Van Eyk, Dunn - Proteomic and Genomic Analysis of Cardiovascular Disease - 2003 (522919), страница 38
Текст из файла (страница 38)
What is important is that at some point, the PCR productsare checked for cross-contamination.How are the cDNA clones curated, maintained, sequenced, and checked?Another problem can arise when the original cDNA clones are misidentifiedor mislabeled at the source. The only way to asses the true quality of a cDNAarray is to have verified the identity of each clone that is spotted on the filter.This needs to be done both with the actual plates used in array printing andon the master set of glycerol stocks. Ultimately, every cDNA used to spot anarray must be sequence verified. As our experience showed, it is not sufficientto show consistent results with a complex probe, because this failed to detectthe arrays being wrong from the beginning.How often are clones cross-checked with the database and to each other?Some percentage of the clones from an array need to be selected randomlyand sequence-verified for fidelity to the original plate, and to the database.
Ideally, this would be a random selection of spots selected from each printed lotof the arrays and should vary from time to time.How does the robot manipulate each array slide, or filter?Is there a micropipette or capillary transfer, how is the tip cleaned betweenspots, how many pins are actually employed for each array grid, how is thepin alignment checked and how often?What are the controls for batches of glass slides, nylon membranes etc?This is perhaps the most critical aspect of array manufacture, and should becompletely understood prior to purchasing a given array.
Are random arrayspulled from each lot for testing or are the first and last arrays from a printingrun always used? How many individual tests does each array undergo? Whatare the exact conditions of quality control testing and do they closely matchyour own hybridization conditions? How much error can be tolerated for anarray to “pass”? When a test array is visualized, what criteria are used to compare one lot to a different lot?What probes are used as the controls for a given lot of array filters?Ideally, a series of arrays would be pulled from each lot and hybridized to aset of complex probes. The critical issue here is what the probes consist of.
Isthe test probe ever varied? How are the test probes themselves generated andtested? If the same set of quality control spots are being checked each time,then by definition the other spots on the array are never checked. Additionally,the arrays tested should be randomly selected from throughout the printingrun and not just always be the first or last (or both) arrays from each lot.What is the mechanism for reporting errors and additions to you?Is there a web site that is updated with information about the arrays? Will youreceive notice of problems or updates via email, fax, phone, sales representative, or will you have to search for updates and corrections to the array data-7.7 Final Thoughtsbases yourself? A responsible manufacturer should notify the user any time aproblem is detected.8. What is the mechanism for reporting errors to the company?Do you have a direct contact other than the sales representative? Is there aspecific person who you can call or send an email to? Ideally there will be acompany contact that is a supervisor who has the complete knowledge of yourarray product and not a group of technical support staff each with varyingknowledge about the arrays.
It would be a good idea to check out the qualityof the tech support by calling the toll free number and asking some of theabove questions prior to purchasing an array.9. What is explicitly stated in the guarantee?Will arrays be replaced and under what conditions, will allowance be made forpersonnel costs and the costs of all associated reagents or just for the arraysthemselves? Who is responsible for making the decision to reimburse you fora defective array and how will a determination be made?7.7Final ThoughtsThe main problem with using arrays to analyze the relative expression level ofthousands of cDNAs is that it is impractical to cross check all of or even most ofthe data.
If fact, one of the main attractions of using microarrays is the ability toexamine expression levels of more cDNAs than possible by any other means. It iscritical that the cDNAs on the array are what they claim to be, and this gets to theheart of quality control issues at the array manufacturer.
Perhaps in no other molecular endeavor does a user have to place so much trust in an outside manufacturer to obtain meaningful results. As we found through our own experiments,this is not a perfect, “off the shelf” technology. There is a tendency to place moreconfidence in arrays than they may merit.Although our experience is specific to the ResGen gf300 rat array marketed byInvitrogen, the issues and problems we encountered are applicable to any cDNAmicroarray technology.
We would like to point out that the underlying technologybehind robotic spotting of cDNA clones to nylon filters and subsequent hybridization methodologies is generally sound. We have only examined a subset of cDNAson the gf300 array. Although statistically unlikely, it is formally possible that fullsequence verification may identify a significantly smaller or larger set of misidentified cDNAs on the gf300 arrays than we predict based on simple extrapolation.Additionally, ResGen produces other arrays representing cDNAs from human,mouse, and yeast and we have no evidence for errors in any of these products.Our experience left us with the difficult position of having detailed data on over5,100 loci but not being able to unequivocally identify any of these clones.
Amongother problems, this made grouping of functional categories of gene products impossible. Our experience with the gf300 rat cDNA Arrays from ResGen has ultimately led to our changing our experimental system from rat to mouse, necessi-1231247 A Cautionary Taletating many months of set up work and the purchase of arrays from a differentsource. Furthermore, personnel costs in excess of $100,000 were incurred, and approximately 18 months of research had to be abandoned.
It is our hope that arrayprinting and misidentification errors will be identified and corrected prior to thewidespread use of the arrays in scientific experimentation, and as the technologyof array manufacturing progresses, methods to validate the results of array experiments will continue to evolve.7.8AcknowledgementsWe would like to thank Isa Werny and Ronald Hanson for their expert advice andtechnical assistance in carrying out these experiments. This work was supportedin part by NIH grants, R01 HL61553, PO1 HL-03174, and NSF grant ERC9529161.Note added in proof: As of October, 2002 ResGen has discontinued sale of thegf300 rate array.7.9References123456Schena, M., Shalon, D., Davis, R.
W.,Brown, P. O., Quantitative monitoring ofgene expression patterns with a complementary DNA microarray. Science 1995,270, 467–470.Paoni, N. F., Lowe, D. G., Expressionprofiling techniques for cardiac molecular phenotyping. Trends Cardiovasc. Med.2001, 11, 218–221.Hwang, J. J., Dzau, V. J., Liew, C. C.,Genomics and the pathophysiology ofheart failure. Curr. Cardiol.
Rep. 2001, 3,198–207.Balasubramaniam, J., Del Bigio, M. R.,Analysis of age-dependant alteration inthe brain gene expression profile following induction of hydrocephalus in rats.Exp. Neurol. 2002, 173, 105–113.Mziaut, H., Korza, G., Elkahloun,A. G., Ozols, J., Induction of stearoylCoA desaturase is associated with highlevel induction of emerin RNA. Biochem.Biophys. Res. Commun. 2001, 282, 910–915.Irwin, L.
N., Gene expression in the hippocampus of behaviorally stimulated rats:7891011analysis by DNA microarray. Brain Res.Mol. Brain Res. 2001, 30, 163–169.Kuo, C. T., Morrisey, E. E., Anandappa,R., Sigrist, K., et al., GATA4 transcription factor is required for ventral morphogenesis and heart tube formation.Genes Dev. 1997, 11, 1048–1060.Lin, Q., Schwarz, J., Bucana, C., Olson, E.
N., Control of mouse cardiacmorphogenesis and myogenesis by transcription factor MEF2C. Science 1997,276, 1404–1407.Lyons, I., Parsons, L. M., Hartley, L.,Li, R., et al., Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box geneNkx2-5. Genes Dev. 1995, 9, 1654–1666.Aplin, A. C., Adams, L., Schwartz,S. M., Murry, C.
E., Defining MolecularPhenotypes of Cardiovascular Cells byMicroarray Analysis. Circulation 2000,102, Suppl., Abs. #139.Flynn, T. G., The elucidation of thestructure of atrial natriuretic factor, anew peptide hormone. Can. J. Physiol.Pharmacol. 1987, 65, 2013–2020.7.9 References1213141516van Groningen, J.
J., Bloemers, H. P.,Swart, G. W., Rat desmin gene structureand expression.Biochim. Biophys. Acta.1994, 1217, 107–109.Adamcova, M., Pelouch, V., Isoforms oftroponin in normal and diseased myocardium. Physiol. Res. 1999, 48, 235–247.Ruppert, C., Kroschewski, R., Bahler,M., Identification, characterization andcloning of myr 1, a mammalian myosinI. J. Cell Biol. 1993, 120, 1393–1403Solway, J., Seltzer, J., Samaha, F. F.,Kim, S., et al., Structure and expressionof a smooth muscle cell-specific gene,SM22 alpha. J. Biol. Chem. 1995, 270,13460–13469.Giachelli, C., Bae, N., Lombardi, D.,Majesky, M., Schwartz, S., Molecularcloning and characterization of 2B7, a ratmRNA which distinguishes smooth muscle cell phenotypes in vitro and is identical to osteopontin (secreted phosphopro-1718192021tein I, 2aR). Biochem.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
