Kelly and Ted were two budding Ph.D. students from different labs in fierce competition and required the use of qPCR (quantitative PCR) to support results generated from microarray and systems biology data. They were both working on two genes implicated in neural regeneration and a compound postulated to induce proliferation of neurons. It was clear from the posters they presented at a recent conference that they were working on the same project and that the associated data would be well received by the scientific and medical community with a publication in a high-profile journal. This also meant that one of the students was going to get scooped, which only served to highlight the need to produce the qPCR data quickly, and unfortunately neither student had experience with this technique.
Upon his return, Ted went to his PI to discuss the situation and as expected it was agreed that Ted would work day and night to produce the key data. Although the lab had access to a common lab with a CFX-384 qPCR machine from Bio-Rad Laboratories, no one had ever performed the experiments. However, since time was of the essence, Ted’s supervisor strongly suggested that he get help from Rubin, a post-doc who had performed recent experiments in the neighboring lab and was therefore considered to be a qPCR guru. Since Rubin had proclaimed on several occasions that qPCR was easy and always worked for him, Ted was anxious to meet with him.
Kelly actually called her PI during the conference to discuss the issue and the best approach to produce the crucial qPCR data quickly. Coincidentally, a friendly lab down the hall was just in the process of upgrading the software of their CFX-96 qPCR machine to version 3.0 and the Bio-Rad Field Application Scientist (FAS) had organized to provide a refresher training the following week. There was also a qPCR guru in the lab, but Kelly was always a little suspicious of the “it always works for me” statements from the post-doc guru and was hoping to get more objective information from the FAS, so preferred to wait for the Bio-Rad training and read up a little on qPCR from the literature in the meantime.
Ted was elated when he sat down with Rubin, who showed him how to search the literature and other sources for primers to his genes using Google and how important it was to use published sequences whenever possible. Although Ted did ask about double checking these sequences to be sure that they annealed to the proper target, Rubin confidently stated that this was a waste of time, that he always had amplification with primers from the literature and not to worry so much. Ted had also heard about reference genes and Rubin explained that these were just housekeeping targets like GAPDH and that he would provide Ted the primers that he had used for GAPDH and 18S RNA. Ted was very excited because Rubin had been so helpful with simple instructions.
Kelly read up on qPCR and came across several articles referring to the MIQE guidelines published by Bustin et al. in 2009.1 She also came across an article published the following year by Taylor et al.,2 which described a practical approach to qPCR and, coincidentally, it was the lead author from this paper who was the Bio-Rad FAS giving the upcoming refresher training on the CFX-96.
The articles described a rigorous stepwise approach to qPCR that involved four key steps: sample isolation, RNA extraction, reverse transcription, and qPCR reaction. Each step required careful procedures to assure that the final data gave interpretable results and correct conclusions. Kelly had also read a couple of papers by Taylor et al. and Lanoix et al. showing how the final data and associated interpretations could dramatically change based on poor sample preparation, primer validation, and reference gene selection. She now had many questions for the Bio-Rad FAS to ensure that the design of her experiments would result in accurate conclusions.
Ted quickly ordered his primers and prepared the RNA samples from the mouse brain tissue, which had been flash frozen in liquid nitrogen. He opted to use the Trizol method and associated buffers supplied by Rubin, who assured Ted that he always extracted good RNA with this method. Although Ted did inquire about testing the RNA to make sure it was good, Rubin told Ted that this was not necessary and to just go ahead and reverse transcribe the samples and run the associated cDNA samples with the two target and the reference gene primers. Anxious to get some data, Ted started the experiments and expected to have results in a few days.
The following week, the Bio-Rad FAS arrived for the CFX-96 training. It was much more than Kelly had expected, with a full day of hands-on, theory, and experimental design information. With several pages of notes taken and questions answered, Kelly was confident with her experimental design plan, which would involve a kit-based methodology to extract the RNA from her flash frozen brain samples. She chose to use the Aurum Fatty Fibrous RNA extraction kit from Bio-Rad, which would not only guarantee a clean prep, but would be free of genomic DNA. The iScript Reverse Transcription kit from Bio-Rad was her choice to convert the RNA to cDNA because it used a blend of Oligo DTs and random primers to provide complete coverage of the transcriptome with a robust and sensitive reverse transcriptase and RNAseH to digest the RNA. Finally, the SsoFast Advanced qPCR supermix from Bio-Rad would be used for the qPCR reaction because of the unique Taq Polymerase enzyme that was highly resistant to potential qPCR inhibitors. The other advantage to using these products was the ongoing support that would be provided by the FAS with whom Kelly could regularly consult during her experiments.
Results had come quickly for Ted, especially with the CFX-384, where he was able to complete his entire qPCR assay on a single plate and avoid any interplate variability. He was now ready to analyze the data, and Rubin was once again very helpful. He gave Ted an EXCEL Spreadsheet that he used to automatically calculate all the normalized expression results simply by importing the raw Cq values. Although Ted had no idea how the calculations were done, Rubin assured him they were correct. Finally, Ted was able to critically assess the data; for the first time in this process, he was concerned with the results.
First, there seemed to be a high level of variability between the samples. Even more disconcerting was that there was an opposite change in expression between the samples depending on which reference target he used for normalization. Ted started to look for help in the literature and came across the same papers that Kelly had researched referring to the MIQE guidelines. He realized quickly that he had overlooked several steps in the design of his experiments, including the testing of his RNA samples for purity and quality, the validation of his primers for annealing temperature and reaction efficiency with standard curves, and the testing of his reference gene targets for stability.
Kelly had just completed her experiments with RNA samples that she had tested for purity (OD 260/280) using a Nanodrop and quality using an Experion System from Bio-Rad. Her primers had been validated for annealing temperature using the thermal gradient option on the CFX-96 qPCR machine followed by standard curves at the optimized annealing temperatures where she obtained reaction efficiencies between 90% and 110% as recommended. She had also carefully chosen 10 potential reference genes from literature references and microarray data for which she ordered the primers, which she then validated and finally tested for stability over her sample sets using qPCR and the software programs GeNorm and NormFinder. She opted to use the three most stable reference genes for normalization, which gave a geNorm M-value below 0.5.
After consulting with the Bio-Rad FAS, she decided to design her plates such that all the samples for a given target were on the same plate, which would eliminate interplate variability. Finally, Kelly used the latest version of CFX Manager software for data analysis, which included the algorithms to perform relative gene expression with normalization to multiple reference genes over multiple plates. The data were publication ready and confirmed the results already collected from microarrays. To Kelly’s surprise, the reviewers of the journal to which the article was submitted were not only supportive of the paper, but also provided very positive feedback for the qPCR data.
Unfortunately, Ted had to restart his experiments and finally produced the same results as Kelly when he followed the MIQE guidelines. He was disappointed to discover that he had been scooped—Kelly’s paper was quickly accepted in the targeted, high-profile journal, but Ted did manage to get his paper into a solid journal after a few revisions. Of course, Ted was thankful that his paper was eventually published, but he was relieved that he had not submitted his paper with the qPCR results he had originally produced. By following the key criteria of the MIQE guidelines as Kelly did, Ted was confident that the results accurately reflected the conclusions drawn, as did the reviewers.
- Bustin, S.A.; Benes, V. et. al.; http://www.rdml.org/clinchem.2008.112797v1.pdf.
- Taylor, S.; Wakem, M. et. al.; http://gene-quantification.eu/Taylor-et-al-The-ongoing-evolution-of-qPCR-Methods-2010.pdf.
The State of RT-Quantitative PCR: Firsthand Observations of Implementation of Minimum Information for the Publication of Quantitative Real-Time PCR Experiments (MIQE): http://bit.ly/JMMB_Taylor (paper) and http://bit.ly/JMMB_video (video).
For more information on MIQE, see http://www.americanlaboratory.com/914-Application-Notes/149612-A-Map-to-More-Reliable-RT-qPCR-Results-Using-MIQE-Guidelines/. For more information on qPCR instruments and reagents, please visit http://www.bio-rad.com/en-us/category/amplification-pcr.
Sean Taylor, Ph.D., is Field Application Specialist, Bio-Rad Laboratories (Canada) Ltd., 1329 Meyerside Dr., Mississauga, Ontario L5T 1C9, Canada; tel.: 905-364-3435; fax: 905-364-3434; e-mail: Sean_Taylor@bio-rad.com . Joey St-Pierre is a lab member, and Cathy Vaillancourt is Associate Professor, Institut National de la Recherche Scientifique Centre INRS-Institut Armand Frappier, Québec, Canada.