part iii institutions and society

Part III

Institutions and Society

Chapter 10. The Question of Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

Chapter 11. Regulation of Genetic Engineering. . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . .211

Chapter 12. Patenting Living Organisms. . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . .237

Chapter 13. Genetics and Society . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .257

Chapter 10

The Question of Risk

Chapter 10

PageIntroduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . 197The Initial Fear of Harm . . . . . . . . . . . . . . . . . . . 197

Classification of Potential Harm . . . . . . . . . . . . 198Identification of Possible Harm . . . . . . ., . . . . zooEstimates of Harm: Risk. . ................200

The Status of the Current Assessment ofPhysical Risk . . . . . . . . . . . . . . . . . . . . . . . 201

Perception of Risk . . . . . . . . . . . . . . . . . . . . . .203Burden of Proof . . . . . . . . . . . . . . . . . . . . .. 203

Other Concerns . . . . . . . . . . . . . . . . .........204Concerns Raised by Industrial Applications .. .204Concerns Raised by the Implications of the

Recombinant DNA Controversy for GeneralMicrobiology. . . . . . . . . . . . . . . . . . . . . .. ..204

Concerns Raised by the Implications of theRecombinant DNA for Other GeneticManipulation . . . . . . . . . . . . . . . . . . . . .. ..206

Page

Ethical and Moral Concerns. , . . . . . . . . . . . . .207Conclusion . . . . . . . . . . . . . . . . . . . . .........207

Figures

Figure No. Page

35. Flow Chart of Possible Consequences of UsingGenetically Engineered Micro-Organisms , . . 199

36. Flow Chart to Establish Probability of HarmCaused by the Escape of a Micro-OrganismCarrying Recombinant DNA . .............201

37. Alternative Methods for Transferring DNA FromOne Cell to Another. . . . . . . . . . . . . . . . . . . . 206

Chapter 10

The Question of Risk

IntroductionThe perception that the genetic manipulation

of micro-organisms might give rise to unfore-seen risks is not new. The originators of chem-ical mutagenesis in the 1940’s were warned thatharmful uncontrolled mutations might be in-duced by their techniques. In a letter to theRecombinant DNA Advisory Committee (RAC) ofthe National Institutes of Health (NIH) in Decem-ber of 1979, a pioneer in genetic transformationat the Rockefeller University, wrote: “. . . I didin 1950, after some deliberation, perform thefirst drug resistance DNA transformations, andin 1964 and 1965 took part in early warningsagainst indiscriminate ‘transformations’ thatwere then being imagined. ”1

Yet none of this earlier public concern led toas great a controversy as has research with re-combinant DNA (rDNA). No doubt it was en-couraged because scientists themselves raisedquestions of potential hazard. The subsequentopen debates among the scientists strengthenedthe public’s perception that there was legitimatecause for concern, This has led to a continuingattempt to define the potential hazards and thechances that they might occur.

‘ Rollin 1). Ho[chkiss, Recombinant DNA Research, vol. 5, NIH pub-lication No. 80-2130, March 1980, p. 484.

The initial fear of harmFor the purposes of this discussion, harm (or

injury) is defined as any undesirable conse-quence of an act. Such a broad definition is war-ranted by the broad targets for hypotheticalharm that genetic manipulation presents: injuryto an individual’s health, to animals, to the en-vironment.

The inital concern involved injury to humanhealth. Specifically, it was feared that combin-ing the DNA of simian virus 40, or SV40, with anEscherichia coli plasmid would establish a newroute for the dissemination of the virus. Al-though the SV40 is harmless to the monkeysfrom which it is obtained, it can cause cancerwhen injected into mice and hamsters. Andwhile it has not been shown to cause cancer inhumans, it does cause human cells to behavelike cancer cells when they are grown in tissueculture. What effect such viruses might have ifthey were inserted into E. coli, a normal in-habitant of the human intestine, was unknown.This uncertainty, combined with an intuitive

judgment, led to a concern that somethingmight go wrong. The dangerous scenario wentas follows:

● SV40 causes cells in tissue culture to be-have like cancer cells,

● SV40-carrying E. coli might be injected ac-cidently into humans,

● humans would be exposed to SV40 in theirintestines, and

● an epidemic of cancer would result.

This chain of connections, while loose, wasstrong enough to raise questions in at least somepeople’s minds.

The virus SV40 has never actually beenshown to cause cancer in humans; but the po-tential hazards led the Committee on Recombi-nant DNA Molecules of the National Academy ofSciences (NAS) to call in 1974 for a deferment ofany experiments that attempted to join the DNAof a cancer-causing or other animal virus to vec-tor DNA. At the same time, other experiments,

197

76-565 0 - 81 - 14

198 ● Impacts of Applied Genetics— Micro-Organisms, Plants, and Animals

that were thought to have a potential for harm–particularly those that were designed totransfer genes for potent toxins or for resist-ance to antibiotics into bacteria of a differentspecies—were also deferred. Finally, one othertype of experiment, in which genes from higherorganisms might have been combined with vec-tors, was to be postponed. The fear was that la-tent “cancer-causing genes” might be inadver-tently passed on to E. coli.

Throughout the moratorium, one point wascertain: no evidence existed to show that harmwould come from these experiments. But it wasa possibility, The scientists who originally raisedquestions wrote in 1975: “. . . few, if any, believethat this methodology is free from risk. ”2 It wasrecognized at that time that “. . . estimating therisks will be difficult and intuitive at first butthis will improve as we acquire additionalknowledge. ”3 Hence two principles were to befollowed: containment of the micro-organisms(see table 35, p. 213) was to be an essential partof any experiment; and the level of containmentwas to match the estimated risk. These prin-ciples were incorporated into the Guidelines forResearch Involving Recombinant DNA Mole-cules, promulgated by NIH in 1976.

But the original fears surrounding rDNA re-search progressed beyond concern that humansmight be harmed. Ecological harm to plants, ani-mals, and the inanimate world were also consid-ered. And other critics noted the possibility ofmoral and ethical harm, which might disruptboth society’s structure and its system of values.

Classification of potential “physical harm

Some combinations of DNA may be harmfulto man or his environment—e.g., if an entireDNA copy of the poliovirus genetic material iscombined with E. coli plasmid DNA, few wouldargue against the need for careful handling ofthis material.

For practical purposes, the potential harmassociated with various micro-organisms is

Viecombimm[ DNA Research, vol. 1, DHEW publication No. (NIH)76-1138, August 1976, p. 59.

31bid.

shown in figure 35. Each letter (A through L)represents the consequence of a particular com-bination of events and micro-organisms. For ex-ample, the letters:

A)C

B,D

E, I

F)J

H,L

G)K

represent the intentional release of micro-organisms known to be harmful to theenvironment or to man—e.g., in biologi-cal warfare or terrorism.represent the inadvertent release ofmicro-organisms known to be harmful tothe environment or to man—e.g., in acci-dents at high-containment facilitieswhere work is being carried out withdangerous micro-organisms.represent the intentional release of micro-organisms thought to be safe but whichprove harmful—when the safety of orga-nisms has been misjudged.represent the intentional release of micro-organisms which prove safe as expected—e.g., in oil recovery, mining, agriculture,and pollution control.represent the inadvertent release ofmicro-organisms which have no harmfulconsequences— e.g., in ordinary accidentswith harmless micro-organisms.represent the inadvertent release ofmicro-organisms thought to be safe butwhich prove harmful—the most unlikelypossible consequence, because both anaccident must occur and a misjudgmentabout the safety must have been made.

Discussions of physical harm have recognizedthe possibility of intentional misuse but haveminimized its likelihood. The Convention on theProhibition of the Development, Production,and Stockpiling of Bacteriological (Biological)and Toxin Weapons and on their Destructionwhich was ratified by both the Senate and thePresident in 1975, * states that the signatorieswill “never develop . . . biological agents or tox-ins . . . that have no justification for prophylac-tic, protective, or other peaceful purposes. ”Such a provision clearly includes micro-orga-nisms carrying rDNA molecules or the toxins

4[;onvention of the Prohibition of the Development, Production,and Stockpiling of Bacleriolo~ical (Biological) and ‘1’oxin Weaponsand on Their Destruction, Washin@on, Umdon, and Moscow,Apr. 10, 1972; entered into force on Mar. 26, 1975 (26 [J. S.”r., 583).

*As of 1980, 80 countries hate ratified the treaty; another 40have signed but not ratified.

Ch. 10—The Question of Risk . 199

Figure 35.—Flow Chart of Possible Consequences of Using Genetically Engineered Micro-Organisms

SOURCE: Office of Technology Assessment.

produced by them. It must be assumed thatthose who signed did so in good faith.

While there is no way to judge the likelihoodof developments in this area, the problems thatwould accompany any attempt to use pathogen-ic micro-organisms in warfare—difficulties incontrolling spread, protection of one’s owntroops and population—tend to discourage theuse of genetic engineering for this purpose. *Similarly, the danger that these techniquesmight be used by terrorists is lessened by thescientific sophistication needed to construct amore virulent organism than those that can

* A 11 hou~h storkpil in~ of” biological ma rf’are agents is prohibited,l’(’S(’ill’(’tl into 11(W’ il#lllS is Ilot.

1

already be obtained—e.g., encephalitis virusesor toxin-producing bacteria like C. botulinum orC. tetani.

Some discussions have centered around thepossibility of accidents caused by a break in con-tainment. Construction of potentially harmfulmicro-organisms will probably continue to beprohibited by the Guidelines; exceptions will bemade only under the most extraordinary cir-cumstances. To date, no organism known to bemore harmful than the organism serving as thesource of DNA has been constructed.

However, the biggest controversy has cen-tered around unforeseen harm—that micro-organisms thought safe might prove harmful.

200 . Impacts of Applied Genetics—Micro-Organisms, Plants, and Animals

Discussion of this kind of harm is hindered bythe difficulty not only of quantifying the prob-ability of an occurrence but also of predictingthe type of damage that might occur. The differ-ent types of damage that can be conjured up arelimited only by imagination. The scenarios haveincluded epidemics of cancer, the spread of oil-eating bacteria, the uncontrolled proliferationof new plant life; and infection with hormone-producing bacteria.

The risk of harm refers to the chance of harmactually occurring. In the present controversy,it has been difficult to distinguish the possiblefrom the probable. It is, for instance, possiblethat an individual will be killed by a meteor fall-ing to the ground, but it is not probable. Analog-ous situations exist in genetic engineering. It isin this analysis that debate over genetic engi-neering has some special elements: the uncer-tainty of what kind of harm could occur, the un-certainty about the magnitude of risk, and theproblem of the perception of risk.

Identification of possible harm

The first step in estimating risk is identifyingthe potential harm. It is not very meaningful toask: How much risk does rDNA pose? The con-cept of risk takes on meaning only when harm isidentified. The question should be: What is thelikelihood that rDNA will cause a specific dis-ease such as in a single individual or in an entirepopulation? The magnitude of the possible harmis incorporated in the question of risk, but dif-fers in the two cases. A statement about the riskof death to one person is different than oneabout the risk of death to a thousand. The rightquestions must be asked about a specific harm.

Since no dangerous accidents are known tohave occurred, their types remain conjectural.Identifying potential harm rests on intuition andarguments based on analogy. Even a so-calledrisk experiment is an approximation of subse-quent genetic manipulations. That is why ex-perts disagree. No incontestable “scientificmethod” dictates which analogy is useful or ac-ceptable. By their very nature, all analogiesshare some characteristics with the event underconsideration but differ in others. The goal is to

discover the one that is most similar and toobserve it often. This process then forms thebasis for extrapolation.

For example, it has been argued that ecologi-cal damage can be caused by the introduction ofplants, animals, and micro-organisms into newenvironments. Scores of examples from historysupport this conclusion. The introduction to theUnited States of the Brazilian water hyacinth inthe late 19th century has led to an infestation ofthe Southern waterways. Uncontrolled spreadof English sparrows originally imported to con-trol insects has made eradication programs nec-essary. Countless other examples are confirma-tion that biological organisms may, at times,cause ecological damage when introduced into anew environment. Yet there is no agreement onwhether such analogies are particularly rele-vant to assessing potential dangers from genet-ically engineered organisms. It could be ar-gued-e.g., that a genetically engineered orga-nism (carrying less than 1 percent new genes) isstill over 99 percent the same as the original,and is therefore not analogous to the “totallynew” organism introduced into an ecosystem.Some experts emphasize the differences be-tween the situations; others emphasize the simi-larities.

Other analogies have been raised. Newstrains of influenza virus arise regularly. Somecan cause epidemics because the population,never before exposed to them, carries no pro-tective antibodies. Yet can this analogy suggestthat relatively harmless strains of E. coli mightbe transformed into epidemic pathogens? Thereis disagreement, and debates continue aboutwhat “could happen” or what is even logicallypossible.

Estimates of harm: risk

Assuming that agreement has been reachedon the possibility of a specific harm, what can bedone to ascertain the probability? What is thelikelihood that damage will occur?

Damage invariably occurs as the result of aseries of events, each of which has its own par-ticular chance of occurring. Flow charts havebeen prepared to identify these steps. A typical


analysis determines a probability value for eachstep-e. g., in figure 36 step II the probability ofescape can be estimated based on the historicalrecord of experiments with micro- organisms.Depending on the degree of containment, theprobability varies. It is almost certain thatexperiments on an open bench top, using noprecautions, will result in some escape to thesurrounding environment—a much less likelyevent in maximum containment facilities. (Seetable 35.)

Two points should be noted. First, each prob-ability can be minimized by appropriate controlmeasures. Second, the probability that the finalevent will occur is equal to or less likely than theleast likely link in the chain. Because the prob-abilities must be multiplied together, if theprobability of any single step is zero, the prob-ability of the final outcome is zero; the chain ofevents is broken.

THE STATUS OF THE CURRENT ASSESSMENTOF PHYSICAL RISK

A successful risk assessment should provideinformation about the likelihood and magnitudeof damage that might occur under given cir-cumstances. It is clear that the more types ofdamage that are identified, the more risk assess-ments must be carried out.

Figure 36.— Flow Chart to Establish Probability ofHarm Caused by the Escape of a Micro-Organism

Carrying Recombinant DNA

Event Probability

L Inadvertent incorporation of hazardous gene P I

into micro-organism

11. Escape of micro-organism into environment P,

Ill. Multiplication of micro-organism and P3establishment in ecological niche

V. Production of factor to cause disease

P*

NOTE: P, will always be smaller than any of the other probabilities.


Although the original charter of RAC under-scored the importance of a risk assessment pro-gram, it was not until 1979 that the details of aformal program were published. For 5 years,risks were assessed on a case-by-case basisthrough: 1) experiments carried out under con-tract from NIH, 2) experiments that were de-signed for other purposes but which proved tobe relevant to the question of risk, and 3) con-ferences at which findings were examined.

From the start, it was difficult to design ex-periments that could supply meaningful infor-mation-e.g., how does one test the possibilitythat “massive ecological disruptions mightoccur?” Or that a new bacterium with harmfulunforseen characteristics will emerge? Stillsome experiments were proposed. But becausethese experiments had to be approximations ofthe actual situation, the applicability of theirfindings was debated. Here too, experts couldand did disagree—not about the findings them-selves, but about their interpretation.

For example, in an important experiment de-signed to test a “worst case situation, ” a tumorvirus called polyoma was found to cause notumors in test animals when incorporated intoE. coli.5* Since just a few molecules of the viralDNA are known to cause tumors when injecteddirectly into animals, it was concluded thattumor viruses are noninfectious to animalswhen incorporated into E. coli. If polyoma virus,which is the most infective tumor virus knownfor hamsters, cannot cause tumors in the rDNAstate in E. coli, it is unlikely that other tumorviruses will do so. This conclusion has had wide-spread, but not unanimous, acceptance. It hasbeen argued that there might be “somethingspecial” about polyoma that prevents it fromcausing tumors in this altered state; othertumor viruses might still be able to do so. At onemeeting of RAC, in fact, it was suggested thatexperiments with several other viruses be car-ried out to confirm the generality of the finding.But how many more viruses? What is enough?

‘M. A. Israel, H. W. Chan, W. P. Rowe, and M. A. Martin, “Molec-ular Cloning of Polyoma Virus DNA in Eacherichki Cofi: PlasmidVector Systems,” Science 203:883-887, 1979.

*Some combinations of free plasmid and tumor virus DNA didcause infections.

202 “ Impacts of Applied Genetics —Micro-Organisms, Plants, and Animals

For some, one carefully planned experimentusing the most sensitive tests is sufficient toallay fears. But for others, significant doubtabout safety remains, regardless of how manyviruses are examined. The criteria depend onan individual’s perception of risk.

Many experiments carried out for purposesother than risk assessment have providedevidence that scenarios of doom or catastropheare highly unlikely. This is the general consen-sus of specialists, not only in molecular biology,but in population genetics, microbiology, infec-tious diseases, epidemiology, and public health.

Experiments have revealed that the structureof genes from higher organisms (plants andanimals) differ from those of bacteria. Con-sequently, those genes are unlikely to be ex-pressed accidentally by a bacterium; the originalfears of “shotgun” experiments have becomeless well-founded. Hence, data gathered to datehave made the accidental construction of a newepidemic strain more unlikely.

Conference discussions have also contributedto a better understanding of the risks. At onesuch conference,6 which was attended by 45 ex-perts in infectious diseases and microbiology, itwas concluded that:

●

●

●

A

E. coli K-12 (the weakened form of E. coli,used in experiments) does not flourish inthe intestinal tract of man;the type of plasmid permitted by the Guide-lines has not been shown to spread from E.coli K-12 to other E. coli in the gut; andE. coli K-12 cannot be converted to a harm-ful strain even after known virulence fac-tors were transferred to it using standardgenetic techniques.

workshop sponsored by NIH7 provided aforum for scientists to discuss the risks posed byviruses in rDNA experiments. They concludedthat the risks were probably fess when a viruswas placed inside a bacterium in rDNA form

‘W’orkshop on Studies for Assessment of Potentiul Risks Associ-iit(?d With Recombinant DNA Kxperimenlal ion, ” k’iiln]outh, Miiss.,June W-21 , 1977.

7’(Workshop to Assess Risks for Recomhinilnt DNA ExperimentsItlvoh’ing \’il’ill (k+nornes,” C(MpOIISOIWl h~ Iht? Niltiolliil Inst i tutes

(d’ li[~illth ai~(i t h e Iiuropeiin Moleculi I I . tliolo~v ol’~illliZillioll,”

AS(X)I , h;t~~liind, Jiin. 26-2&J, 1978.

than when it existed freely. * Experts in infec-tious disease have stressed repeatedly that theability of a micro-organism to cause diseasedepends on a host of factors, all working togeth-er. Inserting a piece of DNA into a bacterium isunlikely to suddenly transform the organisminto a virulent epidemic strain.

Careful calculations can also allay fears aboutthe damage a genetically engineered micro-or-ganism might cause. Doomsday scenarios ofescaped E. coli that carry insulin or otherhormone-producing genes were recently exam-ined in another workshop.8 Prior to this work-shop, newspaper accounts raised the possibilitythat an E. coli carrying the gene for human in-sulin production might colonize humans andthus upset the hormonal balance of the body.

The participants calculated how much insulincould be produced. First, it was assumed that aseries of highly unlikely events would occur—accidental release, ingestion by humans, stablecolonization of the intestine by E. coli K-12. E.coli constitutes approximately 1 percent of theintestinal bacterial population, and it wasassumed that all the normal E. coli would bereplaced by the insulin-producing E. coli. Insulinis made in the form of a precursor molecule,proinsulin. It was assumed that 30 percent of allbacterial protein production would be devotedto this single protein, another highly unlikelysituation. If so, 30 micrograms (ug)-or 0.6units—would then be made in the intestine.Although proteins are very poorly absorbedfrom the intestinal cavity, it was assumed forthe sake of argument that 100 percent of theproinsulin would be absorbed into the circula-tion. Thus, 0.6 units of insulin would be addedto the normal daily human production of 25 to30 units—an imperceptible difference.

Calculations like these have been carried outfor several other hormones. Even with the mostimplausible series of events, leading to thegreatest opportunity for hormone production,

“(hi the other hiil)d, it hils heen iil’gue[i thiit this has providedlriruses with ii new IXILIte for IIissell]illilti(]!l. Netw?rl ht?less, thel’e is

110 fn’idence thiit viruses Cilll readily escilpe tll)[ll the t)ilcleriil illld

Sul)sequellll.v CilUS(? intection.“<Niilk)tliil lnslilute of Allergy iind infectious l)iseiises Workshop

011 Re[:onlhil~i]tlt DNA Risk AssessIIN?Ilt ,“ PilSildellill (Iillif., API..

11-12, 1980.


the conclusion is that normal hormone levelswould change by less than 10 percent. Similarconditions for interferon production couldrelease approximately 70ug or the maximumdaily dose currently used in cancer therapy.Long-term effects of such exposure are current-ly unknown; therefore, experiments using high-producing strains (106 molecules per cell ormore) are likely to be monitored if such strainsever become available.

The NIH program of risk assessment, whichwas formally started in 1979, continues to iden-tify possible consequences of rDNA research.Under the aegis of the National Institute ofAllergy and Infectious Diseases, the programsupports research studies designed to elucidatethe likelihood of harm. * In addition, it collatesgeneral data from other experiments that mightbe relevant to risk assessment. Other risk as-sessments are being conducted by Europeanorganizations * * and by the U.S. EnvironmentalProtection Agency to assess the consequences ofreleasing micro-organisms into the environ-ment.

Thus far, there is no compelling evidence thatE. coli K-12 bacteria carrying rDNA will be morehazardous than any of the micro-organismswhich served as the source of DNA. Never-theless, all the experiments have dealt with onegenus of bacterium. Unless the conclusionsabout E. coli can be extended to other organismslikely to be used in experiments (such as Bacillussubtilis and yeast), other assessments maybe ap-propriate.

“Extramural efforts were first conceived in the summer of 1975to develop and test safer host-vector systems based on E. coli, theinteragency agreement entered into with the Naval BiosciencesLaboratory tested E. coli systems in a series of simulatedaccidental spills in the laboratory. At the University of Michiganthe survival of these systems was tested in mice and in culturalconditions simulating the mouse gastrointestinal tract. TuftsUniversity tested these systems in both mice and humanvolunteers. Finally, the survival of host-vector systems in sewagetreatment plants was tested at the University of Texas. The peakyear for costs of supporting research contracts was 1978; over ahalf-million dollars were required. currently, the cost ofmaintaining the high containment facility at Frederick, Md., isbetween $200,000 and $250,000 annually.

● ● First Report to the Committee on Genetic Experimentation , ascientific committee of the International Council of ScientificUnions, from the Working Group on Risk Assessment, July 1978.

Perception of risk

The probability of damage can be estimatedfor various events. The entire insurance in-dustry is based on the fact that unfavorableevents occur on a regular basis. The number ofpeople dying annually from cancer, or automo-bile accidents, or homicides can be predictedfairly accurately. These estimates depend onthe availability of data and the assumptions thatthe major determinants do not change fromyear to year.

But even if the probability of damage is fairlywell known, a gap often exists between this“real” probability of occurrence and the “per-ceived” probability. Two factors that tend to af-fect perceptions are the magnitude of the possi-ble damage and the lack of individual controlover exposure to the risk. Both of these are sig-nificant factors in the fears associated withrDNA and the manipulation of genes. Becauseintuitive evaluations can contradict analyticalevaluations, the question of risk cannot be re-solved strictly on an analytical basis. Its resolu-tion will have to come through the politicalprocess.

BURDEN OF PROOFThe possibility of inadvertently creating a

dangerous organism does exist, but its prob-ability is lower than was originally thought.Nevertheless, an important principle emergesfrom the debate. Society must decide whetherthe burden of proof rests with those who de-mand evidence of safety or with those who de-mand evidence of hazard. The former wouldhalt experiments until they are proved safe. Thelatter would continue experiments until it isshown that they might cause harm.

A significant theoretical difference exists be-tween the two approaches. Evidence can almostalways be provided to show that somethingcauses harm—e.g., it can be demonstrated that apoliovirus causes paralysis, that a Pneumococcuscauses pneumonia, that a rhinovirus causes thecommon cold. However, it cannot be demon-strated that a poliovirus can never cause thecommon cold, It cannot be demonstrated thatrDNA molecules will never be harmful. It can

204 . Impacts of Applied Genetics—Micro-Organisms, Plants, and Animals

only be demonstrated that harmful events are level of uncertainty it is willing to accept.unlikely. Hence, society must determine what

Other concernsConcerns raised by industrialapplications

Originally concerns involved hazards thatmight arise in the laboratory. Now that thereare industrial applications of genetic engineer-ing, the concerns include:

. risks associated with the laboratory con-struction of new strains of organisms,

. risks associated with industrial productionor consumer use of the new strains, and

risks associated with the products obtainedfrom the new strains.

Many similar considerations apply to the as-sessment of the first two kinds of risks. Unlessthe organisms used in an industrial productionscheme are thoroughly characterized, conjec-tured fears about their ability to cause diseasewill continue. Even with a recombinant orga-nism that has a well-defined sequence of DNA, abreak in containment would leave its behaviorin the environment questionable. Experiencewith substances such as asbestos gives rise tofears that exposure to the new biological sys-tems might also cause unforseen pathologicalconditions at some future time.

Hazards associated with products raise dif-ferent questions. The growing consensus inFederal regulatory agencies appears to be thatthese products should be assessed like allothers—e,g., human growth hormone (hGH)produced by genetically engineered bacteriashould be tested for purity, chemical identity,and biological activity just like hGH from humanpituitary glands. The possibility of productvariation due to mutation of the bacteria,however, suggests that batch testing and certifi-cation might be warranted as well. (For furtherdiscussion see ch. 11.)

Concerns raised by the implications ofthe rDNA controversy for generalmicrobiology

Questions about the potential harm fromgenetically engineered micro-organisms haveled to questions about the efforts currentlyemployed to protect the public from work beingdone with micro-organisms known to be hazard-ous. These viruses, bacteria, and fungi arehandled daily in laboratory experiments, in theroutine isolation of infectious agents from pa-tients, and in the production of vaccines in thepharmaceutical industry.

Questions have been raised about the efficacyof regulations established for these variouspotentially hazardous agents. A full-scale assess-ment is not within the scope of this study, but itis clear that the questions are pertinent. Twoconclusions have been reached.

First, there is a growing belief that the mereexistence of a classification scheme for hazard-ous agents by the Center for Disease Control(CDC) is not enough to ensure their safe han-dling. The Subcommittee on Arbovirus Labora-tory Safety was formed recently because of con-cerns expressed in academic circles. Represent-atives from universities, the Public Health Serv-ice, the U. S. Department of Agriculture, andthe military, who constituted the subcommit-tee, are preparing a report based on an interna-tional survey of laboratory practices and infec-tions. They found wide variation in the waysdifferent agents were handled. Most of theirrecommendations are identical with those ap-plicable to rDNA–that appropriate containmentlevels be used with different viruses, that thehealth of workers be monitored, and that an In-stitutional Biosafety Committee be appointed toserve each institution.


Second, little is known about the healthrecord of workers involved in the fermentationand vaccine industries. For most industrialoperations the evidence of harm is almost en-tirely anecdotal. Most industrial fermentationsare regarded as harmless; representatives of in-dustry characterize it as a “non-problem” thathas never merited monitoring. Comprehensiveinformation on the potential harmful effectsassociated with research using rDNA-carrying

. micro-organisms will not be available becausethe Guidelines consider it the responsibility ofeach institution or company to “determine, inconnection with each project, the necessity formedical surveillance of recombinant-DNA re-search personnel. ” Hence some institutionsmight decide to keep records of some or all ac-tivities; others might not.

To be sure, some companies have exceededthe minimal medical standards set by NIH forfermentation using rDNA-carrying micro-orga-nisms–e.g., Eli Lilly & Co. requires that allillnesses be reported to supervisors and that anyemployees who are ill for more than 5 daysmust report to a physician before being allowedto return to work. Any employee taking antibi-otics (which might make it easier for bacteria tocolonize) is restricted from areas where rDNAresearch is being done until 5 days after the dis-continuance of the antibiotic. At Abbott Labora-tories, a physician checks into the illness of anyrecombinant worker who is off more than 1day–a precaution taken only after 5 days offfor workers in other areas. Lilly maintains acomputer listing of all workers involved inrDNA activities. Lilly, the Upjohn Co., andMerck, Sharp and Dohme have been in theprocess of computerizing the health records ofall their employees over the past several years.

Work with rDNA has focused attention onbiohazards and medical surveillance—an aware-ness that had arisen in the past but had not beensustained. * Consequently, several documentson the subject either have been or will be pub-lished:

● As of September 1980, the National Institutes of occupationalSafety and Health and the Ent’ironnlental Protection Agency wereplanning to fund assessments of the adequacy of current medicalsurveillance technology,

●

●

●

●

CDC is preparing a complete revision of itslaboratory safety manual, which is widelyused as a starting point by other labora-tories.The Classification of Etiologic Agents on theBasis of Hazard, which was last revised in1974, has been expanded by CDC in collab-oration with NIH into a Proposed BiosafetyGuidelines for Microbiological and Biomedi-cal Laboratories. These guidelines serve thepurpose fulfilled by the Dangerous Patho-gens Advisory Group (DPAG) in the UnitedKingdom, although they lack any regula-tory strength.A comprehensive program in safety,health, and environmental protection wasdeveloped in 1979 by and for NIH. It is ad-ministered by the Division of Safety, whichincludes programs in radiation safety, oc-cupational safety and health, environmen-tal protection, and occupational medicine.The Office of Biohazard Safety, NationalCancer Institute has just completed a 3-year study of the medical surveillance pro-grams of its contractors; a report is beingdrafted.

Although the academic, governmental, andindustrial communities have shown growing in-terest in biosafety, * no Federal agency regulatesthe possession or use of micro-organisms exceptfor those highly pathogenic to animals and forinterstate transport. * * Whether such regula-tions are necessary is an issue that extends be-yond the scope of this study. Nevertheless,other countries—for instance the United King-dom, with its DPAG—have acted on the issue.This organization functions specifically to guardagainst hazardous micro-organisms, by moni-toring and licensing university and industriallaboratories and meting out penalties whennecessary.

“Curiously, there is no formal society or journal, hut there has

been an annua l B io log ica l Safety Conference s ince 1955, con-

ducted on a round-rohin hasis primarily h~ close associates of the

late Arnold G. Wedum, M. D.—fornler Director of Industrial Health

anci Safet-v at the LI. S. Army Bioiogica] Research Laboratories, FortDetrick, Md., who is regarded as the “Father of MicrobiologicalSafety. ”

● “In some States and cities, licensing is required for all facilitieshandling pathogenic micro-organisms.

206 ● Impacts of Applied Genetics— Micro-Organisms, Plants, and Animals

Concerns raised by the implications of outside the cell. If the DNA is combined withinthe rDNA controversy for other living cells, the Guidelines do not pertain. Figuregenetic manipulation 35 shows several methods that achieve the same

goal–transfering genetic material from one cellAltering the hereditary characteristics of an to another, bypassing the normal sexual

organism-by usingseveral methods ofdefinition of rDNAcombination of the

rDNA is just one of the mechanisms of mating. It is particularly signifi-genetic engineering. The cant that DNA from different species can berefers specifically to the combined by all these mechanisms, only one of

DNA from two organisms which is rDNA. Different species of bacteria,

Figure 37.-Alternative Methods for Transferring DNA From One Cell to Another

A. The two cells are fused in totoB. A microcell with a fragmented nucleus carries the DNAC. Free DNA can enter the recipient cell in a number of ways: by direct microinjection, through

calcium-mediated transformation, or by being coated with a phospholipid membrane inorder to fuse with the recipient cell

D. The free DNA can be joined to a plasmid and transferred as recombinant DNA


Ch. 10—The Question of Risk ● 207

fungi, and higher organisms can all be fused ormanipulated. *

Opponents of rDNA have stated that combin-ing genes from different species may disturb anextremely intricate ecological interaction that isonly dimly understood. Hence, such experi-ments, it is argued, are unpredictable and there-fore hazardous. If so, all the other methodsrepresented in figure 35 should be included inthe Guidelines. Yet they are not.

The most acceptable explanation for this in-consistency is that rDNA is currently the most

● For example, antibiotic resistant plasmids have been trans-ferred from Staphylococcus aureus to Bacillus subtilis acrossspecies barriers by transformation, not by rDNA. Foreign genesfor the enzyme amylase have also been introduced into B. subtilis.

ConclusionThus far, no demonstrable harm associated

with genetic engineering, and particularlyrDNA, has been found, But although demonstra-ble harm is based on evidence that damage hasoccurred at one time or another, it does notmean that damage cannot occur.

Conjectural hazards based on analogies andscenarios have been addressed and most haveproved less worrisome than previously as-sumed. Nevertheless, there is agreement thatcertain experiments, such as the transfer ofgenes for known toxins or venoms into bacteria,should still be prohibited because of the reallikelihood of danger. Still other experimentscannot clearly be shown to be hazardous orreadily dismissed as harmless. Hence, a politicaldecision is likely to be required to establishwhat constitutes acceptable proof and whomust provide it.

Given that potential harm can be identified insome cases, its probable occurrence and magni-tude quantified, and perceived risk taken intoaccount, a decision to proceed is usually basedon society’s willingness to take the risk. Thistriad of the physical (actual risk), psychological(perception of risk), and political (willingness totake risk) plays a role in all decisions relating togenetic engineering.

efficient and successful method of combininggenes from very diverse organisms. It is reason-able to ask, however, what would happen if anyof the other methods become equally success-ful. Will a profusion of guidelines appear? Willone committee oversee all genetic experiments

Ethical and moral concerns

The perceived risk associated with geneticengineering includes ethical and moral hazardsas well as physical ones. It is important torecognize that these are part of the generaltopic of risk. To some, there is just as much riskto social values and structure as to humanhealth and the environment. (For further dis-cussion see ch. 13.)

The potential benefits must always be con-sidered along with the risks. Decisions made byRAC have reflected this view—e.g., when itapproved the cloning of the genetic material ofthe foot-and-mouth disease virus. The perceivedbenefits to millions of animals outweighed thepotential hazard.

Recombinant DNA techniques represent justone of several methods to join fragments ofDNA from different organisms. The currentGuidelines do no extend to these other tech-niques, although they share some of the sameuncertainties. Ignoring the consequences of theother technologies might be viewed as an incon-sistency in policy.

While the initial concerns about the possibili-ty of hazards at the laboratory level appear tohave been overstated, other types of potentialhazards at different stages of the technologyhave been identified. Emphasis has shiftedsomewhat from conjectured hazards that mightarise from research and development to thosethat might be associated with production tech-nologies. As a consequence, there is a clearermandate for existing Federal regulatory agen-cies to play a role in ensuring safety in industrialsettings.

part iii institutions and society

Documents