An Introduction to Molecular Medicine and Gene Therapy - part 10 pot

DNA PRODUCTION AND QUALITY CONTROL Introduction The large-scale, commercial production and quality control of DNA and viralvectors to be used in clinical research protocols is critically

CONSIDERATIONS IN CHOOSING A TARGET DISEASE

FOR GENE THERAPY

A variety of approaches have been utilized for the introduction of nucleic acids(principally DNA) into cells These include the use of viral and nonviral methodsfor gene delivery Modified retroviruses, adenovirus, adenoassociated virus (AAV),and herpes virus have been investigated for virally based delivery (see Chapter 4).Naked DNA, cationic lipids, liposomes, and cationic polypeptides are being pursued

as nonviral approaches for gene therapy (see Chapters 4 and 5) Matching a genetherapy methodology to a target disease involves a number of factors The techni-cal issues that must be considered include determining the tissue and cell specificityneeded for expression of the therapeutic gene, the number of cells that need to betargeted, and therapeutic level and duration of transgene expression The deliveryvehicle identifies the tissues and cell types that the therapeutic DNA can be deliv-ered If the choice of delivery vehicles is limited, then target diseases or genes willalso be limited The delivery vehicle will dictate the number of cells targeted andthe duration of expression of the transgene (therapeutic gene) Therapies thatrequire high levels of gene expression or require targeting a large percentage ofcells likely require viral delivery vectors rather than nonviral delivery vectors This

is because, at present, viral vectors are more efficient at delivery The delivered genemay be integrated into the host chromosome using AAV or retrovirus vectors (seeChapter 4) These may give longer duration of expression of the transgene thanwould be expected with adenovirus or nonviral delivery vectors However, if thegene is to be delivered multiple times during the course of treatment, nonviralvectors may avoid the development of immune responses that can occur with viraldelivery systems

Regulation of the therapeutic gene is another factor to consider when choosing

a target gene How gene expression is regulated may determine which and howmany cells need to be targeted At present, gene expression regulated at the level

of transcription is less problematic than gene expression regulated tionally Posttranscriptional gene regulation is, in most cases, less well understood.The consideration of posttranscriptional regulatory mechanisms could complicate

posttranscrip-or slow the development of gene therapy As discussed later, the levels of transcription of a gene can be manipulated by modification of the plasmid or viralvector DNA

Unusual requirements for gene product processing needed for activity of theexpressed gene must be considered when choosing a target disease Many genes can

be expressed in cell types other than the normally expressing cell types and still betherapeutic However, other gene products require special processing in a particu-lar cell type or in a particular organelle Thus, such genes would not be effectivelyexpressed in other cell types Still other proteins may have cofactors (proteins) thatare essential for activity and must be made in close proximity (same cell ororganelle) as the cofactor

Another key factor in choosing a target gene is the availability of the gene.Questions that should be asked are:

• Is the gene sequenced and cloned?

• Is it a cDNA clone or full-length gene (containing introns)?

• Does the cloned gene also contain the native promoter and regulatorysequences at the 5¢ and possibly 3¢ ends?

The commercial development process is faster when maximal information isknown about a targeted gene (regulation, sequence, etc.) The overall size of thegene to be delivered is also an important consideration since many viral vectors arelimited in the size of DNA that can be packaged The nonviral delivery systems areless restricted in the size of DNA that can be delivered

The development of a commercial gene therapy product is also facilitated by theavailability of an animal model of the genetic disease being targeted Although notall human genetic diseases currently have animal models of disease, the number oftransgenic and knock-out mouse strains (see Chapter 3), as well as larger animalmodels, has increased exponentially in the last few years These animal models provevaluable in developing effective gene therapy treatment approaches for manysingle-factor genetic disorders and possibly some multifactor diseases as well

As for any commercial venture, patent and licensing issues for a particular genewill necessarily be important factors in choosing a target The size of the potentialpatient population and the accessibility of patients for a particular product are alsocrucial There are numerous genes that could be targeted for gene therapy, however,many of the single-factor genetic diseases are relatively rare (see Chapter 1) Dis-eases currently treated with recombinant proteins (severe immune deficiency, hemo-philia A and B) provide larger markets where gene therapy could have an impact

As with any new therapy, gene therapy approach for a disease state would need tohave advantages over treatments currently in use

DNA PRODUCTION AND QUALITY CONTROL

Introduction

The large-scale, commercial production and quality control of DNA and viralvectors to be used in clinical research protocols is critically important Assurance ofpurity must be provided to investigators who purchase or contract for reagents to

be used in basic or clinical research As can be seen from the recent events, poorquality control of reagents can lead to the cessation of clinical trails of gene therapyprotocols (see Chapter 13)

Laboratory Scale Purification

As the clinical aspects of gene therapy continue to grow, one of the challenges facingindustry is the large-scale purification of plasmid DNA Within the typical researchlaboratory, plasmids continue to be routinely obtained by the standard method

of CsCl–ethidium bromide density gradient ultracentrifugation CsCl–ethidiumbromide gradients are popular since large numbers of different plasmid prepara-tions can be processed simultaneously The approach applies to both plasmid andviral DNA of varying sizes; and a single band in the density gradient contains themonomeric, supercoiled form of the DNA partially resolved from the intrinsic hostcell contaminants (protein, DNA, and RNA) But there are numerous drawbacks

DNA PRODUCTION AND QUALITY CONTROL 357

and limitations to this process For the researcher at the lab bench, it is time suming, labor intensive, and expensive For the biotechnology company, how-ever, this method is completely unacceptable for the production of clinical-gradematerials because of its use of mutagenic reagents and its inherent inability to be aprocess of scale.

con-Recently, a number of companies have initiated market-adapted preparative methods for the production of larger quantities of plasmid DNA Thesemodified “mini-prep” kits, make use of the alkaline lysis method for cell disruptionfollowed by a chromatographic cartridge purification The composition of the sta-tionary phase used in these kits varies Some kits use a silica-based stationary phase,while others are based on an agarose stationary phase In most cases, the mecha-nism of binding is anion exchange These kits are aimed at a particular market niche:the production of small quantities (milligram or less) of research-grade material formolecular biology applications They do not meet the rigorous requirements for thedevelopment of a highly controlled drug manufacturing process and most do nothave a Drug Master File (DMF) Purity in these applications is usually evaluated

micro-by agarose gel electrophoresis Trace impurities such as endotoxin and host DNAare not as thoroughly investigated as needed for human clinical use

The Food and Drug Administration’s (FDA’s) “Points to Consider” on plasmidDNA was drafted in October, 1996, and provides the U.S approach to regulation ofplasmid preventative vaccines (see Chapter 13) The same general criteria that guidethe manufacture of recombinant protein pharmaceuticals apply to the development

of processes for the production of plasmid DNA for human clinical investigations.The common thread linking these processes is the basis of well-documentedresearch This basis allows for the final product to meet defined quality standardssupported by validated analytical methods and controlled unit operations All com-ponents of the process must be generally recognized as safe and must meet allapplicable regulatory standards It is precisely because of these reasons that plasmidDNAs used for clinical investigations are not produced using kits intended for laboratory research

LARGE-SCALE PRODUCTION: AN OVERVIEW

To proceed to advanced clinical trials and ultimately gain regulatory approval, thepharmaceutical development of gene therapy products need to meet the require-ments for cGMP (current good manufacturing practices) production While the “c”ostensibly stands for “current,” when actually following the spirit and intent of the FDA guidelines the “c” represents control of the process and characterization

of the product GMP is defined as “the part of quality assurance that medicinal products are consistently produced and controlled to the quality standards appro-priate to their intended use and as required by the Marketing Authorization orproduct specification.”

There are two main components of cGMP, comprising both production andquality controls Production control is concerned with manufacturing This includesthe suitability of facility and staff for the manufacture of product, development ofstandard operating procedures (SOPs), and record keeping Quality control is con-cerned with sampling, specifications, testing, and with documentation and releaseprocedures ensuring satisfactory quality of the final product

For a typical production conducted under the principles of cGMP, the majorpoints to consider in the manufacture of plasmid DNA are:

• SOPs are in place to ensure the control and consistency of the entire tion cycle starting from the initial receipt of raw materials to the final formu-lated drug product

produc-• All raw materials used in the manufacturing process are put on a testing

program based on the U.S Pharmacopeia.

• A master cell bank (MCB) and manufacturer’s working cell bank (MWCB) hasbeen prepared under conditions of quarantine to ensure the purity and iden-tity of the fermentation seed pools Thorough vector characterization has beencarried out, including a detailed history on the construction of the vector, com-plete nucleic acid sequence determination, and plasmid stability within the hoststrain MCBs should be shown to be free of adventitious agents

• Details of the fermentation process must be elucidated and consistency datagenerated Several commercial media have been designed for plasmid produc-tion, but a defined medium that has been empirically developed for a specificstrain plasmid is preferable This should assist in achieving a reproducible well-controlled process Bacterial strains should be compatible with high copynumber plasmids, high biomass fermentations, and the selection system cannot

be ampicillin based

• Purification processes must be developed to meet the challenges inherent with

a high cell density fermentation process Plasmid DNA purification kits tinely fail when challenged with high cell density starting feed streams Recov-ery and purification must be controlled and validated Special attention must

rou-be paid to the removal of host cell proteins, DNA, and endotoxin Documentedreproducible removal of key host-cell-derived impurities is essential for settingaccurate limits and specifications on the bulk drug product

• Appropriate analytical assays must be developed for both the monitoring ofthe production cycle as well as for final quality control release criteria TheFDA’s “Points to Consider” lists some of the tests needed to confirm purity,identity, safety, and potency of plasmid DNA A functional in vivo or in vitrobioassay that measures the biological activity of the expressed gene product,not merely its presence, should be developed Measuring the relative purity andconcentration of plasmid DNA by agarose gel electrophoresis or by high-pressure liquid chromatography (HPLC) is only a small part of the battery ofanalytical measurements necessary to confirm product quality All assays must

be fully validated

• Ongoing stability and efficacy testing must be conducted on the product insupport of the ongoing clinical trials This data is critical in eventually deter-mining product shelf life for the approved drug

LARGE-SCALE PRODUCTION: THE PURIFICATION PROCESS

The basic unit operations for the manufacture of plasmid DNA are basically thesame as those for the production of any recombinant biopharmaceutical (see Fig.A.1) Typical process steps for the production of plasmid DNA include initial vector

LARGE-SCALE PRODUCTION: THE PURIFICATION PROCESS 359

design, fermentation, cell harvesting, alkaline lysis precipitation, chromatographicpurification, formulation, and filling The process cannot rely on the use of animal-derived enzymes such as lysozyme, proteinase K, and RNAase Use of these reagents

in any manufacturing process for a drug substance raises regulatory concerns aboutresiduals in the final product Disregarding such purity issues would increase thedifficulty in process validation and ultimately putting final regulatory approval atrisk The process should also not include toxic organic extractions The various forms

of plasmid DNA including supercoiled, relaxed, and concatamers should be rated The final product must be free of contaminating nucleic acids, endotoxins, andhost-derived proteins

sepa-Fermentation is generally considered the starting point in designing the cation process By careful selection and control of the variables associated with thefermentation process, the subsequent purification may be greatly simplified Variousfermentation feed strategies (batch, fed-batch, continuous) should be explored.While somewhat more difficult to optimize, as well as document, continuous fer-mentations may offer several advantages in terms of production cycle times Nor-mally, fed-batch fermentations allow quicker process development times, simplerprocess control and sufficiently high biomass The growth stage at which the fer-mentation is harvested must also be tightly controlled since it will greatly impact onthe final yield of purified plasmid Harvesting too early will result in suboptimal finalyields Harvesting too late in the fermentation cycle will not only result in low yieldsbut also plasmid of poor quality The optimal stage of harvest is late log phase.The monitoring of fermentation process parameters including temperature,glucose addition, dissolved oxygen, and carbon dioxide evolution are critical for thedevelopment of a reproducible process By manipulation of these parameters orthrough the use of an inducible plasmid system, the growth characteristics of a straincan be effectively changed, resulting in an increase in the plasmid-to-biomass ratio.Any increase in this ratio will aid in the design of the purification process As well,

purifi-it can result in higher final yields of plasmid Chloramphenicol has been tradpurifi-ition-ally used just for this purpose

tradition-The host cell and plasmid are the most important starting materials in the duction fermentation The key parameters in choosing a host strain are a lowendogenous endotoxin, the capability of growing to high biomass, and relevant

pro-genotypic markers These markers could be recA1, endA1, and deoR: recA1

pre-Cell culture MCB/MWCB Shake flask Fermentation

FIGURE A.1 Steps in a typical large-scale biotechnology process.

vents recombination and improves stability of plasmid inserts; deoR allows for the uptake of large plasmids; endA1 improves plasmid quality The plasmid should be

structurally as well as segregationally stable and have a high copy number origin ofreplication Typically it is pUC derived

Scale-up for the purification of plasmid DNA is a definite issue phy is the tool that has enabled the biotechnology industry to achieve the puritylevels required for today’s biotherapeutics, diagnostics, and other biologicals Theseinclude enzymes and plasma products Chromatographic purification of DNA pre-sents a novel set of problems These are based on the physical characteristics of thebiomolecule as well as the intrinsic impurities derived from the host cell of choice,

Chromatogra-Escherichia coli The chief culprits that hinder the purification of plasmid DNA are

the large amounts of polymers of similar structure (chromosomal DNA and RNA)and high levels of endotoxin

Plasmid DNA is a highly anionic polymer that is sensitive to shear and to dation by nucleases Plasmids are as large or larger than the pores of almost all chro-matographic resins Several chromatographic procedures for the purification of biologically active plasmid DNA (without the use of CsCl–ethidium bromide ultra-centrifugation) have been developed, at least at laboratory scale They include gelfiltration chromatography, hydroxyapaptite chromatography, acridine yellow affin-ity chromatography, anion exchange chromatography, reversed phase chromatog-raphy, silica membrane binding, and binding to glass powder Unfortunately, many

degra-of these methods are not well suited to the purification degra-of large quantities degra-of DNA

In choosing the method of purification for large-scale production of plasmid DNA,there is a most important physical characteristic of the biomolecule to consider It

is that DNA is a highly anionic polymer that is sensitive to shear and to tion by nucleases Any large-scale manufacturing process must address all of thesecharacteristics Currently, the most successful methods of extraction and purifica-tion involve large-scale alkaline lysis in sodium deodecyl-sulfate (SDS) This stepefficiently removes chromosomal DNA, nuclease enzymes, and other contaminants.Therefore, cell lysis conditions must be carefully optimized Low shear mixing mustalso be used during this step Large-scale tangential flow systems, which are rou-tinely used for the processing of recombinant proteins, can easily nick the super-coiled form of the plasmid Cross flow rates, pump design, as well as the mixer’simpeller design must all be carefully scrutinized Plasmid extracts are primarily con-taminated with low-molecular-weight cell components, process chemicals, and RNA.These contaminants and trace host protein contamination may be removed by acombination of selective precipitation, anion exchange chromatography, and a finalpolishing step

degrada-A major drawback of using anion exchange chromatography as the sole resolution purification step in the purification of plasmid DNA is that a portion ofendotoxin and pyrogen contaminants will co-purify with the plasmid Given the lim-itations of currently available commercial matrices and the similar structure andcharge profile of biomolecule species passing over the column, anion exchange chro-matography is best used as a primary capture and initial purification step A secondpolishing step, which is orthogonal to the principles of anion exchange, is prudentand ensures rigorous process control

high-Historically, gel filtration has been used in the biotechnology industry as a ishing step Plasmid DNA, host cell DNA, and endotoxin resolve using gel filtration

pol-LARGE-SCALE PRODUCTION: THE PURIFICATION PROCESS 361

chromatography This is a simple and reproducible method that also offers theadvantage of simultaneously incorporating a buffer exchange step within the chro-matographic process Contaminating salts and/or residual metals can thus beremoved allowing for the careful control of the counter ion in the final drug product.However, the main drawback in using gel filtration is that it is a very slow andvolume-dependent method It is not a high throughput method and often becomesthe bottleneck within a given process.

Reversed phase chromatography (RPC), on the other hand, can also offer excellent separation and resolution of trace contaminants as well as the removal

of endotoxin It is commonly the method of choice for the purification of small pharmaceutical compounds When purifying biologically active molecules, care must

be taken so that biological activity is retained Through its use of volatile solvents,RPC can also serve the function of a buffer exchange step But it is precisely thispoint that contributes to reversed phase’s own set of unique problems The use ofcombustible organic solvents (acetonitrile or ethanol) requires explosion-prooffacilities This safety factor can dramatically increase the cost of waste management.With the heightened awareness of environmental issues in today’s industrial nations,the cost and feasibility of waste disposal are major considerations when designing

or deciding on a purification process Ion-pair RPC, while again providing excellentseparation, resolution, and endotoxin removal, introduces ion-pair reagents thatmust be assayed for in the final product Their removal must be assured by validatedmethods

The final crucial aspect in deciding on a chromatographic support is the sity of cleaning in place and sanitization by cycles of caustic washing The ability to

neces-withstand repeated cycles of regeneration, sterilization, and sanitization with 0.5 N

NaOH while maintaining run-to-run reproducibility of the column profile is animportant consideration in manufacturing pharmaceutical-grade plasmid DNA inaccordance with cGMP manufacturing guidelines

LARGE-SCALE PRODUCTION: QUALITY CONTROL

Recombinant proteins and plasmid DNA are both derived from E.-coli-based

expression systems This results in a fair degree of similarity in their contaminantprofiles The FDA has presented a general list of contaminants that should be quan-tified in all biopharmaceutical products They include pyrogen, nucleic acid, antigen,and microbial and residual contamination Most assays that have been developedfor the quality control of recombinant protein drug substances need only slightadaptation for the quality control of plasmid DNA production The most challeng-ing assays in terms of unique or specific analytical tests for plasmid DNA bulk prod-ucts are the measurements of protein (antigen), nonplasmid DNA, and RNA tracecontaminants Documentation and validation of all assays must adhere to cGMPguidelines

Fermentation cultures need to be routinely monitored for microbial tion Sterility checks should be performed on inoculation flasks, the fermentor, andthe fermentation media The presence of contaminating organisms will alter the production levels of plasmid produced and thereby invalidate data on the levels ofcontaminating impurities within the final DNA product

contamina-The most common and routine analysis of plasmid DNA is through the use ofethidium-bromide-stained agarose gels In research settings, this assay is usuallyused as a standalone technique for determining RNA contamination, residualgenomic DNA, as well as quantifying the relative amounts of supercoiled plasmid

in relation to the relaxed or nicked form It is well known, though, that ethidiumdifferentially stains linear, nicked, and supercoiled plasmid DNA as well as host cellRNA Thus, care must be used when using this assay as the sole tool for judging relative amounts of DNA or in determining residual RNA levels To accurately characterize the purified product (and monitor in-process samples) an array of electrophoretic, chromatographic, and spectrophotometric assays should beemployed In particular, the use of analytical high-resolution HPLC can avoid thedetection and quantitation problems associated with ethidium bromide staining ofplasmid DNA since detection is based on ultraviolet absorption

Another common quality control test for plasmid DNA used in most researchlaboratories is the A260/A280absorbance ratio assay It highlights the discrepanciesbetween true cGMP production and laboratory-scale purification The test was originally designed to measure enzyme concentrations in the presence of low levels

of nucleic acid contamination The original usage has been corrupted, however, andnow it is routinely used in molecular biology laboratories to assess DNA purity An

A260/A280absorbance ratio of 1.8 to 2.0 is generally considered “pure.” In fact, whenone does the actual calculation using the true extinction coefficients of nucleic acidsand proteins (nucleic acids have extinction coefficients on the order of 50 timeshigher than proteins), it becomes obvious that an A260/A280= 1.8 can contain as much

as 60% protein contamination Therefore, this method can only be used as a tional test and cannot in itself be used to determine DNA purity

func-Equally critical for achieving pharmaceutical-grade plasmid DNA is the toring of any chemical reagents introduced into the manufacturing process Ifalcohol is used in a precipitation step in the process, an assay must be included todetermine the residual trace levels of alcohol that remain in the final product Ifantifoam (a common fermentation additive) has been used, an analytical assay must

moni-be in place for its determination as well as a final release specification for its centration Choosing the appropriate analyses in this area requires careful controland sourcing of all raw materials One of the hallmarks of a fully FDA-compliantproduction process is the use of well-characterized reference standards These arenecessary for the completion of analytical assay assessment and for use in ongoingvalidation studies The most critical reference standard is the plasmid DNA Ideally,the plasmid should be fully characterized and be derived from a manufacturingbatch that has been clinically evaluated Having a well-characterized reference stan-dard greatly aids in the successful evaluation of product stability testing

con-With the proper appropriate supporting data, background information and plementary studies, the development of a minimal panel of characterizing assays can

sup-be put in place These would provide the necessary level of confidence to reliablydetermine identity, purity, potency, and stability of the manufactured plasmid DNA(Table A.1) The foundation that makes this possible is rooted in the compliance tocGMP throughout the entire plasmid production cycle While there may be differ-ences in the specific physiochemical assays for the determination of identity andpurity, pharmaceutical-grade plasmid DNA and recombinant-protein-based thera-peutics share a similar quality control characterization strategy

LARGE-SCALE PRODUCTION: QUALITY CONTROL 363

TECHNOLOGY ADVANCEMENT

Techniques for Profiling Proteins and mRNAs

Two general approaches can be used to take a census of the proteins in a cell ortissue: direct analysis of the proteins and indirect analysis of cDNAs reversely tran-scribed from mRNAs The advantages of the former are straightforwardness andthe ability to detect protein modifications The advantages of the latter are sensi-tivity and the ability to tap into the awesome power of molecular genetics throughDNA databases At the moment, nucleic-acid-based techniques are more widelyused However, techniques for analyzing proteins are rapidly advancing

Galactic-Scale cDNA Techniques Methods for analyzing cDNA populationsreversely transcribed from mRNAs are expanding in number and variety They areroughly quantitative because cDNAs are synthesized in proportion to the amounts

of individual mRNAs present in the population A partial list of cDNA methodsincludes differential display, direct sequencing and counting of “tags,” and hybridiza-tion to DNA arrays Selected methods are briefly described below to illustrate thechoices available The latest techniques allow the expression patterns of thou-sands of genes to be monitored simultaneously, generating a rough outline of the

“transcriptome”—the complete set of genes expressed in a particular cell Thesepowerful techniques are generating a tsunami of information They promise to revolutionize the way biology is studied and the way drug development is carriedout

Differential Display In old-fashioned differential display methods, cDNAs areprimed at the 3¢ end of mRNA, with one of three oligo-dT (A, C, G) oligonu-

TABLE A.1 Sample Plasmid Specifications and Test Methods

electrophoresis Anion exchange HPLC

by 1% agarose gel electrophoresis

Protein contamination Optical density scan l min = 230 nm

reporter gene

cleotides, and are anchored at the 5¢ end by a specific (but arbitrary) primer In modified versions of that method, other types of primers, such as oligonucleotidesoptimized to detect coding sequences, are used In either case, about 400 cDNAbands can be resolved by DNA gel electrophoresis in ultrathin polyacrylamide gels.Typically, cDNAs from two populations of mRNAs are compared to each other Toidentify the differentially expressed genes, cDNAs are recovered from gels, cloned,and sequenced.

SAGE (Serial Analysis of Gene Expression) SAGE is based on two principles:(1) that a gene can be identified by a short sequence tag (9 to 11 bases long) pro-vided that a second piece of information about the tag sequence is known, such asits position in the mRNA relative to that of another short sequence, for example, arestriction site and (2) that many tags can be concatenated into a single moleculeand sequenced to determine the abundance of each tag and to identify the gene (if

it is present in a database, such as GenBank)

Using this method, more than 300,000 cDNA tags representing a minimum of45,000 different genes have been examined and compared with mRNA populations

in human intestinal tumor cells and control cells Contrary to expectations, two

widely studied oncogenes, c-fos and c-erb3, were expressed at much higher levels in

normal colon epithelium than in colorectal cancers Such surprising results size the value of studies carried out on clinical samples (patient specimens) Thisapproach provides insight into the gene expression patterns of human malignancyand helps to identify genes that may be useful targets for gene therapy

empha-Expressed Sequence Tags (ESTs) and DNA Arrays ESTs provide gene tifiers.” They are made by copying mRNA populations into cDNA clones, which arethen partially sequenced and entered into databases About 800,000 ESTs of humangenes are currently available in public databases and at various Web sites (see alsoChapter 15) These represent 40,000 to 50,000 of the estimated total of 70,000 to100,000 human genes ESTs can be expressed as DNA molecules and hybridized tocDNAs under a variety of conditions Recently, ESTs have been combined into high-density DNA microarrays (Figure A.2) DNA microarrays consist of thousands ofindividual gene sequences attached to a surface in a precise and reproduciblepattern Because these arrays are minute, they can be used with tiny quantities ofcDNA

“iden-In a tour de force that paints a scintillating and detailed picture of a eukaryoticcell’s inner workings, microarrays were used to study changes in yeast gene expres-sion during the shift from anaerobic to aerobic metabolism (Fig A.3) This studyshowed the feasibility of a small group of researchers to PCR amplify more than 6000

open reading frames (ORFs), representing all the genes of Saccharomyces cerevisiae,

in about 4 months In two days, 110 microarrays, each containing 6400 elements, wereproduced and were ready for hybridization to fluorescently labeled cDNA copiedfrom mRNAs extracted from cells at various time points According to the in-vestigators, preparation of fluorescently labeled cDNA probes, hybridization, andimagine analysis proceeded quickly When analyzed, the data revealed that duringthe metabolic shift, mRNA levels for approximately 710 genes were induced by atleast a factor of 2, and the mRNA levels for approximately 1030 genes declined by afactor of at least 2 Since the equipment for DNA microarrays was chosen for its rel-

TECHNOLOGY ADVANCEMENT 365

atively modest cost, it may be feasible for small academic groups strongly committed

to comprehensive expression analysis to establish this technique for local use

Gene Expression Microarray (GEM) This technique is useful for probingexpression patterns in human cells Their microarrays contain PCR products over

100 bases in length, which permits the use of stringent hybridization conditions.GEM’s two-color competitive hybridization process is reported to detect twofoldchanges in the level of expression The first commercial product, a chip that con-tains tags from 10,000 human genes, is being replaced by a microarray containingtags for 55,000 genes Although the 3-year subscriptions are expensive—from

$300,000 to $9,000,000 depending upon usage—the establishment of academic laborations will reduce costs In addition, several additional commercial hybridiza-tion formats are available for “expression profiling.”

col-FIGURE A.2 Molecular analysis of the shift from anaerobic to aerobic metabolism using

a yeast genome microarray Fluorescently labeled cDNAs were prepared from mRNA from anaerobic cells by reverse transcription in the presence of Cy3-dUTP and from aerobic cells

in the presence of Cy5-dUTP Hybridization of the Cy3-dUTP-labeled cDNA appears in green and that of the Cy5-dUTP-labeled cDNA appears in red Genes up-regulated after the metabolic shift appear in red, while those down-regulated appear in green Genes expressed

at roughly equal levels appear in yellow The actual size of the microarray is 18 by 18 mm The image was obtained with a scanning confocal microscope (Reproduced with kind

permission from Science, 278, 1997.)

High-Resolution Two-Dimensional (2D) Gel Electrophoresis of Proteins, and ProteinChip Microarrays Efforts to define the “transcriptome” are paralleled

by efforts to define the human “proteosome,” the total set of proteins within a ticular cell Direct information about cellular proteins is needed for two reasons.First, protein function is often altered by posttranslational modifications, whichcannot be discerned from mRNA analyses Second, mRNA concentrations andprotein concentrations do not have a strict one-to-one relationship To investigatethe magnitude of the disparity, investigators compared mRNA and protein abun-dance in extracts of human liver They found that of the 50 most abundant livermRNAs, 29 encoded secretory proteins; none of the 50 most abundant proteinsappeared to be secretory.The correlation coefficient of RNA and protein abundancewas only 0.48 These results underscore the need to ensure that cDNA-basedmethods sample both high and low abundance mRNAs and demonstrate the value

par-of protein analyses

High-Resolution 2D Gels High-resolution 2D gels separate proteins according

to charge (in the first dimension) and size (in the second) Scanned images can beanalyzed by computer programs Small changes in the concentration of individualproteins (±15%) can be quantified This sensitivity may be important when measuring the effects of medical interventions In addition, these gels can detectposttranslational protein processing events, including proteolytic cleavage and

TECHNOLOGY ADVANCEMENT 367

FIGURE A.3 Two-dimensional gel master image of liver proteins The image is available from the URL http://expasy.hcu.ge.ch/cgi-bin/map1 Two-dimensional spots can be selected and the annotation linked to the spot displayed The spot marked with the white cross is cata- lase The annotation indicates how it was identified (by immunoblotting) and how it alters

in disease states (decreased in acatalasia) and provides a futher link to the entry in the SwissProt database (Reproduced with kind permission from Biochemical Biophysical Research Communications, 231, 1997.)

phosphorylation The introduction of immobilized pH gradient electrophoresis hasextended the pH range and improved reproducibility Progress is being made towardthe development of fully automated 2D systems A World-Wide-Web (www) feder-ation is being established to facilitate the exchange of 2D gel images A commoninterface for data accession already allows 2D gel databases from all www federa-tion sites to be searched and may soon allow gel images to be matched over thenetwork The SWISS-2DPAGE database (Fig A.4) contains 2D master gel images

of cells representing the “normal” state, while the complementary SWISS-2Diseasedatabase consists of annotated gel images from cells and tissues of various diseasestates, such as renal failure and myeloma Both are available on www Typical 2Dgel patterns stained with silver contain about 1000 to 2000 spots, about 75% of whichcontain less than 500 femtomoles of protein Large-format gels allow up to 10,000spots to be detected

High-resolution 2D protein gels are used in combination with various techniquesfor identifying the proteins comprising the spots, such as microsequencing and anti-body binding Furthermore, new “soft” ionization techniques in mass spectrometryare creating a revolution in spot identification Matrix-assisted laser desorption and ionization (MALDI) and electrospray ionization allow minute quantities ofprotein to be analyzed The powerful new techniques for obtaining reproducible gelpatterns, for analyzing the gel images, and for identifying individual proteins areyielding detailed snapshots of cellular protein populations

Microarrays (Chips) This technology for protein analysis is being developed.Chip technology will examine protein expression and structure The ProteinChipmicroarrays are comprised of molecules that bind proteins, such as antibodies,receptors, or ligands Cellular proteins are incubated with the array, and then laserpulses are used to probe each site Proteins can be released from the surface of thechip and analyzed by mass spectrometry

FIGURE A.4 Effect of overexpression of antioxidative enzymes on life span and protein

oxidative damage in D melanogaster Survival curves (solid) and protein carbonyl content

(dashed curves) at different ages for a control group (blue triangles) and three different lines

(remaining symbols) of transgenic D melanogaster overexpressing both Cu,Zn-superoxide

dismutase and catalase (Reproduced with kind permission from Science, 273, 1996.)

Public-Access, Systematic Database of Functional Genomics Informationabout gene expression patterns is accumulating at a blistering pace In the study ofyeast gene expression, it is perhaps the greatest challenge to develop efficientmethods for organizing, distributing, and interpreting the large volume of data gen-erated with microarrays and related technologies.Adequate methods for storing andanalyzing these data are essential Something akin to a four-dimensional GenBankneeds to be constructed: (1D) primary sequence of the mRNA or protein, (2D) post-translational modifications, (3D) concentration, and (4D) temporal changes Thegene expression database will need to be densely annotated with information aboutthe physiological state of the organism, tissue, or cell and about experimental treat-ments Listing of publicly available bioinformatics resources, which give a usefulstarting point, are currently available In the long-run, the information about expres-sion patterns will need to be correlated with information about the genetic hetero-geneity of the human population.

Elucidating the Human Genome Through DNA Analysis

The human genome is the nuclear composition of genetic material It is estimatedthat 50,000 to 100,000 genes encompass the human genome Elucidation of the com-position of the human genome would result in information and spin off technolo-gies that would revolutionize the study of disease A map or descriptive diagram ofeach human chromosome would involve dividing each chromosome into smallerfragments that can be propagated and characterized as well as ordering the frag-ments to correspond to their location on the chromosome After mapping, the nextstep is to determine the nucleotide sequence of each chromosome fragment Theultimate goal would be to locate and assign a biological function to all the genes inthe DNA sequence

A research effort, entitled the Human Genome Project, is an international effortdesigned to construct detailed genetic and physical maps of the human genome,

to determine the nucleotide sequence of human DNA, to localize all the genes ofthe human genome, and to perform a similar analysis on the genomes of severalother organisms as model systems Ninety percent of the human genome nucleotidesequencing project has been completed Most recently, a highly public announce-ment was made that the human genome has been sequenced Chromosome 22 hasbeen fully characterized and the data put in the public domain The elucidation

of the entire genome and public disclosure of the data would almost certainly identify most if not all of the major genes involved in common diseases Already, asnoted above, correlations between genetic mutations and disease susceptibility arebeing established with the hope that such information will lead to novel therapiestargeted at defined patient populations Genetic based diagnosis and treatment ofdisease have the potential to radically improve the practice of medicine (seeChapter 15)

Human Genome Project The Human Genome Project began in the 1980 as part

of a national scientific research effort supported by the Department of Energy.Recently, an institute, The National Human Genome Research Institute (NHGRI),has been created at the National Institutes of Health to lead this research effort The scientific priorities of the NHGRI can be broken down into the areas

TECHNOLOGY ADVANCEMENT 369

of (1) genetic mapping of the human genome, (2) physical mapping of the human genome, (3) DNA sequencing of the human genome, (4) new technologiesfor interpreting human genome sequence, and (5) analysis of genomes of modelsystems.

The Human Genome Project is driven by technology As mentioned previously,new techniques are constantly being developed and introduced into the researchenvironment A long-term objective of the Human Genome Project is to identifyall coding sequences, including genes and regulatory elements in the humangenome This once unimaginable goal is now feasible through new methods of DNAanalysis

Superfast DNA Sequencing Faster methods for DNA sequencing are beingdeveloped Speed is increasing as a result of both incremental improvements incurrent methods and from the introduction of entirely new approaches, such as

“sequencing by hybridizing” on microarrays or chips In July, 1997, the first of suchsystems was launched and called GeneChip (p53) assay for research applications.This assay is capable of analyzing the full-length coding sequence of the human p53tumor suppresser gene, frequently mutated in human cancers The chip contains amatrix of more than 50,000 DNA molecules and is designed to detect more than

400 distinct mutations in the p53 gene Success of the p53 chip is certain to spawnsimilar chips for heritable diseases, such as cystic fibrosis, which can result from any

of a number of mutations within the CFTR gene.

In addition to sequencing by hybridization, entirely different approaches arebeing explored For example, the possibility of sequencing DNA molecules by mea-suring effects on conductance as DNA passes through ion channels in membranes

is being explored Many additional initiatives are underway with a major through in this field likely It is unclear the form that superfast DNA sequencingwill take, but most would agree that it will depart from the current techniques asdramatically as the current techniques depart from the depurination fingerprintingmethod that started the field

break-Making use of human genetic information will be challenging Genetic ologists and statisticians will be needed to validate genetic linkages in conditionswith polygenic inheritance patterns Information needs to be gathered about thefactors affecting expression of genes and gene mutations, so that predictions can bemade about the outcome in particular individuals

epidemi-Model Systems The mapping and sequencing of the genomes of other organisms

as model systems is fundamental to the elucidation of the human genome Modelsystems are also useful in the testing of new technologies to be applied to the human

genome Currently, five organisms are being used as model systems: E coli terium), S cervisiae (yeast), C elegans (round worm), D melanogaster (fruit fly) and

(bac-M musculus (the laboratory mouse) The physical map and genomic sequence of

E coli, S cervisiae, and C elegans are completed The current goal of the project is

to complete a genetic map, an STS content map of 300 kilobase resolution andsequence regions of the mouse genome in a side-by-side comparison with humangenomic sequences Given the conservation of genetic information and the use

of the mouse in animal models of disease, these data are anticipated to be highlyinformative

Commercial development of pharmaceuticals for gene therapy is a burgeoning field.Discovery and clinical applications of novel genes is expected to continue at anaccelerating pace Gene manipulations to increase expression levels and to providecellular specificity and control mechanisms will lead to added safety and efficacy.There does not appear to be many limitations to the accomplishments of molecu-lar biologists with regard to gene discovery and engineering Methods and inven-tions to deliver these genes in vivo to specific cell types is an area needingimprovement, and the variety of approaches presently being investigated bodes wellfor future breakthroughs In particular, present delivery systems are often lacking

in both specificity and efficiency Since the active ingredient, DNA, is by far the mostexpensive, improvements in delivery will be beneficial on the bases of both cost andpotency It is expected that both components of gene therapy (plasmid DNA anddelivery vehicles) will see large improvements in the near future Large-scale man-ufacturing methods for production and purification will fall into place as the utility

of gene therapies is demonstrated in many of the ongoing clinical trials

KEY CONCEPTS

• In the commercial environment, basic research efforts are evaluated for bothscientific merit and economic potential The economic potential usually isdefined by a clinical application Large companies will typically focus on diseasestates with large markets Smaller companies tend to be open to any oppor-tunity that fits their strategic intent

• Large pharmaceuticals companies may have both gene discovery efforts andgene delivery programs in place and can integrate them to create a proprietarypharmaceutical In contrast, most small companies or budding entrepreneurswill only have one of the two main components in hand

• The research and development needed to advance a proprietary technology

is largely defined by the expected clinical applications For formulated DNA

or nonviral DNA delivery systems, manufacturing concerns about the ponents are not different from what has been developed for protein and drugpharmaceuticals

com-• As for any commercial venture, patent and licensing issues for a particular gene will necessarily be important factors in choosing a target The size of thepotential patient population and the accessibility of patients for a particularproduct is crucial In addition, as with any new therapy, gene therapy approachfor a disease state would need to have advantages over treatments currently

in use

• One of the challenges facing industry is the large-scale purification of plasmidDNA A number of companies have begun to market adapted microprepara-tive methods for the production of larger quantities of plasmid DNA Thesemodified mini-prep kits generally make use of the alkaline lysis method for celldisruption followed by a chromatographic cartridge purification These kitshave been aimed at one particular market niche: the production of small quan-

KEY CONCEPTS 371

tities (milligram or less) of research-grade material for molecular biology cations They do not meet the rigorous requirements for the development of ahighly controlled drug manufacturing process and most do not have a DrugMaster File (DMF).

appli-• FDA’s “Points to Consider” on plasmid DNA was drafted in October, 1996, andprovides the U.S approach to regulation of plasmid preventative vaccines.The same general criteria that guide the manufacture of recombinant proteinpharmaceuticals apply to the development of processes for the production ofplasmid DNA for human clinical investigations

• To proceed to advanced clinical trials and ultimately gain regulatory approval,the pharmaceutical development of gene therapy products will have to meetthe requirements for cGMP (current good manufacturing practices) produc-tion There are two main components of cGMP, comprising both the produc-tion and quality controls This includes the suitability of facility and staff for the manufacture of product, development of standard operating procedures(SOPs), and record keeping Quality control is concerned with sampling, spec-ifications, testing, and with documentation and release procedures ensuring satisfactory quality of the final product

• A group of promising new tools is emerging that will allow patterns of geneexpression to be compared in healthy and diseased tissue On the one hand,these gene profiling techniques will detect gene therapy targets—genes whoseproducts contribute to disease On the other hand, they will identify geneswhose products may be useful when delivered as replacement genes

SUGGESTED READINGS

Commercial Implications

Chang PL (Ed.) Somatic Gene Therapy CRC Press, Boca Raton, FL, 1995.

FDA (CBER) Points to consider on plasmid DNA vaccines for preventive infectious disease indications, Docket No 96N-0400 Food and Drug Administration, Washington, DC 1996.

Friedman T The future for gene therapy—a reevaluation Ann NY Acad Scc 265:141–152, 1976.

Gene therapy therapeutic strategies and commercial prospects TIBTECH 11 (5, Special Issue), May 1993.

Hines RN, O’Conner KC, Vella G, Warren W Large scale purification of plasmid DNA by anion exchange high performance liquid chromatography Biotechniques 12:430–433, 1992.

Horn NA, Meek JA, Budahazi G, Marquet M Cancer gene therapy using plasmid DNA: Purification of DNA for human clinical trials Hum Gene Therapy 6:565–573, 1995 Ledley FD Pharmaceutical approach to somatic gene therapy Pharmaceut Res 13:1595–1614, 1996.

Mahato RI, Smith LC, Rolland A Pharmaceutical perspectives on nonviral gene therapy Adv Genet 41:95–156, 1999.

Maniatis T, Fritsch EF, Sambrook J Molecular Cloning: A Laboratory Manual, 2nd ed Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1989.