Document XOwnqOqqyMrw79RYe6mD86VNJ

Science / / 'v' A. Theory BNPiainiiig cancer partly retracted fnabttty to repeat certain experiments, poaitoMty that others were doctored force researchers to wNhdraw some of biochemical cancer theory Jeffrey L. Fox C&EN, Washington A new theory to explain the deranged biochemistry at work in tumor cells--a theory so elegant that it had much of the cancer research commu nity buzzing with excitement early this year--now is being partly re tracted under admittedly painful circumstances. Despite persistent efforts, certain experiments on which the theory is based cannot so far be repeated and, worse, a few of them to contain doctored mate- "The situation is not a good one," says Cornell University biochemist Efraim Hacker, the senior investiga tor involved in the research. Right now everything is "very, very con fusing," be ados. "We are withdraw ing some of our papers. A good deal of it may be wrong ... but it isn't all wrong. I have no choice but to with drawIt until we establish what we can and Vogt's graduate student Robert B. Pepinsky. Two other papers, which appeared during July in Science [213, 303 (1981)] an?in Cell (25,9(1981)], are being retracted by means of let ters to the two publications. Meanwhile, graduate student Spector has been barred at least temporarily from doing research in Racket's lab. "I have asked him to wait before coming back," Racket says. No other official action has been taken yet. Spector, who is called "brilliant" and "gifted" by thoee dose to him, including Racket, was sup posed to receive nis doctoral degree about now, according to Cornell bio chemistry department chairman Richard E. McCarty. But the process was.held up for technical reasons (not all of the necessary documents had been submitted) before any of the current doubts about this collabora tive research effort surfaced. "I can't say what will happen," McCarty says. "I imagine it will take weeks; wheels grind slowly in academe." In February, Racket agreed to discuss the Cornell team's research with CAEN for an article summariz ing recent progress in tumor virology and toward understanding funda mental biochemical mechanisms in cancer cells (CAEN, March 16, page 16). Though much of the Cornell work then was unpublished, news of it was coming rapidly into public view. Both Raeker aria graduate student Spector were giving public seminars. And the still-faster telephone news circuit also was at work. Enthusiasm was induc ing even skeptical scientists to drop hints at press conferences about the Cornell work and its exciting impli cations. Some of those scientists admit to feeling stung now. But most of"them also are quick to defend the healthi ness of cancer research in general. "The field is still doing nicely ... people have been very careful," says one scientist "This work appeared to give an unexpected intellectual unity to the field whether it holds up, who knows? If our thinking has been in fluenced by these claims, we must Elegant protein caecado explaining cancer now doubted Tuner virus Infects eel, era Induce are protein syndwte and activate Unase cascade Osnadc Insutl iaay Indues er tecraasa synthesis ei src-llka protein. acttvaSng tens-- cssoade reproduce." Hacker, some of his immediate colleagues, and other scientists who have collaborated with than are only reluctantly willing to discuss openly the circumstances leading up to the decision to retract eons ofthepapers They cite fear oflegal reprisal as one reason for this reluctance. But many of them also clearly are shaken by the Coil membrane mere possibility that some experi ments! results might have been faked. Nonetheless, a few unequivocal actions are being taken. Two manu scripts, one submitted to the Journal ofBiological Chemistry and the other to have appeared as a Odd Spring Harbor Symposium paper, are being withdrawn, according to Volker M. Vogt, an assistant professor in the Cornell department who was part of a team that included Raeker, Rack- er's graduate student Mark Specter, 8ept 7. ieei G4 URL 05145 Bctonco figure out how much to jettison," he adds. "But some of It has been very stimulating. " The Cornell claims, before the that the phosphorylated amino acid on the $ subunit is tyrosine." Racker also writ--that he has repeated parts of other experiments that involved one is confident about the worth of testing them further. Racker, Vogt, and others agree that they must start "from scratch" to test what will hold current difficulties arose, appeared to the exchange of material between his up. bring a new unity to cancer research lab and that of George Todero, who is "Everybody is confused, disap They tied together much earlier chief of the Uboratory of viral carci pointed, and intensely skeptical," one findings in the scientific literature nogenesis at the National Cancer In scientist not at Cornell says, adding . about abnormal energy metabolism stitute. Their recent experiments that it's important to k--p straight in cancer cells with many of the re confirmthat a low-molecular-weight that much of the recent excitement sults from the past decade or so of activator protein, obtained by Toaa- that the Cornell scheme contributed research on tumor-causing viruses. ro, increases activity of kinase en to is based on very solid research. The The Cornell team said that a series zymes in the cascade, Racker says. Cornell work gave a new focus to the of enzymes in cells act in a cascade in But, he notes: "The-- experiments findings of others. which the enzymes are phoephoryl- will have to be repeated with enzyme "From my point of view, scientifi ated by adenosine triphosphate preparations and an activator of cally everything in the Cornell papers molecules (ATP). An important re known purity." is extraordinarily complex ana diffi sult of that cascade, they also said, Oth-- experiments have not been cult" says another scientist in the was that a sodium-potassium-ion- verified so tar. They include experi field. "Trying to reproduce something pumping, ATP-using protein in cells ments with antibodies, attempting to that complex is extremely difficult becomes phoephoryUted and thus is identify cell kinase proteins (or their Just be cautious, and rive them a rendered inefficient in cancer cells. subunits) -- bring junilar -- identical chance to work it out" He and others The Cornell team also presented ev to turner virus proteins. "1 am also not urge that Racker and bis colleagues idence, much of it based on complex certain of the correctness of some of be given ample time to repeat and to experiments using antibodies as an the physical-chemical properties as sort through all of the pertinent data. alytic tools, that some of the proteins cribed to the protein kin--," Racker However, a few scientists also say in that cascade (all of whose members says. "It will take us many months privately that it is unfortunate that were enzymes that act as kinases) before we know what is correct." Racker ignored early warning were similar or the same as certain "We have been sifting through the along the way. Racker says that be proteins made by tumor viruses. rubble and ash-- over the past five performed some of the early experi "A gene product, either a kinase weeks," says Vogt of the Cornell ments himself and so grew to trust in itself or an activator of s kinase, starts group. "One* you distrust anything, the rapid progress that was being a cascade of phosphorylation which you tend to distrust all ofit--it's very made. Some cntics say that progress inereosee glycolysis fan anerobic form painful for us." was much too fast to be trusted. This of energy metabolism) and other The serious troubles began when criticism, however valid it may be, is s metabolic pathways ana alters com Vogt got involved in doing more of the of course offered with the clarity of r~ ponents of the cytoskrieton, which in experiments with his own hands. hindsight The same critic admits: ^ turn affects structure and membrane He'd been precluded from such in "You can't live in science without r* processes such as ion transport," volvement before the summer, he attributing honesty to your cowork- ^ Hacker and Spector wrote in their says, because of teaching duties and ers. You can be skeptical, but you Science paper. And the case also was the need to spend time preparing can't live looking at everything -- made there that protein kin-- acted grant applications. "We haa had e lot being made up. There is general -- "pleiotypic mediators," meaning of troubto reproducing experiments," agreement that Racker and Vogt are they exerted and controlled several he says, "so I wanted to nave a per innocent victims. activities at once in cancer cells. sonal hand in them." Other scientists in the field also are That same paper in Science began Vogt participated in one experi viewing them--lv-- -- the willing be with a quotation from G. K Ch--- ment involving the separation and lievers, and thus victims of a power terton that now sounds hauntingly analysis of radioactively Ubaled fully attractive intellectual idea that ironic: "There are no nit-- of archi proteins. "Right away I came up with now is being cast into doubt. Ironi tecture for a castle in the clouds." a discrepancy," he says. cally, that insight helped to stimulate Unfortunately, the irony was not in supposedly contained proteins la those scientists tremendously during tentional. Rack-- especially, but also beled with radioactive isotop-- of the past few months, even giving his coUeaguee, now is paying dearly -- phosphor-- and sulfur. But instead of some of them ideas for new projects. he combs through the ruins of that phoephorue-32, Vogt found radioac "That's the Catch 22 to this whole castle trying to see what remains in tive iodine in the samples. "When I business," says one of those scientists. tact started to analyze, I found the wrong "Thoae projects don't rest on these "I must take the scientific respon thing there," he says. [now believed to be faulty] data, but sibility," Racker tells CAEN. Thus Other rimil-- evidence, obtained in on the biochemical and intellectual far, during an intense effort that other laha, also giv-- hints that in approach. And it has proved to be began late m July, a few ofthe results correct isotop--were used in various rewarding." that the Cornell group claimed earlier experiments. But everyone involved For the time being, the coherent have been confirmed, he says. is ex--rising extreme caution before framework pieced together by the "I have confirmed the phos making any public accusations of in Cornell scientists must be set aside. phorylation of the 0 subunit of the tentional wrongdoing. Indeed, there One seemingly promising scientific sodium-potassium ATPaee by pro is a great deal of confusion about care-- probably will be abruptly V tein kinase from Ehrlich asdt-- various mat-isle. Nonetheless, be truncated. Rut the scientific enter tumor cells prepared by Mark Spec- cause at least some sampl-- are un prise, despite its ability to attract tor," Racker writ-- m a letter sub equivocally mislabeled, and possibly adversity,also is amaringly able to get mitted to Science. "I have established deliberately doctored, virtually no back on course. a as ca 8--l 7, issi t i - -if rUtWWV' iV*-' ,' ! i \ i m. ^ Science explains Eschenmoser, that despite their apparent complexity, they readily form by themselves, provided the necessary precursor molecules are available. Eschenmoser and his coworkers also are investigating other corrinoid structural features that may have "autogeneous character." All natural porphyrinoids, as well as vitamin B^, are known to be biosynthesized from one particular isomer of a tetrapyrrolic macrocycle called uroporphyrinogen. Living or ganisms such as bacteria can assem ble this uroporphyrinogen isomer from a simple monopyrrolic precur sor. Moreover, Mauzerall and others have shown that this same precursor can be converted easily to a mixture of four uroporphyrinogen isomers under simple chemical conditions-- conditions that might have existed on primeval Earth. Of these four iso mers, the one that forms in the greatest amount turns out to be the one of biological importance. This coincidence may not be accidental, according to Eschenmoser and other scientists. It is possible, he suggests, that the statistically preferred iso mer--the one formed in largest amount in an abiotic environment-- may have been "selected to become biosynthesized" after organisms found uroporphyrinogens to be bio logically "useful and important for survival." These rudimentary or ganisms may simply have chosen the uroporphyrinogen isomer that was most abundant, and incorporated it into their biochemistry. Studies of cancer risk in chemists assessed 1981 ACS ANNUAL MEETING Various epidemiological studies over the past dozen years have been probing the connection between the occupation of chemist and the risk of cancer. Now studies are being made and existing ones are being extended and expanded. But as yet very little in the way of a conclusion can be drawn from them. Shelia K. Hoar of the environ mental epidemiology branch of the National Cancer Institute sums up the current situation: "We can con- Swedish mortality study finds chemists show excess of leukemias/fymphomas and brain tumors r cnn e Mh Oini>U Swclaa oaaamd imnfii Otnwil tiaioidd JUleauees - AM cancer PlgdUy Pancreatic 63 103.6 49 74.9 59 87 8 32 24.2 17 17.5 11 20.5 a 9.2 4 9.7 3 7.9 2 1.6 0 1.1 2 1.4 Si^lraBty 3 4.0 4 3.6 5 34 Praetebc 3 15 0 1.1 1 1.4 Urinary 3 1.8 2 1.2 0 1.5 Brain 5 1.2 2 0.9 1 1.0 LymphaHc/hematopoietic 7 3.2 2 2.3 0 2.7 Hodgkin's Circulatory 3 0.7 0 05 0 0.5 22 40.4 10 29.5 20 35.1 AeaMems 15 21.9 16 15.5 15 17.6 Other ceueee 14 17.3 6 12.4 13 14.6 elude that there are some excess risks associated with the occupation of chemist. At this point, we cannot differentiate between effects of chemical exposures and socioeco nomic or other factors related to being a chemist." Hoar discussed the studies for the Division of Chemical Health & Safe ty. Included was a recent study of her own, conducted with Sidney Pell of the epidemiology section in the medical division at Du Pont, Wil mington. That effort was a retro spective cohort study of mortality and cancer incidence *nnng men and women employed at Du Pont as chemists in 1959 (C&EN, Aug. 17, page 34). All told there have been eight studies, including the Du Pont one. The other seven: Death certificates of some 3600 members of the American Chemical Society whose deaths were reported in C&EN from 1948 to 1967 were ex amined in 1969 by epidemiologists from the National Cancer Institute. A study of occupational mortal ity in the state of Washington from 1950 to 1971 included chemists. There was a similar study of oc cupational mortality in the state of California from 1959 to 1961. In Sweden, a retrospective co hort study was made of mortality among graduates during three dec, ades from two Swedish schools of chemical engineering. Registrar General's occupational mortality study in England and Wales in 1970 to 1972 included chemists. Also from Great Britain, pre liminary reports were made of a pro portional mortality study of 938 de- ceased members of the Royal Ins' tute of Chemistry. A study of mortality by occupation level and cause of death among men 20 to 64 years of .age in the U.S. in 1950 included chemists. Together, the studies suggest that excess cancer risk is associated with the occupation of chemist. But, Hoar notes, all eight studies have their strengths ana weaknesses. For example, professional chemist societies include both persons heavily exposed to a wide variety of chemicals and persons with little or no chemical exposure. Occupations given on death certificates are often inaccurate. Several of the studies are based on proportional mortality rates, which can appear elevated as a result of deficits in other causes of death. Also, some studies are based on small numbers. Thus, Hoar says, little can be con cluded from any study alone. Still, she says, when all eight studies are con sidered, patterns in the morbidity and mortality data become apparent. The studies show that chemists generally have lower mortality overall and lower mortality due to lung can cer. Half the studies show excess mortality for all cancers combined, half do not. Lymphatic and hematopoietic cancers (leukemia), colon cancer, cerebrovascular disease, and suicide occur more frequently than expected in most of the studies. The studies. Hoar notes, have very sparse data for bladder cancer, a site often associated with chemicals. The two reports that have data for the disease show that bladder cancer occurs more fre quently among chemists. Hoar points out that better expo- Sept. 14. 1981 CAEN 41 URL 05147 URL 05148 sure histories are needed to determine the factors responsible for the asso ciations seen in the chemist studies. When more time has elapsed, she suggests, and additional subjects have developed illnesses and died, the Du Pont data could be reanalyzed by department and process. Efforts should be made, she says, to charac terize occupational exposures beyond job title and department affiliation through industrial hygiene data. One effort that continues is the Swedish study, first published in 1976 (Lancet, 7991,916 (1976)1. The study is being carried out by G. Robert Olin, director of the occupational health and safety department at the phar maceutical firm Astra Inc., Sodertalje, Sweden, and a professor in the preventive occupational medicine unit in the department of heat tech nology at the Royal Institute of Technology, Stockholm. Olin told the Division of Chemical Health & Safety that since the origi nal study appeared it has been ex panded in several ways. The original cohort, he says, has been widenea and the followup period lengthened. The original study was of chemistry graduates for three decades--1930's, 1940's, and 1950's--from the Royal Institute of Technology. Olin now has included two other academic cohorts for comparison. The first consists of architects, a group of academics who have been educated in the same city, the same university, during the same period of time, and who can generally be con sidered to belong to the same social strata as the chemists. The only ap parent difference between the two groups, Olin explains, is the chemical exposure that occurred among ail chemists, at least during their four or five years of education. The second additional cohort con sists of male mining engineers and metallurgists, who also were educated at the Royal Institute of Tedmology. The idea behind including this group is that it holds an intermediate posi tion in relation to chemical exposure, at least during the education years. The total number of chemists in volved from the three decades is 820. To this now have been added data for 581 mining engineers/metallurgists and 657 architects. Besides adding the two groups to the study, Olin has extended and an alyzed data from a followup to the study until the end of 1977. OUn points out the "very positive results" occurring among the cohorts. For all three, the total mortality is low relative to the general population and there is no significant difference be metallurgists, there is no single cancer tween them. The main explanation, type that can account for the increase he says, seems to be a lowered rate of in cancer cases, relative to those of the deaths due to circulatory diseases--a architects. However, among the particularly interesting finding, he chemists two types of malignancies notes, since it has been believed until account for most of the relative in recently that the higher social classes crease of cancer deaths: leukemias/ in industrial societies were more lymphomas and brain tumors. prone to circulatory diseases. Olin says that preliminary data for As for cancer mortality, it accounts 1978 and 1979 are now available. for a fifth of the deaths among the During these two years, 13 more architects--identical to the propor chemists have died, four from cancer. tion found in the general population. One of these cancer cases was glioma, For chemists, however, the cancer bringing the total among the cohorts rate is higher in comparison to the to six. And one death was due to a teneral population and significantly leukemia, bringing the total to igher in comparison to the archi eight. tects. Mining engineers fall between The followup of the three cohorts the two groups and their cancer rate is continuing, Olin says. During 1977 is significantly elevated when com and 1978, the surviving chemists an pared to the architects. swered retrospective questions con Olin notes that there is no increase cerning life habits and chemical ex of lung cancers in any of the three posure. The participation rate was an cohorts, thus diminishing the possi acceptable 85%, Olin says, so there is bility that smoking could be the the possibility for a better evaluation major explanation for the differences in the future of the relationship be among them. tween life style, or chemical exposure, Among the mining engineers and and various causes of death. ^ Lanthanide, actinide oraanometallics studied 1981 ACS ANNUAL MEETING The organometallic chemistry of the lanthanide and actinide group ele ments is under study in many differ ent laboratories. New information about how these elements interact with organic molecules and the po tential chemical applications for these interactions was presented in several different sessions of the Inorganic Chemistry Division. Roger E. Cramer, chemistiy pro fessor at the University of Hawaii, Manoa, for example, described the organouranium complexes he and colleagues have prepared. These re searchers have been studying urani um phosphoylide complexes for sev eral yean, and now have prepared one of these complexes, (ip-CsHa)- UCHPlCHsMCiHs), with a carbonuranium bond so short (2.29 A) that it should be considered a multiple bond. Complexes with multiple car bon-metal bonds, called alkylidenes, are known for many transition met als, though they are not common, Cramer points out. Their compound is the tint alkylidene of an actinide element to be observed. The chemists have begun to in vestigate how similar the chemistry of their uranium alkylidene complex is to that of better-studied transition metal alkylidenes. They rind, not unexpectedly, that the uranium-car bon bond is a very reactive one. For instance, carbon monoxide can insert into this bond under mild reaction conditions. (In toluene solution, the reaction can occur at 25 C and 1 atm Eressure of carbon monoxide.) Simiir insertion reactions have been seen for alkylidenes of zirconium and tantalum, Cramer says. The Hawaii chemists have other evidence of the reactivity of the car bon-uranium bond in their com pound: It reacts with methyl iodide to produce an organouranium iodide salt and a phosphonium salt, and it reacts with other halogens, including halo- carbons. The other organouranium complexes that the group has inves tigated--those that contain carbonuranium single bonds--do not react with halocarbons. The uranium al kylidene complex also reacts with carbon dioxide, stannous chloride, and several other mild reagents. Another group studying the organometallic chemical reactions of organoactinide complexes is that of Tobin J. Marks, Robert L. Burwell Jr., and associates at Northwestern University. This group has been in vestigating the effect of supporting 42 C&EN Sept. 14, ieei