to BioTechniques free email alert service to receive content updates.
Universal reference method for real-time PCR gene expression analysis of preimplantation embryos
Neil Ivan Bower1, Ralf Joachim Moser1, Jonathan Robert Hill2, and Sigrid Arabella Lehnert1
1CSIRO Livestock Industries, Queensland Bioscience Precinct, St. Lucia, QLD
2CSIRO Livestock Industries, Chiswick Laboratories, Armidale, NSW, Australia
BioTechniques, Vol. 42, No. 2, February 2007, pp. 199–206
Full Text (PDF)

Real-time PCR expression profiling in individual preimplantation embryos poses two main challenges. First, the amount of RNA from blastocysts (between 100 and 200 cells) is too small to quantify, and secondly, a reference gene with stable expression across preimplantation embryos produced by different reproductive technologies is required. We have developed a method using RNA and DNA spikes that allows for accurate normalization of gene expression without the use of an internal housekeeping gene in preimplantation blastocysts. Prior to the simultaneous extraction of RNA and DNA, plant-specific RNA and DNA spikes are added to the tissue. After synthesis of cDNA, target gene transcript and the exogenous RNA spike are measured using real-time PCR. To account for differences in the number of cells in each sample, the genomic gene copies of 18S-DNA are measured by quantitative PCR and normalized to the DNA spike. While the DNA spike accounts for extraction efficiency, the 18S genomic target indicates the number of cells prior to extraction. The values obtained from normalizing the target gene to the RNA spike can be adjusted for cell number, allowing the RNA spike to be used as reference gene. This universal reference approach allows the use of an exogenous spike as a pseudo-housekeeping gene for normalization of gene expression data.


Real-time PCR is a powerful tool routinely used in gene expression analysis and quantification (1). To control for experimental and technical variation, appropriate normalization strategies have to be implemented. The most popular quantification strategy among other approaches is typically the normalization of the expression level of a particular target gene to the expression level of an internal housekeeping or reference gene (2,3,4). By normalizing the expression of a target gene to that of a reference gene, differences in the quality and quantity of starting RNA and in efficiencies of the reverse transcription reaction between samples can be accounted for. This allows the direct comparison of normalized expression values between samples.

For each new experimental system (new tissue samples and treatments), the various reference genes available need to be assessed to determine the most stable housekeeping gene (5). However, a stable housekeeper may not be available in certain experimental situations. For instance, the comparison of somatic cell nuclear transfer (SCNT) embryos with those produced by artificial insemination (AI) or in vitro fertilization (IVF) is confounded by a number of issues. SCNT embryos have been shown to have aberrant gene expression when compared with IVF embryos (6,7,8). Genes that have been used as housekeepers [e.g., glyeraldehyde-3-phosphate dehydrogenase (GAPDH), actin β (ACTB)] in the past may therefore be unsuitable as reference genes when comparing nuclear transfer (NT) and IVF embryos. In addition, controversial results on the gene expression stability of the histone H2a (H2a-615) gene and its use as internal reference gene in preim-plantation embryos are present in the literature (9,10).

The study and quantification of gene expression in single preimplantation embryos is further complicated by the small amounts of starting material. The advancement and refinement of the Eberwine technique (11) to amplify RNA in a linear fashion, in combination with real-time PCR, has overcome these limitations but introduces a number of other issues and different sources of variation. The quality (integrity) and quantity (efficiency) of the resulting RNA/cDNA is dependent on many processes such as the extraction and amplification of the RNA, the reverse transcription reaction, and the amplification efficiency of the PCR.

To account for some of the problems listed above, researchers have used exogenously added spike-in RNAs, which can be used as an exogenous reference gene (12,13,14). These spikes are usually added to the samples during cDNA synthesis and therefore rely heavily on the accurate quantification of the RNA samples. As this exogenous spike is added at the end point in the process, it does not account for the integrity of the RNA sample or the reliability of any technical processing that has previously occurred. Addition of an exogenous reference RNA at the beginning of the entire process could account for any of these potential problems. Two reports describe the use of a commercially available messenger RNA (mRNA) as an exogenous reference for target gene normalization that was spiked into cell culture or tissue samples during the extraction process with only approximate or no knowledge of the starting material (15,16). Yet, in order to use a spike-in RNA as a potential pseudo-housekeeping gene in small samples, the amount of starting material relative to the spike needs to be addressed. This is especially crucial if the amount of cells in the starting material is ranging from 100 to 300 cells, which would profoundly affect the spike-to-RNA ratio after extraction. Determining the number of cells in a given sample would allow the target gene to be normalized to the spike and this value subsequently adjusted for cell number. This would enable the spike to be used as an internal reference gene for the normalization of gene expression data. A current method to count cells in embryos involves the staining of the DNA in the nucleus and the assessment of the cell number under the microscope (17,18). However, the time requirement and the fluorescent dye treatment are likely to have an effect on the transcriptome of these samples and are compromising the ultimate goal to measure gene expression in small biological material. Kanno et al. (19) report the use of Molecular Probes’ PicoGreen® dye (Invitrogen Australia, Mount Waverley, VIC, Australia) to determine the number of cells in starting material that can then be used to calculate an appropriate amount of spike input for gene normalization. Using this method, the smallest sample size producing reliable data are approximately 5000 cells.

  1    2    3    4    5