Presented at:
Advanced Microarray Data Analysis, Bethesda, MD, Jan 20, 2003
Use of a low quantity of RNA for Affymetrix GeneChip arrays
Shirley Zhu, Donglei Hu, Lisa Wong, Scott Huntsman, Erin Heinemeyer and Jeffrey Bluestone
University of California, San Francisco, San Francisco, California, 94103.
Current GeneChip technology allows for gene expression profiling of the entire genome using a variety of biological materials. However, current standard protocols require using five micrograms of total RNA as starting material. There are many instances where only a limited amount of RNA can be purified for analysis by GeneChips, including laser capture micro-dissection, sorted cells from tissue or blood, or samples from patients. Sparse starting material remains a major limiting factor in gene expression profiling, placing severe limits on the technology due to tissue availability.
We have established a reproducible procedure that reduces the total RNA required as starting material by 50 percent, compared to the standard Affymetrix protocol, without compromising data quality. Moreover, we have developed amplification methods that allow for even greater reductions in starting material. Both human peripheral blood mononuclear cells and murine T cells sorted from NOD mice were used as starting material. A two run amplification of anti-sense RNA was performed using a dilution series of 500 ng to 1 ng of RNA as starting material. Comparing the GeneChip data generated by the standard protocol to data generated by our small sample protocol, we found that as little as 1 ng of starting total RNA generated enough material to serve as target for hybridization to Affymetrix GeneChips.
The analyses of small sample preparations were compared with the standard protocol. A similar percentage of “present” calls, high R square values and high reproducibility were observed. A majority of “present” genes were observed in both the standard and amplification protocols. Statistical analysis of two groups of GeneChips produced from the same RNA sample, comparing the standard vs. small sample protocols, indicated that significant genes exhibited high degrees of similarity.
In conclusion, we were able to generate gene expression profiles with
only a few hundred cells. Thus, it is now possible to investigate biological
mechanisms with a low quantity of starting material, using Affymetrix
GeneChip arrays.