Clinical Chemistry
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 52: 1161-1167, 2006. First published April 20, 2006; 10.1373/clinchem.2005.062406
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow 062406.Supplemental Data
Right arrow All Versions of this Article:
clinchem.2005.062406v1
52/6/1161    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wilhelm, J.
Right arrow Articles by Fink, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wilhelm, J.
Right arrow Articles by Fink, L.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Automation and Analytical Techniques
(Clinical Chemistry. 2006;52:1161-1167.)
© 2006 American Association for Clinical Chemistry, Inc.

Automation and Analytical Techniques

Systematic Comparison of the T7-IVT and SMART-Based RNA Preamplification Techniques for DNA Microarray Experiments

Jochen Wilhelm1,a, Jai Prakash Muyal1, Johannes Best1, Grazyna Kwapiszewska1, Maria Magdalena Stein1, Werner Seeger2, Rainer Maria Bohle1 and Ludger Fink1

Departments of1 Pathology and 2 Internal Medicine, Justus-Liebig-University Giessen, Giessen, Germany.

aAddress correspondence to this author at: Department of Pathology Justus-Liebig-University Giessen, Langhansstrasse 10, 35392 Giessen, Germany. Fax 49-641-9941164; e-mail jochen.wilhelm{at}patho.med.uni-giessen.de.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: Small biological samples obtained from biopsies or laser microdissection often do not yield sufficient RNA for successful microarray hybridization; therefore, RNA amplification is performed before microarray experiments. We compared 2 commonly used techniques for RNA amplification.

Methods: We compared 2 commercially available methods, Arcturus RiboAmp for in vitro transcription (IVT) and Clontech BD SMARTTM for PCR, to preamplify 50 ng of total RNA isolated from mouse livers and kidneys. Amplification factors of 3 sequences were determined by real-time PCR. Differential expression profiles were compared within and between techniques as well as with unamplified samples with 10K 50mer oligomer-spotted microarrays (MWG Biotech). The microarray results were validated on the transcript and protein levels by comparison with public expression databases.

Results: Amplification factors for specific sequences were lower after 2 rounds of IVT than after 12 cycles of SMART. Furthermore, IVT showed a clear decrease in amplification with increasing distance of the amplified sequences from the polyA tail, indicating generation of smaller products. In the microarray experiments, reproducibility of the duplicates was highest after SMART. In addition, SMART-processed samples showed higher correlation when compared with unamplified samples as well as with expression databases.

Conclusions: Whenever 1 round of T7-IVT does not yield sufficient product for microarray hybridization, which is usually the case when <200 ng of total RNA is used as starting material, we suggest the use of SMART PCR for preamplification.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Microarray expression profiling experiments usually require RNA in larger quantities than are available from small biological samples (1)(2). Consequently, methods have been developed to amplify the RNA before its use in microarray experiments. The most frequently used method (3) introduces a 5'-terminal promoter (complementary) sequence for the T7 RNA polymerase attached to an oligo(dT) primer into the produced cDNA. After second-strand synthesis, the sense cDNA strand is used as a template for in vitro transcription (IVT). 1 The RNA polymerase produces multiple copies of antisense RNA (aRNA) with amplification factors of up to 1000-fold. Because of random priming to synthesize the second strand, the IVT products are not full-length copies, and the 5'-terminal sequences of the transcripts are underrepresented in the aRNA. This procedure works well when the initial amount of total RNA is 1–3 µg and is used extensively in experiments with cDNA and short-oligonucleotide microarrays (4)(5). When less RNA is available, the aRNA yield can be increased by a second round of amplification (6)(7).

PCR-based amplification [switching mechanism at the 5' end of the RNA transcript (SMARTTM)] was originally developed to amplify full-length cDNAs for construction of clone libraries (8)(9), and more recently has been used in microarray experiments (10)(11)(12). cDNA synthesis is performed with an oligo(dT) primer with an attached binding site for the amplification primer. The technique takes advantage of the ability of reverse transcriptase to add stretches of dC to the synthesized cDNA and thus introduce a binding site for the second amplification primer at the 5' end, yielding products that can be amplified by PCR.

Because IVT is part of the standard labeling protocol for the Affymetrix platform, the fidelity and reproducibility of this method have frequently been analyzed with Affymetrix GeneChip assays (13)(14)(15) and self-made cDNA spotted microarrays (6)(7)(16)(17). The performance of SMART has been evaluated with cDNA-spotted (12)(18) and oligonucleotide-spotted microarrays (19). Starting from low nanogram amounts of RNA, we previously used real-time PCR to determine the amplification factors of several target sequences and found a remarkably higher power of amplification for SMART PCR than for IVT (10).

Published direct comparisons of commercial reagent sets for both of the above methods are few. In these studies, the techniques were applied under different conditions, e.g., with different amounts of starting material (16) or with unknown initial amounts of RNA (20). The application of in-house–produced cDNA microarrays makes repetition of the experiment and confirmation of the results impossible. We performed a direct comparison of 2 commercially available preamplification reagent sets based on SMART (BD Clontech) and T7-IVT (Arcturus).


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
For a detailed description of the materials and methods used, see the Data Supplement that accompanies the online version of this article at www.clinchem.org/content/vol52/issue6 (including the supplemental file containing the description of the materials and methods, supplemental Fig. 1, and supplemental Table 1). Briefly, total RNA from livers and kidneys of BALB/c mice was isolated according to the protocol of Chomczynski and Sacchi (21) and reverse-transcribed by use of the BD Atlas SMART Fluorescent Probe Amplification Kit (Clontech Laboratories) with slight modifications. Unamplified samples for microarray hybridizations were labeled directly during reverse transcription. Amplification of 50 ng of total RNA by T7-IVT was performed with the RiboAmp RNA Amplification Kit (Arcturus). The workflow for the analysis of amplification factors is shown in Fig. 1A . PCR-based amplifications were performed with the BD Atlas SMART Fluorescent Probe Amplification Kit (Clontech). To generate Cy-labeled double-stranded DNA (dsDNA), we amplified 50 ng of total RNA for 18 cycles and labeled it with monofunctional reactive Cy dyes (Amersham).


Figure 1
View larger version (32K):
[in this window]
[in a new window]
  		Figure 1. Flow charts for the assays.

(A), analysis of product-specific amplification rates. RT, reverse transcription (BD Clontech protocol). For details, see the text. (B), flow chart for the microarray experiments. Thick arrows indicate the 40 µg of total RNA that was used for labeling. Thin arrows starting from the circles represent the 50 ng of total RNA introduced into the amplification reactions.

Real-time PCR was performed, as described in the online Data Supplement (22), on the ABI7700 Sequence Detection System (Applied Biosystems) with the SYBR® Green I detection format. Relative concentrations of the target sequences before and after preamplification by IVT or SMART (i.e., the factor of preamplification) were expressed as {epsilon}{Delta}Ct, where {Delta}Ct is the difference between the threshold cycle (Ct) values of the samples being compared (Ctamplified – Ctunamplified), and {epsilon} is the determined amplification factor in the real-time PCR reactions.

Two hybridizations on 10K 50mer oligonucleotide-spotted microarrays (MWG Biotech) were performed for unamplified, T7-IVT (2 rounds)–amplified, and SMART (18 cycles)–amplified samples. The flow chart for the experimental setup is given in Fig. 1BUp . The data generated from the microarrays were evaluated by use of R software (http://www.cran.r-project.org/) and the limma package (23) from BioConductor(24) (http://www.bioconductor.org/), as described in the online Data Supplement (25)(26)(27).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
amplification factors
In preliminary experiments, amplification factors were determined for specific target sequences Mus musculus glyceraldehyde-3-phosphate dehydrogenase gene 3' untranslated region (Gapd 3'-UTR), Gapd, and Mus musculus porphobilinogen deaminase (Pbgd). The results shown in Table 4 in the online Data Supplement and in Fig. 2 demonstrate a dependence of the amplification factors on the distance of the target sequence from the polyA end of the transcripts. After 2 rounds of T7-IVT, average {Delta}Ct values between 10 and 4 were observed, which clearly decreased with increasing distance of the preamplified sequence from the 3' end of the transcripts (ANOVA, P <0.001; Spearman r = –0.988, P <0.001). After 12 cycles of SMART amplification, for the Gapd 3'-UTR and Pbgd target sequences, the expected average {Delta}Ct values of 12 were obtained, whereas the {Delta}Ct value for Gapd was ~8. Additional amplification was seen for all targets after 24 SMART cycles: the average {Delta}Ct values reached 17 for Gapd 3'-UTR and Pbgd and 14 for Gapd (Gapd showed consistently lower values than the other 2 sequences).


Figure 2
View larger version (20K):
[in this window]
[in a new window]
  		Figure 2. {Delta}Ct values of amplified vs unamplified samples for Gapd and Pbgd (n = 6).

The inset is a schematic of the location of the target sequences (white, gray, and black boxes) on the transcripts quantified by real-time PCR. Figure 2, Gapd 3'-UTR, Figure 2, Gapd; {blacksquare}, Pbgd. Shown are the mean values and SDs (error bars).

For a more systematic analysis of these effects, we selected 4 additional genes and determined the amplification factors of 3 target sequences, each with different distances to the polyA tail. The targets were classified according to this distance as short- (130–320 bp), medium- (760–940 bp), and long-distance (1500–1700 bp) targets, respectively. The results shown in Table 2 and Fig. 2 of the online Data Supplement confirmed the initial findings. Even after 1 round of T7-IVT, all sequences with a distance >1 kb to the polyA tail (large distance in Table 2 and Fig. 2A of the online Data Supplement) were amplified with a considerably lower efficiency than the other sequences (paired t-test on average Ct values: short vs long, P <0.005; medium vs long, P = 0.011; short vs medium, P = 0.36). This effect was even more pronounced after the 2nd round of T7-IVT (P <0.005, P <0.01, and P = 0.25, respectively). In contrast, this effect was very small for the SMART-amplified samples (see Table 2 and Fig. 2A in the online Data Supplement; ANOVA, P >0.5; paired t-test, all P values >0.2).

With respect to the distances of the target sequences to the polyA tail, after 1 round of T7-IVT, we obtained an average 200-fold amplification for sequences with a distance <1 kb (short and medium). In contrast, sequences with a distance >1 kb (large) were amplified only 15-fold (Fig. 3A ). After the 2nd round, the amplification factor for short-distant sequences increased to 22 000, and that for medium-distant sequences (760–940 bp) increased to 1500. Amplification of the large-distance sequences (1400–1700 bp) was not significantly greater after the 2nd round of T7-IVT (55-fold). Independent of the distances, 12 and 24 cycles of SMART gave average amplification factors of 1500 and 166 000, respectively (Fig. 3B ).


Figure 3
View larger version (16K):
[in this window]
[in a new window]
  		Figure 3. Average {Delta}Ct values of amplified vs unamplified samples.

Results (shown in Fig. 2Up in the online Data Supplement) were averaged over the distance of the target sequence to the polyA tail. Mean values and SDs (error bars) are shown. Short, medium, and long indicate the distance of the target sequence from the polyA tail: 130–320, 760–940, and 1500–1700 bp, respectively.

It is noteworthy that 2 rounds of T7-IVT, including the incorporation of Cy-UTP in the 2nd round, yielded ~50 µg of aRNA with ~2000 pmol of dyes (40 pmol dye/µg of aRNA), whereas 18 cycles of SMART gave 1.5–2.0 µg of dsDNA with ~40 pmol of dyes (20 pmol dye/µg of dsDNA). With samples taken after 12, 15, 18, 21, and 24 SMART cycles, we found a constant efficiency up to cycle 18. The efficiency between cycles 18 and 21 was decreased and was minimal between cycles 24 and 21 (data not shown). This indicates that the SMART reaction had not reached the plateau phase after 18 cycles.

within- and between-assay comparisons
We assessed the technical reproducibility by evaluating the correlation between 2 duplicate microarray experiments performed independently (see Fig. 3 of the online Data Supplement). MA-Plots of averaged M values are shown in Fig. 4 of the online Data Supplement. The correlation of the M values between the different assays is illustrated in Fig. 4 .


Figure 4
View larger version (19K):
[in this window]
[in a new window]
  		Figure 4. Correlation between preamplified and unamplified samples.

Fitted M values from duplicate microarray experiments were compared. (A), T7-IVT-amplified vs unamplified; (B), SMART-amplified vs unamplified. The gray line indicates the shape of a perfect correlation. Each square represents one spot of the microarray, the combined weight factors (W; product of the 2 weight factors) are shaded from light gray (W = 0.1) to black (W = 1.0). The corresponding correlation coefficient (Pearson r of the weighted and scale normalized data) is given in each panel.

Among the top 200 genes considered as differentially expressed, 32 were common between the assays using unamplified and T7-IVT–amplified RNA, whereas 66 were common between the assays using unamplified and SMART-amplified RNA, and 17 (see Table 5 in the online Data Supplement) were common in all 3 assays. Among the top 100 genes, 20 genes were common between the using unamplified and T7-IVT–amplified RNA and 21 between the assays using unamplified and SMART-amplified RNA. The complete findings for all possible intersections are shown in Fig. 5 .


Figure 5
View larger version (14K):
[in this window]
[in a new window]
  		Figure 5. Venn diagrams for the 200 (A) and 100 (B) genes, respectively, with the highest odds ratios for differential expression.

validity of the microarray results
We compared the 200 (100) genes with the highest odds ratios for differential expression as found by the microarray experiments with the status of differential expression of the genes from public databases. Not all genes were found in the databases because they are either not recorded or not assigned as differentially expressed between the liver and kidney. The comparisons of the results for the genes found are given in Fig. 6 .


Figure 6
View larger version (19K):
[in this window]
[in a new window]
  		Figure 6. Databank queries.

Numbers of correctly and incorrectly assigned differentially expressed genes among the top 200 (filled columns) and top 100 (gray hatched columns) according to the test statistic.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Complex preamplification reactions may alter the relative abundance of specific transcripts and thus yield erroneous results in the subsequent microarray analyses. The IVT reaction may introduce differences caused by incomplete transcription attributable to secondary structures of the mRNA and the possible susceptibility of the RNA polymerase to terminate transcription at specific sequences. IVT has been shown to yield shortened products (13)(28), which is particularly problematic when the probe sequences are not designed to hybridize to the 3' proximal part of the transcript. On the other hand, PCR-based techniques such as SMART may be sensitive to changes in reaction conditions. Any small differences in amplification efficiencies may possibly lead to considerable alterations of the transcript profile. dsDNA obtained by the SMART protocol better represents full-length copies of the original transcripts than aRNA obtained from IVT (19) and is more stable than aRNA, consisting of both sense and antisense strands, which allows hybridization on microarrays spotted with antisense and/or sense oligonucleotides. SMART is considerable less labor- and time-intensive than T7-IVT.

One round of IVT has been reported to yield up to ~2 µg of aRNA from 100 ng of total RNA (6), a result that matched well with our results (data not shown), assuming that ~1%–5% of total RNA is polyA RNA, which corresponds to a 400- to 1000-fold amplification. Starting from as few as 50 ng of total RNA, 1 round of T7-IVT does not yield sufficient labeled material for microarray experiments. Furthermore, after 1 round of IVT, preferential amplification of sequences closer to the polyA tail was already recognizable. Sequences more than 1 kb away from the polyA tail were particularly susceptible to low amplification (see Fig. 2Up and Table 2 in the online Data Supplement).

Two rounds of IVT yielded ~50 µg of aRNA starting from 50 ng of total RNA. Corresponding to 2 x 104- to 105-fold amplification of the polyA RNAs, this amount is less than theoretically expected (~106-fold amplification). This may be explained by the fact that Cy-labeled ribonucleotides are incorporated during the 2nd round of IVT, reducing the efficiency of the RNA polymerase. Additionally, a loss-prone purification step had to be performed. Interestingly, amplification factors for specific target sequences as determined by real-time PCR were considerably lower. This may be attributable to additional amplification of RNA species other than mRNA.

IVT amplification factors for different sequences verified by real-time PCR were clearly dependent on the distance of the detected target sequence from the polyA tail of the transcript (see Fig. 3AUp ; also see Table 2 in the online Data Supplement). The observed decreases in the amplification factors corresponded well to the ratios of 3' to 5' signals reported for Affymetrix arrays (13). In contrast, amplification factors obtained by SMART show no correlation to the distance of the amplified sequence from the polyA tail (Fig. 3Up B). These results demonstrate the ability of SMART to amplify longer products than T7-IVT; these longer products are more suited for investigation of splice variants because differentially spliced exons may be located within the whole transcript sequence and are not restricted to the 3' end.

The high intraassay correlation of the differential expression profiles generated from the unamplified material (r = 0.86; see Fig. 3A in the online Data Supplement) is in line with the results obtained by Klur et al. (29), who reported that a random primed PCR amplification technique was more reproducible than 2 rounds of IVT. Other authors have demonstrated higher reproducibility with 2 rounds of IVT (7)(15)(17)(20) on cDNA arrays. In agreement with several published reports (6)(16)(17)(30), we found that 1 round of IVT starting with >200 ng of total RNA gave expression profiles that correlated very well with those of unamplified samples (data not shown). However, as already expected from the minor correlation of technical repeats, the averaged profiles of the IVT samples amplified for 2 rounds correlated only weakly with those of the unamplified samples (r = 0.65; Fig. 4Up A). In contrast, the SMART-amplified samples correlated noticeably better with the unamplified samples (r = 0.85; Fig. 4Up B). It is noteworthy that both comparisons show that the M values of the preamplified samples are compressed relative to the unamplified samples. This phenomenon was also reported by others (6)(13). Such compression has an impact on the list of candidate genes when these are selected by the M values. Relying only on the average M values would lead to fewer candidate genes for preamplified samples than for unamplified samples. However, when appropriate test statistics are used to select differentially expressed genes, the effect of the compression is largely compensated by lower standard errors.

Many different procedures exist for the selection of genes considered to be differentially expressed, such as controlling the individual or family-wise error rates and/or applying some threshold criteria after explorative data analysis (31). Obviously, different selection criteria will produce different numbers of overlapping candidate genes. We therefore analyzed the overlaps for different selection criteria and found similar patterns of overlap (data not shown). Selection of the 200 genes with the highest test statistics (Bayes estimate for M != 0 expressed as odds ratio) (23) produced a typical pattern (Fig. 5AUp ). Only a few genes (17; 8.5%) were selected in all 3 groups (unamplified, IVT-amplified, and SMART-amplified), and more selected genes were common between the unamplified and SMART-amplified assays (66; 33%) than between the unamplified and IVT-amplified assays (32; 16%). These findings indicate that the SMART-amplified samples better reflect the status of the unamplified samples than do the IVT-amplified samples. However, when we focused on the genes with increased evidence for being differentially expressed (top 100, Fig. 5BUp ), the differences among the amplification methods vanished, which indicates that the data from the IVT-amplified samples still provide some correct results for the genes with high evidence for differential expression.

The better performance of SMART was also supported by the results of our databank queries to validate the selected genes (Fig. 6Up ). The SMART-amplified samples revealed that most genes were correctly assigned as differentially expressed, whereas we found only a relatively few genes selected from the IVT-amplified samples in the databases. The comparison solely with the unamplified sample (Figs. 4Up and 5Up ) shows how many genes are identified in both techniques. However, it does not show whether those genes identified exclusively by either technique are false positives, whereas the comparison with the database results (Fig. 6Up ) showed that a set of genes that were found to be differentially expressed after preamplification and not detected after analysis of the reference data are obviously not false positives.

In this study, oligomer-spotted microarrays were used for expression profiling. The results and conclusions obtained basically refer to this type of array, which is widely available in commercial forms. Nevertheless, the preferential amplification of sequences close to the polyA tail may influence the results obtained with other microarray formats as well. These potential influences remain to be investigated for other microarray formats.

In conclusion, we showed that starting with as little as 50 ng of RNA, SMART was superior to 2 rounds of T7-IVT with respect to the power of amplification, length of the synthesized products, reproducibility, and validity of the microarray results. Moreover, SMART is less laborious and less time-consuming, and the products are more stable. Whenever 1 round of T7-IVT is not sufficient for microarray experiments, which is usually the case when <200 ng of total RNA is available as starting material, we suggest the use of SMART PCR for preamplification.


   Acknowledgments
 
We thank Norbert Weissmann, Ralph Schermuly, and Karin Quanz for kindly providing the tissue specimens, and the Willy-Robert Pitzer Foundation for granting the Agilent Bioanalyzer and Axon Scanner. The study was supported by the Deutsche Forschungsgemeinschaft (DFG), SFB 547, Project Z1.


   Footnotes
 
1 Nonstandard abbreviations: IVT, in vitro transcription; aRNA, antisense RNA; SMART, switching mechanism at the 5' end of the RNA transcript; dsDNA, double-stranded DNA; Ct, threshold cycle; Gapd, Mus musculus glyceraldehyde dehydrogenase gene; UTR, untranslated region; and Pbgd, Mus musculus porphobilinogen deaminase gene.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Jenson SD, Robetorye RS, Bohling SD, Schumacher JA, Morgan JW, Lim MS, et al. Validation of cDNA microarray gene expression data obtained from linearly amplified RNA. Mol Pathol 2003;56:307-312.[Abstract/Free Full Text]
  2. Fink L, Bohle RM. Laser microdissection and RNA analysis. Methods Mol Biol 2005;293:167-185.[Medline] [Order article via Infotrieve]
  3. Van Gelder RN, von Zastrow ME, Yool A, Dement WC, Barchas JD, Eberwine JH. Amplified RNA synthesized from limited quantities of heterogeneous cDNA. Proc Natl Acad Sci U S A 1990;87:1663-1667.[Abstract/Free Full Text]
  4. ’t Hoen PA, de Kort F, van Ommen GJ, den Dunnen JT. Fluorescent labelling of cRNA for microarray applications. Nucleic Acids Res 2003;31:e20.[Abstract/Free Full Text]
  5. Luzzi V, Holtschlag V, Watson MA. Expression profiling of ductal carcinoma in situ by laser capture microdissection and high-density oligonucleotide arrays. Am J Pathol 2001;158:2005-2010.[Abstract/Free Full Text]
  6. Schneider J, Buness A, Huber W, Volz J, Kioschis P, Hafner M, et al. Systematic analysis of T7 RNA polymerase based in vitro linear RNA amplification for use in microarray experiments. BMC Genomics 2004;5:29.[CrossRef][Medline] [Order article via Infotrieve]
  7. Feldman AL, Costouros NG, Wang E, Qian M, Marincola FM, Alexander HR, et al. Advantages of mRNA amplification for microarray analysis. Biotechniques 2002;33:906-1214.[Web of Science][Medline] [Order article via Infotrieve]
  8. Apte AN, Siebert PD. Anchor-ligated cDNA libraries: a technique for generating a cDNA library for the immediate cloning of the 5' ends of mRNAs. Biotechniques 1993;15:890-893.[Web of Science][Medline] [Order article via Infotrieve]
  9. Wellenreuther R, Schupp I, Poustka A, Wiemann S. SMART amplification combined with cDNA size fractionation in order to obtain large full-length clones. BMC Genomics 2004;5:36.[CrossRef][Medline] [Order article via Infotrieve]
  10. Fink L, Kohlhoff S, Stein MM, Hanze J, Weissmann N, Rose F, et al. cDNA array hybridization after laser-assisted microdissection from nonneoplastic tissue. Am J Pathol 2002;160:81-90.[Abstract/Free Full Text]
  11. Seth D, Gorrell MD, McGuinness PH, Leo MA, Lieber CS, McCaughan GW, et al. SMART amplification maintains representation of relative gene expression: quantitative validation by real time PCR and application to studies of alcoholic liver disease in primates. J Biochem Biophys Methods 2003;55:53-66.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  12. Rox JM, Bugert P, Muller J, Schorr A, Hanfland P, Madlener K, et al. Gene expression analysis in platelets from a single donor: evaluation of a PCR-based amplification technique. Clin Chem 2004;50:2271-2278.[Abstract/Free Full Text]
  13. Wilson CL, Pepper SD, Hey Y, Miller CJ. Amplification protocols introduce systematic but reproducible errors into gene expression studies. Biotechniques 2004;36:498-506.[Web of Science][Medline] [Order article via Infotrieve]
  14. King C, Guo N, Frampton GM, Gerry NP, Lenburg ME, Rosenberg CL. Reliability and reproducibility of gene expression measurements using amplified RNA from laser-microdissected primary breast tissue with oligonucleotide arrays. J Mol Diagn 2005;7:57-64.[Abstract/Free Full Text]
  15. Luzzi V, Mahadevappa M, Raja R, Warrington JA, Watson MA. Accurate and reproducible gene expression profiles from laser capture microdissection, transcript amplification, and high density oligonucleotide microarray analysis. J Mol Diagn 2003;5:9-14.[Abstract/Free Full Text]
  16. Puskas LG, Zvara A, Hackler L, Jr, Van Hummelen P. RNA amplification results in reproducible microarray data with slight ratio bias. Biotechniques 2002;32:1330-13346, 8, 40.[Web of Science][Medline] [Order article via Infotrieve]
  17. Wang J, Hu L, Hamilton SR, Coombes KR, Zhang W. RNA amplification strategies for cDNA microarray experiments. Biotechniques 2003;34:394-400.[Medline] [Order article via Infotrieve]
  18. Rihl M, Baeten D, Seta N, Gu J, De Keyser F, Veys EM, et al. Technical validation of cDNA based microarray as screening technique to identify candidate genes in synovial tissue biopsy specimens from patients with spondyloarthropathy. Ann Rheum Dis 2004;63:498-507.[Abstract/Free Full Text]
  19. Petalidis L, Bhattacharyya S, Morris GA, Collins VP, Freeman TC, Lyons PA. Global amplification of mRNA by template-switching PCR: linearity and application to microarray analysis. Nucleic Acids Res 2003;31:e142.[Abstract/Free Full Text]
  20. Wadenback J, Clapham DH, Craig D, Sederoff R, Peter GF, von Arnold S, et al. Comparison of standard exponential and linear techniques to amplify small cDNA samples for microarrays. BMC Genomics 2005;6:61.[Medline] [Order article via Infotrieve]
  21. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156-159.[Web of Science][Medline] [Order article via Infotrieve]
  22. Wilhelm J, Pingoud A. Real-time polymerase chain reaction. [Review]Chembiochem 2003;4:1120-1128.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  23. Smyth GK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 2004;3Article 3. http://www.bepress.com/sagmb/vol3/iss1/art3/..
  24. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, et al. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 2004;5:R80.[CrossRef][Medline] [Order article via Infotrieve]
  25. Carter DA. Selecting candidate genes from DNA array screens: application to neuroscience. Methods 2003;31:263-264.[Medline] [Order article via Infotrieve]
  26. Smyth GK, Speed T. Normalization of cDNA microarray data. Methods 2003;31:265-273.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  27. Wang X, Ghosh S, Guo SW. Quantitative quality control in microarray image processing and data acquisition. Nucleic Acids Res 2001;29:E75-5.
  28. Dumur CI, Garrett CT, Archer KJ, Nasim S, Wilkinson DS, Ferreira-Gonzalez A. Evaluation of a linear amplification method for small samples used on high-density oligonucleotide microarray analysis. Anal Biochem 2004;331:314-321.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  29. Klur S, Toy K, Williams MP, Certa U. Evaluation of procedures for amplification of small-size samples for hybridization on microarrays. Genomics 2004;83:508-517.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  30. Kenzelmann M, Klaren R, Hergenhahn M, Bonrouhi M, Grone HJ, Schmid W, et al. High-accuracy amplification of nanogram total RNA amounts for gene profiling. Genomics 2004;83:550-558.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  31. Kooperberg C, Aragaki A, Strand AD, Olson JM. Significance testing for small microarray experiments. Stat Med 2005;24:2281-2298.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]



The following articles in journals at HighWire Press have cited this article:


Home page
 Am. J. Physiol. Lung Cell. Mol. Physiol.Home page
M. Cabanski, J. Wilhelm, Z. Zaslona, M. Steinmuller, L. Fink, W. Seeger, and J. Lohmeyer
Genome-wide transcriptional profiling of mononuclear phagocytes recruited to mouse lungs in response to alveolar challenge with the TLR2 agonist Pam3CSK4
Am J Physiol Lung Cell Mol Physiol, October 1, 2009; 297(4): L608 - L618.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow 062406.Supplemental Data
Right arrow All Versions of this Article:
clinchem.2005.062406v1
52/6/1161    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wilhelm, J.
Right arrow Articles by Fink, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wilhelm, J.
Right arrow Articles by Fink, L.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Automation and Analytical Techniques

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS