Gene expression predicts overall survival in paraffin-embedded tissues of diffuse large B-cell lymphoma treated with R-CHOP
Blood Rimsza et al.
112: 3425
Supplemental materials for: Rimsza et al
Description of methodology (adapted from Roberts et al, Laboratory Investigation, 20071)
Three 5 micron unstained tissue sections were lysed, denatured, and permeabilized by heating in HTG Lysis Buffer. The samples were then frozen and sent to HTG for analysis. At HTG, the 50-mer probes specific for the genes of interest were incubated with the samples, forming specific probe-mRNA duplexes, then unhybridized probes were digested by S1 Nuclease. Next, alkaline hydrolysis destroyed the mRNA in the duplexes, leaving intact probes with stoichiometric concentrations proportional to the amounts of specific mRNA originally present. After neutralization, samples were transferred to ArrayPlates for probe detection. The ArrayPlates contained a universal array of 16 unique, covalently-bound, 25-mer anchor oligonucleotides spotted in a 4 x 4 grid on the bottom of each well. This universal array was modified to bind 50-mer probes for the genes of interest at pre-selected positions by exposing the array to a mixture of 50-mer Programming Linker oligonucleotides that contained a 25-mer sequence to bind one of the probes at one end, and a 25-mer sequence to bind one of the anchor oligonucleotides on the other end. Three different mixtures of Programming Linker oligonucleotides distributed across 3 ArrayPlate wells were required to measure all the genes of interest in our assay.
After hybridization, probes from the sample were bound to array elements by the Programming Linker oligonucleotides. A mixture of Detection Linker oligonucleotides was added. The 50-mer Detection Linkers contained a 25-mer sequence that bound sample probe on the end not bound by the Programming Linker probe on one end, and a common 25-mer sequence to bind a Detection Probe on the other. Detection Probe was added, which bound to all the Detection Linkers. The Detection Probe contained bound horseradish peroxidase. Upon the addition of chemiluminescent peroxidase substrate (Lumigen PS-atto, Lumigen, Inc., Southfield, MI) each array element gave off light proportional to the amount of sample probe bound at that position.
The signals for all 1,536 elements in an ArrayPlate were recorded simultaneously by imaging the plate from the bottom with a CCD-based Omix Imager (HTG). Images were analyzed using Vuescript software (HTG) which calculated average pixel intensity for each element to determine expression levels for each gene. Expression levels were normalized to the housekeeping gene TBP.
As previously, we used the key genes identified as prognostically important in 4 previous papers in DLBCL which accounted for 36 genes of interest.25 Because of the heterogeneity of cellular composition in human tumor samples, we also included probes designed to test the tumor composition for B cells (CD19, CD20), T cells (CD3) and histiocytes (CD68). Two housekeeping genes, TBP and PRKG1, were chosen based previously published work assessing the utility of different endogenously expressed genes as housekeeping genes, which identified these 2 genes as stably expressed at moderate or low levels in different types of lymphomas by qRT-PCR.6 These 2 housekeeping genes were repeated at diagonal corners in each of the 3 wells used to create the assay. An oligo dT probe was added in order to assess the quantity of mRNA in the sample (since an oligo dT probe should detect all mRNA which has a poly-A tail). However, for technical reasons due to the stringency of the assay, this probe was non-functional and not further utilized. A probe for cytochrome oxidase was also initially included because it is coded in mitochondrial DNA, and should be expressed at high levels. This turned out to bind both DNA and RNA, and so gave an extremely bright and generally oversaturated signal and was therefore not further considered, except that it could be used to distinguish whether there was insufficient material for the assay, or whether, if it had disappeared entirely, the sample was too degraded for use.
For each of the 44 genes of interest, four specific probes were designed though not all were synthesized. ArrayBuilder 2.0 software (HTG) was used to design the oligonucleotides required for the assay to measure target transcripts in groups of 16. Briefly, with the user providing the accession numbers for the target genes and assigning their position in the array, the software retrieved each mRNA sequence from GenBank and ranked successive 50-mer stretches of the target gene sequences according to the melting temperature (Tm) of their 5′- and 3′- constituent 25mers, giving preference to those 50-mers for which the Tm of each of the two 25-mer halves was nearest to 68°C. The four highest ranked and non-overlapping 50-mer sequences for each of the 16 target mRNA species were subjected to BLAST to identify homologous sequences. Sequences with homology to other genes were rejected and replaced with the next highest-ranking 50-mer sequence that was in turn submitted to BLAST. Sequences without significant homology were retained. The software then created output files containing the sequences of the four oligonucleotides (Programming Linker, Protection Probe, Detection Linker and Attenuation Fragment) required to measure a given 50-mer target in the assay.
REFERENCES
(1) Roberts RA, Sabalos CM, LeBlanc ML et al. Quantitative nuclease protection assay in paraffin-embedded tissue replicates prognostic microarray gene expression in diffuse large-B-cell lymphoma. Laboratory Investigation. 2007;87:979997.
(2) Lossos IS, Czerwinski DK, Alizadeh AA et al. Prediction of survival in diffuse large-B-cell lymphoma based on the expression of six genes. N Engl J Med. 2004;350:18281837.
(3) Rosenwald A, Wright G, Chan WC et al. The use of molecular profiling to predict survival after chemotherapy for diffuse large-B-cell lymphoma. N Engl J Med. 2002;346:19371947.
(4) Shipp MA, Ross KN, Tamayo P et al. Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning. Nat Med. 2002;8:6874.
(5) Tome ME, Johnson DBF, Rimsza LM et al. A redox signature score identifies diffuse large B-cell lymphorna patients with a poor prognosis. Blood. 2005;106:35943601.
(6) Lossos IS, Czerwinski DK, Wechser MA, Levy R. Optimization of quantitative real-time RT-PCR parameters for the study of lymphoid malignancies. Leukemia. 2003;17:789795.
