz earlier abouthemoglobin, whether he had thought of tak-ing the ideas in the film and translating theminto a quantitative model. Anfinsen said thathe did not really know how he would do this,but to me it suggested an approach to themechanism of protein folding. When DavidWeaver joined my group at Harvard, whileon a sabbatical leave from Tufts, we developedwhat is now known as the diffusion-collisionmodel for protein folding (70, 71). Althoughit is a simplified coarse-grained description ofthe folding process, it showed how the searchproblem for the native state could be solvedby a divide-and-conquer approach. Formu-lated by Cy Levinthal, the so-called LevinthalParadox points out that to find the nativestate by a random search of the astronomi-cally large configuration space of a polypep-tide chain would take longer than the age ofthe earth, while proteins fold experimentallyon a timescale of microseconds to seconds.In addition to providing a conceptional an-swer to the question posed by Levinthal, thediffusion-collision model made possible theestimation of folding rates. The model wasahead of its time because data to test it werenot available. Only relatively recently haveexperimental studies demonstrated that thediffusion-collision model describes the fold-ing mechanism of many helical proteins (48),as well as some others (49).

Protein folding is an area that has con-tinued to interest me and has led to numer-ous collaborations in addition to that withDavid Weaver. When David and I developedthe diffusion-collision model in 1975, protein


ANRV275-BB35-01 ARI 10 April 2006 15:34

folding was a rather esoteric subject of interestto a very small community of scientists. Thefield has been completely transformed in re-cent years because of its assumed importancefor understanding the large number of proteinsequences available from genome projects andbecause of the realization that misfolding canlead to a wide range of human diseases (22);these diseases are found primarily in the olderpopulations that form an ever-increasing por-tion of humanity. Scientists, both experimen-talists and theoreticians—physicists, as wellchemists and biologists—now study proteinfolding. Over the past decade or so the mech-anism of protein folding has been resolved,in principle. It is now understood that thereare multiple pathways to the native state andthat the bias on the free-energy surface, dueto the greater stability of native-like versusnonnative contacts, is such that only a verysmall fraction of the total number of confor-mations is sampled in each folding trajectory(24). This understanding was achieved by thework of many scientists, but a crucial elementwas the study of lattice models of protein fold-ing. Such toy models, as I like to call them,are simple enough to permit many foldingtrajectories to be calculated to make possiblean analysis of the folding process and free-energy surface sampled by the trajectories(104). However, they are sufficiently complexso that they embody the Levinthal problem,i.e., there are many more configurations thancould be visited during the calculated foldingtrajectory. The importance of such studies wasin part psychological, in that even though thelattice model uses a simplified representation,“real” folding was demonstrated on a com-puter for the first time. An article based on alecture at a meeting in Copenhagen (60) de-scribes this change in attitude as a paradigmof scientific progress.

The mechanism of protein folding andthe development of methods to predict thestructure of a protein from its amino acid se-quence continue to be subjects under intenseinvestigation. It is likely that calculations withprograms such as CHARMM will permit fold-

ing simulations to be done at an atomic levelof detail owing to the ever-increasing speedof computers (either localized multiprocessorsupercomputers or delocalized grid-based ac-cess to many individual processors) before toolong. However, this type of brute force ap-proach is of less interest to me than solving theconceptual problem of protein folding, whichhas been accomplished by more approximatetechniques.

ORIGINS OF THE CHARMMPROGRAM

When I visited Lifson’s group in 1969 therewas considerable interest in developing em-pirical potential energy functions for smallmolecules. The novel idea was to use a func-tional form that could serve not only forcalculating vibrational frequencies, as did theexpansions of the potential about a known orassumed minimum-energy structure, but alsofor determining that structure. The so-calledconsistent force field (CCF) of Lifson andhis coworkers, particularly Arieh Warshel,included nonbonded interaction terms sothat the minimum-energy structure could befound after the energy terms had been ap-propriately calibrated (84). The possibility ofusing such energy functions for larger sys-tems struck me as potentially very importantfor understanding biological macromoleculeslike proteins, though I did not begin workingon this immediately.

Once Attila Szabo had finished the statisti-cal mechanical model of hemoglobin cooper-ativity, I realized that his work raised a numberof questions that could be explored onlywith a method for calculating the energy ofhemoglobin as a function of the atomic po-sitions. No way of doing such a calculationexisted. Bruce Gelin, a new graduate student,had begun theoretical research in my groupin 1967. He started out by studying the ap-plication of the random-phase approximationto two-electron systems, such as the heliumatom. This was still the Vietnam War era andafter two years at Harvard, Gelin was drafted.

34 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

He was assigned to the military police in alaboratory concerned with drug usage (e.g.,LSD). Paradoxically, this work aroused his in-terest in biology, and when he returned tofinish his degree Gelin wanted to change hisarea of research to a biologically related prob-lem. We decided the time was ripe to try todevelop a program that would make it possi-ble to take a given amino acid sequence (e.g.,that of the hemoglobin alpha chain) and a setof coordinates (e.g., those obtained from theX-ray structure of deoxy hemoglobin) and touse this information to calculate the energyof the system and its derivatives as a functionof the atomic positions. This could be usedfor perturbing the structure (e.g., by bindingoxygen to the heme group) and finding a newstructure by minimizing the energy. Develop-ing the program was a major task, but Gelinhad the right combination of abilities to carryit out (33).

The result was pre-CHARMM, althoughit did not have a name at that time. While nottrivial to use, the program was applied to a va-riety of problems, including Gelin’s pioneer-ing study of aromatic ring flips in the bovinepancreatic trypsin inhibitor (BPTI) (34), aswell as his primary project on hemoglobin.The idea was to introduce the effect of lig-and binding on the heme group as a pertur-bation (undoming of the heme) and to useenergy minimization to determine the re-sponse of the protein to the perturbation. Toattempt to do such a calculation on the avail-able computers (an IBM 7090 at ColumbiaUniversity was our work-horse at the time,because computing at the Harvard ComputerCenter was too expensive) required consid-erable courage, but Gelin’s efforts were suc-cessful. His work introduced a new dimensionto theoretical approaches to understandingprotein structure and function. Gelin showedhow the effect of undoming of the heme in-duced by the binding of oxygen was transmit-ted to the interface between the hemoglobinsubunits. The analysis provided an essen-tial element in the cooperative mechanism inits demonstration at an atomic level of de-

tail how communication between the sub-units occurred (35). Another application ofpre-CHARMM was Dave Case’s simulationof ligand escape after photodissociation frommyoglobin (16); a study that was followed bythe work of Ron Elber (25), which gave riseto the locally enhanced sampling (LES) andmultiple copy simultaneous search (MCSS)methods. The latter was developed by AndrewMiranker as a fragment-based approach todrug design (90).

Gelin would have faced an almost insur-mountable task in developing pre-CHARMMif there had not been prior work by oth-ers on protein energy calculations. Althoughmany persons have contributed to the devel-opment of empirical potentials, the two ma-jor inputs to our work came from SchneiorLifson’s group at the Weizmann Institute andHarold Scheraga’s group at Cornell Univer-sity (107). As I already mentioned, Warshelhad come to Harvard and had brought hisCFF program with him. His presence and theavailability of the CFF program were impor-tant resources for Gelin, who was also awareof Michael Levitt’s pioneering energy calcu-lations for proteins (82).

Gelin’s program has been considerably re-structured and has continued to evolve overthe intervening years. In preparing to pub-lish a paper on the program in the early1980s, mainly to give credit to the domi-nant contributors at the time, we felt weneeded a name. Bob Bruccoleri came up withHARMM (HARvard Macromolecular Me-chanics), which seemed to me not to be theideal choice. However, Bob’s suggestion in-spired the addition of a “C” for chemistry,resulting in the name CHARMM. I some-times wonder if Bruccoleri’s original sugges-tion would have served as a useful warningto inexperienced scientists working with theprogram. The CHARMM program is nowbeing developed by a wide group of contrib-utors, most of whom were students or post-doctoral fellows in my group; the program isdistributed worldwide in both academic andcommercial settings.


ANRV275-BB35-01 ARI 10 April 2006 15:34

THE FIRST MOLECULARDYNAMICS SIMULATION OFA BIOMOLECULE

Given that pre-CHARMM could calculatethe forces on the atoms of a protein, the nextstep was to use these forces in Newton’s equa-tion to calculate the dynamics. This funda-mental development was introduced in themid-1970s when Andy McCammon joinedmy group. An essential element that encour-aged us in this attempt was the existence ofmolecular dynamics simulation methods forsimpler systems. Molecular dynamics had fol-lowed two pathways, which come together inthe study of biomolecule dynamics. One path-way concerns trajectory calculations for sim-ple chemical reactions. My own research inthis area had served as preparation for themany-article problem posed by biomolecules.The other pathway in molecular dynamicsconcerns physical rather than chemical inter-actions and the thermodynamic and dynamicproperties of large numbers of particles ratherthan detailed trajectories of a few particles.Although the basic ideas go back to van derWaals and Boltzmann, the modern era beganwith the work of Alder and Wainright (1) onhard-sphere liquids in the late 1950s. The pa-per by Rahman (103) in 1964 on a molecu-lar dynamics simulation of liquid argon witha soft-sphere (Lennard-Jones) potential rep-resented an essential next step. Simulations ofmore complex fluids followed; the now classicstudy of liquid water by Stillinger and Rah-man was published in 1974 (114), shortly be-fore our protein simulations.

The background I have outlined set thestage for the development of molecular dy-namics of biomolecules. The size of an in-dividual molecule, composed of 500 or moreatoms for even a small protein, is such thatits simulation in isolation can serve to obtainapproximate equilibrium properties, as in themolecular dynamics of fluids. Concomitantly,detailed aspects of the atomic motions are ofconsiderable interest, as in trajectory calcu-lations. A basic assumption in initiating such

studies was that potential functions could beconstructed which were sufficiently accurateto give meaningful results for systems as com-plex as proteins or nucleic acids. In addition,it was necessary to assume that for these inho-mogeneous systems, in contrast to the homo-geneous character of even complex liquids likewater, simulations of an attainable timescale(10 to 100 ps) could provide a useful sam-ple of the phase space in the neighborhoodof the native structure. There was no com-pelling evidence for either assumption in theearly 1970s. When I discussed my plans withchemistry colleagues, they thought such cal-culations were impossible, given the difficultyof treating few atom systems accurately; biol-ogy colleagues felt that even if we could dosuch calculations, they would be a waste oftime. By contrast, the importance of molec-ular dynamics simulations in biology wassupported by Richard Feynman’s prescientstatement in the well-known volumes basedon his physics lectures at Caltech:

“Certainly no subject or field is mak-ing more progress on so many fronts at thepresent moment, than biology, and if we wereto name the most powerful assumption of all,which leads one on and on in an attempt tounderstand life, it is that all things are made ofatoms (italics in the original), and that every-thing that living things do can be understood interms of the jigglings and wigglings of atoms. (27;italics added)

More than 25 years have passed sincethe first molecular dynamics simulation of amacromolecule of biological interest was pub-lished (Figure 4) (88). This study has stoodthe test of time, and, perhaps more signif-icantly, it has served to open a new fieldthat is now the focus of the research of anever-growing number of scientists (10). Theoriginal simulation, published in 1977 (88),concerned the bovine pancreatic trypsin in-hibitor (BPTI), which has served as the“hydrogen molecule” of protein dynamics be-cause of its small size, high stability, and arelatively accurate X-ray structure (21); inter-estingly, the physiological function of BPTI

36 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

Figure 4The peptidebackbone (α carbons)and disulfide bondsof the bovinepancreatic trypsininhibitor drawn byBruce Gelin.(a) X-ray structure.(b) Time-evolvedstructure after 3.2 psof dynamicalsimulation. Thefigure has beenreproduced, withpermission, fromReference 88.

remains unknown. In the mid-1970s it wasdifficult to obtain the computer time requiredto do such simulations in the United States.However, CECAM (Center Europeen CalculAtomique et Moleculaire), whose foundingdirector and guiding spirit was Carl Moser,had access to a large computer that was avail-able for scientific research. (Equivalent com-puters in the United States were found only inthe defense agencies and were not generallyaccessible.) A CECAM Workshop (a two-month workshop worthy of its name) was orga-nized by Herman Berendsen in 1976 with thetitle “Models for Protein Dynamics.” As hisintroduction to the workshop states: “Thus,the simulation of water was a first topic tobe studied. The application to proteins wasthen not foreseen in five or ten years to come.”Realizing that the workshop was a great op-portunity to do the required calculations,Andy McCammon and Bruce Gelin workedextremely hard to prepare and test a programfor the molecular dynamics simulation for the

workshop. The initial simulation (88), whichwas performed at the workshop, introducedmany others now active in the field (includingH. Berendsen, W. van Gunsteren, M. Levitt,and J. Hermans) to the possibility of doingsuch calculations (5).

Although the original simulation was donein a vacuum with a crude molecular mechan-ics potential and lasted for only 9.2 ps, the re-sults were instrumental in replacing the viewof proteins as relatively rigid structures [in1981, Sir D. L. Phillips commented, “Brassmodels of DNA and a variety of proteins dom-inated the scene and much of the thinking”(99)] with the realization that they were dy-namic systems whose internal motions playa functional role. Of course, there were al-ready experimental data, such as the hydrogenexchange experiments of Linderstrom-Langand his coworkers (44, 85), pointing in thisdirection. It is now recognized that the X-ray structure of a protein provides the aver-age atomic positions but that the atoms ex-


ANRV275-BB35-01 ARI 10 April 2006 15:34

hibit fluid-like motions of sizable amplitudesabout these averages. Protein dynamics sub-sumes the static picture. The average posi-tions are essential for the discussion of manyaspects of biomolecular function in the lan-guage of structural chemistry, but the recogni-tion of the importance of fluctuations openedthe way for more sophisticated and accurateinterpretations of functional properties.

The conceptual changes resulting from theearly studies make one marvel at how much ofgreat interest could be learned with so little—such poor potentials, such small systems, solittle computer time. This is, of course, one ofthe great benefits of taking the initial, some-what faltering steps in a new field in which thequestions are qualitative rather than quantita-tive and any insights, even if crude, are betterthan none at all.

APPLICATIONS OFMOLECULAR DYNAMICS

Molecular dynamics simulations of proteinsand nucleic acids, as of other systems com-posed of particles (e.g., liquids, galaxies), canin principle provide the ultimate details of mo-tional phenomena. The primary limitation ofsimulation methods is that they are approxi-mate. Here experiment plays an essential rolein validating the simulation methods; that is,comparisons with experimental data serve totest the accuracy of the calculated results andprovide criteria for improving the method-ology. Although the statistical errors can becalculated (122), estimates of the systematicerrors inherent in the simulations have notbeen possible, e.g., the errors introduced bythe use of empirical potentials are difficultto quantify. When experimental comparisonsindicate that the simulations are meaningful,their capacity for providing detailed results of-ten makes it possible to examine specific as-pects of the atomic motions far more easilythan by using laboratory measurements.

Two years after the BPTI simulation, itwas recognized (3, 30) that thermal (B) factorsdetermined in X-ray crystallographic refine-

ment could provide information about the in-ternal motions of proteins. Plots of estimatedmean-square fluctuations versus residue num-ber [introduced in the original BPTI paper(88)] have become a standard part of paperson high-resolution structures, even thoughthe contribution to the B factors of overalltranslation and rotation and crystal disor-der persist as a concern in their interpre-tation (79). During the subsequent decade,a range of phenomena were investigated bymolecular dynamics simulations of proteinsand nucleic acids. A plethora of experimen-tal data were just waiting for molecular dy-namics simulations to elucidate them. Mostof these early studies were made by my stu-dents at Harvard and focused on the phys-ical aspects of the internal motions and theinterpretation of experiments. They includethe analysis of fluorescence depolarization oftryptophan residues (45), the role of dynam-ics in measured NMR parameters (23, 83, 97)and inelastic neutron scattering (20, 113), andthe effect of solvent and temperature on pro-tein structure and dynamics (11, 29, 95). Thenow widely used simulated annealing methodsfor X-ray structure refinement (14, 15) andNMR structure determination (12, 96) alsooriginated in this period. Simultaneously, anumber of applications demonstrated the im-portance of internal motions in biomolecularfunction, including the hinge bending modesfor opening and closing active sites (9, 18),the flexibility of tRNA (38), the induced con-formation change in the activation of trypsin(13), the fluctuations required for ligand en-trance and exit in heme proteins (16, 25), andthe role of configurational entropy in the sta-bility of proteins and nucleic acids (8, 47).Many of these studies, which were done abouttwo decades ago, seem to have been forgotten;at least, they are rarely cited in the current lit-erature. Of course, when the studies are re-done, the more accurate potential functionsand much longer simulations (nanosecondsinstead of picoseconds) now possible yield im-proved results, but generally they confirm theearlier work.

38 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

Two attributes of molecular dynamics sim-ulations have played an essential part in the ex-plosive growth in the number of studies basedon such simulations. As already mentioned,simulations provide the ultimate detail con-cerning individual particle motions as a func-tion of time. For many aspects of biomoleculefunction, it is these details that are of interest(e.g., by what pathways does oxygen enter intoand exit from the heme pocket in myoglobin).The other important aspect of simulations isthat, although the potentials employed in sim-ulations are approximate, they are completelyunder the user’s control. By removing or al-tering specific contributions, their role in de-termining a given property can be examined.This is most graphically demonstrated by theuse of computer alchemy—transmuting thepotential from that representing one systemto another during a simulation—in calculatedfree-energy differences (32, 112, 121).

There are three types of applications ofsimulation methods in the macromoleculararea, as well as in other areas involving meso-scopic systems. The first uses the simulationsimply as a means of sampling configura-tion space. This is involved in the utilizationof molecular dynamics, often with simu-lated annealing protocols, to determine orrefine structures with data obtained fromexperiments. The second uses simulationsto determine equilibrium averages, includingstructural and motional properties (e.g.,atomic mean-square fluctuation amplitudes)and the thermodynamics of the system. Forsuch applications, it is necessary that thesimulations adequately sample configurationspace, as in the first application, with the ad-ditional condition that each point be weightedby the appropriate Boltzmann factor. Thethird application employs simulations to ex-amine the actual dynamics. Here not only isadequate sampling of configuration space withappropriate Boltzmann weighting required,but it must be done so as to properly rep-resent the time development of the system.Monte Carlo simulations, as well as molecu-lar dynamics, can be utilized for the first two

applications. By contrast, in the third appli-cation, where the motions and their time de-velopments are of interest, only molecular dy-namics can provide the necessary information.

FUTURE OF MOLECULARDYNAMICS

Most of the motional phenomena examinedduring the first 10 years after the BPTI sim-ulation paper was published continue to bestudied both experimentally and theoretically.The increasing scope of molecular dynam-ics due to improvements in methodology andthe tremendous increase of the available com-puter power is making possible the studyof systems of greater complexity on ever-increasing timescales. There are many recentexamples of the use of molecular dynamicsto obtain information concerning the func-tion of biological macromolecules. At presentI am most interested in molecular machines,such as GroEL (87, 118), which require sim-ulations for their understanding. Nature hasdesigned these machines to function throughligand-induced conformational changes en-coded in the structure. One of the mostfascinating machines is F0F1-ATPase, whichsynthesizes ATP and H2O, the energy cur-rency of most living cells, from ADP andH2PO4

− (Figure 5) (32a). Numerous appli-cations by my group and others have been re-viewed recently, so that I will not detail themhere (61, 65, 67).

It is my hope that molecular dynamics sim-ulations will become a tool, like any other, tobe used by experimentalists as part of theirarsenal for solving problems. The simu-lated annealing method for determining X-ray structures originally proposed in 1987 byAxel Brunger, John Kuriyan, and me (14, 15),and so ably developed by Brunger, is now anessential part of structural biology. Withoutthis method, the high-throughput structuredetermination initiatives would be in con-siderable difficulty. However, a universal ac-ceptance of molecular dynamics simulationsmethods for extending experimental data to


ANRV275-BB35-01 ARI 10 April 2006 15:34

Figure 5Structural model of F0F1-ATP synthase. The figure has been reproducedfrom Senior AE, Nadanaciva S, Weber J. 2002. Biochim. Biophys. ActaBioenerg. 1553:188–211

learn how biomolecules function is still inthe future. The determination of the reac-tion path involved in conformational changeby simulations, is a clear case where, given thestructural end point from experiment, simula-tions are essential, as some crystallographershave already recognized (123).

I still marvel at the insights that simulationmethods can provide concerning the func-tions of biomolecules. Claude Poyart, a dearfriend with whom I worked on hemoglobin(81), characterized molecular dynamics sim-ulations of proteins with a beautiful image.He likened the X-ray structures of proteins toa tree in winter, beautiful in its stark outlinebut lifeless in appearance. Molecular dynam-ics gives life to this structure by clothing thebranches with leaves that flutter because of thethermal winds.

EPILOGUE

As I read through what I have written, I seewhat a fragmentary picture it provides of mylife, even my scientific life. Missing are in-numerable interactions, most of which wereconstructive but some not so, that have playedsignificant roles in my career. The more than200 graduate students and postdoctoral fel-lows who at one time or another have beenmembers of the group are listed in Table 1.Many have gone on to faculty positions andbecome leaders in their fields of research.They in turn are training students, so I nowhave scientific children, grandchildren, andgreat-grandchildren all over the world. I trea-sure my contribution to their professional andpersonal careers, as much as the scientific ad-vances we have made together.

Contributing to the education of so manypeople in their formative years is a cardinal as-pect of university life. My philosophy in grad-uate and postgraduate education has been toprovide an environment where young scien-tists, once they have proved their ability, candevelop their own ideas, as refined in discus-sions with me and aided by other membersof the group. This fostered independence hasbeen, I believe, an important element in thefact that so many of my students are nowthemselves outstanding researchers and fac-ulty members. My role has been to guide themwhen problems arose and to instill in them thenecessity of doing things in the best possibleway, not to say that I succeeded with all ofthem.

Discussing my scientific family makes merealize that another missing element is mypersonal family, an irreplaceable part of mylife. Reba and Tammy, my two daughterswhose mother, Susan, died in 1982, bothbecame physicians (thereby fulfilling mydestined role); Reba lives in Jerusalem andTammy lives on the West Coast. My wife,Marci, and our son, Mischa, who is presentlyin law school at Boston University, completemy immediate family. As many peopleknow, Marci also plays the pivotal role as the

40 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

Table 1 Karplusians: 1955–2005

R. J. Aerni Kevin Gaffney Jianpeng Ma Andrej SaliDavid H. Anderson Jiali Gao Alexander D. MacKerell, Jr. Michael SchaeferIoan Andricioaei Yi Qin Gao Christoph Maerker Michael SchlenkrichYasuhide Arata Bruce Gelin Paul Maragkakis David M. SchraderGeorgios Archontis R. Benny Gerber Marc Martı-Renom John C. SchugGabriel G. Balint-Kurti Paula M. Getzin Jean-Louis Martin Klaus SchultenChristian Bartels Debra Giammona Carla Mattos Eugene ShakhnovichPaul Bash Martin Godfrey J. Andrew McCammon Moshe ShapiroDonald Bashford Andrei Golosov H. Keith McDowell Ramesh D. SharmaOren M. Becker David M. Grant Jorge A. Medrano Isaiah ShavittRobert Best Daniel Grell Morten Meeg Henry H.-L. ShihAnton Beyer Peter Grootenhuis Marcus Meuwly Bernard ShizgalRobert Birge Hong Guo Olivier Michielin David M. SilverRyan Bitetti-Putzer Robert Harris Stephen Michnick Jeremy SmithArnaud Blondel Karen Haydock Fredrick L. Minn Sung-Sau SoStefan Boresch Russell J. Hemley Andrew Miranker Michael SommerJohn Brady Jeffrey C. Hoch Keiji Morokuma Ojars J. SoversBernard Brooks Gary G. Hoffman A. Mukherji Martin SpichtyCharles L. Brooks III L. Howard Holley Adrian Mulholland David J. StatesThomas H. Brown Barry Honig David Munch Richard M. StevensRobert E. Bruccoleri Victor Hruby Petra Munih Roland StotePaul W. Brumer Rod E. Hubbard Robert Nagle John StraubAxel T. Brunger Robert P. Hurst Setsuko Nakagawa Collin StultzRafael P. Bruschweiler Vincent B.-H. Huynh Eyal Neria Neena SummersMatthias Buck Toshiko Ichiye John-Thomas C. Ngo Henry SuzukawaAmedeo Caflisch K.K. Irikura Lennart Nilsson S. SwaminathanWilliam J. Campion Alfonso Jaramillo Dzung Nguyen Attila L. SzaboWilliam Carlson Diane Joseph-McCarthy Iwao Ohmine Kwong-Tin TangDavid A. Case Sunhee Jung Barry Olafson Bruce TidorLeo Caves C. William Kern E.T. Olejniczak Hideaki UmeyamaThomas C. Caves Burton S. Kleinman Kenneth W. Olsen Arjan van der VaartJohn-Marc Chandonia G.W. Koeppl Neil Ostlund Wilfred van GunsterenTa-Yuan Chang H. Jerrold Kolker Emanuele Paci Herman van VlijmenRob D. Coalson Yifei Kong Yuh-Kang Pan Michele VendruscuoloFrancois Colonna-Cesari Lewis M. Koppel C.S. Pangali Dennis VitkupMichael R. Cook J. Kottalam Richard W. Pastor Shunzhou WanQiang Cui Felix Koziol Lee Pedersen Iris Shih-Yung WangAnnick Dejaegere Christoph Kratky David Perahia Ariel WarshelPhilippe Derreumaux Serguei Krivov Robert Petrella Masakatsu WatanabeAaron Dinner Krzysztof Kuczera B. Montgomery Pettitt David WeaverUri Dinur John Kuriyan Ulrich Pezzeca Paul WeinerRoland L. Dunbrack, Jr. Joseph N. Kushick Richard N. Porter Michael A. WeissChizuko Dutta Peter W. Langhoff Jay M. Portnow Joanna Wiorkiewicz-K.Nader Dutta Antonio C. Lasaga Carol B. Post George Wolken

(Continued )


ANRV275-BB35-01 ARI 10 April 2006 15:34

Table 1 (Continued )

Claus Ehrhardt Frankie T.K. Lau Lawrence R. Pratt Youngdo WonRon Elber Themis Lazaridis Martine Prevost Yudong WuByung Chan Eu Fabrice LeClerc Blaise Prod’hom Robert E. WyattJeffrey Evanseck Angel Wai-mun Lee Dagnija Lazdins Purins Wei YangErik Evensen Irwin Lee Lionel M. Raff Robert YelleJeffrey Evenson Sangyoub Lee Mario Raimondi Swarna Yeturu ReddyThomas C. Farrar Ronald M. Levy Walter E. Reiher III Darrin YorkMartin Field Xiaoling Liang Nathalie Reuter Hsiang-ai YuStefan Fischer Carmay Lim Bruno Robert Vincent ZoeteDavid L. Freeman Xabier Lopez Peter J. Rossky Yaoqi ZhouThomas Frimurer Paul Lyne Benoıt Roux

Laboratory Administrator, adding a spirit ofcontinuity for the group and making possi-ble our commuting between the Harvard and

Strasbourg labs. Without my family, my lifewould have been an empty one, even with sci-entific success.

ACKNOWLEDGMENTS

Much of the research described here was supported in part by grants from the National ScienceFoundation, the National Institutes of Health, the Atomic Energy Commission, and the De-partment of Energy, as well as the royalty income from the CHARMM Development Project.As is evident from the references, the research would not have been possible without my manycoworkers, only a few of whom are mentioned in the text. Frontispiece was kindly providedby Nathan Pitt, Department of Chemistry, University of Cambridge, Cambridge CB2 1EW,United Kingdom.

LITERATURE CITED

1. Alder BJ, Wainwright TE. 1957. Phase transition for a hard sphere system. J. Chem. Phys.27:1208–9

2. Anfinsen CB. 1973. Principles that govern folding of protein chains. Science 181:223–303. Artymiuk PJ, Blake CCF, Grace DEP, Oatley SJ, Phillips DC, Sternberg MJE. 1979.

Crystallographic studies of the dynamic properties of lysozyme. Nature 280:563–684. Balint-Kurti GG, Karplus M. 1969. Multistructure valence-bond and atoms-in-molecules

calculations for LiF, F2, and F2−. J. Chem. Phys. 50:478–885. Berendsen H. 1976. Report of CECAM Workshop: models for protein dynamics. Orsay,

May 24–July 176. Beuhler RJ, Bernstein RB, Kramer KH. 1966. Observation of reactive asymmetry of

methyl iodide crossed beam study of reaction of rubidium with oriented methyl iodidemolecules. J. Am. Chem. Soc. 88:5331–32

7. Bradshaw WH, Conrad HE, Corey EJ, Gunsalus IC, Lednicer D. 1959. Microbiologicaldegradation of (+)-camphor. J. Am. Chem. Soc. 81:5507

8. Brooks BR, Karplus M. 1983. Harmonic dynamics of proteins: normal modes and fluc-tuations in bovine pancreatic trypsin inhibitor. Proc. Natl. Acad. Sci. USA 80:6571–75

9. Brooks BR, Karplus M. 1985. Normal modes for specific motions of macromolecules:application to the hinge-bending mode of lysozyme. Proc. Natl. Acad. Sci. USA 82:4995–99

42 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

10. Brooks CL III, Karplus M, Pettitt BM. 1988. Proteins: A Theoretical Perspective of Dynamics,Structure, and Thermodynamics. New York: Wiley

11. Brunger AT, Brooks CL III, Karplus M. 1985. Active site dynamics of ribonuclease. Proc.Natl. Acad. Sci. USA 82:8458–62

12. Brunger AT, Clore GM, Gronenborn AM, Karplus M. 1986. Three-dimensional struc-ture of proteins determined by molecular dynamics with interproton distance restraints:application to crambin. Proc. Natl. Acad. Sci. USA 83:3801–5

13. Brunger AT, Huber R, Karplus M. 1987. Trypsinogen-trypsin transition: a moleculardynamics study of induced conformational change in the activation domain. Biochemistry26:5153–62

14. Brunger AT, Karplus M. 1991. Molecular dynamics simulations with experimental re-straints. Acc. Chem. Res. 24:54–61

15. Brunger AT, Kuriyan J, Karplus M. 1987. Crystallographic R factor refinement by molec-ular dynamics. Science 235:458–60

16. Case DA, Karplus M. 1979. Dynamics of ligand binding to heme proteins. J. Mol. Biol.132:343–68

17. Caves TC, Karplus M. 1969. Perturbed Hartree-Fock theory. I. Diagrammatic double-perturbation analysis. J. Chem. Phys. 50:3649–61

18. Colonna-Cesari F, Perahia D, Karplus M, Ecklund H, Branden CI, Tapia O. 1986. In-terdomain motion in liver alcohol dehydrogenase: structural and energetic analysis of thehinge bending mode. J. Biol. Chem. 261:15273–80

19. Conroy H. 1960. Nuclear magnetic resonance in organic structural elucidation. Adv. Org.Chem. Vol. II, p. 265

20. Cusack S, Smith J, Finney J, Karplus M, Trewhella J. 1986. Low frequency dynamics ofproteins studied by neutron time-of-flight spectroscopy. Physica 136B:256–59

20a. Dalton L. 2003. Karplus Equation. Chem. Eng. News 81:37–3921. Deisenhofer J, Steigemann W. 1975. Crystallographic refinement and the structure of

the bovine pancreatic trypsin inhibitor at 1.5 A resolution. Acta Crystallogr. B 31:238–5022. Dobson CM. 2003. Protein folding and misfolding. Nature 426:884–9023. Dobson CM, Karplus M. 1986. Internal motion of proteins: nuclear magnetic resonance

measurements and dynamic simulations. Methods Enzymol. 131:362–8924. Dobson CM, Sali A, Karplus M. 1998. Protein folding: a perspective from theory and

experiment. Angew. Chem. Int. Ed. 37:868–9325. Elber R, Karplus M. 1990. Enhanced sampling in molecular dynamics: use of the time-

dependent Hartree approximation for a simulation of carbon monoxide diffusion throughmyoglobin. J. Am. Chem. Soc. 112:9161–75

26. Farkas A, Farkas L. 1935. Experiments on heavy hydrogen. V. The elementary reactions oflight and heavy hydrogen. The thermal conversion of ortho-deuterium and the interactionof hydrogen and deuterium. Proc. R. Soc. London A 152:124–51

27. Feynman RP, Leighton RB, Sands M. 1963. The Feynman Lectures in Physics. Addison-Wesley

28. Field MJ, Bash PA, Karplus M. 1990. A combined quantum mechanical and molecularmechanical potential for molecular dynamics simulations. J. Comp. Chem. 11:700–33

29. Frauenfelder H, Hartmann H, Karplus M, Kuntz ID Jr, Kuriyan J, et al. 1987. Thermalexpansion of a protein. Biochemistry 26:254–61

30. Frauenfelder H, Petsko GA, Tsernoglou D. 1979. Temperature-dependent x-ray diffrac-tion as a probe of protein structural dynamics. Nature 280:558–63


ANRV275-BB35-01 ARI 10 April 2006 15:34

31. Freeman DL, Karplus M. 1976. Many-body perturbation theory applied to molecules:analysis and calculation correlation energy calculation for Li2, N2, and H3. J. Chem. Phys.64:2461–59

32. Gao J, Kuczera K, Tidor B, Karplus M. 1989. Hidden thermodynamics of mutant pro-teins: a molecular dynamics analysis. Science 244:1069–72

32a. Gao YQ, Yang W, Karplus M. 2005. A structure-based model for synthesis and hydrolysisof ATP by F1ATPase. Cell 123:195–205

33. Gelin BR. 1976. Application of empirical energy functions to conformational problems in bio-chemical systems. PhD thesis. Harvard Univ.

34. Gelin BR, Karplus M. 1975. Sidechain torsional potentials and motion of amino acids inproteins: bovine pancreatic trypsin inhibitor. Proc. Natl. Acad. Sci. USA 72:2002–6

35. Gelin BR, Karplus M. 1977. Mechanism of tertiary structural change in hemoglobin.Proc. Natl. Acad. Sci. USA 74:801–5

36. Gilardi R, Karle IL, Karle J, Sperling W. 1971. Crystal structure of visual chromophores,11-cis and all-trans retinal. Nature 232:187

37. Godfrey M, Karplus M. 1968. Theoretical investigation of reactive collisions in molecularbeams: K+Br2. J. Chem. Phys. 49:3602–9

38. Harvey SC, Prabhakaran M, Mao B, McCammon JA. 1984. Phenylalanine transfer RNA:molecular dynamics simulation. Science 223:1189–91

39. Hemley RJ, Dinur U, Vaida V, Karplus M. 1985. Theoretical study of the ground andexcited singlet states of styrene. J. Am. Chem. Soc. 107:836–44

40. Hirschfelder JA, Eyring H, Topley B. 1936. Reactions involving hydrogen molecules andatoms. J. Chem. Phys. 4:170–77

41. Honig B, Hudson B, Sykes BD, Karplus M. 1971. Ring orientation in β-ionone andretinals. Proc. Natl. Acad. Sci. USA 68:1289–93

42. Honig B, Karplus M. 1971. Implications of torsional potential of retinal isomers for visualexcitation. Nature 229:558–60

43. Honig B, Warshel A, Karplus M. 1975. Theoretical studies of the visual chromophore.Acc. Chem. Res. 8:92–100

44. Hvidt A, Nielsen SO. 1966. Hydrogen exchange in proteins. Adv. Protein Chem. 21:287–86

45. Ichiye T, Karplus M. 1983. Fluorescence depolarization of tryptophan residues in pro-teins: a molecular dynamics study. Biochemistry 22:2884–93

46. Imai K, Osawa E. 1989. An extension of multiparameteric Karplus equation. TetrahedronLett. 30:4251–54

47. Irikura KK, Tidor B, Brooks BR, Karplus M. 1985. Transition from B to Z DNA: contri-bution of internal fluctuations to the configurational entropy difference. Science 229:571–72

48. Islam SA, Karplus M, Weaver DL. 2002. Application of the diffusion-collision model tothe folding of three-helix bundle proteins. J. Mol. Biol. 318:199–215

49. Islam SA, Karplus M, Weaver DL. 2004. The role of sequence and structure in proteinfolding kinetics: the diffusion-collision model applied to proteins L and G. Structure12:1833–45

50. Karplus M. 1952. Bird activity in the continuous daylight of arctic summer. Ecology 33:12951. Karplus M. 1956. Charge distribution in the hydrogen molecule. J. Chem. Phys. 25:605–652. Karplus M. 1959. Contact electron-spin interactions of nuclear magnetic moments. J.

Chem. Phys. 30:11–1553. Karplus M. 1959. Interpretation of the electron-spin resonance spectrum of the methyl

radical. J. Chem. Phys. 30:15–18

44 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

54. Karplus M. 1960. Theory of proton coupling constants in unsaturated molecules. J. Am.Chem. Soc. 82:4431

55. Karplus M. 1960. Weak interactions in molecular quantum mechanics. Rev. Mod. Phys.32:455–60

56. Karplus M. 1963. Vicinal proton coupling in nuclear magnetic resonance. J. Am. Chem.Soc. 85:2870

57. Karplus M. 1968. Structural implications of reaction kinetics. In Structural Chemistryand Molecular Biology: A Volume Dedicated to Linus Pauling by His Students, Colleagues, andFriends, ed. A Rich, N Davidson, pp. 837–47. San Francisco: Freeman

58. Karplus M. 1982. Dynamics of proteins. Ber. Bunsen-Ges. Phys. Chem. 86:386–9559. Karplus M. 1996. Theory of vicinal coupling constants. In Encyclopedia of Nuclear Magnetic

Resonance. Vol. 1: Historical Perspectives, ed. DM Grant, RK Harris, pp. 420–22. New York:Wiley

60. Karplus M. 1997. The Levinthal Paradox: yesterday and today. Fold. Des. 2:569–7661. Karplus M. 2002. Molecular dynamics simulations of biomolecules. Acc. Chem. Res.

35:321–2362. Karplus M, Fraenkel GK. 1961. Theoretical interpretation of carbon-13 hyperfine inter-

actions in electron spin resonance spectra. J. Chem. Phys. 35:1312–2363. Karplus M, Godfrey M. 1966. Quasiclassical trajectory analysis for the reaction of potas-

sium atoms with oriented methyl iodide molecules. J. Am. Chem. Soc. 88:533264. Karplus M, Kuppermann A, Isaacson LM. 1958. Quantum-mechanical calculation of

one-electron properties. I. General formulation. J. Chem. Phys. 29:1240–4665. Karplus M, Kuriyan J. 2005. Molecular dynamics and protein function. Proc. Natl. Acad.

Sci. USA 102:6679–8566. Karplus M, Lawler RG, Fraenkel GK. 1965. Electron spin resonance studies of deuterium

isotope effects. A novel resonance-integral perturbation. J. Am. Chem. Soc. 87:526067. Karplus M, McCammon JA. 2002. Molecular dynamics simulations of biomolecules. Nat.

Struct. Biol. 9:646–5268. Karplus M, Porter RN. 1970. Atoms and Molecules: An Introduction for Students of Physical

Chemistry. Menlo Park, CA: Benjamin Cummins69. Karplus M, Porter RN, Sharma RD. 1965. Exchange reactions with activation energy. I.

Simple barrier potential for (H,H2). J. Chem. Phys. 43:3259–8770. Karplus M, Weaver DL. 1976. Protein-folding dynamics. Nature 260:404–671. Karplus M, Weaver DL. 1994. Folding dynamics: the diffusion-collision model and ex-

perimental data. Protein Sci. 3:650–6872. Karplus R, Kroll NM. 1950. Fourth-order corrections in quantum electrodynamics and

the magnetic moment of the electron. Phys. Rev. 77:536–4973. Karplus S, Karplus M. 1972. Nuclear magnetic resonance determination of the angle ψ

in peptides. Proc. Natl. Acad. Sci. USA 69:3204–674. Deleted in proof75. Kirkwood JG, Oppenheim I. 1961. Chemical Thermodynamics. New York: McGraw Hill76. Kline AD, Braun W, Wuthrich K. 1988. Determination of the complete 3-dimensional

structure of the alpha-amylase inhibitor tendamistat in aqueous-solution by nuclear mag-netic resonance and distance geometry. J. Mol. Biol. 204:675–724

77. Kuppermann A, Karplus M, Isaacson LM. 1959. The quantum-mechanical calculation ofone-electron properties. II. One-and two-center moment integrals. Zeit. Nat. 14a:311–18

78. Kuppermann A, Schatz GC. 1975. Quantum-mechanical reactive scattering: accurate3-dimensional calculation. J. Chem. Phys. 62:2502–4


ANRV275-BB35-01 ARI 10 April 2006 15:34

79. Kuriyan J, Weis WI. 1991. Rigid protein motion as a model for crystallographic temper-ature factors. Proc. Natl. Acad. Sci. USA 88:2773–77

80. Lawler RG, Bolton JR, Karplus M, Fraenkel GK. 1967. Deuterium isotope effects in theelectron spin resonance spectra of naphthalene negative ions. J. Chem. Phys. 47:2149–65

81. Lee A-W, Karplus M, Poyart C, Bursaux E. 1988. Analysis of proton release in oxy-gen binding by hemoglobin: implications for the cooperative mechanism. Biochemistry27:1285–301

82. Levitt M, Lifson S. 1969. Refinement of protein conformations using a macromolecularenergy minimization procedure. J. Mol. Biol. 46:269–79

83. Levy RM, Karplus M, Wolynes PG. 1981. NMR relaxation parameters in molecules withinternal motion: exact Langevin trajectory results compared with simplified relaxationmodels. J. Am. Chem. Soc. 103:5998–6011

84. Lifson S, Warshel A. 1969. Consistent force field for calculations of conformations vi-brational spectra and enthalpies of cycloalkanes and n-alkane molecules. J. Chem. Phys.49:5116–29

85. Linderstrom-Lang K. 1955. Deuterium exchange between peptides and water. Chem. Soc.Spec. Publ. 2, p. 1.

86. Liu B. 1973. Ab-initio potential-energy surface for linear H-3. J. Chem. Phys. 58:1925–3787. Ma J, Sigler PB, Xu Z, Karplus M. 2000. A dynamic model for the allosteric mechanism

of GroEL. J. Mol. Biol. 302:303–1388. McCammon JA, Gelin BR, Karplus M. 1977. Dynamics of folded proteins. Nature

267:585–9089. Mierke DF, Huber T, Kessler H. 1994. Coupling-constants again: experimental restraints

in structure refinement. J. Comp. Aided Mol. Des. 8:29–4090. Miranker A, Karplus M. 1991. Functionality maps of binding sites: a multiple copy si-

multaneous search method. Proteins Struct. Funct. Genet. 11:29–3491. Moffitt W. 1954. Atomic valence states and chemical binding. Rep. Prog. Phys. 17:173–20092. Monod J, Wyman J, Changeux JP. 1965. On nature of allosteric transitions: a plausible

model. J. Mol. Biol. 12:88–11893. Morokuma K, Eu BC, Karplus M. 1969. Collision dynamics and the statistical theories of

chemical reactions. I. Average cross section from transition-state theory. J. Chem. Phys.51:5193–203

94. Morokuma K, Karplus M. 1971. Collision dynamics and the statistical theories of chemicalreactions. II. Comparison of reaction probabilities. J. Chem. Phys. 55:63–75

95. Nadler W, Brunger AT, Schulten K, Karplus M. 1987. Molecular and stochastic dynamicsof proteins. Proc. Natl. Acad. Sci. USA 84:7933–37

96. Nilsson L, Clore GM, Gronenborn AM, Brunger AT, Karplus M. 1986. Structure refine-ment of oligonucleotides by molecular dynamics with nuclear Overhauser effect inter-proton distance restraints: application to 5′ d(C-G-T-A-C-G)2. J. Mol. Biol. 188:455–75

97. Olejniczak ET, Dobson CM, Levy RM, Karplus M. 1984. Motional averaging of protonnuclear Overhauser effects in proteins. Predictions from a molecular dynamics simulationof lysozyme. J. Am. Chem. Soc. 106:1923–30

98. Perutz M. 1971. Stereochemistry of cooperative effects in haemoglobin. Nature 232:408–13

99. Phillips DC. 1981. Closing remarks. In Biomolecular Stereodynamics, ed. RH Sarma, 2:497–48. Guilderland, NY: Adenine

100. Porter RN, Karplus M. 1964. Potential energy surface for H3. J. Chem. Phys. 40:1105–15101. Post CB, Dobson CM. 2005. Meeting review frontiers in computational biophysics: a

symposium in honor of Martin Karplus. Structure 13:949–52

46 Karplus

ANRV275-BB35-01 ARI 10 April 2006 15:34

102. Purins D, Karplus M. 1969. Spin delocalization and vibrational-electronic interaction inthe toluene ion-radicals. J. Chem. Phys. 50:214–33

103. Rahman A. 1964. Correlations in motion of atoms in liquid argon. Phys. Rev. 136:A405–11104. Sali A, Shakhnovich E, Karplus M. 1994. How does a protein fold? Nature 369:248–51105. Schatz GC. 2000. Perspective on “Exchange reactions with activation energy. I. Simple

barrier potential for (H, H2)”: Karplus M, Porter RN, Sharma RD (1965) J. Chem. Phys.43:3259–3287. Theor. Chem. Acc. 103:270–72

106. Schekkerman H, Tulp I, Piersma T, Visser GH. 2003. Mechanisms promoting highergrowth rate in arctic than in temperate shorebirds. Ecophysiology 134:332–42

107. Scheraga HA. 1968. Calculations of the conformations of small molecules. Adv. Phys.Org. Chem. 6:103–84

108. Schulten K, Karplus M. 1972. On the origin of a low-lying forbidden transition inpolyenes and related molecules. Chem. Phys. Lett. 14:305–9

109. Shavitt I, Karplus M. 1962. Multicenter integrals in molecular quantum mechanics. J.Chem. Phys. 36:550–51

110. Shavitt I, Stevens RM, Minn FL, Karplus M. 1968. Potential-energy surface for H3. J.Chem. Phys. 48:2700–13

111. Shulman RG, Glarum SH, Karplus M. 1971. Electronic structure of cyanide complexesof hemes and heme protein. J. Mol. Biol. 57:93–115

112. Simonson T, Archontis G, Karplus M. 2002. Free energy simulations come of age:protein-ligand recognition. Acc. Chem. Res. 35:430–37

113. Smith J, Cusack S, Pezzeca U, Brooks BR, Karplus M. 1986. Inelastic neutron scatteringanalysis of low frequency motion in proteins: a normal mode study of the bovine pancreatictrypsin inhibitor. J. Chem. Phys. 85:3636–54

114. Stillinger FH, Rahman A. 1974. Improved simulation of liquid water by molecular-dynamics. J. Chem. Phys. 60:1545–57

115. Szabo A, Karplus M. 1972. A mathematical model for structure-function relations inhemoglobin. J. Mol. Biol. 72:163–97

116. Szent-Gyorgyi A. 1948. Nature of Life, A Study of Muscle. New York: Academic. 102 pp.117. Taylor EH, Datz S. 1955. Study of chemical reaction mechanisms with molecular beams:

the reaction of K with HBr. J. Chem. Phys. 23:1711–18118. van der Vaart A, Ma J, Karplus M. 2004. The unfolding action of GroEL on a protein

substrate. Biophys. J. 87:562–73119. Wall FT, Porter RN. 1963. Sensitivity of exchange-reaction probabilities to potential-

energy surface. J. Chem. Phys. 39:311120. Warshel A, Karplus M. 1974. Calculation of ππ∗ excited state conformations and vibronic

structure of retinal and related molecules. J. Am. Chem. Soc. 96:5677–89121. Wong CF, McCammon JA. 1986. Dynamics and design of enzymes and inhibitors.

J. Am. Chem. Soc. 108:3830–32122. Yang W, Bitetti-Putzer R, Karplus M. 2004. Free energy simulations: use of reverse

cumulative averaging to determine the equilibrated region and the time required forconvergence. J. Chem. Phys. 120:2618–28

123. Young MA, Gonfloni S, Superti-Furga G, Roux B, Kuriyan J. 2001. Dynamic couplingbetween the SH2 and SH3 domains of c-Src and hck underlies their inactivation byC-terminal tyrosin phosphorylation. Cell 105:115–26


martin karplus

Documents