Time to reassess eugenics?

Eugen­ics is an ancient idea with a ter­ri­ble mod­ern rep­u­ta­tion. Is that rep­u­ta­tion now irrrel­e­vant?

[A dis­cus­sion paper for a talk I gave in 2015: repro­duced here for my records, and per­haps your inter­est]

TL; DR

• Nine­teenth cen­tu­ry lib­er­als —- con­tem­po­raries of Dar­win —- revived the idea, which com­pris­es two relat­ed objec­tives, one neg­a­tive, one pos­i­tive. They are: the elim­i­na­tion of defects and dis­eases in the pop­u­la­tion and the improve­ment of desir­able pop­u­la­tion char­ac­ter­is­tics such as intel­lec­tu­al or phys­i­cal capac­i­ty.

• Although enthu­si­as­ti­cal­ly sub­scribed by lead­ing thinkers and adopt­ed by gov­ern­ments in the first few decades of the 20th cen­tu­ry, eugenic ideas, then, had no plau­si­ble pro­gram. Nazi the­o­ries of racial hygeine and their mur­der­ous pur­suit by Hitler’s gov­ern­ment per­vert­ed the ‘neg­a­tive’ eugenic objective.The idea of eugen­ics, pos­i­tive or neg­a­tive was appar­ent­ly taint­ed for­ev­er.

• Then, at the end of the twen­ti­eth cen­tu­ry, advances in cel­lu­lar biol­o­gy and genom­ic sci­ence pro­vid­ed some hope of a pro­gram for (‘neg­a­tive’) eugenic improve­ments to which there are few oblec­tions: the elim­i­na­tion of genet­ic defects and dis­eases. But the mech­a­nisms — already in selec­tive use — are expen­sive, unre­li­able and do not nec­es­sar­i­ly elim­i­nate defec­tive genes. 

• In the past year or two [mid-2015] bio­log­i­cal lab­o­ra­to­ries around the world have rushed to acquire a new, sim­ple and effec­tive genome edit­ing tool, based on “nat­ur­al” mech­a­nisms used by bac­te­ria to stave off infec­tion. Last month, Chi­nese researchers announced they used CRISPR-CAS9 to edit the genome of human embryos to delete genes respon­si­ble for ß‑thalassemia. 

• The rapid spread of a com­par­a­tive­ly sim­ple mech­a­nism that could final­ly achieve eugenic objec­tives sug­gests it is high time to restart the dis­cus­sion (lest gov­ern­ments again rush into unwar­rant­ed action).

A brief historical review

Eugen­ics has a rep­u­ta­tion no bet­ter than racism or slav­ery.

It was an ancient idea (approved by Pla­to in The Repub­lic) when revived in the UK in the 19th cen­tu­ry. An ear­ly, and dis­tin­guished, advo­cate was the Vic­to­ri­an poly­math (cousin of Charles Dar­win), Fran­cis Dal­ton, who coined the term ‘eugen­ics’.¹ 

Dal­ton began the debate about “nature ver­sus nur­ture”, and gave us the phrase. He devised inno­v­a­tive exper­i­ments to inves­ti­gate the the­sis that intel­li­gence (‘Genius’) was her­i­ta­ble. In stud­ies of twins (now a stand-by of psy­cho­log­i­cal tech­nique) he con­clud­ed that it was nature, rather than nur­ture, that account­ed for supe­ri­or intel­lec­tu­al abil­i­ty. Dal­ton, like his cousin, was also a thought­ful con­t­trib­u­t­or to evo­lu­tion­ary the­o­ry and an effec­tive oppo­nent of Lamar­ck­’s pro­pos­als — to which Dar­win had been some­what attract­ed — for the inher­i­tance of acquired traits.² Although his eugenic the­o­ries had no bio­log­i­cal foun­da­tion (Gre­gor Mendel’s genet­ic exper­i­ments were not redis­cov­ered until the turn of the 20th cen­tu­ry), Dal­ton believed he had estab­lished a sound basis for eugenic prac­tice.

The enthu­si­as­tic embrace of empir­i­cal dis­cov­ery was what made 18th and 19th cen­tu­ry Britain great. Eugen­ics became an aca­d­e­m­ic dis­ci­pline. Intel­lec­tu­al and polit­i­cal lead­ers formed learned soci­eties to pro­mote eugenic ideas and pro­grams. In keep­ing with inter­na­tion­al­ist spir­it of the first age of ‘glob­al­iza­tion’, an Inter­na­tion­al Fed­er­a­tion of Eugen­ics Orga­ni­za­tions coor­di­nat­ed sci­en­tif­ic research. The Carnegie foun­da­tion estab­lished a Eugen­ics Records Office. Gov­ern­ments in many coun­tries adopt­ed eugenic poli­cies and pro­grammes includ­ing birth con­trol and abor­tions for class­es of intel­lec­tu­al­ly or phys­i­cal­ly dis­abled peo­ple; they pro­mot­ed mar­riages and births for class­es of ‘desir­able’ indi­vid­u­als. 

But these prac­tices were, like many gov­ern­ment pro­grammes, often free of good sup­port­ing evi­dence, not mon­i­tored for suc­cess, and eas­i­ly sub­vert­ed by the prej­u­dices, hunch­es and even mal­ice of the par­ties in pow­er. The Nazi the­o­ries of racial hygeine and the geno­cide, mur­der and muti­la­tions car­ried out against Jews, (overt) homo­sex­u­als, “gyp­sies” and the dis­abled or indi­gent rep­re­sent­ed the deep­est cor­rup­tion of egu­genic ideas. A series of post-war UN con­ven­tions on human rights and geno­cide con­demned and pro­hibt­ed actions tak­en in the name of eugen­ics. Still, gov­ern­ment eugenic pro­grams have nev­er quite dis­ap­peared; wit­ness the eugenic objec­tives of the baby bonus­es for grad­u­ates that were a fea­ture of Lee Kuan Yew’s Sin­ga­pore pop­u­la­tion pol­i­cy.

Eugenic pro­grams have often been described as either “pos­i­tive” or “neg­a­tive”. The pos­i­tive goal is to favour the repro­duc­tion of indi­vid­u­als whose desir­able traits are con­sid­ered her­i­ta­ble. The goal of a “neg­a­tive” pro­gramme is to elim­i­nate the unde­sir­able traits observed in a pop­u­la­tion by pre­vent­ing their repro­duc­tion. The neg­a­tive pro­gramme has been the focus of con­dem­na­tion because, until the end of the 20th cen­tu­ry, the most direct mech­a­nism was mur­der or ster­il­i­sa­tion of indi­vid­u­als whose unde­sir­able traits were (thought to be) her­i­ta­ble. But there were also more benign social and tech­no­log­i­cal trends that qual­i­fied. For exam­ple, the renowned biol­o­gist J.B.S. Hal­dane observed that the arrival of wide­spread bus trans­port in rur­al eng­land in the 1930s “by break­ing up inbred vil­lage com­mu­ni­ties, was a pow­er­ful eugenic agent.”

Still, there have been poten­tial­ly desir­able tar­gets for neg­a­tive eugen­ics for some decades since the recog­ni­tion that some dis­eases and dis­or­ders have a genet­ic basis (which is not to say ’cause’). As of 2013, there were 3788 dis­eases or dis­or­ders asso­ci­at­ed with a fault in a known gene. They include achon­dropla­sia (dwafism) and autism, some breast, colon and prostate can­cers, Crohn’s dis­ease, cys­tic fibro­sis, Down syn­drome, Duchenne mus­cu­lar dys­tro­phy, hemochro­mato­sis, hemo­phil­ia, Mar­fan syn­trome, myoton­ic dis­tro­phy, Parkin­son’s, retini­tis pig­men­tosa, sick­le cell dis­ease, tha­lassemia and many, many oth­ers. Some of these — such as dwarfism — are also due to spon­ta­neous chro­mo­so­mal abnor­mal­i­ties, but all in this list are also her­i­ta­ble.

So, despite its poor rep­u­ta­tion, most of us would prop­bably sub­scribe to a neg­a­tive eugenic pro­gramme if it elim­i­nat­ed the her­i­ta­ble risk of these dis­eases and dis­or­ders and could be achieved with­out doing oth­er harm.³ Indeed, eugenic manip­u­la­tion of the ‘gene pool’ has been tak­ing place at the mar­gins for a decade with­out any con­tro­ver­sy. In all high-income economies and in many emerg­ing economies too (Chi­na, India), pre-implan­ta­tion genet­ic test­ing of IVF-ini­ti­at­ed pre­gan­cies is avail­able for prospec­tive par­ents with known genet­ic risks. Tech­ni­cians remove one or two cells from the blas­tomere some three to five days after fer­til­iza­tion of the ovum. They test these cells for the tar­get genet­ic defect and only those embryos that, thanks to the Mendelian lot­tery, are free of the dis­ease are implant­ed. 

But this pro­ce­dure does not nec­es­sar­i­ly elim­i­nate genet­ic risk from the germ line. On the con­trary, in the case of an auto­so­mal reces­sive (sin­gle-) gene defect affect­ing both par­ents, any embryo has a 50% chance of hav­ing one defec­tive gene (but no dis­ease) and only a 25% chance of hav­ing no defec­tive gene in its germ cells. Three-quar­ters of all embryos will be ‘suc­cess­ful’ from the view­point of the par­ents (and the child), but only a quar­ter of them will suc­ceed from an eugenist’s view­point. In any case, the incen­tives of pre-implan­ta­tion selec­tion are all wrong for the eugeni­cist’s pur­pos­es. The rewards the par­ents seek are at strong odds, while the rewards the eugenist seeks are unlike­ly by com­par­i­son.  As for pos­i­tive eugen­ics, until recent­ly it seemed at best dot­ty, at worst a repul­sive racist fan­ta­sy. In any case, it lacked cred­i­bil­i­ty. The length of a human gen­er­a­tion — say, 16 to 20 years — was too long to allow repeat­ed exper­i­ment based on Mendelian genet­ic com­bi­na­torics. 

Pre-implan­ta­tion test­ing, for exam­ple, would be of lit­tle use in a “pos­i­tive” eugenic pro­gram. Assum­ing the genet­ic basis of desir­able traits such as skill in math­e­mat­ics were known (an hero­ic assump­tion) a rig­or­ous pro­gram of pre-implant selec­tion might ensure that every infant car­ried the opti­mum mix of alle­les from two “fit” par­ents. If the odds of pos­i­tive out­comes were bet­ter than in the case of reces­sive auto­so­mal dis­ease, there would be no runt in any lit­ter. If the desir­able char­ac­ter­is­tic were more uni­form­ly her­i­ta­ble than reces­sive genet­ic dis­ease (and not espe­cial­ly vul­ner­a­ble to vari­able epi­ge­net­ic expres­sion), pre-implan­ta­tion pro­ce­dures might actu­al­ly change the com­po­si­tion of the gene pool. But those are demand­ing pre-con­di­tions. It might be more effec­tive sim­ply to sub­sidise the nat­ur­al repro­duc­tion of the fittest indi­vid­u­als and sub­mit to the Mendelian lot­tery on the the­o­ry that, among a large num­ber of off­spring, the pro­por­tion of less-fit would be small.⁴ 

Nor is ‘cloning’ an effec­tive mech­a­nism of pos­i­tive eugen­ics. It is cer­tain­ly fea­si­ble and hap­pens both nat­u­ral­ly (monozy­got­ic twins) and, poten­tial­ly, by inter­ven­tion in an IVF preg­nan­cy to divide the blas­tomere and implant both embryos. But a pos­i­tive eugenic pro­gram would seek to clone indi­vid­u­als whose fit­ness was already demon­strat­ed; adults, in oth­er words. The avail­able cloning mech­a­nism for adults is Somat­ic Cell Nuclear Trans­fer (SCNT), which extracts the full (‘dou­ble’) set of chro­mo­somes nec­es­sary for the via­bil­i­ty of the organ­ism from a somat­ic (usu­al­ly ‘stem’) cell of the adult sub­ject and implants it as a ‘replace­mem­nt nucle­us’ in the cyto­plasm of an ovum from anoth­er indi­vid­ual. This is the mech­a­nism that gave us “Dol­ly” the cloned sheep. But a decade of expe­ri­ence with cloning of large mam­mals (dairy cat­tle, race­hors­es) shows that clones are gen­er­al­ly less robust than oth­er ani­mals and that preg­nan­cies with cloned embryos are more dif­fi­cult and have a low­er suc­cess rate.

Then in 2012 the out­look for both pos­i­tive and neg­a­tive eugenic pro­grammes changed, dra­mat­i­cal­ly. Researchers in the USA and France dis­cov­ered a rapid, uncom­pli­cat­ed, ‘nat­ur­al’ way to manip­u­late genes in both germ and somat­ic cells that works in vivo.⁵ The break­through came from study­ing a tech­nique used by bac­te­ria to fight viral infec­tions. Just a decade ear­li­er, researchers began to sus­pect that cer­tain repeat­ed sequences (so-called CRISPR sequences) of DNA code in bac­te­r­i­al DNA were real­ly “libraries” of viral DNA code, stored by the bac­te­ria in their own DNA. The bac­te­ria used these libraries to rec­og­nize and then to destroy invad­ing viral infec­tions using a set of enzyme ‘scis­sors’ (them­selves cod­ed by CAS genes or ‘CRISPR-asso­ci­at­ed’ genes) to chop-up and dis­rupt viral DNA. In bac­te­ria, the CRISPR-CAS sys­tem is a sort of immune mem­o­ry because bac­tir­i­al insert the CRISPR sequences into their own DNA (using the same CAS scis­sors). When it repro­duces, the bac­teri­um pass­es-on its ‘library’ to the next gen­er­a­tion along with copies of the rest of its DNA.⁶ 

It turned out that the CRISPR-CAS mech­a­nism for dis­rupt­ing viral DNA was par­tic­u­lar­ly sim­ple and effi­cient because it relied only on two RNA ‘mes­sen­ger’ strands cre­at­ed from the CRISPR ‘library’ tem­plates to pro­gram the CAS enzyme gene-scis­sors. The CAS enzyme ‘cradling’ these two CRISPR RNA sequences locks onto to the start of a seg­ment of the viral DNA iden­ti­fied as a sequence of bases match­ing its RNA guides. The CAS ‘scis­sors’ then cuts both strands of the DNA mol­e­cule at the end of the match­ing sequence.  In 2012, Jen­nifer Doud­na and her col­leagues showed that they could fur­ther sim­pli­fy and improve this nat­ur­al sys­tem, turn­ing it from a find-and-destroy to a find-and-replace gene edi­tor. They syn­the­sised a sin­gle strand of RNA to use for the ‘find’ half of the oper­a­tion. They implant­ed their RNA strand in a CAS enzyme that cut the dou­ble strand of DNA at the ‘found’ loca­tions (more than one loca­tion, if nec­es­sary, clip­ping out a length of the strand). They also demon­strat­ed that if they intro­duced a short sequence of DNA with the appro­pri­ate form into the nuclear ‘soup’ the cel­l’s repair mech­a­nisms would use this sequence to ‘repair’ the DNA they had just cut. This is the ‘replace’ side of the oper­a­tion. The end result is a revised ‘dou­ble-helix’ DNA mol­e­cule that can take its place with oth­er such mol­e­cules in the genet­ic mate­r­i­al of the cell.

The sim­plic­i­ty of this sys­tem, bor­rowed from an ancient, uni­ver­sal bac­te­r­i­al sur­vival mech­a­nism was imme­di­ate­ly attrac­tive to biol­o­gists inter­est­ed in bac­te­r­i­al, plant and ani­mal genome edit­ing. Even before Doud­na’s lab had demon­strat­ed the arti­fi­cial RNA sys­tem for direct­ed edit­ing, food tech­nol­o­gists at Danis­co had begun to manip­u­late CRSIPR sequences in bac­te­ria found in yoghurts and cheeses to improve their resis­tance to viral infec­tion (Eat­en yoghurt or cheese rcent­ly? You’ve almost cer­tain­ly eat­en CRISPRized cells). After the pub­li­ca­tion of the Doud­na paper, lit­er­al­ly hun­dreds of biol­o­gy research and indus­tri­al lab­o­ra­to­ries rushed to repli­cate and apply the CRISPR-CAS method. 

Soon groups of sci­en­tists were writ­ing por­ten­tious warn­ings in the New York Times, as well as the jour­nals Nature and Sci­ence, demand­ing that sci­en­tists every­where abjure the use of this tech­nol­o­gy to alter the human genome.⁷ They acknowl­ged the val­ue of genome edit­ing tech­nolo­gies of somat­ic cells as a defense against dis­ease. They warned, rea­son­ably enough, of unin­tend­ed con­se­quences if researchers applied the tech­niques to germ cells. They warned against attempts to devel­op of “non-ther­a­peu­tic genet­ic enhance­ment”, but they did not explain why this would be a bad thing.

Then, in April, a team of researchers at a uni­ver­si­ty in Guangzhou report­ed the ambigu­ous and most­ly dis­ap­point­ing results of an attempt to mod­i­fy the gene respon­si­ble for β‑thalassaemia in human embryos. The jour­nals Nature and Sci­ence refused to pub­lish the research (although they did not say why): the online jour­nal Pro­tein and Cell car­ried the paper.⁸ 

Discussion and questions

• Gov­ern­ments have been rep­sonible for the worst (the only?) excess­es com­mit­ted in the name of eugen­ics. They alone have had the coer­cive pow­er, in the past, to enforce a eugenic pro­gram. But is the genome a “pub­lic good”? Is it a sort of “com­mons” whose exploita­tion is like­ly to give rise to the “mar­ket fail­ures” that jus­ti­fy gov­ern­ment reg­u­la­tion?  

• Do pub­lic vac­ci­na­tion pro­grams (so far, not eugenic because they do not impact the genome) sug­gest that the genome must be a pub­lic good? If there were a genom­ic change that, say, would give indi­vid­u­als a her­i­ta­ble resis­tance to HIV (an infec­tious dis­ease) would gov­ern­ments have the right to enforce genom­ic treat­ment on all indi­vid­u­als on the basis that the genome is a pub­lic good ? Or, is the genome mere­ly an abstract col­lec­tive term for a pri­vate good — the genet­ic her­itage of indi­vid­u­als — where gov­ern­ment has no pos­i­tive role (and will only make a mess of things as it has in the past)?

• OECD juris­dic­tions strict­ly pro­tect an indi­vid­u­al’s genet­ic infor­ma­tion under pri­va­cy laws. Does this sug­gest that the genome is mere­ly a col­lec­tive term for a pri­vate prop­er­ty? What is the appro­pri­ate response to the “unin­tend­ed con­se­quences” threat? A “ban”? A UN con­ven­tion? Should it cov­er the edit­ing of both somat­ic cells and germ cells? > Would a gov­ern­ment ban — say, co-ordi­nat­ed by an UN agency — on germ-line edit­ing work? Would it be uni­ver­sal­ly respect­ed? Or would it be like a ‘car­tel’ restric­tion; open to covert cheat­ing? Is there a ‘first mover’ advan­tage in eugenic inter­ven­tion that would favor such cheat­ing? Devel­op­ing supe­ri­or war­riors? 

  • Con­sid­er­ing the risks of edit­ing the genome, would a UN-coor­di­nat­ed restric­tion on germ-line edit­ing even be nec­es­sary? Con­sid­er that none of the dozens of states that have nuclear weapons has used one in the past 70 years or even placed itself in a posi­tion where that choice might be immi­nent. Is there a prob­lem in prin­ci­ple with “non-ther­a­peu­tic” inter­ven­tion (“pos­i­tive” eugen­ics when her­i­ta­ble) as the sci­en­tists writ­ing in Nature claim? 

• Sup­pose it became pos­si­ble to select for cer­tain desir­able char­ac­ter­is­tics (not mere­ly the absence of dis­ease), whether on a her­i­ta­ble basis or not. Sup­pose par­ents could choose to have chil­dren that had a high chance of being a “super ath­lete” or a high chance of being a prodi­gal math­e­mati­cian or musi­cian, would par­ents take the option? 

• Sup­pose, fur­ther, (more real­is­ti­cal­ly?) that the out­come of such a non-ther­a­peu­tic action entailed mixed-risks. Say, due to unpre­dictable epi­ge­net­ic fac­tors, there were a 90 per­cent chance the child born would be a genius and a ten per­cent chance the child would become dis­turbed recluse or, pos­si­bly, psy­chot­ic. What would par­ents choose?

• What would guide par­ents in the “design” of their chil­dren, should that prove pos­si­ble? Would the con­sid­er­a­tions be dif­fer­ent in the case of a her­i­ta­ble (eugenic) inter­ven­tion? PWG 30 April, 2015

The Guangzhou researchers set up their exper­i­ment with care. They used only ‘non-viable’ embryos (dou­bly-fer­til­ized ova with three sets of chro­mo­somes that can­not result in a live birth). But the exper­i­ment did not suc­ceed. The pro­grammed ‘find and replace’ had a low hit rate and that there were ‘off-tar­get’ muta­tions appar­ent­ly due to the CRISPR-CAS9 enzymes attach­ing to the wrong genes. The oppo­nents of any human genome edit­ing acclaimed the messy result as a con­fir­ma­tion of their warn­ings. But, it might be more rea­son­able to see the Chi­nese result as a “fruit­ful fail­ure” that shows where research should now focus to achieve the “100%” suc­cess that an acu­tal inter­ven­tion would need.