HumanInsulin from
`Recombinant DNA Technology
`
`Irving S. Johnson
`
`Summary. Human insulin produced by recombinant DNA technology is thefirst
`commercial health care product derived from this technology. Work on this product
`wasinitiated before there were federal guidelines for large-scale recombinant DNA
`work or commercial development of recombinant DNA products. The steps taken to
`facilitate acceptance of large-scale work and proofof the identity and safety of such a
`product: are described. While basic studies in recombinant DNA technology will
`continue to have a profound impact on research in the life sciences, cormmercial
`applications may well be controlled by economic conditions and the availability of
`investmentcapital.
`
`growing more rapidly than the total pop-
`ulation, While the U.S. population is
`increasing at a rate of about 1 percent per
`year and the world population at slightly
`more than 2 percent, the annual rate of
`increase of insulin-using diabetics in this
`country has been5 to 6 percentin recent
`years, and a similar pattern may hold
`true worldwide (2).
`Several factors contribute to this ac-
`celerated growth of the insulin-using dia-
`betic population. One factor, of course,
`is the availability of insulin, which cn-
`ables diabetics, who often did not sur-
`vive beyond their teens,
`to live long,
`During 1982, human insulin of recom-
`productive—and—reproductive—lives..
`University of Toronto to develop a stan-
`dardized and clinically acceptable insulin
`Because of the genetic etiologic compo-
`binant DNA origin was approved by the
`nent of diabetes, the offspring of diabet-
`product. Banting had just begun to ex-
`appropriate drug regulatory agencies in
`ics are likely to suffer from the disease as
`the United Kingdom,
`the Netherlands,
`tract relatively crude insulin from ani-
`well. Other factors that contribute to
`mals and inject it into his diabetic pa-
`West Germany, and the United. States.
`tients.
`growth of the diabetic insulin-using pop-
`This new source guarantees a reliable,
`expandable, and constant supply of hu-
`In the early 1970’s we began to be
`ulation include improved methods of de-
`man insulin for diabetics around the
`tection, greater public awareness of the
`concerned about a possible shortage of
`world.
`insulin. Until now,
`the world’s insulin
`disease and its symptoms, less reliance
`needs have been derived almost exclu-
`on oral forms of therapy, and changes in
`The research, development, and pro-
`dietary habits.
`sively from pork and beef pancreas
`duction of human insulin by recombinant
`Becauseof the uncertainty of the insu-
`glands, which were collected as by-prod-
`DNA technology ushers in a new era in
`ucts from the meat industry. This supply
`lin supply and the forecasts of rising
`pharmaceuticals, agricultural products,
`insulin requirements, it seemed not only
`prudent but a responsibility as well for
`the scientific community and insulin
`manufacturers to develop alternatives to
`animal sources for supplying insulin to
`the world’s diabetics. Lilly established
`several internal committees of scientists
`to examine various solutions to the prob-
`lem. They considered augmentation of
`insulin production from pancreas glands,
`transplantation of islet of Langerhans
`cells, chemical synthesis, beta cell cul-
`ture, directed-cell synthesis, and cell-
`free biosynthesis, as well as insulin re-
`placements. These discussions touched
`on the technology called genetic engi-
`neering.
`is to
`The function of DNA in a cell
`serve as a stable repository of coded
`information that can be replicated at the
`time of cell division to transmit the ge-
`netic information to the progeny cells
`and to encode the information necessary
`to synthesize proteins and. other cell
`components. There are several ways of
`performing genetic engineering, some of
`which have been practiced for many
`years by geneticists. The first is muta-
`tidn. Mutations in DNA can be either
`spontaneous, due to environmental fac-
`tors and errors in DNA replication, or
`they can be induced in the laboratory by
`physical and chemical agents. Mutations
`can lead to a changein the structure of
`the product coded for by the gene in
`question;
`sometimes
`this
`change
`in
`structure is so great that the product is
`SCIENCE, VOL. 219
`
`and industrial chemicals by establishing
`the feasibility of commercial production
`of a gene productinitiated at a laboratory
`level of expression.
`| shall review how
`human insulin became the first human
`health product of this technology. I will
`also discuss some of the special prob-
`lems, in terms of regulatory environment
`and public opinion, that had to be over-
`come in order to bring it to the current
`stage of development.
`
`Sources of Insulin
`
`Eli Lilly and Company has been in-
`volved in the development and manufac-
`ture of insulin and other products for
`diabetics since 1922.
`In that year our
`scientists began working with Frederick
`G. Banting and his associates at
`the
`
`Irving S. Johnson is vice president of research,
`Lilly Research Laboratories, a division of Eli Lilly
`and Company, Indianapolis, Indiana 46285.
`632
`
`changes with the demand for meat andis
`not
`responsive to the needs of
`the
`world’s diabetics. Indeed, from 1970 to
`1975, the supply of pancreas glands in
`the United States declined sharply (/)
`and remained on a plateau at that lower
`level
`in succeeding. years. There is no
`accurate way to predict availability of
`future supplies of glands, although we
`predicted that
`the demand for insulin
`would continueto increase. Our concern
`was whether or not there would be a time
`when the supply of bovine and porcine
`pancreas glands might not be sufficient
`to meet the needs of insulin-dependent
`diabetics. Although it is difficult to ob-
`tain substantiated figures for a nonreport-
`able disease, we estimate that there are
`60 million diabetics in the world—more
`than half of them in less developed coun-
`tries. In the developed countries, some 4
`million diabetics, 2 million in the United
`States, are treated with insulin.
`Today,
`the diabetic population is
`
`Page 1
`
`KASHIV EXHIBIT 1016
`IPR2019-00797
`
`Page 1
`
`KASHIV EXHIBIT 1016
`IPR2019-00797
`
`

`

`.
`
`scientific community could evaluate the
`vastly different. Other mutations may
`clear, and broader participation in re-
`risks and benefits associated with it (6).
`affect the regulatory clements that con-
`search decisions by those expert in infec-
`tious diseases, containment, and risk-
`Work washalted——including work in my
`trol the expression of the structural gene,
`leading, in some instances, to increased
`assessment, as well as more practical
`own organization—until after the Asilo-
`mar Conference in 1975 (7).
`or decreased production of gene prod-
`experience,
`led inevitably to revisions
`and continued relaxation of restrictions
`The majority of scientists invited to
`ucts. A key point is that mutagenesis is
`around the world.
`the Asilomar Conference were molecular
`an essentially random technique.
`In June 1976, the National Institutes of
`biologists from government and acade-
`A second type of genetic engineering,
`recombination, has also been used for a
`Health (NIH) announced guidelines for
`mia;
`those with expertise in infectious
`long time. Recombination refers to ex-
`disease or those from industry with ex-
`recombinant DNA work, marking the
`end of the 2-year moratorium on this
`changeof a section of DNA between two
`tensive knowledge of large-scale fermen-
`
`DNA_molecules. Recombination of
`type of research. Only research financed
`tation processes and other techniques
`DNAfragmentsfrom different organisms
`by the federal government was subjectto
`that require careful methods of contain-
`ing potentially harmful materials were
`the guidelines, for which Lilly, with oth-
`can occur by the mating of two orga-
`er companies, NIH, the Pharmaceutical
`underrepresented. As a result, many of
`nisms——a process called conjugation—
`Manufacturers Association (PMA),
`the
`where DNA is physically transferred
`us believe that the guidelines defined at
`Food and Drug Administration (FDA),
`from one organism to another. This is a
`Asilomar were unnecessarily restrictive.
`process occurring in nature, which can
`An example wasthe establishment of the
`and the Department of Health, Educa-
`tion and Weifare (HEW), was actively
`10-liter limit. To those of us in industry,
`be duplicated in the laboratory. Two
`this restriction was never considered
`other natural processes whereby cells
`involved in developing compliance pro-
`cedures.
`reasonable. We were accustomed to han-
`exchange DNA in nature are transforma-
`tion and transduction. Recombination
`dling containment problems at much
`larger volumes. But few of the scientists
`at Asilomar conceived of performing
`large-scale fermentation with recombi-
`nant organisms; as noted in the British
`journal Nature,
`‘‘[In 1975] even opti-
`mists would have predicted that it would
`be a decade before genetic engineering
`would be commercially exploited’”’ (8).
`This volume limit was established be-
`cause it was regarded as probably the
`largest volume that could be convenient-
`ly handled in an experimental laboratory
`by conventional laboratory centrifuges.
`It was clear that the 10-liter limit exclud-
`ed industrial-level activity. It was not
`suggested that
`increased volume of a
`culture of a safe organism would result in
`any increased risk. The conjectural na-
`ture of the early concerns soon became
`
`At Lilly, work on DNA recombina-
`tion, which had been under way before
`Jackson, Symons, and Berg (3) pub-
`lished their work, resumed vigorously.
`We contracted with a new California
`company, Genentech, Inc., for specific
`work on humaninsulin. Genentech sub-
`contracted the synthesis of the human
`insulin gene to the City of Hope Medical
`Center, which succeeded in its mission.
`Scientists at Genentech inserted the
`genes for both chains ofinsulin into a K-
`12 strain of Escherichia coli and after
`isolation and purification, the A and B
`chains were joined. by disulfide bonds to
`produce human insulin. In the Lilly Re-
`search Laboratories, we have also used
`recombinant DNA technology to pro-
`duce human proinsulin, the insulin pre-
`cursor.
`
`Pork
`
`Rabbit
`Human
`
`LoDettttttttttet
`
`may also occur following the use of a
`technique known as protoplast fusion,
`where oneliterally strips off the outer
`cell wall of celis of fungi and bacteria.
`This phenomenon only occurs in the
`laboratory and allows the remaining pro-
`toplasts, which now have just a cell
`membrane enclosing the components of
`the cell,
`to fuse together. The fused
`protoplasts contain the DNA molecules
`of both parents, and exchange of sec-
`tions of DNA can now occur as these
`celis regenerate and divide. All
`these
`recombination processes involve ran-
`dom exchange of DNA sequences, and
`this exchange is generally, but not al-
`ways,
`limited to members ofa single
`species of organism.
`In 1972, Jackson, Symons, and Berg
`(3) described the biochemical methods
`for cutting DNA molecules from two
`different organisms, using restriction en-
`zymes, and recombining the fragments
`to produce biologically functional hybrid
`DNA molecules. In 1973, Cohen, Chang,
`and Boyer (4) reported that they could
`make a hybrid molecule that would ex-
`press the foreign DNA within it as
`though it were a part of the original
`molecule’s natural heritage (5). That pro-
`foundly significant accomplishmentalso
`generated major concern over potential
`biohazards.
`
` BthtRdkk
`
`t. HPLC chromato-
`Fig.
`gramsofinsulins whichdif-
`fer by one or more amino
`acids.
`
`Beef
`
`1.00 338.5 676
`
`1013
`
`13951 1688 2026 2363 2701
`Seconds
`
`633
`
`Page 2
`
`Regulating DNA Research
`
`The Berg Committee was formed, and
`it responded to concerns about conjec-
`tural risk associated with recombinant
`DNAresearch in 1974 by calling for a
`moratorium and deferral of certain types
`of recombinant DNAresearch until the
`11 FEBRUARY 1983
`
`Page 2
`
`

`

`Human Insulin
`
`The successful expression of human
`insulin (recombinant DNA) in E. coli
`was announced on 6 September 1978.
`This was a first step. Although we had
`been successful in obtaining expression
`of the hormone under laboratory condi-
`tions and scale, we still faced the equally
`difficult challenge of achieving satisfac-
`tory production of the purified product
`on a commercial scale. The process we
`used in accomplishing large-scale pro-
`duction has been described (9~//), but it
`may be useful to touch on some of the
`methods that we employed to prove that
`the product produced was indeed human
`insulin.
`High-performanceliquid chromatogra-
`phy (HPLC)
`techniques developed at
`Lilly can detect proteins that differ by a
`single amino acid (/0), and HPLCtests
`showed that human insulin (recombinant
`DNA) is identical to pancreatic human
`insulin and thatit is close to, but not the
`same as, pork insulin, which differs from
`the human by one amino acid: beef,
`which differs by three amino acids; and
`sheep, which differs at four residue posi-
`tions (Fig. 1). A chromatogram of human
`insulin (recombinant DNA), pancreatic
`humaninsulin, and a mixture of the two,
`showed that they were superimposable
`and identical (Fig. 2). HPLC has become
`an important analytical tool to determine
`structure and purity and is now consid-
`ered to be a more precise measurement
`of potency than the rabbit assay, al-
`though most government
`regulatory
`agencies around the world still empha-
`size the rabbit potency assay.
`tertiary
`A measure of the correct
`structure and appropriate folding is the
`circular dichroic spectrum. The spec-
`trum for porcine insulin and for human
`insulin (recombinant DNA) were found
`to be identical. X-ray crystallographic
`studies further revealed the structural
`
`integrity of the recombinant molecule
`(12). We also found the amino acid com-
`position of human insulin (recombinant
`DNA) and pancreatic human insulin to
`be identical (Table 1). In addition, we
`compared polyacrylamide gel electro-
`phoresis for human insulin (recombinant
`DNA), pancreatic human insulin, and
`pork insulin, as well as isoelectric focus-
`ing gels for these three insulins.
`Another technique that we found use-
`ful for ensuring that we had the appropri-
`ate disulfide bonds and lacked other
`types of protein or peptide contaminants
`was HPLC of a specifically degraded
`sample. There is a staphylococcal prote-
`ase that cleaves insulin in a specific way
`at
`five sites—always next
`to glutamic
`acid, except for one site between serine
`and leucine. After treating the insulin
`with the protease, we looked for and
`identified the various peptide fragments
`by HPLC (Fig. 3) and found none that
`were not derived from insulin.
`In the end, we employed 12 different
`tests to establish that what we had pro-
`duced was human insulin. We believe
`the correlation amongthree of the tests
`was particularly important—the radio-
`receptor assay,
`the radioimmunoassay,
`and HPLC. Moreover, the pharmacolog-
`ic activity of humaninsulin (recombinant
`DNA), as demonstrated by a rabbit hy-
`poglycemia test, showed a response es-
`sentially identical
`to pancreatic human
`insulin.
`Anotherserious question remained to
`be answered—namely that of the poten-
`tial contamination of the product with
`trace amounts of antigenic E. coli pep-
`tides. Relevant to this question is the
`difference in starting materials between
`human insulin of recombinant DNA ori-
`gin and pancreatic animal insulins. The
`glandulartissue is collected in slaughter-
`houses, with no control over bacterial
`contamination. The desired gene product
`is isolated from a few cells of the islets of
`
`A Humaninsulin (recombinant
`
`DNA)
`
`B Pancreatic human insulin
`
`\
`
`
`
`
`von
`J
`C Pancreatic + human insulin
`(recombinant DNA) mix
`
`ard1. 4 L
`
`L
`Ct
`
`1.00
`188.5
`376
`663.5
`761
`Seconds
`
`!
`938.5
`
`4
`1126
`
`4.
`1313
`
`1501
`
`634
`
`Fig. 2. HPLC chromato-
`grams of human insulin
`(recombinant DNA), pan-
`creatic human insulin, and
`mixtures of the two show-
`ing identity.
`
`Page 3
`
`Langerhans, which make up less than |
`percent of the glands; thus more than 99
`percent of the tissue represents tissue
`contaminants and undesirable materials.
`The common protein contaminants of
`the animal insulins are other pancreatic
`hormones orproteins, many of which are
`highly immunogenic.
`In contrast, with recombinant DNA
`production of human insulin, almost 100
`percent of the cells (E. coli) produce the
`desired gene product. Because of the
`method of manufacture, none of the pan-
`creatic contaminants ofthe animal insu-
`lins are found in the human insulin of
`recombinantorigin. The issue of protein-
`aceous contamination derived from the
`bacterial host cell was addressed through
`some experiments that were made possi-
`ble by running large-scale fermentations
`of the production strain of E. coli, which
`contains the production plasmid with the
`code for the insulin chain sequence de-
`leted. The small quantities of peptides
`isolated after applying the chain purifica-
`tion and disulfide linking process to the
`“blank’’ preparation were shown notto
`be antigenic except in complete Freund’s
`adjuvant (/3); in addition, no changes in
`amount of antibody to E. coli peptides
`were detected in serum from patients
`who had been treated with humaninsulin
`for more than a year (/4).
`
`Commercial Production
`
`As we were scaling up this new tech-
`nology for commercial production, we
`recognized that there would be external
`problems and forces with which to con-
`tend. Because this would be the first
`human health care product
`resulting
`from recombinant DNA techniques, we
`expected that many people would per-
`ceive that
`there were risks associated
`with this new scientific tool. We also
`
`recognized that the existing regulatory
`systems had not been designed to cope
`with the new technology. The public’s
`concern reached such levels that some
`communities, most notably Cambridge,
`Massachusetts, passed ordinances regu-
`lating recombinant DNA research (/5).
`In Congress, several bills were intro-
`duced to regulate the research. Someof
`these would have subjected all recombi-
`nant DNA research, public or private, to
`federal regulation (/6). It was probably
`fortunate that none of the bills was en-
`acted into law, as former Representative
`Paul Rogers (D~Fla.) noted:
`‘‘]
`think
`Congress was right
`[in not regulating
`rDNA research]. Congress did a good
`service in airing the issue, but
`there
`wasn’t a necessity to pass a law” (/7).
`SCIENCE, VOL. 219
`
`Page 3
`
`

`

`too, adapted
`The regulatory system,
`well to this unexpected challenge to its
`flexibility. On 22 December 1978,
`the
`FDA hadpublished in the Federal Regis-
`ter a ‘‘Notice of Intent to Propose Regu-
`lations’? governing recombinant DNA
`work. But, by the time that the FDA’s
`Division of Metabolism and Endocrine
`Drug Products convened a- conference
`on the development of
`insulin and
`growth hormone by recombinant DNA
`techniques (//) in mid-1980, attitudes
`had changed, and the regulations were
`never promulgated.
`the containment of
`Concerns about
`potentially harmful organisms fell under
`the purview of the Recombinant DNA
`Advisory Committee (RAC) and the Na-
`tional. Institute of Occupational Safety
`and Health (NIOSH). RAC, established
`in 1974 by the secretary of HEW, had 11
`members, all of whom werescientists. In
`December 1978, 14 more members were
`added to RAC; all of the new members
`were nonscientists. It was apparent that
`the nonscientists would have to rely
`heavily on the scientists to develop their
`understanding of the new technology:
`Lilly scientists participated actively in all
`aspects of public discussion, throughtes-
`timony in both houses of Congress, par-
`ticipation in the open forum of the Na-
`tional Academy of Sciences (78), and in
`meetings of RAC, and by submitting-
`comments and amendments to NIH for
`its guidelines.
`In June 1979, Lilly made thefirst ap-
`plication to RAC for an exception to the
`rule limiting recombinant DNA work to
`10-liter volumes. At its meeting in Sep-
`tember 1979, RAC recommended that
`our request to scalé up production of
`bacteria-derived insulin be approved,
`and a month later the director of the
`NIH granted us permission to use 150-
`liter containers. In 1980, permission to
`expand to 2000-liter containers was
`granted. This was a major step toward a
`production type of operation;
`the sub-
`mission to RAC contained detailed engi-
`neering specifications for equipment and
`monitoring systems as well as descrip-
`tions of the proposed operating proce-
`dures. Because of the unprecedented
`volume increase in the handling of cul-
`tures of
`recombinant organisms,
`the
`scale-up request was preceded bya visit
`to our plant by a group consisting of
`RAC representatives and NIH officials;
`they cameto see for themselves how we
`could handle containment problems.
`With the experience gained at these in-
`termediate levels, we are now routinely
`using 10,000-gallon fermentors.
`Throughout 1980, there were several
`other positive developments. NIH pub-
`11 FEBRUARY 1983
`
`3.525
`
`4.700 -
`[
`be
`L
`P
`[
`r
`>»
`r
`=
`@ 2.350
`6
`L
`=
`L
`~
`
`AG-4)
`
`At5-12)
`
`
`
`AM18-21)
`B(14-21)
`™~
`
`
`A(13-17)
`i
`
`Semisynthetic
`
`human insulin
`
`
`
`B(22-30)
`
`> 100
`1
`4
`4
`:
`+ 75
`A(5-17)
`4
`|
`4
`B(1-13)
`A(S-21))
`508
`-21}
`LS
`BO 21
`o
`eee~~
`+ 256
`
`
`
`
`
`1.175
`
`9.000 Co
`1,00
`
`
`——~ Human insulin
`
`(recombinant DNA)
`at
`1
`boa a i
`mn
`i
`338.5
`676
`1013
`1351
`1688
`Seconds
`
`]
`
`J
`A]
`bo
`7°
`2363
`2701
`
`Po
`2026
`
`Fig. 3. HPLC chromatograms of peptide fragments from the A and B chains of semisynthetic
`human insulin and humaninsulin (recombinant DNA) after treatment with a specific staphylo-
`coccal protease. These chromatograms indicate the correctness of the disulfide bridges and the
`lack of any other major peptide components.
`
`lished in the Federal Register draft
`guidelines on physical containment rec-
`ommendations for large-scale uses of or-
`ganisms containing recombinant DNA
`molecules. This draft was not formally a
`part of the guidelines, but it did serve as
`a model for persons preparing submis-
`sions to RAC for large-scale fermenta-
`tions with
`recombinant
`organisms.
`About the same time, the NationalInsti-
`tute of Allergy and Infectious Diseases
`sponsored a workshop on risk assess-
`ment. Among the issues discussed were
`risks associated with pharmacological
`action of hormones from recombinant
`
`organisms populating the humanintesti-
`nal tract, medical surveillance of work-
`ers involved in large-scale fermentation
`
`Table 1. Amino acid compositions of human
`insulins. Molar aminoacid ratios with aspartic
`_ acid as unity [actual aspartic acid yields were
`160 nanomoles per milligram for human insu-
`lin (recombinant DNA) and 156 nanomoles
`per milligram for pancreatic human insulin
`0).
`
`Recom-
`Pan-
`Amino acid
`binant
`:
`
`DNA
`creatic
`
`Aspartic acid
`Threonine
`Serine
`Glutamic acid
`Proline
`Glycine
`Alanine
`Half-cystine
`Valine
`Isoleucine
`Leucine
`Tyrosine
`Phenylalanine
`Histidine
`Lysine
`Ammonia
`Arginine
`
`3.00
`2.77
`2.56
`7A
`1.03
`3.98
`0.97
`5.31
`3.76
`1.66
`6.16
`3.91
`2.99
`1.97
`0.97
`6.89
`1.00
`
`3.00
`2.77
`2.63
`7.10
`0.99
`3.98
`0.99
`5.43
`3.71
`1.61
`6.14
`3.90
`2.91
`1.99
`0.97
`6.95
`1.00
`
`of recombinant organisms, pathogenesis
`of approved recombinanthosts, and con-
`tainment practices in commercial-scale
`fermentation facilities. Most participants
`indicated that there waslittle or no risk
`involved in these practices. A few
`months later,
`the industrial practices
`subcommittee of the Federal Interagen-
`cy Advisory Committee (FIAC), a work-
`ing group of representatives from all the
`cabiret-level departments as well as all
`federal agencies that are in any way
`affected by recombinant DNA issues in-
`vited Lilly to make a formal statement.
`Bernard Davis of the Harvard Medical
`School and I submitted a document on
`the safety of FE. coli K-12, the reliability
`of commercial-scale equipment, opera-
`tor training, and other topics; this was
`favorably received. NIOSH also pub-
`lished a favorable report of its on-site
`inspection of Lilly Research Labora-
`tories’ recombinant DNA researchfacili-
`ties and procedures for large-scale fer-
`mentations of recombinant organisms.
`In July 1980, we beganclinical trials of
`our human insulin in the United King-
`dom. Within weeks, similar tests were
`under way in West Germany and Greece
`and, finally, in the United States. Plants,
`specifically designed for the large-scale
`commercial production of human insulin
`(recombinant DNA), were built at India-
`napolis (Fig. 4) and at Liverpool in the
`United Kingdom. On 14 May 1982, we
`filed our new drug application for human
`insulin with the FDA.
`Clinical studies with human insulin
`(recombinant DNA)indicateits efficacy
`in hyperglycemic control. It appears to
`have a slightly quicker onset of action
`than animal
`insulins.
`In double-blind
`
`transfer studies with animal insulins, pa-
`tients previously treated with mixed
`635
`
`Page 4
`
`Page 4
`
`

`

`PETE TL chess ea
`
`
`cet ae
`
`Fig. 4. The new production plant in Indianapolis for human insulin produced by recombinant
`DNA technology.
`
`beef-pork insulin had a 70 percent de-
`crease of bound insulin in comparison
`with a base line. Species-specific binding
`of human, pork, and beef insulin at 6
`months decreased by 61, 58, and 57
`percent, respectively. In patients previ-
`ously treated with pork insulin,
`the
`bound insulin decreased by 30 percent in
`control subjects treated with pork insu-
`lin, and by 51 percent in patients trans-
`ferred to humaninsulin. Species-specific
`binding of beef and humaninsulins de-
`creased equally whether patients were
`maintained on purified pork insulin or
`switched to humaninsulin. Species-spe-
`cific binding for pork insulin. however,
`remained constant in both groups (/9).
`The clinical importance of these findings
`remainsto be clarified in long-term stud-
`ies. Occasional patients hypersensitive
`to animal insulins and semisynthetic hu-
`man insulin derived from pork insulin
`tolerated human insulin (recombinant
`DNA) well. Recombinant
`technology
`now permits us to study humanproinsu-
`lin and mixtures of human proinsulin and
`insulin much as they are secreted by the
`beta cell. These studies may provide an
`improved modality of therapy in diabe-
`tes.
`
`The power of recombinant DNA tech-
`nology resides in its high degree of speci-
`ficity, as well as the ability it provides to
`splice together genes from diverse orga-
`nisms—organisms that will not normally
`exchange DNA in nature. With this tech-
`nology, it is now possible to cause cells
`to produce molecules they would not
`normally synthesize. as well as to more
`efficiently produce molecules that they
`do normally synthesize. The logistic ad-
`vantages of synthesizing human insulin,
`growth hormone, orinterferon in rapidly
`636
`
`dividing bacteria, as opposed to extract-
`ing these from the tissues in which they
`are normally produced, are obvious.
`We have shown the practicality of
`using -recombinant
`technology to pro-
`duce proteins of pharmacological inter-
`est as fermentation products. This was
`accomplished without adverse environ-
`mental impact or increased risk to work-
`ers. At this point it seems reasonable to
`speculate about the future of this new
`technology.
`
`Impact of the Technology on Industry
`
`A whole growth industry largely de-
`pendent on investor interest has devel-
`oped. Through newsletters, conferences
`to develop research strategies, market
`estimates, and so forth, these investors
`supposedly predict which projects will
`be brought to fruition through this new
`biotechnology. It is difficult to estimate
`the extent to which these prognostica-
`tions will reflect economic and scientific
`reality, but there are some items offact
`that appear to be supported by fairly
`simple logic.
`In the biomedical area there will cer-
`tainly be other proteins and peptides of
`pharmacological
`interest
`produced.
`Some of these are likely to result from
`new discoveries as additional genes are
`cloned. As an example, perhaps the most
`interesting aspect of the cloning of the
`interferon genes is that they represent a
`family of genes that code for a large
`number of interferons,
`leading to the
`possibility of producing hybrid mole-
`cules that have not been seen in nature.
`It seems unlikely that interferons should
`be unique in this respect among cyto-
`
`kines or other biologically interesting
`messengers.
`The technology will probably permit
`the mappingof the entire human genome
`during the next decade. Medical geneti-
`cists have laboriously mapped human
`genes by studying electrophoretic vari-
`ants or phenotypic expression of disease
`tracked through family trees. It is now
`possible to isolate individual human
`chromosomes on a preparative scale,
`followed by establishment of gene banks
`or libraries for each chromosome. The
`work should advance rapidly with an
`enormous potential impact on new medi-
`cal research and the understanding of
`human biology.
`In addition,
`it seems
`likely that eventually we will understand
`the mechanism of gene control and regu-
`lation which, combined with information
`now being unraveled concerning potent
`tumor-specific
`oncogenic DNA se-
`quences. clearly suggests major applica-
`tions in our understanding of oncology
`and differentiation. Consider, for exam-
`ple,
`the recent
`finding that
`the point
`mutation in a normal human gene that
`leads to the acquisition of transforming
`properties is due to a single nucleotide
`change from guanylate to thymidylate.
`This codon change results in a single
`aminoacid substitution of valine for gly-
`cine in the 12th amino acid residue of the
`T24 oncogene encoded p25 protein;
`it
`appears to be sufficient to confer trans-
`forming properties on the T24 human
`bladder oncogene (20).
`Assumptions can be made about appli-
`cations to agriculture as well. It seems
`incontrovertible that in some areas, for
`example, the amount of productive land
`is decreasing becauseofthefall of water
`tables and sometimes increasing salinity
`of ground water. Moreover, the number
`of people producing crops is decreasing
`while the population dependent upon
`them continues to increase. Recombi-
`nant
`technology.
`in combination with
`conventional plant breeding, plant cell
`culture, and regeneration, may well re-
`sult
`in the production of new plants.
`Such plants could increase the produc-
`tivity of existing farmland as well as
`permit farming on land currently consid-
`ered to be nonproductive. Equally im-
`portant applications are technically fea-
`sible in the animal husbandry area, and
`manyother types of applications—in the
`fermentation industry, industrial chemi-
`cals,
`environmental
`clean-up—have
`been suggested.
`Wecan certainly debate how rapidly
`these further developments will occur
`and whether ornot they will be economi-
`cally feasible. However. we mustall be
`SCIENCE, VOL. 219
`
`Page 5
`
`Page 5
`
`

`

`impressed with the speed with which the
`technology has progressed since 1974
`and can be confident that if we invest
`wisely,
`this rate will be maintained or
`even increased.
`
`References and Notes
`1. U.S. Department of Agriculture estimates, Live-
`stock and Slaughter Reports (Bulletin of Statis-
`tics 522, Economic, Statistic and Cooperative
`Services, Washington, D.C., 1980).
`2. National Jastitiies of Health Publ. 78-1588
`(April 1978), p.
`. D. A. Jackson, R H. Symons, P. Berg, Proc.
`Natl. Acad. Sci. U.S.A. 69, 2904 (1972).
`S.N, Cohen, A. C. Y. Chang, H. W. Boyer, R.
`B. Helling, ibid. 70, 3240 (1973).
`. U.S. patent number 4,237,224.
`. P. Berg et al., Science 185, 303 (1974).
`. P. Berg, D. Baltimore. Ss. Brenner, R. O. Roblin
`III, M. F. Singer, ibid. 188, 991 (1975).
`
`SDMSBWw
`
`8. B. Hartley, Nature (London) 283, 122 (1980).
`9. R. E. Chance ef al.,
`in Peptides: Synthesis-
`Structure-Function, D. H. Rich and E. Gross,
`Eds,
`(Proceedings of the Seventh American
`Peptide Symposium, Pierce Chemical Compa-
`ny, Rockford, IIl., 1981), pp. 721-728.
`10. R. E. Chance, E. P. Kroeff, J. A. Hoffmann, B.
`H. Frank, Diabetes Care 4, 147 (1981).
`11. ¥. S. Johnson,
`in Insulins, Growth Hormone;
`and Recombinant DNA_ Technology,
`J. L.
`Gueriguian, Ed. (Raven, New York, 1981), p.
`183
`12. S. A. Chawdhury, E. J. Dodson, G. G, Dodson,
`C. D. Reynolds, S. Tolley, A. Cleasby, in Hor-
`mone Drugs: Proceedings of the FDA-USP
`Workshop on Drugs and Reference Standards
`for Insulins, SOMatOnropins, and Thyroid-axis
`Hormones (U.S. Pharmacopeia, Inc., Rockville,
`Md., in press).
`13, R.S. Baker, J. M. Ross, J. R. Schmidtke, W. C.
`Smith, Lancet 1981-11, 1139 (1981).
`14. J. W. Ross, R. S. Baker, C. S. Hooker, I. S.
`‘Johnson, J. R. Schmidtke, W. C. Smith,
`in
`Hormone Drugs: Proceedings of the FDA-USP
`
`Workshop on Drugs and Reference Standards
`for Insulins, Somatotropins, and Thyroid-axis
`Hormones (U. S. Pharmacopeia, Inc., Rockville,
`Md., in press).
`15. Bull. At. Sci. 33, 22 (1977).
`16. 95th Congress, 2d sess. amended to $.1217;
`Calendar No. 334,H. Rep. No. 95359 (1977).
`17. A, J. Large, Wall ‘Street Journal, 25 January
`1982, p. 18,
`18.
`I. S, Johnson,
`in Research with Recombinant
`DNA, an Academy Forum (National Academy
`of Sciences, Washington, D.C., 1977), p. 156.
`I. S. Johnson, Diabetes Care 5 (Suppl. 2), 4
`19.
`(November-December 1982).
`20. E. P. Reddy, R. K, Reynolds, E. Santos, M.
`Barbacid, Nature (London)