Measurement of DNA Biomarkers for the Safety of Tissue-Engineered Medical Products, Using Artificial...

transcript

TISSUE ENGINEERINGVolume 10, Number 9/10, 2004© Mary Ann Liebert, Inc.

Measurement of DNA Biomarkers for the Safety of Tissue-Engineered Medical Products, Using Artificial

Skin as a Model

HENRY RODRIGUEZ, Ph.D., M.B.A.,1 CATHERINE O’CONNELL, Ph.D.,1PETER E. BARKER, Ph.D.,1 DONALD H. ATHA, Ph.D.,1 PAWEL JARUGA, Ph.D.,2

MUSTAFA BIRINCIOGLU, M.D.,3 MICHAEL MARINO, Ph.D.,4 PATRICIA MCANDREW,4and MIRAL DIZDAROGLU, Ph.D.1

ABSTRACT

To test the hypothesis that the process of tissue engineering introduces genetic damage to tissue-en-gineered medical products, we employed the use of five state-of-the-art measurement technologiesto measure a series of DNA biomarkers in commercially available tissue-engineered skin as a model.DNA was extracted from the skin and compared with DNA from cultured human neonatal controlcells (dermal fibroblasts and epidermal keratinocytes) and adult human fibroblasts from a 55-year-old donor and a 96-year-old donor. To determine whether tissue engineering caused oxidative DNAdamage, gas chromatography/isotope-dilution mass spectrometry and liquid chromatography/iso-tope-dilution mass spectrometry were used to measure six oxidatively modified DNA bases as bio-markers. Normal endogenous levels of the modified DNA biomarkers were not elevated in tissue-engineered skin when compared with control cells. Next, denaturing high-performance liquidchromatography and capillary electrophoresis-single strand conformation polymorphism were usedto measure genetic mutations. Specifically, the TP53 tumor suppressor gene was screened for mu-tations, because it is the most commonly mutated gene in skin cancer. The tissue-engineered skinwas found to be free of TP53 mutations at the level of sensitivity of these measurement technolo-gies. Lastly, fluorescence in situ hybridization was employed to measure the loss of Y chromosome,which is associated with excessive cell passage and aging. Loss of Y chromosome was not detectedin the tissue-engineered skin and cultured neonatal cells used as controls. In this study, we havedemonstrated that tissue engineering (for TestSkin II) does not introduce genetic damage above thelimits of detection of the state-of-the-art technologies used. This work explores the standard for mea-suring genetic damage that could be introduced during production of novel tissue-engineered prod-ucts. More importantly, this exploratory work addresses technological considerations that need tobe addressed in order to expedite accurate and useful international reference standards for theemerging tissue-engineering industry.

1Chemical Science and Technology Laboratory, National Institute of Standards and Technology, Gaithersburg, MD 20899.2Department of Chemical and Biochemical Engineering, University of Maryland, Baltimore County, Baltimore, MD 21250.3Department of Pharmacology, Medical School, Adnan Menderes University, Aydin, Turkey.4Applied Genomics and Molecular Genetics Division, Transgenomic, Gaithersburg, MD 20878.

INTRODUCTION

TISSUE ENGINEERING is an emerging biotechnology sec-tor that will provide replacement tissues for patients,

as well as complex, functional biological systems for re-search and testing in the pharmaceutical industry. Thefirst tissue-engineered organ that has progressed fromlaboratory bench to patient care has been skin. The needfor alternative, immediate and permanent wound closurematerials has created a multimillion dollar industry tomanufacture skin substitutes for wound coverage andwound healing. Estimates for hospitalizations from burnsrange from 60,000 to 80,000 annually, and costs for re-covery from acute injuries range from US$36,000 toUS$117,000 per patient.1,2 The knowledge gained fromcultured epidermal autografts, in which a small skin bi-opsy is expanded in vitro to produce large epidermalsheets, led to the development of skin equivalents, inwhich donor tissue with limited immunogenicity is thesource of cell material. Commercially available productsfall into two categories: dermal skin substitutes and com-bined epidermal and dermal skin substitutes. For exam-ple, regarding dermal skin substitutes, AlloDerm (Life-Cell, Branchburg, NJ) is a dermal matrix lackingimmunogenic cells, whereas Integra (Integra Life-Sciences, Plainsboro, NJ) is a combination of dermal fi-broblasts and bovine collagen. Dermagraft (Smith &Nephew-Wound Management, Largo, FL) consists ofnonimmunogenic neonatal fibroblasts cultured on apolyglactin mesh and has been used to treat burns3 anddiabetic foot ulcers.4 Regarding combined epidermal anddermal skin substitutes, Apligraf (Organogenesis, Can-ton, MA) is the first mass-produced skin product com-posed of fibroblasts and keratinocytes (derived from hu-man neonatal foreskin) in a type I bovine collagenextracellular matrix as approved by the Food and DrugAdministration for treatment of partial-thickness and full-thickness skin loss due to venous stasis ulcers.5,6 Apligrafis a bilayered living skin analog with appearance andhandling characteristics similar to normal human skin.7

TestSkin II (Organogenesis) is equivalent to Apligraf, butsold for in vitro research.

The integration of tissues within the body presents ma-jor challenges that tissue engineering must address.Specifically, cells within these tissues must cope with pe-riods of anoxia, inflammation, and possible immune sys-tem attack. Approaches to promote angiogenesis and as-sist cells to withstand inflammation and oxidative stresspostintroduction to the body will be required. In addition,today a variety of materials have been considered forscaffolding, the most common being the biodegradablepolyesters and some natural biopolymers such as colla-gen and a variety of polysaccharides.8 Unfortunately, ma-terial degradation can have broad consequences for med-ical devices, such as the loss of structural integrity of a

MEASUREMENT OF DNA BIOMARKERS

biomaterial or affecting a tissue-engineered medicalproduct (TEMP), either desirably or adversely, when un-intended. In particular, a key feature of a TEMP in theresponse to a degrading material is inflammation.9 As aresult, because of the rapid emergence of TEMPs and thedirection toward products that will be permanent in thepatient, it is important to test the safety of such products.

Our hypothesis is that the process of tissue engineer-ing introduces genetic damage to the final tissue-engi-neered product. To test this hypothesis, we took a novelapproach, working with tissue-engineered skin as a modelthat may be applied to the analysis of the safety of TEMPsbased on DNA damage biomarkers that reflect geneticdamage. Our work focuses on DNA damage biomarkersbecause many pathological conditions are initiated by anincreased incidence of DNA damage that may be mani-fested in the form of base modifications, adduct forma-tion, DNA strand breaks, and DNA–protein cross-links.10,11 Because analytical techniques in DNA researchhave undergone revolutionary improvements in sensitiv-ity, throughput, and accuracy, we believe that the mea-surement of DNA damage biomarkers to assess geneticdamage in a TEMP will likely become an integral partof the tissue-engineering process, especially in the areasof quality assurance/quality control and clinical safety.Biomarkers that reflect genetic integrity include DNAmodifications, mutations, and age-related chromosomalloss. These changes may occur during the manufactur-ing, storage, and shipment of the TEMP, possibly af-fecting the efficacy and safety of the product. Measure-ment of biomarkers may help expedite the regulatoryprocess and thus be critical in the development of thisemerging field of biotechnology.

The primary focus of the present work was to identifyand develop state-of-the-art measurements that could beused for the detection of novel DNA damage biomarkersin order to test the hypothesis that the process of tissueengineering introduces genetic damage to TEMPs. Tis-sue-engineered skin (TestSkin II from Organogenesis)was used as a model system. Five diverse measurementtechnologies were used. Gas chromatography/isotope-dilution mass spectrometry (GC/IDMS) and liquid chro-matography/isotope-dilution mass spectrometry (LC/IDMS) were used to measure oxidatively modified DNAbases to determine whether the tissue-engineered mater-ial underwent inflammation/oxidative DNA damage. Denaturing high-performance liquid chromatography(DHPLC) and capillary electrophoresis-single strandconformation polymorphism (CE-SSCP) were applied tomeasure mutations in the TP53 tumor suppressor genebecause of the high frequency of such mutations in skincancer. Loss of Y chromosome associated with excessivecell passage and aging was measured by fluorescence insitu hybridization (FISH). In this article we report on theuse of these measurement techniques and corresponding

biomarkers to test the hypothesis that the process of tis-sue engineering introduces genetic damage to tissue-en-gineered materials.

MATERIALS AND METHODS

Cell culture and DNA isolation

HeLa cells (ATCC, Manassas, VA) were placed in antibiotic-free Dulbecco’s modified Eagle’s medium(DMEM) containing 10% (v/v) fetal bovine serum andgrown in 175-cm2 flasks at 37°C in a tissue culture in-cubator (5% CO2/95% air). Medium was aspirated andcells were rinsed with 20 mL of 1� phosphate-bufferedsaline (PBS). Cells were detached by adding 3 mL oftrypsin–EDTA solution. Human dermal neonatal fibro-blasts (HDFn; Cascade Biologics, Portland, OR) and fi-broblasts obtained from 55- and 96-year-old donors(Coriell Cell Repositories, Camden, NJ) were cultured inmedium 106 supplemented with low-serum growth sup-plement in the absence of antibiotics and antimycotics(Cascade Biologics). Medium was aspirated and cellswere rinsed with 20 mL of 1� PBS. Cells were detachedby adding 3 mL of trypsin–EDTA solution. Trypsin–EDTA solution was immediately removed and an aliquotof 3 mL of trypsin neutralizer solution (Cascade Biolog-ics) was added to each flask. Human epidermal neonatalkeratinocytes (HEKn; Cascade Biologics) were culturedin EpiLife medium supplemented with human keratino-cyte growth supplement in the absence of antibiotics andantimycotics (Cascade Biologics) and processed as de-scribed above. Lymphocytes were donated by a 50-year-old male volunteer and cultured for 72 h in RPMI medium(20% [v/v] fetal calf serum, 2 mM glutamine, and peni-cillin–streptomycin [100 U; 100 �g/mL]) at 37°C with5% CO2.

RODRIGUEZ ET AL.

DNA was isolated with a blood and cell culture DNAkit (Qiagen, Valencia, CA), recovered by spooling,washed once in 70% ethanol, and air dried. DNA wasdissolved in 10 mM sodium phosphate buffer at a con-centration of 0.3 mg/mL and dialyzed against water for18 h at 4°C. Water outside the dialysis bags was changedthree times during the course of the dialysis. Subse-quently, DNA concentration was determined by ultravi-olet (UV) spectroscopy. Aliquots containing 100 �g ofDNA were dried under vacuum in a SpeedVac (ThermoSavant, Holbrook, NY). Calf thymus DNA (ctDNA) waspurchased from Sigma-Aldrich (St. Louis, MO).

Tissue-engineered skin (TestSkin II):Fibroblast/keratinocyte separation and DNA isolation

Separating cell layers. A TestSkin II disk (Organo-genesis) was placed in a 150-mm2 sterile petri dish con-taining 35 mL of 1� PBS and incubated at room tem-perature for 30 min in a cell culture hood (Fig. 1). PBSwas aspirated and replaced with 40 mL of protease typeX solution (0.5 mg/mL 1� PBS) and incubated at 37°Cfor 2 h in a tissue culture incubator (5% CO2/95% air).The epidermal layer (keratinocytes; gray top layer) wasseparated from the dermal layer (fibroblasts; white bot-tom layer) using sterile forceps.

Separating individual cells and DNA isolation. Theepidermal layer was placed in a 50-mL polypropyleneconical tube containing 40 mL of trypsin–versene andincubated in a 37°C water bath for 2 h. The dermal layerwas placed in a 175-cm2 flask containing 180 mL ofcollagenase solution (60 mL of collagenase [4.17mg/mL H2O] and 120 mL of collagenase premix [120mL of 1� PBS, 14 mL of 2.5% trypsin solution, and 1mL of filtered 0.45% glucose solution]) and incubated

FIG. 1. (A) TestSkin II skin product measuring approximately 75 mm in diameter, with an area of 44.2 cm2. (B) TestSkin IIskin product separated into its two skin layers: neonatal fibroblast and neonatal keratinocyte cell layers.

at 37°C in a tissue culture incubator (5% CO2/95% air)for 2 h. Each respective tube was agitated every 15 min.Cells were pelleted by centrifugation (1300 � g at 4°Cfor 5 min) and rinsed twice with 20 mL of 1� PBS withcentrifugation (1300 � g 4°C for 5 min) between eachrinse. Cells were eventually suspended in 4 mL of 1�PBS and counted with a cell culture hemocytometer.DNA was isolated with a blood and cell culture DNAkit (Qiagen).

Analysis of oxidatively modified DNA bases bygas chromatography/mass spectrometry and liquid chromatography/mass spectrometry

Details of the analysis of oxidatively modified DNAbases by GC/MS, using acidic hydrolysis or formami-dopyridamine-DNA glycosylase (Fpg) hydrolysis ofDNA, and LC/MS, using enzymatic hydrolysis of DNA,have previously been described.12–14 Statistical valueswere derived by the analysis of variance (ANOVA) sin-gle-factor test. Any value of probability greater than 0.05(p � 0.05) was not considered to be significantly dif-ferent.

Capillary electrophoresis-single strandconfirmation polymorphism analysis of the TP53 gene

Amplification of TP53 exons 5 to 9. Polymerase chainreaction (PCR) primers to amplify a region of the TP53gene containing exons 5 to 9 have been previously de-scribed.15 The 2-kilobase fragment was amplified in a 20-�L reaction volume containing the following: 250 ng ofgenomic DNA, 1� PCR buffer H (Invitrogen, Carlsbad,CA) primers (500 nM each), 2.0 mM MgCl2, 2.5 U ofAmpliTaq Gold DNA polymerase (Applied Biosystems,Foster City, CA), and dNTPs (200 �M each). Thermalcycling conditions were as follows: preamplification de-naturation: (1 cycle), 94°C for 5 min; amplification (35cycles): denaturation, 94°C for 30 s; annealing, 66°C for30 s; elongation, 72°C for 40 s; final elongation (1 cy-cle): 72°C for 7 min. A PerkinElmer 9700 thermal cy-cler was used for amplification.

Exon-specific amplification. For mutation detection, 1�L of the 2-kilobase amplification products was ream-plified with exon-specific fluorescently labeled PCRprimers (Table 1). National Institute of Standards andTechnology (NIST) TP53 reference materials were usedfor both positive and negative exon-specific controls at50 ng/PCR.16 Primers labeled with FAM (5-carboxyflu-orescein) or JOE (2�,7�-dimethoxy-4�,5�-dichloro-6-carboxyfluorescein), and 5� primer and 3� primer, wereprovided by Applied Biosystems. A PerkinElmer 9700thermal cycler and GeneAmp PCR core reagents

(PerkinElmer, Norwalk, CT) were used for all amplifi-cations. PCRs contained 1� PCR buffer II, 200 nM flu-orescently labeled primers, 1.5 mM MgCl2, 0.5 unit ofAmpliTaq DNA polymerase, and dNTPs (200 �M each).Thermal cycling conditions were identical to those usedin the amplification of TP53 exons 5 to 9. One microliterof a 1:10 dilution of these amplification products was an-alyzed by CE-SSCP.

CE-SSCP analysis. Control samples considered ofgenomic clones containing the exon 5 to 9 region of thehuman TP53 gene. The wild-type (WT) clone containsa normal TP53 gene sequence for this region. A clone(M4) containing a CGG-to-CAG mutation in amino acid248, in exon 7 of the human TP53 gene, was used as amutant control.16 Fluorescently labeled PCR sampleswere prepared for electrophoresis by combining 10.5�L of deionized formamide with 0.5 �L of 0.3 N NaOH,1 �L of PCR sample (diluted 1:10), and 0.5 �L ofGENESCAN-500 TAMRA (6-carboxy-tetramethylrho-damine)-labeled internal size standard. The mixture washeated for 2 min at 95°C and chilled on ice. SSCP sep-arations were performed with an Applied Biosystemsmodel 310 genetic analyzer.17 All separations were per-formed with an Applied Biosystems GENESCAN cap-illary and polymer system (41 cm � 50 �m capillary,3% GENESCAN polymer containing 10% glycerol in1� Tris–borate–EDTA [TBE]. Samples were electroki-netically injected (10 s, 3 kV) and separated at 10–13 kV. Data were collected and analyzed with AppliedBiosystems PRISM and GENESCAN software, version2.0.2.

Denaturing high-performance liquidchromatography analysis of the TP53 gene

Amplification of TP53 exons 5 to 9. Normal genomiclymphocyte DNA was purchased from Promega (Madi-son, WI). Tissue-engineered samples and appropriate con-trols were analyzed for mutations in exons 5 to 9 of thehuman TP53 gene. Control DNA samples for exon 7 (Fig.4) consist of a wild-type (WT) clone containing a normalTP53 gene sequence for this region and a mutant clone(M10) containing an ATG-to-ACG mutation in amino acid237, in exon 7 of the human TP53 gene.16 PCR amplifi-cations were performed in a 50-�L volume containing 0.25�L of Amplitaq Gold DNA polymerase, 5 �L of 10� PCRbuffer II, 4 �L of 25 mM MgCl2, 4 �L of 10 mM dNTPmix, 3 �L of the 5 �M forward and reverse primer set(Table 1), and 100 ng of DNA template. The final reac-tion volume was adjusted with distilled H2O. PCR wasperformed in 0.2-mL tubes in an MJ-200 thermocycler.Samples were denatured at 95°C for 10 min; heated at 95°Cfor 30 s, 56°C for 30 s, 72°C for 30 s for 30 cycles; witha final extension at 72°C for 7 min.

DHPLC analysis. Samples were analyzed with aWAVE 3500HT DNA fragment analysis system (Trans-genomic, Omaha, NE) under conditions described inTable 2. The stationary phase consists of alkylated non-porous polystyrene-divinylbenzene particles, and themobile phase consisted of buffer A (0.1 M triethylam-monium acetate [TEAA] and buffer B (0.1 M TEAA,25% acetonitrile [v/v]). The PCR products were dena-tured for 4 min at 94°C and cooled to room tempera-ture at a rate of 1°C/min. Ten to 20 microliters of PCRproduct was applied to the DNASep preheated reversedphase column (Transgenomic). As previously de-scribed,18,19 the temperature for optimal resolution ofheteroduplex and homoduplex DNA detection was de-termined by analyzing the melting of a PCR fragmentof each exon while the temperature was increased by1°C increments from 50°C until the fragment was com-pletely melted.20 The analysis temperature for eachfragment was chosen as the temperature at which about�75% of the DNA was present as an � helix.20 The ex-

RODRIGUEZ ET AL.

perimental DNA melting data were analyzed withWavemaker software (Transgenomic).

Analysis of Y chromosome loss by fluorescence in situ hybridization

Cells from cell lines or TestSkin II were collected byscraping, centrifuged (5 min at 1400 rpm in a tabletopcentrifuge), treated with hypotonic solutions, fixed, andair dried, and slides were prepared according to standardprocedures.21 Lymphocytes were treated with 75 mMKCl for 10 min at 37°C, and keratinocytes and fibroblastswere treated with 24 mM trisodium citrate for 20 min at37°C before fixation. Cells were collected from these hy-potonic solutions by mild centrifugation, fixed (metha-nol–acetic acid, 3:1 [v/v]), and spun out of fresh, ice-coldfixative three times before air-dried slides were prepared.FISH methods for the Y chromosome-painting probe, a120-kb locus specific to the SRY gene, was used ac-cording to supplier instructions (probe no. 32-190079,LSI SRY SpectrumOrange DNA FISH probe; Vysis,

TABLE 1. PCR PRIMER SEQUENCES FOR CAPILLARY ELECTROPHORESIS-SINGLE STRAND CONFORMATION

POLYMORPHISM AND DENATURING HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY

AmpliconTarget length

Exon Primer Sequence region (bp)

CE-SSCP

5 Forward: 5�-TTCCTCTTCCTACAGTACTC-3� 13040–13281 242Reverse: 5�-CTGGGCAACCAGCCCTGTCGT-3�

6 Forward: 5�-CCCCAGGCCTCTGATTCC-3� 13285–13494 210Reverse: 5�-CGGAGGGCCACTGACAACC-3�

7 Forward: 5�-CAGGTCTCCCCAAGGCGCAC-3� 13948–14166 219Reverse: 5�-GCAAGCAGAGGCTGGGGCAC-3�

8 Forward: 5�-ACTGCCTCTTGCTTCTCTTTTCC-3� 14417–14647 230Reverse: 5�-AATCTGAGGCATAACTGCACCC-3�

9 Forward: 5�-GGGTGTAGTTATGCCTCAGATT-3� 14626–14815 190Reverse: 5�-CGGCATTTTGAGTGTTAGACTGG-3�

5 Forward: 5�-CACTTGTGCCCTGACTTTCAAC-3� 13009–13277 269Reverse: 5�-GCAACCAGCCCTGTCGTCTC-3�

6 Forward: 5�-GGTTGCCCAGGGTCCCCAG-3� 13272–13489 218Reverse: 5�-GGCCACTGACAACCACCCTTAACC-3�

7 Forward: 5�-GCCACAGGTCTCCCCAAGGC-3� 13944–14154 211Reverse: 5�-TGGGGCACAGCAGGCCAGTG-3�

8 Forward: 5�-GGACCTGATTTCCTTACTGCC-3� 14402–14648 247Reverse: 5�-GAATCTGAGGCATAACTGCACC-3�

9 Forward: 5�-GGGTGCAGTTATGCCTCAGATT-3� 14626–14815 190Reverse: 5�-CGGCATTTTGAGTGTTAGACTGG-3�

Downers Grove, IL). Imaging was performed on a ZeissAxioplan2 imaging photomicroscope equipped withHamamatsu C-4880-80 CCD camera. Images (DAPI orSpectrumOrange) were captured and dual-color imageswere merged with ASI EasyFish version 1.1 (AppliedSpectral Imaging, Carlsbad, CA).22

RESULTS

Oxidative DNA damage

Using GC/MS with either acidic hydrolysis or Fpg hy-drolysis of DNA, and LC/MS with enzymatic hydrolysisof DNA, six oxidatively modified DNA bases were mea-sured as biomarkers in genomic DNA isolated from tis-

sue-engineered skin (TestSkin II) and from cultured con-trol cells. The levels of modified DNA bases were com-pared with those found in control cells, that is, neonatalfibroblasts and neonatal keratinocytes. The modifiedDNA bases measured by GC/MS were 8-hydroxyguanine(8-OH-Gua), 2,6-diamino-4-hydroxy-5-formamidopyri-midine (FapyGua), 4,6-diamino-5-formamidopyrimidine(FapyAde), 5-hydroxycytosine (5-OH-Cyt), and 5-hy-droxyuracil (5-OH-Ura) (for structures of these com-pounds see Dizdaroglu et al.11). 8-OH-Gua was also measured by LC/MS as its nucleoside form 8-hydroxy-2�-deoxyguanosine (8-OH-dGuo). Figure 2 illustrates thelevels of these compounds in DNA samples isolated fromtissue-engineered skin and two neonatal control cell lines(HDFn and HEKn). DNA from HeLa cells and ctDNA

TABLE 2. DENATURING HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY OVEN TEMPERATURE AND GRADIENT

CONDITIONS USED FOR EACH AMPLICON TO SCAN FOR MUTATIONS IN THE TP53 GENEa

Exons 2 � 3: Gradient at a column temp. of 61.1°C Exon 8: Gradient at column temps. of 59.7 and 62.7°C

Step Time %A %B Step Time %A %B

Loading 0.0 44 56 Loading 0.0 48 52Start gradient 0.1 41 59 Start gradient 0.1 45 55Stop gradient 2.0 31 69 Stop gradient 2.1 35 65

Exon 4: Gradient at a column temp. of 62°C Exon 9: Gradient at a column temp. of 58.2°CStep Time %A %B Step Time %A %B

Exon 5: Gradient at column temps. of 61.5 and 65.5°C Exon 10: Gradient at a column temp. of 62.2°CStep Time %A %B Step Time %A %B

Exon 6: Gradient at a column temp. of 60.3°C Exon 11: Gradient at a column temp. of 60°CStep Time %A %B Step Time %A %B

Exon 7: Gradient at a column temp. of 62.2°CStep Time %A %B

Loading 0.0 48 52Start gradient 0.1 45 55Stop gradient 2.1 35 65

aRun time per sample was 2.5 min; WAVE 3500HT mode.

RODRIGUEZ ET AL.1338

FIG. 2. Level of oxidative DNA damage as measured by GC/IDMS with Fpg hydrolysis (A, C, and D), GC/IDMS with acidichydrolysis (E and F), and LC/IDMS with enzymatic hydrolysis (B), A and B each represent three independent TestSkin II lay-ers combined into one (either fibroblast cell layer or keratinocyte cell layer). Values represent means � standard deviation fromthree independently performed measurements.

were also used as references. Figure 2A shows no sig-nificant difference among the values of the tissue-engi-neered samples (TE Skin Fibro A and B, and TE SkinKera A and B) and controls (HDFn, HEKn, and HeLa).The ctDNA was significantly different from the valuesof the other samples (p � 0.001). Figure 2B shows nosignificant difference among the values of the tissue-en-gineered samples compared with controls (HDFn andHEKn). HeLa was significantly different from the valuesof the other samples (p � 0.01) and the value of ctDNAwas also significantly different from the values of theother samples (p � 0.001). Figure 2C shows no signifi-cant difference among the values of the tissue-engineeredsamples compared with controls (HDFn, HEKn, andHeLa), whereas the value of ctDNA was significantly dif-ferent from the values of the other samples (p � 0.001).Figure 2D shows no significant difference among the val-ues of the tissue-engineered samples compared with con-trols (HDFn, HEKn, HeLa, and ctDNA). Figure 2E showsno significant difference among the values of the tissue-engineered samples when compared with controls(HDFn, HEKn, HeLa, and ctDNA). Figure 2F shows nosignificant difference among the values of the tissue-en-gineered samples compared with controls (HDFn, HEKn,HeLa, and ctDNA). Last, GC/MS and LC/MS yieldedsimilar results for 8-OH-Gua and its nucleoside 8-OH-dGuo.

TP53 mutation screening

CE-SSCP. As more than approximately 95% of all hu-man TP53 mutations are found within the exon 5-to-9 re-gion of the gene, this region was selected for analysis.23

Results for all five exons (exons 5 to 9) examined areshown in Table 3. PCR products were examined by CE-

SSCP at three different temperatures (20, 30, and 40°C).No significant changes in peak position or shape relativeto TP53 wild-type controls could be detected under theseconditions. For example, CE electropherograms obtainedat 30°C for the tissue-engineering and control samplesare shown in comparison with the TP53 wild-type sam-ple and a mutant control (Fig. 3). In general, differencesin migration for the TP53 mutant controls were most pro-nounced at this temperature specific mutations were notdetected at either 20 or 40°Ci (data not shown). The slightchanges in peak shape and position, caused by small fluc-tuations in the capillary system, were within experimen-tal error.24 All tissue-engineered samples and controlsmigrated with the normal control (WT). The mutant con-trol (M4) had both sense and antisense strands shifted in

TABLE 3. TP53 MUTATION ANALYSIS

BY CE-SSCP AND DHPLC

Sample 5 6 7 8 9

TE Total –/– –/– –/– –/– –/–TE Fibro –/– –/– –/– –/– –/–TE Kera –/– –/– –/– –/– –/–HDFn –/– –/– –/– –/– –/–HEKn –/– –/– –/– –/– –/–55-YO Fibro –/– –/– –/– –/– –/–96-YO Fibro –/– –/– –/– –/– –/–

Abbreviations: 55-YO Fibro, fibroblasts from a 55-year-olddonor; 96-YO Fibro, fibroblasts from a 96-YO donor; CE-SSCP, capillary electrophoresis-single strand conformationpolymorphism; DHPLC, denaturing high-performance liquidchromatography; Kera, keratinocytes; TE, tissue engineeredproduct.

FIG. 3. Tissue-engineered products (TE Total, TE Fibro, andTE Kera) analyzed by CE-SSCP at 30°C in comparison withcontrol neonatal cells (HDFn and HEKn), and normal fibro-blasts from 55-year-old (55-YO) and 96-year-old (96-YO)donors. TE Total, DNA isolated from three TestSkin II sam-ples that were not separated into their respective cell layers; TEFibro, DNA isolated from fibroblast cells of three TestSkin IIsamples after cell layer separation; TE Kera, DNA isolated fromkeratinocyte cells of three TestSkin II samples after cell layerseparation. Each panel contains sense and antisense SSCP prod-ucts (peaks 1 and 2, respectively) from exon 7 of the TP53 gene.WT, normal control; M4, mutant control.

migration with respect to normal and tissue-engineeredproducts.

DHPLC. A heteroduplex technology, DHPLC, wasused to analyze the DNA obtained from the tissue-engi-neered skin material and appropriate control cells. Eachexon and flanking intronic regions of the TP53 gene wereamplified by PCR and analyzed to detect the presence orabsence of nucleotide variants. Exons 2 and 3 were am-plified together as a single product because of the smallsize of exon 3. Should mutations be present in the testsamples at levels greater than approximately 5 to 10%, aheterozygote peak would be detected (Fig. 4A). Fully mu-tated samples may not be detectable when examined in-dependently of a wild-type control, as the peaks maycomigrate (Fig. 4A). For example, in exon 7, homozy-gous products comigrate and elute at approximately 2.15min. Heterozygote products elute faster (approximately1.9 to 2.0 min). Homozygous products are also evidentin the mixed (heterozygous) products. Results of the tis-sue-engineered samples in comparison with the controlDNA materials for exon 7 are shown in Fig. 4B. No het-

RODRIGUEZ ET AL.

eroduplex peak was detected for any of the skin productsin this or the other exons examined (Table 3).

Loss of the Y chromosome

The rates of loss of the Y chromosome were estimatedin control and aged fibroblasts in comparison with thetissue-engineered skin material (Table 4). Normal lym-phocytes in short-term culture show no instances of lossof the Y chromosome. Concomitantly, in neonatal cells(fibroblasts and keratinocytes), fibroblasts from a 55-year-old subject, as well as the tissue-engineered skin fi-broblasts and keratinocytes, no loss of the Y chromosomewas seen (Table 4 and Fig. 5). However, loss of the Ychromosome was detected at a rate of 8% in fibroblastsobtained from the 96-year-old control donor (Table 4 andFig. 5F).

DISCUSSION

The future of tissue engineering lies in the develop-ment of key technologies in the biological sciences and

FIG. 4. DHPLC analysis at 62.2°C of exon 7 of the TP53 gene. (A) Elution of homozygous normal (WT), homozygous mu-tant (M10), and heterozygous (WT and M10) PCR products. (B) Elution of tissue-engineered products (TE Total, TE Fibro, andTE Kera) in comparison with control neonatal cells (HDFn and HEKn), and normal fibroblasts from 55-year-old (55-YO) and96-year-old (96-YO) donors. TE Total, TE Fibro, and TE Kera samples were the same as those used in the CE-SSCP analysis.Elution profiles are within expected experimental error for the normal, wild-type sample (WT).

engineering. Although the initial focus was on the use ofone’s own cells, harvesting of cells from others will con-tinue and technology will develop beyond skin cells intopermanent products such as liver, pancreas, muscle, andnerve. To do this, however, requires a thorough under-standing of cell biology concerning biomaterial interac-tions. As a result of this evolutionary trend toward per-manent TEMPs and the need for consistent guarantees ofsafety, the primary focus of the present work was to iden-tify and develop state-of-the-art measurements that couldbe used for the detection of novel DNA biomarkers toassess damage in tissue-engineered materials. Using thestate-of-the-art measurement technologies available, wehave shown that the process of tissue engineering doesnot cause damage to tissue-engineered skin at the limitof detection of the technologies used in this study. Thisarticle reports the use of five complementary state-of-the-art measurement techniques that quantitatively addressthe hypothesis that the process of tissue engineering in-troduces genetic damage to TEMPs. The measurementtechnologies consisted of GC/IDMS, LC/IDMS, DHPLC,CE-SSCP, and FISH. Each technique was used to mea-sure a specific type of genetic damage biomarker in tis-sue-engineered skin: GC/IDMS and LC/IDMS for themeasurement of oxidatively modified DNA bases, DHPLC and CE-SSCP for the measurement of DNA mu-tations in the TP53 tumor suppressor gene, and FISH tomeasure loss of the Y chromosome associated with ex-cessive cell passage and aging.

The process of tissue engineering often involves themixing of cells with polymers that may cause inflamma-tion/oxidative stress to the cell, and thus elevate the levelof endogenous free radical production with subsequentoxidative damage to DNA. Free radicals are produced in

aerobic cells by normal metabolism (endogenously) andby exogenous sources (reviewed in Halliwell and Gut-teridge25). Normal endogenous levels of oxidative DNAdamage in a cell may vary between approximately 1 and10 modified bases per 106 normal DNA bases (reviewedin Dizdaroglu et al.11). Studies have shown that free rad-icals are involved in many diseases such as cancer andAlzheimer’s disease (reviewed in Halliwell and Gut-teridge25). Thus, it is important to determine whether anyelevated level of free radical-induced damage to DNAhas occurred to the TEMP as a result of its processing.

Oxidative damage to DNA can be measured by a va-riety of analytical techniques, which have their own advantages and drawbacks (reviewed in Collins, Diz-daroglu, et al.11,26,27). Most of these techniques measureonly a single product with no spectroscopic evidence foridentification. Techniques that use mass spectrometryprovide unequivocal identification and quantification ofmultiple products of oxidative DNA damage (reviewedin Dizdaroglu27). Our laboratory has developed method-ologies using GC/MS with the isotype-dilution techniquefor the measurement of oxidative DNA damage in cellsand in vitro.27 Liquid chromatography/tandem mass spec-trometry (LC/MS/MS) and LC/MS with the isotope-di-lution technique have emerged as new measurement tech-niques for this purpose (reviewed in Dizdaroglu et al.11).Both GC/MS and LC/MS possess a sensitivity level ofdetection that permits them to detect and quantify ox-idatively modified lesions in DNA at cellular backgroundlevels.11 Whereas GC/MS requires a minimum of 1 �gof DNA sample per analysis of the modified bases de-tected in this report, LC/MS requires 20 �g of DNA fordetecting modified DNA nucleosides such as 8-OH-dGuo. Thus, pooling of three tissue-engineered skin sam-

TABLE 4. CHROMOSOME IN PERIPHERAL BLOOD CULTURES, FIBROBLASTS, AND TE SAMPLES

Percentage of cellswith Y chromosomedetected by FISH Cells scored

Age-dependent experimentControl PBLs 100 15855-year-old fibroblasts 100 16496-year-old fibroblasts 92 245

Tissue culture experimentControl PBLs 100 216HDFn 100 176HEKn 100 128

TestSkin II experimentControl PBLs 100 150TE skin fibroblasts 100 214TE skin keratinocytes 100 160

Abbreviations: Control PBLs, peripheral blood lymphocyte cells; FISH, fluorescence in situhybridization; TE, tissue-engineered material obtained from TestSkin II.

ples from the same production batch of the manufacturerwas required for analysis. Because in performing qualityassurance of commercial products, samples are oftenpooled for analysis, there is no reason to expect that oneof the three products would have accumulated DNA mod-ifications at a different frequency than the other two. Inthe present work, six commonly studied oxidatively mod-ified DNA bases were investigated by GC/MS and LC/MS.The levels of these modified DNA bases were not elevatedin the tissue-engineered skin relative to normal controlcells. These results suggest that the tissue-engineering pro-cess to produce artificial skin and/or storage of the finalproduct does not cause a significant elevation of oxidativeDNA damage over the background levels.

RODRIGUEZ ET AL.

Alterations of the TP53 tumor suppressor gene are themost frequent genetic event found in human cancer.28 Be-cause of the high frequency of TP53 mutations in humanskin carcinomas,29 and their occurrence in chronicallysun-exposed skin,30 TP53 mutations are a relevant bio-marker for cancer. Using CE-SSCP and DHPLC, no mu-tations in exons 5 to 9 of the TP53 gene were detected.In one study,31 the limits of detection of DHPLC, CE-SSCP, two-dimensional gene scanning (TDGS), andconformational sensitive gel electrophoresis (CSGE)were compared. DHPLC detected 100% of mutationscompared with TDGS (91%), CSGE (76%), and SSCP(72%). DHPLC was also determined superior to singlestrand conformational analysis (CE-SSCP) in a number

FIG. 5. (A–F) Cells are stained with 4�, 6�-diamidino-2-phenylindole hydrochloride (DAPI), which strains genomic DNA (nu-clear staining) blue (light gray in monochrome), and SpectrumOrange dye probe, which hybridizes to the Y chromosome (whitein monochrome). Designations: Tissue-engineered products (TE Fibro and TE Kera), control neonatal cells (HDFn and HEKn),and normal fibroblasts from 55-year-old (55-YO) and 96-year-old (96-YO) donors. Arrows in (F) designate cells that have lostthe Y chromosome.

of studies referenced in one article.19 The authors re-ported that CE-SSCP detected only 50 to 97.5% of themutations discovered by DHPLC. In addition, DHPLCwas credited with finding 95 to 100% of mutations foundby other methods including sequencing. Thus, shouldmutations be present at sufficient levels for detection (ap-proximately 5% for DHPLC),32 we would expect to de-tect them. DHPLC analysis of the PCR amplicons fromthese products gave elution profiles similar to the normalcontrol samples, indicating that no sequence variantswere present in the exon 5-to-9 region of the TP53 gene.As was performed in oxidative DNA damage analysis,three tissue-engineered skin products were pooled fromthe same production batch for analysis of TP53 muta-tions. There is no reason to expect that one of the threeproducts would have accumulated mutations at a fre-quency different from the other two. These results sug-gest that TP53 mutations do not occur during productionand/or storage of artificial skin at a level above 5%, thedetection limit of DHPLC.

Karyotype assays, unlike other biochemical analyses,can usually be performed on minute samples containinga few intact cell nuclei. Such assays are robust, with ahigh degree of precision.33 Somatic cells from individu-als over age 50 years have been demonstrated to exhibitan elevated loss of the Y chromosome when comparedwith the overall male population, for reasons yet un-clear.34 This normal progressive loss is proportional toincrease in age.35,36 The skin products are derived fromneonatal cells and are thus expected to have little to noY chromosome loss. If present, such loss could indicatenot only a high number of cell passages, but also pre-mature aging or potential transformation events. The lossof Y chromosome was measured in the tissue-engineeredskin cells and compared with neonatal control cells andcells obtained from 55- and 96-year-old donors. Y chro-mosome loss was observed only in fibroblast cells ob-tained from the 96-year-old donor (aged cells). The results suggest that the neonatal cells used in the manu-facturing of tissue-engineered skin do not undergo an ex-cessive and/or detrimental number of passages. Althoughpathological consequences of Y chromosome loss in vivohave not been determined,36 chromosomal aneuploitymay prove an accurate biomarker for small populationsof cells and may prove a useful cutoff point during themanufacturing of TEMPs.

Although this study investigated only tissue-engi-neered skin and showed that no adverse genetic effectsof TEMP processing occurred, if the hypothesis is to becarried further, a greater sample size and additional typesof tissue-engineered materials are warranted. This addi-tional information would then be collected and made intoa reference database to be shared with the research com-munity and regulatory agencies. More importantly, thisexercise led to the following considerations that should

be addressed by the research and regulatory communityin the future development of clinical safety measurementstrategies and related industry standards: (1) target an-alytes—the biological effect of lesions measured, the ef-fective concentrations of such lesions, and limit of de-tection for the technologies chosen will guide the choiceof biomarker and their measurement technologies; (2)practical assay implementation features—this includesassay throughput, cost, and value to producer and cus-tomer; and (3) the extent of applicability—in otherwords, will the assays and biomarker be general enoughto apply across the spectrum of TEMP types and prod-uct formats?

In conclusion, future efforts are likely to increase focus on the development of TEMPs under consensussafety and efficacy standards, including sourcing of cellsand tissues, characterization and testing of the materials,quality assurance and control, and preclinical and clini-cal evaluation. As a pilot study of DNA-based analyticaltechnologies for TEMP quality assurance/quality control,this work constitutes a preliminary consideration of thetechnologies for measuring DNA damage. These consid-erations should, it is hoped, catalyze discussion of a gen-eral approach to TEMP safety. Such discussions amongthe private sector, governmental agencies, and researchlaboratories set the stage for development of standardconsensus protocols and reference materials to promoteproduct safety and better value for end users of thisemerging area of biotechnology.

ACKNOWLEDGMENT

Certain commercial equipment or materials are iden-tified in this article in order to specify adequately the ex-perimental procedures. Such identification does not im-ply recommendation or endorsement by the NationalInstitute of Standards and Technology, nor does it implythat the materials or equipment identified are necessarilythe best available for the purpose.

REFERENCES

1. Saffle, J.R., Davis, B., and Williams, P. Recent outcomesin the treatment of burn injury in the United States: A re-port from the American Burn Association Patient Registry.J. Burn Care Rehabil. 16, 219, 1995.

2. Brigham, P.A., and McLoughlin, E. Burn incidence andmedical care use in the United States: Estimates, trends,and data sources. J. Burn Care Rehabil. 17, 95, 1996.

3. Hansbrough, J.F. Dermagraft-TC for partial-thicknessburns: A clinical evaluation. J. Burn Care Rehabil. 18,S25–S28, 1997.

4. Gentzkow, G.D., Iwasaki, S.D., Hershon, K.S., Mengel,M., Prendergast, J.J., Ricotta, J.J., Steed, D.P., and Lipkin,

S. Use of dermagraft, a cultured human dermis, to treat di-abetic foot ulcers. Diabetes Care 19, 350, 1996.

5. Falanga, V., Margolis, D., Alvarez, O., Auletta, M., Mag-giacomo, F., Altman, M., Jensen, J., Sabolinski, M., andHardin-Young, J. Rapid healing of venous ulcers and lackof clinical rejection with an allogeneic cultured human skinequivalent. Human Skin Equivalent Investigators Group.Arch. Dermatol. 134, 293, 1998.

6. Sabolinski, M.L., Alvarez, O., Auletta, M., Mulder, G., andParenteau, N.L. Cultured skin as a “smart material” forhealing wounds: Experience in venous ulcers. Biomateri-als 17, 311, 1996.

7. Wilkins, L.M., Watson, S.R., Prosky, S.J., Meunier, S.F.,and Parenteau, N.L. Development of a bilayered living skinconstruct for clinical applications. Biotechnol. Bioeng. 43,747, 1994.

8. Hubbell, J.A. Biomaterials in tissue engineering. Biotech-nology 13, 565, 1995.

9. Hunt, J.A., Rhodes, N.P., and Williams, D.F. Analysis ofthe inflammatory exudates surrounding implanted poly-mers using flow cytometry. J. Mater. Sci. Mater. Med. 6,839, 1995.

10. Bernstein, C., and Bernstein, H. Aging, Sex and DNA Re-pair. San Diego, CA: Academic Press, 1991.

11. Dizdaroglu, M., Jaruga, P., Birincioglu, M., and Rodriguez,H. Free radical-induced damage to DNA: Mechanisms andmeasurement. Free Radic. Biol. Med. 32, 1102, 2002.

12. Dizdaroglu, M. Chemical determination of oxidative DNAdamage by gas chromatography-mass spectrometry. Meth-ods Enzymol. 234, 3, 1994.

13. Rodriguez, H., Jurado, J., Laval, J., and Dizdaroglu, M.Comparison of the levels of 8-hydroxyguanine in DNA asmeasured by gas chromatography mass spectrometry fol-lowing hydrolysis of DNA by Escherichia coli Fpg proteinor formic acid. Nucleic Acids Res. 28, E75, 2000.

14. Dizdaroglu, M., Jaruga, P., and Rodriguez, H. Measure-ment of 8-hydroxy-2�-deoxyguanosine in DNA by high-performance liquid chromatography-mass spectrometry:Comparison with measurement by gas chromatography-mass spectrometry. Nucleic Acids Res. 29, E12, 2001.

15. O’Connell, C.D., Tian, J., Juhasz, A., Wenz, H.M., andAtha, D.H. Development of standard reference materialsfor diagnosis of p53 mutations: Analysis by slab gel sin-gle strand conformation polymorphism. Electrophoresis 19,164, 1998.

16. O’Connell, C.D., Tully, L.A., Devaney, J.M., Marino,M.A., Jakupciak, J.P., and Atha, D.H. Renewable standardreference material for the detection of TP53 mutations.Mol. Diagn. 7, 85, 2003.

17. Atha, D.H., Wenz, H.M., Morehead, H., Tian, J., and O’Connell, C.D. Detection of p53 point mutations by sin-gle strand conformation polymorphism: Analysis by capil-lary electrophoresis. Electrophoresis 19, 172, 1998.

18. Gross, E., Kiechle, M., and Arnold, N. Mutation analysisof p53 in ovarian tumors by DHPLC. J. Biochem. Biophys.Methods 47, 73, 2001.

19. Spiegelman, J.I., Mindrinos, M.N., and Oefner, P.J. High-accuracy DNA sequence variation screening by DHPLC.Biotechniques 29, 1084, 2000.

RODRIGUEZ ET AL.

20. Keller, G., Hartmann, A., Mueller, J., and Hofler, H. De-naturing high pressure liquid chromatography (DHPLC) forthe analysis of somatic p53 mutations. Lab. Invest. 81,1735, 2001.

21. Barch, M.J., Knutsen, T., and Spurbeck, J.L. The AGT Cytogenetics Laboratory Manual. New York: Lippincott-Raven, 1997.

22. Barker, P.E., and Schwab, M. Junction mapping of translo-cation chromosomes by fluorescence in situ hybridization(FISH) and computer image analysis in human solid tu-mors. Methods Mol. Genet. 2, 129, 1993.

23. Olivier, M., Eeles, R., Hollstein, M., Khan, M.A., Harris,C.C., and Hainaut, P. The IARC TP53 database: New on-line mutation analysis and recommendations to users. Hum.Mutat. 19, 607, 2002.

24. Wenz, H.M., Ramachandra, S., O’Connell, C.D., and Atha,D.H. Identification of known p53 point mutations by cap-illary electrophoresis using unique mobility profiles in ablinded study. Mutat. Res. 382, 121, 1998.

25. Halliwell, B., and Gutteridge, J.M.C. Free Radicals in Bi-ology and Medicine. Oxford: Oxford Science, 1999.

26. Collins, A., Cadet, J., Epe, B., and Gedik, C. Problems inthe measurement of 8-oxoguanine in human DNA. Reportof a workshop, DNA oxidation, held in Aberdeen, UK,19–21 January, 1997. Carcinogenesis 18, 1833, 1997.

27. Dizdaroglu, M. Mechanisms of free radical damage toDNA. In: Aruoma, O.I., and Halliwell, B., eds. DNA andFree Radicals: Techniques, Mechanisms, and Applications.Saint Lucia: OICA International, 1998.

28. Hollstein, M., Sidransky, D., Vogelstein, B., and Harris, C.C.p53 mutations in human cancers. Science 253, 49, 1991.

29. Basset-Seguin, N., Moles, J.P., Mils, V., Dereure, O., andGuilhou, J.J. TP53 tumor suppressor gene and skin car-cinogenesis. J. Invest. Dermatol. 103, 102S, 1994.

30. Ren, Z.P., Hedrum, A., Ponten, F., Nister, M., Ahmadian,A., Lundeberg, J., Uhlen, M., and Ponten, J. Human epi-dermal cancer and accompanying precursors have identi-cal p53 mutations different from p53 mutations in adjacentareas of clonally expanded non-neoplastic keratinocytes.Oncogene 12, 765, 1996.

31. Eng, C., Brody, L.C., Wagner, T.M., Devilee, P., Vijg, J.,Szabo, C., Tavtigian, S.V., Nathanson, K.L., Ostrander, E.,and Frank, T.S. Interpreting epidemiological research:Blinded comparison of methods used to estimate the preva-lence of inherited mutations in BRCA1. J. Med. Genet. 38,824, 2001.

32. Liu, M.R., Pan, K.F., Li, Z.F., Wang, Y., Deng, D.J.,Zhang, L., and Lu, Y.Y. Rapid screening mitochondrialDNA mutation by using denaturing high-performance liq-uid chromatography. World J. Gastroenterol. 8, 426, 2002.

33. Dewald, G., Stallard, R., Al Saadi, A., Arnold, S., Bader,P.I., Blough, R., Chen, K., Elejalde, B.R., Harris, C.J., Hig-gins, R.R., Hoeltge, G.A., Hsu, W.T., Kubic, V., Mc-Corquodale, D.J., Micale, M.A., Moore, J.W., Phillips,R.M., Scheib-Wixted, S., Schwartz, S., Siembieda, S., Strole, K., VanTuinen, P., Vance, G.H., Wiktor, A., Zins-meister, A., et al. A multicenter investigation with inter-phase fluorescence in situ hybridization using X- and Y-chromosome probes. Am. J. Med. Genet. 76, 318, 1998.

34. United Kingdom Cancer Cytogenetics Group (UKCCG).Loss of the Y chromosome from normal and neoplastic andbone marrows. Genes Chromosomes Cancer 5, 83, 1992.

35. Bukvic, N., Gentile, M., Susca, F., Fanelli, M., Serio, G.,Buonadonna, L., Capurso, A., and Guanti, G. Sex chro-mosome loss, micronuclei, sister chromatid exchange andaging: A study including 16 centenarians. Mutat. Res. 498,159, 2001.

36. Stone, J.F., and Sandberg, A.A. Sex chromosome aneu-ploidy and aging. Mutat. Res. 338, 107, 1995.

Address reprint requests to:Henry Rodriguez, Ph.D., M.B.A.

Group Leader, Cell & Tissue Measurements GroupBiotechnology Division

National Institute of Standards and Technology (NIST)100 Bureau Drive, Mail Stop 8313

Gaithersburg, MD 20899-8311

E-mail: henry.rodriguez@nist.gov

Measurement of DNA Biomarkers for the Safety of Tissue-Engineered Medical Products, Using Artificial...

Documents