|
|
|||||||||||||
|
|||||||||||||
|
|||||||||||||
|
Experimental data1. Optimization of the amount of amino allyl-dUTP in cDNA synthesisThe ratio of amino allyl-dUTP (aa-dUTP) to dTTP in the synthesis of amino-modified cDNA for post-labeling purposes, is important for two main reasons. First, it determines the labeling density at which CyDye fluors are attached to cDNA. It is important to achieve a high labeling density to ensure bright signal, but at the same time the distance between each CyDye molecule must be kept such that fluorescent quenching does not take place. The second effect the aa-dUTP:dTTP ratio influences is the yield of cDNA. As shown in Figure 1, different ratios of aa-dUTP to dTTP were tested in cDNA synthesis reactions using 1 g of mRNA. Higher concentrations of aa-dUTP resulted in a decrease in cDNA yield. When cDNAs prepared with different concentrations of aa-dUTP were labeled with Cy3-NHS ester and used in a microarray hybridization, better signal-to-noise ratio was obtained from cDNAs containing medium to low amounts of aa-dUTP. Taken together, this data was used to select the ratio of aa-dUTP to dTTP included in the kit, so that both the yield of cDNA and signal-to-noise from microarrays are maximized. The results were also verified using Cy5-NHS ester labeling (data not shown).
Fig 1. Optimization of aa-dUTP:dTTP ratio in first-strand cDNA synthesis reactions. The dark blue column shows the ratio chosen for inclusion in the CyScribe Post-Labeling Kit.
Fig 2. Hybridizations with Cy3-labeled probes prepared using cDNA synthesized with varying aa-dUTP:dTTP ratios. cDNAs prepared with different aa-dUTP:dTTP ratio were labeled with an excess of Cy3-NHS ester and then hybridized onto a microarray slide. 50 ng of cDNA probe was used on each slide. 2. Purification of amino-labeled cDNARemoval of free amino groups from the amino-modified cDNA is essential for the success of labeling with reactive CyDye. Both unincorporated aa-dUTP and the reaction buffer contain free amino groups that can reduce the labeling efficiency of cDNA. We found that the removal of these reagents is best achieved with GFX purification columns. Ethanol precipitation can be used as an alternative. Table 2 summarises purification results obtained with these methods. Both methods yield cDNA that is suitable for post-labeling with reactive CyDye.
Table 2. Summary of results for removal of free amine groups from amino-modified cDNA using GFX PCR DNA and Gel Band Purification Kit and Ethanol precipitation. 3. The amount of mRNA starting materialThe yield of amino-modified cDNA with CyScribe Post-Labeling Kit was evaluated by using varying amounts of mRNA in first-strand cDNA synthesis reactions. The results for using 50–1000 ng of mRNA to synthesize aa-dUTP containing cDNA are shown in Figure 3. cDNA synthesis was primed with 1 ml of anchored oligo(dT) and 1 ml of random nonamers. This combination results in the highest yield of cDNA. The standard protocol for this kit recommends the use of 500 ng of mRNA in each labeling reaction.
Fig 3. The yield of cDNA from varying amounts of mRNA 4. The amount of total RNA as starting materialThe yield of amino-modified cDNA with CyScribe Post-Labeling Kit was evaluated by using varying amounts of total RNA in first-strand cDNA synthesis reactions. The results for using 0.5–25 g of total RNA to synthesize aa-dUTP containing cDNA are shown in Figure 4. cDNA synthesis was primed with anchored oligo(dT) alone using 3 ml of anchored oligo(dT) in each reaction because this amount was found to increase the yield of cDNA compared to 1 ml of the primer. It should be noted that because it is possible to use only the anchored oligo(dT) primer with total RNA, to avoid copying all ribosomal RNA, the yield of cDNA is much less than achieved with the dual priming method. The data in Figure 4 shows that from 1 to 25 mg of total RNA can be used as template to synthesize aa-dUTP modified cDNA.
Fig 4. The yield of cDNA from varying amounts of total RNA. 5. Optimization of the amount of reactive CyDye in the labeling reactionIn each labeling reaction, it is necessary to have a sufficient excess of reactive CyDye to ensure quantitative labeling of most aa-dUTP nucleotides incorporated into cDNA. This is necessary to ensure good incorporation of CyDye into probe but excessive amounts of CyDye can cause background problems if not removed efficiently. Varying amounts of CyDye were tested in initial experiments. The incorporation of CyDye per mg of cDNA was measured with spectrophotometry and the fluorescent signal from a constant amount of cDNA was tested in a microarray application. Figure 5 shows a summary of these experiments for Cy3. The amount chosen for inclusion in the kit is that used in the middle panel. This is because it consistently resulted in bright signal and did not show signs of quenching. The range of CyDye amounts shown here were relatively similar to the amounts that were selected for final evaluation of the dye amount, and hence do not show drastically different properties. As an equal amount of cDNA was used on each slide, the data demonstrates that near equal labeling intensity is achieved with most of these amounts of reactive Cy3. These results also demonstrate that in the chosen range, moderate variations in the amount of CyDye can be tolerated. This is important as it means that small amount of amino groups can be tolerated as impurities in the labeling reaction.
Fig. 5. Titration of the amount of reactive CyDye in the labeling reaction. The data shown is raw slide images set at equal contrast and not normalized data. 50 ng of cDNA was used on each slide. Figure 6 shows the final results for both Cy3 and Cy5 of this optimization process.
Fig 6. Signal from equal amounts of probe (15 pmol CyDye) labeled with Cy3 and Cy5 using the CyScribe Post-Labeling Kit. Raw slide images without normalization of data are shown. 6. Incorporation of CyDye into cDNADifferent batches of reactive CyDye were tested with the CyScribe Post-Labeling Kit. Results from this experiment are shown in Figure 7. No statistically significant difference was seen between the performance of any of the batches. Moreover, the incorporation of both Cy3 and Cy5 into cDNA was equal. The specification of the kit has been set so that the Cy3 and Cy5 reactive dyes included in the kit are matched for their chemical properties.
Fig. 7. Incorporation of CyDye into cDNA. Each bar corresponds to a different manufacturing batch of reactive CyDye that had differed in their amount of reactive groups. Darker colour indicates higher NHS ester content. 8. Comparison of the performance of CyScribe First-Strand Labeling and CyScribe Post-Labeling KitsThe performance of standard labeling reactions of each of the CyScribe kits is detailed in Table 3. The recommended amount of mRNA to be labeled in a standard labeling reaction with each of the two kits differs; whereas 1 ug of mRNA is recommended to be used with the CyScribe First-Strand cDNA Labeling Kit, only 0.5 ug of mRNA per reaction is recommended for use with the CyScribe Post-Labeling Kit. This is because the CyScribe Post-Labeling Kit is significantly more efficient in synthesizing cDNA (see 11.1 for details).
Table 3. Comparison of the performance of CyScribe First-Strand Labeling Kit and CyScribe Post-Labeling Kit. Typically, CyScribe First-Strand cDNA Labeling Kit synthesizes 450–650 ng of cDNA from 1 ug of mRNA. The CyScribe Post-Labeling Kit can yield 1.3–1.5 ug of cDNA from 0.5 ug of mRNA per reaction. Since the cDNA needs to be purified before labeling takes place, some of this cDNA will be lost. Typical recovery from the GFX PCR column purification step is about 60%; however, shorter DNA strands are lost more efficiently, resulting in enrichment for longer cDNAs. From two standard labeling reactions, 1.4–1.8 ug of cDNA is typically recovered for the labeling step. The labeling efficiency achieved with both kits is of similar magnitude. On average every 12–20 nucleotides contains a CyDye fluor in a cDNA prepared with CyScribe First-Strand cDNA Labeling Kit, whereas the corresponding value for the CyScribe Post-Labeling Kit is every 9–30 nucleotides. CyScribe First-Strand cDNA Labeling Kit typically incorporates 90–150 pmol of both Cy3 and Cy5 fluors into first-strand cDNA per reaction. The incorporation tends to be slightly higher when CyDye-labeled dUTP nucleotides are used. The CyScribe Post-Labeling Kit incorporates 120–160 pmol of Cy3 and Cy5 into cDNA in one labeling reaction. Hence, from 1 ug of mRNA starting material, 240–160 pmol of CyDye is incorporated. Free CyDye must be removed before the fluorescently labeled cDNA can be used in a microarray hybridization,. If CyDye-labeled nucleotides are present in hybridizations they will give a strongly fluorescent speckled background. In our experience, the recovery of fluorescently labeled cDNA can be variable. Typical recoveries with AutoSeq™. G-50 columns range from 30 to 65% of labeled cDNA retained. The performance of the Qiagen PCR purification columns is similar. Thus the probe yield can vary from 30 to 100 pmol per 1 ug of mRNA with the CyScribe First-Strand cDNA Labeling Kit and from 80 to 200 pmol per 1 ug of mRNA with the CyScribe Post-Labeling Kit. Therefore, the number of slides that can be hybridized starting with 1 ug of mRNA is twice as much with the CyScribe Post-Labeling Kit.
Fig 8. Dual-color hybridization performed with 15 pmol of Cy3 and Cy5 probes, prepared from human skeletal muscle mRNA with CyScribe Post-Labeling Kit. Related productsMicroarray Labeling KitsCombination kitsCyDye nucleotidesValue packsPurification productsProbe quantificationIncludes built-in software application module with CyDye probe-specific algorithm |