Development and application of highly specific PCR for detection of ...

Eur Food Res Technol (2013) 236:129–134 DOI 10.1007/s00217-012-1868-7

ORIGINAL PAPER

Development and application of highly specific PCR for detection of chicken (Gallus gallus) meat adulteration Nagappa S. Karabasanavar • S. P. Singh Deepak Kumar • Sunil N. Shebannavar

•

Received: 2 June 2012 / Revised: 25 October 2012 / Accepted: 29 October 2012 / Published online: 16 November 2012 Ó Springer-Verlag Berlin Heidelberg 2012

Abstract In order to prevent fraud in the sale and strengthen quality assurance, authentic identification of chicken meat is essential. In the present investigation, a chicken (Gallus gallus)-specific polymerase chain reaction (PCR) was developed for the unambiguous identification of chicken meat. The PCR assay employs pair of primers designed against chicken nuclear 5-aminolevulinate (ALA) synthase gene. Highly chicken-specific diagnostic amplicon of 288 bp was established upon PCR and was evident in all the nine breeds/strains of chicken species. Sensitivity of PCR in detecting chicken meat adulteration was established to be at 0.1 % in the foreign meat matrix, while limit of detection (LOD) of chicken DNA was 10 pg. Suitability of the developed chicken-specific PCR was validated and confirmed in raw, cooked/heat treated (60, 80, 100, and 121 °C), and micro-oven cooked meat samples. Possibility of cross-amplification of adulterating DNA was excluded by cross-checking the developed PCR assay with several animal and avian species. The PCR assay developed in this

N. S. Karabasanavar (&) Department of Veterinary Public Health and Epidemiology, Veterinary College, Shimoga, Karnataka 577 204, India e-mail: [email protected] S. P. Singh Department of Veterinary Public Health, College of Veterinary and Animal Sciences, G. B. Pant University of Agriculture and Technology, Pantnagar, Uttarakhand 263145, India D. Kumar Department of Veterinary Public Health and Epidemiology, College of Veterinary Science and Animal Husbandry, R. K. Nagar, Agartala, Tripura (West) 799008, India S. N. Shebannavar Gennova Biopharmaceutical, Pune, Maharashtra, India

study is highly promising for applications involving circumstances that require authentic identification of chicken meat. Keywords

Meat Chicken Adulteration DNA PCR

Introduction Meat is often presented wrongly because of economic reasons, and fraudulent misrepresentation of meat is an offense owing to social and public health considerations. The food offered to public should not be deceptive or fraudulent and should carry proper label [1]. So as to keep consumer’s confidence, the quality assurance should meet the challenge of authentic identification of origin of meat species. Further, there has also been growing incidence of allergy against meat food; 41 % of such human cases have been attributed to chicken [2]. Hence, there is a need for rapid, sensitive, and reliable technique for the authentic identification of chicken meat [3, 4]. Numerous analytical methodologies have been developed for this purpose such as physicochemical, electrophoretic, chromatographic, immunological [5, 6]; however, none of these conventional approaches have been conclusively proved for authentic identification due to one or the other inherent limitation. On the other hand, owing to their repeatability and reproducibility, the DNA-based molecular techniques have promisingly emerged in the recent past [7]. For identification of poultry meat, various DNA-based techniques such as DNA hybridization [8], polymerase chain reaction (PCR)-generated satellite probes [9], restriction fragment length polymorphism (RFLP), and various kinds of PCR assays have been developed [3, 4, 10, 11].

123

130

However, such techniques require either post-PCR handling (restriction digestion in RFLP), software for analysis (RAPD fingerprinting), sophisticated machine or costly chemicals (real-time PCR) to achieve precise identification and differentiation of species. On the other hand, species-specific PCR assays have certain advantages over other DNA-based methods in terms of rapidity and specificity. The PCRs employing specifically designed oligonucleotide primers under restrictive conditions are in vogue for meat speciation, thereby obviating the necessity of further product sequencing or digestion of the PCR products with restriction enzymes. Keeping these considerations in view, need for a chicken-specific PCR that would authentically chicken meat and do not require much post-PCR processing and which would enable detection of chicken even in cooked (heat treated) and adulterated samples was envisaged.

Eur Food Res Technol (2013) 236:129–134

temperatures for a fixed time period. Thermal processing was carried out at different temperatures, that is, in water bath at 60, 80, and 100 °C for 30 min and autoclaving at 121 °C (15 lbs pressure) for 30 min. Similarly, micro-oven cooking was performed to ensure thorough cooking followed by DNA extraction. Preparation of adulteration mixtures A base adulteration meat mixture (BAM) was prepared by mixing meats of cattle, buffalo, sheep, goat, pig, turkey, duck, quail, and guinea fowl in equal proportions; to this mixture, the chicken meat was incorporated in different levels (10, 5, 1, 0.5, and 0.1 %). A total of 100 mg meat mixture from each sample (chicken meat incorporated to BAM) was then subjected to DNA isolation. DNA isolation

Materials and methods Collection of samples Meat or blood samples were collected from various animal and avian species viz. cattle (Bos indicus and Bos taurus n = 52), buffalo (Bubalus bubalis n = 46), sheep (Ovis aries n = 37), goat (Capra hircus n = 42), pig (Sus scrofa domesticus n = 34), camel (Camelus dromedarius n = 1), horse (Equus caballus n = 15), dog (Canis familiaris n = 40), rabbit (Oryctolagus cuniculus n = 10), tiger (Panthera tigris n = 2), leopard (Panthera pardus n = 3), barking deer (Muntiacus muntjak n = 3), sika deer (Cervus nippon n = 3), goral (Naemorhedus goral n = 2), sambar (Cervus unicolor n = 1), elephant (Elephas maximus n = 8), human (Homo sapiens n = 1), and fish (Tor putitora, Schizothorax richardsonii, Raiamas bola, Hypophthalmichthys molitrix, and Garra gotyla gotyla n = 10). Six avian species namely duck (Anas platyrhynchos n = 10), turkey (Meleagris gallopavo n = 4), guinea fowl (Numida meleagris n = 30), Japanese quail (Coturnix japonica n = 10), kite (Milvus migrans n = 1), and parakeet (Psittacula krameri n = 1); as many as 9 breed/strains of chicken (Gallus gallus n = 58) were included in the study, that is, White Leghorn, Columbian, Aseel, Kadaknath, Australorp, Rhode Island Red, White Cornish, New Hampshire, and a cross-bred. Approximately, 50 g of tissue/1–2 ml of blood/semen/milk samples were collected from local markets, post-mortem halls, veterinary clinics, organized farms, or zoo. Thermal processing of samples To replicate kitchen cooking and to evaluate robustness of PCR, raw meat samples were cooked at different

123

DNA extraction was undertaken using WizardÒ Genome DNA purification kit (Promega, Madison, Wisconsin, USA) following manufacturer’s instructions for all samples. Phenol–chloroform method [12] was also used for extraction of DNA for some blood samples. Primer designing Chicken-specific primers were designed against chicken nuclear 5-aminolevulinate (ALA) synthase gene. Target region (Accession no. X03517) and similar DNA sequences were downloaded from GenBank database and aligned using ‘‘MegAlign’’ program (Lasergene software; DNAStar, Inc., Madison, Wisconsin, USA). Primer sites were selected considering intra-species homogeneity and interspecies heterogeneity. Primers were designed against such unique sequence using ‘‘Primer-Select’’ program (Lasergene software; DNAStar, Inc.). Selected primers were screened for species-specificity to chicken and crossamplification in other species using local nucleotide alignment tool Basic Local Alignment Search Tool (‘‘BLAST’’) (http://www.ncbi.nlm.nih.gov/blast). Selected primers were custom synthesized (IDT, Inc. Coralville, Iowa, USA) and used for PCR amplification viz. forward (VPH-ChF) 50 -CCG GCA GCA GAG CAG ACT AAC AAC-30 and reverse (VPH-ChR) 50 -AGG GGA TAC GCC GAC TGC TGA G-30 . PCR amplification The PCR mixture, that is, primer concentration, dNTPs, MgCl2, and program (importantly the annealing temperature) were standardized for chicken-specific amplification. The optimized PCR was set up in a 25 ll reaction mixture


consisting of 2.5 ll of 10X assay buffer [160 mM (NH4)2SO4, 670 mM Tris–HCl, pH 8.8, 0.1 % tween-20, 25 mM MgCl2, Bioron GmbH, Ludwigshafen, Germany], 0.5 ll (200 lM) of dNTP mix [Sodium salts of dATP, dCTP, dGTP, and dTTP 10 mM each in water i.e., 40 mM total pH 7.5, Promega], 0.5 ll (20 Pico moles) each of forward and reverse primers, 1U Taq DNA polymerase (DFS-Taq DNA polymerase, Bioron GmbH), 50 ng of purified DNA and nuclease free water (Merck, Darmstadt, Germany) to make the volume. The tubes were flash spun, and the PCR was performed in a Thermal cycler (GeneAMPÒ PCR System 9700; Applied Biosystems, Foster City, California, USA). The cycling conditions consisted of an initial denaturation (95 °C for 5 min) followed by 30 cycles of denaturation (95 °C, 30 s), primer annealing (55 °C, 30 s), and extension (72 °C, 30 s). After the final extension (72 °C, 5 min), the PCR products were held at 4 °C until subjected to electrophoresis. Agarose gel (2 %) was prepared in 1X TBE (Tris: Boric acid: EDTA) buffer, and the PCR products (5 ll) stained with 6X gel loading dye (1 ll) were electrophoresed at 40–60 V for 1–3 h. The amplified products were visualized and confirmed over a gel documentation system (AlphaImagerÒ HP; Alpha Innotech Corp., San Leandro, California, USA). The relative molecular weight of the amplicon was calculated against a 100 bp DNA ladder (Bangalore Genei, Bangalore, India). Sensitivity and specificity of chicken-specific PCR The sensitivity of the chicken-specific PCR was tested using tenfold serial dilutions of the template DNA starting from 10 ng downward per reaction and PCR was performed. The specificity or selectivity of the chicken-specific PCR was evaluated using non-target (other than chicken) species DNA; for excluding, chances of crossamplification as many as 18 animal and 6 avian species were considered.

131

Chicken-specific PCR assay Primers were designed against chicken nuclear 5-aminolevulinate (ALA) synthase gene. The structural gene is of size 6.9 kb and contains 10 exons; this target (exon 1–7, size 5,282 bp) sequence was downloaded from National Centre for Biotechnology Information (NCBI) (Accession no. X03517) for primer designing. After PCR amplification, a chicken-specific product of 288 bp was obtained corresponding to a discrete amplicon between positions 6 and 293. Upper primer annealed between target nucleotides 6-29, while the lower primer annealed between 272 and 293 positions leading to a highly chicken-specific amplicon of 288 bp. This amplicon was evident in all the 9 breeds/ strains of chicken species studied (Fig. 1). For confirmation, amplified products of Aseel and Kadaknath breeds were custom sequenced and submitted to EMBL database with accession numbers AM933753 and AM933752, respectively. Highly chicken-specific PCR developed for the identification of chicken meat developed in this study exhibited optimum amplification of target DNA at a primer concentration of 20 pico moles and an annealing temperature of 55 °C. Previously, DNA-based identification of chicken species has been accomplished by targeting mitochondrial as well as nuclear sequences such as nuclear actin genes [3, 4, 13], avian CR1 gene [14, 15], satellite DNA [16], and mitochondrial cytochrome-b gene [10, 17–19], 12S rRNA gene [20], and 28S rRNA gene [21]. However, the aminolevulinate synthase gene targeted in this investigation is a new target offering promising advantages of chicken species identification suitable for detection of raw, cooked, and admixed meat samples. Although a cyt-b gene-based multiplex PCR [10] simultaneously identifies 6 animal species including chicken,

Results and discussion DNA isolation Recovery of total DNA from meat samples varied between 84.4 and 188 (mean, ng DNA/mg meat) in raw meat. However, upon thermal processing, recovery of DNA increased at 60, 80, 100, and 121 °C DNA yield of 84.4, 115, 168.6, and 173.8, respectively, was recorded; highest yield of 205 was recorded in micro-oven-processed chicken meat. This rise in DNA recovery could be due to dehydration of meat leading to rise in cell number per unit weight.

Fig. 1 Chicken-specific amplification of 5-aminolevulinate synthase gene from 9 chicken breed/strains. Lane M 100 bp marker, WLH White Leghorn, Col Columbian, Ash Aseel, Kdk Kadaknath, Ast Australorp, RIR Rhode Island Red, Cor Cornish (White), NHS New Hampshire, CB cross-bred (synthetic broiler) and NTC no template control

123

132


advantage of our chicken-specific PCR lies in its sensitivity, that is, 10 pg. Developed PCR delineates a single-step PCR for authentic identification of chicken offering greatest ease and simplicity. Mane and co-workers [22] reported a mitochondrial D-loop-based PCR for identification of chicken meat with detection of less than 1 % in admixed meat and meat products. This assay amplified a chicken-specific amplicon of 442 bp in chicken without cross-amplification in cattle, buffalo, sheep, goat, pig, and duck; however, the PCR assay failed to differentiate chicken meat from turkey and quail meat. PCR developed in the present study was free from such cross-amplifications as proved against 18 animal and 6 avian species including quail and turkey. Hence, targeting of chicken 5-aminolevulinate (ALA) synthase gene in this study was advantageous owing to high specificity of chicken species identification as proved in several breed/strains and without co-(cross)-amplification of signal in other closely related avian species.

Fig. 2 Ascertaining of chicken-specific PCR amplification in adulterated meat mixtures (chicken meat was incorporated at various levels into a base adulteration mixture consisting of meat derived from different animal species). M 100 bp marker, NTC no template control

Sensitivity of PCR

mixtures showed better sensitivity than the previous reports of endpoint PCR [10].

After serially diluting (tenfold) the target DNA (10 ng downward) when PCR was performed, 30 cycles (25 ll reaction volume) PCR detected as low as 10 pg of DNA. This lowest dilution of DNA yielding detectable endpoint PCR signal was taken as limit of detection (LOD) at 20 pico moles of primer concentration and 55 °C of annealing temperature. It is known that endpoint PCR differs from real-time PCR in attaining LOD, the latter has exceptionally high sensitivity to the tune of 0.09 pg [14, 21]. Nevertheless, detection of chicken meat in field conditions with the LOD of 10 pg (as achieved in this study with endpoint PCR) could solve issues related to authentication or forensics of chicken. Apart from spectrophotometric measurement of template DNA [23], sensitivity could also be assessed by incorporating specific meat sample at different proportions into non-target species [24, 25]. The PCR amplification with chicken-specific primers was evaluated in chicken meat experimentally adulterated with other meats. It was found that the PCR was highly sensitive to detect chicken meat adulteration to the extent of 0.1 % (Fig. 2); similar studies detected \1 % meat mixtures [10, 26]; further, in some studies, sensitivity differed in raw (0.01 %) and processed (1 %) meat samples [27]. Further, the primers that amplify short target amplicons yield better sensitivity, and sensitivity progressively declines with increasing amplicon size [14, 23]. Diagnostic chicken-specific amplicon of 288 bp obtained in this study was relatively short, thereby bearing high sensitivity (because smallersized fragments survive better during extreme processing conditions). Also, detection of chicken meat in adulteration

123

Specificity of PCR A total of 18 animal species were checked for crossamplification (Fig. 3), and the primers were found highly specific to the chicken; and no cross-amplification was recorded in any of the species studied. When 6 avian species (turkey, duck, kite, parakeet, guinea fowl, and Japanese quail) were evaluated for amplification, the diagnostic signal was evident only in chicken and not in any of the closely related avian species, making the assay highly specific to chicken (Fig. 4). Even in experimentally

Fig. 3 Chicken-specific amplification of part of 5-aminolevulinate synthase gene from chicken. M 100 bp marker; 1 man, 2 cattle, 3 buffalo, 4 sheep, 5 goat, 6 barking deer, 7 sika deer, 8 chicken, 9 rabbit, 10 horse, 11 pig, 12 leopard, 13 Royal Bengal Tiger, 14 Siberian tiger, 15 dog, 16 goral and NTC no template control


133

processed at high temperatures, that is, 60, 80, 100 °C, autoclaving (cooking under pressure at 121 °C, 15 lbs pressure) for 30 min and micro-oven cooking for 5 min (Fig. 5). Suitability of chicken-specific PCR assay was also evaluated in 20 raw meat samples (intra-laboratory validation study with blind samples). Even up to autoclaving temperature of 121 °C (15 lbs pressure), the PCR amplification was evident. On the other hand, other investigators observed degradation of target DNA leading to loss of PCR signal in samples processed near 121 °C [16, 28]. The Wizard kit used in the present study for DNA isolation was able to extract DNA from thermally exposed meat. Fig. 4 Achieving of chicken-specific PCR amplification using species-specific primers. M 100 bp marker, NTC no template control, Trk turkey, Duk duck, JQI Japanese quail, GFI guinea fowl, Kit kite, Prk parakeet, and Ch chicken

Repeatability The PCR amplification of target DNA with chickenspecific primers was repeated 20 times (n = 20), and consistent results were recorded. Apart from raw meat samples, repeatability was also assessed in cooked, adulterated and meat products and similar findings were established.

Conclusion

Fig. 5 Agarose gel electrophoresis of PCR products from cooked chicken meats. M 100 bp marker, Rw raw meat, Mo micro-oven cooked meat, 60, 80,100 °C chicken thermally cooked at 60, 80, and 100 degree centigrade, At autoclaved (121 °C, 15 lbs, 30 min steam cooking in a pressure cooker), and NTC no template control

adulterated meat (a mixture containing foreign DNA), amplification was evident up to 0.1 % inclusion level (of chicken into foreign meat matrix) and possibility of misidentification of chicken was excluded. Furthermore, diagnostic signal was evident in all the nine breed/strains of chicken (Fig. 1). When the amplicon sequences were analyzed through the ‘‘BlastN’’ of NCBI, they showed highest homology with chicken indicating higher primer specificity to the target sequence. Strength of PCR in detecting chicken in processed meats Successful amplification of target was observed in all the thermally processed chicken meat samples viz. samples

In the present study, authentic identification of chicken species was accomplished through the specific amplification of chicken aminolevulinate synthase gene. The PCR product of 288 bp was highly specific to chicken species, and chances of cross-amplification were excluded through the inclusion of related avian and animal species. Further, the robust PCR technique was established for its application in not only raw but also cooked (up to autoclaving conditions) and adulterated samples. Relative merits of our PCR included: (1) single-step PCR and gel electrophoresis procedure for the conclusive identification of chicken species, (2) simplicity of the operating protocol and result interpretation without requiring any post-PCR analysis, (3) ease of technique application for routine analysis of field samples, (4) any molecular biology laboratory involved in PCR work would undertake sample analysis task, (5) the assay was robust for even cooked (up to autoclaving) and adulterated (up to 0.1 %) meats, and (6) high sensitivity (LOD 10 pg) of the assay qualify the task of routine sample analysis. In the milieu of social, religious, economic, forensic, and public health concerns relating to chicken species identification, the present research would be highly promising in solving issues related to chicken meat authentication. Acknowledgments The authors thank i. the Department of Biotechnology (DBT), Government of India, New Delhi, ii. The Dean, College of Veterinary and Animal Sciences, G.B.P.U.A.T., Pantnagar, and iii. The officials of High Altitude Zoo, Nainital (Uttarakhand, India) for their support rendered in accomplishing this work.

123

134


References 1. Ghovvati S, Nassiri MR, Mirhoseini SZ, Heravi MA, Javadmanesh A (2009) Fraud identification in industrial meat products by multiplex PCR assay. Food Control 20:696–699 2. Ayuso R, Lehrer SB, Tanaka L, Ibanez MD, Pascual C, Burks AW, Sussman GL, Goldberg B, Lopez M, Reese G (1999) Ig-E antibody response to vertebrate meat proteins including tropomyosin. Ann Allergy Asthma Immunol 83:399–405 3. Hopwood AJ, Fairbrother KS, Lockley AK, Bardsley RG (1999) An actin gene-related polymerase chain reaction (PCR) test for identification of chicken in meat mixtures. Meat Sci 53:227–231 4. Lockley AK, Bardsley RG (2002) Intron variability in an actin gene can be used to discriminate between chicken and turkey DNA. Meat Sci 61:163–168 5. Hitchcock CHS, Crimes AA (1985) Methodology for meat species identification: a review. Meat Sci 15:215–224 6. Patterson RLS, Jones SJ (1990) Review of current techniques for the verification of the species of origin of meat. Analyst 115: 501–506 7. Stamoulisa P, Stamatisa C, Sarafidoua T, Mamuris Z (2010) Development and application of molecular markers for poultry meat identification in food chain. Food Control 21:1055–1061 8. Chikuni K, Ozutsumi K, Koishikawa T, Kato S (1990) Species identification of cooked meats by DNA hybridization assay. Meat Sci 27:119–128 9. Janssen FW, Hagele GH, Buntjer JB, Lenstra JA (1998) Species identification in meat using PCR-generated satellite probes. J Indust Microbiol Biotech 21:115–120 10. Matsunaga T, Chikuni K, Tanabe R, Muroya S, Shibata K, Yamada J, Shinmura Y (1999) A quick and simple method for the identification of meat species and meat products by PCR assay. Meat Sci 51:143–148 11. Partis L, Croan D, Guo Z, Clark R, Coldham T, Murby J (2000) Evaluation of a DNA fingerprinting method for determining the species origin of meats. Meat Sci 54:369–376 12. Sambrook J, Russell DW (2001) Molecular cloning: a laboratory manual, 3rd edn. Cold Spring Harbour Laboratory Press, Cold Spring Harbour 13. Bellis C, Ashton KJ, Freney L, Blair B, Griffiths LR (2003) A molecular genetic approach for forensic animal species identification. Forensic Sci Int 134:99–108 14. Walker JA, Hughes DA, Anders BA, Shewale J, Sinha SK, Batzer MA (2003) Quantitative intra-short interspersed element PCR for species-specific DNA identification. Anal Biochem 316:259–269 15. Walker JA, Hughes DA, Anders BA, Shewale J, Sinha SK, Batzer MA (2004) Quantitative PCR for DNA identification based on

123

16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

26.

27.

28.

genome-specific interspersed repetitive elements. Genomics 83:518–527 Buntjer JB, Lamine A, Haagsma N, Lenstra JA (1999) Species identification by oligonucleotide hybridization: the influence of processing of meat products. J Sci Food Agri 79:53–57 Hird H, Goodier R, Hill M (2003) Rapid detection of chicken and turkey in heated meat products using the polymerase chain reaction followed by amplicon visualization with vistra green. Meat Sci 65:1117–1123 Dooley JJ, Paine KE, Garrett SD, Brown HM (2004) Detection of meat species using TaqMan real-time PCR assays. Meat Sci 68:431–438 Tanabe S, Hase M, Yano T, Sato M, Fujimura T, Akiyama H (2007) A real-time quantitative PCR detection method for pork, chicken, beef, mutton and horseflesh in foods. Biosci Biotechnol Biochem 71:3131–3135 Karabasanavar NS, Singh SP, Shebannavar SN, Kotresh AM, Girish PS, Umapathi V (2010) Identification of avian species using polymerase chain reaction and sequence analysis of mitochondrial 12 s rRNA gene. Vet Arhiv 80:653–661 Lo´pez-Andreo M, Lugo L, Garrido-Pertierra A, Prieto MI, Puyet A (2005) Identification and quantitation of species in complex DNA mixtures by real-time polymerase chain reaction. Anal Biochem 339:73–82 Mane BG, Mendiratta SK, Tiwari AK (2009) Polymerase chain reaction assay for identification of chicken in meat and meat products. Food Chem 116:806–810 Mendoza-Romero L, Verkaar EL, Savelkoul PH, Catsburg A, Aarts HJ, Buntjer JB, Lenstra JA (2004) Real-time PCR detection of ruminant DNA. J Food Prot 67:550–554 Krma´ø P, Rene`ova´ E (2001) Identification of bovine-specific DNA in feedstuffs. J Food Prot 64:117–119 Lahiff S, Glennon M, O’Brien L, Lying J, Smith T, Maher M, Shilton N (2001) Species-specific PCR for the identification of ovine, porcine and chicken species in meat and bone meal (MBM). Mol Cell Probe 15:27–35 Meyer R, Hoefelein C, Luethy J, Candrian U (1995) Polymerase chain reaction-restriction fragment length polymorphism analysis: a simple method for species identification in food. J AOAC Int 78:1542–1551 Calvo JH, Osta R, Zaragoza P (2002) Quantitative PCR detection of pork in raw and heated ground beef and pate. J Agric Food Chem 50:5265–5267 Brodmann PD, Moor D (2003) Sensitive and semi-quantitative TaqManTM real-time polymerase chain reaction systems for the detection of beef (Bos illia) and the detection of the family Mammalia in food and feed. Meat Sci 65:599–607